This is second paper in series in which we propose computer-aided molecular design (CAMD) methodology for the optimal design of adsorbents for adsorption of natural occurring radioactive material (NORM) of natural gas fracking waste. The adsorbent selection optimization problem is posed as a mixed integer nonlinear programming problem and a new algorithm called efficient ant colony optimization (EACO) is used to solve the problem. The algorithm maximizes the absorption capacity of the adsorbent subject to structural feasibility, thermodynamic property correlations, process conditions derived from produced water and mass balance on the adsorption process constraints. New adsorbents for removal of radium and barium together and individually are proposed. These adsorbents show far greater capacity to remove radium and barium from produced water than state of the art adsorbents found in literature. (C) 2015 Published by Elsevier Ltd.