Nonequilibrium Effects in Self-Assembled Mesophase Materials

Apr 5, 2010 - The chemical structure of GMO and PHYT, and a simplified generic phase diagram for their behavior in excess water is shown in Figure 1. ...
0 downloads 8 Views 3MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

)

Nonequilibrium Effects in Self-Assembled Mesophase Materials: Unexpected Supercooling Effects for Cubosomes and Hexosomes Yao-Da Dong,† Adam J. Tilley,† Ian Larson,† M. Jayne Lawrence, Heinz Amenitsch,‡ Michael Rappolt,‡ Tracey Hanley,§ and Ben J. Boyd*,† †

)

Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University (Parkville Campus), 381 Royal Pde, Parkville, Victoria, Australia, ‡ Institute of Biophysics and Nanosystems Research, Austrian Academy of Sciences, c/o Sincrotrone Trieste, 34012 Basovizza, Italy, §Bragg Institute, Australian Nuclear Science and Technology Organisation, Menai, New South Wales 2234, Australia, and Pharmacy Department, King’s College London, The Franklin-Wilkins Building, Stamford Street, London SE1 9NH, U.K. Received December 21, 2009. Revised Manuscript Received March 11, 2010 Polar lipids often exhibit equilibrium liquid crystalline structures in excess water, such as the bicontinuous cubic phases (QII) at low temperatures and inverse hexagonal phase (HII) at higher temperatures. In this study, the equilibrium and nonequilibrium phase behavior of glyceryl monooleate (GMO) and phytantriol (PHYT) systems in excess water were investigated using both continuous heating and cooling cycles, and rapid temperature changes. Evolution of the phase structure was followed using small-angle X-ray scattering (SAXS). During cooling, not only was supercooling of the liquid crystalline systems by up to 25 °C observed, but evidence for nonequilibrium phase structures (not present on heating; such as the gyroid cubic phase only present at low water content in equilibrium) was also apparent. The nonequilibrium phases were surprisingly stable, with return to equilibrium structure for dispersed submicrometer sized particle systems taking more than 13 h in some cases. Inhibition of phase nucleation was the key to greater supercooling effects observed for the dispersed particles compared to the bulk systems. These findings highlight the need for continued study into the nonequilibrium phase structures for these types of systems, as this may influence performance in applications such as drug delivery.

Introduction The equilibrium lyotropic phase behavior of liquid crystalline (LC) systems based on amphiphiles such as glyceryl monooleate (GMO),1,2 glyceryl monolinoleate (MLO),3,4 and phytantriol (PHYT)5,6 in excess water has been well studied. The chemical structure of GMO and PHYT, and a simplified generic phase diagram for their behavior in excess water is shown in Figure 1. These systems are known to form LC structures such as inverse bicontinuous cubic (QII) phase at ambient temperatures, and inverse hexagonal (HII) and inverse micellar (L2) phases at higher temperatures, and are considered to have application in the field of drug delivery7-9 due to the potential to provide a persistent drug reservoir for sustained release.10,11 These systems can often be dispersed to form submicrometer particles, which retain the *Corresponding author. E-mail: [email protected]. (1) Landh, T. J. Phys. Chem. 1994, 98, 8453–8467. (2) Gustafsson, J.; LjusbergWahren, H.; Almgren, M.; Larsson, K. Langmuir 1996, 12, 4611–4613. (3) Yaghmur, A.; de Campo, L.; Sagalowicz, L.; Leser, M. E.; Glatter, O. Langmuir 2005, 21, 569–577. (4) de Campo, L.; Yaghmur, A.; Sagalowicz, L.; Leser, M. E.; Watzke, H. Langmuir 2004, 20, 5254–5261. (5) Dong, Y.-D.; Dong, A. W.; Larson, I.; Rappolt, M.; Amenitsch, H.; Hanley, T.; Boyd, B. J. Langmuir 2008, 24, 6998–7003. (6) Dong, Y. D.; Larson, I.; Hanley, T.; Boyd, B. J. Langmuir 2006, 22, 9512– 9518. (7) Chang, C. M.; Bodmeier, R. J. Pharm. Sci. 1997, 86, 747–752. (8) Chang, C. M.; Bodmeier, R. J. Controlled Release 1997, 46, 215–222. (9) Sadhale, Y.; Shah, J. C. Int. J. Pharm. 1999, 191, 51–64. (10) Shah, J. C.; Sadhale, Y.; Chilukuri, D. M. Adv. Drug Delivery Rev. 2001, 47, 229–250. (11) Drummond, C. J.; Fong, C. Curr. Opin. Colloid Interface Sci. 1999, 4, 449–456. (12) Barauskas, J.; Svedaite, I.; Butkus, E.; Razumas, V.; Larsson, K.; Tiberg, F. Colloids Surf., B 2005, 41, 49–53.

