I t would appear t h a t the statistical treatment employed herein is capable of defining the primary sources t h a t contribute to roadway dust and of achieving semiquantitative estimates of those anthropogenic sources of primary environmental significance. T h e resolving power of t h e analysis is, however, insufficient t o distinguish between sources t h a t produce particulate matter having closely similar physical characteristics. Better definition of test vectors and/or more extensive sample fractionation is suggested as a means of improving resolving power. Quantitative assessment of the major source components is generally poor, and it is clear that analytical data for bulk matrix elements are necessary for improved quantitation. Finally, it must be stressed that, while statistical analysis provides a more detailed and more quantitative description of the data, graphical presentation of t h e mass and concentration distributions of individual elements in a fractionated particulate sample is frequently more useful in assigning physical significance. I t is strongly recommended, therefore, that both interpretive approaches be employed in a n interactive configuration.
Acknowledgment We wish to thank the operating staff of the Illinois Advanced TRIGA Reactor facility for their help in making the irradiations and the Illinois State Geological Survey for access to their counting equipment on which most of t h e reported analyses were performed.
L i t e r a t u r e Cited (1) Lee, R. E., Jr., Goransen, S. S., Enrione, R. E., Morgan, G. B., Enuiron. Sci. Technol., 6,1025-30 (1972). ( 2 ) Rahn, K. A,, “The Chemical Composition of the Atmospheric Aerosol”, Technical Report, Graduate School of Oceanography, University of Rhode Island, 1976. ( 3 ) Gladney, E. S., Small, J. A,, Gordon, G. E., Zoller, W. H., Atmos. Enuikon., 10, 1071-7 (1976). ( 4 ) Mroz, E. J.,Ph.D. Thesis, University of Maryland, 1976, unpublished. (5) Ondov, J . M., Ph.D. Thesis, University of Maryland, 1974, unpublished. (6) Greenberg, R. R., Zoller, W. H., Gordon, G. E., Enuiron. Sci. Technol., 12,566-73 (1978);Greenberg, R. R., Zoller, W. H., Jacko, R. B., Neuendorf, D. W., Yost, K. J., ibid., 12, 1329-32 (1978). ( 7 ) Hopke, P. K., Gladney, E. S., Gordon, G. E., Zoller, U’. H., Jones, A. G., Atmos. Enoiron., 10,1015-25 (1976). (8) Gaarenstroom, P. D., Perone, S. P., Poyers, J. L., Enuiron. Sci. Techno/., 11,795-800 (1977). (9) Gatz. D. F., J . Appl. M e t e o r . , 17,600-8 (1978). (10) Bogen. J., Atmos. Enuiron., 7, 1117-25 (1973). (11) Gartrell, G., Jr., Friedlander, S. K., Atmos. Enuiron., 9,279-99 (1975).
Photooxidation of the Propylene-NO,-Air Transform Infrared Spectrometry
(12) Friedlander, S. K., Enuiron. Sci. Technol., 7,235-40 (1973). (13) Miller, M. S., Friedlander, S. K., Hidy, G. M., J . Colloid Interface Sci., 39, 165-76 (1972). (14) Kowalczyk, G. S.,Choquette, C. E., Gordon, G. E., Atmos. Enuiron., 12, 1143-53 (1978). (15) Lagerwerff, J. V., Specht, A. W., Enuiron. Sci. Technol., 4,583-6 (1970). (16) Rahn, K. A., Harrison, R. P., Proceedings of the Conference on Atmosphere-Surface Exchange of Particulate and Gaseous Pollution, CONE-740921, NTIS, 1974, p p 557-69. (17) Solomon, R. L., Hartford, J . W., Enciron. Sci. Techno/., 10, 773-7 (1976). (18) Draftz, It. G., “Types and Sources of Suspended Particulates in Chicago”, Report No. IITRI-C9914-C01, IIT Research Institute, Chicago, 1975. (19) Lepow, M. L., Brickman, L., Rohino, R. A,, Markowitz, G., Gillete, M., Kapish, J., Enuiron. Health Perspect.. 7, 99-102 (1974). (20) Beeton, A. M., Limnoi. Oceanogr., 10,240 (1965). (21) Maney, J. P., Fasching, J. L., Hopke. P. K., Comput. Chem., 1, 257-64 (1977). (22) Rozett, R. W., Peterson, E. M., Anal. Chem., 47, 1301-8 (1975). (2:i) Ritter, G. L., Lowery, S.R., Isenhour, T. L., Wilkins, C. L., Anal. (‘hem., 48, 591-w
I*
Figure 2. Product spectrum in the photooxidation of the C3H6 (3.05 ppm)-NO (1.48ppm)-N02 (0.02ppm)-dry air system: resolution, 1 cm-'; number of scans, 512; irradiation time, 257 min; kl = 0.27 min-'
174
Environmental Science & Technology
0
100
200
Irradiation Time
300
3
(mid
Figure 3. Concentrations of reactants and products vs. irradiation time for the C3Hs-NO-dry air system (run 1)
c E
-8
i
*.
