mb,

anjie's Rpind

Phosphoric acidAmorphous celluloseEmulsion stabilization

usue h

increasing the cellulose content to 0.83%, the emulsions remain completely stable for months. All pre-

zed bylymers

hydrophilic as well, but it is insoluble in water because of strongintermolecular hydrogen bonds (McClements, 2005).

Recently, the role of hydrogen bonding in the solubility andinsolubility properties of cellulose has been challenged (Glasseret al., 2012; Medronho, Romano, Miguel, Stigsson, & Lindman,

emulsions if theylloidal microcrys-carboxymethylcel-d water-in oil-ina network aroundfunction of MCC isding a mechanicalum carboxymeth-tive colloid for the

MCC (Oza & Frank, 1986). Recent studies have demonstrated thatcellulose nanocrystals, without any dispersing agent, can alsoeffectively stabilize oil-in-water emulsions (Kalashnikova, Bizot,Cathala, & Capron, 2011, 2012). Moreover, high internal phaseemulsions (HIPEs) with oil volume fractions more than 0.9 can bestabilized by less than 0.1% cellulose nanocrystals (Capron &Cathala, 2013). These nanocrystals are needle- or rod-shapedcrystalline cellulose with widths of a few nanometers and lengthsvarying from tens of nanometers to several micrometers. They are

* Corresponding author. Tel./fax: 86 25 84395671.

Contents lists availab

Food Hydr

els

Food Hydrocolloids 43 (2015) 275e282E-mail addresses: twu@njau.edu.cn, wutao007@hotmail.com (T. Wu).pectin, chitosan, and galactomannans, are capable of stabilizingemulsions, although there is debate regarding the underlyingmechanisms and whether they involve surface activity or thick-ening effects (McClements, 2005). Traditionally, natural starchesand cellulose are not considered as amphiphilic biopolymers withgood emulsion stabilization ability, unless chemical modicationsare made to introduce surface active groups (McClements, 2005;Xhanari, Syverud, & Stenius, 2011). Starches are essentially hydro-philic molecules with poor surface activity, whereas cellulose is

crystalline cellulose is capable of forming stableare dispersed well. Earlier studies have found cotalline cellulose (MCC) (11% MCC 1% sodiumlulose) can stabilize oil-in-water emulsions anwater (w/o/w) multiple emulsions by formingthe emulsied oils (Oza& Frank,1986,1989). Theto orient at the oil-in-water interface thus provibarrier to droplet coalescence, whereas the sodiylcellulose functions as a dispersing and protecpolysaccharides, and nano/micro-particles (Dickinson, 2009). Anumber of naturally occurring polysaccharides, such as gum arabic,

et al., 2012; Parthasarathi et al., 2011; Yamane et al., 2006). As aresult of this amphiphilic character, it is not surprising to see thatShear-thinningGelAttractive emulsion

1. Introduction

Food emulsions are usually stabiliactive emulsiers, amphiphilic biopohttp://dx.doi.org/10.1016/j.foodhyd.2014.05.0240268-005X/ 2014 Elsevier Ltd. All rights reserved.viously gravitationally unstable liquid-like emulsions were transformed into stable gel-like emulsionsand persisted in this form during storage. Optical and uorescence microscopy demonstrated theadsorption of cellulose on the surface of oil droplets. Rheology study indicates that the resultingemulsions are attractive emulsions with typical shear-thinning gel characteristics. The underlyingemulsion stabilization mechanism is a combination of Pickering and network mechanisms. The ndingsof this research explore a more practical way of utilizing non-derivative cellulose in the food industry as anovel food ingredient.

2014 Elsevier Ltd. All rights reserved.

small molecular surface, including proteins and

2012). It is believed that the properties of cellulose are alsosignicantly inuenced by hydrophobic interactions, and the sur-face of crystalline cellulose has both hydrophobic and hydrophilicplanes (Cho, Gross, & Chu, 2011; Glasser et al., 2012; MedronhoKeywords:Available online 9 June 2014obtained non-derivative amorphous cellulose can effectively stabilize oil-in-water emulsions. Freshlyprepared emulsions with cellulose contents of 0.07e0.56% are not stable against creaming. AfterStabilizing oil-in-water emulsion with a

