Hydrogen Distribution And Onboard Storage

11.3.1 Hydrogen Distribution Systems

It is arguable that in no other area is the scale of the challenge faced by a transition to a hydrogen economy greater than in the transport sector. This is because the

Centralized production


I On-site I production


Liquid dispenser


Centralized production


Distribution m

Dewar m

Dewar r

Compressor J

Compressed dispenser r

Compressor J


FIGURE 11.4 Scheme of an integrated model of a hydrogen distribution system.

Compressed dispenser


FIGURE 11.4 Scheme of an integrated model of a hydrogen distribution system.

difficulties posed are not only the technical problems of production, common to all sectors, but there is the added complication of establishing a hydrogen storage/distribution network. Figure 11.4 shows an integrated model of the possible solutions to such an H2 distribution network. As can be inferred from the previous section, off-board hydrogen generation represents an easier solution to the problem of H2 production. Steam reforming of methane represents the best option for off-board generation, utilizing the existing natural gas network.

Massive investment is needed to realize this model. The needs are especially acute in the case of the personal automobile. For fleet vehicles, centralized refueling could be used, which would limit the number of sites necessary. In fact, H2/fuel-cell buses have already been introduced, albeit on a limited scale. Bigger vehicles are also more flexible in the choice of technology. In the case of automobiles, it would be necessary to replace all or most of the existing infrastructure, which gives rise to a classical chicken-or-egg problem: car manufacturers are unwilling to fully commit to H2 until the network exists, yet it is unlikely that investment will be stimulated in the absence of a ready market. It seems likely that the resolution will be a combination of factors — gradual replacement, using niche markets to stimulate investment, and with a combination of public and private investment and consumer payment. This process has in part been started in the United States, Japan, and Europe, with the introduction of more than 70 hydrogen filling stations worldwide and the so-called Hydrogen HyWays programs.

11.3.2 Hydrogen Storage

It has been estimated that to have a driving autonomy of 400 km, an average personal car, "optimized for mobility and not for prestige," requires about 8 kg of H2 for a hydrogen internal-combustion engine (H2 ICE) or 4 kg for a fuel-cell-driven car [26]. The problems of storing such large amounts of H2 are considerable.

There are numerous criteria that a hydrogen storage material must ideally satisfy before being considered suitable for use in automobiles. These include gravimetric storage density, volumetric storage density, safety, storage stability, and reproduc-ibility (all as high as possible); price, rehydrogenation and equilibrium pressure, enthalpic effects, and toxicity (all as low as possible); hydrogenation rate (as quick as possible); and dehydrogenation rate (as quick as necessary). To make any discussion more manageable, these criteria can be distilled into four main categories:

Gravimetric storage density

Volumetric storage density

Operating temperature (which is linked to pressures, enthalpy, and rates)


The criteria set by the USDOE [138], which are generally accepted as the basic targets, include a gravimetric density of 6.5 wt%, a volumetric density of 65 gl-1, and a decomposition temperature of between 333 and 393 K. Clearly, the storage method should be reversible, as safe as possible, and nontoxic. It should be noted that the densities refer to system characteristics, not material characteristics.

For the purpose of the present discussion, the potential methods of hydrogen storage will be subdivided into the following categories:

High-pressure gas cylinders

Liquid hydrogen in cryogenic tanks

Adsorbed hydrogen on high-surface-area sorbents (at T < 100 K)

Hydride systems Compressed Hydrogen

In order to achieve the required storage densities, pressures must exceed 50 MPa, and a target of 70 MPa is generally considered to be the objective. At present, 35-MPa tanks are routinely available. For comparison, laboratory steel hydrogen cylinders are normally pressurized to ca. 20 MPa. To withstand such high pressures, advanced cylinder materials, constructed from carbon-fiber-reinforced composite materials, have been developed. However, these are quite costly. A further cost-related problem associated with this technology is that of compression, which also considerably diminishes the specific energy content [139]. Other costs include the necessity of advanced pressure-control mechanisms. Finally, there is the problem of safety both during gas compression and use. The potential hazards of such highly pressurized cylinders are enormous, and in fact they have been ruled out as an option for personal automobiles in Japan. From the point of view of structural stability, tanks should ideally be spherical. However, this is rather inconvenient for transport applications, where space is at a premium. Thus, cylinders are usually used, although this results in higher structural strain.

