Abstract
The degree of crosslinking in encapsulants is a critical parameter in photovoltaic (PV) module production, significantly influencing module performance and reliability. Despite its importance, the industry-standard Soxhlet extraction method for assessing crosslinking is offline, time-intensive, and unsuitable to implement for real-time process monitoring. This study explores the application of near-infrared (NIR) spectroscopy as a faster, nondestructive alternative for determining encapsulant crosslinking. Test laminates using an ethylene vinyl acetate (EVA) encapsulant with varying crosslinking times were analyzed using both Soxhlet extraction and NIR spectroscopy. The NIR spectra were processed using multivariate data analysis methods for qualitative classification and quantitative prediction. The classification model demonstrated clear separation between encapsulants with high and low degrees of crosslinking. The prediction model achieved a high accuracy prediction of the degree of crosslinking. These findings highlight the potential of NIR spectroscopy for rapid, inline classification and quantification of encapsulant crosslinking. Future work will expand the calibration models to include polyolefin (POE) and co-extruded POE–EVA encapsulants to verify robustness across different chemistries, and optimizing measurement setups to accommodate double-glass module designs.
Funding Information
  • COMET-program
  • Austrian Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology
  • Federal Ministry for Labour and Economy

This publication has 25 references indexed in Scilit: