Importance of Unimolecular HO2 Elimination in the Heterogeneous OH Reaction of Highly Oxygenated Tartaric Acid Aerosol

Article abstract of DOI:10.1021/acs.jpca.6b05289

Oxygenated organic molecules are abundant in atmospheric aerosols and are transformed by oxidation reactions near the aerosol surface by gas-phase oxidants such as hydroxyl (OH) radicals. To gain better insights into how the structure of an organic molecule, particularly in the presence of hydroxyl groups, controls the heterogeneous reaction mechanisms of oxygenated organic compounds, this study investigates the OH-radical initiated oxidation of aqueous tartaric acid (C₄H₆O₆) droplets using an aerosol flow tube reactor. The molecular composition of the aerosols before and after reaction is characterized by a soft atmospheric pressure ionization source (Direct Analysis in Real Time) coupled with a high-resolution mass spectrometer. The aerosol mass spectra reveal that four major reaction products are formed: a single C₄ functionalization product (C₄H₄O₆) and three C₃ fragmentation products (C₃H₄O₄, C₃H₂O₄, and C₃H₂O₅). The C₄ functionalization product does not appear to originate from peroxy radical self-reactions but instead forms via an α-hydroxylperoxy radical produced by a hydrogen atom abstraction by OH at the tertiary carbon site. The proximity of a hydroxyl group to peroxy group enhances the unimolecular HO₂ elimination from the α-hydroxylperoxy intermediate. This alcohol-to-ketone conversion yields 2-hydroxy-3-oxosuccinic acid (C₄H₄O₆), the major reaction product. While in general, C-C bond scission reactions are expected to dominate the chemistry of organic compounds with high average carbon oxidation states (OS_C), our results show that molecular structure can play a larger role in the heterogeneous transformation of tartaric acid (OS_C = 1.5). These results are also compared with two structurally related dicarboxylic acids (succinic acid and 2,3-dimethylsuccinic acid) to elucidate how the identity and location of functional groups (methyl and hydroxyl groups) alter heterogeneous reaction mechanisms.

Full text of DOI:10.1021/acs.jpca.6b05289

Article
pubs.acs.org/JPCA
Importance of Unimolecular HO2 Elimination in the Heterogeneous
OH Reaction of Highly Oxygenated Tartaric Acid Aerosol
Chiu Tung Cheng,† Man Nin Chan,*,†,‡ and Kevin R. Wilson*,§
‡
†Earth System Science Programme, Faculty of Science and The Institute of Environment, Energy and Sustainability, The Chinese
University of Hong Kong, Hong Kong, China
§Chemical Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, California, United States
S
* Supporting Information
ABSTRACT: Oxygenated organic molecules are abundant in
atmospheric aerosols and are transformed by oxidation reactions
near the aerosol surface by gas-phase oxidants such as hydroxyl (OH)
radicals. To gain better insights into how the structure of an organic
molecule, particularly in the presence of hydroxyl groups, controls the
heterogeneous reaction mechanisms of oxygenated organic com-
pounds, this study investigates the OH-radical initiated oxidation of
aqueous tartaric acid (C4H6O6) droplets using an aerosol flow tube
reactor. The molecular composition of the aerosols before and after
reaction is characterized by a soft atmospheric pressure ionization
source (Direct Analysis in Real Time) coupled with a high-resolution
mass spectrometer. The aerosol mass spectra reveal that four major
reaction products are formed: a single C4 functionalization product
(C4H4O6) and three C3 fragmentation products (C3H4O4, C3H2O4,
and C3H2O5). The C4 functionalization product does not appear to originate from peroxy radical self-reactions but instead forms
via an α-hydroxylperoxy radical produced by a hydrogen atom abstraction by OH at the tertiary carbon site. The proximity of a
hydroxyl group to peroxy group enhances the unimolecular HO2 elimination from the α-hydroxylperoxy intermediate. This
alcohol-to-ketone conversion yields 2-hydroxy-3-oxosuccinic acid (C4H4O6), the major reaction product. While in general, C−C
bond scission reactions are expected to dominate the chemistry of organic compounds with high average carbon oxidation states
(OSC), our results show that molecular structure can play a larger role in the heterogeneous transformation of tartaric acid (OSC
= 1.5). These results are also compared with two structurally related dicarboxylic acids (succinic acid and 2,3-dimethylsuccinic
acid) to elucidate how the identity and location of functional groups (methyl and hydroxyl groups) alter heterogeneous reaction
mechanisms.
The structure of an organic molecule (e.g., branching of the
carbon backbone and average carbon oxidation state (OSC))
plays a key role in controlling heterogeneous chemistry.7,9
Recently, we investigated the heterogeneous OH oxidation of
succinic acid and its more branched analogue (2,3-dimethyl-
succinic acid (2,3-DMSA)) to examine how branched methyl
groups alter reaction mechanisms.10,11 Figure 1 shows reaction
products with new oxygenated groups (i.e., “functionalization”)
are formed in the OH oxidation with succinic acid. The
formation of these products is explained by the self-reactions of
two peroxy radicals to form stable products (e.g., ketones and
alcohols) via the Russell12 and/or Bennett−Summer reaction
mechanisms.13 When the hydrogen atoms on succinic acid are
substituted with methyl groups, as in the OH reaction with 2,3-
DMSA (Figure 1), the two methyl groups sterically hinder the
appropriate arrangement of two peroxy radicals into a cyclic
1. INTRODUCTION
Atmospheric aerosols are suspensions of small particles (solid
or liquid) in air. These particles interact directly and indirectly
with sunlight, affecting the Earth’s radiation balance. Directly,
aerosols absorb and scatter sunlight, thus affecting the global
radiative balance. Indirectly, they modify the size and number
distribution of cloud droplets, altering how cloud droplets
reflect and absorb sunlight.1 Moreover, atmospheric aerosols
can have adverse effects on our respiratory and cardiovascular
systems.2
Atmospheric aerosols are chemically complex; composed of
inorganic salts and a large variety of organic compounds.
