DETERMINATIONS OF HERBICIDE

SPREAD FACTORS

K. YoshidaMember CSAE

INTRODUCTION

One of the inherent problems inpesticide spraying is that of controlling droplet size. A few attempts havebeen made to determine the dropletsizes, or number of droplets per unitarea receiving surface, which wouldbe biologically most effective for thevarious types of treatments.

The importance of droplet size onherbicide efficiency has been investigated. With low-volume applications,Ennis and Williamson (9) observedhigher effectiveness with decreasingdroplet size. On the other side, Beh-rens (2) reported that droplet spacing was of major importance in herb-icidal effectiveness on mesquite, withno direct influence due to the dropletsize or the volume. The minimumnumber of droplets needed to obtainmaximum effectiveness was 72 persquare inch, regardless of size withinthe range of 200 to 800 microns diameter.

Excessively large droplets result ina serious waste of material, and verysmall droplets present a serious driftproblem. While investigating the factors affecting the use of airblastsprays, Brann Jr. (4) indicated thatin New York state the cost of spraymaterials represented about 75% ofthe total cost of insect control spraying. For these reasons, spray dropletsize and the range of sizes or dropletspectrum in a given spray have become important factors in evaluatinga spray distribution and in predictinga spray drift.

A number of attempts were madein different ways to determine spraypatterns, and various kinds of collecting surfaces were investigated in thisconnection. The impression or stain ofdroplet on a receiving surface cannotbe used directly to determine dropletsize distribution without a knowledgeof the relationship between the diameters of the spherical droplets andof the stains produced. This relation

66

Atmospheric Physics LaboratorySaskatchewan Research Council

Saskatoon, Saskatchewan

ship, or the ratio of spot size todroplet size, the spread factor orspreading efficiency, can only be determined by producing droplets ofknown or measurable spherical diameters on an insoluble matrix of collecting medium.

With the spread factor known, itsimplifies determination of the sizesof spherical" droplets in the atmosphere immediately prior to hitting thetarget.

REVIEW OF VARIOUS METHODS

In any spread factor study, themagnesium oxide method developedby May (17) is usually a startingpoint. The ratio of diameters from theoil-gelatine matrix method to theMgO method was obtained at average 0.858 for water in the range ofover 20 to 200 microns diameter.

From calibration of waterdrops atterminal velocity, Magono (15) obtained a relationship between the raindrop diameter and the stain.

d = 10-'/3.D% 1

where d = equivalent diameter ofrain drops (mm),

D = mean diameter of wetarea produced by raindrops on the photographic paper (mm).

Supposing that a droplet, regardless of the size, spreads into a pelletof uniform thickness of (h),

If)'h - Hi)' -volume of droplet 2

,.d^3D°hHowever, with mist, assuming asingle droplet deposits on a photographic paper in a lens shape, h ~D/3, and equation 3 becomes

)3 — 0.79D (or D —2

1.27d) 4In the study on the break-up

d=^/lD-

J. Maybank

mechanism of free-falling drops produced from a tube and deposited ona filter paper, Fournier D'Albe andHidayetulla (10) empirically deriveda spread factor equation over arange of droplet sizes of one to 10mm diameter. Their equation was

d = 0.47 D% 5

The index of % is to be expectedfrom geometry, while the figure 0.47is a constant depending on the typeof filter paper used. Jarman (12)modified this equation for his experiments with red-dyed tap water,power kerosene, spindle oil, and othersolutions. Droplets were produced bya microburette in a range of 139 to2,000 microns diameter. The growthof stains was also measured as afunction of the time elapsed sincedeposition. The constants determinedby Fournier D'Albe and Hidayetulla(10) were re-examined and the valueof the exponential power in this testwas 0.84 ± 0.02 as compared to0.667. It appeared that the value ofthe exponential power depends onthe period of spreading, the properties of the liquid and paper used, andthe ratio of the drop diameter to thethickness of the absorbent paper.

Roth and Reins (19) reported a result of spread factor determinationof red-dyed water and diesel oil overa parchment paper and on a gelatinous matrix. Droplets of uniformsize were produced by a spinningdisc device and the spread factorover a range of 100 to 1600 micronsvaried from one to four with dropletdiameter.

