当前位置:首页 >> 能源/化工 >>

Emerging concepts in solid-state hydrogen storage


Energy & Environmental Science
Cite this: Energy Environ. Sci., 2012, 5, 5951 www.rsc.org/ees
Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/

09/2013 04:07:08.

View Article Online / Journal Homepage / Table of Contents for this issue

Dynamic Article Links <

C

PERSPECTIVE

Emerging concepts in solid-state hydrogen storage: the role of nanomaterials design
Hazel Reardon,a James M. Hanlon,a Robert W. Hughes,a Agata Godula-Jopek,b Tapas K. Mandalac and Duncan H. Gregory*a
Received 11th November 2011, Accepted 26th January 2012 DOI: 10.1039/c2ee03138h This perspective highlights the state-of-the-art solid-state hydrogen storage and describes newly emerging routes towards meeting the practical demands required of a solid-state storage system. The article focuses both on the physical and chemical aspects of hydrogen storage. Common to both classes of storage material is the concept of nanostructure design to tailor kinetics and thermodynamics; whether this be control of functionalised porosity or crystalline growth on the nanoscale. In the area of chemical storage, different processing and nanostructuring techniques that have been employed to overcome the barriers of slow kinetics will be discussed in addition to new chemical systems that have emerged. The prospects of porous inorganic solids, coordination polymers (metal organic frameworks; MOFs) and other polymeric matrices for physical storage of hydrogen will be highlighted. Additionally the role of inorganic nanostructures as evolving materials ‘‘intermediate’’ between physical and chemical storage systems will be discussed and their place within the ?ne thermodynamic balance for optimum hydrogen uptake and release considered.

1. Introduction
The adverse effects of global warming and consequent climate change demands the international utilisation of green and renewable sources of energy. The burgeoning need for energy coupled with the rapid depletion of fossil fuels pose serious threats for sustainable development. Hydrogen is a primary candidate as a future energy carrier but before its potential can be
a West CHEM, School of Chemistry, Joseph Black Building, University of Glasgow, Glasgow, UK. E-mail: Duncan.Gregory@Glasgow.ac.uk; Fax: +44 (0)141 330 4888; Tel: +44 (0)141 330 6438 b EADS Innovation Works, Energy& Propulsion, 81663 Munich, Germany c Department of Chemistry, Indian Institute of Technology Roorkee, Roorkee 247667, Uttarakhand, India

realised, fundamental research including components of invention and discovery, subsequent implementation of new technology and socio-economic acceptance must occur. These are the key steps to the hydrogen energy transition. The production and storage of hydrogen are the two most important steps that currently represent a bottleneck to utilization of hydrogen more widely in, for example, fuel cell systems. It must be realized that unlike coal or oil, hydrogen is not naturally available. It is an energy carrier rather than a source of energy itself. Thus, stable large-scale hydrogen production from non-polluting sources is essential to reduce global CO2 emission. However a safe, light-weight and inexpensive storage medium is likely to be a crucial prerequisite before hydrogen would be

Broader context
Increases in CO2 emissions and probable causal links to climate change demand the adoption of renewable sources of energy on a global scale. Hydrogen is a leading candidate as an energy vector but its production and storage represent major challenges to its utilization. The ef?cient storage of hydrogen in a solid medium is probably the most demanding and challenging part of realising the hydrogen economy as far as mobile applications are concerned. For a storage material for mobile applications reversibility/cyclability is vital and the thermodynamics and kinetics of hydrogen release are key. In this context, the strength of the interaction by which the hydrogen is held within the materials is a vital parameter to be able to tune as is a structural-chemical control over the rate at which hydrogen is taken up and released. In this perspective we aim to demonstrate how nanodesign principles can begin to address the challenges of practical solid-state storage solutions. Such approaches could involve simple physical size reduction, the addition of nanosized additives and catalysts or the systematic chemical design of nanomaterials through nanoscale functionalisation, nanocon?nement or chemically-directed fabrication.
This journal is ? The Royal Society of Chemistry 2012 Energy Environ. Sci., 2012, 5, 5951–5979 | 5951

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

globally acceptable as a fuel, particularly in the automotive industry. It should be emphasised here that while gaseous or liquid hydrogen is currently an option for prototype personal vehicles (cars) or larger commercial transport, solid-state storage of hydrogen is potentially far superior with regard to its storage capacity (both gravimetric and volumetric), energy ef?ciency and safety.1–8 Nevertheless compact storage of hydrogen in a solid medium is the most demanding and challenging part of realising the hydrogen economy as far as mobile applications are concerned.1 In this perspective, we highlight the state-of-the-art in hydrogen storage materials covering both physical and chemical means of storage and where these two categories of materials may interface. For a material for mobile applications reversibility/cyclability is vital and the thermodynamics and kinetics of hydrogen release are key. In this context, the strength of the interaction by which the hydrogen is held within the materials is a vital parameter to be able to tune. The measure of this interaction, the ‘heat of adsorption’, is directly related to the temperature at which the hydrogen can be released. The underpinning theme of this perspective is the potential of nanometric materials design and how newly emerging materials

and systems constructed along nanodesign principles begin to address the challenges of practical storage solutions. In many respects, nanochemical concepts and nanostructuring offer potential improvements to both the thermodynamic and kinetic storage performance of existing materials and in designing new materials with tailored sorption properties. As will be seen in the following sections, this could be achieved by physical size reduction, by the addition of nanosized additives and catalysts, by the chemical design and functionalisation of nanometric cavities or channels, by the encapsulation or nanocon?nement of active storage species or by the chemically directed fabrication of speci?c nanostructured materials. (For an excellent account of recent developments in the nanocon?nement of hydrides, the reader is also directed to the review by Nielsen et al.6) An appreciation of the nanoscale and the regime between surface and bulk is also vital in understanding the process of uptake, retention and release of hydrogen by solids on a fundamental level. We thus highlight how a nanomaterials design approach aids such a comprehension and evaluate how such an approach applies to a cross section of modern storage materials from porous solids through inorganic nanoparticles to metal and complex hydrides.

Hazel Reardon

Hazel Reardon graduated from the University of Strathclyde with an MSci in Analytical Chemistry in 2008. She then took up a position as a technical specialist in industry whilst undertaking a Royal Society of Chemistry accredited graduate programme. In 2010 she began a PhD at the University of Glasgow with Prof. Duncan H. Gregory on the development of hydrogen storage materials. Her research interests are focused on new materials for modern energy and fuel solutions.

Robert W: Hughes

Robert W. Hughes graduated in Chemistry at University College London in 1999. He then moved to the University of Southampton where he obtained a PhD in solid state chemistry in 2004 and remained as a postdoctoral research assistant until 2007. He is currently a WestCHEM Research Fellow at the University of Glasgow. His research interests include microwave-assisted nanomaterials synthesis and structural investigations using neutron diffraction. Agata Godula-Jopek studied at the Technical University in Cracow, completing her PhD thesis in 1999 in the Department of Electrochemical Oxidation of Gaseous Fuels at the Institute of Physical Chemistry of the Polish Academy of Sciences in Cracow. Presently she is fuel cells expert at EADS Innovation Works in Munich. Her research interests centre on fuel cells, hydrogen storage and fuel processing for fuel cells.

James Hanlon

James Hanlon graduated from the University of Glasgow in 2005 with a BSc (Hons) in Chemistry with Medicinal Chemistry. He then returned to the University of Glasgow in 2008 to undertake a PhD with Prof. Duncan H. Gregory on Hydrogen Storage Materials under the ESPRC SUPERGEN program. His research interests include nanostructured hydrogen storage materials, hydrogen release systems and ammonia storage materials.

Agata Godula-Jopek

5952 | Energy Environ. Sci., 2012, 5, 5951–5979

This journal is ? The Royal Society of Chemistry 2012

View Article Online

2. Physical hydrogen storage
The concept of physical storage requires hydrogen to be bound to/included in a host - a surface or porous solid - by relatively weak interactions/forces commonly understood as a process of physical adsorption or physisorption. In physisorption the forces of attraction between the hydrogen and the host originate mainly from weak van der Waals interactions, thereby restricting any signi?cant hydrogen adsorption to that at very low temperatures and/or at very high pressures. Coordination polymers, a.k.a. metal organic frameworks (MOFs), covalent organic frameworks (COFs), polymers with intrinsic microporosity (PIMs) and zeolites are among the widely investigated examples of materials for hydrogen storage that operate via physisorption. Common to each group of materials are internal chemicallyfunctionalisable spaces that can be engineered at the nanoscale. Carbon has also been extensively investigated as a hydrogen storage material in its many different forms. However, a review of this topic is beyond the scope of this perspective and a number of excellent review articles consider the many aspects of this ongoing work.9–14

2.1 Zeolites Zeolites belong to a large group of porous, aluminosilicate minerals of different chemical composition, properties and crystalline form. Aside from potential applications in hydrogen storage, zeolites are gaining attention for their ability to reduce environmentally harmful species like, for example, nitrogen oxides and carbon dioxide, produced by the combustion of fossil fuels. The most important feature of zeolites and that which determines their selective properties, is the presence of microand/or mesopores within their structures. The con?guration of these pores governs the transport phenomena of guest species in zeolites. Zeolites were among the ?rst inorganic porous solids to be investigated for gas storage and thus in many respects offer well understood models of how hydrogen can be contained in channels designed at the nanoscale. Hydrogen is stored in

zeolites under cryogenic conditions by physisorption and encapsulation (trapping).15,16 In the latter case, molecules are forced into the normally inaccesible zeolitic cages at elevated temperature and pressure. Upon cooling to room temperature, hydrogen is trapped inside the pores and can be released by raising the temperature or applying force. A factor limiting the storage capacity of zeolites, however, is the relatively high mass of the framework (containing Si, Al, O and heavy cations).17 Nijkamp et al. investigated hydrogen storage capacities for numerous aluminosilicates at 77 K and 1 bar, with compounds representing signi?cant variations in surface areas and pore volumes.18 Experimental studies showed that mesoporous silicas and aluminas yielded low H2 storage capacities. MCM-42 with a BET surface area of 1017 m2g?1 had a total hydrogen uptake of 65 ml (STP) g?1 while zeolites with microporous structures like ZSM-5 or ferrierite exhibited enhanced capacities of 80 and 65 ml (STP) g?1, respectively. A review of hydrogen storage capacities by Vitillo et al. showed gravimetric capacities oscillating between ca. 0.0 wt% (by encapsulation) and 1.8 wt% (by adsorption) depending on the material framework type and loading conditions (T, P).15 For example, ZSM-5 (MFI type) uptakes 0.24 wt% H2 at 77 K and 0.8 bar and NaY (FAU type) uptakes 1.81 wt% H2 under the same conditions. By contrast, ZK-5 (KFI type) was only able to store 0.018 wt% H2 at 573 K, 100 bar. Further, atomistic simulations to obtain realistic models of hydrogen packing in micropores by progressive ?lling with hydrogen were evaluated for twelve zeolitic frameworks with pure SiO2 composition: CHA, DDR, FAU, FER, KFI, LTA, LTL, MEL, MFI, MOR, RHO and SOD. Calculated zeolite gravimetric capacities were in the range 2.86 wt% H2 (for RHO and FAU) 1.50 wt% H2 (for FER). The investigation revealed it was possible to predict the level of structural ?exibility in zeolitic frameworks a ?nding which could be extrapolated to the design of materials for gas separation and catalysis. The use of zeolites as potential hydrogen storage materials was also studied in depth by Langmi et al. with particular emphasis on zeolite type A, X, Y and RHO.19,20 With these, a maximal hydrogen storage capacity (at 15 bar and 77 K) of 1.8 wt% was

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Tapas K Mandal received his Ph. D. from the Indian Institute of Science, Bangalore, India in 2005. From 2005–2008, he was a Post-Doctoral Associate at the University of New Orleans and Rutgers, The State University of New Jersey, USA. From 2008– 2010, he worked as PostDoctoral Research Associate at the University of Glasgow. He returned to India as an Assistant Professor at Sikkim University Tapas K Mandal (2010–2011). He is currently an Assistant Professor at the Indian Institute of Technology Roorkee. His interests include inorganic solid-state and materials chemistry, nanomaterials chemistry and hydrogen energy.
This journal is ? The Royal Society of Chemistry 2012

Duncan H Gregory

Duncan H Gregory studied at the University of Southampton completing his PhD in 1993 under Prof. Mark Weller. He was an EPSRC Advanced Fellow, Lecturer and Reader in Materials Chemistry at the University of Nottingham until 2006. He then took up the WestCHEM Chair in Inorganic Materials at the University of Glasgow and is Head of Inorganic Chemistry. His research interests centre on materials with potential applications in areas including sustainable energy.

Energy Environ. Sci., 2012, 5, 5951–5979 | 5953

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

obtained for NaY, followed by MgY with 1.74 wt%, MgX 1.61 wt%, CdA 1.14 wt% and others until a value of 0.0 wt% was reached for NaCsRHO. As can be seen in Fig. 1, hydrogen uptake followed a type I BET isotherm, indicating physisorption type processes at low temperatures. In a later study by the same authors, zeolite X, Y, A and RHO were studied with various exchangeable cations at 77 K and up to 15 bar H2.20 The highest gravimetric storage capacity of 2.19 wt% was reported for CaX. Volumetric storage densities of up to 31.0 kg H2 m?3 (143 H2 molecules/unit cell) and 30.2 kg H2 m?3 (144 H2 molecules/unit cell) were achieved for CaX and KX, respectively. The authors noted that large numbers of alkali metal cations can affect the available void volume, therefore limiting the pore space for hydrogen adsorption. Kazansky et al. had previously studied hydrogen adsorption at 77 K in sodium faujasites, NaX, with different Si:Al framework ratios of 1.05, 1.2 and 1.4 and compared these to NaY with a Si:Na ratio of 2.4.21 From the adsorption isotherms it was found that sodium ions act as adsorption sites with the optimum soprtion in NaX with an Si:Al ratio of 1.05 (ca.4.3 ? 1021 H2 molecules g?1) and the poorest sorption in NaY (ca.1.1 ? 1021 H2 molecules g?1), as can be seen in Fig. 2. More recently hydrogen sorption in organic ion-exchanged zeolite-Y (OZ) was investigated by Bae et al.22 By introducing pyridine hydrochloride and pyridinium chlorochromate into the porous structure of zeolite-Y, hydrogen uptake increased, to 0.15 and 0.34 wt% respectively at 298 K and 100 bar (10 MPa) H2. Based on these results Bae et al. speculate, that the sorption ability of zeolite-Y could be tailored with the use of different organic molecules and hence hydrogen storage capacity increased. Further, Kayiran and Darkrim exchanged lithium and potassium for sodium in zeolite NaA and observed that hydrogen adsorption decreased with the extent of exchange for lithium.23 The potassium-exchanged zeolite, however, adsorbed more hydrogen than the other zeolites. Kayiran and Darkrim

Fig. 2 Isotherms of hydrogen adsorption at 77 K on faujasites with different Si:Al ratios in the framework. Reprinted from Micro. Meso. Mater.,1998, 22, 251. Low temperature hydrogen adsorption on sodium forms of faujasites: barometric measurements and drift spectra, V.B. Kazansky, V.Yu. Borovkov, A. Serich and H.G. Karge. Copyright 1998, with permission from Elsevier.

believe that the ability of zeolites to tailor H2 sorption zeolites via ion exchange augurs well for prospects for room temperature storage, although appreciable uptake at ambient temperature is yet to be observed. Du and Wu reported marked contrasts in hydrogen uptake in A, X and ZSM-5 zeolites (at 77–293 K; up to 7 bar H2), based on both the framework structure and the constituent cations.24 At 77 K hydrogen adsorption reached up to 2.55 wt% H2 for NaX and inevitably decreased with increasing temperature (Table 1). Weitkamp et al. focused on relationships between H2 sorption, zeolite pore architecture and composition across temperatures from 293–573 K and pressures from 25–100 bar.25 A relationship between the amount of entrapped hydrogen and the ionic radius of the cation in zeolite A was apparent with the stored hydrogen volume per gram of zeolite increasing from Na+ to K+ and then decreasing radically from K+ to Rb+ and Cs+. It was proposed that by increasing the cation size and reducing the effective pore width to a critical value, it is possible to encapsulate H2 more effectively. Further, storage capacity was generally higher for zeolites with a larger number of smaller pores. 2.2 Metal–organic frameworks (MOFs) Metal – Organic Frameworks (MOFs) are a class of physical storage materials that have been researched extensively for hydrogen storage purposes in recent years.26 MOFs consist of inorganic clusters that are connected through strong bonds via organic linkers. The frameworks of MOFs are interconnected allowing an ordered network of channels or pores. The porosity of MOFs allows gas storage of small molecules, for example CO2, CH4 and H2. Gas can be captured and released from the pores of these structures and this process is reversible. The structure of MOFs can be engineered to the speci?c function required of that MOF: pore size and surface area can be tailored during synthesis. MOFs are suitable for gas adsorption purposes as they possess high porosity, high surface areas, high crystallinity, low density and are relatively cheap and easy to synthesise. They offer fast kinetics and high reversibilty for H2 storage. MOFs can be synthesised by a number of techniques and recently most commonly by solvothermal and hydrothermal
This journal is ? The Royal Society of Chemistry 2012

Fig. 1 Hydrogen adsorption/desorption isotherms at 77 K for CdA zeolite with 0.1 bar pressure steps. Reprinted from J. Alloys Compd., 356– 357, 710. Hydrogen adsorption in zeolites A, X, Y and RHO, H.W. Langmi, A. Walton, M.M. Al-Mamouri, S.R. Johnson, D. Book, J.D. Speight, P.P. Edwards, I. Gameson, P.A. Anderson and I.R. Harris Copyright 2003, with permission from Elsevier.

5954 | Energy Environ. Sci., 2012, 5, 5951–5979

View Article Online
Table 1 Surface area of selected zeolites and measured hydrogen adsorption at different temperatures24 SBET(m2 g?1) 565 445 428 36.2 32.0 H2 adsorption at 77 K, wt (%) 2.55 1.97 1.74 1.64 0.07 H2 adsorption at 195 K, wt (%) 0.98 0.65 0.61 0.44 N/A H2 adsorption at 293 K, wt (%) 0.40 0.38 0.31 0.27 N/A

Zeolite NaX ZSM-5 CaA NaA KA

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

techniques. They can also be synthesised by microwave synthesis, sonochemical and mechanical synthesis as reviewed in more detail by Meek.27 MOF nano?bres have also been recently synthesised by electrospinning (nanoparticles of the MOF are electrospun with carrier polymers to produce nano?bres between 150–300 nm in diamater).28 MOFs ?rst came to prominence as candidate hydrogen storage materials in 2003 when Yaghi et al. synthesised MOF-5 which consists of an inorganic {OZn}4 framework linked by 1.4-benzenedicarboxylate groups.29 MOF-5 was found to absorb 4.5 wt% H2 at 78 K and 20 bar, although H2 uptake decreased to 1.0 wt% H2 at room temperature. IRMOF-8, consisting of cyclobutylbenzene and naphthalene linkers was also synthesised and was found to absorb 2.0 wt% H2.29 A number of MOFs have since been synthesised with high H2 % uptakes at low temperatures. One of the most encouraging is MOF-177 which consists of {Zn4O}6+ linked with 1,3,5 benzetribenzoate and can absorb 7.5 wt% H2 at 77 K and 70 bar.30,31 MIL-101 consisting of benzene1.4-dicarboxylate (bdc) and trimeric chromium(III) octahedral clusters has been found to absorb 6.1 wt% H2 at 77 K with a heat of adsoption of 10 kJ mol?1 at 77 K.32 From the study of MIL100 and MIL-101, ‘‘small’’ pores in MOFs were found to be bene?cal for H2 storage. Larger cages lead to larger pores and reduced interaction with H2 and despite a more ?exible framework, this is a feature that MOFs share with zeolites and other porous inorganic solids. Further, MIL-101 illustrates the importance of high surface area (5500 m2g?1) in the H2 storage capacity of MOFs, as is the case in zeolites. To further empahsise this point, recently synthesised MOF-210 has extremely high porosity (pore size of 3.60 cm3 g?1 and a BET surface area of 6240 m2g?1).33 The excess H2 uptake of MOF-210 (86 mg g?1) is the highest at the time of writing (Fig. 3). Selected MOFs with signi?cant H2 storage properties are summarised in Table 2. More comprehensive tables can found in recent reviews by Hu et al.,34 Murray et al.35 and Scully et al.36 From Table 2 it is evident that there is a strong correlation between the gravimetric capacity and the surface area of MOFs. The relationship between uptake and pore size, however, is not necessarily a straightforward one. For example, Zhong et al. synthesised a 3D microporous cadmium(III) MOF with small pore size39 but found that both small pore volumes and the tetrazoyl-ring decorated inner surfaces of the pores contributed to the value of the heat of H2 adsorption of 13.3 kJ mol?1 at 77 K. This result, among others, illustrates not only the importance of pore size in the design of MOFs but also the the design of the pores themselves. The key role of binding sites in MOFs was demonstrated in an elegant powder neutron diffraction (PND) study by Yan et al.40 Two MOFs, NOTT-112 and NOTT-116
This journal is ? The Royal Society of Chemistry 2012

were investigated and although the latter possesed a higher surface area the former has a higher maximum excess H2 uptake (7.07 wt% H2 at 77 K and 35 bar). In NOTT-112 it was found by PND that D2 binds strongly to the exposed Cu(II) sites (denoted site A1) in the smallest cuboctahedral cages – denoted CuA – rather than to copper located in truncated tetrahedral cages – denoted CuB (Fig. 4). Importantly, this result demonstrates both that optimum pore size need not correlate directly to high surface area and that creation of exposed metal sites can directly in?uence H2 - MOF interaction strength. The concept of metal site exposure is dicussed further below (section 2.2.1). The major drawback to MOFs being considered as candidates to meet the US Dept. of Energy (DOE) targets for use in hydrogen powered vehicles is the temperature at which uptake occurs. High wt% H2 uptake occurs at 77 K whereas at room temperature uptake decreases to ca. 1 wt% H2.29 The key hurdle to H2 storage in MOFs, therefore, is the small hydrogen adsorption enthalpy at room temperature. Bhatia and Myers have suggested that an H2 binding energy of 15.1 kJ mol?1 is required at room temperature for a porous material to be considered as a possible hydrogen store.41 Palomino et al., using unsaturated Mg2+ in Mg-MOF-74 and unsaturated Co2+ in Co-MOF-74, suggest that for effective hydrogen delivery at room temperature and pressures between 30–1.5 bar, the H2 adsorption enthalpy needs to be between 22–25 kJ mol?1.42 If the adsorption energies can be increased so that the H2 storage

Fig. 3 Excess H2 uptake of MOF-210 at 77 K compared to other selected high performing MOFs. From H. Furukawa, N. Ko, Y.B. Go, N Aratani, S.B. Choi, E. Choi, A.O. Yazaydin, R.Q. Snurr, M.O’Keeffe, J. Kim and O. M. Yaghi Science, 2010, 329, 424. Reprinted with permission from AAAS.

