-------
63
23 I XZt
*
X 23
» X
I
a x
tx »
2 «
ax * ODD ooooooooooo
z« o .0 o o o
0021
. . ° * a 2
.'°,
00° .
.
..' " »*
» 00
X *0
It
tc-M . »nxiM « M na cnujmuTiM nciia ncnm io»
FIGURE 111-42
r
r
« i
rrrrrrrr»rrrr
r rr rrrrrrr*
r r r r rr rr
r r » «
r r
r r r
rr r
r
r r r
/ 'r .
' r
r
r r «
rr
r «
r r r
rr r r «
r r r
r r r r« »
" ' ' 'r rr
M.M 1M-M M«.e» 400.M
B-«t . rmm rnor rowi eoncnmuTiw cc«u rtcmi i*^
FIGURE II1-43
-------
64
t - PAR
1 - ALU
-**»
4 ACTIAlDtHrOE
P P P P P P P PP
p p p * *
r p p «
r P
P TT *
P
P P
P
r r
p
p p
pp p
tr
rrrr *pp pp p
M.M I04.M
cc-f* . grtcta PA* ALM oom^rnurt«K BCAIX
FIGURE I11-44
-------
65
« * »
0 0
0
00
OD
«o *
00 0
000000000*00 oooo on «
4 * * < . , * , «
t 5«. * IM.M 19*. M 2*«.Mt 2SC.OO a**.** 3M.M 4OO.M
TIME (nifl/TFS)
BOft* . VTCCIE9 OS COUCCTTIUTI OH RCAJ^ r*CTMI I**»
FIGURE II1-45
n
>*
« MM* RBM II « **» !«>*>
IM.M IM.M TIK'7JiJJtM, » M».M 1M.M
. nxim m M c**nrf»ATiii« nc*ix r«cn» «4«
FIGURE I11-46
-------
66
*rr r
»r
r r
« f T
* r
«r r
. r
EC-** . vrccio nwr cowcwmuTiop RCAIX rAcrow i**«
FIGURE II1-47
r
T
r
r
r
« r
t
n
n T
« m
r n
« r r r
r*r n TTTT rr rr r r
COMXHTIUTIOII BCMX r*cmi «
FIGURE III-48
-------
67
r - rom
A - ALBS
,***
A A A A A
AA AA
AAA A
AAA rrrrrrrrr r
A A rrr r
AA rrr r
AA IT r
AA Arr
AAATF *
AAFFT
A A F
A r
30*. * 33*. 4(MI.I
tc-*« . Brretts nmn ALH CDWcarnuTiM KALE TACTMI i»*«
FIGURE 111-49
-------
68
- 03
I - 10
a - i01
e
o
o
222 I
11 «
2 22
2 0
1 X X O
X 22
* X e 2
a x a
o 2
12 o i
X 2
XI 0 * XI
O 22
K * a
2 O XI
a * xii
a
o x a
a a
x a
2
X 1
e x a a
» x a a
o «> x a a a a
o » xx
M * 9 +n 9* *!K »« XX]
IM.M
. ftrccm M M« awceimMVio* KALE rtcrmi
FIGURE 111-50
F
f FT FTF HIT
m f T TT IT F r
F fr r rr'
r F T *
T * T
T r i
TT T T
r f r
r TT T
T-T
T T * T T
r r , r T T
T * T T
T t r r
r IT
r r
tr * *
TT T -
IT
rr r
T T T r * T * T *T T* m
IM.W IM.«« 30*.M XM.0* ». 09«.O* «M.«f
TIBE tninvmn
. cram mr nm COHCTHTIIATIOH teuc ruena, i»-«
FIGURE 111-51
-------
69
-»* A
r r
PPF
P
ACETALOEHTDE
tC-M . SfKCltM PA1 ALM COOCEITTMTIM SCALE FACTO*
FIGURE 111-52
PPP FF F PF FTP PTTTF F P*
F F F FF F F F
. FF" ' ' ' F F F ,
84. » !*. 1M.M 2**.A* 2M.M
FIGURE 111-53
-------
70
0
FIGURE 111-54
-------
71
O - OS
- KO
1- to*
ooo
a o
o o
13 »
1 210
20 *
i a .
XX 2
X X 2
* X 02 *
i n
10
2 2 X
IX 0 2 X
2 2
22 X
2 X O 2
0 2 X
I 2
2 X
O 2
2 X
O « 2
2 X
2 2
0 « X2
a xa
O.1 - 222
OS** X2222
OR* 222222222
» R K RR 3IWR*R H *flR ****#** XXX
srrcitM ea 10 mi coacumuTw cuu menu ie*e
FIGURE 111-55
r - rrae
r - roBH
t.M*
r nrr r rrr TTT
T TT TTT
FT « f f
F F f F »
F » F
F » F F
F F F
F F F F
F » F F
F F F F
F F F F F
F F F F F
F F * TT TT
F F ' F
F FF
F F
F FF «
F F F
IF F F «
IF F F
1 FFFFFF«F*F«««»*a
o M.oo
to-*t . trecit* nor ronn oncnmuTim ouu FACTOR io>o
FIGURE II1-56
-------
72
* AAA AAA *
I A *
* A
A
A
A
AA
A A
A A
A A
A A
T p r » rr r
r. r r »
p -' ' ACETALOEHra
p p«
. rr
» r «
A
M.M !M.*t ». 20«.M 25».«* 3**.*t 384.M «««.00
TlfK CHIIIimiil
BC-M . cnciDi ru Ataa concnmuTHUi KCALE r«cmi !*
FIGURE II1-57
rrmr r rrf rff
r IT r T r r r
p p - P p p p »
p p r r p p
p p p r p p PP
p p p p PP p? p
199,99 IBV.M X**.M K9.99
TlfK (RIVtfmt
nrattcs «LD9 amct?rnMTim »OUJB PMCTOM
FIGURE 111-58
-------
73
0 0 004
» OO 00 O
*
5».« !»*.*
«t . srccia
OOfODrniATIOII KCALX FACTOIl l»*t
FIGURE 111-59
-------
74
9 - OS
- so
2 - noa
«oo*
* x
2 X
22 O
2 O
> 0
0 2
X O 2
X XX
* I
«,« !«.« IBO.O* VH.M S39.M 3M.OO WB. M 4O4.t*
ec-tzi . Bnxm «a ra na UMCUiMTTCT KMX rwrnin i^«
FIGURE II1-60
ic-111 . tracm nor
»
I K»U MCTMI
M*.«* IU.M WV.M
FIGURE 111-61
-------
75
« TT
T T
rrr
I T
rr
S4.M IM.M 190.M 3
Tim
rc-ui . crtein rM concnmuTtm K«ti FACTOH
FIGURE 111-62
AA A A A A AAAA
AA r rr A A A
AArrrr rrrr AA
AA rr rrr r A
ATT r r A
AT r
A r r A
' r * . .
A? ' » '*.
' r ' J
*
' r
' ' r
A I
A
A ' ACETALOtHYOe
M.M IM.» 19*.M M«.M «3*.W >M.W «9t.*> 4M.M
TIRR (fnxirnvT
to-iti . imin- row am ' eoneamierim KALK r«cn> »«
FIGURE 111-63
-------
76
B B B B
B B B
B B 8
BBB
inn
B B N «
BB
B B * *
pa BP F n r r
PPFP « a p rrrp r r »*
» P B
r r HI
Tf B
r> n n
PF n B *
FP> B
X »
B m B
m.M 28«.M 3M.M 9M.M
', IJtlfVTTft
TIBC
ujn Ren emeomiaim nuu FACTO* i»<
FIGURE 111-64
-------
77
TABLE 111-7. PERCENT CARBON RECOVERY FOR
PROPYLENE/NO SYSTEMS
/\
Elapsed Time
UO [MJM
Number
5
11
13
16
17
18
21
5127
55
56
59
60
95
96
121
60
57%
82
*
99
97
93
91
43
35
*
*
*
73
*
*
120
76%
81
88
90
*
85
*
42
36
*
52
*
68
*
53
180
76%
80
75
73
*
*
*
42
.35
35
52
69
64
58
*
240
75%
*
76
63
87
81
76
43
41
38
52
68
*
57
38
300
70%
*
77
70
72
73
83
39
39
*
49
60
55
56
27
* Data not available.
10. Discussion of the Propylene/NO Systems
A
The set of propylene experiments was performed at UCR over a period
of some 25 months, in contrast to the other sets discussed below, each of
which was completed within a few weeks. Propylene runs EC-5 to EC-21,
however, were performed in a short time period, and later runs were mostly
repeats of the standard run, EC-13. During the next year we will analyze
these repeat runs more closely as part of a computer modeling study to
seek out and explain chamber effects. In this report we present what
-------
78
we consider our best simulations of these runs, using the same overall
mechanism throughout and varying only the light spectrum and initial
HN02 concentration. As explained in Section III-B-5, the light
spectrum was varied in a systematic way and only when spectral data
were not available. The initial HN02 concentration used in the simula-
tions was usually about one-third of the equilibrium concentration
possible from the reaction of FLO, NO, and NOp.
In our final fitting procedure we began with propylene runs EC-95,
EC-96, and EC-121, because detailed spectral measurements were available
for these runs. We used our photolysis constant program to calculate the
carbonyl photolysis constants (relative to H0? photolysis). We then varied
the percent of ozonide formation in the mechanism until the total maintenance
source of radicals from aldehydes, propylene-0 atom reactions, and
ozone-propylene reactions gave a good fit to the UCR data. Note that the
percent of ozonide formation is only a fitting parameter. If the quantity
is measured, many other parameters or added reactions could be used as
fitting parameters. Some of these are as follows:
*v.
> Reactions of ozoniaes with radicals to act as sinks, or
reactions that produce or release radicals.
> A slower or faster rate constant for-the ozone-propylene reaction.
> Larger or smaller quantum yields for carbonyl photolysis.
> Fundamental changes in the mechanism, such as PAN chemistry,
NO loss chemistry, and aldehyde formation.
^
I '-' , ;"''
During the fitting procedure for propylfirie runs, we Discovered that
the photolysis constants calculated from the spectral data for runs EC-95
and EC-121 differed. (Of course, it was still possible to produce good
fits by arbitrarily varying the photolysis constants in the simulations
of these runs.) The reported relative spectral data lead to a formaldehyde
5 _T f :
photolysis constant for radical production of 1.1 x 10 min for EC-95
and 7.5 x 10"4 min"1 for EC-121.
-------
79
On one hand, both of these values are rather high. In the simulations
this causes an abundance of maintenence radicals from the photolysis of
formaldehyde and other aldehydes. Thus, we had to I-swer radical production
from the ozone-propylene reaction by setting the tuning parameter, the
fraction of ozonide formation, to a high value, 0.33. This high value
in turn makes the kinetic mechanism more sensitive to carbonyl photolysis
than to the ozone-olefin reaction. On the other hand, the formaldehyde
photolysis constants for EC-95 and EC-121 are quite differentthe EC-95
constants are almost 50 percent larger than those of EC-121. Despite
this outcome, the data from EC-95 and EC-121 are very similar. Therefore,
the amounts of maintenance radicals required in simulating these runs must
be similar. At this point we examined the reported spectral data for the
period between runs EC-95 and EC-124. These are plotted in Figure 111-65,
which shows that some experimental scatter may be possible. Therefore,
we chose the EC-120 curve to select the optimum value of the fitting para-
meter and used the value in all simulations. Figure 111-66 shows the
product of the acetaldehyde absorption spectrum and the reported relative
spectral data for run EC-120. This figure shows that the region of the
spectrum having the largest effect on radical production from aldehyde
photolysis is from 300. nm to 340 nm. The same region shows the most
apparent scatter in the UCR spectral data. Figures 111-50 to 111-59
show measurements and predictions for runs EC-95 and EC-96. These
simulations used photolysis constants calculated from the relative
spectral data reported by UCR. From the decay curves for propylene
it is evident that too many radicals are present in the simulation.
We continued the fitting procedure by studying earlier experiments
for butane as well as propylene to ensure that the spectra and HN02
concentrations required for different simulations were consistent. At
this time we lowered the photolysis quantum yield for aldehydes larger
than formaldehyde to 0.5. This change produced consistency between the
butane, propylene, and 1-butene simulations.
-------
80
i.o
NORMALIZED EC-95
EC-106
EC-Ill
EC-114
EC-116
EC-121
EXTRAPOLATED EC-121
0.001
280
340 ' 360
Wavelength '(rnn)
400
420
FIGURE 111-65. RELATIVE INTENSITY FOR DIFFERENT UCR RUNS
-------
81
280
300 320
Wave Length--(rim)
340
FIGURE 111-66.
ACETALDEHYDE PHOTOLYSIS CROSS SECTIONS FOR
TWO DIFFERENT SIMULATIONS
-------
82
As with all simulations, the total NO loss is not duplicated.
/\
The overall smog chemistry is dominated by conversion of NO to N0p> and
this conversion has been accurately simulated by providing a combination
of chemical reactions and rate parameters. As discussed in the butane
section above, we feel that our worst simulation is for EC-21. The
mechanism does not provide enough maintenence radicals to sustain the
rate of chemical reactivity observed at UCR. The compound most poorly
simulated is PAN. PAN concentrations can be simulated in runs with high
concentrations of NO and propylene, such as EC-5, EC-13, and EC-16.
/\
For the low NO runs (EC-17, EC-11, and EC-18), however, the predicted
/\
PAN concentrations vary from too low (EC-17) to too high (EC-18) as
the hydrocarbon concentration increases. We feel that the shapes of the
simulated ozone curves will more closely resemble the observed data when
the NO loss and PAN chemistry are elucidated.
/\
The shapes of some ozone curves can be partially explained by loss
of ozone to the walls and the use of the controversial reaction of HO*
and NOp. In the original simulations of EC runs 55, 56, 59, and 60 the
predicted ozone concentrations fit jthe observations through most of the
experiments but were tod low at the ends of the experiments. In repeat
simulations, the removal of these two reactions from the mechanism
improved the fit at the ends of the experiments without degrading the
remainder. The rate of ozone loss to the smog chamber walls is treated
as a constant in the mechanism, although it is probably not constant in
reality. However, we did not feel justified in using different ozone
loss rates for different experiments because this quantity is infrequently
measured at UCR. Therefore, we continued to use a constant ozone loss
rate as reported by Hecht et al. (1974b). The rate constant for the
reaction of HO* and N0? was arbitarily adjusted to give a minimal effect,
but the slowness of smog formation in runs EC-50 to EC-60, which is
presumably due to a severe deterioration in the light source, exaggerates
the overall effect of this reaction.
-------
83
In Table III-8 we present a comparison between the predicted and
measured maximum one-how-average ozone concentrations. Percent differ-
ences are calculated relative to the observed data. The overall average
of +0.93 percent represents the average difference between predictions
and measurements. This average difference does not signify a bias because
the standard deviation is much larger (15.9 percent in the propylene system)
The measured ozone concentrations for EC runs 5, 11, and 18 all
peak and then decrease, but the predicted ozone concentrations do not-
We found that a much better fit could be obtained by using different PAN
chemistry, but the different chemistry caused a poorer fit between measure-
ments and predictions for PAN and NO-- The simulations shown were chosen
for their overall resemblance to the data, not necessarily for the best
fit to ozone.
B. BUTANE/NO CHEMISTRY
/\
Subsequent to our report of last year (Durbin et a!., 1975), we
revised the "explicit" butane mechanism to improved the simulations of
UCR data for the butane/NO system. The following areas were studied:
/\
> Butane oxidation.
> Alkoxyl radical chemistry.
> Secondary products important in overall reactivity
(butyraldehyde, methyl ethylketone).
These areas are discussed below.
-------
84
TABLE III-8.
PREDICTED AND MEASURED MAXIMUM ONE-HOUR-AVERAGE
OZONE CONCENTRATIONS FOR PROPYLENE/NOv SYSTEMS
A
Maximum One-Hour-Average Ozone
Concentration (ppm)
EC Run Number
5
11
13
1-6
17
18
21
5127
55
56
59
60
95
96
121
Measured
0.46
0.23
0.37
0.50
0.14
0.18
0.006
0.26
0.33
' 0.36
0.38
0.25
0.41
0.413
0.503
Predicted
0.47
0.27
0.374
0.45
0.16
0.21
0.0067
0.245
0.28
0.285
0.31
0.21
0.51
0.513
0.45
Percent
Difference
2
17
1.1
-10
14.3
16.7
11.7
-5.8
-15.2
-20.8
-18.4
-16
24
24
-10.5
Note: Average difference +0.93 percent; standard deviation 15.9 percent.
-------
85
1 . Hydrogen Abstraction from Butane by Hydroxyl Radicals
The products of the hydrogen abstraction from butane in air by
hydroxyl radicals are a peroxyalkyl radical and water. Hydroxyl radicals
can abstract a hydrogen from either the terminal carbon or the secondary
carbon as follows:
°2
OH
°2 ?2
C4H1Q + OH- -» CH3CH2CHCH3 +
A ratio of 6:1 for secondary to primary attack was determined from the
empirical formula developed by Greiner (1970) for
Alkane + OH» * Products
The rate constant is:
k = [1.0 exp(-.820/T)Np + 2.3 exp(-430/T^s
+ 2.1 exp(95/T)NT] x 1.476 x 103 ppm^min"1
where Np, N,,, NT are the number of primary, secondary, and tertiary hydro-
gens, respectively. The initial alky! radicals immediately react with
molecular oxygen to form the peroxyalkyl radicals shown, which will react
with nitric oxide (NO) to produce alkoxyl radicals and N02- From here
on, the chemistry of the alkoxyl radicals plays a central role in the
overall system.
-------
86
2. Alkoxyl Radical Chemistry
a. Decomposition Versus Reaction with Molecular Oxygen
In the propylene/NO system, alkoxyl radicals are not as important
A
as they are in the butane/NO system because the hydroxyl -alkoxyl radicals
A
that form in the propylene system are apparently less stable than the
alkoxyl radicals in the butane system. After reviewing the literature
and studying the thermokinetics of the decomposition of the alkoxyl
radicals, we have concluded that the decomposition pathways are unim-
portant (except for sec-butoxyl radical) compared with their reactions
with molecular oxygen to form carbonyl compounds. This is discussed in
the propylene chemistry section.
Formation of butyraldehyde occurs from onTy one reaction, that of
n-butoxyl radicals with molecular oxygen. Therefore, we need to consider
the unimolecular decomposition of sec-butoxyl radicals over the reaction
with molecular oxygen in order to limit the formation of methyl ethyl -
ketone,
0-
-0
CH3CH2CHCH3 + 02 +,CH3C
due to the difference in secondary to primary ratios between the initial
attack ratio (from Greiner's formula) and the product ratio measured
at UCR. UCR data show the need for a ratio of 18:1 of secondary to
primary products, that is, the ratio of MEK to butyraldehyde.
Although neither of these ratios is highly certain, it appears that
they are not equal. Additional evidence for some other path to make the
products ratio different from the production ratio comes from the results
of trial calculations with all the alkoxyl radicals reacting with molecular
-------
87
oxygen. In these experiments we found that too much MEK was formed,
not enough acetaldehyde was formed, and far too much butyraldehyde was
formed. Adding the unimolecular decomposition of secondary butoxyl radi-
cals to the mechanism is therefore very tempting because the main stable
product is acetaldehyde (Benson and O'Neal, 1970). However, the ratio of
MEK to butyraldehyde is then even lower than the 6:1 production ratio.
If the decomposition of the primary butoxyl radicals is postulated, then
the production of propyl radicals must be accepted based on thermokinetic
considerations:
HCHO
Then one would have:
NO
HCHO
In other words, the formation of more formaldehyde and propionaldehyde
is predicted. In all the UCR butane experiments, propionaldehyde was
never reported, yet acetaldehyde and butyraldehyde were seen. In some
propylene experiments, very low amounts of propionaldehyde were re-
ported in the presence of substantial amounts of acetaldehyde. For
instance, in run EC-121, propionaldehyde was reported at 0.003 ppm in
the presence of 0.14 ppm acetaldehyde. We feel that the negative UCR
experimental results for propionaldehyde imply that little, if any,
-------
88
primary butoxyl radicals were thermally decomposing in the butane system.
Also there is supporting evidence from our trial calculations with
exclusive decomposition that the decomposition chemistry would lead to
greatly enhanced formaldehyde production, a result that is not in keep-
ing with the low amounts reported in the UCR data.
As we reported last year, for alkoxyl radicals the rates of unimole-
cular decomposition and reaction with molecular oxygen span several orders
of magnitude. Recently, Barker et al. (1976) have shown that most reac-
tions with molecular oxygen proceed at similar rates virtually independent
of the size of the radical. However, thermokinetic and experimental evi-
dence suggests that unimolecular decomposition rates increase rapidly with
increasing molecular weight. Thus somewhere in the alkyl chain-length
progression, unimolecular decomposition would be competitive wtth reac-
tion with molecular oxygen. Future smog chamber experiments with longer
chain alkanes, such as n-hexane and n-octane, are necessary to help
elucidate this chemistry.
In this study we tried to test the assumption that either one or the
other process is dominant in the butane system. Our finding is that,
N.
overall, the dominance of reaction with oxygen is more in keeping with
the UCR data. The one unimolecular decomposition that seems necessary
to retain at this time is that of the secondary butoxyl radical.
b. Internal Isomerization of the Alkoxyl Radicals
The information above suggests that reaction with oxygen may domi-
nate the alkoxyl radical chemistry of butane. However, the observed
product ratio of MEK to butyraldehyde and especially the carbon mass
loss in the butane experiments cannot be fitted by this conclusion alone.
A carbon mass loss could be explained by formation of bifunctional
compounds, such as hydroxy-carbonyls, which would have long retention
times for chromatographic analysis as used at UCR.
-------
89
Carter, et. aj.. (.1976) have suggested that internal hydrogen rearrangement
may occur via a six-member cyclic intermediate for n-butoxy radicals:
CH3CH2CH2CH20.
H
/
H2C
\
H2<
.--0
\
CH9
/ ^
3 CH2
The hydroxyperoxybutyl radical would then react with NO:
02CH2CH2CH2CH2OH + NO + N02 + -
OCH2CH2CH2CH2OH
0
,1
(4-hydroxybutyraldehyde)
II 20
HCCH2CH2CH2OH + hv ->-2 HCOg + '
0
H!cH2CH2CH2OH + OH-
°2
02CCH2CH2CH2OH + NO + N02 + C02 +
3-Peroxy-l-propanol radicals may also form; these would undergo reactions
similar to the above:
02CH2CH2CH2OH + NO -> N02 +
-------
90
(3-hydroxypropionaldehyde)
II 0?
HCCH2CH2OH + OH- + .0C
HCCH2CH2OH + hv -> HCOg +
S
02CCH2CH2OH + NO
The 2-peroxyethanol radicals formed might then react with NO
NO
OCH2CH2OH + 0? -> H0| + HCCW2OH .(glycolic aldehyde)
HCCH2OH + OH- + HCHO + H02
o 2Q
HCCH2OH + hv ->2 HCOJ + FORM + H02
The sec-butoxyl radical may also undergo rearrangement processes. These
processes would occur more slowly than for the n-butoxyl radical, however,
because five-membered ring intermediates probably form more slowly than
six-membered ring intermediates (Hendry, 1976):
-------
91
CH3CH2CHCH.
