Implantable neuroprosthetic devices are widely used to restore and replace lost bodily functions in a number of neurological impairments including neuromotor deficit, hearing loss, chronic pain, and epilepsy (Hambrecht, 1979; Merrill et al., 2005). In 2015, the total market size for various neural stimulation devices that target spinal cord, cochlear, cerebral cortex, and other peripheral nerves (e.g., sacral, vagus nerve), exceeded $4.9 billion with the annual growth rate of 17% (Pikov, 2015). This is partly driven by advances made in basic neuroscience to better understand underlying neural mechanisms linked to various ailments and aging population that demands more advanced therapeutic options for chronic neurological disorders.

With advances in neurostimulation technologies, the demand for more precise targeting of neural substrate has fueled the development of higher density electrode arrays to improve the resolution of stimulation outcomes while minimizing unwanted side effects (Patil and Thakor, 2016; Rossi et al., 2016). The race to increase the number of input/output channels with smaller electrodes have resulted in a number of different microelectrode arrays (MEA) for various stimulation and recording applications (Jimbo, 1992; Jimbo et al., 2003; Luan et al., 2017). However, there has not been thorough investigation on subsequent consequences of delivering the same electrical current through smaller electrodes. This is especially important because chronic overstimulation is known to cause nerve damage via a number of different mechanisms.

Prevailing hypotheses on neurostimulation-induced tissue damage include hyperactivity of many neurons firing simultaneously that can lead to metabolic overload; and generation of toxic electrochemical byproducts during stimulation including dissolved metallic ions from the electrode, reactive oxygen species, water electrolysis byproducts (Merrill et al., 2011). In most neurostimulation systems, Platinum (Pt) is used as the material of choice due to its relatively high charge injection capacity via redox reactions, chemical resistance, and biocompatibility (Cogan, 2008). However, dissolution of Pt electrodes during stimulation is also known to lead to corrosion of Pt electrodes which is undesirable (Kovach et al., 2016; William F. Agnew et al., 1977).

With smaller electrodes, the charge injection limit plays an important role in minimizing generation of toxic electrochemical byproduct during neurostimulation. The increased perimeter-to-surface area (PSA) in microscale electrodes can lead to significant non-uniformity in current density, which can alter charge injection capability and generation of noxious electrochemical byproducts. Although the impact of pulse parameters, electrode surface finish, and stimulation environment on charge transfer capacity of stimulating electrodes have previously been examined in Pt MEAs, very little work has been done to investigate the effects of electrode geometry and the inter-electrode distances in charge transfer capability of microscale stimulating electrodes (Cogan et al., 2014; Kumsa et al., 2016).

Here we examined the role of electrode geometry (PSA) using custom microfabricated MEAs. The inter-electrode distances were varied to examine the impact of diffusion limit of electrochemical reactions. Four types of electrodes were created: circular, fractal, serpentine I, and serpentine II (Fig. 1, Table 1). Electrical contact to the electrodes was made using a custom 3D printed platform. Each electrode was characterized using cyclic voltammetry and electrochemical impedance spectroscopy. Cyclic voltammetry was measured in phosphate buffered solution (pH: 7.4) between potential limits from -0.65 to 0.85 V (Ag/AgCl) with sweep rate of 50 mV s^-1(Fig. 2a). The cathodal and the total charge storage capacity of each electrode were calculated from the time integral of the current in cyclic voltammogram, which positively correlated with increasing PSA (Fig. 2b). The electrode impedance was measured using electrochemical impedance spectroscopy in 10 Hz to 100 kHz range, which showed elevated impedance value for higher PSA (Fig. 2c, 2d).

Our preliminary results suggest that increasing PSA may alter the charge injection capability of electrode due to highly non-uniform current density facilitated by the edge effects. However, the increased non-uniformity and current density may also adversely affect the electrode surface integrity and facilitate the release of noxious byproducts. Our future work will focus on quantifying the dissolution rate of the platinum and other materials from the different shaped electrodes with different inter-electrode spacing at various electrode potentials. Moreover, we will compare voltage transient response from these electrodes in a biphasic stimulation pulse experiments to correlate our results.