Engineering topological superconductivity in composite systems

7/2/2020 8:47:33 AM

Steph Adams for the Materials Research Laboratory

Illinois Physics Professor Tai-Chang Chiang. Photo by Steph Adams, University of Illinois at Urbana-Champaign
Illinois Physics Professor Tai-Chang Chiang. Photo by Steph Adams, University of Illinois at Urbana-Champaign
Illinois Physics Professor Jim Eckstein. Photo by Steph Adams, University of Illinois at Urbana-Champaign
Illinois Physics Professor Jim Eckstein. Photo by Steph Adams, University of Illinois at Urbana-Champaign
Illinois Physics postdoctoral researcher Joe Hlevyack. Photo by Steph Adams, University of Illinois at Urbana-Champaign
Illinois Physics postdoctoral researcher Joe Hlevyack. Photo by Steph Adams, University of Illinois at Urbana-Champaign
New findings from physicists at the University of Illinois, in collaboration with researchers at The University of Tokyo and others, clarify the physics of coupling topological materials with simple, conventional superconductors.

Through a novel method they devised to fabricate bulk insulating topological insulator (TI) films on superconductor (SC) substrates, the researchers were able to more precisely test the proximity effect, or coupling when two materials contact one another, between TIs and SCs. They found that when the TI film is bulk insulating, no superconductivity is observed at the top surface, but if it is a metal, as in prior work, strong, long-range superconducting order is seen. The experimental efforts were led by physics Professor Tai-Chang Chiang and Joseph Andrew Hlevyack, postdoctoral researcher in Professor Chiang’s group, in collaboration with Professor James N. Eckstein’s group including Yang Bai, Professor Kozo Okazaki’s Lab at The U. of Tokyo, and five other institutes internationally. The findings are published in Physical Review Letters, which has been highlighted as a PRL Editors’ Suggestion.

Materials that are both topological and superconducting, called “topological superconductors,” have attracted great interest. They can host exotic particles called Majorana bound states that are their own antiparticles, which are promising for technological applications in spintronics, quantum gates, and topological quantum computation. Unfortunately, to date, such materials have only been theorized and yet to be experimentally verified. Alternatively, interfacing a TI film with a SC substrate lends itself as a promising route for realizing the topological superconducting phase. Such engineered composite systems offer flexibility in choosing a wide variety of combinations of topological and superconducting components.

“Following their experimental confirmation in 2008, TIs have been widely studied in condensed-matter physics both experimentally and theoretically,” notes Hlevyack. “These materials, which are insulating in the bulk but conducting at the surface, possess topologically-protected surface states that are robust to disorder or impurities in the system.”

The few experiments thus far on TIs coupled to SCs have claimed amazing successes with strong, long-range superconducting order. However, such studies are questionable as the purported TIs are rigorously speaking metals, not insulators, due to heavy doping by nature of the sample fabrication process, which makes the sample conducting.

“This fact led us to question: ‘If a true TI is fabricated, does the TI/SC system still show strong, long-range superconducting order in both topological surface layers?’ This question is crucial for applications, which require a truly isolated topological superconducting sheet on the surface with no interfering bulk carriers,” Hlevyack explains.

The work highlighted here yields that the answer is no. “As shown by ultrahigh-resolution angle-resolved photoemission spectroscopy (ARPES), bulk insulating TI films of bismuth antimony telluride on superconducting niobium substrates exhibited a massive suppression of surface superconductivity, even for TI films that are just two-layers thick, in stark contrast from bulk conducting TI bismuth selenide films on niobium studied previously,” adds Chiang.

Chiang continues: “When the TI film is bulk insulating, no superconductivity is observed at the top surface, but if it is a metal, as in prior work, strong, long-range superconducting order is seen. That is very good, since the comparison between the two cases identifies one mechanism for the superconducting proximity effect in TI/SC systems. Our prior work for bulk conducting bismuth selenide films on niobium showed that the bulk and surface states have the same superconducting gap, suggesting that these states are quantum-mechanically coupled, and the superconductivity is transferred to the surface by the bulk carriers over a fairly long range.”

To fabricate the insulating TI/SC systems, Hlevyack, Eckstein, and Chiang employed a novel “flip-chip” technique, preparing the systems using molecular beam epitaxy (MBE) and DC magnetron sputtering instruments at the Illinois Materials Research Laboratory. “In this technique, TI films (2-10 layers in thickness) are grown on graphene-terminated silicon carbide or sapphire substrates,” says Eckstein, “The thickness is precisely chosen during the growth, and for the bismuth antimony telluride films, the bismuth-to-antimony ratio is controlled so that all films are bulk insulating regardless of thickness. A polycrystalline niobium film is then sputter-deposited onto the TI films. Lastly, each sample is flipped over, glued onto a sample plate, and then topped with a cleave pin. A fresh TI surface is exposed by striking the cleave pin, which removes the substrate to reveal a TI film of a specific thickness predetermined by the growth on superconducting niobium.”

Eckstein comments, “The flip-chip method is a very powerful preparation technique. It overcomes many of the problems encountered when preparing TIs on niobium, adding greater flexibility in the choice of materials we can have in the system. Also, it allows one to fabricate relatively simple TI/SC systems—particularly those involving simple superconductors such as lead and niobium, which permits one to identify the fundamental physics more easily.”

These “flip-chip” samples were introduced into an ultrahigh-vacuum system and, upon cleavage, were probed by ultrahigh-resolution ARPES at the Institute for Solid State Physics, The University of Tokyo, in the lab of Professor Kozo Okazaki. Chiang and Hlevyack credit their international collaborators at The University of Tokyo for overseeing these main measurements.

Hlevyack stresses, “The project as designed was a very sophisticated, difficult problem. Not only is the preparation of these TI/SC systems very tricky but so are the ARPES  measurements themselves. Detecting superconducting features at such very small energy scales requires an ultrahigh-resolution, ultralow-temperature apparatus. The system used for this research at The University of Tokyo is arguably the best in the world. We are indebted to Professor Okazaki's group for their hard work and for the very nice collaboration.”

“Though we cannot dismiss the possibility of a tiny superconducting gap and thus a weak topological superconductor surface state in the bismuth antimony telluride system, these results are very profound, because they settle the question of whether TIs interfaced with niobium are robust topological superconductor candidates. Our work also underlines the limitations of employing the proximity effect in realizing topological superconductivity. We expect that further research in other TI/SC systems will be conducted to test the generality of our conclusions and to create further optimized TI/SC heterostructures, but for TIs on niobium, the picture seems very simple: No bulk states, then no strong superconducting order on the surface,” concludes Chiang.