Production of biodiesel from the novel non-edible seed of Chrysobalanus icaco using natural heterogeneous catalyst: Modeling and prediction using Artificial Neural Network

• Published in: Journal of cleaner production 2023-01, Vol.385, p.135631, Article 135631
• Main Authors: Okonkwo, C.P., Ajiwe, V.I.E., Obiadi, M.C., Okwu, M.O., Ayogu, J.I.
• Format: Article
• Language: English
• Subjects: acid value; Artificial neural network; biodiesel; Catalyst; catalysts; Chrysobalanus icaco; diesel fuel; energy; Fourier transform infrared spectroscopy; gas chromatography; methanol; neural networks; pour point; prediction; seed oils; spectrophotometers; surface area; Trans-esterification; viscosity
• ISSN: 0959-6526; 1879-1786
• DOI: 10.1016/j.jclepro.2022.135631

Biodiesel has been referred to as a perfect substitute for diesel fuel because of its numerous promising properties. They are renewable, clean, increase energy security, and improve the environment. The seed oil of Chrysobalanus icaco was characterised using Gas Chromatography-Mass Spectrophotometer (GCMS) and Fourier Transform Infrared Spectroscopy (FTIR). The heterogeneous solid catalyst of periwinkle shell ash was prepared in 3 forms: raw, calcined and acid-activated. They were characterised using Scanning Electron Microscope (SEM) and FTIR. The results of the SEM analysis revealed the calcined samples to be a better choice because of their larger surface area. The result showed that the oil yield of the used crop was promising for commercial biodiesel production, with Chrysobalanus icaco having a yield of 51.90%. The reusability of the catalyst for continuous reaction runs showed that biofuel yield was still high after five cycles: 92.25–80.60% for calcined periwinkle shell ash (PSA) catalyst and 89.26–78.50% for acid-activated PSA catalyst. The result of the fuel properties of the biodiesel and their blend indicated their suitability for biodiesel production. Methyl ester blends of 20:80 had viscosity that placed them in 2D grade diesel (2.0–4.3 mm2/s), helpful in powering stationary equipment. Other fuel properties, including acid value, pour point, flash point and density, were within the ASTM D6751 limits for biodiesels. Artificial Neural Network (ANN) was used to compare the experimental value to the simulated value using MATLAB 2020. The seed oil of Chrysobalanus icaco trans-esterified with methanol using Periwinkle Shell Ash (PSA) catalyst was proven to be a good source of biodiesel. [Display omitted] •Extraction and characterisation of Chrysobalanus icaco (C. icaco) oil.•Biodiesel production using prepared periwinkle shell ash (PSA) as a heterogeneous catalyst.•Investigation of the process parameters of the trans-esterification process.•Determination of the fuel properties of the biodiesel and diesel blends.•Modelling and optimisation using Artificial Neural Networks (ANN).