EMSL: Russell Tonkyn's Publications

Grieves GA, NG Petrik, J Herring-Captain, B Olanrewaju, A Aleksandrov, RG Tonkyn, SE Barlow, GA Kimmel, and TM Orlando. 2008. "Photoionization of Sodium Salt Solutions in a Liquid Jet." Journal of Physical Chemistry C 112(22):8359-8364. doi:10.1021/jp7102534 Abstract A liquid microjet was employed to examine the gas/liquid interface of aqueous sodium halide (Na+X-, X=Cl, Br, I) salt solutions. Laser excitation at 193 nm produced and removed cations of the form H+(H2O)n and Na+(H2O)m from liquid jet surfaces containing either NaCl, NaBr or NaI. The protonated water cluster yield varied inversely with increasing salt concentration, while the solvated sodium ion cluster yield varied by anion type. The distribution of H+(H2O)n at low salt concentration is identical to that observed from low-energy electron irradiated amorphous ice and the production of these clusters can be accounted for using a localized ionization/Coulomb expulsion model. Production of Na+(H2O)m is not accounted for by this model but requires ionization of solvation shell waters and a contact ion/Coulomb expulsion mechanism. The reduced yields of Na+(H2O)m from high concentration (10-2 and 10-1 M) NaBr and NaI solutions indicate a propensity for Br- and I- at the solution surfaces and interfaces. This is supported by the observation of multiphoton induced production and desorption of Br+ and I+ from the 10-2 and 10-1 M solution surfaces.

Disselkamp RS, RG Tonkyn, YH Chin, and CHF Peden. 2006. "Developing Multiple-Site Kinetic Models in Catalysis Simulation: A Case Study of 02+2N0 ↔ 2 NO2 Oxidation-Reduction Chemistry on Pt(100) Catalyst Crystal Facets." Journal of Catalysis 238(1):1-5. doi:10.1016/j.jcat.2005.11.031 Abstract It is generally recognized that developing a kinetic model for a supported catalyst is difficult since multiple site types exist. These sites can arise from a distribution of crystal facets (e.g., (100), (110), etc.) each with their unique intrinsic site types (e.g., atop, bridge, hollow, etc.). Additional complexities arise from non-basel plane site types (defect, edge, corner, etc.), all whose differing lateral interaction energies may be coverage dependent for each site pairwise interaction. To demonstrate the complexities that develop even for a greatly simplified system, we examine a multiple site kinetic model of the reaction 2NO + O2 - 2NO2 on an ideal Pt(100) catalyst. A model of the Pt(100) surface is adopted where atop, bridge, and 4-fold hollow sites are responsible for O2, NO, and NO2 chemisorption to form Pt-O, Pt-NO, and Pt-NO2 species. In our kinetic scheme, equilibrium is assumed for O2, NO, and NO2 chemisorption due to their high sticking coefficients (all > 0.1). A single rate determining step of the Langmuir-Hinshelwood type was chosen to describe the oxidation of NO on platinum via the reaction PtH,A,B-O + PtH,A,B-NO - PtH,A,B + PtH,A,B-NO2, where H, A, and B represent 4-fold hollow, atop, and bridge sites. Equal kinetic parameters for all site combinations were assumed to exist and, in part, taken from the literature to be AH+=83 kJ/mol and AS+=20 J/K mol. The exercise here is largely hypothetical but offers insight into how more detailed kinetic models may be developed, such as through the use of reaction velocity matrices, a concept introduced here. Specifically for this system, the model yielded insight into NOx chemistry on Pt(100) in that it predicted that the greatest reaction velocities (forward and reverse) occurred via the reaction Pt-O(atop) + Pt-NO(bridge) A Pt(atop) + Pt-NO2(bridge). We believe the framework of a site-specific modeling scheme presented here is an important starting point for future site-specific microkinetic modeling. In particular, a definition and description of use of surface coverages, reaction rate coefficients, and computed reaction velocity matrices are presented.

