Stability of the Metastable Î±-Polymorph in Solid ... - ACS Publications

Article pubs.acs.org/Langmuir

Stability of the Metastable α‑Polymorph in Solid Triglyceride DrugCarrier Nanoparticles Sonja Joseph,† Michael Rappolt,‡,§ Martin Schoenitz,∥ Vera Huzhalska,∥ Wolfgang Augustin,∥ Stephan Scholl,∥ and Heike Bunjes*,† †

Institut für Pharmazeutische Technologie, Technische Universität Braunschweig, Mendelssohnstr. 1, D-38106 Braunschweig, Germany ‡ School of Food Science & Nutrition, University of Leeds, Woodhouse Ln, Leeds LS2 9JT, United Kingdom § Institute of Inorganic Chemistry, Graz University of Technology, Stremayrgasse 6/IV, 8010 Graz, Austria ∥ Institut für Chemische und Thermische Verfahrenstechnik, Technische Universität Braunschweig, Langer Kamp 7, D-38106 Braunschweig, Germany S Supporting Information *

ABSTRACT: Colloidal dispersions of crystalline nonpolar lipids are under intensive investigation as carrier systems in pharmaceutics and nutrition. In this context, the controlled preparation of particles in a metastable polymorphic state is of some interest for the delivery of active substances. In the present study, tristearin particles stabilized with three αpolymorph-preserving emulsifier regimes ((I) sodium glycocholate/saturated long-chain phospholipids, (II) sodium glycocholate, and (III) poly(vinyl alcohol) (PVA)) were investigated concerning the stability of the metastable αpolymorph after controlled crystallization of the particles from the melt. Upon long-term storage, the α-polymorph was preserved best in PVA-stabilized dispersions, followed by those stabilized with the glycocholate/phospholipid mixture and finally those stabilized solely with the bile salt. In particular for rapidly crystallized nanoparticles, the formation of an α-polymorph with highly reduced lamellarity was observed. According to time-/ temperature-resolved synchrotron X-ray diffraction analysis with simultaneous DSC (differential scanning calorimetry) studies, this less-ordered α-polymorph transformed into the common, lamellar α-form upon heating. Although the presence of the lessordered form is probably related to the extraordinarily high stability of the metastable α-polymorph observed in some of the dispersions, it could not completely prevent the transition into the stable β-polymorph. The higher the transition temperature of the less-ordered α-form to the ordered one, the slower was the polymorphic transition to the stable β-polymorph. To estimate the polymorphic stability of the differently stabilized particles upon isothermal long-term storage, standard DSC measurements on samples stored at 23 °C for 4 weeks seem to be of predictive value.

1. INTRODUCTION The use of colloidal solid lipid carrier systems for poorly watersoluble substances is gaining increasing attention in pharmaceutical, cosmetic and food applications.1,2 In particular, triglyceride-based nanoparticles may display high physiological compatibility, which is advantageous for different routes of administration, including injection. Solid triglyceride nanoparticles consist of a core of crystallized triglycerides which is surrounded by an emulsifier shell to render them stable in aqueous dispersion media. Often, they are prepared by highpressure homogenization of the molten matrix lipid in an emulsifier-containing aqueous phase followed by cooling to crystallize the colloidal liquid lipid droplets into solid particles.2 In such nanoparticles, some physicochemical peculiarities of crystalline glycerides have to be taken into consideration, in particular, their potential for ongoing structural alterations of © 2015 American Chemical Society

the colloidal particles upon storage due to the monotropic polymorphism of the triglyceride matrix.3−5 Triglycerides can exist in three key crystal modifications.6,7 Usually, saturated monoacid triglycerides occur in the stable β-polymorph after extended storage in the bulk phase.8,9 Upon crystallization from the melt, the rate of nucleation is lowest for β, intermediate for β′, and highest for α.10,11 After crystallization, the transition into the stable polymorph often occurs faster in colloidally dispersed solid triglycerides than in the bulk phase.8,12,13 The type of matrix triglyceride, the stabilization regime, and also the cooling conditions during crystallization as well as the storage conditions influence the stability of the metastable Received: March 8, 2015 Revised: May 18, 2015 Published: June 1, 2015 6663

DOI: 10.1021/acs.langmuir.5b00874 Langmuir 2015, 31, 6663−6674

Article

Langmuir

2. MATERIALS AND METHODS

polymorphic forms in triglyceride nanoparticles and thus the time course of polymorphic transitions.14−16 Longer-chain triglycerides (e.g., tristearin) transform more slowly into the stable β-polymorph than shorter-chain triglycerides (e.g., trimyristin).14 As amphiphilic molecules, emulsifiers are preferably adsorbed at the lipid−water interface and interact with the lipid matrix.17 For triglyceride nanoparticles stabilized (a) with a blend of the bile salt sodium glycocholate (SGC) and saturated long-chain phospholipids or (b) solely with the bile salt or (c) in particular, with the polymeric stabilizer poly(vinyl alcohol) (PVA), a high stability of the metastable α-polymorph was observed after crystallization.15,16,18,19 Furthermore, in rapidly cooled SGC/saturated long-chain phospholipid-stabilized particles the lamellarity of the α-polymorph appeared to be reduced19 whereas in SGC-stabilized particles a fast crystallization led to the formation of an uncommon form of the α-polymorph displaying virtually no sign of lamellarity.15 Independent of the cooling conditions upon crystallization, in PVA-stabilized tristearin particles an α-polymorph lacking distinct indications for lamellarity was observed, which was stable over several months upon storage at refrigerator temperature.18 It may be hypothesized that the presence of the less-ordered α-form caused the high stability of the metastable α-polymorph and hampered the formation of the stable β-polymorph in these dispersions. Because there are indications that the drug-loading capacity of the metastable polymorphs (α- and β′-forms) might be higher than that of the β-form,13,20 the controlled preparation of triglyceride nanoparticles that are storage-stable in the metastable α-polymorph would be an interesting option from an application point of view. In the field of pharmaceutics, it is, however, very important to know about and to have control over the type of crystalline structure of a drug-carrier system during its whole shelf life. Apart from the drug-loading capacity, other important properties such as the viscosity or the colloidal stability of the dispersions may be affected by the type of crystal modification present in the nanoparticles.21 In this study, a systematic investigation of the polymorphic transitions of colloidal tristearin dispersions emulsified with the three α-form-preserving stabilizer regimes mentioned above was performed. Apart from the type of stabilizer, the influence of the tristearin concentration (5, 10, 15%), the presence of the model drug ubidecarenone, which can be incorporated up to high concentrations into triglyceride nanoparticles,22 and the storage temperature (5, 23 °C) on the long-term stability of the α-polymorph were points of interest. Because rapid cooling rates favor the formation of the less-ordered α-polymorph,15,18 rapidly cooled dispersions might be expected to display a slower transformation into the stable β-polymorph than dispersions prepared at slow cooling rates. To investigate this assumption, very rapidly cooled (10, 70 °C/s) dispersions were compared to those prepared with a slow cooling rate (0.5 °C/ min). To get detailed information on the mechanism of polymorphic transformation and the melting behavior of the particles displaying the metastable α-polymorph, time-/temperature-resolved simultaneous small- and wide-angle synchrotron radiation X-ray diffraction coupled to DSC (differential scanning calorimetry) was applied at different heating rates. Finally, a DSC-based feasibility analysis was tested with the aim of predicting the α-form stability of the differently stabilized tristearin nanoparticles in order to estimate the application period of the dispersions containing the metastable α-form.

2.1. Materials. Tristearin (Dynasan 118, Condea Chemie GmbH, Witten, Germany; melting range 70−73 °C) was used as a matrix lipid for solid nanoparticles. Partially hydrolyzed poly(vinyl alcohol) (PVA) (Mowiol 3-83, Kuraray Specialties GmbH, Frankfurt am Main, Germany), sodium glycocholate hydrate (SGC) (Sigma-Aldrich Chemie GmbH, Steinheim, Germany), or a combination of SGC with hydrogenated soybean lecithin (S100-3) (Lipoid S100-3, Lipoid GmbH, Ludwigshafen, Germany) were used as emulsifiers, ubidecarenone (Q10) (Kyowa Hakko Kogyo Co., Ltd., Tokio, Japan) was used as a model drug, thiomersal (Caesar & Loretz GmbH, Hilden, Germany) as a preservative, and glycerol 85% (Caesar & Loretz GmbH, Hilden, Germany) to adjust the tonicity. Unless otherwise stated, water was freshly prepared by double distillation (doubledistillation water still 2102, GFL, Burgwedel, Germany). 2.2. Preparation of Solid Lipid Nanoparticle Dispersions. Tristearin was used in concentrations of 5, 10, 15, and 20% (w/w) and stabilized with PVA, SGC, or a combination of SGC/S100-3. The mass ratios of matrix lipid to emulsifier depended on the stabilizing agent used. For the above-mentioned tristearin concentrations, (a) PVA was used in concentrations of 2.5, 5, 7.5, and 10%, (b) SGC in concentrations of 1.2, 2.4, 3.6, and 4.8%, and (c) SGC/S100-3 in concentrations of 0.3/1.2, 0.6/2.4, 0.9/3.6, and 1.2/4.8%, respectively (all values are given in w/w concentrations; Table S1, cf., Supporting Information, gives a detailed overview of the dispersions’ composition). For the preparation of drug-loaded dispersions, 5% Q10 (related to the amount of matrix triglyceride) was incorporated into solid lipid nanoparticle dispersions of all stabilizer regimes containing 10% tristearin. The aqueous phase of all samples contained 2.6% glycerol 85% to adjust the tonicity and 0.01% thiomersal as a preservative (both w/w). 150 mL of drug loaded dispersions and 20% matrix lipid-containing dispersions as well as 300 mL of dispersions containing 15% tristearin were prepared by high-pressure melt homogenization. Tristearin was melted at 85 °C on a temperature-controlled magnetic stirrer (IKA RCT basic, IKA-Werke GmbH & Co. KG, Staufen, Germany), and Q10 was dissolved in the molten matrix lipid (for drug-containing dispersions). The stabilizers were dissolved/dispersed in the aqueous phase, and the mixtures were heated to 85 °C and added to the molten lipid phase. The hot mixtures were predispersed with an ultraturrax (Ultra-Turrax IKA T18, IKA-Werke GmbH & Co. KG, Staufen, Germany) at 16 000 rpm for 3 min. The resulting pre-emulsions were processed for 90 (300 mL) or 45 min (150 mL, both corresponding to 30 cycles) in a Niro Soavi homogenizer (Panda 3649, Parma, Italy) equipped with a thermostatic water bath (Haake D1, Thermo Haake GmbH, Karlsruhe, Germany) at 85 °C and a pressure of approximately 350−400 bar. The resulting colloidal emulsion particles were subsequently crystallized by dipping the sample containers under magnetic stirring into an ice−water bath until a temperature of 5 °C was reached. After the cooling step, evaporated water was recharged by reweighing with double-distilled water. Fractions of the dispersions containing 15% tristearin were diluted with preserved and tonicityadjusted water (0.01% thiomersal, 2.6% glycerol 85% (w/w)) to obtain unloaded dispersions containing 5 and 10% matrix lipid. Finally, all colloidal suspensions were stored in a refrigerator. 2.3. Particle Size Determination. PCS (photon correlation spectroscopy) measurements were performed in a Zetasizer Nano ZS (Malvern Instruments Ltd., Worcestershire, U.K.) at 25 °C and an angle of 173° using DTS0012 disposable polystyrol cuvettes (10 × 10 × 45 mm3, Sarstedt AG & Co., Nümbrecht, Germany). The dispersions were diluted with purified, filtered water (EASYpure LF compact ultrapure water system, Barnstead Thermolyne Corporation, Dubuque, IA, USA) to an appropriate concentration of particles to avoid multiple scattering. Four measurements of 300 s each were made after an equilibration period of 5 min. The last three measurements were used to calculate a mean z-average diameter (z-ave) and the polydispersity index (PDI) assuming a viscosity of 0.8872 kg/(m·s) for the aqueous dispersion medium. 6664


