Cite this: CrystEngComm, 2013, 15, 1989
Single crystal growth and characterization of tetragonal
FeSe
12xsuperconductors
Received 14th November 2012, Accepted 4th January 2013 DOI: 10.1039/c2ce26857d www.rsc.org/crystengcomm
Dmitriy Chareev,aEvgeniy Osadchii,aTatiana Kuzmicheva,bJiunn-Yuan Lin,c Svetoslav Kuzmichev,dOlga Volkovadand Alexander Vasiliev*d
The plate-like single crystals of tetragonal (P4/nmm) FeSe12xsuperconductors were grown using the KCl–
AlCl3flux technique which produced single crystalline tetragonal samples of about 4 6 4 6 0.1 mm3
dimensions. The energy dispersive X-ray spectroscopy established a ratio of Fe : Se = 1 : 0.96 ¡ 0.02. The resistivity and magnetization measurements revealed a sharp superconducting transition at Tc= 9.4 K.
Multiple Andreev reflections spectroscopy pointed to the existence of two-gap superconductivity with the gap values DL= 2.4 ¡ 0.2 meV and DS= 0.75 ¡ 0.1 meV at 4.2 K.
Introduction
Iron-based superconductors are extremely sensitive to impur-ity phases and defects in the crystal structure, which could significantly change the physical and chemical properties both in normal and superconducting states.1 This vulnerability is due to the fact that both microscopic and macroscopic inhomogeneities could be magnetically active, therefore influencing bulk and surface superconductivity. In the basic families of iron superconductors, the 11-type materials stand apart, owing to the absence of intermediate charge reservoir layers. The FeSe12x compounds consist only of a stack of
electronically active layers weakly coupled by the van der Waals interaction. These layers are based on the edge-shared FeSe4
tetrahedra. FeSe12xdemonstrates superconductivity with Tc¢ 8 K,2–4but its single crystals can be used for either production of intercalated materials AxFe2Se2(A = Li, Na, Ba, Sr, Ca, Yb
and Eu) with Tc= 30–46 K5 or superconducting monolayers
with Tc= 65 K.6–8 In this paper we describe the synthesis of
FeSe12x single crystals from the halide eutectic flux under
steady temperature gradient conditions and provide proofs of their high quality by the measurements of transport and thermodynamic properties.
The superconducting tetragonal phase of FeSe12x exists
only below 730 K (457 uC) in a rather narrow composition
range.9 If these conditions are not fulfilled, the samples are contaminated by Fe, Fe7Se8, and Fe3O4impurities.
The attempts to grow tetragonal FeSe12x single crystals
from alkali–halide flux have been undertaken in ref. 10–13. The synthesis of iron selenides from KCl flux with the melting temperature 776uC was described in ref. 10. The ampoule with Fe, Se and flux was heated up to 840uC and sustained at this temperature for 30 hours (h) to homogenize the solution. It
was then cooled down to 820 uC for 1 h to provide the
necessary supersaturation for nucleation. Further cooling was done with the rate of 0.3–0.5uC h21from 820 to 770uC. After
that, the ampoule was cooled rapidly down to 400uC and held for 24 h to stabilize Fe and Se. The single crystals with the hexagonal shape were produced in this synthesis. Ref. 11 presented the synthesis based on NaCl/KCl flux with the eutectic temperature 657 uC. The ampoule with preliminary obtained FeSe0.89powder and flux was heated to 900uC. Three
steps of cooling were employed: firstly with 3uC h21rate down
to 740uC, secondly with 1 uC h21rate down to 600uC, and,
finally, the furnace was cooled rapidly down to room temperature. Structural measurements revealed that both tetragonal (a) and hexagonal (b) phases coexist in the sample. A more promising method with LiCl–CsCl mixture which melts at 326uC was used in ref. 12. The ampoule with elemental Fe and Se mixed with the flux was heated to 715uC, and kept for 1 h at this temperature before shifting to a preheated furnace at 457uC. After slow cooling down to 300 uC it was quenched in water. The main disadvantage of this method is that it likely results in either a or b phase depending on the various growth conditions. Application of vapor transport method to the synthesis of tetragonal FeSe12xin ref. 13 also had the same
problem as above. Therefore, to grow tetragonal FeSe12xsingle
crystals of high quality, it is necessary to use a eutectic flux which melts at low temperatures (preferably below 250 uC),
aInstitute of Experimental Mineralogy, Russian Academy of Sciences, 142432
Chernogolovka, Moscow District, Russia
bP.N. Lebedev Physical Institute, Russian Academy of Sciences, 119991 Moscow,
Russia
cInstitute of Physics, National Chiao Tung University, Hsinchu 300, Taiwan, R.O.C dLow Temperature Physics and Superconductivity Department, M.V. Lomonosov
Moscow State University, 119991 Moscow, Russia. E-mail: [email protected]; Fax: +07 495 9329217; Tel: +07 495 9329217
PAPER
Published on 04 January 2013. Downloaded by National Chiao Tung University on 28/04/2014 02:08:49.
