Absence of superconductivity in topological metal ScInAu 2
J. M. DeStefano, G. P. Marciaga, J. B. Flahavan, U. S. Shah, T. A. Elmslie, M. W. Meisel, J. J. Hamlin
AAbsence of superconductivity in topological metal ScInAu J. M. DeStefano, G. P. Marciaga, J. B. Flahavan, U. S. Shah, T. A. Elmslie, M. W. Meisel,
1, 2 and J. J. Hamlin ∗ Department of Physics, University of Florida, Gainesville FL, 32611, USA National High Magnetic Field Laboratory, University of Florida, Gainesville, FL 32611-8440, USA (Dated: December 14, 2020)The Heusler compound ScInAu was previously reported to have a superconducting ground statewith a critical temperature of 3 . β -PdBi .In an effort to explore the interplay between the superconducting and topological properties prop-erties, electrical resistance, magnetization, and x-ray diffraction measurements were performed onpolycrystalline ScInAu . The data reveal that high-quality polycrystalline samples lack the super-conducting transition present samples that have not been annealed. These results indicate the earlierreported superconductivity is non-intrinsic. Several compounds in the Au-In-Sc ternary phase space(ScAu , ScIn , and Sc InAu ) were explored in an attempt to identify the secondary phase respon-sible for the non-intrinsic superconductivity. The results suggest that elemental In is responsible forthe reported superconductivity in ScInAu . I. INTRODUCTION
Many recent studies in condensed matter physics andmaterials science have been focused on the investigationof symmetry-protected topological states [1, 2]. On top ofthe initial efforts to identify and classify different topolog-ical states, increasing efforts have been spent on exploringthe interplay between these states and other electronicand magnetic phases [3, 4]. One such avenue of partic-ular interest is materials systems exhibiting both non-trivial topological states and superconductivity [5, 6].These compounds are candidates for being realized astrue topological superconductors which are predicted tohost Majorana fermions.One such candidate, the 5 . β -PdBi , attracted attention when it was found to havetopologically non-trivial surface states [7]. Ensuing re-search of the compound revealed a variety of interest-ing properties including complex spin textures [8] and apossible spin-triplet order parameter [9, 10]. Further-more, spectroscopic measurements on thin films of β -PdBi were claimed to have shown evidence of non-trivialsuperconductivity and Majorana fermions [11]. However,other measurements have shown that the topological sur-face states likely play no role in the compound’s bulksuperconductivity [12, 13]. Clearly, it would be inter-esting to compare these results to those for a differentcompound with a similar combination of superconduct-ing and topological properties.The search for candidate materials with certain com-binations of properties has recently been facilitated bythe accessibility of new databases of both experimentaland computationally predicted properties. In this case,we searched for materials that exhibited an intersectionof two properties: 1. Previous experimental reports ofsuperconductivity, and 2. Computational prediction of ∗ Corresponding author: jhamlin@ufl.edu a topologically non-trivial band structure. The list ofexperimental T c values was taken from the SuperCondatabase [14]. Topological classification for these com-pounds were obtained from the the Topological QuantumChemistry Project [15–17]. The compound ScInAu wasamong a small number of materials that indicated super-conductivity at readily accessible temperatures (above ∼ further. The topological classification “topo-logical insulator - split electronic band representation” isthe same as that for β -PdBi [15–17].Given the facts above, we thus sought to characterizethe potential interplay of superconductivity and topo-logical properties in ScInAu . Polycrystalline ScInAu was synthesized via arc-melting. Annealing the samplesyielded nearly single phase ScInAu that displayed nosuperconducting transition down to 1 . with a critical tempera-ture of 3 K. Measurements reveal that only unannealedsamples present the previously reported superconductingtransition at 3 K, though the shielding in the magneticsusceptibility is incomplete. These results indicate thatScInAu is not superconducting down to 1 . T c ) of 3 K islikely due to a secondary phase. Based on these resultsseveral other compounds in the Au-In-Sc system wereprobed in search of a potential superconducting phasethat could explain the partial shielding of unannealedScInAu leading to the conclusion that elemental indiumis responsible. II. METHODS
Arc-melted samples were prepared by combining theraw elements in stoichiometric ratios and melting on awater-cooled copper hearth under Ar atmosphere. Each a r X i v : . [ c ond - m a t . s up r- c on ] D ec S c I n A u D a t a - a n n e a l e d B a c k g r o u n d F i t
