Contributions to the selection of solvents in the extraction of phenolic acids

• Autor(a) principal: Alves, Rebeca S.
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10198/22925
• topic: Solubility; Phenolic compounds; NRTL-SAC model; Abraham solvation model; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Química

Naturally occurring phenolic acids are well-known and studied for their bioactive properties and wide distribution in plants, where they can be found in free form, or conjugated to other molecules. The study of the solubility of phenolic compounds in water and organic solvents is thus fundamental for the design of extraction, separation, crystallization and purification processes of great importance in the pharmaceutical, cosmetics and food industries. In this context, the main objective of this work is to measure the solubility of trans-cinnamic, p-coumaric and ferulic acids in water and in seven organic solvents (methanol, ethanol, 1-propanol, 2-propanol, 2-butanone, ethyl acetate and acetonitrile) at 298.2 K and 313.2 K and test the ability of the NRTL-SAC model, with or without the Reference Solvent Approach (RSA), and the Abraham solvation model to correlate, and preferably, predict the solubility data. To accomplish the objectives above the shake-flask method experimental method was combined with UV-Visible spectroscopy and gravimetric methods of analysis to perform the solubility measurements. In general, the results obtained were in close agreement with the very scarce information available in literature. After, the NRTL-SAC segment descriptors of each solute were fitted to solubility data in seven solvents, obtaining average relative errors (ARD) between 23% and 39%. The model was then applied to predict the solubility in other eight solvents, with ARD between 42% and 61%. The RSA was also applied, but no significant improvements were obtained relatively to the first approach. The optimization parameters of the Abraham solutes were also obtained by fitting the solubility data in six solvents for the trans-cinnamic and p-coumaric acids, and seven solvents for the ferulic acid, obtaining ARD between 7% and 24% for correlations and between 4% and 33% for the predictions in the remaining solvents. These values indicate Abraham's solvation model as the most suitable and very satisfactory model to predict the solubility of the selected solutes at 298.2 K.