Adsorption isotherm models for dye removal by cationized starch-based material in a single component system: Error analysis

• Published in: Journal of hazardous materials Vol. 157; no. 1; pp. 34 - 46
• Main Authors: Gimbert, Frédéric, Morin-Crini, Nadia, Renault, François, Badot, Pierre-Marie, Crini, Grégorio
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 30-08-2008; Elsevier
• Subjects: Acid Blue 25; Adsorption; Applied sciences; Chemical engineering; Chemical Sciences; Coloring Agents - analysis; Error analysis; Exact sciences and technology; Freundlich; Generalized model; Isotherm; Kinetics; Langmuir; Models, Chemical; Pollution; Polymers; Redlich–Peterson; Starch; Starch - chemistry; Tempkin; Thermodynamics; Toth; Water Pollutants, Chemical - analysis; Water Purification - methods; Redlich–Peterson; Freundlich; Error analysis; Adsorption; Acid Blue 25; Starch; Tempkin; Langmuir; Generalized model; Isotherm; Toth; Correlation coefficient; Discoloration; Dyes; Non linear regression; Linear regression; Bias; Adsorption isotherm; Adsorption capacity; Redlich-Peterson; Modeling; Physicochemical purification; Aqueous solution; Linearization
• ISSN: 0304-3894; 1873-3336
• DOI: 10.1016/j.jhazmat.2007.12.072

This article describes the adsorption of an anionic dye, namely C.I. Acid Blue 25 (AB 25), from aqueous solutions onto a cationized starch-based adsorbent. Temperature was varied to investigate its effect on the adsorption capacity. Equilibrium adsorption isotherms were measured for the single component system and the experimental data were analyzed by using Langmuir, Freundlich, Tempkin, Generalized, Redlich–Peterson, and Toth isotherm equations. Five error functions were used to determine the alternative single component parameters by non-linear regression due to the bias in using the correlation coefficient resulting from linearization. The error analysis showed that, compared with other models, the Langmuir model described best the dye adsorption data. Both linear regression method and non-linear error functions provided the best-fit to experimental data with the Langmuir model.