Injectable Nanoelectronic Sensors for Brain-Machine Interfacing

Abstract

Facile interfacing of electrogenic tissues to electronics has been an elusive goal in neuroscience and medicine. Microfabricated arrays of electrodes have emerged as an important tool for studying the brain, but problems related to invasiveness, mechanical mismatch, and electrode impedance have limited their success in long-term studies and in probing anatomy with micro- and nano-scale resolution. This dissertation seeks to address these limitations by building upon recent advances in nanotechnology, microelectronics, and materials science to develop injectable nanoelectronics sensors for brain-machine interfacing. First, I present a plug-and-play electrical input/output interfacing scheme for syringe-injectable electronics. Injectable electronics present special challenges for input/output interfacing because the electronics cannot be tethered as they pass through a needle. The scheme solves this problem by incorporating a self-aligning input/output region that can be blindly inserted into a commercially available connector. Injection and interfacing experiments demonstrate rapid and scalable electrical connection to arrays of electrodes and field-effect transistors with a contact resistance of only 3 Ω. Second, I discuss efforts to design and fabricate syringe-injectable electronics containing field-effect transistors with silicon channels synthesized in a bottom-up vapor-liquid-solid process. Optical and electrical characterization confirms successful fabrication of injectable arrays of p-type nanowire field-effect transistors, notably with suspended channel regions up to 30-μm long. Third, I demonstrate an entirely top-down fabrication process for injectable, suspended field-effect transistors based on silicon-on-insulator wafers. Significantly, the top-down fabrication process is compatible with commercial wafer-scale production, allows for implementation of transistors with junctions, and is an important step towards scaling injectable electronics to higher channel counts and smaller feature sizes. Finally, I apply the plug-and-play interfacing scheme and injectable field-effect transistor probes to live brain recordings in mouse models. In vivo characterization demonstrates successful delivery and transconductance measurements of nanowire field-effect transistors inside the brain. Recording experiments highlight the unique spectroscopic features of signals from field-effect transistors compared to electrodes. Together, the work presented here advances the state-of-the-art in brain-machine interfacing and presents exciting new opportunities for studying the brain such as recording from sub-cellular anatomy and monitoring low-frequency brain activity.