Fast-scan cyclic voltammetry (FSCV) using carbon fiber electrodes has proven to be a reliable method for in vivo detection of phasic dopamine signal. The FSCV measurement produces two overlapping current signals: (i) a large non-informative background current due to the charge and the discharge of the double layer capacitance, and (ii) a small faradaic signal due to the redox reactions of dopamine molecules. In FSCV, applying a fast voltage ramp increases the sensitivity and the temporal resolution. However, the non-informative background current also becomes larger at faster ramp rates, thereby making the detection of the small faradaic signal more difficult. Furthermore, the environmental noise is a limiting factor for the detection of dopamine at the physiological levels of interest, particularly in animals with large brains. Finally, the large cylindrical shape of carbon fiber electrodes complicates the development of multielectrode arrays, necessary for large-scale measurements of dopamine concentration. To address these technical challenges, we combine advances in nanofabrication with silicon chip manufacturing and create a heterogeneous integrated CMOS-graphene sensing platform that would enable accurate measurements of dopamine with high spatiotemporal resolution.
To accurately measure the small faradaic current, we developed an integrated readout circuit using a standard 65nm process. Leveraging the high specificity of FSCV, the readout chip is configured to capture the signal of interest at about the redox potentials. Also, a fixed low-noise current continuously subtracts a large portion of the non-informative background current near the redox potentials, while maintaining the overall shape of the signal. Our readout approach, therefore, makes the CMOS chip compact and power-efficient.
We replaced carbon fiber with a few-layer graphene film because the thin planar structure of graphene facilitates the fabrication of a multielectrode array. Furthermore, significant advances in large-area synthesis of graphene using chemical vapor deposition offer a manufacturable solution for large-scale deployment of this integrated sensing technology. These features enabled the direct integration of graphene sensor arrays on top of the CMOS readout chip. We validated the performance of our integrated sensing platform through in vitro measurements.
In conclusion, we introduced and implemented a heterogeneous integrated CMOS-graphene sensor for measuring dopamine concentration using fast-scan cyclic voltammetry. Our core sensing technology offers a new pathway for developing the next generation of neural probes that can accurately monitor dopamine signaling with high spatiotemporal resolution in the brain.