Organic compounds are a significant fraction (∼50%) of the
mass of atmospheric aerosols3 and can be oxidized at the
aerosol surface by gas-phase hydroxyl radicals (OH), ozone
(O3), and nitrate radicals.4 These heterogeneous oxidative
processes continuously alter the surface and bulk composition
of the aerosol, and thus modify aerosol properties such as light
extinction, hygroscopicity, and cloud condensation nuclei
activity.5−8
Received: May 25, 2016
Revised: July 4, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
DMSA) to examine more broadly how the identity of
functional groups (methyl and hydroxyl groups) alter
heterogeneous OH reaction mechanisms in these three
structurally related dicarboxylic acids (Figure 1).
2. METHOD
2.1. Aerosol Flow Tube Reactor. The heterogeneous OH
oxidation of tartaric acid aerosols is investigated using an
aerosol flow tube reactor. Experimental details have been
described elsewhere.10 In brief, aqueous droplets are first
generated using a constant output atomizer (TSI Inc. model
3076). The aerosol stream is then mixed with humidified
nitrogen (N2), oxygen (O2), trace hexane and O3, and
introduced to the reactor. The relative humidity (RH)
measured at the inlet of the reactor is ca. 85.8 0.26%. The
temperature inside the reactor is maintained at 20 °C using a
circulating water jacket surrounding the reactor. Since neither
deliquescence nor crystallization is observed for tartaric acid
aerosols,16 we assume that the tartaric acid aerosols exist as
aqueous droplets before oxidation. The size distribution of the
droplets leaving the reactor is determined by a scanning
mobility particle sizer (SMPS; TSI, 3936). Before oxidation, the
mean surface weighted diameter for the droplets is measured to
be 188.4 0.5 nm. The initial aerosol mass loading is ∼2800
μg m−3. At RH = 85.8%, tartaric acid droplets are predicted to
have a molarity of 3.27 M and a mass fraction of solute (mf) of
0.49.17 The average density of the droplets (ρ) is estimated to
be 1.38 g cm−3 using an additivity rule with the known density
and mass fraction of tartaric acid and water.
Tartaric acid aerosols are oxidized by gas-phase OH radicals,
which are generated by the photolysis of O3 using ultraviolet
lamps at 254 nm. The OH concentration is adjusted by varying
the amount of O3 added to the reactor and is quantified by
measuring the consumption of a gas phase tracer (i.e., hexane)
using a gas chromatography equipped with a flame ionization
detector.18 With a total flow rate of 2 L/min (an aerosol
residence time of 1.3 min), the OH exposure (OH
concentration × aerosol residence time) ranges from 0 to
4.63 × 1011 molecules cm−3 s. (The method of measuring the
OH exposure is described in Supporting Information.) After it
exits the reactor, the aerosol stream passes through an annular
Carulite catalyst denuder to remove O3 and an activated
charcoal denuder to remove any gas-phase products. A portion
of the aerosol stream is sampled by an SMPS for aerosol size
measurements. The remaining flow is introduced into an
ionization region, which is a narrow open space between a
DART ionization source and an atmospheric inlet of a high-
resolution mass spectrometer (ThermoFisher, Q Exactive
Orbitrap), for real-time chemical characterization of the
aerosols.
Figure 1. Molecular structure and proposed reaction mechanisms of
2,3-dimethylsuccinic acid (A), succinic acid (B), and tartaric acid (C)
during heterogeneous OH oxidation. Major functionalization and
fragmentation products are shown. The percentage listed indicates the
abundance of total fragmentation and functionalization products at a
similar OH exposure (∼4 × 1011 molecules cm−3 s).
tetroxide intermediate, which is necessary for the formation of
stable ketone and alcohol products. Instead, alkoxy radicals are
formed from peroxy radical self-reactions and subsequently
abstract hydrogen atoms from neighboring molecules to form
alcohol functionalization products. Moreover, the abundance of
these functionalization products is larger than the C−C bond
scission fragmentation products. An opposite result is observed
for succinic acid. These results illustrate how the differences in
molecular structure can alter reaction pathways.
Recent studies suggest that fragmentation becomes more
favorable when the degree of oxygenation (e.g., OSC) of an
organic molecule increases.9,14 However, a robust relationship
between the degree of oxygenation and volatilization (a proxy
for the fragmentation process) has not been established. For
example, little aerosol carbon mass loss has been reported for
more oxygenated citric acid (OSC = 1), while a significant
production of volatile products has been observed for less
oxygenated erythritol (OSC = −0.5) during OH oxidation.15
Depending on the number, types, and positions of the
functional groups, organic compounds with the same or a
similar degree of oxygenation (e.g., OSC) can chemically evolve
via distinct mechanisms.
