An Electronic Imaging System for Determining Droplet Size

8/6/2019 An Electronic Imaging System for Determining Droplet Size

http://slidepdf.com/reader/full/an-electronic-imaging-system-for-determining-droplet-size 1/6

An Electronic Imaging System for Determining Droplet Sizeand Dynamic Breakdown of Protein Stabilized Emulsions

J.L. KLEMASZEWSKI, 2. HAQUE, and J.E. KINSELLA

ABSTRACT

A computerized imaging system, consisting of a light microscope anda light sensitive diode array interfaced with a computer, was devel-oped and evaluated for monitoring emulsion droplet size distributionand rate of emulsion droplet coalescence. This system compared fa-vorable with electron microscopic analysesand light scattering studiesfor determining emulsion droplet size. Oil in water emulsions stabi-lized with 0.05, 0.1, 0.2, 0.5, and 1.0% /3-lactoglobulin r caseinwere studied. It was observed that P-lactoglobulin stabilized emul-sions ormedsmallerdroplets hatcoalesced t a slower ate hen hosestabilized by acid casein. Emulsions stabilized by less than 0.2% pro-tein were extremely unstable.

INTRODUCTION

THE I NCREASED PRODUCTION of formulated and fabri-cated foods requires the availability of reliable functional in-gredients, and the increased demand for functional ingredientshas dramatized a corresponding need for methods to charac-terize the functional properti es of ingredi ents (Kinsella, 1984).Emulsifying agents are of primary importance in impartingstability to many food systems, such as milk, mayonnaise ,salad dressings, whipped toppings, and sauces. Emulsifiersmay function by one of several mechanisms; these includelowering the interfacial tension between the phases, surround-ing the emulsion droplet with a charged layer and providing aphysical barrier to the approach of other droplets (Walstra,1986). Proteins, which possessan amphiphilic nature by virtue

of their primarystructurewith polar and apolardomains, reuseful emulsifyingagents. n additi on o the electrostatx e-pulsion of charged groups, polypeptide loops and trains in aninterfacial fi lm provide a barrier to the close approach of oildroplets (Phillips, 1981). The breakdown of emulsions via co-alescence f oil droplets, esulti ng n decreasedurfaceareaand flocculation is retarded by interfacial protein films (Dick-inson and Stainsby, 1982). Creaming is slowed by homoge-nizing the droplets to a small size to minimize the effects ofthe density differences between the phases (Walstra and Oot-wijn, 1975). The wide range of hydrophobic and hydrophilicproperties available in different proteins can be exploited informulated food emulsion systems. Proteins also provide nu-tritional benefits not available from lipid-based emulsifiers,and they are not subject to the concern and scrutiny provoked

by various classes of chemical emulsifiers.The use of proteins as emulsion stabili zers has been the

subject of many studies (Halling, 1981). The properties ofemulsions stabilized by proteins vary with the equipment usedfor their preparation (Thorberg and Lundh, 1978). A standardprocedure has been improved by interfacing a single pistonvalve homogenizer with a computer to control energy input(Haque and Kinsella, 1989).

Several methods to characterize the stabilizing effects ofemulsifiers have been used. These include light scattering

Authors Klemaszewski and Kinsella are with the Inst. of FoodScience, Cornell Univ., Ithaca, NY 14853. Author Haque’s pres-ent address i s Dairy Science Dept., Mississippi State Univ., DrawerDD, Mississippi State, MS 39762.

440-JOURNAL OF FOOD SCIENCE-Volume 54, No. 2, 1989

(Walstra 1965, 1968) and (Pearce and Kinsella, 1978), elec-trical conductivity measurements Webb et al, 1970), centri-fugation (Sherman, 1971), emulsion capacity tests (Swift etal., 1961), solubility studies (Voutsinas et al., 1983), the Coul-ter counter (Walstra and Oortwijn, 1969) and electron micros-copy (Liboff et al., 1988). However, these methods lack thereal time reliability of studying emulsion breakdown with lightmicroscopy (Halling, 1981). Many of these available methodsare inadequate for protein stabil ized emulsions. For example,the Coulter counter uses salts that can alter the conformationof t he interfacial protein and may interfere with the state ofdispersion of the emulsion (Tadros and Vincent, 1986). Many

researchers now use light scattering methods to study emul-sions because of their rapidity and reproducibility (Kinsella,1984).

