Determination of Polymorph Enthalpy Difference by Solution Calorimetry

Introduction & Background

Polymorphism is a common phenomenon that exists in many pharmaceutical like drug molecules.  Depending on the thermodynamics and kinetics of the conversion, form changes may occur during many stages of drug manufacturing. Understanding the thermodynamic relationship between polymorphs can be critical to drug development since polymorphs may exhibit different physical & chemical properties, such as stability [1], solubility [2], dissolution rate [3] and bioavailability [4].

The Burger Rule, which compares the melting temperatures and heats of fusion (ΔHf) of the forms (measured by differential scanning calorimetry (DSC), is often used to determine the thermodynamic relationship between polymorphs.  However, if a material sublimes upon heating, decomposes immediately after heating, or if the melting endotherms are not well resolved (concurrent multiply phase changes), it may not possible to obtain an accurate value for the heat of fusion.

Because enthalpy is a state function, solution calorimetry (SolCal) can be used to determine the heat of solution of individual polymorphs and, thereby, the difference in enthalpy between forms can be accurately determined. SolCal measures the heat generated or consumed  when a solid or liquid sample  is dissolved or diluted into a solvent.  The absolute values of heats of solution of polymorphs are different  for each solvent, but the enthalpy difference between the two forms (ΔHtrans) remains constant.  The enthalpy difference between the heats of fusion of the two forms is the same as what is determined using DSC.

For this work, ROY (5-methyl-2-[(2-nitrophenyl)amino]-3-thiophenecarbonitrile), was selected as a model compound.  Polymorphism in ROY has been well studied and at least eight solvent-free polymorphs (e.g. Y, OP, ON, R, YN, ORP, YT04, and Y04) have been identified [5]. The naming convention follows the color differences between the polymorphic forms.  The ROY system is ideal for showcasing the power of solution calorimetry to determine thermodynamic relationships as there are a large number of forms and form conversions during heating in the DSC can make direct interpretation of the data difficult.

The thermodynamic relationship between polymorphs of ROY are complicated: many of the  polymorphs can crystallize simultaneously from the same solution and are kinetically stable under the same condition; some pairs of forms are monotropically related and some are enantiotropically related (stable form changes with temperature). Prof. Lian Yu et.al. have reported free-energy differences, and enthalpy differences for ROY polymorphs using melting and eutectic melting data from DSC [5].  However, some of the forms, such as form YN, undergoes rapid solid-solid transition upon heating and therefore its melting temperature and heat of fusion cannot be measured directly or accurately.  In the literature, there have been attempts to calculate the energy difference between ROY polymorphs using computational models, but there is no satisfactory agreement between these calculations and experimental data [6].  Therefore, it is necessarily to acquire experimental data using a different analytical technique to determine the enthalpy differences between polymorphs in the ROY system.

Objective

The goal of this study was to generate pure polymorphs of ROY and perform SolCal analysis on them to measure the heats of solution in order to determine the enthalpy differences between the forms.

This poster presents SolCal data on four ROY polymorphs (Y, ON, R and YN) in two solvents (DMA and DMF) using a TA TAMIII Precision Solution Calorimeter at 25 ℃.  Forms ON, R and YN were generated in-house using form Y as-received as the starting material. The generated materials were analyzed by XRPD and optical microscopy to evaluate the phase purity. Enthalpy differences between ROY polymorphs were calculated from heats of solution in both solvents.   The enthalpy difference between Y and ON was also calculated from heats of fusion using Hyper DSC. These data were compared to the literature data from DSC that had been collected at 10 ℃/min.

Preparation & Characterization of ROY

Generate pure polymorphs of ROY at large scale for SolCal tests is challenging and required numerous crystallization attempts. In many cases polymorphs were observed to crystallize simultaneously from the same liquid yielding mixtures of forms. Among them, Form YN was particular difficult to isolate since it converts to R and Y in hours to days at ambient temperature.

Pure forms of ON, R and YN were generated and isolated in bulk as below:

  • Form ON– Temperature cycled between 112.5 and 114.5 ℃, or slow cooled IPA solution from 95 ℃ to ambient temperature
  • Form YN– Crash-cooled from IPA solution, isolated by vacuum filtration in small batch and dried by Npurge
  • Form R– Converted from YN using spatula scraping solids on the filter paper

The polymorphs used in this work were phase pure by XRPD and the observations of representative samples under optical microscopy were consistent with the XRPD results.

Results & Discussion

Using DSC at a 10 ℃/min heating rate, Form YN undergoes rapid solid-solid conversion upon heating [5] making direct measurement of its melting point using this heating rate impossible (melting point of YN was estimated as 98 ℃ by extrapolating the GYN curve to the liquid curve [5]). By using a higher heating rate, 50 ℃/min, the solid-solid transition was successfully suppressed and the melting of YN solids was experimentally observed at 94 ℃.  However, the heat of fusion of YN could not be accurately measured due to the concurrent melting-recrystallization events.  Attempts were made to use Hyper DSC at a heating rate of 300 ℃/min to suppress form conversion between Y and ON (for which enthalpy data exists). However, the measured enthalpy difference between the two forms (1.9 kJ/mol) was not in good agreement with the literature and further solution calorimetry work (see below) and the method was abandoned.

