UCLA

Topological materials are materials whose electronic band structures are described by certain non-trivial topological invariants. Forty years ago the importance of band topology in condensed matter physics was first recognized when the quantum Hall effect (QHE) was found to be related with the integer Chern number in two-dimensional (2D) electron gas. Since 2008, the discovery of three-dimensional (3D) topological insulators (TI) with a non-trivial topological invariant and gapless surface state has taken the field into a new era. Various new topological phases were proposed and band topology has become a new way to classify the state of matter. The design, synthesis and characterization of new topological materials pave essential basis to uncovering novel physics arising from non-trivial band topology and its interplay with various degrees of freedom such as spin, orbital and charge. Today, with more sought-after novel topological phases, emergent phenomena such as surface Fermi arcs, chiral anomaly, quantum anomalous Hall effect were discovered and enable future technological advances including topological quantum computation. A new topological phase can be created when additional symmetry breaking is introduced into an existing topological phase. For example, by breaking the time reversal symmetry of a 3D TI through ferromagnetism (FM), one can get a Chern insulator in its 2D limit, where QHE can be realized without external magnetic field and gives topologically-protected dissipationless chiral edge states. This phenomenon, the so-called quantum anomalous Hall effect (QAHE), has been long sought since its early proposal in the yet-to-be-realized Haldane model for graphene lattice with opposite magnetic field at neighboring atoms in 1988. Therefore, the realization of QAHE in magnetically-doped TI Cr0.15(Bi0.1Sb0.9)1.85Te3 thin films in 2013 was revolutionary. However, the unavoidable sample inhomogeneity in doped materials restrains the investigation of associated emergent phenomena in mK-regime. Ideally, magnetism from intrinsic magnetic atoms in a crystal can provide more homogeneous electronic and magnetic properties than the magnetism from dopants. To realize QAHE at higher temperatures, the intrinsic magnetic TIs with only clean topological bands but no other bands at the Fermi level are strongly desired. In 2018, MnBi2Te4 was discovered to be the first of such kinds, as an antiferromagnetic (AFM) TI with intrinsic magnetic Mn site. It is a layered van der Waals (vdW) material. When the magnetism orders below 24 K, the spins are FM aligned in the ab plane but AFM coupled along the c axis. In 2D limit, MnBi2Te4 films can have a net magnetization either in odd-layer devices, or when the even-layer devices are in the spin-flop state above ~ 3.5 T and the forced FM state above ~ 8 T. These time-reversal-symmetry breaking states are ideal for realizing the Chern insulator state. Indeed, QAHE was experimentally observed at 0 T and 1.6 K in a 5-layer device and quantized Hall conductance was realized when the even-layer devices enter the forced FM state above the saturation field of 8 T.

Following this line, for QAHE to be realized at zero field and higher temperature, it is strongly desirable if the FM alignment of Mn spins can be accessed at a lower or even zero field. To do so, one must weaken the interlayer AFM interactions between [MnBi2Te4] layers. We thus propose to introduce n-1 nonmagnetic TI [Bi2Te3] layers between [MnBi2Te4] layers to get natural heterostructures of MnBi2nTe3n+1. By this rational design, we can increase the distance between the neighboring [MnBi2Te4] layers and thus reduce the interlayer AFM interaction. Under such a design principle we successfully grew single crystals of MnBi4Te7 (n=2), MnBi6Te10 (n=3) and MnBi8Te13 (n=4). Then with the physical property characterization, first-principles calculations and angle-resolved photoemission spectroscopy measurements, for the first time, we demonstrated that MnBi4Te7 is an intrinsic AFM TI with saturation field 40 times smaller than that of MnBi2Te4, and that MnBi8Te13 is the first realization of an intrinsic FM axion insulator, proving the success of our material design principle.

The manipulation of magnetism is crucial to access different magnetic topological phase and novel physics. In MnBi2nTe3n+1, the control of the magnetism from AFM to FM by n is only discrete. To achieve a fine and continuous control of the magnetic transition, we doped Sb to MnBi4Te7 where the interlayer AFM coupling is weak and more tunable. Through single crystal growth, transport, thermodynamic, neutron diffraction measurements, we show that under Sb doping, MnBi4Te7 evolves from AFM to FM and then ferrimagnetic. We attribute this to the formation of Mn_(Bi,Sb) antisites upon doping, which results in additional Mn sublattices that modify the delicate interlayer magnetic interactions and changes the overall magnetism. We further investigate the effect of antisites on the band topology using the first-principles calculations. Without considering antisites, the series evolves from AFM topological insulator (x = 0) to FM axion insulators. In the exaggerated case of 16.7\% of periodic antisites, the band topology is modified and type-I magnetic Weyl semimetal phase can be realized at intermediate doping. Therefore, this doping series provides a fruitful platform with rich and continuously tunable magnetism and topology.

After we achieve FM in MnBi2nTe3n+1, for practical applications especially in the pursuit of high temperature QAHE when fluctuations become important, the study on magnetic dynamics is indispensable too. We investigated the magnetic dynamics in FM MnBi8Te13 and Sb doped MnBi4Te7 and MnBi6Te10 using AC susceptibility and magneto-optical imaging. Slow relaxation behavior is observed in all three compounds, suggesting its universality among FM MnBi2nTe3n+1. The origin of the relaxation behavior is attributed to the irreversible domain movements since they only appear below the saturation fields when FM domains form and evolve. These FM domains are very soft, as revealed by the low-field fine-structured domains and high-field sea-urchin-shaped remnant-state domains imaged via the magneto-optical measurements. Finally, we attribute the rare “double-peak” behavior observed in the AC susceptibility under small DC bias fields to the very soft FM domain formations. This study provides a thorough understanding of the soft FM in highly anisotropic magnets.

As the first intrinsic antiferromagnetic topological insulator, MnBi2Te4 is still the major material platform to search for QAHE, so its material optimization is very urged. We develop the chemical-vapor-transport (CVT) growth for of MnBi2Te4, which has a higher success rate in observation of the field-induced quantized Hall conductance in 6-layer devices. Through comparative studies between our CVT-grown and flux-grown MnBi2Te4, we find that CVT-grown MnBi2Te4 is marked with higher Mn occupancy on the Mn site, slightly higher Mn_Bi antisites and smaller carrier concentration. On the device end, thin film from CVT-grown sample shows by far the highest mobility of 2500 cm2 V s in MnBi2Te4 devices with the quantized Hall conductance appearing at 1.8 K and 8 T. This study provides a route to obtain high-quality single crystals of MnBi2Te4 that are promising to make superior devices and realize emergent phenomena.

In summary, we have discovered and established MnBi4Te7 and MnBi8Te13 as new intrinsic magnetic topological insulators. In particular, we provide deep understanding of the importance of material design, synthesis and chemical doping to the magnetism and topology in the series. The growths of high-quality single crystals and the study of magnetic dynamics provide essential basis for the search of QAHE in MnBi2nTe3n+1. Our works will shed light on future endeavors in finding novel magnetic topological materials as well as searching for QAHE and the associated emergent phenomena in the condensed matter field.