Aromatic hydrocarbon oxidation 2. Uncertainty analysis

1. Aromatic hydrocarbon oxidation2. Uncertainty analysis Mike Pilling University of Leeds, UK MCM meeting Leeds

2. Oxidation mechanisms initiated by OH • Most NMHCs: • Abstraction or addition followed by addition of O2 to form peroxy radical. • Peroxy radical reacts with NO to form an alkoxy, which reacts to form HO2. Reaction with NO regenerates OH. • NO  NO2 from peroxy reactions  O3 • Aromatics: • HO-aromatic-O2 is short lived, regenerating reactants. Adduct is too short lived to react with NO under normal atmospheric conditions. • Intermediates are generally short lived cf parent MCM meeting Leeds

3. + OH/O2 HO2 HO2 NO2 NO HO2 O2 Products Products + O2 b-Hydroxy Peroxy Phenol Oxepin NO O2 HO2 O2 NO2 HO2 Epoxy-Oxy Peroxide Bicyclic Oxidation Routes for Toluene Also routes to: Benzaldehyde via abstraction MeBenzoquinone via 1,4 addition MCM1: Peroxide Bicyclic Route, predominantly producing butenedial + methylglyoxal MCM2: Oxepin Route, predominantly producing muconaldehyde type products MCM3.1 – latest mechanism MCM meeting Leeds

4. Mechanism for toluene oxidation • Reviewed up to ~2000 by Calvert et al. • Main contributions to mechanism from the groups of Atkinson, Becker and Seinfeld • ~6% via abstraction from –CH3 • Rest via HO-toluene-O2 adduct: • ~12% via retention of aromatic ring • Addition of O2 to form bridged bicyclic compound that leads to ring opening. • Role of NO not clear • Mass balance only ~50% MCM meeting Leeds

5. Subsequent key results from EUPHORE: Spectroscopic expts using DOAS • Glyoxal, formed from the formation of the bicyclic compound and ring opening is a primary product only. (Volkamer et al, J Phys Chem, 2001, 105, 7865) • Measurements of substituted phenol yields and demonstration of the need to work at low NOx (Volkamer et al , PCCP, 2002, 4, 1598) MCM meeting Leeds

6. EXACT consortiumEffects of oxidation of aromatic compounds in the troposphere (EU, Framework 5) • Kinetics of elementary reactions in initial stages (Bordeaux, Hanover) • Kinetics and mechanisms of secondary chemistry (Cork, Wuppertal) • Development and testing of overall mechanism. Design of EUPHORE experiments. (Leeds, Imperial College) • Synthesis of intermediates (Newcastle) • Secondary aerosol formation (Wuppertal, Imperial College) • Photochemical chamber studies at EUPHORE (Valencia, Wuppertal, Cork) • Coordination (Leeds) MCM meeting Leeds

7. Chamber measurements at Wuppertal and Cork on kinetics of intermediates a :10-12 cm3 molecule-1 s-1 vs butyl ether and 1,2,4 trimethylbenzene b : 2s errors c : 10-15 cm3 molecule-1 s-1 vs tetrahydrofuran NO3 reactions, relative to 2,3 dimethyl 3 butene (17, 15, 10)x10-11 cm3 s-1 Also extensive measurements of products in reactions of intermediates, especially of hydroxyarenes (Olariu et al) Photolysis rates and mechanisms for g –dicarbonyls (Thuener et al) MCM meeting Leeds

8. Toluene Oxidation Routes in MCMv3.1 Low ozone formation route Little ring opening along phenol route Successive addition of OH, NO3. Leads to formation of nitrophenols Ring opening routes are most active photochemically and dominate ozone formation MCM meeting Leeds

9. CEAM Labs, Valencia FTIR: Aromatics, O3, HCO2H,HCHO, HNO3 Absorption spectroscopy:O3 Chemiluminescence:NO DOAS:NO2, Glyoxal LIF:OH, HO2 GC-ECD:PAN, Methylglyoxal, PAN GC-FID: Aromatics HPLC/UV:Cresols, Benzaldehyde CO-Monitor:CO 2D-GC: carbonyls PFBHA:intermediates Derivatisation: Oxepin (Triazolin) Glyoxal, Methylglyoxal (Diaminobenzol) Filterradiometer:J(NO2) SMPS: Particle size distribution MCM meeting Leeds

