Hydrides for Energy Storage: Proceedings of an International Symposium Held in Geilo, Norway, 14 - 19 August 1977

Hydrides for Energy Storage documents the proceedings of an International Symposium held in Geilo, Norway on August 14-19, 1977. This book discusses the thermodynamics of metal, alloy and intermetallic/hydrogen systems; localization and diffusion of hydrogen in lanthanum-nickel compounds; kinetics of hydrogen absorption and desorption; and nuclear magnetic resonance studies of metal hydrides. The calculated heats of formation of metal and metal alloy hydrides; hydrogen absorption into rare earth intermetallic compounds; plateau pressure of RE Ni5 and RE Co5 hydrides; and hydride formation of C14-type Ti alloy are also elaborated. This text likewise covers the mixing effects of two different types of hydrides; hydrogen storage electrode systems; and applications of metal hydrides. This publication is intended for chemists concerned with the fundamental properties of hydrides.

• Language: English
• Publisher: Elsevier Science
• Release date: Oct 22, 2013
• ISBN: 9781483188447

Book preview

Japan

THE PROSPECTS OF HYDROGEN AS AN ENERGY CARRIER FOR THE FUTURE

George G. Libowitz,     Materials Research Center, Allied Chemical Corporation, Morristown, New Jersey, U.S.A. 07960

ABSTRACT

In order to evaluate the possibilities of achieving a Hydrogen Economy, scientific problems involved in the production, storage, transmission, and utilization of hydrogen are discussed. This includes such topics as catalysis, solid state electrolysis, photo-electrolysis, thermochemical generation of hydrogen, and metal-hydrogen interactions. The importance of the last topic is emphasized.

INTRODUCTION

The term Hydrogen Economy has been adopted to describe the use of hydrogen as an energy carrier. In recent years, there have been many articles published on a possible Hydrogen Economy, both in the technical literature [1] and in the popular press [2]. Therefore, a detailed description of a Hydrogen Economy will not be given in this paper. However, one point, which is not always clearly presented in some of the more popular articles, should be emphasized. Namely, that hydrogen is not a primary source of energy, but rather it is a convenient and environmentally desirable way of storing, transporting, and using energy. Consequently, hydrogen must be generated from other sources of energy such as nuclear power, solar energy, etc.

In order to determine the prospects of a Hydrogen Economy in the future, it is necessary to become familiar with some of the problems which must be overcome before hydrogen can be used efficiently as an energy carrier. An indication of some of the scientific problems and possible solutions are given in this paper. Since this is a symposium of physical scientists, economic or political considerations related to a Hydrogen Economy are not discussed. The emphasis is on materials problems which may be associated with (1) the generation of hydrogen, (2) its utilization and (3) transmission and storage.

GENERATION OF HYDROGEN

Catalysts for Production from Coal

Although coal itself can be easily shipped and stored, the advantage of converting coal to hydrogen would be to obtain cleaner burning fuel. Also, hydrogen is a more convenient form of energy for some applications such as automobile fuel.

One method of producing hydrogen from coal is by reaction with steam as shown:

(1)

(2)

The relative amounts of the components of the synthesis gas formed in the first reaction depend upon the type of coal used, the temperature, and other conditions of reaction. The water shift reaction (2) requires catalysts in order to proceed at a sufficiently rapid rate. However, one problem is that most heterogeneous catalysts which could be used for this reaction tend to become poisoned by the sulfur in the coal. With the increased use of high sulfur coals, it will be necessary to find new catalysts which, in addition to having good catalytic properties, must not be poisoned by sulfur or sulfur oxides. Various possible sulfide catalysts are being investigated including sulfospinels, layered transition metal sulfides, and rare earth sulfides.

Water Electrolysis

An established method of generating hydrogen, which should become more important with the increased availability of nuclear energy, is the electrolysis of water. This method will also be significant in the development of newer sources of energy such as solar, wind, and ocean thermal gradients.

