T ESTING OF D IFFERENT E LECTRODE R ADII , E LECTRODE S ETUPS AND M EASURING

To find an optimised measurement setup for determination of diffusion coefficients in ILs and their binary mixtures, Pt disk UMEs with four different electrode radii, as described in Chap. 3.2.1.2, were tested in a solution of 0.00604 mol L^-1 ferrocene (ABCR, purity of 100%) and 0.199 mol L^-1 tetraethylammonium tetrafluoroborate (TEABF4, Merck, selectipur^®) in acetonitrile (AN, Merck, selectipur^®). The results obtained from the two different measurement methods explained in Chap. 3.2.2, steady-state cyclic voltammetry

and chronoamperometry, were compared as well as the application of three different measurement setups; two three electrode measurement setups, with an Ag/Ag⁺-cryptate reference electrode and a Pt wire as pseudo-reference electrode respectively, and a two electrode measurement setup.

Reference electrodes are necessary to control the applied potential during voltammetric and chronoamperometric measurements for which they have to fulfil several requirements [126]. To inhibit appearance of liquid junction potentials and to prevent contamination of the measuring solution with traces of solvent or salt, the solution within the reference electrode has to be as similar as possible to the measuring solution. That is why an Ag/Ag⁺-cryptate reference electrode [64,115] was used for measurements in organic electrolytes. Due to difficult and time consuming handling of an iodine/iodide reference electrode for measurements in ILs [127-130], and the above mentioned problem of contamination of the investigated ILs with traces of organic solvent or water by application of the Ag/Ag⁺-cryptate or a common aqueous reference electrode, the application of a Pt wire as pseudo-reference electrode was examined as well as the application of a two electrode measurement setup. Utilisation of a two electrode measurement setup and a Pt wire pseudo-reference electrode respectively are common and adequate procedures for steady state measurements at UMEs in ILs [131,132] due to the almost negligible IR-drop (see Chap. 3.2.2 or Refs. [89,114,117]).

To exemplify the relation between limiting current and electrode radius, four steady-state cyclic voltammograms recorded at Pt disk UMEs with four different electrode radii are shown in Figure 3-9.

0.00 0.25 0.50 0.75 1.00 1.25 0

5 10 15 20 25 30

I / nA

E / V vs. Ag/Ag⁺-cryptate

Figure 3-9: Comparison of steady-state cyclic voltammograms recorded at Pt disk UMEs with four different radii in a solution of 0.00604 mol L^-1 ferrocene and 0.199 mol L^-1 TEABF4 in AN using an Ag/Ag⁺-cryptate reference electrode; r0: (▬) 0.3 µm, (▬) 0.5 µm, (▬) 1 µm, (▬) 5 µm.

The diffusion coefficients of ferrocene in AN show, within the measurement accuracy, a very good agreement for the four different electrode radii and with the values from literature (Table 3-1). Reproducibility of diffusion coefficients within one measurement series was typically over 98% for all electrodes, but for several consecutive measurement series, including polishing after each series, it stayed only for the 5 µm electrodes at a comparably high value (> 97%).

Table 3-1: Diffusion coefficients of ferrocene in a solution of 0.00604 mol L^-1 ferrocene and

0.199 mol L^-1 TEABF4 in AN determined at electrodes with different radii and compared with values from literature.

r0 ·10⁴ [cm] D ·10⁵ [cm² s^-1] Dlit ·10⁷ [cm² s^-1]

0.3 2.06 0.5 2.34

1 2.40

5^a 2.31 5^b 2.45 5^c 2.44

2.4 [124]

2.17 [120]

a Two electrode setup.

b Three electrode setup with a Pt wire as pseudo-reference electrode.

c Three electrode setup with an Ag/Ag⁺-cryptate reference electrode.

