Journal of Colloid and Interface Science 403 (2013) 7783

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

A new reverse worm-like micellar system from a lecithin, multivalent carboxylic acid and oil mixture
Miko Imai, Kaname Hashizaki , Hiroyuki Taguchi, Yoshihiro Saito, Shigeyasu Motohashi
School of Pharmacy, Nihon University, 7-7-1 Narashinodai, Funabashi, Chiba 274-8555, Japan

a r t i c l e

i n f o

a b s t r a c t
We developed new lecithin organogels composed of reverse worm-like micelles with lecithin/multivalent carboxylic acid/oil systems, and discussed their phase behavior and rheological properties. The most important ndings in this study are the following. From a screening test of many carboxylic acids for gelation, it was found that the number and position of the carboxyl groups of the multivalent carboxylic acids are the determinants for the formation of reverse worm-like micelles, and appropriate carboxylic acids such as citric acid and 1,2,3-propanetricarboxylic acid can change the lecithin/oil solution into a gel. Furthermore, upon addition of these carboxylic acids, the zero-shear viscosity of solutions increased monotonically until phase separation or cloudiness occurred. For example, when citric acid was used, the maximum zero-shear viscosity of the solution was 70,000,000 times larger than that of n-decane. From studies on the scaling of rheological parameters, it was found that further addition of multivalent carboxylic acids not only induced the formation of linear reverse worm-like micelles but also brought about their branching. 2013 Elsevier Inc. All rights reserved.

Article history: Received 5 March 2013 Accepted 13 April 2013 Available online 30 April 2013 Keywords: Lecithin organogel Reverse worm-like micelle Lecithin Multivalent carboxylic acid Small angle X-ray scattering Phase diagram Rheology

1. Introduction Lecithin, which is a key component of cell membranes, is an amphiphilic molecule composed of both hydrophilic and hydrophobic groups. Luisi et al. [1] reported that organic solvents can be transformed into highly viscous gel-like solutions by the addition of both lecithin and a small amount of water, and this research revealed that the combination of lecithin and water acts as an oil gelator. In general, lecithin forms spherical or ellipsoidal reverse micelles when added alone to oil. When trace amounts of water are then added, the water molecules attach to the phosphate groups of neighboring lecithins via hydrogen bonds and reduce the interface curvature of the molecular assemblies, thus inducing the formation of reverse worm-like micelles. The micelles get tangled in oil and form a three-dimensional network throughout the solution. As a result, they turn the solution into an organogel (also called a lecithin organogel) [16]. In this system, water is the key ingredient for the formation of reverse worm-like micelles. Additionally, this simple system is in a thermodynamic equilibrium state, which is different from that for low-molecular-weight gelators. The advantages of lecithin organogels are their biodegradability, biocompatibility, and long-term stability [1,2,7,8]. These organogels are also easy to prepare and can solubilize molecules having various polarities [1]. These advantages might enable
Corresponding author. Fax: +81 47 465 6895.
E-mail address: hashizaki.kaname@nihon-u.ac.jp (K. Hashizaki). 0021-9797/$ - see front matter 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.04.033

lecithin organogels to be applied to a variety of commercial products such as cosmetics, medicines, foods, and ink. Topical drug delivery applications of lecithin organogel-based lecithin/water/ oil systems have also been studied by many researchers [7,8]. The lecithin that is included in the gel acts as a skin permeation enhancer for drugs [9]. Therefore, the lecithin organogel is a material in which the expansion rate is high. While applied research on lecithin organogels continues to advance, key ingredients that can be used as substitutes for water in organogel preparations have actively been explored by many researchers. Shuchipunov reported that glycerol, ethylene glycol, and formamide can induce the formation of lecithin organogels [10]. It should be noted that all the substances mentioned so far are liquids at room temperature. Recently, some very interesting results in which bile salts, which are solid substances at room temperature, induced the formation of a lecithin organogel were reported by Raghavan et al. [1113]. The most remarkable aspect of their report was the method used for sample preparation. If lecithin, a bile salt, and a nonpolar organic solvent are mixed together simultaneously, reverse worm-like micelles do not form because the bile salts do not dissolve in nonpolar organic solvents. Therefore, these researchers homogeneously mixed the lecithin and bile salts in methanol in advance, and then removed the solvent to obtain a solid mixture. Next, they added a nonpolar solvent such as cyclohexane to the solid with agitation to obtain a gel-like substance. They also reported the formation of organogels with p-coumaric acid [14] and multivalent cations of inorganic salts