Article

Langmuir The absence of coarse particles was verified by laser diffraction (LD) with PIDS (polarization intensity differential scattering) technology and optical microscopy. More details concerning these methods are given in the Supporting Information (section 1.2). 2.4. Controlled Crystallization of Solid Lipid Nanoparticles. Prior to use in stability studies and for synchrotron radiation measurements, the lipid nanoparticles were remelted at 85 °C and crystallized under different, highly defined cooling conditions. To achieve high cooling rates of 10 and 70 °C/s, the micro heat exchanger E1 (Karlsruher Institut für Technologie, Karlsruhe, Germany) was utilized in a continuous process as described earlier23,24 (cf. Supporting Information, section 1.3). The continuously crystallized dispersions were collected in small glass vials (volume 10 mL) and were constantly kept below 5 °C. For comparison, 10 mL samples of the same feed solid lipid nanoparticle dispersions were crystallized (in the same 10 mL glass vials) in a controlled batch process without stirring in a thermostat (ministat 230-cc, Peter Huber Kältemaschinenbau GmbH, Offenburg, Germany) at a conventional cooling rate of 0.5 °C/min from 85 to 5 °C. Solid lipid nanoparticle dispersions containing 5, 10 (unloaded and Q10-loaded), and 15% matrix lipid were stored in a refrigerator (at 4.5−6.5 °C, referred to as “5 °C” in the following text), and at 23 °C, respectively. The samples were characterized directly after well-controlled crystallization as well as after 1, 3, 6, 12, and 24 months by PCS, DSC, and X-ray diffraction. For synchrotron radiation X-ray diffraction studies, dispersions containing 20% matrix lipid were crystallized with 70 °C/s and 0.5 °C/min 7 days prior to the measurements (referred to as “recently crystallized” in the following text) and were stored on an ice−water mixture at 1 to 6 °C until investigation. 2.5. Determination of Matrix Lipid Concentration. To investigate potential dilution with water upon processing, the matrix lipid concentration of all dispersions crystallized in the micro heat exchanger was determined. For this purpose, DSC, refractive index (RI), and total organic carbon (TOC) analysis (cf. Supporting Information, section 1.4) were tested. 2.6. DSC Analysis. Measurements were usually performed with the DSC 1 Stare System equipped with a full range sensor (FRS 5) and a sample robot (Mettler Toledo GmbH, Schwerzenbach, Switzerland). For some analyses in the feasibility study (aimed at predicting the storage stability), a DSC 1 Stare System equipped with a highsensitivity sensor (HSS7, also Mettler Toledo GmbH) was employed. The devices were calibrated with indium and zinc (Mettler Toledo GmbH) by measuring the corresponding melting point (indium 156.6 °C, zinc 419.5 °C) and transition enthalpy (indium 28.5 J/g, zinc 107.5 J/g). The calibration was checked by measuring indium before and after a series of measurements, respectively. Twenty microliters of the dispersions were weighed into 40 μL aluminum crucibles (Mettler Toledo GmbH) and cold sealed. Before heating in the relevant temperature range from 5 to 85 °C at different rates (standard measurements at 10 °C/min; stability scans at 0.25, 0.5, 1, 2.5, 5, and 10 °C/min), an isothermal step of 2 min at 5 °C was inserted to equilibrate the sample temperature. An isothermal step of 5 min at 85 °C was also inserted before the cooling step from 85 to 5 °C (20 °C/ min). All measurements were performed against an empty reference crucible and under nitrogen purge (nitrogen gas 5.0, Westfalen AG Göttingen, Germany). The melting and crystallization enthalpies (normalized to the sample mass) obtained in the heating runs were utilized to determine the fraction of particles in the different polymorphic forms. However, it has to be considered that the evaluation of the integrals of the αendotherm (melting of the α-polymorph) and the immediately following β-exotherm (recrystallization of the β-form with a potential contribution of the β′-form) and β-endotherm (melting of the βpolymorph) might be somewhat inaccurate because all three peaks were overlapping as exemplified in Figure 1. The size of the αendotherm corresponded approximately to the fraction of particles in the α-polymorph. The determination of the fraction of β-form particles was complicated by the fact that recrystallization into the β-form was often observed after the melting of the α-form. Thus, the sum of the enthalpies of the β-exotherm (negative) and β-endotherm (positive)

Figure 1. DSC curves of an SGC-stabilized tristearin dispersion (cooled at 70 °C/s) displaying particles in the different polymorphic forms. (Bottom) α-polymorph (t = 0). (Middle) α-polymorph/βrecrystallization/β-polymorph (t = 4 weeks, 5 °C). (Top) βpolymorph (t = 4 weeks, 23 °C). corresponds to the melting of the fraction of β-polymorph particles which were present before DSC analysis (eq 1). enthalpyβ ‐form[J/g] = enthalpyβ ‐endotherm[J/g] + enthalpy β ‐exotherm[J/g]

(1) To determine the fraction of particles in the β-polymorph in relation to the total tristearin content of the sample, the enthalpyβ‑form values were compared to the melting enthalpy determined for reference samples which contained only particles in the β-polymorph (eq 2). To obtain the β-form reference dispersion, a small amount (1 mL) of each dispersion was tempered in a sealed 2 mL glass vial under stirring for 96 h at 41 ± 1 °C. fractionβ ‐form sample[%] =

enthalpyβ ‐form sample[J/g] × 100 enthalpyβ ‐form reference sample[J/g]

(2)

For the phospholipid-containing samples, the phase transition of the phospholipid at about 45 to 55 °C might lead to an overlap with the melting and crystallization peaks of the lipid particles.16 Because the phospholipid phase transition could not be separated from the melting and crystallization peaks of the lipid particles, the phospholipid phase transition was always included in the integration of the melting and crystallization peaks. Thus, this method might have some uncertainties, and the values obtained should be regarded as approximations. 2.7. Isothermal Wide-Angle X-ray Diffraction Analysis. The polymorphic status of the lipid particles was additionally verified by standard isothermal wide-angle X-ray diffraction (WAXD) analysis at room temperature (25 °C) in order to get an estimation of the crystal structure of the particles in the colloidal dispersions. WAXD diffractograms were obtained with a theta−theta goniometer (PW3050/60 MPD system) equipped with a Pre FIX X′ Celerator detector and installed on a PW3040/60 X′pert Pro generator with a PW3373/00 X-ray tube equipped with a copper anode (all: Panalytical B.V., Ea Almelo, The Netherlands) and a nickel Kβ filter (Philips GmbH, Kassel, Germany). The holder of the stainless steel THCchamber RH 200 (VTI Corporation, Hialeah, FL, USA) was filled with the samples, and measurements were taken at a voltage of 40 kV and an anode current of 40 mA from 3 to 45° 2θ for 20 min. The peak positions were specified using the Bragg equation. 2.8. Time-Resolved Synchrotron X-ray Scattering Coupled to Differential Scanning Calorimetry. Simultaneous small- and wide-angle X-ray scattering (SAXS and WAXS = SWAXS) patterns were recorded at the Austrian SAXS Beamline,25 with a photon energy of 8 keV at the synchrotron light source ELETTRA (Trieste, Italy), which allows time-resolved X-ray scattering experiments.26−28 Two linear 1D position-sensitive detectors (Gabriel type) were used (see 6665


Article

Langmuir Figure S3 for a scheme of the experimental setup) which determined the measured angular range. The SAXS detector was mounted at a sample-to-detector distance of 1.76 m corresponding to an s range (s = 2 sin θ/λ) of about 0.003−0.06 Å−1. The WAXS detector was mounted to cover an s range of 0.11−0.31 Å−1. Silver behenate with a d-spacing value of 58.38 Å was used as a standard to calibrate the angular scale of the measured intensity of the SAXS detector,29 and para-bromo benzoic acid was used for the WAXD detector.30 The inline microcalorimeter developed by the group of Michel Ollivon (CNRS, Paris, France)31 allowed simultaneous DSC measurements of the sample. All samples were filled over a length of 10 mm into capillaries (special-purpose glass mark capillaries, Müller & Müller OHG, Schönwalde-Glien, Germany) with a diameter of 1.5 mm which were sealed with a drop of sealing wax. Before each time-resolved SWAXS experiment, static exposures of the samples were taken for 120 s at 10 °C. Heating runs were performed at different scan rates (1, 2.5, 5, 10 °C/min) from 10 to 85 °C. The subsequent cooling runs from 85 to 10 °C were recorded at a scan rate of 10 °C/min. After the cooling run, an equilibration step from 10 to 11 to 10 °C at a rate of ±0.2 °C/min was added to compensate for the slow response of the calorimeter during cooling and to ensure complete recrystallization of the lipid particles. The detailed time frame protocols for the timeresolved X-ray measurements are listed in Table S3. In addition to the recently crystallized samples containing 20% matrix lipid (cf. section 2.4), rapidly crystallized (70 °C/s) PVA-stabilized lipid dispersions containing 10% (drug-loaded) and 15% (drug-unloaded) tristearin were investigated after 9 months of storage at 5 °C at heating rates of 10 and 1 °C/min.