To better understand how the molecular structure governs
the heterogeneous chemistry of oxygenated organic aerosols,
we investigate the OH-initiated oxidation of tartaric acid
(C4H6O6) aerosols using an aerosol flow tube reactor. The
composition of the aerosols before and after oxidation is
characterized by a soft atmospheric pressure ionization source
(direct analysis in real time (DART)) coupled with a high-
resolution mass spectrometer in real time. On the basis of the
identification of reaction products from exact mass measure-
ments, we measure the oxidation kinetics and investigate
reaction mechanisms to examine how the presence of hydroxyl
groups controls the competition between the functionalization
and fragmentation pathways. Results from this work are
compared with our previous studies (succinic acid and 2,3-
2.2. Atmospheric Pressure Aerosol Mass Spectrom-
eter. The DART ionization source (IonSense: DART
SVP)19,20 is operated in negative ionization mode with
metastable helium as the ionizing reagent. The heater inside
the DART ionization source is operated at 500 °C. Prior to
entering the ionization region, the aerosols are vaporized in a
heater to generate gas-phase species, which are then ionized by
the DART source. Proton abstraction from the carboxyl group
of tartaric acid and its reaction products produces the
deprotonated molecular ions [M−H]− observed in the aerosol
mass spectra.21,22 Detailed ionization mechanism is given in the
Supporting Information. The resulting ions are sampled by the
high-resolution mass spectrometer. Mass spectra are collected
B
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
at 1 s intervals over a scan range from m/z 70−700. Each mass
spectrum is averaged for 5 min at a mass resolution of
∼140 000. Mass calibration is performed with standard
solutions before experiments. The mass spectra are analyzed
using the Xcalibar software (Xcalibar Software, Inc., Herndon,
VA, USA).
3. RESULTS AND DISCUSSIONS
In the following sections, we first analyze the heterogeneous
oxidative kinetics of tartaric acid (Section 3.1). On the basis of
the aerosol mass spectra, reaction pathways are proposed
(Section 3.2) to explain the formation of major reaction
products and to explore the role of hydroxyl groups in the
reaction mechanism. Finally, we discuss how different types of
functional groups control the heterogeneous OH chemistry of
three small structurally related dicarboxylic acids (succinic acid,
2,3-DMSA, and tartaric acid; Section 3.3).
3.1. Oxidation Kinetics. Figure 2 shows that the reaction
of OH with tartaric acid produces four major new peaks in the
Figure 3. Heterogeneous OH oxidation of tartaric acid aerosols. (A)
Normalized decay or tartaric acid. (B) Chemical evolution of the
reaction products as a function of OH exposure. (C) The total relative
abundance of the functionalization and fragmentation products. Error
bars of 1 σ are shown.
where D0 and ρ are the mean surface-weighted diameter and
the density of the droplets before oxidation, respectively. Mw is
the molecular weight of tartaric acid, mf is mass fraction of
solute in the droplets, NA is Avogadro’s number, and cOH is
the mean speed of gas-phase OH radicals. Using eq 1, γOH is
Figure 2. Aerosol mass spectra (A) before oxidation and (B) after
oxidation (at OH exposure = 4.63 × 1011 molecules cm−3 s). A
computed to be 1.16
(0.40
0.13) reported by Kessler et al.15 One possible
explanation for this discrepancy is that the marker ion
0.19, which is larger than the value
chemical formula is assigned for individual ion with an error of
5
ppm. The ion peaks for the major reaction products are shown in (C)
over an enlarged scale. The unlabeled peaks contribute less than 1% of
the total signal and were detected in the background.
15
+
(C4H2O3 , m/z = 98) used by Kessler et al. is not unique
for tartaric acid but rather has contributions from reaction
products that fragment during the electron impaction
ionization. This may lead to an overestimate of the
concentration of unreacted tartaric acid, leading to smaller
spectrum. The largest product peak corresponds to C4H4O6,
which is a C4 functionalization product. The remaining peaks
correspond to three distinct C3 fragmentation products
(C3H4O4, C3H2O4, and C3H2O5). Figure 3A shows the reactive
decay of tartaric acid (I/Io) at different OH exposures. At the
highest OH exposure (4.63 × 1011 molecules cm−3 s), ∼60% of
parent tartaric acid is consumed. The decay of tartaric acid is fit
to an exponential function to obtain a second-order
heterogeneous OH reaction rate constant (kOH).18 The kOH is
determined to be (2.04 0.16) × 10−12 cm3 molecule−1 s−1.
Using the fitted kOH value, the effective OH uptake coefficient,
γOH, defined as the fraction of OH collisions that yields a
reaction, is computed using the following equation:23
k
OH. However, a more likely explanation for this difference is
the experimental conditions. Here the reaction is measured at a
higher relative humidity (RH = 85.8%) than those of Kessler et
al.15 (RH = 30%). As shown by Davies and Wilson,23 Houle et
al.,24 Slade and Knopf,25 and Shiraiwa and co-workers,26,27 the
OH reactive uptake exhibits a complex dependence on aerosol
viscosity, which in turn is controlled by RH. At high RH,
tartaric acid aerosols are likely aqueous droplets with molecular
components that can diffuse more rapidly to the aerosol surface
for reaction with OH, leading to higher overall reaction
rates.10,25 At low RH, the aerosols become more concentrated
and viscous.28 In this case, diffusion of the species from the bulk
to the surface is slower, and thus tartaric acid in the interior of
the aerosol is less likely to undergo oxidation at the surface. For
2D0ρmf N
3Mw cOH
γOH
=
A kOH
(1)
C
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
a
Scheme 1. Proposed Reaction Mechanisms for the Heterogeneous OH Oxidation of Tartaric Acid
a
The observed reaction products are shown in boxes. Major reaction pathways are highlighted in black, and minor ones are shown in grey.
