Fluid Sterilization Filtration

ANALYTICAL CHEMISTRY AND MICROCHEMISTRY
HYDRIDES
TYPES, BONDS AND APPLICATIONS

ANALYTICAL CHEMISTRY
AND MICROCHEMISTRY
Additional books in this series can be found on Nova’s website

under the Series tab.
Additional e-books in this series can be found on Nova’s website

under the e-Books tab.
ANALYTICAL CHEMISTRY AND MICROCHEMISTRY
HYDRIDES
TYPES, BONDS AND APPLICATIONS
PATRICK C. DAM
EDITOR
Copyright © 2018 by Nova Science Publishers, Inc.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted
in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying,
recording or otherwise without the written permission of the Publisher.
We have partnered with Copyright Clearance Center to make it easy for you to obtain permissions to
reuse content from this publication. Simply navigate to this publication’s page on Nova’s website and
locate the “Get Permission” button below the title description. This button is linked directly to the title’s
permission page on copyright.com. Alternatively, you can visit copyright.com and search by title, ISBN,
or ISSN.
For further questions about using the service on copyright.com, please contact:
Copyright Clearance Center
Phone: +1-(978) 750-8400 Fax: +1-(978) 750-4470 E-mail: info@copyright.com.
NOTICE TO THE READER

The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or
implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is
assumed for incidental or consequential damages in connection with or arising out of information
contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary
damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Any parts
of this book based on government reports are so indicated and copyright is claimed for those parts to the
extent applicable to compilations of such works.
Independent verification should be sought for any data, advice or recommendations contained in this
book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons
or property arising from any methods, products, instructions, ideas or otherwise contained in this
publication.
This publication is designed to provide accurate and authoritative information with regard to the subject
matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in
rendering legal or any other professional services. If legal or any other expert assistance is required, the
services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS
JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A
COMMITTEE OF PUBLISHERS.
Additional color graphics may be available in the e-book version of this book.
Library of Congress Cataloging-in-Publication Data
ISBN: H%RRN
Published by Nova Science Publishers, Inc. † New York

CONTENTS
Preface vii
Chapter 1 Crystal Structures of Alkali and Alkaline Earth
Tetra Borohydrides 1
P. Vajeeston
Chapter 2 Bonding Nature and Its Implications with the
Energy Storage and Ionic Transport in Hydrides 81
P. Vajeeston and H. Fjellvåg
Chapter 3 The Application of Metal Hydride Based on
Zr-V Alloy in Hydrogen Plasma 149
Ihor Sereda, Alexandr Tseluyko
and Nikolay Azarenkov
Chapter 4 Preparation and Application of Metal Hydrides 193
N. E. Galushkin, N. N. Yazvinskaya
and D. N. Galushkin
Index 231
PREFACE
Hydrides: Types, Bonds and Applications first proposes metal hydrides

as a fascinating class of compounds due to the small mass and size of
hydrogen. Its medium electronegativity causes a large flexibility in terms of
metal-ligand interactions, resulting in a vast variety of possible
compositions, chemical bonding, crystal structures and physical properties.
However, numerous unsolved problems remain on our way towards a
sustainable, carbon free energy system based on renewable energy and on
hydrogen as a future energy carrier. Thus, the authors present the structural
details of alkali, alkali earth based selected tetra-boro hydrides. Selected
hydrides have recently been suggested for applications in optoelectronics
and as solid electrolytes for battery applications. Their use in optoelectronic
devices depends on their stability with respect to doping, solubility of
shallow donors and acceptors, electrical and optical properties. The authors
describe the nature of the bonding in hydrides, and show how these affect
the properties of these materials, focussing on application in the energy
storage and in the transportation sector. Next, the features of gas discharge
and plasma sources based on Penning trap with metal hydride cathodes are
presented. In such devices, metal hydrides fulfill the functions of both a
cathode and the solid-state generator of working gas. The authors determine
that hydrogen desorbed from metal hydride significantly changes the
properties of the discharge. This is expressed, for example, in the fact that
viii Patrick C. Dam
the plasma source based on Penning trap with metal hydride cathode appears
to generate current-compensated ion beams with the ability to control the
energy of the extracted ions. Lastly, the book discusses metal hydrides
obtaining in sintered electrodes of nickel-cadmium batteries with
electrochemical methods by the way of electrolyte decomposition onto
hydrogen and oxygen. It was shown that as a result of electrolyte
decomposition, oxygen releases from batteries, while hydrogen partly
releases and is partly (in virtue of its high diffusion permeability)
accumulated in sintered matrices of electrodes of nickel-cadmium batteries
in the metal hydrides form.
Chapter 1 - Hydrogen gas is particularly relevant to energy applications
as it is a naturally abundant source of clean energy, with a high density per
mass unit. Moreover, it is the lightest fuel and, unlike electricity, it can be
easily stored. Hydrogen gas is the most promising fuel for a large number of
applications, and it is already in use for space programs around the world.
Metal hydrides are a fascinating class of compounds because the small mass
and size of hydrogen and its medium electronegativity causes a large
flexibility in terms of metal-ligand interactions, resulting in a vast variety of
possible compositions, chemical bonding, crystal structures and physical
properties. Metal borohydrides rank among the most hydrogen-rich groups
of compounds (e.g., LiBH4 has an extreme gravimetric hydrogen content of
18.5 wt %) and attract attention as potential chemical hydrogen stores.
However, the borohydrides explored so far reveal serious drawbacks, e.g.,
either too high or too low temperature of hydrogen release, difficult
reversibility and contamination of evolved hydrogen with B2H6, which
obstruct their application but also triggers the research of new systems.
Consequently, various novel borohydride-based materials have been
prepared and investigated. Numerous unsolved problems remain on our way
towards a sustainable, carbon free energy system based on renewable energy
and on hydrogen as a future energy carrier. Novel materials form the back-
bone in most emerging energy related technologies. In this chapter, the
authors presented the structural details of alkali, alkali earth based selected
tetra-boro hydrides.
Preface ix
Chapter 2 - Hydrides are an important class of materials due to their

potential application for the storage of energy and for the transportation
sector. The main focus of studies on hydrides has been so far the acceleration
of kinetics for the hydrogenation/dehydrogenation processes at moderate
temperatures, and the increase of the weight percentage of hydrogen. The
hydrides characterized by relatively slow kinetics for the hydrogenation/
dehydrogenation processes and by a high decomposition temperature are
therefore of scarce relevance for hydrogen economy. However, selected
hydrides have recently been suggested to find unique exciting applications
in optoelectronics and as solid electrolytes for battery applications. Their use
in optoelectronic devices depends on their stability with respect to doping,
solubility of shallow donors and acceptors, and electrical as well as optical
properties. In general, the hydrides are widely available in powder form,
which is required for hydrogen storage and not preferable for electronic
device technology. For the latter applications, crystalline or amorphous
hydrides in solid/thin film forms are preferable. Moreover, the amphoteric
behavior of hydrogen and its interaction with the environment are known to
hinder the optimization of optoelectronic properties of hydrides from an
experimental perspective. These are currently open problems and
comprehensive theoretical and experimental studies are required to develop
a further understanding. Classically, the chemical bonding in solids are
classified as ionic, covalent, and metallic. However, the bonding in a real
system is a combination of these three classes. The results of theoretical
electronic-structure calculations can be readily used to study the nature of
chemical bonds. In general, the distribution of electron bands in reciprocal
space and the distribution of states in energy [the density of states (DOS)]
provide insight on the bonding in a given material. By analyzing the DOS,
decomposed according to crystal site and angular momenta, one can
distinguish between ionic and covalent bonds in solid systems. The nature
of the bonding can be analyzed using several other tools such as valence-
charge density plot, charge-transfer plot, electron-localization function
(ELF), crystal-orbital Hamilton population (COHP), Bader and Mulliken
atomic charges that are derived from the electronic structure calculations. In
this chapter the authors will describe the nature of the bonding in hydrides,
x Patrick C. Dam
and show how these affect the properties of these materials, focussing on
application in the energy storage and in the transportation sector.
Chapter 3 - The features of gas discharge and plasma sources based on
Penning trap with metal hydride cathodes are presented. In such devices,
metal hydrides fulfill the functions of both a cathode and the solid-state
generator of working gas. Their advantages are high purity of gas injected
(99.99 – 99.999%), along with the safety and compactness in storage.
Hydrogen is injected (desorbed) locally under the influence of ion
bombardment of metal hydride surface, which fact provides return coupling
between the intensity of gas desorption and the parameters of gas discharge.
The rate of sputtering for those materials by plasma ions significantly
reduces as well as heat loads. Above effect is achieved due to the creation
of protective gas target as a result of both the thermal decomposition of metal
hydride and ion stimulated desorption. The feature of metal hydride cathode
under the conditions of gas discharge is a decrease in the ionization potential
of desorbed hydrogen by 0.3-0.5 eV due to the molecules desorption in the
vibrationally/rotationally excited state. This permits a substantially increase
in ionization efficiency and the formation of negative ions by the mechanism
of dissociative attachment in plasma volume. However, hydrogen desorbed
from metal hydride significantly changes the properties of the discharge.
This is expressed, for example, in the fact that the plasma source based on
Penning trap with metal hydride cathode appears to generate current-
compensated ion beams with the ability to control the energy of the extracted
ions. There is also the opportunity of longitudinal extraction of negative
hydrogen ions against the traditional method of extraction across the
magnetic field.
Chapter 4 - This chapter discusses metal hydrides found in sintered
electrodes of nickel-cadmium batteries with electrochemical method by the
way of electrolyte decomposition onto hydrogen and oxygen. It was shown
that as a result of electrolyte decomposition, oxygen is released from
batteries, while hydrogen is partly released and partly – in virtue of its high
diffusion permeability – is accumulated in sintered matrices of electrodes of
nickel-cadmium batteries in metal hydrides form. It was established that in
a case of a long-term operation of batteries (more than five years), in their
Preface xi
electrodes, a lot of hydrogen is accumulated. Along with this, a gravimetric

capacity of a sintered nickel matrix of oxide-nickel electrode as a hydrogen
accumulating unit becomes equal to 20.1 wt%, while its volumetric capacity
is 400 kg m-3. These values exceed thrice all the previous data obtained with
traditional thermo-chemical methods for any reversible metal hydrides
including hydrides of magnesium and complex hydrides.
However, the hydrogen desorption from these hydrides by the traditional
thermo-chemical method runs at the high temperature (800°C) and very
slowly. Also in this chapter, it is proved by experiments that the thermal
runaway can be used as the new high-performance method of hydrogen
desorption from any metal hydrides. On its kinetic and thermodynamic
parameters, this method is considerably superior to requirements set by the
US Department of Energy for hydrogen storage systems. Notably, this
method of hydrogen desorption can work at any ambient temperature and
pressure. Besides, with use of this method, the hydrogen desorption takes
place due to the electrochemical reactions of the thermal runaway, this
process is easily controllable by electrotechnical methods. Also it should be
noted that electrochemical processes are far less inertial than the thermal
processes used in the traditional thermo-chemical method.
In: Hydrides: Types, Bonds and Applications ISBN: 978-1-53613-581-7
Editor: Patrick C. Dam © 2018 Nova Science Publishers, Inc.
Chapter 1
CRYSTAL STRUCTURES OF ALKALI AND

ALKALINE EARTH TETRA BOROHYDRIDES
P. Vajeeston*
Center for Materials Science and Nanotechnology,
Department of Chemistry, University of Oslo, Oslo, Norway
ABSTRACT
Hydrogen gas is particularly relevant to energy applications as it is a

naturally abundant source of clean energy, with a high density per mass
unit. Moreover, it is the lightest fuel and, unlike electricity, it can be easily
stored. Hydrogen gas is the most promising fuel for a large number of
applications, and it is already in use for space programs around the world.
Metal hydrides are a fascinating class of compounds because the small
mass and size of hydrogen and its medium electronegativity causes a large
flexibility in terms of metal-ligand interactions, resulting in a vast variety
of possible compositions, chemical bonding, crystal structures and
physical properties. Metal borohydrides rank among the most hydrogen-
rich groups of compounds (e.g., LiBH4 has an extreme gravimetric
hydrogen content of 18.5 wt %) and attract attention as potential chemical
hydrogen stores. However, the borohydrides explored so far reveal serious
* Corresponding Author Email: ponniahv@kjemi.uio.no; Web: folk.uio.no/ponniahv.

2 P. Vajeeston
drawbacks, e.g., either too high or too low temperature of hydrogen release,
difficult reversibility and contamination of evolved hydrogen with B2H6,
which obstruct their application but also triggers the research of new
systems. Consequently, various novel borohydride-based materials have
been prepared and investigated. Numerous unsolved problems remain on
our way towards a sustainable, carbon free energy system based on
renewable energy and on hydrogen as a future energy carrier. Novel
materials form the back-bone in most emerging energy related
technologies. In this chapter, we presented the structural details of alkali,
alkali earth based selected tetra-borohydrides.
Keywords: complex hydrides, borohydrides, crystal structure, tetra-boro

hydrides
INTRODUCTION
The world’s population has surpassed seven billion and is estimated to

reach more than nine billion by this mid-century. This massive demographic
growth and the worldwide advance of industrialisation will result in an
increased demand of energy, predicted to double within 50 years [1]. As a
direct result, there is an increased demand to develop renewable energy
resources as the traditional nonrenewable energy sources are exhausting.
Hydrogen, the most abundant element in the universe, can be combusted to
produce water and it is therefore regarded as a future promising renewable
energy resource when this process can be obtained more efficiently and at a
lower cost.
Hydrogen is also more efficient than conventional energy sources, as it
is known to produce 2.5 times more energy per unit mass than fuel [2].
Nowadays, it takes a substantial amount of non-renewable energy to produce
hydrogen at industrial scale for practical applications. It remains very
challenging to achieve highly efficient production of hydrogen.
Future sustainable energy systems call for abundant, cheap and high-
energy density carriers. Hydrogen meets these requirements and its storage
has suggested being most viable in the solid-state chemical hydrides.
Lightweight complex hydrides are the most competitive contenders due to
Crystal Structures of Alkali and Alkaline Earth … 3
their high gravimetric and volumetric capacities. However, their direct

application is hampered by slow kinetics, poor thermodynamics and high
hydrogen-release temperatures [3]. Borohydrides store hydrogen in covalent
bonds and hydrogen is released by thermolysis or hydrolysis of the tetra-
hydroborate anion [BH4]⎺ and chemical reactions involving it. Crystal
structures amongst borohydrides range from packed predominantly ionic
compounds [4] to three-dimensional [5, 6], and open nanoporous [7]
frameworks. There exists a multitude of topologically intermediate
structures, most of them based on the building principle of complex anions
of the type [Mnm+(BH4)p](p-n*m)-, that are counterbalanced by alkali cations.
Aside solid-state hydrogen storage, two energy-related fields are being
ventured into by the complex hydrides’ community. These concern solid-
state electrolytes [8], where the electrochemical stability as well as structural
dynamics of the BH4 anion are assumed to be favourable properties. On the
other hand, the hydridic nature of hydrogen in the tetra hydroborate anion
has been suggested as a means of enhancing the selective physical gas
storage [7]. These arising applications call for a more precise and extensive
characterization of cations that may act as mobile species in superionic or
framework builders in porous networks in different coordination polyhedra
defined by the BH4 ligand. At the same time, many potentially useful ‘real
life’ energy materials present major challenges concerning their structural
characterization. Such material systems are often nano-crystalline,
multiphase and poorly crystallized. The above problems are considered
major issues when it comes to the structural characterization of new
compounds, which forms the basis of crystal design. As the examined
systems become more complex, many cases only low-quality data are
available.
Group I and II salts of [AlH4]-, [NH2]-, and [BH4]⎺ (alanates, amides, and
borohydrides) have recently received considerable attention as potential
hydrogen storage materials. These materials are currently referred to as
“complex hydrides,” although only the alanates contain anionic metal
complexes. However, like the alanates, amides and borohydrides are saline
materials in which hydrogen is covalently bonded to central atoms in
“complex” anions (in contrast to interstitial hydrides) [9]. These
4 P. Vajeeston
borohydrides have high hydrogen gravimetric densities (Table 2) and are, in

some cases, commercially available. Thus, a priori, they would seem to be
viable candidates for application as practical, onboard hydrogen storage
materials. Many of these “complex hydrides” have, in fact, been utilized in
“one-pass” hydrogen storage systems in which hydrogen is evolved from the
hydride upon contact with water. However, the hydrolysis reactions are
irreversible and could not serve as the basis for rechargeable hydrogen
storage systems. The thermodynamics of the direct, reversible
dehydrogenation of some complex hydrides lie within the limits that are
required for a practical, onboard hydrogen carrier. All of these materials are,
however, plagued by high kinetics barriers to dehydrogenation and/or
rehydrogenation in the solid state. It was traditionally thought that it would
be impossible to reduce the barrier heights to an extent that would give
reaction rates that even approached those that would be required for
vehicular applications. Complex hydrides were thus not considered as
candidates for application as rechargeable hydrogen carriers. This situation
changed recently, with the pioneering work of Bogdanovic´ and
Schwickardi demonstrating that, upon doping with selected titanium
compounds, the dehydriding of anionic aluminum hydrides could be
kinetically enhanced and rendered reversible under moderate conditions in
the solid state [10]. This breakthrough has led to a worldwide effort in
developing-doped complex hydrides as practical hydrogen storage
materials, study that was quickly expanded to amides and borohydrides.
Research over the past decade has revealed a range of new light metal
borohydrides possessing extremely high hydrogen densities, which may be
considered for possible energy storage applications. However, these
compounds generally suffer from poor thermodynamic and kinetic
properties, i.e., they are often too thermally stable and the hydrogen release
and uptake reactions are too slow for practical applications [3, 11]. A new
approach is to allow multiple hydrides, denoted as reactive hydride
composites, to react during the release of hydrogen to form a different
dehydrogenated state. These composites change the hydrogen release
pathway and may improve the thermodynamic and kinetic properties,
enabling hydrogen release and uptake under more favourable conditions [11-
14]. The concept aims to lower the overall reaction enthalpy by introducing
extra (endothermic) chemical reaction(s), i.e., by utilizing the well-known
Hess law from chemistry, also denoted as Medima principles.
Metal borohydrides are also a multifunctional class of materials that may
also be used as fast ion conductors for new types of batteries or for gas
adsorption [7, 15]. They may also have optical, electronic and magnetic
properties and can be used as reducing agents in organic synthetic chemistry
[16-18]. For this reason, mono-, bi- and tri-metallic borohydrides have been
the object of intense research efforts aimed at developing new synthesis
strategies, at performing structural characterization, and at investigating
physical and chemical properties [9, 19]. In this chapter, we critically review
the structures of selected metal borohydrides (only alkali, alkaline earth, and
their mixed tetrahydroborate) and their derivatives, which may form the
basis of future technological applications. The focus of this chapter is to
discover structures and provide perspectives for further research that may
act as inspiration for the development of novel materials towards rational
materials design.
DIBORANE
As the simplest stable boron hydride in its condensed phase, diborane

exhibits an interesting structural chemistry with uniquely bridged hydrogen
bonds. Diborane (B2H6) has an interesting ring-type molecular structure with
peculiar bridged hydrogen bonds. Due to the extremely reactive nature of
single borane (BH3) which rapidly dimerizes into stable B2H6, diborane thus
represents the simplest stable boron hydride. Although kinetically stable,
diborane is an endothermic compound (ΔHf = +36 kJ mol−1) [20] and thus
has been used as a versatile reagent in a large number of industrial and
laboratory applications [21]. The reactivity of diborane is largely associated
with hydrogen loss, which characterizes this material as a promising
lightweight candidate for hydrogen storage, with 22 wt % hydrogen, the
highest hydrogen content among the considered candidates. Its higher
hydride form, pentaborane (B5H9), has already been investigated as a
6 P. Vajeeston
possible hydrogen-based fuel for high-speed jets. In addition, it has been

reported that diborane can undergo polymerization into higher boron
hydrides [22], which makes borane chemistry an intriguing area for both
experimental and theoretical studies.
Crystal Structure of B2H6
The molecular, electronic, and crystal structures as well as the special

hydrogen bridge bonding of diborane have all been extensively
characterized by various approaches, including electron diffractions [23], X-
ray diffraction [24], and vibrational spectroscopy [25-29], as well as several
theoretical studies [30-32]. In addition to the ambient pressure studies,
Barbee et al. [33] investigated the structures and stabilities of boron hydrides
at elevated pressures by computational methods. Stable molecular boranes
at ambient pressure were found to become unstable, and novel structures
with extended networks characterized by covalent or even metallic
interactions may be formed. The importance of the new boranes formed at
high pressures is that these structures may be suitable as hydrogen storage
media if they are metastable and thus recoverable on the release of pressure.
Experimentally, Nakano et al. [34] reported high pressure studies of
decaborane (B10H14) using Raman, IR, and visible transmission
spectroscopy. A non-molecular phase with a loss of covalent B–H bonding
was found to form above 100 GPa. Song et al. [35] report the first high-
pressure IR spectroscopic study of diborane compressed in a diamond anvil
cell up to 50 GPa.
B2H6 solidifies at 108 K [36]. At least four crystalline forms of diborane
appear in the literature, with a pivotal state single crystal structure
determination of one of them. As early as 1925, on the basis of the powder
diffraction pattern, Mark and Pohland reported a crystalline phase of B2H6
in liquid air, assigned four B2H6 molecules to an ortho-hexagonal unit cell,
got an approximate B−B separation (1.8−1.9 Å), but could not resolve the
atomic positions [37]. The double-bridging structure of the B2H6 molecule
was not known at that time. In 1959, Bolz, Mauer, and Peiser [38] reported
two crystalline phases of B2H6 in the temperature region from 4.2 to 100 K,
naming them α and β. The α phase was formed by deposition from gaseous
B2H6 at 4.2 K. It transformed slowly to the β phase above 60 K, which in
turn was obtained by deposition at 77 K and annealing to 90 K. Interestingly,
neither of the α and β phases nor the additional phase found by passing B2H6
through a microwave discharge showed a diffraction pattern that
corresponded to the phase reported earlier by Mark and Pohland [37]. The
refinement of the diffraction data for the three phases they found, was,
however, not carried out by Bolz, Mauer, and Peiser. In 1965, using single-
crystal X-ray diffraction, Smith and Lipscomb [39] successfully
characterized the β phase in atomic detail (see Figure 1). The structures of
all other crystalline phases of B2H6 remain unknown. The β phase is
monoclinic (space group P21/n, Z = 2, a = 4.40, b = 5.72, c = 6.50 Å, and γ
= 105.1°). Recently, Yao and Hoffmann [36] identified a new polymorph of
B2H6 from the theoretical simulation with lattice parameters for the P21/n
structure are, a = 4.46, b = 8.68, c = 4.55 Å, and β = 120.5° (monoclinic
cell), in good agreement with the measured lattice parameters a = 4.54, b =
4.54, c = 8.69 Å, and γ = 120.0° (hexagonal cell) [37]. Recent theoretical
simulation shows that various molecular diborane structures, including β
diborane and the P21/c structure, are all within 5 meV/molecule of each
other in the calculations at P = 1 atm. It is therefore not surprise that there
are several polymorphs of diborane [36].
Figure 1. Crystal structures of β-B2H6.

8 P. Vajeeston
GENERAL TRENDS IN THE STRUCTURAL CHEMISTRY

OF METAL BOROHYDRIDES
Metals (M) with increasing Pauling electronegativity have increasing degrees

of directionality in the M–BH4 coordination, in particular alkaline earth and d-
block metals, which mostly form framework structures. These structures may
have unexpected structural topologies and contain interpenetrated frameworks or
some degree of porosity. These metal borohydrides often exist as several
polymorphs (metastable in a wide temperature range), with the extreme case of
Mg(BH4)2 showing at least seven polymorphs. These observations highlight the
structural flexibility mainly assigned to directionality and some degree of
covalence in the M–BH4 interaction.
The majority of the bimetallic compounds containing an alkali metal and
an alkaline earth metal, or a d-block metal, tend to form structures built from
discrete composite complex anions formed by a metal and [BH4]⎺, which is
due to a significant difference in Pauling electronegativity. In all cases, the
more electronegative metal coordinates relatively strongly to [BH4]⎺ with a
dominantly covalent interaction, while the less electronegative metal,
usually an alkali, acts as a charge-balancing ion. The number of [BH4]⎺
complexes in the first coordination sphere of metals, in the structures of
metal borohydrides, appears to correlate with the ionic radius of the metal.
As expected, the number of [BH4]⎺ complexes increases with the increasing
cationic radius of the metal and the number varies from three to thirteen in
o-RbY(BH4)4 [40]. In general, the M–[BH4]⎺ distances also increase with
increasing cationic radii of the metal.
BOROHYDRIDES (TETRAHYDROBORATE)
The first report of a pure alkali metal borohydride (tetrahydroborate or

tetra borides) appeared in 1940 by Schlesinger and Brown [41], who
synthesized lithium borohydride (LiBH4) by the reaction of ethyl lithium
with diborane (B2H6). The direct reaction (eq. 1)
2MH + B2H6 2MBH4, where M = Li, Na, K (1)
of the corresponding metal hydride with diborane in etheral solvents under

suitable conditions produces high yields of the borohydrides [42].
ALKALI BASED TETRA BORIDES
LiBH4
1. -LiBH4:
At ambient conditions LiBH4 (-LiBH4) crystallizes [43, 44] with an
orthorhombic (KGaH4-type) structure in which each [BH4]⎺ anion is
surrounded by four lithium Li+ and each Li+ by four [BH4]⎺, both in
tetrahedral configurations (Figure 2a). The calculated positional parameters
for -LiBH4 gives a nearly ideal tetrahedral [BH4]⎺ complex (with H–B–H
bond angles ranging between 108° and 112°) whereas the experimental
results suggest that the complex is highly distorted (bond angles: 85–120
[43]). The theoretically derived B–H distances in -LiBH4 turned out to be
almost equal (ca. 1.22 Å), whereas experimental studies gave different
results (1.28–1.44 Å [45], 1.01–1.28 Å [43]). Assuming that the thus
exposed mutual discrepancy is real it seems likely that the structural
arrangements of LiBH4 as a whole or in parts are sensitive to temperature.
Cell parameters: a = 7.1786 Å; b = 44.4369 Å; c = 6.8032 Å; Atomic
positions: Li(4c): 0.1568, ¼, 0.1015; B(4c): 0.3040; ¼, 0.4305; H1(4c):
0.90, ¼,0.9560; H2(4c): 0.4040, ¼, 0.28; H3(8d): 0.1720, 0.0540, 0.4280.
2. β-LiBH4:
At ca. 381 K, -LiBH4 undergoes a structural transition to β-LiBH4, but
conflicting results have been reported concerning the symmetry of the phase
formed. First β-LiBH4 was believed [46, 47] to belong to the tetragonal
class, but the recent experimental study of Soulié et al. shows that the
symmetry is hexagonal [43]. The structure of the hexagonal high-
10 P. Vajeeston
temperature polymorph of LiBH4 is closely related to that of the

orthorhombic room-temperature polymorph (see Figure 2b). Compared to
the room-temperature structure the [BH4]⎺ tetrahedra in the high-
temperature structure are more symmetric (point symmetry 3m) and less
distorted with respect to bond lengths (B–H = 1.27–1.29 Å) and bond angles
(H–B–H = 106–112°). Furthermore, they are differently oriented with
respect to the room-temperature modification and all point in the same
direction c. As a consequence, one of the four nearest [BH4]⎺ tetrahedra is
shifted away from lithium (Li–B = 3.11 Å) and the other three are shifted
towards lithium (Li–B = 2.50 Å) compared to the room-temperature
structure (Li–B = 2.48–2.54 Å). Thus lithium is no longer coordinated with
9 but with up to 13 hydrogen atoms of which twelve are bonded in groups
of three to the four closest [BH4]⎺ anions (Li–H = 2.30–2.87 Å), and one to
a fifth, more distant [BH4]⎺ anion (Li–H1 = 2.57 Å). The closest distance
between hydrogen atoms (H1–H3 = 2.1 Å) is consistent with repulsive
interactions.
positions: Li(2b): 1/3, 2/3, 0; B(2b):1/3,2/3, 0.5530; H1(2b): 1/3, 2/3,
0.3700; H2(6c): 0.1720, 0.3440, 0.6240.
3. hp1-LiBH4:
The structure of phase hp1-, observed between 1.2 and 10 GPa, has been
determined ab initio from high-quality powder diffraction data. hp1- has a
new structure type in the ABX4 family [48]. Moreover, no isomorphous
derivatives of this Ama2 structure exist. The Li and [BH4]⎺ sublattices
interpenetrate so that Li atoms occupy tetrahedral voids in the [BH4]⎺
network (see Figure 2c). Most interestingly, the [BH4]⎺ anions have a nearly
square-planar coordination comprising four Li atoms, in striking contrast to
the -phase. A square-planar coordination of [BH4]⎺ anions has no
analogues in the crystal chemistry of metal borohydrides. The shortest H···H
distance between two neighboring [BH4]⎺ anions is 1.92 Å, as obtained from
the crystallographic data at 2.4 GPa. These short contacts serve as links
within chains formed by borohydride anions. This geometry is unique, and
such a short H···H distance is unprecedented in metal borohydrides.
Cell parameters: a = 6.449 Å; b = 5.3070 Å; c = 5.2919Å; Atomic

positions: Li(4a): 0,0,0; B(4b): ¼, 0.2711, 0.1730; H1(8c): 0.0954, 0.3338,
0.0743; H2(4b): ¼, 0.0497, 0.1831; H3(4b): ¼, 0.3513, 0.3803.
4. hp2-LiBH4:
The high-pressure phase hp2 was identified by comparing experimental
data with theoretical powder patterns calculated for known ABX4 structure
types [49]. The best description of the experimental diffraction pattern has
been achieved with the cubic Fm3̅m structure having a disordered
arrangement of [BH4]⎺ anions (-NaBH4-type) [49], with a = 5.109 Å at 18.1
GPa [48]. This phase exhibits the shortest H···H contacts between [BH4]⎺
anions (1.67 Å in the optimized model; see Figure 2d). However,
considering the high pressure (greater than 10 GPa) of the transition into
hp2, it is less likely that this phase can be stabilized at ambient conditions.
Figure 2. Crystal structures of (a) -LiBH4, (b) β-LiBH4, (c) hp1-LiBH4, and (d) hp2-
LiBH4 (partial H site occupancy is marked as dark brown colour).

positions: Li(4a): 0,0,0; B(4b): ½, ½, ½; H1(32f): 0.3779, 0.3779, 0.3779
(50% occupancy).
12 P. Vajeeston
NaBH4
At ambient pressure NaBH4 can exist in two structural modifications

[50]. The high-temperature phase of NaBH4 is cubic; it contains
orientationally disordered [BH4]⎺ groups [51], and its structure should be
described in the space group Fm3̅m [52, 53]. On cooling below T0 ≈ 190 K,
the cubic phase transforms to the closely related ordered tetragonal phase
[54]. The same tetragonal phase can also be obtained by applying a pressure
of ∼6 GPa at room temperature [55, 56].
1. -NaBH4:
The crystal structure of sodium borohydride was first proposed by
Soldate in 1947 [57] who predicted that it is based on a face centered cubic
lattice and suggested that the structure consisting of tetrahedral [BH4]⎺ ions
and Na+ ions. Abrahams et al. [49] repeated the experiment below the
transition point (−83°C) and demonstrated that sodium borohydride
becomes tetragonal at −95°C. This transition point was according to
Stockmayer and Stephenson [52], the result of an order–disorder transition
involving different orientations of [BH4]⎺. Later on, in 1985, Davis and
Kennard [58] determined the structure of NaBD4 using neutron diffraction,
which was found to have the NaCl type structure, in space group F-43m with
D atoms tetrahedrally oriented about B and along all cube diagonals. This
gives a random distribution of [BD4]⎺ tetrahedra in two different
configurations. Although in a recent study the structure of NaBD4 was found
to belong to space group P421c [59] but a later experimental [55] and
theoretical [44] study showed that the high pressure of 6.3 GPa can alter the
crystal structure of -NaBH4 (cubic; Fm3̅m; see Figure 3a) to β-NaBH4
(tetragonal; P-421c), which further undergoes with a transition to an
orthorhombic phase (Pnma) at 8.9 GPa, which is stable up to 30 GPa. The
structural arrangement of -NaBH4 is similar to the hp2-LiBH4 phase.
positions: Na(4a): 0,0,0; B(4b): ½, ½, ½; H1(32f): -0.1030, -0.3970, 0.1030
(50% occupancy).
2. β-NaBH4:
On cooling below 190 K [54] or upon a compression to 6 GPa at room-
temperature (RT) [55] a phase with closely related ordered tetragonal
structure appears (β-NaBH4; see Figure 3b). It was first reported in the space
group P-421c [54] but later revised in a higher symmetry P42/nmc [60].
Comparing to the cubic phase, the [BH4]⎺ groups in the tetragonal phase are
ordered in two different orientations. As a result, the number of the shortest
H···H contacts in the tetragonal phase is reduced by one third comparing to
the disordered cubic phase, and the network of the repulsive H···H contacts
changes from isotropic three-dimensional to two-dimensional oriented in the
ab plane. Consequently, the c/a ratio changes discontinuously from 1 to
0.964 upon the cubic-to-tetragonal transition at 186 K [60] due to the
contraction of the c-axis.
positions: Na(2a): ¾, ½, ¾; B(2b): ¾, ½, ½; H(8g): ¾, 0.0290, 0.1444.
Figure 3. Crystal structures of (a) -NaBH4, (partial H site occupancy is marked as

dark brown colour at the H site) (b) β-NaBH4, and (c) -NaBH4. Both KAlH4 (-;β-; -
KAlH4) and RbAlH4 (-; hp1-; hp2-RbAlH4) also have the same structure types as
NaAlH4. -CsBH4 also has the similar structure type of -NaBH4.
3. -NaBH4:
Above 9 GPa a new phase was detected [55] and its structure has been
solved from synchrotron powder diffraction data measured at 11.2 GPa in
BaSO4 structure type [61]. Both the cubic and the tetragonal phases are
strongly textured in diamond anvil cells [54, 62]. For the successful solution
of the tetragonal structure, it was essential to model the texture, including
14 P. Vajeeston
one parameter in the global optimization. The Pnma phase has a BaSO4-type
structure see Figure 3c. This structure type rarely occurs in the family of
light metal hydrides. It is important to note that orthorhombic LiBH4 is
sometimes erroneously attributed to the KGaH4 and BaSO4 structure type.
These two structure types have the same symmetry and Wyckoff sequence
and similar unit cell volumes. Nevertheless, they differ significantly with
respect to the coordination numbers (CN) of the cations and the anions sites
CN=6 with octahedral coordination for the BaSO4-type structure and CN=4
with tetrahedral coordination in LiBH4. In the -NaBH4 phase, the Na atom
is surrounded by six borohydride anions at Na-B distances of 2.7633–2.8493
Å. They are slightly shorter than Na-B distances in the P-421c phase at 10
K 2.93–3.06 Å and in the -NaBH4 phase at ambient conditions 3.07 Å. 16
H2 and H3 atoms form nine short contacts Na-H at 2.09–2.35 Å and H1
atom forms three longer contacts at 2.48–2.52 Å. Thus, with respect to
hydrogen atoms, CN for Na is 12, and [BH4]⎺ units are connected to Na+
cations via tetrahedra edges. The same CN is observed in the β- and -
phases and the Na-H distances, 2.44–2.59 and 2.58 Å, respectively. The
[BH4]⎺ units are very close to the tetrahedral geometry, with B-H distances
centered on 1.17 Å. The shortest H···H distance between [BH4]⎺ anions is
2.31 Å.
positions: Na(4c): 0.1540, ¼, 0.2107; B(4c): 0.5924, ¼, 0.7841; H1(4c):
0.4676, ¼, 0.9161; H2(4c): 0.7298,¼, 0.8926; H3(8d): 0.4135, 0.9825,
0.3363.
KBH4
1. -KBH4:
This KBH4 compound shows a similar behaviour to NaBH4. Its cubic
phase has been studied at room temperature by X-ray diffraction on single
crystals and by neutron powder diffraction [63] in both cases it was
described in the space group Fm-3m.
a = 10.6100 Å; b = 10.6100 Å; c = 10.6100 Å; Atomic positions: K(4a):

0,0,0; B(4b): ½, ½, ½; H1(32f): 0.6053, 0.6053, 0.6053 (50% occupancy).
2. β-KBH4:
The cubic phase transforms to tetragonal (P421c; space group 114) at
65–70 K with a structure equivalent to the tetragonal NaBH4. The K atoms
are located at 2a, B atoms at 2b, and H atoms at 8e positions [64]. The cation
size in the cubic alkali borohydrides, MBH4 increases as M goes from Na to
Cs, thus the unit cell expands and the shortest H···H distances between the
neighbouring [BH4]⎺ anions increase. As a result, the weaker H···H
repulsion leads to lower temperatures of the cubic-to-tetragonal transition
[39].
positions: K(2a): 0,0,0; B(2b): 0,0, ½; H(8e): 0, 0.7906, 0.4150.
3. -KBH4:
Above 6.8 GPa the tetragonal phase transform into Pnma symmetry
(space group 62) with K and B atoms at 4c and H atoms at 4c and 8d [64].
positions: K(4a): 0.2056, ¼, 0.2456; B(4c): 0.4559, ¼, 0.4321; H1(4c):
0.3920, ¼, 0.7070; H2(4c): 0.1490,¼, 0.8200; H3(8d): 0.2070, 0.0459,
0.5400.
RbBH4
At least four polymorphs of RbBH4 are known in the literature. Recently

paper by Fillichuk et al. reported a synchrotron X-ray diffraction study of
the structural evolution of RbBH4 under pressure at room temperature [65].
In the pressure range from 0 to 23 GPa, four different structural phases are
found, with transition pressures of about 3.0, 10.4, and 18 GPa. The phase
transitions have also been observed by in situ Raman scattering experiments
under pressure [65]. The space group symmetry changes in the order Fm-3m
16 P. Vajeeston
→ P4/nmm → C222 → I-42m; none of the three high-pressure structure types

have been previously observed or predicted for the borohydrides.
1. RT()-RbBH4:
The crystal structure at ambient-pressure phase is cubic, with space
group Fm-3m and 4 formula units in the unit cell [49]. It has a NaCl-type
arrangement of Rb cations and [BH4]⎺ anions, where the latter are
orientationally disordered over the two positions around the inversion center
(see Figure 4a). The disordered centrosymmetric structure for the cubic
RbBH4 was assumed by the analogy to the well-established structure of
NaBH4. The [BH4]⎺ group is disordered over two positions. Its partial
ordering at low temperature (LT) was detected at 44 K, [66] without a
reconstruction or a change of the atomic coordination numbers.
positions: Rb(4a): 0,0,0; B(4b): ½, ½, ½; H(32f): 0.410, 0.410, 0.401 (50%
occupancy).
2. HP1-RbBH4:
A transition from the cubic to P4/nmm high-pressure phase starts at 3
GPa and completes at 3.4 GPa. In this high-pressure structure Rb and B
atoms are well ordered. Hydrogen atoms cannot be directly detected from
synchrotron X-ray diffraction data due to the presence of the much heavier
Rb atoms. They can however be located as a part of the rigid tetrahedral
[BH4]⎺ group. Due to the change of the [BH4]⎺ coordination mode, the
corresponding interatomic distances are segregated into four short Rb···B of
3.05−3.2 Å (12 long Rb···H of 2.7−3.1 Å) distances for the η3-BH4
coordination and four long Rb···B of 3.76−3.95 Å (four short Rb···H of
2.6−2.7 Å) distances for the η1-BH4 coordination (η1 and η3 stand
respectively for the coordination via a vertex and a face). The P4/nmm
structure is derived from the cubic by shifting the Rb atom and the [BH4]⎺
group along the c axis of the resulting tetragonal cell. The P4/nmm structure
is intermediate between the NaCl and CsCl types, and the transitions occur
by the mechanism of atomic displacements.

positions: Rb(2c): ¼ ¼ 0.3463; B(2a): ¼,3/4, 0; H(8j): 0.427, ¾, 0.174.
3. HP2-RbBH4:
Above 10.4 GPa the P4/nmm phase starts to transform into a new phase
(C222), and this transition completes at ∼12.2 GPa. The number of [BH4]⎺
neighbors increases to eight in the P4/nmm phase. Rb and B atoms were
reliably located in the special positions. The atomic arrangement represents
an orthorhombically distorted CsCl-type structure (see Figure 4c). The ideal
CsCl-type structure is not met in RbBH4 owing to the better packing of the
anisotropic anions and spherical cations in the distorted structures. The
arrangement of the [BH4]⎺ group inside the Rb8 cube, with η1-BH4 pointing
the B−H bonds toward the midpoint between two Rb atoms, is likely the
densest one, as it is practically the same in the higher pressure C222 and I-
42m phases (see Figure 4b and c).
Figure 4. Crystal structures of (a) -RbBH4 (hydrogen vacancy are marked as dark
brown colour), (b) hp1-RbBH4, (c) hp2-RbBH4, (d) hp3-RbBH4.

positions: Rb(2a): 0, 0, 0; B(2c):1/2, 0, ½; H(8l): 0.380, 0.121, 0.3270.
4. HP3-RbBH4:
Above 17 GPa the I-42m phase is observed. This transition is not fully
completed even at the highest reported pressure (up to 25 GPa). The
octahedral coordination in the ambient-pressure phase gradually transforms
to a more or less distorted cubic arrangement at higher pressures. The
18 P. Vajeeston
average Rb···B distances correlate with the change of the CN: upon the first
transition, CN goes from 6 to 4 + 4 and the average Rb···B distance increases
by ∼7% and then smoothly decreases at higher pressures. A combination of
different [BH4]⎺ coordination modes together with the increase in CN leads
to an increase in density by 8.8% at the cubic-to-tetragonal phase transition,
in spite of the fact that the average Rb···B distance increases. The specific
Rb···B and Rb···H distances show a large spread, related to the occurrence
of the three different Rb···BH4 coordination modes: via a face, an edge, and
a vertex of the tetrahedral anion [65].
positions: Rb(4e): 0,0,0.2329; B(4c):0, ½, 0; H(16j): 0.125, 0.375, 0.080.
CsBH4
1. α-CsBH4
Similar to NaBH4, KBH4 and RbBH4 at room temperature CsBH4
crystallize with a cubic high-temperature (HT) structure having Fm3̄m
symmetry in which the [BH4]⎺ complexes are disordered [49].
positions: Cs(4a): 0,0,0; B(4b): ½, ½, ½; H(32f): 0.5948, 0.5948, 0.5948
[63].
In general, the room temperature structures of MBH4 (M=Na, K, Rb, Cs)
have cubic symmetry and are isotypic with their sodium congener NaBH4.
The alkali cations and [BH4]⎺ anions form a NaCl-type arrangement. The
bond distances those within the disordered [BH4]⎺ complexes increase as the
alkaline radius increases: B-H=1.178 Å (Na), 1.196 Å (K), 1.206 Å (Rb) and
1.217 Å (Cs), effect which is expected in view of matrix effects.
Simultaneously, the M-H distances increases. Calculating a ‘pseudo
hydrogen ionic radius’ from the formula r(H−)=d(M-H)−r(M+) by using the
ionic radius of the alkaline in 12 coordination as given by Shannon [67]
(Na+: 1.39 Å, K+: 1.64 Å, Rb+: 1.72 Å, Cs+: 1.88 Å) yields r(H−)=1.19 Å
(NaBD4), 1.20 Å (KBD4), 1.26 Å (RbBD4) and 1.28 Å (CsBD4) [63].
These values compare well with the tabulated 1.22 Å for hydride ions in
coordination four [67] as in the present compounds (H surrounded by three
M and one B) while illustrating the strong dependence of M–H interactions
on cation electronegativity [67].
ALKALI EARTH TETRA BORIDES
Be(BH4)2
Beryllium borohydride, first synthesized in 1940 by Burg and

Schlesinger [68], has been the subject of many experimental and some
theoretical studies. Both linear and triangular arrangements of B-Be-B have
been proposed, with a variety of arrangements for bonds toward hydrogen
atoms. The only known phase was studied by single crystal X-ray diffraction
[69]. Its tetragonal structure contains helical polymeric chains of BH4Be and
[BH4]⎺ units, where the only independent Be cation is coordinated by two
bridging borohydride anions and one terminal borohydride anion (see Figure
5a). Be atom has a trigonal-planar environment made by three [BH4]⎺
groups, and the bridging [BH4]⎺ group has a linear Be-B-Be geometry. Thus,
the low CN for the Be atom leads to a reduced dimensionality (1D) of the
polymeric structure, where coordination potential of the [BH4]⎺ ligands is
not fully realized. The hydrogen arrangement around Be is approximately a
trigonal prism, but the H atoms are much closer to B (about 1.1 A) than to
Be (1.5-1.6 Å). The crystal structure is tetragonal, the space group is I41/cd,
with a = 13.62 Å and b = 9.10 Å.
positions: Be (16b): 0.1998, 0.0869, 0; B1(16b): 0.1638, -0.0491, 0.0090;
B2 (16b): 0.1386, 0.1935, 0.1204; H1 (16b): 0.0955, -0.0687, 0.0598; H2
(16b): 0.2193, -0.1012, -.0418; H3 (16b): 0.2069, -0.0033, 0.0982; H4
(16b): 0.1413, 0.0083, -0.0841; H5 (16b): 0.1054, 0.1564, 0.0252; H6 (16b):
0.2122, 0.1626, 0.1404; H7 (16b): 0.1450, 0.2714, 0.1015; H8 (16b):
0.0899, 0.1850, 0.2261.
20 P. Vajeeston
Figure 5. Crystal structures of (a) -Be(BH4)2, (b) -Mg(BH4)2, (c) β-Mg(BH4)2, (d) -
Mg(BH4)2, (e) 1-Mg(BH4)2, and (f) -Mg(BH4)2.
Mg(BH4)2
Among the Group I and II metal boro hydrides, magnesium borohydride

(Mg(BH4)2) displays very interesting properties from both fundamental and
applicative points of view [70]. It has the largest number of polymorphs with
the complex structures comprising several hundred atoms in the unit cells.
Some of these polymorphs possess a porosity unique for metal hydrides and
a high specific surface area [7, 71], and others are ultra-dense with one of
the highest volumetric hydrogen densities (147-145 g H2/L at ambient
conditions) among all known metal hydrides [7, 70, 72]. In addition,
Mg(BH4)2 has one of the highest gravimetric hydrogen densities (14.5 wt%
exceeded by only LiBH4 and Be(BH4)2) and a theoretically predicted
hydrogen release at rather mild conditions [73]. It has also the lowest
decomposition temperature (Tdec), and the mildest conditions for partial
rehydrogenation, as demonstrated experimentally. In addition, Mg(BH4)2
could be of interest for batteries applications, since Mg metal holds better
volumetric capacity (3833 mAh cm−3 compared to 2036 mAh cm−3 of Li)
and is more abundant [74]. The published results can be found in more than
hundred articles. The experimentally observed phases of Mg(BH4)2 are

summarized in Table 1. Some of the phases can be synthetized by solvent-
based methods, whereas others have been observed as the result of phase-
transitions. For instance, α-Mg(BH4)2 transforms to the orthorhombic β-
phase (Fddd) upon heating [5, 75-77]. These phases are also referred to as
the LT and HT phase, respectively. The porous γ-Mg(BH4)2 undergoes
thermally-induced phase transitions to ε [78-80] and, subsequently, to β′-
Mg(BH4)2 (allegedly a disordered phase of β [78, 79]). The ε-to-α phase
transition upon cooling of ε-Mg(BH4)2 has been also reported [78, 81],
however, it is not always observed. For example most of the time (ε+β′)-
Mg(BH4)2 mixture measured several hours after being cooled down to RT.
The structures for the ε- and β′-phases have not been published, and the
phases were identified by comparison to the literature [78-80]. The phase-
transitions to the HT-Mg(BH4)2 phases are irreversible. Indeed, β-Mg(BH4)2
or β′ have also been found after re-hydrogenation of Mg(BH4)2 [82-84]. The
high-density δ-Mg(BH4)2 polymorph can be obtained upon compression of
the α and γ-phases [7, 85]. This high pressure phase is much simpler than
the α and γ-polymorphs (Table 1). It is preserved after pressure release but
upon heating to 100°C, it transforms to α either γ starting phase [7]. Some
intermediate phases were also observed upon compression of the α-, and γ-
polymorphs [7]. Besides the crystalline phases, amorphous Mg(BH4)2 can
be obtained via solvent-free synthesis methods, mechanical milling of
crystalline phases [7, 86, 87], and pressure collapse of the porous γ-
Mg(BH4)2 [72].
1. -Mg(BH4)2:
The first polymorph, -Mg(BH4)2, has been reported in P61 space group
symmetry by two independent groups [5, 88] using powder diffraction data.
However, DFT studies suggested a possibly higher P6122 symmetry
(see Figure 5b) [89], which was confirmed by single-crystal diffraction at
100 K [77]. Analysis of the published P61 models shows that the location of
the H-atoms from powder data posed the main problem for the identification
of the correct symmetry. Analysis of the initial P61 models shows that the
determination of H-atom locations from powder data posed the main problem
22 P. Vajeeston
for identification of the correct symmetry. Solid α-Mg(BH4)2 contains an

unoccupied volume of 6.4% within the structure [77]. These voids are, in
principle, large enough (37 Å3) to accommodate a small molecule, such as H2O.
The average B−H distance determined from the single-crystal X-ray
diffraction at 100 K is quite accurate: the standard error of the mean (1.142
Å) is 0.008 Å. The tetrahedral [BH4]⎺ groups are coordinated by Mg atoms
via the two opposite edges. The resulting Mg−H2B bridges are
approximately flat, and this results in relatively long Mg−B distances, which
cover a very narrow range from 2.400 to 2.437 Å, as well as maximizes the
Mg ·Mg distances so that the shortest one exceeds 4.65 Å. However, the
deviation of a Mg atom from the plane defined by the three atoms belonging
to the coordinated H2B edge can be as large as 1 Å, as for instance for Mg
atoms near B1 and B6. Thus, the resulting Mg−H2BH2−Mg fragments are
not strictly linear: the Mg−B−Mg angles range 148−170°. Single-crystal
data show that the H−B−H angles in the Mg−H2B fragments are slightly
more open than for the non-coordinated H2B edge, where two H-atoms do
not bond to the same Mg: the respective angles are 110.8° (average over 12
values) versus 108.8° (average over 24 values). The spread of H−B−H
angles for the Mg-coordinated edges is three times smaller than for the non-
coordinated ones, suggesting that the latter are more prone to distortion,
while the coordinated H2B fragments are more rigid.
Table 1. Experimentally observed polymorphs of Mg(BH4)2 and their

specific (ρ) and volumetric hydrogen (ρv) densities
Phase Space Z Cell parameters, Cell volume, ρ, g Ref.

group Å Å3 cm−3
α-Mg(BH4)2 P6122 30 a = 10.33555 3431.21 0.783 [5, 77]
b = 10.33555
c = 37.08910
α = β = 90°
γ = 120
β-Mg(BH4)2 Fddd 64 a = 37.04892 7439.82 0.76 [90-92]
b = 18.49186
c = 10.85945
α = 90°
Phase Space Z Cell parameters, Cell volume, ρ, g Ref.

