University of Minnesota
University of Minnesota: Department of Mechanical Engineering

Environmental Research Laboratory

Selected Abstracts

Electron Density and Energy Distributions in the Positive DC Corona: Interpretation for Corona-Enhanced Chemical Reactions
Junhong Chen and Jane H. Davidson

Electrons produced in atmospheric pressure corona discharges are used for a variety of beneficial purposes including the destruction of gaseous contaminants, and surface treatment. In other applications, such as electrostatic precipitators and photocopiers, unintended reactions such as ozone production and deposition of silicon dioxide are detrimental. In both situations, a kinetic description of the electron distribution in the corona plasma is required to quantify the chemical processes. In this paper, the electron density and energy distributions are numerically determined for a positiue dc corona discharge along a wire. The electron density distribution is obtained from the 1-D charge carrier continuity equations and Maxwell’s equation. The non-Maxwellian electron kinetic energy distribution is determined from the Boltzmann equation. The effects of wire size (10-1000 μm) and current density (0.1–100 μA/cm of wire) on number density and energy distribution of electrons are presented. With increasing current, the electron density increases, but the thickness of the plasma and the electron energy distribution are not affected. Smaller electrodes produce thinner plasmas and fewer, but more energetic electrons, than larger wires. The effect of electrode size on the electron-impact chemical reaction rate is illustrated by the rates of dissociation and ionization of oxygen and nitrogen.

Model of the Negative DC Corona Plasma: Comparison to the Positive DC Corona Plasma
Junhong Chen and Jane H. Davidson

A numerical model of the negative DC corona plasma along a thin wire in dry air is presented. The electron number density and electric field are determined from solution of the one-dimensional coupled continuity equations of charge carriers and Maxwell’s equation. The electron kinetic energy distribution is determined from the spatially homogeneous Boltzmann equation. A parametric study is conducted to examine the effects of linear current density (0.1-100 μA per cm of wire length), wire radius (10-1000 μm), and air temperature (293-800 K) on the distribution of electrons and the Townsend second ionization coefficient. The results are compared to those previously determined for the positive corona discharge. In the negative corona, energetic electrons are present beyond the ionization boundary and the number of electrons is an order of magnitude greater than in the positive corona. The number of electrons increases with increasing gas temperature. The electron energy distribution does not depend on discharge polarity.

Ozone Production in the Positive Corona Discharge
Junhong Chen and Jane H. Davidson

Corona discharge devices, such as indoor electronic air cleaners, photocopiers and laser printers, are sources of indoor ozone. Numerous experimental studies have sought methods to minimize ozone generation rates, but they do not elucidate the fundamental process of corona-enhanced ozone production. In Part I of this paper, a comprehensive numerical model of the ozone generation in clean, dry air by positive DC corona discharges is presented. This model includes a corona plasma model, a chemistry model, and a fluid dynamic model. The distribution of ozone and other gaseous products (e.g. NOx) is obtained in the neighborhood of the corona discharge wire. It is found that ozone and NOx are mainly distributed in a small region downstream of the wire and excited molecules (N2* and O2*) contribute more than 50% of the ozone production in corona discharges.

Ozone Production in the Negative DC Corona: The Dependence of Discharge Polarity
Junhong Chen and Jane H. Davidson

The rate of production and the spatial distribution of ozone in the negative DC corona discharge are predicted with a numerical model. The results are compared to prior experimental data and to results previously presented by the authors for the positive corona discharge. In agreement with experimental data, ozone production rate in the negative corona is an order of magnitude higher than in the positive corona. The model reveals that this significant difference is due to the effect of discharge polarity on the number of energetic electrons in the corona plasma. The number of electrons is one order of magnitude greater and the chemically reactive plasma region extends beyond the ionization region in the negative corona. The paper also extends our prior modeling effort to lower velocities where the Joule heating reduces ozone production. The magnitude of the reduction is characterized by a new dimensionless parameter referred to as the electric Damkohler's third number (DaIII-e).

