| |
| Proc Natl Acad Sci U S A. 2004 September 28; 101(39): 14007–14010. Published online 2004 September 16. doi: 10.1073/pnas.0405501101. | PMCID: PMC521112 |
Copyright © 2004, The National Academy of Sciences Chemistry, Geophysics Laboratory evidence for a key intermediate in the Venus atmosphere: Peroxychloroformyl radical Holger Pernice, * Placido Garcia, * Helge Willner, * Joseph S. Francisco, † Franklin P. Mills, ‡ Mark Allen, §¶ and Yuk L. Yung ¶*Fachbereich C-Anorganische Chemie, Universität Wuppertal, Gauss-Strasse 20, D-42119 Wuppertal, Germany; †Department of Chemistry, Earth and Atmospheric Sciences, Purdue University, 1393 H. C. Brown Building, West Lafayette, IN 47907; ‡Centre for Resource and Environmental Studies, Australian National University, Canberra, ACT 0200, Australia; §Jet Propulsion Laboratory, California Institute of Technology, Mail Stop 183-401, 4800 Oak Grove Drive, Pasadena, CA 91109; and ¶Division of Geological and Planetary Sciences, California Institute of Technology, Mail Code 150-21, 1200 East California Boulevard, Pasadena, CA 91125 Received May 20, 2004. |
Abstract For two decades, the peroxychloroformyl radical, ClC(O)OO, has played a central role in models of the chemical stability of the Venus atmosphere. No confirmation, however, has been possible in the absence of laboratory measurements for ClC(O)OO. We report the isolation of ClC(O)OO in a cryogenic matrix and its infrared and ultraviolet spectral signatures. These experiments show that ClC(O)OO is thermally and photolytically stable in the Venus atmosphere. These experimental discoveries validate the existence of ClC(O)OO, confirm several longstanding model assumptions, and provide a basis for the astronomical search for this important radical species. |
The chemical composition of the atmosphere of Venus has been analyzed by using ground-based and space-probe observations. These studies have shown that the atmosphere of Venus is composed primarily of CO 2, a small amount of N 2, and trace abundances of CO, H 2O, HCl, and SO 2 ( 1). The central problem in Venus atmospheric chemistry is the remarkable stability of CO 2 despite its dissociation to CO and O 2 by the absorption of solar ultraviolet (UV) radiation. Several investigations have examined the role of chlorine in the catalytic recombination of CO and O 2 ( 2– 4). One proposed oxidation mechanism ( 4) involved the chloroformyl radical, ClCO, which was shown later to have a lesser role ( 5, 6). An alternative mechanism for the catalytic oxidation of CO to CO 2 utilizes the peroxychloroformyl radical, ClC(O)OO, as the key reactive agent ( 6). Inclusion of this chemistry provides an explanation for the low CO and O 2 concentrations in the Venus atmosphere. However, despite its suggested central role in Venus chemistry, ClC(O)OO has been experimentally elusive. Here, we report the laboratory isolation of ClC(O)OO, allowing an examination of its stability at Venus middle-atmospheric temperatures, measurements of its infrared (IR) and UV spectra in low-temperature matrices, and finally, consideration of the implications of these laboratory results for Venus chemistry. |
Isolation of ClC(O)OO The radical ClC(O)OO was prepared from pure ClC(O)OONO 2. The latter species was detected originally in low concentrations in a gas-flow system ( 7) but was synthesized for the present investigation. In separate experiments, ClC(O)OO was generated by means of thermal decomposition of ClC(O)OONO 2 with the products trapped in frozen matrices or by photolysis of a mixture of CO/Cl 2 in O 2 ice matrices. In either case, the resultant ice matrices were analyzed spectroscopically. Details of the matrix apparatus are given elsewhere ( 8, 9). Pyrolysis occurs in a quartz tube with a pinhole 1 mm in diameter at the end of the tube. This tube (the spray-on nozzle) is placed in high vacuum 2 cm in front of the cold matrix support. Ice spectra are recorded by using a Bruker (Billerica, MA) 66v Fourier-transform IR spectrometer in the reflectance mode, with transfer optics from 5,000 to 400 cm –1 and a spectral resolution of 1 or 0.25 cm –1. A deutero-triglycinsulfate detector and KBr/Ge beam