group Å Å3 cm−3
β'-Mg(BH4)2 apparently a disordered phase of β not yet solved [79, 80, 85]
γ-Mg(BH4)2 Id-3a 24 a = 15.7575 3912.57 0.55 [7]
α = 90°
Ia-3d 24 a = 15.8234 3961.86 0.5431
α = 90°
δ1-Mg(BH4)2 P63 (indexed, a = 8.35 283.47 [7]
no structural c = 4.68
refinement.)
δ2-Mg(BH4)2 P42nm 2 a = 5.4361 181.65 0.987 [85]
b = 5.4361
c = 6.1468
α = 90°
-Mg(BH4)2 Not yet solved [79]
ζ-Mg(BH4)2 P3112 9 a = 10.424 1009.7 [90-92]
c = 10.729
α = 90°

positions: Mg1(12c): 0.0548, 0.3907, 0.1351; Mg2(12c): 0.5271, 0.4914,
0.0872; Mg3(6a): -0.1257,-0.1257, 0.1667; B1(12c): 0.5897, 0.2948,
0.0833; B2(12c): 0.0176, 0.5088, 0.0833; B3(12c): 0.3064, 0.4556, 0.1173;
B4(12c): -0.2785, -0.2992, 0.1197; B5(12c): -0.0556, 0.1261, 0.1479;
B6(12c): -0.0142, 0.4642, 0.1901; H1(12c): 0.5167, 0.3031, 0.1069;
H2(12c): -0.1421, -0.2628, 0.1246; H3(12c): -0.3464, -0.4267, 0.1331;
H4(12c): -0.3142, -0.2968, 0.0879; H5(12c): -0.1543, 0.051, 0.1728;
H6(12c): -0.0258, 0.0480, 0.1325; H7(12c): 0.0503, 0.2231, 0.1641;
H8(12c): -0.1045, 0.1811, 0.128; H9(12c): 0.1007, 0.524, 0.1751;
H10(12c): -0.1146, 0.3876, 0.1696; H11(12c): -0.0447, 0.5512, 0.2038;
H12(12c): 0.6694, 0.4167, 0.0679; H13(12c): 0.004,0.3862,0.2127;
H14(12c): 0.1006,0.5779,0.1098; H15(12c): -0.0578,0.3769,0.0912;
H16(12c): 0.2777,0.4970,0.1454; H17(12c): 0.1871,0.3661,0.1007;
H18(12c): 0.3701,0.3886,0.1227; H19(12c): 0.3848,0.5642,0.0977;
H20(12c): -0.3101, -0.2178, 0.1345.
24 P. Vajeeston
2. β-Mg(BH4)2:
The -phase with space group P61, is stable below 453 K. Upon heating
above that temperature it transforms to an orthorhombic phase, with space
group Fddd (β-Mg(BH4)2; Figure 5c), stable to 613 K at which point it
decomposes with hydrogen release [88]. The latter phase is metastable upon
cooling, showing an anomalous temperature dependence of the cell
parameter [77]. The β-Mg(BH4)2 phase has a complicated structure with five
distinct types of MgB4 tetrahedra. This phase is consisting of corner-sharing
tetrahedra with Mg atoms at the center and [BH4]⎺ units at the vertices. Mg3
are located close to the c axis, while the other four Mg atoms are relatively
close to the ab planes. No two Mg atoms of the same type do not share a
[BH4]⎺ bridge. Mg1 tetrahedra have one face nearly parallel to the bc plane,
whereas Mg2 tetrahedra have one edge nearly parallel to a and one nearly
parallel to c. Each type of Mg tetrahedron is bonded to one tetrahedron of
the same type, and to three tetrahedra of the other type. Mg—Mg distances
of linked tetrahedra range from 4.61 to 4.72 Å, and the Mg—B—Mg angles
range from 150 to 171°.
positions: Mg1(32h): 0.07749, 0.2501, 0.5381; Mg2(32h): 0.05290, 0.9298,
0.7464; B1(32h): 0.4460, 0.1533, 0.410; B2(32h): 0.1078, 0.0085, 0.2636;
B3(16e): 0.3124,0, 0; B4(32h): 0.4441, 0.6472, 0.560; B5(16g): 0, 0, 0.773;
H1(32h): 0.4284, 0.1778, 0.3524; H2(32h): 0.4593, 0.1908, 0.4600;
H3(32h): 0.4639, 0.1239, 0.3612; H4(32h): 0.4321, 0.1204, 0.4679;
H5(32h): 0.1220, 0.9662, 0.2304; H6(32h): 0.0950, 0.9938, 0.3422;
H7(32h): 0.0891, 0.0239, 0.2006; H8(32h): 0.1247, 0.0499, 0.2812;
H9(32h): 0.3281, -0.0444, 0.0027; H10(32h): 0.2965, 0.0016, 0.0759;
H11(32h): 0.4199, 0.6724,0.5519; H12(32h): 0.4409, 0.5942, 0.5406;
H13(32h): 0.4532, 0.6526, 0.6473; H14(32h): 0.4620, 0.6692, 0.5009;
H15(32h): 0.0218, 0.00992, 0.8267; H16(32h): 0.0049, -0.0433, 0.7192.
3. β'-Mg(BH4)2
The porous γ-Mg(BH4)2 undergoes thermally-induced phase transitions
to ε [79, 80, 85] and, subsequently, to β′-Mg(BH4)2 (allegedly a disordered
phase of β [79, 85]). The structure for the β′-phase has not been published,
and the phase was identified by comparison with the literature [79, 80, 85].
4. -Mg(BH4)2:
The crystal structure of γ-Mg(BH4)2 has a much lower density (0.5431
g cm−3, as calculated from the crystal structure) compared with the α- and
β-phases, resulting in a porous structure with voids surrounded by [BH4]⎺
groups. Recently, Filinchuk et al. investigated the propensity for nitrogen
and hydrogen gas adsorption into the pores of the γ-phase [7]. The
monoclinic solvate structure of Mg(BH4)2⋅1/2S(CH3)2 is a 3D framework
containing two Mg atoms: one atom is tetrahedrally coordinated to four
[BH4]⎺ groups, and the other to four [BH4]⎺ groups and one S(CH3)2 ligand
to form a trigonal bipyramid. Removal of the S(CH3)2 ligand does not break
the integrity of the framework but leads to a highly symmetric cubic
structure of γ-Mg(BH4)2 (space group Id-3a, no. 230: see Figure 5d), where
a single Mg site has a tetrahedral environment of the [BH4]⎺ groups. Its
structure has a 3D net of interpenetrated channels, thus making γ-Mg(BH4)2
the first reported hydride with a large permanent porosity. The empty
volume in the structure amounts to 33 %. The narrowest part of the channel
is defined by a distance of 5.8 Å between hydrogen atoms, while a point at
(1/8, 1/8, 1/8) is 3.56 Å away from the nearest H atom, 4.12 Å from the B,
and 4.82 Å from the Mg atoms. The framework topology of γ-Mg(BH4)2 is
isomorphous to a hypothetical zeolite-type polymorph of SiO2 [93] and to a
porous zinc imidazolate framework ZIF-72, [94] while no other materials
show any similarities.
positions: Mg(24d): 0.25, 0.125, 0.5; B(48g): 0.3098,0.0598, 0.375;
H1(96h): 0.2857,0.0207,0.4373; H2(96h): 0.2974,0.1355,0.3856.
5. 1-Mg(BH4)2:
α- and γ-Mg(BH4)2 have 6.4% and 33% of empty space, respectively, at
elevated pressures. A remarkable volume collapse of approximately 20 %
upon the transition from the α- to a new high-pressure polymorph of
Mg(BH4)2was observed at 1.1–1.6 GPa when using diamond anvil cells [7].
26 P. Vajeeston
This new polymorph, denoted δ1-Mg(BH4)2 (see Figure 5e), has a tetragonal
structure consisting of two interpenetrated Mg(BH4)2 frameworks. Each
framework resembles the cristobalite structure (a polymorph of SiO2), while
their doubly interpenetrated arrangement has a Cu2O topology, which is
typical for metal organic frameworks (MOF). This structural organization is
very robust, as the δ-phase is stable up to 15 GPa and upon a decompression
to 1 bar, and even upon heating to approximately 373 K at ambient pressure,
where it transforms back to α-Mg(BH4)2. δ1-Mg(BH4)2 possesses no empty
voids and has the second highest volumetric hydrogen density (147 g H2/L
at ambient conditions) among all the known hydrides; this value is slightly
below Mg2FeH6 with the hydrogen density of 150 g H2/L. The latter
compound has a much lower gravimetric hydrogen density of 5.5 %,
compared to 14.9 wt % in Mg(BH4)2. The second highest volumetric
hydrogen density in borohydrides, 127 g H2/L, is recorded for the toxic
Be(BH4)2, which has an extreme gravimetric hydrogen density of 20.7 wt %.
In this structure Mg atoms have a tetrahedral environment of four [BH4]⎺
groups, and the [BH4]⎺ groups are coordinated by two Mg atoms through the
opposite edges. The Mg⋅⋅⋅BH4 interaction is highly directional and results in
linear Mg⋅⋅⋅BH4⋅⋅⋅Mg fragments, which can be considered as fundamental
building units in all the structures. Here the [BH4]⎺ group acts as a directional
ligand, similar to the organic ligands (“linkers”) in MOFs.
positions: Mg(2a): 0, 0, 0; B(4c): 0.2584,0.2584,0.2432; H1(8d):
0.2050,0.4641,0.3048; H2(4c): 0.1122,0.1122, 0.3185; H3(4c):
0.2525,0.2525, 0.0460.
6. 2-Mg(BH4)2:
The XRD patterns of the high-pressure phase were indexed in the
hexagonal P63 space group at 5.5 GPa, with lattice parameters a = 8.35(5)
Å, c = 4.68(4) Å, and V = 283.47(8) Å3 [85]. The complete structural
description was however not solved.
7. -Mg(BH4)2:
The porous γ-Mg(BH4)2 undergoes thermally-induced phase transitions
to ε [79, 80, 85] and, subsequently, to β′-Mg(BH4)2 (allegedly a disordered
phase of β [78, 79]). The ε-to-α phase transition upon cooling of ε-Mg(BH4)2
has been also reported [79, 81]. It is however not always observed. The
structure of ε-phase is unpublished, and it has been identified by comparison
with the literature [79, 80, 85].
8. -Mg(BH4)2:
The structure of ζ-Mg(BH4)2 was refined from SR-PXD data measured
at 224°C [71]. Realizing that the diffraction pattern of ζ-Mg(BH4)2 is similar
to that of α-Mn(BH4)2 (space group P3112) [6]. In this structure, both
independent Mg atoms are surrounded by four [BH4]⎺ tetrahedra (see Figure
5f) in deformed tetrahedral coordinations (Mg-B distances within 2.42-2.52
Å, B-Mg-B angles within 64.3-153.19°). A similar tetrahedral coordination
of a cation by four [BH4]⎺ anions is known for Li+ in three of the four known
phases of LiBH4 [43, 48] and in LiK(BH4)2 [5]. On the other side, each
[BH4]⎺ is approximately linearly coordinated by two Mg2+ (Mg-B-Mg angles
within 159° to 170.0°), similarly to manganese borohydride [6]. The
coordination of manganese to hydrogen atoms cannot be determined
unambiguously because of the limited accuracy of the [BH4]⎺ tetrahedra
orientation, which is strongly affected by Mg-H and Mg-B antibump
restraints used in the Rietveld refinement. The configuration resulting from
such restraint refinement can be understood as the Mg-H2BH2-Mn bonding
scheme giving 8-fold coordination of Mg by hydrogen atoms, with Mg-H
distances varying within 1.86-2.304 Å. A single B-H distance of 1.08-1.22
Å was refined for rigid [BH4]⎺ tetrahedra. It is consistent with those in related
borohydrides such as Mn(BH4)2 (1.13 Å [6]).
positions: Mg1(6c): 0.2313,0.9181, 0.1253; Mg2(3a): 0.562, 0.12544,
0.6667; B1(6c): 0.040, 0.6990, 0.006; B2(3b): 0.471, 0.9415, 0.16667;
B3(3a): 0.204, 0.1021, 0; B4(3a): 0.133, 0.8671, 0.3333; B5(3b): 0.706,
0.2944, 0.8333; H1(6c): -0.082, 0.658, 0.001; H2(6c): 0.071, 0.645, 0.929;
H3(6c): 0.069, 0.669, 0.098; H4(6c): 0.104, 0.823, 0.994; H5(6c): 0.428,
28 P. Vajeeston
0.014, 0.21; H6(6c): 0.39, 0.869, 0.091; H7(6c): 0.132, 0.115, -0.072;
H8(6c): 0.277, 0.064, -0.046; H9(6c): 0.020, 0.827, 0.375; H10(6c): 0.125,
0.787, 0.259; H11(6c): 0.753,0.415, 0.859; H12(6c): 0.72, 0.23, 0.915.
In general, all the experimentally observed polymorphs and the lowest-
energy theoretical Mg(BH4)2 structures contain very similar local
configurations: Mg atoms have a tetrahedral environment of four [BH4]⎺
groups, and these [BH4]⎺ groups are coordinated to two Mg atoms through
the opposite edges. The Mg⋅⋅⋅BH4 interaction is highly directional and
results in linear Mg⋅⋅⋅BH4⋅⋅⋅Mg fragments, which can be considered as the
fundamental building units in all the structures. Here the [BH4]⎺ group acts
as a directional ligand, similar to the organic ligands (“linkers”) in MOFs.
On the other hand, Mg atoms form a limited set of MgH8 polyhedra.
Interestingly, out of all the possible eight-vertex polyhedral [95], only the
less uniform Johnson solids are found in the experimental structures. The
stability of MgH8 coordination polyhedra can presumably be linked to the
relative stability of the polymorphs. Both experiments and theoretical
predictions suggest a vast polymorphism of Mg(BH4)2. Moreover, the
experimentally observed phases are stable over relatively wide temperature
and pressure ranges, in particular under ambient conditions, thus indicating
that the reconstruction of strongly bound Mg(BH4)2 coordination
frameworks is kinetically hindered. This behavior is likely due to the high
stability of the linear Mg⋅⋅⋅BH4⋅⋅⋅Mg units, which link the MgH8 nodes into
various framework structures, similar to the partly covalently bonded MOFs.
The relatively small charge transfer from Mg to [BH4]⎺ makes the bonding
partly covalent and is essentially the reason for the MOF-like behavior of
Mg(BH4)2, namely the rich polymorphism and metastability, the large
pressure-induced volume collapses, and the amorphization under pressure.
A possible bonding scheme involves a formation of molecular orbitals
between Mg, H, and B atoms, similar to those in diborane, B2H6.
Ca(BH4)2
Riktor et al. reported on the existence of several Ca(BH4)2 phases [76].

The structure of the -phase was first reported by Miwa et al. [96] in the
space group Fddd and later revised in F2dd by Filinchuk et al. [97]. Besides
the overall improved fit to the powder diffraction data, the non-
centrosymmetric structure shows a group-subgroup relation with the
symmetry of the high-temperature phase. The α′-phase forms via a second
order phase transition at ~495 K and was described in the space group I-42d
[77]. Above 450 K, the closely related F2dd and I-42d phases kinetically
transform into a completely different β-phase. The determination of its space
group symmetry is ambiguous, and its structure was described in two similar
models, corresponding to the lowest binding energies as evaluated by DFT,
namely in the space groups P42/m [98] and P-4 [77] It is likely that the
[BH4]⎺ group in β-Ca(BH4)2 is intrinsically disordered, similar to the high-
temperature phase of LiBH4 [48, 99]. According to the systematic absences
of the diffraction peaks the true symmetry of the disordered structure might
be P42nm [97]. The HT phase is 4–6% denser than the F2dd and I-42d
phases, it is metastable upon cooling, and at room temperature it slowly
transforms back into the orthorhombic phase [53]. Another phase, labelled
-Ca(BH4)2, obtained by wet chemical synthesis, presents an orthorhombic
structure [100]. The crystal structure of the orthorhombic γ-phase of
Ca(BD4)2 is identified by combined synchrotron X-ray powder diffraction,
neutron powder diffraction, and ab initio calculations. It is metastable at all
temperatures and irreversibly transforms into the β-phase at ~590 K.
1. α-Ca(BD4)2:
The α-Ca(BH4)2 structure crystallizes in the noncentrosymmetric space
group F2dd, containing an ordered [BH4]⎺ anion (Figure 6a). The crystal
structures of α-Ca(BH4)2 contain calcium cations nearly octahedrally
coordinated to six borohydride anions. CN for the alkaline-earth atoms
correlate with their radii, and change from 3 in Be(BH4)2 to 4 in Mg(BH4)2,
and finally to 6 in the three Ca(BH4)2 structures. The six Ca-B distances in
α-Ca(BH4)2 within range of 2.816–2.967 Å at 91 K. The local environments
of the Ca atoms differ between the α-, α’-, β-phases by different [BD4]⎺
orientations and deformation of the [BD4]⎺ octahedral coordination (see
Figure 6). The α-phase exhibits more distorted octahedra with respect to the
more regular and higher-symmetry β-phase. This can be seen from the
30 P. Vajeeston
differences between the shortest and the longest B−B interatomic distances
between the [BD4]⎺ groups surrounding the Ca atoms. These differences are
of 11% (3.91−4.366 Å) for the β-phase and 14% (3.807−4.442 Å) for the
α-phase. It is worth noting that, in spite of the similarities in coordination
and calculated formation energy, differences in densities can be as large as
∼4% between the density of the α-phase and the densities of the β-phase and
γ-phase (the density of the β-phase and γ-phase are almost identical. The
coexistence of the α-, β-, and γ-phases as well as the kinetics of the phase
transitions between these phases are very much dependent on the synthesis
method of the samples [76, 101].
positions: Ca(8a): 0, 0, 0; B(16b): 0.0056, 0.2243, 0.0081; H1(16b): 0.0190,
0.1648, 0.1272; H2(16b): 0.0345,0.1817,-0.1256; H3(16b): -0.1180,
0.2553, 0.012; H4(16b): 0.0910, 0.2922, 0.028.
Figure 6. Crystal structures of (a) -Ca(BH4)2, (b) ′-Ca(BH4)2, (c) β-Ca(BH4)2, and
(d) -Ca(BH4)2.
2. α′-Ca(BD4)2:
Ca atoms form a T-shaped arrangement around the B atoms in the α-
and α′-phases (see Figure 6a and b) and a nearly triangular arrangement in
the β-phase (Ca···B···Ca angles 95.5–133.7°). The coordination mode for
the [BH4]⎺ anion with respect to the Ca cations differs correspondingly. In
the α′-phase each Ca atom coordinates two [BH4]⎺ groups via the edge
(Ca…H2B, with Ca–H distances 2.52 Å), and other four via the corner
(Ca…HB, with Ca–H distances 2.02 Å). Coordination via the edge is typical
for alkaline and alkaline-earth borohydrides, such as Be(BH4)2 [69] and
Mg(BH4)2 [7, 88]. The six Ca-B distances in α′-Ca(BH4)2 within range of
2.941–2.976 Å for the α′-phase at 495 K.
positions: Ca(4a): 0, 0, 0; B(8c):0, 0, 0.2250; H1(16e):-0.1608, -
0.035,0.1762; H2(16e): 0.032,-0.1613,0.2752.
3. β-Ca(BD4)2:
The crystal structure of β-Ca(BD4)2 is quite ambiguous. Majzoub and
Rönnebro [102] recently determined the β-polymorph to be best described
in space group P-4 (No. 81) or P42/m (No. 84) by a combined theoretical
and experimental approach, using the Prototype Electrostatic Ground State
method to find the lowest energy structures. A similar conclusion is drawn
from combined experimental and ab initio calculations showning that the
best description of the structure is with the space group P42/m based on
appropriate size/geometry of the [BD4]⎺ tetrahedra, the lowest calculated
formation energy, and real positive vibrational energy, indicating a stable
structure [98]. On the other hand, the Rietveld refinements indicated that the
structure could preferably be described in P-4 [97]. This structure model of
the β-phase was verified by a single-phase sample, obtained by annealing
the α + β mixture at 573 K. At this temperature, the conversion from the α-
to the β-phase is complete. The P-4 model showed slightly better fit than the
P42/m one, but more importantly, more reasonable interatomic distances.
Thus, finally the P-4 structure was confirmed by the Rietveld method with
FullProf [97]. Hence we assume the real β-phase is with P-4 space group
and the second one is assumed as β2-Ca(BH4)2.
32 P. Vajeeston
In the β-phase (P-4; Figure 6c); similarly to the α′-phase, two Ca atoms
coordinate the BH4 groups via edges, and the third one via a tetrahedral
vertex. The experimentally determined Ca–H distances vary from 2.08 to
2.56 Å. The six Ca–B distances in β-Ca(BH4)2 within range of 2.923–2.939
Å for the β-phase at 305 K. The β-phase exhibits more regular octahedra
with respect to the distorted low-symmetry α- and γ-phases. This can be seen
in differences between the shortest and longest B−B interatomic distances
between the BD4 groups surrounding the Ca atoms. These differences are of
11% (3.91−4.366 Å) for the β-phase and 14% (3.807−4.442 Å) to 17%
(3.82−4.618 Å) for the α- and γ-phases.
positions: Ca(2g): 0.5, 0, 0.240, 3.37; B(4h): 0.790, 0.693, 0.259; H1(4h):
0.626, 0.727, 0.217; H2(4h): 0.833, 0.765, 0.491; H3(4h): 0.808, 0.528,
0.262; H4(4h): 0.873, 0.769, 0.062.
4. β2-Ca(BH4)2:
For the structural description please refer above. Cell parameters: a =
6.924 Å; b = 6.924 Å; c = 4.3492 Å; Atomic positions: Ca(2d): 0,1/2, ½;
B(4j): 0.2102, 0.6952, 0; H1(4j): 0.464, 0.68, 0; H2(4j): 0.163,0.548,0;
H3(8k): 0.17000, 0.75500, 0.8090.
5. γ-Ca(BD4)2:
The γ-phase exhibit more distorted octahedra with respect to the more
regular and higher-symmetry β-phase. This can be seen in differences
between the shortest and longest B−B interatomic distances between the
[BD4]⎺groups surrounding the Ca atoms. These differences are of 11%
(3.91−4.366 Å) for the β-phase and 17% (3.82−4.618 Å) for the γ-phase.
The difference between the shortest and the longest interatomic distance
within a tetrahedron is less than 14.5% for the B−D distances (for the two
independent tetrahedra: 1.06−1.24 Å and 1.17−1.33 Å), and less than
10.8% for D−D distances (for the two independent tetrahedra: 1.86−1.95 Å
and 1.90−2.13 Å). Like for the α- and the β-phases of Ca(BD4)2, in the
γ-phase, the Ca atoms are surrounded by six [BD4]⎺groups.

positions: Ca(8c): 0.8634, 0.1275, 0.1240; B1(8c): 0.1740, -0.0033, 0.1572;
B2(8c): 0.5221, 0.2296, 0.1009; H1(8c): 0.218, -0.006, 0.280; H2(8c):
0.267, -0.045, 0.086; H3(8c): 0.170, 0.078, 0.114; H4(8c): 0.039, -0.039,
0.149; H5(8c): 0.520, 0.223, -0.031; H6(8c): 0.648, 0.260, 0.141; H7(8c):
0.415, 0.281, 0.142; H8(8c): 0.505, 0.152, 0.152.
Sr(BH4)2
1. o-Sr(BH4)2:
The structure of Sr(BH4)2 was solved simultaneously and independently
from SR-PXD data measured for the sample prepared by solvent-mediated
synthesis [103]. Strontium borohydride, crystallizes in an orthorhombic unit
cell with space group symmetry Pbcn (see Table 2). The structural drawing
of the strontium borohydride Sr(BH4)2 is shown in Figure 7a. The basic
building unit for the structure of Sr(BH4)2 is the borohydride anion [BH4] ⎺,
which is a regular tetrahedron. This structure is similar to the α-PbO2-type
[104]. The structure is built from half-occupied brucite-like layers of slightly
distorted [Sr(BH4)6] octahedra stacked along the a-axis direction [103]. The
octahedra share two edges and all vertices with eight other octahedra in total.
The [BH4]⎺ units are coordinated by three Sr in a nonplanar triangle. There
are three different strontium–boron distances, each appearing twice within
an octahedron either adjacent or opposite to each other; these are 2.96
(adjacent), 3.058 (opposite), and 3.12 Å (adjacent). These strontium–boron
distances are consistent with the distance of 2.960 Å reported by Bremer et
al. in Sr(BH4)2·2THF [105]. There are 3 B–Sr–B angles within 169.2–173.9°
and 12 within 83.7–97.3°. The Sr–B–Sr angles within the nonplanar [BSr3]
triangle range from 96.3° to 129.9°, and the distance of the B atom from the
Sr plane is 0.36 Å. Four [BH4]⎺ units coordinating via the tetrahedral edge
and two via the tetrahedral face. Thus, Sr is coordinate to 14 hydrogen atoms.
positions: Sr(4ca):0.5, 0.3495, 0.25; B(8d):0.7751, 0.1313, 0.0796; H1(8d):
34 P. Vajeeston
0.8264, 0.2458, 0.0188; H2(8d):0.6461, 0.1556, 0.1648; H3(8d):0.7353,

0.0452, -0.0284; H4(8d):0.8928, 0.0787, 0.1631.
Figure 7. Crystal structures of (a) o-Sr(BH4)2 [isostructural with O1-Ba(BH4)2], (b) t-

Sr/Ba(BH4)2, (c) c-Sr/Ba(BH4)2, and (d) O2-Ba(BH4)2.
2. t-Sr(BH4)2:
Upon heating the - polymorph of Sr(BH4)2 transformed into the
tetragonal t-Sr(BH4)2 polymorph increasing its coordination number from 6
to 8. It then transformed at 445°C into the P41212 polymorph [106]. The
latter is a non-centrosymmetric superstructure (doubled along the c-axis) of
the ht-ZrO2 type (P42/nmc), a deformation of the cubic CaF2 type which is
stabilized by even higher temperatures in ZrO2. The Sr atoms in this
polymorph are coordinated to eight [BH4]⎺ (Figure 7b) forming a cube, and
each borohydride is surrounded by four Sr atoms in a tetrahedral
coordination. Sr–B distances are between 3.0911 and 3.1424 Å.
positions: Sr(4a): 0.06484, 0.06484, 0; B(8b): 0.9401, 0.4249, 0.3648;
H1(8b): 0.0998, 0.4164, 0.4199; H2(8b): 0.959, 0.5845, 0.3101; H3(8b):
0.9311, 0.255, 0.3159; H4(8b): 0.7705, 0.4436, 0.4133.
3. c-Sr(BH4)2:
At temperatures above 450°C, t-Sr(BH4)2 transformed into the cubic
CaF2 type polymorph (c-Sr(BH4)2) [106]. Wtihin this structure the [BH4]⎺
units are embedded into the face centered Sr lattice (see Figure 7c) and the
cation coordination number increases from 6 to 8. The calculated Sr–B, B–
H, and Sr–H distances are 3.078, 1.13, and 2.904 Å, respectively.
positions: Sr(4a): 0, 0, 0; B(8c): 0.25, 0.25, 0.25; H(32f): 0.1583, 0.3419,
0.1583.
Ba(BH4)2
1. o1-Ba(BH4)2:
Ba(BH4)2 crystalize in orthorhombic Pnnm (o1-Ba(BH4)2) and it
transforms at 395°C into the o2-Ba(BH4)2 polymorph, isostructural to the
orthorhombic modification in Eu- and Sr-borohydrides [106]. At 445°C it
further transforms into the tetragonal t-Ba(BH4)2 polymorph. At 460°C the
cubic polymorph also appears. On cooling down to RT the tetragonal and
cubic polymorphs with 8-fold coordination of Ba2+ do not transform to the
orthorhombic polymorphs with octahedral coordination, due to the slow
kinetics as the cooling was rather fast (20°C/min).
positions: Ba(2a): 0, 0, 0; B(4g): 0.1308, 0.2460, 3/6; H1(4g): 9582, 0.2160,
3/6; H2(4g): 0.2204, 0.0998, 1/2; H3(8h): 0.1689, 0.3344, 0.2913.
36 P. Vajeeston
2. o2-Ba(BH4)2:
In the o2-Ba(BH4)2 polymorph the octahedra share all six vertices and
two edges, thus building zigzag chains along the c-axis. Each chain is again
linked by vertices to four other chains. The borohydride tetrahedron is
surrounded by three Ba atoms and forms a nearly perfect trigonal planar
coordination. The Ba–B distances in this polymorph are between 3.127 and
3.289 Å. This structure type is the room temperature polymorph of Eu(BH4)2
and Sr(BH4)2 [106].
positions: Ba(2a): 0.5,0.25, 0.1622; B(4g): 0.7617, 0.0768, 0.3906; H1:
0.7949, 0.1644, 0.4864; H2: 0.8525, -0.0352, 0.3976; H3: 0.6158, 0.0367,
0.4000; H4: 0.7835, 0.1414, 0.2784.
3. t-Ba(BH4)2:
Upon heating the - polymorph of o2-Ba(BH4)2 transformed into the
tetragonal t-Ba(BH4)2 polymorph, thus increasing its coordination number
from 6 to 8. It further ansformed at 718K into the P41212 polymorph [106].
This structure is a non-centrosymmetric superstructure (doubled along the
c-axis) of ht-ZrO2 type (P42/nmc), a deformation of the cubic CaF2 type
which is stabilized at even higher temperatures in ZrO2. The Ba atoms are
coordinated in this polymorph by eight [BH4]⎺ (Figure 7b) forming a cube,
and each borohydride is surrounded by four Ba atoms in a tetrahedral
coordination. Ba–B distances are between 3.052 and 3.499 Å [106].
positions: Ba(4a): 0.0282, 0.0279, 0; B(8b): 0.9401, 0.4249, 0.3648; H1(8b):
0.0998, 0.4164, 0.4199; H2(8b): 0.9590, 0.5845, 0.3101; H3(8b): 0.9311,
0.2550, 0.3159; H4(8b): 0.7705, 0.4436, 0.4133.
4. c-Ba(BH4)2:
At higher temperatures above 773K, t-Ba(BH4)2 transformed into the
cubic CaF2 type polymorph (c-Ba(BH4)2). In this structure [BH4]⎺ units are
embedded into the face centered Sr lattice (see Figure 7c) and the cation
coordination number increases from 6 to 8 with the temperature. The
calculated B-H, and Ba-H distances are 1.13, and 1.941 Å, respectively.

positions: Ba(4a):0, 0, 0; B(8c):0.25, 0.25, 0.25; H(96k):0.1578, 0.3423,
0.3415.
ALKALI METAL MIXED TETRA BORIDES
Most of the following alkali metal mixed tetra borides are characterized
with the help of XPD experiments for structure determination, and the final
structural refinement proceeds via first principles solid-state calculations,
that allowed us to identify minute structural details, such as the hydrogen
coordination mode in all compounds.
LiK(BH4)2
Nickels et al. report [107] the first synthesis and crystal structure
determination of a mixed alkali metal borohydride, LiK(BH4)2 from 1:1 ratio
mixture of LiBH4 and KBH4. Importantly in this new material, the observed
decomposition temperature lies between that of the constituent phases.
LiK(BH4)2 crystalize in space group Pnma with lattice parameters a = 7.9134
Å, b = 4.4907 Å, and c = 13.8440 Å (see Figure 8a). The b-axis lattice
parameter is very similar to that of orthorhombic LiBH4 (4.4368 Å [43]),
suggesting a degree of structural similarity between these phases. The
[BH4]⎺ units in LiK(BH4)2 form an approximately tetrahedral coordination
around the lithium ion, which is similar to that found in orthorhombic LiBH4.
The Li-B bonds are more elongated in LiK(BH4)2 than in LiBH4 but with a
narrower range of angles. The larger Li⋅⋅⋅B separations observed in the new
phase may originate from the presence of potassium cations in the structure,
which are considerably larger than their lithium counterparts (Li+ ionic
radius 0.59 Å, K+ ionic radius 1.38 Å in tetrahedral coordination). The
arrangement of the [BH4]⎺ units in LiK(BH4)2 and KBH4 differs
considerably. KBH4 has an octahedral arrangement of [BH4]⎺ units, whereas
those in LiK(BH4)2 might be best described as monocapped trigonal prisms.
The K⋅⋅⋅B distances in KBH4 are 3.364 Å, whereas in LiK(BH4)2 the (seven)
distances are 3.404 Å (twice), 3.409 Å (twice), 3.431 Å (twice), and 3.475
38 P. Vajeeston
Å (once). It is thought that these larger separations arise because of the

greater number of [BH4]⎺ units present around the potassium cation.
positions: K(4c): 0.3381, 0.25, 0.86851; Li(4c): 0.1885,0.25,0.1349; B(4c):
0.2508,0.25,0.5460; B2(4c): 0.9292, 0.25, 0.2397; H1(8d):
0.324,0.456,0.5288; H2(4c): 0.206, 0.25, 0.614; H3(4c): 0.140, 0.25, 0.498;
H4(8d): 0.976, 0.452, 0.2057; H5(4c): 0.789, 0.25, 0.218; H6(4c): 0.960,
0.25, 0.316.
Figure 8. Crystal structures of (a) LiK(BH4)2; (b) m-LiRb(BH4)2; (c) o-LiRb(BH4)2 and
(d) Li2Rb(BH4)3 (isostructural with Li2Cs(BH4)3).
LiRb(BH4)2
1. m-LiRb(BH4)2:
The formation of monoclinic m-LiRb(BH4)2 is observed in the ball-
milled 1:1 ratio mixture of LiBH4 and RbBH4, together with unreacted
RbBH4 [108]. In the m-LiRb(BH4)2 structure, (4,4)-connected chains
[Li(BH4)2] are made from two 4-rings which share one Li in a square
coplanar coordination, the other Li atoms are in a tetrahedral [BH4]⎺
environment (see Figure 8b). Such chains do not occur in several other
oxides. In particular, the rare square planar coordination is observed in
intercalated lithium ion electrode materials [109]. The same environment is
found in the perovskite-type ionic conductor La0.5Li0.5TiO3 [110].
Substituting for the small Li+ here creates A-site vacancies which account
for the high mobility, while Li itself occupies a square planar position face-
centred on the cubic unit cell. This phase might be a potential candidate for
the Li-ion battery as a solid electrode.
Cell parameters: a = 23.5991 Å; b = 4.6178 Å; c = 17.893 Å; β=97.776o;
Atomic positions: Rb1(2c):0, 0, 0.5; Rb2(4i):0.2862, 0, 0.3584; Rb3(4i):
0.0684,0, 0.2099; Rb4(4i): 0.35229,0.5, 0.0686; Li1(2b): 0, 0.5, 0; Li2(4i):
0.435, 0, 0.264; Li3(4i): 0.203, 0.5, 0.144; Li4(4i): 0.163, 0.5, 0.429; B1(4i):
0.2495, 0, 0.1477; B2(4i): 0.1673, 0, 0.4844; B3(4i): 0.4706, 0, 0.1492;
B4(4i): 0.0314, 0.5, 0.3878; B5(4i): 0.4014,0.5, 0.3164; B6(4i): 0.1910, 0.5,
0.2985; B7(4i): 0.1107, 0.5, 0.083; H1(4i): 0.214,0, 0.185; H2(4i): 0.231, 0,
0.084; H3(8j): 0.2769,-0.199,0.161; H4(4i): 0.140, 0, 0.425; H5(4i): 0.214,
0, 0.477; H6(8j): 0.157, -0.199, 0.518; H7(4i): 0.5186, 0, 0.166; H8(8j):
0.452, -0.199, 0.174; H9(4i): 0.001, 0.5, 0.434; H10(4j): 0.006, 0.5, 0.327;
H11(8j): 0.0594, 0.301, 0.395; H12(4i): 0.365, 0.5, 0.266; H13(4i): 0.444,
0.5, 0.292; H14(8j): 0.399,0.301, 0.354; H15(4i): 0.145, 0.5, 0.269;
H16(4i): 0.221, 0.5, 0.252; H17(8j): 0.199,0.301, 0.3366; H18(4i):
0.067,0.5, 0.104; H19(4i): 0.105, 0.5, 0.017; H20(8j): 0.135, 0.301, 0.106;
H21(4i): 0.460, 0, 0.083.
40 P. Vajeeston
2. o-LiRb(BH4)2:
The monoclinic polymorph m-LiRb(BH4)2 transforms to o-LiRb(BH4)2
at 403 K, which then melts at 414 K. Recrystallization takes place upon
cooling at 409 K, and the high-temperature phase transforms back to m-
LiRb(BH4)2 at 375 K, and it is stable at room temperature [108]. The
hysteresis of 28 K of the m–o transformation is compliant with a first-order
character of the transformation in LiRb(BH4)2. In this structure puckered
[Li(BH4)2] ⎺ sheets built from vertex-sharing tetrahedra of both Rb+ counter-
cations (Figure 8c).

positions: Li(4a): 0, 0.10471, ¼; Rb(4a): 0, 0.3684, 0.272; B1(4a): 0,
0.3124, 0.692; B2(4a): 0, 0.0103, 0.013; H1(8b): -0.2026, 0.3141, 0.782;
H2(4a): 0, 0.3702, 0.612; H3(4a): 0, 0.2512, 0.622; H4(8b): -0.2026,
0.0423, 0.073; H5(4a): 0, 0.0183, -0.133; H6(4a): 0, -0.0614, 0.053.
3. h-Li2Rb(BH4)3
The hexagonal phase Li2Rb(BH4)3 is observed in the ball-milled mixture
of LiBH4 and RbBH4 with 2:1 ratio [108]. The hexagonal phase Li2Rb(BH4)3
is stable up to 392 K where it melts and this is the stable polymorph of
Li2Rb(BH4)3 at room temperature. In this phase each Rb+ is surrounded by 8
H atoms (4 with 2.87Å and 4 with 2.2913 Å; Figure 8d) and Li is surrounded
by 6 H atoms with the distance vary from 1.864 to 2.256 Å.
Cell parameters: a = 7.474; b = 7.474Å; c = 11.585 Å; Atomic positions:
Rb(3c):0.5, 0.5, 1/3; Li(6i):0.85588, 1.7118, 0; B1(3b):0, 0, 0.5; B2(6f):0.5,
0, 0.2943; H1(12k):-0.03678, 0.10050, 0.55574; H2(12k):0.6366, 0.1022,
0.2386; H3(12k):0.5391,-0.0988, 0.3501.
4. o-LiCs(BH4)2
The nature of the system Li–Cs–BH4 is much more complex and
presumably subject to various metastable phase equilibria [108]. Schouwink
et al. studied the thermal phase stabilities of Li–Cs-BH4 phases and are
determined from in-situ powder diffraction are schematized for three
different samples LiBH4 and CsBH4 with 1:1, 2:1 and 3:1 ratios. LiBH4 and
CsBH4 with 4:1 mixture, produced a single phase sample of the most Li-rich
phase Li3Cs(BH4)4 [108]. This phase is already a major phase in the mixture
3:1. Hence, this is the only of the stable phase in the Li-Cs-BH4 system.
LiCs(BH4)2 crystalize in orthorhombic Cmc21 space group [o-LiCs(BH4)2].
Similar to the o-LiRb(BH4)2 phase this structure also have puckered
[Li(BH4)2]– sheets built from vertex-sharing tetrahedra of both Cs+ counter-
cations (see Figure 9a).
Figure 9. Crystal structures of (a) LiCs(BH4)2; (b) Li2Cs(BH4)3; (c) Li3Cs(BH4)4; (d)
LiCs2(BH4)3; and (e) Li2Cs3(BH4)5 and (f) Li3Cs2(BH4)5.

positions: Li(4a): 0.5, 0.093, 0.2714; Cs(4a): 0, 0.1374, 0.7550; B1(4a): 0,
0.1816, 0.2825; B2(4a): 0.5, -0.0046, 0.5039; H1(8b): 0.2126, 0.1904,
0.3720; H2(4a): 0, 0.1113, 0.2216; H3(4a): 0, 0.2339, 0.1698; H4(8d):
0.2843, -0.0330, 0.5683; H5(4a):0.5,-0.0233, 0.3544; H6(4a): 0.5, 0.0717,
0.5207.
42 P. Vajeeston
5. h-Li2Cs(BH4)3
Li2Cs(BH4)3 crystalize in hexagonal [h-Li2Cs(BH4)3; P6222] and
monoclinic [m-Li2Cs(BH4)3; Cc] form (see Figure 8d and Figure 9b) [108].
The 3-connected triangular [Li2(BH4)3] – sheet in m-Li2Cs(BH4)3 has the
topology of a graphene-like Li sheet expanded by borohydride linkers.
Topologically this is called a honeycomb layer. In m-Li2Cs(BH4)3 the
triangular sheets are not linked by layers of larger coordination polyhedra.
Here the connectivity between sheets is ensured by Cs+ cations occupying
sites within the plane defined by the [Li2(BH4)3]– sheet. The symmetry of the
honeycomb layer [Li2(BH4)3] – is lowered over an orthorhombic c-centered
cell down to the monoclinic Cc, monoclinic angle β = 90o, by the rotation
and translation of [BH4]⎺ groups. Strong pseudosymmetry of the sheet in m-
Li2Cs(BH4)3 as a vertical mirror plane passing through Li atoms and leading
to the orthorhombic supergroup Cmc21. In m-Li2Cs(BH4)3 all BH4 groups
are linear linkers, bridging two Li atoms. This situation does not allow for a
general Li–H coordination scheme where all H–H distances are maximized.
Therefore, the monoclinic distortion has its origin in different rotations of
the three H–H edges from the vertical direction, and two symmetrically
independent Li(BH4)3 triangles. The [Li2(BH4)3]– net may be rationalized as
a frustrated net with respect to the ideal conrotation of the three vertical H–
H edges. Such a situation may promote specific dynamics of the [BH4]⎺
groups, such as 180o jump motions around the twofold operation of the 4
axes.
Cell parameters: a = 7.6634; b = 7.6634; c = 12.1279; Atomic positions:
Cs(3c): 0.5, 0.5, 1/3; Li(6i): 0.8614, 0.723, 0; B1(3b): 0, 0, 0.5; B2(6I): 0.5,
0, 0.3076; H1(12k): -0.0363, 0.099, 0.5542; H2(12k): 0.602, 0.134,0.2534;
H3(12k): 0.596, -0.041, 0.3618.
6. m-Li2Cs(BH4)3:
Cell parameters: a = 7.574; b = 7.948; c = 12.880; β = 89.97o; Atomic
positions: Cs(4a): 0.5, 0.3418, 0.83210; Li1(4a): 0.062, 0.657, 0.135;
Li2(4a): 0.07, 0.246, -0.013; B1(4a): 0.811, 0.600, 0.042; B2(4a): 0.360,
0.694, 0.114; B3(4a): 0.561, 0.794, 0.811; H1(4a): 0.88, 0.58, -0.038;
H2(4a): 0.88, 0.52, 0.108; H3(4a): 0.66, 0.56, 0.036; H4(4a): 0.82, 0.745,
0.062; H5(4a): 0.27, 0.75, 0.178; H6(4a): 0.508, 0.71, 0.141; H6(4a): 0.34,
0.77, 0.035; H7(4a): 0.33, 0.550, 0.10; H8(4a): 0.52, 0.70, 0.747; H9(4a):
0.44, 0.82, 0.866; H10(4a): 0.61, 0.92, 0.771; H11(4a): 0.68, 0.74, 0.861.
7. m-LiCs2(BH4)3
LiCs2(BH4)3 is prepared mechano-chemically by ball-milling and
crystalize in monoclinic C2/c symmetry (m-LiCs2(BH4)3; see Figure 9d)
[108]. In this structure isolated chains [Li(BH4)3]2- of tetrahedra sharing two
vertices are found in m-LiCs2(BH4)3. The vertex-sharing tetrahedral chain
[Li(BH4)3]2- in m-LiCs2(BH4)3 is an analogue of the 2D-periodic (zweier)
chain underlying the architecture of inosilicates. This chains [Li(BH4)3]2- of
tetrahedra sharing two vertices. The average H···H distances is 2.1315 Å.
The Cs-H bond lengths vary from 2.654 to 3.48Å; Cs–B varies from 3.59 to
3.80Å.
Cell parameters: a = 7.6021; b = 15.160; c = 17.243; β = 92.693o;
Atomic positions: Cs1(8f): 0.2444, 0.3917, 0.0689; Cs2(8f): 0.7563, 0.2004,
0.1495; Li1(8f): 0.2422, 0.4902, 0.3285; B1(4e): 0.0, 0.0062, 0.25; B2(8f):
0.7453, 0.4078, 0.0448; B3(8f): 0.3060, 0.1763, 0.1324; B4(4e): 0.0, 0.4250,
0.25; H1(8f): 0.1299, 0.0530, 0.2563; H2(8f): 0.0065, -0.0409, 0.1928;
H3(8f): 0.8418, 0.3462, 0.0296; H4(8f): 0.8366, 0.4727, 0.0612; H5(8f):
0.6582, 0.3879, 0.1002; H6(8f): 0.6448, 0.4247, -0.0107; H7(8f): 0.3883,
0.2007, 0.0768; H8(8f): 0.3188, 0.2333, 0.1828; H9(8f): 0.3742, 0.1081,
0.1579; H10(8f): 0.1516, 0.1644, 0.1118; H11(8f): 0.1135, 0.4720, 0.2238;
H12(8f): -0.0622, 0.3785, 0.1978.
8. o-Li3Cs2(BH4)5
Li3Cs2(BH4)5 is prepared mechano-chemically by ball-milling and
crystalizes in orthorhombic (Fddd) symmetry [o-Li3Cs2(BH4)5; see Figure
9f] [108]. O-Li3Cs2(BH4)5 contains (4,3)-connected tetrahedral–triangular Li
sheets. The phase O-Li3Cs2(BH4)5, built from the sheets of [Li3(BH4)5]2-
containing tetrahedra and triangles in the ratio 1:1, and sharing all vertices
[108]. No other phase with this composition and topology was observed for
Rb containing borohydrides. The (4,3)-connected tetrahedral–triangular
sheet in o-Li3Cs2(BH4)5 may be compared to the tetrahedral–triangular
44 P. Vajeeston
framework in LiZn2(BH4)5 [111] and Li3MZn5(BH4)15 [112]; the first

compound contains tetrahedral Li and triangular Zn, the second contains
tetrahedral Li, Mg and triangular Li, Zn. Mixed triangular and tetrahedral
coordination may be found among the borates. Unlike silicon and carbon,
boron has the ability to bind to either three or four O atoms to form a BO3
triangle or a BO4 tetrahedron. The same ability is observed in this compound
for Li with the borohydride ligand. In this phase the average H···H distances
is 1.95 Å. The Cs–H bond lengths vary from 2.538 to 3.406 Å and Cs–B
distance varies from 3.685 to 3.928Å.
Cell parameters: a = 13.5838 Å, b = 17.3001 Å, c = 23.9151 Å; Atomic
positions: Cs1(16g):0.375, 0.375, 0.72791; Cs2(16g):0.625, 0.625, 0.50563;
Li1(16f):0.125, 0.824, 0.125; Li2(16f):0.125, 0.266, 0.125; Li3(16e):0.407,
0.125, 0.125; B1(16e):0.006, 0.625, 0.625; B2(32h):0.628, 0.9093, 0.7209;
B3(32h):0.783, 0.997, 0.1371; H1(32h):-0.042, 0.613, 0.663;
H2(32h):0.055, 0.573, 0.617; H3(32h):0.594, 0.859, 0.697; H4(32h):0.579,
0.962, 0.717; H5(32h):0.704, 0.923, 0.703; H6(32h):0.635, 0.893, 0.767;
H7(32h):0.710, 1.00, 0.160; H8(32h):0.775, 1.024, 0.094; H9(32h):0.81,
0.934, 0.132; H10(32h):0.84, 1.03, 0.162.
9. m-Li2Cs3(BH4)5
Li2Cs3(BH4)5 is prepared mechano-chemically by ball-milling and
crystalize in monoclinic (P21/c) symmetry [m-Li2Cs3(BH4)5; see Figure 9e]
[108]. Increasing the Li content leads to the phase m-Li2Cs3(BH4)5
containing isolated chains [Li2(BH4)5]3- of tetrahedra sharing one edge and
one vertex. The edge- and vertex-sharing tetrahedral chain [Li2(BH4)5]3- in
m- Li2Cs3(BH4)5 does not find a match amongst inorganic compounds. In
this phase the average H···H distances is 1.89 Å. The Cs–H bond lengths
vary from 2.363 to 3.152 Å and Cs–B distance varies from 3.211 to 3.699
Å.
Cell parameters: a = 14.444 Å, b = 7.4218 Å, c = 14.7907 Å, β=104.470;
Atomic positions: Cs1(4e):0.2691, 0.747, 0.6341; Cs2(4e):0.0016, 0.784,
0.8502; Cs3(4e):0.5020, 0.248, 0.0988; Li1(4e):0.79, 0.75, 1.02;
Li2(4e):0.68, 0.83, 0.20; B1(4e):0.759, 0.68, 0.350; B2(4e):0.504, 0.85,
0.156; B3(4e):0.73, 0.55, 0.13; B4(4e):0.75, 0.03, 0.095; B5(4e):0.041,
0.26, 0.874; H1(4e):0.69, 0.63, 0.354; H2(4e):0.815, 0.64, 0.41;

H3(4e):0.757, 0.84, 0.346; H4(4e):0.778, 0.63, 0.29; H5(4e):0.51, 0.73,
0.11; H6(4e):0.52, 0.98, 0.12; H7(4e):0.56, 0.84, 0.22; H8(4e):0.43, 0.85,
0.17; H9(4e):0.71, 0.49, 0.05; H10(4e):0.66, 0.54, 0.16; H11(4e):0.75, 0.70,
0.12; H12(4e):0.79,0.47, 0.17; H13(4e):0.75, 0.17, 0.11; H14(4e):0.77,
0.02, 0.03; H15(4e):0.68, -0.0, 0.09; H16(4e):0.81, -0.0, 0.15; H17(4e):-
0.04, 0.27, 0.837; H18(4e):0.066, 0.38, 0.915; H19(4e):0.084, 0.24, 0.821;
H20(4e):0.052, 0.14, 0.923.
10. m-Li3Cs(BH4)3
Li3Cs(BH4)3 is prepared mechano-chemically by ball-milling of LiBH4
and CsBH4 with 4:1 ratio, which produced a single phase sample of the most
Li-rich phase of Li3Cs(BH4)4 [crystalize in monoclinic (P21/c) symmetry m-
Li3Cs(BH4)3; see Figure 9c] in the Li-Cs-BH4 mixture [108]. This phase is
already a major phase in the 3LiBH4:1CsBH4 mixture. This implies that the
maximum Li content in the phase diagram of LiBH4–CsBH4 is reached with
this phase. The framework [Li3(BH4)4]⎺ observed in the most Li-rich
compound m-Li3Cs(BH4)4 can be derived from the o-LiBH4 framework by
replacing a quarter of the Li atoms by Cs, thus breaking the infinite stripes
of edge-sharing tetrahedra into fragments of three tetrahedra. In this phase
the average H···H distances is 1.89 Å. The Cs–H bond lengths vary from
2.910 to 3.647 Å and Cs–B distance varies from 3.71 to 3.929 Å.
Cell parameters: a = 8.1992 Å, b = 12.3044 Å, c = 11.9009 Å,
β=123.767; Atomic positions: Cs(4e):0.1766, 0.17545, 0.1542;
Li1(4e):0.297, 0.927, 0.485; Li2(4e):0.718, 0.333, 0.276; Li3(4e):0.691,
1.010, 0.185; B1(4e):0.992,0.403, 0.261; B2(4e):0.654, 0.0902, 0.955;
B3(4e):0.568, 0.6284, 0.956; B4(4e):0.226, 0.8667, 0.677; H1(4e):0.976,
0.475, 0.196; H2(4e):0.136, 0.361, 0.299; H3(4e):0.867, 0.344, 0.200;
H4(4e):0.991, 0.433, 0.350; H5(4e):0.713, 0.023, 0.922; H6(4e):0.599,
0.159, 0.880; H6(4e):0.774, 0.120, 0.058; H7(4e):0.529, 0.057, 0.959;
H8(4e):0.51, 0.67, 0.01; H9(4e):0.54, 0.68, 0.87; H10(4e):0.49, 0.55, 0.92;
H11(4e):0.73, 0.61, 0.03; H12(4e):0.241, 0.935, 0.621; H13(4e): 0.373,
0.850, 0.774; H14(4e):0.17, 0.791, 0.612; H15(4e):0.116, 0.890, 0.702.
46 P. Vajeeston
ALKALI AND ALKALINE EARTH BASED MIXED

TETRA BORIDES
Metal borohydrides have been obtained by synthesis approaches that

are hard to control and the often rather complex new materials have little
potential for structural engineering and optimization of useful properties.
Rational structure-property design is most readily accomplished in very
simple and stable structure types, which are capable of accommodating a
wide range of different elemental compositions. Recently, Schouwink et
al. [113] identified the first perovskite-type metal borohydride,
KMn(BH4)3, during thermal decomposition of K2Mn(BH4)4. Indeed, metal
borohydrides usually crystallize with structures closely related to
isoelectronic metal oxides. For instance, the polymorphs of Ca(BH 4)2 are
structurally related to the different modifications of TiO2 [114]. New
facile and easily scalable synthesis strategies, using a combination of
solvent-based methods and mechano-chemical addition reactions, are
developed for the synthesis of homoleptic perovskite-type metal
borohydrides, whereas mechano-chemical metathesis reactions using
metal halides provide anion-substituted metal borohydride-halides [115].
Most of the alkali and alkaline earth based mixed tertra borides are made
using the above mentioned methods.
A series of complex hydrides (based on the tetrahydroborate anion
[BH4]⎺ and crystallizing in the ABX3 type lattice was reported very recently
[115], presenting the first genuine tunable host lattice within iono-covalent
hydrides and drawing interest not only for solid state hydrogen storage but
also physical properties related to heavy metals and lattice instabilities. It
was found, that, at the structural level, mixedmetal borohydride perovskites
undergo polymorphic transformations, where the HT polymorph has lower
space group symmetry relative to the LT polymorph.
K2Mg(BH4)4
K2Mg(BH4)4 is prepared mechano-chemically by ball-milling [113]. At

room temperature, the crystal structure of K2Mg(BH4)4 (P21/n) is a
monoclinically distorted derivative of β-K2SO4, with Mg on the S position
and [BH4]⎺ groups on the O positions, all occupying general position 4e.
Similar observations are made for new compounds found in the systems K–
M–BH4 (M = Zn or Cd) [111, 116]. The most prominent building block is
the complex tetrahedral anion [Mg(BH4)4]2– with bidentate Mg···H2B
bonding. The environment of [Mb(BH4)4]2– described by its nearest
neighbors is a pentacapped trigonal prism, formed by a trigonal prism
(monoclinically distorted) with its 3-fold axis lying parallel to c, and a
strongly distorted equatorial plane of 5 K atoms. The inner B–H distances
are longer than the outer ones, while the inner H–B–H angle of the Mg–H2B
fragment is smaller than that of the opposing K–H2B fragment, yet with both
being larger than the angles of the noncoordinating H2B edges.
Cell parameters: a = 8.1791 Å; b = 9.8842 Å; c = 12.7492 Å; β =
100.735o; Atomic positions: K1(4e): 0.2188, 0.8541, 0.4473; K2(4e):
0.8055, 0.5003, 0.3001; Mg(4e): 0.774, 0.2180, 0.9125; B1(4e): 0.9974,
0.7822, 0.1923; B2(4e): 0.4943, 0.6642, 0.1432; B3(4e): 0.792, 0.4641,
0.5771; B4(4e): 0.396, 0.1522, 0.5741; H1(4e): 0.039, 0.8459, 0.1284;
H2(4e): 0.00, 0.853, 0.261; H3(4e): 0.08, 0.692, 0.211; H4(4e): 0.87, 0.741,
0.161; H5(4e): 0.446, 0.5583, 0.151; H6(4e): 0.63, 0.671, 0.1725; H7(4e):
0.4335, 0.7335, 0.1948; H8(4e): 0.4661, 0.7007, 0.061; H9(4e): 0.821,
0.5049, 0.501; H10(4e): 0.7976, 0.5508, 0.6367; H11(4e): 0.8899, 0.3866,
0.6104; H12(4e): 0.6654, 0.4166, 0.5646; H13(4e): 0.2684, 0.1268, 0.5883;
H14(4e): 0.4917, 0.1259, 0.6476; H15(4e): 0.4039, 0.2652, 0.561; H16(4e):
0.4236, 0.0946, 0.5038.
48 P. Vajeeston
Figure 10. Crystal structures of (a) K2Mg(BH4)4; (b) LT-K3Mg(BH4)5 (c) and HT-
K3Mg(BH4)5. The partial occupancy was marrkes as duel colour.
K3Mg(BH4)5
1. LT-K3Mg(BH4)5:
Similarly to K2Mg(BH4)4, K3Mg(BH4)5 is also prepared mechano-
chemically by ball-milling [113]. The structure of K3Mg(BH4)5 is tetragonal
(P42bc) and has an intriguing nature that can be described as an intergrowth of
KBH4 and K2Mg(BH4)4. Structure of K3Mg(BH4)5 is closely related to
Cs3CoCl5, which crystallizes in space group I4/mcm. The loss of body
centering for K3Mg(BH4)5 is probably caused by the non-spherical nature of
[BH4]⎺ occupying the chlorine positions of the prototype Cs3CoCl5.
Interestingly, the [BH4]⎺ group occurs in two different environments,
forming similar complex anions [Mg(BH4)4]2– as observed in the structure
of K2Mg(BH4)4, on one hand, while the second boron position only
coordinates to potassium. This [BH4]⎺ is octahedrally coordinated, strongly
resembling the local environment of [BH4]⎺ in KBH4. The B–Mg–B angle
with its bisector along c is 96.64° and the average H···H distances is 2.1315
Å. Average bond lengths Mg–B of the complex anion is 2.4369 Å.
positions: K1(8h): 0.32714, 0.67704, 0; K2(8f): 0, 0.5, 0.72739 (50%
occupancy); Mg(4b): 0, 0, 0.75; B1(16i): 0.1181, 0.835, 0.1484; B2(8f): 0,
0.5, 0.482 (50% occupancy); H1(16i): 0.225, 0.76, 0.16; H2(16i): 0.021,
0.796, 0.192; H3(16i): 0.08,0.823,0.08; H4(16i): 0.146, 0.96, 0.16; H5(16i):

-0.104, 0.473, 0.525 (50% occupancy); H6(16i): -0.027, 0.604, 0.440 (50%
occupancy).
It should be noted that LT-K3Mg(BH4)5 can also be described as I4/mcm.
Both of these structures are very close to each other.
Space group: I4/mcm; Cell parameters: a = 9.09783 Å, b = 9.09783 Å,
c = 15.9456 Å; Atomic positions: K1(8h): 0.67840,1.1784, 0; K2(8f): 0, 0,
0.2229 (50% occupancy); Mg(4b): 0, 0.5, 0.25; B1(16l): 0.138, 0.638,
0.148; B2(8f): 0, 0, 0.101 (50% occupancy); H1(32m): 0.0140, 0.663, 0.16;
H2(16l): 0.16,0.66, 0.077; H3(16l): 0.21, 0.71, 0.190; H4(32m): -0.1053, 0,
0.143 (25% occupancy); H5(32m): 0, 0.1053, 0.0581 (25% occupancy).
2. HT-K3Mg(BH4)5:
At higher temperature (at 367K) tetragonal K3Mg(BH4)5 (P42bc)
transform into another tetragonal high symetric P42/mbc phase. The B–Mg–
B angle with its bisector along c is 95.02° and the average H···H distances
is 2.1 Å. Average bond lengths Mg–B of the complex anion is 2.4079 Å.
positions: K1(8h):0.3271, 0.6770,0; K2(8f):0, 0.5, 0.7274; Mg(4b):0, 0,
0.75; B1(16i):0.1181, 0.8351,0.1484; B2(8f):0, 0.5, 0.4821; H1(16i):0.2259,
0.761, 0.161; H2(16i):0.021, 0.796,,0.1928; H3(16i):0.081, 0.8235, 0.081;
H4(16i):0.1463, 0.961, 0.161; H5(16i):-0.1042, 0.4739, 0.5251; H6(16i):-
0.0279, 0.6042, 0.4401.
KCa(BH4)3
1. LT-KCa(BH4)3:
At room temperature, KCa(BH4)3 stabilizes in an orthorhombic structure
(LT-KCa(BH4)3; Pba2; BH4 orientationally ordered) and crystallizes with
the octahedral rotation pattern a0a0c+, which is characteristic of
centrosymmetric tetragonal P4/mbm perovskites [117]. In the symbol a0a0c+,
superscripts refer to the sense of rotation and letters to the axis of rotation.
The space group symmetry P4/mbm, however, is lowered because of the
ordering of hydrogen atoms to the polar space group Pba2.
50 P. Vajeeston
2. HT-KCa(BH4)3
Before the temperature-induced phase transition at 340 K, KCa(BH4)3
approaches tetragonal symmetry, following the intuitive structural
evolution of perovskites. The transition, however, is orthorhombic–
orthorhombic, hence without symmetry increase. The unit cell axis c is
doubled in the HT-polymorph (HT-KCa(BH4)3; space group Pbn21),
introducing a further octahedral tilt (a−a−c+), characteristic of the
orthorhombic Pnma perovskites. Space group symmetry is lowered to
Pbn21, again due to hydrogen ordering [115].
The lattice instabilities and structural dynamics properties of LT-/HT-
KCa(BH4)3 described in the literature [115, 118]. But the structural details
(atomic positions) are not reported in the literature including the structural
data bases [119, 120].
3. α-KSr(BH4)3
The KSr(BH4)3 was synthesized via an addition reaction between Sr(BH4)2
and KBH4 by mechanochemical treatment. At RT KSr(BH4)3 crystalizes into
orthorhombic P21cn [121]. The structures of α-KSr(BH4)3 are built from
distorted octahedra of [Sr(BH4)6], which contain tetrahedral borohydride anions,
[BH4]−. Each [Sr(BH4)6] octahedron shares all corners with other [Sr(BH4)6]
octahedra, thereby forming a three-dimensional framework, while BH4 units are
linearly coordinated to two Sr atoms by edge sharing, as imposed by the
perovskite structure type, see Figure 11a. In KSr(BH4)3, the Sr–B distance varies
between 2.710–3.537 Å; the bond angles Sr–B–Sr are vary from 119.001° to
158.792° and B–Sr–B vary from 70.619° to 176.924°.
positions: K(4a): 0.23979, 0.54281,0.96526; Sr(4a): 0.46099, 0.50716,
0.48652; B1(4a): 0.19014, 0.99238, 0.09103; B2(4a): 0.55245, 0.88349,
0.32359; B3(4a): 0.48618, 0.21503, 0.69107; H1(4a): 0.12999, 0.97787,
0.20842; H2(4a): 0.16173, 0.10055, 0.01475; H3(4a): 0.18850, 0.88432,
0.01575; H4(4a): 0.28312, 1.01438, 0.14054; H5(4a): 0.55954, 0.81557,
0.44048; H6(4a): 0.62484, 0.85705, 0.24562; H7(4a): 0.54852, 0.01656,
0.36290; H8(4a): 0.47102, 0.84315, 0.25868; H9(4a): 0.51842, 0.30082,
0.78652; H10(4a): 0.42949, 0.27241, 0.60449; H11(4a): 0.44412, 0.10684,

