Structural and electrical transport properties of Cu-doped Fe_1–xCu_xSe single crystals^*

The structurally simplest FeSe exhibits distinctive normal-state properties among the iron-based superconductors and is therefore important for investigating the underlying physics of the superconductivity.^[1] Its superconducting transition temperature T_C = 8 K at ambient pressure,^[2] which can be raised to 37 K by high pressure of 8 GPa,^[3] to ∼ 40 K by alkali-metal^[4–7] or small molecule intercalation,^[8,9] and even above 65 K in one-unit-cell film on SrTiO₃.^[10] FeSe undergoes a nematic transition at T_s = 90 K which is rapidly suppressed by pressure and can be tuned continuously by isoelectronic S or Te substitution at Se site of FeSe.^[11–13] Furthermore, with increasing pressure, the low-T_C superconducting phase transforms into the high-T_C phase, where the normal-state Hall resistivity changes sign from negative to positive, demonstrating a hole dominance in contrast to other FeSe-derived high-T_C systems.^[14] Recently, the observation of non-Fermi liquid property in FeSe_{1 – x} S_x and the topological nature of FeSe_{1 – x} Te_x have triggered renewed interest.^[15,16] Besides isoelectronic S and Te substitution at Se site, transition elements substitution at Fe site of FeSe would be also of interest for their comparable ionic sizes to Fe and the potential to tune the carrier type and concentration or investigate magnetic or non-magnetic impurity effect on superconductivity.^[17–24]

Earlier studies of Cu-substituted FeSe powder samples show that Cu substitution at Fe site in FeSe suppresses superconductivity and provokes a metal–insulator transition.^[21,22] Under a relatively small pressure of 1.5 GPa, superconductivity in Cu_0.04Fe_0.97Se is restored below 6.6 K; and T_C is increased to its maximum of 31.3 K upon an applied pressure at 7.8 GPa.^[23] Detailed Mössbauer spectroscopy studies on the ⁵⁷Fe-enriched Cu_0.04Fe_0.97Se powder sample reveal that part of the iron sites are magnetically ordered at low temperature, and the static magnetic moments destroy the superconducting pairing. Raising pressure leads to a collapse of the static magnetism and restoration of superconductivity with the maximal T_C value at 8 GPa.^[24] In contrast to the extensive studies on the intrinsic properties of isoelectronically substituted FeSe_{1 – x} S_x and FeSe_{1 – x} Te_x due to the availability of high-quality single crystals,^[1] few study of transition elements substitution effect is reported in FeSe single crystals. In the present paper, we report the structural and electrical transport properties of Fe_1–x Cu_x Se (x = 0–0.10) single crystals. Cu substitution suppresses both the nematicity and superconductivity of the FeSe single crystal, provokes a metal–insulator transition, changes an electron dominance at low temperature of un-doped FeSe to a hole dominance of Cu-doped Fe_1–x Cu_x Se at x = 0.02 and 0.1, and reduces the sign-change temperature (T_R ) of the Hall coefficient (R_H).

A series of Fe_1–x Cu_x Se (x = 0, 0.02, 0.05, 0.10) single crystals were grown by a chemical vapor transport method, using high-purity Fe, Cu, Se powders as raw materials and mixture of AlCl₃ and KCl as transport agent.^[25] The temperature of the hot and cold positions was kept at 400 °C and 350 °C, respectively. After a duration of 30 days, single crystals with a size of 3–5 mm were grown around the cold part of the quartz tube. The experimental details are given in Table 1.

The actual chemical composition of Fe_1–x Cu_x Se single crystals was determined by inductively coupled plasma atomic emission spectroscopy. The x-ray diffraction (XRD) measurements were performed on the x-ray diffractometer (MXP18A-HF) used copper K_α radiation. Electrical transport measurements up to 9 T were performed on a Quantum Design PPMS-9 system.

The results of powder XRD at room temperature demonstrate that the Fe_1–x Cu_x Se single crystals are of single phase and all the diffraction peaks can be well indexed with a previously reported tetragonal structure,^[21,22] as shown in Fig. 1(a). Figure 1(b) shows the single crystal XRD patterns for all the Fe_1–x Cu_x Se (x = 0, 0.02, 0.05, 0.10) single crystals. Only (0 0 l) reflections are observed, indicating that the single crystals are in perfect (0 0 1) orientation. Figure 1(c) shows the lattice parameters a and c as functions of the Cu doping level. With increasing Cu content, the lattice parameter a increases monotonically, while c decreases monotonically, consistent with previous reports.^[21,22] As displayed in Fig. 1(d), the unit cell volume V = a² c increases linearly with increasing Cu content.

