Hydrogen release properties of lithium alanate for application to fuel cell propulsion systems

In this paper the results of an experimental study on LiAlH₄ (lithium alanate) as hydrogen source for fuel cell propulsion systems are reported. The compound examined in this work was selected as reference material for light metal hydrides, because of its high hydrogen content (10.5 wt.%) and interesting desorption kinetic properties at moderate temperatures. Thermal dynamic and kinetic of hydrogen release from this hydride were investigated using a fixed bed reactor to evaluate the effect of heating procedure, carrier gas flow rate and sample form. The aim of this study was to characterize the lithium alanate decomposition through the reaction steps leading to the formation of Li₃AlH₆ and LiH. A hydrogen tank was designed and realized to contain pellets of lithium alanate as feeding for a fuel cell propulsion system based on a 2-kW Polymeric Electrolyte Fuel Cell (PEFC) stack. The fuel cell system was integrated into the power train comprising DC-DC converter, energy storage systems and electric drive for moped applications (3 kW). The experiments on the power train were conducted on a test bench able to simulate the vehicle behaviour and road characteristics on specific driving cycles. In particular the efficiencies of individual components and overall power train were analyzed evidencing the energy requirements of the hydrogen storage material.     

Electrochemical reaction of lithium alanate dissolved in diethyl ether and tetrahydrofuran

We present electrochemical properties of lithium alanate (LiAlH₄) dissolved in aprotic ethers – diethyl ether (Et₂O) and tetrahydrofuran (THF) – under an Ar atmosphere of 1 atm at 298 K. Specific conductivities of both LiAlH₄–THF and LiAlH₄–Et₂O solutions are measured by AC four-terminal method. Cyclic voltammetry is performed with using a beaker-type electrochemical cell consisted of a Ni wire, Ni mesh and Li wire as a working, counter and reference electrode, respectively. In order to clarify the electrochemical behavior, anodic polarization of LiAlH₄–THF solution is measured. The current density of 1.0 M LiAlH₄–THF solution reaches to 1 A cm⁻², which is higher than the LiAlH₄–Et₂O solution. Quantitative analysis of H₂ gas generated on the working electrode during the potentiostatic electrolysis tells that the number of electrons involved in the anodic reaction at the limiting current is one in case of the LiAlH₄–THF solution. We propose conceivable electrochemical reactions of LiAlH₄ in the non-aqueous ethereal solutions.     

Reversible hydrogen storage in LiBH4/Ca(AlH4)2 systems

To improve the hydrogen storage property of LiBH₄, the LiBH₄/Ca(AlH₄)₂ reactive systems with various ratios were constructed, and their de-/hydrogenation properties as well as the reaction mechanisms were investigated experimentally. It was found that the sample with the LiBH₄ to Ca(AlH₄)₂ molar ratio of 6:1 exhibits the best comprehensive hydrogen storage properties, desorbing hydrogen completely (8.2 wt.%) within 35 min at 450 °C and reversibly absorbing 4.5 wt.% of hydrogen at 450 °C under a hydrogen pressure as low as 4.0 MPa. During the first dehydrogenation process of the LiBH₄/Ca(AlH₄)₂ systems, the CaH₂ and Al particles were in situ precipitated via the self-decomposition of Ca(AlH₄)₂, and then reacted with LiBH₄ to form CaB₆, AlB₂ and LiH. Whereafter, the sample can cycle between LiBH₄ + Ca(BH₄)₂ + Al in the hydrogenated state and CaB₆ + AlB₂ + LiH in the dehydrogenated state.     

Enhanced volumetric hydrogen density in sodium alanate by compaction

Powder compaction is a potential process for the enhancement of the volumetric and gravimetric capacities of hydrogen storage systems based on metal hydrides. This paper presents the hydrogen absorption and desorption behaviour of compacts of sodium alanate material prepared under different levels of compaction pressure. It is shown that even at high compaction levels and low initial porosities, hydrogen absorption and desorption kinetics can proceed comparatively fast in compacted material. Furthermore, experimental hydrogen weight capacities of compacted material are higher than the experimental values obtained in case of loose powder. It is demonstrated that the kinetic behaviour of the compacted material during cycling is directly associated to the volumetric expansion of the compact, which is quantitatively measured and analyzed during both hydrogen absorption and desorption processes. The cycling behaviour and dimensional changes of compacted sodium alanate material are a key consideration point if it is used as hydrogen storage materials in practical tank systems.     

