Liquid-core capsules as a tool in biotransformations

The aim of this thesis was to develop a new type of liquid-core capsules obtained by co-extrusion using the laminar jet break-up technique. These capsules, composed of an organic phase as the core and a hydrogel membrane, were meant to be applied as an in-situ product recovery technique for hydrophobic products and therefore required a fast mass transfer rate. Furthermore capsules had to be highly resistant to chelating agents, thermal sterilization process and mechanical stirring. The first part of the thesis was devoted to improvement of the mechanical stability of calcium alginate liquid-core capsules. A covalent structure could be added to the ionic calcium alginate membrane of the capsule. Acrylamide, methylene-bis-acrylamide and tert-butyl hydroperoxide were mixed with the sodium alginate solution and co-extruded with dibutyl sebacate into a gelling bath composed of 8% calcium alginate and sodium pyrosulfite. Capsules with a cross-linker versus monomer ratio of 5% were found to have the highest elasticity and bursting force after treatment with a citrate solution. Capsules were shown to be stable in the range of pH 4 to 9 and resistant to a sterilization process at 121°C. The first application of these reinforced capsules was their use as an in situ product recovery technique for pesticide removal from water. Atrazine, ethyl parathion, methyl parathion and 2,4 D were chosen as examples of pesticides and their mass transfer towards the capsules under different conditions was studied. This revealed that the main resistance was found in the stagnant organic phase within the liquid-core capsules, rather than through the bulk (well mixed) fluid or capsule membrane. In order to recycle the capsules, extracted pesticide was biodegraded by a specific bacteria. The experiments were performed in shake flasks, where the capsules filled with atrazine were suspended in a medium free of nitrogen and with citric acid as the carbon source. Pseudomonas sp. strain ADP was found to degrade atrazine as sole nitrogen source. By using the liquid-core capsules as a reservoir for the nitrogen source, high concentrations of pesticide could be degraded in the absence of inhibition as opposed to when pure aqueous systems were used. In addition, the results have shown that the immobilization of the organic phase could protect the cells from phase toxicity caused by dibutyl sebacate. The second application of liquid-core capsules was their use as reactive perstraction system for lipase catalyzed reactions. Candida rugosa lipase was immobilized by covalent attachment or by adsorption on the capsules membrane. Hydrolysis of tripropionin and nitrophenyl laurate could be performed with this immobilized lipase and the system could easily be recycled for repeated uses. In the case of nitrophenyl laurate hydrolysis, nitrophenol and lauric acid were the products of the reaction. Combined with the reaction, lauric acid was extracted directly by the organic liquid-core while nitrophenol was released into the aqueous phase thereby reducing inhibition of the enzymatic activity. The hydrolysis of penicillin G by penicillin acylase was the second enzyme catalyzed reaction chosen as a model to demonstrate the feasibility to use liquid-core capsules as a reactive perstraction system. The reaction was performed at pH 3.5, where phenyl acetic acid, one of the two products of the reaction, was partially extracted by the liquid-core capsules. The second product of the reaction, 6-aminopenicillanic acid, was released into the aqueous phase. The system, composed of immobilized penicillin acylase on liquid-core capsules, was successfully recycled up to five times by back extracting the phenyl acetic acid with an alkaline solution. However, the extraction efficiency of this system in stirred tank reactors was limited by the important water content of the liquid-core capsule membrane. Therefore the latter were packed in a column in order to obtain a chromatographic reactor. This set-up allowed the simultaneous reaction and separation of the products throughout the column. Parameters such as flow rate and column height were shown to greatly influence the product separation and reaction yield. Further improvement in this field could be by the use of a simulating moving bed chromatographic reactor