|
Project Description and Objectives: |
Most technical applications of hydrogen storage materials require a fast reaction rate for the hydriding - dehydriding cycles. Surface contamination caused by impurity gases present in hydrogen (O2, CO, H2O, H2S ..) impede hydrogen absorption and desorption kinetics drastically [1,2]. Numerous experiments performed with various reactive metal films have shown, that chemisorbed molecules impede the process of hydrogen dissociation on the metal surface. Since reactive species like CO or H2S are strongly bound at chemisorption sites, an amount corresponding to one monolayer (1 ML = 1015 molecules/cm2) decreases the reaction rate by many orders of magnitude. However, on a closed oxide surface hydrogen dissociation is possible, but with increasing amounts of absorbed oxygen the hydrogen absorption rate is retarded continuously. The reaction finally stops when the oxid layer thickness has exceeded a critical value.
For storage alloys there is still a lack of information about the relations between absorbed amount of poisoning gases and the hydrogen absorption rate. Cracking of brittle hydride powder particles and/or spalling of the contamination layers create new, unpoisoned surfaces. Deviations from isothermal conditions during loading or unloading, caused by the reaction heat evolved affect the reaction rate as well. Therefore a standardization of absorption measurements is much more difficult with powder samples than with films.
|
|
Technical Goals: |
Hydrogen absorption of LaNi5 and LaNi4.7Al0.3 powder samples is investigated at ambient temperatures under isothermal and isobaric conditions. First the samples are activated and the absorption rate for high purity hydrogen is determined. After dehydriding the powder is exposed for ten minutes to a known amount of contamination gas and loaded again with pure hydrogen. For different precoverages the initial absorption rate, v, is compared with the rate, v0, of the uncontaminated sample. Each experiment is carried out with a new sample in order to guarantee identical initial conditions. The standardized activation procedure produces an approximately constant specific surface area of the powder. Thus, the reduction of the reaction rate v/v0 and the impurity precoverage (in monolayer equivalents ML) can be compared with results obtained for film samples [3-5].
Furthermore, the time laws measured for the absorption reactions may indicate which of the individual partial steps of the processes is rate determining. For the limiting case of one rate controlling reaction step simple reaction models can be used [6]. They predict typical time laws for the normalized amount M of hydrogen absorbed under isobaric and isothermal conditions.
If diffusion is the slowest step a time law M(t) t0.5 should be observed at the beginning of absorption. A surface controlled reaction, on the other hand, obeys a linear time law M(t) t. These time laws are based on the simplifying assumption, that the rate limiting step is much slower than all other reaction partial steps, which are assumed to be in equilibrium. If this approximation is not justified, more advanced model calculations must be performed, which contain the forward and backward reaction rates of all individual partial steps. Such a general model for the kinetics of hydrogen absorption and desorption of storage materials has been developed in order to gain a better understanding of intermediate regimes of real absorption/desorption processes. Transitions from one limiting time law to another are simulated and compared with experimental investigations, for example the transition from diffusion control to surface control caused by increasing contamination [2].
|
|
Preliminary or Final Results: |
The decrease of the initial hydrogen absorption rate of LaNi5 powder precovered with five typical gas impurities is compared. Less than two monolayers of CO or three monolayers of H2S respectively are poisoning the hydrogen reaction completely. These gases are strongly bound to chemisorption sites and prevent the hydrogen dissociation. Deviations in the critical layer thickness may be attributed to an acummulation of sulphur atoms. In contrast to this the effect of N2 exposure is negligible. Even a gas exposure corresponding to 500 ML N2 reduces
the absorption rate by less than 10%. Nitrogen is probably only weakly bound at physisorption sites and no chemisorption layer is formed on the powder surface. The effect of the oxide layer thickness on the hydrogen absorption rate seems to run in parallel to the oxygen absorption rate. The dissociation probability of an oxygen or hydrogen molecule is decreased by the presence of electrons in the antibonding molecular level. Since low temperature oxidation models [7-9] predict an electron transfer through thin oxide layers the hydrogen absorption is not completly stopped after one monolayer of oxygen is absorbed. Severe poisoning is observed only after the oxide layer thickness has reached a critical value of about six monolayers. H2O contamination of LaNi5 powder could not be investigated after precoverage, but by exposure to H2/H2O gas mixtures. The results correspond to findings for the experiments with O2 contamination. CO2 on LaNi5 causes only retardation but no passivating of the hydrogen absorption reaction. The addition of one ML CO2 decreases the initial absorption rate by about one order of magnitude. More extended CO2 exposure shows no effect. The bonding to the chemisorption sites is less strong than for CO or H2S. The chemisorbed molecules are in equilibrium with the gas phase and dissociation of hydrogen molecules is still possible on the remaining unoccupied sites.
The time laws of the experimental data measured by stepwise contamination indicate that kinetics of hydrogen absorption is dissociation controlled for all impurities tested, as soon as the initial absorption rate v/v0 is decreased below 10-1. For unpoisoned LaNi5 diffusion control has not been observed. This result agrees with predictions of model calculations for the intermediate regime. For a transition from one limiting case to another the reaction rate has to be changed by about two orders of magnitude.
|
|
Related Reference Papers and Other Publications: |
[1] E. Fromm, M. Martin, F. Schweppe, C. Gommel Einfluß des Oberflächenzustandes auf die Kinetik der Wasserstoffaufnahme und -abgabe von Speicherwerkstoffen und Metallen in: ""Wasserstoff als Energieträger"", VDI-Verlag Düsseldorf, 1994, 173-190
[2] F. Schweppe Einfluß von Gasverunreinigungen auf die Kinetik der Wasserstoffaufnahme von Hydridspeicherwerkstoffen Doctoral Thesis, University of Stuttgart 1995 and Fortschritt-Berichte VDI, Reihe 5, in press
[3] H.H. Uchida, E. Fromm Reaction Kinetics of Hydrogen and Oxygen Absorption of Titanium Films Journal of Advanced Science (Tokio, Japan) 2 (1990) 153-158
[4] N. Hosoda, H.H. Uchida and E. Fromm Effect of O2, N2 and CO Impurities on the H2 Absorption Rate of Ti Films at 300K J. Less.-Common Met., 172-174 (1991) 824-831
[5] N. Hosoda Einfluß von Kontaminationseffekten auf die Kinetik der Wasserstoffaufnahme von Filmproben aus Titan und anderen Metallen Doctoral Thesis, University of Stuttgart 1992
[6] M. Martin, C. Gommel, C. Borkhart, E. Fromm Absorption and Desorption Kinetics of Hydrogen Storage Alloys J. Alloys Comp., in press
[7] V. Grajewski, E. Fromm Low-Temperature Oxidation of Metals in J. Nowotny (ed.) ""Interface Segregation and related Processes in Materials"" Trans. Tech. Publications, Zürich 1991, 337-399
[8] H.H. Uchida, V, Grajewski, E. Fromm Model Calculations of Metal Film Oxidation Bulletin of Japan Institute of Metals, 29 (1990) 990-998
[9] M. Martin, E. Fromm Kinetics of Aluminium Film Oxidation measured by a modified Quartz Crystal Microbalance Proc. Int. Conf. on Metallurgical Coatings and Thin Films, ICMCT93, San Diego, California. Thin Solid Films, 236 (1993) 199-203
|