OH
OH
I
02CH2CH2CHCH3 + NO
OH
OH
i
-OCH2CH2CHCH3
OH
HO'
20,
0 OH
i! I <-u?
HCCH2CHCH3 + hv v
OK
0 OH
II I
HCCH2CHCH3
OH
OH
0 OH .
NO * N0
OH
OH
NO
OH
I
OH
I
-OCH2CHCH3.
OOH
II I
HCCHCH
OOH
20,
OH
HCCHCH3 + hv + HC03
-------
92
OOH OOH
I! I o i |.
HCCHCH3 + OH- _4 -
OOH OH
III 20 i
02CCHCH3 + NO _4 N02 + C02 + -
OH 0
I ^ II
-* HOA ~f" HCCH-
Private communications with K. Darnell of UCR confirmed the existence
of some peaks in their GC analyses that are consistent with hydroxyaldehydes.
Mill and Montorsi (1973) found the major products of gas-phase oxidation of
n-pentane and higher homologs to be cyclic ethers. The corresponding
reaction of the 1-butoxy radical to form tetrahydrofuran is probably
slower, since the transition state is a strained five-membered ring, but
intramolecular hydrogen abstraction, as discussed above, is reported to
occur with facility in the liquid phase. Furthermore, Carter et al .
(1976) have reported evidence for alkoxyl radical isomerization occurring
in C.-C,. alkanes in. NO /air systems in smog chambers. We have chosen ttte
tO X
rate constant for the reaction:
02CH2CH2CH2CH2OH
to fit UCR's butyraldehyde data.
A carbon mass balance, as described earlier in this chapter, was per-
formed on UCR's butane data (Table III-9). The carbon mass balance in general
showed approximately a 40 percent loss early in each run. In the butane/NO
7\
system the reactive losses are proportionally smaller than in the other systems.
Since butane is a four-carbon molecule, the error in its measurement is a
-------
93
TABLE III-9. PERCENT CARBON MASS RECOVERY
FOR BUTANE/MO SYSTEMS
Time
(minutes)
60
120
180
240
300
EC-39
37%
79
*
81
77
EC-41
55%
62
57
57
54
UCR
EC-42
*
83%
80
*
84
Run Number
EC-43
*
53%
46
54
54
EC-44
53%
*
*
*
62
EC-45
33%
48
*
*
54
EC-48
79%
44
*
73
79
Data not available.
large contributor to error in the mass balance calculation. Table 111-10
indicates this effect in EC-44, in which the reported initial concentra-
tion of butane seems anomalously high. Included in the table are mass
balances calculated using a gross extrapolation and a semi-log least-
squares fit to the data. The reactive loss of butane results almost exclu-
sively from oxidation by the hydroxy radical, which is present in steady-
state amounts. The assumption that its concentration is fairly constant
during the run appears justified by the more reasonable carbon recoveries
and the high correlation coefficients for the least-squares fit, which are
usually 0.99 or larger. The qualitative and quantitative differences of the
TABLE III-TO. PERCENT CARBON MASS RECOVERY FOR RUN EC-44
Initial Butane Concentration
Time
(minutes)
60
300
3.92 ppm
(reported)
25%
49
3.80 ppm
(extrapolated)
759%
76
3,90 ppm
(fitted)
53%
62
-------
94
recovery in the butane/NO system indicate an undetected carbon-containing
/\
species other than CO or C02- The early and fairly constant loss of
carbon implicates a primary product of butane oxidation as the prodigal
species.
The mass loss expressed in ppmC is shown in Figure 111-67- It is
closely similar in shape and magnitude to the MEK data (uncorrected for
dilution) plotted as ppmC in Figure 111-68. All carbon products measured
in the butane/NO system exhibit similar concentration-time profiles, so
A
that systematic gross underestimation of one or more of these carbon con-
taining product concentrations (such as MEK) is a possibility. Concentra-
tions of the four-carbon nitrates (which undoubtedly form a portion of the
lost carbon) in the propylene/butane runs are about 20 percent of the MEK
concentration. However, these data, which also show a similar carbon loss
from butane, are more recent and presumably have improved analytical cali-
bration, thus confirming the large mass loss in the butane system. The
production of nydroxyatdehydes could account for some of the undetected
carbon mass.
3. Formation and Destruction of Methyl ethyl ketone
Our initial mechanism for the formation and destruction of MEK was
presented by Durbin et al . (1975):
0- 0
HO*
CH3CH2CCH + hv
OH- -* CH3COH
-------
95
0.7
,EC-48
/ADDED CH.CHO
180 240 300
Tinte--minutes
360
FIGURE 111-67. UNRECOVERED CARBON AS A FUNCTION OF TIME FOR
VARIOUS RUNS (NOT CORRECTED FOR DILUTION)
-------
96
0.7
60 120 180 240
Time--nrinutes
300
360
FIGURE 111-68. MEASURED MEK CONCENTRATION AS A FUNCTION
OF TIME FOR VARIOUS RUNS
-------
97
We now believe that OH- radicals will not attack the carbonyl part
of the molecule to give acetic acid, but rather that OH« radicals will
abstract the secondary hydrogen:
0 n °oO
II 02 |2n
+ OH- 4 C
OpO -0 0
rn i II
CH3CHCCH3 + NO -> CH3CHCCH3
oo
CH3CHCCH3 + 02 + CH3CCCH3 + HO"
This mechanism has been postulated by Demerjian et al. (1974). The rate
of the hydroxyl radical reaction has recently been reviewed by Lloyd et
al. (1976), and we have used their recommended value of 5.2 x 10 ppm~
min~ . The biacetyl formed from the above reaction will photolyze
rapidly to give two radicals:
00 n 0
« °2 H
CH3CCCH3 + hv £ 2CH3C02
The formation of these two radicals may lead to increased reactivity in
the latter stages of simulation of the butane system (see Figure 111-69).
Introduction of this reaction leads to approximately 25 percent more
ozone with approximately a 45 percent increase in PAN concentrations
for EC-39.
4. Simulation Results
/
The explicit butane mechanism is presented in Table 111-11. Simula-
tions were performed on UCR runs EC-39 through EC-48 with the initial
conditions listed in Table III-12a and the photolysis rate constants
-------
98
c
0
N
G
E
N
T
II
A
T
I
0
H
P
P
fl
. 100+
I
I
I
I
I
I
I
I
.075+
I
I
I
I
I
I
I
I
I
.050+
I
I
I
I
I
I
I
I
I
.025+
I
0.00+-
0
SPECIES EXPT. SIM.
03 * 0
03
EC-39
8iacetyl
100
ISO
200
TIME (MINUTES)
250
300
350
400
FIGURE 111-69. EFFECT OF INCLUDING BIACETYL PHOTOLYSIS
-------
99
TABLE 1 1 1-11. THE BUTANE/NOY MECHANISM
A
Rate Constant
___ _ Reaction _ (ppm-1 mi'n-1 )
N02 + hv * NO + 0(3P) Experimental*
0(3P) + 02 + H » 03 + M 2.08 x 10"5t
0(3P) + N02 + NO + 02 1.34 x TO4
03 + NO * N02 + 02 25.2
0(]D) + M -» 0 + M 8.6 x 104
O^D) + H20 -* 20H- 5.1 x 105
O + OH- * H0 + 0 87.0
03 + H02 * OH- + 202 1.2
0 + N0 -> N0 + 0 5 x 10"2
03 + hv-*- OD) + 02 Experimental*
03 + hv * 0(3P) + 0., Experimental*
03 -» wall 1 x 10"3
N03 + NO -> 2N02 1.3 x 104
N03 + N02 -» N205 5.6 x 103
N205 -f H20 -» 2HN03 5 x 10"6
N205(+ M) * N03 + N02(+ M) 24.0*
NO + N02 + H20 -* 2HN02 2.2 x 10"9t
2HN02 + NO + N02 + H20 1 .3 x 10"3
-------
100
TABLE I IT-IT (Continued)
Reaction
HN02 + hv -*- OH- + NO
N09 + OH-(+ M) + HNOJ+ M)
, NO + OH- * HN02
H02 + NO + OH- + N02
H02
HN0
hv-*- 20H-
CH3CH2CH2CH3 + 0 -+ CH3CH2CH(02)CH3 + OH-
°2
°2
HOCH2CH2CH2C{0)02 + NO
HOCH2CH2C(0)02 + NO ->-
+ NO -
°2
CH3CH2C(0)02 + NO -? N02
o2
CH3C(0)02 + NO -£ N02 +
°
HC(0)02' + NO
H02
NO
CH3CH(02)C(0)CH3 + NO
2 + CH3C(0)C{0)CH3
Rate Constant
(ppm~1 min-l)
Experimental*
9.2 x TO3
9 x 103
2 x 103
20
Experimental*
64.0
5.0 x 102
2.9 x TO3
2 x 103
2 x 103
2 x 103
2 x 103
2 x 103
2 x 103
2 x 10
3
2 x 10
3
NO
2 x 10-
-------
101
TABLE iIi-11 (Continued)
Reaction
Rate Constant
min"')
)2 + NO
: + NO H
2 x 10"
2 x 10"
+ NO H- N02 + H02 + HOCH2CHO
2 x
CH3CH2CH202 + NO + N02
2 x 10-
+ NO -» N0
2 x
NO
2 x
6.7 x 10
4*
CH3CH2CH2CH20- - HOCH2CH2CH2CH202
CH,CH,CH(0-)CH, + 0, * CH,CH,C(0)CH, + HOi
323 c. J c. J i
CH3CH2CH2CH20. + 02 + CH^CH.CHO * HO,
fu /*ij /*ij rt« x f\ -* pu f*u r*un 4. wn*
Ln^UnnUnoU T u« Ln-Ln«UnU ^ nU^
322 2 3 i i
CH,CH90- + 09 * CILCHO + HO:,
3 t t 3 <:
CH,0- + 0, * H7CO + HOX
O b k> W
20
H2CO + hv ? HO;, + HC(0)02
H2CO + hv » H2 -i- CO
20
CH.CHO + hv ? CHJk + HC(0)09
. O O t ^
20
CH,CH;,CHO + hv ^ CH-CH70;, + HC(0)Oj
5 -t J t * <-
?n
2
7.5 x 10
0.95
- 0.95
0.95
0.95
0.95
Experimental*
Experimental *
Experimental*
Experimental*
Experimental*
-------
102
TABLE III-ll (Continued)
Reaction
CH3CH2CH2CHO + hv + CH-CHO + C-H.
CH3CH2C(0)CH3 + hv
HOCH2CHO + hv -* HC(0)0^ + HjCO + HO^
HOCH2CH2CHO + hv -* HC(0)02
HOCH2CH2CH2CHO + hv + HC(0)02
20?
CH3C(0)C(0)CH3 + hv $ 2CH3C(0)02
H2CO + OH- -* HC(0)02 + H20
°?
CH3CHO + OH- -* CH3C(0)02 + H20
°2
CH3CH2CHO + OH- 4. CH3CH2C(0)02 + H20
°2
CH3CH2CH2CHO + OH- -4 CH3CH2CH2C(0)02' + HgO
CH3CH2C(0)CH3 + OH- * CH3CH(02)C(0)CH3 + H20
HOCH2CH2CH2CHO + OH- * HOCH2CH2CH2C(0)0^ + H
HOCH2CH2CHO + OH-
HOCH2CHO + OH- * H2CO +
HOCH2CH2CH2C(0)02 + HO^ * HOCH2CH2CH2C(0)OOH
HOCH2CH2C(0}02 + H02
Rate Constant
(ppm'l min~')
Experimental*
Experimental*
1 x 10
-3
1 x 10
-3
1 x 10
-3
2 x 10"3
1 x 104
1 x 104
1 x 104
1 x 104
5.2 x 103
1 x 104
1 x 104
1 x lb4
1 x 104
1 x 104
CH3CH2CH2C(0)02 + H02
1 x 10
-------
103
TABLE III-1T (Continued)
Rate Constant
_ Reaction _ (ppnr| min-T )
CH3CH2C(0)02 + H02 -" CH3CH2C(0)OOH + 02 1 x 104
CH3C(0)02' + H02 -» CH3C(0)OOH + 02 1 x 104
HC(0)02 + H02 -* HC(0)OOH + 02 1 x 104
HOCH2CH2CH2CH202 + H0£ -^ HOCH2CH2CH2CH2OOH + 02 4 x 103
CH3CH(02)C(0)CH3 + H02 * CH3CH(OOH)C(0)CH3 + 02 4 x TO3
i
02 4 x 103
2 4 x TO3
4 x 103
CH3CH202" + H02' -» CH3CH2OOH + 02 4 x TO3
CH302« + H02' -» CHgOOH + 02 4 x TO3
H02 + H02 » H202 + 02 4 x 103
CH3CH2CH2C(0)02 + N02 ^ CH3CH2CH2C(0)02N02 4 x 102
CH3CH2C(0)02 + N02 -« CH3CH2C(0)02N02 4 x 102
CH3C(0)02 + N02 » CH3C(0)02N02 4 x 102
02 -» N02 + CH3C(0)02 1.7 x 10~2*
CH3CH2C(0)02N02 -> N02 + CH3CH2C(0)02 2 x 10"2*
CH3CH2CH2C(0)02N02 -^ N02 + CH3CH2CH2C(0)02 2.5 x 10~2*
-------
104
TABLE III-1T (Concluded)
Rate Constant
Reaction (ppm'^ min"1)
CH30- + N02 * CH3ON02 2 x 104
n ,n3
CH-,0- + N00 -»- H0CO + HN00 2.2 x 10
3 i i 2
CH3CH20- + N02 -» CH3CH2ON02 " 2.0 x 104
CH3CH20- + N02 -» CH3CHO + HN02 2.2 x 103
2.0 X 104
CH0CH0CH00- + NO, -* CH0CH0CHO + HN00 2.2 x 10
322 2 3 i 2
CH,CH,CH9CH,0- + N09 * CH,CH9CH9CH9ON09 2 x 104
^ C. C. L. £. 3 £ L. L. C.
2.2 x 103
CH0CH0CH(0-)CH, + NO, + CH-.CH0CH(ON00)CH0 2 x 10
J L J L J C C 5
CH3CH2CH(0-)CH3 + N02 * CH3CH2C(0)CH3 + HN02 2.2 x 104
CO + OH- -^ H02 + C02 2.06 x 102
t Units of ppnf rain" .
* Units of min~ .
-------
105
TABLE III-12a. INITIAL CONDITIONS FOR SIMULATIONS OF BUTANE/NOV SYSTEMS
Run
Number
EC-39
EC-41
EC-42
EC-43
EC-44
EC-45
EC-48
NO
(PPm)
0.547
0.542
0.542
0.126
1.14
0.552
0.535
NO?
(ppm)
0.06
0.068
0.059
0.013
0.132
0.062
0.059
Butane
(ppm)
2.2
4.03
0.385
0.38
3.92
1.94
1.94
HN02
(ppm)
_ i .-pi _
0.01
0.01
0.03
0.01
0.03
0.006
0.0
kl -,
(miff1')
0.24
0.238
0.235
0.233
0.23
0.227
0.221
* HNOp data and aldehyde photolysis rate constants were chosen to provide
the best fit of predicted concentrations to measured concentrations;
all other data are UCR measurements.
TABLE III-12b. PHOTOLYSIS RATE CONSTANTS FOR SIMULATIONS OF BUTANE/NOV SYSTEMS
}\
EC Run - !-
Number 3
39'
41
42'
43
44 >
> 31.0
45 22.0
48 14.0
82
82
81
500
500
500
7.4
HCHO+H-+HCO-
9.6
6.9 8.4
6.7 7.4
24.0
24.0
21.0
7.2
6.5
5.5
16.0
14.0
11.0
12.0
10.0
12.0
MEK-*
5.8
5.0
3.6
-------
106
listed in Table III-12b. The factorial block for this system is pre-
sented in Figure 111-70. Figures 111-71 through III-103 show the
simulation results for the butane/NO system.
A
5. Discussion of the Butane/NO Systems
~" """ "lu " ' '" ~" "r"~Jr/\
In the butane/NO and propylene/NO systems, we assumed that the
A A
light source at UCR was deteriorating, especially in the short-
wavelength end of the spectrum. This is consistent with UCR data
that show a decay in nitrogen dioxide photolysis during these experi-
ments. Unfortunately, no spectral data were taken; so in the simula-
tions of runs EC-39 to EC-60 we varied the intensity of the,short
wavelengths using an exponential decay function. This function was
reported by UCR for the Pyrex filter.used in conjunction with its
light source. This procedure was reported by Durbin et al. (1975). The
filter function was merely shifted to longer wavelengths for later
experiments, thus progressively attenuating the light intensity between
300 and 350 nm. This procedure affects the simulations by producing
a consistent decrease in the amount of radicals arising from photolysis
of carbonyl compounds.
As mentioned in the preceding sections on butane mechanism
development, a considerable uncertainty is associated with the butane
system because of the substantial carbon mass not accounted for in the
UCR data. Another unfortunate aspect of the data is that in none of the
experiments did ozone reach a peak concentration. In all the experiments
the concentrations of products such as MEK and the aldehydes increased
continuously to the end of the experiment. Runs EC-42 and EC-43 show the
largest relative reaction of butane, about 35 percent. We therefore
recommend more experiments that run longer, to 12 hours instead of 6; that
use low NO to hydrocarbon ratios, (e.g., 0.1 ppm NO to 2 ppm butane) for
A A
faster ozone production; or that use even less hydrocarbon than the 0.4
ppm butane used in EC-42 and EC-43, for more complete reaction of butane.
-------
107
1.0
CL
0.5
0.1
- x42
43
0.4
39
x45
48
x 44
2.0 4.0
Butane--ppm
FIGURE 111-70. FACTORIAL BLOCK FOR BUTANE/NO SYSTEMS
/\
-------
108
l> *
to oo oo oo o o o o o o
0.60* » *
o so.08 loo.oa
* *
300.00 350.00 404.00
CC-39 . SPECIES 03
coacFjrnutioa SCALE FACTOR 10+0
FIGURE 111-71
* a >
00.00 I**.
a;*.08 400.
EC-M . Fi'F-niKH no2 M
FIGURE III- 72
-------
109
a
-------
FORMALDEHYDE
rr FF IT r r r F r r F
Bti.ee
F.c-3*
SOU.OB 330.OO 404.1
FIGURE 111-75
10*. *» t3lt.fi
sot.** a»«.*e -to*.*o
FIGURE 111-76
-------
Ill
0
30.00 IOV.M
CC-41 . 8FECIK9 03
CQBCE!rnUTICHI~9CALX FACTO* !»*
FIGURE 111-77
a a a a
R*lf * H
cc-41 . irccm Ka no
CfiKCFjrniATialt RCALE FACTOn !»
FIGURE 111-78
-------
112
COnCEKTHATIOH fiCUZ FACTOR 10**
FIGURE 111-79
EC-4I . srccita FA*
concuniuTioa BCMX FACIXM i**«
FIGURE 111-80
-------
113
r r r
r r
r r
r
r
r r r
r
r
r r
r
IT r T *T T
r
r
r
80.00 I6».»a IEO.M 200.00 258.00*
TIKK tnitm-ru;}
re-4i . OPEC its pawn COHCENTHATIUH SCAIX FACTOR i&*»
FIGURE 111-81
B
K
A A
* ' » * * ACETAL3EHYK
B
A
B A
H
K
B
H <
A
R
H A
n A
II A
KAA
230.0» 3W.W» 950.** 4M.M
EC-41 . BTCCIK0 HEX *U>3 COnCEKTTUTlOII FCAIX FACIOA !»*
FIGURE III-82
-------
114
o o
0
0 0
. Ffi.ctF.ii OJ
FIGURE 111-83
2 - H02
P - RO
«.«»*
FIGURE 111-84
-------
115
FIGURE 111-85
00.00 ICO.00
FIGURE III-B6
-------
116
H - KEK
A AJJW!
«.*2*
* ACETALDEHYOE
FIGURE 111-87
-------
CIGURE III-B8
T r" _RQ
iij. C.7
-------
118
FORMALDEHYDE
60.00
r,c-«
FIGURE 111-90
H A A
n A
n * A
H AA
> ACETALDEHYDE
A H
A
A n
. STCCICS I»X AI.U2
l*t (XAIT. rACTHH !&*
FIGURE 111-91
-------
119
O - 03
r - FOBS
».**
OZONE
r
r o
so.ca toe.oft ir.o.pw ji.'i.n* 2Se.oa 3'oo.n* nio.oo
Tint: (f!i>urt'*i>
FIGURE 111-92
K 0+4
wn +
o +
0 * +
o + * *
0 + V
0 * «
« * It
EC-44 . Krrcim fo vnz Lo>crjniiATto?i KCAU: rAirron
FIGURE 111-93
-------
120
OO
0 * *
0 ex
re-4+ . emits B'/T cirtCK.
-------
121
r r o
OOOOO 00 O O O
F.C- 4$ . pri£ i EM 03
3GO.04 *<*«,«»
FIGURE 111-96
cop-cr.irnuTios KCA
FIGURE 111-97
-------
122
JC1A.OO I3O.OO
DOO.OO O.'0.00
KCAIX FACTOR l»»»
FIGURE 111-98
If - MKC
A - AJ-JJ*
r - fiat
ACtTALDEHYDE
PAN
r
x x rxrrxrr rr
230.M 30*.B« 3SO.fKi
FIGURE 111-99
-------
1/.3
A A A +
AA AAA A AA
A AA A +A
AAA
A -A A
A AA AA A A
0 A A A
o on o
fie. eo
oo o «o * o
RCALT. rAcmn i*««
FIGURE III-100
2 > s a
FIGURE III-101
-------
124
0 *
o
o o
O « » *
§».** IM.ftn
ec-4« . vnxica ID
KCALE FACTOR |**e
FIGURE 111-102
Tax
r - row
r - rAx
HEC
T
R
« t.W
rr r rxrrrrrrrr r r
* r *
r
.' r
. r
Tim: i
ec-40 . mrccite ua rvw rut cwcomuTitMi KCAIX rAcnm »
FIGURE III-103
-------
125
Table 111-13 presents a comparison of the experimental and simulated
maximum one-hour-average ozone concentrations. (In all UCR butane/NO
/\
experiments, the highest ozone concentration was measured in the last
hour.) The average difference is approximately +17 percent with a stan-
dard deviation of 34 percent. These estimates are made excluding EC-44.
Including EC-44, the average difference is +36.7 percent with a standard
deviation of 60 percent.
TABLE II1-13. PREDICTED AND MEASURED MAXIMUM ONE-HOUR-AVERAGE OZONE
CONCENTRATIONS FOR THE BUTANE/NOV SYSTEMS
A
EC Run
No.
