Relaxation, Incubation, and Dissociation in CO 2 ² Saumitra Saxena and John H. Kiefer* Department of Chemical Engineering, UniVersity of Illinois at Chicago, Chicago, Illinois 60680 Robert S. Tranter Argonne National Laboratory, Argonne, Illinois 60439 ReceiVed: October 12, 2006; In Final Form: December 18, 2006 The dissociation and relaxation of CO 2 has been reexamined in the incident shock wave with the laser- schlieren technique. These new experiments covered 1377-6478 K, and 42-750 Torr, and improvements partly described herein have permitted accurate determination of both rate and incubation time. In general the steady rate measurements are in agreement with other recent determinations. The one anomaly is that the new rates are not fully second order; they vary about 50% over 70-600 Torr. This unexpected feature is actually quite consistent with the recent literature, which shows a similar trend. However, attempts to produce this result with RRKM calculations were unsuccessful. Relaxation times are in agreement with available literature, and incubation time to relaxation time ratios lie between 1.5 and 3 over 4000-6600 K, consistent with findings for other molecules. These ratios are much smaller than those recently derived from reflected- shock experiments by Oehlschlaeger et al. (Z. Phys. Chem. 2005, 219, 555). A simple argument suggests such large values are indeed anomalous, although why they are too large is not clear. Introduction and Background The dissociation of carbon dioxide, CO 2 + (M) f CO + O + (M) {1}, has a very large ∆H° 298 ) 127.12 kcal/mol, 1 and an even larger barrier arising from the need to cross to a triplet surface 2,3 to reach the O( 3 P) ground state product. The triplet states add another 5.9 kcal/mol 4 to the barrier resulting in an E o ) 131.59 kcal/mol. With such a large barrier, the thermal reaction can only be seen in the shock tube, and a fairly large number of studies have now been done this way. 5-19 These have employed various methods: IR emission/absorption, 5-14 single-pulse product analysis, SPST, 15 laser schlieren, LS, 16 O-atom ARAS, 17,18 and UV absorption of the CO 2 . 19 All these have involved various dilutions in rare gases, but a few studies have actually been carried out in the pure gas. 20-22 A problem with many of the earlier experiments is an anomalously low E a , even as low as 70 kcal/mol, now believed to be a consequence of rapid secondary loss of CO 2 via the abstraction O + CO 2 f CO + O 2 {2}, when this is catalyzed by H-atom through H + CO 2 f CO + OH, and OH + O f H + O 2 . 23 In recent efforts this problem has been resolved, and reasonable apparent activation energies greater than 100 kcal/ mol have been obtained. In future consideration we will confine our discussion to these current examples. 16-19 Vibrational relaxation in CO 2 is none too fast, and has been widely observed, at high temperatures, most notably by Simpson, 24-27 using LS and interferometry. The early concern over whether all modes relax serially, i.e., with a single relaxation time, has now been resolved 28 positively, and the process may be considered well understood. The relaxation measurements indicate it should be possible to observe incuba- tion delay times, 29 defined as the delay in the onset of a steady dissociation reaction until vibrational energy accumulates suf- ficiently through relaxation after shock-heating of the cold gas. In fact, theoretical predictions of incubation delay times for CO 2 have been proposed in two instances, 30,31 and times like these should be easily observable. Most recently Oehlschlaeger et al. 19 have presented incubation times measured in the reflected shock, observing UV absorption of the CO 2 . The times they derive are quite long, about an order-of-magnitude greater than the cited theory, 30,31 and there are some rather simple reasons to believe they are indeed much too long (see the Results and Discussion Section). The extraction of incubation times using the alternative LS method in the incident shock, is the primary motivation for this work. The dissociation of CO 2 was already studied by one of us (JHK) some years ago 17 using the LS method, but no incubation times were reported. The pressures used were too high for a good resolution of relaxation, and it was also felt that the LS time origin location was too uncertain for accurate determination of incubation times. Since then the LS time-origin location has been treated and reasonable accuracy established both through theory 32 and experiment. 33 In addition, the experimental resolu- tion and sensitivity have been much improved in the intervening 32 years, so that accurate rate constants, relaxation and incubation times, are easily determined. Experimental Section The essentials of the LS method and our experimental apparatus have been presented many times, 34,35 and only the recent improvements will be discussed here. Recently a number of changes and improvements have been made to the experiment which have resulted in improved sensitivity, time resolution and reduced noise. The changes can be considered in two parts. The first concerns the data acquisition hardware and software and the second the detector hardware. ² Part of the special issue “James A. Miller Festschrift”. * Corresponding author. E-mail: kiefer@uic.edu. 3884 J. Phys. Chem. A 2007, 111, 3884-3890 10.1021/jp066717b CCC: $37.00 © 2007 American Chemical Society Published on Web 02/13/2007