Response surface modeling with Box-Behnken design for strontium removal from soil by calcium-based solution.

Owing to its physicochemical similarity to strontium (Sr), calcium (Ca) was tested as a key component of a soil washing solution for Sr-contaminated soil collected near a nuclear power plant. A four-factor, three-level Box-Behnken experimental design combined with response surface modeling was employed to determine the optimal Sr washing condition for Ca-based solution. The Ca concentration (0.1-1 M), liquid-to-soil ratio (5-20), washing time (0.5-2 h), and pH (2.0-7.0) were tested as the independent variables. From the Box-Behnken design, 27 sets of experimental conditions were selected, and a second-order polynomial regression equation was derived. The significance of the independent parameters and interactions was tested by analysis of variance. Ca concentration was found to be the most influential factor. To determine whether the four variables were independent, three-dimensional (3D) response surface plots were established. The optimal washing condition was determined to be as follows: 1 M Ca, L/S ratio of 20, 1 h washing, and pH = 2. Under this condition, the highest Sr removal efficiency (68.2%) was achieved on a soil contaminated with 90.1 mg/kg of Sr. Results from five-step sequential extraction before and after washing showed that 84.0% and 82.9% of exchangeable and carbonate-bound Sr were released, respectively. In addition, more tightly bound Sr, such as Fe/Mn oxides-bound and organic matter-bound Sr, were also removed (86.2% and 64.5% removal, respectively).