Synthesis and superconductivity in yttrium superhydrides under high pressure

Introduction

The search for high-temperature superconductors (HTS) with superconducting transition temperature (T_c) above liquid-nitrogen temperature has long been recognized as an intriguing topic since the discovery of Hg with T_c = 4.2 K.^[1] According to the Bardeen–Cooper–Schrieffer theory,^[2] metallic hydrogen (MH) is one of the best candidates for achieving HTS; however, the quest for MH has proven extremely challenging due to the requirements of ultrahigh pressure conditions. Satterthwaite et al. discovered ∼ 8 K superconductivity in thorium hydride in 1970, implying that hydrogen-rich metal hydrides would be HTS.^[3] Then, Gilman^[4] and Ashcroft^[5] further proposed that MH could be achieved in hydrogen-rich compounds at lower pressures because the heavier atoms played a chemical precompression role in hydrogen, ushering in a new era of HTS research in hydrogen-rich compounds at high pressures. However, despite significant efforts, there were no experimental breakthroughs for a long time until the observation of 203 K superconductivity at 155 GPa in covalent H₃S,^[6] which further inspired the search for HTS in conventional phonon-mediated hydride superconductors.

In contrast to covalent superhydrides such as H₃S, ionic metal hydrides offer more options for finding HTS. Wang et al. (2012) predicted the first CaH₆ clathrate hydride with a very high T_c of 235 K at 150 GPa.^[7] Following this study, a long list of clathrate REH₆, REH₉, and REH₁₀ superhydrides (RE: rare earth metal) were predicted to have high T_c values close to or even above room temperature.^[8–10] Stimulated by these predictions, a series of clathrate superhydrides, such as CaH₆,^[11] LaH₁₀,^[12,13] CeH₉, CeH₁₀,^[14] ThH₉, ThH₁₀,^[15] (La, Y)H₁₀,^[16] were successfully synthesized with T_c ranging from 57 K–260 K. Among ionic superhydrides, yttrium superhydrides piqued the interest of researchers due to their abundant stoichiometries, they are predicted to have high T_c, e.g., 84 K–95 K at 120 GPa in YH₄,^[10,17] 251 K–264 K at 120 GPa in YH₆,^[10] 21.5 K–43 K at 165 GPa in YH₇,^[18] 253 K–276 K at 150 GPa in YH₉,^[8] and 305 K–326 K at 250 GPa in YH₁₀.^[9] Recently, Kong et al.^[19] successfully synthesized YH₄ and clathrate structured YH₆ and YH₉ with observed T_cs of 220 K at 183 GPa and 243 K at 201 GPa for the last two yttrium superhydrides, respectively. Meanwhile, Troyan et al. also independently synthesized a clathrate YH₆,^[18] with an observed T_c of 224 K at 166 GPa. Following that, Snider et al. synthesized YH₉ with a T_c of up to 262 K using catalytic hydrogenation at about 182 GPa.^[20] Furthermore, recent research has successfully observed 88 K superconductivity of YH₄ at 155 GPa.^[21]

Besides binary yttrium superhydrides, yttrium-bearing ternary hydrides, where the introduction of a third element other than hydrogen considerably expands the phase space, have attracted extensive attention. Liang et al.^[22] and Xie et al.^[23] predicted a clathrate CaYH₁₂ with an estimated T_c of 258 K at 200 GPa and 230 K at 180 GPa, respectively. Then, Liang et al. predicted a ternary YSH₆ with a T_c of 91 K at 210 GPa.^[24] (La, Y)H₆ and (La, Y)H₁₀^[16] were synthesized experimentally at high pressures with T_cs of 237 K and 253 K, respectively.

Previous research has primarily concentrated on HTS (T_c > 200 K), even though more superhydrides have been synthesized. Thus far, there has been a dearth of efforts to systematically investigate the superconductivity of all experimentally reported unconventional superhydrides. In this work, we first conducted detailed structure and superconductivity studies of YH₄, which was chosen as an example due to its rare previous investigation. X-ray diffraction measurements revealed the successful synthesis of predicted I4/mmm-YH₄ at about 167 GPa and 1600 K, and its measured T_c of 82 K was evidenced by a sharp drop in resistance and a characteristic decrease in superconducting transition under a magnetic field up to 8.5 T. Further electrical transport measurements revealed a series of additional superconducting transitions at 29 K (162 GPa), 218 K (165 GPa), and 230 K (300 GPa), which arise from YH₇ and clathrate structured YH₆, and YH₉, respectively, inferred from T_cs consistency with previous studies.

