385018 Analysis of OH Radical Scavengers to Assess Chemical Reactions in Electrical Discharge Plasma Formed at a Gas-Liquid Interface

Sunday, November 16, 2014: 4:15 PM
A706 (Marriott Marquis Atlanta)
Kevin Hsieh, Chemical and Biomedical Engineering, Florida State University, Tallahassee, FL, Bruce R. Locke, Department of Chemical and Biomedical Engineering, Florida State University, FAMU-FSU College of Engineering, Tallahassee, FL, Robert Wandell, Department of Chemical and Biomedical Engineering, Florida State University, Tallahassee, FL and Huijuan Wang, School of the Environment and Safety Engineering, Jiangsu University, Zhenjiang, China

Low power pulsed electrical discharge (250 – 1000 mW) reactors that utilize water sprays with argon carrier gas have been shown to produce hydrogen peroxide (H2O2) from the recombination of hydroxyl radicals (OH∙) which are produced by the dissociation of water.  In reactors where the electrical discharge propagates at and along the interface between the gas and liquid phases, it is unknown whether the reactions to produce hydrogen peroxide occur in the gas or liquid phase. Of the many OH radical scavengers available for gas and liquid phases, there are only a few scavengers that will not alter the physical properties (e.g., conductivity, pH, water solubility, and viscosity) of the solution which in turn affect the discharge.  Dyes such as methylene blue are potentially good radical scavengers for the liquid phase because the decoloration of the dye is an in-direct estimate of the amount of OH radicals scavenged, the dye is non-volatile, and the dye does not interfere with the hydrogen peroxide measurements using UV-Vis spectroscopy.  However, liquid conductivity is affected by the formation of degradation products of the dye at very high dye concentrations.

Alcohols are also good candidates for OH radical scavengers.  Methanol, ethanol, and propanol are highly miscible with water and the liquid conductivity is not strongly affected by the degradation products.  However, lower molecular weight alcohols can diffuse into the gas phase.

In this work, methylene blue and ethanol were used as OH radical scavengers.  The concentration of methylene blue was 0.1 mM which had an initial liquid conductivity of 32 μS/cm.  The liquid flow rates for the dyes ranged from 0.75 to 5 ml/min.  The decoloration of the methylene blue was measured at 664 nm with UV-Vis spectroscopy.  The H2O2 was detected only in the liquid phase and was measured by a colorimetric test with TiOSO4 and UV-Vis spectrometry.  Ethanol concentrations ranging from 0.1 mM to pure ethanol were introduced into the reactor at 0.75 ml/min.  Proton nuclear magnetic resonance (1H NMR) showed the presence of an aldehyde product in both gas and liquid streams.  Ethanol and acetaldehyde were measured by gas chromatography with a flame ionization detector (GC-FID).  The liquid streams leaving the reactor for the cases with ethanol were injected directly into the GC-FID.  The gas stream exiting the reactor went through two cold traps and the condensed liquids were analyzed with GC-FID.  Methylene blue results showed that the decoloration rate increased with increasing liquid flow rate and the addition of methylene blue did not change the H2O2 concentration in comparison with pure water at all flow rates.  However, in alcohol solutions the amount of H2O2 decreased with increasing alcohol concentration.  Acetaldehyde was determined to be the major product from water-ethanol solutions and it was found primarily in the gas phase.  Ethanol feed concentrations at and above 1 M showed complete scavenging of the OH radical with no measurable H2O2.  Both OH radical scavenger results suggest that the majority of the OH radical is produced in the gas phase very close to the gas-liquid interface.  The OH radical production rate from acetaldehyde formation and H2O2 is approximately 2.72 x 10-7 ± 7.17 x 10-8 mol/s which is within experimental error of the OH radical production rate from H2O2 in the pure water cases, 2.27 x 10-7 ± 1.58 x 10-8 mol/s.  This work was supported by the National Science Foundation grants CBET-0932481 and CBET-1236225.


Extended Abstract: File Not Uploaded