print pageprinter friendly
Reports: ND952025-ND9: Activation of Chemical Reactions Through Corona Discharge

Alexandre Yokochi, PhD, Oregon State University

This project’s goal is to demonstrate that high energy efficiency chemical transformations processes can be driven by non-thermal electrical discharges.  While high power arc discharges have been used to drive chemical oxidations, energy efficiencies (e.g., moles reactant/J) need to be improved for practical implementations.  We hypothesized that higher efficiency corona discharge processes can be designed by operating in a mode where most of the electrical energy injected into the system is used for chemical conversions rather than converted to heat. This is achieved by ensuring that only electrons have high kinetic energies while ionized gases have low kinetic energies (i.e., a non-thermal plasma), and by minimizing excess electron energy by lowering voltages required for sustained corona discharge. Specific objectives include characterizing the electrical discharge through gases and condensed fluids and initial characterization of the products from two reactions, specifically C1 activation chemistry of methane and the kinetically fast oxidation of dibenzothiophenes by dissolved oxygen, and initial modeling of the reactions taking place. Potential applications of these processes include easier routes to convert natural gas to chemicals, including capture of stranded gas instead of flaring, and the development of practical extractive desulfurization processes with high performance towards recalcitrant sulfur compounds. Other related processes like gas phase CO2 reduction or advanced oxidation in water are potentially beneficial applications, but beyond the scope of the current study. Following initial work on proof of principle confirmation and reactor development work (described in the previous report), multi-emitter reactors were constructed, shown schematically in Figure 1, and the full electrical performance of a corona enabled reactor was determined, shown in Figure 2.  Several regimes of operation can be seen as reactor power is increased resulting in simultaneous changes in voltage and current.  At low total power, short lived pulsed discharges at high voltages are observed (Trichel pulse mode).  This is followed by a regime in which sustained current flows through the ionized gas (a glow discharge).  As applied power increases a mode in which a region of conductive gas containing high concentrations of electrons and ions (an arc discharge) is observed.  At the high end of the arc discharge regime, a large fraction of hot heavy ions is present, resulting in intense (white) thermal emission from the ionized gas, and likely decreasing the efficiency of the chemical conversion process. Figure 3 shows microphotographs of a) an incipient stable glow discharge developing into a more intense glow discharge b), and finally an arc discharge c). The operational parameters are dependent on the density and the electric polarizability of the matter in the region between the electrodes. Therefore, whilst for gases we can readily enter the glow discharge region, for discharges through liquids it is difficult to achieve stable operation due to the formation and collapse of bubbles.  Therefore, in this work, for gases, we employ operation in the high end of the glow and arc discharge regimes for gases, with high power arc discharges avoided due to the high plasma temperature (wasting energy as heat), and operation in pulsed mode is avoided due to the excessive energy of the free electrons.  For liquids, because we can’t readily enter a stable glow or arc discharge, work focuses on the high end of the pulse regime.  A proof of concept liquid reactor in pulsed discharge mode through a liquid is shown in Figure 4. Additionally, a systematic study of the conversion of methane in an electrical discharge reactor was carried out and is summarized in Figure 5.  The results present the conversion of a mixture of methane and air (25% CH4/75% air) at a flow rate of 60 SCCM, at room temperature. Initial examination of the data shows high conversion of methane to products, with an expected monotonically growing dependence of reactant consumption on applied power, and good selectivity of the products to ethylene and acetylene, similar to results shown in the previous progress report; the balance of the carbon forms longer hydrocarbons. Closer examination of the data shows the effects of the different operating regimes:  Very low conversion is achieved at low power, since the reactor is in pulsed discharge mode, therefore yielding only a few electrons with excessive and therefore wasted energy, and driving a limited number of chemical conversions.  At intermediate power levels conversion grows rapidly.  Since the reactor in glow discharge mode, many lower energy electrons are available, most of which drive chemical conversions.  Finally, a region of high conversion of methane with lower selectivity to ethylene/acetylene occurs when the reactor is in high power arc discharge mode, likely due to the locally high temperature of the gas.  This data is used for modelling of the system as a well-mixed CSTR with a significant bypass (0-D model) with the plasma reaction zone modelled through a 2-D Comsol model. Completion of the work on this project will focus on characterizing reactor performance when employing liquid phase reactants through the dibenzothiophene oxidation reaction.  This work has been delayed by the fact that the LC system we employ to characterize such systems has malfunctioned. To date, this work has formed the basis of several presentations including three at the 246th ACS and one at the 249th ACS national meetings, two at the 2013 and two at the 2014 AIChE annual meetings, one at ISCRE 2014, and one invited at WCCE9, and one review paper submitted to the Industrial and Engineering Chemical Research journal, in all of which ACS-PRF support was acknowledged.

Figure 1.  a) Diagram of a multi-emitter reactor and b) a photograph of three simultaneous discharges in the reactor.

Figure 2.  a) Reactor current and b) the resulting Current-Voltage performance of a microstructured corona discharge reactor. Insert shows current on a logarithmic scale.

Figure 3.  a) and b) glow discharges and c) arc discharge for a 1mm gap reactor.

Figure 4.  A (Trichel) pulse discharge through decane/DBT solution.

Figure 5.  Conversion of Methane (Circles) and selectivity towards ethylene (filled squares) and acetylene (open triangles) as a function of applied power to a multi-emitter reactor.