High-Accuracy Semiempirical Models for Organic Materials under Extreme Conditions

In previous work, we have developed high accuracy semiempirical models for organic molecules by leveraging a machine-learned force field based on Chebyshev polynomials [1]. The benefits of our approach are: (1) many-body interactions can be corrected for in a systematic and rapidly tunable process, and (2) high-level quantum accuracy for a broad range of compounds can be achieved with significant less data compared to other machine learning potentials. Here, we expand our effort to study organic materials under extreme conditions, which exhibit complicated chemistry. We show that our model has capability of reproducing the structural properties and chemistry at high accurate quantum level for a wide range of thermodynamic conditions. Therefore, our method can allow for quantum-accurate simulations on time- and length- scales inaccessible to DFT with relatively small training data.

[1] C.H. Pham, R.K. Lindsey, L.E. Fried, N. Goldman, J. Phys. Chem. Lett. 13, 2934–2942 (2022)