Spatially and angularly resolved spectroscopy for in situ estimation of particle size and concentration in particle suspensions

Tools

Ferreira, Carla and Cardona, Javier and Agimelen, Okpeafoh and Tachtatzis, Christos and Andonovic, Ivan and Sefcik, Jan and Chen, Yi-Chieh (2026) Spatially and angularly resolved spectroscopy for in situ estimation of particle size and concentration in particle suspensions. Analytical and Bioanalytical Chemistry, 418 (21). pp. 7003-7017. ISSN 1618-2642 (https://doi.org/10.1007/s00216-026-06696-8)

[thumbnail of Ferreira-etal-ABC-2026-Spatially-and-angularly-resolved-spectroscopy-for-in-situ-estimation-of]
Preview
 Text. Filename: Ferreira-etal-ABC-2026-Spatially-and-angularly-resolved-spectroscopy-for-in-situ-estimation-of.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (3MB)| Preview

Abstract

Accurate in-line monitoring of particle size and solid concentration in turbid, multiphase systems remains a significant challenge in pharmaceutical manufacturing. Most particle measurement techniques are developed for off-line analysis, while recent in-line methods mainly rely on imaging and chord length distribution (CLD) measurement. Complementing these approaches, the spatially and angularly resolved diffuse reflectance measurement (SAR-DRM) system serves as a process analytical technology analyzer, capturing detailed, configuration-dependent spectral responses of particle suspensions. This study systematically evaluates SAR-DRM’s effectiveness in analyzing a wide range of particle diameters (≤90–800 µm) and concentrations (1–10 wt%) for polystyrene suspensions. The visible-NIR and NIR measurements from its multiple spatial-angular fiber configurations are further analyzed to assess their potential in complementing other in-line techniques, improving interpretability and analytical performance. The results show that combining specific SAR-DRM configurations through data augmentation markedly enhances model performance compared to individual configurations or co-adding methods, reducing error by over 50% in some cases. Moreover, data fusion of SAR-DRM with CLD measurements yields the most accurate models, particularly for mid-sized particles, decreasing the prediction error to 10–12 µm. The study highlights SAR-DRM as a promising process analytical technology analyzer for particulate process monitoring, paving the way for advanced hybrid modelling in pharmaceutical and chemical manufacturing. It also underscores the complementary sensing capabilities of SAR-DRM and CLD measurements, providing a robust multisensory platform for real-time, quantitative analysis in complex, high-turbidity systems.

ORCID iDs

Ferreira, Carla ORCID logoORCID: https://orcid.org/0000-0002-0592-8540, Cardona, Javier ORCID logoORCID: https://orcid.org/0000-0002-9284-1899, Agimelen, Okpeafoh ORCID logoORCID: https://orcid.org/0000-0002-0844-965X, Tachtatzis, Christos ORCID logoORCID: https://orcid.org/0000-0001-9150-6805, Andonovic, Ivan ORCID logoORCID: https://orcid.org/0000-0001-9093-5245, Sefcik, Jan ORCID logoORCID: https://orcid.org/0000-0002-7181-5122 and Chen, Yi-Chieh ORCID logoORCID: https://orcid.org/0000-0002-8307-0666;
CORE (COnnecting REpositories)