Xuejuan Jia a, Ranran Xu a, Wei Shen b, Mengxia XiePeng Wang a, Xiaoxiong Zeng a, Tao Wu a, *

a College of Food Science and Technology, Nanjing Agricultural University, Weigang 1, Nb College of Science, Nanjing Agricultural University, Weigang 1, Nanjing 210095, Peoplc Department of Food Technology, Pir Mehr Ali Shah Arid Agriculture University, Rawal

a r t i c l e i n f o

Article history:Received 17 November 2013Accepted 29 May 2014

a b s t r a c t

Non-derivative cellulose istion ability. In this study, w

journal homepage: www.orphous cellulose

Muhammad Abid a, c, Saqib Jabbar a,

ng 210095, People's Republic of Chinaepublic of Chinai, Pakistan

ally not considered an amphiphilic biopolymer with emulsion stabiliza-ave demonstrated, using a dissolution and regeneration process, that the

le at ScienceDirect

ocolloids

evier .com/locate/ foodhyd

colloprepared by acid hydrolysis of the native cellulose ber (Habibi,Lucia, & Rojas, 2010). It is believed that the amphiphilic characterof cellulose nanocrystals resides in the crystal organization at theelementary brick level, and their emulsion stabilization mechanismis Pickering stabilization (Kalashnikova, Bizot, Cathala, & Capron,2012). In Pickering stabilization, nano/micro-particles are absor-bed at the oil-water interface and protect emulsion droplets againstocculation and coalescence by a steric barrier (Dickinson, 2013).Both oil-in-water and water-in-oil emulsions can be formeddepending on the particle wettability (Dickinson, 2010).

The supramolecular structure of cellulose has both crystallineand amorphous domains (Habibi et al., 2010). The preparation ofcrystalline cellulose products, such as MCC or cellulose nano-crystals, usually employs hydrochloric or sulfuric acid to hydrolyzethe amorphous domain and release the crystalline domain (Habibiet al., 2010). The hydrolysis of amorphous domain results in a sig-nicant loss of the cellulose biomass, and thus, yields of crystallinecellulose are usually low. For example, the yield of cellulosenanocrystals is as low as 30%, even after optimization of thepreparation method (Bondeson, Mathew, & Oksman, 2006).Although mechanical methods can be used to prepare crystallinecellulose, such as microbrillated or nanobrillated cellulose, suchmethods suffer from extremely high energy consumption (Klemmet al., 2011). Compared with the enormous research on crystallinecellulose for food applications, amorphous cellulose, which can beprepared in high yield by solvent dissolution and anti-solventregeneration, has received little research attention in the food in-dustry. This under appreciation is caused by a lack of a good cel-lulose solvent and little exploration of its functional properties forfood applications. Current solvents are either expensive, such asionic liquids, or organic-solvent based, such as N-methylmorpho-line-N-oxide (NMMO) and N, N-dimethylacetamide (DMAc)/LiCl,and thus, their practical applications in the food industry arelimited. Recently, a new solvent system consisting of NaOH/urea,NaOH/thiourea or LiOH/urea has been developed, but it is onlyeffective in dissolving cellulose with a weight-average molecularweight below 1.2 105 (Luo & Zhang, 2013). Nevertheless, thefunctional properties of the resulting cellulose as bers, mem-branes, microspheres and hydrogels have also been studied butmainly in the areas of packaging, separation science, textile andbiomedicine (Luo & Zhang, 2013). Another study found thatamorphous cellulose prepared by regeneration from ionic liquidscan stabilize both oil-in-water and water-in-oil emulsions (Rein,Khaln, & Cohen, 2012). Other than the aforementioned cited ref-erences, little research can be found in the current literatureregarding the food applications of functional amorphous cellulose.

In our previous works, we have successfully prepared amor-phous cellulose by using a benign solvent e phosphoric acid anddemonstrated that prepared amorphous cellulose has excellentrheological properties (Jia et al., 2013; Jia et al., 2014). The disso-lution of cellulose in phosphoric acid is believed to be an esteri-cation process: celluloseeOH H3PO4 / celluloseeOePO3H2 H2O. The regeneration of cellulose in water results inamorphous cellulose by the reverse reaction: celluloseeOePO3H2 H2O/ celluloseeOH H3PO4 (Zhang, Cui, Lynd, &Kuang, 2006). Compared with other solvents, phosphoric acid isinexpensive and has good solubility for cellulose from differentsources with molecular weights up to 5.5 105 (Zhang & Lynd,2005). In this study, we demonstrated the prepared amorphouscellulose has excellent emulsion stabilization ability as well.Although previous studies have demonstrated the emulsion stabi-lization of amorphous cellulose (Jia et al., 2013; Rein et al., 2012),the characteristics of the resulting emulsions, such as their micro-structure and rheology properties has not been studied to date.

X. Jia et al. / Food Hydro276Because of the sensory perception of food emulsion, such asmouthfeel and texture, essentially rely on these properties, theybecome the top priorities of this investigation.

