Band-Edge Prediction of 2D Covalent Organic Frameworks from Molecular Precursor via Machine Learning

The band-edge positions of two-dimensional (2D) covalentorganicframeworks (COFs) play a crucial role in their applications in photocatalystsand nanoelectronics. However, massive amounts of 2D COFs with targetedband-edge positions from high-level first-principles calculationsbased on their composition are time-consuming due to the diversityand complexity of unit cell structures. Here, we report a strategyto predict the band-edge positions of 2D COFs by combining first-principlescalculations with machine learning (ML). The root-mean-square error(RMSE) of the predicted valence band maximum (VBM) and conductionband minimum (CBM) between ML prediction and first-principles calculatedvalues at the Perdew-Burke-Ernzerhof (PBE) level are0.229 and 0.247 eV in test data set, respectively. In addition, alinear relationship is established between the PBE results and theHSE06 results with RMSE values of 0.089 and 0.042 eV for VBMs andCBMs in the test data set. Finally, a workflow is developed to determinethe band-edge positions of the 2D COFs.

Automated summary

We propose a strategy to predict the band-edge positions of two-dimensional covalent organic frameworks (COFs) by combining first-principles calculations with machine learning (ML). The root-mean-square error (RMSE) between ML prediction and first-principles calculated values for the valence band maximum (VBM) and conduction band minimum (CBM) is 0.229 and 0.247 eV, respectively, in the test data set. Additionally, a linear relationship is established between the PBE and HSE06 results with RMSE values of 0.089 and 0.042 eV for VBMs and CBMs in the test data set. Finally, a workflow is developed to determine the band-edge positions of the 2D COFs.