Unique lamellar lyotropic liquid crystal phases of nonionic phytosterol ethoxylates in glycerol

Abstract

As one type of biocompatible surfactant, phytosterol ethoxylates (BPS-n, n is the oxyethylene chain length) have attracted more and more attention for their characteristic molecular structure. In this paper, the aggregation behaviors of BPS-5 or BPS-10 with five or ten oxyethylene units were investigated in glycerol to explore the solvent molecular structure effect, especially at higher surfactant concentration regions. Surface tension measurements were adopted to analyze the thermodynamic process for micelle formation. Polarized optical microscopy and small-angle X-ray scattering were then used to identify and characterize the formed lyotropic lamellar liquid crystalline (LLC) phase structures. An interesting coexistence of two kinds of lamellar phases in the BPS-5/glycerol system was found, which was attributed to different hydrogen bonding interactions between BPS-5 and glycerol. More intermolecular interaction information was mapped through the rheological measurement and Fourier transformed infrared (FTIR) spectroscopy. The relatively high self-assembling capability of glycerol was recognized to result from its molecular structure with more hydroxyl groups. Its higher Gordon parameter benefitting from the stronger hydrogen-bonded networks made the glycerol exhibit obvious enhancement on the LLC phase formation of BPS-5 or BPS-10, compared to those in water or amide solvents (wider concentration range of LLC phase and higher viscoelasticity). Such a stronger organized solvent structure in glycerol might originate from the special three-dimensional hydrogen-bonding patterns and much higher lifetimes compared to those for water.

---

1. Introduction

Sterol ethoxylate surfactants have attracted much attention due to their good biocompatibility, low toxicity, and many applications in the fields of detergents, cosmetics, medicine and food industry, etc.¹ Because of the rigid sterol ring structures, they usually exhibit a much stronger segregation tendency between the hydrophilic and hydrophobic parts compared to the conventional alkyl ethoxylated nonionic surfactants. The aggregates formed in solutions could therefore exhibit certain characteristics of “rigidity”. Various aggregate morphologies including micelles,^2–4 vesicles,^5–8 and liquid crystals^9–16 have been observed in different solvents. In aqueous systems, Abe et al. have carried out systematic investigations on phase behaviors of phytosterol ethoxylate surfactants (BPS-n, n is an oxyethylene chain length).^17–21 They disclosed a network structure of micelles formed by polyoxyethylene cholesteryl ether and lauryl diethanolamide.¹⁸ By adding monoglyceride (monolaurin or monopalmitin) in aqueous solution, however, the temperature-responsive wormlike micelles could be obtained.²⁰ Folmer et al. have also studied the effect of polyoxyethylene chain length on BPS-n aggregation properties²² and found the decreased critical micelle concentration (CMC) values in water with the reduction of the hydrophilic oxyethylene chain length.²³ In aqueous solution of a short chain polyoxyethylene cholesterol ether (ChEO₃), Kunieda et al. observed the formation of a lamellar phase (L_α) when the concentration was above 75%, which could be re-dispersed into vesicles in excess water.²⁴ Richer morphologies were then mapped in BPS-15/H₂O system with the successively formation of micellar, micellar cubic, hexagonal, lamellar, and solid phases with increasing concentration of BPS-15.²⁵

To broaden the solvent media for BPS-n self-assembling and explore more potential properties of their aggregates, researchers have also paid their attention to non-aqueous systems of BPS-n. In room temperature ionic liquid (RTIL) of 1-butyl-3-methylimidazolium hexafluorophosphate ([Bmim]PF₆), Abe et al. studied the phase behavior of BPS-n with different PEO chains.²⁶ Compared to the BPS-n/H₂O system, the notable difference in phase types was found and higher BPS-n concentrations were required to induce phase transitions. We have also compared the aggregation behaviors of BPS-10 in two ionic liquids, a protic ethylammonium nitrate (EAN) and an aprotic 1-butyl-3-methylimidazolium tetrafluoroborate ([Bmim]BF₄).^27,28 The lyotropic liquid crystal (LLC) phase regions in EAN were found to appear at lower surfactant concentrations than those in [Bmim]BF₄. Instead of only one L_α phase formation in [Bmim]BF₄, an extra hexagonal phase (H₁) was observed in EAN due to its higher Gordon parameter and stronger hydrogen-bonded network. Therefore, the protic solvent behaves more effective than the aprotic one as the self-assembling media. This point has been verified in other polar organic solvents with water-like structures, like glycerol, formamide, and ethylene glycol.^29–32 In our study on BPS-10 phase behaviors using three organic amide compounds (FA, NMF and DMF) as solvents, the H₁ and L_α phases could also be obtained due to the existence of similar hydrogen-bonded networks.³³

