M. P. MCLOUGHLIN ET AL.

A

Development of a Field-Portable Time-of-Flight MassSpectrometer System

Michael P. McLoughlin, William R. Allmon, Charles W. Anderson, Micah A. Carlson,Daniel J. DeCicco, and Nicholas H. Evancich

PL has developed and demonstrated a miniature time-of-flight (TOF) massspectrometer that can identify biological warfare agents. Its small size and low powerconsumption make the miniature TOF mass spectrometer an ideal candidate for fieldapplications, overcoming limitations of the current generation of biodetection systems.Portable applications require additional features such as ruggedized packaging, battery-powered electronics and data acquisition, automated sample collection, and automatedsignal processing and control software. In this article we describe the design of a TOFmass spectrometer system that incorporates these capabilities and integrates supportingtechnologies under development. Design goals include short analysis times (<5 min),high sensitivity, wide agent bandwidth, portability, low power consumption, minimaluse of fluids, unattended operation, and automated detection and classification.(Keywords: Biodetection systems, Time-of-flight mass spectrometer.)

BACKGROUNDThe Miniature Time-of-Flight (TOF) Mass Spec-

trometer Program is a 3-year Advanced TechnologyDemonstration jointly funded by the Defense AdvancedResearch Projects Agency (DARPA) and the DefenseThreat Reduction Agency Chem/Bio Directorate. Twomajor objectives of this program are the demonstrationof a TOF mass spectrometer system suitable for use onthe battlefield to protect forces and rapid transition ofthis technology to the warfighter. Although this initialapplication focuses on battlefield (stationary, point de-tection) applications, the TOF mass spectrometer tech-nology developed in this program provides a basis for

326 JOH

developing systems to support counterterrorism, do-mestic preparedness, law enforcement, and low-cost/low-power systems for remote standoff applications. Inthis article we describe the development of a prototypeTOF mass spectrometer system that is suitable for fieldapplications and show its current status. Other articlesin this issue describe particular aspects of the TOF massspectrometer technology used here.

Currently deployed biodetection systems are severe-ly limited owing to low sensitivity, limited agent band-width, and slow response. For example, immunoassaysystems that sense the binding of an antibody with a

NS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1999)

A FIELD-PORTABLE TOF MASS SPECTROMETER SYSTEM

target agent typically require 20–30 min to identify abiological agent and are limited to detecting only thoseagents for which an antibody is available. The biolog-ical detector used in the Army’s Biological IntegratedDetection System has several limitations. First, the sys-tem is limited to the detection of only eight of the manyknown potential threat agents. Second, it requires thestorage and replenishment of perishable reagents.

The program described in this article will demon-strate the ability to use TOF mass spectrometer tech-nology to overcome these and other limitations. Designgoals include short analysis times (<5 min), highsensitivity, wide agent bandwidth, portability, low pow-er consumption, minimal use of fluids, extendedunattended operation, and automated detection andclassification.

During the first year of this program (FY97), effortswere focused on developing system requirements andconcepts and selecting and evaluating technologies forfurther refinement. Basic technology selections includ-ed air sampling and particle capture methods, laser andoptics technology, mass analyzer design, vacuum inter-face concepts, and signal processing. Laboratory effortsfocused on identifying specific simulant mass spectralbiomarkers used for evaluation. Sample preparation formass spectral analysis traditionally involves a skilledtechnician and requires extensive use of fluids, makingthe techniques difficult to automate. As various designalternatives are considered, a high level of coordinationbetween system design and laboratory activities is es-sential to ensure that the methodsdeveloped in the laboratory canbe translated to the field-portablesystem.

