1655_02.pdf

SECTION 2

Cleaning Systems

© 2001 by CRC Press LLC

CHAPTER 2.1

Cleaning Equipment Overview

Barbara Kanegsberg

CONTENTS

Why Do We Need Cleaning Equipment?Performance

The Chicken or the Egg?Cleaning ActionDrying ActionMaterials CompatibilityFixturing, Parts Handling, Automation

Process Efficiency, Process Costs, Environmental and Safety ConcernsPlant Facilities, Floor Space, GrowthEmployee Involvement, Employee EducationIn-House Equipment Design

Ultrasonic CleaningOther Cleaning Systems

Spray Cleaning SystemsReel-to-Reel or Continuous Web CleaningCentrifugal CleanerSpinnersMicroclustersIndustrial Cleaners/Cabinet Washers, Dishwashers, Spray CabinetsSemiaqueous and Cosolvent SystemsWet BenchesImpingement Cleaning

ConclusionReferences

WHY DO WE NEED CLEANING EQUIPMENT?

Anyone tasked with choosing new cleaning equipment asks this question, at least face-tiously. Cleaning equipment often requires a substantial capital investment, and processchange may be accompanied by comments and input from company management, the


insurance company, the fire department, governmental regulatory agencies, the companyfacilities/maintenance group, the in-house environmental health and safety department,and last but not least from the technicians and assemblers who have to watch over theprocess.

So why do we need cleaning equipment? There are a number of reasons, the mostimportant of which should be maximizing cleaning performance. Other reasons for choos-ing a particular type of equipment include decreasing process time, protecting the envi-ronment, and protecting the individual worker.

In any cleaning system, it is important to consider both the cleaning agent and thecleaning action. A cleaning agent may have very high solvency and may be effective in dis-solving the soil. However, it is important to have appropriate cleaning action to assure thatthe cleaning and rinsing agents reach all surfaces and to assure that the soil is carried awayfrom the surface. As a chemist, this author’s first thought in terms of cleaning action isplopping a magnetic stirrer into a beaker of cleaning agent, and cranking up the rheostat.This is not practical for large-scale processes. Similarly, pilot or specialized operations mayinvolve hand-wiping parts with a soft cloth, hand-dipping each part, or scrubbing indi-vidual portions with a cotton swab. For more systematic cleaning, ultrasonics, megasonics,spray in air, spray under immersion, and turbulation all add effectiveness to the process.In addition, weirs and filters1 prolong the life of the cleaning agent in general industrialprocesses. In high-precision applications, filtration with in-line particle monitoring may berequired for adequate contamination control.

PERFORMANCE

Setting aside for a moment worker safety, environmental regulatory issues, and mini-mization of cleaning agent loss (notice, this says “for a moment,” not permanently), per-formance involves matching the cleaning agent to the cleaning equipment, cleaning action,drying action, materials compatibility, and fixturing (including sample handling andautomation). In addition, one must take into account a host of additional site-specific con-siderations, including process costs, safety, and regulatory issues.

The Chicken or the Egg?

Integrating the cleaning system, the drying system, and overall sample handling canbe crucial in maintaining process control and process efficiency. For this reason it is impor-tant to match the cleaning agent or agents with the cleaning equipment. The question isoften asked: Should I first look at the cleaning agent or at the cleaning equipment? Cleaningagent manufacturers typically say: Look first at the cleaning agent. Cleaning equipmentmanufacturers, on the other hand, often say: Don’t worry, just buy the cleaning system, youcan use many different sorts of cleaning agents in it. In the opinion of this author the mostproductive approach is to consider both factors at the same time.

Looking at the cleaning agent without considering the cleaning system can lead one torule out what could be the optimal approach to cleaning. Pragmatic experience leads thisauthor to conclude that attempting to emulate a sophisticated cleaning system at thebenchtop level, by cleaning coupons or scrap parts in beakers, is all too often not informa-tive and can lead the user to discard as unworkable what could be an effective, economicalcleaning process. At the same time, some engineers have been known to contact cleaningequipment manufacturers, demand cleaning equipment be fabricated to exacting specifi-cations, and then, only after the equipment has been built and delivered, inquire about the


appropriate cleaning solution to be used. Such scenarios have led one colleague to aban-don a career as manager of an applications laboratory.

Cleaning Action

If soils are considered in a general sense as “matter out of place,” then cleaning actionis inherently important to assist in:

• Bringing the cleaning agent in contact with the parts to be cleaned• Assisting in solubilization• Assisting in removal of particles• Keeping the soil away from the parts or components• Maintaining cleanliness of the cleaning solution• In rinsing, assisting in removal of cleaning agent and/or continuation of soil

removal

Cleaning action is often discussed in terms of aqueous systems, as a means of assisting inemulsifying soils. However, cleaning action cannot be ignored in solvent, semiaqueous,and cosolvent systems.

Drying Action

Drying is considered a separate topic and is discussed in detail elsewhere. In consid-ering possible drying techniques, it must be emphasized that the drying technique must beintegrated with the cleaning system.

Again, drying is most often considered in terms of aqueous systems. Often, in solventsystems drying is assumed to be an inherent part of the process. However, particularlywith complex, ornate components, the capability of the system to remove residual solventrapidly and effectively without damage to the component must be considered.

Materials Compatibility

Typically, cleaning agent compatibility is considered in static systems, in beakers thatare perhaps heated. However, as is pointed out in Chapter 3.3 covering compatibility by E.Eichinger, the interaction of cleaning agent(s) with materials of construction can bemarkedly enhanced by cleaning action. Ultrasonic cleaning can produce a sonochemistryeffect that may both enhance cleaning and adversely impact materials of construction(Chapter 2.2 by J. Fuchs). Such effects are reportedly particularly pronounced with someaqueous surfactant packages.

It should be pointed out that other types of cleaning action, such as forceful sprays, alsohave the potential to damage parts by deformation and erosion. Such erosion may not bevisible, but may show up in altered tolerances or gravimetric changes. For example, in test-ing alternative defluxing systems, electronics component simulators were designed withbrass coupons used to simulate components raised to varying heights.2 Effectiveness ofremoval of rosin flux from under the components was determined gravimetrically. In a fewcases, over 100% apparent removal of flux was observed; this was determined to be due toerosion of the coupons.


Fixturing, Parts Handling, Automation

Choosing the appropriate fixturing and sample handing techniques are important in:

• Maximizing contact with the cleaning agent• Draining and removal of the cleaning agent• Avoiding parts deformation or damage

For example, even in solvent-based systems, parts rotation is often used in conjunctionwith ultrasonic cleaning to boost cleaning effectiveness. Inattention to fixturing and partshandling can result in inadequate cleaning and drying, as well as in parts damage. It mustbe continuously emphasized that the use of fixture and appropriate fixture design is impor-tant for all cleaning systems, including solvent and aqueous systems. Some of the factorsinvolved in appropriate fixture design are discussed in other chapters of this section and inChapter 4.2 by Callahan, and are summarized in the overview of drying (Chapter 2.18 byKanegsberg).

In addition, one must consider several other factors. Long-term compatibility of mate-rials of construction with the cleaning agent and with the cleaning technique (ultrasonics,spray systems) are essential. If worn or damaged fixtures corrode or oxidize or shed parti-cles, soil can be inadvertently introduced during the process.

Automation of the process can improve performance and consistency, minimize loss ofcleaning agent, minimize worker exposure to chemicals, and may have desirable environ-mental impacts.2,3 The concept of automation is a matter of degree. Certainly, even a processconsisting of a simple open-top degreaser with a manual hoist could be said to be more auto-mated than one where assemblers hand-scrub each part or component individually.

However, one generally thinks of automation as being a bit more sophisticated. In gen-eral, parts handling is either batch or in-line. In classic in-line processes, samples are carriedalong a conveyor belt through various cleaning and rinsing solutions; in such cases, dryingis typically through air or nitrogen nozzles. In batch-automated cleaning, overhead hoists orrobotics are used to carry parts from to various cleaning, rinsing, and drying chambers.

A batch automation system generally consists of a mechanical superstructure, a drivesystem, a control package, and an operator interface.3 Engineers accustomed to in-lineprocesses may be reluctant to try batch cleaning, in part, because they may associate batchcleaning with inconsistent, nonautomated processes. Batch cleaning actually can providemore control and more flexibility than in-line cleaning because with in-line cleaning onecan either vary the length of the process chambers (and this is predetermined during equip-ment design) or the speed of the conveyor belt. Each step of the process is therefore inher-ently tied to the next. Because the duration of each phase of the process can be variedseparately, batch automation can actually provide much more process flexibility. By usingseveral robotic arms, loads can be processed sequentially to maintain process flow (i.e., oneload of parts can be at the drying stage while another is in the wash stage). Cleaning is alsotypically automated in more-sophisticated single-chamber tanks. In such batch processes,the part remains in the chamber for all cleaning steps. Such processes can be very flexible,but, since only one load can be processed per chamber, throughput of parts is limited bythe size of the chamber and by the number of chambers.

An automated system may or may not provide a more rapid cycle time. Often, assem-blers are accustomed to speeding up the process, by quickly submerging and then remov-ing the part to be cleaned. However, there are benefits to automation, such as:


• Improved process control• Improved cleaning consistency• Regulatory, safety, or quality compliance• Lower consumption of cleaning agent• Lower solvent emissions• Lower exposure of employees to potentially toxic cleaning agents

In designing the automation process, it is important to consider not only the immedi-ate cleaning process but also those before and after. In a number of manufacturing facili-ties, the cleaning process itself is automated, but other aspects of assembly are carried outby hand. Or, the fixturing and sample handling for the cleaning process may not mesh wellwith the surrounding processes. One then observes technicians spending appreciable timere-racking and re-fixturing samples to go from one piece of equipment to the next. A bit ofadvance planning in designing the entire build process can alleviate the problem.

Even in automation, the human factor is critical. Automated systems can reduce expo-sure of employees to cleaning agents in terms of both inhalation and skin adsorption. Thereare other safety issues. One of my colleagues consistently showed up at meetings with abump on his head with ever more interesting stories of attempts to install an overheadhoist. The systems must be designed to prevent employee injury during operation, andthey are best designed and installed by experts in the field.

The other aspect of the human factor is employee education and training. All thethoughtful planning and programming can be undone if the technicians accelerate theprocess speed to undesirable levels. This is a real problem, particularly where costs andproduction pressure have built up.

PROCESS EFFICIENCY, PROCESS COSTS, ENVIRONMENTALAND SAFETY CONCERNS

Overall process costs and efficiency are very difficult to determine. It may be necessaryto take educated guesses. In terms of costs, one must consider such factors as initial capitalcosts, costs of disposables, cleaning agents costs (concentrated vs. effective dilution), bathlife, loss of cleaning agent through evaporation/dragout/dragin, disposal costs for thespent cleaning agent, costs of safety and environmental controls, regulatory costs, energyefficiency, and rework costs.

Process costs are site specific. While efficiency claims are prevalent, hard data are not.Further, studies of process costs, whether by governmental agencies or by cleaning agentor cleaning equipment suppliers, are inevitably influenced by the economic interests andpolitical agendas of those sponsoring the studies. All studies have validity; all must belooked at in context. A summary of factors in process costs has been reported;4 studies areongoing.

Environmental requirements often inherently determine the menu of available processoptions in a given area. In addition, one must consider the impact not only of the cleaningagents but also of the process on worker safety. The same increased cleaning forces thatboost cleaning efficiency can magnify materials compatibility concerns—including notonly the materials that make up the part to be cleaned but also the materials that make upthe workers in the area. Safety has to be considered in terms of the chemical, the process,and the interaction with other processes conducted in the workplace.


Plant Facilities, Floor Space, Growth

In addition to all the other considerations, one must not forget the physical limitationsof the production plant, anticipated growth, and flexibility. In evaluating cleaning systems,it is important to consider utilities (e.g., water, electricity, nitrogen lines). In addition, onemust consider floor space. It is important to look at overall equipment dimensions. Manyare careful to look at length and width; fewer consider height, and fewer still consider theadditional three-dimensional space needed for the robotic arm. One must also considerwall and door dimensions relative to cleaning equipment as well as total weight. Placementis important in terms of maintaining and cleaning the cleaning equipment itself. Based ona number of process remodeling anecdotes (e.g., “skylights” to accommodate the unantic-ipated height of robotic arms), which are amusing only in retrospect, the most realisticadvice is to involve the facilities/maintenance department at the beginning of the antici-pated process change.

One must also consider growth and flexibility. Purchasing a marginally–sized cleaningsystem is a false economy. A cleaning system that just barely manages current throughputwill not be effective if business improves. All estimates of process throughput with a givensystem, particularly vendor-generated estimates, must be critically examined to see if theyare overly optimistic. In addition, while it is unrealistic to expect a cleaning system to han-dle all possible chemistries, one should avoid a cleaning system designed for only a singlecleaning chemistry. The situation may change, and it may be necessary to use anothercleaning agent. Possible changes include:

• Product design• Material modification in the component• Cleanliness standards modification• Cost of cleaning agent• Composition of cleaning agent• Availability of cleaning agent• Regulatory changes

If several cleaning systems are essentially equivalent in performance, in general, it is betterto select the more flexible system.

Employee Involvement, Employee Education

The newer, sophisticated automated cleaning systems provide very thorough, consis-tent cleaning. However, employee education (as opposed to attempted rote training) is nec-essary to maintain process quality and to assure employee safety. Equipment maintenanceis also more complex with sophisticated equipment.4 Sophisticated computer program-ming or at least reasoned following of prearranged steps can degenerate into a bravado ofdesperate button-pushing by a terrified employee—this can be disastrous. Initial educationwhen the equipment is installed is helpful; follow-up training at intervals is the key toongoing success.

In addition, in this author’s observation, the dedicated and resourceful employeeintent on cutting corners and speeding up the process can override even the most-sophis-ticated interlock system. Ongoing monitoring of employee performance and behavior isessential.


In-House Equipment Design

In the course of evaluating sophisticated but deceptively simple equipment, manyengineers, in an attempt to control costs and/or achieve customization, are tempted todesign the equipment themselves. The general advice in such cases is that if the equipmentis similar in design to the product being manufactured, there is a chance of success.Occasionally, in-house equipment design can be very successful (see Chapter 2.8 byPetrulio).

In most cases, however, the advice is: Don’t design your own equipment. Some factorsinvolve:

• Experience: You don’t have to be a rocket scientist to design a quality vapordegreaser, but you do need practical experience in the area. Most rocket scientistsprobably do not have this experience.

• Cost: In looking at a half-million dollar box, you may be tempted to design one.However, you must consider your time and research effort. Each commerciallyproduced cleaning system may be cost-effective for the manufacturer to produce,but the engineering effort to produce the initial model is typically significant.

• Product support: If the equipment breaks, who is going to fix it? If you designed it,and if you break it, you have to fix it.

• Permitting, regulatory issues: Often, regulators and fire inspectors are more com-fortable with a standard equipment design, which may be certified to meet cer-tain design standards.

ULTRASONIC CLEANING

Ultrasonic and megasonic cleaning provide manufacturers very powerful cleaningtechniques. The question why so many chapters are needed in this book arises, particularlybecause manufacturers may consider the choice of ultrasonic equipment to be generic. Thisis probably, in part, due to the large number of variables to consider.

There are several reasons ultrasonic cleaning should be emphasized. For one thing,ultrasonic cleaning is useful over a very wide range of applications. It is a technique that isnearly certain to become increasingly important both to extend the range of applicationsamenable to aqueous cleaning and to clean products with increasing miniaturization, com-plexity, and close tolerances. In addition, despite the advances in theoretical understand-ing of ultrasonics, much needs to be learned about the mechanism of action. Ultrasonictechniques are controversial; opinions differ markedly and are also influenced by theexpert’s view of the relative efficacy of aqueous vs. solvent cleaning. Therefore, some of thestatements in various chapters may be contradictory. If this makes readers somewhatuneasy, the unease reflects the fact that ultrasonic and megasonic cleaning is an exciting,dynamic, and sometimes contentious area of technology.

The chapters provide a snapshot of at least a portion of current understanding of ultra-sonic and megasonic techniques. New techniques are constantly under development. Forexample, processes have been proposed that do not involve added chemistry or that usewater alone.5 Very specialized ultrasonic techniques involving no cleaning agent, only con-tact between the ultrasonics source and the component to be cleaned, are also possible, inspecialized situations.


*The description of this technology was contributed by Rebecca Overton, who has had many years of experiencein this technology, most recently at Cobehn.

Lack of standardization is a real issue. One reason that ultrasonic systems are toooften thought of as generic is the absence of any standardized way of comparing perform-ance of various systems. There is always a concern with transducer degradation over timeas well as a lingering questions of comparisons of systems produced by differentvendors or even of two identical models produced by a given equipment supplier.Detection and quantitation of ultrasonic systems are very difficult. Holding an alu-minum foil coupon in the tank and then inspecting it for the characteristic orange peeldesign that indicates reasonable cavitation (and hoping not to see erosion of the foil, indi-cating a hot spot or a potentially undesirable cleaning solution) can provide hours of playvalue. However, results are not quantitative. The slurry wand provides a simple andeconomically accessible means of determining that cavitation is occurring, as well asrelative frequency, hot spots, and dead zones (see Chapter 2.4 by Pedzy). Quantitativeprobes to estimate ultrasonic performance within the tank have been proposed. One suchprobe is coming into favor. The meter can be used to confirm and monitor cavitation energyand frequency within the tank.6 Meters of this type would not necessarily be cost-effectivefor smaller, less critical applications. However, the meter does show promise for monitor-ing behavior of ultrasonic tanks as well as determining relative efficiency of cavitation ofvarious cleaning agents. In one study, efficacy of wafer cleaning was related to cavitationenergy as quantitated with the quantitative meter.7

OTHER CLEANING SYSTEMS

Cleaning systems undergo constant changes, refinement, and development. This bookdoes not cover all cleaning systems available. Instead, some key cleaning techniques havebeen highlighted. It is anticipated that from the examples, principles, and reasoningprocesses, the practicing engineer will be able to extrapolate to evaluate other cleaning sys-tems. A few additional cleaning systems are summarized.

Spray Cleaning Systems*

The spray cleaning process has been in use for many years with high reliability, smallmechanical, optics, microelectronics, and other extremely sensitive and demanding com-ponents. Most often, it is used in those areas of production, such as military, aerospace, andmedical applications, where components and subassemblies must be totally free of allorganic and inorganic contaminants. Spray systems have successfully met exacting clean-ing requirements for delicate wire bonds without damage, to clean both blind and through-holes of extremely small diameters (e.g., 0.0005), and the removal of both organic and inor-ganic contaminants. Such systems utilize a gentle agitation via the fine, usually venturitype, spray and the molecular weight of the chosen solvent.

Often in this type of process, when more traditional volatile of solvents are utilized,there is little or no waste product and virtually no effluent to handle. The solvents are gen-erally used at a rate of 0.2 fluid oz/s, a minimal amount at best, and because the cleaningcycles tend to be very short in duration, an average of 5 s of solvent spray, the solvent isevaporated when these types of systems are properly used. This provides several advan-tages, the most important of which may be the potential to add carbon absorption in gasphase to minimize solvent emissions. This fume and solvent extraction process can be con-


tained in a small unvented chamber, thereby eliminating the exhausting of waste to theatmosphere through existing fume exhausting systems, or the expense of modifying thephysical plant.

The spraying process can totally replace all other processes only in small R&D typefacilities and is generally not intended for high-production-requirement facilities. Whererequirements are less than 100 components per day it is often the ideal solution to an oth-erwise highly expensive problem.

Reel-to-Reel or Continuous Web Cleaning

Reel-to-reel systems are used for such diverse processes as cleaning metal wire, stripsof lead frames, and motion picture film. In such systems, it is helpful to picture giant spoolsof thread (with the thread being made of the material to be cleaned). The part is a continu-ous filament running from one spool, through a cleaning solution, through a drier, and thenonto another spool. Ultrasonics or other agitation may assist in cleaning; rollers or brushesmay assist in removing excess cleaning agent prior to drying. Such cleaning systems can beaqueous or solvent-based. Often, 1,1,1-trichloroethane or another chlorinated solvent wasused.

Where chlorinated solvents are used, manufacturers initially found it difficult tocomply with the Halogenated Solvents NESHAP (National Emissions Standard forHazardous Air Pollutants). The initial NESHAP simply did not consider specific issuesassociated with reel-to-reel cleaning. National regulations specifically covering such appli-cations are being developed.8

Motion picture film cleaning is a very specialized, fascinating challenge in reel-to-reelcleaning, involving cleaning performance, speed, materials compatibility, and long-termstorage. Motion picture film has high aesthetic, technical, and informational value. Thesubstrate is complex, composed of multiple layers of plastics and other synthetics. Nearlyall current film-cleaning equipment is designed for use with solvents. Classically, motionpicture film was cleaned with TCA.

Because of the ODP production phaseout, most film is currently cleaned with per-chloroethylene. Regulatory and technical issues have generated great interest in findingother cleaning alternatives. The delicacy and complexity of materials of construction, andlong-term storage issues make acceptance of appropriate cleaning systems for motion pic-tures a difficult task.

A list of solvent options for film cleaning which have acceptable physical propertiesand maintain acceptable image stability9 is updated periodically. Water-based cleaning inredesigned equipment has been tested, but acceptance has been limited at best because ofproblems with materials compatibility and lengthy, possibly incomplete, drying.Alternative cleaning systems, some with redesigned equipment, have been introduced anduse such solvents as isopropyl alcohol, hydro-treated naphtha, and the newer engineeredfluorinated solvents.10 One promising system was discussed at a recent technical confer-ence. It uses HFE 7200 (packaged for film cleaning as 8200) in Lipsner-Smith cleaningequipment which has been redesigned with additional, powerful ultrasonic transducers toboost cleaning power of this very mild solvent along with more effective solvent contain-ment. Results of initial testing by an end user were very encouraging.11

After over a decade of effort, a cleaning agent matching all of the desirable propertiesof TCA has not yet been found. Many end users would prefer a cleaning agent that couldbe readily adapted to existing cleaning equipment. In addition, there are ongoing issues ofwhether video or some other format will totally replace motion picture film. If this were to


happen, cleaning issues could become a moot point. However, motion picture film remainsa medium of choice, at least for master copies.

Centrifugal Cleaner

In centrifugal systems, parts are cleaned in a single centrifuge chamber filled withcleaning agent. Cleaning at low centrifugal force promotes agitation due to Coreolis mix-ing. Initial costs may be high, and throughput is often lower than conventional methods,but equipment is often designed for use with many different cleaning agents. Centrifugalcleaners may use water- or solvent-based cleaning agents.

Spinners

Spinners are used in cleaning and surface preparation of wafers and optics. The com-ponent is placed on a rapidly moving turntable and various cleaning and rinsing solutionsare applied to the surface to be cleaned. Sequential cleaning with various cleaning agentsas well as rinsing and spot-free drying is carried out in place. These systems are typicallydesigned to be used in clean rooms where contamination is an issue. Avoiding particles,residue from additive packages, and excess foaming are considerations. Spinners may usewater and/or solvent-based cleaning agents within a single process.

Microclusters

Microcluster cleaning is a specialized line-of-sight technique that has been demon-strated to remove submicron particles. Microclusters are produced by atomizing a con-ducting liquid (typically a mixture of n-methyl pyrollidone and water), which is thenexposed to high electric fields. The microcluster dimensions are in the range of that of thecontaminants to be removed. The technique is proposed for wafer cleaning, but may haveother high-end applications.12

Industrial Cleaners/Cabinet Washers, Dishwashers, Spray Cabinets

This category includes a vast array of cleaning systems most often associated with gen-eral industrial cleaning.13 Except where indicated, they are primarily used with aqueouscleaning agents. Some resemble consumer-variety dishwashers, perhaps with stainlesssteel interiors.

Sink-on-a-drum cleaning systems are used in general metals-cleaning applicationswith mineral spirits, or, particularly in areas that are heavily regulated because of poor airquality, with an aqueous-based cleaning agent or a VOC-exempt solvent. They can be con-structed of metal or plastic. Some solvent-based systems have been adapted to provide on-board recycling of the cleaning agent (Chapter 2.7 by Skelly).

Cabinet washers are typically tall, cylindrical systems. The parts (engines, etc.) are typ-ically placed on a turntable and sprayed with hot surfactant solution.

In spray cabinets, the part to be cleaned is placed inside a boxlike container and sprayedtypically with an aqueous-based cleaning chemistry. Spray cabinets may be manual or auto-mated. The manual models resemble a glove box. Automated models, such as are used bythe automotive industry, can be quite large and sophisticated and may include in-line (con-veyor belt) or overhead robotics, with automated monitoring of the cleaning solution. Someaqueous systems have even been fitted with high-pressure spray for tube-cleaning.14


As with other cleaning systems, the cleaning chemistry, temperature, force of cleaningaction, filtration, rinsing, and, in some cases, drying all impact cleaning quality. It is impor-tant to match the equipment to the application. In many general metals-cleaning applica-tions, a single cleaning tank with either solvent or aqueous-based material is sufficient. Thefeatures, capabilities, prices, and quality of construction vary by orders of magnitude.Particularly because such systems will be subjected to extremes of temperature and ofcleaning chemistry, purchasing a well-designed system of high quality can result in long-term benefits.

The user must consider initial costs, ongoing costs (including disposables such as fil-ters), and labor and rework costs. Some of the more-sophisticated systems require a highlevel of ongoing maintenance.4

Semiaqueous and Cosolvent Systems

Semiaqueous and cosolvent systems are related. In semiaqueous systems, a high-boil-ing solvent-based cleaning agent is used for primary removal of soils. This step is followedby several rinsing and drying steps. Cosolvent cleaning can imply two or more solvents,used in a single tank or sequentially. Technically, any solvent blend or azeotrope could beconsidered cosolvent cleaning. This discussion, however, considers cosolvent cleaning tobe a process in which a high-boiling solvent blend is rinsed in a second solvent, solventblend, or azeotrope. In cosolvent cleaning, typically a lower-boiling solvent is chosen toserve as a rinsing and vapor-phase drying agent. In some cases, a supplier may offer twovery similar blends based on hydrocarbon, d-limonene, or ester, which differ in subtlechanges in the additive packages to make them more readily removable with water or withsolvent. Other products are based on complex, modified alcohols. Many high-boiling sol-vent blends are considered competition sensitive by the manufacturers and, therefore,unfortunately are shrouded in mystery. This, of course, makes rational process design achallenge. In such situations, the end user would be well advised to set up comprehensiveproduct support arrangements.

While water is the rinsing agent in semiaqueous cleaning, the rinse/vapor phase/dry-ing agent can be any of a number of lower-boiling solvents, such as HFC, HFE, isopropylalcohol, even isopropyl alcohol/cyclohexane azeotrope (which would, one supposes, con-stitute a co-cosolvent process).

Both semiaqueous and cosolvent systems have potential advantages and drawbacks.In both cases, because compounds and mixtures with widely different solvency character-istics are used for cleaning and rinsing (we are considering rinse water as a cleaning agent),the rinse phase can in a sense be considered part of the cleaning phase. Both types of sys-tems can extend the range of soils that can be removed; both can allow the use of high-boil-ing cleaning agents that may themselves leave residue on the part and/or may not drysufficiently rapidly.

Both types of systems can use agitation, including ultrasonic cleaning, to improve per-formance. In some cases, the cleaning agent is designed to form an emulsion with the rins-ing agent, either for initial cleaning or as final rinsing. The emulsion can be stable ortransient (i.e., the emulsion exists only during agitation). Alternatively, the cleaning andrinsing agents may be miscible. For stable emulsions as well as with miscible cleaning andrinsing agents, the issues of recovering the cleaning and rinsing agents and of waste streammanagement become more complex. Depending on the situation, reverse osmosis may beneeded for recovery. Multiple filtration of the waste stream is typically needed in semi-aqueous systems.


Initially, semiaqueous and cosolvent systems were widely and enthusiaticallyadopted. However, these systems are relatively complex and sophisticated. Like othernewer cleaning systems, they require maintenance, employee education, and process mon-itoring. In both cases, proper fixturing of the product is crucial to assure optimal cleaningand to avoid excessive carryover of cleaning agent into the rinse tank. This author recallsimplementing a semiaqueous cleaning system with in-line automation that initially per-formed very acceptably. Then came the phone call: the system does not work; the cleaningagent is gone; all of the filters for the wastewater are “dead.” The reason turned out to becultural and historical, and involved the legendary third shift. As it happened, the facilitywas located in a town that held monthly auto racing. Over the years, the third shift hadbecome accustomed to cleaning their carburetors in the vapor degreaser. With vapordegreasing, because final cleaning is in freshly distilled solvent, the extraneous soils didnot present a problem. A semiaqueous system is not as forgiving. When carburetors wereplaced on the conveyor belt of the semiaqueous system, there was carryover of excessiveamounts of cleaning agent into the rinse tank, resulting in system failure. The scope ofcleaning allowed in semiaqueous, cosolvent, or any other cleaning system is the preroga-tive of management. The point is that semiaqueous and cosolvent systems are not forgiv-ing of mediocre process control.

It should also be noted that the physical properties, including flammability, have to becarefully considered in designing a system. A semiaqueous cleaning system requires mul-tiple rinse tanks (for in-line systems, a fairly long portion of the conveyor belt should bedevoted to rinsing). Depending on the product and the next step in the process, drying willbe required. Even though the primary cleaning occurs below the boiling point of the clean-ing agent, a typical cosolvent system is designed more like a multitank degreaser, to allowfor vapor-phase rinsing and drying. In some cases, vapor degreasers have been convertedto cosolvent systems. However, it should be remembered that if a low-flash-point solventis employed, a standard vapor degreaser would not be suitable and would pose a poten-tial fire hazard.

Wet Benches

Biomedical devices, optics, semiconductors, and microelectronics are often processedand cleaned in wet benches. This is a broad, generic term for a series of cleaning tanks thatmay contain aqueous, semiaqueous, or solvent-based cleaners.

In addition, etching with strong acids and bases may take place. While such processesmight better be considered as surface modification, they are certainly related to cleaning. Itshould be noted that appropriate process controls, such as titrators, are desirable and maybe crucial to maintain process control. In addition, with strong acids such as hydrofluoricacid, process automation, vapor monitors, and controlled bath neutralization may beneeded for adequate employee protection.

Impingement Cleaning

Impingement cleaning covers an array of processes, including line-of-sight high-forcesolvent and aqueous sprays, CO2 snow, and dry bicarbonate. Impingement cleaning hasbeen used in critical applications for many years.

In metal finishing and deburring, a variety of nonsolvent, nonaqueous media are used.The equipment manufacturer often considers this equipment as separate from cleaning,but there is overlap. For example, when faced with the choice of using chemical stripping


or mechanical stripping of paint, a facility that repairs pumps chose media blast for a vari-ety of safety, environmental, and economic considerations.15

Increasingly, other forms of impingement cleaning are being adapted from generalcleaning to meet critical cleaning and surface finishing requirements. Examples includeplastic pellets, walnut shells, sand, diamond dust, nails, aluminum oxide, garnet, smallnails—the list goes on and on. The impingement or rubbing action of the media itself maybe achieved with blast, agitation, centrifugation, or ultrasonics, and cleaning may be dry orin an aqueous or aqueous-surfactant media.16

Although some forms of impingement cleaning are becoming widely adopted, othermedia cleaning may become more important in the future for high-precision, critical appli-cations. For example, selective removal of high-value coatings in optics and wafer fabrica-tion bears some relationship to selective removal of paint. As with other cleaningapproaches, there will be provisos in expanding media blast. Speed, potential part damage,and preventing residual blast media from recontaminating the part are among the considerations.

CONCLUSION

Cleaning applications and requirements are exceedingly diverse. To meet these needs,an array of cleaning agents and cleaning equipment has been developed. This chaptersummarizes a few aspects of cleaning systems; other chapters discuss the specifics of clean-ing systems in much greater detail.

In evaluating various cleaning processes, the reader must remember that, inherently,cleaning involves a melding of cleaning agent, cleaning action, and overall process equip-ment. In developing processes, some argue that the cleaning agent should be selected priorto the cleaning equipment, or vice versa. There is some validity to either approach, as longas it moves us along the path to ADS (actually doing something). However, given the com-plexity of performance, economics, and environmental requirements, it is often more pro-ductive to consider the cleaning agent and the cleaning system in parallel, then making thechoices.

In addition, even though some cleaning processes may not be widely used in a partic-ular industry, the reader is urged to at least skim through all of the available processchoices. By looking at processes creatively and in a more encompassing manner, it may bepossible to adapt processes from one area of manufacturing to another. This kind of adap-tation fosters overall progress and, to the thoughtful manufacturing company, specificcompetitive advantage.

REFERENCES

1. E. Kanegsberg, Liquid filtration in critical cleaning, A2C2 Mag., April, 2000.2. B. Kanegsberg, Successful Cleaning/Assembly Processes for Small to Medium Electronics

Manufacturers, half-day workshop, Nepcon West ‘97, Anaheim, CA, February 27, 1997.3. J. Aries, Moves toward automation, Parts Cleaning Mag., February, 1998.4. B. Kanegsberg and C. LeBlanc, The cost of process conversion, CleanTech ‘99, Rosemont, IL, May,

1999.5. J. Baker and J. Durkee, Rethinking cleaning processes III, A2C2 Mag., 3, 39–40, 2000.6. L. Azar, ppb Inc., personal communication.


7. Y. Wu, C. Franklin, M. Bran, and B. Fraser, Acoustic Property Characterization of a Single WaferMegasonic Cleaning, presentation and proceedings, Electrochemical Society 196th Meeting,Honolulu, HI, October, 1999.

8. Fed. Regist., 64, (232), 67793–67803; 1999 DOCID:fr03de99-30, amendments to the NationalEmission Standards for Hazardous Air Pollutants: Halogenated Solvent Cleaning, DocketA-92-39. Available at http://www.epa.gov/fedrgstr/EPA-AIR/1999/December/Day-03/a31356.htm.

9. Eastman Kodak, Film Cleaning Solvents, Film Cleaning Solvent Options, available athttp://www.kodak.com/US/en/motion/hse/solvent.jhtml?id=0.1.4.5.16.6&lc=en.

10. L. Smith, Clean Technology News, available at http://www.rti-us.com/newsletters/clnthnws.html.

11. J. Banks, An Alternative to Chlorinated Solvents for Deep Film Cleaning in Telecine Suites,Archives, and Film Laboratories, presentation at the 142nd SMPTE Technical Conference andExhibition, Pasadena, CA, October 20, 2000.

12. J. Perel, C. Sujo, and J.F. Mahoney, Microclusters make an impact on wafer cleaning, PrecisionCleaning Mag., 7, 18–24, 1999.

13. J.B. Durkee, The Parts Cleaning Handbook, Gardner Publishers, Inc., Cincinnati, OH, 1994.14. S.J. Adam, Aqueous Tube Cleaning Advances at McDonnell Douglas Aerospace, in Proceedings of

the 2nd Aerospace Environmental Technology Conference, Huntsville, AL, August. 6–8, NASAConference Publication 3349, 1996, 145.

15. P. Maluso and B. Kanegsberg, Hydrostatic Pump Rebuild: Implementing Aqueous, Steam andSolvent Free Processes, in Proc. Tenth Annual International Workshop on Solvent Substitution and theElimination of Toxic Substances and Emissions, Scottsdale, AZ, September 13–16, 1999.

16. E. Kanegsberg and B. Kanegsberg, Cleaning by abrasive impact, A2C2 Mag., May 2000.


http://www.epa.gov/fedrgstr/EPA-AIR/1999/December/Day-03/a31356.htm

http://www.kodak.com/US/en/motion/hse/solvent.jhtml?id=0.1.4.5.16.6&lc=en

http://www.rti-us.com/newsletters/clnthnws.html

http://www.epa.gov/fedrgstr/EPA-AIR/1999/December/Day-03/a31356.htm

http://www.rti-us.com/newsletters/clnthnws.html

CHAPTER 2.2

The Fundamental Theory and Applicationof Ultrasonics for Cleaning

F. John Fuchs

CONTENTS

IntroductionWhat Is Ultrasonics?The Theory of Sound Waves

Sound Wave GenerationThe Nature of Sound Waves

Benefits of Ultrasonics in the Cleaning and Rinsing ProcessesUltrasonics Speeds Cleaning by DissolutionUltrasonic Activity Displaces ParticlesComplex ContaminantsA Superior Process

Ultrasonic EquipmentUltrasonic Generator

Square Wave OutputPulseFrequency Sweep

Frequency and AmplitudeUltrasonic Transducers

MagnetostrictivePiezoelectric

Ultrasonic Cleaning EquipmentMaximizing the Ultrasonic Cleaning Process Parameters

Maximizing CavitationImportance of Minimizing Dissolved Gas

Ultrasonic PowerUltrasonic Frequency

Maximizing Overall Cleaning EffectUltrasonic Power

Conclusion


INTRODUCTION

Cleaning technology is in a state of change. Vapor degreasing using chlorinated andfluorinated solvents, long the standard for most of industry, is being subjected to increasedregulatory requirements in the interest of the ecology of our planet. At the same time, clean-ing requirements are continually increasing. Cleanliness has become an important issue inmany industries where it never was in the past. In industries such as electronics wherecleanliness was always important, it has become more critical in support of growing tech-nology. It seems that each advance in technology demands greater and greater attention tocleanliness for its success. As a result, the cleaning industry has been challenged to deliverthe needed cleanliness and has done so through rapid innovation over the past severalyears. Many of these advances have involved the use of ultrasonic technology.

The cleaning industry is currently in a struggle to replace solvent degreasing withalternative “environmentally friendly” means of cleaning. Although substitute water-based, semiaqueous, and petroleum-based chemistries are available, they are often some-what less effective as cleaners than the solvents and may not perform adequately in someapplications unless a mechanical energy boost is added to assure the required levels ofcleanliness. Ultrasonic energy is now used extensively in critical cleaning applications toboth speed and enhance the cleaning effect of the alternative chemistries. This chapter isintended to familiarize the reader with the basic theory of ultrasonics and how ultrasonicenergy can be most effectively applied to enhance a variety of cleaning processes.

WHAT IS ULTRASONICS?

Ultrasonics is the science of sound waves above the limits of human audibility. The fre-quency of a sound wave determines its tone or pitch. Low frequencies produce low or basstones. High frequencies produce high or treble tones. Ultrasound is a sound with a pitch sohigh that it cannot be heard by the human ear. Frequencies above 18 kHz are usually con-sidered to be ultrasonic. The frequencies used for ultrasonic cleaning range from 20,000cycles per second or 20 to over 100 kHz. The most commonly used frequencies for indus-trial cleaning are those between 20 and 50 kHz. Frequencies above 50 kHz are more com-monly used in small tabletop ultrasonic cleaners, such as those found in jewelry stores anddental offices.

THE THEORY OF SOUND WAVES

To understand the mechanics of ultrasonics, it is necessary first to have a basic under-standing of sound waves, how they are generated and how they travel through a conduct-ing medium. The dictionary defines sound as the transmission of vibration through anelastic medium which may be a solid, liquid, or a gas.

Sound Wave Generation

A sound wave is produced when a solitary or repeating displacement is generated in asound-conducting medium, such as by a “shock” event or “vibratory” movement (Figure1). The displacement of air by the cone of a radio speaker is a good example of vibratorysound waves generated by mechanical movement. As the speaker cone moves back andforth, the air in front of the cone is alternately compressed and rarefied to produce soundwaves, which travel through the air until they are finally dissipated. We are probably most


Figure 1 Vibratory and shock events.

Figure 2 Coils of a spring representing individual molecules of a sound-conducting medium.

familiar with sound waves generated by alternating mechanical motion. There are alsosound waves that are created by a single “shock” event. An example is thunder, which isgenerated as air instantaneously changes volume as a result of an electrical discharge(lightning). Another example of a shock event might be the sound created as a woodenboard falls with its face against a cement floor. Shock events are sources of a single com-pression wave that radiates from the source.

The Nature of Sound Waves

Figure 2 uses the coils of a spring similar to a Slinky® toy to represent individual mol-ecules of a sound-conducting medium. The molecules in the medium are influenced byadjacent molecules in much the same way that the coils of the spring influence one another.The source of the sound in the model is at the left. The compression generated by the soundsource as it moves propagates down the length of the spring as each adjacent coil of the


spring pushes against its neighbor. It is important to note that, although the wave travelsfrom one end of the spring to the other, the individual coils remain in their same relativepositions, being displaced first one way and then the other as the sound wave passes. As aresult, each coil is first part of a compression as it is pushed toward the next coil and thenpart of a rarefaction as it recedes from the adjacent coil. In much the same way, any pointin a sound conduction medium is alternately subjected to compression and then rarefac-tion. At a point in the area of a compression, the pressure in the medium is positive. At apoint in the area of a rarefaction, the pressure in the medium is negative.

In elastic media such as air and most solids, there is a continuous transition as a soundwave is transmitted. In nonelastic media such as water and most liquids, there is continu-ous transition as long as the amplitude or “loudness” of the sound is relatively low. Asamplitude is increased, however, the magnitude of the negative pressure in the areas of rar-efaction eventually becomes sufficient to cause the liquid to fracture because of the nega-tive pressure, causing a phenomenon known as cavitation.

As shown in Figure 3, cavitation “bubbles” are created at sites of rarefaction as the liq-uid fractures or tears because of the negative pressure of the sound wave in the liquid. Asthe wave fronts pass, the cavitation “bubbles” oscillate under the influence of positive pres-sure, eventually growing to an unstable size. Finally, the violent collapse of the cavitationbubbles results in implosions, which cause shock waves to be radiated from the sites of thecollapse.

The collapse and implosion of myriad cavitation bubbles throughout an ultrasonicallyactivated liquid result in the effect commonly associated with ultrasonics. It has been cal-culated that temperatures in excess of 10,000°F and pressures in excess of 10,000 psi aregenerated at the implosion sites of cavitation bubbles.

Figure 3 Cavitation and implosion.

Cycle repeatsnew bubble growth

Bubbles collapsein compression

Maximumbubble size

Cavitation bubblegrowth in negative

pressure


BENEFITS OF ULTRASONICS IN THE CLEANINGAND RINSING PROCESSES

Cleaning in most instances requires that a contaminant be dissolved (as in the case ofa soluble soil), displaced (as in the case of a nonsoluble soil), or both dissolved and dis-placed (as in the case of insoluble particles being held by a soluble binder such as oil orgrease). The mechanical effect of ultrasonic energy can be helpful in both speeding disso-lution and displacing particles. Just as it is beneficial in cleaning, ultrasonics is also benefi-cial in the rinsing process. Residual cleaning chemicals are removed quickly andcompletely by ultrasonic rinsing.

Ultrasonics Speeds Cleaning by Dissolution

In removing a contaminant by dissolution it is necessary for the solvent to come intocontact with and dissolve the contaminant. The cleaning activity takes place only at theinterface between the cleaning chemistry and the contaminant. (See Figure 4.) As the clean-ing chemistry dissolves the contaminant, a saturated layer develops at the interfacebetween the fresh cleaning chemistry and the contaminant. Once this has happened, clean-ing action stops as the saturated chemistry can no longer attack the contaminant. Freshchemistry cannot reach the contaminant (Figure 5). Ultrasonic cavitation and implosioneffectively displace the saturated layer to allow fresh chemistry to come into contact withthe contaminant (Figure 6) remaining to be removed. This is especially beneficial whenirregular surfaces or internal passageways are to be cleaned.

Ultrasonic Activity Displaces Particles

Some contaminants comprise insoluble particles loosely attached and held in place byionic or cohesive forces. These particles need only be displaced sufficiently to break theattractive forces to be removed (Figure 7).

Cavitation and implosion as a result of ultrasonic activity displace and remove looselyheld contaminants such as dust from surfaces. For this to be effective, it is necessary thatthe coupling medium be capable of wetting the particles to be removed (Figure 8).

Figure 4


Figure 7 Figure 8

Complex Contaminants

Contaminants can also, of course, be more complex in nature, consisting of combina-tion soils made up of both soluble and insoluble components. The effect of ultrasonics issubstantially the same in these cases, as the mechanical microagitation helps speed both thedissolution of soluble contaminants and the displacement of insoluble particles.

Ultrasonic activity has also been demonstrated to speed or enhance the effect of manychemical reactions. This is probably caused mostly by the high energy levels created ashigh pressures and temperatures are created at the implosion sites. It is likely that the supe-rior results achieved in many ultrasonic cleaning operations may be at least partially attrib-uted to the sonochemistry effect.

A Superior Process

In the above illustrations, the surface of the part being cleaned has been represented asflat. In reality, surfaces are seldom flat, instead, they comprise hills, valleys, and convolu-tions of all description. Figure 9 shows why ultrasonic energy has proved to be more effec-tive at enhancing cleaning than other alternatives, including spray washing, brushing,turbulation, air agitation, and even electrocleaning in many applications. The ability ofultrasonic activity to penetrate and assist the cleaning of interior surfaces of complex partsis also especially noteworthy.

Figure 5 Figure 6


ULTRASONIC EQUIPMENT

To introduce ultrasonic energy into a cleaning system requires an ultrasonic transducerand an ultrasonic power supply or “generator.” The generator supplies electrical energy atthe desired ultrasonic frequency. The ultrasonic transducer converts the electrical energyfrom the ultrasonic generator into mechanical vibrations.

Ultrasonic Generator

The ultrasonic generator converts electrical energy from the line which is typicallyalternating current at 50 or 60 Hz to electrical energy at the ultrasonic frequency. This isaccomplished in a number of ways by various equipment manufacturers. Current ultra-sonic generators nearly all use solid-state technology (Figure 10).

There have been several relatively recent innovations in ultrasonic generator technol-ogy which may enhance the effectiveness of ultrasonic cleaning equipment. These includesquare wave outputs, slowly or rapidly pulsing the ultrasonic energy on and off, and mod-ulating or “sweeping” the frequency of the generator output around the central operatingfrequency. The most-advanced ultrasonic generators have provisions for adjusting a vari-ety of output parameters to customize the ultrasonic energy output for the task.

Square Wave Output

Applying a square wave signal to an ultrasonic transducer results in an acoustic out-put rich in harmonics. The result is a multifrequency cleaning system that vibrates simul-taneously at several frequencies which are harmonics of the fundamental frequency.Multifrequency operation offers the benefits of all frequencies combined in a single ultra-sonic cleaning tank.

Figure 9

Figure 10 Generation of ultrasonics.


Pulse

In pulse operation, the ultrasonic energy is turned on and off at a rate that may varyfrom once every several seconds to several hundred times per second. The percentage oftime that the ultrasonic energy is on may also be changed to produce varied results. Atslower pulse rates, more rapid degassing of liquids occurs as coalescing bubbles of air aregiven an opportunity to rise to the surface of the liquid during the time the ultrasonicenergy is off. At more rapid pulse rates the cleaning process may be enhanced as repeatedhigh energy “bursts” of ultrasonic energy occur each time the energy source is turned on(Figure 11).

Frequency Sweep

In sweep operation, the frequency of the output of the ultrasonic generator is modu-lated around a central frequency, which may itself be adjustable (Figure 12).

Various effects are produced by changing the speed and magnitude of the frequencymodulation. The frequency may be modulated from once every several seconds to severalhundred times per second with the magnitude of variation ranging from several hertz toseveral kilohertz. Sweep may be used to prevent damage to extremely delicate parts or toreduce the effects of standing waves in cleaning tanks. The frequency of sweep may be var-ied randomly to prevent damage to parts susceptible to resonating at or near the sweep ratefrequency. Sweep operation may also be found especially useful in facilitating the cavita-tion of terpenes and petroleum-based chemistries. A combination of pulse and sweep oper-ation may provide even better results when the cavitation of terpenes and petroleum-basedchemistries is required.

Frequency and Amplitude

Frequency and amplitude are properties of sound waves. Figure 13A to C demonstratefrequency and amplitude using the spring model introduced earlier. If Figure 13A is thebase sound wave, Figure 13B with less displacement of the media (less intense compres-sion and rarefaction) as the wave front passes represents a sound wave of less amplitude

Figure 11 Pulse operation.

Figure 12 Frequency sweep.


or “loudness.” Figure 13C represents a sound wave of higher frequency indicated by morewave fronts passing a given point within a given period of time.

Ultrasonic Transducers

There are two general types of ultrasonic transducers in use today: magnetostrictiveand piezoelectric. Both accomplish the same task of converting alternating electrical energyto vibratory mechanical energy but do it using different means.

Magnetostrictive

Magnetostrictive transducers utilize the principle of magnetostriction in which certainmaterials expand and contract when placed in an alternating magnetic field.

Alternating electrical energy from the ultrasonic generator is first converted into analternating magnetic field through the use of a coil of wire. The alternating magnetic fieldis then used to induce mechanical vibrations at the ultrasonic frequency in resonant stripsof nickel or other magnetostrictive material that are attached to the surface to be vibrated.Because magnetostrictive materials behave identically to a magnetic field of either polar-ity, the frequency of the electrical energy applied to transducer is one half of the desiredoutput frequency. Magnetostrictive transducers were the first to supply a robust source ofultrasonic vibrations for high-power applications, such as ultrasonic cleaning (Figure 14).

Because of inherent mechanical constraints on the physical size of the hardware as wellas electrical and magnetic complications, high-power magnetostrictive transducers seldomoperate at frequencies much above 30 kHz. Piezoelectric transducers, on the other hand,can easily operate well into the megahertz range.

Magnetostrictive transducers are generally less efficient than their piezoelectric coun-terparts. This is due primarily to the fact that the magnetostrictive transducer requires adual energy conversion from electrical to magnetic and then from magnetic to mechanical.Some efficiency is lost in each conversion. Magnetic hysteresis effects also detract from theefficiency of the magnetostrictive transducer.

Piezoelectric

Piezoelectric transducers (Figure 15) convert alternating electrical energy directly tomechanical energy through use of the piezoelectric effect in which certain materials changedimension when an electrical charge is applied to them.

Electrical energy at the ultrasonic frequency is supplied to the transducer by theultrasonic generator. This electrical energy is applied to piezoelectric element(s) in the

Figure 13 Demonstration of frequency and amplitude using the spring model. (A: base, B: loweramplitude, C: higher frequency.)


transducer, which vibrate. These vibrations are amplified by the resonant masses of thetransducer and directed into the liquid through the radiating plate.

Early piezoelectric transducers utilized such piezoelectric materials as naturally occur-ring quartz crystals and barium titanate, which were fragile and unstable. Early piezoelectric

Figure 14 Magnetostrictive transducer.

Mechanical OutputFrequency = 2F

F

Output Face

Laminated nickelstrips attached tooutput diaphramby silver brazing

Electrical coilwrapped around

nickel strips

Oscillatingmagnetic field

Figure 15 Piezoelectric transducer.

Ultrasonically Active Liquid

Stainless steelnose piece

Groundconnection

Electrical insulator

Electrode

Compression bolt

Steel back mass

Piezoelectricdriving elements

Aluminumcoupling mass

Attachment of thenose piece by vacuum

brazing with copper


transducers were, therefore, unreliable. Today’s transducers incorporate stronger, moreefficient, and highly stable ceramic piezoelectric materials, which were developed as aresult of the efforts of the U.S. Navy and its research to develop advanced sonar transpon-ders in the 1940s. The vast majority of transducers used today for ultrasonic cleaning uti-lize the piezoelectric effect.

Ultrasonic Cleaning Equipment

Ultrasonic cleaning equipment ranges from the small tabletop units often found in dentaloffices or jewelry stores (Figure 16) to huge systems with capacities of several thousand gal-lons used in a variety of industrial applications. Selection or design of the proper equip-ment is paramount in the success of any ultrasonic cleaning application.

The simplest application may require only a small heated tank cleaner with rinsing tobe done in a separate container. More sophisticated cleaning systems include one or morerinses, added process tanks and hot air dryers. Automation is often added to reduce laborand guarantee process consistency.

The largest installations utilize immersible ultrasonic transducers that can be mountedon the sides or bottom of cleaning tanks of nearly any size. Immersible ultrasonic trans-ducers offer maximum flexibility and ease of installation and service.

Small, self-contained cleaners are used in doctor’s offices and jewelry stores. Heatedtank cleaning systems are used in laboratories and for small batch cleaning needs (Figure17). Console cleaning systems integrate ultrasonic cleaning tank(s), rinse tank(s), anda dryer for batch cleaning (Figure 18). Systems can be automated through the use of a

Figure 16 Small, self-contained cleaner.

Figure 17 Heated tank cleaning system.


PLC-controlled material handling system. A wide range of options may be offered in cus-tom-designed systems.

Large-scale installations or retrofitting of existing tanks in plating lines, etc., can beachieved by using modular immersible ultrasonic transducers. Ultrasonic generators areoften housed in climate-controlled enclosures (Figure 19).

MAXIMIZING THE ULTRASONIC CLEANING PROCESS PARAMETERS

Effective application of the ultrasonic cleaning process requires consideration of anumber of parameters. While time, temperature, and chemical remain important in ultra-sonic cleaning as they are in other cleaning technologies, there are additional factors thatmust be considered to maximize the effectiveness of the process. Especially important arethose variables that affect the intensity of ultrasonic cavitation in the liquid.

Maximizing Cavitation

Maximizing cavitation of the cleaning liquid is obviously very important to the successof the ultrasonic cleaning process. Several variables affect cavitation intensity.

Figure 18 Console cleaning system.

Figure 19 Large-scale installation.


Temperature is the most important single parameter to be considered in maximizingcavitation intensity. This is because so many liquid properties affecting cavitation intensityare related to temperature. Changes in temperature result in changes in viscosity, the solu-bility of gas in the liquid, the diffusion rate of dissolved gasses in the liquid, and vaporpressure, all of which affect cavitation intensity. In pure water, the cavitation effect is max-imized at approximately 160°F.

The viscosity of a liquid must be minimized for maximum cavitation effect. Viscous liq-uids are sluggish and cannot respond quickly enough to form cavitation bubbles and vio-lent implosion. The viscosity of most liquids is reduced as temperature is increased.

For most effective cavitation, the cleaning liquid must contain as little dissolved gas aspossible. Gas dissolved in the liquid is released during the bubble growth phase of cavita-tion and prevents its violent implosion, which is required for the desired ultrasonic effect.The amount of dissolved gas in a liquid is reduced as the liquid temperature is increased.

Importance of Minimizing Dissolved Gas

During the negative-pressure portion of the sound wave, the liquid is torn apart andcavitation bubbles start to form. As negative pressure develops within the bubble, gases dis-solved in the cavitating liquid start to diffuse across the boundary into the bubble. As neg-ative pressure is reduced by the passing of the rarefaction portion of the sound wave andatmospheric pressure is reached, the cavitation bubble starts to collapse because of its ownsurface tension. During the compression portion of the sound wave, any gas that diffusedinto the bubble is compressed and finally starts to diffuse across the boundary again toreenter the liquid. This process, however, is never complete as long as the bubble containsgas since the diffusion out of the bubble does not start until the bubble is compressed. Andonce the bubble is compressed, the boundary surface available for diffusion is reduced. Asa result, cavitation bubbles formed in liquids containing gas do not collapse all the way toimplosion but rather result in a small pocket of compressed gas in the liquid. This phe-nomenon can be useful in degassing liquids. The small gas bubbles group together untilthey finally become sufficiently buoyant to come to the surface of the liquid (Figure 20).

The diffusion rate of dissolved gases in a liquid is increased at higher temperatures. Thismeans that liquids at higher temperatures give up dissolved gases more rapidly than thoseat lower temperatures, which aids in minimizing the amount of dissolved gas in the liquid.

A moderate increase in the temperature of a liquid brings it closer to its vapor pressure,meaning that vaporous cavitation is more easily achieved. Vaporous cavitation, in whichthe cavitation bubbles are filled with the vapor of the cavitating liquid, is the most effectiveform of cavitation. As the boiling temperature is approached, however, the cavitationintensity is reduced as the liquid starts to boil at the cavitation sites.

Ultrasonic Power

Cavitation intensity is directly related to ultrasonic power at the power levels gener-ally used in ultrasonic cleaning systems. As power is increased substantially above the cav-itation threshold, cavitation intensity levels off and can only be further increased by usingfocusing techniques.

Ultrasonic Frequency

Cavitation intensity is inversely related to ultrasonic frequency. As the ultrasonicfrequency is increased, cavitation intensity is reduced because of the smaller size of the


cavitation bubbles and their resultant, less violent implosion. Higher frequencies are usedto eliminate cavitation erosion on delicate parts.

Ultrasonic frequencies above the traditional 25 and 40 kHz have also been demon-strated more effective at removing submicron-sized particles from silicon wafers andcoated optics. Other applications may also benefit from the use of higher frequencies.

Maximizing Overall Cleaning Effect

Cleaning chemical selection is extremely important to the overall success of the ultra-sonic cleaning process. The selected chemical must be compatible with the base metalbeing cleaned and have the ability to remove the soils that are present. It must also cavitatewell. Most cleaning chemicals can be used satisfactorily with ultrasonics. Some are formu-lated especially for use with ultrasonics. However, the nonfoaming formulations normallyused in spray washing applications should be avoided. Highly wetted formulations arepreferred. Many of the new petroleum-cleaners, as well as petroleum and terpene-basedsemiaqueous cleaners, are compatible with ultrasonics. Use of these formulations mayrequire some special equipment considerations, including increased ultrasonic power, tobe effective.

Temperature was mentioned earlier as being important to achieving maximum cavita-tion. The effectiveness of the cleaning chemical is also related to temperature. Although thecavitation effect is maximized in pure water at a temperature of approximately 160°F, opti-mum cleaning is often seen at higher or lower temperatures because of the effect that

Figure 20

Negative pressurecavitation bubblegrowing

Atmospheric pressurebubble starts tocollapse

Positive pressurebubble continuesto collapse

Maximum pressure


temperature has on the cleaning chemical. As a general rule, each chemical will performbest at its recommended process temperature regardless of the temperature effect on theultrasonics. For example, although the maximum ultrasonic effect is achieved at 160°F,most highly caustic cleaners are used at a temperature of 180 to 190°F because the chemi-cal effect is greatly enhanced by the added temperature. Other cleaners may be found tobreak down and lose their effectiveness at these high temperatures; for example, someshould not be used above 140°F. The best practice is to use a chemical at its maximum rec-ommended temperature, but not exceeding 190°F. (Figure 21).

Degassing of cleaning solutions is extremely important in achieving satisfactory clean-ing results. Fresh solutions or solutions that have cooled must be degassed before proceed-ing with cleaning. Degassing is done after the chemical is added and is accomplished byoperating the ultrasonic energy and raising the solution temperature. The time required fordegassing varies considerably, based on tank capacity and solution temperature, and mayrange from several minutes for a small tank to an hour or more for a large tank. An unheatedtank may require several hours to degas. Degassing is complete when small bubbles of gascannot be seen rising to the surface of the liquid and a pattern of ripples can be seen.

