- Email: [email protected]
Electronic structure of the unoccupied electron energy states in FeSe1−xTex
Solid State Communications 219 (2015) 48–52
Contents lists available at ScienceDirect
Solid State Communications journal homepage: www.elsevier.com/locate/ssc
Electronic structure of the unoccupied electron energy states in FeSe1 xTex Pramita Mishra a, Himanshu Lohani a, M. Maniraj b, Jayita Nayak b, R.A. Zargar c, V.P.S. Awana d, Sudipta Roy Barman b, Biju Raja Sekhar a,n a
Institute of Physics, Sachivalaya Marg, Bhubaneswar 751005, India UGC-DAE Consortium for Scientific Research, Khandwa Road, Indore 452001, Madhya Pradesh India c Department of Physics, Jamia Millia Islamia, New Delhi-110025, India d CSIR-National Physical Laboratory, K S Krishnan Marg, New Delhi-110012, India b
art ic l e i nf o
a b s t r a c t
Article history: Received 5 May 2015 Received in revised form 26 June 2015 Accepted 29 June 2015 Communicated by Ralph Gebauer Available online 6 July 2015
Inverse photoemission spectroscopic (IPES) measurements along with LDA based band structure calculations have been used to investigate the unoccupied electronic structure of FeSe1 xTex system. The observed doping and temperature dependent pseudogap in this system is found to be linked to the change in the chalcogen height in their geometric structure. The depletion in spectral weight from the near EF states at low temperature in IPES has been correlated with the enhancement of the 3z2–r2 orbitals in the photoemission spectroscopy (PES). The Coulomb correlation energy U, estimated from the combined PES and IPES spectra, signifies the enhancement in electron correlations in FeSe1 xTex, with doping. The formation of pseudogap in PES and IPES confirms the importance of correlations in the 11 family of Fe superconductors. & 2015 Elsevier Ltd. All rights reserved.
Keywords: A. Noncuprate superconductors C. Electronic structure D. Pseudogap E. Inverse photoemission spectroscopy
1. Introduction Since the discovery of superconductivity in Fe pnictides [1], there has been a plethora of research activities to understand the nature of superconductivity in them. Many exotic properties of these superconductors originate from some of the key elements like magnetic fluctuations, unconventional electron–phonon coupling and electron correlations. Further, the multiorbital correlation in combination with Hund's coupling caused by the participation of all five Fe 3d electrons result in some distinctive electronic properties like orbital selectivity [2–4], Mottness [5], quantum phase transition [6], nematicity [7], and pseudogap [8]. The presence of pseudogaps is a characteristic feature of the family of unconventional superconductors such as high Tc cuprates [9,10], heavy fermion superconductors [11] and the recently discovered iron pnictides and chalcogenides [12–15]. In last few years, correlation of the pseudogap with the superconducting state has been under intense study. Although, the origin of pseudogap in these materials is still under debate, a general consensus is prevailing among researchers about the existence of two types of pseudogaps, located at low and high binding energies in high Tc superconductors.
The low energy pseudogap which originates from superconducting fluctuations [16,17] persists just above Tc whereas the origin of high energy pseudogap existing far above Tc is still a puzzle [18,19]. Nevertheless, electron correlation is considered to play a vital role in the existence of the high temperature pseudogap in unconventional superconductors [15,20] like the case of manganites [21]. Electron spectroscopic methods are considered to be quite suited to probe the near EF electronic structure and thereby the nature of pseudogaps. Although a number of photoemission studies [15,22–24] were focussed on the electronic structure of valence band (occupied states) and the pseudogap formation in the Fe(Se,Te) system, only few have been reported on the unoccupied states. In this study, inverse photoemission spectroscopy (IPES) in conjunction with LDA based band structure calculations has been used to study the unoccupied states above EF. The observed temperature dependent pseudogap has been discussed in correlation with its analogue in the valence band reported earlier [15]. Also, from a comparative study of photoemission and inverse photoemission we have estimated the Coulomb correlation energy in Fe(Se,Te) systems.
2. Experimental methods n
Corresponding author. E-mail address: [email protected] (B.R. Sekhar).
http://dx.doi.org/10.1016/j.ssc.2015.06.023 0038-1098/& 2015 Elsevier Ltd. All rights reserved.
Polycrystalline samples of FeSe1 xTex (x¼1, 0.5, 0) used in this study were synthesised via solid state reaction route [25] and
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
15
x=0
A1
x=0
49
10 B1
5
x=1
B1
DOS (states / eV)
Intensity (a.u)
x = 0.5 0 15
x = 0.5
Total Te Se Fe
10 5 0 15
A1
A1 10
Α1′
x=1
Α1′′ B1
5 0 0
2
4
6
8
Energy (eV)
0
2
4
6
Energy (eV)
Fig. 1. (Color online) Panel (a): IPES spectra for FeSe1 xTex (x¼ 1, 0.5, 0 shown in black, red and blue respectively). Dashed vertical lines correspond to the Fermi level. Panel (b): theoretically calculated total unoccupied density of states (DOS). The theoretical features which match with experiment have been marked as A1 and B1 in panels a and b. A1 composed of two sub-features A10 and A1″ , which are not resolved in experimental spectra, is marked in panel b.
