(22) Leith, D., Licht, W., AIChE S y m p . Ser. #126, 68,196 (1972). (23) Leith, D., Mehta, D., Atmos. Environ., 7,527 (1973). (24) Beeckmans, J. M., J. Aerosol Sci., 4,329 (1973); 5,221 (1974). (25) Soo, S. L., Int. J . Multiphase Flow, 1,89 (1973). (26) Mercer, T. T., “Aerosol Technology in Hazard Evaluation”, pp 297-304, Academic Press, New York, N.Y., 1973. (27) Melandri, C., Prodi, V., Am. Ind. Hyg. Assoc. J., 32, 52 (1971). (28) Beeckmans, J. M.,paper presented at the Aerosol Measurement Workshop, Gainesville, Fl., March 1976, to be published in a proceedings. (29) Spertell, R. B.,Lippmann, M., A m . Ind. Hyg. Assoc. J., 32,734 (1971). (30) Stober, W., Flachibart, M., Environ. Sci. Technol., 3, 1280 (1969). (31) Beeckmans, J. M., private communication, 1974. (32) Jotaki, T., Trans. J p n . Soc. Mech. Eng., 23,640 (1957).
(33) Mori, Y., et al., J . Chem. Eng. J p n . , 1,82 (1968). (34) Dahneke, B., J . Colloid Interface Sci., 50,194 (1975). (35) Dahneke, B., ibid., 51,58 (1975). (36) Cleaver, J. W., Yates, B., Chem. Eng. Sci., 31,147 (1976). (37) Bernstein, D., et al., J . Air Pollut. Control Assoc., 26, 1069 (1976). (38) Hochstrasser, J. M., P h D thesis, University of Cincinnati, Cincinnati, Ohio, 1976.
Received for review January 19, 1976. Accepted October 18, 1976. Presented in part at the 25th Canadian Chemical Engineering Conference in Montreal, Canada, November 1975. Research supported by research contracts #HSM-O99-71-48and CPE-70-NEG-128 from the National Institute for Occupational Safety and Health, #RP117-1 from the Electric Power Research Institute, and is part of a center program supported by the National Institute o f Environmental Health Sciences-Grant #ES-00260.
Mechanism for Olefin-Ozone Reactions Theodore A. Walter, Joseph J. Bufalini”, and Bruce W. Gay, Jr. U S . Environmental Protection Agency, Environmental Research Center, Environmental Sciences Research Laboratory, Research Triangle Park, N.C. 2771 1
The gas-phase reaction of 2-butene with ozone is investigated. By use of experimentally obtained disappearance and formation rates, intermediates are proposed that when modeled explain the nonstoichiometric behaviors of olefinozone reactions. The oxidation of SO2 in the presence of 0 3 and 2-butene is also investigated. A biradical is not needed in the oxidation of SOz. Instead, the SO2 can be explained solely by its reaction with intermediates such as OH, OnH, and CH:r02 that are produced in ozone-olefin reactions.
