JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 108, NO. D4, 4145, doi:10.1029/2002JD002432, 2003
Development of a chlorine mechanism for use
in the carbon bond IV chemistry model
Paul L. Tanaka,1 David T. Allen,1 Elena C. McDonald-Buller,2 Sunghye Chang,2
Yosuke Kimura,2 C. Buddie Mullins,3 Greg Yarwood,4 and James D. Neece5
Received 9 April 2002; revised 12 October 2002; accepted 21 November 2002; published 27 February 2003.
[1] Chlorine chemistry has been incorporated into the carbon bond IV mechanism and
employed in a regional photochemical model (the Comprehensive Air Quality Model with
Extensions (CAMx)) for preliminary use in assessing the regional impact of chlorine on
ozone formation in Houston, Texas. Mechanisms employed in regional photochemical
models do not currently account for chlorine chemistry. However, when chlorine
chemistry is accounted for, predicted ozone levels are enhanced by up to 16 ppbv in the
Houston area, with the greatest enhancement predicted for morning hours after sunrise.
Thirteen reactions have been added to the chemical mechanism used by CAMx to
describe chlorine chemistry in the urban atmosphere. The reactions include photolysis of
chlorine radical (Cl) precursors, Cl+ hydrocarbon reactions, and Cl+ ozone reactions.
The hydrocarbon reactions include the reactions of Cl with isoprene and 1,3-butadiene
that yield unique reaction products, or marker species. The development of this
mechanism is presented along with a discussion of the initial set of predictions of
chlorine-based ozone enhancement in the Houston area. Of significant interest is that
methane may be activated by chlorine to contribute significantly to the predicted ozone
enhancement in the Houston area. Such behavior suggests that the impact of chlorine
chemistry would be proportional to the availability of Cl precursor. In urban areas with
anthropogenic sources of chlorine radical precursors, chlorine radical chemistry may be
INDEX TERMS: 0345 Atmospheric
important to more accurately predict ozone formation.
Composition and Structure: Pollution—urban and regional (0305); 0365 Atmospheric Composition and
Structure: Troposphere—composition and chemistry; 0368 Atmospheric Composition and Structure:
Troposphere—constituent transport and chemistry; KEYWORDS: regional photochemical modeling, chlorine
chemistry, CAMx, ozone formation, regional air quality
Citation: Tanaka, P. L., D. T. Allen, E. C. McDonald-Buller, S. Chang, Y. Kimura, C. B. Mullins, G. Yarwood, and J. D. Neece,
Development of a chlorine mechanism for use in the carbon bond IV chemistry model, J. Geophys. Res., 108(D4), 4145,
doi:10.1029/2002JD002432, 2003.
1. Introduction
[2] Urban air pollution is typically characterized by high
concentrations of ozone produced by the reaction of volatile
organic compounds (VOCs) with hydroxyl radicals (OH) in
the presence of nitrogen oxides (NOx). Ozone formation
chemistry initiated by hydroxyl radicals is well understood
[Finlayson-Pitts and Pitts, 2000; Seinfeld and Pandis,
1998]. Because ozone is a secondary air pollutant (formed
1
Center for Energy and Environmental Resources and Department of
Chemical Engineering, University of Texas at Austin, Austin, Texas, USA.
2
Center for Energy and Environmental Resources, University of Texas
at Austin, Austin, Texas, USA.
3
Department of Chemical Engineering, University of Texas at Austin,
Austin, Texas, USA.
4
ENVIRON International Corporation, Novato, California, USA.
5
Texas Commission on Environmental Quality, Austin, Texas, USA.
Copyright 2003 by the American Geophysical Union.
0148-0227/03/2002JD002432$09.00
ACH
as a product of chemical reactions involving pollutants
emitted directly to the air), ozone formation is influenced
not only by chemistry, but also by emissions and physical
processes such as meteorology and transport. A regional
photochemical model such as the Comprehensive Air Quality Model with Extensions (CAMx) may be employed to
simultaneously account for all of these factors. Models such
as CAMx are routinely employed to develop plans to
mitigate ozone formation in urban areas.
[3] Despite the sophistication of regional photochemical models, predicted ozone levels can be significantly
lower than observed ozone levels [Texas Register, 2000].
One hypothesis investigated by Tanaka et al. [2000] is
that chlorine radical (Cl) chemistry, which is not currently accounted for in regional photochemical models,
can lead to enhanced ozone formation in urban industrial
areas.
[4] Similar to OH, Cl reacts quickly with VOCs [Atkinson, 1997; Atkinson et al., 1999, 2000] and has been
implicated as a cause of enhanced ozone formation in
6-1
ACH
6-2
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
Houston, Texas [Tanaka et al., 2000] and Telemark, Norway [Hov, 1985]. Tanaka et al. observed rapid enhancement
of ozone formation when molecular chlorine (Cl2), a Cl
precursor, was injected into simulated urban air mixtures
[Tanaka et al., 2000] and captive Houston-area air [P.
Tanaka et al., An environmental chamber investigation of
chlorine-enhanced ozone formation in Houston, Texas,
submitted to Journal of Geophysical Research, 2002;
Tanaka et al., 2003].
[5] Plans by the State of Texas to bring ambient ozone
levels in Houston and the nearby Galveston area into federal
regulatory compliance include deep reductions in NOx
emissions. For example, 90% reduction in point-source
and 60% reduction in mobile sources of NOx emissions
will be required by 2007. The estimated cost of achieving
these cuts is several billion (109) dollars [Texas Register,
2000]. Because chlorine chemistry is not currently
accounted for by regional photochemical models, there is
no method to determine the impact of Cl chemistry on
regional ozone formation or the potential benefit of instituting controls on anthropogenic chlorine emissions.
[6] Chang et al. [2001, 2002] (the report by Chang et al.
[2001] can be found online at http://www.tceq.state.tx.us)
have compiled an emissions inventory of chlorine for the
11-county area of Houston and surrounding areas. They
suggest that approximately 12 tons per day of chlorine (as
Cl2) are emitted in the 11-county region.
[7] To evaluate the potential impact of Cl chemistry on
ozone formation in the Houston-Galveston, Texas area, and
to provide a tool to evaluate how Cl impacts regional ozone
formation, a condensed form of Cl chemistry has been
incorporated into the carbon bond IV (CB-IV) chemical
mechanism used in CAMx. CAMx is currently being used
by the Texas Commission on Environmental Quality
(TCEQ, formerly the Texas Natural Resource Conservation
Commission) to evaluate plans for mitigating ozone formation in the Houston-Galveston area. This photochemical
model is used here because several historical modeling
periods are available to estimate the effect of chlorine
chemistry on ozone formation, and the model is publicly
available [ENVIRON, 2002] (available online at http://
www.camx.com). The historical modeling periods have also
undergone extensive scrutiny by the TCEQ and U.S. Environmental Protection Agency.
[8] The 6 – 11 September 1993 modeling episode was
used in simulations using the modified chemical model.
We present here the development of the Cl mechanism
incorporated into the CB-IV mechanism that is used by
CAMx and a number of other models that are employed
for regulatory assessment and research. It should be noted
that all chemical mechanisms used in regional photochemical modeling have weaknesses. It is not the intent of this
manuscript to justify the accuracy of the CB-IV mechanism currently used by the TCEQ, but rather to provide a
condensed chlorine radical mechanism that can be used in
a regional photochemical model that is available for
immediate incorporation of chlorine radical chemistry
and evaluation of the effects of incorporating such chemistry. Evaluations of the performance of CB-IV and other
mechanisms can be found in the literature [Liang and
Jacobson, 2000; Bergin et al., 1998; Simonaitis et al.,
1997].
2. Chlorine Chemical Mechanism for Use in
CAMx
[9] CAMx is an Eulerian photochemical grid model
developed by ENVIRON Corporation [ENVIRON, 2002].
It allows for integrated assessment of gas and particulate
phase air pollution from urban to superregional scales.
