1854
J. phys. Chem. 1882, 86, 1854-1858
Study of the Reaction of OH wlth HNO,: Klnetlcs and NO, Yleld A. R. Ravhhartkara;
F. L. Elsole, and P. H. Wlne
AM8ouk?r Schces &OW, EnBkwrerlng Expenrment Stsbbn, Georgia InstiMe of Technology, Atlanta, Georgia 30332 (Received: August 19, 1981; In F h 1 F m : Januery 4. 1982)
-
The kinetics of the reaction OH + HN03 products (kJ were investigated at 298 and 251 K. OH was produced by 248-nm laser photolysis of HNOP The temporal profile of NO3, a reaction product, was monitored by using long-path laser absorption at 662 nm. The value of k2 obtained agrees well with our previous measurements. The yield of NO3 in reaction 2 was directly measured to be near unity at both 298 and 251 K.
Introduction Nitric acid, HN03, is an important reservoir for both HO, and NO, species in the stratosphere where it is principally formed via the combination reaction of NOz with OH OH
+ NOz + M L H N O 3 + M
(1)
and removed by solar photolysis as well as reaction with OH OH
+ HN03
kz
products
(2)
Recently, we measured kzl and found it to be faster than previously believed?+ In addition, kz was found to exhibit an unexpected negative temperature dependence. Even though it is known that NO3 is formed in reaction 2 at 298 K6 and it has been generally assumed that NO3 and HzO are the only products of reaction 2, we suggested the possibility of reaction 2 proceeding through multiple pathways leading to different sets of products,' i.e.
the amount of NO3 produced from a known concentration of OH, we also determined the yield of NO3 and hence the branching ratio, kz,/k,. While the work reported here was in progress, six other groups were studying reaction 2. Nelson et al.' measured k2 to be (8.2 f 1.8) X cm3molecule-' s-' at 298 K using the flash photolysis-resonance fluorescence (FPRF) technique and (10.6 f 3.4) X cm3molecule-' s-' using the flash photolysis-laser absorption technique. In addition, they also found the NO3 free radical to be a major product of reaction 2. Kurylo et a1.F Margitan and Watson? Molina et al.,l0and Marinelli and Johnston'' have also studied reaction 2 using the flash photolysis-resonance fluorescence technique. Their results are in reasonable agreement with our previous measurements in terms of both the observation of a negative temperature dependence and the value of kz at 298 K (and other temperatures). Connell and Howard12have measured kz using a discharge flow system and observed a negative temperature dependence. However, their absolute values of kz are somewhat lower than those obtained from the flash photolysis studies.
plaining the unusual temperature dependence. In stratuspheric model calculations it was found that the estimated ozone concentration changes due to anthropogenic emissions were sensitive to not only the new value of kz but also the identity of the reaction products. Therefore, we carried out the experiments reported here to measure the yield of NO3 in reaction 2 as well as to remeasure k2 by a different experimental technique. The rate coefficient was measured by monitoring the temporal profile of NO3 formed from reaction 2 following the production of OH by laser photolysis of HN03 By measuring
Experimental Section The experimental approach used in the present investigation of reaction 2 was to create OH via pulsed laser photolysis of HN03 and subsequently to monitor the temporal concentration profie of the NO3product by using the long-path laser absorption technique. Even if NO3 is a minor product, as long as it is formed directly from reaction 2, its temporal profile reflects the rate of reaction 2. The ratio of [NO,] formed to initial [OH] is a direct measure of the branching ratio for NO3 formation, ke/k2 A schematic diagram of the apparatus is shown in Figure 1. A jacketed Pyrex cell - 5 cm in diameter and -45 cm long was maintained at a fixed temperature by using an ethylene glycol-water mixture circulated from a thermostated reservoir. HN03 mixed with a large amount of an inert gas (Nz,Ar, or SF6)was passed into the reactor and then into a (76-cm long) absorption cell. The concentration of HN03 in the mixture was directly measured in the absorption cell at 298 K by using the 228.8-cm Cd line. The 228.8-nm line was isolated from the neighboring
(1)P. H. Wine, A. R. Raviahaukara, N. M. Kreutter, R. C. Shah, J. M. Nicovich, R. L. Thompson, and D. J. Wuebbles, J. Geophys. Res., 86, 1105 (1981). (2) R. Zellner and I. W. M. Smith, Chem. Phys. Lett., 26, 72 (1974). (3)I. W. M. Smith and R. Zellner, Int. J . Chem. Kinet. Symp., 1,341 (1975). (4)J. J. Margitan, F. Kaufman, and J. G. Andereon, Znt. J. Chem. Kinet. Symp., 1, 281 (1975). (5)D. Hudson and E. I. Reed, Ede., NASA Ref. Publ., 1049 (1979). (6)D. Husain and R. G. W. Norrish, h o c . R. SOC.London, Ser. A,, 273, 166 (1963).