Tonkyn RG, RS Disselkamp, and CH Peden. 2006. "Nitrogen Release from a NOx Storage and Reduction Catalyst." Catalysis Today 114(1):94-101. doi:10.1016/j.cattod.2006.02.005 Abstract In a NOx storage and reduction (NSR) catalyst the release and reduction of NOx occurs over a very short period. The speed of the NOx release and reduction creates difficulties in analyzing the chemistry using normal analytical techniques, which are typically better suited to slower, steady state studies. We have investigated the time dependence of NO, NO2, NH3, N2O and N2 released by an NSR catalyst using a combination of FTIR and gas chromatographic techniques. Nitrogen was detected with the GC by using He rather than N2 as the background gas. The FTIR was used not only to monitor NO, NO2, NH3 and N2O, but also to establish cycle-to-cycle reproducibility. Under these conditions we used the GC to sample the effluent at multiple times over many lean-rich cycles. To the extent that the chemistry was truly periodic and reproducible, we obtained the time dependence of the release of nitrogen after the lean-to-rich transition. Similar information was obtained for O2, H2 and N2O. Combining the FTIR and GC data we obtained good cycle averaged nitrogen balances, indicating that all the major products were accounted for.

Zhao H, RG Tonkyn, SE Barlow, BE Koel, and CHF Peden. 2006. "Catalytic oxidation of HCN over a 0.5% Pt/Al2O3 catalyst." Applied Catalysis. B, Environmental 65(2006):282-290. doi:10.1016/j.apcatb.2006.02.009 Abstract The adsorption of HCN on, its catalytic oxidation with 6% O2 over 0.5% Pt/Al2O3, and the subsequent oxidation of strongly bound chemisorbed species upon heating were investigated. The observed N-containing products were N2O, NO and NO2, and some residual adsorbed N-containing species were oxidized to NO and NO2 during subsequent temperature programmed oxidation. Because N-atom balance could not be obtained after accounting for the quantities of each of these product species, we propose that N2 and was formed. Both the HCN conversion and the selectivity towards different N-containing products depend strongly on the reaction temperature and the composition of the reactant gas mixture. In particular, total HCN conversion reaches 95% above 250 C. Furthermore, the temperature of maximum HCN conversion to N2O is located between 200 and 250 C, while raising the reaction temperature increases the proportion of NOx in the products. The co-feeding of H2O and C3H6 had little, if any effect on the total HCN conversion, but C3H6 addition did increase the conversion to NO and decrease the conversion to NO2, perhaps due to the competing presence of adsorbed fragments of reductive C3H6. Evidence is also presented that introduction of NO and NO2 into the reactant gas mixture resulted in additional reaction pathways between these NOx species and HCN that provide for lean-NOx reduction coincident with HCN oxidation.

Zhao H, RG Tonkyn, SE Barlow, CHF Peden, and BE Koel. 2006. "Fractional factorial study of HCN removal over a 0.5% Pt/Al₂O₃ catalyst: effects of temperature, gas flow rate, and reactant partial pressure ." Industrial and Engineering Chemistry Research 45(3):934-939. doi:10.21/ie048777e Abstract Fractional factorial design was used to determine which factors have significant effects on the HCN (hydrogen cyanide) oxidation reaction over 0.5% Pt/Al₂O₃ under lean conditions. We conclude that the reaction temperature and gas-hourly space velocity (GHSV) have significant effects on the HCN conversion, while no significant effects are caused by the presence of either NO (nitric oxide) or C₃H₆ (propene). A central composite design was used to study the effects of temperature and GHSV on HCN conversion, C₃H₆ conversion and NOx selectivity. Based on a second polynomial equation model, regression analysis was used to study the significance of each variable term and derive equations for each response. Our results show that HCN conversion was significantly affected by temperature (X3), GHSV (X4), a temperature polynomial term (X32), and a temperature and GHSV interaction term (X3X4). HCN conversion decreased with increasing values of GHSV and increased with increasing temperature, up to a transition temperature that depends on the GHSV value. The variables of temperature (X3), GHSV (X4), and the temperature polynomial term (X32) have significant effects on both C₃H₆ conversion and NOx selectivity, but in these two cases the interaction of temperature and GHSV was not significant. Contour plots of HCN conversion, C₃H₆ conversion, and NOx selectivity versus temperature and GHSV were constructed from an analysis of the measured data, and these plots can be utilized to estimate HCN conversion, C₃H₆ conversion, and NOx selectivity over the range of temperatures and GHSV investigated. Optimum catalyst operation is described by high HCN conversion and low NOx selectivity. These results show C and o that the highest HCN conversion was achieved at temperatures above 250 relatively low GHSV values, while low NOx selectivity was best achieved at a C.o temperature of 215