[10] As previously mentioned, the chemical mechanism
employed by CAMx has been modified to include chlorine
radical chemistry. The intent is to provide a preliminary
evaluation of the importance of chlorine to urban ozone
formation. A complete description of the original photochemical model is provided elsewhere [ENVIRON, 2002].
We describe the implementation of chlorine chemistry in
the CB-IV chemical mechanism in the remainder of this
manuscript.
[11] The original chemical model used by CAMx for
regional photochemical simulations is based on the CB-IV
mechanism [Gery et al., 1989]. The base chemical mechanism (the mechanism without chlorine chemistry) employed
during our CAMx simulations includes updated radical
termination reactions and the ‘‘one-product’’ isoprene
chemistry proposed by Carter [1996]. The complete base
mechanism can be found in the CAMx User’s Guide
[ENVIRON, 2002]. In summary, the base mechanism
accounts for parameterized and real species in a combination of 96 reactions. Radical, inorganic, and a few organic
species (e.g., ethene and isoprene) are explicitly represented. However, the different types of functional groups
contained in complex organic species are individually
represented by parameterized species. For example,
‘‘PAR’’ refers to a paraffinic (or saturated) carbon bond,
whereas ‘‘OLE’’ refers to a terminal olefinic (or unsaturated) C = C bond.
[12] Because of the parameterized representation of complex organics, only 13 reactions were incorporated into the
chemical mechanism employed by CAMx to describe
chlorine chemistry relevant to an urban atmosphere such
as in Houston. Thus the chemistry employed during our
CAMx simulations included 109 reactions. The 13 reactions and the corresponding rate constants are provided in
Table 1. Table 2 contains a key to the species included in
Table 1. To minimize the additional computational requirements of the additional chlorine chemistry, 9 of the 13
additional reactions are assigned rate constants that are
scaled to reactions in the original chemical mechanism. A
description of these reactions, suggested rate parameters,
and rationale for excluding other classes of reactions are
provided subsequently.
2.1. Establishing Rate Parameters and
Reaction Stoichiometry
2.1.1. Photolysis
[13] Two photolytic sources of Cl, were incorporated into
CB-IV. The two species, molecular chlorine (Cl2) and
hypochlorous acid (HOCl), were chosen because there are
documented emissions sources of Cl2 in the HoustonGalveston area, and HOCl is a volatile form of active
chlorine in aqueous solution. Although the model can
accommodate HOCl as a photolytic Cl precursor, Cl2 is
assumed to be the sole source of photolytic Cl precursors in
the modeled episode (6 – 11 September 1993) presented in
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
ACH
6-3
Table 1. Chlorine Chemistry Incorporated Into CAMx
Reaction
Reaction
k, cm3 molecule1 s1
(R1)
(R2)
(R3)
(R4)
(R5)
(R6)
(R7)
(R8)
(R9)
(R10)
(R11)
(R12)
(R13)
Cl2 = 2Cl
HOCl = OH + Cl
Cl + PAR = HCl + 087XO2 + 0.13XO2N + 0.11HO2 + 0.11RCHO + 0.76ROR 0.11PAR
Cl + OLE = FMCL + RCHO + 2XO2 + HO2 1PAR
Cl = HCl + XO2 + FORM + HO2
Cl + ETH = FORM + 2XO2 + FMCL + HO2
Cl + ISOP = 0.15HCl + XO2 + HO2 + 0.28ICL1
OH + ICL1 = ICL2
Cl + BUTA = XO2 + HO2 + 0.70BCL1
OH + BCL1 = BCL2
Cl + O3 = ClO + O2
ClO + NO = Cl + NO2
ClO + HO2 = HOCl + O2
0.264kp,NO2a
143kp,ISPDa
78kOH,PAR
20kOH,OLE
6.6 1012 exp(1240/T)
12.6kOH,ETH
4.5kOH,ISOP
0.19kOH,ISOP
4.2kOH,ISOP
0.36kOH,ISOP
2.9 1011 exp(260/T)b
6.2 1012 exp(295/T)b
4.6 1013 exp(710/T)b
a
The rate of these photolysis reactions is dependent on calculated sunlight intensity.
Source: Atkinson et al. [2000].
b
this paper. The rate of photolysis for Cl2 is scaled directly to
the rate constant for NO2 photolysis. This scaling was done
to simplify execution and minimize additional calculational
complexity with the introduction of the chlorine mechanism. This photolytic scaling is valid because the ratio of
photolysis rates of the two species (Cl2 and NO2) is similar
across the actinic spectrum, and more precisely, across
different zenith angles.
[14] With values for the absorption cross section and
quantum yield obtained from literature [Atkinson et al.,
1997], the photolysis rates for the Cl precursors were
discretely calculated in the same fashion as all other photolytic species in CAMx. Photolytic rates were calculated for
some typical conditions: zenith angles, 0, 10, 20, 30,
40, 50, 60, 70, 78, and 86; elevation, 640 m above
ground; ozone column, 300 Dobson units; UV surface
albedo, 0.06; and optical depth due to haze, 0.1.
[15] The following general expression was used to calculate the rate of photolysis of each species:
kp;x s1 ¼
li
X
sðlÞfðlÞJ ðlÞ;
ð1Þ
l¼290 nm
where s(l) is the absorption cross section (cm2), f(l) is the
quantum yield, J(l) is the actinic flux of photons (cm2
s1), and li is the threshold wavelength for photolysis to
occur (nm).
[16] Values of J(l) were calculated by the TUV model for
the typical conditions above. Photolysis rates for CAMx are
usually prepared using TUV [Madronich and Flocke, 1998].
Tabulated values of s(l) were obtained from Atkinson et al.
[1997]. The quantum yield for Cl2 photolysis is unity for all
wavelengths of concern [Calvert and Pitts, 1966].
[17] Ratios of the photolysis rates were calculated for
different pairs of photolytic species to identify which
species in the original mechanism had photolysis rates that
scaled with Cl2 and HOCl. The surrogates were accepted if
the ratio of calculated photolysis rates over the test conditions had a standard deviation of less than 10% of the
mean. The surrogates for Cl2 and HOCl were NO2 and
ISPD, respectively. ISPD is an oxidation product of isoprene in the base chemical mechanism. The standard
deviations of the ratios of calculated photolysis rates for
the chosen photolytic pairs (Cl2/NO2 and HOCl/ISPD) are
8.1 and 6.4%, respectively. Although NO2 was evaluated as
a surrogate for HOCl photolysis (kp,HOCl = 0.0262kp,NO2),
the standard deviation was calculated to be 18.7% of the
mean.
[18] The ratio of the calculated rates of photolysis of Cl2
and HOCl that are used in the model are 0.264kp,NO2 and
143kp,ISPD, respectively, where kp,X denotes the rate for
photolysis of species X in units of s1.
2.1.2. Methane
[19] The reaction rate constant for the chlorine-methane
reaction is almost 2 orders of magnitude greater than that of
Table 2. Key to Chemical Species Abbreviationsa
Name
Description
BUTA
BCL1
BCL2
ETH
Cl
Cl2
ClO
CO
FORM
FMCL
HCl
HOCl
HO2
ISOP
ISPD
ICL1
ICL2
OH
OLE
1,3 butadiene
4-chlorocrotonaldehyde (CCA)
CCA + OH reaction products
ethene
chlorine atom
molecular chlorine
chlorine oxide
carbon monoxide
formaldehyde
formyl chloride
hydrochloric acid
hypochlorous acid
hydroperoxyl radical
isoprene
oxidation product of isoprene
1-chloro-3-methyl-3-butene-2-one
ICL1 + OH reaction products
hydroxyl radical
terminal olefinic bond (carbon
double bond)
paraffinic carbon
higher aldehyde
organic nitrate forming peroxy
radical operator used in the
PAR mechanism
universal peroxy radical operator
used to represent NO-to-NO2
conversions in multistep
mechanisms
nitrate forming peroxy radical
operator
PAR
RCHO
ROR
XO2
XO2N
a
Example: propene would be represented as 1 PAR + 1 OLE.
