pubs.acs.org/Langmuir © 2009 American Chemical Society
Complex Hydrogel Systems Composed of Polymers, Liposomes, and Cyclodextrins: Implications of Composition on Rheological Properties and Aging† Spyridon Mourtas,‡ Christos A. Aggelopoulos,§ Pavlos Klepetsanis,‡,§ Christos D. Tsakiroglou,§ and Sophia G. Antimisiaris*,‡,§ ‡
Laboratory of Pharmaceutical Technology, Department of Pharmacy, School of Health Sciences, University of Patras, 26510 Rio, Greece, and §Foundation for Research and Technology Hellas, Institute of Chemical Engineering and High Temperature Chemical Processes, 26504 Rio, Greece Received December 30, 2008. Revised Manuscript Received April 24, 2009
Rheological properties of complex hydrogels containing different amounts of liposomes and/or cyclodextrin (CD) were evaluated. Sonicated unilamellar vesicles (SUV) were loaded in a hydrogel composed of Carbopol 974 NF and hydroxyethylcellulose (Natrosol 250 HX). Phosphatidylcholine (PC) and hydrogenated-PC (HPC) liposomes, both mixed with cholesterol in a 2:1 lipid/chol mol ratio, were used. In some cases, hydroxypropyl-β-cyclodextrin was also added (100 or 400 mg/mL). Gels were incubated at 40 C/75% humidity for 7 days or 1 month to evaluate the effect of aging on their rheological properties. FTIR and DSC studies were performed to investigate possible interactions between the polymers and CD molecules at different CD concentrations. Static and dynamic rheological measurements were carried out. All gels had shear-thinning behavior (fitted well by the Cross model) with the exception of gels containing high concentrations of CD that were transformed into nonflowing elastic sticky solids, especially after aging. The more pronounced elastic behavior of gels containing 400 mg/mL CD is reflected by the higher values of relaxation strengths over all relaxation times. Complete interaction between polymers and CD, in the high-CD-content gels, as proven by FTIR and DSC studies, explains the dominating contribution of CD on gel characteristics. The addition of liposomes to such CD-containing gels has a substantial effect on their rheological properties, which are dependent on the liposome type (HPC/chol liposomes > PC/chol) and the lipid/CD ratio. This is explained by the “neutralization” of some CD molecules that prefer to interact with chol molecules that they extract from the lipid membranes. Gels with a high CD concentration (400 mg/mL) are almost insensitive to aging, whereas all other gels become slightly more elastic and less viscous as aging proceeds.
1. Introduction When mucosal or topical (especially vaginal) delivery of liposomal formulations is considered, the rheological and/or mucoadhesive properties of liposomes should be adjusted.1 This can be managed by adding gelling agents to liposome dispersions, in which case complex drug-in-liposome-in-gel formulations are formed.2-5 Such gelling agents can be polymer blends consisting of Carbopol 974 and hydroxyethylcellulose (Natrosol).6 Carbopol 974, an acrylic acid-based polymer, and hydroxyethylcellulose (HEC), a cellulose-based polymer, the structures of which are presented in Figure 1, are the main components of many semisolid formulations (commercially available or under preclinical evaluation). It was recently proposed that mixtures of the above two polymer types have improved rheological properties for the vaginal administration of drugs.6 Indeed, such mixture gels were found to be more stable toward temperature and pH changes † Part of the Molecular and Polymer Gels; Materials with Self-Assembled Fibrillar Networks special issue. *Corresponding author. Tel: 0030-2610-969332. Fax: 0030-2610-996302. E-mail:
[email protected].
(1) Kieweg, S. L.; Katz, D. F. J. Pharm. Sci. 2007, 96, 835–85. (2) Moldovan, M.; Leucuta, S. E.; Bakri, A. J. Drug Delivery Sci. Technol. 2006, 16, 127–132. (3) Mishra, V.; Mahor, S.; Rawat, A.; Dubey, P.; Gupta, P. N.; Singh, P.; Vyas, S. P. Vaccine 2006, 24, 5559–5570. (4) Boulmedarat, L.; Grossiord, J. L.; Fattal, E.; Bochot, A. Int. J. Pharm. 2003, 254, 59–64. (5) Pavelic, Z.; Skalko-Basnet, N.; Filipovic-Grcic, J.; Martinac, A.; Jalsenjak, I. J. Controlled Release 2005, 106, 34–43. (6) Wang, Y.; Lee, C. H. Contraception 2002, 66, 281–287.
8480 DOI: 10.1021/la804305z
compared to gels composed of each polymer alone when used individually.7 For the above stated reasons, we chose to use this specific type of mixture gel in our study. It was recently demonstrated that when liposomes are dispersed in such gels they are protected from the disruptive effects of specific excipients because of the higher viscosity of the gel dispersion compared to that of aqueous dispersions, which prevents (or delays) contact between the various components of the formulation.8 Nevertheless, the rigidity of liposomal membranes determines their integrity,8 and the release of liposomal drugs from drug-in-liposome-in-gel complex systems is determined by different factors according to the physicochemical properties of each drug.9 Indeed, hydrophilic drug release is retarded when rigid membrane liposomes are used, but release is not affected by the amount of lipid loaded into the gels. Oppositely, in the case of amphiphilic/lipophilic drugs (that have the ability to permeate the lipid membrane), the drug release rate is primarily determined by the amount of lipid loaded into the gel. When a large amount of drug (compared to its aqueous solubility) is loaded into the gel, the drug is released at a constant rate that is not affected by the liposome type and is primarily determined by the solubility of the drug in the aqueous environment. In the later (7) Owen, D. H.; Peters, J. J.; Lavine, M. L.; Katz, D. F. Contraception 2003, 67, 57–64. (8) Mourtas, S.; Fotopoulou, S.; Duraj, S.; Sfika, V.; Tsakiroglou, C.; Antimisiaris, S. G. Colloids Surf., B 2007, 55, 212–221. (9) Mourtas, S.; Duraj, S.; Fotopoulou, S.; Antimisiaris, S. G. Colloids Surf., B 2008, 61, 270–276.
Published on Web 06/04/2009
Langmuir 2009, 25(15), 8480–8488
Mourtas et al.
