Download as pdf or txt
Download as pdf or txt
You are on page 1of 65

Review

Cite This: Chem. Rev. 2018, 118, 10775−10839 pubs.acs.org/CR

Nanostructured Metal Hydrides for Hydrogen Storage


Andreas Schneemann,†,⊥ James L. White,†,⊥ ShinYoung Kang,‡ Sohee Jeong,§ Liwen F. Wan,‡
Eun Seon Cho,§,∥ Tae Wook Heo,‡ David Prendergast,§ Jeffrey J. Urban,§ Brandon C. Wood,‡
Mark D. Allendorf,† and Vitalie Stavila*,†

Sandia National Laboratories, Livermore, California 94551, United States

Lawrence Livermore National Laboratory, Livermore, California 94550, United States
§
Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States

Department of Chemical and Biomolecular Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

34141, Republic of Korea


Downloaded via UNIV NACIONAL DE SAN LUIS on August 21, 2021 at 22:42:49 (UTC).

ABSTRACT: Knowledge and foundational understanding of phenomena associated


with the behavior of materials at the nanoscale is one of the key scientific challenges
toward a sustainable energy future. Size reduction from bulk to the nanoscale leads to a
variety of exciting and anomalous phenomena due to enhanced surface-to-volume ratio,
reduced transport length, and tunable nanointerfaces. Nanostructured metal hydrides
are an important class of materials with significant potential for energy storage
applications. Hydrogen storage in nanoscale metal hydrides has been recognized as a
potentially transformative technology, and the field is now growing steadily due to the
ability to tune the material properties more independently and drastically compared to
those of their bulk counterparts. The numerous advantages of nanostructured metal
hydrides compared to bulk include improved reversibility, altered heats of hydrogen
absorption/desorption, nanointerfacial reaction pathways with faster rates, and new
surface states capable of activating chemical bonds. This review aims to summarize the
progress to date in the area of nanostructured metal hydrides and intends to understand
and explain the underpinnings of the innovative concepts and strategies developed over the past decade to tune the
thermodynamics and kinetics of hydrogen storage reactions. These recent achievements have the potential to propel further the
prospects of tuning the hydride properties at nanoscale, with several promising directions and strategies that could lead to the
next generation of solid-state materials for hydrogen storage applications.

CONTENTS 4.1. Experimental Structural Analyses 10789


4.1.1. Transmission Electron Microscopy 10789
1. Introduction 10776 4.1.2. X-ray and Neutron Techniques 10789
2. Classes of Nanostructured Metal Hydrides 10779 4.1.3. Vibrational Spectroscopy 10790
2.1. Bonding in Metal Hydrides 10779 4.2. Theoretical Structural Determination 10790
2.2. Nanostructured Binary Hydrides 10780 4.2.1. Wulff Construction 10790
2.3. Nanostructured Intermetallic Hydrides 10781 4.2.2. Prototype Electrostatic Ground State
2.4. Nanostructured Complex Metal Hydrides 10782 Method 10790
3. Synthetic Routes 10783 4.2.3. Microstructure Modeling 10791
3.1. Ball Milling 10783 4.3. Thermodynamic and Kinetic Analysis 10791
3.2. Gas-Phase Synthesis 10784 4.3.1. Thermodynamic Measurements 10791
3.2.1. Gas-Phase Condensation 10785 4.3.2. Thin Film Hydrogen Content Measure-
3.2.2. Plasma Deposition 10785 ments 10792
3.2.3. Thin Film Syntheses 10786 4.3.3. Rate Analysis and Kinetic Models 10793
3.3. Reductive Methods 10786 4.3.4. Thermodynamic Models 10794
3.3.1. Chemical Reduction 10786 5. Effects of Morphology on Hydrogen Storage
3.3.2. Electrochemical Reduction 10787 Properties 10795
3.3.3. Thermal Decomposition 10787 5.1. Free-Standing Nanoparticles 10795
3.4. Nanoconfinement 10787 5.2. Thin Films 10797
3.4.1. Host Materials 10788 5.3. Nanoconfined Metal Hydrides 10800
3.4.2. Impregnation Methods 10788
3.4.3. Synthesis Inside the Pores 10789
4. Methods to Analyze Structure and Storage Received: May 17, 2018
Properties 10789 Published: October 2, 2018

© 2018 American Chemical Society 10775 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

5.3.1. Confinement in Carbons 10800 world energy problems by delivering and storing energy using
5.3.2. Other Rigid Porous Hosts 10806 hydrogen. The Hydrogen Economy offers a potential solution
6. Mechanistic Effects of Nanosizing 10807 to satisfying global energy needs while reducing (with the
6.1. Effects of Surfaces 10811 ultimate goal of completely eliminating) carbon dioxide and
6.1.1. Effects on Reaction Enthalpy 10811 other greenhouse gas emissions and improving energy security.
6.1.2. Effects on Reaction Pathways 10812 Hydrogen storage is an essential component of a viable
6.1.3. Surface Entropy and Disorder 10812 hydrogen economy.4 Whether the application is stationary or
6.1.4. Chemical Kinetics at Defective Surfaces 10813 geared for the transportation sector, volumetric and
6.1.5. Chemical Impurities and Oxidation 10813 gravimetric efficiencies often dictate feasibility. The mobile
6.2. Effects of Internal Interfaces and Micro- vehicular application is one of the most challenging
structure 10814 applications, as it requires materials with very large gravimetric
6.2.1. Effects of Grain Boundaries 10815 and volumetric hydrogen capacities and also mandates other
6.2.2. Types of Interphase Boundaries 10815 stringent properties such as extended cycle-life, low impurity
6.2.3. Effects of Reactive Phase Boundaries 10817 levels, and fast kinetics of hydrogen uptake and release.
6.2.4. Effects of Nonreactive Phase Bounda- Hydrogen-powered fuel cell vehicles (HFCVs) are now
ries 10817 commercially available and store the fuel in fiber-reinforced
6.2.5. Nucleation and Phase Transformation tanks at 700 bar. Although sufficient to achieve reasonable
Mechanisms 10818 driving range (∼350−450 km), they still fall short of the
6.3. Other Size-Related Effects 10820 ultimate U.S. Department of Energy (DOE) capacity targets
6.3.1. Shortened Diffusion Pathways 10820 for on-board storage for light-duty vehicles (0.065 kg H2/kg
6.3.2. Altered Phase Coexistence Behavior 10820 system and 0.050 kg/L, Table 1). The resulting overall fuel
6.4. Metal Hydride-Host Interactions 10820
6.4.1. Physical Confinement Effects 10821 Table 1. Summary on the Targets for Hydrogen Storage
6.4.2. Electronic Coupling Effects 10821 Systems Set by the United States Department of Energy for
6.4.3. Chemical Interactions with the Host 10821 on-Board Vehicular Applications7
7. Conclusions 10822
storage parameter units 2020 2025 ultimate
Author Information 10823
Corresponding Author 10823 system gravimetric kWh kg−1 1.5 (4.5) 1.8 (5.5) 2.5 (6.5)
capacity: usable, (wt %)
ORCID 10823 specific-energy
Author Contributions 10823 from H2
Notes 10823 system volumetric kWh L−1 1.0 (0.03) 1.3 (0.04) 1.7 (0.05)
capacity: usable (kg H2 L−1)
Biographies 10823 energy density
Acknowledgments 10824 from H2
List of Abbreviations 10825 storage system cost $ kWh−1 net 10 (333) 9 (300) 8 (266)
References 10825 ($ kg−1 H2)
min/max delivery °C −40/85 −40/85 −40/85
temperature
cycle-life (uptake/ cycles 1500 1500 1500
1. INTRODUCTION release cycles)
The ever-growing fossil fuel-based economy has led to a minimum delivery bar (abs) 5 5 5
pressure
number of new challenges facing our civilization in the 21st
cycle life (1/4 tank cycles 1500 1500 1500
century, such as pollution and irreversible climate change due to full)
to excessive drilling, mining, and growing amounts of system fill time min 3−5 3−5 3−5
greenhouse gases. The dominance of fossil fuels as prevailing fuel purity (H2 % H2 99.97% 99.97% 99.97%
energy sources is clearly unsustainable in the long run because from storage)
the adverse impact of a single energy source is additive and
repetitive, and once the harmful consequences have accumu- cost is a significant additional concern, exceeding the target
lated beyond a critical threshold, permanent damage ensues.1,2 due to costs of the tank, compressor, and associated fueling
Diversification of energy sources and the development of station hardware. The ultimate hydrogen storage material
alternatives to fossil fuels such as hydrogen-powered and should have a high gravimetric capacity (≥10 wt % hydrogen),
electric vehicles are not only highly desirable, but some of have high reversibility (≥1500 cycles), and release hydrogen
them are currently technologically feasible and cost-compet- below 85 °C, which would allow the utilization of waste heat
itive. Advanced energy materials with improved properties are from a PEM fuel cell for desorption. The storage medium must
a prerequisite for most technological innovations and are also be “kinetically fast,” meaning that the material is capable of
mandatory to meet the increasing demands of a growing global releasing hydrogen at the demanded rate (up to 2 g H2 s−1)
population. As a consequence, it becomes apparent that and also can uptake hydrogen fast enough so the tank can be
alternative energy solutions and materials are critical for a refueled with hydrogen in the desired time (∼15 g H2 s−1).
sustainable future.1 Consequently, on-board storage remains a factor limiting the
Hydrogen is considered an attractive alternative fuel, or widespread adoption of HFCVs. Metal hydrides, which bind
more precisely an alternative energy carrier, particularly when hydrogen chemically, and sorbents, which store H2 through
used in hydrogen polymer electrolyte membrane (PEM) fuel physisorption, comprised of light elements could, in principle,
cells.3 The concept of an ecologically clean “Hydrogen meet the DOE storage targets, yet each of these two classes of
Economy” was first introduced in the mid-1970s and has materials has its own advantages and drawbacks. For instance,
been gaining momentum as a viable remedy for the growing materials such as 170 Mg(BH4)2 (14.9 wt %) and LiBH4 (18.5
10776 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 1. Processes accompanying hydrogen storage reactions in bulk (a) and nanostructured (c) metal hydrides. Despite differences in chemical
properties, the local bonding (b) can be similar.

Figure 2. Examples of nanostructured metal hydrides, including NPs, core−shell structures, nanowires, nanotubes, thin films, and multilayers: (a)
palladium NPs (adapted with permission from ref 25. Copyright 2009 Elsevier), (b) Mg@MgO core−shell NPs (adapted with permission from ref
26. Copyright 2014 Elsevier), (c) magnesium nanowires (adapted from ref 23. Copyright 2007 American Chemical Society), (d) palladium
nanotubes (adapted with permission from ref 27. Copyright 2000 Elsevier), (e) Pd-capped Mg thin film (adapted with permission from ref 28.
Copyright 2008 Elsevier), and (f) Mg/Cr multilayer film (adapted with permission from ref 29. Copyright 2014 Elsevier).

wt %) nominally can meet both the gravimetric and volumetric considered for improving these properties with varying degrees
targets.5 Unfortunately, other factors limit successful use of of success.5,8,9 These include entropic effects,10 altering the
hydrides as storage materials, and none currently can morphology of hydride particles,11 mechanical strain,12 and
simultaneously meet the DOE targets for minimum delivery dopants or other chemical additives.13 Although one cannot
pressure, charging/discharging rates, and capacity. Sorbents, in say that the potential of any of these has been exhaustively
contrast, have rapid adsorption/desorption rates and are evaluated, the improvements achieved so far are modest.
readily reversible, but their capacities are very low except at A promising approach that has garnered substantial
cryogenic temperatures, typically 77 K.6 attention in the past decade is nanostructuring of the metal
The reasons for the inadequate performance of any given hydrides. Many recent experiments and models that will be
hydride are complex, but in general these can be broken down described in this review suggest that nanostructured metal
into thermodynamic and kinetic factors. Considering thermo- hydrides have significantly, rather than incrementally, different
dynamics, hydrogen storage materials (including both sorbents thermodynamics and kinetics than their bulk counterparts. For
and light metal hydrides) fall on either side of an ideal, but example, a substantial decrease in the H2 desorption enthalpy
effectively empty, range of hydrogen binding energies, with is predicted for very small MgH2 particles.14 Similarly, force
sorbents binding H2 too weakly through physical adsorption field modeling predicts a decrease in thermodynamic stability
and hydrides binding it too strongly with chemical bonds. of MgH2 as particle size decreases from 2.0 to 0.6 nm.15
Metal hydrides suffer additionally from slow H2 uptake and Barriers for H2 desorption from NaAlH4 are also predicted and
release kinetics and limited reversibility, due in the case of experimentally found to depend on particle or cluster size.16−18
some complex hydrides to the intricate reaction pathways and Nanoconfinement of Li3N was shown to fundamentally alter
myriad intermediates involved. Many strategies have been both the hydrogenation and dehydrogenation reaction path-
10777 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

ways as a direct consequence of solid−solid nanointerfaces recesses into which the confined material is introduced, bound,
within the material.19 These are just a few examples to show and then restricted from movement or agglomeration with
that there is ample evidence for altered physical and chemical other particles. The pores, therefore, determine the shape and
properties of nanostructured metal hydrides compared to bulk dimensionality of the nanoscale material. Nanoencapsulation is
(Figure 1). defined here as incorporation of a nanoscale material inside a
Nanostructured materials have sizes intermediate between secondary material which is not necessarily porous, named as
molecules and microscopic structures. They can be a wide the matrix or shell (e.g., graphene,30−33 reduced graphene
variety of shapes and sizes (Figure 2), but they share the oxide,34−36 and polymers22,37−39). Encapsulation can also
characteristic of having at least one dimension be on the occur in core−shell NPs26,40,41 and multilayered thin
nanoscale, which is typically defined as being between 1 and films29,42 and involves a preformed nanostructure that acts as
100 nm (1 nm = 10−9 m). Nanoparticles (NPs), which have a barrier to particle/grain growth and serves as protection
three nanodimensions, are among the most common against disruptive conditions that could cause agglomeration or
nanostructures in materials science due to their relative ease pulverization upon cycling.
of synthesis, often through nucleation and restricted growth Theory and modeling are playing a critical role in elucidating
processes, display large fraction of surface atoms, which makes nanoscale phenomena associated with thermodynamics and
them attractive for numerous applications, particularly in kinetics of metal hydrides. The advances in computational
catalysis and energy-related applications.20 Some NPs are power and the emergence of supercomputers enabled quantum
composed of two different materials, either in a Janus structure, mechanical, empirical, mesoscale, and continuum approaches
in which each covers a portion of the exterior of the particle, or to modeling various phenomena in metal hydrides.43 Classical
a core−shell structure, with one of the materials completely molecular dynamics (MD) simulations are widely used to
encapsulating the other (typically designated as core material@ discover mechanisms and diffusion pathways, examining
shell material).21,22 Hollow nanotubes and solid nanorods moderate-to-large sized systems (i.e., ∼103 to 106 atoms).
(also called nanowhiskers, nanowires, or nanopillars, often Density functional theory (DFT) calculations are used to
depending on the aspect ratio of length to width) are elongated quantify energy barriers and end-state phase energies for
in one dimension.23 Two-dimensional materials, which have various phases and reactions. Combining MD and DFT
only one nanoscale dimension, include planar and sometimes calculations enable direct simulations of time-evolution (pico-
curved structures including nanosheets, nanoflakes, nanorib- to nanoseconds) to identify mobile species and energy barriers.
bons, and thin films.24 Some films that have been prepared Kinetic Monte Carlo simulations can cover much longer time
possess alternating layers of two or more discrete materials; scales (up to seconds), whereas continuum methods can, in
principle, cover time scales from seconds to hours.
even though these multilayers may be grown thicker than 1
Experimental confirmation of predictions is challenging for
μm, the discrete layers typically are individually less than 100
several reasons, although there is abundant evidence indicating
nm and thus remain classified as nanostructured. A further
that nanoscale metal hydrides have lower H2 desorption
distinction among 2D materials is that thin films are generally
temperature than bulk hydrides.24,44−49 First, the sizes
synthesized by deposition onto a flat (occasionally mono-
suggested by the calculations at which changes in properties
crystalline) substrate of another material, whereas nanosheets,
occur are much smaller than those produced by mechano-
nanoflakes, and nanoribbons are usually prepared independ- chemical approaches, which are fraught with problems,
ently of a synthetic foundation. including the presence of impurities, lack of size control, and
Nanoconf inement (Figure 3) is a broad term that involves poor reproducibility. Alternatively, bottom-up approaches to
either the formation of a nanostructured material inside a host nanoscaling, such as the formation of dendrimers, aerogels, and
(nanoscaffolding) or the encapsulation or coating of a material inorganic templates (e.g., zeolites), are unsuitable due either to
with a rigid matrix (nanoencapsulation). We define nano- the highly reactive nature of storage materials or the
scaf folding as confinement of a material inside of a scaffold with substantially detrimental impact on gravimetric capacity.44−49
permanent porosity. The host material has nanoscale pores or Second, tuning particle size can sometimes be difficult. Pores in
host materials are not always monodisperse, making it hard to
achieve uniform thermodynamic behavior and identify the
transition to nonbulk properties. For example, NaAlH4
particles supported on carbon nanofibers have broad size
ranges (e.g., 19−30 nm), resulting in hydrogen desorption
temperatures from 70 to 200 °C.50 Molten-metal infiltration of
carbon aerogels, which have a distribution of pore sizes, yields
2−5 nm Mg particles,51 but wetting agents (e.g., Ni) degrade
the scaffold, and blocking of the outermost pores appears to
limit infiltration. Third, the chemical environment surrounding
the particle may play an important role. For example, colloidal
stabilization of Mg particles yields 5 nm diameter particles that
desorb H2 as low as at 85 °C (>150 °C lower compared to
bulk).52
Figure 3. Schematic depicting the two major subclasses of In general, the factors leading to improved properties at the
nanoconfinement discussed in this review: nanoscaffolding (left), nanoscale are still poorly understood, however, and there are
with the hydride inside a host with permanent porosity, and numerous challenges to be resolved or overcome. Free NPs are
nanoencapsulation (right) with the hydride wrapped by a nonporous not stable, so some form of stabilization or hosting is required.
material, such as graphene. Unfortunately, nanoconfinement adds dead mass and volume
10778 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

that reduces storage capacity. From a synthetic aspect, some Metal hydrides in general, and nanostructured metal
modeling suggests that nanoscale effects do not present hydrides in particular, can be grouped into three large classes:
themselves until particles are one nanometer in size or even (1) binary hydrides MHx (M= main-group or transition metal,
smaller. Thin films lack capacity and the low-cost fabrication such as in LiH, MgH2, PdH0.6, TiH2); (2) intermetallic
techniques required for large-scale production. Core−shell hydrides, ABxHy (A is typically the hydriding metal and B is
architectures may avoid this problem but are still very much at the nonhydriding metal, such as TiFeHx, TiMn2Hx, and
the research stage. In nanoconfined materials, it is unclear LaNi5Hx); and (3) complex metal hydrides, MEHx (E = boron
which is the critical factor: surface-to-volume ratio, particle- (borohydrides, e.g., LiBH4, Mg(BH4)2); nitrogen (amides, e.g.,
host interaction energy, or cluster geometry (“magic number” LiNH2, Mg(NH2)2), aluminum (alanates, e.g., NaAlH4,
effects are observed when metals such as gold are nanoscaled,53 Mg(AlH4)2)). In binary ionic hydrides such as NaH and
for example), though some scaffolds may also have catalytic CaH2, hydrogen exists as the negatively charged hydride ion
effects independent of the size of the hydride.54 Carbon, (H−). In contrast, interstitial hydrides, such as palladium
probably the most economical nanoconfinement material, is hydride, are usually nonstoichiometric compounds, with
difficult to characterize in terms of its pore chemistry, and thus hydrogen existing as atoms dissolved in the lattice of host
it is problematic to control and synthesize reproducibly. metal atoms. Complex metal hydrides are compounds
Nevertheless, nanoscaling is a promising synthetic strategy composed of metal cations and polyatomic hydrogen-
that in recent years has attracted considerable atten- containing anions such as alanates (AlH4−), borohydrides
tion.48,49,55−58 Several review articles with a narrower focus (BH4−), and amides (NH2−). These compounds possess
were published in recent years.44,46−48,59−62 Given the renewed hydrogen atoms covalently bound to aluminum, boron, or
interest in hydrogen storage in general and metal hydrides in nitrogen, with the hydrogen atoms having either δ+ or δ−
particular (for storage as well as other applications), a more polarity.
current and general review is warranted. In this article, we
2.1. Bonding in Metal Hydrides
review new experimental and theoretical approaches to
synthesizing nanostructured metal hydrides and probing their Metal hydrides exhibit a great diversity of chemical bonds,
hydrogen storage properties. spanning from metallic bonding, as in interstitial hydrides such
This review aims to describe, summarize, and analyze as PdHx, ionic bonding in some binary and complex hydrides,
reports in the scientific literature on nanostructured metal and partial or full covalent bonding as in the lightest binaries
hydrides, both simple and complex, and their use for hydrogen (LiH and MgH2) and complex hydride anions (AlH4− and
storage, including their capacities and their thermodynamic BH4−). A detailed discussion of these is provided recently by
and kinetic properties in H2 uptake and release. Techniques for Jensen et al.74 However, to orient the reader we present a brief
making nanostructured metal hydrides will be presented, as summary here.
will computational methods for probing their properties. Extensive experimental thermodynamic data for metal
Previous review articles and book chapters on metal hydrides hydrides are available, typically including the enthalpy and
typically discuss nanostructuring only briefly and not as a entropy of H2 desorption. Heats of formation for binary
comprehensive study, and some have focused specifically on a hydrides of the alkali and alkaline earth metals and main-group
narrow class of nanostructured metal hy- elements have also been measured; for example, see Bourgeois
drides.8,24,44−46,55,56,58,60,61,63−69 In addition, the present re- et al. and references therein.75 Calorimetry and other
view covers the recently discovered novel classes of nano- experimental techniques have been used to estimate M-H
structured hydrogen storage materials, such as graphene- bond dissociation enthalpy (BDE) in main group and
encapsulated metal hydrides and core−shell nanostructures, transition metal hydrides; the latter have received considerable
which present a major leap forward in improving both the attention in this regard.76−81 For example, the “hydricity” of
thermodynamics and kinetics of metal hydrides for hydrogen transition metal complexes in solution, which is the energy
storage applications. required to form a hydride ion (MH → H− + M+), has been
measured in a number of cases;81,82 these data are useful for
2. CLASSES OF NANOSTRUCTURED METAL understanding chemical reactivity and bond-breaking events
HYDRIDES within catalytic cycles. Critical assessments of the literature and
Bulk metal hydrides have been known for more than two associated thermodynamic phase modeling (CALPHAD
centuries: Gay-Lussac and Thenard reported the synthesis of method) have provided phase diagrams for a number of
potassium hydride by reaction of potassium metal and gaseous hydrides,83 and these are a valuable tool for evaluating the
hydrogen upon heating in 1811.70 In 1874, Hautefeuille and reactivity of these materials.
Troost measured the equilibrium plateau pressures for the However, although thermodynamic data for H2 desorption
reactions of metallic lithium, sodium, and potassium with reactions are plentiful, they do not always constitute
hydrogen.71,72 However, bulk metal hydrides started to be dissociation energy (ΔH°BDE) for an individual M-H bond
considered as media for solid-state hydrogen storage in the but instead are the average bond energies relative to the
1960s, when several intermetallic hydrides (e.g., Mg2NiHx, condensed-phase products. Consequently, one must turn to
LaNi5Hx, and TiFeHx)73 were discovered and shown to absorb other approaches, in particular theory, to obtain detailed
and desorb hydrogen gas reversibly in a wide temperature insight into the bonding of metal hydrides. In addition to
range (i.e., −50 to 400 °C). While many intermetallic hydrides reaction energies, first-principles calculations can provide
can absorb substantial quantities of hydrogen on a molar basis, densities of states, atomic charges, and orbital compositions
most of these alloys, composed of heavy rare-earth or that reveal the underlying nature of the metal−hydrogen
transition metals, have a maximum storage of about 1−2% interaction. The most common approach is DFT, which has
hydrogen by weight, with the exception of Mg alloys, which been applied to virtually all classes of metal hydrides of interest
have a capacity up to 7.6 wt % hydrogen. for hydrogen storage75,84−92 and used to screen hydrides for
10779 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 4. Isothermal hydrogenation profiles of Pd octahedrons (red) and cubes (blue) at 303 K at a hydrogen pressure of 101.325 kPa (1.0 bar
H2). (b) Schematic potential energy diagrams of the Pd octahedrons (red) and cubes (blue). (Reproduced from ref 111. Copyright 2014 American
Chemical Society.)

specific applications.93−96 An important conclusion obtained (AlH4− and BH4−) is bound ionically to an alkali- or alkaline-
from these studies is that, for binary hydrides, the primary earth metal. The B−H and Al−H bonds have strong covalent
factors influencing M-H BDE are (1) the difference in the character,92 with X-H bond orders from 0.5 to 0.75 and similar
electronegativities of the metal and hydrogen79 and (2) the ionicities for the four hydrides. The predicted bond order, B−
cohesion energy of the metal relative to the energy cost for H ≈ N−H > Al−H, indicates that B−H and N−H are more
lattice expansion upon formation of the hydride. 75,92 covalent than Al−H. This is consistent with a later
Consequently, the stability of binary hydrides from alkaline investigation by Sholl and co-workers, who considered the
metal hydrides through the early transition metals (Sc and Y) bonding in Mg(BH4)2 and Mg(AlH4)2 and concluded that B−
is dominated by the charge transfer between the metal atom H bonds possess a significantly higher degree of covalency than
and hydrogen.75 In the case of complex metal hydrides, the Al−H bonds.108 The B−H BDE predicted by molecular cluster
bond energy of the M-H diatomic (M typically being B or Al) calculations is a high 384 kJ mol−1,86 which is consistent with
is also an important parameter that can be used to characterize the higher thermal stability of the borohydrides. These
the bonding.92 predictions are all consistent with what is known empirically
A few specific cases are illustrative. For two of the lightest about the stability of these hydrides.
hydrides, LiH and MgH2, extensive charge transfer leads to Overall, these results suggest that it may be difficult to
negative enthalpies of formation: −89.3 to 99.1 kJ mol−1 for influence the thermodynamics of metal hydrides by nano-
LiH97,98 and −76.2 to −77.0 kJ mol−1 for MgH2.99−104 The structuring alone (i.e., steric effects that may alter the
bonds in these hydrides have mixed ionic−covalent character, arrangement of ions), particularly for strongly ionic hydrides
whereas in hydrides of the heavier elements, such as CaH2 and for which Coulombic forces may be difficult to alter.
SrH2, the bonding is primarily ionic.74 Proceeding down the Alternatively, the kinetics of hydrogen desorption and uptake
periodic table, hydride stability begins to decrease beyond the could be dramatically affected. As seen subsequently in this
early transition metals as a result of the competition between article, however, such conclusions are overly simplistic and
metal cohesion energy and electronegativity difference. Miwa other factors, such as surface energy, interactions with “non-
and Fukumoto computed heats of formation for first-row innocent” pore walls, internal nanointerfaces, and mechanical
transition metal dihydrides and showed, consistent with strain can exert large effects that can change the behavior of
experimental results, that at Group 8 (Fe), both the metal these materials.
cohesion energy and the energetic penalty for the lattice
2.2. Nanostructured Binary Hydrides
expansion upon M-H bond formation reach a maximum.
Beyond this point, the dihydrides become unstable.105 A Most of the metals on the periodic table form stable binary
similar conclusion was reached in an earlier study by Smithson metal hydrides. When the surface of a metal, e.g., palladium, is
et al.88 Deeper into the periodic table, the bonding eventually exposed to hydrogen gas, the hydrogen molecules are initially
becomes metallic, with hydrogen located at interstitial sites and adsorbed then split into hydrogen atoms, which diffuse on the
allowing variable compositions (effectively, solid solutions) surface and into the bulk. The absorption and desorption of
without a phase change. In this case, bonding is much weaker. hydrogen by Pd occurs via cycling between two phases; at low
For example, experimental formation enthalpies for bulk PdHx hydrogen concentration, the α-Pd phase is stable, whereas at
(x = 0.25, 0.75, and 0.8) are only −17.7 to 19.5 kJ mol−1.106,107 higher hydrogen concentrations the β-Pd phase is the
It should be noted that DFT calculations consistently dominant species. The effect of size on hydrogen storage
underestimate hydride formation enthalpies; the disagreement properties has been known for a while, and cycling under
with experiment is greatest for alkaline and alkaline earth hydrogen gas near the equilibrium pressures needed for the
hydrides (as much as 20 kJ mol−1 for KH).75 Nevertheless, formation of Pd hydrides was reported to change with particle
correct trends are predicted, even for complex transition metal size.107,109,110 Recent work by Li et al. has demonstrated that
hydrides.95 the morphology of Pd may play a critical role in the absorption
The picture does not change dramatically for complex main rates of hydrogen, and that temperature as well is critical for its
group metal hydrides. Yoshino et al. performed a DFT study uptake, absorption, and diffusion. Figure 4 shows the
for the hydrides LiBH4, LiAlH4, LiNH2, and NaAlH4.92 Their isothermal hydrogenation profiles of Pd octahedrons (red)
analysis indicates that the factors controlling the bonding and cubes (blue) at 303 K, under a hydrogen pressure of
trends in these hydrides are similar to those for metal 101.325 kPa (1.0 bar H2).111 In situ electron energy-loss
dihydrides, with the added factor that the hydride anion spectroscopy (EELS) on individual palladium nanocrystals
10780 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

revealed that palladium nanocrystals undergo dramatic changes ionically, are not of substantial interest for hydrogen storage.
during the phase transition between the α and β phase, and While some, such as NaH, CaH2, and LiH, have fairly high
surface effects dictate the size dependence of hydrogen gravimetric hydrogen capacities (4.2, 4.8, and 12.5 wt %,
absorption pressures.112 respectively), especially compared to the heavy metals, the
Among bulk noble metals, palladium is the only metal that strength of the ionic bonds and the high energetic costs
can absorb and desorb hydrogen at near-ambient conditions of associated with forming metallic sodium, calcium, or lithium
temperature and pressure. Downsizing metal particles to a few lead to desorption temperatures above 400 °C, significantly
nanometers changes the ratio of surface to bulk atoms and can greater than the desired release temperatures for near room
introduce fundamental changes in the thermodynamics of temperature applicaitons.5 However, these and other metal
hydrogen uptake. For example, bulk Ir and Rh form hydride hydrides are of significant interest for the development of high-
phases only at impractical H2 pressures in excess of 4 GPa. temperature thermochemical energy storage systems, for
Nanoscaling Ir and Rh to <10 nm particles results in hydrogen instance, as storage media for concentrated solar thermal
being absorbed near or at room temperature to form the applications.122−129
corresponding hydride phases.113−117 These results indicate 2.3. Nanostructured Intermetallic Hydrides
that the thermodynamics of the Rh−H and Ir−H hydride
systems are highly dependent on the particle size and, at the Intermetallic hydrides have found many applications, which in
smallest size tested (1 to 2 nm), can result in 7 to 8 orders of addition to hydrogen storage systems include nickel metal
magnitude decreases in the equilibrium plateau pressure hydride (NiMH) battery electrodes, hydrogen purification
required to form the hydride phase. For example, bulk Rh systems, hydrogen sensors and catalysts, heat pumps, and
absorbs hydrogen at 4 × 109 Pa, whereas 2.3 nm particles of cooling systems.68 Although metal hydrides composed of two
Rh have a midplateau pressure of only 30 Pa. or more metals store relatively low quantities of hydrogen per
Magnesium hydride has been extensively studied as a unit mass, the kinetics of hydrogen absorption and desorption
convenient model system to explore the effect of nano- are relatively fast in those of them that store the hydrogen in
structuring on the thermodynamics and kinetics of hydrogen interstitial sites in the metal lattice. Most interstitial metal
storage reactions. Although many different synthetic ap- hydrides are intermetallic metal hydrides of general formula
proaches to Mg nanostructures have been proposed, synthesis AmBnHx. The properties of these materials are largely
of pure Mg NPs is complicated by the high reactivity of Mg. determined by the interaction between the interstitial hydro-
Polymer wrapping was proposed as an efficient way to reduce gen atoms and the metal atoms. However, not all intermetallic
the reactivity; for example, 5 nm Mg NPs coated with polymers hydrides store hydrogen interstitially. One notable exception is
were produced in solution by Jeon et al.21 The authors used Mg2NiH4, which is not an interstitial hydride but rather a
poly(methyl methacrylate) (PMMA) as a capping ligand, with complex metal hydride with NiH44− ions; it does, however,
bis(cyclopentadienyl)magnesium (Cp2Mg) as the magnesium form the intermetallic compound Mg2Ni upon dehydrogen-
source. During the reduction of the organometallic precursor ation.130,131 The most important stoichiometries of interme-
by lithium naphthalide, burst-nucleation and growth processes tallic metal hydrides are AB5 (CaCu5 structural type), AB2
occur and PMMA-coated Mg nanocrystals are produced. The (Laves phase), AB (CsCl structural type), and A2B (AlB2
valuable finding in this study is that PMMA acts not only structural type). Less common intermetallic hydrides include
merely as a capping ligand but also as a gas-selective barrier. A2B17, A6B23, A2B7, AB3, and A3B stoichiometries.132 The first
Generally, Mg metal readily forms MgO and Mg(OH)2 layers intermetallic hydrides were reported in the early 1960s;
upon exposure to air or water, which consequently decreases currently, several thousands of these materials are known. A
the hydrogen storage ability because those oxides are inactive rather comprehensive list of intermetallic metal hydrides is
materials for storage and prevent the penetration of H atoms to available as part of the Hydrogen Storage Materials Database,
the remaining metal. PMMA prevented the oxidation of Mg http://hydrogenmaterialssearch.govtools.us/.133
due to the higher H2/O2 permeability ratio of PMMA Nanostructuring has been widely used to improve the
compared to those of other polymers such as poly- properties of intermetallic hydrides. For instance, the structural
(dimethylsiloxane) and polycarbonate, which resulted in and hydrogen storage properties of ball-milled nanostructured
enhanced air stability of PMMA-coated Mg NPs even after 3 TiFe0.5Ni0.5/graphite composite are significantly altered
days of air exposure. compared to bulk. 134 Both bulk and nanostructured
Other synthetic methods such as mechanochemistry,118 gas- TiFe0.5Ni0.5 materials crystallize in a cubic CsCl-type structure.
phase condensation,119 and plasma-metal reactions120,121 (see However, after hydrogenation, the formation of the hydrogen-
section 4) generally produce nano-MgH2 with a larger particle poor β phase is observed for the bulk compound, whereas the
size distribution. In addition to transmission electron nanostructured TiFe0.5Ni0.5/graphite composite forms the
microscopy (TEM), powder X-ray diffraction (XRD), both hydrogen-rich γ phase with improved reversible hydrogen
lab- and synchrotron-based, is often used to determine the capacity. Mössbauer spectroscopy on Fe centers indicate
crystallite size in nano-MgH2 and shows significant broadening significant changes in the isomer shift toward the γ phase,
for particles <50 nm. Dopants and additives are sometimes which indicates a decrease of the s-electron density at the Fe
incorporated into nano-MgH2 to enhance the kinetics; these nuclei due to the charge transfer from the metal to the nearby
are typically transition metals or transition metal compounds. hydrogen atoms.134
It must be noted that the kinetics of MgH2 NPs and thin films TEM investigation of nanostructured Mg-10Ni-2Mm (Mm
are improved even at particle/grain sizes of 50 nm or larger. In = 99.7% lanthanum-rich mischmetal) showed a significant
contrast, thermodynamic improvements were only observed reduction in grain size during ball-milling. Additionally,
for particle/cluster sizes of <10 nm (section 5). nucleation of intermetallic MmMg12 at the grain boundaries
Several other binary metal hydrides, particularly those of the of Mg and Mg2Ni occurred. The interface between MmMg12
alkali and alkaline-earth metals, which bind hydride more and Mg2Ni was semicoherent and features an ordered
10781 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

repetition of the consistent atomic arrangements.135 The perspective on other promising materials and strategies. Figure
kinetics of H absorption and desorption are improved upon 5 shows a schematic depiction of hydride confinement in three
ball-milling due to the fast hydrogen diffusion in the
nanograins, thus providing paths for fast H-exchange.
Pressure−composition−temperature (PCT) diagrams show
the presence of two plateaus, one each for the Mg + H2 ⇌
MgH2 and Mg2Ni + 2H2 ⇌ Mg2NiH4 reactions. The MgH2
plateau showed no hysteresis and practically no slope, while
the plateau for Mg2NiH4 exhibited both a pronounced
hysteresis and a slope, particularly for the nanocrystalline
sample.135 Other examples of microstructure refinement in
intermetallic hydrides can be found in a review by Varin et al.57
2.4. Nanostructured Complex Metal Hydrides
Complex metal hydrides consist of (predominantly alkaline
and alkaline earth) metal cations and “complex” hydrogen-
containing anions, for instance alanates (AlH 4 −) and Figure 5. Schematic representation of hydride confinement in three
borohydrides (BH4−). The appeal of this class of hydrides classes of materials, namely, metal−organic frameworks (left), carbon
arises from their gigantic gravimetric hydrogen content, due to aerogels (center), and mesoporous carbons (right).
their composition of mostly light elements (e.g., LiBH4 and
Mg(BH4)2 contain 18.4 and 14.9 wt % H2, respectively).
However, some drawbacks of these materials are the very classes of materials, namely, carbon aerogels, mesoporous
endothermic hydrogen release and the lack of reversibility. carbons, and metal−organic frameworks. Apart from these,
Strategies are being successfully developed to overcome these other porous materials commonly used are carbon nanofibers
limitations, for example by catalyst doping or nanosizing and and porous oxides.
confinement. A comprehensive review article by Orimo et al.5 Lithium aluminum hydride (LiAlH4) has a high hydrogen
discusses the general feasibility of complex hydrides for content of 10.6 wt %. The material releases H2 in a three-step
hydrogen storage application in great detail. Within the next process, via the intermediates Li3AlH6 and LiH. The three
few paragraphs, we will summarize the properties of some reaction steps take place at (1) 160−180 °C, (2) 180−225 °C,
prototypical complex hydrides, namely LiBH4, LiAlH4, Mg- and (3) above 400 °C, with enthalpies of −10, +25, and +140
(BH4)2, and NaAlH4. kJ mol−1 H2, respectively.139 The last step of the reaction is
Table 2 summarizes some of the important dehydrogenation energetically unfeasible, lowering the effective capacity to only
properties of bulk complex metal hydride materials. Even about 7 wt %. Additionally, it was calculated by Jang et al. that
the rehydrogenation process to form LiAlH4 from Li3AlH6
requires high H2 pressures, above 1000 bar.140,141 However,
Table 2. Summary of Some Critical Values That Describe
due to the excellent desorption properties of the bulk material,
the Dehydrogenation Properties of the Discussed Hydrides
a lot of work is still invested into preparing nanoscale or
H2 ΔH nanoconfined LiAlH4.
hydride wt % Ea (kJ mol−1) Tdec (°C) (kJ mol−1 H2) Sodium aluminum hydride, also called sodium alanate, has a
LiAlH4 10.6 82−115; 160; 180; −10; 25; 140b slightly lower hydrogen capacity than LiAlH4, amounting to 7.4
86−90a 400b
wt % H. NaAlH4 decomposes to release hydrogen in three
NaAlH4 7.4 118, 120a 180; 190; 40.9; 15.6; 120b
400b steps according to the following reactions:
LiBH4 18.4 146a 483−492b 74d 3NaAlH4 F Na3AlH6 + 2Al + 3H 2
Mg(BH4)2 14.9 310.7; 160.9c 320b 53b
a
Data from Christian and Aguey-Zinsou.136 bData from Klebanoff et Tdec = 180 − 190 °C; ΔH = 40.9 kJ mol−1 H 2 (1)
al.137 cData from Fichtner.66 dData from Mauron et al.138
Na3AlH6 F 3NaH + Al + H 2

though they feature very high gravimetric hydrogen contents, Tdec = 190 − 225 °C; ΔH = 15.6 kJ mol−1 H 2 (2)
their activation energies and decomposition temperatures
(Tdec) are unfeasible for practical use. Downsizing of the NaH F Na + 1/2H 2
hydrides through nanostructuring can reduce the enthalpy or Tdec ≥ 400 °C; ΔH = 120 kJ mol−1 H 2 (3)
activation energy needed to desorb hydrogen, in some cases
rather dramatically, in part because the diffusion limitations The last step occurs at very high temperatures (Tdec ≥ 400
within bulky highly crystalline, covalent networks are over- °C), making it infeasible for practical use. A fundamental
come,45 though a number of other reasons have been proposed breakthrough came with the discovery by Bogdanović and
and will be discussed in section 6. Schwickardi that doping sodium aluminum hydride with
The confinement of hydrides in porous materials, e.g., titanium not only improves the kinetics of dehydrogenation
aerogels, porous carbons, or metal−organic frameworks, not reaction but also makes the reaction reversible.13 Since then,
only allows the stabilization of NPs and hinders their the decomposition and reabsorption of hydrogen in the
recombination and agglomeration but also forces interactions NaAlH4 system has been one of the most intensively
between the pore walls and the hydride particles. The effect of investigated reactions in the field of solid hydrogen storage
confinement of four light-metal hydrides, LiAlH4, NaAlH4, materials.142−149 Much of this research has been devoted to the
LiBH4, and Mg(BH4)2, is described, along with a short confinement of this complex hydride in different host matrices,
10782 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 6. Reduction of (a) the crystallite size and (b) the breadth of the particle size distribution as a function of ball milling time of LiH.
(Reproduced with permission from ref 183. Copyright 2008 Elsevier.)

with some interesting effects which are discussed in the reducing the particle size, ball-milling is a versatile technique to
following section. introduce dopants and catalysts.159−168 Plasma processing
LiBH4 attracts substantial research as a hydrogen storage techniques have been also applied to synthesize small hydride
material, as it has one of the highest gravimetric hydrogen NPs by using both hot and cold plasma techniques, but these
content of any material, 18.4 wt %. Differential thermal analysis synthetic methods are less general and generally produce
(DTA) studies suggest very high dehydrogenation temper- samples that are inhomogeneous, poorly crystalline, and with
atures starting at 380 °C, with the majority of H2 released ill-defined particle size.169 Bottom-up synthetic approaches
above 420 °C.150−152 The dehydrogenation process is not yet were developed in recent years to overcome these limitations.
fully understood, and different proposed reaction routes These approaches include gas-phase condensation,23 wet-
include intermediates such as LiB3H8153 and Li2B12H12.154,155 chemical synthesis,52 and nanoconfinement18 inside porous
Hence, substantial research has been devoted into lowering the scaffolds through liquid-phase infiltration. Although these
dehydrogenation temperature by downscaling of the crystallite methods tend to offer better control over particle size, the
size and embedding the material into nanoscaffolds. homogeneity and solvent contamination can be an issue.
Mg(BH4)2 has exceptional gravimetric hydrogen content of 3.1. Ball Milling
14.9 wt %. The desorption onset of this material is very high,
and hydrogen release is usually observed at temperatures above One of the most-used techniques to reduce the particle size in
320 °C. The release of hydrogen most likely occurs in a 3-step metal hydrides is ball milling, which involves the mechanical
reaction cascade: grinding of a solid sample with one or more inert and hard
balls (e.g., ceramic, flint, tungsten carbide, or stainless steel)
6Mg(BH4)2 F 5MgH2 + Mg(B12H12) + 13H 2 (4) moving at a high speed and crushing the powder. Under
certain regimes, the resulting materials are mostly composed of
5MgH2 F 5Mg + 5H 2 (5) small NPs or nanocrystallites.170−180 The “NPs” derived from
mechanochemical processes are different compared to particles
5Mg + Mg(B12H12) F 6MgB2 + 6H 2 (6) formed via solvent-based chemistry approaches and tend to
contain gross internal grain boundaries, fractures, dislocations,
It must be noted that other reaction pathways have been
or internal disorder, whether the crystals they contain are
proposed and experimentally determined for bulk and
nanocrystalline or not. For brittle materials, particle fracture is
nanoscale Mg(BH4)2, as summarized in a recent review article
well-described by Griffith theory, which shows the relationship
by Zavorotynska et al.156 The reversible hydrogen uptake by
between applied nominal stress and crack length at fracture.
MgB2 was demonstrated by Severa et al. under 950 bar H2
The stress, σF, at which crack propagation leading to particle
pressure at 400 °C.157 At the current research stage, this
fracture occurs is approximated by the following equation:
material seems still very attractive, if the hydrogen absorption
and desorption properties can be enhanced. γE
σF ≈
3. SYNTHETIC ROUTES c (7)

The synthetic approaches toward nanostructured metal where c is the length of the crack (in m), E is the modulus of
hydrides can be divided into top-down and bottom-up elasticity (in N/m2), and γ is the surface energy of the milled
approaches. The top-down approach involves breaking down material (in J/m2).181 In ductile materials, a plastic zone
bulk metal hydride particles until NPs or nanocrystallites are accompanies the tip of the crack, which leads to the energy
formed, whereas bottom-up approaches involve building up dissipation as heat. Therefore, additional energy is needed for
NPs from atoms, molecules, clusters, and other small structural crack growth in ductile materials as compared to brittle
blocks. It is important to distinguish between methods that materials. In practice, as particles decrease in size upon ball
create isolated NPs and methods which generate microscopic milling, a certain limit is reached beyond which no reduction in
powders composed of nanograins or nanocrystallites. Ball- size occurs.79,182,103 Typically, the grain size of metal hydrides
milling has been extensively explored for nanosizing and decreases from a few microns to a few nanometers (Figure 6).
nanostructuring, due to its relative simplicity and general Ç akmak et al. showed that mechanical milling of Mg with
applicability of the process.118,158 Due to air- and water- 10% Ti yields large Mg agglomerates up to 100 μm in size,
sensitivity of metals and metal hydrides, the milling is typically with embedded Ti fragments of micron and submicron size
performed under inert gas or under hydrogen. In addition to distributed within the agglomerates.184 The Mg agglomerates
10783 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

are made of coherently diffracting crystallites, which can be as milling partially hydrolyzed MgH2 for 25 h reduces the
small as 26 nm after 30 h of milling. Pang et al. developed a activation energy of hydrogen desorption from 217 to 140 kJ
mechanical-force-driven physical vapor deposition (PVD) mol−1 due to removal of the oxide/hydroxide layer from the
approach for the synthesis of 1D complex hydrides in which surface of MgH2 particles.190
Mg(AlH4)2 nanorods with diameters of 20−40 nm and 10−40 Vijay et al. employed reactive mechanical milling to
LiBH4 nanobelts were formed.185 The procedure involves synthesize Mg−Ti−Fe nanocomposites, which can be hydro-
high-energy ball milling of ether precursors [Mg- genated to form MgH2.191 The hydrogenation reaction starts at
(AlH 4 ) 2 (Et 2 O)] n (where Et 2 O = diethyl ether) and 80 °C, and the material absorbs 4 wt % H2 and desorbs 3.8 wt
[LiBH4(MTBE)]n (MTBE = methyl tert-butyl ether) to form % H2 at 300 °C. Wang et al. produced Mg−Ce alloys by
the corresponding nanostructures (Figure 7). The final melting, which were further milled under hydrogen and finally
mixed with 10 nm Ni NPs.192 This treatment results in a
CeMg12 intermetallic, which can absorb 2.9 wt % H2 at 120 °C
within 30 min, and can desorb hydrogen as low as 180 °C. The
low desorption temperature of 180 °C reported for this
composite is puzzling, as the PCT results for the Mg−Ce/
nano-Ni composite indicate a plateau pressure of 1 atm
hydrogen at 280 °C. Milling the Mg/Mg2Ni0.8Cr0.2 with 40 nm
TiO2 NPs yields a material reported to desorb up to 4.2 wt %
H2 at 240 °C.193
Ball-milling complex metal hydrides is less efficient in
reducing the particle size, compared to doing so for binary and
intermetallic hydrides. Nevertheless, high-energy milling of
LiAlH4 was reported to produce grain sizes below 100
nm.194,195 For high-energy ball-milled LiAlH4, the activation
energy for hydrogen release to form Li3AlH6 and LiH is
reduced from 111 to 100 kJ mol−1.196 In addition, ball milling
is a convenient technique to introduce additives in complex
metal hydrides. For instance, ball milling was used to introduce
metal and metal halides additives to LiAlH4 and LiBH4,
significantly reducing the activation energy for hydrogen
release.163,195−201
However, despite significant progress, ball milling has
important limitations, such as yielding nonuniform particle
sizes and shapes as well as issues with contamination. The large
amount of strain developed in metal hydride particles during
the milling process makes it virtually impossible to obtain
defect-free crystals via this method. In addition, metal and
metal hydride ball milling is believed to create highly reactive
surfaces that readily react with oxygen, even residual oxygen
present on the surfaces of the milling balls and vessel. Although
Figure 7. (a) Mg(AlH4)2 nanorod synthesis procedure. SEM images
the exact role of such oxide layers is not very well-understood,
of (b) the Mg(AlH4)2 etherate precursor, (c) the as-deposited
Mg(AlH4)2 etherate precursor, and (d) as-synthesized Mg(AlH4)2. excessive oxidation can cause capacity loss and create issues
The scale bar in panel b of 2 μm; the scale bar in panels c and d is 200 with hydrogen diffusion. On the other hand, the resulting high
nm. (e) HRTEM image of the LiBH4 nanobelts. Scale bar: 10 nm. (f) density of grain boundaries can, to some extent, facilitate
Selected-area electron diffraction pattern of region in panel e. hydrogen diffusion and thus improve the overall kinetics.
(Adapted with permission from ref 185. Copyright 2014 Springer Finally, even though the hydride particles are in the desired
Nature.) nanoscale regime initially, the material can undergo significant
recrystallization and particle growth upon heating during
Mg(AlH4)2 nanorods and LiBH4 nanobelts are generated by hydrogen absorption and desorption, eventually leading to
heat-treating the deposited material under vacuum. The 1D bulk-like behavior.
Mg(AlH4)2 and LiBH4 nanostructures show superior hydrogen
3.2. Gas-Phase Synthesis
storage properties compared to bulk counterparts.
Ball milling of MgCl2 and LiH mixtures results in MgH2 NPs Gas-phase synthesis is a bottom-up method for producing
as small as 7 nm in size embedded in a LiCl matrix. In metal and metal hydride NPs from individual atoms or
comparison to bulk MgH2, the mechanochemically produced molecules. This method allows synthesis of complex,
MgH2 with the smallest particle size showed a 2.8 kJ mol−1 H2 sophisticated NPs, owing to the fast kinetics and non-
decrease in the decomposition reaction enthalpy. However, equilibrium processes it entails.202,203 Thus, control over the
only a small reduction in the desorption temperature was composition and morphology can be achieved by growth
observed, as the decrease in ΔS partially counteracted the parameters such as pressure, cooling rate, NP drift velocity,
effect from the change in ΔH.186 Ball milling is efficient in and aggregation zone length. Postsynthetic techniques
reducing the activation energy of hydrogen absorption/ employed after NP growth, such as in situ mass filtration
desorption from MgH2.174,187−189 However, this improvement and deposition temperature, offer additional possibilities for
is not always due to the small particle size. For instance, ball the synthesis of tailored NPs. In the following sections, we
10784 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 8. (a) TEM image of Mg-TiH1.971 composite. HR-TEM images of (b) Mg particle and (c) TiH1.971 particle. (Adapted with permission from
ref 212. Copyright 2014 Elsevier.)