Scheme 2. Mechanism of the Unimolecular HO2 Elimination Process in an α-Hydroxylperoxy Radical
D
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
example, Davies and Wilson23 investigated the relationship
between RH and the reactive uptake of OH radicals onto citric
acid (C6H8O7) droplets. They observed that the γOH decreases
by about a factor of 2 (∼0.112 to ∼0.052), when the RH
decreases from 90% to 20%, and suggested that at low humidity
the reactive OH decay of citric acid is controlled by aerosol
viscosity, with reactions occurring near the aerosol/air interface.
More work is needed to elucidate the effect of humidity on the
heterogeneous oxidative kinetics of tartaric acid.
detected the formation of oxalic acid in the OH oxidation of
succinic acid.10 Furthermore, the effective vapor pressure (c*)
of tartronic acid and oxalic acid is estimated to be ∼12 and 797
μg m−3, respectively, using the saturation vapor pressures
reported in the literature.33,34 As a first approximation,
assuming an equilibrium gas/aerosol partitioning, a significant
mass fraction of tartronic acid (99.6%) and oxalic acid (78.2%)
would remain in the aerosol phase under our experimental
conditions. The absence of these two products in the aerosol
mass spectra further supports that rapid unimolecular HO2
elimination is the dominant reaction pathway for the α-
hydroxylperoxy radical.35
When hydrogen abstraction occurs on the hydroxyl group
(Scheme 1A: Path IB), the resultant alkoxy radical can
decompose to form 2-hydroxy-3-oxopropanoic acid
(C3H4O4), which is one of the major fragmentation products
seen in Figure 2 and contributes to ∼6.6% of the total ion
signal at OH exposure = 4.63 × 1011 molecules cm−3 s (Figure
3B). Although the bond energy of O−H is typically larger than
that of C−H,36 the detection of 2-hydroxy-3-oxopropanoic acid
suggests that the hydrogen abstraction on the hydroxyl group
likely occurs. In competition with decomposition, the alkoxy
radical can react with O2 to form the functionalization product,
2-hydroxy-3-oxosuccinic acid (C4H4O6), or abstract a hydrogen
atom from a neighboring molecule.
3.2. Reaction Mechanisms. On the basis of the aerosol
mass spectra and previously reported aerosol-phase reaction
pathways,5,29 reaction schemes are proposed to explain the
formation of the major products observed in Figure 2. Figure
3B shows the kinetic evolution of these reaction products at
different OH exposures. The relative abundance of all products
increases with OH exposure and reaches a maximum at the
highest OH exposure. To the best of our knowledge, authentic
standards or surrogates are not available for these reaction
products. As a first approximation, the ionization efficiencies of
tartaric acid and all products are assumed to be equal. We
acknowledge that certain products (e.g., organic peroxides and
oligomers) may not be detected efficiently by the DART.10,11
3.2.1. OH Oxidation of Tartaric Acid: First Generation
Products. Scheme 1A shows the OH oxidation with tartaric
acid can be initiated by hydrogen abstraction at either a tertiary
backbone carbon site (Scheme 1A: Path IA) or a hydroxyl
group (Scheme 1A: Path IB). When the hydrogen abstraction
occurs on the tertiary carbon site (Scheme 1A: Path IA), an
alkyl radical forms and reacts quickly with O2 to form an α-
hydroxylperoxy radical, which can undergo unimolecular HO2
elimination to form 2-hydroxy-3-oxosuccinic acid (C4H4O6),
the major functionalization product observed in the aerosol
mass spectrum. The relative abundance of this product is 16.5%
at an OH exposure = 4.63 × 1011 molecules cm−3 s (Figure
3B). As illustrated in Scheme 2A, an α-hydroxylperoxy radical
can adopt a cyclic transition state and yield a carbonyl product
together with a HO2 radical, which can further dissociate into
H+ and O•2− in an aqueous solution.30 Bothe et al.30−32
examined the OH reaction with a series of alcohols with
differing numbers of hydroxyl groups in dilute aqueous
solution. They found that alcohols with multiple hydroxyl
groups (e.g., meso-erythritol, methyl α-D-glucoside, and D-
As a first approximation, the concentration of dissolved O2
([O2]aq) is in the order of 1 × 1017 molecules cm−3 under one
atmospheric pressure based on the Henry’s law constant.
Taking the rate constant of a second order O2 reaction with an
alkoxy radical as 1 × 10−14 cm3 molecule−1 s−1, suggested by
Atkinson et al.,37 the pseudo-first-order rate constant
(kO )[O2]aq) is estimated to be ∼1 × 103 s−1. For the
2
intermolecular hydrogen abstraction, the concentration of the
molecules available for hydrogen abstraction, [RH], is assumed
to be equal to that of tartaric acid before oxidation ([RH] is in
order of 1 × 1021 molecules cm−3). The hydrogen abstraction
rate of an alkoxy radical is estimated to be 1 × 10−16 cm3
molecule−1 s−1.38 This gives the pseudo-first-order rate constant
(kRO+RH [RH]) of ∼1 × 105 s−1. The unimolecular
decomposition rate of an alkoxy radical, kdes, can be estimated
using a structure-activity relationship (SAR) proposed for the
decomposition of gas-phase alkoxy radicals.39,40 The SAR
model predicts a large kdes (1 × 1013 s−1), because the hydroxyl
and carbonyl groups on the leaving radicals greatly enhance
radical stability and lower the energy barrier for decomposition.
