Process options to minimise the environmental impact and improve the efficiency of biodiesel production have been investigated. The process options considered include the use of heterogeneous catalysts and used cooking oil (UCO). An esterification pre-treatment reaction was investigated using an ion-exchange resin (Purolite D5082) and an immobilised enzyme (Novozyme 435). Another immobilised enzyme (Amano Lipase PS-IM) was investigated for transesterification. The fresh and used catalysts have been characterised. The catalytic activity of Purolite D5082, Novozyme 435 and Amano Lipase PS-IM have been investigated using a jacketed batch reactor with a reflux condenser. Purolite D5082 has been developed for the esterification pre-treatment process and is not commercially available. Novozyme 435 has been shown to be an effective esterification catalyst for materials with high concentrations of free fatty acid but it has not been investigated for the esterification pre-treatment reaction. It was found that a high conversion was possible with both catalysts. The optimum reaction conditions identified for Purolite D5081 were a temperature of 60 C, a methanol to free fatty acid (FFA) mole ratio of 62:1, a catalyst loading of 5 wt% resulting in a FFAs conversion of 88% after 8 h of reaction time. The optimum conditions identified for Novozyme 435 were a temperature of 50 C, a methanol to FFA mole ratio of 6.2:1 and a catalyst loading of 1 wt% resulting in a conversion of 90% after 8 h of reaction time. These catalysts were compared to previously investigated Purolite D5081 and it was found that the highest conversion of 97% was achieved using Purolite D5081, however there were benefits to using Novozyme 435 because the reaction could be carried out using a much lower mole ratio, at a lower temperature and in much shorter reaction time. During the Novozyme 435 catalysed esterification pre-treatment reactions it was found that the amount of free fatty acid methyl esters (FAME) formed during the reaction was greater than the amount of FFAs consumed. In order to investigate further an ultra-performance liquid chromatography mass spectrometry (UPLC-MS) method was developed to monitor the monogclyeride (MG), diglyceride (DG) and triglyceride (TG) concentrations. This analytical method was used to show that Novozyme 435 would catalyse the esterification of FFAs as well as the transesterification of MGs and DGs typically found in UCO. With the UPLC-MS method it was possible to separate the 1, 2 and 1, 3 DG positional isomers and from this it could be seen that the 1, 3 isomer reacted more readily than the 1, 2 isomer. The results from the UPLC-MS method were combined with a kinetic model to investigate the reaction mechanism. The kinetic model indicated that the reaction progressed with the sequential hydrolysis esterification reactions in parallel with transesterification. Commercially available Amano Lipase PS-IM was investigated for the transesterification reaction. Enzymes are not affected by FFAs and as a result the optimisation was carried out with UCO as the raw material. An optimisation study for the transesterification of UCO with Amano Lipase PS-IM has not previously been reported. The conditions identified for the Amano Lipase PS-IM catalysed transesterification step are addition of 5 vol% water, a temperature of 30 C, a methanol to UCO mole ratio of 3:1 and a catalyst loading of 0.789 wt% resulting in a TG conversion of 43%. An overall enzyme catalysed process was proposed consisting of Amano Lipase PS-IM catalysed transesterification (stage 1) followed by Novozyme 435 catalysed esterification (stage 2). The previously identified optimum conditions identified for each catalyst were used for above stages. It was found that when the oil layer from stage 1 was dried the final TG conversion was 55%.

Pelchem NF3 plant produces an ammonium acid fluoride waste stream. The material of construction for the piping and stirrer fabrication in the plant is Monel. As a predominantly nickel-copper alloy, with minute quantities of carbon, manganese, silicon, sulfur and iron, these may leach into process fluids involved. The two biggest constituents of Monel contaminate the ammonium acid fluoride waste stream. Despite being the lesser of the two in terms of the composition of the Monel, copper is higher in concentration than nickel in the waste stream: the solubility of copper (II) cation in ammonium fluoride is higher than that of nickel (II) cation. Additionally, the ammonium acid fluoride is stored in steel barrels because of the relatively high process temperature that preclude the use of polymeric drums. This results in the leaching of iron from the steel drum to the solution. Pelchem expressed an interest in a suitable method of purification of ammonium fluoride, with specific interest of removing nickel (II) cation, copper (II) cation as well as iron (II) cation. The constraints to consider when selecting the appropriate methods are operating costs as well as the capital costs, but the most important factor to consider is the effectiveness of the method in removing the contaminant. In this regard, cationic exchange resins are very suitable, and they are very practical for industrial applications. In its simplest form, ammonium fluoride solutions are prepared by bubbling ammonia gas through solutions of hydrofluoric acid. Quite a few interesting uses of ammonium fluoride are available, these include as a chemical modifier in lead analysis, synthesis of beta zeolites, etc. The most prominent use is as a technical grade etchant in the electronics industry. The main aim of this research was to investigate ion exchange as a method of removing contaminants from Pelchem ammonium acid fluoride. Static equilibrium/selectivity experiments reveal that Purolite S930 Plus and Lewatit TP207 show a great affinity for the copper cation. For the limiting step of the reaction, the analysis includes apparent kinetics modelling contrasted with mass transfer modelling. In the case of reaction kinetics, Arrhenius and Van’t hoff equations were used to determine reaction parameters: the activation energies are 14 368 J∙mol-1 and 24 116 J∙mol-1, for Purolite and Lewatit respectively. The pre-exponential constants are 2 213 and 269 682 L2∙min-1∙mol-2 for Purolite and Lewatit in that order. The heats of reaction are -26 555 and -4 696 J∙mol-1 for Purolite and Lewatit respectively. Whilst the equilibrium pre-exponential constants are 75 057and 150 for Purolite and Lewatit respectively. Diffusivities for the two resins were found to be in reasonable agreement with those recorded in literature. They follow a temperature dependency trajectory. Weisz-Prater analysis of the observed reaction rate and the diffusion rate, in the two resins, reveals that intraparticle diffusion is the limiting step in the reaction. / Dissertation (MEng)--University of Pretoria, 2017. / Chemical Engineering / MEng / Unrestricted