Energy Environ. Sci., 2012, 5, 5951–5979 | 5955

View Article Online
Table 2 Selected MOFs with H2 gravimetric capacities at 77 K Compound MOF-5 IRMOF-6 IRMOF-11 Cu2(QPTC) HKUST-1 MIL-101 wt% H2 uptake 4.5 4.8 (50 bar) 3.5 (35 bar) 6.06 (20 bar) 3.6 (10 bar) 6.1 Speci?c Surface Area / m2g?1 2000 3300 2340 2932 1958 5550 Reference 29 29 29 37 38 32

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

emphasised recently by Gedrich et al.46 Supercritically dried DUT-9, possessing a high concentration of exposed nickel sites in conjunction with a high surface area and high porosity (the   structure contains pores of 13 A and 25 A diameters), demonstrated an H2 uptake of 4.99 wt% H2 at 45 bar at 77 K which compares to otherwise similar Ni-MOFs with unexposed metal sites that adsorbed 1.1 wt% at 77 K, 1 bar H2.47 Moreover, Farha et al.48 have very recently designed a MOF (NU-100) with exposed Cu(II) sites which has a BET surface area of 6134 m2 g?1 and a gravimetric uptake of 9.05% H2 at 56 bar and 77 K. 2.2.2 Metal doping. Doping MOFs with metal ions is another method for increasing the H2 binding energies and generally involves taking advantage of the large pores in MOFs to ?ll them with metal ions. Doping MOFs with Li+ ions has been studied to the greatest extent to date and Yang et al. doped with Li+ in order to trap H2 in MOFs consisting of In(III) centres and tetracarboxylic ligands (NOTT-200; undoped and NOTT-201; doped).49 As a result of doping the adsorption enthalpy was increased by 1 kJ mol?1 in NOTT-201 (10.1 kJ mol?1 at zero surface coverage) over NOTT-200. Moreover Klontzas et al. used Monte Carlo simulations to show that doping IRMOF-8 with Li (through use of a Li-alkoxide linker) leads to hydrogen adsorption reaching 1.57 wt% H2 for (IRMOF-8) which is 75 times more than the undoped (IRMOF-8) compound.50 The doping strategy proposed by Klontzas et al. was employed experimentally in the study of hydrogen adsorption by MIL-53.51 The MOF was doped with lithium diisopropylamide (LDA) which increased the H2 uptake from 0.5 wt% to 1.7 wt% at 77 K, 1 bar H2. The isosteric heat of adsorption also improved from 5.8 kJ mol?1 to 11.6 kJ mol?1 on Li doping. DFT and grand canonical ensemble Monte Carlo (GCMC) calculations were used to predict optimum structures for Li-doped zinc carboxylate MOFs and their hydrogen uptake respectively. The studies show dramatic enhancement of H2 uptake at 300 K, 10 bar on Li doping of these MOFs and also a correlation of uptake with surface area (Fig. 5).52

Fig. 4 Structure of NOTT-112 from neutron diffraction (a) D2 positions in the cuboctahedral cage at D2 loading as 0.5 D2/Cu; (b) D2 positions in the cage A and cage B at 2.0 D2/Cu D2 loading; (c) view of ?ve D2 positions (A1, A2, A3, A4, and A5) at 2.0 D2/Cu D2 loading; gray, carbon; red, oxygen; turquoise, copper. The D2 positions are represented by coloured spheres: A1, lavender; A2, blue; A3, yellow; A4, orange; A5, green. Reprinted with permission from Y. Yan, I. Telepeni, S. Yang, X. Lin, W. Kockelmann, A. Dailly, A.J. Blake, W. Lewis, G.S. Walker, D.R. Allan, S.A. Barnett, N.R. Champness and M. Schrder, J. Am. o Chem. Soc., 2010, 132, 4092. Copyright 2010 American Chemical Society.

values are favourable at room temperature (and hence in line with the US DOE targets) then given their fast physisorption kinetics, MOFs would be ideal hydrogen storage materials. A number of strategies to increase the H2 binding energy are currently being investigated.35,43 The remainder of section 2.2 will focus brie?y on these. 2.2.1 Metal site exposure. As highlighted above, exposing metal sites has been considered in some depth in order to improve the H2 binding energy in MOFs.37 Exposing metal sites effectively involves the removal of a ligand from the framework leaving the metal site on the MOF exposed for H2 binding. In order to expose a metal site, a terminal linker is typically removed, exposing the metal ion and Lochran et al., for example, used computational energy decomposition analysis to determine that incorporating transition metal groups into organic linkers can improve H2 binding energies at room temperature.44 Vitillo et al. compared sorption properties of MOF-5, with inacessible Zn2+ metal sites, to MOFs with exposed Cu2+ sites (HKUST-1) and Ni2+ sites (CPO-27-Ni) and although adsorption temperatures (and thus adsorption enthapies) were highest for CPO-27-Ni, at high H2 pressure, MOF-5 (with a larger surface area and higher free pore volume) outperformed the exposed metal MOFs. Hence, it was concluded that to exploit the stronger adsorption from exposed metal sites, MOF design needs to improve the surface density of the sites.45 This has been
5956 | Energy Environ. Sci., 2012, 5, 5951–5979

Fig. 5 Comparison of Li-doped and undoped Zn-carboxylate MOFs at 300 K and 100 bar from GCMC calculations. Reprinted with permission from S.S. Han and W. A. Goddard III, J. Am. Chem. Soc., 2007, 129, 8422. Copyright 2007 American Chemical Society.

This journal is ? The Royal Society of Chemistry 2012

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

2.2.3 Impregnation. Impregnation, by analogy to metal doping, takes advantage of the range of pore sizes available in various MOFs and involves the insertion of either another framework or a non-volatile guest. Depending on the guest, additional binding sites may also be created for H2 uptake. Impregnation was investigated by Chae et al. for MOF-17753  which, with a pore diameter of 11.8 A, made it suitable for impregnation by fullerenes. No H2 uptake measurements were performed on this sample. The authors also used Astrazon Orange R, Nile Red and Reichardt’s dye to determine if MOF177 could absorb large organic molecules, with only Reichardt’s dye being too large to penetrate into the inner volumes. Subsequently, and following earlier experimental evidence that metal decorated fullerenes can improve the adsorption of H2,54 the impregnation of magnesium decorated fullerenes was investigated computationally using Lennard-Jones parameters.55 The H2 adsorption energy predicted was 11 kJ mol?1 with H2 uptake up to 7.6 wt% at 77 K, 10 atm (for IRMOF-10). Similarly, GCMC simulations found that by impregnating an Li-doped MOF with Li coated fullerenes resulted in an H2 storage capacity of 6.3 wt% at 243 K and 100 bar.56 Prasanth et al. extended these impregnation concepts by incorporating single wall carbon nanotubes in MIL-101 (a strategy also employed with polymer matrices – see section 2.4.2).57 A carbon nanotube doping level of 8 wt% inside MIL-101 resulted in the H2 uptake increasing from 6.37 wt% to 9.18 wt% at 77 K and 60 bar and an improved H2 uptake at 298 K from 0.23 wt% to 0.64 wt%. 2.2.4 Hydrogen spillover. Hydrogen spillover is a phenomenon that has been well known in heterogeneous catalysis for several decades.58 More recently the phenomenon has found relevance in hydrogen storage in a number of sorbate systems including MOFs. The phenomenon involves the dissociation of H2 on to a metal surface. The hydrogen is then ‘‘spilled over’’ on to the porous material with the aid of a catalyst.59 Accordingly, catalytic nanoparticles are commonly used for hydrogen spillover in MOF systems. Zlotea et al. found that doping MIL-101 with Pd resulted in a hydrogen uptake of 0.35 wt% H2 at room temperature via hydrogen spillover compared with 0.19 wt% H2 in undoped material.60 In other examples, doping SNU-3 with Pd resulted in an H2 uptake increase from 1.03 wt% to 1.48 wt% at 77 K and 1 bar,61 while Pt doping in MOF-177 results in an H2 uptake of 2.5 wt% at 298 K, 144 bar in the ?rst cycle, but with capacity dropping to 0.5 wt% on cycling.62 Solution in?ltration of Pd into MOF-5 results in H2 adsorption increasing from 1.3 wt% to 1.86 wt% at 77 K, 1 atm although the surface area decreases when the MOF is doped.63 For the ‘‘secondary spillover’’ of H2 onto the physical surface of MOFs, close contact is required between the two materials participating in the process due to the energies involved. Using a bridge building approach, H2 is dissociated at the catalyst and is then diffused on to the MOF (receptor) with the aid of a carbon bridge (Fig. 6). One method of constructing bridges to increase the contact between the materials is illustrated by Liu and Yang.64,65 Pt / activated carbon (AC) was used as the bridge on IRMOF-8 and this resulted in an increase in gravimetric capacity from 0.5 wt% H2 to 4 wt% H2 at 10 MPa. The H2 uptake was also reversible at room temperature. HKUST-1 and MIL-101 have also been
This journal is ? The Royal Society of Chemistry 2012

Fig. 6 (a) Primary spillover of atomic hydrogen from Pt metal to the activated carbon support and a small degree of secondary spillover to the MOF (b) Facilitated primary and secondary spillover to the MOF by using carbon bridges. Reprinted with permission from Y. Li and R.T. Yang. J. Am. Chem. Soc., 2006, 128, 8136. Copyright 2006 American Chemical Society.

investigated by using bridged spillover; H2 uptake at 298 K, 10 MPa is enhanced to 0.35 wt% for HKUST-1 (a factor of 3.2 increase) and 0.51 wt% for MIL-101 (increased by a factor of 2.8).66 Increasing the Pd content in the bridge from 10% to 20% in MIL-101 resulted in an H2 uptake of 1.14 wt% at 293 K, 5 MPa compared with 0.37 wt % for the unmodi?ed MOF.67 In the same study, the H2 uptake of MIL-53 was increased to 0.63 wt% from 0 wt% H2 uptake for the unmodi?ed MOF at room temperature. The mechanism of hydrogen spillover using bridges has been studied in order to understand and improve uptake capacity and kinetics. The control of lattice defects in IRMOF-8 crystals and the pore network are important factors in improving adsorption at room temperature.68 The rate limiting step for the spillover is the surface diffusion, as shown by Stuckert et al.69 2.2.5 Catenation. Catenation of MOFs has also been investigated to examine how this phenomenon might impinge on H2 uptake and kinetics. Catenation occurs when two separate frameworks self assemble within each other. This leads to a smaller pore size which in turn could lead to an increase in heats of adsorption closer to the values predicted as necessary for room temperature storage. Ryan and co-workers investigated catenation in IRMOFs via MC simulations.70 The heat of adsoption improved at 77 K and low H2 pressure due to the smaller pore size but at higher pressures the gravimetric capacity of the IRMOFs decreased and at room temperature the pore size effect is offset by the loss of free volume. The authors summise that in light of reduced room
Energy Environ. Sci., 2012, 5, 5951–5979 | 5957

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

temperature gravimetric capacity, catenation should not be the favoured strategy towards meeting DOE targets in MOFs. Farha et al. however described controlling catenation in MOFs without sacri?cing performance71,72 and Ma et al. were able to directly investigate the effect of catenation by synthesising catenated (PCN-6) and non-catenated (PCN-60 ) isomers of a copper 4,40 ,400 -s-triazine-2,4,6-triyltribenzoate.73 They found the catenated isomer had a higher capacity at high loadings than the non-catenated isomer. Catenation may also confer other advantageous properties to the material (even if it is at the expense of ultimate capacity). In PMOF-3, a two fold interpenetrating framework is formed from 1,3-bis(3,5-dicarboxylphenylethynyl)benzene.74 Whilst the gravimetric capacity is not outstanding (3.4 wt% at 77 K, 20 bar) the stability of PMOF-3 is exceptional; the material is able to withstand re?uxing in water overnight with no loss of crystallinity. To summarise section 2.2, although major developments and advances have been made, the single largest challenge remains to elevate hydrogen release temperatures in MOFs to near-ambient values. An alternative, however, is to consider MOFs for cyroadsoption.75 Cryoadsorption (operating between 60–120 K) permits fast refuelling and reversibility using the high surface area of MOFs without compromising gravimetric capacity (excess hydrogen uptake; Fig. 7). Although this approach also reduces heat evolution, the low operating temperatures involved are ultimately unfavourable in terms of cost and ease of refuelling unless associated engineering advances can be made.

2.3 Covalent organic frameworks (COFs) Porous COFs are emerging as both very promising and interesting materials for gas storage. In COFs the organic building units are held together by strong covalent bonds between light elements, C, O, B, Si, N rather than by metal ions as found in MOFs. These materials thus exhibit relatively rigid structures, high thermal stabilities (to temperatures up to 873 K), low densities and permanent porosity and speci?c surface areas surpassing those of zeolites and porous silicates. Due to their low

density and ultra-high porosity, for example 3472 m2g?1 for COF-102 and 4210 m2g?1 for COF-103, they have been found to be very promising candidates for hydrogen storage.76 C^te et al. report a procedure for synthesis of crystalline COFs o where the one-step condensation of discrete molecules known to produce six-membered and ?ve-membered rings leads to constuction of their extended analogues.77 Uribe-Romo et al. demonstrated that the synthesis of ordered COFs is possible and that by careful selection of building components and their conditions, preselected properties and structures can be achieved.78 This was proved by synthesis of COF-300, linked by C–B, B–O, C–C and C–Si. This material was stable up to 763 K and insoluble in water and common organic solvents like hexane, methanol, acetone, tetrahydrofuran and N, N-dimethylformamide. The BET speci?c surface area of the COF-300 was 1360 m2g?1. Based on GCMC simulations with prototypical COFs, Han et al. predicted that three dimensional frameworks should be the most promising candidates for practical hydrogen storage.79 The highest predicted excess hydrogen uptakes at 77 K are 10.0 wt% for COF-105 at 80 bar and 10.0 wt% for COF-108 at 100 bar. From the total adsorption isotherms it was found that COF-108 may be able to store up to 18.9 wt% of hydrogen at 77 K and 10 MPa, followed by COF-105 with 18.3 wt%, COF-103 with 11.3 wt%, COF-102 with 10.6 wt%, COF-5 with 5.5 wt%, and COF-1 with 3.8 wt% (Fig. 8). To extend high uptake to 300 K, the authors suggest ?rst doping with electropositive elements such as Li, Na, K or Cu, Ag and Au in order to increase the H2 adsorption enthalpy and second attempting to increase both the COF surface area and free volume. Furukawa and Yaghi performed isotherm measurements with hydrogen, methane and carbon dioxide at 1–85 bar and 77–298 K on several COFs with different structural dimensions and corresponding pore sizes. They observed that 3D COF-102 and  COF-103 with medium pores (12 A) performed better than 2D  COF-1 and COF-6 with 1D small pores (9 A) and 2D COF-5,  COF-8 and COF-10 with 1D large pores (16–32 A) (Fig. 9).80 Comparing the values with high-pressure hydrogen isotherms for MOFs at the same temperature, it is apparent that COFs have the ability to outperform MOFs signi?cantly.81 In order to increase hydrogen adsorption further in COFs, a chemical modi?cation strategy with two-step doping at ambient conditions was proposed by Zou et al. using COF-1 as

Fig. 7 Excess hydrogen uptake at 77 K for selected high surface area MOFs. Reprinted with permission from ref. 75. Copyright 2011 John Wiley & Sons.

Fig. 8 Simulated hydrogen adsorption for COFs in gL?1 at 77 K. Excess H2 uptake is shown on the left and total H2 uptake on the right. Reprinted from J. Am. Chem. Soc., 2008, 130, 11580 with permission from S.S. Han, H. Furukawa, O.M. Yaghi and W.A. Goddard III. Copyright 2008 American Chemical Society.

5958 | Energy Environ. Sci., 2012, 5, 5951–5979

This journal is ? The Royal Society of Chemistry 2012

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Fig. 9 High-pressure H2 isotherms for COFs measured at 77 K. Data for BPL carbon (activated carbon) are shown for comparison. Reprinted from J. Am. Chem. Soc., 2009, 131, 8875 with permission from H. Furukawa and O.M. Yaghi. Copyright 2009 American Chemical Society.

