EVALUATION OF OPTIMAL CONDITIONS FOR CRUDE PALM OIL BLEACHING ON LOCALLY PREPARED ACTIVATED CARBON USING RSM-GA AND A HYBRID ANN-GA

Summary

The acceptability of palm oil as edible cooking oil, as well, as for industrial uses is limited to the associated impurities and the need to
improve on its quality to meet the required standards for both human and industrial uses through bleaching, form the basis for this
work. Optimal conditions for colour removal from crude palm oil on locally prepared activated carbon were investigated. Palm oil
was obtained from local producers at Uli, Ihiala Local Government of Anambra State, Nigeria. The obtained palm oil was subjected
to pretreatment and physico-chemical analysis before bleaching. The activated carbon used as bleaching material was locally
prepared from animal bone through carbonization and chemical activation processes. The prepared activated carbon (ABAC) was
characterized using Fourier Transform Infrared (FTIR), Scanning Electron Microscopy (SEM), Braummer-Emett-Teller (BET) and Xray Diffraction (XRD). Bleaching of the prepared palm oil on ABAC was done at different process conditions. A bench scale
experiment was conducted and the data so obtained were optimized using Response Surface Methodology and Hybrid Artificial
Neuron Network- Genetic Algorithm. The characterization results of BET analysis show that ABAC has a surface area of
593.270m2
/g, micro pore surface area 595.56 m2
/g, micro pore volume o.212cm3
/g, Langmuir surface area 1.38e+04m2
/g and
adsorption energy 3.998KJ/mol. XRD analysis shows the presence of Gypsum, Brushite, Calcium Hydrogen Phosphate Hydroxide and
Ardealite with Gypsum indicating its highest peak at 2θ angle 120 and X-ray count above 2500 which indicates its dominance in the
ABAC. The ANN gave a regression coefficient (R2
) value of 0.987, while that of RSM was 0.861. This shows that ANN can account for
more than 98.7% of the variability in the system while RSM can only account for 86.1% for the system. The ANN model shows good
fitting with the experimental data and gave better performance than RSM as seen from the results. Optimum conditions which
guaranteed maximum 92.31% color reduction were selected after observing the results of the GA. The optimized conditions were
obtained as follows; 25g adsorbent dosage, 46mins adsorption time and 104.32oC adsorbent temperature. It could be established that
model equations were developed, optimal adsorption parameters were estimated for palm oil bleaching on a non-toxic, low cost,
biodegradable activated carbon produced from waste animal bone.

Index Terms

Browse additional articles authored by one or more contributors to this paper.