Advanced research into this technology is in the area of materials engineering. Research into new tank designs can be divided into two categories. One approach is to design new methods of using the carbon fiber more efficiently, thereby using less. More recently, the idea of conformable or replicant tanks has been proposed. In these, the tank is made from many repeatable structures and thus the tank shape can be chosen to suit space considerations. This approach is still at a very early stage of development [140, 141]. Liquefied Hydrogen

Cryogenic storage of H2 in liquid form (21 K) is a relatively advanced technology [142], and it offers good storage H2 density (70.8 kgm-3). However, as with the gas compression, the liquefaction process is energy consuming and costly. A second drawback is that of evaporation losses. The critical temperature of H2 is very low (32 K). Even with the most advanced cryogenic tanks, some heat transfer occurs, which means that, to prevent very hazardous buildup of pressure, the system must be open or vented. This results in boil-off of the H2. A potential solution to this loss of fuel is to capture the boil-off with hydrides.

As above, research in this area is focused on the cryogenic tank. Progress is being registered, and the latest tanks significantly reduce boil-off to less than 3% per day after the first 3 days, boil-off being negligible during the first 3 days [143].

Although they do not meet the ideal specifications, gas or cryogenic storage at present represent the most advanced hydrogen storage technologies. High-Surface-Area Sorbents

The adsorption of a gas on a solid surface depends on the attractive forces between the two. These forces can in principle be divided into two types: weak van der Waals interactions that give rise to physical adsorption, or physisorption (binding energy ca. 0.1 eV); or stronger interaction that results in a chemical bond between the adsorbate (gas) and the adsorbent (surface), or chemisorption (binding energy ca. 2 to 3 eV). Both processes can potentially be exploited for onboard H2 storage, and adsorption on high-surface-area sorbents is a hugely active research field. The adsorption capacity is dependent on the type of adsorption process, with chemisorp-tion resulting in higher uptake.

In the specific case of H2, it is important to note that, above the critical temperature of 32 K, only monolayer adsorption is possible during the physisorption process. This means that the physisorption capacity (intended as the maximum adsorption) is proportional to the surface area of the material. Within this framework, the adsorption observed depends on the temperature and pressure: at high temperature it decreases rapidly, while at a given temperature, the storage is a function of the pressure. This pressure dependence in turn gives rise to the reversibility, with hydrogen released as the pressure decreases and vice versa [144, 145].

The ability of various gases to adsorb on carbon-based materials is well known. As it is possible to obtain carbon as a very-high-surface-area, highly porous material (activated carbon), and as carbon is a lightweight solid at room temperature, carbon-based materials appear to have good characteristics as H2 storage materials. Recently, reports of highly specific carbon materials, collectively known as nanostructured carbon materials, appear to have opened new possibilities in the field.

Various types of nanostructured carbon materials have been investigated [144-149]. Carbon nanotubes (CNT) were first reported by Ijima [150]. Singlewalled nanotubes (SWNT) consist of a graphene sheet rolled into a cylinder with an inner diameter 0.4 to 3 nm a length of 10 to 100 nm. They typically bunch together to form so-called nanoropes (10 to 100 parallel tubes). Multiwalled nanotubes (MWNT) consist of a series of concentric graphite cylinders (2 to 50 tubes, 30 to 50 nm). The interlay distance is similar to that of graphite. Graphite nanofibers (GNF) are stacks of graphite platelets (length: 5 to 100 pm, diameter: 5 to 200 nm) in various formations, with an interlayer spacing similar to that of graphite. Finally, carbon can also be nanostructured mechanically, usually by ball milling of graphite, which increases the surface area.

