at SciVerse ScienceDirect

Food Hydrocolloids 30 (2013) 358e367

Contents lists available

Food Hydrocolloids

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

Interfacial and emulsifying properties of sucrose ester in coconut milk emulsionsin comparison with Tween

Suwimon Ariyaprakai a,*, Tanachote Limpachoti a, Pasawadee Pradipasena b

aDepartment of Food Biotechnology, Faculty of Biotechnology, Assumption University, Ram Khamhaeng Rd. Soi 24, Hua Mak Campus, Bangkok 10240, ThailandbDepartment of Food Technology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand

a r t i c l e i n f o

Article history:Received 2 January 2012Accepted 6 June 2012

* Corresponding author. Tel.: þ66 2 300 4553x3796E-mail address: suwimona@yahoo.com (S. Ariyapr

0268-005X/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.foodhyd.2012.06.003

a b s t r a c t

In this study, sucrose esters were presented as a promising alternative to petrochemically synthesizedTweens for application in coconut milk emulsions. The interfacial and emulsifier properties of sucroseester (SE), mainly sucrose monostearate, had been investigated in comparison with Tween 60 (TW), anethoxylate surfactant. The interfacial tension measurement showed that SE had a slightly better ability tolower the interfacial tension at coconut oilewater interface. These surfactants (0.25 wt%) were applied incoconut milk emulsions with 5 wt% fat content. The effects of changes in pH, salt concentration, andtemperature on emulsion stability were analyzed from visual appearance, optical micrograph, dropletcharges, particle size distributions, and creaming index. Oil droplets in both SE and TW coconut milkemulsions extensively flocculated at pH 4, or around the pI of the coconut proteins. Salt addition inducedflocculation in both emulsions. The pH and salt dependence indicated polyelectrolyte nature of proteins,suggesting that the proteins on the surface of oil droplets were not completely displaced by either addednonionic SE or TW. TW coconut milk emulsions appeared to be thermally unstable with some coalescedoil drops after heating and some oil layers separated on top after freeze thawing. The change intemperature had much lesser influence on stability of SE coconut milk emulsions and, especially, it wasfound that SE emulsions were remarkably stable after the freeze thawing.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Sucrose esters, or sugar-based surfactants, are in current interestbecause they are produced from natural resources such as sucroseand vegetable oil. They are biodegradable and more biocompatiblewhen compared to other petrochemically synthesized surfactants.Sucrose esters are non-toxic and safe for food and are approved asa food additive under food regulations and laws in several countriesi.e. Japan, USA, and EC. It had been reported that sucrose esters haveexcellent emulsifying properties and be able to apply in various foodproducts (Garti, 2001). Their wide range of hydrophilicelipophilicbalance (HLB), depending on degree of esterification of fatty acidsand sucrose, provides ultimate application of sucrose esters to eachproduct type.

Coconut milk is one of food emulsions that require additionalsurfactants or emulsifiers to improve emulsion stability. Several recentworks have been carried out on physicochemical characterization of

; fax: þ66 2 300 4553x3792.akai).

All rights reserved.

coconut milk employing other small molecule surfactants such asTweens and sodium dodecyl sulfate (Tangsuphoom & Coupland,2009a,b), however, data on sucrose esters remain fewer. In thosestudies, coconut milk emulsions after addition of surfactants werereported to smaller in average droplet size, decrease in total surfaceprotein concentration, and change in droplet surface charges(Tangsuphoom & Coupland, 2008b, 2009a).

This study aims to gain more understanding on the relationshipbetween interfacial properties and stability of coconut milk emul-sions with addition of sucrose ester. We described the comparativeinterfacial and emulsifier properties between sucrose ester (SE),mainly sucrose monostearate, and Tween 60, or polyoxyethylenesorbitan monostearate (TW). Their structures were displayed inFig. 1. The difference between the carbohydrate and ethoxylateheadgroups provided an interesting comparison. Their adsorptionbehaviors at the coconut oil and aqueous interface were investi-gated by the interfacial tension measurement. The coconut milkemulsions prepared with SE and with TW were processed atdifferent temperatures (121 �C, 100 �C and �20 �C) and theirstability was investigated at different pH (2e8) and salt concen-trations (0 and 20 mM CaCl2).

O

HO

O

HO

O

CH OH

HO

O

OH

O

HO HC

OH

HO

O

O

O

OOOOO

OH

OOOO

OH

OOOO

OOOOO

O

O

Fig. 1. Chemical structures of (a) sucrose monosterate (SE) and (b) Tween 60 or ethoxylated sorbitan monostearate (TW).

