As illustrated in Fig. 1(a), the core part of the CAC consists of six anvils converging onto a central cubic gasket from three orthogonal directions; the synchronized movement of these six anvils is driven by a pair of guide blocks that convert the applied uniaxial loading force to a three-axis compression.^[10] In the center of the cubic gasket (see Fig. 1(b)), the sample of interest is hung by electrical leads in a Teflon capsule filled with a liquid pressure-transmitting medium (PTM). In comparison with the opposed-anvil-type HP devices such as DAC, such a three-axis compression geometry together with the adoption of liquid PTM can ensure an excellent pressure homogeneity inside the Teflon capsule. This is very important to obtain intrinsic pressure responses. Because all PTMs will be solidified under HPLT conditions, we should minimize its influence by warming up the CAC to room temperature before increasing to a higher pressure.

Fig. 1.

	Figure Option ViewDownloadNew Window
	Fig. 1. (color online) (a) Illustration of the operation principle of the CAC in which six anvils converge onto the cubic gasket from three orthogonal directions. (b) Sample configuration in the cubic gasket with the typical dimensions of a gasket and Teflon capsule used for the 4 mm anvils.

Currently, most CACs employ 4 mm × 4 mm square-top anvils and a 6 mm-edge-length cubic gasket together with the Teflon capsule with an outer diameter of 2 mm (see Fig. 1(b)). These dimensions of the anvil/gasket/Teflon capsule have been proven to be reliable in terms of the pressure efficiency and the surviving rate of electrical leads over the past 30 years. Under these conditions mentioned above, the maximum pressure can reach approximately 8 GPa using the tungsten-carbide (WC) anvils.^[10]

The CAC apparatus for in situ HPLT measurements was originally developed by N. Mori in the 1990s at the Institute for Solid State Physics (ISSP), University of Tokyo. A detailed description of the CAC apparatus can be found elsewhere.^[10] As shown in Fig. 2(a), this apparatus is featured by a specially designed top-loading HP cryostat fitted in a 250-ton hydraulic press, which is used to maintain a constant loading force over the CAC device placed inside the cryostat. The low-temperature environment is realized by filling the cryostat with liquid nitrogen and helium, and the lowest temperature of ∼ 1.5 K can be achieved using a high-speed vacuum pump. The development of such a CAC apparatus with excellent hydrostatic pressure condition enabled Mori et al. to obtain remarkably distinct results from others; for instance, the complete suppression of the Verwey transition around 8 GPa for Fe₃O₄,^[11] and the observation of SC at 14.2 K around 7 GPa in β′-(BEDT-FFT)₂ICl₂.^[12] Such CAC apparatus has been successfully operated over 30 years and remains in use in the ISSP and other institutions in Japan.

Fig. 2.

	Figure Option ViewDownloadNew Window
	Fig. 2. (color online) (a) Conventional CAC apparatus with a liquid 4He cryostat combined with 250-ton hydraulic press used in the ISSP, University of Tokyo. (b) Palm CAC with a top-loading 3He cryostat used in the ISSP, University of Tokyo, comparison of the (c) guild block and (d) WC anvil for the palm CAC and the conventional CAC. (e) Schematic drawing of the integrated-fin gasket.

However, the massive cell body of the conventional CAC apparatus prevents an efficient cooling to well below 2 K and also renders it difficult to integrate with a strong magnetic field. These issues were overcome by the recent development by Y. Uwatoko using a clamp-type, “palm”-sized CAC.^[13] The operation principle is almost identical to the conventional type, except that the applied loading force is locked by tightening the nut of the beryllium-copper clamp cell, which can sustain a maximum loading force of 100 ton for the current design. As illustrated in Figs. 2(c) and 2(d), the weight of the guide block and WC anvil are reduced remarkably to be, respectively, 1/16 and 1/7.5 of the conventional type, albeit the pressure capacity remains nearly unchanged, i.e., ∼ 9 GPa for the 4 mm WC anvils. Such a miniature palm CAC has been subsequently integrated with a ³He refrigerator equipped with a five-Tesla magnet (see Fig. 2(b)), thus enabling the combination of multiple extreme conditions including hydrostatic pressures up to ∼ 9 GPa, the lowest temperatures down to 0.5 K, and the magnetic field up to 5 T.^[14] These developments were crucial for the discovery of pressure-induced heavy-fermion SC in the nonmagnetic quadrupolar system PrTi₂Al₂₀ with a maximum T_c = 1.1 K at around 9 GPa.^[15]

