Evaluating and Developing a Reliable Reference Electrode for Choline Chloride Based Deep Eutectic Solvents

Ethaline preparation

Ag/AgCl RE preparation

[Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ RE preparation

Electrochemical test

Surface characterization

As shown in Fig. 1a, a typical Ag/AgCl RE was prepared via the deposition/stripping square wave method (50 cycles) and was used to study the behavior of [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ redox couple in ethaline. The dark gray area clearly displays the coating of AgCl film onto the silver wire (Fig. 1a). A cyclic voltammetry (CV) experiment in Fig. 1b shows the redox peaks of the [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ couple on a Pt disk electrode in ethaline using an as-prepared Ag/AgCl wire as reference electrode which was directly inserted into the ethaline electrolyte. Noticeably, over the 800 CV cycles (∼14.22 h), the peak position exhibits a significant shift (∼0.1 V), implying the reference point of potential measurement is shifting since the [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ redox couple is highly stable in ethaline, particularly in the oxygen free environment of the glovebox.^18,19 The color of the used Ag/AgCl RE turned white during the test, another indication of its instability as shown in Fig. 1b. Moreover, when checking the potential of this used Ag/AgCl RE with respect to a commercial Ag/AgCl (3 M NaCl) RE in aqueous 1 M NaCl electrolyte, a potential difference around −5 to −10 mV was found, as compared to the freshly prepared electrode with an expected +24.5 mV difference when accounting for the salt concentration difference (see Table SI). This further suggests the instability of AgCl film when operating in ethaline.

Zoom In Zoom Out Reset image size

In order to understand the degradation of the Ag/AgCl RE in ethaline, we developed a reference electrode by utilizing a soluble and stable redox couple in ethaline, i.e. [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻. It was expected that this reference electrode might be highly effective for ethaline system considering that (1) the reversible potential of [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ couple on Pt is a reasonable reference point within the potential window of ethaline, (2) the high viscosity of ethaline leads to the low diffusivity of [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻, thus minimizing diffusion of the ions across a porous frit, (3) the liquid junction potential drop should be negligible since the electrolyte/solvent in and outside the reference electrode are the same, (4) both the oxidized and reduced forms of the [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ couple are commercially available as compared with other redox couples like ferrocene/ferrocenium where the ferrocenium is unavailable, and (5) both the oxidized and reduced forms are soluble in ethaline (up to at least 40 mM at room temperature). Figures 1c and 1d show the sketch and image of the assembled FCN RE.

We utilized this FCN RE during two weeks of heavy usage (e.g., long-time electrolysis in ethaline, propaline and reline) to investigate its stability and the results are shown in Fig. 1e. We used the redox peak position of ferrocene—a typical internal standard—to investigate the possible potential shift of the FCN RE. Since the redox potential of ferrocene is independent on the electrolyte due to the fact that its redox site is screened from the influence of solvent by its sandwiched molecular structure,^20–22 we can determine if the reference potential is changing by comparing ferrocene's redox peak with respect to as-prepared and used FCN RE. We found that the measured redox potential of ferrocene was identical relative to either as-prepared or used FCN RE, confirming the long-term stability of the reference electrode. In the supplementary material we also estimate the diffusion of [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ across the glass frit and demonstrate that the loss of indicator ions from the internal solution of FCN RE is not expected to be a limiting factor of the reference electrode stability. Furthermore, the amount of [Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ entering the working electrode compartment is very minimal (sub-micromolar concentration) over a continuous five-day experiment (estimated in supplementary material). Thus, this reference electrode is recommended for future electrochemical studies in DES systems, especially those involving choline chloride.

Using the stable FCN RE as a benchmark, we investigated the degradation mechanism of the Ag/AgCl RE. A series of Ag/AgCl REs with different AgCl film thicknesses were prepared via the aforementioned deposition/stripping wave method as shown in Fig. S1a (available online at stacks.iop.org/JES/167/086509/mmedia). Based on the coated area and net deposition charge, we calculated the AgCl film thickness that is reported in Fig. S1b.

The open circuit potentials of these Ag/AgCl REs vs the stable FCN RE in ethaline were recorded to monitor the potential change, and the potential transients are shown in Fig. 2. All of these Ag/AgCl(m) REs show a similar degradation trend: an initial potential plateau at about −50 mV vs FCN RE (except for the Ag/AgCl(5) RE) that lasts for a short period of time(t_s, stable time), then the potential decreases and reaches a final plateau at a terminal potential (E_t). But noticeably, when the AgCl film was thicker, t_s increases from ∼2.8 h for Ag/AgCl(20) RE to ∼15.4 h for Ag/AgCl(100) RE. This degradation phenomenon was also observed for a Ag/AgCl wire taken from a commercial Ag/AgCl(3 M NaCl) RE as shown in Fig. S2, indicating the degradation in ethaline is not the result of the particular preparation method used here. As for the Ag/AgCl(5) RE, since there's no initial stable potential, it quickly degrades to a terminal plateau value. This might be the reason for some previous reports where the instability of Ag/AgCl RE was not noticed, which could be attributed to an overly-thin coating of AgCl film causing the potential to quickly drop to E_t, and thus no transient is observed.¹⁷

