LETTERS PUBLISHED ONLINE: 1 JUNE 2015 | DOI: 10.1038/NMAT4302

High-temperature superconductivity in potassium-coated multilayer FeSe thin films Y. Miyata1, K. Nakayama1*, K. Sugawara2, T. Sato1 and T. Takahashi1,2 The recent discovery of possible high-temperature (T c ) superconductivity over 65 K in a monolayer FeSe film on SrTiO3 (refs 1–6) triggered a fierce debate on how superconductivity evolves from bulk to film, because bulk FeSe crystal exhibits a T c of no higher than 10 K (ref. 7). However, the difficulty in controlling the carrier density and the number of FeSe layers has hindered elucidation of this problem4,8 . Here, we demonstrate that deposition of potassium onto FeSe films markedly expands the accessible doping range towards the heavily electron-doped region. Intriguingly, we have succeeded in converting non-superconducting films with various thicknesses into superconductors with T c as high as 48 K. We also found a marked increase in the magnitude of the superconducting gap on decreasing the FeSe film thickness, indicating that the interface plays a crucial role in realizing the high-temperature superconductivity. The results presented provide a new strategy to enhance and optimize T c in ultrathin films of iron-based superconductors. High-temperature superconductivity has been realized as a consequence of controlling the carrier concentration9,10 . For instance, whereas parent bulk compounds of copper-oxide and iron-based superconductors are typically non-superconducting and exhibit antiferromagnetic or spin-density-wave (SDW) order, superconductivity arises when the ordered phases are suppressed or destroyed by hole or electron doping into the parent compounds. On increasing the doping, the Tc value gradually increases, reaches the maximum at an optimal doping, and then decreases, forming a characteristic dome-shaped superconducting region in the electronic phase diagram. Similarly, in the case of one-monolayer (1 ML) FeSe film on a SrTiO3 substrate—which has various fascinating properties, such as a very high Tc value1–6 , tunable heterostructure11,12 , and a possible link with novel topological phenomena13 —electron doping is a prerequisite for realizing the high-temperature superconductivity3,4,8 . However, in contrast to the well-established electronic phase diagram for bulk iron-based superconductors10 , that for thin films still remains elusive. This is because the electron doping so far examined uses electron transfer from the substrate3,4,8 , which, however, does not provide a sufficient carrier concentration to fully cover the superconducting dome in the phase diagram. This problem becomes more serious in the case of multilayer films, where the number of electrons transferred from the substrate to each FeSe layers is substantially reduced and cannot drive the superconductivity4,8 . In fact, there has been no report of multilayer films showing a signature of superconductivity1,4,8,11 . This, in return, provides a challenging

question as to whether superconductivity emerges in multilayer films when one can dope enough electrons. All these issues definitely call for a new method to dope extra electrons into single- or multilayer FeSe films in a well-controlled manner to unveil the unexplored superconducting regions. Here we propose a novel approach to solve the problem in a rather convenient manner, that is, an in situ deposition of potassium (K) atoms onto the film. Using this technique, we have succeeded in accessing the highest doping level ever achieved in FeSe thin films, and further converting non-superconducting films with various thicknesses into superconductors with Tc values as high as 48 K, although the bulk crystal exhibits a Tc of no higher than 10 K at ambient pressure7 . We first compare the electronic structures of pristine 1 ML and 3 ML FeSe films. The electronic band structure was determined by angle-resolved photoemission spectroscopy (ARPES) (see Methods for details of the film growth and ARPES measurements). In the 1 ML film (Fig. 1a–c), we immediately recognize that the Fermi surface consists of only a large electron pocket centred at the M point (γ pocket) of the Brillouin zone (Fig. 1b and right panel of Fig. 1c), in contrast to bulk FeSe, in which both hole- and electron-pockets are simultaneously observed14,15 . The electron carrier concentration (ne ) estimated from the Fermi-surface volume is ∼0.12 electrons per Fe, close to the reported upper limit for 1 ML film3,4 . We observed superconductivity with an onset temperature (T ∗ ) of ∼60 K for this 1 ML FeSe film, as evidenced by opening of a superconducting gap (Supplementary Fig. 1). As shown in Fig. 1e, the Fermi-surface topology of the 3 ML film is drastically different from that of the 1 ML film; the Fermi surface consists of a hole pocket at the 0 point (α hole pocket) and a small electron pocket located slightly away from the M point (ε electron pocket). From the experimental facts that the emergence of the ε pocket is a hallmark of the parent phase around the non-doped region4 , and no signature of the superconducting gap is found in the 3 ML film (shown later), we infer that the ground state of the pristine 3 ML film would be in an ordered phase, probably the SDW phase4 , in sharp contrast to the superconducting ground state in the 1 ML film. An important discovery manifests itself when we deposit K atoms onto the surface of the 3 ML film. As shown in Fig. 1h, deposition of a small amount of K atoms results in the disappearance of the α pocket, accompanied by a downward shift of the hole-like band at the 0 point (Fig. 1i). Simultaneously, an electron pocket (γ ) emerges at the M point, as also seen from the band dispersion (right panel of Fig. 1i). This evolution of the electronic structure is definitely caused by the electron doping from K atoms into FeSe

