Artificial neural network (ANN) modeling of dynamic adsorption of crystal violet from aqueous solution using citric-acid-modified rice (Oryza sativa) straw as adsorbent

Abstract

Rice straw, an abundant, lignocellulosic agricultural residue worldwide, was thermochemically modified with citric acid to develop a biodegradable cationic adsorbent. The morphological and chemical characteristics of rice straw and acid-modified rice straw were investigated by scanning electron microscopy, surface area, and porosity analysis by the BET (Brunauer, Emmett, and Teller) nitrogen adsorption method and Fourier transform infrared spectroscopy. The modification process leads to the increase in the specific surface area and pore size of rice straw. In order to investigate the application potential of the prepared adsorbent to remove a cationic dye (Crystal violet) from its aqueous solution, a continuous adsorption study was carried out in a laboratory scale fixed-bed column packed with acid-modified rice straw. Effect of different flow rates and bed heights on the column breakthrough performance was investigated. Results show that with increasing bed height and decreasing flow rate, the breakthrough time was delayed. In order to determine the most suitable model for describing the adsorption kinetics of Crystal violet in the fixed-bed column system, the Bed Depth Service Time model as well as the Thomas model was fitted to the experimental data. An artificial neural network (ANN) based model for determining the dye concentration in the column effluent was also developed. An extensive error analysis was carried out between experimental data and data predicted by the models using the following error functions: correlation coefficient (R 2), average relative error (ARE), sum of the absolute error (SAE), and χ2 statistic test. Based on the values of the error functions, the ANN model was most appropriate for describing the dynamic dye adsorption process.