Van der Waals epitaxial growth of air-stable CrSe2 nanosheets with thickness-tunable magnetic order

This work is the first to show that interlayer distance can be used as a new degree of freedom to regulate two-dimensional magnetism, highlighting the importance of non-metallic elements and their interlayer Pauli repulsion not considered in previous models.

We found that the above magnetic coupling mechanism is also applicable to the layers. With the increase of layer thickness, the kinetic energy of electrons moving along the direction of CrSe₂ layer increases, gradually represses the interlayer antiferromagnetic coupling caused by Pauli and Coulomb repulsion, and macro ferromagnetism gradually takes the upper hand. In cooperation with the experimental groups, we confirmed the above theoretical image, which is the first international confirmation that CrSe₂ is a two-dimensional magnetic material with thickness-dependent magnetic coupling transition between layers and is highly stable in air^[2] (FIG. 1 c-e). It has initially overcome the difficulty of insufficient air stability of two-dimensional magnetic materials that had puzzled the scientific community for a long time. As soon as this work was published, it quickly gained academic attention. Professor Andrew T.S. Wee from the National University of Singapore quoted and positively appraised this work for a long time, and Ki Kang Kim from Sungkyunkwan University in Korea wrote a review article affirming this work.

A family of high-temperature ferromagnetic monolayers with locked spin-dichroism-mobility anisotropy: MnNX and CrCX (X = Cl, Br, I; C = S, Se, Te)

Based on the above thought, 12 kinds of single-layer ferromagnetic materials with high Tc are predicted: MnNX and CrCX (where X=Cl,Br,I; C=S,Se,Te)^[1]. Based on the anisotropic Heisenberg model and renormalized spin wave theory, Tc of this series of materials is predicted to range from 100K to 500K (Figure 1a). Eight members among them show semiconducting bandgaps varying from roughly 0.23 to 1.85 eV. These semiconducting monolayers also show extremely large anisotropy along the two in-plane lattice directions of these layers. Additional orbital anisotropy leads to a spin-locked linear dichroism, in different from previously known circular and linear dichroisms in layered materials. Together with the mobility anisotropy, it offers a spin-, dichroism- and mobility-anisotropy locking. These results manifest the potential of this 2D family for both fundamental research and high performance spin-dependent electronic and optoelectronic devices (Figure 1b).

The work was published in Sci.Bull. 2019, making it one of the 10 most downloaded articles of that year. The material predictions has been verified and cited by several different experimental groups in Nat. Mat (1 paper), Nat. Nanotech. (2 papers), Adv. Mater. (2 papers) and Nano Letters (1 paper). The magnetic properties and anisotropic electronic structure were confirmed to be consistent with our theoretical prediction.

On this basis, the researchers, in collaboration with experimental collaborators, determined the bulk phase magnetic ground state and magnetic field evolution of CrOCl^[2] (FIG. 1c-e). Combined with the magnetic transport measurements of experimental collaborators, the layered dependent magnetic field evolution behavior of the CrSBr system was investigated^[3].

Spin mapping of intralayer antiferromagnetism and field-induced spin reorientation in monolayer CrTe2

A question then arises of how do strongly overlapped interlayer wave functions affect interlayer magnetism and whether there are any generalized spin-exchange coupling mechanisms solely for such a noncovalent interaction. The interlayer differential charge density shows that the interlayer electron coupling of of CrX₂(X = S, Se, Te) systems is significantly stronger than that of CrI₃. It is expected that different interlayer magnetic coupling mechanisms exist in CrX₂. The bulk phase of this kind of material shows ferromagnetism, and many studies speculate that the single layer should also be the ferromagnetic ground state. However, the theoretical prediction of the researchers found that the single layer of this kind of material is striped antiferromagnetic ground state. The in-plane magnetic orders of CrX₂ mono and few layers are tunable between striped antiferromagnetic (sAFM) and ferromagnetic (FM) orders by manipulating charge transfer between Cr t2g and eg orbitals. Such charge transfer is realizable through interlayer coupling, direct charge doping, or substituting S with Cl atoms^[1,2].

However, due to the fact that antiferromagnetic orders usually do not exhibit macroscopic magnetic moments, experimental detection is very difficult, and the antiferromagnetic ground state of CrX₂(X=S,Se,Te) materials mentioned in the above theoretical prediction are still lacking of experimental evidence in the two-dimensional limit. By combining spin polarised scanning tunnelling microscopy and first-principles calculations, we investigate the magnetism of vdW ML CrTe₂, which has been successfully grown through molecular-beam epitaxy. We observe a stable antiferromagnetic (AFM) order at the atomic scale in the ML crystal, whose bulk is ferromagnetic, and correlate its imaged zigzag spin texture with the atomic lattice structure. confirming the theoretical prediction of the researchers (FIG. 1 c). Previous studies of two-dimensional magnetic materials usually assumed that the material’s easy magnetization axis was along the conventional lattice base vector or out of plane direction. The researchers calculated that the magnetic moment direction of the CrTe₂ single layer was in the y-z plane at an Angle of 70 degrees from the z direction (FIG. 1 d), resulting in a novel spin reorientation transition mode under magnetic field (FIG. 1 e-g). The real space intrinsic antiferromagnetic order discovered and confirmed in this study, the magnetic order transformation in the single-layer limit, and the spin reorientation under external magnetic fields have not been previously reported before^[3].

Stacking tunable interlayer magnetism in bilayer CrI₃

Bulk CrI3 exhibits a vdW structure and possesses a rhombohedral structure with the R3 space group symmetry at low temperature (the LT phase). When temperature increases to 210–220 K, it undergoes a structural phase transition to a monoclinic lattice with the C2/m space group symmetry (the HT phase) It is generally believed that the experimental low temperature measurement is a low temperature stable structure. However, the CrI3 bilayer prepared at room temperature may not undergo structural transformation during the rapid cooling process and maintain the high-temperature phase structure. The researchers speculate that different stacking in the CrI3 system may correspond to different interlayer magnetic ground states. Based on the above prediction, we found that the interlayer AFM coupling results from a different stacking order with the C2/m space group symmetry, rather than the graphene-like one with R3 as previously believed^[2].

This work solves the problem of the source of the AFM coupling between the two layers of CrI3 in a series of recent experiments and proposes a magnetic coupling mechanism in the weak non-covalent coupling limit. In the past, it was generally assumed that the van der Waals (vdW) interaction dominated the interlayer interaction of two-dimensional materials, and the effect of interlayer coupling on magnetic regulation was significantly underestimated. This work realizes from a new perspective that after the p orbitals of I are polarized by Cr into pxy and pz orbitals with opposite magnetization directions, different interlayer stacking orders leads to different direct exchange between p orbitals of interlayer I atoms, which ultimately determines the magnetic coupling between layers.

This work has been cited positively by Professor Cheung Xiang and Professor Novoselov. The theoretical predictions of this work have been verified and cited by many different experimental groups published in Nature, Science, Nature sub-journals and PRL. Further, in collaboration with experimental collaborators, the theoretical image of orbital splitting of I-pxy and pz was verified by combining scanning tunneling microscopy (STM) and density functional theory (DFT) calculations^[3].