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367 359

2. Material and methods

2.1. Materials

Ryoto sucrose esters (S1670) were supplied by Mitsubishi-Kagaku Foods Corporation (Tokyo, Japan). Ryoto S1670 containedmainly of 50e53% sucrose monostearate with sucrose monop-almitate 18e20%, sucrose disterate 12e14 %, sucrose dipalmitate5e6%, sucrose alkylate 5e10% and other ash and moisture. Tween60 (Polyoxyethylene sorbitan monostearate), hydrochloric acid,sodium hydroxide, sodium azide (NaN3), calcium chloride (CaCl2),and Oil Red O were purchased from Sigma Chemical Company (St.Louis, MO, USA). Coconut oil and soybean oil were supplied byTropicana Oil (Nakhonpathom, Thailand) and Thai Vegetable OilPublic Company Limited (Nakhonpathom, Thailand), respectively.Ground coconut meat was purchased locally. Distilled water wasused for emulsion preparation throughout experiment.

2.2. Interfacial tension measurement

Coconut oilewater interfacial tensions were measured accord-ing to Du Nouy ring method at 25 �C using a digital tensiometer(K10, Kruss Scientific, Hamburg, Germany). A series of SE and TWaqueous solutions at different concentrations were prepared andkept in plastic bottles prior measurements. The SE solutions had tobe preheated to 70 �C due to its difficulty to dissolve in water.Before each measurement, the solution was poured intoa measurement vessel and coconut oil was overlaid on top of theaqueous layer. The ring was cleaned with ethanol and distilledwater and flamed with a Bunsen burner before each measurement.Each measured point was an average from at least two replicates.

2.3. Coconut milk emulsion preparation

Coconut milk was produced by adding distilled water to groundcoconut meat at a weight ratio of 2:1. The mixture was manuallypressed and filtered through filter cloth to remove solid residues.The fat contents of the obtained coconut milk were determined byusing Rose-Gottlieb method (AOAC, 2000). The extracted coconutmilk was further diluted with either distilled water or surfactantaqueous solutions to a final fat content of 5 wt% and a finalsurfactant concentration of 0 or 0.25 wt%. The surfactantconcentration of 0.25 wt% was chosen because this concentrationhad been proven to provide sufficient stability to coconut milkemulsions (Tangsuphoom & Coupland, 2009a). Sodium azide0.02 wt% was added to prevent microbial growth. Each emulsion

was further homogenized by using a high speed homogenizer(Ultra-turrax, IKA Labortechnik, Germany) for 3 min at the speed of11,200 rpm.

2.4. Measurement of emulsion properties

2.4.1. Droplet size measurementThe particle size distributions of emulsions were measured by

using a laser diffraction particle size analyzer (Mastersizer 2000;Malvern Instruments Ltd., Worcestershire, UK), with a dual-wavelength detection system. Emulsion samples were droppedand diluted in the test chamber that filled with distilled water inorder to prevent multiple light scattering effects. The size distri-butions of emulsions were obtained via a best fit using Mie theory.The refractive indices of 1.33 for water and 1.15 for coconut oil wereemployed as optical properties of emulsions. The particle sizeswere reported as surface-volume average diameters,d32 ¼ P

nid3i =P

nid2i , and volume-weighted average diameters,d43 ¼ P

nid4i =P

nid3i , where di is the midpoint of the size interval iand ni is the number of particles in that interval.

2.4.2. z-potential measurementThe electrical charges on emulsion droplets were determined by

using a particle electrophoresis instrument (Zeta-Meter System3.0þ, Zeta-Meter, Inc., Staunton, VA, USA). The emulsion sampleafter 1/100 dilution was put into an electrophoresis cell that hadelectrodes at both ends. The direction and velocity of emulsiondroplets that move in the applied electrical field were observedunder a microscope and the charge or z-potential on emulsiondroplets was calculated. Each reported z-potential was an averagefrom five replicates.

2.4.3. Optical microscopyThe emulsion microstructure was examined by using a standard

optical microscope (Alphaphot-2, YS2-H, Nikon Corporation, Japan)equipped with a microscope eyepiece camera (AM423 Dino-EyeUSB, AnMo Electronics Corporation, Taipei, Taiwan) with a soft-ware (DinoCapture Software) installed on a computer. A few dropsof emulsionwere put on a glass slide and fully covered with a coverslid. The microscope magnification of 100� was employed.

2.4.4. Visual appearance and creaming index measurementThe emulsion sample (10 g) was transferred into a standard test

tube and tightly sealed with a cap. The appearance of the emulsionsamplewas closely captured by a digital camera. After gently mixedby inversion, the tubewas left on a tube stand for 10min. The height

Fig. 2. Coconut oilewater interfacial tensions vs. logarithm of concentration (log C)isotherms for (D) sucrose ester (SE) and (B) Tween 60 (TW).

