D. H. Stedman and H. Niki
2604 (10) Although the capacitance manometer indicates presslire changes to an accuracy greater than 1 mTorr, the initial background pressure, largely air. shown in Figure 2 is nominal. (11) Recent determinations of LHr”o (CS) range from -63 to -70 kcalimol: (a) G. Hancock, C . Morley, and I , W. M. Smith, Chem.
Phys. Lett.. 12, 193 (1971); (b) S. Beli, T. L. Ng, and C. Suggitt, J. M o l . Spectrosc., 44, 267 (1972); (c) D. L. Hildebrand, Chem. Phys. Lett., 12, 193 (1971). (12) “JANAF Thermochemical Tables,” 2nd cd, National Bureau of Standards, Washington, D. C., 1971.
Kinetics and Mechanism for the Photolysis of Nitrogen Dioxide in Air’ C).
W . Stedman”
instirute for Envircnmentai Quaitty and Deoertment of Chemistry. University of Michigan. Ann Arbor Mich!gan 481 C4
and H. Niki f o r d Motor Company, Science Research Laboratory, Dearborn, Mlchigan 48121
(Received March 30, 79731
Publication costs assisted by the Ford Motor Co
The photolysis of low concentrations (0.9-100 ppm) of nitrogen dioxide in air was investigated using vo/o3 chemiluminescence detectors. The rate constants of three of the elementary reactions were di2x02, hlo = 4.0 f 0.2 X 10 -38 cm6 sec-l; NOz O3 KOs rectly determined at 298°K: 2 N 0 + 0 2 0 2 . k q = 6.5 f 0.8 X cm3 sec-I; NO + 0 3 NO2 + 0 2 , h3 = 1.73 f 0.1 x cm3 sec-I. The rate constant ratio for 0 NO2 0 2 + NO to 0 f 0 2 + M 03 M was also determined.
-
+
- -
Introduction The reactions occurring in the photolysis of low concentrations of NO:, in air are important in the chemistry of both photochemical smog and the upper atmosphere. Previous investigations of reactions occurring in this system are summarized in recent reviews.2-4 There is general agreement that the following reactions are involved in the mechanism of the photolysis of KO2 in dry air. NO
+
+
hv-NO
--
+
0
(1)
0 0 , + ?if 0,+ M N O + 0, NO, + 0, NO, 0, NO, + 0, + NO, - + N O 0, --j
+
6
0
+
NO, + M - N O , +M + NO, -;=lt N,O, NO + NO, 2N0,
NO,
0
+
+
NO
+
hf
2 N 0 -I- O2
--
NO,
+
M
2N0,
(2) (3)
(4) (5)
(6) (7) ( - 7 )
(8) (9) (10)
In the present report, this system of reactions was studied using the high sensitivity and fast response speed of chemiluminescent N O / 0 3 detector^.^^^ The results have provided some new rate constants for the individual reactions and a rapid method of measuring the effective light intensity ( k l ) in a photochemical smog reactor.? Experimental Section In these experiments the concentration/time profiles of The Journal of Physical Chemistry. Vol. 77. No. 22, 1973
-+
-
+
+
NO, NOz, and 0 3 in a photochemical reactor were determined a t atmospheric pressure and room temperature, 298 f 2°K. The gases used to make up the bulk of the air were ultra-high-purity nitrogen (American Cryogenics) and ultra-high-purity oxygen (Matheson), without further purification. The only hydrocarbon impurity detected was 4.2 ppm of methane. The relative humidity in these sample gases was less than 1.570,which was the lowest detection limit of the electrical sensors used (Hewlett-Packard and Phys. Chem. Research Corp.). For the reactant gases, nitric oxide was trapped over liquid Nz from cylinder gas (Matheson) and then distilled from liquid 0 2 to liquid Nz, rejecting the first and last fractions. Nitrogen dioxide was prepared from this nitric oxide by adding an excess of oxygen in a 2-1. flask a t r0o.m temperature and then trapping the product at 77°K. These oxides of nitrogen were stored as solids at 77°K. Known concentrations of ozone in oxygen were prepared by irradiating the 0 2 by a mercury penlight in a flow system. Nitric oxide and nitrogen dioxide were introduced into transfer vessels of known volume (IO-I000cc) by warming the traps to obtain the required vapor pressure, which was read using a pressure transducer (Pace Whittaker). The contents of the transfer vessels were then flushed first with nitrogen and then with oxygen, into the photochemical reactor, which was a 44-1. Pyrex bell jar. The first flush with K, was t o minimize oxidation of N O t o NO2 in the transfer process. Most experiments were carried out a t 760 Torr total pressure with 150 Torr of 0 2 . Mixing of the reactant gases was achieved in less than 1 min with a