Experimental methods

According to the different target pressures, symmetric diamond anvil cells (DACs) outfitted diamond anvils with a culet size of ∼ 30 μm–60 μm beveled at 8.5° to a diameter of ∼ 250 μm. The composite gasket was composed of rhenium outer annulus and a mixture of epoxy resin and Al₂O₃ powder. The insulating gasket was pre-indented to a thickness of 10 μm, and the corresponding sample chamber with a diameter of 20 μm–30 μm was drilled using a laser drilling system. Commercially available yttrium ingot (Alfa Aesar, 99.9% purity) and NH₃BH₃ (AB) powder (Sigma-Aldrich, 97%) were loaded into the sample chamber inside a glovebox filled with Ar atmosphere with O₂ and H₂O contents of < 0.01 ppm. The Y foil and Au electrodes with thicknesses of 2 μm and 1 μm, respectively, were sandwiched between the AB layers. The application of Au electrodes can effectively avoid the chemical reaction^[25] between the electrodes and hydrogen, which can result in the formation of undesirable superconductors, as well as help to maintain a hydrogen-rich environment. AB serves as a hydrogen source while also acting as thermal insulation layers. Subsequently, the samples were compressed to the required synthesis pressure. The pressure in the sample chamber was calibrated using the high-frequency edge of the diamond Raman line.^[26] The laser heating of the sample was performed using a pulsed YAG infrared laser, and the temperature was determined using the black-body radiation fit within the Planck function. In situ high-pressure angle-dispersive x-ray diffraction (ADXRD) experiments were performed at the Shanghai Synchrotron Radiation Facility’s BL15U1 beamline (5 μm × 12 μm) with a monochromatic beam wavelength of 0.6199 Å and an average acquisition time of 120 s. Before the experiment, the relevant geometric parameters were calibrated using a CeO₂ standard. Diffraction patterns were collected using a Mar165 CCD detector and analyzed using DIOPTAS software, yielding one-dimension profiles.^[27] The Le Bail profile matching refinements were performed using the GSAS + EXPGUI programs.^[28] Based on the four-probe van der Pauw method,^[29] the resistance measurements were performed with currents of 10⁻⁶–10⁻⁴ A (Keithley 2182A nanovoltmeter and 6221 AC and DC source) and the selected data were warming cycles with a controlled rate of approximately 1 K ⋅ min⁻¹. Furthermore, non-magnetic DACs made of Be–Cu alloy were used for resistance measurements in an external magnetic field of up to 8.5 T.

Results and discussion

In this work, we prepared 11 samples, labeled as samples 1 through sample 11, to synthesize yttrium superhydrides from a mixture of Y and AB, and explore their superconductivity. Previous excellent results have shown AB to be a reliable H₂ source.^{[11,13,18,19,21]} At high temperatures, AB would decompose into H₂ plus c-BN, the latter avoiding the problem of poor contact between the synthesized product and electrodes. The diagram of the assembly used for synthesis and four-probe electrical resistance measurements is shown in Fig. 1(a). In sample 1, the reactants were compressed to 167 GPa [Fig. S1(a)] before being heated to about 1600 K. The clear H–H vibration from H₂ molecular [Fig. S1(b)] demonstrates a hydrogen-rich environment. The sample turned black after laser heating, indicating that a chemical reaction occurred [inset in Fig. 1(b)]. Representative electrical resistance measurements as a function of temperature reveal a superconducting transition at 82 K, as evidenced by the sharp drop in the resistance, as shown in Fig. 1(b). This superconducting transition can be perfectly reproduced in several independent experiments (Fig. 2 and Fig. S2), further confirming the reliability of our results. To determine the highest value of T_c, we evaluated the pressure dependence of T_c, as shown in Fig. 2(b). T_c fluctuates in the pressure range of 145 GPa–170 GPa in different experiments and the highest T_c of 84 K at 162 GPa is consistent with the previous theoretical estimate of 84 K–95 K for YH₄. Furthermore, as pressure decreases, the superconducting transition disappears at about 143 GPa [Figs. S2(a) and S2(b)], indicating a possible superconducting phase decomposition.

