Nonthermal-Plasma Reactions of Dilute Nitrogen Oxide Mixtures: NO x -in-Argon and NO x + CO-in-Argon Xudong Hu, Gui-Bing Zhao, Ji-Jun Zhang, Linna Wang, and Maciej Radosz* Department of Chemical & Petroleum Engineering, University of Wyoming, Laramie, Wyoming 82071-3295 Analysis of conversion mechanisms for NO and N 2 O in Ar plasma suggests that NO is converted through the reaction Ar + + NO + e - f Ar + N + O, whereas N 2 O is converted through the reaction Ar + + N 2 O + e - f Ar + N 2 + O. A time-averaged lumped model developed on the basis of this analysis matches the experimental data. CO inhibits N 2 O conversion but not NO conversion. However, parts-per-million levels of CO affect neither N 2 O nor NO conversion. Compared to N 2 plasma, which produces a weak streamer glow discharge and a small temperature increase along the reactor, Ar plasma produces a strong streamer discharge and a small temperature decrease along the reactor. Introduction Flue gas streams contain parts-per-million levels of pollutants, such as NO x and SO x , that ideally are removed or converted to benign species prior to dis- charge. One way to convert NO x to nitrogen and oxygen is to expose the flue gas stream to electric discharges capable of generating radicals, ions, and excited mol- ecules, which, in turn, activate the pollutants and convert them to benign stable species. Such a reactive mixture containing radicals, ions, and excited molecules in an otherwise neutral gas is referred to as plasma. If a potential difference is applied to plasma, the electric field will impart energy to the charged particles. The electrons, because of their small mass, will be im- mediately accelerated to a higher degree between the collisions than the heavier ions. If the pressure is low or the electric field is high or both, the electrons and the ions will, on average, have a kinetic energy higher than the energy corresponding to the random motion of the molecules. In plasma in such a state, usually referred to as nonequilibrium plasma, the highly ener- getic electrons are capable of ionizing and dissociating the neutral species at high rates even though the bulk gas temperature is quite low. Thus, it is said that such “cold nonequilibrium” discharges are capable of high- temperature chemistry at low temperatures. If, on the other hand, the pressure is so high that the charged particles do not move far between the collisions or the electric field is weak or both, the kinetic energy of the charged particles is not significantly different from the kinetic energy of the neutral species. Such plasma is called equilibrium plasma. In this work, we use a reactor in which nonthermal, nonequilibrium plasma is generated by a pulsed corona reactor (PCR). A PCR converts dilute NO, NO 2 , and N 2 O in nitrogen into the environmentally benign gases N 2 and O 2. 1-4 The electrons generated in a PCR collide with the carrier gas and create chemically active species that initiate NO x reactions. Although much experimental and modeling work has been done on NO and N 2 O conver- sion in a diatomic background gas, such as N 2 , little work has been done on the conversion of NO x in a single- atom background gas, such as Ar, 5,6 which produces fewer active species. Thus, using argon as a background gas might facilitate understanding of the chemical reactions in a corona discharge reactor, especially electron-molecule impact reactions. An additional justification for exploring argon plasma is that it has found a practical application in the treatment of flue gas by a radical injection technique as reported by Chang et al. 12 Their results showed a very high rate of NO x destruction (85%) in the combus- tion exhaust gas. Ohkubo et al. 25 used a small amount of Ar introduced into the flue gas stream through the corona discharging zone. Their results showed that the corona discharge characteristics and modes are signifi- cantly influenced by argon. Such work requires kinetic models. An example of a kinetic model is a time-averaged lumped model, initially proposed by Hu et al. 6 and then improved by Zhao et al., 7 that was found to represent the NO, NO 2 , and N 2 O concentration evolution in nitrogen. The primary goal of this work is to extend this model to another type of carrier gas, such as argon, also studied by Maier, 5 and to test it on new experimental data taken in this work. The secondary goal of this work is to understand the effect of another minor component, such as CO, as an example of a flue gas component that can alter the NO x conversion. 2,9 The NO x conversion study in this paper refers to either NO or N 2 O conversion. Experiment The experimental apparatus and measurement pro- cedures were described in detail previously. 8 In a brief overview, the four-tube PCR used in this work consists of a high-voltage power supply, a control unit, and a pulser/reactor assembly. The high-voltage supply con- trols the pulsed power delivered to the reactor. The pulser/reactor assembly contains the pulsed power generator and the pulsed corona discharge tubes. The reactor has UV-grade quartz windows for diagnostics and plasma observation. In all of the experiments described in this work, only four of 10 tubes are wired for plasma generation. The corona power is calculated as the product of the pulse voltage (V) and current (I); the energy is the time integral of power (∫VI dt). The power consumed can also be calculated as the product * To whom correspondence should be addressed. E-mail: radosz@uwyo.edu. Tel.: 307-766-2500. Fax: 307-766-6777. 7456 Ind. Eng. Chem. Res. 2004, 43, 7456-7464 10.1021/ie0495731 CCC: $27.50 © 2004 American Chemical Society Published on Web 10/12/2004