Some refinements have been conceived for a proposed apparatus that would generate hydrogen (for use in a fuel cell) by means of chemical reactions among magnesium, magnesium hydride, and steam. The refinements lie in tailoring spatial and temporal distributions of steam and liquid water so as to obtain greater overall energy-storage or energy-generation efficiency than would otherwise be possible.

Water Would Be Injected at one end of a reactor bed containing MgH2 particles. Initially, the water would be in the form of steam as needed to start the reactions. Thereafter, liquid water would be injected, and one of the reactions, which is exothermic, would supply the necessary heat.
A description of the prior art is prerequisite to a meaningful description of the present refinements. The hydrogen-generating apparatus in question is one of two versions of what was called the “advanced hydrogen generator” in “Fuel-Cell Power Systems Incorporating Mg-Based H2 Generators” (NPO-43554), NASA Tech Briefs, Vol. 33, No. 1 (January 2009), page 52. To recapitulate: The apparatus would include a reactor vessel that would be initially charged with magnesium hydride. The apparatus would exploit two reactions:

  • The endothermic decomposition reaction MgH2 → Mg + H2, which occurs at a temperature ≥300 °C, and
  • The exothermic oxidation reaction MgH2 + H2O → MgO + 2H2, which occurs at a temperature ≥330 °C.

Once the initial heating was complete and both reactions under way, the endothermic reaction would be sustained by the heat generated from the exothermic reaction. For every mole of MgH2 oxidized, sufficient waste heat is generated to decompose an additional three moles of the hydride. As a consequence of these reaction ratios, the major reaction product is Mg, and the minor one MgO. Both have extremely low toxicity. MgH2 is easily recycled to Mg. In theory, no energy is required because regeneration produces enough heat to power the process. A practical system would not be 100-percent efficient so it would be expected that there would be a modest energy cost. The MgO can be safely and easily recycled in a magnesium-refining plant for less than the cost of smelting Mg because MgO is an intermediate product of that process. This concludes the description of the prior art.

The present refinements reflect the following ideas: In principle, the design and operation of the reactor could be optimized in the sense that the rate of generation of heat in the exothermic reaction, the rate of consumption of heat in the endothermic reaction, and the flow of heat from the exothermic to the endothermic reaction could be tailored to minimize the input energy needed to produce a given amount of H2. In turn, the reaction rates could be tailored by tailoring gradients of temperature and chemical composition, which, in turn, could be tailored through adjustments in rates of flow of steam and liquid water into the reactor, recognizing a need to adjust the rates because the gradients of temperature and chemical composition evolve as MH2 is consumed and MgO is generated.

For the purpose of illustrating the refinements, the figure schematically depicts a reactor that has a simple shape and inlets for steam and liquid water at one end only. One of the refinements would be to inject steam only at the beginning of operation, in no more than the quantity needed to initiate the exothermic reaction. The combination of heating and water vapor provided by steam would initiate both exothermic and endothermic reactions. With initiation of the exothermic reaction, sufficient heat is produced to allow the use of liquid water feed instead of steam. Some of the heat from the exothermic reaction would be consumed in heating the liquid water to steam. As operation continued, the steam front and the hydrogen-generating region would move farther into the region initially containing MgH2, leaving behind the MgO and Mg waste product. The rate of the exothermic reaction would be adjusted by adjusting the rate of injection of liquid water. If the reactor were of a more-complex configuration featuring multiple injection points, then the rates of injection of water at those points could be adjusted individually to obtain a more nearly optimum spatial and temporal distribution of temperature throughout the reactor.

This work was done by Andrew Kindler and Yuhong Huang of Caltech for NASA’s Jet Propulsion Laboratory. For more information, contact This email address is being protected from spambots. You need JavaScript enabled to view it.. NPO-46064