the polymorphic transition. In the following text, all dispersions will be termed according to the concentration of the respective feed dispersion (nominal concentration) for simplicity; cooling conditions are also specified to differentiate between samples. 3.2. Crystalline Structure of the Particles after Controlled Crystallization. The polymorphic status of the particles was determined by DSC directly after controlled crystallization. As expected from earlier investigations,15,18,22,24 there was an influence of the incorporated model drug, the stabilization regime, and of the cooling rate. A small to moderate fraction of particles in the β-polymorph (β-form particles) was determined in all drug-loaded and SGC/S100-3stabilized samples as well as in all slowly cooled (0.5 °C/min) samples (Figure 2). The incorporation of the model drug had a

3. RESULTS AND DISCUSSION 3.1. Results of Sample Preparation Procedures. After melt homogenization followed by crystallization, colloidal particles with z-average sizes of between about 155 and 190 nm and rather narrow particle size distributions (polydispersity index between 0.07 and 0.13) were obtained (Table S4, after preparation). Subsequently, the particles were recrystallized from the melt under well-controlled conditions either by conventional batchwise cooling in a thermostat (0.5 °C/min) or at two very high rates in a micro heat exchanger (10, 70 °C/ s; the dispersions with 20% lipid content were only cooled at 70 K/min). The particle sizes remained stable after controlled crystallization (t = 0) without obvious influence of the cooling rate (Table S4, t = 0). LD/PIDS measurements confirmed the particle size data obtained by PCS. Neither optical microscopy images nor LD/PIDS measurements gave any indication of particles in the micrometer range (data not shown). Continuous melt crystallization of solid lipid nanoparticles in the micro heat exchanger worked generally well without plugging or particle aggregation. However, water remaining in the microfluidic device from the establishment of thermal equilibrium led to the dilution of the dispersions in spite of discarding a considerable volume of dispersion before sample collection. Thus, the matrix lipid concentration had to be determined after crystallization. DSC, RI, and TOC analysis were tested for this purpose, with DSC turning out to be the most reliable method (cf. Supporting Information, section 2.1). According to DSC determination the dispersions contained only 2.9−4.8% (nominal concentration “5%”), 4.4−8.7% (“10%”), 5.2−11.5% (“15 %”), and 18.3−19.6% (“20%”) matrix lipid, respectively (Figure S5). Because of a larger discarded dispersion volume prior to sample collection, dispersions containing 20% matrix lipid were less diluted than dispersions containing 5, 10, and 15% matrix lipid. The strong and variable dilution limited the possibility to study the influence of matrix lipid concentration on the time course of

Figure 2. Fraction of nanoparticles in the β-polymorph directly after crystallization of the dispersions as determined by DSC employing eqs 1 and 2.

distinct promoting influence on the formation of the stable βpolymorph upon cooling. Moreover, the fraction of β-form particles increased from PVA- to SGC- to SGC/S100-3stabilized dispersions with a rather large amount of β-form particles being detected in SGC/S100-3-stabilized dispersions after slow cooling. An influence of the cooling rate could predominantly be observed for the SGC/S100-3-stabilized dispersions. Slow cooling (0.5 °C/min) promoted the formation of β-form particles, whereas rapid cooling (10 and 70 °C/s) favored the formation of the metastable α-polymorph. Such behavior is generally observed upon crystallization of triglycerides from the melt due to a higher nucleation rate of the least stable α-form at high supercooling.17,32−34 With the exception of two dispersions (15%, PVA-stabilized, 10 °C/s; 15%, SGC-stabilized, 70 °C/s), all rapidly cooled (10 and 70 °C/s) unloaded PVA- and SGC-stabilized dispersions displayed only particles in the metastable α-polymorph. According to isothermal wide-angle X-ray diffraction analysis, all dispersions (including those loaded with Q10) contained particles in the α-polymorph without a detectable fraction of βform particles independent of the cooling rate. As an example, Figure 3 shows the WAXD diffractograms of slowly cooled (0.5 °C/min) unloaded and drug-loaded dispersions (10% matrix lipid) of each stabilization regime in comparison to unloaded reference β-form particles (SGC-stabilized dispersion with 10% tristearin). Isothermal SWAXS measurements clearly revealed differences in the size of the small-angle reflection of the αpolymorph of unloaded dispersions crystallized at different cooling rates (Figure 4). For rapidly cooled (70 °C/s) 6666


Article

Langmuir

α- and sub-α-structures occurring during the crystallization of emulsified milk fat,39,40 the lamellar order seemed to be disturbed by the decreasing size of the emulsion particles.40 The results of these studies indicate that it may be rather difficult to unambiguously detect the additional structural feature of the sub-α-form in the wide-angle diffractograms of dispersed samples. In none of these studies was there, however, a complete absence of small-angle reflections for the αpolymorph. Tentatively, the absence of ordered fatty acid structures (as part of the emulsifiers) in the interface may suppress the formation of lamellar triglyceride structures upon crystallization of PVA- and SGC-stabilized nanoparticles. In recent synchrotron radiation X-ray microbeam studies it was observed that templating effects exerted by surfactants in the oil−water interface may play an important role for the orientation of the triglyceride lamellae within crystallizing oil droplets: in the presence of surfactants containing fatty acid chains, the triglyceride lamellae tended to orient parallel to the surface.41 The effect was particularly pronounced when saturated chains were present. A parallel order of the triglyceride lamellae with respect to the particle interface has also been shown for S100-3/SGC-stabilized α-form tristearin nanoparticles by cryo-TEM and has been attributed to a templating effect of the saturated acyl chains of the phospholipids.19 Moreover, it cannot be completely ruled out that certain structural features of the triglyceride melt are retained in the crystallized droplets exhibiting the less-ordered α-form. The (supra)molecular order of melted triglycerides is under some debate, and three different models for the association of liquid triglycerides have been proposed:42 (I) the smectic liquid-crystal model (small pseudo-lamellar domains) introduced by Larsson,43 (II) the nematic liquidcrystal model (triglyceride molecules in a chairlike (h*) conformation in a nearly parallel, nematic arrangement),44 and (III) the discotic model (triglyceride molecules with splayed chains (y conformers) orientating in stacks).45 Thus, the less-ordered α-form might be crystallized from a liquid structure without or with only little lamellarity as e.g. represented by the discotic model. However, these phenomena are still poorly understood and have to be investigated in more detail in the future. 3.3. Long-Term Investigations of the Polymorphic Transition. After controlled cooling from the melt, tristearin nanoparticles were in the metastable α-polymorph in all dispersions with no or only a small amount of β-form particles according to DSC (Figure 5, t = 0). Because of the high stability of the metastable α-polymorph in the presence of the corresponding stabilizer systems,15,16,18 the transition into the stable β-polymorph took an extended period of time (Figure 5). As expected, the incorporation of model drug Q10, the type of stabilization regime of the colloidal dispersion, and the storage temperature had a large influence on the stability of the metastable α-polymorph upon storage and thus on the time course of the polymorphic transition. Drug loading and storage at 23 °C promoted the transition into the stable β-polymorph in all dispersions. The SGC-containing dispersions displayed a faster polymorphic transition upon storage than the corresponding dispersions stabilized with the blend of SGC/S100-3. Stabilization with PVA led to the slowest transition. Although rapid cooling (70 °C/s) induced the formation of the lessordered α-polymorph to at least some extent in the SGC- and SGC/S100-3-stabilized particles (Figure 4), this did not slow down the transition to the stable β-polymorph upon storage as