Ultrasonic Power

The ultrasonic power delivered to the cleaning tank must be adequate to cavitate thatentire volume of liquid with the workload in place. Watts per gallon is a unit of measureoften used to measure the level of ultrasonic power in a cleaning tank. As tank volume isincreased, the number of watts per gallon required to achieve the required performance isreduced. Cleaning parts that are very massive or that have a high ratio of surface to massmay require additional ultrasonic power. Excessive power may cause cavitation erosion or“burning” on soft metal parts. If a wide variety of parts is to be cleaned in a single cleaningsystem, an ultrasonic power control is recommended to allow the power to be adjusted asrequired for various cleaning needs (Figure 22).

Part exposure to both the cleaning chemical and ultrasonic energy is important for effec-tive cleaning. Care must be taken to ensure that all areas of the parts being cleaned areflooded with the cleaning liquid. Parts baskets and fixtures must be designed to allow pen-etration of ultrasonic energy and to position the parts to assure that they are exposed to theultrasonic energy. It is often necessary to individually rack parts on a specific orienta-tion or rotate them during the cleaning process to clean internal passages and blind holesthoroughly.

Figure 21 Temperature effect.


CONCLUSION

Properly utilized, ultrasonic energy can contribute significantly to the speed and effec-tiveness of many immersion cleaning and rinsing processes. It is especially beneficial inincreasing the effectiveness of today’s preferred aqueous cleaning chemistries and, in fact,is necessary in many application to achieve the desired level of cleanliness. With ultrason-ics, aqueous chemistries can often give results surpassing those previously achieved usingsolvents. Ultrasonics is not a technology of the future—it is very much a technology oftoday.

Figure 22


CHAPTER 2.3

Ultrasonic Cleaning Mechanism

Sami B. Awad

CONTENTS

IntroductionCavitation Formation MechanismMatching the Frequency to the ProcessTransducersEnhanced TransducersPrecision CleaningUltrasonic Cavitation and Surface CleaningUltrasonic Cleaning EquipmentCleaning ChemistryContaminantsMechanism of CleaningCleaning Chemistry and ParticlesConclusionReferences

INTRODUCTION

Ultrasonic cleaning is used in such diverse applications as automotive components,optics, disk drives, semiconductors, electronics, medical/pharmaceutical products, surfacepreparation for plating and precision coating, aerospace, general metals cleaning, precisionbearings, and a variety of consumer products from jewelry to guns.

To understand the power and utility of ultrasonics, it is important to understand cavi-tation implosion.1 This unique phenomenon occurs when high-energy ultrasonic waves(20 kHz to about 500 kHz, at about 0.3 to 1 W/cm2) travel in a liquid or a solution.Ultrasonic waves interact with the liquid media to generate a highly dynamic agitatedsolution, producing microvapor/vacuum bubbles. The bubbles grow to maximum sizesinversely proportional to the applied ultrasonic frequency and then implode, releasingenergy. The higher the frequency, the smaller the cavitation size and the lower the implo-sion energy.


Figure 1 Scrubbing forces.

CAVITATION FORMATION MECHANISM

The ultrasonic cleaning model (Figure 1) illustrates generation of cavitation throughnucleation, growth, and violent collapse or implosion. Transient cavities (also referred to asvacuum bubbles or vapor voids), ranging from 50 to 150 �m in diameter at 25 kHz, are pro-duced during the half cycles of the sound waves. During the rarefaction phase of the soundwave, the liquid molecules are extended outward against and beyond the liquid naturalphysical elasticity/bonding/attraction forces, generating vacuum nuclei, which continueto grow. A violent collapse occurs during the compression phase. It is believed that thecompression phase is augmented by the enthalpy of the medium, the degree of mobility ofthe molecules, and the hydrostatic pressure of the medium.

Cavitation generates high forces in very brief bursts. Generation time of cavitation is inthe order of microseconds. At 20 kHz, pressure is estimated at approximately 35 to 70 kPa,transient localized temperatures are about 5000°C, with the velocity of microstreamingaround 400 km/h (Figure 2). A number of factors influence the intensity and abundance ofcavitation in a given medium The ultrasonic waveform, frequency, and the power ampli-tude are important. Other influential factors include physical properties of the liquidmedium (viscosity, surface tension, density, and vapor pressure); temperature; and liquidflow (static, dynamic, or laminar); and dissolved gases.

High-intensity ultrasonics can grow cavities to the maximum diameter prior to implo-sion in the course of a single cycle. At 20 kHz the bubble size is roughly 170 �m in diame-ter (see Figure 2). The vacuum bubble size becomes smaller at higher frequencies as afunction of the wavelength. For example at 132 kHz it is estimated to be about half the sizeof cavitations generated at 68 kHz. At 68 kHz, the total time from nucleation to implosionis estimated to be about one third of that at 25 kHz.


Figure 2 Ultrasonic frequency and cavitation size and population.

At higher frequencies, the minimum amount of energy required to produce ultrasoniccavities is higher and must be above the cavitation threshold. In other words, the ultrasonicwaves must have enough pressure amplitude to overcome the natural molecular bondingforces and the natural elasticity of the liquid medium in order to grow the cavities. Forwater at ambient temperature the minimum amount of energy needed to be above thethreshold was found to be about 0.3 and 0.5 W/cm2 of the transducer radiating surface for20 and 40 kHz, respectively.

MATCHING THE FREQUENCY TO THE PROCESS

Selecting the proper frequency for a particular application is critical. Estimates of cav-itation abundance at various ultrasonic frequencies have shown that the number of cavita-tion sites is directly proportional to the ultrasonic frequency. For example, about 60 to 70%more cavitation sites per unit volume of liquid are generated at 68 kHz than at 40 kHz. Theaverage size of cavities is inversely proportional to the ultrasonic frequency. Therefore, onewould expect that at the higher frequency, at a given energy level, the scrubbing intensitywould be milder, particularly on soft and thin or delicate surfaces.

Because a lower number of cavitations of larger size and higher energy are generatedat frequencies of 20 to 35 kHz, systems with lower frequencies are appropriate for cleaninglarge or heavy components. As the frequency increases, denser cavitation with moderateor low energies is formed. Therefore, frequencies of 60 to 80 kHz are recommended for del-icate surfaces; frequencies of 132 and 200 kHz are recommended for cleaning ultradelicateand tiny components. The guidelines hold for both cleaning and rinsing.

TRANSDUCERS

The transducers most commonly used for generating ultrasonic vibrations are piezo-electric, magnetostrictive, electromagnetic, pneumatic, and other mechanical devices. Thepiezoelectric transducer (PZT) is the most widely used technology in cleaning and weld-ing applications. It offers a wide range of frequencies from about 20 kHz to the megasonicrange.


PZT transducers are typically mounted on the bottom and/or sides of the cleaningtanks. The transducers can be mounted in various designs and sizes of sealed stainless steelcontainers or immersed in the cleaning solution/liquid (immersibles). Ultrasonic trans-ducers should be placed on the longer sides and/or on the bottom of the tank, to providemaximum distribution of the sonic energy through the cleaning solution.

A new transducer design developed by Crest Ultrasonics2 provides greater soundenergy transmission with very low acoustic impedance at high frequencies. Benefitsinclude high-quality surface cleanliness and efficient submicron particle removal.

Another recent design, the push–pull transducer rod, is an immersible transducer. Thepush–pull is made of two PZT transducers mounted on the ends of a titanium rod. Thegenerated ultrasonic waves propagate perpendicularly to the resonating surface. The wavesinteract with liquid media to generate cavitation implosions.

ENHANCED TRANSDUCERS

Since its inception about 40 years ago, the conventional PZT transducer assembly hasconsisted of sandwiching a PZT crystal under compression between two metals. A newerdesign2 was recently developed in which one or both metals are replaced with a ceramicmaterial having twice or higher acoustic inductance. One important benefit is that the newtransducer assembly produces sharply defined primary and tertiary resonant frequencies,including new ones not available using the classic design. A second improvement is ahigher transmission coefficient of ultrasonic waves into liquid, estimated at 20 to 30%.

The enhanced transducer design has been shown to improve cleaning efficiency. Astudy at Clarkson University, (New York) by A. Busnaina et al.3 compared one system withthe conventional transducer design with a second with the enhanced transducer. Resultsindicate that, at 68 kHz, efficiency of removal of small particles from wafer substratesincreases from 84 to 93% with the enhanced transducer system. Efficiency was also influ-enced by frequency; higher efficiency of particle removal (�97%) was observed at 132 kHz.

PRECISION CLEANING

Precision or critical cleaning of components or substrates is the complete removal of unde-sirable contaminants to a preset level, without introducing new contaminants in the process.1 Thispreset level is typically the minimum level at which no adverse effects take place in a sub-sequent operation. In attempting to clean, it is critical not to introduce new contaminant(s).For example, in an aqueous cleaning process, it is important to have high-quality rinsewater and a minimum of two rinse steps. Otherwise, new contaminants will be introducedby residual detergent and/or ionics in the rinse water. Recontamination of cleaned partswith outgassed residues produced from packaging or storing materials is another sourceof contamination.1

To meet production and quality demands, choosing the appropriate cleaning chem-istry and process is essential. Rejected parts are the curse of the assembly line and impropercleaning methods are often to blame. Even beyond the factory floor, improper or inade-quate cleaning of a component could directly affect warranty claims.4,5

ULTRASONIC CAVITATION AND SURFACE CLEANING

The energy released from an implosion in close proximity to the surface collides withand fragments or disintegrates the contaminants, allowing the detergent or the cleaning


solvent to displace them at a very fast rate. The implosion also produces dynamic pressurewaves, which carry the fragments away from the surface. The implosion is also accompa-nied by high-speed microstreaming currents of the liquid molecules. The cumulative effectof millions of continuous tiny implosions in a liquid medium is what provides the neces-sary mechanical energy to break physically bonded contaminants, speed up the hydrolysisof chemically bonded ones, and enhance the solublization of ionic contaminants. Thechemical composition of the medium is an important factor in speeding the removal rateof various contaminants.

Cleaning with ultrasonics offers several advantages over other conventional methods.Ultrasonic waves generate and evenly distribute cavitation implosions in a liquid medium.The released energies reach and penetrate crevices, blind holes, and areas that are inacces-sible to other cleaning methods.6 ,7 The removal of contaminants is consistent and uniform,regardless of the complexity and the geometry of the substrate.

ULTRASONIC CLEANING EQUIPMENT

Ultrasonic aqueous batch cleaning equipment consists of at least four steps: ultrasonicwash, a minimum of two ultrasonic separate (or reverse cascading) water rinse tanks, andheated recirculated clean air for drying. The last drying step is not included if the post-cleaning operation includes an aqueous process, as in electroplating or electroless plating.Ultrasonic transducers are bonded to the outside bottom surface, or to the outside of thesidewalls, or they are provided as immersibles and placed inside the tanks. Immersibles areusually the preferred method for large tanks. Two types of immersibles are commerciallyavailable in various sizes and frequencies. The first is the traditional sealed metal box con-taining a multitransducer system. The second is the cylindrical push–pull immersible,powered by two main transducers, one at each end.

Prior to selecting equipment, it is imperative that an effective cleaning process bedeveloped. Then the number and size of the stations are determined based onrequired yield, total process time, and space limitation.

Typical tank size ranges from 10 to 2500 liters based on the size of the parts, productionthroughput, and the required drying time. The tanks are typically constructed of corrosionresistant stainless steel or electropolished stainless steel. Titanium nitride or a similar coat-ing such as hard chrome or zirconium is used to extend the lifetime of the radiating surfacein the tanks or the immersible transducers.

Advantages of automation are numerous, including consistency, achieving desiredthroughput, and full control of process parameters.8 Automation includes a computerizedtransport system able to run different processes for various parts simultaneously as well asdata monitoring and acquisition.

The entire cleaning system can be enclosed to provide a clean room environmentmeeting Class 10,000 down to Class 100 clean room specifications. Process control andmonitoring equipment consists of flow controls, chemical feed pumps, in-line particlecounter, TOC (total organic carbon) measurement, pH, turbidity, conductivity, refractiveindex, etc.

The power requirement for most ultrasonic cleaning applications using PZTs,expressed in terms of electrical-input wattage to the transducers, ranges from 50 to 100W/gal of cleaning fluid, or 2.8 to 3.6 W/in.2 of transducer radiating surface.


CLEANING CHEMISTRY

It is important to realize that the use of ultrasonics does not eliminate the need for the propercleaning chemicals and implementing and maintaining the proper process parameters.9

Cleaning fluids are selected on the basis of the chemical and physical nature of the con-taminants, substrate material(s), environmental considerations, and cleanliness specifica-tions. Aqueous and solvent cleaning have advantages and disadvantages. Withappropriate additives, aqueous cleaning is universal and achieves better cleaning results.

Cleaning with ultrasonics using only plain water is workable, but only for short time.The question then is how long a system will work before cleaning action stops. The chem-ical composition of the cleaning medium is a critical factor in achieving the completeremoval of various contaminants, without inflicting any damage to the components. Infact, cleaning is more complex than just extracting the contaminants from the componentand moving them away from the surface. Soil loading and encapsulation/dispersion ofcontaminants are determining factors in the effective lifetime of the cleaning medium andtherefore in effective cleaning of the part.

Requirements for the selected chemistry are many and no one chemistry is universal.For example, solvents are appropriate for removing organic contaminants but not forremoving inorganic salts.11 The solvent must cavitate well with ultrasonics and be compat-ible with components to be cleaned. Other properties such as wettability, stability, soil load-ing, oil separation, effectiveness, dispersion or encapsulation of solid residues, ability torinse readily, and disposal considerations must be all addressed in choosing the appropri-ate chemistry. With so many factors to consider, an expert in the field may be better able tomake this decision.

The role of additives in aqueous chemistries is multifaceted: to displace oils, to solubi-lize or emulsify organic contaminants, to encapsulate particles, and to disperse and pre-vent redeposition of contaminants. With appropriately formulated aqueous cleaningchemistries, ferrous and nonferrous metals (for example, aluminum, copper, brass, steel,and stainless steel) can be cleaned in the same bath without interaction.

Special additives are used to assist in the process of breaking chemical bonding,removal of oxides, preventing corrosion, enhancing the physical properties of the surfac-tants, and enhancing the surface finish. Ultrasonic rinsing with deionized water or reverseosmosis (RO) water is important to achieve spot-free surfaces. A minimum of two rinsesteps is recommended. Drying and protection of steel components are valid concerns.However, the current available technologies offer effective ways to alleviate these concerns.

CONTAMINANTS

Three general classes of common contaminants are organic, inorganic, and particulatematter (organic, inorganic, or a mixture). Contaminants of any class may be water solubleor water insoluble.

Most organic contaminants such as oils, greases, waxes, polymers, paints, print, adhe-sives, or coatings are hydrophobic. Organic contaminants can be classified into three gen-eral classes: long-chain, medium-chain, and short-chain molecules. The physical andchemical characteristics are related to their structure and geometry.

Insoluble particulate contaminants can be divided into two groups, hydrophilic andhydrophobic. Examples of the first group include water-wettable particles, such as metals,metal oxides, minerals, and inorganic dusts. Examples of the second include non-water-wettable particles such as plastics, smoke and carbon, graphite dust, and organic chemical


Figure 3 Liquid soil removal.

dusts. Similarly, substrate surfaces can be divided into hydrophilic and hydrophobicgroups.

With few exceptions, inorganic materials or salts are insoluble in water-immiscible sol-vents. However, water-insoluble inorganics, such as polishing compounds made of oxidesof aluminum, cerium, or zirconium, require a more elaborate cleaning process.

MECHANISM OF CLEANING

Two main steps take place in surface cleaning. The first is contaminant removal; thesecond is prevention of re-adherence. The removal of various contaminants involves dif-ferent mechanisms, based on the nature and/or the class of the contaminant.

Organic contaminants are removed by two primary mechanisms. The first is sol-ublization in an organic solvent. The second is by displacement with a surfactant film fol-lowed by encapsulation and dispersion.

The mechanism of removal of organic contaminants by detergent involves wettingboth contaminant and substrate. According to Young’s equation, wetting increases the con-tact angle (�) between the contaminant and the surface, thus decreasing the surface areawetted with the hydrophobe, and reducing the scrubbing energy needed for removal(Figure 3).

COS � � ��SB

��OB

�SO�

Aqueous additives contain one or more surfactants. Surfactants are long-chain organicmolecules with polar and nonpolar sections. Surfactants may be ionic or nonionic. Whendiluted with water, surfactants form aggregates called micelles at a level above the criticalmicelle concentration (CMC). The micelles, composed of aggregates of hydrophilic andhydrophobic moieties, act as a solvent encapsulating the contaminants, thus preventingredeposition.

Ultrasonic cavitation plays an important role in removal of hydrophobic contaminants.The shock wave (and the microstreaming currents) greatly speed up the breaking ofadhered contaminants, enhancing displacement with the detergent film. The contaminants


are then encapsulated in the micellic aggregates, thus preventing redeposition. The netresult is that ultrasonic cavitation accelerates displacement of contaminants from the sur-face of the substrate and facilitates their dispersion.

CLEANING CHEMISTRY AND PARTICLES

Theoretically, adhesion forces, including van der Waals, electrical double layer, capil-lary, and electrostatic, are directly proportional to the size of the particle. One would expectthe energy of detachment to decrease with the size of particles. However, smaller particlesare always more difficult to detach, mainly because small particles tend to get trapped inthe valleys of a rough surface.

According to the Gibbs adsorption equation, the mechanism of particle removalinvolves shifting the free energy of detachment to slightly above or less than zero.Surfactants play a very important role in decreasing the adsorption at particle and sub-strate interfaces.

Ultrasonic cavitation provides the agitation energy for detachment (i.e., the removalforce). At 40 kHz, the detachment or removal efficiency of 1-�m particles is 88%. Efficiencyincreases to 95% at high frequencies (60 to 70 kHz), equalling the efficiency of megasonicsof approximately 850 kHz. This is expected in light of the fact that cavitation size is smallerat higher frequencies and can reach deeper into the surface valleys. One would then antic-ipate that a combination of high-frequency ultrasonics at 65 to 70 kHz with appropriatechemistry would further improve efficiency of particle removal.

Inhibiting redeposition of contaminants involves formation of a barrier between thesuspended contaminant and the cleaned surface. In solvent cleaning, a film of solventadsorbed to both substrate and contaminant forms the barrier. In aqueous cleaning, aneffective surfactant system encapsulates contaminants in the micellic structure as depictedin Figure 4. Redeposition of the encapsulated contaminants (soils) is prevented via stearichindrance (nonionic surfactants) or via electrical repulsion (anionic surfactants).

Depending on the surfactant system, encapsulation can be permanent or transient.Transient encapsulation is preferable to emulsification, as it allows better filtration and/orphase separation of contaminants. Allowing soil loading to reach the saturation pointsignificantly decreases cleaning agent efficiency; cleaning action may cease. To ensure

Figure 4 Antiredeposition.


consistent cleaning, the dispersed contaminants must be removed by continuous filtrationor separation of contaminants, and the recommended concentration of the cleaning chem-ical must be maintained.

The physical properties of the substrate, including surface finish, are important factorsin submicron particle removal.13,14 For example, a silicon wafer surface differs from that ofan aluminum disk, in their physics, topography, and finish. The inherent static charges ofplastics are another challenge when dealing with submicron particles.

CONCLUSION

Cleaning with the assistance of ultrasonic cavitation has numerous advantages, mostimportantly consistency in results.

Advantages:• Efficient cleaning in recessed areas and blind holes• Capability of cleaning assemblies or devices• Removal of micro- and submicrocontaminants• The proper chemistry → exceptional and consistent cleaning• Shorter process time• Full automation and controls, batch and continuous processes

For best cleaning results, selection of the ultrasonic frequency or the cleaning medium (sol-vent or aqueous) for an application must be precise and specific.

REFERENCES

1. S.B. Awad, Ultrasonic cavitations and precision cleaning, Precision Cleaning, Nov. 1996, p. 12.2. J.M. Goodson, U.S. Patent 05,748,566.3. A. Busnaina et al., Microcontamination Research Lab, Clarkson University, Potsdam, NY, 1998,

1999, results to be published elsewhere.4. H.A. Bhatt, How now, Parts Cleaning, May 1998, p. 17.5. J.B. Durkee, The Parts Cleaning Handbook, Gardner Pub. Inc., Cinncinati, OH, 1994.6. M. O’Donoghue, The ultrasonic cleaning process, Microcontamination, 2 (5), 1984.7. F.J. Fuchs, Ultrasonic cleaning principles for parts cleaning potential, Parts Cleaning Mag.,

December 1997, p. 14.8. J. Harmon, Ultrasonic applications in the life sciences, A2C2 Mag., March 1999, p. 7.9. S.B. Awad, Ultrasonic cleaning of medical and pharmaceutical devices and equipment, A2C2

Mag., Feburary 2000.10. S.S. Seelig, The chemical aspects of cleaning, Precision Cleaning, 1995, p. 33.11. B. Kanegsberg, Aqueous cleaning for high-value processes, A2C2 Mag., September 1999, p. 25.12. Figures 1, 2, and 5 were reproduced from Precision Cleaning Mag., Witter Publishing Co., Inc.,

1966.13. S.B. Awad, Ultrasonic aqueous cleaning and particle removal of disk drive components, Datatech,

1999, p. 59.14. K.L. Mittal, Surface contamination concepts and concerns, Precision Cleaning, 3 (1), 17, 1995.


CHAPTER 2.4

Higher-Frequency andMultiple-Frequency Ultrasonic Systems

Michael Pedzy

CONTENTS

IntroductionUltraprobe and High Frequency SystemsDevelopment of a Multiple-Frequency Ultrasonic SystemHurdles to OvercomeUser ExperiencesConclusion

INTRODUCTION

Multiple-frequency ultrasonics systems incorporate the use of more than one opera-tional frequency in the same cleaning bath. They provide the cleaning characteristics ofmore than one frequency, allowing a significantly broader range of particle sizes and con-taminant types to be addressed than single-frequency ultrasonic systems. Thus, multiple-frequency ultrasonics systems represent a significant development in the field of ultrasoniccleaning technology.

Although these systems were developed for the requirements of high-technologyindustries such as computer hard disk, semiconductor, and optical manufacturing, multi-ple-frequency systems are now also finding utility in the heavy-duty manufacturing. Forexample, the technique has found application in manufacture of ball screw mechanisms,ball bearings, and solenoid valve bodies.

Since more than one frequency is present in the liquid medium, the part being cleanedis exposed to the cleaning characteristics of all frequencies included in the system. Eachultrasonic frequency has its own set of characteristics, which are produced by the physicaldevelopment of cavitational energy, the means by which all ultrasonic systems clean. Useof any particular frequency has advantages and disadvantages. However, by combiningmore than one operating frequency, the negative characteristics of both frequencies areoften reduced or eliminated, thereby drastically improving cleaning performance.


The development of multiple-frequency ultrasonic systems actually began during thetime which will be termed The High Frequency Revolution. Prior to 1992, many consideredthat one of the major negative side effects of all ultrasonic cleaning systems, standing waves,had been eliminated with the development of 40-kHz ultrasonics, and piezoelectric trans-ducer elements capable of operating at these frequencies. Standing waves are produced bythe ultrafast compression and expansion cycles produced by the ultrasonic transducerstypically mounted on the cleaning tank bottom. This causes a cleaning action that is dis-tributed as thin bands oriented perpendicular to the stroke direction of the transducers. Inareas where standing waves occur, the bands do not move, and only a small percentage ofcavitation is produced between these bands. The end result is a cleaning pattern of hotspots, where high-intensity cavitation is produced, and dead spots, where little energy ispresent. The consequence of using systems with standing waves is inefficient cleaning;some areas on the parts are cleaned, others are not. Smaller holes or detailed part areas maybe missed, or receive only a small amount of cavitational activity. This is one of the mostcommon complaints about single-frequency ultrasonic cleaning systems in the lower-fre-quency ranges.

ULTRAPROBE AND HIGH FREQUENCY SYSTEMS

The ultraprobe, a patented device that can be used to indicate the presence and qual-ity of cavitation, was first demonstrated in September 1992 at the InternationalManufacturing Technology Show in Chicago, Illinois. Edward Pedzy developed theUltraprobe as a means of visualizing frequency and cavitation and of differentiating higherfrequency (80-kHz) systems from those with lower frequency to components manufactur-ers and other end users. As with many significant developments, the Ultraprobe was dis-covered accidentally. The vibration created by an ultrasonic tank caused a container filledwith a particular metal powder to fall into the bath. The particles of powder immediatelyseparated, indicting a strange bandlike pattern. Noticing the faint bands of patterns in thetank, Pedzy filled a quartz test tube with the fluid, and submerged the tip into an ultrasonicbath; the bands were clearly visible. Within 2 h, the first Ultraprobe was produced.

The probe consists of a 24-in. quartz test tube, filled with a proprietary compound com-posed of ultrafine metallic particles, and an opaque carrier fluid. When the tip of the instru-ment is submerged into an ultrasonic cleaning bath, the alternating compression andexpansion pulses cause the metallic particles to suspend in areas of cavitational activity,creating a visual “picture” of the cleaning action that parts will receive in the cleaning tank.

Tests with the Ultraprobe indicate that the sweep frequency circuit does not eliminatestanding waves in many systems. Tests also indicate that the cleaning action produced byelevated-frequency systems was more evenly distributed within the fluid, with less deadarea, and a greater number of standing waves. The 80-kHz system produces standingwaves only �

14� in. apart, while the 40-kHz sweep frequency system produces standing waves

�12� in. apart. Over the next 2 years, many major ultrasonic cleaning system manufacturersoffered cleaning systems with frequencies above 40-kHz. Industrial ultrasonic cleaningsystems are now available with operating frequencies of 200-kHz and higher. One mightconsider that ultrasonic cleaning technology has seen more improvements over the past 8years than at any other time in the history of the cleaning process.

High-technology industries began investigations of higher-frequency ultrasonics,because the cleaning action produced by the elevated frequencies allowed cleaning of theirsensitive components without damage. Some of the early studies produced by these high-technology companies indicated that elevated-frequency systems not only produced a


more evenly distributed cleaning pattern, but they also clearly had greater overall particleremoval counts. Many high technology manufacturers utilize a particle removal test to pro-vide cleanliness data. Essentially, parts are artificially contaminated with particles ofknown size to represent the expected contaminant. The parts are then cleaned in ultrasonicbaths with different operating frequencies, and a liquid particle counter is used to tally thenumber of particles in the liquid bath. The results indicate which system removes moreparticles in general, as well as providing information on how well each frequencyaddressed particles of a given size.

In the applications tested, high-frequency systems removed substantially greater num-bers of particles than did lower-frequency systems. This might be expected given thenature of the cleaning action produced by high-frequency systems. Most of the contami-nant removed during the ultrasonic cleaning process is not actually removed by the implo-sion of cavities themselves, but rather by the blast area produced by the imploding cavity.The actual implosions strike the part, causing the liquid jet to spread out over the surfaceof the part, thereby removing neighboring contamination. Since higher-frequency systemsproduce significantly greater numbers of imploding cavities, these systems would beexpected to remove a greater amount of contamination more rapidly than their low-fre-quency counterparts.

In addition, each frequency tends to remove particles within a particular size range. The40-kHz system tested demonstrated a tendency to remove particles larger than 0.7 �m insize in greater number than the 80-kHz system, while the 80-kHz system demonstrated atendency to remove particles 0.2 �m and smaller in greater number than the 40-kHz system.

Further elevated frequency ultrasonic systems tend to produce less cavitational erosion,both on parts being cleaned and on the transducer radiating diaphragm, the surface to whichthe transducers are mounted. Cavitational erosion is damage caused by the cavitationalaction produced by an ultrasonic cleaning system. The scrubbing action produced by theultrasonic system has the potential to erode the surface of the part itself. The radiatingdiaphragm is continuously eroding in all ultrasonic cleaning systems, which releasesmicroscopic particles of stainless steel into the bath, particles that could potentially resultin part failure or system failure if left on the surface of a computer hard disk or other highlysensitive hardware. Higher-frequency systems, however, produce a much more evenly dis-tributed cleaning pattern, with less energy produced at standing wave locations.

An additional advantage of higher-frequency systems is increased efficacy of cleaningin small spaces and with very complex geometries. This is because high-frequency systemshave a greater number of standing waves and have shorter compression/expansion cycles,producing cavities that are much smaller in size.

DEVELOPMENT OF A MULTIPLE-FREQUENCY ULTRASONIC SYSTEM

Because ultrasonic cleaning systems with different operating frequencies are expectedto produce vastly different cleaning characteristics, development of a multiple-frequencyultrasonic system was undertaken. It was hypothesized that an ultrasonic system capableof producing more than one ultrasonic frequency would result in more rapid and effectivecleaning. It was further expected that by combining different frequencies, some of the neg-ative effects of each frequency could be eliminated. A system with both low- and high-fre-quency transducers would not only remove heavy contamination, but would be capable ofcleaning ultrafine detail.

Designing a multiple-frequency ultrasonic system has involved overcoming a num-ber of hurdles. At the time of original conception, it was thought impossible to mount


transducers of differing frequencies onto the same radiating diaphragm without destroy-ing ultrasonic components, or eliminating some of the ultrasonic energy by simple cancel-lation of wave energy. To overcome this potential problem, the first multiple-frequencyultrasonic systems devised included ultrasonic submersible transducer packs, each with adifferent operating frequency. To avoid transducer damage, transducers of differing fre-quencies are isolated from one another.

Testing of the newly developed systems was performed by manufacturers in the criti-cal coating and plating industries and compared multiple-frequency with single-frequencysystems. Their test results indicate that cleaning with the multiple-frequency systemsresulted in significantly fewer rejected parts and significantly more consistent results.

HURDLES TO OVERCOME

Tests of the early multiple-frequency systems did expose shortcomings in the originalsystem design. Since each frequency was being emitted from different sides of the cleaningtank, one side of the part would be cleaned by only one ultrasonic frequency, while theother side would be exposed to a different frequency. When cleaning a large batch of denseand heavy parts, shadowing of the ultrasonic energy would occur, and one side of the bas-ket received the effects of one frequency, the other side would receive the effects of theother.

The best multiple-frequency ultrasonic system would be one that could emit allincluded frequencies from the same radiating surface. If the system design were possible,the benefits would be significant. Each radiating surface would emit all frequencies, pro-viding all parts with an evenly distributed exposure to all included ultrasonic frequencies.Smaller tanks could also be equipped with multiple-frequency ultrasonic systems, increas-ing utility to the disk and semiconductor industries.

Development of such a system was another story. Several major design hurdles imme-diately became apparent. First, transducers are typically mounted to radiating surfaceswith differing thickness. The lower the frequency, the thicker the diaphragm. In fact, some20-kHz systems utilize diaphragms up to �

38� in. thick to overcome the effects of cavitational

erosion. However, higher-frequency systems, having transducers with extremely shortstroke lengths during expansion, must be mounted to thinner diaphragms, since thickermaterials would completely prevent the transducer from oscillating effectively.

Interference between neighboring transducers was also a hurdle. Since each transduceroperates at a different frequency, the effects of one transducer would cancel the effects ofneighboring transducers at specific moments. The intense interaction would potentiallyloosen the bonds used to attach the transducers to the radiating surface, or crack the trans-ducer itself.

In one early system designed to address these problems, transducers of identical fre-quency are mounted to a tank, and connected to a generator capable of producing signalswith more than one frequency. During activation, all transducers are activated with thesame frequency of ultrasonic energy for a given period of time, after which all transducersare activated with a second frequency. This cycling of frequencies does indeed produce thecleaning effects of all included frequencies.

Although these systems produce multiple-frequency cleaning effects, this author, asthe manufacturer and patent holder of a different system, considers that the above systemshave certain shortcomings. Each transducer has a very specific optimum operating fre-quency, based on the size of the transducer itself. Even the smallest difference in transducersize will affect the optimum operational frequency. In addition, the frequency range of


effective operation is very narrow. For example, if an 80-kHz transducer is activated withultrasonic energy at 40-kHz, ultrasonic efficiency is reduced to less than 40%. The fartherthe input signal is from the optimum operating frequency of the transducer, the greater thepower loss. Only one of the frequencies will operate at optimum efficiency.

To overcome this loss of power, manufacturers have included devices that increase thepower to the ultrasonic transducer to make up for the loss of energy that is produced whentransducers are powered by signals far from their optimum operating frequency.

Another side effect produced by this system is limited individual frequency exposure.Each frequency operates for only a fraction of the total cleaning time. For example, if twofrequencies are present, and a 10-min cycle is used, each frequency is activated for only 5min. Although a drastic improvement over single-frequency systems, transducers and gen-erators used in these systems are exposed to electrical extremes to overcome efficiencylosses inherent in this design.

To overcome the shortcomings of the modified-frequency design mentioned earlier,this author felt it would be necessary to manufacture a system that would incorporatetransducers operating at different frequencies to provide superior performance. The twomajor challenges were to eliminate destructive transducer interaction and to resolve theproblems associated with transducer diaphragm thickness.

Overcoming the thickness issue was rather easy. The key was to develop transducermounts with significantly less cavitational erosion potential, allowing them to be mountedto thinner diaphragms next to high-frequency transducers. The other possibility was togrind away material under the higher-frequency transducers, thereby allowing each trans-ducer to emit its energy through a diaphragm with a thickness most efficient for that par-ticular operating frequency. The final patented design includes both developments.

Eliminating the destructive interaction between transducers with different frequencieswas more difficult, and more important to overcome, than the diaphragm thickness issue,because of the possibility of permanent damage to components. However, after more than2 years of experimentation, a patented, proprietary technology was developed that allowssimultaneous operation of transducers with a variety of operating frequencies withoutdestructive interference, even when mounted directly next to one another. Today, trans-ducers of differing frequencies are staggered between one another, allowing the diaphragmto produce multiple-frequency output from any location on the radiating surface.

To improve the design even further, ultrasonic frequencies were selected to take advan-tage of harmonic frequencies. Ultrasonic systems, regardless of frequency, produce notonly sound waves of the primary operating frequency of the transducers, but harmonicfrequencies, which are multiples of the original operating frequency. The first harmonic isby far the most powerful, and produces energy sufficiently powerful to produce cavitation.Transducer combinations can be selected to prevent overlapping of harmonics. For exam-ple, originally, a common multiple-frequency combination for industrial use once was 40and 80-kHz. However, since 40-kHz has its first harmonic at 80-kHz, better cleaning actionwould be produced if the 80-kHz were changed to a slightly different frequency to preventthis overlapping of frequencies, thus increasing the number of frequencies producing cav-itation in the cleaning bath.

USER EXPERIENCES

The high-technology sector was largely responsible for the development of multiple-frequency ultrasonic systems, and represents the largest share of users of multiple-fre-quency systems. The disk industry utilizes these systems to remove diamond slurry, which


is used to texture the disks prior to a successive plating/coating operation. Should a singleparticle remain on the surface, the disk fails, and becomes essentially waste material. Infact, if one of the multiple-frequency ultrasonic systems used to remove this slurry fails toremove every single particle from the surface of at least 10,000 disks, the system is takenoff-line and serviced. Hundreds of these systems are in use 24 h a day, with long-term reli-ability that equals or exceeds that of single-frequency systems. As a result of their installa-tion, the disk yields at these locations have improved 2 to 3%, saving the corporationsmillions of dollars each year.

The systems are also beginning to find utility in industrial processes, particularly forapplications with high levels of contamination, requiring low-frequency ultrasonics, andrecessed areas, more efficiently cleaned with elevated frequencies. Other applicationsrequiring the removal of thick layers of contamination are also better addressed by multi-ple-frequency ultrasonic systems. The multiple-frequency system includes the lower-fre-quency components to remove the bulk of the contaminant, while also includingtransducers with slightly higher operating frequency to better address contaminants thatare hidden in blind holes or missed by the standing waves of the lower-frequency system.

Yet another segment that has discovered the benefits of high-frequency ultrasonic sys-tems is the coal-processing industry. Ultrasonics at the 80-kHz frequency have been usedfor years to process coal slurry to remove hazardous by-products from this material,thereby allowing the slurry to be disposed of easily. After experimenting with multiple-fre-quency ultrasonics at 40/90-kHz combinations, a significant improvement in waste extrac-tion was produced. The slurry is passed over a series of transducers at both frequencies, atan unbelievable flow rate of over 400 gal/min, through a reactor with only 12 gal of liquidvolume. The multiple-frequency system had the ability to remove contaminant that wasmissed by the 80-kHz transducers, producing an effluent that was significantly cleaner.

CONCLUSION

Development of the Ultraprobe has enabled visualization of differences in frequencyand of potential problems in ultrasonic cleaning, such as dead zones.

Development of multiple-frequency ultrasonic systems has been a major challenge, buthas yielded many realized and potential benefits to the end user. Although in many casessingle-frequency ultrasonics are appropriate, in other cases ultrasonic cleaning processescan be greatly improved by the use of multiple-frequency ultrasonics.


CHAPTER 2.5

Megasonic Cleaning Action

Mark Beck

CONTENTS

IntroductionOverview of Megasonic CleaningMegasonic Cleaning Compared with Ultrasonic CleaningApplication of Megasonic and Ultrasonics

Discussion of Underlying PhysicsProperties of Piezoelectric TransducersParticle Attraction and Removal Forces

Principal Mechanisms of Megasonic CleaningAcoustic CavitationAcoustic StreamingSignificance of the Boundary Layer

Cleaning Chemistry and Other FactorsCleaning ChemistriesOther Cleaning Factors

Design Considerations for Megasonic SystemsConclusionReferences

INTRODUCTION

Megasonics has been a widely accepted cleaning method for contamination-sensitiveproducts for nearly 20 years. Megasonics was initially developed in the early 1940s as aresult of U.S. Navy research into advanced sonar instrumentation for antisubmarine war-fare. In the late 1970s, RCA adapted this technology for wafer cleaning, and by 1982 com-mercial megasonic cleaning equipment was being delivered to the semiconductor industry.

More recently, advances have been made in this acoustic cleaning technology, througha better understanding of high-frequency acoustic streaming, and controlled acoustic cav-itation, the megasonic cleaning technique has proved effective for removing submicronparticles from silicon and other substrates without damage. As a result, growing numbersof manufacturers in the integrated circuit, hard drive, raw silicon, mask, flat panel


Figure 1 Megasonic cleaning uses the piezoelectric effect to produce acoustic waves that movethrough the cleaning liquid.

display, and other industries have been turning to megasonic cleaning to help meet strin-gent cleaning requirements. Megasonic cleaning is now increasingly accepted by industryas a cost-effective, efficient, and safe method for the removal of nanoscale particles fromcontamination-sensitive products.

Overview of Megasonic Cleaning

Megasonics utilizes the piezoelectric effect at high frequencies to generate controlledacoustic waves in a liquid bath to enable removal of submicron particles from substrates.

In megasonic cleaning (Figure 1), a piezoelectric crystal array transducer convertsalternating electrical energy directly to mechanical energy using the piezoelectric effect, inwhich certain materials change dimension when an electrical charge is applied. A ceramicpiezoelectric crystal is excited by high-frequency AC voltage, between 500 and 2000 kHz,causing the ceramic material to change dimension rapidly, or vibrate. These vibrations aretransmitted by the resonant masses of the transducer, and directed into the liquid througha resonating plate, producing acoustic waves in the cleaning fluid. Acoustic cavitation, pro-duced by pressure variations in the sound waves moving through the liquid, and theeffects of acoustic streaming cause particles to be removed from the material being cleaned.


Megasonic Cleaning Compared with Ultrasonic Cleaning

There are two types of acoustic cavitation: transient cavitation and stable, or controlled,cavitation. Ultrasonic cleaning frequencies, between 20 and 350 kHz, produce transientacoustic cavitation. Transient cavitation is characterized by transient bubbles that exist foronly a few acoustic cycles, after which they collapse violently, producing very high localtemperature and pressure. Transient acoustic cavitation generates shock waves that arepowerful enough to erode solid surfaces nearby,1 and to damage some substrate surfaces.

Megasonic cleaning operates at much higher frequencies, 500 to 2000 kHz, which pro-duce controlled acoustic cavitation. Controlled cavitation is characterized by stable bub-bles that are relatively permanent, can exist for many acoustic cycles, and do not causedamage to substrate surfaces,2 because the cavitation radii are much smaller at higher fre-quencies and have less energy upon collapse. Thus, megasonic-controlled acoustic cavita-tion is best suited for sensitive substrate surfaces that cannot withstand the heat andpressure of transient cavitation.

In addition, ultrasonics simultaneously cleans all surfaces of a submerged object. Thismeans that ultrasonic cleaning subjects all areas of the substrate, including areas that maynot need to be cleaned to the previously described effects of transient acoustic cavitation.Megasonics accomplishes line-of-sight cleaning; it affects only those surfaces of the objectthat are in the path of the acoustic wave.

Application of Megasonics and Ultrasonics

The mechanical effects of both ultrasonic and megasonic cleaning can be helpful inspeeding particle dissolution and in displacing particles. Both ultrasonics and megasonicshave also been demonstrated to speed or enhance the effect of many chemical reactions. Inaddition, residual cleaning chemicals can be removed quickly and completely by eitherultrasonic or megasonic rinsing. However, there are applications for which megasoniccleaning clearly would be favored.

The effects of ultrasonics and megasonics on substrate surfaces and particle removalresults provide the basis for identifying the best applications for each process. Ultrasoniccleaning is most appropriate for strong, heat-tolerant substrate materials requiring multi-surface cleaning. Ultrasonics is also well suited for the removal and/or dissolution of largeparticles from chemically tolerant substrates.

Megasonics is most appropriate for heat- or chemical-sensitive substrates that cannotwithstand the heat and pressure of transient cavitation and for applications requiring line-of-sight-dependent cleaning. Parts that cannot be cleaned with ultrasonics, because theyare sensitive to the frequency or transient cavitation effects can often be cleaned withmegasonics. Megasonics cleaning is also the application of choice for the removal and/ordissolution of small particles (less than 0.3-�m, Figure 2). For example, this cleaning tech-nique has been proved effective for removing 0.15-�m particles from silicon wafers andother cavitation-sensitive products, without causing substrate damage.

Table 1 summarizes the relative strengths of megasonic and ultrasonic cleaning.

Positive Environmental Effects of Megasonics

The use of megasonic cleaning yields several positive environmental results. The highpressure and temperatures produced by ultrasonic cleaning result in the evaporation oflarge volumes of both chemicals and ultrapure water. This has two negative effects. First,


Figure 2 Particle size vs. frequency for megasonic and ultrasonic cleaning.

Table 1 Strengths of Megasonics and Ultrasonics

Applications for Megasonic Cleaning Applications for Ultrasonic Cleaning

Cavitation-sensitive substrates Strong substratesSmall particle removal/dissolution (�0.3 �m) Larger particle removal/dissolutionChemically sensitive substrates Chemically tolerant substratesLine-of-sight-dependent cleaning Multisurface cleaningHeat-sensitive material Heat-tolerant material

the chemical compositions of cleaning solutions cannot be maintained at constant levels.Second, large amounts of chemical vapors are released, increasing the loads on clean-airexhaust systems.

The lower pressures and temperatures produced by megasonic cleaning enableprocesses that drastically reduce both chemical vapor evaporation and the load on airexhaust and replacement systems.

In addition to their environmental benefits, megasonic cleaning methods optimize theuse of cleaning fluids and reduce the costs associated with the acquisition and disposal oftoxic substances.

DISCUSSION OF UNDERLYING PHYSICS

Properties of Piezoelectric Transducers

A basic characteristic of the piezoelectric crystal is that when a sine wave is applied toit, through the application of AC voltage, it expands. For the purpose of megasonic clean-ing, the molecules in the piezoelectric crystal have been aligned, or poled, in the thicknessextension mode. Upon application of the sine wave, the first expansion takes place to theside of the crystal, a second expansion takes place to the end of the crystal, and the thirdexpansion takes place in the thickness of the crystal (Figure 3).

The frequency at which this third expansion, which is the first thickness expansion,takes place is known as the fundamental frequency. The fundamental frequency occurs atapproximately 1000 kHz (Figure 4), with harmonic frequencies at 3 and 5 MHz.

1655_CH02.5I 11/21/00 10:11 AM Page 236


Figure 4 Piezoelectric transducer characteristic of impedance as a function of frequency.Impedance changes near fundamental frequencies.

Figure 3 Piezoelectric crystal transducer rapidly changes dimensions or vibrates with the applica-tion of high-frequency AC voltage.

The most efficient transfer of the energy generated by crystal expansion would occurfrom direct contact between the piezoelectric crystal and the liquid bath. However, clean-ing solutions can damage the piezoelectric crystal, and the piezoelectric crystal is not pureand can add impurities to the cleaning solution. To prevent this, a resonator is adhered tothe top of the crystal, between the crystal and the fluid.

Ideally, the resonator should have no effect on the energy being transferred, thatis, there would be no energy loss or frequency distortion. The velocity of sound in theresonator is an important factor in approaching this ideal. Based on acoustic KLM


transmission theory, the resonator should be designed to be half the wavelength thicknessand should operate optimally at the fundamental frequency.

� � �Vf

L�

where� � wavelength of sound in resonator

VL � velocity of sound in the resonator, mm/sf � frequency of sound, MHz

Particle Attraction and Removal Forces

Megasonics cleaning is able to overcome the attraction forces that hold very small par-ticles to a surface. Particle adhesion force is a function of the type of medium surroundingthe particle and the surface. In general, it is weaker in liquid media than in gas media,3 andit increases linearly with an increase in particle diameter1 (Figure 5).

Although adhesion force is lower at smaller particle diameters, small particles are moredifficult to remove. The weight of the particle decreases as a function of the diametercubed, and for small particles the adhesion force can easily exceed the gravitational forceby a factor of 103 or more.1,3 In addition, van der Waals attractive forces (Fvw) vary depend-ing upon the composition of the particle. They are about ten times larger for silicon parti-cles than for polystyrene latex (PSL) particles, for example, indicating that some particlesmay require much greater removal forces than others.1

PRINCIPAL MECHANISMS OF MEGASONIC CLEANING

An understanding of exactly how particles are removed when megasonic cleaningtechniques are used has been the subject of increased investigation during the past decade.

Figure 5 van der Waals forces vs. particle size.


To date, researchers have not been able to explain precisely why megasonics works, andthere is disagreement on whether controlled acoustic cavitation or acoustic streaming is themore effective mechanism in removing particles. The effect of various parameters on par-ticle removal can be determined, but whether removal is the result of acoustic cavitation oracoustic streaming, or both, is not always clear.2

However, it is accepted that controlled acoustic cavitation, acoustic streaming, andreduction of the boundary layer are the principal particle removal mechanisms in mega-sonic cleaning.

Acoustic Cavitation

Acoustic cavitation is the generation and action of cavities, or bubbles, in a liquid.Acoustic waves moving through a liquid produce variations in the liquid pressure. Whenthe liquid pressure drops momentarily below the vapor pressure during the low-pressureportion of the acoustic wave, small evacuated areas, or cavities, are formed that quicklybecome filled with gas (a foreign contaminant such as dissolved oxygen or air) and/orvapor (a gaseous form of the surrounding liquid).4 These tiny bubbles are set in motion bythe acoustic wave. The bubbles may be suspended in the liquid medium, or they maybecome trapped in voids either in the boundary surface of the liquid or in solid particlessuspended in the liquid.

The tiny bubbles can expand and contract in the liquid. Bubble expansion can becaused by reducing the ambient pressure in the liquid, either by static or dynamic means.The bubbles can then become large enough to be seen by the unaided eye. The bubbles maycontain gas or vapor or a mixture of both. If the bubbles contain gas, then their expansioncan be caused by rectified diffusion, pressure reduction, or an increase in temperature.3

Rectified diffusion is the diffusion of dissolved gas from the liquid into the bubble, andvice versa, with the pressure oscillations resulting in a net diffusion into the bubble. This netinward diffusion occurs because the bubble surface area increases during inward diffusionand decreases during outward diffusion; a higher surface area leads to more diffusion.2 Ifthe ambient liquid is not saturated with gas, then rectified diffusion must compete withordinary diffusion from the bubble to the liquid. In that case, the sound pressure amplitudemust exceed a certain value in order for the bubbles to increase significantly in size.2

The pressure oscillations that created the bubbles can also cause them to expand andcontract. If the pressure variation is great enough to reduce the local liquid pressure downto, or below, the vapor pressure in the negative parts of the acoustic cycle moving throughthe liquid, any minute cavities or bubbles that are present will grow larger. If the range ofthe pressure variation is increased to produce zero and then negative pressures locally inthe liquid, then bubble growth is increased. Gas from the liquid diffuses into a bubble dur-ing expansion, and leaves the bubble during contraction.

When the bubble reaches a size that can no longer be sustained by its surface tension,the bubble will expand and then collapse, or implode, which is an important action of thecavitation phenomenon. The bubble action of cavitation has sufficient energy to overcomeparticle adhesion forces and to dislodge particulates attached to substrates in the stream ofbubbles. Essentially, imploding cavitation bubbles generate shock waves that dislodge par-ticles from substrate surfaces. Cavitation breaks down the molecular force by which a par-ticle is held to a surface either by direct impact from bubble implosion or by the fatiguingaction caused by repeated bombardment.3

Cavitation implosion force varies with the size and contents of the bubble. Larger bub-bles are unstable and implode with larger force; smaller bubbles are stable and collapsewith less force. Vapor collapses more quickly, resulting in larger implosion force, whereasgas cushions and slows the collapse, resulting in smaller implosion force.


Cavitation does not occur until a specific threshold is reached.3 The cavitation thresh-old is defined as the minimum pressure amplitude required to induce cavitation.2

A number of methods have been developed for detecting cavitation, including acousticemissions, visual observations, sonoluminescence (SL), and surface erosion. Of these, SL isbelieved to be the most suitable method for characterizing cavitation in a megasonic tank,1

because it is related to the cavitation collapse of bubbles.The intensity and effect of cavitation on materials being cleaned are related to the type

of acoustic cavitation produced. Two types of acoustic cavitation have been identified andstudied: transient cavitation and stable cavitation. Transient acoustic cavitation is pro-duced by ultrasonic cleaning frequencies, between 20 and 350 kHz, which transform low-energy-density sound waves into high-energy-density collapsing bubbles. In transientcavitation, the mostly vapor-filled bubbles exist for only a few acoustic cycles, followed bya rapid and violent collapse. This type of cavitation is likely to produce violent events inthe acoustic field, such as radiation of light (SL) and shock waves. The level of violence pro-duced is believed to be dependent on the maximum size of transient bubbles, which isrelated to the acoustic frequency.1 Because transient cavitation concentrates energy intovery small volumes and tends to produce very high local temperatures and pressure, it cancause surface erosion and damage to sensitive substrates.

Bubble size decreases as acoustic frequency increases, and the smaller the maximumbubble size, the less violent the cavitation produced.1 The high frequencies used in mega-sonic cleaning, 500 to 2000 kHz, produce controlled acoustic cavitation, which is charac-terized by mostly small, gas-filled cavities. Unlike the violent implosion associated withvapor-filled cavities in transient cavitation, controlled cavitation bubbles exhibit less vio-lent collapse,4 producing lower temperatures and pressure. As a result, megasonic cleaningsubstantially minimizes surface erosion and damage to substrates being cleaned. Stablecavitation produces light in the visible range (violet), while the light produced by transientcavitation is primarily in the ultraviolet range (with a peak at 270 to 290 nm).4

The bubble action of controlled acoustic cavitation is believed to be a primary particleremoval mechanism in megasonic cleaning.

Acoustic Streaming

Acoustic streaming is considered another primary particle removal mechanism ofmegasonic cleaning. Acoustic streaming is time-independent fluid motion generated by asound field. This motion is caused by the loss of acoustic momentum by attenuation orabsorption of a sound beam. Acoustic streaming enhances particle dissolution and thetransport of detached particles away from surfaces,4 thereby decreasing particle redeposi-tion. It also produces a much thinner boundary layer (less than 1 �m) than would be foundin a cleaning tank without megasonics.

Acoustic streaming velocity is a function of energy intensity, geometry, energy absorp-tion, liquid density and viscosity, and sound speed in the liquid. Streaming velocity hasbeen found to increase linearly with acoustic intensity (power). Velocity also increases lin-early with frequency. Streaming velocity also decreases with distance from the source, dueto attenuation.2

Acoustic streaming comprises several important effects: (1) bulk motion of the liquid,(2) microstreaming, and (3) streaming inside the boundary layer.

The primary effect of acoustic streaming is bulk motion of the liquid, the strong local-ized flow of cleaning solution. The shear force of the bulk liquid motion is the primary par-ticle removal agent. In a closed tank, forces due to sound pressure variation create this bulkfluid motion, which carries particles away from the substrate once the molecular attraction


of the particle to the surface is broken and the particle is dislodged. Bulk fluid motionincreases linearly with acoustic intensity. The bulk fluid motion shear force combines withthe other effects of acoustic streaming to increase particle removal.

A second effect of acoustic streaming is microstreaming. Microstreaming, also knownas Eckart streaming, occurs near oscillating bubbles, or any compressible substance in theliquid. Microstreaming occurs at the substrate surface, outside the boundary layer, becauseof the action of bubbles as acoustic lenses that focus sound power in the immediate vicin-ity of the bubble. This is a powerful type of streaming, in which the bubbles scatter soundwaves and generate remarkably swift currents in localized regions. The currents are mostpronounced near bubbles that are undergoing volume resonance and are located alongsolid boundaries. Microstreaming aids in dislodging particles and contributes to mega-sonic cleaning.4

Most of the flow induced by acoustic streaming occurs in the bulk liquid outside theboundary layer. However, there is a third effect of acoustic streaming, called Schlichtingstreaming, which is associated with cavitation collapse and is believed to assist in theremoval of small particles and their transport away from surfaces. Schlichting streamingoccurs outside the boundary layer and is characterized by very high local velocity and vor-tex (rotational) motion. The vortices are of a scale much smaller than the wavelength.Schlichting streaming results from interactions with a solid boundary. Steady viscousstresses are exerted on the boundaries where this type of rotational motion occurs, andthese stresses may contribute significantly to removal of surface layers.2

The combined effects of acoustic streaming produced in megasonic cleaning may slide,roll, or lift a particle from its initial position on a substrate, depending on the size and shapeof the particle, as well as the nature of the hydrodynamic force being applied. Acousticstreaming, both inside and outside the boundary layer, clearly enhances cleaning and otherchemical reactions. Particle transport is aided significantly by the strong currents and smallboundary layer thicknesses that result from acoustic streaming.2

Significance of the Boundary Layer

During megasonic cleaning, the cleaning solution flows swiftly past the substratebeing cleaned, forcing chemistry into contact with contaminant particles, removing themfrom the surface, and carrying them away. On a microscopic scale, during acoustic clean-ing, fluid friction at the surface of the substrate being cleaned causes a thin layer of solu-tion to move more slowly than the bulk solution. This layer of slow-moving fluid at thesurface is called the boundary layer (Figure 6). The boundary layer effectively shields thesubstrate surface from fresh chemistry and shields contaminant particles from the removalforces of the bulk fluid.

Within the boundary layer, van der Waals attractive forces have been shown to be sub-stantially stronger than the removal forces that result from acoustic pressure oscillations,acoustic velocity oscillations, or bulk fluid motion associated with acoustic streaming.

Megasonic cleaning has proved especially effective at removing submicron particles inpart, because it reduces the boundary layer. The higher frequencies of megasonic cleaningreduce the boundary layer to less than 0.5 �m, compared to the boundary layer of 2.5 �mproduced by ultrasonic cleaning frequencies. The primary effect of acoustic streaming isthe bulk fluid motion of the cleaning solution. The thickness of the boundary layerdecreases as the velocity of bulk fluid motion increases.

Reduction of the boundary layer yields several benefits. It allows fresh chemistry tocome closer to the substrate,2 and come into contact with smaller particles. This higherchemistry refresh rate results in faster cleaning. Boundary layer reduction increases the


Figure 6 Comparison of boundary layer in ultrasonic and megasonic cleaning.

effectiveness of the acoustic streaming removal forces by allowing the cleaning solution torush past the substrate closer to the substrate surface, forcing chemistry onto particles,removing them from the surface, and carrying them away. The small, controlled cavitationbubbles generated by megasonics are able to remove contaminants within the thinnerboundary layer. This effect is especially important in removing small particles and access-ing small surface features. Reducing the boundary layer results in increased removal ofsubmicron particles, particles that were previously protected by the boundary layer, as wellas increased particle removal overall.

In megasonic cleaning, the combined results of boundary layer reduction, acousticstreaming, and controlled acoustic cavitation are very effective at enabling smaller parti-cles to be removed.

CLEANING CHEMISTRY AND OTHER FACTORS

Several additional factors contribute to the effectiveness of megasonic cleaning. Theseinclude cleaning chemistries, fluid temperature, process time, and power.

Cleaning Chemistries

Megasonics cleaning may be used with a variety of chemistries, including water, neu-tral aqueous solutions, alkaline aqueous solutions, acidic aqueous solutions, ethyl lactate,alcohol, acetone, N-methyl pyrollidone, dibasic esters, and glycol ethers. Although mega-sonic cleaning is used primarily for particle removal, it can also be used to increase the effi-ciency of chemical cleaning with surfactants or detergents. Efficacy of removal of othercontaminants depends on the solution in the tank.

Cleaning chemistries play a significant role in megasonic cleaning, because the chemi-cal composition of the cleaning solution may affect how quickly the cavitation threshold isreached. In megasonic cleaning (as contrasted with ultrasonic cleaning), it is believed thatoperating at or below the cavitation threshold produces better cleaning results.

The cavitation threshold, defined as the minimum pressure amplitude required toinduce cavitation, has been found to increase with increasing hydrostatic pressure (under


most conditions) and to decrease with increasing surface tension, with increasing temper-ature, and with an increasing number of solid contaminants. A reduction in the number ofhydrophobic ions (such as C� and F�) will also decrease cavitation threshold, since theseions collect at bubble surfaces and prevent cavitation bubbles from dissolving.2

A lower cavitation threshold allows cavitation to occur more readily. This suggests thatcavitation could be mitigated under the following conditions: low surface tension, highhydrostatic pressure, low temperature, and the presence of as few solid surfaces and con-taminants as possible.2

The original RCA Standard Clean consists of sequential immersion in two chemicals:Standard Clean 1 (SC-1) and Standard Clean 2 (SC-2). The formula for SC-1 is one partH2O2, one part NH34OH, and five parts H2O. The formula for SC-2 is one part H2O2, onepart HCl, and five parts H2O. The addition of megasonic cleaning to the SC-1 solution sub-stantially enhances particle removal.5 Chemists have succeeded in getting very dilute solu-tions to clean effectively with the addition of megasonics to the cleaning process.

For example, in statistically designed experiments on semiconductor wafer cleaning,megasonic power was observed to be the dominant factor for particle removal using SC-1type chemistries. Both the bath temperature and the ratio of ammonium hydroxide tohydrogen peroxide were found to modify the effect of megasonic power on particleremoval. Using substantially diluted chemistries, together with high megasonic inputpower and moderate to elevated temperatures, resulted in very high cleaning efficienciesfor small particle removal.6

Table 2 presents typical chemicals used in the wet cleaning of silicon wafers.