78

M. Imai et al. / Journal of Colloid and Interface Science 403 (2013) 7783

(Ca2+, Mg2+, La3+, and Ce3+) [15]. The method presented by Raghavan et al. may be of use in searching for other solid substances that induce the formation of lecithin organogels. Indeed, we adopted their methods and discovered that urea [16], sucrose fatty acid esters [17], D-ribose [18], 2-deoxy-D-ribose [18], polyglycerols [19], and ascorbic acid [20] also induce the formation of lecithin organogels. In this study, we focused on multivalent carboxylic acids that have some carboxyl groups as polar groups for binding to lecithins. Raghavan et al. [14] already reported that p-coumaric acid, which is a monocarboxylic acid, can induce the formation of lecithin organogels; however, other isomers of coumaric acids, o- and mcoumaric acid, did not possess gel-forming ability. It is relevant to note that the possibility of using carboxylic acids has not been examined extensively in the literature. In this paper, we show that multivalent carboxylic acids such as citric acid and 1,2,3-propanetricarboxylic acid can induce highly viscoelastic lecithin organogels composed of reverse worm-like micelles. The present study investigated the phase state and rheology to understand the reverse worm-like micelles of lecithin/multivalent carboxylic acid/oil systems. Additionally, to describe the entangled state of reverse worm-like micelles, we examined the scaling theory of rheological parameters. 2. Experimental 2.1. Materials Soybean lecithin (PHOSPHOLIPON90G; PC content = min. 94%) was purchased from H. Holstein GmbH & Co. (Hamburg, Germany). Citric acid, glycolic acid, L(-)-malic acid, propionic acid, benzoic acid, salicylic acid, succinic acid, o-phthalic acid, tartaric acid, 1,3,5-trimesic acid, acetic acid, fumaric acid and n-decane were purchased from Kanto Chemical Co., Inc. (Tokyo, Japan). 1,2,3Propanetricarboxylic acid and 3,4-dihydroxybenzoic acid were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). trans-Aconitic acid, pyromellitic acid and mellitic acid were purchased from SigmaAldrich Co. (MO, USA). 1,4,5-Trimellitic acid was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). All the chemicals were used without further purication. 2.2. Sample preparation The required amounts of lecithin and carboxylic acid were dissolved in methanol in a vial and then the solvent was completely removed in a desiccator equipped with a vacuum pump. Following addition of n-decane to the vial, the solution was mixed with a magnetic stirrer. The resulting solutions were stored at 25 C for several days to allow equilibration. 2.3. Phase diagrams Phase diagrams of the lecithin/carboxylic acid/n-decane were obtained by visual observation through crossed polarizers and by small-angle X-ray scattering (SAXS) analysis. SAXS was performed using a Nano-STAR instrument (Bruker AXS Inc., WI, USA) with a Cu Ka radiation source operating at 45 kV/120 mA. All measurements were performed at 25 C. 2.4. Rheological measurements Steady and dynamic rheological measurements were performed using a stress-controlled rheometer (MARS III, Thermo Fisher Scientic Inc., MA, USA). Samples were run on a doublecone geometry (60 mm diameter, 1 cone angle), a cone-plate

geometry (60 mm and 35 mm diameters, 1 cone angle), a parallel-plate geometry (35 mm and 20 mm diameters, 1 mm gap), or a serrated parallel-plate geometry (20 mm diameter, 1 mm gap). A Peltier-based temperature control device was used for setting the temperature at 25 C. For steady-ow viscosity measurements, the samples were subjected to the desired shear stress for a sufciently long time to achieve a steady state. Dynamic rheological measurements were performed using the strain (c) value from the linear viscoelastic region. A solvent trap was used to prevent evaporation.