1 Department of Physics, Tohoku University, Sendai 980-8578, Japan. 2 WPI Research Center, Advanced Institute for Materials Research, Tohoku University,

Sendai 980-8577, Japan. *e-mail: [email protected]

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

© 2015 Macmillan Publishers Limited. All rights reserved

1

NATURE MATERIALS DOI: 10.1038/NMAT4302

LETTERS ky (π/a)

1 ML FeSe SrTiO3

e

3 ML FeSe

SrTiO3

High

M

Γ Low

f ε

α 0.0

M

Γ

−0.2

h

i

0.2

γ 0.0

M

Γ

−0.2 0.2

l

γ

0.0

M

Γ

−0.2 0.2

o

γ

0.0

−0.2

High

α

100 Low

EF

Γ α

M

Γ

M

50 100

EF

α

γ

50 100

EF

M

Γ

γ

α

50 100 150

B.E. (meV)

n ky (π/a)

m

50

150

B.E. (meV)

k ky (π/a)

j

γ

M

Γ

150

B.E. (meV)

K

EF

150

0.2

ky (π/a)

g

γ

0.0

−0.2

ky (π/a)

d

c

0.2

B.E. (meV)

b

B.E. (meV)

a

M

Γ

EF

1 kx (π/a)

M

γ

0.0

0.2

α

50 100 150

0

Γ

0.0 ky (π/a)

0.2

ky (π/a)

Figure 1 | Evolution of electronic structure on K deposition on a 3 ML FeSe film. a, Schematic view of a 1 ML FeSe film grown on a SrTiO3 substrate. b, ARPES intensity mapping at T = 30 K for the 1 ML film plotted as a function of a two-dimensional wavevector measured for the He–Iα line. The intensity is obtained by integrating the spectral intensity within ±5 meV of EF . Blue and green lines indicate the momentum cuts where the ARPES intensities in c were obtained. c, ARPES intensity plot near EF as a function of binding energy (B.E.) and wavevector measured along the cuts near 0 (left panel) and M (right panel) points, which correspond to blue and green lines in b, respectively. d–f, Same as a–c, but for a 3 ML film. g–o, Same as d–f, but for lightly (g–i), nearly optimally (j–l) and heavily (m–o) K-deposited 3 ML films, respectively.