cover gas-phase condensation and plasma deposition techni- variation of the plasma processing techniques.209 The method
ques for the synthesis of metal hydride NPs. was initially developed by Ohno and Uda to produce metal
3.2.1. Gas-Phase Condensation. Gas-phase condensa- NPs in the presence of hydrogen and argon gas but was later
tion is a variety of gas-phase synthesis that involves rapid adapted to create metal hydride NPs.
cooling of metal vapor to produce NPs. For instance, Mg NPs decorated with nickel were isolated by arc plasma
magnesium nanocrystals of different sizes can be synthesized evaporation of pure Mg followed by the electroless plating of
by condensation under inert gas and then the surface Ni on the Mg particles in a NiCl2 solution.41 The PCT
decorated with metallic nickel.204 Synchrotron radiation measurements indicate a hydrogenation enthalpy of the Mg−
powder X-ray diffraction and Sieverts measurements reveal Ni composite of −73.6 kJ mol−1 H2, similar to bulk Mg. Core−
that Ni significantly improves the hydrogen release and uptake shell structured binary Mg@Ti and ternary Mg@Ti@Ni core−
kinetics of the NPs. The enhancement of hydrogen sorption shell nanocomposites were synthesized using an arc plasma
properties is believed to be due to the formation of Mg2Ni and method followed by electroless plating in solutions.41 The
Mg2NiH4 phases in the nanocrystals.204 A similar approach was hydrogenated composites with core−shell structures contain a
used to isolate Mg−Pd composite NPs by inert gas MgH2 core and Ti or Mg−Ni hydride shells. Based on PCT
condensation of Mg vapors followed by vacuum evaporation measurements, the hydrogen absorption enthalpy (−67.1 kJ
of Pd clusters. Irreversible formation of the Mg 6 Pd mol−1 H2) of the ternary composite was slightly higher than
intermetallic phase takes place upon vacuum annealing, that of the binary composite (−77.2 kJ mol−1 H2). In addition,
resulting in Mg/Mg6Pd composite NPs. At 573 K and 1 both the binary and ternary hydride display lower activation
MPa hydrogen pressure, the metal-hydride transition leads to energies of H2 absorption than bulk MgH2.
the formation of Mg3Pd and Mg5Pd2, whereas at 5 MPa H2 The hydrogen plasma metal reaction (HPMR) method was
pressure the Pd-richer MgPd intermetallic is obtained.205 employed to synthesize various Mg-based nanocomposites.
Mg−Ti nanostructured samples were prepared via compac- Shao et al. reported hexagonal 50−500 nm Mg particles with
tion of NPs grown by inert gas condensation. The formation of slightly improved hydrogen storage properties.210 By varying
a solid solution of Ti in hexagonally close-packed (hcp) Mg the HPMR reaction conditions, Kooi et al. were able to further
was detected up to 15 at. % Ti, while after in situ hydrogen reduce the Mg particle size to between 10 and 80 nm, leading
absorption, the alloy phase separates into MgH2 and TiH2. to a more significant improvement in kinetics of hydrogen
When Ti content reaches 22 at. %, a metastable Mg−Ti−H desorption.211 The HPMR method was used to generate a
face-centered cubic (fcc) phase was observed after hydrogen Mg−Ti−H nanocomposite composed of 40−180 nm
absorption. Although the enthalpy of hydride formation is hexagonal Mg NPs, with 13 nm spherical TiH1.971 NPs
unchanged compared to bulk MgH2, the addition of Ti uniformly dispersed on the surface of Mg particles (Figure
enhances the kinetics with hydrogen absorption/desorption 8).212 The presence of TiH1.971 on the surface seems to limit
times of only several minutes at 300 °C.206 the sintering upon cycling. Addition of TiF3 to MgH2 NPs
Nanostructured metal hydrides of compositions Ti−H, Zr− synthesized through the HPMR method showed a beneficial
H, Ti−N−H, Zr−N−H, and Ti−Zr−N-H were synthesized effect on kinetics and hydrogen absorption capacity at low
using combustion wave under hydrogen. The self-propagating temperatures but was found to have no effect on H2
high-temperature reaction produced hydride nanograins as desorption.120 Other nanostructured metal hydrides synthe-
small as 50 nm.207 They exhibited improved hydrogen storage sized using the HPMR method include Zr55V29Fe16Hx,213
characteristics, including faster kinetics and longer cycle-life. Mg2NiH4,214−217 Mg2CoH5,214 Mg2FeH6,214 Mg2CuHx,218
3.2.2. Plasma Deposition. Hot and cold plasma TiFeHx,219 Mg−Al−H,220 Mg−Zn−H,221 and several Mg−
processing techniques169,208 are suitable for both metal and La-TM-H (TM = Ti, Fe, and Ni)222 nanocomposites.
metal hydride nanoparticle synthesis, because plasma provides Zhang et al. developed a similar method called acetylene
distinctive advantages, including high enthalpy, fast response, plasma metal reaction for the synthesis of 15−85 nm Mg
and high chemical reactivity. The arc plasma method for NPs.121 By using acetylene, the growth of large Mg particles is
nanoparticle synthesis achieves supersaturation of NPs by limited by the formation of carbon on the metal surface from
vaporizing metals and metals by charging electrodes until the the decomposition of acetylene. Some control over the Mg
breakdown voltage is reached. The arc of spark formed across particle size can be achieved by varying the amount of
electrodes vaporizes the metal, which is transported to the tail acetylene in the plasma, with a clear decreasing trend with
of the arc where the vapors fall into a highly supersaturated increasing acetylene content. Figure 9 shows the samples
state, which achieves effective formation of NPs by nucleation prepared by acetylene plasma metal reaction with 28%, 21.7%,
and condensation. Hydrogen plasma metal reaction is a 14.3%, and 5.3% concentrations of acetylene, respectively. The
10785 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

3.2.3. Thin Film Syntheses. The vast majority of thin


films for hydrogen storage are prepared starting with the pure
metal, often in ultrahigh vacuum (UHV) conditions. Physical
vapor deposition (PVD) is the most common technique.
Different subcategories of PVD include vacuum thermal
evaporation,225 molecular beam epitaxy (MBE),226,227 elec-
tron-beam deposition,28,228,229 ion-beam sputtering (IBS),230
and magnetron sputtering.42,231−233 These methods are highly
conducive to making films composed of more than one metal
by adding additional sources. Co-deposited metals can be in
intimate contact with each other even when they are
immiscible on the bulk scale, and alloys and intermetallic
compounds can be formed by annealing miscible metals.
Layered films are also quite feasible, particularly with
immiscible metals, though the formation of intermetallics can
be limited even between alloying metals by avoiding or limiting
annealing, leading to distinct layers with minimal blending.
Wet chemical methods have also been employed, using
preformed NPs and casting them into a thin film.234 Metal
hydrides are too sensitive for the ablative techniques described
above, and they would decompose into hydrogen and the other
constituent elements instead. Therefore, hydrogen is intro-
duced to the metallic films after they are formed.
3.3. Reductive Methods
3.3.1. Chemical Reduction. Chemical reduction with
hydrogen or other reducing agents has been widely used to
synthesize NPs of inorganic materials.235−240 However, many
of the surfactant-based approaches developed for other
inorganic nanoclusters of NPs are not suitable for synthesizing
nanostructured metal hydrides due to their high reactivity with
various classes of organic molecules, including thiols, alcohols,
and amines. Organic ethers and hydrocarbons are usually the
Figure 9. TEM images of Mg NPs prepared by the acetylene plasma solvents/surfactants of choice, as metals and metal hydrides are
reaction with acetylene content of (a) 28%, (b) 21.7%, (c) 14.3%, (d) relatively stable in these compounds. For instance, Mg
5.3%, and (e) 0%, and (f) size distribution of Mg NPs synthesized nanocrystals of controllable sizes can be prepared by chemical
with 21.7% acetylene. (Reproduced with permission from ref 121. reduction of bis(cyclopentadienyl)magnesium using a solution
Copyright 2010 Elsevier.) of potassium with an aromatic hydrocarbon in glyme.
Depending on the aryl anion used, Mg nanocrystals of 25
optimized acetylene concentration of 21.7% yields Mg NPs nm (potassium biphenyl), 32 nm (potassium phenanthrene),
with particle sizes from 15 to 85 nm with an average size of or 38 nm (potassium naphthalene) were isolated. The
about 40 nm. The equilibrium plateau measurements reveal hydrogen sorption kinetics were shown to be dramatically
anomalously low values for particles of this size ΔH and ΔS faster for nanocrystals with smaller diameters (Figure 10),
values of hydrogen absorption of −65.5 kJ mol−1 H2 and 122.7 although the activation energies calculated for hydrogen
J mol−1 K−1, respectively.121 absorption (115−122 kJ mol−1) and desorption (126−160 kJ
In addition to binary and intermetallic hydrides, complex mol−1) were within previously measured values for bulk Mg.
metal hydrides can also be synthesized using plasma This large rate enhancement cannot be explained by the
techniques. For instance, reactive plasma involving the
evaporation of lithium metal in an Ar/NH3 atmosphere leads
to hollow Li2NH nanospheres with particle sizes between 100
and 400 nm.223 A similar approach can be used to synthesize
Mg3N2 nanocubes, which can be converted into 100−200 nm
Mg(NH2)2 nanospheres.224
Despite significant progress, the practical fabrication of the
metal hydride NPs and other nanostructures by plasma
methods is a complicated phenomenon which involves close
interactions among the thermofluid field, the induced electro-
magnetic field, and the processed particle phases, with
numerous process variables that need to be optimized Figure 10. Left: Mg nanocrystals synthesized by
independently. The plasma methods are in principle scalable, bis(cyclopentadienyl)magnesium reduction with potassium in aro-
but are currently limited to only a few examples of metal matic solvents. Right: Kinetics of hydrogen uptake for 25 (black), 32
hydride NPs, for which optimized synthesis conditions have (red), and 38 nm (blue) Mg nanocrystals at 300 °C. (Reproduced
been established. from ref 241. Copyright 2011 American Chemical Society.)

10786 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

decrease in particle size alone but is likely due to an increase in


the defect density present in smaller nanocrystals.241
Liu et al. reported the possibility of in situ catalytic doping of
Mg nanocomposites with Ni during chemical reduction with
lithium naphthalide in tetrahydrofuran (THF).242 The
approach generated small Ni NPs deposited on the surface
of 10−20 nm Mg NPs. Upon hydrogenation of the
nanocomposite at 225 °C, the γ-MgH2 phase appears. PCT
measurements reveal that the Mg−Ni nanocomposite has
superior hydrogen storage properties over the pure Mg
prepared using the same method. The Mg−Ni nanocomposite
can absorb 85% of its maximum hydrogen capacity within 45 s
at 125 °C, and a hydrogen capacity of 5.6 wt % can be
obtained within 10 h at room temperature. Through the use of
Ca as the reducing agent with di-n-butylmagnesium, the
formation of Mg nanofibers with a width of 40 nm and a length
of 0.4−4 μm was also reported when 1-dodecanethiol was used
as a surfactant.243
3.3.2. Electrochemical Reduction. Electrochemical re-
duction is a simple method for the preparation of nano-
structured metal hydrides. The electrodeposition process
involves the use of a two- or three-electrode cell system, Figure 11. Scanning electron microscopy (SEM) images of Mg
where the electrolyte generally serves both as the source of the nanowires electrodeposited from (a) methylmagnesium chloride and
metal, often Pd or Mg, and as the conductive medium. By (b) ethylmagnesium chloride. (c) Diameter size distribution of NWs
prepared from EtMgCl. (Adapted with permission from ref 246.
controlling the potential or current density of an electro- Copyight 2012 The Royal Society of Chemistry.)
chemical cell, deposition occurs from an oxidized form in
solution to a metallic state at the surface. The most
straightforward is the synthesis of interstitial hydrides, such Similarly, alkyl derivatives of Li, Na, and Al can be used to
as palladium hydride by using various electrochemical baths.244 generate nanoscale LiH, NaH, and AlH3. A variation of this
The synthesis of Mg NPs has also been reported through the approach is to perform the reaction under hydrogen pressure;
electrochemical reduction of organomagnesium compounds. Jensen and coauthors reported MgH2 NPs inside carbon
Haas and Gedanken proposed a sono-electrochemical aerogels using this approach.252 Figure 12 shows a schematic of
synthetic approach to prepare 4 nm magnesium NPs deposited the process, which initially involves MgBu2 (with Bu = n-butyl)
at the tip of a sonicating probe with ethyl- and butylmagnesium infusion into the aerogel scaffold using heptane as a solvent.
chloride as magnesium precursors.245 Mg nanowires (NWs) After solvent removal, the nanoconfined MgBu2 is thermally
were electrodeposited onto a Pt electrode with ethyl- or decomposed at 170 °C under 55 bar H2, which simultaneously
methylmagnesium chloride as a precursor (Figure 11).246 hydrogenates the Mg to give scaffold-incorporated MgH2 NPs.
Using magnesium acetate as a precursor and appropriate De Jongh and coauthors reported a direct synthesis of LiH
modification of the electrochemical synthetic method by Reetz NPs confined inside meso- and macroporous carbon xerogel
and co-workers,247 Mg NPs of 5 nm were electrochemically pores using n-butyllithium as a precursor.253 Since the xerogel
synthesized in THF with tetrabutylammonium bromide as a contains both meso- and macropores, the resulting LiH
stabilizer. In this process, the cationic surfactant adsorbed at particles have a wide particle size distribution from a few
the counter electrode, where Mg2+ is reduced, prevents direct nanometers to a few microns. Nevertheless, the H2 desorption
adsorption of the Mg nuclei onto the electrode surface in favor kinetics are improved, with the onset of hydrogen release as
of the nucleation and growth of surfactant stabilized Mg NPs low as 100 °C. Organometallic precursors of other metals have
in the electrolyte.248,249 also been used, for example, organoaluminum compounds are
Tarascon et al. studied the electrochemical reactivity of suitable for alane nanoparticle synthesis; however, under the
MgH2 with Li and demonstrated a reversible capacity of 1,480 reaction conditions used, AlH3 is believed to decompose
mAh g−1 at an average voltage of 0.5 V versus Li+/Li°. The quickly into Al and gaseous hydrogen.254 Hydrogenation of
electrochemical reaction results in formation of a composite Et3Al at 150 °C under 20 MPa H2 pressure in the presence of a
containing Mg NPs embedded in a LiH matrix, which can be Mg catalyst results in 99% triethylaluminum conversion, but
converted electrochemically into MgH2 NPs of 10 to 40 nm.250 metallic Al is isolated instead of AlH3.255 In contrast, Aguey-
Lithium hydride NPs have also been synthesized electro- Zinsou and co-workers found that carrying out the Et3Al
chemically by reductive hydrogenation of metallic lithium hydrogenation (10 MPa H2) in the presence of a Mg catalyst
intercalated into graphite.251 TGA-MS measurements reveal and tetraoctylammonium bromide as a surfactant, NPs of α-
that the onset of hydrogen desorption for the sub-100 nm LiH AlH3 could be produced.256
NPs is >500 °C lower compared to bulk LiH and starts at 3.4. Nanoconfinement
around 200 °C. Different approaches have been used to prepare nanosized
3.3.3. Thermal Decomposition. Metal and metal hydride hydrides in porous host materials and the feasibility of different
NPs can also be synthesized via thermal decomposition of methods depends on the nature of the host and the hydride.
metal salts and organometallic precursors. Thermal decom- Impregnation methods, commonly used for the preparation of
position of di-n-butylmagnesium leads to β-hydrogen elimi- catalyst/support materials, are frequently also used for the
nation with formation of MgH2 and the corresponding alkene. preparation of hydride@host composites. We will introduce
10787 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 12. Schematic of in-pore synthesis of MgH2 NPs via thermal decomposition of Mg(n-C4H9)2 under hydrogen gas.252

the commonly used host materials and discuss typical methods


for incorporating the hydrides into the systems.
3.4.1. Host Materials. There are few prerequisites for
materials to be suitable host materials for confining hydride
NPs, apart from being stable and nonreactive toward the
hydride NPs and the hydrogenation/dehydrogenation pro-
cesses. Furthermore, for potential real-world applications,
having a low density for high gravimetric and volumetric
loading is advantageous. The most commonly used materials
for these stabilizations are usually carbon-based scaffolds with
an inherent porosity. Quite often, aerogels prepared by the
polymerization and carbonization of resorcinol and form-
aldehyde are used. Their pore sizes are determined by the pH
of the solution from which they are synthesized,257 and surface
areas and pore volumes can also be modulated by
postsynthetic treatments, such as carbon dioxide activa-
tion.16,258 Additionally, monolithic, rather than powdered,
aerogels can be mechanically abraded if a portion of the
hydride material meant for infiltration adheres, typically as bulk
hydride rather than nanoscale, to the exterior, allowing for
more accurate assessment of solely the nanoconfined
hydride.252,259 Another frequently used carbon material is
CMK-3,260 which is prepared by polymerization of carbon
sources (e.g., sugars) inside the pores of the ordered Figure 13. TEM micrographs (a,b) and electron diffraction pattern
(c) of Mg encapsulated in a poly(methyl methacrylate) nanowrapper.
mesoporous silica template SBA-15.261 The large surface area
(Adapted with permission from ref 21. Copyright 2011 Nature
and pore volume, which allow the deposition of a large volume Publishing Group.)
ratio of hydride in this material, make it a frequent choice.
Apart from that, the tunability of the pore space and the
possibility to dope the host structure with heteroatoms also 3.4.2. Impregnation Methods. Impregnation methods
make this particular material highly interesting for the comprise two major synthetic approaches of hydride confine-
controlled inclusion of hydrides. The use of high surface area ment. On the one hand, there are solvent impregnation
graphite (HSAG) is also very common, since the material itself methods that rely on the solubility of the hydride in an organic
is commercially available. More exotic materials include, for solvent. Second, there are melt infiltration methods, which
require the hydride to have a melting point under reasonable
instance, metal−organic frameworks, which consist of metal-
conditions (i.e., LiBH4 and NaAlH4 melt at 268 and 183 °C,
containing building blocks bridged by organic ligands.262 The
respectively, under ambient conditions).137 However, mixtures
tunability of this class of materials makes them interesting
of two or more hydrides can form eutectics, which have lower
hosts by allowing discrete tuning of pore geometry and melting points than the constituent components and can thus
functionality. However, the weak metal−ligand coordination also be melt-infiltrated into host materials.61,263 Several binary
bonds and the high weight penalty from the different inorganic borohydride eutectics have been confined and examined in this
clusters are considerable drawbacks to employment as stable way.264−268 Importantly, melt infiltrations are often carried out
and high-capacity hosts. under hydrogen backpressure to avoid desorption during the
More and more commonly, materials without any inherent process. Mg(BH4)2 decomposes prior to melting even at
porosity are used for the encapsulation of hydrides; in some typical H2 pressures employed (up to 200 bar).269,270
cases these hosts are also referred to as nanowrappers (Figure In general, melt infiltration is a method that achieves very
13).21,30,34 Materials that have been used for the nano- high loadings. The porous host is usually ground with the
encapsulation of particles are graphene, graphene oxide, and hydride to ensure good mixing and afterward put in a tube
poly(methyl methacrylate) polymers.21 Essentially, hydrogen furnace with H2 gas overpressure. De Jongh and co-workers
has a small enough kinetic diameter to diffuse through these additionally pressed the hydride/porous host mixture (MgH2
wrappers, while other gases are blocked out. These wrappers and carbon), in order to get the two compounds into more
protect the materials from external contamination, stabilize intimate contact, and they ended up with loadings of 15−
small particles, and tune the hydrogenation and desorption 20%.51 For hydrides such as MgH2, LiBH4, and NaAlH4, there
behavior. are numerous reports describing loadings of up to 50−65%. In
10788 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

one of these, at 60% loading, the average NaAlH4 crystallite 4.1. Experimental Structural Analyses
size was found to be greater than the average pore diameter of 4.1.1. Transmission Electron Microscopy. Transmission
the aerogel hosts, suggesting that larger pores fill first or more electron microscopy is one of the most used techniques to
readily during melt infiltration than smaller pores.257 characterize nanostructured metal hy-
In the case of solvent impregnation, the host material is drides.23−25,33−42,146,164,277−279 One of the most important
soaked by a solution of the precursors. For hydrides, the most recent developments is the spherical aberration correctors for
common solvents are ethers, such as (methyl tert-butyl ether both conventional TEM and scanning TEM (STEM) modes,
(MTBE), diethyl ether, or THF. In the case of NH3BH3, enabling atomic scale imaging and chemical analyses of hydride
MeOH is also a frequently used solvent. The loadings achieved samples.280 A critical requirement is making the sample
by this technique are often not as high as for melt infiltration, thickness below about 500 nm in order to have a sufficient
even though in the case of NH3BH3 there are several reports fraction of the electron beam transmitted through the sample.
highlighting over 50 wt % of hydride loading.271 However, The typical resolution of an image in contemporary TEM
repeated instances of solvent impregnation can be employed to instruments is in the range of 0.2 nm without using a spherical
enhance hydride infiltration. For instance, Zheng and co- aberration corrector, and one can achieve a resolution of 0.05
workers used multiple soaking steps with a 1 wt % NaAlH4 nm by employing various corrections. Metal hydride samples
solution to impregnate SBA-15 and achieved an actual loading are typically beam sensitive; that is, they decompose to
of 20 wt %.272 hydrogen and the metal under the high flux and energy of the
3.4.3. Synthesis Inside the Pores. There have been standard electron beam, and they typically require cryogenic
several reports highlighting the possibility of hydride synthesis conditions and low-dose or low-energy electron beams to
inside the pores. In order to realize this, the host must be stable maintain their original structure.
against the very reactive precursor molecules, e.g., Mg(Bu)2. 4.1.2. X-ray and Neutron Techniques. X-ray and
There are several reports on carrying out the synthesis of neutron techniques can provide detailed information about
MgH2 inside the pore space of different host materials. Usually, the structure and dynamics of nanostructured metal hydrides.
the host is exposed to a solution of Mg(Bu)2 in heptane at low The sensitivity of a material to X-rays depends on the number
temperatures. The solvent is then removed with vacuum, of electrons and thus the atomic number of its constituent
creating finely dispersed Mg(Bu)2, which is converted into elements; as such, hydrogen is undetectable. However, the
MgH2 and butane gas by hydrogenation.273 In this study, 22 wt interactions of neutrons with nuclei is not proportional to the
% of MgH2 were generated inside carbon nanofibers. Nielsen number of nucleons; thus, hydrogen is readily measured, and
and co-workers were able to incorporate 18.2 wt % inside a protium and deuterium can even be distinguished.281 Current
carbon aerogel by this method.259 However, potentially higher efforts are focused on developing new beams and optics to
loading could have been achieved, as the used carbon aerogel produce appropriate and tunable wavelengths/energies and
was a single monolith and excess MgBu2 that crystallized on detecting their interactions with materials. This is particularly
the outside surface was removed mechanically. The highest helpful to ascertain the arrangements, both physical and
reported value for MgH2 loading by this method was achieved chemical, of atoms and to track their movements or
by Jia and co-workers, amounting to 37.5 wt %. In their study, rearrangements during hydrogen storage reactions.23−25,33−42
doped CMK-3 frameworks were exposed to Mg(Bu2) and The arrangement of atoms inside nanoscale metal hydrides
afterward dried over several days.274 The hydrogenation step can be determined using several techniques based on
was realized inside of a PCT instrument. In the literature, the scattering, diffraction, spectroscopy, and imaging. Since both
in-pore synthesis has not been limited to MgH2; Mg(BH4)2 X-rays and neutrons are in principle nondestructive, such
was synthesized inside the pore of a carbon aerogel by Züttel et techniques can be used for in situ studies of nanostructured
al.275 They first melt-infiltrated the aerogel with MgH2 and metal hydrides during hydrogen absorption and desorption. X-
afterward ball milled the MgH2-containing aerogel in a B2H6/ ray diffraction is one of the most common analytical methods,
H2 atmosphere for 3 days, yielding Mg(BH4)2 in carbon though neutron diffraction has also occasionally been
aerogel. employed in hydrogen storage materials.282 Small-angle
scattering measurements, with both X-rays (SAXS) and
4. METHODS TO ANALYZE STRUCTURE AND neutrons (SANS), are highly useful in analyzing nanoscale
STORAGE PROPERTIES materials, especially those that are nanoconfined.283−288
For optimization of hydrogen storage properties of nano- Synchrotron radiation techniques have proved versatile in
structured metal hydrides, it is critical to establish a elucidating the mechanisms of hydrogen storage reactions,
quantitative feedback loop concerning particle size, shape, with techniques involving “soft” X-ray (∼0.1 keV, correspond-
composition, internal structure, and their relationship to the ing to a wavelength of 12.4 nm) to “hard” X-rays (∼100 keV,
myriad of synthetic variables. Without a tight feedback loop or a wavelength of 0.0124 nm). For many of the experimental
between the structure and function, the synthetic effort is methods mentioned above, the primary advantages of
random with respect to whether changes in synthetic protocol synchrotron X-ray sources are the high intensities and energies
are improving the final material. Unfortunately, extending accessible, allowing for rapid sampling with high signal-to-noise
standard analytic methods to determine the structure of and measurements under non-UHV conditions, as in ambient-
nanoscale metal hydrides is not straightforward, as these pressure XPS (APXPS). X-ray diffraction, in particular, benefits
materials are composed of light elements and pose all sort of from the large penetration depth of short wavelength photons,
issues related to sample reactivity and sensitivity to X-rays and/ which permits high pressure and high temperature operations,
or high-energy electrons. Experimental methods to determine including for the in situ measurement of hydrogen absorption
the structure of nanostructured metal hydrides are mainly and desorption in hydrides.289 However, some important
based on X-ray, neutron, and electron microscopy techniques also exploit the tunability of synchrotron radiation.
tools.23−25,33−42,276 X-ray absorption spectroscopy (XAS), which can be divided
10789 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

into X-ray absorption near edge structure (XANES) and determine the shape of a crystal by performing energy
extended X-ray absorption fine structure (EXAFS), involves minimizations to identify the preferred crystal planes.302 The
the excitation of a core electron of a particular element that is Wulff construction, coupled with Gibbs energy minimizations,
dependent on the oxidation state and the coordination can then be used to gain insights into thermodynamic stability
environment of atoms in the sample.290 EXAFS in particular of metal and metal hydride nanoclusters.
extends its reach beyond the first coordination sphere to more 4.2.2. Prototype Electrostatic Ground State Method.
distant neighbors, allowing a more extensive structural analysis Another powerful tool to predict the structures of both bulk
of which atoms are near each other in the material, such as and nanoscale complex metal hydrides, called the prototype
which sites hydrogen occupies in a Pd−Rh nanoalloy.291 electrostatic ground state (PEGS), was introduced by Majzoub
Modeling of XAS spectra of pure materials has proven useful, and Ozoliņs.̌ 303 This Monte Carlo approach identifies ground
particularly in metal hydrides, which oxidize readily even in state structures by minimizing the electrostatic energy of a
low oxygen environments and for which the presence of oxide system of spherical cations and geometrically complex anions,
phases may distort accurate analysis.36,292 In situ XAS, with subject to constraints of interatomic distances. The energy
both heating and hydrogen pressure, has been employed to functional used this Monte Carlo approach is given by the
probe the hydrogenation and dehydrogenation of rGO- following equation:
wrapped Ni-doped Mg NPs.35 Scanning transmission X-ray
microscopy (STXM) is an application of XAS in which an Q iQ j ϵss
absorption image is obtained at varying photon energies, Etot = ∑ + ∑
i>j
R ij i>j
R ij12 (8)
resulting in a spectrum for every pixel. The resulting set of
images can be deconvoluted into component phases based on
known XAS spectra, yielding a compositional map of the where Etot is the total energy of the system. The first
material.293 summation is the Coulombic energy, with Qi and Qj denoting
4.1.3. Vibrational Spectroscopy. While X-ray methods the ionic charges and Rij is the interion distance, whereas the
can investigate the oxidation states of elements, vibrational second term, in which ϵss is the soft-sphere interaction
spectroscopy is able to study the bonding among atoms, potential, is taken from the repulsive part of the Lennard-
especially in the polyatomic anions of complex metal hydrides. Jones potential, which only contributes to the overall energy
Infrared (IR) spectroscopy can readily assess the presence of, when the ions physically overlap. Both sums involve all ions in
for example, aluminum−hydrogen,294 boron−hydrogen,292,295 the material, including both the metal cations and the rigid
and nitrogen−hydrogen bonds, and the changes between complex anions. The main advantage of the PEGS approach is
hydrogenated and dehydrogenated states are apparent. Diffuse that it can generate crystal structures quickly, without the need
reflectance infrared Fourier-transform spectroscopy (DRIFTS) for the extremely time-consuming development of interatomic
in particular has been employed for in situ desorption potentials. In addition to accurately predicting the crystal
measurements, using heating capabilities.296 Raman spectros- structures of a number of bulk alanates and borohy-
copy has also been employed to detect vibrational modes in drides,304−309 the nano-PEGS variation of the code was used
complex hydrides, both ex situ and in situ.297,298 Using X-ray to generate ground-state structures for NaAlH4 and LiBH4
Raman spectroscopy (XRS), a synchrotron technique, nanoclusters.144,310,311 The model takes advantage of the ionic
Miedema et al. monitored the in situ dehydrogenation and nature of metal hydrides by fixing the local arrangement of
rehydrogenation of nanoconfined LiBH4 in carbon, since the complex anion groups, for example, the BH4− anion in LiBH4.
high-energy X-rays permit the use of elevated pressures of Not only does the PEGS method accurately predict many
hydrogen.299 Neutron vibrational spectroscopy (NVS), which known crystal structures, but it can also predict the structure of
is very sensitive to hydrogen-containing species, has been unknown materials. The predicted structure can then be
measured for both bulk300 and nanoconfined19 complex metal refined using more accurate DFT calculations to establish the
hydrides. A disadvantage to NVS is that 10B absorbs neutrons atomic coordinates of all atoms, including hydrogen.
well and drastically decreases the signal in B−H-containing Majzoub et al. used the predicted nano-PEGS structures to
materials; however, the use of 11B-enriched samples overcomes evaluate the thermodynamic stability of LiBH 4310 and
this problem.300 NaAlH4144 nanoclusters via the grand-canonical free-energy
minimization approach based on total energies and vibrational
4.2. Theoretical Structural Determination frequencies obtained from DFT calculations. Nanoclusters of
4.2.1. Wulff Construction. Theoretical modeling is a key LiBH4 containing fewer than 12 formula units are predicted to
component of understanding the structural features of dehydrogenate through intermediates of the form LinBnHm (m
nanostructured metal hydrides. DFT is the technique of ≤ 4n) into (LiB)n clusters (Figure 14). In the sodium
choice to predict the crystal structures and atomic arrange- aluminum hydride system, the stability of NaAlH4 and AlH3
ments in metal hydride NPs and nanoclusters. Johnson and clusters increases as cluster size shrinks. Interestingly, AlH63−
Sholl introduced a “top-down” approach to predicting the nanoclusters are unstable due to a Jahn−Teller distortion
shape and surface energy of metal and metal hydride arising from degenerate highest occupied molecular orbitals,
nanoclusters.301 The crystal structure of each nanocluster is leading to decomposition of the normally octahedral species.
assumed to be identical to the bulk crystal, aside from This instability of the dehydrogenation intermediate Na3AlH6
relaxation of the atoms in the first few layers near the surface. yields a one-step decomposition in nanoscale NaAlH4. The
For each surface exposed on a nanocluster, the energy calculated PCT isotherms for both LiBH4 and NaAlH4 exhibit
difference between edge and bulk atoms is characterized by a sloping plateaus due to finite size effects, and the
surface energy. The equilibrium crystal shape of a material can decomposition temperatures of free-standing clusters are
be then predicted using the calculated surface energies and the found to decrease slightly with size due to thermodynamic
Wulff construction. The Wulff construction is a method to destabilization of reaction products.
10790 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

boundary effects,327 and elastic inhomogeneous effects328,329 in


particles and polycrystals. The method has been actively
utilized to study phase evolution in materials for electrical
energy storage,330−332 but has only recently been applied to
hydrogen storage materials. Voskuilen et al.333 developed a
phase-field model for describing hydrogen transport and
operating reactions for hydriding LaNi5 and TiCrMn. In
addition, Ulvestad et al.334,335 employed the steady-state phase-
field calculations to analyze the internal strain field and
hydriding phase morphology in Pd NPs. Further development
of phase field-models for hydrogen storage will provide a
means of more systematically exploring the interplay between
microstructure and reaction kinetics.
4.3. Thermodynamic and Kinetic Analysis
Accurate determination of the thermodynamics and kinetics of
metal hydrides is critical to ascertaining the effect of structural
nanoscaling on the hydrogen storage properties. Many of the
Figure 14. Atomic geometries of various B−H clusters determined experimental techniques employed for bulk hydrides are the
from nano-PEGS and first principle calculations. (Reproduced from same for nanostructured materials and are described below.
ref 306. Copyright 2012 American Chemical Society.) Computational models used to deconvolute energetic con-
tributions from various size-related properties are discussed in
section 6.
Since the PEGS approach uses a simple energy model 4.3.1. Thermodynamic Measurements. The general
consisting of purely electrostatic interactions, it has some reaction involving reversible hydrogen storage in metal
limitations in predicting the structures of metal hydrides with a hydrides can be written as
high-degree of covalent character. Other methods, such as x
constrained evolutionary algorithms,69,312−316 have been ABHx(s) F AB(s) + H 2(g)
2 (9)
proposed for the prediction of metal hydride crystal structures
consisting of well-defined molecular units. In this approach, The equilibrium state of this reaction can be described in terms
each structural unit is treated as a rigid body with fixed bond of the equilibrium constant K and enthalpy change by the van’t
lengths, and then the positions of preformed molecules are Hoff equation:
found through constrained global optimization, a process of
finding the most stable crystal packing given a fixed bond d ln K ΔH 0
=
connectivity. This evolutionary approach to structure deter- dT RT 2 (10)
mination has been successfully implemented in the USPEX Since
(Universal Structure Predictor: Evolutionary Xtallography)
code.317−322 However, despite progress, prediction of com- ΔG 0 = −RT ln K (11)
pounds with large unit cells is still challenging. For instance, and
none of the existing theory methods are able to accurately
predict the structure of γ-Mg(BH4)2 with the P61 symmetry, ΔG0 = ΔH 0 − T ΔS 0 (12)
which has as many as 330 atoms in the unit cell.323 and, because the gaseous hydrogen is the only species with an
4.2.3. Microstructure Modeling. Beyond the crystal activity not equal to one, eq 10 can also be represented as
structure and nanoscale geometry of the particles, modeling
the impact of microstructural features on hydrogen storage at d ln peq ΔH 0
=
the nanoscale is likewise important. Nevertheless, there have dT RT 2 (13)
been relatively few attempts to directly model the micro-
or
structure in hydrogen storage materials. One example can be
found in the work of Michel and Ozoliņs,̌ who investigated ΔH 0 ΔS 0
microstructural effects on diffusion kinetics in the bulk Na− ln peq = − −
RT R (14)
Al−H system by considering different model arrangements of
NaAlH4, Na3AlH6, and Al.169 Another example comes from the where p is the equilibrium pressure for reaction 1; ΔH , ΔS0,
eq 0

work of Wood et al., who considered the thermodynamics of a and ΔG0 are the standard enthalpy, entropy, and Gibbs free
few different model microstructures in the nanoscale Li−N−H energy change of the reaction; T is the absolute temperature;
system.19 Although illustrative, these attempts have relied on and R is the universal gas constant.
static models without explicitly considering how the micro- Equation 14 indicates that a plot of ln peq vs 1/T has a slope
structure evolves during the reaction. of −ΔH0/R and an intercept of ΔS0/R. For an equilibrium
The phase-field method has emerged as one of the most pressure of 1 bar, the decomposition temperature Tdec is
powerful methods for mesoscale modeling of material ΔH 0
microstructural evolution.324 This method is based on the Tdec =
diffuse-interface description that naturally connects the ΔS 0 (15)
essential thermodynamic and kinetic ingredients of relevant For a typical standard entropy change of a metal hydride of
chemical and materials processes.325 Recent advances have 120 J K−1 mol−1 H2, at 350 K ΔH0 = 30 kJ mol−1 H2. ΔH0
extended the method to account for surface effects,326 grain values of ∼30 ± 10 kJ mol−1 H2 are considered within the
10791 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 15. Changes in the rates of absorption (a,d) and desorption (b,e) and the activation energy (c,f) as functions of the hydrogen content in Ni-
doped (a−c) and undoped (d−f) rGO-Mg composites. (Adapted with permission from ref 35. Copyright 2017 John Wiley and Sons.)

range of interest for near-ambient temperature reversible any semblance of a plateau if the disorder is particularly
hydrogen storage applications. For a hydride with an substantial, as found in amorphous and some nanostructured
equilibrium pressure of 1 bar, a 10 kJ mol−1 H2 variation in materials.338 If a material has a multistep (de)hydrogenation
ΔH0 results in about 80 K change in the decomposition reaction, as in NaAlH4/Na3AlH6, or contains two hydrides in a
temperature. mixture, such as MgH2 and Mg2NiH4, multiple plateaus,
The thermodynamics, particularly the equilibrium pressure corresponding to each reaction, will be present in the PCT
peq, are typically measured using a volumetric Sieverts-type curves.
instrument. A sample of known initial mass and composition, 4.3.2. Thin Film Hydrogen Content Measurements. In
either with or without hydrogen, is loaded into a pressure general, the total mass of a deposited film is very small and
vessel, and the volume of the sample holder is calibrated using below the threshold for gravimetric or volumetric measure-
a typically inert gas and the calibrated volumes and pressure ments described above that are commonly used in determining
transducers inside the instrument. The two calibrated volumes, hydrogen uptake.230 Therefore, alternative analysis methods
both that exclusively within the instrument manifold and the must be employed to gain insight into the hydrogen content of
total volume including the pressure vessel, are utilized to thin films.339 Nuclear reaction analysis (NRA) bombards the
determine the amounts of hydrogen dosed and the final sample with 15N ions at different energies to sample different
remaining gas not adsorbed or absorbed by the sample, depths, and emitted γ rays, which are proportional to the
respectively. Frequently, numerous measurements are obtained hydrogen concentration at each depth, are detected. Elastic
at a variety of pressures and, often, temperatures to sample a recoil detection analysis (ERDA) employs heavy ions, such as
representative portion of the pressure−composition−temper- Cl7+ or Ag9+, to obtain depth profile concentrations of all
ature space. In general, the pressure is modulated sequentially, elements, including H.232,340 Neutron reflectometry or
gradually adding more or removing more hydrogen from the reflectivity relies in part on the large neutron cross-section of
system to shift the composition, while remaining at constant hydrogen and measures the scattering angle of neutrons by a
temperature, and several such isotherms are measured to yield sample based on the scattering lengths of the pure metal
the van’t Hoff plot. Gravimetric instruments yield identical compared to that of the metal with hydrogen.341
information as volumetric setups, but the compositional shifts In addition to the more direct measurements of the presence
are determined by changes in the measured mass of the of hydrogen, other techniques have been developed that sense
material, rather than from differences in the dosed and final changes in the parent material as the hydrogen content evolves.
hydrogen amounts, as it absorbs and desorbs H2. In hydrogenography, visible light is used on materials that
In a simple interstitial hydride such as Pd, a flat plateau in a transition from reflective metals to opaque or transparent
pressure−composition isotherm is established, with little hydrides upon hydrogenation, utilizing a property that also has
pressure change corresponding to a substantial change in the applications in switchable mirrors and windows.342,343 IR
hydrogen content in the material as the solid solution of α and emission also has been employed, monitoring changes in
β phases evolves from mostly α to mostly β.336 The midpoint intensity over time and even over area as a material is cycled.28
of this plateau, as long as it remains at constant composition, is The IR emissivity depends on the electrical resistivity, which
typically chosen as the value of peq. In more complex systems, decreases as a metal is hydrogenated to a semiconducting or
however, a sloping, rather than flat, plateau is obtained; the insulating hydride, and is itself often used as a measure of
slope may result from atomic disorder, such as defects or uptake.230,234 XRD is another commonplace indirect analysis
noncrystallinity, or from nonequilibirium processes, as found in method and is sensitive to both small changes in lattice
irreversible reactions.337 Some hydride systems completely lack parameter, at low hydrogen contents, to complete structural
10792 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 16. (a) Computed activation barrier for early hydrogenation of MgB2 from Arrhenius analysis. (b) Reaction energy landscape for the two-
step mechanism of dissociation and diffusive adsorption, with adsorption energies derived from ab initio calculations. (c) Experimental and
simulated uptake curves showing the change in rate-limiting step. (d) Schematic of the proposed mechanism. (Reproduced with permission from
ref 292. Copyright 2017 The PCCP Owner Societies.)

transformations from the metal crystal structure to that of the method to analyzing multiple isothermal data for rGO-
hydride. For thin films, grazing angle, also known as glancing encapsulated Ni-doped Mg NPs to extract the activation
incidence, XRD is employed so that the X-ray beam samples energy.36 By interrogating the varying activation energy with
the film preferentially, rather than the substrate. the extent of the hydrogen absorption/desorption reactions
4.3.3. Rate Analysis and Kinetic Models. Kinetic (see Figure 15), they could identify the rate-limiting
modeling and analysis of hydrogen storage materials is critical mechanisms for different reaction stages of (de/re)-
to developing an understanding of rate-limiting processes hydrogenation.
Ea ij 1 yz
jj zz + ln(A)
during (re/de)hydrogenation. In many instances, kinetic

R kT {
insights can be obtained by fitting time-dependent reaction ln(k) = −
data to well-known models of (de)hydrogenation that exhibit (16)
characteristic uptake behavior.344 Dominant rate-limiting Another common approach to obtaining Ea is the Kissinger
mechanisms can then be straightforwardly identified from method.348 This method is usually used to determine the
model parameters derived from the fit.345 Alternatively, more activation energy from nonisothermal (de)hydrogenation
complex mechanisms may require advanced approaches that reaction data with different heating rates, typically using
directly extract activation energies for relevant kinetic differential scanning calorimetry (DSC) systems, and has been
processes. One way to obtain kinetic data for a metal hydride widely used in hydrogen storage research.173,176,194,294,344,349
is to measure the time-dependence of the hydrogen uptake or The linearized Kissinger eq 17, in which A is a constant,
desorption during acquisition of a PCT isotherm. Small assumes a constant heating rate, β, leading to a temperature at
perturbations in the dosed gas pressure, generally though not which the maximum rate occurs, Tm, seen as a peak in a DSC
necessarily while on the equilibrium plateau, lead to recovery curve which shifts to higher values as β increases. Kissinger
over time toward the initial sample pressure, as the stored analysis has the advantage of requiring fewer measurements
capacity and gaseous H2 content shift relative to each other, overall. However, it assumes only a single rate-limiting process,
restoring equilibrium according to Le Chatelier’s principle.337 so the composition range of interest must be chosen carefully if
The rate of this restorational change can then be used, along multiple processes are kinetically competitive. Though many
with those measured at other temperatures, to fit the Arrhenius standard DSC instruments operate at ambient pressure with
eq 16, in which k is the rate constant, A is the pre-exponential gas flow, high-pressure variants are available to maintain a H2
factor, and Ea is the activation energy.346,347 The activation overpressure and thus interrogate compositional spaces
energy for the (de)hydrogenation reaction can thus be otherwise inaccessible.
ij β yz ij AR yz E ij 1 yz
lnjjj 2 zzz = lnjjj zzz − a jjj zzz
extracted from the slope of the plot of the natural logarithm

jT z
k m {
j Ea z
k { R jk Tm z{
of the rate (or rate constant) vs the reciprocal absolute
temperature. In principle, this process can be extended across (17)
the entire range of (de)hydrogenation, yielding Arrhenius-
derived activation barriers for each degree of hydrogen content Kinetic models and analyses have been applied to analyze a
in the metal hydride. The activation energy as a function of wide scope of kinetic processes ranging from the surface-
reaction progress can then be utilized to analyze dynamic related initial stage of (re/de)hydrogenation to the deeper
changes in rate-limiting processes and/or to inform the more hydriding/dehydriding phase transformations. For deeper
detailed kinetic models. For example, Cho et al. applied this hydrogenation processes, one of the most common approaches
10793 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

applies the Johnson-Mehl-Avrami−Kolmogorov (JMAK) if rotational modes associated with molecular anions are
model, which describes the hydrogenation process as a phase sufficiently anharmonic at operating temperatures. Indeed, in
transition operated by simultaneous nucleation, growth, and NPs, the larger free volume associated with surfaces, interfaces,
impingement of a new phase inside the parent phase.344 A and grain boundaries is likely to exacerbate anharmonic effects.
recent review article by Pang and Li documented the JMAK Zarkevich et al.357,358 considered the anharmonic contributions
model, as well as a number of other predefined kinetic models, to the enthalpies of phase transformations and hydrogen
such as geometrical contraction models for interface- or storage reactions in LiBH4. By computing the degrees of
diffusion-controlled processes.344 Their review also provides a freedom originating from anharmonic motion using ab initio
comprehensive overview of the assumptions and derivations of molecular dynamics and classical statistical mechanics, they
these kinetic models, as well as several examples of application showed that including these effects brought phase transition
to hydrogen storage. temperatures and melting enthalpies into better agreement
Within certain reaction regimes, it may be difficult to with experiments.
identify the specific rate-limiting steps. One example is the DFT methods for computing enthalpy and entropy can be
earliest stages of (de)hydrogenation, in which governing extended to predict the thermodynamic stability of NPs by
processes may differ significantly from later stages dominated combining calculations of bulk and surface regions. The most
by diffusion, nucleation, and/or growth. A recent example of straightforward method involves determining the effective
this can be found in the work of Ray et al.292 They employed a surface energy of the particle by considering the surface
multiscale approach that combines quantum-mechanical energies of various individual orientations and terminations,
calculations, the Arrhenius analysis method, and a kinetic then weighting them relative to their expression within an
model to verify the relevant heterogeneous multistep chemical equilibrium Wulff shape. The effective surface energy is then
and materials processes during the initial stages of hydro- scaled according to the surface area and combined with the
genation of MgB2. In their integrated model, the energy bulk energy to estimate the Gibbs free energy of the particle.19
landscape for the entire process (Figure 16b) was constructed Note that this approach assumes that the particle assumes an
by combining the hydrogen binding energetics from DFT ideal Wulff shape with well-defined facets, and that the core
calculations and the activation energies (Figure 16a) from the region behaves like the perfect bulk material. As a result, it is
Arrhenius analysis of isothermal hydrogen uptake curves for likely better suited for larger NPs than for smaller NPs, of
three different temperatures. The energy landscape para- which disordered surface regions comprise a larger fraction of
metrized the corresponding kinetic model consisting of the material and core regions are difficult to delineate.
coupled differential equations that describe coevolving kinetic Recognizing this difficulty, a number of authors have turned
processes, which verified the rate-limiting processes (Figure to explicit modeling of cluster geometries within DFT.
16c). For clusters, the size-dependent particle geometry of
4.3.4. Thermodynamic Models. Fundamental phase nanoscale metal hydrides has been predicted by a variety of
diagrams for several metal−hydrogen systems have been methods, including molecular dynamics simulations,15 cluster
constructed by conventional CALculation of PHAse Diagrams expansion,359 and PEGS combined with a genetic algorithm.310
(CALPHAD) modeling method,83,350,351 which relies on For example, using cluster expansion on NaAlH4, Mueller et al.
inputs from experimental measurements and ab initio proved that particles larger than 5 nm in diameter expose
calculations of thermodynamic quantities. Other similar prominent low energy surfaces, as expected, but surface facets
techniques informed by DFT enthalpy calculations have also and structures become less clear in smaller particles.359
been proposed for computing thermodynamic or thermochem- In addition to surface energies, the energy of internal
ical equilibria of multicomponent hydrogen storage materials. interfaces and grain boundaries can contribute meaningfully to
Kim et al. demonstrated a thermochemical equilibrium the thermodynamics of metal hydride NPs, particularly when
calculation approach using data obtained from DFT phases coexist during (de)hydrogenation. Interface energies
computations.352 They investigated the thermodynamically can be computed by explicitly attaching two phases within
stable solid and gas phases of LiNH2, LiBH4, Mg(BH4)2, as periodic boundary conditions in DFT360−362 or by using
well as their mixtures with destabilizing compounds. appropriately parametrized classical force fields. However,
Akbarzadeh et al. determined the hydride phase diagrams of because determining lattice commensurability and preferred
the Li−Mg−N−H system353 by combining DFT calculations interface orientation can be difficult, this approach is best
with the grand canonical linear programming (GCLP) method, suited for simpler intermetallic and binary hydrides. For
which determines phase expression by minimizing the grand- complex hydrides and other systems featuring highly
canonical Gibbs free energy across several simultaneous disordered interfaces, the chemical composition and structure
competing reactions.354 This method was also used to of the phase boundaries can be difficult to model explicitly (see
investigate the phase equilibria of the Li−Ca−B−N−H section 6.2). To this end, Wood et al. introduced a simpler
system.355 Wood et al. introduced an extension of the GCLP method for approximating the internal interface energetic
method that further incorporates phase coexistence by penalty in the Li−N−H system as a weighted sum of surface
considering the entropy of mixing within an ideal mixing energies of involved phases.19 Because the weights of
model. individual surface energies were difficult to discern a priori,
Within DFT, the temperature dependency of the Gibbs free the authors instead employed a statistical sensitivity analysis to
energy is usually included by employing the quasi-harmonic determine the likelihood of phase expression as a function of
approximation, which extrapolates thermal effects from zero- the weighting factor. From this analysis, it was possible to
temperature phonon frequencies.306,353,356 This approximation demonstrate the critical roles of internal nanointerfaces in
is well suited for considering thermal effects of high-frequency determining the (de)hydrogenation reaction pathways of the
modes such as bond stretching; however, it has been suggested Li−N−H system. Further developments in accurate modeling
that inaccurate results may be produced for complex hydrides of interface energetics are recommended.
10794 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 17. Illustration of supporting materials for nanoconfinement: (a) porous host matrix (porous carbon, silica, MOF, carbon fiber, and
aerogel), (b) ligands (surfactants and polymers), (c) wrapping carbon (graphene and graphene oxide) and synthetic methods for producing
nanomaterials: melt infiltration, solvent impregnation, and solution-based synthesis.

