ABSTRACT

Citric acid has high economical value and as wide range applications in various industry. Some of agro-industrial wastes used for citric acid production include pineapple peel, potato peel and groundnut shell. The aim of this study is to optimize the effect of plant oils required to maximize citric acid yield from a tri-substrate feedstock using Aspergillus niger. Mixture and Response Surface designs were respectively used to study the effect of the composite feedstock and plant oils (Castor oil, Coconut oil and Olive oil) on the production of citric acid using Aspergillus niger. Optimization was carried out using Response Surface Methodology (RSM) and Artificial Neural Network (ANN). A quadratic model was obtained to predict the concentration of citric acid as a function of the composite feedstock and plant oils. The optimum composition of the tri-substrate feedstock obtained from the mixture design was 13.948g of pineapple peel, 0g of potato peels and 1.05g of groundnut shells. The maximum citric acid yield was gotten to be 26.93g/l. For ANN, Quick Propagation (QP) with Sigmoid for both hidden and output layers was the best model for predicting citric acid production. The performance of both models was evaluated based on their R² and Root Mean Square Error (RMSE). ANN with R² and RMSE of 0.99988 and 0.0511 respectively was superior to RSM with R² and RMSE of 0.9081 and 0.8279 respectively. Optimization studies produced maximum citric acid yield at/26.93g/l and 48.6953g/l for RSM and ANN respectively. ANN was also considered to be more efficient because of the optimal plant yield. The plant oils were able to enhance citric acid production. At maximum yield of citric acid concentration of 48.6953ml, optimum levels of factors were Castor oil of 2.999% w/w, coconut oil of 2.9654% w/w and Olive oil of 2.99467% w/w. It was therefore concluded that ANN is better than RSM in the modeling and optimization of the effects of plant oils on citric acid production.