2. Experimental details

2.1. Materials

MCC powder, 85% phosphoric acid (H3PO4) and other commonchemicals were purchased from Sinopharm Chemical Reagent Co,Ltd. (Shanghai, China). Dodecane (99%) obtained from SigmaAldrich (Shanghai, China) was puried through water extractionextensively. Calcouor White was purchased from SigmaeAldrich(Shanghai, China). Deionized water was used throughout theexperiment.

2.2. Preparation of amorphous cellulose

The phosphoric acid was equilibrated to 4 C in a refrigerator(Haier, China). Then, 2.00 g of MCC was wetted with 6 ml ofdeionized water and mixed with 50 ml of 85% phosphoric acid toreach a homogenous state, followed by the addition of another50 ml of 85% phosphoric acid. The obtained cellulose suspensionswere incubated in a shaking bath with temperature set at 0 C and aspeed of 150 rpm for 24 h to obtain a clear solution. Then, 500 ml ofdeionizedwater was used to dilute the cellulose solution to obtain amilky dispersion, followed by centrifugation at 16,700 g (Beckman,J-30I, USA) for 20 min. The supernatant was discarded, and thepellet was washedwith deionizedwater repeatedly until a constantpH was obtained. The concentration of cellulose in the naldispersion was determined gravimetrically to be 1.38% w/v.

2.3. Emulsions preparation

An amorphous cellulose dispersion (1.38%w/v) was dilutedwithdeionized water to obtain cellulose dispersions at a series of con-centrations, i.e., 1.10, 0.83, 0.55, 0.28, and 0.07% w/v. Four millilitersof puried dodecane was then added into a 50 ml plastic tubecontaining 16 ml of cellulose dispersions and then dispersed by adisperser (IKA T18 homogenizer, Germany) at 10,000 rpm for 3 minin a cold water bath to avoid overheating of the sample.

2.4. Visual inspection of emulsion stability

Ten milliliters of prepared emulsion was immediately trans-ferred to a 10 ml glass tube. The stability of emulsions was visuallychecked after storage at 1, 7 and 90 days at room temperature. Afterpreparation, emulsions with low concentration of cellulose addedwere phase separated. The clear bottom phase is designated asserum and the top phase, which is a concentrated layer of droplets,is designated as emulsion cream. The volumes of the emulsioncream are calculated from their thickness, which were measuredwith a digital caliper. And these volumes are used to calculate thedodecane fraction and nal concentration of cellulose in emulsioncream.

2.5. Optical and uorescence microscopy

The optical and uorescence photographs of o/w emulsionswere captured using an Eclipse microscope equipped with a digitalcamera (Nikon 80i, Japan). Calcouor White was used as uores-cent dye of cellulose at a concentration of 0.1% w/v. A drop of Cal-couor White solution was added on the slide followed by theaddition of a drop of emulsion. The diameters (d32) of the emulsionswere measured using Image J software by taking the average of at

ids 43 (2015) 275e282least 500 droplets. For emulsion with volume of v and internal

colloids 43 (2015) 275e282 277phase volume fraction of , the total interfacial area of droplets iscalculated as 6v/d32.

To image the distribution of amorphous cellulose on the surfaceof emulsion droplets, Styrene was mixed with radical initiatorazobisisobutyronitrile (AIBN) at a ratio of 100/1 (w/w). 2 ml of thismixture was added to 8.0 ml 0.4% amorphous cellulose suspension,dispersed by a disperser (IKA T18 homogenizer, Germany) at10,000 rpm for 3 min to prepare an emulsion. Then the emulsionwas polymerized at 50 C for 12 h to synthesize solid polystyreneparticles. These particles were deposited on the slide, air-dried andobserved by optical and polarized microscopes.

2.6. Rheological measurements

Fig. 1. Dispersion states of amorphous cellulose as a function of cellulose concentra-tion. From left to right, the concentrations of cellulose are 0.07, 0.28, 0.56, 0.83 and1.10% (w/v), respectively.

X. Jia et al. / Food HydroThe rheologicalpropertiesof the freshlypreparedemulsions (1daystorage) and emulsions stored for 7 days were analyzed using arotational PhysicalMCR301rheometer (AntonPaar,Graz,Austria) anda50mmdiameterparallel plate (PP50)withagapxedat0.5mmgap.All the rheological measurements were carried out at 25 C. Shearviscositywas testedwith the shear rate ranging from0.01 to 1000 s1.A pre-shear process of 100 s at 100 s1 and a time-waiting process of50 s were applied before each test. A linear viscoelastic region (LVR)testwas conductedwithstrain from0.01 to1000%at axed frequencyof 1Hz. Frequencysweep testswas conductedwitha strainof0.1%andfrequencies ranging from 0.1 to 100 Hz.