View Article Online
and to synthesize at temperatures below 457 uC to obtain reproducible growth results.
Experimental section
The FeSe12xsingle crystals were grown from the flux under a
permanent gradient of temperature. The chemical composi-tions of tetragonal FeSe12xand hexagonal Fe7Se8were studied
by energy-dispersive X-ray spectroscopy performed on a CAMECA SX100 (15 kV) analytical scanning electron micro-scope. The X-ray diffraction spectra were collected by a DRON diffractometer (Co Ka-radiation, Fe filter) in the 10u , 2h , 90u range on the powder samples prepared by grinding small or medium sized single crystals in acetone. To compare the quality of obtained single crystalline samples an additional amount of polycrystalline iron selenide was prepared by the conventional solid-state reaction.13 The FeSe single crystals X-ray diffraction spectra were collected along the c-axis using a BRUKER diffractometer (CuKa1radiation, graphite
monochro-mator).
The FeSe12xsingle crystals were grown in evacuated quartz
ampoules using a KCl–AlCl3 flux.14 The temperature of
synthesis and phase compositions were chosen in accord with the Fe–Se phase diagrams.9 It is noted that the tetragonal FeSe12xphase easily coexists with hexagonal Fe7Se8phase and
pure iron. The decomposition of FeSe12xtetragonal phase to
Fe7Se8 and Fe occurs at temperatures above 457 uC (730 K).
The homogeneity range of tetragonal FeSe12xis very narrow,
i.e., from FeSe0.96to FeSe0.975.15
The starting chemicals were carbonyl iron powder or iron spheres with the diameter of about 0.2 mm, crystalline selenium purified by the floating zone method, potassium chloride KCl and anhydrous aluminum chloride AlCl3(Fluka,
98%). The iron powder and pieces of selenium were mixed in the ratio Fe : Se = 1 :y0.94 and loaded into silica ampoules with an inner diameter of 8 mm and length of 110 mm. Then hygroscopic AlCl3was promptly weighed in air and loaded into
the ampoules. Finally, KCl taken in the ratio AlCl3: KCl = 2 : 1
was weighed and added into the ampoules which were
evacuated down to 1024 bar pressure and sealed. The
ampoules with chemicals were loaded into a horizontal tube furnace, heated to the temperature of synthesis for 1–2 h, and then kept at this temperature for 40–50 days. The temperature of the hot part of the ampoules was about 700–740 K. That of the cold part was 650–680 K. The volume fraction of the melted eutectic from the volume of sealed ampoules was about 60– 65%.
The ampoules were eventually extracted from the furnace and cooled in air. The cold part of the ampoules with crystals of iron selenide was cut off and the eutectic retrieved from it was dissolved in distilled water. Then, the crystals were rinsed in an ultrasonic bath twice in alcoholand twice in acetone. The rinsed crystals were dried in a muffle furnace at 70uC for a few minutes. Tetragonal crystals were separated using a steel magnetic filling knife which can attract metallic iron and
hexagonal Fe7Se8. Cleaning the crystals in water and alcohol
did not produce any visible oxidation of the surface. The crystals were placed into evacuated silica glass ampoules for further physical characterization.
