Application of Terahertz Attenuated Total Reflection Spectroscopy to Detect Changes in the Physical Properties of Lactose during the Lubrication Process Required for Drug Formulation

Abstract

Manufacturing the solid dosage form of an orally administered drug requires lubrication to enhance manufacturability, ensuring that critical quality attributes such as disintegration and dissolution of the drug product are maintained during manufacture. Here, to evaluate lubrication performance during manufacture, we used terahertz attenuated total reflection (THz-ATR) spectroscopy to detect differences in the physical characteristics of the lubricated powder. We applied a simple formulation prepared by blending granulated lactose as filler with magnesium stearate as lubricant. A flat tablet was prepared using the lubricated powder to acquire sharp THz-ATR absorption peaks of the samples. First, we investigated the effects of lubricant concentration and compression pressure on preparation of the tablet and then determined the effect of the pressure applied to samples in contact with the ATR prism on sample absorption amplitude. We focused on the differences in the magnitudes of spectra at the lactose-specific frequency. Second, we conducted the dynamic lubrication process using a 120-L mixer to investigate differences in the magnitudes of absorption corresponding to the lactose-specific frequency during lubrication. In both studies, enriching the lubricated powder with a higher concentration of magnesium stearate or prolonging blending time correlated with higher magnitudes of spectra at the lactose-specific frequency. Further, in the dynamic lubrication study, the wettability and disintegration time of the tablets were compared with the absorption spectra amplitudes at the lactose-specific frequency. We conclude that THz-ATR spectroscopy is useful for detecting differences in densities caused by a change in the physical properties of lactose during lubrication.

The manufacture of the tablet or capsule form of an orally administered drug requires blending of a lubricant with granules or other excipients, including the active pharmaceutical ingredient (API). This important process enhances the flow of the powder,¹⁾ and thereby helps optimize manufacturing processes such as tableting and encapsulation. Factors such as the type of blender, rotation speed, lubrication time, batch and equipment size can affect the quality attribute of a drug product.^2,3) The factors are generally determined according to scale-up theories to maintain the same physicochemical properties of a lubricated powder when changes are made to the equipment or production scale of the lubrication process.^4–7) However, even when using the same manufacturing conditions, we must consider variations in lubrication performance among batches of lubricated powders that are caused, for example, by differences in the physical characteristics of the lubricant or the physical characteristics of granules that are manufactured before the lubrication process.⁸⁾

To conduct real-time monitoring of the uniformity of the blended powder during production (e.g. uniformity of the API), near-infrared spectroscopy (NIRS) is utilized as a nondestructive process analytical technology (PAT) tool. PAT facilitates the design, analysis and control of the manufacture of pharmaceuticals through real-time measurements (i.e., during processing) of critical quality and performance attributes of raw and in-process materials and processes, with the goal of ensuring final product quality⁹⁾ (U.S. Food and Drug Administration (FDA)). Thus, combined with real-time monitoring and process control, PAT systems help minimize the variability of critical quality attributes caused by variation in material characteristics.⁹⁾

The in-line NIRS system is used to determine the end of the blending process so that blending can be terminated at the optimum time for each production batch according to in-line data.^10–15) Previous studies have used an NIRS system to monitor the homogeneity of a lubricant.^16,17) However, the blend uniformity of lubricant as well as the change in the flatting state of the lubricant must be monitored during the lubrication process. Prolonging the lubrication time leads to an increased flow of powders, which decreases adhesion to the equipment because of the enhanced flatting state of the lubricant. In contrast, prolonged lubrication times cause a wider flatting state of the lubricant through the hydrophilic excipients; this wider flatting decreases hardness of the solid dosage,^2,3,18,19) or delays disintegration or dissolution of the API, or both.^18–20) The flatting state of a lubricant is therefore a critical quality attribute of an oral dosage form, and should be monitored and controlled during lubrication. Although the uniformity of the lubricant can be monitored during the process using an NIRS, few studies have addressed the applications used to evaluate lubricant performance.²¹⁾