Because of problems associated with corrosion and variation in concentration of aqueous electrolytes, the use of solid state electrolytes are being explored. For the electrolysis of water, the migrating species must be either hydrogen or oxygen.

An example of a solid electrolyte, in which ionic transport is via hydrogen, is a perfluorinated sulfonic acid polymer developed at General Electric [3]. The behavior of this electrolyte is shown schematically in Fig. 1. Water is introduced at the anode and is decomposed to form oxygen which is evolved, electrons which move through the external circuit, and H+ ions which migrate through the electrolyte as hydrated ions passing from one sulfonic acid group to the next, and finally evolving as H2 gas at the cathode. Since the sulfonic acid groups are fixed in the electrolyte, the concentration of electrolyte remains constant. Other advantages of this electrolyte include its ability to operate at higher pressures, the fact that it is non-corrosive, and reduced power requirements.

Fig. 1 Perfluorinated sulfonic acid polymer electrolyte for electrolysis of water (Ref. 3).

Inorganic defect solids capable of ionic conduction such as yttria, zirconia, and thoria are also being investigated as possible solid state electrolytes. One such electrolyte system [4] (also developed at G.E.) using calcia stabilized zirconia is illustrated in Fig. 2. Some of the Zr⁴ + ions in the ZrO2 lattice are substituted by Ca² +, and in order to maintain electroneutrality, oxygen vacancies VO, are formed in the lattice. Water, that has been vaporized by the neat of coal oxidation (which also may be used to generate the electrical power), is introduced at the cathode and reduced to form hydrogen gas, while oxygen fills the lattice vacancies to form oxygen on normal lattice sites, OO. At the anode, CO reacts with the lattice oxygen to re-form the vacancies, as shown. The oxygen migrates through the electrolyte as lattice vacancies. In addition to some of the advantages mentioned above for the polymer electrolyte, such cells may operate at temperatures as high as 800–1000°C, leading to increased efficiencies.

Fig. 2 Water electrolysis system using calcia stabilized zirconia as an electrolyte (Ref. 4).

Thermochemical Production

It is possible to thermally decompose water by direct application of heat; however, temperatures in excess of 2500°C would be required. Although one such scheme has recently been proposed [5] using solar energy, high temperatures are difficult to obtain by the usual methods of energy production. However, water can be thermally decomposed at lower temperatures by using a series of reactions in which all the reactants (except water) are re-generated, such that the overall result is the decomposition of water to hydrogen and oxygen.

One such system, suggested by Wentorf and Hanneman [6], is illustrated by the set of Eqs. (3) to (7):

(3)

(4)

(5)

(6)

(7)

It can be seen that the sum of these equations is merely:

(8)

Note that the maximum temperature required for any of these reactions is 750°C. Therefore, lower grade heat, such as that available from nuclear reactors, may be used to thermally decompose water. This method is referred to as thermochemical water splitting.

Many such sets of reactions have been proposed and investigated [6]. However, there are also many problems to be solved. The relative kinetics of the reactions are important and side reactions must be avoided. These require appropriate catalysts. Methods of separating the intermediate products must be developed. Also, materials compatibility is important when corrosive intermediates such as HCl and Cl2 are present.

Photoelectrolysis

A relatively new concept for producing hydrogen from solar energy has received a considerable amount of interest recently; the electrochemical photolysis of water, or photoelectrolysis. Although the energy required to decompose water is 2.46 eV (corresponding to light of wavelength of about 500 nm), direct solar photolysis does not occur because water does not absorb light until well into the UV portion of the spectrum where the solar irradiance is weak. However, by using light absorbing semiconductor electrodes immersed in an aqueous solution, as shown schematically in Fig. 3, the normal electrochemical potential of 1.23 eV is required to dissociate water. This corresponds to about 1000 nm; therefore, the visible range of the spectrum can be used.