This simple analysis suggests that when the hydrogen
abstraction occurs on the hydroxyl group, the resultant alkoxy
radical will most likely decompose to form 2-hydroxy-3-
oxopropanoic acid (C3H4O4, Scheme 1A: Path IB), while
hydrogen abstraction at the tertiary carbon site forms 2-
hydroxy-3-oxosuccinic acid (C4H4O6) produced via unim-
olecular HO2 elimination from the α-hydroxylperoxy inter-
mediate (Scheme 1A: Path IA).
xylitol) exhibit faster unimolecular HO2 elimination rates (kHO
2
= 2000−3000 s−1) than alcohols with a single hydroxyl group
(e.g., methanol, ethanol, and propan-2-ol; kHO = < 10−650
2
s−1). They proposed that for polyols, the hydroxyl group
adjacent to the peroxy group can stabilize the resultant carbonyl
group by forming strong intramolecular hydrogen bond, thus
enhancing the unimolecular HO2 elimination rate. As shown in
Scheme 2B, the unimolecular HO2 elimination from the α-
hydroxylperoxy radical formed on tartaric acid may be similarly
enhanced due to hydrogen bonding to the hydroxyl group
located at the β-position of the peroxy group.
It is also potentially possible that hydrogen abstraction might
occur on the hydroxyl group of the carboxylic acid. This
process could lead to the decarboxylation of the acid and after
subsequent addition of O2 to the resulting radical form a
primary α-hydroxylperoxy intermediate, which can in turn
eliminate HO2 to produce 2-hydroxy-3-oxopropanoic acid
(C3H4O4). However, abstracting a hydrogen atom from the
carboxyl group is expected to be less favorable than from the
hydroxyl group.41
In addition to unimolecular HO2 elimination, the α-
hydroxylperoxy radical can in principle react with another
peroxy radical. However, reaction products (see Scheme 1)
formed from these self-reactions are not detected. For example,
the self-reactions of two peroxy radicals can yield two alkoxy
radicals, which can decompose to form tartronic acid (C3H4O5)
and oxalic acid (C2H2O4). Similar to tartaric acid, these two
dicarboxylic acids should be easily ionized by DART if
produced in significant quantities. For example, we previously
E
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
a
Table 1. Molecular Structures and Volatilities of Major Reaction Products Formed in the Heterogeneous OH Oxidation of
Tartaric Acid
a
Saturation vapor pressure vp and the effective vapor pressure c*. The proposed hydrogen abstraction site is either on the tertiary carbon site
(indicated by “−CH”) or the hydroxyl group (indicated by “−OH”).
3.2.2. OH Oxidation of 2-Hydroxy-3-oxosuccinic Acid:
Second-Generation Products. When the reaction proceeds
further, 2-hydroxy-3-oxosuccinic acid (C4H4O6), the most
abundant first-generation product, can be oxidized to form
two minor fragmentation products (C3H2O4 and C3H2O5). As
shown in Scheme 1B, for the OH reaction with 2-hydroxy-3-
oxosuccinic acid, there are two hydrogen atoms available for
abstraction: one at the tertiary carbon site (Scheme 1B: Path
IIA) and one from the hydroxyl group (Scheme 1B: Path IIB).
When the abstraction occurs on the tertiary carbon site
(Scheme 1B: Path IIA), an O2 molecule can quickly react with
an alkyl radical to form an α-hydroxylperoxy radical. However,
in contrast to the α-hydroxylperoxy radical formed from the
OH oxidation with tartaric acid (Scheme 1A: Path IA), the
product of unimolecular HO2 elimination, dioxosuccinic acid
(C4H2O6), is not observed (Scheme 1B: Path IIA). Instead, an
ion in the aerosol mass spectrum corresponding to mesoxalic
acid (C3H2O5) is detected (Figure 2). The most likely
formation route for mesoxalic acid (C3H2O5) is the reaction
of the α-hydroxylperoxy radical with another peroxy radical to
form two tertiary alkoxy radicals, which in turn decompose to
form mesoxalic acid (C3H2O5).
This difference between first- and second-generation
products originating from the same type of α-hydroxylperoxy
intermediate can be explained by differences in molecular
structure. The carbonyl, on the second-generation α-hydrox-
ylperoxy radical (Scheme 1B: Path IIA), adjacent to the peroxy
group, cannot form an intramolecular hydrogen bond with
resulting carbonyl group in the reaction product (shown in
Scheme 2C). Thus, it is likely that the peroxy radical formed on
2-hydroxy-3-oxosuccinic acid exhibits much slower unimolec-
ular HO2 elimination rates than that of tartaric acid. For OH
reactions with alcohols with single or multiple hydroxyl groups,
Bothe et al.30−32 reported that the unimolecular HO2
elimination rates are slowed by 1 to 3 orders of magnitude in
an absence of a hydroxyl group adjacent to the peroxy group.
These results illustrate how molecular structure (i.e., the
potential for intramolecular hydrogen bond formation) plays a
key role in governing the competition between unimolecular
(HO2 elimination) and bimolecular (RO2 + RO2) reaction
steps in aqueous tartaric acid aerosol reactions.
radical is 2,3-dioxopropanoic acid (C3H2O4). This is expected,
since the decomposition rate of the activated alkoxy radicals is
estimated to be very fast (∼1 × 1011 s−1).39,40 Like tartaric acid,
the intermolecular hydrogen abstraction (∼1 × 105 s−1) and O2
reaction with the alkoxy radical (∼1 × 103 s−1) are predicted to
be much less competitive with the decomposition of the alkoxy
radical for 2-hydroxy-3-oxosuccinic acid.
