Chemical and Biomolecular Engineering, Department of
Date of this Version
1-26-2023
Document Type
Article
Citation
Digital Chemical Engineering 7 (2023) 100089. https://doi.org/10.1016/j.dche.2023.100089
Abstract
Acrylate-based polymers are commonly used in the chemical industry. Consistent manufacturing of these polymers with the help of Process Analytical Technology (PAT) is very desirable. The capability of monitoring polymers’ molecular weight in real-time reduces operation time and eliminates the frequent samplings needed for quality control. Herein, molecular weight (Mw) of glycidyl methacrylate-co-methyl methacrylate (GMA-co- MMA) copolymer was monitored in real-time using mid-infrared ATR-FTIR spectroscopy. The Principal Component Analysis (PCA) and Partial Least Square (PLS) models were then utilized to examine, improve the latent space, and select high-quality spectra. We show that acquiring highly correlated spectra enhances the robustness of the regression model. The developed models demonstrate a satisfied R2 correlation up to ~87%. This study suggests the potential of robust data-driven techniques for the development of predictive Mw analytics tools.
Included in
Biochemical and Biomolecular Engineering Commons, Biomedical Engineering and Bioengineering Commons
Comments
Open access.