Rate constants for the reactions of methyl vinyl ketone, methacrolein, methacrylic acid, and acrylic acid with ozone

Article abstract of DOI:10.1002/(SICI)1097-4601(1998)30:10<769::AID-KIN9>3.0.CO;2-T

Rate constants for the reaction of ozone with methylvinyl ketone (H₂C=CHC(O)CH₃), methacrolein (H₂C=C(CH₃)CHO), methacrylic acid (H₂C=C(CH₃)C(O)OH), and acrylic acid (H₂C=CHC(O)OH) were measured at room temperature (296 ± 2 K) in the presence of a sufficient amount of cyclohexane to scavenge OH-radicals. Results from pseudo-first-order experiments in the presence of excess ozone were found not to be consistent with relative rate measurements. It appeared that the formation of the so-called Criegee-intermediates leads to an enhanced decrease in the concentration of the two organic acids investigated. It is shown that the presence of formic acid, which is known to react efficiently with Criegee-intermediates, diminishes the observed removal rate of the organic acids. The rate constant for the reaction of ozone with the unsaturated carbonyl compounds methylvinyl ketone and methacrolein was found not to be influenced by the addition of formic acid. Rate constants for the reaction of ozone determined in the presence of excess formic acid are (in cm³ molecule^-1 s^-1): methylvinyl ketone (5.4 ± 0.6) × 10^-18; methacrolein (1.3 ± 0.14) × 10^-18; methacrylic acid (4.1 ± 0.4) × 10^-18; and acrylic acid (0.65 ± 0.13) × 10^-18. Results are found to be consistent with the Criegee mechanism of the gas-phase ozonolysis.

Full text of DOI:10.1002/(SICI)1097-4601(1998)30:10<769::AID-KIN9>3.0.CO;2-T

Rate Constants for the
Reactions of Methylvinyl
Ketone, Methacrolein,
Methacrylic Acid, and
Acrylic Acid with Ozone
PETER NEEB, ANTJE KOLLOFF, STEPHAN KOCH, GEERT K. MOORTGAT
Max-Planck-Institut f u¨ r Chemie, Division of Atmospheric Chemistry, P.O. Box 3060, D-55020 Mainz, Germany
Received 14 November 1997; accepted 12 May 1998
ABSTRACT: Rate constants for the reaction of ozone with methylvinyl ketone (H C"
2
CHC(O)CH ), methacrolein (H C"C(CH )CHO), methacrylic acid (H C"C(CH )C(O)OH), and
3
2
3
2
3
acrylic acid (H C"CHC(O)OH) were measured at room temperature (296 Ϯ 2 K) in the pres-
2
ence of a sufficient amount of cyclohexane to scavenge OH-radicals. Results from pseudo-
first-order experiments in the presence of excess ozone were found not to be consistent with
relative rate measurements. It appeared that the formation of the so-called Criegee-interme-
diates leads to an enhanced decrease in the concentration of the two organic acids investi-
gated. It is shown that the presence of formic acid, which is known to react efficiently with
Criegee-intermediates, diminishes the observed removal rate of the organic acids. The rate
constant for the reaction of ozone with the unsaturated carbonyl compounds methylvinyl
ketone and methacrolein was found not to be influenced by the addition of formic acid. Rate
constants for the reaction of ozone determined in the presence of excess formic acid are (in
cm molecule s ): methylvinyl ketone (5.4 Ϯ 0.6) ϫ 10 18; methacrolein (1.3 Ϯ 0.14) ϫ
3
Ϫ
1
Ϫ
1
Ϫ
1
0
Ϫ18; methacrylic acid (4.1 Ϯ 0.4) ϫ 10Ϫ18; and acrylic acid (0.65 Ϯ 0.13) ϫ 10Ϫ18. Results are
found to be consistent with the Criegee mechanism of the gas-phase ozonolysis. ᭧ 1998 John
Wiley & Sons, Inc. Int J Chem Kinet 30: 769–776, 1998
INTRODUCTION
of the gas-phase reactions of organic compounds have
recently been comprehensively reviewed by Atkin-
son [1].
In recent years rate constants for a large number of
hydrocarbons and oxygenated hydrocarbons with OH,
O , and NO have been reported in the literature. These
Among the various species known to be constitu-
ents of the troposphere, isoprene, and monoterpenes
emitted from plants are thought to represent the most
important contributors to nonmethane-hydrocarbon
(NMHC) emissions [2]. These compounds, containing
one or more double bonds, are rapidly oxidized in the
troposphere and serve as an important source of par-
tially oxidized compounds such as CO, aldehydes, ke-
tones, and organic acids.