Studies using light microscopes are limited by t he lack ofmagnification that hinders accurate sizing of the smallest emul-sion droplets and is time consuming and tedious. The methodreported herein provides further magnification of the micro-scopic image and permits the sizing of smaller particles usinga light sensitive diode array. Because he emulsion is not ma-nipulated, real time studies of emulsion breakdown can beaccurately performed.

Protein based surface active agents are being increasinglydeveloped for use in food emulsions, and hence rapid andreliable methods are needed o asses heir emulsifying activity(EA) and emulsion stabilizi ng properties. The present researchdescribes the development of a method for determining EAand monitoring he breakdown f emulsions tabilizedby j3-lactoglobulin ndacid casein.

MATERIALS & METHODS

Materials

The emulsions in this study were stabilized by either p-lactoglogulinor casein. Pure D-lactoglogulin (A & B from bovine milk, crystallizedan lyophilized) ‘and iGda\ole (crystalline) were both obtaiied fromthe Simna Chemical Co. (St. Louis. MO\. Casein was nreoared bvthe hydrochloric acid precipitation of fresh raw skimmed’miik (Da<-ies, 1936). Emulsions were made with pure peanut oil that did notcontain added emulsifiers. The specific gravity of this oil was 0.91.

The Staining solution of 0.1% Congo red, a hydrophilic dye, con-tained 5% gelatin (Bovine skin, type?II) which-was+buffere-d o pH

7.0 with 0.05M imidazole buffer. The staining solution was main-tained as a liquid by keeping the solution at 90°C. The gelatin wasobtained from Si gma Chemical Co., and pure Congo red was obtainedfrom Electron Microsconv Science (Fort Washington. PAL All other

< _ I

chemicals used were reagent grade and distilled deionized water wasUsed hroughout this study.

The imaging system consistedof a light sensitivediode array mountedatop the right eyepiece of a n American Optical AO-150 light micro-scope. The light microscope was equipped with a 15X eyepiece forexamining fine emulsions and a 10X eyepiece for examining coarseemulsions. The di ode array is composed of a series of 32768 pixels,3.56per horizontal line and 128 per vertical line. The diode array wasinterfaced to a Turbo-XT computer via an input/output board. Thecomputer was used to set the threshold light level for the diodes, setthe exposure time for the pixels and was used to gather the imagedata from the diode array. The computer set the exposure time foi thepixels such that 25 to 30% of the pixels were transmitting 5 volts to

the computer. The exposure ime was changed as needed by the com-



fig. 1-A typical micrograph of an oil (light) in water (dark)

emulsion stabilized by 0.2% casein obtained using a comput-erized diode array imaging system to magnify light microscopeimage (2940 x).

puter. The lower refractive index of the oil, compared to water, pro-vided t he minimum contrast required for droplets to be distinguished.The use of Congo red accentuated he contrast. Data from the diodearray consisted of binary data reflecting the threshold light levels oneach of the pixels. The computer converted these data to an image onthe screen. The image ont he screen was a magnification of the emul-sion as seen by an experimenter through the eyepiece (Fig. 1). Thisimage was printed out for manual analyses. The magnification of t hediode array and the light microscope was approximately 55 00 foldwith an approximate resolution of 0.18~ when the 15X eyepiece wasused.

The imaging system was calibrated using an American Optical st agemicrometer marked in 0.1~ divisions. The calibration was confirmedusing latex microspheres of a known diameter. 2.04~. The micro-sphereswere obtained from t he Epics division of Coul ter Corp. (Hi-aleah, FL).

Methods

Emulsion preparation. A series of 8-lactoglogul in or acid caseinsolutions. 0.05, 0.1. 0.2. 0.5. 0.75 a nd 1.0% (w/w). was made LIDin 50 mM imidazole’to pH 7.01 The emulsions were made by initiallydispersing 12 g of protein solution and 8 g of oil with a Janke-KunkelTP 18-10 turbo blender for 5 set prior to passage hrough the valvehomogenizer, as described by Haque and Kinsella (1989). The ho-mogenizing system was kept at a constant 25°C using a circulatingwater bath. Emulsions were obtained by passing the protein solutionand oil mixture through a valve homogenizer for 180 strokes, corre-

sponding to an energy input of 640 X 10” J.rne3. The emulsion wastransferred to a beaker and slowly stirred to prevent creaming. Thetemperature of the emulsion was held constant at 25°C.

