Setting up a volumetric method to test H2 storage materials

One of the main issues in the development of the so called “hydrogen economy” is represented by the development of efficient routes to store and transport it in safe and suitable way for mobile applications. The DoE circumscribed the most relevant goals for storage materials suitable for on board transportation of hydrogen. The newly adapted ultimate targets refer to a driving autonomy of 805 km for a medium size car, prescribing 5.5 wt% H2 stored on board as gravimetric target for 2015 and 7.5 wt% as ultimate goal. Such properties refer to the entire storage system and therefore the material should reach approximately twice the desired system capacity. Different methods were used to assess gas sorption capacity, classifiable into gravimetric, volumetric and TPD (temperature programmed desorption). Scarce accuracy in the determination of H2 uptake may explain the inconsistencies frequently found in the literature. It appears that the set up of a simple and reliable method for the quantification of hydrogen storage properties is still an open issue. Therefore, the aim of the present work was to develop a volumetric system to quantify the amount of hydrogen “delivered” from the sample after saturation, which is the most important practical parameter to evaluate the effectiveness of a hydrogen storage material. The method has been standardized under different operating conditions on three active carbons and one graphite sample, characterised by surface area between 300 and 3000 m2/g. The data have been elaborated according to very simple adsorption models, such as the Langmuir one. The results have been compared with a dynamic testing procedure (TPD) here developed to assess the reversibility of adsorption and its kinetics. Some thermal treatments under different atmosphere have been also attempted in order to tune up the textural properties of the materials and to modify their surface composition. CHN analysis has been performed and ash content was determined by heating in air at 1173 K until constant weight. Structural properties were investigated by X-ray diffraction and textural ones by N2 adsorption/desorption. The apparatus set up for static volumetric testing was constituted by a test tube, obtained from ¼” stainless steel tube (5 cm long), closed on one side and connected to a filter. The sample holder may be immersed in an ethylene glycol bath, heated and cooled by a heater/cryostat, or in a Dewar flask containing liquid nitrogen or any other refrigerating mixture. Through a set of valves the material may be pretreated by outgassing under vacuum during sample heating, or connected to the He gas line for the determination of dead volume, or to the H2 line at selected gas pressure. The amount of H2 released (“delivered”) is measured at r.t. after removal of the refrigerating unit. The gas is allowed to freely expand and its volume is measured at ambient pressure by means of a gas burette. Typically, the materials have been rested at 273 K at H2 relative pressure up to100 kgf/cm2 and at 77 K up to 20 kgf/cm2. A conventional TPD apparatus equipped with a thermal-conductivity detector has been also used for the determination of H2 uptake and release under dynamic conditions. H2 adsorption-desorption tests were carried out at atmospheric pressure between 258 and 298 K. Preliminary blank tests have been carried out for the calibration of the instrument. Each sorption cycle has been repeated at least 5 times under every operating condition and during the set up of the method repeated tests with different batches have been carried out to check the repeatability of the whole procedure. The best results achieved with these samples were obtained under cryogenic conditions. For instance, the sample with the highest surface area delivered 6.9  0.3 wt% H2 after adsorption at 77 K and 20 kgf/cm2. By contrast, the best result attained after sorption at 273 K and 100 kgf/cm2 was 2.6  0.2 wt% H2. Such data are better than most literature reports. The application of the simple Langmuir model to our data also allowed to roughly estimate the degree of coverage of the surface, so to understand if an increase of the operating pressure would be convenient or not, in case the sample was nearly saturated yet. Finally, the comparison with the dynamic adsorption/desorption method allowed to exclude significant kinetic or mass transfer limitations under the selected operating conditions. The latter may therefore be adopted for the routine characterisation of the materials, to select the most promising ones for deep characterisation with the more accurate static method.