characterised for their electrical, magnetic and structural properties [26]. In order to avoid a common source of error in the spectroscopic analysis of these compounds, it was ensured that the samples contain no excess Fe and are of single phase in nature. Inverse photoemission spectroscopic (IPES) measurements were performed in the isochromat mode with a mean photon energy of 9.9 eV using a combination of electrostatically focussed Stoffel Johnson type electron gun with a contact potential [27] and an acetone gas filled photon band pass detector with a CaF2 window [28,29]. The total energy resolution was 600 meV. IPES spectra were obtained by normalising the measured photon counts by the sample current at each energy step. The binding energy has been determined with reference to the Fermi level of a sputter cleaned silver surface. Angle integrated ultraviolet photoemission spectroscopy (UPS) measurements were performed by using another system equipped with a high intensity vacuumultraviolet source and a hemispherical electron energy analyser (SCIENTA R3000). At the He I (hν ¼21.2 eV) line, the photon flux was of the order of 1016 photons/s/steradian with a beam spot of 2 mm in diameter. Fermi energies for all measurements were calibrated by using a freshly evaporated Ag film onto the sample holder. The total energy resolution, estimated from the width of the Fermi edge, was about 27 meV for the He I excitation. Both UPS and IPES measurements were performed at a base pressure of 5.0 10 11 mbar. The samples were repeatedly scraped using a diamond file inside the preparation chamber with a base vacuum of 5.0 10 10 mbar and the spectra were taken within an hour, so as to avoid any surface degradation. All the measurements were repeated many times to ensure the reproducibility of the spectra. The low temperature measurements at 77 K were performed immediately after cleaning the sample surfaces followed by the room temperature measurements. Band structure calculations using TBLMTO-ASA [30] were performed using the experimental lattice parameters reported earlier [26]. The correlation effects of
the Fe-3d orbitals were taken into account by using the LDA þ U formalism with J as 0.9 eV [31] and U as 3.5 eV for FeTe and 4.0 eV for FeSe [32]. For FeSe0.5Te0.5, U was taken as 3.8 eV, a value intermediate between those of FeTe and FeSe.
3. Results and discussion Panel (a) of Fig. 1 shows the inverse photoemission spectra of the FeSe1 xTex samples in the energy range 1 to 9 eV taken at room temperature and panel (b) shows our theoretical band structure calculations using TB-LMTO ASA for the FeSe1 xTex depicting the total unoccupied density of states (DOS). The experimental as well as calculated spectra show two prominent features A1 and B1 located at binding energy positions 1 and 5.5 eV respectively. By comparing these two panels, it can be identified that A1 arises primarily from the unoccupied Fe-3d states with a small contribution from the unoccupied chalcogen (Se/Te) p states and B1 consists predominantly of the hybridised Fe3d–Se4p/Te5p unoccupied states. The feature A1 comprises two sub features A10 and A1″ which are not resolved in the experimental spectra. B1 is less prominent due to the electron scattering in inverse photoemission process [33]. These assignments of features are in agreement with other's calculations also [32,34]. It can be seen from the experimental spectra (panel (a)) that with a decrease in the doping value x, the spectral weight near the Fermi level (EF) decreases and correspondingly the intensity of A1 increases. A similar change was earlier reported by Yokoya et al. [34] also. Such a shift in the spectral weight is a signature of the presence of a pseudogap in the case of FeSe. Panel (b) shows that with a decrease in x value, the feature A10 moves away from the EF while the feature A1″ moves towards the EF in addition to an intensity enhancement of both A10 and A1″ . As a result, the plateau like structure seen in the case of FeTe transforms into a prominent peak. The feature B1 moves to
50
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
Α1′
x=0
Α1′′
A1
Intensity (a.u)
4
2
x = 0.5
Fe 3d PDOS (states / eV)
0
x = 0.5
Total xy yz
4
2
3z -r xz 2
x -y
x=1
2
2
-1
0
1
2
3
Energy (eV) Fig. 3. (Color online) Temperature dependent IPES spectra taken at 300 K (black) and 77 K (red) for FeSe1 x Te1 x (x ¼ 1, 0.5, 0).
2
0
x=1 4
Α1′′
3
Α1′ 2 1 0 -0.5
x=0
0
0.5
1
1.5
2
2.5
Energy (eV) Fig. 2. (Color online) Fe 3d unoccupied partial density of states (PDOS), using LDA þ U, for FeSe1 xTex (x¼ 1, 0.5, 0). The PDOS contribution (panel (b)) for xz (magenta) and yz (blue) orbitals are the same for FeTe and FeSe. For FeSe0.5Te0.5, the xz and yz PDOS split owing to the different bond lengths of Fe–Te and Fe–Se.
