lable at ScienceDirect

Food Hydrocolloids 25 (2011) 1521e1529

Contents lists avai

Food Hydrocolloids

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

Oil-in-water emulsions stabilized by chitin nanocrystal particles

Maria V. Tzoumaki a, Thomas Moschakis a, Vassilios Kiosseoglou b, Costas G. Biliaderis a,*

aDepartment of Food Science and Technology, School of Agriculture, P.O. Box 235, Aristotle University of Thessaloniki, GR-54124 Thessaloniki, Greeceb Laboratory of Food Chemistry and Technology, Department of Chemistry, Aristotle University of Thessaloniki, GR-54124 Thessaloniki, Greece

a r t i c l e i n f o

Article history:Received 25 November 2010Accepted 8 February 2011

Keywords:Pickering emulsionsOil-in-water emulsionsChitin nanocrystalsRod-like particlesCreamingRheologyMicrostructure

* Corresponding author. Tel./fax: þ30 2310 991797.E-mail address: biliader@agro.auth.gr (C.G. Biliade

0268-005X/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.foodhyd.2011.02.008

a b s t r a c t

The aim of the present study was to investigate the oil-in-water emulsion stabilizing ability of chitinnanocrystals (colloidal rod-like particles) and the factors that may influence the properties of suchsystems. Chitin nanocrystal aqueous dispersions were prepared by acid hydrolysis of crude chitin fromcrab shells and oil-in-water emulsions were generated by homogenizing appropriate quantities ofa chitin nanocrystal stock aqueous dispersion with corn oil, using an ultra-sonic homogenizer. Theresulting emulsions were visually evaluated for their creaming behaviour upon storage. Additionally, thesamples were studied with static light scattering, small deformation oscillatory rheometry and opticalmicroscopy, under different conditions of nanocrystal concentration, ionic strength, pH and temperature.The chitin nanocrystals were proven quite effective in stabilizing o/w emulsions against coalescence,over a period of one month, as evidenced by static light experiments and microscopy, and this could beattributed to the adsorption of the nanocrystals at the oilewater interface. The rheological data providedevidence for network formation in the emulsions with increasing chitin nanocrystal concentration. Sucha gel-like behaviour was attributed to an inter-droplet network structure and the formation of a chitinnanocrystal network in the continuous phase. The stability of the emulsions to creaming increased withan increase in nanocrystal concentration. Finally, by raising the temperature (20e74 �C), NaCl concen-tration (up to 200 mM) or pH (from 3.0 to 6.7) there was an enhancement of the emulsion elasticcharacter and creaming stability.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Emulsions are of great practical interest because of theirextensive occurrence in food, cosmetics and pharmaceuticalindustries (Vignati, Piazza, & Lockhart, 2003). Emulsion preparationand stabilization can conventionally be achieved by prolongedmechanical agitation and addition of surfactants or other surfaceactive polymers (Ashby & Binks, 2000; Vignati et al., 2003). Earlyworks have shown that emulsion stability does not necessarilyrequire amphiphilic surfactants to reduce the interfacial tension,but can also be efficiently achieved by dispersed solid particles ofcolloidal dimensions (Pickering, 1907; Ramsden, 1904). The emul-sions stabilized by such particles are usually referred as Pickeringemulsions.

The conventional explanation for emulsion stabilization by solidparticles is their accumulation at the oilewater interface in theform of a densely packed layer, which may prevent droplet floc-culation and coalescence by a steric mechanism (Binks & Horozov,

ris).

All rights reserved.

2006; Dickinson, 2006, 2010). The extent of steric barrier dependson how difficult it is to remove particles from the interface, and isgreater when most of the particles’ surfaces lay on the outer side ofthe oil droplets (Dickinson, 2010). Therefore, the contact angle q

made by the stabilizing colloidal particles at the water�oil contactline, determines the particle location at the interface and the natureof the emulsion. Contact angles less that 90�, imply the hydrophilicnature of the colloidal particles that gives rise to o/w emulsions,whereas contact angles greater than 90� imply the hydrophobicnature of the particles that favours w/o emulsion formation (Lagaly,Reese, & Abend, 1999; Vignati et al., 2003). Moreover, the energy ofdesorption per particle is related to the contact angle q and as longas q is not close to 0� or 180�, it is predicted to be of the order ofseveral thousand kT (Aveyard, Binks, & Clint, 2003; Dickinson,2010; Yusoff & Murray, 2011). This denotes that once the particlesare at the interface, they are effectively and irreversibly adsorbed.Hence, one of the most striking features of the particle-stabilizedemulsions is that, in most of the cases, they are extremely stable tocoalescence even when the droplets are quite large (Binks, Dyab, &Fletcher, 2007; Dickinson, 2010).

In addition to the particle layer formation around the droplets,some other ordering mechanisms responsible for the prevention of

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e15291522

droplet coalescence have been proposed for Pickering emulsions.One of these is aggregation of the particles, where the stericparticle-based barrier is not a simple bilayer or monolayer whichis densely packed, but a network of particles, adsorbed at theoilewater interface, with the whole aggregated structure heldtogether by attractive inter-particle forces (Denkov, & Lips, 2008;Gautier et al., 2007; Tcholakova). The properties of Pickeringemulsions are usually determined by particle size (Binks &Lumsdon, 2001), particle wettability (Binks & Clint, 2002),particle concentration (Binks, Philip, & Rodrigues, 2005), oil/waterratio (Binks & Lumsdon, 2001; Binks & Whitby, 2004), pH(Midmore, 1998), salt concentration (Whitby, Fornasiero, & Ralston,2009), and solvent type (Tambe & Sharma, 1994a).