Jarman (13) investigated the deposition mechanism of wind-born oildroplets on various horizontal surfaces. By measurement, plain andsilicone grease coated glass plateswere found to produce a constantspread factor for a range of dropletsizes below 250 microns diameter.MgO coated plates appeared preferable, from the result of this test, for

CANADIAN AGRICULTURAL ENGINEERING, VOL. II, No. 2, NOVEMBER 1969

droplets of 10 to 200 microns diameter.

Courshee and Valentine (7), aspart of a study on spray samplingtechnique, discussed the advantagesof plaster coated glass as a collectingsurface. The spread factor was calculated with a liquid containing 1% ofblack dye within a size range of 700to 800 microns diameter and thevalue obtained indicated a high reliability and stability. In general, theincrease of surface roughness causedan increase in the spread factor. Theuse of a surface on which the spreadfactor is largely independent of impact velocity would be recommendedin this respect.

Droplet distortion and shatter canoccur on impact on some collectingpapers. This produces considerabledifficulty in automatic scanning processes. Middleton and Lowe (18) developed a kaolin-coated samplingsurface in order to reduce shatteringeffects. The spread factor with sucha surface was as high as four over arange of 500 to 5,000 microns diameter.

Gebhardt and Bode (11) developed an improved method to determine the spread factor of water droplets by means of a micro-electricpump which was previously reportedby Atkinson and Miller (1). A photographic negative of the spots wasformed when spray droplets rangingfrom 50 to 1,000 microns diameterwere deposited on a special papersurface. This negative was processedby the USDA flying spot particleanalyzer as described by Brazee andIrons (5).

Carleton (6) reported a new technique for recording water spray spotimages continuously across a sprayswath on a special 35 mm film strip.The dyed stains on this film stripwere printed by a 35 mm motionpicture contact printer. This negativefilm was processed with the flyingspot analyzer and spread factor determined. An empirical equation wasderived for this spray material onthis specific collecting medium,

d = A.D 6

where the constants A and C werefound to have values of 1.80 and 0.81respectively in their particular studies.

Size distribution data processedthrough the analyzer were automati

cally fed into a computer system byBouse (3). The result, using bi-fluidnozzles on white Kromekote-castpaper also showed that drop spreadwas dependent upon the drop size.From a series of helicopter applications with bi-fluid nozzle systems,Lehman, Haas, and Robinson (14)discussed the spread factor determination on a poster paper in comparisonwith a surface filled with Cedarwoodoil. A regression equation of spreadfactor against droplet size was calculated, giving the appropriate spreadfactors for 2,4-D water-in-oil emulsions over a range of 200 to 1,800microns diameter. The spread factor,f, varied linearly with the dropletdiameter and could be expressed as

f — 1.72 + 0.0007 d 7

The spread factor at a diameter of400 microns was approximately 2.0.

Summarizing these various experiments, it appears that the ratio ofstain diameter to drop diameter cannot in general be represented by aconstant spread factor, f. This factorwill increase with time, as the stainspreads, and the value of f tends tobe larger for the larger drops. Furthermore, it is dependent on theproperties of the collecting surfaceand of the fluid. It would also appearthat the lower the value of f, the lessthe variation with droplet diameter.An advantage of choosing a smoothcollecting surface on which spreadingis likely to be slow is that stain edgeswill remain more clearly defined,thereby improving measurement precision. In the present work, investigations of such surfaces were made toexamine the behavior of herbicidedroplets on them.

MICRODROPLET APPLICATORS

For spread factor determinations,microdroplet applicators of varioustypes were developed to producespray droplets of predetermined uniform sizes at required rates of emission. The devices used were basedon capillary tube, capillary tube withvibration, capillary tube with air blastnozzle, spinning disk, a bubble breaking device, micro hydraulic pump,and microsyringe with dispenser.

Of these instruments, the spinningdisk type is most widely used in fuelatomization study and in aerosolgeneration because of the narrowersize ranges produced by it. Thecapillary tube method is well de

CANADIAN AGRICULTURAL ENGINEERING, VOL. 11, No. 2, NOVEMBER 1969

veloped among entomologists and inrain drop studies. Methods usingmicro hydraulic pumps and micro-syringes are recent developments inthis field. Reference is made hereonly to capillary tube methods withenforced vibration.