3. Results and discussion

3.1. Effects of cellulose concentration on the stability of oil-in-wateremulsion

In our previous work, we have determined that cellulose re-generated from phosphoric acid is primarily amorphous with littleresidual crystallinity of cellulose II, and its yield is high, reaching87% consistently, which is an advantage over the production ofcrystalline cellulose (Jia et al., 2013). As shown in Fig. 1, amorphouscellulose can transform between a uid and solid-gel state as afunction of the cellulose concentration. When the concentration isincreased to 1.10%, amorphous cellulose becomes a self-supportinggel as demonstrated by turning the test tube upside-down.

Fig. 2. Effects of amorphous cellulose concentration on the stability of oil-in-wateremulsions (a e 1 day of storage; b e 7 days of storage). From left to right, the con-centrations of cellulose are 0.07, 0.28, 0.56, 0.83 and 1.10%, respectively.

Fig. 3. Effects of initial concentration of amorphous cellulose added to the emulsion onserum fraction, dodecane fraction in emulsion creamand nal concentration of cellulosein emulsion cream(%w/v). Values are represented asmeans standarddeviation (n3).

However, rheology tests demonstrate a gel-like characteristic atlower concentrations (Jia et al., 2014). At concentrations of 0.07 and0.28%, the sedimentation of cellulose is observed, which indicatesthat amorphous cellulose is denser than water. These results alsoindicate that the surface charge density of amorphous cellulose isnot high enough to stabilize itself in aqueous medium by electro-static repulsion. With respect to cellulose nanocrystals prepared byacid hydrolysis, their surfaces are usually more charged and thuspresent better stability in water (Araki, Wada, Kuga, & Okano,1998). However, this property of less charge density may be ad-vantageous in term of emulsion stabilization. It is known thatstrongly charged cellulose nanocrystals prepared by H2SO4 hydro-lysis cannot stabilize emulsions unless the surface charges arescreened by salt (Kalashnikova et al., 2012).

When the concentrations of cellulose are less than 0.83%, thefreshly prepared emulsions (1 day storage) are not stable againstcreaming as dodecane is less dense than water (Fig. 2a). Theextent of creaming decreases with the amount of celluloseadded. According to Stoke's law, the creaming stability of emul-sion can be improved by reducing droplet size, increasing vis-cosity of the continuous phase, or minimizing density differencesbetween droplets and the continuous phase (McClements, 2005).With increasing cellulose concentration, the viscosity is un-doubtedly increased. In addition, as demonstrated by the mi-croscopy studies described later, the adsorption of cellulose atthe emulsion droplet surface should increase the effective den-sity of the droplets, thus minimizing the density difference be-tween the droplets and continuous phase. Both factors may

X. Jia et al. / Food Hydrocolloids 43 (2015) 275e282278Fig. 4. Typical optical micrographs of emulsions stabilized by amorphous cellulose at variedtheir corresponding droplet size during storage (e). The scare bar is 50 mm.cellulose concentrations (a-0.07%, b-0.28%, c-0.55%, d-0.83% w/v; 1 day of storage) and

contribute to the improved stability against creaming. It ispossible to create buoyancy neutral droplets (data not shown) toimprove the emulsion stability by changing the cellulose con-centrations and internal phase fraction to match the density ofdroplets with that of continuous phase, as indicated in thePickering emulsions stabilized by quinoa starch granules (Rayner,Timgren, Sjoo, & Dejmek, 2012). At higher concentrations of0.83% and 1.10%, a complete retardation of creaming up to 3months is achieved, which can be explained by the formation ofa three-dimensional network in the continuous phase that pre-vents the oil droplet from freely moving. It should be noted thatthe color of the 1.10% emulsion is not as whitish as others, and alayer of pure oil oats on the top of the emulsion at such aconcentration. This is most likely because the emulsion viscosityat this concentration is too high to be emulsied well in ourshear condition. Nevertheless, no oiling-off is observed for allother emulsions. These results reveal that amorphous cellulose isvery effective in preventing coalescence, which is the majorcause of oiling-off. The freshly prepared emulsions are stillowable, but with 7 days storage, this uidity is lost completelyeven at the lowest concentration, as demonstrated by turning thetest tubes upside-down (Fig. 2b). It appears that the amorphouscellulose is realigned and associated together to form a three-dimensional network during storage. This is similar to thebehavior of nanocrystalline cellulose (Derakhshandeh, Petekidis,Sabet, Hamad, & Hatzikiriakos, 2013). During aging, the rear-rangement of the microscopic nanocrystalline cellulose particleoccurs and leads to the increase of the storage modulus(Derakhshandeh et al., 2013).