Article
drug-binding CDs to the aqueous phase of such formulations.15,16 It was recently demonstrated that the presence of liposomes in such polymer blend gels modifies their rheological properties, which are greatly affected by the lipid composition of the liposomes used.17 Thereby, it is important to know if the addition of another component to the gels, such as CDs, will cause additional changes to their rheological profile. Furthermore, it is important to evaluate the effect of aging on the rheology of such complex systems, especially if they are intended for commercial use. In particular, aging may have important implications on the characteristics of such gels because the occurrence of interactions between their main components (polymers, lipids, and CDs) has been previously reported.18-22 Herein, we investigate the effect on a hydrogel’s rheological properties of adding increasing amounts of CDs and/or liposomes to a polymer blend hydrogel. Two different concentrations (100 and 400 mg/mL) of hydroxypropyl-β-cyclodextrin (HP-β-CD) were added to the gel in order to study the effect of CD concentration. This specific CD was selected because of its known ability to complex many drug molecules and produce complexes with very high aqueous solubility (compared to other CDs). Additionally, two different types of small unilamellar vesicles (SUV) consisting of phosphatidylcholine PC or hydrogenated PC (HPC) were used. Because of the complexity of the system and to understand the implications of the various components on the properties of the final system, in a systematic way, various control gels containing (i) only liposomes or (ii) only CDs were constructed and evaluated under identical experimental conditions. Furthermore, the effect of aging on the rheological properties of all gels was studied by incubating the gels under constant temperature and humidity conditions (ICH accelerated stability test) for 1 week or 1 month.
2. Material and Methods Figure 1. Structure sof (A) hydroxyethyl cellulose, (B) hydroxypropyl-β-CD, and (C) polyacrylic acid.
case, another important factor is the degree of dilution of the liposome dispersion. Depending on the specific therapeutic need, it may be required to control the drug release kinetics. Because the dilution factor of liposome dispersions is connected with the physiology of the drug administration site and therefore ranges between specific values that cannot be considerably modified, the solubility of the drug in the aqueous environment of the site is possibly the most important, and perhaps the easiest to modulate, factor. Cyclodextrins (CDs)10 are cone-shaped oligosaccharides (Figure 1) that are known to increase the solubility of amphiphilic/lipophilic drugs by forming soluble complexes that incorporate the drug within their lipophilic “cave”, mainly by hydrophobic and van der Waals interactions.11-13 Therefore, the addition of CDs to such formulations may increase the release rate of the drugs from the dispersed liposomes (which act as lipid-phase reservoirs).14-16 Furthermore, it may be possible to control the release rate by adding different amounts and different types of (10) Loftsson, T.; Duchene, D. Int. J. Pharm. 2007, 329, 1–11. (11) Harries, D.; Rau, D. C.; Parsegian, V. A. J. Am. Chem. Soc. 2005, 127, 2184–2190. (12) Lui, L.; Guo, Q. X. J. Inclusion Phenom. Macrocyclic Chem. 2002, 42, 1–14. (13) Rekharsky, M. V.; Inoue, Y. Chem. Rev. 1998, 98, 1875–1917. (14) Boulmedarat, L.; Piel, G.; Bochot, A.; Lesieur, S.; Delattre, L.; Fattal, E. Pharm. Res. 2005, 22, 962–971. (15) Joguparthi, V.; Anderson, B. D. Pharm. Res. 2008, 25, 2505–2515. (16) Cal, K.; Centkowska, K. Eur. J. Pharm. Biopharm. 2008, 68, 467–478.
Langmuir 2009, 25(15), 8480–8488
Phosphatidylcholine (PC, egg lecithin) and hydrogenated-PC (egg) (HPC) were purchased from Lipoid Gmbh (Ludwigshafen, Germany). The chemical purity of the phospholipids was verified by thin layer chromatography, as described before.23 In brief, the lipids were developed on silicic acid-coated plates (Merck, Darmstandt, Germany) using chloroform/methanol/water (65:25:4 v/v/v) as the solvent system, and both lipids gave single spots. Cholesterol (99%) (chol) was purchased from Sigma-Aldrich Hellas (Chemilab, Athens, Greece). Hydroxyethylcellulose (HEC), as Natrosol 250 HX (Hercules Inc.) was kindly provided by Unipharma (Athens, Greece). Carbopol 974 P NF (CRB) was kindly provided by Chemix S.A. (Athens, Greece). Hydroxypropyl-β-cyclodextrin was purchased from TCI Europe N.V. All solvents used were of analytical or HPLC grade and were purchased from Merck (Darmstandt, Germany). All other materials were of analytical grade and were purchased from SigmaAldrich (Chemilab, Athens, Greece). (17) Mourtas, S.; Haikou, M.; Theodoropoulou, M.; Tsakiroglou, C.; Antimisiaris, S. G. J Colloid Interface Sci. 2008, 317, 611–619. (18) Hatzi, P.; Mourtas, S. G.; Klepetsanis, P.; Antimisiaris, S. G. Int. J. Pharm. 2007, 333, 167–176. (19) Piel, G.; Piette, M.; Barillaro, V.; Castagne, D.; Evrard, B.; Delattre, L. J. Inclusion Phenom. Macrocyclic Chem. 2007, 57, 309–311. (20) Alexanian, C.; Papademou, H.; Vertzoni, M.; Archontaki, H.; Valsami, G. J. Pharm. Pharmacol. 2008, 60, 1433–1439. (21) Zhang, L.; Hsieh, Y.-L. J. Nanosci. Nanotechnol. 2008, 8, 4461–4469. (22) Melzak, K. A.; Bender, F.; Tsortos, A.; Gizeli, E. Langmuir 2008, 24, 9172– 9180. (23) New, R. R. C., Ed. Liposomes: A Practical Approach; IRL Press: New York, 1990; Chapter 2.