Fig. 3 Schematic representation of a p-n photoelectrolysis cell.

The cell in Fig. 3 may be viewed as a semiconductor p-n junction, separated by an electrolyte, so that band bending occurs near the semiconductor-electrolyte interface as shown. If the semiconductors are irradiated with light whose wavelength is such that hv> band gap, electron-hole pairs will be formed in each semiconductor electrode. Excess electrons will flow from the p-type semiconductor (cathode)into the semiconductor-electrolyte interface to reduce the H+ ions in the electrolyte according to the reaction (in acidic electrolyte):

(9)

Similarly, holes, h+, from the n-type semiconductor electrode (anode) will oxidize the water as follows:

(10)

The two electrodes are, of course, connected through an external circuit to permit current flow.

If the electrolyte is alkaline, then the reactions corresponding to Eqs. (9) and (10) are

(11)

and

(12)

The sum of Eqs. (9) and (10) or of Eqs. (11) and (12) correspond to the decomposition of water [(Eq. 8)].

The concept of photoelectrolysis was first proposed and partially demonstrated by Fujishima and Honda [7] using TiO2 as the n-type semiconductor anode and platinum metal as the cathode. This type of cell, with only one semiconductor electrode, has been referred to as a Schottky barrier analogue cell [8], and is illustrated in Fig. 4.

Fig. 4 Schematic representation of a Schottky-type photoelectrolysis cell.

In the Schottky-type cell, the band gap of the semiconductor must be greater than 1.23 eV in order that the excited electrons have sufficient energy to decompose water. However, as can be seen in Fig. 5, semiconductors with band gaps greater than 2.5eV will absorb only a relatively small portion of the solar spectrum. For example, TiO2 which has a band gap of 3eV absorbs only about 8 to 10% of the solar spectrum. Therefore, for maximum efficiency, the band gap of the semiconductor electrode in a Schottky-type cell should be higher than 1.3eV (additional energy is needed to overcome irreversible losses in the cell) and less than 2.5eV in order to absorb a sufficient portion of the solar spectrum.

Fig. 5 The solar irradiance curve.

In a p-n cell, the total energy ideally available for photoelectrolysis is the sum of the band gaps of the n-type and p-type semiconductors (if two different materials are used) [8]. Therefore, the band gap of each semiconductor may be less than leV, which means a larger percentage of the solar spectrum could be absorbed with correspondingly greater efficiencies of operation. However, there are other requirements of a semiconductor electrode.

First, the semiconductors must be electrochemically stable. This is particularly important for n-type semiconductors which tend to become oxidized when acting as an anode. For example, CdS will oxidize to Cd² + ions in solution and free sulfur [9] and GaP will oxidize to Ga+3 ions and phosphoric acid [10].

Secondly, the positions of the energy levels in the semiconductors relative to the redox levels in the electrolyte are also important. For example, the bottom of the conduction band, EC, in the p-type semiconductor must be at a higher energy than the H+/H2 redox level (see Fig. 3) so that the photo-excited electrons do not have to overcome an energy barrier in order to reduce the H+ ions [Eq. (9)]. Similarly, the top of the valence band, EV, in the n-type semiconductor should be below the OH−/O2 redox level (since holes flow up). In a Schottky-type cell, a bias voltage can be used to overcome the mismatch of energy levels [11]. However, the energy difference between EC and the H+/H2 redox level in the p-type semiconductor (or between EV and the OH−/O2 level for the n-type semiconductor) must not be too large because this energy difference is not available for dissociation of water, and therefore the efficiency of the cell is decreased [12].

Finally, the relative positions of the flat band potentials (positions of the original Fermi levels before the semiconductor equilibrates with the electrolyte) in the two semiconductor electrodes should not differ too much because this would lead to a large degree of band bending and a corresponding loss of energy [12]; i.e. the energy of the electron at the electrolyte interface would be much less than its energy in the bulk of the semiconductor.

