Rate constants for the OH reactions with oxygenated organic compounds in aqueous solution

Article abstract of DOI:10.1002/kin.20405

The kinetics of the oxidation of functionalized organic compounds of atmospheric relevance by the hydroxyl radical (OH) was measured in the aqueous phase. Competition kinetics, using the thiocyanate anion (SCN-) as competitor, was applied using both a las

Full text of DOI:10.1002/kin.20405

Rate Constants for the OH
Reactions with Oxygenated
Organic Compounds in
Aqueous Solution
S. GLIGOROVSKI,1,2 D. ROUSSE,2 C. H. GEORGE,2 H. HERRMANN1
1Leibniz-Institut fu¨r Tropospha¨renforschung, Permoserstraße 15, 04318 Leipzig, Germany
2Universite´ de Lyon, Lyon F-69626, France; Institut de Recherches sur la Catalyse et l’Environnement de Lyon,
UMR 5256, CNRS, Universite´ de Lyon, Villeurbanne, F-69626, France
Received 24 January 2008; revised 30 July 2008, 6 November 2008; accepted 6 November 2008
DOI 10.1002/kin.20405
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: The kinetics of the oxidation of functionalized organic compounds of atmospheric
relevance by the hydroxyl radical (OH) was measured in the aqueous phase. Competition
kinetics, using the thiocyanate anion (SCN−) as competitor, was applied using both a laser
flash photolysis long path absorption (LP-LPA) setup and a Teflon AF waveguide photolysis
(WP) system. Both experiments were intercompared and validated by measuring the rate
coefficients for the reaction between OH and acetone where values of k1 = (1.8 0.4) ×
108 M−1 s−1 were obtained with the WP system, which agrees very well with the rate
constant of k1 = (2.1 0.6) × 108 M−1 s−1 determined before by LP-LPLA [1]. The following
temperature dependencies of the rate constants (M−1 −1) for the reactions of OH with
s
ketones, dicarboxylic acids, and unsaturated compounds were obtained in Arrhenius’ form:
Reaction of OH with: methylisobutyl ketone (MIBK): k2 = (1.0
0.1) × 1012 exp(−1572 583 ),
T
ꢀ−1062
ꢁ
808
N-methyl pyrrolidone (NMP): k3 = (1.8
0.2) × 1011 exp
, malic acid (pH = 1):
T
k4 = (7.9
0.8) × 1010 exp(−1575 787 ), maleate monoanion (pH = 4.3): k5 = (2.9
0.4) ×
T
1011 exp(−1701 1204 ), maleate dianion (pH = 9): k6 = (1.2
0.2) × 1011 exp(−1449 1193 ),
T
T
mesoxalic acid (pH = 1): k7 = (3.8
0.8) × 1010 exp(−1588 1524 ), triethyleneglycol divinyl
T
Correspondence to: H. Herrmann; e-mail: herrmann@tropos.de.
Correspondence to: C. H. George; e-mail: christian.george@
celyon.univ-lyon1.fr.
Present address of S. Gligorovski: Laboratoire Chimie Provence,
Equipe Instrumentation et Re´activite´ Atmosphe´rique, Universite´s
d’Aix-Marseille I, II, III – CNRS, UMR 6264, Case Courrier 29, 3
Place Victor Hugo, 13331 Marseille Cedex 03, France.
Contract grant sponsor: European Commission (project “Multi-
phase Chemistry of Oxygenated Species in the Troposphere”).
Contract grant number: EVK2-CT-2001-00114.
Contract grant sponsor: Centre National de la Recherche Scien-
tifique (“Programme National de Chimie Atmosphe´rique).
c
ꢀ
2009 Wiley Periodicals, Inc.

310
GLIGOROVSKI ET AL.
ether (DVE-3): k9 = (2.3
0.1) × 1013 exp(−2179 346 ) methacrolein: k10 = (3.0
0.1) ×
T
1013 exp(−2289 168 ). The Ea /A values are given with full precision to avoid rounding errors.
T
Equations were derived from the regression line of the Evans–Polanyi plot in the form activation
energy (Ea) versus bond dissociation energy (BDE) and from the regression line for the region
380 ≤ BDE ≤ 412 kJ mol−1 of the Evans–Polanyi plot in the form log kH versus BDE. Both
equations can be used as a tool for predicting unknown rate coefficients and activation energy
for the reactions of OH with different organics in the aqueous phase. ꢀ 2009 Wiley Periodicals,
C
Inc. Int J Chem Kinet 41: 309–326, 2009
To better understand the atmospheric multiphase
oxidation process of oxygenated organic compounds,
in the present study the attention was focused on the
temperature dependencies of the rate constants for the
reactions of OH radicals with various oxygenated com-
pounds such as ketones, dicarboxylic acids, aldehydes,
and some unsaturated compounds. Aqueous phase ki-
netics was investigated using laser flash photolysis and
a Teflon AF waveguide photolysis (WP) system in the
two contributing laboratories, both applying the com-
petition kinetics method based on the thiocyanate anion
(SCN−) system.
The aim of this study is not only to provide rate
coefficients but also to strengthen already existing [4,5]
correlations based on the Evans–Polanyi relationship
as a tool for predicting unknown rate coefficients and
associated activation parameters in the aqueous phase.
INTRODUCTION
Free radicals such as OH, NO3, Cl, and Br are key
compounds in many (if not all) atmospheric reaction
cycles producing, among other, ozone, acidity, or even
particles. In the troposphere, the hydroxyl radical (OH)
is the most important daytime oxidant. This radical is
the key oxidant involved in the self-cleaning capacity
of the atmosphere and in the transformation of primary
pollutants into secondary species. It is important to
underline that this is the case for the gaseous environ-
ment found not only in the troposphere but also in the
condensed phase such as cloud droplets, deliquescent
particles, and hydrometeors.
Oxidation kinetics and mechanisms in the aqueous
phase have been the focus of many studies over the
last years (e.g., see the recent review by Herrmann [1]
and references therein) where a large amount of data
has been compiled. However, a careful inspection of
these studies reveals that most of the work has focused
either on very simple molecules with just a few car-
bon atoms and only one functional group or, on the
other hand, the complex molecules such as aromatic
compounds. In between these families, however, there
is only little information about the reactivity of OH
toward ketones, hydroxyketones, aldehydes, or dicar-
boxylic acids. For these species there is a clear need for
further research, considering that oxygenated species
may more widely be used as “new” solvents, in a huge
number of industrial processes including cleaning and
degreasing applications, paints, pharmaceutical pro-
cess synthesis, polymer disposal agents, and adhesives
[2], the technical application of which will result in
increased emissions into the atmosphere.
EXPERIMENTAL
Competition Kinetics Method
Because the OH radical in aqueous solution absorbs
only weakly in the deep UV region of the spectrum with
a small extinction coefficient of ε ∼ 600 L mol−1 cm−1
at λmax ∼235 nm [1] and due to the overlapping of the
organic peroxyl radicals formed in parallel to the OH
decay, an OH radical absorption measurement is diffi-
cult. Hence, the concentration of OH was not directly
determined but derived from the titration reaction re-
sulting from the competition −reaction of the OH with
SCN−, which produces (SCN2 ) radical anions.
Also, a number of organic compounds may be pro-
duced in situ in the atmosphere. For instance, the OH
oxidation of aldehydes and ketones both in the gas
phase and in the aqueous phase leads to the accumu-
lated acidity in the aqueous phase. Among these, the
carboxylic acids containing four carbon atoms (e.g.,
malic acid) form important contributions to the or-
ganic mass increase (organic mass production) in tro-
pospheric particles following precursor transfer from
the gas phase [3].
H2O2 + hv → 2OH
(R-A)
(R-B)
(R-C)
(R-D)
OH + SCN− → OH− + SCN
SCN + SCN− ↔ (SCN)−2
(SCN)−2 → products
(R-B) competes with
OH + RH → products
(R-E)
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
311
Therefore, adding any organics to an aqueous solu-
tion containing the chemicals given in the reaction se-
quence (R-A) to (R-D) will lower the yield of (SCN)2−
and such changes will allow the determination of the
ratio of kB/kE (see below).
For this competition kinetics system the tem-
perature dependent rate constant reported by Chin
and Wine has been applied as a reference: ln
in the reaction cell. The amplified electrical signal was
transferred to a digital oscilloscope (Gould 9500A)
connected to a PC controlling the experiment. The er-
rors stated here are statistical errors for a confidence
interval of 95% derived from linear regression analysis
of the experimental data.
Teflon Waveguide Photolysis System
k(T ) = (29.614 0.636)–(1900 190)/T [M−1 s−1
]
[6]. The competition kinetic system based on SCN−
as a competitor for the study of OH reactions in aque-
ous solution has already been applied and discussed in
some previous studies [4,5,7,8].