The temperature dependence of magnetization of FeSe12x
single crystals was measured by a ‘‘Quantum Design’’ SQUID magnetometer MPMS – 9 T. Transport properties of FeSe12x
were measured by a standard four-probe method. The super-conducting gap values in FeSe12xwere determined by
super-conductor–normal metal–superconductor (SnS) Andreev
spectroscopy. This method is based on multiple Andreev reflections effect16 occurring in ballistic junctions of a
SnS-type. ‘‘Ballistic’’ implies the contact diameter a to be larger than the quasiparticle mean free path l.17 To realize the experimental method, SnS contacts were formed in FeSe12x
samples by a ‘‘break-junction’’ technique.18The samples (thin plates with dimensions 2 6 1 6 0.1 mm3) were attached with two current and two potential leads to the sample holder by a liquid In–Ga alloy; the holder was then cooled down to liquid helium temperature T = 4.2 K. Forthcoming precise bending of the sample holder causes a microcrack generation into the sample: the microcrack was in fact a contact between two cryogenic clefts separated with a constriction. The layered structure of FeSe12ximplies exfoliation of the single crystals
along the ab-planes, which provides easy mechanical read-justment of the contact point, forming tens of contacts during an experiment. The microcrack location deep into the sample prevented any impurity presence at clean cryogenic clefts and the overheating of the junction.
Composition, crystal structure and
morphology
The shape of the crystals was characterized by scanning electron microscope images shown in Fig. 1–3. Tetragonal single crystals with dimensions 4 6 4 6 0.1 mm3 with well
shaped habitus were obtained at a temperature gradient of 710–660 K, as shown in Fig. 1 and 2. On the contrary, the samples prepared at higher temperatures, i.e. at a temperature gradient of 730–680 K, had an imperfect shape, nonstable chemical composition, and were mixed with the hexagonal Fe7Se8 impurity phase, as shown in Fig. 3. The chemical
composition of tetragonal FeSe12xphase was found to be xy
0.04 ¡ 0.02. The hexagonal phase was identified to be Fe7Se8
from the energy-dispersive X-ray analysis. No additional impurity ions, i.e. K, Al or Cl, were found in iron selenides.
The phase purity and quality of FeSe12xsingle crystals were
established by indexing the X-ray diffraction patterns. All peaks in the X-ray diffraction pattern of the single crystals ground in an agate mortar, shown in Fig. 4a, were ascribed to tetragonal P4/nmm (129) unit cell, a = 3.765 Å, c = 5.518 Å, V = 78.2(2) Å3. These values are in correspondence with those reported in ref. 19. The X-ray diffraction pattern of FeSe12x
single crystal represents only the (00l) reflexes, as shown in Fig. 4b. While the X-ray diffraction pattern of the
line sample synthesized for comparison by the solid-state reaction contains the peak attributed to an Fe3O4 impurity
phase in Fig. 4c.
A possible explanation for the absence of Fe3O4impurities
in the samples prepared from KCl–AlCl3is in its
transforma-tion to soluble form by the reactransforma-tion:
3Fe3O4+ 3Fe + 8AlCl3(liq)= 4Al2O3+ 12FeCl2(liq)
Therefore, the high quality of FeSe12xsingle crystals was
achieved primarily due to the application of the eutectic melt of salts including fusible aluminum chloride, and secondly due to the growth conditions of a permanent gradient of
temperature. The use of AlCl3–KCl eutectic mixture with the
melting temperature lower than 150uC allows the transport of both iron and selenium at temperatures much lower than that
Fig. 2 The scanning electron microscope image of the layered structure of a tetragonal FeSe12xsingle crystal.
Fig. 3 The scanning electron microscope images of the products obtained at high temperatures. 1, 2, 3, 6 – hexagonal Fe7Se8crystals; 4 – hexagonal Fe7Se8
crystals 5 – a tetragonal FeSe12xcrystal.
Fig. 4 The X-ray diffraction patterns of tetragonal FeSe12x. Panel a represents
the X-ray diffraction pattern of the single crystals ground in an agate mortar (Co Ka). Panel b represents the X-ray diffraction pattern along the c-axis (Cu Ka1).
Panel c represents the X-ray diffraction pattern of the polycrystalline sample synthesized by the solid-state reaction (Cu Ka1).
Fig. 1 The scanning electron microscope image of a tetragonal FeSe12xsingle
crystal.
of the dissociation of tetragonal FeSe (457 uC). Besides, application of AlCl3–KCl mixture instead of CsCl–LiCl flux
prevents lithium from reacting with quartz ampoules. The permanent gradient of temperatures allows growth of single crystals under well defined conditions as compared to that upon slow cooling of the melt.
Transport and magnetic properties
The quality of FeSe12xsingle crystals was examined by physical
measurements. The temperature dependence of resistance R(T), shown in Fig. 5, demonstrates a metallic behavior with Tc
= 9.4 K. This R(T) also indicates the high quality of the FeSe single crystal. For example, the kink at 90 K manifests the structural phase transition. The width of superconducting transition amounts to 1.5 K (see upper inset to Fig. 5). Magnetic measurements confirm the presence of the super-conducting phase transition around 9.4 K (lower inset to Fig. 5). Magnetic properties show a nearly 100% Meissner effect.