0.75591; H12(4a): 0.56567, 0.16378, 0.61463.
4. Rb3Mg(BH4)5
The diffraction peaks originating from Rb3Mg(BH4)5 have been indexed
in tetragonal unit cells with an extinction class of I4cm. K3Mg(BH5)5
crystallizes in a similar unit cell, however, in this case a loss of body
centering is indicated by a set of weak Bragg signals observed below 367 K
[122]. This structure has been described as an intergrowth of RbBH4 and
Rb2[Mg(BH4)4]; it contains the [Mg(BH4)4]2⎺ anions (where Mg2+ is
coordinated tetrahedrally by four bidentate [BH4]⎺ groups) and the
disordered [BH4]⎺ anions (coordinated octahedrally by Rb+ cations, like in
some RbBH4 phases). The geometry of the [Mg(BH4)4]2⎺ anions for Rb-
containing compounds closely resembles that found in the K analogue; the
shortest Mg–B distance is 2.434 Å. The equatorial M1–B2 distance is 3.4402
Å and the apical Rb2–B2 distance is significantly larger than the equatorial
one, 3.9983 Å.
Cell parameters: a =9.2996 Å, b = 9.2996; c = 15.993 Å; Atomic
positions: Rb1(8h): 0.67310, 1.17310, 0; Rb2(4a): 0, 0, ¼; Mg(4b): 0, ½,
¼; B1(16l): 0.1400, 0.6400(3), 0.1499; B2(4c): 0, 0, 0; H1(32m): 0.02416,
0.66693, 0.16947; H2(16l): 0.1509; 0.6509; 0.0785; H3(16l): 0.21825;
0.71825; 0.1821; H4(32m): -0.10097(4), 0.00000, 0.0415 (50% occupancy).
5. RbCa(BH4)3
The ionic radius of Rb+ lies in between those of K+ and Cs+, and the
transformation-sequence of RbCa(BH4)3 unifies atomic displacements and
order–disorder mechanisms. Identically to CsCa(BH4)3, the RT-polymorph
of RbCa(BH4)3 is described as cubic Fm-3c (mistakenly described as P-43m
in [115]) [118]. The cubic-orthorhombic transformation at 370 K, results in
a 16-fold supercell metric of HT1-RbCa(BH4)3, space group A2122, which
has no precedence in the literature and it is stable only in a small temperature
range of approximately 30 K (370–400 K). The commensurate modulation
in the HT1-polymorph is attributed to H···H interactions, and is
accompanied by a negative volume anomaly of –1.2%. The subsequent
52 P. Vajeeston
orthorhombic-tetragonal and tetragonal-cubic transformations are again in

agreement with the general temperature-dependency of the lattice type.
Interestingly, it is possible to stabilize the modulated HT1-phase at room
temperature by substituting Ca2+  Mg2+ on the B-site. Supposedly the
smaller size of the Mg2+ cation stabilizes the displacements present in the
modulation of the superstructure [118]. Above 400K HT1-phase transform
into tetragonal I4/mcm phase (HT2- RbCa(BH4)3) [118].
The lattice instabilities and structural dynamics properties of
RbCa(BH4)3 described in the literature [115, 118]. But the structural details
(atomic positions) are not found in the literature including the structural data
bases [119, 120].
6. RbSr(BH4)3
RbSr(BH4)3 wasa synthesized via an addition reaction between Sr(BH4)2
and RbBH4 by mechanochemical treatment [121]. At RT RbSr(BH4)3
crystalizes into orthorhombic Pbn21. The structures of RbSr(BH4)3 are built
from distorted octahedra of [Sr(BH4)6], which contain tetrahedral
borohydride anions, [BH4] ⎺. Each [Sr(BH4)6] octahedron shares all corners
with other [Sr(BH4)6] octahedra, thereby forming a three-dimensional
framework, while [BH4]⎺ units are linearly coordinated to two Sr atoms by
edge sharing, as imposed by the perovskite structure type, see Figure 11b.
In RbSr(BH4)3, the Sr–B distance varies between 2.878–3.165 Å; the bond
angles Sr–B–Sr are vary from 147.141° to 158.107° and B–Sr–B vary from
74.5° to 178.606°.
Cell parameters: a = 8.3341; b = 8.0835; c = 11.6600; Atomic
positions: Rb(4a): 0.5136, 0.0126, 0.7428; Sr(4a): 0, 0,
0;B1(4a):0.3246,0.7129,0.0665; B2(4a): 0.71156, 0.1951, 0.01258; B3(4a):
0.99067, 0.1057, 0.23879; H1(4a): 0.4506, 0.7579, 0.0995; H2(4a): 0.2606,
0.6499, 0.1385; H3(4a): 0.2536, 0.8139, 0.0325; H4(4a): 0.3536, 0.6179,
0.9955; H5(4a): 0.7512, 0.1787, -0.0839; H6(4a): 0.7594, 0.0962,4 0.0662;
H7(4a): 0.7565, 0.3132, 0.0501; H8(4a): 0.5771, 0.1929, 0.0217; H9(4a):
0.9946, -0.0323, 0.2587; H10(4a): 0.9210, 0.1727, 0.30901; H11(4a):
0.1183, 0.1564, 0.2347; H12(4a): 0.9287, 0.1259, 0.15271.
7. CsMg(BH4)3
The CsMg(BH4)3 structure is identical to that found for a nano-
crystalline monoclinic high-pressure polymorph of ReO3 [123] and an
intermediate low-temperature modification of BaTiO3 [124]. The supercell
metric (√6/√2/√2) have been reported for CsMg(BH4)3 and no subcell to the
superstructure is found in the investigated temperature range between 100 K
and its decomposition [118]. In the CsMg(BH4)3 phase the Mg-atom is
shifted off-centre towards a square pyramidal coordination. The lattice
instabilities and structural dynamics properties of CsMg(BH4)3 are described
in the literature [115, 118]. The structural details (atomic positions)
are however not found in the literature including the structural data bases
[119, 120].
8. Cs3Mg(BH4)5
The diffraction peaks originating from Cs3Mg(BH4)5, have been indexed
to a tetragonal unit cells with an extinction class of I4cm [122]. K3Mg(BH5)5
also crystallizes in a similar unit cell, however, in this case a loss of body
centering is indicated by a set of weak Bragg signals observed below 367 K
[113]. The geometry of the [Mg(BH4)4]2⎺ anions for Cs-containing
compounds closely resembles that found in the K and Rb analogue; the
shortest Mg–B distance is 2.410 Å. The equatorial Cs1–B2 distance is
3.6003 Å and the apical Cs2–B2 distance is significantly larger than the
equatorial one, 4.0631 Å.
Cell parameters: a = 9.7115; b = 9.7115(4); c = 16.2540; Atomic
positions: Cs1(8h): 0.6711, 0.1711,0; Cs2(4a): 0, 0, 0.25; Mg(4b):0, 0.5,
0.25; B1(16l): 0.1370,0.6370, 0.1581; B2(4c): 0,0,0; H1(32m): 0.02031,
0.6570, 0.1583; H2(16l): 0.1852, 0.6852, 0.1012; H3(16l): 0.1852, 0.6852,
0.2162; H4(32m): -0.0967, 0.0408, 0.5.
54 P. Vajeeston
CsCa(BH4)3
1. LT-CsCa(BH4)3:
LT-CsCa(BH4)3 crystallizes in space group Fm-3c in the as-milled
samples [118]. This corresponds to the symmetry of the high-temperature
modification of Y(BH4)3 [125]. The high-temperature polymorph is
described as fully disordered Pm-3m, explaining the absence of
superstructure reflections in the low-temperature phase [115]. The ordered
polymorph of CsCa(BH4)3 is an 8-fold supercell of the fully disordered one.
The space-groups Fm-3c and Pm-3m are thus related by doubling of the
cubic cell parameter. The relationship between the F-centred and the
primitive cubic unit cell can account for an identical x-ray diffraction
pattern, in spite of different systematic extinctions, as was the case for
Y(BH4)3. The origin of the unit-cell axis doubling is therefore exclusively
related to the positional ordering of [BH4]⎺ groups, which occupy a
disordered position 4/mm.m (3d) in the Pm-3m HT-phase, and order on
position -4m.2 (24c) in the Fm-3c LT-phase [115, 118].
The lattice instabilities and structural dynamics properties of
CsCa(BH4)3 are described in the literature [115, 118]. The structural details
(atomic positions) are however not found in the literature including the
structural data bases [119, 120].
2. CsSr(BH4)3
CsSr(BH4)3 was synthesized via an addition reaction between Sr(BH4)2
and CsBH4 by mechano-chemical treatment [121]. At RT, CsSr(BH4)3
crystalize into orthorhombic P22121 [121]. The structures of α-CsSr(BH4)3
are built from distorted octahedra of [Sr(BH4)6], which contain tetrahedral
[BH4]⎺ anions, each [Sr(BH4)6] octahedron shares all corners with other
[Sr(BH4)6] octahedra, thereby forming a three-dimensional framework,
while [BH4]⎺ units are linearly coordinated to two Sr atoms by edge sharing,
as imposed by the perovskite structure type, see Figure 11d. In RbSr(BH4)3,
the Sr–B distance varies between 2.84–3.24 Å; the bond angles Sr–B–Sr are
vary from 149.7° to 180° and B–Sr–B vary from 82.4° to 180°. The bond
angles found in CsSr(BH4)3 are more similar to the ones in Sr(BH4)2, while
MSr(BH4)3 (M = K, Rb) presents features from both Sr(BH4)2 and

Sr(BH4)Cl.
Cell parameters: a = 6.0761; b = 8.2068; c = 8.1793; Atomic positions:
Cs(2b): 0.5360, 0.5, 0; Sr(2a): 0.9945, 0, 0; B1(4c): 0.0586,0.6968, 0.1838;
B2(2a): 0.47315, 0.5, 0.5; H1(4c): 0.1226,0.6518, 0.0608; H2(4c):
0.0496,0.8358, 0.1868; H3(4c): 0.8946,0.6338, 0.2068; H4(4c):
0.1786,0.6608, 0.2848; H5(4c): 0.5891, 0.4086, 0.44063; H6(4c): 0.3672,
0.5548, 0.39867.
Figure 11. Crystal structures of (a) KSr(BH4)3, (b) RbSr(BH4)3, (c)Cs3Mg(BH4)5

(isostructural with Rb3Mg(BH4)5), and (d) α-CsSr(BH4)3.
3. LiKMg(BH4)4
LiKMg(BH4)4 is built up of single-layered sheets organized in an ABAB
stacking sequence along the [010] direction of the standard setting Aba2
[126]. Ignoring K+ counter-cations we can rationalize a single sheet as
[LiMg(BH4)4]−. Li and Mg atoms are arranged in a pseudohexagonal
symmetry (Figure 12a) and are both tetrahedrally coordinated by [BH4]⎺.
This leads to two different boron sites, both on the general position 8b in
Aba2. One position links Mg/Li(BH4)4 tetrahedra via common corners and
56 P. Vajeeston
the other via edges, forming dimeric units. Since double-bonds are counted
as a single link in topology, these sheets correspond to honeycomb layers.
The edge-sharing BH4 groups are located outside the plane defined by the
sheet of pseudohexagonally arranged metal atoms. The corner-linking
[BH4]⎺ groups come to lie within the plane. If we consider a dimeric unit of
[LiMg(BH4)6]3−, then the distances between metal atom centers and those
Bb-atoms that form bridging [BH4]⎺ groups within the unit are longer than
those to the terminal groups Bt, which link to the next dimer and in this way
form the rings from which the sheet is built. For the experimental structure
the refined distances Li–Bb and Li–Bt are 2.56 and 2.54 Å, and the respective
values for Mg–Bb and Mg–Bt are 2.52 and 2.37 Å. The K position is located
in the center of the rings in a distorted cubic coordination by [BH4]⎺. The
single-layered sheets are interconnected by K–BH4 interactions with the
[BH4]⎺ groups lying out of plane.
Cell parameters: a = 8.16633; b = 9.1679; c = 12.5395; Atomic
positions: Li1(4a):0, 0, 0.0538(85.3% occupancy); Li2(4a):0, 0, 0.33645
(14.7% occupancy); K(4a):0, 0, 0.69254; Mg1(4a): 0, 0, 0.0538 (14.7%
occupancy); Mg2:0, 0, 0.0538(85.3% occupancy); B1(8b): 0.7384, 0.5090,
0.4444; B2(8b): 0.60323, 0.8235,0.6973; H1(8b): 0.6741, 0.5952, 0.3954;
H2(8b):0.8444, 0.4629, 0.3972; H3(8b):0.6495, 0.4186, 0.4643;
H4(8b):0.7858, 0.5594, 0.5206; H5(8b):0.5334, 0.7823, 0.6256;
H6(8b):0.6486, 0.9379, 0.6812; H7(8b):0.7111, 0.7496, 0.7133; H8(8b):0
0.5198, 0.8245, 0.7692.
Figure 12. Crystal structures of (a) LiKMg(BH4)4; (b) Li2K3Mg2(BH4)9 (c) and
LiRbMg(BH4)4.
4. Li2K3Mg2(BH4)9
Li2K3Mg2(BH4)9 is made up of double-layered sheets which are stacked
along the monoclinic axis of the space group P2/c (Figure 12b). The new
compound is formed by a reaction between KLiMg(BH4)4 and the KBH4
from molten/decomposed LiK(BH4)2. The Li : Mg ratio is still 1 : 1, which
means that the cation-sequence within a sheet should still correspond to Li–
Mg–Li–Mg [126]. The structure contains one Li and one Mg atom, both
occupying the general position 4g. The sheets contain 4-fold rings that are
interrupted by terminal ligands. The 4-fold rings are interconnected between
both layers of the same double-layered sheet via a Li–B–Mg sequence.
However, within the layer, rings are connected via Li–B–Li within the layer,
while the Mg-atoms have terminal BH4 groups. The potassium atoms occupy
three different positions, 2a, 2c and 2f. 2a and 2f are distorted cubes, sharing
edges in chains along c, while 2c is octahedrally coordinated and forms one-
dimensional octahedral chains connected by vertices in the c direction. Each
octahedron is connected to 2 adjacent octahedra in trans-position and shares
common edge with four cubes in the equatorial plane. There are five
crystallographically distinct [BH4]⎺ groups in the structure, all of which are
bridging ligands, with the exception of B2.
Cell parameters: a = 11.3307; b = 10.20598; c = 11.74449; β = 121.210o;
Atomic positions: Li(4g): 0.2136, 0.7561, 0.7986; K1(2f): 1/2, 0.0005, 3/4;
K2(2c): 0, 1/2, 0; K3(2a): 0, 0, 0; Mg(4g): 0.30230, 0.71402, 0.4039;
B1(4g): 0.2359, 0.7963, 0.5496; B2(4g): 0.2097, 0.2229, 0.6604; B5(2e): 0,
0.1428, 1/4; B3(4g): 0.7725, 0.4899, 0.1175; B4(4g): 0.5342, 0.8060,
0.5194; H1(4g): 0.3332, 0.7392, 0.5782; H2(4g): 0.2242, 0.8042, 0.6391;
H3(4g): 0.1432, 0.7443, 0.4662; H4(4g): 0.2443, 0.8978, 0.5153; H5(4g):
0.1573, 0.190, 0.7142; H6(4g): 0.2094, 0.1401, 0.5963; H7(4g): 0.3202,
0.2515, 0.7352; H8(4g): 0.1545, 0.3103, 0.5963; H9(4g): 0.8518,
0.5651,0.1832; H10(4g): 0.8231, 0.3901, 0.1382; H11(4g): 0.7292, 0.5172,
0.0096; H12(4g): 0.6861, 0.4882, 0.1391; H13(4g): 0.5122, 0.7013, 0.4832;
H14(4g): 0.5812, 0.8602, 0.4682; H15(4g): 0.6092, 0.8071, 0.6305;
H16(4g): 0.4351, 0.8551, 0.4962; H17(4g): -0.0262, 0.2067, 0.1615;
H18(4g): 0.09166, 0.0789, 0.2752.
58 P. Vajeeston
5. LiRbMg(BH4)4
LiRbMg(BH4)4 is built up of single-layered sheets organized in an
ABAB stacking sequence along the [010] direction of the standard setting
Aba2. Ignoring Rb+ counter-cations we can rationalize a single sheet as
[LiMg(BH4)4]−. Li and Mg atoms are arranged in a pseudohexagonal
symmetry and are both tetrahedrally coordinated by [BH4]⎺. This leads to
two different boron sites, both on the general position 8b in Aba2. One
position links Mg/Li(BH4)4 tetrahedra via common corners and the other via
edges (see Figure 12c), forming dimeric units. Since double-bonds are
counted as a single link in topology, these sheets correspond to honeycomb
layers. The edge-sharing [BH4]⎺ groups are located outside the plane defined
by the sheet of pseudohexagonally arranged metal atoms. The corner-linking
[BH4]⎺ groups come to lie within the plane. If we consider a dimeric unit of
[LiMg(BH4)6]3−, then the distances between metal atom centers and those
Bb-atoms that form bridging [BH4]⎺ groups within the unit are longer than
those to the terminal groups Bt, which link to the next dimer and in this way
form the rings from which the sheet is built. For the experimental structure
the refined distances Li–Bb and Li–Bt are 2.56 and 2.54 Å, and the respective
values for Mg–Bb and Mg–Bt are 2.44 and 2.42 Å. The Rb position is located
in the center of the rings in a distorted cubic coordination by [BH4]⎺ The
single-layered sheets are interconnected by Rb–BH4 interactions with the
[BH4]⎺ groups lying out of plane.
Cell parameters: a=8.2342; b= 9.4852; c= 12.5767; Atomic positions:
Rb(4a): 0, 0, 0.70087; Li1(4a): 0, 0.0637, 0.606; Li2(4a): 0, 0.3411, 0.394;
Mg1(4a): 0, 0.0637, 0.394; Mg2(4a): 0, 0.3411, 0.606; B1(8b):0.59950,
0.8333, 0.6960; B2(8b): 0.7726, 0.5366, 0.4453; H1(8b): 0.854, 0.6221,
0.4781; H2(8b): 0.752, 0.5552, 0.3576; H3(8b): 0.6522, 0.538, 0.4884;
H4(8b): 0.832, 0.4305, 0.4571; H5(8b): 0.529, 0.80, 0.7700; H6(8b): 0.524,
0.81, 0.6233; H7(8b): 0.72, 0.7720, 0.6932; H8(4a): 0.63, 0.9505, 0.697.
6. RbMgxCa1-x(BH4)4
The compound RbMgxCa1-x(BH4)3 was identified as a by product of the
RbBH4+Mg(BH4)2+Ca(BH4)2 mixture the cell dimension and the space
group number is reported in Ref. [115]. The structural details (atomic
positions) are however not found in the literature including the structural
data bases [119, 120].
Other Known Monometallic Borohydrides
α-Al(BH4)3; [127] β-Al(BH4)3; [127] Ti(BH4)3; [128] α-Mn(BH4)2; [6] γ-

Mn(BH4)2; [71] δ-Mn(BH4)2; [129] δ′-Mn(BH4)2; [129] α-Y(BH4)3; [130] β-
Y(BH4)3; [125] Zr(BH4)4; [131] α-Cd(BH4)2; [111] β-Cd(BH4)2; [111]
La(BH4)3; [132] Ce(BH4)3; [132] o-Sm(BH4)2; [133] α-Sm(BH4)3; [134] β-
Sm(BH4)3; [134] o-Eu(BH4)2; [133] t-Eu(BH4)2; [106] c-Eu(BH4)2; [106]
Gd(BH4)3; [130] Tb(BH4)3; [134] Dy(BH4)3; [130] α-Ho(BH4)3; [135] β-
Ho(BH4)3; [135] α-Er(BH4)3; [134] β-Er(BH4)3; [134] α-Yb(BH4)2; [130] β-
Yb(BH4)2; [136] γ-Yb(BH4)2; [136] Hf(BH4)4; [137] Th(BH4)4; [138] α-
U(BH4)4; [139] β-U(BH4)4 [140] and Np(BH4)4 [141].
Other Known Bimetalic Mixed Borohydrides
LiSc(BH4)4; [142] LiY(BH4)4; [143] LiYb(BH4)4; [136] LiLu(BH4)4;

[134] LiZn2(BH4)5; [144] LiAl(BH4)4; [145] Li4Al3(BH4)13; [146]
NaAl(BH4)4; [145] NaSc(BH4)4; [147] NaY(BH4)4; [143] NaLa(BH4)4; [148]
NaYb(BH4)4; [149] NaZn(BH4)3; [144] NaZn2(BH4)5; [144] KSr(BH4)3;
[121] KSc(BH4)4; [150] o-KY(BH4)4; [151] m-KY(BH4)4; [152]
K3Ce(BH4)6; [153] KGd(BH4)4; [154] K2Gd(BH4)5; [154] K3Gd(BH4)6;
[154] KHo(BH4)4; [135] lt-KYb(BH4)3; [115] ht-KYb(BH4)3; [115]
KYb(BH4)4; [143] K2Mn(BH4)4; [113] KMn(BH4)3; [113] KZn(BH4)3; [116]
K2Zn(BH4)4; [116] K3Zn(BH4)5; [116] KCd(BH4)3; [111] K2Cd(BH4)4; [111]
KAl(BH4)4; [155] o-RbY(BH4)4; [152] m-RbY(BH4)4; [156] Rb3Y(BH4)6;
[115] Rb3Ce(BH4)6; [115] RbEu(BH4)3; [115] RbAl(BH4)4; [145]
CsY(BH4)4; [156] Cs3Y(BH4)6; [115] CsEu(BH4)3; [115] Cs3Gd(BH4)6; [115]
CsMn(BH4)3; [115] ht-CsAl(BH4)4; [145] CsPb(BH4)3 [115].
Table 2. Crystal structure data (space group and unit cell parameters), hydrogen density (in wt %), stability
range, and number of formula units (Z) in the structure for selected borohydrides
S. Compound ρm Space group Crystal Stable at Z Lattice parameters in (Å) Ref.

no (wt% H) system a b c Angle (o)
1) -B2H6 22 Hex. 4.2K 4.540 8.690 γ=120 [37]
2) β-B2H6 P21/n [14] Mono. 60K 2 4.400 5.720 6.500 γ =105.1 [39]
3) m-B2H6 P21/n [14] Mono. 2 4.460 8.680 4.550 β=120.5 [36]
4) -LiBH4 18.5 Pnma [62] Ortho. RT 4 7.1786 4.4369 6.8032 [44,
157]
5) β-LiBH4 P63mc [186] Hex. >380 K 2 4.2763 6.9484 [157]
6) HP1-LiBH4 Ama2 [40] Ortho. RT-1.2-10 GPa 4 6.449 5.3070 5.2919 [48]
7) HP2-LiBH4 Fm-3m [225] Cubic RT, >10 GPa 4 5.109 [48]
8) -NaBH4 10.7 Fm-3m [225] Cubic RT 4 6.1308 [49]
9) β-NaBH4 (LT) P42/nmc [137] Tet. <190K or >6GPa 2 4.3706 5.9509 [158]
@RT
10) -NaBH4 (HP) Pnma [62] Ortho. >9 GPa 4 7.2970 4.1166 5.5692 [61]
11) -KBH4 7.5 Fm-3m [225] Cubic RT 4 10.6100 [49]
12) β-KBH4 P42/nmc [137] Tet. <70K or 3.8-6.8 2 4.4754 6.3630 [63,
GPa @RT 64]
13) -KBH4 Pnma [62] Ortho. 4 6.9760 4.9330 5.1110 [64]
14) -RbBH4 4.0 Fm-3m [225] Cubic RT 4 7.0290 [49]
15) HP1-RbBH4 P4/nmm Tet. 5.5 GPa 2 5.6122 4.0887 [65]
(choice2; [129])
16) HP2-RbBH4 C222 [21] Ortho. 16.1GPa 2 5.3679 5.1343 3.9098 [65]
Table 2. (Continued)

17) HP3-RbBH4 I-42m [121] Tet. 23.5 GPa 4 5.0057 7.8161 [65]
18) CsBH4 2.7 Fm-3m [225] Cubic RT 4 7.4190 [49]
19) Be(BH4)2 20.8 I41/cd [88] Tet. RT 16 13.6200 9.100
20) -Mg(BH4)2 14.9 P6122 [178] Hex. RT 30 10.3540, 37.0550 [5, 77]
21) β-Mg(BH4)2 Fddd [70] Ortho. RT, metastable, 64 37.072 18.648 10.912 [90-92]
HT polymorph
22) β'-Mg(BH4)2 allegedly a disordered phase of β not yet solved [79,
80, 85]
23) -Mg(BH4)2 Ia-3d [230] Cubic RT, metastable 24 15.7575 [7]
Id-3a
24) 1-Mg(BH4)2 P42nm [102] Tet. RT, >2.1 GPa 2 5.4361 6.1468 [7]
25) 2-Mg(BH4)2 P63 [173] Hex. HP 14-22 GPa 8.35 4.68 [85]
26) -Mg(BH4)2 Not yet solved Above 190°C [79]
27) -Mg(BH4)2 P3112 [151] Hex. ∼500 K 9 10.424 10.729 [71]
28) -Ca(BH4)2 11.6 F2dd [43] Ortho. RT 8 8.7759 13.0234 7.4132 [97]
29) '-Ca(BH4)2 I-42d [122] Tet. >495 K 4 5.8446 13.2279 [97]
30) β-Ca(BH4)2 P-4 [81] Tet. RT, metastable 2 6.9189 4.3471 [97]
31) β2-Ca(BH4)2 P42/m [84] Tet. HT polymorph 2 6.924 4.3492 [98]
32) -Ca(BH4)2 Pbca [61] Ortho. RT, metastable 8 7.5250 13.1090 8.4030 [100]
33) o-Sr(BH4)2 6.9 Pbcn [60] Ortho. RT 4 6.978 8.937 7.559 [103]
34) t-Sr(BH4)2 P41212 [92] Tet. 450-490 oC 4 5.4208 11.709 [106]
no (wt% H) system o
a b c Angle ( )
35) c-Sr(BH4)2 Fm-3m [225] Cubic 450-500 oC 4 7.10870 [106]
36) O1-Ba(BH4)2 4.8 Pnnm [58] Ortho. RT 2 6.98476 7.2284 4.8494 [106]
37) O2-Ba(BH4)2 Pbcn [60] Ortho. >668 K 12 7.4043 8.0935 8.8201 [106]
38) t-Ba(BH4)2 P41212 [92] Tet. >718 K 4 5.5805 12.5727 [106]
39) c-Ba(BH4)2 Fm-3m [225] Cubic >773 4 7.0930 [106]
40) LiK(BH4)2 10.6 Pnma [62] Ortho. RT 4 7.9134 4.4907 13.8440 [159]
41) m-LiRb(BH4)2 6.6 C2/m [12] Mono. RT-403K 14 23.5931 4.6178 17.1893 β=97.776 [108]
42) o-LiRb(BH4)2 Cmc21 [36] Ortho. 403-414K 4 4.5711 15.4847 7.9355 [108]
43) h-Li2Rb(BH4)3 8.4 P6222 [180] Hex. RT-392K 3 7.474 11.585 [108]
44) o-LiCs(BH4)2 4.8 Cmc21 [36] Ortho. RT-430K 4 4.6817 15.950 7.9438 [108]
45) h-Li2Cs(BH4)3 6.3 P6222 [180] Hex. RT-420 3 7.6634 12.1279 [108]
46) m-Li2Cs(BH4)3 Cc [9] Mono. RT-331K 4 7.574 7.948 12.880 β=89.97 [108]
47) m-LiCs2(BH4)3 3.8 C2/c [15] Mono. RT-450K 8 7.6021 15.160 17.243 β=92.693 [108]
48) o-Li3Cs2(BH4)5 5.6 Fddd [70] Ortho. 345-430K 16 13.5838 17.3001 23.9151 [108]
49) m-Li2Cs3(BH4)5 4.1 P21/c [14] Mono. RT-345K 4 14.444 7.4218 14.7907 β=104.470 [108]
50) m-Li3Cs(BH4)4 7.6 P21/c [14] Mono. RT-398K 4 8.1992 12.3044 11.9009 β=123.767 [108]
51) K2Mg(BH4)4 10.0 P21/c [14] Mono. RT 4 8.1791 9.8842 12.7492 β=100.74 [113]
52) LT-K3Mg(BH4)5 9.3 P42bc [106] Tet. RT 4 8.9706 15.9486 [113]
53) HT-K3Mg(BH4)5 9.3 P42/mbc [135] Tet. >370K* 4 8.9693 15.9501 [113]
54) LT-KCa(BH4)3 9.8 Pba2 [32] Ortho. 110-345K 2 7.939 7.837 5.559 [115]
55) HT-KCa(BH4)3 Pna21 [33] Ortho. 345-500K 4 8.003 8.195 11.474 [115]
56) -KSr(BH4)3 7.1 Pna21 [33] Ortho. RT 4 7.8967 8.2953 11.508 [121]
57) Rb3Mg(BH4)5 5.7 I4/mcm [140] Tet. 4 9.2996 15.9930 [122]
58) LT-RbCa(BH4)3 7.1 Fm-3c [226] Cubic 298K 5.728 [115]
Table 2. (Continued)

59) HT1-RbCa(BH4)3 C2221 [20] Ortho. 343K 8.097 16.165 22.850 [115]
60) HT2-RbCa(BH4)3 I4/mcm [140) Tet. 408K 4 11.500 11.541 [115]
61) RbSr(BH4)3 5.6 Pna21 [33) Ortho. RT 4 8.0835 8.3341 11.6600 [115,
121]
62) -CsMg(BH4)3 6.0 Cc [9) Mono. 298K 13.5871 8.3542 19.2559 β=96.59 [115]
63) Cs3Mg(BH4)5 4.1 I4/mcm [140] Tet. 4 9.7115 16.2540 [122]
64) LT-CsCa(BH4)3 5.6 Fm-3c [226] Cubic 298K 11.492 [115]
65) HT-CsCa(BH4)3 5.6 Pm-3m [221] Cubic 510K 5.863 [115]
66) -CsSr(BH4)3 4.6 P21212 [18] Ortho. RT 4 8.2068 8.1793 6.0761 [115,
121]
67) LiKMg(BH4)4 12.4 Aba2 [41] Ortho. RT 4 8.1662 9.1683 12.5391 [126]
68) Li2K3Mg2(BH4)9 11.6 P2/c [13] Mono. RT 2 11.321 10.1811 11.7372 β= 121.2o [126]
69) LiRbMg(BH4)4 9.2 Aba2 [41] Ortho. RT 4 8.234 9.4852 12.5772 [126]
70) RbMgxCa1-x(BH4)3 C2221 [20] Ortho. 298K 8.057 16.093 22.751 [115]
64 P. Vajeeston
Other Known Trimetallic Mixed Borohydrides
LiKMn(BH4)4; [126] Li3K3La2(BH4)12; [153] Li3K3Ce2(BH4)12; [153]

LiRbMn(BH4)4; [126] Rb2LiY(BH4)6; [156] Cs2LiY(BH4)6; [156]
Cs2LiCe(BH4)6; [115] Cs2LiGd(BH4)6; [115] Li3MgZn5(BH4)15; [112]
Li3MnZn5(BH4)15; [112] RbMgxCa1−x(BH4)3 [115].
Remark
The crystal structures of listed compounds in this chapter are

downloadable from the following link (http://folk.uio.no/ponniahv/
Database/BH-str).
REFERENCES
[1] Lee R. The Outlook for Population Growth. Science.

2011;333(6042):569-73.
[2] Christopher K, Dimitrios R. A review on exergy comparison of
hydrogen production methods from renewable energy sources.
Energy Environ Sci 2012;5(5):6640-51.
[3] Ley MB, Jepsen LH, Lee YS, Cho YW, Bellosta von Colbe JM,
Dornheim M, et al. Complex hydrides for hydrogen storage – new
perspectives. Mater Today. 2014;17(3):122-8.
[4] Filinchuk Y, Chernyshov D, Dmitriev V. Light metal borohydrides:
crystal structures and beyond. Z Krist 2008. p. 649.
[5] Černý R, Filinchuk Y, Hagemann H, Yvon K. Magnesium
Borohydride: Synthesis and Crystal Structure. Angew Chem Int Ed
2007;46(30):5765-7.
[6] Černý R, Penin N, Hagemann H, Filinchuk Y. The First
Crystallographic and Spectroscopic Characterization of a 3d-Metal
Borohydride: Mn(BH4)2. J Phys Chem C. 2009;113(20):9003-7.
[7] Filinchuk Y, Richter B, Jensen TR, Dmitriev V, Chernyshov D,

Hagemann H. Porous and Dense Magnesium Borohydride
Frameworks: Synthesis, Stability, and Reversible Absorption of
Guest Species. Angew Chem Int Ed 2011;50(47):11162-6.
[8] Unemoto A, Matsuo M, Orimo S-i. Complex Hydrides for
Electrochemical Energy Storage. Adv Funct Mater.
2014;24(16):2267-79.
[9] Orimo S-i, Nakamori Y, Eliseo JR, Züttel A, Jensen CM. Complex
Hydrides for Hydrogen Storage. Chem Rev. 2007;107(10):4111-32.
[10] Bogdanović B, Schwickardi M. Ti-doped alkali metal aluminium
hydrides as potential novel reversible hydrogen storage materials.
Invited paper presented at the International Symposium on Metal–
Hydrogen Systems, Les Diablerets, August 25–30, 1996,
Switzerland.1. J Alloys Compd. 1997;253-254(Supplement C):1-9.
[11] Callini E, Atakli ZÖK, Hauback BC, Orimo S-i, Jensen C, Dornheim
M, et al. Complex and liquid hydrides for energy storage. Appl Phys
A. 2016;122(4):353.
[12] Vajo JJ, Mertens F, Ahn CC, Bowman RC, Fultz B. Altering
Hydrogen Storage Properties by Hydride Destabilization through
Alloy Formation: LiH and MgH2 Destabilized with Si. J Phys Chem
B. 2004;108(37):13977-83.
[13] Bösenberg U, Doppiu S, Mosegaard L, Barkhordarian G, Eigen N,
Borgschulte A, et al. Hydrogen sorption properties of MgH2–LiBH4
composites. Acta Mater. 2007;55(11):3951-8.
[14] Alapati SV, Johnson JK, Sholl DS. Identification of Destabilized
Metal Hydrides for Hydrogen Storage Using First Principles
Calculations. J Phys Chem B. 2006;110(17):8769-76.
[15] Matsuo M, Orimo S-i. Lithium Fast-Ionic Conduction in Complex
Hydrides: Review and Prospects. Adv Energy Mater. 2011;1(2):
161-72.
[16] Schouwink P, Didelot E, Lee YS, Mazet T, Černý R. Structural and
magnetocaloric properties of novel gadolinium borohydrides.
Journal of Alloys and Compounds. 2016;664:378-84.
66 P. Vajeeston
[17] Zhu Y, Hosmane NS. Nanocatalysis: Recent advances and

applications in boron chemistry. Coordin Chem Rev 2015;293-
294:357-67.
[18] Schlesinger HI, Sanderson RT, Burg AB. Metallo Borohydrides. I.
Aluminum Borohydride. J Am Chem Soc 1940;62(12):3421-5.
[19] Sakintuna B, Lamari-Darkrim F, Hirscher M. Metal hydride
materials for solid hydrogen storage: A review. Int J Hydrog Energy.
2007;32(9):1121-40.
[20] Yu CL, Bauer SH. Thermochemistry of the Boranes. J Phys Chem
Ref Data. 1998;27(4):807-35.
[21] Mikhailov BM. The chemistry of diborane. Russ Chem Rev.
1962;31:207-24.
[22] Lipscomb WN. Electron correlation effects in boron hydride
structures, intermediates and reactions. Pure Appl Chem 1983. p.
1431.
[23] Hedberg K, Schomaker V. A Reinvestigation of the Structures of
Diborane and Ethane by Electron Diffraction. J Am Chem Soc.
1951;73:1482-7.
[24] Jones DS, Lipscomb WN. X‐Ray Evidence for Bonding Electrons in
Diborane. J Chem Phys. 1969;51:3133-4.
[25] Fred S. Infra Red and Raman Spectra of Polyatomic Molecules XV.
Diborane. J Chem Phys. 1941;9:780-5.
[26] Webb AN, Neu JT, Pitzer KS. The Infra‐Red and Raman Spectra and
the Thermodynamic Properties of Diborane. J Chem Phys.
1949;17:1007-11.
[27] Lord RC, Nielsen E. The Vibrational Spectra of Diborane and Some
of its Isotopic Derivatives. J Chem Phys. 1951;19:1-10.
[28] Freund I, Halford RS. Motions of Molecules in Condensed Systems.
XIV. Polarized Infrared Spectra of Single Crystals of Diborane.
J Chem Phys. 1965;43:3795.
[29] Duncan JL, McKean DC, Torto I, Nivellini GD. Diborane: Infrared
spectra of gaseous and crystalline phases, and a complete vibrational
assignment. J Mol Spectrosc. 1981;85:16-39.
[30] Lipscomb WN. Electron correlation effects in boron hydride

structures, intermediates and reactions. Pure Appl Chem.
1983;55:1431-8.
[31] Duncan JL, Harper J, Hamilton E, Nivellini GD. The empirical
harmonic potential function of diborane. J Mol Spectrosc.
1983;102:416-40.
[32] Stanton JF, Bartlett RJ, Lipscomb WN. A coupled-cluster and MBPT
study of B2H6 and BH3. Chem Phys Lett. 1987;138:525-30.
[33] W. BT, K. MA, E. KJ, M. VS. High-pressure boron hydride phases.
Phys Rev B. 1997;56:5148-55.
[34] S. N, J. HR, A. GE, F. GA, K. MH. Pressure-induced transformations
of molecular boron hydride. J Phys: Condens Matter.
2002;14:10453-6.
[35] Song Y, Murli C, Liu Z, L. YC, H. BS. In situ high-pressure study of
diborane by infrared spectroscopy. J Chem Phys. 2009;131(17):
174506.
[36] Yao Y, Hoffmann R. BH3 under Pressure: Leaving the Molecular
Diborane Motif. J Am Chem Soc. 2011;133(51):21002-9.
[37] Mark H, Pohland E. The lattice structure of ethane and diborane
(B2H6). Z Kristallogr. 1925;62:103-12.
[38] Bolz LH, Mauer FA, Peiser HS. Exploratory Study, by Low‐
Temperature X‐Ray Diffraction Techniques, of Diborane and the
Products of a Microwave Discharge in Diborane. J Chem Phys
1959;31(4):1005-7.
[39] Smith HW, Lipscomb WN. Single‐Crystal X‐Ray Diffraction Study
of β‐Diborane. J Chem Phys. 1965;43(3):1060-4.
[40] Cerny R, Schouwink P. The crystal chemistry of inorganic metal
borohydrides and their relation to metal oxides. Acta Crystallogr B.
2015;71(6):619-40.
[41] Schlesinger HI, Brown HC. Metallo Borohydrides. III. Lithium
Borohydride. J Am Chem Soc. 1940;62(12):3429-35.
[42] Schlesinger HI, Brown HC, Hoekstra HR, Rapp LR. Reactions of
Diborane with Alkali Metal Hydrides and Their Addition
68 P. Vajeeston
Compounds. New Syntheses of Borohydrides. Sodium and

Potassium Borohydrides1. J Am Chem Soc. 1953;75(1):199-204.
[43] Soulié JP, Renaudin G, Černý R, Yvon K. Lithium boro-hydride
LiBH4: I. Crystal structure. J Alloys Compd. 2002;346(1):200-5.
[44] Vajeeston P, Ravindran P, Kjekshus A, Fjellvåg H. Structural
stability of alkali boron tetrahydrides ABH4 (A = Li, Na, K, Rb, Cs)
from first principle calculation. J Alloys Compd 2005;387(1):97-104.
[45] Züttel A, Wenger P, Rentsch S, Sudan P, Mauron P, Emmenegger C.
LiBH4 a new hydrogen storage material. J Power Sources.
2003;118(1):1-7.
[46] Pistorius Carl WFT. Melting and Polymorphism of LiBH4 to 45 kbar.
Z Phys Chem 1974. p. 253.
[47] Semenenko KN, Chavgun AP, Surov VN. Reaction of sodium
borohydride with potassium and lithium borohydrides. Russ J Inorg
Chem. 1971;16:271-3.
[48] Filinchuk Y, Chernyshov D, Nevidomskyy A, Dmitriev V. High-
Pressure Polymorphism as a Step towards Destabilization of LiBH4.
Angew Chem Int Ed 2008;47(3):529-32.
[49] Abrahams SC, Kalnajs J. The Lattice Constants of the Alkali
Borohydrides and the Low‐Temperature Phase of Sodium
Borohydride. J Chem Phys 1954;22(3):434-6.
[50] Demirci UB, Miele P. Sodium borohydride versus ammonia borane,
in hydrogen storage and direct fuel cell applications. Energy Environ
Sci. 2009;2(6):627-37.
[51] Johnston HL, Hallett NC. Low Temperature Heat Capacities of
Inorganic Solids. XIV. Heat Capacity of Sodium Borohydride from
15-300°K.1. J Am Chem Soc. 1953;75(6):1467-8.
[52] Stockmeyer WH, Stephenson CC. The Nature of the Gradual
Transition in Sodium Borohydride. J Chem Phys. 1953;21:1311-2.
[53] Filinchuk Y, Hagemann H. Structure and Properties of NaBH4·2H2O
and NaBH4. Eur J Inorg Chem. 2008;2008(20):3127-33.
[54] Fischer P, Zuttel A. Order-Disorder Phase Transition in NaBD4.
Mater Sci Forum. 2004;443-444:287-90.
[55] Kumar RS, Cornelius AL. Structural transitions in NaBH4 under

pressure. Appl Phys Lett. 2005;87(26):261916.
[56] Sundqvist B, Andersson O. Low-temperature phase transformation
in NaBH4 under pressure. Phys Rev B. 2006;73(9):092102.
[57] Soldate AM. Crystal Structure of Sodium Borohydride. J Am Chem
Soc. 1947;69(5):987-8.
[58] Davis RL, Kennard CL. Structure of sodium tetradeuteroborate,
NaBD4. J Solid State Chem. 1985;59:393-6.
[59] Fischer P, Zuttel A, editors. Order-Disorder Phase Transition in
Na[BD4]. Proceeding of EPDIC-8; 2002: Trans. Tech. Publication
Ltd.
[60] Babanova OA, Soloninin AV, Stepanov AP, Skripov AV, Filinchuk
Y. Structural and Dynamical Properties of NaBH4 and KBH4: NMR
and Synchrotron X-ray Diffraction Studies. J Phys Chem C.
2010;114(8):3712-8.
[61] Filinchuk Y, Talyzin AV, Chernyshov D, Dmitriev V. High-pressure
phase of NaB4: Crystal structure from synchrotron powder diffraction
data. Phys Rev B. 2007;76(9):092104.
[62] Chernyshov D, Bosak A, Dmitriev V, Filinchuk Y, Hagemann H.
Low-lying phonons in NaBH4 studied by inelastic scattering of
synchrotron radiation. Phys Rev B. 2008;78(17):172104.
[63] Renaudin G, Gomes S, Hagemann H, Keller L, Yvon K. Structural
and spectroscopic studies on the alkali borohydrides MBH4 (M = Na,
K, Rb, Cs). J Alloys Compd 2004;375(1):98-106.
[64] Kumar RS, Kim E, Cornelius AL. Structural Phase Transitions in the
Potential Hydrogen Storage Compound KBH4 under Compression. J
Phys Chem C. 2008;112(22):8452-7.
[65] Filinchuk Y, Talyzin AV, Hagemann H, Dmitriev V, Chernyshov D,
Sundqvist B. Cation Size and Anion Anisotropy in Structural
Chemistry of Metal Borohydrides. The Peculiar Pressure Evolution
of RbBH4. Inorg Chem 2010;49(11):5285-92.
[66] Stephenson CC, Rice DW, Stockmayer WH. Order‐Disorder
Transitions in the Alkali Borohydrides. J Chem Phys.
1955;23(10):1960-.
70 P. Vajeeston
[67] Shannon R. Revised effective ionic radii and systematic studies of

interatomic distances in halides and chalcogenides. Acta Crystallogr
A. 1976;32(5):751-67.
[68] Burg AB, Schlesinger HI. Metallo Borohydrides. II. Beryllium
Borohydride. J Am Chem Soc. 1940;62(12):3425-9.
[69] Marynick DS, Lipscomb WN. Crystal structure of beryllium
borohydride. Inorg Chem 1972;11(4):820-3.
[70] Zavorotynska O, El-Kharbachi A, Deledda S, Hauback BC. Recent
progress in magnesium borohydride Mg(BH4)2: Fundamentals and
applications for energy storage. Int J Hydrog Energy.
2016;41(32):14387-403.
[71] Richter B, Ravnsbaek DB, Tumanov N, Filinchuk Y, Jensen TR.
Manganese borohydride; synthesis and characterization. Dalton
Trans. 2015;44(9):3988-96.
[72] Ban V, Soloninin AV, Skripov AV, Hadermann J, Abakumov A,
Filinchuk Y. Pressure-Collapsed Amorphous Mg(BH4)2: An
Ultradense Complex Hydride Showing a Reversible Transition to the
Porous Framework. J Phys Chem C. 2014;118(40):23402-8.
[73] Ozolins V, Majzoub EH, Wolverton C. First-Principles Prediction of
a Ground State Crystal Structure of Magnesium Borohydride. Phys
Rev Lett. 2008;100(13):135501.
[74] Muldoon J, Bucur CB, Gregory T. Quest for Nonaqueous
Multivalent Secondary Batteries: Magnesium and Beyond. Chem
Rev. 2014;114(23):11683-720.
[75] Chłopek K, Frommen n, Zabara O, Fichtner M. Synthesis and
properties of magnesium tetrahydroborate, Mg(BH4)2. J Mater
Chem. 2007;17:3496-503.
[76] Riktor MD, Sorby MH, Chlopek K, Fichtner M, Buchter F, Zuttel A,
et al. In situ synchrotron diffraction studies of phase transitions and
thermal decomposition of Mg(BH4)2 and Ca(BH4)2. J Mater Chem
2007;17(47):4939-42.
[77] Filinchuk Y, Černý R, Hagemann H. Insight into Mg(BH4)2 with
Synchrotron X-ray Diffraction: Structure Revision, Crystal
Chemistry, and Anomalous Thermal Expansion. Chem Mater.