higher energy with a decrease in x resulting in a larger separation between the features A1 and B1 in case of FeSe in comparison to FeTe. The depletion of spectral weight near EF with Se substitution has been reported by Saini et al. [35] based on their studies using Xray absorption spectroscopy (XAS) in which the electronic structure of unoccupied states of FeSe1 xTex are probed. This study showed that the Fe-3d unoccupied states display a gradual decrease in spectral intensity with Se substitution. The experimental observation, supported by cluster model calculations, correlates the spectral weight depletion to the enhanced hybridisation of chalcogen p orbitals with the Fe 3d orbitals. The increased chalcogen p–Fe-3d hybridisation for FeSe in comparison to FeTe is also evidenced from the Fe K edge X ray absorption near edge spectra (XANES) [36] which probes the unoccupied Fe–chalcogen hybridised states. Thus, the change in hybridisation strength of the Fe-3d–Se-4p/Te-5p states govern the electronic structure of the FeSe1 xTex system and result in a reduced density of unoccupied states near Fermi level with Se doping. The nature of bands involved in the formation of the pseudogap and in the spectral weight shifts could well be identified from our Fe 3d partial density of states (PDOS) calculations using the TB-
LMTO ASA method. Fig. 2 shows the Fe 3d PDOS of the FeSe1 xTex system. The xz/yz and x2–y2 orbitals are the highest populated ones corresponding to the feature A10 while the xy orbital is the highest for A1″ . The intensity of xy and x2–y2 orbitals could be seen to be increasing with the reduction in the doping value x, leading to the increase in the intensity of A1. Fig. 3 shows the IPES spectra of FeSe1 xTex taken at 300 K (black) and 77 K (red). Quite similar to the spectral changes observed as a function of doping, the intensity of the feature A1 increases for higher temperature. Further, such a change in A1 could be seen for all the three compositions. This shows that the increase in the doping value and increase in the temperature have similar bearings on the spectral weight shifts which lead to the pseudogap. Moreover, such a temperature dependent change also shows that these spectral weight shifts are inherent to the samples and are not artefacts arising from spectral normalisation or sample surface cleaning. In order to understand the origin of the observed temperature dependent pseudogap in IPES which depict the unoccupied states, we have compared our results with the photoemission spectra (PES) representing the occupied states. A comparison between the temperature dependent photoemission spectra taken with He I (21.2 eV) source [15] and the present IPES spectra is shown in Fig. 4. The figure shows an enhancement of states for the feature A in the occupied region ( 0.5 eV – EF energy range) and a complimentary depletion of states for feature A1 in the unoccupied region (EF – 1 eV range) at low temperature. This temperature dependent enhancement in PES was attributed to the change in the chalcogen height (Z) from the Fe plane at low temperature [15]. Reduction in Z at low temperature causes change in the hybridisation between Fe 3d and chalcogen p orbitals. This results in the enhancement of spectral weight for feature A due to the increased occupancy of the 3z2–r2 orbitals at low temperature. The spectral weight in IPES has a direct correspondence to the same in PES. Thus, the enhancement in spectral weight of 3z2–r2 orbitals in the occupied side manifests as depletion of states in the unoccupied side. This explains the observed changes in features A10 and A1″ with temperature. Our result is consistent with the temperature dependent Fe K edge XAS measurements [37], showing an enhancement of pre-peak A, at low temperature in FeSe1 xTex, which has been correlated to the enhanced hybridisation of Fe-3d and chalcogen-p orbitals due to the decreased chalcogen height (h) at low temperature. By combining the spectra from PES and IPES we can estimate some of the energy scales involved in the electronic transport in
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
by using theoretical calculations [32]. The gap in the density of unoccupied states with doping of Se leads to a shifting of the Upper Hubbard band B1 towards higher energy which signifies the enhancement of electron correlation with doping.
x=0 A1
Intensity (a.u)
A
4. Conclusion x = 0.5
x=1
-0.4
0
0.4
0.8
Energy (eV) Fig. 4. (Color online) PES spectrum (solid lines) measured using He I (21.2 eV) below the Fermi level (dashed vertical line) and IPES spectra (dotsþlines) at 9.2 eV photon energy above Fermi level for FeSe1 xTex (x¼ 1. 0.5, 0). The black and red spectra correspond to data taken at 300 K and 77 K respectively.
In conclusion, we have made a systematic study of the unoccupied states in FeSe1 xTex, using IPES and interpreted the results using our theoretical band structure calculations. A doping and temperature dependent pseudogap formation reflecting the role of chalcogen height in this system is observed. The temperature dependency of the pseudogap in the unoccupied region is complementary to the earlier observed pseudogap in the occupied region. We have also estimated the Coulomb correlation energy for different doping concentrations which demonstrate that the electron correlations in Fe(Se,Te) superconductors do have a bearing on the change in their geometric structure.
References [1] [2] [3] [4]
A1
B
x=0
[5] [6]
Intensity (a.u)
[7]
x = 0.5
[8] [9] [10] [11] [12] [13]
x=1
[14] [15] [16] [17]
-4
51
-3
-2
-1
0
1
2
3
Energy (eV) Fig. 5. (Color online) The PES and IPES spectra below and above the Fermi level respectively, for FeTe (black), FeSe0.5Te0.5 (blue) and FeSe (red). The black ticks correspond to the positions of A1 and B.