To monitor stability, i.e., changes n droplet size with time, aliquots(0.1 mL) of the emulsion were removed at hourly intervals and mixedin a testubewith 0.2 mL of the staining solution at 90°C. The stainedemulsion was immediately placed on a microscope slide and spreadthinly to obtain a uniform layer of emulsion. A cover slip was placedon the slide and the stained emulsion sample was then cooled on iceto solidify the warm gelatin and prevent further coalescence of theemulsion. Emulsions were then viewed using the imaging system, andthe image was printed out for measurement of the droplet diameters.Coalescencewas observed by a decrease n the emulsion surface areaand an increase in the average size of t he droplets with time.

Sample preparation for electron microscopy. To check the ac-curacy of the imaging system, casein stabilized emulsions were viewedusing scanning electron microscopy @EM). Emulsions samples for

the electron microscope were prepared n the same manner as for theimaging system. The methodology for sample preparation for SEMhas been decribed by Liboff et al. (1988).

Sample preparation for light scattering studies. Light scatteringmethods, based on the Mie theory (Van De Hulst, 1957), have beenapplied to t he study of emulsions (Walstra, 1965) and Pearce andKinsella (1978). Emulsion samples for t he light scattering studies wereprepared in t he same manner as for the imaging system. For t hesestudies, the emulsion, from the homogenizer, was immediately dilutedfrom 50 to 5000 fold in a 0.1% SDS solution buffered to pH 7.0 by0.05M imidazole-HCl to give an absorbance between 0.2 and 0.7.The d, and emulsion surface area were determined by the spectro-turbidimetric method of Walstra (1968) using a Gary 219 dual beamspectrophotometer with automatic base-line correction to scan thesamples from 350 to 870 mm. The cuvettes used or this study all oweda small acceptance angle of light to be passed through the sample.Surface area of the emulsion was also determined from the turbidity

at 550 nm using the equations of Pearce and Kinsella (1978).Calculations. By measuring the diameter of all the droplets on an

image, a histogram of the droplet size distribution for each view ofan emulsion was prepared. In order to make the experiment statisti-cally valid, over 100 droplets were measured for each sample. Theaverage droplet si ze of the emulsions were determined from the his-tograms prepared from the data obtained using the imaging system.The average droplet size was determined using equation (l), whered, is

d, =2

N(1)

the average droplet di ameter, nt is the number of droplets of size imicrons and N is the total number of droplets measured. The d, ofthe emulsion, described by Walstra (1965) as the average volume tosurface ratio, was determined from the histograms obtained by the

imaging system using Eq. (2).dx3

= X i3 ni

X i2 ni

The d, of t he emulsion was also determined using the light scat-tering method of Walstra (1965). This method involves gathering lightscattering data over a range of wavelengths and fitting the absorbanceobtained to equations for ideal theoretical turbidometric data for bestfit lines as described by Walstra (1965).

Surface area of emulsions were calculated by three different meth-ods. The method of Walstra (1968) uses the &s obtained from li ghtscattering data n equation (3), where 6 is the volume fraction of oil.

6.0S. Area = -

4s

The surface area obtained from Eq. (3) is in units of meter2 per

Volume 54, No. 2, 1989-JOURNAL OF FOOD SCIENCE-441



DETM. DROPLET SIZE DYNAMIC BREAKDOWN OF EMULSIONS. . .

Fig. 2-Scanning electron microscope view of emu(sion stabi-lized by 0.3% casein obtained using the methods of Liboff et al.(1988) (2704 x).

gram of emulsion. The method of Pearce and Kinsella (1978) gives

the surface area of the emulsion in units of meter* per gram of roteindirectly from Eq. (4), where A is the absorbance at 550 nm, 8. 1s hevolume

fraction of oil, C is the protein concentration of the solution used tomake the emulsion and 1 s the pathlength of the cuvette. The surfacearea as calculated using Eq. (4) is also known as the emulsifyingactivity (EA) for proteins.

The surface area of an emulsion can also be calculated from thehistogram obtained from the imaging system data. The total volumeand total surface area of all the droplets observed was determinedusing Eq. (5) and (6), where r is droplet radius. The “dilution factor”was determined using EQ. (7).

