Comparison of Different Chemometric and Analytical Methods for the Prediction of Particle Size Distribution in Pharmaceutical Powders

• Tipo: Artigo em Revista Científica Internacional
• Data: 2011

Autores

Mafalda C. Sarraguça (Autor) FFUP; Ana V. Cruz (Autor) FFUP; Helena R. Amaral (Autor) FFUP; Paulo C. Costa (Autor) FFUP; João A. Lopes (Autor) FFUP

Revista

Título: Analytical and Bioanalytical Chemistry

Vol. 399 Nº 6

Páginas: 2137-2147

ISSN: 1618-2642

Editora: Springer Nature

Indexação

ISI Web of Knowledge - 12 Citações

ISI Web of Science

Scopus - 14 Citações

Pubmed / Medline

Classificação Científica

FOS: Ciências médicas e da saúde > Ciências da saúde

CORDIS: Ciências da Saúde > Ciências farmacológicas

Outras Informações

ID Authenticus: P-002-VAX

Resumo (PT): This work compares the estimation of the particle size distribution of a pharmaceutical powder using near-infrared spectroscopy (NIRS), powder flowability properties, and components concentration. The estimations were made by considering the former data blocks separately and together using a multi-block approach. The powders were based on a formulation of paracetamol as the pharmaceutical active ingredient. The reference method used to determine particle size distribution was sieving. Partial least squares methods were used to estimate the multivariate regression models, and the results were compared in terms of figures of merit. It was shown that the partial least squares methods gave similar prediction errors. Regarding the data blocks used, the NIRS block was proven the most advantageous to estimate the particle size distribution. The prediction error of the NIRS block was similar to the other data blocks with additional advantages such as less generalization problems and the possibility of its use to predict additional physical and chemical properties with an improvement to analysis time. The multi-block approach produced the worst results but nevertheless allowed a deeper understanding of the individual contributions of the data blocks in the prediction of the particle size distribution. <br> <br> Keywords Pharmaceutical powders – Particle size distribution – Near-infrared spectroscopy – Flowability properties – Partial least squares 1 – Partial least squares 2 – Multi-block partial least squares <br> <a target="_blank" href="http://www.springerlink.com/content/5334n8841274j333/ "> Texto integral </a> <br> <br>

Abstract (EN): This work compares the estimation of the particle size distribution of a pharmaceutical powder using near-infrared spectroscopy (NIRS), powder flowability properties, and components concentration. The estimations were made by considering the former data blocks separately and together using a multi-block approach. The powders were based on a formulation of paracetamol as the pharmaceutical active ingredient. The reference method used to determine particle size distribution was sieving. Partial least squares methods were used to estimate the multivariate regression models, and the results were compared in terms of figures of merit. It was shown that the partial least squares methods gave similar prediction errors. Regarding the data blocks used, the NIRS block was proven the most advantageous to estimate the particle size distribution. The prediction error of the NIRS block was similar to the other data blocks with additional advantages such as less generalization problems and the possibility of its use to predict additional physical and chemical properties with an improvement to analysis time. The multi-block approach produced the worst results but nevertheless allowed a deeper understanding of the individual contributions of the data blocks in the prediction of the particle size distribution. <br> <br> Keywords Pharmaceutical powders – Particle size distribution – Near-infrared spectroscopy – Flowability properties – Partial least squares 1 – Partial least squares 2 – Multi-block partial least squares <br> <a target="_blank" href="http://www.springerlink.com/content/5334n8841274j333/ "> Full text</a> <br> <br>

Idioma: Inglês

Tipo (Avaliação Docente): Científica