splitter are used. For these experiments, 64 scans were coadded for each spectrum. The UV spectrum is measured in reflectance with a Perkin–Elmer Lambda 900 UV spectrometer with a 1-nm resolution using two quartz optical fibers and special condenser optics. In the thermolysis experiments, ClC(O)OO was formed in flowing mixtures of ClC(O)OONO 2/Ar or Ne heated at 200, 250, 300, 350, 400, and 450°C: ClC(O)OO also was produced by photolysis at λ > 305 nm of a frozen composition of Cl 2/CO/O 2 in a ratio of 1:20:400. The formation of ClC(O)OO results from the following reaction sequence: Fig. 1 compares the IR matrix spectrum of the starting material ClC(O)OONO 2 ( Fig. 1 A), resulting from photolysis of the CO/Cl 2/O 2 ice mixture ( Fig. 1B) thermolysis of ClC(O)OONO 2 at 250°C ( Fig. 1C). There are common features in Figs. 1 B and C that are not shown in Fig. 1 A, which are attributed to the ClC(O)OO radical. As expected from Eq. 1, the other expected thermolysis product, NO 2, is trapped and is clearly in evidence in the spectrum shown in Fig. 1C. The thermolysis and photolysis results are consistent with previous work suggesting the existence of ClC(O)OO [( i) thermolysis of ClC(O)OONO 2 yielding NO 2 and CO 2 ( 7) and ( ii) measurements of the oxidation of ClCO by O 2 ( 7, 10)], but the existence of ClC(O)OO had not been demonstrated explicitly.  | Fig. 1.IR spectra of CIC(O)OONO2 and products in frozen matrices. IR spectra of ClC(O)OONO2 isolated in an argon matrix (A), Cl2/CO mixture in an O2 matrix (relative abundances 1:20:400) after 15 min of irradiation at λ > 305 nm (B), and thermolysis (more ...) |
|
Spectroscopy The ClC(O)OO radical has as its ground state the  2A″ electronic configuration. There are two rotamers for the ClC(O)OO radical: cis and trans conformations defined by the ClCOO dihedral angle. Vibrational spectra were calculated theoretically by using the single and double excitation-coupled cluster with a perturbation estimate of triple excitations [CCSD(T)] theory ( 11) based on unrestricted Hartree–Fock wavefunctions. Structures of the local minima were fully optimized by using the Dunning augmented correlation-consistent basis set (aug-cc-pVDZ) ( 12). The optimized geometries for trans-ClC(O)OO and cis-ClC(O)OO are as follows: CCSD(T)/aug-cc-pVDZ optimized geometry for trans-ClC(O)OO is r(C  Cl) = 1.747 Å, r(C  O) = 1.188 Å, r(C O) = 1.434 Å, r(O O) = 1.344 Å, θ(O C O) = 126.5°, θ(ClC O) = 127.7°, θ(COO) = 110.4°, and τ(ClCOO) = 180.0°; and CCSD(T)/aug-cc-pVDZ optimized geometry for cis-ClC(O)OO is r(C Cl) = 1.741 Å, r(C O) = 1.192 Å, r(C O) = 1.438 Å, r(O O) = 1.337 Å, θ(O C O) = 117.8°, θ(ClC O) = 127.0°, θ(COO) = 114.8°, and τ(ClCOO) = 0.0°. The observed vibrational band positions and band intensities for ClC(O)OO obtained by thermolysis of ClC(O)OONO 2 and quenched in an argon matrix are given in Table 1. For comparison, vibrational frequencies for the CCSD(T) level of theory calculated numerically by using the above geometries are also shown. To check the consistency and reliability of these results, an independent calculation using density functional theory with the Becke three-parameter functional method ( 13) was used with the aug-cc-pVDZ basis set. All bands belonging to the cis or trans species are easily assigned by comparison with the predicted IR ab initio calculation results. In the 1,850 cm –1 region, the deviations from the predictions are not satisfying, because the ν(C O) bands are disturbed by Fermi resonances with the 2ν 3 band. Note that the ratio of the observed summed relative intensities reported in Table 1 for cis/trans is ≈0.6, which is close to the thermodynamic equilibrium value of ≈0.7 at 250°C expected because the relative energy of the trans rotamer is ≈1.5 kJ·mol –1 below that of the cis rotamer.  | Table 1. IR band positions and intensities of 35ClC(O)OO isolated in an argon matrix |
