UV-VIS spectra as potential process analytical technology (PAT) for measuring the density of compressed materials: Evaluation of the CIELAB color space

Abstract

Measurement methods for determining the density of compressed materials, being a critical quality attribute, provide an important parameter for successful processing. In this study, a novel approach was developed for determining the density of compacts using ultraviolet-visible spectrophotometry. The assumption within this context was that a change in density affects the corresponding color information of the compact. From the obtained spectra of the visible range, the color information of the compact was calculated which turned out to be directly proportional to the density of the compact. In comparison, the obtained spectra were analyzed using partial least square regression. The results of this study showed that both methods could be used predicting the density of a compact from the corresponding visible spectrum at identical accuracy. In contrast to the partial least square regression, the correlation of the color information as a direct output parameter of the spectrophotometer with the density required no excessive data pre-processing. Subsequently, the easier and faster data processing of the color information over the partial least square regression, conceives using this novel approach as potential process analytical technology tool for implementation into a compression process e.g., tableting or roller compaction.

Keywords: Color space; Compression; Partial least square regression (PLS); Principal component analysis (PCA); Process analytical technology (PAT); Spectrophotometry; Tablet density.