The nal compositions of the emulsions are calculated andplotted in Fig. 3. With increasing cellulose concentration in totalliquid from 0.07 to 0.83%, the serum fraction is decreased from 0.68to zero, and the corresponding dodecane fraction in the emulsioncream is decreased from 0.62 to 0.20. It shows glass-systememulsions with an internal phase more than 60% can be stabi-lized by amorphous cellulose at concentration as low as 0.17%,which is corresponding to 9 mg cellulose per m2 of the interfacialarea stabilized. Although it is not as impressive as that of cellulosenanocrystal stabilization of HIPEs (Capron & Cathala, 2013), theseresults demonstrate excellent emulsion stabilization ability incomparison with MCC and modied starch, with which a concen-tration of 1.5e2.5% is required to achieve a total emulsion stabili-zation (Kargar, Fayazmanesh, Alavi, Spyropoulos, & Norton, 2012).

3.2. Microstructures and droplet sizes of the emulsions

As shown in Fig. 4, the prepared emulsions all have similarmicrostructures and droplet sizes between 20 and 40 mm regardlessthe initial concentrations of cellulose added. Usually the dropletsizes are decreased with increasing amount of particle added tostabilize the emulsions (Rayner et al., 2012). This may be caused bythe less powerful dispersing tool we used and the high viscosity inour emulsion. A combination of the mechanical limit and highviscosity disables the further reduction of particle size to demon-strate the effects of initial cellulose concentration. During storage,the droplet sizes were found to have little change from day 1 to day7. Some droplet sizes were even decreased, probably caused by themeasurement variability between replicates. Similar trend has

X. Jia et al. / Food Hydrocolloids 43 (2015) 275e282 279Fig. 5. Typical uorescence micrographs of emulsions stabilized with amorphous cellulose astorage). The scare bar is 10 mm.t varied cellulose concentrations (a e 0.07%; b e 0.28%; c e 0.55%; d e 0.83%, 1 days of

been observed during the storage of starch granules stabilizedemulsions (Rayner et al., 2012). Nevertheless, our emulsions havevery good long term stability even after threemonths storage (dataof 90 days not shown).

With respect of cellulose nanocrystals, the Pickeringmechanismis believed to be the main mechanism of emulsion stabilization(Kalashnikova et al., 2011). Our uorescence study also demon-strates the role of the Pickering mechanism in emulsion stabiliza-tion of amorphous cellulose (Fig. 5). Isolated drops with uorescentsurfaces are observed as well as a uorescent signal from theexternal phase. This suggests that cellulose is adsorbed on thedrops surfaces as well as being present in the continuous phase.Previously, the amphiphilic character of cellulose was believed toreside in the surface of cellulose crystals (Kalashnikova et al., 2012).Our study indicates that the amorphous cellulose chain can beamphiphilic as well. It is possible that the amorphous cellulosechain presents the hydrophilic hydroxyl groups to water and themore hydrophobic planes of glucopyranose rings toward the oil(Rein et al., 2012), which thus leads to the adsorption behavior onthe oil droplet surface. However, network stabilization undoubt-edly also plays a key role in emulsion stabilization, as shown by thecomplete retarding of creaming at higher concentrations of amor-phous cellulose and loss of uidity after storage (Fig. 2).

Due to the problem of out-of-focus signal associated with con-ventional uorescence microscope, we have employed anothermethod to conrm the distribution of amorphous cellulose on thesurface of emulsion droplets. A styrene-in-water emulsion stabi-lized by amorphous cellulose was prepared and polymerized usingan oil-soluble radical initiator AIBN. The synthesized solid poly-styrene particles were observed by conventional optical andpolarized optical microscopes. Air-dried amorphous cellulose

demonstrate birefringence under polarized lights (Fig. 6b). Thisbirefringence has only been observed in the dried samples becausefresh-prepared amorphous cellulose lost birefringence completelyin wet status (Jia et al., 2013). The presence of these amorphouscellulose aggregates on the surface of polystyrene particles can beclearly observed in Figs. 6c and d, suggesting the absorption ofamorphous cellulose on the original emulsion. Although scanningelectron microscope may provide a better image, these polystyreneparticles are quite large so that the conventional optical micro-graphs already have very good resolution.

Therefore, our study has demonstrated that the emulsion sta-bilization mechanism of amorphous cellulose is a combination ofPickering and network mechanisms. This stabilization mechanismis similar to what has been reported for microbrillated celluloseextracted from mangosteen rind (Winuprasith & Suphantharika,2013). Given that the Pickering mechanism is believed to be moreeffective than the network mechanism in emulsion stabilization(Ghosh, Tran, & Rousseau, 2011), their combination is most likelyresponsible for the excellent emulsion stabilization effects.

3.3. Rheological properties of emulsions

The emulsion viscosity as a function of shear rate with variedcellulose concentrations is shown in Fig. 7. For clarity, the viscositydata are multiplied by a factor of 0.2, 0.5, 2.0 or 5.0 for celluloseconcentrations of 0.07, 0.28, 0.55 or 0.83, respectively. The insetsare the original data. Only the 0.07% sample displays typical shearthinning behavior. For high concentrations, a three-region (shearthinning e plateau or shear thickening e shear thinning) viscosityprole is observed, which has been observed for amorphous cel-lulose dispersions alone. This prole is explained by the shear-

X. Jia et al. / Food Hydrocolloids 43 (2015) 275e282280forms solid aggregates with irregular shape (Fig. 6a), and theyFig. 6. Typical bright-eld optical (a, c and d) and polarized optical micrographs (b) of air-dby amorphous cellulose and polymerized using AIBN as initiator. The scare bar is 100 mm.induced occulation of cellulose particle (Karppinen et al., 2012).ried amorphous cellulose (a and b), and styrene Pickering emulsion (c and d) stabilized

colloX. Jia et al. / Food HydroOur result indicates that the rheology behavior of an amorphouscellulose stabilized emulsion is largely controlled by the rheologybehavior of the continuous phase. When we compared theapparent viscosity values of one-week stored and freshly preparedemulsions (Fig 7 e insets), we found a signicant increase of thesevalues after storage, although the values are not very easy todistinguish in the log scale plot. For the freshly prepared emulsions,the 0.07% sample displays a lower apparent viscosity than theothers (Fig. 7a e inset). After storage, the emulsion's viscosityincreased to match higher concentration emulsions (Fig. 7b einset). These results, combined with the visual inspection results,conrm that rearrangement of amorphous cellulose is occurringduring storage, leading to this increase of apparent viscosity.

The emulsion yield was studied bymeasuring the g dependenceof G0 and G00. G0 decreases slowly for g > 1% before falling sharply asg > 10% (Fig. 8). G00 exhibits a single, well-dened peak at g 10%. Arecent study demonstrated that the G00 of repulsive emulsions ex-hibits a single peak only when the droplet volume fraction is abovethe random close packing (fRCP) of spheres at 0.64, whereasattractive emulsions exhibit one peak below fRCP and two peaksabove fRCP (Datta, Gerrard, Rhodes, Mason, & Weitz, 2011). Thesetwo peaks are explained by the break-up of the weakest bonds inthe stress-bearing connected network composed of compactdroplet aggregates and droplet congurational rearrangements,respectively (Datta et al., 2011). As the droplet volume fractions of

Fig. 7. Viscosity as a function of shear rate for o/w emulsions stabilized with amor-phous cellulose at varied concentrations (a e 1 day of storage; b e 7 days of storage).For clarity, viscosity data are multiplied by a factor of 0.2, 0.5, 2.0 or 5.0 for celluloseconcentrations of 0.07, 0.28, 0.55 or 0.83, respectively. The insets are the original data.ids 43 (2015) 275e282 281our emulsions are all below fRCP (Fig. 3), these results indicate thatthe oil-in-water emulsions stabilized by amorphous cellulose arecharacterized by droplets with attractive interactions. The nature ofthis attractive interaction could be hydrogen bonding or hydro-phobic interactions, and further studies are required.

As shown in Fig. 9, the elastic behavior of an oil-in-wateremulsion is characterized by frequency sweep. All emulsionsdisplay a higher storage modulus (G0) values than loss modulus(G00), and both moduli are almost frequency independent, pre-senting a gel-like behavior. For freshly prepared emulsions, both G0

and G00 are increased with the increase of cellulose concentration.The G0 and G00 of 0.07% sample are signicantly lower than those athigher concentrations. After 7 days of storage, the G0 and G00 of0.07% sample increase to match with that of the 0.28% sample,whereas the moduli of all other concentrations remain ratherconstant, which conrms the structure rearrangement of the 0.07%emulsion during storage as we discussed earlier.

