(I-348) A Measurement Model Approach to Electrochemical Impedance Spectroscopy Analysis Across Variable Ultramicroelectrode Sizes

Introduction::
One of the present challenges to effective chronic neurostimulation is the body’s inflammatory immune response. This challenge is particularly acute for microelectrode stimulation. The foreign body response results in device encapsulation by immune cells, producing a glial scar formation which increases the impedance surrounding the devices. To mitigate these effects, ultramicroelectrodes have been designed with the intention of being comparable to or smaller than neuron somas in order to improve biocompatibility and therefore improve long-term recording stability through reducing the immune response. In the field of neurostimulation, the magnitude of the impedance is often measured and reported at 1 kHz over time. However, electrochemical impedance spectroscopy (EIS) can offer more insight into changes in the electrode and tissue. In this study we evaluate EIS of various electrode sizes ranging from microelectrode to ultramicroelectrode. Further, we conduct detailed analysis and modeling of the data. The overarching goal of the work is to develop multichannel ultramicroelectrode arrays for suitable chronic neurostimulation.




Materials and Methods::
Ultramicroelectrode devices were custom designed with 32 electrodes ranging in size from 5µm to 50µm diameters. To improve charge storage capacity, electrodes were coated with a sputtered iridium oxide film (SIROF) and were characterized using EIS both in vitro and in vivo. For in vitro experiments, we used a three-electrode setup consisting of reference, counter, and working electrodes immersed in saline solution. The frequency sweeps ranged from 10 kHz to 0.1 Hz using a potentiostat. For in vivo experiments, a two-electrode set up was used where the working electrode was implanted into the somatosensory cortex of anesthetized Sprague-Dawley rats and the counter and reference electrodes were shorted together. During the procedure, the anesthetized rats were stimulated across the various electrode diameters and EIS was performed before, during, and after surgical implantation to assess the effect of stimulation and implantation on the device performance. Using the Measurement Model Program created by the Orazem group and the University of Florida [1], we computed the error structure of our EIS measurements and developed a custom interpretation model with fitting parameters representing the device-tissue interface.




Results, Conclusions, and Discussions::
EIS data from our first electrode design iteration showed that electrode size and quality had reproducible effects on the electrode-tissue interface. The Measurement Model Program was used to assess the error structure and the data was regressed against a custom model to fit impedance data. Since the program satisfies the Kramers-Kronig relations, unsuccessful fitting allowed us to identify which portions of the measurements were inconsistent either due to non-stationary phenomena or instrument artifacts. Additionally, the Measurement Model provides an estimate of the frequency above which the ohmic impedance associated with the electrode geometry affects the measurement. This computational approach to EIS analysis allows us to fit the data to an interpretation model with parameters representing the electrode and the electrolyte/tissue properties. We then can normalize according to electrode site size to investigate the effects of geometry on the measurements. In conclusion, miniaturizing stimulation electrodes has the potential to enhance their biocompatibility, leading to improved long-term recording stability. EIS can serve as a valuable technique to evaluate the functionality of ultramicroelectrodes and to investigate their tissue response in vivo, but careful analysis and interpretation of EIS data are essential to ensure accurate and reliable results.




Acknowledgements (Optional): :

References (Optional): :
[1] W. Watson and M. E. Orazem, “EIS: Measurement Model Program.” ECSarXiv, Sep. 16, 2020. doi: 10.1149/osf.io/kze9x.