39
41
42
43
44
45
48
Maximum One-Hour Average
Ozone Concentration (ppm)
Measured
0.056
0.21
0.0042
0.113
0.011
0.12
0.14
Predicted
0.048
0.33
0.005
0.097
0.028
0.13
0.22
Percent
Difference
-14.3
57.1
-19.8
-14.7
154.5
8.3
57.1
In our opinion the poorest fit to the data is the simulation for
EC-42, even though EC-44 has the worst fit to the ozone behavior (Table
111-13). This experiment is discussed in the section on development of
the carbon-bond mechanism (Section IVC-1). It has the highest NO to
J\
hydrocarbon ratio and the six-hour experiment shows development of smog
formation only to the NO/NO^ crossover point. Either the mechanism lacks
the proper number of NO to N02 conversions per butane reacted when the
NO to hydrocarbon ratio is this high, or some other source of reactivity
^\ ,
is present. In simulations of the propylenc system, run EC-21, a
-------
126
similar problem exists. In both EC-42 and EC-21 we have shown simulations
in which we used the maximum initial HNCL concentration possible from
the equilibrium between NO, N02, and H20. In both cases it is obvious
that too many radicals are present initially and that the maintenance
source of radicals in the mechanism is inadequate. We could not justify
changing the light spectrum for these two simulations without changing
the light source for adjacent experiments. We speculate that the walls
of the chamber in some way supply radicals to the peroxy-oxyl radical
pool. The effect of such a process would be greatest when the concen-
tration pf normal radicals was the lowestin a low activity and low
hydrocarbon experiment such as EC-42 or EC-21. That the walls are
supplying radicals is supported by the similar need for high HN02
(relative to equilibrium) to simulate the overall chemistry of EC-43
and EC-17. In EC-43 as in EC-42 it is clear that the initial radical
pool in our simulations is too high compared with the rate of reactions
seen in the data.
We feel that the worst aspect of the butane simulations in general
is associated with the PAN chemistry. The PAN chemistry in the various
iT,echanisms was adjusted to give the best overall fit to the measurements
of the olefin systems, where- PAN is more abundant than in butane systems.
PAN is formed from the reaction of peroxyacetyl radicals and NCL:
0
CH3COO- + N02 * PAN
In the butane mechanism a major source of these radicals is the transfer
reaction of hydroxyl radicals and acetaldehyde in air:
0
OH- + CH3CHO * H20 + CH3COO-
-------
127
According to our present formulation, these radicals compete primarily
with N02> as shown above, and with NO:
CH3COO- + NO -» CH30£ + C02 + N02
The reaction with NO leads to formaldehyde formation via the following
reactions:
CH30^ + NO * CH30. + N02
CH30- + 02 -*. HCHO + H0|
The simulation of EC-48-clearly shows that our PAN chemistry is
incorrect and suggests what can be changed during our efforts in tne
near future. Acetaldehyde and formaldehyde were added initially to
systems otherwise similar to EC-39 to make up EC-48 and EC-45, respectively.
In our simulation of EC-48, far too much formaldehyde is formed, too
little PAN is formed (even though it is much more than in EC-45), and "too
much ozone is formed. If the rate constants in the mechanism were changed
so tnat-the reaction of peroxyacetyl radicals with NO was much slower or
their reaction with N02 was much faster or both, the mechanism would
simulate the measurements of formaldehyde, PAN, and ozone more closely.
However, this change would result in poorer simulations of olefin systems.
This is discussed in the preceding section on propylene chemistry.
Another implication of this change is that formaldehyde predictions
would be lower than at present; many of the formaldehyde predictions
are already much lower than measurements.
In summary, the explicit mechanism for butane provides simulations
that follow the observed NO to NQ2 conversions per butane reacted. PAN
and formaldehyde are exceptionsthe .predicted concentrations of those
species are generally low. As discussed above, the PAN chemistry in the
mechanism is probably incorrect.
-------
128
C. PROPYLENE/BUTANE/NOX CHEMISTRY
UCR performed a set of experiments on a propylene/butane/NOv system.
A
To simulate this system, we combined the explicit propylene mechanism
and the explicit butane mechanism. Initial simulations with the "com-
bined" mechanism showed a rapid loss of NO during the simulation.
A
The experiments with propylene and butane have a high hydrocarbon
to NO ratio and, because of the propylene, a high reactivity. Such a
A
combination rapidly produces peroxy radicals in the presence of rela-
tively low concentrations of nitrogen oxides. If the peroxy radical
concentrations get too high, the NO concentration is driven so low
(through the reaction R0£ + NO) that NOX then reacts exclusively with
N0~ to give N90,-. The increasing N90,- concentration in turn drives NO
L. C. 0 L. D X
out of the system via the reaction with H^O producing nitric acid (HNO-,).
Owing to such high concentrations of peroxy radicals, the radical-
radical reactions
R0| + R'O^ -* Products
V *
may be significant reactions fn this system. The reaction rates and
products are not certain. Deirterjian et al. (1974) estimated relative
rates for primary and secondary radicals based on some liquid phase
data by Bennett et al. (1970). In our mechanism, the values for the
radical-radical reaction rate constants estimated by Demerjian et al.
(a/IOC ppm~ min" ) seem too low to account for the observed removal
of the peroxyalkyl radicals. These rate constants lead to unrealisti-
calTy high levels of peroxyalkyl radicals (^ 10~ ppm) and unrealisti-
A -5
cally low NO levels (^10 ppm). Faster rate constants for radical-
A
radical reactions must be used in our mechanism because the peroxy
radical reactions with nitric oxide have a faster rate constant in
our mechanism (2000 ppnf min" ) than in Demerjian's mechanism (100-
500 ppm" min" ). We are using a value of 4 x 10 ppm" min" for the
-------
129
peroxy-peroxy radical reactions, which gives adequate fits to the
observed data. However, the products of these reactions are unclear.
For hydroperoxy radicals the formation of hydrogen peroxide is
accepted, but the reactions of pero-xyalkyl and peroxyacyl radicals
might lead to different radicals:
H02 + R02 * ROOH
R02 + R02 + RO- + RO
An alternate pathway for this last reaction is the formation of ^an alcohol
and an aldehyde. Weaver et al. (1975) suggested approximately an equal
split, between these oaths-, thev also have data for the alkoxyl path
occurring between 18 and 75 percent of the time. (See Chapter IV for
further discussion of these reactions.)
1. Simulation Results
We performed carbon mass balance analysis on UCR's propylene/butane/
NO system. Results of the analysis are presented in Table 111-14. The
/\
carbon recovery values for the propylene/butane/NO system fall logically
J\
between the results for the propylene/NO and the butane/NO systems.
A X
The better balance of EC- 11 4 is expected from its larger propylene/butane
ratio.
Simulations of the propylene/butane/NO system were performed on
/\
UCR runs EC-97, 99, 106, 113, 114, 115, and 116 with the initial condi-
tions listed in Table III-15a and the photolysis rate constants for
these simulations listed in Table III-15b. The factorial block for
this system is shown in Figure III-104. The kinetic mechanism for
the simulation is presented in Table 1 11-16. Simulation results for
this system are shown in Figures 1 1 1-105 through 1 1 1-167.
-------
130
TABLE 111-14. PERCENT CARBON MASS RECOVERY FOR
PROPYLENE/BUTANE/NO SYSTEMS
/\
Time
(minutes)
60
120
180
240
300
360
390
*
Data not
EC-97 EC-99
65% 81%
* 75
74 75
73 77
67 73
* *
* *
available.
ULK Kun Number
EC-106 EC-113 EC-114 EC-115
86% 85% * *
82 74 87% *
76 66 88 *
72 58 80 *
* * * *
* * * 74%
* * * 68
EC-116
76%
64
75
72
66
*
*
TABLE III-15a. INITIAL CONDITIONS FOR SIMULATIONS OF
PROPYLENE/BUTANE/NO₯ SYSTEMS*
A
Run
Number
EC- 97
EC- 99
EC-106
EC-113
EC-114
EC-115
EC-116
NO N02
(ppm) (ppm)
a 397 0. 088
0.407 0.09
Q401 0.102
0.091 0.02
0.794 0.204-
0.402 0.104
0.391 0.104
Propylene Butane HNO£
(ppm) (ppm) (ppm)
0.5 2.1 0.005
0.4 2.0 0.01
0.402 2.0 0.001
0.41 2.08 0.001
0.766 3.66 0.005
0.310 2.94 '0.002
0.824 4.00 0.002
k1
(min'1)
0.351
0.351
0.351
0.351
0.351
0.351
0.351
HN02 data and aldehyde photolysis rate constants are chosen to provide
the best fit of predicted concentrations to measured concentrations;
all other data are UCR measurements.
-------
131
TABLE III-155. PHOTOLYSIS-RATE CONSTANTS FOR SIMULATIONS
OF PRORYLENE/BUTANE/NO₯ SYSTEMS
/\
(104 min"1)
97
99
106
113
114
115
116
46 350 670 8.1
8.8
CH3CH2CH(>*
25.0 6.65 7.4
MDt*
10.5
5.2
1.0
E
CL
CL
i
X
i 0.5
0.1
*m
97
X "
106 x!15 X116
x!13
i i i
2 4
Hydrocarbon--ppm
FIGURE III-104. FACTORIAL BLOCK FOR PROPYLENE/
BUTANE/NOY SYSTEMS
/\
-------
132
TABLE II1-16. THE PROPYLENE/BUTANE/NOX MECHANISM
Rate Constant
_ Reaction _ (ppnrl nrin-1)
N02 + hv -» NO + 0(3P) Experimental *
0(3P) + 02 + M) -» 03 + M 2.08 x 10"5
0(3P) + N02 - NO + 02 1 .34 x 104
03 + NO ->- N02 + 02 25.2
O^D) + M + 0 .+ M 8.6 x TO4
0(TD) + H20 -> 20H- 5.1 x 105
0 + OH- -» H0 + 0 87.0
03 + H02 * OH- + 202 1.2
03 + N02 -»- N03 -f Og ,5 x 10"2
03 + hv * 0(3P) + 02 Ex-perimental *
03 -» wall 1 x 10"3
N03 + NO * 2N02 1.3 5( 104
5.6 x 103
2HN03 5 x 10"6
N205(+ M) * N03 + N02(+ M) 24.0
*
Q
NO + N02 + H20 * 2HN02 2.2 x 10"S
-------
133
Table 111-16 (Continued)
Reacti on
2HN02 * NO + N02 + H20
HN02 + hv -> OH- + NO
N02 + OH- (+ M) -» HN03(+ M)
NO + OH- + HN02
H02 + NO * OH- + N02
H02 + N02 * HN02 + 02
H202 + hv-*- 20H-
°?
CH3CH2CH2CH3 +0-4 CH3CH2CH(02)CH3 + OH-
CH3CH=CH2 + 0 * CH3CH2CHO
CH,CH=CH, + 0 * CH,Oj ^~CH.C(0)OA
3 i $ e. 3 t.
202 ,
CH.CH-CH, + 0 * CH,CH,0; + HC(0)0,
3 L. -J L. f. C.
^ CH CH CH CH 0- + H 0
CH3CH2CH2CH3 + OH- S C«3CH2CH(02)CH3 * H20
CH3CH=CH2 + OH- -£ CH3CH(02)CH2OH
CH3CH=CH2 + N03 -" N02 + Products
°2
CH-CH=CH, + 0, -^ H,CO + CH-,C(0)OA + OH-
«5 t O C. J L.
CI^CH^^ + 03 * CH3CHO + HC(0)02 + OH-
CH3CH=CH2 + 03 CH3CHOOCH2
Rate Constant
(ppm-1 min-1)
1.3 x 10"3
Experimental
9.2 x 103
9 x 103
2 x 103
20
Experimental
64.0 -
1 .77 x TO3
1.77 x 103
1.77 x 103
5.0 x 102
2.9 x 103
3.8 x 104
7.82
0.005
0.005
0.005
-------
134
TABLE 111-16 (Continued)
Rate Constant
Reaction _ (ppm"1 nrin"1)
HOCH2CH2CH2C(0)02 + NO * N02 + HOCH2CH2CH202 + C02 2 x 103
CH3CH(OH)CH2C(0)02 + NO -» N02 + CH3CH(OH)CH202 + C02 2 x 103
HOCH2CH2C(0)02 + NO -> N02 + HOCH2CH202 + C02 2 x 103
CH3CH2CH2C(0)0^ + NO -£ N02 + CH3CH2CH202 + O>2 2 x 103
^9 3
CH,CH-C(0)OA + NO 4 N09 + CH.CH^O, + C09 2 x 10
*3£ L* £ o££ c.
+ NO -+ N02 + CH302 + C02 2x10
HC(0)02 + NO -2 N02 + H02 + C02 2 x 103
HOCH2CH2CH2CH202 + NO -» N02 + HO^ + HOCH2CH2CH2CH2CHO 2 x TO3
CH3CH(0|)C(0)CH3 + NO » N02 + H02 + CH3C(0)C(0)C«3 2 x 103
CH,CH(Or)CH,CH, + NO -» NO, + CH,CH(0- )CH0CH, 2 x 1C3
O L c o tO tj **»»
CH3CH2CH2CH202 + NO -> NOg + CH3CH2CH2CH20- 2 x 103
CH3CH(02)CH2OH + NO -»- N02 + Ch'3CH(0-)CH2OH 2 x 103
HOCH2CH2CH202 + NO * N02 + H02 + HOCH2CH2CHO 2 x 103
HOC^CH^ + NO * N02 + H02 + HOCH2CHO 2 x 103
CH3CH2CH202 + NO + N02 + CH3CH2CH20- 2 x 103
CH,CH,OA + NO -» NO, + CH-CH,O- 2 x io3
O L, £. c. -j t
CH302 + NO -» N02 + CH30- 2 x TO3
CH.CH,CH(0-)CH. -^ CH,CH,Oi + CH.CHO 6-7 x ll)4
3 £ o w
-------
135
TABLE III-16 (Continued)
Rate Constant
Reaction
09
CH,C(0-)CH9OH -£ CH.CHO + H9CO + HO;
J C O c c
°2
CH,CH9CH(0-)CH- + 0* ->- CH,CH,C(0)CH, + HO;
O t O <- O c O £
CH,CH9CH,CH,0. + 0, -v CH.CH,CH,CHO + HO:
3 c. c c 2 322 2
r*u ^ii r*ij r*» j. A _*. PLJ f*\j /*un x un*
wnov»no**noU u« «" ownownu ~ nuo
322 2 32 2
CH,CH,0- + 0, -»- CH.CHO + HOo
J t & O <-
CH,0- + 09 -»- H,CO + HO;
O £ t C.
20?
H£CO + hv * HO^ + HC(0)02
H2CO + hv -»- H2 + CO
CH..CHO + hv 1 CHJk + HC(0)09
<3 j C. £
20
CH,CH,CHO + hv -3 CH,CH,OA + HC(0)Oj
3 C. 3 t. £m £
209
CH,CH9CH9CHO + hv * CH,CH?CH9Oi + HC(0)0;,
0 ^ ^ j t t i, t
CH,CH9CH,CHO + hv » CH.CHO + C-H.
0 t c O t H
CH,CH«C(0)CH, + hv * CH0C(0)Ox + CH,CH,Oi
a e. a J f- 3 £ c.
HOCH,CHO + hv -*- HC(0)Ox + H9CO + HOA
£. C C. £.
HOCH,CH9CHO + hv * HC(0)OA + HOCH7CH-0?
c e. c. f- f- e.
HOCH9CH9CH9CHO + hv -» HC(0)OA + HOCH9CH9CH?0$
.£.£. c. e. c e. c
CH,C(0)C(0)CH. +. hv I 2CH,C(0)0:
5 *
3 x 10°
c *
7.5 x 10°
0.95
0.95
0.95
0.95
0.95
*
Experimental
*
Experimental
Experimental
*
Experimental
^
Experimental
*
Experimental
Experimental
1 x 10"3
_»
1 x 10 J
_o
1 x 10 J
2 x 10"3
-------
136
TABLE 111-16 (Continued)
Rate Constant
_ Reaction __ (ppm-1 min"1)
H2CO + OH- -£ HC(0)02 + H20 ] x 10
0 4
CHgCHO + OH- -2 CH3C(0)02 + HgO 1 x 10
CH3CH2CHO + OH- -£ CH3CH2C(0)02 + H20 ] * 10
Oo 4
CH3CH2CH2CHO + OH- -i CH^CH^CO^ + H20 1x10
CH3CH2C(0)CH3 + OH- -v CH3CH(02)C(0)CH3 + H20 5.2 x 103
HOCH2CH2CH2CHO + OH- * HOCH2CH2CH2C(Q)0;, + H20 1 x 104
HOCHCHCHO + OH- * HOCHCHCOjO + H0 4
22 + OH- * HOCHgCHgCOOg + H20 1 x 10
4
HOCH2CHO + OH- * H2CO + HO; 1 x 10
HOCH0CH,CH,C(0)0; + HO; + HOCH,CH0"cH,C(0)OOH +0, 1 x 104
£.£.£. c. L --*£ C £.-. ^ c.
HOCH2CH2C(0)02 + H02 -» HOCH2CH2C(0)OOH + 02 1 x 104
^
CH3CH2C(0)02 + H02 *'CH3CH2C(0)OOH + 0£ 1 x 104
CH3C(0)02' + H02 -»- CH3C(0)OOH + 02 1 X 104
HC(0)02 + H02 -» HC(0)OOH + 02 1 X 104
HOCH2CH2CH2CH202 + HO^ -* HOCH2CH2CH2CH2OOH + 02 4 x 103
CH3CH(02')C(0)CH3 + H02 -» CH3CH(OOH)C(0)CH3 + 02 4 x 103
CH3CH2CH(02')CH3 + H02' * CH3CH2CH(CH3)OOH + 02 4 x 103
! ^H ptl t\ ^- MO * *^ f^M f*H ^H f*M Of\H -4-0 Jl ** irt
-------
137
TABLE 111-16 (Continued)
Rate Constant
_ Reaction _ (ppm-t nrin-1)
CH3CH(02')CH2OH + HO^ -* CH3CH(OOH)CH2<3H + QZ 4 x TO3
CH3CH2CH20£ + H0£ -> CH3CH2CH2OOH + 02 4 x 103
CH3CH202* + H02' -» CH3CH2OOH + 02 4 x TO3
CH302- + HO^ * CH3OOH + 02 4 x TO3
H02 + H02 -* H202 + 02 4 x }03
CH3CH2CH2C(0)02 + N0£ * CH3CH2CH2C(0)02N02 4 x TO2
CH3CH2C(0)02 + N02 -» CH3CH2C(0)02N02 4 x 10
CH3C{0)02 + N02 * CH3C(0)02N02 4 x 102
CH3C(0)02N02 * N02 + CH3C(0)0.2 1.7 x 10"2 *
N02 + CH3CH2C(0)02 2 x 10-
*
CH3CH2CH2C(0)02N02 * N02 + CH3CH2CH2C(0)02 2.5 x 10"2
CH30- + N02 -» CH3ON02 2 x TO4
CH30- + N02 * H2CO + HN02 2.2 x 103
CH3CH20- + N02 * CH3CH2ON02 2.0 x 104
CH3CH20- + N02 -» CH3CHO + HN02 2.2 x TO3
Cti3CH2CH20- + N02 + CH3CH2CH2ON02 2.0 x 104
CH3CH2CH20- + N02 * CH3CH2CHO + HNOg 2.2 x 103
2CH20- + N02 * CH3CH2CH2CH2ON02 2 x 104
-------
138
TABLE 111-16 (Continued)
Reaction
Rate Constant
(ppnT1 min"1)
2-2
CH3CH2CH(0-)CH3
CH3CH2CH(0-)CH3 + N02
CO + OH- -* H02 + C02
2CH302- -v 2CH30- + 02
2 x 10
4
2-2 x 10
2-06 x 10
4 x TO3
2CH3CH202' -> 2CH3CH20- + .0£
4 x TO
4 x 10
4 x 10
2CH3CH2CH(0^)CHx:»; 2pf,CH2CH(0-)CH3
4 x TO
4 x 10
4 x 10
^ + CH3CH2CH2CH20^ * CH30- +
2 + CH3CH2CH(02')CH3
4 x TO
4 x 10
4 x TO
CH3CH202 + CH3CH2CH2CH20^
^ + CH3CH2CH(02)CH3
4 x 10
4 x 10
+ 0
4 x 10
4 x 10
-------
139
TABLE 111-16 (Concluded)
Rate Constant
Reaction (ppm"1 min"1)
CH2OH + 02 4 x TO3
CH3CH(02')CH2OH + CH3CH(0^)CH2OH -*- CH3CH(0-)CH2OH + CH3CH(0-)CH2OH + 02 4 x 103
-2 -1
t Units of ppm min .
_i
* Units of min .
-------
140
0-03
X - SO
3 - K02
O O 0 & 0
O CO O 0 O O
0 »
33 22 a O
X 0 2
2 *
0 2
X 2
1C S
X
3 X
2
x z a
2
X 2 .!
O 00 0 O > HH KB K * JtflHtRX
FIGURE 111-105
CU..IUUHATIOJ KALE rACTOS
FIGURE III-106
-------
141
one
ooo
* o e
oo o
* O 00
o o
*OOOO
oo o
* oo
50.00
EC-97
100.09
5TECIES HOT
seo.ot sso.
OS. SCALE FACTDa 10*»
FIGURE II1-107
. m PAK
E B H K
II K
r r p trrrr
*KEX
n n *
H n i
n » »rr rp
C6.OO IOC.«0
'/? . ercc i ts r«v
KCAU: FACTOR !«
FIGURE II1-108
-------
142
A A A A AAAA^AA A AA A A A
A FF *
A 7
A F
*ACETALOEHTDE
A f
A *
A r
A r
F
r
r *
SC.QO
EC-* 7
100.00
150. «J £00.M 230.« 3C3.CO
TIKE fEI"yTZS)
FTSC1ES ALS2 FOIUI CO'{C"rrmATIO,i FCALE FACTOR 10*0
FIGURE III-109
r rp p r r rrrrrrr r r?
n L r P."
aoo.ii L a-o. *o
FIGURE 111-110
-------
143
n n B n n
"METHYL NITRATE
. FTZCIES ncra caro concGmurio* SCALE FACTOR i»*«
FIGURE Ill-Ill
*ETHYL NITRATE
300.CO 330.ftO 490.1
tiK.SE 6 6 SSSEG
EC-*? . erne i rji sou
U.! KCUX FA.CT)W
FIGURE III-112
-------
144
0-03
* - r:o
2 - tea
t.to-
o e o
o t> a Q ooooooo
00
22
222 22 2
2 2 22 O
I 2X20
X X 32 O
n 2 u 2
X 0 22
2 X 0 X 2
H O » X
3 P* 0 2
2X
2 II 0 * 2
X 0 X
2 » 22
K 0 * X2
^ * o x
no 2 x
30 S 2
t.* * X 2 5
Of-' X 2 2
] 0 *I.ff *+ K 2X 2 2 2 2
i o ooo ««>* i,«s i;n n^* rn *sa *xn ** »* »* *» ** *x x
oa KG tuts . t^&cLimuTib:; ttcux FA r r r>r
«oo.to
ic-<« . mciEs AUQ nor COCCOTTIUTIK: CCAU mam i»i
-HGURE-III-T14
-------
145
0*O * c
O 0
O 0
0 O
M.OC JCtt.M
CC-+9 . BTtCIKS DOT
3Oa.Ofl SSO.60 40S.CO
FIGURE 111-115-
r rr r r rr
nr 77 FF
FTT r r
r 7 *
r
r
r «
FIGURE III-116
-------
146
Bl>7YRALDEHYD£
pp r*pr 7 r PP PP PP P
p PP P P PP P P PPPPP * P
B * PP i
B PP*?
