Volume 223, Issue 10 - Hydrogen Storage Materials

The methods of structural characterization of metal and chemical hydrides are reviewed. The existing difficulties and problems are outlined and possible solutions presented. It is shown that powder diffraction, and especially the Direct Space Method, is essential component of hydride research. Crystal structures containing as many as 55 independent atoms (including hydrogen) have been fully characterized using powder diffraction. This is of great importance, because rapid collection of powder data thanks to modern synchrotron and neutron time-of-flight sources opens the possibility for fast in-situ studies, mapping of phase transitions induced by the temperature, pressure, hydrogen content, and chemical reactions. The progress in structural characterization of hydrides goes hand in hand with the progress in the powder diffraction methodology.

Total X-ray and neutron scattering experiments aim to measure both diffuse scattering and, in the case of crystalline samples, Bragg scattering on an absolute scale. The total scattering data can be transformed to the Pair Distribution Function (PDF) which shows short-range correlations between the atoms. Reverse Monte Carlo (RMC) modelling is a technique to make three-dimensional model of the atomic arrangements from total scattering data or PDFs. These models are not constrained by space group symmetry, and can give structural insight that is not available from the time- and space averaged models obtained from Bragg scattering alone. RMC modelling has been used to investigate the short-range structure correlations of the disordered crystalline metal deuterides α -VD 0.8 }, β -VD {0.5} and ZrCr 2 D 4 . It was found that the local structures are similar to those found in corresponding ordered low-temperature phases at length scales up to 3–4 Å and that abnormally short D–D separations do not occur despite very short separation between possible D sites. For amorphous metal hydrides, RMC modelling shows that the local atomic configurations can differ substantially from those found in the corresponding crystalline phases.

Multinary hydrides based on mixtures of s - and p -block metals are often not fully hydrogenated, but possess a polymeric anion which is formed by the p -block component upon reduction by the s -block metals. Hydrogen may occur hydridic and exclusively coordinated by s -block metal ions (Zintl phase hydrides), or be incorporated in the polymeric anion where it acts as a covalently bonded terminating ligand (polyanionic hydrides). The current state of Zintl phase hydrides and polyanionic hydrides is reviewed with an emphasis on crystal structures and bonding principles.

Light weight hydrides based on aluminium are among the most promising materials for hydrogen storage applications since the hydrogen content is up to 11 wt%. The crystal structures of many promising aluminium-based compounds have been determined and refined during the last years. This review gives an overview of structures of tetra-alanates (with AlH 4 tetrahedra), hexa-alanates (with AlH 6 octahedra) and different modifications of AlH 3 . The structure determinations are based on powder diffraction with X-rays, synchrotron X-rays and neutrons. In most structural work neutron diffraction has been used in order to determine accurate hydrogen/deuterium positions. In addition, density functional theory has been used both to predict new structures among these compounds and in combination with experimental work.

Experimental structures of M(BH 4 ) n , where M is a 2 nd –4 th period element, are reviewed with a particluar emphasize on crystal chemistry. It is shown that except certain cases, the BH 4 group has a nearly ideal tetrahedral geometry. Correction of the experimentally determined H-positions allows to compare directly the results obtained by different diffraction techniques and by theoretical calculations. Analysis of coordination geometries for M and BH 4 , and of mechanisms of phase transitions in LiBH 4 , suggest that the directional BH 4 … M interaction is at the origin of structural complexity of borohydrides. The ways to influence their stability by chemical modification are discussed. Study of structural evolution with temperature and pressure is shown to be the way to access fundamental information on structural stability of these systems.

We discuss crystal structure of amides and imides, which are focusing on promising hydrogen storage materials. They are ionic crystal with anion of (NH 2 ) – or (NH) 2– and cation of Li + , Na + , Mg 2+ , Ca 2+ , and so forth. We also discuss reaction mechanisms on hydrogen ab/desorption of amide/imide hydrogen storage materials.

Intermetallic compounds are able to store reversibly large amount of hydrogen. The hydride formation usually takes place through a two-phase domain for which an equilibrium plateau is observed between the metallic phase and its hydride. Such phase transformation induces important structural changes such as volume expansion (either isotropic and anisotropic), symmetry lowering, hydrogen ordering and phase transitions. In some cases, change of the electronic structure, from metallic to iono-covalent state, is reported. Finally, amorphization can occur leading to disproportionation of the parent compound. In this overview we give some examples of these different behaviors to illustrate the fascinating structural properties of these materials that are of potential interest from the fundamental point of view and for practical applications like energy storage either by solid-gas or electrochemical routes.