The FY98 follow-on effort fo-cused on developing and demon-strating each of the TOF mass spec-trometer subsystems and testingmajor system interfaces such assample vacuum interface, high-speed data acquisition, automatedsample collection, sample process-ing, and signal processing algo-rithms. Key accomplishments of theFY98 systems engineering effortincluded the following:

• Demonstration of the “end-to-end” process of sample collectionthrough automated identification

• Development of system re-quirements for the TOF massspectrometer

• Development of a transition plan• Completion of a conceptual de-

sign review Figure 1. Fie

Stepper motoand drive sha

VitaSample

spot

Drive shaftside view

JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1

One of the most important aspects of the FY98 effortwas the verification that laboratory sample preparationtechniques could be translated into an automated field-able technique using minimal amounts of fluids. Whilewe are continuing to improve the efficiency of acquir-ing spectra from our dry collections, demonstrationof the feasibility has enabled a fundamental shift inbiodetection technology away from traditional tech-niques based on “wet chemistry.”

SYSTEM OVERVIEWFigure 1 is a conceptual diagram of the TOF mass

spectrometer system under development. Air is drawninto a collector that separates particles in the range of0.5–10.0 mm from the airflow and deposits them ontoa movable tape. After collection is completed, the tapetransports the sample into the TOF mass spectrometeranalyzer. A microsprayer may then add small amounts ofMALDI (matrix-assisted laser desorption/ionization)matrix to the sample to increase sensitivity. (The use ofMALDI for biodetection is described in the article byScholl et al. elsewhere in this issue.) A laser then ionizesthe sample, and the resulting mass spectrum is analyzedfor specific biomarkers that indicate the presence andidentity of a biological agent. One advantage of thisapproach is the minute volume of fluids required forsample processing, eliminating the need for large storagereservoirs, stationary and level mounting configurations,or large power-hungry heating and cooling systems.

ldable time-of-flight mass spectrometer biodetection system.

Air inletMALDI microsprayer Second-

stageimpactorConcentrator

Sample spots

Laser

Turbo pump

Massanalyzer

Taperollersr

ft

Laptop user interface

Video tapecassette

Laserfiber

Columnator/attenuatordeope

999) 327

M. P. MCLOUGHLIN ET AL.

Multiple samples are collected concurrently, allowingboth the application of different analysis protocols andthe archiving of samples for later confirmatory analysis.

Figure 2 illustrates the system configuration underdevelopment. The sample subsystem consists of a VHStape cassette controlled by several stepper motors. Thetape is positioned both inline with an array of fourimpaction nozzles and at the inlet to the mass spectrom-eter. Power, electronics, blowers for air collection, anultraviolet laser, and data acquisition and processingelectronics are located underneath. In the remainder ofthis article we discuss efforts to develop the tape drive,laser and optics, vacuum interface, ruggidized TOF massspectrometer, electronics, and software components ofthe system.

To rapidly detect biological agents at these lowconcentrations, the air collection system must havehigh throughput and high collection efficiency to ob-tain a sufficient number of particles for detection.Because for some pathogens (such as Yersinia pestis, thecausative agent of plague) inhalation of a single organ-ism can result in infection, the system must be able todetect concentrations on the order of 10 cells or sporesper liter of air. The air sampling subsystem includes animpactor to separate particles from the airflow and atape drive to transport and position the tape (see thearticle by Anderson and Carlson, this issue).

TAPE DRIVE SYSTEMA proof-of-concept design of the tape drive mech-

anism used commercially available VHS tape players.

Figure 2. Current system configuration.

Tape cassette

Rough pump valvemotor assembly

Roughing pump Massanalyzer

Turbo pumpcontroller

Turbo pumpSteppermotors

Vacuum interface(platen not shown)

Virtualimpactor

Sprayer

328 JOH

Current designs incorporate the basic concepts con-tained in commercially available units while simplifyingand ruggedizing many of the components for militaryapplications. This approach eliminates the variabilitybetween commercial models or components and simpli-fies field service. For example, the tape interface plateuses a simple snap-down approach to loading the tape,eliminating many of the in-and-down mechanisms andcomponents featured in commercially available units.