Other Cleaning Factors

Cleaning fluid temperature, process time, and power are additional factors that can affectmegasonic cleaning results. In general, sound speed decreases with increasing tempera-ture. The optimum temperature for the cleaning fluid will vary with the type of substratebeing cleaned and with the type of particle that must be removed. The choice of tempera-ture will also depend on the specific cleaning solution being used and how effective it is tobegin with at room temperature. In megasonics cleaning, exposure time and megasonicpower are the most significant variables. The combination of megasonic controlled cavita-tion and acoustic streaming enables typical substrate exposure times of 1 to 30 min, withmost exposure times between 10 to 30 min. As megasonic power or exposure time increases,particle redeposition decreases. Increasing the power level directly affects bulk streaming.Higher power levels increase the microstreaming component of megasonic cleaning,

Table 2 Typical Chemicals for Wet Cleaning ofSilicon Wafers

Contaminants Chemicals

Organics SPM (H2SO4/H2O2)APM (NH4OH/H2O2) � SC-1

Particles APM (NH4OH/H2O2) � SC-1Metallics HPM (HCl/H202/H2O) � SC-2

SPM (H2SO4/H2O2)DHF (HF/H2O)

Native oxides DHF (HF/H2O)BHF (NH4F/HF/H2O)


reducing the boundary layer, and can shorten the process time required. Megasonic poweris affected by array geometry, manufacturing method, and bath geometry.

DESIGN CONSIDERATIONS FOR MEGASONIC SYSTEMS

Initial megasonic systems developed for general industry use had transducer arraylifetimes of a few months. Current technology has increased reliabilities to tens of thou-sands of hours (years).

For overall megasonic cleaning effectiveness, one must take tank design, fluid circula-tion and filtration, and system electronics into consideration. The tank size should be small,to minimize the amount of cleaning fluid used. Additional fixturing may be required toposition the substrate accurately in the bath; only those surfaces located within the acousticstream will be cleaned.

The system should incorporate efficient fluid circulation and filtration, to assist in finalparticle removal from the fluid. Acoustic cavitation dislodges the particles, and acousticstreaming carries them away, but they must be removed from the fluid to prevent theirredeposition on the surface being cleaned.

The electronics that drive the resonator are crucial. Piezoelectric impedance is verydynamic over frequency, temperature, and age. If computer-controlled electronics withpositive feedback are not used to supply the RF power source to the piezoelectric material,reliability is seriously impaired.

Additional important considerations when choosing a megasonic cleaning system arechoosing the appropriate power level and resonator for the fluid type to be used and thetype of particle to be removed.

CONCLUSION

Megasonics provides several advantages over ultrasonics for damage-sensitive sub-strates. Megasonic-controlled cavitation and high-power acoustic streaming enable sub-strate exposure times of 1 to 30 min and provide effective submicron particle removalwithout the substrate damage typically associated with ultrasonics. The lower pressuresand temperatures produced in megasonic cleaning reduce substrate surface erosion whilealso providing significant environmental benefits.

REFERENCES

1. Gouk, R., Experimental Study of Acoustic Pressure and Cavitation Fields in a Megasonic Tank,M.S. thesis, University of Minnesota, Minneapolis, 1996, 47.

2. Gale, G., Physical and Chemical Effects of High Frequency Ultrasound (megasonics) on LiquidBased Cleaning of Si �100� Surfaces, Ph.D. thesis, 1995, 4.

3. Zhang, D., Fundamental Study of Megasonic Cleaning, Ph.D. thesis, University of Minnesota,Minneapolis, 1993, 18.

4. Gale, G., Busnaina, A., Dai, F., and Kashkoush, I., How to accomplish effective megasonic parti-cle removal, Semiconductor Int., 133, August 1996.

5. Hottori, T., Trends in wafer cleaning technology, in Solid State Technology, Penwell Publishing,Nashua, NH, May 1995, S8.

6. Resnick, P.J., Adkins, C.L.J., Clews, P.J., Thomas, E.V., and Korbe, N.C., A study of cleaning per-formance and mechanisms in dilute SC-1 processing, in Ultraclean Semiconductor ProcessingTechnology and Surface Chemical Cleaning and Passivation, Liehr, M., Heyns, M., Hirose, M., andParks, H., Eds., Materials Research Society, Pittsburgh, 1995, 21.


CHAPTER 2.6

Equipment Design

Edward W. Lamm

CONTENTS

IntroductionChoosing the Correct Equipment PathSolvent Equipment

Open-Top Vapor DegreasersClosed-System Degreasing

Semiaqueous and Aqueous SimilaritiesAqueous EquipmentWater Rinsing

Rinse Tank DesignDryingSemiaqueous

Separation StageAncillary Equipment

Oil Skimming and FiltrationMedia FiltrationMembrane FiltrationPhysical Principles of CoalescingCoalescing EquipmentCoalescing ElementsLiquid/Liquid Systems

Water QualityAutomation

Mechanical SuperstructureReference

INTRODUCTION

In the late 1980s as the concern about the effects of CFCs on the ozone layer came to ahead, solvent cleaning was subjected to severe scrutiny. During the next few years, mostcleaning applications were evaluated regarding their ability to be converted to aqueous or


Figure 1 Equipment path questions.2

semiaqueous. With this new activity, the aqueous marketplace began to explode with newmanufacturers eager to gain a piece of the market that always seemed solvents weregranted by royal decree. The new regulations now offered a slice of the fief that was previ-ously out of reach. The manufacturers of chemicals that could be rinsed with water (semi-aqueous) and had minimal emissions also began to see a crack in the solvent armor. Theybegan probing into this new potential growth market. Those applications that could notundergo the conversion to aqueous or semiaqueous were tested with new solvents that hadalmost no ozone depletion potential, but emissions continued to be an issue. As this mar-ket began to grow, new equipment was required to meet the tighter environmental regula-tions that were instituted to assure solvent loss be kept to a minimum.

All these events changed the distribution of the cleaning market and the availableequipment. The new designs that were born will be explored along with the originaldesigns and the ancillary support equipment that enhances their performance.

CHOOSING THE CORRECT EQUIPMENT PATH

To look at the types of equipment, it is first necessary to understand how the decisionto select a cleaning process is reached. Essentially there are a few questions that must beanswered (Figure 1). The questions basically help the engineer decide the approach to takein regard to the cleaning agent to be used. This will then determine the type of equipmentto be investigated, solvent or aqueous.

Once the selection of the cleaning agent category has been completed, the field of avail-able designs is significantly reduced. This is a great start because there are over 100 equip-ment companies that provide in excess of 200 products.1

SOLVENT EQUIPMENT

Equipment designed for cleaning with solvents is divided into two simple groups, coldbatch cleaning and vapor degreasing. The first group, cold batch, is somewhat a throwbackto the paintbrush and coffee can era with a bit of scale-up and sophistication. A sink on a


Figure 2 Solvent cycle. (1) Heaters boil solvent to make vapors; (2) vapors condense on parts; (3)condensate drips into boil sump; (4) excess vapors condense on coils; (5) condensateflows to water separator where water is removed; (6) dry condensate overflows to ultra-sonic sump; (7) solvent overflows from ultrasonic sump to boiling sump.

barrel is a good example. This system is quite labor intensive, but simple to operate. Withthe addition of an agitation lift, the process is enhanced; however, it is still relegated tofairly noncomplex parts and far from the tight tolerances of precision cleaning.

Vapor degreasers on the other hand can run the gamut of low-end open-top toextremely sophisticated closed designs depending on the cleanliness required. The basicsof the concept are the same for all operations. Vapor degreasers are designed not only tovaporize solvent for cleaning and drying the parts, but also to confine and recycle the sol-vent and solvent vapor to maintain a healthful environment and to keep cleaning costs low.Vapor degreasing equipment must provide for:

• Cleaning to remove soluble and particulate soil• Recovery of solvent by distillation for repeated use• Concentration of the soils

Open-Top Vapor Degreasers

These requirements are met by the basic degreasing unit, which is an open rectangulartank with a pool of solvent in the bottom (Figure 2). The solvent is heated in the boil sumpand vaporized into a dense vapor layer (specific gravity greater than 1, heavier than air)that lies above the liquid and constitutes the vapor cleaning zone. The condensing coils,which are installed high on the inside periphery of the tank, condense the vapor reachingthat level. The solvent condensate is returned to the solvent boil sump via the condensertrough and water separator.

The very important vertical extension of the degreaser wall is the freeboard. The free-board provides a stationary air zone above the vapor level. It shields the normal vapor zone


from moderate drafts, which could carry the vapors into the work environment. In addi-tion, a very thin solvent film evaporates from the work, as it is slowly withdrawn from thevapor zone, and the freeboard area confines these vapors.

The degreaser work opening must be adequate to handle the work dimensions andcleaning cycle, but should be kept to a minimum to maintain economical operation andacceptable working conditions. Even when vapor loss is controlled and no work is passedthrough, every square foot of exposed surface permits loss of a given amount of solventrelated to the equipment design and the solvent used. Since vapor in the center of the tankmust be passed to the walls for condensation, extending the tank width increases turbu-lence, which causes entrainment of air resulting in vapor loss. As narrow a tank as feasibleis recommended.

Freeboard is the distance from the top of the vapor line to the top of the confining sidewall at the top of the tank. The freeboard zone reduces vapor disturbance caused by airmotion in the work area. The freeboard zone also permits drainage of the work beingremoved, evaporation of residual solvent, and drying of the part with a minimum of sol-vent loss as well as reduced solvent emissions into the air. In degreasers of extreme length,the height of the freeboard is increased. Generally speaking, the higher the freeboard, thelower the solvent consumption.

The basis of effective vapor degreaser design is control of the vapor level. The controlof the vapor zone provides for cleaning in freshly distilled solvent and also helps minimizesolvent loss. This control is best done by use of condensing coils. Condensing coils arelocated within the degreaser tank at a height above the boiling solvent equal to the workheight plus allowances for clearance below the work and a 3 to 9 in. vapor layer above thework (Figure 3). The usual design of a degreaser provides for the normal vapor level to beat the midpoint of the vertical span of these coils. Thus, the positioning of the condensingcoils also establisheds the freeboard height in a given tank.

Figure 3 Vapor degreaser.


To prevent excessive condensation of atmospheric moisture on the coil surfaces abovethe vapor line, the temperature of the water leaving the coils should be above the dew pointof the ambient air. To accomplish this, the water should flow into the lowest coil and outthe top coil.

The refrigerated freeboard chiller is designed to reduce solvent emissions at the solventvapor–air interface by placing a cool, dry layer of air above the vaporzone. This cool airblanket assists in confining the solvent vapors. The refrigerated freeboard chiller consistsof a coil placed on the inside perimeter of the unit, immediately above the primary con-densing coils. An external refrigeration unit supplies the coils with the necessary cooling.

The chiller unit will condense, and in some cases freeze, atmospheric moisture onto thecoils. The additional water from these coils should be handled by placing a separate troughunder the coils, draining to its own water separator for a holding tank. If an additionaltrough is not placed on the equipment, the coil should drip into the solvent condensatetrough, but a larger separator would then be considered for effectively separating the waterfrom the solvent. Additional water in the solvent may cause corrosion and shorten equip-ment life.

Water enters a degreaser from several sources:

• Condensation of atmospheric moisture on the condenser coils• Moisture on the work being cleaned• Steam or cooling water leaks• Water-soluble cutting oil

Water can form a boiling mixture with the solvent (an azeotrope) that is vaporized, caus-ing equipment corrosion, decreased solvent life, and increased vapor losses. All degreasersshould be equipped with a properly sized water separator.

In the water separator, the condensed solvent–water mixture drops into a troughbelow the condenser coils and flows by gravity to the separator. The mixture enters the sep-arator below the solvent level. The water with a lower specific gravity and insolubility risesto the top and is discharged through a water drain. Relatively moisture-free solvent is thendischarged through the solvent return line to the degreaser. This separation requires time.Since 5 min is a practical minimum, the separation chamber should have a capacity of atleast �1

12� the hourly solvent condensing rate.

A deeper separator is more efficient to operate than a shallow one of equal volume,because the solvent–water interface area is smaller in the deeper design.

To minimize solvent loss induced by air turbulence over the vapor zone, a degreasershould be placed away from excessive air currents, open windows or doors, heating andventilating equipment, and any device causing rapid, uncontrolled air displacement.Typically, the usual air circulation is sufficient to dilute small quantities of vapor that nor-mally escape from the degreaser. When the degreaser must be placed in an unfavorablelocation, a baffle on the windward side will divert drafts and protect the vapor level.

Closed-System Degreasing

The contained or closed degreaser (Chapter 2.11 by Gray and Durkee) offers all thebenefits of the open-top design, but enhances the process by eliminating the solvent–airinterface. This is accomplished by conducting the cleaning in a sealed chamber, therebypreventing emission. Another benefit of the closed chamber is the addition of a vacuumstep during processing. This feature aids in soil displacement from blind holes and


complete removal of solvent during drying. The major concern when evaluating this tech-nology is the associated cost and low throughput as compared with open-top systems.

The major difference in design between the open and closed systems is the parts move-ment. Unlike the open-top degreaser, when items are placed in the closed chamber to becleaned in closed-system degreasing, the parts never leave the chamber until the process iscomplete. The cleaning solvent is brought to the parts where they are cleaned, rinsed, anddried. This eliminates the possibility of a disturbed or collapsed vapor zone, which con-tributes to solvent consumption and emissions. In addition, a number of agitation options,which include spray, ultrasonics, and rotation, can be employed during any of the processsteps (Figure 4*).

These systems typically include multiple feed tanks for a variety of solvent cleanlinesslevels. For reclamation of the solvent and to provide a fresh uncontaminated rinse source,a still is an integral part of the design. As elimination of emissions is a key principle, heatexchangers are installed to remove solvent from the discharge of the vacuum system.

SEMIAQUEOUS AND AQUEOUS SIMILARITIES

In cleaning applications where water does not have a negative impact, both aqueousand semiaqueous systems have been given substantial consideration. There are a numberof similarities between semiaqueous and aqueous equipment. The major one is that theyboth use water as the medium to remove the wash medium, which provides for a huge sim-ilarity in the last two thirds of the process. With both using the same rinse design, thedrying options for elimination of moisture are identical for these processes. Obviously, theancillary equipment is also similar. Pumping, filtration, and water purification are all han-dled by equipment of identical design.

AQUEOUS EQUIPMENT

The process of aqueous cleaning can be divided into three specific components: clean-ing, rinsing, and drying. This is no different from solvent cleaning except that each processcomponent is conducted in a different piece of equipment, each section is generally moresophisticated than a section of the vapor degreaser. Of course, the cost associated in pro-viding the added detail is significant for aqueous design.

When washing a part, the contaminant is often removed through the introduction ofcleaning chemistry and mechanical force. Rinsing involves the removal of any residual soiland chemistry that remain after washing. It is important to perform this task without intro-ducing new contaminants, such as dust or impurities in the water. Drying is the process bywhich residual rinse liquid is removed without introducing any new contaminants.

In the cleaning step, a detergent that is typically diluted in water actually bonds to thesoil (oil, grease, or particulate). To be effective, the detergent requires temperature andmechanical activity to loosen the dirt. Both of these are important when evaluating thedesign of the equipment.

Mechanical force is typically used in both the cleaning and rinsing stages. There are anumber of options available (Table 1). Spraying is a fairly effective, low-cost method forlarge parts without intricate details or holes. However, for smaller components or partswith blind holes, immersion with an additional source of agitation is required. All theoptions provide a relative level of removal of both soluble and particulate soils and mustbe evaluated for the type of contaminant to be removed. Of course, all come at a price andmust be judged on their need and effectiveness.

*Chapter 2.6 Color Figure 4 follows page 104.© 2001 by CRC Press LLC

When people think of cleaning applications, the focus is typically placed on the wash-ing stage of the operation. In precision cleaning, however, rinsing becomes a much moreimportant step. The allowable contamination levels are lower, and spot-free drying isalmost always a requirement.

WATER RINSING

Rinsing is a technology, just as washing is. It is measurable, controllable, and directlycontributes to the effectiveness of the cleaning process. Effective rinsing can improve yield,reliability, and appearance; it is also an important factor in containing operation costs.

Rinsing removes two basic types of soils: (1) solubles, which encompass washingchemistries and other soils that dissolve in the cleaning media, and (2) insolubles, consist-ing of particulate dispersed throughout the cleaning media. Rinsing is based on the princi-ple of dilution. To develop an effective rinsing process, three questions must be answered:

1. What soils are present?2. How much soil is there?3. How much residue is acceptable?

What soils and how much soil there is can be determined with analytical testing. Howmuch residue is acceptable is a more difficult question—one that must often be answeredempirically by the end user.

Often, acceptable residue levels are defined by testing a cleaned part for acceptableperformance in its next operation or use. If the part performs acceptably after being putthrough the cleaning process, the cleanliness level is assumed to be acceptable.

It is important to minimize contamination in the rinsing steps and to allow the use ofless rinse water. Two rinsing techniques commonly used to minimize rinse water volumeare spray rinsing and countercurrent immersion flow rinsing. Spray rinsing, as the nameimplies, uses spray nozzles to direct the flow of rinse water over the parts. This type of rins-ing can be very effective, using much less water than a typical flowing rinse. Proper appli-cation of spray rinses is necessary to ensure that all areas of the parts can be rinsed and alsothat the spray is only activated when the part is present to be rinsed. Effective immersionflow rinsing is based on the successful completion of two tasks: first, the soils must be sep-arated from the part; then, the soils must be prevented from redepositing onto the part.This can be accomplished by several means—for example, sparging the surface to removebuoyant soils, filtering the solution for particulate, and maintaining continuous dilution ofsolubles and fine particulate.

Table 1 Mechanical Forces—Separating the Soil from the Substrate2

Relative Solubles Particle RelativeMethod Energy Removal Removal Cost

Spray High Good OK LowImmersion Low OK Poor LowAgitations

Bubbler Low OK Poor LowLift Med Good Good MedPropeller High Good Good MedUltrasonics High Excel Excel High


Separating the soil often requires mechanical energy, especially with parts having com-plex shapes, or those that are “nested” in blind holes and/or crevices. Table 1 lists severaloptions to accomplish this.

If ultrasonic agitation is used in the wash, it might also be helpful in the rinse. Manytimes, a higher frequency is used in the rinse than has been used in the wash. This facili-tates removal of smaller particles and reduces the potential for part damage.

Continuous filtration of the rinse baths is very important in precision rinsing. The levelof retention of the filter should reflect the level of cleanliness required. In systems with mul-tiple rinse tanks, the filter retention level is often reduced with each succeeding bath.Continuous dilution is also a method of preventing redisposition and involves four keyelements:

1. Concentration of tank chemistry in dragout (C), which is measured in parts permillion (ppm) (1 oz/gal � approximately 7500 ppm).

2. Volume of dragout (V), which is the volume of water/chemistry moved (with theparts and carrier) from the wash to the rinse stage.

3. Flow rate of rinse water (F), which is measured in gallons per hour (gph).4. Rinse tank equilibrium concentration (E), which is a function of flow rate and

dragout, to the point at which incoming and outgoing chemistry levels are equal.

These four rinsing factors are related by the following formula:

C � V � F � E

C � V defines the amount of chemistry entering the rinses. Precision rinsing generallyrequires low E values, therefore, high F values (or overflow rates) are required. Figure 5illustrates the process of continuous dilution.

One method for improving rinsing is the use of several rinse tanks in a series (Figure6). The rinse formula applies to each successive tank. This allows a significant reduction inequilibrium concentration with a fixed overflow rate (F). This arrangement increases the

Figure 5 Continuous dilution.2


capital costs of achieving a particular cleanliness level by requiring more rinse tanks, but itreduces operating costs by lowering the required overflow rate. It is important to note thatthe water flow is in the opposite direction of the work flow. This is called “counter cascade”rinsing. In applications with high dragout, a spray rinse may be used to remove the grosschemistry before the first immersion rinse, further increasing efficiency.

In precision applications, the quality of the rinse water itself can be a factor in the effec-tiveness of the rinsing stage. In most cases, deionized (DI) water is required. In the deion-ization process, organics are removed by carbon, and special functional exchange resinsremove the ions. Biological growth is controlled with ultraviolet lights and special filtra-tion. One method of measuring rinse water quality is through resistivity or conductivity(Table 2). This is a measure of the electrical insulation properties of the water. Dirty waterand tap water may contain many ions that conduct electricity, lowering the resistivity.

Rinse Tank Design

In addition to the process variables, rinse tank design can impact the effectiveness ofrinsing. The flow pattern of the water can be important in rinsing, and this pattern is a func-tion of the tank design. The most common design for rinse tanks is the single-sided, over-flow weir design. This design depends on dilution for effectiveness and has “dead spots”in the corners where mixing does not take place, thereby reducing its effectiveness.

Figure 6 Continuous dilution—several rinse tanks.2

Table 2 Deionized Water Quality2

Resistance Conductance Total Dissolved(Megohms) (Microsiemens) Solids (ppm)

18.2 0.055 None (higher quality)10.0 0.100 0.1154.0 0.250 0.2881.0 1.000 1.1500.4 2.500 2.875 (lower quality)


Another, more effective rinse tank design, which has become popular in precision rins-ing, is the four-sided overflow model (Figure 7). This design utilizes a laminar upflow ofwater, which improves mixing and eliminates dead spots. This design is also very effectiveat sweeping fine particulate off the surface, preventing redeposition on the parts. The over-flow design is often augmented by the use of high-flow recirculation filtration, which fur-ther increases the sweeping action. The return of the filtered water typically enters thebottom of the rinse tank. Return manifolds that are specifically configured for the fixtureare often used.

One of the important variables in rinsing is the cleanup rate. This is defined as “thetime it takes for the contamination in the rinse tank to return to a steady level, after theparts enter the bath.” An experiment was run to determine the cleanup rate for a four-sidedoverflow with recirculation/filtration and a single-sided rinse with recirculation/filtrationand a sparger. The four-sided overflow rinse had a faster cleanup rate than the single-sidedoverflow rinse. There are several factors that contribute to the improved efficiency. Thehigh internal flow and mixing in the four-sided design enhances solubility. The high-vol-ume laminar flow is efficient for particle removal and minimizes redeposition. In the four-sided design, the distance to the overflow is minimized, improving the sweeping action.

The four-sided overflow design can provide up to 60% advantage in rinsing over theconventional single-sided design. It has improved efficiency for both soluble soils and par-ticulate. For fixed overflow rate, process throughput, and cleanliness level, fewer rinsetanks may be required with a four-sided design. Alternately, for a fixed number of rinsetanks, a higher throughput may be possible. Figures 8 and 9 show the cleanup rate for thetwo designs, using soluble soils and particulate.

DRYING

The drying stage of aqueous cleaning has the objective of removing the residual waterthat is carried over from the last rinse. Removal of the remaining water can be quite diffi-cult depending on the geometry of the part being processed. The two basic additivesrequired to increase water evaporation are temperature and flow of air. The mechanicalmethods used to produce these effects include:3

Figure 7 Four-sided overflow weir with 360° saw-tooth weir design.


Compressed air blow-off Vacuum ovenInfrared lamp bank CentrifugeRecirculating air oven Solvent displacement

If moisture can be tolerated, an air blow-off or centrifugation of the parts may be all that isneeded. If the parts need to be dry to the touch, infrared lamps or oven drying may berequired. If a higher level of dryness is needed prior to subsequent processing, a combina-tion of drying steps may be used.

Parts configuration, the substrate involved, and the degree of dryness will dictatewhich drying method or methods are most suitable for the majority of parts involved.Plastics, copper, and aluminum containing alloys may have temperature restrictions thatneed to be considered. Parts with blind holes, threads, depressions, and narrow cavitiesmay require special handling. Small parts tightly nested together also offer special dryingchallenges when considering the best or most efficient design.

Compressed air is economical and can be used directly over a process tank to minimizedragout. It is especially effective on large flat surfaces. This method is ideal if some mois-ture can be tolerated or if used in conjunction with another drying process. Air velocity

Figure 8 Equipment considerations. Cleanup rate using solubles.

Rinse Analysis - Rate of Removal - Solubles

TIme (minutes)

Co

nd

uct

ance

(mic

rosi

emen

s)

15

10

5

00 1 2 3 4 65

Single overflow

4-Way overflow

1 gpm DI Make-up flow rate

Figure 9 Equipment considerations. Cleanup rate using particulate.

Rinse Analysis - Rate of Removal - Particulate

TIme (minutes)

Par

ticl

es(c

um

ula

tive

/ml)

40000

30000

20000

10000

00 1 2 3 4 65

Single overflow

4-Way overflow

1 gpm DI Make-up flow rate


dictates the percentage of moisture removed as droplets and the percentage removed byevaporation. A lower velocity results in a greater amount of evaporated moisture.

Infrared heat lamps are designed to focus heat where needed. A full line of area andchamber heated lamps are available. Infrared heat is clean, fast, and controllable. Samplepart configurations of delicate construction respond best to this type of drying.

Recirculating hot air ovens are commonly used by industry to dry parts cleaned withwater. Drying times are directly related to air velocity and temperature. These ovens areideal if a high degree of dryness is required.

Vacuum ovens should be used only as a polishing step as required to get the last littlebit of moisture off the parts. Vacuum of 1 T or greater is used at temperatures of 120°F orgreater.

Centrifugal dryers are used for small parts with simple configuration. Parts are spunat speeds approaching 1000 rpm for up to 10 min. The liquid can be recovered and returnedto the process or rinse tank as desired. This type of drying process requires little space, andoperation costs are relatively low. This process is, however, for small parts and it is not ade-quate if a high level of dryness is required.

Parts with complicated internal components or blind cavities may require final mois-ture removal using a water-displacing solvent. Any solvent immiscible with water can beused for this process. Molecular structure and physical characteristics all must be consid-ered carefully.

The primary cost of parts drying is energy. For that reason it is prudent not to dry partsany more than is essential for subsequent processing. Less energy is required to run a cen-trifuge than an air compressor needed for forced air blow-off. If heat is added to the dry-ing process, the cost increases as the temperature rises. In addition, depending oncircumstances, manual labor could be the most costly part of the drying equation.

SEMIAQUEOUS

In semiaqueous cleaning, there are basically two distinct categories of agents. They dif-fer based on the miscibility of the cleaning agent in water or the boiling points, i.e., themethod used to separate the cleaning agent from the rinse water.4 This is conducted eitherby gravity or by difference in boiling point. Separation by gravity is based on the immisci-bility of the solvent and rinse water. Boiling point differences are separated by distillationof the water-soluble solvents and rinse water.

As in any industry, there are a number of semiaqueous formulations created from dif-ferent solvent bases. Typically, the semiaqueous cleaning agent suppliers are the manufac-turers of the main components included in their products.

In the fundamental semiaqueous process, parts are cleaned of soil with a suitable sol-vent that often may contain a detergent. The solvent is then removed from the parts bywashing with progressively cleaner water. The parts are dried with hot forced air. To beeconomical, the cleaning agent must be separated from the rinse water by gravity or distil-lation. The rinse water may be purified further for recycle with membranes that rejectorganic materials.

Amajor advantage of the semiaqueous process is the high degree of waste recovery—theonly direct waste is a concentrate of the soil in the cleaning agent. A major disadvantage isequipment complexity. Relative to a vapor degreaser, semiaqueous equipment is expensive.

The cleaning tank is designed similarly to those for other cleaning agent systems. Theoperating temperature is from ambient to as high as 180°F, because of the high flash pointof semiaqueous cleaning agents. Soil concentration at equilibrium should be no more than


5 to 10 wt%. Cleaning time typically runs from 30 s to 5 min. Ultrasonic cleaning is oftenused for removal of particulates.

The next stage is called emulsion cleaning. The parts are removed from the tank and con-tacted with a rapidly moving stream of air (air knife) to blow off liquid cleaning agent from theparts. This is done for two reasons: (1) to adjust the soil concentration in the cleaning stageand (2) to adjust the cleaning agent concentration in the next emulsion cleaning stage. Too lit-tle blow-off will harm cleaning performance by raising the soil concentration in the clean-ing stage and reducing the cleaning agent concentration in the next emulsion cleaning stage.

Water is sprayed onto the parts in the emulsion cleaning stage. Again, this is done fortwo reasons: (1) to remove cleaning agent and (2) to continue the cleaning process with awater emulsion of the cleaning agent. The water emulsion is often a better cleaner than theconcentrated semiaqueous chemistry used in the cleaning stage because little soil is pres-ent in the emulsion cleaning stage. The temperature is increased slightly. Cleaning time isin the same range.

Cleaning agent concentration in water is from 1 to 10%. This is deliberately low to min-imize organic cleaning agent flow to the final rinsing stages.

Separation Stage

The separation stage is not part of cleaning per se, but refers to recovery of the semi-aqueous cleaning agent or to removal of oils from aqueous cleaning agents. Since the sep-aration stage is the keystone of a semiaqueous process, the opportunity to avoid problemsin that stage is worthwhile.

The term gravity separation refers to the driving force that controls the rate of separa-tion. That is the density difference between water and the cleaning agent, and is typically0.15 to 0.2 g/cc.

The emulsion is fed to a decanter for separation (in the gravity-separation process) andto a distillation column (in the distillation-separation process). Conditions in the decanterare deliberately different from those in the cleaning and rinse tanks; usually the tempera-ture in the decanter is higher by 20 to 40°F. The separation should take place in between 5and 30 min. An interface monitor in the decanter is used to activate pumps that withdrawthe top organic phase and the bottom water phase. Removal is usually done in batch modeto maintain the organic/water interface between prescribed levels.

Problems occur in a decanter system when the withdrawal of one phase becomes con-taminated with the other phase. A change in soil chemistry is a major potential cause ofcontamination. Another potential problem is foaming in the rinse tank, which can occur ifspray nozzles are not correctly sized and positioned.

Distillation separates chemicals based on differences in their boiling points. For mostsolvents of interest, the difference between the boiling point of the solvent and of water ismore than 70°C. That is well above the minimum of the 10 to 15°C acceptable for good oper-ation. Further, boiling points of soil are typically 200°C above the boiling point of water.

The key advantages of a distillation separation system are reproducible and forgivingseparation of soil from the rinse water and of water from the cleaning solvent. Operationcould be with batch or continuous mode, depending on cleaning load. Batch distillationsystems probably are less expensive.

Both types of separation schemes have been used in a variety of industrial situations.Decanters and distillation columns commonly are used in chemical plants and refineries.If the successful cleaning situation is one in which two solvents can be used—one of eachseparation type—the distillation option will work best. Distillation requires more capital($5000 vs. $2000) and consumes more energy than does operation of a decanter. However,


distillation is a more positive separation approach than decantation. It can be more easilymonitored, and is less affected by changes in soil chemistry.

ANCILLARY EQUIPMENT

Cleaning solution chemistry can be as benign as hot water or can be a mixture of waterand cleaning chemicals.5 Cleaning chemicals are typically used where heavier soils such asoils need to be removed. Hot water is used where water-soluble contaminants (such aswater-soluble fluxes) need to be removed from the part. The recovery and reuse techniquesdescribed apply to chemical-based cleaning solutions. Those cleaning solutions compris-ing water only can be dealt with using the techniques applicable to the recovery and recy-cling of rinse water.

The key to minimizing the disposal of cleaning solutions lies in extending their usefullife. At some point, the cleaning solution becomes too concentrated in contaminants for thecleaner to perform adequately. The contaminants that cause a cleaning solution to becomespent include both organic compounds such as free and emulsified oils and inorganic com-ponents such as dissolved metal, which are introduced into the solution as part of theprocess. They may also be components inherent in the cleaning chemistry or makeupwater, which build up over time. Processes that are used for recovering aqueous cleaningsolutions include oil skimming, media/membrane filtration, and coalescing.

Oil Skimming and Filtration

Oils removed from parts during cleaning can either be emulsified or “free,” dependingupon the cleaning chemical formulation. Some cleaners are formulated to reject soils,which allows the soils (typically oils) to float on the surface of the solution. Skimmers areused to remove these free oil layers. For those cleaners that are formulated to emulsify oils,the oil can be removed via a coalescing-type filter or membrane filtration.

Media Filtration

Media filtration (e.g., cartridges, bags, and sand) is used to remove suspended solidsfrom cleaning solutions and associated wastewater. No dissolved materials are removedand these total dissolved solids (TDS) remain in the water.

Membrane Filtration

Membrane filtration processes are pressure driven and are used for various aqueousseparations. Several types of membrane processes are used (microfiltration, ultrafiltration,nanofiltration, and reverse osmosis) depending upon the size of the contaminant toremove. The two most important membrane separation processes used in the recovery andreuse of aqueous cleaning solutions are microfiltration and ultrafiltration. The limitationson these processes are those created by the presence of material that can foul, scale, or dam-age the membrane.

Physical Principles of Coalescing

Liquid/liquid coalescing technology is used to accelerate separation of an emul-sion. The principal driving force for coalescing action in either a gas or liquid stream is the


interfacial tension of the droplets. Interfacial tension is the excess free energy due to theexistence of an interface at the surface of a droplet, arising from unbalanced molecularforces. A relatively small interfacial tension value is typically required to obtain a coales-cence rate low enough for practical application.

In a carrier stream of dispersed liquid droplets, the total interfering effect of surfaceactive agents, particulate masking, or electrical charge is not great enough to render the dis-persion permanent. The interfacial tension value between the two liquids is neither drasti-cally reduced nor destroyed. Therefore, the dispersed droplets can be physically inducedto agglomerate and the natural process of fluid coalescing can be mechanically acceleratedto separate economically the liquids making up the emulsion. This provides the basis forliquid/liquid coalescing technology.

There are several different methods available to promote coalescence in an industrialprocess. Three primary mechanisms of coalescence are generally observed: impaction,Brownian diffusion, and turbulent field coalescence. Impaction occurs when the momen-tum of a droplet in the carrier stream causes it to collide with a droplet attached to a fiberor surface media, resulting in coalescence. The second mechanism occurs when theBrownian motion of a droplet in the carrier stream causes it to collide either with anotherdroplet in the carrier stream or with a droplet attached to a fiber or surface media. In tur-bulent field coalescence, drops that have associated in pairs are pushed through the smallcapillary passage of the bed or barrier, resulting in turbulence in the carrier stream. Theassociated droplets eventually coalesce as a result of their relative motion when passingthrough the capillary.

Coalescing Equipment

Industry uses a variety of mechanical means to effect fluid coalescing. A settling tankreduces the velocity of a liquid emulsion and provides a quiescent zone. At low velocity,the dispersed droplets agglomerate and form a second continuous phase because of dif-ferences in specific gravity.

Additional techniques are used to improve the coalescing rate in settling tanks, includ-ing directional flow inducers and baffles. System modifications may include recycling theexcess dispersed phase and flowing the emulsion through beds of coarse, porous media,such as wire mesh or fiberglass.

Similar methods are used to effect gas/liquid coalescing. Surge tanks are used toreduce the velocity of the gas stream, encouraging the agglomeration of liquid droplets.After the droplets settle, they are removed from the system. In many instances, vessels usedevices to induce centrifugal flow and create abrupt changes in the direction of flow.

Coalescing Elements

Using elements with a medium of engineered surface and pore-size characteristics canaugment coalescing of fluids. Several factors need to be considered when selecting themost effective fluid coalescing element.

1. The size and range of the openings (pores) in the porous material.2. The relative surface tension value of the fluids.3. The degree of wetting of the porous material exhibited by the fluid. (This is

related to the surface tension value between the liquid and porous media.)4. The fluid pressure drop across the coalescing media.5. The chemical compatibility of the fluid system and the coalescing element.


Liquid/Liquid Systems

A liquid/liquid system that is a candidate for coalescing is generally in the form of anunstable emulsion. An emulsion is a dispersion of fine droplets of one liquid in a second inwhich the first liquid is completely immiscible or incompletely miscible. Generally, emul-sions are formed by the mixing or mechanical agitation of liquids.

The dispersed fine droplets will rise or fall in the continuous liquid column as a resultof differences in liquid densities. The droplets may impact other droplets, agglomerate,and become larger (coalesce). However, interfering factors usually retard or prevent natu-ral coalescing at an acceptable rate.

WATER QUALITY

Water of defined quality is needed for controlled cleaning. High-purity water is usu-ally needed for precision cleaning; 18.3 M�-cm is considered the measure of perfectionmost commonly sought the world over when talking about water purity. The only com-monly available way to achieve this resistivity level is by use of deionization. To appreci-ate fully what deionization is and how it works, one must first look at the contaminantsfound in water and what purification processes are needed, in addition to deionization, toprovide water purity for a specific application.

Because pure water is the “supreme” solvent, it actively gathers contaminants fromeverything it passes over or through, including, potentially, the parts that are trying to becleaned. Dissolved ionized solids such as sodium (Na), calcium (Ca), and chloride (Cl) arestripped from rock and soil. Organic molecules are gathered from decaying debris andenvironmental pollutants. Particulates include organic debris, dirt and rust from soil andpiping; bacteria and microbials (including pyrogens) from normal growth in water; dissolvedgases such as chlorine (Cl) and carbon dioxide (CO2) from water treatment and organicdecay, and colloids from rock and sand. All these contaminants are present in varying con-centrations in water. Each presents different problems depending upon the application.

Deionization alone can allow achievement of 18.3 M�-cm resistivity, guaranteeingwater free of ionic contaminants, but it does not remove organics, particulate, bacteria, ormicrobials. To remove these contaminants, other types of purification are used in conjunc-tion with deionization. Activated carbon is used to remove organics and chlorine gas.Filtration is used to remove particulate and bacteria. Ultrafiltration is used to removemicrobials, including pyrogens.

Resistivity is the measure of how much electrical current will pass between two elec-trodes at a specific distance. When an electrical current is passed through a solution suchas water, ionic molecules are used as stepping stones by the electrical current. The fewerstepping stones, the more difficult the passage becomes, and the higher the resistivity read-ing. Most organic and bacteria are not adequate stepping stones to change the resistivity ofwater appreciably.

The temperature of water will also have an impact on its resistance. For this reasonmost water systems incorporate a meter that will automatically compensate temperaturesto 25°C, the standard for water purification. The maximum achievable resistivity readingof water at 25°C is 18.3 M�-cm.

Ionic contaminants exist dissolved within the chemical structure of water. Dissolvedionized solids and dissolved ionized gases are removed using ion-exchange resins, whichact like tiny magnets stripping ions from water, replacing them with H and OH ions, whichultimately join to form water (H2O).


Ion-exchange resins are for the most part synthetic polymers with several ion-exchange sites attached to the surface. Two basic types of ion-exchange resins are used.Cation removal resins have several hydrogen ions (H � ) attached to their surface, capa-ble of exchanging for positively charged ions. Anion resins have several hydroxylgroups (OH � ) attached to their surface, each capable of exchanging for negatively chargedions.

In a two-bed cartridge, these reactions occur separately with the cation removal resinbeing used first, followed by anion removal resin. A two-bed cartridge is used to removethe bulk of ionic contaminants, because when the two resins are separated, the cartridgehas higher effective capacity for ionic molecules. However, a two-bed system cartridgecannot fully remove all the ionic contaminants because the reaction is never completed.

To achieve totally deionized water, a mixed-bed cartridge is required. The mixed-bedcartridge is configured so that the cation and anion resin are mixed. When a reaction takesplace in a mixed-bed cartridge, the by-products of one reaction are picked up by the corre-sponding reaction, thus taking it to its completion.

As previously explained, deionization alone may not be enough for a specific applica-tion. This is the reason a system should incorporate more than one method of purificationto deliver water free of any and all contaminants. The system should employ a pretreat-ment cartridge that utilizes a combination of macroreticular resin and carbon to prepare thewater for the deionization that takes place in the following steps. The feed water firstpasses through the carbon to remove organics and chlorine. These components couldpotentially reduce the effectiveness of the ion-exchange resin. From the carbon, the waterpasses through a layer of macroreticular colloids. Colloids are very slightly ionized,extremely small particles that both clog conventional filtration and reduce the ability of theresin to produce high-purity water. This would be followed by a two-bed high-capacitycartridge to remove the majority of ionic contaminants as a preparation for the ultrapuremixed-bed cartridge.

An ultrapure mixed-bed cartridge is then employed to remove all remaining ionic con-taminants yielding up to 18.3 M� water. Organics, which are still present after initial car-bon adsorption and deionization, are removed now using high-efficiency synthetic carbon.Membrane filtration is used as the final treatment to remove bacteria and particulate,which have passed through the previous steps. A 0.2-�m hollow fiber filter attached to thefaucet block performs the final filtration. For most applications, water after this step is suf-ficiently pure for use.

AUTOMATION

Whether the cleaning system is aqueous, semiaqueous, or solvent, automated partshandling can add enormous value to the process in terms of throughput, total output, andease of equipment operation. The obvious requirement for a mechanical assist to movingparts through a system is the sheer weight of the load. However, there are other benefitsthat automation provides. In addition to eliminating the labor cost required if the unit wereto be operated manually, automation increases consistency in the process, provides a trace-able process, and permits the use of static process control.6

Automated systems are composed of four main components: the mechanical super-structure, the drive systems, the control package, and the operator interface. All these sub-systems need to mesh with the entire tank line, which includes the tanks themselves, andenvironmental equipment. Consideration should also be given to up- and downstreamproduction.


The most visible feature that differentiates the various automation options of a systemis the mechanical superstructure. A number of standard designs are available anddescribed below.

Mechanical Superstructure

The basic objective for horizontal and vertical travel should be a clean design with min-imal moving parts, especially over the tank line. The automation system needs to be rigid,durable, adaptable to available footprint, and compatible with the chemistry in use.Concerns include overhead clearance and accessibility for the tank line to operators.

Overhead conveyors are chain or belt pulley systems mounted laterally over the cen-terline of the tanks. Stated tank lengths are typically exaggerated to allow for the transi-tions for vertical travel. There is no flexibility in altering processing and the only variabilityin throughput is by altering the speed of the conveyor.

Tank level conveyors use powered rollers to move payloads between stations and ver-tical movement is implemented by lifts in each tank. This can be an efficient approach toautomation. However, processing flexibility is limited, tanks are significantly oversized,and it may not be appropriate for delicate parts.

A walking beam is typically a top- or side-mounted fixture that indexes payloadssimultaneously. It can be advantageous in single-recipe, high-volume applications. It hasthe same limitations in flexibility and throughput described above. Additionally, these sys-tems limit tank design in that all stations must be the same distance apart and all stationprocess times are identical.

I beam or cable systems employ suspended independent head(s), which travel overthe centerline of the tanks. The only advantage to these systems is where ceiling clearancesare an issue. By design, the moving parts of the heads inherently create potential forcontamination of the payload. Alternatives for low-ceiling applications include motionmultipliers, or where footprint constraints require front-to-back tank layout, three-axisautomation. However, given the potential for contamination, caution is required in cleanroom installations.

Cantilevered design has one horizontal frame mounted behind the tank line alongwhich one or more heads travel and execute vertical movement. Properly designed, thisconcept is considered optimal for general applications since it creates the least contamina-tion, uses the smallest footprint, and affords unimpeded operator access to the front of thetank line. Multiple heads, which overlap travel zones, can be an efficient way to increasethroughput, especially during “dead travel” with no payload. Any head can lift more thanone payload at a time for simple high-throughput applications (use of a “gang fixture”),although as in a walking beam the distance and processing time between stations must beequal.

Gantry/rim runners are two horizontal frames, one along each long axis of the tankline. From here the system can be essentially two I beam systems with associated contam-ination concerns, or mated cantilevered heads sharing weight distribution of the payload.The main disadvantage to this concept is that access to the front of the tank line is limited.


REFERENCE

1. Reynolds, R., Cleaning Equipment Directory, Precision Cleaning Magazine, Witter Publishing,February 1997.

2. Genet, C., Key requirements for proper rinsing in precision applications, CleanRooms East ‘99,Philadelphia, PA, Penn Well Publishing, Nashua, NH, March 1999, 125–143.

3. Quitmeyer, J.A., Aqueous cleaning process challenges, in Precision Cleaning ‘96 Proceedings,Anaheim, CA, Witter Publishing, Flemington, NJ, May 1996, 275–284.

4. Durkee, J.B., The Parts Cleaning Handbook: How to Manage the Challenge without CFCs, Section II,Semi Aqueous Cleaning, Gardner Publications, Cincinnati, OH 1994, 36–42.

5. Riley, C.T., Reduction/recycle/reuse concepts for aqueous cleaning process, in CleanTech ‘98Proceedings, Rosemont, IL, Witter Publishing, Flemington, NJ, May 1998, 128–136.

6. Aries, J., Automation: designing the right system for your cleaning equipment and productionintegration, Precision Cleaning ‘97 Proceedings, Cincinnati, OH, Witter Publishing, Flemington, NJ,April 1997, 296–305.


CHAPTER 2.7

Cold and Heated BatchSolvent Cleaning Systems

P. Daniel Skelly

CONTENTS

IntroductionThe Ideal SolventCold Cleaning

Pail and Scrub BrushHand WipeAerosol SprayRecirculating Overspray(“Sink-on-a-Drum”) Parts WasherImmersion Cleaning, Single-Dip Tank, with Manual Parts HandlingAutomated Immersion Cleaning, Multiple-Dip Tanks

Heated Solvent Cleaning MethodsHeated Dip TankVapor Degreasing

SummaryReferences

INTRODUCTION

In light of current and expected regulations, the trend of the 1990s has been to adoptaqueous (water-based) cleaning systems. In some applications, this may be the best choice.However, in other applications, water just does not work. Some considerations and prob-lems include requirements for pretreatment of water supply, waste stream handlingrequirements and costs, limited efficacy of cleaning due to low solvency for many soils ofinterest and high surface tension, energy costs of heating and drying, requirements for rins-ing and drying, high total cycle time, compatibility/flash rusting, complicated bath main-tenance, high capital equipment costs, high maintenance costs, and large equipmentfootprint.


THE IDEAL SOLVENT

When evaluating a new or replacement cleaning system, the ideal solvent would havethe following properties:

• Environmentally friendly—Does not create air or water pollution—Biodegradable

• Not regulated at the federal, state, or local levels—Not implicated in ozone depletion—Exempt from VOC regulations—Not a HAP—Not implicated in global warming—Not on the SARA 313 or other regulatory lists—Not a RCRA Hazard

• Solubility parameters match those of the contaminant to be removed• Works well as a single-component solution to avoid complex proprietary blends• Widely available at a reasonable cost• Compatible with all construction materials in the operation• Stable, does not readily break down in the presence of heat, metals, or chemical

contact, and does not require the addition of stabilizers to achieve this goal• Nonflammable at operating and handling temperatures• Easily (and inexpensively) distilled or recycled• Low toxicity (a high PEL), with extensive animal testing and a long application

history• Low or pleasant, yet detectable odor• Worker exposure easily controlled under the prescribed conditions of use• Fast evaporation rate for quick dry times• Low vapor pressure to minimize solvent losses

Unfortunately, no chemicals have every desirable property, and development of an idealsolvent is unlikely. Therefore, the end user must evaluate the particular cleaning require-ments as well as specific regulatory constraints.

Solvents are often characterized by their degree of perceived toxicity and rated as low,moderate, or severe. However, it is possible that the largest category, especially for the new-generation products, should be “unknown” or “unsure.” Classic solvents, including thealiphatic and aromatic hydrocarbons, alcohols, ketones, and chlorinated solvents haveeach been studied by numerous organizations and testing laboratories. Even with this siz-able database, scientists, toxicologists, and regulators seldom agree on the significance oftheir results. It is wise to assume that all chemicals have some degree of toxicity and a pri-ority should be to minimize emissions and worker exposure.

Once the solvent options have been reviewed, the cleaning method must be chosen. Sincethere are no completely nontoxic solvents available for cleaning applications, the system mustbe designed to minimize hazards to the worker and the environment. This may includemechanical controls such as tank covers and auxiliary cooling coils to condense solventvapors, or fans and exhaust hoods to remove solvent vapors from the workstation. Each ofthe solvent alternatives can be used safely with an appropriately controlled cleaning system.

With organic solvents, the choice of cleaning methods generally falls into one of three cat-egories: ambient temperature (cold cleaning), elevated temperature (hot liquid dip), or vapordegreasing (cleaning in boiling solvent vapors and often immersion in the liquid solvent).


COLD CLEANING

Cold cleaning with organic solvents and solvent blends is often used when water isdetrimental or ineffective, when the soils are of an oily or greasy nature, or when the capi-tal costs of vapor degreasing cannot be justified. Generally speaking, the majority of theindustrial cleaning applications can be accomplished in a cold solvent system. If cold clean-ing provides results that meet expectations, use it. This method will ordinarily be the sim-plest, most trouble-free, have the lowest utility requirements, and be the least capitalintensive of the cleaning system options.

Cold cleaning methods are as varied as the solvent choices that go with them. The mostsignificant limitations to cold cleaning are decreased cleaning efficiency as a function ofworkload, absence of a drying system, difficulty in controlling flammability, potentialworker exposure hazards, and regulatory compliance. However, these limitations can becountered by the solvent selection and by equipment design.

Pail and Scrub Brush

This method is very basic and has a low capital investment. However, solvent lossesand worker exposure may be excessive, particularly with solvents having a high vaporpressure and low allowable exposure limits. Brushing provides some abrasive action, butis generally not effective on small or intricate parts. A rinse in clean solvent is often neces-sary after brushing, and there is generally no means for reclaiming the solvent once itbecomes contaminated.

Hand Wipe

Hand-wipe cleaning can be accomplished by carefully pouring solvent on a reusablerag, or the purchase of presaturated disposable wipers. Mechanical rubbing with the wipeprovides some abrasive action, but unless the soil loading is low, it is likely to leave a thinresidue film.

Aerosol Spray

Aerosol cleaning is effective for removing soluble soils and the spray action helps toflush away insoluble particulates mechanically. However, it is generally inefficient in sol-vent utilization and is therefore reserved for small bench-scale and precision cleaningapplications. Depending on the solvent selected, there is a potential concern for flamma-bility and/or worker exposure to high levels of the atomized solvent.

Recirculating Overspray (“Sink-on-a-Drum”) Parts Washer

This is a standard method for garage and maintenance shops, and has reasonablecleaning potential until the solvent becomes dirty. The solvent (traditionally a mineral spir-its blend) is often replaced under a service contract, but it is necessary to assure that the sol-vent will be replaced often enough to meet the soil loading requirements. In addition, theconvenience of this service generally comes at a high price. These systems are not gener-ally suitable for high-vapor-pressure, low-flash-point solvents.


Figure 1 Recirculating overspray parts washer with vacuum distillation. (Machine by SystemOne,Miami, FL. With permission.)

It may be worth considering some of the new systems with built-in vacuum distillationfor on-site solvent recovery (Figure 1). Such a system can reduce overall solvent usage andminimize off-site waste disposal. In addition, freshly distilled solvent is available on a reg-ular basis and the need for frequent solvent change-out is eliminated, a particular consid-eration in heavy-duty operations. With solvents and solvent blends where there areconcerns for worker exposure and odor, the unit should be equipped with a hood andexhaust fan for proper ventilation. In areas of poor air quality, recent regulations havefocused on the VOC content of solvents traditionally used in sink-on-a-drum systems. Asa result, water-based cleaners have been the suggested replacement. Where organic sol-vents are required for performance, using a recirculating system with a hood and exhaustfan and with exempt solvents such as parachlorobenzotrifluoride (PCBTF) or volatilemethyl siloxanes (VMSs) may provide an additional option.

Immersion Cleaning, Single-Dip Tank, with Manual Parts Handling

Immersion cleaning is often the most economical cold cleaning method. These are sim-ple cleaning systems where the workload is lowered and raised hydraulically, mechani-cally, or manually into liquid solvent. Agitation generally increases efficiency. Air agitationis not recommended because of high solvent losses to the atmosphere, but ultrasonic agi-tation is often recommended because of its powerful scrubbing action. Mechanical agita-tion can be supplemented with a pump and filter.

Standard single-dip cleaning systems are offered by many equipment manufacturersfor aqueous cleaning. With only minor modifications, these units can sometimes beadapted for use with organic solvents (Figure 2). Where worker inhalation exposure andodor must be controlled, top enclosures and side workload entry can be added.


Figure 2 Single immersion dip cleaning system.2 (Machine design by Magnus Equipment,Willoughby, OH. From BCG-OX-36, Occidental Chemical Corp., May 1996. With permis-sion.)

Figure 3 Automated multiple dip cleaning system with cascade overflow. (Machine design byFinishing Equipment, Eagan, MN. From BCG-OX-36, Occidental Chemical Corp., May1996. With permission.)

In Figure 2, the parts are manually loaded on a roller conveyor, fed through a sideopening on the machine, then immersed and hydraulically agitated in solvent. At the endof the cleaning cycle, the deck is raised to the top position and the parts are allowed to dry.Drying is accomplished by passing a stream of ambient or heated air over the basket. Thisdesign is useful for light workloads and is adaptable to a wide variety of parts. Addition ofa still would enhance removal of oil.

Automated Immersion Cleaning, Multiple-Dip Tanks

A multiple-dip system (typically two to four tanks) is recommended for applicationshaving high soil loading. Sequential dipping into progressively cleaner dip tanks providesfor efficient solvent usage, and the final rinse is in the cleanest solvent. Automated partshandling is recommended to maximize process control and reduce worker exposure(Figure 3). The system is generally unsuitable for containing solvents that have a high


vapor pressure and low boiling point. Depending on the regulatory and toxicological pro-file of the solvent, additional controls may be needed.

In the design in Figure 3, an automated hoist controlled by a microprocessor picksup the workload at the operator station (far left) and processes the part(s) through a clean-ing cycle, a hot air drying chamber, and then returns clean dry parts to the operator station.The operator has the option of controlling duration of immersion, number of immersions,rotation, drying time, and drying temperature. A distillation system could be added toremove oils and keep the final dip tank supplied with fresh clean solvent.

HEATED SOLVENT CLEANING METHODS

In applications where parts are not adequately cleaned with a cold solvent, a combi-nation of temperature and solvency may be required. For example, buffing compounds,spinning compounds, and waxes are solids at room temperature and must be converted toa liquid for effective removal.

Heated Dip Tank

Although solvents are generally more effective cleaners when they are heated, thereare a significant number of disadvantages. Depending on the solvent selected, flammabil-ity may be a concern, solvent losses increase, and there is an increased potential for workerexposure. To address these issues, extensive safeguards may be required, equipmentdesign becomes more complex, and costs increase.

Vapor Degreasing

Although the capital investment can be significant, vapor degreasing (Figure 4) is avery effective and forgiving technology. Cleaning can be accomplished by immersion inhot solvent with agitation and ultrasonics. The final cleaning takes place in freshly distilledsolvent. This vapor blanket also helps to minimize solvent loss. The most important prob-lems relate to the additional engineering controls required to comply with environmentalregulations and to control solvent loss, to minimize worker exposure, and to use specificequipment design for low-flash-point solvents. In addition, for certain solvents, buildup ofwater and acidity must be controlled, so the process has to be monitored.

SUMMARY

Aqueous cleaning is not suitable for all applications; some solvent cleaning is appro-priate. There is not now and there is never likely to be an ideal solvent. With appropriatecontrols and subject to the particular regulatory climate, solvents can be used responsiblyin a variety of cold cleaning, heated cleaning, and vapor degreasing systems. The end usermust consider his or her specific application to select the best option.


Figure 4 Open-top vapor degreaser with still, hood, automated conveyor, and inert atmosphere.(From BCG-OX-36, Occidental Chemical Corp., May 1996. With permission.)

REFERENCES

The following references were taken from technical bulletins produced by OccidentalChemical Corporation:1. OXSOLs for Metal Cleaning, BCG-OX-19, January 1995.2. Cleaning Systems for OXSOL 100, BCG-OX-36, May 1996.


CHAPTER 2.8

Flushing Systems

Richard Petrulio

What makes flushing so different from the mainstream of cleaning processes? It is trulya unique process with subtleties that can make it difficult to develop. So what is flushing?Webster’s says it is to “cleanse with a rush of water.” From this simple definition the basicidea of a fluid moving across a surface to remove soils mechanically can be visualized.However, what is missing is the idea of enclosed, basically inaccessible surfaces that needto be cleaned. This goes beyond blind holes and slots to the fact that parts requiring flush-ing to clean internal surfaces are unable to utilize surface inspection to verify cleanliness.Thus, the process must be reliable.

The inability to verify cleanliness of a part without dissecting it is what drove themanufacturer of custom equipment for airline galley refrigeration to develop its own flush-ing process and equipment. Refrigeration systems such as these circulate relatively smallvolumes of refrigerant (hydrofluorocarbons) and oil through a closed-loop system. Both aprecision compressor and an electric motor, which drives it, are sealed within the systemand thus are exposed to air that is circulated with the refrigerant. Valves with small orificesand heat exchangers with many feet of tubing join in as main parts of the fluid loop. Soilssuch as metal chips, moisture, or incompatible chemicals can do catastrophic damage to arefrigeration system. Unfortunately, poor cleaning in this application will not show upuntil the equipment has been assembled and operated. Having a compressor lock up or anelectric motor burn up is a very expensive method of cleanliness checking.

The question may still be asked, so why is flushing required? The answer comes in twoforms: because you know what’s inside a part or you don’t know what’s inside a part.Mostly, removal of soils, known or unknown, from complex internal surfaces is driven byreliability. As with refrigeration equipment, metal fines and solid particles can damagemoving parts or foul critical passages. Other soils may combine chemically within enclosedsurfaces and slowly degrade the material. This activity is seen within a refrigeration sys-tem when water is left inside the parts. Water, when exposed to heat and refrigerant, canreact to form acid. The acid will remove copper from the walls of the tubing and redepositit on the surfaces of the compressor parts. Deposits of copper will grow on the movingparts and violate the clearances required for proper operation. Thus, after the equipmenthas been in service for a short period of time, it grinds to a painful halt, leaving the cus-tomer hot and the manufacturer with a tarnished reputation.

Soils left inside parts may not cause catastrophic failure; instead, they could limit per-formance. Passages that require fluid flow can become partially or totally blocked. Small


valves with delicate or precise sealing surfaces can fail to seal. Surfaces that transfer heatmay become insulated by soils causing loss of heat transfer rate. For many products theseproblems will cause annoying or embarrassing performance situations.

Other products perform tasks that are safety-critical. For these, performance and relia-bility are not just desires; they are requirements. Examples of such products can be foundaboard the U.S. space shuttle in the liquid oxygen systems. Of course, the issue driving theneed for flushing may have nothing to do with safety, performance, or reliability. It mayonly hinge on aesthetics.

Any issue that forces the need to flush must be addressed for its own requirements.However, if the product can operate with full performance, reliability, and safety whilelooking good doing it without flushing, then stop there. Do not flush time and moneydown the drain.

Now, if the decision to use a flushing process has been put in place, the details must befleshed out. Determining a level of cleanliness is the next logical step in developing theprocess. The goal is to clean only as much as required to consistently meet the productneed.