The kinetic measurements show that γOH (1.16 0.19) is
slightly larger than 1, suggesting the possibility of secondary
chemistry. One possibility for radical chain propagation is the
intermolecular hydrogen abstraction by an alkoxy radical (RO
+RH). However, the time-constant analysis presented above
suggests that this reaction does not compete with decom-
position, since in each case activated alkoxy radicals are formed.
This is in contrast to the formation of alkoxy radical in chemical
reduced systems (e.g., alkanes), which preferentially react via
intermolecular hydrogen abstraction (i.e., RO + RH).42
Another possibility for radical chain propagation chemistry
that is consistent with the mechanism outlined above is the
formation and subsequent reaction of O•2− ions. These species
are formed from the dissociation of HO2 radicals and may
abstract hydrogen atoms from tartaric acid and its reaction
products.43 However, it is noted that the in situ acidity of
tartaric acid droplets is not known. At 85% RH, the
concentration of tartaric acid is ∼3.27 M. From the acidity
constants (pKa1 = 2.89 and pKa2 = 4.40), the pH of the droplets
is calculated to be 1.03. While HO2 radicals (pKa = 4.8) can
dissociate into H+ and O2•− in a dilute aqueous solution, under
highly acidic environment, it is unclear that HO2 will fully
dissociate in the concentrated tartaric acid droplets. Since
hydrogen abstraction rates are not well-constrained for tartaric
acid and its observed products, more work is needed to
investigate the role of O•2− reactions in the OH oxidation of
aqueous tartaric acid droplets. This reaction pathway is not
considered in reaction schemes.
3.2.3. OH Reaction Site: Tertiary Backbone Carbon Site
Versus Hydroxyl Group. As discussed in the preceding sections
3.2.1 and 3.2.2, hydrogen atoms on the tertiary backbone
carbon site and the hydroxyl group can be abstracted by OH
radicals. As shown in Table 1 and Scheme 1, depending on the
OH reaction site, reaction products with distinct chemical
formula are formed and can be assembled into two separate
groups (tertiary carbon site vs hydroxyl group). The ratio of the
abundance of reaction products in these two groups
When the hydrogen atom on the hydroxyl group is removed
by the OH radical (Scheme 1B: Path IIB), an activated tertiary
alkoxy radical forms. The decomposition product of this alkoxy
F
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
●
▼
Figure 4. Chemical evolution in the aerosol composition for the heterogeneous OH oxidation of tartaric acid ( ), 2,3-DMSA ( ), and succinic acid
( ): (A) carbon oxidation state (OSC) vs carbon number (NC) space; (B) van Krevelen plot.
■
Figure 5. Distributions of molecular average carbon oxidation states (OSC) and carbon number (NC) for individual products at the points (a−c)
marked in Figure 4. For the OSC distribution data of 2,3-DMSA, only the major products are shown (the abundance of other minor individual
products is less than 1%).
corresponds to the ratio of OH attacks at the tertiary carbon
site and the hydroxyl group. When the hydrogen abstraction
occurs at the tertiary carbon site of tartaric acid, 2-hydroxy-3-
oxosuccinic acid (C4H4O6) is formed as the major function-
alization product (Scheme 1A: Path IA). Since 2-hydroxy-3-
oxosuccinic acid (C4H4O6) can be further oxidized to yield two
fragmentation products (2,3-dioxopropanoic acid (C3H2O4)
and mesoxalic acid (C3H2O5); Scheme 1B), these two second-
generation products originate from the OH reaction at the
tertiary carbon site of tartaric acid. 2-Hydroxy-3-oxopropanoic
acid (C3H4O4) is the only product that can be attributed to
hydrogen abstraction at the hydroxyl group (Scheme 1A: Path
IB) in tartaric acid. As shown in Figure 3B, the total abundance
of reaction products originating from the tertiary carbon site is
larger than that originated from the hydroxyl group. For
example, at the maximum OH exposure, the tertiary carbon site
versus hydroxyl group products ratio is ∼3.3. The results of this
simple analysis suggest that the tertiary carbon site is the
preferential OH reaction site and are consistent with differences
in relative strengths of C−H bond (tertiary carbon site) and
O−H bond (hydroxyl group).36
3.2.4. Volatilization of Oxidized Tartaric Acid Aerosols.
Using the detailed molecular information obtained for the
reaction products, further insight into how oxidation alters the
properties of the aerosol (e.g., volatility) can be investigated.
Kessler et al.15 measured the volatilization of tartaric acid
aerosols during the heterogeneous OH oxidation. At the OH
exposure of ∼1 × 1011 molecules cm−3 s, which is close to the
maximum OH exposure investigated in our study, a small
quantity of aerosol carbon mass loss (<2%) was observed.
