Ionic-liquid-gating induced protonation and superconductivity in FeSe, FeSe0.93S0.07, ZrNCl, 1T-TaS2, and Bi2Se3

TL;DR

This study demonstrates ionic-liquid gating induces protonation in various compounds, significantly enhancing their superconducting transition temperatures and enabling detailed post-gating measurements, revealing insights into proton placement and superconductivity mechanisms.

Contribution

It introduces a method for protonating multiple materials via ionic-liquid gating, achieving high-temperature superconductivity and enabling post-gating analysis.

Findings

Superconducting transition temperature of FeSe increased to 43.5 K after protonation.

Protonation induced superconductivity in ZrNCl, 1T-TaS2, and Bi2Se3 with specific T_c values.

Proton placement near anions confirmed by NMR and theoretical calculations.

Abstract

We report protonation in several compounds by an ionic-liquid-gating method, with optimized gating conditions. This leads to single superconducting phases for several compounds. Non-volatility of protons allow post-gating magnetization and transport measurements. The superconducting transition temperature $T_C$ is enhanced to 43.5~K for FeSe$_{0.93}$S$_{0.07}$, and 41~K for FeSe after protonation. Superconductivity with $T_c$$\approx$15~K for ZrNCl, $\approx$7.2~K for 1$T$-TaS$_2$, and $\approx$3.8~K for Bi$_2$Se$_3$ are induced after protonation. Electric transport in protonated FeSe$_{0.93}$S$_{0.07}$ confirms high-temperature superconductivity. Our $^{1}$H NMR measurements on protonated FeSe$_{1-x}$S$_{x}$ reveal enhanced spin-lattice relaxation rate $1/^{1}T_1$ with increasing $x$, which is consistent with LDA calculations that H$^{+}$ are located in the interstitial sites close to the anions.

Tables

Table 1. Table 1: T C subscript 𝑇 𝐶 T_{C} of the materials before and after protonation

Compound	FeSe	FeSe0.93S0.07	ZrNCl	1T-TaS2	Bi2Se3
$T_c$ before protonation	9 K	8 K	0	0	0
$T_c$ after protonation	41 K	43.5 K	15 K	7.2 K	3.8 K

Full text

Carrier doping is an effective method for tuning metal-insulator transitions and superconductivity. In addition to chemical substitution, electric gating also emerged as an efficient method for tuning carrier density in thin films. Ahn_RMP_2006 ; Ueno_NM_2008 ; Saito_Science_2015 With the development of various room-temperature ionic liquids, the transistor-like gating method Ye_NM_2010 ; Bollinger_Nature_2011 ; Ye_Science_2012 ; Li_Nature_2015 ; Lu_Science_2015 ; Miyakawa_PRM_2018 was found to induce a large carrier density for thin films or crystal flakes, through charge polarization. Lithium doping by ionic solid gating was also found to enhance the superconducting transition temperature of thin flakes of FeSe ChenXH_PRB_2017 . Recently, tuning of proton or oxygen concentration, using ionic-liquid-gating as a medium, was introduced to modify the lattice structure and magnetism of SrCoO2.5 Yu_nature_2017 . This H+ implantation method was later applied in iron-based superconductors, to induce superconductivity or enhance the superconducting transition temperature in bulk crystals due to an electron doping effect Cui_SciBull_2018 .

It is important to note that H+ originates from water contamination in the ionic liquid by this method Yu_nature_2017 . The advantages of this technique are that H+ is nonvolatile and the gating is performed near the ambient conditions, which allow various post-gating measurements. However, multiple superconducting phases in protonated FeSe1-xSx emerge, indicating that proton concentration is inhomogeneous across the bulk crystals. The magnetization data in the protonated FeSe0.93S0.07 show that the volume fraction of the superconducting phase is very low under the reported gating conditions Cui_SciBull_2018 .