DOI: 10.1021/la804305z
8481
Article A Shimatzu UV-1205 spectrophotometer was utilized for the measurement of liposomal lipid. Rheological measurements were performed on a stress rheometer (Rheometrics SR-200). 2.1. Preparation of Liposomes. Multilamellar vesicles (MLV) were prepared by the thin film hydration method.24 In brief, the appropriate weights of lipid and chol were dissolved in a chloroform/methanol (2:1 v/v) mixture and subsequently evaporated in a round-bottomed flask connected to a rotory evaporator until a thin lipid film was formed on the sides of the flask. The lipid film was hydrated with the appropriate volume of citrate buffer (pH 5.0) at 40 C in the case of PC/chol liposomes and at 60 C in the case of HPC/chol. After complete lipid hydration and the formation of liposomes, the vesicle dispersion was placed in a probe sonicator (Sonics, Vibra Cell, U.K.) for the reduction of vesicle size. Sonicated unilamellar (SUV) liposomes were prepared by subjecting the MLV dispersions to probe sonication for one or two 10 min cycles or until the dispersions became completely clear. After this, the SUV dispersions were centrifuged at 10 000 rpm (Heraeus Biofuge 28RS, Germany) for 10 min in order to precipitate any titanium fragments released from the probe during sonication. Finally, all of the liposome dispersions were incubated at the temperature of preparation for 1 to 2 h in order to anneal structural defects. 2.2. Characterization of Liposome Preparations. The lipid concentration of liposomes was measured by the Stewart colorimetric assay. In this assay, phospholipids form a colored complex with ammonium ferrothiocyanate (OD 485 nm) that is subsequently extracted in chloroform.25 An appropriate calibration curve using known concentrations of lipid was constructed. After measurement, the lipid concentrations of the liposome dispersions were adjusted to the desired value in order to prepare the liposome-containing gels with 5 or 20 mg of lipid/mL of gel, as described below. The size distribution (mean diameter and polydispersity index) and ζ potential of liposomes were measured by dynamic light scattering (DLS) and laser Doppler electrophoresis (LDE), respectively, on a Nano-ZS nanosizer Nanoseries (Malvern Instruments, U.K.), which enables the mass distribution of particle size as well as the electrophoretic mobility to be obtained. Measurements were made at 25 C at a fixed angle of 173. Sizes quoted are the z average means (dz) for the liposomal hydrodynamic diameter (nm). The zeta potential (mV) was calculated by the instrument (from electrophoretic mobility). 2.3. Preparation of Gels. The different types of gel formulations used in this study are presented in Table 1. For their preparation, appropriate amounts of Carbopol 974 NF and Natrosol 250-HX were weighted and added slowly to a citrate buffer solution (pH 5.0) for the control gel preparation or to the appropriate liposomal dispersion for complex gel preparations, under constant stirring by a paddle stirrer (100-150 rpm). In the case of CD-containing gels, the appropriate amount of HP-β-CD was weighed and dissolved in the liposomal dispersion (if liposomes were also added) or in a small volume of buffer. After the addition of the full amount of solid materials, the gels were allowed to swell under moderate stirring (50 rpm) for at least 2 h. In all formulations, sodium benzoate (0.02% w/v) was included in the buffer used, as a preservative, and also 1.0% (v/v) glycerin was added to all of the gels at the end of their preparation for the prevention of dehydration.
2.4. Measurement of Rheological Properties and Rheological Models Applied. All measurements were performed on cone-and-plate geometry (diameter = 4 cm, slope = 2) at constant temperature (T = 37 ( 0.5 C). In steady stress-sweep tests, a range of shear stresses at constant amplitude (0.1-1000 Pa) was applied to the sample, the shear rate was recorded, and the (24) Bangham, A D.; Standish, M M.; Watkins, J C. J. Mol. Biol. 1965, 13, 238– 252. (25) Stewart, J. C. M. Anal. Biochem. 1980, 104, 10–14.
8482 DOI: 10.1021/la804305z
Mourtas et al. shear viscosity was calculated as the ratio of shear stress to shear rate. In dynamic stress-sweep tests, a range of sinusoidal stresses at constant frequency (0.005-20 rad/s) were applied to the sample, and the storage and loss moduli, G0 and G00 , respectively, were measured. The storage modulus (or elasticity modulus) G0 is a measure of the elastic behavior of a gel, which is associated with energy storage, and the loss modulus (or viscosity modulus) G00 is a measure of the viscous behavior of a gel, which is associated with viscous energy dissipation.26,27 The shear viscosity values of the gels were fitted with the Cross model,26 given by eq 1 μ ¼ μ¥ þ ðμ0 -μ¥ Þ½1 þ ðγ=γ0 Þ2 ðn -1Þ=2
ð1Þ
where μ¥ is the infinite shear rate viscosity, μ0 is the zero shear rate viscosity, γ0 is the critical shear rate where the slope of the relationship μ(γ) drops (namely, the fluid transitions from Newtonian to power law behavior), and n is the power law index (at high shear rates, n - 1 tends asymptotically to the slope of the regression line when μ is plotted vs γ on a logarithmic scale). To estimate parameters μ0, γ0, and n ( μ¥ was set equal to the viscosity of water) of the Cross model, nonlinear regression analysis was done using the Bayesian estimator of Athena software package (Stewart and Associates). For an estimation of the elastic modulus (relaxation strength) and relaxation times, previously reported equations were used26-29 (Supporting Information). 2.4.1. Gel Aging Studies. All rheological measurements performed on the various gel types were repeated after 7 and 30 days of aging. For this, the gels were placed in sealed containers and incubated in a homemade constant temperature/ humidity oven at 40 ( 2 C/75 ( 5% (ICH accelerated testing conditions).
2.5. Differential Scanning Calorimetry Experiments. DSC experiments were carried out, in duplicate, in order to investigate the interactions between HP-β-CD molecules and polymers. Three aqueous solutions of cyclodextrin in buffer citrate at pH 5.0 were prepared (10, 50, and 100 mg/mL). To these solutions, appropriate amounts of Carbopol (4 mg/mL) and Natrosol (15 mg/mL) were added in order to have the same ratio of the two polymers as that in the gels. The mixtures were stirred until the polymers were completely dispersed. Each solution was freeze dried. A Star DSC1 (Mettler-Toledo) system with a refrigerated cooling accessory was used. Nitrogen was used as the purge gas at a flow rate of 20 mL min-1. The calorimeter was calibrated for the baseline using empty pans and for the cell constant and temperature using indium (melting point 156.61 C, enthalpy of fusion 28.71 J g-1). The samples were heated from 50 to 250 C at a rate of 10 C min-1.
2.6. Fourier Transform Infrared Spectroscopy (FTIR). FTIR was used to investigate possible interactions between HP-βCD molecules and polymers. The same samples that were prepared for DSC experiments were also evaluated by FTIR. Solids of cyclodextrin, Carbopol, and Natrosol were also used as references. Spectra were recorded with an FTIR spectrometer (Digilab Excalibur, Randolph, MA) at a resolution of 2 cm-1. Scans were run over the range of 400-4000 cm-1 using the KBr pellet technique. To obtain good-quality spectra, a minimum of 20 scans were accumulated. (26) Macosco, C. W. Rheology: Principles, Measurements, and Applications; Wiley-VCH: New York, 1994. (27) Bird, R. B., Armstrong, R. A., Hassager, P., Eds. Dynamics of Polymer Liquids; John Wiley & Sons: New York, 1977; Vol. 1. (28) Robb, I. D.; Smeulders, J. B. A. F. Polymer 1997, 38, 2165–2169. (29) Tan, H.; Tam, K. C.; Jenkins, R. D. J. Colloid Interface Sci. 2000, 231, 52– 58.