9000 DOI: 10.1021/la904803c

internal structure of the nondispersed phase.12-14 In the case of inverse cubic and inverse hexagonal phases, these particles have been termed cubosomes and hexosomes, respectively. Pluronic F127 (F127) is a common polymeric colloidal stabilizer employed in their preparation. In contrast to the equilibrium phase behavior, the nonequilibrium phase behavior of lipid-based LC systems has received very little attention. Qiu and Caffrey (1999)15 and Clogston et al. (2000)16 reported significant supercooling for nondisperse (bulk) GMO-based systems; the QII to lamellar crystalline (Lc) transition was up to 17 °C below that observed during heating. However, supercooling has not been investigated for phase transitions between liquid crystalline states at higher temperatures than ambient temperature, such as between the QII and HII phases. It is often stated in the literature that the phase behavior and internal structure of cubosomes and hexosomes reflects that of the parent nondispersed systems. Previously we have reported significant differences between the phase behavior of the bulk and dispersed PHYT-water-F127 system during heating.6 The dispersed system showed an absence of the HII phase which was present in the phase behavior of the bulk system.6 Also, de Campo et al. reported that the MLO-water-F127 dispersed system has a lower HII to L2 transition temperature range than that of the nondispersed system.4 The study also demonstrated that the MLO dispersed system showed similar phase structure during heating (13) Spicer, P. T. Dekker Encycl. Nanosci. Nanotechnol. 2004, 881-892. (14) Fong, C.; Krodkiewska, I.; Wells, D.; Boyd, B. J.; Booth, J.; Bhargava, S.; McDowall, A.; Hartley, P. G. Aust. J. Chem. 2005, 58, 5. (15) Qiu, H.; Caffrey, M. Biomaterials 2000, 21, 223–234. (16) Clogston, J.; Rathman, J.; Tomasko, D.; Walker, H.; Caffrey, M. Chem. Phys. Lipids 2000, 107, 191–220.

Published on Web 04/05/2010

Langmuir 2010, 26(11), 9000–9010

Dong et al.

Article

Figure 1. Chemical structures of glyceryl monooleate (GMO) and phytantriol (PHYT), together with the generic phase behavior observed for these amphiphiles when exposed to excess water.

and cooling and consequently concluded that the internal structure of dispersed LC particles are independent of thermal history and are in thermodynamic equilibrium. The nonequilibrium phase behavior of dispersed GMO-based systems as well as the bulk nondispersed and dispersed phytantriol-based systems during cooling has to our knowledge, not yet been studied. Elevated temperatures are often required or generated in the preparation of lipid-based lyotropic LC systems. The bulk LC systems are often subjected to repeated heating and cooling cycles, in addition to mechanical mixing, to accelerate the lipid/solvent mixing process.3,17 The processes involved in the production of the LC dispersion systems often involve high energy input, which may result in higher than ambient temperatures in the systems.2,6,18,19 Furthermore, to minimize the formation of vesicles after the dispersion process, the dispersed systems are often subjected to heating.19,20 The resulting material may then be used for experimentation at physiological or lower temperatures. A recent study by Lee et al. (2008) showed that differences in liquid crystalline internal structure can have a profound effect on drug release characteristics.21 As such, it is important to understand both the differences between the phase behavior of the bulk and dispersed systems, as well as evaluating their nonequilibrium phase behavior under the influence of temperatures changes, as potential differences in lyotropic phase behavior during cooling compared to heating may affect particle structure and delivery of incorporated agents. Hence, the major aim of this study was to investigate the temperature dependent equilibrium and nonequilibrium lyotropic phase behavior of GMO and phytantriol-based systems using small-angle X-ray scattering (SAXS) on heating and cooling. Furthermore, discovery of significant supercooling effects in the initial part of this study stimulated further investigation of kinetic stability of nonequilibrium phase structures using temperature jump/drop experiments using SAXS and small angle neutron scattering (SANS).