Kinetics and IVlechanism for the Photolysis of
2605
NO2 in Air
multiperforated Pyrex inlet tube designed so as to direct incoming gas tangentially throughout the height of the bell jar. The bell jar rested on the Teflon-coated stainless steel plate with Viton gasket. The vessel surface was 93% Pyrex and 7% Teflon with the surface-to-volume ratio 0.2 c m - l . The vessel was pumped down between runs to 7-10 iu. pressure by means of a liquid N2 foreline trap and a mechanical pump. Illumination was provided by a single air-cooled F40 BLHO black-light fluorescent tube mounted on the axis of the bell jar, or externally by twelve air-cooled black-light fluorescent lamps. The ultraviolet light intensity was qualitatively measured by a sodium salycilate-sensitized Cd/Se photocell which was mounted behind a Corning glass 7-60, 3000-4000 filter and diffuser. This light monitoring device indicated that it could take as long as 6 sec for the twelve lamps to all switch on. Furthermore, the total light output reached a stable level only after 20-30 min of warm-up. Therefore, in order to conduct experiments in a constant light intensity the lamps were warmed up for a period of 30-60 min, and the bell jar was covered with a black cloth shutter, This shutter could be lifted off in less than 1 sec for the purpose of observing the transient behavior of NO and O3 under constant-illumination conditions. This arrangement was used for the majority of the experiments. Two movable 10-p diameter, 5-cm long capillaries inside the reaction vessel were used to sample nitric oxide and ozone into the chemiluminescent analyzers with flow rates of 35 cc m i n - l each. The chemiluminescent analyzers have been described p r e v i ~ u s l y These .~ analyzers were operated with a noise level -1 ppb and overall response time > [03]the stoichiometry does not enter directly into the evaluation of the rate constant. The expected secondorder nature of the reaction is shown in Figures 2 and 3. Wulf, Daniels, and Karrer in 192213 showed that the stepwise addition of 2.3-7.3% [O,] in ozonized oxygen t o N02-N204 removed 2.04 0.05 equivalents of NOz. At 8.15, 11.89, and 17.04% [O,] their stoichiometry increased to 2.5, 3.1, and 2.8, respectively. They suggested that this might have been caused by errors in their analytical procedure for determining the larger [O,]. If the stoichiometry of 2 is accepted, then the mechanism would seem to be NO, + 0 , S O , + 0, ('1)
-
-
NO,
+
NO,
+
M *N,O,
(7)
as suggested by Johnston and Yost.ll Reaction 4 is 25.5 kcal mol-1 exothermic.9 Another possible exothermic pathway is NO2 0 3 NO 2 0 2 , which is exothermic by 20.4 kcal mol-l. This reaction would be followed by KO O3 NO2 0 2 and would result in lowering of the stoichiometry. This second mechanism may be related to the first in terms of the fact that two possible structures for NO3 exist, and that some of the NO3 produced in reaction 4 may decompose before being sufficiently stabilized. In order to determine if this mechanism is important, a study of the stoichiometry of the NO2 + 0 3 reaction under varying conditions is necessary. I t is worth noting that reaction 0 NO2 going via some form of NO3, which is 45.9 kcal mol-I exothermic, leads to stabilized KO3 at atmospheric pressure a t 1h of the rate a t which it decomposes to S O and 0 ~ . ~ a
+
-
-
+
+
+
+
The Journal of Physical Chemistry, Voi. 77, No 22. 7973
D.H . Stedman and H. Niki
2608
TABLE I:
____
Rate Constants at 298°K for the Reaction NO 4- 03
+
NO2
+02 _ I _ _ _ _
Rate constant, cm3 sec ’ Author
Ref
Johnston, et a/. Phillips, et a / Marte, e f ai. Clyne, et ai. This work
15 16 17 18
50-500 4-1 1 1000
0 3 0 3
NO NO
i
TABLE I I :
computed f i t to the 03 and NO data in the photoiysis of 4 . 2 p p m NO2 in air using the reaction mechanism of Table 1 1 : (-) experimental; ( 0 ) calculated. Note the break in the time axis at 0 . 5 min.
