Due to the difficulty of H 2 storage, development of real time H 2 generators would be advantageous for portable fuel cells. In this paper, the real time production of H 2 using microdischarge devices is discussed with results from a computational investigation. Ar/NH 3 mixtures were studied using plug flow and two-dimensional models. Dissociation of NH 3 by electron impact and thermal processes produces H atoms which recombine to form H 2 . We found that for sandwich type microdischarges with a diameter of 300 µm, dissociation of NH 3 is approximately 95% by electron impact and 5% by thermal processes for a NH 3 mole fraction of 5%. Efficiency of conversion of NH 3 to H 2 is dependent on residence time in the discharge, mole fraction and geometry, as these properties determine the eV/molecule deposited into NH 3 . Conversion efficiencies (fraction of H in NH 3 converted to H 2 ) in excess of 83% are predicted for optimum conditions.

Articles you may be interested inLaser induced fluorescence of the ferroelectric plasma source assisted hollow anode dischargeIn chemical oxygen iodine lasers ͑COILs͒, oscillation at 1.315 m in atomic iodine ͑ 2 P 1/2 → 2 P 3/2 ͒ is produced by collisional excitation transfer of O 2 ͑ 1 ⌬͒ to I 2 and I. Plasma production of O 2 ͑ 1 ⌬͒ in electrical COILs ͑eCOILs͒ eliminates liquid phase generators. For the flowing plasmas used for eCOILs ͑He/ O 2 , a few to tens of torr͒, self-sustaining electron temperatures, T e , are 2 -3 eV whereas excitation of O 2 ͑ 1 ⌬͒ optimizes with T e = 1 -1.5 eV. One method to increase O 2 ͑ 1 ⌬͒ production is by lowering the average value of T e using spiker-sustainer ͑SS͒ excitation where a high power pulse ͑spiker͒ is followed by a lower power period ͑sustainer͒. Excess ionization produced by the spiker enables the sustainer to operate with a lower T e . Previous investigations suggested that SS techniques can significantly raise yields of O 2 ͑ 1 ⌬͒. In this paper, we report on the results from a two-dimensional computational investigation of radio frequency ͑rf͒ excited flowing He/ O 2 plasmas with emphasis on SS excitation. We found that the efficiency of SS methods generally increase with increasing frequency by producing a higher electron density, lower T e , and, as a consequence, a more efficient production of O 2 ͑ 1 ⌬͒.

Chemical oxygen-iodine lasers (COILs) oscillate on the P1∕22→P3∕22 transition of atomic iodine at 1.315μm by a series of excitation transfers from O2(Δ1). In electrically excited COILs (eCOILs), the O2(Δ1) is produced in a flowing plasma, typically He∕O2, at a few to tens of Torr. Many system issues motivate operating eCOILs at higher pressures to obtain larger absolute densities of O2(Δ1) for a given yield and to provide higher back pressure for expansion. In this paper, we discuss results from a computational investigation of O2(Δ1) production in flowing plasmas sustained at moderate pressures (⩽50Torr). Power deposition and flow rates were scaled such that in the absence of second order effects, yield should be constant and absolute O2(Δ1) production should scale linearly with pressure. We found in many cases that absolute O2(Δ1) production scaled sublinearly with pressure. Ozone is found to be one of the major quenchers of O2(Δ1) and its production increases with pressure. Gas heating also increases with increasing pressure due to exothermic three-body reactions. The gas heating reduces O3 production, increases O3 destruction and, for certain conditions, restores yields. With increasing pressure and increasing absolute densities of atomic oxygen and pooling reactions of O2(Δ1), quenching by these species also becomes important, though the influence of O-atom quenching can be controlled by managing the density of O atoms with additives. The yield of O2(Δ1) is also determined by discharge stability which becomes problematic at higher pressure.

Microsatellites with masses of tens of kilograms require only hundreds of micronewtons of thrust for station keeping and attitude control. Microdischarges (MDs) offer a compact way of generating such thrusts without using complex propulsion systems. In this paper, results from computational investigations of MDs sustained in Ar with tens of Torr back pressure are discussed with emphasis on conversion of discharge power into gas heating which can then be expanded in a nozzle to generate thrust. Typical cylindrical geometries have diameters of a few hundred micrometres and lengths of a few millimetres. We found that the gas temperature can exceed 1000 K for power densities of tens of kilowatts per cubic centimetre at back (upstream) pressures of tens of Torr. The nozzle length and location of the discharge in the MD channel are important from a gas dynamics viewpoint and so influence the incremental thrust (above that of the cold flow). Confining the discharge in the nozzle typically increases the peak gas temperature and the flow velocity, thereby potentially improving the performance of a MD used as a microthruster.

In conventional chemical oxygen-iodine lasers ͑COIL͒ the 1.315 m transition in atomic iodine is pumped by a sequence of reactions of I 2 and I with O 2 ͑ 1 ⌬͒ which is generated using liquid chemistry. Ongoing studies are investigating means to produce the O 2 ͑ 1 ⌬͒ precursor with an electric discharge ͑eCOIL͒ to enable a totally gas phase system. Due to the thermodynamic and power loading requirements, the plasma in eCOIL systems is sustained in a flow of a rare-gas diluent and the O 2 . In previous investigations, the scaling of production of O 2 ͑ 1 ⌬͒ was investigated using global-kinetics and one-dimensional ͑1D͒ models. It was found that the production of O 2 ͑ 1 ⌬͒ scaled linearly with energy deposition for moderate loadings ͑a few eV/ O 2 molecule͒. In this paper, these previous investigations are extended to two-dimensions using a plasma hydrodynamics model. The goal of this investigation is to determine if multidimensional considerations affect energy scalings for production of O 2 ͑ 1 ⌬͒. We found that O 2 ͑ 1 ⌬͒ production generally does scale linearly with energy loading, however, the saturation of O 2 ͑ 1 ⌬͒ production occurs at lower-energy loadings than predicted with global and 1D models. This trend is a result of the more accurately depicted and more localized energy deposition afforded by the two-dimensional model, and emphasizes the need for volumetrically uniform power deposition to optimize O 2 ͑ 1 ⌬͒ production.