this family of Fe superconductors, like the onsite Coulomb energy (U). The feature B in the valence band spectra and A1 in the IPES spectra are ascribed to the lower and the upper Hubbard band respectively [31]. The energy separation between B and A1 provides a rough estimate of the onsite Coulomb energy, U [38]. Fig. 5 shows the PES spectra taken using He II (40.8 eV) source and IPES spectra at 9.9 eV photon energy, of the FeSe1 xTex system plotted together by aligning the corresponding Fermi energies. From the plot the value of U was estimated to be 3.5, 3.4, 3.2 eV for FeSe, FeSe0.5Te0.5 and FeTe respectively. The estimated U for FeSe is comparable to the result obtained experimentally by Yokoya et al. [34]. Chen et al. [39] reported a U value of 4.0 using X-ray Absorption Spectroscopy (XAS) and X-ray Emission Spectroscopy (XES). Further, it should be noted that U increases in FeSe1 xTex with decreasing x. Such a trend was earlier predicted
[18] [19] [20]
[21] [22]
[23] [24] [25]
[26] [27] [28] [29] [30] [31] [32] [33] [34]
Y. Kamihara, T. Watanabe, M. Hirano, H. Hosono, J. Am. Chem. Soc. 130 (2008) 3296. T. Hanaguri, S. Niitaka, K. Kuroki, H. Takagi, Science 328 (2010) 474. H. Lohani, P. Mishra, B.R. Sekhar, Physica C 512 (2015) 54. N. Lanatá, U.R.H. Strand, G. Giovannetti, B. Hellsing, dè Medici Luca, M. Capone, Phys. Rev. B. 87 (2013) 045122. dè Medici Luca, G. Giovannetti, M. Capone, Phys. Rev. Lett. 112 (2014) 177001(R). Julián Rinćon, Adriana Moreo, Gonzalo Alvarez, Elbio Dagotto, Phys. Rev. B 90 (2014) 241105(R). J.K. Glasbrenner, I.I. Mazin, Harald O. Jeschke, P.J. Hirschfeld, Roser Valenti, arXiv:1501.04946, 2015. L. Craco, M.S. Laad, arXiv:1001.3273, 2010. T. Timusk, B. Statt, Rep. Prog. Phys. 62 (1999) 61. M.R. Norman, D. Pines, C. Kallin, Adv. Phys. 54 (2005) 715. V.A. Sidorov, M. Nicklas, P.G. Pagliuso, J.L. Sarrao, Y. Bang, A.V. Balatsky, J. D. Thompson, Phys. Rev. Lett. 89 (2002) 157004. Y.S. Kwon, J.B. Hong, Y.R. Jang, H.J. Oh, Y.Y. Song, B.H. Min, T. Lizuka, S.I. Kimura, A.V. Balatsky, Y. Bang, New J. Phys. 14 (2012) 063009. T. Sato, S. Souma, K. Nakayama, K. Terashima, K. Sugawara, T. Takahashi, Y. Kamihara, M. Hirano, H. Hosono, J. Phys. Soc. Jpn. 77 (2008) 063708. D.C. Johnston, Advances in Physics 59 (2010) 803. P. Mishra, H. Lohani, R.A. Zargar, V.P.S. Awana, B.R. Sekhar, J. Phys.: Condens. Matter 26 (2014) 425501. K.K. Gomes, A.N. Pasupathy, A. Pushp, S. Ono, Y. Ando, A. Yazdani, Nature 447 (2007) 569. J. Lee, K. Fujita, A.R. Schmidt, C.K. Kim, H. Eisaki, S. Udhida, J.C. Davis, Science 325 (2009) 1099. M.J. Lawler, K. Fujita, J. Lee, A.R. Schmidt, Y. Kohsaka, C.K. Kim, H. Eisaki, S. Uchida, J.C. Davis, J.P. Sethna, E.-A. Kim, Nature 466 (2010) 347. T.D. Stanescu, P. Phillips, Phys. Rev. Lett. 91 (2003) 017002. M. Yi, D.H. Lu, R. Yu, S.C. Riggs, J.-H. Chu, B. Lv, Z.K. Liu, M. Lu, Y.-T. Cui, M. Hashimoto, S.-K. Mo, Z. Hussain, C.W. Chu, I.R. Fisher, Q. Si, Z.-X. Shen, Phys. Rev. Lett. 110 (2013) 067003. P. Pal, M.K. Dalai, R. Kundu, B.R. Sekhar, Phys. Rev B 77 (2008) 184405. Y. Zhang, F. Chen, C. He, L.X. Yang, B.P. Xie, Y.L. Xie, X.H. Chen, M. Fang, M. Arita, K. Shimada, H. Namatame, M. Taniguchi, J.P. Hu, D.L. Feng, Phys. Rev. B 82 (2010) 165113. P.-H. Lin, Y. Texier, A. Taleb-Ibrahimi, P. LeFèvre, F. Bertran, E. Giannini, M. Grioni, V. Brouet, Phys. Rev. Lett. 111 (2013) 217002. K. Okazaki, Y. Ito, Y. Ota, Y. Kotani, T. Shimojima, T. Kiss, S. Watanabe, C.-T. Chen, S. Niitaka, T. Hanaguri, H. Takagi, A. Chainani, S. Shin, Sci. Rep. 4 (2014) 4109. V.P.S. Awana, A. Pal, A. Vajpayee, M. Mudgel, H. Kishan, M. Husain, R. Zeng, S. Yu, Y.F. Guo, Y.G. Shi, K. Yamaura, E. Takayama-Muromachi, J. Appl. Phys. 107 (2010) 09E128. R.A. Zargar, A. Pal, A.K. Hafiz, V.P.S. Awana, J. Supercond, Nov. Magn. 27 (2014) 897. M. Maniraj, S.R. Barman, Rev. Sci. Instrum. 85 (2014) 033301. S. Banik, A.K. Shukla, S.R. Barman, Rev. Sci. Instrum. 76 (2005) 06612. M. Maniraj, B.N. Raja Sekhar, S.R. Barman, Rev. Sci. Instrum. 83 (2012) 046107. O.K. Anderson, Phys. Rev. B. 12 (1975) 3060. M. Aichhorn, S. Biermann, T. Miyake, A. Georges, M. Imada, Phys. Rev. B. 82 (2010) 064504. T. Miyake, K. Nakamura, R. Arita, M. Imada, J. Phys. Soc. Jpn. 79 (2010) 044705. Keith W. Goodman, Victor E. Henrich, Phys. Rev. B 49 (1994) 4827. T. Yokoya, R. Yoshida, Y. Utsumi, K. Tsubota, H. Okazaki, T. Wakita, Y. Mizuguchi, Y. Takano, T. Muro, Y. Kato, H. Kumigashira, M. Oshima, H. Harima, Y. Aiura, H. Sato, A. Ino, H. Namatame, M. Taniguchi, M. Hirai, Y. Muraoka, Sci. Technol. Adv. Mater. 13 (2012) 054403.
52
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
[35] N.L. Saini, Y. Wakisaka, B. Joseph, A. Iadecola, S. Dalela, P. Srivastava, E. Magnano, M. Malvestuto, Y. Mizuguchi, Y. Takano, T. Mizokawa, K.B. Garg, Phys. Rev. B 83 (2011) 052502. [36] B. Joseph, A. Iadecola, L. Simonelli, Y. Mizuguchi, Y. Takano, T. Mizokawa, N. L. Saini, J. Phys.: Condens. Matter 22 (2010) 485702. [37] L. Simonelli, N.L. Saini, Y. Mizuguchi, Y. Takano, T. Mizokawa, G. Baldi, G. Monaco, J. Phys.: Condens. Matter 24 (2012) 415501.
[38] K. Shimada, T. Mizokawa, K. Mamiya, T. Saitoh, A. Fujimori, K. Ono, A. Kakizaki, T. Ishii, M. Shirai, T. Kamimura, Phys. Rev. B 57 (1998) 8845. [39] C.L. Chen, C.L. Dong, J.L. Chen, J.-H. Guo, W.L. Yang, C.C. Hsu, K.W. Yeh, T. W. Huang, B.H. Mok, T.S. chan, J.F. Lee, C.L. Chang, S.M. Rao, M.K. Wu, Phys. Chem. Chem. Phys 13 (2011) 15666.