The tape drive uses commercially available steppermotor drivers to position the tape. Current designsinclude a three-axis stepper motor driver to controlthree different motors independently. One motor po-sitions the tape in the aerosol collector, the secondpositions a sample in the mass spectrometer vacuuminterface, and the third is used to load the tape. Move-ment is accomplished by pinning the tape between arubber-coated roller and a grooved drive shaft. Theshaft actually drives the rubber rollers using the frictioncreated between the two materials. The grooved shaftmakes contact with the rubber roller both above andbelow the tape without contacting the sample location(see Fig. 1). The groove prevents any possible contam-ination between samples. As the drive motors turn, amotor attached to the tape interface plate providestension to the reel while moving forward or back alongthe length of the tape. A single motor mounted directlybelow the tape cassette uses two opposing unidirection-al couplings to operate both the left and right reels fortape take-up. This eliminates the need for dual motorsor complex mechanisms such as those found in com-mercial units. Elimination of motors and actuators sig-

nificantly lowers the power re-quirements of the system, reducingthe size as well as increasing itsportability.

A tape loop between two of thestepping motors permits the portionof the tape on the analyzer side ofthe system to be moved indepen-dently of the portion on the sampleside. This provides the ability toanalyze several spots on the tapewithout interrupting the collectionprocess. The loop isolates the activ-ities between the aerosol collectionand spectrometry operations, pre-venting any potential fouling or slip-ping of the tape. The design hasbeen maximized for flexibility inoperation and in consideration ofthe transition to manufacturing.Once the system design is tested andoptimized, many currently available“options” may be removed to furtherreduce size, power, weight, and cost.

NS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1999)

VACUUM INTERFACEIntroduction of a sample to the mass analyzer

requires repeated breaking and restoration of a high-vacuum seal as each tape sample is repositioned overthe sample port. It would take many tens of minutes torestore the mass analyzer chamber to a high vacuum ifthe whole chamber were exposed to the atmosphere.The vacuum interface for this system reduces the vac-uum pump loading by isolating the main vacuum cham-ber from the sample port around the tape sample whensamples are being changed, while it provides a clearpassage for the ions during the measurement. The in-terface consists of a seal between the sample tape heldby a platen and a port to the external ionization grid,shown in detail in Fig. 3. Before the tape is positioned,a closed high-vacuum ball valve isolates the largechamber containing the mass spectrometer analyzerfrom the small volume containing the external ioniza-tion grid. During measurements of mass, the ionizationgrid acts as a ground plane electrode in parallel to ahigh-voltage electrode located in the platen behind thesample tape. The grid consists of a screen that allowsthe sample to be illuminated by an ionizing laser beam.When the laser ionizes the sample, an electric fieldbetween the electrodes extracts and accelerates the ionstoward the mass analyzer. The external grid assemblycontains an auxiliary port that connects to a vacuumfore-pump, which evacuates most of the air trapped inthe ionization grid before the main high-vacuum valveopens. Bench tests have shown that we can restorevacuum within 30 to 40 s after closing the platen, andhelium leak tests of the tape seal have consistentlymeasured less than 10–7 cm3/s, well within the capabil-ity of the turbo pump to maintain microtorr vacuums.

Figure 3. Ionization grid vacuum interface.

+ +

++

+ +

++

Platen, –4.5 kV

Insulator

Samplespot

Vacuumroughingport

Ball valveport to massspectrometer

analyzer

Sample spot

Video tape

Ions to massspectrometer

vacuum chamber

Photodetector

Mirror

Laserbeam

Fiber outputfocuser

Ground plane

JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1

A FIELD-PORTABLE TOF MASS SPECTROMETER SYSTEM

LASER AND OPTICSThe portable mass spectrometer system uses a 300-

mJ pulsed ultraviolet laser in the ionization process. Theoutput of this laser is focused onto a circular spot be-tween 0.5 and 1.0 mm in diameter. The energy densitydelivered is variable between 20 and 80 mJ/cm2. As aresult of mechanical considerations associated with aportable and rugged device, the laser power is deliveredvia a fiber-optic transmission channel.

