Pseudogap-less high T c superconductivity in BaCo x Fe 2−x As 2
F. Massee, Y. K. Huang, J. Kaas, E. van Heumen, S. de Jong, R. Huisman, H. Luigjes, J. B. Goedkoop, M. S. Golden
PPseudogap-less high T c superconductivity in BaCo x Fe As F. Massee, Y.K. Huang, J. Kaas, E. van Heumen, S. de Jong,
1, 2
R. Huisman, H. Luigjes, J.B. Goedkoop, and M.S. Golden Van der Waals-Zeeman Institute, University of Amsterdam, 1018XE Amsterdam, The Netherlands SLAC National Accelerator Center, Stanford University, Menlo Park, California 94025, United States
The pseudogap state is one of the peculiarities of the cuprate high temperature superconduc-tors. Here we investigate its presence in BaCo x Fe As , a member of the pnictide family, withtemperature dependent scanning tunneling spectroscopy. We observe that for under, optimally andoverdoped systems the gap in the tunneling spectra always closes at the bulk T c , ruling out thepresence of a pseudogap state. For the underdoped case we observe superconducting gaps overlarge fields of view, setting a lower limit of tens of nanometers on the length scale of possible phaseseparated regions. With the discovery of the pnictide family of hightemperature superconductors [1], a new window tothe physics behind high temperature superconductivity(HTSC) has been opened. At first glance, the phase di-agrams of the cuprates and the pnictides are strikinglysimilar. In both systems, electron or hole doping sup-presses the magnetic ground state of the parent com-pound and produces a superconducting dome, see for in-stance [2]. However, whereas the cuprates are Mott in-sulators at low dopings and upon doping first completelylose their long range magnetic order before superconduc-tivity emerges, the pnictide parent compounds are met-als and the doped systems can directly cross from a longrange ordered magnetic to a superconducting phase uponcooling. The nature of this transition is still under de-bate, as some studies find that the two regions can coex-ist [3–5] while others indicate a phase separation scenario[6, 7].At low doping concentrations, the cuprates further-more display a second energy scale, characterized in boththe charge and spin sectors by a depression of spectralweight at the Fermi level which is observed both in thesuperconducting and normal state: the so called pseu-dogap [8, 9]. Since the role of the pseudogap phase inthe cuprates is still under heavy debate, the establish-ment of the presence or absence of such a phase in theiron pnictides is of particular importance. There are sev-eral indications that at least some of the pnictides have apseudogapped region at the underdoped side of the phasediagram. For instance, anomalous resistivity characteris-tics in LaFeAsO F x and SmFeAsO F x have been in-terpreted in a pseudogap-like scenario [10]. Nuclear mag-netic resonance (NMR) investigations [11, 12] and an-gle resolved photoemission (ARPES) measurements [13]have also found signs of a pseudogap in the 1111 systems.In Ba K x Fe As ARPES [14], optical conductivity [15]and time resolved, pump-probe spectroscopy [16] mea-surements showed pseudogap-like behavior, and claims ofpossible pseudogap behavior have also been made in theFeSe Te x system [17]. In the case of BaCo x Fe As ,ARPES measurements have reported a slight depression in signal above the superconducting transition tempera-ture [18], possibly caused by a weak pseudogap.A further experimental observation that could belinked to the presence of a pseudogap phase is the largespread in the peak-to-peak separation (2∆ p-p ) seen intunneling spectra of the pnictides [19, 20]. Such a large2∆ p-p variation is difficult to reconcile with the sharpnessof the superconducting transition measured using resis-tive or magnetic means. In underdoped cuprates, theanalogous tunneling spectra are dominated