Linear viscoelastic properties of sucrose ester-stabilized oil-in-water emulsions

 View Online  Export Citation CrossMark RESEARCH ARTICLE | NOVEMBER 01 1998 Linear viscoelastic properties of sucrose ester-stabilized oilin-water emulsions A. Guerrero; P. Partal; C. Gallegos Journal of Rheology 42, 1375–1388 (1998) https://doi.org/10.1122/1.550965 Related Content Study of geometrical and electronic structure of lanthanum doped PbTiO3 and PbZrTiO3: First principles calculation AIP Conference Proceedings (November 2018) Linear viscoelastic behavior of commercial and model mayonnaise Journal of Rheology (April 1992) Remarks on high energy muon collider AIP Conference Proceedings (August 2000) Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf Linear viscoelastic properties of sucrose ester-stabilized oil-in-water emulsions A. Guerrero Departamento de Ingenierı ´a Quı ´mica, Universidad de Sevilla, C/P. Garcı ´a Gonza ´lez s/n, 41012 Sevilla, Spain P. Partala) and C. Gallegos Departamento de Ingenierı ´a Quı ´mica, Universidad de Huelva, Escuela Polite ´cnica Superior, Ctra. de Palos de la Frontera s/n, 21819, Huelva, Spain (Received 22 December 1997; final revision received 3 August 1998) Synopsis This paper deals with the influence of composition @sucrose ester SE and oil O concentrations#and temperature on the linear viscoelasticity properties of highly concentrated oil-in-water emulsions, using a hydrophilic sucrose palmitate as emulsifier. Different oil-in-water emulsions were prepared using a sucrose palmitate (SE 51–5 wt %) with a hydrophilic–lipophilic balance of 15, sunflower oil (O 560–80 wt %) and water. Oscillatory measurements were carried out in the linear viscoelasticity region. The relaxation time spectra of the emulsions were calculated using regularization techniques and were fitted to an empirical model proposed by Madiedo ~1996!. The results obtained are explained on the basis of the relationship among linear viscoelasticity properties, droplet size distribution, and characteristics of the continuous phase. Thus, an increase in sucrose palmitate concentration produces higher values of the linear viscoelasticity functions and a broader plateau region in the relaxation time spectrum, which suggests a certain enhancement of the elastic network. © 1998 The Society of Rheology. @S0148-6055~98!01206-1# I. INTRODUCTION Oil-in-water ~o/w!emulsions are manufactured by dispersing the oil phase into the aqueous continuous phase in the form of small droplets ~most commercial emulsions have droplet sizes in the range of 0.1–10 m m!@Walstra ~1983!#. In order to obtain stable emulsions with suitable droplet size distributions ~DSD!, the emulsification process requires a considerable amount of mechanical energy ~in the order of 1–100 MJ/m3de- pending on emulsification machine!@Walstra ~1993!#. It is worth pointing out that the emulsification process is a highly complex unit operation which depends on many variables ~temperature, residence time, agitation speed, etc.!@Gallegos and Franco ~1998!#.It is involved in many different industries to produce a wide variety of emulsions ~foodstuffs, cosmetics, pharmaceuticals, petrochemicals, agrochemicals, etc.!@Walker ~1984!; Yuki et al. ~1990!; Desai, 1990#, which may be formulated to cover a wide range of disperse phase volume fractions f from dilute to highly concentrated emulsions, with f approaching a value of 1. a!Author to whom all correspondence should be addressed. © 1998 by The Society of Rheology, Inc. J. Rheol. 42~6!, November/December 1998 13750148-6055/98/42~6!/1375/14/$20.00 Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf Stability is a primary requirement for the commercial application of emulsions. However, emulsions are thermodynamically unstable because of their large interfacial area and droplets tend to aggregate spontaneously. An emulsion is considered stable when the number, size distribution, and arrangement of droplets do not undergo any discernible change over the storage time scale. An emulsifier is typically added to improve both the emulsification process and the emulsification stability. The emulsifier molecules adsorb at the oil–water interface and reduce the interfacial tension, thereby favoring emulsification and forming a protective barrier around the oil droplets @Dickinson ~1994!