Assessment and optimization of an ultrasound-assisted washing process using organic solvents for polychlorinated biphenyl-contaminated soil.

The goal of this work was to evaluate a washing process that uses organic solutions for polychlorinated biphenyl (PCB)-contaminated soil, and includes an ultrasound pre-treatment step to reduce operational times and organic solvent losses. In a preliminary trial, the suitability of 10 washing solutions of different polarities were tested, from which three n-hexane-based solutions were selected for further evaluation. A second set of experiments was designed using a three-level Taguchi L27 orthogonal array to model the desorption processes of seven different PCB congeners in terms of the variability of their PCB concentration levels, polarity of the washing solution, sonication time, the ratio washing solution/soil, number of extraction steps and total washing time. Linear models were developed for the desorption processes of all congeners. These models provide a good fit with the results obtained. Moreover, statistically significant outcomes were achieved from the analysis of variance tests carried out. It was determined that sonication time and ratio of washing solution/soil were the most influential process parameters. For this reason they were studied in a third set of experiments, constructed as a full factorial design. The process was eventually optimized, achieving desorption rates of more than 90% for all congeners, thus obtaining concentrations lower than 5 ppb in all cases. The use of an ultrasound-assisted soil washing process for PCB-contaminated soils that uses organic solvents seems therefore to be a viable option, especially with the incorporation of an extra step in the sonication process relating to temperature control, which is intended to prevent the loss of the lighter congeners.