layers and the resultant suppression of the parent ordered phase. On depositing further K atoms (Fig. 1j–o), the α band systematically shifts downwards and simultaneously the volume of γ Fermi surface increases gradually. The electron concentrations estimated from the Fermi-surface volume in Fig. 1h,k,n are ne ∼ 0.07, 0.11 and 0.15 (electrons per Fe), respectively. It is noted that the maximum value achieved here (∼0.15) exceeds the reported upper limit (∼0.12) for 1 ML film3 . These results unambiguously demonstrate that deposition of K is an effective way to dope electron carriers into FeSe film in a well-controlled manner. The success in precisely tuning the carrier concentrations by K deposition allows us to explore the superconductivity in multilayer films. Figure 2a shows high-resolution ARPES spectra at each Fermi vector (kF ) measured at low temperature (30 K for ne ∼ 0, and 13 K for others) for pristine (ne ∼ 0) and K-deposited (ne ≥ 0.07) 3 ML films. Whereas the midpoint of the leading edge is located at around EF for the pristine film (ne ∼ 0), that for ne ∼ 0.07, 0.11 and 0.13 is obviously shifted towards higher binding energy, indicative of gap opening at EF . This is more clearly seen in Fig. 2b, where the ARPES spectra are symmetrized to eliminate the influence from the Fermi–Dirac distribution function. The observed energy gap 2

is attributed to the superconducting gap, because the band shows bending back behaviour with the top of dispersion at the kF point (Fig. 2c–e), in good agreement with the Bogoliubov-quasiparticle dispersion in the superconducting state16 . Note that no clear gap is observed for the film with ne ∼ 0.15 (Fig. 2a,b), suggesting the suppression of superconductivity in the heavily doped region. To discuss quantitatively the difference in the superconducting-gap size (∆), we have performed numerical fittings to the symmetrized spectra (Fig. 2b) with the Bardeen– Cooper–Schrieffer (BCS) spectral function17 . Extracted ∆ values are 6.3 ± 1.0, 8.0 ± 0.8 and 6.3 ± 1.0 meV for ne ∼ 0.07, 0.11 and 0.13, respectively, whereas it almost vanishes for ne ∼ 0.15, suggesting the existence of a dome-like superconducting phase, as illustrated in Fig. 3. Thus K deposition onto 3 ML film enables the first observation of the suppression of superconductivity in the heavily overdoped region, which was not accessible using the conventional techniques3 . The present temperature-dependent ARPES measurement demonstrates that the superconducting gap survives at relatively high temperatures. Figure 2g shows the temperature dependence of the symmetrized ARPES spectrum for the 3 ML film near optimal

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

© 2015 Macmillan Publishers Limited. All rights reserved

NATURE MATERIALS DOI: 10.1038/NMAT4302 α

ne =

ε

b

ne = ∼0

∼0

c

e

d

∼0

γ γ

0.07

0.11

0.11

3 ML 40

20

EF −20

20

Binding energy (meV)

g

ne = 0.11

h

T (K) =

31 23

Δ (meV)

41

i

0 −20 −40

40 20

0 −20 −40

4

ne = 0.11

θ= 90°

128° 150°

2

ky

γ θ M

Γ

kx 10

High Low

Δ (meV)

8 6 4

θ

2 0

2

4

6

8 10

211°

13 0 20 0 −20 Energy relative to EF (meV)

0.05π/a

40

j

108°

6

30

T = 71 K

Energy relative to EF (meV) Energy relative to EF (meV)

8

61

46

40 20

−20

Energy relative to EF (meV)

51 Intensity (a.u.)

0

20

ne = 0.11

ne = 0.11 T = 13 K

0.15

0.15

13 K 71 K

10

f

0.13

0.13

γ

60

0.07

EF

ky

γ

Intensity (a.u.)

Intensity (a.u.)

∼0

Binding energy (meV)

a

LETTERS

0

20 40 60 Temperature (K)

20 0 −20 Energy relative to EF (meV)