As a continuation of our research in this paper, we extended the study of BPS-n aggregation behavior to a new solvent medium, the glycerol. This is not only because of its excellent solvent characteristics of a high cohesive energy or dielectric constant, but also due to its potential applications in pharmaceutical and cosmetic products, where it is usually used as co-solvent to improve or modulate the surfactant aggregations.^34–40 For examples, in the study of self-assembly of alkyl-propoxy-ethoxylate surfactants, Alexandridis et al. found the increased surfactant solubility with introducing such alcohol cosolvents into water and thus an increased CMC.³⁴ The liquid crystalline phases (hexagonal and lamellar), isotropic and gel-like phases could also be obtained in glycerol-containing solvent system as reported by Dorfler et al.³⁷ for long-chain carboxylic acid potassium (K-soap) and by Auvray et al. for ionic surfactants in polar solvents.^38,39 Usually, only the lamellar phase was formed in systems with glycerol or ethylene glycol as solvents. Therefore, it is expected to explore the phase behaviors of BPS-n surfactants in such alcohol solvents. The capabilities of adjusting such low toxic and environmentally friendly surfactant interactions using these non-aqueous polar solvents should practically lead to extension of their applications in the pharmaceuticals and cosmetics formulations.

2. Experimental

2.1. Materials

The nonionic ethoxylated phytosterol surfactants BPS-n (n = 5 and 10) with the chemical structures shown in Fig. 1, were supplied as kindly gifts by Nikko Chemicals from Japan. The hydrophobic portion of the BPS-n surfactants was consisted of β-sitosterol, campesterol, and stigmasterol in a ratio of 2 : 1 : 1 and less than 3 wt% polyoxyethylene was contained as a hydrophilic impurity. Glycerol (A.R. grade) was purchased from Sinopharm Chemicals Reagent Co. Both the surfactants and glycerol were used without further purification.

Fig. 1 Chemical structure of studied ethoxylated phytosterol surfactants BPS-n (n = 5, 10).

2.2. Sample preparation

All samples were prepared by mixing the BPS-n surfactant and glycerol with exact mass ratios. The mixtures were heated to about 80 °C and stirred until the samples became completely transparent without any un-dissolved ingredients. After several cycles of heating, vortex-mixing and centrifugation, the samples were fully mixed and then equilibrated at 25 °C for more than three weeks.

2.3. Characterization

2.3.1. Surface tension measurement. The Wilhelmy plate method was used to measure surface tension by a surface tensiometer QBZY-2 (Shanghai fangrui Co. Ltd.) with an accuracy of ±0.01 mN m⁻¹ at a temperature of 25.0 ± 0.1 °C. All measurements were repeated at least twice until the uncertainty of the surface tension value was less than ±0.2 mN m⁻¹.

2.3.2. Polarized optical microscopy (POM) measurement. The textures of formed LLC phases were taken by a Motic B2 polarized optical microscope (POM) with a CCD camera (Panasonic Super Dynamic II WVCP460). Samples at high concentrations were firstly heated to their isotropic status and then cooled to observation temperature.

2.3.3. Small-angle X-ray scattering (SAXS) measurement. The structural characterizations of LLC phases were performed using a SAXSess MC2 high flux small-angle X-ray scattering instrument (Anton Paar, Austria) with a Ni-filtered Cu Kα radiation (0.154 nm), operating at 40 kV and 50 mA. The distance between the sample and detector was 27.8 cm. A standard temperature control unit (Anton-Paar TCS 120) connected with the SAXSess was used to control the temperature at 25 °C.

2.3.4. Rheology measurement. The LLC sample rheological properties were measured at 25.0 ± 0.1 °C on a Haake Rheostress 6000 rheometer equipped with a cone-plate geometry (diameter of 25 mm with cone angle of 0.1 rad). The oscillatory experiments were carried out to determine the viscoelasticities of all samples within the linear viscoelastic region.

2.3.5. Fourier transformed infrared (FTIR) spectroscopy measurement. FTIR spectra were recorded from 400 to 4000 cm⁻¹ with a resolution of 4 cm⁻¹ using an Alpha-T spectrometer (Bruker). The LLC samples were directly coated on a dried KBr plate.

