The following section is taken from an article I am writing for Annual Reports C: Physical Chemistry, and published by the Royal Society of Chemistry. I can give the actual references if anyone wants them. My conclusion is the future doesn't look good for the putative Hydrogen Economy", given the difficulties encountered in storing the gas. I don't see any realistic use of H2 on the large scale, as I explain.
In an effort to address the twin-problems of dependence of nations upon imported hydrocarbon fuel and forcing climate change through global warming induced by emissions of CO2 from burning fossil fuels, hydrogen is being investigated as a clean, carbon-free fuel that could be made on a national (or regional) basis.4 However, hydrogen is not actually a "fuel" but an energy transfer (storage) medium. That is to say that hydrogen is not available in an aboriginal form as are oil, coal or gas, (which are known as "primary fuels"), but rather it must be "made" by some artificial means. Most of the hydrogen currently used in the world (mainly for chemical purposes, such as the wholesale manufacture of ammonia for fertilisers) is produced from natural gas by a process known as "steam reforming" sometimes with the use of a zeolite as a catalyst or to separate24 the carbon monoxide formed when the carbon is "extracted" from methane by its reaction with steam: CH4 + H2O → CO + 3H2, or to separate carbon dioxide when that CO is used to squeeze-out another molecule of H2 by adaptations of the "water-gas shift reaction": CO + H2O → CO2 + H2. Zeolites are also used more generally to remove CO2 from natural gas.24 Ideally, hydrogen should be "green", i.e. made by electrolysis of water using electricity produced from renewable sources, e.g. wind power, but it is arguable that those electrons would be more effectively used in forms of battery technology for driving vehicles and other electrical appliances. Nonetheless, efforts toward the putative hydrogen economy continue, and of greatest concern is the development of materials in which hydrogen might be effectively stored, including zeolites.4 A major advance has been made, which it is proposed may help address the vexed problem of storing hydrogen on the enormous scale which will be required if it is to be used to power vehicles to any significant extent. Researchers in Spain25 have found that a zeolite-Y partially exchanged with magnesium cations (Mg2+) has an unprecedented high adsorption enthalpy of -17.5 kJ/mol, which is close to the value of -15 kJ/mol recently proposed as optimum26 for a material that will efficiently both bind and release H2 according to the demand of its supply: i.e. the gas should neither be too strongly adsorbed otherwise it will not be released in a "fuelling station" situation, nor too weakly adsorbed otherwise the material is ineffective for storing it in the first place. The effect is attributed to the high polarising power (e/r) of Mg2+ cations. The effect of polarisation (see discussion in previous section) both induces an I.R. active vibration from adsorbed H2 and reduces the fundamental H-H stretching frequency from 4163 cm-1 measured by Raman spectroscopy in the gas phase for unperturbed molecules, in this case to 4056 cm-1. The value of -17.5 kJ/mol is significantly greater than those previously reported for the adsorption of H2 in alkali-metal cation exchanged zeolites,22,23,27 and is around 20x higher than the liquefaction enthalpy of H2 of -0.9 kJ/mol (at 20.45 K). Nonetheless, there does remain the issue of exactly how much hydrogen can be imbibed by a zeolite for practical purposes. For commercial applications, an acceptable energy density for a hydrogen storage tank is deemed to be that it can efficiently hold an amount of hydrogen equal to 6.5 wt.