The structural situation

Let us here focus on the special structural situations we are likely to meet in studying crystalline proton conductors, and on how these can best be treated by conventional crystallographic refinement techniques. Reasonably, the same types of local situation occur in amorphous solids and in liquids, but their lack of translational symmetry renders them inaccessible to accurate study. We shall see that access to high quality single-crystal data (X-ray or neutron) gives no guarantee of a satisfactory result; careful thought must also be given to the method of refinement. Since many structural examples of proton conducting materials can be found elsewhere in this book, we shall here approach the problem from a general standpoint. Examples will subsequently be taken from the author's own work to illustrate the structural situations discussed.

What then is the structural feature most characteristic of a proton conductor system? It is generally believed that a proton is transferred through a solid in one of two distinct ways: by a vehicular mechanism, whereby the proton rides on a carrier molecule of type NH4+ or H3O+ ion, or by a Grotthuss mechanism, in which the proton jumps from a donor to a suitably placed acceptor molecule (typically, from H3O+ to H2O, or from H2O to OH–). How then is such a process sensed in a conventional diffraction experiment?

Introduction

The a.c.-impedance method described in the previous chapter is often employed to characterize the electrical properties of solid proton conductors. The method gives a rapid answer to the question ‘does the material under investigation conduct electricity’ and the method is not critical with respect to size, shape and quality of the sample used.

If the a.c.-impedance method shows that the material is a conductor there still remain the following questions.

1. (1) Is the material an electronic or an ionic conductor or a mixture of both?

2. (2) Which kind of ion is responsible for the conduction in the case of several possible moving ions, e.g. is CsHSO4 a cesium ion conductor, a protonic conductor, a sulphate/hydrogensulphate conductor or a mixture of some or all of these possibilities?

3. (3) Are the materials in the case of polycrystalline materials true bulk conductors or do they possess surface conductivity, where the conduction takes place in the boundary phase between the individual grains?

4. (4) If proton conduction is verified, is the conduction mechanism then proton hopping between partially occupied proton sites accompanied by a reorganization of the anion lattice or is the mechanism proton transport by a carrier (e.g. water, hydrazine or ammonia) which possibly counter-diffuses when the proton is discharged at the cathode (vehicle mechanism)?

Only the third question can be answered directly on the basis of simple a.c. measurements.

Compounds containing loosely bonded molecules form by far the majority of all known solid proton conductors. Most of them are hydrates and their conductivity generally strongly depends on their state of hydration (see Chapter 3). Therefore, conduction mechanisms similar to the ones discussed for aqueous solutions have been suggested to hold also for this family of solids. But this is still a matter of ongoing research and in the following, a recent concept will be put forward using, as a background, the previous models. It provides directly a framework for explaining proton conduction in solid hydrates.

The peculiarity of proton transport in aqueous solutions was recognized early on by the exceptionally high limiting equivalent conductivity of the proton (350 cm2 Q–1 at 25 °C1) compared to that of any other monovalent cation, data for which fall into a very narrow range (39, 50, 74, 78, 77, 74 cm2 Ω–1 for Li +, Na +, K +, Cs +, Rb +, NH4+ at 25 °C1). The latter similarity is a direct consequence of cation solvation by water molecules, which diffuse at approximately the same rate (2.26 × 10–5cm2s–1 at 25 °C).

The earliest ideas about proton conduction in aqueous solutions were stimulated by considerations of the electrolytic decomposition of water in 1806. Grotthuss postulated chains of water dipoles along which electricity may be transported.

Introduction

Although there are many protonic conductors which are stable at low temperatures (< 200 °C), they are not suitable for high electrolytic current applications because of their large electrode polarization. Therefore, their usage is probably limited to sensors in which only a voltage signal is valuable.

In contrast to this, high temperature protonic conductors are applicable both to sensors and high electrolytic current usage, e.g. electrolysis, galvanic cells for power supply, etc. since, at high temperature, electrode reactions are able to proceed smoothly and, in general, polarization is small. In Table 34.1, possible applications of high temperature proton conducting solid electrolytes are listed, classifying them as having either an electromotive function or a preferential permeation of protons.

In this chapter, the distinctive features of proton conductors as an electrolyte for use in a fuel cell or in a steam electrolyser are discussed in comparison to those of oxide ion conductors. In addition, the possibility of using a proton conductor in a hydrogen gas separator is also described. As examples, the experiments on fuel-cells, steam electrolysers and hydrogen gas separators using proton conducting, perovskite-type oxides are described.

Fuel-cells

The use of a proton conducting solid electrolyte in a fuel-cell has distinctive features compared with that of oxide ion conductors. illustrated in Fig. 34.1, when a protonic conductor is used, fuel circulation is unnecessary in a hydrogen fuel-cell because no water molecules are generated at the fuel electrode.

Introduction

The energy consumption of electronic devices has decreased gradually with progress in integrated circuit technology. The stable working voltage required for power supplies is today considerably lower than in the first transistor devices. Furthermore, the development of amorphous silicon films makes possible a new design for many portable systems. The demand for miniaturized high energy density and low leakage current capacitors as, for example, a stand-by power source for RAM devices, for actuators or sensors as well the substitution of batteries by long-life power supplies offers new opportunities (pace-makers, watches where the supercapacitor is associated with a balance-wheel…) for capacitors using the electric double layer developed at a highly polarizable, blocking electrode/ electrolyte interface.

