JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. D5, PAGES 5347-5353, APRIL 30, 1986 The CH3CHO-NO 3 Reactionand Possible Nighttime PAN Generation CHRISTOPHER A. CANTRELL, JAMES A. DAVIDSON, KERRY L. BUSAROW, AND JACK G. CALVERT National Centerfor Atmospheric Research, Boulder, Colorado Kinetic studies have been made of the CH3CHO-NO 3 reaction in dilute mixturesof CH3CHO, NO•_, and N,_O5in synthetic air at 700 torr and room temperature. During the course of the reaction, reactant and product analyses were made in situ, usinglong path length Fourier transform infrared spectroscopy. The major products of the reactionfor these conditions are HONOr_ and CH3COO•_NO•_. All evidence suggests that the first step in the reaction sequence, NO 3 + CH3CHO• HONO,_ + CH3CO , proceeds with a rate constant, k = (2.1 + 0.4) x 10 -•5 ½m 3 molecule -• s -• at 299 + 1 K. Theseresults, combined with other reaction rate data pertinent to tropospheric chemistry, are used to computer simulate the homogeneous chemistry of the polluted troposphere and to evaluate theoretically the nature of the nighttime chemistry. INTRODUCTION N20 5 •-- NO 3 + NO 2 (3) Interest in the chemical reactions of the NO 3 free radical in the atmospherehas grown in recent years, following the first spectroscopic detection of this species in the stratosphere by Noxon et al. [1978] and subsequent observations in the tropo- sphere [Platt et al., 1980b; Noxon et al., 1980]. The reactions of the NO 3 radical with aldehydes are of specialinterest in atmospheric chemistry, sincein theory they can be effectivein producing free radicals (HO2, RO2, etc.) and important prod- ucts such as nitric acid, peroxyacetylnitrate, hydrogen perox- ide, etc., during the nighttime hours [Stockwell and Calvert, 1983; Calvert and Stockwell,1983; Cantrell et al., 1985]. How- ever, in the first estimates of the significance of theseprocesses [Stockwell and Calvert, 1983; Calvert and Stockwell, 1983] the rate constant estimates for the NO3-aldehyde reactionswere uncertain, and the results were necessarily qualitative in nature. Stockwell and Calvert assumed a rate constant for the CH20-NO 3 reaction (reaction (1)), k• = 1.38x 10 -•5 cm 3 molecule-• s-• (302.9 K), which was derived fromthat report- ed by Morris and Niki [1974] for the CH3CHO-NO3 reaction (reaction (2)) at 300 K (k 2 = (1.2 _ 0.3) x 10 -15 cm 3 molecule-x s- x, using an estimated value NO 3 + CH20--} HONO 2 + HCO (1) NO 3 + CH3CHO--} HONO: + CH3CO (2) of E 2 = 8.7 kcal/mol. Subsequent measurements by Atkinson et al. [1984] gavea kx value(kx = 3.23_+ 0.26) x 10 -•6 cm 3 molecule-• s-•, very much lower than that assumed in the previouswork. Hence the earlier conclusions as to the impor- tance of (1) in the nighttime chemistryof the polluted tropo- sphere must be questioned [Pitts et al., 1984]. However, recent measurements of the rate constantfor the NO3-CH20 reac- tion by Cantrell et al. [1985] led to a value about a factor of 2 higher(k• = (6.3 _+ 1.1)x 10 -•6 cm -3 molecule -• s -1) than that of Atkinsonet al. [1984]. Recently, Atkinson et al. [1984] also measured the NO3-CH3CHO rate constant and found k2 = (1.34 +_ 0.28) x 10 -•s cm 3 molecule -• s -•. Boththe k• and k 2 estimates are dependentupon the choicemade for the equilibrium con- stant K 3. In view of the large differences observed in k• esti- mates to date, we felt that it is important to measure k2 again as well. Copyright 1986 by the American Geophysical Union. Paper number 5D0865. 0148-0227/86/005D-0865 $05.00 We report here on the results of our measurements and use these to judge, through computer simulations,the theoretical significance of (1) and (2) in the nighttime chemistry of the polluted troposphere. EXPERIMENTAL PROCEDURES Product Analysis The experimental techniques and laboratory equipment usedin the present study have beendiscussed previously[Can- trell et al., 1985], and only a brief description need be given here. The large glass reaction cell (6.3-m length,448-L volume) was equippedwith a multiple-reflection, modified White opti- cal system with a 5.31-m basepath. In this work we employed 32 passes with a total optical path of 170 m. In each of the experiments, sufficient acetaldehyde (Fisher ScientificReagent grade) and nitrogen dioxide (prepared from high-purity nitric oxide and ultra-high-purity oxygen, Linde) were added to the cell to bring the concentration to the (0.05-4.7)x 10 x½ mol- ecule cm -3 range. Dinitrogen pentoxide(preparedby the method of Davidson et al. [1978]) was added in the same concentration range as the other reactants. Then a N2/O2 mixture was flushed into the cell rapidly until a total pressure of about 700 torr was achieved. The concentrations of these reactants and the products formed were determined by inte- gration of characteristic bands and, for selected compounds, usingmeasured characteristic absorptiontogether with appro- priateextinction coefficients (1 cm-x resolution). The integrat- ed band intensities (units cm 2 molecule-• cm-•, base e) and the limits of integration employed (cm -x) are as follows' HNO3, 1.36 x 10 -•7 (840-930) and 2.81x 10 -•7 (1270- 1360); N20•, 4.63x 10 -• (725-760) and4.40x 10 -• (1225- 1270); NO2, 4.15x 10 -•7 (1580-1650); CH3CHO, 2.95 x 10 -•8 (1075-1140) and 1.72x 10 -• (2630-2890). The ex- tinction coefficients (cm 2 molecule-•, base e) are' NO2, 8.25 x 10-•9 (1597) and 1.07x 10-•8 (1627); the peakheights for NO2 weredetermined by drawinga smooth curvethrough the fine structure for each half of the band near 1600 cm-• and taking the maximum absorbance of this smooth curve as the height; CH3CHO, 5.44 x 10 -20 (1095), 6.58 x 10 -20 (1415), and 1.26x 10 -•9 (2745); CH3COO2NO 2, 1.01x 10 -•8 (793), 1.31 x 10 -•8 (1162) and 9.44 x 10 -•9 (1838).The extinction coefficients of PAN are the average of those reported by Ste- phens [1969] and H. Niki (private communication, 1980) which agree to + 3-10%. Acetaldehyde standards were ob- tained by introducinginto the cell accurately measured pres- 5347