Search papers, labs, and topics across Lattice.
This paper introduces an embedded correlated wavefunction transfer learning (ECW-TL) framework to improve the accuracy of molecular dynamics simulations by incorporating high-level electron exchange and correlation effects. The framework fine-tunes a baseline DFT model using embedded DFT and correlated wavefunction data, enabling efficient transfer of chemical accuracy to large-scale simulations. Applied to Ca2+-CO32- ion pairing in aqueous solution, ECW-TL demonstrates the importance of exact electron exchange and correlation for accurately predicting ion-pair stability and structure.
Transfer learning can bring computationally expensive, high-accuracy quantum chemistry methods to large-scale molecular simulations of complex chemical processes.
Achieving chemical accuracy for molecular simulations remains a central challenge in computational chemistry. Here, we present an embedded correlated wavefunction transfer learning (ECW-TL) framework for accurately simulating molecular dynamics in the condensed phase. ECW-TL incorporates high-level electron exchange and correlation effects in ECW theory while preserving training and computational efficiency of machine learned interatomic potentials. We demonstrate the framework on Ca2+-CO32- ion pairing in aqueous solution, a key process underlying CO2 mineralization in seawater. As proof of principle, we first show that finetuning a DFT-revPBE-D3(BJ) baseline model with embedded-DFT-SCAN data reproduces the DFT-SCAN free-energy surface within 1 kcal/mol across all solvation states. Extending the framework to embedded MP2 and localized natural-orbital CCSD(T) further refines the free-energy profile, revealing the crucial role of exact electron exchange and correlation in determining ion-pair stability and structure. ECW-TL thus provides a general, data-efficient route for transferring CW accuracy to large-scale simulations of complex aqueous and interfacial chemical processes.