A variety of particles of different size, shape, and surfacechemistry have been used to stabilize emulsions including hydro-phobic silicas (Binks & Lumsdon, 2001; Binks & Whitby, 2004;Midmore, 1998), clays (Ashby & Binks, 2000), carbon nanotubes(Shen & Resasco, 2009), and latexes (Ashby, Binks, & Paunov, 2004;Binks & Lumsdon, 2001). Although there is a lot of research con-ducted on particle-stabilized emulsions, fairly few of them aredirectly related to foods (Dickinson, 2010). This is probably due tothe different type of particles that are used, or the modificationsapplied to them, which are not allowed in foods. On the other hand,food systems frequently include particle-type material that mayplay a role to the stabilization of emulsions (Dickinson, 2010). Someexamples of Pickering-type food emulsions are homogenized andreconstituted milks (o/w emulsions stabilized by casein micelles),margarines and fat spreads (i.e. w/o emulsions stabilized bytriglyceride crystals) (Dickinson, 2006; Hunter, Pugh, Franks, &Jameson, 2008) Moreover, some potentially food-compatibleparticulate materials that have been used in emulsion stabilizationinclude plant or bacterial cellulose fibres (Andresen & Stenius,2007; Paunov et al., 2007; Wege, Kim, Paunov, Zhong, & Velev,2008), and lately modified starch granules (Yusoff & Murray, 2011).

Natural biopolymers, like polysaccharides, can be an attractivesource of particulate material for potential food use. One example ischitin, which is a structural biopolymer found in shellfish, insects,and microorganisms and is the second most abundant poly-saccharide in nature. It has been previously reported that acid-hydrolyzed chitin preparations spontaneously disperse into rod-likecrystalline particles of nanodimensions (Belamie, Davidson, &Giraud-Guille, 2004; Marchessault, Morehead, & Walter, 1959;Revol & Marchessault, 1993). These chitin nanocrystals possesspositive charges at their surface due to protonation of the aminogroups (Li, Revol, Naranjo, & Marchessault, 1996). It has also beenshown that such colloidal dispersions of acid-degraded chitin canundergo an isotropiceanisotropic nematic transition when theirconcentration is increased (Marchessault et al., 1959; Revol &Marchessault, 1993; Belamie et al., 2004). Regarding chitin nano-crystal (ChN) aqueous dispersions, it has been reported in ourprevious works, that they shift towards a nematic gel-like behaviourwith increasing solid particle concentration, ionic strength, pH andtemperature (Tzoumaki, Moschakis, & Biliaderis, 2010), as well as byaddingwhey proteins (Tzoumaki,Moschakis, & Biliaderis, 2011). TheChN dimensionswere previouslymeasured by TEM (Tzoumaki et al.,2010) and found to be on average 240 nm in length and 18 nm indiameter. There are very few studies on Pickering emulsions stabi-lized by particles with different shape other that spherical. Someexamples include laponite platelets-stabilized emulsions (Ashby &Binks, 2000), and calcium carbonate needle-like particles (Paunovet al., 2007). Moreover, it has been found that the particle shapehas a strong impact on the stability and properties of the emulsions(Madivala, Vandebril, Fransaer, & Vermant, 2009); i.e., emulsionsthat cannot be stabilized by spherical particles yield very stableemulsions when particles of the same surface chemistry and size

range, but with a sufficiently large aspect ratio, are used, even at lowvolume fractions. Additionally, surface shear rheology has been usedto demonstrate that the shape anisotropy leads to monolayers withpronounced viscoelastic properties (Madivala et al., 2009).

The aim of the present work was to investigate the preparationand characterization of corn oil-in-water (o/w) emulsions stabi-lized by chitin nanocrystals and the factors that may influence theproperties of such systems, like chitin nanocrystal concentration,ionic environment, pH and temperature. The emulsion character-izationwas carried out by visual observations, static light scatteringexperiments, rheometry and optical microscopy in an attempt tounderstand the underlying stabilization mechanism(s).

2. Materials and methods

2.1. Materials

Crude chitin from crab shells was obtained from Sigma Chem-icals (St Louis, MO). Hydrochloric acid (concentrated 37% v/v),sodium acetate, glacial acetic acid, potassium hydroxide, sodiumchlorite and sodium chloride were all of reagent grade andpurchased from Sigma Chemicals (St Louis, MO). Corn oil wasobtained from a local supermarket and used without furtherpurification. Double distilled water was used in all the experiments.

2.2. Chitin nanocrystals preparation

Aqueous stock dispersions of chitin nanocrystals (ChN) wereprepared by acid hydrolysis (3 N HCl, 95 �C, 90 min) of the originalrawmaterial of crude chitin fromcrab shells. Detailed information onbleaching (with sodium chlorite) and acid hydrolysis are given else-where (Tzoumaki et al., 2010). The solid chitin content of the stockdispersion was determined gravimetrically by drying the samples at50 �C until a constantweightwas obtained; the total solids content ofthe stock dispersionwas approximately 2.7%w/w. The final pH of thestock ChN dispersion was adjusted to 3.0 with 1 N HCl.

2.3. Emulsion preparation

An oil-in-water emulsion was prepared by mixing appropriatequantities of ChN stock dispersion, corn oil and an aqueous solutionadjusted to pH 3.0, using an ultra-sonic homogenizer (Sonics &Materials, Inc. Danbury, Connecticut, USA, Power 50/60 Hz) for2 min with 20 s intervals, to avoid overheating of the samples. Thecorn oil concentration was always adjusted to 10% w/w, while theChN concentration varied; all the concentrations refer to the wholeemulsion and not just to the aqueous phase. A series of emulsionswere prepared by adjusting them to different pH and ionic strengthvalues using HCl (1 N) and NaCl solutions, respectively.

2.4. Visual assessment

Freshly prepared emulsions were poured into 5 ml glass tubes(height 75 mm, diameter 9 mm) after preparation. The tubes werethen inverted carefully several times to ensure thorough mixing.Subsequently, the tubes were sealed to prevent evaporation. Theemulsion samples were stored quiescently at ambient temperatureand the movement of any creaming boundaries was followed withtime.