Zoom In Zoom Out Reset image size

The correlation between the stable time (t_s) and the film thickness was found to be linear as shown in Fig. 3a. This implies that a Ag/AgCl RE may be useful for short-time experiments if the AgCl coating is thick enough. On the other hand, the t_s—thickness correlation with stability also suggests that the reference electrode is continuously losing AgCl from the surface of the Ag wire, and as there were no obvious solid residues formed in ethaline, this suggests that AgCl is dissolving into ethaline. It has been found that the AgCl would readily dissolve in concentrated HCl solution forming AgCl_n¹⁻ⁿ complexes, as reported by Zelyanskii et al., and the solubility of AgCl increased about 29 times from 0.925 mmol l⁻¹ in 2 M HCl to 26.9 mmol l⁻¹ in 10 M HCl.²³ This behavior has also been confirmed in a non-aqueous solution by Skompska et al., where they found that AgCl would dissolve in acetonitrile containing high concentrations of chloride ion forming AgCl₂⁻ species similar to the aqueous system.²⁴ As for ethaline, since it's a high Cl⁻ concentration solvent, the AgCl species would react with Cl⁻ to form AgCl₂⁻ complex as well (AgCl + Cl⁻ → AgCl₂⁻) with a solubility around 0.2 mol kg⁻¹ as reported by Abbott et al.²⁵ The loss of AgCl from the Ag wire surface can also be confirmed by the XPS analysis as shown in Fig. S3. We analyzed the used and unused portions of a typical Ag/AgCl RE and estimated corresponding surface compositions. The unsubmerged portion of the coated wire shows nearly equal compositions of surface Ag and Cl (∼1:1) consistent with AgCl. However, for the submerged portion of the coated wire, the Cl content is almost zero and the oxidation state of Ag shifts back to the metallic state (Fig. S3c) indicating the surface is barely covered by AgCl, and the underlying Ag surface is exposed. Noticeably, the dissolved AgCl species would also influence the ethaline electrolyte as evidenced in Fig. 3b. When placing a fresh bare Ag wire into the used ethaline electrolyte containing the dissolved species, the potential difference vs FCN RE is dramatically different as compared with a fresh Ag wire placed in fresh ethaline. The potential of the fresh Ag wire placed in used electrolyte approaches the E_t of the tested Ag/AgCl RE in ethaline. Such a large potential difference in fresh and used ethaline suggests that the potential measured on the Ag wire in used ethaline (or E_t) is the potential of the (Ag/AgCl₂⁻: AgCl₂⁻ + e ↔ Ag + 2Cl⁻) couple with the AgCl₂⁻ dissolved in the ethaline. This also explains the different E_t values of Ag/AgCl(m) (m = 5, 20, 50, 100) REs seen in Fig. 2. Different starting amounts of AgCl on the electrode will ultimately result in different dissolved AgCl₂⁻ concentrations at the end of the experiment as estimated in Table SII. The different potential plateaus at the end of the tests reached by each Ag/AgCl RE tested (see Fig. 2) are in reasonable agreement with those expected based on a Nernst correction as shown in Table SII. These results suggest that the so-called Ag/AgCl RE used in many ChCl based DESs is actually the Ag/AgCl₂⁻ RE after the AgCl coating dissolves. Therefore, similar to the FCN RE, another possible reference electrode could be developed by sealing a fixed concentration (or even saturated)of dissolved AgCl₂⁻ in ethaline electrolyte into a glass tube with a silver wire, where the AgCl₂⁻ ethaline electrolyte can be obtained by simply dissolving AgCl in ethaline. This reference electrode is also expected to be stable since the bare Ag wire in used ethaline shows a constant potential in Fig. 3b, which will be studied in the future work.

Zoom In Zoom Out Reset image size

In this work, we found the widely used Ag/AgCl RE for ChCl based DES was unstable and showed significant potential shifts in long-time electrochemical tests. To address this problem, we developed a Pt∣[Fe(CN)₆]³⁻/[Fe(CN)₆]⁴⁻ RE and demonstrated its good stability in ethaline. By using FCN RE as the benchmark, we studied the degradation mechanism of Ag/AgCl RE in ethaline, where we found that the coated AgCl film on Ag/AgCl RE would dissolve in ethaline due to the high concentration of Cl⁻ anion and form soluble AgCl₂⁻ species. Although a thicker AgCl film would postpone the potential shift of the Ag/AgCl RE and create a stable period, thus making short-time experiments possible where the contamination of AgCl₂⁻ is insignificant, it's strongly suggested to use the FCN RE (or possibly a Ag/AgCl₂⁻ RE) instead of Ag/AgCl RE to obtain reliable and repeatable electrochemical results.

This work was supported as part of the Breakthrough Electrolytes for Energy Storage (BEES), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences under Award # DE-SC0019409.