at SciVerse ScienceDirect

Food Hydrocolloids 33 (2013) 320e326

Contents lists available

Food Hydrocolloids

journal homepage: www.elsevier .com/locate/ foodhyd

Modification of emulsion properties by heteroaggregation ofoppositely charged starch-coated and protein-coated fat droplets

Yingyi Mao, David Julian McClements*

Department of Food Science, University of Massachusetts, Amherst, MA 01003, USA

a r t i c l e i n f o

Article history:Received 20 February 2013Accepted 27 March 2013

Keywords:HeteroaggregationEmulsionsModified starchWhey proteinStabilityReduced fat

* Corresponding author. Tel.: þ1 413 545 1019.E-mail address: mcclements@foodsci.umass.edu (D

0268-005X/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.foodhyd.2013.03.014

a b s t r a c t

The manuscript describes the formation and characterization of emulsions formed by controlled het-eroaggregation of oppositely charged fat droplets coated by either cationic whey protein isolate (WPI) oranionic modified starch (MS). Heteroaggregation was induced by mixing two the oppositely charged20 wt% oil-in-water emulsions together (d43 z 380 nm) at pH 3.5. The mean particle diameter, electricalcharge, and micro-structure of the heteroaggregates formed were measured as a function of negative-topositive particle ratio (0e100%) and pH (2.5e6.5). WPI-coated fat droplets were positive below pH 5.5and negative above this value, while MS-coated fat droplets were negatively charged across the entire pHrange. Upon mixing the two types of fat droplets together we found that the largest aggregates andhighest viscosity occurred at a particle ratio of 70% MS and 30% WPI, which was attributed to strongelectrostatic attraction between the oppositely charged droplets. The heteroaggregates partly dissociatedat certain pH values, which was attributed to weakening of the electrostatic attraction between thedifferent droplets. Heteroaggregates formed by oppositely charged fat droplets may be useful for creatingspecific food structures that lead to desirable physicochemical properties, such as enhanced viscosity atlow fat content.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The high hedonic rating of many commercial food products isrelated to their relatively high fat contents, e.g., mayonnaises,sauces, dressings, and desserts (Drewnowski, 1997a, 1997b). How-ever, fat has a higher caloric value and a weaker satiety/satiationeffect (on a per weight basis) than other major food nutrients(Blundell et al., 2005; Cooling & Blundell, 1998), which may lead topassive overconsumption and an increase in overweight andobesity (Micha & Mozaffarian, 2010). The development of reducedfat foods is therefore a major goal of many food scientists (Nehir El& Simsek, 2012; Rogers, 2009). Unfortunately, it is difficult toremove fats from fatty foods without adversely affecting theirdesirable quality attributes. Fats play an important role in deter-mining the appearance, texture, and flavor of foods, and when theyare removedmany of the desirable qualities are lost (McClements &Demetriades, 1998). The food industry is therefore searching foreffective strategies to reduce the fat content of foods, withoutadversely altering the desirable quality attributes. Various fat-reduction strategies have been developed, including the use of

.J. McClements).

All rights reserved.

indigestible fats, reduced calorie fats, thickening agents, andcolloidal particles (Williams & Buttriss, 2006).

One approach that we have recently studied in our laboratory isthe utilization of heteroaggregation of oppositely charged fatdroplets to create highly viscous or paste like materials withreduced fat contents (Mao & McClements, 2011, 2012a, 2012b,2012c). Heteroaggregation is defined as the aggregation of dis-similar particles, which may differ in their size, shape, charge,chemical composition, or other properties (Lopez-Lopez, Schmitt,Moncho-Jorda, & Hidalgo-Alvarez, 2006; Yates, Franks, Biggs, &Jameson, 2005). This approach has been widely used in non-foodscience applications for a variety of reasons, such as controllingthe rheological properties of ceramics (Piechowiak et al., 2012),creating ion exchange columns (Han, Daniels, Sudol, Dimonie, &Klein, 2013), removing colloidal particles from solutions (Findlay,Thompson, & Tipping, 1996), and encapsulating and targeting bio-molecules (Lemmers, Sprakel, Voets, van der Gucht, & Stuart, 2010;McParlane et al., 2012; Ohsugi, Furukawa, Kakugo, Osada, & Gong,2006; Spruijt et al., 2011). Our previous studies have shown thatheteroaggregates can be formed from food ingredients by mixingtwo emulsions together containing oppositely charged protein-coated fat droplets (Mao & McClements, 2011, 2012a). Previously,we used lactoferrin (LF) and b-lactoglobulin (b-Lg) at neutral pH tocreate droplets with different charges, since LF has an isoelectric