Volume = I: F (5)

Surface Area = I: 47rr2 (6)

Dilution Factor = Volume “Og OilResult of Eq. (5)

This dilution factor gave the factor with which the result of equation(6) was multiplied to calculate the surface area of l .Og of oil. Valuesof surface area obtained were all converted to meters2 per gram ofprotein, as these were the units used for comparing the emulsifyingactivity (EA) of proteins.

RESULTS

THE EMULSIONS studied in these experiments were com-

posed of a heterogenous population of di spersed spherical oildroplets, as shown by electron microscopy (Fig. 2). Theseemulsions, as observed using the imaging system, are shownas printouts of what was observed through the light microscope(Fig. 1). By measuring the diameter of the droplets on animage printout, a histogram of droplet size distribution for eachemulsion was prepared. The droplet sizes observed ranged from0.41~.or an emulsion stabili zed by 1% P-lactoglobulin to drop-lets larger than 40 p, in emulsions stabilized by dilute, 0.05%P-lactoglobulin.

The histograms revealed that the droplet size distributionswere normally distributed with one mode (Fig. 3). This typeof distribution is usual for a simple emulsion system (Walstra,1968). Emulsions stabilized by P-lactoglobulin were less het-erogeneous than those stabilized by acid casein, as shown by

comparing the range of droplet sizes of 1.6~ to 7.6~ for an

40

30

f 20

E

10

0

a

1.6 2.2 2.7 3.3 3.8 4.3 4.9 5.4 6.0 6.5 7.0 7.6

Dmplu Si (Micmns)

b

1.4 2.4 3.4 4.4 5.4 6. 5 7.5 a.5 9.5 10.5

DmplU sii (haicmns)

Fig. J-Histograms of emulsion droplet size distribution inemulsions stabilized by (a) 0.2% plactoglobulin after 0 hr; (b)

0.2% casein after 0 hr.

emulsion stabilized by 0.2% P-lactoglobulin (Fig. 3a) to therange of 1.4~ to 10.2~ for an 0.2% acid casein stabilizedemulsion (Fig. 3b). These two emulsions had comparable av-erage diameters of approximately 5~.

The change in emulsion stabil ity with time was observed as

the tendency for the droplet size distribution to increase withtime, as shown in Fig. 4, which are the combined histogramsobtained for 0.5% P-lactoglobulin stabilized emulsions 2 and6 hr after the emulsion was prepared. The distribution modefor the sample taken at 2 hr was l.Op,, as compared to 1.4~for the sample at 6 hr. The droplet size distribution of thesample taken after 2 hr is skewed to the right slightly morethan the 6 hr sample. Since the droplet size distribution of thisemulsion should be normal (Walstra, 1968), skewness indi-cates that some droplets were smaller than the resolving powerof the imaging system. The smallest droplets observed usingthe imaging system have diameters around 0.36~.

Data were gathered for similar emulsions stabilized withspecified concentrations of &lactoglobulin and acid casein asemulsifying agents, and the droplet size distribution of emul-

sions were studied by electron microscopic, light scattering




0.3 0.6 1.0 1.4 1.7 2.1 2.4 2.8 3.2 3.5

‘Fig. 4-Histograms of emulsion droplet size distribution inemulsions stabilized by 0.5% plactoglobulin after 2 hr and 6 hr

of coalescence.

8.0j T

3

8

6.0 -

z 4.0 -

d2.0 -

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Protein Concentration %)

Fig. 5-Effect of protein concentration and method of determi-nation on d,, of emulsions (as defined by Eq. (2). U-Q plac-toglobulin stabilized emulsion, observed using imaging system;u plactoglobulin stabilized emulsion, monitored using themethod of Walstra (1965); A-A casein stabilized emulsion, ob-served using imaging system; A-A casein stabilized emulsion,monitored using the method of Walstra (1965).

methods and the imaging system. The imaging system wasalso used to study the brea kdown of these emulsions.

Comparison with light scattering method. The averageemulsion droplet volume to surface diameter (&), determinedby light scattering (Walstra, 1968), was compared with the 4,of the same emulsions obtained using the imaging system. Thed,, of the emulsions was very comparable for emulsions sta-bilized by 0. 2% protein or greater (Fig. 5). However, whenprotein concentrations were reduced to less than 0.2%, thislight scattering method gave lower values for the &,. The 4%of an emulsion stabilized with 0.1% casein was determined tobe 25.8t~,according to the imaging syst em, but this light scat-tering method gave a d,, of 4.91.L.

It has been reported that the light scattering method used isnot accurate in studying coarse emulsions (Walstra, 1968).