Thus, it can be seen that the requirements of semiconductors for this application are rather stringent and there is need for much further research in order to find appropriate materials [13].

UTILIZATION

One major advantage of a Hydrogen Economy is the ability to conveniently store electricity. Excess electricity may be used to electrolyze water and the hydrogen thus formed is stored. The hydrogen may then be transformed back to electricity via fuel cells. In order that this concept be economically feasible, the efficiency of presently available fuel cells must be improved.

To some degree a hydrogen fuel cell may be viewed as the opposite of an electrolytic cell; instead of electrolyzing water, H2 and O2 are re-combined to generate electricity. Possible new electrolytes for such a cell were discussed under the section Water Electrolysis above. However, in developing new fuel cells it is also necessary to find new electrode materials and electrocatalysts. An electro-catalyst is a substance which activates the reacting molecules (H2 and O2 in this case) such that electron transfer will occur rapidly at the electrode-electrolyte interface. The catalyst can be incorporated into the electrode, or in some cases, the electrode material itself may act as a catalyst.

Other requirements of electrode materials are that they have high electronic conductivities and yet be corrosion resistant. These requirements are frequently mutually exclusive. Oxide layers will usually protect a metal from corrosion, but it will also decrease the conductivity of the material. Types of materials under investigation are carbides such as WC [14], conducting spinels such as NiCo2O4 [15] and heavily doped oxide semiconductors such as Li-doped nickel oxide [16]. Some of the new metallic conducting polymers [e.g. polythiazyl, (SN)x]are also being considered as possible electrode materials [17].

A significant advantage in using hydrogen as a fuel is its versatility; besides direct combustion, and conversion to electricity via fuel cells, hydrogen can be catalytically oxidized at relatively low temperatures. The advantages of this method of utilizing hydrogen include safety, since there is no open flame, and no formation of oxides of nitrogen. Therefore, catalytic oxidation would be desirable for home heating and in appliances such as space heaters and camp food warmers. One problem in using this method however, is the limited life of available catalysts. Therefore, new catalysts for this application also must be developed.

TRANSMISSION AND STORAGE

Hydrogen Embrittlement

Proponents of a Hydrogen Economy have suggested that existing natural gas pipelines may be used to transport hydrogen gas. It has been estimated [18] that, over long distances, the cost of transmitting hydrogen by pipeline will be almost an order of magnitude less costly than transmitting the same amount of energy by electricity. However, in using this method of transporting hydrogen, the problem of hydrogen embrittlement must be considered.

There are three general types of hydrogen embrittlement of metals [19], (1) hydrogen reaction, (2) internal, and (3) hydrogen environment embrittlement. Hydrogen reaction embrittlement is due to the reaction of hydrogen to form internal phases. For example, in hydride forming metals, the formation of hydrides which have volumes 15 to 25% greater than the corresponding metal, will cause stresses and tend to crack the metal. In carbon steels, the hydrogen may react with the carbon to form methane gas which can cause cracking or blistering. In internal hydrogen embrittlement, hydrogen, which is formed from water during melting, casting, pickling, welding, plating or by corrosion, becomes dissolved in the metal. The hydrogen then concentrates at the tips of existing cracks in the metal and tends to propagate the crack through the metal.

The first two types of hydrogen embrittlement may be avoided by eliminating the conditions which cause the embrittlement. For example, in the case of carbon steels, the thermodynamic activity of carbon may be reduced by adding molybdenium so that the carbon no longer reacts with hydrogen.

The third type, hydrogen environment embrittlement, is more difficult to control because its nature is not yet fully understood. In this case, the metal degrades only when in the presence of hydrogen. It is a temperature dependent process with maximum embrittlement usually occurring at room temperature. Small amounts of oxygen impurity in the hydrogen gas will usually inhibit embrittlement, and this is also frequently true for SO2 and CO2 impurities.