5. EFFECTS OF MORPHOLOGY ON HYDROGEN use supporting materials, which can stabilize NPs to <10 nm.
STORAGE PROPERTIES Various supporting materials such as porous matrices, ligands,
and wrapping carbon materials have been proposed, and
For nanostructured materials, morphology has special sig-
different synthetic routes to create these supported metal
nificance since it dictates observed physical and chemical
hydride nanomaterials have been experimentally studied
properties. Morphology is attained during particle growth
(Figure 17). In the following section, we will discuss the
through a self-assembling process dictated by the interplay of
most common morphologies found in nanoscale metal
size and molecular interactions.363 Deviations from bulk
hydrides.
properties become prominent as the sizes of nanomaterials
start to be comparable to the size of constituent blocks. 5.1. Free-Standing Nanoparticles
Properties of metal hydride nanocomposites are controlled not Isolated metal hydride NPs are difficult to synthesize, and most
only by morphology of individual nanomaterials, but also by of the work so far has been focused on metals, which can form
the nature of interactions, which, in turn, are determined by metal hydride NPs upon exposure to hydrogen. Metal hydride
the distribution of the nanomaterials in the hydride matrix. NPs are convenient model systems to investigate the effect of
Reducing the particle size of metal hydrides to nanoscale size on reaction pathways, reversibility, as well as the
dimensions offers new possibilities to optimize the perform- thermodynamics and kinetics of the process. Polymer-coated
ance of hydrogen storage. Ball milling, one method of Pd NPs with diameters of 2.6 ± 0.4 and 7.0 ± 0.9 nm were
nanoscaling, yields freshly prepared surfaces that are highly investigated by PCT and show a narrowing of the two-phase
unstable, such that the newly created NPs (10−50 nm or regions of solid-solution and hydride phases. The smaller
larger) can easily aggregate into much larger particles (200− nanoparticles exhibited a smaller ΔH, indicating the weakening
300 nm).11 An alternative approach to creating stable NPs is to of the Pd−H bond as size decreases.107 However, it was found
10795 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

that the hydrogen binding energy differs for Pd atoms on the Nanowires or NPs of Mg with diameters >50 nm23 and
surface compared to those in the core. Kitagawa et al. hexagonal 50−500 nm Mg particles210 show no change in
investigated the Pd/Pt core/shell-type bimetallic NPs and thermodynamic properties, but dramatic improvements in the
found a new hydrogen absorption site in the nanointerface kinetics, with activation energies being decreased from 120 to
between the Pd core and Pt shell.364 38 kJ mol−1 H2.23 Urban et al. reported air-stable 5 nm Mg
Yamauchi et al. observed the size dependencies of the NPs with reduced activation energies for hydrogen absorption
hydrogen-storage properties of coated Pd NPs with diameters (25 kJ mol−1) and desorption (79 kJ mol−1), and a cycling
of 2.6 ± 0.4 and 7.0 ± 0.9 nm.107 The authors chose poly(N- temperature of 200 °C.21 Sub-10 nm Mg NPs show an even
vinyl-2-pyrrolidone) (PVP) to passivate the Pd NPs because it more dramatic improvement in H2 storage properties, with fast
binds well to metal surfaces. The carbonyl groups of PVP kinetics and improved thermodynamics which lead to
partly coordinated to the surface Pd, which limited the growth hydrogen cycling below 120 °C.52,367 Aguey-Zinsou reported
of the Pd NPs. As a result, the particle size of Pd increased as that LaNi5 NPs of about 26 ± 8 nm have an enthalpy decrease
the amount of PVP decreased. The Pd NPs showed of 5 kJ mol−1 H2 for absorption and 8 kJ mol−1 H2 for
significantly different isotherms from those of Pd bulk; the desorption, while the entropy decreased by 15 J K−1 mol−1 H2
NPs displayed higher hydrogen uptake at low pressure and a for absorption and 19 J K−1 mol−1 H2 for desorption.368
smaller change in enthalpy (Figure 18). Methods to synthesize free-standing NPs of complex metal
hydrides are scarce and have only been demonstrated for a few
materials. Varin and Parviz explored the influence of the
addition of TiN, TiC, and ZrC NPs on the activation energy of
LiAlH4 dehyrogenation.369 LiAlH4 was mixed with separately 5
wt % of each of the compounds in a high impact ball mill.
Interestingly, the activation energies for the first dehydrogen-
ation step are not strongly influenced and remain in the known
range for the bulk material, amounting to 87.7, 89.5, and 96.4
kJ mol−1 H2 for the composites containing TiN, TiC, and ZrC,
respectively. However, a drastic decrease is observed for the
second dehydrogenation step, with activation energies of 76.6,
79.5, and 63.4 kJ mol−1 H2 for the TiN-, TiC-, and ZrC-
containing composites, respectively. Liu et al. studied the effect
of doping LiAlH4 with micro- and nanosized TiH2.370 The
doping was achieved by ball milling and resulted in a marked
decrease of the activation energy by 20 and 24 kJ mol−1 for the
micro- and nano-TiH2 doped LiAlH4, respectively.
Christian and Aguey-Zinsou used an antisolvent approach to
synthesize NPs of sodium borohydride coated with a shell of
nickel.371 In contrast to bulk NaBH4 which releases hydrogen
>500 °C, the NaBH4/Ni nanocomposite can be cycled as low
as 350 °C with a 5 wt % capacity; 80% of hydrogen can be
desorbed and absorbed in less than 60 min, and full capacity is
reached within 5 h. A similar approach used to LiH/Ni core−
shell particles resulted in a dramatic improvement in the
kinetics of hydrogen desorption with about 10 wt % hydrogen
being absorbed and desorbed at 350 °C, a significant
improvement compared to bulk LiH (700 °C). Pang et al.
Figure 18. TEM images of Pd NPs with diameters of (a) 2.6 ± 0.4 developed a mechanochemistry/PVD approach for the syn-
nm (Pd−S) and (b) 7.0 ± 0.9 nm (Pd-L). (c) Absorption pressure− thesis of 20−40 nm Mg(AlH4)2 nanorods and 10−40 nm
composition isotherms of the Pd−S (red), Pd-L (blue), and bulk Pd
black (black) at different temperatures from 303 to 393 K. (Adapted
LiBH4 nanobelts185 with improved kinetics of hydrogen
with permission from ref 107. Copyright 2008 American Chemical desorption; importantly, the morphology of the Mg(AlH4)2
Society.) nanorods persisted after hydriding/dehydriding cycles.
Li2NH hollow nanospheres prepared by plasma metal
reaction of lithium with ammonia show a surface area of 79
Magnesium hydride has been studied extensively as it has m2 g−1. The porosity coupled with particle morphology leads
one of the highest capacities of all binary metal hydrides. NPs to enhanced hydrogen storage kinetics. The Li2NH hollow
of MgH2 typically display accelerated kinetics of hydrogen nanospheres absorb 6.0 wt % hydrogen in 1 min at 200 °C; in
release and/or uptake.23,121,21,365,241 Liu et al. isolated 8−25 addition, the desorption temperature is decreased by about 115
nm Mg NPs using an electroless approach and found that the °C compared with the bulk sample (Figure 19).223 The same
thermodynamics significantly deviate from that of bulk MgH2 group reported nanocubes of Mg3N2 which can be converted
with a significant decrease of both enthalpy (63.5 kJ mol−1 H2) to Mg(NH2)2 hollow nanospheres by reacting with ammonia
and entropy (118.4 J K−1 mol−1 H2) for particle sizes below 25 gas.224 The Mg(NH2)2 nanospheres milled with MgH2 release
nm.366 Experiments by Paskevicius et al. on MgH2 NPs about 6.5 wt % hydrogen be heating to 400 °C, with the initial
embedded in a LiCl matrix show that the enthalpy decreases 2.5 wt % hydrogen released between about 75 and 225 °C.
by 2.8 kJ mol−1 H2, coupled with an entropy decrease of 3.8 J Table 3 summarizes the synthesis and hydrogen storage
K−1 mol−1 H2.186 properties of numerous freestanding metal hydride NPs.
10796 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 19. (a) Hydrogen sorption curves of Li2NH hollow nanospheres (marked by N) and micron-sized particles (marked by M). (b) TEM
image of the as-prepared hollow nanospheres. (Adapted from ref 223. Copyright 2008 American Chemical Society.)

5.2. Thin Films lattice parameter, and the electronic effect of making them less
Thin films differ from NPs in both the synthetic methods and metallic had a lesser absolute effect.234
the resulting structure, necessitating unique analytical methods Several other transition metals have been investigated in thin
for determining hydrogen sorption properties. Unlike free NPs, film format. Palladium is often used as a capping layer in these
though similar to confined or encapsulated particles, thin films systems because it facilitates hydrogen activation from the gas
have a rigid substrate upon which they are deposited. phase and protects the highly oxophilic metals against
Additionally, films have two macroscopic dimensions, with oxidation from oxygen and moisture. Iron−titanium was
evaporated onto silicon and covered with 20 nm of Pd.228
only the thickness being on the nanoscale. These materials,
The resulting material was partially amorphous and could be
therefore, are dominated by the surface facing away from the
charged to 0.35 H per metal atom, with a discharge down to
substrate, leading to highly anisotropic communication with
0.12 H/M, indicating deep trapping sites. In addition, it was
the gas phase. However, thin films are rarely monocrystalline
possible to charge the amorphous film with hydrogen at higher
but instead typically highly textured, with small crystallites,
temperatures and lower pressures than microcrystalline films
numerous grain boundaries, and occasionally even some
due to the high disorder. Zirconium was cosputtered with Ni,
porosity, depending on deposition conditions and method.
Co, Fe, and Cu onto glass and cycled from vacuum to 1 bar.230
A wide variety of metals have been formed as thin films for The change in electrical resistance upon hydrogenation was
numerous different hydrogen-based applications, including most pronounced in the Ni-containing samples, and higher
hydrogen-switchable windows and mirrors339,372−374 and desorption temperatures restored the low resistivity better due
metal-hydride batteries.375 While these are intriguing fields, to increased ripening and growth of the grain sizes. Hamm et
these properties are not germane to this review, which instead al. aimed to destabilize metal hydrides using the clamping
will focus on the changes in H2 sorption relevant to hydrogen stress from the substrate (estimated to be −1.1 kJ mol−1 H
storage, particularly in lighter materials that are of potential GPa−1) but needed the strong adhesion of niobium on
interest for on-board vehicular storage. sapphire Al2O3 to prevent delamination and dislocation
Palladium has often been studied due to its simplicity, fast formation.378 Only films thinner than 6 nm were able to
kinetics, and isostructural hydride. Di Vece et al. prepared avoid plastic deformation and achieve the high stress desired
MBE and nanocluster Pd films on glass and investigated their up to 1 H/M, though 10 nm films could deform elastically after
behavior with both extended X-ray absorption fine structure an initial plastic deformation cycle due to hardening from
(EXAFS) and optical transmission spectroscopy to determine interactions of dislocations.
the effect of hydrogen cycling.226,227 The authors found that Magnesium has received much attention in thin film
hydrogenation led to Ostwald ripening, even at ambient research because of its lightweight and high gravimetric
temperature, due to reduction of the binding energy in the capacity (7.6 wt % H); it also absorbs hydrogen very poorly
hydride compared to the metal. Additionally, the smaller in the bulk due in part to the slow H diffusivity in Mg, making
nanoclusters lacked the miscibility gap, evidenced by a plateau research into nanostructured magnesium especially rele-
in the pressure as H content changes, between the α and β vant.67,379−381 Dura et al. prepared 100 nm Mg thin films
Pd−H phases that was evident for the MBE film and is more with 50 nm of Pd as a cap on monocrystalline Al2O3 and
typical of the bulk material. Similarly sized films electron-beam found, using both scanning electron microscopy (SEM) and
deposited on quartz showed a narrowing of the plateau, the scattering length from neutron reflectivity, that the porosity
particularly for 10 nm thick films, due to the induced stress increased from 4% to 25% upon cycling.382 Gautam et al.
from clamping of the film to the substrate.376 Khanuja et al. sputtered rather thick films (500 and 1000 nm) of Mg capped
compared both thin films and layers of discrete NPs, deposited with Pd.231,232 They found that the γ-MgH2 phase forms as a
on Al on glass, to investigate differences in resistivity result of compressive stresses while annealing and that the
changes.377 Hydrogenation was slowed in the films due to absorption amount, mean crystallite size, and number of
in-plane strain and clamping effects that were absent in the NP defects all increase with absorption temperature in the 150 to
layers which had interparticle gaps. Using sub-10 nm Pd NPs 250 °C range. A significant drawback to the capping of Mg
synthesized by reduction from PdCl42− with ethylene glycol with palladium is the formation of a variety of intermetallic
and dropcast onto glass, Gupta et al. found that the change in species at the interface between the two.
resistivity was dominated by the geometric effect of the Pd Transition metals have been added into thin films of
particles expanding to form the hydride, which has a larger magnesium to act as catalysts as well, either codeposited or in
10797 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Table 3. Hydrogen Storage Properties of Freestanding Metal Hydride NPs


Ea |ΔH| |ΔS|
material size (nm) shape synthesis method H2 wt % (kJ mol−1) (kJ mol−1)a (J mol−1 K−1)a ref
Pd 2.6 sphere solution-based chemistry/PVP ∼0.4 35b 83b 107
Pd 7.0 sphere solution-based chemistry/PVP ∼0.5 31b 67b 107
Pd 14 cube solution-based chemistry/CTAB 18.0b 27.4b 61.5b 109
47.1c 33.4c 75.8c
Pd 32 cube solution-based chemistry/CTAB 20.2b 28.5b 66.9b 109
50.6c 35.5c 82.3c
Pd 65 cube solution-based chemistry/CTAB 22.6b 29.9b 72.6b 109
54.4c 37.6c 88.2c
Pd 110 cube solution-based chemistry/CTAB 25.3b 35.4b 82.4b 109
60.5c 40.9c 94.4c
Pd 18 cube solution-based chemistry 28−36b 64−86b 418
Pd 24 cube solution-based chemistry 27−32b 61−78b 418
Pd 30 cube solution-based chemistry 32−36b 79−90b 418
Pd 32 cube solution-based chemistry 31−37b 74−94b 418
Pd 40 cube solution-based chemistry 30−35b 74−86b 418
Pd 42 cube solution-based chemistry 31−35b 75−87b 418
Pd 43 cube solution-based chemistry 29−32b 70−80b 418
Pd 45 cube Solution-based chemistry 28−34b 65−85b 418
Pd 54 cube solution-based chemistry 32−36b 79−92b 418
Pd 60 cube solution-based chemistry 35−41b 88−107b 418
Pd 63 cube solution-based chemistry 33−42b 83−108b 418
Pd 85 octahedron solution-based chemistry 27−31b 62−75b 418
Pd 85 octahedron solution-based chemistry 30−38b 71−96b 418
Pd 85 octahedron solution-based chemistry 29−32b 68−78b 418
Pd 38 × 137 nanorod solution-based chemistry 33−38b 81−96b 418
Pd 38 × 215 nanorod solution-based chemistry 34−38b 84−98b 418
Pd 47 × 185 nanorod solution-based chemistry 33−39b 78−97b 418
Rh 2.3 irregular solution-based chemistry ∼0.15 113
Rh 2.4−7.1 irregular ∼0.1 114
Rh 7.1−10.5 irregular ∼0.1 115
Ir 1.5 irregular solution-based chemistry, PVP ∼0.09 116
Ir 1.5 irregular solution-based chemistry ∼0.09 117
Li 4 spherical solution-based chemistry 2.0 419
Mg 2−7 sphere ball milling in LiCl 71.2b 129.6b 186
Mg 15 irregular electroless reduction 142.8c 63.5b 118.4b 273
Mg 30−50 nanowire gas phase condensation 33.5b 65.3c 23
38.8c
Mg 80−100 nanowire gas phase condensation 38.7b 65.9c 23
46.5c
Mg 150−170 nanowire gas phase condensation 70.3b 67.2c 23
81.1c
Mg 6.0 irregular Mg(n-Bu)2 hydrogenation 5.7 64.8c 420
Mg 40 hexagon acetylene plasma metal reaction 5.5 61.6b 65.5c 122.7 121
114.0c
Mg 5 sphere Cp2Mg reduction with Li-naphthalide in 5.97 25b 21
PPMA 79c
Mg 7−15 irregular spark discharge 4.2 60−120b 365
Mg 25 irregular Cp2Mg reduction with potassium 6.7 122b 241
126c
Mg 32 irregular Cp2Mg reduction with potassium 6.6 118b 241
131c
Mg 38 irregular Cp2Mg reduction with potassium 6.1 115b 241
160c
Mg 8−25 spherical electroless reduction ∼6.2 63.5 118.4 366
MgH2 11−20 fibers stabilization with azide-terminated poly ∼2.6 421
(styrene)
Ni-doped 5.4 irregular Mg(n-Bu)2 hydrogenation 5.4 22.7b 62.1b 420
MgH2 64.7c
Mg/Ti 10−20 irregular gas phase condensation 68.1 119 119
Mg/Ti 12 irregular spark discharge 4.0 45 84 422

10798 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Table 3. continued
Ea |ΔH| |ΔS|
material size (nm) shape synthesis method H2 wt % (kJ mol−1) (kJ mol−1)a (J mol−1 K−1)a ref
Mg/Ti 50 irregular spark discharge 4.0 170.9c 73.0b 130.9b 423
75.8c 134.8c
Mg/V 50−150 hexagons hydrogen plasma metal reaction 5.2 71.2b 74.3b 155.8b 424
119.4c 74.4c 152.9c
Mg/Co 50−150 hexagons hydrogen plasma metal reaction 4.9 82.3c 138.8c 425
Mg/Ti/Pd 60 nanodots molecular beam epitaxy 70 129 426
Mg/Ni 10−20 irregular chemical reduction 6.0 139.1c 70.0b 242
57.4b 70.7c
Mg@Ni hexagons arc plasma 5.1−6.1 88.9b 73.7b 105.5b 427
81.3c 104.5b
Mg/Ti,Fe,Ni irregular mechanical milling 4.2 45.7 428
MgH2/TiH2 5−10 irregular mechanical milling 4.6 16.4 429
MgH2/TiH2 5−10 irregular mechanical milling 6.1 68.2c 127c 430
MgH2/Nb2O5 irregular mechanical milling 6.3 431
Mg2FeH6 100 irregular reduction of precursor 4.95 89c 145c 432
Mg2FeH6 100 irregular sintering Mg/Fe NPs 5.0 165.7c 82.4c 140.2c 433
Li/Mg 8−40 irregular chemical reduction 159b 66.9 122.1 434
Na/Mg 7−52 irregular chemical reduction 148b 71.1 130.9 434
K/Mg 8−32 irregular chemical reduction 134b 63.2 117.8 434
LiAlH4 2−16 spherical stabilization with 1-dodecanethiol 5.7 417
Mg(AlH4)2 20−40 nanorods physical vapor deposition 9.0 97.3c 185
Mg(NH2)2/ 100 hollow plasma metal reaction 6.5 224
MgH2 nanospheres
Li2NH 100 hollow plasma metal reaction 6.0 223
nanospheres
MgB2 irregular milling under H2 4.0 435
LiBH4 10−40 nanobelts physical vapor deposition 185
LiBH4 30−50 spherical stabilization with PMMA 436
NaBH4 50−120 spherical stabilization with PMMA 436
Ca(BH4)2 100−300 hollow stabilization with PMMA 436
nanospheres
NaBH4/Ni 10−30 spherical solvent precipitation 5.0 371
a
The absolute magnitude of ΔH and ΔS. bValue obtained from absorption measurements. cValue obtained from desorption measurements.

discrete layers. Nickel forms Mg2Ni, which can be hydro- kinetics and capacity under cycling because the Mg/MgH2
genated to Mg2NiH4. Oguchi et al. synthesized 100 nm thick particles could not grow as large and limit the transport of
combinatorial Mg1−xNix thin films capped with Pd.28 The hydrogen. Ti, like Fe, is immiscible with Mg and, like Pd, has
compositions with more Mg showed greater hydrogenation via been used to protect the Mg and dissociate H2, while forming
IR emission imaging, and using XRD, the authors found that a TiH2. Baldi et al. made films with layers of 10 nm Ti and 20
nickel-stabilized fcc Mg phase enabled faster kinetics and a nm Mg, capped with Pd, on glass.343,384 Magnesium
lower temperature of reaction by transforming to a similarly sandwiched between two layers of titanium absorbed H2 at
structured fluorite-type hydride. Fry et al. made a multibilayer lower pressures than Mg between Ti and Pd (Figure 20),
film, capped with Pd, with 150 pairs of 16.5 nm Mg and 2.5 nm which they explained as due to clamping stress and alloying
Ni−Fe−Cr (predominantly nickel).42 PCT measurements, with the palladium. Mooij and Dam conducted follow-up
made possible by the large amount of material, showed faster experiments, again with Ti-sandwiched Mg, that showed that,
hydrogenation than ball-milled Mg and two different plateaus: in a 10 nm Mg film, the hydride phase nucleates in the same
one for MgH2 and one for Mg2NiH4, indicating mixing locations repeatedly, but upon dehydrogenation, the metallic
between the layers. After cycling, the delaminated films broke phase has no visibly obvious nucleation points but appears to
down into powder containing Fe NPs as well as Mg−Ni and fade in because the barrier to Mg nucleation is small and thus
Mg−Pd intermetallics. A 4% Fe-doped Mg thin film was the number of nuclei is large.385,386 Bannenberg et al.
deposited on silicon in a wedge shape to study the effect of film continued the study with neutron reflectivity and hydro-
thickness on hydrogenation.229 The rate of hydrogenation, as genography measurements on deuterated Ti−Mg films,
determined from IR emissivity, was limited by the diffusion of showing that the nucleated β-MgH2 regions grow at a constant
hydrogen through MgH2; however, PCT and van’t Hoff studies radial rate, with plastic deformation of the encompassing Ti
showed that the enthalpy and entropy for the reaction were and Pd layers to fit the larger phase.341 As in the study by Dura
less than those for pure magnesium films. et al., the Mg layer was thicker after dehydrogenation,
A pair of multilayered Mg/FeTi thin films with different indicating increased porosity and enabling faster rehydrogena-
layer thicknesses retained the intact FeTi layers upon cycling, tion kinetics. An X-ray photoelectron and Auger spectroscopy
while the Mg layers formed NPs confined between the investigation on cosputtered Mg−Ti films was conducted by
sandwiching layers.383 The thinner layers had more stable Jensen et al. and found, in addition to some partially oxidized
10799 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

ultrasmall pore size (from 2 to 0.5 nm) that can be achieved


via various synthetic methods.50,51,349,388−393 De Jongh et al.
have prepared high-purity and oxygen- or nitrogen-modified
ACs in order to synthesize Mg/C nanomaterials.51 The high-
purity carbon support has a nonpolar nature, which might
impede the infiltration of molten magnesium into the pores.
To improve the wetting on the surface of the porous carbon,
nitrogen, or oxygen was introduced. Mg nanomaterials were
prepared by melt infiltration, and the average size of Mg
decreased from a few nanometers for hierarchical micro and
mesoporous AC with ∼10 at. % O), to ∼2 nm for high surface
area graphite with mainly 2−3 nm pores to less than 1−2 nm
for microporous AC-based samples (Figure 21a). It is
Figure 20. Mg sample deposited on a glass wafer and partially covered
demonstrated that AC is able to mechanically confine Mg
with a 2 nm thin Ti layer. (a) Sketch of the sample architecture. (b) nanomaterials inside the pores and prevent oxidation of Mg
Pressure−temperature isotherms (PTIs) measured during loading at shortly after preparation (Figure 21b).
333 K. (Reprinted with permission from ref 384. Copyright 2009 AIP The use of AC as a confinement medium has been extended
Publishing.) to other materials, such as metal alanates and borohy-
drides.349,388−391,394 Nanosized NaAlH4 materials in high-
purity porous carbon HSAG-500 with relatively small pores
Mg, the presence of H atoms trapped in interstitial sites (mainly 2−3 nm) were prepared by the melt infiltration
between Ti atomic clusters and the Mg matrix.233 method.388 After melt infiltration, the NaAlH4/C nano-
Multilayers of 18 nm magnesium with 1 nm chromium or composites were examined by physisorption measurements
vanadium as catalysts showed different mechanisms from each because the metal alanate and borohydride materials
other, though the two activation energies were similar.29 Both decompose under the electron beam in TEM. The
catalyzed systems were much faster than the Mg control, but Brunauer−Emmett−Teller (BET) surface area and the
the thermodynamics were comparable. A set of Mg2Si−Mg Barrett−Joyner−Halenda (BJH) total pore volume decreased
thin films, each capped with palladium, was prepared by from 500 to 156 m2/g and from 0.57 to 0.32 cm3/g,
sputtering to determine the limitations on hydrogenation of respectively, after infiltration, indicating that the NaAlH4
Mg2Si, especially compared to ball-milled particles.387 efficiently filled (or blocked) most of the pores. In a follow-
Adsorbed hydrogen diffused through Mg2Si and reacted with up study by the group of de Jongh, the possibility of changing
metallic Mg without substantially hydrogenating the silicide; the reaction pathway by nanosizing it in HSAG and AC was
only regions immediately proximal to either the Mg or Pd were presented.395 In addition, NaBH4 and LiBH4 nanostructures
affected, so very small particles (<2 unit cells) would be were also confined in HSAG-500, which was confirmed by
required for full conversion to MgH2 and Si. broadening and reduction in the intensity of XRD peaks and
While thin films are of interest in fundamental research on the decrease in the total pore volume of the carbon, which is in
nanostructuring in controlled and tunable ways, they have good agreement with the materials completely filling in the
some drawbacks as well. The total hydrogen capacity of a thin carbon pores.389,390 By modifying a conventional method
film is minuscule due to the small area typically employed for through changing the pH of the initial solution, Nielsen et al.
deposition; this limitation significantly curtails the use of thin produced several distinct nanoscale pore dimensions in carbon
films in practical applications and restricts their utility to more aerogels, which distinctly influenced hydrogen release and
fundamental research into the effects of nanostructuring. In uptake properties in the NaAlH4 system.257
addition, some films are prone to either Ostwald ripening or Nielsen et al. explored the influence of the pore size of
pulverization upon cycling, breaking down the original resorcinol-formaldehyde carbon aerogels on the hydrogen
nanostructures and limiting the clamping effects from the desorption kinetics of nanoconfined NaAlH4.16 With increas-
substrate, further diminishing the possibility of their long-term ing pore size of the NaAlH4-loaded scaffold, the onset
deployment. Table 4 contains details about synthesis and temperature increases, linking the desorption kinetics to the
hydrogen storage properties of nanostructured thin films. particle size, which, in turn, is controlled by the pore size.
5.3. Nanoconfined Metal Hydrides Stephens et al. were able to alter the hydrogen desorption
The limitation of metal hydride nanoparticle growth by a temperature of NaAlH4 by melt-infiltrating the hydride into an
physical barrier that controls the ultimate nanocrystal size is a aerogel with 13 nm pores.396 Dehydrogenation to NaH was
robust strategy that has been demonstrated in a number of achieved at 150 °C with reasonable kinetics. Additionally, it
studies.45,46 A wide variety of metal hydride nanostructures was possible to rehydrogenate the material fully at 160 °C
have been synthesized via these approaches. Porous matrixes under 100 bar H2 pressure. Jensen et al. infiltrated a TiCl3-
such as porous carbons, carbon fibers, aerogels, silicas, and doped resorcinol-formaldehyde aerogel with NaAlH4 via melt
MOFs have been utilized as supporting materials because the infiltration.397 The resulting nanocomposite hosted NaAlH4
pore sizes are readily modulated, thus providing more particles with a size of 37 nm.
homogeneous size distributions and fewer aggregations of Gross and co-workers described the synthesis of LiBH4
metal hydride nanomaterials. inside pyrolyzed resorcinol-formaldehyde aerogels with varying
5.3.1. Confinement in Carbons. Activated nanoporous pore size (13 and 25 nm) by melt intrusion.398 The activation
carbons (ACs) and carbon nanofibers (CNFs) are two of the energy was lowered to 103 and 111 kJ mol−1 for the 13 and 25
most commonly used templates in the synthesis of metal nm LiBH4-loaded aerogels, respectively. In comparison, the
hydride nanomaterials due to their excellent accessibility and activation energy for bulk LiBH4 amounts to 146 kJ mol−1.
10800 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Table 4. Hydrogen Storage Properties of Metal Hydride Thin Films
H2 |ΔS|
metal(s)a thickness (nm)b substrate synthesis method wt % Ea (kJ mol−1) |ΔH| (kJ mol−1) (J mol−1 K−1) ref
Pd 10 glass vacuum evaporation 377
Pd 60 glass vacuum evaporation 377
Pd 16 Si/SiO2 laser vaporization 226
Chemical Reviews

Pd 50 Si/SiO2 MBE 226


Pd 2.5 15 nm Au, 2.5 nm Cr on laser vaporization 39c 227
glass
Pd 4 15 nm Au, 2.5 nm Cr on MBE 53.4c 227
glass
Pd 10 quartz e-beam evaporation 376
Pd 15 quartz e-beam evaporation 376
Pd 20 quartz e-beam evaporation 376
Pd 10 glass nanoparticle 234
deposition
Pd/C 300 Al2O3 PVD 437
Pd/C 300 SiO2 PVD 438
FeTi+Pd 100 + 20 Si e-beam evaporation 228
Zr+Co 100 + 50 glass ion-beam sputtering 230
Zr+Cu 100 + 50 glass ion-beam sputtering 230
Zr+Fe 100 + 50 glass ion-beam sputtering 230
Zr+Ni 100 + 50 glass ion-beam sputtering 230
Nb+Pd 5 + 20 Al2O3 (112̅0) cathode beam 378

10801
sputtering
Nb+Pd 10 + 20 Al2O3 (112̅0) cathode beam 378
sputtering
Nb+Pd 40 + 20 Al2O3 (112̅0) cathode beam 378
sputtering
Mg+Pd 400 + 25 Si magnetron sputtering 387
Mg+Pd 100 + 50 Al2O3 sputtering 382
Mg+Pd 500 + 80 glass magnetron sputtering 231
Mg+Pd 1020 + 10 glass magnetron sputtering 232
Mg+Pd 100 + 5 Si(100) magnetron sputtering 233
Mg+Pd 200 + 1.5 Si(100) magnetron sputtering 233
Mg+Pd 3000 + 14 glass PVD 5.4 141.6d 73.9d 134.8d 29
Pd+Mg+Pd 20 + 2000 + 20 Si PLD 68c 439
Mg+Pd NPs 25 + 2.5 Si magnetron sputtering 440
Mg+Pd NPs 25 + 5 Si magnetron sputtering 440
Mg+Pd 400 + 25 glass magnetron sputtering 441
Mg+Pd 400 + 25 Al2O3 magnetron sputtering 441
Mg+Pd 400 + 25 LiGaO2 magnetron sputtering 441
Cr+Mg+Pd 6 + 160 + 30 Si magnetron sputtering 61 (1st cycle),d 77 (2nd 442
cycle)d
Mg+Al 1500 + 100 quartz magnetron sputtering 340
Mg+Ti 1500 + 100 quartz magnetron sputtering 340
Review

Mg/NiFeCr+Pd 150×(16.5 + 2.5)+15 glass PVD 4.6 71.6d 80.5d, 66.6d 142d, 123d 42

Chem. Rev. 2018, 118, 10775−10839


DOI: 10.1021/acs.chemrev.8b00313
Table 4. continued
H2 |ΔS|
metal(s)a thickness (nm)b substrate synthesis method wt % Ea (kJ mol−1) |ΔH| (kJ mol−1) (J mol−1 K−1) ref
Mg8.5Cu1.5+Ta+Pd 1500 + 7.5 + 7.5 Si magnetron sputtering 443
Mg8.5Ni1.5+Ta+Pd 1500 + 7.5 + 7.5 Si magnetron sputtering 3.5 79.6 (MgH2)d, 78.6 147.8 443
(Mg2NiH4)d (Mg2NiH4)
Chemical Reviews

Mg8.5Ni1Cu0.5+Ta+Pd 1500 + 7.5 + 7.5 Si magnetron sputtering 3.5 443


MgxNi+Al 1500 + 100 quartz magnetron sputtering 340
MgxNi+Ti 1500 + 100 quartz magnetron sputtering 340
MgxNi1‑x+Pd (0.4 < x < 0.9) 100 + 5 oxidized Si(100) e-beam evaporation 28
Mg2.9Ni 1160 Si(001) magnetron sputtering 4.45 59.5 (Mg2NiH4),c 444
69.5 (MgH2)c
Mg6Ni 30−40 Si magnetron sputtering 280
Mg9.6Fe0.4+Pd (0−500)+6 Al2O3 (0001), Si (001) e-beam evaporation 229
Mg9.6Fe0.4+Pd 600 + 6 Si(001) e-beam evaporation 42.7,c 66.9d 86,c 102d 229
Mg/FeTi+Ta+Pd 15×(85 + 15)+7.5 + 7.5 Si magnetron sputtering 4 383
Mg/FeTi+Ta+Pd 45×(28 + 5)+7.5 + 7.5 Si magnetron sputtering 4.5 79d 140d 383
Ti+Mg+Nb+Pd 10 + 20 + 10 + 10 glass, quartz magnetron sputtering 343
Ti+Mg+Ni+Pd 10 + 20 + 10 + 10 glass, quartz magnetron sputtering 343
Ti+Mg+Pd 10 + 20 + 20 glass, quartz magnetron sputtering 343
Ti+Mg+Pd 10 + 10 + 40 glass, quartz magnetron sputtering 384
Ti+Mg+Pd 10 + 20 + 40 glass, quartz magnetron sputtering 384
Ti+Mg+Pd 10 + 30 + 40 glass, quartz magnetron sputtering 384

10802
Ti+Mg+Pd 10 + 40 + 40 glass, quartz magnetron sputtering 384
Ti+Mg+Ti+Mg+Pd 10 + 20 + 10 + 20 + 10 glass, quartz magnetron sputtering 343
Ti+Mg+Ti+Pd 10 + 20 + 10 + 10 glass, quartz magnetron sputtering 343
Ti+Mg+Ti+Pd 10 + 10 + 10 + 40 glass, quartz magnetron sputtering 384
Ti+Mg+Ti+Pd 10 + 20 + 10 + 40 glass, quartz magnetron sputtering 384
Ti+Mg+Ti+Pd 10 + 30 + 10 + 40 glass, quartz magnetron sputtering 384
Ti+Mg+Ti+Pd 10 + 40 + 10 + 40 glass, quartz magnetron sputtering 384
Ti+Mg+Ti+Pd 10 + 10 + 10 + 10 quartz magnetron sputtering 386
Ti+Mg+Ti+Pd 10 + 10 + 10 + 10 quartz magnetron sputtering 341
Ti+Mg+V+Pd 10 + 20 + 10 + 10 glass, quartz magnetron sputtering 343
Mg+Ti+Pd 1500 + 7.5 + 7.5 Si(100) magnetron sputtering 445
Mg+Fe+Pd 1500 + 7.5 + 7.5 Si(100) magnetron sputtering 445
Mg+Nb+Pd 1500 + 7.5 + 7.5 Si(100) magnetron sputtering 445
Mg+Ta+Pd 1500 + 7.5 + 7.5 Si(100) magnetron sputtering 445
Mg80Ti20+Pd 100 + 5 Si(100) magnetron sputtering 233
Mg80Ti20+Pd 200 + 1.5 Si(100) magnetron sputtering 233
Mg/Cr+Pd 150×(18 + 0.8)+14 glass PVD 6.1 65.7d 73.6d 133.7d 29
Mg/V+Pd 150×(17 + 1.2)+14 glass PVD 5.4 67.6d 73.3d 133.4d 29
Mg+Mg2Si+Pd 200 + 100 + 25 Si magnetron sputtering 387
Mg2Si/Mg+Pd 50×(3 + 5)+25 Si magnetron sputtering 387
Mg2Si/Mg+Pd 25×(6 + 5)+25 Si magnetron sputtering 387
Mg2Si/Mg+Pd 16×(9 + 5)+25 Si magnetron sputtering 387
Review

Chem. Rev. 2018, 118, 10775−10839


DOI: 10.1021/acs.chemrev.8b00313
Mg2Si/Mg+Pd 12×(12 + 5)+25 Si magnetron sputtering 387
Chemical Reviews Review

Metals are listed in order of deposition, with those closest to the substrate first. Alternating layers in a multilayer are denoted with a slash (/), whereas a subsequent layer is denoted with a plus sign (+).
Thickness values correspond to the listed order of metals, separated with a plus sign. Multilayers with alternating compositions have the number of bilayers first, multiplied by the two component
387
387
387
446
ref
(J mol−1 K−1)
|ΔS|

117c

thicknesses in parentheses. A range of thicknesses is denoted with a dash (−). cValue obtained from absorption measurements. dValue obtained from desorption measurements.
|ΔH| (kJ mol−1)

Figure 21. (a) Dark-field mode transmission electron micrographs for


Mg/activated carbon nanocomposites. (b) Energy-dispersive X-ray
spectroscopy (EDX) results for and Mg/activated carbon samples.
(Reprinted from ref 51. Copyright 2007 American Chemical Society.)

Furthermore, the kinetics of the hydrogen release are strongly


influenced by the pore size and thus the final LiBH4 particle
53c

size. The H2 release rate at 300 °C was determined to be 12.5


wt % h−1 for the aerogel with smaller pores compared to 7.8 wt
% h−1 for the aerogel with 25 nm pores (Figure 22). Yan et al.
wt %Ea (kJ mol−1)
H2

4.5
sputtering
sputtering
sputtering
sputtering
synthesis method
magnetron
magnetron
magnetron
magnetron

Figure 22. TGA curves at 10 °C min−1 and dehydrogenation


temperatures for LiBH4 (a) with nonporous graphite (453 °C), (b) in
2 nm activated carbon (375 °C), (c) in 13 nm carbon aerogel (381
°C), and (d) in 25 nm carbon aerogel (390 °C). (Reproduced from
ref 398. Copyright 2008 American Chemical Society.)
substrate

were able to prepare a Mg(BH4)2@aerogel (resorcinol-


formaldehyde) nanocomposite by ball milling a nanocompo-
site of MgH2@aerogel in an B2H6/H2 atmosphere.275 It was
possible to lower the activation energy of the first dehydrogen-
Si
Si
Si
Si

ation step even further to 102 kJ mol−1. The hydrogen


7.5 + 7.5 + 1500 + 7.5 + 7.5

desorption peaked at 160 °C, which is also significantly lower


than that of the bulk material.
thickness (nm)b

Zhang et al. reported the synthesis of LiBH4 NPs deposited


15×(7.5 + 10)+25

in an amorphous mesoporous carbon.399 They achieved 5 nm


100 + 100 + 25

LiBH4 particles homogeneously distributed throughout the


carbon by ball milling of commercially available LiBH4 with
72 + 25

CMK-3. The ordered structure of CMK-3, originally formed by


templating with the SBA-15 silica, was destroyed by the ball
milling process. The final product showed a hydrogen uptake
Pd+Nb+Mg0.75Nb0.25+Nb+Pd

of 16 wt % (in relation to the metal hydride) and a


rehydrogenation capacity of 6 wt %. Furthermore, this
Table 4. continued

approach leads to a decrease in the enthalpy of hydrogen


metal(s)a

desorption by 27 kJ mol−1 H2. Using CMK-3 as a host, Cahen


Mg2Si+Mg+Pd

et al. were able to stabilize 5 nm LiBH4 particles by incipient


Mg2Si/Pd+Pd

solvent impregnation.400 The confinement had a striking effect


Mg2Si+Pd

on the dehydrogenation properties, lowering the desorption


enthalpy to 54.9 kJ mol−1. The combination of nanoconfine-
ment and a catalyst was studied by Shao et al.401 The porous
a
b

10803 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Figure 23. Temperature-programmed desorption of (a) bulk NaAlH4, LiAlH4, and LiBH4 and (b) their composites after wet impregnation.
(Reproduced with permission from ref 136. Copyright 2010 The Royal Society of Chemistry.)

carbon CMK-3 was first loaded with NbF5 and afterward melt mediated by the Lewis donor nitrogen in the porous host,
infiltrated by LiBH4. This combination reduced the onset which increased the dehydrogenation temperature by 25 °C.
temperature of hydrogen desorption to 150 °C (compared to Metal borohydrides can also be incorporated into porous
375 °C for bulk LiBH4) and the activation energy is reduced to carbons through solvent impregnation.267,349,394 For instance,
97.8 kJ mol−1. Moreover, the material can be rehydrogenated Fichtner et al. demonstrated the use of solvent impregnation to
at 200 °C under 60 bar of H2 pressure. These findings were synthesize unsolvated Mg(BH4)2 NPs inside the pores of AC.
ascribed to the particle size and to the presence of active Nb- The nanoscaling of the hydride by confinement led to a
containing species. decrease of the desorption onset temperature. Furthermore,
Liu et al. showed the effect of nanoconfinement in porous distinct changes of the activation energy of the two-step
carbons (4.4 nm pores) on the hydrogen release of desorption process were observed. The first step shows a
Li4BN3H10.402 The bulk material has a hydrogen content of striking decrease of the activation energy from 310.7 kJ mol−1
approximately 11 wt %; however, it also possesses some H2 for bulk Mg(BH4)2, to 176.2 kJ mol−1 H2 for Mg(BH4)2
significant drawbacks, such as the irreversibility of the reaction confined in the activated carbon. In contrast, an increase of the
and the release of NH3 and B2H6 which accompanies activation energy was observed for the second dehydrogen-
dehydrogenation. It was shown that confining this material ation step and a value of 189.5 kJ mol−1 H2 was determined for
in a nanoporous carbon host drastically alters the decom- the nanocomposite, while the bulk material had an activation
position pathway from a three-step to a two-step mechanism. energy of 160.9 kJ mol−1 H2. Comănescu et al. used a similar
The onset temperature of dehydrogenation was lowered by approach for creating Ca(BH4)2 nanostructures with the
surface area and pore size of the carbon supports further
more than 160 °C and furthermore, the NH3 and B2H6
optimized through a chemical activation and a heat treatment
emission was reduced immensely. Another rising star among
step.391 A new activator, KOH, was used instead of the
storage materials is Li3N, which hydrogenates by a two-step
traditional chemical activator, CO2, because physical activation
reaction via Li2NH + LiH to LiNH2 + 2 LiH. The theoretical
of CO2 was shown to randomly expand the pores resulting in a
hydrogen content amounts to 10.4 wt % and the initial heterogeneous structure with various pore sizes.404,405
hydrogen release has its onset at 150 °C. However, the very Carbon nanotubes have also been the subject of some
stable intermediates LiH and Li2NH do not proceed to release nanoconfinement studies. Christian et al. were able to stabilize
hydrogen until a temperature of 430 °C is reached, rendering LiAlH4 NPs of roughly 10−20 nm in size by encapsulating
this class of material unfeasible. Interestingly, Wood and co- them in carbon nanotubes via wet impregnation with a
workers were able to prepare LiH/LiNH2 in situ in the pores precursor solution.136 The encapsulation had a marked effect
of an amorphous nanoporous carbon by infiltration with on the hydrogen release properties, lowering the release
metallic Li dissolved in liquid NH3.19 This confinement had a temperature to 120 °C (Figure 23). Furthermore, the reaction
profound impact on the hydrogen release properties, making pathway was altered by this method to a single step of
the whole reaction fully reversible at 250 °C. Furthermore, hydrogen release, decreasing the activation energy to 64 kJ
DSC data of the nanohydride revealed a single-step reaction mol−1 (instead of 82−115 kJ mol−1 H2 and 86−90 kJ mol−1 H2
pathway. for the first and second dehydrogenation steps, respectively).
Another interesting strategy was shown by Majzoub’s group, In the same study, sodium alanate NPs and LiBH4 particles
reversing the concept of lowering the dehydrogenation were deposited on carbon nanotubes, resulting in a strong
temperature and presenting a system which actually increases decrease of the dehydrogenation activation energy to 88 kJ
the stability of the hydride.403 In their study, nanoporous mol−1 for LiBH4.136 The resulting NaAlH 4 NPs had
carbons doped with pyridinic nitrogen were prepared and dimensions of 10 × 20 nm on average. The encapsulation
infiltrated with LiBH4. The NPs formed a BH3 capping layer, caused a marked decrease of the activation energies for the
10804 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

two-step dehydrogenation from NaAlH4 via Na3AlH6 to NaH. structure, which provides the optimum environment for
The resulting values for the first and second step were 45 and hydride encapsulation, i.e., high potential loading of nanoma-
122 kJ mol−1, respectively. terials and small dead mass (i.e., inactive material mass).
Tan et al. revealed the positive effect of Pd and Pt doped Moreover, graphene-based nanomaterials have shown high
multiwalled carbon nanotubes (MWCNTs) on the dehydro- catalytic activity, stability, thermal and electrical conductivity.
genation properties of LiAlH4.406 By adding 20 wt % of Pd- or Chong et al. successfully demonstrated NaBH4 encapsula-
Pt-doped MWCNTs, the dehydrogenation temperature was tion inside graphene using a solution of NaBH4 in n-
lowered by 28 and 15 °C, respectively, while the amount of propylamine sonicated with graphene.30 The system demon-
hydrogen released (per LiAlH4 unit) was not affected. At strates good reversibility and rapid uptake and release, albeit at
temperatures above 300 °C, when the third dehydrogenation relatively high temperatures. Figure 25 shows TEM images of
step takes place, the formation of Li5PtH3 was observed via in
situ powder X-ray diffraction (PXRD), confirming an alteration
in the high temperature dehydrogenation pathway.
Activated carbon nanofibers are also versatile hosts
successfully used to confine metal hydrides. Using the melt
infiltration method, Lohstroh et al. successfully embedded
NaAlH4 NPs into the pores of CNFs (pore size ranging from
0.5 to 4 nm) under high hydrogen pressure.392 After the
insertion of NaAlH4, the surface area of CNFs was reduced
from 1815 to 242 m2/g and the pore volume decreased from
0.66 to 0.09 cm3/g. To produce different sizes of infiltrated
nanomaterials, Baldé et al. used different drying temperature
conditions and metal hydride loading.50 By decreasing the
drying temperature and initial loading of NaAlH4, the size of
NaAlH4 particles decreased from 1 to 10 μm to 19−30 and 2−
10 nm (Figure 24). The particle growth can be limited by

Figure 25. TEM images of (a) as-synthesized and (b) dehydro-


genated NaBH4@GNs. Low (c) and high (d) magnification images of
rehydrogenated NaBH4@GNs. (Adapted from ref 30. Copyright 2015
John Wiley and Sons.)