2.5. Microscopy

Optical micrographs of the ChN-stabilized o/w emulsions werecaptured by an Olympus BX 51 optical microscope fitted witha digital camera (Olympus, DP 50). The emulsion samples were

Fig. 1. Effect of ChN concentration on the macroscopic phase separation of the o/wemulsions after 24 h of storage at room temperature.

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e1529 1523

slightly diluted with a stock HCl solution of pH 3.0, and the dilutedemulsion samples were placed directly in a cavity microscope slideand covered with a coverslip. Additionally, polarized opticalmicroscopy was employed for viewing samples prepared by mildhand shaking of corn oil and a 2.7% w/w ChN stock dispersion.

2.6. Emulsion droplet-size analysis

Droplet-size distributions of the emulsions were determined byusing a Mastersizer 2000 (Malvern Instruments, Malvern, UK).Emulsion droplets were characterized under high dilution condi-tions bydispersing the samples in a hydrochloric acid solution-filledtank at pH 3.0. The refractive indices of water and corn oil weretaken as 1.330 and 1.47, respectively, and the Mie theory was usedfor the analysis. Averagedroplet sizeswere characterized in terms ofthe volume mean diameter d43 ¼ P

i$ni$d4i =

P

i$ni$d3i , where ni is

the number of droplets of diameter di. The d43 parameter is a usefulmean diameter value and it ismore sensitive to the presence of largedroplets. All measurements were made at ambient temperature onat least three separately prepared samples.

2.7. Rheological measurements

Rheological measurements of the samples were performed bya rotational Physica MCR 300 rheometer (Physica MesstechnikGmbH, Stuttgart, Germany) using a double-gap geometry (internaland external gap 0.42 and 0.47 mm, respectively) in a controlledshear-stress mode. The temperature was regulated by a PaarPhysica circulating bath and a controlled peltier system (TEZ 150P/MCR) with an accuracy of �0.1 �C. To minimise dehydration, thesamples were covered with a thin layer of light silicone oil. Thefreshly prepared emulsion samples were initially hand shakenbefore placing them in the rheometer. All the rheologicalmeasurements were completed before any visual phase separationin the emulsions took place.

The linear viscoelastic region (LVR) was assessed at 1 Hz byamplitude sweep experiments; for all the chitin nanocrystaldispersions a constant deformation of g ¼ 0.001 was used, whichwas within the linear viscoelastic region of all the samples. Smalldeformation oscillatory measurements for evaluation of theviscoelastic properties, G0 (storage modulus), G00 (loss modulus),and tand (G00’/G0), were performed over the frequency range of0.01e10 Hz at 20 �C.

In order to explore the behaviour of the ChN-stabilized o/wemulsions upon heating, samples were heated from 20 �C to 74 �Cat a scan rate of 3 �C/min and then cooled back to 20 �C at the samescan rate. All the experiments mentioned above were carried out ata constant frequency of 1 Hz.

Amplitude sweeps were performed at a frequency of 1 Hz and at20 �C, in the range of 0.01e10 Pa of applied shear stress. Deviationfrom the linear viscoelastic region occurs when the sample starts topermanently deform, implying the destruction of the transientnetwork structure. The yield stress was calculated as the stressamplitude at which G0 starts decreasing by 10% from the LVR wherethe G0 remains constant.

Flow curves were obtained at 20 �C and the controlled stresswas varied in the range of 0.001e10 Pa.

2.8. Surface tension measurements

The surface tension of ChN dispersions of various concentrationswas measured by using the Du Nouy ring method via a Krusstensiometer (Hamburg, Germany) at 25 �C. A platinum ring wasused, rinsed with distilled water and if necessary with acetone and

then passed through a Bunsen flame and left to cool before eachmeasurement. The samples were placed in the apparatus and leftovernight before the measurements were carried out.

3. Results and discussion

3.1. Emulsion stability to creaming

The effect of different concentrations of ChN on the visualappearance of ChN-stabilized o/w emulsions, stored quiescently atambient temperature for 24 h, is illustrated in Fig. 1. As can be seen,the % serum release decreased with increasing ChN concentrationand no creaming was observed for the emulsion containing 1.0%w/w ChN. It is worth mentioning that the extent of creaming after 6months of storage was exactly the same as after 24 h of storage,indicating outstanding stability.

3.2. Surface tension of the ChN dispersions

In order to further understand the ability of ChN to stabilize theoil droplets, the surface tension of ChN aqueous dispersions ofvarious concentrations wasmeasured. Table 1 shows the values of gfor ChN dispersions at five particle concentrations, from 0.01 to 1%w/w. After 24 h of ageing, a g value of 55 mN/m was measured forthe 0.01% w/w sample and the surface tension was furtherdecreased, with increasing ChN concentration. These values areclose to some proteins that usually stabilize emulsions and indicatethat ChN are indeed surface active nanoparticles. Similar g resultshave been found for aqueous dispersions of modified starch gran-ules and silica particles (Binks & Lumsdon, 2000b; Yusoff & Murray,2011). The low surface tension values obtained might be alsoattributed to impurities, small surface active molecules or proteins,even though during the ChN preparation process (deproteinizationand acid hydrolysis) most of these substances are likely removed;however, this remains a possibility which would require furtherexperimentation. On the other hand, during the surface tensionmeasurements it was noticed that the surface tension was a littlebit less than that of water after 1 h ageing, and only at long times ofparticle adsorption, the surface tension reached a plateau value,suggesting reorganization of the ChN particles at the surface andconcomitantly a better surface packing.

3.3. Droplet-size distributions of the emulsions

Fig. 2 shows the volume mean diameter (d43) of the oil dropletsin the emulsions with different ChN particle concentration,measured at the day of their preparation and after 30 days of

Table 1Surface tension (g) of ChN dispersions (aireaqueous dispersion of ChN) (mN/m) ofdifferent concentrations at 25 �C (n ¼ 3); the samples were left in the apparatusovernight before measurements were taken.