Materials and Methods Materials. Phytantriol used for the SAXS studies was from Roche (Grenzach-Wyhlen, Germany) with nominal purity of >96.6% purity (GC Assay from certificate of analysis no. 01444062). Phytantriol donated by DSM (Basel, Switzerland) (17) Boyd, B. J.; Whittaker, D. V.; Khoo, S. M.; Davey, G. Int. J. Pharm. 2006, 309, 218–226. (18) Gustafsson, J.; Ljusberg-Wahren, H.; Almgren, M.; Larsson, K. Langmuir 1997, 13, 6964–6971. (19) Barauskas, J.; Johnsson, M.; Johnson, F.; Tiberg, F. Langmuir 2005, 21, 2569–2577. (20) W€orle, G.; Siekmann, B.; Koch, M. H. J.; Bunjes, H. Eur. J. Pharm. Sci. 2006, 27, 44–53. (21) Lee, K. W. Y.; Nguyen, T.-H.; Hanley, T.; Boyd, B. J. Int. J. Pharm. 2008, 365, 190–199.

Langmuir 2010, 26(11), 9000–9010

was used for the SANS study with nominal purity >98.3% (certificate of analysis 01724827). Vitamin E acetate (VitEA) and glyceryl monooleate (GMO) were purchased from SigmaAldrich Chemie (Steinheim, Germany). Myverol 18-99K was donated by Kerry Bio-Science (Norwich, NY). The analytical data indicates that Myverol 18-99K (certificate no. 31500455) contains 58.3% glyceryl monooleate (C18:1), 12.2% glyceryl monolinoleate (C18:2), 5.1% glyceryl monolinolenate (C18:3), 3.9% glyceryl monopalmitate (C16:0), 1.7% glyceryl monostearate (C18:0), 0.96% glyceryl monogadoleate (C20:1), 0.2% glyceryl arachidonate (C20:4), 0.1% free fatty acids, and 0.4% glycerol. Trace amounts of unquantified diglycerides are also believed to be present. Pluronic F-127 was purchased from BASF (New Jersey). These chemicals were used as received without further purification. Glass capillaries for SAXS experiments were purchased from Charles Supper (Natick, MA). Quartz cuvettes with 14 mm diameter aperture and 2 mm path length for SANS experiments were supplied by Hellma (Southend on Sea, Essex, U.K.). Milli-Q grade water (0.05 μS cm-1 at 25 °C) purified through a Millipore system (Sydney, Australia) was used for the preparation of samples for the SAXS study. D2O was purchased from Sigma-Aldrich (Sydney, Australia) was used for the preparation of samples for the SANS study. Sample Preparation. The preparation process for both the dispersed and nondispersed liquid crystalline systems have been reported previously.6 Briefly, the nondispersed liquid crystalline systems were prepared by weighing appropriate amounts of lipid and water (50:50 w/w) into glass HPLC vials, the samples were homogenized by repeated cycles of heating to 70 °C, vortex mixing and centrifugation. The samples were then stored at room temperature for at least 1 week prior to being heated to form the low viscosity L2 phase once more and injected into glass capillaries. For dispersed systems, 1.0 g of lipid was weighed into a 20 mL glass vial, then 9.0 g of water or equivalent volume of D2O containing F127 (1% w/w) was added immediately prior to dispersion. Dispersion was achieved by ultrasonication (Misonix XL 2000, Misonix Incorporated, Farmingdale, NY) for 20 min in pulse mode (0.5 s pulses interrupted by 0.5 s breaks) at 40% of maximum power, resulting in a milky dispersion with monomodal particle size distribution and average particle size of 200-300 nm and polydispersity index of T1) between known phase structures confirmed the effectively instantaneous heat transfer into the solution using this approach (see Supporting Information Figure S6). Exposures were taken continuously at 1 s per frame to capture kinetic effects during and after the temperature jump/drop. The delivery rate of solution from T1 to T2 was set to 10 mL/min and the capacity of the needle was 9.25 μL. The delivery quantity was set at 160 μL to allow sufficient sample to be heated, cooled (for T-drop) and reach the X-ray beam to be recorded within 1 s. Data was collected using a one-dimensional position sensitive detector24 with sample to detector distance of approximately 33 and 120 cm, respectively. The kinetics and stability over long time scale for the nonequilibrium dispersed systems were also investigated using a laboratory based SAXS system (Bruker Nanostar). The instrument source was a copper rotating anode (0.3 mm filament) operating at 45 kV and 110 mA, fitted with cross-coupled Montel mirrors, resulting in Cu KR radiation of wavelength 1.54 A˚. The SAXS camera was fitted with a Vantec 2000 2D detector (effective pixel size 60 μm). The optics and sample chamber were under vacuum to minimize air scatter. The sample to detector distance was chosen to be 710 mm, which provided a q-range of 0.010.42 A˚-1. Samples were contained in 2 mm quartz capillaries and temperature controlled by use of a Peltier system accurate to (0.1 °C. Scattering files were integrated using Bruker AXS software v4.1.30 then background subtracted and normalized to sample transmission. The samples were heated to and held at T1 for 15 min (predetermined to be sufficient for equilibration to the nonliquid crystalline L2 phase) then decreased to T2 at 20 °C min-1 and held at the target temperature for 10 min before recording the X-ray diffraction patterns every 15 min for 15 h or more. All SAXS patterns were integrated to the one-dimensional scattering function I(q), where q is the length of the scattering vector, defined by q = (4π/λ)(sin 2θ)/2, λ being the wavelength and 2θ the scattering angle. Applying the appropriate reflection laws the mean lattice parameter a of each LC phase was deduced from the corresponding set of observed interplanar distances d (d = 2π/q). For the L2 phase, which shows only one broad peak, d is termed the characteristic distance. The q scales were calibrated using silver behenate powder at room temperature.25 (22) Amenitsch, H.; Bernstorff, S.; Kriechbaum, M.; Lombardo, D.; Mio, H.; Rappolt, M.; Laggner, P. Appl. Crystallogr. 1997, 30, 872–876. (23) Ollivon, M.; Keller, G.; Bourgaux, C.; Kalnin, D. J. Therm. Anal. Calorim. 2006, 85, 219–224. (24) Petrascu, A.-M.; Koch, M. H. J.; Gabriel, A. J. Macromol. Sci., Part B: Phys. 1998, 37, 463–483. (25) Huang, T. C.; Toraya, H.; Blanton, T. N.; Wu, Y. J. Appl. Crystallogr. 1993, 26, 180–184.

