Modeling and Optimization of the Solvent Extraction of Oil from Africa Locust Bean (Parkia biglobosa) using Response Surface Methodology and Regression Analysis: A Comparative Study
Akintola, J. T. *
Department of Chemical Engineering, Lagos State University, Lagos, Nigeria.
Ezeka, F. C.
Department of Chemical Engineering, University of Lagos, Lagos, Nigeria.
Ayoola, A. I.
Department of Chemical Engineering, University of Lagos, Lagos, Nigeria.
Ebuehi, O. N.
Department of Chemical Engineering, University of Lagos, Lagos, Nigeria.
Asama, M. O.
Department of Chemical Engineering, University of Lagos, Lagos, Nigeria.
Obialor, M. C.
Department of Chemical Engineering, University of Uyo, Akwa Ibom, Nigeria.
Abdulkareem, Y. T.
Department of Chemical Engineering, Lagos State University, Lagos, Nigeria.
Akintola, O. E.
Department of Statistics, Yaba College of Technology, Lagos, Nigeria.
Ekpotu, W. F.
Department of Chemical Engineering, University of Uyo, Akwa Ibom, Nigeria.
Sodeinde, O. A.
Department of Chemical Engineering, Lagos State University of Science and Technology, Lagos, Nigeria.
Iwube, P.
Department of Chemical Engineering, University of Calabar, Crossriver, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
The tree of Locust bean which is often regarded as Parkia Biglobosa tree is a perennial deciduous tree mostly grown in West Africa. Owing to its edible property as suitable for eating, it is classified to be a member of the leguminous family. Oil extracted from this seed is found to be medicinal and of great importance to human health. In this paper, optimization (maximization) of oil yield extracted from Parkia Biglobosa (locust bean) via solvent extraction was carried out, using Boxbehken Design of Response Surface Methodology (RSM) for the design of solvent extraction experiment; and comparing the predicting strength of fit regression model (FRM), response surface methodology and poisson regression analysis (PRA) on the oil yield. The solvent employed in the extraction process is ethanol, and the effect of predicting variables (temperature, extraction time and solvent: seed) are studied on the response variable (oil yield). The statistical optimization with RSM reveals optimum oil yield of 21.9161% obtained from Locust Beans at extraction temperature, extraction time and solvent-to-seed ratio (solvent:seed) of 64.8515 0C, 2.99659 hr and 8.85821 ml/g, respectively. While statistical optimization with PRA reveals optimum oil yield of 21.3123% obtained from Locust Bean at extraction temperature of 65 0C, time of 3 hrs and solvent:seed of 9 ml/g; respectively. Hence, showing approximate optimization performance of RSM and PRA. This reveals that locust bean belongs to the class of oil bearing seeds. Owing to the comparative strength of the different modeling tools, it is observed that Response Surface Methodology gives best prediction of the oil yield with the highest R squared value of 97.78%, compared to 88.36% and 87.37% of Fit Regression Model and Poisson Regression Analysis, respectively.
Keywords: Callus, epicotyl, Parkia biglobosa, hypocotyl,, extraction, regeneration, response surface methodology, fit regression analysis, poisson regression analysis