The issue of H2 storage on carbon-based materials is rather controversial. Two reports in particular triggered a flurry of research activity into nanostructured carbon materials. The first claimed a storage capacity of 5 to 10 wt% for SWNTs under low-pressure/high-temperature adsorption (p = 40 kPa, T = 273 K) conditions [151]. Subsequently, under high-pressure/high-temperature conditions (P = 11,365 MPa, T = 298 K), storage values of 4.52 wt% for graphite, 11.26 wt% for CNTs, 53.68 wt% for platelet GNFs, and 67.55 wt% for herringbone GNFs were reported [152]. These results were hailed as the breakthrough that would usher in a new era of fuel-cell-driven cars.

Other groups have reported similar excellent, or at least very high, uptakes for carbon nanostructures or modified carbon nanostructures [153-156]. In some of these cases, rather high temperature was required to obtain maximum adsorption or desorption, which can be attributed to a chemisorption process. The applicability of such systems to hydrogen storage for mobile applications is doubtful. High storage capacities have also been reported for nanostructured graphite [157, 158]. It should also be noted that many of the groups that reported high uptakes later published downward revisions of the storage amounts.

Against these promising results, a number of groups have not been able to independently reproduce high H2 uptakes. Uptake as low as 0.2 wt% [159], 1 wt% [160], 1 wt% [161], and 0.3 wt% [162] have been reported. Zuettel et al. concluded that adsorption of hydrogen at 77 K is due to physisorption and is therefore proportional to the specific surface area of the CNT and limited to 2 wt% for SWNT (theoretical surface area = 1315 m2g-1) [145]. A similar conclusion was reached by Strobel et al. [162] and Schimmel et al. [163]. If physisorption were indeed responsible, then relatively high uptakes should be possible only at low temperature, and the comparative room temperature (RT ) adsorption should be significantly smaller, or even negligible.

There are a number of reasons for the scatter in the data obtained. It has been rightly pointed out that the definition of "uptake" is sometimes rather vague and is not always used in the same way [144]. However, this is not the core of the problem, as large discrepancies still exist. It has been clearly shown that, in some cases, adsorption of water may give erroneous results [164]. Furthermore, large-scale production of nanotubes is notoriously difficult (although improving), and their purification is complicated. Impurities may contribute to or be responsible for the uptake observed. More specifically, it has been claimed that the presence of Ti-alloy, introduced as an impurity from a sonic probe habitually used to open SWNTs, is responsible for large H2-uptake values at room temperature [165]. This has led to interest in the H2-uptake properties of intentionally metal-decorated nanotubes and fullerenes.

Chemisorption as the adsorption process is also a possibility. However, like physisorption, this also creates a theoretical difficulty: if the adsorption energies associated with physisorption are too low to explain the high capacities reported for CNTs, then the adsorption energies associated with chemisorption are too high to explain reversible storage at ambient temperature. An intermediate state between physisorption and chemisorption has been hypothesized [166]. This suggestion has received support from theoretical calculations [167], but this has not been demonstrated. Another possibility is that an impurity catalyzes the chemisorption process.

Other high-surface-area sorbent materials have been investigated for their hydrogen storage capacities. Inorganic nanotubes, essentially boron and nitrogen analogues of carbon nanotubes, have received much attention [168, 169]. These are also referred to as collapsed nanotubes. Another class of materials that is beginning to attract attention as a hydrogen storage medium is that of metal-organic frameworks (MOFs) [170-172], which are distinguished by very high surface areas. In both cases, sufficient storage has not yet been reported. The reports on the latter materials have come under some criticism on the grounds that the isotherms reported are not consistent with any kind of adsorption process [173]. Hydride Systems Reversible Metal Hydride Systems

