TY - JOUR
T1 - Comparative prediction of red alga biosorbent performance in dye removal using multivariate models of response surface methodology (RSM) and artificial neural network (ANN)
AU - Mokhtar, Nadiah
AU - Aziz, Edriyana A.
AU - Aris, Azmi
AU - Ishak, W. F.W.
AU - Majeed, Anwar P.P.Abdul
AU - N.Moni, Syazwan
AU - Sa'at, Siti Kamariah Md
N1 - Publisher Copyright:
© 2018 Authors.
PY - 2018
Y1 - 2018
N2 - Red algae species, Euchema Spinosum (ES) in Malaysia possesses excellent biosorbent properties in removing dyes from aqueous solutions. In the present study, the experimental design for the biosorption process was carried out via response surface methodology (RSM-CCD). A total of 20 runs were carried out to generate a quadratic model and further analysed for optimisation. Prior to the evaluation, the characterisation study of the ES was performed. It was observed that the maximum uptake capacity of 399 mg/g (> 95%) is obtained at equilibrium time of 60 min, pH solution of 6.9-7.1, dosage of 0.72 g/L and initial dye concentration of 300 g/L through statistical optimisation (CCD-RSM) based on the desirability function. It is demonstrated in the present study that the ANN model (R 2 = 0.9994, adj-R 2 = 0.9916, MSE = 0.19, RMSE = 0.4391, MAPE = 0.087 and AARE = 0.001) is able to provide a slightly better prediction in comparison to the RSM model (R 2 = 0.9992, adj-R 2 = 0.9841, MSE = 1.95, RMSE = 1.395, MAPE = 0.08 and AARE = 0.001). Moreover, the SEM-EDX analysis indicates the development of a considerable number of pore size ranging between 132 to 175 μm. From the experimental observations, it is evident that the ES can achieve high removal rate (> 95%), indeed become a promising eco-friendly biosorptive material for MB dye removal.
AB - Red algae species, Euchema Spinosum (ES) in Malaysia possesses excellent biosorbent properties in removing dyes from aqueous solutions. In the present study, the experimental design for the biosorption process was carried out via response surface methodology (RSM-CCD). A total of 20 runs were carried out to generate a quadratic model and further analysed for optimisation. Prior to the evaluation, the characterisation study of the ES was performed. It was observed that the maximum uptake capacity of 399 mg/g (> 95%) is obtained at equilibrium time of 60 min, pH solution of 6.9-7.1, dosage of 0.72 g/L and initial dye concentration of 300 g/L through statistical optimisation (CCD-RSM) based on the desirability function. It is demonstrated in the present study that the ANN model (R 2 = 0.9994, adj-R 2 = 0.9916, MSE = 0.19, RMSE = 0.4391, MAPE = 0.087 and AARE = 0.001) is able to provide a slightly better prediction in comparison to the RSM model (R 2 = 0.9992, adj-R 2 = 0.9841, MSE = 1.95, RMSE = 1.395, MAPE = 0.08 and AARE = 0.001). Moreover, the SEM-EDX analysis indicates the development of a considerable number of pore size ranging between 132 to 175 μm. From the experimental observations, it is evident that the ES can achieve high removal rate (> 95%), indeed become a promising eco-friendly biosorptive material for MB dye removal.
KW - Artificial Neural Network
KW - Biosorbent
KW - Decolourization
KW - Euchema spinosum
KW - Methylene Blue
KW - Response Surface Methodology
UR - http://www.scopus.com/inward/record.url?scp=85059223464&partnerID=8YFLogxK
U2 - 10.14419/ijet.v7i4.35.22908
DO - 10.14419/ijet.v7i4.35.22908
M3 - Article
AN - SCOPUS:85059223464
SN - 2227-524X
VL - 7
SP - 551
EP - 558
JO - International Journal of Engineering and Technology(UAE)
JF - International Journal of Engineering and Technology(UAE)
IS - 4
ER -