The production of ozone from the photooxidation of simple hydrocarbons in the presence of oxides of nitrogen is well recognized (1). The product, ozone, is also recognized as a principal reactant with the olefinic hydrocarbons in polluted atmospheres. Interest in recent years has been the determination of the rate constants and stoichiometry of olefin+zone reactions. Knowledge of the stoichiometry of these reactions is necessary if a photochemical model is to be written to mathematically describe the manifestations arising from the production of photochemical smog. The reaction investigated in this study was that of ozone k i t h trans-2-butene. This was done for two reasons: the 2butene is the simplest of the internally double-bonded olefins, and the compound is symmetrical through the double bond. This gives rise to only one biradical, thus simplifying the mechanism. In attempting to arrive a t the overall mechanism for the ozonolyses of 2-butene, the results and postulates of Demerjian et al. ( 2 )as well as those of Niki et al. ( 3 )were considered. A major difference between these two groups is the relative importance of HO2 or OH resulting from the reaction of the biradical with oxygen. Experimental Two separate experimental systems were used for the study of the ozonolysis of trans-2-butene. The first system employed a Teflon bag of 300-400 1. for the reaction chamber. The bag was filled with a measured volume of zero grade tank air which had passed through a pen ray 382 Environmental Science & Technology
mercury lamp ozonator. Thus, the volume of air and concentration of ozone could be controlled. Once the specific concentration of ozone and volume was established, trans-2butene was injected into the bag with a gas tight syringe. The puncture hole was immediately sealed with tape, and the bag was kneaded to assure complete mixing. The amount of trans- 2-butene to be injected was calculated from the volume of air in the bag and concentration that was to be obtained. The trans-2-butene was research grade and was used without further purification. The ozone concentration was monitored on a Bendix ozone-ethylene chemiluminescence instrument. The trans-2-butene was monitored by gas chromatography with flame ionization detection. A 10-ml air sample was chromatographed on a 12 f t X $ in. outside diameter stainless steel column containing 60-80 mesh Pora pac Q, which was maintained at 115 “C, with a helium carrier gas flow of 40 ml/min. Trans-2-butene was monitored every 2.5 min while the ozone was monitored continually. Ozone decay in the Teflon bag in the absence of olefin was less than l%/h. The only product analysis with this system was acetaldehyde by use of the above column. The second system used two Teflon bags of 400-1. volume. One contained ozone, and the other trans-2-butene each in zero tank air. The contents of these bags were drawn into a large cylindrical chamber of 700 1. volume via two manifold systems of the chamber. The chamber consisted of six 1.25m-long glass sections, 0.31 m in diameter. Manifold inlets were located between the sections. The chamber was evacuated to less than 1 torr pressure, and both ozone in air and trans-2butene in air filled the cell at about the same rate. The chamber was filled in about 4.5 min. Time “0” was taken when the chamber had filled to 1 atm. In the experiment with SO2, a measured amount of trans2-butene and SO2 mixed in air was injected into the chamber at atm. Ozone in air from the Teflon bag was added to bring the chamber to 1 atm. When the chamber pressure reached 1atm, time “0” was recorded. The chamber contained internal optics for multipass path reflections that enabled in situ infrared measurements. Measurements were made with a rapid scan Fourier transform spectrometer that uses a Michelson interferometer with liquid nitrogen-cooled detectors covering
the spectral range of 700-3400 cm-l(4). Path lengths of 500 m were used. Products observed during the ozonolysis of trans-2-butene were carbon monoxide, acetaldehyde, formaldehyde, formic acid, methanol, and ketene. Carbon dioxide may also be a product. Small changes of carbon dioxide in the chamber could not be detected since some of the optical system between the interferometer and chamber and between the chamber and detectors is open to room air in which the carbon dioxide concentration does not remain constant. The ozone decay in the LPIR cell was 4.5%/h. Sulfur dioxide was measured from its absorption bands at 1135 and 1165 cm-].