IB BP P
CO. CO *1M.OO 139. OS 2C9.CO 25O.CO 3OO.OC 35*. «* 4OO.OO
T1BE (KlTtTES)
EC-M . SPECIES UJD9 ALIH OUICSTntATIM SCAUi TAC7DR !#*»
FIGURE II1-117
BC-4V . SI'tX I fti XTJC
Tine (hlM-TiJC*
ATim KCAU: fAcron UK*
FIGURE III-118
-------
147
s IT n n K M ft n m
E E K E E E K R K E EC EE
FT f.
n r,
K £
KETHYl NITRATE
n E
E
HE
n E
ETHYL HITRATE
SO.CO 100.00
TIKE CK1RLTIS)
250.00 SCO.00 330.OO 409.»
FIGURE III-1ig
6 b- 66 6
K Kfcfigs t S 6S
FSS BSS fiS
M.to* " 1AA.O*
*9 . vrecicac sen
.0* 3Sb.M 440. (in
FIGURE III-120
-------
148
22Z 22
22 22 O
2 2 2 O
-XX =2
X X X O 2
2X o:c a
2 O X 2
W.OO
TEC-IW
360.09 420.
SPECIES O3 BO KOS COTCEflTTHATICJI SCALE FACTOR I»*0
FIGURE II1-121
r +
P T
r +
r rr »
p
p r p »
r p p» r»r P-PP
to. »o IW.M iu
EC-ICC - *T£CILS ALII2 PROP
fiCJUX KACTri.t f**0
FIGURE II1-122
-------
149
oo *
e o
tao.oa 24o-*ru aMt.ao a**.** +20.w>
"Tir*; iKiffims)
. SPECIZB OUT
FIGURE III-123
FORHW.DEHTM
PWI
F *
F F
* r r
« !rr re '
«rttir.
-------
150
prr r;p;';* rr re r." ** ! p r r r
i- p r ?? r ? P r f F r
n » r? r
n; - rr r
E a ]!'!
\ rrr
re-1 IK. . KI-..C i rs «UH
FIGURE III-125
?? »
120.00 leo.oo
420.09 'X30.WJ
ZC-106 . SPTCIES UK
SCALE FACTOR 1»*0
FIGURE III-126
-------
151
KS. K n n H
E E E E
E E EE E E £ E E E E E
K E £
H E £
* *
METHYL HI-TRAT-;
K L
E
r. L
ETHYL
£ »
. BCALL FACJT»H 10*0
FIGURE III-127
9 SS SS B
S SS S S *8 S S S $3 G 3 S 9SSS5
9 S5WJ WSS
420. OO
-------
152
0 0 0 00 O 0
O 0 O 0 O 00
P 0
P 0
* 0
IBO.O6 ISU.OO
SPKCIKS 03 rnor cojicr.mi.'.Tio^ s .ALE FACTOR MH«
FIGURE III-129
IP
J ..«-
» n » * ** * «
iMt.M £4*A.nb 250. OO
Tint IKJKUTLS)
-113 BTHCltK
FIGURE Ill-ISO
-------
153
*D 0
O*OO
o oo
0*00
0 00
00
* 000
0 0
» 0*0
*O 0*
0*0 0
0 0-
0 O*
*o « o * * *
O 0
o *
Co. eo lOo.&o 150.00 3*-f. nn ^r.o.iio aoo.oo ario.oo o.n&
EC-113 . srteiru tarr
FIGURE III-131
r P r
p f P
p
FIGURE III-132
-------
154
A A A AA A A A AA
A A
A A *
Q CO.CU KM).Mr
EC-iia . FranF* ALDS cnscramuTKi;; SCALE FACnm 10*0
FIGURE III-133
r vr r , ' P '
r p p pp i- p
p P f p
P r r p r
pp r p
ITFP t
p PP ppp pp
Co.tto IPO.OQ ISO.*0 2*10.nn n5o n« «oe M .5- _."'
Tlni-.tBIMrn^) -. UI.M
EC-Ill . CPCC1ER ALTO CIMCrjTniATIlia SCAU1. rACTtll I9.O
FIGURE III-134
-------
155
B B
B B D
BOD
BB B D
BBC
B
D 8
D
B
ED
CUkCKjiT4t.'.TJ(rii KUU.C FACTOR |(t*0
FIGURE III-T35
T: r
K K
*T (1
H
n
n
n *
n
OM.M 990.M ecu,
FIGURE III-136
-------
156
O - 03
R - HO
a - KC2
ooooooo
X 2
2
X 2
2 O
X 2 O
O 2
0 K * *
OO Ktl
Ki D OOOO O» 1!
SU.OO IOO.O»
X 2 S
FIGURE II1-13?
rr
IT
r
r s
r
r
SO.C3 100.M 150.00 yit. ;...
TIB*. «;.;:.i.Tr.s
ii-j . Kr>ru>* r»o?
FIGURE II1-138
-------
157
o*o o» *
ooo
ooo «
00 *
0 0
o o * - *
0 O *
O 0
O 6
SO. 00
EC-114
lea.tte iso.o
or. SCALE FACIT:;I KHO
FIGURE III-139
* P
P
r
* rr
p
p P
rrr rr
EC-114 - CrCCtf* PAD
FIGURE III-140
-------
158
A «.9O
T
1
FORKAIDEHYDE
r F
r r
r F
ft w » n n r. r.
n K x
ir r Hrcd-m"!
EC-M'* . SPEC I m FORK K«
FIGURE III-141
* *.
SO.CO 1OO.O0-
EC-114 . CTI9C11?! ALH3
KC*U: rACTOtt I***
FIGURE III-142
-------
159
BUTYRALBEHYD:
ma
p isa
pp u
* P 3
pr nil
p p n
ico. oo ir-o.oo
2*0. £3 aofl.ftO 3"O.C*t
FIGURE III-143
co.ea IOO.
EOII4 .
M ttw.<
FIGURE III-144
-------
160
322223122
2X2
3 2
2 2
n-BUm KITRATt
NITRATC
2 =
2 « 44
! 32 4 4444 4444 44 444
SO. O3 IOO.«3
r^cTa w»*a
FIGURE III-145
CSS
SS K
K KS
«S
ico.co iso.ea s~i<>.f..i xso.co
Tlia: i/;t^trn,«.t
. cr/xio; BCTO cwtcumuTKUi KCAU: r*cr»r ii»«
FIGURE III-146
-------
161
o - i
K - !
3 - !
O 0 0 O ti
oo oo *o o o
OOOO O O 0
22 22i 0
22 2
22 O
! 2
XXX 220
X 02
X 2
0 2 i
X 2
0 2
X 2
0 2
* X 2
2 + fl
0
O X
XX V. 9 2 2
co.no 11:0. o<» i!;«.i'fl
FIGURE III-147
p rrr FT r rr
* r rrr FJTF F rr
|*P FIT ' T₯F 7
rr* r FT r
r rr* ? r r
p
P *
*r * *
ec- nc . ffFfccn:s pnw ronn
Tios HT.M.I: KACIIIU i6«
FIGURE III-148
-------
162
c KC/.LJ; KACiun 10*0
FIGURE III-149
PAN
* p
p;-
* r P
r r
P PP PPTP
FIGURE III-150
-------
163
BBBBBBBBBQ
* <*
rr I'piT r rr?ri*
r? ri- P p r r
C P IT
. Rprnrs AUC AJJM con'jFjn-L*,TJTMATIti:< *CA1
:HMI . (ni
r KACiur.
FIGURE III-152
-------
164
t »
''if' .00 SC.i.Crt
FIGURE ITI-153
E E r E E E
i."- »r. i;;-. Kim:1, it i;;i ];:,i;
34.0. »a *^y. Gti
FIGURE III-154
-------
165
s H B s s a s fi fr -e e 6
ssss FW us
120.00
. sr-cc t K>> u
FIGURE III-155
-------
166
0 O OO OOO OOOOOOOOOO 0 O O » *
00 0 o o* o* o *
« O 00 '
O 0
O 0
PO
or
0 P*
O P
P+
IT r?
» P P r ;
- r P r
FIGURE III-156
222
2 2
232
S 2
222
toe.oo ICO.do 200.nn enc
TIB: (eivrr;:-;)
FIGURE III-157
-------
167
i o*
1 CO *
D+ O 0
0 O *
0
O 0
oo * *
O 0
0 0
FIGURE III-158
cor;ci.Tn>jvT!o>; M:.!.:: r/.craf.
FIGURE IIJ-159
-------
168
co.oa
EC-i ie>
3..00
KCA1X fACTuR 1**0
FIGURE III-161
-------
169
p r i
r? rr .-PPP 7 PPPP p r
FP rr rerr
r r
* p
p
p p
rr pp
EC-116 . SPECIES A1JB3 COSCr.BTW.TlOS SCAI.K VACTOfl IO*O
FIGURE II1-162
C B £
DODOBBU B B t
LBB B E
B D B
D
a
D
11
C0.»» IOO.>0
AUrt
FIGURE III-163
-------
170
CO,60 icn.n
EC-lit . SPECIES
Tin*, if:'1.'.'_>
FIGURE III-164
EC- 111 . PF W: IKS
FIGURE III-165
-------
171
n-BUm NITRATE
E ti£5. E CEEE K E
K E K E E
ETHYL K11P.AU
err. r. r. nrn r.r-.n nn nn D it nnn im mum r.i hi htirji n 5;m'.i u n n p. H n n BE c n n r. n r.
FIGURE III-166
J»
* m
SS KS, G ft&SK S 6 S S
300. w* ar.o.OO
FIGURE II1-167
-------
172
2. Discussion of the Propylene/Butane/NO,, Systems
-n , ;'J{ -^-.
This set of experiments represents the only test to date of more than
. one hydrocarbon using the present explicit mechanisms. It also represents
the highest levels of ozone formed in all the experiments reported to us
by UCR--0.74 ppm ozone was measured in runs EC-114 and EC-116. The combined
mechanism for propylene/butane totaled over 120 reactions, and the computer
program carried some 54 different species; so this set of simulations was
the most severe test of numerical efficiency. A typical simulation required
about 11 seconds of-computing time on a CDC 7600 using our present version
of the CHEMK- computer program.
The only reactions added to the sum of the explicit mechanisms for
propylene and for butane were the cross peroxy radical/peroxy radical
reactions-.- AS"expected from the-high hydrocarbon concentrations and
resulting high ozone concentrations in these experiments, the simulations
indicated~.vepy».hlgh,..per-oxy,_ra.dical concentrations (e.g., HO;, was 2.7 x 10
ppm at two hours in run EC-116). The peroxy-peroxy radical-reactions are
more important in this series of experiments than in other series because
their rate of reaction depends oft the square of their concentration. In the
simulation for EC-1T6, the highest predicted ozone concentration i-s lower
than that observed. Thus, it appears that the rate constant used for
these reactions, 4 x 10 ^,is too fast. If a slower rate constant is used
or if these reactions are omitted in simulating EC-116, the fits for ozone
and NO, are .improved. EC-113 shows a similar effect. However, we continued
3-1-1
to use 4 x 10 ppm min as the rate constant for all peroxy-peroxy
reactions (except peroxyacyVperoxy) because of the lack of gas-phase data
4 _i _i
on these reactions. We used a faster rate, 1 x 10 ppm min , for the
peroxyacyl/peroxy reaction to improve the fit of PAN data in the propylene
system.
Another distinguishing feature of the propylene/butane experiments
is the considerable number of additional products measured by UCR. Most
of these extra products were nitrates. If one assumes that all nitrate
-------
173
formation rate constants are equal, then the amounts of nitrate products
reflect the concentrations of the various alkoxy radicals. UCR found no
nitrate for which the explicit mechanism did not contain the corresponding
alkoxy radical. In addition, the shapes of the observed and calculated
nitrate concentration-time profiles were similar this provides some
validation for the mechanism. At this time it is not clear whether the UCR
data or the mechanism or the assumption of equal formation rates causes
the differences in absolute concentrations.
Table I I 1-17 presents a comparison between the experimental and simu-
lated maximum one-hour-average ozone concentrations near the observed
ozone peak. The combined system shows the most consistent set of percent
differences, with an average difference of +5.6 percent and a standard
deviation of 7.4 percent.
TABLE 111-17. PREDICTED AND MEASURED MAXIMUM ONE-HOUR-AVERAGE OZONE
CONCENTRATIONS- -FOR- -THE PROPYLENE/BUTANE/NO SYSTEMS
Maximum One-.Hour-Average
Ozone Cb/fcgntra t1 on (gpm )
EC Run Percent
No. Measured Predicted Difference
97 0.56 0.61 8.9
99 0.55 0.59 7.3
106 0.55 0.62 12.0
113 0.342 0.36 5.3
114 0.73 0.77 5.5
115 0.58 0.64 10.3
116 0.72 0.646 -10.3
-------
174
The simulations of this set of experiments seem to justify our choice
of the ratio of the rate constants for the propylene/OH- and butane/OH-
reactions. One apparent exception is run EC-113, where the predicted butane
decay rate fits the data but the prop.ylene decay rate does not. Run EC-113
had the lowest initial NOV concentration of the set of propylene/butane exper-
A
iments. Perhaps an NO species reacting with propylene is giving rise to
X
a fortuitous fit in all runs except EC-113, and the ratio of the rate con-
stants for propylene/OH' and butane/OH- is in error. A confirmation
experiment may be necessary at some future time. We also feel that an
experiment with minimal hydrocarbon and NO , for example, 0.1 ppm NO ,
A A
0.1 ppm propylene as in EC-17, and 0.2 ppm butane as in EC-43, would be
useful and closer to atmospheric concentrations. The most puzzling aspect
of these experiments which the mechanism does not follow is the radical
concentration of the peroxy-oxyl pool in the later stages of some experi-
ments. This is-shown most severely in run EC-115, in which the decay
rates of both propylene and butane decrease markedly between two and three
hours. It is possible that this decrease was caused by a sudden dete-
rioration of the light source at UCR. (The light source exploded after
run EC-116). However, runs EC-99r*Et-106, EC-114, and EC-116 showttie
same effect, yet runs EC-97 and EC-113 do not.
D. TQLUENE/NO CHEMISTRY
A
UCR performed ten experiments on toluene/NO chemistry. Our
A
initial analysis for this system is the carbon balance. Table 111-18
shows the carbon mass recovery for the toluene/NO system. The
A
low mass recovery indicates that many species were not observed by UCK.
The chemistry of the aromatic hydrocarbons in smog is still not
known. Finlayson and Pitts (1976) and Calvert and McQuigg (1975)
reported that they were not aware of the details of any of the
important products of the oxidation of aromatic hydrocarbons, although
-------
175
tbess hav^&een detailed investigations of the rate constants of the
oxidation reactions of certain aromatic hydrocarbons.
TABLE 111-18. PERCENT CARBON MASS RECOVERY
FOR TOLUENE/NO SYSTEMS
/\
Time
(minutes) EC-77 EC-78 EC-79 EC-80 EC-81 EC-82 EC-83 EC-84 EC-85 EC-86
60
120
180
240
300
3%
5
5
5
*
11%
19
18
21
18
8%
11
8
8
7
4%
5
6
7
7
1%
5
6
6
5
2%
2
3
4
5
4%
5
6
6
7
*
0.5%
*
2
2
2%
7
7
8
7
1%
1
2
3
2
Data not avail able
Davis (1976) recently measured rate constants of reactions, involving
hydroxyl radicals with aromatic hydrocarbons. Table 111-19 presents a
list of his results. Hansen et al. (1975) also measured the rate
constants for reactions-involving hydroxyl radicals with aromatic
hydrocarbons. Both Hanson et al. and Davis agreed on a value of
the rate constant for the toluene/OH- reaction consistent with
measurements by Doyle et al. (1975).
Reaction of toluene with atomic oxygen is slow compared with
reactions involving hydroxyl radicals. Recent measurements by
Furuyama and Ebara (1975) show good agreement with Atkinson and
-------
176
Pitts' work (1974) on toluene + 0( P) reactions. They found that
the rate constant for the reaction of toluene with atomic oxygen is
approximately 120 ppm min .
TABLE 111-19. RATE CONSTANTS FOR REACTIONS INVOLVING HYDROXYL
RADI-CALS AND AROMATIC HYDROCARBONS
Hydrocarbon
Benzene
T.ol uene
o-Xylene
ntXyJene
Ethyl benzene
n-Propylbenzene
rso-Propylbeniene
Source: Davis (1976).
Rate Constant
(ppnHmin"1)
2.9 x 103
1 x 104
1.8 x 104
3.04 x 10
1,2 x 104
9.2 x 103
*^. 1.1 x 104
,4
Low Pressure
Measurements
(200 torrV
Tnere have-been several reaction pathways postulated for the reaction
of toluene with hydroxyl radicals:
CH.
CH;
OH
H20
(III-D
-------
177
(in-2)
(IH-3)
(III-4)
Reactions (III-2) and (III-4) also have ortho analogs. Finlayson and
Pitts (1976) and Davis (1976) believe the main pathway for OH- attack
is addition of OH- to the ring rather than abstraction of a hydrogen
from the methyl group. This conclusion is based on the similarity
of the toluene/OH- rate constant to those of the reactions of hydroxyl
radicals and many other alkyl benzenes. We estimated heats erf -,
enthalpy for the above four reactions using the techniques of Benson
(1968) and data taken from Domalski (1972), with the following results;
AH? = -34.4 kcal mol"1 »
AH° = -0.47 kcal mol
-1
AH° = -10.7 kcal mol"1
AH° = -46.6 kcal mol"1
-------
178
The large enthalpy change associated with Reaction (III-4) indicates that
formation of a hydroxymethylcyclohexadienyl radical may be the major path-
way. The smallest enthalpy change is for formation of cresols. The
hydroxymethylcyclohexadienyl radical reacts readily with molecular oxygen.
The peroxyhydroxymethylcyclohexacienyl radical may react with NO in three
general ways:
CH3 CH
+ NO
N0? AH = -37.8 kcal mol , (III-5)
CH.
II I + NO -* Nitrotoluene + Products »
o;
OH '
(111-6)
+ NO ->- 'Nitrocresdl + Products
(HI-7)
The products of Reactions (III-6) and (III-7) determine the enthalpy
changes for the reactions. Possible products and corresponding enthalpies
for Reaction (III-6) are:
CH.
II 'I + NO + Nitrotoluene + 20H- , AH° =38.9 kcal mol"1
°2
OH *
-------
179
CH.
II 'J + NO + Nitrotoluene + H00 + 0 , AH° = 21.7 kcal mol"1
OH
CH,
[p] + NO + Nitrotoluene + H00- + H- , AH° = 77.1 kcal mol'1
OH
CH.
+ NO -» Nitrotoluene + \\ , AH° = -12.7 kcal mol"
OH
We see that the formation .of H2024S tne most favored reaction thermo-
dynanrically. For Reaction (IIT-7) possible products are:
CH
OH
CH.
3 ' 3
+ NO -»
. OH
+ OH- + H- , AH° = 38.5 kcal mol"1
CH.
CH.
NO
H20 ,
-1
NO,
OH
OH
-------
180
The main reaction pathway from this analysis appears to be the formation
of nitrocresols. Nitrocresols have been reported by Schwartz (1974),
who also reported nitrotoluene and aerosol formation in toluene/NOv
X
systems.
Akimoto et al. (1976) investigated the products of the toluene-OH-
reaction. They found the products in an oxygen-free system were
o-cresol, a-nitrotoluene, and m-nitrotoluene. They concluded that the
production of a-nitrotoluene, which occurred only in the absence of
oxygen, suggests the presence of benzyl radicals and implies hydrogen
abstraction from the methyl group of toluene by oxygen atoms. Indeed,
in the presence of oxygen benzaldehyde was found instead of ot-nitro-
toluene. The m'trocresol and nitrotoluene may photolyze as follows:
+ hv * R
This reaction is discussed by Calvert and Pitts (1966). The aromatic
nitroso compound may decompose to form a peroxyphenyl radical. Owing to
the lack of detailed product analyses in the UCR data and the lack of
supportive experimental evidence, we have not included an explicit
toluene kinetic mechanism. We have simulated UCR's toluene runs with
the new carbon bond mechanism (see Chapter IV). We hope that in future
UCR experiments on aromatic NO systems more products will be identified.
J\
E. 1-BUTENE/NO₯ CHEMISTRY
A i
;
Near the end of the work reported in the foregoing sections, we
received data on three 1-butene/NO experiments done at the UCR evacuable
J\
smog chamber. An explicit mechanism for 1-butene systems was constructed
-------
181
by expanding the propylene mechanism. Working on 1-butene experiments
has been beneficial to our understanding of the propylene experiments
because of their analogous product distributions.
1. Oxidation of l~Butene
The reaction of 1-butene with atomic oxygen apparently produces the
following products:
CH3CH2CH=CH2 -t- C + CH3CH2CH2CHO
2°2 »
-4 CH3CH2COO- + CH302
£ pu pu pu n 4. HPH *
The total measured rate constant reported by Japar and Niki (1975) is
5.3 x 10 pprrf min~ . The n-peroxypropyl radical will react with nitric
oxide to form n-propyl radicals- as in the pT0pylene/NO mechanism. The
A
peroxypropionyl radical will react with N02 to produce peroxypropionyl
nitrate (PPN). The path to butyraldehyde is included in analogy to the
propylene system and because in two similar experiments (EC-123 and
EC-124), the observed initial rate of butyraldehyde production increased
when the initial nitrogen dioxide concentration was increased--the latter,
of course, increases the initial oxygen atom concentration. The initial
rates of 1-butene disappearance (due primarily to attack by OH-) in the
two experiments were similar, so the hydroxyl radical did not seem to
play a role in butyraldehyde production.
Reaction of hydroxyl radicals with 1-butene is assumed to lead to
the production of propionaldehyde, formaldehyde, and a hydroperoxy
radical (HOi). The mechanism for hydroxyl addition is:
-------
182
02 92
CH3CH2CH=CH2 + OH- -+ CHgCHgC
0-
CH0CH0CHCH?OH + NO -* CH-CH?CHCH9OH + N00
^ *3 £ C £_
r
,CH(
CH3CH2CHCH2OH
HCHO
1-butene can also react with ozone:
CH3CH2CH=CH2
A
0 N0
CH3CH2CHC
CH3CH2
02 0-
CH~CH/
The blradical will presumably decompose to.jfonri'tfce peroxypropionyl radical,
formaldehyde, and a hydroxyl radical. As discussed earlier, .Niki e± al.
(1976b) found that the alky! ozonide is also a stable product of the ozone-
olefin reaction:
I I
CH.CHpCH CH«
,CH0CH CH0
3 2 \ / 2
0
As in the propylene system, we have not fully investigated the reactions
of the ozonide or its effects on the kinetic mechanism.
2. Simulation Results
Table II1-20 presents the kinetic mechanism for 1-butene/NO chemistry.
This mechanism is similar to the propylene/NO kinetic mechanism except for
/\
-------
183
the chemistry of the four carbon atoms. Figure 111-168 shows the
factorial block for this system, Table III-21a shows the initial condi-
tions for the computer simulations, and Table II1-21b shows the photo-
lysis rate constants. We also performed a carbon mass recovery analysis
of the 1-butene/NO system, the results of which are presented in Table
X
II1-22. Analogous to the propylene/NO carbon mass recovery, there is
/\
only a 60 to 80 percent carbon recovery for the three 1-butene runs.
