Annealing Effect Reversal by Water Sorption–Desorption and Heating

Article pubs.acs.org/molecularpharmaceutics

Annealing Effect Reversal by Water Sorption−Desorption and Heating above the Glass Transition TemperatureComparison of Properties A. Saxena,†,⊥ Y. C. Jean,‡ and R. Suryanarayanan*,† †

Department of Pharmaceutics, University of Minnesota, Minneapolis, Minnesota 55455, United States Department of Chemistry, University of MissouriKansas City, Kansas City, Missouri 64110, United States

‡

ABSTRACT: Our objective is to compare the physical properties of materials obtained from two different methods of annealing reversal, that is, water sorption−desorption (WSD) and heating above glass transition temperature (HAT). Trehalose was annealed by storing at 100 °C for 120 h. The annealing effect was reversed either by WSD or HAT, and the resulting materials were characterized by differential scanning calorimetry (DSC), water sorption studies, and positron annihilation spectroscopy (PAS). While the products obtained by the two methods of annealing reversal appeared to be identical by conventional characterization methods, they exhibited pronounced differences in their water sorption behavior. Positron annihilation spectroscopy (PAS), by measuring the fractional free volume changes in the processed samples, provided a mechanistic explanation for the differences in the observed behavior. KEYWORDS: amorphous, trehalose, annealing, glass transition temperature, water sorption, differential scanning calorimetry, free volume, positron annihilation spectroscopy

■

INTRODUCTION A number of new chemical entities under pharmaceutical development are characterized by limited aqueous solubility. Numerous strategies have been developed to enhance the solubility and, more specifically, the dissolution rate of pharmaceuticals. Transforming a crystalline compound into its amorphous counterpart has been an effective approach to enhance the dissolution rate.1 Amorphous compounds, which are characterized by short-range lattice periodicity, have higher free energy than the corresponding crystalline forms.2 This energy difference drives the difference in solubility and, from a practical viewpoint, the dissolution rate. It can also bring about differences, often pronounced, in stability. In many pharmaceuticals, the stability (chemical or physical) has been coupled to the molecular mobility.3−7 In the supercooled liquid region (i.e., above the glass transition temperature), the global mobility is believed to be responsible for the reactivity and instability of amorphous phases.8,9 This mobility dramatically decreases below the glass transition temperature (Tg). However, at temperatures substantially below Tg, local molecular motions can bring about physical and chemical instability.10−12 In addition to the effect on stability, at temperatures close to Tg, the mobility in the glassy state is evident from the decrease in enthalpy, entropy, and volume as a function of time. This is often referred to as physical aging. The structural relaxation time is a measure of the molecular mobility. In pharmaceuticals, enthalpic recovery quantified by differential scanning calorimetry is widely used as a measure of structural relaxation. In polymers, the refractive © 2013 American Chemical Society

index, extent of sorption, and mechanical properties are some of the response parameters used to measure the extent of physical aging. The change in refractive index brought about by polymer densification was further related to the fractional free volume.13,14 The change in physicochemical properties brought about by physical aging can significantly influence the product performance. One possible consequence can be the loss of the advantages conferred by the high energy amorphous state, for example, the enhancement in dissolution rate.15 Different approaches have been used to reverse the aging effect. One of the conventional methods is heating the aged amorphous sample above its Tg (HAT). However, this can cause degradation in thermolabile compounds. In polymers, an alternate method to reverse the aging effect is the sorption of soluble vapors, usually under elevated pressure, but at ambient temperature. While aging causes densification of glass, exposure to soluble gas under high pressure dilates it.16 When the pressure is removed, this dilated glass fails to collapse back to its original state, thereby reversing the aging effect. Thus soluble vapor at high pressure, by causing swelling or dilation, opposes the tendency of volume reduction during aging.17 Both of these methods of aging effect reversal, that is, heating above Received: Revised: Accepted: Published: 3005

February 22, 2013 May 16, 2013 June 12, 2013 July 8, 2013 dx.doi.org/10.1021/mp400099r | Mol. Pharmaceutics 2013, 10, 3005−3012

