FeSe-based superconductors with a superconducting transition temperature of 50 K

Due to the strong reactivity of alkaline metals and the easy formation of the impurity phase, the superconducting transition temperature (Tc) of alkaline metals intercalated FeSe is usually limited to 45 K. To avoid the formation of impurity and improve the Tc, we intercalate a more chemically inert organic ion (rather than the chemically reactive alkaline metals) into FeSe single crystal in this report. A new FeSe-based superconductor, namely (TBA)0.3FeSe, with Tc of 50 K, is synthesized by intercalating FeSe single crystal with organic ion tetrabutyl ammonium (TBA+) via an electrochemical intercalation method, which has the highest Tc among FeSe-based bulk superconductors. The structure of the organic ion intercalated product consists of the alternate stacking of monolayer FeSe and the organic molecule. The superconductivity of (TBA)0.3FeSe is confirmed by both the magnetic susceptibility and the transport measurement. It is suggested that the chemically inert organic ion should play a key role in the enhancement of Tc by avoiding the formation of impurity and disorder in FeSe plane as possible. We also suggest that the TBA+ intercalated FeSe with well defined shape and higher Tc offer a good playground for further bulk measurement investigation.


Introduction
Simple crystal structure [1], large pressure effect [2], and the highest superconducting transition temperature (T c ) above 65 K in monolayer FeSe/SrTiO 3 interface [3] make FeSe a fascinating system in iron-based superconductors. Under ambient pressure, the key factor to improve the T c of FeSe-based superconductors is to dope electrons to the FeSe plane to form intercalated structure or charge-transfer interface. The former includes the AFe 2 Se 2 (A is alkaline metals) obtained by high temperature solid state reaction [4], A x (NH 3 ) y Fe 2 Se 2 obtained by liquid-ammonia method [5][6][7], (Li, Fe)OHFeSe obtained by hydrothermal method [8][9][10], A x (M) y Fe 2 Se 2 (M is the organic solvent molecule, C 6 H 16 N 2 or C 2 H 8 N 2 ) obtained by electrochemical intercalation method [11][12][13] and alkaline metals and organic amine co-intercalated FeSe obtained by sonochemical insertion method [14]. The latter includes FeSe/SrTiO 3 interface [3,15], potassium coating at ultrahigh vacuum [16,17], and gating method [18][19][20]. In addition, the highest T c of the system containing interfaces is usually higher than that of bulk materials [3,6,20,21]. How to achieve the similar high T c in FeSebased bulk superconductors as that in the interface system, for example, at 50 K, is a promising and challenging work, which will help us to understand the real role that the interface plays.
Monolayer FeSe/SrTiO 3 interface system has been wildly investigated by STM and ARPES [22][23][24]. However, limited by its intrinsic small size and air sensitivity [25], traditional bulk characterization methods (xray diffraction, magnetic susceptibility, transport measurement, nuclear magnetic resonance, specific heat measurement, etc) are very hard to be conducted on these interface systems. In fact, consistent results between Original content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence.
Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. the magnetic susceptibility and transport measurement on monolayer FeSe/SrTiO 3 interface remain to be reported [25,26]. As a result, synthesizing a high T c FeSe-based bulk superconductor is of high importance and will offer a new platform for bulk measurement. Our motivation is to find new FeSe-based bulk superconductors with high T c so as to provide new research opportunity. Therefore, we concentrate our attention on the doped FeSe with intercalated structure rather than the interface system.
There are two ways to synthesize FeSe-based superconductors with intercalated structure. The first is the high temperature solid state reaction method [4]. However, there exists many Fe vacancies in FeSe plane or even phase separation problem in AFe Se 2 2 obtained from high temperature solid state reaction [27,28]. The above problems lead to a lower T c of only 32 K and make it a dirty system for physical measurement [4,27,28]. The second method using low temperature to synthesize the target compound avoids the above problems. Liquidammonia method [5][6][7], hydrothermal method [8][9][10], and electrochemical intercalation method [11,12] are good ways to obtain high T c sample with intercalated structure. However, it is very hard to be conducted on FeSe single crystal. In fact, although many alkaline metals can be intercalated into the interlayer of FeSe polycrystalline sample by liquid-ammonia method or electrochemical method, only less reactive metal Li rather than Na or K intercalated FeSe single crystal sample is reported [21]. We attribute it to the easier formation of A 2 Se (A is the akaline metals) impurity for Na and K with the increase of reactivity. Recently, the hydrogen intercalated FeSe 0.93 S 0.07 single crystal is also reported to have an improved T c although the intercalation is not uniform [29]. The above results inspire us to intercalate new intercalator that is chemically inert in the sequence table of metal activity. For example, the less reactive metal ion or aprotogenic ion.