View original graphic| Download| PPT slide

Fig. 1.   (a) Schematic of the experimental setup for synthesis and four-probe superconducting electrical resistance measurements. (b) Temperature dependence of resistance in sample 1 (S1) at 167 GPa. The insets show an optical micrograph of the sample before and after laser heating. The value of the T_c is defined as the crossing point of the resistance slopes before and after the resistance drop. (c) Synchrotron XRD pattern of S1 at 167 GPa. The inset displays a two-dimensional XRD pattern. Unidentified weak reflections are marked by asterisks. (d) Crystal structures of I4/mmm-YH₃ and I4/mmm-YH₄. Big and small balls represent Y and H atoms, respectively.

View original graphic| Download| PPT slide

Fig. 2.   (a) Temperature dependence of resistance in sample 3 (S3) at 162 GPa. Inset: crystal structures of Imm2-YH₇. Big and small balls represent Y and H atoms, respectively. (b) Pressure dependence of T_c for I4/mmm YH₄ (circle) and Imm2-YH₇ (star). Different colors represent different samples. The cited experimental data for YH₄ are represented by open circles.^[21] Dark cyan symbols depict the calculated data from Troyan et al.^[18]

To further determine the structure of the high-temperature superconducting phase, we performed in situ high-pressure ADXRD measurements of sample 1, which revealed that the products were dominated by I4/mmm-YH₃ and I4/mmm-YH₄ as shown in Fig. 1(c) and the refined structural information is listed in Table S1. The tetragonal YH₃, which possessed a new high-pressure phase in addition to the conventional fcc phase, was synthesized for the first time after prediction.^[17] Moreover, no superconductivity was predicted in I4/mmm-YH₃ up to 200 GPa. Consequently, the observed-superconducting transition in sample 1 should be attributed to YH₄.

Due to the small sample size, measuring the Meissner effect in ultra-high-pressure experiments remains a significant challenge to this day. An applied external magnetic field can break the Cooper pairs, reducing the value of T_c; thus, the suppression of superconducting transitions by an applied magnetic field can be used to investigate the nature of the superconducting states. Figure 3(a) shows the measured resistance of sample 2 under different magnetic fields at 170 GPa. The T_c decreased from 77 K to 53 K as the magnetic field increased to 8.5 T, indicating the superconducting nature of the transition. The extrapolated upper critical field μ₀H_c2(T) and coherence length toward T = 0 K are 14.9 T and 47 Å, as well as 18.7 T and 42 Å, respectively, as shown in Fig. 3(b), and were fitted by the Ginzburg–Landau (GL)^[30] and Werthamer–Helfand–Hohenberg (WHH)^[31] models. Furthermore, besides the superconductivity of YH₄, we observed another low-temperature superconductivity of 17 K [inset in Fig. 3(a)] in this experiment, which can be attributed to the element yttrium based on the agreement with the T_c of the unheated sample (Fig. S3). Similar results for YH₄ were independently reported by another group.^[21]

View original graphic| Download| PPT slide

Fig. 3.   (a) The temperature dependence of the resistance for I4/mmm-YH₄ under external magnetic fields of μ₀H = 0 T, 1 T, 3 T, 5 T, 7 T, and 8.5 T at 170 GPa in sample 2 (S2). Inset: the temperature-resistance curve without external magnetic fields. (b) Upper critical field versus temperature, μ₀H(0) was fitted with the GL and WHH models.

