Abstract
[Display omitted]
•Maximum removal of Congo Red on black cardamom activated carbon occurs at pH 6.•Langmuir adsorption capacity was found to be 69.93 mg/g.•Uptake of Congo Red follows pseudo-second-order kinetics.•Congo Red adsorption was found to be spontaneous, random, and exothermic.•Predictions of Congo Red adsorption by Machine Learning models were accurate and generalized (R2 > 0.99).
The present work utilizes waste black cardamom (BC) as an inexpensive and environmentally friendly adsorbent for sequestering the Congo Red (CR) dye from aqueous media for the first time. Following a carbonization process at 600 °C, chemical activation with KOH was carried out for waste BC and subsequent black cardamom activated carbon (BCAC) was employed as an absorbent for CR eradication. The effect of experimental factors, including pH, adsorption time, dose and CR initial concentration, was investigated. 96.21 % of CR dye removal was achieved at pH 6 for 100 mg/L of CR concentration having 0.1 g dose at 30 °C. Maximum Langmuir adsorption capacity of BCAC was found to be 69.93 mg/g at 30 °C. The kinetic analyses showed that the CR adsorption over BCAC behaved in accordance with a pseudo-second order kinetic model as high R2 values (0.997–1) were obtained. Thermodynamic parameters (ΔH°, ΔS°, and ΔG°) demonstrated that the CR adsorption over BCAC was feasible, spontaneous and exothermic in nature. In addition, the state-of-the-art machine learning (ML) approaches namely, support vector regression (SVR) and artificial neural network (ANN) were employed for modeling the BCAC adsorbent for CR removal. The statistical analysis revealed high prediction performance of SVR model with AARE value of 0.0491 and RMSE value of 0.4635 while the corresponding values for the ANN model were 0.0781 and 0.5395, respectively. Furthermore, the plots between experimental CR data and ML forecasted data were closely matched (R2 > 0.99). Thus, it can be concluded that BC, an agro waste could be utilized for CR removal and that the adoption of ML approaches can benefit users by providing them with a tool to enhance the design and performance of wastewater treatment operations.