Fig. 3. Visual appearance of phase separation of coconut milk emulsions without addition oof 0.25 wt% sucrose ester (SE) after thermal treatments at different temperatures. (a) Whit

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367360

of aqueous layer left after the emulsion droplets creamed to the topand the height of initial total sample in the tubeweremeasured. Thecreaming index was determined from the percentage of the heightof the aqueous layer over the height of the total sample.

2.5. Coconut milk emulsion stability

2.5.1. Emulsion stability against pH and salt concentrationThe emulsion sample (10 g) was transferred into a standard test

tube and adjusted to the specified pH values (2, 4, 6, and 8) by using0.1 and 1 M HCl or 0.1 and 1 M NaOH. Noted that the normal pH ofextracted coconut milk was w6 before the adjustment. For the saltexperiment, CaCl2 salt was also added to a concentration of 20 mM.The charges on the emulsion droplets were determined fromz-potential measurement and the stability of emulsion wasanalyzed by z-potential measurement, creaming index measure-ment, and optical micrograph.

2.5.2. Emulsion thermal stabilityThe emulsion samples (10 g each) were transferred into stan-

dard test tubes and kept in an autoclave at 121 �C for 10 min, or ina boiling water bath at 100 �C for 20 min, or in a freezer at �20 �C

f any surfactant (Control), with addition of 0.25 wt% Tween 60 (TW), and with additione creamy emulsion. (b) Coagulated solid particles. (c) Destabilized oil layer (free oil).

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367 361

for 14 h. After removal from each storage conditions, the emulsionswere stored to reach room temperature prior further analysis. Thestability of emulsion was analyzed from droplet size measurement,visual appearance, and optical micrograph.

In addition, freeze thaw stability of emulsions was assessed bymeasuring amount of free oil (destabilized oil) after three freeze-thaw cycles (each cycle: �20 �C for 22 h and 30 �C for 2 h) usingthe principle of dye dilution method (Palanuwech, Potineni,Roberts, & Coupland, 2003; Thanasukarn, Pongsawatmanit, &McClements, 2004b). The employed Oil-Red O is colored oil thatonly dissolves in free oil but does not dissolve in emulsified oildroplets. A stock solution of 0.0015 wt% Oil-Red O in soybean oilwas initially prepared. A series of coconut oil solutions were ob-tained by adding different amounts of coconut oil to the stock Oil-Red O solution. The absorbance of the solution series was measuredby using an ultravioletevisible spectrophotometer (SpectronicGenesys 5, Milton Roy company, Rochester, NY, USA) at a wave-length of at 520 nm. The plot between the measured absorbanceand the coconut oil concentrations was used as a calibration curve.To determine amount of free oil in emulsions, the emulsion sample(35 g) was added with 3 g of the stock Oil-Red O solution, vortexedfor 30 s, and centrifuged at the speed of 5500 rpm for 5 min(Universal centrifuge, Model PLC-012, Gemmy Industrial Corpora-tion, Taipei, Taiwan). The colored oil on the upper layer wastransferred by a micro-pipette into a plastic cuvette and measuredits absorbance. The measured absorbance was converted to weightpercent of free oil by using the prepared calibration curve.

Fig. 4. Optical micrographs of coconut milk emulsions without addition of any surfactant (Coester (SE) after thermal treatments at different temperatures. (a) Extensively coalesced oil

3. Results and discussion

3.1. Coconut oilewater interfacial behavior

The basic information about the interfacial properties of SE andTW at coconut oilewater interface was obtained from the interfa-cial tension measurement. Fig. 2 is the plot between the measuredinterfacial tensions of SE and TW aqueous solutions at variousconcentrations. The gradually decrease in the measured interfacialtension was observed as surfactant concentrations increased. Afterthe interfacial tensions reduced to a minimum value, it approacheda constant as surfactants started forming micellar aggregates. Theminimum interfacial tensions for SE and TW were <2 mN/m and8 mN/m, respectively. Unfortunately, the minimum interfacialtension for SE was too small to be measured precisely from ourexisting instrument. It had reported that the lowest interfacialtension of SE at the rapeseed oilewater interface was 1.5 mN/m(Soultani, Ognier, Engasser, & Ghoul, 2003). SE tended to havea slightly better ability than TW in reducing the oilewater inter-facial tension. The critical aggregation concentrations (CAC) atthose minimum interfacial tensions in Fig. 2 were approaching to0.05wt% for SE and around 0.11wt % for TW. The lower value of CACindicated more hydrophobicity of SE, most likely due to thiscommercial SE also contained some proportion of diesters. Ourresults agreed with the reported CAC value of 0.05 wt% SE at theliquid paraffinewater interface (Yanke, Shufen, Jinzong, & Qinghui,2004).