Recent advances in technology indicate that terahertz (THz) pulsed imaging and THz pulsed spectroscopy may provide a suitable alternative PAT tool, and THz instruments have been evaluated and utilized in pharmaceutical applications.^22,23) The THz region of the electromagnetic spectrum spans the frequencies between the mid-infrared and microwave wavelengths (0.1–4 THz, 3.3–133 cm⁻¹). Many pharmaceutical excipients are transparent or semi-transparent to THz radiation, whereas many crystalline materials have characteristic spectral features.²⁴⁾ The ability of THz radiation to penetrate many pharmaceutical materials allows characterization of the structural features of coating components, and THz pulsed imaging technology is used in drug research to quantitatively characterize the coating quality of pharmaceutical tablets.^25,26) Moreover, absorption features in the THz region are dominated by intermolecular vibrations. Therefore, THz pulsed spectroscopy serves as an excellent technique for characterizing the crystal properties of solids, as well as other techniques such as powder X-ray diffractometry, differential scanning calorimetry, IR spectroscopy, Raman spectroscopy and solid state NMR.²³⁾ For example, THz pulsed spectroscopy detects crystal transition rates and assesses polymorphic compositions, which contribute to the determination of a compound’s structure.^27–31)

To date, the transmission method of THz pulsed spectroscopy systems has been most frequently used to obtain high-quality and reproducible spectra by mixing samples with polyethylene powder.²⁴⁾ An alternative to the transmission method is attenuated total reflection (ATR). Total internal reflection occurs when a THz beam irradiates the interface between the crystal and sample at an angle greater than the critical angle, and total interface reflection occurs only when the refractive index of the crystal is larger than that of the sample. These total internal reflections form evanescent waves that extend into and interact with the sample. The absorbance spectra of samples in the THz region are acquired by placing the sample onto the crystal. Previous studies show that THz-ATR spectral signatures agree with those obtained using the THz-transmission system.^32,33) Further, THz-ATR measurements are useful for screening polymorphs, because only a small amount of sample is required.

Note that the penetration depth of THz radiation into samples using the ATR method varies with wavelength and the incident angle of the THz beam. The exponential dependence of electromagnetic field intensity is observed at a distance from the interface. Compared with IR-spectroscopy, THz-ATR utilizes much longer wavelengths, and the depth of the evanescent wave in the THz frequency-range is much deeper than that of the IR-region. We suggest therefore that small differences in the physical characteristics of powders might be detected as differences in the absorption amplitude of the samples in the deep evanescent wave detected by THz-ATR analysis. The use of THz-ATR techniques in the analysis of polymorphs, crystallization, and other forms has been evaluated,^32–34) but we are unaware of any research into the use of THz-ATR to determine differences in the physical characteristics of powders generated by differences caused by lubrication performance.

Here, we used the simple formulation of a blend of granulated lactose as filler and magnesium stearate as lubricant to prepare a flat tablet. THz-ATR spectroscopy of these tablets generated sharp absorbance peaks. We then used THz-ATR spectroscopy to measure the absorption amplitudes of these samples to determine the effects of varying the lubricant concentration, the compression pressure used to form the tablet, and the pressure applied to samples placed on the ATR prism. Dynamic lubrication was conducted in a 120-L mixer to investigate differences in the magnitude of the spectra as a function of blending time. The wettability and disintegration times of tablet samples were then compared with the spectral data (Fig. 1).