3. Results and discussion

3.1. Micellization behavior of BPS-n in glycerol

As a polar organic solvent, the glycerol exhibited good self-assembling capability, which could be certified by the CMC value for a surfactant determined by the surface tension (γ) measurement. The measured γ profiles with concentrations (weight percentage wt%) for both BPS-5 and BPS-10 in glycerol were shown in Fig. 2. At low concentrations, γ was decreased gradually with increasing concentration until a plateau was reached with a sharp break point, indicating the interfacial saturation of surfactant and the initialization of micelle formation. The CMC values were assigned from the break points with the results of about 4.0 × 10⁻³ wt% (7.5 × 10⁻⁵ mol L⁻¹) and 5.6 × 10⁻³ wt% (9.2 × 10⁻⁵ mol L⁻¹) respectively for BPS-5 and BPS-10. Both CMC values in glycerol were larger than that of BPS-10 in water (∼1.0 × 10⁻⁵ mol L⁻¹),²³ but lower than those in EAN or [Bmim]BF₄ (∼3 × 10⁻⁴ mol L⁻¹),²⁸ reflecting the better capability of glycerol as self-assembling medium than ionic liquids. Meanwhile, the similar CMC values of BPS-5 and BPS-10 indicated that the solvophobic interaction should dominate the aggregate formation because they were of the same solvophobic parts. The a little larger CMC of BPS-10 than BPS-5 was perhaps due to its longer solvophilic EO chains.

Fig. 2 Surface tension profiles of BPS-5 and BPS-10 in glycerol.

From the surface tension measurements, several other thermodynamic parameters, including the effectiveness of the surface tension reduction (Π_CMC), the surface excess at the air/glycerol interface (Γ_max) and the minimum occupied area per surfactant molecule adsorbed at the interface (A_min), were also obtained based on the following equations:

Π_CMC = γ₀ − γ_CMC (1)

where γ₀ and γ_CMC are the surface tensions of solvent and the equilibrium solution at CMC. T is the absolute temperature and dγ/dln C is the slope of γ vs. ln C profile when the concentration is near CMC. N_A is Avogadro's number, and R (= 8.314 J mol⁻¹ K⁻¹) is the gas constant. The calculated data were all listed in Table 1. Little difference between BPS-5 and BPS-10 were found for values of γ_CMC and Π_CMC, reflecting their similar surface activities. The increased Γ_max and decreased A_min values of BPS-10, however, suggested an enhanced solvophilic interaction with longer EO chains. In addition, this A_min of BPS-10 in glycerol was even smaller than that in EAN,²⁸ indicating a more dense packing of molecules between glycerol solvent and EO chains, which should be therefore more extended in glycerol.

Table 1 Surface properties of BPS-5 and BPS-10 in glycerol at 25 °C

	CMC (wt%)	γCMC (mN m^−1)	ΠCMC (mN m^−1)	Γmax (μmol m^−2)	Amin (Å^2)
BPS-5	4.0 × 10^−3	23.2	40.1	7.19	23.1
BPS-10	5.6 × 10^−3	24.6	38.7	10.08	16.47

---

The behavior of glycerol as a solvent here was consistent with its self-assembling capability as indicated by Gordon parameter, which could be used to measure the cohesive energy density of solvent. Generally, the higher the G values, the stronger the driving force for the self-assembly process. For water, glycerol, EAN and [Bmim]BF₄, the Gordon parameter values are respectively 2.8, 1.51, 1.4 and 0.79 J m⁻³.⁴¹ The existed data difference reflects the ordering difference between these solvent bulk structures. Considering the protic nature of these solvents, it was also an indication for the ordering of the hydrogen bonding network formed inside the bulk solvent. At higher surfactant concentrations, however, more complicated intermolecular actions might reduce such differences as reflected by following results.

3.2. LLC formation in BPS-n/glycerol systems

3.2.1. BPS-5. At concentrations higher than CMC, the BPS-5 micelles in glycerol would aggregate further into more organized assemblies. For convenience of discussion, the BPS-5 concentration (C_BPS-5) was expressed in its mass percentage (wt%) with 5% corresponding to about 0.1 mol L⁻¹. Fig. 3 presents the BPS-5 sample appearance at four concentrations. As can be seen that, far below C_BPS-5 = 5%, only a transparent fluid existed with no birefringence under the crossed polarizers (Fig. 3a). When C_BPS-5 was increased up to 5–20%, the sample began to separate into two distinct phases, a transparent fluid at the bottom and a viscous gel-like phase of birefringence on the top. Such a phase separation became more obvious at C_BPS-5 higher than 20% and the up phase proportion expanded gradually (Fig. 3b and c). When the BPS-5 concentration reached 50% or more, the sample became uniformly sticky and exhibited a very strong birefringence with no fluidity (Fig. 3d).