3
3
data provide valuable information to assess the tro-
pospheric lifetimes and the relative importance of the
various degradation pathways of these compounds
which are emitted either by vegetation or by anthro-
progenic activities into the atmosphere. The kinetics
Correspondence to: P. Neeb
᭧
1998 John Wiley & Sons, Inc.
CCC 0538-8066/98/100769-08
Since isoprene is the most important emitted com-

7
70
NEEB ET AL.
pound, investigations have focused on the tropo-
spheric degradation of isoprene and its oxidation prod-
ucts. Rate constants for the various degradation
pathways of isoprene, methylvinyl ketone (MVK), and
methacrolein (MAC) are known. No rate constant has
been reported for methacrylic acid (MAA) which has
been proposed as a reaction product of the isoprene
oxidation [3,4] but has not been reported in field mea-
surements [5]. Since MAA contains a double bond it
is expected to have a short tropospheric lifetime yield-
ing presumably pyruvic acid as a major oxidation
product [4,6].
MAA and ACA by volumetric methods were merely
used to determine the amount of ozone needed to ob-
tain pseudo-first-order conditions and the evaluation
of the data was therefore not affected by the accuracy
of the calibration factors.
After filling the reactor to about 670 Torr with syn-
thetic air, ozone was produced by the photolysis of
oxygen at 185 nm using a Penray UV-lamp mounted
inside the reactor as source of light. After approxi-
mately 10 ppmv O3 was admitted to the reactor,
cyclohexane was added as an OH scavenger in all
experiments. A mixing ratio of 54–65 ppmv cy-
clohexane was sufficient to scavenge more than 90%
of the OH-radicals formed during the ozonolysis [12].
In some experiments HCOOH was added as a scav-
enger for the Criegee-intermediates. Prior to the ad-
dition of the alkene, FTIR spectra were recorded to
determine the exact amount of ozone, c-C H , and
In this work we have determined the rate constants
for the gas-phase reactions of MVK, MAC, MAA, and
acrylic acid (ACA) with O . To the best of our knowl-
3
edge this is the first time that rate constants for the
reaction of unsaturated organic acids with ozone are
reported. Because previous studies have shown that
6
12
OH radicals are generally formed in the O reaction
HCOOH, if present, in the reaction chamber.
3
with alkenes [1], experiments were carried out in the
presence of cyclohexane to minimize alkene con-
sumption through the alkene-OH reaction. In pseudo-
first-order experiments with excess ozone concentra-
tion the observed decay rates for MAA and ACA were
found to be influenced by the presence of HCOOH.
The experimental results of this study are explained
by the reaction of the Criegee-intermediates formed in
the ozonolysis of MAA and ACA.
The decrease of the alkene concentration was mon-
itored by FTIR spectroscopy using an MCT detector
Ϫ1
at the spectral resolution of 0.5 cm . 64 or 128 indi-
vidual scans were coadded for each spectrum resulting
in a temporal resolution of 65 or 120 s, respectively.
To evaluate the absorbance-time profiles only the first
8–10 datapoints, corresponding to an alkene con-
sumption of 20–80%, were used for the linear regres-
sion. The systematic error due to the assumption of
first-order-conditions and the formation of OH-radi-
cals during ozonolysis were estimated by numerical
simulation of a simplified reaction scheme using the
FACSIMILE program [13]. Treating the simulated ab-
sorbance-time profiles in similar manner as the exper-
imental data, the evaluated rate constant showed a de-
viation of less than 2% compared to the rate constant
used in the simulation.
EXPERIMENTAL
Experiments were carried out in an evacuable, 570 l
spherical glass reactor in 730 Ϯ 5 torr (1 torr ϭ
1
33.3224 Pa) synthetic air. The reaction temperature
was kept constant at 296 Ϯ 2 K by the laboratory air
conditioner. Details on the experimental design of the
reactor have been given elsewhere [7,8] and only a
brief description will be given here.
Chemicals were used as purchased commercially:
MVK (Aldrich 99.5%), MAC (Aldrich 99%), ACA
(Aldrich 99%) and MAA (Fluka 99.5%).
The initial mixing ratios of the organic acids and
carbonyl compounds were 0.8–1.2 ppmv (1 ppmv ϭ
1
3
Ϫ3
2
.38 ϫ 10 molecule cm at the above temperature
RESULTS AND DISCUSSION
Methacrylic Acid
and pressure) as determined by standard volumetric
methods. The IR-absorption coefficient of ozone at
Ϫ
1
1
034 cm was determined by comparison with the
Rate constants for the alkene-ozone reactions were de-
termined by employing the pseudo-first-order tech-
nique, namely by measuring the decay of the alkene
absorbance in the presence of excess ozone. With an
ozone concentration in excess of the concentration of
MAA, the decay rate of MAA should obey the pseudo-
first-order expression:
known UV-absorbance at 253 nm [9]. MAC, MVK,
and cyclohexane were calibrated using standard vol-
umetric methods. Due to the formation of organic acid
dimers [10] in the gas-phase, calibrations of organic
acids based on volumetric methods inherently under-
estimate their concentrations. Formic acid was there-
fore calibrated using an in situ pyrolytic method [11].
No attempts have been made to avoid possible uncer-
tainties due to the formation of MAA- and ACA-di-
mers in the gas-phase. Calibration factors obtained for
dln[MAA]
Ϫ
ϭ k2eff[O ] ϭ k
(1)
3
obs
dt

RATE CONSTANTS FOR REACTIONS
771
resulted in a clearly noticeable decrease in the ob-
served removal rate.
The formation of hydroperoxymethyl formate
which is known to be formed in the reaction of the
CH OO intermediate (reaction 5) with HCOOH [14]
2
was also observed together with a decrease in the
HCOOH concentration.
H COO ϩ HCOOH !:
2
H C(OOH)9O9CHO (5)
2
To further investigate the influence of HCOOH on
kobs, a series of experiments with constant initial con-
centrations of MAA and ozone and varying amounts
of added HCOOH were performed. Rate constants
k2eff obtained from the individual experiments are plot-
ted against the HCOOH initial concentration in Figure
Figure 1 Semilogarithmic plot of the decrease of the char-
acteristic absorbance of MAA at 1124 cm in pseudo-first-
order experiments. Initial concentration of the reactants were
Ϫ1
[
O ] ϭ 10 ppmv, [MAA] Ϸ 1 ppmv and [C H ] ϭ
3
0
0
6
12 0
5
4 ppmv. Filled circles represent the data in the absence of
2. It appears that the effective rate constants are con-
HCOOH and crosses those in the presence of 19 ppmv
HCOOH. Solid lines are linear regressions, considering the
first eight data points.
stant under conditions where HCOOH was added in
excess. Under these conditions it can be assumed that
all the thermalized Criegee-intermediates have reacted
with HCOOH and hence the loss of MAA due to re-
action (4) can be neglected. At sufficient high HCOOH
where k2eff is the effective second-order rate constant
concentrations the value of k2eff Ϸ k Ϸ (4.2 Ϯ
2
for the reaction of O with MAA.
Ϫ18
3
Ϫ1
Ϫ1
3
0.2) ϫ 10 cm molecule
s
obtained is therefore
Figure 1 shows a semilogarithmic plot of the IR-
absorbance of MAA, which decayed exponentially in
accordance with eq. (1). From the slope of the semi-
suggested to represent the rate constant of reaction (2).
logarithmic plot, the observed decay rate k was de-
Acrylic Acid
obs
Ϫ3
Ϫ1
termined to be (2.03 Ϯ 0.04) ϫ 10 s , with the er-
ror limit being twice the standard deviation obtained
in the least-squares fitting procedure.
According to the Criegee mechanism the ozon-
olysis of MAA yields two carbonyl compounds
From the results of the experiments with MAA, it was
expected that the decay rates of the reaction of ACA
with ozone exhibit a similar behaviour towards ad-
(HCHO and pyruvic acid) and two Criegee-interme-
diates [4]:
H C"C(CH )COOH ϩ O
3
2
3
!
: [H COO]* ϩ CH C(O)COOH
(2a)
(2b)
2
3
!: HCHO ϩ [CH C(OO)COOH]*
3
It was recently shown that the collisionally thermali-
zed Criegee-intermediate CH OO reacts with hydrox-
2
ylic compounds to form hydroperoxides [7]:
[
H COO]* ϩ M !: H COO ϩ M
(3)
2
2
H COO ϩ ROH !: H C(OOH)9O9R (4)
2
2
Figure 2 Dependence of the effective rate constant k2eff of
the reaction of MAA with ozone on the amount of added
HCOOH. Initial concentration of the reactants were
To test whether MAA is consumed in a reaction anal-
ogous to reaction (4), pseudo-first-order experiments
were repeated in the presence of varying amounts of
HCOOH. In Figure 1 the semilogarithmic plot of the
MAA absorbance in an experiment with 19 ppmv
HCOOH added is shown. The addition of HCOOH
[
O ] ϭ 10 ppmv, [MAA] Ϸ 1 ppmv and [C H ] ϭ 54
3
0
0
6
12 0
ppmv. Symbols represent individual experiments and the
line shows the the fit according to equation 16 (see text).
Error bars (Ϯ2␴) describe the statistical uncertainties from
the linear regressions of the absorbance-time profiles.

7
72
NEEB ET AL.
in inert solvents the secondary ozonides are the main
products [17]:
O
O
R COO ϩ R CO
(7)
2
2
R C
CR2
2
O
In the gas phase, under experimental conditions where
carbonyl compounds are in large excess and in the
absence of alternative reaction partners (e.g., H O), it
2
is expected that the Criegee-intermediate reacts with
carbonyl compounds. With respect to the systematic
errors of the determination of rate constants which
may result from neglecting an additional sink reaction
for one of the reactants, effective rate constants for the
reaction of MVK and MAC with ozone were deter-
mined in the presence of HCOOH.
Figure 3 Semilogarithmic plot of the decrease of the char-
acteristic absorbance of ACA at 1139 cm in pseudo-first-
order experiments. Initial concentration of the reactants were
Ϫ1
[
O ] ϭ 10 ppmv, [ACA] Ϸ 1 ppmv and [C H ] ϭ
3
0
0
6
12 0
5
4 ppmv. Filled circles represent the data in the absence of
H C"CHC(O)CH ϩ O
3
2
3
HCOOH and crosses those in the presence of 19 ppmv
HCOOH. Solid lines are linear regressions, considering the
first ten data points.
!: [H COO]* ϩ HC(O)C(O)CH
(8a)
(8b)
2
3
!
: HCHO ϩ [CH C(O)CHOO]*
3
H C"C(CH )CHO ϩ O
3
2
3
!
: [H COO]* ϩ CH C(O)CHO
(9a)
(9b)
2
3
dition of HCOOH. Expected primary products from
the ozonolysis are HCHO and 2-oxoacetic acid (glyox-
ylic acid, HC(O)C(O)OH) and two Criegee-interme-
diates.
!
: HCHO ϩ [HC(O)C(CH )OO]*
3
The semilogarithmic plots of the IR-absorbances of
Ϫ1
Ϫ1
MVK and MAC (at 953 cm and 933 cm respec-
tively) are shown in Figure 5. In experiments with
added HCOOH, no change in the determined removal
rates was observed (Figs. 5 and 6). From the data
H C"CHCOOH ϩ O
2
3
!
: [H COO]* ϩ HC(O)COOH
(6a)
(6b)
2
!
: HCHO ϩ [HC(OO)COOH]*
shown in Figure 6 the rate constants k ϭ (5.4 Ϯ
8
Figure 3 shows semilogarithmic plots of the IR-ab-
Ϫ1
sorbance of ACA at 1139 cm in pseudo-first-order
experiments in the absence and presence of HCOOH.
As in the case of the MAA-O system, the observed
3
removal rate decreases upon addition of HCOOH. The
dependence of k6eff in the reaction of ACA with ozone
on the initial HCOOH concentration is shown in Fig-
ure 4. The first-order rates kobs showed the same be-
havior as in the case of MAA, although the decrease
was less pronounced. Hence, the experimentally ob-
tained absorbance-time profiles were treated in the
same way as discussed above in order to determine k6.
From the results shown in Figure 4, the value of k
6
3
Ϫ18
was estimated to be (0.73 Ϯ 0.05) ϫ 10 cm
Ϫ1
Ϫ1
molecule s .
