Towards Digital Twin Implementation in Roll-To-Roll Gravure Printed Electronics: Overlay Printing Registration Error...

Area of Science:

• Materials Science and Engineering
• Electrical Engineering
• Manufacturing Technology

Background:

• Roll-to-roll gravure (R2Rg) is crucial for cost-effective printed electronics, particularly thin-film transistors (TFTs).
• Achieving high overlay printing registration accuracy (OPRA) is a significant challenge for multilayer TFT structures in R2Rg processes.
• Integrating TFTs with high precision is essential for advanced active matrix devices.

Purpose of the Study:

• To introduce a digital twin paradigm for optimizing OPRA in R2Rg systems.
• To develop a predictive model for machine directional (MD) error in OPRA using key input variables.
• To enhance the precision and integration of TFTs manufactured via R2Rg.

Main Methods:

• Implementation of a digital twin for the R2Rg system.
• Collection of OPRA error data correlated with web tension, nip force, and printing speed.
• Utilizing a deep learning long-short term memory (LSTM) model trained on statistical features for MD error prediction.

Main Results:

• The trained LSTM model achieved a 77% prediction accuracy for MD error in OPRA.
• A correlation was identified between MD error in OPRA and printing speed.
• Higher printing speeds showed a tendency to reduce the amplitude of MD error in OPRA.

Conclusions:

• The digital twin approach effectively models and predicts OPRA in R2Rg systems.
• Optimizing printing speed is a viable strategy to improve OPRA and reduce MD errors.
• This methodology facilitates higher integration of TFTs in printed electronics.