Results and Discussion Runs were made for olefin-to-ozone ratios of 1:l (Figure l), 1:s (Figure 2), and 3:l (Figure 3). Initial concentrations were on the order of a few parts per million (v/v), The experimental results illustrate: when olefin and ozone were present in equal amounts, more ozone than olefin was consumed; when ozone was in excess, slightly more ozone was consumed; and when the olefin was in excess, more olefin than ozone was consumed. These observations are at least in qualitative agreement with those of Japar et al. ( 5 ) and Williamson (6). The measured products arising from the ozonolyses of trans-2-butene are shown in Figure 4.Products acetaldehyde, formaldehyde, carbon monoxide, and methanol were all followed with the long path IR system. Ketene was also observed spectroscopically, but its concentration was not quantified. Trace amounts of formic acid observed experimentally are not shown in this figure. Model simulations for the three olefin-to-ozone ratios were made. The modeled curves are also shown in Figures 1-3. Although the modeling was not specific for either cis- or trans-2-butene, the rate constants for the initial ozone-2butene reaction were the same as for the experimentally determined trans-2-butene-ozone reaction. The products arising from the model are shown in Figure 4.Ketene, carbon dioxide, and formic acid were plotted separately as modeled (Figure a) since these values were not measured experimentally. From the model the concentration of C o nwas considerable, but its correctness with the experiments could not be confirmed. Also
not plotted but of considerable importance are hydrogen peroxide, and to a lesser extent, water. Small concentrations of H202 were obscured by H20 absorption, and small concentration changes in H20 could not be determined since the external path containing H20 changed more drastically than the changes expected in the reaction cell. In accordance with the Criegee mechanism, the proposed overall mechanism between ozone and 2-butene initially gives rise to a biradical and acetaldehyde. A molozonide was first invoked as an intermediate specie which unimolecularly rearranges in accordance with the O’Neal-Blumstein mechanism (7). This involves alpha-H and beta-H abstraction and a sequence leading to Criegee products and products characteristic only of the O’Neal-Blumstein mechanism. Upon mod-
g
0.2
-
TIME, min
Flgure 2. Experimental and modeled results of trans-2-butene ozonolysis, 0.5 ppm ozone, and 0.17 ppm trans-2-butene
-
EXPERIMENTAL
0
10
20
30
-
40
TIME. min
Figure 1. Experimental and modeled results of trans-2-butene ozonolysis, ozone, and trans-2-butene initially both 0.4 ppm
TIME. min
Flgure 3. Experimental and modeled results of trans-2-butene ozonolysis, 0.14 ppm ozone, and 0.4 ppm trans-2-butene Volume 11, Number 4, April 1977
383
CH3CHOO- ( 0 2 ) CHsCO3 3.15 X l o 2 min-' (3)
0 0 8 C
i
CH3COOH* CH30H 8.4 X 1013min-' (Calculated)
CARBON MONOXIDE MODELED
METHANOL MODELED
+
/ ,
1!
OM
e
FORMALDEHYDE MODELED
-
FORMALDEHYDE
I
,
0 0 2 F
0
f
-
METHANOL
5
10
15
20
25
35
30
40
TIME min
Figure 4. Products measured experimentally and modeled from ozonolysis of 0.4 ppm ozone and 0.4 ppm trans-2-butene
I
I
I
I
I
I
I
I
CH2=C=O*
+
+ CO
(5)
+ (M)
CH2=C=O 1.2 x 1 O l o min-I (Calculated) +
(4a)
(4b)
-
HC02CH3* CH30H 1.8 X 10l3min-' (Calculated)
(3b)
+ HzO
(02) HC02H 1.5 X 1 O l o min-' (Calculated)
+ (M)
+ OH
+ CO
CHSCOOH* ---* CH2=C=O* 4.4 x 10'3 min-' (Calculated) CH2=C=O
0
0.10
-
+
010,
(6)
+ CO
(7)
+
CH3C02 --* CH3 CO2 1.3 X 10l2min-'
(8)
(2)
0.01
-
FORMIC ACID
+
KETENE
0 0
I
I
I
I
6
10
15
20
I
I
30
25
I 35
40
CH3CHCHCH3 OH CH3CHOHCHCH:i 7.5 x 104 ppm-1 min-I +
(10)
(2)
+
eling the experimental results, the Criegee mechanism was all that was required to explain the results when utilizing the O'Neal-Blumstein and Criegee mechanisms together as advocated in ref. 7 . This, however, in no way negates the O'Neal-Blumstein mechanism. It only implies that our results could be explained without introducing this mechanism. T o obtain a fast decay for ozone, it was necessary to introduce a fast reaction between CH302 and 0 3 which produces CH:IO and 0 2 . The rate constants for this reaction and others are given in the Appendix. The rate constants of some of the reactions when not available were derived from the techniques of Benson (8). The reaction mechanism and rates that best describe the experimental results are shown below:
+
-
CH3CHOO. CH3C02H* 1.0 x IO5 min-I (Calculated)
(2a)
CHxcHOO. HC02CH2* 2.3 X lo4 min-' (Calculated)
(2b)
+
-
CH3CHOO- ( 0 2 ) CH:iC02 1.37 X lo5 min-' (2) 384
Environmental Science & Technology
+ OeH
+ OH +
0 2
(11)
(2)
CH:
