Understanding the Failure of Novel MEMS Microplasma Reactors

Compact fuel cells are the logical progression for portable power as they have much higher energy densities than battery technology. However, fuel cells have not dominated the portable power market as they require a readily available source of hydrogen. Hydrogen gas is generally produced at large scale steam reforming plants. Transporting the hydrogen from these plants is impractical for portable applications due to its large volume even when compressed. Instead a better option is to produce the hydrogen locally and create a fuel reformer-fuel cell system. Micro-reactors offer the unique advantages of being able to process hydrocarbons adjacent to a fuel cell.
Microplasma reactors developed from standard semiconductor fabrication techniques have been tested in reforming various hydrocarbons. Plasmas were used as the reaction media as they have many advantages over common catalytic reactions. In contrast to conventional catalytic reactions, plasmas have instantaneous startup times, can operate at room temperatures, and eliminate the possibility of sulfur poisoning and coking.
The microplasma reactors analyzed in this study were microhollow cathode discharge (MHCD) devices. The reactors each have three key layers to produce a plasma. The three layers are an n-doped silicon cathode, a silicon dioxide dielectric, and a nickel/titanium anode. The silicon has a trench etched into it where the plasma reaction occurs. A high voltage (~500V) is established across the anode and cathode to develop the plasma in the trench. 36 variations of reactors were fabricated, with deviations in the oxide thickness (2, 4 μm), trench depth (25, 100, 250 μm), trench width (25, 100, 250 μm), and with and without a patterned anode.
Experiments with these reactors were successful in showing the feasibility of a portable microplasma fuel reformer, but the devices regularly had lifetime issues. Devices would stop working with in an operation time of a few seconds to a few hours. The inconsistency as well as short lifetime pressed an investigation to why these reactors fail.
The electrical and physical properties of the thirteen failed microplasma reactors have been investigated with the goal of determining what caused their malfunction. These devices have been tested by recording their electrical characteristics under loads ranging from -20 to 20 volts. Then they were imaged using an optical microscope as well as a scanning electron microscope (SEM) to determine the point of failure. Energy-dispersive X-ray spectroscopy (EDS) was used to identify the distribution of elements in the channel, anode surface, and cross-section of the microreactor. Devices that hadn’t performed a reaction were also investigated in the same manner as the failed thirteen reactors.

Preliminary SEM imaging suggests a number of chemical and physical changes occurring during the operation of these reactors. The most prominent of these occurrences was the formation or re-deposition of silicon and silicon dioxide on the inner walls of the microchannel. Further inspection using EDS revealed that the nickel anode was also dispersed throughout the channel, indicating an electrical bridge is likely being developed from the cathode to the anode.
Further investigation of the reactors is required to determine if the deposition of these materials is a function of the device dimensions, system parameters (reactants, products, flow rates, etc.), and/or the power applied during the plasma reaction. This analysis will enable us to improve on the design of the microplasma reactor and furthermore allow for the development of a portable fuel reformer-fuel cell system.