<p>The potential use of enzymes in industrial synthesis of epoxidized soybean oil has been limited through the high cost of the enzyme catalyst, in this work we evaluate the effectiveness of chemo enzymatic epoxidation of high oleic soybean oil (HOSBO) using lipase B from <i>Candida antarctica </i>(CALB) on immobilization support Immobead 150 and H₂O₂in a solvent-free system. Additionally, we evaluated the production decay rates for hydrolytic activity and epoxide product formation over consecutive batches to determine half-life of the enzyme catalyst. </p> <p> Batch epoxidation of HOSBO using CALB on 4wt% loading shows yields higher than 90% after 12 hrs. of reaction, and with a correlation to the consumption of double bonds suggesting that the reaction is selective and limiting side product reactions. Non-selective hydrolysis of oil was not found beyond the initial hydrolysis degree of raw HOSBO. Evaluations of decay given by epoxide product formation and released free fatty acids shows a half-life of the enzyme catalyst on these activities is of 22 ad 25 hrs. respectively. Finally, we evaluated the physical parameters influencing this decay, and found that H₂O₂ presence is the most important parameter of enzyme inactivation with no significant effect from its slowed addition. We propose a new reactor configuration for the analysis of the specific steps on epoxide formation through peracid intermediates. </p>

No description available.

<p dir="ltr">With the global push for sustainable energy solutions, microalgae have gained attention as a promising biofuel source, with new methods like virus-assisted lipid extraction making algae-based biofuels more cost-effective. This thesis investigated the interaction between <i>Chlorella</i> algae and <i>Paramecium bursaria chlorella virus</i>-1 (PBCV-1) using variations of the SID (Susceptible-Infected-Dead) model to improve algae biofuel production. By leveraging epidemiological modeling, the thesis aimed to understand viral infection dynamics in algae and optimize the viral infection for efficient and effective lipid extraction for biofuels. Two distinct experimental setups were conducted to evaluate how environmental conditions impact infection outcomes. Mathematical models, including SID and SEID (Susceptible-Exposed-Infected-Dead) versions, were calibrated using Bayesian inference and nonlinear optimization to estimate critical parameters such as infection and mortality rates. Key findings reveal that the Susceptible-Infected-Dead model with variable initial infection (SIDVII), which included the initial concentration of infected algae as a model parameter, offered the most reliable fit by balancing biological relevance and statistical performance. In contrast, the SEID-based models, while able to capture a latent infection phase, posed risks of overfitting with too many parameters. The study emphasized the complexity of applying modeling tools to characterize viral infections at scale, suggesting that future research should address potential trade-offs between lipid productivity and infection dynamics. The findings offer potential insights not only for biofuel production but also for broader ecological management strategies, including the targeted control of harmful algal blooms.</p>