Fig. 1. THz-ATR Spectroscopy

Results and Discussion

THz-ATR Spectroscopy of Granulated Lactose and Magnesium Stearate

We first acquired the spectra of thin flat tablets composed of granulated lactose (53.1 mg, 0.502-mm thick) or magnesium stearate (47.8 mg, 0.623-mm thick) (Fig. 2a). The theoretical maximum value is 100% (Y-axis), indicating undetectable absorbance. Therefore, a smaller value represents a larger ATR, indicating strong absorbance. The peaks observed at 0.53 and 1.37 THz for granulated lactose, which are consistent with those of lactose monohydrate reported previously,³²⁾ Accounted for 90% of the attenuated total reflection at 0.53 THz suggests that absorbance by lactose was approximately 10% of the initial value. In contrast, an absorbance peak for magnesium stearate was undetectable at THz frequencies.

Fig. 2. THz Absorption Spectra of Granulated Lactose and Magnesium Stearate in Tablets

a) Attenuated total reflection. b) Transmission.

As shown in Fig. 2b, the baseline absorbance of the transmission measurement becomes uniformly higher at higher frequency (lower wavelength), which is attributed to the incident THz wave irradiating the tablets to cause diffusion and diffraction, rather than the decay of absorption from specific inherent materials.^22,35) In contrast, there is little change in the values of the baselines of the ATR measurements. We assume that the flat baseline of an ATR measurement is attributed to the much shallower depth of the evanescent wave compared with that of the THz wavelength, which is not detectably diffused and diffracted by the samples.³²⁾ Because absorption by magnesium stearate detected using THz-ATR was nearly zero (Fig. 2a), the absorbance (0.85) by granulated lactose at 0.53 THz was calculated by subtracting scattering and diffraction. The calculated absorption coefficient of granulated lactose is 0.004 µm⁻¹.

THz-ATR Spectroscopy of Lubricated Powders

Because there was no detectable specific absorption peak for magnesium stearate in the THz frequency region, we focused on the absorption peaks of granulated lactose to detect the differences in the physical properties of lubricated powder. Tablets composed of different ratios of granulated lactose and magnesium stearate were applied to the ATR prism under 100 N pressure to acquire sharp spectra (Fig. 3). To remove the effect of baseline variation, second derivative treatment with Savitzky–Golay method was applied using 15 smoothing points (Fig. 4).

Fig. 3. Raw THz-ATR Spectra of Tablets Composed of Granulated Lactose and Different Concentrations of Magnesium Stearate

Compression pressure used to form a tablet, 10 kN. Pressure on samples on the ATR prism, 100 N.

Fig. 4. Second Derivative of Absorption of Tablets Composed of Granulated Lactose and Different Concentrations of Magnesium Stearate

a) 0.53 THz. b) 1.37 THz. Compression pressure used to form tablet, 10 kN. Pressure on samples on the ATR prism, 100 N.

We found it interesting that higher concentrations of magnesium stearate corresponded to increased absorption intensities from lactose at 0.53 and 1.37 THz. Therefore, we determined to acquire the spectra of tablet samples compressed at 5 and 10 kN. The pressures applied to the samples on the ATR prism were varied from 100 to 20 and 50 N for tablets prepared using pressures of 5 or 10 kN. The second derivatives of absorbance at 0.53 and 1.37 THz are shown in Fig. 5. Higher compression pressures used to prepare the tables correlated with the higher absorption of lactose at 0.53 and 1.37 THz. We attribute these findings to the higher density of lactose on the surface of the tablet formed using higher compression pressures. Further, when we applied pressures of 20, 50 and 100 N to samples on the ATR prism, we found that the higher pressures corresponded to an increase in the lactose absorption peak, suggesting that closer contact between a sample and the ATR prism led to an increased lactose absorption peak. Further, higher concentrations of magnesium increased the lactose absorption peaks. We found it interesting that the absorption amplitude of granulated lactose on the ATR prism increased with the concentration of magnesium stearate despite the lower weight ratio of lactose in the tablets. Because a significant magnesium stearate-specific absorption peak was not detected in the THz region, the differences in amplitude of lactose absorption may be explained by the change in the physical properties of granulated lactose caused by the degree of lubrication associated with different concentrations of magnesium stearate.

