Epitaxial growth of a 2D transition metal dichalcogenide lateral heterojunction

Transition metal dichalcogenides (TMDs) have previously been recognized as a new class of semiconducting 2D layered materials.^{1, 2} TMDs—such as molybdenum disulfide (MoS₂) and tungsten diselenide (WSe₂)—thus open up new opportunities in semiconductor technology for the development of future 2D electronics and optoelectronics. Furthermore, monolayer TMDs feature a direct energy band gap, good carrier mobility, and excellent ‘on’/‘off’ current ratios when they are fabricated in field effect transistors. These are all advantageous properties for future low-power electronic and optoelectronic devices. To achieve additional applications of these materials in advanced circuits, however, it is necessary to develop a 2D p–n junction (i.e., the interface between a p-type and an n-type semiconductor).

In earlier studies of such p–n junctions, the focus has mainly been on vertically stacked van der Waals heterojunctions in which the interface properties and stacking angle strongly affect the junction performance.^{3, 4} Alternatively, it has been shown that chemical vapor deposition (CVD) can be used reliably for the growth of wafer-scale 2D materials and is feasible for lateral heterojunction growth.⁵ The thermodynamically favored formation of TMD alloys at the hetero-interface during the so-called one-pot synthetic process, however, remains an obstacle to the formation of lateral heterojunctions via direct CVD growth.^6–8 Moreover, the one-pot synthetic process only allows the growth of heterostructures with either different metals or different chalcogens, and this limits the materials that can be selected for practical applications.^6–8

To provide a more reliable and robust growth method for 2D heterojunctions, we have therefore developed a new synthesis approach for lateral junctions with improved structure and availability.⁹ We have demonstrated our two-step epitaxial method for the growth of monolayer heterojunctions with the fabrication of a WSe₂/MoS₂ lateral heterojunction that has an atomically sharp p–n interface. In this process, we first synthesized the monolayer WSe₂ single crystal on a sapphire substrate and then grew the MoS₂ layers in a separate furnace. An optical microscope image of our WSe₂/MoS₂ lateral heterojunction is shown in Figure 1(a).

Figure 1. (a) Optical microscope image of the tungsten diselenide/molybdenum disulfide (WSe₂/MoS₂) lateral heterojunction fabricated using the two-step epitaxial growth method. (b) A high-resolution scanning transmission electron microscope image of the heterojunction, showing the atomically sharp interface between the WSe₂and the MoS₂.

Careful control of the vapor source is a crucial step in our synthesis method. This step allows us to avoid the formation of alloys at the interface during the MoS₂ growth. In our demonstration of the technique we used scanning transmission electron microscopy to carefully examine the interface structure. The atomically sharp interface between the WSe₂ and MoS₂ can be seen in the annular dark field image—see Figure 1(b)—and illustrates the feasibility of our two-step growth process.

We have also performed polarization-resolved second-harmonic generation microscopy measurements to characterize the 2D crystalline structure of our heterojunction. The extracted orientation angle (between the incident laser polarization and the direction of crystal orientation) is shown in Figure 2(a). This information reveals that the MoS₂ had grown epitaxially out from the edge of the WSe₂ with the same crystal orientation, rather than following the substrate orientation. In addition, the optical microscope images in Figure 2(b) show that the MoS₂ was successfully grown from a large area of pre-patterned WSe₂. This implies that there is potential for further versatile electronics design with our two-step growth method.

Figure 2. (a) Map of the orientation angle (θ) that is extracted from the polarization-resolved second-harmonic generation measurement. The white dashed line indicates the interface of the heterojunction. (b) Optical microscope images showing the MoS₂growth from a pre-patterned area of WSe₂. Scale bars indicate 5μm.

To investigate the electrical properties of our WSe₂/MoS₂ lateral heterojunction, we transferred the fabricated structure—see Figure 3(a)—onto a silica/silicon substrate and we deposited two contact metals onto the WSe₂ and MoS₂. The electrical transport curve of this heterojunction—see Figure 3(b)—shows that the junction exhibits good rectification behavior, photoresponses, and photovoltaic effects. These results thus indicate that our WSe₂/MoS₂ lateral heterojunction is a native monolayer p–n junction, which means it can be used as an essential and fundamental building block for future 2D electronics and optoelectronic components.

Figure 3. (a) Optical microscope image of the WSe₂/MoS₂lateral heterojunction used to characterize the electrical properties of the structure. The structure is placed on a silica/silicon substrate, and two contact metals, D and S, are deposited on the WSe₂and MoS₂, respectively. Pd: Palladium. Ti: Titanium. (b) The electrical transport curves of the WSe₂/MoS₂p–n junction with (light) and without (dark) illumination. The results clearly show the rectification behavior and photovoltaic effect of the device. I: Current. V: Voltage. I_SC: Short-circuit current. V_OC: Open-circuit voltage. FF: Fill factor.

In summary, we have successfully demonstrated the growth of a TMD lateral heterojunction. We have also proposed a novel two-step growth method that can be applied to the construction of heterostructures that are formed from various 2D layered materials for a variety of versatile monolayer electronics. Our work is part of the essential fabrication of large-scale and high-quality monolayer components for the development of next-generation 2D electronics and circuits. In addition, the atomically sharp interface of our lateral heterojunction offers an interesting platform for the study of fundamental materials science. In the next stages of our work we intend to explore the fundamentals of our growth mechanism and to interface it with other materials to form various heterostructures.

The authors acknowledge support from King Abdullah University of Science and Technology (Saudi Arabia), Academia Sinica (Taiwan), the Ministry of Science and Technology, and the Taiwan Consortium of Emergent Crystalline Materials.