Optimization of oleic-estolide fatty acid synthesis using response surface methodology and artificial neural networks

Abstract

Depleting petroleum resources, increasing environmental concerns, and stringent government regulations have led to a global paradigm shift towards bio-lubricants. Vegetable oils are seen as a potential substitute for petroleum-based lubricants but are limited by their low resistance to oxidative degradation, low hydrolytic stability, and poor cold-flow properties. Chemical modification in the form of estolide synthesis from oleic acid leads to increased oxidative & hydrolytic stability and better cold flow properties. Reaction temperature, catalyst, and reaction time of the synthesis of estolide fatty acid were optimized with the aid of response surface methodology (RSM) using acid value (AV), pour point (PP) & kinematic viscosities (KV) at 40 °C & 100 °C as response variables. A low reaction temperature of 40 °C, catalyst amount of 0.569 mol equivalents and the reaction time of 11.857 h were found to be optimum for the synthesis of estolide fatty acid with PP as low as − 9 °C, KV around 203 cSt & 22.5 cSt at 40°C & 100 °C respectively. The results provided by RSM were validated using Artificial Neural Networks (ANN) created with the aid of Matlab & TensorFlow platforms. The generated models were found to be efficient in predicting the known as well as unknown inputs with error less than 5 %. • Optimization of estolide fatty acid synthesis using response surface methodology. • Artificial Neural Networks were utilised to validate optimal solutions. • Estolide fatty acid synthesized at optimal conditions had a pour point of − 9 °C. • Average Estolide Number of synthesized at oleic-estolide fatty acid is 2.13.