#. This increases interdroplet interactions and provides enhanced stability against coalescence. The role of these interactions is fundamental in highly concentrated emulsions, in which very close packing of droplets is found and the continuous phase consists of a thin film separating adjacent droplets. This produces highly viscoelastic responses @Otsubo and Prud’homme ~1994!# and may lead to the development of an extensive flocculation of droplets to form a weak gel-like particulate network depending on the nature and concentration of the emulsifier @Dickinson ~1989!#. The extensive flocculation process provides enhanced stability to the emulsion since the creaming rate is significantly decreased @Franco et al. ~1995a,b, 1997!#. The viscoelastic properties of the emulsions depend on the state of flocculation. Unflocculated or weakly flocculated emulsions show a crossover point between G8and G9, which is related to a characteristic relaxation time for the onset of the terminal or flow region. This relaxation time increases sharply with f because of the stronger interactions among droplets @Tadros ~1993!#. However, emulsions using a mixture of sucrose ester ~SE!and egg yolk proteins as emulsifiers usually show extensive flocculation, depending on the energy input and temperature during the emulsification process, which has been previously related to the development of a plateau region in the relaxation time spectrum @Franco et al. ~1995b, 1997!#. Other highly concentrated emulsions stabilized only by proteins also show gel-like responses (G8is much higher than G9, both are almost parallel!with a wide plateau region @Guerrero and Ball ~1994!; Gallegos et al. ~1992!#. In a previous paper, the ability of a highly hydrophilic ~hydrophilic–lipophilic balance, HLB 515) sucrose palmitate, used as the only emulsifier, to form highly stable emulsions was established @Partal et al. ~1994!#. The flow behavior and droplet size of o/w emulsions stabilized by this sucrose palmitate has also been studied @Partal et al. ~1997!#. An evolution towards smaller sizes and narrower distributions was observed as sucrose palmitate and oil concentrations were raised, which led to an increase in the emulsion viscosity. The objective of the present paper is to study the influence of composition ~sucrose palmitate and oil concentrations!and temperature on the linear viscoelasticity properties of highly concentrated o/w emulsions using a hydrophilic sucrose palmitate as an emulsifier. The relationship among the linear viscoelasticity functions, droplet size distribution, and stability is also analyzed. II. EXPERIMENT Different oil-in-water emulsions were prepared using a Ryoto sucrose palmitate ~P- 1570, HLB 515) from the Mitsubishi–Kasei Food Corporation ~Tokyo, Japan!. This emulsifier was used as received. Sunflower oil was purchased in a local store. Oil-in- water emulsions were prepared using an Ultra-Turrax T-50 homogeneizer. Sunflower oil was added to a solution of the emulsifier in distilled water at 55 °C, and mixed at a speed of 5000 rpm for 3.5 min. The sucrose palmitate concentration ~SE!ranged from 1% to 1376 GUERRERO, PARTAL, AND GALLEGOS Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf 5% ~wt!and the oil content ~O!from 60% to 80% ~wt!( f 50.59–0.77). Emulsions were stored at 5 °C. Oscillatory shear measurements were performed in a Haake RS100 RheoStress rheometer, using a cone-and-plate sensor system C35/4 ~4°, 35 mm!. The stress range for linear viscoelasticity was determined through a dynamic stress sweep at a fixed frequency ~6.28 rad/s!. All samples had the same recent past thermal and rheological history. Rheological tests were carried out in the range of 5–35 °C. The relaxation time spectrum for each emulsion at each temperature was calculated by using a commercial software program ~RHEOSPEC!, which is based on regularization techniques. III. RESULTS AND DISCUSSION A. Influence of sucrose palmitate concentration The evolution of the storage and loss moduli with frequency as a function of the sucrose palmitate concentration is shown in Figs. 1~a!and 1~b!. Both moduli show a continuous increase with emulsifier concentration. The storage modulus G8is always higher than the loss modulus G9within the experimental frequency range. Hence, the emulsions present a predominantly elastic response, although a crossover point tends to appear in the low-frequency regime. Moreover, there is a tendency for G8to exhibit a plateau region @Fig. 1~a!