Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

2.1 Introduction

NOTE: Throughout this chapter recommendations by the International Union of Pure and Applied Chemistry (IUPAC) [1, 2] for representing electrochemical cells will be followed stating that a single vertical bar (|) should be used to represent a phase boundary, a dashed vertical bar (¦) to represent a junction between miscible liquids, and a double dashed bar (¦¦) to represent a liquid junction, in which the liquid junction (potential) is assumed to be eliminated. The subscripts account for the following: (s), solid; (l), liquid; (aq), aqueous; and (Hg), amalgam. A slash (/) is used for cases where both forms of the redox couple are assumed to be in solution.

As electrochemistry moved into mixed and nonaqueous electrolytes it became of interest to compare potentials in different media. Serious problems preventing comparison are the liquid junction potentials between different electrolytes. Such liquid junction potentials also occur in the measurement in aqueous systems, but they are generally suppressed by a salt bridge. Salt bridges for aqueous systems usually consist of (saturated) solutions of KCl or NH4NO3. For both KCl and NH4NO3 similar mobilities for the cation and the anion of the respective salt were measured in aqueous solutions. Thus the liquid junction potential between two aqueous electrolytes connected via such a bridge should be smaller than the experimental error (see Chap. 1). Data in aqueous systems without liquid junction potentials are obtained from measurements in cells without transference such as:

A liquid junction potential will, however, occur even in water whenever the activity of HCl [a ±(HCl)] on the right differs from the activity of HCl on the left.

An electrochemical cell consisting of an aqueous and a nonaqueous half-cell or two different nonaqueous solvents in the half-cells always includes a liquid junction potential. Quite frequently the liquid junction potential is ignored and aqueous reference electrodes such as the silver–silver chloride electrodes in various concentrations of KCl or calomel electrodes are used. Care was at best taken that water cannot diffuse into the nonaqueous system. It is important to note that the liquid junction potential is a transport—not a thermodynamic—property (see Chap. 3). This is generally ignored or not fully understood. Thus the literature is full of papers for conversion factors to aqueous reference electrodes [3]. Experimental evaluations of liquid junction potentials were carried out, but always refer to a given setup of the junction [4]. Its value depends on the design of the diaphragm and on time. Thus a given reference cell with a special diaphragm may yield reproducible data when used in electrochemical measurements. However, the reported potential includes the unknown liquid junction potential. Thus using the very same reference electrode may—due to a different separator—yield different potential values. Electrochemical measurements in the same electrolyte, on the other hand, always present thermodynamic values and thus reproducible values (within the experimental skills of the researcher).

Two different approaches are used to establish electrochemical series in nonaqueous and mixed aqueous–nonaqueous systems: (1) a reference redox system [5] or (2) a reference electrode in combination with a bridge to suppress the liquid junction potential [6].

2.2 Reference Redox Systems

Pleskov proposed the redox systems Rb+|Rb and Rb+|Rb(Hg), respectively, as reference redox systems (pilot ion) to correlate data in aqueous and nonaqueous systems [7]. He proposed this system, because of the large ionic radius of Rb+, assuming that interactions of solvent molecules with the Rb+ cation would be small. His assumption was strongly influenced by the Born model [8]. The Born model is based on purely electrostatic considerations and considers the change in Gibbs energy of an ion by the transfer from vacuum into water. The dominating properties in such transfer are the ionic charge and the ionic radius. The larger the radius and the smaller the charge, the weaker is the ion solvent–ion interaction according to the Born model. Such consideration made Pleskov to propose the redox systems Rb+|Rb and Rb+|Rb(Hg) as reference (pilot ion).

Studies on Gibbs energies of transfer of Rb+ [9–11] showed that the interaction of Rb+ with different solvents is not negligible. The Rb+|Rb couple turned out to be too much solvent dependent to be used as a reference redox system to relate electrochemical properties in different solvents.