ntrol), with addition of 0.25 wt% Tween 60 (TW), and with addition of 0.25 wt% sucrosedrops. (b) Coalesced oil droplets. (c) Coagulated solid particles.

a

b

c

d

e

Fig. 5. Particle size distributions of coconut milk emulsions without addition of any surfactant (Control), with addition of 0.25 wt% Tween 60 (TW), and with addition of 0.25 wt%sucrose ester (SE) after thermal treatments at different temperatures. (a) At room temperature. (b) At 121 �C for 10 min. (c) At 100 �C for 20 min. (d) At �20 �C for 14 h. (e) Allthermal treatment results for coconut milk emulsions with addition of 0.25 wt% sucrose ester (SE). The arrows indicate extra peaks from coagulated solid particles.

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367362

Table 1The average particle sizes (d3,2 and d4,3) of coconut milk emulsions without addition of any surfactant (Control), with addition of 0.25 wt% Tween 60 (TW), and with addition of0.25 wt% sucrose ester (SE) after thermal treatments at different temperatures.

Thermal treatments Control coconut milk TW coconut milk SE coconut milk

Temperature Holding time d3,2 (mm) d4,3 (mm) d3,2 (mm) d4,3 (mm) d3,2 (mm) d4,3 (mm)

28 �C (Room temp.) e 10.001 � 1.228 33.327 � 5.624 5.535 � 0.175 13.785 � 0.482 8.800 � 0.326 25.891 � 1.467121 �C 10 min 22.057 � 3.866 114.155 � 6.522 12.746 � 0.264 125.284 � 5.600 8.340 � 0.541 58.275 � 8.155100 �C 20 min 16.103 � 1.839 72.344 � 16.996 7.336 � 0.298 23.710 � 5.894 7.967 � 0.441 29.078 � 1.698�20 �C 14 h 7.285 � 0.964 42.272 � 3.222 3.533 � 0.112 11.702 � 0.274 7.858 � 0.266 25.628 � 0.366

Means � standard deviation of at least two replicates.

a

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367 363

3.2. Emulsion thermal stability

3.2.1. Coconut milk emulsions at room temperatureCoconut milk emulsions were prepared with and without

addition of SE or TWand kept at room temperature (w28 �C) beforedetermining their characteristics from visual appearance (Fig. 3),optical micrographs (Fig. 4), and particle size measurements(Fig. 5a and Table 1). Fig. 3 shows that all three types of emulsions:Control coconut milk (without addition of any surfactant), TWcoconut milk, and SE coconut milk, appeared as white creamyemulsions. Their average sizes (d3,2) were 10.0 microns, 5.5microns, and 8.8 microns, respectively (Table 1). The results fromthe optical micrographs (Fig. 4) confirmed that the particle sizes ofcoconut milk emulsions containing SE or TW were smaller thanControl coconut milk. More droplet flocculation in SE emulsionswas suggested to be the reason of SE emulsions having largeraverage size than TW emulsions. The bulky headgroup of TWseemed to effectively provide steric barriers that prevented dropletflocculation in TW emulsions. Even though SE had a better abilitythan TW in reducing the interfacial tension between coconut oiland water interface, this was not enough to observe its effect onemulsion droplet size.

3.2.2. Stability of coconut milk emulsions after heatingFurther investigation was assessing thermal stability of coconut

milk emulsions by heating the emulsions at 121 �C for 10 min. Afterthe heating, the coagulatedwhite solid particles were observed in all

% D

esta

biliz

ed o

il (fr

ee o

il)

Number of freeze-thaw cycles

0

20

40

60

80

100

1 2 3

Control coconut milkTW coconut milkSE coconut milk

Fig. 6. Percentage of destabilized oil (free oil) in coconut milk emulsions withoutaddition of any surfactant (Control), with addition of 0.25 wt% Tween 60 (TW), andwith addition of 0.25 wt% sucrose ester (SE) after 1, 2 and 3 freeze-thaw cycles.

three coconut milk emulsions (Fig. 3). The presence of such solidaggregates had also been reported in other studies (Chiewchan,Phungamngoen, & Siriwattanayothin, 2006; Tangsuphoom &Coupland, 2009b). Those particles were denatured proteins whichwere detected as solid substance in the optical micrographs (arrowsin Fig. 4) and as an extra population on the right to the size distri-bution of emulsion population (arrows in Fig. 5b). From the opticalmicrographs, some large coalesced oil drops were observed inControl coconutmilk thatwere significantly larger thanoil droplets in

b

Fig. 7. Emulsion droplet charge (or z-potential) of (�) coconut milk emulsions withoutaddition of any surfactant (Control), (B) with addition of 0.25 wt% Tween 60 (TW),and (D) with addition of 0.25 wt% sucrose ester (SE). (a) After only pH adjustment. (b)After pH adjustment and addition of 20 mM CaCl2 salt.

