Weintroduce a general class of orbital-free density functionals(OF-DFT)decomposed into a local part in coordinate space and a local part in reciprocal space.As a demonstration of principle,we choose for the former the...Weintroduce a general class of orbital-free density functionals(OF-DFT)decomposed into a local part in coordinate space and a local part in reciprocal space.As a demonstration of principle,we choose for the former the Thomas-Fermi-von Weizsäcker(TFW)kinetic energy density functional(KEDF)and for the latter a form derived from the Lindhard function,but with the two system-dependent adjustable parameters.These parameters are machine-learned from Kohn-Sham data using Bayesian linear regression with a kernel method,which employs moments of the Fourier components of the electronic density as the descriptor.Through a number of representative cases,we demonstrate that our machine-learned model provides more than an order-of-magnitude improvement in the accuracy of the frozen-phonon energies compared to theTFWKEDF,with negligible increase in the computational cost.Overall,this work opens an avenue for the construction of accurate KEDFs for OF-DFT.展开更多
基金the support of the Quantum Science and Engineering Center(QSEC)at George Mason UniversityP.S.gratefully acknowledges funding from the U.S.Department of Energy,Office of Science,under Grant No.DE-SC0023445+1 种基金T.O.acknowledges support provided by the George Mason University Provost Dissertation Completion GrantM.E.has been partially supported by the Simons Foundation.
文摘Weintroduce a general class of orbital-free density functionals(OF-DFT)decomposed into a local part in coordinate space and a local part in reciprocal space.As a demonstration of principle,we choose for the former the Thomas-Fermi-von Weizsäcker(TFW)kinetic energy density functional(KEDF)and for the latter a form derived from the Lindhard function,but with the two system-dependent adjustable parameters.These parameters are machine-learned from Kohn-Sham data using Bayesian linear regression with a kernel method,which employs moments of the Fourier components of the electronic density as the descriptor.Through a number of representative cases,we demonstrate that our machine-learned model provides more than an order-of-magnitude improvement in the accuracy of the frozen-phonon energies compared to theTFWKEDF,with negligible increase in the computational cost.Overall,this work opens an avenue for the construction of accurate KEDFs for OF-DFT.