Initial computer simulations of the three UCR runs (EC-122, 123,
124) are shown in Figures III-169 through Ill-ISO.
1.0
Q.
Q_
0.5
Q.I
x!24
X122 x!23
I
I
0.1 0.5 1.0
1-butene--ppm
FIGURE III-168. FACTORIAL BLOCK FOR
1-BUTENE/NO SYSTEM
-------
1S4
TABLE 111-20. THE 1-BUTENE/NOV MECHANISM
/\
Rate Constant
; Reaction (ppnrl im'n~1)
N02 + hv -, NO + 0(3P) Experimental*
0(3P) + 02 + M » 03 + M 2.08 x 10~5 f
0(3P) + N02 * NO + 02 1.34 x 104
03 + NO » NOg + 02 25.2
O^D) * M" + 0 + M 8.6 x TO4
O^D) + H20+ 20H- 5.1 x 105
03 + OH- -*- HOg + 02 87.0
03 + H02 * OH' + 202 1.2
03 + N02 * N03 + 02 5 x ICT2
03 + hv *'0(1D) + 02 Experimental*
03 + hv-»- 0(3P) + 02 Experimental*
03 -»- wall 1.x 10"3
N03 + NO + 2N02 - l.SxtO4
N03 + N02 + N205 5.6 x id3
N205 + H20 + 2HN03 5 x 10"6
N205(+ M) * N03 + N02(+ M) . . 24.0
*
NO + NOg * H20 * 2HN02 2.2 x 10"9
-------
185
TABLE 111-20 (Continued)
Reacti on
2HN02 + NO + N02 + H20
HNO£ + hv -* OH- + NO
N02 + OH- (+ M) -*- HN03(+ M)
NO + OH- -»- HN02
H02 + NO * OH- + N02
H02 + N02 -*- HN02 + 02
H202 + hv -» 20H-
ru ru ru-ru j. n . ru ru ru run
wno^novii~Crio ^ U *"*" v/n,*Uno'-'nol*nU
202
CH-CH,CH=CH9 + 0 £ CH-OA + CH.CH9C{0)0:
6 c c. 3 e. it. f.
20,
CH9CH,CH=CH, + 04 CH-CH,CH90; + HC(0)Oi
22 2 3 2 c c f-
°9
CH3CH2CH=CH2 + OH- 4 CH3CH2CH(02)CH2OH
CH3CH2CH=CH2 + N03 -> N0£ + Products
CH-CH,CH=CH, + 07 -£ H,CO + CH,CH,C(0)0; + OH-
o t £ j C u t t
CH3CH2CH=CH2 + 03 -> CH3CH2CHO + HC[0)02 + OH-
A
CH0CH0CH=CH, + 0- * CH,CH9CHOOCH»
Rate Constant
(ppnr' min"l)
1.3 x 10"3
Experimental
9.2 x 103
9 x 103
2 x 103
20
*
Experimental
1.77 x 103
1.77 x 103
1.77 x TO3
4.8 x 104
12.0
5 x 10"3
5 x 10"3
5 x 10"3
CH3CH2CH2C(0)02 + NO
2 x ID
g + NO -4 N02 +
+ C0
2 x
^ + NO -* N02
2 x icr
-------
186
TABLE I11-20 (Continued)
Reaction
Rate Constant
HC(0)0| + NO
2 x 10
CH3CH(OpCH2OH + NO - NOg + CH3CH(0-)CH2OH
2 x 10
CH3CH2CH(02)CH2OH + NO
2 x TO
+ NO -» N02 + CH3CH20-
2 x TO
+ NO -^ N02 + CH30-
2 x 10
CH3CH2CH(0- )CH2OH
3 x 10
CH3CH2CHO
CH3CHO
0.95
0.95
CH30- * 02 + H2CO + H02
20y
H2CO + hv £ H02 + HC(0)02
H2CO + hv -» H2 + CO
CH,CHO + hv £ CH.Oi + HC(0)0,
,3 <9 C. t
202
CH,CH,CHO + hv * 6H.CH90
O fc J t
HC(0)0
20,
CH3CH2CH2CHO + hv
)2 + HC(0)02
0.95
*
Experimental
*
Experimental
*
Experimental
Experimental
Experimental *
CH3CH2CH2CHO + hv -* CH-jCHO +
H2CO + OH-
^ + H20
Experimental
1 xlO4
OH- -» CH3C(0)02*
1 x 10
-------
187
TABLE 111-20 (Continued)
Rate Constant
Reaction (ppnH
09 4
- CH_CH-C(o)o:r + H.O 1 x Mr
CH3CH2CHO + OH- -5- CH3CH2C(0)0£
0
CH3CH2CH2CHO + OH- ^ CH3CH2CH2C(0)02 + H£ 1 x 10'
yi
i*noV*nrtl*nrti/ \ U )'\Jn "* HUo "*" wn^Cno^nrtC \ 0) OOH "J" 0*^ 1 X 1 0
CH3CH2C(0)02 + H02 * CH3CH2C(0)OOH + 02 1 x 104
CH3C(0)02 + H02 » CH3C(0)OOH + 0£ 1 x TO4
HC(0}0£ + H02 * ftC(0)OOH + 02 1 x TO4
CH,CH9CH(0;)CH,OH + HOX + HO; -»- CH,CH,CH(OOH)CH9OH 4 X TO3
6 c c. c f. £. 3 c. L.
CH3CH2CH202 + HOj -» CH3CH2CH2OOH + 02 4 x TO3
CH,CHoer^+ R0^-* CH,CH,OOH +0, 4 x 103
3 t 2 Z... J f. c.
CH302- + HOg- * CH3OOH + 0£ 4 x 103
A v in3
HO^ + HOx * H,0, +0- 4 x 10
£ C. C t £
CH3CH2CH2C(0)0£ + N02 * CH3CH2CH2C(0)02N02 4 x 102
CH3CH2C(0)02 + N02 * CH3CH2C(0)02N02 4 x 102
CH3C{0)02 + N02 * CH3C(0)02N02 4 x 102
-2 *
CH3C(0)02N02 -^ N02 + CH3C(0}0^- 1.7 x 10
2*
CH3CH2C(0)02N02 * N02 + CH3CH2C(0)0| 2 x 10"
-------
188
TABLE II1-20 (Concluded)
Rate Constant
Reaction (ppra"1 min"1)
fl *
CH,CH9CH,C(0)09NO, + CH-CH,CH9C(0)0: + NO, 2.5 x 10"Z
w££ c £ O £ c C. c.
CH30- + N02 * CH3ON02 2 x TO4
CH30- + N02 -* H2CO + HN02 2.2 X 103
CH3CH20- + N02 * CH3CH2ON02 " 2.0 x 104
CH3CH20- + N02 * CH3CHO + HN02 2.2 x 103
CH3CH2CH20- + N02 -» CH3CH2CH2ON02 2 x 104
CH3CH2CH20- + N02 * CK3CH2CHO + HN02 2.2 x 103
°2 2
CO + OH* -* HO, ,+ C00 2.06 x 10*
CH3Cri2CH(02)CH2OH
* Units of min .
-2 -1
t Units of ppm min
2 , 2
3
4 x 10
-------
189
TABLE 111-21 a. INITIAL CONDITIONS FOR SIMULATIONS
OF l-BUTENE/NOv SYSTEMS
A
EC Run
Number
122
123
124
NO
0.398
0.401
0.608
N02
0.103
0.106
0.385
1-Butene
0.217
0-404
0.424
HN02
0.01
0.01
0.01
kl
0.29
0.28
0.27
TABLE III-21b. PHOTOLYSIS RATE CONSTANTS FOR SIMULATIONS
OF 1-BUTENE/NOV SYSTEMS
J\
(104 mirf1)
EC Rurt Qy+^D 0,-*03P HNO,-- H,0,- urun_^ ..urn HCHO+CO+H, CH,CHO* CH,CH,CHO-" CH,CH,CH,CHO*
nuTTujci ,5 j d. £. c. nL.nlr*Tl"'T'nLU» L. J 0 c. o £. C
122)
> 38.0 93.0 540 6.7 7.3 20 5.0 5.5 9.4
123 \
124 34.0 82.0 500 6.2 6.5 18 4.5 5.0 7.7
-------
190
TABLE 111-22. PERCENT CARBON MASS RECOVERY FOR 1-BUTENE/NOY SYSTEMS
y\
Time
(minutes)
60
120
180
240
300
360
480
540
600
EC- 122
76
70
63
63 -
61
*
*
*
*
EC-123
8CT
71
69
64
60
*
*
*
*
EC-124
85
81
82
81
78
78
75
-Zf
73"
* Data not available.
-------
T9T
- 00
I - 10
* - Ml
* XX
IX X
.;, * *,
I 0
UK * 0 O*
* V 0
K *
*0 0 X I *
9 I
OO
<*.M IM.I
TIKE (HIKVIEBt
concamurioB
M.M 4fi«.*ff
FIGURE III-169
rtrr
t » IT
rr
r
r
r x
ra rrrrrrfr
» rrrr r r r r r 7
* r IT i r F »
*» »»»
' t » r r
r r x n r r
rx.nx
r r x xo
rr B x
r r > a x
r » o B
r x « » x
r » x
r >
. » DD B
B
» B B
1 B
k.M «*.
FIGURE II1-70
-------
192
z m
PAN
A X
. AA* - . * ,r"
I A X * F f
r rrr « o o
* ^ f *F r r FT rm o o o o o o c
.M 13*. fM.M »M.W 0M.M
Tim; minrnxi
ce-iaa . imiix ALU rAi m oaKramiriiui ECUZ r»cro»
FIGURE III-171
II
T
ft
A
T
I
* ETHYL HI71UITE
B B BB B
* .IB
B I E
« X
B X <
B X
B> e
BB X
B X
«*.M IM.M
FIGURE III-172
-------
193
oo moo
aa a a
» a x xx
0
xo
00 1C
X X 22
x a
.!* a
jr o
OttK»
* i
oo
XX X
. 12*.M IO».M
M.M «M.M 4U.M
03 M M3 COUCCfcTMATlWI CCALC rACTOft !*«
FIGURE III-173
IT r rrrr mr
BT F rrr rr r rr rr
ra r rr i
r r
F FF »
F F
FJF *
F F >
F » X X
m a
to
r '
r T
* if
x IT rr
x rr F
B D *
ona *
I3>.M 101.1
recia BVTI FOHI *LM canoomuTioR KULC FAcraa !*
FIGURE III-174
-------
194
tin rrrr »
rt I
r F
r
t
r A * AA
AX A
» r A *
IT A A «
? X A
/ ' AA * *
r x AA
r A
r x AA
X P A
r AAA
r A
Z PA A
P 0000 «
O OO
« 00
« 0
_. 000
A P O 0
AA P a O
A P 00 O 0
X A IT OO OOO
AAA AA-A P PP PPP 00 000 OO 0 00 OO OO .
M.M I30.M IO*.M
U3 . OPGCIIB PAH PP> ALB3 OOHC£ITTttATMNI CCALC PACTDA !»
FIGURE III-175
« XX XBZ X X
XXI
XXX
X X
X
I
X X
zz
n
K » * *
XU U> DO » BO IU)
ED X HUH I
B I XBBH
mi B B
B I « X B B
mi > x
B X B B
B IB
B I . B B
I EE B B
B X
B t
B B I
B X
B E
X
EC X [
IW.W UB.M XM.O*
TIIE IIItVHSl
FIGURE III-176
-------
195
* XX
IK X
X
> X I
XXX
XX X X X' X
XX X X X X
X X
xx > a
X
X 1
X X
X
00°0<0000 0.0.
KKKKPII
IM.M
rccica «a
Ma coKcnrnuTioi* MALE FACTOR. !*
FIGURE III-177
T t T T r T
T
r
r r r +1 T r r
r '
T r «
r *
.r r «
r x x x>
rxx ..xxx
r r B
r r« -
r -i
FCBWIDEHYCE
M.M M*.l
Tl» IHlUVTEEl
caMjurnuTiMi SCALE r*crM ««
FIGURE III-178
-------
196
X A
* A
A
» * FWI
X A
PAX
X A
A A «
AA AX A A* *
TD-at . irecin
FIGURE III-179
, .
. »
* "
x .» x x
1
m
' rnm. KITRATE
IX*
I X
it * n
H ff *
*.»* IM.M
EO-IM . I'ECIIV Al
TIIK dU«VTTSl
I.M M*.M
KCALC FAC1W\ !*«
FIGURE III-18Q
-------
197
3. Discussion of 1-Butene/NO Systems
/\
Only three 1-butene experiments were simulated, yet the results are
very helpful to the understanding of olefin behavior in smog formation.
Modifying the propylene mechanism seems to give the overall pattern
of products. By studying the observation from runs EC-123 and EC-124
for the production of butyraldehyde, we learned that the rate of its
formation correlated with the initial N02 concentration. The propylene/OH-
reaction with a hydride shift postulated by Durbin et al. (1975) leads
to propionaldehyde formation. Similarly, one would expect that butyralde-
hyde would form from hydroxyl attack on 1-butene. But butyraldehyde
formation correlates with initial N02> implying oxygen atoms from NC^
photolysis attack 1-butene. Thus, the hydride shift reaction may
be incorrect.
A comparison between the predicted and measured maximum one-hour-
average ozone concentrations is shown in Table 111-23. The average
difference is +6.6 percent with a standard deviation of 24.8 percent.
The problems involve'^ in NOV loss and PAN chemistry (in this case
A
PPN), which occurred in all simu-lations in this report, appear to be
more severe in the 1-butene experiments. We hope to use these experi-
ments to help evaluate some changes in the mechanisms that may resolve
these problems.
-------
198
TABLE 111-23. PREDICTED AND MEASURED MAXIMUM ONE-HOUR-
AVERAGE OZONE CONCENTRATIONS FOR THE
1-BUTENE/NO SYSTEMS
Maximum One-Hour-Average
Ozone Concentration (ppm)
EC Run
No.
122
123
124
Predicted
0.208
0.43
0.283
Measured
0.216
0.48
0.21
Percent
Difference
-3.5
-11.5
34.9
-------
199
IV A NEW GENERALIZED LUMPED MECHANISM
Thus far in this document, we have discussed the development of
explicit kinetic mechanisms. In earlier works (Hecht et al., 1974a,b),
a generalized kinetic mechanism was developed for use in modeling urban
air pollution. That mechanism groups each individual hydrocarbon into a
general class; for example, propylene, 1-butene, and ethylene are classi-
fied as olefin. The use of that mechanism has highlighted certain diffi-
culties, such as:
> The average carbon number for each class of hydrocarbon
must be known; such determinations are sometimes not
available.
> If the average carbon numbers for the different general
classes are quite different, then the stoichiometric
coefficient that relates to the overall average chain
length is difficult to determine and remains -untested,,
> It is difficult to model "high concentrations of less,
reactive species and low concentrations of highly
reactive species in one -group with one ^et of reaction
rate constants.
The original Hecht, Seinfeld, and Dodge (HSD) mechanism is based on
the concept that the concentrations to be used as input data for each
general class are converted into units of ppm by volume of some repre-
sentative compound with an average carbon number. Experimental data
for mixtures of hydrocarbons are often reported in units of ppm as car-
bon (ppmC), and additional measurements that characterize individual
hydrocarbons, the number of double bonds, carbonyl bonds, or aromatic
-------
200
rings might be known. In order to utilize the HSD mechanism, one must
proceed through a series of conversions and assumptions to obtain the
relative amounts of the hydrocarbon classes (olefins, paraffins, aro-
matics). One must also estimate an average carbon number for each class.
Another difficulty with the original HSD mechanism comes from a
noninteger stoichiometric coefficient used to estimate the number of
times that ROA radicals normally react with NO to produce N0£ and RO-
radicals, which can lead to carbonyl compounds. The carbonyl compounds
in reality react to form an ROA or RCO; with fewer carbon atoms and HOx.
However, in the HSD mechanism some fraction of the original R-reappears
at the end of each of these cycles, the fraction depending on the
stoichiometric coefficient, &. This coefficient acts like an amplifier
in that any number of reaction cycles up to infinity is possible
depending on the choice of e.
With the original HSD mechanism, it is difficult to simulate a
mixture of olefins of widely different reactivities. In the actual
atmosphere, the most abundant olefin measured has been-ethylene
(Calvert, 1976). Even though ethylene is the least reactive -o-lefiIT,
its high concentration makes it an important factor Tn smog chemistry.
The other olefins are much less abundant, but much more reactive. Mix-
ing all olefins together as one lumped species makes it impossible in
principle to model the ethylene/other-olefin chemistry (Hecht et al.,
1974b).
In the sections that follow, we first discuss the specific revisions
we have made to the originaLHSD mechanism, we then describe the develop-
ment of a new generalized mechanism, and finally we present and analyze
the results of simulations using this new mechanism.
-------
201
A. REVISION OF THE HS.D MECHANISM
The original HSD mechanism, as shown in Table IV-1, has been updated
to include recent findings from our work on explicit kinetic mechanisms.
The main changes are as follows:
> Formyl radicals are treated as forming peroxy-
formyl radicals rather than breaking down to HO;, and CO.
> The HOA-NO;, reaction has been included; this reaction
may be important in the late afternoon and evening
chemistry.
> An aldehyde photolysis path leading to stable products
has been added.
> PAN decomposition has been added.
> A path to form nitroaromatics has been added.
> Decomposition of RO- radicals has been incorporated into the
mechanism.
B. THE NEW-GENERALIZED MECHANISM
Because of the association of reactions and reactivities with carbon
bonds, the range of reactions and the range of rate constants in a kinetic
mechanism can be harrowed considerably when each carbon atom is treated
according to its bond type. This concept is the basis for the new carbon-
bond mechanism, which is listed in Table IV-2. In this mechanism.
hydrocarbons are divided into four groups: single-bonded carbon atoms,
fast double bonds (i.e., relatively reactive double bonds), slow double
bonds, and carbonyl bonds. Single-bonded carbon includes not only paraffin
molecules but also the single-bonded carbon atoms of olefins, aromatics,
and aldehydes. Double bonds are treated in pairs of carbon atoms.
Therefore, the concentrations of ethylene, olefins, and aromatics are
taken as the concentrations of the double bonds in those species, which
are one-half of the ppmC concentrations after the single-bonded carbons
are subtracted. An activated aromatic, ring is considered as three double
bonds in the present formulation of the mechanism, and because of a
-------
202
TABLE IV-1. THE ORIGINAL HECHT, SEINFELD, DODGE MECHANISM
N02 + hv -* NO * 0
0 + 02 + M + 03 + M
03 + NO * N02 + 02
>The N02-NO-03 Cycle
0 + NO + M + N02 + M
0 + NO,
NO + 0.
tn ^ I1W ' U«
0 + N02 + M -» N03 + M
°3 * N02 ~" N03 * °2
N03 + NO -t 2N02
N03 + N02 -J N205
N205 i* N02 + NO,
N2°5 * H2° " 2HN03
NO + HNO
3 " HN02
13
HN02 + HN03
, Important Reactions of 0
'with Inorganic Species
.The Chemistry of NO.,,
'N«0.c. and HNO, J
c. o o
1 Reactions of HNO- with
/Inorganic Species
NO +- N02 + H20 * 2HN&2
2HN02 " NO + N02 + H20 ^ Chemistry of HN02
HN02 + hv " OH + NO
-------
'203
TABLE IV-1 (Continued)
OH + N0
17
18
HNO.
OH + NO + M * HNO- + M
OH + CO + (02) - C02 + H02
Important Reactions of
'OH with Inorganic Species
20
H02 + NO * OH + N02
hv * 20H
)Oxidation of NO by H02
>Photolysis of
22
0 * ROO
HC
23
(l-a)HO
RCOO + RO + HC
5 - '
1 T W3
»
HC, + OH * J?00 +.HC4
0 -* ROO
HC2 +OH
26
ROO +,H20
HC3 + 0 -* ROO + OH
HC3 + OH
HC4 + hv
28
ROO
(2-3)H0
HC
OH
30
Organic OxIdaThjn Reactions
H^ = Olefins
HC2 = Aromatics
HC3 = Paraffins
HC4 = Aldehydes
-------
204
TABLE VI-1 (Concluded)
ROO + NO
RCOO
+ NO
RCOO
it
0
RO
RO
RO
+(o2)
+ NO,
2
4 O,
* N02
+ NO
31
* RO + N02
32
-* ROO + N02 + C02
33
* RCOONO,
it L
3. 0
* H02 + HC4
35
* RON02
36
-> RONO
Reactions of Organic
)Free Radicals with NO,
, and 0
H02 + H02
37
38
H02 + ROO ^ RO + OH + 02
*~v
2ROO f 2RO + A,
Other Peroxy Radical
Reactions
-------
205
TABLE IV-2. THE CARBON-BOND MECHANISM
Rate Constant
Reaction (ppm" min )
N02 + hv -> NO + 0- k*
0- + 09(+ M) -» 0, (+ M) 2.08 x 10~5
03 + NO -* N02 + 02 25.2
0- + N02 -> NO + 02 1.34 x TO4
03 + N02 -> N03 + 02 5 x 10~2
N03 + NO -*- N02 + N02 1.3 x 104
NO, + NO, + H50 - 2HNO, 1.66 x 10"3t
3 £. c. o
NO + N02 + H20 -» 2HN02 2.2 x lO"9*
HN09 + hv -»- NO + OH ki,jn
c. ' niilUo
N02 + OH- * HN03 9 x 103
NO + OH- -> HN02 9 x 103
CO + OH- » C02 + HO^ 2.06 -X 102
OLE + OH- 2 HCHO + CH^ 3.8 x TO4
PAR + OH- I CH30^ -i- H£0 1.3 x 103
0* o
ARO + OH- 4 HCHO + CH30^ 8 x 10J
OLE + 0- -? HC(0}02- + CHjO^ 5.3 x TO3
PAR + 0- -S CH302 + OH- 20
20,
ARO + 0- -* HC(0)02 + CH30^ 37
OLE + 03 5 HC(0)0^ + HCHO + OH- 0.01
-------
206
TABLE IV-2 (Concluded)
Reaction
Rate Constant
(ppm" min" )
ARO + 03 * HC(0)0/> + HCHO + OH-
0.002
OLE + 0, -* ozonide
20-
HCHO + hv +* HC(0)02 +
0.005
KHCHO -> radicals
HCHO + hv -*- CO + H
HCHO + OH- $ HC(O)O£ +
NO -*- OH- + N0
k:
1 x
HCHO -*- CO
,4
2 x 10-
-f NO -v N02 + HCHO +
HC{0)0^ + NO -* N02 + C02 +
+ hv f OH- + OH-
2 x 10*
2 x 10*
HSCOOH
HC(0)02- + H02 -* HC(0)OOH
HC(0)02 + N02 -»- PAN
PAN * HC(0)02 + N02
ARO + N03 '* PRODUCTS
4 x JO0
4 x 103
1 x 104
150
0.02
50
20.
* Photolysis rate constants in units of min"1.
t Units of ppm" min" .
-------
207
similarity in reactivities, aromatfcs are lumped with the slow (ethylene)
double bonds rather than with the fast double bonds.
