MBE Growth and Element-distinctive Atomic-resolution Characterization of High Temperature Superconductors

1. Introduction

Superconductivity is a phenomenon in quantum materials that allows for lossless transportation of electrical current, and thus of energy and information. The quantum phenomenon, superconductivity, had been observed only below –140°C (~130 K) until very recent reports on hydrides, that is, sulfur hydride (H₃S) and lanthanum hydride (LaH₁₀). These materials show significantly higher superconducting transition temperatures (T_cs): –70°C (~200 K) for H₃S [1] and even close to room temperature for LaH₁₀ [2]. Synthesizing these hydrides, however, requires extremely high pressure, specifically, ~2 million times higher (~200 GPa) than the atmospheric pressure. To make matters more complicated, their crystal structures are altered when the pressure is reduced after synthesis. In other words, the superconducting phases can exist only under an extremely high pressure comparable to that at the core of the Earth. Accordingly, practical applications using these hydrides remain elusive despite the significance of their discoveries from an academic point of view.

In contrast, cuprates are superconductors that exhibit the highest T_c under ambient pressure. Among them, YBa₂Cu₃O_7-δ (yttrium barium copper oxide) as well as Bi₂Sr₂Ca₂Cu₃O_10+δ (bismuth strontium calcium copper oxide), both of which show higher T_c (~90 K and ~110 K, respectively) than the boiling point of nitrogen (77 K), have already yielded practical applications as superconducting cables and microwave filters. However, the mechanism of high-T_c superconductivity in cuprates remains unclarified despite immense efforts for over three decades, which has been hampering the strategic search of novel superconducting materials with a higher T_c.

Nonetheless, the fundamental structural ingredient of high-T_c cuprates is fortunately well known: copper peroxide (CuO₂) planes, as exemplified by Bi₂Sr₂Ca₂Cu₃O_10+δ (Fig. 1(a)). Two-dimensional CuO₂ planes are made up of copper (Cu) and oxygen (O) and are separated by bismuth (Bi), strontium (Sr), calcium (Ca), and O. The formal valences of O and Cu are 2– and 2+, respectively, so the CuO₂ plane is not charge-neutral and cannot exist independently. Instead, the minimal structural unit that is compassable is charge-neutral calcium copper oxide (CaCuO₂) (Fig. 1(b)), which is commonly included in cuprates whose T_cs are above 100 K; the crystal structure shown in Fig. 1(b) is called an infinite-layer structure. That is why we have been focusing our research efforts on infinite-layer cuprates. Because the synthesis of bulk specimens of CaCuO₂ requires high pressure levels (3–5 GPa), single-crystalline CaCuO₂ specimens can be prepared exclusively by using thin-film growth methods, in our case, molecular beam epitaxy (MBE). Unlike the above-mentioned hydrides, the infinite-layer structure is stable at ambient pressure once it is formed.

Fig. 1. Crystal structures of cuprate superconductors.

2. Fabrication of ultrahigh-quality infinite-layer cuprate superconductors

While MBE has been widely used for the growth of semiconductors, we have extended it to the growth of complex transition metal oxides. Our customized MBE method is well equipped not only to synthesize novel complex transition metal oxides, for example, Sr₃OsO₆, a new ferromagnetic insulator with a Curie temperature > 1000 K [3], but also to push the limits of common crystal growth methods [4], and this is illustrated in Fig. 2. The vacuum recipient is 70 cm in diameter and 150 cm in height. A total of 10 metal sources are mounted at the bottom of the vacuum recipient where the elements can be evaporated by electron guns operated at 10 kV. The evaporant flux of each element is monitored and controlled by electron impact emission spectroscopy (EIES). The details of EIES are explained in another article in this issue [5].

Fig. 2. Oxide MBE setup.

Molecular oxygen (O₂) is insufficient for the growth of complex transition metal oxides, and therefore, the oxide MBE system is equipped with stronger oxidizing agents. This is distinct from conventional MBE. In this work, radio-frequency radical oxygen (O) was used for the growth of infinite-layer cuprates. Oxidation/reduction involves the gain/loss of electrons. Radical oxygen has two unpaired electrons and therefore a strong tendency to deprive the neighboring metal atoms of electrons. The absence of kinetic barriers in atomic oxygen is what drives the growth of complex transition metal oxides in an ultrahigh vacuum (~10^–9 atm). In addition, the growth orientation is defined by the substrate; this phenomenon is called epitaxy (in Greek, order on top). We used a single-crystal substrate with a lattice constant closer to one for infinite-layer cuprates. This mechanism enables the formation of single-crystalline infinite-layer cuprates.