The first attempts to achieve such a device were made in the 1970s. The cell had a non-symmetric Ag/RbAg4I5/C structure with a double-layer capacitance at the carbon electrode in the range of 10–40 μF cm–1 interface area. However, due to the redox reaction, the working voltage was too low (<0.7 V) for the electronic devices of the time. Furthermore, the reversibility of the Ag electrode was poor and it was difficult to use fully the surface area of the ultrafine carbon. More recently, double-layer capacitors using acidic solution (H2SO4) as liquid electrolyte were developed by NEC (Nippon Electric Co. Ltd). A liquid electrolyte allows most of the surface area of the carbon electrode to be used. The electrical characteristics of the devices can be classified in relation to the properties of each material as follows.

A knowledge of the true, bulk proton conductivity has been both desirable and difficult to achieve in the characterization of most known solid proton conductors. It was almost 100 years after the first electrical measurements on ice that von Hippel demonstrated that single crystals of pure ice must have a proton conductivity far below the values reported previously.

Quantitative data on bulk proton transport are required especially for the understanding of proton transport mechanisms (see Chapters 29 & 31) including the implications for the use of solid proton conductors in operational electrochemical cells (see Chapters 32 & 39).

In this Chapter, proton conductivity refers to the displacement of protonic species (e.g. H +, H3O+, O H–) in small electric fields across a sample close to thermodynamic equilibrium. This conductance relates directly to the self diffusion coefficient of the corresponding species, which may be significantly smaller than the chemical diffusion coefficient, in particular in the presence of another highly mobile species (e.g. conduction electrons in hydrogen bronzes; see Chapter 7).

There is no standard procedure for the measurement of proton conductivity and experimental techniques as well as structural and chemical considerations have to be adapted to the material under investigation.

The title compounds have received a considerable amount of attention because of their high protonic conductivities at 100–300 °C. However, because of their complex characters there remain many areas of uncertainty. The current status is reviewed here.

Synthesis

It is well known that the β- and β″-aluminas are sodium ion conductors with promising applications as an electrolyte for the sodium–sulphur secondary battery. They have been widely studied because of their technological and scientific interest.

The sodium ions in the conduction planes of β- and β″-alumina type structures are exchangeable with many other monovalent and divalent cations. The main host crystals are β- and β″-alumina, and the Ga, Fe analogues: β- and β″-gallate, β- and β″-ferrite. The guest ions which contain protons include H+, H3O+ (in this review the use of H3O+ does not necessarily mean a hydronium ion), and NH4+. Many combinations of host crystals and guest ions have been prepared; details are given in Table 13.1.

The crystals, even of β-alumina, are attacked corrosively by hot sulphuric acid. Furthermore, crystals are inclined to cleave during ion exchange because the ionic radii of NH4+ and H3O+ are much larger than that of Na+14. These factors have serious implications for the fabrication of high density, polycrystalline, flaw-free ceramics (see Chapter 33).

Introduction

As reported in a previous chapter, under special conditions, perfluorosulphonic ion-exchange polymers behave as proton conducting materials. They possess a set of specific properties which make them suitable for use in various fields.

These ion-exchange materials are permeable only to cations-more specifically to protons – and also to water. Some polar organic solvents of low molecular weight can also permeate, to a much lower extent, through this perfluorosulphonic material. The easy water transport through the polymer prevents local drying which could otherwise occur during specific applications. At sufficiently high degrees of swelling, the electrical conductivity of perfluorosulphonic polymers in the protonated form is high and exceeds that of the material in other cationic forms.

NafionR is the leader of perfluorosulphonic proton conductors. Produced by Dupont de Nemours, it is commercially available in different forms: homogeneous or reinforced membranes, powders, tubes and solutions. From solutions of perfluorosulphonic acid or salt in organic solvents, gels and films can also be made. From NafionR, many kinds of composite material may also be prepared either by including small, dispersed particles inside the polymeric phase or by using it to coat solid electronic conductors. The adhesive power of this material is important when it is maintained in contact with a wide variety of supports.

Introduction

Hydrogen bonding is considered a prerequisite for proton conduction. The underlying rationale is that protons can jump with relative ease only when the energy barriers between donor and acceptor sites are low. The corollary is that, when energy barriers are high, proton conductivity is not expected to occur.

Since the idea of low energy barriers is intuitively appealing, the search for proton conductors has concentrated worldwide on systems with recognizable H-bonding, both in inorganic systems and biological membranes. This has created a bias which may have prevented the recognition of some important proton transport mechanisms that appear to be present in non H-bonded system.

In this chapter, I shall discuss highly ionic hydroxides that are probably the least likely candidates for proton conduction. However, on closer inspection, they reveal an interesting fundamental behaviour that may be important for understanding proton transport in other systems, for instance across bilayer membranes containing inner sections that are considered to be proton impermeable.