Barlow SE, JH Kwak, CHF Peden, J Szanyi, RG Tonkyn, G Singh, K Stork, JW Hoard, B Cho, DJ Brooks, and S Nunn. 2004. "Plasma Catalysis for NOx Reduction from Light-Duty Diesel Vehicles." Chapter III.G. in Advanced Combustion Engine R&D: 2003 Annual Progress Report, ed. G Singh, pp. 129-136. Department of Energy, Washington, DC. Abstract This annual report reviews FY 2003 progress of a program aimed at the development of a novel plasma/catalyst technology for the remediation of NOx under lean (excess oxygen) conditions, specifically for compression ignition direct injection (CIDI) diesel engines that have significant fuel economy benefits over conventional stoichiometric gasoline engines. Our previous work has shown that a non-thermal plasma in combination with an appropriate catalyst can provide NOx emission reduction efficiency of 60-80% using a simulated diesel exhaust. Based on these levels of NOx reduction obtained in the lab, a simple model was developed in this program that allows for the estimation of the fuel economy penalty that would be incurred by operating a plasma/catalyst system. Results obtained from this model suggest that a 5% fuel economy penalty is achievable with the then current (FY2000) state-of-the-art catalyst materials and plasma reactor designs. In this last year, we have continued to focus on (1) improving the catalyst and plasma reactor efficiencies for NOx reduction, (2) studies to reveal important details of the reaction mechanism(s) that can then guide our catalyst and reactor development efforts (focus 1), and (3) evaluating the performance of prototype systems on real engine exhaust. While studies of the effects of the plasma on PM in real diesel engine exhaust is meant to be part of the program, this year we did not conduct any experiments along these lines due to the major effort required to carry out the engine testing (focus 3).

Barlow SE, JH Kwak, CHF Peden, J Szanyi, RG Tonkyn, K Howden, JW Hoard, B Cho, SJ Schmieg, DJ Brooks, S Nunn, and P Davis. 2004. "Plasma Catalysis for NOx Reduction from Light-Duty Diesel Vehicles." In Advanced Combustion Engine Research & Development, 2004 Annual Progress Report, pp. 180-183. US Department of Energy, Energy Efficiency and Renewable Energy, Washington DC. Abstract The control of NOx (NO and NO2) emissions from so-called ‘lean-burn’ vehicle engines remains a challenge. In this program, we have been developing a novel plasma/catalyst technology for the remediation of NOx under lean (excess oxygen) conditions, specifically for compression ignition direct injection (CIDI) diesel engines that have significant fuel economy benefits over conventional stoichiometric gasoline engines. Program efforts included: (1) improving the catalyst and plasma reactor efficiencies for NOx reduction; (2) studies to reveal important details of the reaction mechanism(s) that can then guide our catalyst and reactor development efforts; (3) evaluating the performance of prototype systems on real engine exhaust; and (4) studies of the effects of the plasma on particulate matter (PM) in real diesel engine exhaust. Figure 1 is a conceptual schematic of a plasma/catalyst device, which also shows our current best understanding of the role of the various components of the overall device for reducing NOx from the exhaust of a CIDI engine. When this program was initiated, it was not at all clear what the plasma was doing and, as such, what class of catalyst materials might be expected to produce good results. With the understanding of the role of the plasma (as depicted in Figure 1) obtained in this program, faujasite zeolite-based catalysts were developed and shown to produce high activity for NOx reduction of plasma-treated exhaust in a temperature range expected for light-duty diesel engines. These materials are the subject of a pending patent application, and were recognized with a prestigious R&D100 Award in 2002. In addition, PNNL staff were awarded a Federal Laboratory Consortium (FLC) Award in 2003 “For Excellence in Technology Transfer”. The program also received the DOE’s 2001 CIDI Combustion and Emission Control Program Special Recognition Award and 2004 Advanced Combustion Engine R&D Special Recognition Award.