Mason, Jayaratne, and Woods (16)developed a microdrop applicator byusing a magnetic earphone as asource of vibration. A No. 30 hypodermic needle produced droplets of160 to 400 microns diameter at frequencies of about 300 Hz. The constant supply of liquid was maintainedby pressurizing the liquid container.By adjusting the amplitude of vibration and the flow rate of the liquid,it was possible to obtain a singlestable droplet stream of any desireddiameter between 280 to 400 microns.

EXPERIMENTAL PROCEDURE

Microdroplet Applicator

For the production of droplet sizesranging from 100 to 1,000 micronsdiameter, the vibrating capillary wasselected for its mechanical simplicityand precision (Figure 1). The vibrator was a commercially-availablemagnetic type earphone and thepower was supplied by a frequencygenerator-amplifier system. Controlof both frequency and amplitude ofthe needle vibration was possiblefrom 180 to 850 Hz. The supply ofliquid was maintained at a predetermined rate by a 12 RPH synchronousmotor with rack and pinion system.Feed rate was adjusted over a rangeof 0.2 to 1.5 microlitre per second. A0.25 ml syringe with 27 gauge hypodermic needle was mounted horizontally on a fixed frame and a point 5mm off the luer end of the needle wasconnected to a vibrating arm extending to the earphone diaphragm.

For larger drops ranging beyond1,000 microns diameter, another typeof applicator was developed (Figure2). A 25 microlitre syringe with a 27gauge square - tipped needle wasmounted on a microdispenser fromwhich uniform drops of 1,100 micronsdiameter or larger could be ejected,depending on the plunger setting.

Collecting Media

After considering the various collecting surfaces described above, apaper with glossy surface was selected. The spread factor calibrationis aimed at establishing a method of

67

Figure 1. Vibrating capillary type microdrop applicator for the production of uniformsized droplets of 100 to 1,000 microns diameter.

quick determination of size distribution both in indoor test and fieldtrials. Thus, simplicity in handlingand stability of impression after deposition is a basic requirement. Instead of non-absorbing surface, whichproduces rather complex phenomenadue to surface wettability and otherfactors, a medium absorbing surfacewas used in order to reduce error inspreading activity due to evaporationof deposited droplets. For photographic contrast, a white surface waspreferred. It is commercially available as "Mead Mark I" manufacturedby E. B. Eddy Co.

A paper of 20 x 40 mm wasmounted on a microscope slide,alongside a second slide containing a20 x 4 mm well, partially filled withparaffin oil, in which the true diameters of any deposited dropletscould be measured. This latter slidewas kept in a cool"cabinet before exposing to droplet application in orderto avoid any spreading of the droplets beneath the oil layer. Then, bothslides were simultaneously exposed toa spray with predetermined dropletsize produced by a micro dropletapplicator.

Close-up Photography

For higher photographic contrast,a water soluble black dye, "NigrosineG140", was selected after examiningthe color contrast of black spots atvarious concentrations of dye. Thedye concentration was fixed through

out the tests at 1.0% on weight basis.Droplet samples of paper and oil

suspension were placed on a speciallydesigned close-up camera base andnegatives at 2 x magnification wereobtained giving transparent spots ona black background. Kodalith OrthoType III 35 mm film was foundsatisfactory for this purpose. Thenegative strips of smaller samplesprocessed were sent to USDA automatic particle analyzer at Wooster,Ohio, for sizing, while negatives oflarger size were scanned with amicroscope having an eyepiece rec-ticle, and size distribution was recorded on a blood cell counter.

Chemicals

2,4-D spray solution was tested atpractically applicable concentrations.Acid equivalent content was calculated from volume percentage ofcommercial formulation. Butoxy etha-nol ester formulation and amine saltformulation both at 2,4-D acid equivalent contents of 80 oz./Imp. gal.were separately mixed with distilledwater and dye.

Surface Tension and Viscosity

Fisher's torsion balance-type surface tension meter and Cannon-Fenske Boutine viscometers of sizenumbers 50 and 25 were used. Theester formulations were carefully prepared to maintain their stable stateof emulsion by stirring every 10 minutes before use. Five consecutive

measurements were made with eachspecimen of various concentrations ofone, two, three, four and five percent and a mean value for each concentration was obtained.