2009;21(5):925-33.
[78] David WIF, Callear SK, Jones MO, Aeberhard PC, Culligan SD,
Pohl AH, et al. The structure, thermal properties and phase
transformations of the cubic polymorph of magnesium
tetrahydroborate. Phys Chem Chem Phys. 2012;14(33):11800-7.
[79] Paskevicius M, Pitt MP, Webb CJ, Sheppard DA, Filsø U, Gray EM,
et al. In-Situ X-ray Diffraction Study of γ-Mg(BH4)2 Decomposition.
J Phys Chem C. 2012;116(29):15231-40.
[80] Zavorotynska O, Deledda S, Vitillo J, Saldan I, Guzik M, Baricco M,
et al. Combined X-ray and Raman Studies on the Effect of Cobalt
Additives on the Decomposition of Magnesium Borohydride.
Energies. 2015;8(9):9173.
[81] Vitillo JG, Bordiga S, Baricco M. Spectroscopic and Structural
Characterization of Thermal Decomposition of γ-Mg(BH4)2:
Dynamic Vacuum versus H2 Atmosphere. J Phys Chem C.
2015;119(45):25340-51.
[82] Newhouse RJ, Stavila V, Hwang SJ, Klebanoff LE, Zhang JZ.
Reversibility and Improved Hydrogen Release of Magnesium
Borohydride. J Phys Chem C. 2010;114(11):5224-32.
[83] Severa G, Ronnebro E, Jensen CM. Direct hydrogenation of
magnesium boride to magnesium borohydride: demonstration of 11
weight percent reversible hydrogen storage. Chem Comm.
2010;46(3):421-3.
[84] Saldan I, Hino S, Humphries TD, Zavorotynska O, Chong M, Jensen
CM, et al. Structural Changes Observed during the Reversible
Hydrogenation of Mg(BH4)2 with Ni-Based Additives. J Phys Chem
C. 2014;118(40):23376-84.
[85] George L, Drozd V, Saxena SK, Bardaji EG, Fichtner M. Structural
Phase Transitions of Mg(BH4)2 under Pressure. J Phys Chem C.
2009;113(1):486-92.
[86] Li HW, Kikuchi K, Nakamori Y, Miwa K, Towata S, Orimo S.
Effects of ball milling and additives on dehydriding behaviors of
well-crystallized Mg(BH4)2. Scr Mater. 2007;57(8):679-82.
72 P. Vajeeston
[87] Guo S, Chan HYL, Reed D, Book D. Investigation of

dehydrogenation processes in disordered γ-Mg(BH4)2. J Alloys
Compd 2013;580(Supplement 1):S296-S300.
[88] Her JH, Stephens PW, Gao Y, Soloveichik GL, Rijssenbeek J,
Andrusb M, et al. Structure of unsolvated magnesium borohydride
Mg(BH4)2. Acta Crystallogr. 2007;B63:561-8.
[89] Dai B, Sholl DS, Johnson JK. First-Principles Study of Experimental
and Hypothetical Mg(BH4)2 Crystal Structures. J Phys Chem C.
2008;112(11):4391-5.
[90] Bird PH, Churchill MR. The crystal structure of zirconium(IV)
borohydride (at -160[degree]). Chem Comm. 1967(8):403-.
[91] Pitt MP, Webb CJ, Paskevicius M, Sheptyakov D, Buckley CE, Gray
EM. In Situ Neutron Diffraction Study of the Deuteration of Isotopic
Mg11B2. J Phys Chem C. 2011;115(45):22669-79.
[92] Her JH, Stephens PW, Gao Y, Soloveichik GL, Rijssenbeek J,
Andrus M, et al. Structure of unsolvated magnesium borohydride
Mg(BH4)2. Acta Crystallogr B. 2007;63(4):561-8.
[93] Foster MD, Delgado Friedrichs O, Bell RG, Almeida Paz FA,
Klinowski J. Chemical Evaluation of Hypothetical Uninodal
Zeolites. J Am Chem Soc. 2004;126(31):9769-75.
[94] Banerjee R, Phan A, Wang B, Knobler C, Furukawa H, O'Keeffe M,
et al. High-Throughput Synthesis of Zeolitic Imidazolate
Frameworks and Application to CO2 Capture. Science.
2008;319(5865):939-43.
[95] Casanova D, Llunell M, Alemany P, Alvarez S. The Rich
Stereochemistry of Eight-Vertex Polyhedra: A Continuous Shape
Measures Study. Chem Eur J. 2005;11(5):1479-94.
[96] Miwa K, Aoki M, Noritake T, Ohba N, Nakamori Y, Towata S-i, et
al. Thermodynamical stability of calcium borohydride Ca(BH4)2.
Phys Rev B. 2006;74(15):155122.
[97] Filinchuk Y, Rönnebro E, Chandra D. Crystal structures and phase
transformations in Ca(BH4)2. Acta Materialia. 2009;57(3):732-8.
[98] Buchter F, Łodziana Z, Remhof A, Friedrichs O, Borgschulte A,
Mauron P, et al. Structure of Ca(BD4)2 β-Phase from Combined
Neutron and Synchrotron X-ray Powder Diffraction Data and

Density Functional Calculations. J Phys Chem B.
2008;112(27):8042-8.
[99] Filinchuk Y, Chernyshov D, Cerny R. Lightest Borohydride Probed
by Synchrotron X-ray Diffraction: Experiment Calls for a New
Theoretical Revision. J Phys Chem C. 2008;112(28):10579-84.
[100] Buchter F, Łodziana Z, Remhof A, Friedrichs O, Borgschulte A,
Mauron P, et al. Structure of the Orthorhombic γ-Phase and Phase
Transitions of Ca(BD4)2. J Phys Chem C. 2009;113(39):17223-30.
[101] Barkhordarian G, Jensen TR, Doppiu S, Bösenberg U, Borgschulte
A, Gremaud R, et al. Formation of Ca(BH4)2 from Hydrogenation of
CaH2+MgB2 Composite. J Phys Chem C. 2008;112(7):2743-9.
[102] Majzoub EH, Rönnebro E. Crystal Structures of Calcium
Borohydride: Theory and Experiment. J Phys Chem C.
2009;113(8):3352-8.
[103] Ravnsbæk DB, Nickels EA, Černý R, Olesen CH, David WIF,
Edwards PP, et al. Novel Alkali Earth Borohydride Sr(BH4)2 and
Borohydride-Chloride Sr(BH4)Cl. Inorg Chem 2013;52(19):10877-
85.
[104] Zaslavskii AI, Kondrashev YD, Tolkachev SS. New modification of
lead dioxide and the texture of anodic deposits Dokl Akad Nauk
SSSR. 1950;75:559– 61.
[105] Bremer M, Nöth H, Thomann M, Schmidt M. Preparation and
Molecular Structures of Tetrahydrofuran, Diethylene Diglycol
Dimethyl Ether and 18-Crown-6 Complexes of Strontium and
Barium Tetrahydridoborate. Chem Ber. 1995;128(5):455-60.
[106] Sharma M, Didelot E, Spyratou A, Lawson Daku LM, Černý R,
Hagemann H. Halide Free M(BH4)2 (M = Sr, Ba, and Eu) Synthesis,
Structure, and Decomposition. Inorg Chem 2016;55(14):7090-7.
[107] Nickels EA, Jones MO, David WIF, Johnson SR, Lowton RL,
Sommariva M, et al. Tuning the Decomposition Temperature in
Complex Hydrides: Synthesis of a Mixed Alkali Metal Borohydride.
Angew Chem Int Ed 2008;47(15):2817-9.
74 P. Vajeeston
[108] Schouwink P, Smrcok L, Cerny R. Role of the Li+ node in the Li-
BH4 substructure of double-cation tetrahydroborates. Acta
Crystallogr B. 2014;70(5):871-8.
[109] Perez-Flores JC, Baehtz C, Hoelzel M, Kuhn A, Garcia-Alvarado F.
Full structural and electrochemical characterization of Li2Ti6O13 as
anode for Li-ion batteries. Physical chemistry chemical physics:
PCCP. 2012;14(8):2892-9.
[110] Alonso JA, Sanz J, Santamaría J, León C, Várez A, Fernández-Díaz
MT. On the Location of Li+ Cations in the Fast Li-Cation Conductor
La0.5Li0.5TiO3 Perovskite. Angew Chem Int Ed 2000;39(3):619-21.
[111] Ravnsbæk DB, Sørensen LH, Filinchuk Y, Besenbacher F, Jensen
TR. Screening of Metal Borohydrides by Mechanochemistry and
Diffraction. Angew Chem Int Ed 2012;51(15):3582-6.
[112] Černý R, Schouwink P, Sadikin Y, Stare K, Smrčok Lu, Richter B, et
al. Trimetallic Borohydride Li3MZn5(BH4)15 (M = Mg, Mn)
Containing Two Weakly Interconnected Frameworks. Inorg Chem
2013;52(17):9941-7.
[113] Schouwink P, D’Anna V, Ley MB, Lawson Daku LM, Richter B,
Jensen TR, et al. Bimetallic Borohydrides in the System M(BH4)2–
KBH4 (M = Mg, Mn): On the Structural Diversity. J Phys Chem C.
2012;116(20):10829-40.
[114] Aeberhard PC, Refson K, Edwards PP, David WIF. High-pressure
crystal structure prediction of calcium borohydride using density
functional theory. Phys Rev B. 2011;83(17):174102.
[115] Schouwink P, Ley MB, Tissot A, Hagemann H, Jensen TR, Smrčok
Ľ, et al. Structure and properties of complex hydride perovskite
materials. Nat Commun. 2014;5:5706.
[116] Černý R, Ravnsbæk DB, Schouwink P, Filinchuk Y, Penin N,
Teyssier J, et al. Potassium Zinc Borohydrides Containing Triangular
[Zn(BH4)3]⎺ and Tetrahedral [Zn(BH4)xCl4–x]2– Anions. J Phys
Chem C. 2012;116(1):1563-71.
[117] Glazer A. The classification of tilted octahedra in perovskites. Acta
Crystallogr B. 1972;28(11):3384-92.
[118] Schouwink P, Hagemann H, Embs JP, Anna VD, Černý R. Di-

hydrogen contact induced lattice instabilities and structural dynamics
in complex hydride perovskites. J Phys Condens Matter.
2015;27(26):265403.
[119] Belsky A, Hellenbrandt M, Karen VL, Luksch P. New developments
in the Inorganic Crystal Structure Database (ICSD): accessibility in
support of materials research and design. Acta Crystallogr B.
2002;58(3 Part 1):364-9.
[120] Groom CR, Bruno IJ, Lightfoot MP, Ward SC. The Cambridge
Structural Database. Acta Crystallogr B. 2016;72(2):171-9.
[121] Moller KT, Ley MB, Schouwink P, Cerny R, Jensen TR. Synthesis
and thermal stability of perovskite alkali metal strontium
borohydrides. Dalton Trans. 2016;45(2):831-40.
[122] Wegner W, Jaron T, Dobrowolski MA, Dobrzycki L, Cyranski MK,
Grochala W. Organic derivatives of Mg(BH4)2 as precursors towards
MgB2 and novel inorganic mixed-cation borohydrides. Dalton Trans.
2016;45(36):14370-7.
[123] Kanishka B, Muthu DVS, Sood AK, Kruger MB, Chen B, Rao CNR.
Pressure-induced phase transitions in nanocrystalline ReO3. J Phys
Condens Matter. 2007;19(43):436214.
[124] Ogata Y, Tsuda K, Akishige Y, Tanaka M. Refinement of the crystal
structural parameters of the intermediate phase of h-BaTiO3 using
convergent-beam electron diffraction. Acta Crystallogr A.
2004;60(6):525-31.
[125] Frommen C, Aliouane N, Deledda S, Fonneløp JE, Grove H,
Lieutenant K, et al. Crystal structure, polymorphism, and thermal
properties of yttrium borohydride Y(BH4)3. J Alloys Compd
2010;496(1):710-6.
[126] Schouwink P, Ley MB, Jensen TR, Smrcok L, Cerny R.
Borohydrides: from sheet to framework topologies. Dalton Trans.
2014;43(21):7726-33.
[127] Aldridge S, J. Blake A, J. Downs A, O. Gould R, Parsons S, R.
Pulham C. Some tetrahydroborate derivatives of aluminium: crystal
structures of dimethylaluminium tetrahydroborate and the [small
76 P. Vajeeston
alpha] and [small beta] phases of aluminium tris(tetrahydroborate) at

low temperature. J Chem Soc, Dalton Trans. 1997(6):1007-12.
[128] Fang ZZ, Ma LP, Kang XD, Wang PJ, Wang P, Cheng HM. In situ
formation and rapid decomposition of Ti(BH4)3 by mechanical
milling LiBH4 with TiF3. Appl Phys Lett. 2009;94(4):044104.
[129] Tumanov NA, Roedern E, Łodziana Z, Nielsen DB, Jensen TR,
Talyzin AV, et al. High-Pressure Study of Mn(BH4)2 Reveals a
Stable Polymorph with High Hydrogen Density. Chem Mater.
2016;28(1):274-83.
[130] Sato T, Miwa K, Nakamori Y, Ohoyama K, Li HW, Noritake T, et
al. Experimental and computational studies on solvent-free rare-earth
metal borohydrides R(BH4)3 (R=Y, Dy, and Gd). Phys Rev B.
2008;77(10):104114.
[131] Rude LH, Corno M, Ugliengo P, Baricco M, Lee YS, Cho YW, et al.
Synthesis and Structural Investigation of Zr(BH4)4. J Phys Chem C.
2012;116(38):20239-45.
[132] Ley MB, Jørgensen M, Černý R, Filinchuk Y, Jensen TR. From
M(BH4)3 (M = La, Ce) Borohydride Frameworks to Controllable
Synthesis of Porous Hydrides and Ion Conductors. Inorg Chem
2016;55(19):9748-56.
[133] Humphries TD, Ley MB, Frommen C, Munroe KT, Jensen TR,
Hauback BC. Crystal structure and in situ decomposition of
Eu(BH4)2 and Sm(BH4)2. J Mater Chem A. 2015;3(2):691-8.
[134] Olsen JE, Frommen C, Jensen TR, Riktor MD, Sorby MH, Hauback
BC. Structure and thermal properties of composites with RE-
borohydrides (RE = La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Er, Yb or Lu)
and LiBH4. RSC Adv. 2014;4(4):1570-82.
[135] Wegner W, Jaroń T, Grochala W. Polymorphism and hydrogen
discharge from holmium borohydride, Ho(BH4)3, and KHo(BH4)4. Int
J Hydrog Energy. 2014;39(35):20024-30.
[136] Olsen JE, Frommen C, Sorby MH, Hauback BC. Crystal structures
and properties of solvent-free LiYb(BH4)4-xClx, Yb(BH4)3 and
Yb(BH4)2-xClx. RSC Adv. 2013;3(27):10764-74.
[137] Broach RW, Chuang IS, Marks TJ, Williams JM. Metrical
characterization of tridentate tetrahydroborate ligation to a transition-
metal ion. Structure and bonding in Hf(BH4)4 by single-crystal
neutron diffraction. Inorg Chem 1983;22:1081-4.
[138] Hoekstra HR, Katz JJ. The Preparation and Properties of the Group
IV-B Metal Borohydrides. J Am Chem Soc. 1949;71(7):2488-92.
[139] Bernstein ER, Hamilton WC, Keiderling TA, La Placa SJ, Lippard
SJ, Mayerle JJ. 14-Coordinate uranium(IV). Structure of uranium
borohydride by single-crystal neutron diffraction. Inorg Chem
1972;11(12):3009-16.
[140] Charpin P, Nierlich M, Vigner D, Lance M, Baudry D. Structure of
the second crystalline form of uranium(IV) tetrahydroborate. Acta
Crystallogr C. 1987;43(8):1465-7.
[141] Banks RH, Edelstein NM, Spencer B, Templeton DH, Zalkin A.
Volatility and molecular structure of neptunium(IV) borohydride. J
Am Chem Soc. 1980;102(2):620-3.
[142] Hagemann H, Longhini M, Kaminski JW, Wesolowski TA, Černý R,
Penin N, et al. LiSc(BH4)4: A Novel Salt of Li+ and Discrete
Sc(BH4)4− Complex Anions. J Phys Chem A. 2008;112(33):7551-5.
[143] Jaroń T, Wegner W, Fijałkowski KJ, Leszczyński PJ, Grochala W.
Facile Formation of Thermodynamically Unstable Novel
Borohydride Materials by a Wet Chemistry Route. Chem Eur J.
2015;21(15):5689-92.
[144] Černý R, Chul Kim K, Penin N, D’Anna V, Hagemann H, Sholl DS.
AZn2(BH4)5 (A = Li, Na) and NaZn(BH4)3: Structural Studies. J Phys
Chem C 2010;114(44):19127-33.
[145] Dovgaliuk I. PhD thesis: Synthesis, structure and properties of Al-
based borohydrides for hydrogen storage. http://hdl.handle.net/
2078.1/169194: Prom.: Filinchuk, Yaroslav; Devillers, Michel; 2015.
[146] Lindemann I, Domènech Ferrer R, Dunsch L, Filinchuk Y, Černý R,
Hagemann H, et al. Al3Li4(BH4)13: A Complex Double-Cation
Borohydride with a New Structure. Chem Eur J. 2010;16(29):8707-
12.
78 P. Vajeeston
[147] Eustis SN, Whiteside A, Wang D, Gutowski M, Bowen KH.

Ammonia−Hydrogen Bromide and Ammonia−Hydrogen Iodide
Complexes: Anion Photoelectron and ab Initio Studies. J Phys Chem
A. 2010;114(3):1357-63.
[148] Payandeh GharibDoust S, Heere M, Sorby MH, Ley MB, Ravnsbaek
DB, Hauback BC, et al. Synthesis, structure and properties of new
bimetallic sodium and potassium lanthanum borohydrides. Dalton
Trans. 2016;45(47):19002-11.
[149] Wegner W, Jaron T, Grochala W. MYb(BH4)4 (M = K, Na)
from laboratory X-ray powder data. Acta Crystallogr C.
2013;69(11):1289-91.
[150] Černý R, Ravnsbæk DB, Severa G, Filinchuk Y, D’ Anna V,
Hagemann H, et al. Structure and Characterization of KSc(BH4)4.
J Phys Chem C. 2010;114(45):19540-9.
[151] Jaron T, Grochala W. Probing Lewis acidity of Y(BH4)3 via its
reactions with MBH4 (M = Li, Na, K, NMe4). Dalton Trans.
2011;40(48):12808-17.
[152] Sadikin Y, Stare K, Schouwink P, Brix Ley M, Jensen TR, Meden A,
et al. Alkali metal – yttrium borohydrides: The link between
coordination of small and large rare-earth. J Solid State Chem.
2015;225(Supplement C):231-9.
[153] Brighi M, Schouwink P, Sadikin Y, Černý R. Fast ion conduction in
garnet-type metal borohydrides Li3K3Ce2(BH4)12 and
Li3K3La2(BH4)12. J Alloys Compd 2016;662(Supplement C):388-95.
[154] Schouwink P, Didelot E, Lee YS, Mazet T, Černý R. Structural and
magnetocaloric properties of novel gadolinium borohydrides. J
Alloys Compd 2016;664(Supplement C):378-84.
[155] Dovgaliuk I, Ban V, Sadikin Y, Černý R, Aranda L, Casati N, et al.
The First Halide-Free Bimetallic Aluminum Borohydride: Synthesis,
Structure, Stability, and Decomposition Pathway. J Phys Chem C.
2014;118(1):145-53.
[156] Jaron T, Wegner W, Grochala W. M[Y(BH4)4] and M2Li[Y(BH4)6-
xClx] (M = Rb, Cs): new borohydride derivatives of yttrium and their
hydrogen storage properties. Dalton Trans. 2013;42(19):6886-93.

[157] Soulié JP, Renaudin G, Černý R, Yvon K. Lithium boro-hydride

LiBH4: I. Crystal structure. J Alloys Compd 2002;346(1):200-5.
[158] Ravnsbæk Dorthe B, Filinchuk Y, Cerný R, Jensen Torben R.
Powder diffraction methods for studies of borohydride-based energy
storage materials. Zeitschrift für Kristallographie Crystalline
Materials 2010. p. 557.
[159] Nickels EA, Jones MO, David WIF, Johnson SR, Lowton RL,
Sommariva M, et al. Tuning the Decomposition Temperature in
Complex Hydrides: Synthesis of a Mixed Alkali Metal Borohydride.
Angewandte Chemie International Edition. 2008;47(15):2817-9.
Chapter 2
BONDING NATURE AND ITS IMPLICATIONS

WITH THE ENERGY STORAGE AND
IONIC TRANSPORT IN HYDRIDES
P. Vajeeston* and H. Fjellvåg

Center for Materials Science and Nanotechnology,
Department of Chemistry, University of Oslo, Oslo, Norway
ABSTRACT
Hydrides are an important class of materials due to their potential

application for the storage of energy and for the transportation sector. The
main focus of studies on hydrides has been so far the acceleration of
kinetics for the hydrogenation/dehydrogenation processes at moderate
temperatures, and the increase of the weight percentage of hydrogen. The
hydrides characterized by relatively slow kinetics for the hydrogenation/
dehydrogenation processes and by a high decomposition temperature are
therefore of scarce relevance for hydrogen economy. However, selected
hydrides have recently been suggested to find unique exciting applications
in optoelectronics and as solid electrolytes for battery applications. Their
use in optoelectronic devices depends on their stability with respect to
*
Corresponding Author address. Email: ponniahv@kjemi.uio.no Web: folk.uio.no/ponniahv.
82 P. Vajeeston and H. Fjellvåg
doping, solubility of shallow donors and acceptors, and electrical as well

as optical properties. In general, the hydrides are widely available in
powder form, which is required for hydrogen storage and not preferable for
electronic device technology. For the latter applications, crystalline or
amorphous hydrides in solid/thin film forms are preferable. Moreover, the
amphoteric behavior of hydrogen and its interaction with the environment
are known to hinder the optimization of optoelectronic properties of
hydrides from an experimental perspective. These are currently open
problems and comprehensive theoretical and experimental studies are
required to develop a further understanding. Classically, the chemical
bonding in solids are classified as ionic, covalent, and metallic. However,
the bonding in a real system is a combination of these three classes. The
results of theoretical electronic-structure calculations can be readily used
to study the nature of chemical bonds. In general, the distribution of
electron bands in reciprocal space and the distribution of states in energy
[the density of states (DOS)] provide insight on the bonding in a given
material. By analyzing the DOS, decomposed according to crystal site and
angular momenta, one can distinguish between ionic and covalent bonds in
solid systems. The nature of the bonding can be analyzed using several
other tools such as valence-charge density plot, charge-transfer plot,
electron-localization function (ELF), crystal-orbital Hamilton population
(COHP), Bader and Mulliken atomic charges that are derived from the
electronic structure calculations. In this chapter we will describe the nature
of the bonding in hydrides, and show how these affect the properties of
these materials, focussing on application in the energy storage and in the
transportation sector.
Keywords: chemical bonding, nano phase hydrides, electronic structure,

ionic conductors
INTRODUCTION
Though the periodic table has only 118 or so elements, there are
obviously more substances in nature than 118 pure elements. Compounds
are formed when two or more atoms chemically bond together, the resulting
compound is unique both chemically and physically from its parent atoms.
A stable compound occurs when the total energy of the combination has
lower energy than the constituent atoms. The properties of a compound are
determined by the chemical bonds that hold the constituents together. A
Bonding Nature and Its Implications with the Energy Storage … 83
chemical bond is an attractive force that exists between certain atoms in a

substance. However, in addition to these attractive forces, there must be
repulsive forces because solids are not easily compressible. The
attractive/repulsive forces between the constituents in solids are basically of
an electronic origin and the classifications of the different types of bonding
are strongly dependent on the electronic structure.
According to the strength and directionality, chemical bonds are
grouped into primary and secondary. By virtue of their nature, primary
bonds are inter-atomic and secondary bonds are intermolecular. Experience
from numerous studies of solids has revealed that there are three strong
principal types of primary bonds: ionic, covalent, and metallic. The bonding
in hydrides are quite complex, hence in the following section we
demonstrate how one can understand the chemical bonding behaviour
microscopically, first some simple representative systems are studied then
the same approach is applied to more complicated real systems.
In an idealized ionic bonding, electrons are completely transferred from
one atom to another. The electron transfer process goes from electro-positive
atoms to electro-negative atoms. The electronegativity difference between
the constituents is commonly taken as a measure of the ionicity in solids. In
the process of either losing or gaining negatively charged electrons, the
reacting atoms form ions. The oppositely charged ions are attracted to each
other by electrostatic forces, which constitute the basis of the ionic bond.
The often considered example of this type of bonding is NaCl, where Na
donates one electron to Cl (see Figure 1a, and 1g). It should be noted that,
when sodium loses its one valence electron it becomes smaller in size, while
chlorine grows larger when it gains an additional valence electron. This is
typical for the relative sizes of ions compared to that in atoms. Positive ions
tend to be smaller than their parent atoms while negative ions tend to be
larger than their parent atoms. After the charge transfer reaction takes place,
the charged Na+ and Cl ions are held together by electrostatic forces, thus
forming an ionic bond. Hydrogen is a peculiar element and it has amphoteric
behaviour in solids. So, it becomes anion or cation in ionic solids depending
upon the chemical environment. However, hydrogen mainly having cationic
behaviour in hydrogen storage materials considered in this chapter. It may
be noted that a system with strong ionic bond usually possesses high
symmetric structures.
Another major type of bonding in hydride is the covalent category where
electrons are equally shared between atoms. As opposite to ionic bonding in
which a complete transfer of electrons occurs, covalent bonding occurs when
two (or more) atoms share their electrons. An example of this type of
bonding is found in diamond (Figure 1b, h). Hydrogen is covalently bonded
to the neighbours in most of the organic molecules. The covalent bonding
has a quantum mechanical origin and occurs because the atoms in a given
compound have a similar “desire” for electrons. In order to have covalent
interaction between electrons they should be energetically and spatially
degenerate and also their spins should be “opposite”. Unlike ionic
compounds, covalent molecules exist as true molecules. Because electrons
are shared in covalent bond, it is the strongest bond in solids. Further, due to
the sharing of electrons, there is no net charge build-up in atomic sites and
hence covalent molecules are not strongly attracted to one another. As a
result, covalent molecules move about freely and tend to exist as liquids or
gases at room temperature. As the covalent bond is having directional
character, the covalently bonded solids often possess low symmetric
structures.
The third type of bonding is called metallic, where the valence electrons
are delocalized, i.e., homogeneously distributed in the space in between the
atoms. The properties of metals suggest that their atoms possess strong
bonds, yet the ease of conduction of heat and electricity suggest that
electrons can move freely in all directions in a metal. The general
observations give rise to a picture of “positive ions in a sea of electrons” to
describe metallic bonding. Examples of this type of bonding is found most
typically in alkali metals where each atom contributes only one electron to
the common pool (see Figure 1c, i, o for Na as an example). The materials
with metallic bond usually have metallic behaviour i.e., they possess good
thermal and electrical conductivity as well as Fermi surface in contrast to
ionic or covalent solids where semiconducting or insulating behaviour
prevails.
The bonding nature of materials generally becomes more complex when

the number of constituents increases. For binary compounds one can easily
calculate the approximate degree of ionic character from their
electronegativities using Gordy’s approximation. [1] For ternary and
quaternary systems it is usually very hard to estimate the appropriate ionic
character. If we consider hydrides, it is even more complicated, because of
the small atomic size and the only one valence electron of the hydrogen.
Moreover, hydrogen takes the valence of +1, 0, or 1 depending upon the
chemical environment which complicates further attempts to quantify the
ionicity in hydrides. Several attempts have been made to more rigorous and
quantifiable way classify the bonding in materials using, e.g., the
electronegativity scale of Pauling [2], the population analysis of Mulliken
[3], the classification scheme using gradient regions of the electron density
by Bader [4], Born effective charges [5], visualization by electron
localization function (ELF)[6], bond strengths according to the crystal
orbital Hamiltonian population (COHP) [7], and bond overlap population
(BOP), etc. When one considers hydrides, bonding analyses are very
important to understand unusual behaviors, like anisotropic volume
expansion during the hydrogenation; hydrogen induced structural
transitions, amorphization, metal-to-insulator or insulator-to-metal
transition, magnetic transitions, etc. By applying only one bonding analysis
tool alone (density of states, charge distribution, charge transfer, etc.) one
cannot expect to obtain the complete picture of the chemical bonding. In
order to have a more precise impression of the bonding nature between
constituents in multi-component hydrides several approaches have to be
involved.
CLASSIFICATION OF THE HYDRIDES
In general, hydrides are classified according to the nature of the bonding

of hydrogen to the host lattice as covalent or volatile, saline or ionic, and
metallic, each of these classes bearing distinct characteristics. Such a
classification does not always clarify the typical features of the compound
in question. For example, lithium hydride is classified as a saline hydride,
whereas it actually exhibits some covalent characteristics. The rare-earth
hydrides are usually classified under the metallic hydrides, whereas they
exhibit some characteristics similar to those of volatile and saline hydrides
(e.g., noticeable charge transfer interaction with high heat of formation). In
solids, hydrogen can occupy the interstitial or substitutional sites depending
upon the chemical environment which changes the bonding behaviour
substantially.
Ionic Hydrides
Ionic hydrides are formed by reaction between a strongly electro-

positive cation metals (e.g., alkali or alkaline-earth) and hydrogen, which
becomes strongly electronegative because of the electron transfer. In
contrast to the covalent hydrides, the bonding in the ionic hydrides results
from the strong electrostatic forces between the dissimilar charges of the two
kinds of ions. The ionic hydrides are usually well crystalline, exhibit high
heats of formation and high melting points. They are all insulators and
commonly colorless. There is little or no solid solubility of hydrogen in
saline hydrides, keeping the stoichiometry of these substances strictly
valence composed (in contrast to metal hydrides). The lattice parameters for
the structures of the alkali-metal hydrides range from 3.14 (LiH) to 5.41 Å
(CsH) where the original parent metal lattices have been substantially
contracted by the insertion of the hydride anion into its octahedral holes. All
alkali-metal hydrides contain fcc cation lattices revealing the NaCl-type
structure. The ionic alkali and alkaline earth hydrides are denser than the
pure alkali (45 to 75%) and alkaline-earth metals (20 to 25%). Consequently,
the bulk modulus of the hydride phases are higher than the corresponding
pure metals.
Covalent Hydrides
Covalent hydrides are found in the solid, liquid or gaseous phases. The
bonding between hydrogen and the other constituent(s) in this class of
hydride is characterized by sharing of the valence electrons on a fairly equal
basis. In general, molecules with covalent bonding are not strongly attracted
to each other and the absence of strong intermolecular forces results in a
degree of volatility and a low melting point. Covalent hydrides are generally
somewhat thermally unstable and this instability increases with increasing
atomic weight of the non-hydrogen element(s). In covalent hydrides,
hydrogen can formally be regarded to appear in the oxidation states +1 or –
1 of which +1 is the most abundant occurrence (as in the single bond between
O and H in H2O). Electron deficient molecules usually have hydrogen in the
formal valence state –1 and these are also known to form hydrogen bridge
bonds similar to those found in diborane (B2H6). Typical covalent hydrides
are carbo-hydrides, borohydrides, germanium hydrides, etc. The kinetics of
hydrogenation and dehydrogenation reactions among members in this group
are usually slow and generally follow a few specific mechanisms if at all
any. Metastable forms of covalent hydrides are to be considered common
observations rather than exceptions. Covalent hydrides generally exhibit low
symmetric structures. In general, most of the complex hydrides (with higher
H wt%) are having dominant covalent characters.
Metallic Hydrides
Metallic hydrides are normally formed by the transition metals (e.g.,

FeTiHx, VHx, ScH2). They generally exhibit the characteristic metallic
properties like high thermal and electrical conductivity, hardness, luster etc.
Because of the wide homogeneity ranges adopted by metallic hydrides, they
have sometimes been considered as solid solutions of hydrogen in the
interstitials of metal, alloy or intermetallic matrixes. The crystal structures
of metallic hydrides may also be distinctly different from those of the parent
matrices. Metal hydrides are characterized by a rather high hydrogen bulk
diffusion rate and hence they find application in rechargeable batteries.
Many of these materials are exceedingly complicated, showing
compositional non-stoichiometry and sometimes complex phase relations.
Several metal hydrides exhibit a low-hydrogen-concentration phase where
the hydrogen ions occupy a small portion of the available interstitial sites.
The lattice strain in such phases is low enabling the metal to retain its
macroscopic shape. However, larger percentage of hydrogen filling into the
interstitial sites brings lattice stresses which results in the formation of
micro-meter sized hydride particles. Since the metals which form the metal
hydrides are heavy and take low valence states (due to the low
electronegativity of hydrogen), the result is a low weight percentage of
stored hydrogen. The energy storage devices based on metal hydrides are
consequently not suitable for mobile hydrogen storage applications.
CHARACTERIZATION OF BONDING IN
HYDRIDES EXPLORED BY THEORETICAL TOOLS
The chemical bond arises in reality from a redistribution of the valence

electrons of the constituent atoms in a way that it is energetically favourable
(nature seeks to minimize energy). As mentioned in the previous section, the
real bonding nature of hydrides is quite complex. In order to understand this
behaviour in detail, we will start with an analysis of well-known classical
examples: NaCl for the ionic case, C in the diamond structure for the
covalent case and Na for the metallic case and then extend the analysis to
simple hydrogen containing compounds. In the hydride cases, we have
chosen LiH for ionic, B2H6 for the covalent, and ScH2 for metal hydride as
examples.
Figure 1. Calculated valence-electron-density maps (a), (b), (c), (d), (e), and (f); charge
transfer plots (g), (h), (i), (j), (k), and (l); and ELF plots (m), (n), (o), (p), (q), and (r) for
NaCl, C (diamond structure), Na, LiH, B2H6, and ScH2, respectively. Atoms are labelled
on the illustration scales are given on the right hand side.
Charge Density
An understanding about the nature of the chemical bond can be aided by

the studies of the distribution of charges in real space. The real space charge
density can also be used to understand features of the electronic properties
of materials. One has to remember that the charge density very much
depends on the crystal structure and different bonding characteristics may
be formed for one and the same compound in different structural
arrangements.
The charge density distribution in the model substances NaCl, C, and
Na are shown in Figure 1 a, b, and c, respectively. The charge-density
distribution for the simple hydrides LiH (ionic), B2H6 (covalent), and ScH2
(metallic) are also shown in Figure 1 d, e, and f, respectively, which clearly
reflect the overall picture of ionic (spherically symmetric charge distribution
at the atomic sites with negligible charge distribution between atoms),
covalent (charges are also present in between the B-B and polarised
electronic distribution around the H site) and metallic bonding (uniform
charge distribution at the interstitial region), respectively. However, the
charge-density distribution in ScH2 cannot be simply viewed as that for a
pure metallic phase as Figure 1f conveys significant ionic and covalent
bonding contributions.
(a)
(b)
Figure 2. The valence electron charge-density plot between the Th and H atoms for
Th2AlH4 in the ac plane (through H) with 40 contours drawn between 0 and 0.25
electrons/a.u.3.
Charge density analysis has also been used to understand the anisotropic
volume expansion, the high decomposition temperature of certain hydrides,
and to explain the unusual short H-H separations and novel electrical, optical
and magnetic properties. In particular the charge density analysis is useful
to understand the formation of nanophases of solids and their
surface/interface characteristics. We have used charge density analysis along
with structure analysis to explain the anisotropic volume expansion in
Th2Al. The experimental [8, 9] and theoretical studies showed high
anisotropic changes in lattice expansion on hydrogenation of Th2Al.
According to the crystal structure of Th2Al, the inter-atomic distance
between the interstitial where one can accommodate H in the ab plane is
2.4 Å. Hence, a large flexible space is available for accommodation of H
atoms in the ab plane without a need to expand the lattice. In contrast, the
inter-atomic distance between the interstitial regions in the ac plane is only
1.65Å. So, a large expansion of the lattice along c is necessary to
accommodate H atoms within the ac plane. As a result, even with a short
H-H separation of 1.95Å, a lattice expansion of 12.41% along c-axis is
required when Th2AlH4 is formed from Th2Al. The experimental
observation of 0.105% lattice expansion along a and 12.15% along c is
found to be in excellent agreement with these considerations. The bonding
between Th and H is predominantly covalent as evidenced by the finite
charge between these atoms. The H-s electrons are tightly bound to the
Th-d states and the arrangement of Th and H atoms forms an H-Th-H dumb-
bell pattern (see Figure 2).
Charge Transfer
The charge density displays the chemical bonds in a given system, but
remains rather featureless with charge maxima around the atoms similar to
the charge density of overlapping free atoms. The charge-transfer contour is
the self-consistent electron density in a particular plane, comp, minus the
electron density of the corresponding overlapping free atoms, o.f.a as
defined below,
(r) = (r)comp-(r)o.f.a, (1)
which allows one to visualize how electrons are redistributed in a particular

plane of the compound compared to free atoms.
If one displays the charge transfer (instead of charge density) i.e., the
difference between the electron density of the compound and the electron
density of the constituent atoms, one obtains an impression of how electrons
are shifted around when the solid is formed. This allows one to see how the
actual chemical bond is formed in real space. So, the Figs.1g, 1h, and 1i
display the charge transfer plot of the representative compounds NaCl, C,
and Na. Similarly, Figure 1j, 1k, and 1l show the charge redistribution in the
hydrides LiH, B2H6, and ScH2 respectively. In these figures, we clearly see
the difference between the three principle types of bonding. In the Figs.1h
and 1k we see the nature of the ionic bond with large charge transfer from
the Na/Li (cations) to Cl/H (anions) sites. In Figure 1i we can see the
covalent bonding in diamond, with charges from surrounding atoms
redistributed in between the atoms with the characteristics of directional
bonds. Similarly, in B2H6 the charges are shared in B-B and B-H bonds like
that in organic molecules. In Figure 1j and m one can see well-dispersed
electrons around the metal atoms and in particular somewhat “uniform”
distribution in the interstitial regions. In Figure 1j charges are also depleted
from Sc site which clearly indicate that a finite degree of ionicity is also
present in this material. A similar type of bonding (mixture of metallic and
ionic) nature was identified in the series of RNiInH1.333 (R = La, Ce, Pr, Nd)
metal hydrides and most of the complex hydrides (like LiAlH4, LiBH4, etc.)
systems.
Electron Localization Function (ELF)
ELF is a ground-state property which discriminates between different

kinds of bonding interaction for the constituents of a solid. [6, 10, 11] In the
implementation for density functional theory (DFT), this quantity depends
on the excess of local kinetic energy tp originating from the Pauli principle.
1
 t  2 
ELF  1    
p(r )
(2)
  t p ,h  
 
where:
tp= - 1/8 [()2/] (3)
is the Pauli kinetic energy density of a closed-shell system. ()2/(8) is the

kinetic energy density of a bosonic-like system, where orbitals proportional
to  are occupied. tp is always positive, and for an assembly of fermions,
it describes the additional kinetic energy density required to satisfy the Pauli
principle. The total electron density ()
n 2
   i (4)
i
As well as the kinetic energy density ()
2
1 n
    i (5)
2 i
are computed from the orbitals,  i . In both equations, the index i runs over
all occupied orbitals.
According to Eqn.1, the ELF takes the value one either for a single-
electron wave function or for a two-electron-singlet wave function. In a
many-electron system, ELF is close to one in regions where electrons are
paired such as covalent bonds, or for unpaired lone electrons of dangling
bonds, while the ELF is small in low-density regions. In a homogeneous
electron gas system ELF equals 0.5 at any electron density and ELF values
of this order in homogeneous systems indicates regions where the bonding
has a metallic character. The ELF distributions of typical examples for the
bonds in the three test cases are given in Figure 1n-p. For the NaCl case ELF
is around 0.9 at the Cl site and only a non-significant ELF value is found at
the Na site, thus clearly reflecting ionic bonding. Further, the spherically
symmetric distribution of ELF at the atomic sites along with negligible ELF
in between Na and Cl are further confirmation for ionic bonding. In the
diamond case, a ELF value of around one is found where the shared electrons
are present (in regions between the C atoms). Moreover, the ELF distribution
around and between the C atoms are anisotropic and this is the typical feature
of covalently bonded solids. In the third case where the metallic bonding is
dominant, a uniform distribution of ELF is present around the interstitial
region with the value of nearly 0.5 where the delocalized free electrons are
found. From the above examples, one can clearly visualize different types
of bonding in simple solids based on results from DFT calculations. We
extended the corresponding analyses to hydrides with different bonding
interactions and found almost similar ELFs for LiH as NaCl and for B2H6
like C. In the metallic ScH2 hydrides, we obtain lower value of ELF
compared to that in metallic Na. It may be noted that our test calculations
for the transition metals such as Ni, Co, and Cu also show such a low ELF
values and this is the general characteristic for materials with d electrons in
the valence state. We have used ELF to interpret the short H-H separation in
the RTInH1.333 (R = La, Ce, Pr, Nd; T = Ni, Pd, Pt) series [12-14]. ELF based
analysis is also conducted for the complex hydrides in order to identify the
reason for their high decomposition temperature (for example see ref. [15]).
Prediction of Site Preference of Hydrogen in Metal Matrix

by ELF
The ELF analysis is a powerful method to identify the position of H

atoms within a metal matrix. From our systematic study on 95 members of
the ZrNiAl-family we have constructed a site-preference rule. [16] A short
introduction about the origin of this rule is provided here. In metal hydrides
the occupation of hydrogen at the interstitial sites depends upon various
parameters such as the size of the interstitial void, the coordination of the
interstitial sites, chemical behaviour of the surrounding atoms, crystal
structure of the lattice etc. If we visualize the ELF distribution in a

nonhydrated metallic phase, we can see maximum value of the ELF
(larger than 0.5) in a particular interstitial site as an indication that the
chemical environment in that place is such that those electrons have
relatively more paired nature (i.e., nonbonding localized electrons) than
other interstitial sites. A hydrogen atom diffuses into the metal matrices with
one unpaired electron. The average value of the ELF indicates that the
chemical environment of such an interstitial site encourages electrons to
have relatively more paired nature (i.e., nonbonding localized electrons)
compared to other interstitial sites. Hydrogen therefore tries to occupy sites
in the metal matrix where it can find electrons to participate in the bonding
interaction such that it can achieve a more stable 1s2 paired electron. The
maximum value of ELF (larger than 0.5) in the interstitial regions indicates
that nonbonding localized electrons at these regions can easily enter to the
H sites than electrons from other interstitial sites. This is the physical origin
for the site preference of H in interstitial sites where ELF has the maximum
value. Another mechanism for such a process could be that hydrogen atom
on entering the interior of the metal matrix gives up its own electron to the
collective interstitial electron well [thus forming an H+ ion (viz a bare
proton)], undergoes lattice diffusion, and arranges itself on an appropriate
interstitial site with high ELF, where it can attract nonbonding localized
electrons and establishes a paired state. The net outcome of the two
processes is the same, H gains one electron from the host lattice and
establishes H− with a paired s2 state configuration and thereby completes its
valence shell. This complies with our finding that charge-transfer plots in
metal hydrides always show electron transfer from the host lattice to H. The
higher ELF values in certain interstitial regions signal that electrons in these
regions have appropriate spin available for completion of the valence shell
of the incorporated hydrogen atoms. If more than one interstitial sites show
large nonbonding localized electrons in a particular matrix, it seems likely
that first hydrogen atoms will go to the site with the largest amount of
nonbonding localized electrons. As a result of the filling of this site, electron
localization on other sites may either increase or decrease. In the former case
a new site may be available for hydrogen filling, say, when exposed to high
hydrogen pressures. This type of analysis has been used to identify the H
position in metal matrix in several systems. [17-21]
The ZrNiAl-type structure (see Figure 3a; maintaining the general
formula RTM) has eight different interstitial sites centered around 2d, 2e, 3f,
3g, 4h, 6i, 6k, and 12l. Of these, the tetrahedral holes around 4h (surrounded
by 3R and 1T), the trigonal-bipyramidal holes around 2d (surrounded by 3R,
1T, and 1M), and the distorted octahedral holes around 3g (surrounded by
3R, 1T, and 2M) are the most interesting in relation to hole size [22] and
symmetry. We have searched the site preference of hydrogen in all these
possible interstitial sites for around a hundred ZrNiAl-type compounds.
Even though we have examined a large number of ZrNiAl-type phases, it is
convenient to concentrate the presentation of the findings on two typical
examples, where hydrogen prefers to occupy different sites. From the total
energy calculation one can identify the likely hydrogen position. In a
particular framework which contains two or more possible sites for location
of hydrogen, we must perform total energy calculations for all the
alternatives. The configuration with the lowest total energy will determine
the actual H position in the framework under consideration. The electron
localization function is another useful tool to spot hydrogen positions in a
particular matrix. [10] Recently we have investigated hydrogen position in
ZrNiAl-type structures, where H occupies different interstitial positions
depending upon the chemical environment (see Figure 3a-f). For example,
H occupies the 2d position in ZrNiAl (Figure 3g and 3h) with the
composition ZrNiAlH0.6667 whereas H occupies the 4h position in LaNiIn
(Figure 3i and 3j) with the composition LaNiInH1.3333. We have correctly
reproduced the experimental findings for these phases and also predicted
hydrogen position in several other compounds of the ZrNiAl series. This
type of theoretical exploration is sometimes very useful for experimentalists
working on crystal structure determination (or to settle contradictions
between different experiments). [23]
Figure 3. The ZrNiAl-type crystal structure (RTM). Legends for the different kinds
of atoms are given in the illustration. The 2d (trigonal-bipyramidal interstice), 4h
(tetrahedral interstice), and 3g (distorted octahedral interstice) sites are indicated by open
circles and pointed by arrows. (b) The empty trigonal-bipyramidal interstice. (c) The
empty tetrahedral interstice. (d) The trigonal-bipyramidal interstice filled with hydrogen.
(e) Two face-sharing tetrahedral interstices filled with hydrogen. (f) The empty
(distorted) octahedral interstice. Calculated ELF in the (001) plane for (g) ZrNiAl, (h)
ZrNiAlH0.667, (i) LaNiIn, (j) LaNiInH1.333. Different crystallographic sites (2d, 4h, and
3g) are marked by arrows. The iso-surface values correspond to 0.6 ELF value. Legends
for the different kinds of atoms are given in the illustration.
Density of States (DOS)
According to the band theory of solids [24, 25], when atoms are brought
together to form a crystal, the discrete electronic energy states of the isolated
atoms merge into energy bands which represent the allowed energies for
electrons in the crystal. These bands may be separated by forbidden regions
or gaps. Based on distribution of electrons in the allowed energy bands the
conductivity of solids changes and hence they can be classified as metal,
semiconductor or insulator. Electrons contained in a filled band make no

contribution to the electrical conductivity. Thus, if the valence electrons
exactly fill one or more bands leaving others empty, the crystal will be an
insulator; if the valence electrons partially fill one or more bands the crystal
will be a conductor. In an insulator at temperatures above 0 K, some
electrons from the highest valence band are thermally excited into the lowest
empty band following the Fermi-Dirac distribution function and conduction
becomes possible (intrinsic semiconduction). The number of electrons
excited into the conduction band is a function of both the temperature and
the magnitude of the energy band gap (Eg), which is defined as the separation
between the maximum energy in the valence band and the minimum energy
in the conduction band. Materials with an energy band gap in the range 0 <
Eg < 4 eV are called semiconductors and those having a gap Eg > 4 eV are
classified as insulators. Semiconductors with gaps below or near 0.5 eV are
called narrow band-gap semiconductors; materials with gaps between 2 and
4 eV are called wide band-gap semiconductors. In an insulator or
semiconductor there should be no electrons at the Fermi level (EF) at 0K and
the valence and conduction bands are separated by a finite Eg. Materials with
Eg close to zero are called semi-metals.
One can find correlation between bonding interactions and the electrical
conductivity. For example, covalent type of interaction manifests (usually
found in semiconducting phases) themselves as localized and energetically
degenerate energy levels in certain region of the DOSs. In an ionic
compound, the DOS feature is energetically well separated and localized
(often insulators). In a metallic compound the DOS will have non-localized
electron viz. a finite number of electrons at EF.
The band gap values for hydrides obtained from ab initio calculation are
given in Table 1. It is well known that the band gap values of solids obtained
from usual DFT calculations are systematically underestimated due to
discontinuity in the exchange correlation potential. Thus the calculated band
gap values are commonly 30–50% smaller than those measured
experimentally. The results given here do not include rigid shift technique,
quasiparticle or other corrections were used to reduce the DFT-related band
gap error. Thus, the real values of the band gaps of the studied hydrides are
larger than the calculated values presented in Table 1. Systematic

experimental studies are needed to clarify this issue.
Table 1. The calculated and experimentally observed fundamental

band gap (Eg, in eV) for selected hydrides. Most of the hydrides exist
in different polymorphs hence the structure-type is given along with
the space group number in the table
No. Compound Space group Band gap (Eg) Band gap type
1 -CaB2H2 Cmc21 (36) 0.32 indirect
2 SrSiAlH P3m1 (156) 0.65 (0.63)[26] indirect
3 Mg2CoH5 P4/nmm (129) 0.71 indirect
4 Sr2FeH6 FM-3M (225) 1.24 direct
5 Ca2FeH6 FM-3M (225) 1.41 direct
6 Mg2NiH4 C12/C1 (12) 1.65 indirect
7 NaBeH3 Pm-3m (221) 1.79 indirect
8 Mg2FeH6 FM-3M (225) 1.97 direct
9 -CaB2H2 P-3m1 (164) 1.98 indirect
10 Sr2RuH6 FM-3M (225) 2.19 direct
11 CsH FM-3M (225) 2.3 direct
12 -AlH3 R-3c (167) 2.53 (3.5)[28] indirect
13 SrLiH3 Pm-3m (221) 2.74 indirect
14 Na3AlH6 P 1 21/n1 (14) 3.0 direct
15 Mg2RuH6 FM-3M (225) 3.02 direct
16 KMgH3 Pm-3m (221) 3.13 indirect
17 BaAlH5 Pna21 (33) 3.38 direct
18 NaMgH3 Pnma (62) 3.45 direct
19 LiMgH3 R-3c (167) 3.98 indirect
20 NaAlH4 I41/aZ (88) 4.98 (6.41)[28] indirect
21 LiAlH4 P121/c1 (14) 4.76 indirect
22 MgH2 P42/mnm (136) 4.83 (5.6)[29] indirect
23 BeH2 Ibam (72) 5.30 direct
24 KH FM-3M (225) 5.51 direct
25 Mg(AlH4)2 P-3m1 (164) 5.60 indirect
26 KAlH4 Pnma (62) 5.97 indirect
27 KGaH4 Pnma (62) 6.05 indirect
28 NaBH4 Pnma (62) 6.21 direct
29 Mg(BH4)2 Pmc21 (26) 6.54 indirect
30 -LiBH4 Pnma (62) 7.0 indirect
Figure 4. The calculated total density of states for NaCl, C, Na, LiH, B2H6, and ScH2.
The valence bands are shaded; the Fermi levels are set at zero energy and marked by
dotted vertical lines.
The calculated total DOS for the studied test cases are shown in
Figure 5. The magnitude of the Eg for these systems shows that both ionic
(NaCl, LiH) and the covalent compounds (C, B2H6) are of the non-metallic
type (the valence band and conduction bands are separated by 3.2 eV for
B2H6 and 4.5 eV for LiH). In Na and ScH2 systems a finite number of
electrons are at the EF in accordance with their metallic character. A good
example for the DOS of metallic hydride is found in LaNiInH1.333 and such
results are presented in the literature (see [12-14]). SrSiAlH is a typical

example for semiconducting hydride with an experimentally-identified band
gap value of 0.63 eV (the corresponding theoretical value is 0.65 eV).
[26, 27] The best example for the insulating case from the complex hydrides
is LiAlH4, where the valence and conduction bands are separated by 4.7 eV
(see Figure 5a). Analysis of Table 1 shows that fundamental band gap of the
hydrides can be varied in a wide range, from infrared to ultraviolet. Table 1
shows that fundamental band gap values for some selected hydrides are in
the visible range (1.8–2.8 eV). Moreover, few hydrides possess fundamental
band gap values in the near infrared region (1.20–1.8 eV). The hydrides with
band gap values mentioned in the above ranges can be classified as wide
band gap hydrides. There are few hydrides having band gap value in the
infrared region 0.19–1.2 eV and hydrides with band gap values in the energy
range 0.01–1.2 eV that can be classified as narrow band gap hydrides.
Figure 5. The calculated total density of states for LaNiInH1.333 (metallic; bottom
panel), SrSiAlH (semiconductor; middle panel), and LiAlH4 (insulator; top panel).
Fermi levels are set at zero energy and marked by dotted vertical lines.
In order to demonstrate the formation of insulating, semiconducting, and

metallic hydrides in multicomponent systems the representative examples
for these three cases are given in Figure 5. The calculated total and partial
density of states for α-Li3AlH6 are shown in Figure 6. There is a finite energy
gap between the valence and conduction bands similar to the other complex
hydrides. The partial DOS (PDOS) for Li shows (see Figure 6) that the s and
p states of -Li3AlH6 are energetically degenerate in the whole energy range,
while Al-s and -p states are well separated. Apart from the ionic contribution
to the bonding, the energetic degeneration of the Al-p and H-s states together
with the spatially favourable constellation of Al and H facilitates covalent
bonding between Al and H. Due to the electron transfer from the Li site to
the AlH6 unit the magnitude of the Li PDOS is much smaller than the Al
PDOS and the interaction between Li and H is ionic.
Figure 6. Calculated total and partial DOS for -Li3AlH6. The Fermi level is set at zero
energy and marked by the vertical dotted line and s states are shaded.
Crystal Orbital Hamiltonian Population (COHP)
COHP is an extremely useful tool to analyze bonding, antibonding and

nonbonding interaction between atoms. The simplest approach is to
investigate the complete COHP between the atoms concerned, taking all
valence orbitals into account. For a detailed description of the COHP
method, the reader is referred to the original literature. [7] Implications of
COHP is that a negative value indicates bonding, positive antibonding and
zero value together with a finite DOS represent nonbonding states. The
COHP values depend on the number of states in a particular energy interval.
The integrated COHP (ICOHP) curve up to the EF represents the total
overlap population of the bond between atoms in question and gives
information about the relative strength of bonds. The calculated COHP curve
for the model cases and the simple hydrides are shown Figure 8. Even
though, the ionic and covalent types of bonds have higher bond strength, the
COHP study gave maximum ICOHP for covalent compounds. For example,
in the studied model systems, diamond has maximum ICOHP value and
similarly, in considered hydrides B2H6 has maximum ICOHP. This is due to
non-directional nature of the ionic bonding, because of the overlap between
the orbital is very small in the ionic compounds.
In our research work, the bond strength of the RTInH1.333 (R = La, Ce,
Pr, Nd; T = Ni, Pd, Pt) series have been examined to see how efficiently one
can pack hydrogen in the RTIn matrix (see ref. [13]). An interesting
observation from the COHP analysis is that, when one includes hydrogen in
the RTIn matrix, the estimated ICOHP value for the T- In bond is changed
from ca. -0.8 to -1.1eV. This appears to reflect transfer of electrons from
both T and In to the H site which considerably reduces the ionic interaction
between T and In and to compensate the charge transfer effect the covalency
between T and In during hydrogenation. Moreover, one can explain the
reason for the short H-H separation in the RTInH1.333 matrix and
quantitatively evaluate different types of bonding situations in these phases
using this approach.
Figure 7. Calculated COHP curves for (a) NaCl, (b) LiH, (c) diamond, (d) B2H6, (e)
Na, and (f) ScH2. Fermi levels are set at zero energy and marked by dotted lines.
We have also extensively studied the COHP in complex hydrides in

order to find suitable hydride phases with low decomposition temperature
(Ref. [15]). One should weaken the bonding interaction between H with
neighbours within the complex to use them for mobile applications. We have
made extensive studies on AXH4 (A=Li, Na, K, Rb, Cs; X = B, Al, Ga) series.
The most notable feature from this study is the remarkable strength of the
Ga-H interaction [-2.6 to -3.4eV in ICOHP on going from Li to Cs; similar
strength is observed for all X-H (X = B, Al, Ga) interactions in this series]
compared to the other interactions. Moving from A = Li to Cs the X-H
interaction (covalent bond strength) is reduced (both in the Al and Ga series).
According to the calculated ICOHP for these phases, the strength of the Al-
H bond is stronger than that of the Ga-H bond. This indicates that the
stability (the decomposition temperature) in the Al series is higher than that
in the Ga series. It implies that the partial substitution of Li by other alkaline
elements in materials like Li1-xNaxAlH4, Li1-xKxAlH4, or Li1-xKxAl1- yGayH4
may give rise to phases with lower decomposition temperatures (for more
details see Ref. [15]).
QUANTIFICATION OF CHARGES
The quantification of the charges associated with atoms/ions in

multicomponent systems is one of the most challenging tasks in theoretical
chemistry. The net atomic charge in each site is not directly observable in
quantum mechanical calculation, it is an arbitrarily observable quantity, but
one that is nonetheless useful in the development of understanding from
quantitative results. There are many different definitions of charge obtained
from ab initio calculations, the most famous one is that due to Mulliken, [3]
where overlap density is apportioned equally to both nuclear centers. More
recently, there has been considerable interest in Bader analysis [4] as a
means of partitioning electron density. Both of these aforementioned
schemes are based on the analysis of the density matrix, or the density itself.
However, there is another family of charge definitions that define the atomic
charges in terms of the response of the dipole moment of the system with
respect to perturbations. Since the dipole is a genuine observable, such
charge definitions are particularly attractive for force-fields as they describe
the charges that are consistent with the response of the material to atomic
displacements. In the following section we are going to see how one can use
these methods for analysing the bonding nature in materials.
Table 2. Mulliken population and Bader analysis for selected

compounds. The Mulliken effective charges (MEC) and Bader charges
(BC) are given in terms of e.
Compound Atom MEC Overlap population BC

C C 0.0 2.54 (C-C)
NaCl Na +0.94 0.01(Na-Cl)
Cl -0.94
ScH2 Sc +0.73 0.47 (Sc-H)
H -0.37
B2H6 B -0.28 0.52-0.99(B-H) +3.0
H +0.06 to 0.15 0.27 (B-B) -1.0
LiH Li +0.98 -0.003 (Li-H) +0.88
H -0.98 -0.88
CH4 C -0.26 0.384 (C-H) -0.34
H +0.06 +0.08
MgH2 Mg +1.87 -0.040 Mg-H +1.66
H -0.98 -0.83
BeH2 Be +1.63 0.045 Be-H +1.85
H -0.82 -0.92
AlH3 Al +2.22 0.091 Al-H +2.9
H -0.74 -0.97
LiAlH4 Li +1.01 0.171 Al-H +0.92
Al +2.01 -0.021 Li-H +2.98
H -0.75 -0.97
Li3AlH6 Li +1.01 0.105 Al-H +0.88
Al +2.08 -0.020 Li-H +2.80
H -0.85 -0.89
SrSiAlH Sr 1.85 2.72 (Al-Si)
Si -0.98 0.59 (Al-H)
Al -0.35 -0.23 (Sr-H)
H -0.52
Mulliken Population and Bader Charge Analyzes
In order to make a quantitative analysis, it would be useful if one could

identify the amount of electrons on a particular atom and the population of
electrons between all surrounding atoms concerned. Although there is no
unique definition of how to extract how many electrons are associated with
an atom in a molecule or a structural sub-unit of a solid, it has nevertheless
proved useful in many cases to perform population analyses. Due to its
simplicity the Mulliken population scheme has become the most familiar
approach to count electrons associated with a given atom. However, this
method also rather qualitative than quantitative; the results obtained from
this approach are sensitive to the atomic basis. Mulliken effective charges
(MEC; defined as the difference between the Mulliken charge and the atomic
charge) are listed in Table.2 for representative materials which may provide
benchmarks for systems with well recognized chemical bonding. For a pure
ionic NaCl the calculated MEC clearly reflect that Na donates one electron
to the Cl atom and the bond overlap population (BOP) between Na and Cl is
almost zero. LiH is a reasonably pure ionic compound and the calculated
Mulliken charges reflect approximately the expected ionic picture with Li+
and H-. The overlap population between Li+ and H- is also close to zero, as
expected for an ionic compound. Similarly, in MgH2, BeH2, and AlH3 the
bonding interaction is mainly ionic, but the degree of ionicity is reduced
from MgH2 to BeH2 and further to AlH3, viz. these compounds exhibit some
covalent character as evidenced by the non-zero overlap population.
Similarly, for the covalently bonded diamond, the calculated BOP value is
much higher. For the CH4 molecule, the overlap population takes a value of
0.384 consistent with the well known covalent character of this molecule.
The Mulliken effective charges for Li, Al, and H in LiAlH4 and Li3AlH6
indicate that the interaction between the Li and AlH4/AlH6 is ionic
(one electron transferred from Li to AlH4/AlH6).
In the Bader charge (BC) analysis, each atom of a compound is
surrounded by a surface (called Bader regions) that run through minima of
the charge density and total charge of an atomic site is determined by
integration within the Bader region. The calculated BC for the selected
compounds is given in Table 2. For LiH phase, similar to the Mulliken
analysis, BC also reflects an almost pure ionic picture i.e., around one
electron transferred from Li to H. This finding is consistent with the MEC
and charge density analyses. However, in most of the cases the magnitudes
of the charges obtained from MEC and BC analysis are considerably
different from each other. This is partly due to the artefact of Bader’s “atom
in molecule” approach that involves making the boundaries to integrate
charges in each atomic basin and, also the basis set dependent behaviour of
MEC. However, consistent with the charge, charge transfer, and ELF
analyses, the BC always qualitatively shows the bonding nature in materials.
We thus find that the bonding nature of the hydrides (both metallic and
complex) does not exhibit a simple ionic or covalent character. In fact, the
bonding present in these compounds have mixed bonding character. In the
complex hydrides like LiAlH4 and Li3AlH6 the interaction between Li and
AlH4/AlH6 is ionic whereas the interaction between Al and H comprises of
both ionic and covalent character. A similar type of bonding situation
prevails in all complex hydrides, but the magnitude of the ionic/covalent
mixture will exhibit considerable individual variation.
Born-Effective-Charge Analysis
As we have mentioned in the beginning of this section, Born-effective-

charge (BEC) analysis is another tool to evaluate bonding characteristics.
The most widely used definition for the Born effective charge is:
 
q Born
 (6)
 i
i
were  and  are the Cartesian directions. The charge of an atom is a

symmetric (3 × 3) tensorial quantity, since it describes the derivative of the
three Cartesian components of the dipole moment µ with respect to the three
Cartesian atomic displacements. While the definition of the dipole moment
is complicated by the choice of the particular images of each ion, this
complication does not affect the derivatives of the dipole moment within the
present model. Differentiation of the dipole moment leads to the following
expression for the Born effective charge:
qiBorn  qicore   ( Dcoreshell Dshell