One of the goals in the development of the carbon-bond mechanism has
been a mechanism responsive to the range of analytical detail in field
measurements. At one end of the range the only hydrocarbon data available
might be the total ppmC of hydrocarbons at one time. At the other extreme,
the time dependence of many hydrocarbons may be known. The carbon-bond
mechanism will accept these data and any others concerning the con-
centrations of double bonds, carbonyl groups, and aromatic rings. These
data might be available through either spectral analysis or gas chromato-
graphy. Thus, the carbon-bond mechanism is directly responsive to the
most basic measurement (nonmethane hydrocarbon), yet will accept data at
any level of sophistication. The HSD mechanism requires a knowledge of
the average carbon numbers in addition to'overall concentrations. Of'course,
using the new mechanism with minimum data requires making an assumption as
to the percentages of single, double, and carbonyl bonds.
The present formulation and parameterization of the carbon-bond
mechanism has not been validated for complex hydrocarbon mixes in either
smog chambers or the actual atmosphere. We-expect that-such an effort
will be undertaken in the near future. The actual guidelines for deter-
mining the proper rate parameter averages should be part of the validation
effort. We hope that validation efforts will confirm the= fundamental
theoretical foundation of the mechanism and that the averaging pr
-------
208
a. Single-Bonded Carbon Atoms
Paraffins apparently undergo the following oxidation reactions:
PAR + OH- -v R0£ + H20 , (IV-D
R02 + NO + HO^ + ALD + N02 , (IV-2)
HO^ + NO -* OH- + N02 . (1V-3)
In this set of reactions, one hydroxyl radical reacts and is regenerated
while two nitric oxide molecules are converted to nitrogen dioxide and
th.e paraffin is converted into an aldehyde.
From Greiner's (1970) formula for calculating alkane-OH- rate con-
stants, primary carbon atoms at 300°K have a reactivity of 290 ppm" min" ;
secondary carbon atoms of normal alkanes have a reactivity of 1600 ppm" min"
and tertiary carbons have a reactivity of 4200 ppm" min" . The creation
of a side chain converts two secondary carbons into one primary carbon
and one tertiary carbon, for an .average reactivity of 2240 ppm" min .
*" "» .-**
Table IV-3 lists the reactivity per eSTlx>R atom for several common alkanes.
Many of the alkanesexcept propane-and those containing the tertiary-
butyl grouphave reactivities within about 35 percent of 1300 ppmC" min .
The compounds with tert-butyl groups may be present in Tow concentrations
in typical atmospheric mixes. They were reported by Calvert (1976) to be
less than 8 percent of the alkanes in observations made during the Los
Angeles Reactive Pollutant Program (LARPP).
b. Fast Double Bonds
Olefin chemistry is represented by the following reactions (for
ethylene):
OLE + 0 + HCO^ + CHgO^ , (IV-4)
-------
209
OLE + OH- + HCHO + CH^ , (IV-5)
OLE + 0- -> HCH0 + HCOr +--OH-. . (IV-6)
Olefins are treated with greater difficulty than paraffins. The rate
constant for OH' oxidation differs greatly between ethylene and the
longer chain olefins, especially the internal olefins. The mechanism
is further complicated because olefins react with ozone to a significant
extent.
TABLE IV-3. REACTIVITIES PER CARBON ATOM FOR SEVERAL ALKANES
Reactivity
Alkane (ppm"'min" )
Propane 730
n-Butane 950
i-Butane 1250
n-Pentane 1080
neo-PentaTre ~ '232
i-Pentane 1340
n-Hexane 1170
i-Hexane 1380
2,3-Dimethy!-butane 1590
2,2-Dimethyl-butane 460
n-Heptane N 1220
1-Heptane 1410
2,3-Dimethylpentane 1590
n-Octane 1280
i-Octane 1440
2,3,4-Trimethylpentane 1760
n-Nonane 1310
i-Nonane 1460
-------
210
Per double bond, the reaction rates of olefins with OH- jump from
8000 ppm min for ethylene to 38,000 ppm" min" for propylene (Davis,
1976) and stay constant within a factor of 2 for longer olefins. The "
reactivity increases for 1-butene (to 4.3 x 10 ppm" min" ) and for
internal olefins such as cis-2-butene (6.3 x 10 ppm" min ). Table IV-4
lists reactivities in ppm" min" for five olefins. Because of the low
variation in rate constants of olefins (except ethylene), we treat all
olefins except ethylene as one lumped species in the carbon-bond mechanism.
In the near future efforts will be made to develop methods for estimating
"averaged" rate constants for olefin reactions. Currently, we are using
rate constants for the specific olefin in the simulation runs (see
Section C-2).
TABLE IV-4. RATE CONSTANTS PER DOUBLE BOND FOR THE REACTIONS
-OF SORE OLEFINS WITH OH- AT 300°K
Rate Constant
Olefin (ppm" min )
Ethylene 7.9 x 103
* Tk
Propylene " ^-3T?T x 10^
1-Butene 4.3 x 10^
cis-2-Butene 6,3 x 104
Tetramethyl ethylene 8.4 x 104
The olefin-03 reactions are more complicated than the olefin-OH-
reactions. Table III-3 shows rate constants for the same olefin-0^
tJ
reactions. The propylene reaction with ozone is roughly seven times
faster than the ethylene reaction, and nonterminal olefins such as cis-2-
butene react even faster (terminal olefins such as isobutene have reac-
tion rate constants comparable to that of propylene). Substituted
internal olefins, such as 2,3-dimethyl-2-butene, react with ozone roughly
10 faster than does ethylene.
-------
211
Thus, we are confronted with the task of devising lumping parameters
capable of spanning a 10 difference in rate constants. This task has
not yet been performed; it seems possible only if the slower 03~olefin
reactions are neglected. In future work, we hope to incorporate the
reactions of olefins with NO;. These reactions are insignificant for
smaller olefins (ethylene, propylene, and 1-butene), but are fast for
longer chain olefins (Japar et al., 1975). At present, we do not include
this reaction in the carbon-bond mechanism because little is known about
the products of this reaction.
c. Slow Double Bonds
In the carbon-bond mechanism, the aromatic oxidation reactions pro-
~duce~the-same productsas-^the olefin oxidation reactions:
ARO + 0 -> HC03 + CHgO^ , (IV-7)
ARO + OH' + HCHO + CH^ , (IV-8)
ARO-+ 03 -* HCHO + HCO^ + OH- . (IV-9)
In these equations, ARO represents one slow double bond; thus, the toluene
molecule equals three AftOs. We have found that an aromatic ring activated
by the addition of an alkyl group, such as toluene, can be treated by the
carbon-bond mechanism as three ethylenic double bonds (olefins). Thus,
we treat ethylene and the aromatics together as one lumped species. This
lumping is used primarily because the oxidation rate of ethylene by OH-
is closer to that of aromatics than to the other olefins. In addition,
*
ethylene typically makes up 50 percent of the atmospheric olefins, as com-
pared to only 10 to 20 percent of the aromatics (on the basis of total
double bonds).
-------
212
One question we had to resolve was whether ozone-aromatic reactions
similar to ozone-O'lefin reactions should be included in the mechanism
for toluene systems. As the aromatic ring decomposes, it surely must
lead to some olefin species which, of course, would react with ozone.
We have tentatively included ozone-olefin reactions for two reasons:
> The rate constant for the reaction of ethylene with ozone
is very slow. It may be similar to the rate constant for
the ozone-aromatic reaction, which has not been measured
accurately but is known to be slow (Finlayson and Pitts,
1976).
> We wanted to maintain the treatment of aromatic -bonds and
ethylene together as one lumped species.
We hope that some ethylene/toluene smog chamber experiments can be
carried out in the near future to provide a better test for the mechanism.
The choice of rate constant for the reaction of slow double bonds
with oxygen atoms is more difficult because the rate constants for aro-
matics and ethylene reacting with oxygen atoms differ greatly. We did
not use- ttre~ rate constant for the reaction of oxygen atoms with ethylene
because that rate constant caused too high a radical concentration early
in a simulatiorrof a toluene system. We plan to develop a procedure for
estimating a rate constant in accordance with the relative amounts of
aromatics and ethylene in the system being simulated. This is not a
great difficulty because the reactions of slow double bonds with oxygen
atoms are not very important. We did include a reaction between NO;, and
aromatics to account for the internal olefins that are formed as the
aromatic rings open. The rate constant of 50 ppm min~ is adequate at
present for most of the smog chamber runs to date. It is faster than that
measured for the reaction of NO., with ethylene (1.4 ppnf «iin~ ) by Japar
and Niki (1975), but far slower than their value for the reaction with sub-
stituted internal olefins (5 x 10 pprrf min ). The inclusion of the reac-
tion of NO* with aromatics is partially justified by the observation that
no N01 Is formed in smog chamber simulations until some toluene has reacted.
-------
213
As mentioned above, the aromatic ring is represented by three double
bonds (ethylene groups) in the carbon-bond mechanism. Therefore, observed
aromatic concentrations in ppm are multiplied by three to give input
data for the mechanism. Any alkyl groups or carbonyl groups on the
aromatics are considered separately. For toluene, we use three times
the measured toluene concentration of slow double bonds and the toluene
concentration of single-bonded carbon atoms.
d. Carbonyl Bonds
The carbon-bond mechanism treats carbon-oxygen double bonds, whether
in ketones or aldehydes, as a single group. We are currently using
4 -1 -1
a value of 1 x 10 ppm min for the aldehyde-OH- reaction rate. This
3 -1 -1
compares to an estimate of 9.8 x 10 ppm min as the lower limit of the
reaction rate (Herron and Penzhorn, 1969). Our present value falls mid-
way between the value of 2.1 x 10 ppm" min" measured by Morris and Niki
Oil
(1971) and the Volman and Gorse (1972) estimate of 5.6 x 10 ppm min"
for the reaction of propionaldehyde-OH«. Data on the rate constants of
larger aldehydes are sketchy, but seem to be in accord with our choice
of rate constant.
2. Sources of Radicals in the Carbon-Bond^ Mechanism
In computer simulations of smog chamber experiments, the initial
concentration of hydroxyl radicals is produced by photolysis of nitrous
acid:
HN02 + hv -» OH- + NO . (IV-10)
The amounts of HN02 that give the best fits between predictions and
measurements are generally about one-third of the equilibrium concentration
of HN02 from the reaction
H20 + NO + N02 + 2HN02 . (IV-11)
-------
214
Is only an initial radical source because its half-life due to
photolysis is about 15 minutes.
Some radicals can also come from oxygen atom attack on hydrocarbons:
0 + PAR -v RO^ + OH- , (IV-12)
0 + OLE -> R0 + RCO . (IV-13)
This reaction is slower for paraffins, particularly for butane, than it
is for olefins (Hampson and Garvin, 1975). Both of these radical sources
are used in the carbon-bond mechanism.
The third source of radicals in the mechanism is the photolysis of
carbonyls:
RR'CO + hv -*- R0£ + RCO^ . (IV-14)
Maintaining the total concentration of radicals in the mechanism by
carbonyl photolysis is necessary because of the numerous temporary and
permanent radical sinks. For example, both nitrous .acid and hydrogen
peroxide are temporary radical sinks because they are formed rapidly
via reactions that consume radicals:
OH- + NO + -H.NQ-2 , (IV-15)
However, both species can reproduce radicals at a later time by photolysis.
The most important permanent radical sink in the UCR smog chamber experi-
ments on butane/NO systems also appears to be the most important sink
A
for oxides of nitrogen:
OH- + N02 -> HON02 . (IV-17)
-------
215
Carbonyl photolysis is most important in maintaining the radical
concentration, but it is also the most complex to treat in the mechanism.
The complexity arises- from the three reaction pathways of carbonyls
photolysis that produces radicals IReaction (IV-18)], photolysis that does
not produce radicals IReaction (IV-19)], and reaction with hydroxyl radi-
cals IReactions (IV-20) and (IV-21 )], which leads to oxidation of the
carbonyl without net production of radicals:
RR'CO + hv -» RO^ + R'CO^ , (IV-18)
RR'CO + hv + RR' + CO , (IV-19)
RCHO + OH-, * H20 + RCO^ , (IV-20)
0 0
RCR + OH- * H20 + RCRO^ . (IV-21)
The magnitude of radical production from carbonyl photolysis is a function
of the carbonyl concentration and the photolysis constant. In the explicit
mechanism study we found that the principal carbonyl compound was MEK, but
an important source of radicals was the photolysis of biacetyl (which pre-
sumably has a large photolysis constant). In the UCR data, the actual
light spectrum was uncertain, as was the total carbonyl concentration
because of the considerable mass recovery problem. Even if the explicit
mechanism and the light spectrum were known completely and the mass recov-
ery were perfect, the photolysis constants would still be uncertain,
primarily because of the uncertainties in the quantum yield data.
The relative importance of HNOp and carbonyl photolysis as sources
of radicals in the atmosphere has yet to be determined. Available data--
ambient concentrations of the species, solar spectral data, quantum yields
are insufficient to allow anything but the crudest estimates. The selection
of initial conditions and photolysis rates for the initiation of the smog
formation process in computer simulations is arbitrary.
-------
216
C. SIMULATION RESULTS
The use of this lumped mechanism to model data on a single hydro-
carbon species presents a paradoxical situation. Optimal performance for
a mixture of hydrocarbons may not be the best for certain specific hydro-
carbons taken one at a time. However, this mechanism is based on the
explicit chemistry of specific hydrocarbons (except aromatics). Thus,
the rate parameters used to optimize the present mechanism results may
not be optimal for use in the atmosphere.
The carbonyl photolysis constants used in the carbon-bond simulations
were derived from tire constants used in the corresponding explicit studies
using average values weighted according to the explicitly simulated
fraction of each carbonyl compound at the time of maximum total carbonyl
concentration. The initial values of nitrous acid were in all cases
identical to the values used in the corresponding explicit studies.
1. Butane/NO Systems
"' ""*Jn""~ J\ '"""'~
Simulations of the butane/NG syst-em with the caH>cfs-bofKl mechanism
-y\
are presented in Figures IV-1 through IV-17. As discussed above, paraffins
are the simplest species to generalize since the reactions involve only 0
atoms and OH- radicals. In the factorial design set of butane systems, the
simplest NO data are in run EC-42; these data are shown in Figure IV-8.
A
Essentially no ozone formed in this experiment; thus, complications from
species such as NO^ and N205 that form late in the smog reaction were mini-
mized. The NO crossover in EC-42 was reached just before the end of the
J\
six-hour experiment. The N0x/carbon ratio in this run was higher than in
any other run, thus exaggerating the radical-NO sink effect. Although the
J\
carbon-bond mechanism does not follow the data for EC-42 closely, it is no
worse than the explicit mechanism. If the initial concentration of nitrous
acid is raised to or above the equilibrium value of 0.03 ppm, the initial
slope of NO disappearance is too steep and then it flattens out. If the main-
tenance source of radicals from carbonyl photolysis is increased by using
-------
217
1M.M
ten* M
IM.«* 3M.M
TIME miwran
FIGURE IV-1
FIGURE IV-2
-------
218
r
p
» p p r P _
M.M tM.M 1M.«* 20*. M 1M.M
^TME tnuumn
FIGURE-IV-3
-------
219
0
owcomuTiM KILE ncn» MM
FIGURE IV-4
i a * i n » a a
* * a
, ' * " ,
«.* 4**.M
FIGURE IV-5
-------
220
r- MI
r-
-t
T
cc-«i . ram ni CMCEHTIUTIM nuu r
r
r
» P T T
FIGURE IV-6
-------
221
H.«* !.
«c-«a . mem
TIIB
COTCCimuTIM KALI TtCmk
FIGURE IV-7
2 a >
a *
» ».« IM.M
nm itimrat
OMcnruTim «ou FICTM
FIGURE IV-8
-------
222
»- M
-
» * M*
».!*»
» si
. ' '
*.M> I « X . »
*. * X * » X X
' >
* X ' " X
a » » x « x- x
»- x»
. * . . ' . ' * x
* m O
# H 0
* II 0
m 0 V
o o * jr»
«0 * R I
O O . ***](***
III**
M.M IM.M u».t» »>.*
TIIK im«umn
FIGURE IV-9
rru
.!*
M.M IM.M IU.M M*.M
TIME cninrrm
FIGURE IV-10
-------
223
0
*" * M.M !.* !». *». 3*».t» Mt.M «*.
T11B lltiSUT£B>
FIGURE IV-11
* " . * *
...*
* * *
» H.M IM.M IM.M 9M.M 9M.M IM.M «U.M
nm « ».
FIGURE IV-12
-------
224
*.H I«*.M
FIGURE IV-13
«.«
a a a
a a a a
FIGURE IV-14
-------
225
M.M IM.M IM.M »*.* . 2H.M M*.M IU.M 4M.M
TIK I minimi
FIGURE IV-15
* > a it i i
* >
1
1
i
Tint imwinn
FIGURE IV-16
-------
226
X
X
*
x
X
X
r r
-f
tf
r r ' ' '
TIH
r«
FIGURE IV-17
-------
227
o -I
a photolysis constant larger than about 1.5 x 10 min , a somewhat
better fit is obtained for EC-42. However, a poorer fit is then obtained
for EC-45 because the aldehydes present initially in this run are pre-
dicted to produce radicals at too fast a rate. Finally, if the rate of
the main radical and NO sink [Reaction (IV-17)] is lowered, the total
A
NO data cannot be followed.
A
Thus, the mechanism for oxidation of butane as presented in Table
IV-2 is not complete. For many runs, one can achieve a reasonable fit
of predictions to measurements by using initial and maintenance sources
of radicals within the bounds of uncertainty of the various reaction
rates. However, it is not yet possible to obtain a good fit to measure-
ments of systems in which the hydrocarbon/NO ratio and the rate of
A
hydrocarbon oxidation ar-@- low.
The simulations of UCR's butane/NO data (Figures IV-1 through
A
IV-17) were performed assuming paraffins behave chemically as a methane
molecule. Therefore, the input data for paraffins would be four times
the total butane concentration. All the symbols R and R' are represented
by the symbols CHj and H-, respectively. Therefore, R0£ and R'COg are
written as CH^DX and HCOX. Formaldehyde is the-representative carbonyl
in the carbon-bond mechanism. The format4eji. of -peroxyacyl nitrates is
represented as peroxyformyl nitrate in the mechanism with a rate constant
estimated to fit UCRrs PAN {peroxyacetyl nitrate) data. The single-bonded
carbon atom rate constant for the reaction with OH- was lowered to 1100
ppm min to account for butane having a particularly slow value.
2. Propylene/NO.. Systems
/\
The basic factorial design block of experiments for propylene, runs
EC-11 through EC-21, was simulated with the carbon-bond mechanism. Many
of the features of the results are the same as those obtained using the
explicit mechanism, which is discussed in Chapter III. EC-21 is similar
-------
228
the butane run (EC-42) discussed -above,. These low hydrocarbon to NO ratio
A
runs are also discussed in Chapter III. The carbon-bond mechanism results
are presented in Figures IV-18 through IV-39.
3 . Propyl ene/Butane/NO Systems
A
This series of experiments performed at UCR combined high activity and
high hydrocarbon concentrations. Therefore, the reaction of OH- with N02,
which is the main radical sink in the butane system, is not as important in
these experiments. Instead, radical-radical reactions are more important
as radical sinks:
H02
H02 + R0£ -> Products
^-Products
We found that removing these reactions from the mechanism produced sub-
stantial differences between predictions and measurements. In other
series of experiments, the removal of such reactions had little effect.
Other radical -radical reactions are:
R0£ + R0£ -» ALD + ALD + HO^ + H0£
HCOg + HCO^ + C02 + C02 + HO' .+ HO^
HCO; + RO; -* C00 -i- ALD + HO' + HO;
0 C. £. i i
The use of such reactions in a large airshed model would cause considerable
computational problems in solving the steady-state equations for radical
concentrations (Reynolds, 1976). We felt that the propylene/butane experi-
ments would be most appropriate for testing whether these last three
-------
229
tM.M SO*.** 44*.»
BO-ii . raoia «
FIGURE IV-18
*
a*
*
*
* *-
t
a
H.M IM.M IM.M ' 1M.M aM.W M«.M H«.M
FIGURE v-i 9
-------
230
T
r
T
r
r
r
T
r
r
r
r
r
F
r
M.M IM.M IM.M SM.M
TIK immnBA
FIGURE IV-20
r r r
r r
T
T
T *
r
v
r
T
T
r
rr
r
M.M IM.M IM.M >M.<*
n« iiin
n-n . matt n*
FIGURE IV-21
-------
231
a v
a «
. e
o
o
0 X
o
» I
i O I
*
M.M IM.M tU.«
IC-li . >r(CI
FIGURE IV-22
r
r
r
M.M M*.W
BO-M . frtcin
FIGURE IV-23
-------
232
F « F F F F»F
F F
»r
r r
» M.M U*.M
co-ii . snett* ra
FIGURE IV-24
-------
233
0
o
00* 0 B 0 *
co-it . arccm o»
FIGURE IV-25
- xn
I - M
!.<*>
O
O 0
I 0 *
»
' 0 .
» .0
»
n .
«»» » n _»« »»»»» »«
FIGURE IV-26
-------
234
F
r
r *
r
F
r
r
r
r
L
F
F
F
r
F
F
« F
T
F
* F« * r * r
FIGURE IY-27
F
F
F
. r
F
F F
IT
F F-FFFFF
M.M IH.M 1M.M JM.M 2W.M «M.M U*.~M tM.M
TIIB iniiiRm «.«
FIGURE IV-28
-------
235
» * o
,1» t a aa O
* **_ * a a a
»««**« °° a «
IM.M IM.M 2W.M'
TUB (HfaVTES)
FIGURE IV-29
r
r
i
r
r
r
IM.M IM.M Mft.*
. TIME (HI HUTU*
FIGURE IV-30
-------
236
*.«»
no ra
FIGURE IV-31
-------
237
M.M IM.M IM.M
Tin iftimnm.
FIGURE IV-32
*
m
a
11
un 11 I ra > *
M.M IM.M IM.M SM.M
. TUB minm
FIGURE IV-33
-------
238
r
r .
r m
r r
IM.M IH.M
_
TUB munmn
ncin
-------
239
0
.M 4M.M
FIGURE IV-36
I - «0
a - M*
-*».
* *
a
* a »
a a*
1
M.M IM.M IM.M * aM.M
TIIB (HIjura)
ai . mcita M MI
FIGURE IV-37
-------
240
r-m*
T
r
r
r
* * T
T
M.M
FIGURE IV-38
r r »
r »
F
TI« innnn
FIGURE IV-39
-------
241
reactions must be included in the kinetic mechanism. We found that remov-
ing them had no noticeable effect on predictions. They are not included
in the carbon-bond mechanism. Figures IV-40 through IV-60 show the pre-
dictions and measurements for the propylene/butane series of experiments.
One disturbing aspect of the carbon-bond mechanism simulations for
this set of experiments was that we had to lower the carbonyl photolysis
-4 -1
constant to 3 x 10 min for radical production to produce the simula-
tions shown. The calculated value from the explicit mechanisms was
-4 -1
7.2 x 10 min , yet the use of this value produced too many maintenance
radicals in the simulations. During the coming year, we hope to elucidate
the reasons for this discrepancy.