(a) CO
20-
C
0
c
2
L
C 0,
61
U
IO I
Irradiation
Time
I
1
0
(min)
Irradiation Time
Figure 4. Concentrations of reactants and products vs. irradiation time for the C3H6-NO-humid air system (run 2)
C:{He-NO-humidified air mixture (run 2) are shown in Figures 3a and b. In Figures 3b and 4b, the concentrations of NO and O:{are those monitored by the chemiluminescent analyzers. T h e concentration of N O monitored by t h e IR absorption in the dry air systems agreed within 5% with t h a t monitored by the chemiluminescent analyzer. In t h e humidified system, NO? concentration was not determined directly in this study, since the IR absorption band of NO2 a t 1603 cm-l is masked by t h e overwhelming absorption of H20. The NO2 concentration estimated from NO, - (NO PAN PGDN) is shown by a dashed line in Figure 4b. Figures 5 and 6 show the variations of the concentrations of reactants and products for runs 3 and 4.
+
1
+
Di,scu.s.sio n
Effect of Water Vapor on Formic Acid Formation. I t has been generally recognized that addition of water vapor to the photochemical system of hydrocarbon-NO,-air accelerates the overall photooxidation process due mainly t o the thermal reaction to form "02 from NO, NOP, and HzO, and (2-5). However, the effect the subsequent photolysis of "02 ot'water vapor o n the product distribution has not been clarified yet. Comparison of Figures 3a and 4a clearly indicates that the addition of water vapor enhances t h e yield of HCOOH very markedly. The same result was obtained for the C : I H ~ - N Osystem ~ (runs 3 and 4). T h e time profile of t h e formation of HI:'OOH suggests that this compound is mainly
(min)
Figure 5. Concentrations of reactants and products vs irradiation time for the C3H6-N02-dry air system (run 3)
formed in the reaction of 0 3 and C3H6. Formic acid is known to be one of the final products of ozone-olefin reactions (20-24). Although the formation mechanism has not been well established yet. isomerization of the Criegee intermediate diradical, .CH200., has been suggested (23-25) to give HCOOH as follows:
-
.CH~OO.
CH/Y
+
HCOOH
(1)
' 0
Another possibility for the formation of HCOOH in ozoneolefin reactions was suggested to be hydrolysis of some peroxidic reaction products ( 2 1 ) .Oxidation of ketene is also proposed to be a formation path to HCOOH (26). In our recent study of the dark reactions of the 03-C2H4 and C:jHc,system, it was found t h a t t h e addition of water vapor does increase the yield of HCOOH (27). Therefore, the water vapor effect observed in the present study on formic acid formation in the photooxidation of the C:3He-NO,-air system can be ascribed to that for the C:IHB-O:Jreaction. Water vapor effect in ozone reactions has been reported by Cox and Penkett (2S),who observed an inhibition of the oxidation of SO? by H20 in the czs-2-C4H8-O:j-SO~reaction. Calvert et al. (29) proposed a competitive reaction:
+ SO2 + H20
CH:1CH00. CH:$H00.