Figure 3. WAX diffractograms of dispersions containing 10% tristearin after crystallization with 0.5 °C/min in comparison to reference particles (SGC-stabilized dispersion containing 10% matrix lipid) in the β-polymorph.

Figure 4. SWAXS diffractograms of recently crystallized (0.5 °C/min, 70 °C/s) dispersions containing 20% matrix lipid in comparison to reference particles (an SGC-stabilized dispersion containing 20% matrix lipid) displaying the β-polymorph.

dispersions stabilized with SGC/S100-3 or solely with SGC, a weaker small-angle reflection was observed than for the corresponding slowly cooled (0.5 °C/min) dispersions. The small-angle reflections of PVA-stabilized dispersions were very small and diffuse after cooling at 0.5 °C/min or even hardly observable for dispersions cooled at 70 °C/s. Confirming the results of earlier studies,15,18 the occurrence of these small and diffuse reflections pointed to the presence of an uncommon structure of the α-polymorph which is less ordered than the usual α-polymorph displaying a typical, sharp small-angle reflection as observed for slowly cooled (0.5 °C/min) dispersions stabilized with SGC/S100-3 or SGC. Although the SAXS diffractograms of particles in the unusual less-ordered α-polymorph gave no or little indication of lamellarity, the WAXS diffractograms clearly indicated a crystalline order of the acyl chains within the single triglyceride layers in agreement with earlier results.15,18 On the basis of the X-ray diffraction patterns and transmission electron micrographs of PVA-, SGC-, and SGC/S100-3-stabilized dispersions, different molecular arrangements of the triglycerides have been proposed for the structure of the less-ordered α-polymorph, which has, however, not yet been clarified.15,18 Although less-common forms of the α-polymorph have been found for other triglyceride systems, those were typically observed for triglycerides with more complex structure, in particular, those containing a mixture of saturated and unsaturated triglyceride chains.35−38 For transient 6667


Article

Langmuir

Figure 5. Evolution of the fraction of β-polymorph in unloaded (left) and Q10-loaded dispersions (right) containing 10% tristearin and crystallized at different cooling rates (0.5 °C/min, 10 °C/s, 70 °C/s). Dispersions were investigated over 24 months of storage at 5 and 23 °C. The fraction of βpolymorph was calculated according to eqs 1 and 2 after DSC analysis (samples marked with an asterisk displayed a double endotherm for the αand/or β-polymorph; see also Figure 6).

(10 and 70 °C/s; Figures 4 and 5), the fraction of the βpolymorph increased more rapidly in these dispersions than in those stabilized with SGC/S100-3. Furthermore, SGC/S100-3-

compared to the slowly cooled (0.5 °C/min) dispersions (Figure 5). In spite of unloaded SGC-stabilized particles being completely in the less-ordered α-polymorph after rapid cooling 6668


Article

Langmuir

Figure 6. Exemplified DSC curves of rapidly cooled (70 °C/s), unloaded (stored at 23 °C) (A) and Q10-loaded (stored at 5 °C) (B) PVA-stabilized dispersions (containing 10% tristearin) which might display the presence of the less-ordered and the ordered α-polymorph (A) as well as endotherms of the β′- and β-polymorph (A and B).

6B). Tentatively, the occurrence of two maxima in the endotherm of the α-polymorph may be related to the melting of the two α-forms (less-ordered and ordered) and thus reflect the ongoing transition from the less-ordered to the ordered lamellar α-polymorph in the dispersion. The fact that also for PVA-stabilized tripalmitin nanoparticles different melting points could be observed for the α-polymorph depending on the thermal history supports the assumption that the melting temperature of the less-ordered α-form may deviate slightly from that of the ordered α-form.24 The additional small shoulder around 64 °C in the region of the β-endotherm which appeared upon storage in some samples (Figure 6) might indicate the presence of the β′form, which was also observed for PVA-stabilized triglyceride dispersions in previous studies.24,46 For triglyceride nanoparticles, the transition from the α- to the β-polymorph is expected to be accompanied by a change in particle shape from isometric (spherical to cubelike) to more anisometric (blocklike or plateletlike).15,18,19 Because anisometric particles tend to yield larger PCS particle sizes than corresponding spherical ones,15,18,19 the small differences in PCS particle size observed beyond experimental error were probably caused by changes in the particle shape due to the αto β-transformation (Table S4). In particular for dispersions stabilized with SGC/S100-3 there was a clear influence of the storage temperature on the polymorphic transition and thus on the particle size observed upon storage. Particles stored at 23 °C displayed larger z-average and PDI values than cold-stored dispersions, in agreement with a faster polymorphic transformation at higher temperatures. All dispersions were colloidally stable over a period of 24 months and did not contain larger particles according to LD/PIDS measurements or optical microscopy (data not shown). Information on the dispersions’ chemical stability upon 24 months of storage was obtained by monitoring the pH and can be found in the Supporting Information (sections 1.6 and 2.2). 3.4. Investigations of the Melting Behavior. The high storage stability of the metastable α-polymorph was rather remarkable, in particular for PVA-stabilized nanoparticles (Figure 5), and calls for elucidation of the underlying mechanism. At least for PVA-stabilized samples, the high stability of the α-polymorph may be correlated with the presence of the less-ordered α-form. Because it has been shown that the orientation of the lamellar arrangement in triglycerides

stabilized dispersions displayed a smaller structural difference between corresponding rapidly and slowly cooled samples than samples stabilized solely with SGC (Figure 4) and even contained a small fraction of β-form particles after rapid crystallization (10 and 70 °C/s; Figure 5). The incorporation of the saturated phospholipid as a stabilizer into the tristearin dispersion which is assumed to form a rigid shell around the nanoparticles19 seems to “freeze” their α-crystalline structure and thus slows down the polymorphic alterations. In PVAstabilized dispersions, the less-ordered α-polymorph was formed independently of the cooling rate (Figure 4). In the rapidly (10 and 70 °C/s) crystallized unloaded PVA-stabilized dispersions, a very small fraction ( SGC-stabilized particles) or