Develop the criteria and final results desired of the flushing process. The example refrig-eration system required flushing to remove particles that could do mechanical damage toa compressor or block critical valve passages. In addition, fluids that could form sludge oracid had to be removed. The next problem is how to determine the goal has been met.

Since the nature of parts that require flushing is that they have inaccessible internal sur-faces, verification is difficult. Two directions can be taken for flushing process verification.The first is continuous inspection of parts for results. The other is to develop the process sothat it will obtain the desired results without inspection. Although these methods couldapply to almost any process, flushing presents a substantial inspection challenge. To gainvisibility of internal surface cleanliness would require special equipment such as a boro-scope, methods such as chemical analysis of cleaning fluid, or dissection of the part.

Dissection of a part can only be used for spot or batch inspection because of its destruc-tive nature, although for high-volume, low-cost parts this can be effective. Fluid samplingcan give results for 100% inspection via analysis of samples for each part. Such a methodwould be suitable for small quantity parts, which require high precision and consistency ofcleanliness. The downside of this method lies in the tracking required for each sample andthe potentially long turnaround time for results. Immediate results can be obtained byusing sophisticated equipment such as a boroscope. However, such equipment is expen-sive and requires proper training to be used effectively. Additionally, the interiors of someparts are not conducive to accepting the boroscope. For these reasons, developing theflushing process to do the job right every time becomes an attractive method. In the case ofrefrigeration equipment heat exchangers and plumbing, developing the flushing processto ensure consistent results proved to be the best method.

Now that the forest view of flushing needs has been seen, it is time to look at sometrees. The soils to be removed play an important part in developing the verificationmethod. All known and potential soils should be listed. Each of the soils should be evalu-ated to determine if it actually needs to be removed. This is another opportunity to choosea no-clean option. Again, some of the soils may have no impact on the performance or reli-ability of a part. For those soils that do need to be removed, the process to flush them maynot remove them all in one step. Aqueous flushing will require multiple steps to clean outthe soils as well as rinse out the cleaning solution. Solvent flushing may also require addi-tional steps to rinse out the main solvent. An early flushing technique used on the refrig-eration heat exchangers employed a Stoddard solvent-based cleaner that had beenpunched up with perchloroethylene and methylene chloride. The solvent cleaned well but


was incompatible with the refrigerant and oil used in the final product. Thus, it required arinse step using HCFC-141b to flush out the Stoddard solvent. Both aqueous and solventflushing chemistries will likely require a drying step. The drying step is most difficult foraqueous and solvents that do not readily evaporate. As mentioned earlier, moisture is anenemy of refrigeration systems. Therefore, flushing with water did not look like a goodcandidate for the heat exchangers and plumbing.

As the methods and chemistry for flushing are being narrowed down, it is importantto consider material compatibility. In the same manner that soils were listed and theirpotential effects on the parts explored, compatibility of the cleaning compounds andprocess with the materials of construction must be evaluated. Substantial effort should beput forth to ensure that all materials that will be exposed to the flushing process would notbe degraded in either the short or long term.

Consideration for cleaning level, soils to remove, and material compatibility has nowbeen given. But, what thoughts have been given to the environment? Remember that clean-ing is a dirty business. A manufacturing facility must be able to provide a safe and appro-priate environment for the flushing process to be performed. Further, the process must bedesigned to have as minimal impact on its surroundings as possible. The solvent flushingprocess used for the heat exchangers required a location with ample space, electricalpower, and ventilation. Safety for the operators as well as governmental emission limitsrequired that the process maintain a tight lid on release of the chosen solvent. These con-siderations were made a part of the design for the equipment as well as the process steps.Even so, upon initial operation of the system, air monitoring was conducted to ensure thesafety of the operators and nearby employees. Long-term monitoring consists of emissionlogs to track any losses of solvent. Ultimately, common sense and sound ethics will dictatewhat equipment and steps are required to build a safe and environmentally sound process.

Before the equipment is built and dropped on the production floor, put the wholeprocess together virtually. Nail down the chemistry desired to fit with the projectedrequirements. Do not forget to have a backup. Next, envision the equipment needed to usethe chosen chemistry. From these write down each step of the procedure from start to fin-ish. Decide what skill level of operator is required to perform that procedure. Then, projectahead to when the process has matured some and look at who will be in charge of theprocess and equipment. Will the equipment be reliable? Who will be responsible for main-tenance? What about record keeping and follow-on training?

Building a virtual process will help shake out some of the bugs and shed light onpotential pitfalls. Now it is time to make the process a reality. Equipment can be obtainedoutside or developed in-house. Since flushing is unique even to the cleaning industry, find-ing a turnkey system that is off the shelf is nearly impossible. Custom-designed systemscan be fabricated but are generally very costly. In addition, the fabricator may not fullyunderstand the process, thereby making it difficult to get the system desired. For these rea-sons, it may be justified to develop a system in-house. In support of this phase of theprocess, it is worth employing outside help to provide industry contacts and keep the salesglitz to a minimum. Time spent interviewing the industry with a well-versed consultantalong is well worth the cost.

The flushing process developed for cleaning heat exchangers and plumbing in the gal-ley refrigeration equipment was designed and built completely in-house. The uniquerequirements of refrigeration eliminated the ability to use an aqueous process comfortablyand not flushing could not be an option. Thus, the use of a solvent-based flushing processwas the only option. The level of cleaning for the flushing process was difficult to deter-mine. It was clear that moisture of any magnitude could not be trapped in the system. Pastfailures due to copper fines being ingested into the compressor led to the discovery that


forming oils within heat exchangers and complex tube shapes held these fines in place.These were the main challenges for flushing of new equipment parts. However, the repairof used equipment presented the challenge of cleaning out coked oil following a burnoutof a compressor motor. A burnout occurs when the electric motor overheats or shorts outand causes the refrigerant and oil to burn. This generally will coat the entire refrigerant sys-tem with the tough black residue of the burned oil. To meet these cleaning needs, a veryaggressive flushing system was going to be required.

Fortunately, there had been a history of flushing activity used for the cleaning of refrig-eration equipment parts. The downfall of these past methods was caused by the increasedenvironmental awareness. Without the creation of new processes to take over for the oldunacceptable methods, product performance and reliability suffered severely. One of thefirst methods employed for flushing heat exchangers was to blast liquid refrigerant R12through the tubes and out to the parking lot through a hole in the wall. Verification of clean-liness required allowing some of the R12 to flow through a white towel. If the towelremained white, the flushing was done. Although this method worked very well, it wasdestined to be eliminated. The replacement method used the previously mentionedpunched-up Stoddard solvent to break down the oils and loosen debris. A flush of 141busing the spray wand within a degreaser rinsed out the Stoddard solvent residue. Finally,shop air was blown through the heat exchanger to remove and evaporate the 141b. Again,the method worked but it was costly and rather unsound environmentally. It was at thispoint that a systematic approach to providing a flushing process was initiated.

The understanding of soils and materials had been investigated and the time had comefor professional assistance. With the help of a top consultant, numerous solvent chemistriesand equipment options were investigated. As a result, n-propyl bromide was chosen to per-form the cleaning task. However, that still left the equipment end open. After numerousconversations, demonstrations, and some hard-to-swallow quotes, it was clear that theonly way to obtain the flushing equipment needed was to design and build it in-house. Theresult would be a safe, effective system with a reasonable price tag.

During the virtual process phase, at least five different systems were penned out. Eachfocused on the need to introduce solvent into the parts, flush out the soils, and then removethe solvent without exposing the operator to the chemical. The largest challenge wasremoval of the solvent from the part such that it was not vented to the atmosphere and yetthe part was left clean and dry. A sophisticated system was prototyped and given to therepair department for evaluation. Although it worked, a couple of the premises needed tobe revisited. The system was modified on paper for creation of the first full-scale system.Its operation employs a vapor degreaser to provide clean hot solvent. Pumps draw the sol-vent out of the degreaser and direct it through the part being flushed. Solvent is passedthrough filters to collect particulate soils while the degreaser separates out the oils from thesolvent. Once the automatic pump timers shut off the solvent circulation cycle, the solventleft in the part can be pushed out with carefully controlled nitrogen flow. The mixture ofnitrogen and solvent is delivered to a separating tank to allow the solvent to be capturedand the nitrogen expelled.

Each of the components needed for this process was selected from commercially avail-able hardware to ensure reasonable cost and allow for future replacement. Electrical con-trols and safeties were employed to allow for simple operation and control. As the systemtook physical shape, a comprehensive procedure and maintenance manual was written tobe available as soon as the equipment was put into operation. Initial use of the systemdemonstrated the success of the process. The development effort had paid off. However,not all was rosy. A couple of valves did not operate as expected, technicians found ways tolet small parts be ingested into the pumps, and the solvent separation left something to be


desired. Minor changes corrected the valve and debris issues; however, the separationissue took some thought. Because the temperature of the solvent was close to its boilingpoint when it entered the separation tank, some of it would exhaust with the nitrogen. Thesolution turned out to be a significant feature of the total flushing system. Refrigerationwas used to subcool the solvent prior to reaching the separation tank. With the solvent nowin a fully liquid state, it separated from the nitrogen effectively.

Two complete systems have been in operation with continuous use. Cleaning resultsfrom the flushing process have been consistent and effective. The equipment combinedwith thorough procedures, training, and a maintenance program has allowed an effectiveand reliable flushing process.


CHAPTER 2.9

Solvent Vapor Degreasing—Minimizing Waste Streams

Joe McChesney and Joe Scapelliti

CONTENTS

IntroductionVapor DegreasingFugitive Solvent Emissions Reclamation Using Carbon Adsorption

Overview of Fugitive Solvent EmissionProcess AdsorptionDesorptionCool DownCalculations of Solvent Emissions/System Design

Size of CAS SystemOperational Cost EstimatesElectricalWaterReturn on Investment

SummaryDistillation Process—Requirements and Calculations

FormulasExamples

INTRODUCTION

Every industry, from the huge automobile plant down to the smallest specialty manu-facturer, is impacted by the need to comply with environmental regulations, reduceamounts of chemical fluids used, and decrease waste streams.

With today’s ever-changing rules and regulations concerning solvent cleaning sys-tems, the cleaning industry has been greatly influenced by process control, from the stand-point of both equipment design and operational and maintenance procedures. Theinclusion of these tighter processes on solvent vapor degreasers has created a substantialprice increase in the original cost of the equipment. However, the savings resulting from


Figure 1 Typical vapor degreaser.

Flexible HoseVapor LevelSpray Lance

CondensingCoils

Water Separator

CondensateReservoir

Spray Pump

FreeBoard

WaterJacket

CondensateTrough

BoilingSump

Heater

reduced chemical usage and reduced waste streams will offset the price increase, usuallywithin the first year of operation.

Waste streams typically consist of fugitive emissions and/or dragout losses as well ascontaminated solvent residue in the distillation system commonly referred to as “still bot-toms.” This chapter first describes the solvent cleaning process. Guidelines and calcula-tions for reduction of fugitive solvent emissions by means of carbon adsorption arediscussed as is recycling of the contaminated residue by means of distillation.

VAPOR DEGREASING

Vapor degreasing is a process that has been used in the United States since the early1930s. It became a popular method for cleaning metal parts for both precision and generalmetal cleaning during the World War II and still remains widely used today.

In the beginning, the solvents used in the process were basically the chlorinated sol-vents such as methylene chloride, trichloroethylene, and perchloroethylene. In the 1960sthrough the mid-1990s, other solvents such as 1,1,1-trichloroethane, CFCs, HCFCs, HFCs,and HFEs helped expand the process into electronic assembly cleaning, the cleaning ofmedical implant devices (i.e., artificial joints and pacemakers), and aerospace hardware. Inthe last few years a new solvent, n-propyl bromide, has become a part of the process forapplications in both the general metal and precision cleaning market.

The vapor degreasing process is very efficient. It provides excellent cleaning and dry-ing in one tank. The distillation of the cleaning solvent within the system is continuous.This means the solvent is continually cleaned of soluble contaminants and used again inthe degreasing system. The life of the chemistry in the system is potentially infinite so longas it is managed properly. The floor space and energy required to run a degreasing processare generally less than running other cleaning processes.

Parts are cleaned in the degreasing process by vapor alone or vapor in combinationwith spray, or immersion and vapor, or immersion and vapor in combination with spray.Adding ultrasonics or mechanical agitation to the process can enhance certain cleaningoperations. The parts to be cleaned can be manually processed through a degreaser by anoperator using a hoist or they can be processed automatically using a robot or conveyingsystem. A typical vapor degreaser is shown in Figure 1.


Figure 2 Immersion/vapor degreaser.

Vapor Level

Condensing Coils

Water Separator

Condensate Reservoir

Free Board Water JacketCondensateTrough

Boiling Sump

Steam

The solvent used to clean the dirty parts is placed in the boiling sump and condensatereservoir. Heat is applied to the solvent in the boil sump bringing the solvent to its boilingpoint. As the solvent is boiled, a vapor is created and fills the machine. The vapor is main-tained in the machine by condensing coils. The condensing medium normally is recircu-lated tower water. For low-boiling–point solvents, recirculated chilled water or directrefrigeration is used. The condensing coils convert the solvent vapor into liquid. The liquidis collected in the condensate trough and flows into the water separator. The condensedsolvent flows from the water separator into a condensate reservoir. Excess solvent over-flows from the condensate reservoir into the boil chamber completing the distillation cycle.

Degreasing in vapor is a simple process. The parts to be cleaned are lowered into thedegreaser and allowed to dwell in the vapor. The part that enters the degreaser must becooler than the vapor temperature of the solvent. The solvent vapor begins to condense onthe cool surface of the dirty part. As the solvent condenses on the part, it dissolves the oilsor greases on the part. The dissolved oils and greases flow off the part and into the boilsump of the degreaser. This process continues until the work temperature and the vaportemperature are equal and then the cleaning stops. Spraying the part with distilled solventfrom the condensate reservoir will enhance the cleaning process. Spraying the partremoves insoluble debris such as chips, fines, and dirt. Spraying the part also cools the partbelow the vapor temperature allowing additional vapor rinsing to take place until thevapor and part temperature once again reach equilibrium. When the part is removed fromthe degreaser, it will be both clean and dry.

This process is suited for parts of simple geometry that can be racked so that the partsto be cleaned do not have surface contact with each other. Examples of parts that can becleaned in this process are flat sheets of metal, bar stock, simple stampings, molds, dies,machinery parts, transmission parts, engine parts.

Parts of a more complex geometry or parts that are nested together in baskets or carri-ers will require more than vapor or vapor/spray cleaning. For applications such as thesean immersion/vapor degreaser is required. See Figure 2.

Examples of parts that would best be cleaned in an immersion degreaser are screwmachine parts, heater cores, tubing, medical implants (i.e., artificial joints and pacemakercomponents), electronic assemblies, cosmetic cases, and fasteners.

Both cleaning cycles can be augmented with features to enhance the performance of thedegreasing system. Features such as ultrasonics can be added to assist in removing solids


Figure 3 Conveyor and transporter.

Figure 4 Conveyor and transporter.

from the surface of the parts being cleaned. Conveyors and transporters (Figures 3 and 4)can be used to move the work through the cleaning cycle.

Sometimes it is desirable to tumble the baskets of parts as they move through thedegreaser. Tumbling allows cleaning solvent to fill and then drain from cavities within thepart. Tumbling is also helpful in dislodging chips and fines on the surface of a part.Filtration equipment will remove solid debris. Distillation equipment will remove oil fromthe system on a continuous basis. Carbon adsorption recovers solvent from airstreams,keeping it out of the environment and returning it to the degreasing system for reuse.


Figure 5 Solvent recovery.

FUGITIVE SOLVENT EMISSIONS RECLAMATIONUSING CARBON ADSORPTION

Overview of Fugitive Solvent Emissions

Most manufactured products must be cleaned to remove lubricants, cutting oils, draw-ing compounds, miscellaneous contaminants, etc. used in the fabrication process.

When the cleaning process involves typical solvents, it is practical, efficient, and some-times mandatory that the emissive solvent vapors be recovered and possibly reclaimed.Carbon adsorption is one of the most efficient and cost-effective pollution control/solventrecovery processes available today. Carbon adsorption reclaims solvent vapors that wouldnormally be dissipated to the atmosphere.

Carbon is the preferred material used in adsorption systems because it exhibits uniquesurface tension properties. Because of its nonpolar surface, activated carbon will preferen-tially attract other nonpolar materials such as organic solvents rather than polar materialslike water. The granular multifacet geometry of carbon also possesses tremendous surfacearea (with 1 lb having an area greater than 750,000 ft2). This characteristic allows carbon toadsorb up to 30% of its own weight in solvent.

Solvent recovery consists of passing solvent-laden air through an activated carbon bed(Figure 5). The activated carbon captures the solvent molecules allowing residual denudedair to be exhausted to the atmosphere.

Process Adsorption

Solvent-laden air is directed from the exhaust source to the activated carbon bed by ablower/fan assembly (Figure 6, left). The carbon adsorbs the solvent vapor and residualpurified air is exhausted through the ventilation duct. This process continues until theentire carbon bed is near saturation.


Figure 6 Solvent adsorption/desorption.

Adsorption

Damper

Damper

Solvent Vapor

Laden Air In

ActivatedCharcoal Bed

Clean Air

Out

Desorption

Damper Open

Damper Closed

ReclaimedSolvent

Water Separator

WasteWater

Steam In

Condenser

Desorption

At the end of the time allowed for adsorption, the unit will automatically switch theincoming airflow from the first carbon bed to a second carbon bed. This will allow incom-ing solvent-laden air to flow through a fresh activated carbon bed while the first bed is des-orbed or stripped (Figure 6, right). The first bed is now injected with steam, which passesthrough the carbon bed vaporizing the adsorbed solvent. Additionally, the physical char-acteristics of the steam condensate passing through the carbon assist in removing solventresidue.

The mixture of steam condensate and solvent then passes through a water-cooled heatexchanger, which cools the solution. This allows for gravity separation to occur in a waterseparator device because of the difference in specific gravity of the liquids. The reclaimedsolvent is now ready for reuse or disposal. The water discharge is channeled for treatmentor disposal.

Cool Down

As hot wet carbon will not readily adsorb solvent, the carbon must be dried and cooledbefore the next adsorption cycle. Ambient air or process air is drawn through the bed for apreset period of time, which dries and cools the carbon. At the end of this cycle, the unitshifts into a standby mode ready for the next adsorption cycle.

Calculations of Solvent Emissions/System Design

The proper sizing of a carbon adsorption system (CAS) for a particular solvent appli-cation depends on two main factors:

1. The volume of recoverable solvent that is to be directed to the CAS2. The amount of air mixture that is to be directed to the CAS

With respect to the first factor, obviously, if the fugitive emissions that are lost cannot bepicked up in an airstream and directed toward a carbon adsorption system, these fugitive


emissions are not recoverable by normal means. The best way to determine the recoverableloss is to measure the amount of solvent in the airstream that is at present being emittedfrom the source, which is usually through a vent duct. This can be done with a variety ofmeters on the market today, but is most accurately performed with a recording device con-nected to a properly calibrated meter. It should be emphasized that solvent losses may notbe constant. This necessitates either continuous monitoring via a recorder connected to themeter or periodic sampling. For the best accuracy, all factors should be gathered, startingwith the total solvent purchases, and then determination made of how much is used, dis-posed as waste, and lost as a fugitive emission.

With respect to the second factor, accurate equipment measuring solvent/air mixtureis useless without knowing the exact amount of airflow. This can be best determined byusing a precise air-measuring instrument. In most cases, lower airflow rates are preferredto reduce the amount of fugitive emissions being generated and to increase concentrationof the mixture. This will allow the CAS to be more efficient.

Calculation examples for determining CAS size make use of the carbon capacity withthe solvent of interest (Table 1), and the rate of loss of the solvent (Table 2).

Size of CAS System

Customer vent system airflow is measured to be 2500 cfm. The solvent is trichloroeth-ylene. The daily stack loss average measurement is 1500 ppm for the first 8 h and for thenext 16 h is 800 ppm; the operation runs 6 days a week.

Referring to Table 2:

Table 1 Capacity of Activated Carbon (lb solvent/lb carbon at 80°F)

Solvent Level (ppm) TCE MC PCE CFC 113

20 0.052 0.00413 0.069 0.03350 0.066 0.00935 0.092 0.041

100 0.077 0.0154 0.099 0.050200 0.091 0.022 0.112 0.060500 0.110 0.037 0.129 0.077

1000 0.124 0.049 0.142 0.0912000 0.138 0.0633 0.152 0.1025000 0.153 0.08 0.167 0.125

ppm � parts per million; TCE � trichloroethylene; MC � methylene chloride; PCE � perchloroethylene.

Table 2 Loss Factors

Trichloroethylenelb/h loss � ppm � cfm � 0.00002175

Perchloroethylenelb/h loss � ppm � cfm � 0.00002756

CFC 113lb/h loss � ppm � cfm � 0.0000286

Methylene chloridelb/h loss � ppm � cfm � 0.00001414

cfm � cubic feet per minute.


lb/hr loss � ppm � cfm � 0.00002175For first 8 h 1500 � 2500 � 0.00002175 � 81.375 lb/hFor next 16 h 800 � 2500 � 0.00002175 � 43.4 lb/hOr for total day 81.375 � 8 � 651.0

43.4 � 16 � 694.41345.4 lb/day

The user’s records indicate machines connected to this vent are using 70 drums/monthand they operate 26 days/month, which means they are losing through the stack 26days/month � 1345.4 lb/day � 34,980.4 lb/month but are using 70 drums/month � 660lb/drum � 46,200 lb/month. Therefore, 34,980.4/46,200 or 76% is directly available forrecovery. A CAS is approximately 95 to 98% efficient, so one can expect to recover a maxi-mum of 0.95 � 34,950.4 � 33,231 lb/month.

Referring to Table 1, at the rate of 1500 ppm, a unit will hold approximately 0.13 lb ofsolvent per pound of carbon. Therefore, with a system containing 1200 lb of carbon,0.13 � 1200 � 156 lb of solvent can be recovered before desorption is required.

At this high rate, the CAS will require desorption in a little over 2 h so the timers canbe set to desorb every 2 h during the first 8 h. At the lower rate, i.e., 800 ppm, approximately0.12 lb per pound or 0.12 � 1200 � 144 lb of solvent can be recovered between desorbs andsince only 43.6 lb/h is lost during the remainder of the day, the timers can be set for 4 hbetween desorption for the 16 h the CAS is on low rate, thus saving water and steam.

Thus, by calculating or obtaining the following information:

• Operational hours• Amount of solvent to be adsorbed• Adsorption characteristics of applied solvent• Pounds of carbon required to adsorb incoming solvent amount• Airstream velocity/volume• Discharge limits to atmosphere• Safety factor

a carbon adsorption system size can be determined that will safely handle the solventemissions.

Operational Cost Estimates

The CAS system described above, with 1200 lb of carbon, needs 500 lb/h of steam, 2400gal/h of water, a 5 HP motor for the steam blower, and a 3 HP motor for the condenserwater to recover 4.5 gal/h of trichloroethylene.

Electrical

Steam power:

500 lb/h � 945 BTU/lb � 138 kW3415 BTU/kW

A 5 HP blower motor (460 V/3-phase/60 Hz) needs 6.04 kW; similarly, a 3 HP motor willhave an energy use rate of 3.8 kW. Therefore,


138 kW � 6.04 kW � 3.8 kW � 52 kW (misc.) � 200 kW total loading

Using the Tennessee Valley Authority (TVA) general power rate for a 200 kW totalload � 24 h operation � 31 days/month � 148,800 kWh at 100% loading.

GP-12 rating for these conditions per TVA area is $8.708/h average cost.

Water

40 gpm at 85°F inlet � 2400 gph

At a rate of $0.0013/gal, water cost is $3.20/h.

Return on Investment

Electrical costs: 8.708/hWater: 3.20/h

11.91/h operating costs

Using trichloroethylene at $5.87/gal × 4.6 gal/h (recovered) � $27.00/h (recovered) less$11.91 operation cost � $15.09/h net payback. Therefore,

$33,275.00 (system cost) � 2205 h payback or 138 days operating 16 h/day × $15.09/h

The recovery system can therefore pay for itself in well under half a year.

Summary

Carbon adsorption systems are ideal for solvent recovery. Numerous systems exist inthe field today reclaiming various solvents with high efficiency. Recovery efficiency as highas 95 to 98% of the incoming solvent-laden airstream can be achieved or exceeded.

In reclaiming this solvent, the system quickly pays for itself in solvent savings. Typicalreduction in gross solvent purchases due to reclamation is 50%. In addition, carbon adsor-bers help comply with EPA and OSHA regulations while providing a better workplaceenvironment for employees.

DISTILLATION PROCESS—REQUIREMENTS AND CALCULATIONS

The process of distillation occurs when a fluid is heated to its boiling point and con-verted to a gas. The gas is then condensed back to a liquid and contained in such a manneras to remove it from the area of the mixture containing the liquid phase and other contam-inants of fluids. Distillation is therefore a method through which contaminated or mixedprocess fluids may be separated, purified, and reused. Although it is not feasible in allcases, distillation should be given due consideration when practical.

The main factor to be considered when determining the feasibility of distillation is thecost/benefit ratio. Components to the analysis should include the following:

• Fluid purchase costs—How much can the purchased volumes be reduced? Will thisincrease the fluid unit cost by eliminating volume discounts?


• Waste disposal costs—What part of the waste stream can be reduced or eliminated?• Capitol equipment expenditures—What will distillation equipment cost? Unit price

can vary depending on the size and complexity of the unit and process.• Installation costs—Installation costs can be a significant percentage of the capital

equipment purchase price; 25% is usually a good budgetary estimate.• Floor space—What is the floor space required and what will it cost?• Maintenance costs—Will maintenance costs increase, decrease, or remain the

same? When associated with a cleaning process, maintenance costs can actuallygo down in some cases.

• Energy costs—Is there an energy source readily available? Will this source add tothe energy requirements or can it be supplied from another process or source thatmight otherwise be wasted? Additional energy costs will be associated with con-densing the vapor back to a liquid.

• Process control—Will the process be in better control if online distillation is addedor if fluids are changed more frequently?

• Environmental—In some instances the environmental issues outweigh all theother factors combined.

These are the main factors to be considered when evaluating a distillation process; theremay be others for a particular situation. These are included here as a starting point.

Types of process fluids are essentially divided into two classes:

• Flammables, which include alcohols, acetone, petroleum distillates, and manyothers.

• Nonflammables, which include water and other types of solvents.

Flammable fluids have special considerations when selecting equipment and processes.Nonflammable fluids by their very nature pose less risk in most instances.

Processes of distillation include three types. They are classed by the pressure at whichthey operate:

• Conventional or atmospheric stills• Pressure stills• Vacuum stills

Some systems operate with a combination of pressures or levels of vacuum in order toextract different fluids.

All distillation systems require at least four basic components:

• Containment vessel—To provide for containment of the solution. This vessel can beof any one of a number of configurations.

• Heat source—To boil the solution. This energy can be provided by any one of anumber of sources including steam, electrical, hot water, natural gas, solar, andothers.

• Condenser—To provide condensing of the gas back to liquid. This condenser canbe cooled by air, water, brine, refrigeration, and many other sources of conduc-tive media.

• Process controls—To ensure the process is producing the desired results. Controlscan include temperature controls, pressure controls, fluid flow controls, andother analytical devices as necessary.


In addition, some systems may incorporate one or more of the following:

• Vacuum pumps• Fractional distillation columns• Distillate analyzer• Feed pumps• Internal agitators• Thin-film applicators

For purposes of discussion, this chapter deals with atmospheric stills.Upon initial start-up, the system requires sufficient time to heat the contents and the

containment vessel to the boiling point before any appreciable amount of vapor is pro-duced. Additionally, the air volume in the vessel must be displaced by vapor before distil-lation flow is stabilized.

Formulas

The basic formula for calculating heat input for a given distillation rate is as follows:

H � D � ((�T � Sh) � Lhv) � (Hrl � A)

whereH � total heat input in BTUs/hourD � total amount of solution in pounds distilled per hour

Sh � specific heat of the solution in BTUs/lb/°F (This factor is readily available for puresolutions; however, the specific heat of a mixture will vary. This must be taken intoconsideration if the distillation rate is critical; otherwise using the Sh of the major-ity component is usually acceptable.)

�T � the difference in feed temperature and the boiling point of the solution in °FLhv � the latent heat of vaporization of the solution in BTUs/lb (As with specific heat, the

actual Lhv may vary depending on the different components of the feedstock.)Hrl � the heat lost from radiation to the surroundings in BTUs/ft2/h. (The radiation

losses may vary due to the containment vessel, boiling point of the solution,whether the vessel is insulated or not.)

A � the total area radiating heat to the surroundings in square feet.

This must include all the heated surfaces capable of radiating to their surroundings.Distillation rate for a given heat input can be obtained from the same equation as

D � (H � (Hrl � A))/(�T � Sh � lhv)

Examples

Table 3 contains density and boiling point data for the four solvents listed in Tables 1and 2.

A distillation unit made of 12 gauge steel designed to distill 2200 lb of trichloroethyl-ene/h will require a certain amount of heat. The heat requirement is calculated asfollows:


Table 3 Density and Boiling Point Datafor Four Solvents

lb/gal Boiling Point, °F

12.22 TCE 18813.55 PCE 24011.07 MC 10413.16 CFC 113 117

H � D((�T � Sh) � Lhc) � (Hrl � A)H � (2200 lb � (((188°F � 75°F � 0.225 BTU/lb h °F) � 103 BTU/lb))

� ((315 BTU/ft2) � (64 ft 2))� 302,695 BTU� 302,695 BTU/3414 BTU/kW � 89 kW

A similar unit designed with a fixed heat input of 60 kW of electric heat should distill a spe-cific amount of trichloroethylene. Expected distillation rate can be calculated as follows:

D � (H � (Hrl � A))/(�T � Sh � Lhv)D � (60 kW � 3414 BTU/kW) � ((315 BTU/ft2) � (64 ft 2))/

(188°F � 75°F) � (0.225 BTU/lb/°F) � (103 BTU/lb))� 1438 lb/h

Addition of insulation can lower the BTU loss from radiation significantly. Radiation lossin this example is from a bare metal surface and is approximately 7% of the total heatrequired. This can be lowered to less than 1% with proper insulation.

The formulas should work for most fluids considered recoverable by distillation pro-vided the necessary factors are available.

Payback calculations for a distillation unit can be made in a similar manner to thosemade for carbon adsorption systems.


CHAPTER 2.10

Vapor Degreaser Retrofitting

Arthur Gillman

CONTENTS

IntroductionEconomicsSafetyRetrofitting

Freeboard RatioMajor Emission Reduction Devices

Freeboard ChillerCarbonSuperheat

Additional Emission Control DevicesCoverControlled Speed Hoist

Retrofit Sources

INTRODUCTION

The first question would be why? If a unit is in good working order, and there are noparticular complaints, why make costly changes? There are good reasons, and they includeregulation, including federal (NESHAP), state, and regional. There are also economic andsafety issues.

The Halogenated Solvents NESHAP (National Emission Standard for Hazardous AirPollutants) is a federal regulation that specifically regulates vapor degreasers usingtrichloroethylene, perchloroethylene, l,l,l-trichloroethane, and methylene chloride (twoadditional solvents not typically used in vapor degreasing are also part of the NESHAP). Itis made up of a series of emission reduction choices. Assuming the vapor degreaser is inrelatively constant use, retrofitting is most often the best choice.

ECONOMICS

Where low-cost chlorinated solvents either cannot be used or are not a good choice,


there are new so-called exotic solvent and solvent blends to choose from. One of thecommon threads among many of these newer solvents is cost. They are expensive per pound,per gallon, or per drum. They are so expensive, costing perhaps $10,000 to $15,000/drum,that unnecessary solvent losses are worth preventing. Proper operating procedure, combinedwith a decent retrofit, can produce operating costs on a par with the older solvents.

SAFETY

Each solvent has a toxicity listing called a threshold limit value (TLV). Many of thenewer solvents are more toxic (lower TLV) than the solvents they replaced. Further, thechlorinated solvents are being reevaluated and may see lower limits set. One of the newersolvents, normal-propyl bronide (nPB), has a recommended exposure rate but it has notbeen firmly established. The government-approved rate has not been set as of this date.This all means that reducing operator exposure makes good sense.

RETROFITTING

Retrofitting means making physical changes and additions to the vapor degreaser.Although there are theoretically many things to be done, here are the tried-and-true “bestof the list.”

Freeboard Ratio

This is defined as the distance from the point where the boiling solvent vapor idles(usually around the middle of the cooling coils or cooling jacket) to the top of the machineopening. This dimension must be at least equal to the narrowest width of the overall vaporarea. Example: If vapor depth measures 20 in. and the tank measures 24 � 48 in., then thefreeboard ratio must be raised to the narrowest dimension of 24 in. or an increase of at least4 in. This is a standard that has been changing. The early vapor degreasers were typicallymanufactured with a freeboard ratio of 50%. Later, the federal government dictated thatthis ratio should be raised to 75%. The federal NESHAP rules now dictate a ratio of 100%.The question is often asked, “Will the ratio be moved higher?” This author’s opinion is no.The reason is that a ratio above 100% does not improve idle losses by much and the con-tinual raising of the freeboard causes operating problems, including interference with hoistmounting and ceiling heights. There remains the question of how much freeboard is “best.”The answer is, the more the better, but be practical! Even if one is not affected by NESHAP,a freeboard ratio of 100%, or greater, is going to reduce idle solvent losses.

Freeboard ratio can be accomplished by installing a stainless steel collar of the appro-priate height. Make certain that the collar is sealed and that the top is flat and sturdyenough to support a proper cover. We recommend that, where possible, the top shouldhave a lip that is turned in horizontally toward the tank opening and then formed downtoward the vapor.

Major Emission Reduction Devices

Freeboard Chiller

This consists of a second set of cooling coils, powered by a separate refrigeration com-pressor condenser, using an EPA-approved refrigerant such as 404A. The coils are mounted


as close to the primary cooling coils, or jacket, as practical. Freeboard coils can be mountedone side of the tank wall or around all sides. The combination of refrigeration power(motor horsepower) and coil surface area must produce a temperature at the center of thetank, and center of the coil system, that does not exceed 40% of the boiling point of the sol-vent. The larger the opening of the vapor degreaser, and the higher the boiling point of thesolvent, the more power and surface area is required to achieve the desired temperature.We recommend using finned tubing and mounting on all four sides for best results. Thereare situations where this is not practical because the addition of either finned tubing, or themounting on all four sides, chokes off the tank area too severely. In this case, do what youneed to get results. This might include mounting finned coils on one side only or usingstraight, nonfinned tubing, and raising the number of coil wraps to increase the surface area.

Carbon

Activated carbon systems have been used for decades and can be quite effective. A sys-tem consists of one or two specifically sized canister(s), a lip exhaust, a heat source torelease the solvent from the carbon, and a condensing system to collect the condensate. Theidea is to draw off vapor from the top of the vapor degreaser and trap (adsorb) it in the car-bon canister. When loaded, the canister is desorbed by heating the carbon, causing thetrapped solvent to turn to vapor where it is condensed. A two-canister system allows forcontinuous operation. Carbon systems typically are chosen only for very large vapordegreasers. Cost is the reason. Carbon systems can easily cost $100,000 and up. For this rea-son, carbon is warranted only when there is no other economic choice.

Super Heat

Superheat involves the addition of a heated surface placed in the vapor zone. By rais-ing the temperature of the solvent vapor above the boiling point of the solvent, liquid sol-vent that is entrained in the parts is boiled off and the result is less solvent dragout. Heatingmust be carefully controlled because surface temperatures that are too high can damage thesolvent. Each solvent has its own limits. The most common heating method is circulatinghot oil. The problem with super heat as a retrofit device is that it significantly reduces tankarea. For that reason it is not as popular as a retrofit choice and is most often considered amajor emission control device on new equipment.

Additional Emission Control Devices

Cover

Here is a simple test to determine if the cover style is adequate. Can one open and closethe cover quickly and not disturb the vapor? If not, then a non interfering-style cover isnecessary. Cover styles that do not disturb the vapor include roll top, sliding, and pivot.Some of these styles are available in both manual and automatic versions. Power covers arebest in two situations. The first is with large vapor degreasers where reaching across theopening presents a risk and unnecessary exposure. The second situation is any vapordegreaser utilizing a programmable hoist or material handling system. The reason isbecause it is often possible to integrate the automatic open/close function as part of thehoist or material-handling controller. The bottom line advantage is solvent saving, know-ing that the cover will be closed after each cycle and during idle periods.


Controlled Speed Hoist

The most overlooked emission control device is the controlled speed hoist. This deviceguarantees that the speed of the workload will always be at or below the 3.3 m/min max-imum mandated by NESHAP and other regional rules. This is the speed limit determinedto be necessary to prevent dragging out vapors. With large vapor degreasers and heavyloads, a hoist seems obvious. But with smaller systems it is often overlooked. The problemis that an operator has no concept of what 3.3 m/min means. Even if the operator did, thereis another problem and that is that the typical basket handle puts an operator’s arm at anawkward and uncomfortable angle. Going that slow is almost impossible. The result is thatmost solvent losses occur during load insertion and withdrawal. That means nothing anyone cando will save more solvent than automating the speed of the parts in and out of the vapordegreaser. Hoist systems can be as simple as pendent-controlled chain hoists, costingaround $1000, to microprocessor hoists (Figure 1) that can automate all of the movementinvolved in the cleaning cycle, as well as automatically turn on/off various vapordegreaser accessories, such as automatic covers, ultrasonics, and pump/filter systems.Cost of these automated systems can range from approximately $12,000 to $50,000,depending on weight and complexity of function.

Retrofit Sources

Contact the vapor degreaser supplier, as well as the solvent supplier. In addition, thereare independent retrofit suppliers who specialize in this area.


Figure 1 Automated hoist. (Courtesy of Unique Equipment Corporation.)


CHAPTER 2.11

Enclosed Cleaning Systems

Don Gray and John Durkee

CONTENTS

Background and DefinitionsAnother DistinctionA Dynamic FieldRationale

Principles of System DesignAirtight SystemsAirless SystemsExternally Sealed SystemsSummary, Enclosed SystemsRegulation of Enclosed Systems

FederalRegional

CostsAnalysisHidden Costs

Why Purchase an Enclosed System?SummaryReferences

BACKGROUND AND DEFINITIONS

While enclosed cleaning systems have been used for specific applications for severalgenerations, their popularity as a solution to broader industrial cleaning problems has onlyemerged in the United States in the 1990s.

At the heart of every enclosed cleaning system is a cleaning process. The general pur-pose of the enclosure is to protect the environment from emissions from the cleaningprocess. A second purpose is to implement some unique cleaning process within the enclo-sure, which could not be implemented outside the enclosure.


Enclosed cleaning systems are of three general types. They differ by the degree andmethod by which they are sealed from the ambient environment. As one would expect,sealing is the key issue in defining enclosed cleaning systems.

The three types are

Airtight—These systems are sealed to contain a light pressure above ambient.Typically, maximum pressure is around 0.5 psig.

Airless—These systems are sealed to contain either full vacuum (�1 mmgHg) or apressure significantly elevated above ambient (�800 to 10,000 mmHg). The wordhas been used generically, and later as a trademark.

Externally Sealed—These systems are not sealed to contain either pressure or vac-uum. Rather they are sealed to restrict interaction of the internal environmentwith the ambient environment.

Another Distinction

All enclosed cleaning systems bring the value of keeping the cleaning solvent “in thetank.” There are two very different methods by which this is accomplished. Usually bothare incorporated in any enclosed cleaning system—however, one is the dominant methodof emission control. Reliance on each method has very different consequences for users.

The first method is described by environmental engineers as “tailpipe control.”Generally this means solvent vapors leave the cleaning system in a stream of air and passthrough a bed of activated carbon to be adsorbed prior to discharge to the environment.Nearly every enclosed system uses carbon treatment for “tailpipe control” to meet envi-ronmental standards.

The second method would be similarly described as “pollution prevention.” Thismeans that the operating process has steps through which solvent liquid and vapor arerecovered and not allowed to leave the cleaning systems. Although not necessarily practi-cal, excellent internal recovery of solvent could mean that external carbon treatment is notrequired.

A Dynamic Field

This subject is a “moving target.” The commercial application of enclosed cleaning sys-tems is affected by environmental regulations, investment at purchase, perception of eco-nomics in use, competitive offerings, customer needs, availability and price of solvents,and quality of design. As this is written, all factors are in flux—especially environmentalregulations and investment at purchase. New and existing firms are providing new offer-ings of enclosed cleaning systems. Existing firms, currently offering enclosed cleaning sys-tems, are retrenching. Prices in the United States and attitudes about enclosed cleaningsystems are also in a state of flux.

Consequently, a comparison by supplier of offerings would be obsolete within a yearor so. For example, such a comparison written in 1997 would not have included the impacton the marketplace of the LAER/BACT regulations (lowest achievable emission rate/bestachievable control technology) and would be of little value to current readers.

So, this chapter will focus on basic differences among enclosed cleaning systems, gen-eral principles of operation, common process steps, and lasting disadvantages and advan-tages of their use.


Rationale

In a sense, the initial applications of enclosed systems were chemical reactors, auto-claves, or storage vessels. Only very seldom would a process engineer consider completionof a chemical reaction in an open system, and depend on controlling atmospheric diffusionrates to keep the feeds and products of reaction out of the ambient environment.

Similarly, when it became necessary to contain emissions from cleaning systems it wasnatural for process engineers to turn to sealed vessels. In traditional, liquid/vapordegreasers, diffusion-based controls (high side walls with refrigeration) are used withopen-top cleaning systems because they are low cost, not because they produce high con-tainment efficiency. Here the establishment of effective sealing mechanisms offers muchhigher efficiency of containment. When this degree of containment is demanded by envi-ronmental regulations, or for other reasons, cleaning experts turn to enclosed systems.

PRINCIPLES OF SYSTEM DESIGN

Design of enclosed systems is partially based on what is known about the equivalentcleaning process in an open-top vapor degreasing system. The designer of any enclosedsystem must consider the following principles:

1. Most of the same processes as practiced in an open-top vapor cleaning system canbe well used in an enclosed cleaning system. That is, almost any cleaning process(immersion, sonics, hot rinse, superheat, dry, etc.) practiced in an open-top systemcan be converted to an enclosed system for the purpose of emission reduction.

2. In addition, other process features may be added or subtracted. For example,vapor spray onto parts often leads to unacceptable emissions in an open-top sys-tem, but is a normal cleaning technique practiced in enclosed systems.

3. Environmental contaminants must not be allowed to enter the cleaning chamberof the enclosed system, or additional precleaning process steps will be necessary.Basically, the items of concern in the outside environment are humidity (water),noncondensables (nitrogen and possibly oxygen), and airborne particulates.Because the system is sealed, the process designer must be careful to eliminatethe entry of impurities that are not normally purged from the enclosed system.For example, retention of water and oxygen can lead to rapid deterioration of thesolvent. Finally, any process in an enclosed system must allow for separation ofthe solvent from the internal environment prior to release of that environment.

4. The environment inside the chamber of the enclosed cleaning system must not beallowed to enter the ambient atmosphere. Since this environment is rich (or pos-sibly saturated) in solvent, the result would be significant air pollution. One can-not simply “open the door” in the enclosed chamber when the cleaning cycle iscomplete, because the chamber is loaded with solvent.

This chapter discusses each of the types of enclosed cleaning systems, and provides someexamples where they have been successfully and unsuccessfully used.

AIRTIGHT SYSTEMS

Compared with other enclosed cleaning systems, airtight systems are simpler todesign, cheaper to construct, and more inexpensive to operate. Usually, the cleaning cycle


in an airtight system is rapid. In fact, a cleaning cycle in an airtight system may well beshorter than the same cycle in a open-top system.

Here is an example of a cleaning cycle for parts with a low thermal mass (ballpoint penrefills, catheter wires, electronic connectors, gold foils, eyeglass frames, etc.).

Example—Airtight 1

Total Elapsed Time (min:s)

A. Parts loaded in racks; racks loaded in chamber 0:00B. Chamber sealed; cycle selected and started 0:15C. Hot liquid solvent sprayed on cold parts, parts heat rapidly 1:45D. Superheated solvent vapor sprayed on hot parts to dry them 3:15E. Solvent vapors displaced with dry forced hot air 4:30F. Hot air displaced with cool air to cool parts 5:15G. Cycle complete; chamber unsealed automatically 5:30

The equivalent process in an open-top vapor degreaser would be immersion in a singlesump followed by spray-drying with hot vapor.

Steps E and F are required to satisfy the fourth principle, to keep solvents from escap-ing. The process equipment required for step F is a holdup chamber followed by a hugecarbon absorption column. The holdup chamber is mandatory because the hot air solventmixture is forced from the cleaning chamber at a rate higher than solvent can be adsorbedby the carbon absorber. This is an important point. It can be used to distinguish low-costenclosed cleaning systems of poor design from enclosed cleaning systems offering real value.

Those who would choose this process:

1. Desire solvent cleaning to avoid mineral residues (water spotting)2. Require extraordinarily low solvent emissions3. Greatly value short and controlled cycle time4. Have a low level of soil on their parts5. Have a high throughput and a highly automated process6. Are able to rack parts for exposure to liquid, vapor, and air sprays7. Require good to excellent drying

The first example illustrates a short cleaning cycle. As a second example, envision a sit-uation where the soil is difficult to remove and the parts have a high thermal mass. Such aprocess is similar to the first in respect to points 1, 2, 6, and 7, but could take nearly 50 min,because long soaking in hot solvent is required to remove soil (e.g., wax/gum, soot, or buff-ing compound). Ultrasonic cleaning may be required during the immersion process. In thiscase, step C, from the previous case, is replaced by a step lasting 34:45 where the parts areimmersed in hot boiling liquid solvent. The remainder of this process is similar to ExampleAirtight 1 except that the parts are elevated above the immersion vessel prior to the start ofdrying steps D and E. In step F, 45 s rather than 15 s might be needed to cool parts with ahigher thermal mass. These users would also need to monitor solvent quality frequentlybecause, in contrast to design principle 3, water and oxygen enter the sealed chamber.

While equipment is quite flexible, there is one common operation that cannot be com-pleted in an enclosed cleaning system: true boiling of the solvent. True boiling happenswhen the vapor pressure of the solvent equals the total pressure of the atmosphere. The for-mer is a property of the solvent molecules. The latter is atmospheric pressure in open space,


or some pressure in an enclosed chamber. Air has a partial pressure. The sum of partialpressures of the two vapors (air and solvent) must equal the total pressure. The presence ofa diluent (air) means that the partial pressure of the solvent can never equal the total pres-sure. Thus, the solvent can never truly “boil.” If the temperature in the chamber is raised,the partial pressure of air is raised per Dalton’s law; and the partial pressure of solvent fol-lows the vapor pressure curve upward.

Is it possible to have high solvent evaporation rates without boiling? Absolutely! Butthe net rate of vapor generation will not be as high as that in true boiling. The vapor con-densation rate on the parts is likewise impeded by the presence of air, which, with no placeto go, cannot be displaced by a vapor blanket as in an open-top degreaser. The solvent nowmust diffuse to the solid surface through the air surrounding the part—thus adding a sig-nificant resistance to condensing heat transfer.

There are no inherent limitations to which solvent can be used. Naturally, the solventshould be chosen to match the soil. Newer, engineered solvents such as hydrofluorocarbon(HFC), hydrofluoroether (HFE), and hydrochlorofluorocarbon (HCFC) are well suited forthe first application (short cycle time) because of their rapid evaporation rate. However,trichloroethylene (TCE), perchloroethylene (PCE), and normal-propyl bromide (nPB) areoften used to remove drawing oils and ink residues from ballpoint pen components.Flammable solvents are not commonly used with air-spray processes because of theabsence of the control of sparks. But they could be used readily with the second applica-tion (soaking).

AIRLESS SYSTEMS

Airless systems offer the most capability and power for customization of cleaningoperations, albeit with the highest price tag. The capability of pressure (vacuum) greatlyexpands the possibilities for use of airless systems.

The initial airless systems were developed for three applications: use of highly regu-lated toxic solvents, cleaning of large parts, and cleaning/drying of complex parts. Airlesssystems are commonly used in Europe, chiefly with hydrocarbon solvents. In addition,there are many untapped, unrecognized applications.

Typically, an airless system is a vacuum system, although a few systems use methylenechloride or other solvents1,2 under pressure in such applications as paint stripping.Temperature and pressure are linked because the boiling point of a solvent decreases as thepressure is reduced. The reverse is also true. For example, it is quite possible to use TCE(boiling point 189°F) at the temperature at which CFC-113 is used (117°F). The vapor pres-sure curve indicates what pressure should be selected to attain that temperature.

Several commercial processes will be described below. The first3,4 is typical of anapproach to cleaning and drying highly porous or complex parts. Examples include metalstructures impregnated with grease for lubrication, multiport injectors, or aluminum/alloyhoneycomb structures used as panels in construction of aircraft. The process schedule isdescribed below.

Example—Airless 1

Elapsed Time

A. Parts loaded in racks; racks loaded in chamber 0:00B. Chamber sealed; cycle selected and started 0:15


Elapsed Time

C. Air removed by vacuum pump-down to 1 mmHg (to remove water andoxygen; see design principle 3) 3:30

D. Parts sprayed with hot solvent vapor (which raises total pressure); thevapor condenses on the cold parts 5:30

E. Solvent vapors are removed by vacuum; parts are naturally cooled 7:30F. Step D is repeated (solvent vapor spray) 9:30G. Step E is repeated (vacuum evacuation) 11:30H. Step D is repeated (solvent vapor spray) 13:30I. Step E is repeated (vacuum evacuation) 15:30J. Chamber is filled and flushed with air (fed to carbon absorption column) 16:30K. Step E is repeated (vacuum evacuation) 18:30L. Chamber is filled and flushed with air (fed to carbon absorption column) 19:30M. Cycle complete; chamber unsealed automatically 20:00

Note that step C reduced the air concentration in the cleaning chamber to 1300 ppm (v/v)(equal to 1 mmHg/760 mmHg), the oxygen concentration to around 250 ppm (20% of1300), and the water concentration in a warm, humid ambient environment (100% relativehumidity at 90°F) to �50 ppm. For additional control of the process environment, the evac-uated chamber could be filled with clean dry nitrogen and again vacuum-evacuated to 1mmHg. That would reduce each concentration by a factor of 759/760.

The above process provides three separate stages of vapor degreasing. Parts are cooledbetween stages by vacuum removal of vapor. Obviously, this process could be shortenedby use of fewer wash stages. A cycle time of 15:00 to 20:00 is typical.

Another example where airless systems can provide excellent value is with hydrocar-bon-based solvents. Hydrocarbon-based solvents are used in Europe because of an aver-sion to chlorinated solvents. Hydrocarbons are high boiling (some are at 400°F), have aflash point above 200°F, are excellent solvents for hydrocarbon-based soils, have low odor,low skin irritation, etc. The drawback is evaporation/drying rates are very low.

Airless systems overcome the drying problem: (1) nitrogen is added to increase the pres-sure and dilute the hydrocarbon concentration, and (2) the chamber is vacuum-evacuated.When the total pressure is quickly reduced to around 10 mmHg, the “oily” hydrocarbons“fly off” the parts. Some vendors have excellent videos showing this effect. If drying is notsufficient, the cycle may be repeated. The process schedule is shown below. Steps A throughF are omitted because the operations and timing are similar to Example Airless 1. At stepD, parts are sprayed, then immersed in hot solvent, and ultrasonic agitation may be used.

Example—Airless 2

Elapsed Time

A–F. As above except step D as noted 9:30N. Step E is repeated (vacuum evacuation; first drying 11:30O. The chamber is filled with hot nitrogen and the pressure increased to

�500 mmHg 12:30P. Step E is repeated (vacuum evacuation); final drying 14:30Q. Chamber is filled and flushed with air (fed to carbon absorption column) 15:30R. Cycle complete; chamber unsealed automatically 16:00


That this technique is only seldom practiced in the United States does not mean it is not ofvalue. Here, improved technology is employed to allow use of an excellent and environ-mentally sound solvent—while overcoming the normal drawback of this solvent.

A recently commercialized process5,8 provides good value for cleaning of small parts.In the previous system,3 air is removed by vacuum before solvent is added. This is thedesirable system (according to principle 4). In this example, the vacuum step is only usedfor final drying. Instead, at step C, air and water are displaced by flushing with hot nitro-gen. Note that this system is not a “true” vacuum system in the sense that the cleaning isnot done at reduced pressure—under vacuum. But if the cleaning process downstream ofstep D is appropriate, the user should receive clean dry parts.

Example—Airless 3

Elapsed Time

A. Parts loaded in racks; racks loaded in chamber 0:00B. Chamber sealed; cycle already started 0:00C. Chamber flushed with hot nitrogen (to displace oxygen) 1:00D. Chamber filled with oxygen-free hot nitrogen after purging 1:30E. Hot solvent liquid introduced for immersion cleaning 2:00F. Immersion cleaning with ultrasonics 4:00G. Drain liquid and flush with hot clean solvent 4:30H. Second immersion cleaning with ultrasonics 6:30I. Drain liquid and flush with hot clean solvent 7:00J. Continuous flushing with hot clean solvent 9:00K. Drain liquid solvent from tank 9:30L. Blow hot nitrogen across parts to dislodge liquid 10:30M. Evacuate chamber to 1 mmHg, to dry parts 13:30N. Replace chamber environment with clean dry air 14:45

O. Cycle complete, chamber unsealed automatically 15:50

A final example involves an unusual solvent—water.9 This technique could also beused with other solvents. Vacuum technology is used to overcome the major limitation ofaqueous cleaning—drying. Evaporation of water at atmospheric pressure is slow, and sol-uble mineral salts are left behind on the parts as imperfections, stains, scars, or spots.

The evaporation rate is raised by the huge partial pressure difference between wateron the part surface and the water in vapor space. Naturally, the partial pressure of water inthe vapor space is low because the vacuum pump is continually removing all vapor fromthe chamber.

Evaporation of water under vacuum is quick, but brings an unexpected problem: ice.Remember, evaporation involves both a transfer of mass (water) as well as a transfer of heat.This is true for evaporation in a vacuum or under pressure. If water (or any other liquid) isevaporated, the heat of vaporization must be supplied. In this case heat comes from thesurroundings. Typically, without process modification, the parts become chilled, and theremaining water becomes frozen. This situation is much more critical with water than withorganic solvents because the heat of vaporization of water is �1000 BTU/lb and that fororganic solvents is �200 BTU/lb. The modification is to add hot air, hot water, or radiantheat, so that the heat of vaporization is supplied externally. Using the system for dryingonly has a cycle time of about 14 min.


Water spots are usually oxides and salts of metal ions. The metal ions are soluble inwater—that is the reason the final rinse is often with metal-free (deionized) water. The oxy-gen component is thought to come from the air. No metal salts are left on the surface sincethe oxygen has been removed from the chamber prior to evaporation of the water, andsince metal-to-nitrogen bonds are exceedingly difficult to form.

What happens to the metal ions left on the parts by the last rinse with water? Theseauthors do not know, but believe that the ions remain on the parts as ionic contamination.

EXTERNALLY SEALED SYSTEMS

Basically, these systems are open-top systems in an isolation chamber. Thus, the sol-vent is effectively separated from the atmosphere. Exhaust is vented through a carbonabsorption trap. In some cases, the isolation chamber is retrofitted on an existing open-topvapor degreaser.9 Additional designs have recently become available. Externally sealedsystems have been designed around traditional vapor degreasers as well as low-flash-pointsystems.

In externally sealed systems, loading and unloading of parts requires either more laboror additional capital for automation. In addition, cycle time may be increased.

SUMMARY, ENCLOSED SYSTEMS

The above information is summarized in Table 1.

REGULATION OF ENCLOSED SYSTEMS

Two major air pollution regulations cover enclosed systems. One is federal, the otheris regional. The federal regulation applies to all users in the United States. The regional reg-ulation, which is more stringent, applies to U.S. firms in that region, or in regions that havebeen defined by the EPA as having similar characteristics.

Federal

The federal regulation is one of many National Emission Standards for Hazardous AirPollutants, known by its acronym NESHAP. The halogenated solvent NESHAP was pub-lished in December 1994, and took effect in December 1997. This NESHAP covered clean-ing operations using the chlorinated solvents 1,1,1-trichloroethane (TCA), trichlor-oethylene (TCE), methylene chloride (MC), and perchloroethylene (PCE), and two othersnot used normally in cleaning operations. The basis for this standard was maximumachievable control technology (MACT) as defined in the 1990 Clean Air Act1 for these sol-vents. This was embodied in the engineering requirements for compliance as 50 to 70%control efficiency. This NESHAP does not cover other halogenated and nonhalogenatedsolvents.

The adjectives used here: airless, airtight, and externally sealed, were not in commonuse when the NESHAP for chlorinated solvents was developed. These terms are not men-tioned in the NESHAP. However, these three types of cleaning systems are covered byinterpretation of other language. In the NESHAP, airless systems are included under “sol-vent systems without an air–solvent interface.” Although airtight systems are full of air


Table 1 Comparison of Types of Enclosed Systems

Item Airtight Airless Externally Sealed

Operating pressure Low pressure Vacuum or Atmosphericpressure

Operating Temperature � Normal boiling Any Normal boiling pointpoint

Choice of solvents Volatile Any (volatile or Volatilenonvolatile)

Type of parts Best with low Any Best with lowthermal mass thermal mass

Best type of process Hot soaking in liquid Hot vapor spray Traditional vapordegreasing

Drying quality Normal with volatile Vacuum quality Normal with volatilesolvents solvents

Estimated investment 2.0 � open top 2.5 � open top 1.25 � open top(smaller units)

Estimated operating costs10 �10% � open top � open top � open top(with TCE and PCE)

Supplies One to two �Five Two (plus two withflammables)

and solvent, as are externally sealed open-top systems, airtight systems are classified in thesame manner as airless systems.1,2 Unfortunately, the NESHAP was written approximately4 years ago when externally sealed systems were not recognized. The EPA only considered“systems without an air–solvent interface” as those “that do not expose the cleaning sol-vent to the ambient air during or between the cleaning of parts.”1,3 The authors’ interpre-tation is that this definition includes only airless (and airtight) systems, but does notinclude externally sealed open-top systems. The EPA has recently confirmed this under-standing.1,4

Regional

Los Angeles and surrounding counties have a significant problem with smog causedby emissions of volatile organic compounds (VOCs). The EPA defines this region, regulatedby the South Coast Air Quality Management District (SCAQMD) as “non-attainment” forfederal VOC guidelines. The concepts of lowest achievable emission rate (LAER) or bestachievable control technology (BACT) apply in all nonattainment areas. These concepts arebeyond the scope of this chapter. The situation is contentious and, as of this time, subject tocontinuing interpretation. Given these extreme problems, manufacturers wishing to usenonexempt solvents in new operations (this includes changing locations and changing sol-vents in a given operation) might do well to consider an enclosed cleaning system withdocumented, demonstrated emissions values.