4. 1-Butene/NO., Systems
""r~r""~1 r^"" " A
As discussed in Chapter III, three experiments using 1-butene performed
at UCR gave us an opportunity to check the olefin chemistry by extending
the propylene chemistry to 1-butene. For the results shown in Figures
IV-61 through IV-69 using the carbon-bond mechanism, we "changed only the
4 -1
OH reaction with double bonds to the value for 1-butene of 4.8 x 10 ppm
min"1.
5. Toluene/NO Systems
~ /\ ~~
Since there is no explicit mechanism to account for the complete oxi-
dation of toluene carbon atoms, we did not have the explicit mechanism as
a basis for photolysis rates or initial nitrous acid values. An overall
carbonyl photolysis constant of 2 x 10" min" was chosen to be consistent
with the low N02 photolysis of 0.16 min" reported by UCR and the deteri-
oration of the solar simulator that we assumed for the EC-50 to EC-60
series. For simplicity we did not include a carbonyl photolysis to stable
products. The results are presented in Figures IV-70 through IV-92.
-------
242
- m
m - 1*9
3 - am
o z a
o x
x *
' ' .
X 91
x
OC-*T . ancle* a* n na
FIGURE IV-40
r
r
* r
r
r
FIGURE IV-41
-------
243
T - ra
.24*
r t T T T
r T T r
' r
T
r
rr
» T *r rr
t W.M
FIGURE IV-42
-------
244
e* »
*
x
xx x , x
-01
* X X X t *
1 1
> IS
I * OX X *
, «
..». .".. ." *
' .
X *
X Xtt
X XI*
OW
« »
MO* to B» imt4ni * »» »
».«!' >
I0*.o» l«.0» 3M.M _ 9B*.0* IM.OO 3M.M
FIGURE IV-43
* r
r
r
r
r
» r
r
F4
r
TT
r
FIGURE IV-44
-------
245
» m r r r
r , ' ' r , r r
T r
p
» f
* r r
r r
T
r r
n
rr
tr r
JCIBi M»
FIGURE IV-45
-------
246
t *
o
M
00
o *
1 1 a n'
* m
* I
' t I 0
." *',.'. '
ft a *
» xi * *
HI XI
ii x a
* o . x
r* * * x * M
XII 0 II
II ft X 1 I
« I « It
* * .
XI
* * X 1
X I *
X
I
M.M IM.M !.
BO-IM . meia MOM*
FIGURE IV-46
P
v
P
P» P» P
,-...
FIGURE IV-47
-------
247
r r r
IT T
IT
r r
r
r
T
T
r
T T
M*.«* 4M.H 4M.M
one tea PJM
F I SORE IV-48
-------
248
. .
i ."
e o
' . "
r
*
r
r
e r
r
r
« r
r
v .«
.
' ,
f
> r
r
v
« »
r
* .v ,
* T r
« » T
IM.M IM.M
FIGURE IV-49
- m
1 - Ml
.IK
HI
1
1 1
s
I
* '
* * t
a
i i
i
*
a
a
a
a
I
a
«
f
** « »
If «««»
m nra n IB n * » i »» » « »
FIGURE IV-50
-------
249
T
r
T
IT P
i-in . tncu* nm
FIGURE IV-51
-------
250
I
+ n
XI «
X **
I «
I* «
V
« *
01*
« » I
i n i IBI»» » IMP » »
!».« lit.I
FIGURE IV-52
n
r
r
r
ra
r
« r
r
r r
r
» trm
T rr* r*r r* t* r i
M.M IM.M IM.M >M.M
I.M 4**.t*
FIGURE IV-53
-------
251
n t T r r r
T r T T T
r « r
r
r
r
r
r *
r
T
r
i m mr
I-IM . men* n*
FIGURE IV-54
-------
252
i a x x x a
» xt x at
».D» a x o a
* e aa«
xa x a
at « aa
. « x a
a- « a
xa « x a
a o * a
a B « « .x a
x ii « a
,* a
§
> x i
*« *
i j,
X
. «« «
tini_>»»gHi » n»in»» **m» ***«****
1M.M IM.M M*.M «M.»* ttW.M
FIGURE IV-55
r
r
r
IT *T*T * rr *t
FIGURE IV-56
-------
253
T T T T -f
T T T T
T
m
p
m
rr
FIGURE IV-57
-------
254
p
p«
p
t
T * »
'p
» 00
r a
r «
o
f *
o r
a p
r »
p
o r
r
r
p *
p p
p
p
p
» p
p p p
* P»P*P»P* * *p p»
M.M US.M 1M.M tS.M m.M S1I.M «W.M
FIGURE IV-58
> a
a i
t *
*
t
n » * i * *
t * t
I t
I
»re»»»iiii»ii»«i»« »« «»« » «» . . « «
M.M !«: in.M Mt.M *».< MM.M IM.M ' «t*.M
TIIB uumm
KIM . *nam m M
FIGURE IV-59
-------
255
r
r.
T
T
T
r
r
T
r
TU
FIGURE IV-60
-------
256
X X
X
X
X
X
n « oo I
« 0 X
ax
a
x>
a
a
a «
MM O«OOO«
FIGURE IV-61
a
a
11
»a»
la
FIGURE IV-62
-------
257
ram. am or pw wo n*
mtaama n oat
I.M *M.*»
FIGURE IV-63
-------
258
a a a
a a
a
a a »
> a aa o «
XX aa
* x xx a
a a
a X a
a x «e a
x x o » a
a o a
i. a » a
x ' a
a x x a
i ex
i « a
a x a
ax 0 »* x a a
.' x 'aaa
a a * oo B x aaa
BB a a
a o » x a a
* B x a
i * B» x a a
x
l» X
1» X
BB» XX
an** xxx
a »» «** *.* * ** * » x
M.M lat.M IM.H M«.M ».« at*.**
!** fwtBnwvjvn
FIGURE IV-64
B - »LC
«.4*»
a
a
a
a '
a
i
* .
*
a «a
i ^ a u BUM a * a a
FIGURE IV-65
-------
259
' - TUTU suiar MB MD PM
MUSUM9 M UCa
ooooo « « a o
t
i
*
FIGURE IV-66
-------
260
1-1 1
SI 1
s
* *
I
I - 1
» x a
* a
I XXX * t
S X X X X X X 2*
XX X X X X
S . X « «
* X X »
,« *
% * .
"x,
- ' * x r
* V
* x
X X .
» V
X * ft ft I
000.
* o * *
+ Oft *
* O 00*) 0 *
* ft ' "
» o
V » ' *
*
* * «-»
ft * » »
» * *'
o » »
ft » * »
ft '
<»»»» »«»»
*.« IM.W Mft.ft* 8».t« 4M.ftft * 4ft».(
TIME '(ttt'inFBB'
EO-IM . BFtXII* «~ M PM
FIGURE IV-67
> - "IX
.**>
i * <
!*. M«.N IM.W
Tt*c
-------
261
,.,». - rant, am or wm ««
mauiMi n ucm
T
t
W
T
rrrrr r rr
, *?"'
FIGURE IV-69
-------
262
I OO M O
I 000
M.M IOO.OO 13*. *
Tint i«»m>
>.00 4M.M
FIGURE IV-70
11
* *
11
2* t »
O M.Ot IM.M 1M.I
BO-TT . craeiB
.O* *0*.l
FIGURE IV-71
-------
263
t
rc-rr . craem MM
FIGURE IV-72
-------
264
*
9
OO
M.M M».M 1».M *M.M tM.M
TIB
ton . mem n
FIGURE IV-73
ff - BO
1 - KO
a. ie>
t
i
t
*
» « » » mom* tug
1M.M *.«
Tim
FIGURE IV-74
-------
ta-n . trtcia ra
FIGURE IV-75
-------
266
O - 09
* - no
2 - Htt
9 0 0 0 OO
* 000* W
» e OB
oo eooooo
o OOBOO
00 0000
O 000
i e
1M.OO- 2W.OO
TW int
FIGURE IV-76
o - ria
'"**
t
T
loo.oo iw.o* MO.OO MO.OO
FIGURE IV-77
-------
267
« a o
O 3 *
O
s
0
o
o «
0 X
o
* o
« o
« 9
I
*
O <
a
21
x < a
a
x
*
X X X S
TlflK imcincn
FIGURE IV-78
. «rtrre»
FIGURE IV-79
-------
268
- 0*
- IW
a - no!
1 I
X *
^
x n
9
X I
Mt.0« 9M.M
FIGURE IV-80
FIGURE IV-81
-------
269
o - oo
- no
s - NO
a
a
* 2 x x x
v a x
n ax
" 2 X
» a x a
2 *
a
IX II
.*
X
000 000 00*0
2 ooo oo x o a o
00 0
oo a
.
/ o a
a x
« 22
2 X
0 22
v o * * a a
no * a a
» * a a a
o * * a a t
01* *
r * * > mn mm* V *********
CO.M 1M.M ISO.** 20».M XM.O* 3O*.*O 3SO.M 4M.M
«a . BPCCIO 09 » MB
FIGURE IV-82
*
TIW fiunmn
FIGURE IV-83
-------
270
ok
000 0
o oa o o a
00 ' . . .
.* >W.M M*.M «*.** <**.»
FIGURE IV-84
*
! t
*.eo» t
3 I
* » » <
* » t
I » i
»
1
*
I
II
t
IM.M IM.M M*.M «M.»» <*.* «M.M 4M.«*
FIGURE IV-85
-------
1 271
FIGURE IV-86
-------
272
- OB
- m
2 - m
.»«
2
X 0
XI
2 10
2 X O
* H 22 02
X 0
as o x2
IT O
IB X
X * US 3
2 0
I O 2 X *
II 2
II 0 X
21 2
x 2 a
t X
3 « 3 X
3 » *
.,
.
* *9 * 21
3 X
B » XXX
> a 3 s x
3
t* . 33 32
0 X « 33 3
MX
« « 33 * » »
o-ooo-oow o in ni n ran n » » » «
FIGURE IV-87
«
e
e
e
o
* t
e
ncia nut
FIGURE IV-88
-------
273
i a
X X
1 O 00 O 00 DO 000 00 00
* IT
X -
1
a IT
02 X
1
2 X
M.M !«.» IM.I
tncaa m m ma
FIGURE IV-89
.* **.*
FIGURE IV-90
-------
274
a x x a
x a
z x oooo»e oo> o
a o o oo
ax M
> « a
o
1 0 9 X
a
x * o
* * O -9 X
x a w e a
t . 2 x
i a a
a o « x
x a i a
a * a
a o ax
> I >
o ax
aa
» za x
aa x
no a a
» » an x
aa x x
a a a * x
* a a
* at*
o n n m w» x* * * * * * * * * .
so.** IM.*« IM.M a
TIIK
. BTECIES us M aoa
FIGURE IV-91
o
*
too.oo a».»»
T1BB ««!»
FIGURE IV-92
-------
275
The run EC-77, shown in Figures IV-70 and IV-71, had the lowest hydro-
carbon to N0x ratio, and'the inadequate*maintenance radical problem shown
in the N0x curves is consistent with the discussion in Chapter III on Runs
EC-21 and EC-42. However, the maintenance radicals in many other runs
were numerous and tended to accelerate in production. The runs with
high toluene concentrations, EC-82 and EC-83, showed this effect most
severely. For Run EC-83 the mechanism was grossly inadequate, but the
initial hydrocarbon concentration for this run was nearly 40 ppmC and the
humidity was near zero. Such extreme hydrocarbon concentrations at low
humidity will require further study to elucidate these effects of hydro-
carbon and water.
Comparison of the simulations of EC Runs 80 and 81 with 84 and 85
reveals a curious finding. These two pairs of runs have similar toluene/
NO concentrations, but the latter pair has somewhat higher NO to N0?
/\ *->
initial values. The mechanism can follow the ma-intenance radicals re-
quired in the former pair, but too many radicals are simulated in the
latter. Perhaps the light source deteriorated between these two and
the carbonyl photolysis should be lower for the latter pair. In the
other hydrocarbon systems, we could check the lumped mechanism agaiJist
the corresponding explicit mechanism. Unfortunately, we do not yet
have an explicit toluene mechanism to help develop and validate the
carbon-bond mechanism for aromatics. Table IV-5 lists the initial
conditions for the toluene set..of experiments.
-------
276
TABLE IV-5. INITIAL CONDITIONS FOR SIMULATIONS
OF THE TOLUENE/NOV SYSTEMS
J\
(ppm)
EC Run
Number
77
78
79
80
81
82
83
84
85
86
NO
0.518
0.069
0.08
0.401
0.408
0.679
1.334
0.388
0.431
0.407
N02
0.059
0.032
0.02
0.095
0.094
0.337
0.674
0.081
0.093
"8,080
Toluene
0.276
0.228
0.976
1. 02
1.96
1.88
5.63
0.968
1.92
1;-092
HN02
O.Ol
0.005
O.OOl
0,006
0.003
O.OOl
0
0.001
0.001
o-
-------
277
D. DISCUSSION OF THE CARBON-BOND MECHANISM
Summarizing the results of this chapter, a generalized kinetic
mechanism based on the explicit mechanisms discussed in Chapter III
has been proposed for future use in airshed models. This generalized
kinetic mechanism separates carbon atoms into groups depending on
their bonding. The mechanism is validated for the simulation of
smog chamber data by explicit mechanisms. Intermediate and overall
species behavior are consistent with predictions made by the explicit
mechanisms. Table IV-6 presents a comparison of the observed and
simulated one-hour-average ozone concentrations for the different sets
of UCR data. Even though the sets of experiments are not identical,
the individual results and the calculated standard deviations are
similar for the explicit and generalized mechanisms.
The carbon-bond mechanism is still in the developmental stage and
certain areas of the mechanism must be investigated further. One of
these areas is the handling of peroxyacyl radicals with more than one
carban. These radicals generate shorter chain alkylperoxy radicals
rather than HOV radicals."The carbon-tond mechantsm as "presented in
Table IV-2 only accounts for alley! radical production fran single-
* ,-,
bonded carbon atoms. Another important requirement is a set of guide-
lines for determining the .best rate constants to use forji mixture of
-hydrocarbons with substantially differenrtr-T'Sact'hnties. The UCR data
available in the contract period included only £ne, set of experiments
with a mixture of initial hydrocarbons, the propylene/butane set. In
the near future smog chamber data sets using mixtures of olefins and
mixtures of paraffins will be used to establish the needed averaging
guidelines.
-------
278
TABLE IV-6. ONE-HOUR-AVERAGE OZONE CONCENTRATIONS MEASURED
AND SIMULATED WITH THE CARBON-BOND MECHANISM
Maximum One-Hour-Average
EC Run
No.
11
13
16
17
18
21
39
41
42
43
44
45
48
97
99
106
113
114
115
116
Measured
0.23'
0.37
0.50
0.14
0.18
0.006
0.056
0.21
0.0042
0.113
0.011
0.12
0.14
0.56
0.55
0.55
0.342
0.73
0.58
0.72
Simulated Difference Comments
(a) Propylene/NO Systems
A
0.26 13
0.38 2.7
0 43 ,. Average = 2.32%
Standard deviation = 11.2%
0.14 1.2
0.21 16
0.006 -5
(b) Butarve/N0v Systems
A
0.041 -26.8
0.24 14.3
0.005 19.1
0.063 -44 Jr^V0'28*
Standard deviation = 2%
0.027 145
0.12 0
0.19 35.7
(c) Propylene/Butane/NO Systems
A
0.57 1.8
0.58 5.5
0.61 10.9
0 got _4 7 Average = 0.834%
Standard deviation = 10.1%
0.75 2.74
0.63 8.6
0.58 -19
-------
279
TABLE IV-6. (Concluded)
EC Run
No.
122
123
124
Maximum One-Hour-Average
Ozone Concentration (ppm)
Measured
0.216
0.48
0.21
Simulated
(d) 1-Buti
0.16
0.47
0.27
Percent
Difference
)v Systems
A
-25.9
-2.1
28.6
Comments
Average =0.2%
Standard deviation = 27.3%
-------
280
V GENERAL DISCUSSION
In the previous chapters we described the details of the explicit
and condensed kinetic mechanisms and presented simulations of some smog
chamber experiments at the University of California at Riverside. In
this chapter we present a general discussion of the following items:
> The perception of smog chemistry on which the mechanisms
are based.
> Some important assumptions "about experimental conditions
common to all simulations.
> How well the simulations with the explicit and con-
densed mechanisms reproduce the measurements.
> A summary of recommendations for future experimental
studies.
The overall perception of smog format ion that is lire basis of our
present mechanisms can be described in terms of t-wo cyclic merles of
reactions. One series is the oxygen atom series; it is associated with
the photolysis of N02 and the production of ozone. The other is the
peroxy-oxyl series. In this series of reactions hydrocarbons are oxygenated
and NO is converted to N02>
In the oxygen atom series N02 photolyzes to give NO and an oxygen
atom, which immediately combines with an oxygen molecule to form 0-. Oo
completes the cyclic series by reacting with NO to regenerate N02:
N02 + hv -»- NO + 0 , (V-l)
0 4- 02 03 , (V-2)
-------
281
03 + NO * N02 + 02 . (V-3)
The cycle time is inversely proportional to the light flux and the NO
concentration ([NO] + [N02]). This cycle time is typically short enough
that the ozone concentration is approximated by a steady-state relationship:
i
' {V'4)
Since the ozone concentration is nearly proportional to the ratio of
[N02] to [NO], an increase in the ozone concentration can be related to
the conversion of NO to N02. An important feature of the oxygen atom
series of reactions is that by itself it does not change the ratio [NOp]
to [NO]. Thus an independent mechanism that changes the ratio of [NO/,]
to [NO] will indirectly control the ozone concentration.
The second cyclic series, the peroxy-oxyl series, is associated with
a pool of chemical species having unpaired electrons. In this series NO
is converted to N02 and hydrocarbons are oxidized. The unpaired electrons
are passed frora.peroxy radical s'tb oxyl radicals and vice versa. Normally
the start of this series is the "rate-control! ing reaction of hydrbxyl
radicals and hydrocarbons:
OH- + tiC * R- . (V-5)
The unpaired electron appears on the resulting R', which rapidly picks up
an oxygen molecule to form a peroxy radical, RO^:
R. + 02 5 R02' . (V-6)
Typically, the next reaction in the series converts NO to N02 and transfers
the electron to an oxyl radical, RO-:
RO^ + NO -> RO. + N02 . (V-7)
-------
282
Then a hydrogen abstraction by molecular oxygen passes the unpaired
electron to a hydroperoxyl radical, HO^. The rest of the RO- radical
typically forms a carbonyl compound, OHC.
RO- + 02 -» OHC + HO^ , (V-8)
Finally, the unpaired electron is returned to a hydroxyl radical by a
second NO to N0£ conversion:
H0£ + NO t- OH- + N02 . (V-9)
Although this description is very simplified, these two cyclic series
contain the essential features of ozone formation. The two series are
partially connected by the reactions of oxygen atoms with hydrocarbons,
the reactions of ozone with olefins, and a few others. This smog mecfianism
depends on the formation and maintenance of a pool of unpaired electrons.
A portion of this pool is formed and maintained by the reactions of
oxygen atoms and ozone that link the two cyclic series:
0 + HC + R. + OH- , (V-10)
0
0 + HC -> R. + R. , " (V-ll)
03 + OLE -v R. + OH- . (V-12)
In computer simulations, however, these reactions alone do not form enough
radicals to reproduce the measurements. The initial rates of hydrocarbon
disappearance and conversion of NO to NOo observed in smog chambers
indicate that the pool of unpaired electrons involved in the peroxy-oxyl
series is formed very rapidly.
We resolved this difficulty by making an assumption that is common
to all simulations discussed in this reportthat some nitrous acid is
present at the start of each smog chamber experiment. In the simulations,
-------
283
the nitrous acid is rapidly depleted by photolysis, which produces NO
and hydroxyl radicals. The concentration of nitrous acid required to
simulate UCR experiments is generally about one-third of the equilibrium
concentration that would form eventually from the initial concentrations
of NO, N02, and H20. Nitrous acid is known to form during the loading of
large vessels (Chan et al., 1976), but whether it is present in the amounts
we have assumed is unknown. The amounts of nitrous acid required to
simulate the propylene/NOY, butane/NO. l-butene/NOv, and toluene/NO
A X X X
systems are consistent.
Another important assumption in all simulations involves the spectral
distribution of the light source. The primary maintenance source of
radicals for the peroxy-oxyl series, the photolysis of carbonyls, is very
sensitive to the intensity of short wavelength UV. The nitrogen dioxide
photolysis rate reported by UCR depends on the intensity of long wave-
length UV. This rate was formerly used to adjust the rates of all
photolysis reactions to account for variations in light intensity. However,
the light source in the UCR chamber is known to deteriorate, and the deter-
ioration is most rapid at short wavelengths. Thus, we chose a spectral
distribution for each simulation that "was consistent witli tHe'lfge of the
light source used in the corresponding experiment. The r-angfe -Qf spectral
distributions used for all simulations was chosen to gtve the J?est overall
fit for all the simulations and still be consistent with th^ range of
spectral distribution measurements.
The objective of these assumptions, and of the entire contract, is
to produce simulations based on scientific knowledge that also fit the
measurements. How well this objective was achieved can best be seen by
examining the many figures in this report that show UCR measurements and
our simulations. The overall fit for most species is good. A notable
exception is NO . In most simulations the fit between predicted and
J\
measured NO concentrations is poor, even though the conversion of NO to
N00 is correctly simulated. The poor fit is usually seen as an
-------
284
overprediction of the N09 concentration. It could be caused by NOV
C~ A
losses in the UCR smog chamber, the absence or incorrect treatment of
some important aspect of the chemistry in the mechanisms, or both.
Nitrogen compounds measured at UCR include N02> NO, PAN, and alkyl
nitrates. In every UCR experiment the sum of the concentrations of these
compounds, after correction for dilution, decreased during the experiment.
A similar but smaller decrease occurs in the corresponding simulation,
primarily because of the formation of nitric acid. Nitric acid is not
routinely measured at UCR, but has been reported in smog chambers in
amounts consistent with our simulations by Spicer and Miller (1976).
Modifications to the mechanisms designed to improve the fit to the UCR
nitrogen data by forming more nitric acid have been unsatisfactory; they
quantitatively remove radicals from the peroxy-oxyl pool and thus degrade
the overall fit.
Some indication of the validity of,the mechanisms presented in this
report for ozone prediction is provided by Figures V-l, V-2, and V-3.
Figures V-l and V-2 show the absolute and percentage differences between
predicted and measured maximum one-hour-average ozone concentrations for
the-explicit and carbon-bond mechanisms, respectively. All UCR systems
except toluene/NO are included. These figures show that the average
absolute difference and the average percentage difference are small and
positive. The indicated positive bias is not statistically significant
for two reasons: the number of experiments is small, and the standard
deviation is much larger than average difference.
