Updated Assessment of Microreactors for Improved Thermal Management of H2O2 Decomposition as a Fuel Cell Oxidant Aboard Unmanned Undersea Vehicles

Abstract

            The catalytic decomposition of hydrogen peroxide (H₂O₂) into water (H₂O) and oxygen (O₂) [H₂O_2(l) → H₂O_(l,g) + ½ O_2(g)] is notoriously susceptible to thermal runaway (Heat of Reaction: ΔH_rxn = -98 kJ/mol). However, liquid H₂O₂ has high oxygen density making it a particularly attractive oxidant for applications where air-independence and space limitations elicit additional power supply design challenges. Microchemical technologies have intrinsically higher surface-to-volume ratios (S/V) to potentially mitigate this problem by facilitating heat and mass transfer. This study further developed previous work on the capabilities of a micro-scale packed bed (MPB) reactor (radius 0.5 mm) to enhance thermal management and oxygen production during the multiphase decomposition of H₂O₂ as an oxidant for an unmanned undersea vehicle (UUV) solid oxide fuel cell (SOFC). Despite an order of magnitude increase in S/V over conventional reactors (S/V:  ~100 m²/m³), the micro-scale reactor channel (SV: 2254 m²/m³) generated significant heat rise (T_rise ~ 100 K) during simulated decomposition under convective cooling using the program COMSOL. The extension of surface area around the reactor channel (SV: 188,439 m²/m³) improved passive cooling, resulting in simulated MPB temperature rises less than 7 K and experimental MPB temperature rises less 4 K. S/Vs at least three orders of magnitude larger than conventional S/Vs were required to achieve thermal control over the reaction zone during H₂O₂ decomposition. Although thermal management was successfully accomplished with the addition of the high area heat exchanger, experimental oxygen production rates were lower than simulated oxygen production rates. The simulated MPB reactor assumed a homogenously dispersed multiphase product stream, whereas the experimentally observed product stream more closely resembled two-phase plug flow (the formation of oxygen bubbles was directly observed within the experimental MPB). In the experimental two-phase plug flow, oxygen bubbles situated on active catalyst sites likely prevented reactant from reaching the active sites, in turn contributing to lower oxygen production during experimentation. The presence of phosphorous based stabilizers in the concentrated H₂O₂ stocks probably caused deactivation of manganese dioxide (MnO₂) active sites over time, also lowering oxygen flow rates observed at the outlet. Recommendations are made to overcome the mass transfer limitations associated with oxygen bubbles occupying active sites. For example, using a thin layer of inert adhesive to hold catalyst particles in place down the length of the reactor channel could encourage uninterrupted flow and more readily dislodge oxygen bubbles. Overall these findings show improved thermal control during H₂O₂ decomposition using a micro-scale reactor relative to conventionally-sized or even miniature (mm-scale) reactors, oxygen production capabilities of micro-scale reactors during H₂O₂ decomposition, and suggests mechanisms for overcoming mass transfer limitations naturally present in multiphase reactions like H₂O₂ decomposition.