3. Results and discussion 3.1. Phase diagrams of lecithin/multivalent carboxylic acid/n-decane systems Fig. 1 shows photographs of the lecithin/citric acid/n-decane solutions at a xed lecithin concentration of 10 wt.% and with increasing concentration of citric acid. At low citric acid concentration (Fig. 1a), the solution has a low viscosity. However, the solution changed into a transparent gel at a certain citric acid concentration (Fig. 1b). Further addition of citric acid changed the transparent gel into a turbid gel (Fig. 1c). Next, to identify the structure of the transparent gel from the lecithin/citric acid/ n-decane system, we carried out SAXS measurements. All samples used for SAXS measurements were diluted ve times to eliminate the structure-factor effect. Fig. 2 shows the SAXS scattering intensity [I(q)] as a function of the scattering vector (q) [q 4p=k sin h, where k is the X-ray wavelength and 2h is the scattering angle] for the lecithin/citric acid/n-decane [2:0.35:97.65 (wt.%)] system. In this SAXS prole, the slope of the double logarithmic plot in the low-q region was 1. This indicated cylindrical particles, i.e., the existence of reverse worm-like micelles. Furthermore, the section radius and the contour length of the reverse worm-like micelles were estimated from the SAXS proles. The section radius can be calculated from cross-sectional plots of the SAXS prole, and the contour length can be estimated by a simulation based on the scattering function of cylindrical particles (see Refs. [16,17,19] for more information). As a result, the section radius and the contour length of the reverse worm-like micelles were determined to be 1.9 nm and 50 nm or more, respectively. Since we found that citric acid is a useful polar substance for forming reverse worm-like micelles having high viscoelasticity, we conducted a screening test for various kinds of carboxylic acids (normal carboxylic acids or hydroxycarboxylic acids). Table 1 summarizes the results of the gelation ability of various carboxylic acids against n-decane at a xed lecithin concentration of 10 wt.%. Here, we classied the gelation ability of the carboxylic acids into four classes (i.e., sol, viscous, gel, or insoluble) by visual observations. The upper half of Table 1 shows the results for normal carboxylic acids. It can be seen that normal carboxylic acids having two or less carboxyl groups did not cause gelation. Interestingly, normal carboxylic acids having three carboxyl groups were able to induce the formation of gels. However, normal carboxylic acids having four or more carboxyl groups tended to self-assemble, and they did not induce the formation of reverse worm-like micelles. The lower half of Table 1 shows the results for hydroxycarboxylic acids. When hydroxycarboxylic acids having one carboxyl group were used, only 3,4-dihydroxybenzoic acid and p-coumaric acid were able to induce the formation of gels. The structure of the hydroxycarboxylic acid, that is, the positions of the hydroxyl and carboxyl groups, might inuence gel formation. Furthermore, all hydroxycarboxylic acids having two or three carboxyl groups used in this study induced oil gelling.

M. Imai et al. / Journal of Colloid and Interface Science 403 (2013) 7783

79

Fig. 1. Photographs of (a) lecithin:citric acid:n-decane = 10:0.7:89.3 (wt.%), (b) lecithin:citric acid:n-decane = 10:1.75:88.25 (wt.%), and (c) lecithin:citric acid:ndecane = 10:3.0:87.0 (wt.%).

10

10

-1

10

-1

10

-2

10

-3

10

-4

10

-1

10

q (nm )
Fig. 2. SAXS scattering intensity [I(q)] as a function of the scattering vector (q) for lecithin/citric acid/n-decane [2:0.35:97.65 (wt.%)] system.

-1

Next, we determined the partial phase diagrams for the lecithin/carboxylic acid/n-decane system for some carboxylic acids that were used for the screening test (Fig. 3). The phase states of the phase diagrams were identied by visual observation through crossed polarizers or through SAXS analysis. In the lecithin/glycolic acid/n-decane and lecithin/succinic acid/n-decane systems, although both systems showed the reverse micellar phase (Om) by the addition of a small amount of these carboxylic acids, highly viscous regions (areas colored red) were not conrmed within the Om regions. This result indicates that the reverse worm-like micelles formed in these systems were not long enough to increase the viscosity. On the other hand, in the lecithin/malic acid/n-decane, lecithin/citric acid/n-decane, and lecithin/1,2,3-propanetricarboxylic acid/n-decane systems, highly viscous regions (areas colored red) were conrmed within the Om regions. In these regions, it is clear that reverse worm-like micelles with sufcient length for entanglement were formed. Further addition of citric acid resulted in turbid solutions that retained their high viscosity (turbid phase). The most plausible explanation for this turbidity is the formation of larger assemblies. In contrast, in systems using carboxylic acids other than citric acid, further addition of carboxylic acid caused the solution to separate into two coexisting liquid

Table 1 Gelation ability of various carboxylic acids against n-decane at a xed lecithin concentration of 10 wt.%. Compound Monocarboxylic acid Acetic acid Propionic acid Benzoic acid Succinic acid Fumaric acid o-Phthalic acid trans-Aconitic acid 1,2,3-Propanetricarboxylic acid 1,3,5-Trimesic acid 1,4,5-Trimellitic acid Pyromellitic acid Mellitic acid Glycolic acid Salicylic acid o-Coumaric acid m-Coumaric acid p-Coumaric acid 3,4-Dihydroxybenzoic acid Tartaric acid
L(-)Malic acid Citric acid

Normalized I(q) (a.u.)

