Neural electrode resilience against dielectric damage may be improved by use of highly doped silicon as a conductive material

Abstract

Background: Dielectric damage occurring in vivo to neural electrodes, leading to conductive material exposure and impedance reduction over time, limits the functional lifetime and clinical viability of neuroprosthetics. We used silicon micromachined Utah Electrode Arrays (UEAs) with iridium oxide (IrO_x) tip metallization and parylene C dielectric encapsulation to understand the factors affecting device resilience and drive improvements.

New method: In vitro impedance measurements and finite element analyses were conducted to evaluate how exposed surface area of silicon and IrO_x affect UEA properties. Through an aggressive in vitro reactive accelerated aging (RAA) protocol, in vivo parylene degradation was simulated on UEAs to explore agreement with our models. Electrochemical properties of silicon and other common electrode materials were compared to help inform material choice in future neural electrode designs.

Results: Exposure of silicon on UEAs was found to primarily affect impedance at frequencies >1kHz, while characteristics at 1 kHz and below were largely unchanged. Post-RAA impedance reduction of UEAs was mitigated in cases where dielectric damage was more likely to expose silicon instead of IrO_x. Silicon was found to have a per-area electrochemical impedance >10×higher than many common electrode materials regardless of doping level and resistivity, making it best suited for use as a low-shunting conductor.

Comparison with existing methods: Non-semiconductor electrode materials commonly used in neural electrode design are more susceptible to shunting neural interface signals through dielectric defects, compared to highly doped silicon.

Conclusion: Strategic use of silicon and similar materials may increase neural electrode robustness against encapsulation failures.

Keywords: Accelerated aging; Dielectric encapsulation; Doped silicon; Electrochemical impedance; Electrode materials; Finite element analysis; Neural interface; Neuroprosthetics; Parylene C; Shunting; Utah array.