We used the oxide MBE system to synthesize superconducting infinite-layer Sr_0.9La_0.1CuO₂ thin films [6–9]. In Sr_0.9La_0.1CuO₂, strontium ions (Sr²⁺ ions) are partially substituted with La³⁺ ions (lanthanum ions) to induce superconductivity. The dependence of resistivity on temperature for the Sr_0.9La_0.1CuO₂ thin film is plotted in Fig. 3. Resistivity decreases as the temperature decreases and abruptly goes to zero at 41 K. The films presented here are superior to those in other reports on superconducting infinite-layer thin films [10], as evidenced by the following characteristics:

• sharp superconducting transition (transition width ΔT_c < 1 K)
• metal-like electronic response between 400 K and 42 K
• low resistivity value in normal state (corresponds to the intrinsic value for defect-free CuO₂ plane).

Fig. 3. Temperature dependence of resistivity for Sr_0.9La_0.1CuO₂ thin film.

It is therefore shown that samples grown with our method are most suitable to reveal the inherent physical properties leading to understanding the mechanism of high-T_c superconductivity.

3. Infinite-layer CaCuO₂—elemental-resolved and atomic-resolution characterization

As already mentioned, CaCuO₂ is an essential structural part of high-T_c cuprates. In contrast to strontium copper oxide (SrCuO₂), however, little is known about the physical properties of infinite-layer CaCuO₂ due to the difficulty of synthesis. We synthesized infinite-layer Ca_1-xNd_xCuO₂ thin films using MBE. In Ca_1-xNd_xCuO₂, calcium ions (Ca²⁺ ions) are partially substituted with Nd³⁺ ions (neodymium ions) by mimicking Sr_0.9La_0.1CuO₂. For a composition of x = 0.06, we found traces of superconductivity around 10 K [11]. It is important to note that a truly superconducting state has not been established. Therefore, it is of significant importance to determine the differences between superconducting Sr_1-xLa_xCuO₂ and non-superconducting Ca_1-xNd_xCuO₂ by using scanning transmission electron microscopy (STEM) combined with electron energy loss spectroscopy (EELS).

STEM is used to visualize atomic positions in matter. The configuration of our STEM apparatus is shown in Fig. 4(a) [12]. Accelerated electrons at 200 kV are focused on a sample. The scattered electron beam contains information on the atomic position. This is shown in Fig. 4(b) for Ca_0.96Nd_0.04CuO₂. Because the contrast of the original bright-field image is inverted, the atomic positions appear bright in Fig. 4(b). The observed atomic column arrangements correspond to the infinite-layer structure with two-dimensional CuO₂ planes. It is evident that Ca_0.96Nd_0.04CuO₂ thin films are single crystalline. We used EELS to identify the constituent atomic species. This is shown in Fig. 4(c). The atomic selectivity of EELS enables an atomic column identification that exactly matches the infinite-layer phase.

Fig. 4. Identification of superconducting state of infinite-layer CaCuO₂.

By measuring the distance between Cu and O (d_Cu-O) from Fig. 4(c), we determined the in-plane lattice constant of Ca_0.96Nd_0.04CuO₂ to be 0.386 nm. This is 0.001 nm longer than for CaCuO₂ [13]. Similar expansion is found for the Sr_0.9La_0.1CuO₂ system, where a substitution of 10% of La results in an expansion of 0.002 nm of the in-plane lattice constant [14]. It is intuitively suggested that a critical in-plane lattice constant (d_Cu-O x 2) for the induction of superconductivity is to be expected. As the ionic size of Ca²⁺ ions is smaller than that of Sr²⁺ ions, CaCuO₂ has shorter d_Cu-O than for SrCuO₂. During the course of our study, we found that a higher concentration of Nd causes instability of the infinite-layer phase. Consequently, it is not possible to prepare the infinite-layer phase of Ca_1-xNd_xCuO₂ with a sufficient amount of Nd to induce a superconducting transition. This result suggests that lattice constant (the Cu-O bond length in the CuO₂ plane) engineering is important for the induction of superconductivity in cuprates.

4. Future outlook

Infinite-layer cuprates are very important to understand the physics of high-T_c superconductivity. We plan to apply our thin film growth method to further understand the mechanism of high-T_c superconductivity commonly emerging in materials containing CaCuO₂. This method is currently being extended to the synthesis of superlattices containing CaCuO₂ [15].