Zoom In Zoom Out Reset image size

Figure 2 shows the temperature dependence of the normalized electrical resistivity ρ (T)/ρ (300 K) and specific heat measured for Fe_1–x Cu_x Se (x = 0, 0.02, 0.05, 0.10) single crystals. Pure FeSe is superconducting at T_C = 8.3 K and the kink at T_s = 90 K in the resistivity curve represents the nematic transition, which is more clearly visible from a dip in the temperature derivative of the resistivity as shown in the inset of Fig. 2(a) and a jump in the specific heat as displayed in Fig. 2(b). A small Cu doping of 2% can completely suppress the superconductivity and nematicity, as no such kink and jump can be found in the resistivity and specific heat measurements, respectively [Figs. 2(a) and 2(b)]. While the resistivity of FeSe in the normal state demonstrates a metallic behavior, an upturn with decreasing temperature appears around 42 K in the resistivity of Fe_0.98Cu_0.02Se, indicating a transition behavior from metallic to semiconducting with Cu doping. At Cu doping level x = 0.05 and 0.10, the resistivity increases quickly with cooling, showing an insulating behavior in the whole temperature range, which is in agreement with previous results of powder sample. The metal–insulator transition in Fe_1–x Cu_x Se can be probably attributed to Anderson localization arising from disorder.^[26,27]

Zoom In Zoom Out Reset image size

To further investigate the effects of Cu doping on the electrical transport properties, we measured the Hall effect of Fe_1–x Cu_x Se (x = 0, 0.02, 0.05, 0.10) single crystals. Figure 3 shows the Hall resistivity ρ_xy at various temperatures with different Cu doping level. As seen in Fig. 3(a), ρ_xy of stoichiometric FeSe shows a linear field dependence for T > 50 K, in accordance with the compensated semimetal character, and the slope changes sign twice from positive to negative and then back to positive upon cooling.^[28] A non-linearity develops for ρ_xy and the initial slope eventually becomes negative for T ≤ 50 K, but tends to change sign again under higher magnetic field. Unlike pure FeSe, the common feature of ρ_xy curves for Fe_1–x Cu_x Se with x = 0.02, 0.05, 0.1 is that no non-linearity develops in the whole temperature and magnetic field range. For Fe_0.98Cu_0.02Se as shown in Fig. 3(b), the slope changes sign twice from positive to negative and then back to positive and remains positive upon cooling to 10 K. For Fe_0.95Cu_0.05Se as displayed in Fig. 3(c), the slope changes sign once from positive to negative at T = 140 K. For Fe_0.90Cu_0.10Se, the slope always remains positive in the whole temperature range without any sign change as presented in Fig. 3(d).

Zoom In Zoom Out Reset image size

We plot in Fig. 4(a) the temperature dependence of the Hall coefficient, defined as the field derivative of ρ_xy , R_H ≡ d ρ_xy /d H, at the zero-field limit. The sign of the Hall coefficient R_H is an indicator of the dominant carrier type. The temperature dependence of R_H is shown in Fig. 4(a). With cooling, the Hall coefficient of FeSe changes sign at three temperatures T_R1, T_R2, and T_R3, from positive to negative at T_R1 = 205 K, then from negative to positive at T_R2 = 120 K with a moderate enhancement of R_H before T_s, and finally reversed again at T_R3 = 59 K, exhibiting a tendency to strongly negative values below T_s. The evolution with Cu doping of the sign-change temperature T_R of the Hall coefficient is summarized in Fig. 4(b). All the three temperatures T_R1, T_R2, and T_R3 are gradually reduced with Cu doping, with T_R1 tending to vanish towards Cu doping x = 0.1, which indicates a hole dominant carrier type.

Zoom In Zoom Out Reset image size

As we know, the valence state of the Fe atom in FeSe and other iron-based superconductors is assigned as Fe²⁺, so the Fe atoms are formally in the 3d⁶ electronic configuration. Also, electronic structure calculations for SrCu₂As₂ and BaCu₂As₂ predicted that the Cu atoms have a formal valence state of Cu^{1 +} and a nonmagnetic and chemically inert 3d¹⁰ electronic configuration.^[29] Anand et al. confirmed this result by measuring the electronic and magnetic properties of SrCu₂As₂ and BaCu₂As₂, suggesting that Cu substitution for Fe in (Ca,Sr,Ba)(Fe_1–x Cu_x )₂As₂ should result in hole doping.^[30] We infer that Cu substitution in FeSe plays a role of hole doping which leads to the suppression of sign change temperatures T_R1, T_R2, and T_R3 of the Hall coefficient and hole dominant carrier type in Fe_1–x Cu_x Se.

We have successfully grown a series of Fe_1–x Cu_x Se single crystals with x = 0, 0.02, 0.05, and 0.1. With increasing Cu doping content, the lattice parameter a increases monotonically, while the lattice parameter c decreases monotonically. Both the resistivity and specific heat measurements indicate that Cu doping suppresses nematicity and superconductivity and provokes a metal–insulator transition in Fe_1–x Cu_x Se. Also, Cu substitution changes an electron dominance at low temperature of un-doped FeSe to a hole dominance of Cu-doped Fe_1–x Cu_x Se at x = 0.02 and 0.1, and reduces the sign-change temperature (T_R ) of the Hall coefficient (R_H).