In this Letter, we report our optimized protonation conditions with this ionic-liquid-gating method, to improve the superconducting volume ratio and the doping homogeneity. The best protonation temperature is found to be 350 K (higher than the room temperature), with a gating period of 12 days. For FeSe0.93S0.07, the superconducting volume ratio is largely enhanced compared to the room temperature gating, as determined by the magnetization measurement. Transport measurement is also succeeded to confirm superconductivity. We also apply the optimized protonation on various layered compounds, including FeSe, insulating ZrNCl, 1$T$-TaS2, and Bi2Se3, where protonation either induces superconductivity or enhances the $T_{\rm c}$ largely. In particular for 1$T$-TaS2, we achieve a $T_{\rm c}$ higher than the regular gating method.

In our experiment, pristine FeSe and FeSe0.93S0.07 single crystals were made by the vapor transport method Bohmer_PRB_2013 ; Shibauchi_PNAS_2016 . ZrNCl powders were made by the high-pressure synthesis ChenX_JPCM_2002 . The 1$T$-TaS2 single crystal was grown by the chemical vapor transport method Kuwabara_1986 . Bi2Se3 was grown by the flux method Sultana_JSNM_2017 . FeS single crystal was made by the hydrothermal method Borg_PRB_2016 . We employ the protonation technique as illustrated in Fig. 1. As shown in Fig. 1, samples are attached to the negative electrodes, and a voltage of 3.0 V is applied as the gating voltage. The ionic liquid EMIM-BF4 is used. The gating temperature is optimized to be 350 K, which improves proton diffusion efficiency in the crystal. Typical gating period is 12 days when water is nearly fully electrolyzed. The dc magnetization is measured with a magnetic property measurement system (MPMS), and the transport is measured with a physical property measurement system (PPMS). These measurements were successfully performed after gating was removed at the room temperature, which indicates nonvolatile protons are inserted, in contrast to conventional ionic-liquid gating where gating cannot be removed during measurements. The proton NMR is performed by the spin-echo method, and the spin-lattice relaxation rate $1/^{1}T_{1}$ is measured by the inversion recovering method.

In the following, we present protonation measurements on these compounds.

FeSe. Recently, FeSe has attracted a great deal of research attention because of its highly tunable superconductivity. Its $T_{c}$ is enhanced from 8.5 K to above 40 K under high pressure Felser_NatMater_2009 ; sun_NC_2016 , by chemical intercalation ChenXL_PRB_2010 ; Hatakeda_JPSJ_2013 ; Dong_JACS_2015 ; ChenXH_NM_2015 , by ionic-liquid/solid gating ChenXH_PRL_2016 ; ChenXH_PRB_2017 , or by dimensional reduction into a single-layer phase XueQK_CPL_2012 .

Fig. 2 shows the dc susceptibility $\chi$(T) of a protonated FeSe single crystal. A rapid drop of $\chi$ are clearly at 41 K seen, indicating the onset of superconductivity. Therefore, the $T_{c}$ of FeSe is also largely enhanced by the protonation technique.

FeSe0.93S0.07. FeSe1-xSx is a series of compounds, whose $T_{c}$ ranges between 8 K and 13 K for x$<$0.12 Takano_JPSJ_2009 . Previously, two superconducting transitions, at 25 K and 42.5 K were reported in the protonated sample. Here we show that with increased protonation temperature at 350 K, a single high-$T_{c}$ phase is realized. Figure 3 shows the dc susceptibility $\chi$(T) and the resistance data R(T) of a protonated FeSe0.93S0.07 single crystal. The susceptibility data shows $T_{c}$ $\approx$ 41 K, seen by the drop of $\chi$ (Fig. 3 (a)). By contrast, the resistance data shows a higher onset $T_{c}$ of 43.5 K as indicted in Fig. 3, by a sudden drop of resistance upon cooling.

ZrNCl. ZrNCl is a layered material with electric gating or lithium doping. Superconductivity can be induced in ZrNCl by electric gating or lithium doping Ye_NM_2010 ; Saito_Science_2015 ; Taguchi_PRL_2006 . We pressed ZrNCl powders into thin pellets and then doped H+ with the current ionic-liquid-gating method. The samples turn from blue into black upon proton doping. Figure 4 shows the dc susceptibility of the proton-doped ZrNCl. The sharp drop of $\chi$ below 15 K shows the onset of superconductivity, with field up to 1000 Oe. The volume ratio of the superconducting phase, estimated from the ZFC data at 10 Oe field, is about 12$\%$. This suggests that proton doping is very efficient. We note that an ionic-liquid gating on ZrNCl at low temperatures is also reported, which proposes that the depletion of Cl*-* concentration causes superconductivity zhangshuai .

