Something in the smog biz that used to drive me nuts was when someone would look at some smog chamber experiment that had some unusual feature to it and remark, “Well, that’s just a chamber effect.” The subtext was “We’re studying gas phase kinetics, and that’s something having to do with a surface phenomenon, so we shouldn’t pay any attention to it.”
I didn’t think that should let us off the hook. What kind of surface effect was it? How did it behave? And were we absolutely sure that such effects didn’t occur elsewhere?
Eventually I wrote a paper, “Background Reactivity in Smog Chambers.” Google scholar tells me that it’s been cited at least 17 times, as recently as last year, so it did okay for a paper published 20 years ago.
In the 60s and 70s, there were a lot of smog chamber experiments done on all sorts of individual compounds; there was a belief that one could produce a “reactivity scale” that would let you reduce those things that had the most smog forming potential. As the complex nature of smog chemistry began to dawn on people, such experiments became less common, because “reactivity” has multiple components, sometimes 2 + 2 = 6 in smog chemistry, making the development of a single scale problematic. There’s a fellow at SAPRC in Riverside, Bill Carter, who has developed a much more complicated way of estimating “incremental reactivity,” which has its own problems, but it’s better than “one size fits all.”
Anyway, one of the “pure compound” experiments involved methyl chloroform, and I found it fascinating.
Methyl chloroform is also called 1,1,1 tri-chloroethane. If you start with ethane (CH3CH3) and replace all the hydrogens on one methyl group with chlorine, you get methyl chloroform. It’s pretty unreactive stuff; the only reaction sites for hydroxyl radicals are the ones on the methyl group and methyl hydrogens are bound pretty tightly. So for the first part of the chamber experiment, using very high concentrations of MCF with some added NOx, the thing just sat there.
Then, after a couple of hours of induction, something began to happen. The NO began to convert to NO2, some of the MCF began to decay, then suddenly, wham! The whole system kicked into high gear, NO went down like a shot, the MCF began to oxidize like crazy, and ozone began to shoot up. Then, just as suddenly, the ozone just disappeared, all of it, in just a couple of measurement cycles.
Everyone who looked at it said, “Ah, chlorine chemistry,” which was a sure guess. Chlorine will pull hydrogen off of even methyl groups with almost collisional efficiency (if a chlorine atom hits the molecule, it pulls off the hydrogen almost every time). Moreover, chlorine atoms destroy ozone; that’s the “stratospheric ozone depletion” thing.
But I was puzzled. Where did the chlorine atoms come from? Yes, there was plenty of chlorine in the MC, but that was bound. To get one off, you need to create a free radical and those ain’t cheap. If you create an HO radical, that can pull off one of the hydrogens, and that, after the usual reactions, gives you chloral, a tri-chlorinated version of acetaldehyde. Put in a high enough rate of photolysis for chloral in your simulation and you can get the whole system to react.
The problem was, it didn’t look right. With a high rate of photolysis for chloral, the simulation kicked off too quickly. Lower the rate and you never got the sudden takeoff. I’m pretty good at fitting the curves, and I could never get it to work.
So I started looking at the other actors in the system. The end result of chloral oxidation is phosgene (see why I was looking up all those post-WWI gas papers?), but phosgene itself didn’t fill the bill. So maybe the phosgene was converting to CO and Cl2 on the chamber surfaces like it does in someone’s lungs. No, that didn’t work either.
I kept returning to the problem over the years, trying yet another idea, each time getting no further.
In 1985, the “ozone hole” over the Antarctic was reported, and everyone in the stratospheric ozone community, including Gary Whitten, my boss at SAI, immediately suspected that it had something to do with the ice clouds that only form in the stratosphere over the Antarctic. In 1987, Mario Molina published a series of papers describing the surface reactions of stratospheric chemical species on ice crystal surfaces. The really critical reaction was the reaction of chlorine nitrate with hydrochloric acid to form nitric acid an molecular chlorine (Cl2). Cl2 photolyzes so rapidly that it might as well be two chlorine atoms.
I’m not sure when I first tried the Molina reaction on the methyl chloroform system, but it worked much better than anything else I’d tried. It makes the whole thing a very strong positive feedback system. It worked well enough to convince me that it was probably the missing factor; if I wanted to get a better simulation, I’d have to get very specific about some details of the original chamber experiment, and that one’s 35 years old. It’s pretty well moot at this point anyway.
Molina won the Nobel Prize for his work on stratospheric ozone depletion, and it was well-deserved. I was just looking at a single smog chamber experiment, one with a surface reaction that no one was interested in. The chance that I would have figured out the right answer to the peculiarities of that experiment is pretty small. The chance that I would have made the leap from the chamber walls to the stratospheric ice clouds is smaller still; I’d never heard of them before Whitten told me about them, and I certainly didn’t make the connection between them and the chamber experiment until Molina worked out the correct surface chemistry. So I’m certainly not trying to say that I coulda been a contenda.
But I will say that we all should have been paying more attention to the chamber wall effects. You don’t get to say beforehand what will turn out to be important.