A 1T-MoS2/FBC composite electrode for efficient removal of Cr(VI) through capacitive deionization: performance, mechanism, and BP neural network modeling.

Hexavalent chromium (Cr(VI)), a highly toxic and mobile heavy metal, poses significant environmental and health risks, making its removal a critical challenge. Capacitive deionization (CDI) has emerged as a promising method for Cr(VI) removal due to its environmental sustainability and energy efficiency. However, the effectiveness of CDI is largely influenced by the electrode materials. In this study, a novel composite, 1T-MoS2/FBC, was synthesized through the in situ growth of metallic 1T-MoS2 on Fenton-treated bamboo-derived biochar. The 1T-MoS2/FBC electrode exhibited a maximum electrosorption capacity of 83.92 mg g-1 for Cr(VI) at 45 °C under an applied voltage of 1.2 V and a flow rate of 20 mL min-1. Notably, the electrode demonstrated excellent reusability, maintaining 84.51% of its initial removal efficiency after 10 consecutive electrosorption-desorption cycles. This superior cycling stability is attributed to its low charge-transfer resistance (0.385 Ω) and the synergistic removal mechanisms involving electrostatic attraction, redox reactions, and surface coordination. Crucially, a machine learning model, specifically a back-propagation (BP) neural network, was successfully developed to simulate and predict the electrosorption behaviors of 1T-MoS2/FBC, achieving high predictive accuracy (R2 = 0.9843, MSE = 0.0114). This BP neural network model not only offers a promising alternative to extensive and tedious experiments for predicting electrosorption capacity but also elucidates the complex non-linear relationships among various influencing factors during the electrosorption process, offering new opportunities for intelligent material design, performance prediction, and process optimization in environmental applications.