(7) H. H. Nelson, W. J. Marinelli, and H. S.Johnston, Chem. Phys. Lett., 78, 495 (1981). (8)M. J. Kurylo, K. Cornett, and J. Murphy, J . Geophys. Res., in preaa. (9)J. J. Margitan and R. T. Watson, 182nd National Meeting of the American Chemical Society, New York, Aug 23-28,1981. (10)L. T.Molina, C. Smith, and M. J. Molina, 28th Congress of the International Union of Pure and Applied Chemistry, Vancouver, British Columbia, Canada, Aug 16-21, 1981. (11)W. MarineUi and H. Johnston, private communication; W. Marinelli, Ph.D. Thesis, University of California, Berkeley, CA, Nov 1981. (12)P. Connell and C. J. Howard, private communication.
OH
+ HN03 -% HzO + NO3 -!% HzOz+ NOz
kac
addition product
(24 (2b) (2C)
It was further implied that the branching ratios, i.e., ke/kz and kzb/kz, could vary with temperature, thereby ex-
0022-365418212086-185480 1.2510
0 1982 American Chemical Society
The Journal of Physical Chemistry, Voi. 86, No. 10, 1982 1855
Study of the Reaction of OH with HN03
I5 Ilk
1
..
-
488 n i
1
-%.
IO
.-.*.
-
,:.,*,*,*5*. 0.2.
*.
-
.
*.
5-. M 1 and M 2 ARE D I E L E C T R I C MIRRORS W H I C H R E F l E C l 6 6 2 nn A N D T R A N S M I T 218 nm.
LE19
I.LI 0.75 MONOCHROMATOR
cd
p 7
1
T O PUMP
.-"*
~ ELIClROMElER
~ 0
I
~I
I
~I
I
I
0 . 2 5 II MONOCHROMRTOR
Figure 1. Schematic diagram of the laser photolysis-longpath laser absorption apparatus.
214-nm line by using a 0.25-m monochromator and de-
tected by a photomultiplier tube (RCA 4837). The current output of the photomultiplier was constantly monitored with an electrometer. Before each run the cell was flushed with the diluent gas and Io, the incident light intensity, recorded. Then, the HN03-diluent gas mixture was introduced and I , the transmitted light intensity, measured. At the end of the run,Io was again measured to check for drifts in light intensity during the experiment. The HNO, concentration was calculated by using an absorption cross section of 6.32 X cm2at 228.8 nm.13J4 The pressure in the system was measured at the exit of the reactor by using a capacitance manometer. The flow rates of all gases were monitored with calibrated mass flowmeters. A pulsed KrF excimer laser (248 nm) was used as the photolysis light source. It was placed -2 m from the reactor in order to produce a more spatially uniform beam. The laser beam entered the reactor through mirror M2and exited through M1 where its energy could be monitored. The CW 662-nm laser beam used for the NO, concentration measurement was generated by pumping a tunable ring dye laser with the all-lines output of a 4-W argon ion laser. The line width of the dye laser output was measured with a Fabry-Perot etalon to be -2 X lo-, nm. The probe beam was multipassed through the reactor by using mirrors M1 and M2 to obtain a path length of 4.5-9 m (mirrors M1 and M2 were specially coated to transmit 86% at 248 nm and reflect 99.5% at 660 nm). The probe beam emerged from the cell, passed through a red filter and a set of collimators, and illuminated a diffuser attached to the entrance slit of a 0.25-m monochromator. A red-sensitive photomultiplier tube (Hamamatsu R928P) was used at the exit slit for light detection. The photomultiplier was wired so as to have a response time of less than 1 ps. The photomultiplier response was found to be linear over a wide range of light intensity. The red filter prevented the detection of the photolysis beam while the combination of apertures, the diffuser, and the monochromator ensured that only the probe beam was detected (i.e., any fluorescence induced by photolysis laser and room light did not interfere with the measured probe beam intensity). The wavelength of the probe beam was periodically checked by using a 0.75-m monochromator operating in the second order. The monochromator was calibrated against known standards, and the wavelength was accurate to 0.05 nm. The diluent gases wed in this study were obtained from Matheson Gas Products and had the following stated (13)H.S. Johnaton and G. Graham, J. Phys. Chem., 77,62 (1973). (14)F. Biaume, J. Photochem., 2, 139 (1973).