Strehlow and coworkers studied various organometallic complexes. They formulated requirements for suitable reference redox systems [12] (1) The ions or molecules forming the reference redox system should preferably be spherical with as large a radius as possible, (2) the ions should carry a low charge, (3) the equilibrium at the electrode should be rapid and reversible, (4) both components of the redox couple should be soluble, (5) no change in the geometry of the ligands should occur upon the redox process, (6) the redox potential should be in a potential range that is accessible in as many solvents as possible, and (7) both forms should be stable enough to permit potentiometric measurements. Strehlow suggested the systems ferrocene/ferrocenium ion (ferrocene: bis(η-cyclopentadienyl)iron(II)) and cobaltocene/cobaltocenium ion (cobaltocene: bis(η-cyclopentadienyl)cobalt(II)). While some of the arguments used in this publication are strongly influenced by the Born concept of ion–solvent interactions, this paper became very influential in establishing reference redox systems. Unfortunately the ferrocenium/ferrocene couple is affected by interactions of water with the ferrocenium ion. Thus the ferrocenium/ferrocene couple should not be used in water, making this couple unsuited for establishing a general electrochemical series versus the standard hydrogen electrode [13]. The ferrocenium ion/ferrocene couple consists of cation and a neutral molecule. It is the typical example of a cation/neutral analogue redox couple. Studies in ionic liquids indicate problems with the ferrocenium ion/ferrocene couple [14]. Cobaltocenium ion/cobaltocene, which was also used in nonaqueous solvents but never properly connected to the ferroceniumion ion/ferrocene scale [15, 16], is currently favored in electrolytes based on ionic liquids, but recent publications indicate that ferrocenium ion/ferrocene and bis(biphenyl)chromium(I)/bis(biphenyl)chromium(0) may also be used in room temperature ionic liquids [17, 18]. Decamethylferrocene was later suggested to be superior to ferrocene due to its larger ionic radius [19, 20] as claimed by the authors.

Other reference redox systems have been proposed and used, such as tris(2,2′-bipyridine)iron(I)/tris(2,2′-bipyridine)iron(0) [21], 4,7-dimethyl-1,10 phenanthroline iron(II) [4], and redox systems based on polynuclear aromatic hydrocarbons and the respective radical ions [22–24].

Most data in nonaqueous and mixed solvents have been published versus the bis(biphenyl)chromium(I)|bis(biphenyl)chromium(0) redox couple [25].

The use of reference redox systems became very popular in connection with polarographic and cyclovoltammetric studies. Generally the respective reference redox system was added to the electrolyte with the studied redox system, thus avoiding any liquid junction potential (internal standard). An additional advantage of using a reference redox system is that only one form of the redox couple may be used. The second partner of the redox couple is generated during the polarographic or cyclovoltammetric study. In the case of the redox couple the oxidized form bis(biphenyl)chromium(I) tetraphenylborate [25] (earlier bis(biphenyl)chromium(I) iodide [26] is added; in case of the ferrocenium ion/ferrocene couple the reduced form, namely, ferrocene, is added.

2.3 Reference Electrodes in Combination with a Bridge

Pleskov, during his investigations in acetonitrile, introduced the system Ag|0.1 mol dm−3 AgNO3 as a reference electrode in this solvent [27]. This electrode, later coined Pleskov electrode [28], is an electrode of the first kind. Reversibility of the Ag+|Ag system in many nonaqueous electrolytes led to reference electrodes based on Ag+|Ag in many solvents. Reversibility of this electrode was proven in some solvents by varying the Ag+ concentration and observing Nernstian behavior of the electrode in a Ag+ salt solution, maintaining constant ionic strength by adding a supporting electrolyte. Measurements of such a reference electrode with the same supporting electrolyte in both half-cells should yield potentials free of liquid junction potentials. Thus one could establish an electrochemical series in the solvent chosen. But such series are solvent specific and do not allow comparison of potential data in different solvents. Thus an assumption is necessary to allow establishing a universal scale of redox potentials.

Such an assumption was proposed, namely that a bridge consisting of a 0.1 mol dm−3 tetraethylammonium picrate in acetonitrile suppresses the liquid junction potential between two different nonaqueous electrolytes [6]. The argument in favor of such a salt bridge for nonaqueous electrolytes is the similar electrical mobility of the tetraethylammonium cation and the picrate anion in acetonitrile. This assumption was later expanded to allow for other nonaqueous solvents [28]. Agreement for the electrochemical data was found if the nonaqueous solvents did not have acidic hydrogen atom(s) in the solvent molecule (aprotic solvents) [29]. 0.1 mol dm−3 solutions of either tetrabutylammonium picrate or pyridinium trifluorosulfonate [30] were also used.