Molecular Pharmaceutics

Article

Tg and vapor sorption−desorption, have been pioneered in polymers and later successfully used in pharmaceuticals.18 In a recent investigation, water sorption followed by desorption (WSD) was found to reverse the aging effect.18 In this method, the aging reversal was brought about by continuously maintaining the material in the glassy state. In contrast, the conventional method requires heating to the supercooled liquid state (T > Tg). As explained earlier, water vapor sorption is believed to cause matrix expansion, while desorption at room temperature did not result in spontaneous densification. This is because of the low molecular mobility at room temperature, which is substantially below the plasticized Tg (dry Tg of trehalose = 117 °C). The amount of water sorbed, and by extension the extent of aging reversal, depends on water vapor pressure (relative humidity) in the atmosphere and the storage time. However, the kinetics of aging reversal, which is of great practical importance, has not been investigated. We hypothesize that the major mechanism of annealing effect reversal is through changes in the fractional free volume. In the first method (i.e., HAT), since the sample is heated to the supercooled liquid state, structural changes affecting the fractional free volume can be expected. In the WSD method, sorption of water vapor can form new “holes” to increase the fractional free volume. Positron annihilation spectroscopy (PAS) will be used to measure changes in the fractional free volume. This technique has been a versatile tool for quantifying small changes in free volume in polymers.19 In PAS, a positron (e+), antimatter to electron (e−), is released from a radioactive source. Positron can annihilate with electrons directly with a lifetime of 0.4−0.5 ns or may pick up an electron from the molecular medium and annihilate in different ways. It can annihilate as a free positron, be trapped into defects, and then annihilate or form a bound state with an electron of the medium and called the positronium (Ps) atom. Ps exists in two different spin states, parapositronium (p-Ps) in which spins of e+ and e− are antiparallel or orthopositronium (o-Ps) with parallel spins. In vacuum, p-Ps has a shorter lifetime of 0.125 ns and annihilates via two photons, while o-Ps annihilates via three photons with a lifetime of 142 ns. However, in molecular media, the positron of o-Ps picks an electron and annihilates faster to shorten its lifetime to a few nanoseconds. Such annihilations, also known as pick-off annihilation, result in two-photon emission. Due to its relatively small size (1.59 Ǻ ) and localization, Ps can be considered a microprobe of free volume holes. The measured o-Ps lifetime and intensity are respectively related to the free volume hole size and number of holes in the system.20,21 Aging of an amorphous pharmaceutical (API or excipient) can bring about undesirable changes in properties. Heating above Tg (HAT) and water sorption−desorption (WSD) are two possible strategies to reverse the aging effect. The products resulting from these two widely different sample treatments can be expected to exhibit pronounced differences in properties. However, there has been no direct comparison of the product properties following aging reversal by these two methods. In summary, while amorphization of pharmaceuticals is an increasingly desired strategy to enhance bioavailability, the instability associated with this high energy state can be a major impediment. Moreover, the extent of relaxation and the associated free energy change as a function of time can influence the pharmaceutically relevant properties (for example, kinetics of water sorption) and thereby affect manufacturability as well as the final product performance. While the aging effect

can be reversed, the specific method used for this purpose can affect the final product properties. Our first objective is to compare the physical properties of materials obtained from two different methods of aging reversal, that is, water sorption−desorption and heating above Tg. In addition to PAS, differential scanning calorimetry (DSC) and water sorption studies were used to characterize the materials. Our second objective was to evaluate the kinetics of aging effect reversal brought about by water sorption. These experiments were conducted at room temperature (RT), under different water vapor pressures. The water sorbed was quantified as a function of the water vapor pressure in the atmosphere and related to the extent of the aging effect reversal.

■

MATERIALS AND METHODS Preparation of Amorphous Trehalose. α,α-Trehalose dihydrate (C12H22O11·2H2O, Sigma, St. Louis, MO, USA) was used as received. Amorphous trehalose was prepared by lyophilization (Unitop 400 L, Virtis, Gardiner, NY, USA). Aqueous trehalose solution (10% w/v) was cooled from RT to −45 °C over a period of 80 min and held for 150 min. The frozen solution was heated at 0.1 °C/min to −30 °C under reduced pressure (100 mTorr), and primary drying was carried out for 38 h. The temperature was then gradually increased to 50 °C, and secondary drying was conducted for 31 h. The final drying was conducted at 60 °C for 24 h. The lyophile was stored at −20 °C in a desiccator over anhydrous calcium sulfate until use. The freeze-dried trehalose was always handled in a controlled humidity environment (12 h), the decrease in available free volume considerably slows down the relaxation. The increase in Tg, observed as a function of aging time, has been predominantly attributed to structural rearrangement and the consequent changes in free volume. During aging of poly(DL-lactide), the density of the cohesional entanglements increases, and the parallel packing of the local neighboring segments becomes stronger, leading to increase in Tg.28 The conformational changes during aging caused a decrease in free volume and mobility of the glassy starch chains and polyimide, raising the Tg.29,30 In the case of maltose, a disaccharide, the observed increase in Tg was attributed to the shift in relaxation endotherm to higher temperatures, brought about by aging.31 A similar effect was reported in annealed fructose, wherein the