Fig. 10 Stored H2 per mass and per volume of COFs vs. selected storage materials. Reprinted from J. Am. Chem. Soc., 2009, 131, 8875 with permission from H. Furukawa and O.M. Yaghi. Copyright 2009 American Chemical Society.

a model.82 In this model, the ?rst step involves substituting carbon with boron in the organic structure of COF-1 to form BCOF-1, the second step then involves doping metal atoms (Ca, Sc, Ti) as active centres for hydrogen adsorption. Ab inito calculations showed that boron substitution increased the binding energies of metal atoms and thus suppressed their aggregation. It was observed that Ca atoms prefer single-sided binding with respect to the organic linkers, whereas Sc and Ti favour double-sided binding. Gravimetric hydrogen storage capacity of such modi?ed fragments of B-COF-1 were estimated to be 6.7, 6.9 and 6.7 wt%, respectively for Ca, Si and Ti. It is assumed that increasing the boron ratio in the COF structure may lead to a further increase of the adsorption energies of metal atoms in boron-doped COFs and therefore increase hydrogen storage capacities. In related experimental work, Hunt et al. synthesised a new robust COF from borosilicate linkages, B–O– Si, with high thermal stability and porosity.83 It is assumed that silicon sites of the borosilicate cages may be functionalised to afford materials with a wide variety of applications. Interesting studies on integrating bonding features of MOFs and COFs were performed by Wu et al. who assumed that such a combination could offer excellent opportunities for the development of new porous materials dubbed ‘‘MOCOFs’’, in which the coordinating bonds of MOFs and covalent bonds of COF could confer unique advantages in hybrid materials.84 In the design of MOCOFs, the lightest zeolite-type tetrahedral framework, RHO, was selected to maximise the number of porous topological combinations. Boron imidazole frameworks (BIF) BIF-9-Li and BIF-Cu with the RHO topology were synthesised giving polyhedral single crystals of BIF-9-Li and BIF-9-Cu. Both structures are thermally stable up to ca. 573 K under N2. Hydrogen adsorption measurements showed that BIF-9-Li can adsorb 1.23 wt% (volumetric uptake of 9.23 kgm?3) at 77 K and 1 bar, whereas BIF-9-Cu reaches 1.06 wt%. (volumetric uptake 8.58 kgm?3) under similar conditions. As a means of evalating the prospects for COFs, Furukawa and Yaghi produced a diagram comparing COF hydrogen uptake capacities with selected hydrocarbons, MOFs and metal
This journal is ? The Royal Society of Chemistry 2012

hydrides (Fig. 10).80 Group 3 COFs demonstrate among the best performances for physical storage materials. 2.4 Polymer matrices Polymers are not a modern phenomenon. They form many of the basic and yet essential components of the natural world and their adaptation by scientists over the years has bred many materials that are now woven into day to day life. Since the intrinsic morphology of polymers and inclusion of functional components have previously been investigated for gas separation, capture and sensing, polymeric materials have inevitably, been subject to examination for use in hydrogen storage. To date research in the use of polymers as hydrogen storage materials has been relatively limited to theoretical studies. These studies will be discussed alongside available experimental data in the following section, where particular polymers and the various aspects of their nanostructures are considered in relation to storage criteria. While the performance of many of the investigated polymers thus far is perhaps unexceptional, there is clearly much scope for systematic experimental studies to understand and optimise such materials. 2.4.1 Polyacetylene. Polyacetylene, (C2H2)n, thin ?lms in the  range of 200–300 A (20–30 nm) were prepared in the early 70’s, with a focus on their use in electronics rather than hydrogen storage.85 Recent computational studies by Lee et al. however, noted a promising maximum hydrogen capacity of 14 wt% for trans-polyacetylene doped with transition metals; Sc and Ti.86 First principles calculations predicted a maximum capacity of 12 wt% for Ti-doped trans-polyacetylene, whilst an earlier combinatorial computational study by Lee et al. suggested a gravimetric capacity of 7.6 wt% in the Ti-doped cis-form of the polymer.87 The structure and conformation of the polymer and the metal dopant and hydrogen atom positions within the matrix have also been calculated (Fig. 11).88 The monomer units were de?ned as (C4H2$2TiH3)n and (C4H2$2ScH2)n for Ti and Sc respectively, which gave the most compact geometries of the doped polymers. This structure
Energy Environ. Sci., 2012, 5, 5951–5979 | 5959

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Fig. 11 Atomic structures of dihydrogen binding to a single Ti attached on cis-polyacetylene. (a) Pristine cis-polyacetylene, (b) Ti-doped cispolyacetylene and (c) 5 hydrogen molecules sorbed on Ti-doped cispolyacetylene. Green, pink, and white dots indicate carbon, titanium, and hydrogen atoms, respectively. Reprinted ?gure with permission from H. Lee, W.I. Choi, M.C. Nguyen, M.-H. Cha, E. Moon and J. Ihm, Phys. Rev. B,76,195110, 2007. Copyright 2007 by the American Physical Society.

provided suf?cent space around the polymer for hydrogen to be adsorbed. Li and Jena used DFT methods (generalised gradient approximation) to show that boron-decorated polyacetylene would not be a suitable hydrogen store since the chemically bound hydrogen would not be easily desorbed.89 2.4.2 Polyaniline (PANI). Porous polyaniline structures display evidence of promising hydrogen sorption properties. The polymer can exist in crosslinked and hypercrosslinked forms, the latter having a more complex morphology than the former.90 There are three variations of polyaniline; pernigraniline, emeraldine, and leucoemeraldine, which are formed as a result of variations in the oxidation conditions of the aniline monomer during polymerisation. PANI ?bres have been produced via electrospinning techniques by Srinivasan et al. Although these ?bres are effectively constructed on the micro- rather than the nanoscale, the authors describe remarkable reversible hydrogen storage with 3 wt% hydrogen adsorbed at 323 K and 6–8 wt% between 373–398 K (both at 80 bar), and desorption of hydrogen between 323– 398 K.91 Using scanning electron microscopy (SEM), the ?bres were shown to have swollen after treatment with hydrogen (Fig. 12) which was postulated to be a contributor to the observed high capacities. The swelling (volume expansion) in these materials is signi?cant given possible in?uences on cycleability and longevity in a storage system. Such tests have not yet been reported in the literature. Such effects, however, may be offset by cross-linking as highlighted below. Rahy et al. have produced PANI nano?bres in the presence of sucrose or sucralose and using two different oxidants; potassium iodate and ammonium peroxydisulfate.92 The authors report a reversible hydrogen storage capacity of 4.3 wt% at 298 K and 20 atm for the PANI generated with sucrose, with desorption possible at room temperature. It is postulated that nanopores in this instance have been formed by sucrose ‘‘pillars’’ that link PANI nano?bres, and it is within these pores that hydrogen is adsorbed. Studies reporting high capacities in PANI materials, however, are not without controversy. Cho et al. reported a hydrogen capacity of HCl-treated PANI in 2002 of ca. 6 wt%93
5960 | Energy Environ. Sci., 2012, 5, 5951–5979

Fig. 12 SEM images showing PANI ?bres (a) before hydrogenating, (b) after hydrogenating, indicating the swelling effect. Reprinted from Int.J. Hydrogen Energy, 2010, 35, 225. Reversible hydrogen storage in electrospun polyaniline ?bers, S. S. Srinivasan, R. Ratnadurai, M. U. Niemann, A. R. Phani, D. Y. Goswami and E. K. Stefanakos, Copyright 2010, with permission from Elsevier.

and after doubt cast over the results by Panella and co-workers,94 further material by Cho et al. was released.95 The effect of HCl addition is discussed later in section 2.4.3 with respect to polypyrrole. Hypercrosslinking of two of the PANI variations (emeraldine and leucoemeraldine) was conducted by Germain et al., with sorption results dependent on the crosslinkers employed (Fig. 13).96 The polymer was found to consist of pore less than 10 nm across and although the porous structure is seemingly not dissimilar among materials (Fig. 13b), variations in the crosslinkers and synthesis methods have signi?cant effects on gravimetric performance (Fig. 13c); diiodomethane samples outperform paraformaldehyde samples by about 0.9 wt%. When PANI was hypercrosslinked with various halogenated aryl-ring compounds it was shown that the resulting material could selectively uptake hydrogen over nitrogen.97 The maximum hydrogen uptake possible was 0.97 wt% for a polyaniline – diaminobenzene/diiodobenzene hypercrosslinked polymer with pore sizes estimated at 0.29–0.36 nm. Similar to the hypothesis advanced by Rahy et al.,92 Germain and co-workers describe the rigid bonds (pillars) formed between the polymer chains as a result of the introduction of aromatic rings within the matrix. There is a marked contrast in the gravimetric capacity, however, of these aromatically linked materials compared to Rahy et al.’s nano?bres. Further studies of these hypercrosslinked matrices to improve the kinetics and hydrogen capacity could establish a complex polymeric matrix for reversible hydrogen storage. Composite-type systems involving PANI have also been investigated. An AB3-type alloy, La0.7Mg0.25Ti0.05Ni2.975Co0.525, was mixed with varying concentrations of PANI to evaluate the effect on hydrogen storage performance of the alloy.98 Through PCT analyses, Qi et al. evaluated that absorption kinetics were enhanced with increasing quantities of PANI. However with this
This journal is ? The Royal Society of Chemistry 2012

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

material, with many adsorbing little or negligible hydrogen at 298 K irrespective of pressure and then absorbing <0.5 wt% in the case of the aluminium powder at 70 bar and 383 K. The reason for the slightly enhanced effects observed with aluminium are uncertain, but it is thought that the introduction of more nanopores with addition of CNTs is responsible in comparison to PANI alone and PANI doped with 10 wt% tin oxide. Kim et al. recently developed PANI composite work which was based on the synthesis used by Pang et al.101,102 The hydrogen storage capacity of vanadium pentoxide was compared with that of PANI and the PANI-V2O5 composite material, the latter forming thin nanosheets, as shown in Fig. 14. The structures are comparable to those synthesised by Pang and co-workers. The capacity was measured for the composite to be about 1.8 wt%, at 70 bar and 77 K, which is a signi?cant increase in comparison to studies at room temperature (0.16 wt%) and of the PANI and V2O5 alone; both 0.2 wt%. This was attributed to the reduced distance between PANI nanosheets in the composite. Research has shown that extremely interesting polyaniline morphologies can be achieved at the nanoscale by different synthesis techniques. These structures have not been tested for hydrogen capacity nor has the surface area or the pore size been quanti?ed, but these developments show potential polymer matrices for hydrogen storage, and so are worth noting. Sun and Deng showed hollow rod-like structures in their work (Fig. 15a), whereas ‘‘sea-cucumber’’ morphologies were exhibited in the work of Liu et al. (Fig. 15b).103,104 2.4.3 Polypyrrole (PPY). Polypyrrole, an electrically conducting polymer, consists of a chain of pyrrole (C4H4NH) monomer units joined at the carbons adjacent to the nitrogen within the pyrrole rings.105 It has been investigated for various applications, e.g. gas and bio- sensing, and is now emerging as a hydrogen storage matrix.106 The structure of the polymer may be designed (molecularly imprinted) for speci?c applications, meaning that it is possible for speci?c species to be encapsulated within pre-designed pores/cavities within the polymer. The publications by Cho and co-workers in 2002 and 2007 suggest substantial hydrogen storage in PPY, the results of which are shown in Fig. 16.93,95 The maximum capacity observed for PPY was better than PANI in the same studies, although the

Fig. 13 (a) SEM micrograph of hypercrosslinked PANI; (b) incremental surface areas with variation of pore sizes of hypercrosslinked PANI via the two methods used by Germain et al.; (c) hydrogen sorption pro?les of hypercrosslinked PANI at 77 K, 2 MPa H2. Reproduced from ref. 96 with permission from The Royal Society of Chemistry.

came a decrease in the overall hydrogen capacity (approximately 1.6 wt% without PANI and 1.4 wt% with PANI at 353 K). From Van’t Hoff plots it was possible to relate this behaviour to the stability of the (A,B)-H bonds, showing PANI would increase the bond stability to an optimum at 2 wt% of the polymer. Carbon nanotubes (CNTs) have also been mixed with PANI, resulting in distortion of the CNTs and interlinking by aniline chains.99 Thermogravimetric analysis (TGA) demonstrated that the thermal stability of the PANI improved in the presence of CNTs. The implications for hydrogen storage are that gas could be stored within the nanotubes themselves and also within any cavities or pores which may have formed in the CNT-PANI structure. Emeraldine PANI was investigated by Jurczyk et al. with various ?llers; multiwall CNTs (MWCNTs), tin oxide, and aluminium.100 Ostensibly, the capacities are not greatly enhanced by the additives, but it is obvious that an increase in temperature and pressure has a signi?cant effect on the capacities of each
This journal is ? The Royal Society of Chemistry 2012

Fig. 14 SEM micrograph of PANI-V2O5 nanocomposite nanosheets. Reprinted from Int. J. Hydrogen Energy, 2010, 35, 1300. Enhancement of hydrogen storage capacity in polyaniline - vanadium pentoxide nanocomposites, B. H. Kim, W. G. Hong, S. M. Lee, Y. J. Yun, H. Y. Yu, S. -Y. Oh, C. H. Kim, Y. Y. Kim and H. J. Kim. Copyright 2010, with permission from Elsevier.

Energy Environ. Sci., 2012, 5, 5951–5979 | 5961

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

1.3 and 0.69 wt% with CH2, CH and B crosslink units respectively.107 It is interesting to note that although crosslinking with B leads to the lowest hydrogen capacity, the pore size is such that the polymer becomes hydrogen selective, and negligible nitrogen uptake was observed. Although the synthesis techniques were different from that used in the PANI hypercrosslinking work above,96 it is interesting to see that similar results may be found in terms of gas selectivity, which could prove exceptionally useful in the design of membranes for puri?cation or indeed hydrogen storage, could the sorption-desorption characteristics be optimised for H2. 2.4.4 Polyurea (P-Urea) & polyamide amine (PAMAM). P-Urea has been recently highlighted by Rehim et al. as a hydrogen storage medium, along with polyamide amine (PAMAM) and a PAMAM-VOx nanocomposite.108 Maximum hydrogen adsorption capacities of 1.4 wt%, 0.9 wt% and 2 wt% (at 77 K, 20 bar) were reported for these three materials respectively, where the posibility of the polymer matrices to capture molecular hydrogen within mesopores (P-urea) or between layered structures (PAMAM-VOx) requires further investigation. Furthermore, information relating to the reversibilty of hydrogen storage within the latter matrix is required. 2.4.5 Polystyrene (PS). Davankov et al. were pioneers in the hypercrosslinking of PS,109 and recent work by Lee et al. showed how nanoporous (0.5–1.5 nm) polystyrene (Hyper Crosslinked Polymer; HCP) could hold up to 3.04 wt% hydrogen within the structure at 77 K, 15 bar.110 Figueroa-Gertenmaier and coworkers indicated how commercial PS aerogel materials may be used to store hydrogen within the nanosized cavities which are formed upon cooling of the melted PS precursor material.111 Their simulation showed a maximum capacity of 4 H2 molecules per cavity, although their experimental results showed only three molecules to have been captured, which is slightly less than 1 wt%. PS gel can be hypercrosslinked to achieve a maximum hydrogen capacity of ca. 1.5 wt% in a material with an average pore width of slightly less than 3 nm.112 If the nanopores in this study are, as indicated, slightly larger than the nanopores in the study by Lee and colleagues,110 this could indicate the sensitivity of modifying hydrogen storage capacity via pore size in PS materials. It remains to be seen if reducing the nanopore sizes below the 0.5–1.5 nm range could enhance the hydrogen storage capacity even further. 2.4.6 Polymers of intrinsic microporosity (PIMs). Polymers of intrinsic microporosity (PIMs) belong to a novel class of materials with large inner surface areas and large free volumes. They are promising for a wide spectrum of applications including heterogeneous catalysis, membrane separation, and adsorption of organic compounds and gases such as hydrogen and carbon dioxide.113,114 PIMs are synthesised by a benzodioxane formation reaction between monomers. One of the monomers must contain either a rigid non-planar unit connected to a rigid backbone or a site of contortion (for example a spiro-center, i. e. a single tetrahedral carbon atom shared by two rings). The products can be obtained as insoluble networks or as soluble polymers, suitable for solution-based purposes.114 Budd et al. have described in detail their investigations and experimental work on generating
This journal is ? The Royal Society of Chemistry 2012

Fig. 15 SEM micrographs of polyaniline showing (a) hollow rod-like and (b) sea-cucumber morphologies (scale bar shown is 1 micron in length). (a) reprinted from Eur. Polym. J., 2008, 44, 3402, Morphology studies of polyaniline lengthy nano?bers from via dimers copolymerization approach, Q. Sun and Y. Deng. Copyright 2008, with permission from Elsevier. (b) reprinted from J. Mater. Sci. Technol., 2010, 26, 39, Sea cucumber-like polyaniline nano?bers synthesized by aqueous solution method, B. Liu, L. Liu, N. Shi, J. Gong and C. Sun. Copyright 2010, with permission from Elsevier.

Fig. 16 Hydrogen adsorption pro?le of PPY; (1) adsorption of H2 at 298 K, (2) H2 adsorption after treatment at 473 K under vacuum and then cooled to 298 K, (3) second adsorption after same temperature and pressure treatment as in (2). Reprinted from Catalysis Today, 120, 336, H2 sorption in HCl-treated polyaniline and polypyrrole, S. J. Cho, K. Choo, D. P. Kim and J. W. Kim. Copyright 2007, with permission from Elsevier.

thermal treatment conditions prior to sorption measurements may have had an effect on the overall capacity differences. The effects of HCl treatment on PPY and PANI were observed via SEM showing a change in the surface morphology of the polymers, which might explain the enhanced hydrogen storage. Unfortunately, the pore size was not reported, but Cho et al. advocate that the differences in H2 uptake observed between their work and that of Panella et al.94 could be a result of the presence or lack of nanopores, respectively. Germain et al. hypercrosslinked PPY using diiodomethane, iodofom, and boron triiodide to yield hydrogen capacities of 1.6,
5962 | Energy Environ. Sci., 2012, 5, 5951–5979

View Article Online

network polymers with high surface area by incorporating several pre-formed monomer combinations with ligands: A1 + B3 (phthalocyanine), A1 + B1 (porphyrin), A1 + B2 (hexaazatrianaphthylene) among others.115–118 These ligands serve as connections to metal ions or cavities able to host organic molecules. For example, HATN-PIM (hexaazatrianaphthylene), CTC-PIM (cyclotricatechylene), Porph-PIM (porphyrin) and Trip-PIM (triptycene) are clasi?ed as insoluble network polymers, whereas the non-network polymers such as PIM-1 and PIM-7 are soluble systems. PIM-1 (polybenzodioxane) was the ?rst of the family of soluble polymers with intrinsic porosity and remains the most characterised (Fig. 17). From low temperature sorption measurements, PIM-1 was shown to have a high surface area (850 m2g?1) and microporous character (pores < 2 nm in diameter).117 Hydrogen sorption behavior on six PIMs was compared with polystyrene-based HCP.110,118 H2 sorption measurements and calculations have shown that hydrgen sorption values for HCP were below the trend set by PIMs. Analyzing hydrogen uptakes at 1–15 bar, plotted against H2-surface area and extrapolating, it was speculated that PIMs with an H2-surface area of 2400 m2g?1 would achieve 6 wt% of H2 uptake at 15 bar and 77 K. However the authors highlighted that the interpretation of Langmuirderived N2- and H2-surface areas must be made with care. Van den Berg and Arean reported that a maximum hydrogen uptake on network PIMs with incorporated triptycene was 2.71 wt% at 77 K at 15 bar.113 PIM-1 and PIM-7 are suitable for use as gas separation membranes and a study of the physicochemical properties of PIMs, such as gas permeation, thermodynamic properties and free volume were conducted on PIM-1 by Budd et al.119 Permeation measurements were performed with He, H2, O2, CO2, CH4 mixtures at 295–328 K and with pure gases at 303– 328 K. It was found that PIM-1 has a high permeability coef?cient among membrane materials as might be expected from its large inner surface area. With further exploration of possible porous systems and optimisation in PIM processing, the prospects for practical application of PIMs in hydrogen adsorption could be healthy. To conclude this section on polymers, tuning porosity and guest binding strength is clearly key to the potential for hydrogen storage. The long-term stability of the polymer itself, however, is

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

also vitally important as is an ability to control volume expansion. Performing and interpreting hydrogen cycling studies will be central to establishing whether polymeric nanostructures, both as active materials and as matrices in composite systems, have a sizeable role to play in reversible hydrogen storage solutions. Understanding the links between polymer conformation, cross linking, short and long range structure and gas sorption properties will be essential in establishing the bases for nanodesign strategies.

2.5 Clathrate hydrates Another way of storing hydrogen is by the encapsulation of the gas inside a solid structure to form a clathrate. It is believed that such materials may be viable for off-board hydrogen storage without a need for high pressures or liquid cylinders.120 Clathrate hydrates are crystalline inclusion compounds of a hydrogenbonded water host lattice and one or more types of guest molecules. They are built when water and guest molecules come into contact and hydrogen can be usually released at low temperatures and high pressures (for example, 274 K and 28 bar are required for methane).121 In clathrates a solid structure is created from a 3D lattice of hydrogen bonded water, which can trap guest molecules like hydrogen, carbon dioxide, methane or methylcyclohexane. The guest gas molecules are of different sizes, therefore solid clathrate hydrates form three common structure types: sI, sII and sH (Fig. 18). Characteristic of the sII structure is the ordered stacking of 16 small pentagonal dodecahedral (512) and 8 larger hexakaidecahedral (51264) cages (i.e. with 12 pentagonal and 4 hexagonal faces). The structure is   formed with hydrate species of ca. 4 A and ca. 6–7 A in diameter.  Structure sI is formed by molecules of 4–6 A and has 2 small (512) and 6 large (51262) cages. Structure sH forms when the largest  guest molecules (7.5–9 A diameter; such as methylcylcohexane) occupy the large pores (51268) and smaller guest molecules (such as CH4), locate in the smaller 512 and 435663 cavities.

Fig. 17 (a) Chemical structure of PIM-1. (b) Molecular model of PIM-1 showing its highly contorted, rigid structure. Reproduced from ref. 117 with permission from The Royal Society of Chemistry.

