We fabricated 100 nm thick films of two-dimensional triangular antiferromagnet Ag2CrO2 using the mechanical exfoliation technique, and performed the transport measurements down to 5 K. As in the case of polycrystalline samples, a large reduction of the resistivity due to the antiferromagnetic ordering was clearly observed at around 25 K. Surprisingly, the resistivity for the thin films is one order magnitude smaller than that for the polycrystalline samples, indicating that the crystalline nature is much better in the former than in the latter. The present result sheds new light on the use of atomic-layer antiferromagnetic materials for device applications.
Since the discovery of graphene in 2004,1 there have been many reports not only on graphene but also on a variety of two-dimensional (2D) materials such as MoS2,2–4 WSe2,5,6 NbSe2,7–11 hexagonal boron nitride,12,13 and so on. These are intrinsically semiconductors, metals (even superconductors at low temperatures), and insulators, respectively. Importantly, for such 2D materials, the material properties can be controlled by applying electric fields to them. In the case of graphene, the carrier polarity (electron or hole) as well as the carrier density can be changed by gating because of its unique band structure.14 MoS2 is originally semiconducting, but the electron density can be enhanced by tuning the electric field, resulting in the superconductivity at low temperatures.4
2D materials can also be used for spintronic devices. Graphene has a weak spin-orbit interaction (SOI). By forming a bridge with graphene between two ferromagnetic metal wires, a lateral spin valve device can be fabricated, where spin current, flow of spin angular momentum, is generated in the graphene channel.15 Owing to the weak SOI in graphene, the spin current has been transferred as far as several micrometers.16
Another important issue in spintronics is to realize efficient generation and detection of the spin current. The spin Hall effect enables one to create and detect the spin current in an electrical manner. For this purpose, strong SOI materials are required. MacNeill et al.17 have demonstrated control of spin-orbit torques in WTe2/ferromagnetic metal bilayers via the spin Hall effect. Even ferromagnetic metals can be replaced by layered ferromagnetic materials.18,19 Synthesis of purely 2D magnetic materials intuitively seems to be impossible because of the Mermin-Wagner theorem.20 However, it has been reported very recently that 2D ferromagnetic layers can be obtained by the exfoliation technique.21,22
In spintronics, antiferromagnetic materials have also attracted much attention because they have tiny leak magnetic fields which lead to robust data storage.23,24 In addition, they enable high speed data processing due to the higher magnetic resonance frequency compared to normal ferromagnetic materials. Thus, exploring 2D antiferromagnets is highly desirable to achieve future 2D spintronic devices.
On the other hand, 2D antiferromagnets have another interesting property, that is magnetic frustrations. When the magnetic sites are arranged in a triangular shape, it has been well-known that the magnetic moments have a large frustration because of the geometrical effect. This topic has been one of the central issues in modern condensed matter physics.
Here we have chosen one of the 2D triangular antiferromagnets, i.e., Ag2CrO2.25 This material shows an antiferromagnetic transition at TN = 24 K and exhibits a complex magnetic state, i.e., partially disordered state, below TN.26–28 Because of this complex state, there exists a small magnetic moment even below TN.26,27 Figure 1 shows the crystal structure of Ag2CrO2. The Cr atom, which has a spin angular momentum of S = 3/2, is arranged in a triangular shape with a distance of 2.93 Å. The Ag layer is inserted into the two CrO2 layers. The lattice constant along the c-axis is 8.66 Å. This Ag layer plays a role in electric conduction. Owing to the Ag layer, the coupling between the conduction electrons in Ag and the 3d electrons of Cr is strong. On the other hand, we found that the exfoliation of Ag2CrO2 was not as easy as that for graphite and transition metal dichalcogenides. This property makes it difficult to fabricate a thin film of Ag2CrO2.
In this work, we have developed the method to obtain thin films of Ag2CrO2 from the polycrystalline samples and have performed the transport measurements. A large resistivity drop at around TN was clearly observed in the fabricated thin film devices, as shown in previous reports25,28 for the polycrystalline Ag2CrO2. Interestingly, the measured resistivities for the thin films are about 10 times smaller than those for polycrystalline Ag2CrO2. This fact clearly shows that the crystalline nature of the thin films is much better than that for the polycrystalline samples.