NaBH4 NPs with diameters of 2 to 16 nm well-dispersed


throughout the entire graphene host, both before and after
dehydrogenation and rehydrogenation. Dehydrogenated mate-
rials show similar sizes as that of the as-synthesized one;
however, the average size of the rehydrogenated one was
increased to 45 nm. Even though the initial morphology of
Figure 24. Kissinger plots for NaAlH4-impregnated CNFs with NaBH4 nanomaterials is not well retained by graphene, this
particles of three different sizes. (Reprinted from ref 50. Copyright approach shows that nanoconfinement with graphene is able to
2008 American Chemical Society.) realize improved hydrogen sorption performance. NaBH4
nanostructured with graphitic nanosheets (NaBH4@GNs)
using a lower concentration, and the viscosity of the solvent has also been prepared from the mechanical milling method
increases at lower temperatures, which yields a greater by Li et al.408 The nanostructure exhibits exceptional stability,
dispersion. de Jongh and co-workers showed the possibility in which the size and morphology are well sustained even after
of substantially lowering the desorption temperatures by 5 cycles of uptake/release. However, residual contaminants
impregnation of CNFs with well dispersed, nanosized such as LiCl produced during the synthesis hamper the
NaAlH4 (9 wt % hydride on carbon).407 Remarkably, it was hydrogen storage ability. Stable LiH was generated during the
possible to render the process reversible, something that was dehydrogenation and consequently decreased the H2 capacity
not observed before for bulk alanates in the absence of and disrupted reversibility. In order to restrict the generation
additives. The sample reabsorbed 26% of its initial hydrogen of stable LiH, MgH2 was added to react with LiCl
content after treatment for 48 h under 90 bar H2. subsequently produce Li3Mg12. Ultimately, a high cyclic
A continuing effort in the development of new nanoparticle capacity is exhibited in the new NaBH4@GNs. Graphene is
systems has employed graphene and its derivatives as a promising candidate for the wrapping matrix with reducing
encapsulating materials. The developments have in turn been the size of metal hydride materials in nanoscale as proven in
driven by the distinct advantages of graphene because it avoids the experimental results.
some of the inevitable disadvantages of other approaches for Gasnier et al. used entangled graphenes, which are prepared
hydrogen storage. Graphene, consisting of a single atomic layer via pyrolysis of a graphene hydrogel.409 The porous graphene
of carbon, has a large surface area and a lightweight and open aerogels was melt infiltrated with LiBH4 at 300 °C under 30
10805 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

bar H2. The pore size, ranging from 6.1 to 26 nm, was reversible capacity is observed. Nanoconfinement of NaAlH4
controlled by the precursor concentration, and the total pore leads to an alteration of the reaction pathway, resulting in a
volume increased by incorporating graphene rather than one-step dehydrogenation process, suppressing the formation
carbonizing precursors, which resulted in the higher loading of the Na3AlH6 intermediate. More recently, other MOF
of LiBH4 nanomaterials in the pores. A similar wrapping structures, including MIL-100(Al), have also been used in
material for the nanoencapsulation of metal hydrides is the confining a metal hydride for yielding well-defined NPs with
functionalized graphene called graphene oxide (GO), which improved H2 sorption properties.412,413
contains oxygen moieties (carboxyl and hydroxyl groups) or Silica can also serve as a nanoscale template which allows
reduced graphene oxide (rGO). Cho et al. first reported the well-tailored size and shape of metal hydride nanomateri-
facile solution-based coreduction method for Mg NPs als.272,414,415 Metal hydrides, such as NaBH4, NaAlH4, and
encapsulated by rGO nanosheets using Cp2Mg and GO LiBH4, infiltrated into the nanoscale channels of mesoporous
precursors in THF.34 The TEM analysis showed monodisperse silica SBA-15 show improved desorption properties through
Mg NPs with diameters of 3.25 ± 0.87 nm. rGO-Mg destabilization.414,415 However, the presence of silanol (Si−
multilaminates exhibit high hydrogen storage capacity and OH) groups in the pore channels of SBA-15 caused the metal
cyclability because Mg NPs were well-encapsulated by rGO hydride to be partially oxidized during the infiltration process.
and consequently protected from oxidation. In addition, Pd/ To prevent oxidation, Zheng et al. utilized SBA-15 function-
rGO nanocomposites have also been prepared via a solution- alized with methyl groups, which shows the enhanced
based method and electrochemical method from the reduction hydrogen storage ability of NaAlH4 (Figure 26).272
of the GO and Pd precursors as well.410
Carbon-based porous matrices confining nanomaterials have
improved the performance of hydrogen sorption because they
are lightweight and act as catalysts. However, inhomogeneous
pore sizes with irregular morphologies in carbon-based porous
matrixes have inspired the creation of hosts with well-defined
pores with narrow size distributions.
5.3.2. Other Rigid Porous Hosts. In addition to pure
carbon materials, MOFs and silicas have seen increasing use in
the nanoconfinement of hydride materials. MOFs are attractive
porous supports for the synthesis of metal hydride nanoma-
terials because their well-ordered crystalline structure and
monodisperse pore dimensions provide a highly controlled
cavity size and chemical functionality. Fischer et al. were the
first to prepare transition-metal nanoclusters inside MOFs by
employing organometallic precursors.411 This approach was
extended to the preparation of metal hydrides with dimensions
of a few nanometers.18,412,413 Bhakta et al. were first to
infiltrate NaAlH4 inside the pores of the MOF HKUST-1
(Cu3BTC2 with BTC3− = benzenetricarboxylate).413 Remark-
ably, the dehydrogenation temperature was substantially
lowered, and hydrogen evolution from the sample was
observed starting at 70 °C. In a follow up study, the kinetics
and thermodynamics, as well as the pathway of the
dehydrogenation process for these nanoconfined hydrides
were unveiled.17 Notably, the two-step reaction from NaAlH4 Figure 26. (a) High resolution TEM micrograph of NaAlH4/OMS
to NaH showed enthalpies of 47.3 kJ mol−1 H2 and 45.6 kJ (ordered mesoporous silica). (b) Dehydrogenation over time for
mol−1 H2 for the nanoconfined hydride, compared to 40.9 and pristine NaAlH4 (black) and for the NaAlH4/OMS composite (blue)
15.6 kJ mol−1 H2 for the bulk material. The activation energy at 150 and 180 °C. (Adapted with permission from ref 272. Copyright
for this process amounted to 53 kJ mol−1 (compared to 118 2008 American Chemical Society.)
and 120 kJ mol−1 for the individual reaction steps of the bulk
material). Core−shell architectures have also been used to encapsulate
Stavila et al. investigated the infiltration into another complex metal hydrides. Preparation of these systems is more
framework, namely CPO-27(Mg) (also known as MOF- complicated because of the extreme reactivity of metal
74(Mg), Mg2(dobdc) with dobdc4− = 2,5-dioxido-1,4- borohydrides, amides, and alanates. Christian et al. first used
benzenedicarboxylate) in the presence of catalytic amounts the liquid-phase method for the synthesis of core−shell NaBH4
of titanium chloride.18 The framework possesses 1D channels NPs, with the Ni shell acting as both a stabilizer and a
lined with open metal sites that are capable of stabilizing the catalyst.371 In addition, the Ni shell prevents agglomeration of
titanium catalyst. After loading the pores with NaAlH4 by melt NaBH4 NPs and evaporation of Na or NaH during the
infiltration, 1.2 nm metal hydride particles are obtained. The desorption step. As a result, the core−shell system shows
activation energy of the dehydrogenation is lowered to 57.4 kJ reversibility under relatively mild conditions and a stable core−
mol−1 and the desorption onset was already observed at 50 °C. shell structure up to 350 °C. However, at high desorption
Remarkably, it is possible to rehydrogenate the material under temperatures (>350 °C), a distinct change in morphology was
mild conditions (10.4 MPa H2 , 160 °C); after four observed due to sintering of the Ni shell. Various other metals
dehydrogenation/rehydrogenation cycles, a 15% loss in can be used as a shell (Co, Cu, Fe, Ni, and Sn) in order to
10806 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

retain the shape of NaBH4 core during desorption.416 The Sn preparation of nanomaterials has been also studied. Table 5
shell (melting point = 232 °C) fell apart at 300 °C, the Ni shell contains details about synthesis and hydrogen storage
was unstable at 600 °C, and the remaining three metals properties of nanoscaffolded and nanoencapsulated metal
maintained their structures even at the higher temperature hydride NPs.
(Figure 27). Wang et al. expanded this approach to LiAlH4@Ti
and showed significant improvements in hydrogen storage 6. MECHANISTIC EFFECTS OF NANOSIZING
properties of LiAlH4.417 Across a vast range of applications, nanomaterials exhibit their
own unique chemical and physical properties compared to
their bulk counterparts.469 It is well known that various kinds
of physicochemical properties can be tuned by manipulating
material dimensions. As to metal hydrides for hydrogen
storage, many studies have reported that their thermodynamics
and kinetics can be improved via nanostructuring. Nanoscale
metal hydrides possess high surface areas, leading to an
increased surface energy, which is expected to improve
hydrogen sorption thermodynamic and kinetic characteristics.
Also, nanostructuring plays a key role in generating and
enhancing the effects of controlled interfaces between different
material phases, which can induce certain forms of lattice
deformation or distortion that can be beneficial for cycling.
Furthermore, confinement effects can fundamentally alter
materials properties, either through chemical or electronic
interactions with the confining medium or else through
mechanical confinement stress.
Potential effects of nanostructuring or nanoconfinement on
the thermodynamics of metal hydrides are usually represented
in terms of the enthalpy change upon cycling. For most metal
hydrides, ΔH is negative (exothermic) upon hydrogenation,
indicating the relative stability of the hydrided form, with the
notable exception being AlH3. Furthermore, in complex metal
hydrides and most light metal hydrides, the magnitude of the
reaction enthalpy tends to be too large, which in turn signifies
impractically high operating temperatures. This fundamentally
originates from the high lattice enthalpy and strong ionic and
covalent bonding that characterizes these materials classes.14
Accordingly, one of the primary goals of nanosizing and
nanoconfinement is to destabilize the hydride, thereby
reducing the magnitude of the reaction enthalpy and
facilitating practical use within a reasonable range of operating
temperature. For complex metal hydrides, the reaction entropy
may be altered in addition to the reaction enthalpy, providing a
second avenue for thermodynamic tuning.
Likewise problematic are the often sluggish hydrogenation
and dehydrogenation kinetics of metal hydrides. Poor kinetics
may be associated with slow diffusion, chemical bond
activation, nucleation of reaction products, or formation of
stable intermediate phases in multistage reactions. As a result,
Figure 27. TEM micrographs of the core−shell structures in the as- cycling of metal hydrides often requires hydrogen release
synthesized (left), 300 °C annealed (center), and 600 °C annealed temperatures and uptake pressures far in excess of what is
(right) particles: (a) NaBH4@Co, (b) NaBH4@Cu, (c) NaBH4@Fe, predicted thermodynamically. Nanoscale and nanoconfined
(d) NaBH4@Ni, and (e) NaBH4@Sn. (f) Temperature-programmed
metal hydrides can offer significant advantages in this regard by
desorption of the core−shell NPs at a rate of 10 °C min−1 monitored
by mass spectrometry. (Adapted with permission from ref 416. reducing diffusion lengths, facilitating chemical kinetics and
Copyright 2013 Royal Society of Chemistry.) nucleation or altering reaction pathways.
Review articles by Berube et al.,11 Pundt and Kirchheim,470
Pundt,471 de Jongh and Adelhelm,45 Vajo,47 and Wang472
These porous matrixes are able to decrease the particle size systematically surveyed earlier efforts exploring the effects of
of metal hydrides to less than 10 nm and prohibit the nanosizing, nanoconfinement, and microstructure on the
aggregation into larger particles. However, there are a few interactions between hydrogen and solid-state materials.
disadvantages, including the decrease in gravimetric and These articles comprehensively documented various effects of
volumetric hydrogen densities, because this approach has surfaces, phase boundaries, grain boundaries, dislocations,
thus far exhibited only limited loading of metal hydride into vacancies, and other structural defects at the nanoscale. In the
the dead weight of inert supports. Thus, the development of following sections, we review some of their conclusions, while
new synthetic methods diminishing the usage of supports for focusing heavily on more recently reported/proposed mech-
10807 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Table 5. Hydrogen Storage Properties of Nanoconfined Metal Hydrides
Vtot Ea |ΔH| |ΔS|
hydride porous host Davg (nm) (cm3 g−1) infiltration method solvent or atmosphere loading/wt % H2 wt % (kJ mol−1) (kJ mol−1) (J mol−1 K−1) ref
PdHx CMK-3-template 3.5 1.14 reduction of PdCl4 500 °C, Ar or 300 °C, H2/Ar 10 2.8−3 447
PdHx CMK-3-CVD 6 1.13 reduction of PdCl4 500 °C, Ar or 300 °C, H2/Ar 10 1.4−1.6 447
PdHx CMK-3 1.1 solvent impregnation 300 °C, H2, Ar 17.2 1.45 448
Chemical Reviews

PdHx MIL-100(Al) 0.87 0.65 solvent impregnation 280 °C, H2, Ar 10 3.1b 412
PdHx MIL-101(Cr) 1.2−1.6 1.64 solvent impregnation pentane/H2O 20 449
PdHx COF-102 1.2 1.35 chemical vapor infiltration Ar 9.5 0.4b 450
Mg6Pd HSAG 2−3 0.69 solvent impregnation/Melt 650 °C, Ar 18.5 1.7 451
infiltration
MgH2 Ni-CMK-3 2−3 0.66 MgBu2 infiltration 200 °C, 80 bar H2 37.5 107.6 274
MgH2 RF-aerogel 22 1.27 MgBu2 170−200 °C, 50−78 bar H2, 18.2 1.14 259
Ar, heptane
MgH2 RF-aerogel 7 0.65 MgBu2 170−200 °C, 50−78 bar H2, 10 0.62 259
Ar, heptane
MgH2 RF-aerogel 13 0.8 MgBu2 Ar 14.8 252
MgH2 activated carbon 0.5−3 0.79 MgBu2 Ar 22 142.8 63.8 117.2 273
fiber
MgH2 RF-aerogel(Ni) 13 0.8 melt infiltration Ar 9.6 452
MgH2 RF-aerogel(Cu) 13 0.8 melt infiltration Ar 12.7 452
MgH2 RF-aerogel 13 0.8 melt infiltration Ar 3.6 452
MgH2 HSAG 2−3 0.65 melt infiltration 625 °C, H2, Ar 10−15 51
MgH2 activated carbon <2 0.45 melt infiltration 625 °C, H2, Ar 15−20 51

10808
MgH2 SNU-90 1.3 1.64 chemical vapor infiltration 30 bar H2 200 °C 10.5 0.71b 453
MgH2 RF-Aerogel 16.6 1.32 MgBu2 150 °C, 150 bar H2 24.8 1.8 454
MgH2 RF-Aerogel 27.1 1.32 MgBu2 150 °C, 150 bar H2 24.3 1.3 454
MgH2 RF-Aerogel 14.7 1.85 MgBu2 150 °C, 150 bar H2 37.1 3.1 454
MgH2 RF-Aerogel 25 1.89 MgBu2 150 °C, 150 bar H2 40.2 2.2 454
MgH2 Ni nanobelt MgBu2 100 °C, 60 bar H2 32 34.4 76.9 69.2 455
MgH2 Ni-doped hollow 3.4 1.53 MgBu2 NH2NH2·H2O 60 6.63 46.9,c 264
carbon −49.1
AlH3 HSAG 2−3 0.66 solvent impregnation Et2O 14.4 0.25 61
LiBH4MgH2 0.13 TiCl/RF- 26 1.3 grinding/melt infiltration 310 °C, 60 bar H2 33.3 3.6 456
aerogel
LiBH4MgH2 RF-Aerogel 2.2 1.27 melt infiltration 50 bar H2,heptane, Ar 35 11.5a 397
LiBH4 CMK-3 4.5 1.3 ball milling 3 bar H2 “50” 14 40 399
LiBH4 pyrolyzed aerogel 13 0.8 melt infiltration 280−300 °C, Ar 25−30 12.5 103 398
LiBH4 pyrolyzed aerogel 25 1.38 melt infiltration 280−300 °C, Ar 45−50 7.8 111 398
LiBH4 CMK-3 4.5 1.3 solvent impregnation MTBE 33 3.4 400
LiBH4 HSAG 2−3 0.65 melt infiltration 295 °C, H2 150 bar 25 1.45 390
LiBH4−Ca(BH4)2 RF-aerogel 30 1.21 melt infiltration 210 °C, 110−130 bar H2 38.4 6.23 156 268
LiBH4−Ca(BH4)2 RF-aerogel 30 3.13 melt infiltration 210 °C, 110−130 bar H2 64.9 7.34 130 268
LiBH4 Activated carbon 1.75−3.2 0.87 solvent impregnation THF 28.4−30 9 457
LiBH4 NbOF5−CMK-3 3.8 1.25 melt infiltration 300 °C, 140 bar H2 6.52a 97.8 401
LiBH4 CNT solvent impregnation THF 1.5 0.27 88 136
Review

HSAG 2−3 0.66 melt infiltration 30 18.1a 458

Chem. Rev. 2018, 118, 10775−10839


DOI: 10.1021/acs.chemrev.8b00313
LiBH4 295 °C, 100 bar H2
Table 5. continued
Vtot Ea |ΔH| |ΔS|
hydride porous host Davg (nm) (cm3 g−1) infiltration method solvent or atmosphere loading/wt % H2 wt % (kJ mol−1) (kJ mol−1) (J mol−1 K−1) ref
LiBH4 SBA-15 4.7−6.5 0.35−0.83 solvent impregnation MTBE 33−50 414
LiBH4Ca(BH4)2 CMK-3 3.5 1.63 ball milling + melt infiltration 200 °C, 3 bar H2 11a 265
2LiBH4−NaAlH4 RF-Aerogel 30 1.21 melt infiltration 310 °C, 110 bar H2 25.3 9.52a 459
Chemical Reviews

LiBH4−LiAlH4 carbon nanofiber 3 1.2 sequential solvent impregnation/ Et2O/310 °C, 110 bar H2 66 6.8a 117 266
melt infiltration
Li/LiH/LiBH4 HSAG 2−3 0.65 melt infiltration 200 °C, 100 bar H2 20 460
LiAlH4 CNT solvent impregnation THF 1.2 0.19 64 136
LiAlH4 HSAG 2−3 0.66 solvent impregnation THF 10a 458
LiAlH4 HSAG 2−3 0.66 solvent impregnation THF 18.9 0.6 461
LiAlH4 TiO2/Porous 4.55 0.28 solvent impregnation THF 37 6.2 47.1 462
carbon
NaH HSAG 2−3 0.66 physical mixing Ar 33 3.8−4.4a 463
NaBH4 graphite (300 ball milling LiBH4, NaCl 6.8 408
mesh)
NaBH4 HSAG 2−3 0.66 solvent impregnation H2O 25 6.7a 389
NaBH4 HSAG 2−3 0.66 melt infiltration 520 °C, 5 bar H2 25 7.9a 389
NaAlH4 RF-aerogel 13 0.8 melt infiltration 189 °C, 183 bar H2 48.6 5.5 396
NaAlH4 HSAG 2−3 0.65 melt infiltration H2 20 7.4a 388
NaAlH4 RF-aerogel 10 0.91 melt infiltration 189 °C, 210 bar H2 34.1 1.3 16
NaAlH4 RF-aerogel 11 1.3 melt infiltration 189 °C, 210 bar H2 41.5 1.9 16
NaAlH4 RF-aerogel 10 2.21 melt infiltration 189 °C, 210 bar H2 51.3 2.4 16

10809
NaAlH4 RF-aerogel 8 2.11 melt infiltration 189 °C, 210 bar H2 56.5 2.7 16
NaAlH4 TiCl3/RF-aerogel 8 1.93 melt infiltration 189 °C, 210 bar H2 48.6 2.1 16
NaAlH4 HSAG 2−3 0.65 melt infiltration H2, 180−190 bar, 180 °C 5−80 6.3 395
NaAlH4 carbon nanofiber 0.36 solvent impregnation THF 9 2.84 407
NaAlH4 carbon nanofiber 0.36 solvent impregnation THF 2 58 50
NaAlH4 RF-aerogel 4 0.36 melt infiltration 189 °C, 210 bar H2 12.8 257
NaAlH4 RF-aerogel 7 0.69 melt infiltration 189 °C, 210 bar H2 27.3 257
NaAlH4 RF-aerogel 10 0.91 melt infiltration 189 °C, 210 bar H2 34.1 257
NaAlH4 RF-aerogel 13 0.86 melt infiltration 189 °C, 210 bar H2 31.8 257
NaAlH4 RF-aerogel 19 1.33 melt infiltration 189 °C, 210 bar H2 41.9 257
NaAlH4 RF-aerogel 22 1.39 melt infiltration 189 °C, 210 bar H2 43 257
NaAlH4 RF-aerogel 26 1.3 melt infiltration 189 °C, 210 bar H2 41.4 257
NaAlH4 RF-aerogel 39 1.09 melt infiltration 189 °C, 210 bar H2 37.2 257
NaAlH4 RF-aerogel >100 0.26 melt infiltration 189 °C, 210 bar H2 32 257
NaAlH4 CNT solvent impregnation THF 5.68 0.45 45 136
NaAlH4 Ti/Carbon 0.36 solvent impregnation THF 8 145
nanofiber
NaAlH4 HKUST-1 1.07 0.41 solvent impregnation THF 4 413
NaAlH4 HKUST-1 1.07 0.41 solvent impregnation THF 4 53 47.3 17
NaAlH4 MOF-74(Mg) 1.2 0.593 melt infiltration 195 °C, 250 barH2 21 4.5a 57.4 18
NaAlH4 SBA-15 10 solvent impregnation THF 20 4.8 272
Review

Chem. Rev. 2018, 118, 10775−10839


DOI: 10.1021/acs.chemrev.8b00313
Table 5. continued
Vtot Ea |ΔH| |ΔS|
hydride porous host Davg (nm) (cm3 g−1) infiltration method solvent or atmosphere loading/wt % H2 wt % (kJ mol−1) (kJ mol−1) (J mol−1 K−1) ref
Na/NaH/ HSAG 2−3 0.65 melt infiltration 180 °C, 120 bar H2 20 460
NaAlH4
Mg(BH4)2 activated carbon 0.61 solvent impregnation Et2O 44 176.2; 349
189.5c
Chemical Reviews

Mg(BH4)2 RF-aerogel 10 0.52 ball milling B2H6/H2 6 (27% 102 275


MgB12H12)
Mg(BH4)2 Ni-RF-Aerogel 9 0.5 melt infiltration (MgH2) 120 °C, B2H6/H2 10 464
Mg(BH4)2 Ni-CMK-1 4.80 0.86 solvent impregnation THF 7.21a 21.3 465
Mg(BH4)2 Cu2S hollow 20 0.128 solvent impregnation THF 289
spheres
Ca(BH4)2 Cu2S hollow 20 0.128 solvent impregnation THF 289
spheres
Mg(BH4)2 • 6 activated carbon 2 1.1 solvent impregnation NH3 50 9.3 466
NH3 (Mg(BH4)2·Et2O))
NH3BH3 CMK-3 4.5 1.29 solvent impregnation MeOH 50 2.1 467
NH3BH3 CMK-3-Li 4.5 1.29 solvent impregnation MeOH 50 3.5 98 467
NH3BH3 RF-carbon-cryogel 9 0.7 solvent impregnation THF 24 9a 120 468
NH3BH3 MCM-41 4 solvent impregnation THF 33−75 468
NH3BH3 SBA-15 7.5 1.2 solvent impregnation MeOH 50 67 271
a
Hydrogen desorption values are normalized as weight percent of the infiltrated hydride. bHydrogen content is determined by absorption measurements on the corresponding metal/host complex.
c
Enthalpies for two consecutive reaction steps occurring at different temperatures.

10810
Review

Chem. Rev. 2018, 118, 10775−10839


DOI: 10.1021/acs.chemrev.8b00313
Chemical Reviews Review

anisms by which thermodynamics and kinetics are altered 0.55 J m−2 for Mg.473 Wagemans et al. studied the surface-
upon nanosizing or nanoconfinement of metal hydrides. induced destabilization effect in MgH2 using DFT calculations
We begin by discussing the effects of surfaces, which play an on clusters, predicting a significant reduction of reaction
increasingly dominant role for nanoscale systems. Next, we enthalpy (>10%) for clusters below 2 nm in size due to surface
give an overview of effects associated with internal interfaces in effects alone.14 These results generally agreed with the later
metal hydride NPs and nanostructures, including the roles of DFT calculations of Kim et al.,301 as well as direct
grain boundaries and internal phase boundaries. Third, we hydrogenation simulations by Cheung et al. using the ReaxFF
introduce some additional effects of nanosizing on kinetics, method.15 Experimentally reported values for reaction enthalpy
including diffusion and plateau pressure behavior. Finally, we reduction in MgH2 upon nanosizing show a somewhat greater
discuss potential effects associated with the confining medium, variation. Paskevicius et al. reported a reduction of <4% for
including hydride-host interactions. Particular emphasis is MgH2 particles of ∼7 nm, similar to the predicted value for
given to experimental and theoretical studies designed to probe that size.186 However, experiments performed by Zhao-Karger
specific mechanisms associated with nanosizing and nano- et al. on nanconfined MgH2 particles with <3 nm diameter
confinement. found much greater reductions, up to twice the predicted value
6.1. Effects of Surfaces based on surface effects alone.273 A possible explanation for
this discrepancy may lie in additional enthalpic contributions
Among the most obvious thermodynamic implications of from internal interfaces or confinement stress, as discussed in
particle size is the high surface energy introduced by sections 6.2 and 6.4.
unsatisfied chemical bonds. At the nanoscale, the relative It is important to note that surface-induced thermodynamic
contribution of surface energies to the reaction enthalpy changes can be activated only if very small particles are
becomes increasingly large due to the high surface-to-volume produced without agglomeration, which tends to reduce
ratio. surface area during synthesis and cycling. Even within a
6.1.1. Effects on Reaction Enthalpy. Certain simple confining medium, particles are likely to interact with one
binary metal hydrides have higher surface energies than their another, particularly as they undergo dramatic changes in
dehydrided phases, which allows some portion of the enthalpy volume and morphology during cycling. Likewise, catalyst
of formation to be stored as an excess surface energy and additives or synthesis byproducts (e.g., solvents) may also alter
contributes to decreasing the magnitude of ΔH. Kim et al. used effective surface energies via additive-particle interactions.
DFT to compute the reaction enthalpy for NPs of several Berube et al.473 explored the reduction of effective surface
archetypical binary metal hydrides.301 The authors mixed bulk areas due to agglomeration by considering the van der Waals
enthalpy calculations with surface energies, invoking a Wulff interactions between MgH2 NPs. In noninteracting ideal MgH2
construction to generate realistic nanoparticle shapes and particles, the reaction enthalpy shift increases monotonically
properly weight energies of different surfaces. Among their with smaller particle size, exceeding 5 kJ mol−1 for particles
tested binary simple metal hydrides (VH2, LiH, ScH2, TiH2, smaller than ∼15 nm radius (compared with the 75 kJ mol−1
AlH3, MgH2, and NaH), only MgH2 and NaH exhibited a reaction enthalpy of bulk MgH2). However, at the same time,
reduction in the magnitude of the reaction enthalpy upon the authors suggested that the effective surface area
nanosizing (Figure 28). In all other cases, the dehydrided exponentially decreases with smaller particle sizes due to van
phases had higher surface energies and led to significant der Waals-induced particle clustering. Considering these two
increases in the magnitude of the reaction enthalpy. competing factors, the enthalpy shift due to surface energies
Perhaps for this reason, much of the effort in surface- has a maximum at ∼7 nm radius, and never exceeds 4 kJ mol−1.
induced hydride destabilization has focused on MgH2, which It is worth noting that not only is there an optimum particle
has an average surface energy of 2.08 J m−2 compared with size to achieve the maximal enthalpy shift, but the magnitude
of energy contribution by particle surface energies is also
constrained to <5−6% of the total reaction enthalpy. Such a
small contribution indicates that even for cases like MgH2 for
which the surface energy ordering favors hydride destabiliza-
tion at the nanoscale, this factor alone may not appreciably
manipulate the hydrogen storage thermodynamics. Never-
theless, it may be possible to mitigate particle clustering within
a properly structured confining medium.
In contrast to MgH2, most complex metal hydrides are
loosely packed molecular salts. These compounds typically
have rather low surface energies, often far below that of their
denser dehydrogenation products. For instance, the average
surface energy of NaAlH4 is only 0.06−0.15 J m−2, far lower
than those of the dehydrogenated products NaH and Al
(∼0.20 and 0.74 J m−2, respectively).359,474,475 This implies
that surface energy contributions within nanoscale NaAlH4
should increase the equilibrium desorption temperature for the
full reaction, contrary to the preferred outcome.
Figure 28. Variation in the enthalpy of the metal/metal hydride Such undesirable surface energy effects on enthalpy have
reaction as a function of metal particle size for seven different metal been generalized to other complex metal hydrides. Hazrati et
hydrides. (Reproduced with permission from ref 301. Copyright 2009 al.476 computed the energy of (LiBH4)n, (Li)n, (B)n, (LiB)n,
IOP Publishing.) and (LiH)n (n = 2 to 12) nanoclusters using ab initio
10811 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

calculations and found that the total energies of the fully equilibrium temperature required to achieve full dehydrogen-
hydrided (LiBH4)n nanoclusters are less influenced by the ation could be reduced to as low as 65 °C for these particles by
particle size n compared to the other phases. In fact, the eliminating Na3AlH6 as a thermodynamic sink. Their argument
stability of (LiBH4)n nanoclusters was comparable to the bulk was based upon comparing the enthalpy of NPs of each
material, inferring a relatively low cohesive energy and a isolated phase for different particle sizes and considering the
negligible surface energy penalty associated with cluster size. thermodynamically preferred products at different temper-
The lower surface energies of complex metal hydrides relative atures.
to their binary metal hydride counterparts likely complicates Experimentally, Gao et al. also concluded that nanoconfine-
strategies for nanoconfinement-induced destabilization of such ment induced apparent single-stage decomposition in
materials via surface energy effects alone. NaAlH4.395 Using thermogravimetric analysis (TGA), they
Another factor that must be considered in determining the found only a single desorption feature in the uncatalyzed
impact of surface energy penalties on reaction enthalpy is material when confined within 2−3 nm pores. To confirm the
possible competition from other polymorphs of the hydrided thermodynamic origin of the behavior, they further hydro-
or dehydrided phase. An example of this phenomenon is the genated the samples under conditions that were within the
stabilization of β-Li3N over the more common α-Li3N variant expected stability range of the Na3AlH6 intermediate, based on
at the nanoscale, which has been attributed to different surface the thermodynamic equilibrium of the bulk material. The
energy contributions for the two phases.19 Such competition nanoconfined samples were than analyzed with 27Al NMR,
between polymorphs in complex metal hydrides will generally which confirmed the lack of any Na3AlH6-like species. Work by
tend to mitigate changes in reaction enthalpy. Lohstroh et al. likewise found possible evidence of nano-
6.1.2. Effects on Reaction Pathways. Surface energies confinement-induced single-stage decomposition from PCT
can also alter the relative stability of various intermediate analysis of NaAlH4 decomposition in 4−5 nm pores.392 In
phases that appear during cycling of nanoscale particles. Such addition, several authors observed suppression of solid−solid
intermediates play an especially critical role in the kinetics of phase transitions and melting of LiBH4 when confined in
many complex metal hydrides, which often feature multistage nanopores.136,477 These studies also found that the undesirable
decomposition pathways. Stable intermediates can be evolution of B2H6 was curtailed. Similar transitions to single-
thermodynamic sinks, making further reaction difficult except stage pathways upon nanoconfinement have been observed in
under the most extreme driving forces. Likewise, nucleation of other metal hydride systems, including LiBH4/MgH2 mix-
intermediates can incur unacceptable kinetic penalties. tures.47
Under the most favorable circumstances, surface effects can 6.1.3. Surface Entropy and Disorder. Unlike their
destabilize otherwise stable intermediates to the point that they influence on the enthalpy of nanoscale hydrides and
will no longer form at all. In this event, the reaction pathway is intermediates, the impact of surfaces on entropy has been far
fundamentally altered, often with significant enhancements in less explored. Nevertheless, it is reasonable that the additional
the overall reaction kinetics. Mueller and Ceder used DFT disorder, dynamics, and free volume introduced within surface
calculations and a cluster expansion model to show that surface regions may provide a significant entropic gain compared to
effects in nanoscale NaAlH4 could alter the dehydrogenation bulk regions. Indeed, there is some evidence to suggest that
reaction pathway from two-stage decomposition to single-stage nanosizing can alter the reaction entropy. Using van’t Hoff
decomposition, bypassing formation of the Na3AlH6 inter- analysis, Zhao-Karger et al. determined that the decomposition
mediate phase altogether for particles <52 nm in diameter of MgH2 in <3 nm pores exhibited a reduction in the reaction
(Figure 29).359 At the same time, they predicted that the entropy with respect to the bulk by 19 J mol−1 K−1.273 Another
study by Paskevicius et al. on ∼7 nm MgH2 nanocrystals
reported a similar effect, with a more modest entropic
reduction of around 4 J mol−1 K−1.186 Analysis by Cho et al.
showed a reduction of 4−13 J mol−1 K−1 for MgH2 in 3−4 nm
pores in reduced graphene oxide.35 In each of these instances,
size reduction also resulted in reaction entropy decrease, with
the largest reductions for the smallest particles. As a result,
these studies had the unintended effect of further stabilizing
rather than destabilizing the hydrided form. Although the
authors did not identify the origin of entropy changes, surface
dynamics and disorder are some of the likely causes. If
nanosizing is to be used as a practical avenue for reducing the
reaction enthalpy of metal hydrides, any accompanying
reduction in entropy should also be curtailed.
In complex metal hydrides, reorientation of molecular ions
can provide an additional potential entropic contribu-
tion.300,478,479 These reorientations are often quite hindered
in the bulk, but are likely to be much freer at surfaces. Beyond
the thermodynamic implications, the dynamical disorder may
Figure 29. Phase diagram of the NaAlH4 (region A)-Na3AlH6+Al
(region B)-NaH+Al (region C) system as a function of NaAlH4 aid kinetic processes, including mass transport, chemical bond-
particle size and temperature, showing the possibility of direct breaking, and phase decomposition kinetics. Using ab initio
conversion from NaAlH4 to NaH+Al at sizes smaller than ∼50 nm. molecular dynamics simulations, Wood and Marzari reported a
(Reproduced from ref 359. Copyright 2010 American Chemical phase transition in (001) surface slabs of NaAlH4 from the
Society.) conventional tetragonal phase to a new orthorhombic phase
10812 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

above 320 K.143 The authors concluded that this new phase is able to investigate phase transformations in individual cubes of
stabilized by entropic contributions from rotationally disor- particular sizes. In conjunction with statistical mechanical
dered AlH4− polyhedra. They further suggested that, because simulations based on an Ising model, the authors claimed to
the transition is accompanied by a significant volume isolate the intrinsic size scaling laws on the thermodynamics
expansion to facilitate AlH4− rotation, it may be difficult to and kinetics of hydriding and dehydriding phase trans-
nucleate in the bulk but should be more easily accommodated formations in NPs. This size scaling behavior was suggested
at surfaces. The new orthorhombic phase was also determined to arise from the thermal fluctuation-driven hydride nucleation
to have a lower surface energy than the conventional tetragonal that preferentially occurs at the particle surfaces. As a result,
phase, suggesting preferential formation in nanoscale systems. decreasing nanocrystal size and increasing temperature would
It was subsequently determined that the larger volume also diminish the nucleation barrier, leading to the reduction of
significantly lowers the barrier for defect mobility, implying it hysteresis of plateau pressures and enhanced phase trans-
may be relevant for facilitating hydrogen diffusion kinetics formation kinetics.
during dehydrogenation.480 Beyond atomic rearrangement, surfaces can also facilitate the
Additional work by Majzoub et al.304 based on the prototype creation of vacancies, which act as another source of
electrostatic ground state method303 demonstrated that the undercoordination that can facilitate chemical bond breaking.
orthorhombic phase found by Wood and Marzari is only one Theoretical work by Du et al. showed that the B−H bonds
example of several competing NaAlH4 polymorphs with very within the BH4− unit next to a Li monovacancy or divacancy
similar free energies of formation. The authors concluded that on the LiBH4 (010) surface become elongated due to charge
many of these phases are likewise stabilized by entropic factors renormalization.484 Based on a nudged elastic band calculation,
associated with rotation of AlH4− polyhedra. they found that the energy barrier for H2 release is reduced
Another example of the possible role of surfaces in from 3.64 eV for the pristine LiBH4 (010) surface to 1.53 and
enhancing the rotational mobility of nanoscale complex 0.23 eV when the surface is defected by a Li monovacancy and
hydrides is found in the work of Shane et al., who studied divacancy, respectively. The same group also studied hydrogen
LiBH4 confined in carbon aerogels using hydrogen NMR.481 desorption from the NaAlH4 (001) surface and claimed that
Based on an analysis of NMR line widths, they concluded that the process is “vacancy-mediated”.485 The authors performed
dynamical motion of hydrogen persisted across a wide static and dynamic ab initio calculations to examine the
temperature range. The intensity of this feature was larger hydrogen recombination process on the NaAlH4 (001) surface
for smaller particles, leading the authors to conclude that its in the presence of a neutral Na vacancy. Their ab initio
contribution was associated with surface regions and grain molecular dynamics simulation showed spontaneous release of
boundaries rather than bulk-like regions (additional kinetic hydrogen molecular on the picosecond time scale at 300 K.
aspects of grain boundaries in metal hydride NPs are covered Although the formation energy to create such a metal vacancy
in Section 6.2.1). is relatively high, it was suggested that dopants and impurities
6.1.4. Chemical Kinetics at Defective Surfaces. In such as TiCl3 might stabilize vacancy formation.
surface regions, additional disorder and looser packing, as well Possible surface defect-mediated kinetic improvements in
as undercoordination, may alter the kinetics of other chemical the decomposition of LiNH2 and LiNH2+LiH mixtures with
processes, including hydrogen dissociation/recombination and decreasing particle size were discussed by Hoang et al.486,487
chemical bond activation. In this regard, it can be expected that They proposed two distinct mechanisms based on systematic
the higher concentration of surface atoms in nanoscale metal DFT calculations of formation energies and migration barriers
hydrides should be advantageous for enhancing the overall for native defects and defect complexes. For both mechanisms,
rates of all surface-active chemical processes. Likewise, NPs highly mobile interstitial and vacancy defects (Lii+,VLi−) on the
express proportionately higher concentrations of corners and Li sublattice create vacancy (VH−) or interstitial (Hi+) defects
edges, where atoms are typically even further undercoordi- in the hydrogen sublattice. In the first mechanism, the
nated and therefore likely exhibit additional reactivity. formation of Frenkel pairs (Hi+,VH−), the separation of the
The relevance of undercoordinated surface atoms for H2 defect species in the pairs, and the migration of Hi+ occur in
dissociation is suggested in DFT calculations by Vegge, who the bulk interior of LiNH2. In the second mechanism, the
showed that the NaAlH4 (110) surface, which is more open formation of VH− occurs predominantly at the surface, where
and has lower coordination than the more common (001) activation energy barriers are lowered with respect to the bulk.
surface, exhibits a 0.9 eV lower H2 dissociation barrier.475 The authors claimed that the relative contributions of these
Based on these calculations, Baldé et al. reasoned that higher bulk and surface mechanisms may vary with the particle size,
dissociation activity at open surfaces, corners, and edges could which successfully explains the reduction of experimental
partially explain the dramatic reduction in decomposition activation energies for the kinetic processes with increasing ball
temperature they observed for 2−10 nm NaAlH4 particles milling time.
supported by carbon nanofibers.50 Similar undercoordination 6.1.5. Chemical Impurities and Oxidation. When
arguments have been invoked to explain predictions of additional chemical impurities are present in the metal hydride,
improved desorption thermodynamics in MgH2 nano- they are often more easily accommodated by surfaces than by
wires.482,483 the bulk material. This is because surfaces already contain a
The additional concentration of surface atoms in nanoscale degree of structural and/or chemical inhomogeneity. The
particles may also alter the nucleation kinetics of the hydrided relative prevalence of surfaces in nanoscale metal hydrides may
or dehydrided phases. Bardhan et al. explored this effect therefore lead to complete segregation of impurity atoms to
systematically for Pd NPs.109 They employed a novel optical the surface regions, which can carry additional consequences
technique based on probing luminescence during hydriding for hydrogenation and dehydrogenation kinetics. One example
and dehydriding of an ensemble of Pd nanocubes with exposed can be found in the work of Cho et al., who reported the
(100) facets. By controlling the size distribution, they were kinetic enhancement of reduced graphene oxide-encapsulated
10813 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Mg NPs by Ni doping.36 They observed the formation of Mg− surface oxides on NaAlH4 during dehydrogenation.148 They
Ni nanoalloy clusters populated preferentially in the surface identified the formation of Al 2 O 3 and Ti 2 O 3 during
regions. The authors proposed that these clusters create decomposition of TiCl3-doped NaAlH4 and attributed the
stresses and local disorder that eliminate or reduce surface enhanced dehydrogenation kinetics to the high hydrogen
reaction and hydrogen diffusion barriers, thereby aiding mobility within these oxides species near the surface. Kato et
hydrogenation. al. and Carrillo-Bucio et al. both found that partial oxidation of
The high concentration of undercoordinated atoms at the milled LiBH4, either with oxygen or oxide additives and limited
surfaces of nanoscale metal hydrides also makes them more to the surface, reduced the activation barrier for dehydrogen-
susceptible to side reactions, such as oxidation. In the past, the ation, lowering the observed Tdec.490,491 The oxidized samples
presence of a surface oxide has been considered as one of the exhibited additional Li2O by XPS following segregation of
main factors that contribute to slow hydrogenation/dehydro- lithium to the surface, which helped to destabilize LiBH4 for
genation kinetics, since it acts as a diffusion barrier between the hydrogen desorption and not release of diborane. These
metal hydride and its environment. Multiple strategies have studies suggest that the role of surface oxidation on metal
been applied toward the mitigation of oxidation. In Mg metal, hydride hydrogenation and dehydrogenation kinetics may be
annealing and high-temperature cycling have been used to far more subtle and complex than has generally been supposed.
crack or destroy these surface oxides.488 For nanoscale A better understanding of these effects may illuminate
hydrides, embedding the metal or metal hydride into a additional avenues for improved kinetics of nanoscale hydrides
carbon-based host material can, to some extent, prevent the via control of surface chemistry.
formation of unwanted oxides and therefore improve the 6.2. Effects of Internal Interfaces and Microstructure
overall kinetics.489 In this regard, a solution-based synthetic
approach may show an advantage over the ball-milling In metal hydrides, hydrogen uptake and release kinetics are
technique because it can limit the access of active material to typically determined by a collective combination of solid-state
air and moisture. mass transport, structural modifications, and chemical
A few recent studies have indicated that the presence of a reactivity. These processes are in turn often associated with
properly controlled surface oxide need not necessarily inhibit complicated interactions at internal interfaces. For instance,
reaction kinetics and may even lead to select enhancements. the presence of grain boundaries or internal phase boundaries
For example, Zhang et al. demonstrated that deliberate can affect hydrogen diffusion, which must take place through
formation of an MgO layer atop Mg2NiH4 NPs dispersed on or across such boundaries. Likewise, the relevant chemical
graphene was successful in keeping the particles from reactions tend to take place at interphase regions between
agglomerating.33 They further showed that the oxide layer hydrided and dehydrided domains created by a structural
exhibited selective gas permeability to H2, leading to enhanced transformation. The presence of these interphase regions can
kinetics and stability. Nielsen and Jensen showed that the also alter the mechanical properties of the material, impacting
Nb2O5 additive, which improves the completeness of H2 cycling behavior.
release and absorption in MgH2, forms the ternary oxide At the nanoscale, the energetic contributions of these
MgxNb1−xO (0.2 ≤ x ≤ 0.6).462 This oxide, which has a larger internal interfaces become increasingly important, maximizing
structure than either of the two binary oxides MgO and NbO, their impact on the overall reaction thermodynamics. Similarly,
serves to break up thick MgO and permit H2 diffusion out of many kinetic processes associated with these interfaces
the hydride. increase in prominence as competing bulk processes, such as
Wan et al. showed that when Mg is wrapped in a reduced diffusion, become less accessible. Beyond their effect on the
graphene oxide encapsulant the nanocomposite prevents the energy landscape, interfaces may themselves provide additional
formation of bulk-like MgO but allows for the presence of a storage sites for hydrogen atoms. Accordingly, understanding
thin surface oxide layer.36 Surface-sensitive X-ray photo- the nature of these interfaces and their potential impacts on the
electron spectroscopy (XPS) and X-ray absorption spectros- phase evolution pathway in nanoscale metal hydrides could
copy (XAS) techniques were combined with first-principles facilitate new strategies for enhancing hydrogen storage
simulations to prove the existence of a thin suboxide on the performance.
metal surface and show that it does not necessarily hinder the In metal hydride systems, two major classes of internal
hydrogen dissociation kinetics. In addition, the authors showed interfaces exist: grain boundaries and phase boundaries. The
that formation of this thin interfacial suboxide can be beneficial former class refers to interfaces between two otherwise
toward enhancing the mechanical stability and cyclability of identical crystal domains with different orientations. Indeed,
the nanocomposite due to stronger charge transfer observed it can be assumed that under operating conditions, most metal
between the reduced graphene oxide and the slightly oxidized hydrides form polycrystals consisting of multiple grains
Mg. The unique properties of the suboxide were attributed to separated by grain boundaries. Examples include twin
its lattice structure, which results from controlled or limited boundaries and generic high-angle grain boundaries in
oxidation of Mg. We emphasize here the distinction between Pd.492,493 The latter class refers to interfaces between different
extensive oxidation and controlled or limited oxidation, which phases or crystal compositions. Such interfaces necessarily
can lead to very different surface structures and local oxygen form as intermediates and products coevolve during the course
coordination environments. Thus, it may be difficult to ascribe of the hydriding/dehydriding reaction. Examples include the
a single functionality to a surface oxide without knowledge of well-defined α/β phase boundary in PdHx, as well as more
composition or morphology. disordered interfaces such as those between reactants,
Apart from simple metals, positive effects of surface intermediates, and products in many complex hydrides.5,46
oxidation have also been proposed in the context of the We begin by outlining different classes of internal interfaces
complex metal hydrides NaAlH4 and LiBH4. Using in situ XPS, present in nanoscale metal hydrides. Next, we explore the
Delmelle et al. examined the formation and evolution of effects of grain boundaries on the thermodynamics and kinetics
10814 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

of hydrogenation and dehydrogenation, followed by a similar assumption of solid-solution characteristics of grain bounda-
discussion of the effects of interphase boundaries. ries. Weissmuller and Lemier also reported solubility variations
6.2.1. Effects of Grain Boundaries. Metal hydrides are in nanograin interiors in Pd, which they attributed to the
generally polycrystalline, particularly upon cycling as multiple mechanical stress generated by grain boundary segregation of
product-phase nuclei form. Moreover, these materials are often hydrogen.495
treated via high-energy mechanical milling processes, which Compared with simple hydrides, the effects of grain
disturb crystalline regions. This prompts a discussion of boundaries in complex hydrides have received considerably
possible impacts that grain boundaries and polycrystallinity less attention. This may be because the structure and
could have on thermodynamics or kinetics. Indeed, for composition of grain boundaries in these materials are
nanoscale metal hydrides with sufficiently small grains, atoms particularly difficult to ascertain. However, the far lower
in boundary regions can comprise a nonnegligible fraction of surface energy of most complex hydrides with respect to simple
the overall atoms. hydrides (section 6.1.1) suggests that the thermodynamic
A theoretical investigation into the possible role of grain implications of grain boundaries may be comparatively minor.
boundaries on the thermodynamics of nanoscale metal Instead, the most significant effects may be on kinetics. One
hydrides was performed by Berube et al.10,473,494 Using a example can be found in the recent work by Ray et al., which
simple model approximating a grain boundary as two surfaces focused on the chemical impact of grain boundaries on the
of spherical particles, the authors found that the presence of initial hydrogenation reaction pathways of MgB2 by combining
7−9 nm grains in nanoscale MgH2 could reduce the reaction spectroscopic characterization with multiscale modeling.292 By
enthalpy by up to 9−10 kJ mol−1.473 However, they concluded comparing reference calculations against spectroscopic meas-
that the real value is likely smaller, since the two-surface model urements, they identified that high-energy edge B−B bonds at
places an upper limit on the real grain boundary energy. The grain boundaries provide reaction sites for hydrogenation
same authors also explored the effects of excess volume, which according to a two-step mechanism (H2 dissociation and
is relevant for high-angle grain boundaries or for similarly “diffusive adsorption”), leading to the direct production of
structured interparticle regions in agglomerated particles. Such Mg(BH4)2-like units without other persistent intermediate
regions may be introduced during high-energy mechanical chemical species. They suggested that a high density of
milling processes. Their model used a general equation of state chemically reactive grain boundaries for nanoscale particles
to consider the effects of excess volume and tensile stress on may alter the hydrogenation reaction pathways and enhance
both the enthalpy and entropy. In MgH2 NPs, this enthalpy kinetics. Similarly, an NMR study by Shane et al. hinted at the
was found to be similar in magnitude to the grain boundary possible role of grain boundaries for enhancing hydrogen
energy computed from the two-surface approximation (Figure dynamics in LiBH4.481
30).473 More generally, the authors further postulated that 6.2.2. Types of Interphase Boundaries. In addition to
entropy of disordered excess volume regions could be a grain boundaries, metal hydrides at intermediate stages of
significant factor in stabilizing many metal hydrides.10 hydrogenation and dehydrogenation necessarily include
boundaries between different phases. A broad classification of
these interphase boundaries is shown in Table 6. The type of
boundary differs according to whether (de)hydriding is
predominantly a diffusive process or whether it involves
breaking and forming of chemical (i.e., ionic or covalent)
bonds.
The phase evolution of most complex hydrides (e.g., Mg−
B−H, Li−N−H, and Na−Al−H) involves reactive phase
boundaries, in which the formation of chemically and
structurally distinct phases occurs via interfacial cleavage and
formation of bonds with hydrogen. On the other hand, the
phase evolution of simple hydrides, which encompass metallic
(interstitial) hydrides such as PdHx as well as ionic hydrides
such as MgH2, involves nonreactive phase boundaries. In these
systems, hydrogen is incorporated into interstitial sites of the
Figure 30. Comparison of the reduction of the enthalpy of formation host lattice through an intercalation-type process. The
of MgH2 predicted by the four different mechanisms described in the resulting hydrided phase may or may not inherit the structural
legend. (Reproduced with permission from ref 473. Copyright 2008 characteristics of its dehydrided parent.
Elsevier.) Broadly speaking, the interfacial energies associated with
reactive and nonreactive phase boundaries arise from different
Additional efforts to elucidate the roles of grain boundaries dominant factors. Reactive phase boundaries usually feature
have focused on their contribution to hydrogen solubility in adjoining phases with little or no commonality in structural
nanosized simple hydrides.470 For example, Mutschele and and chemical features. In this case, the lattice continuity is not
Kirchheim experimentally investigated the hydrogen sorption likely to be maintained, and the phase boundaries become
and solubility in nanocrystalline Pd.493 They observed the incoherent. In this case, the coherency strain energy associated
increased and reduced solubility limits of the hydrogen-poor α with lattice commensurability can safely be neglected, and the
and hydrogen-rich β phases, respectively, resulting in a phase boundary propagation and corresponding phase
narrowed miscibility gap. They explained this behavior based evolution instead become governed by interfacial rearrange-
on a theoretical model incorporating the difference in local ments of chemical bonding and mixing of molecular species. It
solubility between grain interiors and boundaries, as well as the is worth emphasizing that not all complex hydride phase
10815 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Table 6. Categorization of Internal Phase Boundaries in Metal Hydrides

