- Email: [email protected]
Characterization of Pharmaceutical Solvates by Combined Thermogravimetric and Infrared Analysis
+
+
NOTES
Characterization of Pharmaceutical Solvates by Combined Thermogravimetric and Infrared Analysis CHRISTIANE RODRIGUEZ
AND
DAVID E. BUGAYX
Received July 12, 1996, from the Bristol-Myers Squibb Pharmaceutical Research Institute, One Squibb Drive, P.O. Box 191, New Brunswick, NJ 08903. Final revised manuscript received November 4, 1996. Accepted for publication November 4, 1996X. Abstract 0 The combined physical analytical technique of thermogravimetric and infrared analysis (TG/IR) is described for the investigation of pharmaceutical solids. TG/IR provides an unequivocal identification of the volatile content of a pharmaceutical solid. In the case of pharmaceutical solvates, TG/IR provides identification of the evolved gas from the crystalline solid. Variable temperature X-ray diffraction and differential scanning calorimetry techniques provide the required information to ascertain whether the evolved gas was due to a solvent incorporated into the crystal lattice or physically adsorbed onto the solid. This work illustrates three examples of TG/IR as utilized in a multidisciplinary approach to the physical characterization of pharmaceutical solids.
the sample, but often other solvents may be present that are left unidentified. TG/IR complements the information obtained by the aforementioned techniques, but more importantly provides an unequivocal identification of any volatile content within a sample. The combined TG/IR technique obtains spectroscopic data of the gases evolved as the sample is heated in a TG furnace. The resultant spectra obtained as a function of temperature can then be compared with available spectral libraries for potential species identification. Some advantages of TG/IR analysis are the small sampling amounts required (as little as 1 mg), as well as the relatively short sample preparation and subsequent analysis times. Examples are presented in this work that clearly illustrate that TG/IR aids in the multidisciplinary characterization of pharmaceutical solvates.
Introduction Materials of pharmaceutical interest (bulk drug substance and excipients) often exist in various crystal forms, including solvates. The type of solvent present is often of extreme importance with regard to the physical and chemical stability of the drug substance in addition to formulation issues. Thus, characterization of the solvate is very important. Various techniques are currently utilized to characterize solvates including powder X-ray diffraction (XRD), microscopy, thermogravimetric analysis (TGA), gas chromatography (GC), and differential scanning calorimetry (DSC).1,2 However, none of these techniques unequivocally identifies the chemical solvate (e.g., methanoate versus ethanoate). Although other combined techniques, such as GC/infrared (IR) spectroscopy (GC/ IR)3 and GC/mass spectrometry (GC/MS),4 utilize spectroscopic techniques for evolved gas analysis from solids, these techniques have not been applied to crystal form analysis. TG/IR analysis is a relatively new technique to the pharmaceutical industry5 that can provide extremely useful information for the unequivocal identification of the solvent incorporated into a crystal lattice. Although microscopy and X-ray diffraction provide information regarding the morphology and crystalline nature of the sample, they do not provide the information needed to identify the type of crystalline solvate present. DSC provides information on the energetics of how the solvent interacts with the parent molecule, and TGA provides insight about the amount of volatiles actually present.2 However, these techniques are nonselective. Although the type of solvent incorporated into the crystal lattice can be inferred from the DSC and TGA data because the solvent recrystallization system is usually known, these techniques do not unequivocally identify the volatile. Karl Fisher analysis typically can indicate the amount of water in X
Abstract published in Advance ACS Abstracts, December 15, 1996.