Fig. 5. Second Derivative of Absorbance of Tablets Composed of Granulated Lactose and Different Concentrations of Magnesium Stearate

a) 0.53 THz. b) 1.37 THz. Compression pressures used to form the tablets, 5, 10 kN. Pressure on samples on the ATR prism: 20, 50, and 100 N.

Effect of Dynamic Lubrication Time on THz-ATR Spectra

We determined that the values of the compression pressure and the pressure applied to samples on the ATR prism were 10 kN and 100 N, respectively. Therefore, we used these pressures to evaluate the effect of lubrication time on dynamic lubrication process. The second-derivative values at 0.53 and 1.37 THz, which are come from the lactose absorption, are calculated for the samples at different lubrication times. The average second-derivative values and standard deviations of 10 samples at 0.53 and 1.37 THz are shown in Fig. 6. Prolonging the lubrication time increased the lactose-specific absorption intensity at each frequency. The average value increased with blending time, with the large change occurring from 10 to 30 min. There was no significant change in the average values of the second derivatives of absorption from 30 to 60 min (Fig. 6). Standard deviation of the second-derivative values for 10 samples that were taken from the top to the bottom of the mixing chamber was larger for shorter lubrication time. This may be ascribed to the non-uniform distribution of magnesium stearate in the blended powder because of the shorter lubrication time. In addition, second-derivative values for long lubrication time of 60 min at 0.53 and 1.37 THz were 2249±125 and 3909±318, and their coefficients of variations were 5.6 and 8.1%, respectively. These somewhat larger variations may be due to the principle of ATR spectroscopy. Because the electromagnetic field intensity of evanescent wave is decreased exponentially with the increasing distance from the surface on ATR prism, absorption on evanescent wave by the tablet strongly depends on the distances between the ingredient particles at the tablet surface and ATR prism. Larger variations of second-derivative values may be attributable to the variation of the lactose and magnesium stearate distribution at the surfaces of tablets. It is suggested that the average of second derivative values can be index as representative value for lactose-derived absorption, and increasing the number of tablets for calculation of the average value for each lubrication time point would further improve the reproducibility. The reproducibility may be also improved by optimizing the incident angle or refractive index of prism in THz-ATR system that affects the penetration depth of the evanescent wave.

Fig. 6. Effect of Lubrication Time on the Secondary Derivative of Absorbance of Tablets Composed of Granulated Lactose and 1% (w/w) Magnesium Stearate

a) 0.53 THz. b) 1.37 THz. Compression pressure used to form the tablet, 10 kN. Pressure on samples on the ATR prism, 100 N.

When we analyzed granulated lactose alone using the same blending conditions, no significant change in lactose absorption was observed at either wave number. Therefore, these findings demonstrate that the change in the lactose-specific absorption was caused by the addition of magnesium stearate. The physical change in the distribution of magnesium stearate might change the density of lactose during the lubrication process. We conclude that the higher absorbance around the lactose-specific frequency observed after longer lubrication may be explained by a decrease in distance between lactose particles in the powder.

Correlation between THz-ATR Spectra and Wettability and Disintegration Time

The contact angles formed between the trajectory of the water drops and the surface of the tablets formed under 10 kN compression pressure were measured as a function of lubrication time. The average change in contact angle from 100–2000 ms after releasing the water at each time (10 samples each) is shown in Fig. 7a. The contact angle between the tablets and water drops became larger with longer lubrication time, and therefore the wettability of the samples decreased with longer lubrication times. Further, the disintegration time of the tablets was longer with prolonged lubrication times (Fig. 7b). These findings are likely explained by the gradual spreading of hydrophobic magnesium stearate on hydrophilic lactose during blending.

Fig. 7. Wettability and Disintegration of Tablets Composed of Granulated Lactose and 1% (w/w) Magnesium Stearate with Different Blending Times

a) Change in contact angle on wettability measurement. b) Effect of blending time on the disintegration time of tablet.