% of the weight of the tank and 62 kg H2/m3 in terms of volume.27,28 However, although investigations of hydrogen storage methods have been carried out for over 30 years, there has been no single method devolved which fulfils these demanding criteria. Some approaches meet the weight target, but occupy unsatisfactorily large volumes (e.g. tanks of compressed hydrogen gas) yet others achieve the volume target but not the weight ratio (e.g. metal hydride absorbents). To approach the matter from a theoretical perspective, a molecular mechanics study has been made of the thermodynamic limits on hydrogen storage in sodalite framework materials, built up from TO4 (where T = Al, Si, Ge, P) terahedra.30 It is concluded that cation-free sodalite structures could accommodate eight hydrogen molecules per cage as an optimum loading, at which point the density of the hydrogen is almost equal to that in liquid hydrogen, and the calculated densities of 65 kg H2/m3 can theoretically at least be achieved for most structures based on sodalite. For pure liquid hydrogen the figure is 70.8 kg H2/m3 which is the normal density of the liquid at a temperature of 20.28 K. However, to liquefy hydrogen costs around 30% of the energy that might be recovered from the material as a fuel.31 There is however a considerable discrepancy between the loading of sodalite found experimentally32, 0.26 and 0.4 wt.% for all-Si ( Si96O192) and AlP (Al48P48O192), and the calculated30 capacities of 4.8 and 5.2 wt.% respectively. However, the theoretical maximum capacities are based solely on energetic considerations, and do not address effects such as ions, water or other impurities that might act to block access to part of the internal volume of the sodalite crystals. There is also no influence of entropy included in the calculations which are in effect performed at zero Kelvin. In an extension of the theoretical work, adsorption isotherms of H2 in various sodalite materials were calculated using a grand canonical Monte Carlo method.33 It is concluded that at loading capacities of technical interest, 573 K and 100 bar, a storage capacity of around 0.1 wt.% might be achieved for each type of sodalite structure. However, the really technologically desirable capacities of above 4% are likely to only be met under conditions of extremely low temperature and/or extremely high pressure.33 The results make an interesting comparison with theoretically estimated maximal storage capacities for hydrogen in zeolitic materials. In effect, the adsorption can be thought of as a facilitated liquefaction, where the solid-gas interaction causes condensation at conditions of temperature and pressure that are more convenient than those required to form the bulk liquid. One such study34 was made which used the force-field method and performed its calculations within the Discover module of the Materials Studio 2.2 package of Accelerys Inc.35 The progressive filling with H2 of twelve purely siliceous models of common zeolite frameworks was simulated in order to determine the effect of framework properties including flexibility on the maximum adsorption capacity for hydrogen. It was deduced that the flexible non-pentasil zeolites (RHO, FAU, KFI, LTA and CHA)5 show the highest maximal capacities, in the range 2.65-2.86 wt.% of H2. The predicted adsorption capacities were found to correlate well with experimental results obtained at low temperatures (77K), but these materials are well below the 6.5 wt.% target value set for hydrogen storage in a practical device. The zeolite chabazite (CHA) has received particular attention in its context as a potential material for storing H2 since it was rated as having the largest capacity of any zeolite in this regard.31,36 For a H-exchanged (protonic) chabazite, H-SSZ-13 (Si/Al = 11.8), an absorption capacity of 1.28 wt.% was determined at 77K, slightly above that for zeolite-A at 1.24 wt.% and for H-CHA itself (Si/Al ratio = 2.1) at 1.10 wt.%.31 The hydrogen is described as "liquid hydrogen" in the zeolite, and it is shown that the available volume of a chabazite (H-SSZ-13) cage can contain seven hydrogen molecules at the density of liquid hydrogen. Actually in the zeolite, the results indicate that at 77K, 57 K above the boiling point of liquid hydrogen, about five hydrogen molecules are confined to each cage. This implies that conditions close to liquefaction are achieved when hydrogen is adsorbed into H-SSZ-13 zeolite at 77K, a result of sufficient importance that the paper was published in JACS.31 The point was investigated further by similarly measuring the volumetric uptake of H2 at 77K and a transmission I.R. measurement of H2 absorption at 15 K, in H-SSZ-13, (the isostructural silico-aluminophosphate material with the same Bronsted site density) H-SAPO-34, and H-CHA itself. It was found there is an improvement in H2 uptake when the acid strength of the Bronsted sites is increased (moving from H-SAPO-34 to H-SSZ-13), while conversely, increasing the density of Bronsted sites (moving from H-SSZ-13 to H-CHA) impacts negatively on the adsorption process. The latter result is quite counter-intuitive but an explanation is offered that the additional Bronsted sites are in mutual interaction via H-bonds inside the small cages of the chabazite framework and for most of them the energetic cost of displacing the adjacent OH ligand is higher than the adsorption enthalpy gained in forming the OH---H2 complex.36 The record set by H-SSZ-13 for a hydrogen storage capacity in a zeolite of 1.28 wt.