Approximately 10% of total aerosol carbon mass is lost at the
highest OH exposure in their measurements (∼1 × 1012
molecules cm−3 s). They suggested that fragmentation products
may have sufficiently low volatilities to remain in the aerosol
phase. On the basis of proposed reaction schemes and the
molecular structures of the reaction products, the saturation
vapor pressures of the products are predicted using the group
contribution method developed by Pankow and Asher.44 The
c* of the fragmentation products are predicted to be in the
order from 1 × 102 to 1 × 104 μg m−3 (Table 1). These results
support Kessler et al.’s observation that a significant fraction of
fragmentation products remain in the aerosol phase given the
G
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
organic aerosol loading is ∼500−650 μg m−3 in their
measurements. Equilibrium gas/aerosol partitioning calcula-
tions predict here that ∼36−42% of 2-hydroxy-3-oxopropanoic
acid (C3H4O4) and 81−85% of mesoxalic acid (C3H2O5)
remain in the aerosol-phase, while 2,3-dioxopropanoic acid
(C3H2O4) is a volatile product (only <5% remains in aerosol
phase). It is noted that the reaction products might expect to be
less volatile than predicted. For example, the saturation vapor
pressure of tartaric acid is reported to be (1.8−4.9) × 10−8
Pa,45 while the modeled value is 2 × 10−7 Pa.44 Moreover, as
shown in Figure 3C, if we assume the ionization efficiencies of
all observed products are the same and the major reaction
products can be ionized by the DART ionization source, the
abundance of functionalization product is found to be slightly
larger than that of fragmentation products. Our results further
suggest that a small aerosol carbon mass loss reported in the
literature could also attribute to the formation of nonvolatile
functionalization product (c* of 2-hydroxy-3-oxosuccinic acid is
predicted to be 0.315 μg m−3).
3.3. Effect of Different Functional Groups on the
Heterogeneous Reaction Mechanisms. The structure of an
organic molecule plays a key role in controlling the
heterogeneous chemistry. The evolution in molecular compo-
sition of the reaction products formed in the heterogeneous
OH reaction with succinic acid,10 2,3-DMSA,11 and tartaric acid
is used to examine how different functional groups determine
the heterogeneous reaction mechanisms for these three
structurally related dicarboxylic acids (Figure 1).46
have also been suggested to be major formation pathways of
functionalization products in the OH oxidation of C2−C9
unsubstituted dicarboxylic acids.35,46,48 On the other hand,
when the methyl or hydroxyl groups are substituted to the
backbone of succinic acid, the functionalization products are
formed via different pathways, and the relative importance of
functionalization versus fragmentation processes increases.
With the presence of two methyl groups in 2,3-DMSA (Figure
1C), the ketone-to-alcohol functionalization product ratio
drops to nearly 0 (i.e., no ketone functionalization product is
formed). It is hypothesized that the two bulky methyl groups
may sterically hinder the appropriate arrangement of two
peroxy radicals into the correct intermediate, necessary for the
formation of functionalization products proposed in the Russell
mechanism. Alternatively, alkoxy radicals are produced from
these self-reactions and can in turn abstract hydrogen atoms
from neighboring molecules to form the alcohol functionaliza-
tion products. It is also postulated that the methyl groups might
enhance the stability of the tertiary alkoxy radicals. This may
favor the intermolecular hydrogen abstraction and in part
explain the significant formation of functionalization products
(Figures 5E,F).11 However, more work is needed to test this
hypothesis.
For tartaric acid (Figure 1A), the formation of functionaliza-
tion products does not appear to proceed through bimolecular
reactions of two peroxy radicals. In presence of a hydroxyl
group near the peroxy group, the α-hydroxylperoxy radical
tends to undergo the unimolecular HO2 elimination process
(Scheme 2B). This functionalization process converts an
alcohol group to a ketone group (the ketone-to-alcohol
functionalization products ratio tends to infinity), leading to a
high abundance of functionalization products (Figure 5A,B).
Since the functionalization product contributes significantly to
the observed products, the steep and positive slope observed in
the van Krevelen plot arise from the conversion of alcohols to
ketones originating from unimolecular HO2 elimination during
oxidation.35
The change in average aerosol elemental composition upon
oxidation can be expressed in terms of hydrogen-to-carbon
ratio (H/C), oxygen-to-carbon ratio (O/C), carbon oxidation
state (OSC = 2 × (O/C) − (H/C))47 and average carbon
number (NC). These quantities are computed from the relative
abundance and suggested chemical formula of the observed
products:
Q avg
=
Q I
i
i
∑
(3)
i
Overall, for these three structurally related acids, the
fragmentation products appear to be primarily formed through
the decomposition of alkoxy radicals, while the functionaliza-
tion products are formed via distinct pathways. At the same OH
exposure, the relative contribution of fragmentation products to
the total products is less significant for 2,3-DMSA and tartaric
acid when compared to succinic acid. This observation does not
agree with recent hypothesis that the fragmentation process is
more important for more branched and oxygenated species
relative to their less branched and oxygenated analogues.49
These results show that the types and location of the functional
groups (e.g., methyl and hydroxyl groups) can significantly alter
reaction pathways for these two structurally related dicarboxylic
acids.
where Q can be H/C, O/C, OSC, or NC. Qavg is the average
quantity. Qi and Ii are the quantity and relative abundance of
the reaction product i, respectively. As shown in Figure 4A,
when the OH exposure increases, the OSC increases and the NC
decreases for all three acids. However, the acids exhibit different
oxidation pathways in the van Krevelen plot (H/C vs O/C;
Figure 4B). At a similar OH exposure (∼4 × 1011 cm3
molecule−1 s−1), a slope of ca. −1 is observed for succinic
acid (−0.88) and 2,3-DMSA (−0.72), but a steep positive slope
of +13.6 is observed for tartaric acid. Aerosol speciation data
further reveal that the acids have different molecular
distribution of reaction products. For succinic acid (Figure
5C,D), the bulk aerosol elemental composition evolves largely
through the formation of smaller fragmentation products with
higher oxidation states. In contrast to succinic acid, for both
tartaric acid (Figure 5A,B) and 2,3-DMSA (Figure 5E,F), only a
few major products are formed. Both functionalization products
and fragmentation products contribute significantly to the
observed products. All these results suggest that the acids
chemically evolve via distinct mechanisms.
