- Email: [email protected]
Microspectroscopic FT-IR mapping system as a tool to assess blend homogeneity of drug–excipient mixtures
European Journal of Pharmaceutical Sciences 23 (2004) 117–122
Microspectroscopic FT-IR mapping system as a tool to assess blend homogeneity of drug–excipient mixtures Ting-Huei Lee, Shan-Yang Lin∗ Biopharmaceutics Laboratory, Department of Medical Research and Education, Veterans General Hospital-Taipei, Taipei, Taiwan Received 11 July 2003; received in revised form 1 April 2004; accepted 1 April 2004 Available online 12 August 2004
Abstract In order to prepare a controlled-release tablet by direct compression, a Fourier transform infrared (FT-IR) microspectroscopic mapping system was utilized to assess the blend homogeneity of a pharmaceutical powder blend. The model powder blend was a two-component mixture consisting of captopril as a model drug and micronized ethylcellulose (EC) as a direct-compressible model excipient, which was mixed in a laboratory mill. The two-component mixture was mixed in two different weight proportions (captopril:EC, 1:1 or 10:1). By varying the mixing time, different blend mixtures sampled were determined by a reflectance FT-IR microspectroscopic mapping system to collect successively the IR spectra from the actual analysis area. The results revealed that the blend homogeneity increased gradually with increased mixing times, but the powder began to demix or segregate as mixing continued beyond the time when homogeneity was reached. In addition, the statistical results indicated that the sample mixture proportion had an effect on the uniformity of the powder blend. This study demonstrates that microspectroscopic FT-IR mapping technique can be easily used to determine the blend homogeneity in a powder blend mixture. © 2004 Elsevier B.V. All rights reserved. Keywords: FT-IR; Mapping; Blend homogeneity; Micronized ethylcellulose; Captopril; Mixing ratio
1. Introduction In the preparation of pharmaceutical solid dosage forms, the blending process is one of the most common unit operations to uniformly mix the active drug ingredient with one or more excipients to achieve the in-process and product control. The homogeneous blending ensures a uniform distribution and reproducibility of all components in the end product (Berntsson et al., 2002; Lai et al., 2001; Sekulic et al., 1998; Chang et al., 1996). Because blend uniformity analysis has also been requested by good manufacturing practice (GMP) regulation in US Food and Drug Administration’s draft guidance, it should consistently test the blending mixture to ensure adequate blend homogeneity and stability in the blending process. The testing by high-performance liquid chromatography (HPLC) and ultraviolet (UV) spectroscopy has been typically used to evaluate the blend or content uniformity of blend samples or dosage unit samples in pharmaceutical ∗
Corresponding author. Fax: +886-2-2873-7200. E-mail address: [email protected] (S.-Y. Lin).
0928-0987/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ejps.2004.04.015
manufacturing. Recently, vibrational spectroscopic techniques, including near-infrared, mid-infrared, and Raman spectroscopies, have been employed as alternatives to the common destructive methods in order to save time and reduce the requirement for large amounts of toxic and expensive solvents (Blanco et al., 2002; Gustafsson et al., 2003; Johansson et al., 2002). These vibrational spectroscopic methods provide non-destructively physical and chemical characterization of the active and inactive components in the composite mixtures, but the single point of spectroscopic determination cannot directly reflect the spatial distribution of the components in the mixture. Thus, a mapping technique is included with the vibrational spectroscopy to map the location of each spatially resolved component (Smith, 2000). This combination of a mapping system and vibrational spectroscopy for the molecular analysis of components is achieved by scanning techniques. The Fourier transform infrared (FT-IR) microspectroscopic mapping system is a powerful technique that combines the image analysis of light microscopy, chemical analysis of FT-IR spectroscopy, and a mapping stage. A mapping stage can translate the sample along the X- and
118
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
Y-axes thereby moving different areas of the sample into the microscope’s beam path. The IR spectra can then be directly related back to the sample’s microstructure, and can also reflect the existence and content of the sample. This combined system not only reduces the blend times by allowing accurate determination of end-points, but also monitors the homogeneity of all components, since the homogenization process affects not only the active ingredient but also the excipients. The aim of this study was to evaluate the FT-IR microspectroscopic mapping system as a tool to assess the blend uniformity and potency of the blend mixture of captopril and micronized ethylcellulose (EC) used for preparing the directly compressed controlled-release tablet (Lin et al., 2001a,b, 2002). In our preliminary study, we found that two unique IR spectra peaks at 1747cm-1 for captopril and at 3478cm-1 for EC could be used as markers to study the blend uniformity of captopril and EC mixture. Blend homogeneity and optimal mixing times have been examined using a statistical analysis.
