Machine Learning-Driven Discovery of Key Descriptors for CO₂Activation over Two-Dimensional Transition Metal Carbides and Nitrides

Abstract

Fusing high-throughput quantum mechanical screening techniques with modern artificial intelligence strategies is among the most fundamental yet revolutionary science activities, capable of opening new horizons in catalyst discovery. Here, we apply this strategy to the process of finding appropriate key descriptors for CO2 activation over two-dimensional transition metal (TM) carbides/nitrides (MXenes). Various machine learning (ML) models are developed to screen over 114 pure and defective MXenes, where the random forest regressor (RFR) ML scheme exhibits the best predictive performance for the CO2 adsorption energy, with a mean absolute error ± standard deviation of 0.16 ± 0.01 and 0.42 ± 0.06 eV for training and test data sets, respectively. Feature importance analysis revealed d-band center (εd), surface metal electronegativity (χM), and valence electron number of metal atoms (M V) as key descriptors for CO2 activation. These findings furnish a fundamental basis for designing novel MXene-based catalysts through the prediction of potential indicators for CO2 activation and their posterior usage.

A robust Fe-based heterogeneous photocatalyst for the visible-light-mediated selective reduction of an impure CO2 stream

“…90–92 Experimentally, this activation was obtained by light irradiation ( λ = 427 nm), which induced a charge transfer from the catalyst to the molecule, CO 2 chemisorption with the elongation of the C–O bond, bending of O–C–O angle, and finally, dissociation of CO 2 on the catalyst surface into CO and O species. 93,94…”

Machine Learning Accelerated Discovery of Functional MXenes with Giant Piezoelectric Coefficients

“…In recent years, the rapid development of data mining technology has provided a new approach for accelerating the screening of new materials. The combination of machine learning (ML) with DFT calculations or experiments has led to a new wave of prediction of material properties such as the band gap, − free energy, , stability, and piezoelectricity. , The application of machine learning in MXene data sets also could help accelerate the research process of MXenes’ properties. − The quality of the ML models was generally evaluated by train-test split scores or cross-validation results. However, a high score on the testing set does not mean that the model has an excellent extrapolation ability.…”

Advances in MXene surface functionalization modification strategies for CO2 reduction

“…28 In addition, since the standard potential for CO 2 reduction is similar to that of the hydrogenolysis reaction, the involvement of a competing hydrogenolysis reaction (HER) in the catalytic process of MXenes decreases the efficiency of catalytic reduction of CO 2 . 29 Therefore, in order to improve the adsorption and activation of CO 2 molecules, the selectivity of target products and the reduction efficiency during the electrocatalytic reduction process of MXenes, the functionalization and modification of the surface of MXenes to improve their CO 2 reduction performance have become a hot spot in current research.…”