9002 DOI: 10.1021/la904803c

Dong et al.

Small Angle Neutron Scattering (SANS). The kinetic stability of key supercooled dispersed LC systems was also studied using SANS as confirmation of behavior observed in SAXS T-drop experiments. SANS measurements were performed using a 12 mm diameter neutron beam on the LOQ beamline at the ISIS pulsed neutron source (STFC Rutherford-Appleton Laboratory, Didcot, U.K.). LOQ uses pulses of neutrons of wavelengths between 2.2 and 10 A˚, which are separated by time-of-flight and detected by a 64  64 cm, two-dimensional detector at 4.1 m from the sample. Wavelength dependent corrections were made to allow for the incident spectrum, detector efficiencies and measured sample transmissions in order to create a composite SANS pattern as described in detail in Heenan et al. (2007).26 This gives a scattering vector range of q = 0.008 - 0.22 A˚-1. The SANS data were obtained in D2O to reduce incoherent scattering and to maximize the contrast. Quartz Hellma cells filled with LC samples were heated using a Linkam HFS 91 heating stage and a TP-93 temperature programmer (Linkam, Surrey, England) for 5 min at T1 and transferred to a SANS sample holder at T2 with temperature controlled by a circulating water bath. The first pattern was taken for 0-20 or 0-30 min after sample transfer. All SANS patterns were integrated to the one-dimensional scattering function I(q), where q is the length of the scattering vector defined as above. Background subtraction and normalization with respect to a reference standard were performed using the Colette procedure.27