One possible mechanism for hydrogen-environment embrittlement is based upon the strong interaction between hydrogen and transition metals. Gilman [20] has suggested that the strong surface adsorption of hydrogen, particularly near crack tips in the metal, will suppress plastic deformation by increasing the energy necessary to create the surface shear step. Thus the tendency towards cleavage will be enhanced, with resulting embrittlement. However, other mechanisms have been proposed and there is need for a great deal of further research on the nature of hydrogen embrittlement [21].

Hydrogen Storage

Hydrogen may be stored as a gas, as a liquid, or in easily dissociated compounds such as metal hydrides, which is the major topic of this symposium. Storing hydrogen as a gas requires large volumes. Even under compression the volume storage efficiency of gaseous storage is not as high as liquid hydrogen, and the weight of the storage cylinder becomes a major disadvantage. Although the volume efficiency is improved when hydrogen is liquefied, the energy required for liquefaction and the need for well insulated containers are disadvantages. Also, when storing for long periods of time there is still considerable loss of hydrogen due to evaporation.

Storing hydrogen as a metal hydride has several advantages. First, with respect to volume, hydrogen can be stored more efficiently than in liquid, or even solid, hydrogen as illustrated in Table 1, which shows the number density of hydrogen atoms in some representative hydrides. In every case, the number of hydrogen atoms per cm³ is greater than that of liquid, or even solid, hydrogen; and in TiH2, the number density is more than double that in liquid hydrogen. However, it can be seen that water also has a relatively high hydrogen density. This points up the second major advantage of metal hydrides, the ease of reversibility of the formation reaction:

TABLE 1

Hydrogen Densities in Some Hydrogen-Containing Compounds

(13)

The formation of the hydride is an exothermic and usually spontaneous reaction, but the hydrogen can be easily recovered by heating the hydride.

The use of metal hydrides is an unusually safe method of storing hydrogen because hydrides are generally quite stable below their dissociation temperatures. Also, since the reverse of Eq. (13) is an endothermic reaction, the self-cooling effect will suppress any loss of hydrogen if a leak develops in the storage system. This method of storing hydrogen requires no thick-walled containers or heavy insulation, and the possibility of explosion due to high pressures is lessened.

The properties required of an efficient metal hydride storage medium are summarized in Table 2. High hydrogen retentive capacity corresponds to hydrides with high hydrogen-to-metal (H/M) ratios. Low dissociation temperatures are necessary so that the hydrogen will be easily recoverable when needed. Low heats of formation are desirable to minimize energy requirements when recovering the hydrogen, and also because there will be less heat to dissipate during formation of the hydride. Light weights are desirable for applications in which the fuel is portable, such as hydrogen-powered vehicles.

TABLE 2

Desired Properties of a Hydrogen Storage Material

High hydrogen retentive capacity

Low temperature of dissociation (≤100°C)

High rates of hydrogen uptake and discharge

Low heats of formation

Low cost of alloy

Light weight

Stable towards oxygen and moisture

None of the known binary hydrides meet all, or even most, of these requirements. Therefore, it is necessary to develop new alloy hydrides which will have the desired properties listed in Table 2. A knowledge of the fundamental properties of metal hydrides, in general, would be of value in designing new hydride system. Such properties have been reviewed in the past [22], and updated reviews of the fundamental properties are presented in following papers by Maeland, Wallace, Flanagan, and Andresen, among others.

There are two general approaches which can be taken in the development of new alloy hydrides. One is modification of the properties of known hydrides by appropriate alloying or variation of the compositions of intermetallic compounds. This approach is described in the papers by Douglass in the case of magnesium hydride, and Machida et al and Davidov et al for intermetallic compound hydrides. The Rule of Reversed Stability, which states that for a given series of intermetallic compounds, the thermodynamic stabilities of the corresponding hydrides will decrease with increasing stability of the intermetallic compound, can be of value in this latter approach. The rule was proposed by VanMal et al [23] and is discussed in following papers by Buschow and Miedema, Gelatt, and Davidov et al.

