With the increasing demand for biodiesel, the accumulation of byproduct, crude glycerol has become a problem which needs to be solved. Lactic acid is one of the value-added chemical which can be produced from glycerol that has wide uses in food and chemical industry. Although glycerol can be converted to lactic acid with an alkali as the catalyst at high glycerol conversion (100 mol%) and lactic acid yield (around 90 mol%), the high alkalinity would cause severe corrosiveness to a stainless steel reactor. In this study two tasks were performed to convert glycerol to lactic acid with satisfactory conversion and selectivity, and to reduce the corrosiveness of reaction medium. First, CaO was used as solid base catalyst. The highest lactic acid yield achieved was 40.8 mol% with a 97.8 mol% glycerol conversion, when operating at 290°C after 150 min reaction with molar ratio of CaO: glycerol=0.3. Also CaO has advantages such as high lactic acid productivity (3.35 g/(min·L)) and reusability. Meanwhile, CaO can be used as the catalyst for both biodiesel production and the following crude glycerol conversion to lactic acid. Second, for glycerol conversion with NaOH as catalyst, a fed-batch reactor was applied to continuously supply NaOH during reaction process, compensating the OH- neutralized by newly formed lactic acid. The optimal lactic acid yield of 80.5 mol%, with 92.8% glycerol conversion was obtained at 300 °C for 220min, with 1.1 M glycerol initial concentration. A first-order kinetic model for glycerol concentration versus time was developed and verified experimentally under conditions with different initial glycerol concentration and reaction temperature. Although crude glycerol samples contained large amount of impurities, both methods, conversion with solid base catalyst and with fed-batch reactor, were applied successfully to three crude glycerol samples provided by biodiesel manufacturers, and the lactic acid yield reached 52.3 mol% and 72.7 mol% respectively. Finally, the corrosion issue of different methods was compared based on the Fe3+ concentration (analyzed with atomic-absorption spectroscopy) in the products. Both methods of glycerol conversions, with solid base catalyst and fed-batch reactor, can reduce the corrosiveness of glycerol conversion with an alkali as the catalyst.

Rhizopus oryzae is a prominent strain for producing fumarate, where biomass growth precedes fumarate production. The natural bio film growth of R. oryzae as fungal mat was investigated using different glucose addition strategies in a novel fed-batch fermenter. Batch growth was compared through three fed-batch runs, each with a different glucose addition strategy. The fed-batch runs involved a constant glucose feed (CGF) of 0.075 g h-1 and controlled glucose feeds in order to keep the respiration quotient (RQ) at either 1.3 mol CO2 mol-1 O2 (RQ1.3) or 1.1 mol CO2 mol-1 O2 (RQ1.1). Ethanol overflow via the established Crabtree mechanism was completely negated for the CGF and RQ1.1 runs, while the batch and RQ1.3 runs exhibited significant ethanol formation. Biomass yield on glucose was found to be 0.476 g g-1 (RQ1.1), 0.194 g g-1 (RQ1.3), 0.125 g g-1 (CGF) and 0.144 g g-1 (batch). The results indicate a three-fold improvement in biomass yield when comparing the batch run with the RQ1.1 run. In addition, the RQ1.1 run resulted in zero detectable byproducts, unlike the batch scenario where pyruvate and fumarate were associated with ethanol formation. Clear evidence is provided that glucose overflow can be fully eliminated during R. oryzae growth, significantly affecting the biomass yield on glucose. / Dissertation (MEng (Chemical Engineering))--University of Pretoria, 2021. / University of Pretoria postgraduate bursary / CSIR Inter-bursary Programme / Chemical Engineering / MEng (Chemical Engineering) / Unrestricted

Researches are done with the aim of select yeast strains that plays a differentiated role in the process of alcoholic fermentation, with the purpose of improve the performance in the ethanol production and decrease the productions costs. In this work it was studied the application of Saccharomyces cerevisiae yeasts with flocculent features in fed batch reactor. It was evaluated the fermentative capacity of six strains of flocculent yeasts. The yeast C2/00 was involved in higher productivity and yield in ethanol compared to other yeasts tested. After it was studied the alcoholic fermentation in fed batch reactor using the yeast C2/00. The fermentations were performed at 32°C and initial pH adjusted in 4.5. The process was optimized in sucrose concentration of 170 g/L, cell concentration in the inoculum of 40% (v/v) and filling time of 6 hours, it is obtaining a yield of 92.20% in relation to the theoretical one, productivity of 6.01 g/L.h and residual sucrose of 42.84 g/L in 10.5 hours of fermentative process. It was studied the influence of cells recirculation during the fermentative process and the influence