The unwanted joule heat can be reduced by different underlying principles in superconductors and topological insulators, respectively. On the one hand, the superconductors are endowed with zero resistance to the flow of current by forming Cooper pairs from electrons. However, the pairing mechanism to achieve high operating temperatures is still elusive. On the other hand, in a topological insulator, the resistance to the flow of current is not zero but highly attenuated. A topological insulator is characterized by an insulating bulk bandgap but still guaranteed by band topology to support gapless (conducting) boundary states, which are immune to backscattering via the spin-momentum locking relationship. Although the mechanism for topological insulators is fully understood theoretically, it is practically difficult to avoid structural defects that usually shift the Fermi level of the system out of the working regime. Therefore, a material showing both states via doping tunability may provide a way to optimize the circuit design with highly reduced joule heat and, moreover, to realize more exotic applications, as we will briefly see later.
Monolayer FeSe/STO exhibits a superconducting transition temperature (Tc) of nearly 70 K (ref.2). This is a dramatic enhancement of the Tc of 8 K in bulk FeSe (ref. 3), the electronic structure of which features the presence of both hole pockets around the centre (Γ point) of the Brillouin zone and electron pockets around the corner (M point). Although still a matter of debate, this large change in Tc might be driven by the crystal structure, as illustrated in Fig. 1a. Since there is an in-plane lattice mismatch of ~3.6% with the substrate, the FeSe film is expected to suffer a substantial tensile strain under which the bulk hole pockets around Γ could disappear as observed in photoemission spectra. With these results in mind, Zhengfei Wang and co-workers carried out first-principles band structure calculations for various magnetic orders in FeSe/STO. By comparing their theoretical predictions with the photoemission observations, they identified a checkerboard antiferromagnetic (CAFM) order, as schematically depicted in Fig. 1b, to be a good match, adducing the need for the inclusion of effects of electronic correlations beyond those of just strain to explain the absence of hole pockets. Moreover, taking a hint from a previous study4, they found that by including spin–orbit coupling (SOC) within the CAFM order a small gap (~40 meV) opened at the M point, a feature also seen in the photoemission spectra and shown in Fig. 1c. Importantly, the opening of such a gap leads to band inversion, which was monitored by computing a non-vanishing, symmetry-protected (the sz component of the spin is conserved), bulk topological invariant named the spin Chern number. This was confirmed experimentally via scanning tunnelling spectra, which showed significant differences in local densities of states between the bulk and edge portions of the sample. These results are consistent with the theoretical predictions, and justify the claim of the existence of a topological state in FeSe/STO.
a, The crystal structure of monolayer FeSe on the TiO2-terminated SrTiO3(001) substrate. b, The checkerboard antiferromagnetic (CAFM) order proposed by Zhengfei Wang and collaborators. The yellow arrows represent the magnetic moments of the Fe atoms. c, Theoretical band structures (black lines) of FeSe/STO in CAFM order, including the SOC overlapped with the second derivative of photoemission spectra around M in the Brillouin zone, which is proportional to the spectral weights. a, lattice constant. Figure adapted from ref. 1, Nature Publishing Group.
According to Zhengfei Wang and colleagues, FeSe/STO behaves like a superconductor with electron doping ('n-type') but is expected to achieve an antiferromagnetic topological insulator state with extra hole doping (until the Fermi level shifts to an energy position within the gap, dubbed p-type). These features should enable the realization of Majorana fermions (zero modes), which are their own anti-particles, with enormous potential for quantum computing applications. A proposal5is sketched in Fig. 2, where a region in the topological insulating state is in proximity of a region in the superconducting state. The underlying idea is simple: in order to make fermions 'real', the particle–hole symmetry of a superconductor erases the distinction between positive and negative charges; the spin-momentum locking in a topological insulator then makes electrons effectively spinless and eliminates the energy difference from zero-point motion6. Such a configuration could be fabricated within a single FeSe layer as a p–n junction by suitable gating. With ferromagnetic insulators to confine the two Majorana modes at both ends, a two-level state will be formed as a basic quantum qubit, which can non-locally encode useful information.
Figure 2: A schematic of a FeSe/STO-based p–n-junction device for potentially realizing Majorana fermions.
p-type and n-type doping of FeSe/STO via suitable gating could lead to an antiferromagnetic topological insulating state (AFM-TI) and a superconducting state (SC), respectively, and could generate a 1D topological superconductor. Two Majorana zero modes should then be confined by the ferromagnetic (FM) regions in red.
The work of Zhengfei Wang and colleagues, combining theory and experiments, provides a new route for integrating topological and superconducting properties, without appealing to the topological superconductor state directly. Despite this progress, however, more work is needed to definitively pin down the topological phase in FeSe/STO. In particular, the magnetic order, which has been indirectly inferred from first-principles calculations, should be verified more directly via measurements. Also, the coexistence of the antiferromagnetism and superconductivity may need further confirmation. The existence of symmetry-protected edge states and their robustness to symmetry-preserving perturbations should be established via transport experiments. Nevertheless, the study by Zhengfei Wang and co-workers should encourage follow-up work on superconducting materials with non-trivial topological properties, and new strategies for building devices for low-power-consuming spintronic/electronic applications.
2016-07-04, Nature