Results Lyotropic Behavior of Nondispersed Systems during Continuous Heating and Cooling. GMO-Based Systems. The nondispersed bulk GMO-based systems in excess water were heated to 95 °C and then cooled to 50 at 1 °C min-1. The individual SAXS profiles and transition temperatures used in bar charts for phase behavior, were obtained from the contour plots for each system, provided in the Supporting Information, Figure S1. Panels A and B of Figure 2 compare representative SAXS I(q) vs q patterns for the bulk “pure” GMO and “commercial” Myverol 18-99K (MVY) systems at specific temperatures acquired during heating and cooling. These profiles illustrate that a supercooling phenomenon was occurring. In Figure 2B, at 60 °C, the HII phase was not observed on the heating cycle, but it was evident coexisting with the QII phase upon cooling. The differences in the phase transition boundaries for the respective systems are represented in bar chart form in Figure 2, panels C and D. During heating, the “pure” GMO in excess water showed the QII(Pn3m) phase between 20 and 92 °C and coexistence of the QII(Pn3m) and HII at 93-95 °C which agreed with the previous study by Briggs et al. (1996).28 During cooling, the QII(Pn3m) to QII(Pn3m) þ HII phase transition for the GMO system was suppressed from above 92 °C during heating to below 89 °C. MYV contains a mixture of monoacylglycerols of which, only approximately 60% w/w is GMO.16 The transition temperatures for the MYV þ water system were lower than those in the “pure” GMO þ water system. During heating, the MYV þ excess water system showed QII(Pn3m) at 20-62 °C and coexisting QII(Pn3m) þ HII at 63-69 °C and HII alone at 70-95 °C. The MYV þ excess water system showed similar suppression of phase transitions as that for the “pure” GMO þ excess water system on cooling, with (26) Heenan, R. K.; Penfold, J.; King, S. M. J. Appl. Crystallogr. 1997, 30, 1140–1147. (27) King, S. M.; Heenan, R. K., The LOQ Instrument Handbook, Vol. 1. In Technical Report RAL-TR-96036; Rutherford Appleton Laboratories: Didcot, U.K., 1996. (28) Briggs, J.; Chung, H.; Caffrey, M. J Phys II Fr. 1996, 6, 723–751.

Langmuir 2010, 26(11), 9000–9010

Dong et al.

Article

Figure 2. Small angle X-ray scattering profiles at specific temperatures for bulk (A) glyceryl monooleate (GMO) and (B) Myverol (MYV) systems in excess water, as a function of temperature during heating (black lines) and cooling (gray lines) at 1 °C min-1. (C and D), schematics summarizing the differences in the phase transition temperatures due to supercooling for the respective systems. Key: QII(Pn3m), light gray; HII, dark gray; L2, HII þ QII(Pn3m), white diagonal stripes.

the QII(Pn3m) þ HII f QII(Pn3m) phase transition being reduced from above 61 °C during heating to below 54 °C during cooling. In addition to the supercooling effect, Figure 2, panels A and B also illustrates the overlapping of the scattering peak positions of the systems during heating and cooling. This indicates that the lattice size of the phases formed is independent of the direction of the temperature change and is independent of the presence of supercooled phases. Phytantriol-Based System. Figure 3 compares the phase behavior of the bulk phytantriol and PHYT þ 3% VitEA systems in excess water during heating and cooling as for the GMO-based systems in Figure 2. A composition with 3% vitamin E acetate was selected as it shows a transition to the HII phase at close to physiological temperatures, giving it relevance in potential drug delivery applications. The suppression of the phase transition compared to the pure phytantriol þ water system was also similar. During heating, both the phytantriol and PHYT þ 3%VitEA systems showed the phase transition from QII(Pn3m) f HII f L2 during heating, with the PHYT þ 3%VitEA system showing earlier onset of the QII(Pn3m) f HII transition temperature than the pure phytantriol system, in agreement with results reported previously.5,6 During cooling, the HII þ L2 f HII phase boundary was suppressed from 68 to 64 °C in the phytantriol system, indicating that supercooling also occurs for the bulk phytantriol-based systems, not previously reported. The PHYT þ 3% VitEA system also showed a suppression of the HII þ L2 f HII phase transition, from 65 to 61 °C, and for the HII þ QII f QII phase transition from 38 to 34 °C. The SAXS curves in Figure 3, Panels A and B, also showed the presence of extra Bragg peaks during cooling at the HII f QII(Pn3m) phase transition boundary that were not present during heating. The extra Bragg peaks suggests the formation of the Langmuir 2010, 26(11), 9000–9010