only a very low and diffuse (PVA-stabilized particles) SAXS signal of the α-polymorph, indicating the presence of the lessordered α-polymorph (interestingly, small signals that may correspond to the 002 reflection of the α-polymorph could be observed, even in samples with absent or a small 001 reflection). Instead of the melting of the less-ordered αpolymorph upon heating, the transition to the common, lamellar α-polymorph displaying a high small-angle intensity peak was observed in all lipid dispersions. This transition was accompanied by the formation of a much sharper wide-angle reflection leading to a clear WAXS pattern of the hexagonal subcell structure. The transition to the ordered α-polymorph occurred at distinctly higher temperature in the PVA-stabilized nanoparticles than in SGC- and SGC/S100-3-stabilized dispersions. The PVA- and SGC-stabilized particles melted directly from the lamellar α-polymorph at this heating rate, whereas some recrystallization to the stabile β-polymorph was observed for the SGC/S100-3-containing dispersion before final melting. At heating rates lower than 10 K/min, PVA- and SGCstabilized particles also transformed to the stable β-polymorph before final melting as exemplified in Figure 8 for a heating rate of 1 °C/min. These data (Figure 8) also show clearly that the 6670


Article

Langmuir

Figure 8. Contour plots of time/temperature-resolved SWAXS measurements (middle) with selected single SWAXS plots (right; plots were averaged by the sum of five sequent diffractograms) and simultaneously recorded DSC curves (left) of recently crystallized, rapidly cooled (70 °C/ min) lipid dispersions (containing 20% tristearin) obtained upon heating from 10 to 85 °C at 1 °C/min. The green, yellow, and red colors of the contour plots indicate low, medium, and high signal intensity, respectively. The temperature scale of the DSC curve also applies to the contour plot. αlo, less-ordered α-polymorph; αo, ordered α-polymorph.

transformation from the less-ordered to the ordered αpolymorph occurred in a higher temperature range (40 to 50 °C) in the PVA-stabilized dispersion than in the SGC/S100-3(35 to 44 °C) and SGC-stabilized dispersions (32 to 38 °C). Independent of the stabilization regime, an exact transition temperature (around 58 °C) was observed for the formation of the β-polymorph instead of a temperature range indicating differences in the kinetics of the transition of the less-ordered to the ordered α-polymorph and of the ordered α-polymorph to the β-polymorph. The formation of the lamellar α-polymorph was much slower than that of the β-polymorph, indicating that the presence of the less-ordered structure of the α-polymorph might indeed foster the high storage stability of the metastable α-polymorph. The transition ranges/temperatures depended on the heating rate, being slightly higher at faster heating rates. Even at the lowest heating rate (1 °C/min), the formation of the ordered α-polymorph occurred only gradually at elevated temperature (>30 °C, Figure 8). The ordered α-polymorph was apparently most stable in the PVA-stabilized sample for which a clear melting/recrystallization transition into the β-polymorph was detected in the DSC curve when heated at 1 °C/min. For the other two samples (in particular, for the SGC/S100-3-

stabilized one) transformation to the β-polymorph occurred in a less distinct transition. The extraordinarily high stability of the less-ordered αpolymorph in PVA-stabilized particles became even more evident after storage (Figure 9). After 9 months at 5 °C the rapidly cooled (70 °C/s) PVA-stabilized dispersions (containing 15% tristearin in the unloaded sample and 10% in the Q10loaded one) still displayed the characteristic signals of the lessordered α-polymorph in SWAXS diffractograms at room temperature and elevated temperature (approximately up to 40 °C). Upon further heating, no (unloaded) or only a small fraction (Q10-loaded) of particles in the stable β-polymorph was formed. These results are in good agreement with DSC analysis upon storage (Figure 5): no (unloaded) or only a small amount (Q10-loaded) of β-polymorph particles was observed after 6 months in PVA-stabilized dispersions. However, even in unloaded PVA-stabilized particles the presence of the lessordered α-polymorph after crystallization could not completely prevent the formation of the stable β-polymorph in the particles. After 1 year of storage at 5 °C, the first PVA-stabilized particles displaying the stable β-polymorph were observed (Figure 5). 6671


Article

Langmuir

Figure 9. Contour plots of time/temperature-resolved SWAXS measurements (middle) with selected single SWAXS plots (right; plots were averaged by the sum of five sequent diffractograms) and simultaneously recorded DSC curves (left) of stored (9 months), rapidly cooled (70 °C/ min) PVA-stabilized lipid dispersions (containing 15% (unloaded) and 10% (Q10-loaded) tristearin) obtained upon heating from 10 to 85 °C at 10 °C/min. The green, yellow, and red colors of the contour plots indicate low, medium, and high signal intensities, respectively. The temperature scale of the DSC curve also applies to the contour plot. αlo, less-ordered α-polymorph; αo, ordered α-polymorph.

The investigations on the melting behavior confirm that the stabilization regime of the lipid particles has a major influence on the time course of the polymorphic transition. Tentatively, a high transition temperature of the less-ordered to the ordered α-polymorph and a large amount of particles remaining in the ordered α-polymorph upon (slow) heating to its melting point can be an indicator of the high stability of the particles in the αpolymorph upon storage. 3.5. Feasibility of Predicting the Polymorphic Stability. From an application point of view, it would be desirable to predict the stability of a given sample without the requirement of extended storage studies. Therefore, it was investigated whether DSC measurements performed at different scan rates (DSC scan study) could provide information on the stability of the α-polymorph upon long-term storage. Triglyceride polymorphism is monotropic,3−5 and the energy input during DSC scans might thus initiate the polymorphic transition to a different extent depending on the α-form stability of a given sample. For example, the time for polymorphic transitions to more stable forms is increased at decreasing heating rates,33 and melt-mediated transformations occur more rapidly than in the solid state.35 Unfortunately, the DSC scans were not able to reveal the differences between the differently stabilized dispersions in the way as they were observed in the long-term investigations. Furthermore, the DSC scans indicated a potential influence of the cooling rate on the stability of the α-polymorph which was not clearly confirmed in the long-term investigations (Figure 5). Thus, detailed results of the DSC scan study are given only in the Supporting Information (section 2.4, Figures S8 and S9). A DSC scan study as performed here can thus not be used as a screening method to reliably predict the polymorphic stability of differently stabilized tristearin dispersions. One reason may be that the processes occurring during solid-state transformations (isothermal storage) cannot easily be compared to

melt-mediated transformations (thermal treatment) that may be induced during heating.33 As an alternative approach, the evaluation of the standard DSC measurements performed on the dispersions after 4 weeks of storage at 23 °C (cf. Figure 5) can be considered to get an estimate of the stability of the αpolymorph. After 4 weeks it became clear that PVA-stabilized particles were more stable in the α-polymorph than SGC/S1003- and, in particular, SGC-stabilized ones and that drug loading favored the formation of the β-polymorph.

