Physical Review Letters reports Prof. Yuan Li and collaborators’ study of superconducting mechanism in the iron-based Sr1-xNaxFe2As2 compound

Unraveling the mechanism of high-temperature superconductivity in magnetic materials is one of the most important outstanding quests in condensed matter research. Electrons that carry spin moments are usually considered to be localized on magnetic ions, and electrons that carry electric currents are itinerant on the lattice. But in fact they are the same particles, so philosophically, it is an intriguing question how the two sides of the same coin work together towards the formation of superconductivity. Such “unconventional” mechanisms are considered to be relevant to several important classes of materials including the cuprates, the iron pnictides and chalcogenides, and the rare-earth-based heavy-fermion compounds.

In a multi-orbital system such as the iron-based materials, interplay between electrons’ spin and orbital degrees of freedom adds another layer of complexity to the problem. As has been revealed in a series of previous studies [1,2] by Prof. Yuan Li’s group and collaborators, spin-orbit coupling (SOC) plays an important role in shaping the materials’ magnetic properties. Those experiments also suggested a somewhat hybrid nature of the magnetism, in that some of the observations were best explained by considering itinerant as well as localized spins. This was not totally surprising, since theoretical calculations had shown that the iron-based materials can be described as “Hund’s metals” [3]. However, how the effect of SOC plays out in the superconducting mechanism had remained unclear so far.

Now, based on time-of-flight neutron spectroscopy, Prof. Yuan Li’s group and collaborators have found evidence that a certain type of magnetic excitations is particularly important for the formation of superconductivity in an iron-based compound due to SOC effects [4]. The work has been published in Physical Review Letters on Jan. 4^th, 2019.

The study was enabled by the recent discovery of a novel tetragonal magnetic phase in some hole-doped “122” iron pnictides. Unlike the more common orthorhombic magnetic phase, magnetic moments in the tetragonal magnetic phase are primarily along the crystallographic c axis (rather than lying in the ab-plane). In the presence of the tetragonal magnetic phase, the superconducting critical temperature (T_c) is strongly suppressed. The idea behind the study was to correlate the suppression of T_c to spectroscopic characteristics of the tetragonal magnetic phase. The research team chose to study Sr_1-xNa_xFe₂As₂, which has a robust tetragonal magnetic phase, and for which large single crystals can be grown.

Indeed, it is found that the transition from the orthorhombic magnetic phase (Fig. 1a, T = 80 K) to the tetragonal magnetic phase (Fig. 1b, T = 20 K) brings a clear change in the spin excitation spectrum at low energies. Here, due to geometric dependence of the neutron-scattering cross section, the signals at L = 1 detect mostly c-axis excitations of the magnetic moments, whereas signals at L = 3 detect more ab-plane excitations. The fact that the c-axis excitations are suppressed at 20 K compared to 80 K (Fig. 1d) is exactly because of the spin reorientation in the tetragonal magnetic phase – when the static magnetic moments are along the c axis, they are no longer able to change their projection along c in the low-energy excitations. The key result is the following: upon entering the superconducting state (Fig. 1c, T = 6 K, compared to Fig. 1b), a weak change is seen in the excitation spectrum, reflecting the fact that the magnetism is related to the superconductivity, and the change is only visible at L = 1 (Fig. 2). Therefore, the c-axis excitations are suppressed in the tetragonal magnetic phase, but they are actually what is most related to the superconductivity. This result in turn suggests a “compatibility requirement” for realizing high-temperature superconductivity in a multi-orbital Hund’s metal: the local moments must be able to provide the right type of excitations needed by the itinerant electrons. The tetragonal magnetic phase of Sr_1-xNa_xFe₂As₂ is against this requirement, hence the T_c is low.

ICQM Ph.D. students Jianqing Guo and Li Yue are the first authors of this work. The neutron scattering experiment was performed at the MLF, J-PARC, Japan, under a user program.  The work was done in collaboration with ICQM member Prof. Yan Zhang’s group. The project was financially supported by NSFC and MOST.

References:

[1] C. Wang et al., Phys. Rev. X 3, 041036 (2013).

[2] M. Ma et al., Phys. Rev. X 7, 021025 (2017).

[3] Z.P. Yin et al., Nat. Mater. 10, 932 (2011).

[4] J. Guo, L. Yue et al., Phys. Rev. Lett. 122, 017001 (2019).

Figure 1. (a-c) Imaginary part of dynamic spin susceptibility measured at different temperatures. (d) Imaginary part of dynamic spin susceptibility integrated over 4-8 meV based on the data in (a) and (b).

Figure 2. (a-b) Constant-Q cuts measured at (0.5, 0.5, 1) and (0.5, 0.5, 3), with background subtracted.  (c-d) Intensity difference between 6 K and 20 K at L = 1 and 3.