Experimental investigation on lithium borohydride hydrolysis

Lithium borohydride, one of the highest energy density chemical energy carriers, is considered as an attractive potential hydrogen storage material due to its high gravimetric hydrogen density (19.6%). Belonging to borohydride compounds, it presents a real issue to overcome aims fixed by the U.S. Department of Energy in the field of energy, and so crystallizes currently attention and effort to use this material for large scale civil and military applications. However, due to its important hygroscopicity, lithium borohydride is a hazardous material which requires specific handling conditions for industrial aspects.     

Effects of cerium oxide catalyst on the dehydrogenation of lithium alanate using synchrotron XRD and XAFS

Compared with titanium oxide (TiO₂) and zirconium chloride (ZrCl₄), cerium oxide (CeO₂) significantly affects the dehydrogenation properties of lithium alanate (LiAlH₄). CeO₂-doped LiAlH₄ samples obtained in different decomposition states were characterized by in situ X-ray diffraction, synchrotron radiation powder X-ray diffraction (PXD) data, and X-ray absorption fine structure (XAFS). PXD and XAFS analyses reveal that CeO₂ is present as a cubic CeO_2−x phase and behaves as a true catalyst. Ce⁴⁺ reduction affects the ability of Li to donate its charge to AlH₄, thereby weakening the Al–H bond and causing hydrogen to desorb at lower temperatures. Beyond 180 °C, approximately 50% of the O atoms of the CeO₂ catalyst are removed, which results in the formation of long-range disorder rhombohedral Ce₇O₁₂ phase and an opposite effect on the dynamics.     

Desorption kinetics of lithium amide/magnesium hydride systems at constant pressure thermodynamic driving forces

Lithium amide and magnesium hydride are lightweight materials with high hydrogen-holding capacities and thus they are of interest for hydrogen storage. In the present work mixtures with initial molar compositions of (LiNH₂ + MgH₂) and (2LiNH₂ + MgH₂) were ball milled with and without the presence of 3.3 mol% potassium hydride dopant. Temperature programmed desorption, TPD, analyses of the mixtures showed that the potassium hydride doped samples had lower onset temperatures than their corresponding pristine samples. The dehydrogenation kinetics of the doped and pristine mixtures was compared at 210 °C. In each case a constant pressure thermodynamic driving force was applied in which the ratio of the plateau pressure to the applied hydrogen pressure was set at 10. Under equivalent conditions, the (LiNH₂ + MgH₂) mixture desorbed hydrogen about 4 times faster than the (2LiNH₂ + MgH₂) mixture. The addition of potassium hydride dopant was found to have a 25-fold increase on the desorption rates of the (2LiNH₂ + MgH₂) mixture, however it had almost no effect on the desorption rates of the (LiNH₂ + MgH₂) mixture. Activation energies were determined by the Kissinger method. Results showed the potassium hydride doped mixtures to have lower activation energies than the pristine mixtures.     

Numerical modeling and performance evaluation of multi-tubular sodium alanate hydride finned reactor

The optimization of the hydrogen loading process in a multi-tubular sodium alanate hydride reactor equipped with longitudinal fins is investigated numerically. The effect of the number, thickness and tip clearance of the fins on the hydrogen charging rate is assessed, so that the fin optimal geometric properties are determined by the compromise between the hydrogen loading rate and the fin contribution to the weight and the volume of the storage system. Simulation results have shown that the hydrogen loading rate corresponding to this optimized configuration is 41% greater than the case without fins if we suppose a perfect interconnectivity between the fin tips and the internal walls of the hydride tubes. Otherwise, the amount of stored hydrogen decreases drastically. The loading of hydrogen under high charging pressures results in higher hydrogen loading rates and there is an interaction between the geometric and operating parameters leading to the optimized amount of stored hydrogen.     

