Professor Wang Xinran's team at Nanjing University achieves new progress in wafer-scale two-dimensional semiconductor stacking control: Nature Materials

On July 9, 2025, a research paper titled "Homoepitaxial growth of large-area rhombohedral-stacked MoS₂" was published online in Nature Materials by Professor Xinran Wang's research group at Nanjing University. This achievement reports the homoepitaxial growth of wafer-scale rhombohedral (3R) molybdenum disulfide (MoS₂), and demonstrates its applications in ferroelectric devices and storage. This finding marks a significant new advance in the controllable preparation of two-dimensional materials and brings new opportunities for multifunctional heterogeneous integration of two-dimensional materials. The corresponding author of the paper is Professor Xinran Wang, and the co-first authors are Dr. Lei Liu, Associate Professor Taotao Li, Dr. Xiaoshu Gong, and Dr. Hengdi Wen.

Two-dimensional semiconductors, due to their atomic thickness, high mobility, and compatibility with three-dimensional integration, have become important candidate materials for extending Moore's Law and building three-dimensional integrated circuits in the post-silicon era. Professor Xinran Wang's research group has been committed to the controllable growth of two-dimensional transition metal dichalcogenides (TMDCs) for a long time, and has achieved a series of results in in-plane orientation control, layer number control, and stacking control: creating a theory of TMDC oriented epitaxial growth, revealing the TMDC nucleation mechanism induced by atomic steps on the sapphire substrate surface, establishing an oriented epitaxial relationship, and internationally achieving the first breakthrough in wafer-scale two-dimensional semiconductor single-crystal epitaxial preparation (Nature Nanotech., 16, 1201 (2021)); proposing the idea of controlling the number of nucleation layers by step height, breaking through the precise control technology of TMDC layer number, and preparing large-area uniform bilayer MoS2 for the first time (Nature, 605, 69 (2022)). In this work, the stacking control of two-dimensional semiconductors is further broken through, realizing the controllable preparation of 3R-stacked MoS2, which opens up a new research dimension for physical property control and device research.

In two-dimensional materials, "stacking" refers to the arrangement of atomic layers. If we consider one layer of MoS2 as a sheet of paper, stacking is like a stack of papers that can be stacked with different rotation angles or different sliding methods. Different stacking methods change the relative positions of atoms, which significantly affects the electronic structure and physical properties of the material at the nanoscale. The recently popular "magic angle" (i.e., small-angle rotation between layers) is considered a special artificially constructed stacking form. In nature, there are two common stacking methods for MoS₂: hexagonal (2H) and rhombohedral (3R). The latter, due to its lack of centrosymmetry, exhibits excellent nonlinear optical, valleytronic, and ferroelectric properties, making it particularly suitable for building new types of memory and optoelectronic devices. However, because the 2H and 3R structures are almost equally stable thermodynamically, it is difficult to strictly control the stacking method during growth, which is one of the major challenges in the field of two-dimensional material preparation.

To overcome this difficulty, the research team used a homoepitaxial strategy. Using high-quality single-layer single-crystal MoS₂ as the epitaxial substrate, and by precisely controlling the concentration of transition metal precursors, the preparation of multi-layer MoS₂ wafers with pure 3R phase was successfully achieved (Figure 1). Using artificial intelligence image recognition technology to automatically identify and statistically analyze the stacking structure, it was confirmed that the proportion of the 3R phase was close to 100%.

Figure 1: Wafer-scale rhombohedral multi-layer MoS₂.

To further reveal the mechanism of selective formation of the 3R phase, the research team conducted theoretical calculations in collaboration with Professor Jinlan Wang's team at Southeast University. The results show that the Mo substituting S defect (MoS) in crystal defects can significantly increase the difference in formation energy between the two stacking methods from 1 meV/MoS2 to 75 meV/MoS₂. Experimentally, the team focused on the initial stage of nucleation, using high-resolution STEM to observe clusters <10 nm, confirming the promoting effect of MoS defects on the selective growth of 3R (Figure 2). Based on the cross-validation of theory and experiment, the team proposed a growth mechanism of defect-promoted selective nucleation in homoepitaxy. The proposal of this theory provides a new mechanism and idea for the structural control of two-dimensional materials, bringing new breakthroughs and hope to the research in this field.

Figure 2: Growth mechanism of rhombohedral MoS₂.

In addition, this work also revealed the shear ferroelectricity exhibited by 3R-MoS₂ (Figure 3). Experiments using piezoelectric force microscopy (PFM) clearly observed ferroelectric domains and a significant piezoelectric response hysteresis loop, and constructed an ultra-thin ferroelectric transistor array using bilayer 3R-MoS₂ as the channel material. Although the channel material thickness is only 1.3 nm, it still exhibits data retention capability exceeding ten years, excellent conductivity, and 16-bit multi-state writing capability, injecting new impetus into the development of future high-density, low-power, non-volatile memory devices.

Figure 3: Ferroelectricity of rhombohedral MoS₂.

Original link:

https://doi.org/10.1038/s41563-025-02274-y