Inverse design of chemoenzymatic epoxidation of soyabean oil through artificial intelligence-driven experimental approach.

This paper presents an inverse design methodology that utilizes artificial intelligence (AI)-driven experiments to optimize the chemoenzymatic epoxidation of soyabean oil using hydrogen peroxide and lipase (Novozym 435). First, experiments are conducted using a systematic 3-level, 5-factor Box-Behnken design to explore the effect of input parameters on oxirane oxygen content (OOC (%)). Based on these experiments, various AI models are trained, with the support vector regression (SVR) model being found to be the most accurate.

Reduced Order Machine Learning Models for Accurate Prediction of CO Capture in Physical Solvents.

CO sorption in physical solvents is one of the promising approaches for carbon capture from highly concentrated CO streams at high pressures. Identifying an efficient solvent and evaluating its solubility data at different operating conditions are highly essential for effective capture, which generally involves expensive and time-consuming experimental procedures. This work presents a machine learning based ultrafast alternative for accurate prediction of CO solubility in physical solvents using their physical, thermodynamic, and structural properties data.

Simultaneous determination of phthalates and bisphenols from plastic bottled water samples by dispersive solid-phase extraction with multiwalled carbon nanotubes and liquid chromatography/atmospheric pressure photoionization/high-resolution mass spectrometry.

Rationale: Phthalates and bisphenols were reported as endocrine disrupting chemicals and hence a potential threat to human health. Polyethylene terephthalate bottles are being used to store drinking water and the probability of migration of phthalates and bisphenols from the bottles into the water is high. The migration of analytes with respect to different storage conditions need to be studied.

Machine learning aided experimental approach for evaluating the growth kinetics of Candida antarctica for lipase production.

A hybrid machine learning (ML) aided experimental approach was proposed in this study to evaluate the growth kinetics of Candida antarctica for lipase production. Different ML models were trained and optimized to predict the growth curves at various substrate concentrations. Results on comparison demonstrate the superior performance of the Gradient boosting regression (GBR) model in growth curves prediction.