Figure 31. (a) Free energy difference between α- and β-Li3N as a function of particle size. (b) Predicted Li2NH mole fraction after hydrogenation
at 100 bar H2 (red) and dehydrogenation at 1 bar (blue) as a function of particle size, with dashed lines representing the results when interfacial
free energy contributions are neglected. (c) Mole fractions of phases in the Li−N−H system as functions of pressure for bulk (left), 10 nm
(center), and 3.2 nm diameter particles (right). (Reproduced with permission from ref 19. Copyright John Wiley and Sons.)

boundaries fall within this category; for example, David et al. the overall thermodynamics and kinetics. The exact degree of
suggested that the lattice similarity between LiNH2 and Li2NH coherency is usually determined by the volumetric and
may preserve coherency at the interface between these two structural disparity between the transforming phases. For
materials.496 Nevertheless, as a general rule, it can be assumed example, the hydrogen-poor and hydrogen-rich phases in the
that most complex metal hydrides will tend to form incoherent Pd−H system (a nonreactive metallic hydride) share the same
interfaces. fcc-based crystal structure, so the corresponding interfaces may
For nonreactive phase boundaries, the intercalation-type maintain interfacial coherency to a large degree at the
hydriding process usually implies similarity in lattice features. nanoscale.110 On the other hand, the hydrogen-poor and
In this case, some degree of coherency of the lattice planes at hydrogen-rich phases in the Mg−H system (a nonreactive
the interface can be assumed. The lattice mismatch between ionic hydride) have different crystal structures (hcp Mg versus
the constituent phases, which arises from expansion upon body-centered tetragonal (bct) MgH2) and different volumes
incorporation of hydrogen, then becomes the most significant (∼30% difference). Accordingly, it is highly likely that
contributor to the interfacial energy. The resulting long-range interfacial dislocations or defects are introduced in such
elastic field can extend well into the bulk interior, influencing systems in order to reduce the coherency strain energy
10816 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

resulting from the extreme lattice mismatch, forming a and the effects of interfaces become increasingly important,
semicoherent interface.497 Some of the key differences between magnifying the relevance of these conclusions.
the subcategories of the interphase boundaries are also 6.2.4. Effects of Nonreactive Phase Boundaries. For
summarized in Table 6. nonreactive interfaces, the interfacial energies are likely to be
6.2.3. Effects of Reactive Phase Boundaries. As particle dominated by mechanical effects associated with coherency
size decreases, it is natural to assume that the increase in the strain. Beyond the higher overall proportion of interfacial
ratio of interface-to-bulk atoms enhances the energetic atoms found in nanoscale systems, the degree of coherency
contribution of internal interphase boundaries to the overall strain due to lattice mismatch between dehydrided and
enthalpy. However, because reactive interfaces within complex hydrided phases can itself become size dependent. Although
metal hydrides are exceedingly difficult to observe or to model widely studied in the context of other energy storage
explicitly, there is very little information about the specific materials,499 detailed studies of the effects of interfacial
impact of these interfaces on reaction thermodynamics or coherency strain on reaction thermodynamics and kinetics in
pathways. hydrogen storage materials have appeared only recently.
One such study recently performed by Wood et al. used a One way to establish the relevance of coherency strain in
combined experiment-theory approach to examine the possible nanoscale systems has been to employ thin films.343,500,501
effect of internal interphase boundaries (“nanointerfaces”) and Baldi et al. reported experimental results for stacks of thin films
microstructure on the reaction pathways of the Li3N/ containing two Mg layers under different mechanical
[LiNH2+2LiH] hydrogen storage system upon nanoconfine- constraints defined by the neighboring layers.343 In their
ment.19 Their experiments determined that the Li2NH phase, experiment, one Mg layer was sandwiched between Ti layers.
which exists as an intermediate in the decomposition pathway This had the effect of isolating Mg elastically, since the
of the bulk material, is suppressed upon cycling of ∼3 nm NPs. immiscibility of Mg (or MgH2) and Ti (or TiH2) leads to a
The elimination of this intermediate enabled a single-step fully incoherent interface. The other Mg layer was bound to
decomposition pathway, leading to enhanced overall kinetics Pd, which alloys with Mg at the interface. This clamps Mg
elastically, since the Mg/Pd interface becomes fully coherent
and better reversibility. Their computational investigation
(see the inset of Figure 32). Experiments and corresponding
found that surface energy effects could not account for the
destabilization of Li2NH. Instead, they proposed that the
behavior was driven by the interfacial energy penalties incurred
in nanoscale particles. By accounting for the approximate
interfacial energy within a microstructural model, they verified
that thermodynamic penalties associated with interfacial
energies could indeed become catastrophic for small NPs,
altering the reaction pathways in agreement with observations
(Figure 31).
The thermodynamic calculations of Wood et al.19 also
provide insight into other possible nanointerface-driven kinetic
enhancements. First, the predicted equilibrium phase fractions
transitioned much more abruptly with pressure for Li3N NPs
than for the bulk material, suggesting a stronger driving force
for phase growth or depletion. Second, the authors concluded
that the elimination of the undesirable intermediate (Li2NH)
as a solid phase necessarily increases the nonstoichiometry of
the product phase (LiNH2) during intermediate reaction
stages, which they reasoned would lead to likely enhancements Figure 32. Pressure-transmission isotherm measured by optical
in mass transport kinetics. hydrogenography on a multilayer thin film sample shown schemati-
Although not directly applied to nanoscale systems, other cally in the inset. (Reproduced with permission from ref 343.
studies have attempted to unravel the effects of phase Copyright 2009 American Physical Society.)
boundaries and their microstructural arrangements on kinetics
and reaction pathways. Michel and Ozoliņs ̌ studied the effect elasticity modeling results indicated that the elastically clamped
of the microstructural arrangement of NaAlH4, Na3AlH6, and Mg layer exhibited a higher plateau pressure, implying that this
Al directly affects the diffusional fluxes of relevant defects in layer was destabilized by the elastic strain contribution from
these phases, which in turn modifies the overall (de)- the mechanical constraint (Figure 32). These results explicitly
hydrogenation kinetics.498 Microstructure-dependent reaction demonstrate the significant impact that internal strain can have
pathways for the dehydrogenation of (LiNH2+LiH) mixtures on the hydrogen storage thermodynamics. They also imply
have also been explored.486,487 Hoang et al. reasoned that, if that the local thermodynamics of nanoscale particles may be
LiNH2 and LiH are in intimate contact in a particle, the controlled by the coherency of the phase boundary, which
dehydrogenation reaction occurs without forming NH3, since determines the relative strain energy contribution.
hydrogen atoms from LiNH2 can be recombined with The theoretical foundations for the thermodynamics of open
hydrogen atoms from LiH at the LiNH2/LiH interface. On two-phase systems with coherent interfaces were established by
the other hand, if these phases are not in intimate contact, the Schwarz and Khachaturyan, who subsequently applied the
dehydrogenation reaction is dominated by the decomposition methodology to understand phase transformations in metal
of LiNH2, which generates NH3 as an intermediate species. In hydrides.502,503 Their model considered the absorption/
nanoscale particles, microstructure may differ from the bulk, desorption of interstitial atoms by a solid crystalline host
10817 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

with a coherent mixture of the two phases. They identified that Griessen et al. comprehensively compared existing thermo-
the coherency strain associated with lattice mismatch produces dynamic data for hydrogen absorption and desorption in Pd
a macroscopic energy barrier between transforming phases, NPs to those of a freestanding thin film and the bulk
which connects the system thermodynamics to the phase material.110 Their results were used to obtain a general
transformation kinetics. They also reasoned that this energy thermodynamic principle of nanosizing effects (Figure 33).
barrier is proportional to the macroscopic volume, which
means that it cannot be easily surmounted by thermally
assisted processes. Instead, the phase transformation occurs
once the external gas pressure surpasses a critical point, beyond
which the free energy of the entire system can be decreased by
the transformation because the barrier vanishes. In this regime,
the two transforming phases cannot coexist at equilibrium. The
authors also claimed that the barrier associated with the
coherency strain is the thermodynamic origin of the hysteresis
in plateau pressures between hydriding and dehydriding phase
transformations. Within their model, the shift in plateau
pressures could be directly quantified in terms of elastic
parameters of the constituent phases.
Recently, experimental efforts to verify the role and
magnitude of coherency strain have benefited significantly
from advanced characterization techniques. These have
enabled more systematic exploration of thermodynamic and
kinetic impacts of intrinsic particle size, shape, and internal
defects on the phase behaviors of hydriding and dehydriding of
nanoscale metal hydrides. These studies have focused on
simple metals that form semicoherent or coherent interfaces
upon hydriding, including Pd and Mg. The use of these simple
model systems allows for unbiased decoupling of mechanical
strain effects from more complex chemical factors. Figure 33. Plateau pressures for hydrogenation of Pd NPs of varying
Baldi et al. successfully observed the phase changes during sizes. Experimental data from various sources are shown in green and
hydrogen absorption/desorption of a single isolated (100)- blue: Bardhan et al.109 (dots), Wadell et al.505 (triangles), Yamauchi
terminated Pd nanocrystal utilizing in situ EELS in an et al.107 (inverted triangles), Züttel et al.506 (half-shaded triangles),
Sachs et al.507 (diamonds), Syrenova et al.418 (squares), Lässer and
environmental transmission electron microscope.112 Exploiting Klatt508 (stars), and Baldi et al.112 (half-shaded circles). Red squares
differences in the plasmon resonances between the hydrogen- represent the upper spinodal pressure, pink circles the lower spinodal
poor α and hydrogen-rich β phases of PdHx for phase pressure, and orange triangles the midpressure, all calculated with a
identification, they characterized a sharp α-to-β transition and mean-field model. (Reproduced with permission from ref 110.
a large hysteresis in the accompanying uptake and release Copyright 2016 Springer Nature.)
isotherms of small particles. They found that the results
followed the coherent thermodynamic theory of Schwarz and
Comparing against a thermodynamic model, they also
Khachaturyan,502,503 suggesting coherency is maintained
concluded that hydrogen absorption of NPs follows a coherent
during the phase transformation for their particle sizes. The interface phase transformation model. However, they also
authors also observed a positive correlation between the asserted that desorption takes place fully coherently only for
hydrogen loading pressure and particle size. This effect was small NPs (<50 nm). For larger particles, hydrogen desorption
attributed to the stress generated by excess hydrogen occurs as some degree of coherency is released. These results
concentration near the surface. They argued that because the demonstrate that particle size affects the structure of the
surface region forms a coherent interface with the particle interface.
interior, the hydrogen-rich outer region exerts a hydrostatic 6.2.5. Nucleation and Phase Transformation Mecha-
tensile stress on the particle core. This stress modifies the nisms. A number of modern in situ techniques have emerged
equilibrium chemical potential of the particle interior. Broadly that allow direct imaging of the hydriding mechanisms of
consistent results were found by Syrenova et al., who used nanoscale metal hydrides, including product nucleation
plasmonic nanospectroscopy to study the hydride formation mechanisms and growth modes. To date, these have chiefly
thermodynamics of single-nanoparticle Pd for a variety of been applied to Pd, which has sufficiently fast kinetics, is
particle sizes and shapes.418 resistant to beam damage, and can be made into single crystals
Further evidence for interfacial coherency in nanoscale Pd with a high degree of control. One example is the work of
was established by Narayan et al. using in situ environmental Narayan et al., who used dark field imaging and selected-area
STEM and EELS during hydriding of Pd nanocubes.504 They electron diffraction (SAED) to identify reaction intermediates
found that (100) phase boundaries, which minimize the elastic within in situ STEM and EELS measurements of PdHx.504
energy, are always established regardless of the initially formed They explicitly verified that a nucleation-and-growth mecha-
phase boundary orientation. Because loss of interfacial nism dominates, with phase transformations originating from
coherency would be expected to express different phase (111) facets at the corners of the cube. Their imaging was able
boundaries, the conclusion was that interfacial coherency is to directly capture a dynamic two-phase coexistence
maintained in nanosized Pd crystals. throughout the process (Figure 34). This nucleation behavior
10818 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

and surfaces are under tensile strain, whereas the particle


center is compressively strained.510
Ulvestad et al. also carried out a single-particle experiment
employing the coherent X-ray diffractive imaging (CXDI)
technique.334 The authors investigated the hydriding phase
transformation dynamics of Pd nanocubes with (100) facets
containing no dislocations by tracking the lattice constant
variation. Similar to the experimental result of Narayan et
al.,504 they observed the α-to-β phase transformation via a
nucleation-and-growth mechanism initiated from the cube
corners. They combined their experiments with steady-state
phase-field modeling to characterize the strain field distribution
across the nanocubes. Within the hydrogen-poor α phase, the
strain distribution was found to be dominated by the
hydrogen-rich surface layer. On the other hand, the strain
distribution of the hydrogen-rich β phase within the nanocube
was primarily dominated by elastic effects with small
compositional inhomogeneity, which tend to minimize the
surface area.
In another paper, Ulvestad et al. successfully characterized
the structural defects and their impact within hydriding Pd
nanocrystals utilizing Bragg coherent diffraction imaging
(BCDI; Figure 35).335 For particles with 150−300 nm sizes,

Figure 34. SEM brightfield images (left) and phase transformations


monitored by STEM (right) of Pd NPs of (a) 20 nm, (b) 36 nm, and
(c) 43 nm, with the scale bar being 25 nm in each image.
(Reproduced with permission from ref 504. Used under the CC BY
4.0 license.)

was thought to originate from both mechanical and chemical


properties of the corners. The corners could be under tensile
strain, which may lower the enthalpy of the hydriding reaction
in these regions. At the same time, (111) surfaces may have
enhanced chemical reactivity, resulting in faster hydrogen
absorption. The morphology of nucleating phases was found to
follow a spherical cap model509 rather than a spherical shell
mechanism.112 The authors also concluded that the phase
transformation is limited by the motion of the phase boundary
rather than the surface reaction or bulk hydrogen diffusion.
A separate study by Narayan et al. reported in situ
characterization of the spatial distribution of hydride phases
within individual Pd nanocrystals with different structural
features (cube, pyramid, and multiply twinned icosahe-
dron).492 Employing plasmon EELS mapping, SAED, and Figure 35. Changes in displacement field and Bragg electron density
centered dark-field imaging techniques, they reconstructed the of PdHx during hydrogen exposure, demonstrating stress evolution
α and β phase distributions to evaluate the impact of structural and defect formation within the particle. (Reproduced with
defects and internal strain on the phase transformation permission from ref 335. Copyright 2017 Springer Nature.)
thermodynamics. Single crystalline cubes and pyramids of Pd
exhibited abrupt α-to-β phase transformation behaviors, with the relationship between equilibrium transformation and
fully hydrogenated and dehydrogenated crystals exhibiting particle size was found to be consistent with a coherent
homogeneous compositions. On the other hand, multiply interface mean-field theory model.110 In contrast, for particles
twinned icosahedral NPs, which contained twin planes and larger than 300 nm, they imaged dislocation nucleation near
defects, exhibited qualitatively different behavior. These the phase boundary during hydriding. Based on these results,
particles contained both α and β phases when hydrogenated, they estimated the critical particle size above which dislocation
with the former preferentially located near the center and the nucleation might theoretically occur due to coherency strain
latter at the edges. The authors claimed that this inhomoge- energy effects. Their measured critical size (∼300 nm) was a
neous spatial distribution of the phases could be explained in factor of 7−10 larger than the predictions, which were based
terms of a nonuniform internal strain field arising from on a core−shell microstructure model. They attributed this
structural defects and crystallite domain boundaries: the edges discrepancy to differences between the model and the actual
10819 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

hydriding phase morphology. Their experimental character- ature, phase mixing will occur spontaneously, stabilizing a
ization instead suggested that the hydrogen-rich phase exhibits solid-solution regime that exhibits PCT sloping behavior.
a spherical cap geometry on the hydrogen-poor phase, 6.3.2.2. Particle-to-Particle Mechanism. Another possible
consistent with the phase morphology determined by Narayan explanation assumes that phase segregation is maintained and
et al.504 With additional evidence from a phase-field micro- that phase transformations occur abruptly within individual
elasticity calculation, they explained that the spherical cap particles or grains but that the sloping plateau instead arises
geometry produces less interfacial coherency strain energy than from an ensemble of nonuniform particles or grains exhibiting
the conventionally assumed core−shell geometry. These more a dispersion of equilibrium pressures. For reasons discussed in
coherent interfaces raise the threshold particle size for sections 6.1 and 6.2, the equilibrium pressure often varies with
necessary dislocation formation. particle size. Therefore, if the ensemble incorporates a broad
6.3. Other Size-Related Effects range of particles/grain sizes, sloping behavior will occur due
to the different equilibrium pressures as larger particles are
6.3.1. Shortened Diffusion Pathways. Arguably the
continuously activated. This mechanism was invoked by
most oft-referenced effect of nanosizing on kinetics involves
Bardhan et al. to account for the PCT behavior of nanoscale
shortening of the diffusion pathways, since the diffusion time, t,
PdHx with a distribution of particle sizes.109
is related to the distance the diffusing species must travel, l, and
6.3.2.3. Defect-Mediated Mechanism. Sloping behavior
the diffusion coefficient, D (eq 18). In metal hydrides for
may also arise from the introduction of additional defects in
which solid-state diffusion is rate limiting, shorter pathways
nanoscale particles. For instance, the introduction of
may lead to faster uptake or a transition to a different rate-
dislocations or twin boundaries can perturb the stress
limiting process altogether. Despite this conventional wisdom,
systematic studies of size-dependent diffusion kinetics in metal distributions within the particle. For small particles or grains,
hydrides have been rare. differences in the stress distributions may become more
pronounced rather than averaging out, as would be expected
l2 for larger particles. The resulting inhomogeneous stress
t≈ distributions will result in a dispersion of pressure responses,
2D (18)
which can generate a sloped pressure plateau. This explanation
One such study was performed by Yao et al., who was favored by Narayan et al. in the interpretation of their in
parametrized a core−shell model of Mg/MgH2 particles (or situ results of the hydriding kinetics of different Pd
grains) using experimentally derived diffusion coefficients.511 nanoparticle morphologies.492 The authors observed sequen-
Their model assumed that hydride formation kinetics are tial mosaic-style loading behavior of the individual crystallites
determined exclusively by diffusion kinetics and that rate within an icosahedral particle over a range of pressures. They
limitations in hydrogen diffusion are associated exclusively with argued that the different equilibrium pressures of small regions
MgH2 rather than Mg. The authors predicted a substantial (i.e., crystallites) within the particle can be traced to different
effect on the overall kinetics of hydrogenation upon reduction degrees of strain modulated by different defect densities.
of particle/grain sizes, with 3 nm particles exhibiting ∼10× However, it is worth noting that, in explaining the sloped
faster hydrogenation than 30 nm particles at 100 °C. plateau pressure in their measured isotherms, they could not
6.3.2. Altered Phase Coexistence Behavior. Another exclude the possibility of a continuous solid-solution
common feature of nanoscale metal hydrides is the mechanism associated with the smallest tetrahedral crystallites
introduction of sloping behavior in the pressure plateau region (∼10 nm). For these domains, the critical temperature could
in the PCT analysis (see Figure 18). This is directly analogous be lower than their experimental operating temperature.
to the behavior exhibited by many nanoscale Li battery Alternatively, buildup of dislocations or structural defects in
electrodes upon electrochemical voltammetry cycling.512 Such nanoscale particles may arrest or pin the phase boundary
behavior has been observed in NPs of NaAlH4392 and motion, slowing the relaxation kinetics and leading to
PdHx,107,492 for example. Whereas a flat plateau indicates deviations from equilibrium. For reasonable cycling rates,
idealized two-phase thermodynamic coexistence and an abrupt these deviations may manifest as sloping plateau behavior.
transformation between phases, as is generally observed for These explanations are broadly consistent with observations of
bulk metal hydrides, there is no single definitive origin for a the phase transformation behavior of Pd NPs by Ulvestad et
sloping plateau. Instead, at least three mechanisms may be al.335 They reported that, in small particles containing
invoked to explain the behavior in nanoscale metal hydrides. dislocations, full transformation to the hydrogen-rich β phase
6.3.2.1. Solid-Solution Mechanism. The most straightfor- was not observed. The authors attributed this to two factors:
ward explanation for the sloping plateau involves stabilization first, coherency loss due to dislocations and, second, inhibition
of solid-solution behavior, which is associated with higher of phase boundary migration due to the interaction with
solubility of hydrogen in the dehydrided phase and vacancies dislocations. Within this interpretation, a dispersion of
in the hydrided phase. Such behavior tends to be more dislocation densities would translate to a dispersion of
prevalent at surfaces, interfaces, and grain boundary regions equilibrium compositions and pressures, as well as phase
due to the higher degree of disorder.493,513 Furthermore, the transformation rates, resulting in sloping plateau behavior.
energetic penalty associated with the interfacial regions
6.4. Metal Hydride-Host Interactions
becomes more significant at the nanoscale,499 which can
favor a single-phase solid solution in order to minimize the Although technical challenges may exist to stabilize NPs
presence of sharp interfaces. Another important factor is the smaller than 10 nm, recent studies show that by incorporating
enhanced interfacial coherency (discussed further in section hydrides into certain porous media or polymer based host
6.2), which generates long-ranged stress fields that can materials, the particle size of metal hydrides can be stabilized
stabilizes solid solutions.514−516 In the limit that the critical to 5 nm or smaller.34,517 Carbon-based materials are of
temperature falls below the experimental operating temper- particular interest as a choice of host material, because they are
10820 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

lightweight, relatively cheap, and abundant. Most carbon-based phases during cycling. A more detailed discussion regarding the
materials are also good thermal conductors,518,519 which can effect of mechanical stresses in the interior of confined NPs can
improve heat exchange during (de)hydrogenation of metal be found in section 6.2.
hydrides. In general, the hydride−host interaction is expected In addition to the potential confinement stress/strain
to be mostly physical, with the host material acting as passive imposed by the host material, the carbon matrix can also
medium for enforcing small particle size. However, in some physically block certain intermediates from vaporization. For
cases, there is evidence that the host matrix can instead interact example, Zhao-Karger et al. demonstrated that by infiltrating
directly with the active storage material and change its LiBH4−Mg(BH4)2 into porous carbon, the toxic intermediate
properties. In the following sections, we discuss physical, B2H6 can be effectively retained in the composite.523
electronic, and chemical aspects of the metal hydride−host 6.4.2. Electronic Coupling Effects. Beyond physical
interaction and address the potential impact of host materials confinement effects, host materials can also affect the behavior
on the hydrogen storage performance of the nanocomposite. of hydrogen storage materials by inducing electronic charge
6.4.1. Physical Confinement Effects. A host material is transfer. Graphene-based carbon nanostructures in particular
generally expected to bring structural stabilization to hydride have been extensively studied for their interactions with metals.
NPs to avoid agglomeration or to help retain high energy For example, graphene is considered as chemisorbed on
surfaces/phases that might be beneficial to (de)hydrogenation. transition metals such as Co, Ni, Pd, and Ti due to strong
Numerous studies have demonstrated the ability to stabilize charge transfer between the metal and the graphene sheet.524
hydride NPs to 1−3 nm in selected carbon matrix materials, For weaker binding observed for graphene on Al, Cu, Ag, Au,
resulting in substantially enhanced (de)hydrogenation kinetics and Pt, the electronic coupling is still large enough to tune the
and shifts in thermodynamic equilibria.11,34 electronic structure of graphene such that the graphene sheet
One advantage of using these nongraphitic carbons as host may be considered as either hole doped or electron doped.524
materials is their ability to change volume. Metal hydrides Even for nanocomposites made of graphene and large band
generally experience significant volume change during (de)- gap semiconductors, such as TiO2, interfacial charge transfer is
hydrogenation; for instance, converting Mg into MgH2 leads to still expected.525 However, most studies of graphene
a volume expansion of ∼30%, with complex hydrides often interactions with other materials have focused on how the
exhibiting far larger expansion.5 The structural flexibility of electronic properties of graphene are modified. Here, we focus
carbon hosts allows them to accommodate the dramatic instead on how electronic coupling at the graphene−hydride
volume change of hydrides, extending cycle life. Nevertheless, interface may affect the (de)hydrogenation processes of the
expansion may accrue some residual stress exerted by the hydride.
confining medium at the interface with the hydride. Among Berseth et al. used C60 fullerenes, carbon nanotubes, and
other effects, this confinement stress can alter the morphology graphene as models of zero, one, and two-dimensional carbon
of the confined material. For instance, Chong et al. showed nanostructures and studied their interactions with NaAlH4.526
that the NaBH4 nanoparticle encapsulated in a “graphene They found that these carbon nanomaterials are catalytic
wrapper” can maintain a spherical shape after cycling, whereas toward hydrogen uptake and release in NaAlH4. To be able to
a composite created by ball milling a mixture of NaBH4 and manipulate the electron localization in these carbon materials
graphene (NaBH4+graphene) shows an irregular shape of and compare their electronic interactions with NaAlH4, they
NaBH4 distributed within flake-like graphene layers.30 synthesized carbon nanotubes with varying diameters
Confinement stress can also affect the (de)hydrogenation (curvature) using a solution-based preparation technique to
thermodynamics and kinetics. As discussed in section 6.1, the avoid potential metal and carbon contamination. Upon cycling
reaction enthalpy and the pathways should be re-evaluated the nanocomposite, they found C60 had the best performance
when a change of surface energy is expected for the hydride to catalyze NaAlH4. This was attributed to the ability of C60 to
experiencing surface strain. The possibility of using volumetric restrict electron donation from Na to AlH4, thereby weakening
stress to alter hydrogen desorption properties was explored by the Al−H bond. From first-principles simulations, they
Ham et al.520 Using a thin film of a metastable phase of MgH2, correlated the hydrogen removal energy to the electron
they measured stress-dependent hydrogen desorption thermo- affinities of the carbon nanomaterials, which in turn are
dynamics. The authors showed that stressed thin films can related to local curvature. As the curvature increases, stronger
release hydrogen at lower temperatures than stress-free MgH2, charge transfer occurs at the interface, which affects charge
suggesting mechanically induced destabilization. partitioning within the hydride and influences (de)-
A number of first-principles studies have also examined the hydrogenation performance.
strain effect on (de)hydrogenation kinetics of metal hydrides. Nitrogen-doped porous carbons, NCMK-3 and NNPC, were
Lei et al. computed the adsorption and dissociation energy of used by Carr et al. to confine LiBH4 and NaAlH4.403,527 The
H2 on Mg (0001) surface as a function of biaxial strain.521 pyridinic nitrogen atoms acted as Lewis bases on the complex
Zhang et al. studied the dehydrogenation properties of MgH2 hydride anions, greatly increasing the desorption activation
under biaxial strain and found that both tensile and energy of confined LiBH4 and, while decreasing the activation
compressive strain lowers the activation energy for hydrogen energy of confined NaAlH4, increasing the temperature at
diffusion.522 Similarly, Benzidi et al. showed that under which maximum desorption from the alanate occurred (Figure
compression, the energy barrier for hydrogen diffusion in 36).
LiBH4 can be reduced from ∼93 to ∼86 kJ mol−1.12 The 6.4.3. Chemical Interactions with the Host. For
elastic constraint placed by the host material may also assist the nanostructures based on carbon scaffolds or encapsulants,
buildup of internal strain energy at grain boundaries or carbon may also interact directly with hydrides to influence
interphase boundaries, which might otherwise be easily their (de)hydrogenation performance.49,58 It has been shown
released for unconfined particles. This buildup could lead to experimentally that, by adding carbon materials, the hydro-
the creation or elimination of new metastable intermediate genation kinetics of Mg metal can be greatly improved.528 DFT
10821 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

but also for batteries, portable power, and solid electro-


lytes.46,63 There is now overwhelming evidence that nanosizing
affects the reaction pathways, kinetics, and thermodynamics,
with several theoretical and experimental investigations
suggesting that both the enthalpy and entropy of hydrogen
desorption can be altered. Synthetic approaches, both top-
down and bottom-up, are now available for all classes of metal
hydrides that can be used to produce both free-standing and
supported NPs or clusters, thin films, and nanowires, as well as
more complex shapes such as core−shell particles and
multilayers (e.g., Figures 2, 5, and 17).
Control over the nanohydride properties, in particular,
composition, size, and purity, is far greater than what was
previously achieved using conventional mechanical milling.
Supporting these experimental developments is the vast
increase in computing power and theoretical method
innovations that are making structure and property prediction
faster and more reliable. As a result, the fundamental
understanding of nanoscale properties has improved dramat-
ically in the past ten years, such that basic design principles are
beginning to emerge. On the most basic level, particle size
appears to be an almost universal property for predicting the
extent of nanoscale effects. Although this varies from one
material to another, recent investigations in which a high
degree of particle size control and uniformity was achieved
indicate that the most dramatic effects occur when the critical
dimension is 10 nm or less.
Figure 36. (a) Arrhenius plots of NaAlH4 confined in two undoped
carbons (NPC and CMK-3) and two N-doped carbons (NNPC and
The preponderance of the available data indicates that
NCMK-3) and of unconfined bulk NaAlH4 from isothermal nanoscaling leads to significant improvements in the kinetics of
desorption. (b) Ramped-temperature hydrogen desorption from the hydrogen uptake and desorption. However, it has yet to be
NaAlH4-infiltrated composites, with the black dashes representing the determined conclusively as to whether such dramatic effects on
temperature profile. (Adapted with permission from ref 527. the thermodynamics can be achieved. Theory suggests that it is
Copyright 2018 American Chemical Society.) possible, but the most significant impacts from surface energies
are predicted for very small particle dimensions (a few
work also predicts that interstitial C in the Mg lattice can nanometers). Mechanical stresses due to volume expansion
simultaneously lower the activation energy for H2 dissociation upon hydrogenation also merit further theoretical and
(by 0.16 eV) and improve atomic hydrogen diffusion through experimental investigation, since stress effects on both kinetics
the subsurfaces (by 0.33 eV or more).529 In other systems, de/ and thermodynamics have been proposed. These factors could
rehydrogenation may involve carbide formation, indicating a contribute to changes in reaction pathways, such as those
direct reaction with the confining medium. For example, observed in NaAlH4, LiNH2, and LiBH4 for which the number
Zhang et al. showed reversible appearance of Li2C2 phase in of reaction steps is reduced upon nanoscaling. Although theory
the XRD patterns during de/rehydrogenation of nano-LiBH4 generally supports the conclusion that these are thermody-
confined within mesoporous carbon.399 Miyaoka et al., based namically driven, it cannot be ruled out that they may instead
on a Li−C−H system composed of hydrogenated graphite and result from lowering of a particular activation barrier to enable
LiH, demonstrated the formation of both Li2C2 and CH4.530 the reaction to proceed through a different pathway. Indeed,
Although reasonable cycling temperature (350 °C) was distinguishing between thermodynamic and kinetic effects can
achieved, the system suffered from low rehydrogenation be difficult, in part because thermodynamic data for the bulk
capacity due to the formation of CH4. Residual oxygen in material are not always well established. More advanced kinetic
the carbon host, present as CO or CO species, can also models that account for the diversity of physicochemical
react with hydrides or their dehydrogenated forms. It has been processes that govern (de)hydrogenation may help to elucidate
suggested that magnesium, which is very oxophilic, readily which factors control the reaction pathway.
scavenges oxygen from its confining scaffold and forms MgO, Rationalizing the observed improvements in hydride proper-
which no longer participates in hydrogen cycling and decreases ties remains a matter of debate, as a number of factors that can
the overall capacity over time.454 affect the metal hydrides at nanoscale, including surface area,
defects, nanointerfaces, and strain, are difficult to control. The
7. CONCLUSIONS rather simplistic interpretation of the transport phenomena,
It is abundantly clear from the foregoing discussion that based solely on shorter diffusion times and lengths, occurring
research and development of nanoscale metal hydrides is a during H2 absorption/desorption is appealing, but, in fact,
vigorous field. In this review, we considered only work in several other factors can affect the kinetics of nanoscale metal
roughly the past two decades, which still accounts for more hydrides. Examples include hydride−hydride and hydride−
than five hundred cited papers. These materials are no longer a scaffold nanointerfaces, where lattice mismatches and defects
laboratory curiosity but are a growing subclass with potential sites are also likely to occur. Likewise, the effects of surface and
applications not only for hydrogen storage, as discussed herein, interface heterogeneity and disorder have not been studied in
10822 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

sufficient detail, nor has the evolution of internal micro- or more penalties typically observed for other carbon-based
structure. Consequently, understanding of the phenomena nanoconfinement media, such as templated porous carbons or
controlling nanohydride properties remains incomplete. carbon aerogels. Reduced graphene oxide is an excellent
In spite of this rapid progress, important challenges remain encapsulating layer for wrapping Mg34 and Ni-doped Mg36
regarding both fundamental understanding and practical nanocrystals for hydrogen storage. In addition to high
implementation of nanohydrides. Since most of these materials gravimetric and volumetric hydrogen capacities, the materials
are stabilized within some kind of host or shell, understanding display enhanced air-stability, due to the rGO protecting layer.
how the host/shell material affects properties is essential. The In summary, the rapid progress made in the past decade has
presence of reactive sites within a nanoscaffold, such as open created exciting new opportunities for synthesizing nano-
metal sites in MOFs or Lewis acid or base sites in porous hydrides with properties tailored for specific applications.
carbons, make the host “non-innocent,” in the sense that Although nanoscale phenomena may seem quite distant from
chemical interactions would be expected to be stronger than large-scale applications such as vehicular hydrogen storage, the
weak van der Waals forces. Moreover, there is evidence that recent advances are so significant that it is not difficult to
even hosts that might otherwise be considered “passive” may envision progress toward materials that could ultimately meet
affect properties through the creation of solid−solid inter- some of the most challenging DOE technical targets. Proving
faces.60 Focused efforts employing both theory and experi- this true is a worthy scientific goal with profound implications
ments are needed to provide insights for selecting and for future energy storage systems and, more broadly, for
designing host materials to achieve specific properties. countering the effects of climate change.
Similarly, there is evidence that solid−solid interfaces within
the nanoparticle itself that form during uptake or release of H2 AUTHOR INFORMATION
can dramatically influence kinetics, as is the case of Li3N.19
Corresponding Author
Further investigation into the role of solid−solid interfaces and
material-host interactions for other metal hydrides is *E-mail: vnstavi@sandia.gov.
recommended; this may necessitate additional developments ORCID
in theoretical modeling of interfaces.
The interest in using nanostructured metal hydrides for Andreas Schneemann: 0000-0001-6801-2735
hydrogen storage applications has prompted the need for more James L. White: 0000-0002-8216-7212
accurate measurement of hydrogen capacity, thermodynamics, Liwen F. Wan: 0000-0002-5391-0804
and kinetics. The complex nature of the samples (particles of Eun Seon Cho: 0000-0001-9428-1269
various sizes, presence of porous hosts, impurities) have Jeffrey J. Urban: 0000-0002-6520-830X
presented new challenges for traditional hydrogen measure- Brandon C. Wood: 0000-0002-1450-9719
ment techniques. Several factors can affect the accuracy of the Mark D. Allendorf: 0000-0001-5645-8246
measurements, including sample size, temperature gradients, Vitalie Stavila: 0000-0003-0981-0432
types of valves used, leak rates, and must be carefully
Author Contributions
considered.531−533 In addition, many of the materials studied

are porous (e.g., gamma-Mg(BH4)2)323 and can adsorb helium, A.S. and J.L.W. contributed equally to this work.
which has typically been considered as a nonadsorbing species Notes
for free space determination. Helium adsorption has been
shown to impact skeletal density measurements in porous The authors declare no competing financial interest.
samples.531,532 The U.S. Department of Energy Fuel Cell Biographies
Technologies Office provided a comprehensive overview of the
recommended best practices in making measurements of the Andreas Schneemann originates from Sprockhövel, Germany. He
hydrogen storage properties on their Web site.337 Finally, it is obtained his M.Sc. and Ph.D. at the Ruhr University Bochum under
critical for the community to report not only the hydrogen the guidance of Prof. Roland A. Fischer, working on metal−organic
capacity or the rate of hydrogen release and uptake but also frameworks. He was a visiting scholar in the laboratories of Ian A.
thermodynamic and kinetic data, such as enthalpy, entropy, Fallis (Cardiff University), Seth M. Cohen (University of California,
and activation energy (e.g., Tables 3−5). These values may all San Diego), and Shin-ichiro Noro (Hokkaido University). After a
change over the course of multiple dehydrogenation/ short stay as a scientific staff member at Technical University Munich
rehydrogenation cycles, so the stability of the storage materials he joined the group of Mark D. Allendorf at Sandia National
must be assessed, as has been done for some, but not many, Laboratories in Livermore, California on a German Research
systems. Foundation (Deutsche Forschungsgemeinschaft) Fellowship. His
From a practical perspective, it should be noted that that current research interests revolve around hydrogen storage materials,
there are significant (∼20−30%) gravimetric and volumetric particularly metal hydrides confined in porous matrices.
penalties associated with using even optimized nanoscaffolds,44 James L. White, a native of Pennsylvania, earned his Ph.D. in
which make it challenging to consider them for optimal chemistry and materials from Princeton University in 2016, working
hydrogen storage systems. Ideally, a synthetic route to under Prof. Andrew B. Bocarsly investigating electrocatalytic carbon
nanoscale particles in the 1 to 20 nm range would be highly dioxide reduction. Prior to that, he obtained his honors B.S. degree in
desirable, producing NPs with enhanced kinetic behavior but chemistry with minors in biochemistry, mathematics, and economics
without the gravimetric and volumetric penalties associated from the University of Delaware in 2011, where he did undergraduate
with scaffold use. A promising new direction is nano- research on solid oxide fuel cell electrolytes with Prof. Joshua L.
encapsulation of metal hydride NPs inside graphene and Hertz. He is currently a postdoctoral fellow at Sandia National
reduced graphene oxide sheets. In selected cases, the Laboratories, where he has been studying complex metal hydrides,
gravimetric penalty does not exceed 10%, compared to 50% their reactivity under extreme conditions, and their mechanisms, as

10823 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

part of the Hydrogen Materials−Advanced Research Consortium Jeffrey J. Urban is a Staff Scientist and Facility Director at the
(HyMARC). Molecular Foundry located in Lawrence Berkeley National Labo-
ShinYoung Kang is a postdoctoral researcher at Lawrence Livermore ratories in Berkeley, CA. He earned his Ph.D. in physics and
National Laboratory in Livermore, CA. She received her Ph.D. in chemistry in 2004 from Harvard University. His projects in the area of
Materials Science and Engineering at Massachusetts Institute of hydrogen storage focus on the development of new approaches to
nanoscale metal hydrides and metal hydride hybrid materials
Technology in 2014. Her primary research focus is to understand
complexed with organic frameworks and graphene-based materials.
fundamental processes occurring in clean energy technology,
He has also studied the fundamental size-dependence of hydriding
including hydrogen storage, batteries, and solar cells, and to provide
phase transitions in metal hydrides, establishing the first general
design insights using atomistic simulations.
model of confinement in these systems. A central theme in Dr.
Sohee Jeong earned her Ph.D. in Chemistry at Yonsei University in Urban’s research is revealing the fundamental physical origins of
2016. She is currently a postdoctoral fellow in Jeffrey Urban’s research transport and phase transformation in hydrogen storage materials.
group at Lawrence Berkeley National Lab. Her research interests
Brandon Wood is currently a Staff Scientist in the Materials Science
include the development of new metal hydride complexes for gas
Division at Lawrence Livermore National Laboratory, U.S.A. He
storage and the rational design of nanocrystals, with an emphasis on
received his Ph.D. in Materials Science and Engineering from MIT in
transition metal dichalcogenide 2D materials.
2007. His research activities lie in the application of ab initio and
Liwen Wan is currently a postdoctoral researcher at Lawrence mesoscale simulation techniques to materials for energy storage and
Livermore National Laboratory. She received her Ph.D. in Material conversion, with particular emphasis on the use of high-performance
Sciences and Engineering in 2013 from Iowa State University and computing to understand complex interfaces. He is the Theory
then worked as a postdoctoral research fellow at Lawrence Berkeley Director for the U.S. Department of Energy Hydrogen Storage
National Laboratory. Her research focuses on using theoretical Materials--Advanced Research Consortium (HyMARC), for which he
methods to study and probe chemical reactions and processes across leads the development and application of multiscale computational
solid−liquid and solid−solid interfaces for energy storage and methods for probing kinetics and thermodynamics of metal hydrides
conversion applications. and sorbents. He has (co)authored more than 60 peer-reviewed
Eun Seon Cho is an Assistant Professor of Chemical and publications and two patents.
Biomolecular Engineering at Korea Advanced Institute of Science Mark Allendorf is Director of the Hydrogen Materials−Advanced
and Technology (KAIST). She received her Ph.D. in Materials Research Consortium (HyMARC) and a senior scientist at Sandia
Science and Engineering in 2013 from Massachusetts Institute of National Laboratories. He received an AB degree in chemistry from
Technology. Her works in the area of hydrogen storage focus on the Washington University in St. Louis and a Ph.D. degree in chemistry
development of nanostructured metal hydrides to enhance the from Stanford University. His research focuses on the fundamental
hydrogen release and absorption properties. Her major research science and applications of metal−organic frameworks and related
interests include the design and synthesis of functional hybrid materials. His work has been published in more than 170 publications,
nanomaterials with organic and inorganic building blocks for energy including more than 130 journal articles. He is President Emeritus
and environmental applications. and Fellow of The Electrochemical Society.
Tae Wook Heo is a Staff Scientist at Lawrence Livermore National Vitalie Stavila is a Principal Member of Technical Staff at Sandia
Laboratory. He earned his Ph.D. in Materials Science and Engineering National Laboratories in Livermore, CA. He earned his Ph.D. in
in 2012 from The Pennsylvania State University. He specializes in inorganic chemistry in 2002 from State University of Moldova. His
mesoscale modeling of complex phase microstructures of various projects in the area of hydrogen storage focus on the development of
classes of materials including hydrogen storage materials, battery metal hydrides and sorbents with improved kinetics and thermody-
electrodes/electrolytes, and metallic alloys. Within the hydrogen namics of hydrogen release and absorption. A central theme in Dr.
storage research program, his effort is mainly focused on mass Stavila’s research is the rational design of new materials by changing
transport and phase transformations coupled to mechanical stress and the chemical composition, the arrangement of the atoms or molecules
interfacial chemical reactions. He is actively involved in multiscale in crystalline or amorphous configurations, and the size, shape, and
integration connecting atomistic simulations with mesoscale formal- orientation of nano- or macroscopic objects. Dr. Stavila coauthored
isms, as well as code development for phase-field modeling of simple more than 110 journal articles and nine patents.
and complex metal hydrides.
David Prendergast is a Senior Staff Scientist at Lawrence Berkeley ACKNOWLEDGMENTS
National Laboratory, working in its Energy Sciences Area. He is the Sandia National Laboratories is a multimission laboratory
Facility Director for Theory at The Molecular Foundry, a National managed and operated by National Technology and Engineer-
User Facility for Nanoscience, serving academic, government, and ing Solutions of Sandia, LLC., a wholly owned subsidiary of
industrial research. He has a Ph.D. in Physics from University College Honeywell International, Inc., for the U.S. Department of
Cork, Ireland (2002). His research focuses on developments and Energy’s National Nuclear Security Administration under
applications of first-principles electronic structure theory and contract DE-NA-0003525. This work was performed in part
molecular dynamics in the context of energy-relevant phenomena in under the auspices of DOE by Lawrence Livermore National
chemistry and materials, from molecules to condensed phases. He has Laboratory under Contract DE-AC52-07NA27344. Lawrence
particular expertise in the simulation and interpretation of Berkeley National Laboratory efforts were supported by the
synchrotron X-ray spectroscopy measurements as a means of Office of Science of the U.S. Department of Energy under
connecting characterization of complex and dynamic homogeneous Contract No. DE-AC02-05CH11231. The authors gratefully
or interfacial systems to atomistic or molecular-scale mechanistic acknowledge research support from the U.S. Department of
models and the development of the necessary computational methods Energy, Office of Energy Efficiency and Renewable Energy,
to derive such connections efficiently. Fuel Cell Technologies Office through the Hydrogen Storage
10824 DOI: 10.1021/acs.chemrev.8b00313
Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Materials Advanced Research Consortium (HyMARC). A.S. NVS neutron vibrational spectroscopy
gratefully acknowledges the German Research Foundation NW nanowire
(Deutsche Forschungsgemeinschaft) for a Postdoctoral OMS ordered mesoporous silica
Fellowship (SCHN 1539/1-1). E.S.C. is grateful for financial PCT pressure−composition−temperature (measure-
support by International Energy Joint R&D Program of the ment)
Korea Institute of Energy Technology Evaluation and Planning PEGS prototype electrostatic ground state
(KETEP), granted financial resource from the Ministry of PEM polymer electrolyte membrane
Trade, Industry & Energy, Republic of Korea (No. PMMA poly(methyl methacrylate)
20188520000570). This paper describes objective technical PTI pressure−temperature isotherm
results and analysis. Any subjective views or opinions that PVD physical vapor deposition
might be expressed in the paper do not necessarily represent PVP poly(N-vinyl-2-pyrrolidone)
the views of the U.S. Department of Energy or the United PXRD powder X-ray diffraction
States Government. rGO reduced graphene oxide
SAED selected-area electron diffraction
LIST OF ABBREVIATIONS SANS small-angle neutron scattering
APXPS ambient pressure X-ray photoelectron spectros- SAXS small-angle X-ray scattering
copy SEM scanning electron microscopy
BCDI Bragg coherent diffraction imaging STEM scanning transmission electron microscopy
bct body-centered tetragonal STXM scanning transmission X-ray microscopy
BDE bond dissociation enthalpy Tdec decomposition temperature
BET Brunauer−Emmet−Teller surface area measure- TEM transmission electron microscopy
ment TGA thermogravimetric analysis
BJH Barrett−Joyner−Halenda pore volume analysis THF tetrahydrofuran
BTC benzenetricarboxylate UHV ultrahigh vacuum
Bu butyl (typically n-butyl) XANES X-ray absorption near edge structure
CALPHAD calculation of phase diagrams XAS X-ray absorption spectroscopy
CNF carbon nanofiber XPS X-ray photoelectron spectroscopy
CNT carbon nanotube XRD X-ray diffraction/diffractometry
Cp cyclopentadienyl XRS X-ray Raman spectroscopy
CXDI coherent X-ray diffractive imaging
DFT density-functional theory
dobdc 2,5-dioxido-1,4-benzenedicarboxylate REFERENCES
DOE Department of Energy (1) Chu, S.; Majumdar, A. Opportunities and Challenges for a
DRIFTS diffuse reflectance infrared Fourier-transform Sustainable Energy Future. Nature 2012, 488, 294.
spectroscopy (2) Mika, L. T.; Csefalvay, E.; Nemeth, A. Catalytic Conversion of
DSC differential scanning calorimetry Carbohydrates to Initial Platform Chemicals: Chemistry and
DTA differential thermal analysis Sustainability. Chem. Rev. 2018, 118, 505−613.
(3) Wee, J. H. Applications of Proton Exchange Membrane Fuel Cell
EDX energy-dispersive X-ray spectroscopy Systems. Renewable Sustainable Energy Rev. 2007, 11, 1720−1738.
EELS electron energy loss spectroscopy (4) Preuster, P.; Alekseev, A.; Wasserscheid, P. In Annual Review of
ERDA elastic recoil detection analysis Chemical and Biomolecular Engineering; Prausnitz, J. M., Ed., Annual
Et ethyl Reviews: Palo Alto, CA, 2017; Vol. 8.
EXAFS extended X-ray absorption fine structure (5) Orimo, S. I.; Nakamori, Y.; Eliseo, J. R.; Züttel, A.; Jensen, C. M.
fcc face-centered cubic Complex Hydrides for Hydrogen Storage. Chem. Rev. 2007, 107,
GCLP grand canonical linear programming 4111−4132.
GO graphene oxide (6) Allendorf, M. D.; Hulvey, Z.; Gennett, T.; Ahmed, A.; Autrey,
hcp hexagonally close-packed T.; Camp, J.; Cho, E. S.; Furukawa, H.; Haranczyk, M.; Head-Gordon,
HFCV hydrogen fuel cell vehicle M.; Jeong, S.; Karkamkar, A.; Liu, D. J.; Long, J. R.; Meihaus, K. R.;
Navyar, I. H.; Nazarov, R.; Siegel, D. J.; Stavila, V.; Urban, J. J.;
HPMR hydrogen plasma metal reaction
Veccham, S. R.; Wood, B.C. An Assessment of Strategies for the
HR high-resolution Development of Solid-State Adsorbents for Vehicular Hydrogen
HSAG high surface area graphite Storage. Energy Environ. Sci. 2018, DOI: 10.1039/C8EE01085D.
IBS ion-beam sputtering (7) Target Explanation Document: Onboard Hydrogen Storage for
IR infrared Light-Duty Fuel Cell Vehicles. United States Department of Energy:
JMAK Johnson−Mehl−Avrami−Kolmogorov kinetic Washington, DC, 2017; https://www.energy.gov/eere/fuelcells/
model downloads/target-explanation-document-onboard-hydrogen-storage-
MBE molecular beam epitaxy light-duty-fuel-cell.
MD molecular dynamics (8) Churchard, A. J.; Banach, E.; Borgschulte, A.; Caputo, R.; Chen,
Mm mischmetal (rare-earth mixture) J.-C.; Clary, D. C.; Fijalkowski, K. J.; Geerlings, H.; Genova, R. V.;
Grochala, W.; et al. A Multifaceted Approach to Hydrogen Storage.
MOF metal−organic framework
Phys. Chem. Chem. Phys. 2011, 13, 16955−16972.
MS mass spectrometry (9) Kan, H. M.; Zhang, N.; Wang, X. Y.; Sun, H. Recent Advances in
MTBE methyl tert-butyl ether Hydrogen Storage Materials. Adv. Mater. Res. 2012, 512−515, 1438−
MWCNT multiwalled carbon nanotube 1441.
NP nanoparticle (10) Berube, V.; Dresselhaus, M. S.; Chen, G. Entropy Stabilization
NRA nuclear reaction analysis of Deformed Regions Characterized by an Excess Volume for

