arXiv:2603.24467
Second-Order Magnetic Properties in Paramagnetic Molecules From a Current Density Formulation Including Scalar Relativistic Effects
Francesco Ferdinando Summa, Sonia Coriani, Andre Severo Pereira Gomes
Abstract
This work presents the theoretical background for the computation of nuclear magnetic shielding, nuclear hyperfine and magnetizability tensors of paramagnetic molecules, using a magnetically induced current density framework to account for both orbital and spin contributions. The resulting magnetizability tensor is fully consistent with the general Van Vleck formulation, recovering the temperature-dependent Curie contribution through the explicit integration of the magnetically induced spin current density. The methodology proposed herein provides a straightforward computational route that bypasses the complex evaluation of g-tensors and Zero-Field Splitting (ZFS) Hamiltonians. While the theoretical framework is general, we present applications rooted on physically motivated approximations where scalar relativistic effects are incorporated through corrections based on the Zeroth-Order Regular Approximation (ZORA) Hamiltonian within the ground-state spin density. This approach combines a relativistic self-consistent field (SCF) calculation for the ground-state spin density with a non-relativistic, origin-independent current density calculation for the orbital contribution. This hybrid strategy is shown to capture the Heavy-Atom Light-Atom (HALA) effect in 1H and 13C shieldings, particularly in paramagnetic molecular systems containing transition metals up to the 3d series. By restricting the relativistic treatment to the spin density, where scalar relativistic effects are dominant, and neglecting such effects on the orbital contribution of light atoms, this method offers a good compromise between computational efficiency and accuracy for the characterization of large open-shell molecular systems.
AI Summary
The paper addresses computation of nuclear magnetic shielding, hyperfine coupling and magnetizability tensors in paramagnetic molecules using magnetically induced current densities. It derives orbital and spin-current expressions within the zeroth-order regular approximation (ZORA), with thermal averaging over a spin multiplet recovering the Curie contribution proportional to $S(S+1)/T$. The aim is to evaluate these properties without explicitly calculating $g$ tensors or zero-field-splitting Hamiltonians. The implementation combines relativistic ground-state spin densities obtained from Douglas–Kroll–Hess calculations with nonrelativistic, gauge-origin-independent orbital currents. A separate approximate spin–orbit-coupling (SOC) treatment uses field-free noncollinear densities rather than explicit magnetic-field response. Tests cover five 3d metallocenes, rhodocene, a nitroxide radical, and magnetizabilities of $\mathrm{O_2}$, $\mathrm{NO}$ and $\mathrm{NO_2}$. Scalar-relativistic corrections change ligand proton shieldings by up to about 7 ppm and carbon shieldings by roughly 14–24 ppm in several metallocenes. Comparisons with published X2C calculations give $R^2=0.9999$ for proton shieldings, whereas carbon residuals remain approximately 5–10%, without systematic improvement from scalar corrections. Calculated gas-phase magnetizabilities are 3379.3, 1326.9 and 1259.3 ppm cm$^3$ mol$^{-1}$, respectively, versus experimental values of 3330, 1461 and 1249. Stated limitations include neglected relativistic orbital corrections, picture-change effects and explicit spin-dependent field response. The SOC ansatz lacks formal rigor and gives unacceptable accuracy for rhodocene; pseudocontact shifts are omitted. Experimental agreement is mixed, with particularly large discrepancies for manganocene, including the wrong sign of its proton chemical shift. Full coupled-perturbed relativistic response is deferred.
Central claim may not hold: The claimed general Van Vleck consistency and avoidance of g/ZFS calculations are established only for a restricted spin-multiplet response, while the implementation neglects rather than reproduces general SOC- and ZFS-dependent magnetic response. The numerical evidence supports conditional accuracy rather than uniform reliability through the 3d series, with manganocene showing a wrong-sign proton shift and a carbon shift approximately 33.3% above experiment. The proposed spin current has divergence proportional to grad K · curl Q, and neither the atomic argument in the appendix nor the separately computed nonrelativistic orbital current establishes conservation of the implemented molecular hybrid current. The statistical-mechanics appendix confuses fixed-field saturation with zero-field response when concluding that spin shielding and magnetizability vanish at zero temperature.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.