Ionization Energies of 3d Transition Metal Atoms: Insights from Density Functional Theory and Self-Interaction Correction

Основні поняття

Density functional approximations (DFAs) and self-interaction corrected DFAs exhibit varying performance in describing the ionization energies of 3d transition metal atoms, with the self-interaction correction introducing an "energy penalty" for the noded 3d orbitals.

Анотація

The study investigates the performance of various density functional approximations (DFAs) and self-interaction corrected DFAs in describing the ionization energies of 3d transition metal atoms (Sc-Zn).
Key findings:

Standard DFAs (LSDA, PBE, r2SCAN) tend to overbind the 4s electron in the 4s^1 configuration and underbind the first 4s electron in the 4s^2 configuration, leading to errors in the first ionization energies.
The self-interaction corrected DFAs (FLOSIC-DFT) suffer from a self-interaction correction (SIC) energy penalty for the noded 3d orbitals, leading to an underestimation of the number of 4s electrons in some cases.
The ground-state measure of the sd transfer error decreases from 0.65 eV in LSDA to 0.44 eV in PBE to 0.09 eV in r2SCAN, indicating improved performance in balancing the description of s and d electrons.
The SIC energy penalty is largest when removing the first 3d electron from a 3d^5 or 3d^10 configuration, leading to severe underestimation of the corresponding ionization energies in FLOSIC-DFT.
Locally scaling the self-interaction correction (LSIC-LSDA) can effectively cancel out the s and d errors, providing accurate ionization energies.

Статистика

The mean errors (ME) and mean absolute errors (MAE) for the first, second, and third ionization energies of the 3d transition metal atoms are reported.

Цитати

"While, from total energy differences, we see that the standard DFAs (LSDA, PBE, and r2SCAN) overbind the 4s electron in the 4s^1 configuration, and underbind the first 4s electron in the 4s^2 configuration, the self-interaction corrected DFAs suffer from a SIC energy penalty for noded d electrons."
"The SIC energy penalty appears largest when an electron is removed from either a half-filled (3d^5) or completely-filled (3d^10) subshell, as evidenced by a severe underestimation of the ionization energy in these cases."

Ключові висновки, отримані з

Ionization energy: sd transfer error and Perdew-Zunger self-interaction correction energy penalty in 3d atoms

by Rohan Maniar... о arxiv.org 10-03-2024

https://arxiv.org/pdf/2409.07438.pdf

Глибші Запити

How can the self-interaction correction energy penalty associated with the noded 3d orbitals be further reduced or eliminated?

The self-interaction correction (SIC) energy penalty associated with noded 3d orbitals can be further reduced or eliminated through several strategies. One effective approach is the local scaling of the self-interaction correction, as demonstrated in the LSIC (Locally Scaled Self-Interaction Correction) method. This technique involves adjusting the self-interaction correction based on the local electron density, allowing for full correction in one-electron-like densities while applying smaller corrections in regions of slow-varying density. This method has shown significant improvements in the accuracy of ionization energy predictions, particularly for configurations with highly noded orbitals.
Additionally, employing complex localized orbitals could provide another avenue for mitigating the SIC energy penalty. By utilizing complex orbitals, which are designed to have nodeless densities, one can potentially avoid the issues related to the lobed nature of real localized orbitals. This could lead to a more accurate representation of the electronic structure of transition metal atoms, thereby reducing the self-interaction errors associated with the noded 3d orbitals.

What are the implications of the observed self-interaction correction errors on the prediction of physical processes involving changes in the oxidation state of transition metal atoms?

The observed self-interaction correction errors have significant implications for predicting physical processes that involve changes in the oxidation state of transition metal atoms. These errors can lead to inaccurate ionization energies, which are critical for understanding the stability and reactivity of different oxidation states. For instance, the underestimation of ionization energies for configurations with filled 3d shells (such as 3d5 or 3d10) can result in incorrect predictions of the energy required for oxidation or reduction processes.
Moreover, the SIC energy penalty can affect the calculated energies of transition states and intermediates in redox reactions, potentially leading to erroneous conclusions about reaction pathways and mechanisms. This is particularly relevant in catalysis and materials science, where the oxidation state of transition metals plays a crucial role in determining catalytic activity and material properties. Therefore, addressing these self-interaction errors is essential for improving the reliability of computational predictions in systems involving transition metal oxidation states.

Could the use of complex localized orbitals help mitigate the issues related to the lobedness of the 3d orbitals and the associated self-interaction correction energy penalty?

Yes, the use of complex localized orbitals could significantly help mitigate the issues related to the lobedness of the 3d orbitals and the associated self-interaction correction energy penalty. Complex localized orbitals are designed to have nodeless densities, which can reduce the complications arising from the highly lobed nature of traditional 3d orbitals. By avoiding nodes, these complex orbitals can provide a more accurate representation of the electronic structure, particularly in systems where the lobedness leads to significant self-interaction errors.
Implementing complex localized orbitals may also enhance the performance of density functional approximations (DFAs) by improving the description of electron correlation effects in transition metal systems. This could lead to more accurate predictions of ionization energies and other electronic properties, ultimately reducing the SIC energy penalty associated with noded orbitals. As a result, the application of complex localized orbitals presents a promising direction for future research aimed at refining computational methods for transition metal chemistry.