Flgwe 2. Temporal profile of NO3 concentration following the photolysis of HN03 at 248 nm.
purities: Ar 2 99.995%, SF6 L 99.9970, and Nz L 99.9995%. They were used without further purification. HNO, was prepared by reacting concentrated H2S04with NaNO, under vacuum at 290 K and collecting the vapor at 77 K. It was then purified by trap-to-trap distillation (273-77 K) with only the middle fraction being kept. The pure HN03 was stored in the dark at 77 K. In a typical kinetics experiment i016-i017HNO, cm-, in 20-100 torr of diluent gas (N2,Ar, or SF6)was flowed through the cell with a linear velocity of -3 cm s-l. The probe beam intensity was monitored by a signal averager operated in the peak height analysis mode. The signal averager was pretriggered -1 ms before the photolysis beam entered the reactor to allow measurement of Io, the intensity of the probe beam in the absence of NO,. The photolysis laser beam (248 nm, - 5 mJ cm-2) photodissociated HN03 to produce OH
-
-
hv HN03
X=248nm,
OH
+ NO2
(3)
OH then reacted with HN03 to produce NO,: OH
+ HN03 2H 2 0 + NO3
(2a)
The probe beam intensity was attenuated because of absorption by NO, and was measured in real time. The intensity-time profile was converted to an [NO3]-time profile by using the relationship In (Io/l)= [N03]ul (where u is the absorption cross section of NO, at 662 nm, and 1 is the path length). One such profile, obtained by averaging 32 flashes, is shown in Figure 2. In general, 16-100 flashes (repetition rate: 0.03 Hz) were averaged to obtain the NO3 temporal profiie. In the absence of any NO, loss, the temporal profile of [NO,] is given by the relationship [NO,], = (k2a/k2)[OH]o{1 - e-k2[HN031tJ (1) where [NO,], is the concentration of NO3 at time t, [OH], is the initial [OH], and [HN03] is the constant measured concentration of HNO3 A plot of In [[NO,]., - [NO,],) vs. time yields a straight line, such as the one shown in Figure 3; the slope of the line gives k i (=k2[HN0,]). Note that to obtain k i neither u nor 1 needs to be known. k i was measured at various concentrations of HNO,; a plot of k i vs. [HNO,] yields a straight line whose slope is k2 (see Figure 4). In most experiments, the [NO3]-time profile was fitted by using an iterative computer program to the expression "31,
ki =,[OH]o(k,a/kz)[e-k~t - e-kdt1 kd
- k2
1858
The Journal of Physical Chemism, Vol. 86, No. 70, 1982
TABLE I: Rate Constant for Reaction 2 Measured by Monitoring the Appearance of NO, 1oi3h,, cm3 molecule" s-l
I
""I-
Ravishankara et al.