2. Materials and methods 2.1. Material Ethylcellulose powder with a particle size of 4.0mm (grade N-7-F; average molecular weight, 58,000) was obtained from Shin-Etsu Chem. Ind. Co. Ltd. (Tokyo, Japan). Captopril (<149mm) was of pharmaceutical grade and was purchased from Bulk Med. Pharm., Hamburg, Germany. 2.2. Sample preparation All mixing studies were carried out in an analytical mill (Model A-70, Morat, Germany). A two-component mixture consisting of captopril and EC (weight ratio (w/w), 1:1 or 1:10) was powder blended. The blending process was monitored. At predetermined timepoints, an aliquot of the blended powder was carefully sampled and compressed using an IR spectrophotometric hydraulic press (Riken Seiki Co., Tokyo, Japan) under constant pressure at 400kg/cm2 for 1min.
2.3. Microspectroscopic FT-IR mapping study The powder sample compressed on the die was set on the stage of FT-IR microscopic spectrometer (Micro IRT-30, Jasco, Tokyo, Japan) equipped with a MCT detector. The system was operated in a reflectance mode over the range 4600–1000cm-1 . An appropriate sample area was selected, and the reflectance IR spectra were collected successively from the actual analysis area by a mapping process. An automated X–Y stage for mapping was employed in order to obtain full IR spectra from the sample at 18mm intervals across the 35 spots ’ 35 spots. The use of an 18mm step size results in the acquisition of 1225 spectra. The sampling aperture was set at 40mm ’ 40mm, each spectrum was performed at 10 scans with resolution of 4cm-1 . Fig. 1 depicts the concept diagram of data collection method and processing required in the mapping experiments. 2.4. Statistical analysis of blend sample variability The blend uniformity of the captopril–EC mixture was estimated from the mapping spectra by evaluating the consistency of the IR peak intensity ratio of the above unique IR peaks in each reflectance spectra of EC or captopril, as represented by the IR spectral map. Spectral manager software for Windows (Jasco Co., Japan) was used. The sampling process was done in triplicates. The mixture was estimated to be homogeneous when their IR peak intensity ratios of these unique IR peaks was consistent as exhibited by histogram or modified box plot. The mixture was termed homogeneous when the standard deviation of the IR spectrum reached a minimum value. Analyses were performed using SPSS 11.0 for windows, Statistica 6, and SigmaPlot 2000.
3. Results and discussion The representative FT-IR spectra of EC, captopril, and the EC–captopril mixture are shown in Fig. 2. As expected, a number of differences can be observed between the spectra of EC and captopril. Captopril had two unique IR peaks, at 1747 and 1590cm-1 , which can be assigned to the C==O
Fig. 1. Concept diagram of data collection method and processing required in mapping experiments.