800

0.2 0.4 0.6 0.8 1.0

Protein Concentration 96)

0.0 0.2 0.4 0.6 0.8 1.0

Pmein CImcentration 96)

1.2

Fig. 6-Relationship between methods for determination of sur-

face area for emulsions stabilized b y different concentrations ofproteins. (a) casein (b) plactoglobulin. Surface area determined

using a--o imaging system; M-H method of Walstra (1965);A-A method of Pearce and Kinsella (1978).

From the data gathered in these experiments, it appeared hatthe light scattering method was appropriate for emulsions withaveragedroplet diameters ess than 7.0~~.The results obtainedfrom light scattering for finer emulsions were much more var-iable than the data obtained from the imaging system, as thestandard error of the d,, was greater than OSp, for light scat-tering data (Fig. 5), wherea s it was less than 0.1~ for dataobtained from the imaging system. The correlation betweenthe imaging system and the light scattering method was quitehigh, 0.96, for emulsion droplets stabilized by greater than0.3% of either protein. The a,, as determined by these twomethods did not differ by more than 0.5l.~ or smaller droplets(Fig. 5). This agreement was offset by the high variation of

the light sca ttering method.The surface area of the emulsions was determined using themethods of Walstra (1965). Pearce and Kinsella (1978) andthe imaging system. The results of these experiments showedthat the light scattering method of Walstra (1965) again showedgreater disparity compared to the imaging system when study-ing co arse emulsions (Fig. 6). At protein concentrations essthan 0.2%, this method indicated a finer emulsion than wasactually present. Since the method of Walstra (1965) calculatessurface area directly from d.,,, using Eq. (3), it is expected hatthe same imitations apply to surface area calculations. At pro-tein concentrations greater than 0.2%, the imaging systemcompared avorably with the results obtained using the methodof Walstra (1965). The results derived from the light scatteringmethod of Pearceand Kinsella (1978) were comparable to theresults obtained with the imaging system at protein concentra-




DETM. DROPLET SIZE DYNAMIC BREAKDOWN OF EMULSIONS. . .

a

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Protein Concentration (46)

b

0.0 0.2 0.4 0.6 0.8 1.0 1.2

Protein Concentration (%)

Fig. 7-Effect of protein concentration on (A) average dropletdiameter and (B)‘emulsion surface area for emulsions stabilizedby M Plactoglobulin and A-A casein.

tions less than 0.2% and greater than 0.2%. The data obtainedfrom the imaging system and those by the method of Pearceand Kinsella (1978) were comparable at all the casein concen-trations studied (Fig. 6a), but emulsions stabilized with p-lactoglobulin did not compare as favorably (Fig. 6b). The methodof Pearce and Kinsella (1978) had a higher variability than theimaging system, but the variability was similar to that of theWalstra (1965) method. The method of Pearce and Kinsella(1978) is simple and is useful for estimating surface area ofemulsions with mean droplet sizes in the range of 2t.1, o 15p,surface area.

Initial emulsion characteristics. The average droplet sizeof emulsions stabilized by S-lactoglobulin and acid casein, at640 x lo6 J-m-’ of energy input, decreased with increasing

protein concentration (Fig. 7a). Emulsions stabilized by lessthan 0.2% protein were initially unstable, with average dropletsizes much greater than 5t.~. At protein concentrations greaterthan 0.2%, B-lactoglobulin formed emulsions with smalleremulsion droplets than acid casein. The average droplet sizeof the emulsions stabilized by 1.0% P-lactoglobulin was 1.14+,as compared to 1.9111. or 1.0% acid casein stabilized emul-sions. The effects of increasing protein concentration on drop-let size were small at protein concentrations greater than 0.2%.The average droplet size did not differ much between emul-sions stabilized by 0.5% and 1.0% protein for the proteinsstudied. B-Lactoglobulin may form finer emulsions than acidcasein because t may adsorb at the interface more rapidly thanacid casein and perform extensive molecular interactions. Theability of a protein to adsorb at interfaces influences its abilityto form emulsions (Dickinson and Stainsby, 1982), but sub-

0 2 4 6 a 10

Breakdown Tie (bows)

0 2 4 6 8 10

Breakdown Tie (hours)

Fig. 8-Effect of emulsion breakdown time on (a) average drop-let diameter and (b) emulsion surface area for emulsions sta-bilized by U-U 0.5% plactoglobulin; A-A 1.0% plactoglobulin;

W--W 0.5% casein; A-A 1.0% casein.

sequent interactions are also critically important (Graham andPhillips 1979; Kinsella, 1984).

