Quantitative analysis of sulfathiazole polymorphs in ternary mixtures by attenuated total reflectance infrared,near-infrared and Raman spectroscopy

The simultaneous quantitative analysis of sulfathiazole polymorphs (forms I, III and V) in ternary mixtures by attenuated total reflectance-infrared (ATR-IR), near-infrared (NIR) and Raman spectroscopy combined with multivariate analysis is reported. To reduce the effect of systematic variations, four different data pre-processing methods; multiplicative scatter correction (MSC), standard normal variate (SNV), first and second derivatives, were applied and their performance was evaluated using their prediction errors. It was possible to derive a reliable calibration model for the three polymorphic forms, in powder ternary mixtures, using a partial least squares (PLS) algorithm with SNV pre-processing, which predicted the concentration of polymorphs I, III and V. Root mean square errors of prediction (RMSEP) for ATR-IR spectra were 5.0%, 5.1% and 4.5% for polymorphs I, III and V, respectively, while NIR spectra had a RMSEP of 2.0%, 2.9%, and 2.8% and Raman spectra had a RMSEP of 3.5%, 4.1%, and 3.6% for polymorphs I, III and V, respectively. NIR spectroscopy exhibits the smallest analytical error, higher accuracy and robustness. When these advantages are combined with the greater convenience of NIR's “in glass bottle” sampling method both ATR-IR and Raman methods appear less attractive.     

Simultaneous quantitative analysis of mebendazole polymorphs A–C in powder mixtures by DRIFTS spectroscopy and ANN modeling

In the present study, a simple method, based on diffuse reflectance FTIR spectroscopy (DRIFTS) and artificial neural network (ANN) modeling is developed for the simultaneous quantitative analysis of mebendazole polymorphs A–C in powder mixtures. Spectral differences between the polymorphs are elucidated by computationally assisted band assignments on the basis of quantum chemical calculations, and subsequently, the spectra are preprocessed by calculation of 1st and 2nd derivatives. Then ANN models are fitted after PCA compression of the input space. Finally the predictive performance of the ANNs is compared with that of PLS regression. It was found that simultaneous quantitative analysis of forms A–C in powder mixtures is possible by fitting an ANN model to the 2nd derivative spectra even after PCA compression of the data (RMSEP of 1.75% for form A, 1.85% for B, and 1.65% for C), while PLS regression, applied for comparison purposes, results in acceptable predictions only within the 700–1750 cm⁻¹ spectral range and after direct orthogonal signal correction (DOSC), with RMSEP values of 2.69%, 2.68%, and 3.40% for forms A, B, and C, respectively. Application of the ANN to commercial samples of raw material and formulation (suspension) proved its suitability for the prediction of polymorphic content.     

Quantitative on-line vs. off-line NIR analysis of fluidized bed drying with consideration of the spectral background

Quantitative dehydration studies of dibasic calcium phosphate anhydrous (DCPA) in a small-scale cold-model fluidized bed dryer with process air control were conducted. Near-infrared spectroscopy (NIRS) with partial least squares regression (PLSR) was used to predict DCPAs’ residual moisture content. Loss-on-drying (LOD) was employed as a reference method and confirmed the actual moisture content of DCPA. First, dynamic PLSR modeling was carried out, i.e., the NIR spectra were on-line recorded and predicted throughout the drying process. Secondly, PLSR off-line modeling was performed, i.e., samples were consecutively thief-probed from the processor, put into glass vials and analyzed off-line. Furthermore, two background spectra were collected prior to the in- and off-line measurements in an attempt to increase the method’s sensitivity, i.e., (i) dry DCPA that was fluidized at respective process air velocity (on-line) or inside a glass vial (off-line) and (ii) Spectralon® – a highly reflecting standard reference material made of fluoropolymer. Benefits and drawbacks of the in- and off-line approaches with different spectral backgrounds are discussed in detail. The results indicated that (i) the thief-probed sample amount from the processor and thus the sample weight and (ii) the downtime between thief-probing a sample and its actual analysis via NIRS and LOD can bias the moisture content predictions.     