R e s i d u a l
Intensity (arb. units) q ( d e g r e e s ) * FIG. 1. Top: XRD pattern of unannealed ScInAu with ticksindicating expected peaks of In, ScAu , and ScInAu . Bot-tom: XRD pattern of annealed ScInAu . The small residualindicates that a nearly single-phase sample of ScInAu wasgrown. A small impurity peak is marked with an asterisk. sample was melted multiple times, and was flipped inbetween each melting to ensure homogeneity throughoutthe boule. Samples that did not contain In had negligiblemass loss, whereas samples containing In showed masslosses around 3%. In order to compensate for this, extraIn was added and the samples were arc-melted again untilthe mass of the sample indicated the correct stoichiome-try had been reached. The samples were then annealedwhile wrapped in Ta foil under partial Ar atmospheres.The crystal structures were characterized with powderx-ray diffraction (XRD) using a Siemans D500 diffrac-tometer or a Panalytical X’Pert Pro diffractometer, andRietveld refinements using GSAS-II [19] yielded latticeparameters consistent with those given in literature foreach compound unless otherwise noted. Electrical trans-port and magnetization measurements were performedin Quantum Design PPMS and MPMS systems respec-tively, at temperatures down to ∼ III. EXPERIMENTAL RESULTSA. ScInAu Polycrystalline samples of ScInAu were synthesizedvia arc-melting. Samples were measured both before and after annealing at 700 ° C for three days. Figure 1 presentsXRD data for both the annealed and unannealed sam-ples. While the unannealed sample shows a mixtureof phases, including ScInAu , ScAu , and In, the an-nealed data indicates nearly single phase ScInAu . Theannealed sample presents a single unidentified impuritypeak near 34 ° (marked with an asterisk). Electrical resis-tivity measurements performed on the annealed sample(Fig. 2) show metallic behavior from room temperaturedown to the base temperature of 1 . . The data showa clear drop in the susceptibility beginning slightly be-low 3 K. At the base temperature of 2 K the transitionis still incomplete but has reached a shielding fraction ofmore than 50%. In order to estimate the shielding frac-tion, we included the demagnetization correction of theroughly spherical sample. The substantial shielding indi-cates that the secondary phase likely comprises a sizablefraction of the total sample volume. Hence, the XRDdata suggests that either In or ScAu is responsible. Ameasurement of the magnetization vs field at 2 K (in-set of Fig. 3) indicates H c ∼
40 Oe and complete fluxexpulsion by (cid:46)
150 Oe. The critical field of In at 2 Kis only 180 Oe, which is roughly consistent with our ob-servations [20]. The low critical field indicates that thesuperconducting impurity is almost certainly unreactedelemental indium ( T c = 3 . T c observedhere is somewhat lower that that of indium ( ∼ . ) thathad not previously been measured at low temperaturesin order to determine if they could instead be responsi-ble for the superconductivity observed in the unannealedsample. B. ScAu Arc melted and annealed samples of ScAu showdiffraction patterns that matched the expected MoSi -type structure [21]. Electrical resistivity measurementspresent metallic behavior with a residual resistivity ratio(RRR) of ∼
50. No evidence for superconductivity is de-tected down to 1 . ∼
10 K could be due to a Kondo effectarising from magnetic impurities. r ( mW cm) T ( K )S c I n A u a n n e a l e d r ( mW cm) T ( K )
FIG. 2. Resistivity versus temperature of ScInAu down to1.8 K. No indication of superconductivity is observed. shielding fraction (%) T ( K )
S c I n A u Z F C F C H = 1 0 O e
M (103 emu)
H ( O e )
T = 2 K
FIG. 3. Shielding percentage versus temperature on unan-nealed ScInAu . The incomplete shielding suggests that animpurity phase is responsible. The inset shows the magneti-zation as a function of applied field. Very small fields of order100 Oe are sufficient to suppress the superconductivity. C. ScIn Single crystals of ScIn were grown with the moltenflux method: 80:20 atomic % In:Sc were heated in an alu-mina crucible sealed in a quartz ampule under 70 torr Argas to 1000 ° C and then cooled to 400 ° C over 240 hours.After holding at this temperature of 8 hours, the ampulewas centrifuged to remove the flux. This revealed small,cubic crystals, confirmed by xray diffraction to be cubicScIn [22].Magnetic measurements on samples yielded a diamag-netic signal with an onset of around 3 K, but the shield-ing fraction of order 1%. Furthermore, a magnetic field r ( mW cm) T ( K )S c A u r ( mW cm) T ( K )
FIG. 4. Electrical resistivity versus temperature for ScAu measured from 1.8 to 400 K. The sample is non supercon-ducting in this temperature range. of 0 .