Y. Mao, D.J. McClements / Food Hydrocolloids 33 (2013) 320e326 321

point around 8 and is therefore positive, whereas b-Lg has anisoelectric point around 5 and is therefore negative (Mao &McClements, 2011). However, LF and b-Lg are highly purified pro-teins that are too expensive for widespread use as functional in-gredients in many foods. It is therefore important to identifyalternative emulsifiers that are more suitable for commercialapplication in the food industry. Previous researchers have shownthat novel textural attributes could be created in model foods bymixing an emulsion containing b-lactoglobulin-coated fat dropletswith one containing gum arabic-coated fat droplets under acidic(pH 4.2) conditions (Schmitt & Kolodziejczyk, 2009). In this study,we investigated the possibility of using whey protein isolate (WPI)and modified starch (MS) as two food-grade emulsifiers withdifferent electrical characteristics to induce heteroaggregation inmixed emulsions.

Whey protein isolate is a mixture of surface-active globularproteins normally isolated from diary milk, such as b-lactoglobulin,a-lactalbumin and bovine serum albumin (Hu, McClements, &Decker, 2003). WPI-coated droplets are positively charged belowthe isoelectric point of the adsorbed globular proteins (pI z 5), butnegatively charged above this value (Demetriades, Coupland, &McClements, 1997). Modified starch is formed by covalentlyattaching non-polar (octenyl succinic anhydride) groups tohydrophilic starch molecules to form an amphiphilic moleculethat can adsorb to oil-water interfaces and stabilize emulsions(McClements, Decker, & Weiss, 2007; Trubiano, 1995). MS-coatedfat droplets have been shown to be negatively charged over awide pH range due to the presence of anionic groups on the MSbackbone (Charoen et al., 2011). At pH values below the isoelectricpoint of WPI, one would expect heteroaggregation to occur be-tween the anionic MS-coated droplets and the cationic WPI-coateddroplets. We hypothesized that these economical food-gradeemulsifiers may be suitable for producing novel desirable charac-teristics in reduced-fat products through the heteroaggregationprocess.

2. Experimental methods

2.1. Materials

Corn oil was purchased from a commercial food supplier(Mazola, ACH Food Companies, Inc., Memphis, TN). Powderedwheyprotein isolate (97.7 wt% protein) was supplied by Davisco FoodsIntl. (Eden Prairie, MN, U.S.A.). Powdered OSA-modified starch(PURITY GUM� Ultra) was supplied by the National Starch LLC(Bridgewater, N.J., U.S.A.). Glacial acetic acid, sodium acetate, andsodium azide were purchased from SigmaeAldrich (Sigma Chem-ical Co., St. Louis, MO) or Fisher Scientific (Pittsburgh, PA). All otherchemicals were purchased from SigmaeAldrich (St Louis, MO).Double distilled water was used to prepare all solutions.

2.2. Emulsion preparation

2.2.1. Preparation of Single-droplet Type Emulsions:Aqueous emulsifier solutions were prepared by dispersing

either WPI (1 wt%) or MS (4 wt%) powder into acetic acid buffer(10mM, 0.02% sodium azide, pH 7) and then stirring for 3 h at roomtemperature to ensure complete dispersion. These emulsifier levelswere selected based on previous studies that have shown thatemulsions with relatively small droplet diameters can be produced(Charoen et al., 2011). The pH of the aqueous emulsifier solutionswas then adjusted to pH 3.5 using 1 M HCl, which led to the for-mation of clear solutions. Oil-in-water emulsions were prepared byhomogenizing 20 wt% oil phase with 80 wt% aqueous phase atambient temperature. Coarse emulsions were formed using a hand

blender (M133/1281-0, 2 speed, Biospec Products Inc., ESGC,Switzerland) for 2min at level 2. These emulsions were then passedfour-times through a two-stage homogenizer (LAB 1000, APV-Gaulin, Wilmington, MA) at a first-stage pressure of 5400 psi anda second-stage pressure of 600 psi. All emulsions were stored for24 h prior to utilization. The diameters of the WPI- and MS-stabilized emulsions produced were similar: d43 z 0.37 mm.

2.2.2. Preparation of Mixed-droplet Type Emulsions:Initially, two series of emulsions stabilized byeither 1%WPI or 4%

MS were prepared in acetic acid buffer (pH 3.5). Mixed emulsionscontaining a total of 10 wt% oil were prepared by mixing the twosingle emulsions together (each also containing 10 wt% oil), stirringfor 10 min, and then storing them for 24 h before analysis. Thisresulted in mixed emulsions containing different amounts of WPI-coated droplets (0e10 wt%) and MS-coated droplets (10e0 wt%).After these steps, the mixed emulsions contained different massratios of negative-to-positive particles, i.e., MS-coated-to-WPI-coated oil droplets. For convenience, we used the notation“70%MS” to refer to mixtures containing 70 wt% MS-stabilized emulsionand 30% wt% WPI-stabilized emulsion.