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
Fig. 2. FT-IR spectra of captopril (a), EC (b), and the blended captopril–EC (1:1) mixture (c).
stretching vibration of carboxylic acid and amide band, respectively. However, EC did not have these features in this range but exhibited a unique peak at 3478cm-1 due to hydroxyl groups. This clearly suggests that it is possible to distinguish the spectral difference from the mixture by determining the peak intensity ratio of the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ). Fig. 3 presents the IR spectral maps and frequency distribution plots of the blended samples (captopril:EC, 1:1 (w/w)) in the process of mixing. The corresponding maps and plots represent a 3D and 2D distribution of the peak height ratio of the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ). The width of the distribution shown in each frequency distribution plot is an indicator of the heterogeneity of the sample. Results of this analysis clearly indicate that the sample was poorly blended after up to 0.5min of mixing. The variety of ratios presented in the map indicates that the two pharmaceutical powders were not uniformly distributed within the blended sample. A wide distribution was exhibited in the corresponding distribution plot as well. The IR spectral map for the blended sample after mixing for 1min exhibited an improved uniformity compared with that of the 0.1 or 0.5min blended mixture. The frequency distribution plot had a narrower range. With increasing time of mixing (up to 3min), the blended mixture had a more uniform distribution of components in which the distribution of the peak height ratios was grouped together. This reveals that the two pharmaceutical powders are at a nearly constant ratio within the blended sample. From these results, it can be seen that blend homogeneity is achieved rapidly. On the other hand, the high-peak height ratio and low-peak height ratio contributing to the slight extension in tailing of the distribution plot was observed at 5min. The blended samples from 0.1 to 3min showed a progressive decrease in heterogeneity, while the blended samples from 4 to 7min demonstrated a progressive increase in heterogeneity. The segregation occurring in the mixture might be responsible for this observation.
119
The changes in IR spectral maps and frequency distribution plots of blended samples with 1:10 (w/w) ratio of captopril to EC after mixing were also investigated. The optimum blending time for mixing the 1:10 weight ratio of captopril to EC was about 1–6min, as compared with the blending time for the mixing the 1:1 weight ratio of captopril to EC. This also indicates that the proportion of sample mixture also played an important role to affect the uniformity of the powder blend. Fig. 4A illustrates the relationship between the peak height ratio for the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ) versus mixing time. Clearly, the values of peak height ratios were gradually reduced and concentrated. The effect of blending time on homogeneity is shown in Fig. 4B, which depicts the S.D. versus blending time. In common, the homogeneous mixture had aminimum value of S.D. As seen in Fig. 4B, the S.D. decreases as a function of time, indicating that the blends are approaching homogeneity. It was gradually reduced with the blending progress, and reached to a minimum but after that begun to enlarge again. A progressive decrease in the S.D. was observed as the degree of blending developed. Apparently, the blend homogeneity was achieved after approximately 3min in the blending process. This might be attributed to the particle size and size distribution changes arising from attrition in the blending, which is consistent with results from other studies (Wargo and Drennen, 1996; Blanco et al., 2002; El-Hagrasy et al., 2001). Fig. 5A depicts the relationship between the peak height ratio for the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ) versus mixing time. Apparently, the values of peak height ratios were gradually reduced. A progressive decrease in the S.D. was observed as the degree of blending developed. The effect of blend time on homogeneity is evident as shown in Fig. 5B, which depicts the sample standard deviations versus blending time. A blend homogeneity was reached after 6min. But the powder begun to demix or segregate as mixing continued beyond the time when homogeneity was reached, exhibiting increased variability after 6min. This same phenomenon was observed in the 1:1 mixture as discussed above. This result indicates that the proportion of sample mixture also played an important role to affect the uniformity of the powder blend, which is consistent with other reports (Kornchankul et al., 2002). For this study, the optimum blending time for mixing captopril to EC was about 3min for 1:1 weight ratio while it was almost 6min for 1:10 weight ratio. In conclusion, the homogeneity of the blend samples can be determined using mid-infrared microspectroscopic mapping, which indicates that this technique has great potential as an analytical tool for blend uniformity analysis. Qualitative IR analysis can be employed to assess the uniformity of a single production blend or to define the optimal mixing time during the development process. These experiments have demonstrated that both IR spectral maps and frequency distribution plots for equality of
120
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
Fig. 3. Three-dimensional FT-IR spectral maps and frequency distribution plots of the blended captopril–EC (1:1) mixture in the process of blending.