As a highly reactive element, hydrogen can form hydrides or solid solutions with most elements as well as with thousands of metal alloys of various compositions (AB5, AB2, AB, A2B, AB3, A2B7, A2B17, A6B23, etc.) [26, 143, 174]. Although the rich chemistry of hydrides prevents easy subdivision of various hydride species, reversible hydrides for H2 storage are conventionally divided into two categories: (a) interstitial hydrides, in which the hydrogen is absorbed on interstitial sites in a host metal, and (b) complex hydrides, in which the hydrogen is chemically bonded in covalent and ionic compounds. The complex elemental hydrides include mainly hydrides of elements in groups 1 to 3 and some transition metals (TM) hydrides. Reversible hydrides are charged and discharged by thermal means, and the storage behavior of a given material is described by a pressure-concentration-temperature (PCT) plot or, more simply, by van't Hoff plots (lnP vs. 1/Tdec) [26, 143]. The reversibility of metallic hydrides can be influenced by appropriate alloying, usually by combining strong and weak hydride-forming elements to produce compounds with intermediate thermodynamic affinities for hydrogen. A classical example is

Rehydrogenation Borohydrides

FIGURE 11.5 Comparison of the hydrogen stored in metal hydrides, carbon nanotubes, hydrocarbons, and other liquid fuels. (Adapted from Schlapbach, L. and Zuttel, A., Nature, 414, 353, 2001; Zuttel, A. et al., J. Alloys Comp., 356, 515, 2003; and Zuttel, A. et al., J. Power Sources, 118, 1, 2003. With permission.)

FIGURE 11.5 Comparison of the hydrogen stored in metal hydrides, carbon nanotubes, hydrocarbons, and other liquid fuels. (Adapted from Schlapbach, L. and Zuttel, A., Nature, 414, 353, 2001; Zuttel, A. et al., J. Alloys Comp., 356, 515, 2003; and Zuttel, A. et al., J. Power Sources, 118, 1, 2003. With permission.)

LaNi5H6, which shows fast and reversible sorption at room temperature (RT), with a dissociation pressure higher than 0.1 MPa [143]. This material has already been commercialized in hydride-based batteries.

Although hydride chemistry offers a seemingly endless list of storage candidates, when the full set of requirements for automotive applications is considered, the number of suitable materials decreases dramatically (Figure 11.5). To meet the gravimetric density requirements, only light metal hydrides and complex hydrides can be considered. The potentially suitable complex hydrides include the transition metal complexes (e.g., Mg2NiH4 = Mg2+ [NiH4]4- Mg2+) and nonmetal hydrides (LiBH4, NaBH4). However, not all of these are suitable thermodynamically. When reversibility (in general, complex hydrides have greater reversibility problems) or toxicity considerations are evaluated (e.g., Be hydrides, Li3Be2H7), the list is even further reduced [175]. Indeed, when all factors are considered, it may be stated that no hydrogen storage material currently meets all the generally accepted criteria for a H2 storage material [176, 177].

At present, complex metal hydrides are considered to be the most promising onboard storage systems. The alanate family (aluminum hydrides), especially sodium alanate (NaAlH4), have received much attention. Despite the favorable thermodynamics, it was originally thought that NaAlH4 was an unsuitable for hydride storage on the basis of its poor kinetic reversibility. In fact, it was considered impossible under practical conditions, necessitating a rehydrogenation temperature of 473 to 673 K (above the melting point) and a pressure of 10 to 40 MPa [143]. However, the observation that rehydrogenation could be catalyzed by Ti, under conditions acceptable for fuel cell use [178], has strongly renewed research interest in this material [173, 179-181].

Hydrogen is released in a two-step process:

Both steps can be achieved near 373 K, while the first step has an equilibrium pressure of 0.1 MPa at 303 K [143]. The theoretical maximum gravimetric storage density is 5.5 wt%, which is below the target limit but represents a good compromise. Research into this system is mainly directed toward finding efficient dopants to catalyze the rehydrogenation, finding efficient methods of doping, and optimizing the temperature and pressure characteristics of decomposition [182]. Titanium remains the best catalyst thus far reported. Mechanical methods, usually high-energy ball-milling, to improve the kinetics have also been employed with some success. Apart from the specific interest in NaAlH4, the discovery that catalysis can be used to overcome kinetic reversibility has opened up the general possibility that other hydrides can be similarly treated [183].