2.3. Influence of pH on emulsion stability

The influence of pH on the electrical charge, stability, andrheology of WPI-emulsions, MS-emulsions, and mixed-emulsions(70% MS) were examined. Emulsions were adjusted to a partic-ular pH value (pH 2.5e6.5) using either 1 M HCl or 1 M NaOH, andthen stored overnight prior to analysis.

2.4. Particle characterization

2.4.1. Particle Size:The particle size distribution of the emulsions was measured

using a laser diffraction particle size analyzer (Mastersizer 2000;Malvern Instruments, Ltd., Worcestershire, UK). To avoid multiplescattering effects the emulsions were diluted to a droplet concen-tration of approximately 0.005 wt% using pH-adjusted water at thesame pH as the sample. The emulsions were stirred continuouslythroughout the measurements to ensure they were homogenous.Measurements are reported as the volume-weighted mean diam-eter: d4;3 ¼ P

din4i =P

din3i , where ni is the number of droplets ofdiameter di.

It should be noted that particle size measurements made bystatic light scattering on highly flocculated emulsions should betreatedwith caution. First, the theory (Mie theory) used to interpretlight scattering data assumes that the scattering particles are ho-mogeneous spheres with well-defined refractive indices. In prac-tice, flocs are non-spherical and non-homogeneous particles, withill-defined refractive indices. Second, the process of dilution andstirring may have altered the dimensions and structural organiza-tion of the flocs. Consequently, the reported particle sizes shouldonly be treated as an indication of strong droplet association ratherthan a measure of the actual size of any aggregates present in theoriginal non-diluted samples.

2.4.2. Particle Charge:The z-potential of the particles in the emulsions was determined

using a particle electrophoresis instrument (Zetasizer Nano ZS se-ries, Malvern Instruments, Worcestershire, UK). Emulsions werediluted to a droplet concentration of approximately 0.001 wt% us-ing pH-adjusted water to avoid multiple scattering effects. Afterloading the samples into the instrument they were equilibrated forabout 120 s before particle charge data was collected over 20continuous readings.

-10

0

10

20

30

40

50

60

70

0% 20% 40% 60% 80% 100%

ζ-po

tent

ial (

mV

)

Percentage of Negative Droplets

Fig. 1. Change in electrical characteristics of the particles in mixed emulsions con-taining cationic WPI-coated droplets and anionic MS-coated droplets (pH 3.5).

Y. Mao, D.J. McClements / Food Hydrocolloids 33 (2013) 320e326322

2.5. Rheological properties

The rheological behavior of samples was characterized using adynamic shear rheometer (Kinexus Rotational Rheometer, Malvern,U.K.). A cup and bob geometry consisting of a rotating inner cyl-inder (diameter 25 mm) and static outer cylinder (diameter27.5 mm) was used in viscosity and oscillation measurements. Thesamples were loaded into the rheometer measurement cell andallowed to equilibrate at 25 �C for 5 min before beginning all ex-periments. Samples underwent a constant shearing treatment(1 s�1 for 10 min) prior to analysis to remove history effects. Theshear stress of the emulsions was then measured over a range ofshear rates (1e100 s�1), and the apparent viscosity was calculatedfrom this data.

2.6. Microstructure analysis

The microstructures of the emulsions were observed usingconfocal scanning fluorescence microscopy. The oil droplets in theMS emulsions were dyed with Bodipy 493, while the protein phasein the WPI emulsions was stained with Rhodamine B. For the MSemulsions, Bodipy 493 dye (0.1 mg/mL) was added to corn oil, andthis mixture was covered and stirred overnight to ensure the dyewas completely dissolved in the oil. A 20 wt% corn oil-oil-wateremulsion stabilized by 4% MS was then prepared with the dyedoil using the same procedure described previously. To avoidingfluorescence quenching, alumina foil was used to cover and protectthe dye solutions in all procedures. For the WPI emulsions,Rhodamine B was dissolved in distilled water at a concentration of0.05% (w/v). This solution was then added to a 1% WPI solution at aconcentration of 5 ml/g and then a WPI emulsion was prepared asdescribed previously. Bodipy 493 was excited with a 488 nm laserand detected at 515 nm. Rhodamine B was excited with a 543 nmlaser and detected at 605 nm. The pinhole setting was small(33.3mm), and the imagewasmagnified by a factor of 4� using thedigital zoom feature. All pictures were taken using a 10 � eyepiecewith a 60 � objective lens (oil immersion).

