2. Methyl Bromide : Spectroscopic line parameters in the 10-μm region D. Jacquemart 1, N. Lacome 1, F. Kwabia-Tchana 1, I. Kleiner 2 1 Laboratoire de Dynamique, Interactions et Réactivité; Université Pierre et Marie Curie- Paris6, CNRS, UMR 7075, France 2 Laboratoire Inter-Universitaire des Systèmes Atmosphériques; Universités Paris 12 et Paris 7, CNRS, UMR 7583, FRANCE

3. → Atmospheric trace gas (≈ 10 pptv) of both natural and anthropogenic origins (oceanic emission, biomass burning, leaded gasoline, agricultural pesticide …) → Deadly toxic gas for human and animal life when exposed to high concentration → Major contributor to stratospheric bromine which participates to ozone destruction Methyl Bromide (CH 3 Br)

4. → Previous works concern mainly line positions analysis (see Graner JMS 1981;90:394-438) → Two recent works on line positions and intensities in the 7-μm spectral region Spectroscopic line parameters in literature (Kwabia Tchana et al. JMS 2004;228:441-52 ; Kwabia Tchana et al. JMS 2006;235:132-43) → No spectroscopic data is available in atmospheric database such as HITRAN or GEISA → No work on broadening coefficients

5. Experimental conditions for spectra recorded around 10 μm → Rapid scan interferometer Bruker IFS 120 HR (LADIR, Paris) (Δmax = 450 cm; FWHM =1.1  10  3 cm  1 ) Absorbing sample Natural CH 3 Br50.54 % of CH 3 79 Br 49.46 % of CH 3 81 Br Stated purity99.50 % Experimental conditions S/N ratio  100 _________________________________________________________________ _ # CH 3 Br pressure N 2 pressure Temperature Absorption path (mbar) (mbar) (K) (cm) ________________________________________________________________ __ 1 0.4712 0 298.15 415 2 0.8745 0 297.15 415 3 4.738 0 298.15 415 4 7.200 0 298.15 30 5 2.030 25.30 297.55 415 6 3.376 32.90 296.45 415 ________________________________________________________________

8. Preliminary work → Phase correction for each spectrum (Mertz method) → Determination of an average effective iris radius → Wavenumber calibration using NH 3 transitions and HITRAN2004 wavenumbers as etalon = 1.789(40)×10 –6 scattering (1SD) of 0.04×10 –3 cm –1 at 1000 cm –1

9. Line parameters measurement for transitions having J and K ranging from 0 to 55 and from 0 to 9 → Use of a multispectrum fitting procedure (Eur Phys J D 2001;14:55-69.) Position, intensity and broadening coefficients of a same line are constrained to be the same during the simultaneous fit of the six spectra. Use of a Voigt profile. For broadening coefficients we assumed that: → 1200 transitions fitted between 880 and 1050 cm  1 of both CH 3 79 Br and CH 3 81 Br

10. Two models have been used to analyze measured line intensities → Treatment using the eigenvectors as a linear combination of the zero order basis wavefunction (ℓ-type interactions) (Tarrago G, Delaveau M. Triad v n (A 1 ), v t (E), v t’ (E) in C 3v Molecules: Energy and Intensity Formulation (Computer Programs). J Mol Spectrosc 1986;119:418-25 ) → Classical Herman Wallis treatment with |v,ℓ,J,K> as eigenvectors (Watson JKG. Quadratic Herman-Wallis Factors for Symmetric- and Asymmetric- Top Molecules. J Mol Spectrosc 1992;153:211-24.)

11. = 0.2 ± 3.7 % = –0.01 ± 3.84 % ;  R 0  2 = 2.688(6)10  3 Debye 2 A K = 5.3(2)10  3 d 6 (2) =1.41(4)×10 -4 d 6 2 = 2. 691(8)10  3 Debye 2 → weak ℓ-type interactions for the v 6 level A J = 0

14. Strong K-dependence No significant J -dependence

15. Empirical model has been used to compute measured self and N 2 widths → For C 3v molecules the J-and K-dependences of the widths have already been observed for: NH 3 (Nemtchinov V, Sung, K, Varanasi P. Measurements of line intensities and half-widths in the 10-μm bands of 14 NH 3. JQSRT 2004;83:243-65.) CH 3 D ( Predoi-Cross A, Hambrook K, Brawley-Tremblay S, Bouanich JP, Malathy Devi V, Smith MAH. Measurements and theoretical calculations of N 2 -broadening and N 2 -shifting coefficients in the ν 2 band of CH 3 D. J Mol Spectrosc 2006;235;35-53.) Analysis of the measured self and N 2 widths

16. Empirical model used to compute measured self and N 2 widths → For transitions having same value of J inf

17. Empirical model used to compute measured self widths → Fit of the two coefficients a J 0 and a J 2

18. Empirical model used to compute measured N 2 widths → Fit of the two coefficients a J 0 and a J 2

19. Comparison between measured and calculated self-widths

22. Comparison between measured and calculated N 2 -widths

25. Conclusion For positions: The average discrepancy obs-calc is equal to (0.001 ± 0.114)×10 -3 cm -1, The accuracy is estimated to be better than 0.2×10 -3 cm -1. For intensities: The average discrepancy obs-calc is equal to 0.2 ± 3.8 %, The rotational dependence is reproduced with accuracy around 5 %. For widths: The average discrepancy %self is equal to 0.8 ± 6.4 %, The average discrepancy % N 2 is equal to –0.3 ± 3.3 %. The J and K dependence of the measurement is reproduced with accuracy better than 10 % for the self-broadening coefficients, and around 5 % for the N 2 -broadening coefficients. → List of these parameters will be proposed to atmospheric databases

27. Ratio of the two calculations for measured transitions (1200) P Q(1) R Q(1) R Q(2) P P(1) R R(1) P Q(2)

28. Ratio of the two calculations for extrapolated transitions (18000 transitions) → No line intensity cutoff, but J max =60 and K max =30 P Q(1) R Q(1) R Q(2) P P(1) R R(1) P Q(2)

29. P Q(1) branch R Q(1) branch Comparison with measurements

30. Ratio of the two calculations for extrapolated transitions (18000 transitions)