The Teflon WP system has been detailed in two pre-
vious studies [7,8]. Basically, in WP system a “win-
dowless” reaction cell is applied, which can be created
by using a liquid-core waveguide (henceforth called
LCW) made of Teflon AF-2400. The highly flexible
Teflon AF-2400 tubing was loosely coiled (∼4 cm di-
ameter) and placed in a thermostated reactor close to
a photolytic source, that is, an HPK lamp (Cathodeon,
150 W). The OH radicals were generated by photolysis
of water dissolved H2O2 within the LCW. The gen-
erated OH radicals then react with the organic com-
pounds under study, which are also water dissolved.
The concentration of OH radicals in the aqueous phase
was not directly measured but derived according to the
reactions (R-A) to (R-E) as described above.
Note that a recent study [9] has questioned this ref-
erence value while a second one has validated it and
even extended its application to various ionic strengths
[10]. In this context, it has been decided to derive all
rate coefficients by using the reference rate coefficient
reported by Chin and Wine [6] as the latter has been
widely used as mentioned before [4,5,7,8]. Where data
for comparison exist, the results obtained in the present
work are in good agreement with the kinetic data from
those previous studies and various available kinetic
data from the literature demonstrate that the use of
thiocyanate as a competitor and the use of the kinetic
reference data set of Chin and Wine are appropriate
for the experimental conditions applied in the present
work.
To probe the (SCN)−2 concentration in the LCW
tube, the output of a 75 W xenon lamp was focused
onto the entrance of a 200 μm diameter-fused silica
optical fiber held inside the LCW tube. The light leaves
the solid optical fiber and is conducted to the other end
of the liquid waveguide. There, another fused silica
optical fiber (placed in the liquid) collects most of
the transmitted light and leads it to a spectrograph
(Andor, Shamrock) coupled to a charge-coupled device
(CCD) camera (Andor Technology). The CCD camera
monitors the evolution of the UV–visible spectrum as a
function of wavelength and time within the LCW. The
(SCN)−2 anion spectra were recorded in the wavelength
range from 350 to 750 nm, but the kinetics was derived
from absorbance changes at 465 nm.
Aqueous solutions of the organic reactant and the
radical precursors were continuously pumped into the
LCW, which was hence used as a microflowtube. In
all cases, the liquid passed through the LCW in less
than 0.5 s. Under these experimental conditions, steady
state was obtained within the LCW and verified exper-
imentally (see a previous study [7]) and theoretically
using numerical simulations of the chemistry within
the LCW.
To improve the reliability of the kinetic data (and
hence of any correlations making use of them), the
kinetics was derived by two different experimental ap-
proaches as described in the following.
Laser Flash Photolysis Long-Path
Absorption Setup
A laser photolysis long-path absorption (LP-LPA)
setup [1,4,5] was used for indirect kinetic studies of
OH radical reactions. This experimental setup has been
described in detail previously and only a brief summary
is given here. The formation of the OH radicals was
initiated by photolysis of hydrogen peroxide by an ex-
cimer laser (Lambda LPX 100) operated at λ = 248 nm
using KrF as active medium. The mean energy of the
excimer laser was about 150 mJ measured with an en-
ergy monitor.
The time dependency of the concentration of the
(SCN)−2 radical was monitored using a combination
of a high-pressure Hg-Xe lamp (λ = 436 nm) and a
monochromator/photomultiplier unit (Zeiss PMQ 2/
Hamamatsu PMT 1P 28). The emitted analytical light
has been adjusted via White cell optics [11] for 12
passes resulting in an absorption path length of 84 cm
For both experiments, the following relationship de-
scribes the relation between the absorbance of (SCN)2−
in the presence and absence of added organics:
ꢂ
ꢃ
1
Abs
1
Abs0
kE[RH]
kB[SCN−]
=
1 +
(1)
International Journal of Chemical Kinetics DOI 10.1002/kin

312
GLIGOROVSKI ET AL.
In this equation, Abs is the absorbance of the
(SCN)−2 anion at a known concentration of the organic
compound RH and Abs0 is the maximum (SCN)−2 anion
absorbance in the absence of the reactant RH. There-
fore, given the absorbencies, kB and the concentra-
tion of the reactant RH, the rate coefficient kE can be
determined.
Therefore, by means of the two experiments de-
scribed above any changes in the absorbance of the
(SCN)−2 anion as a function of the organics content of
the solution can be measured allowing the determina-
tion of the ratio kE/kB. Knowing the rate coefficient kB
for the reference reaction (R-B), allows the calculation
of the unknown rate constant kE.
The measured data are compiled in Table I in com-
parison with preexisting data. First of all, it can be seen
that the data obtained within this study do agree quite
remarkably validating the experimental approach.
The measured rate coefficients do also compare rea-
sonably with most of the previous measurements. The
very good agreement between the rate constants from
this study with the one from Ervens et al. [5] represents
an additional validation for the two complementary ex-
perimental systems applied in this study.
As can be seen from Table I, the reaction of OH
with acetone was studied over a wide range of experi-
mental methods since this compound is one of the most
abundant carbonyls in the atmosphere. Following the
rate constants for the reactions of OH with acetone as a
function of temperature is out of scope since they were
previously studied, and the corresponding activation
parameters were reported [5]. Studies focusing on the
gas-phase reactions suggested that (R-1) may proceed
by two mechanisms, through an H-abstraction mech-
anism or by addition via intermediate H-bonded com-
plexes [20–25]. The recent kinetic and product studies
for the reactions of OH with acetone in the aqueous
phase [18] identified reaction products that correspond
(i) to the oxidation of the weakest C H bond into a
carbonyl function and (ii) to the oxidation of the weak-
est C H bond into a hydroxyl group. Accordingly, the
main reaction pathway is still C H oxidation through
an H-abstraction process (as depicted by (R-1)) and
not addition on the carbonyl function.
Materials
All the organic compounds used in this study were
purchased from Sigma Aldrich (Seelze, Germany) (pu-
rity ≥ 99%) except DVE-3, which was provided from
BASF (Ludwigshafen, Germany).
Prior to the measurements, the LP-LPA system was
cleaned by continuous flowing (15 min) of ultrapure
MilliQ water through the tank, tubing, and the reaction
cell.
The WP system was cleaned with continuous flow-
ing (10–15 min) of ultrapure MilliQ water throughout
the Teflon WP system.
RESULTS AND DISCUSSION
Methyl Isobutyl Ketone
In this section, a report will be first given on all
the kinetics measurements performed for a number
of oxygenated compounds, before discussing possible
structure–reactivity relationships.
Methyl isobutyl ketone (MIBK) is widely used as a
solvent, a denaturant, in dry cleaning operations and in
the synthesis of methyl isobutyl carbinol [26]. On the
other hand, MIBK could be an important atmospheric
organic compound because of its ozone production po-
tential [27]. As a consequence, the reaction of OH with
MIBK was also investigated at the 298 K and the rate
constant is presented in Table I and compared with the
literature.
Reactions of OH with Ketones
Ketones are an important class of compounds that are
either directly emitted by industrial/commercial appli-
cations or can be formed in the atmosphere by oxida-
tion and photochemical reactions.
OH + (CH3)2CHCH2COCH3 →
H2O + (CH3)2CHCHCOCH3
(R-2)
Acetone
The reaction between MIBK and OH may corre-
spond to an H-abstraction either at the secondary car-
bon atom or at the tertiary carbon atom of the butyl
group. To identify the weakest bond in an H-abstraction
process, Benson’s incremental method [28] was ap-
plied. The calculated bond dissociation energy (BDE)
The reaction of OH with acetone was investigated at
298 K in aqueous solution using the two experimental
techniques.
OH + CH3COCH3 → H2O + CH3COCH2 (R-1)
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
313
Table I The Rate Constants for the Reactions of OH with Acetone, Methyl Isobutyl Ketone, and N-Methyl Pyrolidone,
in Aqueous Solution, at 298 K
Experiment
Type
Analytical
Technique
Source
of OH
Reference
Compound
Compound
Acetone
k [L mol−1 s−1
]
pH
Ref.
8.3 × 107
9.7 × 107
1.4 × 108
1.3 × 108
PR
PR
PR
PR
LFP
UV–vis abs
UV–vis abs
UV–vis abs
UV–vis abs
UV–vis abs
e−aq + N2O
e−aq + N2O
e−aq + N2O
e−aq + N2O
H2O2 + hν
(248 nm)
I−
SCN−
[Fe(CN)6]4−
ABTS2−
Absorption/time
profile at 244 nm
CH3OH
7
6–7
[12]
[13]
[14]
[15]
[16]
6
6
(1.3 0.1) × 108
(1.7 0.5) × 108
(2.1 0.6) × 108
(1.8 0.4) × 108
(2.1 0.5) × 109
(4.6 1.6) × 109
(3.1 1.0) × 109
(4.9 2.3) × 109
CP
FP
FP
CP
FP
CP
FP
Chromatography Photo–Fenton
reaction
UV–vis abs
2
6
[17]
[5]
H2O2 + hν
(248 nm)
H2O2 + hν
(248 nm)
SCN−
SCN−
UV–vis abs
6–7 This work
[18]
6–7 This work
[19]
6–7 This work
Methyl
isobutyl
ketone
Chromatography Photo–Fenton
reaction
UV–vis abs
2-Propanol
SCN−
2
H2O2 + hν
(248 nm)
N-Methyl
pyrrolidone
Chromatography Photo–Fenton
reaction
UV–vis abs
MIBK
2
H2O2 + hν
SCN−
(248 nm)
PR = pulse radiolysis; LFP = laser flash photolysis; FP = flash photolysis; CP = continuous photolysis.
of 376 kJ mol−1indicates that the weakest bond cor-
responds to the tertiary carbon atom (for more details
see section on OH correlations).
.
N
N
H
H
O
O
H O
OH
+
2
+
C
H
(R-3)
Reaction (R-2) was studied as a function of tem-
perature over the range from 278 to 318 K. The corre-
sponding Arrhenius equation is
The rate coefficients for this reaction are listed in
Table I and are described by the following Arrhenius
equation:
k2 = (1.0
0.1) × 1012
k3 = (1.8
0.2) × 1011
ꢂ
ꢃ
−1572
583
−1 s−1
(2)
ꢂ
ꢃ
× exp
M
−1062
808
× exp
M
−1 s−1
(3)
T
T
The measured rate coefficients for R-2 are summa-
rized in Table I.
The Arrhenius plots for the reactions of OH with
MIBK and NMP are shown in Fig. 1.
The oxidation of NMP initiated via H-abstraction by
OH occurs preferentially at the α-carbon atom [30].
This pathway leads to succinimide through the for-
mation of N-methyl succinimide and 2-pyrrolidone,
respectively.
The kinetics and mechanisms of (R-3) in the aque-
ous phase were studied earlier by Monod and co-
authors [19] suggesting three H-abstraction reactions,
(i) at the methyl group; (ii) at the CH2 group adjacent
to the imine function, and (iii) at the other CH2 groups
of the ring. Pathway (iii), which is more speculative,
N-Methyl Pyrrolidone
N-Methyl pyrrolidone (NMP) is suggested to be used
as a solvent in a wide variety of commercial applica-
tions. It may be released to the troposphere as a fugitive
emission during its production or in the effluent of in-
dustrial processes [29].
The reaction of OH with NMP has been investigated
in the temperature range between 278 and 308 K.
International Journal of Chemical Kinetics DOI 10.1002/kin

314
GLIGOROVSKI ET AL.
Figure 1 Arrhenius plot for the reaction of OH with MIBK ( ) and NMP (ꢀ).
suggests that this reaction proceeds via a ring opening
mechanism, leading to the formation of N-methyl-4-
aminobutanoic acid. The associated rate coefficients at
298 K are summarized in Table I and compared with
the literature available data.
From the comparison it can be seen that both rate
constants, obtained in this study and the one from
Monod and coauthors [19] are in good agreement
within the statistical error.
ison to the C4-α,ω-dicarboxylic acid (e.g., succinic
acid) [33].
The LP-LPA setup has been used to study the
reactivity of OH with dicarboxylic acids such as
malic and mesoxalic acids at different pH values and
temperatures.
Malic Acid
Table II presents the resulting kinetic data for the reac-
tion between OH and malic acid at 298 K, compared
with available literature data.
Reactions of OH with Dicarboxylic Acids
Experimental observations have shown that dicar-
boxylic acids are commonly found in snow, rain, cloud
droplets, fog, and aerosols both in urban, rural, ma-
rine, remote, and polar areas [31]. Ervens et al. [32]
in their cloud parcel model coupled to an aqueous
chemistry model have shown that dicarboxylic acids
(≤C6) can be efficiently produced by tropospheric
multiphase oxidation. The formation and accumula-
tion of dicarboxylic acids in aqueous solution is also
observed in the aqueous phase model CAPRAM 3.0
recently developed by Herrmann et al. [3]. As dicar-
boxylic acids may affect the hygroscopic behavior of
the aerosols [32,33], it is especially important to un-
derstand their condensed phase chemistry. Some ob-
servations have demonstrated that the presence of a
double bond (mesoxalic acid) or a hydroxyl group
(malic acid) in the C4 dicarboxylic acids could increase
the water solubility, and thus enhance the decrease of
surface tension in concentrated solutions in compar-
At pH = 1, the reaction proceeds as follows:
OH + HOOCCH(OH)CH2COOH →
H2O + HOOCC(OH)CH2COOH (R-4)
In Table II, it can be seen that the measured rate
constant is about two times lower compared with the
previously reported data. The reason for this is unclear
at present. The reaction of OH with the dianion form
of malic acid was investigated at pH = 9:
OH + −OOCCH(OH)CH2COO−
→
H2O + −OOCC(OH)CH2COOH− (R-5)
At this pH, there is a good agreement with the rate
constant determined by Hesper [16], but still some dis-
crepancy with the data obtained by Logan [35]. This
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
315
Table II The Rate Constants for the Reactions of OH with Malic Acid at Different pH, in Aqueous Solution, at 298 K
Reference
Compound
Experiment
Type
Malic acid
pH
2
k [L mol−1 s−1
6.4 × 108
]
Ref.
[34]
Undissociated
OH + 5-MeU
Absorption/time
profile at 244 nm
OH + SCN−
γ -R
1.5
1
9
(7.1 1.3) × 108
(3.6 1.6) × 108
(8.4 1.0) × 108
LFP
LFP
LFP
[16]
This work
[16]
Dianion
Absorption/time
profile at 244 nm
OH
9
9
4.3
2.3 × 109
Ferrocenylbutyrate ion
OH + SCN−
PR
LFP
LFP
[35]
This work
This work
(8.5 1.1) × 108
(9.7 2.5) × 108
Monoanion
OH + SCN−
PR = pulse radiolysis; LFP = laser flash photolysis, γ -R = gamma radiolysis (pKa = 3.44 and 5.11) [36].
difference can be partly explained by the use of differ-
ent reference compounds. In the study of Logan [35] a
reference reaction of OH + ferrocenylbutyrate ion with
the rate constant of 1.6 × 1010 L mol−1 s−1 was used.
The reaction between OH radical and the monoanion
form of malic acid was investigated at pH 4.3:
coefficient for the monoanion. All data are listed in
Table II.
In addition, the temperature dependence was inves-
tigated from 288 to 328 K, at pH = 1, pH = 4.3, and
pH = 9 (see Table III).
The rate constants for the reaction of OH with
maleate (pH = 4.3) at different temperatures were
again corrected according to the reported dissociation
constants of malic acid pKa = 3.44 and 5.11 [36], as-
suming that the pKa values do not change in the tem-
perature range studied here as T -dependences of the
mentioned pKa values were not available.
OH + HOOCCH(OH)CH2COO−
→
H2O + HOOCC(OH)CH2COO−
(R-6)
However, under these conditions the different an-
ionic forms of the acid may coexist. Therefore, de-
termining the rate coefficient for the monoanion only
requires some correction of the raw data. This can be
achieved by considering the two dissociation constants
for malic acid (pKa = 3.44 and 5.11) [36], which al-
low the speciation of the various forms of the acids.
In the region where the monoanion form of malic acid
exists with a fraction of about 80%, the undissoci-
ated form and the dianion are present with 13% and
8%, respectively. Hence, the obtained rate constant of
(8.9 2.2) × 108 L mol−1 s−1 was properly corrected
correspondingly to the dissociation constants of malic
acid (pKa = 3.44 and 5.11) [36] and by help of the
following equation:
As stated in our previous study [5], the rate co-
efficients for the reactions of OH with undissociated
acids are lower than those of their corresponding an-
ions. This trend is also confirmed in the present study
for (R-5) and (R-6). The possible explanation for such
trends may arise from an electron transfer contribution
to the observed rate constants for anions.