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367364

either SE and TW emulsions (Fig. 4). Table 1 shows that the averageparticle size (d3,2) of Control coconutmilk increased from10.0 to 22.1microns (orbya factorof2.2)andTWcoconutmilk increased from5.5to 12.7 microns (or by a factor of 2.3). However, the average particlesize of SE emulsion was still around 8e9 microns showing nosignificantly change (Table 1 and Fig. 5e). Noted that we chose thesurface-volume average diameter (d3,2) instead of d4,3 to analyze ourresults of multi-modal distributions here. D3,2 was based on areadistribution and better representing the average size of smallerparticles (with more surface area) which supposed to be emulsiondroplets in this case. D4,3,or volume-weighted average diameter, wascalculated from volume distribution and more representing theaverage size of larger particles (with more volume) which could bethe size of flocculated droplets or solid aggregates.

Similarly, after coconut milk emulsions were kept at thetemperature of 100 �C for 20 min, the average sizes of Controlcoconut milk emulsions increased by a factor of 1.6 and the size ofTWemulsions slightly increased by a factor of 1.3, but SE emulsionsremained the same average size of 8e9 microns (Table 1 and Fig. 5cand e). This was accompanied by the results from the opticalmicrographs (Fig. 4) showing that SE and TW emulsions stillappeared as small oil droplets but Control coconut milks became

Fig. 8. Optical micrographs of coconut milk emulsions without addition of any surfactant (Coester (SE) after pH adjustment (a) Extensive flocculation. (b) Some flocculation.

large coalesced oil drops. No coagulated solid particles floated inTW or SE emulsions as observed in TW and SE coconut milk at121 �C (Fig. 3), suggesting that the degree of protein denaturationwas lesser at this lower temperature (100 �C).

Coconut proteins, a natural coconut milk emulsifier, whendenatured by heat lose their ability to stabilize emulsion droplets.We observed large coalesced oil drops in Control coconut milk afterheated at 121 �C or 100 �C as oil droplets that come in closeproximity undergo extensively coalesce. Lesser degrees of dropletcoalescence in SE and TW emulsions when compared with Controlcoconut milk indicated that heating stability of coconut milkenhanced by addition of SE or TW. We found that SE stabilizedemulsions seemed to bemore heat stable than TWemulsions as theaverage sizes of SE emulsions (d3,2) did not significantly alter afterthe heating at both temperatures (Table 1 and Fig. 5e). The prop-erties of SE had been reported to be lesser sensitive to temperaturevariation when compared to other nonionic surfactants with oxy-ethylene groups (Stubenrauch, 2001).

3.2.3. Stability of coconut milk emulsions after freeze thawingCoconut milk emulsions were kept at freezing temperature of

�20 �C for 14 h, and after being thawed at room temperature, their

ntrol), with addition of 0.25 wt% Tween 60 (TW), and with addition of 0.25 wt% sucrose

Fig. 9. Creaming index after ten-minute storage of (�) coconut milk emulsionswithout addition of any surfactant (Control), (B) with addition of 0.25 wt% Tween 60(TW), and (D) with addition of 0.25 wt% sucrose ester (SE) after adjustment to differentpH conditions.

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367 365

characteristics were examined. From their visual appearance(Fig. 3), some oil layers were accumulated on top of Control and TWcoconut milk emulsions, but in a sharp contrast, no oil layer wasobserved in SE coconut milk. The optical micrographs in Fig. 4 showsome coalesced oil droplets in the Control and TW samples. Theparticles size distributions (Fig. 5d) and average sizes (d3,2) (Table 1)of Control and TW emulsions were significantly altered from theiroriginal emulsions at room temperature. This deviation should bereasoned that some oil droplets in emulsions were already coa-lesced and separated out as oil layers and those oil droplets werenot detected during the size measurement. Differently, SE emulsi-fied droplets appeared the same as their originally preparedemulsions (Fig. 4) and its particle size distribution remained thesame (Fig. 5e) with the average size w8 microns closed to itsoriginal size (Table 1), revealing good emulsion stability.