Non-hydrogen bonded systems

Inorganic hydroxides with the highest degree of ionicity are the binary hydroxides of the alkali and alkaline earth metals, MeIOH and MeII(OH)2 with MeI = Na+, K+ …, and MeII = Mg2+, Ca2+ …, respectively. They all contain well-defined, essentially non H-bonded OH– anions.

Most alkali hydroxides undergo a rotational transition prior to melting.

Introduction

Raman scattering spectroscopy is a versatile and powerful technique for studying composition, structure and dynamics in a wide variety of condensed phases. Transitions between vibrational states are measured in the usual Raman scattering experiment, as in an infrared transmission experiment. However, the selection rules governing Raman scattering differ from those of infrared absorption, so the two techniques should be regarded as complementary. Since the scientific literature regarding the application of Raman spectroscopy to studies of proton conductors is quite extensive, this chapter will not attempt to provide a comprehensive review. Rather the different kinds of structural and dynamical information which can be obtained from these studies will be surveyed, with examples given which illustrate the versatility and power of the technique.

In the next section, the theory of the Raman effect will be briefly reviewed, followed by Section 24.3, the application of Raman spectroscopy to studies of proton conductors. The description of structural information will be divided according to data obtained from band frequencies, intensities, bandwidths, and studies of phase transitions. The chapter will then conclude with several illustrations of dynamical information which can result from an appropriate analysis.

The Raman effect

Raman scattering has been the subject of numerous excellent reviews and will be only briefly summarized here. In general terms, an external electromagnetic field interacts with matter, inducing an oscillating electric moment which acts as a source of scattered electromagnetic radiation.

Historical background and development

Electrochemical cells (electrolysers, batteries and fuel cells) require separators, which allow a flow of specific ionic charges but prevent the transfer of chemical species which remain located either in the cathodic or in the anodic compartment. Among the various separators of electrochemical cells, the ion permeable organic membranes are also used for separation processes such as dialysis and electrodialysis.

Ion permeable organic membranes are polymer films made from ion exchange material. Their thickness is between 50 and 200 um. The polymeric matrix contains fixed ionic groups (the functional sites) the charges of which are compensated by mobile ions (the counter-ions) which can be exchanged with adjoining media. In contact with an aqueous electrolyte solution, ion permeable membranes contain also water and sorbed electrolytes. Monofunctional membranes are ion permselective. The fixed sites of cation permeable membranes are sulphonic or carboxylic groups; those of anion permeable membranes are amine or quaternary ammonium groups.

Among the cation permeable membranes, the perfluorinated membranes which have been developed as separators for fuel cells and chlor-alkali electrolysers show the characteristic features of superselectivity, very high thermal stability and chemical resistance, which are not obtained by the other classes of polymeric ion permeable membranes. Three commercial forms of cation permeable perfluorinated membranes have been proposed:

Proton-containing anhydrous materials are interesting proton conductors as they are stable to relatively high temperatures and contain fewer protons that do hydrates. Their conduction mechanism is thus expected to be easier to understand, and the proton diffusion may be investigated over a rather wide temperature range. However, the conduction mechanism can become complicated when the ‘melting’ temperatures of the protonic species and the non-protonic sublattice are close.

Oxonium perchlorate

Oxonium perchlorate H3OClO4 is the first solid proton conductor whose conductivity, σ300K = 3.4 × 10–4 Ω–1 cm–1, has been found to be comparable to that of a liquid electrolyte. It is also one of the first materials in which the oxonium ion has been identified. The compound is very hygroscopic and photosensitive. A glassy state can be obtained by quenching the molten salt (Tm = 55 °C)3. The crystalline phase undergoes a first order transition on cooling, at 249 K, with a slight volume increase. The low temperature phase is monoclinic with P21/n space group and four formula units in the unit cell. The structure consists of layers of oxonium and perchlorate ions hydrogen bonded by two short (O … O = 0.263 and 0.261 nm) and one long (O … O = 0.271 nm) hydrogen bond (Fig. 11.1).

Introduction

Nuclear magnetic resonance (NMR) is a well-established local probe for the study of dynamics as well as structure in condensed matter systems. The various NMR spin Hamiltonian parameters (see below) accurately reflect the symmetry and the nature of motions at and around the site of the nucleus being studied. When combined with the variations of thermodynamic parameters such as temperature and pressure, NMR can provide a wealth of microscopic, structural and dynamical information, such as lattice spacings, potential barriers and jump/attempt frequencies of the motion. Therefore it is not surprising that NMR has played a major role in the study of fast ionic conductors (FICs). Advantages of the NMR technique stem from its nuclear specificity, it being a true local probe in the sense that inequivalent sites within a unit cell can be distinguished and its non-destructive nature. Fast protonic conductors (FPCs) are eminently suited for investigations by NMR since protons or hydrogen nuclei are about the best nuclei that can be studied by most of the NMR techniques.

There are a number of review articles summarizing and critically examining the NMR studies of FICs, and many of them contain some discussion of fast protonic conductors. Many features of the NMR studies are similar for protonic and other fast ionic conductors. However, there are also aspects peculiar to FPCs, an important one being the influence of local dynamics.