Fig. 3 Preliminary phase behavior observation at different BPS-5 concentrations (C_BPS-5) in glycerol without (top) and with polarizers (bottom). From (a) to (d), C_BPS-5 = 0.05, 20, 40, and 60 wt%, respectively.

The POM and SAXS measurements were then employed to identify the phase structures of the samples at higher concentrations (C_BPS-5 > 50%). From the POM photographs shown in Fig. 4, the “oily streaks” textures could be observed. The upper anisotropic one in two-phase separated samples at C_BPS-5 = 25% and 40% were also shown in Fig. S1† with the “Maltese cross” and “oily streaks” textures, which usually indicated the formation of lamellar (L_α) liquid crystalline phase.⁴²

Fig. 4 Polarized optical micrographs for BPS-5 in glycerol at different concentrations at 25 °C. From (a) to (c), C_BPS-5 = 50%, 70% and 90%.

Such possible L_α phase structures were further characterized by their SAXS profiles. As shown in Fig. S1† for upper anisotropic one in two-phase separated samples at C_BPS-5 = 25% and 40%, three Bragg scattering peaks with the relative scattering factor q positions of 1 : 2 : 3 ratio could be obtained, indicating definitely the formation of L_α phases.^43,44 At concentrations higher than 50% in the uniform single phase samples, we still observed three or more Bragg scattering peaks with their relative scattering factor q positions of 1 : 2 : 3 ratio corresponding to the formation of L_α phase as seen in Fig. 5a. Based on the Bragg equation, d = 2π/q₁, the repeated lamellae distance (d) could be derived. When the C_BPS-5 was increased from 50% to 90%, the d values were contracted from 8.02 to 6.02 nm.

Fig. 5 SAXS patterns for uniform single phase of BPS-5 in glycerol at different concentrations (a) and at four concentrations with two sets scattering peaks (b) at 25 °C.

It was interested that two sets of scattering peaks were observed in the sample at C_BPS-5 = 70% (Fig. 5a). Further SAXS characterization at concentrations around 70% were then carried out with the results shown in Fig. 5b. Though not so obvious, it could be still noted that the separation of Bragg peaks was initiated at 65%, especially at the 2^nd peak. In both sets of scattering peaks, their relative scattering factor q positions were still of 1 : 2 : 3 ratios, indicating a coexistence of two kinds of lamellar phases in this concentration region. By carefully checking the 1^st peaks in these profiles, the scattering factors in set 1 were closer to those of lamellar phase before the occurrence of another set SAXS pattern. While the scattering factors in set 1′ were closer to those of lamellar phases at higher concentrations larger than 80%. The relatively big difference between the lamellae repeat spacings (d) at 60% (7.50 nm) and 80% (6.66 nm) provided certain clue to explain this phenomenon. The change of d value with the volume fraction of solvophobic component was plotted according to the method described previously²⁷ with the result shown in Fig. S2.† A disruption was found at C_BPS-5 around 70%. Therefore, the molecular packing inside both lamellar phases should be in different manners. Considering the existence of three-dimensional hydrogen-bonded network in glycerol,⁴⁵ and the possible hydrogen bondings either between glycerol and PEO chains in BPS-5 or just between PEO chain themselves, the lamellar phase at C_BPS-5 less than 60% should be organized with more glycerol incorporated between PEO chains, which were then more extended. With the decreasing of glycerol amount, the interaction between PEO chains was enhanced to induce them more interdigitated and therefore an obvious reduction of d value.

3.2.2. BPS-10. Like BPS-5 in glycerol, the BPS-10 molecules exhibited similar phase behaviors with several typical POM and SAXS results shown in Fig. S3.† Because of the similar CMC values, the BPS-10 samples presented also as transparent fluids at C_BPS-10 below 5% and no birefringence was observed under the crossed polarizers. However, the two-phase (L_α and L₁) separation region was a little larger here and appeared at C_BPS-10 from 5% to 60%. The uniform and gel-like samples of very strong birefringence existed at C_BPS-10 above 60%. The repeated lamellae distance (d) was calculated as 7.31 nm at C_BPS-10 of 90%, larger than that of BPS-5 (6.02 nm) due to longer PEO chains. It was noted that, however, no lamellar phase coexistence or transformation was observed in BPS-10 systems. This should be attributed to longer PEO chains here which dominated the lamellae structure formation at all concentrations.

To get more detailed structural information of the lamellar phase, the volume fraction of solvophobic components (Φ_a), the solvophobic domain thickness (d_c), the solvophilic thickness (d_g), and the effective cross-sectional area per molecule at the surfactant/glycerol interface (A_S) were calculated based on the following equations and relations,^24,43

where W_S (ρ_S) and W_g (ρ_g) were respectively the weights (densities) of surfactant and glycerol. M_S was the molecular weight of surfactant, i.e. 634.97 g mol⁻¹ for BPS-5 and 855.23 g mol⁻¹ for BPS-10. V_a was the molar volume of the solvophobic part. N_A was the Avogadro's number. The structure parameters were thus calculated and listed in Table 2.

Table 2 Lamellar structure parameters for BPS-5 (BPS-10) in glycerol at 25 °C

CBPS-n (wt%)	q1 (nm^−1)		d (nm)		Φa (vol%)		dc (nm)		dg (nm)		AS (nm^2)
60%	0.837	(0.734)	7.50	(8.56)	0.41	(0.32)	1.53	(1.37)	4.44	(5.82)	0.42	(0.47)
70%	0.870	(0.798)	7.22	(7.87)	0.47	(0.37)	1.69	(1.46)	3.84	(4.95)	0.38	(0.44)
80%	0.943	(0.828)	6.66	(7.58)	0.52	(0.41)	1.75	(1.56)	3.16	(4.46)	0.36	(0.41)
90%	1.043	(0.859)	6.02	(7.31)	0.58	(0.46)	1.75	(1.68)	2.53	(3.95)	0.36	(0.38)

---

The change profile of d with Φ_a in BPS-5 system had been plotted in Fig. S2,† which presented two approximately linear relationship between log(d) and log(Φ_a), indicating two lamellar structures as discussed before. For BPS-10, such a profile behaved more like only one linear relationship for a one-dimensional swelling scheme^44,46 as shown in Fig. S4,† consistent with a single lamellar structure in LLC region as discussed before. Changes of three other structural parameters with the volume fraction of glycerol (Φ_g) were also plotted and shown in Fig. S5.† Only little changes for d_c and A_S values could be seen. The a little higher solvophobic domain thickness (d_c) in BPS-5 system and a little bigger solvophilic thickness (d_g) in BPS-10 case reflected respectively an enhanced solvophobic volume in BPS-5 molecular packing and an increased solvophilic interactions due to the longer EO chains in BPS-10.⁴⁷ In addition, the reduced A_S and d values compared to those in EAN⁴⁸ confirmed a more dense packing of BPS-10 in glycerol, which was consistent with the higher Gordon parameter of glycerol than EAN.

3.3 Rheological properties of LLC in BPS-5 and BPS-10 systems

To get more information on molecular interactions inside the aggregate, the viscoelastic properties of formed LLC phases at different concentrations were studied by rheological measurements. The obtained results indicated a shear-thinning rheology behavior for the L_α phase of BPS-5 at higher concentrations (50% to 90%). The apparent viscosity was increased with increasing C_BPS-5 and higher than those of the conventional nonionic surfactant like C_nEO_m,^49,50 at low shear rates, which might be attributed to the stronger interactions between steroid rings.

Fig. 6 presented the results from the dynamic oscillatory experiments. As can be seen that, both the storage (G′) and loss (G′′) moduli were increased with increasing the sweep frequency, indicating a viscoelastic liquid character.³⁴ Meanwhile, the complex viscosity (η*) values were decreased with elevating the sweep frequency. All these phenomena were characteristic for the formation of L_α phase of BPS-5 in glycerol.⁵¹ It could be noted also in Fig. 6a that for the sample at C_BPS-5 of 60%, the G′ was higher in the low frequency region but lower in the high frequency region compared to G′′. However, when C_BPS-5 was increased to 80% or higher, G′ was always larger than G′′ at the whole measured frequency range, indicating only an elastic behavior, which was characteristic for gel-like aggregate.^52,53 Meanwhile, the complex viscosity was also enhanced as seen in Fig. 6b. All these results reflected a concentration induced denser packing of BPS-5 molecules in glycerol, which was in accordance with the deduction from SAXS results.

Fig. 6 Variation of storage and loss moduli and the complex viscosity as a function of shear frequency at different C_BPS-5: (a) 60% and (b) 80%.