ChN concentration (% w/w) Surface tension (g) (mN/m)

0.01 55 � 10.05 52 � 20.1 49 � 10.5 44 � 11.0 44 � 2

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e15291524

storage at ambient conditions. The size of the emulsion dropletswas around 100 mm and decreased with the increase in ChNconcentration. At ChN concentrations >0.05% w/w the volumemean diameter was approximately 10 mm. The decrease in dropletsize with the increase in ChN concentration probably occursbecause more ChN particles are available to stabilize smaller oildroplets. In the low ChN concentration (0.01% w/w) bridging floc-culation and/or coalescence may take place, while when the ChNparticles are in excess of that required for complete coverage, moreparticles may adsorb to form additional layers or remain dispersedin the continuous phase (Binks & Lumsdon, 2000a; Binks &Whitby,2004; Midmore, 1999). Moreover, it was noticed that the ChN-stabilized o/w emulsions have relatively large droplet sizescompared to those produced by low Mw surfactants and proteins,which is mainly due to slower adsorption kinetics. At even higherChN concentrations, a slight increase in droplet size was observed,and this could be attributed to the higher viscosity of the contin-uous phase that may hinder the ChN movement during emulsifi-cation and their final deposition at the oilewater interface. Similarfindings have been observed in other studies for the average dropdiameter in emulsions stabilized by silica nanoparticles (Binks &Whitby, 2004), laponite clay (Ashby & Binks, 2000) and otherparticles (Abend & Lagaly, 2001).

All the samples did not show any change of d43 values overa period of one month, which implies that limited coalescence ofthe oil droplets occurred throughout the entire storage period. It isworth pointing out, that even the emulsions having the largestdroplets (about 100 mm) show complete stability to coalescence,a situation that is not usually noticed in surfactant-stabilizedemulsions. Such outstanding stability, even at this length scales, isprobably one of the most striking features of the ChN-stabilizedemulsions and in general, of particle-stabilized emulsions (Binks &

0

20

40

60

80

100

120

0.01 0.03 0.05 0.10 0.30 0.50 0.70 1.00ChN concentration (% w/w)

d4

,3 μ

m

day 0

day 30

Fig. 2. Mean droplet diameter (d43) of ChN-stabilized o/w emulsions, made of differentChN concentrations, at the day of preparation (day 0) and after 30 days of storage atambient conditions.

Whitby, 2004). This can be attributed to the fact that when particleshave the correct surface energy or contact angle with the interface,then once they attach to the airewater or watereoil interface, theyare effectively and irreversibly adsorbed because the energy ofdesorption per particle is of the order of several thousand kT(Dickinson, 2010; Yusoff & Murray, 2011). Moreover, particlesadsorbed at the oilewater interface can sterically hinder the closeapproach of droplets, thus reducing the extent of coalescence(Binks, Clint, & Whitby, 2005). It has also been shown that, atsufficiently high concentrations of adsorbed particles, a networkformation involving the particles may take place at the oilewaterinterface (Dickinson, Ettelaie, Kostakis, & Murray, 2004; Tambe &Sharma, 1994b, 1995), thereby reducing the extent of filmdrainage between droplets and increasing stability to coalescence.The interfacial elasticity, in particular, increases dramatically withincreasing particle concentration. Therefore, in future studies, itwould be useful to examine the viscoelastic properties of theoilewater interface of these systems in relation to emulsionstability. The absence of oil droplet coalescence with time alsocoincides with the creaming results (Fig. 1), where no separate oilphase was detected even after six months of storage at ambientconditions.

3.4. Microstructure of the emulsions

Freshly made o/w emulsions stabilized by ChNwere first dilutedby approximately 100 times, and then observed under a lightmicroscope. Fig. 3 shows typical optical micrographs, combinedwith droplet-size distributions of the emulsions. As evidenced fromthe micrographs, the ChN concentration has a distinct influence onthe size of the emulsified oil droplets. The samples with low ChNconcentration (0.01 and 0.03% w/w) seem to have larger dropletsizes compared to those at higher particle concentrations(>0.05% w/w). These observations are in good agreement with theresults obtained from the droplet-size distribution determined bystatic light scattering experiments (Fig. 2 and background of Fig. 3).Moreover, the micrographs obtained after one month, were similarwith those obtained at the day of their preparation, indicating againthat no coalescence took place (data not shown).

From all the above, it is obvious that the ChN have the ability tostabilize o/w emulsions and it can be assumed that one of themechanisms responsible for emulsion stabilization is the adsorp-tion of ChN at the surface of oil droplets. Fig. 4a shows amicrographof an emulsion with low ChN concentration (0.01% w/w), where itcan be seen that a layer of ChN particles is present at the dropletsinterface. Additionally, Fig. 4b shows a polarized light opticalmicrograph of a hand made emulsion containing 2.7% w/w ChN.Intense birefringent regions of ChN can be noticed at the oil dropletinterface, implying that ChN aggregates are being involved indroplet stabilization.

3.5. Rheological characterization

3.5.1. Effect of ChN concentrationThe long term physical stability of emulsions was also assessed

rheologically. The effect of the ChN concentration on the rheolog-ical properties of the emulsions was primarily investigated byshear-stress controlled rheometry. Fig. 5a displays the flow curvesof ChN-stabilized o/w emulsions. It can be seen that the emulsionsexhibit pseudoplastic behaviour; i.e., the curves show moderate topronounced shear-thinning behaviour, and high viscosities at lowshear rates. This behaviour is typical of weak associative interac-tions and suggests the formation of a weak droplet networkstructure. Additionally, the samples exhibit a small increase inviscosity at low shear rates, until they reach the steady viscosity

Fig. 4. a) Optical micrograph of a 0.01% w/w ChN o/w emulsion (scale bar 50 mm); b)Optical polarized micrograph of a 2.7% ChN dispersion with a corn oil droplet (scale bar500 mm).