Hoard J, and RG Tonkyn. 2003. "Two-Stage Plasma-Catalysis for Diesel NOx Emission Control." Journal of Advanced Oxidation Technologies 6(2):158-165. Abstract Plasma discharges in diesel simulated exhaust gas oxidize NO primarily to NO2 while formaing aldehydes andother partial-oxidation products from hydrocarbons. Appropriate catalysts can react NOx and aldehydes in the presence of oxygen, producing nitrogen.

Kaspar TC, AC Tuan, RG Tonkyn, WP Hess, JW Rogers, Jr., and Y Ono. 2003. "The Role of O(1D) in the Oxidation of Si(100)." Journal of Vacuum Science and Technology B--Microelectronics and Nanometer Structures 21(2):895-899. Abstract Oxidation of silicon with neutral atomic oxygen species generated in a rare gas plasma has recently been shown to produce high-quality thin oxides. It has been speculated that atomic oxygen in the first excited state, O(1D), is a dominant reactive species in the oxidation mechanism. In this study, we investigate the role of O(1D) in silicon oxidation in the absence of other oxidizing species. The O(1D) is generated by laser-induced photodissociation of N2O at 193 nm. We find that, at 400?C, O(1D) is effective in the initial stages of oxidation, but the oxide growth rate falls dramatically past 1.5 nm. Oxide films thicker than 2 nm were unachievable regardless of oxidation time or N2O partial pressure (0.5-90 mTorr), indicating O(1D) cannot be a dominant reactive species in thicker oxidation mechanisms. We suggest that quenching of O(1D) to O(3P) (ground state) during diffusion through thicker oxides results in drastically slower oxidation kinetics. In contrast, oxidation with a vacuum ultraviolet (VUV) excimer lamp operating at 172 nm resulted in oxide thicknesses up to 4 nm. Thus, other species produced in plasmas and excimer lamps, such as molecular and atomic ions, photons, and free and conduction band electrons, play a dominant role in the rapid oxidation mechanism of thicker oxides (> 2 nm).

Tonkyn RG, SE Barlow, and J Hoard. 2003. "Reduction of NOx in Synthetic Diesel Exhaust via Two-Step Plasma-Catalysis Treatment." Applied Catalysis. B, Environmental 40(3):207-217. Abstract Significant reduction of NOx in synthetic light duty diesel exhaust has been achieved over a broad temperature window by combining atmospheric plasma with appropriate catalysts. The technique relies on the addition of hydrocarbon reductant prior to passing the exhaust simulant through a non-thermal plasma and a catalyst bed. The observed chemistry in the plasma includes conversion of NO to NO2 as well as the partial oxidation of the hydrocarbon. Despite driving the NO oxidation to completion, the overall NOx reduction has a maximum of less than 80%, with this maximum obtained only at high energy input into the plasma, high concentration of hydrocarbon reductant and low space velocity. We present data in this paper illustrating that a multiple stage treatment strategy, whereby two or more plasma-catalyst reactor stages are utilized in series, can increase the maximum NOx conversion obtainable. Alternatively this technique can reduce the energy and/or hydrocarbon requirements for a fixed conversion efficiency. When propene is used as a reductant, the limiting reagent for the overall process is most likely acetaldehyde. The data suggest that acetaldehyde is formed in concert with NO oxidation to NO2 in the plasma stage. The limited NOx reduction efficiency attained in a single stage, even with excess energy, oxygen content and/or hydrocarbon to NOx ration is well explained by this hypothesis, as is the effectiveness of the multiple stage treatment strategy. We present data here illustrating the advantage of this approach under a wide variety of conditions.