1
 shell q
shell
)i (7)
where the first term on the right-hand side is the core charge of the ion and
the second term is the corresponding component of the product of the core-
shell second-derivative matrix with the inverse of the shell-shell second-
derivative matrix, scaled by the vector of shell charges, qshell. Physically, the
second term corresponds to the response of all the shells present to the
atomic displacement of atom i, or, in other words, the electronic
contribution. Hence, for a rigid ion model, the Born effective charge tensor
is equal to a diagonal matrix with all the diagonal elements equal to the core
charge. Increasingly, the Born effective charges are being determined from
ab initio calculations, thus creating a new avenue for determining shell
model parameters beyond the fitting of dielectric constants.
For the LiH and MgH2 cases (see Table 3) the diagonal components of
the effective charges in a Cartesian frame satisfy the relation Zxx = Zyy = Zzz,
whereas the off-diagonal components turn out to be negligible. This is as
expected for ionic compounds due to the spherical character of ionic bonds.
For LiAlH4 the diagonal component of the effective charges at the Li site is
also almost equal (Zxx ≈ Zyy ≈ Zzz) and the off-diagonal components are
negligibly small (see Table 3). At the Al and H sites the diagonal
components are also almost equal, but the off-diagonal components take
small but definite values. This probably originates from some exchange
owing to sharing of electrons between H and Al within the AlH4 units. The
BEC analyses accordingly reconfirms that the Li sites of the LiAlH4 series
give up nearly one electron, whereas the H atoms correspondingly gain 0.52-
0.89 electron each. This conclusion therefore largely concurs with the mixed
iono-covalent picture between the Al and H. A similar type of bonding
character was present in most of the complex hydrides (ex. alanates, boro
hydrides, amides, etc.). A methodological comment seems appropriate: Our
experience from studies of chemical bonding by both Mulliken-population
and Born-effective-charge analyses shows that these two approaches yield
almost the same conclusion. Our personal view is that Born-effective-charge
calculations normally demand larger computations. Hence, we recommend
Mulliken population analysis as a more suitable tool to probe bonding in

complex hydrides. However, it is very important to note that the Born
effective charge is originating from the polarization of charge by electric
field. So it contains both static charge arising from usual ionic charge in the
solid and the dynamical charge (arising from the covalency). BECs are used
to estimate the LO-TO splitting in solids and hence very important to
determine Γ-point phonon frequencies correctly.
Table 3. Calculated Born-effective-charge tensor elements (Z) for the

constituents of LiH, NaH (in FM3-M), -MgH2 (in P42/mnm), -AlH3
(in R-3cH), -LiAlH4 (in P21/c1), -NaAlH4 (in I41/a), Li3AlH6 (in R-
3), Na3AlH6 (in P21/n), and SrAlSiH (in P3m1) polymorphs. The
corresponding crystallographic sites are given in the second column
Position xx yy zz xy xz yx yz zx zy
LiH
ZLi 4a 1.06 1.06 1.06 0 0 0 0 0 0
ZH 4b -1.06 -1.06 -1.06 0 0 0 0 0 0
NaH
ZNa 4a 1.0 1.0 1.0 0 0 0 0 0 0
ZH 4b -1.0 -1.0 -1.0 0 0 0 0 0 0
MgH2
ZMg 2a 1.96 1.96 2.01 0.02 0.0 0.02 0.0 0.0 0.0
ZH 4f -0.98 -0.98 -1.01 -0.02 0.0 0.02 0.0 0.0 0.0
-AlH3
ZAl 6b 2.72 2.72 2.96 -0.6 0 0.62 0 0 0
ZH 18e -0.77 -1.04 -0.99 -0.23 -0.23 -0.23 -0.39 -0.23 -0.39
-LiAlH4
ZLi 4e 1.09 1.14 1.1 0.16 -0.02 -0.01 -0.06 -0.01 0.04
ZAl 4e 1.54 1.69 1.75 -0.13 -0.02 0.13 -0.03 -0.04 0.03
ZH 4e -0.52 -0.66 -0.89 0.05 0.10 0.33 -0.16 0.09 -0.15
ZH 4e -0.66 -0.83 -0.73 -0.25 0.07 -0.24 -0.01 0.06 -0.01
ZH 4e -0.61 -0.67 -0.76 0.04 -0.14 0.0 0.13 -0.11 0.16
ZH 4e -0.84 -0.68 -0.49 0.14 0.03 0.19 0.08 0.02 -0.01
-NaAlH4
ZNa 4a 1.23 1.23 1.12 0.02 0.0 -0.02 0.0 0.0 0.0
ZAl 4b 1.57 1.57 1.87 0.0 0.0 0.0 0.0 0.0 0.0
ZH 16f -0.75 -0.65 -0.75 -0.09 0.13 -0.09 0.09 0.11 0.11
Position xx yy zz xy xz yx yz zx zy
Li3AlH6
ZLi 18f 1.05 1.12 1.02 0.10 -0.08 0.07 0.11 -0.03 -0.01
ZAl 3a 2.11 2.11 2.41 -0.05 0.0 0.05 0.0 0.0 0.0
ZAl 3b 2.38 2.38 1.76 -0.08 0.0 0.08 0.0 0.0 0.0
ZH 18f 1.07 -0.75 -0.88 -0.08 -0.21 -0.12 -0.03 -0.19 -0.08
ZH 18f -0.74 -1.08 -0.88 -0.07 0.08 -0.11 0.12 0.03 0.21
Na3AlH6
ZNa 2b 1.06 1.04 1.07 -0.03 0.01 0.0 0.07 -0.03 0.04
ZNa 4e 1.07 1.1 1.09 -0.06 -0.07 0.04 -0.04 0.07 -0.01
ZAl 2a 2.15 2.20 2.16 0.01 0.03 0.08 -0.06 -0.03 0.02
ZH 4e -0.77 -0.71 -1.20 0.0 -0.17 -0.03 -0.07 -0.17 -0.09
ZH 4e -1.07 -087 -0.74 -0.2 0.07 -0.22 0.03 0.07 0.09
ZH 4e -0.83 -1.11 -0.76 -0.22 0.11 -0.23 0-15 0.10 0.13
SrAlSiH
ZSr 1a 2.64 2.64 2.20 0.0 0.0 0.0 0.0 0.0 0.0
ZAl 1c 1.69 1.69 1.44 0.0 0.0 0.0 0.0 0.0 0.0
ZSi 1b -3.22 -3.22 -2.01 0.0 0.0 0.0 0.0 0.0 0.0
ZH 1c -1.1 -1.1 -1.63 0.0 0.0 0.0 0.0 0.0 0.0
From the experience on analysing chemical bonding in several

compounds/phases we have found that for metal, complex hydrides and
other materials several analysis tools must be consulted in order to make
more firm conclusions regarding the bonding behavior.
Characterizing Chemical Bonding in Multicomponent Hydrides:

A Challenging Task
As we have mentioned in the beginning of this chapter, evaluation of the

chemical bonding in materials becomes difficult when the number of
constituents increases. This section we are going to demonstrate how one
can get the clear picture of chemical bonding in a test case (Li3AlH6) from
various ab initio tools. One can imagine the chemical bonding of Li3AlH6 in
the following manner: Within one Li3AlH6 unit, the three electrons from Li
will fill three of the six half-filled H-s orbitals and the remaining three half-
filled H-s orbitals form covalent interaction with the three electrons from Al,
resulting in the complete filling of VB, which gives the material insulating
behaviour and it has the mixed bonding nature. On the otherhand, one can
also imagine that, within one Li3AlH6 unit, the three electrons from Li will
fill three of the six half-filled H-s orbitals and the remaining three H-s
orbitals are filled by three electrons from the Al atom (i.e., pure ionic
situation). In most of the complex hydrides (like NaAlH4, Li3AlH6, etc.) the
latter type (pure ionic) of bonding was believed. [30, 31] The arguments for
this interpretation are based on DOS analysis and long-range Coulomb
interactions according to calculations using the linearized-augmented plane-
wave method. As more H-s states are present in the valence band than in the
conduction band, it is concluded that the interaction between Al and H must
be ionic.
The calculated DOS, from different codes are perfectly matching each
other. Hence, there is no ambiguity between the different computational
methods, whereas the interpretation/understanding of the results differs. It is
commonly recognized that it is difficult to characterize the nature of the
chemical bonding (in particular for hydrides) from DOS and integrated
charges inside each atomic spheres alone. Each and every theoretical tool
has some additional flexibility/facility to evaluate bonding behavior. We
have therefore now going to see how one can use different approaches to
gather as much information as possible regarding the chemical bonding in
Li3AlH6.
The calculated partial DOS of Al and H in Li3AlH6 are shown in Figure
6, illustrating the following three main features: (1) The VB and CB are
separated by a band gap of 3.8 eV, confirming that this compound is an
insulator. (2) The VB is split into two separate regions by a 1.3 eV energy
gap. The lowest-lying band (at −7.5 to −5.8 eV) mainly originates from
Al-s states with appreciable contributions from H-s states. The second
region, from −4.5 to −2.2 eV, comprises energetically degenerate H-s and
Al-p states, which we focused on as a favorable situation for formation of
covalent bonding. The spherically symmetric nature of s orbitals, together
with the energetic degeneration of H-s and Al-s states, suggests that there
should be a high probability for formation of covalent-type bonding between
these atoms. (3) The very small contribution from H-s in the unoccupied
states above the Fermi level (EF) is explained with the help of COHP
analysis. The COHP, which is the Hamiltonian-population-weighted DOS,

is identical to the crystal orbital overlap population. A negative value of
COHP indicates bonding character and a positive value of COHP points at
an antibonding character. The DOS and COHP are evaluated using different
computer codes and that is the reason for the slight energy shift between the
DOS and COHP curves in Figure 6 and Figure 8. However, the overall
features are the same. The calculated COHP for Al-H (Figure 8) shows that
bonding states are present below −2.8 eV and nonbonding states between
−2.8 eV and EF. The COHP for H-H shows bonding states between −4 eV
and EF. The presence of both bonding and antibonding H-H states within the
VB indicates that the H-s states are almost filled, and this could explain why
there are few H-s states in the CB.
Figure 8. Calculated COHP for H-H (note: multiplied by 6) and Al-H in Li3AlH6.
If the chemical bonding between Al and H was purely ionic one would
expect that Al-p and H-s states should be energetically well separated. [32]
Furthermore, in a purely ionic situation one should not expect the finite
electron-density distribution between Al and H seen in Figure 9. Moreover,
the charge-density distribution should be spherically symmetric around the
Al and H sites if the bonding is purely ionic. Hence, we came to our initial
conclusion that the bonding between Al and H in Li3AlH6 is largely covalent.
If there is strong covalent bonding, one should expect that the H character
should be distributed between the three s-p manifolds. As the H-s character
is very small in the CB (Figure 6), one can conclude that the covalent
contribution is negligible. However, in the first place one cannot judge the
character of chemical bonding in complex materials based on DOS analyses
alone (see, e.g., [12]). Owing to iono-covalent interactions, all H-s orbitals
will be filled and therefore both bonding and antibonding states of the s-p
hybrid are within the VB, as is evident from our COHP analysis (see, e.g.,
Figure 8). If the bonding interaction between all constituents in Li3AlH6 was
purely ionic, one would expect narrow-band features and certainly not the
broad DOS features found for Li3AlH6 (which indicates overlap interaction
between the constituents). In a purely ionic situation, the distinct DOS
manifold around −7 eV with two electrons per formula unit would have to
be contributed by only one of the constituents. However, considerable
amounts of electrons (according to the integrated DOS 36% from Al, 49%
from H, and the remaining 15% from Li) from both Al-s and H-s indicate a
distinct degree of covalent character. If Al was in the 3+ state, negligible
amounts of electrons would be left at the Al site, resulting in a very small
contribution from it to the VB. In fact, the integrated DOS yielded 0.78
electrons at the Al site and shows that the bonding interaction is not purely
ionic (partial conclusion from DOS alone).
In order to gain further understanding of the nature of the bonding, we
have calculated the charge density, charge transfer, energy projected charge
density, and ELF for Li3AlH6. The results are shown in Figure 9, where we
show only the relation between Al and H since there is no ambiguity with
regard to the ionic interaction between Li+ and AlH6−. The charge-density
distribution and charge-transfer plots indicate that a finite number of
electrons are present between Al and H, which means that there is a clear-
cut degree of covalent interaction between Al and H within the AlH6 subunit.
The charge-transfer plot [Figure 9b] clearly indicates that charges are
depleted from Al and Li. The depletion is not spherically symmetric at the
Al site (in a pure ionic case it should be spherically symmetric), which
implies that the bonding can be regarded as ionic with an appreciable
covalent-type interaction (directional bonding) between Al and H
(additional insight from charge-density and charge-transfer plots). The
energy projected charge density [see Figure 9c the electron distribution

within the energy domain region -7 to -5.8 eV] clearly revealed that the
electrons present between Al and H sites are originating from the energy -7
to -5.8 eV regions and are contributed by both Al and H, i.e., confirming the
presence of finite covalent bonding.
Figure 9. Calculated plots of (a) valence-electron charge density, (b) charge transfer,
(c) energy (range -7.5 to -5.8eV) projected charge density, and (d) ELF for Li3AlH6.
The illustrations refer to the (001) plane.
The calculated ELF for Li3AlH6 in Figure 9d shows that the H electrons
are polarized toward the Al site. A similar character of the ELF is found for
the molecules C2H6 and C2H4, where the interaction between C and H is
commonly recognized as covalent. [10] Hence, one finds that there is a
directional (covalent) bonding character between Al and H in Li3AlH6
(subconclusion from ELF). The maximum value of ELF at the H sites along
with the nonspherical distribution of ELF in Figure 9d also confirming the
iono-covalent bond between Al and H in Li3AlH6.
Mulliken charges are reported in Table 2 for a few H-based
representative materials that may provide benchmarks for systems with well-
recognized chemical bonding characters. LiH is a nearly pure ionic
compound and correspondingly the calculated Mulliken charges reflect a

nearly pure ionic picture with Li+ and H−. The overlap population between
Li+ and H− is also close to zero, as expected for an ionic compound.
Similarly, in MgH2, BeH2, and AlH3, the bonding interaction is mainly ionic,
but the degree of ionicity is reduced from MgH2 to BeH2 and further to AlH3,
viz., these compounds exhibit some covalent character as evidenced by the
nonzero overlap population. For the CH4 molecule the overlap population
takes a value of 0.384, consistent with the well-known covalent bonding in
this molecule. The Mulliken effective charges for Li, Al, and H in LiAlH4
and Li3AlH6 indicate that the interaction between the Li and AlH4/AlH6 is
ionic (one electron transferred from Li to AlH4/AlH6). There is a distinct
overlap population between Al and H within the AlH4/AlH6 units, which
reflects a partly covalent character of the Al-H bond, but the magnitude of
the overlap population is smaller than for purely covalent compounds
(subconclusion from Mulliken populations). Moreover, the partial charges
of approximately two electrons transferred from Al to H imply that there is
a significant ionic contribution to the Al-H bond. The calculated integrated
COHP indicates that the covalent Al-H interaction in LiAlH4 is stronger than
that in Li3AlH6. Similarly, the calculated Mulliken effective charges and
overlap population indicates that the covalent Al-H interaction is reduced
when we move from LiAlH4 to Li3AlH6.
In Table 3 we have listed the calculated Born effective charge tensor
elements for the constituents of Li3AlH6. For Li and Al, the diagonal
components of the effective charges in a Cartesian frame are Zxx = Zyy  Zxx,
the off-diagonal components being negligible. However, at the H site,
distinctive effective charges are present along the diagonals as well as along
off-diagonal directions, which, along with their anisotropic distribution,
implies that a finite directional bonding is present in Li3AlH6. If the
interaction had been purely ionic, one would expect that the effective
charges along the diagonal components will be equal and the off-diagonal
components should be negligible. It can be seen from Table 3 that the
dynamical charges for Li are close to the nominal ionic charge (+1 for Li).
For Al, the dynamic charges are +1.74 to +2.24, which are smaller than the
nominal charge of +3. This indicates that part of the Al charges also
participate in directional (viz., covalent) bonding (sub conclusion from Born

effective charges).
HYDROGEN DECOMPOSITION AND CHEMICAL BONDING
Destabilization of Hydrides by Particle Size Reduction
Nanostructured materials are particularly promising for hydrogen

storage applications because of their unique features such as adsorption on
the surface, inter- and intragrain boundaries, and within the bulk [33].
Nanostructured and nanoscale materials strongly influence the
thermodynamics and the kinetics of hydrogen absorption and dissociation
by increasing the diffusion rate as well as by decreasing the required
diffusion length. Additionally, the materials at the nanoscale offer the
possibility of tailoring the material parameters independently from their bulk
counterparts. They also lead to the design of lightweight hydrogen storage
systems with better characteristics. To efficiently use hydrogen as a fuel, a
proper storage material is necessary satisfying two main criteria: (a) high
hydrogen content; (b) reversible binding and releasing of hydrogen at
moderate conditions. The last criterion will require a hydride compound that
is neither too stable nor too unstable. AlH3 (alane) has historically been used
as an energetic component in rocket propellants [34] and more recently has
been considered by many as a good candidate for hydrogen storage
applications. AlH3 is an unique binary hydride presenting at least six
crystalline phases with different physical properties. It is capable of storing
up to 10.1 wt.% of hydrogen [35]. Its gravimetric hydrogen density is two
times higher than liquid hydrogen and much higher than that of most of the
known metal hydrides. It is thus considered as a possible hydrogen storage
material [36]. Freshly prepared alane was shown to decompose (it is not a
stable compound) and to release hydrogen at rates suitable for practical
applications at relatively low temperatures (∼100oC) [37, 38] which is a
beneficial feature for hydrogen storage materials. Unlike complex hydrides,
where one needs to destabilize the system (also decomposition temperature)
as well as enhance sorption kinetics, in alane one should enhance its kinetics
to use it for practical applications. In general, lightweight metal hydrides are
considered as storage media, but the kinetic constraints limit their
application. A promising approach to address this issue is to reduce the
particle size of the metal hydride to the nanometer range, resulting in
enhanced kinetics without the need of a catalyst [39, 40]. For example, in
MgH2, a numerous studies have been focused on improving the problematic
sorption kinetics, including mechanical ball milling [41-43] and chemical
alloying [44, 45]. The smallest particle sizes (20 nm) obtainable by these
methods still primarily display bulk desorption characteristics [46]. Though
AlH3 is one of the promising candidate for hydrogen storage application, it
is not a stable compound and its property changes with time when it is stored
ambient condition and especially releases the hydrogen around 100°C
[37, 38]. In order to increase the stability of this system one must reduce the
particle size below its critical value. In our study we suggested that particle
size of α-AlH3 clusters below 1 nm might have the required
physical/chemical properties for the practical applications. [47]
Stability Enhancement by Particle Size Reduction
From the variation in the interatomic distances compared with bulk

materials it should be anticipated that nano-phase materials have different
physical and chemical property. Once we reduce the particle size beyond
certain range (the so-called critical particle size), most of the atoms will be
exposed at the surface. It is in this region where the properties of the material
begin to differ drastically from that of the bulk. In order to identify the
critical particle size we have calculated the total energy as a function of the
cluster size for AlH3 as shown in Figure 10a. From Figure 10a it is evident
that if the cluster size decreases the total energy becomes more positive. In
particular, there is a steep increase in the total energy when the size of the
cluster is below 1 nm for α-AlH3. Furthermore, the reduction in the total
energy for the nanoparticles suggests that these systems will have different
thermodynamical properties with respect to the bulk materials. The
hydrogen desorption temperature, in particular, is expected to be lower. This

is due to the fact that the surface-to-volume ratio increases upon decreasing
the cluster size. Since the surface atoms have lower coordination (generally
found to occupy the less stable top and bridge sites) than that in bulk
materials, the average number of bonds between constituents is lower for
smaller clusters. This could explain why the decomposition temperatures for
nanoparticles are usually lower than that in bulk materials. If one compares
the variation in total energy with particle size for α-AlH3 and that with the
combination of nanoparticles of Al with H2 molecule (see Figure 10a), even
the particle size below 1 nm the nanoparticle from α-AlH3 is energetically
stable compared to the corresponding decomposed phases (i.e.,
nanoparticles of Al with H2 molecule). Especially, below the critical particle
size, i.e., ca. 1 nm, the total energy gets more positive (i.e., structure becomes
very unstable) for the combination of nanoparticles of Al with H2 molecule.
This is opposite to the conclusion we have drawn for nanoparticles of MgH2
and borohydrides. In the latter case, the nanoparticles decompose and release
hydrogen below the critical particle size. [48, 49] The present result is
thererfore suggesting that, unlike the case of the other hydrides we have
investigated for their nanophase aspects, one can stabilize nanoparticles of
AlH3 even below 1 nm size. [47]
Figure 10. Calculated total energy as a function of particle size (a) for the AlH3 nano-
clusters (in filled circle) with nano particles of Al plus H2 molecule [i.e.,
EAl(nano)+(3/2)EH2 (mol.)] (in open circle) and (b) for NaBH4 and KBH4.
Particle Size Reduction Reduces the Stability
The critical particle size of the nano clusters and nano whiskers of
LiBH4, NaBH4 and KBH4 are identified from the calculated the total energy
as a function of the cluster size and that are shown in Figure 10b (only for
NaBH4 and KBH4). If the cluster size decreases the total energy becomes
more positive (i.e., the formation energy decreases with decreasing cluster
size). In particular, there is a steep increase in the total energy when the size
of the cluster is below 1.35 and 1.8 nm for NaBH4 and KBH4, respectively.
Similarly, the calculated total energy as a function of nano-whisker diameter
shows that when we reduce the diameter below certain critical value (2.6 nm
for NaBH4; the corresponding value for KBH4 is 2.9 nm) the total energy
increase drastically. If one reduces the cluster size and nano-whisker
diameter, the relative energy of the clusters/nano-whiskers becomes less
negative, which indicates the destabilization of small particles. This also
suggests that the thermodynamical properties and in particular the hydrogen
sorption temperature is expected to reduce in nanophases compared with that
in bulk materials. It should be noted that a similar trend has also been
observed in LiBH4 where the critical size of the cluster was 1.75 nm. On the
other hand, in the previous section we have seen that in AlH3 the particle
size reduction enhances the stability. This finding clearly indicates that
particle size plays an important role on decomposition temperature of the
complex hydrides. The reduction of the particle size in some cases enhances
the stability and in some other cases deceases it. In order to identify the
stability upon particle size reduction we have carried out detailed study on
39 complex hydrides and the calculated deformation energies are calculated
using the following equation
E = nE(bulk) –nE(nano) (8)
where, E(bulk) is the total energy of the bulk phase (solid state), E(nano)
total energy of the ultra-small nano-particles, and n represent the number of
formula units of the specific chemical composition. If the deformation
energy is positive then the chosen compound becomes stable upon particle
size reduction. On the other hand, if the deformation energy is negative then
the chosen compound becomes unstable upon particle size reduction. The
calculated deformation energies for the studied 39 compounds are displayed
in Figure 11. From Figure 11 it is evident that most of the complex hydrides
are more stable and only few compounds are unstable upon particle size
reduction. This finding clearly demonstrates that nano particles of the
complex hydrides may not the suitable candidates for the reversible
hydrogen storage application.
Figure 11. Theoretically calculated stability diagram (energy difference between bulk
and the nano-particle; positive and negative values corresponding to the particle size
reduction can stabilize and or destabilize the system, respectively) for ultra-small
nanoparticles made of different high hydrogen content metal/complex hydrides.
Bonding Nature in Nano Phase Materials
For the bulk phases the theoretical charge-density maps clearly

demonstrated the striking ionic character between Na/K and H with
noticeable covalent bond within the BH4 units. [50] The substantial
difference in the electronegativity between Na/K and H suggests the
presence of strong ionic character (i.e., the Na/K valence electrons
transferred to the H sites) and the small difference in the electronegativity
between B and H suggests the presence of strong covalent character. In order
to gain further understanding about the bonding situation in nano objects of
β-Na/KBH4, we turn our attention to the charge density analysis. According
to the charge-density distribution at the Na and K sites in the center of the
nano-phases (Figure 12 (a–f)), it is evident that the highest charge density
resides in the immediate vicinity of the nuclei. The spherical charge
distribution at the Na and K sites (Figure 12 (b, c, e and f)) clearly reflects
the ionic character. On the other hand, due to the covalent interaction a finite
charge density was present in between B and H. In order to get more insight
into the chemical bonding we have visualized the charge density with the
iso-surface value of 0.6 e/volume, which clearly shows the sharing of
electrons in between B and H (see Figure 12 (b and e)). From the Figure 12,
it is evident that the charge density distribution in clusters and whiskers are
almost similar to that in the bulk phase. A closer examination at the edges
of the clusters and whiskers B and H sites shows however a significant
difference in the charge density distribution. In particular, the charges are
more directional dependent at the surface H sites towards the B sites and the
charge distribution is more asymmetric. It can be hence concluded that
considerable more covalent interaction is present between B and H at the
BH4 surface than that in the inner part of the BH4 units.
In order to have a better understanding about the bonding interaction
between the constituents the bond overlap population (BOP) values are
calculated on the basis of Mulliken population analysis. The BOP values can
provide useful information about the bonding property between atoms. A
high BOP value indicates a strong covalent bond, while a low BOP value
indicates an ionic interaction. For the practical use of NaBH4 and KBH4 as
hydrogen storage materials one must weaken their B–H bonds. The
calculated BOP values for the B–H bonds in different modifications of
NaBH4 and KBH4 are listed in Table 4. Table 4 shows that the BOP values
for the B–H bonds are considerably weaker in the nano-phases. It is
interesting to note that the BOP value for B–H bond in the bulk phase is
around 1.13, which is almost equivalent to that of covalent C–C bond in
diamond (the BOP value is 1.08). This might be one of the reasons why these
phases are able to release H only at elevated temperature (above the melting
point). In all the considered phases the calculated Na/K–H BOP values are
almost zero, owing to the high degree of the ionic character.
Figure 12. Charge density distribution in β-KBH4 derived nano-clusters, and nano-
whiskers. Charge distribution in volumetric data view with isosurface value of 0.6
e/volume. [(a) for dot and (d) for whiskers], only the isosurfaces with value of 0.6 e
[(b) for dot and (e) for whiskers], and two dimensional view [(c) for dot and (f) for
whiskers]
It should be noted that when we increase the cluster/nano-whisker size

above the critical size these nano objects have core Na/KBH4 structural units
that makes these systems become stable. [51] Hence, one must reduce the
particle size beyond the critical value to easily remove the H atom from the
Na/KBH4 particles. In order to strengthen this argument we have calculated
the H site energy (HSE; E) in these particles. In nano-clusters/whiskers the
H is situated in four different configurations such as H1 (in between Na/K),

H2 (in between Na/K and B), H3 (H is connected with B) and H4 (center of
the cluster). The H site energy is calculated in the following manner; E =
(EHvac+1/2EHmol.) – Enano, where EHvac and Enano refer to the energy of the nano
object with and without H vacancy and EHmol. is the total energy of a free H2
molecule calculated in a large box (the calculated energy is −6.789 eV). The
calculated HSE for β-NaBH4 and β -KBH4 derived clusters/whiskers are
scattered in a wide energy range (see Figure 13); which is highly depending
upon the environment of the H sites. The calculated H1, H2, H3 and H4 site
energy values are 0.65, 0.82, 1.30, and 2.2 eV, for β-NaBH4 and 0.73, 0.90,
1.32, and 2.31 eV for β-KBH4 derived clusters. Similarly, for nano-whiskers,
the calculated HSE value in β-NaBH4 is 0.78, 1.09, 1.38 and 2.19 and that
in β-KBH4 whiskers 0.68, 1.14, 1.33, and 2.30 eV for H1, H2, H3 and H4
sites, respectively. The corresponding HSE value for the bulk phases are
2.21 and 2.32 eV for β-NaBH4 and β-KBH4, respectively. This finding
clearly indicates that the required energy to remove H from the center of the
nano objects (nano-clusters/nano-whisker) will be similar to that of bulk
phase. Also, a substantial reduction in the energy to remove hydrogen is
expected in nanophases below their critical size. Furthermore, the small
energy requires to remove from the surface layers of complex hydrides
indicates that the nano-phases of hydrogen storage materials will be more
advantageous to use for practical applications since they substantially reduce
the hydrogen sorption temperature and improve the kinetics.
Table 4. Calculated B–H distance (in Å) and bond overlap population

(BOP) values for bulk, low energy surfaces, nano-clusters, and
nanowhiskers of β-NaBH4 and β-KBH4
Modification B-H distance BOP

β-NaBH4 β-KBH4 β-NaBH4 β-KBH4
Bulk 1.207 1.204 1.04 1.13
Surface 1.26-1.23 1.22-1.24 1.01-1.2 0.98-1.03
Cluster 1.20-1.25 1.22-1.25 0.84-0.99 0.82-0.85
Whisker 1.202-1.25 1.23-1.24 0.81-0.98 0.87-0.89
Figure 13. Schematic representation of the calculated hydrogen site energies for the
optimized NaBH4 and KBH4 nano-clusters and whiskers compared with the
corresponding bulk phase. H1, H2, H3 and H4 refers to H in between Na/K; H between
Na/K and B; H connected with B; and H in the center of the cluster, respectively. The
red (left side) and black (right side data) color represent the NaBH4 and KBH4 phases,
respectively.
NANO PARTICLE SIZE IDENTIFICATION BY NMR
Nuclear magnetic resonance (NMR) is often used as an analytical tool

to aid in structure prediction. Even though the general features of the crystal
structure are understood, a detailed analysis of the geometry of nanophases
proves elusive. When the calculated NMR parameters are used, it is possible
to simulate the NMR spectrum for a series of related structures until a match
is found between the computed and experimental results. In this way theory
complements experiment by contributing to the determination of the correct
crystal structure. If one could use both experiment and theory, NMR
parameters can be used not only for structural analysis of bulk materials and
nanophase materials, but also to characterize individual phases in a mixed-
phase system. In this connection we have calculated the isotropic chemical
shielding (σiso), quadrupolar coupling constant (CQ), and quadrupolar
asymmetry parameter (ηQ) for different polymorphs of MgH2 and their
nanophases using the following relations.
(𝜎𝑥𝑥 +𝜎𝑦𝑦 +𝜎𝑧𝑧 )

𝜎𝑖𝑠𝑜 = 3
(9)
𝑒𝑄𝑉𝑧𝑧
𝐶𝑄 = ℎ
(10)
(𝑉𝑥𝑥 −𝑉𝑦𝑦 )
𝜂𝑄 = 𝑉𝑧𝑧
(11)
where σ refers to the chemical shielding tensor in the principal axis frame.
This is an absolute value of the isotropic chemical shielding, not relative to
any standard. Vzz is the largest component of the diagonalized EFG tensor,
Q is the nuclear quadrupole moment, and h is Plank’s constant.
The calculated values of σiso, CQ, and ηQ for the different polymorphs are
displayed in Table 5. The calculated σiso values for Mg are scattered between
549 to 556 ppm. Similarly, for the hydrogen atoms in different polymorphs
these values are scattered between 25.4 to 26.6 ppm and one cannot compare
these values directly to the experimental values. Because the calculated
values are absolute chemical shielding whereas the experimental value are
concerned with shifts relative to a known standard. However, one can
compare the comparative shifts between the different peaks. [49] Similarly
the CQ and ηQ values are also scattered over a wide range and these values
are directly comparable with experimental values. It should be noted that the
NMR parameters are strongly related to the atomic environment (near
neighbors) and quadrupolar coupling constants are directly related to the
bond strength and anisotropy in the charge density distribution around the
probe nuclei. In this connection, we have calculated NMR related
parameters only for the stable α-derived nanowhiskers (for size 1.6 nm) and
β-derived nanoparticles (of 2.4 nm size cluster). As calculation of the NMR
parameters requires higher accuracy for the wave functions, we have
performed very accurate computation to obtain reliable values.
The calculated values of σiso, CQ, and ηQ for the nano-objects are given in
Table 5. As we have shown in the previous section, the bond length values
are very scattered in the nanophases and also the NMR-related parameters
are highly dependent on the local environment of the system. Hence, the
calculated σiso, CQ, and ηQ values for both nano-objects are highly scattered
(see Table 5). For example, the calculated σiso value for Mg in the
nanowhisker varies between 548 to 593 ppm and the corresponding value in
nano particles varies between 542 to 581 ppm. Similarly, the calculated σiso
value for H in the nanowhisker varies between 22 to 26.9 ppm and the
corresponding value in nano particles between 21.4 to 27.2 ppm. To
understand the changes on NMR parameters as a function of particle size,
we have calculated the NMR parameters for the selected nanowhiskers (α-
derived; size: 0.6, 1, and 1.6 nm) and nanoparticles (β-derived; size: 0.6, 1.6,
and 2.4 nm). Due to the close similarity in the chemical shielding of Mg and
H in nanowhiskers and nanoparticles, we have displayed only the
nanowhiskers chemical shielding corresponding to Mg and H in Figure 14.
It is clear that depending on the particle size the calculated chemical
shielding varies considerably and the values are well scattered. If the system
becomes considerably larger, the calculated chemical shielding is well
broadened owing to the bulk-like core structure formation. On the other
hand, in a smaller whisker/particle the chemical shielding becomes very
narrow. The present study suggests that from the shift in the chemical
shielding spectrum one can qualitatively classify the size of nanoparticles.
Figure 14. (Continued).

Figure 14. Calculated chemical shielding of H and Mg in different size of the

nanowhiskers of MgH2.
Table 5. Computed Isotropic Chemical Shielding (σiso; in ppm),

Quadrupolar Coupling Constant (CQ; in MHz), and Quadrupolar
Asymmetry (ηQ) for Different MgH2 Polymorphs
parameter element σiso (ppm) CQ (MHz) ηQ

α-MgH2 Mg 549 2.06 0.31
H 26.6 0.037 0.6
β-MgH2 Mg 556 4.43 0
H 25.4 0.034 0
γ-MgH2 Mg 551 1.17 0.06
H 26.4 0.035 0.73
δ′-MgH2 Mg 552 3.06 0.81
H1 25.7 0.032 0.37
H2 26.1 0.35 0.69
α-MgH2 derived Mg 548–593 1.16–8.67 0.22–0.58
whisker (1.6 nm) H 22–26.9 0.0095–0.038 0.25–0.98
β-MgH2 derived Mg 542–581 1.86–7.21 0.0–0.34
cluster (2.4 nm) H 21.4–27.2 0.0029–0.045 0.0–0.14
CHEMICAL BONDING AND IONIC CONDUCTIVITY IN

HYDRIDE BASED ELECTROLYTES
Complex hydrides consisting of a metal cation and a complex anion are

represented by the formula, M(M’Hn). Here examples of the metal cation,
M, are Li+, Na+, Mg2+, and so forth, and those of the complex anion, M’Hn,
are [BH4]−, [NH2]−, [AlH4]−, [AlH6]3−, and so on. [52] LiBH4 is a typical
complex hydride formed by ionic bonding between Li+ and complex anion,
[BH4]−. Despite the conventional applications of LiBH4 in chemical
processes as a reducing agent, energy storage-related functions such as i)
solid-state hydrogen storage, ii) electrochemical Li storage, and iii) fast Li-
and Na ionic conductions are focused on the class of the materials.
Solid-State Hydrogen Storage
This is the most classical energy-conversion related function. LiBH4

possesses high gravimetric and volumetric hydrogen densities compared to
the other hydrogen storage materials. [52-54] LiBH4 experiences a phase
transition from the low-temperature (LT) phase (orthorhombic structure
with a space group of Pnma, shown in Figure 15a) to the high-temperature
(HT) one (hexagonal structure with a space group of P63mc, shown in
Figure 15b) at approximately 390 K. LiBH4 releases approximately 13.8
mass % hydrogen via the decomposition reaction. [54]
Despite its high hydrogen density, slow dehydriding and rehydriding
kinetics interfere with its practical application. To enhance the slow
dehydriding and rehydriding kinetics, researchers have proposed the
introduction of additives. [55] Confinements in structured carbons [56-59]
as well as the introduction of the intermediate reactions with the reactive
media [60-65] are also reported to be effective measures.
Figure 15. Crystal structures of (a) LT-LiBH4, (b) HT-LiBH4, and (c) HT-
LiBH4+LiI+P2S5.
Electrochemical Li Storage
Oumellal et al. firstly reported the reversible electrochemical conversion

reaction of the redox couple, MgH2/Li. [66] It exhibits a reversible capacity
of 1480 mAh g−1 at an average voltage of 0.5 V versus Li/Li+. [67-69]
Developments have extended not only to other metal hydrides, that is, TiH2
[70] and AlH3, [69] but also to transitionmetal-based complex hydrides such
as Mg2FeH6, Mg2CoH5, and Mg2NiH4. [71, 72] These materials exhibit
higher theoretical gravimetric and volumetric capacities than the
conventional graphite negative electrode. Optimization of the composite
electrode structure need to be addressed to mitigate the over-potential
accompanied by the slow conversion kinetics. [71]
Fast Li- and Na-Ionic Conductions
One of the important challenges in the field of battery research is the

development of fast ionic conductors because this class of materials enables
the assembly from micro-[91] to bulk-type [92-94] all-solid-state
rechargeable cells. With this background, various solid-state electrolytes
have been developed so far, however, the materials showing sufficient ionic
Table 6. Li-, Na-ionic conductivity data for the complex hydrides

based on the pseudo-ternary system,
Li/NaBH4-Li/NaNH2-Li/Na-halides, and the Li/Na-alanates
Complex Hydride Lattice Constant Space Group σ300 K Ea Ref.

[Å] [S cm−1] [eV]
LiBH4 (HT phase) a = 4.27631(5), P63mc(Hexa.) 1 × 10−3 0.53 [73,
c = 6.94844(8) (393 K) 74]
LiBH4 (LT phase) a = 7.17858(4), Pnma(Ortho.) 2 × 10−8 0.69 [73,
b = 4.43686(2), 74]
c = 6.80321(4)
Li(BH4)0.75I0.25 a = 4.354(2), P63mc(Hexa.) 2 × 10−5 0.48 [75,
c = 7.035(5) 76]
Li2(BH4)(NH2) a = 14.3944(3), R-3 (Trigo.) 1 × 10−4 0.66 [77,
c = 9.0522(3) 78]
Li4(BH4)(NH2)3 a = 10.6645(1) I213 (Cubic) 2 × 10−4 0.26 [77,
78]
LiNH2 a = 5.04309(7), I-4 (Tetra.) 4 × 10−10 0.94 [77,
c = 10.2262(3) 79]
Li3(NH2)2I a = 7.09109(5), P63mc(Hexa.) 2 × 10−5 0.58 [78,
c = 11.50958(10) 80]
LiAlH4 a = 4.8254(1), P21/c(Mono.) 2 × 10−9 0.76 [81,
b = 7.8040(1) 82]
c = 7.8968(1)
Li3AlH6 a = 8.07117(10), R-3 (Trig.) 1 × 10−7 0.61 [81]
c = 9.5130(2)
Li2NH a = 5.0742(2) Fm-3m (Cubic) 3 × 10−4 0.58 [83,
84]
Li3N a = 3.6529(2), P6/mmm(Hexa.) 7 × 10−4 0.25 [79,
c = 3.8736(2) 85]
NaBH4 a = 6.1571(2) Fm-3m (Cubic) 2 × 10−10 0.61 [86]
NaNH2 a = 8.9613(3), Fddd (Ortho.) < × 10−10 1.40 [86]
b = 10.454(4)
c = 8.0727(3)
Na2(BH4)(NH2) a = 6.5384(2), Pbcm (Ortho.) 2 × 10−6 0.59 [86,
b = 6.679(2), 87]
c = 9.831(3)
Na4(BH4)3I a = 6.2442(3) Fm-3m(Cubic) 2 × 10−10 0.70 [86]
Na2(NH2)I a = 6.762(3) Ibam (Cubic) 2 × 10−9 0.82 [86]
NaAlH4 a = 5.023(1), I41/a (Tetra.) 2 × 10−10 0.79 [88]
c = 11.347(4)
Na3AlH6 a = 5.21, P21/n (mono.) 5 × 10−7 0.62 [88,
b = 5.42, 89]
c = 7.5,
β = 89.86o
NaI a = 6.4686(2) Fm-3m(Cubic) < × 10−10 0.95 [90]
conductivity as well as good stability in the voltage ranges for battery

operation are limited to a few cases. [95] Therefore, novel solid-state
electrolytes are urgently required for the development of future generation
batteries. Since the discovery of fast Li-ionic conduction in the HT phase of
LiBH4, [74] numerous solid-state electrolytes based on the complex
hydrides and manifesting fast Li- and Na-ionic conduction have been
developed. [96] For example, the pseudo-ternary system consisting of
LiBH4-LiNH2-LiI, [78, 80] the complex hydrides Li2(BH4)(NH2),
Li4(BH4)(NH2)3,59 Li3(NH2)2I (see Table 6).(60) Li-alanates, LiAlH4 and
Li3AlH6. [81]
LiBH4 undergoes a phase transition at approximately 390 K. The HT
phase exhibits fast Li-ionic conductivity of greater than 2 × 10−3 S cm−1. [74]
Large band gaps with approximately 6 eV both in the LT- and HT-phase of
LiBH4 [97, 98] suggests the negligible contribution of the electronic
conductivity in that material. The primary approach for the development of
solid-state electrolytes is to stabilize the HT phase of LiBH4 in the lower
temperature region or even at room temperature. This stabilization could be
achieved through the partial replacement of the complex anion, [BH4]−
(0.205 nm), by iodide-ions, I– (0.211 nm). Accompanied by the increasing
of the I− doping level, the phase transition temperature monotonically
decreased. [99, 100] As a result, the solid-solution, Li(BH4)1−xIx, maintains
the hexagonal phase of LiBH4 (fast Li-ion conduction phase) even at room
temperature when x exceeds 0.25. The solid-solution, Li4(BH4)3I, exhibits
fast Li-ionic conductivity of 2 × 10−5 S cm−1 at 300 K. [75, 99, 100] This
phenomenon might be due to increased neighboring distance of [BH4]− [101]
and induced lattice anharmonicity by the partial substitution of I− for [BH4]−.
[102, 103] Yamauchi and co-workers recently examined the effect of LiBH4
addition to 0.75Li2S-0.25P2S5 glass electrolytes on the Li-ionic conductivity.
[104] The addition resulted in an increase in the electrical conductivity
(1.6 × 10−3 S cm−1) and a decrease in the activation energy (0.21 eV) with
respect to the 0.75Li2S-0.25P2S5 glass electrolyte, whose Li-ionic
conductivity and activation energy are 2.7 × 10−4 S cm−1and 0.26 eV,
respectively. An all-solid-state TiS2/Li cell could be stably operated for at
least 5 charge–discharge cycles with a high TiS2 utilization ratio of
approximately 200 mAh g−1 at 298 K. The study provided a novel design
principle for the development of the electrolytes for use in rechargeable
batteries. All these finding clearly indicates that [BH4]−-[BH4]− interaction
place an important role on Li ion conduction.
The aim of our study is to determine the energy barriers for Li ions
hopping between different sites and how these barriers are influenced by
incorporation of LiI, Li2S and/or P2S5 molecular units and [I-] ions into the
LiBH4 structure. In this work, only the addition of the P2S5 phase to the HT-
LiBH4+LiI system (LPS) at low local concentration is considered, in order
to have better view ability. The calculated activation barriers for the LT and
HT phases of LiBH4 are 0.69 eV and 0.53 eV respectively, in good
agreement with previous theoretical work. [105, 106] Upon introduction of
LiI in the HT-LiBH4 matrix, these activation barriers are lower (0.42 eV to
0.54 eV) and the minimum values are found in proximity of the chemical
Table 7. Calculated BH4-BH4 distance (in Å); Mulliken effective

charges (MEC) and overlap population (BOP) are given in terms of e.
For the simplicity only the Li, B, and H contributions
are given in the table
Modification distance MEC BOP

BH4-BH4
LT-LiBH4 1.78 Li = 1.48 B-H : 1.01 to 1.03
B = -0.9 Li-H : -0.12
H = -0.13 to -0.19 H-H : -0.09
HT-LiBH4 1.636 - 2.27 Li = 1.45 B-H : 0.95 to 1.00
B = -0.81 Li-H : -0.18
H = -0.13 to -0.24 H-H : -0.09
HT-LiBH4 + LiI 1.636 - 2.32 Li = 1.41 to 1.49 B-H : 1.01 to 1.04
B = -0.9 to -1.01 Li-H : -0.21 to -0.43
H = -0.11 to -0.22 H-H : -0.09 to -0.12
HT-LiBH4 + LiI + 1.646 - 2.51 Li = 1.40 to 1.51 B-H : 0.91 to 1.06
Li2S B = -0.9 to -1.01 Li-H : -0.12 to -0.19
(LLS) H = -0.13 to -0.21 H-H : -0.09 to -0.14
HT-LiBH4 + LiI + 1.636 - 2.82 Li = 1.45 to 1.63 B-H : 0.95 to 1.24
P2S5 B = -0.81 to -1.21 Li-H : -0.18 to -0.23
(LPS) H = -0.15 to -0.29 H-H : -0.09 to -0.19
impurity. Similarly, upon introduction of P2S5 units in the HT-Li(BH4)0.75I0.25

matrix (see Figure 16), the activation barriers are considerably lower,
ranging from 0.19 eV to 0.63 eV. [107] The lowest value is found in the
proximity of the substituted P2S5 site, and the maximum one is observed at
the second nearest BH4 units from the P2S5 site. This finding clearly indicates
that the Li mobility in LiBH4 is influenced by the interaction between the
BH4···BH4 molecular units (see Table 7). In both HT and LT phase of LiBH4
the interactions between the BH4···BH4 molecular units are considerably
larger, and they are hindering the transportation of lithium. In the case of the
HT polymorphs the Li are arranged in a layered structure, thus improving Li
transportation. When [I-] or [P2S53-] is substituted into the structure, the
BH4···BH4 interaction is reduced and more space is introduced, boosting the
transportation of Li ions. When the critical LPS ratio is reached, the opposite
effect is observed, as the BH4···BH4 interaction is further reduced and
interaction between P2S5 molecules arises, resulting in poorer Li
transportation properties.
According to the charge-density distribution at the Li, B, and H sites (for
both LT- and HT-LiBH4), it is evident that the highest charge density resides
in the immediate vicinity of the nuclei. As also evidenced from the almost
spherical charge distribution, the bonding between Li is purely ionic and the
one between B and H has a predominantly directional character. The nature
of the charge distribution seen in Figure 17 appears to be typical for iono-
covalent compounds. The electron population between Li and the BH4 units
is almost zero (viz. charges are depleted from this region), which reconfirms
that the interaction between Li and BH4 units is purely ionic. On the other
hand, the charge density plot clearly indicates that the interaction between
the B-H has apredominantly covalent character (electrons are located
between the B and H site; see Figure 17b and e). The calculated ELF plot
(Figure 17c and f) shows a predominant maximum of ca. 1 at the H site and
these electrons have a paired character. The ELF values at the Na and B sites
are very low. The inference from this observation is that charge is transferred
from the inhabitants of these sites to the H sites and there are certainly very
few paired valence electrons left at the Na and B sites. A certain polarized
character is found in the ELF distribution at the H sites in all the complex
hydrides we have investigated earlier. When we introduced LI, Li2S, or P2S5
in the HT-LiBH4 phase the charge distributions are considerably unchanged.
On the other hand, for the LiI added, LLS and LPS systems the calculated
MEC and BOP values are varying considerably. Especially the BOP value
for the B-H interaction is notably different within the BH4 units and this
value is minimum in the pure HT-LiBH4 and maximum in the LPS system.
This finding clearly demonstrated that the BH4 ···BH4 interaction plays a
crucial role on the Li transport. It should be highlighted that for the hydrogen
storage application one might weaken the B-H bonds within the [BH4]− unit.
On the otherhand, for obtaining a good Li-conductors the B-H interaction
should be increase in order to reduce the BH4 ···BH4 interaction.
Figure 16. Li ion migration in different possible path ways in (a) HT-LiBH4 doped with
LiI and P2S5 as obtained from climbing Nudged Elastic Band method and (b) energy
barrier heights as function of the local concentration. The connected lines are given for
visual help of the sequence of images. The symbols represent calculated data points.
Figure 17. Calculated valence electron charge density distribution (a and d; color code:
red-high and blue-low electron density.) charge transfer (b and e; isosurface value +0.2
e) and ELF (c and f; isosurface values of 0.9) for low temperature LiBH4 and high
temperature LiBH4.
CONCLUSION AND OUTLOOK
In general, in the present chapter we have seen that the bonding in the
hydrides like LiAlH4 and Li3AlH6 does not exhibit a simple ionic or covalent
character. In fact, the bonding interaction in these compounds is rather
complicated. The interaction between Li and AlH4/AlH6 is ionic and the one
between Al and H manifests both ionic and covalent characters. We believe
that similar types of bonding situations prevail in all similar complex
hydrides, but the magnitude of the ionic/covalent mixture will exhibit a
considerable individual variation. This study clearly demonstrates that
several analysis tools based on ab initio calculations must be consulted in
order to make more assured conclusions regarding the bonding nature of
multicomponent systems.
Nanomaterials have attracted great interest in recent years because of

the unusual mechanical, electrical, electronic, optical, magnetic and surface
properties. The high surface/volume ratio of these materials has significant
implications with respect to energy storage. Both the high surface area and
the opportunity for nanomaterial consolidation are key attributes of this new
class of materials for hydrogen storage devices. However, in the present
chapter we have seen that in most of the complex hydrides the particle size
reduction/nano-phase enhances the stability and the chemical bonding
nature is changed considerably (strong M-H bond). Hence, the nanophases
materials may not suitable for reversible hydrogen storage application.
Complex hydrides are now considered possible candidates as solid-state
electrolytes in all-solid-state rechargeable batteries. Complex metal
hydrides, and specifically LiBH4-based compounds, have recently been
shown to be a promising class of fast-ionic conductors. Although
intrinsically macrocrystalline LiBH 4 displays high conductivities only at
temperatures above 110°C, more recent strategies such as partial anion
substitution and nanoconfinement have been developed that lead also to
high conductivities at room temperature. Yet, LiBH4 is not unique, and
many other complex metal hydrides containing Li+ as well as Na+ and
probably other cations also exhibit high ionic conductivities. Combined
experimental and DFT calculations suggest that [BH4]- groups are
structurally influenced by the presence of [PS43-], as it does for the [I-]
anions, allowing less hindered effect regarding Li mobility and consequently
facilitating the Li ion conduction in the mixed system at lower temperature
RT-150°C. In general, in the mixed systems the BH4···BH4 interaction is
reduced, generating more space for the Li transportation. It is also crucial to
obtain a better understanding of their chemical and electrochemical
stabilities, most notably pertaining to their interface interactions with the
positive electrode. Nevertheless, it is clear that this rapidly expanding family
of complex metal hydrides with high room-temperature conductivities
provides for a rather intriguing area of fundamental scientific research that
can ultimately lead to exciting practical applications for these materials in
next-generation, all-solid-state, lithium-ion, and other rechargeable
batteries.
REFERENCES
[1] R. Gordy, “Microwave spectroscopy. Introductory paper: quadrupole

couplings, dipole moments and the chemical bond,” Discussions
Faraday Soc., vol. 19, pp. 14-29, 1955.
[2] L. Pauling, The Nature of the Chemical Bond., Correll Univ. Press,
1960.
[3] R.S. Mulliken, “Electronic Population Analysis on LCAO MO
Molecular Wave Functions.,” J. Chem. Phys., vol. 23, pp. 1833-1840,
1955.
[4] R.F.W. Bader, Atom in Molecules: A Quantum Theory., Oxford
Univ. Press,, 1990.
[5] M. Born and K. Huang, Dynamical Theory of Crystal Lattices.,
OxfordUniversity Press, 1968.
[6] B. Silvi and A. Savin, “Classification of chemical bonds based on
topological analysis of electron localization functions,” Nature, vol.
371, pp. 683-686, 1994.
[7] R. Dronskowski and P.E. Blöchl, “Crystal orbital Hamilton
populations (COHP): energy-resolved visualization of chemical
bonding in solids based on density-functional calculations,” J. Phys.
Chem.,, vol. 97, pp. 8617-8624, 1993.
[8] J. Bergsma, J.A. Goedkoop and J. H. N. v. Vucht., “Neutron
diffraction investigation of solid solutions AlTh2Dh,” Acta
Crystallogr., vol. 14, pp. 223-228, 1961.
[9] M. H. Sørby, H. Fjellvåg, B. C. Hauback, A. J. Maeland and and V.
A. Yartys, “Crystal structure of Th2Al deuterides,” J. Alloys Compd.,
vol. 309, pp. 154-164, 2000.
[10] A. Savin, A. D. Becke, J. FLad, R. Nesper, H. Preuss and H.G. von
Schnering, “A New Look at Electron Localization,” Angew. Chem.
Int. Ed., vol. 30, pp. 409-412, 1991.
[11] A.D. Becke and K. E. Edgecombe, “A simple measure of electron
localization in atomic and molecular systems,” J. Chem. Phys., vol.
92, pp. 5397-5403, 1990.
[12] P. Ravindran, P. Vajeeston, R. Vidya, A. Kjekshus and H. Fjellvåg,

“Violation of the minimum H-H separation “rule” for metal
hydrides,” Phys. Rev. Lett., vol. 89, no. 10, 2002.
[13] P. Vajeeston, P. Ravindran, R. Vidya, A. Kjekshus, H. Fjellvåg and
V.A. Yartys, “Short hydrogen-hydrogen separation in RNiInH1.333
(R=La, Ce, Nd),” Phys. Rev. B, vol. 67, no. 1, 2003.
[14] P. Vajeeston, P. Ravindran, H. Fjellvåg and A. Kjekshus, “Search for
metal hydrides with short hydrogen-hydrogen separation: Ab initio
calculations,” Phys. Rev. B, vol. 70, no. 1, 2004.
[15] P. Vajeeston, P. Ravindran, R. Vidya, H. Fjellvag and A. Kjekshus,
“Design of potential hydrogen-storage materials using first-principle
density-functional calculations,” Cryst. Growth Des., vol. 4, no. 3,
pp. 471-477, 2004.
[16] P. Vajeeston, P. Ravindran, R. Vidya, A. Kjekshus and H. Fjellvåg,
“Site preference of hydrogen in metal, alloy, and intermetallic
frameworks,” EPL, vol. 72, no. 4, pp. 569-575, 2005.
[17] C. Duan, Y. L. Liu, H. B. Zhou, Y. Zhang, S. Jin, G. H. Lu and G. N.
Luo, “First-principles study on dissolution and diffusion properties
of hydrogen in molybdenum,” J. Nucl. Mater., vol. 404, no. 2, pp.
109-115.
[18] 18. V. Hlukhyy and T.F. Fassler, “Synthesis, Structure, and Bonding
of Ca2Ni3Ge2 and Comparison with CaNiGe, SrNi2Ge, and
Ca3Ni3Si2,” Z. Anorg. Allg. Chem vol. 636, no. 1, pp. 100-107.
[19] J.N. Chotard, D. Sheptyakov and K. Yvon, “Hydrogen induced site
depopulation in the LaMgNi4-hydrogen system,” Z. Kristallogr.
Cryst., vol. 223, no. 10, pp. 690-696, 2008.
[20] S.F. Matar and A.F. Al Alam, “First principles study of the electronic
and magnetic structures of U2Ni2SnH2,” New J. Phys., vol. 10, 2008.
[21] R. Vincent, A. Lodder, I. Nagy and P.M. Echenique, “Description of
a migrating proton embedded in an electron gas,” J. Phys. Condens.
Matter, vol. 20, no. 28, 2008.
[22] D. G. Westlake, “Rationalizing the hydrogen sites in Pd3P1−xHy,” J.
Less. Common Met., vol. 103, no. 2, pp. 203-209, 1984.
[23] P. Vajeeston, P. Ravindran, R. Vidya, A. Kjekshus and H. Fjellvåg,