As discussed in Chapter IV, the carbon-bond mechanism was designed
as a condensation of the explicit mechanisms. Thus its validity is
determined not by how well it reproduces smog chamber data but by how
well it reproduces the simulations with the explicit mechanisms. Figure
V-3 shows the absolute and percentage differences between the maximum
one-hour-average ozone concentrations predicted by the carbon-bond mecha-
nism and those predicted by the explicit mechanisms. Again, toluene/NO
r\
systems are not included. The small differences and standard deviations
in Figure V-3 demonstrate that the carbon-bond mechanism does retain
-------
285
I V
9
8
7
tn
fti
Q) £
i- **
u
° 5
o
L
Z
3
2
1
0
-
_
-
-
-
, ,[~T,
AVERAGE DIFFERENCE - +4.7*
^M
STANDARD DEVIATION = 20%
1 1 1 1
I 1 I 1
-50 -40 -30 -20
-10 0 10 20
Difference (percent)
30
40
50
60
FIGURE V-la. PERCENTAGE DIFFERENCES BETWEEN PREDICTIONS OF
EXPLICIT MECHANISMS AND MEASUREMENTS OF MAXIMUM
ONE-HOUR-AVERAGE OZONE CONCENTRATIONS
10
9
8
o
o
14
JL
DIFFERENCE » +0.015 ppm
STANDARD DEVIATION = 0.053 ppm
J_
-0.10 -0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.10
Difference (ppra)
FIGURE V-lb. DIFFERENCES BETWEEN PREDICTIONS OF EXPLICIT
MECHANISMS AND MEASUREMENTS OF MAXIMUM ONE-HOUR-
AVERAGE OZONE CONCENTRATIONS
-------
286
IU
9
8
7
/
O
g 6
fc
3
U
o 5
<*-
o
1 4
E
i
3
2
1
n
-
-
_
,,
-
-
" ,n. ,
-60 -50 -40 -30
i
-20
(AVERAGE DIFFERENCE « +0.85S
I
-^
A
|l
» STANDARD DFVIATTON
^^ W 1 Wv\vr\t\tJ L/W f JLrl 1 JhWll
= 271
1 1 l
-10 0 10 20 30 40 SI
Difference (percent)
FIGURE V.-2au PERCENTAGE DIFFERENCES BETWEEN CARBON-BOND MECHANISM
PREDICTIONS AND MEASUREMENTS OF MAXIMUM ONE-HOUR-
AVERAGE OZONE CONCENTRATIONS
10
a
8
| 4
n,
i
i
.AVERAGE DIFFERENCE
I 0.0011 ppm
STANDARD DEVIATION
0.047 ppm
-0.14 -0.12 -0.10 -0.08 -0.06 -0.04 -0.02
Difference (ppm)
0.02 0.04 0.06 0.08
FIGURE V-2b. DIFFERENCES BETWEEN PREDICTIONS OF CARBON-BOND
MECHANISM AND MEASUREMENTS OF MAXIMUM ONE-HOUR
AVERAGE OZONE CONCENTRATIONS
-------
287
-60 -50 -40
-20 -10 0 10
Difference (percent)
20
30
40
50
FIGURE V-3ffv PERCENT DIFFERENCES BETWEEN THE MAXIMUM ONE-HOUR-
AVERAGE OZONE CONCENTRATIONS PREDICTED BY THE
CARBON-BOND MECHANISM AND BY THE EXPLICIT MECHANISMS
20
18
16
14
tft
-------
288
the significant features of the explicit mechanisms even though it
requires only lt)-20 percent as much computing time.
Throughout this report we have made suggestions regarding future
smog chamber experiments. These suggestions are summarized below:
> Light intensity measurements should be more detailed.
> Experiments with slowly reacting hydrocarbons (e.g.
butane) should be carried out for more than the usual
six-hour period.
> More experiments using hydrocarbon mixtures should be
performed to aid in the validation of generalized
mechanisms.
> More detailed product measurements are needed for the
paraffin/NO and aromatic/NO systems so that a mass
/\ x\
balance for carbon can be approached.
> More consistent carbonyl data are needed in the
propylene/NO experiments. UCR runs EC-13, 51, 52, 57,
J\
95, 96, and 121 are propylene/NO experiments with the
A
same initial conditions, but t^he formaldehyde and
acetaldehyde concentrations and their ratios in these
runs varied widely.
> More detailed measurements of rritrogen compounds are
\
needed in all experiments.
> Experiments on aldehyde/N.Ov systems are needed.
A>
As discussed in Chapter III and above, the light source in the UCR
chamber is known to deteriorate, resulting in lower photolysis rate
constants. The light intensity changes mainly between 300 and 340 nm,
which is the part of the spectrum that photolyzes aldehydes. Therefore
detailed intensity measurements in this part of the spectrum are very
important.
As discussed in Chapter III, the ozone concentration did not reach
a peak in any of the butane runs. There is also a large carbon atom loss
-------
289
in these runs. Longer experiments with such hydrocarbons are required
to elucidate the chemistry. As longer chain hydrocarbons (with more than
four carbon atoms in the skeleton) are investigated, more detailed obser-
vations of products are desirable. Product measurements in aromatic/NOv
A
systems at UCR are extremely sparse; in addition, the percent carbon
recovery is too low to validate an explicit mechanism. More of the carbon
atoms in the initial hydrocarbons must be accounted for in order to
improve explicit kinetic mechanisms for aromatics. Also, a consistent
set of data is required for validation of kinetic mechanisms. In the
propylene/NO system, the early runs (EC-5 to EC-21) show equal maximum
A
concentrations of formaldehyde and acetaldehyde, but later runs (EC-95,
96, and 121) show much more formaldehyde than acetaldehyde. Inconsisten-
cies of this nature must be resolved if correct reaction pathways are to
be established. Experiments should be performed with simple mixtures of
similar hydrocarbons. Such experiments, even without detailed product
analyses, will assist in the development of the condensed mechanism because
average rate constants must be derived to account for mixtures of similar
hydrocarbons.
The chemistry of inorganics, aldehydes, arnd PAN is common to all
smog mechanisms. It is difficult to evaluate this chemistry in current
mechanisms because all recent smog chamber experiments are complicated
by the numerous chemical species resulting from the oxidation of hydrocar-
bons. The uncertainties in simulating this chemistry may be reduced most
easily by modeling smog chamber experiments on very simple systems.
Experiments with formaldehyde/NOx systems would be useful in verifying
the basic inorganic and formaldehyde chemistry. Such experiments would
not involve hydrocarbon or PAN chemistry. Experiments with acetaldehyde/
NO would add only PAN chemistry to this basic set. The round of mechanism
development just completed proceeded from parameterized mechanisms to
explicit mechanisms. The next round of mechanism development will involve
reducing uncertainties in individual chemical reactions. We recommend
that experiments be performed on very simple systems in the most advanced
chambers so that some of the basic chemical reactions in smog may be
studied with as few complications as possible.
-------
290
REFERENCES
Akimoto, H., et al. (1976), "Photo-Oxidation of Toluene-N02-02-N2
System in Gas Phase," 12th Informal Conference on Photochemistry,
National Bureau of Standards, Extended Abstracts, pp. K8-1 to K8-4.
Anastasi, C., P. P. Bemand, and I.W.M. Smith (1976a), "Rate Constant
for: OH + N02 (+ N2) -> HNO? (+ N2) Between 220 and 358K," Chem.
Phys. Letters. Vol. 37, p. 370.
Anastasi, C., I.W.M. Smith, and R. Zellner (1976b), "Kinetics of the
Reaction OH + N02 (+ M) » HN03 (+ M) over a Wide Range of Temper-
ature and Pressure," 12th Informal Conference on Photochemistry,
National Bureau of Standards,. Extended Abstracts, pp. P7-1 to P7-4.
Atkinson, R., D. A. Hansen, and J. N. Pitts, Jr. (1975), "Rate Constants
for the Reaction of the OH Radical with H2 and NO (M = Ar and N2),"
J. Chem. Phys.. Vol. 62, pp. 3284-3288.
Atkinson, R., and J. N. Pitts, Jr. (1975), "Rate Constants for the
Reaction of OH Radical with Propylene and the Butanes over the
-Temperature Range 297-425K," J. Chem. Phys., Vol. 63, p. 3591.
(1974), J. Phys. Chem.. Vol. 78, p. 1780.
Baker, G., and R. Shaw (1965), "Reactions of Methoxyl, Ethoxyl, and
t-Butoxyl Radicals with Nitric Oxide and with Nitrogen Dioxide,"
J. Chem. Soc., Vol. 61, pp. 6965-6970.
Barker, J. R., S. W. Benson, and D. M. Golden <1976), "The -Decomposi-
tion of Dimethyl Peroxide and the Rate Constant for QhO + 02 *
CH20 + H02," SRI Project No. 2554, Stanford Research Institute,
Menlo Park, California.
Batt, L., R. D. McCulloch, and R. T. Milne (1975), "Thermochemical and
Kinetic Studies of Alkyl Nitrites (RONO) - D(RONO), the Reactions
Between RO- and NO, and the Decomposition of RO-," Int. J. Chem.
Kinet. Symposium No. 1, pp. 441-461.
Baulch, D. L., D. D. Drysdale, and D. G. Home (1973), unpublished
data sheets prepared for NBS Technical Note 866, University of
Leeds.
Bell, E. R., et al. (1951), Disc. Faraday Soc., Vol. 10, p. 242.
-------
291
Bennett, J. E., D. M. Brown, and B. Mile (1970), "Studies by Electron
Spin Resonance of the Reactions of Alkylperoxy Radicals," Trans.
Faraday Soc.. Vol. 66, pp. 386-396.
Benson, S. W. (1975),-"Current Status of Methods for the Estimation of
Rate Parameters," Int. J. Chem. Kinetics Symposium No. 1,
pp. 359-378.
(1968), Th-ermochemical Kinetics (John Wiley & Sons, New
York, New Yqrk)7
Benson, S. W., and H. E. O'Neal (1970), Kinetic Data on Gas Phase
Unimolecular Reactions, NSRDS-NB521 (U.S. Government Printing
Office, Washington, D*C.).
Bufalini, J. J., S. L. Kopczynski , and M. C. Dodge (1972), "Contaminated
Smog Chambers in Air Pollution Research," Environ. Letters, Vol. 3,
pp. 101-109.
Calvert, J. G. (1976), "Hydrocarbon Involvement in Photochemical Smog^
Formation in Los Angeles Atmosphere , " £nvi ron . Sci . Techno! . ,
Vol. 10, p. 256.
Gal vert, J. 6., and R. D. McQuigg (1975), "The Computer Simulation of
the Rates and Mechanisms of Photochemical Smog Formation," Int.
J. Chem. Kinet. Symposium. No. "U...pp. 113-154.
Calvert, J. G., and J. N. Pitts, Jr. (1966), Photochemistry (John Wiley
& Sons, New York, New' York)".
Carter, W.P.L., et al. (1976), "Evidence for Alkoxy Radical
tion in C/^-Ce Alkanes in NO* -.Air Systems," 12th Informal Confer-
ence on Photochemistry, National Bureau of Standards ,, Extended
Abstracts, pp. N4-1 to N4-5.
Chan, W. H., et al. (1976), "Kinetic Study of ttONO Formation and Decay
Reactions in Gaseous Mixtures of HONO, NO, N02, H20, and NZ,"
Environ. Sci. Techno!., Vol. 10, pp. 674-682.
Cox, R. A. (1975), "The Photolysis of Nitrous Acid in the Presence of
Carbon Monoxide and Sulfur Dioxide," J. Photochem. , Vol. 4,
pp. 291-304.
(1974), "The Photolysis of Gaseous Nitrous Acid," J. Photo-
chemT, Vol. 3, pp. 175-188.
Cox, R. A., and R. G. Derwent (1975), "Kinetics of the Reaction of H02
with Nitric Oxide and Nitrogen Dioxide," J. Photochem.. Vol. 4,
pp. 139-153.
-------
292
Davis, D. D. (1976), private communication with Dr. M. Dodge of EPA.
Davis, D. D., W. Heaps, and T. McGee (1976), "Direct Measurements of
Natural Tropospheric Levels of OH. via an. Aircraft Borne Tunable
Dye Laser," Geophys. Res. Lett.. Vol. 3, pp. 331-333.
Demerjian, K. L., J. A. Kerr, and J. 6. Calvert (1974), "The Mechanism
of Photochemical Smog Formation," in Advances in Environmental
Science Technology, J. N. Pitts and R. L. Metcalf, eds. (John
Wiley & Sons, New York, New York).
Domalski, E. S. (1972), "Selected Values of Heats of Combustion and
Heats of Formation of Organic Compounds Containing the Elements
C, H, N, 0, P, and S," J. Phys. and Chem. Ref. Data, Vol. 1,
pp. 221-277.
Doyle, G. J., et al. (1975), "Gas Phase Kinetic Study of Relative Rates
of Reaction of Selected Aromatic Compounds with Hydroxyl Radicals
in an Environmental Chamber," Environ. Sci . Techno!., Vol. 9,
No. 3, pp. 237-241.."
Durbin, P. A., T. A. Hecht, and G. Z. Whitten (1975), "Mathematical
Modeling of Simulated Photochemical Smog," EPA-650/4-75-026,
Systems Applications, Incorporated, San Rafael, California.
Farrow, L. A., T. E. Graedel, and T. H. Weber (1975), "Heterogeneous
Removal of Free Radicals by Aerosols in the Urban Troposphere,"
Americas Chemical Society Symposium Series, in press.
Finlayson, B. J., and J. N-. Pitts, Jr. (197£}, "Photochemistry of the
Polluted Troposphere," Science, Vol. 192, -pp. 111-1.19.
Furuyama, S., and N. Ebara (1975), "A Kinetic Study of the Reaction of
0(3P) with Tol uene . " Int-, J . Chem! Ki net . , Vol. 7, pp. 689-698.
Graedel, T. E., L. A. Farrow, and T. H. Weber (1975), "The Influence
of Aerosols on the Chemistry of the Troposphere," Int. J. Chem.
Kinet. Symposium No. 1, pp. 581-594.
Graham, R. (1975), private communication.
Gray, P., R. Shaw, and J.C.J. Thynne (1967), "The Rate Constants of
Alkoxyl Radical Reactions," Prog, in React. Kinet.. Vol. 4,
pp. 63-117.
Greiner, N. R. (1970), "Hydroxyl Radical Kinetics by Kinetic Spectro-
scopy: VI Reactions with Alkanes in the Range 300 to 500K,"
J. Chem. Phys., Vol. 53, pp. 1070-1076.
-------
293
Hampson, R. F., and D. Garvin (1975), "Chemical Kinetic and Photo-
chemical Data for Modelling Atmospheric Chemistry," NBS Technical
Note 866, National Bureau of Standards, Washington, D.C.
Hansen, D. A., R. Atkinson, and J. N. Pitts, Jr. (1975), "Rate Constants
for the Reaction of OH Radicals with a Series of Aromatic Hydro-
carbons," J. Phys. Chem.. Vol. 79, No. 17, pp. 1763-1766.
Hecht, T. A., J. H. Seinfeld, and M. C. Dodge (1974a), "Further Develop-
ment of a Generalized Kinetic Mechanism for Photochemical Smog,"
Environ. Sci. Techno!.. Vol. 8, pp. 327-339.
Hecht, T. A., M. K. Liu, and D. C. Whitney (1974b), "Mathematical Simu-
lations of Smog Chamber Photochemical Experiments," EPA-650/4-74-040,
Systems Applications, Incorporated, San Rafael, California.
Heicklen, J. (1968), "Oxidation of Organic CompoundsII," Adv. Chem..
No. 76, p. 23.
Hendry, D. G. (1976), private communication.
Herron, J. T., and R. D. Penzhorn (1969), "Mass Spectrometric Study of
the Reactions of Atomic Oxygen with Ethylene and Formaldehyde,"
J. Phys. Chem., Vol. 73, pp. 191-196.
Japar, S. M., and H. Niki (1975), "Gas-Phase Reactions of the Nitrate
Radical with Olefins," J. Phys. Chem., Vol. 79, No. 16,
pp. 1629-1632.
Japar, S. M., C. H. Wu, and H. Niki (1974), "Rate Constants for the
Reaction of Ozone with Olefins in the Gas Phase," J. Phys. Chem.,
Vol. 78, No. 23, pp. 2318-2320.
Johnston, H. S., and R. Graham (1973), "Gas Phase Ultraviolet Absorp-
tion Spectrum of Nitric Acid Vapor," J. Phys. Chem., Vol. 77,
pp. 62-63.
Lloyd, A. C. (1974), "Evaluated and Estimated Kinetic Data for Gas Phase
Reactions of the Hydroperoxyl Radical," Int. J. Chem. Kinetics,
Vol. 6, pp. 169-228.
Lloyd, A. C., et al. (1976), "Relative Rate Constants for Reactions of
the Hydroxyl Radical with a Series of Alkanes, Alkenes, and
Aromatic Hydrocarbons," J. Phys. Chem., Vol. 80, No. 8,
pp. 789-794.
Mabey, W. R., and A. G. Hendry (1974), "Reactions of Hydroperoxy
Radicals," Stanford Research Institute, Menlo Park, California.
Menfenhall, G., D. M. Golden, and S. W. Bensen (1975), "The Very-Low-
Pressure Pyrolysis (VLPP) of n-Propyl titrate, tert-Butyl Nitrite,
and Methyl Nitrite. Rate Constants for Some Alkoxy Radical Reac-
tions," lnt^_jL_J&&t±J{Ane^., Vol. 7, pp. 725-737.
-------
294
Mill, T., and G. Montorsi (1973), "The Liquid-Phase Oxidation of
2,4-Dimethylpentane," Int. J. Chem. Kinetics, Vol. 5, pp. 119-136.
Morris, E. D., Jr., and H. Niki (1971), "Mass Spectrometric Study of the
Reaction of Hydroxyl Radical with Formaldehyde," J. Chem. Phys.,
Vol. 55, pp. 1991-1992.
(1973), "Reaction of Dinitrogen Pentoxide with Water,"
J. Phys. Chem., Vol. 77, pp. 1929-1932.
Niki, H., et al. (1976a), "IR Fourier-Transform Spectroscopic Studies
of Atmospheric Reactions," 12th Informal Conference on Photochemistry,
National Bureau of Standards, Extended Abstracts, pp. N1-N4
(peroxynitrate data given at meeting).
(1976b), "Fourier Transform Spectroscopic Studies of Gas Phase
Ozone-Olefin Reactions," 171st American Chemical Society National
Meeting.
Niki, H., E. E. Daby, and B. Weinstock (1972), Photochemical Smog and
Ozone Reactions, Advances in Chemistry Series 113 (American Chem-
ical Society, Washington, D. C.).
O'Neal, H. E., and C. Blumstein (1973), "A New Mechanism for Gas Phase
Ozone-Olefin Reactions," Int. J. Chem. Kinet., Vol. 5, pp. 397-413.
Osif, T. L., and J. I. Heicklen (1976), "Oxidation of HCO Radicals,"
" - J- Phys. Chem., Vol.'BO, pp. 1526-1531.
Parkes, D, A., et al. {1973). "The Ultraviolet Absorption by Alkylperoxy
Radicals and Their Mutual B^actiojis," Chem. Phys. Letters, Vol. 23,
pp. 425-429.
Pate, C. T., R. Atkinson, and J. N. Pitts, Jr. (1976), "Rate Constants for
the Gas Phase Reaction of Peroxyffcetyl Nitrate with Selected Atmo-
spheric Constituents," J. Environ, Sci. Health, Vol. A-ll, p. 19-33.
Reynolds, S. D. (1976), private communication.
Schwartz, N. (1974), "Chemical Characterization of Model Aerosols,"
EPA-650/3-74-011, Environmental Protection Agency, Research Triangle
Park, North Carolina.
Simonaitis, R., and J. I. Heicklen (1974), "Reactions of 0^03 with NO
and N02," J. Phys. Chem., Vol. 78, pp. 2417-2421.
Spicer, C. W., et al. (1973), "The Reactions of Methylperoxy Radicals
with NO and N02," J. Amer. Chem. Soc., Vol. 95, p. 13.
Stuhl, F., and H. Niki (1971), J. Chem. Phys., Vol. 55, p. 3904.
-------
295
Tsang, W. (1973), "Comparison between Experimental and Calculated Rate
Constants for Dissociation and Combination Reactions Involving Small
Polyatomic Molecules," Int. J. Chem. Kinet.. Vol. 5, pp. 947-963.
Volman, D. H., and R. A. Gorse (1972), "Photochemistry of the Gaseous
Hydrogen Peroxide-Carbon Monoxide Systems: Rate Constants for
Hydroxyl Radical Reactions by Competitive Kinetics," Final Report,
Project 5-143-2, University of California, Davis, California.
Wang, C. C., et al . (1975), "Hydroxyl Radical Concentrations Measured in
Ambient Air," Science, Vol. 189, pp. 797-800.
Weaver, 0., et al. (1975), "The Photooxidation of CDaNeCDs," J. Photochem.,
Vol. 4, pp. 109-120.
Weinstock, B. (1976), private communication.
Whitbeck, M. R.5 et al . (1976), "A Kinetic Study of CH302 and (CH3)3C02
Radical Reactions by Kinetic Flash Spectroscopy," 12th Informal
Conference on Photochemistry, National Bureau of Standards,
Extended Abstracts, pp. Kl-1 to Kl-5.
Wiebe, H. A., et al. (1973), "Photolysis of Methyl Nitrite in the Presence
of Nitric Oxide, Nitrogen Dioxide, and Oxygen," J. Amer. Chem. Soc.,
Vol. 95, pp. 7-12. ~
Wijnen, M.H.J. (1958), J. Amer. Chem. Soc.. Vol. 80, p. 2394.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/3-77-011
2.
3. RECIPIENT'S ACCESSIOf*NO.
4. TITLE A\D SUBTITLE
MATHEMATICAL MODELING OF
SMOG
SIMULATED PHOTOCHEMICAL
5. REPORT DATE
January 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHCRtSI
G. Z. Whitten and H. H. Hugo
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORG "^NIZATION NAME AND ADDRESS
Systems Applications, Inc.
950 Northgate Drive
San Rafael, California 94903
10. PROGRAM ELEMENT NO.
-1AA603 (1AA008)
11. CONTRACT/GRANT NO.
Contract No. 68-02-0580
12. SPONSORIN'G AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
14. SPONSORING AGENCY CODE
EPA-ORD
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This report deals with the continuing effort to develop a chemical kinetic
mechanism to describe the formation of photochemical smog. Using the technique
of computer modeling to simulate-smog chamber data, several explicit kinetic
mechanisms for specific hydrocarbons were analyzed and a generalized kinetic
mechanism was proposed for use in dynamic urban airshed models. Computer sim-
ulations of propylene/NO" , butane/NO , l-bu£ene./NO , and propylene/butane/NOx
systems were performed to simulate tne smog chamber data collected by the
University of California, Riverside. The general kinetic mechanism, which is
based on the carbon bonding found in the hydrocarbon system, 'is-described.
Simulations using this new carbon-bond mechanism were also compared with the
smog chamber data on the propylene/'NO , butane/NO ,-propytene/butane/NQ , and
toluene/NO systems. The usefulness and validity of the computer modeling is
also discussed relative to the current understanding of the -
process.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
COSATl Field/Group
Air pollution
Reaction kinetics
Photochemical reactions
Test chambers
Mathematical models
Computerized simulation
13B
07D
07E
14B
12A
09B
RELEASE TO PUBLIC
19. SECURITY CLASS (This Report)
UNCLASSIFIED
21. NO. OF PAGES
306
JO. SECURITY CLASS (This page)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
296
-------