Fig. 18 The three common clathrate hydrate structures formed by combining the respective numbers of each type of cavity indicated. Reprinted from Chem. Phys Lett., 2009, 478, 97, Properties of the clathrates of hydrogen and developments in their applicability for hydrogen storage, T. A. Strobel, K. C. Hester, C. A. Koh, A. K. Sum and E. D. Sloan Jr. Copyright 2009, with permission from Elsevier.

This journal is ? The Royal Society of Chemistry 2012

Energy Environ. Sci., 2012, 5, 5951–5979 | 5963

View Article Online

Pure hydrogen hydrate has several advantages as a hydrogen storage material. For hydrogen clathrate, the only byproduct after release of the stored molecular hydrogen is water, making it attractive as a potential green and inexpensive storage material. Hydrogen is stored in molecular form and the binding energy is of a magnitude that heat generation will not be problematic. A clear disadvantage, however, to using hydrogen hydrate as a storage material is the high pressure required for its formation. The pressure could be reduced by adding a stabilizing agent such as tetrahydrofuran. The number of hydrogen molecules that can occupy the cages in the clathrate hydrate host structure is critical in determining the overall gravimetric capacity. Mao et al. described synthesis of hydrogen and water into  a binary clathrate sII type with a lattice parameter of 17.047 A 122 and an H2:H2O molar ratio of of 1 : 2. Different distributions of H2 within the large and small cavities were analyzed for this hydrogen:water ratio and it was concluded that the most probable con?guration was one in which two and four hydrogen molecules were located in the small and large cavities respectively. The resulting structure contained more than 5.0 wt% of hydrogen at 250 K and 2 kbar. Sluiter et al. performed thermodynamic simulations to determine the most thermodynamically favored cage occupancy and calculated con?gurations in agreement with experiment,123 while Florusse et al. observed that hydrogen clusters can be stabilized and stored at low pressures in the sII binary structure.124 Investigations with tetrahydrofuran, THF, as a second guest molecule to promote clathrate formation showed that THF stabilises the clathrate at 279.6 K and 50 bar. This is interesting when compared with values for pure hydrogen clathrate of 280 K and 3 kbar (300 MPa). The hydrogen uptake was dependent on the THF concentration and reached a maximum at 4 wt%. A detailed description of the THF + H2 hydrates is presented by Strobel et al., but when the THF concentration decreased from 5.56 to 0.5 mol %, there was no signi?cant change in storage capacity with a constant formation pressure of 138 bar.121,125,126 Over the range of experimental conditions tested (Fig. 19), the maximum hydrogen storage capacity of the binary sII THF-H2 hydrate was found to be approximately 1.0 wt% at moderate pressure (<600 bar); the stoichiometric THF binary hydrate can only store a maximum of 1.05 wt% H2. These values clearly limit the use of THF-H2 hydrate as a hydrogen storage material.125 Recently Sugahara et al. have proved that acetone is superior as a promoter for clathrate formation compared to THF or cyclohexanone (CHONE).127 Four hydrates: sII-type THF + H2, sII-type CHONE + H2, sII-type acetone + H2 and sH-type methylcyclohexane MCH + H2 were investigated using Raman spectroscopy with respect to storage pressure and promoter concentration. Raman spectra showed that hydrogen occupied the largest (hexakaidecahedral, 51264) cages in THF + H2, CHONE + H2 and acetone + H2 clathrates, but did not occupy the largest (51268) cages in MCH + H2 clathrate, irrespective of the MCH mole fraction. The amount of stored H2 in sII-type clathrates can be moderated by promoter concentration and H2 pressure in THF and acetone hydrates. Comparing the pressure effect of the H2 (51264) cage occupancy in THF + H2 and acetone + H2 clathrates, the required pressures for the H2 clusters in the acetone + H2 hydrate are lower than those in the
5964 | Energy Environ. Sci., 2012, 5, 5951–5979

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Fig. 19 Hydrogen storage capacity of sII THF + H2 hydrates as a function of formation pressure for various THF concentrations using different experimental techniques. pVT - thermodynamic gas release or gas consumption measurement; NMR - calculated from integral 1H NMR intensities; NS - neutron scattering; INS - inelastic neutron scattering measurement; grav.- gravimetric measurement. Reprinted from Chem. Phys Lett., 478, 97, Properties of the clathrates of hydrogen and developments in their applicability for hydrogen storage, T. A. Strobel, K. C. Hester, C. A. Koh, A. K. Sum and E. D. Sloan Jr. Copyright 2009, with permission from Elsevier.

THF + H2 hydrate. Hence, when comparing the occupancy of 51264 cavities it is evident that acetone is a better promoter than THF. Another barrier to practical application of H2 clathrates is the slow kinetics associated with enclathration. Su et al. demonstrated a method for improving kinetics and reusability of gas clathrates, using polymerised phase emulsion, polyHIPE, as a support.128 Experiments were performed on H2-THF-H2O clathrates, usually needing several days for formation.122 Kinetic plots for H2 enclathration in preformed THF-H2O hydrate (with 5. 56 mol. % THF) at 270 K both with and without polyHIPE support were compared and bulk THF-H2O, with no polyHIPE required more than 15000 min. (>11 days) to reach 90% of the potential H2 enclathration capacity (assuming a max. pressure of 10 bar). In the presence of the emulsion, 90% of the max. pressure was reached after 1 h, thus representing more than a 250-fold kinetic enhancement. Similar observations were made at higher initial pressures (>138 bar). Volumetric release experiments where H2 clathrate was stabilized showed H2 storage capacities of 0.4–0.5 wt%. The size and the structure of clathrate hydrates determine hydrogen storage capacity, as recently reported by Chattaraj et al.129 DFT methods were employed to determine how many H2 molecules could be stored in each clathrate cavity. It was revealed that clathrate hydrate cages sI (512) and (51262) could accommodate up to two hydrogen molecules and sH (51268) could accommodate up to six molecules of H2, depending on the exact size and shapes of the cavities. The stability of the clathrates was
This journal is ? The Royal Society of Chemistry 2012

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

enhanced and in most cases the number of hydrogen molecules in the cage structure increased. Binary, ternary and more complex structures like semiclathrates are discussed in detail by Struzhkin et al. where they present an overview of theoretical and practical approaches.121,130,131 Strobel et al. summarized hydrogen storage capacities (gravimetric and volumetric) of different clathrate hydrates, based on numerous literature data, with respect to US DOE targets (Table 3). It can be seen that there are a variety of challenges for clathrate hydrates to overcome before they can compete to reach the targets set by the US DOE.

3. ‘‘Intermediate’’ hydrogen storage
This section considers materials which might be located between classic physical hydrogen storage systems and chemical storage systems and comprise various nanostructured inorganic compounds.

3.1 Nanostructured boron nitride The application of boron nitride in hydrogen storage was ?rst reported by Wang et al. who prepared nanostructured h-BN by ball milling micron sized powders under hydrogen pressure.132 The crystallite size after 80 h was $3 nm. Combustion analysis of the product showed a 2.6 wt% hydrogen concentration. Soon after this, Ma et al. reported the ?rst measurements of hydrogen unptake in boron nitride nanotubes (BNNTs).133 CVD processes produced two different types of nanotubes from the pyrolysis of different B–N–O precursors. Multiwall nanotubes were made from a higher oxygen content starting material (Fig. 20a) whilst a lower oxygen content yielded bamboo nanotubes (Fig. 20b). The highest hydrogen uptake was seen in the bamboo nanotubes at 2.6 wt%, with the multiwall tubes taking up 1.8 wt% (bulk BN absorbed 0.2 wt%) at room temperature between 1–100 bar. The enhanced capacity in the nanotubes was explained by their increased surface area compared to the bulk. The multiwall tubes have less capacity as they are usually capped at both ends, preventing access to the inner surface. When subsequently returned to ambient pressure, around 70% of the adsorbed hydrogen is retained, suggesting a chemisorption process rather than a physisorption one. Complete release is achieved by heating to 573 K. Ma’s group subsequently synthesised BN nano?bres from a mixture of B2O2, BN and B, pyrolysed under N2 at 2000 K.134 These ?bres have widths of 30–100 nm, lengths of several microns, and are of solid cross section. EELS con?rmed that the product was pure BN and transmission electron micropscopy (TEM) showed that they possessed rough outer surfaces. H2 absorption capacity is slightly higher than in bamboo BN
Fig. 20 The morphologies of BN nanotubes produced by Ma et al.: (a) multiwall nanotubes and (b) bamboo-like nanotubes. Scale bar: 100 nm. Reprinted with permission from reference J. Am. Chem. Soc., 2002, 124, 7672. Copyright 2002 American Chemical Society.

nanotubes at 2.9 wt% under the same conditions (298 K, 1– 100 bar). Tang et al. produced a collapsed nanotube morphology whilst studying defects in BNNTs.135 TEM shows that the inner areas of the tube are intact, but the outer surfaces are cracked and have hair-like protuberences projecting outwards. The collapse leads to a surface area increase from 254.2 to 789.1 m2 g?1. Gravimetric measurements under 20–110 bar H2 at 298 K revealed maximum capacity for both pristine and collapsed BNNTs was reached at ca. 60 bar. The pristine tubes absorbed 0.9 wt% of hydrogen whereas the collapsed tubes absorbed 4.2 wt%. Chemisorption is again indicated in this system, as physisorption mechanisms could not account for such values at room temperature. Higher temperatures were needed to fully release the hydrogen (723 K). Subsequently Oku and Kuno were able to access an array of different BN morphologies by varying reagents and ratios, producing nanocapsules, nanotubes and nanocages.136,137 H2 uptake, measured by TGA/DTA was highest in samples synthesised from 1 : 6 LaB6:B under an Ar / N2 mixture (which contained a mixture of capsules and tubes, but no cages). A 3.2 wt% gain was observed between 298–573 K. Terao et al. found that including ZnS in BN synthesis generated SO2 in situ which leads to enhanced etching of the BNNT surfaces.138 When B, MgO, SnO and ZnS were reacted with

Table 3 Hydrogen storage parameters of common clathrates compared with US DOE goals121 sII H2 only Gravimetric capacity / wt Volumetric capacity / g L?1 Formation pressure at 270 K / MPa 3.80 31.33 $200 sII (THF + H2) 1.05 10.44 $70 sI (EO + H2) 0.37 3.84 $70 sH (MTBE + H2) 1.42 13.01 $100 DOE 2010 6.0 45.0 $10.1

This journal is ? The Royal Society of Chemistry 2012

Energy Environ. Sci., 2012, 5, 5951–5979 | 5965

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

ammonia at 1723 K, the resulting BNNTs have a complex surface (Fig. 21). The BNNTs were puri?ed by heating under Ar at 1723–1823 K. Hydrogen uptake was measured at 77 K from 1– 30 bar and was found to be $1.2 wt%. Reddy et al. prepared nanostructured BN using a catalysed CVD process with materials following a VLS growth mechanism.139 Fe2O3 catalyst milled with amorphous B and SiO2 powder was reacted with ammonia and at 1273 K ?ower-like nanostructures were formed. Increasing to 1373 K produced bamboo structures, whereas by 1423 K smooth walled BNNTs were obtained. Hydrogen storage capacity was measured from 1– 100 bar at room temperature and was highest in the bamboo structures (3.0 wt%). The capacity increase over the other nanostructures is attributed to surface area enhancement. Theoretical and computational studies have been performed to complement and understand experimental data. Narita and Oku used MD calculations to compare C60 and B36N36 fullerene cages.140 Calculations showed that H2 enters the cage more easily through hexagonal rings than tetragonal rings and also that BN materials would store H2 more easily than carbon. Wu et al. have used DFT to study chemisorption of H atoms in (8,0) zigzag BNNTs.141 They found that the preferential way for H atoms to absorb is on adjacent B and N atoms forming an armchair chain along the tube axis. They also found an even-odd oscillation behaviour in the adsorption energies with the average adsorption energies of even H atoms greater than that of odd H atoms. The same group also used DFT calculations to study the effect of radial deformations of BNNTs on the chemical adsorption of H atoms.142 Calculations showed that the adsorption is exothermic, and that it can be modi?ed by radial deformations. When the deformation is relatively small, the H atom prefers to adsorb on the B atom, creating an acceptor state in the gap. On the other hand, with larger deformations of the tube, adsorption

on the N site is preferred creating a donor state. DFT was also used to study binding energies of physisorbed molecular H2 in BNNTs.143 The binding energy is increased by up to 40% compared to carbon nanotubes (CNTs). This is attributed to the heteropolar bonding in BN. Modi?cation of the sp2 bonding by substitution in the structure also increases the binding energy. Further DFT studies of the effect of carbon substitution in BN showed that for clean BNNTs, H2 could adsorb on top of a boron, a nitrogen or above the centre of a hexagonal ring, with the N site favoured energetically of the three.144 In carbon doped systems, the carbon can substitute at either the boron site (CB) or the nitrogen site (CN). Calculations show almost no difference in energy between a H2 molecule adsorbed on a CB or CN site. There are however differences in C–H and H–H distances between the two adsorption sites, demonstrating that the adsorption process must be different. Zhang et al. have performed DFT calculations on the effect of metal incorporation into BNNTs.145 The results show that rhodium, nickel and palladium will all chemisorb to the outer surface of the tube at the axial bridge site. The metal incorporation forms sites where H2 molecule binding is favoured, enhancing storage capacity. GCMC studies of physisorption of H2 in single walled BNNTs and CNTs show that when the diameter of both tube types is equal, the BNNT will have a higher capacity.146,147 Moreover BNNTs were predicted to exceed a 6 wt% H2 target at room temperature and 10 MPa (15 MPa) if the tube diameter is $30 nm ($25 nm). Mpourmpakis and Froudakis also concluded from DFT that increasing the diameter of the tube would increase the ef?ciency of the binding energies due to decreased curvature of the wall.148 From theoretical studies of defects in BNNTs where vacancies were introduced into the nanotubes at either the B or N sites, it was demonstrated that the vacancies reconstruct, forming a pentagon ring and a dangling bond.149 New bonds formed are either B–B or N–N and are less stable than B–N bonds. Subsequent reaction with hydrogen should result in bond formation between B and N to give more stable structures. These B–H and N–H defects provide energetically favourable sites for further H2 molecules to pass through the tube wall for storage. 3.2 Chalcogenide and oxide nanostructures Many different nanostructures have been unveiled for group 16 compounds, including nanocoils, nanotubes, fullerene-type structures, etc, with certain analogues to carbon.150–153 MoS2 and TiS2, for example, and a small number of oxides have been investigated as hydrogen stores, but their ab/adsoprtion mechanisms are not clear, as will be discussed.154 3.2.1 Molybdenum disul?de. Chen et al. have synthesised MoS2 as nanocrystals and nanotubes.155,156 The surface area of polycrystalline bulk material was signi?cantly smaller (3.6 m2g?1) in comparison to the non-treated nanotubes (up to 58 m2g?1) and KOH treated nanotubes (up to 66 m2g?1). MoS2 with the highest surface area has the highest hydrogen sorption capacity (Fig. 22). It was proposed that the introduction of defects by the alkali treatment of the nanotubes (observed by SEM), were responsible for the slightly higher surface area and hence enhanced storage performance, in comparison to the non-treated nanotubes.
This journal is ? The Royal Society of Chemistry 2012

Fig. 21 (a) Low-magni?cation TEM image of the crude BNNT surface, (b) SEM image, and (c) HRTEM image of BNNTs; the inset in (c) is an electron diffraction pattern taken from the tube surface. Reprinted from Physica E, 2007, 40, 2551. Effective synthesis of surface-modi?ed boron nitride nanotubes and related nanostructures and their hydrogen uptake. T. Terao, Y. Bando, M. Mitome, K. Kurashima, C. Y. Zhi, C. C. Tang and D. Goldberg. Copyright 2007, with permission from Elsevier.

5966 | Energy Environ. Sci., 2012, 5, 5951–5979

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

the nanotubes uptake.159–161 The mechanism for the bonding of hydrogen in the nanotubes is not well-elucidated, but as with MoS2, it was suggested that either a Ti–H or S–H bond could form, with the latter being the most likely. The electronic structure and defects of TiS2 nanostructures have been investigated theoretically, which show that different structural defects can dictate the intrinsic morphology of the nanotubes.162–164 Computational studies may hold the key as to the role that such defects play in binding hydrogen. 3.2.3 Titanium dioxide. Titanium dioxide already plays an important role in many applications and developments in the synthesis of TiO2 nanostructures over the past decade have broadened the scope and prospects for its use.165–167 Nanotubular arrays (tube diameter, 30–100 nm) of TiO2 can electrochemically store very small amounts of hydrogen ($0.02 wt%), where defects in the nanotubes have a pronounced effect on the hydrogen storage capacity; the higher the defect density, the lower the hydrogen capacity.168 This is contrary to the effects observed for light metal hydrides (section 4.1), where milling has been used to introduce defects into the surface morphology of the metals. Recent DFT calculations performed for single-walled TiO2 nanotubes demonstrated that the binding energy for the zigzag conformation (Fig. 24) was 0.053 eV per molecule of H2. The hydrogen storage capacity was calculated to be 3.2 wt%.169 Since TiO2 is cheap and easily sourced, it could show promise as a hydrogen storage material. However, further experimental investigations are required to determine whether it could achieve capacities that could be used in real applications. 3.2.4 Zinc oxide. Wan et al. have evaluated the hydrogen storage capacity of zinc oxide nanowires (Fig. 25a) and found that the 0.83 wt% H2 that was stored within the structures could only be partially desorbed (Fig. 25b), but no hydrogen sorption mechanisms were reported to suggest a reason for this.170 A more recent study by Pan and co-workers has shown a higher uptake of hydrogen (2.57 wt%, and 2.75 wt% with Mg dopant), but similarly, the hydrogen could not be fully desorbed.171 Alternative synthesis methods have generated plate like ZnO nanostructures.172,173 The hydrogen storage capacity of these platelets were compared with Kadox ZnO (formed by combustion of Zn in the presence of oxygen). The larger surface area of the former enabled a greater degree of hydrogen dissociation in comparison to the latter, but the authors propose that steric

Fig. 22 Adsorption pro?les showing the enhanced performance of MoS2 nanotubes over polycrystalline materials at 298 K. Reprinted from J. Alloys Compd., 2003, 356–357, 413, Novel hydrogen storage properties of MoS2 nanotubes, J. Chen, S. L. Li and Z. L. Tao. Copyright 2003, with permission from Elsevier.

Variation in H2 adsorption performance may be attributed to a lack of available binding sites, and/or the inaccessible nature of these sites. If a suitable poragen could be found to increase the surface area of high purity material then accessibility to bonding sites could be optimised, and H2 capacity increased. Spirko et al. used DFT to propose the bonding mechanisms that occur between hydrogen and molybdenum disul?de. Stable geometries of the hydrided MoS2 unit are shown in Fig. 23.157 Their calculations show that H2 could either be bound to the Mo or S depending upon the size and surface characteristics of the disul?de material; smaller clusters favouring the Mo–H bond and S–H bonding predominating for larger clusters. DFT has also shown that pore formation is possible in the walls of MoS2 nanotubes where defects are present. This may also enhance their capability in uptaking hydrogen, although this is not speci?cally implied in the study by Enyashin and Ivanovskii.158 This hypothesis would seem reasonable since the introduction of pores/defects after treatment of nanotubes with KOH by Chen et al. had this effect. 3.2.2 Titanium disul?de. Titanium disul?de nanotubes can store 2.5 wt% of hydrogen at 298 K, 4 MPa, where a further increase in temperature diminishes the amount of hydrogen that

Fig. 23 (a) Ground state geometry of H2-MoS2, (b) and (c) two metastable geometries of H2-MoS2 proposed by Spirko et al. Grey ? hydrogen, red ? molybdenum, yellow ? sulfur. Reprinted from Surf. Sci., 2004, 572, 191, Electronic structure and reactivity of defect MoS2 II. Bonding and activation of hydrogen on surface defect sites and clusters, J. A. Spirko, M. L. Neiman, A. M. Oelker and K. Klier. Copyright 2004, with permission from Elsevier.

Fig. 24 Calculated zigzag TiO2 nanotube structure. Reprinted from Physica E, 2009, 41, 838, Structures, electronic properties, and hydrogenstorage capacity of single-walled TiO2 nanotubes, J. Wang, L. Wang, L. Ma, J. Zhao, B. Wang and G. Wang. Copyright 2009, with permission from Elsevier.

This journal is ? The Royal Society of Chemistry 2012

Energy Environ. Sci., 2012, 5, 5951–5979 | 5967

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

storage, with an inevitable decrease in gravimetric hydrogen capacity as one descends group 1 (12.68 wt% and 0.75 wt% for LiH and CsH respectively) or group 2 (18.28 wt% and 1.44 wt% for BeH2 and BaH2 respectively) and a relative increase in gravimetric capacity from group 1 to group 2. As will be discussed in later sections, some of these metal hydrides have been complexed with group 13 elements, e.g., aluminium and boron, to form ternary hydrides (complex hydrides) that have higher theoretical hydrogen capacities than the light metal (binary) hydrides themselves. Light metal hydride storage is dominated by Mg/MgH2, which with perhaps the exception of Li/LiH is the only system that has merited signi?cant research from a nanodesign viewpoint. We focus here particularly on magnesium hydride processed or synthesised at the nanoscale and investigate the effects of particle size reduction and shape control on hydrogen storage perfomance. 4.1.1 Lithium hydride. Ortiz et al. established that increased milling times in argon produced porous nanoscale agglomerates, which had either a smooth, round structure or a faceted structure, which were stable up to 558 K.175,176 After ca. 180 min of milling the average crystal sizes were about 25 nm in diameter. Extensive studies of the effect of nanostructuring on Li–H uptake and release are yet to be performed. 4.1.2 Magnesium hydride. In the early 1980s, pure magnesium powder was hydrided to investigate whether smaller particle sizes, and hence larger surface areas may enhance the performance of magnesium as a hydrogen store.177 The results published by Vigeholm et al. showed that when commercially sourced magnesium powders below 100 mm in size were hydrided, the metal was converted entirely to the metal hydride. After further studies into the effects of hydrogen cycling in small particles (<75 microns) of pure magnesium, interesting structural changes were revealed.178,179 Magnesium ‘‘whiskers’’ with a diameter of 500 nm were identi?ed; a phenomenon that had not been observed in previous studies. After an increased number of hydrogen cycles further structural changes in the magnesium were visible; as the number of cycles increased, a signi?cant agglomeration of the particles was observed, and this led to a decrease in adsorption-desorption of hydrogen by the metal.180 Theoretical studies have now shown that Mg nanoparticles must be reduced to much less than 20 nm to be practical for reversible hydrogen storage.181 4.1.2.1 Milling. The main parameters for ball milling of magnesium hydride are milling time, ball:powder mass ratio, milling atmosphere, i.e., pressures of hydrogen or argon, milling aparatus, and mill rotation speed.182 The milling rotation speed is often not stated in the literature and as a result it is dif?cult to optimise rotation speed, and hence energy, which has implications for the consistency of results from milling. There are numerous studies in the literature describing the processing of MgH2 via ball milling, we highlight brie?y here some key points and focus more in the following subsections on chemical methods to design hydride nanostructures. Investigations into the structure of milled magnesium hydride have been conducted to establish a correlation between the particle size and hydrogen sorption-desorption pro?les.
This journal is ? The Royal Society of Chemistry 2012

Fig. 25 (a) ZnO nanowires formed in a tube furnace and (b) the partially reversible hydrogen sorption pro?le of the nanowires. Reprinted with permission from Q. Wan, C. L. Lin, X. B. Yu and T. H. Wang, Appl. Phys. Lett., 2004, 84, 124. Copyright 2004, American Institute of Physics.

hinderance and repulsive forces are introduced by the formation of Zn–OH bonds in addition to Zn–H bonds. The maximum H2 uptake is apparently quashed in the platelets, as observed by a comparison of the FTIR spectra of the hydrided ZnO samples. To conclude this section on intermediate storage, it is already evident from the relatively small number of examples that appreciable gravimetric hydrogen storage capacities can be achieved under relatively mild conditions. These capacities, however, do not yet match those of porous solids at sub-ambient temperatures nor those of the best hydrides at elevated temperature (see below) and it remains to be seen whether such systems can retain their performance over an extended cycle life.

4. Chemical storage
In chemical storage materials, the hydrogen is chemically bound in a compound. Therefore, the loading and unloading of hydrogen (uptake-release process) involves a chemisorption step prior to the absorption of hydrogen in the bulk and the formation of chemical bonds. Therefore, in the chemical storage scenario hydrogen is relatively strongly bound and many of the challenges in utilising such materials centre on the thermodynamics and kinetics of dehydrogenation. The nanostructuring of chemical storage materials may be one approach (perhaps in combination with catalysis) that enables such barriers to be overcome.174 In this section we highlight important developments at the nanoscale in light metal, complex and so called ‘‘chemical’’ hydrides and the implications for understanding and potentially exploiting such nanometric phenomena. 4.1 Light metal hydrides The theoretical hydrogen storage capacities of the alkali and alkaline earth metals hydrides are promising for hydrogen
5968 | Energy Environ. Sci., 2012, 5, 5951–5979

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Hanada et al. reported that optimisation of the milling time is important to establish effective hydrogen capacities and kinetic properties, since the lattice strain and crystallite size must be ?nely balanced to achieve the best hydrogen cycling properties.183 The milled magnesium hydride nanoparticles formed by Huot and co-workers were shown via powder X-ray diffraction (XRD) to be comprised of two different phases of the hydride; b-phase and g-phase.184,185 Further practical and theoretical studies into the signi?cance of the phase changes before and after sorptiondesorption studies were conducted, which con?rmed that reduction of the particle size during milling was predominantly responsible for the improved kinetics rather than presence of polymorphs or changes in defect densities.186 4.1.2.2 Vapour deposition. Chemical approaches to producing nanostructured metal hydrides are emerging as an alternative to conventional milling techniques. Given that particle size distribution (and shape) is dif?cult to control in milling, such chemical routes, in principle, offer prescription in design. Vapour deposition is one such chemical method and was employed by Li et al., to grow MgH2 nanowires (30–50 nm in diameter).187 Upon cycling with hydrogen, the thinnest nanowires performed best, with higher capacities and faster sorptiondesorption kinetics (Fig. 26). Theoretical studies using DFT have been conducted to provide reasoning behind this observation, which show that a destabilisation effect is seen upon reduction in size of the nanowires, and it is this destabilisation that enables enhanced hydrogen cycling capability in the smallest nanowires.188 Crucially, after 50 hydrogen cycles the kinetic performance of the material was unaffected, and nanowires transformed to nanoparticles. Zhu, Hayashi and Saita recently reported synthesis of nano?bres of MgH2 (Fig. 27) also by a vapour deposition method, but hydrogen was used as the (reactive) vapour carrier gas rather than argon, as in the work of Li et al.189,190 Their study highlighted the effects of hydrogen pressure on the morphology of the hydride; higher hydrogen pressures (4 MPa) gave more regular, needle-like ?bres, whereas lower hydrogen pressures (1 MPa) resulted in randomly curved strands. Further to this, a comprehensive study to investigate hydrogen pressure, temperature and deposition duration as a means of understanding the optimum conditions for production of nanowires via CVD was performed with the outcomes shown in Fig. 28.191,192 Recent investigation of the structural changes on transforming from magnesium hydride nano?bres to magnesium nano?bres and vice versa used an electron beam to decompose the hydrided metal nanowires and TEM to investigate the effects, the results of which are shown schematically in Fig. 29.193 No further work has been published to demonstrate how the nano?bres produced can be reversibly hydrided and so the optimum conditions for hydrogen storage in these materials are still not apparent. A very different structure of Mg nanoparticles can be produced by sputtering magnesium on to a substrate comprised of silicon nitride.194 TEM results showed hexagonal crystals of Mg with an MgO layer at the particle surface. The diffusion of oxygen to form the oxide layer was said to play an important role in de?ning the resultant structure of the magnesium held therin. Voids were seen where the diffusion of the magnesium out of the particle exceeded the diffusion of oxygen in, but the electron
This journal is ? The Royal Society of Chemistry 2012

Fig. 26 Adsorption and desorption pro?les of MgH2 nanowires at three different temperatures; 573, 473, 373 K, (blue, red, and black respectively). Reprinted with permission from W. Li, C. Li, H. Ma and J. Chen, J. Am. Chem. Soc., 2007, 129, 6710. Copyright 2007 American Chemical Society.

beam was noted to enhance this effect. Recent results have since shown that these voids are still generated when the nanoparticles are hydrided.195 These interesting results show that there are still many facets to the Mg–H system, including the role of MgO, that must be established to assess its suitablility for practical application.

Fig. 27 MgH2 nanowires produced from magnesium under 4MPa hydrogen. Reprinted from Int. J. Hydrogen Energy, 2009, 34, 7283–7290. Direct synthesis of MgH2 nano?bers at different hydrogen pressures. C. Zhu, H. Hayashi, I. Saita, T. Akiyama. Copyright 2009, with permission from Elsevier.

Energy Environ. Sci., 2012, 5, 5951–5979 | 5969

View Article Online

vapour deposition methods, both of which require specialised equipment. However, the average particle diameter is signi?cantly greater than optimum (ca. 20 nm) when compared to those made by vapour deposition. Synthesis of smaller, more monodisperse nanoparticles of Mg via this method may allow for a better comparison in performance against other synthesis routes.
Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Fig. 28 Guideline for optimal Pressure-Temperature conditions for synthesis of MgH2 micro- and nanowires over 7 h. Reprinted with permission from C. Zhu, S. Hosokai, I. Matsumoto and T. Akiyama, Cryst. Growth Des., 2010, 10, 5123. Copyright 2010 American Chemical Society.

4.1.2.5. Electrochemistry. Electrochemical methods have been used to generate magnesium nanoparticles for hydrogen storage by Aguey-Zinsou and Ares-Fernndez, where 1.34 wt% a hydrogen was stored in the resulting Mg colloid nanoparticles by hydriding at 313 K, 2 MPa. The hydrogen could be desorbed at 358 K.199 The anode and cathode used in the electrolysis were Mg ribbon and a Pt/Rd gauze respectively, which were placed in a magnesium acetate electrolyte and surfactant (tetrabutylammonium bromide, TBA) solution. This formed spheres of the Mg colloid about 5 nm in size. Since the surfactant makes up a proportion of the total mass the gravimetric capacity is relatively small. However, if the surfactant could be removed without affecting the Mg morphology, or if a lower molecular weight surfactant could be found with similar properties to TBA, then this method of magnesium nanoparticle generation could promise much, even on a larger scale. 4.1.2.6. Solvated metal atom dispersion. Kalidindi and Jagirdar generated Mg nanoparticles protected by an organic layer via a solvated metal atom dispersion technique, which is an established method of catalyst formation.200,201 Digestive ripening was adopted to modify the particle size, and it was imperative to select a solvent that would favour smaller particle formation by: (i) maximising the interaction sites between Mg and the solvent, since high surface areas of the metal are formed, and (ii) preventing larger particle formation as a result of steric factors. The colloids were found to be as small as 2 nm in diameter and the Mg contained within the colloid was separated easily using a centrifuge. The magnesium nanoparticles absorb hydrogen at 33 bar, 391 K and desorb at 378 K. This is signi?cant, since this represents a ca. 25% reduction in desorption temperature over bulk MgH2.202 It remains to be seen how these nanoparticles perform over multiple cycles. 4.1.2.7 Nano-additives & composites. Nanoscale additives, including transition metals and transition metal oxides have produced H2 sorption-desorption results in the Mg–H system outperforming nano-MgH2 alone.203 Oelerich et al. showed that even a small incorporation of the oxides tested; Sc2O3 TiO2, V2O5, Cr2O3, Mn2O3, Fe3O4, CuO, Al2O3, with the exception of SiO2, were capable of producing a positive effect on the

Fig. 29 Changes in the unit cell of MgH2 in the form of nanowires upon removal of hydrogen. Reprinted from Int. J. Hydrogen Energy, 2011, 36, 3600. In situ transmission electron microscopy observation of the decomposition of MgH2 nano?ber, C. Zhu, N. Sakaguchi, S. Hosokai, S. Watanabe and T. Akiyama. Copyright 2011, with permission from Elsevier.

4.1.2.3. Laser ablation. Laser ablation has been investigated for the synthesis of Mg nanoparticles, where ablation was conducted in a cell containing either acetone or 2-propanol.196,197 The products were in a size range of 15–100 nm and their structures were non-uniform, with both Mg and MgO being identi?able via XRD. Phuoc et al. demonstrated that different solvents may produce different results, i.e. ablation in de-ionised water, by comparison to that in acetone and 2-propanol, produced ?bre shaped products that comprised Mg(OH)2, but there was no evidence of pure magnesium. No results have yet been published on how the magnesium nanoparticles formed via this method perform under hydrogen atmospheres. 4.1.2.4 Salt reduction. Song, Chen and Zhang reported a simple synthesis of magnesium nanoparticles (ca. 300 nm in diameter) by reducing a magnesium salt in tetrahydrofuran (THF) using lithium as the reducing agent and naphthalene as an electron carrier (Scheme I).198 Song et al. did not include hydrogen sorption-desorption testing in their experimental work, but commented on the simplicity of their technique in comparison to ball milling and
5970 | Energy Environ. Sci., 2012, 5, 5951–5979

Scheme 1 Salt reduction synthesis of Mg nanoparticles. Reprinted from Mater. Charact., 2008, 59, 514. Preparation and characterization of Mg nanoparticles, M. -R. Song, M. Chen, and Z. -J. Zhang. Copyright 2008, with permission from Elsevier.

This journal is ? The Royal Society of Chemistry 2012

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

sorption-desorption pro?les.204 Chromium oxide had the most pronounced effect on the adsorption pro?le, and iron and vanadium oxides were most ef?cient in increasing desorption rate. It is not yet obvious why both the adsorption pro?les and desorption pro?les were not enhanced in each case. Several studies have sought to establish how the size of nanoadditives, such as transition metals, affects hydride storage performance.205,206 The dispersion of these nanomaterials throughout the hydrogen storage material by, e.g., sputtering or milling, etc., is said to have a catalytic effect, and also promote nucleation. Studies with nickel and niobium oxide nanoparticle catalysts (both established previously as enhancing storage in MgH2) showed that Ni nanoparticles were observed on the surface of the magnesium, whereas the niobium oxide was dispersed throughout the metal.207,208 In both cases, the catalyst remained on/in the hydride after dehydrogenation under vacuum at 473 K / 8 h, which is promising since it would imply that retention of catalytic activity was possible over the lifetime of the hydride. Unfortunately, the structural integrity of the hydride was not maintained after dehydrogenation, and repeated cycles were not conducted to test this hypothesis. 4.1.2.8 Carbon scaffolding. Nanoporous materials have been used as scaffolds (or nano-templates) for synthesis of nanoscale light metal hydrides, where the conformation and size of the hydride may be constrained by the structure of the support.209 The pore size may be speci?ed for a particlar application, and depends on the synthesis conditions of the porous material.210 In?ltration of carbon pores with molten magnesium has been conducted, and also by using dibutylmagnesium and subsequent treatment with hydrogen to produce the magnesium hydride.211,212 The effects of these spatial restrictions on the hydrogen storage capacity of the hydrides have been compared and although they show signi?cantly improved hydrogen cycling characteristics when compared to ball milled MgH2, the mechanisms and size effects have not yet been investigated fully. One of the main disadvantages of the smaller particle sizes formed by nanocasting in porous materials was stated by Vajo;210 the detrimental effects of oxide formation on storage reversibility may be greater owing to a larger surface area and hence more sites at which an oxide layer may develop and grow. However, promising results have been published by Jeon et al., showing how the nano con?nement of Mg is possible within a gas permeable, porous polymer matrix (see also section 2.4).213 This new departure in nanomaterials design for hydrogen storage indicates that there are many avenues yet to be explored in terms of hydride composites (see also section 2.4). 4.2 Complex and ‘‘chemical’’ hydrides Aluminium and boron, like magnesium, are light and cheap, and ternary hydrides formed with group 1 or 2 metals, e.g., LiAlH4 and LiBH4, are appealing in principle since hydrogen capacity could be increased over light metal hydrides. (It should be noted that the binary group 13 hydrides such as borane and aluminium hydride have been largely discarded as storage materials mainly owing to their toxic and pyrophoric natures respectively. Further they are limited by their poor ability to store hydrogen reversibly.214)
This journal is ? The Royal Society of Chemistry 2012

4.2.1 Alanates. XRD analysis has shown that it is possible to use ball milling to prepare alanate nanoparticles.215,216 Milling of a group 2 metal halide, e.g., MgCl2, or CaCl2, with an alanate, e.g., LiAlH4 and NaAlH4, generates nanoscale particles of the respective group 2 alanates, i.e., Mg(AlH4)2, Ca(AlH4)2.217 This study provided information on the thermal decomposition of nano- Mg(AlH4)2, Ca(AlH4)2, and LiMg(AlH4)2, showing maximum hydrogen desorption values of up to 5 wt% for magnesium and calcium alanates, and 4 wt% for lithium magnesium alanate. Additives, e.g. TiCl3, have been milled with sodium alanate to attempt to improve the kinetic and thermodynamic hydrogen storage properties.218,219 Nanoparticles of Ti-containing additives have also been investigated via milling to establish whether there is an effect on the performance of the alanate with additive particle size.220 TiO2 nanoparticles improve the performance of sodium alanate, which was accredited to enhanced catalytic activity, and from Fig. 30 it is obvious that smaller particles of TiO2, widely dispersed across the the surface of the alanate, perform better than larger oxide particles.221,222 However, milling of alanates presents the same size distribution and control issues as for light metal hydrides (section 4.1). The high theoretical hydrogen capacity of magnesium alanate, 9.3 wt%, is appealing and Fichtner et al. produced magnesium alanate nanoparticles via a wet synthesis method (30–40 nm).223 The performance of the alanate in this instance was typical of an unmodi?ed metal hydride, i.e. high desorption temperatures and high regeneration pressures, as can be seen in Fig. 31. Research by Varin et al. presented similar hydrogen desorption results with milled magnesium alanate, alluding also to the irreversibility of hydrogen storage within this material.224,225 Templating of sodium alanate in a THF solution using carbon nano?bres has been presented by Balde et al. as a means by which large size distributions can be more effectively restricted.226 The size of alanate nanoparticles could be mediated by the loading of alanate on the carbon ?bres; 1–10 mm, 19–30 nm, and 2–10 nm diameters for 9 wt%, 8 wt%, and 2 wt% loadings of the alanate respectively. The alanate decomposed entirely at about 423 K and temperature programmed desorption (TPD) with XRD showed that the hydrogen desorption temperatures were lower for the smaller particles. This was attributed to a lower activation

Fig. 30 Effect of admixed TiO2 particle size on hydrogen desorption from sodium alanate. Reprinted from Energy, 2010, 35, 5037. Studies on metal oxide nanoparticles catalyzed sodium aluminum hydride, D. Pukazhselvan, M. S. L. Hudson, A. S. K. Sinha and O. N. Srivastava. Copyright 2010, with permission from Elsevier.

