Hydrogen Storage Employing Select, Main-Group-Based Inorganic Materials

Abstract

The use of hydrogen as a fuel is considered a serious option to reduce the long-term environmental impact of global warming. A key challenge of using hydrogen as a fuel is that of employing safe and effective methods by which to store it. One general approach to addressing this challenge is to utilize chemical compounds that release hydrogen gas under highly specified and controlled chemical conditions. This review will discuss said compounds which contain selected main-group inorganic elements, including certain (1) Alkaline-based metals (Li, Na, K, Mg, Ca), (2) Boron and Aluminum, and (3) Silicon. The majority of these compounds release hydrogen gas under mild conditions, typically by hydrolysis. The performance criteria of these compounds will be compared along with commentary on the topics of (1) Synthesis of these materials, (2) Energy requirements, (3) Hydrogen release chemistry, (4) Handling safety, and (5) The challenges of recycling/ reloading these materials.

1. Introduction

The concept of global warming due to the atmospheric build-up of carbon dioxide via the combustion of fossil fuels is firmly established in the scientific community and other spheres that strongly influence energy policy-making [1]. One of the most intensely investigated alternatives to fossil fuels is that of molecular hydrogen, H₂ [1, 2, 3, 4, 5, 6].

Concerned stakeholders generally agree that a major drawback in the use of hydrogen as an energy source is the lack of methods available to effectively store it [7, 8, 9, 10, 11, 12, 13, 14, 15]. Regarding storage, although hydrogen gas has a favorably high weight energy density (120 MJ/kg at 25°C and atmospheric pressure) [16], that advantage is compromised by its unfavorably low volumetric energy density (10.05 kJ/L at 25°C and atmospheric pressure) [16]. Therefore, the current mature storage technology is H₂ gas compression at 70 MPa. (resulting in a volumetric energy density of approximately 3000 kJ/L) or else cryo-compression of the gas to a liquid, at the exceptionally low temperature of 20 deg. K. Unfortunately, these storage methods are hampered by the considerations that, under these conditions; (1) Many metals used in storage tanks become embrittled, (2) Handling either a high-pressure gas or else an exceptionally cold, cryogenic liquid¹ may not compatible with mass-market applications where untrained individuals are involved (such as hydrogen fuel cell-powered vehicles), and (3) Hydrogen is explosively flammable in air under relatively easy-to-achieve conditions in a closed environment (see Footnote #1). These afore-mentioned challenges spurred an intense effort to develop alternative methods of storing H₂.

Generally, when one considers “alternative” storage systems for H₂, the working concept is that of substituting “pure” H₂ within the storage container with a material that releases H₂ at a controlled rate under specific operational conditions. For example, one can consider sodium metal to be a material that fits these criteria. Under typical storage conditions (near room temperature and atmospheric pressure), when protected from moisture, sodium metal is sufficiently stable to be utilized in storage applications. Furthermore, when one seeks to “release” H₂ from sodium, this can be accomplished by exposing the metal to a chemical that contains what is commonly referred to as an “active” (acidic) proton. Common chemicals that contain “active” protons include water, alcohols, silanols, carboxylic acids, ammonia and certain amines, etc. The chemical reaction between sodium and, for example, water, is relatively rapid, highly exothermic and yields hydrogen gas and sodium hydroxide as products.

$$\mathrm{Na}\left(\mathrm{s}\right)+{\mathrm{H}}_2\mathrm{O}➔\mathrm{\text{NaOH}}\left(\mathrm{s}\right)+\mathrm{1/2}{\mathrm{H}}_2\left(\mathrm{g}\right)$$E1

The formation of a by-product, NaOH, hints at another common feature of many (but not all) of these types of H₂ storage systems, that being the need to remove a chemical by-product from the storage system and replace it with more storage material, in this case, sodium metal.²

Now that we have established that “alternative” H₂ storage methodologies involve the “release” or “generation” of H₂ from a defined, stored material, it is convenient to categorize the commonly referred to types of storage processes:

1. Physisorption, Storage, and Desorption (“Release”) of H₂- Reversible adsorption of H₂ on a solid or liquid where the H-H bond in H₂ is preserved throughout the entire process. The “adsorbing” material is usually not altered due to the process and can typically be reused.