Contents lists available at ScienceDirect
Solid State Communications journal homepage: www.elsevier.com/locate/ssc
Electronic structure of the unoccupied electron energy states in FeSe1 xTex Pramita Mishra a, Himanshu Lohani a, M. Maniraj b, Jayita Nayak b, R.A. Zargar c, V.P.S. Awana d, Sudipta Roy Barman b, Biju Raja Sekhar a,n a
Institute of Physics, Sachivalaya Marg, Bhubaneswar 751005, India UGC-DAE Consortium for Scientific Research, Khandwa Road, Indore 452001, Madhya Pradesh India c Department of Physics, Jamia Millia Islamia, New Delhi-110025, India d CSIR-National Physical Laboratory, K S Krishnan Marg, New Delhi-110012, India b
art ic l e i nf o
a b s t r a c t
Article history: Received 5 May 2015 Received in revised form 26 June 2015 Accepted 29 June 2015 Communicated by Ralph Gebauer Available online 6 July 2015
Inverse photoemission spectroscopic (IPES) measurements along with LDA based band structure calculations have been used to investigate the unoccupied electronic structure of FeSe1 xTex system. The observed doping and temperature dependent pseudogap in this system is found to be linked to the change in the chalcogen height in their geometric structure. The depletion in spectral weight from the near EF states at low temperature in IPES has been correlated with the enhancement of the 3z2–r2 orbitals in the photoemission spectroscopy (PES). The Coulomb correlation energy U, estimated from the combined PES and IPES spectra, signifies the enhancement in electron correlations in FeSe1 xTex, with doping. The formation of pseudogap in PES and IPES confirms the importance of correlations in the 11 family of Fe superconductors. & 2015 Elsevier Ltd. All rights reserved.
Keywords: A. Noncuprate superconductors C. Electronic structure D. Pseudogap E. Inverse photoemission spectroscopy
1. Introduction Since the discovery of superconductivity in Fe pnictides [1], there has been a plethora of research activities to understand the nature of superconductivity in them. Many exotic properties of these superconductors originate from some of the key elements like magnetic fluctuations, unconventional electron–phonon coupling and electron correlations. Further, the multiorbital correlation in combination with Hund's coupling caused by the participation of all five Fe 3d electrons result in some distinctive electronic properties like orbital selectivity [2–4], Mottness [5], quantum phase transition [6], nematicity [7], and pseudogap [8]. The presence of pseudogaps is a characteristic feature of the family of unconventional superconductors such as high Tc cuprates [9,10], heavy fermion superconductors [11] and the recently discovered iron pnictides and chalcogenides [12–15]. In last few years, correlation of the pseudogap with the superconducting state has been under intense study. Although, the origin of pseudogap in these materials is still under debate, a general consensus is prevailing among researchers about the existence of two types of pseudogaps, located at low and high binding energies in high Tc superconductors.
The low energy pseudogap which originates from superconducting fluctuations [16,17] persists just above Tc whereas the origin of high energy pseudogap existing far above Tc is still a puzzle [18,19]. Nevertheless, electron correlation is considered to play a vital role in the existence of the high temperature pseudogap in unconventional superconductors [15,20] like the case of manganites [21]. Electron spectroscopic methods are considered to be quite suited to probe the near EF electronic structure and thereby the nature of pseudogaps. Although a number of photoemission studies [15,22–24] were focussed on the electronic structure of valence band (occupied states) and the pseudogap formation in the Fe(Se,Te) system, only few have been reported on the unoccupied states. In this study, inverse photoemission spectroscopy (IPES) in conjunction with LDA based band structure calculations has been used to study the unoccupied states above EF. The observed temperature dependent pseudogap has been discussed in correlation with its analogue in the valence band reported earlier [15]. Also, from a comparative study of photoemission and inverse photoemission we have estimated the Coulomb correlation energy in Fe(Se,Te) systems.
2. Experimental methods n
Corresponding author. E-mail address: [email protected] (B.R. Sekhar).
http://dx.doi.org/10.1016/j.ssc.2015.06.023 0038-1098/& 2015 Elsevier Ltd. All rights reserved.
Polycrystalline samples of FeSe1 xTex (x¼1, 0.5, 0) used in this study were synthesised via solid state reaction route [25] and
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
15
x=0
A1
x=0
49
10 B1
5
x=1
B1
DOS (states / eV)
Intensity (a.u)
x = 0.5 0 15
x = 0.5
Total Te Se Fe
10 5 0 15
A1
A1 10
Α1′
x=1
Α1′′ B1
5 0 0
2
4
6
8
Energy (eV)
0
2
4
6
Energy (eV)
Fig. 1. (Color online) Panel (a): IPES spectra for FeSe1 xTex (x¼ 1, 0.5, 0 shown in black, red and blue respectively). Dashed vertical lines correspond to the Fermi level. Panel (b): theoretically calculated total unoccupied density of states (DOS). The theoretical features which match with experiment have been marked as A1 and B1 in panels a and b. A1 composed of two sub-features A10 and A1″ , which are not resolved in experimental spectra, is marked in panel b.