ChN concentration (% w/w)

Shear rate (s-1

)

Vis

co

sit

y (

Pa

.s)

0.01

0.1

1

10

100

0.0001 0.01 1 100 10000Shear rate (s

-1

)

Vis

co

sit

y (

Pa

.s)

0.1

1

10

100

0 1 2 3

Zero

sh

ear v

isco

sit

y

(P

a.s

) ChN-stabilized emulsions

ChN aqueous dispersions

0.01

0.1

1

10

100

0.0001 0.01 1 100 10000

0.1% w/w

0.3% w/w

0.5% w/w

0.7% w/w

1.0% w/w

2.7% w/w1.3% w/w0.7% w/w0.3% w/w

a

b

Fig. 5. Shear-rate dependence of viscosity for a) ChN-stabilized o/w emulsions withdifferent ChN concentrations and b) aqueous ChN dispersions with varying particleconcentrations (20 �C). Inset: calculated zero shear viscosities for the ChN-stabilized o/w emulsions and the ChN aqueous dispersions.

Fig. 3. Typical optical micrographs combined with their respective oil droplet-sizedistributions obtained by static light scattering experiments of fresh ChN-stabilized o/w emulsions with different ChN concentration, after dilution of 1:100 (scale bar of50 mm in all three frames).

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e1529 1525

region, which might be due to the network formation fromimmediate rearrangement of the ChN that takes place right aftersevere “stirring” and placing the sample in the apparatus. More-over, the viscosity of the samples at low shear rates, increase withthe increase of ChN concentration. The ChN dispersions alone alsoexhibit a shear-thinning behaviour (Fig. 5b), which are pronouncedat ChN concentrations higher than 2.4% w/w, but for the ChNdispersions with concentrations as low as the ones used for theproduction of the emulsions (<1% w/w), the flow behaviour ismoderate shear-thinning, with relatively low viscosities (around

0.001

0.01

0.1

1

10

100

1000

0.01 0.1 1Shear stress (Pa)

G' (P

a)

0.10% w/w 0.30% w/w

0.50% w/w 0.70% w/w

0.460.170.130.05Yield stress (Pa)

0.70.50.30.1ChN concentration

(% w/w)

Fig. 7. Shear-stress amplitude sweeps (frequency 1 Hz, 20 �C) of ChN-stabilized o/wemulsions with different ChN concentrations (inset: yield stress values).

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e15291526

0.1 Pa s) at low shear rates (viscosities calculated at the plateauregion of the low shear-rate zone, inset Fig. 5).

For an ideal gel which behaves elastically, the G0 value isexpected to be independent of frequency and G0 >> G00. As it can beseen from Fig. 6a, for the 1.0% w/w ChN-stabilized emulsion, G0 wasalways higher than G00 in the frequency range explored, by oneorder of magnitude, and the moduli were almost independent offrequency. However, at a lower concentration of 0.1% w/w, G0 wasslightly higher than G00 and a weak frequency dependence wasobserved. Such a frequency dependence is typical of weak gelbehaviour (Ikeda & Nishinari, 2001). Moreover, with increasing ChNconcentration from 0.1 to 1.0% w/w, the G0 is raised by approxi-mately two orders of magnitude, while the tand decreased(frequency 1 Hz), as shown in Fig. 6b. At this point, it is worthmentioning that aqueous ChN dispersions (at pH 3.0) behaveas weak gels when the particle concentration is slightly above2.4% w/w (Tzoumaki et al., 2010) and not at the low concentrationsemployed in the present experiments for the o/w emulsions(<1.0% w/w).

The shear-stress amplitude sweeps for ChN-stabilized o/wemulsions of different ChN concentrations are shown in Fig. 7. It isobserved that the higher the ChN concentration the longer thelinear viscoelastic region, where the G0 value remains relativelyconstant over a broader range of shear stress applied. The shearstress where a deviation from the linear viscoelastic region occurscould be a rough estimate of the sample’s yield stress. Although theexistence of yield stress has been questioned by many authors(Barnes, 1999), it is a good indication of gel network strength. Itshould be noted that the yield stress value provides an indication ofhow well the emulsion can resist sedimentation or creaming. Theyield stress associated with the emulsion network is thought tocounteract the gravitational forces acting upon the emulsiondroplet in order for the emulsion to remain physically stable fora substantial period of time. A very low yield stress may indicatea tendency to phase separation.

0.1

1

10

100

0 0.3 0.6 0.9 1.2 ChN concentration (% w/w)

G' (P

a)

0.01

0.1

1

10

ta

n δ

G' tan δ

0.01

0.1

1

10

100

0.1 1 10 Frequency (Hz)

G'. G

'' (

Pa

)

G' 0.1% w/w G'' 0.1% w/w G' 1.0% w/w G'' 1.0% w/w

a

b

Fig. 6. a) Mechanical spectra (20 �C, g ¼ 0.001) of ChN-stabilized o/w emulsions; b)ChN concentration dependence of elastic modulus (G0) and tangent value (tand) ofChN- stabilized o/w emulsions (20 �C, frequency 1 Hz) (half bars represent S.D.).

All the above findings imply that an inter-droplet networkformation probably takes place, giving the emulsions a gel-likebehaviour, since the ChN network formation in the continuousphase on its own could not explain the enhanced viscositiesobserved at ChN concentrations lower than 1.0% w/w, as previouslydiscussed. A possible explanation for the droplet network forma-tion could be ChN interactions at the oil�water interfaces. More-over, depletion flocculation events caused by the ChN particles inthe continuous phase could be another reason for inter-dropletnetwork formation; i.e., the ChN particles, having an anisotropicrod-like shape with dimensions 240 nm in length and 18 nm indiameter, are relatively smaller than the spherical oil droplets(100e10 mm), a situation that can lead to osmotically induced weakassociative interactions among the spherical droplets, as describedby many authors (Koenderink et al., 1999; Mao, Cates, &Lekkerkerker, 1995; Mao, Cates, & Lekkerkerker, 1997; Moschakis,Murray, & Dickinson, 2005; Moschakis, Murray, & Dickinson,2006; Moschakis, Murray, & Biliaderis, 2010). The formation ofa reversible relatively weak inter-droplet network in the presentemulsion system can be further supported by microscopy of anundiluted emulsion (0.1% w/w ChN) (Fig. 8), where both the oildroplets-ChN rich regions and the continuous phase are seen.Additionally, with increasing the ChN concentration, apart from theincrease in the inter-droplet attractive forces, an extra networkmaybe formed in the continuous phase. The ChN are mechanicallyconstrained by the droplet network, and thus their concentrationlocally increases, causing an increase in the viscosity of the aqueousphase. The gel-like structure can further improve the emulsionstability by immobilizing the oil droplets in the formed structure.Formation of a network in the continuous phase of emulsions byparticles, which traps droplets and impedes the drainage of liquidfilms between them, thus hindering coalescence, has been postu-lated (Abend, Bonnke, Gutschner, & Lagaly, 1998; Lagaly et al., 1999;Midmore, 1999).