© 1997, American Chemical Society and American Pharmaceutical Association
Experimental Section MaterialssThree separate pharmaceutical developmental compounds (A, B, and C) were obtained from the Chemical Process Development laboratories of the Bristol-Myers Squibb Pharmaceutical Research Institute. Purity values for each compound were >95% as determined by high performance liquid chromatography (HPLC). EquipmentsA Bomem TG Plus system was used that interfaces a Dupont 951 TG analyzer with a Bomem Michelson (MB-100) Fourier transform (FT)-IR spectrophotometer. The sample (∼10 mg) was placed in the TG analyzer and, as the sample was heated, ultra-high purity helium (99.9%) was used as a carrier gas (880 cc/min) to introduce the evolved gases from the sample to the IR gas cell. A 16-pass, 27.2-cm long gas cell was used for IR spectral acquisition. The gas cell was maintained at 160 °C to avoid the condensation of gases on the reflectance mirrors. In addition, helium was introduced into the ends of the IR gas cell to help carry the evolved gases through the gas cell and to keep it well purged. Nitrogen (2 L/min) was used to purge the inside of the spectrophotometer and eliminate atmospheric gases (CO2, H2O). Each IR spectrum represents 16 co-added scans acquired at a spectral resolution of 4 cm-1. The IR was equipped with a globar source, a Ge/KBr beamsplitter, and a DTGS detector. Bomem TG Plus integrated software was used to control both the IR and the TG analyzer concurrently. All experiments were run at a TG ramp rate of 10 °C/min from ambient to 300 °C. An IR spectrum was acquired every 1 min. Galactic Industries Corporation GRAMS/386 software was utilized for data analysis. The multifile option was used to create a three-dimensional representation of the IR spectra, which was then incorporated into a view.
Results and Discussion To physically characterize a compound, a multidisciplinary approach is required incorporating microscopy, XRD, thermal analysis, GA, and molecular spectroscopy. In an effort to characterize a new developmental compound, a dopamine reuptake inhibitor currently under development for the treat-
S0022-3549(96)00286-9 CCC: $14.00
Journal of Pharmaceutical Sciences / 263 Vol. 86, No. 2, February 1997
+
+
Figure 1s(a) TG/IR data set for the HBr salt of developmental compound A. (b) The TG curve displays a 2.3% weight loss between 40 and 100 °C that is attributed to water by the presence of IR absorption bands centered at 1600 cm-1. The bracketed area of the spectral dataset highlights the low intensity absorption bands attributed to water vapor.
Figure 2s(a) TG/IR data set for developmental compound B. (b) Display of a 12.2% weight loss. Water and butyl acetate were identified as evolved components.
ment of Parkinson’s disease, TG/IR analysis was used. Various salts of compound A were generated in an effort to acquire the best final form for stability, formulation, and efficacy. During this process, it was necessary to determine if residual solvent was present or if a solvated form existed. The HBr salt exhibited a distinct XRD pattern, indicating a crystalline solid. Upon variable temperature XRD studies, the diffraction pattern did not change as the temperature was increased, eliminating the possibility of a different crystal form, but TGA displayed a weight loss from ∼40 to 100 °C. The results of TG/IR analysis of the sample are shown in Figure 1. The TG curve displays a weight loss of 2.3% (w/w) between 40 and 100 °C. Analysis of the IR spectra during this time/temperature period shows that the weight loss detected in the TG curve can be attributed to water because of the absorption bands centered at ∼1600 cm-1. The weight loss occurring 264 / Journal of Pharmaceutical Sciences Vol. 86, No. 2, February 1997
Figure 3sSpectrum a represents the averaged IR spectra acquired from 60 to 100 °C for developmental compound B. Comparison with the reference IR spectrum of butyl acetate is shown in spectrum b.
+
+
Figure 4sTG/IR data set for (a) developmental compound C and (b) its inherent TG weight loss curve.
