A database approach for materials selection for hydrogen storage in aerospace technology

Abstract

Hydrogen economy has been suggested as a possible green alternative to produce energy, also in the framework of transport applications. According to the specific transport means, different kinds of materials can be adopted. The choice of the most suitable materials should then be addressed according to a systematic analysis of available data. In this paper, together with the major physical storage technologies typically used for aerospace applications, additional possible candidates are suggested, namely clathrates hydrates and metal-organic frameworks (MOFs). They are chosen according to the specific features that are asked in the aerospace industry, such as high storage capacities, low weight and materials cost, high cyclability and full reversibility. To this scope, a comprehensive database based on a large set of information from literature (containing, for example, details on the synthesis processes, the operating temperatures and pressures, volumetric and gravimetric capacities) has been created, and specific tools have been developed to query the database. Indeed, the selection of the materials has been performed via an alternative database approach where the queries can be managed using a user-friendly tool, and potential materials can be selected based on any pool of desirable properties in quantitative terms. Essential information and characterization on theoretical and experimental data about these performing materials are provided and commented. As an example, in this paper, the case of clathrates hydrates is shown, and their potential impact is explored and characterized in this context, suggesting the most suitable synthesis processes.

Appendix: Highly compressed, liquid hydrogen and cryo-compressed storage

The phase diagram of hydrogen shows several interesting behaviours according to the temperature and pressure ranges. At normal condition of temperature and pressure, hydrogen exists in gas state. Decreasing the temperature and/or increasing the pressure leads to the phase transition to liquid and to solid. Four different insulating molecular crystal phases have been observed for increasing pressure. Moreover, metallic hydrogen has been recently found as a new phase of matter, that shows to be superfluid and superconductor; it is expected to be found in the interior of gas giants as Jupiter. The characteristic temperatures for hydrogen (critical temperature, boiling point and melting temperature) are quite low compared to other elements (32.976 K, 20.28 K and 13.81 K, respectively). Eventually, an important role is played by quantum mechanics and statistical mechanics in the description of phase transitions, especially the gas to liquid one. In particular, the technological problems connected to the liquefaction of hydrogen can be understood in the framework of nuclear spins statistics. Historically, the discovery of different nuclear states has been argued and inferred studying the experimental curve of the specific heat of hydrogen. Indeed, molecular hydrogen can be found in two distinct nuclear configurations, labelled as ortho and para and corresponding respectively to odd and even rotational quantum numbers.

Highly compressed hydrogen gas At room temperature (RT), hydrogen gas is described by a van der Waals-like equation of state; indeed, it behaves as a non-ideal gas because of the strong repulsive interaction among $\hbox {H}_2$ molecules.⁴ Hydrogen is stored as highly compressed gas in particular cylinders that can endure pressures up to 80 MPa for composite materials tanks (20 MPa for steel ones). The variation in pressure $\varDelta p$ across the tank wall (also referred to as tank gauge pressure) depends on the geometry of the cylinder (specifically on the thickness of the wall $t_w$ and of the total cylinder diameter $t_o$) and on a parameter ($\sigma _v$, called tensile strength) that is specific of the materials the cylinder is made of. In particular the following relation holds:

$$\begin{aligned} \varDelta p = 2 \sigma _v \left( \frac{t_o}{t_w} - 1\right) ^{-1} \end{aligned}$$

(1)

hydrogen volumetric density around

The tensile strength varies according to the material, assuming values that are of the order of 50 MPa for aluminum, 1100 MPa for high quality steel and 2410 MPa in the case of boron. In designing a cylinder for highly compressed gas, the higher the tensile strength, the better is the material in terms of performances; moreover, the material should be light to prevent the tank to be heavy. At the moment, the maximum tank pressure that has been achieved corresponds to 80 MPa, which corresponds to a hydrogen volumetric density around 36 kg.$\hbox {m}^{-3}$ (Zuttel 2003).

Cryogenic liquid hydrogen At normal pressure condition, hydrogen gas can be liquefied by cooling the system down to the boiling point (20.28 K). The cooling process for gases usually requires cycles in which the system is firstly compressed, isoenthalpically and, subsequently, it is expanded. Typically, at RT, gases (e.g., nitrogen) warm upon compression and cool during the expansion⁵; this is called the Joule-Thomson effect and it is used in the Linde cycle to reach cryogenic temperatures. However, for $\hbox {H}_2$, as well as for He and Ne, the gas warms upon expansion. For this reason, the system must be cooled down below the inversion temperature at the corresponding operating pressure; from that moment on, the regular Linde cycle can be applied to reach lower temperatures. In the case of hydrogen, the maximum inversion temperature is 202 K at 0 atm. For this reason, to start with higher pressure and eventually expand the gas, the hydrogen is pre-cooled down to 78 K by means of liquid nitrogen (that is eventually recycled in the refrigeration loop).

The experimental amount of energy required in the process of hydrogen liquefaction at RT is $W_{exp} \sim$15.2 kWh $\hbox {kg}^{-1}$ (Zuttel 2003). An important aspect that limits and mines the overall efficiency relies on the boil-off rate of hydrogen from a liquid storage vessel; this is due to heat leaks, that depend on the geometry of the tank (size and shape) and on the thermal insulation applied. Since boil-off losses caused by heat leaks are proportional to the surface-to-volume ratio, as the size of the vessel increase, the evaporation rate decreases. In terms of orders of magnitude, in the case of doubly-walled vacuum-insulated tanks with spherical geometry, boil-off losses are typically 0.4%/day, 0.2%/day and 0.06%/day for volumes of 50 $\hbox {m}^3$, 100 $\hbox {m}^3$ and 20000 $\hbox {m}^3$, respectively (Zuttel 2003).

Together with these technological aspects, there are important effects on the gas-liquid phase transition due to the conversion between the ortho- to para- hydrogen species. They represent a special rearrangement of nuclear spins (parallel and antiparallel respectively). The distribution of ortho- and para-hydrogen is a key element to describe possible heat loss deriving from the conversion in between these two nuclear spin arrangements (Dunlap 2014) ; having a proper description of the ortho- to-para ratio would give a more precise estimate of the heat transfer and a better technological design can eventually be suggested (and it will part of following tasks of the present project).

Cryo-compressed storage The cryo-compressed storage is a mixed approach between the highly compressed gas and the cryogenic storage technologies. The volumetric density obtained in such condition can easily reach 87 g/l, alleviating also the boil-off problem and allowing, at the moment, for the longest driving distance with a single tank in the automotive application. This efficiency corresponds to 5.8 wt% and 43 g $\hbox {H}_2$/l.⁶