After the above analysis and consideration, we try to intercalate less reactive metal ion and aprotogenic ion. To our disappointment, the less reactive metal cannot be intercalated into FeSe due to the electrodeposition of the less reactive metal. As for the aprotogenic ion, we do successfully intercalate organic quaternary ammonium ion into FeSe single crystal and obtain exciting results.
Hereafter, we report the successful intercalation of organic ion TBA + into FeSe single crystal. The structure of the intercalated product contains the alternate stacking of monolayer FeSe and the organic molecule, which is evidenced by the x-ray diffraction and TEM. The superconductivity of the intercalated product is confirmed by both the magnetic susceptibility and the transport measurement. Furthermore, by applying external pressure (0-2.46 GPa), the T c of (TBA) 0.3 FeSe is gradually suppressed to 29.3 K at 2.46 GPa from 50 K at ambient pressure, with a negative pressure effect on superconductivity with dT c /dP=−8.4 K GPa −1 . This is the highest T c among FeSe-based bulk superconductors at ambient pressure.
The FeSe single crystal is synthesized using AlCl 3 /KCl as the flux according to a reported paper elsewhere [30]. Firstly, KCl (Aladdin, 99.99%) is dried at 200°C for 10 h in an Argon gas glove box with O 2 and H 2 O amount less than 1 ppm to remove the adsorbed water. Secondly, the KCl and AlCl 3 (Alfa-Aesar, anhydrous, 99%) are mixed and fully grounded in the mole ratio of 1:2 in an argon glove box. Thirdly, Fe powders (Aladdin, AR, 100 mesh) and Se powders (Aladdin, 99.999%, 80 mesh) are mixed and grounded at the mole ratio of 1:0.94 and then mixed and grounded with the flux (AlCl 3 /KCl) with the mole ration of Fe:Se:AlCl 3 :KCl=1:0.94:7:3.5. At last, the above mixture is sealed into a silica ampoule and transported from 390°C to 260°C for 30 d. Shining crystal can be obtained after the remove of flux (AlCl 3 /KCl) by rising products into the deionized water.
The (TBA) 0.3 FeSe is synthesized through an electrochemical intercalation process using FeSe single crystal as the starting materials. Firstly, FeSe single crystal is weighted using a microgram balance (AX 26). Secondly, the weighted FeSe single crystal is fixed onto an indium wire, which is used as the positive electrode. The negative electrode is composed of a silver piece. The electrolyte is obtained by dissolving 6 g TBAB (Aladdin, AR, 99.0%) into 20 ml DMF (Innochem, 99.9%, extra dry with molecular sieves, water less than 50 ppm). At last, the above electrodes are inserted into the electrolyte and a constant current (20-30 μA) is set to pass through the electrolytic cell. During the current passing through the electrolytic cell, the negative electrode loses electrons while the positive electrode obtains electrons. The above electrochemical reaction can be described as the following equations: The doping amount x in (TBA) x FeSe is controlled by adjusting the current passing through the electrolytic cell and the time through the Faraday law. For a fixed doping amount x, the time t needed can be calculated using the following formula where t is the time (s), F is the Faraday constant (96 485.31 C mol −1 ), m is the mass (g) of the FeSe single crystal, x is the doping amount, M is the molar mass of FeSe (g mol −1 ), and I is the current (A). The x is set as 0.3 in this report to obtain a fully intercalated product with pure phase and typical discharge curve is shown in figure S1 available online atstacks.iop.org/NJP/20/123007/mmedia. It is necessary to note that the sample is very sensitive to air and moisture. The x-ray diffraction pattern of the sample is collected on a diffractometer (Rigaku SmartLab 9) equipped with Cu Kα radiation and a fixed graphite monochromator. The Fourier transform infrared (FTIR) spectroscopy is obtained on a Nicolet 8700 infrared spectrometer. The cross-section transmission electron microscopy (TEM) images of FeSe single crystal is captured from a Talos F200X microscope at 200 kV. The TEM images of (TBA) 0.3 FeSe are obtained from a H7700 microscope at 100 kV. The magnetic susceptibility measurement is conducted on a Quantum Design magnetic property measurement system. The resistivity measurement is carried out on a physical properties measurement system with the standard four-terminal method. Resistance measurement under hydrostatic pressure (0-2.46 GPa) is conducted on a Quantum Design high pressure piston-cylinder cell using Daphne 7373 oil as pressure transmitting medium. Figure 1(a) shows the optical images of the FeSe single crystal and the intercalated product. The morphology of the intercalated FeSe keeps a well defined shape as that of FeSe single crystal. The area of ab plane nearly does not change with only the enlargement in c axis, indicating the intercalation of TBA + ion. We note that such electrochemical