05 T removed this feature. Both of these facts in-dicate that the superconductivity is not intrinsic to theScIn but is due to droplets of In flux on the surfacesof the crystals. Superconducting transitions have beenobserved at 0 .
78 K and 0 .
71 K in YIn and LaIn respec-tively [23], suggesting that ScIn probably becomes su-perconducting below 1 K. D. Sc InAu Samples of Sc InAu were synthesized by arc melt-ing. The tetragonal Mo FeB -type structure [24] wasconfirmed by x-ray diffraction, though some unidentifiedsecondary phases were present. Nonetheless, magneticsusceptibility measurements from 2-300 K presented noevidence for superconductivity or any other anomalies. IV. CONCLUSIONS
The previously reported superconducting behavior ofScInAu , a material that shares the same topological clas-sification as β -PdBi , has been re-analyzed. These mea-surements suggest that ScInAu is not intrinsically su-perconducting, but that unannealed samples can exhibitpartial superconducting shielding in the magnetic suscep-tibility due to a secondary phase - most likely unreactedindium. We also investigated the possibility that anotherphase is responsible for the superconductivity in unan-nealed samples of ScInAu . Queries were performed withthe Materials Platform for Data Science [25] and the Su-perconducting Material Database [14] to search for com-pounds in the Au-In-Sc family that are reported to be su-perconducting. However, no other phases with reports of T c ∼ , ScIn , andSc InAu and found that they are all essentially non-magnetic non-superconducting metals with no anomaliesin the resistivity or magnetic susceptibility down to 2 K.With the existence of large databases of experimentaland computational properties, the search for materialswith certain combinations of properties is now straight-forward. In this case we identified an inaccuracy in therecord - ScInAu is non-superconducting, though it hadpreviously been reported to have T c = 3 K [18]. How- ever, it is clear that there are a large number of knownsuperconducting materials with non-trivial band struc-tures awaiting further study. ACKNOWLEDGEMENTS
Work on this project was supported, in part, bythe National Science Foundation (NSF) via CAREERaward DMR-1453752 (JJH), REU Program DMR-1852138 (GPM), DMR-1708410 (MWM), and DMR-1644779 (NHMFL), and the State of Florida. We thankG. R. Stewart for helpful conversations. [1] C.-K. Chiu, J. C. Teo, A. P. Schnyder, and S. Ryu, Classi-fication of topological quantum matter with symmetries,Reviews of Modern Physics , 035005 (2016).[2] H. C. Po, A. Vishwanath, and H. Watanabe, Symmetry-based indicators of band topology in the 230 spacegroups, Nature Communications , 50 (2017).[3] L.-L. Wang, N. H. Jo, B. Kuthanazhi, Y. Wu, R. J. Mc-Queeney, A. Kaminski, and P. C. Canfield, Single pair ofWeyl fermions in the half-metallic semimetal EuCd As ,Physical Review B , 245147 (2019).[4] L. Ye, M. Kang, J. Liu, F. von Cube, C. R. Wicker,T. Suzuki, C. Jozwiak, A. Bostwick, E. Rotenberg, D. C.Bell, L. Fu, R. Comin, and J. G. Checkelsky, MassiveDirac fermions in a ferromagnetic kagome metal, Nature , 638 (2018).[5] M. Sato and Y. Ando, Topological superconductors: a re-view, Reports on Progress in Physics , 076501 (2017).[6] X.-L. Qi and S.-C. Zhang, Topological insulators andsuperconductors, Reviews of Modern Physics , 1057(2011).