Investigations on the local structure and the spin Hamiltonian parameters for Cu2+ in xCuO-(68-x)V2 O5 -32TeO2 glasses.

The defect structure, spin Hamiltonian parameters (SHPs: anisotropic g factors g ‖ and g ⊥ and the hyperfine structure constants A ‖ and A ⊥ ), and their compositional dependence of Cu 2 + in x CuO - ( 68 - x ) V 2 O 5 - 32 TeO 2 ( x = 5, 10, 20, 30 mol%) glasses are quantitatively analyzed by using the higher-order perturbation formula of octahedral complex with tetrahedral elongation distortion. Due to the Jahn-Teller effect, the [ CuO 6 ] 10 - group is subjected to tetragonal elongation distortion of varying degrees. D q , N , ρ , κ , and H show nonlinear changes with the concentrations of Cu 2 + . When x = 10 mol% CuO, the degree of distortion ( ρ ≈ 0 . 1 % ) is the smallest; when x = 30 mol% CuO, the degree of distortion ( ρ ≈ 15 % ) is the largest, which indicates that excessive distortion leads to the appearance of Z -axis oxygen vacancies and the coordination number of copper ions from six to four. The increasing tendency of the evaluated N and H reveals decreasing covalency of the whole glass system. Present theoretical studies would be useful to the explore the structural properties and optical applications of glass with different CuO concentrations.

Achieving vibrational energies of diatomic systems with high quality by machine learning improved DFT method.

When using ab initio methods to obtain high-quality quantum behavior of molecules, it often involves a lot of trial-and-error work in algorithm design and parameter selection, which requires enormous time and computational resource costs. In the study of vibrational energies of diatomic molecules, we found that starting from a low-precision DFT model and then correcting the errors using the high-dimensional function modeling capabilities of machine learning, one can considerably reduce the computational burden and improve the prediction accuracy. Data-driven machine learning is able to capture subtle physical information that is missing from DFT approaches. The results of 12C16O, 24MgO and Na35Cl show that, compared with CCSD(T)/cc-pV5Z calculation, this work improves the prediction accuracy by more than one order of magnitude, and reduces the computation cost by more than one order of magnitude.