To further increase the pressure capacity of the palm CAC, we designed an integrated-fin gasket (see Fig. 2(e)), which shows great improvements in terms of the pressure efficiency and the success rate for the HPLT experiments. By using such a gasket made from MgO and the WC anvils of 2.5-mm square top, we successfully generated pressures over 15 GPa at both room and cryogenic temperatures.^[16] Very recently, the palm CAC has been successfully integrated into a cryogen-free dilution refrigerator with a superconducting magnet, achieving the following multiple extreme conditions: P_max = 15 GPa, T_min = 10 mK, and B_max = 8 T, in the ISSP, University of Tokyo. These improvements are indispensable for many unsolved issues in heavy fermion physics.

Given the excellent pressure homogeneity and large hydrostatic pressure capacity, a CAC apparatus has been approved in the SECUF project. We shall follow the most recent developments of the CAC in the ISSP, University of Tokyo and aim to achieve the state-of-the-art technical targets. We will employ the palm CAC with the integrated-fin gasket and 2.5-mm WC anvils to reach the maximum hydrostatic pressure of 15 GPa. We shall integrate the palm CAC with two different cryostats for different research purposes: (i) a liquid ⁴He sample-in-vapor cryostat equipped with a 9 T room-temperature (RT) bore cryogen-free superconducting magnet for routine HPLT measurements from RT down to 1.5 K under 9 T; (ii) a cryogen-free ³He/⁴He dilution refrigerator equipped with 12 T cold-bore superconducting magnet for ultralow temperature studies. We will perform the following physical property measurements: electrical resistivity, magnetoresistance and Hall resistivity, dielectric constant, ac magnetic susceptibility, and ac specific heat.

One of the common features of unconventional superconducting systems such as the heavy-fermion, high-T_c cuprate and iron-pnictide superconductors is that SC emerges on the border of long-range magnetically ordered state. The close proximity of SC to a magnetic instability suggests that the critical spin fluctuations would be crucial for mediating the Cooper pairs. In addition to doping charge carriers, the application of HP is an effective approach for searching novel classes of unconventional superconductors near the magnetic quantum critical point (QCP). After we discovered SC with T_c ≈ 2 K in CrAs by suppressing its helimagnetic order via the application of hydrostatic pressure of ∼ 1 GPa,^[17] we explored the pressure-induced SC in MnP by following the similar strategy.^[18]

At the ambient pressure, MnP is an itinerant-electron helimagnet with an orthorhombic B31-type structure similar as CrAs. In the absence of magnetic field, MnP undergoes two successive magnetic transitions upon cooling:^[19] a transition from the paramagnetic (PM) to ferromagnetic (FM) state at T_c = 291 K, and subsequently the second transition to a double helical state at T_s ≈ 50 K.^[19] Earlier HP studies on MnP have revealed that both T_c and T_s decrease with pressure, but the highest hydrostatic pressure of ∼ 3 GPa in the previous studies was far from the magnetic QCP.^[20]

Figure 3 displays the T–P phase diagram of MnP constructed based on measurements of resistivity and ac magnetic susceptibility up to 11 GPa by using the palm CAC.^[18] As shown, both T_s and T_c decrease with pressure and T_s is suppressed completely around 1.4 GPa. Above 3 GPa, the FM transition at T_c changes to a new-type antiferromagnetic (AFM) order marked as T_m, which was later confirmed to be also a helical magnetic order with the propagation vector along the b-axis.^[21] Surprisingly, SC emerges below T_sc ≈ 1 K near the critical pressure of P_c ≈ 8 GPa where the long-range AFM order at T_m just vanishes. Because the majority of the density of states near the Fermi level for MnP is attributed to the Mn-3d states, this discovery renders MnP the first Mn-based superconductor, and the close proximity of SC to a magnetic instability suggests an unconventional pairing mechanism. Moreover, the detailed analysis of the normal-state transport properties evidenced non-Fermi liquid behavior and the dramatic enhancement in the quasiparticles effective mass near P_c associated with the magnetic quantum fluctuations.^[18] The discovery of the first Mn-based superconductor dispels the general wisdom about Mn’s antagonism to SC and has thus opened a new avenue for investigating more Mn-based superconductors.^[22]

Fig. 3.