The measured Arrhenius equations for the reactions
of OH with the undissociated malic acid, monoan-
ion, and dianion are presented with Eqs. (5)–(7),
Table III The Observed Rate Constants for the
Reactions of OH with Malic Acid at Different pH and
Temperatures in Aqueous Solution
kobs = 0.8 k(HA−) + 0.13 k(H2A) + 0.08 k(A2−
)
T
pH = l
pH = 4.3
pH = 9
(K) k [L mol−1 s−1
]
k [L mol−1 s−1
]
k [L mol−1 s−1
]
(4)
where kobs is the measured rate constant for the
reaction of OH with the monoanion form at pH = 4.3,
k(H2A) is the rate constant observed for the reaction of
OH with malic acid at pH = 1, k(A2−) is the rate con-
stant obtained for the reaction of OH with the dianion
form of malic acid at pH = 9, and k(HA−) is the rate
288 (3.6 0.8) × 108 (8.3 3.6) × 108 (8.3 1.1) × 108
298 (3.6 1.6) × 108 (9.7 2.5) × 108 (8.5 1.1) × 108
308 (4.7 0.1) × 108 (1.0 0.2) × 109 (9.9 2.8) × 108
318 (5.6 1.2) × 108 (1.6 0.5) × 109 (1.1 0.2) × 109
328 (6.7 1.9) × 108 (1.6 0.4) × 109 (1.6 0.4) × 109
International Journal of Chemical Kinetics DOI 10.1002/kin

316
GLIGOROVSKI ET AL.
Figure 2 Arrhenius plot for the reaction of OH with malic acid: undissociated ( ), monoanion ( ), and dianion ( ).
◦
ꢁ
respectively.
The hydration constant of mesoxalic acid is reported
in the literature [37] to be KHydr = 102 = [hydrate
form]/[carbonyl form] at 298 K and pH = 1.7, indicat-
ing that the hydrated form is predominant. However,
for the dianion, the carbonyl form is as abundant as
10% at 293 K [37]. The obtained rate constant for the
reaction of OH with the hydrated form of mesoxalic
acid at pH = 9 (dianion)
k4 = (7.9
0.8) × 1010
ꢂ
ꢂ
ꢂ
ꢃ
−1575
787
× exp
k5 = (2.9
× exp
k6 = (1.2
× exp
M
−1 s−1
(5)
(6)
(7)
T
0.4) × 1011
ꢃ
−1701
1204
M
−1 s−1
T
OH + −OOCC(OH)2COO− → products (R-8)
0.2) × 1011
ꢃ
is k8 = (2.2 0.6) × 109 L mol−1
s
−1. A rate constant
−1449
1193
M
−1 s−1
of k8 = 1.0 × 108 L mol−1 s−1 has been reported by
Schuchmann et al. [37] measured by pulse radiolysis
at pH = 9. Accordingly, there is a large discrepancy
of about one order of magnitude between both studies
that is still unclear to the authors.
T
Figure 2 shows the corresponding plots for these
Arrhenius equations.
The reactions of OH with mesoxalic acid were
investigated in the temperature range between 288 and
328 K, only at pH = 1. The measured rate coefficients
for the undissociated acid at different temperatures are
summarized in Table IV. In this case the rate constants
were not corrected since the pKa value for this acid en-
ables exclusive existence of its dianion form at pH = 9.
Mesoxalic Acid
The reaction of OH radical with mesoxalic acid
OH + HOOC(CO)COOH → products
(R-7)
has been investigated in aqueous solution at pH = 1.
The dissociation constants of mesoxalic acid are
pKa = 1.6 and 3.9 as reported by Schuchmann et al.
[37]. At pH = 1, we measured a rate constant of
Table IV The Observed Rate Constants for the
Reactions of OH with Mesoxalic Acid at Different
Temperatures in Aqueous Solution at pH = 1
k7 = (1.8 0.3) × 108 L mol−1 s−1
.
k (Mesoxalic Acid) [L mol−1 s−1
]
The same reaction has been studied by pulse radi-
T (K)
olysis by Schuchmann et al. [37] at pH = 3 and the
(1.7 0.8) × 108
(1.8 0.3) × 108
(1.8 0.7) × 108
(3.0 0.9) × 108
(3.1 0.4) × 108
reported rate constant is k7 = 5.7 × 107 L mol−1 s−1
.
288
298
308
318
328
Obviously the rate constant from this study is some-
what different than the rate constant measured by
Schuchmann et al., which arise due to the different
pH values of the solution.
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
317
Figure 3 Arrhenius plot for the reaction of OH with mesoxalic acid, at pH = 1.
The Arrhenius plot corresponding to Eq. (8) is
shown in Fig. 3.
The reaction of OH with DVE-3 has been investi-
gated in the temperature range between 288 and 328 K
(Table V). The measured rate constants of this reaction
are similar to the calculated diffusion rate constants for
the corresponding temperature. In this case according
to the kinetic theory [39] the investigated reactions are
fully diffusion controlled reactions. Not surprisingly,
the rate constant from this study is not significantly
different compared with the one determined in the pre-
vious study by Moise et al., in 2005 [8], where a value
k7 = (3.8
0.8) × 1010
ꢂ
ꢃ
−1588
1524
× exp
M
−1 s−1
(8)
T
In aqueous solutions, the mesoxalate dianion is in
equilibrium with its hydrate [37]. The carbonyl form
is present with 10% at 293 K and the ratio [hydrate
form]/[carbonyl form] increases linearly with temper-
ature. Therefore, the reaction of OH with mesoxalic
acid dianion was not studied in an extended tempera-
ture range due to the lack of data for the T -dependence
of KHydr for the mesoxalic acid dianion.
of k9 = (4.8 2.4) × 1010 L mol−1
s
−1 was found.
From the temperature-dependent rate constants for
this reaction, the following Arrhenius equation has
been derived:
k9 = (2.3
0.1) × 1013
ꢂ
ꢃ
−2179
346
× exp
M−1 s−1
(9)
Reactions of OH with Unsaturated
Compounds
T
Vinyl ethers arewidely usedas solvents in the industrial
applications, as motor oil additives and for coatings
(see e.g., website of BASF) [38]. Upon application
they may partly be released into the atmosphere where
they are available for atmospheric processing.
In this work, the reaction of OH with triethyleneg-
lycol divinyl ether (DVE-3)
The corresponding Arrhenius plot is shown in
Fig. 4.
Table V The Observed Rate Constants for the
Reactions of OH with DVE-3 and Methacrolein at
Different Temperatures in Aqueous Solution
k (DVE-3)
[L mol−1 s−1
k (Methacrolein)
[L mol−1 s−1
T (K)
]
]
OH + H2C = CH(OCH2CH2)3OCH = CH2 →
278
288
298
308
318
328
–
(7.9 1.6) × 109
(1.1 0.5) × 1010
(1.4 0.3) × 1010
(1.8 0.3) × 1010
(2.5 0.7) × 1010
–
products
(R-9)
(1.2 0.2) × 1010
(1.6 0.2) × 1010
(2.1 0.2) × 1010
(2.4 0.7) × 1010
(3.1 1.8) × 1010
has been investigated in aqueous solution at 298 K us-
ing the LP-LPA setup. The obtained second-order rate
constant for the reaction (R-9) is k9 = (1.6 0.2) × 1010
L mol−1 s−1
.
International Journal of Chemical Kinetics DOI 10.1002/kin

318
GLIGOROVSKI ET AL.
Figure 4 Arrhenius plot for the reaction of OH with DVE-3 ( ) and methacrolein (ꢀ).
Mechanistic studies have shown that the gas-phase
Methacrolein
reactions of OH with methacrolein proceed via addi-
tion to the double bond (80% of the addition occurs at
the carbon) and by H abstraction of the aldehyde group
[44]. Since there are no available mechanistic data for
(R-10) in the aqueous phase, it can be simply assumed
that reaction of OH with methacrolein might proceed
through the same mechanism as in the gas phase.
Methacrolein is used in numerous pharmaceutical and
agricultural applications. It is also produced in signif-
icant yields from the gas-phase reactions of isoprene
[40] and directly emitted from vegetation [41]. Un-
der atmospheric conditions, the major loss process for
methacrolein in the gas phase is reaction with the OH
radical [42]. In this work, the reactivity of OH with
methacrolein, that is,
OH Reactivity Correlations
OH + H2C = C(CH3)CHO → products (R-10)
Evans–Polanyi Type Correlations: An Overview. The
basic principle of the Evans and Polanyi [45] correla-
tion relies on the relation between the activation energy
(Ea) and the reaction enthalpy (ꢀHR0) for a series of
reactions:
has been investigated in aqueous solution at 298 K.
This reaction proceeds by addition of the OH radical
to the carbon atoms of the >C C< double bond. The
measured rate coefficient of k10 = (1.4 0.3) × 1010
L
Ea = E0 + α ꢀHR0
(11)
mol−1 s−1 is higher than the OH addition rate constant
for the reaction of OH with acrolein of k = 7.0 × 109 L
mol−1 s−1 reported in the literature [43], which might
be attributed to the enhancement of the reactivity of the
>C C< double bond due to the presence of a methyl
group in methacrolein.
For reactions proceeding through the same mechanism
(e.g., H-abstraction mechanism), the reaction enthalpy
can be correlated to the BDE (C H) of the weakest
extractable bond:
The kinetics of the reaction of OH with
methacrolein has also been investigated as a function
of temperature as summarized in Table V. Again, this
reaction is shown to be fully diffusion controlled. The
Arrhenius plot corresponding to Eq. (10) is shown in
Fig. 4.
ꢀHR ∼ BDE (C H)
(12)
Based on the Arrhenius equation, it emerges that
BDE can also be correlated to the measured rate con-
stants of the H-abstraction reactions:
Ea
log k = log A −
RT
k10 = (3.0
0.1) × 1013
ꢂ
ꢃ
−2289
168
E0
αBDE(C H)
RT
× exp
M−1 s−1
(10)
= log A −
−
(13)
T
RT
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
319
If “nH” equivalent H atoms in the reactive molecule
exist, all of them contribute to the same extent (kH) to
cal and saturated organic aliphatic compounds in the
gas phase and in aqueous solution is the abstraction of
the most loosely bonded hydrogen atom (e.g., see the
reviews by Mellouki and coauthors [20], Herrmann [1],
and references therein). In previous kinetic studies of
OH radical reactions in aqueous solutions performed
by Ervens et al. [5], Gligorovski and Herrmann [4],
and Morozov et al. [46], the correlations between the
BDE of the weakest C H bond, activation parame-
ters, and kinetic data have been established using the
Evans–Polanyi equation.
the observed rate constant kobs
:
kobs = nH · kH
(14)
The combination of Eqs. (13) and (14) leads to a
correlation between the logarithm of the rate constant
per abstractable H atom (log kH (M−1 −1)) and BDE
s
(C H):
ꢂ
ꢃ
Such correlations could further be strengthened
by including the kinetic information from this study
(Fig. 5).
kobs
nH
log
= log kH
E0
α × BDE(C H)
(15)
Hence, the measured rate coefficients for the reac-
tions of OH with acetone, MIBK, malic acid (pH = 1),
maleate (monoanion, pH = 4.3), and maleate (dianion,
pH = 9) were added to the Evans–Polanyi correlation
previously developed by Gligorovski and Herrmann
[4]. The measured value of the rate coefficient for the
reactions of OH with NMP was excluded from these
correlations due to the missing value of the BDE. The
rate coefficients per abstractable hydrogen atoms (kH)
and BDEs are summarized in Table VI.
= log A −
−
RT
RT
The rate constant per abstractable hydrogen atom is
the ratio of the observed second-order rate constant
and number of the identical most easily abstractable
hydrogen atoms in the given molecule. Following this
treatment it has to be assumed that the OH radical is
predominantly reacting with the weakest abstractable
H atom.
The BDEs for the malic acid and MIBK were esti-
mated by Benson’s method [28].
For the regression line through the aqueous-phase
data points in Fig. 5 for the region 380 ≤ BDE ≤ 412
kJ mol−1 the following equation holds:
Comparison of Evans–Polanyi Correlations in the
Form of log kH = f (BDE) Between the Aqueous
Phase and Gas Phase. Several studies suggest that
the prevailing reaction mechanism between OH radi-
Figure 5 Comparison of the Evans–Polanyi plots in the form of log (k ) versus BDE between aqueous phase ( ), this work,
◦
H
ꢂ
( ), and gas phase ( ). Numbering refers to the numbers in Table VI.
•
International Journal of Chemical Kinetics DOI 10.1002/kin

320
GLIGOROVSKI ET AL.
Table VI Rate Constants Per Most Easily Abstractable Hydrogen Atom (kH) in the Aqueous Phase and in the Gas
Phase
Aqueous Phase
Gas Phase
◦ kH
log kH
(298 K)
(L mol−1 s−1
ꢀkH
Log kH
(298 K)
(L mol−1 s−1
(298 K)
(298 K)
(L mol−1 s−1
BDEa
Ref. (kJ mol−1
)
No.
Reactant
(L mol−1 s−1
)
)
Ref.
)
)
1
2
3
4
5
6
7
8
Acetone
tert-Butanol
Acetonylacetone
Ethyl formate
Ethanol
MEK
Malic acid
Maleate (monoanion)
Maleate (dianion)
1-Propanol
3.0 × 107
5.3 × 107
1.9 × 108
3.3 × 108
1.0 × 109
7.5 × 108
3.6 × 108
9.7 × 108
8.5 × 108
1.6 × 109
2.1 × 109
2.1 × 109
3.5 × 109
2.3 × 109
2.4 × 109
3.9 × 109
2.8 × 109
3.9 × 109
2.9 × 109
7.48
7.73
8.28
8.52
9.02
8.88
8.56
8.99
8.93
9.20
9.32
9.32
9.55
9.36
9.38
9.59
9.45
9.59
9.46
This work
[5]
8.9 × 107
5.4 × 107
2.2 × 108
5.2 × 108
1.1 × 109
1.0 × 109
–
7.95
7.73
8.35
8.72
9.06
9.00
–
[48]
[49]
[50]
[50]
[51]
[52]
–
411
410
401
396
389
386
386
386
386
385
385
381
381
376
374a
372
366
361
358
[4]
[4]
[5]
[4]
This work
This work
This work
[16]
–
–
–
–
–
–
9
10
11
12
13
14
15
16
17
18
19
1.7 × 109
4.7 × 109
3.3 × 109
4.9 × 109
4.2 × 109
1.0 × 1010
1.6 × 1010
1.2 × 1010
1.4 × 1010
1.1 × 1010
9.23
9.67
9.52
9.69
9.63
10.00
10.20
10.07
10.16
10.02
[53]
[53]
[51]
[54]
[42]
[55]
[56]
[42]
[56]
[57]
1-Butanol
2-Propanol
2-Butanol
MIBK
[16]
[16]
[16]
This work
[47]
Acetaldehyde
Valeraldehyde
Propionaldehyde
Butyraldehyde
Isobutyraldehyde
[18]
[16]
[16]
[4]
aRefs. [58] and [59].
log (kH (M−1s−1)) = (30
− (0.06
with n = 14 and r = 0.95.
4)
In this plateau (Fig. 5) formed by the aldehydes, we
added the rate constants values for the reactions of
OH with MIBK from this study and valeraldehyde
from the literature [18]. In the BDE range from 358
to 376 kJ mol−1, the rate constant corresponds to
0.01) × BDE (kJ mol−1
)
(16)
The attack of OH radical on dicarboxylic acids is
directed primarily to the C H bonds at different posi-
tions. The corresponding mono and dianions are react-
ing faster, which probably comes from the difference
in the inductive effect of COOH and COO− groups
and the electrophilic nature of the OH radical.
kH = 3.0 × 109 L mol−1 s−1. The formation of the
¯
plateau by aldehydes in both phases indicates that the
reaction mechanism for these compounds toward OH
radical is not a H-abstraction but probably addition to
the >C O double bond or formation of the hydrogen-
bonded complexes followed by H-abstraction [20].
However, the existence of hydrogen-bonded com-
plexes is less probable in the aqueous phase, since all
the compounds are already bonded to water molecules
in the first solvation shell (see next section). To the
best of the authors knowledge the existence of such a
plateau remains unclear.
This regression line found here may be used to esti-
mate rate constants at 298 K for H-abstraction reactions
of the OH radical in aqueous solution when measured
kinetic data are not available.
The same series of reactions in the gas phase were
plotted and compared with the corresponding one in the
aqueous phase in Fig. 5. Identical plateau formed by
the group of aldehydes and MIBK is observed for the
gas-phase data as well. In the BDE range between 358
Although electron transfer may contribute to the
observed rate constant of OH-oxidation of anions
such as maleate (monoanion) and maleate (dianion),
H-abstraction seems to be the dominating reaction
pathway also for these species.
In case of aldehydes, as discussed in the previous
study [4], the number of most easily abstractable equiv-
alent H atoms is n = 1 in both forms, the aldehyde and
in its corresponding diol form. Therefore, the rate con-
stant per abstractable H atom (kH) is equal to the ob-
served rate constant in both cases. The rate constants
for the reaction of OH with the equilibrium mixtures
of aldehydes and their hydrates are considered for the
correlation shown in Fig. 5.
On the other hand, the BDEs applied in the present
correlation are those for the unhydrated aldehydes.
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
321
Table VII Preexponential Factors and Activation Energies for H-Abstraction Reactions of OH Radical in Aqueous
Solution
No.