State sol sol sol sol viscous sol gel gel gel viscous insoluble insoluble viscous sol viscous [14] viscous [14] gel [14] gel gel gel gel

Dicarboxylic acid

Tricarboxylic acid

Tetracarboxylic acid Hexacarboxylic acid Hydroxymonocarboxylic acid

Hydroxydicarboxylic acid Hydroxytricarboxylic acid

80

M. Imai et al. / Journal of Colloid and Interface Science 403 (2013) 7783

Lecithin 0 0.3

0.2

Om

0.1

0.1

2-Phase

0.2
O HO OH

0.3

Lecithin 0

n-Decane 0

Om
1.0

0.3 0.7

Glycolic acid

0.2

Om

0.1

0.1

0.2

2-Phase
HO

OH OH

0.1

2-Phase

0.2
O HO OH O

n-Decane 0

Om
1.0

0.3 0.7

L(-)-Malic acid

n-Decane 0

Om
1.0

0.3 Succinic acid 0.7

line A line B
0.1

line A
0.2
O HO O OH O OH OH

line B
0.1

line C Turbid

line C 2-Phase

0.2
O HO O HO O OH

n-Decane 0

0.3 1.0 0.9 0.8 0.7

Citric acid

n-Decane 0

1.0

0.9

0.8

0.3 1,2,3-Propanetri0.7 carboxylic acid

Fig. 3. Partial phase diagrams of lecithin/carboxylic acid/n-decane systems in the dilute region at 25 C. The notation Om represents the reverse micellar phase. The region of high viscosity within the Om phase is colored red. Lines AC0 within the phase diagrams are used later to investigate the scaling theory of reverse worm-like micelles.

phases (2-phase region). It is reasonable to think that the phase separation resulted from the shrinking of reverse worm-like micelles by the addition of excess carboxylic acid. Similar phase behavior has been conrmed in other systems [17,18,20]. As a side note, lines AC0 within the phase diagrams are used later to investigate the scaling theory of reverse worm-like micelles. 3.2. Formation mechanism of reverse worm-like micelles in lecithin/ multivalent carboxylic acid/n-decane systems Carboxylic acids cause the formation of reverse worm-like micelles as follows. It is well known that lecithin forms spherical or ellipsoidal reverse micelles that have a large interface curvature in oil. For these reverse spherical micelles to grow into reverse worm-like micelles, the interface curvature of the molecular assembly must decrease slightly. In the conventional reverse worm-like micellar systems using polar substances, such as water [1] or urea [16], hydroxyl groups or amino groups of polar substances attach to the phosphate groups of lecithins via electrostatic interactions, and the gap between the head groups of the neighboring lecithin molecules widens. Thus, the interface curvature of the molecular assembly decreases, causing reverse spherical micelles to transform into reverse worm-like micelles. On the other hand, it is likely that the carboxylic acids used in this study bind not only to the phosphate groups but also to the quaternary ammonium group of choline via electrostatic interactions, and induce the formation of reverse worm-like micelles. Furthermore, it was found that the number and position of the carboxyl groups of the carboxylic acids are the determinants for the formation of reverse wormlike micelles. Further consideration will be needed to determine the binding positions of the carboxylic acid and lecithin.

3.3. Rheological behavior of lecithin/multivalent carboxylic acid/ndecane systems Based on the above results, the rheological properties of the reverse micellar solutions, which were prepared at a xed lecithin concentration of 10 wt.% and varying multivalent carboxylic acid concentrations, were examined.

10

Citric acid L(-)-Malic acid Glycolic acid

10

0 (Pas)

10

10

10

-2

0.0

0.5

1.0

1.5

2.0

2.5

Concn. of carboxylic acid (wt%)

Fig. 4. Zero-shear viscosity (g0) of lecithin/carboxylic acid/n-decane systems as a function of carboxylic acid concentration at 25 C. Lecithin concentration was xed at 10 wt.%.