intercalation process can be conducted on very big FeSe single crystal with the mass of FeSe larger than 10 mg, which is helpful for bulk measurement. As a demonstration, figure S1 shows the specific discharge curve of a FeSe single crystal with the mass 12.62 mg. The doping amount in (TBA) x FeSe is determined by the discharge curve with the integral of current versus time, and x is 0.3 in our report. We could not intercalate more TBA + ion into the FeSe single crystal by extending discharging time due to the decomposition of the electrolyte. The well defined morphology of the intercalated FeSe is helpful for further physical characterization including resistance measurement. Figure 1(b) shows the FTIR spectroscopy of FeSe, TBAB and (TBA) 0.3 FeSe. There is no obvious absorption band for FeSe, but the FTIR spectroscopy of (TBA) 0.3 FeSe is very similar to that of TBAB, indicating the intercalation of TBA + ion into FeSe. The x-ray diffraction pattern of FeSe and (TBA) 0.3 FeSe is shown in figure 1(c). Before the intercalation of TBA + ion, FeSe single crystal shows a sharp diffraction peak at around 16°, which reflects the distance of the adjacent FeSe layers (0.55 nm) [1]. However, after the intercalation of FeSe with TBA + ion, the diffraction pattern of (TBA) 0.3 FeSe shows series (00l) diffraction peaks and can be indexed with a lattice parameter of 1.55 nm for c axis. In view of the adjacent distance of FeSe layers (0.55 nm) and the size of TBA + ion (0.84-1.19 nm, depending on the orientation of the TBA + , more discussion is shown in the supporting information) [31][32][33][34][35][36], a crystal model with the stacking of monolayer FeSe and TBA + organic molecule is shown in figure 1(d). The structure model of (TBA) 0.3 FeSe will be confirmed using TEM later. At last, it is necessary to note that the (TBA) 0.3 FeSe sample is not stable in the air. More specific data (figure S3) and discussion is shown in the supporting information. More discussion on the structure of the organic ion intercalated FeSe is also shown in the supporting information. The above results indicate that an organic ion intercalated FeSe-based superconductor has been synthesized.

Results and discussions
The structure evolution process from FeSe to (TBA) 0.3 FeSe is investigated using TEM, as is shown in figure 2. Before the intercalation of TBA + ion, the distance of the adjacent FeSe layers ( figure 2(a)) is measured to be 0.55 nm, which is consistent with the lattice parameter of FeSe crystal in c axis [1]. After the intercalation of TBA + , the distance of the adjacent FeSe layers obtained from the TEM image is 1.49 nm ( figure 2(b)), which is close to the value 1.55 nm obtained from XRD pattern. The smaller value in TEM image is possible due to the small angle deviation between the e-beam and c axis of (TBA) 0.3 FeSe. We note that the organic ion TBA + is not observed in the TEM image due to the small diffraction contrast and the possible disorder of the organic molecule. These TEM images confirm the structure model shown in figure 1(d).
The magnetic susceptibility and the transport measurement results of (TBA) 0.3 FeSe are shown in figure 3.   Such results indicate that impurities and defects in FeSe plane should be very small and the intercalation process is uniform because a small difference means that the fraction of sample volume where the magnetic flux is pinned due to defects or impurities is small. Figure 3(b) shows the magnetic susceptibility as a function of magnetic field (M-H curve) at 15 K. The M-H curve shows a typical magnetic hysteresis profile of type-II superconductors and the lower critical field H c1 is about 890 Oe. It is necessary to emphasize that the T c of pristine FeSe is about 8.9 K [1], which is lower than 15 K. Furthermore, the transport measurement of (TBA) 0.3 FeSe is shown in figure 3(c). The resistance of (TBA) 0.3 FeSe drops sharply at 50 K, and goes to zero at 42 K. The mid-point superconducting transition temperature (T c mid ) obtained from the resistance curve is 48 K, consistent with the T c obtained from the magnetic susceptibility curve ( figure 3(a)). It is necessary to note that the temperature where the resistance departs from linear is more than 55 K, a little higher than the value observed on the transport measurement result conducted on monolayer FeSe film grown on undoped SrTiO 3 (54.5 K) [37]. The above results suggest that a new FeSe-based superconductor with T c onset of 50 K has been synthesized by intercalating FeSe with TBA + ion. The transport measurement of (TBA) 0.3 FeSe under different magnetic field is also conducted and the result is shown in figure 3(d).Under external magnetic field, the T c onset of (TBA) 0.3 FeSe nearly does not change but the transition width is enlarged, more systematic measurements are necessary to be conducted in TBA + intercalated FeSe.