2.7. Statistical analysis

All experiments were carried out in either duplicate or triplicateusing freshly prepared samples. Results are reported as the calcu-lated means and standard deviations.

3. Results and discussion

3.1. Properties of single-droplet type emulsions

Initially, we measured the particle diameter and electricalcharacteristics of the two single-droplet type emulsions. At pH 3.5,emulsions containing MS-coated fat droplets had a mean particlediameter (d43) of 0.37 mm and a z-potential of �1.8 mV. The nega-tive charge on the MS-coated lipid droplets can be attributed to thepresence of carboxyl groups (eCOO-) on the OSA-modified starchmolecules (Shogren, Viswanathan, Felker, & Gross, 2000). The WPI-coated fat droplets had a mean particle diameter (d43) of 0.37 mmand a z-potential ofþ62mV. The positive charge on theWPI-coatedlipid droplets can be attributed to the fact that the acidic pH usedwas well below the isoelectric point of the adsorbed proteins(pI z 5) (Charoen et al., 2011). The relatively small size of theparticles in these emulsions can be attributed to the fact thatthe repulsive forces acting between the droplets outweighed theattractive forces, thereby inhibiting aggregation. In the case of theWPI-coated droplets, the most important repulsive force iselectrostatic repulsion due to their high positive charge, but in the

case of MS-coated droplets it is steric repulsion due to the presenceof a relatively thick hydrophilic adsorbed layer (McClements, 2005).Overall, these results show that the WPI- and MS-stabilizedemulsions contained droplets that had similar dimensions butopposite charges at pH 3.5.

3.2. Influence of particle ratio on particle properties in mixedemulsions

The initial objective was to understand the influence of the ratioof negative-to-positive particles on heteroaggregation in mixedemulsions containing MS-coated and WPI-coated droplets. Thesemixed emulsionswere formed bymixing together two 5wt% oil-in-water emulsions stabilized by either MS or WPI at different massratios. A lower total oil droplet concentration was used in thisexperiment to avoid gelation of the mixed emulsions, since thiswould have made particle size measurements difficult to perform.

As expected, the z-potential in the mixed systems went fromhighly positive to slightly negative as the concentration of nega-tively charged droplets (MS-coated) increased (Fig. 1). The point ofzero charge was around 80% negative droplets, which can beattributed to the fact that the magnitude of the positive charge onthe WPI-coated droplets (þ62 mV) was much higher than that ofthe negative charge on the MS-coated droplets (�1.8 mV). Hence, agreater number of anionic MS-coated droplets were required toneutralize the cationic WPI-coated droplets. A similar result wasfound in our previous work on the formation of microclusters inprotein-stabilized emulsions containing fat droplets with oppositecharges of different magnitude (Mao & McClements, 2011, 2012a).

The influence of negative-to-positive particle ratio on the size ofthe particles in themixed emulsions was alsomeasured (Fig. 2). Themean particle diameter of the single-droplet type emulsions wereboth relatively small (d43 z 0.37 mm) at pH 3.5, indicating that theywere stable to aggregation under these conditions. The dependenceof mean particle diameter on particle ratio can be divided into threeregimes: (i) from 0 to 30% MS, the particle size was relatively small(d43 < 0.5 mm); (ii) from 40% to 80% MS, the particle size wasrelatively large (d43 > 1 mm); (iii) from 90% to 100% MS, the particle

0

1

2

3

4

5

0% 20% 40% 60% 80% 100%

d 43(µ

m)

Percentage of Negative Droplets

Fig. 2. Change in mean particle diameter of the particles in mixed emulsionscontaining cationic WPI-coated droplets and anionic MS-coated droplets (pH 3.5).

Y. Mao, D.J. McClements / Food Hydrocolloids 33 (2013) 320e326 323

size was again relatively small (d43 < 0.5 mm). These measurementssuggest that relatively small clusters of droplets were formed at lowand high particle ratios, but that large clusters were formed at in-termediate values. The largest clusters (d43 > 5 mm) were formed at70% MS-coated droplets and 30% WPI-coated droplets (Fig. 2),which was close to the composition where the net charge on theclusters was zero (Fig. 1).

