(151a) Advanced Manufacturing of Thin-Film Devices from Laser-Induced Graphene

Laser-Induced Graphene (LIG) has excellent electrical properties with great potential for in-space manufacturing applications, such as wearable flexible sensors and energy storage devices. However, LIG in-space synthesis faces the problem of material scarcity that requires advanced manufacturing techniques. LIG is produced via laser irradiation of graphene oxide (GO) thin films or Kapton polyimide to pattern conductive lines on the substrate, a process that consumes valuable time and material if optimized through traditional development methods. We utilize a state-of-the-art machine learning approach called Bayesian optimization, which explores parameter combinations within set boundaries to identify promising areas for optimization. This approach is sample efficient and accelerates production of high-quality LIG to shave off weeks of human experimentation time to days. We demonstrate this approach on GO and Kapton substrates to optimize structural and electrical properties, and present work on transferring the machine-learning-trained parameters to pattern low resistance lines for device applications.