ACH
6-4
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
of HCl in the case of hydrogen abstraction by Cl. The
product stoichiometries are as follows:
ðR2Þ
OH þ PAR ¼ 0:87XO2 þ 0:13XO2 N þ 0:11HO2
þ 0:11RCHO þ 0:76ROR 0:11PAR
ðR3Þ
Cl þ PAR ¼ HCl þ 0:87XO2 þ 0:13XO2 N þ 0:11HO2
þ 0:11RCHO þ 0:76ROR 0:11PAR:
Figure 1. Mechanism for the reaction between chlorine
radical (Cl) and methane in air.
the hydroxyl radical-methane reaction. Because of this
significantly higher reactivity, we have included the methane-chlorine radical reaction in the chlorine mechanism. The
reaction kinetics are well known and the reaction mechanism proceeds as in Figure 1.
[20] Methane is represented implicitly within the CB-IV
mechanism. A global average of 1.85 ppm methane is
assumed and used to provide an ‘‘unimolecular’’ rate
expression for OH. The reaction stoichiometry included
for chlorine radicals is similar to the representation of the
OH-methane reaction in the original CB-IV mechanism.
However, HCl is formed as a product:
ðR1Þ
Cl ¼ HCl þ XO2 þ FORM þ HO2 :
[21] Note that methane does not appear explicitly as a
reactant because the background level of 1.85 ppm is
assumed and accounted for in the rate constant. The CBIV mechanism notation is described in Table 2 [Gery et al.,
1989]. The rate constant for the reaction of chlorine radicals
with methane is k = 6.6 102 [exp(1240/T )] cm3
molecule1 s1 [Atkinson et al., 1999].
2.1.3. Paraffins
[22] Paraffins, or saturated hydrocarbons more complex
than methane, are parameterized in the CB-IV chemical
model. According to known reaction mechanisms [Atkinson, 1997], chlorine radicals initiate oxidation of alkanes by
abstraction of hydrogen, in a manner similar to hydroxyl
radicals. The abstraction of hydrogen from alkanes by
chlorine radicals is included in the mechanism. The specific
reaction products would be expected to differ because the
branching ratios for abstraction from different sites is different for abstraction by Cl and OH. However, the difference
is expected to have an insignificant impact on ozone
productivity compared to approximations in the CB-IV
representation of alkanes.
[23] Because the base CB-IV chemical model provides
the reaction stoichiometry of the OH + PAR reaction, and
since Cl and OH react with paraffins in a similar manner
(through hydrogen abstraction, with the caveat stated
above), the proposed overall product stoichiometry is very
similar for both the hydroxyl- and chlorine radical-initiated
oxidation of paraffins. The only difference is the formation
[24] The reaction rate constant for Cl + PAR is 78kOH+PAR.
This rate is based on the ratio of the average of Cl and OH
reaction rate constants [Atkinson, 1997] for 20 alkanes (Table
3). Thus the rate constant for reaction of Cl with a paraffinic
carbon is represented as kCl,PAR = 78kOH+PAR.
2.1.4. Ethene and Olefins
[25] CAMx handles ethene as an explicit species. To
simplify calculation of the chlorine radical-ethene reaction
rate constant, the rate constant is scaled to the hydroxyl
radical-ethene reaction rate expression at 298 K. At 298 K,
the reaction rate constant of the chlorine radical-ethene
reaction is 1.07 1010 cm3 molecule1 s1 [Atkinson,
1997]. This rate constant is 12.6 times the reaction rate
constant of the hydroxyl radical-ethene reaction suggested
by Atkinson [1997] of 8.52 1012 cm3 molecule1 s1.
Therefore the chlorine radical-ethene rate constant is represented as kCl,ETH = 12.6kOH,ETH.
[26] The chlorine radical-ethene reaction stoichiometry
employed in this mechanism is based on Carter’s mechanism [Carter et al., 1997a, 1997b]:
ðR4Þ
Cl þ ETH ¼ FORM þ 2XO2 þ FMCL þ HO2 :
Subsequent to incorporating this reaction, other references
came to our attention that provided experimental evidence
for different primary reaction products. Chlorine-initiated
oxidation of ethene produces chloroacetaldehyde [Wallington et al., 1990; Yarwood et al., 1992; Orlando et al.,
1998], which undergoes oxidation by OH, photolysis
[Balestra-Garcia et al., 1992], and Cl. Although the
mechanism proposed above is not explicitly accurate, the
final oxidation products of the proposed reaction and
those resulting from the intermediate formation of
chloroacetaldehyde are believed to be nearly the same.
To ensure that the proposed reaction adequately captures
the chemistry important to ozone formation resulting from
the chlorine-ethene reaction, a series of box model
simulations were performed using a more detailed
chemical mechanism contained in SAPRC-99 [Carter,
2000]. The results of this sensitivity analysis are presented
in section 2.2.
[27] The base CB-IV mechanism employed by CAMx
gives the following stoichiometry for the hydroxyl radicalolefin reaction:
ðR5Þ OH þ OLE ¼ FORM þ RCHO þ XO2 þ HO2 1PAR:
A similar condensed mechanism was employed to represent
the more general Cl + OLE reaction. The differences
between the Cl + OLE mechanism and the chlorine-ethene
mechanism were the substitution of larger aldehydes
6-5
ACH
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
Table 3. Reaction Rate Constants of OH and Cl With Selected Paraffins and Olefins
Compound
(Paraffins)
1011 k298,Cl, cm3
molecule1 s1a
1011 k298,OH, cm3
molecule1 s1a
Number
of Carbons
1011 k298,Cl,C,
cm3 C1 s1
1011 k298,OH,C,
cm3 C1 s1
Ethane
Propane
n-Butane
2-Methylpropane
n-Pentane
2-Methylbutane
2,2-Dimethylpropane
n-Hexane
2-Methylpentane
3-Methylpentane
2,3-Dimethylbutane
n-Heptane
2,4-Dimethylpentane
2,2,3-Trimethylbutane
n-Octane
2,2,4-Trimethylpentane
2,2,3,3-Tetramethylbutane
n-Nonane
n-Decane
cis-Bicyclo[4,4,0]decane
Average
5.9
13.7
21.8
14.3
28
22
11.1
34
29
28
23
39
29
20
46
26
17.5
48
55
48
0.0254
0.112
0.244
0.219
0.4
0.37
0.0848
0.545
0.53
0.54
0.578
0.702
0.5
0.424
0.871
0.357
0.105
1
1.12
1.9
2
3
4
4
5
5
5
6
6
6
6
7
7
7
8
8
8
9
10
20
3.0
4.57
5.45
3.58
5.6
4.4
2.22
5.7
4.8
4.7
3.8
5.6
4.1
2.9
5.8
3.3
2.19
5.3
5.5
2.4
0.0127
0.0373
0.061
0.0548
0.08
0.074
0.0170
0.0908
0.088
0.09
0.0963
0.100
0.07
0.061
0.109
0.0446
0.0131
0.11
0.112
0.095
Compound
(Olefins)
Propenea
1,2-Propadienea
1-Buteneb
1-Penteneb
1011 k298,CL*,
cm3 mol1 s1
1011 k298,OH*,
cm3 mol1 s1
28
42
35.2
49.3
2.63
0.982
3.14
3.14
Number of 1011 k298,Cl,DB, cm3
double bond1 s1
Double Bonds
1
2
1
1
28.0
21.0
35.2
49.3
1011 k298,OH,DB, cm3
double bond1 s1
k298,Cl,C/k298,OH,C
232
122
89
65
70
59
131
62
55
52
40
56
58
47
53
73
167
48
49
25
78
k298,Cl,DB/k298,OH,DB
2.63
0.491
3.14
3.14
Average
11
43
11
16
20
a
From Atkinson et al. [1997].
b
From Coquet and Ariya [2000].