The second approach to developing new hydrides for hydrogen storage is to synthesize new intermetallic compounds capable of forming hydrides with appropriate properties. This approach has led to several promising systems such as FeTi hydride developed at Brookhaven [24] and the rare earth-transition metal compounds discovered at Philips-Eindhoven [25].

In general, the properties of intermetallic compound hydrides appear to have little, or no, resemblance to those of the constituent metal hydrides. For example, Table 3 shows some typical intermetallic compound hydrides which take up more hydrogen than would be expected on the basis of the constituent metal hydrides. In Table 4, the properties of ZrNiH3 are compared to those of ZrH2 in more detail. It can be seen that the crystal structures are different and that although the intermetallic compound hydride is less stable (dissociation pressure is higher by a factor of 10¹¹), the Zr-H and H-H distances are smaller in that compound [29]. It may be convenient to consider intermetallic compound hydrides as pseudo-binary hydrides.

TABLE 3

Intermetallic-Compound Hydrides

TABLE 4

Comparison of ZrNiH3 With ZrH2

Since there is a very large number of possible intermetallic compounds and an infinite number of compositional variations, it would be desirable to have some way of predicting which intermetallic compounds will react with hydrogen to form hydrides having the properties required of a good hydrogen storage medium. The Rule of Reversed Stability could have some degree of success in this respect [30], but at present, it appears to be of greater value in the first approach; i.e. in predicting the effect of alloying elements on the thermodynamic stability of known hydrides [23].

Certainly, the relationship between the electronic band structure of an intermetallic compound and its behavior with hydrogen is important. Therefore, a better understanding of the electronic structures of intermetallic compounds and how they are modified by interaction with hydrogen would be of value in predicting new intermetallic compound hydrides. There are many papers at the symposium which cover that aspect, including those by Wallace, Pedersen, Korn, Griessen et al, and Gelatt.

The importance of electronic structure relative to crystal structure can be studied by investigating the hydrogen uptake of a metallic glass (sometimes called amorphous) alloy, whose composition is identical to that of a known intermetallic compound. Such studies on Ti-Cu alloys are reported by Maeland in a following paper.

CONCLUSION

The scientific problems discussed in this paper are an indication of the technical difficulties which must be overcome before hydrogen may be efficiently utilized as an energy carrier. Nevertheless, I believe that there will be a Hydrogen Economy in the future. However, it will be attained gradually over a period of time, and probably not all aspects of the Hydrogen Economy will be achieved. Fleet vehicular systems (such as busses) look promising, but the use of hydrogen in private autos appears unlikely in the near future. Off-peak power storage is another promising possibility. Also, as the newer intermittent sources of energy such as solar and wind are developed, the use of hydrogen for energy storage will become more attractive.

However, it is obvious that there is need for a great deal of further research before the Hydrogen Economy becomes a reality.

REFERENCES

1. Gregory, D.P. Sci. Am. January 1973; 228:13. [For example:]. Winsche, W.E., Hoffman, K.C., Salzano, F.J. Scienc. 1973; 180:1325. Bamberger, C.E., Braunstein, J. Am. Sci. 1975; 63:438. [In addition the International Journal of Hydrogen Energy provides articles concerned with various aspects of the Hydrogen Economy in more detail.].

2. For example: The Coming Hydrogen Economy Fortune, November 1972 and Here Comes the Hydrogen Era Readers Digest, December 1973.

3. L. J. Nuttall, A. P. Fickett, and W. A. Titterington, Proc. Hydrogen Economy Miami Energy Conf., T. N. Veziroglu, Ed. pp. S9–33 to S9–37, Univ. of Miami, Coral Gables, Fla. (1974).

