A study of amorphous molecular dispersions of indomethacin and its sodium salt.

Amorphous solid dispersions of indomethacin (IMC) and sodium indomethacin (NaIMC) over a range of compositions were prepared by physically mixing amorphous IMC and amorphous NaIMC, as well as by coprecipitation from methanol solution. Measurement of glass transition temperatures, T(g), for the physical mixtures revealed two values indicating, as expected, phase separation. In contrast, all samples of coprecipitated materials exhibited one value of T(g), which was greater than that predicted for ideal miscibility in the formation of a molecular dispersion. Such nonideality suggests a stronger acid-salt interaction in the amorphous state than that between acid-acid and salt-salt. FTIR spectroscopic analysis provides evidence for interactions between NaIMC and IMC through a combination of hydrogen bonding and ion-dipole interactions between the carboxylic group of the acid and the carboxylate anion of the salt. The inhibition of isothermal crystallization of IMC by NaIMC only when in molecular dispersion is believed to result from the interaction between the acid and the salt, which prevents the formation of hydrogen-bonded carboxylic acid dimers for IMC, required for the formation of crystal nuclei and crystallization.

Water vapor absorption into amorphous sucrose-poly(vinyl pyrrolidone) and trehalose-poly(vinyl pyrrolidone) mixtures.

Previous studies from this laboratory suggested that a solution model (Flory-Huggins equation) modified by a free volume model (Vrentas equation) could satisfactorily describe water absorption into an amorphous solid composed of a sugar or a polymer. This paper has extended the studies of single solutes to binary mixtures of trehalose-and sucrose-poly(vinyl pyrrolidone) (trehalose-PVP and sucrose-PVP, respectively) either co-lyophilized or individually lyophilized and then physically mixed. Water vapor absorption isotherms of the binary mixtures were determined at 30 degrees C. Co-lyophilized PVP-sugar mixtures take up essentially the same amount of water as predicted by the weight average of individual isotherms, whereas sugar crystallization is significant retarded in the molecular dispersions. The sugar-PVP interaction, as reflected in the Flory-Huggins chi interaction parameter, was estimated by fitting the high relative pressure (p/p(0)) region of the isotherm, at which the system is in a liquid state, with a three-component Flory-Huggins-type model. The estimated sugar-water PVP-water, and sugar-PVP interaction parameters suggest that the solute-water interactions are not significantly affected by the sugar-PVP interaction; that is, the solute-water interaction parameters in a binary solute system are similar to those in the corresponding single solute systems. Based on these interaction parameters, the sucrose-PVP interaction appears to be stronger than that of trehalose-PVP. Manipulation of the interaction parameters suggest that the water vapor absorption isotherm is not a sensitive indicator of possible sugar-PVP interactions. Density, glass transition temperature, T(g), and the heat capacity change, DeltaC(p), at T(g) were determined to estimate the excess water absorption energy due to the plasticizing effect of water using the structural relaxation model, as described by Vrentas. Results suggest that PVP is a better antiplasticizer for sucrose than for trehalose. Consequently, the excess free energy arising from structural relaxation was disproportionally reduced by the presence of PVP in these molecular dispersions. Finally, the entire isotherms of co-lyophilized sugar-PVP mixtures are reasonably described with an extended three-component Flory-Huggins model and Vrentas glass structural relaxation model.

Fourier transform Raman spectroscopic study of the interaction of water vapor with amorphous polymers.

Water associated with amorphous polymers is known to affect their chemical and physical properties. The purpose of this study was to investigate the nature of water-polymer interactions for some polymers of pharmaceutical interest. Using Raman spectroscopy, polymer-water hydrogen bond interactions were probed for two molecular weight grades of poly(vinylpyrrolidone), namely PVP K90 and PVP K12, and also for poly(vinylacetate) and poly(vinyl pyrrolidone-co-vinyl acetate). Water vapor absorption isotherms were obtained for the polymers, and the effect of the absorbed water on the glass transition temperature was determined. A knowledge of the water content and physical state of the polymer was used to aid interpretation of Raman spectral changes. The strength of the hydrogen bond formed with water was found to depend on the chemistry of the polymer, with the pyrrolidone group interacting more strongly than the acetate group. However, minor differences were also observed between the degree of interaction of water and polymer for PVP K12 and PVP K90 at some water contents. This result is attributed to differences in the structural relaxation changes accompanying plasticization by water for the two molecular weight grades. Using principal components analysis of the spectral data, it was also possible to differentiate between samples in the rubbery state and samples in the glassy state. In conclusion, water sorbed into polymers causes changes in the polymer Raman spectra not only because of hydrogen bonding, but also as a result of the plasticizing effect of water on polymer mobility.

