How to Use IR Spectroscopy in Pharmaceutical Starting Material Test?

Published: Wednesday, 10 June 2015

The pharmaceutical guidelines from International Conference on Harmonization (ICH) describes the importance of drug identify tests. The pharmaceutical starting materials analysis and test methods mentioned in the pharmacopoeia, includes: Optical rotation, HPLC, and chromatography. In addition, infrared spectroscopy (IRS) is a kind of common analytical method, which are used in pharmaceutical starting materials identification test technique, especially in qualitative analysis. In the Chapter 2 of European Pharmacopoeia, you can find the introductions of infrared spectroscopy analytical technology.

The IRS methods are commonly used in pharmaceutical starting material incoming inspection, because of the relative easy sample preparation requirements among the other methods, or even without any treatment. There are many thesis related NIR Spectroscopy in the control of excipients, active pharmaceutical ingredient (API) and medicine final products.

The modern technologies in this field, are mainly use the NIR spectroscopy with the software of principal component analysis and clustering algorithms, in order to differentiate different types of materials, such as starches, sugars, celluloses, intermediates and active ingredients.

Identification of Microcrystalline Celluloses 

NIR Spectroscopy is often used to identify Avicel® microcrystalline celluloses products, such as PH-101, PH-102, and PH-200. NIR spectroscopy has significant advantage in statistical analysis. But for the separation of methylcellulose and cellulose ethers with methyl or hydroxyalkyl groups, NIR spectroscopy is not a good solution.

Classification of Polyvinyl Pyrrolidone

The current technology is able to classify various types of povidones by a NIR spectrometer along with Soft Independent Modelling of Class Analogies (SIMCA) in the incoming material inspection process. Those NIR spectrometers, which are designed for factory environment with rugged conditions, are the proper device for this purpose.

Quality Control of Pharmaceutical Intermediates 

The quaintly control of pharmaceutical intermediates can be executed by a NIR spectrometer with the comparison to wet chemical methods. For example, the quality control of 7-Aminocephalosporanic Acid.

Identify The Composition of Excipients

By using the Transflectance NIR Spectroscopy can distinguish the types of solvent effectively. In a correlation coefficient of collected discrete data, the optimum conditions for the clustering can be obtained by the second derivative over the wavelength from NIR spectrometer. There are several solvents and benzodiazepine (BZD),in which the correlation coefficient were public. By this method, there are more than one hundred of excipients can be identified by NIR spectrometer.

Control & Manage The Quality Conditions

In addition to the applications of the classification of raw material, the qualification of raw materials is also important. To monitor the raw material quality, which is within the normal variability range or is subject to over limit deviations, is also the advantage of a NIR spectrometer. By collecting the Big Data, the tradition random pick up inspection to be upgraded to machine full inspection, be possible.

Machine learning

The topics of Industrial 4.0 became popular recently, in addition to the investments of hardware facilities in the pharmaceutical companies, the core technologies in software and efficient processes are important. In addition to those hardware facilities that are also available for the competitors, these software and process management technologies are the unique core value to generate advantages (the soft power). Thus, the trends of some public domain distributed software, is worth to be monitored and look into their progress. For example, Scikit-learn, Shogun, are well known software modules for classification, clustering, forecast, in the machine learning technologies.

Scikit-learn Classification
Scikit-learn has several intelligent classification modules for usage
Scikit-learn Learns The Trands of Spectrum
Scikit-learn learns the trends from the spectrum sets and find a trend