Langmuir 2009, 25(15), 8480–8488
Mourtas et al.
Article
Table 1. Compositions of the Gels Studieda gel name
polymers
HP-β-CD lipid conc lipid compb conc (mg/mL) (mg/mL)c
BL (blank) no lipid 0 0 HPC1 all gels: 0, 40% HPC 0 5 (w/v) Carbopol HPC HPC2 0 20 (974 NF) and CD1 no lipid 100 0 1,5% (w/v) CD2 no lipid 400 0 Natrosol CG-1-H HPC 100 5 (250-HX) CG-2-H HPC 400 5 CG-3-H HPC 100 20 CG-4-H HPC 400 20 PC1 PC 0 5 PC2 PC 0 20 CG-1-P PC 100 5 CG-2-P PC 400 5 CG-3-P PC 100 20 CG-4-P PC 400 20 a All gels contain Carbopol 974 NF and hydroxyethyl cellulose (Natrosol) at the concentrations reported as well as citrate buffer pH 5.00 (0.10 M), glycerin (1% w/v), and sodium benzoate (0.20% w/v). b All liposomes contain chol at lipid/chol 2:1 (mol/mol). c Lipid concentrations were remeasured after gel preparation and were found to be (5.013 ( 0.022)-(20.045 ( 0.050).
Table 2. Mean Diameter and Zeta Potential Values of the Liposomes Used in Gelsa lipid comp
mean diameter (nm)
PIb
ζ potential (mV)
PC/chol (2:1) 100.9 ( 7.5 0.201 -2.1 ( 4.1 HPC/chol (2:1) 109.6 ( 5.2 0.163 2.05 ( 2.75 a Each value is the mean, calculated from at least three separate preparations, and the standard deviation of the mean is presented. b PI is the polydispersity index for the measurements (mean ( SD).
3. Results 3.1. Liposome Physicochemical Properties. The size distribution (mean diameter and polydispersity index) and ζ potential of some of the liposome dispersions are presented in Table 2. As anticipated, SUV liposomes are small with mean diameters ranging between 101 and 110 nm, slightly larger when compared to SUV liposomes without chol in their lipid membranes, as previously reported.17 The ζ-potential values show that these vesicles have no surface charge. This was expected because the lipids used for their preparation are not charged. 3.2. Gel Rheological Properties. 3.2.1. Shear Viscosity versus Shear Rate. 3.2.1.1. Blank Gel: Effect of Temperature. The shear viscosity as a function of shear rate for the blank gel is shown in Figure 2. Measurements were performed at different temperatures ranging from 20 to 37 C. It is evident that there is practically no temperature effect on the viscosity of this gel, as anticipated by previous studies.6,7 Furthermore, the blank gel demonstrates shear-thinning behavior that is fitted well by the Cross model (the fitting of the experimental data (points) is presented as lines in the graphs). 3.2.1.2. Complex Gels. In Figures 3 and 4, the viscosity/rate graphs are shown for some of the HPC/chol- and the PC/cholcontaining gels, respectively. The measurements performed on the different types of gels are presented as symbols, and the curves predicted by the Cross model are presented as lines. All graphs in Figures 3 and 4 are plotted on identical x-axis/y-axis scales in order to permit direct comparison between the different cases. In all gels, the fitting of the experimental data (points) to the Cross model (lines) was very good, especially over the shear-thinning flow regime. In Table 3, the parameters estimated by the Cross model, zero shear rate viscosity μ0, and power law index n are Langmuir 2009, 25(15), 8480–8488
Figure 2. Viscosity versus shear rate graph of the blank gel (BL), measured at various temperatures ranging between 20 and 37 C. The symbol key is included in the graph inset.
given for all of the gel types evaluated. The zero shear rate viscosity is a measure of the fluidity of gels under stress-free conditions. The power law index, n, of the Cross model is a measure of the slope of the μ(γ) curve at high shear rates where the gel tends to behave as a power law fluid. Actually, the slope of μ(γ) at high γ values tends to become equal to n - 1, and as the value of n decreases, the drop in viscosity with γ becomes sharper . When observing the modulation of the shear viscosity of the gels as a function of shear rate in Figures 3 and 4, we realize that almost all of the gels are clearly shear-thinning fluids and have the tendency to become Newtonian at asymptotically low shear rates. The gel that contains PC/chol liposomes (5 mg/ mL) together with 400 mg/mL CD (plotted as triangles in Figure 4A) has a very high zero-rate viscosity, and no shear-thinning behavior is observed over the range of shear stresses applied by the rheometer. However, gel CD2, which contains the same concentration of HP-β-CD but no lipid (plotted as triangles in Figure 3A), is shearthinning, indicating that the addition of PC/chol liposomes may be responsible for the reduction of the fluidity of the previous gel (CG-2-P). Nevertheless, when a higher number of liposomes is added to the same gel, in which case gel CG-4-P is formed, it again becomes shear-thinning (triangles in Figure 4B). This indicates that the ratio of CD to lipid in the gel may have an effect on the rheological behavior of the final product. Additionally, by comparing the viscosity graphs of similar gels (that contain the same amounts of CD and lipid but different types of liposomes, as in Figures 3B and 4A; for more data see Supporting Information), one can easily conclude that the flow behavior of the final product is also affected by the lipid composition of the liposomes added to the complex gels, in agreement with earlier observations.17 3.2.1.3. Effect of Aging. The rheological data measured for the gels after 1 week (7 days) and 1 month (30 days) of aging are presented in the Supporting Information. For the blank gel, the viscosity increases with aging. Indeed, as seen in Table 3, the calculated zero-rate viscosity μ0 is increased by factors of 2 and 16 after 1 week and 1 month of aging, respectively. However, the gel remains shear-thinning with no significant modification of the shear rate value γ0 at which the viscosity of the gel starts to decrease. In general, the effect of CD molecules on the rheological profile of the complex gels during aging seems to be influenced by the presence of liposomes and also by the number and type of liposomes added to the gels. Indeed, gel CD2, which contains 400 mg/mL CD and no liposomes, has a very high zero-rate viscosity after 1 month of aging (see Supporting Information for data) and abolishes its shear-thinning behavior (over the full shear stress range of the rheometer). Nevertheless, when HPC/chol DOI: 10.1021/la804305z
8483
Article
Mourtas et al.