3 ML ne = 0.11

Figure 2 | High-temperature superconductivity in a 3 ML FeSe film. a, Representative ARPES spectra measured at the kF point of the α and ε bands at T = 30 K for pristine 3 ML film (upper two spectra) and those of the γ band at T = 13 K for K-deposited 3 ML films with different dopings (lower four spectra). ne represents the electron carrier concentration in units of electrons per Fe. b, Symmetrized spectra of a. Triangles are a guide to the eyes to trace the peak position. c,d, Symmetrized ARPES spectra near EF at T = 13 K (c) and 71 K (d), measured along the green line in f. The spectrum at the kF point is shown in blue. Dashed curves are a guide to the eyes to trace the band dispersions. e, Temperature dependence of the band dispersion extracted from the peak positions in c,d. f, ARPES intensity at EF measured at 30 K around the M point in a K-deposited 3 ML film (ne = 0.11) plotted as a function of two-dimensional wavevector, together with the definition of Fermi-surface angle (θ ). The data are the same as those in Fig. 1k. White open circle represents the schematic Fermi surface. Filled circles indicate the momentum location where the symmetrized ARPES spectra in g,i were obtained. g, Temperature dependence of the symmetrized ARPES spectrum at the kF point (denoted by the red filled circle in f) of the γ band. h, Temperature dependence of the superconducting-gap size ∆(T) extracted by fitting the data in g with the BCS spectral function17 . Error bars are estimated from the standard deviation of ∆ in the numerical fitting. Black line represents the BCS mean-field calculation with T ∗ = 48 K and ∆(0) = 8 meV. i, Representative symmetrized ARPES spectra at 13 K measured at various kF points shown by filled circles in f. j, Momentum dependence of superconducting-gap size as a function of θ defined in f. Dashed red circle is a guide to the eyes to highlight the nearly isotropic character of the superconducting gap.

doping (ne ∼ 0.11) measured at the kF point (red filled circle in Fig. 2f). A two-peaked structure, indicative of superconducting-gap opening, is clearly seen up to T = 46 K, and seems to vanish at around 51 K. The estimated gap size as a function of temperature (Fig. 2h) closely follows the BCS mean-field form with a pairformation temperature T ∗ of 48 ± 3 K. Measurement of the gap size at various kF points indicates that the superconducting gap is nodeless and almost isotropic (Fig. 2i,j), as in the case of 1 ML film2,3 . We have confirmed that K deposition to dope electrons is applicable to multilayer films with various thicknesses. In fact, we found that the parent ordered phase in 2–20 ML films is destroyed by K deposition (Supplementary Fig. 2) and superconductivity emerges in K-deposited multilayer films with up to 4 ML thickness (Supplementary Fig. 2). The present result thus provides the first experimental evidence for superconducting pairing in multilayer FeSe films. It is inferred that the previously reported abrupt disappearance of superconductivity in multilayer FeSe films1 is due simply to insufficient electron doping from the substrate. A next important question is how the superconductivity evolves as a function of film thickness. Actually it has been a target of

intensive debate whether the superconductivity is enhanced8,18,19 or suppressed1,4,11 in multilayer films compared with a single-layer film. To get an insight into this problem, we have clarified the dopinginduced evolution of the superconducting gap for 1, 3 and 20 ML films. Figure 3 shows the symmetrized ARPES spectra at the γ Fermi surface for these three films, in which the doping level is optimized to show the maximum superconducting gap. It is noted that the 20 ML film does not show superconductivity at any K-deposition rate, even with the lowest accessible temperature of 13 K. The gap values estimated by numerical fittings for 1 and 3 ML films are ∆ = 19.2 ± 1.5 and 8.0 ± 0.8, respectively, signifying a marked enhancement of the pairing strength in the thinner film, in contrast to the conventional ultrathin-superconductor films18,19 . This was confirmed by extracting the intrinsic coherence-peak weight with the background-subtraction method (Supplementary Fig. 3). These results definitely demonstrate that the interface between the FeSe layer and the SrTiO3 substrate plays a crucial role in realizing hightemperature superconductivity. The next challenge is to pin down the source of the interfacial force responsible for the high-temperature superconductivity. The existence of a dome-shaped superconducting phase in 3 ML film

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

© 2015 Macmillan Publishers Limited. All rights reserved

3

NATURE MATERIALS DOI: 10.1038/NMAT4302

LETTERS 250 Intensity (a.u.)