To investigate the relationship between the absorption amplitude of the lactose-specific peak and the disintegration time of tablets, the average values of the second derivatives of absorption for 0.53 and 1.37 THz were plotted vs. the disintegration time of tablets in Fig. 8. The regression coefficients (R²) for the correlation between these parameters were 0.9361 and 0.9237 for 0.53 and 1.37 THz, respectively. Therefore, the disintegration times of the tablets could be estimated using nondestructive THz-ATR spectroscopy data, focusing on lactose-specific absorption. The higher value of the lactose-specific absorption peak might be explained by the shorter distance between lactose particles. Therefore, we estimated that the delayed disintegration time was caused by the hydrophobicity of the lubricant as well as the higher density of lactose in the tablets.

Fig. 8. Relationship between the Second Derivatives of Absorbance vs. Disintegration Time of Tablets Composed of Granulated Lactose and Magnesium Stearate with Different Lubrication Times

a) 0.53 THz. b) 1.37 THz.

Relationship between Lactose-Specific Absorbance and Tablet Density

We show here that increased lubrication increased lactose-specific absorption, which increased with higher concentrations of lubricant and dynamic lubrication time. To identify the factors that increased lactose-specific absorbance, we calculated the densities of core tablets according to their weights, diameters, and thicknesses at different concentrations of magnesium stearate and different dynamic lubrication blending times. The average second-derivative values of absorption at 0.53 and 1.37 THz were plotted vs. average tablet density. We observed a correlation between lactose-specific absorbance at 0.53 and 1.37 THz and the density of the core tablet (Fig. 9), indicating that THz-ATR measurements detected the difference in density of a core tablet according to the absorbance of the tablet’s surface (R² values for the correlations between tablet density and the second derivatives=0.9117 and 0.621 at 0.53 and 1.37 THz, respectively).

Fig. 9. Relationship between the Second Derivative of Absorbance vs. Tablet Density for Tablets Composed of Granulated Lactose and Different Concentrations of Magnesium Stearate and Different Lubrication Times

a) 0.53 THz. b) 1.37 THz.

We conclude that the differences in accuracy of predictions using absorption at 0.53 and 1.37 THz were caused by the respective differences in the depths of the evanescent wave on the ATR prism. The theoretical depths of the evanescent waves are 41 µm at 0.53 THz and 16 µm and 1.37 THz, suggesting that better correlation between tablet density and amplitude of absorbance of the lactose-specific peak was observed for the deeper evanescent wave at 0.53 THz. Therefore, we conclude that 0.53 THz should be utilized to improve the accuracy of detection of differences in the density of lactose.

Conclusion

We used THz-ATR spectroscopy to show that lactose-specific absorption increased as a function of the degree of lubrication, including a flatting state change that was affected by lubrication time. Moreover, lactose-specific absorption correlated with the wettability and disintegration time of tablets. These findings indicate that THz-ATR spectroscopy facilitates analysis of the physical characteristics of lubricated powders consisting of granulated lactose and magnesium stearate. We suggest that the deeper evanescent wave propagated at 0.53 THz detects changes in the physical characteristics of lactose. Therefore, ATR-THz spectroscopy may facilitate the detection of differences in physical characteristics, such as the density of lactose, as well as the estimation of disintegration and dissolution of orally administered drugs.

Experimental

Materials

Lactose and magnesium stearate were selected because these are frequently used in the preparation of solid dosage forms (e.g. tablets, capsules) of orally administered drugs. The lubricant consisted of a blend of granulated lactose, which served as the filler (Dilactose R; Freund Corporation, Japan), and magnesium stearate (vegetable origin, fine type; Taihei Chemical, Japan).