Preparation of a new Ti catalyst for improved performance of NaAlH4

Three effective Ti catalysts for NaAlH₄ were made by stoichiometrically reacting TiCl₃ with LiAlH₄ in tetrahydrofuran (THF), NaAlH₄ in THF, and LiAlH₄ in diethyl ether (Et₂O). The solid products produced after drying were named ex situ catalysts and designated respectively as Ti(Li)T, Ti(Na)T and Ti(Li)E. NaAlH₄ was dry doped with 2 mol% of these ex situ catalysts, and for comparison, NaAlH₄ was conventionally wet doped with 2 mol% TiCl₃ in THF that made in situ catalyst (designated as TiCl₃). All four doped samples were dry ball milled, and hydrogenation and dehydrogenation studies were carried out over five cycles. Temperature programmed desorption, constant temperature desorption, and constant temperature cycling curves showed that the effectiveness of these catalysts decreased as Ti(Li)T > Ti(Na)T > TiCl₃ > Ti(Li)E. Ti(Li)T ex situ catalyst, being the best Ti catalyst, markedly decreased the dehydrogenation temperature, improved both the hydrogenation and dehydrogenation kinetics with sustained rates over cycling, and exhibited the least loss of hydrogen storage capacity over cycling. Ti(Li)T ex situ catalyst exhibited properties commensurate with some of the best NaAlH₄ catalysts to date, such as CeCl₃, ScCl₃ and Ti nanocluster. It is easy to make, readily available and relatively inexpensive.     

Decomposition of lithium magnesium aluminum hydride

The quaternary aluminum hydride LiMg(AlH₄)₃ contains 9.7 wt% hydrogen, of which 7.2 wt% can be released in a two-step decomposition reaction via first formation of LiMgAlH₆ and then the binary hydrides MgH₂ and LiH. In-situ synchrotron radiation powder X-ray diffraction and thermal desorption spectroscopy measurements were performed to analyze the product distributions formed during the thermal decomposition of LiMg(AlD₄)₃. The first decomposition step occurs at about 120 °C and the second at about 160 °C for the as-milled sample, while for a purified sample of LiMg(AlD₄)₃, the decomposition temperatures involving release of hydrogen increase to 140 and 190 °C, respectively, suggesting that pure samples of LiMg(AlD₄)₃ are kinetically stabilized. Studies of the purified LiMg(AlD₄)₃ also showed that the second decomposition step can be divided into two reactions: 3LiMgAlD₆ → Li₃AlD₆ + 3MgD₂ + 2Al + 3D₂ and Li₃AlD₆ → 3LiD + Al + 3/2D₂. Addition of TiCl₃ to LiMg(AlD₄)₃ under a variety of ball milling conditions consistently led to decomposition of LiMg(AlD₄)₃ during milling. Correspondingly, all attempts to rehydrogenate the (completely or partially) decomposed samples at up to 200 bar hydrogen pressure failed. Decomposition of MgD₂ was observed at relatively low temperatures. This is ascribed to thermodynamic destabilization due to the formation of different Al_xMg_y phases, and to kinetic destabilization by addition of TiCl₃. A thermodynamic assessment was established for the calculation of phase stability and decomposition reaction relationships within the Li-Mg-Al−H system. The calculations confirmed the metastability of the LiMg(AlH₄)₃ phase and the irreversibility of the Li-Mg alanate phase decomposition reactions. The Li-Mg alanate decomposition pathways followed experimentally could be explained by the endothermicity of the calculated decomposition enthalpies, in that an impure or catalyzed LiMgAlH₆ intermediate phase could more directly access an endothermic decomposition reaction at lower temperatures, while a kinetically-hindered, purified LiMgAlH₆ would require higher temperatures to initiate the two-step decomposition through an exothermic reaction.     

