At-line near-infrared and Raman spectroscopy methods for determining the thermal decomposition of sodium hydrogen carbonate in a fluidized bed process

• Published in: Journal of pharmaceutical and biomedical analysis Vol. 219; p. 114918
• Main Authors: Frenkel, Kevin, Opel, Christian, Walter, Reinhard, Imming, Peter
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 20.09.2022
• Subjects: Calcination; Multivariate calibration; Near-infrared spectroscopy; Process Analytical Technology (PAT); Raman spectroscopy; Sodium hydrogen carbonate decomposition quantification; Calcination; Process Analytical Technology (PAT); Raman spectroscopy; Sodium hydrogen carbonate decomposition quantification; Multivariate calibration; Near-infrared spectroscopy
• ISSN: 0731-7085; 1873-264X
• DOI: 10.1016/j.jpba.2022.114918

Near-infrared (NIR) and Raman spectroscopy are analytical methods which are used increasingly for qualitative and quantitative at- or in-line measurements in the pharmaceutical industry. With the published quality guidelines of the Food and Drug Administration (FDA), these technologies are becoming popular, especially for ensuring high product quality and process monitoring. After a development of multivariate methods in this study, the best models were selected based on various quality parameters and the applicability of these in the routine process was presented. Calcination is a partial conversion to sodium carbonate, which increases compressibility and stability of effervescent tablets. Fluid bed granulation and drying oven processing were available as calcination technologies in this study. To determine the decomposition level spectroscopically, an analytical method was required. For this purpose, two standard methods (hydrochloric acid titration and a thermogravimetric analysis method) were compared. Thermogravimetric analysis was chosen due to a better determination of separately weighed binary mixtures of sodium hydrogen carbonate and sodium carbonate. Changes of water content, powder density and particle size showed influences on spectra and chemometric models. This impact was consistently avoided or reduced by reproducible sample handling and by using pre-processing operations. With a well understood standard method available, the main part of this study deals with diverse model developments of NIR and Raman data. The Raman technology was found to be superior to the NIR to determine the decomposition level of sodium hydrogen carbonate. The NIR demonstrated a low robustness and routine capability, higher number of factors, and poorer Root Mean Square Errors. [Display omitted] •Thermogravimetric method for determining a sodium hydrogen carbonate decomposition.•Elucidation of factors influencing near-infrared and Raman method development.•Use of PLS regression to build near-infrared and Raman multivariate models.•Evaluation and comparison of the models considering different quality attributes.