Figure 3. Viscosity versus shear rate graph of various types of control and complex gels, measured immediately after preparation. Graph A is for the gels without lipid; B, for 5 mg/mL HPC/cholliposome-containing gels. (For the results of gels with higher liposome concentration and the results obtained after gel aging, see Supporting Information.) The symbol key is included in the graph inset.
liposomes are added to that same gel at 5 or 20 mg/mL, the initial shear-thinning behavior of the gels is maintained even after the 1 month aging period (Figure 1 in Supporting Information). In contrast, the addition of 5 mg/mL PC/chol liposomes to the gel seems to affect the gel negatively because it loses its shear-thinning ability even at time 0 (Figure 4A) and the same behavior is also observed after aging. However, when PC/chol liposomes are added at a higher concentration (20 mg/mL), the gel regains it shear-thinning behavior initially, but after 1 week of aging, it stops flowing and becomes very sticky again (Figure 2 in Supporting Information). This behavior indicates that when high amounts of PC/chol liposomes are added to the gel that contains high amount of HP-β-CD the interactions that occur between gel and liposome components are slow and time is needed for the system to equilibrate. As shown in Table 3, the effect of adding HPC/chol liposomes to the blank gel on the zero shear rate viscosity μ0 and power law index n is similar to that observed previously for HPC liposomes;17 μ0 values increase by an order of magnitude when 5 mg/ mL liposomes is included (comparison between BL and HPC1, gels) and furthermore when the lipid content is increased to 20 mg/mL (comparison between BL and HPC2, gels), whereas the n index decreases in both cases. However, the addition of PC/chol liposomes to the blank gel (PC1, PC2) also results in an increase in μ0 and a slight decrease in n, although this increase/decrease is lower than that caused by equivalent amounts of HPC/chol liposomes. This was not the case for PC liposomes that were studied previously in the same blank gel system17 and were observed to cause a decrease in μ0 values and a slight increase in n values. Thereby, it is evident that the rigidity of the liposomal membrane is indeed one of the parameters that influence the effect 8484 DOI: 10.1021/la804305z
Figure 4. Viscosity versus shear rate graph of 5 mg/mL PC/cholliposome-containing complex gels measured immediately after preparation. (For the results of gels with higher liposome concentration and the results obtained after gel aging, see Supporting Information.) The symbol key is included in the graph inset.
of liposome addition to polymeric gels on the gel properties; PC/chol liposomes used herein behave differently compared to the PC liposomes used before17 as a result of the addition of cholesterol to their lipid membrane, which results in the formation of a more rigid lipid membrane. 3.2.2. Dynamic Frequency Sweep Tests. From the results of the oscillatory measurements performed (DFS - dynamic frequency sweep), one can extract information about the network structure of gels. For each formulation, the elasticity (storage) modulus (G0 ) and the viscosity (loss) modulus (G00 ) were measured as a function of the oscillatory frequency (hertz). The gel structure was examined over the frequency range of 0.005-20 rad/s. 3.2.2.1. Blank Gel. In Figure 5, the dynamic measurements of the blank gel are compared with corresponding predictions of the multimodal Maxwell model. (See Supporting Information for more details about the model.) As seen, the Maxwell model satisfactorily predicts the experimental frequency responses (predictions are presented as lines and experimental responses as symbols), although some discrepancy is observed over the very low frequencies measured. In general, the storage modulus (plotted as solid symbols) is higher than the loss modulus (plotted as hollow symbols), particularly with reference to the Langmuir 2009, 25(15), 8480–8488
Mourtas et al.
Article
Table 3. Zero Shear Rate Viscosity and Power-Law Index Values, Calculated for All of the Gels Studied at Time 0 (Immediately after Preparation) and after 7 and 30 Days of Aging by the Cross Model Equation (eq 1)a gel name
μ0, (Pa s) time 0/7 days/30 days
BL 137 ( 13/289 ( 21/2215 ( 193 HPC1 1248 ( 178/733 ( 90/1640 ( 167 HPC2 2154 ( 300/1269 ( 126/1147 ( 104 CD1 25 475 ( 3300/879 ( 65/2073 ( 240 CD2 902 ( 83/1395 ( 92/45 707 ( 4511 CG-1-H 1399 ( 179/2181 ( 202/4052 ( 414 CG-2-H 1638 ( 101/1566 ( 126/1732 ( 131 CG-3-H 4938 ( 569/5786 ( 1021/14 662 ( 1736 CG-4-H 3619 ( 331/3971 ( 365/11 460 ( 1824 PC1 327 ( 29/623 ( 44/11 570 ( 1530 PC2 831 ( 34 /1317 ( 85/5593 ( 540 CG-1-P 719 ( 66/709 ( 75/4563 ( 500 CG-2-P 37 444 ( 22 000/18 270 ( 14 000/58 816 ( 36 000 CG-3-P 992 ( 87/2046 ( 181/8127 ( 905 CG-4-P 2335 ( 233/VH/VH a ND - not determined; VH - very high.
n, time 0/7 days/30 days 0.528/0.519/0.228 0.187/0.331/0.383 0.176/0.404/0.503 0.425/0.394/0.275 0.406/0.383/ND 0.322/0.467/0.146 0.342/0.373/0.333 0.574/0.701/0.361 0.438/0.669/ND 0.518/0.394/0.222 0.489/0.469/0.441 0.445/0.422/0.078 ND 0.413/0.595/0.322 0.416
Figure 5. Storage or elasticity (G0 ) (closed symbols) and loss or
viscosity modulus values (G0 ) (corresponding open symbols) measured for the blank gel (BL) immediately after preparation or after 7 or 30 days of aging. The symbol key is included in the graph inset.