4. CONCLUSIONS Continuous melt crystallization of tristearin nanoparticles in the micro heat exchanger generally worked well. However, residual water from the adjustment of thermal equilibrium led to high dilution effects which require discarding a considerable amount of dispersion prior to sampling. In all dispersions, high cooling rates upon crystallization favored the formation of a lessordered α-polymorph, which was identified by a diffuse and weak SAXS signal and which also displayed a broader WAXS signal than the ordered α-polymorph. Moreover, there were indications that the less-ordered α-polymorph has a slightly different melting point than the ordered one. The stabilizer regime of the particles had a larger influence on the stability of the α-polymorph upon storage than the applied cooling rate upon crystallization for which a clear influence on the transition rate could not be observed in this study. In particular in PVAstabilized particles, the α-polymorph displayed an extraordinarily high stability upon storage over 24 months. Loading with Q10 generally enhanced the α-to-β transition process. The thermal stability of the less-ordered α-polymorph increased from the SGC-stabilized over the SGC/S100-3stabilized to the PVA-stabilized dispersions. The higher the thermal stability of this α-polymorphic form, the slower the αto-β transition upon storage, indicating that the presence of the less-ordered α-polymorph may induce a slow rate of transition 6672


Article

Langmuir into the stable β-polymorph. The structure of the less-ordered α-polymorph and the terms of its occurrence remain to be elucidated in more detail. To predict the long-term stability of the α-polymorph in a normal laboratory setting, short-term stability studies at moderate temperature seem to lead to more reliable results than the evaluation of DSC scans performed at different heating rates. However, the alterations of the αpolymorph upon storage (and heating in DSC) are still poorly understood and call for more detailed investigations, also with regard to particles consisting of other types of matrix lipids.

■

(7) Larsson, K. Classification of glyceride crystal forms. Acta Chem. Scand. 1966, 20, 2255−2260. (8) Bunjes, H.; Westesen, K. Influences of Colloidal State on Physical Properties of Solid Fats. In Crystallization Processes in Fats and Lipid Systems; Garti, N., Sato, K., Eds.; Marcel Dekker: New York, 2001; pp 457−483. (9) Lutton, E. S.; Fehl, A. J. The polymorphism of odd and even saturated single acid triglycerides, C8-C22. Lipids 1970, 5, 90−99. (10) Bayés-García, L.; Calvet, T.; Cuevas-Diarte, M. À .; Ueno, S.; Sato, K. In situ synchrotron radiation X-ray diffraction study of crystallization kinetics of polymorphs of 1,3-dioleoyl-2-palmitoyl glycerol (OPO). CrystEngComm 2011, 13, 3592−3599. (11) Sato, K.; Ueno, S. Crystallization, transformation and microstructures of polymorphic fats in colloidal dispersion states. Curr. Opin. Colloid Interface Sci. 2011, 16, 384−390. (12) Siekmann, B.; Westesen, K. Thermoanalysis of the recrystallization process of melt-homogenized glyceride nanoparticles. Colloids Surf., B 1994, 3, 159−175. (13) Westesen, K.; Bunjes, H.; Koch, M. H. J. Physicochemical characterization of lipid nanoparticles and evaluation of their drug loading capacity and sustained release potential. J. Controlled Release 1997, 48, 223−236. (14) Bunjes, H.; Westesen, K.; Koch, M. H. J. Crystallization tendency and polymorphic transitions in triglyceride nanoparticles. Int. J. Pharm. 1996, 129, 159−173. (15) Bunjes, H.; Koch, M. H. J.; Westesen, K. Influence of emulsifiers on the crystallization of solid lipid nanoparticles. J. Pharm. Sci. 2003, 92, 1509−1520. (16) Bunjes, H.; Koch, M. H. J. Saturated phospholipids promote crystallization but slow down polymorphic transitions in triglyceride nanoparticles. J. Controlled Release 2005, 107, 229−243. (17) Douaire, M.; Di Bari, V.; Norton, J. E.; Sullo, A.; Lillford, P.; Norton, I. T. Fat crystallisation at oil-water interfaces. Adv. Colloid Interface Sci. 2014, 203, 1−10. (18) Rosenblatt, K. M.; Bunjes, H. Poly(vinyl alcohol) as emulsifier stabilizes solid triglyceride drug carrier nanoparticles in theαmodification. Mol. Pharmaceutics 2009, 6, 105−120. (19) Bunjes, H.; Steiniger, F.; Richter, W. Visualizing the structure of triglyceride nanoparticles in different crystal modifications. Langmuir 2007, 23, 4005−4011. (20) Jenning, V.; Schäfer-Korting, M.; Gohla, S. Vitamin A-loaded solid lipid nanoparticles for topical use: Drug release properties. J. Controlled Release 2000, 66, 115−126. (21) Awad, T. S.; Helgason, T.; Kristbergsson, K.; Decker, E. A.; Weiss, J.; McClements, D. J. Effect of cooling and heating rates on polymorphic transformations and gelation of tripalmitin Solid Lipid Nanoparticle (SLN) suspensions. Food Biophys. 2008, 3, 155−162. (22) Bunjes, H.; Drechsler, M.; Koch, M. H. J.; Westesen, K. Incorporation of the model drug ubidecarenone into solid lipid nanoparticles. Pharm. Res. 2001, 18, 287−293. (23) Jasch, K.; Barth, N.; Fehr, S.; Bunjes, H.; Augustin, W.; Scholl, S. A microfluidic approach for a continuous crystallization of drug carrier nanoparticles. Chem. Eng. Technol. 2009, 32, 1806−1814. (24) Schoenitz, M.; Joseph, S.; Nitz, A.; Bunjes, H.; Scholl, S. Controlled polymorphic transformation of continuously crystallized solid lipid nanoparticles in a microstructured device: A feasibility study. Eur. J. Pharm. Biopharm. 2014, 86, 324−331. (25) Amenitsch, H.; Rappolt, M.; Kriechbaum, M.; Mio, H.; Laggner, P.; Bernstorff, S. First performance assessment of the small-angle X-ray scattering beamline at ELETTRA. J. Synchrotron Radiat. 1998, 5, 506− 508. (26) Kalnin, D.; Garnaud, G.; Amenitsch, H.; Ollivon, M. Monitoring fat crystallization in aerated food emulsions by combined DSC and time-resolved synchrotron X-ray diffraction. Food Res. Int. 2002, 35, 927−934. (27) Relkin, P.; Yung, J.-M.; Kalnin, D.; Ollivon, M. Structural behaviour of lipid droplets in protein-stabilized nano-emulsions and stability of α-tocopherol. Food Biophys. 2008, 3, 163−168.

ASSOCIATED CONTENT

S Supporting Information *

Preparation of solid lipid nanoparticle dispersions. Particle size determination. Controlled crystallization of solid lipid nanoparticles. Determination of matrix lipid concentration. Timeresolved synchrotron X-ray scattering coupled to differential scanning calorimetry. pH measurements. Results of samplepreparation procedures. Long-term investigations of the polymorphic transition. Investigations on the melting behavior. Feasibility of predicting the polymorphic stability. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b00874.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +49 531 3915657. Fax: +49 531 3918108. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by grants from the Deutsche Forschungsgemeinschaft (DFG) for the research group F856 mikroPART. The authors acknowledge support from the Forschungszentrum Karlsruhe by providing the micro heat exchanger. The authors are grateful to Ursula Jahn, Kirsten Nebelung, Juliane Schildt, and Manuela Handt (Institut für Pharmazeutische Technologie, Technische Universität Braunschweig, Germany) for technical assistance with isothermal X-ray diffraction analyses and DSC measurements, as well as to Condea Chemie GmbH, Kuraray Specialities GmbH and Lipoid GmbH for providing raw materials.