To ensure our results, numbers of freeze-thaw cycles wereincreased up to three and the amounts of destabilized oils weredetermined after each cycle. Fig. 6 clearly demonstrated that theamounts of destabilized oil coming out from SE emulsions aftereach freezing cycle were much lesser than from Control and TWemulsions. After the first cycle, almost no destabilized oil wasdetected in SE emulsions, while 93% and 68% of destabilized oilswere obtained from Control and TW emulsions, respectively. Afterthe third cycle, the percentage of destabilized oils in SE emulsionswas three times lesser than in Control or TW emulsions. Noted thateven though this dye-dilution method (Palanuwech et al., 2003)might not be very precise to quantify amounts of oil, the resultsshould be accurate enough for comparing relative amounts of oilcoming out from each emulsion.

Many other studies have shown that their emulsions alsodestabilized into oil layer after freeze thawing; for example, coconutmilks prepared from Tween 20 (Tangsuphoom & Coupland, 2009b),palm oil emulsions prepared from Tween 20 (Thanasukarn et al.,2004b; Thanasukarn, Pongsawatmanit, & McClements, 2006), andn-alkanes emulsions prepared from SDS and Tween 20 (Cramp,Docking, Ghosh, & Coupland, 2004). We noticed that those emul-sions that prepared from Tweens were prone to be unstable simi-larly to our TW result. Several researchers aim to understand thedestabilizationmechanism inorder to improve freeze-thawstabilityof emulsions (Cramp et al., 2004; Ghosh & Coupland, 2008). Thus, itis very interesting here to discover that emulsions prepared from SEwere more freeze-thaw stable than TW emulsions.

A number of studies have reported that the presence of sucroseincreased freeze-thaw stability of emulsions (Ghosh, Cramp, &Coupland, 2006; Thanasukarn, Pongsawatmanit, & McClements,2004a; Thanasukarn et al., 2006). Sucrose has been known to havea cryo-protective property coming from sucrose ability to lower thefreezing point of water. Such unfrozen water protects othercompounds dissolving in the aqueous phase from the freeze damage.Theunfrozenwater inemulsionsprotectsoildroplets frompenetrationof growing ice crystal thatwoulddisrupt the interfacialmembrane andcausedroplet coalescence.We found that therewas also a report aboutthe presence of unfrozenwater in the aqueous solution of sugar-basedsurfactant and its possible cryo-protective effect (Ogawa, Asakura, &Osanai, 2009). We suspected that our SE coconut milk containedsome unfrozen water and this was the reason that SE coconut milkemulsions were still stable after the freeze thawing.

3.3. Emulsion stability against pH and salt concentrations

The charge characteristics of coconut protein on oil dropletswere investigated byexposing emulsions to different pH conditions.Fig. 7a shows that the z-potential of Control coconut milk dropletschanged from a positively charge of þ16 mV to a negatively chargeof�28 mVwhen increasing the pH from 2 to 8. From the graph, the

net charge gradually reduced to zero when pH approached to thevalue of 4, corresponding to the PI of coconut protein reported inother literatures (Onsaard, Vittayanont, Srigam, & McClements,2005; Tangsuphoom & Coupland, 2008a). When proteins losetheir net charge at this pH, the Control emulsion droplets becameextensively flocculated (Fig. 8) and creamed rapidly with thecreaming index of 42% after ten-minute storage (Fig. 9). At other pHvalues, the electrostatic repulsion provided by coconut proteins wassufficient to prevent aggregation of droplets and no creaming wasobserved.

The result of pH effect for SE and TW coconut milk was relativelyinvariant to Control coconut milk. The z-potential of SE and TWemulsions decreased from a positively charge at pH 2 to a nega-tively charge at pH 8with a net zero charge at pH close to 4 (Fig. 7a).SE and TW emulsions only creamed at pH 4 with the creamingindex after ten-minute storage of 8% for SE emulsions and of 50% forTW emulsions (Fig. 9). The much lower in creaming index of SEemulsions suggested that SE might also alter rheological propertyof emulsions and beneficial as a creaming stabilizer. The highestcreaming rate was accompanied by the highest flocculation shownin the micrographs at pH 4 in both emulsions (Fig. 8). It was a littlesurprising that, after addition of nonionic surfactants such as SEand TW, the charges on droplets were still dominant and sensitiveto the pH condition since nonionic surfactants would provide stericstability and remove electrostatic effect of proteins. If the proteinson emulsion droplets were completely displaced then the proper-ties of the emulsion should be the same as the replacing surfactants(McClements, 2004).