Other regions of the United States are nonattainment areas. If solvent cleaning innonattainment areas is contemplated, one should insist on a commercial-scale demonstra-tion or a supplier certification prior to purchase of a new or rebuilt solvent cleaning system.In addition, the regional environmental regulatory agency should provide writtenapproval of this evidence to satisfy compliance requirements.


In summary, although all enclosed systems should easily meet the NESHAP require-ments, in areas determined by the EPA to have poor air quality, performance-basedevidence should be required from the supplier and accepted by the regional regulatorybody before enclosed systems are purchased for use in nonattainment regions.

COSTS

Enclosed systems of all types are more expensive than open-top liquid/vapordegreasers, in part because there are relatively few producers of enclosed cleaning systems.Technical justification for choosing the more complex systems, beyond cleaning and emis-sions control, may include better drying and operation at a lower temperature with lesspotential damage to parts.

The authors estimate that the average ratio of initial investment costs relative to open-top systems is 2.5 for airless, 2.0 for airtight, and 1.3 for externally sealed systems for com-monly used, smaller systems. In general, the larger the cleaning chamber, the more costlythe system. However, the differential relative to open-top systems begins to converge forlarge systems (approximately 75 ft3 chamber volume). These are estimates; additionalprocess control and parts handling may increase the initial investment.

Analysis

Although some data have been provided by equipment suppliers, the analysis, andconclusions are those of the authors.

Are the costs of enclosed systems justified? Do enclosed systems pay for themselves?Over what period? Unbiased answers are difficult to obtain. Nearly all studies are based onestimates or forecasts.1,5 With chlorinated solvents, the decrease in solvent purchases with99.X% control efficiency system over 70% control efficiency probably will not pay for theadditional �2.5 times greater investment needed from an enclosed system.

For more costly solvents, a stronger case may be made for enclosed cleaning systems.What if a solvent costing $15/lb is used? For HFC-43-10, HFE-7100, or AK-225, the authorsestimate that the enclosed airless system costs less to operate than does the open-top clean-ing system.

Studies based on customer experience are difficult to apply to other situations becauseof the narrow focus of the customer’s application, the customer’s youth on the learningcurve, and the small number of systems constructed to date in the United States (�75).Additional cost issues do not normally register on a cleaning cost sheet: reduction of haz-ardous solvent-based waste and reduced cost of obtaining and complying with an envi-ronmental permit.

One reason to purchase an enclosed system is to reduce labor costs. The main compo-nent of operating costs (approximately 80% for very large systems) for all enclosed systemsis capital payback. Additional costs include, in decreasing order of significance, labor, sol-vent, and miscellaneous (power/waste disposal). In contrast, for open-top systems, capitalpayback is much less significant, and operating labor is much more significant.

In the authors’ surveys, capital investment was found to be the main barrier to imple-mentation of enclosed airless systems. Total cost of ownership (capital payback � labor �solvent purchase � solvent disposal � miscellaneous) should be significant in the minds ofusers. In all of the analyses, open-top cleaning systems show less annual cost of ownershipthan do airless enclosed cleaning systems. The authors do not have adequate experience


with airtight and externally sealed enclosed systems. However, because they require lessinvestment than airless enclosed systems, it seems reasonable, at worst case, to assumetheir cost of ownership is no different from that of open-top systems.

Hidden Costs

Not all costs are readily quantified.First, there are the environmental/regulatory costs. What does it cost in time and legal

fees to get a permit for an open-top system in locations where no permit is required for anenclosed airtight system because emissions are below a de minimis value? What does it saveto avoid the need for environmental monitoring, where it might not be required with anexternally sealed system? What if one must use PCE in a process, and only an airless sys-tem will meet the regulatory emissions requirements? Can the operating costs of an open-top system be justified if a solvent costing $15/lb is necessary?

Costs of quality are also difficult to quantify. What is vacuum drying worth, if it comes“free” with purchase of an airless system? Will an enclosed system allow one to match thesolvent precisely with the soil rather than compromise based on what can be readily con-tained in an open-top system? What is it worth to be able to clean and dry repeatedly a com-plex structure that would retain residual solvent when processed in a nonvacuum system?

WHY PURCHASE AN ENCLOSED SYSTEM?

That question puzzles users, especially since the cost of ownership (with low-pricedsolvents) is slightly higher for an enclosed system. Regulators easily find an acceptableanswer: to control solvent loss. However, opinions vary. The view of the authors is that anenclosed system is preferred over open-top systems for nearly all cleaning operations.

For low-price solvents, which pose no environmental concern, the open-top system isslightly more cost-effective. However, because most of these solvents dry very slowly, whynot purchase an airless enclosed system that provides vacuum drying? For low-price sol-vents, with environmental issues, the open-top system is slightly cheaper to operate. But,either now or in the near future some type of enclosed system is likely to be needed to meetenvironmental regulations Why not purchase an enclosed system and save the second pur-chase investment? For medium- and high-price solvents, enclosed systems are readily jus-tified on the basis of reduced total operating cost. What if an operation is located in an areaof poor air quality that is heavily regulated? An enclosed system is nearly certainly needed.Even where the predicted cost of ownership of an open-top system is less than that of anenclosed system, what about the hidden costs?

In summary, the strength of the case for the open-top cleaning system is diminishingrelative to the strength of the case for some type of enclosed cleaning system.

SUMMARY

The sole purpose of an enclosed system is to conduct some cleaning process within acontained environment. Enclosed cleaning systems are relatively new and not commonlyused commercially. They were developed in the United States for a few specific applica-tions that were difficult to compete satisfactorily with open-top cleaning systems.Operation with reduced emissions is the principal motivation for purchase. Other factors


include excellent drying performance and cleaning of unusual shapes or structures. Airlesssystems also have significant potential to conduct additional cleaning processes.

The three types of systems include airtight, airless (a vacuum chamber), and externallysealed (an open-top system with an airlock). The airless system is the most costly and issuitable for certain unique applications. The airtight system operates with reducedemissions, cleans components, and is priced intermediately. The externally sealed systemoperates with reduced emissions and is the lowest priced.

High purchase investment has restricted commercial adoption of enclosed cleaningsystems. The authors’ cost analysis indicates that total operating costs for enclosed systemsare close to those for open-top systems, even with low-cost solvents. With high-cost sol-vents, users rapidly recover the higher capital investment.

REFERENCES

1. Grant, D.C.H., Solvent Recovery and Reclamation System, U.S. patent 5,232,476, August 3, 1993.Assignee is Baxter International.

2. Grant, D.C.H., Method for Cleaning with a Volatile Solvent, U.S. patent 5,304,253, April 19, 1994.Assignee is Baxter International.

3. Gray, D.J. and Gebhard, P.T.E., Cleaning Method and System, U.S. patent 5,469,876, November28, 1995. Assignee is Serec.

4. Gray, D.J. and Gebhard, P.T.E., Solvent Cleaning System, U.S. patent 5,538,025, July 23, 1996.Assignee is Serec.

5. Tanaka, M. and Ichikawa, T., Cleaning System Using a Solvent, U.S. patent 5,193,560, March 16,1993. Assignee is Tiyoda.

6. Tanaka, M. and Ichikawa, T., Cleaning Method Using a Solvent While Preventing Discharge ofSolvent Vapors to the Environment, U.S. patent 5,051,135, September 24, 1991. Assignee isTiyoda.

7. Grant, D.C.H., Emission Control for Fluid Compositions Having Volatile Constituents, andMethod Thereof, U.S. patent 5,106,404, April 21, 1992. Assignee is Baxter International.

8. Turicco, T., Pressure Controlled Cleaning System, U.S. patent 5,449,010, September 12, 1995.9. Nafzifer, C.P., Single Chamber Cleaning, Rinsing, and Drying Apparatus, and Method Therefor,

U.S. patent 5301701, April 12, 1994. Assignee is Hyperflo.10. Grant, D.C.H., Vacuum Airlock for a Closed-Perimeter Solvent Conversation System, U.S. patent

5,343,885, September 6, 1995. Assignee is Baxter International.11. Basis: 5-year use of capital at 8% annual interest; cost components are capital use, solvent, oper-

ating labor, disposal, and miscellaneous, which includes power.12. Durkee, J.B., NESHAP recap—”Dr. PC” explains all, Precision Cleaning, April 1995, p. 39.13. Almodovar, P., U.S. EPA, personal communication, August 14, 1997.14. U.S. EPA, Guidance Document, Part 2, Section 1.2, 1995.15. Almodovar, P., U.S. EPA, personal communication, April 16, 1999.16. Office of Air Quality Planning and Standards at Research Triangle Park, Impact Analysis of the

Halogenated Solvent Cleaning NESHAP, U.S. EPA-453/D93-058, November 1993. This analysiswas based on 15-year recovery of capital cost and 10% cost of capital.


CHAPTER 2.12

Precision Cleaning and Drying UtilizingLow-Flash-Point Solvents

Matt Bartell

CONTENTS

Solvent OverviewOther SolventsCosolventsSolvent Costs

Process OverviewApplicationsEquipment Configuration

Heating SystemCooling SystemAutomationSolvent Containment

Safety FeaturesFinal Thoughts

SOLVENT OVERVIEW

A low-flash-point solvent is defined by the NFPA (National Fire ProtectionAssociation) as any solvent having a flash point below 100°F. Precision cleaning and dry-ing with low-flash-point solvents such as isopropyl alcohol (IPA), cyclohexane, and ace-tone can be a very effective cleaning strategy. However, properly configured equipmentand a well-designed cleaning process are essential to the successful and safe implementa-tion of low-flash-point solvent processes.

The three most common low-flash-point solvents used in precision cleaning processesare IPA, cyclohexane, and acetone. These solvents are extremely effective cleaning agents,and are priced well below engineered solvents such as azeotropes or blends of hydrofluo-rocarbons or hydrofluoroethers.

• IPA is typically used for the removal of particle contamination and inorganicfilms such as salts, fingerprints, and highly activated fluxes.


• Cyclohexane is effective in removing particle contamination and heavy organicfilms including oils and greases.

• Acetone has been proved effective in the removal of many types of inks andadhesives. Acetone also has an advantage in areas of poor air quality in that it isnot federally regulated as a volatile organic compound (VOC). Although a well-designed solvent cleaning system should lose only a very small amount of sol-vent to fugitive emissions, the VOC status of acetone is a distinct andenvironmentally friendly advantage, especially in highly regulated geographiclocations.

Other Solvents

An azeotrope of alcohol and cyclohexane has been demonstrated to be effective inremoving most contaminants, and is highly effective in removing resin-activated fluxes.Azeotropes are solvent mixtures that have a constant composition in the liquid and vaporphases over a certain temperature range, and so take on common cleaning properties. Oncecombined, they are inseparable via normal distillation during cleaning processes.

Many other solvents, such as heptane, ethanol, and volatile methyl siloxanes, may alsobe used in low-flash-point systems. Use of such solvents should first be discussed with theequipment manufacturer to ensure safe and effective operation.

Cosolvents

A low-flash-point solvent may be used to rinse a higher-boiling solvent within a singlespecially designed cleaning system. An example of such a process, deemed cosolvent,would be the use of N-methyl pyrillodone to remove thermally or ultraviolet cured adhe-sives from various substrates, followed by a rinse of IPA or acetone. These cosolventprocesses require specific temperature parameters and control to prevent the solvents frommixing within the cleaning system.

Solvent Costs

Solvent cost is typically $3 to 4/gal (IPA). Solvent waste may be incinerated for energygeneration. Disposal cost for IPA is typically $2 to 3/gal. This low solvent cost provides asignificant benefit, especially when compared to many of the engineered solvents currentlyavailable.

PROCESS OVERVIEW

The vaporization and condensation of the solvent act as the driving force to move sol-vent throughout the system. This process is commonly known as the “reflux cycle.” In awell-designed system, the dirtiest solvent is concentrated in an offset boil sump, and partsare never exposed to this offset boil sump. The condensed distillate drains via gravitythrough the immersion sump or sumps and into the offset boil sump. The final rinse anddry take place in a saturated superheated solvent vapor blanket. A superheated zone isdefined here to be a vapor zone maintained at approximately 20 to 50°F above the boilingtemperature of the solvent being used. Any liquid solvent remaining on the parts in thevapor zone will flash dry due to the superheated zone, allowing dry parts to emerge. This


Figure 1 Reflux of a solvent in a well-designed system.

TRANSPORT SYSTEM

INNERDOOR

PROCESSPART

OUTERDOOR

SUPERHEATER

SUPERHEATEDVAPOR ZONECONDENSING

COILS

SUBZEROFREEBOARD

COOLINGCOILS

OFFSET BOILSUMP

IMMERSION

SUMP 1

IMMERSION

SUMP 2

FREEBOARD ZONELOAD LOCKTM

ACCESSCHAMBER

VAPOR LINE

on-board distillation process not only provides the solvent vapor needed for complete dry-ing of the parts being processed, but also acts to clean the solvent continuously in theimmersion tanks. A well-designed system will reflux the solvent at a minimum rate of oneimmersion tank volume per hour. A schematic view of this type of process is shown inFigure 1.

A typical low-flash-point precision cleaning and drying process includes two solventimmersion tanks, one or both of which may include ultrasonics and filtered recirculationof solvent and a superheated vapor drying zone. Additional options to enhance cleaningperformance include spray under immersion, vertical basket oscillation, and a final sprayflush of distillate solvent.

APPLICATIONS

Low-flash-point solvent precision cleaning processes have been in use for many years.A multitude of companies have found solvents such as the azeotrope of IPA/cyclohexaneto be a low-cost and effective alternative to engineered solvent for precision cleaning oper-ations. Typical applications include the removal of contaminants from the following:

• Printed circuit boards/hybrid circuits/MCMs/C4 packages• Disk drive components• Precision mechanical/electromechanical components• Optical instruments• Medical devices and components

In addition to cleaning, low-flash-point solvents may be used to dewater parts followingaqueous cleaning. IPA is widely used throughout the computer hard disk industry to pro-vide spot-free drying following wet-bench cleaning processes.


EQUIPMENT CONFIGURATION

The safe and effective use of low-flash-point solvent requires careful attention to spe-cific system design criteria.

Heating System

Indirect heating systems hold a distinct advantage in that circulating hot waterthrough heat exchange coils in the boil sump and immersion tanks provides an effectivemeans of heating the solvent, without the possibility of exposing the solvent to a runawayheater element, either immersed or mounted to the side of the tank. Indirect heating elim-inates both the safety concerns and the possibility of thermally degrading the solventbecause of contact with the high surface temperature associated with most contact heaters.

Cooling System

Every BTU (British Thermal Unit) introduced into a solvent cleaning system must beremoved by the cooling system of the system. Two options for cooling media are chilledwater and refrigerant. Chilled water, typically run at 40°F, may be used to condense the sol-vent vapor, and is especially effective when solvents with a relatively high freezing pointare used.

Refrigerated systems provide a higher level of emissions control, but require a higherlevel of design expertise to handle the varying heat load encompassed within most preci-sion cleaning equipment.

Automation

The use of PLCs (programmable logic controllers) and mechanical transport systemsadds a valuable layer of process control and repeatability when compared with a manualsystem. A well-designed automated system will allow the operator simply to load the bas-ket into the system and start the cycle. From this point forward, all other machine actions,such as moving the baskets from tank to tank, and controlling system temperatures, shouldbe controlled by the system PLC. This is true for all cleaning systems; however with low-flash-point systems, design considerations are of particular importance.

The transport system itself must be designed to comply with all NFPA guidelines asso-ciated with the zone in which it is operating. Typically, the area immediately above theprocess tank will be classified as NFPA Class One, Division One. This will require all com-ponents used in this area to be classified as explosion-proof. Beyond the safety concernsassociated with the transport system, other points of consideration in the choice of an auto-mated system include payload capacity, maintainability, particle generation, and the over-all ruggedness of the design.

Solvent Containment

One of the most important considerations when choosing a solvent cleaning system isthe reduction of solvent emissions and operator solvent exposure. A well-designed solventsystem will be based on solid vapor degreaser principles, utilizing the reflux cycle of boil-ing and then condensing the solvent to minimize solvent vapor losses, combined with ahigh freeboard ratio. If solvent is used at ambient temperature, in an open-top immersion


tank with no saturated vapor blanket, the solvent will quickly evaporate. The use of the sol-vent reflux cycle greatly slows down this evaporation process. A well-designed “open top”cleaning system, engineered to use a low-flash-point solvent such as IPA can be expectedto emit approximately 1 gal/day.

In an attempt to improve on the already impressive emissions reduction results of theaforementioned open-top cleaning systems, many other methods of solvent containmenthave been incorporated into various systems. One particularly effective solution has beenthe addition of a sealed upper enclosure and load-lock loading chamber to the existingvapor degreaser type design. This overall machine design offers a single solvent emissionloss point based on the number of loads passing through the load-lock loading chamberper hour. This load-lock loading chamber acts as a true air lock, separating the process envi-ronment from the bay environment surrounding the cleaning system. This single exhaustpoint emits one load-lock volume of air contaminated with solvent vapors per basket. Thissolvent-contaminated air is directed to the facility exhaust stream where it may be furthertreated before being emitted to the environment. With a properly designed cleaning anddrying cycle, such a cleaning system can yield low solvent emissions of 1 lb/day.

SAFETY FEATURES

The single item differentiating a system designed for low-flash-point solvents fromother solvent-based cleaning systems is the level of safety features. When considering anysystem for use with these solvents, safety should be the first consideration.

The safety design of a well-designed low-flash-point system should be based on theprinciples found in the fire safety triangle. To support combustion, three ingredients mustbe present, fuel, ignition sources, and oxygen. Any system designed to comply with NFPAregulations must eliminate two of these components in all areas of the system. The systemswith the highest levels of safety will eliminate two of the three combustion requirements inall areas of the system.

As a minimum, the following safety features should be included in any systemintended for use with low-flash-point solvents.

• Compliance with all applicable NFPA guidelines.• Indirect heating system, to prevent heater “runaway.”• Electrical signals wired through intrinsic barriers, when located in the classified

area.• High-voltage items must be wired in sealed, explosion-proof conduits, in the

classified areas.• Safety checks in a given area of the machine should include redundant backups.• The secondary containment pan should include a monitored leak-detection sys-

tem.• An integral fire detection and CO2 fire suppression system should be installed as

part of the system.• Only systems bearing independent third-party approval, such as by Factory

Mutual Research Corporation, should be considered.

FINAL THOUGHTS

Although there has been a trend toward aqueous cleaning, use of solvents offersadvantages over aqueous cleaning in many applications. Some factors include a greatersolvency range for soils of interest such as activated solder flux, lower viscosity for


components with blind holes and complex geometries, improved spot-free drying foroptics, compatibility with readily oxidized metals, and low conductivity in cleaning motorwindings. In addition, compared with aqueous cleaning, which requires multiple rinsingand drying steps, solvent systems tend to have a smaller footprint. Finally, a well-designedlow-flash-point system minimizes solvent losses due to dragout and emissions, resultingin very low solvent consumption, compared with hundreds of gallons of deionized rinsewater usage and waste processing.

Familiar solvents such as IPA, cyclohexane, and acetone can solve a broad spectrum ofcleaning challenges. Examples include cleaning titanium medical components, defluxingcircuit boards, cleaning sputtering targets, water removal, industrial degreasing (TCEreplacement), debonding, and precision ink and lubricant removal.

Why do low-flash-point solvents work? Low-flash-point solvents have advantagesthat make them ideal for a broad spectrum of cleaning and drying challenges.

Effective solvency for a range of soils. Low-flash-point solvents efficiently removefluxes, oils, most inks and dyes, particulates, and other impurities.

Low cost. Low-flash-point solvents are very inexpensive ($2 to $3/gal) and have lowdisposal costs. Some can also be used as fuels.

Low toxicity. Many low-flash-point solvents are common chemistries found inhomes and businesses. This wide acceptance makes low-flash-point solvents anexcellent choice where there are considerations of employee safety and low usagerestrictions.

Low residual. Low residue (solvent or otherwise) on the part delivers spot-free dry-ing and a high level of cleanliness.

So why doesn’t everyone use low flashpoint solvents?There are two main objections to low-flash-point solvents. The first, surprisingly

enough, centers on price. “It can’t work; it’s too cheap.” The answer to this concern is thatlow-flash-point solvents do work (even if they are inexpensive).

The second centers on safety concerns. Although a flammable substance, the hazardsinvolved are virtually nonexistent with properly designed equipment and are comparableto driving a car fueled with gasoline.

Low-flash-point solvents are an excellent choice for many applications. Their estab-lished track record including excellent cleaning, low cost, low regulatory requirements,and accepted safety makes them the ideal choice for many applications. Give them a look;you may be pleasantly surprised with what you find.


CHAPTER 2.13

Dense-Phase CO2 as a Cleaning Solvent:Liquid CO2 and Supercritical CO2

William M. Nelson

CONTENTS

IntroductionNeed for New SolventsLiquid and Supercritical Carbon Dioxide

Carbon DioxidePhases of CO2

Liquid CO2

Supercritical CO2

Industrial CleaningCondensed-Phase CO2 Cleaning

Supercritical Fluid ExtractionSupercritical CO2

Engineering Condensed-Phase CO2 CleaningDetails of Condensed Phases

SolubilityCosolventsSubstrate CompatibilityEffect of Mixing on Supercritical CO2 Cleaning

EconomicsConclusionsReferences

INTRODUCTION

The use of carbon dioxide (CO2), either liquid CO2 (LCO2) or supercritical CO2 (SCCO2),cleaning has been shown through extensive laboratory and pilot testing to be potentialalternatives for manufacturers looking for new precision and parts cleaning systems. Theuse of carbon dioxide as an extraction solvent has slowly found acceptance, and also close


to commercialization are applications for supercritical and liquid CO2 in the electronics industry. Although much of the potential for supercritical or liquid carbon dioxide lies infairly exotic applications, CO2 solvent technology is already commercial in the compara-tively mundane fields of dry cleaning and spray painting. The technology is especially wellsuited for precision cleaning applications in which parts have intricate geometries or forapplications in which parts are sensitive to water or high temperature. CO2 cleaningappears to be compatible with a majority of substrates encountered in manufacturing,including most metal and glass, and many plastics. Mixing generally appears to improvethe efficiency of the CO2 cleaning process, as does the incorporation of cosolvents with con-densed-phase CO2.

Condensed-phase CO2 cleaning has relatively low operation and maintenance costs,and does not generate additional waste streams as a result of operations. These low oper-ation and maintenance costs are offset by a relatively high initial capital cost that has pre-vented the technology from being truly cost-competitive with other options. If the savingsfrom regulatory fees and/or wastewater treatment costs from manufacturing processes areconsidered, the economics become more appealing.

NEED FOR NEW SOLVENTS

Increasing concern regarding the dissemination of chemical waste (both aqueous andorganic) into the environment has prompted considerable interest in new technologiesaimed at reducing current waste streams. Cleaning technologies (preparing the surface ofa material for subsequent steps in an industrial process) can have a significant effect on itsoverall environmental impact.

Over the past decade, concern for the environment, economic competitiveness, andtechnological advances all converged to cause both industry and government to reevalu-ate manufacturing processes. Changes from traditional solvent cleaning to alternativemethods set into motion recent trends toward zero discharge of pollutants into the air,water, and soil. With ozone-depleting chemicals (ODCs) being phased out, many manu-facturers are struggling to find efficient and effective replacement solvents and cleaningagents. The ODC phaseout and a host of other environmental and safety concerns haveprompted the development of alternative cleaning agents. However, there are no drop-insubstitutes for ODCs.

The quality and suitability of the cleaning process is heavily dependent upon the qual-ity of the solvent utilized. The solvent is either an active agent in the process or is the stageon which the process occurs.1 Solvents “clean” by producing species from contaminantsthat have a higher affinity for the cleaner than for the surface to which they adhere. In sodoing, the cleaning agent can separate the contaminant from the surface.

Solvent use in cleaning will continue to be pervasive. The challenge to the cleaningindustry will be to adopt the most environmentally benign and efficacious technology.Probably one of the most critical components of a cleaning process is the identity of the sol-vent(s) used. The available solvents have given individuals in the cleaning industry a greatset of tools. This is, however, only the beginning. Careful reasoning must enter into thechoice of cleaning technologies. Condensed-phase CO2 is a true solvent and it can serve asthe dissolving media for cleaning processes. This chapter examines: (1) the solubility char-acter of CO2 in its liquid and supercritical phases, (2) the use of cosolvents and entrainersto enhance solubility, and (3) the equipment and engineering necessary to implement thiscleaning technology. Within the discussion, references to case studies and test results areprovided.


LIQUID AND SUPERCRITICAL CARBON DIOXIDE

Carbon Dioxide

Carbon dioxide is a colorless gas, which was first recognized in 1577 by Van Helmontwho detected it in the products of both fermentation and charcoal burning. CO2 is used insolid (dry ice), liquid, and gaseous form in a variety of industrial applications such as bev-erage carbonation, welding, chemicals manufacture, and cleaning. It occurs in the productsof combustion of all carbonaceous fuels and can be recovered from them in a variety ofways. CO2 is also a product of animal metabolism and is important in the life cycles of bothanimals and plants. It is present in the atmosphere in small quantities (0.03%, by volume).

CO2 is not very reactive at normal temperatures. It does, however, form carbonic acid,H2CO3, in aqueous solution. This will undergo the typical reactions of a weak acid to formsalts and esters. A solid hydrate, CO2

. 8H2O separates from aqueous solutions of CO2 thatare chilled at elevated pressures. It is very stable at normal temperatures but forms CO andO2 when heated above 1700°C.

CO2 has several advantages: environmental acceptability, nonflammability, and non-corrosivity. Additionally, CO2 has no ozone-depletion potential, and while it does havesome global warming potential, its use in cleaning operations would contribute insignifi-cantly to global warming in comparison with, for example, automobile emissions, coal-burning electric generation, steel smelting, etc. In addition, as Frank Cano points out inChapter 2.14 on CO2 snow cleaning, commercially produced CO2 is recycled from otherindustrial processes and repesents a delayed emission rather than a new source.

Condensed-phase CO2 cleaning systems assume several forms. Cleaning with CO2 isadvantageous in that, after cleaning, the only waste streams generated are the isolated con-taminants that were removed from the part that was cleaned. There are no large, liquidstreams to treat (as there are with aqueous cleaning) or airstreams to treat (as is the casewith some solvent cleaning solutions).

Phases of CO2

A pressure–temperature (P–T) phase diagram, shown in Figure 1*, illustrates the phasechanges of CO2, where the three phases of solid, gas, and liquid are indicated.

The substance remains a liquid, as long as the temperature and pressure fall within theCO2(1) region. CO2 is supercritical when its pressure and temperature are beyond the crit-ical point. Notice that the triple point of carbon dioxide is well above 1 atm. Notice also thatat 1 atm CO2 can only be the solid or the gas. Liquid CO2 does not exist at 1 atm. Dry ice(solid CO2) has a temperature of �78.5°C at room pressure, which is why one can get a seri-ous burn (actually frostbite) from holding it.

Liquid CO2

Although CO2 liquid does not exist at normal room pressures, it does exist at slightlyelevated pressure. A laboratory cylinder of CO2 will contain LCO2 at a pressure of about 75psi at room temperature. On a particularly hot day (above 88°F) the liquid CO2 will passthough its critical point and the contents of the cylinder will exist as the supercritical fluid.

The LCO2 cleaning technology used alone under these conditions is a solvent muchlike room-temperature 1,1,1-trichloroethane (TCA), As such, it will remove many butnot all types of contaminants. Contaminants that are not soluble in LCO2 alone can be

*Chapter 2.13 Color Figure 1 follows page 104.


solubilized or otherwise separated by employing proprietary additives, modifiers, ormechanical adjuncts in the process.2

Liquefied CO2 can be used as a solvent for cleaning. Process temperatures generallyrange between 50 and 70°F, and process pressures range from 750 to 1200 psi. The area inthe CO2 phase diagram where the substance is liquid is shown in Figure 1.

Finally, the LCO2 immersion cleaning process can meet a variety of cleanliness require-ments, ranging from visually clean to more rigorous quality standards requiring suchsophisticated test methods as nonvolatile residue analysis, infrared spectroscopy, or scan-ning electron microscopy. Functional testing, such as that measuring weld joint porosityand adhesive strength, has also been conducted to evaluate the technology.

Supercritical CO2

Gases become “supercritical” when they are heated above their critical temperature—the point beyond which they cannot be liquefied—and compressed (see Figure 1). CO2

becomes supercritical at temperatures above 87.8°F (31.1°C) and pressures above 1072 psi(73.8 bar).3 Supercritical CO2 applications typically operate at temperatures between 90 and120°F (32 and 49°C) and pressures between 1070 and 3500 psi.

There are two unique points on this phase diagram. The lower point is called the “triplepoint” and is the unique combination of temperature and pressure at which all three phasesexist simultaneously. The easiest way to imagine this is to think of boiling ice water. If weremember that the word boiling has nothing to do with “hot,” then it is easy to imagine low-ering the pressure far enough to have ice water boil. In fact, the pressure need only be about4.58 mm, which is quite easy for a simple vacuum pump. Notice that this temperature isslightly above the normal melting point because of the retrograde nature of the meltingpoint curve.

The second unique point is stranger than the first. It can be shown experimentally thatfor every liquid there is a point along the boiling point curve where the line between theliquid and gaseous phases disappears. This is called the critical point. At temperatureshigher than this point, we can no longer think of two phases; there is only a single phasethat is a very dense gas, or frequently called a critical fluid. Another way of thinking aboutthis is to remember that at the critical temperature or above we can no longer compress thematerial to a liquid no matter how much pressure we apply. These criticalfluids have extremely unique properties and are now used for many commercial processes.For example, supercritical CO2 is used for the extraction of caffeine from coffee and tea.

Industrial Cleaning

A cleaner, by definition, removes dirt or other extraneous material from a surface. Aneffective cleaner is able to perform its task because of its ability to:

• Wet the surface• Penetrate the soil• Lift and remove the soil• Hold soil in suspension (so a surface can be wiped or rinsed)

Cleaning produces a clean surface that contains no significant amounts of undesiredmaterial. The degree of surface cleanliness must meet the following two criteria: (1) it mustbe sufficient for subsequent processing, and (2) it must be sufficient to ensure the future


Table 1 Drivers Influencing Adoption of Condensed-Phase CO2 Cleaning

Driver Description

Environmental The environmental problems associated with common industrial solvents(mostly chlorinated hydrocarbons)

Economic The increasing cost of regulated solvent useTechnological The inability of traditional techniques to provide the necessary separations

needed for emerging new industries (microelectronics, biotechnology, etc.)

reliability of the device or system. Beyond this, there are numerous factors that affect thequality of the cleanliness. In terms of cleaning, a solvent is a substance, single or multi-component, capable of dissolving other substances to form a homogeneous system (solu-tion). The criteria for what determines a good solvent will vary, depending upon the useand ultimate level of cleanliness sought during the process. The development of knowl-edge of solutions reflects to some extent the development of chemistry itself. A solvent maybe defined in rough terms as any liquid that serves as a carrier for another substance or asa means of extracting or separating other substances.

In practice, many solvents are mixtures rather than pure compounds. The most com-mon solvent is water, but next in importance comes a group of organic liquids and theirmixtures. At present, there is a concern for use of environmentally benign solvents, andthere is increasing interest in alternatives to the more traditional solvents. Precision andparts cleaning of manufactured metal parts has relied heavily on the use of conventionalchlorofluorocarbon (CFC) solvents.

Condensed-Phase CO2 Cleaning

The growth in interest in liquid and supercritical CO2 in industrial applications overthe past two decades has resulted from several key characteristics, which are relevant toboth academic and industrial communities. Three drivers, or forces, have contributed tothe recent attention given to these solvents (Table 1). The availability of inexpensive, non-toxic solvents such as liquid or supercritical CO2 and their attractive properties hasrenewed interest in the applicability of these solvents, especially in the area of cleaning.

Supercritical Fluid Extraction

Supercritical fluid (SCF) extraction has been developed extensively as both a cleanuptechnique and as an analytical technique for liquid and solid environmental samples.4,5 Wecan regard cleaning with condensed-phase CO2 as an extraction process: the extraction ofcontaminants from the surface of interest. Condensed CO2, when regarded as a solvent,may benefit from the enormous data of the science of solvents.6

SCF extraction (SFE), like any cleaning technology, works better on certain classes ofsoils. SFE has been employed successfully to remove oils such as hydrocarbons, esters, sil-icones, perfluoropolyethers, halocarbon substituted triazines, etc. SFE is an elegant clean-ing method with many environmental benefits, but one that is not at present cost-effective.

Metal cleaning in an industrial setting is typically an environmentally hazardous activ-ity. Until recently, vapor degreasers utilizing halogenated solvents were prevalent through-out industry. It is readily apparent that the reliance of industry on halogenated solvents asmetal cleaners has resulted in a technologically easy way to clean metals. A change to a


Table 2 Contaminants NotRemoved by Supercritical CO2

• Rust• Scale• Lint or dust• Ionic Species• Metal salts• Many (but not all) fluxes

different technology will necessitate some adaptations. As in any reaction medium it mustbe engineered for greater productivity by manipulating the temperature and pressure.

Supercritical CO2

While liquid CO2 has come to see increased use,2 further discussion of supercritical CO2

will illuminate characteristics of condensed-phase CO2. An extensive discussion of super-critical media is to be found in the review by Savage et al.7 SCFs are effective cleaningagents because of their ability to penetrate substrates and small interstitial spaces rapidly.After dissolving any contaminants, the critical fluid is easily and completely removedbecause it lacks surface tension. The critical fluid of choice for surface cleaning applicationsis most often CO2, either pure or in combination with a small amount of cosolvent. Solventproperties can be adjusted by small changes in temperature and pressure, allowing CO2 todissolve a range of organic compounds.

Supercritical CO2 actually has physical properties somewhere between those of a liq-uid and a gas. SCFs are able to spread out along a surface more easily than a true liquidbecause they have lower surface tensions than liquids. At the same time, an SCF maintainsthe ability of a liquid to dissolve substances that are soluble in the compound, which a gascannot do. In the case of supercritical CO2, this means oil and other organic contaminantscan be removed from a surface even if the surface has an intricate geometry or includescracks and crevices. In general, this process cannot remove contaminants that do not dis-solve in CO2 (Table 2).

It has been shown that high-pressure supercritical CO2 shows potential as a cleaningmedium for removing hydrocarbon machine coolants from metal substrates.8 In addition,supercritical CO2 alone can be tuned to remove the contaminants listed in Table 3.

As was mentioned previously, these condensed-phase fluids have both liquid and gas-like properties. This property will confer highly desirably solubility characteristics. Thisallows the liquid or fluid to penetrate very small gaps and complex assemblies, which willenhance the potential range of substrate geometries (Table 4).

The supercritical phenomenon has been known since the middle of the 19th century,but industrial applications—mostly extraction based—emerged only in the early 1980s.Today, a number of companies operate extraction facilities based on supercritical CO2.These include General Foods, which runs a coffee decaffeination plant in Houston; com-panies in Washington’s Yakima Valley, such as Yakima Chief and John I. Haas, whichextract flavor from hops; and natural flavor producers such as Finland’s Cultor andGermany’s SKW Trostberg.9 As the range of chemicals soluble in CO2 expands, wider fron-tiers will become available for supercritical and liquid CO2. Chemical manufacturing usesof CO2, as well as cleaning and materials processing applications, are likely to become theinitial industrial-scale uses of condensed-phase CO2.


Table 3 ContaminantsRemoved by Supercritical CO2

• Silicone oils• Flux residues• Petroleum oils• Machining oils• Dielectric oils• Lubricants• Adhesive residues• Plasticizers• Fats and waxes

Table 4 Examples ofSubstrates for Supercritical CO2

Cleaning

• Missile gyroscopes• Accelerometers• Thermal switches• Nuclear valve seals• Electromechanical assemblies• Polymeric containers• Special camera lenses• Laser optics components• Porous ceramics• Metal parts

ENGINEERING CONDENSED-PHASE CO2 CLEANING

To achieve condensed-phase CO2, the CO2 gas needs to be pressurized and heated (seeFigure 1). Parts are placed into a pressure vessel into which CO2 gas is introduced. The tem-perature and pressure are then raised until the supercritical state is reached. A basic systemconsists of six components.

1. Compressor2. Heat exchanger (heating)3. Extraction vessel (pressure vessel)4. Pressure control valve (expansion)5. Heat exchanger (cooling)6. Separation vessel

Figure 2 shows the basic components that comprise a condensed-phase CO2 cleaningsystem. CO2, which may be stored as a gas or in liquid form, is compressed above its liquidor critical pressure by a pump. The compressed CO2 is then heated to its liquid phase or toabove its critical temperature in a heater, or sometimes in the cleaning chamber. Any partsin the cleaning chamber are cleaned by exposure to the liquid or fluid. Typically, the clean-ing chamber will include an impeller to promote mixing. Further work is being done on


cosolvents and entrainers, which will enhance the solubility characteristics of CO2 (eitherliquid or supercritical).

Condensed-phase CO2 (containing dissolved contaminants) is then bled off to a sepa-rator vessel, where the liquid/fluid is decompressed and returned to a gaseous state. Thecontaminants remain in liquid form and are collected out the bottom of the separator, whilethe gaseous CO2 is sent through a chiller to return it to a liquid form for storage to be reusedagain. This closed-loop recycling of the CO2 means only a small portion of the cleaningsolution has to be replaced over time due to system leakage. The now-clean parts can beremoved from the chamber and are usually immediately ready for the next step in the man-ufacturing process, since no drying or rinsing is required to remove residual cleaning solu-tion. With some plastics, which can absorb CO2, a bakeout may be needed.

Process temperatures may range from 95 to 149°F (35 to 65°C). Pressures vary fromabout 1070 to 4000 psi. Nonmetallic materials must be tested for compatibility. The processworks well for removing trace fluids. Some suppliers also claim effective removal of parti-cle contamination. It may be possible to fine-tune the operating pressure and temperatureto match the soil being removed. Parts that cannot be subjected to elevated atmosphericpressures cannot be cleaned with SCFs. This process has been developed for the precisioncleaning industry. With further development it may become more broadly applicable.

DETAILS OF CONDENSED PHASES

Solubility

An experimental determination of the solubilities of solid mixtures in supercritical flu-ids was made.10 Examples of the solubility measurements are becoming more common inthe scientific literature (Table 5).

Cosolvents

If condensed-phase CO2 is going to be implemented on a commercial scale to removecontaminants from material surfaces, the costs associated with this process must bereduced and the range of solubilities must be broadened. These goals can be made more

Figure 2 Components of a condensed-phase CO2 cleaning system.

Heatingunit

Pump LiquidCO2

Coolingunit

Separator

Cleaningvessel CO2

Recycle


attainable with the use of an entrainer (or cosolvent) to increase the solubility of the solutein the supercritical phase, reducing the size of the required extractor and/or lowering thepressure needed to effect the desired extraction.

An example of this is the cleanup of soils contaminated with organics by extractionwith supercritical CO2. It can be influenced decisively by additional substances or entrain-ers. In most cases, the contaminated soil already contains water as a substance that can alterthe extractability of these contaminants. In particular, the effects of soil moisture, as a kindof discontinuous addition of water, on the extraction of polycyclic aromatic hydrocarbons(PAHs) from soil with supercritical CO2 were examined. On the other hand, humidifyingthe supercritical CO2 used a continuous addition of water by humidifying the supercriticalCO2. The improvement of the extraction yield by moisture indicates additionally that theextraction is limited by adsorption and not by diffusion effects. However, the contaminantis more accessible and is transported faster out of the soil with water.18

Substrate Compatibility

While case studies from successful industry implementation of supercritical CO2 clean-ing are helpful, the number of successful industrial case studies for supercritical CO2 clean-ing is small. Although the majority of published matter is laboratory- or pilot-scale testresults, the lessons learned from true, long-term implementation of a technology arealways the most valuable to others considering applying the technology themselves(Table 6).

Although test results show the technology works quite effectively in many cases, thereare some published case studies from industry that resulted in decisions not to pursue thetechnology for full-scale application. In one instance, tests of supercritical CO2 cleaningwere performed for a manufacturer on metal disks contaminated with oil-type residues ina one-liter system at 180°F (82°C) and 2000 psi.28

• The cleaning was acceptable (below the specified “clean” level and comparablewith what a trichloroethylene vapor degreaser could obtain).

• Carbon residue on parts was reduced by 50%.• Mixing did not improve the cleaning efficiency in this test.

Table 5 Examples of Solubility Measurements in Supercritical CO2

Issue Result Ref.

Improving polymer solubility Surfactants improved solubility 11Low solubility of organophosphorous All compounds solubility improved 12

compoundsSolubilities of amorphous polymers Describes conditions for solubility 13Separation of vegetable cuticular Enhanced separation 14

waxesEffect of cosolvents Measured effects 15Determine effects of methanol as Measured effects 16

cosolventDetermine effects of n-pentanol as Measured effects 17

cosolvent


Table 6 Examples of Compatibility Tests

Issue Result Ref.

Concerns over supercritical CO2 Supercritical CO2 cleaning can be 19cleaning polymers adjusted to have no detrimental

effect on crystalline polymersOil removal from rings, washers, and Supercritical CO2 cleaning removed 20

plates 97 to 99.95% of the oilDifferent substrates, including Supercritical CO2 good for water 21

aluminum, glass, copper, brass, sensitive or high-temperature-stainless steel, and epoxy boards sensitive parts

Soils and other solid materials, Supercritical fluids can be applied to 22containing residual pesticides these substrates

Various metals, plastics, and Supercritical CO2 used; excellent 23epoxies needed to be precision payback and �90% reductioncleaned, using CFC-113 in ODS

Vapor degreaser used to clean Supercritical CO2 works well for 24gyroscope parts contaminated parts with complex shapeswith machining coolants, siliconeoils, and damping fluids

High cleanliness standards required Supercritical CO2 shown to work 4, 8, 24–27for precision cleaning more efficiently for contaminated

plastic than cleaning withCFC-113

Replace perchloroethylene (PERC) Supercritical CO2 is equally effective 9as a dry cleaning solvent

• Supercritical CO2 cleaning was not pursued because of the predicted long pay-back time (10 years) of the investment, and the high operating pressure of the sys-tem (which was viewed as a safety hazard).

In another study, AT&T researchers tested a number of different cleaning alternatives toreplace 1,1,1-trichloroethane vapor degreasing.

• Aluminum parts contaminated with cutting oils and protective fluids were tested.• Supercritical CO2 cleaning efficiency was more than 98% for several of the con-

taminants.• The requirement for additional additives and cosolvents to obtain acceptable

cleaning results led the researchers to eliminate supercritical CO2 from consider-ation.29

As can be seen from these examples, while supercritical CO2 cleaning appears to beeffective in a number of instances from a technological standpoint, some technical limita-tions combined with the economies of the process, which will be discussed more below,have resulted in a slow rate of implementation in the private industrial sector.

Effect of Mixing on Supercritical CO2 Cleaning

Test results indicate that increasing the internal agitation and temperature in the clean-ing chamber reduces the time needed to clean the metal parts. The supercritical cleaning


process, when run at optimum conditions, appears to use less energy than conventionalvapor degreasing operations. Furthermore, cleaning results attained with supercritical CO2

plus mixing compare favorably with conventional solvent cleaning.27

Two studies that discuss the effect of mixing on supercritical CO2 cleaning resulted insomewhat different conclusions. As mentioned earlier, a manufacturer of metal disks wasnot able to determine that mixing had any effect on cleaning efficiency, but that a largersample size may have given a more definitive result.28

A more in-depth study specifically designed to examine the effect of fluid turbulenceon supercritical CO2 cleaning concluded that mixing has an effect. Researchers at PacificNorthwest National Laboratory completed the study. The researchers recommended thatagitation be used whenever possible for supercritical CO2 cleaning applications to helpmaximize cleaning efficiency, and that mixing rates can be optimized to minimize powercosts.30

Previous studies by Phasex Corporation, the company with the longest experiencewith supercritical fluids for cleaning, have demonstrated that supercritical CO2 is an excel-lent solvent for oils such as hydrocarbons, esters, silicones, perfluorpolyether, halocarbon-substituted triazines, and organosilicones with various reactive functionalities; many ofthese oils are associated with the manufacture of precision components such as gyroscopesand accelerometers.4 The ability to dissolve a particular oil or polymer at any given pres-sure will greatly depend on the molecular weight and structure of the material. The abilityof supercritical fluids to dissolve many types of oils and organic materials, coupled withthe ability to penetrate minuscule pores and interstices of metal, ceramic, and compositeparts, suggests that these fluids could partially replace CFCs.

As the use of CFCs and other halogenated solvents is phased out because of their rolein stratospheric ozone depletion, condensed-phase CO2 is the most promising of the alter-native technologies because of its low cost, low toxicity, nonflammability, and environ-mental acceptability. The CO2 cleaning process, the CO2 cleaning system operation,technical factors and life-cycle costs, commercial applications, and future developmentshave been considered.4

ECONOMICS

As has been pointed out, the economics behind condensed-phase CO2 use in cleaninghas been largely prohibitive when applied in the cleaning industry. Equipment for super-critical fluids cleaning tends to be costly, and process development is very application-specific.31

Researchers at Pacific Northwest National Laboratory during a supercritical CO2

cleaning market assessment completed in 1994 identified some of the reasons for lack ofdemand in the private sector for supercritical CO2 cleaning (Table 7).32

The initial cost of a condensed-phase CO2 system might seem prohibitive, but when theenvironmental benefits (health and safety and regulatory) and the energy savings overallare considered, the return-on-investment looks rosier. The high capital cost of supercriticalCO2 cleaning systems can be attributed, in part, to the high-pressure cleaning chamber andthe valves and instrumentation required for the system. The system cost is also highbecause there are no vendors that mass-produce supercritical CO2 systems. Lack of massproduction does not allow vendors to realize the economies of scale that could be obtainedif demand for the system were higher. Unless demand increases, it will be difficult toreduce the purchase costs to a point where the basic payback for supercritical CO2 cleaningis rapid enough to attract a significant number of manufacturers.


Table 7 Potential Barriers to Acceptance of Supercritical CO2

• Higher capital costs for supercritical CO2 systems relative to other cleaning technologies• Lack of awareness of supercritical CO2 cleaning technology• Substrate to be cleaned lacks compatibility with supercritical CO2 or with high pressures• A perception that supercritical CO2 cleaning does not remove particulates effectively• The requirement for a continuous process (supercritical CO2 cleaning is a batch process)• The existence of established aqueous-cleaning technologies to replace solvent vapor degreasers

CO2 cleaning systems have lower capital costs, require less labor, do not release anyCFCs, take up less space, and, above all, the solvents cost less. The time required for clean-ing and stripping processes can be reduced by as much as 80 to 90%.

Published investigations into the economics of supercritical CO2 cleaning show thatoperational costs of supercritical CO2 cleaning are quite reasonable and often lower thansolvent vapor degreasers or aqueous cleaning systems. CO2 is relatively inexpensive andcan be reused in most supercritical CO2 cleaning systems. Waste treatment is minimal, asno waste stream is generated beyond the actual contaminants removed from the partsbeing cleaned. Unfortunately, the initial capital costs for supercritical CO2 systems are usu-ally higher than other alternatives—sometimes by a significant amount. Findings on eco-nomics include:

• Researchers at Los Alamos Laboratory conducted a study to measure the totalelectrical costs associated with operation of a 60-l capacity supercritical CO2 sys-tem at temperatures between 30 and 50°C and pressures between 1500 and 3500psi. The researchers concluded that utility costs for the unit will be relativelyinsignificant when compared with operations and maintenance labor costs.33

• An economic analysis done after bench-scale testing based on cleaning 150 bear-ings/year showed annual savings of only $1400 on a $75,000 to $100,000 invest-ment. It appears that a much higher volume application would be required tojustify the supercritical CO2 system.34

CONCLUSIONS

In terms of the triad of drivers affecting the adoption of any cleaning technology (tech-nological, regulatory, and economic), condensed-phase CO2 is favorably positioned in allbut the economic arena. As the economics of this cleaning technology become more accept-able, cleaning with liquid and/or supercritical CO2 will become more prevalent.

REFERENCES

1. Nelson, W.M., Art in science: utility of solvents in green chemistry, in Green Chemistry: Frontiersin Benign Chemical Syntheses and Processes, Anastas, P.T. and Williamson, T.C., Eds., OxfordUniversity Press, Oxford, UK, 1998, 200.

2. Jackson, D. and Carver, B., Liquid CO2 immersion cleaning, Parts Cleaning, April, 1999.3. Smith, J.M. and Van Ness, H.C., Introduction to Chemical Engineering Thermodynamics, 4th ed.,

McGraw-Hill, New York, 1975, 54.4. Gallagher, P.M. and Krukonis, V.J., Precision parts cleaning with supercritical carbon dioxide, in

Solvent Substitution for Pollution Prevention, U.S. Department of Energy and U.S. Air Force,Noyes Data Corporation, Park Ridge, NJ, 1993, 76.


5. Yamauchi, Y., Supercritical fluid extraction and chromatography, J. Syn. Org. Chem. Jpn., 54, 395,1996.

6. Connors, K.A., Chemical Kinetics: The Study of Reaction Rates in Solution, VCH Publishers, Inc.,New York, 1990, 480.

7. Savage, P.E., Gopalan, S., Mizan, T.I., Martino, C.J., and Brock, E.E., Reactions at supercritical con-ditions—applications and fundamentals, AICHE J., 41, 1723, 1995.

8. Salerno, R.F., High pressure supercritical carbon dioxide efficiency in removing hydrocarbonmachine coolants from metal coupons and components parts, in Solvent Substitution forPollution Prevention, U.S. Department of Energy and U.S. Air Force, Noyes Data Corporation,Park Ridge, NJ, 1993, 98.

9. McCoy, M., Industry intrigued by CO2 as solvent: “green” processes based on supercritical car-bon dioxide are moving out of the lab, Chem. Eng. News, June 14, 11, 1999.

10. Liu, G.T. and Nagahama, K., Solubility of organic solid mixture in supercritical fluids, J.Supercritical Fluids, 9, 152, 1996.

11. McClain, J.B., Betts, D.E., and Canelas, D.A., Design of nonionic surfactants for supercritical car-bon dioxide, Science, 274, 2049, 1996.

12. Meguro, Y., Iso, S., and Sasaki T., Solubility of organophosphorus metal extractants in supercrit-ical carbon dioxide, Anal. Chem., 70, 774, 1998.

13. O’Neill, M.L., Cao, Q., Fang, R., Johnston, K.P., Wilkinson, S.P., Smith, C.D., Kerschner, J.L., andJureller, S.H., Solubility of homopolymers and copolymers in carbon dioxide. Ind. Eng. Chem.Res., 37, 3067, 1998.

14. Stassi, A. and Schiraldi, A., Solubility of vegetable cuticular waxes in supercritical CO2 isother-mal calorimetry investigations, Thermochim Acta, 246, 417, 1994.

15. Anitescu, G. and Tavlarides, L.L., Solubilities of solids in supercritical fluids 2. Polycyclic aro-matic hydrocarbons (PAHs) plus CO2/cosolvent, J. Supercritical Fluids, 11, 37, 1997.

16. Souvignet, I. and Olesik, S.V., Solvent-solvent and solute-solvent interactions in liquidmethanol/carbon dioxide mixtures, J. Phys. Chem., 99, 16800, 1995.

17. McFann, G.J., Johnston, K.P., and Howdle, S. M., Solubilization in nonionic reverse micelles incarbon dioxide, AICHE J., 40, 543, 1994.

18. Schleussinger, A.O. and Ingo, B.R., Moisture effects on the cleanup of PAH-contaminated soilwith dense carbon dioxide, Environ. Sci. Technol., 30, 3199, 1996.

19. Sawan, S.P., Evaluation of the Interactions between Supercritical Carbon Dioxide and PolymericMaterials, Los Alamos National Laboratory, Los Alamos, NM, 1994.

20. Novak, R.A., Cleaning of precision components with supercritical carbon dioxide, InternationalCFC and Halon Alternatives Conference, U.S. Environmental Protection Agency, Washington,D.C., 1993.

21. Williams, S.B., Elimination of Solvents and Waste by Using Supercritical Carbon Dioxide inPrecision Cleaning, LA-UR-94-3313, Los Alamos National Laboratory, Los Alamos, NM, 1994.

22. Knez, Z., Riznerhras, A., Kokot, K., and Bauman, D., Solubility of some solid triazine herbicidesin supercritical carbon dioxide, Fluid Phase Equilibria, 152, 95, 1998.

23. Hunt, D., How one of the largest Air Force users is getting out of CFCs, in Proceedings of the 1992International CFC and Halon Alternatives Conference: Stratospheric Ozone Protection for the 90’s,Washington, D.C., 1992.

24. Weber, D.C., McGovern, W.E., and Moses, J., Precision surface cleaning with supercritical carbondioxide: issues, experience, and prospects, Metal Finishing, 93, 22, 1995.

25. McGovern, W.E., Moses, J.M., and Weber, D.C., The use of supercritical carbon dioxide as analternative for chlorofluorocarbon (CFC) solvents in precision parts cleaning applications, inProceedings Air Pollution Control Association Annual Meeting, 1994.

26. Purtrell, R., Rothman, L., Eldridge, B., and Chess, C., Precision parts cleaning using supercriticalfluids, J. Vac. Sci. Technol., 11, 1696, 1993.

27. Silva, L.J., Supercritical fluid for cleaning metal parts, Haz. Waste Consultant, 13, 1.25, 1995.28. Supercritical Fluid Extraction Cleaner Application: Texas Instruments Incorporated, Toxics Use

Reduction Institute, Lowell, MA, 1994.


29. Gillum, W.O., Replacement of chlorinated solvents for metal parts cleaning, in Precision Cleaning‘94, Witter Publishing, Rosemont, IL, 1994.

30. Phelps, M.R., Waste Reduction Using Carbon Dioxide: A Solvent Substitute for PrecisionCleaning Applications, Pacific Northwest National Laboratory, Richland, WA, 1994.

31. Kanegsberg, B., Precision cleaning without ozone depleting chemicals, Chem. Ind., 787, 1996.32. Pacific Northwest National Laboratory, Richland, WA, 1994.33. Barton, J.C., The Los Alamos Super Scrub: Supercritical Carbon Dioxide System Utilities and

Consumables Study, Los Alamos National Laboratory, Los Alamos, NM, 1994.34. Licis, I.J., Pollution Prevention Possibilities for Small and Medium-Sized Industries—Results of

the WRITE Projects, U.S. Environmental Protection Agency, Washington, D.C., May, 1995, 127.


CHAPTER 2.14

Carbon Dioxide Dry Ice Snow Cleaning

Frank Cano

CONTENTS

IntroductionPropertiesCleaning with CO2

Momentum TransferSolvent EffectThermophoresisFreeze-Fracture EffectApplicationsCondensationElectrostatic DischargeRedistribution of ContaminantsThermal ShockToo Much Blast for Sensitive PartsSurface PropertiesCost of CO2 CleaningSafety/EnvironmentalConclusion

INTRODUCTION

Jan Baptist van Helmont, the 17th-century Belgian chemist who first recognized carbondioxide, could never have imagined the varied applications of carbon dioxide today. Inchemical extraction, decaffeination of coffee, carbonation of beverages, uses as a refriger-ant, use in fire extinguishers, use in cleaning applications, and many other commercialuses, carbon dioxide has important applications in the modern world.

PROPERTIES

Carbon dioxide is a minor component of Earth and makes up about 0.04% of the totalatmosphere. Plants employ it in the photosynthesis process. CO2 is formed in various ways:


in the combustion of carbon-containing materials, in fermentation, and in the respirationof animals. Commercially it is recovered from flue gasses, as a by-product in the synthesisof ammonia, from the cracking of petroleum, and other processes.

CO2 is colorless in its liquid and vapor state, is about 53% heavier than air, and will dis-place oxygen. Heavy concentrations must be avoided in areas not sufficiently ventilated.Asphyxiation is a real danger when CO2 is used in large amounts in confined areas. It isnontoxic, which makes it ideal for many commercial uses. CO2 is inert at normal atmos-pheric temperatures; however, potassium will burn violently if heated in a CO2 atmos-phere. Under specific conditions, pressurized CO2 can combine with water to formcarbonic acid (H2CO3), a relatively weak dissociated acid.

CO2 can exist in three phases—vapor, liquid, or solid. The state of CO2 is dependentupon the temperature and the pressure at which it exists. At atmospheric temperatures andpressure CO2 exists as a gas but when cooled and pressurized it converts to its liquid form.Liquid CO2 cannot exist at atmospheric pressures. At 21°C (69.8°F) it must be compressedto 850 psia (58.6 bar) to remain a liquid. Thus, pressurized cylinders of CO2 can be storedfor an indefinite time at normal room temperature and still maintain a liquid phase. Thehighest temperature at which liquid CO2 can exist is 31.1°C (87.9°F). This is known as its“critical temperature.” The highest pressure at which CO2 will liquefy is 1070.6 psia (73.8bar). This is known as its “critical pressure.”

Below �56.7°C (�69.8°F) and below 75 psia (5.2 bar) CO2 exists as a solid. The condi-tion at which all three forms of CO2 exist simultaneously is known as its “triple point.” Thisphenomenon occurs when CO2 is at �56.7°C (�69.8°F) and 75 psia (5.2 bar). If pressurizedliquid CO2 is allowed to expand to atmospheric pressure through a constriction, it passesthrough its triple point and a portion of the liquid is converted to dry ice particles. Whenthe “dry ice” warms, it converts directly to vapor bypassing the liquid phase. This is knownas sublimation. It leaves no residue in this process. The temperature of “dry ice” at atmos-pheric pressure is �78.5°C (�109.3°F). This low temperature accounts for its wide use as arefrigerant.

CLEANING WITH CO2

One of the latest uses of CO2 is in cleaning applications. CO2 is an effective cleaningagent in its liquid and supercritical states and as dry ice particles, flakes, or pellets. Thischapter will deal with dry ice snow cleaning only.

When the liquid CO2 is released to atmospheric pressure, flocculent particles of “snow”are formed. “Snow flakes” are formed when the particles of dry ice agglomerate. The par-ticles of dry ice snow are central to the cleaning process of CO2. Dry ice particles and flakesclean in a variety of ways. These mechanisms combine to create an effective means to cleanparticulate and light hydrocarbon contamination from surfaces quickly and efficiently. Thesolid phase of CO2 sublimes directly to a vapor at atmospheric pressures leaving noresidues.

MOMENTUM TRANSFER

Contaminating particles in the 2-�m range and smaller are difficult to remove by con-ventional blowoffs with compressed air or dry nitrogen. The smaller the particle, thegreater the percentage of its total area in contact with the surface to be cleaned and there-fore the lesser the percentage of its surface exposed to aerodynamic drag forces generatedby gas or liquid flowing over the contaminated surface. There are also electrostatic and


Figure 1 Particle removal by momentum transfer.

bonding adhesive forces that hold contaminants to surfaces. Even if there were no adhesiveforces holding the contaminant to the surface, there would still be removal problemsbecause of the boundary-layer phenomenon.