The fiber-optic transmission channel consists of fourmain components: the output coupler, the power atten-uator, the optical fiber, and the focuser. The outputcoupler is a series of lenses that focuses the 5 × 7 mmbeam produced by the laser into the optical fiber core(see Fig. 4). Power coupling efficiency varies from 20%to 90%, depending on the lens configuration and sizeof the fiber core. A large-diameter multimode or spe-cialized fiber core has a much greater ability to accepthigher power than a small-diameter single-mode opti-cal fiber core. The mass spectrometer uses a 100-mmoptical fiber core, with a projected input power cou-pling efficiency of 80%. This selection represents agood compromise between coupling efficiency and thefiber flexibility needed for packaging.

In addition to the output coupler, the mass spec-trometer incorporates a commercially available vari-able attenuator controlled by a stepper motor for vary-ing the output power (Fig. 4). The attenuation rangeis continuously variable from 0 to 30 dB. The attenu-ator contains a variable position screw, which is adjust-ed by a stepper motor to partially block the output beamprior to the coupling optics. In the TOF mass spectrom-eter optics, the attenuator and laser output couplerhave been combined into a single device.

The focuser consists of a variable position biconvexlens. The sample to be ionized is constrained to a fixeddistance from the focuser. The magnification of the

9

Figure 4. Laser-to-fiber delivery system.

99) 329

M. P. MCLOUGHLIN ET AL.

focuser is nominally 6.5 at 76.0 mm. The output spothas a nominal diameter of 0.65 mm as a result of thesize of the fiber core and the distance of the core fromthe focusing lens. The output spot diameter can beeasily focused from 0.5 to 1.0 mm. The focuser will becustomized to allow computerized focusing.

The optical fiber used for the mass spectrometer isa fused silica multimode fiber with a 100-mm core and140-mm cladding. This optical fiber is commerciallyavailable with industry standard FC/PC-type opticalconnectors on both ends. Both the attenuator and fo-cuser are also provided with FC/PC connectors. Themultimode fiber was chosen to maximize the powertransmission. The output beam pattern of a multimodefiber from a highly coherent light source is not Gauss-ian, as is the case for a single-mode fiber. The beampattern is a time- and position-varying “speckle” pat-tern that depends on the number of propagating modes.The large number of propagating modes will most likelyminimize any problems associated with this effect. Theexact beam patterns produced will be analyzed furtherusing a beam profiling system.

ANALYZER DESIGNThe heart of the system shown in Fig. 2 is the TOF

mass spectrometer analyzer described in this issue byCornish and Bryden. The TOF mass spectrometer hasbeen redesigned to fit in the ruggedized vacuum housingshown in Fig. 5. The housing usespiston-type O-ring seals for the endcap and ISO-NW flanges for theaccess ports, thereby providinghighly reliable sealing of the vacu-um chamber. The external housingis made of 304 stainless steel; theinternal support structure shown inFig. 6 is made of polycarbonate.The polycarbonate serves as aninsulator between charged platesfound in the reflector and was se-lected for its excellent impactstrength, ease of machining, lowcost, and relatively low outgassingproperties. All materials are com-patible with common disinfectantssuch as formaldehyde. The sampleexchange pump-down time is min-imized by adopting a cylindrical de-sign with as little internal volumeas possible.

The reflectron is made up of31 circular plates with a 3.3-cm-diameter hole through the centerto allow ions to travel through thereflector. Slots in the rails hold

Vacuum valve

Rough

Ionizationsourceregion

F

330 JO

these plates in place with the rails positioned 90° apart.The plates step down in equal voltage steps starting at6000 V on the plate farthest from the reflectron detec-tor to ground for the plate nearest the reflectron de-tector. Between each plate is a resistor to step down thevoltage. The resistors are held in place on the top andbottom rails via resistor stocks. The rails also includecutouts that act like springs. These “springs” makecontact with the inside of the vacuum chamber toensure a snug fit over a wide temperature range. Theend of the cylinder has a flange with a bolt hole patternfor connecting the end cap, which contains all theelectrical feedthroughs. Five of these feedthroughs arewelded in place, with four carrying high voltages (6, 8,and 10 kV). Four other feedthroughs are double-endedSMA type, which are isolated electrically from thechamber via polycarbonate inserts. The SMAs andinserts are glued in place using a low-outgassing epoxysealant. Two SMAs are used to carry signals fromeach of the detector plates, and two SMAs control theion gate.