1 ML

Torder

200 3 ML

150

30 0 −30 Energy (meV)

150

200

100

15 10

10

3 ML

0

5

20 ML

0 0.05 0.10 ne (electrons per Fe)

0

5

SC

0 0.00

5

SC

1 ML

Δ (meV)

50

Parent phase

Torder (K)

150

100

20

Parent phase

200

Parent phase

20 ML

100

25

Δ

0.15

Figure 3 | Electronic phase diagrams for different film thickness. Experimentally determined superconducting-gap size (∆) for 1, 3 and 20 ML films plotted as a function of electron doping (red filled circles). Error bars are estimated from the standard deviation of ∆ in the numerical fitting. Previously reported values for a 1 ML film3 are also shown by red open circles. Blue filled squares and diamond represent the transition temperature of the parent ordered phase (Torder ) reported in refs 4,8, respectively. Upper left panel shows the symmetrized ARPES spectra at 13 K measured at the kF point on the γ Fermi surface for 1, 3 and 20 ML films with ne ∼ 0.12, 0.11 and 0.09, respectively. Triangles are a guide to the eyes to trace the peak position.

thinnest limit, as supported by the experimental fact that the largest superconducting gap is observed for the 1 ML film. In this context, the suppression of superconductivity in the heavily electron-doped region is attributed to the weakening or disappearance of the antiferromagnetic fluctuations. The present result thus suggests a common pairing mechanism between FeSe films on SrTiO3 and bulk iron-based superconductors. Finally, we comment briefly on the implications of present result in relation to a recent in situ resistivity measurement for 1 ML film5 , which reported a high Tc value of 109 K and has triggered many discussions23 . In the present ARPES study, we have demonstrated that the highest Tc for 3 ML film is realized at ne ∼ 0.12. If the optimal-doping level remains unchanged in 1 ML film, the highest Tc expected from the ARPES data would be as high as 65 K, apparently in contradiction to the resistivity data. On the other hand, if the optimal-doping level in 1 ML film is higher than 0.12, as inferred from the monotonic increase of Tc in heavily doped 1 ML film (see Fig. 3), the highest Tc may reach or exceed 100 K as a consequence of an extended parent ordered phase and a resultant enhancement of antiferromagnetic fluctuations. In such a case, however, it remains unclear why an extreme electron doping could be achieved with a conventional doping technique in the resistivity measurement5 . To summarize, we have demonstrated that in situ K deposition onto the surface of FeSe film on SrTiO3 significantly expands the accessible doping range, enabling the first systematic investigation on the evolution of superconductivity as a function of film thickness and doping. We have demonstrated definitive evidence for the emergence of high-temperature superconductivity in multilayer FeSe films, together with the dome-shaped nature of the superconducting region in the phase diagram. The present result indicates that interaction with the substrate is a prerequisite for realizing high-temperature superconductivity, cross-interfacial electron–phonon coupling is not the primary interaction that mediates the superconducting pairing, and the parent ordered phase is closely related to the high-temperature superconductivity.

Methods (Fig. 3) puts a severe constraint on the proposed models. For instance, one of the leading candidates to trigger the enhanced pairing at the interface may be electron–phonon coupling which uses the cross-interfacial interactions (that is, interactions between electrons in FeSe and phonons in SrTiO3 ; refs 1,20,21). A signature of electron–phonon coupling has been recently reported by an ARPES study of 1 ML film, which observed a main-band replica separated by ∼100 meV ascribable to the phonon shake-off effect21 (note that such a replica band is not clearly seen in our ARPES data, probably due to the photoelectron matrix-element effect). However, the substantial reduction of ∆ and T ∗ in the heavily doped region of the 3 ML film (see Fig. 3) is not explained well by the cross-interfacial electron–phonon coupling alone, because the phonon modes of SrTiO3 , as well as its coupling strength to the FeSe electrons, are expected to be rather insensitive to the small variation in the ne value (from 0.11 to 0.15). It is thus inferred that the cross-interfacial electron–phonon coupling is not important (or plays a secondary role) in strengthening the pairing, and other primary interactions may possibly mediate the pairing. The present ARPES observation supports the models based on the close relationship between the parent ordered phase and the superconductivity. It has been argued that interfacial effects, such as tensile strains4,22 and/or non-trivial interactions at the interface11 , would enhance the antiferromagnetic interaction in the FeSe layer, leading to the higher onset temperature of the parent ordered phase (Torder ) in the thinner film4,8 (schematically shown in Fig. 3). We think that the enhanced antiferromagnetic fluctuations in the properly doped region would give rise to the highest Tc at the 4