10825 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Hydrogen Storage Applications. Int. J. Hydrogen Energy 2009, 34, (30) Chong, L.; Zeng, X.; Ding, W.; Liu, D.; Zou, J. NaBH4 in
1862−1872. ″Graphene Wrapper:″ Significantly Enhanced Hydrogen Storage
(11) Bérubé, V.; Radtke, G.; Dresselhaus, M.; Chen, G. Size Effects Capacity and Regenerability through Nanoencapsulation. Adv. Mater.
on the Hydrogen Storage Properties of Nanostructured Metal 2015, 27, 5070−5074.
Hydrides: A Review. Int. J. Energy Res. 2007, 31, 637−663. (31) Ji, L.; Meduri, P.; Agubra, V.; Xiao, X.; Alcoutlabi, M.
(12) Benzidi, H.; Lakhal, M.; Benyoussef, A.; Hamedoun, M.; Graphene-Based Nanocomposites for Energy Storage. Adv. Energy
Loulidi, M.; El kenz, A.; Mounkachi, O. First Principle Study of Strain Mater. 2016, 6, 1502159.
Effect on Structural and Dehydrogenation Properties of Complex (32) Xia, G.; Tan, Y.; Wu, F.; Fang, F.; Sun, D.; Guo, Z.; Huang, Z.;
Hydride LiBH4. Int. J. Hydrogen Energy 2017, 42, 19481−19486. Yu, X. Graphene-Wrapped Reversible Reaction for Advanced
(13) Bogdanovic, B.; Schwickardi, M. Ti-Doped Alkali Metal Hydrogen Storage. Nano Energy 2016, 26, 488−495.
Aluminium Hydrides as Potential Novel Reversible Hydrogen Storage (33) Zhang, J.; Zhu, Y.; Lin, H.; Liu, Y.; Zhang, Y.; Li, S.; Ma, Z.; Li,
Materials. J. Alloys Compd. 1997, 253−254, 1−9. L. Metal Hydride Nanoparticles with Ultrahigh Structural Stability
(14) Wagemans, R. W. P.; van Lenthe, J. H.; de Jongh, P. E.; van and Hydrogen Storage Activity Derived from Microencapsulated
Dillen, A. J.; de Jong, K. P. Hydrogen Storage in Magnesium Clusters: Nanoconfinement. Adv. Mater. 2017, 29, 1700760.
Quantum Chemical Study. J. Am. Chem. Soc. 2005, 127, 16675− (34) Cho, E. S.; Ruminski, A. M.; Aloni, S.; Liu, Y.-S.; Guo, J.;
16680. Urban, J. J. Graphene Oxide/Metal Nanocrystal Multilaminates as the
(15) Cheung, S.; Deng, W.-Q.; van Duin, A. C. T.; Goddard, W. A. Atomic Limit for Safe and Selective Hydrogen Storage. Nat. Commun.
ReaxFFMgH Reactive Force Field for Magnesium Hydride Systems. J. 2016, 7, 10804.
Phys. Chem. A 2005, 109, 851−859. (35) Cho, E. S.; Ruminski, A. M.; Liu, Y.-S.; Shea, P. T.; Kang, S.;
(16) Nielsen, T. K.; Javadian, P.; Polanski, M.; Besenbacher, F.; Zaia, E. W.; Park, J. Y.; Chuang, Y.-D.; Yuk, J. M.; Zhou, X.; et al.
Bystrzycki, J.; Skibsted, J.; Jensen, T. R. Nanoconfined NaAlH4: Hierarchically Controlled inside-out Doping of Mg Nanocomposites
Prolific Effects from Increased Surface Area and Pore Volume. for Moderate Temperature Hydrogen Storage. Adv. Funct. Mater.
Nanoscale 2014, 6, 599−607. 2017, 27, 1704316.
(17) Bhakta, R. K.; Maharrey, S.; Stavila, V.; Highley, A.; Alam, T.; (36) Wan, L. W. F.; Liu, Y. S.; Cho, E. S.; Forster, J. D.; Jeong, S.;
Majzoub, E.; Allendorf, M. Thermodynamics and Kinetics of NaAlH4 Wang, H. T.; Urban, J. J.; Guo, J. H.; Prendergast, D. Atomically Thin
Nanocluster Decomposition. Phys. Chem. Chem. Phys. 2012, 14, Interfacial Suboxide Key to Hydrogen Storage Performance Enhance-
8160−8169. ments of Magnesium Nanoparticles Encapsulated in Reduced
(18) Stavila, V.; Bhakta, R. K.; Alam, T. M.; Majzoub, E. H.; Graphene Oxide. Nano Lett. 2017, 17, 5540−5545.
Allendorf, M. D. Reversible Hydrogen Storage by NaAlH4 Confined (37) Gosalawit-Utke, R.; Meethom, S.; Pistidda, C.; Milanese, C.;
within a Titanium-Functionalized MOF-74(Mg) Nanoreactor. ACS Laipple, D.; Saisopa, T.; Marini, A.; Klassen, T.; Dornheim, M.
Nano 2012, 6, 9807−9817. Destabilization of LiBH4 by Nanoconfinement in PMMA-co-BM
(19) Wood, B. C.; Stavila, V.; Poonyayant, N.; Heo, T. W.; Ray, K. Polymer Matrix for Reversible Hydrogen Storage. Int. J. Hydrogen
G.; Klebanoff, L. E.; Udovic, T. J.; Lee, J. R. I.; Angboonpong, N.; Energy 2014, 39, 5019−5029.
Sugar, J. D.; et al. Nanointerface-Driven Reversible Hydrogen Storage (38) Plerdsranoy, P.; Wiset, N.; Milanese, C.; Laipple, D.; Marini,
in the Nanoconfined Li-N-H System. Adv. Mater. Interfaces 2017, 4, A.; Klassen, T.; Dornheim, M.; Gosalawit-Utke, R. Improvement of
1600803. Thermal Stability and Reduction of LiBH4/Polymer Host Interaction
(20) Boles, M. A.; Engel, M.; Talapin, D. V. Self-Assembly of of Nanoconfined LiBH4 for Reversible Hydrogen Storage. Int. J.
Colloidal Nanocrystals: From Intricate Structures to Functional Hydrogen Energy 2015, 40, 392−402.
Materials. Chem. Rev. 2016, 116, 11220−11289. (39) Pentimalli, M.; Imperi, E.; Zaccagnini, A.; Padella, F.
(21) Jeon, K.-J.; Moon, H. R.; Ruminski, A. M.; Jiang, B.; Nanostructured Metal Hydride - Polymer Composite as Fixed Bed
Kisielowski, C.; Bardhan, R.; Urban, J. J. Air-Stable Magnesium for Sorption Technologies. Advantages of an Innovative Combined
Nanocomposites Provide Rapid and High-Capacity Hydrogen Storage Approach by High-Energy Ball Milling and Extrusion Techniques.
without Using Heavy-Metal Catalysts. Nat. Mater. 2011, 10, 286− Renewable Energy 2017, 110, 69−78.
290. (40) Callini, E.; Pasquini, L.; Piscopiello, E.; Montone, A.; Antisari,
(22) Bardhan, R.; Ruminski, A. M.; Brand, A.; Urban, J. J. M. V.; Bonetti, E. Hydrogen Sorption in Pd-Decorated Mg-Mgo
Magnesium Nanocrystal-Polymer Composites: A New Platform for Core-Shell Nanoparticles. Appl. Phys. Lett. 2009, 94, 221905.
Designer Hydrogen Storage Materials. Energy Environ. Sci. 2011, 4, (41) Lu, C.; Zou, J. X.; Shi, X. Y.; Zeng, X. Q.; Ding, W. J. Synthesis
4882−4895. and Hydrogen Storage Properties of Core Shell Structured Binary
(23) Li, W.; Li, C.; Ma, H.; Chen, J. Magnesium Nanowires: Mg@Ti and Ternary Mg@Ti@Ni Composites. Int. J. Hydrogen Energy
Enhanced Kinetics for Hydrogen Absorption and Desorption. J. Am. 2017, 42, 2239−2247.
Chem. Soc. 2007, 129, 6710−6711. (42) Fry, C. M. P.; Grant, D. M.; Walker, G. S. Improved Hydrogen
(24) Baldi, A.; Dam, B. Thin Film Metal Hydrides for Hydrogen Cycling Kinetics of Nano-Structured Magnesium/Transition Metal
Storage Applications. J. Mater. Chem. 2011, 21, 4021−4026. Multilayer Thin Films. Int. J. Hydrogen Energy 2013, 38, 982−990.
(25) Narehood, D. G.; Kishore, S.; Goto, H.; Adair, J. H.; Nelson, J. (43) Michel, K. J.; Ozoliņs,̌ V. Recent Advances in the Theory of
A.; Gutierrez, H. R.; Eklund, P. C. X-Ray Diffraction and H-Storage in Hydrogen Storage in Complex Metal Hydrides. MRS Bull. 2013, 38,
Ultra-Small Palladium Particles. Int. J. Hydrogen Energy 2009, 34, 462−472.
952−960. (44) de Jongh, P. E.; Allendorf, M.; Vajo, J. J.; Zlotea, C.
(26) Pasquini, L.; Sacchi, M.; Brighi, M.; Boelsma, C.; Bals, S.; Nanoconfined Light Metal Hydrides for Reversible Hydrogen
Perkisas, T.; Dam, B. Hydride Destabilization in Core-Shell Storage. MRS Bull. 2013, 38, 488−494.
Nanoparticles. Int. J. Hydrogen Energy 2014, 39, 2115−2123. (45) de Jongh, P. E.; Adelhelm, P. Nanosizing and Nanoconfine-
(27) Badri, V.; Hermann, A. M. Metal Hydride Batteries: Pd ment: New Strategies Towards Meeting Hydrogen Storage Goals.
Nanotube Incorporation into the Negative Electrode. Int. J. Hydrogen ChemSusChem 2010, 3, 1332−1348.
Energy 2000, 25, 249−253. (46) Mohtadi, R.; Orimo, S.-i. The Renaissance of Hydrides as
(28) Oguchi, H.; Heilweil, E. J.; Josell, D.; Bendersky, L. A. Infrared Energy Materials. Nat. Rev. Mater. 2017, 2, 16091.
Emission Imaging as a Tool for Characterization of Hydrogen Storage (47) Vajo, J. J. Influence of Nano-Confinement on the
Materials. J. Alloys Compd. 2009, 477, 8−15. Thermodynamics and Dehydrogenation Kinetics of Metal Hydrides.
(29) Fry, C. M. P.; Grant, D. M.; Walker, G. S. Catalysis and Curr. Opin. Solid State Mater. Sci. 2011, 15, 52−61.
Evolution on Cycling of Nano-Structured Magnesium Multilayer (48) Nielsen, T. K.; Besenbacher, F.; Jensen, T. R. Nanoconfined
Thin Films. Int. J. Hydrogen Energy 2014, 39, 1173−1184. Hydrides for Energy Storage. Nanoscale 2011, 3, 2086−2098.

10826 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(49) Niemann, M. U.; Srinivasan, S. S.; Phani, A. R.; Kumar, A.; (68) Rusman, N. A. A.; Dahari, M. A Review on the Current
Goswami, D. Y.; Stefanakos, E. K. Nanomaterials for Hydrogen Progress of Metal Hydrides Material for Solid-State Hydrogen Storage
Storage Applications: A Review. J. Nanomater. 2008, 950967. Applications. Int. J. Hydrogen Energy 2016, 41, 12108−12126.
(50) Baldé, C. P.; Hereijgers, B. P. C.; Bitter, J. H.; de Jong, K. P. (69) Harrison, D.; Thonhauser, T. Thermodynamics of Hydrogen
Sodium Alanate Nanoparticles − Linking Size to Hydrogen Storage Release in Complexed Borohydrides. Phys. Rev. Materials 2018, 2,
Properties. J. Am. Chem. Soc. 2008, 130, 6761−6765. 065403.
(51) de Jongh, P. E.; Wagemans, R. W.; Eggenhuisen, T. M.; (70) Gay-Lussac, J. L.; Thenard, L. J. Recherches Physico-
Dauvillier, B. S.; Radstake, P. B.; Meeldijk, J. D.; Geus, J. W.; de Jong, Chimiques; Paris: Paris, 1811.
K. P. The Preparation of Carbon-Supported Magnesium Nano- (71) Hautefeuille, P.; Troost, L. Alliages De H Avec Les Metaux.
particles Using Melt Infiltration. Chem. Mater. 2007, 19, 6052−6057. Ann. Chim. Phys. 1874, 2, 273.
(52) Aguey-Zinsou, K.-F.; Ares-Fernández, J.-R. Synthesis of (72) Hautefeuille, P.; Troost, L. Chaleurs De Combustion De
Colloidal Magnesium: A Near Room Temperature Store for Varietes De P Rouge. Compt. Rend. 1874, 78, 809.
Hydrogen. Chem. Mater. 2008, 20, 376−378. (73) Bowman, R. C.; Fultz, B. Metallic Hydrides I: Hydrogen
(53) Li, H.; Li, L.; Pedersen, A.; Gao, Y.; Khetrapal, N.; Jónsson, H.; Storage and Other Gas-Phase Applications. MRS Bull. 2002, 27, 688−
Zeng, X. C. Magic-Number Gold Nanoclusters with Diameters from 1 693.
to 3.5 nm: Relative Stability and Catalytic Activity for Co Oxidation. (74) Jepsen, L. H.; Paskevicius, M.; Jensen, T. R. In Nanostructured
Nano Lett. 2015, 15, 682−688. and Complex Hydrides for Hydrogen Storage; Raj, B., Voorde, M. V. d.,
(54) Rossin, A.; Tuci, G.; Luconi, L.; Giambastiani, G. Metal− Mahajan, Y., Eds.; Wiley: Weinheim, Germany, 2017.
Organic Frameworks as Heterogeneous Catalysts in Hydrogen (75) Bourgeois, N.; Crivello, J. C.; Cenedese, P.; Joubert, J. M.
Production from Lightweight Inorganic Hydrides. ACS Catal. 2017, Systematic First-Principles Study of Binary Metal Hydrides. ACS
7, 5035−5045. Comb. Sci. 2017, 19, 513−523.
(55) Yu, X. B.; Tang, Z. W.; Sun, D. L.; Ouyang, L. Z.; Zhu, M. (76) Belt, S. T.; Scaiano, J. C.; Whittlesey, M. K. Determination of
Recent Advances and Remaining Challenges of Nanostructured Metal Hydride and Metal-Ligand (L = CO, N2) Bond-Energies Using
Materials for Hydrogen Storage Applications. Prog. Mater. Sci. 2017, Photoacoustic Calorimetry. J. Am. Chem. Soc. 1993, 115, 1921−1925.
88, 1−48. (77) Elkind, J. L.; Armentrout, P. B. Transition-Metal Hydride
(56) Kan, H. M.; Sun, M.; Zhang, N.; Wang, X. Y.; Long, H. B. Bond-Energies - 1st and 2nd Row. Inorg. Chem. 1986, 25, 1078−1080.
Nanomaterials for Hydrogen Storage. Appl. Mech. Mater. 2014, 587− (78) Kiss, G.; Nolan, S. P.; Hoff, C. D. Direct Solution Calorimetric
589, 216−219. Measurements of Enthalpies of Proton and Electron-Transfer
(57) Varin, R. A.; Zbroniec, L.; Polanski, M.; Bystrzycki, J. A Review Reactions for Transition-Metal Complexes - Thermochemical Study
of Recent Advances on the Effects of Microstructural Refinement and of Metal-Hydride and Metal-Metal Bond-Energies. Inorg. Chim. Acta
Nano-Catalytic Additives on the Hydrogen Storage Properties of 1994, 227, 285−292.
Metal and Complex Hydrides. Energies 2011, 4, 1−25. (79) Labinger, J. A.; Bercaw, J. E. Metal Hydride and Metal Alkyl
(58) Sepehri, S.; Liu, Y.; Cao, G. In Advances in New Catalytic Bond Strengths - the Influence of Electronegativity Differences.
Materials; Wang, J. A., Cao, G. Z., Dominguez, J. M., Eds.; MDPI: Organometallics 1988, 7, 926−928.
Basel, Switzerland, 2010; Vol. 132. (80) Stevens, A. E.; Beauchamp, J. L. Determination of Metal-
(59) Lai, Q.; Wang, T.; Sun, Y.; Aguey-Zinsou, K. F. Rational Design Hydrogen Bond-Dissociation Energies by the Deprotonation of
of Nanosized Light Elements for Hydrogen Storage: Classes, Transition-Metal Hydride Ions - Application to MnH+. Chem. Phys.
Synthesis, Characterization, and Properties. Adv. Mater. Technol. Lett. 1981, 78, 291−295.
2018, 3, 1700298. (81) Wang, D. M.; Angelici, R. J. Metal-Hydrogen Bond
(60) Callini, E.; Aguey-Zinsou, K. F.; Ahuja, R.; Ares, J. R.; Bals, S.; Dissociation Enthalpies in Series of Complexes of Eight Different
Biliskov, N.; Chakraborty, S.; Charalambopoulou, G.; Chaudhary, A. Transition Metals. J. Am. Chem. Soc. 1996, 118, 935−942.
L.; Cuevas, F.; et al. Nanostructured Materials for Solid-State (82) Wiedner, E. S.; Chambers, M. B.; Pitman, C. L.; Bullock, R. M.;
Hydrogen Storage: A Review of the Achievement of COST Action Miller, A. J. M.; Appel, A. M. Thermodynamic Hydricity of Transition
MP1103. Int. J. Hydrogen Energy 2016, 41, 14404−14428. Metal Hydrides. Chem. Rev. 2016, 116, 8655−8692.
(61) Wang, L.; Rawal, A.; Aguey-Zinsou, K.-F. Hydrogen Storage (83) Joubert, J. M. Calphad Modeling of Metal-Hydrogen Systems:
Properties of Nanoconfined Aluminium Hydride (AlH3). Chem. Eng. A Review. JOM 2012, 64, 1438−1447.
Sci. 2018, DOI: 10.1016/j.ces.2018.02.014. (84) Cai, Y. Y.; Sun, X. L.; Huang, X. R. Transition Metal Hydrides
(62) Qiu, S.; Chu, H.; Zou, Y.; Xiang, C.; Xu, F.; Sun, L. Light Metal MH+/0/‑ (M = Sc - Zn): Benchmark Study and Periodic Trends.
Borohydrides/Amides Combined Hydrogen Storage Systems: Com- Comput. Theor. Chem. 2018, 1134, 15−21.
position, Structure and Properties. J. Mater. Chem. A 2017, 5, 25112− (85) Haussermann, U.; Blomqvist, H.; Noreus, D. Bonding and
25130. Stability of the Hydrogen Storage Material Mg2nih4. Inorg. Chem.
(63) Callini, E.; Atakli, Z. Ö . K.; Hauback, B. C.; Orimo, S.-i.; 2002, 41, 3684−3692.
Jensen, C.; Dornheim, M.; Grant, D.; Cho, Y. W.; Chen, P.; (86) Li, X.-H.; Ju, X.-H. Density Functional Theory Study on
Hjörvarsson, B.; et al. Complex and Liquid Hydrides for Energy (Mg(BH4))(n) (n = 1−4) Clusters as a Material for Hydrogen
Storage. Appl. Phys. A: Mater. Sci. Process. 2016, 122, 353. Storage. Comput. Theor. Chem. 2013, 1025, 46−51.
(64) Mao, J.; Gregory, D. H. Recent Advances in the Use of Sodium (87) Li, Y. W.; Liu, J. H.; Hou, C.; Shao, Y. X.; Qu, L. B.; Zhao, C.
Borohydride as a Solid State Hydrogen Store. Energies 2015, 8, 430− Y.; Ke, Z. F. Elucidating Metal Hydride Reactivity Using Late
453. Transition Metal Boryl and Borane Hydrides: 2c-2e Terminal
(65) De Rango, P.; Chaise, A.; Fruchart, D.; Miraglia, S.; Marty, P. Hydride, 3c-2e Bridging Hydride, and 3c-4e Bridging Hydride.
In Physics, Chemistry and Applications of Nanostructures; Borisenko, V. Catal. Sci. Technol. 2018, 8, 3395−3405.
E., Gaponenko, S. V., Gurin, V. S., Kam, C. H., Eds.; World Scientific: (88) Smithson, H.; Marianetti, C. A.; Morgan, D.; Van der Ven, A.;
Singapore, 2013. Predith, A.; Ceder, G. First-Principles Study of the Stability and
(66) Fichtner, M. Properties of Nanoscale Metal Hydrides. Electronic Structure of Metal Hydrides. Phys. Rev. B: Condens. Matter
Nanotechnology 2009, 20, 204009. Mater. Phys. 2002, 66, 144107.
(67) Sadhasivam, T.; Kim, H. T.; Jung, S.; Roh, S. H.; Park, J. H.; (89) Takagi, S.; Iijima, Y.; Sato, T.; Saitoh, H.; Ikeda, K.; Otomo, T.;
Jung, H. Y. Dimensional Effects of Nanostructured Mg/MgH2 for Miwa, K.; Ikeshoji, T.; Aoki, K.; Orimo, S. True Boundary for the
Hydrogen Storage Applications: A Review. Renewable Sustainable Formation of Homoleptic Transition-Metal Hydride Complexes.
Energy Rev. 2017, 72, 523−534. Angew. Chem., Int. Ed. 2015, 54, 5650−5653.

10827 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(90) Tang, S. Y.; Fu, Y.; Guo, Q. X. Ab Initio Calculation of M-H (110) Griessen, R.; Strohfeldt, N.; Giessen, H. Thermodynamics of
Bond Dissociation Energies of Cr-Group Metal Hydrides. Huaxue the Hybrid Interaction of Hydrogen with Palladium Nanoparticles.
Xuebao 2012, 70, 1923−1929. Nat. Mater. 2016, 15, 311.
(91) Xu, D. D.; Shan, C. H.; Li, Y. Z.; Qi, X. T.; Luo, X. L.; Bai, R. (111) Li, G. Q.; Kobayashi, H.; Dekura, S.; Ikeda, R.; Kubota, Y.;
P.; Lan, Y. Bond Dissociation Energy Controlled Sigma-Bond Kato, K.; Takata, M.; Yamamoto, T.; Matsumura, S.; Kitagawa, H.
Metathesis in Alkaline-Earth-Metal Hydride Catalyzed Dehydrocou- Shape-Dependent Hydrogen-Storage Properties in Pd Nanocrystals:
pling of Amines and Boranes: A Theoretical Study. Inorg. Chem. Front. Which Does Hydrogen Prefer, Octahedron (111) or Cube (100)? J.
2017, 4, 1813−1820. Am. Chem. Soc. 2014, 136, 10222−10225.
(92) Yoshino, M.; Komiya, K.; Takahashi, Y.; Shinzato, Y.; Yukawa, (112) Baldi, A.; Narayan, T. C.; Koh, A. L.; Dionne, J. A. In Situ
H.; Morinaga, M. Nature of the Chemical Bond in Complex Detection of Hydrogen-Induced Phase Transitions in Individual
Hydrides, NaAlH4, LiAlH4, LiBH4 and LiNH2. J. Alloys Compd. Palladium Nanocrystals. Nat. Mater. 2014, 13, 1143.
2005, 404-406, 185−190. (113) Zlotea, C.; Oumellal, Y.; Msakni, M.; Bourgon, J.; Bastide, S.;
(93) Alapati, S. V.; Johnson, J. K.; Sholl, D. S. Large-Scale Screening Cachet-Vivier, C.; Latroche, M. First Evidence of Rh Nano-Hydride
of Metal Hydride Mixtures for High-Capacity Hydrogen Storage from Formation at Low Pressure. Nano Lett. 2015, 15, 4752−4757.
First-Principles Calculations. J. Phys. Chem. C 2008, 112, 5258−5262. (114) Kobayashi, H.; Morita, H.; Yamauchi, M.; Ikeda, R.; Kitagawa,
(94) Kim, K. C.; Kulkarni, A. D.; Johnson, J. K.; Sholl, D. S. Large- H.; Kubota, Y.; Kato, K.; Takata, M. Nanosize-Induced Hydrogen
Scale Screening of Metal Hydrides for Hydrogen Storage from First- Storage and Capacity Control in a Non-Hydride-Forming Element:
Principles Calculations Based on Equilibrium Reaction Thermody- Rhodium. J. Am. Chem. Soc. 2011, 133, 11034−11037.
namics. Phys. Chem. Chem. Phys. 2011, 13, 7218−7229. (115) Kusada, K.; Kobayashi, H.; Ikeda, R.; Morita, H.; Kitagawa, H.
(95) Nicholson, K. M.; Sholl, D. S. First-Principles Screening of Changeover of the Thermodynamic Behavior for Hydrogen Storage in
Complex Transition Metal Hydrides for High Temperature Rh with Increasing Nanoparticle Size. Chem. Lett. 2013, 42, 55−56.
Applications. Inorg. Chem. 2014, 53, 11833−11848. (116) Zlotea, C.; Morfin, F.; Nguyen, T. S.; Nguyen, N. T.; Nelayah,
(96) Siegel, D. J.; Wolverton, C.; Ozoliņs,̌ V. Thermodynamic J.; Ricolleau, C.; Latroche, M.; Piccolo, L. Nanoalloying Bulk-
Guidelines for the Prediction of Hydrogen Storage Reactions and Immiscible Iridium and Palladium Inhibits Hydride Formation and
Their Application to Destabilized Hydride Mixtures. Phys. Rev. B: Promotes Catalytic Performances. Nanoscale 2014, 6, 9955−9959.
Condens. Matter Mater. Phys. 2007, 76, 134102. (117) Kobayashi, H.; Yamauchi, M.; Kitagawa, H. Finding
(97) Gunn, S. R.; Green, L. G. The Heats of Formation at 25- Hydrogen-Storage Capability in Iridium Induced by the Nanosize
Degrees of the Crystalline Hydrides and Deuterides and Aqueous Effect. J. Am. Chem. Soc. 2012, 134, 6893−6895.
Hydroxides of Lithium, Sodium and Potassium. J. Am. Chem. Soc. (118) Huot, J.; Ravnsbæk, D. B.; Zhang, J.; Cuevas, F.; Latroche, M.;
1958, 80, 4782−4786. Jensen, T. R. Mechanochemical Synthesis of Hydrogen Storage
(98) Messer, C. E.; Fasolino, L. G.; Thalmayer, C. E. The Heats of Materials. Prog. Mater. Sci. 2013, 58, 30−75.
Formation of Lithium, Sodium and Potassium Hydrides. J. Am. Chem. (119) Patelli, N.; Calizzi, M.; Migliori, A.; Morandi, V.; Pasquini, L.
Soc. 1955, 77, 4524−4526. Hydrogen Desorption Below 150 °C in MgH2−TiH2 Composite
(99) Gross, K. J.; Spatz, P.; Züttel, A.; Schlapbach, L. Mechanically Nanoparticles: Equilibrium and Kinetic Properties. J. Phys. Chem. C
Milled Mg Composites for Hydrogen Storage - the Transition to a 2017, 121, 11166−11177.
Steady State Composition. J. Alloys Compd. 1996, 240, 206−213. (120) Xie, L.; Liu, Y.; Wang, Y. T.; Zheng, J.; Li, X. G. Superior
(100) Pedersen, A. S.; Kjoller, J.; Larsen, B.; Vigeholm, B. Hydrogen Storage Kinetics of Mgh2 Nanoparticles Doped with TiF3.
Magnesium for Hydrogen Storage. Int. J. Hydrogen Energy 1983, 8, Acta Mater. 2007, 55, 4585−4591.
205−211. (121) Zhang, X.; Yang, R.; Yang, J.; Zhao, W.; Zheng, J.; Tian, W.;
(101) Post, M. L.; Murray, J. J.; Taylor, J. B. Metal Hydride Studies Li, X. Synthesis of Magnesium Nanoparticles with Superior Hydrogen
at the National Research Council of Canada. Int. J. Hydrogen Energy Storage Properties by Acetylene Plasma Metal Reaction. Int. J.
1984, 9, 137−145. Hydrogen Energy 2011, 36, 4967−4975.
(102) Reilly, J. J.; Wiswall, R. H. Reaction Hydrogen with Alloys (122) Dong, D.; Humphries, T. D.; Sheppard, D. A.; Stansby, B.;
Magnesium and Nickel and Formation of Mg2NiH4. Inorg. Chem. Paskevicius, M.; Sofianos, M. V.; Chaudhary, A. L.; Dornheim, M.;
1968, 7, 2254−2256. Buckley, C. E. Thermal Optimisation of Metal Hydride Reactors for
(103) Stampfer, J. F.; Holley, C. E.; Suttle, J. F. The Magnesium Thermal Energy Storage Applications. Sustainable Energy Fuels 2017,
Hydrogen System. J. Am. Chem. Soc. 1960, 82, 3504−3508. 1, 1820−1829.
(104) Tanguy, B.; Soubeyroux, J. L.; Pezat, M.; Portier, J.; (123) Sheppard, D. A.; Paskevicius, M.; Humphries, T. D.;
Hagenmuller, P. Improvement of Conditions for Synthesizing Felderhoff, M.; Capurso, G.; von Colbe, J. B.; Dornheim, M.;
Magnesium Hydride Using Adjuvants. Mater. Res. Bull. 1976, 11, Klassen, T.; Ward, P. A.; Teprovich, J. A.; et al. Metal Hydrides for
1441−1447. Concentrating Solar Thermal Power Energy Storage. Appl. Phys. A:
(105) Miwa, K.; Fukumoto, A. First-Principles Study on 3d Mater. Sci. Process. 2016, 122, 395.
Transition-Metal Dihydrides. Phys. Rev. B: Condens. Matter Mater. (124) Javadian, P.; Sheppard, D. A.; Jensen, T. R.; Buckley, C. E.
Phys. 2002, 65, 155114. Destabilization of Lithium Hydride and the Thermodynamic
(106) Vons, V. A.; Leegwater, H.; Legerstee, W. J.; Eijt, S. W. H.; Assessment of the Li-Al-H System for Solar Thermal Energy Storage.
Schmidt-Ott, A. Hydrogen Storage Properties of Spark Generated RSC Adv. 2016, 6, 94927−94933.
Palladium Nanoparticles. Int. J. Hydrogen Energy 2010, 35, 5479− (125) Corgnale, C.; Hardy, B.; Motyka, T.; Zidan, R. Metal Hydride
5489. Based Thermal Energy Storage System Requirements for High
(107) Yamauchi, M.; Ikeda, R.; Kitagawa, H.; Takata, M. Nanosize Performance Concentrating Solar Power Plants. Int. J. Hydrogen
Effects on Hydrogen Storage in Palladium. J. Phys. Chem. C 2008, Energy 2016, 41, 20217−20230.
112, 3294−3299. (126) Qu, X. H.; Li, Y.; Li, P.; Wan, Q.; Zhai, F. Q. The
(108) Dai, B.; Sholl, D. S.; Johnson, J. K. First-Principles Study of Development of Metal Hydrides Using as Concentrating Solar
Experimental and Hypothetical Mg(BH4)2 Crystal Structures. J. Phys. Thermal Storage Materials. Front. Mater. Sci. 2015, 9, 317−331.
Chem. C 2008, 112, 4391−4395. (127) Paskevicius, M.; Sheppard, D. A.; Williamson, K.; Buckley, C.
(109) Bardhan, R.; Hedges, L. O.; Pint, C. L.; Javey, A.; Whitelam, E. Metal Hydride Thermal Heat Storage Prototype for Concentrating
S.; Urban, J. J. Uncovering the Intrinsic Size Dependence of Solar Thermal Power. Energy 2015, 88, 469−477.
Hydriding Phase Transformations in Nanocrystals. Nat. Mater. (128) Corgnale, C.; Hardy, B.; Motyka, T.; Zidan, R.; Teprovich, J.;
2013, 12, 905. Peters, B. Screening Analysis of Metal Hydride Based Thermal Energy

10828 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Storage Systems for Concentrating Solar Power Plants. Renewable (150) Fedneva, E. M.; Alpatova, V. I.; Mikheeva, V. I.
Sustainable Energy Rev. 2014, 38, 821−833. Termicheskaya Ustoichivost Borogidrida Litiya. Zh. Neorg. Khim.
(129) Harries, D. N.; Paskevicius, M.; Sheppard, D. A.; Price, T. E. 1964, 9, 1519−1520.
C.; Buckley, C. E. Concentrating Solar Thermal Heat Storage Using (151) Züttel, A.; Wenger, P.; Rentsch, S.; Sudan, P.; Mauron, P.;
Metal Hydrides. Proc. IEEE 2012, 100, 539−549. Emmenegger, C. LiBH4 a New Hydrogen Storage Material. J. Power
(130) Zolliker, P.; Yvon, K.; Jorgensen, J. D.; Rotella, F. J. Structural Sources 2003, 118, 1−7.
Studies of the Hydrogen Storage Material Magnesium Nickel Hydride (152) Mosegaard, L.; Møller, B.; Jørgensen, J.-E.; Filinchuk, Y.;
(Mg2NiH4). 2. Monoclinic Low-Temperature Structure. Inorg. Chem. Cerenius, Y.; Hanson, J. C.; Dimasi, E.; Besenbacher, F.; Jensen, T. R.
1986, 25, 3590−3593. Reactivity of LiBH4: In Situ Synchrotron Radiation Powder X-Ray
(131) Aizawa, T.; Kuji, T.; Nakano, H. Non-Equilibration of Diffraction Study. J. Phys. Chem. C 2008, 112, 1299−1303.
Nanostructured Mg2ni by Bulk Mechanical Alloying. Mater. Trans. (153) Ohba, N.; Miwa, K.; Aoki, M.; Noritake, T.; Towata, S.-i.;
2001, 42, 1284−1292. Nakamori, Y.; Orimo, S.-i.; Züttel, A. First-Principles Study on the
(132) Okada, M.; Kuriiwa, T.; Kamegawa, A.; Takamura, H. Role of Stability of Intermediate Compounds of LiBH4. Phys. Rev. B: Condens.
Intermetallics in Hydrogen Storage Materials. Mater. Sci. Eng., A 2002, Matter Mater. Phys. 2006, 74, 075110.
329−331, 305−312. (154) Li, H. W.; Miwa, K.; Ohba, N.; Fujita, T.; Sato, T.; Yan, Y.;
(133) U.S. Department of Energy Fuel Cell Technologies Office Towata, S.; Chen, M. W.; Orimo, S. Formation of an Intermediate
Hydrogen Storage Materials Database, 2018; http:// Compound with a B12H12 Cluster: Experimental and Theoretical
hydrogenmaterialssearch.govtools.us/. Studies on Magnesium Borohydride Mg(BH4)2. Nanotechnology 2009,
(134) Martinez, M. A. R.; Andre-Filho, J.; Felix, L. L.; Coaquira, J. A. 20, 204013.
H.; Garg, V. K.; Oliveira, A. C.; Mestnik-Filho, J. Structural and (155) Orimo, S.; Nakamori, Y.; Kitahara, G.; Miwa, K.; Ohba, N.;
Mossbauer Spectroscopy Characterization of Bulk and Nanostruc- Towata, S.; Züttel, A. Dehydriding and Rehydriding Reactions of
tured TiFe0.5Ni0.5/Graphite Compounds and Their Hydrides. Hyper- LiBH4. J. Alloys Compd. 2005, 404-406, 427−430.
fine Interact. 2015, 232, 149−156. (156) Zavorotynska, O.; El-Kharbachi, A.; Deledda, S.; Hauback, B.
(135) Wu, Y.; Lototsky, M. V.; Solberg, J. K.; Yartys, V. A. C. Recent Progress in Magnesium Borohydride Mg(BH 4)2 :
Microstructural Evolution and Improved Hydrogenation-Dehydro- Fundamentals and Applications for Energy Storage. Int. J. Hydrogen
genation Kinetics of Nanostructured Melt-Spun Mg-Ni-Mm Alloys. J. Energy 2016, 41, 14387−14403.
Alloys Compd. 2011, 509, S640−S645. (157) Severa, G.; Ronnebro, E.; Jensen, C. M. Direct Hydrogenation
(136) Christian, M.; Aguey-Zinsou, K.-F. Destabilisation of Complex of Magnesium Boride to Magnesium Borohydride: Demonstration of
Hydrides through Size Effects. Nanoscale 2010, 2, 2587−2590. > 11 Weight Percent Reversible Hydrogen Storage. Chem. Commun.
(137) Klebanoff, L. Hydrogen Storage Technology: Materials and 2010, 46, 421−423.
Applications; CRC Press: Boca Raton, FL, 2012. (158) Liang, G. X.; Huot, J.; Schulz, R. In Metastable, Mechanically
(138) Mauron, P.; Buchter, F.; Friedrichs, O.; Remhof, A.; Bielmann,
Alloyed and Nanocrystalline Materials; Ma, E., Atzmon, M., Koch, C.
M.; Zwicky, C. N.; Züttel, A. Stability and Reversibility of LiBH4. J.
C., Eds.; MRS Bulletin, 2002; Vol. 386−3.
Phys. Chem. B 2008, 112, 906−910.
(159) Huang, X.; Xiao, X. Z.; Zhang, W.; Fan, X. L.; Zhang, L. T.;
(139) Stavila, V.; Klebanoff, L. E.; Vajo, J. J.; Chen, P. In Hydrogen
Cheng, C. J.; Li, S. Q.; Ge, H. W.; Wang, Q. D.; Chen, L. X.
Storage Technology; Klebanoff, L. E., Ed.; Taylor and Francis: Boca
Transition Metal (Co, Ni) Nanoparticles Wrapped with Carbon and
Raton, FL, 2012.
Their Superior Catalytic Activities for the Reversible Hydrogen
(140) Jang, J.-W.; Shim, J.-H.; Cho, Y. W.; Lee, B.-J. Thermody-
Storage of Magnesium Hydride. Phys. Chem. Chem. Phys. 2017, 19,
namic Calculation of LiH↔Li3AlH6↔LiAlH4 Reactions. J. Alloys
Compd. 2006, 420, 286−290. 4019−4029.
(141) Graetz, J.; Wegrzyn, J.; Reilly, J. J. Regeneration of Lithium (160) Huang, X.; Xiao, X. Z.; Shao, J.; Zhai, B.; Fan, X. L.; Cheng, C.
Aluminum Hydride. J. Am. Chem. Soc. 2008, 130, 17790−17794. J.; Li, S. Q.; Ge, H. W.; Wang, Q. D.; Chen, L. X. Building Robust
(142) Baldé, C. P.; Stil, H. A.; van der Eerden, A. M. J.; de Jong, K. Architectures of Carbon-Wrapped Transition Metal Nanoparticles for
P.; Bitter, J. H. Active Ti Species in TiCl3-Doped NaAlH4. Mechanism High Catalytic Enhancement of the 2LiBH4-MgH2 System for
for Catalyst Deactivation. J. Phys. Chem. C 2007, 111, 2797−2802. Hydrogen Storage Cycling Performance. Nanoscale 2016, 8, 14898−
(143) Wood, B. C.; Marzari, N. Dynamics and Thermodynamics of a 14908.
Novel Phase of NaAlH4. Phys. Rev. Lett. 2009, 103, 185901. (161) El-Eskandarany, M. S.; Shaban, E. Contamination Effects on
(144) Majzoub, E. H.; Zhou, F.; Ozoliņs,̌ V. First-Principles Improving the Hydrogenation/Dehydrogenation Kinetics of Binary
Calculated Phase Diagram for Nanoclusters in the Na−Al−H System: Magnesium Hydride/Titanium Carbide Systems Prepared by
A Single-Step Decomposition Pathway for NaAlH4. J. Phys. Chem. C Reactive Ball Milling. Materials 2015, 8, 6880−6892.
2011, 115, 2636−2643. (162) Liu, S.-S.; Li, Z.-B.; Jiao, C.-L.; Si, X.-L.; Yang, L.-N.; Zhang,
(145) Baldé, C. P.; Leynaud, O.; Barnes, P.; Pelaez-Jimenez, E.; de J.; Zhou, H.-Y.; Huang, F.-L.; Gabelica, Z.; Schick, C.; et al. Improved
Jong, K. P.; Bitter, J. H. Towards a Structure-Performance Reversible Hydrogen Storage of LiAlH4 by Nano-Sized TiH2. Int. J.
Relationship for Hydrogen Storage in Ti-Doped NaAlH4 Nano- Hydrogen Energy 2013, 38, 2770−2777.
particles. Chem. Commun. 2011, 47, 2143−2145. (163) Varin, R. A.; Zbroniec, L. The Effects of Ball Milling and
(146) Pitt, M. P.; Vullum, P. E.; Sorby, M. H.; Blanchard, D.; Sulic, Nanometric Nickel Additive on the Hydrogen Desorption from
M. P.; Emerich, H.; Paskevicius, M.; Buckley, C. E.; Walmsley, J.; Lithium Borohydride and Manganese Chloride (3LiBH4 + MnCl2)
Holmestad, R.; et al. The Location of Ti Containing Phases after the Mixture. Int. J. Hydrogen Energy 2010, 35, 3588−3597.
Completion of the NaAlH4 + xTiCl3 Milling Process. J. Alloys Compd. (164) Hanada, N.; Hirotoshi, E.; Ichikawa, T.; Akiba, E.; Fujii, H.
2012, 513, 597−605. Sem and Tem Characterization of Magnesium Hydride Catalyzed
(147) Fan, X.; Xiao, X.; Chen, L.; Zhang, L.; Shao, J.; Li, S.; Ge, H.; with Ni Nano-Particle or Nb2O5. J. Alloys Compd. 2008, 450, 395−
Wang, Q. Significantly Improved Hydrogen Storage Properties of 399.
Naalh4 Catalyzed by Ce-Based Nanoparticles. J. Mater. Chem. A 2013, (165) Varin, R. A.; Czujko, T.; Wasmund, E. B.; Wronski, Z. S.
1, 9752−9759. Hydrogen Desorption Properties of MgH2 Nanocomposites with
(148) Delmelle, R.; Gehrig, J. C.; Borgschulte, A.; Züttel, A. Nano-Oxides and Inco Micrometric- and Nanometric-Ni. J. Alloys
Reactivity Enhancement of Oxide Skins in Reversible Ti-Doped Compd. 2007, 446-447, 63−66.
NaAlH4. AIP Adv. 2014, 4, 127130. (166) Varin, R. A.; Czujko, T.; Wasmund, E. B.; Wronski, Z. S.
(149) Frankcombe, T. J. Proposed Mechanisms for the Catalytic Catalytic Effects of Various Forms of Nickel on the Synthesis Rate
Activity of Ti in NaAlH4. Chem. Rev. 2012, 112, 2164−2178. and Hydrogen Desorption Properties of Nanocrystalline Magnesium