The polycrystalline Ag2CrO2 samples were obtained by encapsulating a mixture of Ag, Ag2O, and Cr2O3 powders in a gold cell, and then by baking it at 1200°C for 1 hour under a pressure of 6 GPa.25 As mentioned above, the mechanical exfoliation was not as easy as for other typical 2D materials such as graphite and transition metal dichalcogenides. Therefore, we took the following recipe to fabricate the thin films of Ag2CrO2. We first pounded Ag2CrO2 samples on a glass plate, in order to obtain small pieces of Ag2CrO2. We picked up the small grains of Ag2CrO2 with a scotch tape and then pasted them onto a thermally oxidized silicon substrate with several 100 nm thick gold marks. After removing the scotch tape from the substrate, we prepared another silicon substrate without any gold marks and pushed it onto the silicon substrate with the Ag2CrO2 flakes and the 100 nm thick gold marks [see Fig. 2(a)]. In this process, some of the Ag2CrO2 flakes become thinner than 100 nm and relatively thick Ag2CrO2 flakes are attached to the substrate without the gold marks. As a result, a few thin Ag2CrO2 flakes remain on the substrate with the gold marks, as shown in Fig. 2(b). These thicknesses have been confirmed by a commercially available atomic force microscope, as shown in Fig. 2(d).
For such thin films of Ag2CrO2, electrodes for transport measurements were attached using a standard electron beam lithography and a subsequent lift-off process. We first coated a polymethyl-methacrylate resist for electron beam lithography and irradiated a certain amount of electron beam. After the development of the resist, 150 nm thick Cu electrodes were deposited by a Joule heating evaporator using a 99.9999% purity source. The final device was obtained after the lift-off process, as shown in Fig. 2(c). While the exfoliated surface of Ag2CrO2 has not been identified in this experiment, we presume that the Ag layer is exposed to the surface since the contact resistance between Ag2CrO2 and Cu is less than 1 Ω, which is comparable to a contact resistance for normal metallic junctions with almost the same junction area.
Figure 3(a) shows the resistivity of the thin Ag2CrO2 film in Fig. 2(c) as a function of temperature T. We note that this is a typical result among four different thin Ag2CrO2 film devices. As in the case of normal metals, the resistivity decreases with decreasing T, but there is a large resistivity reduction at around T = 25 K. At 5 K, the resistivity ρxx of the thin Ag2CrO2 film reaches about 3 μΩ⋅cm. Surprisingly, this resistivity is about ten times smaller than that (∼ 36 μΩ⋅cm) of the polycrystalline samples.25
In order to see the difference between the thin film and the polycrystalline sample more clearly, the temperature dependence of ρxx(T) normalized with ρxx(T = 5 K) is plotted in Fig. 3(b). Apparently, the normalized resistivity for the thin film is much larger and sharper at TN compared to the polycrystalline sample. To characterize the sudden reduction of ρxx at around TN, we plot the derivative of the normalized resistivity for both the thin film and the polycrystalline sample in Fig. 3(c).28 The peak positions indicated by arrows are located at almost the same temperature (about 25 K), which is very close to TN = 24 K determined from heat capacity measurements.25 The peak width for the thin film is narrower and more asymmetric with respect to 25 K compared to that for the polycrystalline sample. All the above results clearly show that the crystalline nature of the exfoliated thin Ag2CrO2 film is much better than that of the polycrystalline samples. Thus, the demonstrated fabrication of the thin film Ag2CrO2 with improved quality paves the new way for device applications of layered antiferromagnetic materials.
In summary, we have fabricated thin films of 2D triangular antiferromagnet Ag2CrO2 using the mechanical exfoliation technique. As in the polycrystalline samples, a large reduction of the resistivity was observed near TN. This sudden reduction near TN was much sharper for the thin film rather than for the polycrystalline sample. The resistivity of the thin film is one order magnitude smaller than that of the polycrystalline sample. These results indicate a much better quality for the thin film device. Such a thin film of 2D antiferromagnetic material could be useful for future applications, for example in spintronic devices.
We thank M. Hagiwara, T. Kida, and A. Okutani for the fruitful discussions. This work was supported by Grants-in-Aid for Scientific Research (Nos. 16H05964, 17K18756, 26103002, 26220711, and 15K17680), Mazda foundation, Shimadzu Science foundation, Yazaki foundation, SCAT foundation, and Murata foundation.