Methods and any associated references are available in the online version of the paper. Received 29 December 2014; accepted 21 April 2015; published online 1 June 2015

References 1. Wang, Q. Y. et al. Interface-induced high-temperature superconductivity in single unit-cell FeSe films on SrTiO3 . Chin. Phys. Lett. 29, 037402 (2012). 2. Liu, D. F. et al. Electronic origin of high-temperature superconductivity in single-layer FeSe superconductor. Nature Commun. 3, 931 (2012). 3. He, S. L. et al. Phase diagram and electronic indication of high-temperature superconductivity at 65 K in single-layer FeSe films. Nature Mater. 12, 605–610 (2013). 4. Tan, S. Y. et al. Interface-induced superconductivity and strain-dependent spin density wave in FeSe/SrTiO3 thin films. Nature Mater. 12, 634–640 (2013). 5. Ge, G-F. et al. Superconductivity above 100 K in single-layer FeSe films on doped SrTiO3 . Nature Mater. 14, 285–289 (2015). 6. Sun, Y. et al. High temperature superconducting FeSe films on SrTiO3 substrates. Sci. Rep. 4, 6040 (2014). 7. Hsu, F. C. et al. Superconductivity in the PbO-type structure α-FeSe. Proc. Natl Acad. Sci. USA 105, 14262–14264 (2008). 8. Liu, X. et al. Dichotomy of the electronic structure and superconductivity between single-layer and double-layer FeSe/SrTiO3 films. Nature Commun. 5, 5047 (2014). 9. Lee, P. A., Nagaosa, N. & Wen, X-G. Doping a Mott insulator: Physics of high-temperature superconductivity. Rev. Mod. Phys. 78, 17–85 (2006). 10. Stewart, G. R. Superconductivity in iron compounds. Rev. Mod. Phys. 83, 1589–1652 (2011). 11. Peng, R. et al. Tuning the band structure and superconductivity in single-layer FeSe by interface engineering. Nature Commun. 5, 5044 (2014).

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

© 2015 Macmillan Publishers Limited. All rights reserved

NATURE MATERIALS DOI: 10.1038/NMAT4302 12. Peng, R. et al. Measurement of an enhanced superconducting phase and a pronounced anisotropy of the energy gap of a strained FeSe single layer in FeSe/Nb:SrTiO3 /KTaO3 heterostructures using photoemission spectroscopy. Phys. Rev. Lett. 112, 107001 (2014). 13. Hao, N. & Hu, J. Topological phases in the single-layer FeSe. Phys. Rev. X 4, 031053 (2014). 14. Maletz, J. et al. Unusual band renormalization in the simplest iron-based superconductor FeSe1−x . Phys. Rev. B 89, 220506(R) (2014). 15. Nakayama, K. et al. Reconstruction of band structure induced by electronic nematicity in an FeSe superconductor. Phys. Rev. Lett. 113, 237001 (2014). 16. Matsui, H. et al. BCS-Like Bogoliubov quasiparticles in high-Tc superconductors observed by angle-resolved photoemission spectroscopy. Phys. Rev. Lett. 90, 217002 (2003). 17. Norman, M. R., Randeria, M., Ding, H. & Campuzano, J. C. Phenomenology of the low-energy spectral function in high-Tc superconductors. Phys. Rev. B 57, R11093–R11096 (1998). 18. Haviland, D. B., Liu, Y. & Goldman, A. M. Onset of superconductivity in the two-dimensional limit. Phys. Rev. Lett. 62, 2180–2183 (1989). 19. Song, C. L. et al. Molecular-beam epitaxy and robust superconductivity of stoichiometric FeSe crystalline films on bilayer graphene. Phys. Rev. B 84, 020503(R) (2011). 20. Xiang, Y. Y., Wang, F., Wang, D., Wang, Q. H. & Lee, D. H. High-temperature superconductivity at the FeSe/SrTiO3 interface. Phys. Rev. B 86, 134508 (2012). 21. Lee, J. J. et al. Interfacial mode coupling as the origin of the enhancement of Tc in FeSe films on SrTiO3 . Nature 515, 245–248 (2014).