Reactant
A [L mol−1 s−1
]
E [kJ mol−1
]
Ref.
a
1
2
3
4
5
6
7
8
Acetone
Acetone
Acetone
Acetone
Acetone
(3.4 0.4) × 1011
(8.4 0.4) × 1010
(2.1 0.3) × 1011
(6.6 0.2) × 1010
–
(18 11)
(16 3)
(17 8)
(16 3)
(12 4)
(10 3)
(12 5)
(10 4)
(10 5)
(7 1)
(15 8)
(4 0.3)
(13 3)
(8 6)
(7 2)
(8 1)
(8 2)
(8 3)
(11 3)
(8 3)
(6 3)
(13 5)
10
(13 7)
(14 10)
(12 10)
[4]
[16]
[60]
[17]
[18]
[5]
[4]
[4]
[5]
[18]
[4]
[60]
[18]
[5]
[18]
[16]
[16]
tert-Butanol
Acetonylacetone
Ethyl formate
Ethanol
(3.3 0.1) × 1010
(1.1 0.1) × 1011
(1.8 0.1) × 1010
(1.0 0.1) × 1011
–
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Ethanol
MEK
MEK
MEK
1-Propanol
1-Propanol
1-Butanol
2-Propanol
(5.1 0.6) × 1011
(6.6 4.9) × 109
–
(5.6 0.6) × 1010
–
(1.0 0.1) × 1011
(6.1 0.3) × 1010
(7.4 0.3) × 1010
(2.6 0.1) × 1011
(8.1 0.3) × 1010
(3.0 0.1) × 1010
(1.0 0.1) × 1012
1.3 × 1011
2-Butanol
[16]
[16]
[16]
[4]
Propionaldehyde
Butyraldehyde
Isobutyraldehyde
MIBK
This work
[18]
This work
This work
This work
MIBK
Malic acid (pH = l)
Maleate monoanion (pH = 4.3)
Maleate dianion (pH = 9)
(7.9 0.8) × 1010
(2.9 0.4) × 1011
(1.2 0.2) × 1011
and 374 kJ mol−1 the rate constant for the gas-phase
reactions with aldehydes and MIBK is best represented
In Tables VII and VIII Ea for the reactions of OH
with different organic compounds are listed for the
aqueous phase and gas phase, respectively.
The regression line for the aqueous phase repre-
sented in Fig. 6 corresponds to
by kH = 1.2 × 1010 L mol−1 s−1. The correlation found
¯
in the gas phase is much alike to the one in the aqueous
phase for the BDE between 380 and 412 (kJ mol−1
and corresponds to the following equation:
)
Ea (kJ mol−1) = − (42
+ (0.14
31)
log (kH(M−1s−1)) = (32
4)
0.08) × BDE (kJ mol−1
)
(18)
− (0.06
0.01) × BDE
(17)
with n = 26 and r = 0.58
The correlation found here might be used to estimate
activation energies for H-abstraction reactions of OH
radical reactions in aqueous solution. The potential
user of the correlation Eq. (18) has to be aware of the
uncertainties of the derived activation energies.
with n = 10; r = 0.98
Evans–Polanyi Correlation in the Form of
Ea = f (BDE): Comparison Between the Aqueous
Phase and Gas Phase. This form of Evans–Polanyi
plot has also been previously developed by Ervens
et al. [5] and further confirmed by Gligorovski and
Herrmann [4] in 2004 for the H-abstraction reactions
of OH with saturated organic compounds. The afore-
mentioned correlation was expanded with the Ea values
from this study. Using the available Ea data set for the
OH reactions in the gas phase a comparative analysis
is undertaken.
The equation derived from the regression line for
the reactions of OH in the gas phase corresponds to
Ea (kJ mol−1) = − (58
18)
+ (0.15
0.05) × BDE (kJ mol−1
)
(19)
with n = 25 and r = 0.79
The very similar slopes in both plots presented
in Fig. 6 indicate that the same reaction mechanism
International Journal of Chemical Kinetics DOI 10.1002/kin

322
GLIGOROVSKI ET AL.
Table VIII Preexponential Factors and Activation Energies for H-Abstraction Reactions of OH Radical in the Gas
Phase
No.
Reactant
A [L mol−1 s−1
]
E [kJ mol−1
]
Ref.
a
1
2
3
4
5
6
7
8
Acetone
Acetone
Acetone
1.7 × 109
(7.5 1.4) × 108
1.3 × 109
(6.3 1.6)
(4.6 0.5)
(5.7 0.8)
(5.0 0.6)
(2.6 1.3)
[61]
[62]
[63]
[64]
[65]
[66]
[50]
[67]
[61]
[68]
[63]
[62]
[61]
[64]
[61]
[61]
[61]
[61]
[61]
[69]
[70]
[71]
[69]
[70]
[62]
Acetone
(1.0 0.3) × 109
(2.0 1.0) × 109
(1.6 0.3) × 109
(9.0 2.6) × 108
(3.4 0.7) × 108
2.5 × 109
tert-Butanol
tert-Butanol
Acetonylacetone
Ethyl formate
Ethanol
(2.2 1.1)
−(3.7 0.7)
−(1.1 0.5)
(0.6 1.7)
(2.0 0.9)
(2.0 0.8)
(0.5 0.5)
(0.2 1.7)
(1.4 1.0)
−(0.6 0.8)
−(1.2 1.7)
−(1.6 1.7)
−(3.4 1.7)
−(3.4 2.1)
−(3.2 0.6)
−(3.3 1.0)
−(2.6 1.2)
−(2.5 0.5)
−(3.8 0.9)
−(6.9 0.4)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Ethanol
Ethanol
MEK
MEK
MEK
1-Propanol
1-Butanol
(4.5 1.9) × 109
4.2 × 109
(9.1 1.7) × 108
7.8 × 108
(1.4 0.7) × 109
2.8 × 109
3.2 × 109
2-Propanol
Propionaldehyde
Butyraldehyde
Isobutyraldehyde
Isobutyraldehyde
Isobutyraldehyde
Valeraldehyde
Valeraldehyde
MIBK
1.6 × 109
3.1 × 109
(3.6 1.5) × 109
(4.4 1.1) × 109
(4.1 0.2) × 109
9.5 × 109
(6.0 1.1) × 109
(3.8 0.1) × 109
(4.6 0.8) × 108
Figure 6 Comparison of Evans–Polanyi plot in the form of E versus BDE: aqueous phase ( ), this work ( ), and gas phase
◦
a
•
(ꢀ). Numbering refers to the numbers in Table VII for the aqueous-phase and Table VIII for the gas-phase data. For BDE see
Table VI.
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
323
(hydrogen abstraction) occurs in the gas phase and in
the aqueous phase. Yet one can notice the higher dis-
persion of the data in the aqueous phase in comparison
with the gas phase. The latter probably comes from
the limited temperature range in the aqueous phase
(3◦C–80◦C).
a non-Arrhenius behavior, does not show a negative
temperature dependence for the overall reaction.
In the aqueous phase the formation of H-bonded
complex can be hindered due to the already existing
hydrogen bond between the water molecules and the
organic compounds. The presence of hydrogen bonds
between the water molecules and the organics may
influence their reactivity [75]. For instance, the ab initio
calculations performed by Vassilev et al. [76] have
revealed that OH radical in the first solvation shell
is characterized with three H-bonded water molecules
leading to enlargement of its effective size [77].
It should be noted that the negative temperature
dependencies for the above mentioned reactions have
not been observed in the aqueous phase. The latter can
be due to the limited range of temperature studied or
different reaction pathways compared to the gas phase.
Monod et al. [18] speculated that although the overall
mechanism in both phases can be the abstraction of the
hydrogen atom, different reactivity of the OH radical
in both phases comes from the different intermediary
steps.
Moreover, one can notice the negative temperature
dependence for the reaction of the OH radical with 1-
propanol, 1-butanol, 2-propanol, acetonylacetone, and
ethyl formate and with the aldehydes performed in
the gas phase. This observation suggests the possi-
bility that those reactions and especially those of the
aldehydes with OH in general do proceed through an
addition–elimination mechanism [72], since many ad-
dition reactions of OH show overall negative temper-
ature dependence. Atkinson [73] postulated that the
reactions of OH with aldehydes proceed via overall H-
atom abstraction, although the initial reaction possibly
involves the OH radical addition to the >C O function.
However, it is not clear why a hydrogen abstraction re-
action presents negative Ea and why the overall OH
addition to the double bond does not occur. Singleton
and Cvetanovic [74] propose a complex mechanism
and explain the occurrence of these negative activa-
tion energies as being due to the reversible formation
of a loosely bound prereactive complex that is formed
without an Ea, followed by a second reaction, which is
irreversible, and whose transition-state energy is lower
than the energy of the separated reactants. In the recent
review of Mellouki et al., [20], the reaction pathway
of aldehydes toward OH in the gas phase is regarded
unclear. The authors in this review consider all the
possibilities: the reaction can proceed via H-atom ab-
straction, addition to the >C O double bond or initial
formation of a hydrogen-bonded complex followed by
H-atom abstraction.
Smith and Ravishankara [23] concluded that the
dynamics of the reactive encounters between OH and
aldehydes is affected by the hydrogen bond that is
formed between them and that this could be the reason
for the negative temperature dependencies manifested
in those reactions.