diluent Ar, 50 torr N,, 50 torr SF,, 60 torr
298 K 1.25 * 0.13 1.39 0.28 +_
251 K 1.94 ?: 0.24 2.09 * 0.17
to be unity15). Equation 111 is valid since the system is optically thin at 248 nm. The concentration of NO3 produced was calculated by using the relationship 0
0.2
0.4 0.6
0.8
1.0
TIME, ins
Figure 3. Firstorder NO3 growth plot; slope of the line gives k i . r
5t
where Io and I,,,are measured intensities of the 662-nm beam in the absence and the presence (at maximum) of NO3, ug& is the absorption cross section for NO, at 662 nm taken to be 1.70 X lo-'' cm2 (ref 16-18), and I is the path length. The ratio [NO,],/[OH], gives the yield of NO, in reaction 2.
I
I
/
[HNOJ
cm3
Figure 4. Plot of k,' vs. HNO, concentration; slope gives the bimolecular rate coefficient k,.
"he value of k2/ thus obtained was identical with the slopes of In {[NO,] - [NO,],\ vs. t plots since kd, the first-order rate constant for NO, loss, was usually -20 s-l while k2' ranged from 3000 to 20000 s-'. Experiments carried out to measure the yield of NO3 in reaction 2 were very similar to the kinetics runs. Careful attention was given to ensure reproducibility in photolysis laser energy and beam quality. Unlike the kinetics runs, only 10 flashes were averaged. The laser energy density was measured before and after the run by using a calibrated energy monitor. In preliminary experiments it was determined that the energy density was constant to within 10% over the length of the cell. The transmission of the mirror M2 and cell windows were directly determined by measuring the attenuation of the 248-nm laser beam. As in the kinetics runs, the concentration of HNO, was constantly monitored. The maximum in absorption due to NO, was measured for various concentrations of HNO, and laser energies. Since there was no loss of NO, in the time scale of the experiment, the maximum absorbance gave the net concentration of NO3 produced. The initial concentration of photolytically produced OH, [OH],, was calculated by using the equation [OHIO= P[HN031gk%03@'
(111)
where P = photon flux at 248 nm in photons cm-2, [HNO,] is the measured HNO, concentration, is the absorption cross section of HNO, at 248 nm (taken to be 2.01 X cm2 molecule-' (ref 13 and 14), and 9 is the quantum yield of OH in HN03photolysis at 248 nm (taken
Results and Discussion Some initial tests were performed to ensure that the absorption at 662 nm was indeed due to NO,. When the laser was tuned off the band, no absorption was measured. When the laser was tuned to -664 nm (the side of the absorption feature), an absorption that is to be expected (within 20%) based on the reported cross sections was measured.ls-l8 Therefore, it is clear that NOs was indeed the transient that was measured. Kinetic Studies. All experiments were carried out under pseudo-first order conditions with HNO, in very large excess (lo3< [HN03]/[OH]o< 5 X lo4). In preliminary experiments it was determined that a variation of 10 in [HNO,]/[OH], ratio at a fixed [HNO,] did not affect the measured rate coefficients. Also, as long as the contents of the reactor were completely replaced between consecutive photolysis laser flashes, the NO3decay rate was very small. The position of the absorption cell (i.e., in front or back of the reactor with respect to the direction of gas flow) did not have any effect on the measured rate coefficient. As seen from Figure 3, the decay of [OH] as reflected by the [NO,] rise is exponential and the dependence of k i (=k2[HN03])on [HNO,] is linear. These observations in conjunction with the invariance of k 2 to changes in [OH],, Le., photolysis laser energy, demonstrate that chemical complications did not affect the measured value of kz. One systematic error which can produce the observed negative temperature dependence of k2 is the occurrence of reaction 1 due to the presence of NO2 in the HNO, sample or in the reactor. Concentrations of NO2 in our HN03 samples, as determined by long-path absorption at 366 nm, were very low. (Note that one needs -20% NOz in the sample to see the observed increase with decrease in temperature while the measured upper limit for NO, was 0.5%.) In addition, we checked for the presence of reaction 1 by taking advantage of the fact that k, is strongly dependent on both pressure and the identity of the third body.le As seen from Table I, the pressure and (15) H. S. Johnston, S.Chang, and G. Whitten, J . Phys. Chem., 78, l(1974). (16) R. A. Graham and H. S.Johnston, J. Phys. Chem., 82,254 (1978). (17)D.N.Mitchell, R. P. Wayne, P. J. Allen, R. P. Harrison, and R. J. Twin, J . Chem. SOC.,Faraday Tram. 2,76,785 (1980). (18)F. Magnotta and H. S. Johnston, Geophy. Res. Lett., 7 , 769 (1980). (19)P.H.Wine, N. M. Kreutter, and A. R. Ravishankara, J . Phys. Chem., 83,3191 (1979).
