Wednesday, July 01, 2009

Science lurches forward again

About two years ago Stan Sanders group at JPL tossed a bombshell into our understanding of stratospheric ozone depletion, with a new, and much lower measurement of the chlorine peroxide (ClOOCl) absorption cross-section. This was a bombshell because it meant that there would be a lot less free chlorine in the stratosphere available for ozone depletion. Drew Shindell blogged on it at Real Climate, there was a bit of triumphalism on the denial beat.

The basic idea is that Cl catalyzes conversion of ozone, O3 to molecular oxygen, O2

(1) Cl + O3 --> ClO + O2
(2) ClO + O3 --> Cl + 2O2

ClO can dimerize

(3) ClO + ClO --> ClOOCl

which removes it from the catalytic cycle (1) and (2), but it can also be photolyzed

(4) ClOOCl + hv --> ClO + ClO

so the rate of photolysis, which is proportional to the absorption cross-section, determines how much ClO is available for the cycle. Because O3 absorbs most of the light at shorter wavelengths than 305 nm, for all practical purposes, only the absorption cross-section of ClOOCl above ~305 nm is important.

The rub, of course, is that to measure the absorption cross-section of a molecule, the lab bunnies need to know how much of the molecule is in the light path. This is not easy. As a matter of fact it is very tough, especially for ClOOCl because you cannot prepare a pure sample, there will always be Cl2, O2, and maybe other stuff hanging around. Before Pope, Hansen, Bayes, Friedl and Sander, others had tried to figure out the amount of ClOOCl in the light path by mass balance, starting from the initial reactant concentrations and using absorption spectroscopy in various regions of the spectrum to assign concentrations of the stable species where known. Sanders group did a spectral subtraction of the know Cl2 spectrum from the observed spectrum in the UV, which they assumed was a combination of ClOOCl and Cl2 (O2 does not absorb much until 200 nm).

The Academica Sinica group (H. Y. Chen, C. Y. Lien, W. Y. Lin, Y. T. Lee and J. J. Lin) chose a different path. They created a molecular beam containing the equilibrium mixture of ClOOCl and Cl2 and used a laser to photodisocciate (shoot out) each. Because they knew the absorption coefficient of Cl2 at the laser wavelengths, they could get the ratio of the ClOOCl to Cl2 cross-sections at each wavelength. The results are shown in the figure and agree well with earlier measurements. As they point out all three methods have their difficulties. They think that Pope, et al, overcorrected for Cl2, but that really is speculation.

This is the way that science auditing works. Comments?


Anonymous said...

It's usually good to be novel in science. If you can't be first, though, it's helpful to be right. Hurray for improving methods, predictions and outcomes.

Anonymous said...

It can't be an audit without snide and repeated accusations of fraud and incompetence.

sss said...

There had been 2 recent works on the measurement of ClO dimer absorption cross section since the paper by Pope et el, (Marc von Hobe et al and Hsueh-Ying Chen et al)but both of them used the same procedure for preparation of the ClO dimer as Pope et al. The problem with that method is - it does not produce ClO dimer in situ, but rather cools it in a trap and desorbs it later. This time lagg between the production and absorption spectroscopy can lead to formation of different isomers of ClO, with totally different optical properties. Hence, before any conclusion can be drawn about the ClO photochemistry, it needs to be prepared in situ under atmospheric conditions, and its absorption cross section measured directly. Also, only Hsueh-Ying Chen et al had a comprehensive way of detecting the Cl atoms and subtracting their contribution, the other methods need to have more conclusive approach for Cl detection. Lots of more work is needed and some interesting papers should follow....