2. Chemisorption, Storage, and Desorption of H₂- Reversible adsorption of H₂ on a solid or liquid where the H-H bond in H₂ is broken during the adsorption process, the “breakage” is maintained through the “storage” process, and then is reformed during the desorption (“release”) process. The “adsorbing” material may or may not altered due to the process and can, in favorable cases, be reused.

3. The storage of a material which may, or may not, contain hydrogen atoms, followed by a specifically-defined, chemical process which results in the controlled formation of H₂ gas. In the case where the stored material does not contain hydrogen atoms, then the stored material is exposed to a 2nd material which does contain the hydrogen atoms which ultimately form H₂- During this process there is typically a substantial degree of breakage and reformation of chemical bonds, highlighted by the breakage of X-H bonds (where X can be virtually any element in the periodic table) and the subsequent formation of H-H bonds. Along with the generation of H₂ gas, chemical by-products are formed that need to be ultimately removed from the storage container. An example of this process is shown in Eq. (1).

This chapter will focus on examples of Process #3 listed above. The primary reason for this focus is that materials undergoing Process # 3 potentially produce more H₂ per storage/release cycle than do the materials that undergo Processes #1 and #2 [13, 20]. Within the broad range of materials and processes that meet the definition of Process #3, this chapter will focus on storage materials where the central element X of the material is referred to as a “main group” element of the Periodic Table of the Elements (Groups # 1, 2, 13–18)³ [21]. Furthermore, within the “main group” elements, we will narrow our scope of review to those groups where, for the element X, the X-H bond is hydridic in nature. A “hydridic” X-H bond is one where the electron distribution in the bond is unbalanced with the majority of it located near the hydrogen atom. Main-group compounds with hydridic character constitute the majority of examples associated with the H₂ storage application.⁴ Therefore, this chapter will cover materials which contain chemical elements from Groups # 1, 2, 13 & 14 of the Periodic Table which also meet certain criteria.⁵ The result of subjecting the Periodic Table to these limitations leaves us with seven elements: Lithium (Li), Sodium (Na), Potassium (K), Magnesium (Mg), Boron (B), Aluminum (Al), and Silicon (Si).

2. Main group element-based hydrogen storage materials (MGE-HSMs)

2.1 General trends relevant to the H₂ storage application

2.1.1 Hydridic X-H bonds react with protonic X-H bonds to form hydrogen

Regarding the seven elements selected to be further reviewed, for the great majority of them, the materials which find applicability for H₂ storage contain one or more hydridic X-H bonds; in fact, the hydridic bond is usually the source of generation of H₂ gas. The polarity of X-H bonds within the materials covered in this chapter range from X⁺H⁻ (where the electron charge distribution is so localized at the H atom that the bond is “ionic”), through a range of X-H bond polarities where the bond has substantial covalent character.

Regarding the H₂ storage application, the key working chemistry principle is that: “Hydridic X-H bonds react with Protonic X-H bonds to form H₂”. A “protonic” X-H bond is the case where the hydrogen atom in the bond has a degree of positive charge. A relevant and highly common example is given in Eq. (2). In this case, the hydridic X-H bond exists in NaH⁶ and the protonic X-H bond exists in H₂O.

$$\mathrm{NaH}\left(\mathrm{s}\right)+{\mathrm{H}}_2\mathrm{O}➔\mathrm{\text{NaOH}}\left(\mathrm{s}\right)+{\mathrm{H}}_2\left(\mathrm{g}\right)$$E2

To extend this unifying concept even further, the polarity of X-H bonds, all else being equal, correlates to the position of an element within the Periodic Table of the Elements [23]. Regarding this correlation, one can apply the following “rules of thumb”: (1) As one descends down a Group (or alternatively a “column”) in the Table the X-H bond becomes more hydridic (therefore, the K-H bond in potassium hydride is more hydridic than the Na-H bond in sodium hydride), and (2) As one moves from right-to-left along a row in the Table, the X-H bond becomes more hydridic (therefore, the Mg-H bond in magnesium hydride is more hydridic than the B-H bond in boron hydrides).