The Journal of Physical Chemlstry, Vol. 86, No. 10, 1982
Study of the Reaction of OH with HN03
OH + HN0,-
-
Products
30 -
-'In
-a
'W
$20-
E
J
IO (1.52t0.38)~ d 4 e x p [(649*69)/T]
072.6
3.0
34
3.0
4.2
4.6
1000 T(K) Figure 5. Arrhenius plot of k,. Open circles obtained by monitoring NO3. Closed circles from our prevlous study (ref 1).
identity of the third body did not affect the measured value of kP It is easy to show that, if reaction 1had indeed been important, substantial differences in measured values of k2 would have been observed. Figure 5 shows the Arrhenius plot for reaction 2. As seen from this figure, the present measurements yield k2 values which are in excellent agreement with our previous measurements. We could not measure k2 at temperatures below 251 K because the vapor pressure of HN03 is very low and hence sufficient concentrations of HN03 could not be introduced into the cell to perform the measurement. It should be noted that the concentrations of HN03needed in the current method are -20 times larger than those required in our previous flash photolysis-resonance fluorescence experiments.' Discrepancies between our values of k2 and other previous (pre-1980) measurements have already been described.' Hence, no discussion of these discrepancies needs to be presented here. Since our measurement of k2 using the flash photolysis-resonance fluorescence technique,' there have been six measurements of k2, one at 298 K only and the others as functions of temperature. Nelson et al.' produced OH by laser photolysis of HN03 and monitored the course of the reaction by detecting NO3 via long-path laser absorption. They measured k2(298K) to be (1.06 f 0.34) X cm3 molecule-' s-l, which falls within the lower limit of the k2 value of our two studies, (1.25 f 0.28) X and (1.32 f 0.30) X cm3molecule-' s-'. However, their k2(298 cm3 molecule-' 8, obtained K) value, (8.2 f 1.8) X by using a technique nearly identical with our previous FPRF, is much lower than the results of recent FPRF studies by Margitan and Watson: Kurylo et al.,9Molina et a1.,l0 and Marinelli and Johnston." Margitan and Watson observed a change in k2 when the diluent gas (He) pressure was changed from 20 to 100 torr. They believe that the pressure dependence of k2 is real since there was insufficient NO2 in their system for reaction 1 to be responsible for this observation. A linear extrapolation of their pressure-dependent data to zero pressure yields k2(298 K) = 1.1 X cm3 molecule-' s-'. Kurylo et al did not observe a pressure dependence, and they report k2(298 K) to be 1.38 X cm3molecule-'s-l. The only recent low-pressure discharge flow tube study of reaction 2 has been carried out by Connell and Howard.12 They measure k2(298 K) to be -8.6 X cm3 molecule-' s-'. In ad-
1857
dition, they did not observe any pressure effects. The temperature dependence of k2 observed by Kurylo et a1.,8 Margitan and Watson: Connell and Howard,12 Molina et al.,', and Marinelli and Johnston'l is similar to that observed by us (both here and previously); i.e., k2 increases when the temperature is decreased. The results of the former three i n v e s t i g a t ~ r s ~support ~ ~ J ~ a non-Arrhenius behavior of k2, k2 decreasing more slowly with 1/T at T > 298 K. Even though our data, within the errors of our measurements, are best described by a linear Arrhenius expression, it is not inconsistent with a curved Arrhenius plot, particularly if the lowest temperature point is excluded. Kurylo et al. have fitted their low-temperature (i.e., T < 298 K) data to an Arrhenius form k2 = (1.05 f 0.20) X exp[(760 f 50)/T] cm3 molecule-' s-' which is nearly identical with that reported by us previously. Marinelli and Johnston obtained the following Arrhenius expression: k 2 = 1.52 X exp(644/T) cm3 molecule-' s-' The other investigators have not provided explicit Arrhenius expressions. It is, however, clear that the results of Connell and Howard12 (obtained in a low-pressure discharge tube) are decidely lower than those of Wine et al.,' Margitan and Watson: Kurylo et al.,8 Marinelli and Johnston," and this