high-frequency range. At time 0 (square symbols), the BL gel has elastic behavior that is similar to that observed before.17 3.2.2.2. Complex Gels. For the complex gels (Figure 6), only the storage (elasticity) modulus is presented, for clarity. The viscoelastic properties of the various gel formulations studied are different. When adding CD to the gel at concentration 400 mg/mL, the storage modulus G0 increases dramatically (CD2 gel, plotted as solid diamond symbols in Figure 6A). Indeed, the gel that contains 400 mg/mL CD is strongly elastic (G0 is >104 over the full frequency range studied) compared to the blank gel, whereas at 100 mg/mL the effect of CD addition in the gel is negligible (solid down-triangles in Figure 6A). Oppositely, the variation of G0 due to the addition of liposomes to the BL gel (symbols for gels HPC1-HPC2 and PC1-PC2 given in the figure insets) is much less pronounced (Figure 6A,B). When liposomes are added to the 400 mg/mL CD-containing gel (CD2), their effect on the G0 values is different depending on the type and number of liposomes loaded. The highest effect is observed in the case of adding 5 mg/mL HPC/chol liposomes (hollow circles in Figure 6. A). In this case, the G0 values are substantially decreased (compared to those of the CD2 gel) and range from 5 101 to 5 103. However, when greater amounts of HPC/chol liposomes are added to the same gel (gel GC-4H, presented as hollow downtriangles in Figure 6.A), their effect on the G0 values is lower (compared to that of the 5 mg/mL lipid-containing gel), and G0 values range from 5 102 to 5 104. Furthermore, the gels that contain 400 mg/mL CD together with PC/chol liposomes have high G0 values as presented in Figure 6B (as hollow circles and hollow down-triangles), which are only slightly lower compared to those of the CD2 gel at all frequencies measured. Interestingly, Langmuir 2009, 25(15), 8480–8488
Figure 6. Storage or elasticity (G0 ) modulus values of control gels (blank and gels with no lipid) and gels that contain (A)HPC/chol liposomes or (B) PC/chol liposomes, measured immediately after preparation. For the values measured after gel aging, see the Supporting Information. The symbol key is included in the graph insets.
in the case of PC/chol liposome addition in to the CD2 gel, the amount of lipid loaded into the gel has no effect on the gel elasticity. The above-mentioned comparisons indicate that the elastic properties of the complex gels are affected by several different factors, including structural modifications caused by the addition of lipid particles to the gels (that have been previously reported to be influenced by the rigidity of the liposomes added) and modifications caused by interaction between the various gel components. 3.2.2.3. Effect of Aging on Dynamic Rheological Values of Gels. For the blank gel, as seen in Figure 5 the variation of G0 (ω) and G00 (ω) during increased periods of aging is nonlinear (by comparing the triangle symbols (7 days of aging) and the diamond symbols (30 days of aging) with the square symbols (time 0)). Indeed, it is seen that both G0 and G00 values of the blank gel are significantly decreased after 7 days of aging, compared to DOI: 10.1021/la804305z
8485
Article
Mourtas et al.
Figure 8. Correlation between relaxation modulus and relaxation times for gels that contain PC/chol liposomes at 5 mg/mL, measured immediately after preparation. (For gels with higher liposome concentration and data after gel aging, see the Supporting Information.) The symbol key is presented in the graphs.
Figure 7. Correlation between relaxation modulus and relaxation times for (A) control gels (blank and gels with no lipid) and (B) gels that contain HPC/chol liposomes at 5 mg/mL, measured immediately after preparation. (For gels with higher liposome concentration and data after gel aging, see the Supporting Information). The symbol key is presented in the graphs.
the values measured at time 0, whereas after 30 days of aging there is a significant increase. This behavior may be associated with changes in the structure of the polymer networks. In the case of the complex gels, it is observed that the different gel-types studied have different behaviors during aging. (See Supporting Information for data (Figures 3 and 4)). In summary, the initial drop in G0 values after 7 days of aging followed by an increase after 30 days was observed in the case of gels that contain polymers and liposomes only, which have behavior similar to that of blank gel (Figure 5), but not when HP-β-CD is also present. More analytically, when comparing G0 values for the different gels at different time points of incubation, the following points are observed: (i) For all of the gels that contain 400 mg/mL CD, with or without liposomes, the elasticity modulus remains more or less unchanged during the full aging period. (ii) For gels that contain 100 mg/mL CD, with or without liposomes, a gradual, slight increase in G0 values occurs during aging, irrespective of the number or type of liposomes that they contain. The storage modulus increases by almost 1 order of magnitude after 1 month of aging. (iii) In almost all other cases, there is a slight modulation of G0 values that consists of an initial drop (after 1 week of aging) followed by a slight increase (after 1 month of aging). Finally, the G0 values are higher than the initial value by about 1 order of magnitude after 1 month of aging. The only exception is the case of gel PC1 (the gel that contains 5 mg/mL of PC/chol liposomes), for which an increase in G0 values after 1 week of aging was observed and then after 1 month of aging a decrease, or better, a return to the initial value was observed. 3.2.2.4. Maxwell Model Fitting. The four pairs of relaxation times and relaxation strengths (I = 4), estimated for each gel (for more details about the method used, see Supporting Information) are shown in Figures 7 and 8. The more pronounced elastic behavior of gels containing 400 mg/mL of CD, compared to that of all of the other gels studied, is reflected by the higher values of relaxation strength, Gi, calculated over all of the relaxation times used (easily observed by comparing the star 8486 DOI: 10.1021/la804305z
symbols with the squares and triangles in Figures 7 and 8). However, the relaxation strengths decrease substantially when 5 mg/mL HPC/chol liposomes is added to the gel (Figure 7B), correlating well with the significant drop in the G0 values measured for this gel compared to that of the CD2 gel (Figure 6A), as noted above. When comparing the gels that contain 100 mg/mL CD (hollow triangles) and no CD (solid squares), it is evident that the 100 mg/mL CD concentration does not cause significant changes in Gi , λi (i = 1,.., 4) values. This means that such a CD concentration is so low that the gel structure remains “unsaturated” without any respectable changes. In contrast, at higher CD concentration (equal to 400 mg/mL) the gel structure is “saturated” and the excess CD results in a more elastic structure. Even when no lipid is present in the gels, the 100 mg/mL CD content is not enough to cause substantial changes in the gels, denoting that this CD quantity may be fully “neutralized” by interactions occurring between the CD molecules and the polymer components of the gels, Carbopol and Natrosol. With respect to relaxation times, the estimated relaxation times span a range of at least 5 orders of magnitude (0.01-1000s), as seen in Figures 7 and 8. Each relaxation time is associated with a process and depends on the size of the cluster or network that relaxes (the larger the cluster, the longer the relaxation time). The very broad range of relaxation times estimated for all gel types studied is indicative of a wide range of cluster sizes. The structural changes occurring on the gels after the addition of liposomes and CD molecules seem to cause more changes in the rigidity of the clusters (most possibly by creating cross links of particles with polymers), but it seems that the cluster size distribution is more weakly affected. Regarding the effects of aging on the relaxation strength and time spectrum (corresponding data in Supporting Information), the gels containing the highest CD concentration (400 mg/mL) are almost insensitive to aging, whereas the other gel types studied seem to become more elastic and less viscous as aging proceeds. As also mentioned above for liposome and CD addition, the effects of aging on the gels seem to be more dramatic with respect to relaxing cluster rigidity and weaker with respect to cluster size distribution. 3.3. DSC and FTIR Studies. 3.3.1. DSC Studies. The possibility that interactions occur between the polymers of the blank gel and HP-β-CD molecules, under the specific conditions applying in the complex gels, was investigated because this would explain the substantial modulations observed in the rheological characteristics of the blank gel after CD addition. Indeed, it has been previously reported that a strong interaction occurs between Langmuir 2009, 25(15), 8480–8488
Mourtas et al.