Additional evidence for the identification of ClC(O)OO comes from isotope-substitution experiments. For cis-ClC(O)OO produced by photolysis in an 18O 2 matrix, the bands at 1,106, 952, 637, and 567 cm –1 are shifted to lower wavenumbers by 52, 16, 10, and 19 cm –1, respectively. The first three measured shifts are consistent in magnitude and sign with the predicted shifts ( 14) for the assignment of those bands attributed to the O O, C O, and C Cl stretch vibrations of cis-ClC(O)OO, respectively. The UV spectrum of the thermolysis-produced matrix-isolated products is shown in Fig. 2. In this spectrum, the ClC(O)OO radical is a strong absorber around 230 nm. Photolysis of the matrix at λ = 255 nm results in the simultaneous decrease of all ClC(O)OO IR bands and an increase in CO 2 IR bands ( Fig. 1D). In addition, the resultant UV spectrum shows a decrease in the 230-nm band and the appearance of a band due to ClO, which suggests that the photodissociation of ClC(O)OO results in the elimination of ClO: Although the dissociation limit of the ground electronic state of ClC(O)OO favors formation of ClCO and O 2, the above result shows that excited-state photochemistry does not mimic necessarily the ground state, nor is the ground state structure a good predictor of excited-state photochemistry.  | Fig. 2.The UV spectrum of the thermolysis products of ClC(O)OONO2 at 250°C isolated in a Ne matrix. Computation of a cross section from absorbance is based on an assumption that the yield for FC(O)OO and ClC(O)OO from thermal decomposition of FC(O)OONO (more ...) |
|
Thermal Stability of ClC(O)OO Fig. 1E shows the IR spectra from all the thermolysis experiments. The matrix abundance of ClC(O)OO is largest in the 200°C experiment, which is consistent with the threshold of 115 ± 3.2 kJ·mol –1 ( 15) for dissociation of ClC(O)OONO 2 yielding ClC(O)OO. Note that some CO 2 also can result from the self-reaction of ClC(O)OO. At and above a thermolysis temperature of ≈350°C, the matrix abundance of ClC(O)OO is seen to decrease, and secondary dissociation into ClCO [reported IR spectra ( 16)] and O 2 is observed. Because the dissociation threshold for the ClC(O) OO bond is higher than the threshold for the ClC(O)OO NO 2 bond by a factor of ≈1.5, the bond energy for ClC(O)OO dissociation should be ≈169 kJ·mol –1. |
Implications for Venus Chemistry These experimental results provide important confirmations for assumptions adopted in Venus chemistry simulations for more than two decades. To explain the stability of the CO 2 atmosphere and reproduce the observed O 2 observations, chemical reactions involving ClC(O)OO were proposed ( 6, 17). The inclusion of ClC(O)OO chemistry reduces the model prediction for the O 2 column density from 3 × 10 19 to 2 × 10 18 cm –2 ( 17), the latter closely approximating the observational upper limit of <1.5 × 10 18 cm –2 ( 18, 19) (see Table 2 for summary). Previous model work ( 20) predicted a smaller O 2 column abundance (<1.5 × 10 18 cm –2), assuming that CO oxidation occurs primarily by ClCO + O 2 → CO 2 + ClO as proposed ( 21). More recent measurements ( 22) suggest, however, that ClCO is much less stable than what seems to have been assumed in ref. 20, which would reduce the rate of CO oxidation through this channel and increase the predicted O 2 abundance in the absence of another mechanism [such as reactions involving ClC(O)OO]. In light of existing uncertainties in the interpretation of the available Venus O 2 observations ( 19), a definitive confirmation of the CO oxidation mechanism also awaits more precise measurements of the total abundance and altitude distribution of O 2 above the cloud tops. | Table 2. Comparison of Venus observation of column O2 abundance with model simulations |
In the earlier models ( 6, 17), ClC(O)OO was assumed to be stable at Venus middle-atmospheric temperatures of –100 to 0°C, and no thermal dissociation reactions were included in the simulations. The experimental results reported here that the threshold for ClC(O)OO thermal decomposition is 350°C validates this critical model assumption. In the absence of any laboratory measurements of UV cross sections, previous models also did not include ClC(O)OO photodissociation. A recent Venus photochemical simulation (the “one-sigma” model described in ref. 17) was updated to include ClC(O)OO photodissociation using the UV cross sections of Fig. 2. The California Institute of Technology/Jet