Impedance Biosensor: Randles Circuit, Nyquist Plot and Binding Calibration

A screen-printed electrode under an electrolyte droplet: probes on the gold disc catch analyte and the coverage sets the charge-transfer resistance. Drag to orbit.

Switch to 2D
Controls

Presets

Parameters

Quality

Camera

Model checks

The science

Impedance as a sensor signal. In electrochemical impedance spectroscopy a small AC voltage is applied at many frequencies and the complex impedance Z(ω) is recorded. For a label-free biosensor a redox probe (ferri/ferrocyanide) is added; when the target binds the surface probes, it blocks electron transfer and the charge-transfer resistance Rct rises. The signal is the change of Rct, read off the diameter of the semicircle in the Nyquist plot.

Randles circuit. Z = Rs + (Rct + ZW) / (1 + jωCdl(Rct + ZW)), with the Warburg element ZW = σω-1/2(1 - j) for diffusion of the redox probe. At high frequency Z → Rs; without diffusion the low-frequency end is Rs + Rct and the semicircle has its apex at ω = 1/(RctCdl) where Z = Rs + Rct/2 - jRct/2. With diffusion the plot continues as a 45° line: -Z'' = (Z' - Rs - Rct) + 2σ2Cdl.

Binding. Probe-analyte binding is Langmuir: dθ/dt = konc(1 - θ) - koffθ, equilibrium θ = c/(c + Kd), time constant 1/(konc + koff), and Rct = Rct0 + ΔRmaxθ. The calibration curve is therefore hyperbolic with half-saturation at Kd; a concentration can be read back as c = Kdθ/(1 - θ).

Simplifications. Ideal Randles circuit (real electrodes need a constant-phase element instead of a capacitor), a linear dependence of Rct on coverage, a single binding site type, no non-specific adsorption and no drift (the biofouling and drift of implantable sensors is a different effect). Parameter values are typical for a screen-printed gold electrode, not measurements. Checks: the two frequency limits, the exact semicircle apex, impedance form against admittance form, the Warburg line, RK4 binding kinetics against the closed form, and the calibration inverse.

Loading model...
Loading Blender model...
-
-
Made with 🟠 Blender · Three.js