Quantitative solid-state analysis of three solid forms of ranitidine hydrochloride in ternary mixtures using Raman spectroscopy and X-ray powder diffraction

The aim of the study was to develop a reliable quantification procedure for mixtures of three solid forms of ranitidine hydrochloride using X-ray powder diffraction (XRPD) and Raman spectroscopy combined with multivariate analysis. The effect of mixing methods of the calibration samples on the calibration model quality was also investigated. Thirteen ternary samples of form 1, form 2 and the amorphous form of ranitidine hydrochloride were prepared in triplicate to build a calibration model. The ternary samples were prepared by three mixing methods (a) manual mixing (MM) and ball mill mixing (BM) using two (b) 5 mm (BM5) or (c) 12 mm (BM12) balls for 1 min. The samples were analyzed with XRPD and Raman spectroscopy. Principal component analysis (PCA) was used to study the effect of mixing method, while partial least squares (PLS) regression was used to build the quantification models. PCA score plots showed that, in general, BM12 resulted in the narrowest sample clustering indicating better sample homogeneity. In the quantification models, the number of PLS factors was determined using cross-validation and the models were validated using independent test samples with known concentrations. Multiplicative scattering correction (MSC) without scaling gave the best PLS regression model for XPRD, and standard normal variate (SNV) transformation with centering gave the best model for Raman spectroscopy. Using PLS regression, the root mean square error of prediction (RMSEP) values of the best models were 5.0–6.9% for XRPD and 2.5–4.5% for Raman spectroscopy. XRPD and Raman spectroscopy in combination with PLS regression can be used to quantify the amount of single components in ternary mixtures of ranitidine hydrochloride solid forms. Raman spectroscopy gave better PLS regression models than XRPD, allowing a more accurate quantification.     

基于近红外光谱技术和人工神经网络算法的佐匹克隆片定量分析

目的 将近红外光谱与人工神经网络算法相结合建立佐匹克隆片的定量分析模型,用于佐匹克隆片的快速检验。方法 采用近红外光谱分析方法获得不同批次佐匹克隆片的光谱数据,借助主成分分析方法对数据进行降维,通过人工神经网络算法建立佐匹克隆片的定量分析模型。结果 主成分分析方法结果显示,选取前10个主成分的累计贡献率为99.07%,可以解释原始光谱99.07%的信息。建立的10-6-1三层人工神经网络模型,对整个样本集的回归系数达到0.994,模型预测结果与实测结果很好吻合,其最大偏差为2.85%,最小偏差为0.02%,平均偏差为0.70%。结论 基于近红外光谱技术与人工神经网络算法所建立的佐匹克隆片定量分析模型快速、准确、绿色环保,可以用于佐匹克隆片的快速检验。     

近红外漫反射光谱法结合CP-ANN和PLS高通量分析草麻黄药材

目的:建立草麻黄药材的近红外漫反射光谱高通量分析方法。方法:测量草麻黄样品的近红外漫反射光谱(near infra-red diffuse reflectance spectra,NIRDRS),应用化学计量学技术进行光谱处理和数据预处理,分别建立并验证草麻黄药材的产地和采摘时间判别对向传播人工神经网络(counter-propagation artificial neural network,CP-ANN)模型及麻黄碱和伪麻黄碱含量预测偏最小二乘(partial least square,PLS)模型。结果:草麻黄药材的产地和采摘时间判别CP-ANN模型的验证样品预测准确率分别为100.0%和80.0%;麻黄碱和伪麻黄碱含量预测PLS模型的验证样品预测均方根误差(root mean square errors ofprediction,RMSEPs)小,分别为1.12和0.236,预测值与参考值的相关系数(correlation coefficients)大,分别为0.9721和0.9309。结论:采用所建方法能同时对草麻黄药材的产地和采摘时间进行准确判别,对其麻黄碱和伪麻黄碱的含量进行准确预测。该方法准确、快速,无需特殊的样品处理,也不使用化学试剂。     

A review of near infrared spectroscopy and chemometrics in pharmaceutical technologies

Near-infrared spectroscopy (NIRS) is a fast and non-destructive analytical method. Associated with chemometrics, it becomes a powerful tool for the pharmaceutical industry. Indeed, NIRS is suitable for analysis of solid, liquid and biotechnological pharmaceutical forms. Moreover, NIRS can be implemented during pharmaceutical development, in production for process monitoring or in quality control laboratories.This review focuses on chemometric techniques and pharmaceutical NIRS applications. The following topics are covered: qualitative analyses, quantitative methods and on-line applications. Theoretical and practical aspects are described with pharmaceutical examples of NIRS applications.