Polyimide-Based Neural Interfaces: From Implantation to Microscopy: Stability Across Fixation and Storage Conditions.

Abstract

Decades of studies have enabled a constant optimization of thin-film-based neural implants, with technologies designed to adapt to the biological surroundings through biocompatible and miniaturized materials. Careful assessment of these materials is essential prior to photolithographic fabrication of multilayered devices, as delamination from adhesion mismatches between neighboring layers can compromise recording or stimulation and lead to channel crosstalk or device failure. Multi-layer interactions are therefore closely examined to tailor devices for long-term stability, yet post-implantation examinations still reveal failure modes that are not fully addressed and understood. Here, we propose an in vitro investigation of polyimide (PI), a widely used substrate material, in a simplified single layer configuration, to isolate substrate-related effects and evaluate long-term stability under combined conditions that cover not only the implantation period, but also post-implantation steps. Accelerated ageing was used to mimic processes occurring in the brain during implantation while aged samples were also exposed to aldehyde-based fixation and storage conditions commonly encountered during post-mortem processing. Surface morphology was assessed using Scanning Electron Microscopy (SEM). Chemical microstructure was analyzed in a dual approach using Time-of-Flight-Secondary-Ion Mass-Spectroscopy (ToF-SIMS) and Fourier Transform Infrared Spectroscopy (FTIR). PI remains morphologically and chemically stable under the harshest conditions, in both implantation scenarios and subsequent post-implantation steps. These findings provide critical validation for the use of PI as a durable substrate in long-term neural interfaces. They further indicate that damage observed after implantation may arise from factors beyond intrinsic PI degradation, helping prevent misinterpretation during post-implantation analysis.