4. Conclusions

Our study has demonstrated that amorphous cellulose preparedby an inexpensive solvent, phosphoric acid, can effectively stabilizeoil-in-water emulsions at a concentration less than 1% throughcombined Pickering and network stabilization mechanisms. Theresulting emulsions are shear-thinning with typical gel character-istics and can be described as emulsions with attractive in-teractions. Compared with cellulose derivatives, the use ofcrystalline cellulose or amorphous cellulose in emulsion stabiliza-tion may be advantageous in terms of regulation and consumeracceptance, although it may be argued that the cellulose

Fig. 8. Strain sweep test of oil-in-water emulsions stabilized with amorphous celluloseat varied concentrations (1 day of storage).

colloids 43 (2015) 275e282X. Jia et al. / Food Hydro282nanocrystals and amorphous cellulose are still native cellulose, asacid hydrolysis or solvent regeneration has been used in theirpreparation. Nevertheless, these celluloses are essentially insolubledietary bers, which not only can function as thickening agents oremulsion stabilizers but also can bring about health benets. Inparticular, amorphous cellulose can be obtained with a high yieldby a simple dissolution and regeneration process and can performthe same or better functions as that of crystalline cellulose.Therefore, amorphous cellulose has considerable potential to beused as a new hydrocolloid in the food industry in addition tocellulose derivatives.

Acknowledgments

This work is supported by a startup fund (804080) from NanjingAgricultural University and projects funded by the Priority Aca-demic Program Development of Jiangsu Higher Education In-stitutions (PAPD), the General Program of National Natural ScienceFoundation of China (31271829), and the Natural Science Founda-tion of Jiangsu Province (BK2012770).

References

Araki, J., Wada, M., Kuga, S., & Okano, T. (1998). Flow properties of microcrystallinecellulose suspension prepared by acid treatment of native cellulose. Colloids andSurfaces, A: Physicochemical and Engineering Aspects, 142, 75e82.

Fig. 9. Storage modulus (lled symbols) and loss modulus (empty symbols) as afunction of frequency for oil-in-water emulsions stabilized with amorphous celluloseat different concentrations (a e 1 day of storage; b e 7 days of storage).Bondeson, D., Mathew, A., & Oksman, K. (2006). Optimization of the isolation ofnanocrystals from microcrystalline cellulose by acid hydrolysis. Cellulose, 13,171e180.

Capron, I., & Cathala, B. (2013). Surfactant-free high internal phase emulsions sta-bilized by cellulose nanocrystals. Biomacromolecules, 14, 291e296.

Cho, H. M., Gross, A. S., & Chu, J.-W. (2011). Dissecting force interactions incellulose deconstruction reveals the required solvent versatility for over-coming biomass recalcitrance. Journal of the American Chemical Society, 133,14033e14041.

Datta, S. S., Gerrard, D. D., Rhodes, T. S., Mason, T. G., & Weitz, D. A. (2011). Rheologyof attractive emulsions. Physical Review E, 84.

Derakhshandeh, B., Petekidis, G., Sabet, S. S., Hamad, W. Y., & Hatzikiriakos, S. G.(2013). Ageing, yielding, and rheology of nanocrystalline cellulose suspensions.Journal of Rheology, 57, 131e148.

Dickinson, E. (2009). Hydrocolloids as emulsiers and emulsion stabilizers. FoodHydrocolloids, 23, 1473e1482.

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

Dickinson, E. (2013). Stabilising emulsion-based colloidal structures with mixedfood ingredients. Journal of the Science of Food and Agriculture, 93, 710e721.

Ghosh, S., Tran, T., & Rousseau, D. (2011). Comparison of pickering and networkstabilization in water-in-oil emulsions. Langmuir, 27, 6589e6597.

Glasser, W. G., Atalla, R. H., Blackwell, J., Brown, R. M., Jr., Burchard, W., French, A. D.,et al. (2012). About the structure of cellulose: debating the Lindman hypothesis.Cellulose, 19, 589e598.

Habibi, Y., Lucia, L. A., & Rojas, O. J. (2010). Cellulose nanocrystals: chemistry, self-assembly, and applications. Chemical Reviews, 110, 3479e3500.

Jia, X., Chen, Y., Shi, C., Ye, Y., Abid, M., Jabbar, S., et al. (2014). Rheological propertiesof an amorphous cellulose suspension. Food Hydrocolloids, 39, 27e33.

Jia, X., Chen, Y., Shi, C., Ye, Y., Wang, P., Zeng, X., et al. (2013). Preparation andcharacterization of cellulose regenerated from phosphoric acid. Journal ofAgricultural and Food Chemistry, 61, 12405e12414.