Regarding the H₂ storage application, one can apply the “rules” above in order to make qualitative predictions about the chemistries required to yield H₂ gas under various conditions. To wit, regarding the uppermost row in The Periodic Table (Na, Mg, B, C, N, O, F), one might expect the following outcomes: (1) In order to harvest H₂ from either NaH or MgH₂, both highly hydridic, one would use a reagent with a protonic X-H bond, and a weakly protonic agent such as ammonia, NH₃, (2) The boron hydrides are less hydridic than either NaH or MgH₂; therefore, a reagent that is moderately protonic (such as water or, if necessary, acetic acid, CH₃COOH)⁷ might be needed to efficiently release H₂, (3) The C-H bond, particularly in molecules not containing many electronegative atoms (typically nitrogen or oxygen), is close to being non-polar and, therefore, is quite resistant to the reaction with either hydridic or protonic X-H bonds to form H₂,⁸ and (4) Finally, the N-H, O-H and F-H bonds are progressively more protonic, and therefore require hydridic X-H reagents, such as NaH, to release H₂ gas.

The concepts developed above can be generalized to expand the “chemistry set” available to produce H₂ gas. For example, returning to Eq. (2) above, one can also categorize this reaction as a reduction/oxidation (“redox”) process. Specifically, water can be the “oxidizing” agent, formally removing an electron from the hydride (H⁻) ion, yielding the hydrogen atom, which rapidly dimerizes to form H₂. Alternatively, one can describe this process as the reduction of water (by the hydride “reducing agent”), yielding the hydroxide (OH⁻) ion (as NaOH). The utility of carrying out this categorization is that one can conceive of a much wider range of redox chemistry which can potentially yield H₂ gas “on demand”. For example, in Eq. (3), it is problematic to consider the chemical reaction as one between a “hydridic” compound and a “protonic” compound to yield H₂; specifically there is no “hydridic” compound present as a reagent.

$$\mathrm{Na}\left(\mathrm{s}\right)+{\mathrm{H}}_2\mathrm{O}➔\mathrm{\text{NaOH}}\left(\mathrm{s}\right)+\mathrm{1/2}{\mathrm{H}}_2\left(\mathrm{g}\right)$$E3

Equation 3 is more accurately described as a “redox” reaction. Sodium metal has an extensive chemical portfolio as a strong reducing agent [21]. Within Eq. (3), sodium reduces water,⁹ yielding the reaction products including H₂ gas.

2.1.2 The terrestrial forms of MGE-HSMs

It is well-known that carbon-centered “fossil fuels” were created geochemically, over the millennia, due to the decomposition of animal and plant residues in anaerobic conditions [25]. The anaerobic conditions were necessary for the formation of their signature C-C and C-H bonds. Under aerobic conditions, in the presence of an ignition source, these bonds will rapidly react with oxygen, forming C-O and O-H bonds. These are the essential steps in the conventional energy-creating, fossil fuel-based, combustion process. In the absence of an ignition source, under conditions commonly found on the surface of the earth, fossil fuels have significant chemical stability.¹⁰ Furthermore, physical forms of the element carbon existing under terrestrial conditions,¹¹ such as graphite, are quite stable to chemical change.

Unlike the example of carbon discussed above, many of the main group elements¹² considered in this chapter, in forms relevant to the H₂ storage application (MGE-HSM), whether in compounds containing X-H bonds or else the element in its native state,¹³ are not chemically stable under terrestrial conditions. These elements are usually harvested as highly-oxygenated compounds such as the oxides, hydroxides and silicates. This is consistent with an earth atmosphere whose most reactive gases are oxygen and water. The chloride ion is also quite abundant in the earth’s crust, resulting in the presence of the chloride forms of many of these elements.

2.1.3 Storage and reaction

When one considers the MGE-HSMs within this chapter, in contrast to the case of fossil fuels where the stored fuel (for example, that in one’s gas tank) is directly combusted to yield the energy content, it is necessary to carry out a chemical process on the stored material in order to yield the combustible H₂. This is the necessary “trade-off” for utilizing a storage material that is, by choice, easier to handle than either gaseous or liquid H₂.

Therefore, when one considers utilizing these MGE-HSMs, they have two “high-level” choices: (1) Transport the MGE-HSM from the storage container to a separate reactor compartment to carry out the H₂ release chemistry, or (2) Carry out the said chemical process within the “storage” container.

