Dissolution Rate Characterization of LOR10C Resist on Polyimide Substrates Using Coherence Scanning Interferometry
Brandon Chacon
Advisors: Shawn Sanctis, Ph.D., Arvind Dasgupta, KJ Seo, Ph.D., and Jonathan Reeder Ph.D.(Science Corporation)
Dissolution Rate Characterization of LOR10C Resist on Polyimide Substrates Using Coherence Scanning Interferometry
Brandon Chacon
Advisors: Shawn Sanctis, Ph.D., Arvind Dasgupta, KJ Seo, Ph.D., and Jonathan Reeder Ph.D.(Science Corporation)
Bilayer lift-off lithography using LOR10C sacrificial resist is the dominant technique for producing controlled undercut profiles in flexible brain-computer-interface device fabrication, yet develop-time selection currently relies on operator experience rather than on a predictive physical model. Against a 1.0 µm to 2.0 µm undercut specification window, the baseline 45-second develop recipe produces a process capability index Cpk of −5.25, placing the process mean entirely outside specification. This project applies a Six Sigma DMAIC framework to characterize LOR10C dissolution rate as a function of exposure dose and MLA150 defocus on a custom copper-on-polyimide witness wafer stack, with activation energy extracted via a two-point Arrhenius analysis across the two production bilayer variants. A non-destructive Zygo ZeGage Pro coherence scanning interferometry measurement chain was formally qualified at 9.75 percent equipment variation, and spin-curve data on the Cu/polyimide stack agreed with Kayaku datasheet reference values within ±7 percent. Ongoing work will complete the Design of Experiments, extract Arrhenius kinetic parameters, and deliver a predictive develop-time equation embedded in a Python-based lookup tool. A confirmation run executed at the model-predicted develop time is expected to demonstrate process-capability improvement from Cpk = −5.25 to Cpk ≥ 1.33, establishing industrial-grade discipline in low-volume implantable-device fabrication.