A carregar...
Publicação
Optimizing blending processes: establishing and refining NIR and transmission raman methods
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.11 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Garantir a uniformidade da mistura de pós e grânulos farmacêuticos é crucial na obtenção
de produtos farmacêuticos, especialmente, formas farmacêuticas sólidas, com qualidade e de
acordo com os critérios regulatórios. A espetroscopia de infravermelho próximo e a espetroscopia
de Raman por transmissão têm surgido como métodos que permitem alcançar e determinar quando
a mistura está homogénea de forma não destrutiva, rápida e eficaz. Os principais objetivos deste
trabalho foram utilizar a espetroscopia de infravermelho próximo como ferramenta de
monitorização da mistura em tempo real e avaliar a sensibilidade desta técnica na deteção de
pequenas variações na formulação. Assim como usar a espetroscopia de Raman por transmissão
para quantificar a substância ativa em amostras e avaliar o seu potencial como uma técnica
alternativa de referência analítica para controlo de qualidade.
De modo a alcançar estes objetivos foram realizados estudos com duas formulações
diferentes. Primeiro desenvolveu-se e validou-se um modelo de Mínimos Quadrados parciais
usando a cafeína como substância modelo e posteriormente aplicou-se esta mesma abordagem a
outra formulação utilizando uma outra substância ativa, cuja identidade é confidencial.
A espetroscopia de infravermelho próximo mostrou conseguir identificar pequenas
variações até 0.5% nas formulações de cafeína. Os modelos de Mínimos quadrados parciais
desenvolvidos foram robustos, permitiram que mesmo com alterações ao tamanho do lote se
mantivesse uma boa capacidade preditiva para determinar a uniformidade da mistura. Embora
sejam de realçar, estatisticamente, diferenças significativas entre as previsões e os resultados de
HPLC/UPLC em grandes intervalos de concentração.
Quanto à espetroscopia de Raman por transmissão, permitiu fazer boas previsões na
quantificação da substância ativa em ambas as formulações estando de acordo com expetativa
quando comparado com os resultados de HPLC/UPLC. De forma geral, as previsões obtidas
através do modelo de Mínimos quadrados parciais para o Raman de transmissão mostraram não
ser estatisticamente diferentes dos resultados de HPLC/UPLC, garantindo que este é o método que
mais se aproxima destas técnicas de referência analíticas e a partir do qual de obtêm resultados
mais exatos e fidedignos.
Ensuring blend uniformity of pharmaceutical powders and granules is crucial in obtaining pharmaceutical products, especially solid pharmaceutical forms, with quality and in accordance with regulatory criteria. Near-Infrared Spectroscopy and Transmission Raman Spectroscopy have emerged as methods that enable the determination of when the mixture is homogeneous in a non destructive, fast and effective way. The main objectives of this work were to use Near-Infrared Spectroscopy as a tool for monitoring the mixture in real time and to evaluate the sensitivity of this technique in detecting small variations in the formulation. Also to use Transmission Raman Spectroscopy to quantify API in samples and evaluate its potential as an alternative analytical reference technique for quality control. In order to achieve these objectives investigation with two different formulations were carried out. First a Partial Least Squares model was developed and validated using caffeine as the model substance and then this same approach was applied to another formulation using an API from the development pipeline. Near-Infrared Spectroscopy was able to identify small variations down to 0.5% in the caffeine formulations. The Partial Least Squares models developed were robust and allowed changes to be made to the batch size while maintaining a good prediction capability for blend uniformity. However, there are statistically significant differences between the predictions and the HPLC/UPLC results over wide concentration ranges. As for Transmission Raman Spectroscopy, it made good predictions for API quantification in both formulations and was in line with expectations when compared with the HPLC/UPLC results. Transmission Raman Spectroscopy predictions showed no significant difference to HPLC/UPLC results, proving this method to give closer results to the latter reference analytics techniques and to be a more accurate and reliable method than Near-Infrared Spectroscopy.
Ensuring blend uniformity of pharmaceutical powders and granules is crucial in obtaining pharmaceutical products, especially solid pharmaceutical forms, with quality and in accordance with regulatory criteria. Near-Infrared Spectroscopy and Transmission Raman Spectroscopy have emerged as methods that enable the determination of when the mixture is homogeneous in a non destructive, fast and effective way. The main objectives of this work were to use Near-Infrared Spectroscopy as a tool for monitoring the mixture in real time and to evaluate the sensitivity of this technique in detecting small variations in the formulation. Also to use Transmission Raman Spectroscopy to quantify API in samples and evaluate its potential as an alternative analytical reference technique for quality control. In order to achieve these objectives investigation with two different formulations were carried out. First a Partial Least Squares model was developed and validated using caffeine as the model substance and then this same approach was applied to another formulation using an API from the development pipeline. Near-Infrared Spectroscopy was able to identify small variations down to 0.5% in the caffeine formulations. The Partial Least Squares models developed were robust and allowed changes to be made to the batch size while maintaining a good prediction capability for blend uniformity. However, there are statistically significant differences between the predictions and the HPLC/UPLC results over wide concentration ranges. As for Transmission Raman Spectroscopy, it made good predictions for API quantification in both formulations and was in line with expectations when compared with the HPLC/UPLC results. Transmission Raman Spectroscopy predictions showed no significant difference to HPLC/UPLC results, proving this method to give closer results to the latter reference analytics techniques and to be a more accurate and reliable method than Near-Infrared Spectroscopy.
Descrição
Tese de mestrado, Engenharia Farmacêutica, 2024, Universidade de Lisboa, Faculdade de Farmácia.
Palavras-chave
Near-infrared spectroscopy Transmission raman spectroscopy Blend uniformity Bulk powder Teses de mestrado - 2024