Energy Environ. Sci., 2012, 5, 5951–5979 | 5971

View Article Online

amide, is readily reversible. Therefore it is possible to store theoretically ca. 6.5 wt% of hydrogen reversibly below 573 K. Following the initial discovery above, large numbers of similar systems have been studied.238,239 Following the DOE guidelines, the most attention has been paid to lightweight systems containing Li, Mg, Al, B and Na. On the nano-scale several different approaches have been reported, and are detailed below.
Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Fig. 31 Thermogravimetric analysis-mass spectrometry (TGA-MS) data for decomposition of magnesium alanate nanoparticles. Reprinted from Mater. Sci. Eng., B, 2004, 108, 42. Studies on metal oxide nanoparticles catalyzed sodium aluminum hydride, M. Fichtner, J. Engel, O. Fuhr, O. Kircher and O. Rubner. Copyright 2004, with permission from Elsevier.

energy of the nanoscale particles. More recently, desorption of sodium alanate melt-in?ltrated into a porous carbon aerogel (to form nanopores $13 nm in diameter) was compared to alanate cataysed with TiCl3.227 The desorption pro?les of the catalysed alanate are superior to the aerogel infused alanate and according to this study, a catalysed alanate aerogel is the next step in establishing a suitable alanate hydrogen storage material. 4.2.2 Borohydrides. Destabilisation of borohydride systems has been investigated to establish how operating parameters may be improved for hydrogen storage, e.g. by milling with one or multiple additives, with promising results.228–230 Further, nanocon?nement of borohydrides in porous carbon has been reported over the past decade and has shown encouraging results for magnesium, lithium, and sodium borohydrides; all compounds displaying a lower hydrogen desorption temperature than their respective bulk borohydride compounds.231–234 Signi?cantly, it was shown that the smaller the pore into which lithium borohydride was placed, the more signi?cant the decrease in desorption temperature and LiBH4 constrained in 13 nm pores dehydrogneated up to 50 times faster than non-con?ned material.235 Liu et al. demonstrated, through XRD analyses, that differences can be seen in the structure of lithium borohydride upon constraining it in a porous carbon matrix, i.e. crystalline lithium borohydride is observed at higher pore sizes (ca. 25 nm), but amorphous borohydride is present as the pore size decreases.236 Further, they proposed that the degradation path of the LiBH4 is modi?ed when contained in the nanopores with the expulsion of toxic diborane, a signi?cant safety and performance improvement over the bulk material. 4.2.3 Amides. The publication in 2002 by Chen et al. on hydrogen storage in lithium nitride signi?ed a new direction in the research into hydrogen storage in complex hydrides.237 The storage process proceeds via two steps: Li3N + 2 H2 / Li2NH + LiH + H2 4 LiNH2 + 2 LiH (1) The theoretical gravimetric capacity of this system is 10.3 wt% but in fact only the second step, cycling between imide and
5972 | Energy Environ. Sci., 2012, 5, 5951–5979

4.2.3.1 Ball milling. Ball milling has become an almost ubiquitous technique for amide based systems. The decomposition of LiNH2 has been shown to be induced by ball millling.240 Some more detailed experiments have attempted to quantify how the ball milling improves performance. Markmaitree et al. showed that ball milling for 180 min reduced the crystallite size to 5.5 nm.241 This had the effect of increasing the surface area by an order of magnitude over the as-received LiNH2 (of >100 nm particle size) and also introduced internal strain, suggesting defects are present in the milled material. The decomposition onset temperature in the milled sample is reduced from 393 K to 298 K. 6Li MAS NMR measurements have shown that the mechanical activation also leads to change of local electronic structure around lithium and 1H NMR shows that dehydrogenation is more favourable than ammonia release in the mechanically activated sample.242 Varin et al. studied the effects of both ball milling and reagent ratio on the lithium amide - lithium hydride reaction.243 They also found that increasing milling time reduces crystallite size and increases surface area, although after 100 h, the surface area begins to decrease again due to agglomeration effects. Ball milling is often employed in preparing amide/hydride mixtures without detailed analysis of the crystallite size post-milling. It is therefore quite probable that many more of these reactions are enhanced by the presence of nano-crystallites. 4.2.3.2 Nanocomposites. The reaction of LiNH2 with LiBH4 in a 2 : 1 ratio (eqn (2)) offers a theoretical capacity of 11.9 wt%. It has been subsequently found to proceed via an intermediate quaternary phase (eqn (3)):244 2 LiNH2 + LiBH4 / Li3BN2 + 4 H2 (2)

2 LiNH2 + LiBH4 / Li3(BH4)(NH2)2 / Li3BN2 + 4 H2 (3) The drawback of the system is that normally the Li3BN2 is not rehydridable. This process does however become reversible when the material is incorporated into composite systems. Wu et al. incorporated Li3(BH4)(NH2)2 into nanoporous carbon scaffolds either by melting it with carbon aerogel, or ball milling it with activated carbon.245 The particle size of the ternary nitride was calculated to be 13 nm in the aerogel. Overall gas release on the ?rst cycle is 11.1 wt% with the release mid-point at ca. 593 K. The material can subsequently be rehydrided to a level of 3.8 wt%. (at 50 bar, 573 K) which can then be completely dehydrided again. In the activated carbon sample (particle size 2 nm) the dehydrogenation onset temperature is reduced to 438 K. 10.7 wt% gas was evolved on cycle 1 and 4% could be re-absorbed. Further investigation however, revealed scaffold incorporation increases the amount of NH3 release compared to the bulk. This implies that different dehydrogenation mechanisms may be
This journal is ? The Royal Society of Chemistry 2012

View Article Online

occurring (cf. MgH2 section 4.1.2.8). Therefore, the overall hydrogen release of the carbon scaffold-hosted Li3(BH4)(NH2)2 is less, and this may also explain why only $4 wt% is able to be re-absorbed. These systems might be improved by incorporation of transition metal catalysts into the structure by analogy to noncomposite amide-borohydrides (see below).
Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

4.2.3.3 Nanocatalysts. The hydrogen storage properties of many nitride based systems can be improved by the addition of catalyst materials, which are usually nanoparticles themselves. Ichikawa et al. studied the addition of metal or metal chloride nanoparticles to the lithium amide / lithium hydride reaction.246 The best improvement was shown in the reaction of 1 : 1 molar LiNH2:LiH with 1 mol% of TiCl3 added. The hydrogen desorption rate was increased, indicating a reduced activation energy. The catalysed reaction also showed improved cyclability. Further study by Isobe et al. compared micro- vs. nanoscale titanium based catalysts.247 The results show that the particle size of the catalyst is important – no improvement is shown in the samples with micro-sized additives. They also showed that in addition to TiCl3, nano-sized titanium metal and TiO2 are also able to catalyse the dehydrogenation. Pinkerton et al. studied the effect of various catalysts on the dehydrogenation of the quaternary material Li3(BH4)(NH2)2.248 The best results obtained were with an 11 wt% doping with NiCl2. The particle size of the additive was approximately 8 nm and a large (?112 K) shift in the midpoint of the hydrogen desorption exists compared to the additive free sample (Fig. 32). Kojima et al. observed improved properties in a nano-nickel catalysed reaction between Li3AlH6 and LiNH2.249 The addition of the catalyst improves the cyclability of the system. Initial desorption releases 6.9 wt% hydrogen. Subsequent cycling only absorbs and desorbs 1–2 wt% in the uncatalysed material whereas addition of 5 wt% of nickel catalyst (20 nm particle size) doubles the cycling capacity to 3–4 wt%. 4.2.3.4 Synthesis of nanomaterials by design. Chemical synthesis can lead to nanoscale materials with improved kinetics in both hydrogen uptake and release. Xie et al. synthesised hollow nanospheres of Li2NH through a plasma metal reaction (Fig. 33).250 Lithium was evaporated using an arc in an Ar/NH3 mix. The spheres produced range from 100–400 nm (with 90% in the 100–200 nm range) with a shell thickness of ca. 20 nm. H2 storage properties were investigated and compared to micron sized Li2NH. The latter started to absorb at 550 K, peaking at 610 K, whereas the hollow spheres began uptake at 298 K and peak at 470 K. The rate was also enhanced with the nanospheres absorbing 6 wt% in <1 min at 473 K. Desorption showed an onset temperature reduced by 120 K compared to the micron sized material (to 452 K from 571 K). Via similar methods, Xie et al. also prepared nanospheres of magnesium amide.251 Mg3N2 nanocubes prepared by plasma metal methods were subsequently treated with ammonia to form the amide spheres. Reacting these with MgH2 released about 2.5 wt% hydrogen. Further reactions performed with LiH, using Mg(NH2)2 of 100, 500 and 2000 nm diameter showed that peak desorption temperatures decrease with decreasing particle size; 2000 nm, 500 nm and 100 nm diameter spheres desorb at 539.8 K, 514.8 K and 482.5 K respectively.252 Differential Scanning
This journal is ? The Royal Society of Chemistry 2012

Fig. 32 The effect of Ni additions on dehydrogenation of Li3(BH4)(NH2)2. The temperature was ramped at 5  C min?1 in 100 kPa Ar. The dots indicate the T1/2 values for additive-free and 11 wt% TiCl2added Li3(BH4)(NH2)2. The error bar indicates the typical standard deviation of the total weight loss due to the small sample size. Reprinted from J. Alloys Compd. 433, 282. F. Pinkerton, M. Meyer, G. Meisner and M. Balogh, Improved hydrogen release from LiB0.33N0.67H2.67 with metal additives: Ni, Fe, and Zn. Copyright 2007, with permission from Elsevier.

Calorimetry (DSC) measurements demonstrated that the desorption activation energy decreases on reduction of particle size with values of 182.0, 134.7 and 122.2 kJ mol?1 for the 2000, 500 and 100 nm particles respectively. Metal plasma reactions with ammonia, therefore, seem a promising route to nanoscale imide and amides, which offer enhanced hydrogen desorption performance over bulk materials. 4.2.4. Ammonia borane. Ammonia borane (AB; NH3BH3) has a high gravimetric content of hydrogen (19.6 wt%). In practice, up to two thirds of the hydrogen may be released by

Fig. 33 (a) SEM image, (b) TEM image (inset: magni?ed TEM image) of the as-prepared Li2NH hollow nanospheres, (c) TEM image of vacuum annealed Li2NH hollow nanospheres and (d) TEM image of the hollow nanospheres after hydrogenation at 573 K, 35 bar. Reprinted with permission from L. Xie, J. Zheng, Y. Liu, Y. Li and X. Li, Chem. Mater., 2007, 20, 282. Copyright 2007 American Chemical Society.

Energy Environ. Sci., 2012, 5, 5951–5979 | 5973

View Article Online

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

decomposition up to 423 K. Higher temperatures tend to promote borazine formation, which can poison a fuel cell. Ammonia borane is poorly reversible, which limits its application, although off-board regeneration may be possible.253–257 In 2005, Gutowska et al. reported the con?nement of ammonia borane within a nanoporous host.258 Ammonia borane was dissolved in methanol and impregnated into SBA-15 mesoporous silica with a pore diameter of 7–8 nm; 50 wt% was loaded and then dried. TPD / MS pro?les of bulk vs. the nanoscaffolded products show that the con?nement reduces the onset temperatures for the ?rst two equivalents of hydrogen. Borazine formation is also retarded in the SBA-15 composite (Fig. 34). 11B NMR measurements show that the borazine is not trapped within the silica–therefore implying that the decomposition pathway has been altered by the con?nement. DSC provided activation energies for hydrogen release which are reduced from 184 kJ mol?1 in the bulk to 67 kJ mol?1 in the SBA-15 material. A similar effect was seen by Feaver et al. who produced nanocomposite materials with AB and carbon cryogels.259 The pore size in these materials is in the range 2–20 nm. In this system the hydrogen desorption is reduced from a two step process at 383 K and 423 K in the bulk, to a single-step hydrogen release at around 363 K. The difference is attributed to the size-dependent surface energy of the AB in the pores, and also possible catalytic enhancement from the AB-carbon interface. Recently initial reports have been made of another method for con?nement of AB. Kurban et al. have used co-axial electrospinning in polystyrene to encapsulate AB.260 Polystyrene is stable up to 473 K, so should remain intact for the thermolysis of the ammonia borane. In the co-electrospinning process, core and shell precursor solutions are delivered independently through concentric nozzles. When the electric ?eld is high enough to overcome the surface tension of a droplet, the ?uids are drawn to a point and a jet is created. The jet spirals, stretches and thins into nano?bers. Variation of the solvents employed and spinning

conditions creates a variety of different nano?bre structures (Fig. 35). Initial investigations indicate that as for silica and carbon cryogel encapsulation, the H2 desorption temperature in the ?bres is decreased compared to bulk AB (by 15–20 K) with an associated borazine retardation.

5. Kinetics and thermodynamics in nanostructured storage materials
Before closing, it is useful to consider the relative kinetic and particularly, thermodynamic (i.e. enthalpies of adsorption and desorption) properties of various nanometric storage materials. Since the interaction strength of hydrogen with the solid is re?ected in the heat of adsorption then this is a direct measure of the temperature at which hydrogen adsorption-desorption occurs. Therefore, the optimisation of this interaction strength is crucial for achieving near room temperature uptake and release of hydrogen. In addition, the uptake and release reaction pathways can be modi?ed via the relative activation energies of these processes and faster kinetics (uptake and release rate) are also desirable from a practical application perspective. It is very evident that a nanodesign approach can make a major impact in both improving sorption kinetics and in enhancing the interaction strength (enthalpy of ad(de)sorption) and thereby signi?cantly reducing the temperature of hydrogen release as compared to the respective bulk analogues. The schematic (Fig. 36) summaries some of the nanomaterials discussed in previous sections in terms of the relative sorption (hydrogenation) enthalpy values. It is eminently noticeable in Fig. 36 that physisorbed and chemisorbed materials vary widely in their enthalpy of adsorption and desorption and both deviate from the ‘green bar’ which indicates the ideal value of the enthalphy (15–25 kJ mol ?1 of H2) for near-ambient temperature hydrogen sorption.41,42 Great efforts are continuing with the design of physisorbed materials especially MOFs and COFs - to enhance the interaction strength

Fig. 34 TPD-MS (1 K min?1) of volatile products from heating ammonia borane (solid line) and AB:SBA-15 (dashed line); m/z ? 2 (H2) and m/z ? 80 (borazine, c-(NHBH)3). The value m/z ? 2 (H2) is normalized to the area under the curve for ammonia borane (solid line). The corresponding scalar was used to normalize the m/z ? 80 borazine data for AB:SBA-15. Reprinted with permission from ref. 258. Copyright 2005 John Wiley and Sons.

Fig. 35 Various electrospun AB-polystyrene nano?bers. Reprinted with permission from Z. Kurban, A. Lovell, S. M. Bennington, D. W. K. Jenkins, K. R. Ryan, M. O. Jones, N. T. Skipper and W. I. F. David, J. Phys. Chem. C, 2010, 214, 21201. Copyright 2010 American Chemical Society.

5974 | Energy Environ. Sci., 2012, 5, 5951–5979

This journal is ? The Royal Society of Chemistry 2012

View Article Online

so that higher storage capacities at room temperature can be realised. To obtain high H2 uptake under these conditions, the introduction of exposed metal sites and doping with electropositive elements (such as Li, Na, K) have been employed to increase H2 binding energy. Theory has predicted that this approach should reap rewards and some success has already been achieved experimentally.35,52 By way of example, for MOFs with exposed Mg and Cu sites an adsorption enthalpy of 9.5 and 10.1 kJ mol ?1 of H2, respectively, could be reached.35 Likewise, in Lidoped MOF-C30 much stronger hydrogen binding has been realised giving rise to a hydrogen uptake of 3.89 wt% at room temperature (at 20 bar).52 Among other physisorbed materials, nanoporous polymers have recently emerged as promising materials for hydrogen storage. For example, hypercrosslinked PANI exhibits strikingly high af?nity for hydrogen as re?ected in the enthalpy of adsorption with values as high as 9.3 kJ mol?1.97 With PANI nano?bres, hydrogen capacity and kinetics improves with increasing temperature due to the presence of both physisorption and chemisorption sites.91 Nanocon?nement strategies and use of nano-catalysts have been shown to be effective methods for enhancing hydrogen release kinetics. For example, the activation energy for hydrogen release from nanoporous SBA-15 con?ned ammonia borane was reduced to 67 kJ mol ?1 as compared to 184 kJ mol ?1 in the bulk.258 Controlled (prescribed) nanostructuring methods in the light metal and complex hydrides are challenging and at a relatively early stage, yet already yield encouraging results.186,187 For example, the dehydriding activation energies for thinner Mg nanowires were much less as compared to their bulk counterparts.187 Although nanostructuring through ball milling of MgH2, for example, may not alter the thermodynamics of the system, kinetic enhancements can be achieved.186 From the growing evidence available, materials design at the nanoscale can have dramatic effects on both the thermodynamics and the kinetics of H2 uptake and release. The challenges now are both to extend nanostructuring approaches that are effective in light metal hydrides (MgH2) to complex hydrides and to stabilise the interactions of hydrogen with porous solids at the nanoscale.

6. Closing remarks
Nanomaterials design represents an emerging approach to challenge the limitations of ‘‘classical’’ hydrogen storage in bulk solids. Both physical and chemical means of storage can prosper from nano-modi?cation. At one extreme, for example, MOFs bene?t from nanoparticle addition (to generate catalytic centers for hydrogen spillover) and metal doping; two step substitution/ doping in COFs can have signi?cant results as can metal incorporation in polymer matrices. At the other extreme, the nanostructuring of the host material itself can have profound effects, e.g. polymer nano?bres and light and complex hydride nanomaterials display a wide range in performance in storing hydrogen. Between these two extremes lies the relatively uncharted territory of inorganic nanomaterials; compounds which normally show no appreciable uptake of hydrogen in their bulk forms. Despite these growing studies, challenges remain to be addressed if one is to employ such materials in viable power systems, especially those where the application demands are the most stringent such as in a mobile, onboard store coupled to a fuel cell (FC) stack for automotive usage. The MOFs possess many advantages as superior hydrogen storage materials due to their favourable kinetics and reversibility. However, the key issue for MOFs to address is the low adsorption of H2 at normal operating conditions (and low heat of H2 adsorption at room temperature). In some cases also, the volumetric hydrogen density would need to be improved and relative costs of production would have to be considered. The modi?cation of the internal nanostructure in MOFs and control of potentially bene?cal effects such as hydrogen spillover are the ?rst priorities in overcoming the major challenge of retaining hydrogen below FC operating temperatures. COFs face many of the same issues but have the apparent advantage (from P-C-T isotherm evidence) of outperforming MOFs at high hydrogen pressures and similar temperatures. Similar nanostructuring strategies as those adopted for MOFs should be feasible for COFs. Studies of polymer matrices (such as polyacetylene, polyaniline, polypyrrole and polystyrene) as hydrogen stores, although still at relatively early stages of development, are very encouraging.85–87,91–95,109,110Attempts have been made to enhance hydrogen storage capacities of polymer matrices, for example, through metal particle doping in polyacetylene85,86 and addition of various additives, e.g. alloys, Al, tin oxide and carbon nanotubes etc., in PANI,98–100 though with somewhat limited success to date. Hypercrosslinking of polymer matrices, especially with PANI and PPY, has been employed extensively to alter nano?bre morphologies and encouragingly this approach could result in hydrogen selective nano?bres.96,97,107 The challenge in these systems is then to increase the storage capacity further. In fact, two main hurdles in using polymeric materials should be highlighted; the weight penalty and the expense. The former is more relevant when utilising polymer matrices as a component in a composite since unless the polymer itself contributes signi?cantly to the potential to store hydrogen then it simply detracts gravimetrically from the performance of an active hydridic component. The latter (cost) is commented upon by Checchetto et al. in their experimental work on the encapsulation of LaNi5, where commercial availability and synthesis costs would certainly have an impact on later, commercial stages in product
Energy Environ. Sci., 2012, 5, 5951–5979 | 5975

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Fig. 36 Schematic showing the enthalpy of adsorption ranges for various types of storage materials.

This journal is ? The Royal Society of Chemistry 2012

View Article Online

development.261 However, there are tremendous possibilities for nanostructural modi?cation by design in complex polymer matrices. Recent studies with boron nitride and many other inorganic nanotubes, such as, MoS2, TiS2, TiO2 and ZnO have not yet resulted in a high capacity material. At present mobile applications appear to be beyond the reach of these materials but their near-ambient temperature gravimetric capacities could ?nd applications in stationary power generation. The use of nano-additives as catalysts has been shown to improve the hydrogen cycling ability of light metal hydrides.203–208 Similar effects have also been observed when nano-scaffolding magnesium hydride.211,212 In addition to ball milling,182–186 nanoscale synthesis by vapour deposition,187,189,190 solvated metal atom dispersion200,201 and electrochemical199 techniques have emerged as methods of generating metal hydride nanoparticles and nanowires. Initial hydrogen uptake-release studies involving Mg nanoparticles200,201 have shown promise with improved hydrogen sorption-desorption kinetics and with a signi?cant decrease in the hydrogen desorption temperature (358–388 K) as compared to that of the bulk.202 Although ball milling has been used extensively to improve performance of simple and complex hydride and amide systems,240–242 appropriate correlation and optimisation of thermodynamic and kinetic parameters with regard to milling speed, milling time and sample to ball ratio is still lacking. This is where an as yet immature chemical approach to nanostructuring would be a distinct advantage. Nanostructuring in complex hydrides has also been investigated to optimise the thermodynamics and kinetics of hydrogen uptake-release cycles. Interestingly, studies involving nanoscale lithium imide and magnesium amide have revealed much faster kinetics (adsorption in less than 1 min) at a reduced temperature (473 K) as compared to their micron sized counterparts.250,251 The particle size of Mg(NH2)2 for example, has a dramatic effect on the desorption temperature when reacted with LiH for hydrogen release.252 Striking performances are also seen via the nanocon?nement of ammonia borane into SBA-15 mesoporous silica, with a reduced activation energy for hydrogen release.258 Interestingly also, a single step hydrogen release at 90  C was observed for a nanocomposite of carbon cryogel and ammonia borane.259 The con?nement of ammonia borane in nanopores of SBA-15 and carbon cryogels lowers the desorption temperature and suppresses the formation of borazine, a fuel-cell poison. Overall, there are clearly immense opportunities for the design and modi?cation of materials at the nanoscale towards superior hydrogen storage materials. Already these are yielding major changes in the way in which one understands storage of hydrogen compared to that in bulk materials and nanomaterials offer the possibility to bridge the chemistry between hydrogen-surface interactions and bulk chemical bond formation. Despite the many recent advances, one feels that much of the potential as yet remains untapped.