2.2 Chemical reactivity patterns

2.2.1 Homogeneous (single phase) reactions

Assuming that (1) The H₂ generation process occurs within a single phase, and (2) The rate of the reaction is not reagent-diffusion-rate-controlled; then the rate of reaction between an MGE-HSM and its “activator” (reagent used to generate H₂ gas) will often be a second-order rate process with rate dependencies on the concentration of both the MGE-HSM and its activator. One result of this set of initial conditions is, if the MGE-HSM and its activator are mixed together with a mutual solvent, and none of either reagent is added to the mixture afterwards, then the rate of evolution of H₂ gas will exponentially decrease with time. Assuming that under operational conditions (such as supplying H₂ gas to a working fuel cell) one desires a constant rate of evolution of H₂ gas, one would need to implement either mechanical and/or chemical modifications to the storage/reactor/gas transfer assembly. For example, in terms of a chemical modification, one could potentially control the rate of introduction of activator into the storage/reactor in such a method that the product of the two reagent concentrations remain constant throughout the course of the reaction.

2.2.2 Heterogeneous (multiple phase) reactions

Assuming that (1) The H₂ generation process occurs in a multiple phase system, and (2) The rate of the reaction is not reagent-diffusion-rate-controlled, the rate of reaction between an MSG-HMS and it’s “activator” will, along with potential rate dependencies on the concentration of both reagents, will potentially have rate dependencies on the total interfacial area where reaction occurs. In the context of MSG-HMS chemistry, a common example of this situation is one where the MSG-HMS is being stored as a solid and the activator is introduced into the storage compartment as a liquid or a vapor. The H₂-generating reaction occurs at the surface of the solid MSG-HMS. While the rate of the reaction may depend upon the amounts of MGE-HMS and activator, it will also depend on the total surface area of the solid MGE-HMS. Furthermore, for a given amount of solid, the total surface area will be inversely proportional to the particle size (overall, smaller particles ➔ larger surface area ➔ faster reaction). Overall, this is a complex kinetic situation, and establishing a constant rate of H₂ evolution would be a non-trivial technical challenge.

2.2.3 Heat release and exchange

Another common challenge in generating H₂ gas from the “activation” of an MGE-HSM is that often the activation reaction is both rapid and highly exothermic. Although the rate of reaction can be controlled by a plethora of technical options, efficient heat removal from the storage/reaction container is necessary (if nothing else, to safely operate the engine consuming the H₂ gas) and practical utilization of the heat generated would be a highly-desired outcome.

2.2.4 Removal of non-H₂, non-volatile, reaction products

Typically, the generation of H₂ from MGE-HSM materials will yield non-volatile reaction products that will need to be removed from the storage/reaction container at the end of a fueling/combustion cycle.

3. The transformation of terrestrial forms of MGE-HSMs to useful hydrogen storage materials and their subsequent performance

A summary of the relevant MGE-HSMs is displayed in Table 1 including (1) Their terrestrial forms, (2) The chemical transformations necessary to convert them to storable materials, (3) The chemistry necessary to release H₂ gas from the stored material and, (4) Their relative abundances of their central element in the earth’s crust [42].

Element	A common terrestrial form of the element	Element transformation to H2 storage material	A common example of H2 release chemistry	RAECa
Li	LiAlSi2O6 (Spodumene)	LiAlSi2O6 ➔ LiCl ➔ Li(s) ➔ LiH	LiH + H2O ➔ LiOH + H2	7.09(10^–5)
Na	NaCl (Rock Salt)	NaCl ➔ Na(s) ➔ NaH	NaH + H2O ➔ NaOH + H2	0.08
K	KCl/ MgCl2/ 6 H2O (Carnalite)	Carnalite ➔ KCl ➔ K ➔ KH	KH + H2O ➔ KOH + H2	0.07
Mg, Ca	CaCO3/ MgCO3 (Dolomite)	Dolomite ➔ MgO/CaO ➔ Mg, Ca➔ MgH2, CaH2	CaH2 + H2O ➔ Ca(OH)2 + H2	0.08 (Mg); 0.15 (Ca)
B	B2O3 (Boric oxide), B(OH)3 (Boric acid) and their various polymeric forms	Boric oxide/ Boric acid ➔ B(OMe)3 ➔ NaBH4; Borate/Boric acid ➔ BF3; NaBH4 + BF3 ➔ B2H6	NaBH4 + (2 + x) H2O ➔ 4 H2 + NaBO2·x H2O.; B2H6 + 3 H2O ➔ 2 B(OH)3 + 3 H2	3.55(10^–5)
Al	Al(OH)3 (Gibbsite-one component of Bauxite)	Al(OH)3 ➔ Al2O3 ➔Al ➔ NaAlH4 ➔ LiAlH4	Li[AlH 4] + 4 H2O ➔ LiOH + Al(OH)3 + 4 H2	0.29
Si	SiO2 (Silica)	SiO2 ➔ Si➔SixHy; SiO2 ➔ SiH4	SiH4 + O2 ➔ SiO2 + 2 H2	1

Table 1.

