Large-Scale Capacitive Neural Interface

Microelectrode arrays (MEAs) allow simultaneous neural recording from multiple electrodes, and the number of concurrent measurement channels is one of the most critical considerations for MEAs in studying complex neuronal networks.

This work proposes a neural interface that senses the electrical double layer (EDL) capacitance as a function of the ion concentration produced by neurons firing action potentials (AP). Unlike conventional microelectrode arrays (MEAs) detecting voltage, capacitance sensing allows access to multiple recording sites with a single wire using code-division multiplexing (CDM), thereby significantly reducing the number of required interconnects. In this work, we implemented 32 drivers and 32 analog front-end circuits (AFEs) to realize 1,024 channel concurrent neural recordings while using a total of 64 interconnects and improving area efficiency for large-scale integration. This work achieves 9.7μW power/ch and 0.005mm2 area/ch efficiency with the highest electrode density of 10,000mm-2, and the fewest interconnects to the authors’ best knowledge.