Some explanation on the carbonyl vibrations of the solid state of Fusidic acid文献综述

 2023-01-02 06:01

BackgroundThe FTIR spectra of two solid states of fusidic acid different greatly in carbonyl vibrations: form I of fusidic acid shows one carbonyl vibrational peaks in 1720 cm-1 while form III exhibits two peaks in 1748 cm-1 and 1688 cm-1 but the reason is unknown. Although lack of the crystal structures of both forms, we hypothesize a conformational difference between the crystal forms and investigate their vibrational spectra using computational chemistry method.Vibrational spectroscopic techniques such as infrared, near-infrared and Raman spectroscopy have become popular in detecting and quantifying polymorphism of pharmaceutics since they are fast and non-destructive. This study assessed the ability of three vibrational spectroscopy combined with multivariate analysis to quantify a low-content undesired polymorph within a binary polymorphic mixture. Partial least squares (PLS) regression and support vector machine (SVM) regression were employed to build quantitative models. Fusidic acid, a steroidal antibiotics, was used as the model compound. It was found that PLS regression performs slightly better than SVM regression in all the three spectroscopic techniques. Root mean square errors of prediction (RMSEP) were ranging from 0.48% to 1.17% for diffuse reflectance FTIR spectroscopy and 1.60% to 1.93% for diffuse reflectance FT-NIR spectroscopy and 1.62% to 2.31% for Raman spectroscopy. The results indicate that diffuse reflectance FTIR spectroscopy offers significant advantages in providing accurate measurement of polymorphic content in the fusidic acid binary mixtures while Raman spectroscopy is the least accurate technique for quantitative analysis of polymorphs. IntroductionMany chemical compounds can exist in more than one crystal form. This phenomenon is referred as polymorphism and each crystalline form is known as a polymorph[1]. Different polymorphs of the same compound may have different physicochemical properties (including dissolution rate, solubility, melting point, crystal habit, density, flowability, compressibility and even color). Differences in physicochemical properties among these crystalline forms in turn could have an impact on the stability, bioavailability and processability of pharmaceutical products[2, 3]. Since polymorphism can have serious consequences for the quality or performance of drugs, it is essential to check for the existence of polymorphism and quantitate the content of the effective crystal form in order to ensure high-quality products.A wide variety of analytical techniques have been employed to characterize and quantify the pharmaceutical solids, among which, X-ray diffraction (XRD), differential scanning calorimetry (DSC) and vibrational spectroscopy (mid-infrared, near-infrared, Raman spectroscopy) are most popularly used. X-ray powder diffraction (XRPD) provides structural data to identify different polymorph. However, the appearance of XRPD patterns can be affected by sample particle size, preferred orientation and so on, which limits its use for quantitative analysis[4, 5]. Vibrational spectroscopic techniques such as mid-infrared, near-infrared and Raman spectroscopy are powerful because they allow investigating differences at the molecular level and, more importantly, they are fast, non-destructive, required no or minimal sample preparation and able to simultaneously determinate multiple components[6-8]. In particular, diffuse reflectance FTIR spectroscopy (DRIFTS) is better than the conventional IR spectroscopy because it avoids compression of sample that could induce solid-state phase transformations[9, 10]. Quantitative determinations with vibrational spectroscopy are typically based on multivariate calibration models that established a relation between sample concentrations and measured intensity for a set of samples at a range of wavelengths. Multivariate regression methods such as partial least square regression can extract qualitative and quantitative information from the spectrum, overcoming problems related to overlapped signals and non-linearities between spectral response signals and the concentration of the samples[11, 12].The model compound used in the study is fusidic acid, an antibiotics used clinically for the treatment of infectious diseases[13]. It is reported that fusidic acid exits in an amorphous and four crystal forms, designated Ⅰ-Ⅳ[14]. Form Ⅲ is the preferred polymorph in pharmaceutical products since it represents a thermodynamically more stable