investigation especially at lower temperature. From the above discussion it is evident that, even though there is general agreement that reaction 2 exhibits a negative temperature dependence, there are disagreements regarding dependence of k2 on pressure, Arrhenius behavior of reaction 2, and the exact magnitude of k2 at various temperatures. NO3 Yield in Reaction 2. As in the case of kinetics experiments, the yield measurements were carried out under pseudo-first-order conditions. The ratio of the final NO3 concentration to the initial OH concentration, [NO,],/[OH],, was measured. At 298 K the average of seven runs yields [N03],/[OH], to be 0.98 f 0.20 and an average of five runs at 251 K yields the ratio to be 1.17 f 0.19. The quoted errors are 2a and refer to precision only. A propagation of error calculation was carried out to estimate systematic errors. The major sources of systematic errors are inaccuracies in photolysis photon flux measurements and in the absorption cross sections for NO3 at 662 nm and HN03 at 248 and 228.8 nm. The yields are calculated to be 0.98 f 0.35 at 298 K and 1.17 f 0.34 at 251 K. (These errors represent 2 m values.) The lower limits for NO3 yields are, therefore, 0.63 and 0.83 at 298 and 251 K, respectively. The yield measurements demonstrate clearly that NO3 is the major, if not the only, product of reaction 2 (i.e., k,/k2 is close to unity). This result confirms the previous assumption regarding the reaction pathways. Even though the yields at 298 and 251 K are, within quoted errors, identical, a possible source of discrepancy between the two measurements needs to be pointed out. It has been assumed that the absorption cross section of NO3 at 662 nm and that of H N 0 3 a t 248 nm are temperature independent-assumptions that are not proven. On the basis of the origin of the absorption features of HN03 at 248 nm and NO3 at 662 nm, it is unlikely that the cross sections will exhibit significant temperature dependencies. However, if such effects are found, our measured yield values will be directly affected. We have shown that there were no secondary chemistry problems. In this regard, it should be noted that secondary chemistry (i.e., secondary reactions which consume OH or NO3)would tend to reduce
1858
J. Phys. Chem. 1982, 86, 1858-1861
the yields, but the measured yields are around the theoretical maximum. It should also be noted that our measured value of the NO3 yield is in good agreement with the results of Nelson et aL7
Conclusions The results of the present study confirm our previous value of k2 at 298 K as well as the temperature dependence of k2 Four other flash photolysis studies, by Kurylo et al.? Margitan and Watson: Molina et al.,'O and Marinelli and Johnston," are also in approximate agreement with our previous results as well as the present value of k,. However, the resulta of Nelson et al.7at 298 K and Connell and Howard12as a function of temperature do not agree with these six studies. The yield measurements demonstrate that reaction 2a is the major pathway and is in agreement with the results of Nelson et al. Thus, the atmospheric calculation presented earlier' assuming NO, to be the only product can be considered preferable over other alternatives. The maximum effect on perturbation calculations due to the new values of k 2 seems most reasonable. Finally, it should be pointed out that, until a reasonable (i.e., consistent with presently accepted reaction rate
theories) explanation of the origin of the unusual temperature dependence of k2 is available, reaction 2 will not be well understood and a certain level of caution (skepticism?) should be exercised in choosing a value for k , to apply in atmospheric calculations. Acknowledgment. We thank C. A. Gump and N. M. Kreutter for their help in preparing HN03 We also thank Drs. M. J. Kurylo, J. J. Margitan, C. J. Howard, W. Marinelli, and H. S. Johnston for communicating to us their results prior to publication. This work was supported by the US Department of Transportation/Federal Aviation Administration through Contract No. DOT-FA-78WA4259.