■

REFERENCES

(1) McClements, D. J.; Li, Y. Structured emulsion-based delivery systems: Controlling the digestion and release of lipophilic food components. Adv. Colloid Interface Sci. 2010, 159, 213−228. (2) Mehnert, W.; Mäder, K. Solid lipid nanoparticles: Production, characterization and applications. Adv. Drug Delivery Rev. 2012, 64, 83−101. (3) Hagemann, J. W. Thermal Behaviour and Polymorphism of Acylglycerides. In Crystallization and Polymorphism of Fats and Fatty Acids; Garti, N., Sato, K., Eds.; Marcel Dekker: New York, 1988; pp 9− 95. (4) Garti, N. Effects of Surfactants on Crystallization and Polymorphic Transformation of Fats and Fatty Acids. In Crystallization and Polymorphism of Fats and Fatty Acids; Garti, N., Sato, K., Eds.; Marcel Dekker: New York, 1988; pp 267−303. (5) Whittam, J. H.; Rosano, H. L. Physical aging of even saturated monoacid triglycerides. J. Am. Oil Chem. Soc. 1975, 52, 128−133. (6) Lutton, E. S. Review of the polymorphism of saturated even glycerides. J. Am. Oil Chem. Soc. 1950, 27, 276−281. 6673


Article

Langmuir (28) Yaghmur, A.; Rappolt, M. Structural characterization of lipidic systems under nonequilibrium conditions. Eur. Biophys. J. 2012, 41, 831−840. (29) Huang, T. C.; Toraya, H.; Blanton, T. N.; Wu, Y. X-ray powder diffraction analysis of silver behenate, a possible low-angle diffraction standard. J. Appl. Crystallogr. 1993, 26, 180−184. (30) Ohkura, K.; Kashino, S.; Haisa, M. Crystal structure of pbromobenzoic acid. Bull. Chem. Soc. Jpn. 1972, 45, 2651−2653. (31) Ollivon, M.; Keller, G.; Bourgaux, C.; Kalnin, D.; Villeneuve, P.; Lesieur, P. DSC and high resolution X-ray diffraction coupling. J. Therm. Anal. Calorim. 2006, 85, 219−224. (32) Himawan, C.; Starov, V. M.; Stapley, A. G. F. Thermodynamic and kinetic aspects of fat crystallization. Adv. Colloid Interface Sci. 2006, 122, 3−33. (33) Sato, K.; Bayés-García, L.; Calvet, T.; Cuevas-Diarte, M. A.; Ueno, S. External factors affecting polymorphic crystallization of lipids. Eur. J. Lipid Sci. Technol. 2013, 115, 1224−1238. (34) Marangoni, A. G.; Acevedo, N.; Maleky, F.; Co, E.; Peyronel, F.; Mazzanti, G.; Quinnc, B.; Pink, D. Structure and functionality of edible fats. Soft Matter 2012, 8, 1275−1300. (35) Ueno, S.; Minato, A.; Seto, H.; Amemiya, Y.; Sato, K. Synchrotron radiation X-ray diffraction study of liquid crystal formation and polymorphic crystallization of SOS (sn-1,3-distearoyl2-oleoyl glycerol). J. Phys. Chem. B 1997, 101, 6847−6854. (36) Bouzidi, L.; Narine, S. S. Relationships between molecular structure and kinetic and thermodynamic controls in lipid systems. Part II: Phase behavior and transformation paths of SSS, PSS and PPS saturated triacylglycerols - Effect of chain length mismatch. Chem. Phys. Lipids 2012, 165, 77−88. (37) Mykhaylyk, O. O.; Martin, C. M. Effect of unsaturated acyl chains on structural transformations in triacylglycerols. Eur. J. Lipid Sci. Technol. 2009, 111, 227−235. (38) Wassell, P.; Okamura, A.; Young, N. W. G.; Bonwick, G.; Smith, C.; Sato, K.; Ueno, S. Synchrotron radiation macrobeam and microbeam x-ray diffraction studies of interfacial crystallization of fats in water-in-oil emulsions. Langmuir 2012, 28, 5539−5547. (39) Lopez, C.; Lesieur, P.; Keller, G.; Ollivon, M. Thermal and structural behavior of milk fat 1. Unstable species of cream. J. Colloid Interface Sci. 2000, 229, 62−71. (40) Lopez, C.; Bourgaux, C.; Lesieur, P.; Bernadou, S.; Keller, G.; Ollivon, M. Thermal and structural behavior of milk fat 3. Influence of cooling rate and droplet size on cream crystallization. J. Colloid Interface Sci. 2002, 254, 64−78. (41) Arima, S.; Ueno, S.; Ogawa, A.; Sato, K. Scanning microbeam small-angle X-ray diffraction study of interfacial heterogeneous crystallization of fat crystals in oil-in-water emulsion droplets. Langmuir 2009, 25, 9777−9784. (42) Iwahashi, M.; Kasahara, Y. Dynamic molecular movements and aggregation structures of lipids in a liquid state. Curr. Opin. Colloid Interface Sci. 2011, 16, 359−366. (43) Larsson, K. Molecular arrangement in glycerides. Fette, Seifen, Anstrichm. 1972, 74, 136−142. (44) Cebula, D. J.; McClements, D. J.; Povey, M. J. W.; Smith, P. R. Neutron diffraction studies of liquid and crystalline trilaurin. J. Am. Oil Chem. Soc. 1992, 69, 130−136. (45) Corkery, R. W.; Rousseau, D.; Smith, P.; Pink, D. A.; Hanna, C. B. A case for discotic liquid crystals in molten triglycerides. Langmuir 2007, 23, 7241−7246. (46) Schaal G. Untersuchungen einer Befilmungsmöglichkeit fester Arzneiformen mit modifizierten Triglycerid-Dispersionen. Ph.D. Thesis, Jena, 2004. (47) Ueno, S.; Nishida, T.; Sato, K. Synchrotron radiation microbeam X-ray analysis of microstructures and the polymorphic transformation of spherulite crystals of trilaurin. Cryst. Growth Des. 2008, 8, 751−754. (48) Kalnin, D.; Schafer, O.; Amenitsch, H.; Ollivon, M. Fat crystallization in emulsion: Influence of emulsifier concentration on triacylglycerol crystal growth and polymorphism. Cryst. Growth Des. 2004, 4, 1283−1293. 6674


Stability of the Metastable Î±-Polymorph in Solid ... - ACS Publications

Recommend Documents