by pseudo-gap features, and the true superconducting quasiparticlespectrum has been resolved only very recently [21]. Thequestion is therefore whether the tunneling spectra of thepnictide HTSC also show a mix of pseudogap and super-conducting gap features.In this Letter we report temperature and doping de-pendent scanning tunneling spectroscopy measurementson underdoped, optimally and overdoped BaCo x Fe As crystals (x=0.08, 0.14 and 0.21). The main goal is todetermine whether all the gaps seen in tunneling experi-ments are indeed superconducting gaps that close at T c ,or whether there are also pseudogapped tunneling signa-tures for which the gap does not close for T > T c . There issome debate in the literature on the nature of the cleav-age surface of the 122-pnictide family [19, 20, 22–26]. Ourview is that, in order to avoid creation of a polar surface,the crystal cleaves in the Ba layer, exposing half a Balayer on either side of the cleave [25–28]. Although acomparison of VUV and hard x-ray photoemission datasuggests that the cleavage surface does not have a domi-nating influence on the global near-surface electronic en-vironment [29], LDA calculations report electronic prop-erties differing from the bulk for √ × √ × √ × √ × > a r X i v : . [ c ond - m a t . s up r- c on ] S e p FIG. 1: Optimally doped BaCo x Fe As (x=0.14, T c = 22 ± c , the results remain unaltered. ficiency. Such regions of the cleavage surfaces were notused for the spectroscopic surveys reported here. TheSTM/STS microscope, which enables us to track the sur-face with atomic precision as a function of temperature,has been described elsewhere [20].We begin by presenting the data from optimally dopedBaCo x Fe As (x = 0.14) samples which have T c = 22 ± c . The U-shaped background of thespectra makes it difficult to follow the temperature de-pendence in a quantitative manner. Consequently, weadopt the common procedure of dividing the differentialconductance at a temperature T, by that taken in thenormal state, i.e. (dI/dV) T /(dI/dV) T normal . The resultof this analysis is plotted in Fig. 1(e)-(h). Here, unlikein the cuprates, the gap spectrum obtained after normal-ization shows the same 2∆ p-p as is apparent in the rawdata in the superconducting state for all four magnitudesof the peak-to-peak gap sampled. The exact temperatureat which the gap closes, or fills-in, is difficult to estimate,but for all gap sizes it is within 1 K of the bulk T c de-termined from resistivity measurements from the samecrystals.To ensure that the choice of normalization temperature used in the procedure described above does not influencethe conclusions, normalization of a spectrum taken at17.5 K to both the 22 K data, i.e. very close to T c ,and those recorded at 24 K are compared in Fig. 1(j).The result of both normalizations is so similar that wecan safely say that the results of the normalization arerobust, as long a T > T c is chosen for the division. Sum-marizing for the optimally doped system, it is thus safeto say that the gaps observed in our STS data are indeedsuperconducting gaps.We now turn our attention to lower doping concentra-tions, where, in analogy with the cuprates, it could beexpected that pseudogap behavior would show up morestrongly, mediated - possibly - by enhanced magnetic cor-relations. Figure 2(a) shows the average of ∼ field ofview on a BaCo x Fe As crystal with x=0.08 (T c = 14 ± N = 70 ± c . FIG. 2: Underdoped BaCo x Fe As (x=0.08, T c = 14 ± ∼ field of view. (b) Spectrumobtained by normalizing the 5 K trace shown in (a) to the 20K data. (c) A collection of single pixel spectra taken along alinesman through the conduction map whose average is shownin (a), illustrating the point to point variation in the spectra.The gap edges are indicated with circles. Again, these observations strongly support the notionthat the tunneling gaps observed at low temperature areindeed superconducting gaps and not gaps showing thesame phenomenology as the pseudogaps in cuprate su-perconductors [21].Since at 20 K the underdoped system is still well withinthe orthorhombic, magnetically ordered state, a certainsuppression of the LDOS around E F could be expecteddue to the gapping of significant parts of the Fermi sur-face observed in quantum oscillation and ARPES ex-periments [30–32]. For a mean field gap linked to thestructural/magnetic ordering temperature, such featurescould be ∼