Effects of lyophilization on the physical characteristics and chemical stability of amorphous quinapril hydrochloride.

PURPOSE: To prepare amorphous quinapril hydrochloride (QHCl) by lyophilization and to compare its physical characteristics and chemical stability as a function of the initial pH of the pre-lyophilized solution. METHODS: Amorphous QHCl samples were prepared by lyophilization from aqueous solutions. Solid-state characteristics were evaluated by DSC, PXRD, and optical microscopy. Chemical degradation was monitored by an HPLC assay. RESULTS: Amorphous QHCl samples obtained from lyophilization exhibited variable glass transition temperatures, depending on the pH and/or concentration of the starting aqueous solutions. Neutralized quinapril (Q) in the amorphous form, which has a Tg of 51 degrees C, lower than that of its HCl salt (91 degrees C), was significantly more reactive than QHCl at the same temperature. The Tg of lyophilized samples prepared at various initial pH values correlated well with values predicted for mixtures of QHCl and Q. Their different reaction rates were related to their glass transition temperature, consistent with the results from earlier studies obtained with amorphous samples made by precipitation from an organic solution and grinding of the crystal solvate. CONCLUSIONS: Lyophilization of different QHCl solutions produces mixtures of amorphous QHCl and its neutralized form Q, with Tg values intermediate to the values of QHCl and Q. As the fraction of Q increases the overall rate of chemical degradation increases relative to QHCl alone, primarily due to the increase in molecular mobility induced by the plasticizing effects of Q.

The relationship between "BET" and "free volume"-derived parameters for water vapor absorption into amorphous solids.

Water vapor absorption isotherms for amorphous solids with the same chemical composition but differing in molecular weight (i.e., PVP-90, PVP-30, and PVP-12), and for glucose, trehalose, and two molecular weight grades of dextran were obtained at 30 degrees C and analyzed using the Brunauer-Emmett-Teller (BET) equation to obtain the parameters, W(m) and C(B). Similar analyses were carried out for the same molecule (e.g., glucose or fructose) at -10 and 40 degrees C. Within each chemical group, W(m), the apparent BET-like parameter that is generally referred to as the "monolayer-limit of absorption", changed very little. In contrast, C(B), a measure of the free energy of absorption, significantly increased with increasing molecular weight or decreasing temperature, leading to a shift from a Type III to a Type II isotherm. The shift in isotherm shape correlates directly with the glass transition temperature, T(g), of the dry sample relative to the operating temperature, T (i.e., Type III when T > T(g) and Type II when T < T(g). These results are shown to be consistent with the combined Flory-Huggins solution model and Vrentas structural relaxation model; wherein Type II isotherm behavior, observed for T < T(g), reflects nonideal volumetric contributions to the overall free energy of absorption due to plasticization by water, as described by Vrentas, whereas Type III behavior only reflects the Flory-Huggins solution model. These volumetric free energy changes within each chemical group are shown to be correlated to the values of the "BET" parameter C(B).

Acid-catalyzed inversion of sucrose in the amorphous state at very low levels of residual water.

PURPOSE: Factors affecting the solid-state acid-catalyzed inversion of amorphous sucrose to glucose and fructose in the presence of colyophilized citric acid, with less than 0.1% w/w residual water, have been studied. METHODS: Samples of citric acid and sucrose were lyophilized at a weight ratio of 1:10 citric acid:sucrose from solutions with initial pH values of 1.87, 2.03, and 2.43, as well as at a weight ratio of 1:5, at an initial pH of 1.87. Glass transition temperatures, Tg, were measured by DSC and the presence of any possible residual water was monitored by Karl Fischer Titrimetry. The inversion of sucrose was measured by polarimetric analysis after reconstitution of solid samples stored at 50 degrees C under P2O5. RESULTS: Samples of 1:10 citric acid:sucrose at an initial pH of 1.87, 2.03, and 2.43 exhibited the same Tg. The initial rate of reactivity was affected at a 1:10 ratio by the solution pH before lyophilization in the order: 1.87 > 2.03 > 2.43 and by citric acid concentration at pH 1.87 in the order 1:5 > 1:10. CONCLUSIONS: Sucrose, colyophilized with an acid such as citric acid, undergoes significant acid-catalyzed inversion at 50 degrees C despite the very low levels of residual water, i.e., <0.1% w/w. At the same ratio of citric acid to sucrose (1:10), and hence the same Tg, the rate of reaction correlates with the initial solution pH indicating that the degree of ionization of citric acid in solution is most likely retained in the solid state. That protonation of sucrose by citric acid is important is shown by the direct relationship between maximum extent of reaction and citric acid composition. It is concluded that colyophilization of acidic substances with sucrose, even in the absence of residual water, can produce reducing sugars capable of further reaction with other formulation ingredients susceptible to reaction with reducing sugars.