(1)

where ro is equal to (r + Δr) and Δr is an empirical parameter (1.656 Ǻ ) determined by fitting eq 1 to the known hole size in porous materials.22 In this model, free hole is considered as a spherical potential well having an infinite potential barrier of radius (r + Δr) with an electron layer in the region of r < x < ro. Ps is assumed to be localized in this spherical potential well. This equation is valid for o-Ps lifetimes < 10 ns. The average size of free volume holes is calculated using Vf =

4 3 πr 3

(3)

(2)

where r is calculated from eq 1. 3007

dx.doi.org/10.1021/mp400099r | Mol. Pharmaceutics 2013, 10, 3005−3012

Molecular Pharmaceutics

Article

Figure 2. Plot of primary y-axis enthalpic recovery (ER) and secondary y-axis corresponding water sorption (WS) of annealed trehalose as a function of time following storage at different RH conditions at room temperature (n = 3). The annealing effect reversal and water sorption profile for the first 4 h are shown in the inset.

Figure 3. Water uptake profiles of freeze-dried trehalose unaged, annealed (120 h), and annealing effect reversed by WSD and HAT. The samples were exposed to 10% RH (25 °C) in an automated water sorption analyzer.

approach, the annealed trehalose was heated above its Tg (HAT). In the second method, which was described earlier from our laboratory, controlled water sorption followed by desorption (WSD) reversed the effect of annealing. HAT will completely reverse the annealing effect. On the other hand, WSD provides an approach to control the degree of reversal, based on the amount of water sorbed. Kinetics of Annealing Effect Reversal by Water Sorption Desorption. The effect of annealing was completely reversed by WSD.18 The annealing reversal was accomplished by storing at 30% RH (RT) which resulted in 6.4% (w/w) water sorption. Storage at lower RH values resulted in partial reversal, though the extents were not evaluated. By conducting these studies at different RH values, it was possible to determine the minimum RH at which complete annealing reversal could be accomplished in the experimental time scale.

enthalpic recovery over the glass transition region caused the increase in its dynamic Tg.32 Using modulated temperature DSC (MDSC), which separates total heat flow into reversible and irreversible heat flow signals, it was possible to separate the glass transition event from enthalpic recovery. In this technique, where the underlying heating rate was 1 °C/min, the Tg of annealed trehalose (116 °C) continued to be higher than that of unaged trehalose (114 °C). When compared with the MDSC results, the increase in Tg with aging time is much more pronounced in conventional DSC. The reason for this difference is that, in MDSC, the observed Tg is not strongly affected by the enthalpic recovery.33,34 This confirmed that the observed increase in Tg of the annealed sample was not an experimental artifact. Two methods were employed to reverse the annealing effect. In the first method, which is the conventional and widely used 3008

dx.doi.org/10.1021/mp400099r | Mol. Pharmaceutics 2013, 10, 3005−3012

Molecular Pharmaceutics

Article

spectroscopy (PAS) was used to measure small changes in the fractional free volume brought about by the different sample treatments. Positron Annihilation Spectroscopy (PAS). PAS is capable of determining the size, distribution (see Figure 4),

