Solid Electrolyte Electrochemical Cells for Catalyst Sensing

BY IAN S. METCALFE

1 Introduction

The development of sensors for industrial process monitoring and control is an area of increasing importance. In particular, there are relatively few sensors that are capable of monitoring the state of a catalyst despite the fact that catalyst state can have a very significant impact on overall process performance. Consequently, there is a need to develop new sensors for the in-situ monitoring of catalyst state. Solid electrolyte electrochemical cells show promise as sensors which could be used for intermediate and high temperature application (temperatures greater than about 200°C).

A solid electrolyte is a material in which the electrolytic, or ionic, conductivity is much greater than the electronic conductivity (for solid electrolytes to be practically useful the ratio of electrolytic to electronic conductivities should be of the order of 100 or greater). Solid electrolytes with conduction ions of O2- H+, Li+, Na+, Ag+, F-, Cl- have all been reported. Much attention has been devoted to oxygen-ion conducting solid electrolytes, many of which show appreciable oxygen-ion conductivities in the range of 200-1200°C.

At high oxygen partial pressures an equilibrium is established between the gas phase oxygen, interstitial oxygen ions and electron holes and conductivity is predominantly p-type due to electron-hole transfer.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1.1)

Using the notation of Kroger and Vink, the symbol Oi" denotes an interstitial oxygen ion carrying an effective double negative charge and h· denotes an electron hole.

At low oxygen partial pressures an equilibrium is established between lattice oxygen, gas phase oxygen, oxygen ion vacancies and free electrons and conduction is n-type.

2Oo ? O2 + Vo" + 2e' (1.2)

Oo denotes a lattice oxygen, Vo" denotes an oxygen ion vacancy with an effective double positive charge and e' denotes an electron.

Over a large range partial pressures of oxygen ionic conductivity dominates and the material behaves as a solid electrolyte. Under these conditions there is an equilibrium established between oxygen ion vacancies, interstitial oxygen ions and lattice oxygen.

2Oo ? Vo" + Oi" (1.3)

Hence the partial pressure of oxygen and the temperature determine whether the solid will exhibit n-type, p-type or ionic conduction. Although the concentration of defects is important it is also necessary to consider the mobilities of the individual defects; higher ionic mobilities will result in a larger domain for electrolytic conduction. Figure 14 shows the dominant mode of conduction in some mixed oxide materials, exhibiting solid electrolyte behaviour, as a function of temperature and oxygen partial pressure.

Suitable solid electrolytes can be employed as the electrolyte in an electro-chemical cell. The electrolyte is used in the form of a membrane which is impermeable to gas phase transport. Electroactive materials, or electrodes, are deposited on both sides of the electrolyte to increase the rates of charge transfer across the electrolyte interface and it is important that the active molecules in the gas phase have easy access to the electrode/electrolyte interface where they can participate in the charge-transfer reactions. For this reason it is necessary, in most cases, to ensure that the electrode has a high porosity while, at the same time, remaining electrically continuous.

Solid electrolyte electrochemical cells can be operated in a variety of ways (the three modes of operation are illustrated schematically in Figure 2). Such a cell may be operated potentiometrically in order to investigate the behaviour of a catalyst of interest. This technique has become known as solid electrolyte potentiometry (SEP). The catalyst itself is deposited in the form of an electrode and is exposed to reaction conditions. The electrode on the other side of the solid electrolyte membrane is exposed to constant gas phase conditions and acts as a reference. The e.m.f. generated by such a cell is related to the state of the catalyst and can be indicative of the catalyst work function. 5 In the case of an oxygen-ion conducting solid electrolyte, the e.m.f. of the cell is a reflection of the thermodynamic activity of oxygen at the catalyst-electrode.

If an external circuit and a power supply are now used, a current can be passed through the cell resulting in a pumping of oxygen ions (in the case of an oxygenion conducting solid electrolyte) towards the electrode of interest or, if the current is reversed, a removal of oxygen ions from the electrode. Consequently, the surface of the catalyst-electrode can be investigated by the passage of current (amperometric techniques). Recently behaviour under amperometric conditions has been recognised to be more complex because of evidence that electrochemical oxygen pumping (EOP) may modify the behaviour of a catalyst.