#. This zone, at intermediate frequencies, is known as the plateau region and may be related to the formation of an elastic structural network due to interactions among the emulsifier molecules located at the oil–water interface of adjacent droplets @Dickinson ~1989!; Franco et al. ~1995b!#. This behavior is typical of highly concentrated emulsions such as commercial or model mayonnaises @Gallegos et al. ~1992!; Guerrero and Ball ~1994!# and salad dressings @Mun ˜oz and Sherman ~1992!; Franco et al. ~1995b!#. The plateau region has been extensively described in polymer rheology in terms of an entanglement network among polymer chains in which the FIG. 1. Influence of sucrose palmitate @~a!,~b!# on the storage and loss moduli; and normalized linear viscoelasticity functions for sucrose palmitate-stabilized emulsions ~c!(T525°C). 1377OIL-IN-WATER EMULSIONS Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf motion of molecules is constrained by the neighboring molecules @Ferry ~1980!;Wu ~1989!#. When the emulsion is stabilized by proteins @Gallegos et al. ~1992!; Guerrero and Ball ~1994!#, the three-dimensional network is favored by entanglements among protein segments adsorbed at the oil–water interface. The same behavior was reported when a sucrose stearate was used in combination with egg yolk @Franco et al. ~1995b!#. Taking into account that the emulsions studied are stabilized exclusively by a low molecular weight surfactant ~sucrose palmitate!and have high values of disperse phase volume fraction ~oil droplets!, the plateau region was attributed to an extensive flocculation process that favors emulsion stability by forming a gel-like particulate network @Dickinson ~1989!# due to interdroplet interactions. Similar effects were reported by Ebert et al. ~1988!for highly concentrated emulsions stabilized by a cetyltrimethylamonium-bromide surfactant. In the plateau region, the loss tangent passes through a minimum, which has been used to calculate an approximate value of the plateau modulus, GN 0@Wu ~1989!; Arendt et al. ~1994!; Franco et al. ~1995b!#. The plateau modulus is a viscoelastic parameter, defined as an extrapolation of the entanglement contribution to G8or G9at high frequency @Baumgaertel et al. ~1992!#. This parameter may be approximated from the minimum in tan d as follows @Wu ~1989!#: GN 05@G8#tan d →minimum.~1! As a result, the plateau modulus GN 0undergoes an exponential increase when sucrose palmitate concentration is raised @Fig. 2~a!#, as corresponds to a more structured system. The evolution of G8and G9, and also of the other viscoelastic functions with frequency ~i.e., loss tangent!, is rather similar up to 3% sucrose palmitate. This similarity allows one to superpose all the viscoelastic functions ~between 1% and 3% sucrose palmitate!by using the plateau modulus as a normalization factor @Fig. 1~c!#. However, the normalization procedure fails at the highest concentrations of the emulsifier ~4% and 5%!, because an increase in loss tangent was observed from 3% to 5%. As has been previously reported @Partal et al. ~1997!#, an increase in SE concentration produces a decrease in droplet size and polydispersity, which leads to an increase in the specific interfacial area a@Fig. 2~b!#. Assuming spherical droplets, ‘‘a’’ is related to the Sauter mean diameter as follows: a56/dsv,dsv5 (nidi 3 (nidi 2,~2! where niis the number of droplets having diameter di. However, there is a critical concentration ~3% SE!above which the slope of this increase is lower. To understand this change, the properties of the continuous phase used for the preparation of the emulsions should also be taken into consideration. Thus, the emulsions containing 70% oil and 1%–3% SE were prepared from a continuous phase containing 3%–10% SE, which were isotropic micellar solutions (L1) at the temperature of emulsification ~50 °C!, as may be seen in the phase diagram shown in a previous paper @Calahorro et al. ~1992!