More investigation on the role of droplet chargeswas byadditionof 20 mM CaCl2 salt. The micrographs from Fig. 10 show that saltaddition induced droplet flocculation at pH of 6 and 8, similarly inControl, TW and SE emulsions. The flocculation was obvious whencomparing the micrograph before (Fig. 8) and after salt addition(Fig.10). The presence of flocculated droplets had been attributed toability of salt to screen charges of coconut proteins (Fig. 7b) and

Fig. 10. Optical micrographs of coconut milk emulsions without addition of any surfactant (Control), with addition of 0.25 wt% Tween 60 (TW), and with addition of 0.25 wt%sucrose ester (SE) after pH adjustment and addition of 20 mM CaCl2 salt. (a) Extensive flocculation.

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367366

reduced electrostatic repulsion between oil droplets. We noticedthat the oil droplets did not flocculate after addition of salt at pH 2where amino acidsweremostly protonated and proteins had highlynet positively charge. It seemed that oil droplets were most repul-sive at this lowpHcondition as also observed in emulsions stabilizedby other protein (Kim, Decker, & McClements, 2004).

The sensitivity to pH and salt confirmed the presence of proteinsat interfacial droplets in SE and TW emulsions. Thus, in the presentcase, proteins on droplets of coconut milk emulsion were notcompletely displaced by the SE or TW content of 0.25 wt%. Wepostulated that SE or TW molecules stayed on oil droplet interfacetogether with coconut protein stabilizing emulsions by both stericand electrostatic repulsion.

Further investigation on interfacial behaviors when coconutproteins coexist with these surfactants is suggested. The interfacialcharacteristics are governed by structure and thickness ofproteinesurfactant films, the interfacial ratio of protein to surfac-tant, and the proteinesurfactant interactions (Rodriguez Patino,Rodriguez Nino, & Carrera Sanchez, 2007). A previous studyshowed that addition of sucrose ester decreased the interfacialtension and the interfacial elasticity of milk protein membrane in

corresponding to the surfactant/protein ratio (Rouimi, Schorsch,Valentini, & Vaslin, 2005). Some literature also evidenced thatTween 60 only partially displaced sodium caseinate (Dalgleish,Srinivasan, & Singh, 1995) and albumin (Seta, Baldino, Gabriele,Lupi, & de Cindio, 2012) from the oilewater interface even at somehigh surfactant/protein ratios.

4. Conclusions

This work clearly showed that sucrose ester (SE) exhibitedinteresting interfacial and emulsifying properties when comparedwith Tween 60 (TW). SE was slightly better in lowering the inter-facial tension between coconut oil and water interface. The stabilityand surface charge characteristics of coconut milk emulsionsemulsified with 0.25 wt% of nonionic SE or TW strongly dependedon pH and salt environments. Thus, it was likely that the addition ofSE or TW did not completely displace proteins on oil droplets toremove their electrostatic interaction. The presence of either SE orTW had a distinct effect on the thermal characteristics of coconutmilk emulsions: coconut milk emulsions emulsified with SE weremore thermally stable than TW emulsions. Especially, we found

S. Ariyaprakai et al. / Food Hydrocolloids 30 (2013) 358e367 367

that SE coconut milk emulsions weremuchmore stable after freezethawing. The complex between coconut protein and SE seemed tobe an emulsifying membrane that could better protect coconutmilk emulsions from the heat and freeze damages.

Acknowledgment

This researchwas fully supported by the Research Grant for NewScholar (MRG53) provided by the Thailand Research Fund andAssumption University and the mentor of this project was Asst.Prof. Pasawadee Pradipasena fromDepartment of Food Technology,Faculty of Science, Chulalongkorn University.

Special thanks toMs.Treuktongjai Saenghiruna andMs. RungchatKladcharoen, undergraduate students from Assumption Universitywho dedicated in some experimental works and toMs. Martha IntanBudiati, a practical training student from Faculty of AgriculturalTechnology, Soegijapranata Catholic University, Semarang, Indonesia,who helped during sample preparation. Also thanks to Caltech Corp.,Ltd. for kindly donating the sucrose ester used in this study.

References

AOAC. (2000). Official methods of analysis of AOAC international (17 ed.).Gaithersburg, MD: Association of Analytical Communities.

Chiewchan, N., Phungamngoen, C., & Siriwattanayothin, S. (2006). Effect ofhomogenizing pressure and sterilizing condition on quality of canned high fatcoconut milk. Journal of Food Engineering, 73(1), 38e44.

Cramp, G. L., Docking, A. M., Ghosh, S., & Coupland, J. N. (2004). On the stability ofoil-in-water emulsions to freezing. Food Hydrocolloids, 18(6), 899e905.

Dalgleish, D. G., Srinivasan, M., & Singh, H. (1995). Surface properties of oil-in-wateremulsion droplets containing casein and Tween 60. Journal of Agricultural andFood Chemistry, 43(9), 2351e2355.