characterised for their electrical, magnetic and structural properties [26]. In order to avoid a common source of error in the spectroscopic analysis of these compounds, it was ensured that the samples contain no excess Fe and are of single phase in nature. Inverse photoemission spectroscopic (IPES) measurements were performed in the isochromat mode with a mean photon energy of 9.9 eV using a combination of electrostatically focussed Stoffel Johnson type electron gun with a contact potential [27] and an acetone gas filled photon band pass detector with a CaF2 window [28,29]. The total energy resolution was 600 meV. IPES spectra were obtained by normalising the measured photon counts by the sample current at each energy step. The binding energy has been determined with reference to the Fermi level of a sputter cleaned silver surface. Angle integrated ultraviolet photoemission spectroscopy (UPS) measurements were performed by using another system equipped with a high intensity vacuumultraviolet source and a hemispherical electron energy analyser (SCIENTA R3000). At the He I (hν ¼21.2 eV) line, the photon flux was of the order of 1016 photons/s/steradian with a beam spot of 2 mm in diameter. Fermi energies for all measurements were calibrated by using a freshly evaporated Ag film onto the sample holder. The total energy resolution, estimated from the width of the Fermi edge, was about 27 meV for the He I excitation. Both UPS and IPES measurements were performed at a base pressure of 5.0 10 11 mbar. The samples were repeatedly scraped using a diamond file inside the preparation chamber with a base vacuum of 5.0 10 10 mbar and the spectra were taken within an hour, so as to avoid any surface degradation. All the measurements were repeated many times to ensure the reproducibility of the spectra. The low temperature measurements at 77 K were performed immediately after cleaning the sample surfaces followed by the room temperature measurements. Band structure calculations using TBLMTO-ASA [30] were performed using the experimental lattice parameters reported earlier [26]. The correlation effects of
the Fe-3d orbitals were taken into account by using the LDA þ U formalism with J as 0.9 eV [31] and U as 3.5 eV for FeTe and 4.0 eV for FeSe [32]. For FeSe0.5Te0.5, U was taken as 3.8 eV, a value intermediate between those of FeTe and FeSe.
3. Results and discussion Panel (a) of Fig. 1 shows the inverse photoemission spectra of the FeSe1 xTex samples in the energy range 1 to 9 eV taken at room temperature and panel (b) shows our theoretical band structure calculations using TB-LMTO ASA for the FeSe1 xTex depicting the total unoccupied density of states (DOS). The experimental as well as calculated spectra show two prominent features A1 and B1 located at binding energy positions 1 and 5.5 eV respectively. By comparing these two panels, it can be identified that A1 arises primarily from the unoccupied Fe-3d states with a small contribution from the unoccupied chalcogen (Se/Te) p states and B1 consists predominantly of the hybridised Fe3d–Se4p/Te5p unoccupied states. The feature A1 comprises two sub features A10 and A1″ which are not resolved in the experimental spectra. B1 is less prominent due to the electron scattering in inverse photoemission process [33]. These assignments of features are in agreement with other's calculations also [32,34]. It can be seen from the experimental spectra (panel (a)) that with a decrease in the doping value x, the spectral weight near the Fermi level (EF) decreases and correspondingly the intensity of A1 increases. A similar change was earlier reported by Yokoya et al. [34] also. Such a shift in the spectral weight is a signature of the presence of a pseudogap in the case of FeSe. Panel (b) shows that with a decrease in x value, the feature A10 moves away from the EF while the feature A1″ moves towards the EF in addition to an intensity enhancement of both A10 and A1″ . As a result, the plateau like structure seen in the case of FeTe transforms into a prominent peak. The feature B1 moves to
50
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
Α1′
x=0
Α1′′
A1
Intensity (a.u)
4
2
x = 0.5
Fe 3d PDOS (states / eV)
0
x = 0.5
Total xy yz
4
2
3z -r xz 2
x -y
x=1
2
2
-1
0
1
2
3
Energy (eV) Fig. 3. (Color online) Temperature dependent IPES spectra taken at 300 K (black) and 77 K (red) for FeSe1 x Te1 x (x ¼ 1, 0.5, 0).
2
0
x=1 4
Α1′′
3
Α1′ 2 1 0 -0.5
x=0
0
0.5
1
1.5
2
2.5
Energy (eV) Fig. 2. (Color online) Fe 3d unoccupied partial density of states (PDOS), using LDA þ U, for FeSe1 xTex (x¼ 1, 0.5, 0). The PDOS contribution (panel (b)) for xz (magenta) and yz (blue) orbitals are the same for FeTe and FeSe. For FeSe0.5Te0.5, the xz and yz PDOS split owing to the different bond lengths of Fe–Te and Fe–Se.
