Contrast of LiFeAs with Isostructural, Isoelectronic, and Non-superconducting MgFeGe

Abstract

Stoichiometric LiFeAs at ambient pressure is an 18 K superconductor while isostructural, isoelectronic MgFeGe does not superconduct, despite their extremely similar electronic structures. To investigate possible sources of this distinctively different superconducting behavior, we quantify the differences using first principles density functional theory. Total Fe 3d occupations are identical, with individual 3d orbital occupations differing by no more than 0.015. However, a redistribution of bands just above the Fermi level ε F provides an important distinction, with more Fe-derived states within 0.5 eV of the Fermi level and a higher N(ε F ) in MgFeGe. For many mechanisms these features would enhance the tendency toward superconductivity by providing more Cooper pairs (in MgFeGe), but the tendency toward magnetic instability might be more important. Two of the five Fermi surfaces differ between LiFeAs and MgFeGe, but still lead to similar q-dependencies of susceptibilities χ 0 (q) including the familiar broad peak at (π, π). The larger χ 0 (q) in MgFeGe, by 10-15% throughout the zone, leads us to tentatively identify this proximity to magnetic instability as the feature underlying the absence of superconductivity in MgFeGe. Another significant difference is the 2.5% difference of the in-plane lattice constant, positioning LiFeAs as a chemically compressed version of MgFeGe. This has possible significance since certain Fe pnictides display pressure-induced superconductivity.

High spin polarization and the origin of unique ferromagnetic ground state in CuFeSb

“…This is similar to the case in ferromagnetic MgFeGe. 41 Next we calculate the transport spin polarization P t , which is in excellent agreement with the present spinresolved Andreev reflection spectroscopy measurement. Within classical Bloch-Boltzmann transport theory, P t can be defined in terms of spin-dependent current densities, and in general, 42…”

“…[38][39][40] In contrast, N (E F ) is comparable to MgFeGe which is isoelectronic to LiFeAs but non-superconducting similar to CuFeSb. 41 Large number of states at the Fermi level and in the vicinity is conventionally connected to the number of condensed Cooper pairs, which enhance superconductivity. However, for CuFeSb the narrow d band near the Fermi level and very high N (E F ) the magnetic Stoner instability takes over, and the Fe-layer becomes ferromagnetic.…”

Spin dynamics and orbital-antiphase pairing symmetry in iron-based superconductors

“…Within the random phase approximation, the two compounds have very similar spin excitations, with low-energy maximum intensity at q = (1, 0) (ref. 17). The inclusion of both strong electronic correlations and the realistic two-particle vertex function, as done in this study, identifies clear differences between these two compounds.…”

Why MgFeGe is not a superconductor

“…In particular the former is a good superconductor with T c = 18 K 6 , and, like MgFeGe, is a paramagnetic metal in its normal state. Moreover, the electronic structures of both compounds, including their Fermi surfaces, calculated without any account of magnetism (which seems logical, in view of the experimental situation), are nearly identical 5,7 , and so are the calculated non-interacting susceptibilities. By implication, the spin fluctuation spectra in both compounds must be also very close, which seems, at first glance, to invalidate theories ascribing superconductivity in iron pnictides to spin fluctuations.…”