LETTERS 22. Cao, H. Y., Tan, S. Y., Xiang, H. J., Feng, D. L. & Gong, X. G. Interfacial effects on the spin density wave in FeSe/SrTiO3 thin films. Phys. Rev. B 89, 014501 (2014). 23. Bozovic, I. & Ahn, C. A new frontier for superconductivity. Nature Phys. 10, 892–895 (2014).

Acknowledgements We thank Q. Xue, X. Ma, L. Wang, F. Li and W. Zhang for their advice in thin-film growth. We also thank E. Ieki, G. N. Phan, S. Kanayama and E. Noguchi for their assistance in thin-film growth and ARPES measurements. This work was supported by grants from the Japan Society for the Promotion Science (JSPS) and the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan.

Author contributions Y.M., K.N. and T.T. designed the research. Y.M., K.N., K.S. and T.S. carried out the experiment. Y.M., K.N., T.S. and T.T. wrote the manuscript.

Additional information Supplementary information is available in the online version of the paper. Reprints and permissions information is available online at www.nature.com/reprints. Correspondence and requests for materials should be addressed to K.N.

Competing financial interests The authors declare no competing financial interests.

NATURE MATERIALS | ADVANCE ONLINE PUBLICATION | www.nature.com/naturematerials

© 2015 Macmillan Publishers Limited. All rights reserved

5

NATURE MATERIALS DOI: 10.1038/NMAT4302

LETTERS Methods FeSe thin films were grown on a TiO2 -terminated Nb(0.5 wt%)-doped SrTiO3 substrate (SHINKOSHA) with the molecular beam epitaxy (MBE) method in a vacuum better than 2 × 10−10 torr (ref. 1). The substrate was first degassed at 600 ◦ C for 2 h by resistive heating, and then heated to 900 ◦ C for 30 min. FeSe films were grown by co-evaporating Fe and Se in a Se-rich condition while keeping the substrate temperature at 430 ◦ C. After the evaporation, the films were annealed at 450 ◦ C for 2 h, and then transferred to the ARPES-measurement chamber without breaking vacuum. Reflection high-energy electron diffraction (RHEED) and low-energy electron diffraction (LEED) measurements were performed to characterize the substrate and film surfaces. The film thickness was controlled

by varying the deposition time and keeping a constant deposition rate (approximately 0.01 ML s−1 ), and was estimated using a quartz-oscillator thickness monitor as well as the electronic structure determined by ARPES. Deposition of the potassium atoms was carried out using a potassium dispenser (SAES Getters). ARPES measurements were performed in an ultrahigh vacuum better than 5 × 10−11 torr using a VG-Scienta SES2002 spectrometer with a high-flux He discharge lamp at Tohoku University. To excite photoelectrons, the He–Iα resonance line (hν = 21.218 eV) was used. The energy and angular resolutions were set at 4–30 meV and 0.2◦ , respectively. The Fermi level of the films was referenced to that of a gold film evaporated onto the sample holder.

NATURE MATERIALS | www.nature.com/naturematerials

© 2015 Macmillan Publishers Limited. All rights reserved

High-temperature superconductivity in potassium-coated multilayer FeSe thin films.

The recent discovery of possible high-temperature (T(c)) superconductivity over 65 K in a monolayer FeSe film on SrTiO3 (refs 1-6) triggered a fierce ...
2MB Sizes 0 Downloads 7 Views