10. Quantitative GCxGC of Toluene Oxidation Products Difficult to detect appropriate amounts of coproducts of glyoxal and Meglyoxal Measurements also made by GC/ECD Benzaldehyde Angelica lactone /oxopentanal Me-benzoquinone Toluene unknown Maleic anhydride MCM meeting Leeds

11. Test of [OH]LIF calibration Hydrocarbon concentration Inferred [OH] Guggenheim (1926) k' = ln (c2 / c1) / (t2 - t1) [OH]HC = (k' - kdil) / kOH+HC Measured [OH] (LIF) MCM meeting Leeds

12. Test of HO2 calibration HO2 concentration evolution in the dark and second-order decay analysis HCHO photolysis No NOx From decay analysis k(HO2+HO2) = 3.0 x 10-12 molecule-1cm3s-1 Literature (JPL 97-4) k(HO2+HO2) = 2.8 x 10-12 molecule-1cm3s-1 MCM meeting Leeds

13. Toluene Oxidation Routes in MCMv3.1 Low ozone formation route Little ring opening along phenol route Successive addition of OH, NO3. Leads to formation of nitrophenols Ring opening routes are most active photochemically and dominate ozone formation MCM meeting Leeds

14. Design of chamber experiments MCM meeting Leeds

15. Comparison of MCM3.1 to Toluene Chamber Experiment (27/09/01) • Conclusions: • - Ozone overpredicted • but OH is too low. Need • early OH source that • doesn’t produce O3 • NO2 is notrapidly • enough • - Co-products of glyoxal/ • Me glyoxal not detected • in sufficient concn MCM meeting Leeds

16. g-dicarbonyls. Photolysis (NO = 0) and ‘photosmog’ experiments (with NO)(Cork, Valencia measurements) photolysis photosmog MCM meeting Leeds

17. MCM v3.1 photolysis mechanism vs photolysis observations MCM meeting Leeds

18. Searching for an OH production route • Alkyl peroxy radicals isomerise / dissociate to from OH only at high T • Modification of the peroxy can lead to low T production of OH: e.g. CH3CO + O2 → CH3CO3* → OH + CH2CO2 CH3CO3* + M → CH3CO3 • Can such routes operate in aromatic chemistry? MCM meeting Leeds

19. EXACT-1 : Attempts to improve the model performance by including an NO2 aerosol sink /HONO source and an early source of OHAlternative mechanisms are also feasible, e.g. Volkamer, O3 + furanones MCM meeting Leeds

20. Current status of aromatic mechanisms • Mechanism underestimates total radical production rates by a factor of ~2 at short and long times. • At the same time, mechanism overestimates O3 formation – need route to radical formation that doesn’t give NO to NO2 conversion. • NOx removed from system more rapidly than mechanism indicates • Identification of glyoxal co-products ( g dicarbonyls) via GCxGC and GC/ECD with synthesis of targets – yields very low. • Photochemistry and photosmog experiments on g dicarbonyls are incompatible in terms of radical yields – need new chemistry. • Further extensive experiments on other aromatics – benzene, p-xylene, 1,3,5 trimethyl benzene, hydroxy aromatics, using both EUPHORE and small chambers for kinetics. Provide detailed mechanistic and kinetic data, but problems remain. • Need new detailed experiments, e.g. by laser flash photolysis or discharge flow on targeted intermediates: • OH and HO2 formation • O3 + g-dicarbonyls / furanones • Improved detection methods for glyoxal co-products MCM meeting Leeds

21. Comparison of ethene measurements and simulations Ethene experiments used to refine the auxiliary chamber mechanism 2s measurement uncertainty (grey bands) 2s uncertainties from Monte-Carlo simulations (error bars) MCM meeting Leeds

22. MCM meeting Leeds

23. Uncertainty contributions, ethene, low and high NOx MCM meeting Leeds

24. Morris method – (MOAT analysis), high NOx Effects of individual rate constants on peak O3 concentration MCM meeting Leeds