Occasionally also the use of so-called pseudo “reference electrodes” has been reported (see Chap. 14). Such “pseudo reference electrodes” became popular in polarography but especially in cyclic voltammetry, employing three electrode arrangements. They consist of a silver or platinum wire or activated carbon dipping into the electrolyte. They substitute for a reference electrode. Such electrodes were reported to exhibit very stable potentials. The ease of such an arrangement was also used in electrochemical studies in (room temperature) ionic liquids. It must be pointed out that electrode potentials versus such electrodes are meaningless as such arrangements do not constitute thermodynamic values.

2.4 Concentration and Activities

Electrochemical measurements especially in polarography and cyclic voltammetry are frequently carried out in the presence of a supporting electrolyte. Concentrations of the supporting electrolyte in aqueous and nonaqueous solutions are usually 0.1 mol dm−3, but may reach 1 mol dm−3 or more. Analysis of activity coefficients for the salt under study is not possible. Conductance studies in nonaqueous solvents were carried out and equations to analyze 1:1 and later 2:2 and other symmetrical and unsymmetrical electrolytes were developed. From such measurements association constants were derived.

2.5 Summary and Recommendation

Electrode potentials should only be reported in thermodynamic arrangements. The most convenient way in polarography and cyclic voltammetry is the use of a reference redox system in the same electrolyte as the system under study. The Ag+|Ag electrode seems applicable to many solvents and may be used as reference electrode in potentiometric investigations.

There are two aspects to reference redox systems. One point is the possibility of compiling electrode potentials in a variety of solvents and solvent mixtures, which are not affected by unknown liquid junction potentials. Unfortunately very frequently aqueous reference electrodes are employed in electrochemical studies in nonaqueous electrolytes. Such data, however, include an unknown, irreproducible phase boundary potential. Electrode potentials of a redox couple measured in the same electrolyte together with the reference redox system constitute reproducible, thermodynamic data. In order to stop the proliferation of—in the view of the respective authors—better and better reference redox systems, the IUPAC recommended that either ferrocenium ion/ferrocene or bis(biphenyl)chromium(I)/bis(biphenyl)chromium(0) be used as a reference redox system [5].

The second point is the assumption that the potential of a half-cell containing the reference redox system is—within experimental error—independent of the nature of the solvent. This assumption is outside the realm of exact thermodynamics and thus open to discussion. As for any extra-thermodynamic assumption it is impossible to prove its validity. This point should be kept in mind especially when discussing single-ion transfer properties.

2.6 The Relation of Redox Potentials in Nonaqueous or Mixed Electrolytes to the Aqueous Standard Hydrogen Electrode

The conversion to the aqueous standard hydrogen electrode as reference half-cell requires an extra-thermodynamic assumption, either the assumption of a solvent independent reference redox system or other assumptions employed in calculating single-ion transfer properties. Details about the procedure and data for univalent cation|metal systems were published [13]. The redox couple ferrocenium ion/ferrocene as reference electrode system is not very suited for such a conversion as the ferrocenium cation undergoes interactions with water and thus impairs the extra-thermodynamic assumption for aqueous solutions. This becomes apparent when comparing the difference in electrode potentials for the ferrocenium ion/ferrocene couple and the bis(biphenyl)chromium(I)/bis(biphenyl)chromium(0) redox couples (Table 2.1). Most of the organic reference redox couples (at least one form) are practically insoluble in water. This makes measurement of reliable electrode potentials very difficult. In some solvents the electrode potential of the ferrocenium ion/ferrocene couple is more positive than the solvent oxidation (especially in sulfur donor solvents) and thus cannot be measured.

Table 2.1 Difference of the half-wave potentials and ½ (E pa + E pc) potentials, respectively, of the ferrocenium ion/ferrocene and bis(biphenyl)chromium(I)/bis(biphenyl)chromium couples in several solvents
Full size table

Table 2.1 clearly shows that the water value for the relation between ferrocene and bis(biphenyl)chromium (ΔE (Fc−BCr)) is too small and transfer properties using ferrocene as base from water will be incorrect.