#. At concentrations higher than 10% ~as corresponds to emulsions containing 4% and 5% SE!the initial continuous phase was always a dilute dispersion of lamellar liquid-crystalline (L a ) portions in an isotropic micellar matrix, at the emulsification temperature ~50 °C!. Therefore, the above-mentioned critical concentration corresponds to the transition from L1micellar solution to the L11L a dispersion. As a result, a remarkable increase in the low-shear limiting viscosity h oc is observed for the surfactant–water system @Fig. 2~c!#. This increase seems to impair the emulsification process, leading to a lower increase in the interfacial area @Fig. 2~b!#. 1378 GUERRERO, PARTAL, AND GALLEGOS Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf On the other hand, the excess of surfactant over the amount necessary to saturate the interface as a monolayer ( e SE) continuously increases with SE concentration @Fig. 2~b!#. The excess of surfactant has been calculated as follows: e SE 5 ~C r /a f ! Csat 21, ~3! where Cis the sucrose palmitate concentration, r is the emulsion density, ‘‘a’’ is the specific interfacial area, f is the disperse phase volume fraction ~oil droplets!, and Csat is the concentration necessary to saturate the interface as a monolayer. Taking into account the surface pressure-area isotherm obtained by Abran et al. ~1988!for the sucrose palmitate, the Csat value was estimated as 0.36531026g/cm2@Partal et al. ~1997!#. Once the emulsification process is over, the continuous phase of these highly concentrated emulsions consists of a very thin film separating adjacent droplets. The consistency of this film, favored by the increase in e SE, as well as the increase in the specific surface area, explain the large values of the viscoelastic functions and the exponential rise in the plateau modulus shown in Fig. 2~a!. Figure 3 shows the relaxation time spectrum of selected emulsions obtained by means of the Tikhonov regularization procedure using the RHEOSPEC software. As may be observed in Fig. 1, the recalculation of the viscoelastic functions from the relaxation time spectrum obtained by the regularization technique ~RHEOSPEC software!is very accurate. FIG. 2. ~a!Influence of sucrose palmitate concentration on the plateau modulus; ~b!excess of surfactant over the amount necessary to saturate the interface as a monolayer and specific interfacial area as a function of sucrose palmitate concentration; ~c!viscosity of the initial surfactant-containing solution ~SCS!at which oil is added ~please note that the concentration of this solution is SCS 5SE33.3) at 50 °C. 1379OIL-IN-WATER EMULSIONS Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf To express qualitatively the relaxation data, De Rosa and Winter ~1994!proposed an empirical model, derived from the Baumgaertel–Schausberger–Winter ~BSW!spectrum @Baumgaertel et al. ~1992!#, and the Chambon–Winter ~CW!spectrum @Chambon and Winter ~1985!#. Although this model was proposed to describe the relaxation process of polymer melts, it has been also used for emulsions as salad dressings @Franco et al. ~1995b!#. This model requires two functions to describe the transition, plateau, and terminal regions of the relaxation time spectrum. A simple empirical equation, which comprises the three regions of the spectrum, has recently been proposed by Madiedo ~1996! and Madiedo and Gallegos ~1997!: H~l!5 a lm1 b ln 11~l/lt!p,~4! where m,nand (n–p) are the slopes of the transition, plateau, and pseudoterminal zones, respectively; and ltis a characteristic relaxation time related to the boundary of the plateau region. Since the crossover frequency was not reached for the emulsions studied, ltwill be considered a fitting parameter in this paper. As may be noted, these parameters are relatively similar to the parameters of the BSW–CW model @Madiedo and Gallegos ~1997!#. Parameters a and b have the following expressions: a 5mH0 ~m2n!l0 m,~5! b 5nH0 l0 n~n2m!,~6! where l0is a characteristic relaxation time related to the onset of the plateau region and H0is the value of the spectrum corresponding to this relaxation time. If the slope of the plateau region is positive, n.0, Eq. ~4!predicts a maximum value for H(l) for a relaxation time lmax 5lt@n/(n2p)#1/p, which was related by De Rosa and Winter ~1994! FIG. 3. Influence of sucrose palmitate concentration on the relaxation time spectra for sucrose palmitatestabilized emulsions (T525°C). 