The effects of protein concentration on unit surface area ofemulsion were similar, with surface area increasing with in-creased protein concentration. However, this was not the casefor surface area per unit protein at different protein concentra-tions (Fig. 7b). As protein concentration increased, the EAinitially increased to a maximum at 0.5% protein and thendecreased. This effect may be caused by protein forming athicker f ilm at the interface without an increase in surface areadirectly proportional to the amount of available protein. Athigher concentrations, the protein at the interface may not com-

-pletely unfold. At lower protein concentrations, the proteinmay unfold more extensively because it is thermodynamicallyfavorable to coat as much interfacial area as possible by un-folding (Graham and Phillips, 1979; Halling, 1981).

Emulsion stability. Data gathered using the imaging systemshowed that the average droplet size of emulsions increasedand the interfacial surface area decreased with time (Fig. 8).The average droplet diameter of an emulsion stabilized by 1.0%p-lactoglobulin increased from 1.14~ to 1.3111 after 5 hr ofstorage at 25°C. The average droplet diameter of an emulsionstabilized by 1.0% acid casein increased from 1.91b to 2.39p,in the same time period (Fig. 8a). The slopes of the bestfitlines for the breakdown of these two emulsions were 0.03t.r,/hr for 1.0% P-lactoglobulin and O.O9p&r for 1.0% acid casein.By comparing these slopes, it was observed that acid caseinstabilized emulsions had a faster breakdown rate than did p-




lactoglobulin stabilized emulsions. This effect was observed atboth protein concentrations studied. The faster breakdown rateof the acid casein stabilized emulsions may reflect, in part, thelarger droplets of the emulsion initially. Large droplets providea greater driving force for t he creaming of the emulsion andpossibly resulting in hi gher rate of droplet collisions. This couldresult in an increase in the rate of coal escence (Walstra andOortwign, 1975). In addition, caseins may form weaker filmsaround the fat globules which are desorbed or ruptured uponcontact of the droplets.

Increasing the protein concentration resulted in a retardationof t he rate of coalescence. Emulsions stabili zed by 1.0% pro-tein had a slower rate of droplet size increase than those sta-bilized by 0.5% protein (Fig. 8a). The slope of curve ofcoalescenceof an emulsion stabilized by 0.5% acid casein wasO.llpJhr, as compared to O.O9tL/hr or 1.0% acid casein. Theeffects of protein concentration on emulsion stability may alsoreflect, i n part, the slightly larger droplet size present at lowerstabilizing protein concentrations Walstra and Oortwign, 1975).Stability associatedwith concentrationmay also reflect a thickerinterfacial protein layer providing a stronger barrier to the ap-proach of other emulsion droplets and minimizing coalescence(Dickinson and Stainsby, 1982).

The surface area per gram of emulsion decreasedwith in-creasing breakdown t ime (Fig. 8b). Surface area was calcu-

lated in meter* per gram of emulsion to avoid con fusion whencomparing the coalescence rates of emulsions stabili zed bydiffering levels of protein concentration. The EA of an emul-sion stabilized by 0.5% protein would be greater than the EAof an emulsion with 1.0% protein if the emulsions were iden-tical in droplet size distribution. Data for the average diameterof emulsions was more precise than the surface area data ob-tained, even though both were calculated from the same dataobtained from the imaging system. This resulted from the in-ordinate contributions of small droplets to surface area as op-posed to volume in equations (5) and (6). To obtain the samedegreeof precision or surfaceareacalculations, a greaternumberof observations must be t aken as compared to the averagediameter.

Other applications. The imaging system was used to study

emulsions that had the oil phase dyed with Sudan black, ahydrophobic dye. This was-done p;ior to emulsification,-sothe lipid phase gave greater contrast. The imaging system gavesimilar data when either Sudanblack or congo red was used toincrease contrast. This indi cated that the imaging system couldbe used for the study of emulsions where the oil phase wascontinuous. With the use of high and low magnification eye-pieces, the imaging system could be used o study a wide rangeof fine and coarseemulsions Klemaszewski and Kinsella, 1987).