Confocal fluorescent microscopy images of the single and mixedemulsions also showed extensive droplet aggregation at interme-diate particle ratios (Fig. 3). Images of the emulsions containingeither MS-coated or WPI-coated droplets indicated that they con-tained small droplets evenly dispersed throughout the system. Onthe other hand, images of mixed emulsions (70% MS, 30% WPI)showed that they contained large clusters of droplets. The fact thatthese clusters appeared yellow in the fluorescent microscopy im-ages suggests that they contained a mixture of different droplets.

The main driving force for fat droplet association in these sys-tems is presumably electrostatic attraction between the oppositelycharged droplets. At relatively low particle ratios, one would expectany negatively charged droplets to be surrounded by positivelycharged droplets resulting in the formation of relatively small

Fig. 3. Change in microstructure of emulsions containing cationic WPI-coated d

clusters (Fig. 4). Similarly, at relatively high particle ratios, onewould expect any positively charged droplets to be completelysurrounded by negatively charged droplets, again resulting in theformation of small clusters. On the other hand, at intermediateparticle ratios one would expect that the clusters could grow quitelarge due to the possibility of electrostatic attraction betweenmanymore droplets (Fig. 4). In addition, there may be further aggregationof clusters whose net charge is close to zero, because there wouldbe no strong electrostatic repulsion to overcome the van der Waalsattraction between them.

Statistical thermodynamics models of the aggregation state ofparticles in mixed colloidal systems suggest that the maximumamount of aggregation should occur in systems containing 50%positively charged and 50% negatively charged particles (Dickinson,2011). There are a number of possible reasons for the differencebetween the theoretical predictions and our experimental mea-surements: (i) the mixed emulsions were not at thermodynamicequilibrium; (ii) the two different kinds of droplet had verydifferent charge magnitudes; and (iii) non-adsorbed biopolymersplayed a role. In particular, the magnitude of the positive charge onthe WPI-coated droplets was much greater than the magnitude ofthe negative charge on the MS-coated droplets, which may havehad important consequences on the structural organization of thefat droplets within the aggregates.

3.3. Influence of particle ratio on rheology and creaming of mixedemulsions

The rheology and creaming stability of emulsions is known todepend strongly on the aggregation state of the droplets(McClements, 2005). We therefore measured the rheological prop-erties and appearance of mixed emulsions with different particleratios at pH 3.5.

The flow profiles (apparent viscosity versus shear rate) ofselectedmixed emulsionswith different particle ratios are shown inFig. 5a. At low and high particle ratios the mixed emulsions hadrelatively low apparent shear viscosities (<0.03 Pa s) that did notchange appreciably with shear rate. On the other hand, at inter-mediate particle ratios (70% and 80% MS) the mixed emulsions hadrelative high viscosities (>0.5 Pa s) and showed distinct shearthinning behavior. The origin of this shear thinning flow behaviorcan be related to the progressive breakdown of the droplet clustersupon application of increasing external shear forces (Kondaraju,Farhat, & Lee, 2012). The influence of particle ratio on the apparentviscosity determined at a constant shear rate (1 s�1) is compared inFig. 5b. The mixed emulsions containing 70% and 80% MS-coateddroplets had the highest apparent viscosities, which correlates tothe samples with the highest amount of droplet aggregation in the

roplets, anionic MS-coated droplets or mixtures of the two types (pH 3.5).

Fig. 4. Highly schematic representation of proposed structures formed in mixedemulsions containing positively-charged and negatively-charged droplets at differentparticle ratios.

Y. Mao, D.J. McClements / Food Hydrocolloids 33 (2013) 320e326324

light scatteringmeasurements (Fig. 2). The apparent viscosity of themixed emulsions containing 70% MS were over 10-times higherthan either of the single-droplet type emulsions.

Visual observations of emulsion appearance after the test tubescontaining them were inverted (Fig. 5b) supported the rheologymeasurements. Mixed emulsions with relatively low and highparticle ratios were fluid and flowed to the bottom of the tubes,whereas those with intermediate particle ratios were gelled andremained at the top of the tubes after inversion.

3.4. Influence of pH on properties of mixed emulsions

Finally, we examined the influence of pH on the properties ofthe mixed emulsions, since changes in pH will modulate the signand magnitude of the electrical charge on the biopolymer-coateddroplets, thereby influencing the nature of any aggregates formed.

As reported previously (Charoen et al., 2011; Qian, Decker, Xiao,& McClements, 2011), WPI-coated droplets went from highly pos-itive at low pH to highly negative at high pH (with a point of zerocharge around pH 5), while MS-coated droplets were slightlynegative across the whole pH range (Fig. 6a). These measurementssuggest that the droplets in the mixed emulsions should haveopposite charges below about pH 5. The electrical charge of theparticles in the mixed emulsions (70% MS, 30% WPI) was betweenthat of the single-droplet type systems, which suggests that dropletaggregation may have occurred and formed clusters with inter-mediate charges.