QII(Ia3d) phase. This was determined by tentatively assigning the peaks as 221 and 220 reflections of the QII(Ia3d) phase with a lattice parameter of 92.9 A˚ at 55 °C. The initial value of 60.0 A˚ for the coexisting QII(Pn3m) phase at 55 °C gives a lattice parameter ratio of 1.55, close to the Bonnet ratio of 1.58.29 (Note that the second peaks for the QII(Ia3d) and QII(Pn3m) phases are not well resolved on cooling in Figure 3A). While not unequivocally confirming that the gyroid phase was present, the two items of evidence lend strong support to the postulate that the extra Bragg peaks during cooling are due to the formation of coexisting QII(Ia3d) phase. For the phytantriol only system (Figure 3, panel A) the purported QII(Ia3d) phase was present at the HII f QII(Pn3m) phase transition boundary at below 62 °C, and was still apparent at the end of the cooling cycle at 30 °C. For the PHYT þ 3%VitEA system, the purported QII(Ia3d) phase was present over a narrower temperature range of 42 °C (HII f QII(Pn3m) phase transition) to 34 °C. The presence of the purported QII(Ia3d) phase is of special significant as the Ia3d phase only exists at equilibrium in these systems under dehydrated conditions where the water content is ∼20-35%,5 while these experiments were conducted in excess water. Panels A and B of Figure 3 also show peak positions at higher q values for the phytantriol-based systems during cooling compared to heating. This indicates that unlike the GMO-based systems, the lattice spacing of the phytantriol-based systems was slightly smaller during cooling than heating suggesting that, in this case, the water ingress may have been suppressed during cooling. Observations presented so far have confirmed that the bulk LC systems also undergo supercooling, not just for the low temperature transition studied by Clogston et al. (2000)16 and Qiu and (29) Barauskas, J.; Landh, T. Langmuir 2003, 19, 9562–9565.

DOI: 10.1021/la904803c

9003

Article

Dong et al.

Figure 3. Small angle X-ray scattering profiles at specific temperatures bulk (A) phytantriol and (B) phytantriol þ vitamin E acetate (PHYT þ VitEA) systems in excess water, as a function of temperature during heating (black lines) and cooling (gray lines) at 1 °C min-1; (C) and (D), schematics summarizing the differences in the phase transition temperatures due to supercooling for the respective systems. Key: QII(Pn3m), light gray; HII, dark gray; L2, black; QII(Ia3d) þ QII(Pn3m), gray cross-hatched; HII þ QII(Pn3m), white diagonal stripes; HII þ QII(Ia3d), white horizontal stripes; HII þ QII(Ia3d) þ QII(Pn3m), white vertical stripes; HII þ L2, gray diagonal stripes.

Caffrey (2000)15 but for all transitions, irrespective of the actual phases or temperature. This study has also shown that nonequilibrium phase structures not seen on heating may be encountered on cooling. Therefore, it is of interest to determine whether this behavior is also evident in the dispersed form of these LC phases. Lyotropic Behavior of Dispersed Systems during Continuous Heating and Cooling. Dispersed GMO-Based Systems. The dispersed “pure” GMO system displayed the QII(Im3m) phase between 25 - 67 °C, coexisting QII(Im3m) þ QII(Pn3m) phases at 68-87 °C and coexisting QII(Pn3m) þ L2 from 89 to 95 °C (Figure 4, panel C). No HII phase was observed in contrast to the parent nondispersed system (Figure 2, panel C). During cooling, the QII(Pn3m) f L2 transition was suppressed from above 89 °C during heating to below 82 °C during cooling. The dispersed MYV system also displayed the QII(Im3m) phase between 25 and 67 °C during heating and converted from QII(Im3m) f HII at 68-72 °C, and HII f L2 at 82-88 °C, (Figure 4, panel C). The HII phase was detected at a narrower range (68-88 °C) than the parent nondispersed system (6395 °C, Figure 2, panel C). During cooling, the MYV system showed suppression of the HII f L2 phase transition from above 82 °C during heating to below 75 °C and the suppression of the transition to pure QII(Im3m) phase from 68 to 54 °C. The extra Bragg peaks observed in the dispersed MYV system during cooling (Figure 4, panel B) that were not present during heating again suggested the formation of a nonequilibrium structure, in this case the QII(Pn3m) phase, as the system progressed from the L2 f HII f QII(Im3m) phase. This was again determined by assigning the peaks to the 110 and 111 reflections of the QII(Pn3m) phase with a lattice parameter of 9004 DOI: 10.1021/la904803c

