Random forest modeling for the kinetic and isotherm study of malachite green adsorption from aqueous environments using zinc sulfide nanoparticle loaded with activated carbon

In this study, a novel zinc sulfide nanoparticle loaded with activated carbon (ZnS-NP-AC) was used to treat hazardous dye malachite green (MG) from aqueous solution. Further, multiple linear regression (MLR) and random forest (RF) were applied to project the performance of the process. Bruner-Emmet-Teller surface area measurement, field emission scanning electron microscope, X-ray diffraction, Barrett-Joyner-Halenda and UV spectrum analyses were utilized to characterize the prepared adsorbent. Over 98% of the dye was removed under optimal conditions: initial pH 7, contact time 30 min, adsorbent dose 0.02 g and initial dye concentration 15 mg L-1. The adsorption efficacy continued to stay unchanged when pH ranged from 2 to 8. It was found that the dye adsorption followed the pseudo-second-order rate equation. To determine the rate-limiting step of the adsorption process, the intraparticle diffusion model was employed. The Langmuir isotherm model fitted the data significantly better and adsorption capacity was 500 mg g(-1) of the adsorbent. The novel adsorbent investigated in this study can be considered as a suitable alternative for MG removal from aqueous solutions. The optimum tuning variables in the RF model were attained according to n(tree) = 100, m(try) = 2, importance = 1 and nPerm = 3. To test data set, the mean squared error (MSE) values of 7.1e-04 and the coefficient of determination (R-2) value of 0.9826 for the RF model and the MSE value of 0.008 and the R-2 value of 0.9091 for the MLR model were achieved. The findings showed that the RF model is a better model than MLR.