The superconducting gap values in FeSe12x found from
Andreev spectroscopy were analyzed according to Ku¨mmel et al.20 The current–voltage characteristic for an SnS contact exhibits an excess current at low bias and a subharmonic gap structure due to the multiple Andreev reflections effect, which reveals a series of minima for the dynamic conductance dI(V)/ dV at the bias voltages
Vn~ 2D
en, n~1, 2, 3, . . . (1)
where D is a superconducting gap value, e the elementary charge, n the subharmonic order. The relative intensity of the Andreev peculiarities decreases exponentially with increasing n. For a two-gap superconductor, such structures should appear in dI(V)/dV of the SnS contact.21Using eqn (1), one can
directly obtain a gap value from the corresponding Andreev minima positions without any fitting.
The experimental current–voltage characteristic in Fig. 6 (thin blue line) with an excess current at low bias voltages is typical for the clean classical SnS-Andreev contact. Therefore, the constriction exercises normal metal properties, and the theory by Ku¨mmel et al.20is applicable to the present
break-junctions. Dips of the contact dynamic conductance (the bold red lines in Fig. 6) located at VL1# ¡4.6 meV (n = 1; marked
as 2DL) and VL2# ¡2.4 meV (n = 2; marked as DL), in accord
with eqn (1), can be considered as the first and the second Andreev reflexes. Eqn (1) implies a linear dependence of Andreev dip positions versus their reversed number, Vn(1/n),
starting at (0;0) point. Such a dependence was plotted for the large gap dynamic conductance dips (inset of Fig. 6). The experimental points do follow eqn (1) and suggest the large gap DL= 2.4 ¡ 0.2 meV. This large gap value is not far from
the maximum value of 2.0 meV of the anisotropic gap found in the specific heat data.22The next peculiarities located at VS1#
¡1.5 meV (2DSlabels) do not correspond to the expected third
Andreev minima positions for the large gap as their intensity is much higher than that of the second dips for the large gap. Therefore, the minima located at VS1# ¡1.5 meV indicates a
new subharmonic gap structure, associated with the small gap DS = 0.75 ¡ 0.1 meV. Dynamic conductance for several SnS
contacts has been studied in the same way and the results are consistent. The peculiarities sharpness confirms a good quality of the single crystals due to the microscopic homo-geneity.
Fig. 5 The temperature dependence of resistance of a FeSe12xsingle crystal.
The lower inset shows the temperature dependence of magnetic susceptibility measured in the magnetic field BIc = 20 Oe in both ZFC (open circles) and FC (closed circles). The upper inset shows the enlarged superconducting transition region.
Fig. 6 Current–voltage characteristic (thin blue line) and dynamic conductance (bold red line) measured at T = 4.2 K for SnS-Andreev contact. Andreev dip positions define the large gap value DL# 2.4 meV (the corresponding minima
are marked as 2DLand DL) and the small gap DS# 0.75 meV (2DSlabels). Inset
shows a linear dependence of the large gap minima voltages Vnversus 1/n.
The value of 2DL/kBTC# 5.9 ¡ 1 exceeding the BCS value of
3.52 is close to those obtained in other works4,23,24using FeSe monolayers, poly- and single crystals.
Conclusions
We present the method for the growth of high quality FeSe12x
single crystals in KCl–AlCl3eutectic mixture under conditions
of a permanent gradient of temperature of 385–427uC. We also found that better quality of the samples can be achieved at lower temperatures. Physical characterizations indicates a superconducting state below 9.4 K. SnS-Andreev spectroscopy shows two superconducting gaps with DL= 2.4 ¡ 0.2 meV and
DS= 0.75 ¡ 0.1 meV. The large gap BCS-ratio of 2DL/kBTC#
5.9 ¡ 1 points to strong coupling in FeSe.
Acknowledgements
The authors thank T.N. Dokina, K.V. Van, A.A. Viryus, A.N. Nekrasov for technical support and Ya.G. Ponomarev for valuable discussions. This work was supported by a grant of the President of the Russian Federation for State Support of Young Russian Scientists (MK-1557.2011.5), and the Russian Foundation for Basic Research 12-02-90405, 12-02-90823, and Russian Ministry of Science and Education 11.519.11.6012 and 8378.