Physical characteristics and chemical degradation of amorphous quinapril hydrochloride.

This study was designed to investigate the relationships between the solid-state chemical instability and physical characteristics of a model drug, quinapril hydrochloride (QHCl), in the amorphous state. Amorphous QHCl samples were prepared by rapid evaporation from dichloromethane solution and by grinding and subsequent heating of the crystalline form. Physical characteristics, including the glass transition temperature and molecular mobility, were determined using differential scanning calorimetry, thermogravimetric analysis, powder x-ray diffractometry, polarizing microscopy, scanning electron microscopy, and infrared spectroscopy. The amorphous form of QHCl, produced by both methods, has a T(g) of 91 degrees C. Isothermal degradation studies showed that cyclization of QHCl occurred at the same rate for amorphous samples prepared by the two methods. The activation energy was determined to be 30 to 35 kcal/mol. The rate of the reaction was shown to be affected by sample weight, dilution through mixing with another solid, and by altering the pressure above the sample. The temperature dependence for chemical reactivity below T(g) correlated very closely with the temperature dependence of molecular mobility. Above T(g), however, the reaction was considerably slower than predicted from molecular mobility. From an analysis of all data, it appears that agglomeration and sintering of particles caused by softening of the solid, particularly above T(g), and a resulting reduction of the particle surface/volume ratio play a major role in affecting the reaction rate by decreasing the rate of removal of the gaseous HCl product.

Physical properties of solid molecular dispersions of indomethacin with poly(vinylpyrrolidone) and poly(vinylpyrrolidone-co-vinyl-acetate) in relation to indomethacin crystallization.

PURPOSE: To measure solid-state features of amorphous molecular dispersions of indomethacin and various molecular weight grades of poly(vinylpyrrolidone), PVP, and poly(vinylpyrrolidone-co-vinylacetate), PVP/VA, in relation to isothermal crystallization of indomethacin at 30 degrees C. METHODS: The glass transition temperatures (Tg) of molecular dispersions were measured using differential scanning calorimetry (DSC). FT-IR spectroscopy was used to investigate possible differences in interactions between indomethacin and polymer in the various dispersions. The enthalpy relaxation of 5%w/w and 30%w/w polymer dispersions was determined following various aging times. Quantitative isothermal crystallization studies were carried out with pure indomethacin and 5%w/w polymers in drug as physical mixtures and molecular dispersions. RESULTS: All coprecipitated mixtures exhibited a single glass transition temperature. All polymers interacted with indomethacin in the solid state through hydrogen bonding and in the process eliminated the hydrogen bonding associated with the carboxylic acid dimers of indomethacin. Molecular mobility at 16.5 degrees C below Tg was reduced relative to indomethacin alone, at the 5%w/w and 30%w/w polymer level. No crystallization of indomethacin at 30 degrees C was observed in any of the 5%w/w polymer molecular dispersions over a period of 20 weeks. Indomethacin alone and in physical mixtures with various polymers completely crystallized to the y form at this level within 2 weeks. CONCLUSIONS: The major basis for crystal inhibition of indomethacin at 30 degrees C at the 5%w/w polymer level in molecular dispersions is not related to polymer molecular weight and to the glass transition temperature, and is more likely related to the ability to hydrogen bond with indomethacin and to inhibit the formation of carboxylic acid dimers that are required for nucleation and growth to the gamma crystal form of indomethacin.

Solid-state characteristics of amorphous sodium indomethacin relative to its free acid.