Freeze-dried trehalose, annealed for 120 h, was stored at the desired RH. Samples withdrawn at the desired time periods were divided into two aliquots. The water content was determined by TGA (Figure 2). The second aliquot was completely desorbed (water content < 0.5% w/w), and the enthalpic recovery was measured by DSC (Figure 2). Storage at 40% RH for 3 h resulted in complete reversal of the annealing effect. When the RH of storage was reduced to 32 and 21%, storage times of 4 and 16 h, respectively, were required to achieve the same effect. The amount of water sorption required to bring about complete reversal was 5.3, 4.4, and 3.4% w/w at 40, 32, and 21% RH, respectively. Thus, with decreasing storage RH, less water is required to reverse the annealing effect. From a kinetic perspective, the rate of water sorption and the annealing effect reversal increased as a function of RH. Comparison of Annealing Effect Reversed Trehalose Obtained from Different Methods. It was of interest to determine the effect of annealing reversal method (HAT vs WSD) on the physical properties of the resultant product. Likewise, a comparison of the properties of annealing-reversed products obtained by storage under three different RH was also warranted. In this context, we wish to reiterate that all of these samples are “identical” by DSC. They all exhibit nearly identical glass transition temperature (∼117 °C) with no enthalpic recovery following the Tg. Water Vapor Sorption. The thermal history of amorphous trehalose will influence its interaction with water vapor.18 Figure 3 provides the water uptake profiles of trehalose samples subjected to different treatments. As reported earlier, annealing caused a pronounced decrease in the extent of water sorption. This can be attributed to an increase in density and decrease in the free volume during annealing. Next, the water sorption profiles of annealing-reversed trehalose obtained from two different methods (WSD and HAT) were compared. While the unaged and the WSD trehalose exhibited substantially similar water sorption behavior, HAT sorbed less water. Thus the method of annealing reversal appears to influence the water sorption behavior. The annealing effect was reversed by storage at three different RH values (Figure 2). The resultant materials, when exposed to 10% RH (25 °C), exhibited virtually identical water sorption behavior (results not shown). Thus the condition of WSD does not seem to affect the water sorption behavior. As pointed out earlier, the water uptake behavior of the annealing-reversed trehalose obtained from WSD was comparable to that of unaged trehalose, whereas HAT trehalose sorbed less water. Differences in particle morphology could bring about this effect. The particle size and shape of the samples, characterized by SEM, were substantially similar. The surface area values did not exhibit pronounced differences (the specific surface area values: HAT: 0.44 m2/g; WSD: 0.50 m2/ g). The method used for annealing reversal could have affected the matrix. Aging reversal is expected to cause an increase in free volume and a consequent decrease in density. However, the overall effect for the two methods is expected to be different. These differences can be monitored by density measurement. However, the changes in density brought about, first by annealing and then by annealing reversal, are expected to be very small. Quantifying these small differences will not be possible using the conventional techniques for density measurement of powder samples. Instead positron annihilation

Figure 4. Positronium lifetime distribution probability density function (PDF) for trehalose freeze-dried unaged, annealed, and annealing reversed by WSD and HAT.

and fractional free volume in samples. Results of a representative finite-lifetime analysis of a lifetime spectrum are summarized in Table 1. The p-Ps lifetime (τ1) was fixed at 0.125 ns. Free positron annihilated with electrons directly in both unaged and annealed trehalose in about 0.4−0.5 ns (referred as τ2; data not shown). The o-Ps pick-off lifetime (τ3) in trehalose is shorter (1.2 ns) than in common polymers (1.4− 3.0 ns). At room temperature the drop-off in the fractional free volume of trehalose was 1.1−1.4%, while in polymers the reported range is 2.5−10%.20,35 Effect of Annealing on Fractional Free Volume. Annealing caused a small decrease in the o-Ps lifetime (τ3) with no appreciable change in intensity (I3 = 22%). The free volume radius (r) and free volume (Vf), calculated using eqs 1 and 2, respectively, decreased with annealing. As is evident from Table 1, with no WSD treatment, annealing caused a decrease in fractional free volume, and this decrease was brought about by the decrease in τ3 and not by any changes in I3 (eq 3). The decrease in τ3 and r3 can be attributed to shrinkage of free volume size during annealing. However, I3 did not change as there was no change in the number of free volumes. The increase in full width at half-maximum (fwhm) suggested that the free volume distribution became more heterogeneous during annealing due to nonsymmetric shrinkage of the free volume size. Effect of Annealing Reversal by WSD on Fractional Free Volume. As explained earlier, WSD reverses the annealing effect. This “annealing-effect reversed trehalose” sorbed more water than annealed trehalose which can be explained by the free volume difference (Figure 3; Table 1). Penetration of small water molecules during WSD increased the free volume radius and also created new holes, reflected by an increase in τ3 and I3, respectively. The positronium formation probability and the free-volume hole concentration determine the value of I3. The increase in the number of holes 3009