Description of the processes necessary to convert MGE-HSMs into their storage forms and the associated H₂ generation chemical reactions.

^a

RAEC is an acronym for “Relative Abundance in the Earth’s Crust”.

Ref. Groups I and II elements—[29, 30].

Ref. Boron—[31, 32, 33].

Ref. Aluminum—[34, 35, 36, 37].

Ref. Silicon—[26, 27, 28, 38, 39, 40, 41].

As displayed in Table 1, one general result arising from the harvesting of the terrestrial forms of these specific elements is that often a multiple series of chemical transformations is necessary to yield a useful HSM. Furthermore, as a trend, the chemical reactions necessary to yield the HSM tend to be substantially endothermic with high activation energies [21, 23].

3.1 Group I and II elements (Li, Na, K, Mg, Ca)

The MSG-HSMs of Groups I and II of the Periodic Table of the Elements share the characteristic of containing a “hydridic” ionic bond, M⁺H⁻. Therefore, based on the concepts discussed in Section IIIa, the following “protonic source reactivity” scale can be asserted:

MgH₂ < CaH₂, LiH < NaH < KH

For the three Group I elements, the associated hydride is produced from the native state of the element which, itself, is produced from the chloride salt. For Na and K, the chloride salt is the terrestrial form of the element. For Li, the chloride salt is synthesized from the terrestrial mineral Spodumene (LiAlSi₂O₆). The hydride salts of all of the Group I elements are crystalline solids at room temperature and are highly reactive, yielding substantial exothermicity, with water. Therefore, with regards to safety, for the H₂ storage application, they need to reside in an H₂O-free, ventable, storage device.

When assessing the effectiveness of an MGE-HSM, it is also useful to consider the terrestrial abundance of that specific element and the on-going efforts being made to harvest it. The relative abundances of these three Group I elements are displayed in Table 1. Both Na and K have substantial terrestrial abundances while Li is significantly less. All three of these elements are part of materials that are utilized in large and commercially important applications not directly related to the use of hydrogen as a fuel; therefore, substantial infrastructure exists to harvest these elements in some form. In particular, despite its low natural abundance, during the last few decades there have been expansive worldwide efforts to harvest terrestrial lithium. This effort is being driven by large commercial applications, in particular, the fabrication of lithium-ion batteries.

The situations for magnesium and calcium mirror those in place for the Group I MGE-HSMs.

Overall, the main advantages of utilizing Group I and II MSG-HSMs includes: (1) With the exception of lithium, the terrestrial forms of the elements are in relatively high abundance and, including for lithium, substantial infrastructure is currently in place to harvest them, (2) The hydride compounds of these elements are very reactive to protonic reagents; therefore, many options exist to control the H₂ generation processes, and (3) The by-products of the H₂ generation process, the various hydroxide salts, are all chemicals that have an extensive industrial processing history. The disadvantages of using these MGE-HSMs are (1) Their production requires multi-step, highly endothermic processes regarding substantial energy input and often releasing unwanted CO₂ into the environment, (2) In the presence of water they are unstable, even explosive, which is very problematic, particularly when in the presence of untrained personnel.

3.2 Group XIII and XIV elements (B, Al, Si)

For all three of the above elements, substantial infrastructure exists to harvest their terrestrial compound forms. In common with the Group I and II MGE-HSMs, the M-H bonds in boron (B), aluminum (Al) and silicon (Si) MGE-HSMs are hydridic. Therefore, these M-H bonds exhibit substantial reactivity with protonic materials, yielding H₂ gas. However, as expected from an analysis of the Periodic Table, these M-H bonds are not as polar as those of the Groups I/II, and therefore are not as reactive with protonic reagents.

3.2.1 Boron

As displayed in Table 1, the terrestrial forms of boron include boric oxide (B₂O₃), boric acid [B(OH)₃] and various polymeric forms of these compounds. With regards to the H₂ storage application, it is the hydrides of boron which are the materials of interest. Unlike the Group I and II hydrides, which are essentially encompassed by the ionic hydrides M-H, the boron hydrides (boranes) and their relevant derivatives have a rich structural complexity [43, 44]. For the boranes, electron-deficient bonding leads to exceptional reactivity for the lower boranes (di-, tetra- and pentaborane), including their spontaneous inflammability in air. This reactivity is diminished as either (1) The MW of the borane increases (along with degree of delocalized bonding), or else (2) Base adducts are formed, relieving the degree of electron-deficient bonding (ammonia-boranes, borohydrides, etc.).

