Reading the Interface: Impedance, the Double Layer, and Charge Transfer

The electrical double layer

When a metal meets an electrolyte, ions and solvent reorganise into the electrical double layer — a nanometre-thin capacitor at the interface (Helmholtz, refined by Gouy–Chapman–Stern). To first order it behaves as a plate capacitor whose spacing is a single solvation layer, which is why the capacitance per unit area is enormous:

C_{\mathrm{dl}} = \frac{\varepsilon_r\,\varepsilon_0\,A}{d_{\mathrm{H}}}

Idealised double-layer capacitance: area A over a Helmholtz spacing d_H of order one solvation layer. Real interfaces depart from this ideal — see the constant phase element below.

There is also a resting half-cell potential across the interface. Stimulation must swing around it while staying inside the water window; drift of this potential is one early warning of an electrode going bad.

The Randles circuit

The workhorse model of the interface is the Randles equivalent circuit: a solution (spreading) resistance R_s in series with a parallel combination of the double-layer capacitance and a charge-transfer resistance R_{ct} that carries the faradaic (redox) current, often with a Warburg diffusion term at low frequency.

Z(\omega) = R_s + \left(\frac{1}{R_{ct}} + Q\,(j\omega)^{n}\right)^{-1}

Randles impedance with the double layer modelled as a constant phase element (Q, n). At high frequency Z → R_s; at low frequency Z → R_s + R_ct.

The two extremes matter. For recording you want a low impedance magnitude across the spike band (low noise). For stimulation you want a large reservoir of reversible charge, i.e. a big effective capacitance plus reversible faradaic capacity — but no irreversible leakage.

Real interfaces aren't ideal: the constant phase element

Fit a real electrode and it never gives a clean capacitive semicircle. Instead the phase freezes somewhere between 0° and −90° and stays there over decades of frequency. This is modelled by the constant phase element (CPE), whose exponent n encodes how far the surface departs from an ideal capacitor:

Z_{\mathrm{CPE}}(\omega) = \frac{1}{Q\,(j\omega)^{n}},\qquad 0 < n \le 1

The constant phase element. n = 1 is an ideal capacitor; n < 1 reflects surface roughness, porosity, and heterogeneous current distribution.

This is the model people actually fit in practice (Randles with a CPE). The exponent n and the coefficient Q are, in effect, a fingerprint of surface morphology: nanostructuring a surface raises Q (more accessible area) and typically lowers n (more distributed, non-ideal current paths). We will exploit exactly that in the next guide.

Faradaic vs capacitive charge transfer

The faradaic vs capacitive distinction is the heart of safe stimulation. Capacitive transfer just charges and discharges the double layer — perfectly reversible but limited in capacity. Faradaic transfer moves electrons through a redox reaction; iridium oxide exploits a reversible Ir³⁺/Ir⁴⁺ pseudocapacitance to inject far more charge, while PEDOT blends capacitive and reversible faradaic contributions.

Q_{\text{ph}} = \int_0^{t_p} I(t)\,dt, \qquad D = \frac{Q_{\text{ph}}}{A_{\text{geo}}}

Charge per phase is the integral of stimulus current over the pulse; dividing by geometric area gives the charge density that must stay under the Shannon line.

Measuring it — and why the numbers don't compare

The primary tool is electrochemical impedance spectroscopy (EIS): sweep frequency, fit the circuit, and report impedance at 1 kHz, charge storage capacity (CSC), and charge injection capacity (CIC). Cyclic voltammetry gives CSC; voltage-transient measurements under real pulses give CIC.