The Journal of Physical Chemistry, Vol. 77, No. 22, 7973
f0.16 fO.l
Rate constan!, ppm-’ min-’
9.5 X 1.25 x 3.8 x 4.5 x 14 3.2 x 2.5 x 2.8 x
a Units min of air
’
2.5
10.5
Factor of 2.5
NO2 in Air
1.37 x 1.73 X 6.5 X 8.3 X 2.6 X 3.06 X
104 103 103
2.2
104 103
Ref
10-14
This worka 226
10-14 10-17
This work Thiswork 24 7h 25 25 25 23h
lo-”
IO-’*
x IO-”
1.7 X l o - ” 1 . 9 x 10-19
10-4
This work”
Expressed as a biomolecular rate constant in 1 atm
-
TABLE Ill: Ratio of Rate Constants 0 4- 02 -t M (kp[M]) a n d 0 + NO2 0 2 NO ( k 5 ) Author
This work Schuck, et a / . Ford, et a/. Bufaiini, et ai. From Table I 1 Davis. et a / . Clyne, et a / a
2.46
Rate constant. cm3 s e c - ’
0.15 20 25.5
1 2 3 4 5 6 7 -7 8 9 10
Figure 9. A
N O + 0 3 . The dependence of In [03]on time and on [NO] is shown in Figures 4 and 5. Figure 6 shows the determination with 0 3 in excess, and one data point a t which [KO] = [03]decayed as [KO] - I = time. This is one of the three fast reactions setting up the photostationary state.2a.4.7 An accurate value for its rate constant is important for the determination of light intensity in photochemical reactors, or in the atmosphere. The largest systematic error in this rate determination is in the calibration of the NO and O3 detectors. These are both calibrated using tanks of NO in N2, checked against standard mixtures in this laboratory and in the Analytical Services group a t Ford by the Saltzmann technique. The error from this source is estimated to be less than 3%. The limits of error are derived from the least-squares analysis of the slopes of Figures 5 and 6, which combined with the calibration error lead to an overall value of k 3 = 1.73 f 0.1 x cm3 sec-I. Table 115-18 shows that the present constant falls between the two most accurate previous studies.16J8 There is general agreement on an activation energy of 2.5 kcal mol-’ over the range from 198 to 330°K. Thus the expresexp(-2500/RT) can be used to sion k 3 = 1.13 x good accuracy over the range of temperatures likely to he encountered in the atmosphere. The O3 detector was routinely calibrated against NO assuming a unit stoichiometry. This has been established in other s t ~ d i e s . ’ ~ , ~ ~ The Photolysis of NO2 in Air. This reaction was carried out at 10 ppm of NO2 in air by Ripperton and Lillian,lg as a baseline experiment for a photochemical smog study. They observed rapid increase of NO and 0 3 , as expected; but then [os] increased slowly, and [NO] decreased. This contrasts with the behavior observed here between 1 and 10 ppm NO2, shown in Figure 7. Ripperton and Lillian suggest that their observations may be indicative of the presence of an organic impurity.