PURPOSE: Having previously studied the amorphous properties of indomethacin (IN) as a model compound for drugs rendered amorphous during processing, we report on the formation and characterization of its sodium salt in the amorphous state and a comparison between the two systems. METHODS: Sodium indomethacin (SI) was subjected to lyophilization from aqueous solution, rapid precipitation from methanol solution, and dehydration followed by grinding to produce, in each case, a completely amorphous form. The amorphous form of SI was analyzed using DSC, XRD, thermomicroscopy and FTIR. The method of scanning rate dependence of the glass transition temperature, Tg, was used to estimate the fragility of the SI system. Enthalpy relaxation experiments were carried out to probe the molecular mobility of the SI system below Tg. RESULTS: The amorphous form of SI formed by different methods had a Tg equal to 121 degrees C at a scanning rate of 20 degrees C/min. This compares with a Tg for indomethacin of 45 degrees C. Estimation of fragility by the scanning rate dependence of Tg indicates no significant differences in fragility between ionized and unionized forms. Enthalpy relaxation measurements reveal very similar relaxation patterns between the two systems at the same degree of supercooling relative to their respective Tg values. CONCLUSIONS: The amorphous form of SI made by various methods has a Tg that is about 75 degrees C greater than that of IN, most likely because of the greater density and hence lower free volume of SI. Yet, the change of molecular mobility as a function of temperature relative to Tg is not very different between the ionized and unionized systems.

The effects of absorbed water on the properties of amorphous mixtures containing sucrose.

PURPOSE: To measure the water vapor absorption behavior of sucrose-poly(vinyl pyrrolidone) (PVP) and sucrose-poly(vinyl pyrrolidone co-vinyl acetate) (PVP/VA) mixtures, prepared as amorphous solid solutions and as physical mixtures, and the effect of absorbed water on the amorphous properties, i.e., crystallization and glass transition temperature, Tg, of these systems. METHODS: Mixtures of sucrose and polymer were prepared by co-lyophilization of aqueous sucrose-polymer solutions and by physically mixing amorphous sucrose and polymer. Absorption isotherms for the individual components and their mixtures were determined gravimetrically at 30 degrees C as a function of relative humidity. Following the absorption experiments, mixtures were analyzed for evidence of crystallization using X-ray powder diffraction. For co-lyophilized mixtures showing no evidence of crystalline sucrose, Tg was determined as a function of water content using differential scanning calorimetry. RESULTS: The absorption of water vapor was the same for co-lyophilized and physically mixed samples under the same conditions and equal to the weighted sums of the individual isotherms where no sucrose crystallization was observed. The crystallization of sucrose in the mixtures was reduced relative to sucrose alone only when sucrose was molecularly dispersed (co-lyophilized) with the polymers. In particular, when co-lyophilized with sucrose at a concentration of 50%, PVP was able to maintain sucrose in the amorphous state for up to three months, even when the Tg was reduced well below the storage temperature by the absorbed water. CONCLUSIONS: The water vapor absorption isotherms for co-lyophilized and physically mixed amorphous sucrose-PVP and sucrose-PVPNA mixtures at 30 degrees C are similar despite interactions between sugar and polymer which are formed when the components are molecularly dispersed with one another. In the presence of absorbed water the crystallization of sucrose was reduced only by the formation of a solid-solution, with PVP having a much more pronounced effect than PVP/VA. The effectiveness of PVP in preventing sucrose crystallization when significant levels of absorbed water are present was attributed to the molecular interactions between sucrose, PVP and water.

Phase behavior of binary and ternary amorphous mixtures containing indomethacin, citric acid, and PVP.

PURPOSE: To better understand the nature of drug-excipient interactions we have studied the phase behavior of amorphous binary and ternary mixtures of citric acid, indomethacin and PVP, as model systems. METHODS: We have prepared amorphous mixtures by co-melting or coprecipitation from solvents, and have measured glass transition temperatures with differential scanning calorimetry. RESULTS: Citric acid and indomethacin in the amorphous state are miscible up to 0.25 weight fraction of citric acid, equivalent to about 2 moles of citric acid and 3 moles of indomethacin. Phase separation, as reflected by two Tg values, occurs without crystallization leading to a saturated citric acid-indomethacin amorphous phase and one essentially containing only citric acid. PVP-citric acid and PVP-indomethacin form non-ideal miscible systems at all compositions. A ternary system containing 0.3 weight fraction of PVP produces a completely miscible system at all citric acid-indomethacin compositions. The use of 0.2 weight fraction of PVP, however, only produces miscibility up to a weight fraction of 0.4 citric acid relative to indomethacin. The two phases above this point appear to contain citric acid in PVP and citric acid in indomethacin, respectively. CONCLUSIONS: Two components of an amorphous solid mixture containing citric acid and indomethacin with limited solid state miscibility can be solubilized as an amorphous solid phase by the addition of moderate levels of PVP.

The molecular mobility of supercooled amorphous indomethacin as a function of temperature and relative humidity.