Careful consideration was given to many aspects ofthe analyzer design to ensure that virtual leaks causedby gas trapped between tightly fitting components areminimized. Fusion welds are used to eliminate theintroduction of outgassing materials, such as sulfur, thatcould be introduced from weld rods. The feedthroughsare fusion welded to the end cap for sealing and sturdyconstruction, and are placed to avoid arcing between

Access ports

ing port

End capflange

Turbo pumpport

igure 5. Grid ionization valve and vacuum chamber.

HNS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1999)

A FIELD-PO

Resistor

Resistor stock Rail End cap

Contactspring

Contactspring

Reflectrondetector

Reflectronplates

(not all shown)Shielding

tubesO-ringgroove

Figure 6. Internal mass analyzer configuration.

components internal and external to the vacuum.Stainless steel shielding tubes are used to prevent high-voltage feeds from affecting the ion flight paths and toreduce noise in the signal lines. Tight tolerances aremaintained to prevent virtual leaks.

Because of the complexity of the design, a fuseddeposition model (FDM) was made of various iterationsof the design to help find design flaws and to bettercommunicate the design concept to others. The FDMmachine is a cost-effective tool for building models ofless than 25 cm in two of three axes. With minimalprogramming, the machine runs automatically, thus re-ducing labor. The FDM uses ABS plastic thread, avail-able in various colors, that is laid side by side andmelted together to form an approximately 0.025-cm-thick layer. The machine builds one layer at a time,melting each layer together as it builds the part verti-cally. Parts constructed with this process are currentlybeing used for benchtop testing of the mass analyzer.

SOFTWARE AND SIGNALPROCESSING

Software control of the analyzer consists of fourfunctions: data acquisition card control, motor control,data processing, and the user interface. For initialdevelopmental efforts, MATLAB and LABView wereused for the graphical user interface and control, respec-tively, facilitating the rapid testing of critical systemcomponents. A switch to the C programming languagewas made for several reasons, including speed, cost, androbustness. C is inherently faster than MATLAB andLABView and does not have an added abstraction layer.For example, LABView under MS Windows currentlyhas a timing resolution of 1 ms, which is unacceptable

JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1999)

RTABLE TOF MASS SPECTROMETER SYSTEM

for this application. Using C code,the sampling rate is limited only bythe oscillator of the computer.

The data acquisition controlsoftware operates a Gage AppliedSciences, Inc., CompuScope 8500A/D card. This card can sample ata maximum rate of 500 MHz withan 8-bit resolution. Data acquisi-tion is initiated by a trigger from aphotodetector diode, providing areference for the averaging of mul-tiple spectra. The trigger pulse isalso used as a reference for controlof the ion gate. For low-speed dataacquisition required for statusmonitoring (vacuum gauges, tem-perature, tape position, etc.), acommercially available Winsys-tems PC/104 data acquisition cardis used.

The motor control software positions the tape forcollection and analysis, and the software controls aCompumotor controller card and interfaces with thevarious stepper motors in the system. Compumotorprovided LABView and C drivers for its card. The mainrequirement for this piece of software is that the tapemust be repeatedly positioned within 0.1 mm to ensurethat the sample is illuminated with the laser. Laboratorytests have shown that the stepper motors and control-lers have sufficient accuracy to position the tape withinthe required margin of error.