[7] M. Sakano, K. Okawa, M. Kanou, H. Sanjo, T. Okuda,T. Sasagawa, and K. Ishizaka, Topologically protectedsurface states in a centrosymmetric superconductor β -PdBi , Nature Communications , 8595 (2015).[8] T. Xu, B. T. Wang, M. Wang, Q. Jiang, X. P. Shen,B. Gao, M. Ye, and S. Qiao, Nonhelical spin texture inthe normal states of the centrosymmetric superconductor β -PdBi , Physical Review B , 161109 (2019).[9] K. Iwaya, Y. Kohsaka, K. Okawa, T. Machida, M. S.Bahramy, T. Hanaguri, and T. Sasagawa, Full-gap su-perconductivity in spin-polarised surface states of topo-logical semimetal β -PdBi , Nature Communications ,976 (2017).[10] Y. Li, X. Xu, M.-H. Lee, M.-W. Chu, and C. L. Chien,Observation of half-quantum flux in the unconventionalsuperconductor β -Bi Pd, Science , 238 (2019).[11] Y.-F. Lv, W.-L. Wang, Y.-M. Zhang, H. Ding, W. Li,L. Wang, K. He, C.-L. Song, X.-C. Ma, and Q.-K. Xue,Experimental signature of topological superconductivityand Majorana zero modes on β -Bi Pd thin films, ScienceBulletin , 852 (2017).[12] L. Che, T. Le, C. Q. Xu, X. Z. Xing, Z. Shi, X. Xu,and X. Lu, Absence of Andreev bound states in β -PdBi probed by point-contact Andreev reflection spectroscopy,Physical Review B , 024519 (2016). [13] P. K. Biswas, D. G. Mazzone, R. Sibille, E. Pom-jakushina, K. Conder, H. Luetkens, C. Baines, J. L.Gavilano, M. Kenzelmann, A. Amato, and E. Moren-zoni, Fully gapped superconductivity in the topologicalsuperconductor β -PdBi , Physical Review B , 220504(2016).[14] M. Tanifuji, A. Matsuda, and H. Yoshikawa, MaterialsData Platform - a FAIR System for Data-Driven Mate-rials Science, in (2019) pp. 1021–1022.[15] B. Bradlyn, L. Elcoro, J. Cano, M. G. Vergniory,Z. Wang, C. Felser, M. I. Aroyo, and B. A. Bernevig,Topological quantum chemistry, Nature , 298 (2017).[16] T. Zhang, Y. Jiang, Z. Song, H. Huang, Y. He, Z. Fang,H. Weng, and C. Fang, Catalogue of topological elec-tronic materials, Nature , 475 (2019).[17] A. Jain, S. P. Ong, G. Hautier, W. Chen, W. D. Richards,S. Dacek, S. Cholia, D. Gunter, D. Skinner, G. Ceder,and K. A. Persson, Commentary: The Materials Project:A materials genome approach to accelerating materialsinnovation, APL Materials , 011002 (2013).[18] B. Matthias, E. Corenzwit, J. Vandenberg, H. Barz,M. Maple, and R. Shelton, Obstacles to superconductiv-ity in CsCl phases, Journal of the Less Common Metals , 339 (1976).[19] B. H. Toby and R. B. Von Dreele, GSAS-II : the gene-sis of a modern open-source all purpose crystallographysoftware package, Journal of Applied Crystallography ,544 (2013).[20] R. W. Shaw, D. E. Mapother, and D. C. Hopkins, Criti-cal fields of superconducting tin, indium, and tantalum,Phys. Rev. , 88 (1960).[21] A. E. Dwight, J. W. Downey, and R. A. Conner, SomeC11b-type compounds of Sc, Y and the lanthanides withCu, Ag and Au, Acta Crystallographica , 745 (1967).[22] E. Parth´e, D. Hohnke, W. Jeitschko, and O. Schob,Structure data of new intermetallic compounds, Natur-wissenschaften , 155 (1965).[23] R. Sharma, G. Ahmed, and Y. Sharma, Intermediate cou-pled superconductivity in yttrium intermetallics, PhysicaC: Superconductivity and its Applications , 1 (2017).[24] F. Hulliger, On tetragonal M Au In and related com-pounds, Journal of Alloys and Compounds , 160(1996). [25] E. Blokhin and P. Villars, The PAULING FILE Projectand Materials Platform for Data Science: From Big DataToward Materials Genome, in