Aloisio and Francisco [21] performed ab initio cal-
culations by run of the program GAUSSIAN 94 and
have demonstrated the presence of hydrogen bond
adducts formed in the reactions of OH with aldehy-
des in the gas phase. Furthermore, they are giving
the complete picture of the structure and energetics of
the complexes involving OH radical and the carbonyl-
containing molecules.
Finally in a very recent paper by Monod and
Doussin [77], a structure–activity relationship was de-
veloped for the estimation of OH rate constants in aque-
ous solution as an alternative approach to the Evans–
Polanyi correlations presented in this study.
Atmospheric Implications
Oxidation processes of organic compounds both in the
gas phase and in the aqueous phase lead to less volatile
organics, for example, dicarboxylic acids. As shown in
several studies dicarboxylic acid is ubiquitous in tro-
pospheric particles [78–81]. The addition of their for-
mation and decay mechanisms into models can help
to clarify the composition of the organic matter in at-
mospheric particles. Such interpretations could be of
particular interest because the composition of parti-
cles can have influence on their hygroscopic behavior,
that is, the ability to act as cloud condensation nuclei
[82–84]. Furthermore, the degradation processes might
influence the acidity [82] and the radiative properties
[85] of the atmospheric particles.
The rate constant data obtained in the present study
contribute to a better definition of the tropospheric life-
times of the studied ketones, dicarboxylic acids, and
unsaturated compounds that react predominantly with
the OH radical.
Despite the aforementioned theoretical observa-
tions for the reactions of OH with carbonyl compounds
in the gas phase, it is still not clear why the reaction of
OH with acetone, for instance, which is known to have
A remote scenario, which was chosen as a stan-
dard scenario in CAPRAM 3.0 [1], corresponds to OH
concentration of 1.3 × 10−13 mol L−1 in the aqueous
phase. By using this value, first-order loss processes
International Journal of Chemical Kinetics DOI 10.1002/kin

324
GLIGOROVSKI ET AL.
Table IX The First-Order Loss Rates in [s−1] for OH Radical Reactions in the Aqueous Phase and in the Gas Phase
First-Order Loss Process
in the Aqueous Phase (s−1
First Order Loss Process
in the Gas Phase (s−1
)
Reactant
)
Acetone
2.3 × 10−5
6.4 × 10−4
6.0 × 10−4
1.8 × 10−3
2.1 × 10−3
4.7 × 10−5
2.3 × 10−5
3.2 × 10−5 [61]
N-Methyl pyrrolidone
Methyl isobutyl ketone
Methacrolein
DVE-3
Malic acid
Mesoxalic acid
4.8 × 10−4 [61]
1.2 × 10−3 [86]
OH concentration in the aqueous phase (remote scenario) 1.3 × 10−13 M [3].
OH concentration in the gas phase (standard scenario) 6.0 × 10−14 M [27].
of the organics under study toward OH reaction were
The correlations presented in this study cover a lack
calculated.
Concerning a typical gas phase OH concentration
in the literature and they are in good agreement with
previous studies [4,5].
of 1 × 106 molecule cm−3 it is possible to roughly esti-
mate the impact of OH radical on the reactant’s budget
(Table IX). In Table IX the first-order loss processes
for the reactions of OH with the organics under study
in the aqueous phase and in the gas phase are com-
pared. Note that although the gas-phase kinetic data
set is larger than the aqueous phase, the rate constants
for the reactions of OH with NMP, MIBK, DVE-3, and
malic acid gas-phase data are not available. Hence, the
comparison was not possible in these cases.
In Table IX it can be seen that the first-order loss
processes are of comparable magnitude in both phases
at 298 K, which comes from the respective rate con-
stants discussed in the previous sections of this article.
CAPRAM 3.0 [1], the latest development of
CAPRAM series includes a description of the chem-
istry of oxygenated organic compounds that are used
as solvents and fuel additives such as NMP, MIBK,
and DVE-3. It has been shown that these compounds
or their respective oxidation products might play a role
in the tropospheric aqueous phase in a close interaction
with their respective gas-phase chemistry.
For ketones and dicarboxylic acids considered in
this study it was possible to describe the observed rate
constants in view of a H-abstraction mechanism acting
during the oxidation process.
On the other hand, the reactions of OH with
methacrolein and DVE-3 are proceeding by addition–
elimination mechanism, which is clearly diffusion con-
trolled.
To cover a wider range of situations encountered
in the atmosphere, further experimental and model-
ing investigations are needed to understand the overall
aqueous-phase chemistry.
BIBLIOGRAPHY
1. Herrmann, H. Chem Rev 2003, 103, 4691–4716.
2. Wypych, G. Handbook of Solvents, 1st ed.; ChemTech
Publishing: Toronto, 2001.
3. Herrmann, H.; Tilgner, A.; Majdik, Z.; Barzaghi, P.;
Gligorovski, S.; Poulain, L.; Monod, A. Atmos Environ
2005, 39, 4351–4363.
4. Gligorovski, S.; Herrmann, H. Phys Chem Chem Phys
2004, 6, 4118–4126.
5. Ervens, B.; Gligorovski, S.; Herrmann, H. Phys Chem
Chem Phys 2003, 5, 1811–1824.
6. Chin, M.; Wine, P. H. J Photochem Photobiol A 1992,
69, 17–25.
7. George, C.; Rousse, D.; Perraudin, E.; Strekowski, R.
Phys Chem Chem Phys 2003, 5, 1562–1569.
8. Moise, T.; Rudich, Y.; Rousse, D.; George, C. Environ
Sci Technol 2005, 39, 5203–5208.
9. Milosavljevic, B.; La Verne, J. A. J Phys Chem A 2005,
109, 165–168.
10. Gligorovski, S.; Laser-Based Studies of OH Radical Re-
actions in Aqueous Solution, Ph.D. Dissertation, Univer-
sity of Leipzig, Leipzig, Germany, 2005.
CONCLUSION
Rate constants for the reactions of OH with ketones, di-
carboxylic acids, and unsaturated compounds in aque-
ous solution have been determined.
The obtained rate constants and the activation pa-
rameters were used to improve the already estab-
lished Evans–Polanyi correlations in the form of log
(kH) = f (BDE), Ea = f (BDE). Correlations Eqs.
(16) and (18) were provided for the prediction of
second-order rate constants and activation energies for
the reaction of OH with organic compounds reacting
by H-abstraction mechanism.
International Journal of Chemical Kinetics DOI 10.1002/kin

RATE CONSTANTS FOR THE OH REACTIONS WITH OXYGENATED ORGANIC COMPOUNDS
325
11. White, J. U. J Opt Soc Am 1942, 32, 285–288.
12. Thomas, J. K. Trans Faraday Soc 1965, 61, 702–707.
13. Adams, G. E.; Boag, J. W.; Currant, J.; Michael, B. D.
In Pulse Radiolysis; Ebert, M.; Keene, J. P.; Swallow, A.
J.; Baxendale, J. H. (eds.); Academic Press: New York,
1965, 131.
14. Wilson, G. L.; Greenstock, C. L.; Adams, G. E.;
Wageman, R.; Dorfman, L. M. Int J Radiat Phys Chem
1971, 3, 211.
37. Schuchmann, M. N.; Schuchmann, H. P.; Hess, M.; von
Sonntag, C. J Am Chem Soc 1991, 113, 6934–6937.
38. BASF—The Chemical Company. http://worldaccount.
basf.com/wa/NAFTA/Catalog/ChemicalsNAFTA/info/
BASF/PRD/30037191, accessed 24 May 2004.
39. Connors, K. A. Chemical Kinetics, 3rd ed.; VCH Pub-
lishers: New York, 1990.
40. Jenkin, M. E.; Saunders, S. M.; Pilling, M. J. Atmos
Environ 1997, 31, 81–104.
15. Wolfenden, B. S.; Willson, R. L. J Chem Soc Perkin
Trans 1982, 2, 805–812.
41. Rasmussen, R. A. J Air Pollut Control Assoc 1972, 22,
537–543.
16. Hesper, J. Spektroskopishe und kinetische Untersuchun-
gen von Reaktionen der Radikale O2- und OH in
wa¨ssriger Phase, Ph.D. Dissertation, University of
Leipzig, Leipzig, Germany, 2003.
17. Poulain, L.; Sowka, I.; Monod, A.; Wortham, H. Pro-
ceedings from the EC/EUROTRAC-2 CMD Joint Work-
shop, Shaping the Future of Atmospheric Chemistry Re-
search in Europe, Paris, 2002, p. 52.
18. Monod, A.; Poulain, L.; Grubert, S.; Voisin, D.;
Wortham, H. Atmos Environ 2005, 39, 7667–
7688.
19. Poulain, L.; Monod, A.; Wortham, H. J Photochem Pho-
tobiol A Chemistry 2007, 187, 10–23.