polymorphic form, but it is still possible to occur solid-state transformations during any stage of manufacturing operations. Therefore, it is of primary importance for the drug manufacturer to be able to identify and quantify other crystal forms of fusidic acid in form Ⅲ. In this study, we quantify polymorphic impurity form Ⅰ in commercial fusidic acid form Ⅲ using diffuse reflectance FTIR spectroscopy, diffuse reflectance FT-NIR spectroscopy and Raman spectroscopy combined with multivariate calibration techniques (partial least square regression and support vector machine regression). To our knowledge, there is no previous reports on quantifying different solid-states of fusidic acid using the present three vibrational spectroscopic techniques.Method1 Geometry optimize and calculate different conformers of fusidic acid using quantum chemistry program Gaussion09.2 Ab initio molecular dynamics calculations of dipole moment using CP2K package. Then take the Fourier transform of dipole moment to obtain the IR spectra.3 Summary results and compare the difference of the above two methods in calculating IR spectra.Experiment1) Static CalculationWe used the crystal structure of fusidic acid in Protein Data Bank with accession number %JMN as the start conformation and carried out a full geometry optimization, Since we only care concern about the carbonyl stretching, we hypothesis two chemically reasonable conformers: one without intramolecular hydrogen bond, namely conformer A and the other with a intramolecular hydrogen bond, namely conformer H. Density functional theory was applied using the Becks three-parameter exchange functional in combination with the Lee-Yang-Parrs correlation functional (B3LYP). The 6-31G (d, p) basis set was used for all atoms. In both cases, isolated molecules in gas phase were used and the IR frequency and intensity calculation were preceded by a full geometry optimization. All static calculations were carried out using the Gaussian09 suite of programs. The multifunctional wavefunction analyzer Multiwfn program was used for extract IR data from Gaussian out file (.log), which makes it easy for futher replot of IR spectra.2) Molecular Dynamics SimulationThe AIMD simulations were performed using the CP2K, a freely available quantum chemistry and solid physics software package. Density functional theory was used as an electronic structure method employing the BLPY exchange-correlation functional with Gimmes dispersion correction D3. The molecularly optimized double-zeta basis set MOLOPT-DZVR-GHT was applied to all atoms together with the corresponding Goedecker-Teter-Hutter (GTH) pseudopotentials and a plane were cutoff of 280 Ry. The nuclei were propagated with a timestep of 0.5 fs and the temperature was set to 400 K. A Wannier localization was employed to localize the molecular orbitals and obtain Wannier centers that used for calculating the dipole moment of the molecule. TRVIS, a free program package for analyzing and visualizing Monte Carlo and molecular dynamics trajectories, was used to calculate IR spectra.DR-FT-NIR spectra of fusidic acid form Ⅰ and form Ⅲ are shown in Fig. 3c. The two form of fusidic acid showed significant NIR spectral peaks and it is noted that all the peak intensities of the form Ⅰ were stronger than those exhibited by the form Ⅲ. The bands observed in DR-FT-NIR are difficult to assign because a single band may be attributable to vibrational overtone and combination bands but it is still suitable for identification and quantification purposes because two crystal form of fusidic acid showed distinct spectra pattern. The main spectra differences between form Ⅰ and form Ⅲ were observed from 8000 cm-1 to 4000 cm-1 and spectral range between 8000 cm-1 and 6000 cm-1 was selected for quantitative analysis.For quantitative analysis using DR-FT-NIR, a second derivative (second-order finite difference method with a segment size of ten) after multiplicative scatter correction (MSC) was applied to solid-state DR-FT-NIR spectra because of the broad, overlapping and low intensity absorbance peaks[19, 20]. Fig. 5 shows the DR-FT-NIR spectra of two polymorphic form pretreated by MSC and second derivative. Significant differences between spectra of polymorphic form pairs at specific wavelengths can be observed. The optimal selection of pretreatment for DR-FT-NIR data can effectively reduce the noise and therefore improve the model prediction (Table 3). After the best pretreatment, the spectra data were used to multivariate regression. Six PLS factors were used in PLS regression because six