Iron-based Materials

Iron-based superconductors

Fe-based superconductors appear in “1111”, “122”, “111” and “11” type crystal structures which all have the common motif of edge-sharing FeAs or FeSe tetrahedra. Within the “122” type two substitution series EuFe_2-xT_xAs₂ (T = transition metal), containing the strongly magnetic Eu²⁺ species, were investigated. The mother compound EuFe₂As₂ has an antiferromagnetic ground state which is gradually suppressed with substitution of Fe by Co or by Ru. The onset of superconductivity is reported for Co substitution at x » 0.2 and for Ru substitution around x = 0.5. Our investigations together with EPR methods (in co-operation with the University Augsburg) and DFT calculations for the Ru series [1] give a clear picture of a strong reduction of the electronic density of states with increasing substitution with nominally isovalent Ru. This reduction is equally due to structural effects as well as to the direct change of the transition-metal states. Interestingly, the presence of Eu²⁺ spins (S = 7/2) has almost no effect on the electronic and magnetic properties of the Fe(Co)As layers. Therefore, the evolution of magnetic fluctuations of the layers can be successfully probed by Eu EPR even in such a concentrated system [2]. Further, our group contributed to the investigation of the magnetic and superconducting properties of Fe_1+xSe materials [3,4].

Non-magnetic analogs to Fe-based superconductors

Iron-based pnictogen and chalcogenide superconductors often have “non-magnetic” analogs which are superconductors that crystallize in the same structure but consist of only non-magnetic elements. Often these analogs have a much lower superconducting transition temperature T_c, providing support for the argument that the magnetic element, viz. iron, is important for the enhancement of the T_c [5]. For example, SnO crystallizes in the same crystal structure as FeSe, has a similar Fermi surface and exhibits the same pressure dependence of T_c, albeit with a much lower T_c [5]. The analysis of SnO showed that the Fermi surface topology and the degree of nesting are important for the superconductivity, but that spin fluctuations are not essential for the appearance of superconductivity in the “11” structure type; rather, the spin fluctuations increase the coupling and thereby the T_c.

In the case of LiFeAs, the superconductor NaAlSi (T_c= 7 K) [6] can be viewed as its non-magnetic analog. The Fermi surfaces of the two compounds have been shown to have similarities [7]. Our investigations of NaAlSi under high pressure [8] showed that T_c initially increases with p and is then suppressed rather quickly beyond p ≈ 5 GPa. This behavior cannot be attributed to a structural phase transition as proven by x-ray diffraction experiments under pressure. Although pressure has a strong effect on superconductivity, DFT calculations demonstrate that it does not significantly alter the electronic structure. Similarly, a comparison of NaAlSi with non-superconducting NaAlGe showed, that the electronic structure cannot explain the different behavior regarding superconductivity. The fact that the density of states does not change around the Fermi level enforces the idea of a non-BCS model for NaAlSi but does not prove it. Further experimental and theoretical work will be required to determine what the dominant factors are in determining T_c in these analogs of the “111” pnictide superconductors.

In a different project [9], we observed bulk superconductivity with T_c = 0.6 K in the intermetallic compound HfCuGe₂ (Fig. 1) which is structurally related to the “1111” Fe-based superconductors but contains only non-magnetic elements. These findings indicates that superconductivity tends to run in certain structure types, and the very low T_c observed supports arguments that the presence of magnetic Fe is important for obtaining enhanced T_c in this family.