The project was designed to study the removal of color using adsorption techniques. The sorption of three acid dyes, namely, Acid Blue 80 (AB80), Acid Red 114 (AR1 14), Acid Yellow 117 (AYl17) have been studied Equilibrium isotherms have been measured for three single component systems (AB, AR and AY) and three binary component systems (AB+AY, AB+AR and AR+AY). The isotherms were determined by shaking 0.05g activated carbon, particle size ranges 500-710 microns, and with 0.05dm³ dye solution of initial concentrations from 10mg/dm³ to 250mg/dm³. A constant temperature agitating shaking water bath was used at 200 rev/min and the temperature maintained at 20±2°C. A contact time of 21 days was required to achieve equilibrium. Analysis of data has been carried out in three stages....[ Read more ]

The project was designed to study the removal of color using adsorption techniques. The sorption of three acid dyes, namely, Acid Blue 80 (AB80), Acid Red 114 (AR1 14), Acid Yellow 117 (AYl17) have been studied Equilibrium isotherms have been measured for three single component systems (AB, AR and AY) and three binary component systems (AB+AY, AB+AR and AR+AY). The isotherms were determined by shaking 0.05g activated carbon, particle size ranges 500-710 microns, and with 0.05dm³ dye solution of initial concentrations from 10mg/dm³ to 250mg/dm³. A constant temperature agitating shaking water bath was used at 200 rev/min and the temperature maintained at 20±2°C. A contact time of 21 days was required to achieve equilibrium. Analysis of data has been carried out in three stages.

1) Single component analysis: The experimental isotherm data were analyzed using Langmuir, Freundlich, Redlich-Peterson, Toth, Temkin and Dubinin-Radushkevich equations for each individual dye. Based on the Langmuir isotherm analysis, the monolayer adsorption capacities are 112.3mg Acid Blue per g carbon, 103.5mg Acid Red per g carbon and 155.8mg Acid Yellow per g carbon. Correlation coefficients were determined for each isotherm analysis; the Redlich-Peterson isotherm equatilon provides the best fit model for the sorption of the three acid dyes, AB80, AR1 14 and AY 117, onto activated carbon compared to the other five isotherms.

2) Error analysis selection: Due to the inherent bias in using the correlation coefficient resulting from linearization, alternative single component parameters were determined by non-linear regression. Five error functions were used and examined in this study, namely: Sum of the squares of the errors (ERRSQ); Hybrid fractional error function (HYBRID); Marquardt'S percent standard deviation (MPSD); Average relative error (ARE); and Sum of the absolute errors (EABS). It was found that the HYBRID error function method provided the best parameters for the five isotherm equations in the acid dye systems, when the errors were norrnalized for comparison purposes.

3) Multicomponent analysis: Methods of predicting and correlating multicomponent dye isotherms were developed and applied using the single component dye adsorption data to predict bisolute equilibrium data. Simple metholds based on the Langmuir isotherm equation were used to predict or correlate single-component and multicomponent isotherms. They are the Extended Langmuir Equation, the Jam and Snoeyink Modified Extended Langmuir Model, the P-Factor and the Interaction Factor models. Two different parameter sets, based on LTFM and HYBRID error functions, of Langmuir isotherm parameters were applied in tile Langmuir based multicomponent isotherm equations. It was found that the P-factor and interaction factor models are better methods to predict the multicomponent experimental data and the LTFM Langmuir constants gave the best fit model to tile multicomponent data prediction. Moreover, the Extended Freundlich and Redlic h-Peterson isotherm equations were tested to predict the bisolute equilibrium data. Tire Extended Freundlich isotherm proved successful in predicting bisolute isotherm data for the three two-component dye mixtures adsorbing onto activated carbon but the Extended Redlich-Peterson approach failed to give accurate predictions.