Time Evolution of the Skin-Electrode Interface Impedance under Different Skin Treatments

Abstract

A low and stable impedance at the skin-electrode interface is key to high-fidelity acquisition of biosignals, both acutely and in the long term. However, recording quality is highly variable due to the complex nature of human skin. Here, we present an experimental and modeling framework to investigate the interfacial impedance behavior, and describe how skin interventions affect its stability over time. To illustrate this approach, we report experimental measurements on the skin-electrode impedance using pre-gelled, clinical-grade electrodes in healthy human subjects recorded over 24 h following four skin treatments: (i) mechanical abrasion, (ii) chemical exfoliation, (iii) microporation, and (iv) no treatment. In the immediate post-treatment period, mechanical abrasion yields the lowest initial impedance, whereas the other treatments provide modest improvement compared to untreated skin. After 24 h, however, the impedance becomes more uniform across all groups (<20 kΩ at 10 Hz). The impedance data are fitted with an equivalent circuit model of the complete skin-electrode interface, clearly identifying skin-level versus electrode-level contributions to the overall impedance. Using this model, we systematically investigate how time and treatment affect the impedance response, and show that removal of the superficial epidermal layers is essential to achieving a low, long-term stable interface impedance.

Keywords: equivalent circuit model; skin impedance; skin treatment; skin–electrode interface; wearable sensors.

Figure 1

Electrodes and skin treatments used in the present study. (a) A cartoon schematic showing application of the four skin treatments to the upper arm. (b) Electrodes used in the study: Cleartrace^TM 1700-O30 ECG pre-gelled Ag/AgCl disk electrodes (geometric surface area = 5.1 cm², ConMed Co). (c–f) Skin treatment agents: (c) ‘No Treatment’ control (alcohol prep pad for pre-cleaning the skin); (d) 3M’s TracePrep^TM abrasive tape; (e) Neutrogena^® Acne Face Wash (with 2% salicylic acid); (f) AdminPatch^® 0900 microneedle array device.

Figure 2

(a) Schematic of the equivalent circuit model used for fitting the skin–electrode impedance data. At the electrode level (top three elements): R_ct is the charge-transfer resistance, C_dl is the double-layer capacitance, and R_gel is the gelled electrolyte resistance. At the skin level (bottom three elements), R_epi and C_epi are the epidermal resistance and capacitance, respectively, and R_sub is the resistance of the dermis and subcutaneous fat layers. (b,c) Representative Bode plots of the impedance (b) modulus and (c) phase for one subject at the initial timepoint, for all skin treatments. Points represent measured experimental data, while solid black lines denote the fitted curves. (d–f) 10 Hz impedance values for all subjects, for all skin treatments, at the (d) initial (t₀; ANOVA F(14,3) = 68.02, p < 1.5 × 10⁻¹⁵), (e) middle (t_mid~8 h; ANOVA F(10,3) = 19.31, p < 5.0 × 10⁻⁴), and (f) final (t_f~24 h; ANOVA F(10,3) = 19.17, p < 7.5 × 10⁻⁷) timepoints. Significance levels: (**) denotes p < 0.01, (***) denotes p < 0.001.

Figure 3

Area-normalized values of the fitted model parameters at t₀. (a–c) Skin-level results at t₀, for each treatment, including values of the (a) subcutaneous resistance, (b) epidermal resistance, and (c) epidermal capacitance. (d–f) Electrode-level results at t₀, for each treatment, including values of the (d) gelled electrolyte resistance, (e) charge-transfer resistance, and (f) double-layer capacitance. Significance levels: (**) denotes p < 0.0167 and (***) denotes p < 0.00167.

Figure 4

Time evolution of the fitted, area-normalized circuit model parameters over 24 h. (a–c) Skin-level results for each treatment type, showing time evolution of the values of the (a) subcutaneous resistance, (b) epidermal resistance, and (c) epidermal capacitance. (d–f) Electrode-level results for each treatment type, showing time evolution of the values of the (d) gelled electrolyte resistance, (e) charge-transfer resistance, and (f) double-layer capacitance. Bars represent means, errorbars are standard deviations (N = 14); t₀—initial timepoint, t_mid—study midpoint, 8 h. after skin treatment application, t_f—final timepoint, 24 h. after skin treatment application. Significance levels: (*) denotes p < 0.05, (***) denotes p < 0.01 (values are reported Table 2).