PURPOSE: To determine the relaxation times of supercooled indomethacin as a function of temperature and relative humidity above Tg, and to analyze the results in the context of being able to predict such behavior at various storage conditions. METHODS: Dielectric relaxation times were measured in the frequency domain (12 to 10(5) Hz) for amorphous indomethacin equilibrated at 0, 56, and 83% relative humidity. The heating rate dependence of Tg for dry supercooled indomethacin was measured with differential scanning calorimetry and used to determine relaxation times. The results were compared with previously published shear relaxation times and enthalpy recovery data. RESULTS: Very good agreement was observed between dielectric and shear relaxation times, and those obtained from the heating rate dependence of the Tg, for dry indomethacin as a function of temperature above Tg. The introduction of water lowered the dielectric relaxation times of supercooled indomethacin without significantly affecting its fragility. The relaxation times below Tg, found to be lower than those predicted by extrapolation of the data obtained above Tg, were analyzed in the context of the Adam-Gibbs-Vogel equation. CONCLUSIONS: The relaxation times of amorphous indomethacin obtained from the heating rate dependence of Tg were in good agreement with those obtained from shear and dielectric measurements, thus validating a relatively simple approach of assessing molecular mobility. The significant molecular mobility of amorphous indomethacin observed below Tg, and the significant plasticizing effects of sorbed water, help to explain why amorphous indomethacin crystallizes well below Tg over relatively short time scales.

Water vapor sorption by peptides, proteins and their formulations.

The interactions of pharmaceutical peptides, proteins and their formulations with environmental water vapor are reviewed. Particular attention is paid to the importance of the physical structure and chemical diversity of peptides and proteins, and comparisons are made with the mechanisms of water vapor sorption by synthetic macromolecular systems. The influences of formulation processes and additives are also considered and suggestions made for future areas of research.

Mixing behavior of colyophilized binary systems.

The purpose of this study was to investigate the factors which govern the mixing of amorphous sucrose with trehalose, poly(vinylpyrrolidone) (PVP), dextran, and poly(vinylpyrrolidone-co-vinyl acetate) (PVP/VA). These materials were chosen as model systems to represent multicomponent freeze-dried pharmaceutical preparations. Mixtures were prepared by colyophilization of the components from aqueous solutions. The glass transition temperatures (Tg) of these mixtures were measured using differential scanning calorimetry (DSC) and were compared to predictions based on simple mixing rules. FT-Raman spectroscopy was used to probe selected mixtures for evidence of molecular interactions between components. Colyophilized mixtures were confirmed to be amorphous by X-ray powder diffraction. The Tg values of the various mixtures generally were lower than values predicted from free volume and thermodynamic models, indicating that mixing is not ideal. The FT-Raman spectra of colyophilized sucrose-PVP and sucrose-PVP/VA mixtures provided evidence for interaction between the components through hydrogen bonding. Hydrogen bonds formed between components in colyophilized sucrose-additive mixtures are formed at the expense of hydrogen bonds within sucrose and in some cases within the additive. A thermodynamic analysis of these mixtures indicates that mixing is endothermic, which is consistent with a net loss in the degree of hydrogen bonding on mixing. There is also a positive excess entropy of mixing which accompanies the net loss in hydrogen bonds. Despite this gain in excess entropy, the excess free energy of mixing is positive, consistent with the observed deviations in Tg from values predicted using models which assume ideal mixing.

The quantitative analysis of crystallinity using FT-Raman spectroscopy.

PURPOSE: To establish if FT-Raman spectroscopy can be used to quantitate the degree of crystallinity in a model compound. METHODS: Mixtures containing different proportions of amorphous and crystalline indomethacin were prepared. Using the peak intensity ratio 1698 cm(-1) (crystalline) to 1680 cm(-1) (amorphous), a correlation curve was prepared. This correlation curve was validated by testing further samples of known composition. Partially crystalline indomethacin was prepared by milling crystalline indomethacin. RESULTS: A linear correlation curve was obtained across the entire range of 0-100% crystallinity. Using this method, it was possible to detect down to either 1% amorphous or crystalline content. The largest errors were found to result from inhomogeneities in the mixing of the calibration and validation samples. The spectra of the mechanically processed samples were similar to the spectra of the calibration samples, and the degree of crystallinity could be estimated in these samples. CONCLUSIONS: FT-Raman spectroscopy is a potentially useful method to complement existing techniques for the quantitative determination of crystallinity.

Effects of a cationic and hydrophobic peptide, KL4, on model lung surfactant lipid monolayers.

