It has long been suspected that OH is produced in the ozonolysis of olefins. In fact, you can find a published yield for your favorite ozone - olefin reaction. However, to our knowledge, nobody has actually looked for the OH; the yields are based on plausible mechanisms and the observed oxidation of OH scrubbers. We added some olefins (one at a time) to a mixture of ozone, oxygen, and nitrogen in our flow tube. Just a bit downstream we looked for OH. Here is what we saw: OH production is rate limited by kO3[O3][olefin] and removed by kOH[OH][olefin], so the OH yeld is defined as [OH]/[O3] kOH/kO3. The figure shows the observed OH on our 5 LIF axes under four reaction conditions
The expected counts for a unit yield are shown in blue. Just after we turn on a flow of olefin, causing the OH counts to surge, we tune our laser off of the OH line, and the counts drop. We then frequency scan across the line, producing the peak in the data, and finally shift to the peak frequency before averaging up some data; there is no doubt we are looking at OH. We also know that the signal is first order in ozone and largely independent of the olefin, as we expect. These data are not taken sufficently close to reaction onset to guarantee that the OH is a prompt product, but we do not observe any obvious pressure dependence or dependence on oxygen. Most surprisingly, we observe similar yields for ethene, trans-2-butene, isoprene, and tetramethylethylene, and all the yields could easily be unity once everything is properly calibrated. The calibration is not yet so good, and the yield depends on four relatively uncertain numbers, so please don't regard this month's figure as a published result; we are working on the calibration very hard. Many kudos to Dalia and Lana.