SUMMARY

USING THE DIODE assisted maging system described n thispaper, it was possible to study the droplet sizes and breakdown

of emulsions stabilized by acid casein and p-lactoglobulin. TheP-lactoglobulin gave emulsions hat had a smaller averagedropletsize, larger surface area, and a slower rate of breakdown atprotein concentrations greater than 0.2%. At protein concen-trations less than 0.2%, the emulsions were extremely unsta-ble. The data gathered compared favorably with light scatteringstudies and electron microscopy. The imagi ng system can alsobe used on systems other than oil in water emulsions.

REFERENCESDavies, W.L. 1936. The proteins of milk: Casein. In “The Chemistry of

Milk,” E.H. Tri (Ed.), D. Van Nostrand Co., Inc. New York.Dickinson, E. an B .tamsby, G. 1982. “Colloids in Foods,” Applied Science

Publishers, Barking, Essex, England.Graham, D.E. and Phillips, MC. 1979. Proteins at li

%;&;s of adsorptron and surface denaturatron. J. 2uid interfaces. !.011. nterface Ser.

. “ . Z “.

Halling, P.J. 1981. Protein stabilized foams and emulsions. CRC Crit. Rev.Food Sci. Nutr. 13: 166.

Haqu; 2 and Kinsella J E 1969 Emulsip ~~m&,i;;;~~eina De-velo ‘mint of a stand&&d method J 00”

Kinse la, J.E. 1984. Milk proteins: Pherties. CRC Critical Rev. Food Sci. r?

sicochemical and unctional prop-utr. 21(3): 197.

Klemarzewski, J.L. and Kinsella, J.E. 1987. Unpublished data. Instituteof Food Science, Cornell University, Ithaca, NY.

Liboff, M., Goff, H.D., Haque, Z., Jordan, W.K., and Kinsella, J.E. 1988.Changes in the ultrastructure of emulsions as a result of electron mi-

P::X.preparation procedures. Food Microstructure. In press.N. and Kinsella, J.E. 1978. Emulsifying properties of proteins:

evaluation of a turbidimetric technique. J. Agric. Food Chem. 26: 716.

Phillips, MC. 1981. Protein conformation at liquid interfaces and its rolein stabilizing emulsions. Food Technol. 3550.Sherman, P. 1971. Accelerated tests of emulsion stability. Soap Perfumery

Cosmetics 44: 693.Swift, C.E., Lockett, C., and Fryar, A.J. 1961. Comminuted meat emul-

sions: The capacity of meats for emulsifying fat. Food Technol. 15: 468.Tadros, T.F. and Vincent, B. 1986. Emulsion Stability. In “En clopedia

of Emulsion Technology,” P. Becher (Ed.), Marcel Dekker, nc., NewYork.

Tornberg, E. and Lundh, G. 1978. Functional characterization of proteinstabilized emulsions: standardized emulsifying procedure. J. Food Sci.43: 1653.

Van De Hulst, H.C. 1957. “Light Scattering by Small Particles,” JohnWiley and Sons, New York.

Voutsinas, L., Cheung, E., andNakai, S. 1983. Relationships of hydropho-biit2v6 o emulslfylng properties of heat denatured protems. J. Food Sa.

We.&& P.J. 1965. Light scattering by milk fat globules. J. Coil. InterfaceSc-i 97, AOR- _ . - . . _ _ “ .

Walstra! P.J. 1968. Estimating globule-size distribution of oil-in-wateremulsions by spectroturbidimetry. Br. J. Appl. Phys. 6: 1187.

Walstra, P. J. 1986. Formation of emulsions. In “Encyclopedia of Emulsion

Technology.” P. Becher, (Ed.), Marcel Dekker, Inc., New York.Walstraj P.J. and Oort\?ngn, H. 1969. Estimating globule-size distribution

;&&-m-water emulsions by Coulter Counter. Neth. Mdk Dany J. 29:-__.

Walstra, P.J. and Oortwign, H. 1975. Effect of globule size and concentra-tion on creaming in pasteurized milk. Neth. Milk Dairy J. 29: 263.

Webb, N.B., Ivey, F.J., Craiment of emulsion capacity 6

H.B., and Jones, V.A. 1970. The measure-y electrical resistance. J. Food Sci. 35: 601.

MS received 2/19/88; revised 7/11/88; accepted g/30/88.

Suuuort rom he National Dairy Research oard s aclmowledaed.


An Electronic Imaging System for Determining Droplet Size

Documents