Notes and references
1 M. S. Dresselhaus and I. L. Thomas, Nature, 2001, 414, 332. 2 L. Schlapbach and A. Zttel, Nature, 2001, 414, 353. u 3 (a) D. H. Gregory, J. Mater. Chem., 2008, 18, 2321; (b) T. K. Mandal and D. H. Gregory, Annu. Rep. Prog. Chem., Sect. A, 2009, 105, 21; (c) T. K. Mandal and D. H. Gregory, Proc. Inst. Mech. Eng., Part C, 2010, 224(3), 539; J. M. Cameron, R. W. Hughes, Y. Zhao and D. H. Gregory, Chem. Soc. Rev., 2011, 40, 4099. 4 W. Grochala and P. P. Edwards, Chem. Rev., 2004, 104, 1283. 5 S. Orimo, Y. Nakamori, J. R. Eliseo, A. Zttel and C. M. Jensen, u Chem. Rev., 2007, 107, 4111. 6 T. K. Nielsen, F. Besenbacher and T. R. Jensen, Nanoscale, 2011, 3, 2086. 7 J. Graetz, Chem. Soc. Rev., 2009, 38, 73. 8 J. Yang, A. Sudik, C. Wolverton and D. J. Siegel, Chem. Soc. Rev., 2010, 39, 656. 9 R. Strobel, J. Garche, P. T. Moseley, L. Jorissen and G. Wolf, J. Power Sources, 2006, 159, 781. 10 P. Bnard and R. Chahine, Scr. Mater., 2007, 56, 803. e 11 Y. Yr m, A. Taralp and T. N. Veziroglu, Int. J. Hydrogen Energy, uu 2009, 34, 3784. 12 X. Hu, M. Fan, B. F. Towler, M. Radosz and D. A. Bell, Chapter 8Hydrogen Adsorption and Storage, Elsevier, 2011. 13 M. Pumera, Energy Environ. Sci., 2011, 4, 668. 14 J. Biener, M. Stadermann, M. Suss, M. A. Worsley, M. M. Biener, K. A. Rose and T. F. Baumann, Energy Environ. Sci., 2011, 4, 656. 15 J. G. Vitillo, G. Ricchiardi, G. Spoto and A. Zecchina, Phys. Chem. Chem. Phys., 2005, 7(23), 3948. 16 J. Iniguez, in Solid-state hydrogen storage: Materials and chemistry, ed. G. Walker, Woodhead Publishing Ltd, Cambridge, 2008, pp 205–260. 17 M. Fichtner, Adv. Eng. Mater., 2005, 7, 443. 18 M. G. Nijkamp, J. E. M. J. Raaymakers, A. J. van Dillen and K. P. de Jong, Appl. Phys. A: Mater. Sci. Process., 2001, 72, 619. 19 H. W. Langmi, A. Walton, M. M. Al-Mamouri, S. R. Johnson, D. Book, J. D. Speight, P. P. Edwards, I. Gameson, P. A. Anderson and I. R. Harris, J. Alloys Compd., 2003, 356–357, 710. 20 H. W. Langmi, D. Book, A. Walton, S. R. Johnson, M. M. AlMamouri, J. D. Speight, P. P. Edwards, I. R. Harris and P. A. Anderson, J. Alloys Compd., 2005, 404–406, 637. 21 V. B. Kazansky, V. Yu. Borovkov, A. Serich and H. G. Karge, Microporous Mesoporous Mater., 1998, 22, 251. 22 D. Bae, H. Park, Ji. S. Kim, J.- b. Lee, O. Y. Kwon, K.-Y. Kim, M. K. Song and K. T. No, J. Phys. Chem. Solids, 2008, 69, 1152. 23 S. B. Kayiran and F. L. Darkrim, Surf. Interface Anal., 2002, 34, 100. 24 X. M. Du and E. D. Wu, Chin. J. Chem. Phys., 2006, 19, 457. 25 J. Weitkamp, M. Fritz and S. Ernst, Int. J. Hydrogen Energy, 1995, 20(12), 967. 26 J. L. C. Rowsell and O. M. Yaghi, Angew. Chem., Int. Ed., 2005, 44, 4670. 27 S. T. Meek and J. A. Greathouse, Adv. Mater., 2011, 23, 249. 28 R. Ostermann, J. Cravillon, C. Weidmann, M. Wiebcke and B. M. Smarsly, Chem. Commun., 2011, 47, 442. 29 N. L. Rosi, J. Eckert, M. Eddaoudi, D. T. Vodak, J. Kim, M. O. Keeffe and O. M. Yaghi, Science, 2003, 300, 1127. 30 A. G. Wong-Foy, A. J. Matzger and O. M. Yaghi, J. Am. Chem. Soc., 2006, 128, 3494. 31 H. Furukawa, M. A. Miller and O. M. Yaghi, J. Mater. Chem., 2007, 17, 3197. 32 M. Latroche, S. Surble, C. Serre, C. Mellot-Draznieks, P. L. Llewellyn, J.-H. Lee, J.-S. Chang, S. H. Jhung and G. Ferey, Angew. Chem., Int. Ed., 2006, 45, 8227. 33 H. Furukawa, N. Ko, Y. B. Go, N. Aratani, S. B. Choi, E. Choi, A. O. Yazaydin, R. Q. Snurr, M. O. ’Keeffe, J. Kim and O. M. Yaghi, Science, 2010, 329, 424. 34 Y. H. Hu and L. Zhang, Adv. Mater., 2010, 22, E117. 35 L. J. Murray, M. Dinc and J. R. Long, Chem. Soc. Rev., 2009, 38, a 1294. 36 J. Sculley, D. Yuan and H.-C. Zhou, Energy Environ. Sci., 2011, 4, 2721. 37 X. Lin, J. Jia, X. Zhao, K. M. Thomas, A. J. Blake, G. S. Walker, N. R. Champness, P. Hubberstey and M. Schrder, Angew. Chem., o Int. Ed., 2006, 45, 7358.

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

Acknowledgements
DHG thanks EPSRC for funding under grants EP/E040071/1 and EP/H500138/1 and EPSRC, the Materials KTN and EADS Innovation Works for a CASE studentship for HR.
5976 | Energy Environ. Sci., 2012, 5, 5951–5979

This journal is ? The Royal Society of Chemistry 2012

View Article Online
38 B. Xiao, P. S. Wheatley, X. Zhao, A. J. Fletcher, S. Fox, A. G. Rossi, I. L. Megson, S. Bordiga, L. Regli, K. M. Thomas and R. E. Morris, J. Am. Chem. Soc., 2007, 129, 1203. 39 D.-C. Zhong, J.-B. Lin, W.-G. Lu, L. Jiang and T.-B. Lu, Inorg. Chem., 2009, 48, 8656. 40 Y. Yan, I. Telepeni, S. Yang, X. Lin, W. Kockelmann, A. Dailly, A. J. Blake, W. Lewis, G. S. Walker, D. R. Allan, S. A. Barnett, N. R. Champness and M. Schrder, J. Am. Chem. Soc., 2010, 132, o 4092. 41 S. K. Bhatia and A. L. Myers, Langmuir, 2006, 22, 1688. 42 C. O. Aren, S. Chavan, C. P. Cabello, E. Garrone and a G. T. Palomino, ChemPhysChem, 2010, 11, 3237. 43 K. M. Thomas, Dalton Trans., 2009, 1487. 44 R. C. Lochan, R. Z. Khaliullin and M. H. -Gordon, Inorg. Chem., 2008, 47, 4032. 45 G. Vitillo, L. Regli, S. Chavan, G. Ricchiardi, G. Spoto, P. D. C. Dietzel, J. S. Bordiga and A. Zecchina, J. Am. Chem. Soc., 2008, 130, 8386. 46 K. Gedrich, I. Senkovska, N. Klein, U. Stoeck, A. Henschel, M. R. Lohe, I. A. Baburin, U. Mueller and S. Kaskel, Angew. Chem., Int. Ed., 2010, 49, 8489. 47 E. Y. Lee and M. P. Suh, Angew. Chem., Int. Ed., 2004, 43, 2798. 48 O. K. Farha, A. O. Yazaydiin, I. Eryazici, C. D. Malliakas, B. G. Hauser, M. G. Kanatzidis, S. T. Nguyen, R. Q. Snurr and J. T. Hupp, Nat. Chem., 2010, 2, 944. 49 S. Yang, X. Lin, A. J. Blake, G. S. Walker, P. Hubberstey, N. R. Champness and M. Schroder, Nat. Chem., 2009, 1, 487. 50 E. Klontzas, A. Mavrandonakis, E. Tylianakis and G. E. Froudakis, Nano Lett., 2008, 8, 1572. 51 D. Himsl, D. Wallacher and M. Hartmann, Angew. Chem., Int. Ed., 2009, 48, 4639. 52 S. S. Han and W. A. Goddard III, J. Am. Chem. Soc., 2007, 129, 8422. 53 H. K. Chae, D. Y. Siberio-Perez, J. Kim, Y. Go, M. Eddaoudi, A. J. Matzger, M. O’Keeffe and O. M. Yaghi, Nature, 2004, 427, 523. 54 T. Yildrium and S. Ciraci, Phys. Rev. Lett., 2005, 94, 175501. 55 A. W. Thornton, K. M. Nairn, J. M. Hill, A. J. Hill and M. R. Hill, J. Am. Chem. Soc., 2009, 131, 10662. 56 D. Rao, R. Lu, C. Xiao, E. Kan and K. Deng, Chem. Commun., 2011, 47, 7698. 57 K. P. Prasanth, P. Rallapalli, M. C. Raj, H. C. Bajaj and R. V. Jasra, Int. J. Hydrogen Energy, 2011, 36, 7594. 58 A. J. Robell, E. Ballou and M. Boudart, J. Phys. Chem., 1964, 68, 2748. 59 L. Wang and R. T. Yang, Energy Environ. Sci., 2008, 1, 268. 60 C. Zlotea, R. Campesi, F. Cuevas, E. Leroy, P. Dibandjo, C. Volkringer, T. Loiseau, G. Ferey and M. Latroche, J. Am. Chem. Soc., 2010, 132, 2991. 61 K. Gedrich, I. Senkovska, N. Klein, U. Stoeck, A. Henschel, M. R. Lohe, I. A. Baburin, U. Mueller and S. Kaskel, Angew. Chem., Int. Ed., 2010, 49, 8489. 62 S. Proch, J. Herrmannsdrfer, R. Kempe, C. Kern, A. Jess, o L. Seyfarth and J. Senker, Chem.–Eur. J., 2008, 14, 8204. 63 M. Sabo, A. Henschel, H. Frde, E. Klemm and S. Kaskel, J. Mater. o Chem., 2007, 17, 3827. 64 Y. Li and R. T. Yang, J. Am. Chem. Soc., 2006, 128, 726. 65 Y. Li and R. T. Yang, J. Am. Chem. Soc., 2006, 128, 8136. 66 Y. Li and R. T. Yang, AIChE J., 2008, 54, 269. 67 Y.-Y. Liu, J.-L. Zeng, F. Xu and L.-X. Sun, Int. J. Hydrogen Energy, 2007, 32, 4005. 68 C.-S. Tsao, M.-S. Yu, C.-Y. Wang, P.-Y. Liao, H.-L. Chen, U.-S. Jeng, Y.-R. Tzeng, T.-Y. Chung and H.-C. Wu, J. Am. Chem. Soc., 2009, 131, 1404. 69 N. R. Stuckert, L. Wang and R. T. Yang, Langmuir, 2010, 26, 11963. 70 P. Ryan, L. J. Broadbelt and R. Q. Snurr, Chem. Commun., 2008, 4132. 71 O. K. Farha, C. D. Malliakas, M. G. Kanatzidis and J. T. Hupp, J. Am. Chem. Soc., 2010, 132, 950. 72 O. K. Farha and J. T. Hupp, Acc. Chem. Res., 2010, 43, 1166. 73 S. Ma, J. Eckert, P. M. Forster, J. W. Yoon, Y. K. Hwang, J.-S. Chang, C. D. Collier, J. B. Parise and H.-C. Zhou, J. Am. Chem. Soc., 2008, 130, 15896. 74 X. Liu, M. Park, S. Hong, M. Oh, J. W. Yoon, J.-S. Chang and M. S. Lah, Inorg. Chem., 2009, 48, 11507. 75 M. Hirscher, Angew. Chem., Int. Ed., 2011, 50, 581. 76 H. M. El-Kaderi, J. R. Hunt, J. L. Mendoza-Corts, A. P. C^te, e o R. E. Taylor, M. O’Keeffe and O. M. Yaghi, Science, 2007, 316, 268. 77 A. P. C^te, A. I. Benin, N. W. Ockwig, M. O’Keeffe, A. J. Matzger o and O. M. Yaghi, Science, 2005, 310, 1166. 78 F. J. Uribe-Romo, J. R. Hunt, H. Furukawa, C. Klock, M. O’Keeffe and O. M. Yaghi, J. Am. Chem. Soc., 2009, 131, 4570. 79 S. S. Han, H. Furukawa, O. M. Yaghi and W. A. Goddard III, J. Am. Chem. Soc., 2008, 130, 11580. 80 H. Furukawa and O. M. Yaghi, J. Am. Chem. Soc., 2009, 131, 8875. 81 A. G. Wong-Foy, A. J. Matzger and O. M. Yaghi, J. Am. Chem. Soc., 2006, 128, 3494. 82 X. Zou, G. Zhou, W. I. Duan, K. Choi and J. Ihm, J. Phys. Chem. C, 2010, 114, 13402. 83 J. R. Hunt, C. J. Doonan, J. D. LeVangie, A. P. Cote and O. M. Yaghi, J. Am. Chem. Soc., 2008, 130, 11872. 84 T. Wu, J. Zhang, C. Zhou, L. Wang, X. Bu and P. Feng, J. Am. Chem. Soc., 2009, 131, 6111. 85 T. Ito, H. Shirakawa and S. Ikeda, J. Polym. Sci., Polym. Chem. Ed., 1974, 12, 11. 86 H. Lee, W. I. Choi and J. Ihm, J. Alloys Compd., 2007, 446–447, 373. 87 H. Lee, W. I. Choi and J. Ihm, Phys. Rev. Lett., 2006, 97, 056104. 88 H. Lee, W. I. Choi, M. C. Nguyen, M.-H. Cha, E. Moon and J. Ihm, Phys. Rev. B: Condens. Matter Mater. Phys., 2007, 76, 195110. 89 S. Li and P. Jena, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 77, 193101. 90 J. Strejskal and R. G. Gilbert, Pure Appl. Chem., 2002, 74, 857. 91 S. S. Srinivasan, R. Ratnadurai, M. U. Niemann, A. R. Phani, D. Y. Goswami and E. K. Stefanakos, Int. J. Hydrogen Energy, 2010, 35, 225. 92 A. Rahy, T. Rguig, S. J. Cho, C. E. Bunker and D. J. Yang, Synth. Met., 2011, 161, 280. 93 S. J. Cho, K. S. Song, J. W. Kim, T. H. Kim and K. Choo, Prepr. Pap. - Am. Chem. Soc., Div. Fuel Chem., 2002, 47, 790. 94 B. Panella, L. Kossykh, U. Dettlaff-Weglikowska, M. Hirscher, G. Zerbi and S. Roth, Synth. Met., 2005, 151, 208. 95 S. J. Cho, K. Choo, D. P. Kim and J. W. Kim, Catal. Today, 2007, 120, 336. 96 J. Germain, J. M. J. Frchet and F. Svec, J. Mater. Chem., 2007, 17, e 4989. 97 J. Germain, F. Svec and J. M. J. Frchet, Chem. Mater., 2008, 20, e 7069. 98 Y.-N. Qi, H.-L. Chu, F. Xu, L.-X. Sun, Y. Zhang, J. Zhang, S.-J. Qiu and H. Yuan, Int. J. Hydrogen Energy, 2007, 32, 3395. 99 J. Deng, X. Ding, W. Zhang, Y. Peng, J. Wang, X. Long, P. Li and A. S. C. Chan, Eur. Polym. J., 2002, 38, 2497. 100 M. U. Jusczyk, A. Kumar, S. Srinivasan and E. Stefanakos, Int. J. Hydrogen Energy, 2007, 32, 1010. 101 S. Pang, G. Li and Z. Zhang, Macromol. Rapid Commun., 2005, 26, 1262. 102 B. H. Kim, W. G. Hong, S. M. Lee, Y. J. Yun, H. Y. Yu, S.-Y. Oh, C. H. Kim, Y. Y. Kim and H. J. Kim, Int. J. Hydrogen Energy, 2010, 35, 1300. 103 Q. Sun and Y. Deng, Eur. Polym. J., 2008, 44, 3402. 104 B. Liu, L. Liu, N. Shi, J. Gong and C. Sun, J. Mater. Sci. Technol., 2010, 26, 39. 105 S. Machida, S. Miyata and A. Techagumpuch, Synth. Met., 1989, 31, 311. 106 A. Ramanaviius, A. Ramanaviiene and A. Malinauskas, c c Electrochim. Acta, 2006, 51, 6025. 107 J. Germain, J. M. J. Frchet and F. Svec, Chem. Commun., 2009, e 1526. 108 M. H. Abdel Rehim, N. Ismail, A. El-R. A. A. Badaway and G. Turky, Mater. Sci. Eng., B, 2011, 176, 1184. 109 V. A. Davankov and M. P. Tsyurupa, React. Polym., 1990, 13, 27. 110 J.-Y. Lee, C. Wood, D. Bradshaw, M. J. Rosseinsky and A. I. Cooper, Chem. Commun., 2006, 2670. 111 S. Figueroa-Gerstenmaier, C. Daniel, G. Milano, G. Guerra, O. Zavorotynska, J. G. Vitillo, A. Zecchina and G. Spoto, Phys. Chem. Chem. Phys., 2010, 12, 5369. 112 J. Germain, J. Hradil, J. M. J. Frchet and F. Svec, Chem. Mater., e 2006, 18, 4430. 113 A. W. C. van den Berg and C. O. Arean, Chem. Commun., 2008, 668. 114 N. B. McKeown, B. Gahnem, K. J. Msayib, P. M. Budd, C. F. Tattershall, K. Mahmood, S. Tan, D. Book, H. W. Langmi and A. Walton, Angew. Chem., 2006, 118, 1836.

