Enhancing inline quality control: Machine learning for full scale 3D prediction of screen-printed silver contacts
Source abstract
Abstract
In the production of silicon solar cells, screen-printing is the industry standard for the application of silver electrodes. With regard to fine-line printing (<15 μm), however, this process often leads to inefficient silver consumption and increased lateral resistance of the grid due to irregularities such as mesh marks or interruptions. In this work we present insights into a full-scale software approach to evaluate the quality of the contact finger geometry inline and directly after printing. Several fully automatic image processing pipelines are developed to generate training data using images of a 2D microscope and a 3D confocal laser scanning microscope (CLSM). For inline use, another image processing pipeline is being developed to recognize structures fully automatically and provide them in the correct format. The use of machine learning methods enables a prediction of the 3-dimensional finger geometry, which can then be used to draw conclusions about the impact of parameter changes on the finger geometry. After training, the model's predicted height maps are compared with ground truth data, which are the actual measured values used to assess the model's accuracy. The results demonstrate the model's feasibility and reliability, extending its applicability to new, previously unseen data from screen-printed contact fingers. Based on the mean squared error (MSE), a prediction accuracy of 97,4 % was achieved. The 3-dimensional finger structures of a complete wafer (e.g. M2 format, 100 fingers) can be predicted within 338 ms which makes permanent inline use possible within the state-of-the-art cycle time. • Full scale software approach for inline evaluation of contact finger geometry in silicon solar cell metallization. • Creation of fully automatic image processing pipelines for generating training data from 2D and 3D microscopy images. • Using machine learning methods to predict 3-dimensional finger geometry with a prediction accuracy of 97.4 %. • Prediction of full format 3D finger structures within 338 ms for permanent inline use.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: 3D IC and TSV technologies · Electrical Contact Performance and Analysis · Electronic Packaging and Soldering Technologies
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (23)
Show source-provided references
1. Silver‐lean metallization and hybrid contacts via plating on screen‐printed metal for silicon solar cells manufacturing10.1002/pip.3799 ↗
2. AOI for automotive industry - a quality assessment approach combining 2D and 3D sensors
3. Prediction of screen‐printed electrodes with fine‐line and arbitrary structures
4. M. S. Tom Hoger, Numerical Simulations of the Screen-Printing Process for Solar Cell Metallization. [Online]. Available: https://publica.fraunhofer.de/entities/publication/c9305f3d-8c4d-4159-a26b-9dc72f3496af.
5. Deep learning-based prediction of 3-dimensional silver contact shapes enabling improved quality control in solar cell metallization10.1016/j.egyai.2024.100404 ↗
6. Supervised machine learning: a brief primer10.1016/j.beth.2020.05.002 ↗
7. The SIFT algorithm for fundamental frequency estimation10.1109/TAU.1972.1162410 ↗
8. DeTone · 2017
9. Lindenberger · 2023
10. Investigation on the effect of a Gaussian Blur in image filtering and segmentation: zadar
11. Howse · 2013
12. Bressert · 2012
13. Otsu method and K-means
14. Ronneberger · 2015
15. Chen · 2017
16. Huang · 2016
17. Mou · 2018
18. Deconvolution and checkerboard artifacts10.23915/distill.00003 ↗
19. Kingma · 2014
20. Dozat · 2016
21. Mean squared error, deconstructed10.1029/2021MS002681 ↗
22. Printing technologies for silicon solar cell metallization: a comprehensive review10.1002/pip.3674 ↗
23. The problem of overfitting10.1021/ci0342472 ↗