Neuromodulation by electrical excitation of the brain and peripheral nerve is of great interest for treating a host of established and emerging indications that include Parkinson’s disease, epilepsy, depression and blindness in the brain, or pain, epilepsy, hypertension, diabetes and rheumatoid arthritis in peripheral nerve. There is also interest in recording neural signals associated with these disease states to provide adaptive feedback that informs the delivery of the neuro-modulating waveforms to improve treatment efficacy. The electrodes that deliver the electrical excitation reside in a complex environment characterized by biomolecules that absorb onto the electrode surface, pH buffering involving carbonate, phosphate and biomolecules, the formation of tissue encapsulation around the electrode, and a restricted diffusional environment that limits the transport of species to the electrode. In this environment, a stimulation electrode must deliver charge to tissue via interfacial reactions that are not harmful to the electrode or to tissue. In most clinical neuromodulation devices, the charge is delivered with a pulsatile controlled current or controlled voltage waveform in which cathodal and anodal phases are balanced to provide an overall zero net charge. The charge/phase and charge density at an electrode vary over a wide range depending on whether the electrode is large and planar, with a geometric surface area (GSA) greater than approximately 0.01 cm², or whether the electrode is designed to penetrate tissue and has a comparatively small GSA of <10,000 μm². Large GSA electrodes (macroelectrodes) usually deliver 0.1-5 μC/ph at charge densities that vary from about 5 μC/cm² to 100 μC/cm². Microelectrodes, on the other hand, deliver only about 1-10 nC/phase, but at much higher charge and current densities that can exceed 1 mC/cm² and 1 A/cm², respectively. The challenge with neural stimulation electrodes is to deliver charge at these very high rates (high current densities) without inducing electrode corrosion, biomolecule reduction or oxidation, or water electrolysis. Electrodes and charge-injection waveforms capable of delivering charge without inducing obvious electrode and tissue damage are often considered “reversible” although it has to be recognized that some degree irreversibility in the charge-injection reactions is inevitable at these current densities. In practice, the ability of an electrode to deliver charge reversibly is determined by measuring the electrode potential during a charge-injection pulse. If the potential does not exceed the water electrolysis limits for that electrode, the charge-injection is considered reversible. This strategy ignores potentially harmful reactions that might occur within the water electrolysis limits, including oxygen reduction or electrode dissolution, and also ignores the non-uniform distribution of potential across the surface of an electrode. Experimentally, it is also difficult to determine the potential of a stimulation electrode in the presence of large overpotentials and the iR-drop in the electrolyte immediately surrounding the electrode. Nonetheless, this strategy has been used for about four decades to evaluate and compare stimulation electrodes and to assess the electrochemical safety of stimulation waveforms.

In this presentation, the charge-injection processes occurring at different types of stimulation electrodes are described and the limitations to the charge-injection capacities of electrodes identified. Electrodes and electrode coatings that operate predominately by capacitive or faradaic reactions are included and the advantages and disadvantages of each discussed. Most electrodes in clinical use are macroelectrodes employing platinum (Pt or PtIr) or porous titanium nitride (TiN) for charge-injection, although there has been recent interest in sputtered iridium oxide for microelectrodes. Efforts have been made to increase the charge-injection capacity and reduce the impedance of Pt and TiN electrodes by increasing surface roughness and separately increasing perimeter-to-area ratio. These efforts have met with only modest success due to the negative impact of pore resistance, secondary and tertiary current distributions, and the impact of tissue encapsulation. The role of these limitations on electrode performance is discussed, including as assessment of the notable differences between charge-injection properties measured in buffered saline electrolytes and in animals.

Developments in neuromodulation treatments that block neural activity are of increasing interest and are being implemented in clinical devices for pain management. Electrical waveforms for neural blocking are significantly different from conventional pulsatile waveforms and usually employ continuously applied high frequency sinusoidal currents. These waveforms introduce new challenges and opportunities for characterizing electrodes and efficiently delivering charge to tissue. Recently, the use of very small electrodes to evade the body’s immune response has been investigated for recording and stimulation in the brain. These electrodes are sufficiently small, with at least one dimension <10 μm, to exhibit ultramicroelectrode behavior. Electrode processes associated with both these emerging waveforms and devices are discussed.