Thermal monitoring of phenoxyacid herbicide adsoprtion on granular activated carbon

A thermal detection technique, which is based on monitoring the effects of a partial desorption of adsorbed material resulting from an applied heat pulse, has been used to follow the adsorption of phenoxyacid herbicides (MCPA, MCPP, and MCPB) on a carbon packed column. Results obtained show that this approach allows effective prediction of single component breakthrough using a single-sensor configuration and an artificial neural network for signal interpretation. However, the integration of a multiple-sensor array using a modified neural network was found to give improved performance. Although this work has focused mainly on single-solute systems, preliminary results are presented which demonstrate the potential for monitoring binary-solute loadings.