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
121
Fig. 4. Relationship between the spectral peak height ratio and mixing time for the captopril–EC mixture (1:1) (A) and the effect of blending time on homogeneity (B).
Fig. 5. The relationship between the spectral peak height ratio and mixing time for the captopril–EC mixture (1:10) (A) and the effect of blending time on homogeneity (B).
variance are effective methods for qualitative evaluation of powder blend homogeneity. Compared with traditional IR spectroscopy (single point), IR mapping provides the opportunity to investigate a large area within a single sample. A minimal abnormality in a bulk sample may be undetected by using traditional IR spectroscopy but it may be detected using IR mapping. This study indicates the FT-IR microspectroscopic mapping technique for direct analysis of powder blend mixture can be easily used to determine the homogeneity of a pharmaceutical blend.
References Berntsson, O., Danielsson, L.G., Lagerholm, B., Folestad, S., 2002. Quantitative in-line montoring of powder blending by near infrared reflection spectroscopy. Powder Technol. 123, 185–193. Blanco, M., Gozález Bañó, R., Bertran, E., 2002. Monitoring powder blending in pharmaceutical processes by use of near infrared spectroscopy. Talanta 56, 203–212. Chang, R., Shukla, J., Buehler, J., 1996. An evaluation of a unit-dose compacting sample thief and a discussion of content uniformity testing and blending validation issues. Drug Dev. Ind. Pharm. 22, 1031–1035.
El-Hagrasy, A.S., Morris, H.R., D’Amico, F., Lodder, R.A., Drennen III, J.K., 2001. Near-infrared spectroscopy and imaging for the monitoring of powder blend homogeneity. J. Pharm. Sci. 90, 1298–1307. Gustafsson, C., Nyström, C., Lennholm, H., Bonferoni, M.C., Caramella, C.M., 2003. Characteristics of hydroxypropyl methylcellulose infulencing compactibility and prediction of particle and tablet properties by infrared spectroscopy. J. Pharm. Sci. 92, 494–504. Johansson, J., Pettersson, S., Taylor, L.S., 2002. Infrared imaging of laser-induced heating during Raman spectroscopy of pharmaceutical solids. J. Pharm. Biomed. Anal. 30, 1223–1231. Kornchankul, W., Hamed, E., Parikh, N.H., Sakr, A., 2002. Effect of drug proportion and mixing time on the content uniformity of a low dose drug in a high shear mixer. Pharmazie 57, 49–53. Lai, C.K., Holt, D., Leung, J.C., Conney, C.L., Raju, G.K., Hansen, P., 2001. Real time and noninvasive monitoring of dry powder blend homogeneity. AIChE J. 47, 2618–2622. Lin, S.Y., Lin, K.H., Li, M.J., 2001a. Micronized ethylcellulose used for designing a directly compressed time-controlled disintegration tablet. J. Controlled Release 70, 321–328. Lin, K.H., Lin, S.Y., Li, M.J., 2001b. Compression forces and amount of outer coating layer affecting the time-controlled disintegration of the compression-coated tablets prepared by direct compression with micronized ethylcellulose. J. Pharm. Sci. 90, 2005– 2009. Lin, S.Y., Lin, K.H., Li, M.J., 2002. Infulence of excipients, drugs, and osmotic agent in the inner core on the time-controlled disintegration
122
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
of compression-coated ethylcellulose tablets. J. Pharm. Sci. 91, 2040– 2046. Sekulic, S.S., Wakeman, J., Doherty, P., Hailey, P.A., 1998. Automated system for the on-line montoring of powder blending processes using near-infrared spectroscopy. II. Qualitative approaches to blend evaluation. J. Pharm. Biomed. Anal. 17, 1285–1309.
Smith, P.A.M., 2000. Infrared microspectroscopy mapping studies of packaging materials: experiment design and data profiling considerations. Vib. Spectrosc. 24, 47–62. Wargo, D., Drennen, J.K., 1996. Near-infrared spectroscopic characterization of pharmaceutical powder blends. J. Pharm. Biomed. Anal. 14, 1415–1423.