References

[1]	A. P. Drozdov, M. I. Eremets, I. A. Troyan, V. Ksenofontov, and S. I. Shylin, “Conventional Superconductivity at 203 Kelvin at High Pressures in the Sulfur Hydride System,” Nature, Vol. 525, No. 7567, pp. 73–76, 2015.
[2]	M. Somayazulu, M. Ahart, A. K. Mishra, Z. M. Geballe, M. Baldini, Y. Meng, V. V. Struzhkin, and R. J. Hemley, “Evidence for Superconductivity above 260 K in Lanthanum Superhydride at Megabar Pressures,” Phys. Rev. Lett., Vol. 122, No. 2, 027001, 2019.
[3]	Y. K. Wakabayashi, Y. Krockenberger, N. Tsujimoto, T. Boykin, S. Tsuneyuki, Y. Taniyasu, and H. Yamamoto, “Ferromagnetism above 1000 K in a Highly Cation-ordered Double-perovskite Insulator Sr3OsO6,” Nat. Commun., Vol. 10, No. 1, 535, 2019.
[4]	H. Yamamoto, Y. Krockenberger, and M. Naito, “Multi-source MBE with High-precision Rate Control System as a Synthesis Method sui generis for Multi-cation Metal Oxides,” Journal of Crystal Growth, Vol. 378, pp. 184–188, 2013.
[5]	H. Yamamoto and H. Gotoh, “Overview of Novel Materials Creation Research at NTT,” NTT Technical Review, Vol. 17, No. 10, pp. 1–6, 2019. https://www.ntt-review.jp/archive/ntttechnical.php?contents=ntr201910fa1.html
[6]	Y. Krockenberger, K. Sakuma, and H. Yamamoto, “Molecular Beam Epitaxy and Transport Properties of Infinite-layer Sr0.90La0.10CuO2 Thin Films,” Appl. Phys. Express, Vol. 5, No. 4, 043101, 2012.
[7]	S. Karimoto and M. Naito, “Electron-doped Infinite-layer Thin Films with Tc over 40 K Grown on DyScO3 Substrates,” Appl. Phys. Lett., Vol. 84, No. 12, pp. 2136–2138, 2004.
[8]	S. Karimoto, K. Ueda, M. Naito, and T. Imai, “Superconducting Thin Films of Electron-doped Infinite-layer Sr1−xLaxCuO2 Grown by Molecular Beam Epitaxy,” Physica C Supercond., Vol. 378–381, Part 1, pp. 127–130, 2002.
[9]	M. Naito, S. Karimoto, and A. Tsukada, “Epitaxy-stabilized n-type Superconducting Cuprates,” Supercond. Sci. Technol., Vol. 15, No. 12, 1663, 2002.
[10]	L. Maritato, A. Galdi, P. Orgiani, J. W. Harter, J. Schubert, K. M. Shen, and D. G. Schlom, “Layer-by-layer Shuttered Molecular-beam Epitaxial Growth of Superconducting Sr1–xLaxCuO2 Thin Films,” J. Appl. Phys., Vol. 113, No. 5, 053911, 2013.
[11]	A. Ikeda, Y. Krockenberger, and H. Yamamoto, “Molecular Beam Epitaxy of Electron-doped Infinite-layer Ca1-xRxCuO2 Thin Films,” Phys. Rev. Materials, Vol. 3, No. 6, 064803, 2019.
[12]	S. J. Pennycook, C. Li, M. Li, C. Tang, E. Okunishi, M. Varela, Y.-M. Kim, and J. H. Jang, “Material Structure, Properties, and Dynamics through Scanning Transmission Electron Microscopy,” J. Anal. Sci. Technol., Vol. 9, 11, 2018.
[13]	Y. Krockenberger, A. Ikeda, K. Kumakura, and H. Yamamoto, “Infinite-layer Phase Formation in the Ca1–xSrxCuO2 System by Reactive Molecular Beam Epitaxy,” J. Appl. Phys., Vol. 124, No. 7, 073905, 2018.
[14]	N. Ikeda, Z. Hiroi, M. Azuma, M. Takano, Y. Bando, and Y. Takeda, “Synthesis and Superconducting Properties of the Infinite-layer Compounds Sr1−xLnxCuO2 (Ln = La, Nd, Sm, Gd),” Physica C Supercond., Vol. 210, No. 3–4, pp. 367–372, 1993.
[15]	D. Di Castro, M. Salvato, A. Tebano, D. Innocenti, C. Aruta, W. Prellier, O. I. Lebedev, I. Ottaviani, N. B. Brookes, M. Minola, M. Moretti Sala, C. Mazzoli, P. G. Medaglia, G. Ghiringhelli, L. Braicovich, M. Cirillo, and G. Balestrino, “Occurrence of a High-temperature Superconducting Phase in (CaCuO2)n/(SrTiO3)m Superlattices,” Phys. Rev. B, Vol. 86, No. 13, 134524, 2012.