Similar rheological properties were observed for the L_α phase samples of BPS-10 in glycerol with the results shown in Fig. S6 and S7.† But, the apparent viscosity of BPS-10 L_α phase was lower than that of BPS-5 at high concentrations, denoting a looser packing of BPS-10 and being consistent with the increased A_S values. However, these viscosity values were still about ten times higher than those of BPS-10 in EAN and [Bmim]BF₄,²⁸ and the moduli values in glycerol were also much higher than those in both ionic liquids and water.^23,28 The reason should be attributed to the stronger interactions between BPS-10 and glycerol.

Based on the results discussed above, the preliminary phase diagrams of BPS-n/glycerol binary systems with five or ten oxyethylene units could be mapped and shown in Fig. 7. At lower concentrations, micelles coexisted with the L_α phase in both binary systems. With increasing BPS-n concentration, the uniform sticky L_α phases could be formed in both BPS-5 and BPS-10 systems. There existed two kinds of lamellar phases (L_α and L′_α) at high concentrations of BPS-5, while only one lamellar phase could be formed for BPS-10 in glycerol at high concentrations. It was noted that the initial concentration for lamellar phase appearance was reduced in BPS-n/glycerol binary systems in compared with those LLC systems reported for BPS-n in water or [Bmim]BF₄,^23,27 indicating the unusual self-assembling capability presented in glycerol.

Fig. 7 Phase diagram for BPS-n (n = 5, 10)/glycerol binary systems at 25 °C. 2Φ denotes a two-phase coexistence.

3.4 Molecular interactions between BPS-n and glycerol

What causes the presence of two L_α phases of BPS-5 in glycerol and what is the origin of the good self-assembling capability of glycerol? To answer these questions, the main driving forces during the LLC phase formation should be depicted. In the light of the unique structural character of BPS-n molecules with the solvophilic PEO block and the solvophobic steroid nucleus, these molecules should adopt well-ordered molecular arrangements inside the aggregates. In our previous work, the strong solvophobic interactions between bulky ring groups have been demonstrated as the main driving force on the formation of L_α phase in BPS-n/ILs system, and the hydrogen bonding between PEO blocks and ILs was another important driving force.^27,54

Using glycerol as solvent, however, the contribution from hydrogen bonding became more obvious because of the presence of more hydroxyl groups in glycerol. In addition, the H-bonding mean lifetimes of glycerol which was even longer than that in water,^47,55 might induce the stronger solvophobic effect of BPS-n than in water. Thus, we could speculate that the hydrogen bonding between PEO block and glycerol molecules was another source of driving force during the formation of LLCs. This point could be confirmed by the FTIR spectra for BPS-5/glycerol system with the results shown in Fig. 8. In the BPS-5/glycerol binary system, the stretching vibrations of the hydroxyl groups at 3528 cm⁻¹ in pure glycerol and the C–O–C stretching vibration of PEO groups at 1109 cm⁻¹ were both shifted to lower wavenumbers. These frequency variations denoted the formation of hydrogen bonding between the hydroxyl group of glycerol solvent and the oxygen atom of PEO blocks.⁴² Similar FTIR spectra changes were obtained for BPS-10 in glycerol with the results shown in Fig. S8.† Based on the facts that the L_α phase repeat distances at the higher concentration like 90% (6.02 or 7.31 nm for BPS-5 or BPS-10 systems) were smaller than their corresponding bilayer thicknesses (5.28 × 2 or 6.87 × 2 for BPS-5 or BPS-10),⁴⁸ it was reasonable to suppose an interdigitated antiparallel model (shown in Fig. S9†) for the molecular packing of BPS-n in L_α phase, where only PEO chains were overlapped with each other via hydrogen bonds between the hydroxyl group of glycerol and the oxygen atom of PEO blocks.

Fig. 8 FTIR spectra for glycerol (a) L_α phase (C_BPS-5 = 80%) (b) and BPS-5 (c), respectively.

Then, it was a little strange that only the L_α phase was formed in BPS-n/glycerol system. With the Gordon parameter value close to that of formamide and higher than that of EAN, the BPS-n surfactants should form richer morphological aggregates in glycerol but not only the lamellar one. This fact suggested that the moderate solvophilic head-solvent interaction and geometric constraints were other important factors to control the final aggregate morphology.³⁹