Garti, N. (2001). Food emulsifiers and stabilizers. In N. A. M. Eskin, & D. S. Robinson(Eds.), Food shelf life stability: Chemical, biochemical, and microbiological changes.Boca Raton: CRC Press.

Ghosh, S., & Coupland, J. N. (2008). Factors affecting the freeze-thaw stability ofemulsions. Food Hydrocolloids, 22(1), 105e111.

Ghosh, S., Cramp, G. L., & Coupland, J. N. (2006). Effect of aqueous composition onthe freeze-thaw stability of emulsions. Colloids and Surfaces A: Physicochemicaland Engineering Aspects, 272(1e2), 82e88.

Kim, H. J., Decker, E. A., & McClements, D. J. (2004). Comparison of dropletflocculation in hexadecane oil-in-water emulsions stabilized by Î2-lactoglobulinat pH 3 and 7. Langmuir, 20(14), 5753e5758.

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

Ogawa, S., Asakura, K., & Osanai, S. (2009). 117. Freeze-thawing behavior of theoc-tyl-b-D-glucoside/water system. Cryobiology, 59(3), 402e403.

Onsaard, E., Vittayanont, M., Srigam, S., & McClements, D. J. (2005). Properties andstability of oil-in-water emulsions stabilized by coconut skim milk proteins.Journal of Agricultural and Food Chemistry, 53(14), 5747e5753.

Palanuwech, J., Potineni, R., Roberts, R. F., & Coupland, J. N. (2003). A method todetermine free fat in emulsions. Food Hydrocolloids, 17(1), 55e62.

Rodriguez Patino, J. M., Rodriguez Nino, M. R., & Carrera Sanchez, C. (2007). Phys-ico-chemical properties of surfactant and protein films. Current Opinion inColloid & Interface Science, 12, 187e195.

Rouimi, S., Schorsch, C., Valentini, C. l., & Vaslin, S. (2005). Foam stability andinterfacial properties of milk protein-surfactant systems. Food Hydrocolloids,19(3), 467e478.

Seta, L., Baldino, N., Gabriele, D., Lupi, F. R., & de Cindio, B. (2012). The effect ofsurfactant type on the rheology of ovalbumin layers at the air/water and oil/water interfaces. Food Hydrocolloids, 29(2), 247e257.

Soultani, S., Ognier, S., Engasser, J. M., & Ghoul, M. (2003). Comparative study ofsome surface active properties of fructose esters and commercial sucrose esters.Colloids and Surfaces A-Physicochemical and Engineering Aspects, 227(1e3),35e44.

Stubenrauch, C. (2001). Sugar surfactants e aggregation, interfacial, andadsorption phenomena. Current Opinion in Colloid & Interface Science, 6(2),160e170.

Tangsuphoom, N., & Coupland, J. N. (2008a). Effect of pH and ionic strength on thephysicochemical properties of coconut milk emulsions. Journal of Food Science,73(6), E274eE280.

Tangsuphoom, N., & Coupland, J. N. (2008b). Effect of surface-active stabilizers onthe microstructure and stability of coconut milk emulsions. Food Hydrocolloids,22(7), 1233e1242.

Tangsuphoom, N., & Coupland, J. N. (2009a). Effect of surface-active stabilizers onthe surface properties of coconut milk emulsions. Food Hydrocolloids, 23(7),1801e1809.

Tangsuphoom, N., & Coupland, J. N. (2009b). Effect of thermal treatments on theproperties of coconut milk emulsions prepared with surface-active stabilizers.Food Hydrocolloids, 23(7), 1792e1800.

Thanasukarn, P., Pongsawatmanit, R., & McClements, D. J. (2004a). Impact of fat andwater crystallization on the stability of hydrogenated palm oil-in-wateremulsions stabilized by whey protein isolate. Colloids and Surfaces A-Physico-chemical and Engineering Aspects, 246(1e3), 49e59.

Thanasukarn, P., Pongsawatmanit, R., & McClements, D. J. (2004b). Influence ofemulsifier type on freeze-thaw stability of hydrogenated palm oil-in-wateremulsions. Food Hydrocolloids, 18(6), 1033e1043.

Thanasukarn, P., Pongsawatmanit, R., & McClements, D. J. (2006). Impact of fat andwater crystallization on the stability of hydrogenated palm oil-in-wateremulsions stabilized by a nonionic surfactant. Journal of Agricultural and FoodChemistry, 54(10), 3591e3597.

Yanke, L., Shufen, Z., Jinzong, Y., & Qinghui, W. (2004). Relationship of solubilityparameters to interfacial properties of sucrose esters. Colloids and Surfaces A:Physicochemical and Engineering Aspects, 248(1e3), 127e133.