(RCHO) for formaldehyde (FORM) and the subtraction of a
PAR. Therefore the reaction stoichiometry employed for the
chlorine radical-olefin reaction is given as:
ðR6Þ
Cl þ OLE ¼ FMCL þ RCHO þ 2XO2 þ HO2 1PAR:
[28] Although formyl chloride (FMCL) photolyzes in the
troposphere, the lifetime of this species is more than 1
month against tropospheric photolysis. FMCL is therefore
left as an inert species in the incorporated chemistry.
[29] Based on the ratio of average reaction rates [Atkinson, 1997] (per double bond) for Cl to OH with several
alkenes (Table 3), a reaction rate constant for Cl + OLE of
20 kOH+OLE is used.
[30] As shown in Table 3, 1,2-propadiene and 1-pentene
were used to help determine the rate constant for the Cl +
OLE reaction. Although 1,2-propadiene consists of two
terminal olefinic bonds, both are also conjugated. The rate
constant multiplier for 1-pentene was not corrected for
contributions from reaction on paraffinic bonds. If neither
of these compounds is considered, the rate constant for the
Cl + OLE reaction would be 11kOH+OLE. To determine the
importance of this multiplier, a sensitivity analysis was
performed using the box model described in section 2.2.
The results of the analysis are provided in section 2.2.
2.1.5. Isoprene
[31] The chlorine radical-isoprene reaction is included in
this model to track the unique marker species (1-chloro-3methyl-3-butene-2-one) for this reaction. The reaction
mechanism of the chlorine-isoprene reaction is very com-
plex and is not explicitly understood in the presence of NOx.
Because 1-chloro-3-methyl-3-butene-2-one can react with
OH, the corresponding reaction has been added to account
for consumption of 1-chloro-3-methyl-3-butene-2-one after
being produced by the Cl + ISOP reaction.
[32] The following simplifying assumptions were made to
arrive at a stoichiometry for the Cl + ISOP reaction. These
assumptions are as follows: (1) Formation of organic
nitrates and organic compounds other than 1-chloro-3methyl-3-butene-2-one are ignored in this reaction; and
(2) measurements of the rate of reaction of 1-chloro-3methyl-3-butene-2-one with OH are not available. The
stated rate constant is estimated from contributions from
OH addition to the olefinic bond and abstraction of the
methyl and chloromethyl hydrogens. The rationale for
estimating this rate constant is provided subsequently.
[33] The simplified reaction mechanism for Cl + ISOP is
based on the work of Ragains and Finlayson-Pitts [1997]
and Nordmeyer et al. [1997]. In the mechanism incorporated
into CB-IV, the Cl + ISOP reaction results in a 28% yield of
1-chloro-3-methyl-3-butene-2-one (ICL1), resulting in the
following stoichiometry:
ðR7Þ
Cl þ ISOP ¼ 0:15 HCl þ XO2 þ 0:28ICL1:
[34] Because OH reacts with ICL1, the following reaction is included to account for consumption of ICL1 after it
is produced:
ðR8Þ
OH þ ICL1 ¼ ICL2;
ACH
6-6
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
Figure 2. Chemical structure of 1-Chloro-3-methyl-3butene-2-one (ICL1) and methacrolein (MACR).
where ICL 1 represents 1-chloro-3-methyl-3-butene-2-one
and ICL2 represents the reaction products of the OH + ICL1
reaction. None of the other organic products are tracked.
[35] The rate constant for reaction (R8) has not been
reported in literature. Since ICL1 is structurally similar to
methacrolein (MACR), an estimate was made of the rate
constant for ICL1 + OH by examining the reaction of OH
with MACR. For clarity, the structures are provided in
Figure 2.
[36] According to Orlando et al. [1999], OH attack on
MACR proceeds 55% by addition to the double bond and
45% via abstraction of the aldehydic hydrogen. It is
assumed that OH addition contributes to the reaction rate
of OH + ICL1 and OH + MACR similarly. By assuming
that the rate of addition to the double bond is unaffected by
the substitution of the aldehydic hydrogen on MACR, the
contribution of OH addition to the double bond in ICL1 is
0.55 * 3.35 1011 cm3 molecule1 s1 or 1.84 1011
cm3 molecule1 s1.
[37] The five hydrogen atoms available for abstraction on
ICL1 are of two types: three alkyl hydrogens of the 3methyl group and two terminal chloromethyl (CH2Cl)
hydrogens. The contribution of the H-atom abstraction by
OH is scaled to the rate of abstraction of H atoms in the OH
+ ethane and OH + chloroethane reactions. The contribution
of H-atom abstraction to the overall rate of OH + ICL1 is
estimated at 2.9 1013 cm3 molecule1 s1. An overall
reaction rate of 1.87 1011 cm3 molecule1 s1 results
from summing the contribution of H-atom abstraction with
that of OH addition.
[38] Based on the ratio of reaction rate constants for the
chlorine radical-isoprene reaction and the hydroxyl radicalisoprene reaction, the proposed rate constant for the Cl +
isoprene reaction is represented as kCl,ISOP = 4.5kOH,ISOP.
The corresponding reaction rate constant for the OH + ICL1
reaction is represented as kOH+ICL1 = 0.19kOH+ISOP.
[39] The Cl + ICL1 reaction is not included in the
mechanism because it is not expected to be an important
loss pathway. The lifetimes of ICL1 against OH and Cl are 6
and 28 hours, respectively. The radical concentrations and
rate constants used to calculate these lifetimes are provided
below.
[40] The concentration of hydroxyl radicals is equal to the
median during TEXAQS 2000, [OH] = 2.5 106 cm3
[Riemer, 2001] (this document can be found online at http://
www.tceq.state.tx.us). The concentration of chlorine radicals is assumed to be 5.0 104 cm3, the median
estimated by Riemer [2001] during TEXAQS 2000. The
rate constant for the OH + ICL1 reaction is estimated above
to be 1.87 1011 cm3 molecule1 s1. The rate constant
for the Cl + 1-chloro-3-methyl-3-butene-2-one reaction is
assumed to be the same as chloromethyl vinyl ketone. The
rate of the surrogate reaction has been reported by Wang et
al. [2002] as 2 1010 cm3 molecule1 s1.
2.1.6. 1,3-Butadiene
[41] The reaction of 1,3-butadiene (BUTA) with chlorine radicals is included in the modification to CB-IV for
the purpose of tracking the concentration of 4-chlorocrotonaldehyde, identified by Wang and Finlayson-Pitts
[2000] as a species unique to the reaction of chlorine
radicals with BUTA. However, BUTA is not a species
included in the CB-IV inventory used by CAMx. Therefore its contributions are not represented in the simulation
results. If future inventories are developed to include
BUTA, the OH + BUTA chemistry and subsequent
reactions would need to be incorporated into the CB-IV
mechanism to properly account for the contribution of
BUTA to ozone production.
[42] In the presence of NOx, the yield of 4-chlorocrotonaldehyde from the reaction of Cl with BUTA is approximately 70% [Wang and Finlayson-Pitts, 2000]. As with the
isoprene + Cl marker species, 4-chlorocrotonaldehyde
(BCL1) is the only organic reaction product that is retained.
Because this species readily reacts with OH, a second
reaction (OH + BCL1) is added.
[43] The following simplifying assumptions have been
made: (1) Organic nitrate formation from 1,3-butadiene is
ignored; and (2) measurements of the rate of reaction of 4chlorocrotonaldehyde with OH are not available. However,
it is expected that the rate of reaction is similar to that of
crotonaldehyde with OH [Wang and Finlayson-Pitts,
2000]. Therefore the reaction rate for BCL1 + OH is
assumed to be the same as that for crotonaldehyde + OH,
or kOH,BCL1 = 3.6 1011 cm3 molecule1 s1. For
simplicity, the rates are presented as ratios of the OH +
isoprene reaction rate at 298 K.