10829 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Hydride (MgH2) Synthesized by Controlled Reactive Mechanical (186) Paskevicius, M.; Sheppard, D. A.; Buckley, C. E.
Milling (CRMM). J. Alloys Compd. 2007, 432, 217−231. Thermodynamic Changes in Mechanochemically Synthesized Mag-
(167) Hanada, N.; Ichikawa, T.; Fujii, H. Catalytic Effect of nesium Hydride Nanoparticles. J. Am. Chem. Soc. 2010, 132, 5077−
Nanoparticle 3d-Transition Metals on Hydrogen Storage Properties 5083.
in Magnesium Hydride MgH2 Prepared by Mechanical Milling. J. (187) Kodera, Y.; Yamasaki, N.; Miki, J.; Ohyanagi, M.; Shiozaki, S.;
Phys. Chem. B 2005, 109, 7188−7194. Fukui, S.; Yin, J.; Fukui, T. In Advances in Energy Materials; Dogan, F.,
(168) Zaluska, A.; Zaluski, L.; Strom-Olsen, J. O. Structure, Catalysis Manjooran, N., Eds.; John Wiley & Sons: New York, 2009; Vol. 205.
and Atomic Reactions on the Nano-Scale: A Systematic Approach to (188) de Rango, P.; Chaise, A.; Charbonnier, J.; Fruchart, D.; Jehan,
Metal Hydrides for Hydrogen Storage. Appl. Phys. A: Mater. Sci. M.; Marty, P.; Miraglia, S.; Rivoirard, S.; Skryabina, N. Nano-
Process. 2001, 72, 157−165. structured Magnesium Hydride for Pilot Tank Development. J. Alloys
(169) Kortshagen, U. R.; Sankaran, R. M.; Pereira, R. N.; Girshick, S. Compd. 2007, 446-447, 52−57.
L.; Wu, J. J.; Aydil, E. S. Nonthermal Plasma Synthesis of (189) Huot, J.; Liang, G.; Schulz, R. Mechanically Alloyed Metal
Nanocrystals: Fundamental Principles, Materials, and Applications. Hydride Systems. Appl. Phys. A: Mater. Sci. Process. 2001, 72, 187−
Chem. Rev. 2016, 116, 11061−11127. 195.
(170) Biasetti, A.; Meyer, M.; Zelis, L. M. Hydriding Kinetics of Mg- (190) Varin, R. A.; Czujko, T.; Chiu, C.; Wronski, Z. Particle Size
TiH2 Fine Dispersions Obtained by Mechanosynthesis. Powder Effects on the Desorption Properties of Nanostructured Magnesium
Technol. 2017, 307, 145−152. Dihydride (MgH2) Synthesized by Controlled Reactive Mechanical
(171) Zepon, G.; Leiva, D. R.; Kaufman, M. J.; Figueroa, S. J. A.; Milling (CRMM). J. Alloys Compd. 2006, 424, 356−364.
Floriano, R.; Lamas, D. G.; Asselli, A. A. C.; Botta, W. J. Controlled (191) Vijay, R.; Sundaresan, R.; Maiya, M. P.; Srinivasa Murthy, S.;
Mechanochemical Synthesis and Hydrogen Desorption Mechanisms Fu, Y.; Klein, H. P.; Groll, M. Characterisation of Mg−x Wt.% FeTi (x
of Nanostructured Mg2CoH5. Int. J. Hydrogen Energy 2015, 40, 1504− = 5−30) and Mg−40wt.% FeTiMn Hydrogen Absorbing Materials
1515. Prepared by Mechanical Alloying. J. Alloys Compd. 2004, 384, 283−
(172) Nachev, S.; de Rango, P.; Skryabina, N.; Skachkov, A.; 295.
Aptukov, V.; Fruchart, D.; Marty, P. Mechanical Behavior of Highly (192) Wang, X. L.; Tu, J. P.; Wang, C. H.; Zhang, X. B.; Chen, C. P.;
Reactive Nanostructured MgH2. Int. J. Hydrogen Energy 2015, 40, Zhao, X. B. Hydrogen Storage Properties of Nanocrystalline Mg−Ce/
17065−17074. Ni Composite. J. Power Sources 2006, 159, 163−166.
(173) Luo, X.; Grant, D. M.; Walker, G. S. Catalytic Effect of Nano- (193) Wang, X. L.; Tu, J. P.; Zhang, X. B.; Chen, C. P.; Zhao, X. B.
Sized ScH2 on the Hydrogen Storage of Mechanically Milled MgH2. J. Hydrogenation Properties of Mg/Mg2Ni0.8Cr0.2 Composites Contain-
Alloys Compd. 2015, 622, 842−850. ing Tio2 Nanoparticles. J. Alloys Compd. 2005, 404−406, 529−532.
(174) Santos, S. F.; Ishikawa, T. T.; Botta, W. J.; Huot, J. MgH2 + (194) Varin, R. A.; Zbroniec, L. Catalytic Effects of Nanometric Ni
FeNb Nanocomposites for Hydrogen Storage. Mater. Chem. Phys. (N-Ni) on the Dehydrogenation of Ball Milled Sodium Alanate
2014, 147, 557−562. (NaAlH4). Nanosci. Nanotechnol. Lett. 2012, 4, 149−159.
(175) Biasetti, A.; Meyer, M.; Mendoza Zelis, L. Formation Kinetics (195) Varin, R. A.; Zbroniec, L. The Effects of Nanometric Nickel
and Microstructure of Mg-Ti Hydrides Made by Reactive Ball Milling. (N-Ni) Catalyst on the Dehydrogenation and Rehydrogenation
Int. J. Hydrogen Energy 2014, 39, 8767−8771. Behavior of Ball Milled Lithium Alanate (LiAlH4). J. Alloys Compd.
(176) Luo, X.; Grant, D. M.; Walker, G. S. Hydrogen Storage 2010, 506, 928−939.
Properties of Nano-Structured 0.65MgH2/0.35ScH2. Int. J. Hydrogen (196) Varin, R. A.; Zbroniec, L. Decomposition Behavior of
Energy 2013, 38, 153−161. Unmilled and Ball Milled Lithium Alanate (LiAlH4) Including
(177) Leiva, D. R.; Zepon, G.; Asselli, A. A. C.; Fruchart, D.; Long-Term Storage and Moisture Effects. J. Alloys Compd. 2010,
Miraglia, S.; Ishikawa, T. T.; Botta, W. J. Mechanochemistry and H- 504, 89−101.
Sorption Properties of Mg2FeH6-Based Nanocomposites. Int. J. Mater. (197) Amama, P. B.; Grant, J. T.; Shamberger, P. J.; Voevodin, A. A.;
Res. 2012, 103, 1147−1154. Fisher, T. S. Improved Dehydrogenation Properties of Ti-Doped
(178) Nowak, M.; Jurczyk, M. Mg-Based Nanocomposites for Room LiAlH4: Role of Ti Precursors. J. Phys. Chem. C 2012, 116, 21886−
Temperature Hydrogen Storage. Phys. Status Solidi A 2010, 207, 21894.
1144−1147. (198) Varin, R. A.; Zbroniec, L. Fast and Slow Dehydrogenation of
(179) Teresiak, A.; Uhlemann, M.; Gebert, A.; Thomas, J.; Eckert, J.; Ball Milled Lithium Alanate (LiAlH4) Catalyzed with Manganese
Schultz, L. Formation of Nanostructured LaMg2Ni by Rapid Chloride (MnCl2) as Compared to Nanometric Nickel Catalyst. J.
Quenching and Intensive Milling and Its Hydrogen Reactivity. J. Alloys Compd. 2011, 509, S736−S739.
Alloys Compd. 2009, 481, 144−151. (199) Li, L.; Xu, Y. N.; Wang, Y.; Wang, Y. J.; Qiu, F. Y.; An, C. H.;
(180) Wronski, Z.; Varin, R. A.; Chiu, C.; Czujko, T.; Calka, A. Jiao, L. F.; Yuan, H. T. Nbn Nanoparticles as Additive for the High
Mechanochemical Synthesis of Nanostructured Chemical Hydrides in Dehydrogenation Properties of LiAlH4. Dalton Trans 2014, 43,
Hydrogen Alloying Mills. J. Alloys Compd. 2007, 434-435, 743−746. 1806−1813.
(181) Austin, L. G.; Klimpel, R. R. The Theory of Grinding (200) Zhang, B. J.; Liu, B. H. Hydrogen Desorption from LiBH4
Operations. Ind. Eng. Chem. 1964, 56, 18−29. Destabilized by Chlorides of Transition Metal Fe, Co, and Ni. Int. J.
(182) Kim, Y. D.; Chung, J. Y.; Kim, J.; Jeon, H. Formation of Hydrogen Energy 2010, 35, 7288−7294.
Nanocrystalline Fe−Co Powders Produced by Mechanical Alloying. (201) Au, M.; Walters, R. T. Reversibility Aspect of Lithium
Mater. Sci. Eng., A 2000, 291, 17−21. Borohydrides. Int. J. Hydrogen Energy 2010, 35, 10311−10316.
(183) Ortiz, A. L.; Osborn, W.; Markmaitree, T.; Shaw, L. L. (202) Huttel, Y. Gas-Phase Synthesis of Nanoparticles; Wiley-VCH:
Crystallite Sizes of LiH before and after Ball Milling and Thermal Weinheim, Germany, 2017.
Exposure. J. Alloys Compd. 2008, 454, 297−305. (203) Grammatikopoulos, P.; Steinhauer, S.; Vernieres, J.; Singh, V.;
(184) Ç akmak, G.; Károly, Z.; Mohai, I.; Ö ztürk, T.; Szépvölgyi, J. Sowwan, M. Nanoparticle Design by Gas-Phase Synthesis. Adv. Phys.
The Processing of Mg−Ti for Hydrogen Storage; Mechanical Milling X 2016, 1, 81−100.
and Plasma Synthesis. Int. J. Hydrogen Energy 2010, 35, 10412− (204) Callini, E.; Pasquini, L.; Jensen, T. R.; Bonetti, E. Hydrogen
10418. Storage Properties of Mg-Ni Nanoparticles. Int. J. Hydrogen Energy
(185) Pang, Y. P.; Liu, Y. F.; Gao, M. X.; Ouyang, L. Z.; Liu, J. W.; 2013, 38, 12207−12212.
Wang, H.; Zhu, M.; Pan, H. G. A Mechanical-Force-Driven Physical (205) Callini, E.; Pasquini, L.; Rude, L. H.; Nielsen, T. K.; Jensen, T.
Vapour Deposition Approach to Fabricating Complex Hydride R.; Bonetti, E. Hydrogen Storage and Phase Transformations in Mg-
Nanostructures. Nat. Commun. 2014, 5, 3519. Pd Nanoparticles. J. Appl. Phys. 2010, 108, 073513.

10830 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(206) Calizzi, M.; Venturi, F.; Ponthieu, M.; Cuevas, F.; Morandi, Ostwald Ripening at Room Temperature in a Pd Nanocluster Film.
V.; Perkisas, T.; Bals, S.; Pasquini, L. Gas-Phase Synthesis of Mg-Ti Phys. Rev. Lett. 2008, 100, 236105.
Nanoparticles for Solid-State Hydrogen Storage. Phys. Chem. Chem. (227) Di Vece, M.; Kelly, J. J.; Lievens, P. Inhomogeneous Phase
Phys. 2016, 18, 141−148. Transition Upon Hydrogenation of Nanocluster Pd Films. Chem-
(207) Shekhtman, V. S.; Dolukhanyan, S. K.; Abrosimova, G. E.; PhysChem 2009, 10, 512−515.
Abrahamyan, K. A.; Aleksanyan, A. G.; Aghajanyan, N. N.; Ter- (228) Vredenberg, A. M.; Heller, E. M. B.; Boerma, D. O. Hydriding
Galstyan, O. P. The Nanocrystalline Forming by Combustion Characteristics of FeTi/Pd Films. J. Alloys Compd. 2005, 400, 188−
Synthesis of Ti (Zr) Hydrides. Int. J. Hydrogen Energy 2001, 26, 193.
435−440. (229) Tan, Z.; Chiu, C.; Heilweil, E. J.; Bendersky, L. A.
(208) Ostrikov, K.; Cvelbar, U.; Murphy, A. B. Plasma Nanoscience: Thermodynamics, Kinetics and Microstructural Evolution During
Setting Directions, Tackling Grand Challenges. J. Phys. D: Appl. Phys. Hydrogenation of Iron-Doped Magnesium Thin Films. Int. J.
2011, 44, 174001. Hydrogen Energy 2011, 36, 9702−9713.
(209) Ohno, S.; Uda, M. Generation Rate of Ultrafine Metal (230) Agarwal, S.; Jain, A.; Jain, P.; Vyas, D.; Ganesan, V.; Jain, I. P.
Particles in Hydrogen Plasma-Metal Reaction. Nippon Kinzoku Synthesis of Nano-Crystalline Zr-M (M = Ni, Co, Fe, Cu) Bilayer
Gakkaishi 1984, 48, 640−646. Films and Their Thermodynamics of Hydrogen Uptake by Resistance
(210) Shao, H.; Wang, Y.; Xu, H.; Li, X. Hydrogen Storage
Measurement. Int. J. Hydrogen Energy 2010, 35, 9893−9900.
Properties of Magnesium Ultrafine Particles Prepared by Hydrogen
(231) Gautam, Y. K.; Chawla, A. K.; Walia, R.; Agrawal, R. D.;
Plasma-Metal Reaction. Mater. Sci. Eng., B 2004, 110, 221−226.
Chandra, R. Hydrogenation of Pd-Capped Mg Thin Films Prepared
(211) Kooi, B. J.; Palasantzas, G.; De Hosson, J. T. M. Gas-Phase
Synthesis of Magnesium Nanoparticles: A High-Resolution Trans- by Dc Magnetron Sputtering. Appl. Surf. Sci. 2011, 257, 6291−6295.
(232) Gautam, Y. K.; Chawla, A. K.; Khan, S. A.; Agrawal, R. D.;
mission Electron Microscopy Study. Appl. Phys. Lett. 2006, 89,
161914. Chandra, R. Hydrogen Absorption and Optical Properties of Pd/Mg
(212) Liu, T.; Chen, C.; Wang, F.; Li, X. Enhanced Hydrogen Thin Films Prepared by Dc Magnetron Sputtering. Int. J. Hydrogen
Storage Properties of Magnesium by the Synergic Catalytic Effect of Energy 2012, 37, 3772−3778.
TiH1.971 and TiH1.5 Nanoparticles at Room Temperature. J. Power (233) Jensen, I. J. T.; Thogersen, A.; Lovvik, O. M.; Schreuders, H.;
Sources 2014, 267, 69−77. Dam, B.; Diplas, S. X-Ray Photoelectron Spectroscopy Investigation
(213) Lee, G.-G.; Park, J.-S. Hydrogen Sorption Property of of Magnetron Sputtered Mg-Ti-H Thin Films. Int. J. Hydrogen Energy
Zr55V29Fe16 Nanopowder Synthesized by the Plasma Arc Discharge 2013, 38, 10704−10715.
Process. Mater. Trans. 2007, 48, 1566−1570. (234) Gupta, D.; Barman, P. B.; Hazra, S. K. Hydrogen Response of
(214) Shao, H.; Liu, T.; Wang, Y.; Xu, H.; Li, X. Preparation of Mg- Porous Palladium Nano-Films. AIP Conf. Proc. 2015, 1675, 030001.
Based Hydrogen Storage Materials from Metal Nanoparticles. J. Alloys (235) Zhu, K.; Ju, Y. M.; Xu, J. J.; Yang, Z. Y.; Gao, S.; Hou, Y. L.
Compd. 2008, 465, 527−533. Magnetic Nanomaterials: Chemical Design, Synthesis, and Potential
(215) Shao, H.; Xu, H.; Wang, Y.; Li, X. Preparation and Hydrogen Applications. Acc. Chem. Res. 2018, 51, 404−413.
Storage Properties of Mg2Ni Intermetallic Nanoparticles. Nano- (236) Saldan, I.; Semenyuk, Y.; Marchuk, I.; Reshetnyak, O.
technology 2004, 15, 269. Chemical Synthesis and Application of Palladium Nanoparticles. J.
(216) Wirth, E.; Milcius, D.; Filiou, C.; Noréus, D. Exploring the Mater. Sci. 2015, 50, 2337−2354.
Hydrogen Sorption Capacity of Mg−Ni Powders Produced by the (237) Wang, H. L.; Dai, H. J. Strongly Coupled Inorganic-Nano-
Vapour Deposition Technique. Int. J. Hydrogen Energy 2008, 33, Carbon Hybrid Materials for Energy Storage. Chem. Soc. Rev. 2013,
3122−3127. 42, 3088−3113.
(217) Zou, J.; Sun, H.; Zeng, X.; Ji, G.; Ding, W. Preparation and (238) Long, N. V.; Yang, Y.; Thi, C. M.; Minh, N. V.; Cao, Y. Q.;
Hydrogen Storage Properties of Mg-Rich Mg-Ni Ultrafine Particles. J. Nogami, M. The Development of Mixture, Alloy, and Core-Shell
Nanomater. 2012, 2012, 8. Nanocatalysts with Nanomaterial Supports for Energy Conversion in
(218) Shao, H.; Wang, Y.; Xu, H.; Li, X. Preparation and Hydrogen Low-Temperature Fuel Cells. Nano Energy 2013, 2, 636−676.
Storage Properties of Nanostructured Mg2Cu Alloy. J. Solid State (239) Liu, X. W.; Wang, D. S.; Li, Y. D. Synthesis and Catalytic
Chem. 2005, 178, 2211−2217. Properties of Bimetallic Nanomaterials with Various Architectures.
(219) Liu, T.; Shao, H.; Li, X. Synthesis and Characteristics of Ti-Fe Nano Today 2012, 7, 448−466.
Nanoparticles by Hydrogen Plasma-Metal Reaction. Intermetallics (240) Alonso, F.; Riente, P.; Yus, M. Nickel Nanoparticles in
2004, 12, 97−102. Hydrogen Transfer Reactions. Acc. Chem. Res. 2011, 44, 379−391.
(220) Liu, T.; Qin, C.; Zhang, T.; Cao, Y.; Zhu, M.; Li, X. Synthesis (241) Norberg, N. S.; Arthur, T. S.; Fredrick, S. J.; Prieto, A. L. Size-
of Mg@Mg17Al12 Ultrafine Particles with Superior Hydrogen Storage Dependent Hydrogen Storage Properties of Mg Nanocrystals
Properties by Hydrogen Plasma-Metal Reaction. J. Mater. Chem.
Prepared from Solution. J. Am. Chem. Soc. 2011, 133, 10679−10681.
2012, 22, 19831−19838.
(242) Liu, Y.; Zou, J.; Zeng, X.; Wu, X.; Li, D.; Ding, W. Hydrogen
(221) Liu, T.; Zhang, T.; Zhang, X.; Li, X. Synthesis and Hydrogen
Storage Properties of a Mg-Ni Nanocomposite Coprecipitated from
Storage Properties of Ultrafine Mg−Zn Particles. Int. J. Hydrogen
Solution. J. Phys. Chem. C 2014, 118, 18401−18411.
Energy 2011, 36, 3515−3520.
(243) Sun, Y.; Aguey-Zinsou, K. F. Synthesis of Magnesium
(222) Zou, J.; Guo, H.; Zeng, X.; Zhou, S.; Chen, X.; Ding, W.
Hydrogen Storage Properties of Mg-TM-La (TM = Ti, Fe, Ni) Nanofibers by Electroless Reduction and Their Hydrogen Interaction
Ternary Composite Powders Prepared through Arc Plasma Method. Properties. Part. Part. Syst. Char 2017, 34, 1600276.
Int. J. Hydrogen Energy 2013, 38, 8852−8862. (244) Borgese, L.; Zanola, P.; Bontempi, E.; Rossi, D.; Depero, L. E.
(223) Xie, L.; Zheng, J.; Liu, Y.; Li, Y.; Li, X. Synthesis of Li2NH In Situ XRD Characterization of Hydrogen Desorption from
Hollow Nanospheres with Superior Hydrogen Storage Kinetics by Electrochemically Deposited Pd Coating. J. Coat. Technol. Res.
Plasma Metal Reaction. Chem. Mater. 2008, 20, 282−286. 2010, 7, 691−695.
(224) Xie, L.; Li, Y.; Yang, R.; Liu, Y.; Li, X. Superior Hydrogen (245) Haas, I.; Gedanken, A. Synthesis of Metallic Magnesium
Desorption Kinetics of Mg(NH2)2 Hollow Nanospheres Mixed with Nanoparticles by Sonoelectrochemistry. Chem. Commun. 2008, 0,
MgH2 Nanoparticles. Appl. Phys. Lett. 2008, 92, 231910. 1795−1797.
(225) Suzuki, K. Structure and Properties of Amorphous Metal (246) Viyannalage, L.; Lee, V.; Dennis, R. V.; Kapoor, D.; Haines, C.
Hydrides. J. Less-Common Met. 1983, 89, 183−195. D.; Banerjee, S. From Grignard’s Reagents to Well-Defined Mg
(226) Di Vece, M.; Grandjean, D.; Van Bael, M. J.; Romero, C. P.; Nanostructures: Distinctive Electrochemical and Solution Reduction
Wang, X.; Decoster, S.; Vantomme, A.; Lievens, P. Hydrogen-Induced Routes. Chem. Commun. 2012, 48, 5169−5171.

10831 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(247) Reetz, M. T.; Quaiser, S. A. A New Method for the in Activated Carbon Nanofibers: Dehydrogenation Kinetics, Rever-
Preparation of Nanostructured Metal Clusters. Angew. Chem., Int. Ed. sibility, and Mechanical Stability During Cycling. Int. J. Hydrogen
Engl. 1995, 34, 2240−2241. Energy 2017, 42, 1036−1047.
(248) Becker, J. A.; Schäfer, R.; Festag, R.; Ruland, W.; Wendorff, J. (267) Zhao-Karger, Z.; Witter, R.; Bardaji, E. G.; Wang, D.;
H.; Pebler, J.; Quaiser, S. A.; Helbig, W.; Reetz, M. T. Electrochemical Cossement, D.; Fichtner, M. Altered Reaction Pathways of Eutectic
Growth of Superparamagnetic Cobalt Clusters. J. Chem. Phys. 1995, LiBH4-Mg(BH4)2 by Nanoconfinement. J. Mater. Chem. A 2013, 1,
103, 2520−2527. 3379−3386.
(249) Reetz, M. T.; Winter, M.; Breinbauer, R.; Thurn-Albrecht, T.; (268) Javadian, P.; Sheppard, D. A.; Buckley, C. E.; Jensen, T. R.
Vogel, W. Size-Selective Electrochemical Preparation of Surfactant- Hydrogen Storage Properties of Nanoconfined LiBH4−Ca(BH4)2.
Stabilized Pd-, Ni- and Pt/Pd Colloids. Chem. - Eur. J. 2001, 7, 1084− Nano Energy 2015, 11, 96−103.
1094. (269) Hanada, N.; Chlopek, K.; Frommen, C.; Lohstroh, W.;
(250) Oumellal, Y.; Rougier, A.; Nazri, G. A.; Tarascon, J. M.; Fichtner, M. Thermal Decomposition of Mg(BH4)2 under He Flow
Aymard, L. Metal Hydrides for Lithium-Ion Batteries. Nat. Mater. and H2 Pressure. J. Mater. Chem. 2008, 18, 2611−2614.
2008, 7, 916−921. (270) Vitillo, J. G.; Bordiga, S.; Baricco, M. Spectroscopic and
(251) Huang, L.; Bonnet, J.-P.; Zlotea, C.; Bourgon, J.; Latroche, M.; Structural Characterization of Thermal Decomposition of γ-Mg-
Courty, M.; Aymard, L. Synthesis of Destabilized Nanostructured (BH4)2: Dynamic Vacuum Versus H2 Atmosphere. J. Phys. Chem. C
Lithium Hydride Via Hydrogenation of Lithium Electrochemically 2015, 119, 25340−25351.
Inserted into Graphite. Int. J. Hydrogen Energy 2015, 40, 13936− (271) Gutowska, A.; Li, L.; Shin, Y.; Wang Chongmin, M.; Li
13941. Xiaohong, S.; Linehan John, C.; Smith, R. S.; Kay Bruce, D.; Schmid,
(252) Zhang, S.; Gross, A. F.; Van Atta, S. L.; Lopez, M.; Liu, P.; B.; Shaw, W.; et al. Nanoscaffold Mediates Hydrogen Release and the
Ahn, C. C.; Vajo, J. J.; Jensen, C. M. The Synthesis and Hydrogen Reactivity of Ammonia Borane. Angew. Chem., Int. Ed. 2005, 44,
Storage Properties of a MgH2 Incorporated Carbon Aerogel Scaffold. 3578−3582.
Nanotechnology 2009, 20, 204027. (272) Zheng, S.; Fang, F.; Zhou, G.; Chen, G.; Ouyang, L.; Zhu, M.;
(253) Bramwell, P. L.; Ngene, P.; de Jongh, P. E. Carbon Supported Sun, D. Hydrogen Storage Properties of Space-Confined NaAlH4
Lithium Hydride Nanoparticles: Impact of Preparation Conditions on Nanoparticles in Ordered Mesoporous Silica. Chem. Mater. 2008, 20,
Particle Size and Hydrogen Sorption. Int. J. Hydrogen Energy 2017, 42, 3954−3958.
5188−5198. (273) Zhao-Karger, Z.; Hu, J.; Roth, A.; Wang, D.; Kubel, C.;
(254) Ziegler, K.; Nagel, K.; Pfohl, W. Metallorganische Lohstroh, W.; Fichtner, M. Altered Thermodynamic and Kinetic
Verbindungen. XXXVIII Pyrolyse Von Aluminiumtrialkylen. Liebigs. Properties of MgH2 Infiltrated in Microporous Scaffold. Chem.
Ann. 1960, 629, 210−221. Commun. 2010, 46, 8353−8355.
(255) Podall, H. E.; Petree, H. E.; Zietz, J. R. Relative Ease of (274) Jia, Y.; Yao, X. D. Carbon Scaffold Modified by Metal (Ni) or
Hydrogenolysis of Some Organometallic Compounds. J. Org. Chem.
Non-Metal (N) to Enhance Hydrogen Storage of MgH2 through
1959, 24, 1222−1226.
Nanoconfinement. Int. J. Hydrogen Energy 2017, 42, 22933−22941.
(256) Wang, L.; Rawal, A.; Quadir, M. Z.; Aguey-Zinsou, K.-F.
(275) Yan, Y.; Au, Y. S.; Rentsch, D.; Remhof, A.; de Jongh, P. E.;
Formation of Aluminium Hydride (AlH3) Via the Decomposition of
Zü ttel, A. Reversible Hydrogen Storage in Mg(BH4)2/Carbon
Organoaluminium and Hydrogen Storage Properties. Int. J. Hydrogen
Nanocomposites. J. Mater. Chem. A 2013, 1, 11177−11183.
Energy 2018, 43, 16749−16757.
(276) Sterl, F.; Linnenbank, H.; Steinle, T.; Mörz, F.; Strohfeldt, N.;
(257) Nielsen, T. K.; Javadian, P.; Polanski, M.; Besenbacher, F.;
Giessen, H. Nanoscale Hydrogenography on Single Magnesium
Bystrzycki, J.; Jensen, T. R. Nanoconfined NaAlH4: Determination of
Distinct Prolific Effects from Pore Size, Crystallite Size, and Surface Nanoparticles. Nano Lett. 2018, 18, 4293−4302.
(277) Danaie, M.; Mitlin, D. Tem Analysis and Sorption Properties
Interactions. J. Phys. Chem. C 2012, 116, 21046−21051.
(258) Li, W.-C.; Lu, A.-H.; Weidenthaler, C.; Schueth, F. Hard- of High-Energy Milled MgH2 Powders. J. Alloys Compd. 2009, 476,
Templating Pathway to Create Mesoporous Magnesium Oxide. Chem. 590−598.
Mater. 2004, 16, 5676−5681. (278) Surrey, A.; Schultz, L.; Rellinghaus, B. Multislice Simulations
(259) Nielsen, T. K.; Manickam, K.; Hirscher, M.; Besenbacher, F.; for In-Situ HRTEM Studies of Nanostructured Magnesium Hydride
Jensen, T. R. Confinement of MgH2 Nanoclusters within Nanoporous at Ambient Hydrogen Pressure. Ultramicroscopy 2017, 175, 111−115.
Aerogel Scaffold Materials. ACS Nano 2009, 3, 3521−3528. (279) Berti, N.; Cuevas, F.; Zhang, J. X.; Latroche, M. Enhanced
(260) Jun, S.; Joo, S. H.; Ryoo, R.; Kruk, M.; Jaroniec, M.; Liu, Z.; Reversibility of the Electrochemical Li Conversion Reaction with
Ohsuna, T.; Terasaki, O. Synthesis of New, Nanoporous Carbon with MgH2-TiH2 Nanocomposites. Int. J. Hydrogen Energy 2017, 42,
Hexagonally Ordered Mesostructure. J. Am. Chem. Soc. 2000, 122, 22615−22621.
10712−10713. (280) Matsuda, J.; Yoshida, K.; Sasaki, Y.; Uchiyama, N.; Akiba, E.
(261) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Frederickson, G. H.; In Situ Observation on Hydrogenation of Mg-Ni Films Using
Chmelka, B. F.; Stucky, G. D. Triblock Copolymer Syntheses of Environmental Transmission Electron Microscope with Aberration
Mesoporous Silica with Periodic 50 to 300 Angstrom Pores. Science Correction. Appl. Phys. Lett. 2014, 105, 083903.
1998, 279, 548−552. (281) Peterson, V. K.; Kearley, G. J. In Neutron Applications in
(262) Kitagawa, S.; Kitaura, R.; Noro, S.-i. Functional Porous Materials for Energy; Kearley, G. J., Peterson, V. K., Eds.; Springer:
Coordination Polymers. Angew. Chem., Int. Ed. 2004, 43, 2334−2375. Berlin, 2015.
(263) Paskevicius, M.; Ley, M. B.; Sheppard, D. A.; Jensen, T. R.; (282) Walker, G.; Bououdina, M.; Guo, Z. X.; Fruchart, D.
Buckley, C. E. Eutectic Melting in Metal Borohydrides. Phys. Chem. Handbook of Research on Nanoscience, Nanotechnology and Advanced
Chem. Phys. 2013, 15, 19774−19789. Materials; IGI Global, 2014.
(264) Shinde, S. S.; Kim, D.-H.; Yu, J.-Y.; Lee, J.-H. Self-Assembled (283) Bösenberg, U.; Vainio, U.; Pranzas, P. K.; Colbe, J. M. B. v.;
Air-Stable Magnesium Hydride Embedded in 3-D Activated Carbon Goerigk, G.; Welter, E.; Dornheim, M.; Schreyer, A.; Bormann, R. On
for Reversible Hydrogen Storage. Nanoscale 2017, 9, 7094−7103. the Chemical State and Distribution of Zr- and V-Based Additives in
(265) Lee, H.-S.; Lee, Y.-S.; Suh, J.-Y.; Kim, M.; Yu, J.-S.; Cho, Y. W. Reactive Hydride Composites. Nanotechnology 2009, 20, 204003.
Enhanced Desorption and Absorption Properties of Eutectic LiBH4− (284) NaraseGowda, S.; Gold, S. A.; Ilavsky, J.; Dobbins, T. A. In
Ca(BH4)2 Infiltrated into Mesoporous Carbon. J. Phys. Chem. C 2011, Materials Challenges in Alternative and Renewable Energy; Wicks, G.,
115, 20027−20035. Simon, J., Zidan, R., LaraCurzio, E., Adams, T., Zayas, J., Karkamkar,
(266) Plerdsranoy, P.; Javadian, P.; Jensen, N. D.; Nielsen, U. G.; A., Sindelar, R., GarciaDiaz, B., Eds.; Wiley: Weinheim, Germany,
Jensen, T. R.; Utke, R. Compaction of LiBH4-LiAlH4 Nanoconfined 2011; Vol. 224.

10832 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(285) Chumphongphan, S.; Filsø, U.; Paskevicius, M.; Sheppard, D. Application to Hydrogen Storage Materials. Phys. Rev. B: Condens.
A.; Jensen, T. R.; Buckley, C. E. Nanoconfinement Degradation in Matter Mater. Phys. 2008, 77, 104115.
NaAlH4/CMK-1. Int. J. Hydrogen Energy 2014, 39, 11103−11109. (304) Majzoub, E. H.; Hazrati, E.; de Wijs, G. A. First-Principles
(286) Sartori, S.; Knudsen, K. D.; Roth, A.; Fichtner, M.; Hauback, Study of Structural Prototypes for NaAlH4: Elevated Pressure
B. C. Small-Angle Scattering Investigations on Nanoconfined Sodium Polymorph in Symmetry Fmm2 Leads to a Single-Step Decom-
Alanate for Hydrogen Storage Applications. Nanosci. Nanotechnol. position Pathway. J. Phys. Chem. C 2013, 117, 8864−8870.
Lett. 2012, 4, 173−177. (305) Ozoliņs,̌ V.; Akbarzadeh, A. R.; Gunaydin, H.; Michel, K.;
(287) Sartori, S.; Knudsen, K. D.; Zhao-Karger, Z.; Bardaji, E. G.; Wolverton, C.; Majzoub, E. H. First-Principles Computational
Muller, J.; Fichtner, M.; Hauback, B. C. Nanoconfined Magnesium Discovery of Materials for Hydrogen Storage. J. Phys. Conf. Ser.
Borohydride for Hydrogen Storage Applications Investigated by 2009, 180, 012076.
SANS and SAXS. J. Phys. Chem. C 2010, 114, 18785−18789. (306) Zhang, Y. S.; Majzoub, E. H.; Ozoliņs,̌ V.; Wolverton, C.
(288) Iwase, K.; Mori, K.; Tomihira, S.; Oba, Y.; Fukunaga, T.; Theoretical Prediction of Metastable Intermediates in the Decom-
Sugiyama, M. Nano Structure of Metal Hydride by Neutron Small position of Mg(BH4)2. J. Phys. Chem. C 2012, 116, 10522−10528.
Angle Scattering. KURRI Progress Report 2013 2015, 155−155. (307) Ozoliņs,̌ V.; Majzoub, E. H.; Wolverton, C. First-Principles
(289) Lai, Q.; Aguey-Zinsou, K.-F. Destabilisation of Ca(BH4)2 and Prediction of a Ground State Crystal Structure of Magnesium
Mg(BH4)2 via Confinement in Nanoporous Cu2S Hollow Spheres. Borohydride. Phys. Rev. Lett. 2008, 100, 135501.
Sustainable Energy Fuels 2017, 1, 1308−1319. (308) Aidhy, D. S.; Wolverton, C. First-Principles Prediction of
(290) de Groot, F. High-Resolution X-Ray Emission and X-Ray Phase Stability and Crystal Structures in Li-Zn and Na-Zn Mixed-
Absorption Spectroscopy. Chem. Rev. 2001, 101, 1779−1808. Metal Borohydrides. Phys. Rev. B: Condens. Matter Mater. Phys. 2011,
(291) Oumellal, Y.; Provost, K.; Camelia Matei, G.; Yuso, A. M. d.; 83, 144111.
Zlotea, C. Composition and Size Dependence of Hydrogen (309) Aidhy, D. S.; Zhang, Y.; Wolverton, C. Prediction of a
Interaction with Carbon Supported Bulk-Immiscible Pd−Rh Nano- Ca(BH4)(NH2) Quaternary Hydrogen Storage Compound from
alloys. Nanotechnology 2016, 27, 465401. First-Principles Calculations. Phys. Rev. B: Condens. Matter Mater.
(292) Ray, K. G.; Klebanoff, L. E.; Lee, J. R. I.; Stavila, V.; Heo, T. Phys. 2011, 84, 134103.
W.; Shea, P.; Baker, A. A.; Kang, S.; Bagge-Hansen, M.; Liu, Y.-S.; (310) Huang, Z. Q.; Chen, W. C.; Chuang, F. C.; Majzoub, E. H.;
et al. Elucidating the Mechanism of MgB2 Initial Hydrogenation Via a Ozoliņs,̌ V. First-Principles Calculated Decomposition Pathways for
Combined Experimental-Theoretical Study. Phys. Chem. Chem. Phys. LiBH4 Nanoclusters. Sci. Rep. 2016, 6, 26056.
2017, 19, 22646−22658. (311) Wagner, L. K.; Majzoub, E. H.; Allendorf, M. D.; Grossman, J.
(293) Chueh, W. C.; El Gabaly, F.; Sugar, J. D.; Bartelt, N. C.; C. Tuning Metal Hydride Thermodynamics Via Size and
McDaniel, A. H.; Fenton, K. R.; Zavadil, K. R.; Tyliszczak, T.; Lai, W.; Composition: Li-H, Mg-H, Al-H, and Mg-Al-H Nanoclusters for
McCarty, K. F. Intercalation Pathway in Many-Particle LiFePO4
Hydrogen Storage. Phys. Chem. Chem. Phys. 2012, 14, 6611−6616.
Electrode Revealed by Nanoscale State-of-Charge Mapping. Nano
(312) Zhu, Q.; Oganov, A. R.; Glass, C. W.; Stokes, H. T.
Lett. 2013, 13, 866−872.
Constrained Evolutionary Algorithm for Structure Prediction of
(294) Rafi-ud-din; Zhang, L.; Ping, L.; Xuanhui, Q. Catalytic Effects
Molecular Crystals: Methodology and Applications. Acta Crystallogr.,
of Nano-Sized TiC Additions on the Hydrogen Storage Properties of
Sect. B: Struct. Sci. 2012, 68, 215−226.
LiAlH4. J. Alloys Compd. 2010, 508, 119−128.
(313) Fan, J.; Duan, D. F.; Jin, X. L.; Bao, K.; Liu, B. B.; Cui, T.
(295) Wahab, M. A.; Jia, Y.; Yang, D.; Zhao, H.; Yao, X. Enhanced
Structure Determination of Ultra Dense Magnesium Borohydride: A
Hydrogen Desorption from Mg(BH4)2 by Combining Nanoconfine-
ment and a Ni Catalyst. J. Mater. Chem. A 2013, 1, 3471−3478. First-Principles Study. J. Chem. Phys. 2013, 138, 214503.
(296) Zhang, Y.; Farrell, D.; Yang, J.; Sudik, A.; Wolverton, C. (314) Zhou, X. F.; Oganov, A. R.; Qian, G. R.; Zhu, Q. First-
Crystal Structures, Phase Stability, and Decomposition Reactions in Principles Determination of the Structure of Magnesium Borohydride.
the Quaternary Mg−B−N−H Hydrogen Storage System. J. Phys. Phys. Rev. Lett. 2012, 109, 245503.
Chem. C 2014, 118, 11193−11202. (315) Zhong, Y.; Zhou, H. Y.; Hu, C. H.; Wang, D. H.; Oganov, A.
(297) Newhouse, R. J.; Stavila, V.; Hwang, S.-J.; Klebanoff, L. E.; R. Theoretical Studies of High-Pressure Phases, Electronic Structure,
Zhang, J. Z. Reversibility and Improved Hydrogen Release of and Vibrational Properties of NaNH2. J. Phys. Chem. C 2012, 116,
Magnesium Borohydride. J. Phys. Chem. C 2010, 114, 5224−5232. 8387−8393.
(298) Hattrick-Simpers, J. R.; Maslar, J. E.; Niemann, M. U.; Chiu, (316) Hu, C. H.; Oganov, A. R.; Wang, Y. M.; Zhou, H. Y.; Lyakhov,
C.; Srinivasan, S. S.; Stefanakos, E. K.; Bendersky, L. A. Raman A.; Hafner, J. Crystal Structure Prediction of LiBeH3 Using Ab Initio
Spectroscopic Observation of Dehydrogenation in Ball-Milled Total-Energy Calculations and Evolutionary Simulations. J. Chem.
LiNH2−LiBH4−MgH2 Nanoparticles. Int. J. Hydrogen Energy 2010, Phys. 2008, 129, 234105.
35, 6323−6331. (317) Zhu, Q.; Oganov, A. R.; Lyakhov, A. O. Evolutionary
(299) Miedema, P. S.; Ngene, P.; van der Eerden, A. M. J.; Sokaras, Metadynamics: A Novel Method to Predict Crystal Structures.
D.; Weng, T.-C.; Nordlund, D.; Au, Y. S.; de Groot, F. M. F. In Situ CrystEngComm 2012, 14, 3596−3601.
X-Ray Raman Spectroscopy Study of the Hydrogen Sorption (318) Oganov, A. R.; Lyakhov, A. O.; Valle, M. How Evolutionary
Properties of Lithium Borohydride Nanocomposites. Phys. Chem. Crystal Structure Prediction Works-and Why. Acc. Chem. Res. 2011,
Chem. Phys. 2014, 16, 22651−22658. 44, 227−237.
(300) Dimitrievska, M.; White, J.; Zhou, W.; Stavila, V.; Klebanoff, (319) Lyakhov, A. O.; Oganov, A. R.; Valle, M. How to Predict Very
L.; Udovic, T. Structure-Dependent Vibrational Dynamics of Large and Complex Crystal Structures. Comput. Phys. Commun. 2010,
Mg(BH4)2 Polymorphs Probed with Neutron Vibrational Spectros- 181, 1623−1632.
copy and First-Principles Calculations. Phys. Chem. Chem. Phys. 2016, (320) Martonak, R.; Oganov, A. R.; Glass, C. W. Crystal Structure
18, 25546−25552. Prediction and Simulations of Structural Transformations: Metady-
(301) Kim, K. C.; Dai, B.; Johnson, J. K.; Sholl, D. S. Assessing namics and Evolutionary Algorithms. Phase Transitions 2007, 80,
Nanoparticle Size Effects on Metal Hydride Thermodynamics Using 277−298.
the Wulff Construction. Nanotechnology 2009, 20, 204001. (321) Oganov, A. R.; Glass, C. W. Crystal Structure Prediction
(302) Barmparis, G. D.; Lodziana, Z.; Lopez, N.; Remediakis, I. N. Using Ab Initio Evolutionary Techniques: Principles and Applica-
Nanoparticle Shapes by Using Wulff Constructions and First- tions. J. Chem. Phys. 2006, 124, 244704.
Principles Calculations. Beilstein J. Nanotechnol. 2015, 6, 361−368. (322) Glass, C. W.; Oganov, A. R.; Hansen, N. USPEX -
(303) Majzoub, E. H.; Ozoliņs,̌ V. Prototype Electrostatic Ground Evolutionary Crystal Structure Prediction. Comput. Phys. Commun.
State Approach to Predicting Crystal Structures of Ionic Compounds: 2006, 175, 713−720.

10833 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(323) Filinchuk, Y.; Richter, B.; Jensen, T. R.; Dmitriev, V.; (343) Baldi, A.; Gonzalez-Silveira, M.; Palmisano, V.; Dam, B.;
Chernyshov, D.; Hagemann, H. Porous and Dense Magnesium Griessen, R. Destabilization of the Mg-H System through Elastic
Borohydride Frameworks: Synthesis, Stability, and Reversible Constraints. Phys. Rev. Lett. 2009, 102, 226102.
Absorption of Guest Species. Angew. Chem., Int. Ed. 2011, 50, (344) Pang, Y.; Li, Q. A Review on Kinetic Models and
11162−11166. Corresponding Analysis Methods for Hydrogen Storage Materials.
(324) Chen, L. Q. Phase-Field Models for Microstructure Evolution. Int. J. Hydrogen Energy 2016, 41, 18072−18087.
Annu. Rev. Mater. Res. 2002, 32, 113−140. (345) Zavorotynska, O.; Deledda, S.; Hauback, B. C. Kinetics
(325) Cahn, J. W.; Hilliard, J. E. Free Energy of a Nonuniform Studies of the Reversible Partial Decomposition Reaction in
System 1. Interfacial Free Energy. J. Chem. Phys. 1958, 28, 258−267. Mg(BH4)2. Int. J. Hydrogen Energy 2016, 41, 9885−9892.
(326) Heo, T. W.; Chen, L. Q.; Wood, B. C. Phase-Field Modeling (346) Arrhenius, S. Ü ber Die Dissociationswärme Und Den Einfluss
of Diffusional Phase Behaviors of Solid Surfaces: A Case Study of Der Temperatur Auf Den Dissociationsgrad Der Elektrolyte. Z. Phys.
Phase-Separating LixFePO4 Electrode Particles. Comput. Mater. Sci. Chem. 1889, 4U, 96.
2015, 108, 323−332. (347) Arrhenius, S. Ü ber Die Reaktionsgeschwindigkeit Bei Der
(327) Heo, T. W.; Bhattacharyya, S.; Chen, L. Q. A Phase Field Inversion Von Rohrzucker Durch Säuren. Z. Phys. Chem. 1889, 4U,
Study of Strain Energy Effects on Solute-Grain Boundary Interactions. 226.
Acta Mater. 2011, 59, 7800−7815. (348) Kissinger, H. E. Reaction Kinetics in Differential Thermal
(328) Heo, T. W.; Bhattacharyya, S.; Chen, L. Q. A Phase-Field Analysis. Anal. Chem. 1957, 29, 1702−1706.
Model for Elastically Anisotropic Polycrystalline Binary Solid (349) Fichtner, M.; Zhao-Karger, Z.; Hu, J.; Roth, A.; Weidler, P.
Solutions. Philos. Mag. 2013, 93, 1468−1489. The Kinetic Properties of Mg(BH4)2 Infiltrated in Activated Carbon.
(329) Heo, T. W.; Chen, L. Q. Phase-Field Modeling of Displacive Nanotechnology 2009, 20, 204029.
Phase Transformations in Elastically Anisotropic and Inhomogeneous (350) Joubert, J. M.; Thiébaut, S. Thermodynamic Assessment of
Polycrystals. Acta Mater. 2014, 76, 68−81. the Pd−H−D−T System. J. Nucl. Mater. 2009, 395, 79−88.
(330) Tang, M.; Carter, W. C.; Chiang, Y. M. Electrochemically (351) Pinatel, E. R.; Albanese, E.; Civalleri, B.; Baricco, M.
Driven Phase Transitions in Insertion Electrodes or Lithium-Ion Thermodynamic Modelling of Mg(BH4)2. J. Alloys Compd. 2015,
Batteries: Examples in Lithium Metal Phosphate Olivines. Annu. Rev. 645, S64−S68.
Mater. Res. 2010, 40, 501−529. (352) Kim, K. C.; Allendorf, M. D.; Stavila, V.; Sholl, D. S.
(331) Heo, T. W.; Tang, M.; Chen, L. Q.; Wood, B. C. Defects, Predicting Impurity Gases and Phases During Hydrogen Evolution
Entropy, and the Stabilization of Alternative Phase Boundary from Complex Metal Hydrides Using Free Energy Minimization
Orientations in Battery Electrode Particles. Adv. Energy Mater. Enabled by First-Principles Calculations. Phys. Chem. Chem. Phys.
2016, 6, 10. 2010, 12, 9918−9926.
(332) Hong, L.; Li, L.; Chen-Wiegart, Y.-K.; Wang, J.; Xiang, K.; (353) Akbarzadeh, A.; Ozoliņs,̌ V.; Wolverton, C. First-Principles
Gan, L.; Li, W.; Meng, F.; Wang, F.; Wang, J.; et al. Two-Dimensional Determination of Multicomponent Hydride Phase Diagrams:
Lithium Diffusion Behavior and Probable Hybrid Phase Trans- Application to the Li-Mg-N-H System. Adv. Mater. 2007, 19,
formation Kinetics in Olivine Lithium Iron Phosphate. Nat. Commun. 3233−3239.
2017, 8, 1194. (354) Ozolins, V.; Akbarzadeh, A. R.; Gunaydin, H.; Michel, K.;
(333) Voskuilen, T. G.; Pourpoint, T. L. Phase Field Modeling of Wolverton, C.; Majzoub, E. H. First-Principles Computational
Hydrogen Transport and Reaction in Metal Hydrides. Int. J. Hydrogen Discovery of Materials for Hydrogen Storage. Journal of Physics:
Energy 2013, 38, 7363−7375. Conference Series 2009, 180, 012076.
(334) Ulvestad, A.; Welland, M. J.; Collins, S. S. E.; Harder, R.; (355) Poonyayant, N.; Stavila, V.; Majzoub, E. H.; Klebanoff, L. E.;
Maxey, E.; Wingert, J.; Singer, A.; Hy, S.; Mulvaney, P.; Zapol, P.; Behrens, R.; Angboonpong, N.; Ulutagay-Kartin, M.; Pakawatpanurut,
et al. Avalanching Strain Dynamics During the Hydriding Phase P.; Hecht, E. S.; Breit, J. S. An Investigation into the Hydrogen
Transformation in Individual Palladium Nanoparticles. Nat. Commun. Storage Characteristics of Ca(BH4)2/LiNH2 and Ca(BH4)2/NaNH2:
2015, 6, 10092. Evidence of Intramolecular Destabilization. J. Phys. Chem. C 2014,
(335) Ulvestad, A.; Welland, M. J.; Cha, W.; Liu, Y.; Kim, J. W.; 118, 14759−14769.
Harder, R.; Maxey, E.; Clark, J. N.; Highland, M. J.; You, H.; et al. (356) Voss, J.; Hummelshoj, J.; Lodziana, Z.; Vegge, T. Structural
Three-Dimensional Imaging of Dislocation Dynamics During the Stability and Decomposition of Mg(BH4)2 Isomorphs-an Ab Initio
Hydriding Phase Transformation. Nat. Mater. 2017, 16, 565−571. Free Energy Study. J. Phys.: Condens. Matter 2009, 21, 012203.
(336) Züttel, A. Materials for Hydrogen Storage. Mater. Today 2003, (357) Zarkevich, N. A.; Johnson, D. D. Predicting Enthalpies of
6, 24−33. Molecular Substances: Application to LiBH4. Phys. Rev. Lett. 2008,
(337) Gross, K. J.; Carrington, K. R. Recommended Best Practices for 100, 040602.
the Characterization of Storage Properties of Hydrogen Storage Materials; (358) Zarkevich, N. A.; Majzoub, E. H.; Johnson, D. D. Anisotropic
National Renewable Energy Laboratory, 2008. Thermal Expansion in Molecular Solids: Theory and Experiment on
(338) Fukai, Y. The Metal-Hydrogen System: Basic Bulk Properties; LiBH4. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 89, 134308.
Springer-Verlag: Berlin, 2005. (359) Mueller, T.; Ceder, G. Effect of Particle Size on Hydrogen
(339) Remhof, A.; Borgschulte, A. Thin-Film Metal Hydrides. Release from Sodium Alanate Nanoparticles. ACS Nano 2010, 4,
ChemPhysChem 2008, 9, 2440−2455. 5647−5656.
(340) Pranevicius, L.; Milcius, D.; Templier, C. The Effects of (360) Kang, S.; Ogitsu, T.; Bonev, S. A.; Heo, T. W.; Allendorf, M.
Dynamic Structural Transformations on Hydrogenation Properties of D.; Wood, B. C. Understanding Charge Transfer at Mg/MgH2
Mg and Mgni Thin Films. Int. J. Hydrogen Energy 2009, 34, 5131− Interfaces for Hydrogen Storage. ECS Trans. 2017, 77, 81−90.
5137. (361) Tang, J.; Yang, X.; Chen, L.; Zhao, Y. Modeling and Stabilities
(341) Bannenberg, L. J.; Schreuders, H.; van Eijck, L.; Heringa, J. R.; of Mg/MgH2 Interfaces: A First-Principles Investigation. AIP Adv.
Steinke, N. J.; Dalgliesh, R.; Dam, B.; Mulder, F. M.; van Well, A. A. 2014, 4, 077101.
Impact of Nanostructuring on the Phase Behavior of Insertion (362) Tao, S.; Kalisvaart, W.; Danaie, M.; Mitlin, D.; Notten, P.; van
Materials: The Hydrogenation Kinetics of a Magnesium Thin Film. J. Santen, R.; Jansen, A. First Principle Study of Hydrogen Diffusion in
Phys. Chem. C 2016, 120, 10185−10191. Equilibrium Rutile, Rutile with Deformation Twins and Fluorite
(342) Griessen, R. Hydrogenography: Shedding Light on Hydrogen Polymorph of Mg Hydride. Int. J. Hydrogen Energy 2011, 36, 11802−
in Metals. Diffus. Defect Data, Pt. B 2011, 170, 329−329. 11809.