Comparisons between MgH2- and LiH-containing systems for hydrogen storage applications

The present study compares the dehydrogenation kinetics of (2_LiNH2+_MgH2)$(2{\mathrm{LiNH}}_2+{\mathrm{MgH}}_2)$ and (_LiNH2+LiH)$({\mathrm{LiNH}}_2+\mathrm{LiH})$ systems and their vulnerabilities to the NH₃ emission problem. The (2_LiNH2+_MgH2)$(2{\mathrm{LiNH}}_2+{\mathrm{MgH}}_2)$ and (_LiNH2+LiH)$({\mathrm{LiNH}}_2+\mathrm{LiH})$ mixtures with different degrees of mechanical activation are investigated in order to evaluate the effect of mechanical activation on the dehydrogenation kinetics and NH₃ emission rate. The activation energy for dehydrogenation, the phase changes at different stages of dehydrogenation, and the level of NH₃ emission during the dehydrogenation process are studied. It is found that the (2_LiNH2+_MgH2)$(2{\mathrm{LiNH}}_2+{\mathrm{MgH}}_2)$ mixture has a higher rate for hydrogen release, slower rate for approaching a certain percentage of its equilibrium pressure, higher activation energy, and more NH₃ emission than the (_LiNH2+LiH)$({\mathrm{LiNH}}_2+\mathrm{LiH})$ mixture. On the basis of the phenomena observed, the reaction mechanism for the dehydrogenation of the (2_LiNH2+_MgH2)$(2{\mathrm{LiNH}}_2+{\mathrm{MgH}}_2)$ system has been proposed for the first time. Approaches for further improving the hydrogen storage behavior of the (2_LiNH2+_MgH2)$(2{\mathrm{LiNH}}_2+{\mathrm{MgH}}_2)$ system are discussed in light of the newly proposed reaction mechanism.     

Enhanced hydrogen storage properties of LiBH4-MgH2 composite by the catalytic effect of MoCl3

Though LiBH₄-MgH₂ system exhibits an excellent hydrogen storage property, it still presents high decomposition temperature over 350 °C and sluggish hydrogen absorption/desorption kinetics. In order to improve the hydrogen storage properties, the influence of MoCl₃ as an additive on the hydrogenation and dehydrogenation properties of LiBH₄-MgH₂ system is investigated. The reversible hydrogen storage performance is significantly improved, which leads to a capacity of about 7 wt.% hydrogen at 300 °C. XRD analysis reveals that the metallic Mo is formed by the reaction between LiBH₄ and MoCl₃, which is highly dispersed in the sample and results in improved dehydrogenation and hydrogenation performance of LiBH₄-MgH₂ system. From Kissinger plot, the activation energy for hydrogen desorption of LiBH₄-MgH₂ system with additive MoCl₃ is estimated to be ∼43 kJ mol⁻¹ H₂, 10 kJ mol⁻¹ lower than that for the pure LiBH₄-MgH₂ system indicating that the kinetics of LiBH₄-MgH₂ composite is significantly improved by the introduction of Mo.     

Cerium hydride generated during ball milling and enhanced by graphene for tailoring hydrogen sorption properties of sodium alanate

In this study, NaAlH₄?based hydrogen storage materials with dopants were prepared by a two-steps in-situ ball milling method. The dopants adopted included Ce, few layer graphene (FLG), Ce + FLG, and CeH_2.51. The hydrogen storage materials were studied by non-isothermal and isothermal hydrogen desorption measurements, X-ray diffractions analysis, cycling sorption tests, and morphology analysis. The hydrogen storage performance of the as-prepared NaAlH₄ with Ce addition is much better than that with CeH_2.51 addition. This is due to that the impact of Ce occurs from the body to the surface of the materials. The addition of FLG further enhances the impact of Ce on the hydrogen storage performance of the materials. The hydrogen storage capacity, hydrogen sorption kinetics, and cycle performance of NaAlH₄ with Ce + FLG additions are all better than NaAlH₄ materials with the addition of either Ce or FLG alone. The NaAlH₄ with Ce and FLG addition starts to release hydrogen at 85 °C and achieves a capacity of 5.06 wt% after heated to 200 °C. The capacity maintains at 4.91 wt% (94.7% of the theoretical value) for up to 8 cycles. At 110 °C, the material can release isothermally a hydrogen capacity of 2.8 wt% within 2 h. The activation energies for the two hydrogen desorption steps of NaAlH₄ with Ce and FLG addition are estimated to be 106.99 and 125.91 kJ mol^?1 H₂, respectively. The related mechanisms were studied with first-principle and experimental methods.     

