Metal-doped NaAlH₄ is being studied as a “model” reversible hydrogen storage material from the class of materials referred to as alanates.  With few exceptions, only single transition metal and lanthanide series catalysis has been studied; yet, interesting work by the PI on bimetallic catalysis showed some synergy between the two metals.

It was hypothesized that the behavior of bimetallic catalyzed NaAlH₄ could be understood in terms of the electron sharing ability of electropositive early transition metals, such as Sc, Ti, V, Zr and Hf being enhanced when co-doped with the more electronegative electron-rich late transition metals, such as Fe, Co, Ni, Cu, Ru, Rh, Pd and Ag, based on the so called “metal-metal bond polarity” concept. According to that concept, it might be that NaAlH₄ is forming a heterodimetallic complex, for example, with Ti-Fe, which makes it more reactive and hence facilitates the release and uptake of hydrogen more readily than when using Ti or Fe alone as catalysts, as shown recently by the PI.

To shed some light on this hypothesis, the objective of this research was to explore the effect of early and late transition metal bimetallic catalysis on the dehydrogenation and hydrogenation of NaAlH₄ in terms of performance and synergy.  A systematic experimental study was carried out in four phases: I) temperature programmed desorption (TPD) screening, II) dehydrogenation and hydrogenation cycling, III) ancillary experiments and modeling, and IV) analysis and correlation. Some key findings from this study are presented and discussed below.

At the outset, it is necessary to define what is meant be synergistic behavior. In the initial work done by the PI, synergistic behavior was denoted by a bimetallic catalyzed sample of NaAlH₄ exhibiting better dehydrogenation kinetics than a single metal catalyzed sample of NaAlH₄, with the following stipulations. In each case, the total catalyst content was the same, the single metal catalyst was always one of the bimetallic catalysts, and the experimental conditions were the same. For example, the bimetallic catalyzed 3 mol% Zr-1 mol% Fe system behaved synergistically whenever its TPD curve resided to the left of the TPD curve of the single metal catalyzed 4 mol% Zr system, with Zr necessarily being the better of the two metal catalysts. Numerous instances of synergistic behavior were observed based on that definition.  However, that definition was thought to be too restrictive and difficult to interpret.

A less restrictive and easier to understand definition of synergistic behavior was envisioned based on the following provisions. Synergistic behavior occurred still whenever a TPD curve of a bimetallic catalyzed sample resided to the left of a TPD curve of a single metal catalyzed sample. However, the total catalyst content of the more active metal was the same in both samples, but with one of the samples containing only the single metal catalyst and the other one containing a small amount of the second less active metal catalyst. For example, based on this new definition, samples of NaAlH₄ bimetallic catalyzed with Zr and Fe exhibited significant, synergistic and sustained catalytic activity during hydrogenation (charge) and dehydrogenation (discharge) cycling. The results in Figure 1 illustrate this behavior based on the new definition. The upper figure shows TPD curves at 5 ^oC min^-1 and the lower figure shows parity plots for doped, ball milled, and uncycled and cycled samples of NaAlH₄ containing 1 mol% Zr as a single metal catalyst and 1 mol% Zr-3 mol% Fe as a binary metal catalyst. The wt% of hydrogen released was normalized to the weight of NaAlH₄, after accounting for that consumed during doping according to the following reaction: 3NaAlH₄+MCl₃ -->3NaCl+M+3Al+6H₂, where M is the metal catalyst.

Taking into account that the PI has shown even 4 mol% Fe to be a relatively inactive catalyst for NaAlH₄ dehydrogenation, the synergistic behavior observed for this system was very distinct and sustaining, even through five dehydrogenation and hydrogenation cycles.  The parity plot clearly shows this marked synergism associated with adding 3 mol% Fe to the sample. Other combinations of metals and concentrations exhibited similar synergistic behaviors.

Based on evidence from the literature and observed enhanced performance and synergistic trends with the Zr-Fe system in this work, and with the Ti-Fe, Zr-Fe and Ti-Zr co-doped NaAlH₄ systems shown elsewhere by the PI, the following suppositions were offered to explain these behaviors. The dominant effect on synergy was perhaps associated with the propensity of the early and late binary metal pairs to form higher polarity metal-metal bonds as they move apart on the periodic table. The dominant effect on performance was perhaps associated with the propensity of the more electropositive early transition metal to interact more favorably with the NaAlH₄, with its d-orbital electrons being more facile.  In conclusion, the hypothesis was perhaps correct, at least for these bimetallic systems; clearly, more research needs to be done.