Fig. 5. a: Change in apparent viscosity of mixed emulsions containing cationic WPI-coatedb: Change in apparent viscosity (at 10 s�1) of mixed emulsions containing cationic WPI-co

As reported previously (Charoen et al., 2011; Qian et al., 2011),MS-coated droplets were stable to droplet aggregation across thewhole pH range, whereas WPI-coated droplets were only stable atlow and high pH values but aggregated around their isoelectricpoint (pI z 5) (Fig. 6b). The difference in pH-stability betweenthese two emulsifiers can be attributed to their different stabili-zation mechanisms (McClements, 2005). MS-coated droplets areprimarily stabilized by steric repulsion and are therefore notstrongly affected by pH, whereas WPI-coated droplets are pri-marily stabilized by electrostatic repulsion and therefore tend toaggregate around their isoelectric point (Dickinson, 2003, 2010;McClements, 2004). The mean particle sizes were relatively low inthe mixed emulsions at the lowest (pH 2.5) and highest (pH 6.5)pH values studied in this work (Fig. 6b), but were relatively high atintermediate pH values (3.5 � pH � 5.5). The reason that aggre-gation was not observed at pH 2.5 may have been because the MS-coated droplets lost most of their negative charge and thereforecould not interact with the positive WPI-coated droplets (Fig. 6a).Conversely, the reason that aggregation did not occur at pH 6.5may have been because both the MS- and WPI-coated dropletswere strongly negatively charged (Fig. 6a), and therefore therewould be an electrostatic repulsion between them preventingaggregation. At intermediate pH values the electrostaticattraction between the droplets was strong enough to promoteheteroaggregation.

The rheological measurements were in close agreement withthe light scattering measurements. The viscosity of the MS-stabilized emulsions remained relatively low and constant acrossthe entire pH range, suggesting that there was no droplet aggre-gation. The viscosity of the WPI-stabilized emulsions was relativelysmall at low and high pH values, suggesting that the dropletsremained non-aggregated, but was very high around the isoelectricpoint suggesting that extensive droplet aggregation occurred(Fig. 6c). The viscosity of the mixed emulsions was highest at in-termediate pH values because of extensive droplet aggregationdriven by electrostatic attraction of oppositely charged dropletsleading to the formation of large clusters.

The viscosity of non-aggregated MS-stabilized emulsions wasconsiderably higher than that of non-aggregated WPI-stabilizedemulsions. It is likely that a fraction of the emulsifiers in theseemulsions was not absorbed to the droplet surfaces, but wasinstead dispersed within the aqueous phase. These free biopolymermolecules will contribute to the aqueous phase viscosity. TheMS-stabilized emulsions contained a higher overall biopolymer

droplets and anionic MS-coated droplets as a function of applied shear rate (pH 3.5).ated droplets and anionic MS-coated droplets (pH 3.5).

-40

-30

-20

-10

0

10

20

30

40

50

60

70

2 3 4 5 6 7

ζ-po

tent

ial (

mV

)

pH

WPI

MS

Mixed

0

5

10

15

20

2 3 4 5 6 7

d 43

(µm

)

pH

WPI

MS

Mixed

0.001

0.01

0.1

1

10

2 3 4 5 6 7

App

aren

t Vis

cosi

ty (

Pa·

s)

pH

WPI

MS

Mixed

a

c

b

Fig. 6. a: Influence of pH on the electrical characteristics of the particles in emulsions containing WPI-coated droplets, MS-coated droplets or a mixed emulsion (70% MS: 30% WPI).b: Influence of pH on the mean diameter of the particles in emulsions containing WPI-coated droplets, MS-coated droplets or a mixed emulsion (70% MS: 30% WPI).c: Influence ofpH on the mean diameter of the particles in emulsions containing WPI-coated droplets, MS-coated droplets or a mixed emulsion (70% MS: 30% WPI).

Y. Mao, D.J. McClements / Food Hydrocolloids 33 (2013) 320e326 325

concentration than theWPI-stabilized emulsions, and the structureof MS molecules is more open and flexible than that of WPI mol-ecules, which would lead to a higher viscosity enhancement for MS(McClements, 2000).