M. Imai et al. / Journal of Colloid and Interface Science 403 (2013) 7783

81

First, we carried out steady-ow viscosity measurements for the reverse micellar solutions. Fig. 4 shows the relationship between zero-shear viscosity (g0), given by the steady-ow viscosity measurements, and the carboxylic acid (glycolic acid, L(-)-malic acid, and citric acid) concentrations. The g0 values for all systems increased monotonically until phase separation or cloudiness occurred. It follows from this that reverse worm-like micelles can grow upon carboxylic acid addition. L(-)Malic acid and citric acid induced the formation of reverse wormlike micelles with sufcient length for entanglement. Next, we examined the difference among these carboxylic acids. The viscosity curves became steeper as the number of carboxyl groups increased. Namely, although the viscosity increased slightly when glycolic acid was used, a gel was not formed, as shown in Fig. 3. However, a gel having moderate viscosity was formed when L(-)-malic acid having two carboxyl groups was used. Furthermore, citric acid having three carboxyl groups induced the formation of a gel with extremely high viscosity. Signicantly, the maximum zero-shear viscosity of this solution was 70,000,000 times larger than that of n-decane when citric acid was used. This viscosity value is signicantly larger than that of the reverse worm-like micellar solutions that we have reported to date [1620]. As mentioned previously, this reason is probably related to the binding mode of multivalent carboxylic acid to lecithins. Next, to characterize the viscoelasticity of the lecithin/carboxylic acid/n-decane systems, we carried out dynamic viscoelasticity measurements [variation in the storage modulus (G0 ) and loss modulus (G00 ) as a function of frequency (x)]. Here, G0 and G00 represent elasticity and viscosity, respectively. Fig. 5 shows the results of the dynamic viscoelasticity measurements at a xed lecithin concentration of 10 wt.% and varying citric acid concentrations. G0 and G00 intersect at a certain x. At high frequencies, the elasticity component is predominant (G0 > G00 ), while at low frequencies, the viscosity component is predominant (G00 > G0 ). It is well known that the dynamic viscoelastic behavior of reverse worm-like micelles follows the single Maxwell model with a single relaxation time (s). This relaxation time indicates the disentanglement time of reverse worm-like micelles. In general, the relaxation time for linear polymers has a wide distribution because the polymers disentangle via reptation and the relaxation time is proportional to the third power of the molecular weight [2124]. Here, reptation implies the mode of motion by which the entangled polymers gradually disentangle from the chain end

under the inuence of their own Brownian motion. In contrast, the relaxation mode for reverse worm-like micelles is different from that of polymers. It is well known that each surfactant in monomers and micelles is in a thermodynamic equilibrium state, and micelle-monomer exchange occurs constantly. In other words, the micelles break and recombine at a rapid rate. Therefore, the rheological behavior of reverse worm-like micelles is determined by the balance between reversible breakage and reptation of the micelles, and single Maxwellian behavior is generally observed when the reversible breakage time is sufciently shorter than the reptation time [11,25]. If the viscoelastic behavior of reverse worm-like micelles follows the single Maxwell model, Eqs. (1) and (2) t well to the data obtained by the dynamic viscoelasticity measurements. In this situation, the relaxation time (s) is represented by the geometric mean of the breaking time (sb) and reptation time (srep) (Eq. (3)) [25],

G0 G00

x2 s 2 G0 1 x2 s2 xs
1 x2 s 2 G0

1 2 3

s sb srep 1=2

where G0 is the plateau modulus [11,25]. When a sample follows a single Maxwell model, G0 can be calculated from the ColeCole plot (relation between G0 and G00 based on Eq. (4)) and g0 is determined by the balance between G0 and s (Eq. (5))

 2 G0 G2 G0 G002 0 2 4

4 5

g0 G0 s

10

G', G', G',

G'' (Citric acid = 2.25 wt%) G'' (Citric acid = 1.75 wt%) G'' (Citric acid = 1.25 wt%)

As shown in Fig. 5, the rheological behavior of the reverse worm-like micelles showed a good t with the Maxwellian curves at low frequencies. However, the t was not good at high frequencies, which indicated that the relaxation mode of the reverse worm-like micelles formed in this system did not strictly follow the single Maxwell model and that faster relaxation occurred in addition to the relaxation of reversible breakage. This faster relaxation mode is thought to be related to micro-Brownian motion (i.e., the Rouse mode [26]) of the reverse worm-like micelles. Moreover, the entire frequency spectrum moved to the upper left with increasing citric acid concentration. We obtained similar results in our previous studies [16,17,19,20]. This means that citric acid not only induces the growth of reverse worm-like micelles, but also induces a greater number of entangled micelles. 3.4. Scaling theory for some rheological parameters of lecithin/ multivalent carboxylic acid/n-decane systems To describe the entangled state of reverse worm-like micelles of lecithin/multivalent carboxylic acid/n-decane systems, we examined the scaling dependence of rheological parameters. The scaling of rheological parameters for the linear and branched worm-like micelles, as a function of surfactant volume fraction (i.e., wormlike micellar volume fraction), have been studied in detail by various groups [2,25,2732], and those scaling exponents are based on a model developed by Cates. Namely, the rheological behavior of linear worm-like micelles is characterized by the following scaling laws:

10

G', G'' (Pa)

10

10

10

10

-1

10

-3

10

-2

10

-1

(rad/s)

10

10

10

G0  /2:25 ; g0  /3:5 ; s  /1:25

where / is the volume fraction of surfactant. On the other hand, the branched worm-like micelles are characterized by the following scaling laws:

Fig. 5. Variation in G and G as a function of x at different citric acid concentrations in the lecithin/citric acid/n-decane system at 25 C. Lecithin concentration was xed at 10 wt.%. Maxwellian ttings to the experimental data are shown by solid lines.