Hydrides of the i 3 - and R 2 Ni 7 -based (R = light rare earth element) intermetallics exhibit novel structural features. Structures of these hydrides, including CeNi 3 D 2.8 , La 2 Ni 7 D 6.5 , LaNi 3 D 2.8 , and Ce 2 Ni 7 D 4.7 , are formed via a huge volume expansion occurring along a single crystallographic direction. Unique structural features during the formation of the hydrides include: (a) The lattice expansion proceeds exclusively within the RNi 2 slabs leaving the RNi 5 slabs unmodified. Such expansion, about 60% along [001] for the Laves layers, is associated with occupation by D atoms of these slabs; (b) New types of interstitial sites occupied by D are formed; (c) An ordered hydrogen sublattice is observed. In the present work we give (a) a review of the crystal chemistry of the conventional, interstitial type hydrides formed by RT 3 and R 2 T 7 intermetallic compounds (R = rare earths; T = Fe, Co, Ni) as compared to the unusual features of the crystal chemistry of anisotropic hydrides formed by the RNi 3 and R 2 Ni 7 intermetallics and (b) studies of the interrelation between structure and bonding in anisotropic hydrides by performing density functional calculations for CeNi 3 and Ce 2 Ni 7 intermetallic alloys and their corresponding hydrides. These studies allowed obtaining an understanding of the bonding mechanism in the hydrogenated compounds which causes a complete anisotropic rebuilding of their structures. From DOS analysis, both initial intermetallics and their related hydrides were found to be metallic. Bader topological analysis for the non-hydrogenated intermetallics showed that Ce atoms donate in average of almost 1.2 electrons to the Ni sites. Hydrogenation increases electron transfer from Ce; its atoms donate 1.2–1.6 electrons to Ni and H. Charge Density Distribution and Electron Localization Function for Ce 2 Ni 7 D 4.7 phase clearly confirm that the interaction between the Ce and Ni does not have any significant covalent bonding. Ni is bonded with H via forming spatial frameworks –H–Ni–H–Ni– where H atoms accumulate an excess electron density of about 0.5e – . Thus, the tetrahedral or open saddle-type NiH 4 coordination observed in the structures of the hydrides is not associated with the formation of [Ni 0 H 4 1– ] 4– complexes containing a hydrido-ion H –1 . In the structural frameworks there are terminal bonds Ni–H, bridges Ni–H–Ni, and the bonds where one H is bound to three different Ni. These spatial ordered frameworks stand as the principal reason for the anisotropic changes in the structural parameters on hydrogenation. Another unique feature of anisotropic hydrides is the donation of electrons from nonhydrogenated RNi 5 parts to hydrogen in RNi 2 slabs stabilising these fragments.

The LaMgNi 4 -hydrogen system was investigated by in-situ neutron powder diffraction and pressure-composition isotherm measurements at 100 °C and hydrogen (deuterium) pressures of up 50 bar. The system displays three hydride phases that have distinctly different hydrogen plateau pressures and H atom distributions. The cubic α -LaMgNi 4 H 0.75 phase forms below 0.01 bar hydrogen pressure and H atoms fill one type of tetrahedral Ni 4 interstices. The orthorhombic distorted β -LaMgNiH 3.7 phase forms at about 3 bar hydrogen pressure and H atoms fill both tetrahedral LaNi 3 and triangular bi-pyramidal La 2 MgNi 2 interstices. Interestingly, tetrahedral Ni 4 interstices are no longer occupied. Finally, the most hydrogen rich γ-LaMgNi 4 H 4.85 phase forms above 20 bar. It has again cubic symmetry and H atoms continue to occupy triangular bi-pyramidal La 2 MgNi 2 interstices while filling a new type of tetrahedral Ni 4 interstices that are neither occupied in the α - nor in the β -phase. The tetrahedral LaNi 3 interstices occupied in the β -phase are empty. Hydrogen induced depopulations of interstitial sites in metal hydrides are relatively rare and consistent with, but not entirely due to, the onset of repulsive H–H interactions at increasing hydrogen concentrations.

The efficacy of different structural models for describing the observed neutron-powder-diffraction (NPD) measurements of bulk polycrystalline YD 3 as well as other hexagonal rare-earth ( i.e. , Nd, Tb, Dy, Ho, Er, and Tm) trideuteride powders has been investigated via Rietveld refinement. Between the two possible structural configurations, centrosymmetric P -3 c 1 and noncentrosymmetric P 6 3 cm , the latter can be excluded due to very high correlations found between the positions of the D sites. Hence, the true “diffraction-average” structure for YD 3 and all other rare-earth deuterides studied is centrosymmetric ( P -3 c 1). This seems to contrast with the prior evidence from first-principles calculations and various spectroscopic probes suggesting that the true local symmetry is not P -3 c 1, but rather, noncentrosymmetric. A possible way to reconcile the apparently conflicting conclusions from NPD and spectroscopic measurements is by assuming that the real structure is a twinned arrangement of nanosized, noncentrosymmetric configurations. For example, we demonstrate that the diffraction-average centrosymmetric P -3 c 1 structure can result from a superposition of individual, noncentrosymmetric P 3 c 1 twins. A comparison of neutron vibrational spectra for YH 3 and YD 3 confirms that both compounds share similar structural arrangements.

The crystal structure of rutile-type magnesium deuteride, α -MgD 2 , has been studied by in situ neutron powder diffraction at room temperature and pressures up to 9.3 GPa. The bulk modulus B 0 = 49(2) GPa is about half, the linear compressibilities (0.0063/GPa along a and 0.0041/GPa along c ) are about twice and the variation of the free positional parameter x about six times the values for the isostructural magnesium fluoride, indicating the strong polarizability of the deuteride anion. Under pressure the MgD 6 octahedra are more compressed (four Mg–D distances at 199(1), two at 173(1) pm) than at ambient pressure (195.1(1), 194.8(2) pm) and the structural arrangement becomes more CaCl 2 like, although no structural phase transition has been observed.