Microspectroscopic FT-IR mapping system as a tool to assess blend homogeneity of drug–excipient mixtures Ting-Huei Lee, Shan-Yang Lin∗ Biopharmaceutics Laboratory, Department of Medical Research and Education, Veterans General Hospital-Taipei, Taipei, Taiwan Received 11 July 2003; received in revised form 1 April 2004; accepted 1 April 2004 Available online 12 August 2004
Abstract In order to prepare a controlled-release tablet by direct compression, a Fourier transform infrared (FT-IR) microspectroscopic mapping system was utilized to assess the blend homogeneity of a pharmaceutical powder blend. The model powder blend was a two-component mixture consisting of captopril as a model drug and micronized ethylcellulose (EC) as a direct-compressible model excipient, which was mixed in a laboratory mill. The two-component mixture was mixed in two different weight proportions (captopril:EC, 1:1 or 10:1). By varying the mixing time, different blend mixtures sampled were determined by a reflectance FT-IR microspectroscopic mapping system to collect successively the IR spectra from the actual analysis area. The results revealed that the blend homogeneity increased gradually with increased mixing times, but the powder began to demix or segregate as mixing continued beyond the time when homogeneity was reached. In addition, the statistical results indicated that the sample mixture proportion had an effect on the uniformity of the powder blend. This study demonstrates that microspectroscopic FT-IR mapping technique can be easily used to determine the blend homogeneity in a powder blend mixture. © 2004 Elsevier B.V. All rights reserved. Keywords: FT-IR; Mapping; Blend homogeneity; Micronized ethylcellulose; Captopril; Mixing ratio
1. Introduction In the preparation of pharmaceutical solid dosage forms, the blending process is one of the most common unit operations to uniformly mix the active drug ingredient with one or more excipients to achieve the in-process and product control. The homogeneous blending ensures a uniform distribution and reproducibility of all components in the end product (Berntsson et al., 2002; Lai et al., 2001; Sekulic et al., 1998; Chang et al., 1996). Because blend uniformity analysis has also been requested by good manufacturing practice (GMP) regulation in US Food and Drug Administration’s draft guidance, it should consistently test the blending mixture to ensure adequate blend homogeneity and stability in the blending process. The testing by high-performance liquid chromatography (HPLC) and ultraviolet (UV) spectroscopy has been typically used to evaluate the blend or content uniformity of blend samples or dosage unit samples in pharmaceutical ∗
Corresponding author. Fax: +886-2-2873-7200. E-mail address: [email protected] (S.-Y. Lin).
0928-0987/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ejps.2004.04.015
manufacturing. Recently, vibrational spectroscopic techniques, including near-infrared, mid-infrared, and Raman spectroscopies, have been employed as alternatives to the common destructive methods in order to save time and reduce the requirement for large amounts of toxic and expensive solvents (Blanco et al., 2002; Gustafsson et al., 2003; Johansson et al., 2002). These vibrational spectroscopic methods provide non-destructively physical and chemical characterization of the active and inactive components in the composite mixtures, but the single point of spectroscopic determination cannot directly reflect the spatial distribution of the components in the mixture. Thus, a mapping technique is included with the vibrational spectroscopy to map the location of each spatially resolved component (Smith, 2000). This combination of a mapping system and vibrational spectroscopy for the molecular analysis of components is achieved by scanning techniques. The Fourier transform infrared (FT-IR) microspectroscopic mapping system is a powerful technique that combines the image analysis of light microscopy, chemical analysis of FT-IR spectroscopy, and a mapping stage. A mapping stage can translate the sample along the X- and
118
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
Y-axes thereby moving different areas of the sample into the microscope’s beam path. The IR spectra can then be directly related back to the sample’s microstructure, and can also reflect the existence and content of the sample. This combined system not only reduces the blend times by allowing accurate determination of end-points, but also monitors the homogeneity of all components, since the homogenization process affects not only the active ingredient but also the excipients. The aim of this study was to evaluate the FT-IR microspectroscopic mapping system as a tool to assess the blend uniformity and potency of the blend mixture of captopril and micronized ethylcellulose (EC) used for preparing the directly compressed controlled-release tablet (Lin et al., 2001a,b, 2002). In our preliminary study, we found that two unique IR spectra peaks at 1747cm-1 for captopril and at 3478cm-1 for EC could be used as markers to study the blend uniformity of captopril and EC mixture. Blend homogeneity and optimal mixing times have been examined using a statistical analysis.