	Figure Option ViewDownloadNew Window
	Fig. 3. (color online) T–P phase diagram of MnP. PM: paramagnetic, FM: ferromagnetic, AFM: antiferromagnetic, SC: superconductivity; Tc: PM–FM transition temperature, Ts: FM–Screw transition temperature, T*: FM–AFM transition temperature, Tm: PM–AFM transition temperature, Tsc: superconducting transition temperature. This figure is reproduced from Ref. [18] .

Among the iron-based superconductors, the structurally simplest β-FeSe and its derived materials have attracted tremendous attention recently owing to its peculiar electronic properties and the great tunability of T_c.^[23,24] At the ambient pressure, FeSe undergoes a tetragonal-to-orthorhombic structural transition at T_s = 90 K, below which it develops a significant electronic anisotropy or nematicity. Unlike the iron-pnictides superconductors, however, no magnetic order appears within the nematic phase of FeSe, and the SC with T_c ≈ 9 K occurs without chemical doping or HP. Meanwhile, it was found that HP can induce a static magnetic order at T_m around 1 GPa^[25] and promote a high-T_c SC with a T_c of up to 37 K around 7 GPa.^[26] These facts render FeSe an ideal platform to study the interplay of SC with magnetism and nematicity.

Despite the extensive HP investigations on FeSe, unfortunately, all previous studies failed to construct a completed T–P phase diagram for FeSe owing to technical limitations; i.e., the PCC with a maximum pressure of ∼ 3 GPa is insufficient to reach the highest T_c around 7 GPa,^[27] whereas the presence of significant pressure inhomogeneity in the DAC with a solid pressure medium hinders the evolution tracking of the nematic and magnetic orders.^[26] In contrast, the excellent hydrostaticity and large pressure capacity render CAC an ideal choice for this study.

By performing HP resistivity and ac magnetic susceptibility up to 15 GPa with the palm CAC on high-quality FeSe single crystals, we constructed its most comprehensive T–P phase diagram^[28] (see Fig. 4), which maps out the explicit evolutions with the pressure of T_s, T_c, and T_m. As shown, we uncovered a previously unknown dome-shaped T_m(P), having two end points situated on the boundaries separating the three plateaus of T_c(P). Our results provide compelling evidence linking intimately the sudden enhancement of T_c to 38 K to the suppression of the long-rang magnetic order. The fact that the optimal T_c is achieved when the long-range antiferromagnetic order just vanishes is reminiscent of the situations shown frequently in the FeAs-based superconductors.

Fig. 4.

	Figure Option ViewDownloadNew Window
	Fig. 4. (color online) T–P phase diagram of FeSe. Ts for the tetragonal to orthorhombic structural transition, Tm for the antiferromagnetic or SDW transition, Tc for the superconducting transition. This figure is reproduced from Ref. [28].

Previous studies on the FeSe-derived high-T_c (> 30 K) superconductors, including A_xFe_2−ySe₂ (A = K, Cs, Rb, Tl), (Li,Fe)OHFeSe, and monolayer FeSe film, have shown a distinct Fermi surface (FS) topology consisting of only electron pockets near the Fermi level.^[23] In contrast, the FS of FeAs-based superconductors typically consists of hole and electron pockets near the Brillouin zone center and corners, respectively. Thus, the distinct FS topology between FeAs- and FeSe-based materials has challenged the current theories on a unified understanding of the mechanism of iron-based superconductors.

In contrast to the electron-doping approaches, the application of HP does not introduce extra electron carriers, yet can still enhance the T_c of bulk FeSe to ∼ 38 K near 6 GPa. To understand the mechanism of pressure-induced high-T_c SC in FeSe, it is important to obtain the information about the evolution of FS under HP—a regime in which angle-resolved photoemission spectroscopy experiments are impractical, and where quantum oscillation measurements are challenging. Alternatively, we measured the Hall effect under hydrostatic pressures up to 8.8 GPa with the CAC to gain further insights into the electronic structure evolution of FeSe at HP.^[29]

As shown in Fig. 5, our results demonstrate that the electrical transport properties of FeSe at HPs with K are dominated by the hole carriers, which is in contrast with the known FeSe-derived high-T_c superconductors that are typically heavily electron-doped. In addition, we observed an enhancement in the Hall coefficient R_H near the critical pressure where the optimal T_c is realized with a simultaneous suppression of the long-range magnetic order. This implies a strong reconstruction of the FS due to the magnetic order, consistent with the ordering pattern driven by interband scattering. More importantly, our results show a continuous path to high-T_c SC in chalcogenides without electron doping. Our results constitute a significant step forward in presenting a unified picture on the current understanding of iron-based high-T_c superconductors, specifically by demonstrating that high T_c in FeSe can be achieved with an electronic structure and other characteristics similar to the FeAs-based high-T_c superconductors.