after 200 °C is due to decomposition of the compound. Thus, it may be concluded from the XRD studies that the HBr salt of developmental compound A is a single crystal form, and the TG/IR data show that the material retains some residual water. In a second example of the physical characterization of a pharmaceutical active substance, TG/IR was used to study developmental compound B, a substance indicated for thetreatment of hypercholesterolemia. Based on a series of studies, it was determined that the methanesulfonic acid salt was to be used for development. This material was recrystallized from butyl acetate. The XRD pattern showed the material to be crystalline at room temperature but indicated that the material was amorphous after the sample was heated to 110 °C. Combined XRD and thermal analysis results confirmed that a physical transformation occurred upon heating. TG/IR analysis was performed and the results are displayed in Figure 2. The weight loss curve shows a 12.2% (w/w) loss that concurred with the weight loss obtained from TGA alone. Analysis of the evolved gases given off between 20 and 100 °C by IR spectroscopy makes it possible to identify the volatile components. Water can be seen to evolve at 2060 °C by the presence of the peaks centered at 1600 cm-1, but another volatile is detected that begins to evolve at 60 °C. Observation of the IR spectra collected from 60 to 100 °C indicates that varying intensities of the same IR spectrum is obtained for each temperature. The intensity variation is related to the evolved species concentration by Beer’s law. To facilitate the identification of the unknown volatile evolving during this time/temperature period, all of the spectra collected from 60 to 100 °C were coadded together and the resultant spectrum was compared with those found in a search of a vapor-phase IR spectral library. After this process was performed, it was clear that the identity of the volatile was butyl acetate (Figure 3). The two volatiles were concurrently evolving from the sample during the same temperature range (Figure 2), so quantitative analysis of the water content could not be determined from the TGA; therefore, Karl Fisher (KF) analysis was used to determine the water content. KF analysis indicated that the sample contained 1.5% water and it was concluded that the sample contained 10.7% butyl acetate. Based on the molecular weight of the parent compound and the knowledge that a butyl acetate solvate existed,
Figure 5sSpectrum a represents the averaged IR spectrum acquired from 60 to 100 °C for developmental compound C. Comparisons with the reference IR spectra of ethanol and hexane are shown in spectra b and c, respectively.
it was determined that the compound existed as a 1:1 butyl acetate solvate. The final example of solvate characterization by TG/IR analysis involves developmental compound C. This compound is intended for use in the treatment of congestive heart failure. XRD analysis determined that a number of different crystal forms existed, and it was thus necessary to fully characterize each one. One crystal form was crystallized from a cosolvent mixture of ethanol and hexane. The XRD pattern indicated that the material was crystalline, and a melting point of ∼70 °C was determined by DSC and visual observation. Upon heating of the sample (not above the melting point of the material, 70 °C), the XRD pattern changed but remained crystalline. Through analysis of the TG/IR data (Figure 4) it was determined that there was a 9.6% weight loss associated with the sample. Within the IR data set collected, evolved gases were detected from 60 to 100 °C. Spectral averaging during this temperature range provided a clear spectrum for interpretation (Figure 5). Ethanol is clearly detected, with minor contributions from hexane. After analyzing a series of samples that were generated from different ratios of the cosolvent recrystallization, it was clear that this crystal form was an ethanol solvate that also exhibited a small amount of hexane residual solvent. ConclusionssThrough the various examples shown in this paper it is clear that TG/IR analysis is a worthwhile, simple technique beneficial for the physical characterization of
Journal of Pharmaceutical Sciences / 265 Vol. 86, No. 2, February 1997
+
+
pharmaceutical solids. Though not widely used in pharmaceutical laboratories, TG/IR analysis merits consideration whenever a solvate is suspected. TG/IR analysis has the potential to provide the amount of volatile present in the sample from analysis of the TG data and also affords an unequivocal identification of the evolved gases. The technique may also be extended to the study of formulated products (capsules and tablets).
References and Notes 1. Giron, D. Thermochim. Acta 1995, 248, 1-59.
266 / Journal of Pharmaceutical Sciences Vol. 86, No. 2, February 1997
2. Khankari, R. K.; Grant, D. J. W. Thermochim. Acta 1995, 248, 61-79. 3. Gilbert, A. S.; Moss, C. J.; Francis, P. L.; Ashton, M. J.; Ashton, D. S. Chromatographia 1996, 42, 305-308. 4. Jacobsson, S.; Hagman, A. Drug Dev. Ind. Pharm. 1990, 16, 2547-2560. 5. Johnson, D. J.; Compton, A. C. Spectroscopy 1988, 3(6), 47-50.
Acknowledgments The authors acknowledge Imre Vitez and Chris Kiesnowski for TG/ DSC and XRD data acquisition, respectively, and Ann Newman for subsequent discussions.