Synthesis and hydrogen storage properties of lithium borohydride amidoborane complex

A series of mixtures of LiAB/LiBH₄ with different molar ratios were prepared and their hydrogen storage properties were investigated in this study. Among them, a new structure was found in the LiAB/LiBH₄ sample with a molar ratio of 1/1. It is of orthorhombic structure and composed of alternative layers of LiAB and LiBH₄. It shows similar hydrogen desorption behaviors of LiAB–LiBH₄ and LiAB–0.5LiBH₄. For use in hydrogen storage, high hydrogen capacity and low operation temperature are demanded, thus, the dehydrogenation properties of LiAB–0.5LiBH₄ were subsequently measured. Three steps of desorption were observed during the heating process, with a total release of 11.5 wt% H₂ at 500 °C. The reaction path was identified using a combined investigation of XRD and ¹¹B solid state NMR. Dehydrogenation kinetic analyses show that the complex has lower activation energy (61 ± 4 kJ mol⁻¹ H₂) than that of LiAB (71 ± 5 kJ mol⁻¹ H₂). It is likely that dehydrogenation process was promoted due to the presence of LiBH₄.     

Behavior of scaled-up sodium alanate hydrogen storage tanks during sorption

Sodium alanate is being experimentally tested in scaled-up quantities. For this purpose, several tanks have been designed and constructed. The tank functionality during absorption and desorption of hydrogen was demonstrated in a scale of 8 kg of alanate, with a peak technical absorption time below 10 min. The absorption and desorption data show good reproducibility. Neutron radiography was used in another tank to show the powder’s physical behavior during sorption, showing conservation of the macroscopic structure during cycling.     

Improving solid-state hydriding and dehydriding properties of the LiBH4 plus MgH2 system with the addition of Mn and V dopants

The hydriding process of the 2LiH + MgB₂ mixture is controlled by outward diffusion of Mg and inward diffusion of Li and H within MgB₂ crystals to form LiBH₄. This study explores the feasibility of using transition metal dopants, such as Mn and V, to enhance the diffusion rate and thus the hydriding kinetics. It is found that Mn can indeed enhance the hydriding kinetics of the 2LiH + MgB₂ mixture, while V does not. The major factor in enhancing the diffusion rate and thus the hydriding kinetics is related to the dopant's ability to induce the lattice distortion of MgB₂ crystals. This study demonstrates that the kinetics of the diffusion controlled solid-state hydriding process can be improved by doping if the dopant is properly selected.     

Cycling and engineering properties of highly compacted sodium alanate pellets

Sodium alanate powder comprised of NaH and Al was doped with 3 mol% titanium chloride (TiCl₃) and pelletized into highly compacted cylindrical pellets. The pelletization process was performed to improve thermal conductivity and volumetric hydrogen capacity of the metal hydride, compared to loose or tapped powder, which are vital requirements for on-board hydrogen storage applications. The pelletization process was performed over a range of 69–345 MPa (10–50 kPSI) with a 95% increase in density and improvement in thermal conductivity 18 times greater compared to powder at the maximum pelletization pressure (1.60 g/cm³ and 0.82 g/cm³; 9.09 W/m K and 0.50 W/m K, respectively). Hydrogen cycling capacities and kinetics were not adversely affected by the pelletization process although 10 cycles are required to obtain full hydrogen capacity. Pellet cycling capacity maintained a stable 4 wt% H₂ over 50 cycles. Ti-doped NaH + Al pellets exhibited similar thermal cycling expansion as with the loose powder; within 30 cycles there was a 50% loss in pellet density and by 50 cycles the loss in pellet structural integrity made handling problematic.