Flexible electronics (FE) have electronic components and devices for sensing elements, either assembled within or imprinted on the flexible substrate such that they touch the skin or appropriately interface with the body and can be worn on the human body for health monitoring.  The versatility of applications, and the lightweight and compact structure of FE, along with the ever-expanding consumer goods industry, strongly spurred the growth of the FE market and research. The ever-changing FE market demands a high yield and low cost for manufacturing FE products. The emerging roll-to-roll (R2R) FE printing allows millions of nano- or micron-scale patterns to be printed on a continuously moving web across a meter scale width, hence producing a large volume of FE products efficiently. Due to the delicate nature and small scale of the nanomanufacturing process, non-contact optical imaging techniques are an attractive solution for rapidly assessing the online R2R print quality. However, optical limits such as depth of focus, field of view, and resolution, along with challenges such as extremely thin patterns and loss of registration of the pattern hinder the successful implementation of optical imaging for inspecting nanomanufacturing processes. To bridge this gap, we intelligently integrated multimode and multiscale imaging information to monitor and control print pattern quality. We achieved (1) ln-line metrology of print pattern quality in nano-thin monolayer scale using condensation figures (CF), (2) super-resolution and image registration techniques for moving large area nano- and micro-scale flexible patterns, (3) predictive control of roll-to-roll (R2R) printing through neural network (NN) model, and (4) fast and precise autofocus solutions for real-time R2R imaging. The results of the completed research can be applied to predictive regulation of the R2R print process to ensure pattern functionality and integrity.

Under the GOALI initiative, this project involved a collaboration of industrial partners from Carpe Diem. Over the past four years, two of our graduate students have undergone comprehensive training by Carpe Diem engineers at the R2R core facility. This training encompassed the design, utilization, and maintenance of a wide array of industrial-scale coating, cleaning, and printing equipment used in the production of industrial goods. During this period, our graduate students have also closely collaborated with Carpe Diem engineers on testing and expanding the applicability of R2R sensing and control algorithms in other R2R printing processes, such as Gravure printing. Additionally, two graduate students were co-supported by this grant and NSF INTERN for an internship at the National Renewable Energy Lab (NREL). They investigated spectrum imaging and deep learning methods for real-time thickness measurement and defect detection on polymer-electrolyte-membrane (PEM) fuel cells that were produced by R2R coating at NREL.

This project aids in the development of metrology and imaging techniques for inline monitoring and quality control of R2R printing processes. This integration of these techniques into the R2R printing process models allows us to: (1) rapidly detect variations in the printed pattern; (2) and take necessary actions to regulate and maintain the desired quality standards. By incorporating these metrology techniques into various R2R print control systems, we have enhanced the efficiency, reliability, and productivity of the printing processes while ensuring high-quality outputs.

Last Modified: 12/22/2023
Modified by: Xian Du