00

82

M. Imai et al. / Journal of Colloid and Interface Science 403 (2013) 7783 Table 2 Values of the exponents in the scaling laws. Model/experiment Theoretical values Linear micelles Branched micelles Experimental values Lecithin:CA 1:0.1 (line A) 1:0.13 (line B) 1:0.175 (line C) Lecithin:PtcA 1:0.1 (line A0 ) 1:0.125 (line B0 ) 1:0.18 (line C0 ) G0 2.25 2.25

G0  /2:25 ; g0  /2:5 ; s  /0:25

Shchipunov et al. reported [27] the scaling dependence of rheological parameters for lecithin/water/n-decane systems in detail. In their report, at low water content, the scaling exponents of the rheological parameters were quite close to the theoretical values predicted by Cates model for the linear worm-like micelles. At high water content, in contrast, the scaling exponents of the rheological parameters were in satisfactory agreement with those expected from a model of the branched worm-like micelles. We adopted their technique, and studied the variation of rheological parameters with lecithin volume fraction (/) at xed values of the lecithin-to-carboxylic acids weight ratio. In this investigation, citric acid (CA) and 1,2,3-propanetricarboxylic acid (PtcA) were used as the multivalent carboxylic acids of lecithin/multivalent carboxylic acid/n-decane systems. Specically, we studied different weight ratios of lecithin-to-carboxylic acids indicated by the lines (A, B, C, A0 , B0 and C0 ) in the ternary phase diagrams (Fig. 3). Namely, the weight ratios lecithin:CA for line A, line B, and line C are 1:0.1, 1:0.13, and 1:0.175, respectively. The weight ratios of lecithin:PtcA for line A0 , line B0 , and line C0 are 1:0.1, 1:0.125, and 1:0.18, respectively. Lines A and A0 are located in the left part of highly viscous regions, lines B and B0 are located in the central part, and lines C and C0 are located in the right part. As an example, Fig. 6 shows the rheological parameters (G0, g0, and s) for lecithin/CA/ndecane systems as a function of lecithin volume fraction (/) at three different weight ratios. Rheological measurements were carried out within the lecithin volume fractions (/) shown by the solid lines (Fig. 3). The straight lines in the graphs are determined by the experimental points using the least-squares method. The exponents of concentration dependence for each rheological parameter are determined from the slopes of these dependences. The results of these calculations and the values of theoretical exponents obtained for the models of linear and branched worm-like micelles are listed in Table 2. Comparison of the experimental and theoret-

g0
3.5 2.5

s
1.25 0.25

Refs. [27] [27]

2.80 0.01 2.44 0.09 2.31 0.20 3.39 0.08 2.43 0.03 2.04 0.10

3.81 0.12 3.18 0.17 2.54 0.22 4.94 0.12 3.11 0.04 2.13 0.15

1.01 0.12 0.75 0.09 0.23 0.04 1.56 0.07 0.68 0.04 0.09 0.05

10

4 3 2 1 0 2 5 4 3 2 1 2 2 1 0 3 4 5 6 7 8 9 3 4 5 6 7 8 9

G0 (Pa)

10 10 10 10 10

line C (Lecithin : CA = 1 : 0.175) line B (Lecithin : CA = 1 : 0.13) line A (Lecithin : CA = 1 : 0.1)

0.1

line C line B line A

0 (Pas)