Solid electrolyte electrochemical cells can also be operated as fuel cells. One electrode is exposed to a fuel while the other is exposed to an oxidant. The driving force due to the differences in oxygen chemical potential at the two electrodes causes electrons to flow around an external circuit and electrical power is generated. The temperature of operation of fuel cells must be greater because of the increased oxygen fluxes required. Fuel cells are commonly operated at temperatures in excess of 800°C. Fuel cells can also be used for the cogeneration of both useful chemicals and electrical energy if the reaction occurring at the anode is appropriate. Fuel cell technology as such is, however, not relevant for sensor development.

It is important to appreciate that solid electrolyte systems are different from aqueous systems in one important way. In a solid electrolyte system the charge transfer takes place across the catalyst/electrolyte interface whereas the chemical reaction takes place on the gas-exposed surface of the catalyst. In aqueous systems the surface at which charge-transfer occurs is the same as the surface over which any catalysis occurs. As a result care must be exercised when making analogies between the two types of cell.

A number of reviews have been written on the use of solid electrolyte electrochemical cells in catalysis. Vayenas and other workers have reviewed the area of SEP as have Gellings et al., Stoukides and Metcalfe. Vayenas et al. have also reviewed work on the modification of catalytic behaviour.

The aim of this review is to present and discuss recent work on solid electrolyte electrochemical cells relevant to in-situ catalyst sensing. Consequently, the area of SEP will be concentrated upon, however, appropriate closed-circuit or amperometric studies will also be discussed. This review is intended to also introduce the reader familiar with heterogeneous catalysis to the electrochemical concepts and techniques required to fully appreciate the research work in this field.

Much work with solid electrolyte electrochemical cells has involved the use of oxygen-ion conductors and it is therefore the use of these conductors which will be concentrated upon in this review article.

2 Electrochemical Principles

In this section both the open-circuit and closed circuit behaviour of electro-chemical cells will be briefly discussed. The mechanism of the charge-transfer process for oxygen-ion conducting systems will also be discussed.

2.1 Open-circuit Behaviour

2.1.l Single Charge-transfer Process- Consider an electrochemical cell consisting of two porous platinum electrodes contacting an oxygen-ion conducting solid electrolyte membrane. The two electrodes are exposed to two different chemical potentials of oxygen µo2' and µo2". The cell may be represented as,

µo2', Pt|MO|Pt, µo2" (2.1)

where MO represents the metal oxide electrolyte.

To determine the e.m.f. of the cell the idea of electrochemical potential will be introduced. The electrochemical potential of species i is defined as,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.2)

where Φ is the electrostatic potential, zi is the charge on the species and F is the faraday constant.

Under open-circuit conditions there can be no net flow of current through the cell. However, there may be individual currents due to the migration of each charge carrying species, given by,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.3)

where x is the distance co-ordinate and cri is the individual conductivity. The net current must be zero, so,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.4)

The charge-transfer reaction at both electrodes is assumed to be,

O2 + 4e- ? 2O2- (2.5)

Using equations (2.2) to (2.4) it can be easily shown that the e.m.f. of the cell will be given by,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.6)

where tion is the ionic transport number (equal to unity if there is no electronic conduction). If tion is less than unity the relationship between tion and µO2 must be known in order to integrate equation (2.6). In practice, if any appreciable electronic conduction is present, then electrons will migrate in the opposite direction to oxygen ions and, if diffusional processes are too slow, the result may be the build-up of oxide at the negative electrode and the appearance of metal at the positive electrode. In such a case, the oxygen potentials at the interfaces are no longer fixed and the cell will exhibit an unstable e.m.f. For solid electrolytes, as tion approaches unity, this is not a problem and the equation can be easily integrated,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.7)

If the oxygen at the interfaces is in equilibrium with the gas phase oxygen, i.e. there is no chemical reaction, and oxygen behaves as an ideal gas, this reduces to,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.8)

Equation (2.8) is, of course, the Nernst equation. From the e.m.f. of the cell, an unknown oxygen partial pressure can be determined. This has led to the use of solid electrolyte electrochemical cells as oxygen probes.

2.1.2 Mixed Potentials - When reactive gases other than oxygen are also present, it is possible for mixed potentials to occur. As an example, if a mixture of oxygen, carbon monoxide and carbon dioxide are now supplied to the cell,

O2,CO,CO2,Pt|MO|Pt,O2 (2.9)

there may be more than one charge-transfer process which can take place, e.g., both of the following reactions may proceed,

* + Oo ? O* + Vo" + 2e' (2.10)

CO* + Oo ? CO2* + Vo" + 2e' (2.11)

where * denotes an electrode site.