1T-TaS2. 1$T$-TaS2 is a layered compound with a triangular lattice. It goes through a series of charge-density-wave (CDW) transitions upon cooling Wilson_1975 ; Thomson_1994 . Superconductivity can be achieved by chemical doping, where the highest $T_{\rm c}$ is achieved at 3.5 K Liu_APL_2013 . Here we performed protonation on 1$T$-TaS2 single crystals. The dc magnetization of a protonated sample is shown in Fig. 5, measured under FC and ZFC conditions at different fields. The superconducting transition temperature $T_{\rm c}$ is found to be $\sim$7.2 K under 10 Oe field. We note that this transition temperature is higher than that achieved by the chemical doping. With an applied field of 500 Oe, the superconducting transition is still observed.

Bi2Se3. As a topological insulator, Bi2Se3 has caused a lot of research interests zhang_NP_2009 ; Xia_NP_2009 . Superconductivity can be achieved upon Cu or Sr doping into this material Hor_PRL_2010 ; Liu_JACS_2015 . Here we find that by protonation, superconductivity can also be achieved. As seen in Fig. 6, the superconducting transition temperature $T_{\rm c}$ is found to be 3.8 K, which is close to that reported by the chemical doping. With an applied field of 500 Oe, superconductivity is highly suppressed. Since protonation does not induce chemical substitution, our study indicates that chemical doping in the interstitial sites is important for the occurrence of superconductivity. Further studies on the protonation of induced superconductivity in this compound, regarding to possible topological superconductivity, are demanded.

1H NMR studies on HyFeSe1-xSx: In HyFeSe1-xSx compounds, intrinsic 1H NMR spectra was observed. Figure 7 shows the 1H spin-lattice relaxation rates divided by temperatures, versus temperature for protonated FeSe1-xSx with $x$=0, 0.07 and 1. Above 50 K, $1/^{1}T_{1}T$ stays nearly constant but varies with $x$, which indicate that doped protons are detected by the current measurements. Indeed, the increase of $1/^{1}T_{1}$ with increasing $x$ is consistent with the LDA calculations that H+ is inserted in the interstitial sites as discussed below. Since the $c$-axis lattice parameter is reduced with increasing $x$ Takano_JPSJ_2009 , the hyperfine coupling between 1H and the FeSe plane increases with increasing S2- concentration.

Our LDA calculations indicate that H+ is located in the interstitial sites close to the anion Se2-/S2-, as shown by the schematic drawing in the inset of Fig. 7. This can be understood as an effect of coulomb attraction between H*+* and Se2-/S2-. So far, we have found that this doping method is efficient in layered compounds, which indicates that H+ is most likely doped between the layers as in HyFeSe1-xSx.

Discussions and summary. Our XRD measurement did not resolve the change of lattice structure after protonation, which suggests that the chemical pressure effect of proton insertion is possibly very small. As a result, an electron-doping should be primarily responsible for the change of $T_{\rm c}$.

In Table 1, we summarize all the $T_{\rm c}$ of different compounds, before and after protonation under the current optimized conditions. The optimization at 350 K suggests that the efficiency of proton doping is caused by a balance between proton diffusion into the sample and the evasion out of the sample, both of which increase with temperature. The current method supplies a universal electron doping method, which could be widely used in tuning and searching for superconductivity and metal-insulator transitions in the layered compounds.

Work at RUC was supported by the National Natural Science Foundation of China with Grand Nos. 51872328, 11622437, 11574394, 11774423, and 11822412, the Strategic Priority Research Program of Chinese Academy of Sciences (Grant No. XDB30000000), the Ministry of Science and Technology of China with Grand No. 2016YFA0300504, the Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China (RUC) (15XNLQ07, 18XNLG14, 19XNLG17). SJ was supported by the National Natural Science Foundation of China with Grand Nos. 11774007 and U1832214. YC was supported by the Outstanding Innovative Talents Cultivation Funded Programs 2018 of Renmin University of China. JQY was supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Division of Materials Sciences and Engineering.