Furthermore, after laser heating sample 3 to approximately 1750 K at 162 GPa, we observed step-down behavior in electrical resistance measurements at 81 K, 29 K, and 18 K (Fig. 2(a)). As aforementioned, the first and third resistance drops, result from superconducting transitions of YH₄ and element Y, respectively. Based on previous theoretical work,^[18] we hypothesized that the second resistance drop at 29 K may originate from the superconducting transition of Imm2-YH₇, which was also reproduced in sample 7 [Fig. S2(c)]. Figure 2(b) summarizes the pressure dependency of T_c for YH₇ and YH₄. Similar to the variation trend of YH₄, the T_c of YH₇ was relatively stable in the pressure range of 142 GPa–170 GPa. Although both I4/mmm-YH₄ and Imm2-YH₇ have a molecular “H₂” unit [Fig. 1(d) and Fig. 2(a)], the T_c of YH₄ with a high-symmetry structure is higher than that of YH₇ due to stronger electron-phonon coupling.^[18]

In the following work, we tuned the heating temperature and pressure, to synthesize the high-temperature superconducting clathrate YH₆, YH₉, or even YH₁₀. When we increased the heating temperature to 2200 K at 165 GPa for sample 4, a T_c of 218 K was observed, as shown in Fig. 4(a). Subsequently, sample 5 was compressed to a superhigh pressure of 300 GPa [Fig. S1(a)] and heated to about 2000 K, and the electrical resistance measurement curve revealed superconductivity at 230 K (Fig. 4(a)). As shown in Fig. 4(b) the T_cs of samples 4 and samples 5 perfectly match the reported experimental results for clathrate structured YH₆ and YH₉,^[18,19] respectively. The high T_c of Im3¯m-YH₆ and P6₃/mmc-YH₉ was attributed to their hydrogen cage structure, and particularly the significant contribution of the H-derived electronic density of states at the Fermi level.^[8,10] Unfortunately, we found no evidence of clathrate YH₁₀, which may be synthesized at higher pressures.

View original graphic| Download| PPT slide

Fig. 4.   (a) Temperature dependence of resistance in sample 4 (S4) at 165 GPa and sample 5 (S5) at 300 GPa. The large residual resistance in S4 and S5 is mainly from the coexistence of multiple phases. Furthermore, the pseudo-four-electrode method was used in the electrical measurement for S5, thus introducing additional resistance from the electrodes. Inset: crystal structures of Im\overline{3}m-YH₆ and P6₃/mmc-YH₉. Big and small balls represent Y and H atoms, respectively. (b) Pressure dependence of T_c for Im\overline{3}m-YH₆ (star) and P6₃/mmc-YH₉ (hexagon). The symbols of dark cyan, orange, and red correspond to the data of Kong et al.,^[19] Troyan et al.,^[18] and this work, respectively.

A series of superhydrides with high T_c have been synthesized under high pressures; however, the absence of resistive transition broadening with increasing magnetic field in some works^[12,16,32] has led to a debate about their superconductivity.^[33] Using YH₄ as an example, we observed a clear broadening of the resistive transition under applied magnetic fields (Fig. S4), which follows a similar trend to that of typical standard superconductors such as MgB₂^[34] and NbN,^[35] further demonstrating the veracity of our results. As a member of superhydride, the results of electrical transport measurements under external magnetic fields in YH₄ will help clarify the debate on the superconductivity in superhydrides.

Conclusion and perspectives

In summary, we have successfully synthesized YH₄, YH₆, YH₇, and YH₉, which exhibited T_cs of 82 K at 167 GPa, 218 K at 165 GPa, 29 K at 162 GPa, and 230 K at 300 GPa, respectively. Furthermore, a tetragonal phase as a new high-pressure structure of conventional YH₃ was synthesized for the first time at 167 GPa. These findings confirm the original theoretical prediction and provide a foundation for future research into HTS on the doped Y-based polynary superhydrides.

Acknowledgments

XRD measurements were performed at BL15U1 station in Shanghai Synchrotron Radiation Facility (SSRF) and 4W2 station in Beijing Synchrotron Radiation Facility (BSRF). The measurements of superconducting transition under external magnetic fields were supported by the Synergic Extreme Condition User Facility (SECUF) and China’s Steady High Magnetic Field Facility (SHMFF).

Project supported by the National Key Research and Development Program of China (Grant Nos. 2021YFA1400203 and 2018YFA0305900), the National Natural Science Foundation of China (Grant Nos. 52090024, 11874175, 12074139, 12074138, 11874176, and 12034009), the Strategic Priority Research Program of Chinese Academy of Sciences (Grant No. XDB33000000), and Program for JLU Science and Technology Innovative Research Team (JLUSTIRT).

Reference

View Option

• By original order
• By published year
• By cited within times
• By Impact factor