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

This journal is ? The Royal Society of Chemistry 2012

Energy Environ. Sci., 2012, 5, 5951–5979 | 5977

View Article Online
115 P. M. Budd, B. S. Ghanem, S. Makhseed, N. B. McKeown, K. J. Msayib and C. E. Tattershall, Chem. Commun., 2004, 230. 116 N. B. McKeown and P. M. Budd, Chem. Soc. Rev., 2006, 35, 675. 117 P. M. Budd, N. B. McKeown and D. Fritsch, J. Mater. Chem., 2005, 15, 1977. 118 P. M. Budd, A. Butler, J. Selbie, K. Mahmood, N. B. McKeown, B. Ghanem, K. Msayib, D. Book and A. Walton, Phys. Chem. Chem. Phys., 2007, 9, 1802. 119 P. M. Budd, N. B. McKeown, B. S. Ghanem, K. J. Msayib, D. Fritsch, L. Starannikova, N. Belov, O. San?rova, Y. Yampolskii and V. Shantarovich, J. Membr. Sci., 2008, 325, 851. 120 M. U. Niemann, S. S. Srinivasan, A. R. Phani, A. Kumar, D. Y. Goswami and E. K. Stefanakos, J. Nanomater., 2008, 950967. 121 T. A. Strobel, K. C. Hester, C. A. Koh, A. K. Sum and E. D. Sloan Jr., Chem. Phys. Lett., 2009, 478, 97. 122 W. L. Mao, H. K. Mao, A. F. Goncharov, V. V. Struzhkin, Q. Guo, J. Hu, J. Shu, R. J. Hemley, M. Somayazulu and Y. Zhao, Science, 2002, 297, 2247. 123 M. H. Sluiter, H. Adachi, R. V. Belosludov, V. R. Belosludov and Y. Kawazoe, Mater. Trans., 2004, 45(5), 1452. 124 L. J. Florusse, C. J. Peters, J. Schoonman, K. C. Hester, C. A. Koh, S. F. Dec, K. N. Marsh and E. D. Sloan, Science, 2004, 306, 469. 125 T. A. Strobel, C. J. Taylor, K. C. Hester, S. F. Dec, C. A. Koh, K. T. Miller and E. D. Sloan Jr, J. Phys. Chem. B, 2006, 110, 17121. 126 T. A. Strobel, C. A. Koh and E. D. Sloan, Fluid Phase Equilib., 2007, 261, 382. 127 T. Sugahara, J. C. Haag, A. A. Warntjes, P. S. R. Prasad, E. D. Sloan, C. A. Koh and A. K. Sum, J. Phys. Chem. C, 2010, 114, 15218. 128 F. Su, C. L. Bray, B. Tan and A. I. Cooper, Adv. Mater., 2008, 20, 2663. 129 P. K. Chattaraj, S. Bandaru and S. Mondal, J. Phys. Chem. A, 2011, 115, 187. 130 V. V. Struzhkin, B. Militzer, W. L. Mao, H. K. Mao and R. J. Hemley, Chem. Rev., 2007, 107, 4133. 131 Y. H. Hu and E. Ruckenstein, Angew. Chem., Int. Ed., 2006, 45, 2011. 132 P. Wang, S. Orimo, T. Matsushima, H. Fujii and G. Majer, Appl. Phys. Lett., 2002, 80, 318. 133 R. Ma, Y. Bando, H. Zhu, T. Sato, C. Xu and D. Wu, J. Am. Chem. Soc., 2002, 124, 7672. 134 R. Ma, Y. Bando, T. Sato, D. Golberg, H. Zhu, C. Xu and D. Wu, Appl. Phys. Lett., 2002, 81, 5225. 135 C. Tang, Y. Bando, X. Ding, S. Qi and D. Golberg, J. Am. Chem. Soc., 2002, 124, 14550. 136 T. Oku and M. Kuno, Diamond Relat. Mater., 2003, 12, 840. 137 T. Oku, M. Kuno and I. Narita, J. Phys. Chem. Solids, 2004, 65, 549. 138 T. Terao, Y. Bando, M. Mitome, K. Kurashima, C. Y. Zhi, C. C. Tang and D. Goldberg, Phys. E, 2008, 40, 2551. 139 A. L. M. Reddy, A. E. Tanur and G. C. Walker, Int. J. Hydrogen Energy, 2010, 35, 4138. 140 I. Narita and T. Oku, Diamond Relat. Mater., 2002, 11, 945. 141 X. Wu, J. Yang, J. G. Hou and Q. Zhu, J. Chem. Phys., 2004, 121, 8481. 142 X. Wu, J. Yang, J. Hou and Q. Zhu, Phys. Rev. B, 2004, 69(15), 1. 143 S.-H. Jhi and Y.-K. Kwon, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 69(24), 1. 144 R. J. Baierle, P. Piquini, T. M. Schmidt and A. Fazzio, J. Phys. Chem. B, 2006, 110, 21184. 145 L. Zhang, P. Wu and M. Sullivan, J. Phys. Chem. C, 2011, 115, 4289. 146 J. Cheng, L. Zhang, R. Ding, Z. Ding, X. Wang and Z. Wang, Int. J. Hydrogen Energy, 2007, 32, 3402. 147 J. Cheng, R. Ding, Y. Liu, Z. Ding and L. Zhang, Comput. Mater. Sci., 2007, 40, 341. 148 G. Mpourmpakis and G. E. Froudakis, Catal. Today, 2007, 120, 341. 149 K. Ozdoan and S. Berber, Int. J. Hydrogen Energy, 2009, 34, 5213. g 150 R. Tenne, L. Margulis, M. Genut and G. Hodes, Nature, 1992, 360, 444. 151 V. Lavayen, N. Mirabal, C. O’Dwyer, M. A. Santa Ana, E. Benavente, C. M. Sotomayor Torres and G. Gonzlez, Appl. a Surf. Sci., 2007, 253, 5185. 152 K. Du, W. Fu, R. Wei, H. Yang, S. Liu, S. Yu and G. Zou, Mater. Lett., 2007, 61, 4887. 153 F. L. Deepak, A. Mayoral and M. J. Yacaman, Mater. Chem. Phys., 2009, 118, 392. 154 J. Chen and F. Wu, Appl. Phys. A: Mater. Sci. Process., 2004, 78, 989. 155 J. Chen, N. Kuriyama, H. Yuan, H. T. Takeshita and T. Sakai, J. Am. Chem. Soc., 2001, 123, 11813. 156 J. Chen, S. L. Li and Z. L. Tao, J. Alloys Compd., 2003, 356–357, 413. 157 J. A. Spirko, M. L. Neiman, A. M. Oelker and K. Klier, Surf. Sci., 2004, 572, 191. 158 A. N. Enyashin and A. L. Ivanovsk Semiconductors, 2007, 41, 81. i, 159 J. Chen, S.-L. Li, Z.-L. Tao and F. Gao, Chem. Commun., 2003, 980. 160 J. Chen, S.-L. Li, Z.-L. Tao, Y.-T. Shen and C.-X. Cui, J. Am. Chem. Soc., 2003, 125, 5284. 161 M. Hua, T. Zhankliang, G. Feng, C. Jun and Y. Huatang, Front. Chem. China, 2006, 3, 260. 162 V. V. Ivanovskaya and G. Seifert, Solid State Commun., 2004, 130, 175. 163 V. V. Ivanovskaya, G. Seifert and A. L. Ivanovskii, Semiconductors, 2005, 39, 1058. 164 A. N. Enyashin and A. L. Ivanovskii, Inorg. Mater., 2005, 41, 1118. 165 C. M. Teh and A. R. Mohammed, J. Alloys Compd., 2011, 509, 1648. 166 G. Armstrong, A. R. Armstrong, J. Canales and P. G. Bruce, Chem. Commun., 2005, 2454. 167 G. Mogilevsky, Q. Chen, A. Kleinhammes and Y. Wu, Chem. Phys. Lett., 2008, 460, 517. 168 P. Pillai, K. S. Raja and M. Misra, J. Power Sources, 2006, 161, 524. 169 J. Wang, L. Wang, L. Ma, J. Zhao, B. Wang and G. Wang, Phys. E, 2009, 41, 838. 170 Q. Wan, C. L. Lin, X. B. Yu and T. H. Wang, Appl. Phys. Lett., 2004, 84, 124. 171 H. Pan, J. Luo, H. San, Y. Feng, C. Poh and J. Lin, Nanotechnology, 2006, 17, 2963. 172 P. Li, Y. Wei, H. Liu and X. Wang, Chem. Commun., 2004, 2856. 173 D. Scarano, S. Bertarione, F. Cesano, J. G. Vitillo and A. Zecchina, Catal. Today, 2006, 116, 433. 174 V. Berube, G. Chen and M. S. Dresslhaus, Int. J. Hydrogen Energy, 2008, 33, 4122. 175 R. Ren, A. L. Ortiz, T. Markmaitree, W. Osborn and L. L. Shaw, J. Phys. Chem. B, 2006, 110, 10567. 176 A. L. Ortiz, W. Osborn, T. Markmaitree and L. L. Shaw, J. Alloys Compd., 2008, 454, 297. 177 B. Vigeholm, J. Kj?ller and B. Larsen, J. Less Common Met., 1980, 74, 341. 178 B. Vigeholm, J. Kj?ller, B. Larsen and A. Schroder Pedersen, Int. J. Hydrogen Energy, 1983, 8, 809. 179 B. Vigeholm, J. Kj?ller, B. Larsen and A. Schroder Pedersen, J. Less Common Met., 1983, 89, 135. 180 P. J. Herley, W. Jones and B. Vigeholm, J. Appl. Phys., 1985, 58, 292. 181 S. Cheung, W.-Q. Deng, A. C. T. van Duin and W. A. Goddard III, J. Phys. Chem. A, 2005, 109, 851. 182 J. R. Ares, K.-F. Aguey-Zinsou, T. Klassen and R. Bormann, J. Alloys Compd., 2007, 434–435, 729. 183 N. Hanada, T. Ichikawa, S.-I. Orimo and H. Fujii, J. Alloys Compd., 2004, 366, 269. 184 J. Hout, G. Liang, S. Boily, A. Van Neste and R. Schulz, J. Alloys Compd., 1999, 293–295, 495. 185 A. San-Martin and F. D. Manchester, Bull. Alloy Phase Diagrams, 1987, 8, 431. 186 H. G. Schimmel, M. R. Johnson, G. J. Kearley, A. J. RamirezCuesta, J. Huot and F. M. Mulder, J. Alloys Compd., 2005, 393, 1. 187 W. Li, C. Li, H. Ma and J. Chen, J. Am. Chem. Soc., 2007, 129, 6710. 188 L. Li, B. Peng, W. Ji and J. Chen, J. Phys. Chem. C, 2009, 113, 3007. 189 I. Saita, T. Toshima, S. Tanda and T. Akiyama, J. Alloys Compd., 2007, 446–447, 80. 190 C. Zhu, H. Hayashi and I. Saita, Int. J. Hydrogen Energy, 2009, 34, 7283. 191 C. Zhu, S. Hosokai, I. Matsumoto and T. Akiyama, Cryst. Growth Des., 2010, 10, 5123. 192 C. Zhu, S. Hosokai and T. Akiyama, Cryst. Growth Des., 2011, 11, 4166. 193 C. Zhu, N. Sakaguchi, S. Hosokai, S. Watanabe and T. Akiyama, Int. J. Hydrogen Energy, 2011, 36, 3600. 194 B. J. Kooi, G. Palasantzas and J. T. M. De Hosson, Appl. Phys. Lett., 2006, 89, 161914. 195 G. Krishnan, B. J. Kooi, G. Palasantzas, Y. Pivak and B. Dam, J. Appl. Phys., 2010, 107, 053504.

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

5978 | Energy Environ. Sci., 2012, 5, 5951–5979

This journal is ? The Royal Society of Chemistry 2012

View Article Online
196 G. Yang, Prog. Mater. Sci., 2007, 52, 648. 197 T. Phuoc, B. Howard, D. Martello, Y. Soong and M. K. Chyu, Opt. Lasers Eng., 2008, 46, 829. 198 M.-R. Song, M. Chen and Z.-J. Zhang, Mater. Charact., 2008, 59, 514. 199 K.-F. Aguey-Zinsou and J.-R. Ares-Fernndez, Chem. Mater., 2008, a 20, 376. 200 S. B. Kalidindi and B. R. Jagirdar, Inorg. Chem., 2009, 48, 4524. 201 K. J. Klabunde and Y. Tanaka, J. Mol. Catal., 1983, 21, 57. 202 B. Bogdanovi, K. Bohmhammel, B. Christ, A. Reiser, K. Schlichte, c R. Vehlen and U. Wolf, J. Alloys Compd., 1999, 282, 84. 203 W. Oelerich, T. Klassen and R. Bormann, Adv. Eng. Mater., 2001, 3, 487. 204 W. Oelerich, T. Klassen and R. Bormann, J. Alloys Compd., 2001, 315, 237. 205 N. Hanada, T. Ichikawa and H. Fujii, J. Phys. Chem. B, 2005, 109, 7188. 206 N. Bazzanella, R. Checchetto and A. Miotello, J. Nanomaterials, 2011, 865969. 207 N. Hanada, E. Hirotoshi, T. Ichikawa, E. Akiba and H. Fujii, J. Alloys Compd., 2008, 450, 395. 208 N. Hanada, T. Ichikawa, S. Hino and H. Fujii, J. Alloys Compd., 2006, 420, 46. 209 A.-H. Lu and F. Sch th, Adv. Mater., 2006, 18, 1793. u 210 J. J. Vajo, Curr. Opin. Solid State Mater. Sci., 2011, 15, 52. 211 P. E. de Jongh, R. W. P. Wagemans, T. E. Eggenhuisen, B. S. Dauvillier, P. B. Radstake, J. D. Meeldijk, J. W. Geus and K. P. de Jongh, Chem. Mater., 2007, 19, 6052. 212 S. Zhang, A. F. Gross, S. L. Van Atta, M. Lopez, P. Liu, C. C. Ahn, J. J. Vajo and C. M. Jensen, Nanotechnology, 2009, 20, 204027. 213 K.-J. Jeon, H. R. Moon, A. M. Ruminski, B. Jiang, C. Kisielowski, R. Bardhan and J. J. Urban, Nat. Mater., 2011, 10, 286. 214 J. Graetz, J. J. Reilly, V. A. Yartys, J. P. Maelen, B. M. Bulychev, V. E. Antonov, B. P. Tarasov and I. E. Gabis, J. Alloys Compd., 2011, 509, S517. 215 A. Zaluska, L. Zaluski and J. O. Str m-Olsen, J. Alloys Compd., o 2000, 298, 125. 216 K. J. Gross, G. J. Thomas and C. M. Jensen, J. Alloys Compd., 2002, 330–332, 683. 217 M. Mamatha, B. Bogdanovi, Felderhoff, A. Pommerin, c W. Schmidt, F. Sch th and C. Weidenthaler, J. Alloys Compd., u 2006, 407, 78. 218 J. Graetz, J. J. Reilly, J. Johnson, A. Y. Ignatov and T. A. Tyson, Appl. Phys. Lett., 2004, 85, 500. 219 M. Fichtner, P. Canton, O. Kircher and A. Lon, J. Alloys Compd., e 2005, 404–406, 732. 220 T. Vegge, Phys. Chem. Chem. Phys., 2006, 8, 4853. 221 D. Pukazhselvan, M. S. L. Hudson, A. S. K. Sinha and O. N. Srivastava, Energy, 2010, 35, 5037. 222 M. Fichtner, O. Fuhr, O. Kircher and J. Rothe, Nanotechnology, 2003, 14, 778. 223 M. Fichtner, J. Engel, O. Fuhr, O. Kircher and O. Rubner, Mater. Sci. Eng., B, 2004, 108, 42. 224 R. A. Varin, C. Chiu, T. Czujko and Z. Wronski, J. Alloys Compd., 2007, 439, 302. 225 R. A. Varin, C. Chiu, T. Czujko and Z. Wronski, Nanotechnology, 2005, 16, 2261. 226 C. Bald, B. P. C. Hereijgers, J. H. Bitter and K. P. de Jong, J. Am. e Chem. Soc., 2008, 130, 6761. 227 R. D. Stephens, A. F. Gross, S. L. Van Atta, J. J. Vajo and F. E. Pinkerton, Nanotechnology, 2009, 20, 204018. 228 J. J. Vajo and S. L. Skeith, J. Phys. Chem. B, 2005, 109, 3719. 229 J. R. Hattrick-Simpers, J. E. Maslar, M. U. Niemann, C. Chiu, Sesha. S. Srinivasan, E. K. Stefanakos and L. A. Bendersky, Int. J. Hydrogen Energy, 2010, 35, 6323. 230 R. A. Varin and L. Zbroniec, Int. J. Hydrogen Energy, 2010, 35, 3588. 231 S. Sartori, K. D. Knudsen, Z. Zhao-Karger, E. G. Bardaji, J. Muller, M. Fichtner and B. C. Hauback, J. Phys. Chem. C, 2010, 114, 18785. 232 M. Fichtner, Z. Zhao-Karger, J. Hu, A. Roth and P. Weidler, Nanotechnology, 2009, 20, 204029. 233 T. K. Nielsen, U. B senberg, R. Gosalawit, M. Dornheim, o Y. Cerenius, F. Besenbacher and T. R. Jensen, ACS Nano, 2010, 4, 3903. 234 A. Ampoumogli, T. Steriotis, P. Trikalitis, D. Giasafaki, E. G. Bardaji, M. Fichtner and G. Charalambopoulou, J. Alloys Compd., 2011, 509, S705. 235 A. F. Gross, J. J. Vajo, S. L. Van Atta and G. L. Olson, J. Phys. Chem. C, 2008, 112, 5651. 236 X. Liu, D. Peaslee, C. Z. Jost and E. H. Majzoub, J. Phys. Chem. C, 2010, 114, 14036. 237 P. Chen, Z. Xiong, J. Luo, J. Lin and K. L. Tan, Nature, 2002, 420, 302. 238 D. H. Gregory, J. Mater. Chem., 2008, 18, 2321. 239 P. Chen and M. Zhu, Mater. Today, 2008, 11, 36. 240 F. E. Pinkerton, J. Alloys Compd., 2005, 400, 76. 241 T. Markmaitree, R. Ren and L. L. Shaw, J. Phys. Chem. B, 2006, 110, 20710. 242 C. Lu, J. Hu, J. Kwak, Z. Yang, R. Ren, T. Markmaitree and L. Shaw, J. Power Sources, 2007, 419–424, 170. 243 R. A. Varin, M. Jang and M. Polanski, J. Alloys Compd., 2010, 491, 658. 244 F. E. Pinkerton, G. P. Meisner, M. S. Meyer, M. P. Balogh and M. D. Kundrat, J. Phys. Chem. B, 2005, 109, 6. 245 H. Wu, W. Zhou, K. Wang, T. J. Udovic, J. J. Rush, T. Yildirim, L. A. Bendersky, A. F. Gross, S. L. Van Atta, J. J. Vajo, F. E. Pinkerton and M. S. Meyer, Nanotechnology, 2009, 20, 204002. 246 T. Ichikawa, I. Shigehito, N. Hanada and H. Fujii, J. Alloys Compd., 2004, 365, 271. 247 S. Isobe, T. Ichikawa, N. Hanada, H. Y. Leng, M. Fichtner, O. Fuhr and H. Fujii, J. Alloys Compd., 2005, 404–406, 439. 248 F. Pinkerton, M. Meyer, G. Meisner and M. Balogh, J. Alloys Compd., 2007, 433, 282. 249 Y. Kojima, M. Matsumoto, Y. Kawai, T. Haga, N. Ohba, K. Miwa, S. -ichi Towata, Y. Nakamori and S. -ichi Orimo, J. Phys. Chem. B, 2006, 110, 9632. 250 L. Xie, J. Zheng, Y. Liu, Y. Li and X. Li, Chem. Mater., 2008, 20, 282. 251 L. Xie, Y. Li, R. Yang, Y. Liu and X. Li, Appl. Phys. Lett., 2008, 92, 231910. 252 L. Xie, Y. Liu, G. Li and X. Li, J. Phys. Chem. C, 2009, 113, 14523. 253 F. H. Stephens, V. Pons and R. T. Baker, Dalton Trans., 2007, 2613. 254 B. Peng and J. Chen, Energy Environ. Sci., 2008, 1, 479. 255 T. B. Marder, Angew. Chem., Int. Ed., 2007, 46, 8116. 256 P. Wang and X. D. Kang, Dalton Trans., 2008, 5400. 257 Y. S. Chen, J. L. Fulton, J. C. Linehan and T. Autrey, J. Am. Chem. Soc., 2005, 127, 3254. 258 A. Gutowska, L. Li, Y. Shin, C. M. Wang, X. S. Li, J. C. Linehan, R. S. Smith, B. D. Kay, B. Schmid, W. Shaw, M. Gutowski and T. Autrey, Angew. Chem., Int. Ed., 2005, 44, 3578. 259 A. Feaver, S. Sepehri, P. Shamberger, A. Stowe, T. Autrey and G. Cao, J. Phys. Chem. B, 2007, 111, 7469. 260 Z. Kurban, A. Lovell, S. M. Bennington, D. W. K. Jenkins, K. R. Ryan, M. O. Jones, N. T. Skipper and W. I. F. David, J. Phys. Chem. C, 2010, 114, 21201. 261 R. Checchetto, G. Carotenuto, N. Bazzanella and A. Miotello, J. Phys. D: Appl. Phys., 2007, 40, 4043.

Published on 26 January 2012. Downloaded by Nanjing University of Technology on 05/09/2013 04:07:08.

This journal is ? The Royal Society of Chemistry 2012

Energy Environ. Sci., 2012, 5, 5951–5979 | 5979


相关文章:
网络检索整理3
years in the emerging field of molecular magnetism... Basic Concepts and Phenomena - 801.4 Physical ...solid state physicists, an increased effort in ...
集成电路领域会议杂志
IEEE Journal of Solid-State Circuits(JSSC) IEEE...Emerging Technologies ? Programmable Devices ? ...Future methodology needs and concepts ?Impact of ...
Top 10 Emerging Technologies of 2015
Top 10 Emerging Technologies of 2015_互联网_IT/计算机_专业资料。TECH Top ...Innovative hydrogen storage techniques, such as organic liquid carriers that ...
Integrating emerging topics through online team design in a hybrid communication networks course
Integrating emerging topics through online team ...Key Internet safety concepts and educators' related...the 1980s to today's state-of-the-art CSCD....
更多相关标签:
hydrogen storage | solid state storage | hydrogen | concepts | netconcepts china | hydrogen peroxide | netconcepts | hydrogen bonding |