2. Materials and methods 2.1. Material Ethylcellulose powder with a particle size of 4.0mm (grade N-7-F; average molecular weight, 58,000) was obtained from Shin-Etsu Chem. Ind. Co. Ltd. (Tokyo, Japan). Captopril (<149mm) was of pharmaceutical grade and was purchased from Bulk Med. Pharm., Hamburg, Germany. 2.2. Sample preparation All mixing studies were carried out in an analytical mill (Model A-70, Morat, Germany). A two-component mixture consisting of captopril and EC (weight ratio (w/w), 1:1 or 1:10) was powder blended. The blending process was monitored. At predetermined timepoints, an aliquot of the blended powder was carefully sampled and compressed using an IR spectrophotometric hydraulic press (Riken Seiki Co., Tokyo, Japan) under constant pressure at 400kg/cm2 for 1min.
2.3. Microspectroscopic FT-IR mapping study The powder sample compressed on the die was set on the stage of FT-IR microscopic spectrometer (Micro IRT-30, Jasco, Tokyo, Japan) equipped with a MCT detector. The system was operated in a reflectance mode over the range 4600–1000cm-1 . An appropriate sample area was selected, and the reflectance IR spectra were collected successively from the actual analysis area by a mapping process. An automated X–Y stage for mapping was employed in order to obtain full IR spectra from the sample at 18mm intervals across the 35 spots ’ 35 spots. The use of an 18mm step size results in the acquisition of 1225 spectra. The sampling aperture was set at 40mm ’ 40mm, each spectrum was performed at 10 scans with resolution of 4cm-1 . Fig. 1 depicts the concept diagram of data collection method and processing required in the mapping experiments. 2.4. Statistical analysis of blend sample variability The blend uniformity of the captopril–EC mixture was estimated from the mapping spectra by evaluating the consistency of the IR peak intensity ratio of the above unique IR peaks in each reflectance spectra of EC or captopril, as represented by the IR spectral map. Spectral manager software for Windows (Jasco Co., Japan) was used. The sampling process was done in triplicates. The mixture was estimated to be homogeneous when their IR peak intensity ratios of these unique IR peaks was consistent as exhibited by histogram or modified box plot. The mixture was termed homogeneous when the standard deviation of the IR spectrum reached a minimum value. Analyses were performed using SPSS 11.0 for windows, Statistica 6, and SigmaPlot 2000.
3. Results and discussion The representative FT-IR spectra of EC, captopril, and the EC–captopril mixture are shown in Fig. 2. As expected, a number of differences can be observed between the spectra of EC and captopril. Captopril had two unique IR peaks, at 1747 and 1590cm-1 , which can be assigned to the C==O
Fig. 1. Concept diagram of data collection method and processing required in mapping experiments.
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
Fig. 2. FT-IR spectra of captopril (a), EC (b), and the blended captopril–EC (1:1) mixture (c).