It is also worth mentioning that all the rheological responsesdescribed aboveweremaintained over a period of onemonth. Thus,all the samples returned to their initial form obtained right afterstirring, indicative of reversible changes in the emulsion structure.

3.5.2. Effect of temperatureThe influence of temperature on the behaviour of ChN-stabi-

lized emulsions was explored by applying a heating and coolingprotocol on samples containing different ChN concentrations from20 to 74 �C, as shown in Fig. 9. Heating the emulsions resulted in anincrease in the elastic modulus and upon cooling, the value of G0

remained relatively constant at the level obtained upon heating; i.e.the structural change caused by heating was irreversible. The

Fig. 8. Typical optical micrograph of a fresh-made undiluted o/w emulsion with 0.1%w/w ChN concentration.

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e1529 1527

increase in elastic modulus upon heating became higher withincreasing ChN concentration (Fig. 9). The increase of storagemodulus in the above samples should be related to the ChNnetwork “strengthening”, in the continuous phase, with increasingtemperature. Indeed, in our previous work (Tzoumaki et al., 2011) itwas shown that heating causes an increase in storage modulus ofthe aqueous chitin nanocrystals dispersions, which is irreversibleupon cooling, and the rate of increase is depended on temperature.It was also found, that the ChN dispersions were not stable withtime even at 20 �C, and as the particles tended to reorientate to givea stronger network structure, heating seemed to accelerate thisprocess; i.e. the kinetic energy of the particles is enhanced and thusthey become more mobile to rearrange themselves. Additionally,with increasing temperature, the kinetic energy of the counterionsis increased and the diffuse-double layer thickness is reduced. Asa result, the inter-particle distances are decreased and the

0

4

8

12

16

20

24

28

32

20 30 40 50 60 70 80Temperature (

o

C)

G' (P

a)

1.0% w/w heating 1.0% w/w cooling0.7% w/w heating 0.7% w/w cooling0.5% w/w heating 0.5% w/w cooling

Fig. 9. Effect of heating at 74 �C and cooling back to 20 �C (scan rate 3 �C/min) on theelastic modulus (G0) of ChN-stabilized o/w emulsions with different ChNconcentrations.

associative forces among the nanocrystals are enhanced (Tzoumakiet al., 2010).

3.5.3. Effect of ionic strength and pHThe influence of ionic strength in a 0.5% w/w ChN-stabilized o/w

emulsion properties was examined by small deformation rheometry(NaCl concentrations between 50 and 200 mM). Fig. 10a shows theeffect of NaCl concentration in a 0.5% w/w ChN-stabilized emulsionon its rheological and creaming characteristics (inset). The storagemodulus (G0) increased with the NaCl concentration and the tanddecreased, respectively. Furthermore, salt addition seems to reduceor even inhibit creaming. As the salt concentration increases, there isa greater tendency for the ChN to aggregate since the stabilizingrepulsive forces are weakened by the increase in electrolyteconcentration. Therefore, the ChNnetwork in the continuousphase isstrengthened, resulting in further stabilization of the structure.

In the present work, the ChN are positively charged due toprotonation of the amino groups (Li et al., 1996), and thus someparticles, partitioning at the interfaces, behave as emulsion stabi-lizers due to their surface charge. Ashby and Binks (2000) reportedthat the addition of electrolytes reduces the electrostatic repulsionbetween the negatively charged clay particles, permitting theformation of a denser film around the oil droplets, therebyincreasing their stability to coalescence. Indeed, Midmore (1999)has observed that flocculated silica particles produce more stableemulsions. It has also been found that salt addition furtherimproves the stability of kaolin and bentonite particle-stabilizedemulsions (Torres, Iturbe, Snowden, Chowdhry, & Leharne, 2007).

Similar responses to those obtained by varying the ionicstrength were also noted by changing the pH. As can be seen inFig. 10b, with increasing pH, the 0.5% w/w ChN-stabilized emulsion

Fig. 10. Variation of G0 and tand of ChN-stabilized o/w emulsions (0.5% w/w ChN) asa function of: a) NaCl concentration; b) pH (half bars represent � S.D.). The insetsrepresent the creaming results after one month of storage at ambient conditions.

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e15291528

exhibited a stronger gel-like behaviour; i.e. the G0 value increasedand the tand decreased. Additionally, creaming was reduced in thesample with pH 4.6 compared to pH 3.0, while the other twosamples with pH 5.6 and 6.7, respectively, did not exhibit anycreaming after a storage period of one month.

The pK of ChN dispersions is reported to be around 6.3 (Li et al.,1996), therefore as the pH reaches lower values, e.g. pH 3.0, the ChNamino groups should be totally protonated. The electrostaticrepulsive forces at this pH level and a low ChN concentration(0.5% w/w) would hinder formation of a strong network. Incontrast, when the dispersion approaches the pH value of 6.3, therepulsive electrostatic forces are minimized and the nanocrystalscould then easily aggregate and form a gel; this will further stabi-lize the emulsion towards creaming.