81.6 A˚ at 60 °C. The initial value of 102.5 A˚ for the coexisting QII(Im3m) phase at 55 °C, gives a lattice parameter ratio of 1.26, close to the Bonnet ratio of 1.28.30 As was observed for the bulk GMO systems the peak positions overlapped during heating and cooling (Figure 4, panels A and B), again indicating that despite the supercooling phenomenon, the actual lattice size of the LC structure formed is independent from the thermal history. However, the presence of the Pn3m cubic phase during cooling for the MYV dispersion suggests that the mechanism for phase transitions in lyotropic LC systems during cooling is not simply the reverse of that which occurs upon heating. Dispersed Phytantriol Systems. The dispersed phytantriol system showed a direct phase transition from QII(Pn3m) f L2 (Figure 5, panel A) on heating without showing an HII phase, in agreement with our previous study.6 During cooling, the transition back to the pure QII(Pn3m) phase was suppressed to below 40 °C, 21 °C lower than the transition on heating. The extra Bragg peaks in the SAXS plot during cooling again suggests the formation of the purported QII(Ia3d) phase as the system progressed from the L2 phase to the QII(Pn3m) phase, similar to that seen for the parent bulk system in Figure 3. This assignment was again supported by the lattice parameter ratio of 1.57 for the purported QII(Ia3d) and QII(Pn3m) phase.30 The QII(Ia3d) phase was not detected during cooling for the dispersed PHYT þ 3% VitEA system. During cooling, the system transformed from L2 f HII f QII(Pn3m), mirroring that observed during heating, but again with significant supercooling of the (30) Hyde, S. T. Curr. Opin. Solid State Mater. Sci. 1996, 1, 653–662.

Langmuir 2010, 26(11), 9000–9010

Dong et al.

Article

Figure 4. Small angle X-ray scattering profiles at specific temperatures for dispersed (A) glyceryl monooleate (GMO) and (B) Myverol (MYV), systems in excess water, as a function of temperature during heating (black lines) and cooling (gray lines) at 1 °C min-1; (C and D), schematics summarizing the differences in the phase transition temperatures due to supercooling for the respective systems. Key: QII(Im3m), white; HII, dark gray; L2, black; QII(Im3m) þ QII(Pn3m), white cross-hatched; QII(Im3m) þ HII, gray cross-hatched; QII(Pn3m) þ HII, vertical stripes; QII(Im3m) þ QII(Pn3m) þ HII, horizontal stripes; QII(Im3m) þ L2, white diagonal stripes; HII þ L2, gray diagonal stripes.

QII f HII transition. The HII f L2 phase transition boundary for the PHYT þ 3%VitEA was also suppressed from above 64 °C to below 49 °C due to supercooling. In contrast to the bulk phytantriol systems in Figure 3, the peak positions did overlap on cooling for the phytantriol-based dispersions indicating equivalent lattice size (Figure 5, panels A and B). As such, the presence of the purported Ia3d cubic phase during cooling of the dispersed phytantriol system signifies that the supercooling effect and nonequilibrium phase structure cannot be simply explained by the suppression of water ingress into the material. Overall, and rather surprisingly, the magnitude of the supercooling effect was greater for the dispersions than the parent bulk nondispersed systems for both GMO and phytantriol-based systems. In order to better understand the kinetic aspects of the supercooling effect, temperature-drop experiments were conducted and described in the following section. In these experiments, rather than continuous heating and cooling being applied, a rapid temperature drop followed by holding the systems at specific temperatures in the supercooled state was used. Rapid Temperature Drop (T-Drop). The bar charts in Figures 4 and 5 were used as the reference to select parameters for the temperature-drop experiments. The initial temperature, T1 was chosen to ensure the dispersed systems exhibit the L2 phase. For the MYV dispersion, T1 was set at 95 °C, while for phytantriolbased systems, T1 was set in the range 70-90 °C. The variation in T1 for the phytantriol system allowed for the investigation of the effect of initial temperature on phase transition kinetics. T2 was set to a temperature either where supercooling was observed in Figures 4 and 5 (i.e., a nonequilibrium phase structure was Langmuir 2010, 26(11), 9000–9010

