Pervaporation is a selective membrane technique in which a liquid feed mixture is separated by means of partial vaporization through a non-porous perm-selective membrane. This method can be used for the recovery of heat sensitive aroma compounds to avoid them from thermal damage in beverage industries. The main objective of this study was to determine the effects of feed temperature (30, 40, 50&deg / C), composition (different model solutions, strawberry essence), concentration (50, 100, 150 ppm) and permeate pressure (4, 8 mbar) on the recovery of aroma compounds of strawberry by pervaporation in terms of mass flux and selectivity. In addition, it was aimed to optimize the extraction conditions (extraction time, temperature, agitation speed, strawberry matrix) of Solid-phase microextraction (SPME), which is used for the analysis of strawberry aroma compounds. Optimum results for SPME were obtained at 40&deg / C, 700 rpm for 30 min and no matrix effect was observed. Pervaporation experiments were performed using a hydrophobic membrane, PERVAP 1070 (PDMS). As the feed temperature increased, the mass flux and selectivity increased and the total mass flux followed an Arrhenius type relation. Decreasing downstream pressure increased both total flux and selectivity, while increase in feed concentration led to higher organic fluxes but lower selectivities. In general, PERVAP 1070 showed a higher selectivity towards Methyl butyrate (MTB) than Ethyl butyrate (ETB) and MTB flux was affected negatively by the presence of ETB in the feed solution. Pervaporation experiments were also performed with a strawberry essence and strawberry model solution. The selectivities of MTB and ETB were negatively affected by the presence of other aroma compounds.

Orientadores: Maria Regina Wolf Maciel, Mario Eusebio Torres Alvarez / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Quimica / Made available in DSpace on 2018-08-11T16:17:30Z (GMT). No. of bitstreams: 1 Araujo_WilsonAndaleciode_D.pdf: 2993508 bytes, checksum: 29abb03a65a80f6c9835367e0cb49d50 (MD5) Previous issue date: 2008 / Resumo: Os processos de separação com membranas (PSM) têm sido considerados como alternativa a processos clássicos de separação. Esta é uma área de estudo que apresenta uma forte interdisciplinaridade. Há um crescente interesse nestes processos para diversas aplicações como, por exemplo, tratamento de efluentes industriais, desalinização de águas, purificação e concentração de correntes da indústria alimentícia. A separação, em geral, não envolve mudança de fase, o que significa economia no consumo de energia e operações a temperaturas moderadas. Na tecnologia de separação com membranas, os componentes das misturas líquidas ou gasosas são separados ao permearem de forma seletiva através de uma membrana. As membranas podem ser poliméricas ou cerâmicas. A corrente de alimentação é dividida em duas correntes de saída: a que permeou através da membrana (permeado) e a corrente concentrada retida (¿retentate¿). Estes processos têm sido aplicados no processamento de bebidas, sucos, e aromas. Neste trabalho, dois PSM foram estudados, a Osmose Inversa (OI) e a Pervaporação (PV). Experimentos em escala piloto foram realizados empregando-se o processo de OI (membrana de poliamida) para a concentração de suco de laranja a 20°Brix. Avaliou-se a retenção de compostos de voláteis (acetaldeído, metanol e etanol) monitorando-se as correntes de alimentação e permeado. Os resultados de retenção de aromas obtidos não foram satisfatórios. A membrana apresentou baixas retenções para os voláteis monitorados na temperatura usada para realização dos experimentos. Na segunda etapa do trabalho, o processo de PV foi avaliado para recuperação de um importante éster do suco de laranja, o etil butirato. O software PERVAP, um simulador Fortran essencialmente preditivo, foi empregado no estudo de desempenho do processo para duas membranas, polidimetilsiloxano (PDMS) e polioctilmetilsiloxano (POMS). Realizou-se a predição de propriedades de membranas poliméricas para incremento da capacidade preditiva do simulador. Foram empregados métodos de contribuição de grupos para predição das propriedades dos polímeros. Os dados de viscosidade preditos para o POMS viabilizaram a realização de cálculos para obtenção de parâmetros requeridos para operação do simulador. A abordagem proposta proporcionou maior versatilidade ao simulador / Abstract: The membrane separation processes (MSP) have been considered as alternative for conventional separation processes. In this research area a strong interdisciplinarity is observed. There is an increasing of interest for these processes considering many aplications (e.g., industrial wastewater treatment, water desalination, purification and concentration of food industry streams). The separation usually does not requires phase change, which means energy savings and moderate temperatures. A membrane separation system separates an inlet stream into two effluent streams known as the permeate and the retentate. The permeate is the portion of the fluid that has passed through the membrane. Whereas the retentate stream contains the constituents that have been rejected by the membrane. The membrane can be polymeric or ceramic. These processes have been applied for processing beverages, juices and aromas. In this work, two of these processes were studied, Reverse Osmosis (RO) and Pervaporation (PV). Pilot scale experiments were accomplished using RO (poliamide membrane) for concentrating single strength orange juice at 20ºBrix. The retention of volatile compounds (acetaldehyde, methanol and ethanol) was evaluated by monitoring feed and permeate streams. The retention results obtained were unsatisfactory. The membrane presented low retention for monitored volatiles under studied temperature conditions. In the second stage of this work, the PV process was evaluated for recovering an important ester of orange juice, the ethyl butyrate. The PERVAP software, an essentially predictive Fortran simulator, was used for evaluating process performance considering two membranes, polydimethylsiloxane (PDMS) and polyoctylmethylsiloxane (POMS). It was accomplished the prediction of properties for polymeric membranes targeting the software predictivity improvement. Viscosity data predicted for POMS was crucial for calculating parameters required by simulator. The predictive approach proposed improved the software versatility / Doutorado / Desenvolvimento de Processos Químicos / Doutor em Engenharia Química