JS960286Z
+
NOTES
Characterization of Pharmaceutical Solvates by Combined Thermogravimetric and Infrared Analysis CHRISTIANE RODRIGUEZ
AND
DAVID E. BUGAYX
Received July 12, 1996, from the Bristol-Myers Squibb Pharmaceutical Research Institute, One Squibb Drive, P.O. Box 191, New Brunswick, NJ 08903. Final revised manuscript received November 4, 1996. Accepted for publication November 4, 1996X. Abstract 0 The combined physical analytical technique of thermogravimetric and infrared analysis (TG/IR) is described for the investigation of pharmaceutical solids. TG/IR provides an unequivocal identification of the volatile content of a pharmaceutical solid. In the case of pharmaceutical solvates, TG/IR provides identification of the evolved gas from the crystalline solid. Variable temperature X-ray diffraction and differential scanning calorimetry techniques provide the required information to ascertain whether the evolved gas was due to a solvent incorporated into the crystal lattice or physically adsorbed onto the solid. This work illustrates three examples of TG/IR as utilized in a multidisciplinary approach to the physical characterization of pharmaceutical solids.
the sample, but often other solvents may be present that are left unidentified. TG/IR complements the information obtained by the aforementioned techniques, but more importantly provides an unequivocal identification of any volatile content within a sample. The combined TG/IR technique obtains spectroscopic data of the gases evolved as the sample is heated in a TG furnace. The resultant spectra obtained as a function of temperature can then be compared with available spectral libraries for potential species identification. Some advantages of TG/IR analysis are the small sampling amounts required (as little as 1 mg), as well as the relatively short sample preparation and subsequent analysis times. Examples are presented in this work that clearly illustrate that TG/IR aids in the multidisciplinary characterization of pharmaceutical solvates.
Introduction Materials of pharmaceutical interest (bulk drug substance and excipients) often exist in various crystal forms, including solvates. The type of solvent present is often of extreme importance with regard to the physical and chemical stability of the drug substance in addition to formulation issues. Thus, characterization of the solvate is very important. Various techniques are currently utilized to characterize solvates including powder X-ray diffraction (XRD), microscopy, thermogravimetric analysis (TGA), gas chromatography (GC), and differential scanning calorimetry (DSC).1,2 However, none of these techniques unequivocally identifies the chemical solvate (e.g., methanoate versus ethanoate). Although other combined techniques, such as GC/infrared (IR) spectroscopy (GC/ IR)3 and GC/mass spectrometry (GC/MS),4 utilize spectroscopic techniques for evolved gas analysis from solids, these techniques have not been applied to crystal form analysis. TG/IR analysis is a relatively new technique to the pharmaceutical industry5 that can provide extremely useful information for the unequivocal identification of the solvent incorporated into a crystal lattice. Although microscopy and X-ray diffraction provide information regarding the morphology and crystalline nature of the sample, they do not provide the information needed to identify the type of crystalline solvate present. DSC provides information on the energetics of how the solvent interacts with the parent molecule, and TGA provides insight about the amount of volatiles actually present.2 However, these techniques are nonselective. Although the type of solvent incorporated into the crystal lattice can be inferred from the DSC and TGA data because the solvent recrystallization system is usually known, these techniques do not unequivocally identify the volatile. Karl Fisher analysis typically can indicate the amount of water in X
Abstract published in Advance ACS Abstracts, December 15, 1996.