BESULTS

The value of surface tension decreased from 73.5 Dyne/cm/72F forwater to 61.6 by adding 1% dye todistilled water and to 35 Dyne/cm byadding 2,4-D concentrate at differentrates. The surfactant contained in theconcentrate reduced the spray liquidsurface tension; however, within thepractically applicable range of concentration (1 »>« 5% v/v) there wasno variation in the magnitude of thisreduction. The 2,4 - D emulsionshowed lower surface tension thanwater throughout the test. Increasingthe concentration of 2,4-D in theliquid increased kinematic viscosityfrom 0.705 stokes for water at 100Fto 0.830 for emulsion of 5% concentration. The viscosity of the esteremulsion was greater than that of theamine formulation at similar concentrations.

Droplet samples were taken at foursteps of size range on 5 each of paperslide and paraffin oil slide at threedifferent concentrations. Two histograms obtained from samples ofpaper slide and paraffin oil slidewere compared to obtain the ratio oftwo representative diameters. In bothcases there was no remarkable influence of droplet size on spread factor, but only of concentration of2,4-D. For both ester and amineformulations, the spread factor of applicable concentration of 2,4-D onthis sampling paper can be approximately estimated between 1.7 and2.1 at size range of 100 to 1,600 microns diameter (Figure 3).

DISCUSSION

After considering various methodsof determining f values, a glossypaper was found to be satisfactory asa sampling surface. In the case of anabsorbing surface, like filter paper,there is naturally a rapid and higherdegree of spreading activity of liquid.However, this spreading activity canbe understood to terminate at a timeof equilibrium of capillary movementand of evaporation of liquid. An absorbing surface with some roughnesswould result in distorted stains. Besides, edge fuzziness thus producedwould cause difficulty in analysis.

68 CANADIAN AGRICULTURAL ENGINEERING, VOL. 11, No. 2, NOVEMBER 1969

Figure 2. Microsyringe-dispenser type microdrop applicator for the production ofuniform sized droplets of 1,000 microns diameter.

The time of spreading of liquid overa filter paper surface, as discussedby Fournier D'Albe and Hidayetulla(11), would become the controllingfactor in case samples were takenunder different ambient humidity.

On the other hand, a non-absorbingsurface might eliminate this complexity and would result in sharpedged stain images as compared toabsorbing surfaces. There are stillsome possibilities of bouncing orshattering of droplets on impact andthe shape of stains may be distortedto a greater extent. A perfectly non-absorbing surface, like glass plate orElastic sheet, in this sense can easily

e contaminated with grease or filmof other undesirable substances before sampling and it seems rather

difficult to maintain the constantangle of contact against the liquidon this surface. This would bringforth unavoidable confusion in themethod attempted by Magono (15)and would result in modification ofhis empirical expression.

In this view, a medium rangedabsorbing surface with minimumroughness and with highest color contrast was selected. This material metrequirements of close-up photographyand permanent storage of specimens.

The reduction of surface tension onaddition of 2,4-D indicates the effectof the surfactant contained in theconcentrate. Naturally, spreading activity was encouraged by loweringthe surface tension; however, there

CANADIAN AGRICULTURAL ENGINEERING, VOL. 11, No. 2, NOVEMBER 1969

appear to be certain limits beyondwhich no further enhanced spreadingoccurs with further lowering of thesurface tension. This indicates thatthe spread factor is not controlledsimply by surface tension but byother complex factors as well.

Kinematic viscosity seems to berather influential on the spreadingactivity and could be one of the controlling factors. From a spray nozzleanalysis attempted by Dorman (8) afactor affecting the statistical diameter of droplets may be expressedas,

VMD factor = C. tj% . yV4 8

Where y is surface tension and 17is kinematic viscosity, respectively.From this approximation it becomesobvious that the spreading activity isa combined result of both surfacetension and kinematic viscosity ofliquid and also of other factors suchas roughness of surface, relative humidity of ambient air and the hardness of locally available water.

The spread factor determinationdescribed above is aimed at a reproducible calibration of spray depositsand the values obtained will beutilized in forthcoming spray nozzleemission tests and spray depositiontests with wind tunnel analysis. However, this method can be well appliedin field sampling of spray because ofits simplicity in operation and itsreproducibility.

CONCLUSIONS

Diameters of stain produced on aglossy paper surface and of dropletsuspended in oil matrix were measured to determine the spread factorof 2,4-D solution at various application rates of concentration. A vibrating capillary produced droplets ofhighly uniform size ranging from TOOto 1,000 microns diameter. Large sizedroplets were produced by a micro-syringe with dispenser apparatus. Theimages of black dyed solution on thewhite glossy paper were taken byclose-up photography and the negatives were processed with an automatic particle analyzer for size distribution.