“Site preference of hydrogen in metal, alloy, and intermetallic
frameworks,” Europhys. Lett., vol. 72, pp. 569-575, 2005.
[24] G.B. Abdullaev, E.S. Guseinova and B.G. Tagiev, “Study of Electron
Lifetime in p-Si,” Phys, Status Solidi, vol. 20, pp. 423-429, 1967.
[25] S.A. Abagyan, V.K. Subashiev and K. SOy., Sov. Phys. -Solid State,
vol. 6, pp. 2529, 1965.
[26] T. Björling, D. Noréus, K. Jansson, M. Andersson, E. Leonova, M.
Edén, U. Hålenius and U. Häussermann, “SrAlSiH: A Polyanionic
Semiconductor Hydride,” Angew. Chem. Int. Ed., vol. 44, pp. 7269-
7273, 2005.
[27] A. Aguayo and D. J. Singh, Phys. Rev. B, vol. 69, pp. 155103, 2004.
[28] M.J. van Setten, V.A. Popa, a. G.A. de Wijs and G. Brocks,
“Electronic structure and optical properties of lightweight metal
hydrides,” Phys. Rev. B, vol. 75, pp. 035204, 2007.
[29] J. Isidorsson, I.A.M.E. Giebels, H. Arwin and R. Griessen, “Optical
properties of MgH2 measured in situ by ellipsometry and
spectrophotometry,” Phys. Rev. B, vol. 68, pp. 115112, 2003.
[30] A. Aguayo and D. J. Singh, “Electronic structure of the complex
hydride NaAlH4,” Phys. Rev. B, vol. 69, pp. 155103, 2004.
[31] D.J. Singh, “Comment on “Structural stability and electronic
structure for Li3AlH6”,” Phys. Rev. B, vol. 71, pp. 216101.
[32] J. C. D. Gelatt, A. R. Williams, and V. L. Moruzzi,, “Theory of
bonding of transition metals to nontransition metals,” Phys. Rev. B,
vol. 27, pp. 2005–2013, 1983.
[33] M. Jurczyk, “Nanocrystalline materials for hydrogen storage,” J.
Optoelectron Adv. M., vol. 8, no. 2, pp. 418-424, 2006.
[34] Chan ML and Johnson CE Evaluation of AlH3 for propellant
application., Bergamo, Paper 33, 2003.
[35] F. M. Brower, N. E. Matzek, P. F. Reigler, H. W. Rinn, C. B. Roberts,
D. L. Schmidt, J. A. Snover and K. Terada, “Preparation and
properties of aluminum hydride,” J. Am. Chem. Soc., vol. 98, no. 9,
pp. 2450-2453, 1976.
[36] L. Schlapbach and A. Züttel, “Hydrogen-storage materials for mobile

applications,” Nature, vol. 414, pp. 353, 2001.
[37] J. Graetz and J. J. Reilly, “Decomposition Kinetics of the AlH3
Polymorphs,” J. Phys. Chem. B, vol. 109, no. 47, pp. 22181-22185,
2005.
[38] G. Alefeld, Hydrogen in Metals II, Springer Berlin, 1978.
[39] C. P. Baldé, B. P. C. Hereijgers, J. H. Bitter and K. P. de Jong,
“Facilitated Hydrogen Storage in NaAlH4 Supported on Carbon
Nanofibers,” Angew. Chem. Int. Ed., vol. 45, no. 21, pp. 3501-3503,
2006.
[40] A. Gutowska, L. Li, Y. Shin, C. M. Wang, X. S. Li, J. C. Linehan, R.
S. Smith, B. D. Kay, B. Schmid, W. Shaw, M. Gutowski and T.
Autrey, “Nanoscaffold Mediates Hydrogen Release and the
Reactivity of Ammonia Borane,” Angew. Chem. Int. Ed., vol. 44, no.
23, pp. 3578-3582, 2005.
[41] A. Zaluska, L. Zaluski and J. O. Ström–Olsen, “Nanocrystalline
magnesium for hydrogen storage,” J. Alloys Compd., vol. 288, no. 1,
pp. 217-225, 1999.
[42] A. Zaluska, L. Zaluski and J. O. Ström-Olsen, “Structure, catalysis
and atomic reactions on the nano-scale: a systematic approach to
metal hydrides for hydrogen storage,” Appl. Phys. A, vol. 72, no. 2,
pp. 157-165, 2001.
[43] S. Orimo, H. Fujii and K. Ikeda, “Notable hydriding properties of a
nanostructured composite material of the Mg2Ni-H system
synthesized by reactive mechanical grinding,” Acta Materialia, vol.
45, no. 1, pp. 331-341, 1997.
[44] J. Huot, G. Liang and R. Schulz, “Mechanically alloyed metal
hydride systems,” App. Phys. A, vol. 72, no. 2, pp. 187-195, 2001.
[45] J. Huot, J. F. Pelletier, L. B. Lurio, M. Sutton and R. Schulz,
“Investigation of dehydrogenation mechanism of MgH2–Nb
nanocomposites,” J. Alloys Compd., vol. 348, no. 1, pp. 319-324,
2003.
[46] G. Liang, J. Huot, S. Boily, A. Van Neste and R. Schulz, “Catalytic

effect of transition metals on hydrogen sorption in nanocrystalline
ball milled MgH2–Tm (Tm=Ti, V, Mn, Fe and Ni) systems,” J. Alloys
Compd., vol. 292, no. 1, pp. 247-252, 1999.
[47] P. Vajeeston, P. Ravindran and H. Fjellvåg, “Stability enhancement
by particle size reduction in AlH3,” J. Alloys Compd., vol. 509, no.
Supplement 2, pp. S662-S666, 2011.
[48] P. Vajeeston, S. Sartori, P. Ravindran, K. D. Knudsen, B. Hauback
and H. Fjellvåg, “MgH2 in Carbon Scaffolds: A Combined
Experimental and Theoretical Investigation,” J. Phys. Chem. C, vol.
116, no. 40, pp. 21139-21147, 2012.
[49] P. Vajeeston, P. Ravindran, M. Fichtner and H. Fjellvåg, “Influence
of Crystal Structure of Bulk Phase on the Stability of Nanoscale
Phases: Investigation on MgH2 Derived Nanostructures,” J. Phys.
Chem. C, vol. 116, no. 35, pp. 18965-18972, 2012.
[50] P. Vajeeston, P. Ravindran, A. Kjekshus and H. Fjellvåg, “Structural
stability of alkali boron tetrahydrides ABH4 (A = Li, Na, K, Rb, Cs)
from first principle calculation,” J. Alloys Compd., vol. 387, no. 1,
pp. 97-104, 2005.
[51] P. Vajeeston, P. Ravindran and H. Fjellvåg, “Nano-Phases of NaBH4
and KBH4,” J. Nanosci. Nanotechnol., vol. 11, no. 3, pp. 1929-1934,
2011.
[52] S.-i. Orimo, Y. Nakamori, J. R. Eliseo, A. Züttel and C. M. Jensen,
“Complex Hydrides for Hydrogen Storage,” Chem. Rev., vol. 107,
no. 10, pp. 4111-4132, 2007.
[53] A. Züttel, P. Wenger, S. Rentsch, P. Sudan, P. Mauron and C.
Emmenegger, “LiBH4 a new hydrogen storage material,” J. Power
Sources, vol. 118, no. 1, pp. 1-7, 2003.
[54] S. Orimo, Y. Nakamori, G. Kitahara, K. Miwa, N. Ohba, S. Towata
and A. Züttel, “Dehydriding and rehydriding reactions of LiBH4,” J.
Alloys Compd., vol. 404-406, no. Supplement C, pp. 427-430, 2005.
[55] M. Au, A. Jurgensen and K. Zeigler, “Modified Lithium
Borohydrides for Reversible Hydrogen Storage (2),” J. Phys. Chem.
B, vol. 110, no. 51, pp. 26482-26487, 2006.
[56] A. F. Gross, J. J. Vajo, S. L. Van Atta and G. L. Olson, “Enhanced

Hydrogen Storage Kinetics of LiBH4 in Nanoporous Carbon
Scaffolds,” J. Phys. Chem. C, vol. 112, no. 14, pp. 5651-5657, 2008.
[57] T. K. Nielsen, U. Bösenberg, R. Gosalawit, M. Dornheim, Y.
Cerenius, F. Besenbacher and T. R. Jensen, “A Reversible
Nanoconfined Chemical Reaction,” ACS Nano, vol. 4, no. 7, pp.
3903-3908, 2010.
[58] X. Liu, D. Peaslee, C. Z. Jost and E. H. Majzoub, “Controlling the
Decomposition Pathway of LiBH4 via Confinement in Highly
Ordered Nanoporous Carbon,” J. Phys. Chem. C, vol. 114, no. 33,
pp. 14036-14041, 2010.
[59] P. Ngene, P. Adelhelm, A. M. Beale, K. P. de Jong and P. E. de
Jongh, “LiBH4/SBA-15 Nanocomposites Prepared by Melt
Infiltration under Hydrogen Pressure: Synthesis and Hydrogen
Sorption Properties,” J. Phys. Chem. C, vol. 114, no. 13, pp. 6163-
6168, 2010.
[60] J. J. Vajo, S. L. Skeith and F. Mertens, “Reversible Storage of
Hydrogen in Destabilized LiBH4,” J. Phys. Chem. B, vol. 109, no. 9,
pp. 3719-3722, 2005.
[61] G. Barkhordarian, T. Klassen, M. Dornheim and R. Bormann,
“Unexpected kinetic effect of MgB2 in reactive hydride composites
containing complex borohydrides,” J. Alloys Compd., vol. 440, no.
1, pp. L18-L21, 2007.
[62] F. E. Pinkerton, M. S. Meyer, G. P. Meisner, M. P. Balogh and J. J.
Vajo, “Phase Boundaries and Reversibility of LiBH4/MgH2
Hydrogen Storage Material,” J. Phys. Chem. C, vol. 111, no. 35, pp.
12881-12885, 2007.
[63] T. Nakagawa, T. Ichikawa, N. Hanada, Y. Kojima and H. Fujii,
“Thermal analysis on the Li–Mg–B–H systems,” J. Alloys Compd.,
vol. 446-447, no. Supplement C, pp. 306-309, 2007.
[64] S.-A. Jin, Y.-S. Lee, J.-H. Shim and Y. W. Cho, “Reversible
Hydrogen Storage in LiBH4−MH2 (M = Ce, Ca) Composites,” J.
Phys. Chem. C, vol. 112, no. 25, pp. 9520-9524, 2008.
[65] O. Friedrichs, J. W. Kim, A. Remhof, F. Buchter, A. Borgschulte, D.

Wallacher, Y. W. Cho, M. Fichtner, K. H. Oh and A. Zuttel, “The
effect of Al on the hydrogen sorption mechanism of LiBH4,” Phys.
Chem. Chem. Phys., vol. 11, no. 10, pp. 1515-1520, 2009.
[66] Y. Oumellal, A. Rougier, G. A. Nazri, J. M. Tarascon and L. Aymard,
“Metal hydrides for lithium-ion batteries,” Nat. Mat., vol. 7, pp. 916,
2008.
[67] W. Zaïdi, Y. Oumellal, J. P. Bonnet, J. Zhang, F. Cuevas, M.
Latroche, J. L. Bobet and L. Aymard, “Carboxymethylcellulose and
carboxymethylcellulose-formate as binders in MgH2–carbon
composites negative electrode for lithium-ion batteries,” J. Power
Sources, vol. 196, no. 5, pp. 2854-2857, 2011.
[68] Y. Oumellal, A. Rougier, J. M. Tarascon and L. Aymard, “2LiH+M
(M=Mg, Ti): New concept of negative electrode for rechargeable
lithium-ion batteries,” J. Power Sources, vol. 192, no. 2, pp. 698-702,
2009.
[69] N. Hanada, A. Kamura, H. Suzuki, K. Takai, T. Ichikawa and Y.
Kojima, “Electrochemical charge and discharge properties for the
formation of magnesium and aluminum hydrides,” J. Alloys Compd.,
vol. 509, no. Supplement 2, pp. S584-S587, 2011.
[70] Y. Oumellal, W. Zaïdi, J. P. Bonnet, F. Cuevas, M. Latroche, J.
Zhang, J. L. Bobet, A. Rougier and L. Aymard, “Reactivity of TiH2
hydride with lithium ion: Evidence for a new conversion
mechanism,” Int. J. Hydrog. Energy, vol. 37, no. 9, pp. 7831-7835,
2012.
[71] W. Zaïdi, J. P. Bonnet, J. Zhang, F. Cuevas, M. Latroche, S.
Couillaud, J. L. Bobet, M. T. Sougrati, J. C. Jumas and L. Aymard,
“Reactivity of complex hydrides Mg2FeH6, Mg2CoH5 and Mg2NiH4
with lithium ion: Far from equilibrium electrochemically driven
conversion reactions,” Int. J. Hydrog. Energy, vol. 38, no. 11, pp.
4798-4808, 2013.
[72] J. Zhang, W. Zaidi, V. Paul-Boncour, K. Provost, A. Michalowicz,
F. Cuevas, M. Latroche, S. Belin, J.-P. Bonnet and L. Aymard, “XAS
investigations on nanocrystalline Mg2FeH6 used as a negative
electrode of Li-ion batteries,” J. Mater. Chem. A, vol. 1, no. 15, pp.

4706-4717, 2013.
[73] J. P. Soulié, G. Renaudin, R. Černý and K. Yvon, “Lithium boro-
hydride LiBH4: I. Crystal structure,” J. Alloys Compd., vol. 346, no.
1, pp. 200-205, 2002.
[74] M. Matsuo, Y. Nakamori, S.-i. Orimo, H. Maekawa and H.
Takamura, “Lithium superionic conduction in lithium borohydride
accompanied by structural transition,” Appl. Phys. Lett., vol. 91, no.
22, pp. 224103, 2007.
[75] H. Maekawa, M. Matsuo, H. Takamura, M. Ando, Y. Noda, T.
Karahashi and S.-i. Orimo, “Halide-Stabilized LiBH4, a Room-
Temperature Lithium Fast-Ion Conductor,” J. Am. Chem. Soc., vol.
131, no. 3, pp. 894-895, 2009.
[76] H. Oguchi, M. Matsuo, J. S. Hummelshøj, T. Vegge, J. K. Nørskov,
T. Sato, Y. Miura, H. Takamura, H. Maekawa and S. Orimo,
“Experimental and computational studies on structural transitions in
the LiBH4–LiI pseudobinary system,” Applied Physics Letters, vol.
94, no. 14, pp. 141912, 2009.
[77] H. Wu, W. Zhou, T. J. Udovic, J. J. Rush and T. Yildirim, “Structures
and Crystal Chemistry of Li2BNH6 and Li4BN3H10,” Chem. Mater.,
vol. 20, no. 4, pp. 1245-1247, 2008.
[78] M. Matsuo, A. Remhof, P. Martelli, R. Caputo, M. Ernst, Y. Miura,
T. Sato, H. Oguchi, H. Maekawa, H. Takamura, A. Borgschulte, A.
Züttel and S.-i. Orimo, “Complex Hydrides with (BH4)− and (NH2)−
Anions as New Lithium Fast-Ion Conductors,” J. Am. Chem. Soc.,
vol. 131, no. 45, pp. 16389-16391, 2009.
[79] W. I. F. David, M. O. Jones, D. H. Gregory, C. M. Jewell, S. R.
Johnson, A. Walton and P. P. Edwards, “A Mechanism for Non-
stoichiometry in the Lithium Amide/Lithium Imide Hydrogen
Storage Reaction,” J. Am. Chem. Soc., vol. 129, no. 6, pp. 1594-1601,
2007.
[80] M. Matsuo, T. Sato, Y. Miura, H. Oguchi, Y. Zhou, H. Maekawa, H.
Takamura and S.-i. Orimo, “Synthesis and Lithium Fast-Ion
Conductivity of a New Complex Hydride Li3(NH2)2I with Double-
Layered Structure,” Chem. Mater., vol. 22, no. 9, pp. 2702-2704,

2010.
[81] H. Oguchi, M. Matsuo, T. Sato, H. Takamura, H. Maekawa, H.
Kuwano and S. Orimo, “Lithium-ion conduction in complex
hydrides LiAlH4 and Li3AlH6,” J. Appl. Phys., vol. 107, no. 9, pp.
096104, 2010.
[82] T. Sato, K. Ikeda, H. W. Li, H. Yukawa, M. Morinaga and S. Orimo,
“Direct Dry Syntheses and Thermal Analyses of a Series of
Aluminum Complex Hydrides,” Mater. Trans., vol. 50, no. 1, pp.
182-186, 2009.
[83] B. A. Boukamp and R. A. Huggins, “Ionic conductivity in lithium
imide,” Phys. Lett. A, vol. 72, no. 6, pp. 464-466, 1979.
[84] T. Noritake, H. Nozaki, M. Aoki, S. Towata, G. Kitahara, Y.
Nakamori and S. Orimo, “Crystal structure and charge density
analysis of Li2NH by synchrotron X-ray diffraction,” J. Alloys
Compd., vol. 393, no. 1, pp. 264-268, 2005.
[85] A. Rabenau, “Lithium nitride and related materials case study of the
use of modern solid state research techniques,” Solid State Ionics,
vol. 6, no. 4, pp. 277-293, 1982.
[86] M. Matsuo, S. Kuromoto, T. Sato, H. Oguchi, H. Takamura and S.-i.
Orimo, “Sodium ionic conduction in complex hydrides with [BH4]−
and [NH2]− anions,” Appl. Phys. Lett., vol. 100, no. 20, pp. 203904,
2012.
[87] M. Somer, S. Acar, C. Koz, I. Kokal, P. Höhn, R. Cardoso-Gil, U.
Aydemir and L. Akselrud, “α- and β-Na2 [BH4][NH2]: Two
modifications of a complex hydride in the system NaNH2–NaBH4;
syntheses, crystal structures, thermal analyses, mass and vibrational
spectra,” J. Alloys Compd., vol. 491, no. 1, pp. 98-105, 2010.
[88] H. Oguchi, M. Matsuo, S. Kuromoto, H. Kuwano and S. Orimo,
“Sodium-ion conduction in complex hydrides NaAlH4 and
Na3AlH6,” J. Appl. Phys., vol. 111, no. 3, pp. 036102, 2012.
[89] V. Ozolins, E. H. Majzoub and T. J. Udovic, “Electronic structure
and Rietveld refinement parameters of Ti-doped sodium alanates,” J.
Alloys Compd., vol. 375, no. 1, pp. 1-10, 2004.
[90] P. Manoravi and K. Shahi, “Fast ion transport and phase diagram of
the KBr-NaI system,” J. Phys. Chem. Solids, vol. 52, no. 3, pp. 527-
535, 1991.
[91] J. F. M. Oudenhoven, L. Baggetto and P. H. L. Notten, “All-Solid-
State Lithium-Ion Microbatteries: A Review of Various Three-
Dimensional Concepts,” Adv. Energy Mater., vol. 1, no. 1, pp. 10-33,
2011.
[92] M. Armand and J. M. Tarascon, “Building better batteries,” Nature,
vol. 451, pp. 652, 2008.
[93] K. Takada, “Progress and prospective of solid-state lithium
batteries,” Acta Materialia, vol. 61, no. 3, pp. 759-770, 2013.
[94] Y. Kato, K. Kawamoto, R. Kanno and M. Hirayama, “Discharge
Performance of All-Solid-State Battery Using a Lithium Superionic
Conductor Li10GeP2S12,” Electrochemistry, vol. 80, no. 10, pp. 749-
751, 2012.
[95] J. W. Fergus, “Ceramic and polymeric solid electrolytes for lithium-
ion batteries,” J. Power Sources, vol. 195, no. 15, pp. 4554-4569,
2010.
[96] M. Matsuo and S.-i. Orimo, “Lithium Fast-Ionic Conduction in
Complex Hydrides: Review and Prospects,” Adv. Energy Mater., vol.
1, no. 2, pp. 161-172, 2011.
[97] K. Miwa, N. Ohba, S.-i. Towata, Y. Nakamori and S.-i. Orimo,
“First-principles study on lithium borohydride LiBH4,” Phys. Rev. B,
vol. 69, no. 24, pp. 245120, 2004.
[98] E. Orgaz, A. Membrillo, R. Castañeda and A. Aburto, “Electronic
structure of ternary hydrides based on light elements,” J. Alloys
Compd., vol. 404-406, no. Supplement C, pp. 176-180, 2005.
[99] H. Oguchi, M. Matsuo, J. S. Hummelshøj, T. Vegge, J. K. Nørskov,
T. Sato, Y. Miura, H. Takamura, H. Maekawa and S. Orimo,
“Experimental and computational studies on structural transitions in
the LiBH4–LiI pseudobinary system,” Appl. Phys. Lett., vol. 94, no.
14, pp. 141912, 2009.
[100] D. Sveinbjörnsson, J. S. G. Myrdal, D. Blanchard, J. J. Bentzen, T.

Hirata, M. B. Mogensen, P. Norby, S.-I. Orimo and T. Vegge, “Effect
of Heat Treatment on the Lithium Ion Conduction of the LiBH4–LiI
Solid Solution,” J. Phys. Chem. C, vol. 117, no. 7, pp. 3249-3257,
2013.
[101] C. C. Stephenson, D. W. Rice and W. H. Stockmayer, “Order‐
Disorder Transitions in the Alkali Borohydrides,” J. Chem. Phys.,
vol. 23, no. 10, pp. 1960-1960, 1955.
[102] Y. Filinchuk, D. Chernyshov and R. Cerny, “Lightest Borohydride
Probed by Synchrotron X-ray Diffraction: Experiment Calls for a
New Theoretical Revision,” J. Phys. Chem. C, vol. 112, no. 28, pp.
10579-10584, 2008.
[103] F. Buchter, Z. Łodziana, P. Mauron, A. Remhof, O. Friedrichs, A.
Borgschulte, A. Züttel, D. Sheptyakov, T. Strässle and A. J. Ramirez-
Cuesta, “Dynamical properties and temperature induced molecular
disordering of LiBH4 and LiBD4,” Phys. Rev. B, vol. 78, no. 9, pp.
094302, 2008.
[104] A. Yamauchi, A. Sakuda, A. Hayashi and M. Tatsumisago,
“Preparation and ionic conductivities of (100 − x)
(0.75Li2S·0.25P2S5)·xLiBH4 glass electrolytes,” J. Power Sources,
vol. 244, pp. 707-710, 2013.
[105] M. Matsuo, Y. Nakamori, S.-i. Orimo, H. Maekawa and H.
Takamura, “Lithium superionic conduction in lithium borohydride
accompanied by structural transition,” Applied Physics Letters, vol.
91, no. 22, pp. 224103, 2007.
[106] T. Ikeshoji, E. Tsuchida, T. Morishita, K. Ikeda, M. Matsuo, Y.
Kawazoe and S.-i. Orimo, “Fast-ionic conductivity of Li in LiBH4,”
Physical Review B, vol. 83, no. 14, pp. 144301, 2011.
[107] A. E. Kharbachi, Y. Hu, K. Yoshida, P. Vajeeston, M. H. Sørby, S.
Kim, S. Orimoc, H. Fjellvåg and B. C. Hauback, “Lithium ionic
conduction in composites of Li(BH4)0.75I0.25 and amorphous
0.75Li2S·0.25P2S5 for battery applications,” manuscript under review,
2018.
Chapter 3
THE APPLICATION OF METAL

HYDRIDE BASED ON ZR-V ALLOY IN
HYDROGEN PLASMA
Ihor Sereda*, Alexandr Tseluyko and Nikolay Azarenkov

Department of Physics and Technology,
V. N. Karazin Kharkiv National University, Kharkiv, Ukraine
ABSTRACT
The features of gas discharge and plasma sources based on Penning

trap with metal hydride cathodes are presented. In such devices, metal
hydrides fulfill the functions of both a cathode and the solid-state generator
of working gas. Their advantages are high purity of gas injected (99.99 –
99.999%), along with the safety and compactness in storage. Hydrogen is
injected (desorbed) locally under the influence of ion bombardment of
metal hydride surface, which fact provides return coupling between the
intensity of gas desorption and the parameters of gas discharge. The rate of
sputtering for those materials by plasma ions significantly reduces as well
as heat loads. Above effect is achieved due to the creation of protective gas
target as a result of both the thermal decomposition of metal hydride and
*
Corresponding author: Email: igorsereda@karazin.ua.
150 Ihor Sereda, Alexandr Tseluyko and Nikolay Azarenkov
ion stimulated desorption. The feature of metal hydride cathode under the
conditions of gas discharge is a decrease in the ionization potential of
desorbed hydrogen by 0.3-0.5 eV due to the molecules desorption in the
vibrationally/rotationally excited state. This permits a substantially
increase in ionization efficiency and the formation of negative ions by the
mechanism of dissociative attachment in plasma volume. However,
hydrogen desorbed from metal hydride significantly changes the properties
of the discharge. This is expressed, for example, in the fact that the plasma
source based on Penning trap with metal hydride cathode appears to
generate current-compensated ion beams with the ability to control the
energy of the extracted ions. There is also the opportunity of longitudinal
extraction of negative hydrogen ions against the traditional method of
extraction across the magnetic field.
Keywords: metal hydride, hydrogen plasma, PIG, negative ions
1. INTRODUCTION
Metal hydride systems are traditionally used for the energy-

technological processing of hydrogen: its storage, purification from gas
impurities or selective extraction from hydrogen-containing gases with
following delivery to the consumer. Substantial development has been
achieved in the creation of hydrogen supply systems for plasma devices. In
this case, metal hydrides are not used only for hydrogen storage, but, as a
rule, play the role of electrodes for gas-discharge devices [1]. Hydride-
forming getter materials based on Zr-V alloys are a promising base for the
manufacture of such electrodes [2]. These compositions are characterized by
high pumping rates and hydrogen storage capacity, they have low activation
temperatures and easily desorb hydrogen when heated to moderately
elevated temperatures. In the regime of hydrogen desorption under the
pressure of 10-4 – 1 Torr, the operating temperature range of metal hydrides
based on zirconium and vanadium alloys is 400 – 900 K. This circumstance
determines the high efficiency of hydrogen pressure adjusting according to
the scheme of “adjustable parameter – thermal effect”. However, in order to
control the regime of metal hydride element operation, an active thermal
action on the metal hydride is necessary. But it is associated with the
The Application of Metal Hydride Based on Zr-V Alloy … 151
significant complication of supply system control and heat removal. On the

other hand, the possibility of changing the dynamics of hydrogen sorption-
desorption processes in hydrogen plasma points to one of the real methods
of constructing gas supply systems that are not require a complex external
control. The essence of this approach is metal hydrides application as a
material for the electrodes of plasma devices. In this case the rate of
hydrogen desorption is controlled by heat release on the metal hydride
element, caused by high-energy plasma particles bombardment.
On the other hand, desorption processes are accompanied by hydrogen
activation, which leads to two main advantages of metal hydrides application
in plasma. There are a reduction of hydrogen ionization potential by 0.3 –
0.5 eV [3] and an increase in cross section for the dissociative attachment of
slow electrons to neutral hydrogen molecules. For example, the cross section
for the dissociative attachment of electrons to vibrationally/rotationally
excited H2* and D2* molecules rises rapidly with vibrational quantum
number ν increasing and reaches a maximum value σ ≈ 3×10-16 cm2 for ν ≥
5 and for electron energy of several electron volts [4].
Thus, the application of metal hydrides as the electrodes of plasma
devices opens the possibility of increasing the efficiency of ion formation in
plasma devices, and provides an internal local hydrogen supply directly into
the gas-discharge cell.
2. CONDITIONS FOR EXPERIMENTAL STUDIES
Experimental studies were carried out in a discharge with symmetric

Penning configuration of electrodes (Figure 1). The discharge gap was 5 cm
in length. All electrodes were located inside a quartz cylinder placed in the
longitudinal magnetic field of 0.05 – 0.1 T. The cylindrical anode (2), 3 cm
in length, was made from stainless steel and had an internal diameter of 3.7
cm.
Four types of cathodes (1) were used in the experiment. The first one
was a metal hydride cathode (MH-cathode), which was just set in the
discharge. The second one was the same MH-cathode but set in the cathode-
holder (5) with water-cool to keep the temperature unchangeable. The third
one was a copper cathode with balloon hydrogen supply for the simulation
of hydrogen desorption (Figure 2). Balloon hydrogen was injected locally
through a number of thin apertures in the surface of the cathode. The ratio
between the local flow of hydrogen (through the cathode) and the additional
one (in the vacuum chamber) were picked the same as in the discharge with
the MH-cathode. The fourth one was solid copper cathode, and it was the
same as cathode-reflector (3). All the types of the cathodes had the same
spatial dimensions: 2.0 cm in diameter and 0.5 cm thick. The copper
cathode-reflector (3) had an aperture at the center 0.5 cm in diameter. Behind
the aperture a collector (4) was set.
When studying the energy spectra of charged particles, collector (4)
were replaced by multigrid energy analyzer. The high-frequency component
of the registered current was carried out by a spectrum analyzer. The plasma
parameters on the axis of the discharge cell were measured by Langmuir
probes (7) located at half distance between the anode (7) and the cathodes
(1 and 3). The analyzer of negative ions (8) had a collector under the positive
potential of +400 V and a grid under the positive potential of +300 V.
1 – MH-cathode, 2 – anode, 3 – copper cathode-reflector, 4 – collector, 5 – cathode-

holder, 6 – thermocouple, 7 – Langmuir probes, 8 – negative ion analyzer
Figure 1. The scheme of discharge cell.

Figure 2. Copper cathode with hydrogen supply.
The residual pressure in the vacuum chamber did not exceed the value
of 5×10-6 Torr. The pressure higher, than the residual one was achieved by
initial balloon hydrogen supply into the vacuum chamber.
The metal-hydride cathodes were produced from hydride-forming alloy
Zr50V50 by method including its melting, activation and filling with
hydrogen, hydride crushing and mixing it with a binder, and pressing. The
filling of the alloys with hydrogen was performed after its activation. It was
activated in dynamic vacuum at the pressure of 10-2 Torr under the
temperature of 600 K during an hour. Afterwards, the volume with the
sample cooled down to room temperature and was filled with hydrogen
under the pressure of 2 – 4 standard atmospheres. The hydride powder was
loaded in an argon medium, mixed up with copper binder (30 – 60% by
weight) and then it was loaded into the press-form. Pressing was carried out
at room temperature at the pressure of 3 – 5 tf/cm2 during 5 – 10 min.
The saturation coefficient of the alloy with hydrogen was about 180
3
cm /g under normal conditions. The quantity of accumulated in our MH-
cathode hydrogen was about 900 cm3 under normal conditions. Heating the
cathode from 400 to 700 K realizes hydrogen desorption under the pressure
of 10-6 – 1 Torr. The purity of hydrogen desorbed from metal hydride was
better, than 99.999%.
3. THE THERMAL REGIMES OF METAL HYDRIDE

CATHODE OPERATION IN PLASMA
Initially, metal hydrides were considered only as the devices for

compact and safe storage of reversibly chemically bound hydrogen [5]. This
task has not lost the relevance today. However, the area of metal hydride
application has significantly expanded. Now, the delivering of hydrogen
isotopes to a consumer with preset pressure and flow in a spectrally pure
state plays the most important role. The advantages of metal hydride
technology over alternative ones are the possibility of combining various
technological operations in a single device.
One of the promising area of metal hydrides application is their use in
plasma devices. It applies to neutral atoms injectors for fusion [6], ion
sources in accelerator complexes [7, 8], vacuum-plasma process units [9],
neutron generators [10] and other similar objects used hydrogen isotopes as
a working gas. In the cases listed, it is also necessary to solve the tasks of
compact storage, purification and controlled injection of hydrogen isotopes
into plasma.
3.1. Features of Hydrogen Desorption from Metal Hydride

in Plasma
Under the conditions of gas discharge with metal hydride electrodes,

hydrogen pressure is controlled by active thermal impact on metal hydride,
due to its bombardment by high-energy plasma particles. In this case
hydrogen desorption could occur both due to the thermal decomposition of
hydride phases, and due to stimulated hydrogen desorption under the impact
of charged particles bombardment [11].
If a metal hydride electrode plays the role of a cathode, it will be exposed
with ions from plasma. Since the value of ion current enhances, the energy
transferred to the cathode increases, and its temperature rises. The amount
of hydrogen stored in the metal hydride is determined by the area of thermal
desorption curve (Figure 3). As one can see, almost 50% of hydrogen stored
in the metal hydride fits on the high-temperature phase, characterized by the
decomposition temperature of 700 K. Therefore, it was originally expected
that under the conditions of vacuum thermal insulation, the further
enhancement of discharge current will lead to the increase in cathode
temperature with following discharge transition into an arc mode.
Figure 3. The differential curve of hydrogen thermal desorption from Zr50V50 sample.
800 copper
cathode
700
Тc , K
600
500
MH-cathode
400
300
0 10 20 30 40 50 60 70 80
t, min
Figure 4. The dependence of MH-cathode and copper cathode temperatures on

exposure time at Ud = 2,5 kV; Ii = 5 mА; Р = 1×10-3 Torr.
However, even the first studies of the MH-cathode operation in Penning

discharge have shown that the temperature of the MH-cathode stabilizes at
a certain level, which is much lower, than the corresponding values for
materials that do not form hydride phases. Under the conditions of our
experiments, the temperature of the MH-cathode did not exceed Tc ≈ 430 K
at ion current on its surface Ii ≈ 5 mA, while the copper cathode-reflector
was heated above 800 K (Figure 4).
A significant decrease in the temperature of the MH-cathode in plasma
can be explained by following. When the MH-cathode is heated due to ion
bombardment, it is necessary to take into account the total thermal effect of
hydride phases formation/decomposition and convective heat transfer,
which accompanies the processes of hydrogen sorption/desorption.
Therefore, the energy balance equation of the MH-cathode in plasma can be
written as:
Us Ii = S kg P(Tc – T0) + λc (Tc – T0) + S ke σ (Tc4 – T04) + Q (1)
where Us is the space potential of ion formation, Ii is the ion current on the
MH-cathode, S is the surface area of the MH-cathode, kg is the coefficient of
working gas thermal conductivity, Tc is the temperature of the MH-cathode,
T0 is the temperature of discharge chamber walls, σ is the Stefan-Boltzmann
constant, Q is a term that takes into account the total thermal effect of metal
hydride formation/decomposition, as well as convective heat transfer in gas
phase, λc is the thermal conductivity of cathode holder, ke is the emissivity
coefficient of the MH-cathode surface.
The left side of balance equation (1) is the power of heat source, which
is released at the cathode volume when it is bombarded by ions. The first
and the second terms on the right side of the equation describe the loss of
thermal energy due to the thermal conductivity of hydrogen and the elements
of cathode-holder respectively. The third term is the thermal radiation power
of the cathode. The last term takes into account the total thermal effect of
hydride phases formation/decomposition, as well as the convective heat
transfer, which accompanies hydrogen sorption/desorption processes.
The main contribution to the heat balance of the MH-cathode is made

by the thermal effect of hydride phases decomposition during ion
bombardment. Substituting the parameters used in the experiment: Us ≈ 103
V, S = 3 cm2, Ii ≈ 5 mA, σ = 5.67×10-12 W/cm2, Tc ≈ 400 K, T0 ≈ 300 K, one
can get the value of thermal energy loss on the level of 0.05 W due to the
hydrogen thermal conductivity and the thermal conductivity of the cathode-
holder elements. The thermal radiation from the MH-cathode surface is
about 0.3 W, while the total power released on the cathode is about 5 W.
Therefore, as it follows from the energy balance equation (1) the main part
(up to 70%) of the heat power released at the MH-cathode is consumed to
desorb hydrogen due to the decomposition of cathode hydride phases.
Thus, a significant reduction in the thermal loads on such materials is
achieved due to hydrogen desorption as a result of metal hydride thermal
decomposition.
The studies have also shown a significant reduction in the surface
sputtering in comparison with the same materials, but not saturated with
hydrogen. This is explained by the formation of protective gas target under
the metal-hydride surface during the bombardment by high-energy particles
with energies of 0.5 – 5 keV and current densities of 10 – 100 mA/cm2. It
practically eliminates the sputtering of the cathode material during the
experiment (60 min). Such a result was experimentally confirmed in case of
the MH-cathode bombardment by heavy ions (in particular, by argon ones)
[12].
On the other hand, treatment by pulsed plasma streams performed on
QSPA plasma accelerator leads to surface layer melting and significant
weight losses of the sample (up to 0.53 mg/cm2 per pulse) [13].
The duration of plasma stream pulse was 3 – 5 µs. The ion energy was
2 keV, plasma density ~ 2×1014 cm-3, specific power of plasma stream up to
10 MW/cm2, and plasma energy density was varied in the range of 10 – 30
J/cm2.
According to the data of the electron microscopy, the treatment by
powerful pulsed plasma stream leads to melting the sample surface with
following high-speed cooling after the plasma pulse action. Phase analysis
has revealed hydrogen loss on the level of 5% per pulse from total amount
stored in the material.
3.2. The Features of Hydrogen Sorption Processes by Metal

Hydride in Plasma
Along with the well-studied processes of hydrogen desorption under the

influence of bombarding particles from plasma, there are also the processes
of sorption due to the getter properties of the material. Sorption/ desorption
processes are competitive ones and determine the dynamic of neutral
hydrogen concentration in the cell [11].
The dynamic of hydrogen sorption by the MH-cathode was studied
under dynamic vacuum with a monopole mass spectrometer that had its own
pumping system and was connected to the vacuum chamber. In the
experiments the second atomic mass unit (molecular hydrogen) was
scanned. The recording of signals was carried out with a period of 10
seconds. The residual pressure in the vacuum chamber did not exceed
10-2 Torr. The range of investigated pressures was 10-1 – 4 Torr.
Before the discharge was turned on, the whole system had been “blown”
with hydrogen, injected to the chamber from a balloon. After some time
(about 5 minutes), the amplitude of the hydrogen signal on the mass
spectrometer came out to a certain constant value, which we considered as
the “background value” for a given pressure. All the following
measurements were made with respect to the “background value”. The rate
of pump and external gas inlet were not changed during signal recording.
The typical dependences of the molecular hydrogen signal in the
discharge with the MH-cathode on time are shown in Figure 5. One can see,
when the discharge is switched on and the discharge current is set at the level
of Id = 20 mA, the signal amplitude increases. Due to the ion bombardment
the surface contaminations are removed, the sample is heated, hydrogen is

desorbed from it and the pressure and discharge current increase. When the
pressure in the chamber sufficiently rises and the current reaches the value
of Id ≈ 30 mA, we forcibly turn discharge current to the value of 10 mA. At
the same time, the amplitude of the molecular hydrogen signal drastically
falls below the “background value”. A smooth increase in the amplitude to
the “background value” keeps during 15 – 7 minutes depending on the initial
amount of hydrogen stored in the MH-cathode. The signal reaches the
“background value” much faster, when the discharge is forcibly shutdown.
There is about 3 minutes regardless of the initially stored amount of
hydrogen.
When working with copper cathodes, the level of the molecular
hydrogen signal did not fall below the “background value” regardless of the
discharge parameters.
Such a behavior of the signal in the mass spectrum was the same for all
investigated working pressures.
Based on the results obtained, the following explanation of observed
effect seems to be the most probable. A decrease in the signal amplitude
below the “background value”, obviously, is due to the processes of
hydrogen sorption by metal hydride. Under the impact of charged particles
from plasma, hydrogen desorption occurs from the near-surface layer and
leads to its hydrogen depletion. When thermal impact on the metal hydride
decreases, the rate of hydrogen desorption falls and becomes lower, than
sorption one. Since the hydrogenation process takes place most likely on the
surface of the sample, which is placed in the plasma [14], the depleted near-
surface layer is compensated with hydrogen from the gas phase of the
discharge. In our case, when metal hydride used as a cathode of the
discharge, there is an additional gas flow to the cathode in the form of ions.
Therefore, the time when the hydrogen sorption rate exceeds the desorption
one increases in comparison with the case of plasma absence. Since
hydrogen is desorbed from metal hydride, the amount of stored gas
decreases. So, the possible amount of absorbed by such a way hydrogen
decreases unfortunately as well.
Figure 5. The dependence of molecular hydrogen signal on time at P = 0.3 Torr.

a – MH-cathode with 70% of total amount of stored hydrogen.
b – |MH-cathode with 50% of total amount of stored hydrogen.
Thus, when MH-cathode is used in a gas discharge under the conditions

of constant discharge current, the rate of hydrogen desorption substantially
exceeds sorption one. The hydrogen sorption by the MH-cathode in plasma
occurs only with a sharp decrease in discharge current. And a much larger
amount of hydrogen is absorbed under the ion impact from plasma in
comparison with the case of ordinary thermal relaxation of metal hydride.
4. THE FEATURES OF PENNING DISCHARGE

WITH METAL HYDRIDE CATHODE
Reversible hydrogen getters based on Zr-V alloys have a number of

advantages when they used as the materials of structural elements for
plasma-forming devices. A decrease in the ionization potential by 0.5 eV
due to the desorption of hydrogen in the vibrationally/rotationally excited
state, as well as the controlled injection of pure hydrogen (99.999%), along
with the safe and compact hydrogen storage are the main advantages of such
compounds as the elements of plasma-forming units in charged particles
sources.
However, as it turned out, these materials significantly affect the
characteristics of the discharge plasma and its emission properties [15-18].
Moreover, this influence is mutual because the control of hydrogen
desorption from MH-cathode is made by discharge current, which depends

on many external factors, for example, the pressure in the discharge cell, the
composition and the state of injected gas.
4.1. Penning Discharge Modes with Metal Hydride Cathode
In Penning discharge with the MH-cathode the effect of desorbed

hydrogen is foremost an intense yield of electrons in longitudinal direction,
whereas without metal hydride there is no longitudinal electron flux [16].
One can see from the discharge current-voltage characteristics (Figure 6a)
and the collector current dependencies (Figure 6b), that an increase in
discharge voltage Ud forces Penning discharge go through three modes.
In the first mode, at Ud ≈ 1 – 1.7 kV, the electron yield is observed at
both sides of the discharge cell, regardless of the electrodes material. In
earlier studies [19] this was explained by the formation of a large negative
space charge near the anode, which reduced the center potential below the
cathode voltage drop. Collective processes were not taken into account.
Later, in [20], electron yield was associated with the formation of electrons
with anomalously high energy in the unstable anode layer. These electrons
may have so high energy to overcome the potential barrier near the cathode.
Subsequent increase in discharge voltage Ud leads to change the sign of
collector current along with a decrease in discharge current more, than in
two times. The discharge goes to the second mode with an ion flux formation
in the axial direction.
The third mode is observed only when the MH-cathode is used. In this
mode, at Ud ≈ 2.75 – 3 kV discharge current increases again and the
discharge returns to the mode with electrons yield in the longitudinal
direction. In contrast to the transition between the first and the second
modes, here it proceeds more smoothly, mainly due to the compensating
action of ions ejected in axial direction. In this case, electron current is
registered only from the side of the copper cathode-reflector (Figure 6b).
When the MH-cathode is changed in copper one there is only positive
current registered and the third mode is not observed. All of the modes were
observed in the pressure range of 10-5 – 10-4 Torr.
At small Ud the rate of hydrogen desorption from metal hydride is small,
therefore the existence of the first two modes does not depend on the cathode
material, which was observed earlier by a number of authors [19-20]. An
increase in the introduced power leads to the intensification of hydrogen
desorption from the MH-cathode. The interaction of the additional flux of
neutral gas with charged particles substantially changes the discharge
characteristics and it transits to the third mode. The first two modes have
been thoroughly studied, so the main attention was paid to the third mode of
the discharge.
The investigation of the high-frequency oscillation component of
discharge voltage and the axial collector currents revealed a number of
characteristic features. The modulation of collector current is at the same
frequency as the oscillations of discharge voltage. In the experiments with
two copper cathodes, when only the ionic axial current is observed, the
dependence of the oscillations frequency on discharge voltage is close to
linear (Figure 6c). But in case of the MH-cathode, the transition from mode
to mode is accompanied by a sharp change in the oscillations frequency by
the value of 6 MHz (Figure 6c). The frequency decreases during the
transition from the first mode to the second one and increases with the
transition from the second to the third mode. The axial electrons yield from
Penning discharge corresponds to the increased oscillation frequency.
Since a linear dependence of oscillation frequency on discharge voltage
is clearly expressed [16], it could be made an assumption of diocotron
instability development. Taking into account well known ratio for the
frequency of diocotron oscillations f ~ Er/H [21], the dependence of
oscillation frequency (f) on discharge voltage could be explained in the
following way. At transition to third mode the frequency (f) increases due to
both effects. The first one is an increase in radial electric field (Er) due to
discharge voltage rise (linear growth). And second one is owing to
redistribution of axial electric field (Ez) towards Er (frequency jumps,
followed by an increase in discharge current).
Figure 6. The dependence of collector current (a), oscillation frequency (b), discharge
current (c) and floating potential (d) on discharge voltage at P = 3×10-5 Torr, H = 0.1
T.
Figure 6d shows the reason for electrons to yield from the side of copper
cathode-reflector only. One can see, when the discharge goes to the third
mode, a significant decrease in floating potential is observed on the
discharge axis by both cathodes. However, near the copper cathode-reflector
the potential drops to values of 50 V, while near the MH-cathode it drops
only to 450 V. In other words, axial electric field (Ez) falls with the radial
one (Er) rapid growth.
Thus, electrons with energies greater than 50 eV overcome a potential
barrier by the copper cathode-reflector and yield through its central aperture.
Only high-energy electrons can yield from the side of the MH-cathode,
which quantity is very small.
When working with both copper cathodes there is no third mode and
there is no potential drop by the cathodes.
4.2. The Experimental Simulation of Hydrogen Desorption in

Penning Cell
The reason for changing the properties of Penning discharge is the flux
of desorbed hydrogen, which is injected into discharge region locally from
the side of the MH-cathode. To determine its impact on the discharge
properties, the experimental simulation of the MH-cathode was carried out.
It was performed by organizing a balloon hydrogen supply through the one
of the copper cathodes (see Figure 2).
Here hydrogen was injected in two ways: the local one is through the
cathode with H2 supply and the additional one into the vacuum chamber. In
the experiments with both solid copper cathodes hydrogen was injected only
into the vacuum chamber.
The investigations were carried out at the pressure of 10-5 – 10-4 Torr.
The relationship between local and additional hydrogen fluxes, as well as
the magnitude of external magnetic field were chosen to be the same as in
the previous section.
Figure 7. The dependence of collector current (a), discharge current (b) and floating
potential (c) on discharge voltage at P = 3×10-5 Torr, H = 0.1 T.
We note, that in the simulation experiment the negative current on the

collector in the first mode is substantially less, than in case of the MH-
cathode application (Figure 7a). This is associated with different conditions
for organizing the local hydrogen supply. In the simulation experiment the
ratio between local and additional supply was chosen by the third mode of
discharge operation and was maintained the same in the entire range of
discharge voltages. Therefore, for the simulation experiment in the first
mode the relationship between local and additional supply differs from the
case of the MH-cathode. Nevertheless, this fact does not affect the
correctness of the simulation, since the work is devoted to the third mode of
the discharge.
The third mode here starts at 3.0 – 3.5 kV (Figure 7), when at discharge
current increasing (Figure 7b) the current on the collector begins to fall
(Figure 7a). In the simulation experiment collector current also decreases.
But, if in case of the MH-cathode it is because of electrons in the axial flux
of charged particles, then there are no electrons in the axial flux in the
simulation experiment, and a decrease in collector current is associated only
with total discharge current drop (Figure 7b). The absence of electrons in
collector current was proved by the way of a grid under retarding potential
setting in front of the collector Thus, in the simulation experiment, there is
no third mode.
At the same time as the current characteristics change, the distribution
of space potential sufficiently changes as well. In experiments with both
solid cathodes the potentials on the axis by the cathodes were the same and
raised together with discharge voltage (Figure 7c). In the simulation
experiment the potential behavior by the cathode-reflector is close to the
experiment with both solid copper cathodes, while the potential by the
cathode with hydrogen supply sharply increases.
Although the potentials behavior by the cathodes in the simulation
experiments and in the experiments with the MH-cathode is diametrically
opposite, nevertheless, one can see a general regularity. There are much
higher potentials by the cathode with an additional flow of balloon
hydrogen, than by the copper cathode-reflector.
So, the transition to the third mode is possible only if hydrogen is

desorbed from metal hydride, and this is not related to the specific properties
of the MH-cathode surface, since at total hydrogen depletion of the cathode,
third mode disappears and the discharge behaves as in case of solid copper
cathodes [17].
Thus, the experimental simulation of hydrogen desorption in the
Penning discharge indicates on the decisive role of desorbed hydrogen in the
formation of axial electron fluxes at high discharge voltages. Here, the axial
ejection of electrons is not related to the way of hydrogen supply and the
cathode material. The decisive factor for the emission of axial electron
flux is a significant decrease in near-cathodic potentials during the
development of high-frequency instability under hydrogen desorption in
vibrationally/rotationally excited state from metal-hydride.
4.3. Influence of Ion-Stimulated Hydrogen Desorption on

Penning Discharge Properties
The strong dependence of desorption velocity on the temperature of

metal hydrides complicates the application of the MH-cathodes without
additional steps, which are aimed at stabilizing the pressure of the discharge.
To solve this problem, the authors proposed to use the forced water-cool of
the MH-cathode. In this case, the temperature of the cathode will be
maintained at the room level, so the desorption of hydrogen should take
place only due to ion-stimulated processes, and the rate of desorption should
be determined only by the ion current bombardment of metal hydride.
For this purpose, the MH-cathode was placed in a water cool copper
cathode-holder (see Figure 1) with the external diameter of 2.5 cm. To
ensure a good thermal contact between the cathode and the cathode-holder,
the adjoined surfaces were covered with a thermally conductive paste. The
temperature of the cathode was controlled by a thermocouple and do not
exceed 290 K.
Figure 8 shows the typical dependences of pressure (a), discharge
current (b) and collector current on the discharge voltage. The discharge was
ignited at residual pressure P = 5×10-6 Torr. There was no external gas

supply into the vacuum chamber, so the change in pressure during discharge
operation can be uniquely associated with desorption of hydrogen from the
MH-cathode.
One can see a slight change in the pressure when two copper cathodes
are applied (Figure 8a). While the MH-cathodes is being used the pressure
increases due to hydrogen desorption from the cathode under ion-stimulated
processes. It is clearly seen from the value of discharge voltage Ud ≈ 2kV (it
corresponds to the discharge current Id ≈ 0.2 mA (Figure 8b)). Increasing Id
up to 1 mA results in the pressure set on the value of 210-5 Torr. It is
important, that the pressure in the chamber is determined only by discharge
current, and when it changes, the pressure rapidly stabilizes at a new level.
The current-voltage characteristics (Figure 8b) show that for the
discharge with the MH-cathode much higher voltage drops is needed t in
comparison with the case of two solid copper cathodes. This is due to the
following circumstances. The cross section of the dissociative attachment of
low-energy electrons to vibrationally/rotationally excited hydrogen
molecules increases along with the quantum vibrational number of
molecules [4].
Electrons have minimal energy by the cathode due to the trajectory
rotation, as well as due to the emission of low-energy secondary electrons
formed as a result of cathode bombardment by ions. On the other hand, the
MH-cathode desorbs hydrogen molecules already in vibrationally/
rotationally excited state. Therefore, these processes in the near-cathode
region lead to a significant increase in the efficiency of negative ions
formation and, as a consequence, the lack of electrons caused ignition and
discharge operation at higher voltages.
The fact that this phenomenon was not clearly observed in case of
balloon hydrogen supply up to the pressure of 3×10-5 Torr (Figure 6c and
Figure 7b) is explained by insignificant amount of desorbed hydrogen in
comparison with the balloon one. Thus, the processes of dissociative
attachment of excited molecules to slow electrons did not have a significant
impact on the discharge characteristics.
Figure 8. The dependences of pressure (a), discharge current (b) and collector current
(c) on discharge voltage at P = 5×10-6 Torr, H = 0.1 T.
One can see from the Figure 8c that the discharge behaves in the same
way as in the previous experiments with the MH-cathode and with additional
external hydrogen supply into the vacuum chamber. Moreover, an increase
in pressure does not sufficiently shift the voltage of the discharge transition
to the third mode. All these issues give an opportunity to work only on
hydrogen desorbed from the MH-cathode without external supply using the
data carried out in previous experiments [18]. The water-cooling of the MH-
cathode is only the tool for pressure stabilization here.
Thus, the experiments confirm that forced cooling of the MH-cathode
allows to stabilize the pressure in discharge cell and eliminates the hydrogen
waste due to the uncontrolled thermal decomposition of hydride phases. The
low temperature of the MH-cathode (below the decomposition temperature
of hydride phases) ensures the desorption of hydrogen only due to ion-
stimulated processes. This allows to effectively control the rate of hydrogen
desorption by discharge current. It also significantly reduces the
consumption of hydrogen and increases the time of continuous operation of
the cathode. Any qualitative changes are not observed in the discharge
operation in comparison with the uncooled MH-cathode.
4.4. The Influence of Hydrogen Supply Method on Energy

Spectra of Axial Particles
When the discharge works at residual pressure, the electrons have a

weak energy dependence, both on the discharge voltage and on the external
magnetic field. The energy of electron flux lies in the range of 10 – 75 eV
(Figure 9a) with the maximum that corresponds to the value of
approximately 20 eV.
An increase in initial pressure due to balloon hydrogen injection leads
to the broadening of electron energy distribution function (EEDF) toward
higher energy values (Figure 9b). It is important to note that this fact
becomes the most expressed at the pressure P = 3×10-5 Torr, when discharge
operates in the third mode and electrons collect additional energy under the
development of intense diocotron oscillations [19, 20].
Figure 9. The electron energy distribution functions for Ud = 3.5 kV in case of constant
pressure P = 5×10-6 Torr (a) and in case of constant magnetic field H = 0.1 T (b).
The discharge voltages for measuring the ion energy distribution

function (IEDF) were chosen in the following way. Figures for Ud = 2.5 kV
correspond to the second discharge mode, when only ions are registered in
axial direction. Figures for Ud = 3 kV correspond to the transition to the third
mode (electrons start to appear in the axial flux). And figures for Ud = 3.5
kV correspond to the third mode of discharge operation. It should be pointed
out that at residual pressure P = 5×10-6 Torr there is no point in comparing
the simulation experiments, the experiments with both solid cathodes and
the experiments with the MH-cathode, because of different gas environment.
So, there is only data for experiments, when the MH-cathode is applied
(Figure 10 at Ud = 2.5 kV).
One can see qualitatively the same behavior of distribution function in
simulation experiments (curve “cathode with H2 supply” in Figure 10) and
experiments with both solid copper cathodes (curve “copper cathodes” in
Figure 10) regardless of discharge voltage. The only difference is the yield
of particle quantity from the discharge. When additional hydrogen injection
through the one of the cathodes (simulation experiments) is used there are
much more particles registered on the collector.
Figure 10. Ion energy distribution function at H = 0.1 T for different pressures, cathode
types and discharge voltages.
In case of application of the MH-cathode the distribution function shifts

and widens towards the lower values of energy when the discharge transits
to the third mode [22]. If the ion energy is determined by the potential of its
ionization point, a decrease in the energy could be explained by the widening
of intensive ionization field from an anode layer to discharge axis, where
the space potential has a lower value. This phenomenon is the most
pronounced in case of the MH-cathode without water-cool (curve “MH-
cathode” in Figure 10). In this case intensive hydrogen desorption in
rotationally/vibrationally state takes place under the impact of discharge
current that leads to ionization intensification near the axis with lower space
potential. In case of hydrogen desorption only due to ion-stimulated

processes (curve “MH - w/c cathode” in Figure 10) the situation is the same,
but the quantity of desorbed hydrogen is sufficiently lower and the
distribution function widens, accordingly, not so much.
Thus, the analysis of the distribution functions of ejected charged
particles in the longitudinal direction shows the following. Electrons have a
weak energy dependence, both on discharge voltage and on magnetic field.
Ion energy, on the contrary, decreases when the current of axial electrons
rises, which is due to the widening of intense ionization region to discharge
axis.
4.5. Plasma Parameters in Penning Discharge with