%36 at 77K and one atmosphere pressure of H2 gas has been broken using low silica type-X zeolites (LSX, Si/Al = 1) fully exchanged with alkali-metal cations (Li+, Na+, K+).37 Hydrogen adsorption isotherms were determined separately at 77K and a pressure of <>2 and the cations follow the order Li+ > Na+ > K+, in order of the increasing cation radii: 0.068, 0.097, 0.133 nm, respectively. Li-LSX had an adsorption capacity of 1.5 wt.% at 77K and 1 atmosphere pressure, and a capacity of 0.6 wt.% at 298K and 10 MPa pressure, which places it among the highest of known sorbents. The possibility of enhancing the uptake of H2 by bridged hydrogen spillover was also investigated, for which a simple and effective method was found to construct carbon bridges between the dissociation catalyst and the zeolite to facilitate spillover of hydrogen atoms. By this means, the hydrogen storage capacity was enlarged to 1.6 wt.% (i.e. by a factor of 2.6) at 298K and 10 MPa pressure of hydrogen gas. This is by far the greatest hydrogen storage capacity achieved using a zeolite material at ambient temperature.37 A theoretical study was made of the hydrogen adsorption isotherms for a range of clathrasil frameworks. A clathrasil is a framework with Si6O6 as its largest ring aperture. The properties were calculated for twelve known clathrasils and seven hypothetical energetically stable versions. Under all conditions of temperature and pressure, high adsorption energies were predicted for small volume cages (<400a3) in consequence of the larger contact area between the cage wall and H2. Nonetheless, the H2 loading into the material is quite low because of the large internal surface-to-volume ratio which leaves little void space for the H2 molecules to occupy. It is concluded that clathrasils are unlikely to become of any use in practical hydrogen storage applications.38 An experimental study has been reported of the physisorption of H2 into zeolite types A, X and ZSM-5 under moderately high pressures of 2-5 MPa. The highest storage capacity found was 2.55 wt.% for Na-Y zeolite at 77K and 4 MPa pressure. In CaA, NaX and ZSM-5 zeolites, the hydrogen uptake was found to be proportional to the specific surface area of the adsorbent, and which were associated with the available void volumes of the zeolites.39
In conclusion, the prospect of using zeolites for practical hydrogen storage appears limited. More promising appear to be certain zeolite-templated porous carbons, and for one example a hydrogen uptake of 4.5 wt.% and 45 g/L weight and volumetric densities, respectively, were reported at 77K and 20 atmospheres (2 MPa) pressure.40 Still greater capacities for H2 are reported for porous coordination-framework materials giving uptakes of up to 6 wt.%, at 78K and pressures less than 20 atm., which are therefore likely to receive further attention as potential candidates for practical hydrogen storage systems.41 I conclude this section by noting one paper42 entitled "Hydrogen Storage: The major technological barrier to the development of hydrogen fuel cell cars", which provides a useful survey of the whole contentious business. I would also recommend the wikipedia entry on hydrogen storage.43 However hydrogen might be used, either as a pure substance or as adsorbed into zeolites or other porous materials, the energy costs of cryogenic cooling and compression must be born and factored into the energy balance equation for the hydrogen economy All such efforts to find a substitute for hydrocarbon fuels have brought home exactly how ideal the latter are as fuels, both in terms of energy density and their handling properties, and finding a substitute for them will be a hard act if it can be done at all.4