10 mV away from zero bias. Our STS datashow no signs of such an LDOS suppression, which maybe due either to tunneling matrix element effects or tothe difference in LDOS enhancement at the gap edgesbetween partial gapping in k -space due to back-foldingand hybridization of bands in the magnetically ordered,orthorhombic state and the coherence peaks of the super-conducting state. Additional measurements, for instanceto higher energies, should be performed to address thispoint in a conclusive manner. In any case, the presenceof superconducting gaps over the entire field of view setsa lower limit of tens of nanometers on the length scale ofpossible phase separation between superconducting andnon-superconducting regions of the sample surface.To complete our survey of the phase diagram ofBaCo x Fe As , we turn our attention to the overdopedcompound (x=0.21, T c = 13 ± p-p . Figure3(e)-(h) show the traces after normalization to a normalstate spectrum recorded from the same real-space loca-tion. As was the case for the under and optimally dopedsamples, the gaps seen in the normalized T < T c spectrahave identical magnitude to those in the raw data, andall vanish above the bulk T c .To summarize, for underdoped, optimally doped, andoverdoped BaCo x Fe As (x=0.08, 0.14 and 0.21), thepeak-to-peak gap in the local tunneling density of states,2∆ p-p , has been tracked as a function of location on thesurface and as a function of temperature. The gaps for alldoping concentrations studied and all gap sizes observedvanish at the bulk T c , excluding a cuprate-like pseudogapscenario in these pnictide superconductors. Since super-conducting gaps have been seen across the entire fieldof view of low-corrugation surfaces over areas of hun-dreds of square ˚Angstrom, nano-scale phase separationof non-superconducting (magnetic) and superconductingpatches seems to be unlikely, also in the underdoped ma-terial.The experimental indications of the presence of a pseu-dogap in several compounds [10–17] closely related tothe one studied here makes the family of pnictide su-perconductors a theoretical challenge to understand, asthe action of specific dopant atoms clearly can lead tocompletely differing behavior. For the Ba122 systems,one could draw a parallel with the cuprates, in whichthe electron-doped systems seem less prone to pseudo-gap physics than the p-type [33], although the differentbehavior for the 1111 pnictide systems again underlinesthat subtle differences in doping route and crystal chem-istry can lead to large differences in the observed physicalphenomenology. Consequently, our new data presentedhere set a first step with this clear result in the Co-dopedBa122 system. Mapping out the presence or absence ofa pseudogap in other related pnictide compounds - bothelectron and hole doped - will be an important next stepin gaining a better understanding of the role of the pseu-dogap for high temperature superconductivity in the ironand copper-based materials. ACKNOWLEDGEMENTS