4. W. W. Aker, D. H. Broun, H. S. Spacil, and D. W. White, U.S. Patent No. 3,616,334, Oct. 26, 1971.

5. Fletcher, E.A., Moen, R.L. Science. 1977; 197:1050.

6. Wentorf, R.H., Hanneman, R.E. Science. 1974; 185:311.

7. Fujishima, A., Honda, K. Nature. 1972; 238:37.

8. Nozik, A.J. Appl. Phys. Lett. 1976; 29:150.

9. Williams, R. J. Chem. Phys. 1960; 32:1505.

10. A. J. Nozik, Proc. 1st World Hydrogen Energy Conference, Vol. II, Univ. of Miami, Coral Gables, Fla. pp. 5B-31 to 5B-34 (1976).

11. Ohnishi, T., Nakato, Y., Tsubomura, H. Ber. Bunsenges. Physik. Chem. 1975; 79:523.

12. A. J. Nozik, Proc. Conf. on the Electrochemistry and Physics of Semiconductor Liquid Interfaces Under Illumination, A. Heller, Ed., The Electrochemical Soc. Inc., Proceedings Vol. 77–3, Princeton, N.J., pp. 272–289 (1977).

13. Nozik, A.J. J. Cryst. Growth. 1977; 39:200.

14. Bohn, H. Electrochim. Acta. 1970; 15:1273.

15. King, W.J., Tseung, A.C.C. Electrochim Acta. 1974; 19:485.

16. Bevan, H.L., Tseung, A.C.C. Electrochim. Acta. 1974; 19:201.

17. Nowak, R.J., Mark, H.B., MacDiarmid, A.G., Weber, D. J. Chem. Soc., Chem. Commun. 1977; 9.

18. Winsche, W.E., Hoffman, K.C., Salzano, F.J. Science. 1973; 180:1325.

19. W. T. Chandler and R. J. Walter, Proc. Hydrogen Economy Miami Energy Conf., T. N. Veziroglu, Ed., Univ. of Miami, Coral Gables, Fla., (1974) pp. S6–15 to S6–31.

20. Gilman, J.J. Phil. Mag. 1972; 26:801.

21. Thompson, A.W., Bernstein, I.M., eds. Effect of Hydrogen on Behavior of Materials. Metallurgical Soc. of AIME, 1976.

22. Libowitz, G.G.The Solid State Chemistry of Binary Metal Hydrides. New York: W. A. Benjamin Inc., 1965.

Mueller, W.M., Blackledge, J.P., Libowitz, G.G.Metal Hydrides. New York: Academic Press, 1968.

Libowitz, G.G.Roberts, L.E.J., eds. MTP Internatl. Rev. Sci., Inorg. Chem. Ser. 1; 10. Butterworths Ltd., London, 1972:79–116. [Solid State Chemistry].

23. Van Mal, H.H., Buschow, K.H.J., Miedema, A.R. J. Less Common Metals. 1974; 35:65.

24. Reilly, J.J., Wiswall, R.H. Inorg. Chem. 1974; 13:218.

25. van Vucht, J.H.N., Kuijpers, F.A., Bruning, H.C.A.M. Philips Res. Repts. 1970; 25:133.

26. Takeshita, T., Wallace, W.E., Craig, R.S. Inorg. Chem. 1974; 13:2283.

27. Libowitz, G.G., Hayes, H.F., Gibb, T.R.P. J. Phys. Chem. 1958; 62:76.

28. W. L. Korst, U.S.A.E.C. Report No. NAA-SR-6881 (1962).

29. Peterson, S.W., Sodana, V.N., Korst, W.L. J. Phys. (Paris). 1964; 25:451.

30. Buschow, K.H.J., Van Mal, H.H., Miedema, A.R. J. Less Common Metal. 1975; 42:163

SURVEY OF THE DIFFERENT TYPES OF HYDRIDES

Arnulf J. Maeland,     Materials Research Center, Allied Chemical Corporation, Morristown, New Jersey, U.S.A. 07960

ABSTRACT

Binary hydrides, conveniently classified according to bonding as saline, metallic and covalent are reviewed and surveyed with respect to structure and physical properties. Hydrides of inter-metallic compounds, which are of major interest at this meeting, may be considered to be pseudo-binary hydrides and are included in the survey.