higher energy with a decrease in x resulting in a larger separation between the features A1 and B1 in case of FeSe in comparison to FeTe. The depletion of spectral weight near EF with Se substitution has been reported by Saini et al. [35] based on their studies using Xray absorption spectroscopy (XAS) in which the electronic structure of unoccupied states of FeSe1 xTex are probed. This study showed that the Fe-3d unoccupied states display a gradual decrease in spectral intensity with Se substitution. The experimental observation, supported by cluster model calculations, correlates the spectral weight depletion to the enhanced hybridisation of chalcogen p orbitals with the Fe 3d orbitals. The increased chalcogen p–Fe-3d hybridisation for FeSe in comparison to FeTe is also evidenced from the Fe K edge X ray absorption near edge spectra (XANES) [36] which probes the unoccupied Fe–chalcogen hybridised states. Thus, the change in hybridisation strength of the Fe-3d–Se-4p/Te-5p states govern the electronic structure of the FeSe1 xTex system and result in a reduced density of unoccupied states near Fermi level with Se doping. The nature of bands involved in the formation of the pseudogap and in the spectral weight shifts could well be identified from our Fe 3d partial density of states (PDOS) calculations using the TB-
LMTO ASA method. Fig. 2 shows the Fe 3d PDOS of the FeSe1 xTex system. The xz/yz and x2–y2 orbitals are the highest populated ones corresponding to the feature A10 while the xy orbital is the highest for A1″ . The intensity of xy and x2–y2 orbitals could be seen to be increasing with the reduction in the doping value x, leading to the increase in the intensity of A1. Fig. 3 shows the IPES spectra of FeSe1 xTex taken at 300 K (black) and 77 K (red). Quite similar to the spectral changes observed as a function of doping, the intensity of the feature A1 increases for higher temperature. Further, such a change in A1 could be seen for all the three compositions. This shows that the increase in the doping value and increase in the temperature have similar bearings on the spectral weight shifts which lead to the pseudogap. Moreover, such a temperature dependent change also shows that these spectral weight shifts are inherent to the samples and are not artefacts arising from spectral normalisation or sample surface cleaning. In order to understand the origin of the observed temperature dependent pseudogap in IPES which depict the unoccupied states, we have compared our results with the photoemission spectra (PES) representing the occupied states. A comparison between the temperature dependent photoemission spectra taken with He I (21.2 eV) source [15] and the present IPES spectra is shown in Fig. 4. The figure shows an enhancement of states for the feature A in the occupied region ( 0.5 eV – EF energy range) and a complimentary depletion of states for feature A1 in the unoccupied region (EF – 1 eV range) at low temperature. This temperature dependent enhancement in PES was attributed to the change in the chalcogen height (Z) from the Fe plane at low temperature [15]. Reduction in Z at low temperature causes change in the hybridisation between Fe 3d and chalcogen p orbitals. This results in the enhancement of spectral weight for feature A due to the increased occupancy of the 3z2–r2 orbitals at low temperature. The spectral weight in IPES has a direct correspondence to the same in PES. Thus, the enhancement in spectral weight of 3z2–r2 orbitals in the occupied side manifests as depletion of states in the unoccupied side. This explains the observed changes in features A10 and A1″ with temperature. Our result is consistent with the temperature dependent Fe K edge XAS measurements [37], showing an enhancement of pre-peak A, at low temperature in FeSe1 xTex, which has been correlated to the enhanced hybridisation of Fe-3d and chalcogen-p orbitals due to the decreased chalcogen height (h) at low temperature. By combining the spectra from PES and IPES we can estimate some of the energy scales involved in the electronic transport in
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
by using theoretical calculations [32]. The gap in the density of unoccupied states with doping of Se leads to a shifting of the Upper Hubbard band B1 towards higher energy which signifies the enhancement of electron correlation with doping.
x=0 A1
Intensity (a.u)
A
4. Conclusion x = 0.5
x=1
-0.4
0
0.4
0.8
Energy (eV) Fig. 4. (Color online) PES spectrum (solid lines) measured using He I (21.2 eV) below the Fermi level (dashed vertical line) and IPES spectra (dotsþlines) at 9.2 eV photon energy above Fermi level for FeSe1 xTex (x¼ 1. 0.5, 0). The black and red spectra correspond to data taken at 300 K and 77 K respectively.
In conclusion, we have made a systematic study of the unoccupied states in FeSe1 xTex, using IPES and interpreted the results using our theoretical band structure calculations. A doping and temperature dependent pseudogap formation reflecting the role of chalcogen height in this system is observed. The temperature dependency of the pseudogap in the unoccupied region is complementary to the earlier observed pseudogap in the occupied region. We have also estimated the Coulomb correlation energy for different doping concentrations which demonstrate that the electron correlations in Fe(Se,Te) superconductors do have a bearing on the change in their geometric structure.
References [1] [2] [3] [4]
A1
B
x=0
[5] [6]
Intensity (a.u)
[7]
x = 0.5
[8] [9] [10] [11] [12] [13]
x=1
[14] [15] [16] [17]
-4
51
-3
-2
-1
0
1
2
3
Energy (eV) Fig. 5. (Color online) The PES and IPES spectra below and above the Fermi level respectively, for FeTe (black), FeSe0.5Te0.5 (blue) and FeSe (red). The black ticks correspond to the positions of A1 and B.