As mentioned above, steady-state cyclic voltammograms were recorded with a three electrode measurement setup with an Ag/Ag⁺-cryptate reference electrode or a Pt wire as pseudo-reference electrode and with a two electrode measurement setup respectively. The determined diffusion coefficients showed, within the measurement accuracy, a very good agreement for the three different measurement setups (Table 3-1) but their reproducibility and the reproducibility of the shapes of the recorded cyclic voltammograms were best for the three electrode measurement setup with an Ag/Ag⁺-cryptate reference electrode, followed by the three electrode measurement setup with a Pt wire as pseudo-reference electrode.

Chronoamperometric measurements were also performed with electrodes of different radii and with different measuring setups. But here the main focus was set on application of three different methods for evaluating the recorded current-time curves. Method one is similar to evaluation of steady-state cyclic voltammograms. After a sufficiently long time, the limiting current of the chronoamperometric measurements shown in Figure 3-10, is assumed to be strictly faradaic and diffusion controlled and therefore Eq. (3.20) is valid.

0 15 30 45 60 75 90 0

5 10 15 20 25 30 35

I / nA

t / s

Figure 3-10: Current-time curves recorded during chronoamperometric measurements in a solution of 0.00604 mol L^-1 ferrocene and 0.199 mol L^-1 TEABF4 in AN at a 0.5 µm (─) and a 5 µm (─) Pt disk UME using an Ag/Ag⁺-cryptate reference electrode; (○) and (○) corresponding fits of the current according Eq. (3.18).

A second approach is also shown in Figure 3-10, where diffusion coefficients were determined by non-linear-regression of the data according to Eq. (3.18). The third method is shown in Figure 3-11, where according to Eq. (3.17) the diffusion coefficient D can be determined from the y-axis intercept of a plot of i vs. t ^-1/2.

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0

5 10 15 20 25 30 35

I / nA

t^-1/2 / s^-1/2

Figure 3-11: Plot of the current recorded during chronoamperometric measurements in a solution of 0.00604 mol L^-1 ferrocene and 0.199 mol L^-1 TEABF4 in AN at a 0.5 µm (─) and a 5 µm (─) Pt disk UME using an Ag/Ag⁺-cryptate reference electrode vs. t ^-1/2; (○) and (○) corresponding fits of the current according Eq. (3.17) vs. t ^-1/2.

In general, the diffusion coefficients of ferrocene in acetonitrile determined with chrono-amperometry showed, within the measurement accuracy, a very good agreement with the diffusion coefficients determined with steady-state cyclic voltammetry (typically > 95%) and therefore with the values from literature too (Table 3-2). The results of three different evaluation methods showed also very good agreement (> 99%) among each other. Similar to steady-state cyclic voltammetry, reproducibility of the limiting currents was best for the three electrode measurement setup with an Ag/Ag⁺-cryptate reference electrode and a 5 µm electrode as working electrode.

Table 3-2: Diffusion coefficients of ferrocene in a solution of 0.00604 mol L^-1 ferrocene and 0.199 mol L^-1 TEABF4 in AN determined with chronoamperometry at electrodes with 5 μm nominal radius and compared with values from literature.

Reference

a Evaluation according Shoup and Szabo (Eq. (3.18)).

b Evaluation according Baur and Wightman (Eq. (3.17)).

c Assumption of a time independent diffusion controlled current and evaluation according Eq. (3.20).

To sum it up, the best results for both, steady-state cyclic voltammetry and chrono-amperometry were achieved with an Ag/Ag⁺-cryptate reference electrode and a 5 µm electrode as working electrode. As mentioned above, application of a reference electrode in ILs causes some problems, hence for further measurements a Pt wire was used as pseudo-reference electrode instead the Ag/Ag⁺-cryptate reference electrode. Generally, chrono-amperometry is a faster method than steady-state cyclic voltammetry, but the shape of the cyclic voltammograms gives some additional and useful information, e.g. the location of the reaction on the potential scale, the occurrence of additional electrode processes and electrode passivation, which can provide additional direction on reliability and reproducibility of the determined values.

Due to that, for determination of diffusion coefficients in ionic liquids, steady-state cyclic voltammetry was conducted at a three electrode setup with a Pt wire as pseudo-reference electrode and a 5 µm Pt disk UME.