10834 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(363) Rao, C. N. R.; Muller, A.; Cheetam, A. K. The Chemistry of Formation Upon Dehydrogenation of MgH2 Thin Films. J. Appl. Phys.
Nanomaterials: Synthesis, Properties and Applications; Wiley-VCH: 2011, 109, 093501.
Darmstadt, 2004. (383) Kalisvaart, W.; Kubis, A.; Danaie, M.; Amirkhiz, B. S.; Mitlin,
(364) Yamauchi, M.; Kobayashi, H.; Kitagawa, H. Hydrogen Storage D. Microstructural Evolution During Hydrogen Sorption Cycling of
Mediated by Pd and Pt Nanoparticles. ChemPhysChem 2009, 10, Mg−Feti Nanolayered Composites. Acta Mater. 2011, 59, 2083−
2566−2576. 2095.
(365) Vons, V. A.; Anastasopol, A.; Legerstee, W. J.; Mulder, F. M.; (384) Baldi, A.; Palmisano, V.; Gonzalez-Silveira, M.; Pivak, Y.;
Eijt, S. W. H.; Schmidt-Ott, A. Low-Temperature Hydrogen Slaman, M.; Schreuders, H.; Dam, B.; Griessen, R. Quasifree Mg−H
Desorption and the Structural Properties of Spark Discharge Thin Films. Appl. Phys. Lett. 2009, 95, 071903.
Generated Mg Nanoparticles. Acta Mater. 2011, 59, 3070−3080. (385) Mooij, L.; Dam, B. Nucleation and Growth Mechanisms of
(366) Liu, W.; Aguey-Zinsou, K.-F. Size Effects and Hydrogen Nano Magnesium Hydride from the Hydrogen Sorption Kinetics.
Storage Properties of Mg Nanoparticles Synthesised by an Electroless Phys. Chem. Chem. Phys. 2013, 15, 11501−11510.
Reduction Method. J. Mater. Chem. A 2014, 2, 9718−9726. (386) Mooij, L.; Dam, B. Hysteresis and the Role of Nucleation and
(367) Kalidindi, S. B.; Jagirdar, B. R. Highly Monodisperse Colloidal Growth in the Hydrogenation of Mg Nanolayers. Phys. Chem. Chem.
Magnesium Nanoparticles by Room Temperature Digestive Ripening. Phys. 2013, 15, 2782−2792.
Inorg. Chem. 2009, 48, 4524−4529. (387) Kelly, S. T.; Van Atta, S. L.; Vajo, J. J.; Olson, G. L.; Clemens,
(368) Liu, W.; Aguey-Zinsou, K.-F. Synthesis of Highly Dispersed B. M. Kinetic Limitations of the Mg2Si System for Reversible
Nanosized LaNi5 on Carbon: Revisiting Particle Size Effects on Hydrogen Storage. Nanotechnology 2009, 20, 204017.
Hydrogen Storage Properties. Int. J. Hydrogen Energy 2016, 41, (388) Adelhelm, P.; Gao, J.; Verkuijlen, M. H.; Rongeat, C.; Herrich,
14429−14436. M.; van Bentum, P. J. M.; Gutfleisch, O.; Kentgens, A. P.; de Jong, K.
(369) Varin, R. A.; Parviz, R. The Effects of the Nanometric P.; de Jongh, P. E. Comprehensive Study of Melt Infiltration for the
Interstitial Compounds TiC, ZrC and TiN on the Mechanical and Synthesis of NaAlH4/C Nanocomposites. Chem. Mater. 2010, 22,
Thermal Dehydrogenation and Rehydrogenation of the Nano- 2233−2238.
composite Lithium Alanate (LiAlH4) Hydride. Int. J. Hydrogen Energy (389) Ngene, P.; van den Berg, R.; Verkuijlen, M. H. W.; de Jong, K.
2014, 39, 2575−2586. P.; de Jongh, P. E. Reversibility of the Hydrogen Desorption from
(370) Lin, S. S. Y.; Yang, J.; Kung, H. H.; Kung, M. C. Hydrogen NaBH4 by Confinement in Nanoporous Carbon. Energy Environ. Sci.
Storage Properties of Complex Metal Hydride-Carbon Materials. Top. 2011, 4, 4108−4115.
Catal. 2013, 56, 1937−1943. (390) Ngene, P.; van Zwienen, M. R.; de Jongh, P. E. Reversibility of
(371) Christian, M. L.; Aguey-Zinsou, K.-F. Core−Shell Strategy the Hydrogen Desorption from LiBH4: A Synergetic Effect of
Leading to High Reversible Hydrogen Storage Capacity for NaBH4. Nanoconfinement and Ni Addition. Chem. Commun. 2009, 46, 8201−
ACS Nano 2012, 6, 7739−7751. 8203.
(372) Huiberts, J. N.; Griessen, R.; Rector, J. H.; Wijngaarden, R. J.; (391) Comănescu, C.; Capurso, G.; Maddalena, A. Nanoconfine-
Dekker, J. P.; de Groot, D. G.; Koeman, N. J. Yttrium and Lanthanum ment in Activated Mesoporous Carbon of Calcium Borohydride for
Hydride Films with Switchable Optical Properties. Nature 1996, 380, Improved Reversible Hydrogen Storage. Nanotechnology 2012, 23,
231. 385401.
(373) Zhang, W. K.; Gan, Y. P.; Yang, X. G.; Huang, H.; Yu, L. Y. (392) Lohstroh, W.; Roth, A.; Hahn, H.; Fichtner, M. Thermody-
Crystal Structure and Switchable Optical Properties of Yttrium namic Effects in Nanoscale NaAlH4. ChemPhysChem 2010, 11, 789−
Hydride Films Covered by Palladium Layer. Trans. Nonferrous Met. 792.
Soc. China 2003, 13, 1401−1404. (393) Wu, C.; Cheng, H.-M. Effects of Carbon on Hydrogen Storage
(374) Zhang, W. K.; Huang, H.; Gan, Y. P.; Yang, X. G.; Yu, L. Y. Performances of Hydrides. J. Mater. Chem. 2010, 20, 5390−5400.
Surface Morphology and Switchable Optical Properties of Pd/Y Films (394) Capurso, G.; Agresti, F.; Crociani, L.; Rossetto, G.; Schiavo,
During Gas Hydrogen Absorption/Desorption Process. Acta Metall. B.; Maddalena, A.; Lo Russo, S.; Principi, G. Nanoconfined Mixed Li
Sin. 2003, 39, 974−978. and Mg Borohydrides as Materials for Solid State Hydrogen Storage.
(375) Sakai, T.; Ishikawa, H.; Miyamura, H.; Kuriyama, N.; Yamada, Int. J. Hydrogen Energy 2012, 37, 10768−10773.
S.; Iwasaki, T. Thin Film Preparation of Hydrogen Storage Alloys and (395) Gao, J.; Adelhelm, P.; Verkuijlen, M. H. W.; Rongeat, C.;
Their Characteristics as Metal Hydride Electrodes. J. Electrochem. Soc. Herrich, M.; van Bentum, P. J. M.; Gutfleisch, O.; Kentgens, A. P. M.;
1991, 138, 908−915. de Jong, K. P.; de Jongh, P. E. Confinement of NaAH4 in Nanoporous
(376) Clark, N.; Vargas, W. E.; Azofeifa, D. E. Dielectric Function of Carbon: Impact on H2 Release, Reversibility, and Thermodynamics. J.
Pd Hydride Thin Films in Terms of Hydrogen Concentration and Phys. Chem. C 2010, 114, 4675−4682.
Film’s Thickness: A Parametric Formulation. J. Alloys Compd. 2015, (396) Stephens, R. D.; Gross, A. F.; Van Atta, S. L.; Vajo, J. J.;
645, S320−S324. Pinkerton, F. E. The Kinetic Enhancement of Hydrogen Cycling in
(377) Khanuja, M.; Varandani, D.; Mehta, B. R. Pulse Like NaAlH4 by Melt Infusion into Nanoporous Carbon Aerogel.
Hydrogen Sensing Response in Pd Nanoparticle Layers. Appl. Phys. Nanotechnology 2009, 20, 204018.
Lett. 2007, 91, 253121. (397) Nielsen, T. K.; Polanski, M.; Zasada, D.; Javadian, P.;
(378) Hamm, M.; Burlaka, V.; Wagner, S.; Pundt, A. Achieving Besenbacher, F.; Bystrzycki, J.; Skibsted, J.; Jensen, T. R. Improved
Reversibility of Ultra-High Mechanical Stress by Hydrogen Loading Hydrogen Storage Kinetics of Nanoconfined NaAlH4 Catalyzed with
of Thin Films. Appl. Phys. Lett. 2015, 106, 243108. TiCl3 Nanoparticles. ACS Nano 2011, 5, 4056−4064.
(379) Milcius, D.; Lelis, M. In International Conference on Radiation (398) Gross, A. F.; Vajo, J. J.; Van Atta, S. L.; Olson, G. L. Enhanced
Interaction with Materials and Its Use in Technologies 2008; Grigonis, Hydrogen Storage Kinetics of LiBH4 in Nanoporous Carbon
A., Ed., 2008. Scaffolds. J. Phys. Chem. C 2008, 112, 5651−5657.
(380) Jurczyk, M.; Smardz, L.; Okonska, I.; Jankowska, E.; Nowak, (399) Zhang, Y.; Zhang, W.-S.; Wang, A.-Q.; Sun, L.-X.; Fan, M.-Q.;
M.; Smardz, K. Nanoscale Mg-Based Materials for Hydrogen Storage. Chu, H.-L.; Sun, J.-C.; Zhang, T. LiBH4 Nanoparticles Supported by
Int. J. Hydrogen Energy 2008, 33, 374−380. Disordered Mesoporous Carbon: Hydrogen Storage Performances
(381) Nowak, M.; Smardz, L.; Jurczyk, M. Hydrogen Storage in and Destabilization Mechanisms. Int. J. Hydrogen Energy 2007, 32,
Nanostructured Mg-Based Hydrides and Their Composites. Current 3976−3980.
Topics in Electrochemistry 2010, 15, 25−38. (400) Cahen, S.; Eymery, J. B.; Janot, R.; Tarascon, J. M.
(382) Dura, J. A.; Kelly, S. T.; Kienzle, P. A.; Her, J. H.; Udovic, T. Improvement of the LiBH4 Hydrogen Desorption by Inclusion into
J.; Majkrzak, C. F.; Chung, C. J.; Clemens, B. M. Porous Mg Mesoporous Carbons. J. Power Sources 2009, 189, 902−908.

10835 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(401) Shao, J.; Xiao, X. Z.; Fan, X. L.; Zhang, L. T.; Li, S. Q.; Ge, H. (420) Xia, G.; Tan, Y.; Chen, X.; Sun, D.; Guo, Z.; Liu, H.; Ouyang,
W.; Wang, Q. D.; Chen, L. X. Low-Temperature Reversible Hydrogen L.; Zhu, M.; Yu, X. Monodisperse Magnesium Hydride Nanoparticles
Storage Properties of LiBH4: A Synergetic Effect of Nanoconfinement Uniformly Self-Assembled on Graphene. Adv. Mater. 2015, 27, 5981−
and Nanocatalysis. J. Phys. Chem. C 2014, 118, 11252−11260. 5988.
(402) Liu, X. F.; Peaslee, D.; Majzoub, E. H. Tailoring the Hydrogen (421) Aguey-Zinsou, K. F.; Boyer, C. Synthesis and Stabilisation of
Storage Properties of Li4BN3H10 by Confinement into Highly MgH2 Nanoparticles by Self-Assembly. ChemPlusChem 2012, 77,
Ordered Nanoporous Carbon. J. Mater. Chem. A 2013, 1, 3926−3931. 423−426.
(403) Carr, C. L.; Majzoub, E. H. Surface-Functionalized Nano- (422) Anastasopol, A.; Pfeiffer, T. V.; Middelkoop, J.; Lafont, U.;
porous Carbons for Kinetically Stabilized Complex Hydrides through Canales-Perez, R. J.; Schmidt-Ott, A.; Mulder, F. M.; Eijt, S. W. H.
Lewis Acid-Lewis Base Chemistry. J. Phys. Chem. C 2016, 120, Reduced Enthalpy of Metal Hydride Formation for Mg−Ti
11426−11432. Nanocomposites Produced by Spark Discharge Generation. J. Am.
(404) Ryoo, R.; Joo, S. H.; Kruk, M.; Jaroniec, M. Ordered Chem. Soc. 2013, 135, 7891−7900.
Mesoporous Carbons. Advanced Materials; 2001. (423) Liu, Y.; Zou, J.; Zeng, X.; Ding, W. A Co-Precipitated Mg-Ti
(405) Hu, Z.; Srinivasan, M. P.; Ni, Y. Preparation of Mesoporous Nano-Composite with High Capacity and Rapid Hydrogen
High-Surface-Area Activated Carbon. Adv. Mater. 2000, 12, 62−65. Absorption Kinetics at Room Temperature. RSC Adv. 2014, 4,
(406) Tan, C.-Y.; Tsai, W.-T. Catalytic and Inhibitive Effects of Pd 42764−42771.
and Pt Decorated MWCNTs on the Dehydrogenation Behavior of (424) Liu, T.; Zhang, T.; Qin, C.; Zhu, M.; Li, X. Improved
LiAlH4. Int. J. Hydrogen Energy 2015, 40, 10185−10193. Hydrogen Storage Properties of Mg−V Nanoparticles Prepared by
(407) Baldé, C. P.; Hereijgers, B. P. C.; Bitter, J. H.; de Jong, K. P. Hydrogen Plasma−Metal Reaction. J. Power Sources 2011, 196,
Facilitated Hydrogen Storage in NaAlH4 Supported on Carbon 9599−9604.
Nanofibers. Angew. Chem., Int. Ed. 2006, 45, 3501−3503. (425) Shao, H.; Xu, H.; Wang, Y.; Li, X. Synthesis and Hydrogen
(408) Li, Y.; Ding, X.; Zhang, Q. Self-Printing on Graphitic Storage Behavior of Mg−Co−H System at Nanometer Scale. J. Solid
Nanosheets with Metal Borohydride Nanodots for Hydrogen Storage. State Chem. 2004, 177, 3626−3632.
Sci. Rep. 2016, 6, 31144. (426) Molinari, A.; D’Amico, F.; Calizzi, M.; Zheng, Y.; Boelsma, C.;
(409) Gasnier, A.; Gennari, F. Graphene Entanglement in a Mooij, L.; Lei, Y.; Hahn, H.; Dam, B.; Pasquini, L. Interface and
Mesoporous Resorcinol−Formaldehyde Matrix Applied to the Strain Effects on the H-Sorption Thermodynamics of Size-Selected
Nanoconfinement of LiBH4 for Hydrogen Storage. RSC Adv. 2017, Mg Nanodots. Int. J. Hydrogen Energy 2016, 41, 9841−9851.
7, 27905−27912. (427) Zou, J.; Long, S.; Chen, X.; Zeng, X.; Ding, W. Preparation
(410) Zhou, C.; Szpunar, J. A. Hydrogen Storage Performance in
and Hydrogen Sorption Properties of a Ni Decorated Mg Based Mg@
Pd/Graphene Nanocomposites. ACS Appl. Mater. Interfaces 2016, 8,
Ni Nano-Composite. Int. J. Hydrogen Energy 2015, 40, 1820−1828.
25933−25940.
(428) Shahi, R. R.; Tiwari, A. P.; Shaz, M. A.; Srivastava, O. N.
(411) Schröder, F.; Esken, D.; Cokoja, M.; van den Berg, M. W.;
Studies on De/Rehydrogenation Characteristics of Nanocrystalline
Lebedev, O. I.; Van Tendeloo, G.; Walaszek, B.; Buntkowsky, G.;
MgH2 Co-Catalyzed with Ti, Fe and Ni. Int. J. Hydrogen Energy 2013,
Limbach, H.-H.; Chaudret, B. Ruthenium Nanoparticles inside
38, 2778−2784.
Porous [Zn4O(bdc)3] by Hydrogenolysis of Adsorbed [Ru(cod)-
(429) Lu, J.; Choi, Y. J.; Fang, Z. Z.; Sohn, H. Y.; Rönnebro, E.
(cot)]: A Solid-State Reference System for Surfactant-Stabilized
Hydrogenation of Nanocrystalline Mg at Room Temperature in the
Ruthenium Colloids. J. Am. Chem. Soc. 2008, 130, 6119−6130.
(412) Zlotea, C.; Campesi, R.; Cuevas, F.; Leroy, E.; Dibandjo, P.; Presence of TiH2. J. Am. Chem. Soc. 2010, 132, 6616−6617.
Volkringer, C.; Loiseau, T.; Férey, G. r.; Latroche, M. Pd (430) Lu, J.; Choi, Y. J.; Fang, Z. Z.; Sohn, H. Y.; Rönnebro, E.
Nanoparticles Embedded into a Metal-Organic Framework: Synthesis, Hydrogen Storage Properties of Nanosized MgH2−0.1TiH2 Prepared
Structural Characteristics, and Hydrogen Sorption Properties. J. Am. by Ultrahigh-Energy−High-Pressure Milling. J. Am. Chem. Soc. 2009,
Chem. Soc. 2010, 132, 2991−2997. 131, 15843−15852.
(413) Bhakta, R. K.; Herberg, J. L.; Jacobs, B.; Highley, A.; Behrens, (431) Barkhordarian, G.; Klassen, T.; Bormann, R. Fast Hydrogen
R.; Ockwig, N. W.; Greathouse, J. A.; Allendorf, M. D. Metal− Sorption Kinetics of Nanocrystalline Mg Using Nb2O5 as Catalyst.
Organic Frameworks as Templates for Nanoscale NaAlH4. J. Am. Scr. Mater. 2003, 49, 213−217.
Chem. Soc. 2009, 131, 13198−13199. (432) Niaz, N. A.; Ahmad, I.; Khalid, N. R.; Ahmed, E.; Abbas, S.
(414) Ngene, P.; Adelhelm, P.; Beale, A. M.; de Jong, K. P.; de M.; Jabeen, N. Preparation of Mg2FeH6 Nanoparticles for Hydrogen
Jongh, P. E. LiBH4/SBA-15 Nanocomposites Prepared by Melt Storage Properties. J. Nanomater. 2013, 2013, 7.
Infiltration under Hydrogen Pressure: Synthesis and Hydrogen (433) Zhang, X.; Yang, R.; Qu, J.; Zhao, W.; Xie, L.; Tian, W.; Li, X.
Sorption Properties. J. Phys. Chem. C 2010, 114, 6163−6168. The Synthesis and Hydrogen Storage Properties of Pure Nano-
(415) Mulas, G.; Campesi, R.; Garroni, S.; Napolitano, E.; Milanese, structured Mg2FeH6. Nanotechnology 2010, 21, 095706.
C.; Dolci, F.; Pellicer, E.; Baró, M. D.; Marini, A. Hydrogen Storage in (434) Liu, W.; Aguey-Zinsou, K. F. Hydrogen Storage Properties of
2NaBH4 + MgH2 Mixtures: Destabilization by Additives and In-Situ Stabilised Magnesium Nanoparticles Generated by Electroless
Nanoconfinement. J. Alloys Compd. 2012, 536, S236−S240. Reduction with Alkali Metals. Int. J. Hydrogen Energy 2015, 40,
(416) Christian, M.; Aguey-Zinsou, K.-F. Synthesis of Core−Shell 16948−16960.
NaiBH4@ M (M= Co, Cu, Fe, Ni, Sn) Nanoparticles Leading to (435) Gupta, S.; Hlova, I. Z.; Kobayashi, T.; Denys, R. V.; Chen, F.;
Various Morphologies and Hydrogen Storage Properties. Chem. Zavaliy, I. Y.; Pruski, M.; Pecharsky, V. K. Facile Synthesis and
Commun. 2013, 49, 6794−6796. Regeneration of Mg(BH4)2 by High Energy Reactive Ball Milling of
(417) Wang, L.; Aguey-Zinsou, K.-F. Synthesis of LiAlH4 Nano- MgB2. Chem. Commun. 2013, 49, 828−830.
particles Leading to a Single Hydrogen Release Step Upon Ti (436) Li, Y.; Zhang, Q.; Fang, F.; Song, Y.; Sun, D.; Ouyang, L.;
Coating. Inorganics 2017, 5, 38. Zhu, M. Facile Self-Assembly of Light Metal Borohydrides with
(418) Syrenova, S.; Wadell, C.; Nugroho, F. A. A.; Gschneidtner, T. Controllable Nanostructures. RSC Adv. 2014, 4, 983−986.
A.; Diaz Fernandez, Y. A.; Nalin, G.; Świtlik, D.; Westerlund, F.; (437) Diduszko, R.; Kozlowski, M.; Czerwosz, E.; Kaminska, A.;
Antosiewicz, T. J.; Zhdanov, V. P.; et al. Hydride Formation Nietubyc, R. In Applied Crystallography XXII; Stroz, D., Dercz, G.,
Thermodynamics and Hysteresis in Individual Pd Nanocrystals Eds.; Trans Tech Publications, 2013; Vol. 203−204.
With different Size and Shape. Nat. Mater. 2015, 14, 1236. (438) Kaminska, A.; Diduszko, R.; Krawczyk, S.; Czerwosz, E.;
(419) Wang, L.; Quadir, M. Z.; Aguey-Zinsou, K.-F. Ni Coated LiH Sobczak, K. Influence of Hydrogen on the Properties of Nano-
Nanoparticles for Reversible Hydrogen Storage. Int. J. Hydrogen structured C-Pd Films for Sensing Applications. Pol. J. Chem. Technol.
Energy 2016, 41, 6376−6386. 2014, 16, 77−81.

10836 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(439) Barcelo, S.; Rogers, M.; Grigoropoulos, C. P.; Mao, S. S. T.; et al. Nanoconfined 2LiBH4−MgH2−TiCl3 in Carbon Aerogel
Hydrogen Storage Property of Sandwiched Magnesium Hydride Scaffold for Reversible Hydrogen Storage. Int. J. Hydrogen Energy
Nanoparticle Thin Film. Int. J. Hydrogen Energy 2010, 35, 7232−7235. 2013, 38, 3275−3282.
(440) Kumar, S.; Singh, V.; Cassidy, C.; Pursell, C.; Nivargi, C.; (457) Fang, Z. Z.; Wang, P.; Rufford, T. E.; Kang, X. D.; Lu, G. Q.;
Clemens, B.; Sowwan, M. Hydrogenation of Mg Nanofilms Catalyzed Cheng, H. M. Kinetic- and Thermodynamic-Based Improvements of
by Size-Selected Pd Nanoparticles: Observation of Localized MgH2 Lithium Borohydride Incorporated into Activated Carbon. Acta
Nanodomains. J. Catal. 2016, 337, 14−25. Mater. 2008, 56, 6257−6263.
(441) Kelekar, R.; Giffard, H.; Kelly, S. T.; Clemens, B. M. (458) Ngene, P.; Verkuijlen, M. H. W.; Barre, C.; Kentgens, A. P.
Formation and Dissociation of MgH2 in Epitaxial Mg Thin Films. J. M.; de Jongh, P. E. Reversible Li-Insertion in Nanoscaffolds: A
Appl. Phys. 2007, 101, 114311. Promising Strategy to Alter the Hydrogen Sorption Properties of Li-
(442) Siviero, G.; Bello, V.; Mattei, G.; Mazzoldi, P.; Battaglin, G.; Based Complex Hydrides. Nano Energy 2016, 22, 169−178.
Bazzanella, N.; Checchetto, R.; Miotello, A. Structural Evolution of (459) Javadian, P.; Sheppard, D.; Buckley, C.; Jensen, T. Hydrogen
Pd-Capped Mg Thin Films under H2 Absorption and Desorption Desorption Properties of Bulk and Nanoconfined LiBH4-NaAlH4.
Cycles. Int. J. Hydrogen Energy 2009, 34, 4817−4826. Crystals 2016, 6, 70.
(443) Tan, X.; Danaie, M.; Kalisvaart, W.; Mitlin, D. The Influence (460) Gao, J.; Ngene, P.; Lindemann, I.; Gutfleisch, O.; de Jong, K.
of Cu Substitution on the Hydrogen Sorption Properties of
P.; de Jongh, P. E. Enhanced Reversibility of H2 Sorption in
Magnesium Rich Mg-Ni Films. Int. J. Hydrogen Energy 2011, 36,
Nanoconfined Complex Metal Hydrides by Alkali Metal Addition. J.
2154−2164.
(444) Ouyang, L. Z.; Ye, S. Y.; Dong, H. W.; Zhu, M. Effect of Mater. Chem. 2012, 22, 13209−13215.
Interfacial Free Energy on Hydriding Reaction of Mg−Ni Thin Films. (461) Wang, L.; Rawal, A.; Quadir, M. Z.; Aguey-Zinsou, K.-F.
Appl. Phys. Lett. 2007, 90, 021917. Nanoconfined Lithium Aluminium Hydride (LiAlH4) and Hydrogen
(445) Tan, X.; Harrower, C. T.; Amirkhiz, B. S.; Mitlin, D. Nano- Reversibility. Int. J. Hydrogen Energy 2017, 42, 14144−14153.
Scale Bi-Layer Pd/Ta, Pd/Nb, Pd/Ti and Pd/Fe Catalysts for (462) Zhao, Y.; Han, M.; Wang, H.; Chen, C.; Chen, J. LiAlH4
Hydrogen Sorption in Magnesium Thin Films. Int. J. Hydrogen Energy Supported on TiO2/Hierarchically Porous Carbon Nanocomposites
2009, 34, 7741−7748. with Enhanced Hydrogen Storage Properties. Inorg. Chem. Front.
(446) Tan, X.; Wang, L.; Holt, C. M. B.; Zahiri, B.; Eikerling, M. H.; 2016, 3, 1536−1542.
Mitlin, D. Body Centered Cubic Magnesium Niobium Hydride with (463) Adelhelm, P.; de Jong, K. P.; de Jongh, P. E. How Intimate
Facile Room Temperature Absorption and Four Weight Percent Contact with Nanoporous Carbon Benefits the Reversible Hydrogen
Reversible Capacity. Phys. Chem. Chem. Phys. 2012, 14, 10904− Desorption from NaH and NaAlH4. Chem. Commun. 2009, 6261−
10909. 6263.
(447) Dibandjo, P.; Zlotea, C.; Gadiou, R.; Ghimbeu, C. M.; Cuevas, (464) Au, Y. S.; Yan, Y.; de Jong, K. P.; Remhof, A.; de Jongh, P. E.
F.; Latroche, M.; Leroy, E.; Vix-Guterl, C. Hydrogen Storage in Pore Confined Synthesis of Magnesium Boron Hydride Nano-
Hybrid Nanostructured Carbon/Palladium Materials: Influence of particles. J. Phys. Chem. C 2014, 118, 20832−20839.
Particle Size and Surface Chemistry. Int. J. Hydrogen Energy 2013, 38, (465) Wahab, M. A.; Young, D. J.; Karim, A.; Fawzia, S.; Beltramini,
952−965. J. N. Low-Temperature Hydrogen Desorption from Mg(BH4)2
(448) Zlotea, C.; Cuevas, F.; Paul-Boncour, V.; Leroy, E.; Dibandjo, Catalysed by Ultrafine Ni Nanoparticles in a Mesoporous Carbon
P.; Gadiou, R.; Vix-Guterl, C.; Latroche, M. Size-Dependent Matrix. Int. J. Hydrogen Energy 2016, 41, 20573−20582.
Hydrogen Sorption in Ultrasmall Pd Clusters Embedded in a (466) Yang, Y.; Liu, Y.; Li, Y.; Zhang, X.; Gao, M.; Pan, H. Towards
Mesoporous Carbon Template. J. Am. Chem. Soc. 2010, 132, 7720− the Endothermic Dehydrogenation of Nanoconfined Magnesium
7729. Borohydride Ammoniate. J. Mater. Chem. A 2015, 3, 11057−11065.
(449) Malouche, A.; Blanita, G.; Lupu, D.; Bourgon, J.; Nelayah, J.; (467) Li, L.; Yao, X.; Sun, C.; Du, A.; Cheng, L.; Zhu, Z.; Yu, C.;
Zlotea, C. Hydrogen Absorption in 1 nm Pd Clusters Confined in Zou, J.; Smith, S. C.; Wang, P.; et al. Lithium-Catalyzed Dehydrogen-
MIL-101(Cr). J. Mater. Chem. A 2017, 5, 23043−23052. ation of Ammonia Borane within Mesoporous Carbon Framework for
(450) Kalidindi, S. B.; Oh, H.; Hirscher, M.; Esken, D.; Wiktor, C.; Chemical Hydrogen Storage. Adv. Funct. Mater. 2009, 19, 265−271.
Turner, S.; Van Tendeloo, G.; Fischer, R. A. Metal@COFs: Covalent (468) Kim, H.; Karkamkar, A.; Autrey, T.; Chupas, P.; Proffen, T.
Organic Frameworks as Templates for Pd Nanoparticles and Determination of Structure and Phase Transition of Light Element
Hydrogen Storage Properties of Pd@COF-102 Hybrid Material. Nanocomposites in Mesoporous Silica: Case Study of NH3BH3 in
Chem. - Eur. J. 2012, 18, 10848−10856.
MCM-41. J. Am. Chem. Soc. 2009, 131, 13749−13755.
(451) Ponthieu, M.; Au, Y. S.; Provost, K.; Zlotea, C.; Leroy, E.;
(469) Klabunde, K. J. Nanoscale Materials in Chemistry; Wiley-
Fernandez, J. F.; Latroche, M.; de Jongh, P. E.; Cuevas, F.
Interscience: New York, 2001.
Nanoconfinement of Mg6Pd Particles in Porous Carbon: Size Effects
(470) Pundt, A.; Kirchheim, R. Hydrogen in Metals: Microstructural
on Structural and Hydrogenation Properties. J. Mater. Chem. A 2014,
Aspects. Annu. Rev. Mater. Res. 2006, 36, 555−608.
2, 18444−18453.
(471) Pundt, A. Hydrogen in Nano-Sized Metals. Adv. Eng. Mater.
(452) Gross, A. F.; Ahn, C. C.; Van Atta, S. L.; Liu, P.; Vajo, J. J.
Fabrication and Hydrogen Sorption Behaviour of Nanoparticulate 2004, 6, 11−21.
MgH2 Incorporated in a Porous Carbon Host. Nanotechnology 2009, (472) Wang, H.; Lin, H. J.; Cai, W. T.; Ouyang, L. Z.; Zhu, M.
20, 204005. Tuning Kinetics and Thermodynamics of Hydrogen Storage in Light
(453) Lim, D.-W.; Yoon, J. W.; Ryu, K. Y.; Suh, M. P. Magnesium Metal Element Based Systems - a Review of Recent Progress. J. Alloys
Nanocrystals Embedded in a Metal-Organic Framework: Hybrid Compd. 2016, 658, 280−300.
Hydrogen Storage with Synergistic Effect on Physi- and Chem- (473) Berube, V.; Chen, G.; Dresselhaus, M. S. Impact of
isorption. Angew. Chem., Int. Ed. 2012, 51, 9814−9817. Nanostructuring on the Enthalpy of Formation of Metal Hydrides.
(454) Huen, P.; Paskevicius, M.; Richter, B.; Ravnsbæk, D.; Jensen, Int. J. Hydrogen Energy 2008, 33, 4122−4131.
T. Hydrogen Storage Stability of Nanoconfined MgH2 Upon Cycling. (474) Singh-Miller, N. E.; Marzari, N. Surface Energies, Work
Inorganics 2017, 5, 57. Functions, and Surface Relaxations of Low-Index Metallic Surfaces
(455) Sun, Y.; Ma, T.; Aguey-Zinsou, K.-F. Magnesium Supported from First Principles. Phys. Rev. B: Condens. Matter Mater. Phys. 2009,
on Nickel Nanobelts for Hydrogen Storage: Coupling Nanosizing and 80, 235407.
Catalysis. ACS Appl. Nano Mater. 2018, 1, 1272−1279. (475) Vegge, T. Equilibrium Structure and Ti-Catalyzed H2
(456) Gosalawit-Utke, R.; Milanese, C.; Javadian, P.; Jepsen, J.; Desorption in NaAlH4 Nanoparticles from Density Functional
Laipple, D.; Karmi, F.; Puszkiel, J.; Jensen, T. R.; Marini, A.; Klassen, Theory. Phys. Chem. Chem. Phys. 2006, 8, 4853−4861.

10837 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

(476) Hazrati, E.; Brocks, G.; de Wijs, G. A. First-Principles Study of (496) David, W. I. F.; Jones, M. O.; Gregory, D. H.; Jewell, C. M.;
LiBH4 Nanoclusters and Their Hydrogen Storage Properties. J. Phys. Johnson, S. R.; Walton, A.; Edwards, P. P. A Mechanism for Non-
Chem. C 2012, 116, 18038−18047. Stoichiometry in the Lithium Amide/Lithium Imide Hydrogen
(477) Liu, X.; Peaslee, D.; Jost, C. Z.; Majzoub, E. H. Controlling Storage Reaction. J. Am. Chem. Soc. 2007, 129, 1594−1601.
the Decomposition Pathway of LiBH4 Via Confinement in Highly (497) Louchez, M. A.; Besson, R.; Thuinet, L.; Legris, A. Interfacial
Ordered Nanoporous Carbon. J. Phys. Chem. C 2010, 114, 14036− Properties of Hydrides in Α-Zr: A Theoretical Study. J. Phys.:
14041. Condens. Matter 2017, 29, 415001.
(478) Verdal, N.; Hartman, M. R.; Jenkins, T.; DeVries, D. J.; Rush, (498) Michel, K. J.; Ozoliņs,̌ V. Theory of Mass Transport in
J. J.; Udovic, T. J. Reorientational Dynamics of Nabh4 and Kbh4. J. Sodium Alanate. J. Mater. Chem. A 2014, 2, 4438−4448.
Phys. Chem. C 2010, 114, 10027−10033. (499) Wagemaker, M.; Mulder, F. M.; Van der Ven, A. The Role of
(479) Buchter, F.; Łodziana, Z.; Mauron, P.; Remhof, A.; Friedrichs, Surface and Interface Energy on Phase Stability of Nanosized
O.; Borgschulte, A.; Züttel, A.; Sheptyakov, D.; Strässle, T.; Ramirez- Insertion Compounds. Adv. Mater. 2009, 21, 2703−2709.
Cuesta, A. J. Dynamical Properties and Temperature Induced (500) Hjö rvarsson, B.; Andersson, G.; Karlsson, E. Metallic
Molecular Disordering of LiBH4 and LiBD4. Phys. Rev. B: Condens. Superlattices: Quasi Two-Dimensional Playground for Hydrogen. J.
Matter Mater. Phys. 2008, 78, 094302. Alloys Compd. 1997, 253−254, 51−57.
(480) Zhang, F.; Wood, B. C.; Wang, Y.; Wang, C. Z.; Ho, K. M.; (501) Gharavi, A. G.; Akyıldız, H.; Ö ztürk, T. Thickness Effects in
Chou, M. Y. Ultrafast Bulk Diffusion of AlHx in High-Entropy Hydrogen Sorption of Mg/Pd Thin Films. J. Alloys Compd. 2013, 580,
Dehydrogenation Intermediates of NaAlH4. J. Phys. Chem. C 2014, S175−S178.
118, 18356−18361. (502) Schwarz, R. B.; Khachaturyan, A. G. Thermodynamics of
(481) Shane, D. T.; Corey, R. L.; McIntosh, C.; Rayhel, L. H.; Open Two-Phase Systems with Coherent Interfaces. Phys. Rev. Lett.
Bowman, R. C.; Vajo, J. J.; Gross, A. F.; Conradi, M. S. LiBH4 in 1995, 74, 2523−2526.
Carbon Aerogel Nanoscaffolds: An Nmr Study of Atomic Motions. J. (503) Schwarz, R. B.; Khachaturyan, A. G. Thermodynamics of
Phys. Chem. C 2010, 114, 4008−4014. Open Two-Phase Systems with Coherent Interfaces: Application to
(482) Vajeeston, P.; Ravindran, P.; Fjellvåg, H. Theoretical Metal−Hydrogen Systems. Acta Mater. 2006, 54, 313−323.
Investigations on Low Energy Surfaces and Nanowires of MgH2. (504) Narayan, T. C.; Hayee, F.; Baldi, A.; Leen Koh, A.; Sinclair,
Nanotechnology 2008, 19, 275704. R.; Dionne, J. A. Direct Visualization of Hydrogen Absorption
(483) Peng, B.; Li, L.; Ji, W.; Cheng, F.; Chen, J. A Quantum Dynamics in Individual Palladium Nanoparticles. Nat. Commun. 2017,
Chemical Study on Magnesium(Mg)/Magnesium−Hydrogen(Mg− 8, 14020.
H) Nanowires. J. Alloys Compd. 2009, 484, 308−313. (505) Wadell, C.; Pingel, T.; Olsson, E.; Zorić, I.; Zhdanov, V. P.;
(484) Du, A. J.; Smith, S. C.; Yao, X. D.; Lu, G. Q. Role of Lithium Langhammer, C. Thermodynamics of Hydride Formation and
Vacancies in Accelerating the Dehydrogenation Kinetics on a Decomposition in Supported Sub-10nm Pd Nanoparticles of
LiBH4(010) Surface: An Ab Initio Study. J. Phys. Chem. C 2007, Different Sizes. Chem. Phys. Lett. 2014, 603, 75−81.
111, 12124−12128. (506) Züttel, A.; Nützenadel, C.; Schmid, G.; Emmenegger, C.;
(485) Du, A. J.; Smith, S. C.; Lu, G. Q. Vacancy Mediated Sudan, P.; Schlapbach, L. Thermodynamic Aspects of the Interaction
Desorption of Hydrogen from a Sodium Alanate Surface: An Ab Initio of Hydrogen with Pd Clusters. Appl. Surf. Sci. 2000, 162−163, 571−
Spin-Polarized Study. Appl. Phys. Lett. 2007, 90, 143119. 575.
(486) Hoang, K.; Janotti, A.; Van de Walle, C. G. The Particle-Size (507) Sachs, C.; Pundt, A.; Kirchheim, R.; Winter, M.; Reetz, M. T.;
Dependence of the Activation Energy for Decomposition of Lithium Fritsch, D. Solubility of Hydrogen in Single-Sized Palladium Clusters.
Amide. Angew. Chem., Int. Ed. 2011, 50, 10170−10173. Phys. Rev. B: Condens. Matter Mater. Phys. 2001, 64, 075408.
(487) Hoang, K.; Janotti, A.; Van de Walle, C. G. Mechanisms for (508) Lässer, R.; Klatt, K. H. Solubility of Hydrogen Isotopes in
the Decomposition and Dehydrogenation of Li Amide/Imide. Phys. Palladium. Phys. Rev. B: Condens. Matter Mater. Phys. 1983, 28, 748−
Rev. B: Condens. Matter Mater. Phys. 2012, 85, 064115. 758.
(488) Zaluska, A.; Zaluski, L.; Ström-Olsen, J. O. Nanocrystalline (509) Meethong, N.; Huang, H. Y. S.; Speakman, S. A.; Carter, W.
Magnesium for Hydrogen Storage. J. Alloys Compd. 1999, 288, 217− C.; Chiang, Y. M. Strain Accommodation During Phase Trans-
225. formations in Olivine-Based Cathodes as a Materials Selection
(489) Bouaricha, S.; Dodelet, J. P.; Guay, D.; Huot, J.; Schulz, R. Criterion for High-Power Rechargeable Batteries. Adv. Funct. Mater.
Activation Characteristics of Graphite Modified Hydrogen Absorbing 2007, 17, 1115−1123.
Materials. J. Alloys Compd. 2001, 325, 245−251. (510) Howie, A.; Marks, L. D. Elastic Strains and the Energy Balance
(490) Kato, S.; Bielmann, M.; Borgschulte, A.; Zakaznova-Herzog, for Multiply Twinned Particles. Philos. Mag. A 1984, 49, 95−109.
V.; Remhof, A.; Orimo, S.-i.; Züttel, A. Effect of the Surface Oxidation (511) Yao, X.; Zhu, Z. H.; Cheng, H. M.; Lu, G. Q. Hydrogen
of LiBH4 on the Hydrogen Desorption Mechanism. Phys. Chem. Diffusion and Effect of Grain Size on Hydrogenation Kinetics in
Chem. Phys. 2010, 12, 10950−10955. Magnesium Hydrides. J. Mater. Res. 2008, 23, 336−340.
(491) Carrillo-Bucio, J.; Tena-García, J.; Suárez-Alcántara, K. (512) Gibot, P.; Casas-Cabanas, M.; Laffont, L.; Levasseur, S.;
Dehydrogenation of Surface-Oxidized Mixtures of 2LiBH4 + Al/ Carlach, P.; Hamelet, S.; Tarascon, J. M.; Masquelier, C. Room-
Additives (TiF3 or CeO2). Inorganics 2017, 5, 82. Temperature Single-Phase Li Insertion/Extraction in Nanoscale
(492) Narayan, T. C.; Baldi, A.; Koh, A. L.; Sinclair, R.; Dionne, J. A. LixFePO4. Nat. Mater. 2008, 7, 741−747.
Reconstructing Solute-Induced Phase Transformations within Indi- (513) Niu, J.; Kushima, A.; Qian, X.; Qi, L.; Xiang, K.; Chiang, Y.-
vidual Nanocrystals. Nat. Mater. 2016, 15, 768. M.; Li, J. In Situ Observation of Random Solid Solution Zone in
(493) Mütschele, T.; Kirchheim, R. Segregation and Diffusion of LiFePO4 Electrode. Nano Lett. 2014, 14, 4005−4010.
Hydrogen in Grain Boundaries of Palladium. Scr. Metall. 1987, 21, (514) Van der Ven, A.; Garikipati, K.; Kim, S.; Wagemaker, M. The
135−140. Role of Coherency Strains on Phase Stability in LixFePO4: Needle
(494) Berube, V.; Dresselhaus, M. S.; Chen, G. Temperature Crystallites Minimize Coherency Strain and Overpotential. J.
Dependence of the Enthalpy of Formation of Metal Hydrides Electrochem. Soc. 2009, 156, A949−A957.
Characterized by an Excess Volume. Int. J. Hydrogen Energy 2008, 33, (515) Cogswell, D. A.; Bazant, M. Z. Coherency Strain and the
5617−5628. Kinetics of Phase Separation in LiFePO4 Nanoparticles. ACS Nano
(495) Weissmüller, J.; Lemier, C. Lattice Constants of Solid Solution 2012, 6, 2215−2225.
Microstructures: The Case of Nanocrystalline Pd-H. Phys. Rev. Lett. (516) Ichitsubo, T.; Tokuda, K.; Yagi, S.; Kawamori, M.; Kawaguchi,
1999, 82, 213−216. T.; Doi, T.; Oishi, M.; Matsubara, E. Elastically Constrained Phase-

10838 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839
Chemical Reviews Review

Separation Dynamics Competing with the Charge Process in the


LiFePO4/FePO4 System. J. Mater. Chem. A 2013, 1, 2567−2577.
(517) Bugaev, A.; Guda, A.; Lomachenko, K.; Shapovalov, V.;
Lazzarini, A.; Vitillo, J.; Bugaev, L.; Groppo, E.; Pellegrini, R.;
Soldatov, A.; et al. Core-Shell Structure of Palladium Hydride
Nanoparticles Revealed by Combined X-Ray Absorption Spectrosco-
py and X-Ray Diffraction. J. Phys. Chem. C 2017, 121, 18202−18213.
(518) Pop, E.; Varshney, V.; Roy, A. Thermal Properties of
Graphene: Fundamentals and Applications. MRS Bull. 2012, 37,
1273−1281.
(519) Li, A.; Zhang, C.; Zhang, Y. Thermal Conductivity of
Graphene-Polymer Composites: Mechanisms, Properties, and Appli-
cations. Polymers 2017, 9, 437.
(520) Ham, B.; Junkaew, A.; Arróyave, R.; Park, J.; Zhou, H. C.;
Foley, D.; Rios, S.; Wang, H.; Zhang, X. Size and Stress Dependent
Hydrogen Desorption in Metastable Mg Hydride Films. Int. J.
Hydrogen Energy 2014, 39, 2597−2607.
(521) Lei, H.; Wang, C.; Yao, Y.; Wang, Y.; Hupalo, M.; McDougall,
D.; Tringides, M.; Ho, K. Strain Effect on the Adsorption, Diffusion,
and Molecular Dissociation of Hydrogen on Mg (0001) Surface. J.
Chem. Phys. 2013, 139, 224702.
(522) Zhang, J.; Zhou, Y. C.; Ma, Z. S.; Sun, L. Q.; Peng, P. Strain
Effect on Structural and Dehydrogenation Properties of MgH2
Hydride from First-Principles Calculations. Int. J. Hydrogen Energy
2013, 38, 3661−3669.
(523) Zhao-Karger, Z.; Witter, R.; Bardají, E. G.; Wang, D.;
Cossement, D.; Fichtner, M. Influence of Nanoconfinement on
Reaction Pathways of Complex Metal Hydrides. Energy Procedia
2012, 29, 731−737.
(524) Khomyakov, P.; Giovannetti, G.; Rusu, P.; Brocks, G.; van den
Brink, J.; Kelly, P. First-Principles Study of the Interaction and Charge
Transfer between Graphene and Metals. Phys. Rev. B: Condens. Matter
Mater. Phys. 2009, 79, 195425.
(525) Du, A.; Ng, Y.; Bell, N.; Zhu, Z.; Amal, R.; Smith, S. Hybrid
Graphene/Titania Nanocomposite: Interface Charge Transfer, Hole
Doping, and Sensitization for Visible Light Response. J. Phys. Chem.
Lett. 2011, 2, 894−899.
(526) Berseth, P. A.; Harter, A. G.; Zidan, R.; Blomqvist, A.; Araújo,
C. M.; Scheicher, R. H.; Ahuja, R.; Jena, P. Carbon Nanomaterials as
Catalysts for Hydrogen Uptake and Release in NaAlH4. Nano Lett.
2009, 9, 1501−1505.
(527) Carr, C. L.; Jayawardana, W.; Zou, H.; White, J. L.; El Gabaly,
F.; Conradi, M. S.; Stavila, V.; Allendorf, M. D.; Majzoub, E. H.
Anomalous H2 Desorption Rate of NaAlH4 Confined in Nitrogen-
Doped Nanoporous Carbon Frameworks. Chem. Mater. 2018, 30,
2930−2938.
(528) Lototskyy, M.; Sibanyoni, J.; Denys, R.; Williams, M.; Pollet,
B.; Yartys, V. Magnesium-Carbon Hydrogen Storage Hybrid Materials
Produced by Reactive Ball Milling in Hydrogen. Carbon 2013, 57,
146−160.
(529) Du, A.; Smith, S.; Yao, X.; Lu, G. Catalytic Effects of
Subsurface Carbon in the Chemisorption of Hydrogen on a
Mg(0001) Surface: An Ab-Initio Study. J. Phys. Chem. B 2006, 110,
1814−1819.
(530) Miyaoka, H.; Ichikawa, T.; Kojima, Y. The Reaction Process
of Hydrogen Absorption and Desorption on the Nanocomposite of
Hydrogenated Graphite and Lithium Hydride. Nanotechnology 2009,
20, 204021.
(531) Lee, Y. W.; Clemens, B. M.; Gross, K. J. Novel Sieverts’ Type
Volumetric Measurements of Hydrogen Storage Properties for Very
Small Sample Quantities. J. Alloys Compd. 2008, 452, 410−413.
(532) Parilla, P. A.; Gross, K.; Hurst, K.; Gennett, T. Recommended
Volumetric Capacity Definitions and Protocols for Accurate, Stand-
ardized and Unambiguous Metrics for Hydrogen Storage Materials.
Appl. Phys. A: Mater. Sci. Process. 2016, 122, 201.
(533) Broom, D. P.; Hirscher, M. Irreproducibility in Hydrogen
Storage Material Research. Energy Environ. Sci. 2016, 9, 3368−3380.

10839 DOI: 10.1021/acs.chemrev.8b00313


Chem. Rev. 2018, 118, 10775−10839

You might also like