Magnesium-based systems are also potential hydrogen storage candidates. In many ways, Mg-based systems are a microcosm of the whole area of hydrogen storage materials: multiple possibilities exist, but always with at least one problem that makes them impractical for mobile application. Mg2NiH4 (3.6 wt%), Mg2FeH6 (5.5 wt%, volumetric density 150 kgm-3), and MgH2 (7.6 wt%) have all been investigated. The first would be suitable but for its low gravimetric density; the second has reasonable gravimetric density and remarkable volumetric density, but unsuitable thermodynamic properties; the third shows high gravimetric density but poor kinetic reversibility. However, Mg2FeH6 is a strong candidate in applications other than hydrogen storage, such as thermal energy storage. Recently, a method of chemically activating MgH2 has been reported [184], once more illustrating that kinetic problems can be overcome.

A novel hydrogen storage system is based on lithium nitride (LiN3), which can be hydrogenated first to a mixture of lithium imide and lithium hydride and then, in a second step, to lithium amide [153]:

Li3N + 2 H2 ^ Li2NH + LiH + H2 ^ LiNH2 + 2 LiH (11.25)

The second step, with a theoretical gravimetric storage density (with respect to LiNH2) of ca. 7%, was found to be almost completely reversible at 528 K and 558 K. However, it was also reported that addition of ca. 1% TiCl3 substantially improved the kinetics of the system and eliminated the undesirable formation of ammonia [185], which is a poison for the PEMFC membrane. The analogous Mg and Ca systems have also been investigated. These are clearly complicated systems, containing mixtures at both "ends" of the hydrogenation/dehydrogenation reaction. However, this complexity also introduces an element of flexibility. Based on this idea, other composite systems, with mixtures of the light elements/metals, have also been investigated. The general aim of this strategy is to improve the kinetics of the system while paying a minimum penalty in storage capacity.

The complex boron hydrides LiBH4 (18.4 wt%) and NaBH4 have also received some attention as reversible storage materials [176, 177]. In their pure forms, these compounds do not show good promise for mobile application, as they are very stable and thus require high operating temperature. While some progress has been registered in improving the behavior, for example, using SiO2 as a catalyst to lower the temperature of decomposition [177], it is true to say that their use in chemical storage systems at present offers a better alternative (discussed below).

Many other reversible hydride materials, even simple hydrides such as vanadium hydrides, have been investigated [26, 158, 174, 181, 183]. The strong interest is fueled by a widespread belief that hydride chemistry still offers the possibility of significant breakthroughs, either by synthesis of novel hydrides with suitable properties or by finding ways to catalyze the storage behavior of known materials. Chemical Hydride Storage

Chemical hydride storage involves release of H2 by chemical reaction of a hydride. The hydride is usually stored onboard as a solution or in a slurry. This might be considered as onboard production rather than storage; however, it is usually classified and discussed as chemical hydride storage, as hydrides are involved. The approach has the advantage of circumventing some of the reversibility problems of complex hydrides, but the disadvantage is that onboard regeneration is not possible. One option is hydrolysis of light hydrides or reactive metals [143]:

Alkaline solutions of alkali metal borohydrides (NaBH4, LiBH4) in H2O have received particular attention. The decomposition of aqueous solutions of NaBH4 to produce H2 is a well-established process, especially in the presence of a suitable catalyst [186-190]. The Millenium company has commercialized a system for hydrogen generation based on this reaction, and it is the subject of ongoing research for use with automobiles [191, 192].

An interesting recent field of study is that of liquid hydrides, which can be reversibly hydrogenated/dehydrogenated [193, 194]. Such systems are at an early stage of development, but they seem promising.

Was this article helpful?

0 0

Post a comment