Of all the elements discussed in this chapter, MGE-HSMs based on boron have received the most R&D attention. For example, in the 2010–2015-time frame, the US Department of Energy (US-DOE) funded a substantial program to identify the best candidate HSMs, with an emphases on (a) Fuel cell/vehicle transport applications, and (b) The performance metric of direct hydrogenation of the “spent” fuel. The materials screened numbered well into the hundreds [9]. One result of this program was the selection of a boron-containing MGE-HMS, ammonia-borane (H₃B-NH₃) as the top candidate for the H₂ storage application. This selection was the first step in a detailed and comprehensive investigation of this material as an MGE-HSM, with over 300 publications created. Demirci recently published an overview and critique of this large body of work [45].

Ammonia-borane can be classified as a base adduct of “borane” (BH₃), where the “base” is ammonia, NH₃, and the chemical bond between nitrogen and boron is the result of the donation of the lone pair of electrons on the nitrogen atom of ammonia to boron. Base adducts of the boranes are very common and, in the case where the “base” is the hydride ion, H-, the resultant adduct is termed a “borohydride”, BH₄⁻. Borohydride salts, particularly lithium, sodium and potassium, have been intensively investigated in the H₂ storage application, in part due to, in analogy with the case for H₃B-NH₃, their (1) Relatively high hydrogen densities, and (2) Stability in air [20, 33].

With regard to the potential reversible hydrogenation/dehydrogenation of boron-containing materials, Severa and co-workers demonstrated the direct hydrogenation of magnesium boride (MgB₂), to magnesium borohydride [Mg(BH₄)₂], under 950 bar H₂ at 400 deg. C. [46].

To date, to the authors’ knowledge, despite all of this R&D work, no boron-centered MSG-HMS have yet been commercialized for the H₂ storage application. The authors conclude that there are two reasons for that: (1) In agreement with Demirci [45], not enough work has been carried out investigating the processes needed to scale-up the most promising materials to the production environment, and (2) The threat of the unintended creation and release of explosive, low molecular weight boranes into the air.

3.2.2 Aluminum

With aluminum (Al) positioned just below boron in the Periodic Table, one would expect, all else equal, the Al-H bond to be more ionic, and therefore, more hydridic, than the B-H bond. This should, and does, translate into a higher reactivity with protonic sources, such as water [47].

A number of aluminum hydrides have been investigated as HSMs [20, 37, 48]. The most popular ones for study, in part because they are so common, are NaAlH₄ and LiAlH₄. As can be seen in Table 1, as HSMs, they can release H₂ via hydrolysis.

Varying degrees of success were achieved in efforts to demonstrate reversible thermal hydrogenation of NaAlH₄ and LiAlH₄, particularly when the materials were either nanoparticulates, or under the condition of nanoconfinement. First, it was demonstrated that the H₂ storage properties of these aluminum hydrides could be altered by microstructural refinement (particle and grain size) via the process of ball milling [49]. Furthermore, reversible hydrogenation of either NaAlH₄ or LiAlH₄ was demonstrated when: (1) Fullerene-C₆₀ was used as a catalyst [50], (2) NaAlH₄ was confined within the nanoporous structure of titanium-loaded, highly-ordered, mesoporous carbon [51]. (3) Carbon nanotubes and graphitic nanofibers were used as catalysts [52]. (4) When a nanoparticulate-complex hydride mixture, stemming from the ad-mixing of LiAlH₄ and nano-MgH₂, was used as an MGE-HSM [53].

Outside of the methodology of using aluminum-containing compounds as MGE-HSMs, Omran and co-workers demonstrated the release of H₂ when aluminum chips¹⁴ were exposed to aqueous base solutions [54].