The boundary-layer phenomenon pertains to the fact that a fluid flow (gas or liquid) isonly effective at some finite distance from the surface being cleaned. As a flow gets nearera surface, its velocity decreases. The laws of fluid dynamics state that fluid flow velocity ata surface must be zero. Thus, small particles below this boundary layer are minimallyaffected by pressurized gas blowoffs.

One way to penetrate the boundary layer is to introduce a mass into the cleaning flowallowing the mass to transfer its energy to the contaminant and knock it free. Many meth-ods have been used to introduce a mass such as sand, walnut shells, or talc. The residue leftby these methods often outweighs their benefits. CO2 introduces a mass to the cleaningflow but leaves no residue. Contaminants are knocked free by the particles of CO2 snowand are carried away in the vapor flow of the CO2 (Figure 1).

By changing nozzle configurations the spray pattern, size, and force of the dry ice par-ticles can be adjusted. Microparticles can be generated in the �1-mm-diameter range witha velocity from 150 to 1000 fps that produces an aggressive dry ice storm capable of remov-ing light oils, light greases, and hydrocarbons. Fingerprints can be removed from manysurfaces. Other nozzles may be employed generating flakes up to 0.5 cm that produce agentle snow fall capable of cleaning unadhered particles down to 0.1 um without disturb-ing delicate substrates.

The force of the gas pressure is not the only means of transference of energy. As thevery cold particle of dry ice approaches a much warmer surface, the side of the particleclosest to the surface rapidly changes phase. When the CO2 sublimes there is a rapid, vir-tually explosive, expansion as the particle of CO2 changes from a solid to a gas. This is asource of shear stress energy that can dislodge a contaminant.

SOLVENT EFFECT

CO2 in its liquid phase acts as an excellent solvent. The solvent action occurs when athin layer of liquid CO2 forms at the collision interface of the dry ice particle and the sur-face being cleaned. The liquid is generated at the moment of impact when the dry ice par-ticle is deformed (Figure 2). Surface pressure on the dry ice particle rises above the triplepoint pressure of 75 psia (5.2 bar). At this pressure, all three phases of CO2 are present: solid,


Figure 2 Solvent effect from transient liquid due to impact pressures.

liquid, and vapor. It is the liquid that acts as a solvent dissolving organic contaminants andcarrying them off in the vapor flow.

THERMOPHORESIS

Thermophoresis refers to the temperature gradient in a particle of CO2. As the cold par-ticle approaches a warmer surface, the side of the particle closest to the surface warms morerapidly than the side away from the surface. Warm air molecules have the tendency to pushcontaminants toward the colder side of the particle of dry ice. The greater the temperaturedifference, the stronger the force. Contamination becomes entrapped in the dry ice particleand is again carried off in the vapor flow. The contamination remains entrapped in the dryice particle until it has fully sublimed. It is then redeposited at the locations where the dryice sublimed. This factor should be considered when cleaning with CO2 dry ice. Adequatemeasures should be taken to exhaust the vapor and dry ice flow to prevent the redeposi-tion of contaminants.

FREEZE-FRACTURE EFFECT

As the supercold flow of CO2 blankets a surface, certain organics may freeze into ahardened state. Liquid CO2 is forced through surface pores in the organic material. In theexplosive rapid phase change, the frozen, brittle materials are shattered and carried away(Figure 3).

APPLICATIONS

The many applications where CO2 dry ice cleaning can be used are diverse. Some of themore typical include the following:

Optics—From a 400-in. telescope mirror to 20 � 20 mm photoelectric cells, CO2

cleaning removes dust particles and light hydrocarbon contamination.Laser interferometers and mirrors—Cleaning with dry ice particles provides contami-

nant-free optical paths.


Figure 3 Freeze-fracture effect.

Silicon wafers—The speedy cleaning effect allows contamination as small as 0.1 umto be removed in an effective manner.

Ceramics—Light oils, hydrocarbons, and fingerprints can be removed from ceramicsurfaces.

Substrates—Flat-panel-display substrates and glass prior to coatings are effectivelycleaned of the most minute contamination.

Semiconductors—Printed circuit boards and hybrid circuits can be cleaned to a pre-cision level without disturbing delicate wire bonds or damaging substrates.

Read/write heads—Removing microscopic contamination is quick and thorough withCO2 dry ice cleaning.

Medical products—Cardiac and ophthalmic shunts are cleaned to a precision level.Medical tools and instruments have dust, light oils, and fingerprints removed.

Decontamination—Semiconductor instruments can be reclaimed by having arsenicdeposits removed from difficult-to-reach areas (threads, orifices). No measurablearsine gas has been detected after CO2 cleaning.

Micromechanical assemblies—Gyroscopes and microvalving components can becleaned to remove microscopic contamination.

Many other applications where precision cleaning to a microscopic level is necessary maybe served by CO2 dry ice cleaning systems.

CO2 cleaning is not the panacea for all cleaning applications and consideration for sev-eral aspects of CO2 use must be taken.

CONDENSATION

Because of the very cold nature of solid CO2, parts being cleaned that have a small ther-mal mass often can be cooled until they drop below the dew point of the ambient air. Theresult is that moisture is drawn from the surrounding air and condenses on the cooledparts. This may lead to additional cleaning problems.

There are several ways to eliminate or reduce this moisture problem. Parts may becleaned in a “dry box” where ambient air is purged and replaced with a dry inert gas, suchas nitrogen, argon, or CO2 vapor. This allows the ambient humidity to be reduced so thatmoisture does not form on the cleaned surfaces.


Another method is to heat the parts to be cleaned. Raising the temperature with a hotplate, infrared lamp, or heat gun allows the part to be cleaned without dropping its tem-perature below the dew point, thus reducing the condensation effect.

ELECTROSTATIC DISCHARGE

Precautions must be taken to protect parts vulnerable to electrostatic discharge (ESD).CO2 is itself nonconducting, but as particles of CO2 leave the nozzle of a cleaning mecha-nism friction results in an electrostatic charge. Many techniques can be used to reduce oreliminate the ESD. Grounding parts is one method. A voltage probe inserted into the CO2

flow can dissipate the ESD effect. Ionized airflow convergent with the CO2 is also an effec-tive means to reduce the ESD and may have the added benefit of reducing the condensa-tion effect.

REDISTRIBUTION OF CONTAMINANTS

As contaminants are knocked from the surface of a part, they are picked up in the flowof CO2 vapor and solid and can be redeposited wherever the CO2 sublimates. If there areorifices, nooks, crannies, or corners where the CO2 snow can accumulate, this is where thecontaminants will be deposited. There are various solutions to this problem.

Cleaning can be done under a laminar flow allowing the contaminants to be carried off.Often parts are cleaned upside down to allow the particles of dry ice and the contamina-tion to drop free of the part. Sometimes parts are heated to create a chimney effect of risinghot air to carry the contaminants free of the part.

Dry ice cleaning is a line-of-sight cleaning. Any surface that is struck by the CO2 parti-cle will be scoured. Often surfaces not meant to be cleaned are struck by the dry ice parti-cles, cleaned of contaminants, and then those contaminants are deposited on the surfaceoriginally intended to be cleaned. Dry ice particles can rebound off walls around surfacesand deposit contaminants from those walls. Again, adequate exhaust flow or laminar flowwill often prevent this occurrence.

THERMAL SHOCK

Parts that are sensitive to rapid temperature drops or cold temperatures may beaffected by CO2 cleaning. This may not be a problem, however, since cleaning with dry icesnow is very rapid. Particles are removed in an action so fast that parts often do not have achance to drop significantly in temperature.

TOO MUCH BLAST FOR SENSITIVE PARTS

Users may be concerned about the effect of the pressurized flow of CO2 striking theirsensitive parts. Because of the nature of the particle of CO2 at the moment of impact, dam-age rarely happens. A pressure distribution exists across the particle–surface interface.When the local pressure exceeds the yield pressure of the dry ice, the dry ice will yield. Theyield stress point is equal to the triple point pressure of 75 psia. At pressures above thetriple point a liquid phase will form that causes it to yield, resulting in an increased impactarea. Spreading the impact force over a larger area limits the stress and permits thorough


cleaning without destructive forces. By using interchangeable nozzles, users may also varythe spray pattern, force, and speed of the dry ice particles.

SURFACE PROPERTIES

Although CO2 dry ice cleaning can remove light oils and hydrocarbons, surface prop-erties may affect its cleaning potential. If a surface is porous and the contaminants areabsorbed into the surface material, CO2 cleaning may be ineffective. Oils have a tendencyto etch themselves into soft materials and coatings like gold, silver, or aluminum. Unlesscleaned in a timely manner, a more abrasive means to clean these etched surfaces must beused.

COST OF CO2 CLEANING

Cleaning to a precision level requires that no other contaminants be introduced to thecleaning process. With CO2 cleaning this means that ultraclean CO2 must be used.Typically, many grades of CO2 are available from industrial gas suppliers. These are sup-plied under many names (Food Grade, Coleman, Medical, Research, or Laboratory toname a few) but these grades should always be specified by their purity, 99.99% (four nine),99.999% (five nine), etc. Gas suppliers should be able to provide guaranteed specificationsfor the liquid-phase CO2 that includes total hydrocarbon, water, and nonvolatile residuecontent. The price for the purest grade of liquid phase CO2 may be negotiated with the sup-plier but will typically be from $2 to $15 lb. Availability of high-purity CO2 may vary fromregion to region.

CO2 snow cleaning systems, depending upon the manufacturer, may consume from 30lb/h to over 100 lb/h under continuous duty. With CO2 snow the cleaning operation is gen-erally done in short 2- to 3-second bursts. Hydrocarbon and fingerprint removal may takea longer spray duration. Rarely is continuous-duty operation performed.

If operating costs become significant, CO2 purification equipment is available thatallows the user to purchase low-quality “welding grade” CO2 and process it into ultrapure99.999999% quality. Welding-grade vapor CO2 could then be purchased in the $0.08 to$0.18/lb range, thus cutting the operating costs significantly.

The purifiers work through a process of distillation removing contaminants and non-volatile residues. Because of the vast differences in the condensing pressures and temper-atures for CO2 and these impurities, purifiers eliminate these contaminants and onlycondense pure CO2 into its liquid phase.

SAFETY/ENVIRONMENTAL

CO2 is an inert gas and human beings have a very high tolerance of exposure. OSHArequirements effective March 1, 1990 specify a PEL-TWA of 5000 ppm and a STEL of 30,000ppm. In real English, this means that a person can be exposed to an average concentrationof 5000 ppm over an entire 8-h workday. A person can also be exposed to a concentrationof 30,000 ppm when exposure is averaged over 15 min.

The threshold for humans is well below both of these limits. At moderate concentra-tions, a sharp odor or taste will be detected. If there is a concern regarding the accumula-tion of CO2 in the working environment, several high-quality monitors are available forCO2 detection or oxygen depletion. CO2 should only be used in well-ventilated areas whereCO2 vapor will not build beyond the tolerance levels.


The blast of dry ice snowflakes or flow of liquid CO2 should never be aimed at any per-son. The dry ice is very cold (�78.5°C). Momentary contact with human skin is harmlessbut prolonged contact may cause frostbite. Users of dry ice snow cleaning equipmentshould always wear eye protection.

Parts being cleaned should always be secured to prevent them from being blownaround by the vapor pressure of CO2.

Liquid CO2 is stored under high pressure. Cleaning equipment should be manufac-tured to withstand these high pressures. Safety devices like rupture disks and relief valvesshould be standard to prevent overpressurization.

If high-pressure cylinders are used, they should always be adequately restrained in anapproved cylinder rack or securely fastened to a structural wall or bench clamp unit to pre-vent any possibility of tipping over.

CO2 does not contribute to the depletion of the ozone layer. However, it is considereda “greenhouse gas.” Commercially sold CO2 is recycled CO2 that has already been pro-duced as a result of some other industrial process. If CO2 were not recaptured for cleaning,cooling, and other uses, it would have been vented to the atmosphere at the time it wasgenerated from the manufacture of ammonia and hydrogen, from natural gas production,and from the cracking of petroleum to make gasoline and other petroleum products. Its usetherefore is a delayed release and has no net additional effect on the environment. Also theamounts released for this application are exceedingly small compared with the amountsgenerated by the combustion of fossil fuels.

CONCLUSION

CO2 dry ice cleaning is a fast and effective way to remove particulate contaminationand light hydrocarbon contamination down to 0.1 um from silicon wafers, hybrid circuits,optics, disk drive assemblies, medical instruments, metal parts, and many other compo-nents. It leaves no residue.


CHAPTER 2.15

Gas Plasma: A Dry Process for Cleaningand Surface Treatment

Lou Rigali and William Moffat

CONTENTS

HistoryTechnologySurface Effects

Semiconductor ApplicationsWafer FabricationPackaging and Assembly

Die BondingWire BondingEncapsulationMarkingFluxless Soldering

Nonsemiconductor ApplicationsRemoval of Vacuum Grease from Machined Copperand Stainless Steel Parts

EquipmentSummaryReferences

HISTORY

The first commercial application of gas plasma was ashing animal tissue. The tech-nique was used to remove the organic matrix and leave the inorganic residue for subse-quent analysis. A chemist, Steve Irving, at the Signetics wafer processing companyreasoned, “If plasma could remove organic material, and photo resist was an organic mate-rial, then plasma should be able to remove resist from wafers.”1 This was in about 1965 andit did not take long for plasma technology to develop and expand into many areas of waferfabrication and to spawn companies such as Applied Materials, Lam Research, and manyothers whose total sales are hundreds of millions of dollars.


In the 1970s, companies like Raytheon and Rockwell were manufacturing hybrids andfound that plasma cleaning bond pads improved bond strengths by more than 70% andplasma treating the die pad improved adhesion of the die.2 Now, probably more than 80%of all hybrid manufacturers use plasma cleaning as part of the process.

TECHNOLOGY

Plasma is a state of matter, a so-called fourth state of matter along with gases, liquids,and solids. This chapter discusses a low-temperature plasma where a significant but lowfraction of the gas is ionized (an ion is a gas atom that has become charged by losing orgaining an electron). Such a plasma is not in thermal equilibrium because the temperatureof the gas is only about 30 to 50°F above ambient, while the electrons are much hotter. Aneon light is an example of this kind of plasma. The sun would be an example of a high-temperature plasma where the gas is all ionized and the gases and the electrons are both atvery high temperatures.

Typically, a low-temperature plasma is formed under vacuum conditions ranging from100 to 1000 mT, although there are a number of publications that describe an atmosphericprocess (not a corona or dielectric discharge).3 The energy to dissociate the gas can be eitherDC voltage or radio frequency from several thousand hertz up to microwave frequenciesof 10 GHz. It can be argued that the chemistry of the plasma is the same independent of thepower source; however, each method of dissociation has its merits and disadvantages.

Besides ions, a plasma contains free radicals, which are atomic and molecular specie inexcited energy states. Many of these specie can react chemically or physically under rela-tively mild conditions. For example, paper which will “burn” or oxidize rapidly at about800°C, and will undergo the same reaction at about 30 to 60°C in a plasma.

The types and nature of the specie formed will depend on the gases used. The mostimportant active agent is atomic oxygen in an oxygen plasma. This is a free radical and willreact chemically with organic material to form CO2 and H2O. When argon is used as a gas,an argon ion can be accelerated in a field and has enough physical energy to break carbon-to-carbon bonds or to displace by sputtering other elements on a surface. Whichever gasesare used, the reaction is at the surface and material is removed on the molecular level atrates of angstroms per second or minute. There is substantial chemistry that can take placeat the surface, especially with organic and polymeric material. Plasma can be used to intro-duce functional groups into a polymer chain or to actually deposit polymers. These appli-cations are important but not within the scope of this chapter, which attempts to describesome applications where dry plasma is effective and can be used instead of volatile sol-vents and corrosive chemicals.

SURFACE EFFECTS

The plasma process affects the surface of materials usually ranging up to 100 to 300 Å.One can etch material and as the material is removed, the new surface is exposed, and theprocess continues so that plasma can be used to etch (ash) microns of material.

Many applications, however, do not involve removal of the material but the modifica-tion of a surface. Treatment with plasma can change the surface energy. The lower the con-tact angle, the higher the surface energy and bondability. Plasma is an effective techniquefor etching and treatment of material that involves no organic solvents nor any acids orcaustic agents.


Figure 1 Contact angle.

A convenient way to measure surface energy is by measuring the contact angle of adeionized water droplet (Figure 1). When the contact angle is small, the surface energy islarge and the inactivated surfaces are called hydrophobic. Good adhesion can be obtainedwhen the contact angle is less than 10°. Many applications are related to the semiconduc-tor or electronics industry, but the modifications are of broader use and can extend intomany other areas, such as polymer chemistry and material sciences.

Semiconductor Applications

Wafer Fabrication

There are many terms that are used to describe the plasma process. It is difficult todefine these terms in many examples because people will use the same term when describ-ing different processes. In the semiconductor industry, the terms ashing, stripping, and etch-ing can all mean the removal of photoresist from a wafer. This is a major application andreplaces the use of hot sulfuric acid. Removing metal and metal oxides is usually referredto as an etching process. Thin organic and inorganic films such as nitrides can be depositedon wafers (and other surfaces). Most applications involve making significant changes inthe surface, extending several microns into the surface.

Packaging and Assembly

Unlike wafer fabrication, most of the applications in this section involve only the firstfew 100 Å of the surface. The topography of the surface is not drastically changed and mostapplications involve improving adhesion. Adhesion to a surface is usually good if the sur-face has a high surface energy. Terms such as cleaning, ablation, treatment, or roughening gen-erally describe the sample process.


Die Bonding

Plasma cleaning/treatment of substrates will always improve the adhesion of theepoxy and provide a better bond between the die and the substrate. The better bond pro-vides better heat dissipation. Studies have also shown that there is less delamination at thedie with samples that have been plasma treated.4

Wire Bonding

ESCA (electron spectroscopy for chemical analysis) has shown that the presence of car-bon on a surface will limit the quality of the wire bond.5 The relative level of carbon con-tamination on a copper surface, for example, can be determined from the ratio of the areaof the carbon (C) and copper (Cu) peaks in ESCA. Since water drop measurements aremuch easier to perform, it is good to know that the measured ESCA C/Cu ratios correlateextremely well with water drop contact angles measured on the same surface.

Bonding pads on the substrate are also subjected to various and inconsistent levels ofcontamination. One source of contamination on the die surface is fluorine ion, most likelyleft as a residue on the wafer or die during the fabrication process.6 Although the presenceof this contamination may not show wire bond degradation immediately, there is evidenceshowing correlation of the presence of fluorine element and bond failure due to the migra-tion of fluorine, causing embrittlement of the wire.7 Removal of this contaminant mayrequire the use of argon bombardment. This is an important application of plasma, i.e., theremoval of trace amounts of impurities such as inorganic ions by the use of argon.

Encapsulation

In this process the molding compound is asked to provide, in addition to all the otherrequired properties that are important, good adhesion to a number of different surfaces.Depending on the type of package, the molding compound must adhere to the substratematerial, solder mask, die, and the metal bond pads. There are many applications thatinvolve the bonding of one material to another, plastic to plastic, metal to plastic, etc. Ineach of these bonding operations, plasma improves the quality of the bond (Figure 2) asshown by the testing and measurement done in the study by Herard.8

Marking

The replacement of organic solvent-based inks with aqueous-based inks does not pro-vide the same consistency of adhesion. Even heat or hydrogen flame treatment is inconsis-tent. Again, plasma-treated surfaces always provide a better surface. Some situations willshow better results than others because of the nature of the ink, the encapsulant, and themarking process. Plasma treatment ensures uniformity of the process and decreases thevariability of the results. Printing or marking on many different types of surfaces is espe-cially improved with use of aqueous-based inks.

Fluxless Soldering

A plasma process, patented by MCNC, has been shown to give good welding resultswith typical lead and gold solder formulations, without the use of flux, and hence withoutthe requirement of a flux removal step.9


Figure 2 Improved adhesion.

treated

nontreated

Yield point

Composite rupture

Strain

171

65

Str

ess

(MP

a)

Nonsemiconductor Applications

Removal of Vacuum Grease from Machined Copperand Stainless Steel Parts

A manufacturer of a very sophisticated piece of medical diagnostic equipment thatoperates under vacuum conditions requires that the machined metal subassemblies that gointo product must be absolutely clean from contamination, in particular, but not limited to,carbon. In addition many of these same machined parts have to be periodically replaced.The wet cleaning process that was used included the use of acetone and water. The disad-vantage was that there was always some residue of the solvent that compromised the vac-uum, and a company directive indicated that wherever possible the use of all organicsolvents be eliminated.

A specific cleaning process was developed using O2/CF4 plasma in a batch plasma sys-tem (March PX-2400). The test samples were contaminated by taking a finger smear of vac-uum grease and spreading it on the parts. No attempt was made to measure the amount ofcontamination, but it would appear that it was about 1000 times greater than would nor-mally occur during actual use. Process time varied as one would suspect, because of theindeterminate levels of contamination, from about 10 to 30 minutes.

There are many other application that are outside the semiconductor industry. The fol-lowing list gives a brief summary of the types of current applications.

Medical—Treatment of cathetersEnvironmental—Low-temperature combustion toxic agentsManufacturing—Removal of machine oils from various partsPlastic—Treatment of films to improve surface propertiesDisk media—Improving the uniformity of deposited thin filmsFilm—Microvia etching10


EQUIPMENT

One can select manual batch, automated batch, or fully automated processes. Plasmasystems have been developed to meet most requirements from small development toolscosting less than $10,000 to automated systems costing about $250,000. Most systems oper-ate under a soft vacuum and require a vacuum chamber and, of course, a vacuum pump.A relatively new development3 involves the use of an atmospheric plasma, not a corona ordielectric discharge.

SUMMARY

Plasma, a dry nontoxic cleaning/treatment process, has been around for more than 30years and is being used to replace wet chemicals in many industries.

REFERENCES

1. Irving, S.M., Kodak Photoresist Seminar, 2, 26, 1968.2. Bonham, H.B. et al, Plasma cleaning for improved wire bonding on thin-film hybrids, Electr.

Packaging Prod., Feb., 1979.3. Roth, J.R. et al, Experimental generation of a steady-state glow discharge at atmospheric pres-

sure, Proc. IEEE International Conference on Plasma Science, 1992, 170.4. Oren, K., A case study of plastic part delamination, Semiconductor Int., April, 1996, 109.5. Djennas, F. et al., Investigations of plasma effects on plastic packages: delamination and cracking,

presented at Electronic Components and Technology Conference, Orlando, FL, June 1–4, 1993.6. Goodman, J, et al., Fluoride contamination from fluoropolymers in semiconductor manufacture,

Solid State Technology., July, 1990, 65–68.7. Gore, S., Degradation of thick film gold bondability following argon plasma cleaning, ISHM

Proc., 1992, 737–742.8. Herard, L., Surface treatment for plastic ball grid array assembly and its effect on package relia-

bility, Proc. Workshop on Flip Chip and Ball Grid Array, Berlin, Germany, Nov. 11–15, 1995.9. Koopman, N. et al, Solder flip chip developments at MCNC, presented at ITAP, 1996.

10. Fisher, J., ITRI Reports on Micro-Vias, Printed Circuit Fabrication, 1997, 58–60.


CHAPTER 2.16

Super-Heated, High-PressureSteam Vapor Cleaning

Max Friedheim

CONTENTS

IntroductionOperationAdditivesDryingSafetyWaste Stream ManagementEnergy and Water UsageGeneral ApplicationsCost Saving EstimatesCase Studies

Electronics Assembly, General CleaningFinal Surface Preparation prior to Laser Welding, Biomedical ApplicationDetail Cleaning of Refrigeration EquipmentPrinting Equipment

ConclusionsReferences

INTRODUCTION

Steam, a natural phenomenon used in cooking and medication, is known to all. Steamdrives ships, blows whistles, and in general is part of our lives. We take it for granted. Heatup some water above 100°C (212°F) and we get steam. Traditional steam cleaning has beenused successfully for a number of years for janitorial and food service applications. Thischapter, however, discusses a patented cleaning technique based on superheated high-pressure steam vapor.

The basis of steam/vapor cleaning or aqueous/waterless cleaning is readily explainedby analogy with steam from a tea kettle. Almost everyone at one time or another has put a


hand over the tea kettle into the flow of steam vapor, felt the warmth, and marveled in thedelight. But how many have thought to ask, “How come we are not scalded by boilingwater?” Simply put, the steam vapor is composed of single molecules of water, not waterdroplets. Water droplets have a high specific heat and can store much heat energy, whichcan potentially cause the burns or scalds when the common old-fashioned steam jenny orsteamer is used.

OPERATION

This new technology sometimes enables the use of water without an additive packageas the sole primary cleaning technique. In the steam vapor cleaner, distilled or deionizedwater is drawn from a reservoir. A metered amount of liquid is injected into the patentedchamber and instantaneously is converted to steam vapor under pressure. The equipmentis activated via a hand or foot switch, and the equipment discharges high-pressure, drysteam through a wand or handle. Efficiency of cleaning is based on:

• Heat• Pressure• Vapor-phase water

Heat, when applied to the contaminant such as oil or grease, helps to liquefy the soilso that it is more readily removed. In addition, the pressure used in the range of 200 to 300psi mechanically assists in dislodging and removing contamination from the surface. Formany benchtop applications, steam vapor cleaning offers advantages over traditionalhand-wipe cleaning. There is no wipe cloth, swab, or solid applicator. Instead, the combi-nation of hot steam and pressurized blast enables cleaning in tight spaces, complex geome-tries, blind holes, and under closely spaced components.

The equipment can be small and highly portable. As such, the technique can be usedfor small-scale benchtop cleaning or even in field repair, where the wand is handheld. Theduration of the steam blast is controlled by the operator. The operator typically activatesgeneration of steam vapor by pressing an activation switch. The resulting flow of high-pressure steam vapor continues for approximately 10 to 20 s, allowing the part to becleaned. In the smaller models, a short recovery time is required for more steam to be gen-erated. After the chamber has purged itself of all liquid, the user again activates the system.With some practice, the operator is able to time the bursts to provide a nearly continuousflow of steam. In addition, the technology can be adapted for automated, continuous gen-eration of steam vapor for large-scale operations.

ADDITIVES

In some cases, cleaning and drying can be accomplished with water alone. With heav-ily soiled surfaces, cleaning can be enhanced by application of surfactant solution directlyto the part to be cleaned. Steam vapor is then used to heat and remove the cleaning agentand contaminant.

Avoiding corrosion during cleaning of carbon steel is another potential problem. Insuch cases, an appropriately designed rust inhibitor can be added for applications wherepotential corrosion of the part is an issue.


DRYING

While many consider liquid aqueous cleaning to be environmentally preferable interms of pollution prevention, one potential drawback to liquid aqueous cleaning is theneed for a drying step. Drying may add to the cost of the process in terms of:

EquipmentDrying timeDrying temperatureCooldown timeEnergyHold up in production flow

In the case of steam vapor cleaning, because cleaning is accomplished with heated, pres-surized water vapor, the cleaned parts are typically dry and ready for the next step in themanufacturing process, be it plating, painting, or further fabrication.

SAFETY

Because the steam vapor is generated almost instantaneously, no steam is stored underpressure. This provides safer operation for the worker. Because steam vapor cleaning is freeof water droplets, the operator gains a measure of safety with steam vapor cleaning, evenat the high pressures employed. However, gloves may be considered depending on thespecific application, and goggles are recommended to protect the eyes from debris.

In addition, the equipment may be used in a waste management cabinet to minimizeworker exposure to debris from the part being cleaned, as well as entrapment of the residuebeing removed.

WASTE STREAM MANAGEMENT

Waste stream management is a concern in any cleaning process. It is generally under-stood that organic solvents must be handled as hazardous waste. However, liquid aqueoussystems must also be managed as waste streams. The additive package itself may beunsuitable for discharge to a sewer line. In addition, even where the surfactant package issaid to be biodegradable, soils and trace metals may result in the need to treat wash andrinse water as hazardous waste. Even with filtration and evaporative techniques, this canadd significant costs to the process.

In contrast, because the temperature at which the steam vapor is being produced is500°F the vapor evaporates, leaving only the residue of contaminants for disposal. Theresidue is typically collected on rags or absorbency pads, resulting in a relatively concen-trated, manageable waste.

ENERGY AND WATER USAGE

Compared with many other cleaning techniques, steam vapor technology is very lowin energy consumption because only a small mass of water is being heated. Many cleaningsystems require constant heating of a fairly large mass bath. In steam vapor cleaning,


Table 1 Estimated Water Consumption, Steam Vapor Cleaning

Size Steam Vapor Cleaner Water (average) use, gal/8 h

Small 1Medium 2Large, continuous steam capability 5

electrical current is drawn only when the chambers are heating. In addition, one cannottake the availability of water for granted. In many areas, water is a costly commodity. Insteam vapor cleaning, the average water consumption per 8-h shift is relatively low(Table 1).

GENERAL APPLICATIONS

Solvent-based cleaning, for all the safety and environmental problems involved, hasadvantages of rapid solubilization of a wide range of soils. With liquid, aqueous-basedcleaning, it is sometimes necessary to use additional cleaning approaches to achieve thedesired cleanliness. Steam vapor cleaning has utility not just on its own but also as anadjunctive technique to complement solvent or liquid aqueous approaches.

In some applications, a chemical alternative or solvent must be used because the con-taminant is such that only a chemical agent can provide the needed solubility or “soften-ing” power to allow it to be readily removed. Examples include burned-on carbon deposit,paint, and heavy industrial grease. In such cases, steam vapor can be used as a final rins-ing, drying, or detailing process. In such cases, steam vapor is used to remove the finaltraces of cleaning agent residue, soil, or surface oxidation.

In addition, while a large-scale cleaning operation may be adaptable to most of theparts being cleaned, there may be parts that do not lend themselves to the standard clean-ing process, or must be cleaned rapidly, on short notice. Vapor steam cleaning can be usedto add flexibility to the general cleaning system.

No cleaning system is perfect. With many cleaning systems in use today, a certain pro-portion of parts are not cleaned acceptably. This, in turn, necessitates costly time-consum-ing hand-probing-type detailing. Adding steam vapor technology as a final detailing toolaids in final inspection and acceptance of parts, and may allow manufacturers to do theirjobs more efficiently and economically.

Other cleaning systems may rely on applying abrasion or force to remove contami-nants. Pressure washers use water blasting; ultrasonic cleaning uses implosions to loosendirt; enclosed-type cleaners use spray and chemical under high pressure to dislodge soils.Steam vapor technology has essentially no abrasive quality as the vapor consists of indi-vidual water molecules, so it may be preferable for fragile, delicate applications.

COST SAVING ESTIMATES

A U.S. Navy report1 by its Fleet Activity Support Technology Transfer (FASTT–P-2)using this technology as a viable alternative to solvent cleaning and degreasing ofweapons, automotive parts, electronics, printed circuit boards, ground support equipment,and other items estimates a capital cost of approximately $8300, an annual saving of nearly$400,000 with a payback period of under 1 year, actually less than 10 days.


CASE STUDIES

Electronics Assembly, General Cleaning

Steam vapor cleaning has been evaluated for removal of flux and other contaminantsfor surface-mounted assemblies. Test were performed for the U.S. Navy at Crane, IN. Tenmotherboards and 26 interface cards were cleaned with steam vapor technology. No dam-age due to heat or electrostatic discharge (ESD) was detected. The U.S. Navy has author-ized use of the steam vapor technology with avionics and other applications.

Final Surface Preparation prior to Laser Welding,Biomedical Application

A manufacturer of stainless steel needles for biomedical applications used steam vaporcleaning to improve surface cleanliness prior to laser welding.

The overall process of unwinding the roll of stainless steel strip, bending, shaping, andlaser-welding the product requires 4.5 h. Prior to laser welding, the original process calledfor the strip to be run through solvent, then wiped dry between paper towels. It is impor-tant to remove all traces of solvent; any residual solvent interferes with laser welding.Unfortunately, residual solvent produced welding “misses,” resulting in an unacceptablereject rate. An automated steam vapor cleaner was implemented prior to laser welding.The reject rate was reduced to negligible levels, production was increased by over 30%, andsolvent usage was eliminated.

Detail Cleaning of Refrigeration Equipment

B/E Aerospace, Galley Products Group, in Anaheim, CA produces some 90% of the air-line galley refrigeration equipment in use worldwide. B/E also repairs refrigeration in-house and specifies options for field repair. After some process optimization, B/E hasintroduced steam vapor cleaning to replace some mineral spirits cleaning of segments ofrefrigeration tubing.2

Initially, assemblers accustomed to cleaning with mineral spirits were unfamiliar withthe new technology. By making the equipment available in the shop, operators found anumber of applications for steam vapor cleaning. It is currently in regular use in assortedapplications as a final cleaning technique.

Printing Equipment

The Los Angeles Daily News is subject to stringent requirements for solvent eliminationmandated by the South Coast Air Quality Management District (SCAQMD). The LosAngeles Daily News implemented steam vapor cleaning technology for an assortment ofprinting-related cleaning applications. As a result, solvent tank cleaning at the facility waseliminated. Emissions of volatile organic compounds (VOCs) were reduced from 20 to 2.2tons annually.


CONCLUSIONS

Steam vapor cleaning has found wide application in such diverse areas as electronics,aircraft, ground equipment, plant maintenance, and biomedical uses. Because of the diver-sity and relatively low capital investment, it has been adopted by the U.S. military, Fortune-500 companies, and small to medium-sized manufacturers. Still, there are many facets tousing steam vapor technology, some of which are yet to be discovered. An in-depth evalu-ation3 by the Naval Air Warfare Center Aircraft Intermediate Maintenance Facility atCoronado, CA concludes that there are so many potential applications for steam vaporcleaning that “we haven’t even scratched the surface yet.” To date, the manufacturer of thisequipment has observed no injury to products and no personnel injury. Further, plasticssuch as conformal coating on electronics assemblies are not damaged or removed whensteam vapor is applied.

Note: Technology and applications of steam vapor cleaning refer to various models ofthe PDQ Mini-Max Steam Vapor Cleaner.

REFERENCES

1. U.S. Navy report by Fleet Activity Support Technology Transfer (FASTT–P-2), available athttp://web.dandp.com/n451/aboutus/about.cfm.

2. Petrulio, R. and B. Kanegsberg, Practical solutions to cleaning and flushing problems, presentedat CleanTech ‘98, 1998.

3. Naval Air Warfare Center Aircraft Intermediate Maintenance Facility at Coronado, CA.


http://web.dandp.com/n451/aboutus/about.cfm

CHAPTER 2.17

Making Decisions about Water andWastewater for Aqueous Operations

John F. Russo

CONTENTS

IntroductionTypical Cleaning SystemOperational Situations of a Typical User

Determining the Water Purity RequirementsMeasuring Water Purity

Undissolved ContaminantsDissolved ContaminantsUndissolved and Dissolved Contaminants

Other ConditionsDetermining the Wastewater Volume ProducedSource Water Treatment

No TreatmentRemoval by Mechanical Filters, Adsorptive Filters,and an Oxidation Method

Mechanical FiltersAdsorptive FiltersOxidation Method

Water SofteningWater Softener Capacity Calculation

Dissolved Solids and Ionic RemovalReverse Osmosis ProcessDeionization Process DI or RO or BothOther Methods

No-Wastewater-Discharge OptionsClosed-Loop MethodZero-Wastewater-Discharge Method

Wash ChemicalRinse Water


Wastewater Discharge OptionsFat, Oil, and GreaseBiological Oxygen Demand and Chemical Oxygen DemandHazardous Metals

Determining the Wastewater Treatment for a New ProcessSource Water TreatmentNo-Wastewater-Discharge DesignWastewater-Discharge Design

Overcapacity of Current Wastewater Treatment SystemCase Histories

Case 1—Manufacturer Unable to Reduce the High Failure Rateof Plated Parts

Case 2—Large Computer Manufacturer Buys a System from Local SupplierCase 3—Large New England Military Contractor Decides to Build Its Own

System and Makes a Large InvestmentCase 4—Small Contractor

Conclusion References

INTRODUCTION

Water is the essential liquid in aqueous cleaning processes. Purity of the water is anintegral part of the cleaning process. With water, one must be concerned about the condi-tion of the water at each stage of the process to finish with a usable product. Also of con-cern is the condition of the water at the end of the process, i.e., the wastewater. This chapterdiscusses water purification and conditioning techniques both for the cleaning processitself and for the wastewater. In many cases, the wastewater from one stage of an operationis the source water for another stage. It is notable that discussions of water source treatmentprocesses are often integrated with those from wastewater since, in many cases, the prin-ciples and techniques are the same.

Usually this subject is discussed by describing several general water treatment sys-tems. But the author has decided to take the user’s viewpoint to make this chapter a moreusable reference. Even with minimal knowledge of water processes, the reader can refer tothe section “Operational Situations of a Typical User,” review the specific area of interest,and devise a plan of action.

As new technologies are introduced, users have more options in source water andwastewater treatment than ever before. This adds to the complexity of decision making,especially if the most cost-effective solution is necessary. Typical water treatment terms aredefined and various water processes are explained and compared. The main objective ofthis chapter is to introduce new users to the water treatment field and to serve as a quick,easy-to-use reference guide for experienced users

TYPICAL CLEANING SYSTEM

Essentially all cleaning operations use one or more of the sequence of operationsshown in Figure 1 (washer only). The schematic shows a parts washing unit where the


Figure 1 Conveyorized washer schematic. Note. The schematic represents a conveyorized washer.It can also be visualized as a multistage cabinet washer where all of the parts remain sta-tionary and are subjected to each cleaning step or a diptank cleaning process where theparts are moved manually or automatically from one cleaning step to another.

Washer Equipment

Parts

Wash Drag - OutReduction Rinse Final

Rinse

Knife

Air

Knife

Air

parts move along a conveyor to different stages of washing, rinsing, and drying. Also, thissame schematic can be visualized as a cabinet washer in which the parts remain stationarywhile they are subjected to one or more of the same cleaning stages as in a conveyorizedwasher. Many of the following discussions apply to this schematic.

OPERATIONAL SITUATIONS OF A TYPICAL USER

There are seven general, operational situation considerations:

1. Determining the water purity requirement2. Determining the wastewater volume produced3. Source water treatment4. No-wastewater discharge options5. Wastewater discharge options6. Determining the wastewater treatment for a new process7. Overcapacity of the current wastewater treatment system

In most cases, a user may have to consider more than one of the above situations. The firsttwo are the most critical and greatly affect the others. It is not unusual for minor differencesin conditions between one user and another to have a major impact on a user’s finaldecision.

Determining the Water Purity Requirements

In some cases, determining the water purity requirements is not easy and some inves-tigation is necessary. Information from trade associations, competitors, or related processesis helpful. If these sources are inadequate, the user may have to experiment on a small scaleor make the determination during the actual production process. The latter decision has adownside risk of too many part failures. The user may then have to rent a system on shortnotice to reduce the failure rate. In certain cases, purchasing new equipment with a vendorbuyback if the equipment is later found to be unnecessary is a good option.


Measuring Water Purity

In many applications the user must be concerned about measuring those characteris-tics of source water (tap water) from a lake, river, well, or from wastewater that affect thequality of the parts being cleaned. In the great majority of cases, two characteristics aremeasured: undissolved and dissolved contaminants.

Undissolved Contaminants

Undissolved contaminants are contaminants in water that do not affect its electri-cal properties. These contaminants can be measured by different methods, dependingupon specific requirements, and any or none of these might have to be monitored. Fat, oil,grease (FOG) measurements are used to determine whether a user complies with the dis-charge regulations of municipal sewer districts, called publicly owned treatment works(POTWs).

Total suspended solids (TSS) is a measure usually of the amount of suspended particleswith sizes over 0.45 �m. Fat, oil, grease (FOG) is a measure of any compound (vegetable oranimal fats, petroleum and synthetic oils, lubricants and some sulfur compounds)extracted by a fat-soluble solvent.

Dissolved Contaminants

Dissolved contaminants such as ionic compounds including sodium chloride, calciumcarbonate, and many others that form ions in water, are measured by a total dissolvedsolids (TDS), conductivity, or resistivity meter. Dissolved contaminants such as sugar,starches, and other water-solubilizing organic compounds are not ionic, do not conduct anelectrical current, and are not detected by electrical measurements. These measurementsare not usually used by POTWs to determine compliance with discharge regulations butcan interfere with some cleaning processes if not detected and removed.

Typically, measurement of dissolved contaminants is made with a TDS meter to makea quick approximation of the capacity of ion-exchange resins and reject capability ofnanofilter and reverse osmosis (RO) membranes. The readings are in ppm (parts per mil-lion) of ions in water. Each meter manufacturer might use a different algorithm to convertthe electrical measurement to a TDS reading so it is possible that different meters mightgive different results. Without this measurement, a user would need a complete wateranalysis, which is time-consuming and expensive. Such an analysis is done primarily whena high degree of accuracy is required.

The higher the dissolved ionic content of the water, the higher the conductivity. Sourcewater (tap water) typically has a conductivity from 40 to 1000 �S/cm. A conductivity meteris the measurement instrument of choice for water typically above about 10 �S/cm.Conductivity readings of about 1 �S/cm are near the limit of accuracy for this type of meas-urement.

For a conductivity meter to be useful as a TDS meter, the conductivity reading has tobe converted to an approximate amount (ppm) of ions in the water. The conversion factor(0.4 to 0.6) was determined by averaging the readings calculated from a complete wateranalysis of many samples of well, river, or lake water supplies throughout the UnitedStates. For wastewater, which may contain ions that differ substantially from natural watersupplies, this conversion range might be less accurate. For simplicity, all TDS readings usedin this chapter are determined by multiplying the conductivity readings by a conversionfactor of 0.5 (e.g., a conductivity of 1000 �S corresponds to approximately 500 ppm of TDS).


Table 1 Resistivity, Conductivity, and TDS Conversion Chart

Resistivity Conductivity Dissolved Solids ApproximateOhm-cm Microsiemens/cm Parts per Million Grains/Gallon@25°C @25°C (ppm) (GPG) as CaCO3

18,000,000 0.056 0.0277 0.0016415,000,000 0.067 0.0333 0.0019312,000,000 0.084 0.0417 0.0024010,000,000 0.100 0.0500 0.002925,000,000 0.200 0.100 0.005852,000,000 0.500 0.250 0.01461,000,000 1.00 0.500 0.0292

500,000 2.00 1.00 0.0585300,000 3.33 1.67 0.0971200,000 5.00 2.50 0.146100,000 10.0 5.00 0.29250,000 20.0 10.0 0.58530,000 33.3 16.7 0.97120,000 50.0 25.0 1.4610,000 100.0 50.0 2.925,000 200 100 5.853,000 333 167 9.712,000 500 250 14.61,000 1,000 500 29.2

500 2,000 1,000 58.5300 3,330 1,670 97.1200 5,000 2,500 146

100 10,000 5,000 292

Note: Approximate grains/gallon calculated by dividing ppm column by 17.1.

Source: From Owens, D.L., Practical Principles of Ion Exchange Water Treatment, Tall Oaks Publishing,Colorado, 1985. With permission.

For lower conductivities (purer water), the inverse of conductivity, resistivity, is meas-ured. A convenient conversion number to remember is that a conductivity reading of 1�S/cm is equal to a resistivity reading of 1 M�-cm. Table 1 lists typical conversions of con-ductivity and resistivity and TDS. From an accuracy standpoint, readings from 10 to 18.2M�-cm (the highest purity possible) can only be made with a resistivity meter with its cellinserted into a pipe in a flowing stream of water.

Undissolved and Dissolved Contaminants

There are several measurements that are made on water for both undissolved and dis-solved contaminants.

Total organic carbon (TOC) is a measure of the total amount of oxidizable organic mat-ter (oxidized by ultraviolet radiation).


Biological oxygen demand (BOD) is a measure of the amount of oxygen that bacterianeed to oxidize biodegradable organic matter over a given period of time.

Chemical oxygen demand (COD) is a measure of the amount of oxygen required to oxi-dize reducing compounds such as sulfides, salts of metals, etc. and organic com-pounds into carbon dioxide and water.

TOC measurements are usually used for critical high-purity water applications. BOD andCOD measurements are usually used to determine whether a user complies with dischargeregulations of POTWs.

Other Conditions

pH is a measure of the acidity, neutrality, or basicity of water and is expressed as thenegative log of the hydrogen ion concentration, or �log [H�]. A pH reading below 7 is anacid condition, 7 is a neutral condition, and above 7 is a basic condition. The pH of source(tap) water for certain wash chemical preparations and of rinse water in certain applica-tions can be very important.

Determining the Wastewater Volume Produced

Determining the amount of wastewater produced by a cleaning process is very impor-tant because it has a major influence on the user’s strategy and decision making. For exam-ple, for small volumes, cleaning processes generating less than about 25 to 75 gals/week,it is probably best to haul away the wastewater unless there is an existing treatment sys-tem. Depending upon the cost, some form of evaporation, like solar evaporation, might beless expensive. A determination of whether the wastewater is hazardous or not is requiredto comply with federal, state, and local regulations. Hauling a hazardous waste can cost asmuch as $1000/55-gal drum, whereas for a nonhazardous waste the cost can be less than$50/55-gal drum. For large volumes, other wastewater reuse processes should beemployed and are discussed in later sections.

Source Water Treatment

All aqueous processes require a minimum initial charge of water from a well, river,lake, or a transported supply of water (bottled or from a tanker truck). Many operationsmight need a continuous supply. Typically, a closed-loop system uses the lowest amount ofmakeup water, while a cleaning process without any water reuse requires the largestamount of water.

There are five typical options in general order of decreasing amounts of suspendedsolids and dissolved minerals:

1. No treatment2. Mechanical, adsorptive and oxidation3. Water softening4. Reverse osmosis (RO)5. Deionization (DI)

There are exceptions to this ranking, for example, a water supply with no treatment couldbe as good as another water supply that is softened. In some cases, a source water could be


even better than river water treated with RO, if the criterion is ionic content. Also, theamount of particles passing through an RO is far less than from DI, but the ionic contentfrom DI can be far less than RO.

No Treatment

In some cases the source (tap) water is of sufficient quality that no treatment is neces-sary. If a water purity specification is not available, the required purity might be deter-mined by testing on a small or a pilot production scale. If a pilot scale is not practical, it maybe necessary to go to a full production scale with a backup plan to treat the source water asquickly as possible should this option prove to be insufficient.

Removal by Mechanical Filters, Adsorptive Filters,and an Oxidation Method

Mechanical filters depend on a physical barrier for contamination removal. Adsorptivefilters use large surface areas to remove contamination. An oxidation method uses oxygento convert dissolved ions into particles that are removed mechanically.

Mechanical Filters

Mechanical filtration is one of the most common methods used to remove particlesfrom water and wastewater in cleaning processes. They are ranked from coarse to fineremoval with some overlap of removal capability of one method with another. See Table 2for a chart of the different types of contaminants and the separation technology used toremove each one.

Granular media filters are composed of single media or multimedia with various gradesof sand and other minerals, used primarily to remove suspended particles from 20 to 40 �m(micrometers or microns) in size, but can remove finer particles as well. As a referencepoint, a grain of table salt is about 125 �m. Bag filters are manufactured from feltlike mate-rials both woven and nonwoven and typically have a higher contaminant loading and alower cost per pound of contamination removal than cartridge filters. Cartridge filters arecommonly used filters made from a wide variety of plastic and natural fibers, such aspolypropylene and cotton, in a large variety of designs such as molded, fiber wound, andpleated papers.

Generally, cartridge filters are most often used for lower flow rates and higher-effi-ciency applications, whereas bag filters are used for lower-efficiency and high flow rateapplications. For high-flow-rate and high-volume applications, granular filters are mostoften used first, then frequently followed by the other two methods.

Membrane filters are manufactured from a variety of plastic and inorganic materialswith different shapes (flat sheets, tubes, spiral wound tubes). They are designed to removevery small particles and organic molecules from a liquid stream. Microfilters (MFs) arerated at about 0.05 to 1.0 �m. Ultrafilters (UFs) essentially remove all particles and mole-cules from about 10,000 to 1,000,000 MW (molecular weight; sometimes referred to asDaltons) from water.

There is neither an industry-wide micrometer rating that demarcates microfilters andultrafilters nor an industry-wide filtration efficiency rating standard. So it is not uncom-mon for a microfilter from one manufacturer to be called an ultrafilter by another manu-facturer. To compare one membrane with another, a user must determine from the


Differential pressure increases with reduced micron ratings; dirt-holding capacity and relative flow ratesdecrease with reduced micron ratings. Å, Angstrom = 10-8cm; mm, micrometer (micron)=104 Å; 1 mil = 0.001 in. =25.4 mm.

Note: Nanofilters, a newer technology, is between reverse osmosis and ultrafiltration.

Source: Gelman Sciences, Ann Arbor, MI, ©1987.

Gen

eral

filtr

atio

nM

icro

filtr

atio

nU

ltraf

iltra

tion

Rev

erse

Osm

osis

1,000,000

500,000

300,000

100,000

50,000

30,000

10,000

500

50

107

106

105

104

8,000

4,000

1000

800

400

100

80

40

10

1000

800

600

400

200100

80

60

40

2010

8

6

4

21

0.8

0.6

0.4

0.20.1

0.08

0.06

0.04

0.020.01

0.008

0.006

0.004

0.0020.001

Table 2 Relative Size of Small Particles

Mol. Wt. Å �m (micron)

• Sewing needlediameter

• Razor bladethickness

• Human hairdiameter

• Smallest visibleparticle

• Ragweed pollen

• Serratiamarcescens

• Pseudomonasdiminuta; DOP

• Albumin(60,000 MW)

Redbloodcell

Beachsand

Carbonblack

Emulsions(Latex)

Proteins

Metalions

Drizzle

Mist

Pollens

Jewelersrouge

Electronmicros-

copy

Whitelight

micro-scopy

Syrups

Colloids

Mycoplasm

Virus

Bacteria

Tobaccosmoke

Soluble salts(ions)

Yeastsand fungi

Endotoxins(Pyrogen)


manufacturer the test method for the rating. This rating problem can be extreme, for exam-ple, a membrane manufactured from a plastic material, such as polysulfone, polypropy-lene, or nylon, rated at 0.2 �m can reject 99.9999%� of all bacteria, whereas a ceramicmembrane with the same rating may have a far lower removal efficiency.

Neither of these types of membranes removes ions from water, but they do remove col-loids and other high-molecular-weight substances such as surfactants. Microfilter mem-branes have holes and are coarser than ultrafilters, and both are used to recycle washchemicals (alkaline cleaners). Ultrafilter membranes do not have physical holes and areeven more effective than microfilters in removing large organic molecules and low-molec-ular-weight petroleum products.

Adsorptive Filters

Activated carbon is a granular medium made by heating carbon-containing materials,such as coal, coconut shells, and similar substances, in the absence of air, producing aporous material with a large surface area. This large surface area allows the attachment oflarge organic molecules. Typically, it is used as a pretreatment method to remove chlorineand long-chained organic molecules prior to ion-exchange resins and some RO systems. Itacts as a catalyst to eliminate the oxidizing power of the chemicals by reducing them toother ions. It is also used to remove low levels of oil and grease (petroleum and synthetic)products.

Oxidation Method

Oxidation is a chemical process that changes the state of the dissolved species, such asiron or manganese, to a particulate form that is removed by mechanical filtration. This isan important pretreatment process before RO or ion-exchange resins for iron- and man-ganese-bearing water. Oxidation is sometimes used alone to treat water just before a clean-ing process. The oxidation is achieved by a chemical or air.

Water Softening

Water softening is a process of removing hardness minerals such as calcium and mag-nesium cations from water without reducing the TDS content of the water. The key com-ponent of a water softener is the ion-exchange resin contained inside a tank. The tank canhave manual or automatic controls to regenerate the ion-exchange resin (Figure 2).

Ion-exchange resin is manufactured from polystyrene that is cross-linked withdivinylbenzene. It consists of small plastic spheres about the size of the head of a commonpin. The resin has positively charged sodium cations held on the resin surfaces by electro-static charges. The sodium cations are exchanged for cations of calcium, magnesium, anddissolved iron in the water. Once all of the sodium cations are exchanged, the resin isexhausted. It must be replaced with new resin or be regenerated (reversing the process) byflowing concentrated sodium chloride brine through the resin during a multistage process,performed manually or automatically, within the tank. Even though ions are beingexchanged for other ions, there is essentially no change in the TDS of the water as meas-ured by a conductivity or TDS meter.


Water Softener Capacity Calculation

Water treatment chemists can predict the probable number of gallons of soft water awater softener will produce. For example, the “ppm” (expressed as CaCO3) of the water hasto be converted to grains per gallon because most ion-exchange resins are rated on the basisof grains expressed as CaCO3/ft3. The term “grain” is an old unit of weight measurement,originally referring to grains of wheat, and is used in the water industry. There are 7000grains per pound and 1 gr/gal = 17.1 ppm. To convert a reading, for example, 100 ppm ofhardness to grains/gal of hardness, the following proportion is used:

1 gr/gal = X X = 5.8 gr/gal17.1 ppm 100 ppm

Most water softeners with cation resin have a capacity of about 30,000 gr (expressed asCaCO3/ft3) of resin. If water supply has a total hardness of 5.8 gr/gal, a user can expect asoftener with 1 ft3 of cation resin to produce close to 5172 gal/ft3 (30,000 gr/ft3 � 5.8 gr/gal= 5172 gal) of soft water before the cation resin has to be regenerated again. There are fac-tors such as regenerant concentration, iron fouling of the resin, and others that can signifi-cantly influence the actual capacity of the resin.

Dissolved Solids and Ionic Removal

The most common industrial processes used for reducing dissolved solids and ions inwater are deionization (DI) and reverse osmosis (RO). Nanofiltration (NF), a membrane processvery similar to RO, can remove dissolved solids and ions to a much lesser extent, and isused in far fewer applications than RO. While RO removes dissolved solids and ions downto about 200 MW NF removes down to about 300 MW. Distilled water is not as economicalto use unless it is purchased in low volumes. Electrodeionization, a newer ion-exchangeprocess, produces high-purity water of less than about 0.4 ppm (�1 M�-cm resistivity) assodium chloride without the use of chemical regeneration.

Figure 2 Water softener.


Reverse Osmosis Process

RO is a membrane process that removes essentially all particles, and most moleculesand ions about 200 MW and larger from water. RO (Figure 3) is a process in which a pumpis used to force water through a membrane barrier to produce water with a lower dissolvedsolids content. The key component of an RO unit is the membrane, which is made from athin film of plastic, most often in the form of a spiral or “jelly roll.” Membranes vary in sizefrom 2 in. diameter � 10 in. long up to about 12 in. diameter and about 5 ft long. Water pres-sure up to 1000 lb/in2 forces water through the membrane. A complete RO system can con-sist of a pretreatment stage using mechanical filter (cartridge or multimedia filter),adsorptive media (activated carbon), and/or antiscaling (chemical, pH treatment, watersoftening), a high-pressure pump, RO membrane, storage tank (optional), and post-treat-ment (ultraviolet light, repressurization pump, and deionization). The selection of theseprocesses depends upon a source water analysis and the specific objectives of the user. Adouble-pass RO is an RO followed by another RO.

The RO membrane separates the water into two streams: contaminants into a rejectstream (wastewater to a sewer) and lower-ionic-content water into a permeate stream(usable for process). About 25 to 85% of the total water in a single-pass RO becomes a rejectstream containing all of the contaminants. Therefore, 15 to 75% of all source water becomespermeate water ready for use in the process. This percentage range is the practical limit fora single-pass RO and the actual percentage depends upon the RO design and a wateranalysis.

The membrane removes essentially all particles including microorganisms and rejects70 to 99+% of the dissolved solids and ions down to about 200 MW. It rejects essentially thesame percentage of ions whether the incoming stream has thousands or hundreds of partsper million of dissolved solids. For example, if the TDS of the wastewater to the RO dou-bles, the TDS of the permeate water will about double, and if ion-exchange is used as post-treatment after the RO, the ion-exchange cost will about double. The ionic weight, shape,and amount of the charge determine the degree of rejection.

The water purity of the permeate (usable) water typically ranges from 50,000 to 600,000�-cm and can be estimated with a source water analysis. As the membrane ages, its abilityto reject dissolved solids decreases, resulting in a practical life of the membrane of about 3years. Higher-purity water, which has a lower TDS and higher resistivity, can be attainedwith a double-pass RO (replacing the single-pass RO), DI, or electrodeionization of the ROproduct water being required. DI or electrodeionization is necessary as a post-treatmentprocess to RO whenever the user requires a higher water purity than 1 M�-cm resistivity.

Figure 3 Single-pass RO system.


Deionization Process

DI is a process using ion-exchange resin to remove ionized solids (cations and anions)from water. The key component of a DI unit is the ion-exchange resin. A two-bed deionizerconsists of two tanks in series: a tank with cation resin followed by a tank with anion. Thecation resin is the same as the resin used in a water softener except that it has hydrogencations instead of sodium cations on the functional groups of the ion-exchange resin.Another type of deionizer, a mixed-bed, has both cation and anion resins intimately mixedin one tank.

There are three basic deionizer designs:

1. Two-bed2. Mixed-bed3. Tri-bed (two-bed followed by a mixed-bed)

Usually there are three basic operating options:

1. Disposable resin2. Regenerable resin (rental or owned)3. On-site regenerable deionizers

For the disposable resin option, the resin is used once and discarded. For the rental orowned resin option, the user rents or owns the tanks with resin and the vendor takes theexhausted tanks back to its facility and regenerates the resins with strong acid and causticchemicals. For the on-site regenerable deionizer option, the resins are regenerated insidethe tanks with the same chemicals used in the rental or owned tank option, but the usermight have to treat the wastewater produced by the regeneration process for pH and/orheavy metals.

When resin is regenerated repeatedly, its capacity to remove ions after each regenera-tion decreases. The rate of this decrease depends upon a number of variables, such as thetype and amount of foulants, oxidizing power of the contaminants, temperature, and otherfactors in the water. The capacity decrease rate is usually greater for wastewater applica-tions than for source water (tap water) treatment.

With the disposable and rental or owned tank options, there is no waste stream to treatat the user’s facility since the contaminants are held on the resin beads. An RO system, bycomparison, always has a wastewater stream that goes to a sewer. This is the key reasonresin systems lend themselves more easily to closed-loop treatment, whereas membranesystems generally do not.

Occasionally, DI is referred to as demineralization, an older term used infrequentlytoday. Technically, deionized water is any water treated by a deionizer from which dis-solved solids are removed and the water resistivity increases. There is no specific waterpurity measurement that defines the term deionized. DI removes ions, positively chargedcations and negatively charged anions, from water using ion-exchange resins in the hydro-gen and hydroxyl form. Even though RO removes dissolved solids similar to DI and oftencan produce similar water with resistivity below 1 M�-cm, it is not referred to as deionizedwater, but RO water.