-
-
+ SO:) CH:$OOH + H2O CH:$HO
Volume 14, Number 2, February 1980
(2) (3) 175
and PGDN in the photooxidation of the CaHs-NO,-air system strongly suggests the importance of the NO2 reaction in photochemical smog chemistry. As shown in Figures 3b-6b, N ~ O Fstarted , to appear when 03 accumulated to appreciable concentration, while NO2 Concentration was also high. After co ,/i reaching its maximum concentration, N205 disappeared as Y z (a) NO2 was consumed. This kinetic behavior is consistent with that of N205 expected from known reactions:
1
I '
+ NO:) + 0 2 NO2 + NO:{ 2 N205
NO2
0:j
+
wall
NO3
100
0
I
3 00
200
4 00
I
I
I
I
\
A \
-
i
*
*
A
A
.
1
100
0
1
Irradiation Time
4 00
(min)
Figure 6. Concentrations of reactants and products vs. irradiation time for the C3HG-NOp-humidair system (run 4)
to interpret the water vapor effect, and speculated on a complex between t h e Criegee intermediate and a H2O molecule. The marked increase in formic acid formation observed in this study might, a t least in part, be explained by a n analogous homogeneous or heterogeneous reaction: CH200.
+ H20
+
[CH200*H20]
+
HCOOH
+ H2O
(4)
Since the complex between the HO2 radical and H20 was shown to be present (30, 31 1, and the calculated dipole moment of CH202 (3.03 D ) (25) is greater than that of HOz (2.34 D) (32), complex formation between CH202 and HnO that favors the formation of HCOOH may be plausible. On the other hand, as shown in Figures 3a-6a, formation of HCOOH was found to continue even after C3He was entirely dissipated. This fact suggests t h a t HCOOH is also formed from relatively stable products such as peroxidic products or ketene as suggested by Vrbaski and Cvetanovit (21) and Walter et al. (26). The presence of H20 may enhance hydrolysis, decomposition, or oxidation to form HCOOH from these compounds. Actually, enhanced decomposition of ozonide in the presence of H10 vapor was recently found in our laboratory (27). T h e reaction of HCHO with OH radical may also give continuing formation of HCOOH, but the yield o f HCOOH in the photooxidation of HCHO is reported to he very small by Hanst and Gay ( 3 3 ) . Importance of NO3 Reaction. The identification of N20;, Environmental Science & Technology
(7) (8)
T h e NO3 radical is known to react with C:& with a rate cm3 molecule-1 s-l (34).While constant of (5.3 f 0.3) X the reaction path of and C3H6 has not been reported yet, the formation mechanism of PGDN is thought to he as shown in Scheme I. A study of the reaction of N20j and C3H6 which gives t h e identification of the intermediate nitroxyperoxypropyl nitrate (CH3CH(ONO2)CH*OON02 or CH3CH(OON02)CH2ONOz) and elucidates the above reaction scheme will be reported elsewhere. Using the equilibrium constant of Reaction 6, K = h-6/h6 = 0.8 X 10" molecule cm-3 ( 3 4 ) ,the maximum concentration of the NO:i radical can be calculated to be 3.9 X IO9 molecule ~ m from - ~the observed concentration of N 2 0 5and NO2 for the run shown in Figure 3. Since the maximum concentration of the O H radical for the same run has been estimated to be 6.6 X lo6 molecule cm-:' (33),the relative importance of C;jHs decay due to the NO:) radical as compared to that due to the O H radical can be given as:
*
3w
200
+ C : I H -*~ products
h8[N03] - 5.3 X -hlOIOH] 2.5 x
PGDN
176
(6)
H20
N205 *2HN03
0
c
(5)
OH
+ C:&
X
3.9 X lo9 = 0.13 x lo6
X 6.6
-
products
(9) (10)
The values for h s and h 10 used were those given by Japar and Niki ( 3 4 )and Atkinson and Pitts (36),respectively. Equation 9 shows that for a particular phase of C:iHG photooxidation, where 03 and NO2 coexist in appreciable concentrations, the relative importance of NO:)reaction with C : I Hamounts ~ to up to 13% of the O H reaction. Thus, although the average importance of NO3 as compared to OH would be much smaller than this figure, the NO3 reaction with olefins is an important process t h a t not only accounts for the appreciable hydrocarbon consumption but also gives new types of nitrogen-containing secondary pollutants, and should be included in future computer modeling work. Maximum Yield of 03.In our previous study (10) o f t h e photooxidation of the C,