The user interface controls almost every aspect ofthe instrument, from collection parameter to errorcondition response. This software consists of two dialogscreens: the Data Acquisition System (DAS) and theAerosol Collection System (ACS). The DAS controlshow each sample spot on the tape drive is analyzed. Theuser selects the sample spot to analyze, the MALDImatrix to use, the ion gate settings, and the laser powerlevels. The DAS then discretely operates on the indi-vidual samples with the user-supplied options. Analysisprotocols can be scripted and saved, eliminating theneed for repetitive parameter entry. Similarly, the ACSdialog controls the aerosol collection phase of thesystem. The user inputs the total collection time, andthe system automatically initiates collection and dis-plays the current status in the status box.

Signal processing software provides real-time feed-back to the operator about the identity of biologicalagents in the sample. Each sample is repeatedly pulsedwith the laser, and the resulting spectra are averaged.Averaged spectra are processed using a split-windowconstant false alarm rate processor, which sets a vari-able threshold that eliminates broad elevations in thebackground caused by nonspecific ions. The resulting

331

M. P. MCLOUGHLIN ET AL.

threshold crossings are then isolated, and a clustercentroid is found. For each centroid, a list of features,including mass, peak width, and amplitude, is thenpassed to a discriminator. The discriminator then ap-plies a series of rules to determine if any biomarkerlisted in the threat library is present, and the resultsare displayed to the operator.

SYSTEM TESTINGAcceptance testing for all candidate biodetection

systems is performed at the U.S. Army Dugway ProvingGrounds (DPG) test facility using simulants nonpatho-genic to humans. These simulants include Bacillus glo-bigii, a spore-forming bacterium similar to anthrax;ovalbumin, a chicken egg protein developed as a sim-ulant for toxins such as botulinium; MS-2, a simulantfor pathogenic viruses; and Erwinia herbicola, a simulantfor a vegetative bacterium such as plague. Aerosol sim-ulant concentration levels used in the evaluation arepredetermined by DPG and are set on the basis ofexpectations of threat concentration levels.

Tests at DPG comprise both field and chamber tests.Field tests are designed to determine the utility of abiodetection system under simulated battlefield condi-tions. Simulants are released by both ground-based andairborne sprayers and simulated artillery bursts, andsensor performance is compared to ground truth mon-itors. Although these types of tests provide some abilityto evaluate a system’s performance in an operationalcontext, sensor sensitivity cannot be accurately deter-mined because of the difficulty of controlling and mon-itoring aerosol releases.

Chamber testing provides performance data on thesensitivity of the detection/identification systems. Sim-ulants are released into a chamber at the threat con-centration under controlled environmental conditions.

332 JO

Table 1. Simulant concentrations for Dugway ProvingGrounds chamber testing.

Goal for total counts perSimulant liter of air (0.47–20 mm)

Bacillus globigii 150Erwinia herbicola 49,500MS-2 1,500Ovalbumin 1,650

Biodetection systems can be placed in the chamber, anddetection performance can be measured. The goal par-ticle counts per liter of air are shown in Table 1. Testingoperations allow simulants to be collected over a 5-minperiod with the initial simulant concentrations at thelevels indicated in Table 1. In preparation for the DPGchamber tests, we are performing tests using the aerosolresearch facilities at the Johns Hopkins School ofHygiene and Public Health.

CONCLUSIONS AND FUTUREEFFORTS

We have successfully demonstrated the functionalityof all critical subsystems for the TOF mass spectrometersystem and developed a design to fully integrate allcomponents into a field-portable prototype. The field-able system described is currently being fabricated andwill be completed before the end of FY99. In earlyFY00, APL will demonstrate the capability of the sys-tem with chamber testing, possibly at DPG. At the endof FY00 the system will be tested at DPG in evaluatedchamber and field trials, and the results will be used todetermine the suitability of our system for transition tocandidate acquisition programs.

THE AUTHORS

MICHAEL P. MCLOUGHLIN received his B.E.E. and M.S.E.E. degrees fromthe University of Delaware in 1983 and 1985, respectively. He joined APL in1985 as a member of the Submarine Technology Department’s Acoustics Groupand was involved in the development of signal processing algorithms foractive acoustic sonar systems. From 1990 to 1995 he worked at NoiseCancellation Technologies, Inc., and was responsible for the development andcommercialization of active noise control systems. He rejoined APL in 1995 andis currently a Program Manager in the Special Applications Program Officeand involved in system engineering, program management, and programdevelopment for counterproliferation sensor systems. His e-mail address ismike.mcloughlin@jhuapl.edu.

