Date of this Version

8-4-2023

Keywords

Pharmaceutical Materials, Impedance Spectroscopy, Moisture Content, 1DCNN, Electrical Properties

Abstract

Pharmaceutical production requires strict moisture content control in order to ensure quality, stability, and efficacy of medicinal products. Moisture content can be manually tested but a rapid and reliable solution for real-time moisture content monitoring was necessary to improve overall efficiency in pharmaceutical manufacturing. This research delved into the practicality of using electrochemical impedance spectroscopy (EIS) as a dependable technique to determine moisture content. By investigating the electrical properties of materials, we established a robust connection between shifts in electrical properties with variations in moisture levels. We employed an equivalent circuit model to unveil the underlying mechanism which provided valuable insight into the sensitivity of impedance spectroscopy to changes in moisture content. Furthermore, this study incorporates AI techniques, employing a 1D Convolutional Neural Network (1DCNN) model to effectively process the complex spectroscopy data. The proposed model integrates frequency spectrum information, equivalent circuit indices, and material characteristics to estimate moisture content with an average predictive accuracy of 1.06%. Our work represents a significant advancement in the field, promising to enhance process control, ensure product quality, and drive overall efficiency improvements within pharmaceutical manufacturing.

Included in

Engineering Commons

Share

COinS