dx.doi.org/10.1021/mp400099r | Mol. Pharmaceutics 2013, 10, 3005−3012

Molecular Pharmaceutics

Article

Table 1. Positron and Positronium Lifetime (τ3), Intensity (I3), Mean Free Volume Radius (r), Free-Volume Hole Size (Vf), Fractional Free Volume (FFV), and Full-Width at Half-Maximum (FWHM) for Freeze-Dried, Annealed, and AnnealingReversed Trehalose sample freeze-dried trehalose annealed trehalose annealing reversed by WSD annealing reversed by HAT annealing reversed by WSD followed by HAT

τ3 (ns) 1.14 1.13 1.17 1.12 1.16

± ± ± ± ±

0.07 0.10 0.07 0.07 0.08

I3 (%) 22.28 22.35 25.71 24.32 24.71

± ± ± ± ±

r (Å)

1.95 2.57 1.99 2.23 2.26

(I3) coupled with the increase in radius (r3) brought about by WSD, caused the pronounced increase in fractional free volume. As a result, there was an increase in the free volume distribution reflected by an increase in fwhm. Effect of Annealing Reversal by HAT on Fractional Free Volume. Annealing reversal by HAT involves heating the annealed trehalose above its glass transition temperature (to the rubbery state) followed by cooling back to room temperature. Rapid cooling to room temperature enables retention of a part of the free volume generated in the rubbery state. The faster the cooling rate, larger is the fraction of this free volume retained in the glass. Annealing reversal by HAT did not change the free volume size but increased the number of holes, resulting in an increase in the fractional free volume. The variability in the PAS results can be attributed to a number of factors. The positron source, which decays over time, directly affects the absolute numbers of the positron lifetime and intensity. The electronics can be a source of variability since the detector is sensitive to the environment, specifically the RH and temperature. It is therefore more meaningful to look at the trends in the PAS results as a function of various processing steps rather than the absolute numbers. The changes observed in the fractional free volume (Table 1) can be explained by the sample treatment. While annealing caused a decrease in the fractional free volume, the magnitude of effect of the annealing reversal depended on the method used. Only WSD caused a pronounced increase in the fractional free volume which became evident from its tendency to absorb much more water than HAT trehalose. These methoddependent reversal effects will not become evident by the conventional pharmaceutical techniques such as DSC. Reannealing of Annealing-Reversed Trehalose. Annealing-reversed trehalose obtained by HAT and WSD was reannealed at 100 °C for the different time periods to determine the kinetics of enthalpic recovery as a function of annealing time led us to compare the behavior of the annealingreversed sample with that of fresh amorphous trehalose (Figure 5). Irrespective of the annealing reversal method, the kinetics of enthalpic recovery was slower during reannealing. The slower relaxation in annealing-reversed trehalose illustrated the differences in the properties of fresh amorphous (unaged) trehalose from that of annealing-reversed trehalose. Interestingly, the annealing reversal method did not seem to have an appreciable influence over the results obtained. Annealing-reversed trehalose was reannealed for 120 h at 100 °C, and the water sorption behavior of the resulting product was compared with that of freshly annealed trehalose. These three samples exhibited essentially identical water sorption behavior (Figure 6). A decrease in τ3 was observed for the annealed as well as the reannealed trehalose (HAT and WSD) when compared with their respective starting samples (Table 2). Interestingly,

1.88 1.86 1.92 1.85 1.91

± ± ± ± ±

Vf (Å)

0.10 0.14 0.10 0.11 0.11

27.84 26.98 29.80 26.34 28.97

± ± ± ± ±

4.59 6.08 4.50 4.72 5.03

FFV (%) 1.12 1.09 1.38 1.15 1.29

± ± ± ± ±

0.28 0.37 0.32 0.31 0.34

fwhm (ns) 0.26 0.32 0.34 0.31 0.33

± ± ± ± ±

0.06 0.08 0.06 0.06 0.07

Figure 5. Kinetics of enthalpic recovery as a function of annealing time in fresh (unaged) trehalose. Trehalose was annealed at 100 °C for 120 h, and the annealing effect was reversed, either by HAT or WSD. The annealing-reversed samples were reannealed at 100 °C and the kinetics of enthalpic recovery as a function of annealing time was evaluated.