Ion-exchange resins have specific capacities, that is, the ability to remove ions from agiven number of gallons of water and it is inversely proportional to the TDS of the water.For example, if tripling the TDS, the capacity of the resin will be decreased to about onethird of its capacity. If the TDS is too high, the cost of replacing or regenerating the resin canbe uneconomical.


Sometimes the only way to make deionization economical is to use an RO membraneas pretreatment for the deionizer (discussed below with closed-loop systems). The type ofions, ionic charge, and the concentration of each ion can affect the capacity of the resin dif-ferently.

Mixed-bed ion-exchange resin has a nominal capacity of approximately 10,000grains/ft 3 to an end point of 1 M�-cm resistivity. The actual capacity of the resin dependsupon a number of factors such as amount of chemical used to regenerate the resin, the typeand concentration of each ion in the water, the amount and type of foulants in the water,the flow rate, the cross-sectional area of the resin surface in the tank, the depth of the resinin the tank, and the temperature of the water.

For a two-bed deionizer, the resin capacity is calculated from the capacity of the cationand anion resin. The cation resin has a nominal capacity of 30,000 grains/ft 3 and a strongbase anion resin at about 20,000 grains/ft 3. A weak-base anion resin (another option) has acapacity 50 to 100% greater than a strong base because it does not remove dissolved carbondioxide and silica.

Table 3 shows the key differences between the performance of four types of deioniza-tion systems.

DI or RO or Both

Generally, DI is

• Preferred when wastewater has a TDS less than about 100 ppm because operat-ing costs are lower;

• Required when higher water purity is needed than an RO alone can produce;• Able to maintain the same water purity even if the feed water quality varies sub-

stantially;• A simpler system to operate for low-flow-rate applications using rental tanks.

Table 3 Types of Deionizer Designs vs. Water Characteristics

Type of Deionizer Design

Water Two-Bed Two-BedCharacteristics Weak Base Strong Base Mixed-Beda Tri-Bed

Purity (M�-cm)b 0.02–0.6 0.1–1.0 1.0–18.2 (see previouscolumn“Mixed-Bed”)

pH 6 or lower 8.0� 5.5–8.5Carbon dioxide and

silica removal No Yes YesBOD and COD reduction ¨ Essentially none Æ

a A mixed-bed followed by one or more mixed-bed tanks is used when (1) a polisher is necessary to removeresidual ions that might get through only one mixed-bed tank and when 18.2 M�-cm water purity (highest waterpurity available) is required and (2) added capacity is required; the higher the water purity, the closer the pH is to7.0b These are typical ranges for each process.

Source: Otten, G., American Laboratory, July 1972.


Generally, RO is

• Preferred when wastewater has a TDS above about 100 ppm because operatingcosts are lower;

• Preferred when lower water purity is required, unless low flow rates are used;• Not able to maintain the same water purity if the feed water quality varies sub-

stantially unless DI post-treatment is used;• A more complicated system to operate for low-flow-rate applications.

Even though these reasons are typical for choosing DI or RO, there are exceptions:

• Even though the initial cost of an RO system and its operating costs are signifi-cantly higher than DI in a low-TDS case, the capability of an RO system to removemicroorganisms and other fine particles might be more desirable.

• The required use of strong acids and caustics when using a regenerable unit at theuser’s site may be too hazardous.

• Even though RO may be preferable in a higher-TDS application, the simplicity ofrenting a DI system with minimal operating costs may be preferred.

In summary, both of these technologies are used together whenever the water purityrequired is higher than an RO can produce and the TDS of the wastewater is too high tomake DI alone cost-effective. When comparing closed-loop and zero-discharge wastewatertreatment systems, it is important to consider that RO always has a reject stream, whereasDI might have a wastewater stream.

Other Methods

The following methods have limited use in providing high-purity water for cleaningoperations.

• Distillation is a process that heats water until it vaporizes and condenses intowater with a purity up to about 1 M�-cm. Distillation is capable of removingdissolved and undissolved minerals and some organics, but is not generally usedfor industrial water purification of tap water. As compared with RO and DI, it hasa higher operational cost because it is an energy-intensive process. However, it isan inexpensive source for low-volume applications if purchased in bottled quan-tities. Using bottled water is an economical way of testing what water purity isrequired by a cleaning process.

• Electrodeionization is an ion-exchange process that uses an electrical current on amembrane barrier embedded with ion-exchange resin. This process, usuallyrequiring pretreatment of a source water with a membrane process like RO, canincrease the resistivity of the water purity to 1 M�-cm and as high as 15� M�-cm. This is a newer technology primarily used to eliminate safety hazards fromusing strong acids and caustics when regenerating mixed-bed deionizers on site.Flow rates can range from low to high volume.


No-Wastewater-Discharge Options

The key to any “no-wastewater-discharge” option is the reuse of the wastewater.Sometimes wastewater from one application can be considered as acceptable source waterfor another process. Cascade counterflow rinses are used very often and are a good exam-ple of wastewater reuse in the same process. In this method, the purest water is used torinse at the end of the process and the wastewater flows opposite to the parts being cleanedas it cascades to the previous step in the process. As many as four cascade rinses are notunusual. Each time the wastewater is reused, the overall cost of water for the processdecreases as compared with using new water for each rinse stage.

A user may decide not to discharge any wastewater because:

• There is a desire or policy to reduce the chance of future liability for contamina-tion.

• The local community prohibits discharge of any industrial wastewater.• There is a high cost of monitoring contamination to a septic system and/or a pro-

hibitive cost of possible future remediation of the groundwater.• There is uncertainty of water availability.

Closed-Loop Method

A closed-loop process can be defined as a wastewater treatment process that has nowastewater discharged to a sewer, with the wastewater recycled to the same or anotherprocess. A closed-loop is not easily attained, but for some processes it is the most cost-effec-tive, ideal solution.

This is the design standard for the electronics assembly cleaning industry (see Figure4). In this application, the capital cost for a closed-loop system is about 20% more than anon-closed-loop system that discharges all the wastewater to a sewer. However, the oper-ating cost for a closed-loop system is usually so favorable that it has a positive operatingcash flow. The low TDS, below about 20 ppm, is the key to making this process economi-cal. The lower the TDS, the greater the return on the user’s investment. For many non-elec-tronic-assembly applications, the capital cost difference might be similar but the operatingcost for a closed-loop may be prohibitive because of the high dissolved solids in the waste-water.

Figure 4 Washer and closed-loop wastewater system interface schematic.


Several aspects of the electronics assembly application might be applicable to otheruser applications. In this application, a manufacturer takes printed circuit boards, inserts avariety of electronic devices on the boards, fluxes the boards, and then solders the devicesonto the board. The flux might be left as is on the board or sometimes is removed witheither source water or DI water and the wastewater discharged to a sewer or treated witha closed-loop system. This closed-loop process accomplishes the following:

• No water pollution (no wash or rinse water goes down the drain)• No wastewater tests, permits, inspections, and reports• Reduction of energy and water usage by at least 90%• Essential elimination of the continuous need for water• Water purity ranging from low to high depending upon the process requirements• Solid waste contaminants that are not hazardous except in unusual cases• Wastewater converted to hot, deionized water• Wastewater that can be recycled indefinitely• Pretreatment of water performed by equipment

The typical electronics assembly closed-loop design uses a combination of particulateremoval, organic removal media, and ionic removal media to allow the water to be com-pletely reused. Water purity levels start at 15 M�-cm and higher and, as the contaminantsaccumulate on the ion-exchange resin, the water purity decreases to the minimum accept-able water purity. This process operates economically whenever the water purity isallowed to decrease to about 1 M�-cm. However, the operating costs would be about halfas much if the water purity were allowed to decrease to 1000 to 20,000 �-cm. This latterdesign has the highest potential positive cash flow as compared with a non-closed-loopsystem. Once the particulate, granular organic, and ionic removal media are exhausted, thesolid waste generated is usually nonhazardous, according to the federal ToxicCharacteristic Leaching Procedure (TCLP) test.

Figure 5 Washer and zero-discharge-wastewater system interface schematic.

Knife

Air

Knife

Air

Washer Equipment

Parts

Wash Rinse FinalRinse

Closed-LoopDeionizer

Recycled Deionized Water

Drag outReduction

Zero-DischargeEquipment

Make-UpTo Wash

Wastewater Wastewater

Wash Chemical Recycle-Microfiltration

Evaporationand Haul

Pre-Treatment-ReverseOsmosis

SourceWater1

2 3

4


Zero-Wastewater-Discharge Method

Even though the closed-loop process is the ideal process because it has the greatestprobability of yielding the largest return on investment, it can be used only in limitedapplications. For applications where it is not feasible, a zero-discharge method can beused. This design allows no wastewater to be discharged to the drain and uses a combina-tion of microfiltration and reverse osmosis membrane, ion-exchange (closed-loop), andevaporation. When comparing this design with a closed-loop recycling, the additional capi-tal equipment is about double and it is more expensive to operate than a closed-loop system.

Figure 5 shows a possible zero-discharge design that represents the cleaning stages ofa typical conveyorized or batch-type cleaner (parts remain stationary). In this design, thewash chemical might be recycled with a microfiltration membrane system. Most often, thefinal rinse water from the same cleaning process cannot be recycled economically in an ion-exchange closed-loop system because of the excessive TDS, usually above about 75 ppm.This is caused by the dragout from the wash tank. If the same ion-exchange closed-loopprocess discussed above were used, a pretreatment method such as RO would be required;otherwise, the operating costs would be prohibitive. The following paragraphs describe indetail each part of the zero-discharge design and evaluate the user’s decision-making fac-tors, starting with the chemical wash (alkaline) stage, from left to right.

Wash Chemical

There are three methods of handling the wash chemical: hauling, evaporation, andrecycling.

If the wastewater is not hazardous, it can be hauled by a standard commercial vehicle.If it is hazardous, it must be manifested to an authorized facility. This might be used as atemporary measure until other solutions are implemented.

Evaporation may be an alternative, when hauling large volumes of wash chemical is notappropriate and recycling is not cost-effective. The user has more cost-effective options fortreating wastewater from a low-volume than a high-volume application. For example, incleaning processes producing less than about 75 gal/week of wastewater, it is more cost-effective to haul the wastewater unless there is an existing wastewater treatment system.For large volumes of wastewater, the hauling option is not usually cost-effective.

Evaporation is an energy-intensive process and the cost of the energy must be considered.It is a method of separating a liquid from its solids typically by heating the liquid (gas, elec-tricity, solarenergy)orbyusinga vacuum distillation unit. This can greatly reduce the amountof wastewater to be disposed of by 70 to 95%. If there are other processes in a plant producingexcess energy, or if solar energy is available, evaporation can be economical for large vol-umes. After evaporating the volatiles, the remaining contaminant might become a solidwaste containing hazardous metals or have a high pH, which makes it a hazardous waste.

For some of these processes, the vapors might be regulated and a permit might berequired. The water vapor from any of these evaporative devices might have a distilledwater quality (100,000 to 700,000 �-cm resistivity) that can be reused in the process.However, in most instances, the cost of condensing the water vapor is greater than treatingthe source water.

It is noteworthy that the cost of hauling and evaporation usually is not significantlyaffected by the concentration of dissolved minerals or hazardous metals in the wastewater.In addition to evaporating the spent wash chemical, an evaporator can treat the rejectwastewater stream from an RO membrane in the next stage of the treatment process.


Membrane recycling is a relatively new treatment process. As mentioned earlier, allmembrane processes have a wastewater reject stream containing all contaminants in a con-centrated form that usually goes to a sewer. However, some processes, like membrane recy-cling of wash chemicals, can reuse the reject stream in a closed-loop manner. Whenmembrane recycling of wash chemicals is used, the contaminants are continuously con-centrated and eventually must be processed or disposed.

This membrane recycle process uses microfilters or ultrafilters and permits the reuse ofa chemical cleaner by allowing most of it to pass through the membrane, while at the sametime removing the fine particles and emulsified oils. The term oil refers generically to bothpetroleum and synthetic products that are oils, greases, lubricants, and similar products.This separation process is imperfect and sometimes some or many of the key ingredientsof the cleaner are removed. The critical balance of this membrane recycling process is toachieve the separation of the oil from the wash chemical while not removing too many ofthe key ingredients of the chemical cleaner. Even in the best-balanced process, some chem-ical cleaner is removed and the critical ingredients might be replaced periodically withsmall amounts of additional chemical. Experience from operating such systems has shownthat the life of a chemical cleaner can be extended from three to ten times.

There are multiple benefits from this process, including increased life of wash chemicalwith resulting less chemical consumption, less water used, lower hauling costs, and less laborand downtime. There can also be an increased consistency of wash chemical with a muchlower average concentration of emulsified oil and a much lower average particulate level.

Table 4 Zero-Discharge-Wastewater System Designs Using Different TDSs of Wastewater toProduce Low- and High-Purity Rinse Water

Sampling 1 2 3 4Point TDS of Resistivity of Resistivity of Final

Wastewatera Pretreatment Water after Rinse Water(ppm) Equipment Pretreatment (�-cm) (�-cm)

Case A:Low-purity Up to (1) Single-pass RO (1) About 62,000b No DI closed-loop;rinse water about 5000 (2) Double-pass RO (2) About 900,000 same as column 3

Case B:Low- and Less than None Same as column 1 After a DI closed-high-purity about 20 loop, from 1000 torinse water 5,000,000

Case C:High-purity Up to Dragout reductionc (1) About 30,000 After a DI closed-rinse water about 5000 with either (1) single- (2) About 1,000,000 loop (1) 15,000,000

pass RO or (2) (2) 15,000,000double-pass RO

Note: The water sampling points for 1, 2, 3, and 4, are shown in Figure 5.aTDS of the wastewater going to drain from the wash chemical tank (excluding wash chemical) and is thewastewater treated by next column. The conversion from TDS (ppm) to conductivity is 0.5 (as calciumcarbonate) = 1 �S/cm.bIf 98% reduction is used for the second RO, the calculated resistivity would be 3 M�-cm (3,000,000 �-cm).However, the water purity is sensitive to any dissolved solids, which can most likely reduce the resistivity tobelow 1 M�-cm.cRO rejecting 98% of the TDS was used for these calculations. The percentage may as high as 99% but not forall dissolved solids. Dragout refers to mechanical methods used to reduce the amount of dissolved solids goingto the next process. The dragout could be a prerinse section in the in-line cleaner, or a time delay between twodip tanks to allow drainage of the parts or other similar method.


A user’s ability to achieve these benefits depends on a careful evaluation of the user’sprocess and suppliers of the cleaning chemical, parts cleaner, membrane unit, andoil/lubricant/grease contaminants.

Once a user is convinced recycling might be cost-effective, a demonstration test shouldbe performed on the wash chemical to determine its recyclability and the cleanliness of therecycled product. This would be followed by a pilot test at the user’s facility to corroboratethe benefits of recycling. Sometimes, other chemicals and micrometer-rated membranes arerequired to achieve optimum results.

Rinse Water

After the parts are washed, the next step in the cleaning process is rinsing. There areseveral possible methods to treat the wastewater depending upon the volume of waste-water produced. The designation of a low- and high-volume application is arbitrary, andthere can be a large overlap between the two in actual applications.

Hauling (even with evaporation) is usually not economical for processes produc-ing thousands of gallons of rinse wastewater daily. To make hauling cost-effective, reusemethods like RO can reduce the amount of wastewater requiring further treatment by upto about 75%. The RO can provide the additional benefit of treating the source water tomake up for any water losses from drying parts or the reject wastewater from the RO.

The last consideration for a zero-discharge-wastewater design is the effect on thedesign by the user’s requirement for either a low- or high-purity rinse water. Low- andhigh-purity water are arbitrary terms that can have a wide range of meanings dependingupon the user industry. For this discussion, water with a resistivity below 1 M�-cm is con-sidered low purity and water above 1 M�-cm is considered high purity. RO as pretreat-ment is required to attain both levels of water purity, unless the TDS is 20 ppm or lower.For low-purity rinse water, a single-pass RO might produce the required water purity. Fora high-purity rinse water, a dragout reduction step plus a single-pass RO, or double-passRO, might be required before a final rinse DI closed loop.

The amount of pretreatment depends upon the TDS of the wastewater being draggedout from the wash chemical tank by the parts being cleaned, racks, conveyor, and otherhandling equipment used in a dip tank operation, conveyorized in-line cleaner, or a cabi-net washer. As a first step for any type of cleaning process, it is important to orient the partsto allow more time to drain off the wash chemical. These pretreatment methods will assurea lower operating cost for a closed-loop system if used to polish the water up to 15.0 M�-cm and higher. Table 4 provides a guideline for the kind of pretreatment equipment and theexpected water purity for the final rinse. As shown, the lower the TDS of the wastewater,the less extensive the pretreatment equipment required.

Case A: Depending upon the TDS of the wastewater before the RO and the water purityrequirement, a single-pass RO alone might achieve a user’s goal for a low-purity-waterrinse (below 1 M�-cm). If the water purity is not sufficient, a double-pass RO will producea higher water purity than a single-pass RO. To achieve a zero-wastewater-discharge sys-tem, the reject wastewater stream from either RO process is hauled or evaporated andhauled.

Case B: As discussed above, no pretreatment is required for the economical operationof a zero-discharge wastewater system if the TDS of the wastewater is below about 20 ppmjust before a final rinse DI closed-loop. The key difference between operating a closed-loopsystem for a low (below 1 M�-cm) and high (above 1 M�-cm) water purity application isthat for a low-purity application the water purity is allowed to degrade to a resistivity ofabout 1000 to 20,000 �-cm, which is about the range of the purity of source water through-


out the United States. The control of this process can be accomplished simply with a con-ductivity or TDS meter. When the maximum conductivity or TDS allowed by the processis reached, the ion-exchange resin is replaced. This reduces the operating costs of the sys-tem by one half to one third compared with a high-purity application. For a high-purityapplication, the higher the minimum water purity required by the process, the higher theoperating cost, because the ion-exchange media will have to be replaced more frequently.For some applications, high-purity water may be too corrosive to the parts being cleaned,especially steel, galvanized, or brass parts. A DI closed-loop process produces only granu-lar media disposed of or regenerated at a vendor’s plant. The different operating condi-tions of this closed-loop process might compare more favorably to hauling whenever ROis used.

Case C: When the wastewater TDS of the stream feeding the RO membrane is about5000, the RO is followed by a final rinse DI closed-loop with a dragout reduction before theRO to reduce the TDS. Dragout reduction refers to mechanical methods used to reduce theamount of contamination dragged out of a wash chemical tank going to the next process.Its purpose is to concentrate the dragout from the wash chemical tank into the smallest vol-ume of water possible to minimize the size of the RO. For an in-line cleaner, the dragoutreduction step is usually a prerinse. For a dip tank, the amount of dragout can be controlledby letting the wash chemical drain from the parts into the wash tank before going to therinse tank, a brief rinse spray, or by using a still rinse tank of water (even source watermight be adequate). For a conveyorized cleaner, a good design is air spray the parts andconveyor belt to blow off excess wash chemical before it enters the dragout reduction stepthat has a water spray and follow with another air spray to blow off excess water. For thecabinet washer without a conveyor, the most practical way is to let the parts drain off excesswash chemical and, if necessary, have one or more short rinses with a drain-out step.

After the dragout reduction step, the next TDS reduction process is the RO. A double-pass RO without dragout reduction might be an alternative to a dragout reduction with asingle-pass RO. To determine which alternative is most cost-effective, compare the cost ofinstalling a dragout reduction in a washer along with a single-pass RO and evaporating theRO reject wastewater with the cost of using a double-pass RO without dragout reductionin a washer and evaporating the RO reject wastewater. The additional benefit of a double-pass RO is that there is a higher probability of achieving a higher water purity that mighteliminate the need for a final rinse DI closed-loop.

If the wastewater to the RO has significant amounts of oil or surfactants, the life of theRO membrane can be reduced. To protect the RO, pretreatment such as activated carbon orbag filters for low oil concentration applications or an ultrafilter (UF) or microfilter (MF)membrane for higher concentrations can be used. The activated carbon does not have areject stream creating more wastewater to handle while a UF or MF membrane processdoes. There are low-challenge applications for which a UF or MF membrane can be used asa dead-end unit (without any reject stream) and taken off line and cleaned periodically.

In summary, low-purity (below 1 M�-cm) rinse water is sufficient for some applica-tions. A single-pass or double-pass RO is adequate to produce this purity. But for higherpurity (above 1 M�-cm) rinse water, the amount of wastewater treatment depends uponthe TDS of the wastewater and the required resistivity of the final rinse water.

For low-TDS wastewater, no pretreatment is necessary before a DI closed-loop thatproduces a high-purity-water final rinse. For high-TDS wastewater, pretreatment beforethe final rinse DI closed-loop is required. The dragout wastewater from a wash tank alongwith the rinse wastewater might be recycled through a single RO or double-pass RO. Inother cases, a dragout reduction step prior to the RO may be a better choice to achieve therequired water purity. If a dragout reduction step followed by a single-pass RO does not


achieve the desired water purity, a double-pass RO might provide the additional removalof the dissolved solids. For high-purity-water requirements above about 1 M�-cm for thefinal rinse, a DI closed-loop may be necessary.

Wastewater Discharge Options

Federal, state, and local regulations determine a user’s program of action for which ofthe contaminants and how much of the contaminants to treat. Each user must comply withthe federal regulations at a minimum. After this requirement, the state regulations, whichmay be the same or even more stringent than the federal, must be followed. Finally, thelocal community regulations, which must be as restrictive as the state and federal regula-tions at a minimum, might be still more stringent. Local compliance issues can vary greatlythroughout the United States. It is very important for any user planning to discharge anyindustrial wastewater to obtain a permit from the local regulatory agency, a POTW). Eventhough the wastewater is in compliance with the discharge regulations, discharges fromsmall batch-type cleaners, like a household dishwasher, are considered industrial waste-water discharges subject to permitting before any discharge is allowed.

In the past, the testing point usually was the end of the sewer pipe from the building.However, in increasingly more states, the wastewater is tested in the building at the sourceof discharge as it comes from the equipment. This makes compliance more difficult.

The typical regulation requirements pertain to FOG, pH, BOD, and COD, and heavymetals. General treatment methods will be discussed for each of these conditions.

Fat, Oil, and Grease

Most POTWs regulate the amount of these three contaminants in wastewater. Thesecontaminants come from petroleum and synthetic compounds from the parts beingcleaned. The ability to remove them depends on numerous factors and conditions, includ-ing the condition of oil (free, dispersed, chemically or physically emulsified), temperature,amount removed per unit of time, types of petroleum and synthetic compounds, amountof TSS and TDS, available space, maintenance, and other operating conditions.

There are a number of removal methods selected on the basis of the specific applica-tion. Membranes, both MF and UF remove contaminants by preventing them from pene-trating the membrane and allowing water to pass through. Dissolved air flotation (DAF)uses air that attaches to free or dispersed oil and facilitates its rise to the surface of thewastewater for easy removal. Chemicals are often used to enhance the process efficiency.Chemical precipitation causes separation of the contaminants by precipitation.

Centrifugation spins the wastewater at high velocities, forcing the heavier particlesand high-molecular-weight compounds to separate from lighter molecules or particles. Acoalescer is a device constructed of materials that allows the adherence of very smalldroplets of contaminants that grow in size and are released to the surface of the water. Anoil skimmer includes a belt, disk, or other mechanical device or other methods such as a thin-film technology to remove contaminants from the surface of the wastewater. A decanter(gravity separator) allows the separation to the surface of contaminants lighter than water,which then, under low turbulence conditions, spill over a weir into a waste container.

The following guideline shows an approximate order for the effectiveness of eachprocess according to its ability to remove petroleum and synthetic compounds from waste-water.


• Membranes (most effective)—UF—MF

• Dissolved air flotation• Chemical precipitation• Centrifugation• Coalescers• Oil skimmers• Decanters (least effective)

pH

Typically, the wash tank of a cleaning operation contains an alkaline cleaner with a pHhigher than the local discharge limit. This condition can be corrected by using an acid pHchemical control system. When there are regulated hazardous metals, the user must com-ply with the federal, state, and local regulations when disposing of the waste. Treatment ofhazardous metals is discussed below.

Biological Oxygen Demand and Chemical Oxygen Demand

BOD is a test method that uses microorganisms to determine the amount of oxygenrequired to oxidize organic contaminants in water. COD is a test that uses a chemical oxi-dant to determine indirectly the amount of oxygen required to oxidize both organic andinorganic contaminants in water.

Sometimes, state and local regulatory agencies have limits for BOD and COD. Theremoval methods for petroleum and synthetic contaminants may achieve sufficient reduc-tion of these two measurements to meet these discharge limits. However, BOD and CODnot only measure these contaminants, but also other oxidizable compounds that the FOGtest does not.

A packaged biological wastewater treatment system reduces the BOD levels to meetthe discharge limits. It is a natural process that uses microorganisms to achieve the degra-dation of the organic contaminants and is used by essentially all POTWs in the UnitedStates. An equivalent industrial design is based on the amount of wastewater beingprocessed and is usually much smaller than what a POTW would use. Since COD is com-posed of both inorganic and organic contaminants and microorganisms effectively oxidizeonly organic contaminants, an insufficient amount of the inorganic contaminants might beoxidized to meet the discharge limits.

To reduce the COD further, a chemical oxidant, carbon adsorption, ultraviolet oxida-tion, ozonation, or other means is required. A membrane process such as ultrafiltration,nanofiltration, and RO could be used, but they are more often used in a recycling processwhere the permeate would be reused. The reject stream for either of these two processesincreases the concentration of the contamination from 10 to 100 times.

Hazardous Metals

The eight hazardous metals that are federally regulated are cadmium, lead, selenium,mercury, barium, chromium, silver, and arsenic. In addition, some states and local agenciesmight list others. Any of the four following methods might be used to reduce the metal con-centration in the wastewater to meet discharge limits:


1. Mechanical filtration (particulate only)2. Chemical precipitation (particulate and dissolved)3. Ionic removal (dissolved only)4. Membrane (particulate and dissolved)

The choice depends greatly upon the metal and its state (dissolved, particle, colloidal), flowrate, total flow per day, and other factors. For example, if a user is cleaning cadmium-platedparts and must comply with an FOG and cadmium metal regulation, a UF might achieveboth so long as the dissolved cadmium metal is not beyond the regulatory limit. The mem-brane does not effectively remove dissolved low-molecular-weight contaminants. In somecases, the processed water could be reused instead of being disposed. The reject streamcontaining the concentrated metal and oils would be hauled as hazardous waste.

Determining the Wastewater Treatment for a New Process

This is one of the most difficult applications. To reduce the uncertainty of wastewatertreatment decisions, the user should determine the local source water conditions, similarprocesses in the industry (competitors), availability of hauling, potential dischargewaivers, and piloting the process, all of which can aid in limiting overdesigning costs. Theless that is known about a process, the greater the margin of safety that is usually necessaryto ensure a treatment system that meets the user’s requirements. The user should try tomaintain maximum flexibility before buying a permanent system. This section examinesthree possible decision-making areas.

Source Water Treatment

If a water sample is available, it is best to have it analyzed especially if high-puritywater is necessary. It is best to wait for the results of the analysis before renting a long-termsystem or buying a permanent system unless the uncertainty of the treatment process isminimal.

No-Wastewater-Discharge Design

It is difficult to achieve an economical wastewater treatment system for a no-wastewater-discharge design because of the unknowns: type of wash chemical, specificcontamination generated by the process, surface quality of the parts, and other conditions.For small-volume applications, the entire wash tank and rinse water could be hauled. Forlarge volumes of wastewater, where hauling might be a problem and the user is on a munic-ipal sewer, it may be possible to discharge it with minimal treatment on a waiver. If on aseptic system, river, or other body of water, hauling may be the only practical way. Anotheralternative for any of the above could be a temporary treatment system alone or along withhauling until enough data are gathered to define the final permanent treatment system.

Wastewater-Discharge Design

If the user has decided to discharge to a POTW, it is necessary to obtain the dischargeregulations to determine the wastewater conditions that must be met and to obtain a per-mit. It is easier to prepare for this application than for a zero-discharge design because thereare far fewer conditions affecting the final design. For example, for most alkaline cleaningapplications, pH and oil are the two key concerns. For the pH adjustment, equipment is


usually easily obtainable on relatively short notice. The amount of oil in the wastewater ismore difficult to assess and could lead to a large, unnecessary initial expenditure if a largemargin of safety is required, such as considering a UF membrane or chemical treatmentsystem. In such cases, a discharge waiver from a POTW would be of great value until thefinal effluent is tested.

Overcapacity of Current Wastewater Treatment System

In such applications, usually recycling at the source of the discharge can become a pri-mary solution. The reason is that the cost of expanding the entire wastewater treatment sys-tem is usually much more than trying to reduce the amount of wastewater going to thetreatment system by using a point-source treatment system. A careful evaluation of all dis-charge sources is made to determine which are the most viable from a cost standpoint. It isunusual for the expansion of the central wastewater treatment to be the most economicalchoice. For temporary overcapacity applications, hauling may be most economical.

CASE HISTORIES

Case 1—Manufacturer Unable to Reduce the High Failure Rate of Plated Parts

Situation

“We are replacing our wash chemical weekly, but part spotting is still a substantialproblem that causes post-cleaning plating part failures.”

Discussion

The user was replacing the 600 gal of wash chemical weekly because neither coalesc-ing nor skimming was capable of removing the emulsified oil, causing part spotting. Toconsider a wash chemical membrane recycling system, the user had to try another type ofwash chemistry. After a successful match of a new wash chemistry with a recycling system,the incidence of spotting was essentially eliminated. After the new equipment wasinstalled, the user’s costs from product defects, chemical purchases, haulage of spentchemicals, and labor totaling about $120,000/year were eliminated. With the new systemthe concentration and cleanliness of the wash chemical are maintained at a relatively con-stant level where, previously, the emulsified oil would build up toward the end of theweekly cycle of replacement of wash chemical.

Lesson

New technologies can sometimes help solve problems that existing methods cannotand, in addition, can yield additional unforeseen benefits.


Case 2—Large Computer Manufacturer Buys a Systemfrom Local Supplier

Situation

“I have a local wastewater treatment company that said it could do it.”

Discussion

The manufacturing engineer was not familiar with wastewater treatment and believedthe local company. The installed system cost in excess of $50,000 and required essentially afull-time operator trained in chemical wastewater treatment practices and a 20 20 ft floorspace. This type of system is very typical in a printed circuit board fabrication facility. Highoperating costs, floods, and high volumes of water discharged to drain characterized thefirst year’s operation before a major design change eliminated one of the three major prob-lems (floods). Another vendor with extensive experience with these systems had informedthe engineer that it was not economically feasible to operate a closed-loop system withoutmajor changes in the way the cleaner had to operate. Several months later, the engineer leftthe company under unknown circumstances, and a supervisory engineer involved in thedecision making was reassigned. Several months later, the company purchased anotherclosed-loop wastewater recycling system for about $35,000 with a specially designedcleaner specified by the recycling vendor. The system only required an operator once every3 to 4 weeks for 2 h for normal maintenance.

Lessons

1. The engineer lacked the fundamental knowledge necessary to judge the techni-cal merits of the two competing companies.

2. The local vendor had no operating systems experience or knowledge of thesesystems, despite its other wastewater treatment experience.

Case 3—Large New England Military Contractor Decides to Build ItsOwn System and Makes a Large Investment

Situation

“I can do it myself for less money.”

Discussion

A manufacturing engineer believed that a large central system would be more cost-effective than a vendor’s recommendation of several standard packaged systems. The userhired a water treatment consultant to assist with the design of the system. After talking toa number of different water treatment suppliers, none of whom was familiar with boardassembly wastewater recycling, the engineer bought components and integrated them intoa system with the help of the above consultant at a cost of $200,000� and the time of out-side engineering consultants. The footprint was ten times what the packaged systemswould have required, making an addition to the existing building necessary. At this time,


the expensive, overdesigned system does not achieve the specified water quality or theoperating economies used to justify the investment.

Lesson

A careful evaluation of alternative technologies and vendors is important before mak-ing a final decision.

Case 4—Small Contractor

Situation

“If we did it over again, we would have spent less money, and saved 160 hours of engi-neering time and liability concerns with a local water purification company servicing awaste treatment application.”

Discussion

Upper management decided that a local water purification company, not experiencedin waste treatment, could perform the service less expensively. The user purchased a watertreatment system at a substantial cost without the necessary functional features. In addi-tion, the user was not aware of the liability issues concerning the possible misuse of lead-contaminated ion-exchange resin by vendors servicing both waste treatment systems andhigh-purity systems such as medical facilities, laboratories, and other sensitive customers.If these other customers knew that their vendor was supplying them with resins that hadbeen exposed to wastewater containing lead and other contaminants, they would immedi-ately discontinue their business relationship.

Lesson

The engineering and design of closed-loop wastewater recycling systems were seri-ously underestimated and liability issues were completely overlooked.

CONCLUSION

The current general trend is increasing stringency of discharge regulations. Thisrequires continual vigilance by users in maintaining their knowledge of current water andwastewater practices.

Selecting the best source water and wastewater treatment processes for a cleaningapplication requires a methodical approach. In the case of solving an immediate cleaningproblem, it is usually best to take a systems approach by evaluating the entire cleaningprocess each time because of the interdependency of each part of the cleaning process.Sometimes a simple change in the cleaning or water/wastewater process can alter theentire economic equation, transforming a previously uneconomical solution into an eco-nomical or, perhaps, even the best choice.


REFERENCES

1. Water Quality Association, WQA Glossary of Terms, 1993.2. McPherson, L., Correlating conductivity to PPM of total dissolved solids, Water/Engineering &

Management, August 1995.3. Owens, D.L., Practical Principles of Ion Exchange Water Treatment, Tall Oaks Publishing, Colorado,

1985.4. Kunin, R, Ion-Exchange Resins, Robert E. Krieger, 1985.5. American Water Works Association, Water Quality and Treatment, McGraw-Hill, New York, 1990.6. Byrne, W. Reverse Osmosis: Practical Guide for Industrial Users, Tall Oaks Publishing, Colorado,

1995.7. Quitmeyer, J., Sifting through filtration options, Precision Cleaning, December 1997.8. Russo, J.F. and Fischer, M., Operating cost analysis of PWB aqueous cleaner systems: zero dis-

charge water recycling system vs. once-through, presented at Third Int. SAMPE ElectronicsConference, June 20–22, 1989.

9. Kieper, T. and Russo, J.F., Closed-loop alkaline recycling proves an award winning application,Parts Cleaning, May 1999.

10. Rajagopalan, N., Lindsey, T., and Sparks, J., Recycling Aqueous Cleaning Solution, ProductsFinishing, July 1999.


CHAPTER 2.18

Overview of Drying: Drying after SolventCleaning and Fixturing

Barbara Kanegsberg

CONTENTS

IntroductionSolvent DryingFixturingThe Human FactorConclusionReferences

INTRODUCTION

Drying is a critical step in the cleaning process. The drying process must be chosencarefully with consideration of the product, process flow, and ultimate use of the product.In planning a cleaning process, to assure an effective process with optimal process flow,budget some money for an effective drying system. Particularly with aqueous cleaning,many manufacturers have found it to be the most time-consuming step in the process. Thisis because drying requires that the product be heated and agitated. If the product cannottolerate a high temperature, for physical drying, the drying step must be longer and slower.In addition, if a fairly large component must be heated to a high temperature, it may be toohot to handle for longer than may be tolerable to achieve efficient process flow.

This chapter considers drying after solvent cleaning, the importance of fixturing, andthe critical human factor.

SOLVENT DRYING

Two chapters in this section are concerned with drying in aqueous processes, and thisis indeed appropriate because water is, well, wet. Water has a much higher boiling pointand higher surface tension than many solvents, so it is more difficult to get rid of. Perhapsthe reader is thinking: “I am purchasing a highly sophisticated solvent system. Drying willnot be a problem.”


Think again!Those considering solvent systems also need to plan carefully to achieve adequate dry-

ing in an efficient, cost-effective manner. Some of the principles (although not all of thespecifics) discussed in the chapters on removal of water are very applicable to designing agood solvent drying system. The author suggests that particular attention be paid to dis-cussions involving vacuum drying and the need to avoid recontamination of the part dur-ing the drying step.

In general, all drying systems, solvent and aqueous, should be properly vented to min-imize employee exposure to undesirable vapors. Considering that understanding of whatis undesirable is still limited, and considering that many different solvent residues may bepresent, it is best to err on the side of caution even when removing seemingly benign sol-vents. At the same time, solvent traps may be needed to assure environmental compliance,neighborhood safety, and worker safety.

Those experienced in cleaning regularly observe problems associated with solvent dry-ing. Some are equipment related; others are related to employee education and inappro-priate handling of components. Inadequate and improper drying in solvent cleaningprocesses can result in:

• Product damage• Production slowdown• Solvent loss

With aqueous systems, drying is often optional. In solvent systems, the drying portion ofthe process may be so integrated a part of the cleaning process that it is indistinguishableas a separate set of features. It is important that those features related to drying be identi-fied and then which features will be most important in an application be determined. Somefeatures associated with solvent drying include:

• Hoists• Superheated vapor zones• Freeboard• Sample rotation• Fixture design• Vacuum-assisted removal of solvent

Solvent drying is often thought of as automatic. However, drying is relative. One needsto consider not only how clean is clean enough, but also, how dry is dry enough. The con-cept of adequate drying is relative to process requirements. Perhaps it is not necessary todry. In industrial applications, a very light coating of mineral spirits or of a light oil can pro-tect the part from corrosion. However, with components containing plastics and those hav-ing complex geometries, one needs to consider the consequences of not removing residualsolvent. Residual solvent can interfere with subsequent process steps such as coating andphysical assembly.

One must also consider the consequences of outgassing. Outgassing can be thought ofas a subtle form of solvent incompatibility or cleaning agent residue. Minute crevices incomponents can entrap solvent. Plastics and epoxies can adsorb solvent. The extent of sol-vent adsorption depends on the material, the chemical, and the application conditions(such as temperature, pressure, duration of cleaning). At ambient temperature, the solventcan be released over a period of days, weeks, or even years. It is particularly critical toavoid outgassing in sealed systems, which are expected to last for decades. In a sealed


gyroscope, released solvent may chemically react with the flotation fluid, producing amedium that corrodes delicate coil windings, resulting in unexpected, catastrophic prod-uct failure.1 One can think of other sealed, critical systems, such as pacemakers, where nei-ther the doctor nor the recipient wants to deal with catastrophic product failure.2,3 Even innonsealed systems, for biomedical implantables and other critical applications, avoidingsolvent outgassing would seem to be a reasonable policy on general principles. One alsomight want to avoid outgassing where the part is sealed or bagged. If metal parts withresidual water are placed in plastic bags for storage or shipment, they can corrode or dis-color. Similarly, solvents can gradually escape or outgas from parts in plastic bags, per-haps interacting with the plastic, perhaps recondensing on the parts—in general makinga mess.

How does one know when there is outgassing? There are analytical techniques such asresidual gas analysis (RGA) and head-space gas chromatography (GC). In both cases, aprotocol is established to speed up the outgassing process, systematically collect thevapors, and then analyze them. However, only users and their co-workers can make thedecision regarding where outgassing is an important factor. If so, one needs to dry. As withremoval of water, drying to avoid outgassing can be accomplished by both chemical andphysical methods. Users may decide to use both methods.

Removal of solvent can be accomplished by chemical methods. Isopropyl alcohol (IPA)is sometimes used to displace water. IPA can also be used as a second rinsing and dryingstep with high-boiling solvents such as those based on mineral spirits, ester blends, long-chain alcohols, and terpenes,2 However, for some applications, with very complex compo-nents, particularly those containing complex composites and plastics, IPA may leave thecomponent wet, for the purpose of the process under consideration.4,5 In such cases, the IPAwould be classically displaced by a perfluorinated material (PFC). Today, because of con-cerns about using materials with a long atmospheric lifetime, hydrofluorocarbons (HFCs)or hydrofluoroethers (HFEs) are preferred over PFCs in such applications.

Solvent drying (i.e., removal of excess solvent) can also be accomplished by physicalmethods, including spin-drying and vacuum, and vacuum with heat. In manual systems,as might be used with very low volume, high-precision processes, vacuum bakeout in asmall oven can be useful. In larger, airless or airtight cleaning operations, the vacuum dry-ing system may be an integral part of the cleaning system.6 Appropriate vacuum dryingcan complete the process of cleaning under vacuum. In all cases, the process must be care-fully adjusted to the product in question and to the total volume of parts to be processed.Where low-flash-point solvents are involved, the drying system must have appropriateengineering controls to prevent fires or explosions.

In a liquid/vapor-phase cleaning or degreasing system, the components must be heldin the vapor zone and then above the vapor zone to achieve adequate final cleaning, rins-ing, and drying. It is critical to handle solvent properly to avoid solvent loss. In addition totechnical considerations, there are both regulatory and economic factors related to solventdrying. For certain solvents, notably in the Federal Halogenated Solvents NESHAP,7 thedwell time and rate of removal are spelled out. Local regulations may also call out certainrequirements to avoid emissions of air toxics (HAPs) and volatile organic compounds(VOCs). Readers are reminded to explore the fascinating, and often contradictory andByzantine, world of solvent control as it applies to the regulatory climate in their location.Even for relatively unregulated solvents, it is important to remember that cleaning agentsare not free. Some of them cost more than $15/lb. Particularly with large-scale operations,containing the solvent can be an important economic consideration even for relatively inex-pensive solvents. Even in well-designed, automated systems, assuring complete drying ofthe parts is a critical part of solvent containment.


Table 1 Summary: Impact of Fixturing on Drying

Fixturing Factor Positive Impacts on Drying Provisos

Basket design, large Adequate exposure of parts Large solid, nonmesh metalproportion of mesh or to heat, forced air, or areas of fixtures may trapscreen drying solvent water or drying agent

Avoid trapping water or dryingagent

Materials of construction Able to withstand heat Exercise care in adaptingMust have long-term existing baskets to new drying

compatibility with processchemical drying agent

Size, strength relative to Adequate size allows flexible Overloading baskets canproduce load processing mechanically damage

fixturing, producinginadequate drying

Part rotation Achieve thorough drying Parts can be mechanicallyDecrease drying time damaged during rotation

particularly when notimmersed in solvent

Sample positioning Decrease drying time Improper positioning can causewater or solvent to remaintrapped

FIXTURING

Appropriate fixturing and positioning of the components cannot be overemphasizedin achieving adequate drying. This applies to aqueous, semiaqueous, and solvent systems.If the parts are not adequately exposed to the drying agent or the drying environment,there will be problems. In aqueous systems, if the parts are not adequately exposed, watercan recondense on the parts, and drying may not be achieved within a reasonable time.

Designing the fixturing system is very process specific, and the difficulties are not lim-ited to aqueous processes. Time and again, the author has observed that in the first trial runof even very well-contained, sealed solvent systems with vacuum drying, after processingthe first batch of product, visible solvent is present. This does not mean that the system isperforming poorly. Often, it means that the parts have to be positioned so that the solventflows out of the part. Part rotation can be helpful in this regard.

Some of the factors may be summarized as in Table 1.

THE HUMAN FACTOR

Employee education and the input and feedback of production personnel, particularlytechnicians, are critical in optimizing the drying systems. This factor is often neglected insolvent drying processes. In manual systems, there is the tendency for the operator to speedup the degreasing process by dunking the parts in the solvent, yanking the basket out, hold-ing it briefly in the now-destroyed vapor zone, and then racing across the floor while rap-idly shaking the parts. This phenomenon is popularly referred to as the “dunky-do”approach. To counter this, employee training and, more important, employee education arecrucial.7


Certainly, good equipment design and system automation are important. The increas-ing emphasis on cost-containment and rapid manufacturing tends to counteract even themost thoughtfully designed solvent cleaning system. All too often, this author hasobserved that a relatively sophisticated cleaning system will be purchased and cleaningand drying times set appropriately. The employees are then “trained,” and manuals out-lining complex procedures are installed. Then, within a few weeks, there are complaints ofsolvent loss, solvent odor, and inadequately cleaned parts with pockets of solvent. Aninvestigation usually reveals that someone (typically someone on that infamous third shift,the equivalent of “the butler” in mystery stories) has taken shortcuts.

In aqueous processes, employees may remove the parts from the drier sooner thanwould be desirable, or even skip the drying step completely. They may contaminate theparts by attempting to speed up drying using air hoses, or they may damage parts byincreasing drying temperature.

Typical undesirable shortcuts in both aqueous and solvent systems include:

• Speeding up the drying cycle• Shortening the entire process• Overloading baskets of parts

The importance of employee buy-in cannot be overemphasized. One may think one haspurchased a fail-safe, totally automated system. In the author’s experience, the creativeand determined employee can override a fail-safe system and destroy process controlalmost as quickly as it takes the average 4-year-old to remove a childproof medicine cap.

Usually, employees have an immediate reason for shortcutting the system, and it istypically related to total processing time and immediate perceived profits. An employeemay intellectually understand that a new aqueous or solvent process will take, for exam-ple, 20 min for complete cleaning and drying. If the old process only took 10 min, or if therehas been a sudden increase in production volume, that employee may independently oreven under direct orders from the supervisor decide to cut back on the process time. Moreoften than not, this means cutting back on the drying time on the grounds that drying is anextra, nonessential part of cleaning. Often, in cutting back on drying, the employee hascompromised the most important step. It is critical to ask, are we unwittingly rewardingsupposed efficiency at the expense of adequate process control? The only way to combatthe problem of shortcutting the drying step is to reward good process control to at least thesame extent as rapid production rates.

CONCLUSION

In summary, drying is a critical, time-consuming portion of the cleaning process foraqueous, semiaqueous, and solvent cleaning.8 It is also a part of the process that is over-looked and underfunded on a regular basis. Once the question of how dry is dry enough hasbeen answered, it is important to follow through with appropriate choices in drying methodand fixturing, and then with appropriate ongoing employee training and education.


REFERENCES

1. Kanegsberg, B., Abbink, B., Dishart, K.T., Kenyon, W.G., and Knapp, C.W., Development andimplementation of non ozone depleting, non-aqueous high precision cleaning protocols for iner-tial navigation subassemblies. in Microcontamination ‘93 Proceedings, San Jose, CA, 1993.

2. Kanegsberg, B., Mallela, H., Dominguez, H., and Kenyon, W.G., Integrating precision de-oilingand defluxing processes in high volume manufacturing systems, in IPC Proceedings, San Diego,CA, 1995.

3. Kanegsberg, B., Cleaning for biomedical applications, in Proc. Precision Cleaning 97, Cincinnati,OH, 1997.

4. Kanegsberg, B., Cleaning systems for low flashport solvents, Precision Cleaning Mag., 3(3), 21–28,1995.

5. Carter, M., Andersen, E., Chang, S.C., Sanders, P.J., and Kanegsberg, B., Cleaning high precisioninertial navigation systems a case study and panel discussion in Proc. Clean Tech ‘99, Rosemont,IL, 1999.

6. Ohkubo, M., An airtight argument: vacuum solvent cleaning systems work, Precision CleaningMag., 7, 24–29, 1999.

7. Petrulio, R., and Kanegsberg, B., Back to basics: the care and feeding of a vapor degreaser withnew solvents, in Nepcon West ‘98, Anaheim, CA, 1998.

8. Seelig, S., Adequate water removal for aqueous based operations presented at Tenth AnnualWorkshop on Solvent Substitution, Scottsdale, AZ, 1999.

9. U.S. EPA, National Emission Standards for Hazardous Air Pollutants, Fed. Regis., December 2,1994. 40 CFR Parts 9 and 63 (AD-FRL-5111-3) RIN 2060-AC31, available at http://www.epa.gov./fedrgstr/EPA-AIR/1994/December/Day-02/pr-184.html


http://www.epa.gov./fedrgstr/EPA-AIR/1994/December/Day-02/pr-184.html

http://www.epa.gov./fedrgstr/EPA-AIR/1994/December/Day-02/pr-184.html

CHAPTER 2.19

Aqueous Parts Drying

Daniel J. VanderPyl

CONTENTS

Definition of Drying for Aqueous PartsHistorical Perspective on the Drying ProcessTypes of Physical Drying Techniques for Aqueous Cleaning Technology

Centrifugal Spin DryingDesiccant Bulk DryingForced Air Drying without HeatForced Air Drying with HeatHigh-Velocity Air Blowoff (Compressed Air and Blowers)

Compressed AirBlowers

Low-Velocity Air DryingRadiant HeatSpot Drying (Vacuum Hoses and Compressed Air Blowoff Nozzles)

Vacuum HosesCompressed Air Blowoff Nozzles

Vacuum Chamber DryingIntegration of Drying Systems with Cleaning Systems

DEFINITION OF DRYING FOR AQUEOUS PARTS

There are two primary categories of parts-drying technologies for aqueous cleaningsystems. There is chemical displacement drying, where typically organic solvents are usedto displace water from a component, and physical drying, the topic of this chapter. Thischapter focuses on the most widely used methods, i.e., those that use an exchange of airthroughout the component by various means to blast, strip, or evaporate moisture from theexterior and interior of a component.

Air is the primary element interacting with the moisture on the surface of the compo-nent. Air drying is therefore defined as any mechanical action to remove water/moisturefrom the surface that does not rely solely on natural evaporative drying. The followingpages delineate several types of air-drying systems with the related applications and the


impact of component design on dryness standards, production rates, operating costs, andworker safety.

HISTORICAL PERSPECTIVE ON THE DRYING PROCESS

The continuing evolution of manufacturing technology has led to a wide range of pur-pose-built machinery for automated and semiautomated processes in parts cleaning anddrying. Prior to today’s prominence of aqueous cleaning technology, those using sol-vent/chemical cleaning and degreasing processes were unconcerned about drying. Often,the process resulted in cleaning and subsequent evaporation simultaneously. The naturalprocess of chemical cleaning resulted in a clean, spot-free, generally high-quality part,although maintaining product cleanliness was complicated by a continuous need to filterand replenish chemicals. Little was known about the health or environmental effects ofsuch chemical use.

Drying was an afterthought in the early aqueous cleaning systems. As such, approachesto drying consisted of a wide variety of methods that were often inefficient or ill suited tothe application. As the impact of various drying methods on product cleanliness becameapparent, the focus on drying quickly sharpened. The 1987 Montreal Protocol initiated thephaseout of ozone-depleting chemical cleaning processes, and its signers began a move-ment that took aqueous cleaning from a cleaning option to a cleaning standard.

The printed circuit board industry led the way. It pioneered new aqueous cleaningequipment and effective chemical replacement technologies. Throughout the 1980s, prod-uct quality in the printed circuit board industry was the driving force while manufacturingefficiency was a much lower priority. Today, this scenario has reversed itself in the indus-try and most other high-volume production environments; process costs and throughputrates are now scrutinized just as much as quality.

Currently, aqueous cleaning systems are effective and able to match the cleanliness ofmost chemical-based cleaning systems. Customer knowledge of cleaning/dryingprocesses has increased tremendously. Customers have learned that cleaning and dryinggo hand-in-hand; one cannot exist without the other. Today, there is more specific dryingrequired than ever before. As always, the requirements of the drying system are intimatelyconnected to the needs of the subsequent manufacturing steps whether they be partsinspection, component assembly, painting, ink jet coding, or packaging.

Equipment manufacturers have developed machinery and technology to match theincreasingly stringent cleanliness requirements. The U.S. military specifications (mil-specs)were a driving force in many of these refinements, which then spurred equipment manu-facturers to identify drying as the weak link to total product cleanliness.

With total product cleanliness the focus, process costs soared. Drying systems werecumbersome, operating costs were high, and throughput capacity was poor. As the worldeconomy continued to evolve and prices in the technology sector (i.e., computers and com-munications equipment) dropped, competitive forces dictated that improvements in manu-facturing efficiency were the key to tomorrow’s profits. Thus evolved a menu of dryingoptions. The secret of their best use lies in matching the appropriate technology to the clean-ing method and to the part being cleaned to attain the desired levels of cleanliness and pro-ductivity at the least cost.


TYPES OF PHYSICAL DRYING TECHNIQUES FOR AQUEOUSCLEANING TECHNOLOGY

The following drying systems will be covered in this chapter.

• Centrifugal spin drying• Desiccant bulk drying• Forced air drying without heat• Forced air drying with heat• High-velocity air blowoff

—Compressed air—Blowers

• Low-velocity air blowoff• Radiant heat• Spot drying

—Vacuum hoses—Compressed air blowoff nozzles

• Vacuum chamber drying

There are many good options to achieve efficient drying. Making the optimal choice is dif-ficult. Just as there is no ideal, one-size-fits-all cleaning technique, the drying system mustbe suited to the components and to the overall manufacturing processes. Therefore, thepros and cons, the applications and misapplications, are indicated.

Centrifugal Spin Drying

Centrifugal spin drying involves high-speed centrifugal spinning of up to hundreds ofbatched components to remove moisture from component surfaces following cleaning,cooling, and plating. Commercially available spin dryers are rated by loading capacity andrange from one to several hundred pounds per basket. Components are transported in abasket and either manually or automatically hoisted from the final rinse process to the cen-trifugal spin dryer. Cycle time can last from 2 to 20 min, depending on component com-plexity. At speeds up to 1500 rpm, care must be taken not to damage precision or delicatecomponents.

Ideal Application. Centrifugal drying is best suited to components of less than 1 in.3 insize having simple geometry and few if any blind holes or crevices that do not require crit-ical drying, such as those on plating lines.

Misapplication. The technique is not well suited for drying complex machined compo-nents requiring absolutely dry and spot-free surfaces.

Advantages. Advantages include low operating cost and a small footprint for batchapplications with adequate dwell time at the drying step. In addition, the noise level is low.

Disadvantages. On the other hand, because baskets must be constantly loaded andunloaded, the method is relatively labor intensive. With heavily loaded baskets, there is thepotential for worker injury. As production rates increase, the tendency is to load basketsmore and to add more spin-drying units, rather than upgrade drying technology.

Impact of Component Design. Centrifugal drying is more likely to become ineffective forparts with complex shapes or with densely packed batches. It can only be used for durablydesigned components.


Drying Effectiveness. Component geometry influences drying effectiveness in centrifu-gal spin dryers more than most other drying methods, and moisture retention can be aproblem. Forced air heating may be used to supplement drying capacity.

Drying Impact on Cleanliness. The potential for component damage makes spin dryingunsuitable for precision components. Clean components wearing against one another canresult in surface damage, deformation, or dislodging of particles, thus defeating the clean-liness requirement of more critical applications.

Equipment Menu. The primary equipment required is a spin dryer (0.5 to 7.5 HP).Larger dryers equipped with heated air circulation are also available. Ancillary equipmentincludes an automated or semiautomated hoist.

Associated Equipment Requirements; Labor Costs. Loading and unloading time can makelabor costs significant. The number of components per batch impacts the energy cost percomponent dried. As with all forms of drying, there is an optimum point of effectivenesswith the spin dryer; overloading the spin basket leads to diminished drying effectivenessand increased per part drying costs.

Safety and Environmental Issues. The greatest potential issue for employees is the phys-ical handling of baskets of components, both during loading and unloading of the cham-ber and emptying the basket of components for the next process.

Desiccant Bulk Drying

Desiccant bulk drying is one step better than nature’s own evaporative process.Desiccant material draws moisture from ambient air and is often used as both a curing anda drying process. Storage in a large chamber or room with desiccant material would be thefinal drying step to achieve as close to zero moisture content as possible. Frequently, it isused for removing moisture from compressed air sources and in large batch drying of agri-cultural products. It is the least often used method to dry manufactured parts.

Ideal Application. Desiccant drying is ideally suited for products where a moisture con-tent of 1 to 5% as a portion of total weight is needed for maximum part stability or shelf lifeand where heat or other aggressive drying methods may be detrimental to the parts. Forexample, any assembly with a latex component may benefit most from desiccant drying.

Misapplication. Desiccant drying is not appropriate for a continuous manufacturingprocess where drying cycle time must be kept to a minimum. It is not suitable for applica-tions where absorbed moisture is not the issue.

Advantages. It is much easier to control the uniformity of dryness among all the com-ponents with desiccant drying than with radiant heat or forced air drying. Radiant heatand forced air create different rates of drying within the same batch or process.

Disadvantages. The drying chamber typically has a significantly large footprint, thecycle time is long, and labor costs associated with product handling are high.

Impact of Component Design. Desiccant drying performance is impacted by the porosityof the product. The more porous, the longer it will take to draw out the moisture. Typically,the goal is to reduce the moisture content of a given part to between 1 and 5% of moistureas a percentage of the weight of the component.

Drying Effectiveness. The level of drying using this method is measured by the humid-ity in the chamber. Desiccant drying in large bulk processes makes it easy to measure thelevel of retained moisture.

Drying Impact on Cleanliness. Outside elements that could compromise the level ofproduct cleanliness are not usually a factor. The cleanliness of porous products has more todo with the forming of the product, and water may be part of that process. For example, inmolded parts such as latex, water is an integral part of product molding and additionalcleaning is not a required process.


Equipment Menu. The primary equipment consists of a sizable container or room, from1 to 1000 ft3 in volume. Ancillary equipment includes a fan for air circulation and a dehu-midifying unit.

Associated Energy and Labor Requirements. Throughput is lowest with desiccant drying.Labor rates are among the highest because product must be manually moved in and out ofthe chamber or room.

Safety and Environmental Issues. Desiccant drying itself poses relatively few worker haz-ards. Hazards associated with product composition are a potential problem.

Forced Air Drying without Heat

Forced air drying (without heat) is based on the exchange of large volumes of airthrough an enclosed zone to extract ambient moisture. It is an accelerated evaporationprocess that can also use heated air. A dehumidification unit can shorten the drying cycle.

Ideal Application. The most effective use is with simply designed components that havesurface and/or core temperatures higher than ambient. By drawing away air with forced airdrying, humidity and moisture are removed from all but the surface, encouraging moreevaporation. Without air circulation, such a heated product would make the whole chamberhumid. For example, in the tire industry, using a Banberry rubber extrusion process, a con-tinuous ribbon of rubber product exits at 350°F. It is then immersed in a cold water quench.Forced air accelerating the evaporative process can help remove the water on the rubber strip.

Misapplication. Forced air drying is not well-suited to batch-drying applications wherepart configuration and part stacking result in numerous water pockets within the batch. Insuch situations, the drying time may increase unacceptably.

Advantages. Initial capital outlay is minimal, as no specialized equipment is needed.Noise levels are low, and environmental impact is minimal.

Disadvantages. The throughput rate for forced air drying without heat is low, and thedrying zone may be large, resulting in a large footprint. There can be problems in processcontrol. For one thing, there can be difficulty in controlling drying effectiveness whereambient temperatures may fluctuate. In addition, product quality may be adverselyimpacted due to the potential for part recontamination from unclean air.