observed), or where the equilibrium phase structure was observed but at a much lower temperature (Table 1). The specific trends with T1 and T2 are described in the following two sections. Effect of Initial Temperature, T1. The phytantriol dispersion, when subjected to a T-drop from 90-53 °C (where QII(Pn3m) should exist at equilibrium), showed dominance of L2 phase at 1 min post T-drop. By comparison a 75-53 °C T-drop with the same dispersion showed coexisting QII(Pn3m) and the purported QII(Ia3d) phases. Similarly the 75-40 °C T-drop showed coexisting QII(Pn3m) and the purported QII(Ia3d) phases, while 70-40 °C showed the QII(Pn3m) as the dominant phase (see Supporting Information Figure S7 for supporting data). Despite the differences in dominance of phases formed after T-jump, the lattice spacing of the phases agreed with the equilibrium phase structure at that temperature providing further proof that the lattice spacing is independent of thermal history for the dispersions as presented in the previous section using constant heating and cooling. Effect of Final Temperature, T2. Figure 6 shows the X-ray diffractograms of phytantriol (A-C) and MYV (D-F) dispersions over time after T-drop from the L2 phase to various final temperatures T2. Note for all three phytantriol experiments the equilibrium phase at T2 is QII(Pn3m). For panel D, the equilibrium phase for the MYV system is QII(Im3m) þ HII, while for Myverol in panels E and F it is QII(Im3m) alone. The phase structures formed were in good agreement with previous studies using cooling at 1 °C/min. The formation of the nonequilibrium phases, which appeared during supercooling of the respective systems during constant cooling (the purported QII(Ia3d) for phytantriol dispersion in Figure 6, parts A and C, and DOI: 10.1021/la904803c

9005

Article

Dong et al.

Figure 5. Small angle X-ray scattering profiles at specific temperatures for dispersed (A) phytantriol and (B) phytantriol þ vitamin E acetate

(PHYT þ VitEA) systems in excess water, as a function of temperature during heating (black lines) and cooling (gray lines) at 1 °C min-1; (C) and (D), schematics summarizing the differences in the phase transition temperatures due to supercooling for the respective systems. Key: QII(Pn3m), light gray; HII, dark gray; L2, black; QII(Pn3m) þ HII, white diagonal stripes; HII þ L2, gray diagonal stripes; QII(Pn3m) þ L2, gray horizontal stripes; QII(Ia3d) þ QII(Pn3m), gray cross-hatched; QII(Ia3d) þ QII(Pn3m) þ L2, white horizontal stripes. Table 1. Experimental Parameters for T-Drop Studies on Supercooling, and Phases Expected to Form on T-Drop for Selected Systems from T1 to T2 (°C) Based on the Previous Result Using Constant Cooling in the Previous Section (Italicized notations mark non-equilibrium structures) T1

phase at T1

T2

equilibrium phase at T2 (i.e., on heating)

phase at T2 on cooling at 1 °C min-1

phytantriol

90-70

L2

MYV

95

L2

56 53 50 40 30 70 60 50 40

QII(Pn3m) QII(Pn3m) QII(Pn3m) QII(Pn3m) QII(Pn3m) QII(Im3m) þ HII QII(Im3m) QII(Im3m) QII(Im3m)

L2 L2 QII(Pn3m) þ QII(Ia3d) QII(Pn3m) þ QII(Ia3d) QII(Pn3m) HII þ L2 QII(Im3m) þ HII þ QII(Pn3m) QII(Im3m) þ QII(Pn3m) QII(Im3m)

the QII(Pn3m) phase for MYV in Figure 6, parts E and F), were again apparent after the T-drop was performed. Figure 6A shows the changes to SAXS diffractograms for the phytantriol dispersion after the T-drop from 90-53 °C. The L2 phase initially dominated, but slow appearance and growth in intensity of peaks corresponding to the QII(Pn3m) and the purported QII(Ia3d) phases were evident. In contrast Panels C -F show rapid formation of phase structure but little change to relative intensity or position of the Bragg reflections from 1 s to 10 min. This indicated more rapid formation of the liquid crystalline structure from the L2 phase for the Myverol system (