HNS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1999)

J

A FIELD-PORTABLE TOF MASS SPECTROMETER SYSTEM

WILLIAM R. ALLMON is a mechanical engineer in the Engineering and TestGroup of APL’s Submarine Technology Department. He received a B.S. degreein mechanical engineering from Virginia Tech in 1991. He joined APL in 1997after working at the United States Army Research Laboratory (formerly HarryDiamond Laboratory). Mr. Allmon specializes in electromechanical design,development, fabrication, testing, and installation with specific interests instructural analysis and materials. In addition to the Field-Portable Time-of-Flight Mass Spectrometer System, he has been involved in system design of theJoint Biological Remote Early Warning System (JBREWS) and mechanicaldesign for the Flare Genesis Experiment (FGE). His e-mail address iswilliam.allmon@jhuapl.edu.

CHARLES W. ANDERSON received a B.S. in mechanical engineering from theUniversity of Wisconsin-Madison and joined APL in 1978. He currently is amember of the Principal Professional Staff in the Submarine TechnologyDepartment’s Engineering and Test Group. His duties range from detailedengineering design to project management in the development of new systemssuch as the one described in this article. Previously, Mr. Anderson designed anddeployed towed undersea vehicles and associated launch and recovery, seabed-laid instrumentation, and aircraft towed systems. He has served as projectmanager and project engineer in the development of instrumentation systems forthe Submarine Security Program and others. He is active in professionalorganizations and chaired an international workshop on undersea cable andconnector technology for the Marine Technology Society. His e-mail address ischarles.anderson@jhuapl.edu.

MICAH A. CARLSON is a mechanical engineer in the APL SubmarineTechnology Department’s Engineering and Test Group. He received a B.S. inmechanical engineering and a B.A. in music from the University of Colorado in1996. Mr. Carlson’s specialties include structural design and analysis, materialsengineering, system integration, and the development and application ofminiaturized fluidics handling systems. His current work areas include thedevelopment of coastal monitoring buoy systems and the aerosol collectionsystem for use in the field-portable miniature mass spectrometer. Mr. Carlsonis currently pursuing his M.S. in mechanical engineering from The JohnsHopkins University and is also a member of ASME. His e-mail address ismicah.carlson@jhuapl.edu.

DANIEL J. DECICCO received a B.S. in electrical engineering from the GeorgiaInstitute of Technology in 1987 and an M.S. in electrical engineering from TheJohns Hopkins University in 1998. He was employed by the Ocean SystemsDepartment of EG&G Services from 1989 to 1998, where he designed numerousdata collection and telemetry systems for towed array sensor systems and towedbodies for the David Taylor Research Center in Carderock, MD. He joined theEngineering and Test Group of APL’s Submarine Technology Department in1998. Since then, Mr. DeCicco has worked on the electronic systems for theminiature TOF Mass Spectrometer and on the design of remote acoustic sensorsystems for various applications. His e-mail address is daniel.decicco@jhuapl.edu.

OHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1999) 333

M. P. MCLOUGHLIN ET AL.

334 JOHNS HOPKINS APL TECHNICAL DIGEST, VOLUME 20, NUMBER 3 (1999)

NICHOLAS H. EVANCICH received B.S. and M.S. degrees in electricalengineering from Purdue University. From 1996 to 1998, he was a member ofPurdue’s Applied Digital Electronics Research Facility, where he developedconvergence-based products. He joined APL in 1998 in the Engineering and TestGroup of the Submarine Technology Department. Since then, Mr. Evancich hasworked on acoustic processing software, control software for sensor systems forfacility protection, and acoustic sensor software. His e-mail address isnicholas.evancich@jhuapl.edu.