SolCal analysis were performed on ROY polymorphs of Y, ON, R, and YN. Analysis was done at 25 ℃ in two solvents, DMA and DMF. Solids were packed into a 1-mL glass ampoule, which was sealed and placed into a holder equipped with stirring and hammer function. The holder was then inserted into the reaction vessel containing 100 mL of solvent and the vessel was placed into a thermostat at a constant temperature.  The dissolution process was initiated by breaking the glass ampoule to release solids into solvent after the calibration baseline was established.  Based on the SolCal response, the dissolution of ROY is an endothermic process for each polymorph. For each test, two heats of solution were obtained-one calculated using a calibration preceding the sample analysis and one calculated using a calibration following the sample analysis.  The mean values from the two calibrations were calculated and summarized in Table 1.

 

Although the magnitude of the individual heat of solution depends on the solvent used, in both solvents, the heats of solution follows the same order (YN < ON < R < YN). Using form Y as a reference, the enthalpy differences (ΔHtrans) between forms ON, R, YN and Y can be calculated and compared to literature ΔHvalues from fitting melting and eutectic melting data.

Measured enthalpies ΔHY-YN and ΔHY-ON were consistent with literature values [5]. For ΔHY-R, there is a discrepancy between literature data and SolCal measurement.  The SolCal data is self consistent between solvents ( 0.72 kJ/mol in DMA and 0.78 kJ/mol in DMF) and such good agreement provides evidence of the reliability of SolCal analysis compared to the previously estimated values.

Mixture Analysis

Another parameter that can be determined using SolCal is the phase purity of the drug.  Materials composed primarily of Form ON with some Y and Form R with Y were generated and tested in DMA. Because Form Y has a higher heat of solution, the measured heats of solution for both mixtures were higher than the corresponding pure Forms of ON and R.

Assuming a linear relationship for SolCal response in the mixtures, the amount of Form Y in both lots is 22% and 38% respectively.  An estimate of the amount of Form Y in both lots was also done using Rietveld refinement on the XRPD patterns for the mixtures.  Using this analysis the mixtures were 27% and 11% Y, respectively.  However, there are difficulties with both analyses.  Rietveld suffers from preferred orientation issues (especially with needle morphologies) whereas the value of the difference in the enthalpies is limited by the instrumental sensitivity for small amounts of form impurities if the absolute differences between the enthalpies of the pure forms are not sufficiently different.

Conclusions

Pure forms of ROY polymorphs, ON, R, and YN were successfully produced from Form Y as received. The melting endotherm of YN was experimentally observed at 94 ℃ (onset) by DSC at 50 ℃/min.  Solution calorimetry analysis were performed on Forms Y, ON, R, and YN at 25℃ in two solvents, DMA and DMF.  The rank order of enthalpies from both solvents were consistent (Y > R > ON > YN).  The enthalpy differences between the forms, ΔHY-R, ΔHY-ON, and ΔHY-YN were obtained from the heats of solution and was independent of solvent. ΔHY-ON, and ΔHY-YN were in good agreement with the estimations from DSC in the literature, whereas the measured ΔHY-R was about half the DSC estimate.  Mixture analysis on two lots were also done and showed agreement and discrepancies between the percent form impurity (Form Y) using both SolCal and Reitveld analysis.

Acknowledgments

The help and support from Susan Bogdanowich-Knipp, David A Engers, Stefanie Schwab, Claire Gendron, Lisa Edwards, Aaron Atkinson, and SSCI Analytical Resources is greatly appreciated.

Originally presented by Jing Teng, Karen Gushurst, Kevin Leach, Stephan X.M. Boerrigter, and Jon Selbo at the 2013 AAPS Annual Meeting and Exposition in San Antonio, TX.

References

[1]: Byrn SR et.al. Chemical reactivity in solid-state pharmaceuticals: formulation implications. Adv. Drug Deliv. Rev., 2001, 48(1):115–136.
[2]: Pudipeddi M, and Serajuddin AT. Trends in solubility of polymorphs. J. Pharm. Sci, 2005, 94, 929-939
[3]: Tuladhar MD et.al. Thermal behavior and dissolution properties of phenylbutazone polymorphs. J. Pharm. Pharmacol., 1983, 35, 208-214
[4]: Brittain HG, and Grant DJW. Effects of polymorphism and solid-state solvation on solubility and dissolution rate. In Polymorphism in pharmaceutical sciences, drugs and the pharmaceutical sciences; Brittain HG, Ed. New York City, New York: Marcel Dekker, 1999. 279–330.
[5]: Yu L et.al. Thermochemistry and conformational polymorphism of a hexamorphic crystal system. J. Am. Chem. Soc., 2000, 122, 585-591
[6]: Yu L. Polymorphism in molecular solids: an extraordinary system of red, orange, and yellow crystals. Acc. Chem. Res., 2010, 43 (9), 125-1266

 

2018-03-12T13:16:55+00:00
Menu