Chemical Vapor Deposition of Silicon Dioxide by Direct-Current Corona Discharges in Dry Air Containing Octamethylcyclotetrasiloxane Vapor: Measurement of the Deposition Rate
Junhong Chen and Jane H. Davidson

Experiments in a positive-polarity, wire/plate electrode establish the effects of the concentration of octamethylcyclotetrasiloxane (150 to 1100 ppm) and the operating current (0.5 to 2.55 µA per cm of length of wire) on the rate of deposition of silicon dioxide on the high voltage wire. The wire is 100 µm radius tungsten and the wire-to-plate spacing is 1.5 cm. Analyses of the deposit with X-ray diffraction, energy dispersive X-ray spectroscopy, and X-ray photoelectron spectroscopy show that it is amorphous silicon dioxide. The deposition rate increases linearly with increasing silicone concentration and corona current. For the concentrations of silicone likely to present in indoor air, the gas-phase processes limit the rate of deposition.

A Global Model of Chemical Vapor Deposition of Silicon Dioxide by Direct-Current Corona Discharges in Dry Air Containing Octamethylcyclotetrasiloxane Vapor
Junhong Chen and Jane H. Davidson

A model of the electron distribution in direct current corona plasmas is combined with a global chemistry model and a two-dimensional transport model to predict the rate of chemical vapor deposition of silicon dioxide on the discharge wire in both positive and negative discharges in dry air containing octamethylcyclotetrasiloxane. The gas-phase chemistry includes reactions to form atomic oxygen (O) and additional global reactions to form gaseous silicon dioxide precursors by the impact reactions of electrons and atomic oxygen with silicone molecules. Surface chemistry is approximated by a single step global reaction from gaseous to solid silicon dioxide. The rate coefficient between atomic oxygen and octamethylcyclotetrasiloxane is estimated from prior experiments to be on the order of 10-12 cm3/molecule-s. The effects of discharge polarity, current, wire radius and air velocity (Peclet number for mass transfer) on the deposition rate are considered. Deposition rates can be minimized by using positive coronas instead of negative coronas for Peclet number less than 18.5. At higher Peclet numbers, the deposition rate is slightly higher in positive corona discharges, but devices used indoors should continue to use the positive corona in order to minimize the production of ozone. The deposition rate in the positive corona is relatively insensitive to air velocity for velocities from 0.044 to 10 m/s. However, it may be minimized by operating the corona with the lowest current that provides adequate performance (e.g. particle charging) and the smallest wire that provides adequate mechanical strength.

Effects of Low-Concentration Ozone on Activated Carbon and Implications for Removal of VOCs in Indoor Air
Michael L. Kingsleya, Jane H. Davidsona, Andrew Dallasb, Jon Jorimanb, Lefei Dingb

a) Department of Mechanical Engineering, University of Minnesota, 111 Church St. SE, Minneapolis, MN 55455
b) Donaldson Company, Corporate Technology, 9301 James Ave., Bloomington, MN 55431

The effects of ozone adsorption on the pore structure and surface chemistry of activated carbon derived from coconut shell are measured for ozone concentrations of 0.1, 1.7 and 7.0 ppm in dry (less than 5% RH) and humid (55% RH) air. Breakthrough experiments conducted using a UV absorption ozone detector prove that carbons with 60 and 80% activation levels are very effective at adsorbing ozone at low concentration. However, comparisons of x-ray photoelectron spectra of the virgin and exposed carbons indicate that surface functional groups form as a result of ozone reactions with the carbon surface. Evidence of loss of the carbon surface through gasification is found for carbons exposed to 1.7 and 7.0 ppm ozone. The surface area and pore volume, determined from nitrogen sorption isotherms, are reduced as a result of ozone exposure in dry air. In humid air, the formation of surface oxides is reduced but is still sufficient to significantly increase the water uptake of the carbon. Observed surface changes adversely affect the adsorption of toluene. In dry air, the 10% toluene breakthrough times of the exposed CTC60 and CTC80 carbon beds are 25% and 15% less than those of the virgin carbon beds. At 55% RH, the corresponding reductions in toluene breakthrough time are 73 and 78%.