[44] The reaction stoichiometry for Cl + BUTA, included
for the purpose of tracking the marker species BCL1, is
based on the work of Wang and Finlayson-Pitts [2000]:
ðR9Þ
Cl þ BUTA ¼ XO2 þ HO2 þ 0:70BCL1:
Because OH can react with BCL1, the following reaction is
included to account for consumption of BCL1 after it is
produced:
ðR10Þ
OH þ BCL1 ¼ BCL2;
where BCL1 represents 4-chlorocrotonaldehyde and BCL2
represents the reaction products of the OH + BCL1 reaction.
As with the Cl + ISOP reaction, no other organic products
are tracked.
[45] Based on the ratio of reaction rate constants for the
chlorine-1,3-butadiene reaction and the OH + isoprene
reaction, the proposed rate constant for the Cl + BUTA
reaction is represented as kCl,BUTA = 4.2kOH,ISOP. The
corresponding reaction rate constant for the OH + BCL1
reaction is represented as kOH,BCL1 = 0.36kOH,ISOP.
[46] The Cl + BCL1 reaction is not included in the
mechanism because it is not expected to be an important
loss pathway. Assuming [OH] and [Cl] of 2.5 106 and
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
ACH
6-7
Table 4. List of Reactions Between Chlorine and Inorganic Compounds Accounted for in the Modified SAPRC-99 Mechanism
(Mechanism Based on Carter et al. [1997a])
Reaction No.
(R1)
(R2)b
(R3)
(R4)b
Cl2 + hn ! 2Cl
Cl+ NO + M ! ClNO + M
ClNO + hn ! Cl+ NO
Cl+ NO2 ! ClONO
(R5)b
Cl+ NO2 ! ClNO2
(R6)
(R7)
(R8)b
(R9)b
(R10)b
(R11)b
(R12)
(R13)
ClONO + hn ! Cl+ NO2
ClNO2 + hn ! Cl+ NO2
Cl+ HO2 ! HCl + O2
Cl+ HO2 ! ClO+ OH
Cl+ O3 ! ClO+ O2
Cl+ NO3 ! ClO+ NO2
ClO+ NO ! Cl+ NO2
ClO+ NO2 ! ClONO2
(R14)
ClONO2 + hn ! 0.9 {Cl+ NO3}
+ 0.1 {O3P + ClONO}
ClONO2 ! ClO+ NO2
Cl+ ClONO2 ! Cl2 + NO3
ClO+ HO2 ! HOCl + O2
HOCl + hn ! OH+ Cl
Cl+ H2 ! HCl + HO2
(R15)
(R16)
(R17)
(R18)
(R19)
Temperature Dependence of k, cm3 molecule1 s1)
Reaction
a
9.0 1032 (T/300)1.6c
a
k0 ¼ 1:30 1030 ðT =300Þ2 c
ðfalloff kineticsÞ kinf ¼ 1:00 1010 ðT =300Þ1
F ¼ 0:6
k0 ¼ 1:80 1031 ðT =300Þ2 c
ðfalloff kineticsÞ kinf ¼ 1:00 1010 ðT =300Þ1
F ¼ 0:6
a
a
1.8 1011 exp(170/T )d
4.1 1011 exp(450/T )b
2.9 1011 exp(260/T )d
2.40 1011d
6.2 1012 exp(295/T)d
k0 ¼ 1:60 1031 ðT =300Þ3:4 ½N2 ðfalloff kineticsÞ
kinf ¼ 1:5 1011
F ¼ 0:5
d
a
KEQ k13, KEQ = 5.20 1025 exp(12,000/T)(T/300)3.4e
6.5 1012 exp(135/T )d
4.6 1013 exp(710/T )d
a
3.9 1011 exp(2310/T)d
a
The rate of these photolysis reactions is dependent on calculated sunlight intensity. Cross-section data are taken from the work of Atkinson et al. [1997].
Key reaction that can be omitted to effectively eliminate subsequent reactions involving the products of this key reaction.
DeMore et al. [1997].
d
Atkinson et al. [2000].
e
Carter et al. [1997a].
b
c
5.0 104 cm3 and the reaction rate constants kOH,BCL1 =
3.6 1011 cm3 molecule1 s1 and kCl,BCL1 = 1.6 1010
cm3 molecule1 s1 [Wang et al., 2002], the lifetimes of
BCL1 against OH and Cl are 3 and 35 hours, respectively.
2.1.7. Ozone
[47] Although the Cl + ozone reaction is an important
one in the chlorine-catalyzed destruction of stratospheric
ozone [Molina and Rowland, 1974], Cl may result in
significant ozone formation in parts of the troposphere. This
dichotomy occurs because of the higher concentration of
hydrocarbons in the troposphere resulting in the faster
reaction rate of Cl with hydrocarbons versus ozone in the
troposphere. According to Oum et al. [1998], when considering the reaction of Cl with ozone (40 ppb) versus the
reaction of Cl with background methane and 10 ppb of
organics other than methane, approximately 70% of Cl
atoms would react with methane and organics. In the
Houston atmosphere, where nonmethane hydrocarbon levels are high (1000 ppbc) relative to ozone levels (150
ppbv), it is anticipated that a very small fraction of chlorine
atoms would react with ozone. However, the reaction of Cl
+ ozone and subsequent reaction of the corresponding
reaction product (ClO) are included for completeness. The
included reactions are:
ðR11Þ
Cl þ O3 ¼ ClO þ O2
ðR12Þ
ClO þ NO ¼ Cl þ NO2
ðR13Þ
ClO þ HO2 ¼ HOCl þ O2
[48] The rate constants associated with these reactions are
kCl,O3 (T ) = 2.9 1011 exp(260/T ), kCl,NO (T ) = 6.2 Figure 3. Sensitivity run employing the SAPRC-99
model. Ozone mixing ratios are shown over the course of
the run for three different cases. (1) Base case with all
chlorine reactions turned ‘‘on.’’ (2) Same as case 1, but with
all Cl-inorganic reactions omitted. (3) Same as case 1, but
with the Cl-pentane reaction omitted.
ACH
6-8
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
Figure 4. Maximum difference in ozone levels during the 6 – 11 September 1993 modeling episode
between the scenario with all chlorine chemistry turned on and the same scenario, but with the chlorineozone reaction turned off.
1012 exp(295/T ), and kClO,HO2 (T ) = 4.6 1013 exp(710/
T ) cm3 molecule1 s1, respectively [Atkinson et al., 2000].
2.2. Sensitivity Analyses Using SAPRC-99
[49] To help determine which chlorine radical reactions
contribute significantly to urban chlorine radical chemistry
and should be included in the regional photochemical
model, SAPRC-99 [Carter, 2000] was used to perform
sensitivity analyses. SAPRC was used in box model mode
with two different sets of conditions. The results of the
sensitivity runs indicate that the omitted reactions are not
important to the production of ozone under conditions
encountered within the regional photochemical model
domain.
[50] The SAPRC-99 mechanism has assignments for
approximately 400 types of VOCs and has been evaluated
against the results of approximately 1700 environmental
chamber experiments [Carter, 2000]. SAPRC-99 was modified to include chlorine radical chemistry based on the
mechanism incorporated into earlier versions of SAPRC
by Carter et al. [1997a, 1997b]. The chlorine radical
chemistry incorporated into SAPRC-99 includes the reactions of chlorine radicals with NO, NO2, ozone, nitrate
radicals, hydroperoxy radicals, and various organic species.
Table 4 contains the list of reactions between chlorine and
inorganic compounds employed to represent chlorine chemistry in the modified SAPRC-99 mechanism. The reactions
of organic compounds with chlorine were adapted from
Carter et al. [1997a, 1997b], and are not listed here. The
product stoichiometries of reactions between chlorine and
Figure 5. Sensitivity of maximum ozone levels to the
omission of all key Cl-inorganic reactions not included in
CB-IV under conditions representative of those encountered
in the regional photochemical model (0% VOC reduction,
0% NOx reduction). The illustrated difference (0.6 to 0
ppbv ozone) is obtained by subtracting the simulation values
for the case where all Cl-inorganic reactions not included in
the proposed CB-IV mechanism have been turned off in
SAPRC, from the simulation values of the case where all
Cl-inorganic reactions in SAPRC have been left on. The
comparison was made for various reductions in VOC and/or
NOx to simulate the sensitivity under different VOC-NOx
ratios.
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
Figure 6. Sensitivity of ozone production to differences in
how the chlorine-ethene reaction is represented. The
illustrated difference (2 to 2 ppbv) is obtained by
subtracting the maximum ozone levels predicted while
employing the proposed mechanism, from the maximum
ozone levels predicted while employing the more detailed
mechanism. The more detailed mechanism involved the
formation of a chloroacetaldehyde intermediate and its
subsequent oxidation. The comparison was made for
various reductions in VOC and/or NOx to estimate the
sensitivity under different VOC-NOx ratios.
alkanes with odd numbers of carbons that were not specified
by Carter et al. [1997a, 1997b] were based on the corresponding alkane reactions with OH. Reaction rate constants
for all reactions were updated with various references
[Atkinson, 1997; Atkinson et al., 1997, 1999, 2000; Coquet
and Ariya, 2000; DeMore et al., 1997].
2.2.1. Omission of Inorganic Reactions
[51] The first set of box model simulations employed a
relatively simple mixture of Cl2, pentane, ozone, NO, and
NO2 in a simulated environmental chamber. The initial
mixing ratios (in ppb) of Cl2, ozone, pentane, NO, and
NO2 were 5, 0.27, 68.9, 46, and 6, respectively. The
inherent wall reactivity of the environmental chambers
was kept constant.
[52] During each simulation, one of the several key
reactions was omitted from the mechanism. For example,
the omission of reaction (R2) in Table 4 eliminated the
production of ClNO, and in turn, resulted in the effective
removal of subsequent reaction (R3) with ClNO as a
reactant, from the overall mechanism. The key reactions
that were omitted in turn were reactions (R2), (R4), (R5),
(R8), (R9), (R10), and/or (R11).
[53] As each key chlorine-inorganic reaction was omitted
from consideration in the first set of box model simulations, the output from each run was compared to the run
with all chlorine reactions active (referred to here as the
‘‘base case’’). All of these sensitivity runs resulted in no
significant difference in model response. As shown in
Figure 3, the elimination of all chlorine radical reactions
with inorganic species did not result in a noticeable difference in predicted ozone levels. However, if all chlorine
radical reactions were retained except for the pentanechlorine reaction, the ozone levels were underpredicted
ACH
6-9
by 12 ppb by the end of the simulation. This result
suggested that none of the chlorine-inorganic reactions were
important to representing the major aspects of the ozone
formation process. These results were used as the basis for
omitting the reactions between chlorine and the inorganic
reactions stated above. The chlorine-ozone reaction and
subsequent ClO-HO2 and ClO-NO reactions were included
for completeness.
[54] A simulation was performed using CAMx in which
the chlorine-ozone reaction was omitted. The maximum
difference between the regional photochemical modeling
simulation with and without the chlorine-ozone reaction is
plotted in Figure 4. As shown, the difference is negligible
with maximum predicted differences in ozone levels of less
than 0.08 ppbv across the modeled domain.
[55] A second set of sensitivity analyses were performed
using a more complex box model that was developed after
the regional photochemical modeling work was completed.
This new box model not only employed the SAPRC-99
chemical mechanism, but also contained a complex mixture
of hydrocarbons and NOx representative of typical conditions in the Houston area. Representative emissions were
added to a 1 km2 box, while wall effects were omitted and
boundary layer heights were allowed to vary diurnally.
Additional details about this box model are presented by
Y. Kimura et al. (Reactivity of volatile organic compounds
released in industrial process upset events in southeast
Texas, submitted to Environmental Science and Technology,
2002.)
[56] Again, each key chlorine-inorganic reaction was
omitted from consideration. As shown in Figure 5, no effect
was observed on ozone productivity when all chlorineinorganic reactions were omitted. It should be noted that
at extremely elevated chlorine emission levels (20 times
the largest point source in the photochemical model),
Figure 7. Sensitivity of ozone production to differences in
the rate constant for the Cl + OLE reaction. The illustrated
difference (0.6 to 0.4 ppbv ozone) is obtained by
subtracting the maximum levels of ozone predicted when
the employed rate constant is kCl+OLE = 11kOH+OLE from the
corresponding maximum levels of ozone predicted when
the employed rate constant is kCl+OLE = 20kOH+OLE. The
comparison was made for various reductions in VOC and/or
NOx to estimate the sensitivity under different VOC-NOx
ratios.
ACH
6 - 10
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
Figure 8. Maximum difference in ozone levels during the 6 – 11 September 1993 modeling episode.
The difference was obtained as the maximum of the difference between the scenario that included
chlorine chemistry and the scenario that did not include chlorine chemistry. The maximum difference was
predicted for 0800 hours on 11 September 1993.
omission of chlorine nitrate chemistry (reactions (R13) –
(R16) in Table 4) led to an overprediction of ozone
productivity.
[57] The results of this analysis suggest that chlorine
radical chemistry in urban atmospheres such as Houston
is, to a first approximation, adequately described by the
reactions of chlorine radicals with organic species only.
However, for chlorine emission rates significantly above
1 kg h1 km2, chlorine nitrate chemistry may become
important. Future work will examine more closely the
importance of chlorine nitrate (ClONO2) to the overall
production of ozone from chlorine radical chemistry.
2.2.2. Chlorine-Ethene Reaction
[58] As described in section 2.1.4, the proposed Cl +
ethene reaction is not explicitly correct. Rather than producing chloroacetaldehyde as suggested by Wallington et al.
[1990] and Yarwood et al. [1992], the products of the
proposed reaction include formaldehyde and formyl chloride (HCOCl). Formaldehyde is known to undergo oxidation
to form CO2, HO2, and convert NO to NO2. According to
Balestra-Garcia et al. [1992], the primary oxidation pathway for chloroacetaldehyde is through hydrogen abstraction
from the aldehyde group. The product species may then
undergo rapid oxidation to form products such as formyl
chloride and CO2. When the final oxidation products of the
chlorine-ethene reaction via chloroacetaldehyde are compared with the final oxidation products of the proposed
mechanism, similar product stoichiometry is obtained.
[59] Box model comparisons were made between the
proposed mechanism and a more detailed mechanism that
employed a chloroacetaldehyde intermediate. The goal was
to determine the significance of employing the simpler,
proposed reaction as opposed to a more complex represen-
tation that produced chloroacetaldehyde as an intermediate
product. The initial conditions were typical for the Houston
urban area around La Porte (southeast of Houston). The
comparison was made over a range of various VOC/NOx
emissions ratios. As shown in Figure 6, the predicted
difference in maximum ozone level for the two chlorineethene mechanisms was less than 1.3 ppbv. The maximum
ozone level corresponding to this difference was 88 ppbv.
This level of uncertainty was considered acceptable for our
development purposes.
2.2.3. Cl + OLE Multiplier
[60] As decribed in section 2.1.4, the rate constant for the
Cl + OLE reaction was estimated to be 20 times that of the
OH + OLE reaction. This factor of 20 may be high due to
the inclusion of 1,2-propadiene and 1-pentene in calculating
the multiplier (see Table 3). To estimate the difference in
ozone productivity associated with using a factor of 20
instead of 11 (the average using 1-butene and propene), a
sensitivity analysis was performed using the SAPRC-99 box
model. A comparison was made of the maximum ozone
level obtained when a factor of 20 and a factor of 11 were
employed. As shown in Figure 7, the predicted difference in
maximum ozone level for the two different rate constants
was insignificant.