annealing had no effect on the I3 values. These observations suggest that, while annealing decreased the size of free volume, the number of free volumes was unaffected. The effect of annealing on the fractional free volume is evident when the results of the first (unaged; FFV = 1.116%) and second (annealed for 120 h; FFV = 1.086%) samples are compared. Similar effects were observed for the annealing reversed by water sorption (samples 3 and 4) and annealing reversed by HAT samples (samples 5 and 6). Annealing caused a pronounced decrease in the extent of water sorbed by amorphous pharmaceuticals. While DSC can reveal the extent of relaxation caused by annealing, it does not provide any information about microstructural changes in the matrix. Using PAS, we have observed a pronounced decrease in the fractional free volume which provides an explanation for the observed change in sorption behavior. This technique was also very useful to understand the annealing effect reversal. In polymers, soluble gas sorption has been extensively used as a technique to reverse the annealing effect. In pharmaceuticals, on the other hand, HAT has been the preferred method for annealing reversal. For the first time, the annealing effect has been reversed by both HAT and WSD, and the properties of the resulting products were compared with that of fresh unaged trehalose. DSC, a technique of choice to characterize amorphous pharmaceuticals, was unable to distinguish between these samples. While the water uptake behavior of the annealing-reversed trehalose obtained by WSD was comparable to that of unaged trehalose, HAT trehalose had a much lower tendency to sorb water. PAS results effectively explained the observed difference in behaviorWSD caused a pronounced increase in the fractional free volume, while HAT did not. 3010

dx.doi.org/10.1021/mp400099r | Mol. Pharmaceutics 2013, 10, 3005−3012

Molecular Pharmaceutics

Article

Figure 6. Water sorption behavior of trehalose annealed at 100 °C for 120 h following storage at 10% RH (25 °C). The annealing effect was reversed, either by HAT or WSD, and the samples were reannealed at 100 °C for 120 h. The water sorption profiles of these reannealed samples were obtained. The legends refer to trehalose annealed for 120 h, followed by annealing effect reversal by WSD/HAT, followed by reannealing for 120 h (120_WSD_120/120_HAT_120).

Table 2. Positron and Positronium Lifetime (τ3), Intensity (I3), and Fractional Free Volume (FFV) for Annealed and Reannealed Trehalose along with Their Respective Starting Samples sample no.

sample

1 2 3 4 5 6

unaged 120 h 120_WSD 120_WSD_120 120_HAT 120_HAT_120

τ3 (ns) 1.142 1.128 1.171 1.081 1.118 1.099

± ± ± ± ± ±

I3 (%)

0.07 0.10 0.07 0.04 0.07 0.03

While dealing with amorphous pharmaceuticals, the conventional practice is to “erase” thermal history by heating above Tg. Since a significant fraction of amorphous materials are thermolabile, the water sorption−desorption technique can be an effective alternative. While these two methods erase the thermal history, the resulting products can have pronounced differences in their microstructural properties which in turn can have a pronounced impact on product manufacturability, stability, and performance. The conventional techniques used for physical characterization are unlikely to reveal these microstructural differences. We have demonstrated the power and utility of PAS to provide a mechanistic explanation for the differences in the behavior of the unaged (i.e., freshly prepared), annealed, annealing-effect reversed, and reannealed trehalose.

22.28 22.35 25.71 30.08 24.32 24.78

± ± ± ± ± ±

1.95 2.57 1.99 1.43 2.23 0.98

FFV (%) 1.116 1.086 1.379 1.298 1.153 1.120

± ± ± ± ± ±

0.28 0.37 0.32 0.21 0.31 0.14

fractional free volume changes in the processed samples, provided a mechanistic explanation for the differences in the observed behavior.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: 612-624-9626. Fax: 612626-2125. Present Address ⊥

Biocon-BMS R&D Center, Syngene International Limited, Bangalore-560099, Karnataka, India.

Notes

The authors declare no competing financial interest.

■

■

ACKNOWLEDGMENTS Ajay Saxena thanks Upsher-Smith Laboratories, Maple Grove, MN for partial support. The work was partially supported by the William and Mildred Peters Endowment fund. We thank Dr. Hongmin Chen for the help with the experimental work and Micromeritics Analytical Services for the assistance in the Surface Area Measurements.