4. Conclusions

This study has shown that the rheological properties of emul-sions can be controlled by mixing negative polysaccharide-coateddroplets with positive protein-coated droplets to induce dropletaggregation through electrostatic attraction. The rheological char-acteristics of the mixed emulsions could be modulated by varyingthe particle ratio and solution pH. At low and high negative-to-positive particle ratios, low viscosity mixed emulsions were ob-tained because of the small droplet clusters formed, but at inter-mediate particle ratios highly viscous or paste-like emulsions wereproduced due to the formation of large droplet clusters. Overall,this study provides valuable insights into the formation of highlyviscous and paste-like food materials by controlled hetero-aggregation of oppositely charged oil droplets. These materialswere assembled entirely from commonly used food-ingredients(corn oil, modified starch, and whey protein isolate), and there-fore this approach may be useful for modifying the texturalproducts of commercial food products.

Acknowledgments

This material is based upon work supported by the CooperativeState Research, Extension, Education Service, United State Depart-ment of Agriculture, Massachusetts Agricultural Experiment Sta-tion and United States Department of Agriculture, CREES, NRI, andAFRI Grants. We greatly thank Danisco for donating the b-lacto-globulin and Afaf Makarious (National Starch LLC) for donating theOSA-modified starch used in this study.

References

Blundell, J. E., Stubbs, R. J., Golding, C., Croden, F., Alam, R., Whybrow, S., et al.(2005). Resistance and susceptibility to weight gain: individual variability inresponse to a high-fat diet. Physiology & Behavior, 86(5), 614e622.

Charoen, R., Jangchud, A., Jangchud, K., Harnsilawat, T., Naivikul, O., &McClements, D. J. (2011). Influence of biopolymer emulsifier type on formationand stability of rice bran oil-in-water emulsions: whey protein, gum arabic, andmodified starch. Journal of Food Science, 76(1), E165eE172.

Cooling, J., & Blundell, J. (1998). Are high-fat and low-fat consumers distinct phe-notypes? Differences in the subjective and behavioural response to energy andnutrient challenges. European Journal of Clinical Nutrition, 52(3), 193e201.

Demetriades, K., Coupland, J. N., & McClements, D. J. (1997). Physical properties ofwhey protein stabilized emulsions as related to pH and NaCl. Journal of FoodScience, 62(2), 342e347.

Dickinson, E. (2003). Hydrocolloids at interfaces and the influence on the propertiesof dispersed systems. Food Hydrocolloids, 17(1), 25e39.

Y. Mao, D.J. McClements / Food Hydrocolloids 33 (2013) 320e326326

Dickinson, E. (2010). Flocculation of protein-stabilized oil-in-water emulsions.Colloids and Surfaces B-Biointerfaces, 81(1), 130e140.

Dickinson, E. (2011). Simple statistical thermodynamic model of the hetero-aggregation and gelation of dispersions and emulsions. Journal of Colloid andInterface Science, 356(1), 196e202.

Drewnowski, A. (1997a). Taste preferences and food intake. Annual Review ofNutrition, 17, 237e253.

Drewnowski, A. (1997b). Why do we like fat? Journal of the American Dietetic As-sociation, 97(7 Suppl.), S58eS62.

Findlay, A. D., Thompson, D. W., & Tipping, E. (1996). The aggregation of silica andhaematite particles dispersed in natural water samples. Colloids and Surfaces A-Physicochemical and Engineering Aspects, 118(1e2), 97e105.

Han, S. J., Daniels, E. S., Sudol, E. D., Dimonie, V. L., & Klein, A. (2013). Preparation ofanionic ion exchange latex particles via heteroaggregation. Journal of AppliedPolymer Science, 127(5), 3601e3612.

Hu, M., McClements, D. J., & Decker, E. A. (2003). Lipid oxidation in corn oil-in-wateremulsions stabilized by casein, whey protein isolate, and soy protein isolate.Journal of Agricultural and Food Chemistry, 51(6), 1696e1700.

Kondaraju, S., Farhat, H., & Lee, J. S. (2012). Study of aggregational characteristics ofemulsions on their rheological properties using the lattice Boltzmann approach.Soft Matter, 8(5), 1374e1384.

Lemmers, M., Sprakel, J., Voets, I. K., van der Gucht, J., & Stuart, M. A. C. (2010).Multiresponsive reversible gels based on charge-driven assembly. AngewandteChemie-International Edition, 49(4), 708e711.

Lopez-Lopez, J. M., Schmitt, A., Moncho-Jorda, A., & Hidalgo-Alvarez, R. (2006).Stability of binary colloids: kinetic and structural aspects of heteroaggregationprocesses. Soft Matter, 2(12), 1025e1042.

McClements, D. J. (2000). Comments on viscosity enhancement and depletionflocculation by polysaccharides. Food Hydrocolloids, 14(2), 173e177.

McClements, D. J. (2004). Protein-stabilized emulsions. Current Opinion in Colloid &Interface Science, 9(5), 305e313.