We would like to acknowledge J. S. Agema for experttechnical support. This work is part of the research pro-gram of FOM (09PR2657), which is financially supportedby the NWO. [1] Y. Kamihara et al., J. Am. Chem. Soc. , 3296 (2008)[2] S.A. Kivelson and H. Yao, Nature Materials , 927 (2008)[3] M.-H. Julien et al., EPL , 37001 (2009)[4] A.J. Drew et al., Nature Materials , 310 (2009) FIG. 3: Overdoped BaCo x Fe As (x=0.21, T c = 13 ± > T c for the different locations fall perfectly on top of each other, contrary to thedata for T < T c . Panels (e)-(h) show the spectra (a)-(d) normalized to the respective spectra taken at 17 K, i.e. in the normalstate.[5] J.S. Kim et al., J. Phys.: Condens. Matter , 102203(2009)[6] J.T. Park et al., Phys. Rev. Lett. 102, 117006 (2009)[7] S. Takeshita and R. Kadono, New Journal of Physics ,035006 (2009)[8] M. Buchanan, Nature , 8 (2001)[9] T. Timusk and B. Statt, Rep. Prog. Phys. , 61 (1999)[10] C. Hess, A. Kondrat, A. Narduzzo, J.E. Hamann-Borrero, R. Klingeler, J. Werner, G. Behr and B.B¨uchner, EPL , 17005 (2009)[11] Y. Nakai, K. Ishida, Y. Kamihara, M. Hirano and H.Hosono, J. Phys. Soc. Jpn. , 073701 (2008)[12] H.-J. Grafe, D. Paar, G. Lang, N.J. Curro, G. Behr,J. Werner, J. Hamann-Borrero, C. Hess, N. Leps, R.Klingeler and B. B¨uchner, Phys. Rev. Lett. , 047003(2008)[13] T. Sato, S. Souma, K. Nakayama, K. Terashima, K. Sug-awara, T. Takahashi, Y. Kamihara, M. Hirano and H.Hosono, J. Phys. Soc. Jpn. , 063708 (2008)[14] Y.-M. Xu, P. Richard, K. Nakayama, T. Kawahara, Y.Sekiba, T. Qian, M. Neupane, S. Souma, T. Sato, T.Takahashi, H. Luo, H.-H. Wen, G.-F. Chen, N.-L. Wang,Z. Wang, Z. Fang, X. Dai and H. Ding, arXiv:0905.4467(2009)[15] Y.S. Kwon, J.B. Hong, Y.R. Jang, H.J. Oh, Y.Y. Song,B.H. Min, T. Iizuka, S. Kimura, A.V. Balatsky and Y.Bang, arXiv:1007.3617 (2010)[16] D.H. Torchinsky, G.F. Chen, J.L. Luo, N.L. Wang andN. Gedik, Phys. Rev. Lett. , 027005 (2010)[17] L. Craco and M.S. Laad, arXiv:1001.3273 (2010)[18] K. Terashima, Y. Sekiba, J.H. Bowen, K. Nakayama, T.Kawahara, T. Sato, P. Richard, Y.-M. Xu, L.J. Li, G.H.Cao, Z.-A. Xu, H. Ding and T. Takahashi, PNAS ,7330 (2009)[19] Y. Yin, M. Zech, T.L. Williams, X.F. Wang, G. Wu, X.H.Chen, and J.E. Hoffman, Phys. Rev. Lett. , 097002(2009) [20] F. Massee, Y. Huang, R. Huisman, S. de Jong, J.B. Goed-koop, and M. S. Golden, Phys. Rev. B , 220517(R)(2009)[21] M.C. Boyer, W.D. Wise, K. Chatterjee, M. Ji, T. Kondo,T. Takeuchi, H. Ikuta and E.W. Hudson, Nature 3, 802(2007)[22] M.C. Boyer, K. Chatterjee, W.D. Wise, G.F. Chen,J.L. Luo, N.L. Wang and E.W. Hudson, arXiv:0806.4400(2008)[23] D. Hsieh, Y. Xia, L. Wray, D. Qian, K. Gomes, A. Yaz-dani, G.F. Chen, J.L. Luo, N.L. Wang and M.Z. Hasan,arXiv:0812.2289 (2008)[24] V.B. Nascimento, A. Li, D. R. Jayasundara, Y. Xuan, J.ONeal, S. Pan, T.Y. Chien, B. Hu, X.B. He, G. Li, A.S. Sefat, M.A. McGuire, B.C. Sales, D. Mandrus, M.H.Pan, J. Zhang, R. Jin and E.W. Plummer, Phys. Rev.Lett. , 076104 (2009)[25] F. Massee, S. de Jong, Y. Huang, J. Kaas, E. vanHeumen, J.B. Goedkoop and M.S. Golden, Phys. Rev.B , 140507(R) (2009)[26] H. Zhang, J. Dai, Y. Zhang, D. Qu, H. Ji, G. Wu, X.F.Wang, X.H. Chen, B. Wang, C. Zeng, J. Yang, and J.G.Hou, Phys. Rev. B , 104520 (2010)[27] M. Gao, F. Ma, Z.-Y. Lu and T. Xiang, Phys. Rev. B , 193409 (2010)[28] E. van Heumen, J. Vuorinen, K. Koepernik, F. Massee,Y. Huang, M. Shi, J. Klei, J. Goedkoop, M. Lindroos, J.van den Brink, M.S. Golden, arXiv: 1009.3493.[29] S. de Jong, Y. Huang, R. Huisman, F. Massee, S. Thiru-pathaiah, M. Gorgoi, F. Schaefers, R. Follath, J. B.Goedkoop, and M. S. Golden, Phys. Rev. B , 115125(2009)[30] S.E. Sebastian, J. Gillet, N. Harrison, P.H.C. Lau, D.J.Sing, C.H. Mielke and G.G. Lonzarich, J. Phys.: Con-dens. Matter Fisher, Phys. Rev. B , 064507 (2009)[32] S. de Jong, E. van Heumen, S. Thirupathaiah, R.Huisman, F. Massee, J.B. Goedkoop, R. Ovsyannikov,J. Fink, H.A. D¨urr, A. Gloskovskii, H. S. Jeevan, P.Gegenwart, A. Erb, L. Patthey, M. Shi, R. Follath, A. Varykhalov and M.S. Golden, EPL , 27007 (2010)[33] N.P. Armitage, P. Fournier and R.L. Greene, Rev. Mod.Phys.82