this family of Fe superconductors, like the onsite Coulomb energy (U). The feature B in the valence band spectra and A1 in the IPES spectra are ascribed to the lower and the upper Hubbard band respectively [31]. The energy separation between B and A1 provides a rough estimate of the onsite Coulomb energy, U [38]. Fig. 5 shows the PES spectra taken using He II (40.8 eV) source and IPES spectra at 9.9 eV photon energy, of the FeSe1 xTex system plotted together by aligning the corresponding Fermi energies. From the plot the value of U was estimated to be 3.5, 3.4, 3.2 eV for FeSe, FeSe0.5Te0.5 and FeTe respectively. The estimated U for FeSe is comparable to the result obtained experimentally by Yokoya et al. [34]. Chen et al. [39] reported a U value of 4.0 using X-ray Absorption Spectroscopy (XAS) and X-ray Emission Spectroscopy (XES). Further, it should be noted that U increases in FeSe1 xTex with decreasing x. Such a trend was earlier predicted
[18] [19] [20]
[21] [22]
[23] [24] [25]
[26] [27] [28] [29] [30] [31] [32] [33] [34]
Y. Kamihara, T. Watanabe, M. Hirano, H. Hosono, J. Am. Chem. Soc. 130 (2008) 3296. T. Hanaguri, S. Niitaka, K. Kuroki, H. Takagi, Science 328 (2010) 474. H. Lohani, P. Mishra, B.R. Sekhar, Physica C 512 (2015) 54. N. Lanatá, U.R.H. Strand, G. Giovannetti, B. Hellsing, dè Medici Luca, M. Capone, Phys. Rev. B. 87 (2013) 045122. dè Medici Luca, G. Giovannetti, M. Capone, Phys. Rev. Lett. 112 (2014) 177001(R). Julián Rinćon, Adriana Moreo, Gonzalo Alvarez, Elbio Dagotto, Phys. Rev. B 90 (2014) 241105(R). J.K. Glasbrenner, I.I. Mazin, Harald O. Jeschke, P.J. Hirschfeld, Roser Valenti, arXiv:1501.04946, 2015. L. Craco, M.S. Laad, arXiv:1001.3273, 2010. T. Timusk, B. Statt, Rep. Prog. Phys. 62 (1999) 61. M.R. Norman, D. Pines, C. Kallin, Adv. Phys. 54 (2005) 715. V.A. Sidorov, M. Nicklas, P.G. Pagliuso, J.L. Sarrao, Y. Bang, A.V. Balatsky, J. D. Thompson, Phys. Rev. Lett. 89 (2002) 157004. Y.S. Kwon, J.B. Hong, Y.R. Jang, H.J. Oh, Y.Y. Song, B.H. Min, T. Lizuka, S.I. Kimura, A.V. Balatsky, Y. Bang, New J. Phys. 14 (2012) 063009. T. Sato, S. Souma, K. Nakayama, K. Terashima, K. Sugawara, T. Takahashi, Y. Kamihara, M. Hirano, H. Hosono, J. Phys. Soc. Jpn. 77 (2008) 063708. D.C. Johnston, Advances in Physics 59 (2010) 803. P. Mishra, H. Lohani, R.A. Zargar, V.P.S. Awana, B.R. Sekhar, J. Phys.: Condens. Matter 26 (2014) 425501. K.K. Gomes, A.N. Pasupathy, A. Pushp, S. Ono, Y. Ando, A. Yazdani, Nature 447 (2007) 569. J. Lee, K. Fujita, A.R. Schmidt, C.K. Kim, H. Eisaki, S. Udhida, J.C. Davis, Science 325 (2009) 1099. M.J. Lawler, K. Fujita, J. Lee, A.R. Schmidt, Y. Kohsaka, C.K. Kim, H. Eisaki, S. Uchida, J.C. Davis, J.P. Sethna, E.-A. Kim, Nature 466 (2010) 347. T.D. Stanescu, P. Phillips, Phys. Rev. Lett. 91 (2003) 017002. M. Yi, D.H. Lu, R. Yu, S.C. Riggs, J.-H. Chu, B. Lv, Z.K. Liu, M. Lu, Y.-T. Cui, M. Hashimoto, S.-K. Mo, Z. Hussain, C.W. Chu, I.R. Fisher, Q. Si, Z.-X. Shen, Phys. Rev. Lett. 110 (2013) 067003. P. Pal, M.K. Dalai, R. Kundu, B.R. Sekhar, Phys. Rev B 77 (2008) 184405. Y. Zhang, F. Chen, C. He, L.X. Yang, B.P. Xie, Y.L. Xie, X.H. Chen, M. Fang, M. Arita, K. Shimada, H. Namatame, M. Taniguchi, J.P. Hu, D.L. Feng, Phys. Rev. B 82 (2010) 165113. P.-H. Lin, Y. Texier, A. Taleb-Ibrahimi, P. LeFèvre, F. Bertran, E. Giannini, M. Grioni, V. Brouet, Phys. Rev. Lett. 111 (2013) 217002. K. Okazaki, Y. Ito, Y. Ota, Y. Kotani, T. Shimojima, T. Kiss, S. Watanabe, C.-T. Chen, S. Niitaka, T. Hanaguri, H. Takagi, A. Chainani, S. Shin, Sci. Rep. 4 (2014) 4109. V.P.S. Awana, A. Pal, A. Vajpayee, M. Mudgel, H. Kishan, M. Husain, R. Zeng, S. Yu, Y.F. Guo, Y.G. Shi, K. Yamaura, E. Takayama-Muromachi, J. Appl. Phys. 107 (2010) 09E128. R.A. Zargar, A. Pal, A.K. Hafiz, V.P.S. Awana, J. Supercond, Nov. Magn. 27 (2014) 897. M. Maniraj, S.R. Barman, Rev. Sci. Instrum. 85 (2014) 033301. S. Banik, A.K. Shukla, S.R. Barman, Rev. Sci. Instrum. 76 (2005) 06612. M. Maniraj, B.N. Raja Sekhar, S.R. Barman, Rev. Sci. Instrum. 83 (2012) 046107. O.K. Anderson, Phys. Rev. B. 12 (1975) 3060. M. Aichhorn, S. Biermann, T. Miyake, A. Georges, M. Imada, Phys. Rev. B. 82 (2010) 064504. T. Miyake, K. Nakamura, R. Arita, M. Imada, J. Phys. Soc. Jpn. 79 (2010) 044705. Keith W. Goodman, Victor E. Henrich, Phys. Rev. B 49 (1994) 4827. T. Yokoya, R. Yoshida, Y. Utsumi, K. Tsubota, H. Okazaki, T. Wakita, Y. Mizuguchi, Y. Takano, T. Muro, Y. Kato, H. Kumigashira, M. Oshima, H. Harima, Y. Aiura, H. Sato, A. Ino, H. Namatame, M. Taniguchi, M. Hirai, Y. Muraoka, Sci. Technol. Adv. Mater. 13 (2012) 054403.
52
P. Mishra et al. / Solid State Communications 219 (2015) 48–52
[35] N.L. Saini, Y. Wakisaka, B. Joseph, A. Iadecola, S. Dalela, P. Srivastava, E. Magnano, M. Malvestuto, Y. Mizuguchi, Y. Takano, T. Mizokawa, K.B. Garg, Phys. Rev. B 83 (2011) 052502. [36] B. Joseph, A. Iadecola, L. Simonelli, Y. Mizuguchi, Y. Takano, T. Mizokawa, N. L. Saini, J. Phys.: Condens. Matter 22 (2010) 485702. [37] L. Simonelli, N.L. Saini, Y. Mizuguchi, Y. Takano, T. Mizokawa, G. Baldi, G. Monaco, J. Phys.: Condens. Matter 24 (2012) 415501.
[38] K. Shimada, T. Mizokawa, K. Mamiya, T. Saitoh, A. Fujimori, K. Ono, A. Kakizaki, T. Ishii, M. Shirai, T. Kamimura, Phys. Rev. B 57 (1998) 8845. [39] C.L. Chen, C.L. Dong, J.L. Chen, J.-H. Guo, W.L. Yang, C.C. Hsu, K.W. Yeh, T. W. Huang, B.H. Mok, T.S. chan, J.F. Lee, C.L. Chang, S.M. Rao, M.K. Wu, Phys. Chem. Chem. Phys 13 (2011) 15666.