Researchers from Massachusetts Institute of Technology (MIT) and the Chinese University of Hong Kong have developed a digital simulator that mimics the photolithography process, a technique used to manufacture computer chips and optical devices. The project marks the first use of actual data from a photolithography system in the construction of a simulator.
This advancement could assist engineers in the creation of more precise devices for diverse applications including mobile cameras, telecommunication systems, and medical imaging. The simulator taps into machine learning algorithms and neural networks to model the behaviour of the photolithography process as closely as possible.
Bringing the digital simulated model to life required new methods to correctly assemble the software and hardware to build a high fidelity dataset. According to Cheng Zheng, a co-lead author of the paper, these new models with real data are more efficient and accurate than simulators that use analytical equations. The data was sourced by generating numerous designs combining different sizes and shapes, which were then manufactured using a photolithography system.
The digital lithography simulator is made of an optics model that demonstrates how light is projected on a device’s surface and a resist model that illustrates how the photochemical reaction produces features. Within this framework, the research team also connects the simulated model to another system, which predicts how the device, once fabricated, will perform.
To test the system, the team fabricated a holographic element which created a butterfly image when light shone on it. The element, etched using their method, produced a near-perfect butterfly which closely resembled the original design. The research team has announced plans to enhance their algorithms for modelling more complex devices and test the system using consumer cameras. Moreover, the team is aiming to develop the system for application in a broader range of photolithography processes that use deep or extreme ultraviolet light.
The project, supported by the U.S. National Institutes of Health, Fujikura Ltd, and the Hong Kong Innovation and Technology Fund, involved use of MIT’s facilities.