Note Added in Proof. Recently, W. J. Marinelli, D. M. Swanson, and H. S. Johnston ( J . Chem. Phys., 76, 2684 (1982)) have reported c#&to be 8% higher than the value employed in our data analysis. If we use their cross section, our NO3 yields will be 8% lower. In a recently published study of the reaction OH + NO2 + M HNO, + M, J. S. Robertshaw and I. W. M. Smith (J.Phys. Chem., 86,785 (1982)) report a new measurement of k , at 295 f 4 K. Their result, (8.4 f 0.8) X cm3 molecule-' s-', agrees with the earlier work of Smith and Zellner (ref 3) and is -50% lower than our results.
-
A Fourier Transform Infrared Study of the Gas-Phase Reactions of 0, with Chloroethylenes. Detection of Peroxyformic Acid H. Nikl,' P. D. Maker, C. M. Savage, L. P. Breltenbach, Research and Englnwing Staff, Ford Motor Company, Dearborn, Michigan 48 12 1
R. 1. Martlner, and J. T. Herron Center for Chemlcai Physlcs, Chemlcal Klnetics LXvisbn, Natbnal Bureau of Standards, Washington, DC 20234 (Received: October 12, 1981; I n Flnal Form: January 11, 1982)
Using the FT IR spectroscopic method, we identified peroxyformic acid among the products formed in the gas-phase reactions of O3with chloroethylenes of the form CHC1=CH,Cly [0 5 ( x , y ) 5 21. It was concluded that the transient species observed by Hisatsune and Heicklen in the 0,-CHCI=CHCl system was HC(0)OOH and not the anti conformer of HC(0)OH which they had postulated. The results obtained also suggest that the Criegee intermediate H(C1)COO-is the precursor of the HC(0)OOH.
Introduction In an earlier IR spectroscopic study of the gas-phase reactions of O3 with cis- and trans-CHCl=CHCl, Hisatsune and Heicklen observed two transient absorption bands at 1744.7 (C=O stretching) and 1124.6 cm-' (C-0 stretching), both having PQR band structures.' Reportedly, these bands always overlapped and also appeared concomitantly with the formic acid bands at 1776.2 and 1105.4 cm-'. It was presumed that these observations provided the first spectroscopic evidence for the anti conformer of HC(0)OH (0-H bond anti to the C=O bond). However, a recent ab initio study of HC(0)OH by Bock et al. casts doubts on this assignment.2 Moreover, we recently noted that these transient bands reported by
Hisatsune and Heicklen' coincided almost exactly with the corresponding bands of peroxyformic acid [HC(O)OOH] which were recently determined by using the FT IR method: viz., 1744.7 and 1124.5 ~ m - ' . In ~ addition to these two bands, HC(0)OOH has a unique 0-H stretching band centered at 3340.7 cm-' which exhibits a large downward shift (- 200 cm-') caused by intramolecular hydrogen bonding?,* In the present study, all three of these bands, as well as other weaker bands of HC(O)OOH, were detected in the gas-phase reactions of O3 with all chloroethylenes of the form CHCl=CH,CL, [0 I( x , y ) I 21. This provides the first positive identification of this simplest peroxy acid as a product of the gas-phase oxidation of organic compounds. Thus, attempts were made to estab-
(1) Hisataune, I. C.; Heicklen, J. Can. J. Spectrosc. 1973, 18, 135. (2) Bock, C. W.; Trachtman, M.; George, P. J . Mol. Spectrosc. 1980, 80, 131.
(3)Maker, P. D.;Niki, H.; Savage, C. M.; Breitenbach, L.P.Anal. Chem. 1977, 49,1346. (4)Gigusre, P.A.;Olmos, A. W. Can. J . Chem. 1952, 30, 821.
0022-3654/82/2086-1858$01.25/0
@ 1982 American Chemical Society