3.2.3 Silicon (cast form)

Terrestrial silicon consists of its oxide forms including silica (SiO₂) and the silicates. Silica is chemically transformed, on a production scale, into its native state, silicon metal, by its carbothermic reaction with charcoal at around 2500 deg. C [26, 27, 28]. This reaction is displayed in Eq. (4).

$${\mathrm{SiO}}_2+2\mathrm{C}➔\mathrm{Si}+\mathrm{CO}+\mathrm{1/2}{\mathrm{CO}}_2$$E4

Regarding the H₂ storage application, the chemical processes necessary to form Si-H bonds also need to be considered. In analogy to Eq. (4), silica will react with H₂ at high temperature (1400–1600 deg. C), and the silicon atom will go through a sequential reductive process, first yielding SiO, then Si, then compounds containing Si-H bonds. The by-product of these reactions is water, H₂O [40, 41].

3.2.4 Silicon (nanoparticulate)

Nanosilicon, due to its much larger surface area, is substantially more reactive than the cast form. If prepared correctly (yielding small enough particles and a surface free of oxidation), hydrogen is generated by the direct, heterogeneous reaction with water [55] as displayed in Eq. (5).

$$\mathrm{Si}\left(\mathrm{\text{nano}}\right)+2{\mathrm{H}}_2\mathrm{O}➔{\mathrm{SiO}}_2+2{\mathrm{H}}_2$$E5

Erogbogdo and co-workers [56] reported that the rate of H₂ generation from the hydrolysis of 10 nm silicon particles was 1000 times the rate when one used bulk (cast) silicon. A Si-containing material, specifically that of the “alkali metals plus silica gel” composites¹⁵ developed by Dye and co-workers [57] releases H₂ gas when exposed to water under mild conditions. These composites have been commercialized as a H₂ fuel source and are marketed under the trade name “Active Fuel”.

Methods demonstrated to produce water-reactive, H₂-generating, silicon-containing materials include: (1) The ball milling of mixtures; (a) The combination of silicon-containing precursors (such as SiCl₄) and various reducing agents (such as lithium metal) [58], (b) The combination of carbon (graphite) and SiO₂ [59] and (c) The combination of SiO₂ with aluminum [60], (2) The elevated temperature treatment of rice husk with aqueous sodium hydroxide such as the work of Bose et al. [61], and Liu et al. [62], and (3) The laser ablation of silicon wafers [63].

Finally, a process to produce commercial quantities of nanosilicon has been recently report by the firm PyroGenesis of Montreal, Canada [64]. PyroGenesis developed the process for their PUREVAP™ Nano Silicon reactor.

3.2.5 Silanes and polymers featuring a “backbone” containing silicon atoms

Si-H bond containing compounds including (1) Silanes, with the generic molecular structure SiX_aY_b and, specifically, the examples where the X group is a hydrogen atom, and (2) Polymers containing Si-H bonds, can be classified as MGE-HSMs. The polar, covalent, Si-H bond is hydridic and, when exposed to protonic “activators” under defined conditions, will generate H₂ gas. Generally, the rate of generation scales with the density of Si-H bonds in the material.

Monosilane, SiH₄, with the highest Si-H bond density of any of these types of materials, has long been considered a candidate MGE-HSM. Jackson and co-workers reported the synthesis of silane, SiH₄, from the reaction outlined in Eq. (6) [39].

$$3{\mathrm{SiO}}_2+4\mathrm{Al}+2{\mathrm{\text{AlCl}}}_3+6{\mathrm{H}}_2➔6/\mathrm{n}{\left[\mathrm{\text{AlOCl}}\right]}_{\mathrm{n}}+3{\mathrm{SiH}}_4$$E6

A serious problem with utilizing SiH₄ as an MGE-HSM is that, like diborane (B₂H₆), silane is spontaneously explosive in air. Therefore, one method to use Si-H bond-containing materials as H₂-generating agents, while avoiding explosions, is to work with polysilanes, whose linear structure, analogous with linear hydrocarbons, has the general formula Si_xH_2x+2.¹⁶ For example, as displayed in Eq. (7), Simone and co-workers demonstrated the thermal generation of H₂ from low MW polysilanes [65]:

$${\mathrm{Si}}_{\mathrm{x}}{\mathrm{H}}_{\mathrm{\text{2x}}+2}➔\mathrm{Si}\left(\mathrm{\text{metal}}\right)+\left(\mathrm{x}+1\right){\mathrm{H}}_2$$E7

In related work, as displayed in Eq. (8), Brunel demonstrated the hydrolytic generation of H₂ from polysilanes [38]:

$${\mathrm{Si}}_4{\mathrm{H}}_{10}+12{\mathrm{H}}_2\mathrm{O}/\mathrm{\text{NaOH}}➔{}^{“}{{\mathrm{SiO}}_2}^{”}+12{\mathrm{H}}_2$$E8