Manufacture of the Lubricated Powder

We first prepared the blended powder with different concentrations of magnesium stearate as follows: Granulated lactose and magnesium stearate were mixed to prepare 200 g samples of 0.1, 0.5, 1.0, 3.0, and 5.0% (w/w) magnesium stearate. The mixtures were contained in polyethylene bags, which were shaken manually for 3 min to prepare a homogeneous mixture. We next used a rotating-type blender (PM100, Bohre, Germany) to lubricate the mixture for different times. Specifically, 24 kg of powder consisting of 99% granulated lactose (w/w) and 1% magnesium stearate (w/w) was added to a 120-L chamber and blended at 12 rpm for 60 min. During blending, 10 samples were taken from the top to the bottom of the mixing chamber after 2, 5, 10, 30, and 60 min.

THz Spectra Acquirement

A THz spectroscopic system (TAS7500, Advantest, Japan) was used to acquire THz spectra of individual sample components and their mixtures. The absorption spectra of samples were acquired by calculating the differences between spectra with and without the samples (Fig. 1). Measurement conditions for THz-ATR spectroscopy were as follows: Incident angle, 57°; beam diameter, 7 mm, 0.53 THz; pressures applied to samples on the surface of the ATR prism, 20, 50, and 100 N; cumulative background value, 4096; cumulative number of samples, 2048; and frequency resolution, 7.6 GHz (0.13 cm⁻¹).

The depth of the evanescent wave (D_ev) is calculated as follows:   

λ: wavelength (nm), θ: incident angle, n₁: refractive index of prism, n₂: refractive index of sample.

Tablet Sample Preparation

First, thin flat tablets composed of granulated lactose and magnesium stearate were prepared to obtain the THz spectrum of each component. These tablets were used to evaluate the THz-ATR and THz-transmission methods. The blended powders (180.0±3.0 mg) were compressed using an 8-mm diameter flat punch and die using compression forces of 5 or 10 kN. The weight, thickness, and diameter of each tablet were determined.

Wettability and Disintegration Time

To evaluate the wettability of lubricated powders, the contact angle between the drops (2 µL) of purified water relative to a tablet was measured using a contact angle meter (LCD-400S; Kyowa Interface Science Co., Ltd., Japan). The contact angle is conventionally measured through the liquid between a liquid/vapor interface and the solid surface. This quantifies the wettability of a tablet via the Young equation. The contact angle θ was determined using the equation below (r=radius and h=height h of the droplet).   

Further, the disintegration times of the tablets were determined using a disintegration tester (NT-20T, Toyama Sangyo Co., Ltd., Japan). Tablets were added to purified water at 37°C in the disintegration tester and shaken vertically until the tablet was fully disintegrated.

Acknowledgment

The authors would like to thank all members at Astellas Pharma Inc. related to the manufacturing and evaluation of the study.

Conflict of Interest

Wataru Momose, Kazunari Yamashita, Tadashi Hakomori and Masafumi Dohi are employees of Astellas Pharma Inc.