Propulsion Laboratory photochemical model ( 23) was used in its one-dimensional mode. Photodissociation rates were calculated for local noon at 45° latitude and then divided by two to simulate global average conditions between 58 and 112 km in altitude. Because the primary products of UV photolysis of ClC(O)OO are ClO and CO 2, as discussed above, the inclusion in a model of ClC(O)OO UV photolysis based on the spectrum shown in Fig. 2 results in only minor changes in model results for the CO 2-reformation rate and the O 2 column density. [The abundance of, and the rate of reformation of CO 2 through, ClC(O)OO are primarily controlled kinetically, although reformation of CO 2 through ClC(O)OO photolysis is not negligible.] |
Summary Essential features of current models accounting for the chemical stability of the Venus atmosphere are now supported by laboratory experiments. To complete a satisfactory understanding of Venus middle-atmosphere chemistry, what is needed is direct detection of key reactive-species abundances consistent with the predictions of the model simulations. The IR spectrum of ClC(O)OO reported herein will allow a search for this central species. In addition, precise measurements of O2 distribution above the cloud tops are needed. |
Acknowledgments We thank M. F. Gerstell for a critical reading of the manuscript and W. DeMore, H. Hartman, C. Miller, C. Parkinson, V. Krasnopolsky, and the anonymous referees for useful comments. This research was supported partially by a Deutsche Forschungsgemeinschaft grant. This research was carried out in part at the Jet Propulsion Laboratory (California Institute of Technology) under contract to the National Aeronautics and Space Administration. J.S.F. expresses his thanks to the Alexander von Humboldt Foundation for a research grant for senior U.S. scientists. |
Notes |
References 1. Yung, Y. L. & DeMore, W. B. (1999) Photochemistry of Planetary Atmospheres (Oxford Univ. Press, New York). 2. Prinn, R. G. (1971) J. Atmos. Sci. 28:, 1058–1068. 3. Sze, N. D. & McElroy, M. B. (1975) Planet. Space Sci. 23:, 763–786. 4. Krasnopolsky, V. A. & Parshev, V. A. (1981) Nature 292:, 610–613. 5. DeMore, W. B. & Yung, Y. L. (1982) Science 217:, 1209–1213. 6. Yung, Y. L. & DeMore, W. B. (1982) Icarus 51:, 199–247. 7. Spence, J. W., Edney, E. O. & Hanst, P. L. (1978) Chem. Phys. Lett. 56:, 478–483. 8. Schnockel, H. G. & Willner, H. (1995) in Infrared and Raman Spectroscopy, Methods and Applications, eds. Schrader, B. & Bougeard, D. (VCH, Weinheim, Germany), pp. 297–313. 9. von Ahsen, S., Hufen, J., Willner, H. & Francisco, J. S. (2002) Chem. Eur. J. 8:, 1189–1195. 10. Ohta, T. & Mizoguchi, I. (1980) Int. J. Chem. Kinet. 12:, 717–727. 11. Raghavachari, K., Trucks, G. W., Pople, J. A. & Head-Gordon, M. (1989) Chem. Phys. Lett. 157:, 479–483. 12. Dunning, T. H. (1989) J. Chem. Phys. 90:, 1007–1023. 13. Lee, C. T., Yang, W. T. & Parr, R. G. (1988) Phys. Rev. B Condens. Matter 37:, 785–789. [PubMed]. 14. Francisco, J. S. & Williams, I. H. (1988) J. Phys. Chem. 92:, 5347–5352. 15. Kirchner, H. G. L. F. & Zabel, F. (1990) in 11th International Symposium on Gas Kinetics (Univ. of Perugia Press, Assisi, Italy). 16. Schnockel, H., Eberlein, R. A. & Plitt, H. S. (1992) J. Chem. Phys. 97:, 4–7. 17. Mills, F. P. (1998) Ph.D. thesis (California Institute of Technology, Pasadena). 18. Trauger, J. T. & Lunine, J. I. (1983) Icarus 55:, 272–281. 19. Mills, F. P. (1999) J. Geophys. Res. Planets 104:, 30757–30763. 20. Esposito, L. W., Bertaux, J.-L., Krasnopolsky, V., Moroz, V. I. & Zasova, L. V. (1997) in Venus II: Geology, Geophysics, Atmosphere, and Solar Wind Environment, eds. Bougher, S. W., Hunten, D. M. & Phillips, R. J. (Univ. of Arizona Press, Tucson), pp. 415–458. 21. Krasnopol'sky, V. A. & Parshev, V. A. (1983) in Venus, eds. Hunten, D. M., Colin, L., Donohue, T. M. & Moroz, V. I. (Univ. of Arizona Press, Tucson), pp. 431–458. 22. Nicovich, J. M., Kreutter, K. D. & Wine, P. H. (1990) J. Chem. Phys. 92:, 3539–3544. 23. Allen, M., Yung, Y. L. & Waters, J. W. (1981) J. Geophys. Res. Space Phys. 86:, 3617–3627. 24. Sander, S., Pernice, H. & Willner, H. (2000) Chem. Eur. J. 6:, 3645–3653. |
| |