Factor of 2
Mechanism of Photolysis of
Reaction
mi^
Estimated errors
1014
2.0 2.1 2.9 1.56 1.73
2-4 0.05-0.27
Both
Time,
x
Concri range, ppnl
Excess
Activation energy, kcal mol-
+
k2[M]!kj
1.08 5 0.12 x 10-3 1.15 f 0.08 X
1 . 3 3 x 10-3 -6 X 1 . 6 1 0.4 x 1 0 - 3 1.1 x 10-3 2.2 x 10-3
-
03 -t M
Ref
21 14
10 26a 27’1
Using the author’s new values for k Z ,and k2 [ M I from Table I1
Experiments such as those shown in Figure 7 were conducted for the purpose of: (a) demonstrating that an analysis of [NO] and [Os] in the first 30 sec of photolysis of NO2 can lead to a direct measurement of the light intensity; and (b) understanding the details of the reactions in the KO2 photolysis system. The rapid buildup of equal concentrations of NO and 0 3 is predicted by the first three equations in the mechanism. The slower increase of [NO] and decrease of [ 0 3 ] in the later stages, such that (d[NO]/dt - d[Osj/dt) 0: [N02l2 as shown in Figure 8 can only be interpreted with a more extensive mechanism. The best available rate constants were assigned to each step in the mechanism in Table 11. Assuming steady-state [N03], [o,], [O], and [NzO,] in equilibrium, an analytical solution can be derived from the mechanism (dINOlldt - d[O,l/dt) = 2h,h,[N0,l2/hz[O,l[M1 (1) Equation I has the correct functional dependence on [KO12 shown in Figure 8. A value of k2[i\/I]/k5 = 1.08 It: 0.12 X is derived from the position of the line in Figure 8. This is compared with literature data in Table 111.
Photochemistry of Phenylcyclobutane
The data of Table I11 are good evidence that the absolute rate constants are correct, and that the mechanism of Table I1 gives an adequate understanding of the photolysis of 1-10 ppm of NOz. Further verification is provided by computer modeling using the experimental initial concentrations and the rate constants of Table II.20The points in Figure 8 show the results of these calculations over the first 30 sec of reaction. Figure 9 shows as solid circles the results of this calculation applied to 5 min of data from the photolysis of 4.2 ppm of NOz. The agreement is within the errors of the absolute rate constants.
Acknowledgments. The authors acknowledge helpful discussions with Dr. E . E. Daby and other members of the Fuel Sciences Department at Ford Motor Co., where the experimental work was carried out. References arid Notes (1 j Work performed at the Ford Motor Company. ( 2 ) (a) E . A . Schuck and E. R. Stephens, Advan. Environ. Sci., 1, 7 3 ( 1 9 7 0 ) : (b) H. S. Johnston, Project Clean Air 4, Task Force No. 7, Section 4, University of California, 1970. (3) A . tieicklen and N . Cohen. Advan Photochem , 5 , 1 5 7 ( 1 9 6 8 ) (41 (a) P. A . Leighton. "The Photochemistry of Air Pollution. ' Academic Press, New York, N. Y . . 1961; (b1 A. P. Altshuiler and J. J. Bufalini. f n v i r o n Sci Techno/, 5 . 3 9 ( 1 9 7 1 ) . (5) D. H Stedman, E. E Daby, F. S:uhl. and ki Niki. J A i r Po//ut. Contr. A s s . . 22. 260 ( 1 9 7 2 ) . ( 6 ) A. Fontijn, A J. Sabadell, and R . J Ronco. Ana/. Chem , 42. 575 (1970). ( 7 ) D. H Stedman and ti. Niki. Environ. S c i Techno/.. in press.
