Calculating spin crossover temperatures by a first-principles LDA+U scheme with parameter U evaluated from GW.

Area of Science:

• Computational chemistry
• Materials science
• Quantum mechanics

Background:

• Accurate prediction of spin crossover (SCO) temperatures (T1/2) is crucial for designing functional materials.
• The Local Density Approximation combined with the Hubbard U model (LDA+U) offers a balance between accuracy and computational cost for SCO studies.
• Static U values in LDA+U can limit predictive power and fail to account for electronic structure changes.

Purpose of the Study:

• To develop a framework for deriving dynamical U (U_dyn) values from many-body GW calculations for iron(II) complexes.
• To assess the performance of the proposed LDA+U_dyn method in predicting experimental ground spin states and SCO temperatures.
• To overcome the limitations of static U values in computational modeling of SCO phenomena.

Main Methods:

• Derivation of dynamical U (U_dyn) values using many-body GW calculations.
• Application of the LDA+U_dyn method to iron(II) complexes with varying ligand fields.
• Model calculations and comparison with experimental data for SCO properties.

Main Results:

• U values derived from GW calculations are dependent on the local electronic environment.
• The LDA+U_dyn method successfully reproduces experimental ground spin states of selected SCO complexes.
• The LDA+U_dyn approach significantly improves the calculation of thermochemical quantities compared to static U.
• Predicted T1/2 values show a small mean absolute error of 176 K relative to experimental data.

Conclusions:

• The proposed framework provides a more accurate description of local Coulomb correlation for SCO systems.
• Dynamical U values derived from GW calculations enhance the predictive capability of LDA+U for SCO temperatures.
• The LDA+U_dyn method offers a reliable computational tool for the design and understanding of spin crossover materials.