Kalashnikova, I., Bizot, H., Cathala, B., & Capron, I. (2011). New pickering emulsionsstabilized by bacterial cellulose nanocrystals. Langmuir, 27, 7471e7479.

Kalashnikova, I., Bizot, H., Cathala, B., & Capron, I. (2012). Modulation of cellulosenanocrystals amphiphilic properties to stabilize oil/water interface. Bio-macromolecules, 13, 267e275.

Kargar, M., Fayazmanesh, K., Alavi, M., Spyropoulos, F., & Norton, I. T. (2012).Investigation into the potential ability of pickering emulsions (food-gradeparticles) to enhance the oxidative stability of oil-in-water emulsions. Journal ofColloid and Interface Science, 366, 209e215.

Karppinen, A., Saarinen, T., Salmela, J., Laukkanen, A., Nuopponen, M., & Seppala, J.(2012). Flocculation of microbrillated cellulose in shear ow. Cellulose, 19,1807e1819.

Klemm, D., Kramer, F., Moritz, S., Lindstrom, T., Ankerfors, M., Gray, D., et al. (2011).Nanocelluloses: a new family of nature-based materials. Angewandte Chemie-International Edition, 50, 5438e5466.

Luo, X., & Zhang, L. (2013). New solvents and functional materials prepared fromcellulose solutions in alkali/urea aqueous system. Food Research International,52, 387e400.

McClements, D. J. (2005). Food emulsions: Principles, practice, and techniques (2nded.). New York: CRC Press.

Medronho, B., Romano, A., Miguel, M. G., Stigsson, L., & Lindman, B. (2012). Ratio-nalizing cellulose (in)solubility: reviewing basic physicochemical aspects androle of hydrophobic interactions. Cellulose, 19, 581e587.

Oza, K. P., & Frank, S. G. (1986). Microcrystalline cellulose stabilized emulsions.Journal of Dispersion Science and Technology, 7, 543e561.

Oza, K. P., & Frank, S. G. (1989). Multiple emulsions stabilized by colloidal micro-crystalline cellulose. Journal of Dispersion Science and Technology, 10, 163e185.

Parthasarathi, R., Bellesia, G., Chundawat, S. P. S., Dale, B. E., Langan, P., &Gnanakaran, S. (2011). Insights into hydrogen bonding and stacking interactionsin cellulose. Journal of Physical Chemistry A, 115, 14191e14202.

Rayner, M., Timgren, A., Sjoo, M., & Dejmek, P. (2012). Quinoa starch granules: acandidate for stabilising food-grade pickering emulsions. Journal of the Scienceof Food and Agriculture, 92, 1841e1847.

Rein, D. M., Khaln, R., & Cohen, Y. (2012). Cellulose as a novel amphiphilic coatingfor oil-in-water and water-in-oil dispersions. Journal of Colloid and InterfaceScience, 386, 456e463.

Winuprasith, T., & Suphantharika, M. (2013). Microbrillated cellulose frommangosteen (Garcinia mangostana L.) rind: preparation, characterization, andevaluation as an emulsion stabilizer. Food Hydrocolloids, 32, 383e394.

Xhanari, K., Syverud, K., & Stenius, P. (2011). Emulsions stabilized by microbrillatedcellulose: the effect of hydrophobization, concentration and o/w ratio. Journal ofDispersion Science and Technology, 32, 447e452.

Yamane, C., Aoyagi, T., Ago, M., Sato, K., Okajima, K., & Takahashi, T. (2006). Twodifferent surface properties of regenerated cellulose due to structural anisot-ropy. Polymer Journal, 38, 819e826.

Zhang, Y. H. P., Cui, J. B., Lynd, L. R., & Kuang, L. R. (2006). A transition fromcellulose swelling to cellulose dissolution by o-phosphoric acid: evidencefrom enzymatic hydrolysis and supramolecular structure. Biomacromolecules,7, 644e648.

Zhang, Y. H. P., & Lynd, L. R. (2005). Determination of the number-average degree ofpolymerization of cellodextrins and cellulose with application to enzymatic

hydrolysis. Biomacromolecules, 6, 1510e1515.

Stabilizing oil-in-water emulsion with amorphous cellulose1 Introduction2 Experimental details2.1 Materials2.2 Preparation of amorphous cellulose2.3 Emulsions preparation2.4 Visual inspection of emulsion stability2.5 Optical and fluorescence microscopy2.6 Rheological measurements

3 Results and discussion3.1 Effects of cellulose concentration on the stability of oil-in-water emulsion3.2 Microstructures and droplet sizes of the emulsions3.3 Rheological properties of emulsions

4 ConclusionsAcknowledgmentsReferences