2609 (8) L. P. Breitenbach and M. Shelef, J , A i r Poiiut Conrr A S S , 23. 128 (1973). (9) G . Schott and N. Davidson,J. Amer. Chem. Soc . 80. 1841 (1958) ( 1 0 ) J. J. Bufalini and E. R . Stephens, lnt. J A i r Water Po!iut. 9. 1 2 3 (1965). (11) H . S. Johnston and D. M. Yost. J. Chem. Phys.. 1 7 . 3 8 6 ( 1 9 4 9 ) ( 1 2 ) B. Dimitriades, J. A w Pollut. Contr. A S S , 1 7 , 460 ( 1 9 6 7 ) . ( 1 3 ) 0. R . Wulf, F. Daniels, and S.Karrer. J. Amer. Chem Soc.. 44. 2938 (1 9 2 2 ) . ( 1 4 ) ( a i H. W. Ford. G. J. Doyle. and N. Endow J. Chem Phys. 26. 1336 ( 1 9 5 7 ) : (b) H. W Ford and N. Endow, i b i d . 27, 1156. 1277 (1957). ( 1 5 ) H. S. Johnston and H J, Crosby. J . Chem. Phys, 22. 689 11954). ( 1 6 ) L. F. Phiilipsand H. I . Schiff.J. Chem. Phys. 36, 1059 ( 1 9 6 2 ) . ( 1 7 ) J. E. Marte. E. Tschuikow-Roux. and H. W. Ford, J . Chem. Phys 39,3277 (19 6 3 ) , (18) M. A. A . Clyne. B, A. Thrush. and R . P. Wayne, Trans. Faraday Soc.. 60,359 ( 1 9 6 4 ) . (19) L A. Ripperton and D . Lillian. J . A i r Poilut. Ccntr. A s s . . 21. 629 (1971 j . (20) T. E. Sharp and E. E . Daby adapted a chemical kinetics program (D. F . DeTar. 3. Chem E d u c . . 44, 1 9 3 ( 1 9 6 7 ) ) for use on the Philco-Ford timeshare system. Ford Motor Co. Scientific Lab Report No. SR 71-1 1 1 , ( 2 1 ) E. A. Schuck, E. R. Stephens, and R. P Schrock, J . 4rr Poliut. Contr. ASS.. 11. 695 (1966) (22) F. Stuhl and H . Niki. J. Chem. P h y s . . 55. 3943 ( 1 9 7 1 ) . and references therein. ( 2 3 ) F . Stuhl and ti. Niki Chem Phys. Lett.. 7. 197 ( 1 9 7 0 ) . and references therein. ( 2 4 ) M . A. A . Clyne and H. W . Cruse, Trans. Faraday Soc.. 67. 2869 (1971). (25) L. Zafonte, Project Clean Air 4 , task Force No. 7 , Section k , University of California, 1 9 7 0 , and references therein. ( 2 6 ) D. D. Davis, J. T. Herron, R. E. Huie. Paper presented at the 164th Nationai Meeting of the American Chemicai Society. New York. N. Y., Aug 1972 (27) M. A. A. Clyne and H. W. Cruse, J , Chem. Soc.. Faraday Trans 2 68, 1281 ( 1 9 7 2 ) .
Photochemistry of Phenylcyclobutane. I1' Shih Yeng Ho, Robert A. Gorse, and W . Albert Noyes, Jr.* Departmenf of Chemistry. The University of Texas af Austin, Austin, Texas 78712
(Received May 25, 1973)
Pub/icafion costs assisted by the Universlty of Texas at Austm
The products of the photolysis of phenylcyclobutane have been confirmed but yields of styrene and of ethylene are somewhat higher than previously found as are yields of cis- and trans-1-methyl-2-phenylcyclopropane. The reason is probably to be ascribed to wall effects. Since fluorescence yields agree with earlier values, the effect is probably on the excited triplet state. The external heavy atom effect of added xenon is to increase somewhat the yields of styrene and of ethylene but to decrease markedly the yields of the cyclopropanes a t 266 nm. Triplet sensitizers with high fluorescence yields tend to decrease product yields but others with high crossovers to triplet states have little effect. Without question styrene and ethylene must come partially but perhaps not solely from the triplet state. The source of the cyclopropanes is either the high vibrational levels of the ground state or of the triplet state. Neither state can be excluded or confirmed as the sole precursor of isomerization.
Recently Autard2 has shown that phenylcyclobutane gives four principal products upon photolysis: (a) styrene; (b) ethylene; (c) cis-1-methyl-2-phenylcyclopropane; (d) trans-1-methyl-2-phenylcyclopropane. Styrene yields were always lower than ethylene yields and this was ascribed to polymerization of the styrene. All yields decreased with decrease in wavelength of' the incident radiation as did
also the yield of fluorescent emission. Even a t the longest wavelength used (266 nm) an energy balance was not achieved and only about 0.76 of the absorbed photons could be accounted for either by product formation or by light emission. By use of cis-2-butene which presumably quenches triplet but has little effect on singlet states, some evidence relating product formation to the triplet The Journal of Physicai Chemistry, Vol. 77, No. 22, 1973
