Solvent Evaporation Rate as a Tool for Tuning the Performance of a Solid Polymer Electrolyte Gas Sensor

Abstract

Solid polymer electrolytes show their potential to partially replace conventional electrolytes in electrochemical devices. The solvent evaporation rate represents one of many options for modifying the electrode-electrolyte interface by affecting the structural and electrical properties of polymer electrolytes used in batteries. This paper evaluates the effect of solvent evaporation during the preparation of solid polymer electrolytes on the overall performance of an amperometric gas sensor. A mixture of the polymer host, solvent and an ionic liquid was thermally treated under different evaporation rates to prepare four polymer electrolytes. A carbon nanotube-based working electrode deposited by spray-coating the polymer electrolyte layer allowed the preparation of the electrode-electrolyte interface with different morphologies, which were then investigated using scanning electron microscopy and Raman spectroscopy. All prepared sensors were exposed to nitrogen dioxide concentration of 0-10 ppm, and the current responses and their fluctuations were analyzed. Electrochemical impedance spectroscopy was used to describe the sensor with an equivalent electric circuit. Experimental results showed that a higher solvent evaporation rate leads to lower sensor sensitivity, affects associated parameters (such as the detection/quantification limit) and increases the limit of the maximum current flowing through the sensor, while the other properties (hysteresis, repeatability, response time, recovery time) change insignificantly.

Keywords: gas sensor; ionic liquid; noise spectroscopy; solid polymer electrolyte.

Figure 1

Topology of an amperometric NO₂ sensor (a): working electrode contact (1a), carbon working electrode (1b), pseudo-reference electrode contact (2a), pseudo-reference electrode (2b), counter electrode contact (3a), counter electrode (3b), substrate (4), solid polymer electrolyte (SPE) (5), white square indicates the location of Raman spectroscopy measurements of the WE electrode. (b) Sensor with MWCNT working electrode and SPE layer treatment 120 °C for 90 s.

Figure 2

SEM images (back-scattered electron imaging and secondary electron imaging) of the MWCNT WE–SPE interface with the [EMIM][TFSI] ionic liquid prepared for 90 s at 80 °C (a,b), 90 s at 120 °C (c,d), 210 s at 120 °C (e,f) and 600 s at 150 °C (g,h) with view-field 100 μm (a,c,e,g). The boundary lines of WE–SPE are estimated and plotted as dashed lines. (b,d,f,h) show details of WE electrodes on SPE layers of corresponding crystallinities with view-field 10 μm.

Figure 3

(a) Raman spectra of the MWCNT working electrode and solid polymer electrolyte (SPE treated at 80 °C for 90 s) and their mutual interface corresponding to marked positions from #1 to #5 at (b) confocal microscopy image of their acquisition locations.

Figure 4

(a) Equivalent electric circuit of the NO₂ sensor, (b) typical impedance complex plane (Nyquist) plot for a system with two well-separated time constants.

Figure 5

Nyquist plots for the NO₂ sensor (SPE treated at 80 °C for 90 s) with MWCNT-based working electrode for different NO₂ concentrations (operation conditions: 298 K, 40%RH, 1013.25 hPa, analyte flow rate 1 L/min, V_BIAS= −0.5 V): (a) Nyquist plot for the whole tested frequency range: 10 kHz–5 mHz, (b) Detail of the Nyquist plot for higher frequencies.

Figure 6

Nyquist plot for NO₂ sensors with different SPE types for 0 ppm NO₂.

Figure 7

Effect on sensor properties of solvent evaporation rate from solid polymer electrolyte during preparation: (a) sensor sensitivity per unit WE area as a function of electrolyte-spherulite diameter, (b) sensor sensitivity per unit WE area as a function of electrolyte conductivity, (c) signal-to-noise ratio as a function of electrolyte conductivity, (d) limit-of-detection as a function of electrolyte-spherulite diameter.

Figure 8

(a) Saturation current of sensor response as a function of SPE conductivity. (b) Current response as a function of NO₂ concentration for three sensors prepared with various crystallinities (i.e., of various conductivities). The measured points were fitted by a function derived from the Langmuir adsorption isotherm.

Figure 9

Effect of carbon working electrode and polymer electrolyte with various microstructures on current fluctuations of particular sensor: (a) Spectral density of current fluctuations for all measured NO₂ concentrations (sensor with SPE prepared at 120 °C for 90 s), (b) ratio of the spectral density of current fluctuations at 1 Hz to the square of the direct current as a function of electrolyte conductivity, (c) ratio of spectral density of current fluctuations at 1 Hz to the square of the direct current as a function of electrolyte—spherulite diameter.