of initial concentration of ethanol and substrate in inoculum on the fed batch process with the aim of improve the productivity and reduce the residual sugar. From of this study it was obtained 92.75% of yield, 9.26 g/L.h of productivity, 2.9 g/L of residual sucrose concentration and the ethanol concentration produced was 83.37 g/L in 9 hours of fermentative process. The inhibition by the substrate and product model to the kinetics of alcoholic fermentation was proposed. The parameters of model were calculated by means of nonlinear adjust to the experimental results of growth of yeast, substrate consumption and formation of product to the batch reactor. The maximum specific speed of growth was 0.103 h-1 with KI and Ks equal to 109.86 and 30.24 g/L, respectively. With the experimental results of fed batch reactor and fed batch with recycle, it can be noted a good fit to the model proposed, resulting in a maximum specific velocity of growth of 0.080 h-1 to the process in fed batch without recycle and 0.182 h-1 to the fed batch process with recycle of fermentative media. The ethanol concentration in which the production of it is completely inhibited (P max) was 110 gethanol /L , approximately 13.92% (v/v). For the result of maximum concentration of product that inhibits fully the microorganisms growth (Pmax) was 12% (v/v), corresponding to 94.8 g/L of ethanol. The specific velocity of sedimentation (SVS) to the yeast C2/00 in pH 5 was 0.240 min-1 and the sedimentation rate of the test beaker was 0.444 cm/min. The alcoholic fermentation, using the flocculent yeast Saccharomyces cerevisiae in fed batch reactor with recycle of fermentative media provided higher productivity and yields when compared to the reported data by literature, in which used batch reactor or fed batch without recycle of fermentative media. / Com o objetivo de melhorar o desempenho na produção de etanol e diminuir os custos de produção, pesquisas são realizadas no intuito de selecionar linhagens de leveduras que vêm sendo diferenciadas no processo de fermentação alcoólica. Neste trabalho estudou-se a aplicação de leveduras Saccharomyces cerevisiae com características floculantes em reator batelada alimentada. Avaliou-se a capacidade fermentativa de seis cepas de leveduras floculantes. A levedura C2/00 foi a gerou maiores produtividade e rendimento em etanol em comparação às outras leveduras testadas. Posteriormente, estudou-se a fermentação alcoólica em processo batelada alimentada utilizando a levedura C2/00. As fermentações foram realizadas a 32°C e pH inicial ajustado em 4,5. O processo foi otimizado com concentração de sacarose de 170 g/L, concentração celular no inóculo de 40% (v/v) e tempo de enchimento de 6 horas, obtendo-se um rendimento de 92,20% em relação ao teórico, produtividade de 6,01 g/L.h e sacarose residual de 42,84 g/L em 10,5 horas de processo fermentativo. Com o intuito de melhorar a produtividade e reduzir o açúcar residual foi estudada a influência da recirculação de células durante o processo fermentativo e a influência da concentração inicial de etanol e substrato no inóculo no processo em batelada alimentada. A partir deste estudo obteve-se 92,75% de rendimento, 9,26 g/L.h de produtividade, 2,9 g/L de concentração de sacarose residual e a concentração de etanol produzido foi de 83,37 g/L em 9 horas de processo fermentativo. Foi proposto o modelo de inibição pelo substrato e produto para a cinética da fermentação alcoólica. Os parâmetros do modelo foram calculados por meio de ajuste não linear aos resultados experimentais de crescimento de leveduras, consumo de substrato e formação de produto para o reator batelada. A velocidade específica máxima de crescimento foi de 0,103 h-1 com KI e Ks iguais a 109,86 e 30,24 g/L, respectivamente. Com os resultados experimentais do reator batelada alimentada e batelada alimentada com reciclo, constatou um bom ajuste do modelo proposto, resultando em uma velocidade específica máxima de crescimento de 0,080 h-1 para o processo em batelada alimentada sem reciclo e 0,182 h-1 para o processo batelada alimentada com reciclo de meio fermentativo. A concentração máxima de etanol na qual a produção do mesmo foi completadamente inibida (P máx) foi de 110 getanol /L , aproximadamente 13,92% (v/v). No entanto a concentração máxima de produto que cessa totalmente o crescimento do micro-organismo (Pmáx) foi de 12% (v/v), correspondendo a 94,8 g/L de etanol. A velocidade específica de sedimentação (VES) para a levedura C2/00 em pH 5 foi de 0,240 min-1 e a velocidade de sedimentação pelo teste da proveta foi de 0,444 cm/min. A fermentação alcoólica, utilizando a levedura floculante Saccharomyces cerevisiae em reator batelada alimentada com reciclo de meio fermentativo forneceu maior produtividade e rendimentos quando comparados a dados reportados pela literatura, nos quais utilizaram reator batelada ou batelada alimentada sem reciclo de meio fermentativo. / Doutor em Engenharia Química