Tuesday, June 16, 2009

Science lurches backwards

About a year ago, a friend came into Eli's office, hair standing on end, waving a copy of Science. Eli's friend is an atmospheric modeler, but he models air pollution, not global stuff, and the article claimed (it's open source) that excited NO2, which can be written as NO2* can react with water vapor to create OH, something like this

(1) NO2 + hv (> 420 nm) --> NO2*
(2) NO2* + H2O --> OH + HONO (aka ho no as opposed to oh no)
(3) NO2* + H2O --> NO2 + H2O

The last step is de-excitation and it dominates, so Sinha (the boss bunny in the lab that did the experiment) said that they had to really press on the first step to excite a lot of NO2* to see even a bit of OH. In the competition between reactions 2 and 3, 3 is a lot faster. To get a high concentration of NO2* they focused the excitation laser. This is a VERY tricky thing, because you can easily do amusing multiphoton things like

NO2 + 2hv --> NO + O
O + H2O --> 2 OH

Sinha et al. measured the amount of OH they produced and found that it scaled linearly with the number of photons they pumped in. Normally, for a single photon process, the response scales linearly with the photon flux, for a two photon process it scales as the square, etc. However, in a focussed geometry, the power law is 1.5. Anyone who wants an explanation of that needs to hire Eli. If you look at the figure on the right, you see that the response appears linear, but as the dog that did not bark in the night, the intercept is not at the origin. If the reponse really was linear it has to pass through the origin.

Eli spent a lot of his youth doing multiphoton stuff and his reaction to this test was that he was not impressed, and certainly not convinced.

Hopefully gentle bunnies you will continence a short digression here about why air pollution guys hair caught on fire when they saw this. Indeed there was a huge amount of press coverage because it provided another source of OH radicals which could attack hydrocarbons in the atmosphere leading to smog and additional ozone. The paper was splashed all over the world. There was a perspective in Science (Li is the lead author)

Li et al. add another level of complexity to the story. In laboratory experiments, they show that excited-state NO2 (denoted NO2*) may also break the O-H bond in water, yielding OH and HONO. NO2* is produced in the atmosphere when NO2 absorbs sunlight between 400 and 650 nm. Li et al. find that for typical urban conditions, 1 in 10,000 of the NO2* molecules produced reacts with H2O to produce OH and HONO. Despite this low efficiency, the rate of OH production from NO2* in urban atmospheres can be comparable to the classical OH source (ozone photolysis). The amount of OH produced from NO2* scales linearly with the amount of NO2 in the atmosphere. The net result is that the calculated ozone production rate is higher. In addition, the maximum ozone production rate occurs at higher NOx concentrations.
We have performed model simulations of air quality in the Los Angeles basin for a typical summer smog episode (4, 5) both with and without the NO2* source of OH. With the NO2* source, ozone concentrations are calculated to be much higher throughout the city, with increases of up to 55 parts per billion; percentage increases in ozone concentrations are as high as 30 to 40% (see the second figure). The most affected area is downwind of the city, near Riverside, where NOx is most abundant. Aerosol levels are also affected, especially near Riverside, where small particle concentrations (diameter <2.5 src="http://www.sciencemag.org/icons/symbolgifs/hand/mu.gif" alt="mu" border="0">m) increase by 20 mug/m3.
To give credit, the perspective did hedge its bets
The experimental approach used by Li et al. differs slightly from that used by Crowley and Carl, but it seems unlikely that this alone explains the different findings. We find no obvious problem with either study. Given the potential importance of this chemistry and the high sensitivity of atmospheric models to the reaction of NO2* with H2O, further investigation is clearly needed.
Crowley and Carl being an earlier study that found bupkis. So Eli was sitting in the smallest room a couple of weeks ago reading Science (wo sonst?) when he saw buried at the end of the letters a short one from someone he knew (lab bunnies live in an incestuous world).
Li et al. (Reports, 21 March 2008, p. 1657) suggested that the reaction between electronically excited nitrogen dioxide and water vapor is an important atmospheric source of the hydroxyl radical. However, under conditions that better approximate the solar flux, we find no evidence for OH production from this reaction.**
These folk did not focus the excitation laser and they did not see any OH.

Nor did they get any publicity

Eli's friend said he was glad he had not added this to his models.

(**There is still a potential gottcha. NO2 has two different low lying electronically excited states, 2B2 and 2B1. Don't worry if you don't understand the notation. The point is that the 2B1 state is much longer lived than the 2B2. The two spectra overlap so it is possible that Sinha and Heard excited different states....time for a proposal....:)


Aaron said...

Is OH formation a rate limiting step in real world pathways to all bad stuff? Or, are there other reactions that can supply OH? For example, can the combustion processes that produce the NOx, also produce significant OH?

What happens when the plume from a modern coal burning power plant is injected in to an air mass contaminated with PIC (products of incomplete combustion) and VOC (volatile organic compounds) from cooking & heating fires and older industrial processes? Do the power plant plumes also contain metal vapors? Are such organics and metals vapors allowed for in your lab data?

J. Zimmermann said...

Li et al. immediatly replied: http://www.sciencemag.org/cgi/content/full/324/5925/336c
Though I am not competent about this, they seem to have answered well. It remains controversial.

Anyway, when I remember the time I was doing modelling of OH-chemistry, the fit between models and measurements of OH would not improve with another OH-source. Production of a lot of OH in polluted areas is rather not, what was called for. Though I am rather sceptic about Li et al. I would need convincing measurements in polluted area showing high OH concentrations.

C W Magee said...

Is zero zero, or is zero their detection limit, after background subtraction?

Steve Bloom said...

I'm looking forward to your post about recent events in Boulder, Eli. Is there a connection between the coming of Max Boykoff and the termination of Prometheus, or is it just that years of exposure to RP Jr.'s stuff has made me prone to paranoid delusion?

Anonymous said...

Sinha Heard a Who?

Anonymous said...

Isn't the problem here that their conclusions come from logical reasoning rather than from collecting relevant data?

They got a different result, using a different method and different materials. Well, okay.

So wouldn't they also run the experiment without using the concentrating lens, just to reduce that difference?

They used different source chemicals. I don't know if that's something their setup allows them to experiment with or not.

-- hank roberts

Hank Roberts said...

Betcha, Steve.

"Max Boykoff will be joining our Center as a faculty member in August 2009. ... " sciencepolicy.colorado.edu/

This Max Boycoff, serious, heavyweight.


"When journalists and editors do not effectively place the issues in the context of such views, public understanding suffers. As New York Times journalist Andy Revkin writes, "The media seem either to overplay a sense of imminent calamity or to ignore the issue altogether because it is not black and white or on a time scale that feels like news. This approach leaves society like a ship at anchor swinging cyclically with the tide and not going anywhere. What is lost in the swings of media coverage is a century of study and evidence…."