10 10 10 10 10 10

0.1

10 10 10

-1 -2

line C line B line A

2 3 4 5 6 7 8 9 2 3

ical values shows that the scaling law exponents for line A (lecithin:CA = 1:0.1), line B (lecithin:CA = 1:0.13), and line B0 (lecithin:PtcA = 1:0.125) approach the theoretical values for linear micellar models. This means that, in the left and central parts of highly viscous regions, the systems include predominantly linear worm-like micelles. On the other hand, the exponents for both line C (lecithin:CA = 1:0.175) and line C0 (lecithin:PtcA = 1:0.18) are close to that for a branched micellar model. This result means that these systems include predominantly branched worm-like micelles in the right part of highly viscous regions. However, the exponents for line A0 (lecithin:PtcA = 1:0.1) are considerably larger than the theoretical exponents for both models. Similar large exponents were previously reported by Raghavan et al. for the lecithin/bile salt/cyclohexane system [11]. This discrepancy arises because the lecithin volume fraction (/) does not agree with the volume fraction of reverse worm-like micelles actually entangled. That is to say, at a low PtcA content, the system includes not only entangled reverse worm-like micelles but also short reverse worm-like micelles that do not take part in entanglement. It follows from these results that increasing the carboxylic acid concentration can transform the reverse micelles from spherical, to linear worm-like, and nally to branched worm-like. Turning next to the zero-shear viscosity of reverse worm-like micelles, we note the large difference between our systems and other systems. In a typical lecithin/water/isooctane system, the zero-shear viscosity reaches a maximum at a certain water concentration [2]. This behavior has been interpreted in terms of micellar branching [3032]. These joints of branched worm-like micelles can slip along the micellar body easily to release the stress quickly. Thus, the branched worm-like micelles display a reduced viscosity compared to entangled linear worm-like micelles. In a number of surfactant systems in aqueous media, such micellar connections or branching points have been detected [3337]. However, as shown in Fig. 4, the zero-shear viscosity of the lecithin/CA/n-decane system did not reach the maximum although branched worm-like micelles were formed at higher CA concentration. The most plausible explanation for the discrepancy in zero-shear viscosities is the difference of the motility of the joints. Namely, it is suggested that the zero-shear viscosity of the micellar solution does not decrease because the joints of branched worm-like micelles, which we discovered, do not slip along the micellar body easily. More detailed examination of the branched micellar model is underway in our laboratory. 4. Conclusions Lecithin organogel systems composed of reverse worm-like micelles have been studied previously, with a focus on their phase states and rheological properties [1,1020]. Lecithin organogels

(s)

0.1

Fig. 6. Rheological parameters (G0, g0, and s) of lecithin/citric acid/n-decane systems as a function of lecithin volume fraction (/) at three different weight ratios (lecithin:CA).

M. Imai et al. / Journal of Colloid and Interface Science 403 (2013) 7783

83

are generally formed in combination with lecithin, polar substances, and oil. Strong dipolar/electrostatic interactions, such as the hydrogen bonds between lecithins and polar substances in oil, are very important for the formation of reverse worm-like micelles. In conventional reverse worm-like micellar systems using polar substances, such as water [1], glycerol [10], formamide [10], or urea [16], the hydroxyl groups or amino groups of polar substances attach to the phosphate groups of lecithins via hydrogen bonding. However, little is known about the inuence of different polar groups on the properties of lecithin organogels. Hence, in this study, we focused on the carboxyl group as a new polar group, and used multivalent carboxylic acids. Screening results for various carboxylic acids revealed the requirements for the formation of highly viscoelastic reverse worm-like micelles. Specically, multivalent carboxylic acids that can be used as polar substances must have 2 or 3 carboxyl groups in their molecular structures. The zero-shear viscosity of reverse worm-like micelles increased depending on the quantity of multivalent carboxylic acid added. When citric acid was used, the maximum zero-shear viscosity of the solution was 70,000,000 times larger than that of n-decane. This viscosity value was signicantly larger than that of the reverse worm-like micellar solutions we have reported to date [1620]. Scaling-law studies for rheological parameters showed that the spherical reverse micelles transformed into reverse branched micelles via reverse linear micelles as the content of multivalent carboxylic acid increased. Thus, we concluded that multivalent carboxylic acids are useful polar substances because of their ability to induce the formation of lecithin organogels. These results may aid in the search for other polar substances, and help to explain the formation mechanisms of different types of lecithin organogel systems. Areas for future study include applied research in various industrial elds, such as medicine, food, cosmetics, and ink. Acknowledgments Thanks are due to Mr. Naohito Watanabe, Mr. Shinya Morita and Mr. Takahiro Yamada for their technical assistance in the experimental work. This work was supported in part by a Grantin-Aid for Young Scientists (B) 24780134 from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan.