McClements, D. J. (2005). Food emulsions: principles, practice, and techniques (2nded.). Boca Raton: CRC Press.

McClements, D. J., Decker, E. A., & Weiss, J. (2007). Emulsion-based deliverysystems for lipophilic bioactive components. Journal of Food Science, 72(8),R109eR124.

McClements, D. J., & Demetriades, K. (1998). An integrated approach to the devel-opment of reduced-fat food emulsions. Critical Reviews in Food Science andNutrition, 38(6), 511e536.

McParlane, J., Dupin, D., Saunders, J. M., Lally, S., Armes, S. P., & Saunders, B. R.(2012). Dual pH-triggered physical gels prepared from mixed dispersions ofoppositely charged pH-responsive microgels. Soft Matter, 8(23), 6239e6247.

Mao, Y. Y., & McClements, D. J. (2011). Modulation of bulk physicochemical prop-erties of emulsions by hetero-aggregation of oppositely charged protein-coatedlipid droplets. Food Hydrocolloids, 25(5), 1201e1209.

Mao, Y. Y., & McClements, D. J. (2012a). Fabrication of functional micro-clusters byheteroaggregation of oppositely charged protein-coated lipid droplets. FoodHydrocolloids, 27(1), 80e90.

Mao, Y. Y., & McClements, D. J. (2012b). Fabrication of reduced fat products bycontrolled heteroaggregation of oppositely charged lipid droplets. Journal ofFood Science, 77(5), E144eE152.

Mao, Y. Y., & McClements, D. J. (2012c). Modulation of emulsion rheology throughelectrostatic heteroaggregation of oppositely charged lipid droplets: Influenceof particle size and emulsifier content. Journal of Colloid and Interface Science,380, 60e66.

Micha, R., & Mozaffarian, D. (2010). Saturated fat and cardiometabolic risk factors,coronary heart disease, stroke, and diabetes: a fresh look at the evidence. Lipids,45(10), 893e905.

Nehir El, S., & Simsek, S. (2012). Food technological applications for optimalnutrition: an overview of opportunities for the food industry. ComprehensiveReviews in Food Science and Food Safety, 11(1), 2e12.

Ohsugi, A., Furukawa, H., Kakugo, A., Osada, Y., & Gong, J. P. (2006). Catch andrelease of DNA in coacervate-dispersed gels. Macromolecular Rapid Communi-cations, 27(15), 1242e1246.

Piechowiak, M. A., Videcoq, A., Ferrando, R., Bochicchio, D., Pagnoux, C., &Rossignol, F. (2012). Aggregation kinetics and gel formation in modestlyconcentrated suspensions of oppositely charged model ceramic colloids: anumerical study. Physical Chemistry Chemical Physics, 14(4), 1431e1439.

Qian, C., Decker, E. A., Xiao, H., & McClements, D. J. (2011). Comparison ofbiopolymer emulsifier performance in formation and stabilization of orange oil-in-water emulsions. Journal of the American Oil Chemists Society, 88(1), 47e55.

Rogers, M. A. (2009). Novel structuring strategies for unsaturated fats e meetingthe zero-trans, zero-saturated fat challenge: a review. Food Research Interna-tional, 42(7), 747e753.

Schmitt, C., & Kolodziejczyk, E. (2009). Protein-polysaccharide complexes: frombasicsto food applications. In P. A.Williams, &G.O. Phillips (Eds.),Gums and stabilizers forthe food industry 15 (pp. 211e222). Cambridge, U.K.: Royal Society of Chemistr.

Shogren, R. L., Viswanathan, A., Felker, F., & Gross, R. A. (2000). Distribution ofoctenyl succinate groups in octenyl succinic anhydride modified waxy maizestarch. Starch-Starke, 52(6e7), 196e204.

Spruijt, E., Bakker, H. E., Kodger, T. E., Sprakel, J., Stuart, M. A. C., & van der Gucht, J.(2011). Reversible assembly of oppositely charged hairy colloids in water. SoftMatter, 7(18), 8281e8290.

Trubiano, P. C. (1995). The role of specialty food starches in flavor emulsions. FlavorTechnology, 610, 199e209.

Williams, C., & Buttriss, J. (2006). Improving the fat content of foods. Boca Raton, FL:CRC Press.

Yates, P. D., Franks, G. V., Biggs, S., & Jameson, G. J. (2005). Heteroaggregation withnanoparticles: effect of particle size ratio on optimum particle dose. Colloids andSurfaces A-Physicochemical and Engineering Aspects, 255(1e3), 85e90.