INTRODUCTION

Hydrogen with its unique electronic structure of one electron in a 1s orbital forms compounds with most of the elements in the periodic table. Compounds in which there is a metal-hydrogen or metalloid-hydrogen bond are collectively referred to as hydrides. Based on the nature of the metal-or metalloid-hydrogen bond and the resulting physical properties, the hydrides may be classified in three major categories: (1) Saline or ionic hydrides, (2) metallic hydrides, and (3) covalent hydrides.

Saline hydrides have typically, high enthalpies of formation, high melting points, and are electrically conducting in the molten state. The saline hydrides include the binary hydrides of the alkali and alkaline earth (except beryllium) metals. The physical properties of the alkali and alkaline earth hydrides are in many respects similar to the corresponding halides. The similarity extends to the crystal structure as well, particularly in the alkali series. The alkali hydrides have the sodium chloride structure, while the alkaline earth hydrides (except MgH2) have an orthorhombic structure which is related to the structure of the barium halides. The crystal lattices of the saline hydrides consist basically of hydrogen anions and metal cations. This description is not to be construed as exclusive. In lithium hydride, for example, theoretical calculations [1] and diffraction experiments [2] suggest that the electron transfer from lithium to hydrogen is between 0.8 and 1 electron. This implies a strong ionic bond, but with some covalent character. Magnesium hydride occupies a special position. Although classified here as a saline hydride, its physical properties are intermediate between the ionic hydrides and covalent beryllium hydride. MgH2 may thus be regarded as a transition hydride between the saline and covalent hydrides. The dihydrides of europium and ytterbium are isostructural with the alkaline earth hydrides, and may also be regarded as saline hydrides. Ternary hydrides such as LiBaH3, LiSrH3, and LiEuH3 are basically saline.

Metallic hydrides have, as the name implies, metallic properties such as luster, hardness, metallic conductivity (except the higher hydrides of the rare earths), but unlike metals they are quite brittle. Another characteristic of metallic hydrides is their deviation from stoichiometry which in many cases is unusually large. Hydrides of those transition metals which form binary compounds with hydrogen (Groups IIIA through VIIIA) are classified as metallic hydrides. This includes the rare earth hydrides (except Eu and Yb) and the actinide hydrides. Many of the intermetallic compound hydrides which are discussed at this meeting, e.g. TiFeH2, LaNi5H6 and related compounds, have properties which suggest that they be classified as metallic hydrides. For convenience these hydrides may be regarded as pseudo-binary hydrides.

The nature of the chemical bonding in the metallic hydrides has been the subject of much controversy[3–7]. Two opposing models have been proposed: the protonic and the anionic. In the protonic model[8] hydrogen is assumed to donate its electron to the d-band of the transition metal forming essentially an alloy with the metal. Hydrogen may thus be considered to exist as protons, partially screened by the conduction electrons, in the metal sub-lattice. The opposing view[9] asserts that hydrogen accept electrons from the metal to form hydride anions and metal cations, i.e. a saline hydride. Major support for the protonic model has come from the fact that most metallic hydrides are metallic conductors. Libowitz has pointed out, however, that the trihydrides of the rare earths become semiconductors and their electronic properties are more readily explained by the ionic model[10]. The relatively large enthalpies of formation of most metallic hydrides (they are comparable to the enthalpies of formation of the saline hydrides in many cases) appear to favor the anion model. Experimental results from Mossbauer spectroscopy, positron annihilation magnetic susceptibility, and nuclear magnetic resonance as well as theoretical calculations have not been conclusive, but have been interpreted in favor of one or the other of these two models