© 1997, American Chemical Society and American Pharmaceutical Association
Experimental Section MaterialssThree separate pharmaceutical developmental compounds (A, B, and C) were obtained from the Chemical Process Development laboratories of the Bristol-Myers Squibb Pharmaceutical Research Institute. Purity values for each compound were >95% as determined by high performance liquid chromatography (HPLC). EquipmentsA Bomem TG Plus system was used that interfaces a Dupont 951 TG analyzer with a Bomem Michelson (MB-100) Fourier transform (FT)-IR spectrophotometer. The sample (∼10 mg) was placed in the TG analyzer and, as the sample was heated, ultra-high purity helium (99.9%) was used as a carrier gas (880 cc/min) to introduce the evolved gases from the sample to the IR gas cell. A 16-pass, 27.2-cm long gas cell was used for IR spectral acquisition. The gas cell was maintained at 160 °C to avoid the condensation of gases on the reflectance mirrors. In addition, helium was introduced into the ends of the IR gas cell to help carry the evolved gases through the gas cell and to keep it well purged. Nitrogen (2 L/min) was used to purge the inside of the spectrophotometer and eliminate atmospheric gases (CO2, H2O). Each IR spectrum represents 16 co-added scans acquired at a spectral resolution of 4 cm-1. The IR was equipped with a globar source, a Ge/KBr beamsplitter, and a DTGS detector. Bomem TG Plus integrated software was used to control both the IR and the TG analyzer concurrently. All experiments were run at a TG ramp rate of 10 °C/min from ambient to 300 °C. An IR spectrum was acquired every 1 min. Galactic Industries Corporation GRAMS/386 software was utilized for data analysis. The multifile option was used to create a three-dimensional representation of the IR spectra, which was then incorporated into a view.
Results and Discussion To physically characterize a compound, a multidisciplinary approach is required incorporating microscopy, XRD, thermal analysis, GA, and molecular spectroscopy. In an effort to characterize a new developmental compound, a dopamine reuptake inhibitor currently under development for the treat-
S0022-3549(96)00286-9 CCC: $14.00
Journal of Pharmaceutical Sciences / 263 Vol. 86, No. 2, February 1997
+
+
Figure 1s(a) TG/IR data set for the HBr salt of developmental compound A. (b) The TG curve displays a 2.3% weight loss between 40 and 100 °C that is attributed to water by the presence of IR absorption bands centered at 1600 cm-1. The bracketed area of the spectral dataset highlights the low intensity absorption bands attributed to water vapor.
Figure 2s(a) TG/IR data set for developmental compound B. (b) Display of a 12.2% weight loss. Water and butyl acetate were identified as evolved components.
ment of Parkinson’s disease, TG/IR analysis was used. Various salts of compound A were generated in an effort to acquire the best final form for stability, formulation, and efficacy. During this process, it was necessary to determine if residual solvent was present or if a solvated form existed. The HBr salt exhibited a distinct XRD pattern, indicating a crystalline solid. Upon variable temperature XRD studies, the diffraction pattern did not change as the temperature was increased, eliminating the possibility of a different crystal form, but TGA displayed a weight loss from ∼40 to 100 °C. The results of TG/IR analysis of the sample are shown in Figure 1. The TG curve displays a weight loss of 2.3% (w/w) between 40 and 100 °C. Analysis of the IR spectra during this time/temperature period shows that the weight loss detected in the TG curve can be attributed to water because of the absorption bands centered at ∼1600 cm-1. The weight loss occurring 264 / Journal of Pharmaceutical Sciences Vol. 86, No. 2, February 1997
Figure 3sSpectrum a represents the averaged IR spectra acquired from 60 to 100 °C for developmental compound B. Comparison with the reference IR spectrum of butyl acetate is shown in spectrum b.
+
+
Figure 4sTG/IR data set for (a) developmental compound C and (b) its inherent TG weight loss curve.