Figure 9. DSC thermograms of freeze-dried samples of Carbopol, Natrosol, HP-β-CD, and Carbopol/Natrosol/ HP-β-CD blends (S1-S3). (A) First set of thermograms obtained during the initial heating cycle. (B) Second set of thermograms obtained (after the sample pans were cooled to 25 C and heated again). A line key is presented in the graphs.
HP-β-CD and Carbopol (in water) and that the complexes formed show high thermal stability.30,31 As seen in Figure 9A, in the first DSC thermogram of Carbopol, Natrosol, and some of the samples processed (namely, S1 and S2), a broad endothermic peak between 50 and 120 C is always recorded. This peak has been documented in previous reports and is attributed to physically bound water. The fact that the peak of physically bound water observed in the S1 thermogram is considerably lower in S2 and finally disappears in S3 reveals that some interactions between the polymer components and HP-β-CD occur that result in the displacement of bound water from the blends. It is logical that as this interaction increases (with increasing CD concentration) the displacement of the bound water molecules also increases. In Figure 9B, the first and second thermograms obtained for the Carbopol sample and the S1 blend are presented. As seen, Carbopol has a Tg at 163 C, which is not observed in any thermogram of sample S1, proving that HP-β-CD molecules interact with Carbopol in accordance with previous reports32,33 under the specific conditions that apply in the gels. Natrosol did not give a characteristic endothermic peak in the second DSC run and is not presented in this graph in order to avoid confusion. 3.3.2. FTIR Studies. The FTIR spectra suggest that several experimental factors are important for the results obtained. First, the use of buffer citrate pH = 5.0 for sample preparation, had a significant effect on the Carbopol carboxy groups. Additionally, when freeze-dried compounds (Carbopol, natrosol and HP-βCD) or physical mixtures (Carbopol/natrosol/HPβCD) were analyzed, several differences between the two types of samples are detected. (See Supporting Information for data.) An analysis of freeze-dried samples of pure Carbopol in citrate buffer (see Supporting Information for data) shows that the band (30) Rodrı´ guez-Tenreiro, C.; Alvarez-Lorenzo, C.; Concheiro, A; TorresLabandeira, J. J. J. Therm. Anal. Calorim. 2004, 77, 403–411. (31) Rodriguez-Tenreiro, C.; Diez-Bueno, L.; Concheiro, A.; TorresLabandeira, J. J.; Alvarez-Lorenzo, C. J Controlled Release 2007, 123, 56–66. (32) Gomez-Carracedo, A.; Alvarez-Lorenzo, C.; Gomez-Amoza, J. L.; Concheiro, A. Int. J. Pharm. 2004, 274, 233–243. (33) Park, S.-H.; Chun, M.-K.; Choi, H.-K. Int. J. Pharm. 2008, 347, 39–44.
Langmuir 2009, 25(15), 8480–8488
Article
at 1719 cm-1 disappears and is replaced by two different bands at 1704 cm-1 (attributed to the vCdO band of the Carbopol carboxy group (H-bond dimer) and 1587 cm-1 (attributed to the symmetrical CO2 stretching of the carbonyl group)) . A very important finding is that the band at 1587 cm-1 (Carbopol spectrum) is broad and covers the bands of HP-β-CD and Natrosol at approximately 1598 cm-1. When mixtures of Carbopol/Natrosol/HP-β-CD are analyzed (Supporting Information), the characteristic bands of HP-β-CD at 1000-1220 cm-1 are observed in all of the mixtures,34 but the broad Carbopol carbonyl band at 1588 cm-1, which covers the band of HP-β-CD and natrosol35 at 1599 cm-1, is not detected. This indicates that some interaction occurs between Carbopol and hydroxyl groups of HP-β-CD or Natrosol. The fact that the Carbopol carbonyl band at 1704 cm-1 does not disappear proves that the carboxylic acid group of Carbopol does not react but that hydrogen bonds are formed between carboxylic and hydroxyl groups. The effect of increasing concentrations of HP- β-CD on the sample spectra proves that the interaction observed is between Carbopol and CD molecules (and not Natrosol). (See Supporting Information for data and more details.)
4. Discussion Gels that contain mixtures of Carbopol 974 and hydroxyethylcellulose were previously reported to have many advantages in the vaginal delivery of microbicides.6,7 The addition of liposomes in such gels modulates their rheological characteristics.17 Indeed, significant modifications of gel rheological properties were observed when liposomes were loaded at 5 mg/mL or higher concentrations in the case of liposomes composed by HPC. The observed effects on the rheological behavior of such gel formulations may have some practical implications for their in vivo behavior. The explanation given previously to explain the effect of HPC liposomes on the rheological behavior of the blank gel is that the rigidity of the HPC liposomes is such that they do not behave comparably with the gel under stress conditions, resulting in a modification of the flow and the elastic properties of the complex gel, compared to the blank one. This is probably related to the fact that HPC liposome membranes are in the gel state at the temperature at which the rheological measurements were performed (37 C) because the transition temperature of HPC is 50 C.36 However, PC liposomes that are in the fluid state (because the transition temperature of PC is below zero) are easily deformed under stress conditions and thus do not modulate the rheological properties of the blank gel.17 Herein, the same gel system was studied when a large number of rigid membrane liposomes were added together (or not) with an extra component, HP-β-CD. This CD type may be needed in order to control the release of amphiphilic drugs with low aqueous solubility from such complex systems, as explained in the Introduction.14-16 Additionally, the aging of such complex systems was evaluated with respect to the effect of aging on gel rheological properties. In general, the dramatic effect of adding a high concentration of CDs to the initial polymer blend (blank gel) is very interesting and has not been, to our knowledge, reported before. Indeed, when 400 mg/mL HP-β-CD is added to the gel, it seems to be transformed into a very elastic solid that cannot flow even under (34) Valero, M.; Perez-Revuelta, B. I.; Rodrı´ guez, L. J. Int. J. Pharm. 2003, 253, 97–104. (35) Luo, K.; Yin, J.; Khutoryanskaya, O. V.; Khutoryanskiy, V. V. Macromol. Biosci. 2008, 8, 184–191. (36) Antonov, V. F.; Anosov, A. A.; Norik, V. P.;. Korepanova, E. A.; Smirnova, E. Y. Eur. Biophys. J. 2003, 32, 55-59 (HPC).
