Improving high throughput manufacture of laser-inscribed graphene electrodes via hierarchical clustering.

ABSTRACT

Laser-inscribed graphene (LIG), initially developed for graphene supercapacitors, has found widespread use in sensor research and development, particularly as a platform for low-cost electrochemical sensing. However, batch-to-batch variation in LIG fabrication introduces uncertainty that cannot be adequately tracked during manufacturing process, limiting scalability. Therefore, there is an urgent need for robust quality control (QC) methodologies to identify and select similar and functional LIG electrodes for sensor fabrication. For the first time, we have developed a statistical workflow and an open-source hierarchical clustering tool for QC analysis in LIG electrode fabrication. The QC process was challenged with multi-operator cyclic voltammetry (CV) data for bare and metalized LIG. As a proof of concept, we employed the developed QC process for laboratory-scale manufacturing of LIG-based biosensors. The study demonstrates that our QC process can rapidly identify similar LIG electrodes from large batches (n ≥ 36) of electrodes, leading to a reduction in biosensor measurement variation by approximately 13% compared to the control group without QC. The statistical workflow and open-source code presented here provide a versatile toolkit for clustering analysis, opening a pathway toward scalable manufacturing of LIG electrodes in sensing. In addition, we establish a data repository for further study of LIG variation.