3. Implications for Predicted Regional Ozone
Formation
[61] A preliminary view at the effect of chlorine chemistry on ozone formation in the Houston area during the 6 –
11 September 1993 modeling episode is summarized in the
data plotted in Figures 8 – 10. Details of the chlorine
emissions inventory are presented elsewhere [Chang et
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
ACH
6 - 11
Figure 9. Maximum difference in ozone levels during the 6 – 11 September 1993 modeling episode.
The maximum difference is shown between the scenarios with the chlorine-methane chemistry turned on
and with all chlorine chemistry turned off.
al., 2001, 2002]. Figure 8 shows the maximum predicted
difference in ozone levels between the simulation with
chlorine chemistry and without chlorine chemistry. Figure 9
shows the maximum predicted difference in ozone levels
when the photolysis of Cl precursor and the chlorinemethane reaction are the only reactions, relative to when no
chlorine chemistry is accounted for. The chlorine-methane
reaction was targeted here because of the availability of
Figure 10. Predicted mixing ratios of 1-chloro-3-methyl-3-butene-2-one at the time of its maximum
during the 6 – 11 September 1993 modeling episode. The maximum prediction for the modeled episode
was 59 ppt 1-chloro-3-methyl-3-butene-2-one at 0900 hours on 11 September 1993. The inset is a time
series of the predicted 1-chloro-3-methyl-3-butene-2-one mixing ratio for 11 September 1993 at the
location where the maximum is predicted for this episode.
ACH
6 - 12
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
methane on the global scale at mixing ratios greater than 1.5
ppm (106).
[62] As evident in Figure 8, the peak enhancement was
16 ppbv at 0800 hours. In general, during the modeled
episode, ozone enhancement was predicted to be greatest
during the morning hours when chlorine atoms were most
available and hydroxyl radicals were least available.
Although the 16 ppbv ozone enhancement is not very large
compared to the daily maximum (101 ppbv) predicted for
11 September, the predicted ozone level at 0800 hours
without considering chlorine chemistry was 49 ppbv, or a
difference of 30%.
[63] Although maximum enhancement is localized, significant ozone enhancement is predicted for large areas in
the modeled domain. When only the chlorine-methane
reaction is considered, the local maximum ozone enhancement is also approximately 16 ppbv (Figure 9) with a similar
spatial distribution of the maxima. The similarity in spatial
distribution of the maxima in Figures 8 and 9 is due to the
rapid reaction of Cl near precursor emission sources.
[64] Another indicator of where Cl reactions are occurring is the mixing ratio of 1-chloro-3-methyl-3-butene-2one. As previously stated, this species is unique to the
reaction of Cl with isoprene. Because it readily reacts with
OH, elevated levels of 1-chloro-3-methyl-3-butene-2-one
would be expected near emission sources of Cl precursors
when isoprene is present.
[65] Figure 10 shows the mixing ratios of 1-chloro-3methyl-3-butene-2-one across the modeled domain at the
time corresponding to the maximum predicted 1-chloro-3methyl-3-butene-2-one for the entire modeled episode. As
shown in Figure 10, the predicted peak 1-chloro-3-methyl3-butene-2-one mixing ratio is 59 ppt, at 0900 hours on 11
September 1993, near an emission source of Cl precursor.
The time series is provided for the location where the
maximum 1-chloro-3-methyl-3-butene-2-one mixing ratio
was predicted. Ambient measurements of 1-chloro-3methyl-3-butene-2-one performed by Riemer [2001] indicate that daily peak mixing ratios can range from 12 to 126
ppt in the Houston area, with peaks occurring mostly during
the morning hours. Although a quantitative comparison
cannot be made between the 1-chloro-3-methyl-3-butene2-one mixing ratios predicted by the model for 11 September 1993 and those collected in the summer of 2000, the
mixing ratios are expected to be qualitatively similar
because the meteorology and emissions of isoprene and
chlorine are similar for the two periods. A more detailed
analysis of the modeled episode will be presented in a
parallel publication [Chang et al., 2002]. However, based
on the results presented here, chlorine chemistry should be
further evaluated for its role in enhancing ozone formation
in the vicinity of sources of Cl precursors.
[66] As we have shown through the preliminary evaluation presented here, chlorine chemistry can enhance ozone
formation in the Houston-Galveston area by up to 16 ppbv.
The comparable enhancements shown in the two scenarios
(i.e., when all chlorine reactions are included and when only
the chlorine-methane reaction is included) would seem to
suggest that the contribution to ozone enhancement by
chlorine is not dependent on the availability of nonmethane
hydrocarbons. Because methane is ubiquitous at high levels
relative to other hydrocarbons in the troposphere, the impact
of chlorine chemistry would be expected to follow in
proportion to the amount of Cl precursor available.
[67] By employing the proposed chlorine chemistry in
subsequent regional photochemical simulations, it should be
possible to provide refined predictions of ozone formation
in the Houston-Galveston area. Although chlorine chemistry
is not currently accounted for, our regional photochemical
modeling suggests that chlorine radicals may play an
important role in ozone formation in the urban troposphere.
Additional simulations would allow for evaluation of control of chlorine emissions as an alternative to expensive NOx
controls.
[68] Acknowledgments. This work has been supported by the Texas
Commission on Environmental Quality with additional funding from the
State of Texas as part of the Texas Air Research Center under project
129UTA0007A. The contents do not necessarily reflect the views and
policies of the USEPA nor does the mention of trade names or commercial
products constitute endorsement or recommendation for use. The authors
thank William Carter and Matthew Russell for many helpful discussions.
C.B.M. acknowledges the Welch Foundation for their support via grant F1436. P.L.T. acknowledges the USEPA for their support through the STAR
fellowship program.
References
Atkinson, R., Gas-phase tropospheric chemistry of volatile organic compounds, part 1, Alkanes and alkenes, J. Phys. Chem. Ref. Data, 26, 215 –
290, 1997.
Atkinson, R., D. L. Baulch, R. A. Cox, R. F. Hampson Jr., J. A. Kerr, M. J.
Rossi, and J. Troe, Evaluated kinetic, photochemical and heterogeneous
data for atmospheric chemistry: IUPAC subcommittee on gas kinetic data
evaluation for atmospheric chemistry, J. Phys. Chem. Ref. Data, 26,
suppl. V, 521 – 1013, 1997.
Atkinson, R., D. L. Baulch, R. A. Cox, R. F. Hampson Jr., J. A. Kerr, M. J.
Rossi, and J. Troe, Evaluated kinetic and photochemical data for atmospheric chemistry, organic species, J. Phys. Chem. Ref. Data, 28, suppl.
VII, 191 – 393, 1999.
Atkinson, R., D. L. Baulch, R. A. Cox, R. F. Hampson Jr., J. A. Kerr, M. J.
Rossi, and J. Troe, Evaluated kinetic and photochemical data for atmospheric chemistry: Halogen species IUPAC subcommittee on gas kinetic
data evaluation for atmospheric chemistry, J. Phys. Chem. Ref. Data, 29,
suppl. VIII, 167 – 266, 2000.
Balestra-Garcia, C., G. Le Bras, and H. Mac Leod, Kinetic study of the
reactions OH + mono-, di-, and trichloroacetaldehyde and acetaldehyde
by laser photolysis-resonance fluorescence at 298 K, J. Phys. Chem., 96,
3312 – 3316, 1992.
Bergin, M. S., A. G. Russell, W. P. L. Carter, B. E. Croes, and J. H.
Seinfeld, Ozone control and VOC reactivity, in Encyclopedia of Environmental Analysis and Remediation, p. 3355, John Wiley, New York, 1998.
Calvert, J. G., and J. J. N. Pitts, Photochemistry, 899 pp., John Wiley, New
York, 1966.