stretching vibration of carboxylic acid and amide band, respectively. However, EC did not have these features in this range but exhibited a unique peak at 3478cm-1 due to hydroxyl groups. This clearly suggests that it is possible to distinguish the spectral difference from the mixture by determining the peak intensity ratio of the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ). Fig. 3 presents the IR spectral maps and frequency distribution plots of the blended samples (captopril:EC, 1:1 (w/w)) in the process of mixing. The corresponding maps and plots represent a 3D and 2D distribution of the peak height ratio of the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ). The width of the distribution shown in each frequency distribution plot is an indicator of the heterogeneity of the sample. Results of this analysis clearly indicate that the sample was poorly blended after up to 0.5min of mixing. The variety of ratios presented in the map indicates that the two pharmaceutical powders were not uniformly distributed within the blended sample. A wide distribution was exhibited in the corresponding distribution plot as well. The IR spectral map for the blended sample after mixing for 1min exhibited an improved uniformity compared with that of the 0.1 or 0.5min blended mixture. The frequency distribution plot had a narrower range. With increasing time of mixing (up to 3min), the blended mixture had a more uniform distribution of components in which the distribution of the peak height ratios was grouped together. This reveals that the two pharmaceutical powders are at a nearly constant ratio within the blended sample. From these results, it can be seen that blend homogeneity is achieved rapidly. On the other hand, the high-peak height ratio and low-peak height ratio contributing to the slight extension in tailing of the distribution plot was observed at 5min. The blended samples from 0.1 to 3min showed a progressive decrease in heterogeneity, while the blended samples from 4 to 7min demonstrated a progressive increase in heterogeneity. The segregation occurring in the mixture might be responsible for this observation.
119
The changes in IR spectral maps and frequency distribution plots of blended samples with 1:10 (w/w) ratio of captopril to EC after mixing were also investigated. The optimum blending time for mixing the 1:10 weight ratio of captopril to EC was about 1–6min, as compared with the blending time for the mixing the 1:1 weight ratio of captopril to EC. This also indicates that the proportion of sample mixture also played an important role to affect the uniformity of the powder blend. Fig. 4A illustrates the relationship between the peak height ratio for the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ) versus mixing time. Clearly, the values of peak height ratios were gradually reduced and concentrated. The effect of blending time on homogeneity is shown in Fig. 4B, which depicts the S.D. versus blending time. In common, the homogeneous mixture had aminimum value of S.D. As seen in Fig. 4B, the S.D. decreases as a function of time, indicating that the blends are approaching homogeneity. It was gradually reduced with the blending progress, and reached to a minimum but after that begun to enlarge again. A progressive decrease in the S.D. was observed as the degree of blending developed. Apparently, the blend homogeneity was achieved after approximately 3min in the blending process. This might be attributed to the particle size and size distribution changes arising from attrition in the blending, which is consistent with results from other studies (Wargo and Drennen, 1996; Blanco et al., 2002; El-Hagrasy et al., 2001). Fig. 5A depicts the relationship between the peak height ratio for the IR bands of captopril (1747cm-1 ) and EC (3478cm-1 ) versus mixing time. Apparently, the values of peak height ratios were gradually reduced. A progressive decrease in the S.D. was observed as the degree of blending developed. The effect of blend time on homogeneity is evident as shown in Fig. 5B, which depicts the sample standard deviations versus blending time. A blend homogeneity was reached after 6min. But the powder begun to demix or segregate as mixing continued beyond the time when homogeneity was reached, exhibiting increased variability after 6min. This same phenomenon was observed in the 1:1 mixture as discussed above. This result indicates that the proportion of sample mixture also played an important role to affect the uniformity of the powder blend, which is consistent with other reports (Kornchankul et al., 2002). For this study, the optimum blending time for mixing captopril to EC was about 3min for 1:1 weight ratio while it was almost 6min for 1:10 weight ratio. In conclusion, the homogeneity of the blend samples can be determined using mid-infrared microspectroscopic mapping, which indicates that this technique has great potential as an analytical tool for blend uniformity analysis. Qualitative IR analysis can be employed to assess the uniformity of a single production blend or to define the optimal mixing time during the development process. These experiments have demonstrated that both IR spectral maps and frequency distribution plots for equality of
120
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
Fig. 3. Three-dimensional FT-IR spectral maps and frequency distribution plots of the blended captopril–EC (1:1) mixture in the process of blending.
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
121
Fig. 4. Relationship between the spectral peak height ratio and mixing time for the captopril–EC mixture (1:1) (A) and the effect of blending time on homogeneity (B).
Fig. 5. The relationship between the spectral peak height ratio and mixing time for the captopril–EC mixture (1:10) (A) and the effect of blending time on homogeneity (B).
variance are effective methods for qualitative evaluation of powder blend homogeneity. Compared with traditional IR spectroscopy (single point), IR mapping provides the opportunity to investigate a large area within a single sample. A minimal abnormality in a bulk sample may be undetected by using traditional IR spectroscopy but it may be detected using IR mapping. This study indicates the FT-IR microspectroscopic mapping technique for direct analysis of powder blend mixture can be easily used to determine the homogeneity of a pharmaceutical blend.
References Berntsson, O., Danielsson, L.G., Lagerholm, B., Folestad, S., 2002. Quantitative in-line montoring of powder blending by near infrared reflection spectroscopy. Powder Technol. 123, 185–193. Blanco, M., Gozález Bañó, R., Bertran, E., 2002. Monitoring powder blending in pharmaceutical processes by use of near infrared spectroscopy. Talanta 56, 203–212. Chang, R., Shukla, J., Buehler, J., 1996. An evaluation of a unit-dose compacting sample thief and a discussion of content uniformity testing and blending validation issues. Drug Dev. Ind. Pharm. 22, 1031–1035.
El-Hagrasy, A.S., Morris, H.R., D’Amico, F., Lodder, R.A., Drennen III, J.K., 2001. Near-infrared spectroscopy and imaging for the monitoring of powder blend homogeneity. J. Pharm. Sci. 90, 1298–1307. Gustafsson, C., Nyström, C., Lennholm, H., Bonferoni, M.C., Caramella, C.M., 2003. Characteristics of hydroxypropyl methylcellulose infulencing compactibility and prediction of particle and tablet properties by infrared spectroscopy. J. Pharm. Sci. 92, 494–504. Johansson, J., Pettersson, S., Taylor, L.S., 2002. Infrared imaging of laser-induced heating during Raman spectroscopy of pharmaceutical solids. J. Pharm. Biomed. Anal. 30, 1223–1231. Kornchankul, W., Hamed, E., Parikh, N.H., Sakr, A., 2002. Effect of drug proportion and mixing time on the content uniformity of a low dose drug in a high shear mixer. Pharmazie 57, 49–53. Lai, C.K., Holt, D., Leung, J.C., Conney, C.L., Raju, G.K., Hansen, P., 2001. Real time and noninvasive monitoring of dry powder blend homogeneity. AIChE J. 47, 2618–2622. Lin, S.Y., Lin, K.H., Li, M.J., 2001a. Micronized ethylcellulose used for designing a directly compressed time-controlled disintegration tablet. J. Controlled Release 70, 321–328. Lin, K.H., Lin, S.Y., Li, M.J., 2001b. Compression forces and amount of outer coating layer affecting the time-controlled disintegration of the compression-coated tablets prepared by direct compression with micronized ethylcellulose. J. Pharm. Sci. 90, 2005– 2009. Lin, S.Y., Lin, K.H., Li, M.J., 2002. Infulence of excipients, drugs, and osmotic agent in the inner core on the time-controlled disintegration
122
T.-H. Lee, S.-Y. Lin / European Journal of Pharmaceutical Sciences 23 (2004) 117–122
of compression-coated ethylcellulose tablets. J. Pharm. Sci. 91, 2040– 2046. Sekulic, S.S., Wakeman, J., Doherty, P., Hailey, P.A., 1998. Automated system for the on-line montoring of powder blending processes using near-infrared spectroscopy. II. Qualitative approaches to blend evaluation. J. Pharm. Biomed. Anal. 17, 1285–1309.
Smith, P.A.M., 2000. Infrared microspectroscopy mapping studies of packaging materials: experiment design and data profiling considerations. Vib. Spectrosc. 24, 47–62. Wargo, D., Drennen, J.K., 1996. Near-infrared spectroscopic characterization of pharmaceutical powder blends. J. Pharm. Biomed. Anal. 14, 1415–1423.