42. Atkinson, R. Chem Rev 1986, 86, 69–201.
43. Lilie, J.; Henglein, A. Ber Bunsenges Physik Chem
1970, 74, 388.
44. Orlando, J. J.; Tyndall, G. S.; J Phys Chem A 2002, 106,
12252–12259.
45. Evans, M. G.; Polanyi, M. Trans Faraday Soc 1938, 34,
11.
46. Morozov, I.; Gligorovski, S.; Barzaghi, P.; Hoffmann,
D.; Lazarou, Y. G.; Vasiliev, E.; Herrmann, H. Int J
Chem Kinet 2008, 40, 174–188.
47. Schuchmann, M. N.; Von Sonntag, C. J Am Chem Soc
1988, 110, 5698–5701.
48. Bott, J. F.; Cohen, N. Int J Chem Kinet 1991, 23, 1017–
1033.
49. Saunders, S. M.; Baulch, D. L.; Cooke, K. M.; Pilling,
M. J.; Smurthwaite, P. I. Int J Chem Kinet 1994, 26,
113–130.
50. Dagaut, P.; Wallington, T. J.; Liu, R.; Kurylo, M. J. J
Phys Chem 1988, 92, 4375–4377.
51. Overend, R.; Paraskevopoulos, G. J Phys Chem 1978,
82, 1329–1333.
52. Winer, A.; M.; Lloyd, A. C.; Darnall, K. R.; Pitts, J. N.
J Phys Chem 1976, 80, 1635–1639.
53. Nelson, L.; Rattigan, O.; Neavyn, R.; Sidebottom, H.;
Treacy, J.; Nielsen, O. J. Int J Chem Kinet 1990, 22,
1111–1126.
54. Baxley, J. S.; Wells, J. R. Int J Chem Kinet 1998, 30,
745–752.
55. Balestra-Garcia, C.; Le Bras, G.; MacLeod, H. J Phys
Chem 1992, 96, 3312–3316.
56. D’Anna, B.; Andresen, O.; Gefen, Z.; Nielsen, C. J. Phys
Chem Chem Phys 2001, 3, 3057–3063.
57. Audley, G. J.; Baulch, D. L.; Campbell, I. M. J Chem
Soc Faraday Trans 1981, 77, 2541.
20. Mellouki, A.; LeBras, G.; Sidebottom, H. Chem Rev
2003, 103, 5007–5096.
21. Aloisio, S.; Francisco, J. S. J Phys Chem A 2000, 104,
3211–3224.
´
22. Vasva´ri, G.; Szila´gyi, I.; Bencsura, A.; Do´be´, S.; Be´rces,
T.; Henon, E.; Canneaux, S.; Bohr, F. Phys Chem Chem
Phys 2001, 3, 551–555.
23. Smith, I. W. M.; Ravishankara, A. R. J Phys Chem A
2002, 106, 4798–4807.
24. Gierczak, T.; Gilles, M. K.; Bauerle, S.; Ravishankara,
A. R. J Phys Chem A 2003, 107, 5014–5020.
25. Donahue, N. M. Chem Rev 2003, 103, 4593–4604.
26. Stoye, D.; Freitag, W. Paints, Coatings and Solvents,
2nd ed.; Wiley: New York, 1998.
27. Jenkin, M. E.; Saunders, S. M.; Derwent, R. G.; Pilling,
M. J. Atmos Environ 2002, 36, 4725–4734.
28. Benson, S. W. Thermochemical Kinetics, 2nd ed.;
Wiley: New York, 1976.
29. Howard, P. H. Handbook of Environmental Fate and
Exposure Data for Organic Chemicals; CRC Press:
New York, 1998.
30. Friesen, D. A.; Headley, J. V.; Langford, C. H. Environ
Sci Technol 1999, 33, 3193–3198.
31. Lund Myhre, C. E.; Nielsen, C. J. Atmos Chem Phys
2004, 4, 1759–1769.
32. Ervens, B.; Feingold, G.; Clegg, S. L.; Kreidenweis,
S. M. J Geophys Res 2004, 109, 15206–15218.
33. Varga, Z.; Kiss, G.; Hansson, H.-C. Atmos Chem Phys
Discuss 2007, 7, 5341–5364.
58. Kerr, J. A. Strengths of Chemical Bonds. In Handbook
of Chemistry and Physics; Lide, D. R. (ed.); CRC Press:
New York, 1995; pp. 9–51.
59. Niiranen, J. T.; Gutman, D.; Krasnoperov, L. N. J Phys
Chem 1992, 96, 5881–5886.
60. Rousse, D. Chimie Multiphasique des Composes Or-
ganiques Volatils, Ph.D. Dissertation, University of
Lyon, Lyon, France, 2005.
34. Scholes, G.; Willson, R. L. Trans Faraday Soc 1967, 63,
2983–2993.
35. Logan, S. R. J Chem Soc Perkin Trans 1989, 2, 751–754.
36. Lide, D. R. Handbook of Chemistry and Physics; CRC
Press: New York, 1995.
61. Atkinson, R.; Baulch, D. L.; Cox, R.; Hampson, A. R.
F. Jr.; Kerr, J. A.; Rossi, M. J.; Troe, J. J Phys Chem Ref
Data 1997, 26, 521–1011.
62. Le Calve, S.; Hitier, D.; Le Bras G.; Mellouki, A. J Phys
Chem A 1998, 102, 4579–4584.
International Journal of Chemical Kinetics DOI 10.1002/kin

326
GLIGOROVSKI ET AL.
63. DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson,
R. F.; Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.;
Kolb, C. E.; Molina, M. J. JPL Publication 97-4, 1997,
1–266.
74. Singleton, D. L.; Cvetanovic, R. J. J Am Chem Soc 1976,
98, 6812–6819.
75. Rudakov, E. S.; Volkova, L. K.; Tret’yakov, V. P. React
Kinet Catal Lett 1981, 16, 333–337.
64. Wallington, T. J.; Kurylo, M. J. J Phys Chem 1987, 91,
5050–5054.
76. Vassilev, P.; Louwerse, M. J.; Baerends, E. J. Chem Phys
Lett 2004, 398, 212–216.
65. Wallington, T. J.; Dagaut, P.; Liu, R.; Kurylo, M. J.
Environ Sci and Technol 1988, 22, 842–844.
66. Teton, S.; Mellouki, A.; Le Bras, G.; Sidebottom, H. Int
J Chem Kinet 1996, 28, 291–297.
67. Le Calve, S.; Le Bras, G.; Mellouki, A. J Phys Chem A
1997, 101, 5489–5493.
77. Monod, A.; Doussin. J. F. Atmos Environ 2008, 42,
7611–7622.
78. Narukawa, M.; Kawamura, K.; Li, S.-M.; Bottenheim,
J. W. Atmos Environ 2002, 36, 2491–2499.
79. Neusu¨ß, C.; Pelzing, M.; Plewka, A.; Herrmann, H. J
Geophys Res 2000, 105(D4), 4513–4527.
68. Wallington, T. J.; Kurylo, M. J. Int J Chem Kinet 1987,
19, 1015–1023.
69. Thevenet, R.; Mellouki, A.; Le Bras, G. Int J Chem Kinet
2000, 32, 676–685.
70. Semmes, D. H.; Ravishankara, A. R.; Gump-Perkins,
C. A.; Wine, P. H. Int J Chem Kinet 1985, 17, 303–
313.
71. Dobe, S.; Khachatryan, L. A.; Berces, T. Ber Bunsenges
Phys Chem 1989, 93, 847–852.
80. Plewka, A.; Gnauk, T.; Bruggemann, E.; Neusu¨ß, C.;
Herrmann, H. J Atm Chem 2004, 48, 131–156.
81. Neusu¨ß, C.; Gnauk, T.; Plewka, A.; Herrmann, H.;
Quinn, P. K. J Geophys Res 2002, 107(D19).
82. Novakov, T.; Penner, J. E. Nature 1993, 365, 823–826.
83. Saxena, P.; Hildemann, L. M.; McMurry, P. H.; Seinfeld,
J. H. J Geophys Res 1995, 100, 18755–18770.
84. Abdul-Razzak, H.; Ghan, S. J.; Rivera-Carpio, C. A. J
Geophys Res 1998, 103, D6, 6123–6131.
72. Michael, J. V.; Keil, D. G.; Klemm, R. B. J Chem Phys
1985, 83, 1630–1636.
85. Shulman, M. L.; Charlson, R. J.; Davis, E. J. J Aerosol
Sci 1997, 28, 737–752.
73. Atkinson, R. J Phys Chem Ref Data 1994, Monograph
2, 61.
86. Chuong, B.; Stevens, P. Int J Chem Kinet 1985, 36,
12–25.
International Journal of Chemical Kinetics DOI 10.1002/kin