The discovery of graphene has opened the door to two-dimensional materials for human beings, and after more than ten years of research, the various excellent properties of two-dimensional materials have become more and more attractive to scientific researchers. However, there are still many problems in the large-scale application of two-dimensional materials in industry, and the devices made cannot meet industrial standards.
Professor Bao Wenzhong of Fudan University used machine learning (ML) algorithms to optimize the fabrication process of two-dimensional semiconductor (MoS2) top-gate field-effect transistors (FETs), and used industry-standard design processes and processes to fabricate and test wafer-level devices and circuits. The article was published in Nature Communications under the title "Wafer-scale functional circuits based on two dimensional semiconductors with fabrication optimized by machine learning." In this paper, wafer-size device fabrication is optimized using machine learning to guide the fabrication process, followed by a small benchtop maskless lithography machine, MicroWriter ML3, which optimizes performance such as mobility, threshold voltage, and subthreshold swing.
Quantum Design - Small desktop maskless direct writing lithography system - MicroWriter ML3
Illustrated guide
Figure 1. Prepare a general flowchart of MoS2 FETs. (a) Wafer-sized MoS2 is prepared by CVD method. (b) Various cross-sectional views of the MoS2 FET. (c) The relationship between the performance of transistors and various parameters. (d) Optimized feedback loop from material preparation to chip fabrication and testing.
Figure 2. Logic circuit diagram of MoS2 FETs. Effects of voltages (a), (b), (c) and (d) on the device. (e) Positive and negative using the MicroWriter ML3 maskless laser direct writing mechanism and (f) corresponding experimental results, (g) Adders prepared using the MicroWriter ML3 maskless laser direct writing mechanism and (h) corresponding experimental results.
Figure 3. Simulation, memory and optoelectronic circuits prepared using MoS2 FETs. (a) Ring oscillator prepared using a maskless lithography mechanism and (b) corresponding experimental results. (c) Storage arrays prepared based on MoS2 FETs and corresponding experimental results (d-f). (g) Photoelectric circuits prepared using MicroWriter ML3 and corresponding performance results (h-i).
Figure 4. Prepare MoS2 FETs on wafers using the MicroWriter ML3 maskless laser direct-writer. (a) MoS2 FET-based adder prepared on a two-inch wafer. (b), (c), and (d) prepare the results of the adder on the wafer.
conclusion
With the application of two-dimensional materials and the rapid development of artificial intelligence in various fields, how to quickly develop prototype chip structures that meet experimental design has become very important. Since the corresponding parameters need to be modified in time during the experiment to obtain optimized experimental results, it relies heavily on flexible lithographic means. As can be seen from the above, the small desktop maskless lithography machine MicroWriter ML3 can help users quickly realize the development of various logic structures and help research in microelectronics related fields.
MicroWriter ML3, a small desktop maskless lithography machine
In view of the excellent performance and high output of the first small desktop maskless lithography machine MicroWriter ML3, the relevant research team of the research group continued to follow the hot spots, and seized the opportunity to add a new small desktop maskless lithography machine - MicroWriter ML3 Pro+0.4 μm Professional Edition system of the British DMO company to strive for better device performance. Hope to help the research team make important progress!
Equipment consultation
If you are interested in our equipment or technology, please click the link below or send us a private message to get more equipment information!
Quantum Design - Small desktop maskless direct writing lithography system - MicroWriter ML3