Fig. 5.

	Figure Option ViewDownloadNew Window
	Fig. 5. (color online) T–P phase diagram of FeSe superimposed by a contour plot of Hall coefficient RH. This figure is reproduced from Ref. [29].

Although the bulk FeSe displays a relatively low T_c = 8.5 K within the nonmagnetic nematic phase below T_s = 90 K, a high-T_c SC with T_c above 30–40 K have been successfully achieved via intercalating some alkali–metal ions, ammonia, or organic molecules in between the adjacent FeSe layers, such as in A_xFe_2−ySe₂ (A = K, Rb, Cs, Tl),^[30] A_x(NH₃)_yFeSe,^[31] and (Li, Fe)OHFeSe.^[32] We refer these high-T_c superconductors derived directly from FeSe as the SC-I phase. The superconducting mechanism for these SC-I phases has been subjected to extensive investigations, and the observed common FS topology consisting of only electron pockets in the Brillouin zone corners suggests that the electron doping plays an essential role for achieving a high T_c, in agreement with the gate-voltage regulation experiments on the FeSe flakes.^[33]

Starting from the SC-I phase in A_xFe_2−ySe₂, it was reported that pressure can induce the sudden reemergence of a second superconducting phase (denoted as SC-II hereafter) with a higher T_c of up to 48.7 K above ∼ 10 GPa.^[34] A similar SC-II phase has also been observed in Cs_0.4(NH₃)_yFeSe under HP.^[35] Although the reemergence of the SC-II phase with a higher T_c is intriguing and a different pairing symmetry has been proposed theoretically,^[36] the intrinsic superconducting- and normal-state properties have been poorly characterized thus far owing to some sample and technical difficulties. For example, A_xFe_2−ySe₂ superconductors are prone to phase separation accompanied by the intergrowth of the antiferromagnetic insulating A₂Fe₄Se₅ phase. Moreover, HP techniques capable of both large pressure capacity and good hydrostaticity are required to obtain reliable superconducting- and normal-state properties. Therefore, these complexities have hampered the proper understanding of the intriguing SC-II phase of these FeSe-derived materials.

To approach this intriguing problem, we focus on the recently discovered (Li_1−xFe_x)OHFeSe,^[32] which is free from phase separation, relatively stable in air, and more importantly, can be obtained in high-quality single crystals via a specially designed hydrothermal ion-exchange method.^[37] Using the palm CAC, we recently performed detailed magnetotransport measurements on (Li_1−xFe_x)OHFeSe single crystals under hydrostatic pressures up to 12.5 GPa.^[38]

The primary results are summarized in Fig. 6. We found that the ambient-pressure SC-I phase is suppressed gradually by increasing pressure to P_c = 5 GPa, above which a new SC-II phase with a higher T_c over 50 K emerges gradually. More importantly, our high-precision resistivity data enabled us to uncover a sharp transition of the normal state from a Fermi liquid for SC-I to a non-Fermi liquid for the SC-II phase. In addition, the reemergence of a higher T_c SC-II phase is found to accompany with a concurrent enhancement in the electron carrier density. Such information was unavailable in the previous HP studies on the FeSe-derived high-T_c superconductors.^[33,34] The present work thus provides positive correlations between the high-T_c SC in SC-II with an FS reconstruction, which is not induced by a structural transition, as confirmed by our HP structural study.^[38]

Fig. 6.

	Figure Option ViewDownloadNew Window
	Fig. 6. (color online). T–P phase diagram of (Li,Fe)OHFeSe. SC-I and SC-II represents two distinct superconducting regions. The contour color plot of the normal-state resistivity exponent α illustrates a sharp transition of the normal state from the Fermi liquid for SC-I to the non-Fermi liquid for the SC-II phase. This figure is reproduced from Ref. [38].