4. Conclusions

The aggregation behaviors of two phytosterol ethoxylate surfactants (BPS-5 and BPS-10) in glycerol were investigated in this work. Both of them formed micelles with their CMC values in between those of water and ionic liquids (EAN or [Bmim]BF₄), reflecting a better capability of glycerol as self-assembling medium than ionic liquids. The reason was due to its similarity to water and EAN on the ability to donate protons to form hydrogen bonding networks. The L_α phases were formed from BPS-5 and BPS-10. However, BPS-n in glycerol exhibited better aggregation behaviors compared to those in water or ionic liquids, which might be attributed to their stronger hydrogen-bonded network and much longer hydrogen bonding lifetime. The main driving forces on the formation of the L_α phases were the interactions of the steroid rings and hydrogen bonding effects. The geometric constraints and head-solvent interactions were demonstrated to be the other factors. For BPS-5 system, the delicate balance between these interactions induced two kinds of lamellar phases (L_α and L′_α) at concentrations around 70%. The deductions and conclusions derived here should be useful supplements to the aggregation behaviors of steroid surfactant in non-aqueous solvents and therefore extend their applications in water-free self-assemblies, which will increase the potential applications of BPS-n surfactant in the future.

Acknowledgements

We are thankful for the financial supports from the National Natural Science Foundation of China (20973104 and 21373127) and the Specialized Research Fund for the Doctoral Program of Higher Education (20130131110010).

