Determination of magnetic exchange couplings in KS-DFT: decomposition, recomposition method and the role of the spin decontamination

Grégoire David
Univ Rennes, CNRS, ISCR (Institut des Sciences Chimiques de Rennes), Rennes, France

Mercredi 9 Novembre 2022, 11h00



Molecules with remarkable magnetic properties, such as Single-Molecule Magnets, have been attracting much attention over the last decades for their potential technological applications as elementary building blocks in Quantum Technologies or Molecular Spintronics.1 Indeed, molecules have the advantage of being highly versatile, with a possible design of compounds with predefined magnetic properties. In this aim, compounds with multiple magnetic centres, for instance with f- and d-metal ions or radical ligands, are particularly interesting since the couplings between centres may play a key role in the control of the global magnetic behaviour.2
However, the theoretical evaluation of the magnetic exchange couplings J in these systems is particularly challenging. On the one hand, wave-function theory-based methods are the reference approach but their computational cost is most of the time prohibitive in these very large compounds. On the other hand, it is possible to compute these couplings in Kohn-Sham density functional theory (KS-DFT) by means of the broken-symmetry (BS) approach but this strategy is not exempt of important theoretical questions, and among them, one may mention the problem of the spin contamination which is very often neglected.
Whilst KS-DFT is supposed to provide only a numerical evaluation of J, Ferré et al. proposed a decomposition scheme based on selective relaxations of the orbitals involved in different mechanisms to extract the main physical contributions to the magnetic exchange coupling: the direct exchange between the magnetic orbitals, the kinetic exchange corresponding to the relaxation of the magnetic orbitals and the spin polarisation of the core orbitals.3
In this communication, I will show how the decomposition allows to get valuable insight into the physical mechanisms governing the magnetic exchange couplings in some dinuclear copper compounds4 or heteronuclear complexes.5 In addition to this powerful rationalisation tool, I will present how this strategy, the so-called recomposition method, provide a new route to compute magnetic exchange couplings, solving the inherent problem of the spin decontamination in some diradical organic compounds4 or in multicentre compounds.6 For the latter situation, the method is applied on a cubane-like copper ions-based compound (Figure) and will allow to discuss the importance of the spin decontamination.7

[Picture]

Figure. Cubane-like compound studied [Cu4(N-(2-hydroxyethyl)-1,3-propanediamine)4].





_________________________________

References :
[1] E. Coronado, Nat. Rev. Mater., 5 (2019) 87-104.
[2] J. D. Rinehart, M. Fang, W. J. Evans, and J. R. Long, Nat. Chem., 3 (2011) 538-542.
[3] E. Coulaud, J.-P. Malrieu, N. Guihé,ry, and N. Ferré, J. Chem. Theory Comput., 9 (2013) 3429-3436.
[4] G. David, F. Wennmohs, F. Neese, and N. Ferré, Inorg. Chem., 57 (2018) 12769-12776.
[5] G. Duplaix-Rata, B. Le Guennic, and G. David, in preparation.
[6] G. David, G. Trinquier, and J.-P. Malrieu, J. Chem. Phys., 153 (2020) 194107.
[7] G. David, N. Ferré and B. Le Guennic, submitted.