References
[1] R. Scartazzini, P.L. Luisi, J. Phys. Chem. 92 (1988) 829. [2] P. Schurtenberger, R. Scartazzini, P.L. Luisi, Rheol. Acta 28 (1989) 372. [3] P.L. Luisi, R. Scartazzini, G. Hearing, P. Schurtenberger, Colloid. Polym. Sci. 268 (1990) 356. [4] R. Angelico, G. Palazzo, G. Colafemmina, P.A. Giustini, A. Ceglie, J. Phys. Chem. B 102 (1998) 2883. [5] R. Angelico, A. Ceglie, U. Olsson, G. Palazzo, Langmuir 16 (2000) 2124. [6] F. Aliotta, M.E. Fontanella, M. Pieruccini, G. Salvato, S. Trusso, C. Vasi, R.E. Lechner, Colloid Polym. Sci. 280 (2002) 193. [7] R. Kumar, O.P. Katare, AAPS PharmSciTech 6 (2005) E298. [8] S. Raut, S.S. Bhadoriya, V. Uplanchiwar, V. Mishra, A. Gahane, S.K. Jain, Acta Pharm. Sin. B 2 (2012) 8. [9] H. Willimann, P. Walde, P.L. Luisi, A. Gazzaniga, F. Stroppolo, J. Pharm. Sci. 81 (1992) 871. [10] Y.A. Shchipunov, Colloid Surf. A 183 (2001) 541. [11] S.H. Tung, Y.E. Huang, S.R. Raghavan, J. Am. Chem. Soc. 128 (2006) 5751. [12] S.H. Tung, Y.E. Huang, S.R. Raghavan, Langmuir 23 (2007) 372. [13] S.H. Tung, S.R. Raghavan, Langmuir 24 (2008) 8405. [14] R. Kumar, A.M. Ketner, S.R. Raghavan, Langmuir 26 (2010) 5405. [15] H.Y. Lee, K.K. Diehn, S.W. Ko, S.H. Tung, S.R. Raghavan, Langmuir 26 (2010) 13831. [16] K. Hashizaki, T. Chiba, H. Taguchi, Y. Saito, Colloid Polym. Sci. 287 (2009) 927. [17] K. Hashizaki, H. Taguchi, Y. Saito, Colloid Polym. Sci. 287 (2009) 1099. [18] K. Hashizaki, H. Taguchi, Y. Saito, Chem. Lett. 38 (2009) 1036. [19] K. Hashizaki, Y. Sakanishi, S. Yako, H. Tsusaka, M. Imai, H. Taguchi, Y. Saito, J. Oleo Sci. 61 (2012) 267. [20] K. Hashizaki, N. Watanabe, M. Imai, H. Taguchi, Y. Saito, Chem. Lett. 41 (2012) 427. [21] M. Doi, S.F. Edwards, J. Chem. Soc. Faraday Trans. 2 (74) (1978) 1789. [22] M. Doi, S.F. Edwards, J. Chem. Soc. Faraday Trans. 2 (74) (1978) 1802. [23] M. Doi, S.F. Edwards, J. Chem. Soc. Faraday Trans. 2 (74) (1978) 1818. [24] M. Doi, S.F. Edwards, J. Chem. Soc. Faraday Trans. 2 (75) (1979) 38. [25] M.E. Cates, S.J. Candau, J. Phys. Condens. Matter. 2 (1990) 6869. [26] P.E. Rause, J. Chem. Phys. 21 (1953) 1272. [27] Y.A. Shchipunov, H. Hoffmann, Langmuir 14 (1998) 6350. [28] Y.A. Shchipunov, E.V. Shumilina, W. Ulbricht, H. Hoffmann, J. Colloid Interface Sci. 211 (1999) 81. [29] E.V. Shumilina, Y.L. Khromova, Y.A. Shchipunov, Colloid J. 68 (2006) 241. [30] F. Lequeux, Europhys. Lett. 19 (1992) 675. [31] J. Appell, G. Porte, A. Khatory, F. Kern, S.J. Candau, J. Phys. II France 2 (1992) 1045. [32] A. Khatory, F. Kern, F. Lequeux, J. Appell, G. Porte, N. Morie, A. Ott, W. Urbach, Langmuir 9 (1993) 933. [33] A. Khatory, F. Lequeux, F. Kern, S.J. Candau, Langmuir 9 (1993) 1456. [34] P.A. Hassan, S.J. Candau, F. Kern, C. Manohar, Langmuir 14 (1998) 6025. [35] R.D. Koehler, S.R. Raghavan, E.W. Kaler, J. Phys. Chem. B 104 (2000) 11035. [36] S.R. Raghavan, G. Fritz, E.W. Kaler, Langmuir 18 (2002) 3797. [37] D. Angelescu, A. Khan, H. Caldararu, Langmuir 19 (2003) 9155.