4. Conclusions

In the present work, the stabilizing properties of chitin nano-crystals in o/w emulsions and their dynamics were studied undervarying conditions. The chitin nanocrystals aided towards theproduction of stable to coalescence emulsions for a period of onemonth, even when the droplets were of relatively large size. Theincrease in ChN concentration led to smaller droplets, higherstability to creaming and more pronounced elastic responses. Aspossible mechanisms responsible for the o/w emulsion stabiliza-tion were suggested to be the ChN adsorption at the oilewaterinterfaces, the build up of an inter-droplet network and a ChNnetwork in the continuous phase. Additionally, with an increase inNaCl concentration, pH and temperature, there was furtherenhancement of the emulsion elastic structure.

Acknowledgements

M. Tzoumaki thanks the Greek Ministry of Education, LifelongLearning and Religious Affairs for awarding her a graduate schol-arship, Heraklitus II, for her PhD studies.

References

Abend, S., Bonnke, N., Gutschner, U., & Lagaly, G. (1998). Stabilization of emulsionsby heterocoagulation of clay minerals and layered double hydroxides. Colloidand Polymer Science, 276(8), 730e737.

Abend, S., & Lagaly, G. (2001). Bentonite and double hydroxides as emulsifyingagents. Clay Minerals, 36(4), 557e570.

Andresen, M., & Stenius, P. (2007). Water-in-oil emulsions stabilized by hydro-phobized microfibrillated cellulose. Journal of Dispersion Science and Technology,28(6), 837e844.

Ashby, N. P., & Binks, B. P. (2000). Pickering emulsions stabilised by Laponite clayparticles. Physical Chemistry Chemical Physics, 2(24), 5640e5646.

Ashby, N. P., Binks, B. P., & Paunov, V. N. (2004). Bridging interaction betweena water drop stabilised by solid particles and a planar oil/water interface.Chemical Communications, 4, 436e437.

Aveyard, R., Binks, B. P., & Clint, J. H. (2003). Emulsions stabilised solely by colloidalparticles. Advances in Colloid and Interface Science, 100, 503e546.

Barnes, H. A. (1999). The yield stress e a review or ‘pi alpha nu tau alpha rho epsiloniota’ e everything flows? Journal of Non-Newtonian Fluid Mechanics, 81(1e2),133e178.

Belamie, E., Davidson, P., & Giraud-Guille, M. M. (2004). Structure and chirality ofthe nematic phase in alpha-chitin suspensions. Journal of Physical Chemistry B,108(39), 14991e15000.

Binks, B. P., & Clint, J. H. (2002). Solid wettability from surface energy components:relevance to pickering emulsions. Langmuir, 18(4), 1270e1273.

Binks, B. P., Clint, J. H., & Whitby, C. P. (2005). Rheological behavior of water-in-oilemulsions stabilized by hydrophobic bentonite particles. Langmuir, 21(12),5307e5316.

Binks, B. P., Dyab, A. K. F., & Fletcher, P. D. I. (2007). Contact angles in relation toemulsions stabilised solely by silica nanoparticles including systems containingroom temperature ionic liquids. Physical Chemistry Chemical Physics, 9(48),6391e6397.

Binks, B. P., & Horozov, T. S. (2006). Colloidal particles at liquid interfaces: anintroduction. In B. P. Binks, & T. S. Horozov (Eds.), Colloidal particles at liquidinterfaces (pp. 1e73). Cambridge, UK: Cambridge University Press.

Binks, B. P., & Lumsdon, S. O. (2000a). Catastrophic phase inversion of water-in-oilemulsions stabilized by hydrophobic silica. Langmuir, 16(6), 2539e2547.

Binks, B. P., & Lumsdon, S. O. (2000b). Influence of particle wettability on the typeand stability of surfactant-free emulsions. Langmuir, 16(23), 8622e8631.

Binks, B. P., & Lumsdon, S. O. (2001). Pickering emulsions stabilized by mono-disperse latex particles: effects of particle size. Langmuir, 17(15), 4540e4547.

Binks, B. P., & Whitby, C. P. (2004). Silica particle-stabilized emulsions of silicone oiland water: aspects of emulsification. Langmuir, 20(4), 1130e1137.

Dickinson, E. (2006). Interfacial particles in food emulsions and foams. In B. P. Binks,& T. S. Horozov (Eds.), Colloidal particles at liquid interfaces (pp. 298e327).Cambridge, UK: Cambridge University Press.

Dickinson, E. (2010). Food emulsions and foams: stabilization by particles. CurrentOpinion in Colloid & Interface Science, 15(1e2), 40e49.

Dickinson, E., Ettelaie, R., Kostakis, T., & Murray, B. S. (2004). Factors controlling theformation and stability of air bubbles stabilized by partially hydrophobic silicananoparticles. Langmuir, 20(20), 8517e8525.

Gautier, F., Destribats,M., Perrier-Cornet, R.,Dechezelles, J. F., Giermanska, J.,Heroguez, V.,et al. (2007). Pickering emulsions with stimulable particles: from highly- to weak-ly-covered interfaces. Physical Chemistry Chemical Physics, 9(48), 6455e6462.

Hunter, T. N., Pugh, R. J., Franks, G. V., & Jameson, G. J. (2008). The role of particles instabilising foams and emulsions. Advances in Colloid and Interface Science, 137(2), 57e81.

Ikeda, S., & Nishinari, K. (2001). “Weak gel”-type rheological properties of aqueousdispersions of nonaggregated kappa-carrageenan helices. Journal of Agriculturaland Food Chemistry, 49(9), 4436e4441.

Koenderink, G. H., Vliegenthart, G. A., Kluijtmans, S., van Blaaderen, A.,Philipse, A. P., & Lekkerkerker, H. N. W. (1999). Depletion-induced crystalliza-tion in colloidal rod-sphere mixtures. Langmuir, 15(14), 4693e4696.