factors explained 98.00% of the spectra variation and its performance was evaluated with a RMSECV 1.4%, a RMSEC 0.5%, a RMSEP 1.0% for model 1 while for model 2, six factors explained 97.69% of the spectra variation with a RMSECV 1.1%, a RMSEC 0.5%, a RMSEP 2.0%. The results for the multivariate regression are shown in Table 4. Model 1of PLS regression produces the most accurate quantitative model with a RMSEC 0.5%, a RMSEP 1.0%. The plot of observed vs. predicted content of form Ⅰ using two regression methods and each method with two models is presented in Fig. 6. Fig. 3d shows the Raman spectra between 200 and 2000 cm-1 for the pure polymorphic form of fusidic acid. There are some differences in the Raman spectra of the two polymorphs. It is obvious that the Raman spectra of form Ⅰ has a characteristic upward sloping baseline which may due to the scatting properties of the material[19]. Besides, form Ⅲ has one characteristic peak at 758 cm-1 while in form Ⅰ the band is split into two at 764 and 743 cm-1. The peak observed in 1731 cm-1 for form Ⅰ is much stronger than the peak observed in 1746 cm-1 for form Ⅲ. For data analysis, all 45 (153) spectra was pretreated with MSC to eliminate the spectral differences introduced by the non-uniform particle size (Fig. 7). To create the PLS regression, different spectral ranges were chosen, including the range from 2000 cm-1 to 200 -1. The results suggested that the spectra range between 1940 cm-1 and 1460-1 was the most suitable and was therefore selected for the regression. In fig. 8 are plotted the estimated RMSECVs as functions of the number of factors. Four PLS factors give a RMSECV of 1.89 % for model 1 and of 1.82 % for model 2 which have minimum RMSECV. Therefore in PLS regression, 4 factors were used. The best result was obtained using PLS regression of model 1 with a RMSECV of 1.89%, RMSEC of 1.12%, RMSEP of 1.62%. The plot of observed vs. predicted content of form Ⅰ is presented in Fig. 9. References[1] G.R. Desiraju, Polymorphism: The Same and Not Quite the Same, Cryst. Growth Des. 8(1) (2008) 3-5.[2] N. Chieng, T. Rades, J. Aaltonen, An overview of recent studies on the analysis of pharmaceutical polymorphs, J. Pharm. Biomed. Anal. 55(4) (2011) 618-644.[3] H.G. Brittain, Polymorphism in Pharmaceutical Solids, Second ed., CRC Press, 2009.[4] Y. Li, P.S. Chow, R.B. Tan, Quantification of polymorphic impurity in an enantiotropic polymorph system using differential scanning calorimetry, X-ray powder diffraction and Raman spectroscopy, Int. J. Pharm. 415(1-2) (2011) 110-8.[5] M. Tiwari, G. Chawla, A.K. Bansal, Quantification of olanzapine polymorphs using powder X-ray diffraction technique, J. Pharm. Biom. Anal. 43(3) (2007) 865-872.[6] A. Heinz, M. Savolainen, T. Rades, C.J. Strachan, Quantifying ternary mixtures of different solid-state forms of indomethacin by Raman and near-infrared spectroscopy, Eur. J. Pharm. Sci. 32(3) (2007) 182-92.[7] A. Heinz, C.J. Strachan, K.C. Gordon, T. Rades, Analysis of solid-state transformations of pharmaceutical compounds using vibrational spectroscopy, J. Pharm. Pharmacol. 61(8) (2009) 971-88.[8] K. Kipouros, K. Kachrimanis, I. Nikolakakis, V. Tserki, S. Malamataris, Simultaneous quantification of carbamazepine crystal forms in ternary mixtures (I, III, and IV) by diffuse reflectance FTIR spectroscopy (DRIFTS) and multivariate calibration, J. Pharm. Sci. 95(11) (2006) 2419-31.[9] K. Kachrimanis, M. Rontogianni, S. Malamataris, Simultaneous quantitative analysis of mebendazole polymorphs A-C in powder mixtures by DRIFTS spectroscopy and ANN modeling, J. Pharm. Biomed. Anal. 51(3) (2010) 512-20.[10] K. Kipouros, K. Kachrimanis, I. Nikolakakis, S. Malamataris, Quantitative analysis of less soluble form IV in commercial carbamazepine (form III) by diffuse reflectance fourier transform spectroscopy (DRIFTS) and lazy learning algorithm, Anal. Chim. Acta. 550(1-2) (2005) 191-198.[11] B. Mevik, The pls Package: Principal Component and Partial Least Squares Regression in R, J. Stat. Soft. 18(02) (2007) 1--24.[12] G. James, D. Witten, T. Hastie, R. Tibshirani, An Introduction to Statistical Learning: with Applications in R, 2014.[13] M.M. Curbete, H.R. Salgado, A Critical Review of the Properties of Fusidic Acid and Analytical Methods for Its Determination, Crit. Rev. Anal. Chem. 46(4) (2016) 352-60.[14] S.E. Gilchrist, K. Letchford, H.M. Burt, The solid-state characterization of fusidic acid, Int. J. Pharm. 422(1-2) (2012) 245-53.

资料编号：[380533]

剩余内容已隐藏，您需要先支付 10元 才能查看该篇文章全部内容！立即支付

以上是毕业论文文献综述，课题毕业论文、任务书、外文翻译、程序设计、图纸设计等资料可联系客服协助查找。

您可能感兴趣的文章