References
1 D. C. Johnston, Adv. Phys., 2010, 59, 803.
2 F. C. Hsu, J. Y. Luo, K. W. Yeh, T. K. Chen, T. W. Huang, P. M. Wu, Y. C. Lee, Y. L. Huang, Y. Y. Chu, D. C. Yan and M. K. Wu, Proc. Natl. Acad. Sci. U. S. A., 2008, 105, 14262. 3 Y. Mizuguchi, F. Tomioka, S. Tsuda, T. Yamaguchi and
Y. Takano, Appl. Phys. Lett., 2008, 93, 152505. 4 B. Bu¨chner and C. Hess, Nat. Mater., 2009, 8, 615. 5 T. P. Ying, X. L. Chen, G. Wang, S. F. Jin, T. T. Zhou, X.
F. Lai, H. Zhang and W. Y. Wang, Sci. Rep., 2012, 2, 426. 6 R. Schneider, A. G. Zaitsev, D. Fuchs and H. v. Lo¨hneysen,
Phys. Rev. Lett., 2012, 108, 257003.
7 L. Li, Z. R. Yang, Y. P. Sun, J. Y. Zhang, D. Z. Shen and Y. H. Zhang, Supercond. Sci. Technol., 2011, 24, 015010. 8 D. Liu, W. Zhang, D. Mou, J. He, Y.-B. Ou, Q.-Y. Wang, Z. Li,
L. Wang, L. Zhao, S. He, Y. Peng, X. Liu, C. Chen, L. Yu, G. Liu, X. Dong, J. Zhang, C. Chen, Z. Xu, J. Hu, X. Chen, X. Ma, Q. Xue and X.-J. Zhou, Nat. Commun., 2012, 3, 931. 9 H. Okamoto, J. Phase Equilib., 1991, 12, 383.
10 S. B. Zhang, X. D. Zhu, H. C. Lei, G. Li, B. S. Wang, L. J. Li, X. B. Zhu, Z. R. Yang, W. H. Song, J. M. Dai and Y. P. Sun, Supercond. Sci. Technol., 2009, 22, 075016.
11 R. Hu, H. Lei, M. Abeykoon, E. S. Bozin, S. J. L. Billinge, J. B. Warren, T. Siegrist and C. Petrovic, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 83, 224502.
12 U. Patel, J. Hua, S. H. Yu, S. Avci, Z. L. Xiao, H. Claus, J. Schlueter, V. V. Vlasko-Vlasov, U. Welp and W. K. Kwok, Appl. Phys. Lett., 2009, 94, 082508.
13 G. Kullerud, Experimental techniques in dry sulfide research in Research Techniques for High Pressure and High Temperature, ed. G. C. Ulmer, Spinger-Verlag, New York, 1971, pp. 288–315.
14 G. H. Moh and L. A. Taylor, Neues Jahrb. Mineral., Monatsh., 1971, 10, 450.
15 F. Gronvold, Acta Chem. Scand., 1968, 22, 1219. 16 A. F. Andreev, J. Exp. Theor. Phys., 1964, 19, 1228. 17 Yu. V. Sharvin, J. Exp. Theor. Phys., 1965, 48, 984. 18 J. Moreland and J. W. Ekin, J. Appl. Phys., 1985, 58, 3888. 19 G. Ha¨gg and A. L. Kindstrom, Z. Phys. Chem., 1933, 22, 455. 20 B. Ku¨mmel, U. Gunsenheimer and R. Nicolsky, Phys. Rev. B:
Condens. Matter, 1990, 42, 3992.
21 B. A. Aminov, A. A. Golubov and M. Y. Kupriyanov, Phys. Rev. B: Condens. Matter, 1996, 53, 365.
22 J.-Y. Lin, Y. S. Hsieh, D. A. Chareev, A. N. Vasiliev, Y. Parsons and H. D. Yang, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 84, 220507.
23 M. Bendele, S. Weyeneth, R. Puzniak, A. Maisuradze, E. Pomjakushina, H. Luetkens, S. Katrych, A. Wisniewski, R. Khasanov and H. Keller, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 81, 224520.
24 Ya. G. Ponomarev, S. A. Kuzmichev, M. G. Mikheev, M. V. Sudakova, S. N. Tchesnokov, T. E. Shanygina, O. S. Volkova, A. N. Vasiliev and Th. Wolf, J. Exp. Theor. Phys., 2011, 113, 459.