At last, the resistance measurement under external pressure (0-2.46 GPa) is also conducted and the result is summarized in figure 4. Under external pressure, the T c onset of (TBA) 0.3 FeSe is gradually suppressed to 29.3 K at 2.46 GPa from 50 K at ambient pressure, with a negative pressure effect with dT c /dP=−8.4 K GPa −1 . Such negative pressure effect is frequently observed in the T c -pressure phase diagram of the intercalated FeSe-based superconductor. Usually, there are two domes in the intercalated FeSe-based superconductors [38][39][40]. When the pressure is lower than the critical pressure P c , the T c decreases with the increase of pressure. However, when the pressure is above P c , the T c increases as the increase of pressure, going into the SC-II region. The P c for Li x (NH 3 ) y Fe 2 Se 2 and Li 0.8 Fe 0.2 OHFeSe is 5 GPa and 2 GPa, respectively [38,39]. Our negative pressure effect at the pressure range of 0-2.46 GPa indicates that the P c of (TBA) 0.3 FeSe is higher than 2.46 GPa. Measurement under higher pressure needs to be conducted for further investigation.
In the previous work, FeSe-based superconductors obtained from electrochemical intercalation method or liquid-ammonia method shows a smaller T c in magnetic susceptibility curve [5-7, 11, 12, 41, 42], typically limited to 45 K. Most of the intercalated sample contains lots of grain boundary, which makes it difficult to conduct resistance measurement. Furthermore, compared with the organic ion, the alkaline metals should be very reactive and easier to form the impurity phase A 2 Se (A is alkaline metals) because the smaller ionic radius of the alkaline metals results in a stronger interaction with the FeSe layer. At last, the chemically inert organic ion should avoid the disorder in FeSe plane as possible compared with the alkaline metals intercalated FeSe, which is benefit for the emergence of superconductivity.
We attribute the high T c in (TBA) 0.3 FeSe to the electron doping to FeSe plane and a smaller interaction between the FeSe and organic ion avoiding the formation of impurity phase and disorder in FeSe plane as possible. We note that the FeSe layers show a wave like shape to some extent in the HRTEM image ( figure 2(b)), which is possible induced by the disordered distribution and orientation of the TBA + in the organic layer, such disorder should be against superconductivity. However, such disorder in FeSe plane induced by the disorder in organic layer should be much smaller than that in alkaline metals intercalated case due to the smaller interaction between the organic layer and FeSe than that between alkaline metals and FeSe. It suggests that the TBA + intercalated FeSe with well defined shape and higher T c offers a new platform for the investigation of superconducting mechanism and shields new light on the searching for superconductors with higher T c .
At last, we recall that negative ions including small inorganic ion, for example, -AlCl 4 and -ClO 4 , and large organic ion TFSI − have also been reported to be intercalated into the interlayer of 2D layered materials [43][44][45][46]. How to dope hole by intercalating such chemically inert negative ion into FeSe will help us to demonstrate a full electronic phase diagram for FeSe-based superconductors, which is very important for the construction of a universal picture for FeSe-based superconductivity.

Conclusions
Here, we have shown that the T c of FeSe could be improved from 8.9 to 50 K by intercalating FeSe with TBA + ion. The enhanced T c should be due to the charge transfer of electrons into the FeSe plane [16][17][18][19][20]. The intercalated sample keeps a well defined morphology and makes it possible for bulk measurement to be conducted.
In summary, a new FeSe-based bulk superconductor with T c of 50 K, is synthesized by intercalating FeSe single crystal with organic ion tetrabutyl ammonium (TBA + ) via electrochemical intercalation method, which has the highest T c among FeSe-based bulk superconductors. The structure of the organic ion intercalated product consists of the alternate stacking of monolayer FeSe and the organic molecule. The chemically inert organic ion should avoid the formation of impurity and disorder in FeSe plane as possible. Our finding provides a new platform for the understanding of the superconductivity in FeSe-based superconductors.