Lagaly, G., Reese, M., & Abend, S. (1999). Smectites as colloidal stabilizers ofemulsions e I. Preparation and properties of emulsions with smectites andnonionic surfactants. Applied Clay Science, 14(1e3), 83e103.

Li, J., Revol, J. F., Naranjo, E., & Marchessault, R. H. (1996). Effect of electrostaticinteraction on phase separation behaviour of chitin crystallite suspensions.International Journal of Biological Macromolecules, 18(3), 177e187.

Madivala, B., Vandebril, S., Fransaer, J., & Vermant, J. (2009). Exploiting particleshape in solid stabilized emulsions. Soft Matter, 5(8), 1717e1727.

Mao, Y., Cates, M. E., & Lekkerkerker, H. N. W. (1995). Depletion force in colloidalsystems. Physica A, 222(1e4), 10e24.

Mao, Y., Cates, M. E., & Lekkerkerker, H. N. W. (1997). Theory of the depletion forcedue to rodlike polymers. Journal of Chemical Physics, 106(9), 3721e3729.

Marchessault, R. H., Morehead, F. F., & Walter, N. M. (1959). Liquid crystal systemsfrom fibrillar polysaccharides. Nature, 184(4686), 632e633.

Midmore, B. R. (1998). Preparation of a novel silica-stabilized oil/water emulsion.Colloids and Surfaces A-Physicochemical and Engineering Aspects, 132(2e3),257e265.

Midmore, B. R. (1999). Effect of aqueous phase composition on the properties ofa silica-stabilized W/O emulsion. Journal of Colloid and Interface Science, 213(2),352e359.

Moschakis, T., Murray, B. S., & Biliaderis, C. G. (2010). Modifications in stability andstructure of whey protein-coated o/w emulsions by interacting chitosan andgum arabic mixed dispersions. Food Hydrocolloids, 24(1), 8e17.

Moschakis, T., Murray, B. S., & Dickinson, E. (2005). Microstructural evolution ofviscoelastic emulsions stabilised by sodium caseinate and xanthan gum. Journalof Colloid and Interface Science, 284(2), 714e728.

Moschakis, T., Murray, B. S., & Dickinson, E. (2006). Particle tracking using confocalmicroscopy to probe the microrheology in a phase-separating emulsion con-taining nonadsorbing polysaccharide. Langmuir, 22(10), 4710e4719.

Paunov, V. N., Cayre, O. J., Noble, P. F., Stoyanov, S. D., Velikov, K. P., & Golding, M.(2007). Emulsions stabilised by food colloid particles: role of particle adsorptionand wettability at the liquid interface. Journal of Colloid and Interface Science,312(2), 381e389.

Pickering, S. U. (1907). Emulsions. Journal of Chemical Society, 91, 2001e2021.Ramsden, W. (1904). Separation of solids in the surface-layers of solutions and

suspensions (observations on surface-membranes, bubbles, emulsions, andmechanical coagulation). Proccedings of the Royal Society London, 72, 156e164.

Revol, J. F., & Marchessault, R. H. (1993). In-vitro chiral nematic ordering of chitincrystallites. International Journal of Biological Macromolecules, 15(6), 329e335.

Shen, M., & Resasco, D. E. (2009). Emulsions stabilized by carbon nanotube-silicananohybrids. Langmuir, 25(18), 10843e10851.

Tambe, D. E., & Sharma, M. M. (1994a). The effect of colloidal particles on fluidefluidinterfacial properties and emulsion stability. Advances in Colloid and InterfaceScience, 52, 1e63.

Tambe, D. E., & Sharma, M. M. (1994b). Factors controlling the stability of colloid-stabilized emulsion. 2. A model for rheological properties of colloideladeninterfaces. Journal of Colloid and Interface Science, 162(1), 1e10.

Tambe, D. E., & Sharma, M. M. (1995). Factors controlling the stability of colloid e

stabilized emulsions. 3. Measurement of the rheological properties ofcolloideladen interfaces. Journal of Colloid and Interface Science, 171(2), 456e462.

Tcholakova, S., Denkov, N. D., & Lips, A. (2008). Comparison of solid particles,globular proteins and surfactants as emulsifiers. Physical Chemistry ChemicalPhysics, 10(12), 1608e1627.

Torres, L. G., Iturbe, R., Snowden, M. J., Chowdhry, B. Z., & Leharne, S. A. (2007).Preparation of o/w emulsions stabilized by solid particles and their character-ization by oscillatory rheology. Colloids and Surfaces A-Physicochemical andEngineering Aspects, 302(1e3), 439e448.

M.V. Tzoumaki et al. / Food Hydrocolloids 25 (2011) 1521e1529 1529

Tzoumaki, M. V., Moschakis, T., & Biliaderis, C. G. (2010). Metastability of nematic gelsmadeof aqueous chitinnanocrystal dispersions.Biomacromolecules,11(1),175e181.

Tzoumaki, M. V., Moschakis, T., & Biliaderis, C. G. (2011). Mixed aqueous chitinnanocrystal-whey protein dispersions: microstructure and rheological behav-iour. Food Hydrocolloids, 25(5), 935e942.

Vignati, E., Piazza, R., & Lockhart, T. P. (2003). Pickering emulsions: interfacialtension, colloidal layer morphology, and trapped-particle motion. Langmuir, 19(17), 6650e6656.

Wege, H. A., Kim, S., Paunov, V. N., Zhong, Q. X., & Velev, O. D. (2008). Long-termstabilization of foams and emulsions with in-situ formed microparticles fromhydrophobic cellulose. Langmuir, 24(17), 9245e9253.

Whitby, C. P., Fornasiero, D., & Ralston, J. (2009). Effect of adding anionic surfactanton the stability of Pickering emulsions. Journal of Colloid and Interface Science,329(1), 173e181.

Yusoff, A., & Murray, B. S. (2011). Modified starch granules as particle-stabilizers ofoil-in-water emulsions. Food Hydrocolloids, 25(1), 42e55.