Metal Hydride Cathode under the Different Ways of
Hydrogen Injection
The investigation of plasma density depending on the different methods

of hydrogen injection into the cell was carried out with Langmuir probes
placed near the MH-cathode and the cathode-reflector (see Figure 1). Since
the first two modes of the discharge have been thoroughly studied, the main
attention was paid again to the third mode. This mode occurs only when
saturated with hydrogen MH-cathode is used [15-18] and, as it will be shown
below, a great increase in plasma density is observed.
It is interesting that the additional cooling of the MH-cathode not only
leads to the stabilization of hydrogen pressure, but also provides a uniform
profile of plasma density along the discharge axis (curve “by MH - w/c
cathode” in Figure 11) despite the intensive hydrogen injection from the
MH-cathode. The data for plasma density near the cathode-reflector is not
given, because it practically matches with mentioned curve for water-cool
MH-cathode. At the same time, if the MH-cathode are not cooled, then the
gradient of plasma density appears in the direction of the MH-cathode
(Figure 11). This is in accordance with the previous experiments, because in
addition to ion-stimulated processes, the mechanisms of the thermal
decomposition of hydride phases occurs and hydrogen yield sufficiently
increases [3, 15]. In case of hydrogen supply only due to ion-stimulated

processes, the desorption rate is determined only by discharge current, which
adjusts to provide the necessary plasma density in the cell. So, stored
hydrogen is not wasted due to thermal loads on the cathode and desorbed
only in certain quantities for discharge maintaining. The data for the plasma
density with two copper cathodes are not given, since the experiments were
carried out only at residual pressure under hydrogen desorption from the
MH-cathode.
Figure 11. Plasma density on the axis of Penning discharge with different cathodes
depending on discharge voltage, P = 5×10-6 Torr, H = 0.1 T.
Figure 12 shows the characteristic curves of plasma density as a function

of magnetic field. One can see slight difference in curves behavior at
different discharge voltages. It is apparently due to the nonlinear dependence
of discharge current on magnetic field under the conditions of anode layer
instability [21].
depending on magnetic field intensity, P = 5×10-6 Torr.
In Figure 13 the dependences of plasma density on discharge voltage for

the different ways of hydrogen injection into the cell is presented. Here,
hydrogen supply was performed both from the cathode and from a balloon
into the vacuum chamber. The curves 1, 2, and 3 is similar to the
dependences shown in Figure 11 and are explained by the same reasons as
for experiments at residual pressure. The curve 4 corresponds to the
simulation experiment, when balloon hydrogen was injected through thin
apertures in the copper cathode, set instead the metal hydride one. The curve
5 corresponds to the experiment with two solid copper cathodes.
One can see, that regardless of the way of hydrogen supply, the highest
values of plasma density are achieved when the MH-cathode is used. It can
also be concluded that the application of the MH-cathode increases the
efficiency of plasma formation due to the mutual impact of discharge current
on hydrogen desorption.
depending on discharge voltage, P = 3×10-5 Torr, H = 0.1 Т, 1 and 2 – by MH-cathode
and by opposite side correspondingly, 3 – by MH water-cooled cathode, 4 – by cathode
with balloon H2 supply, 5 – by both solid copper cathodes.
Thus, the MH-cathode is not only the source of local hydrogen supply,
but also increases the efficiency of plasma device as a whole [23]. When the
discharge operates on the hydrogen desorbed from the MH-cathode, the
plasma density n sharply increases. Moreover, the largest values of n ≈
1.2×1010 см-3 are achieved by the MH-cathode without water-cool,
regardless of initial gas environment. On the opposite side, by the cathode-
reflector, the plasma density is somewhat lower, so that a density gradient
forms in the MH-cathode direction. The additional cooling of the MH-
cathode reduces hydrogen desorption rate, plasma density takes on lower
values and the density gradient along the axis practically disappears. When
hydrogen supplied only from a balloon, regardless of the method of its
injection, the mode with axial electron yield is not observed and the plasma
density is much lower, than in case of the MH-cathode application.
5. THE IMPACT OF METAL HYDRIDE CATHODE ON

NEGATIVE HYDROGEN IONS FORMATION IN THE
VOLUME OF PENNING DISCHARGE
Traditionally, volume H– sources are important ion sources for

generating high-intensity neutral beams in the problems of plasma heating
and diagnostics [6], as well as proton beams for producing a number of
medical radionuclides used in diagnosis and contact radiation therapy [10].
The production of an initial beam of negative ions is possible in two
ways: by means of sources with surface and volume ionization. In the first
case, significant intensity of H– beam is achieved by cesium addition [7].
But using cesium complicates ion source operation and requires a careful
stabilization of cesium injection and discharge parameters. In the second
case, H– ions are formed in the plasma volume by the mechanism of
dissociative attachment of low-energy electrons to the vibrationally/
rotationally excited molecules of hydrogen [4]. These sources are more
reliable, compact and environmentally friendly (cesium free). However, two
significant drawbacks substantially limit their use. This is an extremely low
intensity of H– beam, as well as the significant leakage of neutral hydrogen
gas together with the extracted beam, which needs to be differentially
pumped to avoid large stripping losses.
Previous studies [24] of H– sources shown that using metal hydrides
could solve this problem. The thermal desorption of hydrogen occurs
already in vibrationally/rotationally excited state. This leads to a sufficient
increase in the cross-section of dissociative attachment. As can be seen
from Table 1, the values of cross sections increase rapidly together
with vibrational quantum number v and reach a maximum value σ ≈ 3
×10-16 cm2 for v ≥ 5 and the electron energy of several electron volts [4].
Thus, the decision to use hydride-forming zirconium-vanadium alloy as
a cathode material could significantly increase the production efficiency of
negative ions. There are two factors responsible for it. Local hydrogen
injection occurs in the cathode region, where are lots of slow electrons and
hydrogen is desorbed already in vibrationally/rotationally excited state.
Table 1. Cross section for the dissociative attachment of slow electrons to

H2 and D2 molecules
Cross section for Cross section for

Vibrational Electron the dissociative Electron the dissociative
quantum energy, attachment of slow energy, attachment of slow
number eV electrons to H2 eV electrons to D2
molecules, cm2 molecules, cm2
0 3.73 1.6∙10-21 3.83 3.0∙10-24
1 3.23 5.5∙10-20 3.45 1.5∙10-22
2 2.73 8.0∙10-19 3.08 3.3∙10-21
3 2.28 6.3∙10-18 2.75 4.2∙10-20
4 1.85 3.2∙10-17 2.43 3.6∙10-19
5 1.45 1.1∙10-16 2.10 2.2∙10-18
6 1.08 3.0∙10-16 1.80 1.0∙10-17
7 0.73 4.4∙10-16 1.53 3.3∙10-17
8 0.40 3.5∙10-16 1.25 9.6∙10-17
9 0.13 4.8∙10-16 1.00 2.3∙10-17
5.1. The Study of H– Ions Formation at Elevated Pressures and

Transverse Extraction
In order to determine the impact of hydrogen supply method on the

efficiency of negative ions production, hydrogen was injected from a balloon
and was also desorbed from the MH-cathode. The pressure during the
experiment was varied from 1×10-2 Torr to 4×10-3 Torr.
The extraction of negative particles was performed perpendicular to
external magnetic field from the near-cathode discharge region (see Figure
1). The registration of negatively charged particles was carried out with an
analyzer (8), consisting of a collector and a pulling grid. The collector was
at the potential of +400 V. The grid was installed in front of the collector
and had the potential of +300 V. The distance between the collector and the
grid was 1 cm.
Figure 14 shows the typical dependences of negative particles current I–
as a function of external magnetic field for the various methods of hydrogen
injection. As one can see from the figure, the dependences have two
characteristic parts. In the first part, the current of negative particles drops
sharply along with magnetic field increasing from 0.02 T to 0.04 T. Such a
behavior was the same for all investigated hydrogen pressures. The sharp
decrease in the current of extracted particles at H < 0.04 T is due to the
retraction of accompanying electrons along the external magnetic field.
To determine the contribution of accompanying electrons to the total
registered current, argon was used as working gas, which does not form
negative ions. To keep the conditions of experiment unchangeable, the MH-
cathode had been not saturated with hydrogen in this case.
In order to analyze the obtained dependences, the current I– was plotted
as a function of hydrogen pressure in the chamber for the different types of
hydrogen injection (Figure 15). As one can see from the figure, the internal
hydrogen injection from the MH-cathode under the impact of high-energy
ions leads to a significant increase in the current I– in comparison with
external hydrogen supply.
One can see from the picture, that the complete cutoff of accompanying
electrons is observed at H > 0.04 T.
Figure 14. The dependence of negative current on external magnetic field at Ud = 700
V, IMH = 4.3mA, Р = 4×10-2 Torr.
Figure 15. The dependence of negative current on the pressure in vacuum chamber at
Н = 0.06 T, Ud = 700 V, IMH = 4.3 mA.
In the range of magnetic field H > 0.04 T and hydrogen pressure from
1.3·10-2 Torr to 4·10-3 Torr the current of negative particles does not
practically change at magnetic field increasing. Therefore, the measured
current at H > 0.04 T can be associated with negative hydrogen ions, which
are formed in this type of discharge.
This is due to the desorption of hydrogen molecules from the MH-
cathode in vibrationally/rotationally excited state. As a consequence, the rate
of H– ions formation increases and their contribution to the total extracted
current of negative particles rises. An increase in the ion current to the MH-
cathode leads to an increase in the yield of negative particles (Figure 16).
Moreover, the higher power transferred to the MH-cathode, the more I–
current registered.
The experiments were carried out only at water-cooled MH-cathode,
because it was the only way to stabilize hydrogen pressure in the cell. If it’s
not, the temperature of metal hydride will rapidly grow up to 400 – 500 K
and the pressure will significantly jump more, than in order of magnitude.
Figure 16. The dependence of negative current on the pressure in vacuum chamber at
different ion current on MH-cathode at Н = 0.06 T, Ud = 700 V.
In the previous studies that have been performed by various authors, it

has been convincingly shown that H– negative ions are formed in a reflective
discharge in hydrogen environment. The yield of negative ions increases
substantially if they are extracted from the near-cathode region [25]. Under
the conditions of our experiments, no special attempts were taken to separate
the H– ions in the extracted beam of negative particles. However, based on
the results obtained by other authors, it can be assumed that the observed
effects of the strong dependence of the current I– on different ways of
hydrogen supply are related to the processes of H– formation.
Thus, hydrogen desorption from the MH-cathode, caused by ion
bombardment of its surface, leads to the significant increase in the yield of
negative ions in comparison with external balloon hydrogen supply. The
yield of negative ions increases together with energy transferred to the MH-
cathode by bombarded particles.
5.2. The Study of H– Ions Formation at Low Pressures and

Longitudinal Extraction
The brightness of ion beam in the sources with volume H– production is

determined mainly by the flow of neutral hydrogen into the interaction
region [26]. From this point of view, using increased hydrogen pressures in
the cell is preferable. On the other hand, the sources of negative ions, as a
rule, are the part of accelerator complex, where the pressure increasing can
lead to breakdowns. Also, increased pressures significantly rise the loss of
H– ions due to collisions.
When a discharge operates at low pressures, the current of extracted
negative ions is extremely low [26]. However, the application of MH-
cathodes can partially solve this problem. First, the desorption of hydrogen
is already in vibrationally/rotationally excited state, and secondly, the
injection occurs locally and automatically adjusts to the discharge current.
Traditionally, H– ions are extracted perpendicularly to external magnetic
field through the aperture in an anode [25, 27]. On the other hand, the
application of MH-cathode sufficiently changes the properties of Penning
discharge. In particular, there is an additional operation mode in the high-
voltage range, where the emission properties of the discharge dramatically
change. Namely, in the axial direction negative current starts extracted [15].
Exactly this fact was the basis for the idea of the creation of negative
hydrogen ions source with longitudinal extraction. The problem arisen here
is the necessity of H– ions separation from the total flux of particles emitted
along the magnetic field. Considering the large difference in mass of a
hydrogen ion and an electron it is conveniently to separate them by
inhomogeneous magnetic field. And the separation of positive ions Н2+ from
negative particles should be done by electric field.
To solve this problem, it was calculated and produced a cathode unit
(Figure 17), which was set instead of cathode-reflector (3) (see Figure 1). It
consists a copper cathode-reflector and a magnetic filter which includes a
grid (1), electrons current collector (2), a coil of magnetic field (4) and a
collector of negative ions (5). The copper cathode-reflector (3) has got an
aperture in the center 0.5 cm in diameter for charged particles extraction.
The magnetic filter was set on the axis of the discharge behind the
aperture in the copper cathode-reflector so, that all reverse magnetic field of
the coil (4) was concentrated outside the discharge cell. For convenience the
distance between the cathode (3), the grid (1) and the electron collectors (2)
were the same and were 0.4 cm. The ion collector (5) were at the distance of
1.8 cm from the copper cathode-reflector (3).
1 – retarding grid, 3 –copper cathode-reflector with an aperture, 2 – electrons collector,

4 – coil of the filter magnetic field, 5 – H– ion collector
Figure 17. The discharge cell with electromagnetic filter.
The cathodes and collectors were under ground potential. The grid (1)
was supplied with +3 kV for positive particles removing. The whole
electrodes system was placed in external uniform longitudinal magnetic field
Hzo0 with intensity that could be changed in the range of Hzo0 = 0 – 0.1 T.
The idea is to create reverse magnetic field in the gap between the
cathode (3) and the collector (5) to deflect electrons on the collector (2), but
not impact on H– ions being registered by collector (5).
We calculated its construction basing on the analysis of the trajectories
of negatively charged particles by numerical solution of a motion equation
in axially symmetric electric and magnetic fields [28]. Taking into account
the configuration of the electrodes, which ensures the registration of only
the paraxial group of particles, the equation of motion can be obtained in the
paraxial approximation in similar way as it was done in [29]. The result is a
paraxial trajectory equation in meridian plane r = r(z):
q rH z20   r02 H zo0  

2
d 2 r 1  o dr
  1   0 (2)
dz 2 2  o dz 8mc 2  o   r 2 H z 0  
 
where φo = φ(r = 0, z) is a potential on the discharge axis relative to the

potential of the particle creation point φo0 = φ(r = 0, z0). In our case, the
potential of the particle creation point (emitter potential) φo0 = 0; Hz0 = H(r
= 0, z) is a magnetic field on the axis at an arbitrary point; Hzo0 = H(r = 0,
z0) is a magnetic field on the axis at the point z0 – the point of the emitter; r0
is the radius of the particle’s entry into the gap. Equation (2) was obtained
under the assumption of a slowly varied electric and magnetic fields in
longitudinal direction.
To solve equation (2) non-uniform magnetic and electric fields in the
investigated gap must be specified. The profile of magnetic field in equation
(2) depends on two variables: Hzo0 = H(0, z0), in our case it is external
magnetic field on the axis; Hz0 = H(0, z), it is nonuniform magnetic field
between the cathode and the ions collector, which is created by reversal
switching of the coils. The values of Hzo0 were fixed and were chosen at the
level of 0.06 T, 0.08 T and 0.1 T. The profiles of Hz0 and φo = φ(0, z) were
calculated in the program femm 4.0 basing on the geometric dimensions of
the cathode unit, the potential of electrodes and the current of the filter coil.
The solution of the equation of paraxial trajectory (2) was carried out
numerically by the fourth order Runge-Kutta method. The result of the
numerical solution is the dependence of the particle position r(z) at a certain
value of the longitudinal coordinate z in the cathode-collector gap. These
results are presented in the Figure 18. For convenience the profiles of total
magnetic field, magnetic coil and the collector of charged particles are
shown in the figure. The position and size of the coil and the collector
correspond to the figure scale. In our calculations, the position z = 0.0 cm
corresponds to the cathode edge and z = 1.8 cm – to the collector edge
(see Figure 17).
One can see that reversal magnetic field has a little effect on the
trajectory of H– ions, whereas, electron trajectories significantly curves, and
at the value of total magnetic field Hz0 = 0 at z = 1.8 cm (collector edge) the
electrons are diverted.
It should be noted that the end of electron trajectories essentially
depends on the initial value of the constant r0, where r0 is the radius of
electron entry in the gap. For example, at zero magnetic field on the collector
edge and r0 > 0.08 cm all the electrons are diverted. Accordingly, electrons
that moves strictly in the coaxial tube 0.16 cm in diameter slightly diverted
by magnetic field and registered by the ions collector. Estimates show that
this current is around an order of magnitude smaller, than the total current
of negative particles registered by the collector. In other words, if the
collector is taken exclusively electron beam, the effect of such a reverse
magnetic field will reduce electron current in about 10 times.
In the same time the trajectories of ions had been little effect on magnetic
field.
An experimental check of the filter operation was carried out using an
electron gun, which simulated the electron flow characteristic for the
discharge cell [22]. The electron gun was set on the axis of the system
instead of the MH-cathode. It created a cylindrical electron beam 1.2 cm in
diameter with a current 10 mA and energy 100 eV. The experimental results
are shown in Figure 19.
One can see that at Hcoil/Hzo0 = 1 the electron beam is diverted almost
completely on the electrons collector. Only a small group of paraxial
particles passes through with a current by order of magnitude smaller than
the total current. Thus, the obtained data are in good agreement with the
calculation. Large errors are caused by using a non-stabilized emission
power source of the electron gun.
So, as a result of numerous calculations, it was built a model that allows
to choose the best external parameters for the efficient separation of H– ions
from the axial flow of charged particles. Good coincidence between the
experimental and the calculated data shown the possibility to apply the
model for the interpretation of following experiments.
This figure is also shown the typical curve in the case of discharge work
(dot line) at Ud = 5 kV and Hz0 = 0.08 T. The registered current at Hcoil/Hzo0
= 1 turns out to be higher than the electron current. Taking into account all
the foregoing, in our opinion, it will be justified to consider the collector
current at Hcoil/Hzo0 = 1 as the current of negative hydrogen ions.
Figure 18. The trajectories of electrons and ions as well as the normalized profile of
magnetic field in the gap cathode (z = 0 cm)-collector (z = 1.8 cm).
At Hcoil/Hzo0 > 1 the effect of the axial electrons decreases, but the
magnetic field of the filter Hcoil is already distorts the external magnetic field
Hzo0 in the discharge cell. Therefore, we have not carried out data analysis
for Hcoil/Hzo0 > 1 in the framework of this paper.
The dependence of the collector current Iion col on the discharge voltage
Ud when the filter is switched on and the filter is switched off is shown in
Figure 20. The filter has the optimal distribution of Hz0 when Hcoil/Hzo0 = 1
and the retarding grid has a potential +3 kV.
One can see when the filter is switched on, the ion current H– is
registered only in those modes when the total extracted current also takes on
negative values. The behavior of the total extracted current is described
previously.
Figure 19. Current on H– ion collector 5 (Iion col) and on electron collector 2 (Iel. col)
depending on Hcoil/Hzo0, solid and dash lines – experiments with electron gun, dot line –
experiment with discharge.
It is of interest that although the total extracted current sufficiently

depends on the magnetic field in the cell Hzo0, this dependence weakens for
H– ions. This is due to the fact that the total current consists the current of
electrons, negative and positive hydrogen ions. Each of them has its own
dependence on discharge parameters, while the current of H– ions apparently
depends mainly on the rate of desorption of activated hydrogen. This clearly
shows a tendency to increase this current when the discharge voltage rises.
Thus, it is possible to increase the brightness of the source by raising the
discharge voltage (increasing the flow of desorbed hydrogen). The weak
dependence of the negative ions current on the magnetic field, apparently,
will allow source to be applied in a wide range of external magnetic fields.
Thus, the application of the MH-cathode in Penning discharge allows to
produce axial H– ion flow along external magnetic field. H– ion production
takes place in the near-cathode region caused by the dissociative attachment
of thermal electrons to hydrogen, desorbed from the MH-cathode, that is
already in vibrationally/rotationally excited state. It is sufficiently increases
the efficiency of H– ion formation due to the energy saving, that would be
spent on the excitation of a H2 molecule in plasma.
Figure 20. Collector current when the filter is switched on Hcoil/Hzo0 = 1 (solid lines)
and switched off (dash lines) depending on discharge voltage.
The configuration of electrodes and electrical field of the discharge

ensures the registration of only the paraxial group of particles. Separation of
H– ions flow from electrons and positive ions could be made with electro-
magnetic filter. Positive ions are retarded by electric field of the grid and
electrons are diverted by magnetic field. Performed calculations together
with experimental verification shows, that there are not more than 10%
electrons in the registered H– current. At the same time, electrical field of
the grid fully reflects positive ions.
CONCLUSION
Thus, the application of the MH-cathode in hydrogen plasma causes

significant reduction in the thermal loads on such materials, which is
achieved due to hydrogen desorption as a result of metal hydride thermal
decomposition. The rate of sputtering for those materials by plasma ions
significantly reduces as well as heat loads. It is due to the creation of
protective gas target under the surface of metal hydride as a result of both
the thermal decomposition of the material and ion stimulated desorption.
Hydrogen sorption by the MH-cathode in plasma occurs only with a
sharp decrease in the ion current on the surface of metal hydride. And much
larger amount of hydrogen is absorbed under the ion impact from plasma in
comparison with the case of ordinary thermal relaxation of metal hydride.
Local hydrogen injection under the impact of ion bombardment of metal
hydride provides return coupling between the intensity of gas desorption and
the parameters of gas discharge. In Penning discharge with the MH-cathode
the effect of desorbed hydrogen is, foremost, an intense yield of electrons
together with ions in longitudinal direction, whereas without metal hydride
there is no longitudinal electron flux. The decisive factor for the emission of
axial electron flux is a significant decrease in near-cathodic potentials during
the development of high-frequency instability under hydrogen desorption in
vibrationally/rotationally excited state from metal-hydride. These electrons
have a weak energy dependence, both on the discharge voltage and on the
magnetic field. The ion energy, on the contrary, decreases when the current
of axial electrons grows, which is due to the widening of intense ionization
region to the discharge axis. Sufficient increase in plasma density is
observed in this case.
The water cool of the MH-cathode allows to stabilize the pressure in the
discharge cell and eliminates the hydrogen waste due to uncontrolled
thermal decomposition of hydride phases. Keeping temperature of the MH-
cathode below the hydride phases decomposition one ensures desorption of
hydrogen only due to the ion-stimulated processes. So, the discharge current
could effectively control the hydrogen desorption rate. It also significantly
reduces the consumption of hydrogen and increases the time of continuous
operation. The qualitative changes in the discharge operation are not
observed in comparison with uncooled MH-cathode.
Hydrogen desorption from the MH-cathode, caused by the ion
bombardment of the surface, leads to a significant increase in the yield of
negative ions in comparison with external balloon hydrogen supply.
Negative ions production takes place in the near-cathode region caused by
the dissociative attachment of thermal electrons to hydrogen, desorbed from
the MH-cathode, that is already in vibrationally/rotationally excited state. It

is sufficiently increases the efficiency of H– ion formation due to the energy
saving, that would be spent on the excitation of a H2 molecule in plasma.
The yield of negative ions increases together with the energy transferred to
the MH-cathode by bombarded particles. Moreover, negative ions could be
simply extracted in longitudinal direction due to the application of metal
hydride in Penning discharge, which operates at low pressure.
REFERENCES
[1] Shmal’ko, YuF; Solovey, VV; Lototsky, MV. Hydrogen Energy

Progress X, 1994, 2, 1311-1319.
[2] Lototsky, MV; Yartys, VA; Klochko, YeV; Borisko, VN; Starovoitov,
RI; Azhazha, VM; V’Yugov, PN. J. Alloys Compounds, 2005, 404-
406, 724-727.
[3] Shmal’ko, YuF; Klochko, YeV; Lototsky, NV. Int. J. Hydrogen
energy, 1996, 21, 1057-1059.
[4] Wadehra, JM. Nonequilibrium vibrational kinetics; Capitelli M.; Ed.;
Topics in Current Physics, 39, Springer: Berlin, 1986, Vol. 39, pp.
191-225.
[5] Sandrock, G. Hydrogen energy system. Production and utilization of
hydrogen and future aspects; Yurum Y.; Ed.; Nato Science Series E,
295. Kluwer Academic Publishers: Norwell, MA, 1995, Vol. 295, pp.
135-166.
[6] Peters, J. Rev. Sci. Instruments, 2002, 71 (2), 1069-1074.
[7] Dudnikov, V. Rev. Sci. Instruments, 2002, 73 (2), 992-994.
[8] Stammbach, Th; Adam, S; Mezger, A; Schmelzbach, PA; Sigg, P.
Proc. 8th Europ. Part. Acc. Conf, Paris, 2002, 159-163.
[9] Borisko, VN; Klochko, YeV; Lototsky, MV; Popov, VV; Shmal’ko,
YuF. Int. J. Hydrogen energy, 1999, 24, 169-174.
[10] Henderson, S. Proc. 25th Linear Accelerator Conf., LINAC10,
Tsukuba, Japan, 2010, 103, 11-15.
[11] Klochko, YeV; Lototsky, MV; Popov, VV; Shmal’ko, YuF; Borysko,
VN. J. Alloys Compounds, 1997, 261, 259-262.
[12] Borisko, VN; Bobkov, VV; Chebotarev, VV; Garkusha, IYe;
Glazunov, GP; Klochko, YeV; Lototsky, MV; Poltavtsev, NS;
Shmal’ko, YuF; Starovoitov, RI; Tereshin, VI. Hydrogen and Helium
Recycling at Plasma Facing Materials; Hassanein A.; Ed.; Nato
Science Series II: mathematics, physics and chemistry. Kluwer
Academic Publishers: Norwell, MA, 2002, Vol. 54, pp 205-212.
[13] Borisko, VN; Garkusha, IE; Chebotarev, VV; Lototsky, MV; Langner,
J; Sadowski, MJ; Tereshin, VI; Shmal’ko, YuF. J. Nucl. Mat., 2003,
313-316, 465-468.
[14] Barosi, T; Giorgi, A. Vacuum, 1973, 23 (1), 15-19.
[15] Afanas’eva, IA; Borisko, VN; Klochko, YeV; Tseluyko, AF. Probl.
Atomic Sci. and Tech. Series: plasma physics, (13) 2007, 1, 191-193.
[16] Klochko, YeV; Ryabchikov, DL; Sereda, IN; Tseluyko, AF. Probl.
Atomic Sci. and Tech. Series: plasma electronics and new acceleration
methods, (7) 2010, 4, 226-229.
[17] Borgun, IV; Ryabchikov, DL; Sereda, IN; Tseluyko, AF. Probl.
Atomic Sci. and Tech. Series: plasma physics, (83), 2013, 1, 228-230.
[18] Agarkov, AV; Ryabchikov, DL; Sereda, IN; Tseluyko, AF. Probl.
Atomic Sci. and Tech. Series: plasma electronics and new acceleration
methods, (86), 2013, 4, 301-303.
[19] Krendel, YuE; Ionov, AS. Tech. Phys., 1964, 34, 1199.
[20] Borisko, VN; Petrushenya, AA. Tech. Phys., 2003, 48, 80-84.
[21] Khauer, W. J. Appl. Phys., 1966, 37(2), 602-611.
[22] Borgun, IeV; Ryabchikov, DL; Sereda, IN; Tseluyko, AF. J. Physics:
Conference Series, 2014, 514, 012051.
[23] Sereda, IN; Tseluyko, AF; Ryabchikov, DL; Babenko, IV; Hrechko,
YaO; Hetman, VA. Probl. Atomic Sci. and Tech. Series: plasma
electronics and new acceleration methods, (98), 2015, 4, 342-344.
[24] Shmal’ko, YuF; Lototsky, MV; Solovey, VV; Yartys’, VA; Strokach,
AP. Z. Physik. Chem., 1994, 183, 479-483.
[25] Ehlers, KW; Gavin, BF; Hubbard, EL. Nucl. Instrum. and Meth. 1963,
22 (1), 87-92.
[26] Bacal, M; Hamilton, GW. Phys. Rev. Lett., 1979, 42, 1538-1540.
[27] Borisko, VN; Borisko, SV; Lapshin, VI; Maslov, VI; Klochko, YeV;
Ryabchikov, DL; Sereda, IN. 29th EPS Conf. Plasma phys. Сontr.
Fusion, Montreux, Switzerland, 2002, 26B, 2.036.
[28] Sereda, IN; Tseluyko, AF; Ryabchikov, DL; Babenko, IV; Hrechko,
YaO; Hetman, VA. Probl. Atomic Sci. and Tech. Series: plasma phys.,
(22), 2016, 6, 241-243.
[29] Brown Ian, G. The physics and technology of ion sources; John Wiley
& Sons: New York, 1989, pp. 85-87.
Chapter 4
PREPARATION AND APPLICATION

OF METAL HYDRIDES
N. E. Galushkin*, N. N. Yazvinskaya
and D. N. Galushkin
Department of Technics and Techology,
Don State Technical University, Laboratory of Electrochemical
and Hydrogen Energy, Shakhty, Russia
ABSTRACT
This chapter discusses metal hydrides found in sintered electrodes of

nickel-cadmium batteries with electrochemical method by the way of
electrolyte decomposition onto hydrogen and oxygen. It was shown that as
a result of electrolyte decomposition, oxygen is released from batteries,
while hydrogen is partly released and partly – in virtue of its high diffusion
permeability – is accumulated in sintered matrices of electrodes of nickel-
cadmium batteries in metal hydrides form. It was established that in a case
of a long-term operation of batteries (more than five years), in their
electrodes, a lot of hydrogen is accumulated. Along with this, a gravimetric
capacity of a sintered nickel matrix of oxide-nickel electrode as a hydrogen
*
Corresponding Author address. Email: galushkinne@mail.ru.
194 N. E. Galushkin, N. N. Yazvinskaya and D. N. Galushkin
accumulating unit becomes equal to 20.1 wt%, while its volumetric

capacity is 400 kg m-3. These values exceed thrice all the previous data
obtained with traditional thermo-chemical methods for any reversible
metal hydrides including hydrides of magnesium and complex hydrides.
However, the hydrogen desorption from these hydrides by the
traditional thermo-chemical method runs at the high temperature (800°C)
and very slowly. Also in this chapter, it is proved by experiments that the
thermal runaway can be used as the new high-performance method of
hydrogen desorption from any metal hydrides. On its kinetic and
thermodynamic parameters, this method is considerably superior to
requirements set by the US Department of Energy for hydrogen storage
systems. Notably, this method of hydrogen desorption can work at any
ambient temperature and pressure. Besides, with use of this method, the
hydrogen desorption takes place due to the electrochemical reactions of the
thermal runaway, this process is easily controllable by electrotechnical
methods. Also it should be noted that electrochemical processes are far less
inertial than the thermal processes used in the traditional thermo-chemical
method.
Keywords: hydrogen storage, metal hydride, thermal runaway,

oxide-nickel, electrode, gravimetric capacity, volumetric capacity
1. INTRODUCTION
The studying of metal hydrides started more than 150 years ago, when
it was found that palladium was capable of accumulating lots of hydrogen.
At the present time, binary hydrides have been obtained virtually for every
metal [1, 2]. However for a long time, metal hydrides were out of a wide use
in practice. For the first time, the metal hydrides obtained an essential
commercial application as the negative electrodes in the nickel-metal
hydride batteries [3, 4]. Now, the metal hydrides also have other
applications.
Firstly, they are used in diverse chemical processes. For example,
sodium borohydride, lithium aluminum hydride and lithium
triethylborohydride are usually used as reducing agents in chemical
synthesis; sodium hydride and potassium hydride are used as strong bases in
organic synthesis; hydride complexes represent catalysts and catalytic
Preparation and Application of Metal Hydrides 195
intermediates in many chemical processes (hydrogenation,

hydroformylation, hydrosilylation, etc.) [5-7].
Secondly, in connection with the cost growth of hydrocarbonic energy
sources and worsening of the ecological situation in the world, the use of
environmentally friendly energy sources becomes more and more actual. In
this regard, an ideal energy source is hydrogen as at its burning, the
environmentally friendly water steam releases. The especially extensive
research of various problems of the hydrogen-based power industry started
beginning in 2000 [1, 8-11]. At present, the main obstacle for a wide
proliferation of the hydrogen-based power industry in the world is an
absence of mobile hydrogen storage systems with the necessary parameters
of a specific capacity. By the US Department of Energy (DOE) in close
collaboration with the automotive industry, the requirements were
developed for mobile hydrogen storage systems so that they were able to
compete with the systems based on carbon energy sources. According to
those requirements, the gravimetric capacity of the hydrogen storage
systems must be more than 5.5 wt%, while the volumetric capacity should
be more than 40 kg m-3 [12]. The liquefied hydrogen meets these
requirements. However the process of hydrogen liquefaction and of the
operation of automotive cryogenic tanks is very expensive and explosion-
hazardous. In the case of making use of tanks with compressed hydrogen, in
order to achieve the requirements set by US DOE, the pressure of more than
70MPa is needed. The tanks operation under such high pressure is very
dangerous, especially in the case of a vehicle accident [13-16].
To the date, the most safe and effective means of hydrogen storage is its
accumulation in hydrides or carbon-based materials (activated carbon,
carbon nanostructures, templated carbons, metal-organic frameworks, etc.).
Among metal hydrides, the most promising substrates for hydrogen storage
systems are: simple hydrides based on magnesium (7.66 wt% and 110 kg m-
3
for MgH2), hydrides of transition metals and complex metal hydrides
containing lithium, natrium, kalium or calcium, aluminium or boron [1,
2,17-21]. Nevertheless, until now, despite all researchers’ efforts, no metal
hydrides were obtained meeting all US DOE requirements, both on the
parameters of specific capacity of hydrogen accumulation and on

thermodynamic and kinetic parameters.
Thirdly, metal hydrides can be used in fuel cells as a source of especially
pure hydrogen. In the frame of modern technologies of fuel cells use, it is
required that the hydrogen purity reaches 99.999% as any by-product gases
(especially CO) poison platinum catalyst. At the present time, gaseous
hydrogen is obtained with the aid of the steam reforming of light
hydrocarbons, which produces a gas mixture containing hydrogen and other
gases such as CO (1-3%), CO2 (19-25%), N2(25%), etc. [22]. Metal hydrides
can be used for hydrogen purification as in the case of hydrogenation where
hydrogen is accumulated in metals. This is connected with the fact that
hydrogen possesses a very high diffusion permeability. For example, at the
temperature 20°C, the diffusivity of hydrogen in nickel is approximately
1010 times higher than that of nitrogen or oxygen [1, 2].
Nowadays, the following three methods exist for obtaining metal
hydrides.
First, we have the chemical method of obtaining metal hydrides. In the
frame of this method, metal hydrides are synthesized with the aid of
chemical reactions [1]. However, obtainment of metal hydrides with aid of
chemical reactions is possible only for a limited number of hydrides [2].
Besides, this process is irreversible. That is why this method is now
considered to be unacceptable for hydrogen storage systems.
The second method is the thermo-chemical one consisting of thermo-
chemical hydrogen accumulation in metal hydrides. In the frame of this
method, the process of hydrogenation/dehydrogenation is tuned by setting
certain values of pressure and temperature, i.e.,
n
MeHn 
 P
P ,T
1
,T
1
 Me
2 2
H2 (1)
2
where Me represents metal or alloy, while P1, T1 and P2, T2 are the pressure
and temperature of the hydrogenation or dehydrogenation processes
respectively.
At the present time, this method is the main one for hydrogen storage
systems. However it features a number of shortages, too. Firstly, it is inertial.
For a hydrogen release start, the metal hydrides must be heated up to a
certain temperature value, after which with a certain delay (depending on
hydride type), the hydrogen release starts. A similar procedure is needed for
the termination of the hydrogen release. However, the processes of heating-
up/cooldown as well as the processes of hydrogen release initiation/
termination from hydrides are very inertial. Secondly, the launching of the
hydrogen release process from hydrides and the termination of this process
requires supplying an additional energy to the system, which decreases the
power-generating efficacy of the hydrogen storage systems. Thirdly, the rate
of the hydrogen release from metal hydrides in frame of the thermo-chemical
method is not high. To increase it and thereby improve the kinetic
parameters of the system, a temperature increase is needed, which would
worsen the thermodynamic parameters of the system. That is why until now
metal hydrides have not been found, which – in the frame of the thermo-
chemical method – would meet the US DOE requirements on both kinetic
and thermodynamic parameters.
The third method is an electrochemical one consisting of
electrochemical hydrogen accumulation in metal hydrides. During an
electrolyte decomposition, for example, in alkaline medium, on a cathode,
the hydrogen is released.
2e− + 2H2O → H2 + 2OH− (cathode) (2)
2OH− → 1/2O2 + H2O + 2e− (anode) (3)
The released hydrogen is adsorbed on a cathode and then the major

portion of the hydrogen escapes to the ambient atmosphere, while a small
portion of the hydrogen penetrates inside the cathode. At first, the hydrogen
stays inside the cathode in its unbound state (  -phase); but after a long
electrolysis, the hydrogen starts forming bound states with a metal of this
cathode (  -phase), i.e., a metal hydride is formed [1, 23]. In the case of
ordinary electrolysis during several hours, the hydride is formed in the thin
surface layer of the cathode [23]. That is why the gravimetric capacity of the
cathode as a hydrogen accumulating unit turns out to be rather small. This
method of hydrogen accumulation in cathodes has been well-known for a
long time. In particular, this method of hydrogen accumulation is used in the
nickel-metal hydride batteries for the hydrogen accumulation in negative
electrodes [1, 4].
In this chapter, we shall study the electrochemical method of hydrogen
accumulation in metals and compare it with the thermo-chemical method of
hydrogen accumulation.
2. PREPARATION OF METAL HYDRIDES
2.1. Introduction
Never before has the electrochemical method of hydrogen accumulation

been considered to be a method of hydrogen accumulation for hydrogen
storage systems.
However, when we started studying the thermal runaway in the alkaline
batteries [23-38], we saw clearly that with the aid of the electrochemical
method, it is possible to accumulate hydrogen in metals in much larger
volumes than with the aid of the traditional thermo-chemical method.
In batteries with the aqueous electrolytes, the thermal runaway can take
place in a case of their charging at constant voltage or in a case of their work
in the floating mode [39]. In this case, the batteries are heated up sharply to
a high temperature where an explosion is also possible. As a result of the
thermal runaway from the batteries, a lot of steam and gas are released. For
example, in the case of the thermal runaway of a battery KSХ-25 (with the
sintered electrodes by capacity 25 Ah), about 297 liters gas and 189 liters
steam are released (which corresponds to 152ml electrolyte) [35]. The gas
analyzer VOG-2M showed that the gas released from the battery consists of
95.6% hydrogen (284 liters) and 4.4% oxygen (13 liters). In total, a battery
KSХ-25 contains about 180 ml electrolyte. One can suppose that the
released oxygen is a result of the electrolyte decomposition (2, 3) during the

thermal runaway, i.e., this oxygen is the result of a 20.9 ml electrolyte
decomposition. Besides, a part of the electrolyte was left in the pores of the
electrodes despite the very high temperature of the battery electrodes in the
course of the thermal runaway [24, 35].
Thus, by the electrolyte decomposition during the thermal runaway, it is
impossible to explain the release of such a great amount of the hydrogen.
Besides, it should be noted that the amount of hydrogen released in the
course of the thermal runaway (284 liters (H2)) is larger than its content in
the entire electrolyte (180 ml) if to decompose it onto hydrogen and oxygen
(about 224 liters (H2)). Hence, the hydrogen was present in the battery
electrodes even before the thermal runaway.
In accordance with the electrochemical method (2, 3), the hydrogen
could be accumulated in the batteries’ electrodes in the course of their
operation. In the course of a battery KSX-25 charging (according to its
Operation Manual), the battery is overcharged 1.6 times in comparison with
its nominal capacity. The overcharge is necessary for a battery to fully
charge. So during a charging session (due to the electrolyte decomposition),
up to 6 liters hydrogen and 3 liters oxygen are released. When needed, the
electrolyte is replenished into the batteries KSX-25 as this kind of battery is
the common vented one. Hence, it is feasible that the battery KSX-25 could
accumulate enough hydrogen during thermal runaway [35] for 284/6 = 48
charge-discharge cycles. In the experiments [35], the batteries were used
with their operating life more than five years; so they passed through at least
ten times more charge-discharge cycles than that. Thus, these batteries could
easily accumulate the found hydrogen amount.
2.2. Electrochemical Method of the Hydrogen Accumulation in

Electrodes of Nickel-Cadmium Batteries
For the hypothesis testing on the possibility of the hydrogen

accumulation in electrodes of nickel-cadmium batteries during their cycling,
the batteries KSX-25 with different operating life were used. In this case, a
hydrogen amount is supposed to be growing in the electrodes in proportion

to the batteries operating life length. For finding out the amounts of
accumulated hydrogen, the electrodes were subjected to thermal
decomposition in a thermal chamber.
The thermal chamber was represented by a tube 1.8 m long and 2 cm in
diameter (Figure 1). One end of the tube was sealed and this end of the tube
was placed into a muffle furnace. Into the other tube end, a rubber plug was
inserted with a tube for gas withdrawal. An electrode in its coiled state was
placed into a cartridge. Then the cartridge was placed into the thermal
chamber. Inside the thermal chamber, above the cartridge, a ceramic plug 20
cm long was inserted. It depressed convection in the thermal chamber, which
ensured a uniform temperature in the cartridge.
Figure 1. The experimental set-up to study the process of gas release from the heated
electrodes. (1) Manometer, (2) intake chamber accumulating gas, (3) thermocouple, (4)
retort, (5) tap, (6) muffle furnace, (7) metal thermal chamber, (8) rubber plug with a
tube for gas extraction, (9) standard coil and (10) heater.
The intake chamber for collecting the gas 2 was a hermetically sealed
glass reservoir filled with water. Its capacity was 5 L. At the bottom of this
reservoir, a number of tubes were inserted through a rubber plug: a long tube
for supplying gas from the thermal chamber to the top of the reservoir, a tube
for removing excess water, and a leg of a manometer tube 1 for controlling
the pressure of the gas accumulated above the water (Figure 1).
The heating up of the electrode in the thermal chamber resulted in the
gas being released from it. This gas was partially cooled down by its passage
through the standard coil 9 prior to its entering into the gas intake chamber.
As the gas entered the chamber, the pressure above the water increased, and
this was registered by the manometer. In order to balance the internal
pressure in the intake chamber and atmospheric pressure, a part of the water
was removed from the intake chamber into the retort 4. The water level in
the retort gave an indication of the amount of gas that had entered to the
intake chamber during the period studied [25-27].
The electrodes were subjected to decomposition inside of the thermal
chamber at the temperature 800°C, as an essential hydrogen release from
electrodes starts only at the temperature 740°C. The electrodes’
decomposition stopped when a daily hydrogen release from the electrodes
became less than 100 ml. On average, the oxide-nickel electrodes were
decomposed during 13 days, while the cadmium ones lasted 7 days. The
electrodes thermal decomposition was conducted in the mode of 11 hours
per day.
To conduct the experiment, eight batteries KSX-25 were chosen with
different operating lives in the same conditions. From each battery there
were taken three cadmium and three oxide-nickel electrodes. Then, the
electrodes were subjected to the thermal decomposition in the installation
shown in Figure 1. In Table 1, the average values are shown of the amount
of the released hydrogen calculated for three electrodes taken from each
battery.
To find the gas composition, the gas analyzer VOG-2M was used. The
absolute error in the percentage concentrations was 0.3–0.5. The analysis of
the gas released from both cadmium and oxide–nickel electrodes showed
that it consisted only of hydrogen.
Table 1. Change of hydrogen content in electrodes of batteries KSX-25

in proportion to their operating life
No. of the battery 1 2 3 4 5 6 7 8

Period of operation New 1 1.5 2.5 4 5 5.5 6.5
(years)
Amount of gas 0 14 17 26 32 36.1 36 35.9
released (Ni) (L)
Amount of gas 0 10 11.4 14.5 17.8 19.9 20.1 20
released (Cd) (L)
Relative error of data in Table 1 makes 5-6%.
From the Table 1, it is seen that the hydrogen amount in the electrodes
grows in the course of the batteries’ operation. However after five years of
operation, the hydrogen amount in electrodes stops growing, i.e., peak
(maximum) capacity of the electrodes is reached as the hydrogen is stored.
In the battery KSX-25, there are 14 cadmium and 15 oxide–nickel
electrodes. That is why in one battery KSX-25 after five years of operation,
approximately 820 liters hydrogen are accumulated.
As for this hydrogen amount, it is quite realistic that a battery KSX-25
was able to accumulate it during five years of operation. Indeed, according
to batteries’ Operation Manual, a battery KSX-25 is charged in two stages:
at first by the current 10A during 3 hours and then by the current 5A during
2 hours. Hence, batteries are overcharged 1.6 times in comparison to their
nominal capacity. The overcharge is necessary for the batteries to fully
charge. So during one charging session of the battery KSX-25, there can be
up to 6 L hydrogen and up to 3 L oxygen released. Thus, for the 820 liters
of hydrogen found in the electrodes, theoretically, any battery KSX-25 was
able to accumulate them in the course of 137 cycles of charge-discharge.
During five years of its operation, any battery KSX-25 passes through a ten
times bigger number of cycles of charge-discharge. That is why,
theoretically, it is able to accumulate the found hydrogen amount.
Now, let us evaluate the gravimetric capacity of the oxide-nickel
electrode as a hydrogen accumulating unit. After five years of operation
in an oxide-nickel electrode, about 36 liters hydrogen is accumulated
(Table 1). The weight of the oxide-nickel electrode is 24 grams.

Consequently, the gravimetric capacity of the oxide-nickel electrode is equal
to 13.4 mass%. The obtained result exceeds twice the obtained earlier results
for any reversible metal hydrides including magnesium hydride and complex
hydrides [1, 40]. For the volumetric capacity of an oxide-nickel electrode,
we obtain the value 400 kg m-3 as the physical sizes of the oxide-nickel
electrode are equal to 7.3×13.6×0.081cm. This result exceeds more than
three times the earlier obtained results for any reversible metal hydrides [1].
Theoretically, hydrogen can be accumulated either in an active
substance of an oxide-nickel electrode (nickel hydroxide) or in the sintered
nickel matrix of an electrode. If we suppose that the hydrogen is
accumulated in the active substance of the oxide-nickel electrode (in its free
or bound state), in this case, during an interaction of the active substance
with acids resulted in a soluble salt formation, the intercalary hydrogen
should supposedly be released in the amounts found earlier in the
experiments (Table1). For this assumption testing, it is possible to use any
acid which forms a soluble salt with the nickel hydroxide and weakly
interacts with the sintered nickel matrix. In our experiments, we used a
22.6% solution of sulfuric acid
Ni(OH) 2  H 2SO 4  NiSO 4  2H 2 O . (4)
As a result, the soluble salt is formed (the nickel sulfate). In our

preliminary experiments, it was found that after 20 minutes of etching in the
sulfuric acid, the mass of the oxide-nickel electrode was not changed. Also
this experiment confirms that the sulfuric acid interacts weakly with the
sintered nickel matrix. That is why in all further experiments, etching of the
oxide-nickel electrodes was conducted during 30 minutes.
To conduct the experiment, there were chosen 8 oxide-nickel electrodes
from different batteries KSX-25 with an operating life of 6 years. After
washing-out in the distilled water and drying-up, the electrodes were
weighed and placed inside of a retort with the sulfuric acid for 30 minutes.
The retort was tightly shut by a rubber plug with a tube for gas withdrawal
into a measuring container 2 (Figure 1) collecting the gas. Then the
electrodes were washed out in the distilled water, dried up and weighed
again. The results of the mass loss of the oxide-nickel electrodes after their
etching in the sulfuric acid are shown in the Table 2.
The experiments showed that during the etching of an oxide-nickel
electrode in sulfuric acid, the hydrogen does not release at all. Hence, in the
active substance of the oxide-nickel electrode, the hydrogen is accumulated
neither in its free state nor in its weakly bound state, which could be
decomposed by the sulfuric acid causing a hydrogen release.
The mass loss of the oxide-nickel electrodes as a result of their etching
in the sulfuric acid makes 33-36% (Table 2). According to the data of the
batteries KSX-25 manufacturer, a positive electrode contains 30-34% nickel
hydroxide and 1.5-2% cobalt hydroxide. The results obtained for the mass
loss of the electrodes fall exactly into this range. Therefore, during batteries’
operation, with the electrodes’ active substance, the hydrogen did not form
any stable chemical compounds unable to be decomposed in the sulfuric
acid.
Table 2. Mass loss of oxide-nickel electrodes

during their etching in sulfuric acid
No. of the electrode 1 2 3 4 5 6 7 8

Initial weight (g) 24.0 23.8 24.1 23.7 23.8 23.9 24.0 23.9
Final weight (g) 15.5 15.4 15.7 15.4 15.5 15.4 15.6 15.6
Weight loss (%) 35.4 35.3 34.8 35.0 34.9 35.6 35.0 34.7
Thus, the hydrogen can be accumulated only in the sintered nickel

matrix of any oxide-nickel electrode. It should be noted that the transition
elements, to which nickel belongs, are capable of accumulating hydrogen
[1, 2, 41].
For the testing of the hypothesis on the hydrogen accumulation in the
sintered nickel matrix of the oxide-nickel electrode, the installation Figure 1
was used.
To conduct the experiment, four batteries KSX-25 with an operating life
of more than 6 years were chosen. From each battery, three oxide-nickel
electrodes were taken. The electrodes were etched in sulfuric acid for 30
minutes for removal of the nickel hydroxide. Then the electrodes were
washed out in the distilled water and dried up. Then the electrodes were
subjected to the thermal decomposition inside of the installation Figure 1 at
the temperature 800°C. In the Table 3, the average hydrogen amount inside
of the electrodes is given. It was calculated for three electrodes from each
battery. The obtained experimental results for the amount of the hydrogen
released from the sintered nickel matrix coincide (in limits of an
experimental error) with the similar results obtained for the oxide-nickel
electrode (Table 1). Thus, during batteries’ operation, the hydrogen is
accumulated in the sintered matrices of electrodes.
Now let us assess the gravimetric capacity of the sintered nickel matrix
of the oxide-nickel electrode as a hydrogen accumulating unit. After five
years of operation in an oxide-nickel electrode, about 36 liters of hydrogen
is accumulated (Tables 1, 3). The weight of the oxide-nickel electrode is 24
g and the weight of the nickel hydroxide is 8 g. Therefore, the gravimetric
capacity of the sintered nickel matrix of the oxide-nickel electrode is equal
to 20.1 wt%. This value exceeds by three times, the earlier data obtained by
the traditional thermo-chemical method for any reversible metal hydrides
including magnesium hydride and complex hydrides [1, 40]. As the physical
sizes of the oxide-nickel electrode and ones of its sintered nickel matrix
are the same, their volumetric capacities are also the same, i.e., 400 kg m-3
[25, 26].
Table 3. Hydrogen content in sintered nickel matrices of oxide-nickel

electrodes of batteries KSX-25
No. of the battery 1 2 3 4

Period of operation (years) 6.4 6.9 7.4 7.7
Amount of gas released (liters)a 36.1 36.0 36.2 35.9
aThe relative error in the data in Table 3 is 5-7%.
At present, further research is needed for reasons of clarification of the

accumulation of such great volumes of hydrogen in the sintered electrodes
of the nickel-cadmium batteries. However, the fact itself of the hydrogen
accumulation in the electrodes has been established by us reliably in our
studies that lasted more than twenty years [20-30]. Nevertheless, it should
be noted that in the oxide-nickel electrode, a number of factors are present,
which promote the hydrogen accumulation. At the present time, those
factors are under an extensive investigation.
Firstly, the oxide-nickel electrode in the batteries KSX-25 is sintered. It
is produced from finely divided nickel powder. This powder has a badly
destroyed metal crystal structure. As it is known [1], any defects of metal
crystal structure (in particular, dislocations) are traps for hydrogen. Besides,
they decrease the energy of hydrogen atoms in comparison with their
location in normal interstices. All this promotes the hydrogen penetration of
the metal. Hence, any imperfections of a metal crystal structure must cause
a drastic increase of its hydrogen capacity. With this purpose, the hydrides
used in the modern hydrogen storage systems are ground down finely in ball
mills [42, 43].
Secondly, the above experiments (Table 1) showed that the hydrogen
released on a cadmium electrode during battery charging can penetrate into
the pores of both oxide-nickel and cadmium electrodes. Most probably, this
is connected with the dense packing of the electrodes in the batteries KSX-
25. Thus, it is possible that the oxide-nickel electrode accumulates the
hydrogen not electrochemically, but instead because of the high capillary
pressure. According to the papers [44, 45], the concentration CH of the
hydrogen accumulated in metal is connected with a hydrogen external
pressure Р in line with the formula
CH  K P . (5)
The capillary pressure in electrode pores is inversely proportional to the

pores radius R, i.e.
P = A/R, (6)
where A = 0.2 μm MPa is proportionality constant for the given electrolyte.