References

1. Novel Surfactants: Preparation Applications and Biodegradability, ed. K. Holmberg, Marcel Dekker, Inc., New York, 2nd edn, 2003.
2. B. M. Folmer and M. Nyden, Langmuir, 2008, 24, 6441.
3. K. Miyajima, T. Lee and M. Nakagaki, Chem. Pharm. Bull., 1984, 32, 3670.
4. D. Meissner, I. Grassert, G. Oehme, G. Holzhüter and V. Vill, Colloid Polym. Sci., 2000, 278, 364.
5. C. Müller-Goymann, Pharm. Res., 1984, 1, 154.
6. I. F. Uchegbu, D. McCarthy, A. Schätzlein and A. T. Florence, STP Pharma Sci., 1996, 6, 33.
7. S. Beugin, K. Edwards, G. Karlsson, M. Ollivon and S. Lesieur, J. Biophys., 1998, 74, 3198.
8. J. Bouwstra, D. van Hall, H. Hofland and H. Junginger, Colloids Surf., A, 1996, 178, 263.
9. E. F. Marques, H. Edlund, C. L. Mesa and A. Khan, Langmuir, 2000, 16, 5178.
10. H. Amenitsch, H. Edlund, A. Khan, E. F. Marques and C. La Mesa, Colloids Surf., A, 2003, 213, 79.
11. H. Edlund, A. Khan and C. L. Mesa, Langmuir, 1998, 14, 3691.
12. H. Soderlund, J. Sjoblom and T. Warnheim, J. Dispersion Sci. Technol., 1989, 10, 131.
13. D. Libster, A. Aserin, D. Yariv, G. Shoham and N. Garti, Colloids Surf., B, 2009, 74, 202.
14. T. Teshigaware, R. Miyahara, T. Fukuhara and T. Oka, J. Oleo Sci., 2009, 58, 27.
15. H. Sakai, T. Saitoh, T. Endo, K. Tsuchiya, K. Sakai and M. Abe, Langmuir, 2009, 25, 2601.
16. K. Sakai, Y. Onuma, K. Torigoe, S. Biggs, H. Sakai and M. Abe, Langmuir, 2011, 27, 3244.
17. S. C. Sharma, L. K. Shrestha, K. Tsuchiya, K. Sakai, H. Sakai and M. Abe, J. Phys. Chem. B, 2009, 113, 304.
18. R. G. Shrestha, L. Abezgauz, D. Danino, K. Sakai, H. Sakai and M. Abe, Langmuir, 2011, 27, 12877.
19. R. G. Shrestha, S. C. Sharma, K. Sakai, H. Sakai and M. Abe, Colloid Polym. Sci., 2012, 290, 339.
20. S. C. Sharma, L. K. Shrestha, K. Sakai, H. Sakai and M. Abe, Colloid Polym. Sci., 2010, 288, 405.
21. R. G. Shrestha, K. Sakai, H. Sakai and M. Abe, J. Phys. Chem. B, 2011, 115, 2937.
22. B. M. Folmer, M. Svensson, K. Holmberg and W. Brown, J. Colloid Interface Sci., 1999, 213, 112.
23. B. M. Folmer, J. Colloid Interface Sci., 2003, 103, 99.
24. C. Rodriguez, N. Naito and H. Kunieda, Colloids Surf., A, 2001, 181, 237.
25. M. K. Hossain, D. P Acharya, T. Sakai and H. Kunieda, J. Colloid Interface Sci., 2004, 277, 235.
26. H. Sakai, T. Saitoh, T. Misono, K. Tsuchiya, K. Sakai and M. Abe, J. Oleo Sci., 2012, 61, 135.
27. X. Yue, X. Chen, X. D. Wang and Z. H. Li, Colloids Surf., A, 2011, 392, 225.
28. X. Yue, X. Chen and Q. T. Li, J. Phys. Chem. B, 2012, 116, 9439.
29. R. Palepu, H. Gharibi, D. M. Bloor and E. Wyn-Jones, Langmuir, 1993, 9, 110.
30. C. Carnero Ruiz, J. A. Molina-Bolívar, J. Aguiar, G. MacIsaac, S. Moroze and R. Palepu, Langmuir, 2001, 17, 6831.
31. C. Seguin, J. Eastoe, R. Clapperton, R. K. Heenan and I. Grillo, Colloids Surf., A, 2006, 282–283, 134.
32. K. M. Glenn, S. Moroze, S. C. Bhattacharya and R. M. Palepu, J. Dispersion Sci. Technol., 2005, 26, 79.
33. X. Yue, X. Chen, Q. T. Li and Z. H. Li, Langmuir, 2013, 29, 11013.
34. B. Sarkar, S. Lam and P. Alexandridis, Langmuir, 2010, 26, 10532.
35. J. Penfold, E. Staples, I. Tucker and P. Cummins, J. Colloid Interface Sci., 1997, 185, 424.
36. M. A. EI-nokaly, L. D. Ford and S. E. Friberg, J. Colloid Interface Sci., 1981, 84, 228.
37. H. D. Dorf1er and A. Senst, Colloid Polym. Sci., 1993, 271, 173.
38. X. Auvray, C. Petipas, R. Anthore, I. Rico and A. Lalles, J. Phys. Chem., 1989, 93, 7458.
39. X. Auvray, T. Perche, R. Anthore, C. Petipas, I. Rico and A. Lattes, Langmuir, 1991, 7, 2385.
40. G. D’Errico, D. Ciccarelli and O. Ortona, J. Colloid Interface Sci., 2005, 286, 747.
41. T. L. Greaves and C. J. Drummond, Chem. Soc. Rev., 2008, 37, 1709.
42. F. B. Rosevear, J. Am. Oil Chem. Soc., 1954, 31, 628.
43. H. Kunieda, G. Umizu and Y. Yamaguchi, J. Colloid Interface Sci., 1999, 218, 88.
44. G. D. Zhang, X. Chen, Y. R. Zhao, Y. Z. Xie and H. Y. Qiu, J. Phys. Chem. B, 2007, 111, 11708.
45. J. A. Padro, L. Saiz and E. Guardia, J. Mol. Struct., 1997, 416, 243.
46. P. Ekwall, L. Mandell and K. Fontell, J. Colloid Interface Sci., 1969, 29, 639.
47. R. Atkin, S. M. C. Bobillier and G. G. Warr, J. Phys. Chem. B, 2010, 114, 1350.
48. M. A. Lopez-Quintela, A. Akahane, C. Rodriguez and H. Kunieda, J. Colloid Interface Sci., 2002, 247, 186.
49. F. M. Ma, X. Chen, Y. R. Zhao, X. D. Wang, Q. H. Li, C. Lv and X. Yue, Langmuir, 2010, 26, 7802.
50. P. Versluis, J. C. van de Pas and J. Mellema, Langmuir, 1997, 13, 5732.
51. R. Mezzenga, C. Meyer, C. Servais, A. I. Romoscanu, L. Sagalowicz and R. C. Hayward, Langmuir, 2005, 21, 3322.
52. T. Durrschmidt and H. Hoffmann, Colloid Polym. Sci., 2001, 279, 1005.
53. J. Zhang, A. X. Song, Z. B. Li, G. Y. Xu and J. C. Hao, J. Phys. Chem. B, 2010, 114, 13128.
54. L. Y. Wang, X. Chen, Y. C. Chai, J. C. Hao, Z. M. Sui, W. C. Zhuang and Z. W. Sun, Chem. Commun., 2004, 24, 2840.
55. E. Guardia, J. Marti, J. A. Padro, L. Saiz and A. V. Komolkin, J. Mol. Liq., 2002, 96–97, 3.

---

Footnote

† Electronic supplementary information (ESI) available: The aggregation behaviors of BPS-10 in glycerol and more structural information of the lamellar phase of BPS-5 were characterized by POM, SAXS, FTIR and rheology measurement with the results shown in supporting information. See DOI: 10.1039/c5ra21446g

---