Impact of Component Design. Typically, forced air drying is used only for simplydesigned components at low production rates.

Drying Effectiveness. Forced air drying is effective for industrial drying. For example, itcan dry rubber products well enough to eliminate potential voids in the stamping or mold-ing phase of production. However, forced air drying without heat is ineffective for partswith complex surfaces requiring a high drying standard.

Drying Impact on Cleanliness. Forced air drying can contaminate the component if theair source is unclean. However, cleanliness is a function of the total manufacturing process,and forced air drying is not used in applications involving critical cleaning.

Equipment Menu. The primary equipment consists of a low-pressure fan assembly adjacentto the product, most commonly an axial or box fan simply blowing large volumes of air acrossthe surface. Ancillary equipment includes a dehumidification unit to decrease drying time.

Associated Energy and Labor Requirements. Where the process is conveyorized (as in mostapplications of this method), labor costs are low. However, forced air drying may also be usedfor batch cleaning, where labor costs may be significant. Energy costs are low since blowersuse much less energy than heaters.

Safety and Environmental Issues. Worker safety and environmental concerns are mini-mal, since humid air can be expelled with a roof-mounted exhaust fan. Noise is typicallynot a problem.


Forced Air Drying with Heat

In such systems, components are indexed into a drying zone using a circulating fanthat warms air up to 200°F above ambient temperature. The heat source is most oftenelectric, but natural gas-fired heat and waste steam are also used. Overhead conveyor sys-tems also use forced air drying with heat prior to electrostatic powder paint zones.

Ideal Application. Forced air drying with heat is the method commonly used in batchultrasonic cleaning systems. Using sliding beam systems, baskets of components are trans-ferred into drying chambers after the final rinse. The forced air with heat provides an accel-erated evaporative drying process.

Misapplication. Misapplications of forced air drying are related to the thermal stabilityof the component as well as to required process time, process flow, and parthandling/worker training problems. In a continuous conveyorized process, forced air dry-ing with heat can be either a benefit or detriment depending on what happens after the dry-ing process. The time taken to stabilize the parts at room temperature is the price paid todry those parts with heated forced air. Attempts to speed up the process can cause prob-lems. For example, in a batch type paint process of cleaning/drying/painting, componentsare often manually handled in inspection or assembly, and high-temperature parts mayburn workers. In addition, increased drying temperatures cause thermal expansion of thecomponent, pushing a precision machined part out of inspection tolerances.

Advantages. Forced air drying with heat is versatile and can be used over a wide rangeof part and component sizes. Because the air is mixed evenly, the method results in uniformdrying. Compared with desiccant drying, it can dry relatively large batches of parts moreeffectively in a much shorter period of time. In some applications, the waste heat fromother processes can be used to create steam for forced heat drying.

Disadvantages. The drying cycle time is directly proportional to variations in batch sizeand component orientation. If an elevated core temperature is required, drying time andcost will increase. Parts may become too hot to handle in the next step of production. Heatmay expand components beyond inspection tolerances. Heat may bake on contaminantsdepending on the particle count in final rinse stage.

Impact of Component Design. Particularly in ultrasonic batch cleaning systems, batchsize component geometry and orientation in each basket must be consistent for repro-ducible results. Complex, ornate components with blind holes present the most difficultchallenge in maintaining drying/cleanliness consistency.

Drying Effectiveness. For both batch and in-line continuous processes, variations incomponent design, volume of parts per batch, drying rate, and any thermal constraints ofthe product directly affect drying quality.

Drying Impact on Cleanliness. Forced air drying with heat results in accelerated evapo-ration of moisture from components. Any particulate suspended in the surface liquid willultimately be baked onto the surface of the component as it is dried. Therefore, the waterquality of the final rinse zone in terms of dissolved contaminants and the filtration neededto maintain water quality and eliminate particulates are integral parts of final productcleanliness.

Equipment Menu. The primary equipment required includes a squirrel cage fan/blower(high volume, low pressure) and an in-line duct heater. The equipment is predominantlyelectric. However, natural gas may be used in a continuous conveyor system. High effi-ciency particulate arrestance (HEPA) filters are commonly used for some precision and allcritical cleaning applications. For batch processes, the drying zone is typically an enclosedchamber placed subsequent to the ultrasonic cleaning and rinsing tanks.


Associated Energy and Labor Requirements. Process costs depend on the drying cyclerequired for each component. Process costs for batch processes can be considerable. Forexample, running 10 batches a day over an 8-h day requires a fan and a heat source. Thelowest temperature likely to result in effective drying would be approximately 200°F. Ifproduction levels were to increase to 20 batches a day, to achieve the same level of dryingeither the temperature would have to increase or the number of drying chambers wouldhave to be doubled. While the energy consumed per component dried might not increase,total energy usage would likely increase considerably. In addition, for batch processes,labor costs are incurred for loading and unloading. Lag time for cooling may add extra han-dling and labor costs. In-line applications generally reduce labor costs. However, conveyorlength and lag time in the production flow may involve added costs. The throughput ratecan be anything desired within the limits of the temperature constraints of the parts. Theutilizable temperature becomes the limiting factor.

Safety and Environmental Issues. Workers must be protected against burns from the heatchamber and from heated components. Environmental impact can vary according to theapplication.

High-Velocity Air Blowoff (Compressed Air and Blowers)

A high-velocity air blowoff is any air stream directed at the surface of a product to cre-ate a sheer force that strips liquid from the product. It is one of the most widely used dry-ing methods. The air source can be generated from high-pressure plant air systems,compressed nitrogen, or self-contained blower systems. High-velocity airstreams can gen-erate static charge. While the moisture being removed from the part counteracts this ten-dency, static buildup in parts-drying applications is of concern.

Compressed Air

Compressed air systems are defined as drying systems using a compressed air gener-ator with a minimum 10 psi (pounds per square inch) and supplying air through air knivesand nozzles.

Ideal Application. Compressed air is best for the small-scale drying of parts. Theapproach is practical for parts of less than 6 in.2 of cross-sectional area and traveling singlefile at less than 5 ft/min.

Misapplication. Compressed air, when used to dry components with more than 36 in.2

of surface area on any one side, will generally result in very high operating costs. Also,components with critical cleanliness requirements must be protected from oil and conden-sation produced by the compressed air system.

Advantages. Compressed air systems of adequate capacity are often already in place.Equipment is available from a range of suppliers and the systems consume relatively littlespace. Air knives and nozzles are compact; piping systems are generally less than 1 in. indiameter.

Disadvantages. Energy consumption can be up to 75% more than with blowers. Becausemost compressors are oil lubricated, filtration of the airstream is required to prevent recon-tamination of the parts. In addition, compressors condense liquid from the airstream,resulting in part contamination. Finally, compressed air produces a low-frequency noisethat is audible at considerable distances from the blowoff zone. Workers may find the noiseuncomfortable or unacceptable.


Impact of Component Design. Component design and fixturing considerations duringcleaning and drying impact the choice of compressed air vs. a blower system. As a rule ofthumb, compressed air can be effectively used for drying parts measuring less than 6 � 6in. of cross-sectional area with blind holes and crevices that measure less than 1/2 in. diam-eter with hole depth a minimum five times hole diameter. Components with large, smoothsurface areas and simple geometry are poor candidates for compressed air, as operatingcosts will skyrocket.

Drying Effectiveness. The ability of air to strip moisture from complex or critical surfacesrelies on the air having a straight path to the respective surface or area. A part can be easyto dry, but if parts are stacked ten on a rack, airflow and effectiveness are restricted.Moisture condensation is a potential problem.

Drying Impact on Cleanliness. Compressed air may carry particles of oil and dirt thatwould compromise precision cleaning. Also, condensation from the high-pressure air canrecontaminate the parts with moisture. Therefore, compressed air is typically used forindustrial processes, but not for precision cleaning.

Equipment Menu. The primary equipment consists of an electrically operated dryer, anoil separator, a piping system, a compressed air knife or nozzle, available plant air, and areceiver tank, which is placed as close to the blowoff point as possible. If the drying processis not continuous, this tank can be used as a reservoir for compressed air. Ancillary equip-ment includes an enclosure, tunnel, or chamber where blowoff can take place, to reduceambient noise and provide an opportunity to exhaust moisture-laden air as the product isdried.

Associated Energy and Labor Requirements; Additional Cost Considerations. A wide range ofcompressors and accessory items, nozzles, knives, and other blowoff devices are available,but the end user is generally left to decide how many of what configuration are required.It is important to note that compressor costs are commonly higher than for blower systemsand use approximately 75% more energy. The decision whether to use compressors orblowers must be made on a trade-off basis, considering existing resources and return oninvestment (e.g., is sufficient compressed air already available?). In general, if neitherblowers nor compressors are currently available, a blower will generally cost less in termsof initial cost and operating cost. In addition, for more exacting processes, while com-pressed air is available with very stringent air-drying capabilities, in most cases the cost offilters, separators, oil filters, and energy use is prohibitive. Finally, for larger parts or preci-sion cleaning, where workers must carefully and individually clean parts, added laborcosts must be considered.

Safety and Environmental Issues. Noise is a problem with compressed air. It generallyoperates at a lower frequency range than high-velocity air blowers, and this allows soundlevels to travel greater distances than blower system air sounds. For example, at 100 ft, ablower system might read in the high 70 dbA range. A compressed air nozzle at 90 dbA,located 100 ft away, would still generate 85 dbA. The high-frequency whistling noise thatoccurs as air is blown off surfaces causes random spikes of impact noise that often exceed100 dbA. The cost of the materials capable of absorbing lower-frequency sound levels isgenerally very high. Small bits of debris blown from blind holes can be a worker safetyproblem, and workers should wear protective goggles. Sound levels often require hearingprotection. Contaminants may be blown into the environment, and condensation can blowinto associated work areas, causing slippery floors or fouling of work sites. If the systemscannot be adequately enclosed or ventilated, operators must also be protected from breath-ing atomized contaminants.


Blowers

High-velocity air blowoff can be achieved using a dedicated blower producingbetween 0.5 and 5.0 psi, with the air directed through air nozzles or air knives. Blower sys-tems with optional infrared lamps are the most widely applied method of drying compo-nents in conveyorized precision cleaning and plating systems. In addition, high-velocityair is often used for critical cleaning applications such as semi-conductor components andmedical devices.

Ideal Application. High-velocity air blow-off with air blowers is most effective with anycomponent size greater than a 6 � 6 in.2 surface area, having simple to moderate surfacecomplexity, and for production rates greater than 100/h.

Misapplication. Blowers are not suitable for extremely small components, complexgeometries and blind holes, low production rates, and short cycle times. Process equipmentdesign may restrict nozzle or air knife access, necessitating small piping and unacceptablyhigh pressure. However, adding heat can increase drying capability.

Advantages. Blowers can be sized to accommodate a wide variety of components andproduction rates . High-velocity blowers provide the most energy efficient of blowoff airmethods. The heat of compression assists in the drying. Air can be supplied free of oil andmoisture, and self-contained air delivery without air pressure fluctuations is achievable.

Disadvantages. Blowers cannot remove liquid from complex parts with unexposed sur-faces. Exhaust air requirements are much higher than for compressed air. Large 2- and 3-in.-diameter feed and connecting piping for air knives and nozzles takes up additionalspace.

Impact of Component Design. Small intricate components and those with blind holesmore than three times the depth of the hole diameter are not dried effectively.

Drying Effectiveness. As compared with compressed air, the greater range of air volumethat can be used in blowers allows greater versatility.

Drying Impact on Cleanliness. Sealed bearing blowers can deliver oil-free air. However,any oil-lubricated blower is vulnerable to seal failure resulting in contamination of theentire cleaning system. On a positive note, because the high velocity strips away all mois-ture, the system does not create enough pressure to produce condensation that mightrecontaminate the part. In addition, use of high-impact air effectively eliminates baking onof particulates entrained in surface liquids.

Equipment Menu. The primary equipment consists of a self-contained motor assemblyblower, an air filter/silencer assembly to filter out ambient air and dirt particles, and airknives. Ancillary equipment includes electric in-line heaters, a recirculating blower airwith piping and filters, and exhaust fans connected to drying chambers. The option of arecirculating blower air should be considered to cut process costs and to avoid contamina-tion. Air introduced into a drying chamber must be properly exhausted to avoid introduc-ing heat, moisture, and contaminants into the surrounding work environment. The heatintroduced can be comparable to leaving a door open in a factory all day long.

Associated Energy and Labor Requirements. High-velocity air blowers are the most effi-cient means of forced air blowoff. While little labor is associated with the drying process,complex parts may require spot blowoff following exit from the drying chamber.Throughput rates for blower systems are generally high and and are based on matchingblower size to throughput demands. Assistance of an experienced equipment vendor canbe very helpful in giving specific application advice in designing the system and evaluat-ing the appropriate size and/or number of blowers to support demand.


Safety and Environmental Issues. Noise with high-velocity blowers can be significant.The blower system operates at the higher end of the audible frequency range, i.e., 90 dbA.But at 100 ft, that same blower system might read in the high 70 dbA range (every point isa multiple of 10, therefore, there is a tremendous difference between 85 and 80 dbA). Properenclosure can reduce sound impact. In-line heated air or recirculated blower air canincrease surface temperature of the components being dried. Subsequent handling byoperators must be considered.

If air from the blower is not recirculated, an exhaust system must be installed. Controlsdepend on specific regulatory requirements.

Low-Velocity Air Drying

Low-velocity air-knife drying systems involve large volumes of air introduced throughan air-knife plenum exiting at velocities no greater than 10,000 ft/min. Although initialvelocities are low, because of the wider air path and greater volume of air, the drop in pres-sure is lower compared with other systems. This effectively results in a longer air path.

Ideal Application. The ideal application for low-velocity air drying is conveyorizedcleaning of components with simple geometries and with low throughput rates. In theearly days of aqueous cleaning, parts cleaners frequently used low-air-velocity drying sys-tems. They worked well because the component geometry was simpler and throughputrates and conveyor speeds were slower. The low velocity was effective for many applica-tions, as the dwell time within the drying zone was correspondingly much greater thanwith the higher production speeds of today. Today, higher-impact velocity will compensatefor higher throughput rates by creating greater shear force for more effective dryingdespite shorter dwell time.

Misapplication. Despite advances in the technique, low-velocity drying is not readilyadaptable to parts with complex geometries. Low-velocity air drying is unsuitable forhigh-speed throughput of anything but the smoothest of surfaces.

Advantages. Advantages include low initial capital outlay, low operating costs, lowmaintenance, and low noise levels.

Disadvantages. The method is slow and is ineffective with complex parts. Equipmentsize is large, particularly relative to drying capacity.

Impact of Component Design. Simple component designs are best suited to low-velocityair drying.

Drying Effectiveness. Low-velocity air drying is effective for simple parts at low speeds.It can work effectively at higher throughput rates if critical levels of drying are not requiredor with supplementary in-line heating.

Drying Impact on Cleanliness. Filters must be increased appropriately to filter air effec-tively while minimizing pressure drop. The design also does not lend itself to overcome thepressure drop associated with high-capacity inlet filters or in-line HEPA filtration.

Equipment Menu. The primary equipment is a centrifugal fan assembly (1 to 10 HP), anair distribution manifold, flex-hose and piping, and air-knife plenum. Ancillary equipmentincludes a shroud enclosing the air-knife assembly with an optional hood to draw mois-ture-laden air from air-knife zone.

Associated Energy and Labor Requirements. While total energy consumption of low-veloc-ity air knives is low, energy usage per component tends to be comparable with higher-velocity systems with higher production rate equipment. In addition, with supplementaryin-line heating, low-velocity air drying then becomes an energy-efficiency problem whencompared with other types of technology and may become an issue.


Safety and Environmental Issues. Even with the low-velocity system, undesirable aerosolsmust be vented and, if required to meet environmental regulations, appropriately trapped.

Radiant Heat

Radiant heat is the process (typically in-line) by which a heat source (generally infraredtube lamps, IR) is used to flash-dry very thin layers of moisture on the surface of compo-nents. It is often used after high-velocity systems that remove the bulk of the liquid. Insome batch processes it is used for final drying following forced air drying.

Ideal Application. The ideal application is for final drying of parts with complex geome-tries where air knives may not completely dry the parts.

Misapplication. IR heat is inappropriate as the primary or sole drying source for prod-ucts having more than a trace amount of liquid residue.

Advantages. Advantages include relatively low capital outlay and the absence of noiseor environmental concerns.

Disadvantages. When used without initial drying to remove gross moisture, operatingcosts and cycle time can be high. The footprint for exclusive IR drying can be high.Dissolved or suspended contaminants can bake onto surface, and excessive heating canresult in throughput, handling, and inspection problems.

Impact of Component Design. Complex shapes and surfaces with close dimensionalclearance between components in an assembly, i.e., circuit boards, are best suited to the IRdrying method. IR provides better control and greater simplicity.

Drying Effectiveness. An IR heater element, like any heater element, is either on or off.The cycle time of the element determines the amount of radiant heat produced. It can bedifficult to maintain consistent heating.

Drying Impact on Cleanliness. When IR is the second stage of drying to an air-knife sys-tem, most contaminants entrained in the surface liquid are carried away with the air veloc-ity. However, if the percentage of surface moisture is more than 5% or the level of entrainedcontaminants is high, the likelihood of undesirable residue increases.

Equipment Menu. Primary equipment consists of IR heater tubes in a shielded orshrouded chamber with insulation to minimize surface heat on the exterior of the chamberalong with a heater control system to ensure even element cycle times. Ancillary equip-ment consists of initial drying with fans, blowers, and/or air knives.

Associated Energy and Labor Requirements. IR heater elements used as the only dryingmethod are likely to result in excessive electrical demands. But as a second stage, they canbe a very energy efficient drying method for complex components. Automated processesare available.

Safety and Environmental Issues. If components are raised above 125°F, part handlingissues must be considered. If preceded by adequate rinsing, there are no obvious environ-mental issues related to IR drying.

Spot Drying (Vacuum Hoses and Compressed Air Blowoff Nozzles)

A wide range of products have drying requirements for specific areas of the product,but do not necessarily require a completely dry part. In such cases, capital outlay may notbe justified relative to labor costs.


Vacuum Hoses

Localized vacuum drying is a manual process used for very low production rates, orfor component inspection where only certain portions require drying. A vacuum dryingsystem could be a central system or a dedicated vacuum unit.

Ideal Application. The ideal application is for simple components in industrial-gradecleaning.

Misapplication. Examples include complex parts and blind holes where air cannot beexchanged adequately to draw liquid to the vacuum source, or critical applications wherecontact with the vacuum nozzle could compromise cleanliness.

Advantages. Advantages include low equipment and energy costs. The technique mayavoid scattering of large amounts of moisture and contaminants. In addition, it may be pos-sible to combine the drying and inspection steps.

Disadvantages. Contact with the vacuum hose and/or bristled pickup nozzles mayresult in contamination. Process control is difficult.

Impact of Component Design. Localized vacuum is ineffective with complex compo-nents.

Drying Effectiveness. Process control is operator dependent; effectiveness is variable.The greatest use of localized drying is in quality control operations adjunctive to criticalinspection of machined components.

Equipment Menu. Equipment consists of vacuum systems or portable vacuum units.Associated Energy and Labor Requirements. Vacuum units require manual handling of

each component. With low production rates, labor costs may be acceptable. The energyconsumption for a vacuum is 1 to 3 HP; energy per component is highly variable. A vac-uum source running continuously with sporadic parts spot drying may be more costly tooperate than an on-demand compressed air nozzle.

Safety and Environmental Issues. When using a portable vacuum, residual chemicalsmay cause problems as potential fire hazards.

Compressed Air Blowoff Nozzles

Compressed air blowoff is the most common spot drying method because of the avail-ability of compressed air and its ability to blow moisture from confined spaces.

Ideal Application. Spot drying with air blowoff is uniquely suitable for blind holes orcrevices where moisture is randomly trapped. The technique is used for low-volume, inter-mittent production and for inspection.

Misapplication. The most common misapplication is where compressed air nozzles areoperated by line workers repeatedly for a nonvariable product line. These applications aremuch better suited to high-velocity blowoff systems.

Advantages. The small handheld nozzles are very maneuverable for working with com-plex parts.

Disadvantages. The equipment is noisy and can scatter moisture and contaminant.Process control is difficult.

Equipment Menu. Equipment consists of localized blowoff nozzles or custom nozzleand tube assemblies connected to a compressed air line.

Associated Energy and Labor Requirements. The method is labor intensive. On-demanduse of compressed air generally minimizes the power requirements of the compressed airsystem applied for spot drying, unless the blowoff nozzles are operated on a continuousbasis.


Safety and Environmental Issues. The potential for debris or moisture to be sprayedthroughout the work area poses potential worker exposure issues (eyes, inhalation).Scattered contaminant on floors may produce slipping hazards. The equipment can pro-duce intermittent, unpleasant noises.

Vacuum Chamber Drying

Vacuum drying is the process by which components are placed into a sealed chamberwhere a vacuum is pulled on the component to lower the water vapor point. The moisturebecomes an aerosol in the chamber. A filtration system pumps the moisture from the air,returning dry air to the chamber to repeat the process.

Ideal Application. Vacuum chamber drying is best suited to complex geometries and toporous metal components that must be completely dried, such as prior to epoxy resinimpregnation. It is also useful for metals that must be completely dried to prevent corrosion.

Misapplication. Where the above considerations are not a concern, vacuum drying maynot be the most efficient or rapid approach.

Advantages. Because high heat is not required, vacuum drying is less likely to damagetemperature-sensitive components. Also, energy costs may be less for critical metal com-ponents.

Disadvantages. Equipment costs are relatively high; with manual systems, componenthandling (loading) increases labor costs and process times.

Drying Effectiveness. With proper equipment maintenance, the technique is very effec-tive for most components.

Drying Impact on Cleanliness. The process does not typically produce any contaminants.Equipment Menu. The primary equipment is a vacuum chamber of appropriate dimen-

sions. Ancillary equipment includes hoists and semiautomatic systems.Associated Energy and Labor Requirements. The vacuum pump and the air circula-

tion/moisture extraction system can require from fractional to 15 to 20 HP, depending onchamber volume. Labor costs must be included in considerations of manually loadedchambers.

Safety and Environmental Issues. The chamber itself, even though under high vacuum,generally presents no immediate worker safety issues. Loading and unloading of thechamber must be done to minimize worker injury.

INTEGRATION OF DRYING SYSTEMS WITH CLEANING SYSTEMS

Integration of drying with the aqueous cleaning process is determined, in part, by thetype of system employed, in-line vs. batch cleaning. In conveyorized in-line cleaning,cleaning occurs in one zone, rinsing in a second zone, and drying in the last zone. Batchcleaning may be single zone, where all steps take place sequentially in a single chamber; ormultiple zone batch, where cleaning, rinsing, and drying steps occur in separate chambers.

Integration is also dependent on the level of drying required. Industrial dryingdemands only visual cleanliness. If it looks clean, then it must be. This level of drying isassociated with a broad range of heavy industry manufacturing requirements. Precisioncleaning is the standard in a wide array of manufacturing, from metals to electronics,where surface contaminants remaining after cleaning and drying are measured by weight,electrical conductivity, optical scanners, and other methods. Critical cleaning has the moststringent cleaning and drying needs. Products made with critical cleanliness standards are


generally those of the highest-technology sector. Any of these drying levels can be achievedusing any of the aforementioned methods of throughput. A single-zone batchcleaning/drying process can work for automotive machine parts as well as for medicalfiber-optic components.

The key to effective integration of cleaning and drying is the correct evaluation of userneeds. This can best be done by starting at the end result (with cleanliness requirementsand throughput) and working backward to develop the proper equipment menu. Withtoday’s fast-paced changes in technology and production capacity, users must continu-ously revisit the process to validate that the method used today still meets current qualityand production criteria. Increases in production and throughput, the ability to measurecontaminants more accurately, and increasing understanding of the effect of contaminationlevel on the quality of the end product all drive cleaning/drying integration.

Unless a component is cleaned and dried in a bubble, where all elements are controlledto absolute values, the potential for contamination exists the moment the component entersa new environment. Critical cleaning applications are performed in clean rooms where thehumidity level is controlled, yet humidity can increase just from people breathing, com-promising the cleanliness of the part. Handling components through conveyorizedprocesses or having air inadequately filtered from the compressed air or blower sourcecompromises the cleaning method. Every engineer responsible for parts cleaning must askthe question, “Does the drying method maintain the results of the parts cleaning method?”If it does, then the integration of the drying and cleaning processes has been successful.


CHAPTER 2.20

Liquid Displacement Drying Techniques

Robert L. Polhamus, Steve R. Henly, and Phil Dale

CONTENTS

IntroductionWater-Soluble Displacement

Vapor-Phase ProcessDual-Vapor Drying SystemsLiquid/Vapor Process

Liquid DisplacementLiquid More Dense Than Water

Displacement Fluid with SurfactanDisplacement Fluid with Alcohol

Liquid Less Dense Than WaterSummary

INTRODUCTION

As the old expression goes, “The job ain’t over ‘til the paper work’s done.” This axiomcan also be applied to cleaning. A great deal of effort has been expended this far in deter-mining the appropriate cleaning chemistry and its suitability for the soil encountered andits environmental impact. In addition, the appropriate rinsing of the chemical has been dis-cussed. Having fully analyzed and tested the cleaning and rinsing aspects, the method ofdrying the part needs equally dedicated review.

It would appear obvious that if a simple cleaning chemistry can be used with a simplerinse system, an equally simple drying technique can be employed. Unfortunately, this isnot always the case. When evaluating a potential drying technology, one must rememberthat the product is clean prior to entering the dryer, and the dryer can only recontaminatethe product. It would be impractical to expend any amount of time in cleaning a partbeyond the capabilities of the drying system. As a result, it is critical that the drying tech-nology be matched to the application.

Various drying technologies offer advantages while suffering from some disadvan-tages. Any of the three basic drying techniques: evaporative, mechanical displacement, and


liquid displacement can be optimized to provide the proper balance of technical perform-ance and cost.

As in the evaluation of the cleaning and rinsing chemical, once a thorough investiga-tion has determined technical requirements of the application, several drying processes canbe evaluated on their technical merit to meet the needs of the application. Having deter-mined the acceptable drying technology, the equipment to implement the process is fairlyself-evident and, subsequently, the cost is appropriate to the need.

The use of liquid displacement dryers has historically centered around value-addedcleaning applications. These applications are those in which a measurable cleaning require-ment is established. In other words, there is a reason to clean the components that providesvalue to the product, whereas if they were not clean they would most likely not performadequately in the final application. As more users become more definitive in their cleaningrequirements, the need for precision drying opportunities will grow. In addition, moreproducts are being designed and fabricated to support the development of small geometryand close tolerances. Once water is entrapped in these devices, it becomes increasingly dif-ficult to remove it by any method.

Although liquid displacement drying is most closely associated with water removal, itcan also be used in areas of solvent cleaning. As with water-based cleaning, a multitank sol-vent system may have one or more chemicals that are good cleaning agents but, with lowvolatility, are difficult to evaporate. These products can undergo a rinsing phase with amore volatile chemical and, eventually, either the rinsing chemical or an additional solventis evaporated from the surface, leaving the product surface dry. For the purposes of thisdiscussion the removal of water from the substrate will be the subject, although these sameprinciples can be applied to solvent cleaning systems.

Two liquid displacement processes are commonly utilized. One process involveschemicals that are soluble with water. The second process uses chemicals that are insolublewith water, and these are characterized as either more or less dense than water. Theseprocesses will be described.

WATER-SOLUBLE DISPLACEMENT

Water-soluble displacement drying is primarily accomplished through two methods.The predominant technique is vapor-phase using isopropyl alcohol (IPA). Another processuses an immersion technology in cold high-purity water with an alcohol vapor layer in thesame process chamber. In either case, it is the molecular interaction between the two chem-icals that makes the process successful. IPA and water are perfectly miscible materials. Thatis, they will dissolve or take each other into solution in a nearly unlimited capacity. Afterthe drying process, the effluent from the system will be a homogeneous mixture of alcoholand water. As a result, the alcohol is considered “consumed” in the process and is inap-propriate for further drying unless the IPA can be purified and returned to the system.There are a variety of techniques suitable for this purpose, and they must be evaluated rel-ative to the potential payback based upon the application.

With a few exceptions, all liquid displacement techniques utilize batch processing. Thefollowing process descriptions center on the batch concept.

Vapor-Phase Process

The vapor-phase process is the most widely used of the liquid displacement processes.The equipment is simple in nature and the process relies very heavily on the principles of


Figure 1 Vapor-phase process.

SLIDING LID

CONTAINMENT COIL

IPA SUPPLY

COMPONENT

CATCHMENT TRAY

IPA

WASTE IPA / WATER

HEATER

physical chemistry. The equipment configuration consists of a single chamber where theentire process is performed. The main features of the process chamber are heating ele-ments, a vapor-containment cooling coil, a liquid effluent capture tray, a freeboard zone,and a robotic lift mechanism (Figure 1).

The equipment must be prepared to induce and contain the process of vapor displace-ment. Liquid IPA is introduced into the process chamber to a preset level. Cooling water iscirculated through the containment coils at an appropriate temperature and flow rate. If allconditions are acceptable, heat is applied to the liquid bath.

The IPA is heated until it reaches its boiling point. IPA vapors are generated from theboiling liquid and begin to rise upward in the process chamber. Through a combination ofcondensation on the sidewalls and a vapor density greater than air, the vapors move slowlyupward from the liquid level displacing air in a “plug flow.” The vapors will continue to riseupward in the chamber until they contact the containment coil. The containment coil willcause the vapors to condense on its surface, and as the vapors condense, more vapor isattracted to the coil to produce a continuous condensation action. Convection of the vaporstoward the coil will produce a limit or ceiling on the height to which the vapors will rise.

As the vapors condense, they are directed back downward along the sidewalls andreturned to the liquid volume. Upon return, the liquid is available to be regenerated intovapor. This process continues indefinitely in the process-ready or idle mode until a dryingcycle is initiated. Once the vapor zone has been fully established and stabilized, the unit isready to process work. Prior to the implementation of a vapor-phase process, careful


consideration should be given to product orientation and fixture design. Both factors musttake into consideration the accessibility of the parts to the vapor and the ability of the partsto drain freely. The critical nature of these factors will become self-evident from further dis-cussion of the process.

The process contains three zones: the liquid level, vapor zone, and freeboard area. Theliquid level is the amount of liquid IPA contained in the process chamber. In most cases thislevel is no more than a few inches deep. The second zone is called the vapor zone. This isthe area from the top of the liquid to the top of the vapor blanket at the point where it iscaptured by the containment coil. The containment coil is located a specified distance fromthe top opening of the process chamber. The third zone is the area from the top of the vaporzone to the top of the process chamber, called the freeboard zone. This dimension shouldbe greater than or equal to the height of the vapor zone. This freeboard plays a critical rolein the containment of the vapors and the cost-effectiveness of the system.

The work and fixture (load) are placed on the robotic hoist platform or suspended fromthe robot arm. The temperature of the load must be below the boiling point of IPA, and, ingeneral, the cooler the better. Once loading is complete, the process is initiated and the loadis lowered from the home position through the freeboard zone into the vapor zone. Thisbegins the first phase of drying called the displacement phase.

Once the load penetrates the vapor zone, it becomes a preferential condensation site forthe vapors vs. the containment coil, and the vast majority of the vapor condenses on theload. The rate of condensation on the coil observed during the idle mode will show signif-icant, if not complete, reduction upon insertion of the load. The IPA will condense and mixwith the water on the surface of the load. As the IPA liquid accumulates on the surface, itwill fall from the surface by gravity, taking dissolved water with it. In many cases, thewater that is on the surface of the load is high-purity deionized water, which has a rela-tively high surface tension. The IPA has a relatively low surface tension so it will “wet” thesurface and also displace water from the load. The liquid condensate and water fall fromthe product and are collected on a liquid effluent capture tray and directed to waste.

Through the combination of solvation and displacement, all water is removed from thesurface. After a period of time, based on water loading and vapor generation rate, the sur-face becomes water free. Even though the surface is water free, as long as the load is belowthe boiling point of the liquid, condensation will continue.

The water layer on the parts has been displaced by a layer of IPA condensate. As theIPA condenses, it transfers energy to the load. Over time, the load temperature begins torise. As the temperature approaches the boiling point of IPA, the rate of condensation slowsand eventually almost stops. This is the second phase of the process and is referred to asthe thermal equilibrium phase. This phase begins when the water is displaced from the sur-face and ends when the load reaches a near equilibrium with the vapor, where equal vol-umes of condensate and liquid are changing state on the surface. An indication thatthermal equilibrium has been reached is when the condensate dripping from the contain-ment coil has reached a rate equal to or close to the precycle idling rate.

At this point in the process the load is covered with a microlayer of IPA liquid.Although the surface may appear to be dry, this microlayer exists on the surface in equi-librium with the vapor. To “dry” the surface, the load is removed from the vapor zone intothe freeboard area. Once in the freeboard area, and even beginning during the transitionalphase from the vapor zone into the freeboard area, the microlayer is allowed to flash-dryfrom the surface.

The IPA vapors that flash from the surface are more dense than air and, if allowed toremain undisturbed, will fall back into the vapor zone. Residence time in the freeboard areawill allow these vapors to be recaptured as well as allow the load to cool prior to removal


from the system. After the freeboard residence has timed out, the load is returned to thehome position.

The vapor-phase drying system is quite effective but, as can be implied from theprocess description, works very well on certain product configurations. Since the vaporphase provides little if any mechanical agitation, the process relies strictly on the affinity ofIPA for water. If product or fixture configurations trap or hold water, the process is lesseffective. This is the most significant drawback to vapor phase and must be taken into con-sideration.

Vapor phase does provide two significant advantages over its competitive liquid dis-placement techniques. First, the process is extremely simple. It contains limited movingparts and simpler component design. It utilizes the laws of physics to transport the chem-ical in contact to the load and, in the vapor phase, has the ability to provide equal contactto all surfaces simultaneously.

The second significant advantage is that it can maintain chemical purity levels supe-rior to other techniques. The process generates contaminant-free chemistry by the produc-tion of the vapor zone. By definition, the vapor that condenses into liquid must contain nononvolatile constituents since all components of the vapor must be volatile to exist in thevapor zone. The liquid in the sump remains pure since any condensate that will be mixedwith water is immediately removed from the process chamber after it has been captured onthe liquid effluent capture tray. This process prevents any build-up of contamination fromthe influent alcohol or from the water carried into the dryer.

Dual-Vapor Drying Systems

The drive to continually improve drying techniques and the ability to design processesfor specific applications are never ending. Recently, a technique has been introduced thatuses two immiscible fluids that create a constant-boiling blend vapor phase. This techniqueis referred to as the dual-vapor drying system.

The chemicals chosen for this process have the commonality of being relatively high involatility with similar boiling points and relative vapor pressures. However, in order forthis process to be successful, the fluids must be essentially immiscible with each other. Byutilizing one chemistry that is soluble with water and one that is not, the amount of chem-ical that can be returned directly to the system without contamination is increased, thusreducing overall chemical consumption while still providing effective drying. An addi-tional advantage of this system is that if one chemical is chosen that is nonflammable, it canbe used with a flammable solvent to make the combined vapor phase nonflammable.

One such system on the market combines a perfluoronated chemical (PFC) with theconventional drying fluid IPA. The process utilized is very similar to conventional IPAvapor-phase drying discussed above with minor differences in chemical management(Figure 2). From a drying perspective, the parts that can be effectively dried and the dura-tion of the process are essentially unchanged relative to conventional IPA drying.

The significant difference between this process and conventional IPA drying is that lessIPA is consumed in the process, the vapor is nonflammable, and gravimetric separationrecovers the PFC for reuse in the process.

A volume of liquid that is approximately 50% PFC and 50% IPA (v/v) is placed in thebottom of the process chamber. Since the liquids are immiscible and have a significant dif-ference in density, the PFC will settle to the bottom and the IPA will form a layer on the topof the PFC. Heat is applied to the liquid. As the PFC liquid begins to boil, the PFC vapormust percolate through the IPA layer, heating the IPA, which although not boiling will


Figure 2 Dual vapor process.

SEALED LID

CHILLED FREEBOARD

CONTAINMENTCOIL

SEPARATOR

HEATER

SETTLING TANK

COMPONENT

CATCHMENT TRAY

PFC

IPA

IPA / PFC VAPOR

contribute vapor to the system. As previously described, the vapor is more dense than airand will displace all air out of the process zone to create a 100% vapor blanket above the liq-uid to the height of the condensing coil. The relative vapor pressures of each chemical con-tribute to a mixture of vapor that is approximately 50/50 by volume and is nonflammable.

During idling conditions, the vapor will condense on the coil with the liquid conden-sate being directed back down the chamber walls to return to the boiling solution. Uponthe introduction of a load, the vapor will preferentially condense on the parts to be dried.The liquid condensate will drip off the parts and be collected on a condensate troughdescribed in the conventional IPA process. The condensation of both vapors will contributeheat to the substrate and, as the product approaches the temperature of the vapor, theprocess will eventually cease. Leaving the product in the vapor zone will eventually pro-duce an equilibrium of vapor/condensate on the surface of the parts with its associatedmicrolayer of liquid coating the parts.

The liquid that condenses on the parts is captured by a saucer tray as described in theearlier discussion of IPA vapor phase. However, in this case the liquid will separate intotwo phases, a PFC and IPA/water mixture. This liquid effluent is sent to a holding tankwhere the liquid is allowed to settle and separate into two layers. The more dense PFC willsettle to the bottom where it can be extracted and recirculated back into the process cham-ber. The IPA/water mixture is decanted off for reclamation or disposal. Since the PFC is notchilled, it retains much of its thermal content and is returned to the process chamber at ele-vated temperatures ready to begin the vapor cycle again. This conserves heat energy by notextracting the thermal value as with the waste IPA/water effluent.

Since IPA is consumed in the process, the level of IPA in the tank will be reduced as afunction of process load volume. A special system of liquid level floats is included in the


design to make up the IPA volume continually from a sealed reservoir. By maintaining arelatively consistent ratio of IPA to PFC in the liquid phase, the composition of the vaporwill remain constant, providing reproducible results.

In a conventional IPA process, the parts are removed from the vapor zone into the free-board area above the condensing coil where final flash evaporation takes place. After a rel-atively short time, the parts are removed from the system. Any residual IPA that is left onthe parts or that is distributed into the atmosphere around the unit is of a sufficiently lowconcentration to avoid any operator exposure or flammability concerns. A significant dif-ference between the conventional and dual-vapor process is that the cost of the PFC chem-ical is such that even minor losses will have cost impact on the process.

The dual-vapor system is designed with a chilled freeboard to increase solvent reten-tion. Above the primary condensing coil is a refrigerated plate operating at a temperaturebelow the freezing point of water. This low temperature over a boiling environment createsa temperature inversion producing a dense layer of cold air above the vapor. The parts areraised from the vapor zone into the freeboard, where the vapor pressure of either con-stituent is essentially zero. As the condensate layer on the parts is flashed off the surface, itis immediately cooled and, as its density is significantly higher than air, it will fall back intothe vapor zone for recapture. The refrigerated freeboard area works to reduce the amountof the expensive PFC that will exit the system as well as to reduce significantly the amountof IPA evolved from the system. Although the cost consideration is not as critical for theIPA, it is a volatile organic compound (VOC) and is tightly regulated by many federal,state, and local environmental laws. The design of this system is such as to produce verylow solvent emissions. The unit can also be supplied with a lid that seals during the processto lower solvent emissions even further.

In addition to lowering overall vaporous emissions, the dual-phase system reduces theamount of chemical consumed in the process. In the conventional IPA dryer, the entirevapor zone is composed of IPA and it condenses on the parts, which solvates and displaceswater from the surface. The volume of liquid IPA generated during the vapor-phase dry-ing process is contaminated with water. In some cases, the water content will be a little as1 to 2% or as high as 12 to 15%. In either case, the IPA is unusable in this state and must beextracted from the system for disposal or reclamation. The difficulty of removing IPA fromthe water is exacerbated because IPA and water will form an azeotrope that makes simple,single-plate distillation inappropriate to provide solution suitable for use. Multiplate dis-tillation or membrane separation is usually required, often with significant cost impact.

The dual-vapor system reduces the amount of water-laden IPA waste by simply mak-ing less IPA available to be contaminated. Since the vapor phase is roughly 50/50 PFC toIPA, and water is immiscible in PFC, all of the water removed must be contained in the liq-uid IPA effluent. This automatically doubles the water loading in the IPA by halving thevolume condensed.

The major advantage of this process is the ability to use IPA efficiently as the dryingfluid, which has established itself as a product of choice. The long history of IPA in thisapplication provides confidence in the performance, whereas this technique eliminatessome of the negative aspects. By providing a vapor phase that is nonflammable the safetyaspects of using IPA are addressed. Also, since the system has emission control technologytargeted at the maximum level of solvent retention, it reduces the potential environmentalimpact from using IPA in communities sensitive to VOC emissions. In addition, the totalamount of chemical consumed in the process can be reduced through effective recircula-tion of the PFC solvent.

One drawback to this process is the potential of increased cost to add the solventretention technology. Another potential drawback may be in extended process times


necessitated by increased residence time in the chilled freeboard to accommodate solventretention criteria.

Liquid/Vapor Process

A second process that uses a combination of liquid immersion followed by a vapordeposition is called the “Marangoni” process. This process is generally used following ahigh-purity deionized (DI) water final rinse stage. The principle for displacement differssignificantly from the previously described technique (Figure 3).

The fundamental principle for displacement of the water relies on the differential sur-face tension between the DI water and the IPA. Following a high-purity rinse stage, thesubstrates are immersed in a bath of DI water at ambient or slightly above ambient tem-perature. The substrates are slowly removed from the water either by gently raising thesubstrates from the bath or, as more generally used in production, slowly draining thewater from the tank. At this point, the substrates are usually removed from the carrierthrough a lift mechanism that suspends them in the liquid bath providing minimal contactpoints. The slow withdrawal of the DI water creates a sheeting effect on the surface to stripaway any residue. A meniscus is formed that trails the liquid withdrawal. As the liquidlevel is being lowered, IPA vapor is introduced into the system. Some of the vapor coats theexposed surface as some of it is absorbed into the surface layer of the DI water.

At this point, an intriguing phenomenon occurs. A discovery by Lord Kelvin’s brother,James Thompson, identified a reaction in fluids of dissimilar surface tensions as related toa teardrop effect witnessed on vessels containing alcoholic beverages. He identified this in

Figure 3 Marangoni process.

COMPONENT

N2 N2

IPA VAPOR

DI IMMERSION


a paper published in 1855. Subsequent investigations by a fluid dynamics investigator,Marangoni, provided the theory and the name for the process.

As the IPA is absorbed into the microlayer of liquid on the surface of the substrate, itlowers the surface tension relative to the bulk liquid in the process tank. The physics of thetwo dissimilar surface tensions drive the system to equilibrium by forcing the low-surface-tension liquid toward the liquid of higher surface tension. This produces a flow that essen-tially strips the water layer from the substrate. The bulk liquid is continually purged withpure DI water to maintain its high surface tension.

Once the substrate has been completely removed from the liquid, a purge of dryingatmosphere, usually nitrogen gas, is admitted to the system and any residual IPA that maybe on the surface of the substrate is evaporated.

Small amounts of IPA are consumed in the process as it is absorbed into the bulk liquidor evaporated. The DI water must be completely drained from the system and refilledbetween cycles since it has become contaminated with the IPA.

Fixturing, substrate topography, and fluid motion are all critical to the success of thisprocess. In most commercial systems, the substrate is removed from the carrier during thewater-removal phase. This is a result of requiring a smooth flow of liquid across the sub-strate surface, which could be impaired or negated by contact points or irregular configu-rations in the carrier. The product surface must also be extremely smooth to guaranteesuccessful water removal. Finally, the motion of the water across the surface must be con-trolled to eliminate any turbulence. Agitation at the interface of the IPA-rich layer and bulkliquid will result in a disruption of the surface tension differential and negate the effects ofthe Marangoni principle.

The technical requirements stated above make the product configuration much morecomplex than other dryer designs. More moving parts and the possibility of disturbancesin the flow of fluid across the surface make the system less forgiving in its operation. Theprocess also consumes large volumes of DI water since the entire tank volume must beexchanged after each cycle. Even though there is only a small amount of IPA dissolved inthe liquid, it is quite difficult to recover in situ and must either be thrown out or recycled.

The major advantage of Marangoni over other principles is that it essentially useswater to dry itself. The small amount of alcohol used is fairly insignificant from a chemicalconsumption perspective. Also, since the process takes place at room temperature orslightly higher, and the water is removed from the surface without evaporation, the resultsshould be spot-free drying and minimal energy consumption.

LIQUID DISPLACEMENT

The previously discussed techniques relied on the solubility of water with the dryingchemical (IPA) for success.As was mentioned, the solubility of water in alcohol enhanced theremoval process, but rendered the IPA unsuitable for reuse in the system. There exist severalchemicals that can successfully remove water from the surface without being soluble. Theadvantage of these chemicals is that they will reject the water down to part per million levelsallowing the drying chemistry to be recycled in the system. Closed-loop recycling of thechemical will significantly reduce the waste stream volume and enhance system economics.

The chemicals that will dry while rejecting the water are also highly volatile and can berecirculated in the system through distillation. Although these various chemicals use thesame principles described below, they differ significantly in their behavior with the wateronce it is removed. Two types of chemicals are used in the liquid displacement technique:one is more dense than water and the other is less dense. Both techniques will be discussed.


Liquid More Dense Than Water

Fluids more dense than water have been used for many years in industrial applica-tions. These products were based on chlorofluorocarbon (CFC) chemicals. Two methodsdominated the process. One process involved using a CFC with a surfactant; the otherblended the CFC with alcohol. The chemical management within the system was slightlydifferent and so was the process.

As a result of the terms of the Montreal Protocol, the CFC chemistries were bannedworldwide for use in this application. Several chemical companies have developedreplacement chemistries for this process. These chemicals are hydrofluorocarbons (HFCs),hydrofluoroethers (HFEs), and some brominated hydrocarbons. All of these chemicals arebeing developed with surfactant additives to assist in the drying process. These productsare also being blended with alcohols to replace CFC products in nonsurfactant dryingapplications.

Displacement Fluid with Surfactant

Unlike the previously discussed techniques that used vapor-phase alcohol, the liquiddisplacement technique requires a minimum of two process tanks. One process tank con-tains the drying chemical and the surfactant, while the second chamber contains pure dry-ing chemical distillate. The chemical management is strikingly similar to the conventionalsolvent cleaning system, and this makes sense if water is considered a contaminant.

In the boil sump, usually situated on the left-hand side of the equipment (Figure 4), thedrying chemical and surfactant are blended. This mixture is boiled to create a vapor asdescribed in previous techniques. The surfactant is nonvolatile or of extremely low vaporpressure at the boiling point of the drying chemical to contribute little if any vapor. The vaporis condensed on a containment coil and the liquid returned to the right-hand immersion

Figure 4 More dense than water—surfactant/displacement fluid.

DRYING FLUID

BOIL SUMP IMMERSION SUMP

SURFACTANT/ DRYING FLUID


sump. This pure distillate stream fills the immersion sump and overflows a weir to returnto the boil sump.

The drying process takes place in a manner similar to conventional solvent cleaning.The product is lowered into the unit over the boil sump. As it enters the vapor zone, sol-vent vapors condense on the part and begin to displace the water on the surface. This liq-uid falls from the surface into the boil sump. The product continues to be introduced intothe equipment until it is fully immersed in the boil sump.

When the product is fully immersed, the combination of the drying chemical and sur-factant displaces the water from the surface, including blind holes and other areas wherewater may be trapped. The water that is displaced rises to the surface of the sump since itis insoluble and less dense than the drying fluid.

The water displaced from the product will accumulate at the surface of the boil sump.To prevent recontamination of the surface of the parts, the surface of the fluid is continu-ally purged by a sparging mechanism to skim the surface layer off the sump and force itinto a water separator. The water separator is a series of baffled chambers that will allowthe water to separate from the drying fluid and surfactant mixture for removal from thesystem. The pure drying fluid/surfactant mixture is returned to the boil sump through thesparging system to continue the process. The water is decanted from one of the baffledchambers and sent to drain or collected for proper disposal.

After the product has been removed from the boil sump and suspended in the vaporzone to allow bulk liquid to drip off the product, the parts are transferred to the pure dis-tillate immersion sump. The parts are coated with a microlayer of drying fluid/surfactant.In the immersion sump, the surfactant is removed from the surface by its solvency in thedrying fluid. The mechanism in this sump will be to remove all surfactant from the surface,disperse it throughout the bath, and redeposit it on the surface at the equilibrium concen-tration of the mixture in the bath. This fluid is continually returned to the boil sump byoverflow of the weir, which returns the surfactant to the boil sump for reuse. Since the sur-factant is essentially nonvolatile, the concentration in the system will remain fairly constantacross a long period of time and will produce consistent performance. After a specifiedperiod of time, the parts are removed from the immersion sump and suspended in thevapor zone above the immersion sump.

The temperature of the immersion sump is below the boiling point of the drying fluid.While the surfactant rinse is occurring, the parts cool down. When they are suspended inthe vapor zone after the surfactant removal, they will experience a final distillate rinse.Pure distillate from the vapor zone will condense on the parts until they reach the micro-layer of thermal equilibrium, previously described, and the residual layer will be flashedoff in the freeboard area for the final “dry” cycle. Water-free and solvent-free parts areremoved from the system.

This process offers several advantages over the previously described techniques. Theprimary advantage is its ability to dry complex geometries by going into full immersionwhile maintaining chemical purity through the ability of the drying fluids to reject thewater for rapid recycling. Another advantage is that it performs at lower temperatures thanvapor-phase alcohol dryers. It also has lower chemical consumption than the previouslydescribed techniques.

The major disadvantage of the system is that the use of a surfactant can raise the sus-picion of residue on the product after drying. Although this can usually be addressed bymultiple rinse sumps, it could be considered unacceptable in some applications. Other dis-advantages are the relatively large footprint of the equipment and the cost of the dryingfluid. The drying fluids are quite expensive relative to alcohol and require sophisticatedequipment design features to minimize emissions.


Displacement Fluid with Alcohol

The fluids discussed above are also being blended with alcohol to eliminate the needfor the surfactant. In this process alcohol is blended with the drying fluid in the boil sump.The formulation is generally 10% IPA and 90% drying fluid. Although the alcohol is solu-ble in the drying fluid, it does not form an azeotrope at that concentration but will form anazeotrope at a very low percentage as it is boiled and generated into a vapor.

The vapor generated from the boil sump will form an azeotrope with the alcohol and,in a manner described earlier, will be condensed on a cooling coil and sent to an immersionsump (Figure 5). The parts to be dried are immersed in the immersion sump. Here, thesmall amount of alcohol with the high-density drying fluid will combine to remove thewater from the surface. The alcohol is preferentially soluble in the water and produces aphase separation creating an insoluble layer of alcohol and water. This liquid layer is lessdense than the drying fluid and therefore rises to the top of the sump as previouslydescribed in the surfactant system. The surface is sparged to force the insoluble layer intothe water separator. The water/alcohol layer is separated from the drying fluid and thedrying fluid is returned to the boil sump via the sparging system. Since no surfactants areused and all constituent chemicals are volatile, an immersion rinse after immersion in thedrying sump is not required.

The drying cycle is essentially identical to the vapor-phase drying cycle described ear-lier. The chemical composition of the vapor is an azeotrope of the drying fluid/alcohol andwill condense on the parts and heat the substrate until condensation ceases. This processwill tend to displace any liquid dragout from the water removal step and leave the prod-uct free of water and liquid solvent.

The major advantage of this process is that it is able to use alcohol as a drying fluidwithout the concern for flammability while also being able to reject the water rapidly fromthe system. Although this process maintains the advantageous aspect of immersion drying,it produces a waste stream of alcohol and water not generated in the surfactant process. It

Figure 5 More dense than water—alcohol/displacement fluid.

SPARGER

ALCOHOL / DISPLACEMENTFLUID AZEOTROPE

WATER SEPARATOR

IMMERSION SUMP

ALCOHOL / DISPLACEMENTFLUID BLEND BOILING

FLUID


does eliminate the concern of surfactant residue but, since alcohol is consumed in theprocess, requires that the alcohol level in the boil sump be monitored and adjusted to pro-duce consistent drying performance. Another aspect of this technique is that the top layerof the immersion must be adequately purged to prevent the redeposition of water from thislayer. If this is a critical concern, an additional sump may be used for safety.

Liquid Less Dense Than Water

Liquid displacement with fluids less dense than water and insoluble with water hasbeen used for many years. The established techniques usually utilize one of an assortmentof hydrocarbons that displaces the water from the surface and then is air-dried. Many ofthese techniques are not able to produce residue-free surfaces because of nonvolatile frac-tions in the hydrocarbon. With water displacing oils, a significant fraction of the solution isnonvolatile with the intended purpose of leaving behind a residual material to protect thesurface from oxidation or other aspects of contamination.

These dryers are usually quite simple in design, generally consisting of a single immer-sion tank (Figure 6). The process is also simple. The parts to be dewatered are placed in theimmersion bath and usually suspended off the bottom with fixturing or a work rest. Someform of mechanical energy is applied, usually vertical agitation in the fluid, air agitation, orultrasonics. The water is displaced from the surface, and since it is more dense than the dis-placement fluid, it sinks to the bottom of the tank. The bottom of the tank below the workrest or below the suspended parts is usually necked down to produce a funnel-shaped sec-tion where the water is collected. The water level in the tank can be observed through theuse of a sight glass or other methods. When the water accumulates to a proper level, it canbe drained from the system off the bottom of the tank. When it is drained, a small layershould be left in the tank to prevent draining of any hydrocarbon into the water effluent.

The advantage of this technology is its simplicity both in process and design. Theequipment is very inexpensive as are the displacement fluids. It has a disadvantage in that

Figure 6 Less dense than water—oil displacement.

WASTE WATER

U/S

ON

IC

U/S

ON

IC

DISPLACING OIL


its ability to remove all water totally from the surface is suspect since it relies strictly onunenhanced chemical incompatibility. Without affecting the substrate affinity for water, itcannot prevent redeposition of water should the opportunity arise.

Displacement chemicals that are volatile, less dense than water, and insoluble in wateroffer an additional drying option. Much of the development work in this area has centeredaround volatile methyl siloxane (VMS) fluids. These chemicals are insoluble in water andare combined with a surfactant to displace water from the surface of the substrate. Thedevelopment of this technology is relatively recent and is still being tested to optimize itsperformance.

The process for these chemicals differs from the more-dense-than water technique.Although the process still uses a vapor-phase drying step, the equipment is considerablydifferent from the systems described above (Figure 7). The equipment is a two-sumpdesign, but the chambers are isolated in that they do not share a common vapor zone. Theprimary water displacement takes place in a chamber using room-temperature or slightlyabove ambient drying fluid. The parts are immersed in a bath of displacement fluid andsurfactant. The water is removed from the surface, and as it is more dense than the dryingfluid, sinks to the bottom of the liquid bath. After a specified period of time, the parts areremoved from the liquid bath and suspended above the fluid.

The parts are coated with the liquid from the immersion sump, which is a combinationof displacement chemistry and surfactant. The bulk liquid is allowed to drip off the parts.The parts are then sprayed with recycled displacement fluid, which will help to knock offany of the dragout layer coating the parts. After the spray has timed out, the parts areremoved from the dewatering chamber and transferred to the drying chamber.

During the dewatering cycle, the liquid immersion sump is continually recycledthrough a water separation system that strips the water from the solution and returns it tothe chamber. This material is also used as the source for the spray following the immersiondewatering.

Figure 7 Less dense than water—surfactant/displacement fluid.

WATER SEPARATOR

TANK "1" DISPLACEMENT FLUID / SURFACTANT TANK "2" VAPOR RINSE

DISPLACING FLUID

HEATER

U/S

ON

IC

U/S

ON

IC

SPRAY BARS C/WFAN NOZZLES CONTAINMENT

COIL


The drying chamber is a single-sump design very similar to the vapor-phase dryingsystem discussed previously, and it essentially performs the same process. In this chamberthe displacement chemical without surfactant is boiled to create a vapor. The vapor is con-tained with a cooling coil to define a vapor zone, and the cooling coils are located at a depthin the chamber to provide acceptable freeboard. The vapor that condenses on the coolingcoil is directed downward along the side of the chamber walls to be returned to the liquidlevel to complete the distillation cycle.

The displacement fluid-wet parts from the dewatering step are introduced into the dry-ing chamber. They are immersed into the vapor zone and the displacement chemical con-denses on the surface. This liquid condensate flushes off the displacement fluid and theliquid falls into the liquid level. The vapors continue to condense on the parts until the partsapproach the boiling point of the displacement chemical, completing the thermal equilib-rium phase of the process. After thermal equilibrium is reached, the parts are elevated intothe freeboard area where the residual microlayer of displacement chemistry is flashed off.The product can be removed from the system free of water and displacement chemical.

During the vapor rinse phase the vapor condenses on the parts and removes the dis-placement fluid, which falls into the sump. This liquid contains displacement fluid and sur-factant as well as the pure distillate from the vapor zone. The surfactant is nonvolatile soas the solution boils it does not contribute to the vapor zone, and as a result the vapor zoneremains pure regardless of how much product is processed.

The introduction of wet parts into the drying chamber, and the collection of the dis-placed fluid in the liquid level, result in a net increase in liquid volume in the drying cham-ber. Liquid level floats monitor the level in the chamber, and periodically liquid from thesump is returned to the dewatering chamber. This fluid contains the displacement chem-istry as well as surfactant that has accumulated in the drying chamber. This action returnsthe surfactant to the dewaterng chamber to prolong its useful life.

The most significant advantage of this process is that the water leaves the surface andfalls to the bottom of the tank. Unlike the more-dense-than-water technique where thewater rises, this process eliminates the necessity to remove dewatered product through alayer of liquid that may contain some water. The process is suited to complex geometriesbecause of its immersion step. These chemicals are also more moderately priced than themore-dense-than-water products and have essentially no adverse environmental impactassociated with global warming or ozone depletion.

The major drawback of this system is that the chemicals are flammable; however, theyare not VOCs which may make them acceptable in specific environmental situations. Theequipment also has a relatively large footprint.

SUMMARY

The attempt in this limited evaluation is not to make inference that all liquid dryingtechniques are applicable across the board, but merely to show that the availability of dry-ing techniques is almost as diverse as the applications they address. In any case, it is theresponsibility of users to provide the necessary insight into cleanliness standards, evalua-tion criteria, costs, environmental impact, and other specific considerations relative to par-ticular requirements. They will utilize this information in collaboration with the chemicaland equipment suppliers to develop their appropriate process. As it is with all things, how-ever, change will occur and any decisions made today must be considered tentative.Careful consideration should also be given to future needs and changes in technology soan optimized position can be established that satisfies today’s needs while being flexibleenough to accommodate tomorrow’s.


Documents

1655_02.pdf