The spread factor of 2,4-D solutions of ester and amine formulationwith black dye was calculated withina range of 1.7 to 2.1 for the ranges ofconcentration normally used in fieldapplications of these chemicals. Thevalues of spread factor obtained were

69

stable and applicable for calibrationof size distribution analysis of bothfield and laboratory sprays.

The influence of kinematic viscosity and surface tension of the solution to the spread factor were discussed and it was found that a combined effect of these two major contributing factors would control tosome extent the spreading activity ofliquid over the surface.

This work was partially supportedby a research grant from the CanadaDepartment of Agriculture. The sizedistribution analysis, carried out atWooster, Ohio on a flying-spot microscope, was authorized by an agreement between the U.S. Departmentof Agriculture and the SaskatchewanResearch Council.

2400 —

REFERENCES

1. Atkinson, W. R. and Miller, A. H.1965. Versatile technique for theproduction of uniform drops at aconstant rate and ejection velocity. Rev. Sci. Instru. 36:846-847.

2. Behrens, R. 1957. Influence ofvarious components on the effectiveness of 2,4,5,-T sprays. Weeds,5:183-196.

3. Bouse, L. F. 1967. Aerial spraypenetration through canopies.A.S.A.E. Tech. Paper 67-657, 27 p.

4. Brann, Jr. J. L. 1964. Factors affecting use of airblast sprayenTrans. A.S.A.E. 200-203.

5. Brazee, R. D. and Irons, F. 1965.The flying spot particle analyzer.USDA, AFED, Pub. C. A. No. 53,7 p.

400 800 1200

SPHERE DIAMETER (microns)

Figure 3. Ratio of stain versus sphere diameters. Receiving surface is a white glossypaper of medium absorbing property.

6. Carleton, J. B. 1967. Continuousrecording of H2O spray spotimages across the sprayed swathon 35 mm film. Jr. Econ. Ent.60:744-748.

7. Courshee, R. J. and Valentine,R. W. 1959. Use of plaster ofParis for recording spray drops.Jr. Agr. Eng. Res. 4:62-65.

8. Dorman, R. G. 1952. The atom-ization of liquid in a flat spray.Brit. Jr. Appl. Physics, 3:189-192.

9. Ennis, Jr. W. B. and Williamson,R. E. 1963. Influence of dropletsize on effectiveness of low-volume herbicidal sprays. Weeds,11:67-72.

10. Fournier D'Albe, E. M. and Hidayetulla, M. S. 1955. The break-upof large water drops falling atterminal velocity in free air. Qt.Jr. Roy. Met. Soc. 81:610-613.

11. Gebhardt, M. R. and Bode, L. E.1967. An improved method of determining the spread factor ofwater droplets. USDA, ARSPaper 42-131, 3 p.

12. Jarman, R. T. 1956. Stains produced by drops on filter papers,Qt. Jr. Roy. Met. Soc. 82:352.

13. Jarman, R. T. 1958. The deposition of wind-borne oil droplets onvarious horizontal surfaces. Jr.Agr. Eng. Res. 3:131-136.

14. Lehman, S. K., Haas, R. H. andRobinson, E. D. 1964. Preliminaryevaluation of droplet size and distribution of water in oil emulsionsprays applied with the bifluidsystem. Proc. S. Weed Conf. 17:253-261.

15. Magono, C. 1953. Volume distribution of the large precipitationelements. Jr. Met. Soc. Japan, 31:286-297.

16. Mason, B. J., Jayaratne, O. W.,and Woods, J. D. 1963. An improved vibrating capillary devicefor producing uniform waterdroplets of 15 to 500 microns inradius. Jr. Sci. Instru. 40:247-249.

17. May, K. R. 1950. The measurement of airborne droplets by themagnesium oxide method. Jr. Sci,Instru. 27:128-130.

18. Middleton, M. R. and Lowe, B.1967. Kaolin coated targets forthe collection of spray deposits.Agr. Avia. 9:46-48.

19. Roth, G. A. and Reins, G. E. 1957.Rotating disk apparatus for theproduction of droplets of uniformsize. Weeds, 5:197-205.

70 CANADIAN AGRICULTURAL ENGINEERING, VOL. 11, No. 2, NOVEMBER 1969