Under open-circuit conditions no net current can flow so that the total rate of the anodic reactions must equal that of the cathodic reactions. Applying this condition allows the e.m.f. to be determined if the current-voltage relationships are known for the charge-transfer processes.

Mixed potentials have been found to be important in reaction systems other than CO oxidation. Michaels and co-workers studied a platinum electrode exposed to nitric oxide, nitrogen dioxide and oxygen in the region of 600-800°C. Using e.m.f. data and closed-circuit work at low overpotentials (to avoid modifying coverages of electroactive species) it was shown that two charge-transfer reactions were important,

* + Oo ? O* + Vo" + 2e' (2.12)

NO* + Oo ? NO2* + Vo" + 2e' (2.13)

Mixed potentials have also been shown for the electrodes exposed to reacting mixtures of methane and oxygen.

2.2 Closed-circuit Behaviour - When a voltage is applied to such a cell there is a tendency for oxygen to migrate from one electrode to the other. As a result such arrangements have been used as oxygen pumps.

When the potential difference across the electrode/electrolyte double-layer is not at its equilibrium value then the interface is said to be polarised. When the interface is polarised a net current will flow, the magnitude of the current being dependent upon the difference in the total anodic and total cathodic currents,

i = iA-iC (2.14)

Take as an example the oxygen charge-transfer reaction

* + Oo ? O* + Vo" + 2e' (2.15)

Oxygen from the electrolyte reacts with a vacant site on the electrode forming adsorbed oxygen and creating an oxygen vacancy. The reverse reaction involves oxygen adsorbed on the electrode undergoing electroreduction. The forward reaction is the anodic process and the reverse reaction is the cathodic process. Using transition state theory, anodic and cathodic current densities can be expressed in terms of the potential difference across the interface, the coverage of reactant and the relevant rate constant,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.16)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.17)

where i is the current density, k is the rate constant, the subscript 'A' refers to the anodic process, the subscript 'C' refers to the cathodic process, θ refers to fractional coverages on the electrode, is the symmetry factor, ΔΦ is the potential difference across the electrode/electrolyte interface. At equilibrium, the anodic and cathodic current densities are equal in magnitude and equal to what is called the exchange current density, i0.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.18)

Therefore, rearranging, the exchange current density can be expressed in terms of the rate constants and coverages,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.19)

The actual potential difference across the electrode/electrolyte interface minus the equilibrium potential difference across the interface is known as the electrode overpotential, 11. and is, in effect, the driving force for net charge-transfer,

η = ΔΦ - ΔΦ0 (2.20)

Equations (2.16) and (2.17) can be rewritten in terms of the exchange current density and the overpotential,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.21)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.22)

The net current can be written as,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.23)

which is an equation of the Butler-Volmer type.

When,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.24)

it is possible to make a low-field approximation for the Butler-Volmer equation,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.25)

and there is a linear relationship between the current density and the over-potential.

When,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.26)

it is possible to make a high-field approximation in a form known as the Tafel equation,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.27)

So far it has been assumed that charge-transfer is the limiting step in the overall process. However, other steps may influence the overall rate. Such steps would include any chemical reactions, adsorption and desorption, and diffusional resistances in the gas phase, on the electrode surface, or in the electrolyte lattice. When mass transfer (here mass transfer refers to any step other than the charge-transfer reaction) between the surface of the electrode and the gas phase is rate limiting, the surface coverages of adsorbed species may be modified by the passage of a current. Under these conditions the current-overpotential relationship is modified. It can easily be shown (assuming that the rate of mass transfer between the three-phase region and the gas phase is directly proportional to the effective oxygen concentration difference between these two regions - a complete derivation has been presented by Bard and Faulkner) that,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.28)

where ilA and ilc are the anodic and cathodic limiting currents. Again this equation can be linearized for low overpotentials,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.29)

Defining resistances,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.30)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.31)

where Rmt can be considered a resistance due to mass transfer and Rct can be considered as a resistance due to charge transfer. Equation

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2.32)

and the interface can be modelled as a series of resistances.