The finely divided powder used for the manufacturing of the electrodes’
sintered matrix has micro fissures sized from a few hundreds angstrom up
to the size of the metal crystal structure cell. Meanwhile the pores’ average
size in the sintered electrodes of the batteries KSX-25 is a few tens microns.
So the hydrogen accumulated in the micro fissures because of the electrolyte
decomposition can stay there under the pressure reaches 100 MPa and more
[24].
Thirdly, in the oxide-nickel electrode, the active substance is nickel
oxide. However, it is well-known that the oxides of transition metals act as
catalysts of hydrogen accumulation [46, 47].
And finally, the hydrogen accumulation process in the batteries’
electrodes runs during their whole operating life. There is more than five
years in this research. In the frame of the traditional thermo-chemical
method, the process of hydrides preparation takes time, from a few minutes
up to several hours. While analyzing the papers on hydrides investigation
for the last 20 years, we did not find studies in which metal hydrides were
subjected to high external pressures during a very long period of time.
However, the exposure duration of an external pressure is an important
factor in any processes of the penetration of one substance into another. This
is especially so, when the penetration process is extremely slow as it is in
our case.
Apparently, the high capillary pressure of hydrogen, acting for an
extremely long time on the strongly destroyed crystalline structure of the
electrodes’ metal over the catalyst (an oxide of a transition metal) allowed
the accumulation of such a large amount of hydrogen inside the electrodes.
Of course, this assumption still requires both experimental and theoretical
studies.
It should be noted that the battery KSX-25 is operated in the floating
mode. That is why during most of the time, it stays under charging. Within
this time, its electrolyte is decomposed onto the hydrogen and the oxygen.
The hydrogen is featured with very high diffusion permeability. For
example, at the temperature 20°C, the diffusion coefficient of the hydrogen
in the nickel is about 1010 times higher than that of nitrogen or oxygen [1,
2]. That is why when the electrolyte is decomposed onto hydrogen and
oxygen, only the hydrogen is accumulated in the electrodes, while the

oxygen escapes into the atmosphere.
Also it should be noted that the oxide-nickel electrode – on its specific
parameters of hydrogen accumulation – significantly exceeds the
requirements of the US Department of Energy [12]. However, because of
the high temperature of hydrogen desorption and the low rate of
absorption/desorption, these hydrides cannot be used as hydrogen
accumulating units in the construction of future hydrogen storage systems
for vehicles. However, as it will be shown below, if the thermal runaways
for hydrogen desorption from hydrides are used instead of the traditional
thermo-chemical method, these problems can be solved [48].
The similar values of specific capacity of hydrogen accumulation in the
sintered electrodes of the nickel-cadmium batteries were obtained by us also
in investigation of other batteries: KSL-15, KSX-6, KSX-3.5, etc. [31, 33,
34]. In all cases, after five years of operation, the specific capacities of the
hydrogen accumulation coincided with the limits of the experimental error
5%.
The obtained experimental results are of a considerable practical
significance, too, as they widen our concept of possibility of hydrogen
accumulation in metal hydrides.
3. METHODS OF HYDROGEN DESORPTION

FROM METAL HYDRIDES
3.1. Introduction
The experimental studies represented in the previous section showed

that it is possible to obtain metal hydrides with the large specific capacities
of hydrogen accumulation – 20.1 wt% and 400 kg m-3 [26]. For this, it is
necessary to use the electrochemical method of hydrogen accumulation
instead of the traditional thermo-chemical method. The obtained values of
the gravimetric capacity exceeded by almost four times the criteria
established by US DOE for hydrogen storage systems [12]. However, in the
case of use of the traditional thermo-chemical method, the hydrogen

desorption from these hydrides runs only at a very high temperature (800°C).
Besides, a full desorption of the hydrogen from these hydrides takes place
during 140 hours, i.e., very slowly [24-26]. So these hydrides do not meet
the criteria of US DOE on their kinetic and thermodynamic parameters.
In this section, we shall give consideration to the thermal runaway as the
feasible, new effective method of hydrogen desorption from hydrides.
First, let us apply a more detailed analysis to the requirements set by the
US DOE [12] for onboard hydrogen storage systems. The main kinetic and
thermodynamic requirements of the US DOE are represented in the Table 4.
Table 4. DOE technical targets: Onboard hydrogen storage systems
Discharging Rates Units Value

Minimum full-flow rate (g s-1) kW-1 0.02
Start time to full flow (20°C) s 5
Start time to full flow (-20°C) s 15
Transient response at operating temperature 10%-90% s 0.75
and 90%-0%
Min/max delivery temperature °C -40/85
Min delivery pressure from storage system* bar (abs) 3FC/35ICE
Max delivery pressure from storage system* bar (abs) 100
* FC = fuel cell, ICE = internal combustion engine.
The parameters 2-4 (Table 4) define limits for the inertness of a

hydrogen storage system. The parameters 5-7 (Table 4) define the
thermodynamic conditions for the processes of hydrogen absorption/
desorption in hydrogen storage systems. The first parameter defines the least
acceptable rate of hydrogen release from a system for a vehicle of the
average class (3000 lb in weight with a powerplant about 150kW). This
parameter is standardized on the power of a powerplant; that is why it is not
convenient for evaluation of hydrogen storage systems. It is better to
standardize the kinetic parameter on a weight or a volume of the hydrogen
storage system. The automotive vehicles of the average class are featured
with the fuel tank volume about 70 liters and the petrol mass of about 52.5
kg. Consequently, the first parameter of the Table 4, standardized on the

system mass will be equal to:
V = 5.714·10-5 wt% s-1. (7)
The peak value for the hydrogen full flow rate from the system was
found out in the paper [49]. It turned to be equal to 8.333·10-4 wt% s-1.
The kinetic parameter (7) is quite reachable for the modern hydrides
based on magnesium [50] and complex hydrides [51]. Nevertheless, these
hydrides do not meet the US DOE requirements on their thermodynamic
parameters. First of all, this is connected with the fact that the hydrogen
desorption from these hydrides runs at high temperatures.
At the present time, two methods of hydrogen desorption exist.
The first method of hydrogen desorption is the chemical one, i.e.
MeHn  nH2O  Me(OH)n  nH2 . (8)
With use of this method, the hydrogen amount turns out to be twice its
content in the initial hydride. However, the reaction (8) is possible only for
a limited number of hydrides [2]. Besides, it is absolutely irreversible. At the
present time, this method of desorption is considered to be unacceptable for
hydrogen storage systems – first of all, because of economical
considerations.
The second method of hydrogen desorption is the thermo-chemical one
discussed in detail in the section 2. At the present time, this method is the
main one for hydrogen storage systems.
In this section we study thermal runaway as the new method of hydrogen
desorption from hydrides.
3.2. Thermal Runaway as the New Method of Hydrogen

Desorption from Hydrides
At first, we shall investigate the hydrogen release rate from the

electrodes of the nickel-cadmium batteries KSX-25 in the case of use of the
traditional thermo-chemical method of hydrogen desorption. With this
purpose we shall make use of the installation Figure 1. In the experiments,
the batteries KSX-25 with their operating life of 6 years were used. The
thermal decomposition of oxide-nickel and cadmium electrodes ran at
different temperatures. The hydrogen release rate was measured every hour
on the hydrogen amount released every five minutes. The process of the
electrodes thermal decomposition stopped when a daily hydrogen release
became less than 100 ml. The results of the experimental studies are
represented in Figure 2.
Figure 2. Rate variations of hydrogen release in a course of thermal decomposition of

oxide–nickel electrodes (a) and cadmium electrodes (b) of battery KSX-25 at heating-
up different temperatures.
From Figure 2, it is seen that despite the high temperature (which

exceeds much the requirements set by US DOE (Table 4)), the rate of
hydrogen desorption from the electrodes is very small.
Now, let us study the thermal runaway as the new possible method of
hydrogen desorption from metal hydrides.
Since the probability of the thermal runaway grows in proportion to the

growth of the battery temperature, its charging voltage and its operating life
[29], the batteries cycling was conducted in the thermal chamber at the
temperature 45°C. Besides, in the experiments, the batteries KSX-25 were
used with their operating life of 6 years and the charging session was
conducted at the high voltage 2.2V for 10 hours. The discharging was
performed by the current 10A down to the voltage 1 V. In this series of
experiments, despite the severe conditions of cycling, we had to perform 353
cycles of charging-discharging until we could see initiation of the thermal
runaway as the thermal runaway is a very rare phenomenon [24, 31].
In the process of thermal runaway on the terminals of a battery, the
voltage was measured. A charge current was measured with a shunt
resistance of 0.001 Ohm and maximum current of 500A.
As a result of the thermal runaway, the steam-gas mixture [24, 48]
released from the battery passed through the standard coil cooler and in so
doing, it cooled down. Along with it, the steam transformed into water,
which was collected in a special storage reservoir, while the gas was
collected in an elastic container, 1070 liters in volume; then this gas amount
was measured. The gas composition was measured with the aid of the gas
analyzer VOG-2M. The absolute error in the concentration percentage was
0.3-0.5. It was found that as the result of the thermal runaway, there were
280 liters hydrogen, 10 liters oxygen and 151 ml water released. The oxygen
release from the battery in the course of the thermal runaway is connected
with the electrolyte decomposition (see the section 2).
The amount (in wt%) of the desorbed hydrogen in proportion to the time
was found out in two stages. Firstly, the volume of the released hydrogen
was measured as a function of time during the thermal runaway. Secondly,
the mass of the electrodes participating in the process of the thermal runaway
was evaluated. This was done based on the area of circles of burned separator
on the electrodes in spots of the thermal runaway emergence (Figure 3).
Based on the obtained data, the amount of desorbed hydrogen in wt%
versus the time was calculated. The obtained experimental data on the
changing of current & voltage on the battery terminals and changes of the
amount of the released hydrogen (in wt%) are represented in Figure 4.
Figure 3. Oxide–nickel electrode of KSX-25 battery after thermal runaway.
Figure 4. Change in parameters of the KSX-25 battery during thermal runaway: I is the
charging current of the battery, U is the voltage of the battery terminals, H is the
hydrogen desorbed (in wt%), T is the temperature of the battery positive terminal.
As a result of the thermal runaway from the battery, which lasted about
4 minutes, 280 liters hydrogen was released. Thus, the thermal runaway
drastically accelerates the hydrogen escape from the hydrides in comparison
with the thermo-chemical method.
Besides, it should be noted that the total hydrogen amount (in wt%)
desorbed during the thermal runaway (Figure 4) was approximately
equal to these electrodes’ gravimetric capacity, experimentally found out in
section 2.
3.3. Discussion
Let us compare the rate of hydrogen desorption from the electrodes of

the nickel-cadmium batteries using the traditional thermo-chemical method
versus the use of the thermal runaway.
In the case of the traditional thermo-chemical method use, the rate of the
hydrogen desorption from the oxide-nickel electrode in its initial moment
was equal to 12.8 ml min-1 (Figure 2). The average weight of the oxide-
nickel electrodes was 24 g. Consequently, the full flow rate of the hydrogen
from the oxide-nickel electrode is equal to 7.94·10-7 wt% s-1. Similarly for
the cadmium electrode, we obtained the value 7.35·10-7 wt% s-1. From the
comparison of the obtained values with the least acceptable value (7) set by
the US DOE, it is seen that these values are approximately one hundred times
less. Besides, the desorption temperature (800°C) greatly exceeds the
thermodynamic requirements set by the US DOE (Table 4).
Now, let us evaluate the rate of hydrogen release from the electrodes of
the battery KSX-25 in the case of use of the thermal runaway as the new
method of the hydrogen desorption from hydrides. The hydrogen desorption
rate was determined using the data (Figure 4) within the scale of 20–80%
completion desorption processes [50, 52]. In this case, for the full flow rate
of hydrogen, we obtained the value 1.2 wt% s-1. From the comparison
between the obtained value and the least acceptable value (7) set by the US
DOE, it is seen that this value is approximately 20 000 times larger.
Now let us give consideration to the thermodynamic parameters of the
thermal runaway as the new method of hydrogen desorption from metal
hydrides. Firstly, the thermal runaway can take place at any ambient
temperature. Secondly, for the thermal runaway initiation, there is no need
for additional energy for the heating-up of metal hydrides. On the contrary,
during the thermal runaway, a very large amount of energy is released (about
5012 kJ [30]), which also can be utilized rationally. So on kinetic and
thermodynamic parameters, the thermal runaway exceeds far behind the
minimum requirements set by the US DOE [12] (Table 4).
However, the thermal runaway, as the method of hydrogen desorption
from hydrides, also contains a good deal of essential shortages. Firstly, under
common conditions, the thermal runaway is very unlikely for the cycling of
nickel-cadmium batteries [28-32]. Secondly, this is an episodic and hardly
repeatable phenomenon. However, based on comprehension of the real
mechanism of the thermal runaway, these problems can be solved.
According to the standard point of view [39, 53], the thermal runaway
takes place after a battery’s long charging session at constant voltage (or in
a case of battery work in the floating mode). Along with it, it is thought that
the reason of the thermal runaway is an acceleration of standard
electrochemical reactions to battery charging due to its self-heating [39].
However in the papers [29, 30], a lot of experimental facts are reported,
which contradict the widely accepted mechanism of the thermal runaway
[39].
Firstly, the widely accepted mechanism of the thermal runaway
supposes that during the thermal runaway, a battery heat-up is going on,
which results in a decrease of its internal resistance and subsequently in
charging current growth, etc. Thus, it is supposed that the thermal runaway
is connected with an intensification of electrochemical reactions running in
the course of a battery charging session. Besides, during thermal runaway,
an intensification is also possible of electrochemical reactions of electrolyte
decomposition (2, 3) as the thermal runaway takes place after a long
overcharge of a battery [39]. In this case, as a result of the thermal runaway
from the battery, a steam-gas mixture must release. The steam is formed
because of electrolyte evaporation due to the high temperature of the thermal
runaway. Meanwhile, gas can be formed during the thermal runaway due to
the electrolyte decomposition onto hydrogen and oxygen. Subsequently, in
the released gas, the ratio of hydrogen/oxygen must be 2:1. However, the
direct experimental testing [24-30] shows that the hydrogen makes not less
than 95% in the composition of the released gas mixture in the thermal
runaway, the remaining negligible part of the gas is oxygen. Notably, in the
released gas mixture, the hydrogen amount is larger than its content in the
full volume of the battery electrolyte to decompose it onto hydrogen and
oxygen [35]. Besides, the amount of the steam released in the thermal
runaway is approximately equal to the electrolyte amount in the battery. This
fact can be explained only if we suppose that the hydrogen was accumulated
in the battery in the course of its operation, which is exactly what
experimental studies confirm (Table 1 and [25, 26]). It should be noted that
in the course of batteries KSX-25 operation, electrolyte is added into them,
when needed.
Secondly, according to the widely accepted mechanism, the thermal
runaway emerges at the expense of that energy, which batteries obtain from
their charger [39]. However, the direct calorimetric measures [30] showed
that as the result of the thermal runaway, a battery releases energy 140 times
more than it obtains from its charger at the same time. This fact can be
explained based only on the assumption that the thermal runaway is
connected with the running of a very powerful exothermic reaction.
Also, other experimental facts exist that contradict the widely accepted
mechanism of the thermal runaway [29, 30].
In the paper [23], it was shown that the reason of the thermal runaway
in nickel-cadmium batteries is an emergence of the exothermic reaction of a
recombination of atomic hydrogen accumulated in battery electrodes in the
course of operation of the battery, i.e.
Hads Cd + Hads Ni → H2 ↑, (9)
which runs in the line of the electrochemical mechanism
H2O + Hads + e− → H2↑+OH− (cathode), (10)

Hads + OH− → H2O + e− (anode). (11)
The hydrogen is present inside the battery electrodes in the form of metal
hydrides – that is in its atomic form [1]. The reaction (9) is the powerful
exothermic reaction with the heat release of 436 kJ/mol (hydrogen). [30, 54].
This heat release is more powerful than in the case of the reaction of
hydrogen burning in oxygen (285.8 kJ/mol) [55, 56].
Also in the paper [23], it was experimentally proved that a rate-limiting
step for the reaction (9) is the step of the metal hydrides decomposition.
MeHx → Me + Hx. (12)
Thus, the real mechanism of the thermal runaway in nickel-cadmium

batteries runs in line with the following scenario [23, 30].
In the course of a battery’s work, two processes of accumulation bring
them step-by-step to the thermal runaway. Firstly, this is the process of the
hydrogen accumulation inside the battery electrodes. Secondly, this is the
process of dendrites accumulation on cadmium electrodes of the nickel-
cadmium batteries.
An initiation of a thermal runaway is connected with a dendrites’
intergrowth through a separator. A distance between a dendrite and an
opposite electrode is much less than the distance between electrodes.
Subsequently, in a spot of the dendrite location, the density of a battery’s
charging current will be much more than in other places on the electrodes.
So this spot will be heated up much stronger than other places on the
electrodes.
With electrodes’ temperature growth, the decomposition rate of the
metal hydrides (12) grows drastically [23, 26, 37]. However, the metal
hydrides decomposition is the rate-limiting step for the exothermic reaction
of the thermal runaway (9). So, in proportion to the metal hydrides
decomposition rate, the intensity of the thermal runaway exothermic
reaction will also grow (9). In its turn, the exothermic reaction (9) will heat
up the electrode even harder in the spot of the dendrite location and in this
way it will increase even faster the metal hydrides decomposition rate (12)
in this location, and so on. This is what the true mechanism is of the thermal
runaway in the nickel-cadmium batteries.
The exothermic reaction (9) requires a new hydride mass for its
continuation. In the dendrite location spot, it will be exhausted and the
reaction will shift radially on the electrode from this point, while burning the
separator in the form of the regular circles. Indeed, in reality, those circles
are observed in experiments (Figure 3 and papers [23-27]).
Two factors prevent an expansion of the exothermic reaction of the
thermal runaway (9) to the entire electrode. Firstly, there is a gas plug, which
is formed between electrodes in the case of the thermal runaway due to the
electrolyte evaporation (because of the high temperature) and to the
hydrogen release (10, 11). Physically, this gas plug can interrupt the
electrochemical reactions of the thermal runaway (10, 11). Secondly, there
is a depletion of the hydrogen in the electrodes in spots of the thermal
runaway.
It should be noted that the emergence of the regular circles indicating
the separator burning during the thermal runaway is impossible to be
explained [30] in the frame of the widely accepted mechanism of the thermal
runaway [39].
At the common operation conditions of the nickel-cadmium batteries
KSX-25, the current generated by the electrochemical reactions (10, 11) is
rather small; approximately it is equal to 0.22A [23]. However, in the case
of the thermal runaway occurrence, the current generated by the
electrochemical reactions (10, 11) becomes equal to hundreds of amperes;
notably, in this case, the voltage on the batteries terminals falls down to
0.5 - 0.6V (Figure 4) in accordance with the voltage for the reactions
(10, 11) [23].
Thus, knowing the true mechanism of the thermal runaway, it is possible
to create it artificially and reliably. Indeed, for the launching of the
electrochemical reactions of the thermal runaway (10, 11), a powerful local
heating up of the electrodes is needed. In the natural conditions of the
batteries KSX-25 work, this is reached due to the dendrites’ intergrowth
through a separator. However a powerful local heating-up of electrodes can
also be caused artificially, by imposing (on battery’ terminals) a voltage
sufficient for punching-through the distance between the electrodes in the
battery.
In our experiments for the artificial and reliable launching of the process
of the thermal runaway, an impact of charging voltage more than100V was
applied onto the terminals of the battery KSX-25 for 0.5 seconds. Then for
the thermal runaway support, it was sufficient to maintain the charging

voltage on the battery terminals more than 1.5V. Usually for this purpose,
we used the voltage 1.87V.
Thus, only 0.5 seconds is needed for the launching of the process of the
thermal runaway. This parameter is ten times better than the US DOE
requirement for initiation of the process of hydrogen desorption from a
system of its storage (Table 4, requirement 2). Besides, the launching time
of the thermal runaway can be further decreased if to increase the charging
impact. In order to stop the process of the thermal runaway, it is enough to
interrupt the electrochemical reactions (10, 11), i.e., to interrupt the process
of battery charging. This can be done almost instantly. Thus, in the case of
this method, the fourth US DOE requirement is also met (Table 4).
With the use of the thermal runaway method, the hydrogen desorption
is performed with the aid of the electrochemical reactions. A rate of the
electrochemical reactions is determined by a current value. Hence, any
electrochemical reactions are easily controllable by electro technical
methods by way of changing of the current value. This feature of the method
of hydrogen desorption from metal hydrides with aid of the thermal runaway
is very important for its practical use.
Let us investigate this feature in more detail. At first, let us evaluate the
parameters of the thermal runaway represented in the Figure 4. In this
experiment, the charging session was made at the constant voltage 1.87V of
the charger. At the moment of the thermal runaway, the voltage on the
battery terminals was equal to about 0.55V, while the peak current was 344
А (Figure 4). Therefore, the resistance of the shunt and the connecting wires
had to be equal to 0.0038Ω. The direct measures of the resistance of the
shunt and the connecting wires confirm the correctness of this value. It
follows therefore that by changing a charger voltage or a resistance of an
external circuit, it is possible to regulate the current of the thermal runaway
and hence the rate of hydrogen release can be controllable.
For this assumption testing, let us investigate the thermal runaway
parameters in the case of another external resistance. We increased twice the
resistance of the shunt and the connecting wires but used the same charger
voltage, which is 1.87V. Besides, we used a battery with the same operating
life that was used in the experiment shown in Figure 4. Along with it, the
initiation of the thermal runaway was conducted in line with the above
described method.
As a result, the peak current of the thermal runaway decreased about
twice the full flow rate of the hydrogen and desorption decreased almost
nine times down to 0.137 wt% s-1 [48].
Thus by regulating the current value of the electrochemical reactions of
the thermal runaway (10, 11), it is possible to set any needed rate of the
hydrogen desorption.
3.4. Conclusion
In conclusion, let us consider the advantages of the thermal runaway as

the new method of the hydrogen desorption from metal hydrides in
comparison with the widely accepted thermo-chemical method.
Firstly, it should be noted that the electrochemical reactions of the
thermal runaway (10, 11) do not depend on a specific metal hydride. That is
why they are applicable for hydrogen desorption from any metal hydrides.
However, beyond doubt, the hydrogen desorption rate depends on a
decomposition rate of specific hydrides and on a diffusion rate of hydrogen
atoms into specific metals [23]. Thus, from the point of view of the scope of
metal hydrides, this method of hydrogen desorption is identical to the
traditional thermo-chemical method.
Secondly, from the point of view of kinetic and thermodynamic
parameters, the thermal runaway method significantly exceeds the
requirements of the US DOE [12]; the same as it does in relation to the best
values obtained in the frame of the traditional thermo-chemical method [1,
17-21]. For its work, the thermal runaway requires no certain values of
temperature or pressure. It can work at any ambient temperature and
pressure.
Besides, for its work (unlike the traditional thermo-chemical method),
the thermal runaway does not require any additional energy. On the contrary,
in the case of the thermal runaway use, a considerable amount of energy is

released, which can be rationally applied.
Thirdly, the hydrogen desorption on the thermal runaway method runs
with the aid of the electrochemical reactions (10, 11). That is why this
process is easily controllable by electro technical methods and hence, it is
far less inertial than the thermal processes used by the traditional thermo-
chemical method.
The conducted experimental researche shows that the thermal runaway
method of the hydrogen desorption from metal hydrides is very promising.
However, for its practical, without a doubt, further experimental and
theoretical researche will be necessary.
REFERENCES
[1] Broom, D.P. Hydrogen Storage Materials; Springer: London, 2011.

[2] Hydrogen in Metals: Basic Properties; Alefeld, G.; Volkl, J.; Eds.;
Springer-Verlag: Berlin, 1978; Vol. 1.
[3] Zhao, X.; Ma, L. Int. J. Hydrogen Energy 2009, 34, 4788-4796.
[4] Crompton T.R. Battery Reference Book; 3rd ed.; Newnes: Oxford,
2000.
[5] Busch, D.H. Inorganic Syntheses; John Wiley & Sons: New York,
2009; Vol. 20.
[6] Brown, H.C. Organic Syntheses via Boranes; John Wiley & Sons:
New York, 1975.
[7] Zweifel, G.S.; Nantz, M.H.; Somfai P. Modern Organic Synthesis:
An Introduction; John Wiley & Sons: New York, 2017.
[8] Ellabban, O.; Abu-Rub, H.; Blaabjerg F. Renew. Sustain. Energy Rev.
2014, 39, 748-764.
[9] Andreasen, K.P.; Sovacool B.K. Renew. Sustain. Energy Rev. 2014,
39, 891-897.
[10] Nikolaidis, P.; Poullikkas, A. Renew. Sustain. Energy Rev. 2017, 67,
597-611.
[11] Mathiesen, B.V.; Lund, H.; Connolly, D.; Wenzel, H.; Østergaard,
P.A.; Möller, B.; Nielsen, S.; Ridjan, I.; Karnøe, P.; Sperling, K.;
Hvelplund, F.K. Appl. Energy 2015, 145, 139-154.
[12] Hydrogen, Fuel Cells & Infrastructure Technologies Program. Multi-
Year Research,Development and Demonstration Plan. Planned
program activities for 2005–2015; U.S. Department of Energy,
Energy Efficiency and Renewable Energy: Washington, 2009.
[13] Gross, K.J.; Carrington, K.R.; Barcelo, S.; Karkamkar, A.; Purewal,
J.; Ma, S.; Zhou, H-C.; Dantzer, P.; Ott, K.; Burrell, T.; Semeslberger,
T.; Pivak, Y.; Dam, B.; Chandra, D. Recommended Best Practices for
the Characterization of Storage Properties of Hydrogen Storage
Materials; V3.34, Feb. 21; EERE: Washington, 2012.
[14] Eihusen, J.A. In 15th World Hydrogen Energy Conf. "WHEC-15";
Yokohama, Japan, June 27-July 2; 2004; pp 301-307.
[15] Energy Technology Analysis; International Energy Agency: Paris,
2005; pp 70-75.
[16] Ahluwalia, R.K.; Hua, T.Q.; Peng, J.K.; Kumar, R. System Level
Analysis of Hydrogen Storage Options; DOE Hydrogen Program
Review: Washington, DC, 2010.
[17] Shahi, R.R.; Yadav, T.P.; Shaz, M.A.; Srivastva, O.N. Int. J.
Hydrogen Energy 2010, 35, 238-246.
[18] Zhang, F.; Zhao, P.; Niu, M.; Maddy J. Int. J. Hydrogen Energy 2016,
41, 14535-14552.
[19] Lototskyy, M.V.; Yartys, V.A.; Pollet, B.G.; Bowman Jr, R.C. Int. J.
Hydrogen Energy 2014, 39, 5818-5851
[20] Li, L.; Jiang, G.; Tian, H.; Wang, Y. Int. J. Hydrogen Energy 2017,
42, 28464-28472.
[21] Pyle, D.S.; Gray, E.M.; Webb, C J. Int. J. Hydrogen Energy 2016, 41,
19098-19113
[22] Hoogers G. Fuel Cell Technology Handbook; CRC Press: Boca
Raton, FL, 2003.
[23] Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N. J. Electrochem.
Soc. 2015, 162, A2044-A2050.
[24] Galushkin, D.N.; Yazvinskaya, N.N.; Galushkin, N.E. J. Power

Sources 2008, 177, 610-616.
[25] Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N. ECS
Electrochemistry Letters 2013, 2, A1-A2.
[26] Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N.; Galushkina,
I.A. Int. J. Hydrogen Energy 2014, 39, 18962-18965.
Soc. 2017, 164, A2555-A2558.
I.A. Int. J. Electrochem. Sci. 2014, 9, 3022-3028.
[29] Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N.; Galushkina
I.A. J. Electrochem. Soc. 2014, 161, A1360-A1363.
Soc. 2015, 162, A749- A753.
[32] Yazvinskaya, N.N.; Galushkin, N.E.; Galushkin, D.N.; Galushkina,
I.A. Int. J. Electrochem. Sci. 2018, 13, 14 – 22.
[38] Yazvinskaya, N.N.; Galushkin, N.E.; Galushkin, D.N. Int. J.
Electrochem. Sci. 2018, 13, 1275 – 1282.
[39] Guo, Y. In Encyclopedia of Electrochemical Power Sources; Garche,
J.; Ed.; Elsevier: Amsterdam, 2009; Vol. 4, pp 241-253.
[40] Sakintuna, B.; Lamari-Darkrim, F.; Hirscher, M. Int. J. Hydrogen
Energy 2007, 32, 1121-1140.
[41] Bernardini, M.; Comisso, N.; Davolio, G.; Mengoli, G. J. Electroanal.

Chem. 1998, 442, 125-135.
[42] Kojima, Y.; Kawai, Y.; Haga, T. J. Alloys Compd. 2006, 424, 294-
298.
[43] Shang, C.X.; Bououdina, M.; Song, Y.; Guo. Z.X. Int. J. Hydrogen
Energy 2004, 29, 73-80.
[44] Borgschulte, A.; Lohstroh, W.; Westerwaal, R.J.; Schreuders, H.;
Rector, J.H.; Dam, B. J. Alloys Compd. 2005, 404–406, 699-705.
[45] Hydrogen Systems; Pergamon Press: Oxford, 1986; Vol. 1, pp 550-
553.
[46] Oelerich, W.; Oelerich, B.; Klassen, T.; Bormann, R. J. Alloys Compd.
2001, 315, 237-242.
[47] Polanski, M.; Bystrzycki, J. J. Alloys Compd. 2009, 486, 697-701.
[48] Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N. Int. J.
Hydrogen Energy 2016, 41, 14813-14819.
[49] Dornheim, M.; Barkhordarian, G.; Bosenberg, U.; Eigen, N.;
Borgschulte, A.; Keller, C; Qi, X.; Metz, O.; Klassen, T.; Bormann R.
In International workshop on functional materials for mobile
hydrogen storage; 20-22 Sept.; GKSS: Geesthacht, 2006.
[50] Dornheim, M.; Doppiu, S.; Barkhordarian, G.; Boesenberg, U.;
Klassen, T.; Gutfleisch, O.; Bormann R. Scr. Mater. 2007, 56, 841-
846.
[51] Puszkiel, J.; Gennari, F.C.; Larochette, P.A.; Troiani, H.E.; Karimi,
F.; Pistidda, C.; Gosalawit-Utke, R.; Jepsen, J.; Jensen, T.R.;
Gundlach, C.; Tolkiehn, M.; Bellosta von Colbe, J.; Klassen, T.;
Dornheim, M. J. Power Sources 2014, 267, 799- 811.
[52] Ma, L-P.; Wang, P.; Cheng, H-M. Int. J. Hydrogen Energy 2010, 35,
3046-3050.
[53] Vincent, C.A.; Scrosati, B. Modern Batteries; 2nd ed.; Butterworth-
Heinemann: Oxford, 1997; pp 312-315.
[54] Blanksby, S.J.; Ellison, G.B. Acc. Chem. Res. 2003, 36, 255-263.
[55] The Hydrogen Economy: Opportunities, Costs, Barriers, and R&D
Needs; National Academies Press: Washington, 2004; pp 240-242.
[56] Suban, M.; Tusek, J.; Uran, M. J. Mater. Process. Technol. 2001, 119,
193-198.
BIOGRAPHICAL SKETCHES
Nikolay Efimovich Galushkin
Affiliation: Don State Technical University, Laboratory of Electrochemical

and Hydrogen Energy, Russia.
Education: He graduated from the Southern Federal University in 1973. He

received Dr.Sc. in Engineering from the South-Russian State
Polytechnical University in 1998.
Business Address: 147 Shevchenko street, town of Shakhty, Rostov

Province, Russia, 346500.
E-mail: galushkinne@mail.ru
Research and Professional Experience:

His research interests include:
First, the research and development of hydrogen storage systems
meeting the criteria for on-board hydrogen storage systems that have been
defined by the US Department of Energy. He received experimentally metal-
hydrides high capacity. The capacity of the metal-hydrides as a hydrogen
absorber was quantified as 20.1 wt% and 400 kg m-3. This value exceed three
times the earlier data obtained by traditional methods for any reversible
metal hydrides, including magnesium hydride or complex hydrides.
Also they far exceed the criteria for hydrogen storage systems established
by US DOE [3, 5] (https://advanceseng.com/chemical-engineering/thermal-
runaway-new-high-performance-method-desorption-hydrogen-hydrides/).
Second, the study of the processes of thermal runaway in alkaline, acid
and lithium-ion batteries. He proved by experiments that the thermal
runaway in alkaline and acid batteries is connected with a powerful
exothermic reaction initiation of the recombination of the atomic hydrogen

accumulated in the electrodes, which runs in line with the electrochemical
mechanism.
Third, the modeling of processes in electrochemical batteries to develop
battery models suitable for practical use in electric vehicles.
Professional Appointments:
He is a professor at Don State Technical University, Russia. He heads a
research laboratory "Electrochemical and Hydrogen Energy".
Nataliya Nikolaevna Yazvinskaya
Affiliation: Don State Technical University, Laboratory of electrochemical

and hydrogen energy, Russia.
Education: She graduated from Don State Technical University in 2004.

She received Cand.Sc. in Engineering from the South-Russian State

Province, Russia, 346500 E-mail: lionnat@mail.ru

Her research interests are focused on the electrochemical processes in
batteries and porous media (metal hydrides, carbon materials,
nanostructures, etc.). She is an active participant in the development of
hydrogen storage systems, high-capacity meeting the criteria for on-board
hydrogen storage systems that have been defined by the US Department of
Energy. She suggested a series of new methods of sorption and desorption
of hydrogen from hydrides and carbon materials.
She is an associate professor at Don State Technical University, Russia.
She is a senior researcher in the research laboratory "Electrochemical and
Hydrogen Energy".
Dmitriy Nikolaevich Galushkin
Affiliation: Don State Technical University, Laboratory of Electrochemical

and Hydrogen Energy, Russia.
Education: He graduated from Don State Technical University in 1998. He

received Dr.Sc. in Engineering from the South-Russian State

Province, Russia, 346500
E-mail: dmitrigall@yandex.ru

His research interests focus on the study of the process of thermal
runaway in alkaline, acid and lithium-ion batteries. The thermal runaway is
encountered in batteries of practically any electrochemical system. Thermal
runaway is the unique phenomenon in electrochemical batteries. It happens
very rapidly, regardless of the attendants with the release of high energy.
Despite of this given phenomenon is very little studied. Studying and
modeling of this dangerous phenomenon is extremely important for the safe
operation of the plurality of electrical facilities and systems. He proved that
the thermal runaway is not associated with the acceleration of known
electrochemical reactions (charge batteries and the electrolyte
decomposition, as thermal runaway occurs during overcharge) as it is
assumed in classical mechanism of thermal runaway. He experimentally
proved that the thermal runaway in alkaline and acid batteries is connected
with a powerful exothermic reaction initiation of the recombination of the
atomic hydrogen accumulated in the electrodes, which runs in line with the
electrochemical mechanism.
He is a professor at Don State Technical University, Russia. He is a
senior researcher in the research laboratory “Electrochemical and Hydrogen
Energy”.
Publications from the Last 3 Years (All authors):

1. Yazvinskaya, N.N.; Galushkin, N.E.; Galushkin, D.N.; Galushkina, I.A.
Int. J. Electrochem. Sci. 2018, 13, 14 – 22.
2. Yazvinskaya, N.N.; Galushkin, N.E.; Galushkin, D.N. Int. J. Electrochem.
Sci. 2018, 13, 1275 – 1282.
3. Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N. J. Electrochem.
Soc. 2017, 164, A2555-A2558.
Int. J. Electrochem. Sci. 2017, 12, 2791-2797.
5. Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N. Int. J. Hydrogen
Energy 2016, 41, 14813-14819.
Soc. 2015, 162, A2044-A2050.
Soc. 2015, 162, A749- A753.
Soc. 2015, 162, A308-A314.
13. Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N.; Galushkina, I.A.

14. Galushkin, N.E.; Yazvinskaya, N.N.; Galushkin, D.N.; Galushkina, I.A.
15. Galushkin N.E., Yazvinskaya N.N., Galushkin D.N. ECS
Electrochemistry Letters. 2015, 4, A94-A96.
INDEX
211, 212, 214, 215, 216, 217, 218, 226,

A
227
benchmarks, 107, 115
accelerator, 154, 157, 182
Boltzmann constant, 156
acid, 203, 204, 226, 227
bonding, vii, viii, ix, 1, 6, 27, 28, 47, 77, 82,
activated carbon, 195
83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93,
activation energy, 133
95, 98, 102, 103, 104, 105, 107, 108,
adsorption, 5, 25, 117
109, 111, 112, 113, 114, 115, 116, 122,
aluminium, 65, 75, 195
128, 134, 136, 137, 138, 140
amplitude, 158, 159
bonds, 5, 17, 19, 37, 55, 58, 83, 84, 87, 92,
argon, 153, 157, 179
93, 103, 109, 119, 122, 135
atomic positions, 6, 50, 52, 53, 54, 59
BOP, 85, 107, 122, 124, 133, 135
atoms, 3, 10, 12, 14, 15, 16, 17, 19, 20, 21,
Borohydrides, v, vii, viii, 1, 2, 3, 4, 5, 8, 9,
24, 25, 26, 27, 28, 29, 31, 32, 35, 36, 39,
10, 15, 16, 20, 26, 27, 31, 35, 43, 46, 59,
40, 42, 44, 45, 47, 50, 52, 54, 55, 57, 58,
60, 64, 65, 66, 67, 68, 69, 70, 74, 75, 76,
82, 83, 84, 88, 89, 90, 91, 92, 94, 97,
77, 78, 87, 109, 119, 142, 143, 148
103, 105, 106, 109, 112, 118, 122, 154
bulk materials, 118, 120, 125
B
C
band gap, 98, 99, 101, 112, 132
Ca2+, 52
batteries, viii, x, 5, 20, 74, 88, 132, 133,
cadmium, viii, x, 193, 199, 201, 202, 205,
137, 144, 145, 147, 193, 194, 198, 199,
206, 208, 211, 214, 215, 216, 217, 218
201, 202, 203, 204, 205, 206, 207, 208,
calcium, 29, 72, 74, 195
carbon, vii, viii, 2, 44, 144, 195, 226
232 Index
catalyst, 118, 196, 207 covalent bond, ix, 3, 82, 84, 87, 90, 92, 93,
charge density, ix, 82, 89, 91, 92, 107, 114, 102, 105, 112, 113, 115, 116, 122
115, 122, 126, 134, 136, 146 critical value, 118, 120, 123
chemical bonding, vii, viii, ix, 1, 82, 83, 85, crystal structure, vii, viii, 1, 2, 3, 6, 7, 11,
107, 109, 111, 112, 113, 115, 122, 137, 12, 13, 16, 17, 19, 20, 25, 29, 30, 31, 34,
138 37, 38, 41, 46, 48, 55, 56, 60, 64, 68, 69,
chemical bonds, ix, 82, 83, 91, 138 70, 72, 74, 75, 76, 79, 87, 89, 91, 95, 96,
chemical method, xi, 194, 196, 197, 198, 97, 125, 130, 138, 145, 146, 206, 207
205, 207, 208, 209, 211, 214, 220, 221 crystalline, ix, 3, 6, 21, 53, 66, 77, 82, 86,
chemical properties, 5, 118 117, 207
chemical reactions, 3, 196 cycles, 133, 199, 202, 212
classification, 74, 85, 86 cycling, 199, 212, 215
clusters, 118, 119, 120, 122, 123, 124, 125
complex hydrides, xi, 2, 3, 46, 87, 92, 94,
D
101, 102, 104, 108, 109, 111, 112, 117,
120, 121, 124, 130, 132, 135, 136, 137,
decomposition, viii, ix, x, 20, 37, 53, 76, 81,
144, 146, 194, 203, 205, 210, 225
91, 94, 104, 117, 119, 120, 129, 155,
composites, 4, 65, 76, 143, 144, 148
156, 157, 170, 189, 193, 197, 199, 201,
composition, 43, 96, 120, 161, 201, 212,
207, 211, 212, 215, 217, 220, 227
216
decomposition temperature, ix, 20, 37, 81,
compounds, vii, viii, 1, 3, 4, 8, 19, 37, 44,
91, 94, 104, 117, 119, 120, 155, 170
47, 51, 53, 64, 84, 85, 88, 92, 96, 100,
deformation, 29, 34, 36, 120
103, 106, 107, 108, 109, 111, 116, 121,
degenerate, 84, 98, 102, 112
134, 136, 137, 160, 204
dendrites, 217, 218
conduction, 78, 84, 98, 100, 101, 112, 132,
density functional theory, 74, 92
137, 145, 146, 148
Department of Energy, xi, 194, 195, 208,
conductivity, 97, 130, 132, 146, 148, 156,
222, 225, 226
157
derivatives, 5, 10, 75, 78, 108
conductor(s), 5, 39, 82, 98, 132, 135
desorption, x, xi, 118, 119, 150, 151, 155,
configuration, 27, 95, 96, 151, 183, 188
157, 160, 161, 162, 167, 168, 170, 172,
constituents, 82, 83, 85, 92, 110, 111, 114,
174, 175, 176, 177, 180, 181, 182, 187,
116, 119, 122
188, 189, 194, 208, 209, 210, 211, 212,
construction, 183, 208
214, 215, 219, 220, 221, 225, 226
consumption, 170, 189
DFT, 21, 29, 92, 94, 98, 137
conversion reaction, 130, 144
dielectric constant, 109
cooling, 12, 13, 21, 24, 27, 29, 35, 40, 158,
diffraction, 6, 10, 11, 12, 13, 14, 21, 27, 29,
170, 173, 176
40, 51, 53, 54, 69, 70, 77, 79, 138
copper, 152, 153, 155, 156, 159, 161, 162,
diffusion, viii, x, 88, 95, 117, 139, 193, 196,
164, 166, 167, 168, 171, 174, 175, 176,
207, 220
182, 183
diffusion permeability, viii, x, 193, 196, 207
correlation, 66, 67, 98
distilled water, 203, 205
covalency, 103, 110
Index 233
distribution, ix, 12, 82, 85, 89, 92, 94, 95, environment, ix, 19, 25, 26, 28, 39, 47, 48,
97, 113, 114, 115, 116, 122, 123, 126, 82, 83, 85, 86, 95, 96, 124, 126, 171,
134, 136, 166, 170, 171, 172, 173, 186 176, 181
distribution function, 98, 170, 171, 172, 173 etching, 203, 204
donors, vii, ix, 82 evaporation, 215, 218
evolution, 15, 50
excitation, 187, 190
E
exothermic reaction, 216, 217, 218, 226,
228
electric field, 110, 162, 164, 182, 184, 188
exposure, 155, 207
electrical conductivity, 84, 87, 98, 133
external magnetic fields, 187
electricity, viii, 1, 84
extraction, x, 150, 178, 182, 200
electrochemical method, viii, x, 193, 198,
199, 208
electrode, xi, 39, 130, 137, 144, 145, 154, F
193, 194, 200, 201, 202, 203, 204, 205,
206, 207, 208, 213, 214, 217, 218, 226, Fermi level, 98, 100, 101, 102, 104, 112
228 flexibility, vii, viii, 1, 8, 112
electrolyte, viii, x, 133, 193, 197, 198, 199, formation, x, 28, 30, 31, 39, 76, 86, 88, 91,
206, 207, 212, 215, 218, 227 101, 112, 120, 144, 150, 151, 156, 157,
electromagnetic, 183 161, 167, 168, 175, 180, 181, 187, 190
electron diffraction, 6, 75 formula, 16, 18, 60, 96, 114, 120, 128, 206
electron microscopy, 157 fragments, 22, 26, 28, 45
electronic structure, ix, 82, 83, 140 free energy, vii, viii, 2
electrons, 83, 84, 87, 88, 90, 91, 92, 93, 95, fuel cell, 68, 196, 209
97, 98, 100, 103, 106, 109, 111, 114,
115, 116, 122, 134, 151, 161, 162, 164,
G
166, 167, 168, 170, 171, 173, 177, 178,
179, 182, 183, 185, 186, 187, 188, 189
gadolinium, 65, 78
emission, 160, 167, 168, 182, 185, 189
geometry, 10, 14, 19, 31, 51, 53, 125
energy, vii, viii, ix, x, 1, 2, 4, 28, 30, 31, 65,
gravimetric capacity, xi, 193, 194, 195, 198,
70, 79, 81, 82, 88, 92, 93, 96, 97, 98,
202, 205, 208, 214
100, 101, 102, 103, 104, 112, 114, 115,
growth, 2, 162, 164, 195, 212, 215, 217
118, 120, 121, 123, 124, 128, 129, 133,
135, 137, 138, 150, 151, 152, 154, 156,
157, 161, 164, 168, 170, 171, 172, 173, H
177, 178, 179, 181, 185, 187, 189, 190,
195, 197, 206, 215, 216, 221, 226, 227 Hamiltonian, 85, 103, 113
energy density, 2, 93, 157 heat release, 151, 217
energy transfer, 154, 181, 190 heat removal, 151
engineering, 46, 225 heat transfer, 156
heavy metals, 46
234 Index
helical polymer, 19 ionicity, 83, 85, 92, 107, 116

holmium, 76 ionization, x, 150, 151, 160, 172, 173, 177,
homogeneity, 87 189
host, 46, 85, 95 ions, viii, x, 12, 19, 83, 84, 86, 88, 105, 133,
hybrid, 114 149, 150, 152, 154, 156, 157, 159, 161,
hydrides, vii, viii, ix, x, xi, 1, 2, 3, 4, 6, 14, 168, 171, 177, 178, 179, 180, 181, 182,
20, 26, 46, 64, 65, 81, 82, 83, 85, 86, 87, 183, 184, 185, 186, 187, 188, 189
88, 89, 91, 92, 94, 98, 99, 101, 103, 104,
108, 109, 111, 112, 117, 119, 120, 121,
K
124, 128, 130, 132, 135, 136, 137, 139,
140, 141, 144, 146, 147, 149, 150, 151,
K+, 18, 37, 51, 55
154, 167, 177, 193, 194, 195, 196, 197,
KBr, 147
203, 205, 206, 207, 208, 209, 210, 212,
kinetic parameters, 196, 197
214, 215, 217, 219, 220, 221, 225, 226
kinetics, ix, 3, 4, 30, 35, 81, 87, 117, 124,
hydrocarbons, 196
129, 130, 190
hydrogen atoms, 10, 14, 19, 25, 27, 33, 49,
95, 126, 206, 220
hydrogen bonds, 5 L
hydrogen gas, 25, 177
hydrogen plasma, 150, 151, 188 lattice parameters, 7, 26, 37, 86
hydrogen storage, ix, xi, 3, 5, 6, 46, 64, 65, ligand, vii, viii, 1, 3, 25, 26, 28, 44
66, 68, 71, 77, 78, 82, 83, 88, 117, 121, linear dependence, 162
122, 124, 128, 129, 135, 137, 140, 141, lithium, 8, 9, 10, 37, 39, 68, 86, 134, 137,
142, 150, 160, 194, 195, 196, 197, 198, 144, 145, 146, 147, 148, 194, 195, 226,
206, 208, 209, 210, 224, 225, 226 227
hydrogenation, ix, 21, 71, 81, 85, 87, 91, local configuration, 28
103, 159, 195, 196 localization, ix, 82, 85, 95, 96, 138
hydrolysis, 3, 4 low temperatures, 117
hydrosilylation, 195 lying, 47, 56, 58, 69, 112
hydroxide, 203, 204, 205
M
I
magnesium, xi, 20, 70, 71, 72, 141, 144,
infrared spectroscopy, 67 194, 195, 203, 205, 210, 225
initiation, 197, 212, 215, 217, 219, 220, 226, magnetic field, x, 150, 151, 164, 170, 171,
228 173, 174, 175, 178, 179, 180, 182, 183,
ion bombardment, x, 149, 156, 157, 158, 184, 185, 186, 187, 188, 189
181, 189 magnetic properties, 5, 91
ion transport, 147 magnetic resonance, 125
ionic conduction, 128, 132, 146, 148 magnetic structure, 139
ionic conductors, 82, 132, 137
Index 235
magnitude, 98, 100, 102, 108, 116, 136, negative ions, x, 83, 150, 152, 168, 177,
164, 180, 185 178, 179, 181, 182, 187, 189
materials, vii, viii, ix, x, 2, 3, 5, 25, 39, 46, neutral, 151, 154, 158, 162, 177, 182
66, 74, 75, 79, 81, 83, 84, 85, 88, 89, 94, NH2, 3, 128, 131, 132, 145, 146
98, 105, 107, 108, 111, 114, 115, 117, nickel, viii, x, 193, 194, 196, 198, 199, 201,
118, 122, 124, 125, 128, 129, 130, 132, 202, 203, 204, 205, 206, 207, 208, 211,
137, 139, 140, 141, 146, 149, 150, 156, 213, 214, 215, 216, 217, 218
157, 160, 188, 195, 224, 226 nickel-cadmium, viii, x, 193, 199, 205, 208,
matrix, xi, 18, 94, 96, 103, 105, 109, 134, 211, 214, 215, 216, 217, 218
193, 203, 204, 205, 207 nitrogen, 25, 196, 207
melting, 86, 87, 123, 153, 157 NMR, 69, 125, 126
metal hydride(s), vii, x, xi, 9, 14, 20, 86, 88, nuclei, 122, 126, 134
92, 94, 117, 130, 137, 139, 140, 141,
149, 150, 151, 153, 154, 156, 157, 159,
O
160, 161, 162, 167, 175, 177, 180, 188,
189, 190, 193, 194, 195, 196, 197, 203,
optical properties, vii, ix, 82, 140
205, 207, 208, 212, 215, 217, 219, 220,
optimization, ix, 14, 46, 82
221, 225, 226
optoelectronic properties, ix, 82
metal ion, 77
optoelectronics, vii, ix, 81
metal oxides, 46, 67
oscillation, 162, 163
metals, 8, 84, 86, 88, 98, 140, 196, 198, 220
oxide-nickel, xi, 193, 194, 201, 202, 203,
Mg2+, 27, 51, 52, 128
204, 205, 206, 207, 208, 211, 214
models, 21, 29, 226
oxygen, viii, x, 193, 196, 198, 199, 202,
modifications, 46, 122, 146
207, 212, 216, 217
molecular orbital, 28
molecular structure, 5, 77
molecules, x, 6, 84, 87, 92, 115, 134, 150, P
151, 168, 177, 178, 180
molybdenum, 139 particle creation, 184
phase diagram, 45, 147
phase transformation, 69, 71, 72
N phase transitions, 15, 21, 24, 27, 30, 70, 75
physical properties, vii, viii, 1, 46, 117
Na+, 12, 14, 18, 83, 128, 137
PIG, 150
NaCl, 12, 16, 18, 83, 86, 88, 89, 92, 93, 100,
polymorphism, 28, 75
104, 106, 107
population, ix, 2, 82, 85, 103, 106, 109, 113,
nano phase hydrides, 82
116, 122, 124, 133, 134
nanocomposites, 141
porosity, 8, 20, 25
nanoparticles, 118, 121, 126
potassium, 37, 48, 57, 68, 78, 194
nanophases, 91, 120, 124, 125, 126, 137
purification, 150, 154, 196
nanoscale materials, 117
purity, x, 149, 153, 196
nanostructures, 142, 195, 226
236 Index
storage, vii, ix, x, xi, 2, 3, 4, 5, 6, 46, 64, 65,

R
66, 68, 70, 71, 77, 78, 79, 81, 83, 88,
117, 121, 122, 124, 128, 129, 135, 137,
radiation, 69, 156, 157, 177
139, 140, 141, 142, 149, 150, 154, 160,
radius, 8, 18, 37, 51, 184, 185, 206
194, 195, 196, 197, 198, 206, 208, 209,
reactions, xi, 4, 46, 66, 67, 78, 87, 129, 141,
210, 212, 219, 224, 225, 226
142, 194, 196, 215, 218, 219, 220, 221,
strontium, 33, 75
227
structural modifications, 12
recombination, 216, 226, 228
structural transitions, 85, 145, 147
reconstruction, 16, 28
structure, ix, 6, 9, 10, 11, 12, 13, 15, 16, 17,
redistribution, 88, 92, 162
18, 19, 22, 24, 25, 26, 27, 28, 29, 31, 33,
renewable energy, vii, viii, 2, 64
35, 36, 37, 39, 40, 41, 43, 46, 48, 49, 50,
requirements, xi, 2, 194, 195, 197, 208, 209,
51, 52, 53, 54, 56, 57, 58, 60, 67, 68, 69,
210, 211, 214, 215, 220
70, 71, 75, 76, 77, 78, 79, 82, 86, 88, 89,
resistance, 212, 215, 219, 220
91, 96, 99, 119, 125, 127, 129, 130, 134,
rings, 39, 56, 57, 58
138, 140, 145, 146, 147, 206, 207
room temperature, 12, 14, 15, 18, 29, 36,
structure formation, 127
40, 46, 49, 52, 84, 132, 137, 153
sulfuric acid, 203, 204, 205
rubber, 200, 201, 203
surface area, 137, 156
surface layer, 124, 157, 159, 198
S symmetry, 9, 13, 14, 15, 18, 21, 29, 32, 33,
42, 43, 44, 45, 46, 49, 50, 53, 55, 58, 96
signals, 51, 53, 158 synthesis, 5, 21, 29, 30, 33, 37, 46, 70, 194
simulation, 7, 152, 164, 166, 167, 171, 175
sintered matrices, viii, x, 193, 205
SiO2, 25, 26 T
sodium, 12, 18, 68, 69, 78, 83, 146, 194
target, x, 149, 157, 189
solid solutions, 87, 138
temperature, viii, xi, 2, 7, 8, 9, 10, 12, 13,
solid state, 4, 46, 120, 146
16, 18, 24, 28, 29, 31, 36, 40, 49, 50, 51,
solubility, vii, ix, 82, 86
53, 69, 76, 98, 105, 119, 120, 123, 124,
solution, 13, 132, 183, 184, 203
129, 132, 136, 137, 148, 150, 152, 153,
sorption, 65, 118, 120, 124, 142, 144, 151,
154, 156, 167, 170, 180, 189, 194, 196,
226
197, 198, 200, 201, 205, 207, 208, 209,
sorption kinetics, 118
211, 212, 213, 214, 215, 217, 218, 221
stability, vii, ix, 3, 28, 68, 72, 81, 105, 118,
temperature dependence, 24
120, 121, 132, 137, 140, 142
terminals, 212, 213, 218, 219
stabilization, 132, 170, 173, 177
testing, 203, 204, 216, 220
state, vii, x, 2, 4, 6, 37, 87, 92, 94, 95, 114,
tetra-boro hydrides, vii, viii, 2
128, 132, 137, 147, 149, 154, 160, 161,
thermal decomposition, x, 46, 70, 149, 154,
167, 168, 172, 177, 180, 182, 187, 189,
157, 170, 173, 188, 189, 200, 201, 205,
190, 197, 200, 203, 204
211
stoichiometry, 86, 88, 145
Index 237
thermal energy, 156, 157

V
thermal properties, 71, 75, 76
thermal relaxation, 160, 189
valence, ix, 82, 83, 84, 85, 86, 87, 88, 89,
thermal runaway, xi, 194, 198, 199, 208,
90, 94, 95, 98, 100, 101, 103, 112, 115,
209, 210, 212, 213, 214, 215, 216, 217,
122, 135, 136
218, 219, 220, 221, 226, 227
vanadium, 150, 177
thermal stability, 75
volumetric capacity, xi, 20, 194, 195, 203
thermo-chemical method, xi, 194, 197, 198,
205, 207, 208, 211, 214, 220, 221
thermodynamic parameters, xi, 194, 197, W
209, 210, 215, 220
thermodynamics, 3, 4, 117 water, 2, 4, 152, 167, 170, 172, 173, 176,
topology, 25, 26, 42, 43, 55, 58 180, 189, 195, 201, 204, 212
total energy, 82, 96, 118, 119, 120, 124
trajectory, 168, 184, 185 Y
transformations, 40, 46, 51, 52, 67
transition elements, 204 yttrium, 75, 78
transition metal, 87, 94, 140, 142, 195, 207
transportation, vii, ix, 81, 134, 137
treatment, 50, 52, 54, 157 Z
zinc, 25
zirconium, 72, 150, 177

Fluid Sterilization Filtration

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Fluid Sterilization Filtration

Загружено:

Авторское право:

Доступные форматы

ANALYTICAL CHEMISTRY AND MICROCHEMISTRY

TYPES, BONDS AND APPLICATIONS

Additional books in this series can be found on Nova’s website

Additional e-books in this series can be found on Nova’s website

TYPES, BONDS AND APPLICATIONS

NOTICE TO THE READER

Library of Congress Cataloging-in-Publication Data

ISBN:  H%RRN

Published by Nova Science Publishers, Inc. † New York

Hydrides: Types, Bonds and Applications first proposes metal hydrides

Chapter 2 - Hydrides are an important class of materials due to their

electrodes, a lot of hydrogen is accumulated. Along with this, a gravimetric

CRYSTAL STRUCTURES OF ALKALI AND

Hydrogen gas is particularly relevant to energy applications as it is a

* Corresponding Author Email: ponniahv@kjemi.uio.no; Web: folk.uio.no/ponniahv.

Keywords: complex hydrides, borohydrides, crystal structure, tetra-boro

The world’s population has surpassed seven billion and is estimated to

their high gravimetric and volumetric capacities. However, their direct

borohydrides have high hydrogen gravimetric densities (Table 2) and are, in

As the simplest stable boron hydride in its condensed phase, diborane

possible hydrogen-based fuel for high-speed jets. In addition, it has been

Crystal Structure of B2H6

The molecular, electronic, and crystal structures as well as the special

Figure 1. Crystal structures of β-B2H6.

GENERAL TRENDS IN THE STRUCTURAL CHEMISTRY

Metals (M) with increasing Pauling electronegativity have increasing degrees

The first report of a pure alkali metal borohydride (tetrahydroborate or

2MH + B2H6 2MBH4, where M = Li, Na, K (1)

of the corresponding metal hydride with diborane in etheral solvents under

ALKALI BASED TETRA BORIDES

temperature polymorph of LiBH4 is closely related to that of the

Cell parameters: a = 6.449 Å; b = 5.3070 Å; c = 5.2919Å; Atomic

Cell parameters: a = 5.109 Å; b = 5.109 Å; c = 5.109 Å; Atomic

At ambient pressure NaBH4 can exist in two structural modifications

Figure 3. Crystal structures of (a) -NaBH4, (partial H site occupancy is marked as

a = 10.6100 Å; b = 10.6100 Å; c = 10.6100 Å; Atomic positions: K(4a):

At least four polymorphs of RbBH4 are known in the literature. Recently

→ P4/nmm → C222 → I-42m; none of the three high-pressure structure types

Cell parameters: a = 5.6122 Å; b = 5.6122 Å; c = 4.0887 Å; Atomic

Cell parameters: a = 5.3679 Å; b = 5.1343 Å; c = 3.9098 Å; Atomic

ALKALI EARTH TETRA BORIDES

Beryllium borohydride, first synthesized in 1940 by Burg and

Among the Group I and II metal boro hydrides, magnesium borohydride

hundred articles. The experimentally observed phases of Mg(BH4)2 are

for identification of the correct symmetry. Solid α-Mg(BH4)2 contains an

Table 1. Experimentally observed polymorphs of Mg(BH4)2 and their

Phase Space Z Cell parameters, Cell volume, ρ, g Ref.

Phase Space Z Cell parameters, Cell volume, ρ, g Ref.

Cell parameters: a = 10.3540 Å; b = 10.3540 Å; c = 37.0550 Å; Atomic

Riktor et al. reported on the existence of several Ca(BH4)2 phases [76].

Cell parameters: a = 7.5250 Å; b = 13.1090 Å; c = 8.4030 Å; Atomic

0.8264, 0.2458, 0.0188; H2(8d):0.6461, 0.1556, 0.1648; H3(8d):0.7353,

Figure 7. Crystal structures of (a) o-Sr(BH4)2 [isostructural with O1-Ba(BH4)2], (b) t-

Cell parameters: a = 7.0930 Å; b = 7.0930 Å; c = 7.0930 Å; Atomic

ALKALI METAL MIXED TETRA BORIDES

Å (once). It is thought that these larger separations arise because of the

Cell parameters: a = 4.5711 Å; b = 15.4847 Å; c = 7.9355 Å; Atomic

Cell parameters: a = 4.6817 Å; b = 15.950 Å; c = 7.9438 Å; Atomic

framework in LiZn2(BH4)5 [111] and Li3MZn5(BH4)15 [112]; the first

0.26, 0.874; H1(4e):0.69, 0.63, 0.354; H2(4e):0.815, 0.64, 0.41;

ALKALI AND ALKALINE EARTH BASED MIXED

Metal borohydrides have been obtained by synthesis approaches that

K2Mg(BH4)4 is prepared mechano-chemically by ball-milling [113]. At

0.796, 0.192; H3(16i): 0.08,0.823,0.08; H4(16i): 0.146, 0.96, 0.16; H5(16i):

ISBN: H%RRN