Carter, W. P. L., Condensed atmospheric photooxidation mechanisms for
isoprene, Atmos. Environ., 30, 4275 – 4290, 1996.
Carter, W. P. L., Documentation of the SAPRC-99 chemical mechanism for
VOC reactivity assessment: Final report to California Air Resources
Board, contract 92-329 and 95-308, p. 213 + appendices, APRC and
CERT, UC Riverside, Riverside, Calif., 2000.
Carter, W. P. L., D. Luo, and I. L. Malkina, Investigation of the atmospheric
ozone formation potential of trichloroethylene: Draft report to the halogenated solvents industry alliance, p. 47 + appendices, Coll. of Eng.,
Cent. for Environ. Res. and Technol., Univ. of Calif., Riverside, 1997a.
Carter, W. P. L., D. Luo, and I. L. Malkina, Investigation of the atmospheric
reactions of chloropicrin, Atmos. Environ., 31, 1425 – 1439, 1997b.
Chang, S., P. Tanaka, E. McDonald-Buller, and D. Allen, Emission inventory for atomic chlorine precursors in southeast Texas: Report, contract
98800776600-18, Univ. of Tex., Tex. Nat. Res. Conserv. Comm., Univ.
of Tex. at Austin, 2001.
Chang, S., E. McDonald-Buller, Y. Kimura, G. Yarwood, J. Neece,
M. Russell, P. Tanaka, and D. Allen, Sensitivity of urban ozone formation
to chlorine emission estimates, Atmos. Environ., 36, 4991 – 5003, 2002.
Coquet, S., and P. Ariya, Kinetics of the gas-phase reactions of Cl atom
with selected C2 – C5 unsaturated hydrocarbons at 283 < T < 323 K, Int.
J. Chem. Kinet., 32, 478, 2000.
DeMore, W. B., et al., Chemical kinetics and photochemical data for use in
stratospheric modeling: Evaluation 12, in NASA Evaluation JPL 97-4, Jet
Propul. Lab., Pasadena, Calif., 1997.
TANAKA ET AL.: CHLORINE MECHANISM IN CARBON BOND IV
ENVIRON, User’s Guide, Comprehensive Air Quality Model with Extensions (CAMx), ENVIRON International Corporation, 101 Rowland Way,
Novato, Calif., 2002.
Finlayson-Pitts, B. J., and J. J. N. Pitts, Chemistry of the Upper and Lower
Atmosphere, 969 pp., Academic, San Diego, Calif., 2000.
Gery, M. W., G. Z. Whitten, J. P. Killus, and M. C. Dodge, A photochemical kinetics mechanism for urban and regional scale computer modeling, J. Geophys. Res., 94(D10), 12,925 – 12,956, 1989.
Hov, O., The effect of chlorine on the formation of photochemical oxidants
in southern Telemark, Norway, Atmos. Environ., 19, 471 – 485, 1985.
Liang, J., and M. Z. Jacobson, Comparison of a 4000-reaction chemical
mechanism with the carbon bond IV and an adjusted carbon bond IV-EX
mechanism using SMVGEAR II, Atmos. Environ., 34, 3015 – 3026, 2000.
Madronich, S., and S. Flocke, The role of solar radiation in atmospheric
chemistry, in Handbook of Environmental Chemistry, edited by P. Boule,
pp. 1 – 26, Springer-Verlag, New York, 1998.
Molina, M. J., and F. S. Rowland, Stratospheric sink for chlorofluoromethanes. Chlorine atom-catalyzed destruction of ozone, Nature,
249(5460), 810 – 812, 1974.
Nordmeyer, T., W. Wang, M. L. Ragains, B. J. Finlayson-Pitts, C. W.
Spicer, and R. A. Plastridge, Unique products of the reaction of isoprene
with atomic chlorine: Potential markers of chlorine atom chemistry, Geophys. Res. Lett., 24(13), 1615 – 1618, 1997.
Orlando, J. J., G. S. Tyndall, M. Bilde, C. Ferronato, T. J. Wallington,
L. Vereecken, and J. Peeters, Laboratory and theoretical study of the
oxy radicals in the OH- and Cl-initiated oxidation of ethene, J. Phys.
Chem. A, 102, 8116 – 8123, 1998.
Orlando, J. J., G. S. Tyndall, and S. E. Paulson, Mechanism of the OHinitiated oxidation of methacrolein, Geophys. Res. Lett., 26(14), 2191 –
2194, 1999.
Oum, K. W., M. J. Lakin, D. O. DeHaan, T. Brauers, and B. J. FinlaysonPitts, Formation of molecular chlorine from the photolysis of ozone and
aqueous sea-salt particles, Science, 279(5347), 74 – 77, 1998.
Ragains, M. L., and B. J. Finlayson-Pitts, Kinetics and mechanism of the
reaction of Cl atoms with 2-methyl-1,3-butadiene (isoprene) at 298 K,
J. Phys. Chem. A, 101, 1509 – 1517, 1997.
Riemer, D. D., Final report to the Texas Natural Resource Conservation
Commission: Confirming the presence and extent of oxidation by Cl in
the Houston, Texas urban area using specific isoprene oxidation products
as tracers, report, Univ. of Miami, Fla., 2001.
ACH
6 - 13
Seinfeld, J. H., and S. N. Pandis, Atmospheric Chemistry and Physics:
From Air Pollution to Climate Change, 1326 pp., John Wiley, New York,
1998.
Simonaitis, R., J. F. Meagher, and E. M. Bailey, Evaluation of the condensed carbon bond (CB-IV) mechanism against smog chamber data at
low VOC and NOx concentrations, Atmos. Environ., 31, 27 – 43, 1997.
Tanaka, P. L., S. Oldfield, J. D. Neece, C. B. Mullins, and D. T. Allen,
Anthropogenic sources of chlorine and ozone formation in urban atmospheres, Environ. Sci. Technol., 34(21), 4470 – 4473, 2000.
Tanaka, P. L., et al., Direct evidence for chlorine-enhanced urban ozone
formation in Houston, Texas, Atmos. Environ., in press, 2003.
Texas Register, Control of Air Pollution from Nitrogen Compounds, Tex.
Regist., 25(34), 8287 – 8293, 2000.
Wallington, T. J., J. M. Andino, and S. M. Japar, FTIR product study of the
self-reaction of CH2ClCH2O2 radicals in air at 295 K, Chem. Phys. Lett.,
165(2), 189 – 194, 1990.
Wang, W., and B. J. Finlayson-Pitts, 4-Chlorocrotonaldehyde as a unique
chlorine-containing compound from the reaction of atomic chlorine with
1,3-butadiene in air at room temperature, Geophys. Res. Lett., 27(7),
947 – 950, 2000.
Wang, W., M. J. Ezell, A. A. Ezell, G. Soskin, and B. J. Finlayson-Pitts,
Rate constants for the reactions of chlorine atoms with a series of unsaturated aldehydes and ketones at 298 K: Structure and reactivity, Phys.
Chem. Chem. Phys., 4, 1824 – 1831, 2002.
Yarwood, G., N. Peng, and H. Niki, FTIR spectroscopic study of the Cland Br-atom initiated oxidation of ethene, Int. J. Chem. Kinet., 24, 369 –
383, 1992.
D. T. Allen, S. Chang, Y. Kimura, E. C. McDonald-Buller, and P. L.
Tanaka, Center for Energy and Environmental Resources and Department
of Chemical Engineering, University of Texas at Austin, Austin, TX 787121062, USA. ([email protected])
C. B. Mullins, Department of Chemical Engineering, University of Texas
at Austin, MC C0400, 1 University Station, Austin, TX 78712, USA.
J. D. Neece, Texas Commission on Environmental Quality, P. O. Box
13087, Austin, TX 78711, USA.
G. Yarwood, ENVIRON International Corporation, 101 Rowland Way,
Novato, CA 94945, USA.