CONCLUSION Heating above the glass transition temperature (HAT) has been the method of choice to reverse the effects of annealing of amorphous pharmaceuticals. We have demonstrated that water sorption−desorption (WSD), at room temperature, is an effective technique to reverse annealing effects. The kinetics of annealing effect reversal, as a function of water vapor pressure in the atmosphere, was evaluated. Interestingly, the properties of the product depended on the method (HAT or WSD) used for annealing effect reversal. Annealing effect reversal by WSD resulted in a product with properties nearly identical to that of the unaged material. Therefore, WSD could become the desired method to reverse the annealing effects. Positron annihilation spectroscopy (PAS), by measuring the

■

REFERENCES

(1) Hancock, B. C.; Parks, M. What is the true solubility advantage for amorphous pharmaceuticals? Pharm. Res. 2000, 17, 397−404. (2) Hilden, L. R.; Morris, K. R. Physics of amorphous solids. J. Pharm. Sci. 2004, 93, 3−12. (3) Abdul-Fattah, A. M.; Dellerman, K. M.; Bogner, R. H.; Pikal, M. J. The effect of annealing on the stability of amorphous solids: chemical

3011

dx.doi.org/10.1021/mp400099r | Mol. Pharmaceutics 2013, 10, 3005−3012

Molecular Pharmaceutics

Article

stability of freeze-dried moxalactam. J. Pharm. Sci. 2007, 96, 1237− 1250. (4) Bhugra, C.; Pikal, M. J. Role of thermodynamic, molecular, and kinetic factors in crystallization from the amorphous state. J. Pharm. Sci. 2008, 97, 1329−1349. (5) Bhugra, C.; Shmeis, R.; Krill, S. L.; Pikal, M. J. Prediction of onset of crystallization from experimental relaxation times. II. Comparison between predicted and experimental onset times. J. Pharm. Sci. 2008, 97, 455−472. (6) Zhou, D.; Zhang, G. G. Z.; Law, D.; Grant, D. J. W.; Schmitt, E. A. Physical stability of amorphous pharmaceuticals: importance of configurational thermodynamic quantities and molecular mobility. J. Pharm. Sci. 2002, 91, 1863−1872. (7) Zhou, D.; Grant, D. J. W.; Zhang, G. G. Z.; Law, D.; Schmitt, E. A. A calorimetric investigation of thermodynamic and molecular mobility contributions to the physical stability of two pharmaceutical glasses. J. Pharm. Sci. 2007, 96, 71−83. (8) Liu, J.; Rigsbee, D. R.; Stotz, C.; Pikal, M. J. Dynamics of pharmaceutical amorphous solids: the study of enthalpy relaxation by isothermal microcalorimetry. J. Pharm. Sci. 2002, 91, 1853−1862. (9) Pikal, M. J. Mechanisms of protein stabilization during freezedrying and storage: the relative importance of thermodynamic stabilization and glassy state relaxation dynamics. Drugs Pharm. Sci. 2004, 137, 63−107. (10) Vyazovkin, S.; Dranca, I. Physical Stability and Relaxation of Amorphous Indomethacin. J. Phys. Chem. B 2005, 109, 18637−18644. (11) Vyazovkin, S.; Dranca, I. Effect of Physical Aging on Nucleation of Amorphous Indomethacin. J. Phys. Chem. B 2007, 111, 7283−7287. (12) Bhattacharya, S.; Suryanarayanan, R. Local mobility in amorphous pharmaceuticals-characterization and implications on stability. J. Pharm. Sci. 2009, 98, 2935−2953. (13) Huang, Y.; Paul, D. R. Physical Aging of Thin Glassy Polymer Films Monitored by Optical Properties. Macromolecules 2006, 39, 1554−1559. (14) Huang, Y.; Wang, X.; Paul, D. R. Physical aging of thin glassy polymer films: Free volume interpretation. J. Membr. Sci. 2006, 277, 219−229. (15) Guo, J. H.; Robertson, R. E.; Amidon, G. L. Influence of physical aging on mechanical properties of polymer free films: the prediction of long-term aging effects on the water permeability and dissolution rate of polymer film-coated tablets. Pharm. Res. 1991, 8, 1500−1504. (16) Chan, A. H.; Paul, D. R. Influence of history on the gas sorption, thermal, and mechanical properties of glassy polycarbonate. J. Appl. Polym. Sci. 1979, 24, 1539−1550. (17) Berens, A. R.; Hodge, I. M. Effects of annealing and prior history on enthalpy relaxation in glassy polymers. 1. Experimental study on poly(vinyl chloride). Macromolecules 1982, 15, 756−761. (18) Surana, R.; Pyne, A.; Suryanarayanan, R. Effect of Aging on the Physical Properties of Amorphous Trehalose. Pharm. Res. 2004, 21, 867−874. (19) Mallon, P. E. Applications to polymers. In Principles and applications of positron & positronium chemistry; Jean, Y. C., Mellon, P. E., Schrader, D. M., Eds.; World Science Publisher: Singapore, 2003; pp 253−277. (20) Jean, Y. C.; Chen, H.; Lee, L. J.; Yang, J. Positron annihilation lifetime and glass transition temperatures in CO2 sorption polystyrene. Acta Phys. Pol. A 2008, 113, 1385−1395. (21) Lue, S. J.; Lee, D.-T.; Chen, J.-Y.; Chiu, C.-H.; Hu, C.-C.; Jean, Y. C.; Lai, J.-Y. Diffusivity enhancement of water vapor in poly(vinyl alcohol)-fumed silica nano-composite membranes: Correlation with polymer crystallinity and free-volume properties. J. Membr. Sci. 2008, 325, 831−839. (22) Vorenkamp, E. J.; Challa, G. Influence of chlorination of poly(vinyl chloride) on miscibility with poly(methyl methacrylate). Polymer 1988, 29, 86−92. (23) Shukla, A.; Peter, M.; Hoffmann, L. Analysis of positron lifetime spectra using quantified maximum entropy and a general linear filter. Nucl. Instrum. Methods Phys. Res., Sect. A 1993, 335, 310−317.

