Supercritical fluid fractionation process development : a techno-economic evaluation of retrofitting an aroma production facility

Abstract

An aroma–rich, aqueous condensate is produced as a by-product from the triple-effect evaporation of apple juice during the production of sugar from the fruit juice. At about 1% content of organic compounds, an opportunity exists to add value to this product by concentrating the aroma compounds and reducing the bulk of the water content. In doing so, the non-aroma imparting organic compounds contained may also be selectively removed, and thus increasing the value of the product. Besides its application as food and beverage, a concentrated aroma can find use is perfumery and other scented products. This work investigated the feasibility of further concentrating an aroma-rich, but highly dilute aqueous condensate by-product using supercritical CO2 (sCO2) as the selective solvent. The compounds of interest were hexenal and trans-2-hexenal (representing the desired, aroma compounds), and hexanol (an organic compound that does not contribute to the desired aroma). Initially, a theoretical approach to the problem was taken to investigate the possibility of enrichment of the organic compounds from their aqueous mixture, as well as separation of the organic compounds. This was achieved using a series of equilibrium flash calculations representing a counter-current column. Experimental vapour-liquid equilibria data of binary solute-CO2 for each of the compounds, obtained from the literature, was used to develop a model to represent the phase behaviour of the system. The SRKKD thermodynamic model in Aspen Plus® was shown to adequately represent the system behaviour. An outcome representing the possibility of concentration, but the unlikelihood of separation of the hexanol from the aroma compounds was obtained at this stage. This outcome led to further investigation at pilot plant, experimental level, using a synthetic mixture of the main components of a typical condensate from a fruit-sugar factory. The sample and the dense gas were fed counter-currently through the column, and the concentrated extract separated from the solvent in a series of separators. Pilot plant experiments were conducted at 40 & 50oC and 70 & 100 bar at a solvent to feed ratio of 5. Flow and composition data was obtained from the series of experiments. The data was used to develop a process model, using the commercial simulator Aspen Plus®. The model parameters were adjusted to fit the flow and composition experimental data. Good prediction of the process performance was obtained from model. The model was thus used to predict the performance of the process at conditions other than those investigated experimentally. Different process lay-out scenarios were investigated. In this way, it was possible to predict the conditions required for optimum operation of the process. The results showed that, in agreement with the theoretical study, the concentration of organic components in the feed was relatively easily achievable. Somewhat unexpectedly, the results showed that hexanol was removed to a greater extent than had been predicted by the model. Optimum conditions were selected at 40oC, 70 bar and S/F ratio of 5. Energy requirements for the complete process, according to the model was 348 MJ/kg product. Both the product quality and yield indicate that the process holds potential for further development, possibly towards achieving a commercial process.