2.2. Impedance Testing Methods

The electrical characterization of the electrodes was divided into two separate test series. In Series I, the skin–electrode impedance, in the following referred to as three-electrode contact impedance, was measured on a human arm, similar to the setup used by Yun-Hsuan et al. (2014), while in Series II, the pressure-dependent impedance, in the following referred to as two-electrode contact impedance, was assessed on an agar dummy similar to the study by Beckmann et al. (2010) [39,40]. In both series, an electrical impedance spectroscope (PGSTAT 204 with FRA 32M module, Metrohm Autolab, Utrecht, Netherlands) was used in potentiostatic mode to measure the impedances of the respective system. A sinusoidal voltage (amplitude 0.01 V) was applied, and the frequency-dependent impedance was automatically calculated from the measurement data using a frequency scan from 1 MHz down to 0.1 Hz with 10 points per decade.

For the analysis, besides comparing the measured impedance values, equivalent circuits were calculated for the experimental systems. For the modelling, the integrated tool in the EIS software NOVA 2.4.1 (Metrohm Autolab, Utrecht, Netherlands) was used. As suggested by Zhou et al. (2015), to represent the skin–electrode system, a basic circuit consisting of a parallel resistor Rp and constant phase element CPE in series with another resistor Rs was chosen, see Figure 3 [31]. Within this, Rs represents the total resistance of electrode, wires, and the body/agar; Rp was used to model the charge transfer resistance; and the CPE represents the double-layer capacitance, according to following equation:

where Y₀ corresponds to the admittance of an ideal capacitance and N is an empirical constant that can be located between 0 to 1. For the circuit fitting, a maximum number of 300 iterations was chosen with a maximum of 50 iterations without improvement. Further, a maximum change of 0.001 in goodness of fit χ² (scaled) was used, and each data point was multiplied by a weight factor, i.e., the inverse of the square of the impedance modulus.

Equivalent circuit to represent the contact impedance of textile electrodes.

In Series I, the different electrode versions were characterized regarding their influence on the impedance behavior of the system when applied to a human forearm using a three-electrode configuration with the textile electrode as working electrode (WE) and conventional, self-adhesive Ag/AgCl electrodes (23 × 34 mm, Fiab SpA, Firenze, Italy) as counter electrode (CE) and reference electrode (RE). The analyzed system consisted of the impedances of the textile electrode, of the skin and body tissue, as well as of the skin–electrode interface, and the measured three-electrode contact impedance is the combination of those individual impedances [40]. The electrode positions were kept constant for all measurements performed in Series I with a distance of 10 cm between WE and CE and a distance of 1 cm between WE and RE, as shown in Figure 4. To reduce variations arising from individual differences in body impedance, all measurements were performed on one subject only (female, 25 years old).

Measurement setup on human forearm (Series I) with the textile electrode indicated in pink and the conventional electrodes indicated in yellow.

One testing cycle included three impedance measurements performed in a row without re-attaching the electrodes. Five replicates of this testing cycle were made per electrode sample and per electrode condition, i.e., dry or wetted with tap water (1 mL/20 cm²), where each replicate was performed on a different day. Tap water was chosen as electrolyte, because advantages were seen regarding the user convenience for future applications in terms of its easy accessibility as well as the possibility of applying it from the back side of the electrode, i.e., the outside of a garment when integrated into a wearable. Thus, in future applications, the electrode can be wetted and re-wetted without taking off the garment. Additionally, compared to when using other electrolytes, no washing of the electrode is required after use.

In Series II, the influence of pressure application to the electrode on the two-electrode contact impedance was investigated. To keep the contact force controlled and uniform, a water-based agar dummy (200 mL deionized water and 7.5 g of high gel-strength agar, Sigma Aldrich A9799, St. Louis, MO, USA) was chosen instead of a human subject. The impedance was measured in a two-electrode setup with the textile electrode as WE, placed on top of the dummy, and a copper plate placed below the dummy as CE/RE (see Figure 5a). Thus, the measured two-electrode contact impedance consisted of the impedances of WE, CE/RE, the dummy impedance, and the interface impedances of the dummy and the electrodes [39]. Potential reaction products occurring at the surface of the CE/RE were removed by sanding the copper plate after every four hours of testing. The dummy was placed in a custom-built testing rig, shown in Figure 5b, which has a 3D printed stamp to apply a controlled pressure to the agar-electrode setup. A new dummy was used for every day of testing, as the dummy is expected to dry out over a longer period of time, thereby changing its electrical and mechanical properties.

(a) Side-view of the dummy setup in Series II. (b) Testing rig for Series II.

To perform one testing cycle, the stamp gave a pre-force of 200 g to the testing setup, followed by three impedance measurements; then, the pressure was increased to the investigated pressure level (indicated by the force at 400, 600, and 1000 g), and again, three measurements were performed. Eight replicates of the testing cycle were performed per electrode sample and pressure level. Only electrodes with the approximate same size were evaluated in this experimental series to ensure a comparable pressure distribution for all tested electrodes. Therefore, the two bigger sizes of electrode construction E1 were excluded.