DOI: 10.1021/la804305z
8487
Article
the influence of very high shear rates. Nevertheless, when in solution (in the buffer used for gel preparation), this high concentration of HP-β-CD is completely soluble, and no increase in the solution viscosity is observed. Thereby, it seems that there is some interaction between one, or both, of the polymers that are present in the system and the CD molecules, which is responsible for the increased viscosity observed and the transformation of the initial gel into the elastic nonflowing solid. Such interactions between CD molecules and cellulose derivatives have been recently mentioned during the formation of new hydrogels based on CDs cross-linked with ethylene glycol diglycidylether in the presence of hydroxypropylmethyl cellulose (HPMC).37 Furthermore, proof of interactions between various Carbopol types and CD molecules have also been reported before.30,31 By performing DSC (Figure 9) and FTIR (Supporting Information Figure 10) measurements, we have shown that in the case of the gels studied herein the most important interactions are occurring between Carbopol and HP-β-CD molecules. Nevertheless, the specific interaction or interactions occurring between the CD molecules and the polymers seem to be influenced by the presence of liposomes in the gels, whereas the type and number of liposomes added are important. A possible explanation for the modulation of the CD2 gel characteristics (the gel that contains only polymers and has a high CD content) upon liposome addition (Figures 3, 4, 6, and 7) is that some of the CD molecules prefer to interact with liposomes, extracting chol molecules from the liposome membranes, and thus the interaction between polymers and CDs is decreased, resulting in a different gel structure. In fact, the interaction between chol and HP-β-CD molecules is well known, and it has been demonstrated in our laboratory that CD molecules can extract chol molecules from lipid/chol liposomes when the liposomes are incubated in the presence of CDs.18 Furthermore, such interactions between CD molecules and liposomal chol molecules were also recently demonstrated to occur when the liposomes were dispersed in gel systems.22 It is logical to suggest that such interactions (between liposomes and CD molecules) are more likely to occur to a larger extent as the CD/lipid ratio increases. Furthermore, such interactions will require more time to occur when the components are dispersed in viscous gel systems,8 explaining the effects of aging on the gel rheological properties. In the gels studied herein, some of which contain a larger number of CDs together with a small number of liposomes, it is most likely that such interactions between the liposomes and the CD molecules occur, especially after incubation during the aging process. When CD molecules interact with liposomal chol and extract chol molecules from the lipid membranes, the remaining liposomes will be composed of PC or HPC, respectively, in the case of gels initially loaded with PC/chol or HPC/chol liposomes. These two types of vesicles were previously seen to have different effects on the blank gel rheological properties (minimum or no effects in the case of PC vesicles and substantial effects in the case of HPC vesicles).17 Thereby, in addition to the modulations caused by lowering the interactions between polymers and CD molecules (as a result of the complexation of some CD molecules with chol), perhaps the final rheological behavior of the complex (37) Rodriguez-Tenreiro, C.; Alvarez-Lorenzo, C.; Rodriguez-Perez, A.; Concheiro, A.; Torres-Labandeira, J. J. Pharm. Res. 2006, 23, 121–130.
8488 DOI: 10.1021/la804305z
Mourtas et al.
gels is also affected by the modulations in liposome membrane rigidity caused by chol depletion from the vesicle. However, we cannot derive any conclusions about the relative importance of the various mechanisms that are implicated in the observed modifications of the rheological behavior of the different gel types studied. Furthermore, when such systems are constructed as drug formulations, the presence of the drug may also affect the interactions between the different components and thus affect the rheological properties of the formulation. In fact, preliminary rheological studies in gel systems that contain the same polymer mixture as the one used herein, with 20 mg/mL liposomal drug, together with 400 mg/mL of the HP-ss-CD/drug complex reveal that the inclusion of CDs and liposomes in gels in the form of CDdrug complexes and liposomal drug formulations, respectively, has a significant effect on the rheological behavior of the system. Thereby, what is most important from a practical point of view is to consider, in advance, the rheological behavior of such complex systems during the design of topical drug formulations and make the appropriate modulations if and when needed.
5. Conclusions The results of this study demonstrate that the presence of HPβ-CD at high concentration (400 mg/mL or at least higher than 100 mg/mL) in a hydrogel consisting of two polymers (Carbopol 974 and Natrosol) has a profound effect on the viscocity and rheological properties of the gel, transforming it into a sticky elastic solid that has no ability to flow, even under high forces, after 1 week of aging under accelerated testing conditions. However, complex systems containing liposomes and such large numbers of CD molecules demonstrate increased flow under high shear rates, when the liposomes loaded in the gels consist of rigid lipid membranes, provided that the lipid/CD ratio in the gel is regulated. Therefore, when designing drug formulations for topical (and especially for vaginal) delivery in which CD molecules will be used together with polymers and liposomes, it is very important to adjust their relative quantities and types in order to achieve appropriate rheological behavior of the formulated systems. Acknowledgment. This work received funding from the European Community’s Sixth Framework Programme under contract number LSHP-CT-2004-503162 (Selection and Development of Microbicides for Mucosal Use to Prevent Sexual HIV Transmission/Acquisition). No guarantee or warranty is given by the EU that the information is fit for any particular purpose. We thank Chemix S. A. (Athens, Greece), Noveon distributors, for providing Carpobol 974 NF and Unipharma (Athens, Greece) for providing Natrosol. Supporting Information Available: Detailed information on the Cross model for fitting rheological curves; on the fourmodal Maxwell model for fitting the measured G0 (ω) and G00 (ω) values; data on the rheological properties of some of the gels immediately after preparation (time 0) and all of the gels after 7 days and 1 month of aging; and details of the FTIR studies performed on samples and control samples. This material is available free of charge via the Internet at http://pubs.acs.org.
Langmuir 2009, 25(15), 8480–8488