Figure 4 Dependence of the effective rate constant k6eff of
the reaction of ACA with ozone on the amount of added
HCOOH. Initial concentration of the reactants were
MVK and MAC
[
O ] ϭ 10 ppmv, [ACA] Ϸ 1 ppmv and [C H ] ϭ 54
3
0
0
6
12 0
Criegee-intermediates are reported not only to react
with hydroxylic compounds but also with carbonyl
compounds in the gas phase [15,16]. This reaction has
been thoroughly investigated in the liquid phase where
ppmv. Symbols represent individual experiments and the
line shows the fit according to equation 16 (see text). Error
bars (Ϯ2␴) describe the statistical uncertainties from the lin-
ear regression only.

RATE CONSTANTS FOR REACTIONS
773
Figure 5 Semilogarithmic plot of the decrease of the char-
Figure 7 Relative rate experiments of MVK and MAA.
Open circles show results obtained from an experiment in
the presence of 19 ppmv HCOOH and filled circles were
obtained under identical conditions but without addition of
HCOOH. The solid line corresponds to the slope which can
Ϫ1
acteristic absorbance of MVK at 953 cm and of MAC at
Ϫ1
9
33 cm in pseudo-first-order experiments. Initial con-
centration of the reactants were [O ] ϭ 10 ppmv,
3
0
[
carbonyl compound] ϭ 1 ppmv and [C H ] ϭ 54–65
0 6 12 0
ppmv. For both MAC and MVK the data of experiments in
the absence of HCOOH (triangles and crosses) and with
be expected based on the absolute rate constants k and k8
2
for the reactions of MAA and MVK with ozone, respec-
tively. Crosses are from an experiment without HCOOH,
where the initial concentration of MVK exceeds that of
MAA by a factor of three.
[
HCOOH] ϭ 19 ppmv (filled and open circles) are shown.
0
Solid lines are the linear regressions lines considering the
first eight data points.
Ϫ
18
3
Ϫ
1
Ϫ
1
0
0
.6) ϫ 10 cm molecule
s
and k9 ϭ (1.3 Ϯ
were determined.
tems do not react with MAC or MVK, respectively,
although a decrease of the HCOOH mixing ratio was
observed. To our knowledge there are no comprehen-
sive product studies of the ozonolysis of MVK or
MAC, so little information concerning possible reac-
tions of the Criegee-intermediates is available. How-
ever, it is known that HCOOH is a product of the
Ϫ18
3
Ϫ
1
Ϫ
1
.14) ϫ 10 cm molecule
s
It was found that the ratio of the consumption of the
unsaturated organic compounds to the consumption of
ozone is unity, indicating that MVK and MAC react
exclusively with ozone in the presence of an OH-scav-
enger. From these results it appears that the Criegee-
intermediates formed in the MAC/MVK-ozone sys-
isomerisation of the H COO Criegee-intermediate[18,
2
19] and may react according to reaction (4). The for-
mation of hydroperoxymethyl formate (HPMF,
CH (OOH)-O-CHO was observed in all experiments,
2
although no attempts were made to quantify this prod-
uct. It was noticed, that more HPMF was formed in
the presence of HCOOH.
Relative Rate Measurements
To test the consistency of the rate constants for MVK
and MAA obtained in pseudo-first-order experiments,
also relative rate measurements were performed. It
was expected that the decrease of MAA would not
only be caused by the reaction with ozone but also
with the thermalized Criegee-intermediate. Relative
rate measurements were therefore repeated in the pres-
Figure 6 Determined rate constants of the reaction of
MVK and MAC with ozone. Symbols represent experiments
performed in the presence of various amounts of added
HCOOH, lines are obtained by linear regression of the data.
Error bars shown (Ϯ2␴) include only the statistical uncer-
tainties from linear regression of the data.
ence of HCOOH and with varying ratios of [MVK]0/
[
MAA] . The results of these experiments are shown
0
in Figure 7. The expression for evaluating relative rate
constants

7
74
NEEB ET AL.
[
MAA]0
k
[MVK]
R R COO ϩ R R C"CRCOOH !:
1 2 1 2
2
ͮ ͩ ͪ ͫ ͭ ͮͬ
0
ln
ϭ
ln
ͭ
products (11)
[
MAA]t
k
8
[MVK]
t
eff
R R COO ϩ HCOOH !: products (12)
1
2
yields in the presence of 19 ppmv HCOOH a ratio of
.84 Ϯ 0.01 for (k /k ) , which is in good agreement
with the value of 0.76 Ϯ 0.09 (1␴) calculated from the
pseudo-first-order experiments in the presence of
HCOOH. In the absence of HCOOH the ratio of
0
2
8 eff
The temporal decrease of the organic acid R R C"
1
2
CRCOOH (hereafter abbrievated as A) can therefore
be described by using reactions (10–12) and the pa-
rameter Y for the extent of stabilization:
(
k /k ) was found to be 2.67. From the comparison
2 8 eff
of the decrease of MVK in the two relative rate ex-
periments, it can be stated that the strong increase in
the ratio (k /k ) in the absence of HCOOH is solely
d[A]
ϭ Ϫk [O ][A] Ϫ k [R R COO][A] (13)
1
0
3
11
1
2
dt
2
8 eff
due to an enhanced removal of MAA. Since the Crie-
gee-intermediates formed in the ozonolysis of MVK
are also expected to react with MAA, the removal rate
of MAA is not only dependent on the degree of sta-
bilization of the Criegee-intermediate but also on the
ratio of the initial concentrations of the two alkenes.
Figure 7 also shows data from an experiment where
the initial MVK concentration was in excess of the
concentration of MAA. In this experiment more than
it follows under the assumption of steady-state con-
ditions for the Criegee-intermediate (d[R R COO]/0:
1
2
d[A]
ϭ Ϫk [O ][A]
1
0
3
dt
2
k k Y[O ][A]
1
0
11
3
Ϫ
(14)
k [A] ϩ k [HCOOH]
1
1
12
9
5% of the MAA was consumed while less than 38%
In experiments without addition of HCOOH one can
assume [HCOOH] ϭ 0 and the equation (14) can be
simplified to:
of the initially present MVK has reacted.
Kinetic Analysis
d[A]
ϭ Ϫk (1 ϩ Y)[O ][A]
The apparent decrease of the effective rate constants
for reactions (2) and (6) in the presence of HCOOH
enables us to determine the degree of stabilization of
the Criegee-intermediates formed. The kinetic descrip-
tion of the decrease in organic acid can be expressed
as follows, where no distinction is made between the
different Criegee-intermediates formed (see eg., reac-
tion (2) and (6)).
10
3
dt
ϭ Ϫk10eff[O ][A] (15)
3
In the presence of excess HCOOH the second part of
eq. (14) can be set to zero and under pseudo-first-order
conditions [A] ϭ [A] exp(Ϫk ϫ t) with kobs
ϭ
0
obs
k [O ]. For the reaction of MAA with ozone, k was
1
0
3
obs
constant within the experimental errors if more than
0 ppmv HCOOH were present (Fig. 3). For the cor-
responding reaction of ACA it appeared that k6eff has
1
R R C"CRCOOH ϩ O !:
1
2
3
[
R R COO]* ϩ RC(O)COOH (10)
1
2
not yet reached the limit of k even at the highest
6
HCOOH concentration employed, resulting in an
overestimation of the rate constant for the reaction
with ozone under these conditions.
[
R R COO]* !: decomposition
(a)
1
2
[
R R COO]* ϩ M !: R R COO ϩ M (b)
1
2
1
2
A more detailed kinetic analysis is available by ex-
trapolating eq. (14) to t : 0.
Defining the degree of stabilization as Y ϭ (k ϫ
b
[
M])/(k ϩ k ϫ [M]), the net reaction can be written
a b
as:
d[A]
ͩdt
ͪ
10 3
t:0
R R C"CRCOOH ϩ O !: Y ϫ R R COO
1
2
3
1
2
Y
ϩ (1 Ϫ Y) ϫ [R R COO]* ϩ RC(O)COOH
1
2
ϭ Ϫk [O ][A] 1 ϩ
(16)
0
ͩ
ͪ
k [HCOOH]
0
1
2
1
ϩ
At initial stages of the reaction (10), where the con-
centration of any product is negligible, the reactions
of the Criegee-intermediate R R COO can be formu-
k11[A]
0
This procedure enables us to determine k , and the
degree of stabilization Y of the Criegee intermediates
1
2
10
lated according to reaction (4):

RATE CONSTANTS FOR REACTIONS
775
R R COO from the experimentally observed rates as
Table I Rate Constants for the Reaction of O with
Methacrylic Acid, Acrylic Acid, Methylvinyl Ketone,
Methacrolein, and Acrolein
3
1
2
a function of HCOOH by least-square fitting to eq.
16).
(
The parameters derived from the fitting procedure
kO3 [ϫ10 cm
Ϫ
18
3
Ϫ
18
3
Ϫ
1
Ϫ
1
Ϫ
1
Ϫ1
were k2 ϭ (4.1 Ϯ 0.4) ϫ 10 cm molecule
s
molecule s ]
Ref.
Ϫ18
3
Ϫ1
Ϫ1
and k ϭ (0.65 Ϯ 0.13) ϫ 10 cm molecule
s
6
a
a
Methacrylic acid
Methacrolein
4.1 Ϯ 0.4
1.3 Ϯ 0.14
1.12 Ϯ 0.2
this work
this work
[21]
for the reaction of MAA and ACA with ozone, re-
spectively. The value obtained for k is in good agree-
2
ment with the value derived directly from experiments
in the presence of more than 10 ppmv HCOOH. For
1
.1 Ϯ 0.2
[22]
1.08 Ϯ 0.2
0.65 Ϯ 0.13
0.28 Ϯ 0.05
0.30 Ϯ 0.04
5.4 Ϯ 0.6
[23]
this work
[21]
[22]
this work
[21]
a
a
ACA the resulting value for k is about 12% lower
Acrylic acid
Acrolein
6
than the experimental value in the presence of
1
9 ppmv HCOOH, indicating that the Criegee-inter-
Methylvinyl ketone
mediate formed in the gas-phase reaction does not
solely react with HCOOH.
Values for the degree of stabilization Y as deter-
mined from eq. (16) are 0.98 Ϯ 0.12 for MAA and
4
.77 Ϯ 0.59
.2 Ϯ 0.4
.84 Ϯ 0.39
4
[22]
[23]
5
a
0
.69 Ϯ 0.16 for ACA. Compared to results of other,
Error limits for the rate constants determined in this study in-
clude a 10% uncertainty in the absolute calibration factor used to
quantify ozone concentrations.
more direct methods for the determination of the de-
gree of stabilization that were performed for various
alkenes, these values seem to be too high [20]. Espe-
cially the calculated degree of stabilization for the
Criegee-intermediate of nearly unity for MAA seems
unrealistic. This discrepancy cannot be explained at
present. Although wall effects cannot be ruled out, an
overestimation of the degree of stabilization may also
result from an additional reaction consuming organic
acids which are not taken into account in the kinetic
analysis given before.
It is unlikely that the OH-radical formed in the
ozonolysis [12] may account for the observed effect.
The rate constants for the reactions of the OH-radical
with the unsaturated organic acids as well as the OH-
yield in the ozonolysis of these compounds have not
been reported, and, therefore, experimental conditions
served rates of the reactions of ozone with MAA and
ACA were strongly influenced by HCOOH, a result
which can be explained on the basis of the Criegee
mechanism.
No such dependency of the effective rate constants
of ozone with MVK and MAC was observed and the
results of this work are in good agreement with data
from the literature [21–23]. Values of the rate con-
stants for the reaction of ozone with the unsaturated
compounds investigated in this study are given in Ta-
ble I. For comparison also data from acrolein are in-
cluded.
In the absence of a scavenger, the thermalized Crie-
gee-intermediate reacts with the unsaturated organic
acid, resulting in an overestimation of the rate constant
with ozone. Together with the recently reported reac-
(eg., amount of cyclohexane) were adjusted on the ba-
sis of the corresponding values of MVK. However
losses of MAA and ACA due to reaction with OH-
radicals would result in an additional increase of keff
without affecting the degree of stabilization Y . There
are some analogous findings from experiments in the
ethene-ozone system in the presence of HCOOH [14].
In these experiments it was observed that the amount
of HCOOH consumed exceeds the expected conver-
sion of HCOOH due to reaction (5), hence, implying
an additional sink of HCOOH.
tion of HCOOH and CH COOH with the Criegee-in-
3
termediate CH OO [7], it can be concluded that the
2
Criegee-intermediates generally react with organic ac-
ids. Although no products were identified in this study,
it is expected that these reactions lead to the formation
of oxygenated hydroperoxides in analogy to reaction
(5):
R R COO ϩ R COOH !:
1
2
3
R R C(OOH)9O9C(O)R
3
(17)
1
2
CONCLUSION
From an analysis of the dependence of the effective
rate constants on the HCOOH concentrations, an es-
timation of the degree of stabilization was obtained
which seems too high to be realistic. The most likely
To date no rate constants have been reported for the
gas-phase reactions of ozone with an unsaturated or-
ganic acid. It was shown that the experimentally ob-

7
76
NEEB ET AL.
reason for this behavior is the neglect of an additional
sink process for organic acids in the simple kinetic
scheme used to derive the degree of stabilization.
Comparing the rate constants in Table I reveals that
the substitution of a carbonyl group by a carboxylic
acid group leads to an increase of the rate constant by
roughly a factor of three. This is in accordance with
the basic concept of the Criegee-mechanism that the
ozonolysis is initiated by an electrophilic cycloaddi-
tion.
8. O. Horie, P. Neeb, and G. K. Moortgat, Int. J. Chem.
Kin., 29, 461 (1997).
9
. W. B. DeMore, S. P. Sander, C. J. Howard, A. R. Ravi-
shankara, D. M. Golden, C. E. Kolb, R. F. Hampson,
M. J. Kurylo, and M. J. Molina, JPL Publication 97-4,
1
997.
1
1
0. R. B u¨ ttner and G. Maurer, Ber. Bunsenges. Phys.
Chem., 87, 877 (1983).
1. M. Finkbeiner, P. Neeb, O. Horie, and G. K. Moortgat,
Fresenius J. Anal. Chem., 351, 521 (1995).
12. R. Atkinson, S. M. Aschmann, J. Arey, and B. Shorees,
J. Geophys. Res., 97, 6065 (1992).
1
3. A. M. Malleson, H. M. Kellett, R. G. Myhill, and W. P.
Sweetenham, FACSIMILE User Guide, Oxfordshire,
Harwell Laboratory, 1990.
BIBLIOGRAPHY
1
4. P. Neeb, O. Horie, and G. K. Moortgat, Chem. Phys.
Lett., 246, 150 (1995).
1
2
. R. Atkinson, J. Phys. Chem. Ref. Data, 26, 215
1997).
15. H. Niki, P. D. Maker, C. M. Savage, and L. P. Breiten-
bach, Chem. Phys. Lett., 46, 327 (1977).
16. P. Neeb, O. Horie, and G. K. Moortgat, Tetrahedron
Lett., 37, 9297 (1996).
17. K. Griesbaum, M. Hilß, and J. Bosch, Tetrahedron, 52,
14813 (1996).
18. J. T. Herron and R. E. Huie, J. Am. Chem. Soc., 99,
5430 (1977).
19. F. Su, J. G. Calvert, and J. H. Shaw, J. Phys. Chem.,
84, 239 (1980).
(
. A. Guenther, N. C. Hewitt, D. Erickson, R. Fall, C.
Geron, T. Graedel, P. Harley, L. Klinger, M. Lerdau,
W. A. McKay, T. Pierce, B. Scholes, R. Steinbrecher,
R. Tallamraju, J. Taylor, and P. Zimmerman, J. Geo-
phys. Res., 100, 8873 (1995).
3
4
5
. D. J. Jacob and S. C. Wofsy, J. Geophys. Res., 93, 1477
(1988).
. D. Grosjean, E. L. Williams, II, and E. Grosjean, En-
viron. Sci. Technol., 27, 830 (1993).
20. S. Hatakeyama, H. Kobayashi, and H. Akimoto, J. Phys.
Chem., 88, 4736 (1984).
. R. W. Talbot, B. W. Mosher, B. G. Heikes, D. J. Jacob,
J. W. Munger, B. C. Daube, W. C. Keene, J. R. Maben,
and R. S. Artz, J. Geophys. Res., 100, 9335 (1995).
. Y. N. Lee, X. L. Zhou, and K. Hallock, J. Geophys.
Res., 100, 25933 (1995).
21. R. Atkinson, S. M. Aschmann, A. M. Winer, and J. N.
Pitts Jr., Int. J. Chem. Kinet, 13, 1133 (1981).
22. J. Treacy, M. El Hag, D. O’Farrell, and H. Sidebottom,
Ber. Bunsenges. Phys. Chem., 96, 422 (1992).
23. E. Grosjean and D. Grosjean, Int. J. Chem. Kin., 30, 21
(1998).
6
7
. P. Neeb, O. Horie, and G. K. Moortgat, Int. J. Chem.
Kin., 28, 721 (1996).