References

• 1) Podczeck F., Mia Y., Int. J. Pharm., 144, 187–194 (1996).
• 2) Bossert J., Stains A., Drug Dev. Ind. Pharm., 6, 573–589 (1980).
• 3) Bolhuis G. K., De Jong S. W., Lerk C. F., Dettmers H., Pharbita B. V., Drug Dev. Ind. Pharm., 13, 1547–1567 (1987).
• 4) Kushner J. 4th, Moore F., Int. J. Pharm., 399, 19–30 (2010).
• 5) Kushner J. 4th, Int. J. Pharm., 429, 1–11 (2012).
• 6) Kushner J. 4th, Schlack H., Int. J. Pharm., 475, 147–155 (2014).
• 7) Mosig J., Kleinebudde P., Powder Technol., 266, 156–166 (2014).
• 8) Kato H., Kimura K., Izumi S., Nakamichi K., Danjo K., Sunada H., J. Drug Del. Sci. Tech., 15, 475–480 (2005).
• 9) Food, Drug Administration, “Guidance for Industry, PAT-A Framework for Innovative Pharmaceutical Development. Manufacturing, and Quality Assurance,” 2004.
• 10) Wargo D. J., Drennen J. K., J. Pharm. Biomed. Anal., 14, 1415–1423 (1996).
• 11) Hailey P. A., Doherty P., Tapsell P., Oliver T., Aldridge P. K., J. Pharm. Biomed. Anal., 14, 551–559 (1996).
• 12) Sekulic S. S., Wakeman J., Doherty P., Hailey P. A., J. Pharm. Biomed. Anal., 17, 1285–1309 (1998).
• 13) Wu H., Tawakkul M., White M., Khan M. A., Int. J. Pharm., 372, 39–48 (2009).
• 14) Liew C. V., Karande A. D., Heng P. W. S., Int. J. Pharm., 386, 138–148 (2010).
• 15) Sulub Y., Wabuyele B., Gargiulo P., Pazdan J., Cheney J., Berry J., Gupta A., Shah R., Wu H., Khan M., J. Pharm. Biomed. Anal., 49, 48–54 (2009).
• 16) El Hagrasy A. S., Chang S., Kiang S., Pharm. Dev. Technol., 11, 303–312 (2006).
• 17) Duong N. H., Arratia P., Muzzio F., Lange A., Timmermans J., Reynolds S., Drug Dev. Ind. Pharm., 29, 679–687 (2003).
• 18) Kikuta J., Kitamori N., Drug Dev. Ind. Pharm., 20, 343–355 (1994).
• 19) Sheskey P. J., Robb R. T., Moore R. D., Boyce B. M., Drug Dev. Ind. Pharm., 21, 2151–2165 (1995).
• 20) Billany M. R., Richards J. H., Drug Dev. Ind. Pharm., 8, 497–511 (1982).
• 21) Nakagawa H., Kano M., Hasebe S., Suzuki T., Wakiyama N., Int. J. Pharm., 441, 402–413 (2013).
• 22) Shen Y. C., Int. J. Pharm., 417, 48–60 (2011).
• 23) Zeitler J. A., Taday P. F., Newnham D. A., Pepper M., Gordon K. C., Rades T., J. Pharm. Pharmacol., 59, 209–223 (2007).
• 24) Walther M., Fischer B. M., Uhd Jepsen P., Chem. Phys., 288, 261–268 (2003).
• 25) Haaser M., Gordon K. C., Strachan C. J., Rades T., Int. J. Pharm., 457, 510–520 (2013).
• 26) Knop K., Kleinebudde P., Int. J. Pharm., 457, 527–536 (2013).
• 27) Ikeda Y., Ishihara Y., Moriwaki T., Kato E., Terada K., Chem. Pharm. Bull., 58, 76–81 (2010).
• 28) Otsuka M., Nishizawa J., Shibata J., Ito M., J. Pharm. Sci., 99, 4048–4053 (2010).
• 29) Siegrist K., Bucher C. R., Mandelbaum I., Hightwalker A. R., Balu R., Gregurick S. K., Plusquellic D. F., J. Am. Chem. Soc., 128, 5764–5775 (2006).
• 30) Strachan C. J., Taday P. F., Newnham D. A., Gordon K. C., Zeitler J. A., Pepper M., Rades T., J. Pharm. Sci., 94, 837–846 (2005).
• 31) Taday P. F., Bradley I. V., Arnone D. D., Pepper M., J. Pharm. Sci., 92, 831–838 (2003).
• 32) Newnham D. A., Taday P. F., Appl. Spectrosc., 62, 394–398 (2008).
• 33) Takebe G., Kawada Y., Akiyama K., Takahashi H., Takamoto H., Hiramatsu M., J. Pharm. Sci., 102, 4065–4071 (2013).
• 34) Nagai M., Yada H., Arikawa T., Tanaka K., Int. J. Infrared Milli., 27, 505–515 (2006).
• 35) Wu H., Heilweil E. J., Hussain A. S., Khan M. A., Int. J. Pharm., 343, 148–158 (2007).