(24) Nakaniski, H.; Wang, S. J.; Jean, Y. C. Microscopic surface tension studies by positron annihilation. In Positron annihilation studies of fluids; Sharon, S. C., Ed.; World Science Publisher: Singapore, 1988; pp 292−298. (25) Tao, S. J. Positronium annihilation in molecular substances. J. Chem. Phys. 1972, 56, 5499−5510. (26) Eldrup, M.; Lightbody, D.; Sherwood, J. N. The temperature dependence of positron lifetimes in solid pivalic acid. Chem. Phys. 1981, 63, 51−58. (27) Surana, R.; Pyne, A.; Suryanarayanan, R. Effect of Preparation Method on Physical Properties of Amorphous Trehalose. Pharm. Res. 2004, 21, 1167−1176. (28) Liao, K.; Quan, D.; Lu, Z. Effects of physical aging on glass transition behavior of poly(DL-lactide). Eur. Polym. J. 2001, 38, 157− 162. (29) Chung, H.-J.; Lim, S.-T. Physical aging of glassy normal and waxy rice starches: effect of aging time on glass transition and enthalpy relaxation. Food Hydrocolloids 2003, 17, 855−861. (30) Fu, Y.-J.; Hsiao, S.-W.; Hu, C.-C.; Qui, H.-z.; Lee, K.-R.; Lai, J.Y. Effect of physical aging on sorption and permeation of small molecules in polyimide membranes. Desalination 2008, 234, 58−65. (31) Schmidt, S. J.; Lammert, A. M. Physical aging of maltose glasses. J. Food Sci. 1996, 61, 870−875. (32) Truong, V.; Bhandari, B. R.; Howes, T.; Adhikari, B. Physical aging of amorphous fructose. J. Food Sci. 2002, 67, 3011−3018. (33) Tonchev, D.; Kasap, S. O. Effect of aging on glass transformation measurements by temperature modulated DSC. Mater. Sci. Eng., A 2002, A328, 62−66. (34) Hutchinson, J. M.; Tong, A. B.; Jiang, Z. Aging of polycarbonate studied by temperature modulated differential scanning calorimetry. Thermochim. Acta 1999, 335, 27−42. (35) Sandreczki, T. C.; Hong, X.; Jean, Y. C. Sub-Glass-TransitionTemperature Annealing of Polycarbonate Studied by Positron Annihilation Spectroscopy. Macromolecules 1996, 29, 4015−4018.

3012

dx.doi.org/10.1021/mp400099r | Mol. Pharmaceutics 2013, 10, 3005−3012