Biosorption of Synthetic Dyes on Spruce Wood Shavings from Binary Solutions: A Comparison of Equilibrium Models

Abstract

Aims: To compare various models describing the sorption of synthetic dyes in multicomponent systems. For selected models, their predictive ability (an ability to predict the sorption behavior in muticomponent systems from the model parameters obtained in the single-component systems) was examined. Place and Duration of Study: University of J. E. Purkyne in Ústínad Labem, Faculty of the Environment, Czech Republic, January 2012 to June 2013. Methodology: Equilibrium sorption experiments were performed in a batch arrangement, the dye concentrations were determined by spectrophotometry and HPLC, and the experimental data were evaluated by the method of non-linear regression. Results: The single-component equilibrium data was evaluated successfully with the three-parameter isotherm equations, such as the Langmuir-Freundlich isotherm. To describe the experimental dependencies for multicomponent systems, various isotherms can be used (derived mostly from the well-known single-component isotherms), but not all of them are capable to predict the dye sorption in muticomponent systems from the model parameters obtained in the single-component system. To compare the predictive ability of various models, an Average Relative Deviation (ARD) and Root Mean Squared Deviation (RMSD) were used as estimators. Conclusions: Untreated wood shavings form Picea abies are applicable as a low-cost non-conventional sorbent for the removal of synthetic dyes form aqueous solutions. The sorption of Egacid Orange (Acid Orange 7), Egacid Yellow (Acid Yellow 11) and Acid Red 88 as representatives of anionic azodyes with strongly acidic sulfonic group was studied in a single-component system, as well as in binary systems. Various isotherm equations were used to fit the equilibrium experimental data in both systems. Among the examined models, the Toth multicomponent isotherm exhibited quite good prediction ability, whereas the predictions based on a widely used competitive Langmuir model were rather poor.