Linear polymers containing -SiH₂X- (X = CH₂, NH, O) monomeric groups could potentially function as MGE-HSMs. Polymers where X = O (“polysiloxanes” or “silicones”) were first reported by Rochow [66], Seyferth and co-workers [67, 68], Nishii and Narisawa [69], Harimoto and co-workers [70] and recently by Lin and co-workers [71]. As depicted in Eq. 10, conversion of -(O-SiH₂)_x- materials to silica (films), presumably with the release of hydrogen gas, occurred under a variety of conditions [70]:

$$-{\left(\mathrm{O}-{\mathrm{SiH}}_2\right)}_{\mathrm{x}}-+\mathrm{\text{2x}}{\mathrm{H}}_2\mathrm{O}➔-{\left[\mathrm{O}-\mathrm{Si}{\left(\mathrm{OH}\right)}_2\right]}_{\mathrm{x}}-+2{\mathrm{H}}_2$$E9

In a related study, Yap and co-workers reported the hydrolytic generation of H₂ from silicones (commercial products) containing the –(SiMeHO)- monomer [72].

Linear polymers containing -SiH₂X- where X = NH (“polysilazanes”) have been prepared and commercialized. The decomposition of select polysilazanes yield high quality¹⁷ silica thin films, with presumably the generation of H₂ under mild conditions [73, 74, 75, 76].

Linear polymers containing -SiH₂X- where X = CH₂ (“polycarbosilanes”) have been prepared [77]. Linear –(H₂C-SiH₂)_x- remains a liquid up to a substantial molecular weight. The heating of this liquid to approximately 300 deg. C results in the liberation of H₂ and the formation of a glassy solid. In a related study, when a stoichiometric amount of methanol was added to a mixture of cyclic organosilane, (CH₂SiH₂)₃ or (CH₂SiH₂CHSiH₃)₂, and 5 mol% NaOMe, rapid hydrogen release was observed at room temperature within 10–15 s. The original cyclic carbosilane could be regenerated via reaction with LiAlH₄ [78].

4. Conclusions

A major challenge with using hydrogen, H₂, as an energy source, is finding methods by which to store it effectively and safely. In some applications, storing “pure” H₂, either as a compressed gas or a cryogenic liquid, is not feasible. This concern leads to the concept of employing storage materials which, via chemical reactions, release H₂ gas under mild, predictable and controlled conditions. Specifically, this review is concerned with main-group element-based hydrogen storage materials, MGE-HSMs, in either the native state of the element or else compounds featuring hydridic X-H bonds, where the central element (X) is one of the following seven; Lithium (Li), Sodium (Na), Potassium (K), Magnesium (Mg), Calcium (Ca), Boron (B), Aluminum (Al) and Silicon (Si).

High energy, often CO₂ generating, chemical processes are necessary to produce these MGE-HSMs from their terrestrial precursors, which are usually oxide, chloride, or silicate-based minerals. Once formed, the reactivity of the hydridic X-H bond to protonic reagents, usually water, a process that efficiently generates H₂ gas, scales with the degree of X⁺H⁻ polarity of the bond. Regarding the Periodic Table of the Elements, this polarity increases either as one: (1) Descends a column of the table, or (2) Traverses from right-to-left across a row.

The Group I and II MGE-HSMs exist as the binary, ionic, hydride salts. The generation of H₂ from hydrolysis of these salts yields their common hydroxide compounds. The MGE-HSMs encompassing the hydrides of boron, aluminum and silicon are structurally complex and, in many cases, feature X-H bonds of significant covalent character which demonstrate reduced reactivity to protonic reagents. Furthermore, within the context of H₂ storage and generation; (1) Nanoscopic silicon metal, and a number of related composite materials, generate H₂ gas upon exposure to water at ambient temperatures, (2) Certain nanoscopic forms of MAlH₄ (M = Li, Na) demonstrate reversible, thermal hydrogenation/ dehydrogenation chemistries, and (3) Silicon-based polymers containing Si-H bonds release H₂ gas under specific hydrolytic/ thermolytic conditions.

Acknowledgments

The authors wish to thank Mr. Hans Haas, Mr. Kirk Ball and Dr. Sarah Snow for many useful discussions and insights.