after 200 °C is due to decomposition of the compound. Thus, it may be concluded from the XRD studies that the HBr salt of developmental compound A is a single crystal form, and the TG/IR data show that the material retains some residual water. In a second example of the physical characterization of a pharmaceutical active substance, TG/IR was used to study developmental compound B, a substance indicated for thetreatment of hypercholesterolemia. Based on a series of studies, it was determined that the methanesulfonic acid salt was to be used for development. This material was recrystallized from butyl acetate. The XRD pattern showed the material to be crystalline at room temperature but indicated that the material was amorphous after the sample was heated to 110 °C. Combined XRD and thermal analysis results confirmed that a physical transformation occurred upon heating. TG/IR analysis was performed and the results are displayed in Figure 2. The weight loss curve shows a 12.2% (w/w) loss that concurred with the weight loss obtained from TGA alone. Analysis of the evolved gases given off between 20 and 100 °C by IR spectroscopy makes it possible to identify the volatile components. Water can be seen to evolve at 2060 °C by the presence of the peaks centered at 1600 cm-1, but another volatile is detected that begins to evolve at 60 °C. Observation of the IR spectra collected from 60 to 100 °C indicates that varying intensities of the same IR spectrum is obtained for each temperature. The intensity variation is related to the evolved species concentration by Beer’s law. To facilitate the identification of the unknown volatile evolving during this time/temperature period, all of the spectra collected from 60 to 100 °C were coadded together and the resultant spectrum was compared with those found in a search of a vapor-phase IR spectral library. After this process was performed, it was clear that the identity of the volatile was butyl acetate (Figure 3). The two volatiles were concurrently evolving from the sample during the same temperature range (Figure 2), so quantitative analysis of the water content could not be determined from the TGA; therefore, Karl Fisher (KF) analysis was used to determine the water content. KF analysis indicated that the sample contained 1.5% water and it was concluded that the sample contained 10.7% butyl acetate. Based on the molecular weight of the parent compound and the knowledge that a butyl acetate solvate existed,
Figure 5sSpectrum a represents the averaged IR spectrum acquired from 60 to 100 °C for developmental compound C. Comparisons with the reference IR spectra of ethanol and hexane are shown in spectra b and c, respectively.
it was determined that the compound existed as a 1:1 butyl acetate solvate. The final example of solvate characterization by TG/IR analysis involves developmental compound C. This compound is intended for use in the treatment of congestive heart failure. XRD analysis determined that a number of different crystal forms existed, and it was thus necessary to fully characterize each one. One crystal form was crystallized from a cosolvent mixture of ethanol and hexane. The XRD pattern indicated that the material was crystalline, and a melting point of ∼70 °C was determined by DSC and visual observation. Upon heating of the sample (not above the melting point of the material, 70 °C), the XRD pattern changed but remained crystalline. Through analysis of the TG/IR data (Figure 4) it was determined that there was a 9.6% weight loss associated with the sample. Within the IR data set collected, evolved gases were detected from 60 to 100 °C. Spectral averaging during this temperature range provided a clear spectrum for interpretation (Figure 5). Ethanol is clearly detected, with minor contributions from hexane. After analyzing a series of samples that were generated from different ratios of the cosolvent recrystallization, it was clear that this crystal form was an ethanol solvate that also exhibited a small amount of hexane residual solvent. ConclusionssThrough the various examples shown in this paper it is clear that TG/IR analysis is a worthwhile, simple technique beneficial for the physical characterization of
Journal of Pharmaceutical Sciences / 265 Vol. 86, No. 2, February 1997
+
+
pharmaceutical solids. Though not widely used in pharmaceutical laboratories, TG/IR analysis merits consideration whenever a solvate is suspected. TG/IR analysis has the potential to provide the amount of volatile present in the sample from analysis of the TG data and also affords an unequivocal identification of the evolved gases. The technique may also be extended to the study of formulated products (capsules and tablets).
References and Notes 1. Giron, D. Thermochim. Acta 1995, 248, 1-59.
266 / Journal of Pharmaceutical Sciences Vol. 86, No. 2, February 1997
2. Khankari, R. K.; Grant, D. J. W. Thermochim. Acta 1995, 248, 61-79. 3. Gilbert, A. S.; Moss, C. J.; Francis, P. L.; Ashton, M. J.; Ashton, D. S. Chromatographia 1996, 42, 305-308. 4. Jacobsson, S.; Hagman, A. Drug Dev. Ind. Pharm. 1990, 16, 2547-2560. 5. Johnson, D. J.; Compton, A. C. Spectroscopy 1988, 3(6), 47-50.
Acknowledgments The authors acknowledge Imre Vitez and Chris Kiesnowski for TG/ DSC and XRD data acquisition, respectively, and Ann Newman for subsequent discussions.
JS960286Z