1380 GUERRERO, PARTAL, AND GALLEGOS Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf with the relaxation time at which H(l) equals GN 0times n. However, this maximum does not always appear. Consequently, the relaxation time spectrum calculated from the regularization method for each emulsion was analyzed according to the Madiedo and Gallegos ~1997!model @Eq. ~4!#. The relaxation time spectra calculated from the Tikhonov regularization method, for each sucrose palmitate concentration, are shown in Fig. 3. As may be seen, the abovementioned model satisfactorily describes the relaxation time spectra obtained for these emulsions. The shape of the spectrum is always similar. A plateau region is always clearly observed. Likewise, a region characterized by a trend to reach the terminal zone or flow region may be noted. The increase in emulsifier concentration produces an increase in the relaxation time spectrum higher than one order of magnitude, which is approximately the same increase undergone by GN 0and indicates that an enhancement of the elastic network is produced. However, Table I shows a maximum value in ltfollowed by a decrease, which corresponds to a displacement of the plateau region towards a shorter relaxation time or higher frequency. This displacement has also been found for salad dressing emulsions stabilized by proteins and sucrose ester @Franco et al. ~1997!#. The slope of the plateau region does not change with SE concentration showing values close to zero. The only exception was the emulsion containing 5% SE, which shows a positive value of n, indicative of a higher degree of structure. A significant increase in the stability of these emulsions was also found as SE became higher @Partal et al. ~1994!#. B. Influence of oil concentration The evolution of the storage and loss moduli with frequency is similar for all the oil concentrations studied, G8always being higher than G9@Figs. 4~a!and 4~b!#. Thus, the loss tangent is roughly independent of the oil concentration. As a result, the plateau modulus @which is shown in Fig. 5~a!# was successfully used to normalize both storage and loss moduli as a function of frequency (a f 8 v ), leading to a master curve for each of them @Fig. 4~c!#. However, some scatter may be observed in G9at high frequencies @Fig. 4~c!#. An empirical horizontal shift factor, (a f 850.4) was necessary for the smallest oil concentration because of a frequency displacement of the plateau region as was previously found for salad dressing emulsions @Franco et al. ~1997!#. Figure 5~a!shows the values of parameter GN 0as a function of the disperse phase volume fraction. An increase in this parameter, with a trend to reach an almost constant TABLE I. Influence of concentration on both the pseudoterminal relaxation time (lt) and slope of the plateau region ~n!, for sucrose palmitatestabilized emulsions. %SE %O nl t~s! 1% 70% 0.017 59 2% 70% 20.006 80 3% 70% 20.013 84 4% 70% 20.025 66 5% 70% 0.045 18 3% 60% 0.003 28 3% 65% 20.035 93 3% 75% 0.004 62 3% 80% 20.002 144 1381OIL-IN-WATER EMULSIONS Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf value above 70% ( f 50.68), was observed. Similarly, the viscoelastic moduli increase with oil concentration but only up to a certain value which depends on frequency. Thus, at low frequencies, the emulsion containing 80% oil has the highest values of G8and G9 @Figs. 4~a!and 4~b!#, while at intermediate and high frequencies there are no differences in G8from 70% to 80% and the highest values of G9correspond to 70% oil. This FIG. 4. Influence of oil concentration @~a!,~b!# on the storage and loss moduli; and normalized linear viscoelasticity functions for sucrose palmitate-stabilized emulsions ~c!(T525°C). FIG. 5. ~a!Influence of disperse phase volume fraction ~oil!on the plateau modulus; ~b!excess of surfactant over the amount necessary to saturate the interface as a monolayer and specific interfacial area as a function of disperse phase volume fraction. 1382 GUERRERO, PARTAL, AND GALLEGOS Downloaded from http://pubs.aip.org/sor/jor/article-pdf/42/6/1375/12481961/1375_1_online.pdf