We report on the surface behavior of a hydrophobic, cationic peptide, [lysine-(leucine)4]4-lysine (KL4), spread at the air/water interface at 25 degrees C and pH 7.2, and its effect at very low molar ratios on the surface properties of the zwitterionic phospholipid 1,2-dipalmitoylphosphatidylcholine (DPPC), and the anionic forms of 1-palmitoyl-2-oleoylphosphatidylglycerol (POPG) and palmitic acid (PA), in various combinations. Surface properties were evaluated by measuring equilibrium spreading pressures (pi(e)) and surface pressure-area isotherms (pi-A) with the Wilhelmy plate technique. Surface phase separation was observed with fluorescence microscopy. KL4 itself forms a single-phase monolayer, stable up to a surface pressure pi of 30 mN/m, and forms an immiscible monolayer mixture with DPPC. No strong interaction was detected between POPG and KL4 in the low pi region, whereas a stable monolayer of the PA/KL4 binary mixture forms, which is attributed to ionic interactions between oppositely charged PA and KL4. KL4 has significant effects on the DPPC/POPG mixture, in that it promotes surface phase separation while also increasing pi(e) and pi(max), and these effects are greatly enhanced in the presence of PA. In the model we have proposed, KL4 facilitates the separation of DPPC-rich and POPG/PA-rich phases to achieve surface refinement. It is these two phases that can fulfill the important lung surfactant functions of high surface pressure stability and efficient spreading.

The interaction of lung annexin I with phospholipid monolayers at the air/water interface.

Lung annexin I (LAI), a calcium-ion-dependent phospholipid-binding protein, has been shown earlier to cause aggregation and fusion of bilayered vesicles containing phospholipids found in lung surfactant, and to be a very likely factor in the assembly of lung surfactant into the lamellar bodies stored in the Type II cell. In this study, we have measured the accumulation of LAI into spread monolayers of some major lipid components of lung surfactant, dipalmitoyl-phosphatidylcholine (DPPC), dipalmitoyl-phosphatidylglycerol (DPPG), palmitoyl-oleyoyl-phosphatidylglycerol (POPG), and selected mixtures, as a function of calcium-ion concentration and surface concentration (degree of packing) of the phospholipid monolayer. The ability of LAI to significantly penetrate such monolayers was calcium-ion-dependent and only occurred in the presence of DPPG or POPG. The relative extent of penetration into DPPG and POPG was directly related to the available free area in the monolayer, penetration being greater with POPG. Fluorescence microscopy measurements revealed that DPPC mixed with either DPPG or POPG caused a change in surface phase behavior in a manner believed to be related to certain types of bilayer fusion. A chemical breakdown product of LAI, LAI-bp, previously found not to cause aggregation and fusion of bilayers, did not exhibit comparable monolayer penetration or surface phase separation to LAI.

Enthalpy relaxation in binary amorphous mixtures containing sucrose.

PURPOSE: To compare the enthalpy relaxation of amorphous sucrose and co-lyophilized sucrose-additive mixtures near the calorimetric glass transition temperature, so as to measure the effects of additives on the molecular mobility of sucrose. METHODS: Amorphous sucrose and sucrose-additive mixtures, containing poly(vinylpyrrolidone) (PVP), poly(vinylpyrrolidone-co-vinyl-acetate) (PVPNA) dextran or trehalose, were prepared by lyophilization. Differential scanning calorimetry (DSC) was used to determine the area of the enthalpy recovery endotherm following aging times of up to 750 hours for the various systems. This technique was also used to compare the enthalpy relaxation of a physical mixture of amorphous sucrose and PVP. RESULTS: Relative to sucrose alone, the enthalpy relaxation of co-lyophilized sucrose-additive mixtures was reduced when aged for the same length of time at a comparable degree of undercooling in the order: dextran approximately PVP > PVPNA > trehalose. Calculated estimates of the total enthalpy change required for sucrose and the mixtures to relax to an equilibrium supercooled liquid state (deltaHinfinity) were essentially the same and were in agreement with enthalpy changes measured at longer aging times (750 hours). CONCLUSIONS: The observed decrease in the enthalpy relaxation of the mixtures relative to sucrose alone indicates that the mobility of sucrose is reduced by the presence of additives having a Tg that is greater than that of sucrose. Comparison with a physically mixed amorphous system revealed no such effects on sucrose. The formation of a molecular dispersion of sucrose with a second component, present at a level as low as 10%, thus reduces the mobility of sucrose below Tg, most likely due to the coupling of the molecular motions of sucrose to those of the additive through molecular interactions.