4. CONCLUSIONS
In this work, we investigate the chemical evolution in the
composition of aqueous tartaric acid droplets during the
heterogeneous OH oxidation. The aerosol speciation data have
revealed that the functionalization products are the most
abundant among the observed products. This can be explained
by that when the tertiary backbone carbon site of tartaric acid is
abstracted by the OH radical, the presence of a hydroxyl group
on the neighboring position of the peroxy group favors the
unimolecular HO2 elimination of the α-hydroxylperoxy radical
over the self-reactions of two peroxy radicals. On the one hand,
For the OH reaction with succinic acid (Figure 1B), we
found that the ketone-to-alcohol (i.e., functionalization
products) ratio is ∼1 and can be explained by self-reactions
of two peroxy radicals via the Bennett−Summers reaction13 or
Russell mechanism.12 On the one hand, these two mechanisms
H
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
Secondary Organic Aerosol: Current and Emerging Issues. Atmos.
Chem. Phys. 2009, 9, 5155−5236.
this functionalization process converts an alcohol group into a
ketone group, yielding 2-hydroxy-3-oxosuccinic acid. On the
other hand, when the hydrogen atom from the backbone
carbon site of 2-hydroxy-3-oxosuccinic acid is abstracted by the
OH radical, in the absence of a hydroxyl group adjacent to the
peroxy group, the bimolecular reaction of two peroxy radicals
becomes competitive with the unimolecular HO2 elimination
process. The alkoxy radical resulted from peroxy radical self-
reactions tend to decompose, leading to a fragmentation of the
carbon backbone. Although higher bond dissociation energy is
expected for the O−H bond than the C−H bond, a significant
fraction of fragmentation products is likely produced through
the decomposition of alkoxy radicals resulted from the
hydrogen abstraction on the hydroxyl group.
Both functionalization and fragmentation products contrib-
ute significantly to the observed products in the OH oxidation
of tartaric acid. This is not consistent with recent hypothesis
that the fragmentation process is expected to be favorable for
oxygenated organic compounds. The results of this work may
suggest that for oxygenated organic compounds, especially
those with multiple hydroxyl groups, the position of the
hydroxyl groups may play a role in determining the
competitions between functionalization and fragmentation
processes during oxidation. In addition to the bulk elemental
composition such as O/C and OSC, results from this work and
our previous studies suggest that the number, types, and
location of the functional groups (e.g., hydroxyl, ketone, and
carboxyl groups) is needed to better understand the
heterogeneous chemistry.
(2) Pope, C. A., III; Dockery, D. W. Health Effects of Fine Particulate
Air Pollution: Lines That Connect. J. Air Waste Manage. Assoc. 2006,
56, 709−742.
(3) Ng, N. L.; Canagaratna, M. R.; Jimenez, J. L.; Chhabra, P. S.;
Seinfeld, J. H.; Worsnop, D. R. Changes in Organic Aerosol
Composition with Aging Inferred from Aerosol Mass Spectra. Atmos.
Chem. Phys. 2011, 11, 6465−6474.
(4) Rudich, Y.; Donahue, N. M.; Mentel, T. F. Aging of Organic
Aerosol: Bridging the Gap between Laboratory and Field Studies.
Annu. Rev. Phys. Chem. 2007, 58, 321−352.
(5) George, I. J.; Abbatt, J. P. Heterogeneous Oxidation of
Atmospheric Aerosol Particles by Gas-Phase Radicals. Nat. Chem.
2010, 2, 713−722.
(6) Cappa, C. D.; Che, D. L.; Kessler, S. H.; Kroll, J. H.; Wilson, K. R.
Variations in Organic Aerosol Optical and Hygroscopic Properties
Upon Heterogeneous OH Oxidation. J. Geophys. Res. 2011, 116, 1
DOI: 10.1029/2011JD015918.
(7) Jimenez, J. L.; Canagaratna, M. R.; Donahue, N. M.; Prevot, A. S.;
Zhang, Q.; Kroll, J. H.; DeCarlo, P. F.; Allan, J. D.; Coe, H.; Ng, N. L.;
et al. Evolution of Organic Aerosols in the Atmosphere. Science 2009,
326, 1525−1529.
(8) Petters, M. D.; Prenni, A. J.; Kreidenweis, S. M.; DeMott, P. J.;
Matsunaga, A.; Lim, Y. B.; Ziemann, P. J. Chemical Aging and the
Hydrophobic-to-Hydrophilic Conversion of Carbonaceous Aerosol.
Geophys. Res. Lett. 2006, 33, 1 DOI: 10.1029/2006GL027249.
(9) Kroll, J. H.; Smith, J. D.; Che, D. L.; Kessler, S. H.; Worsnop, D.
R.; Wilson, K. R. Measurement of Fragmentation and Functionaliza-
tion Pathways in the Heterogeneous Oxidation of Oxidized Organic
Aerosol. Phys. Chem. Chem. Phys. 2009, 11, 8005−8014.
(10) Chan, M. N.; Zhang, H. F.; Goldstein, A. H.; Wilson, K. R. Role
of Water and Phase in the Heterogeneous Oxidation of Solid and
Aqueous Succinic Acid Aerosol by Hydroxyl Radicals. J. Phys. Chem. C
2014, 118, 28978−28992.
ASSOCIATED CONTENT
* Supporting Information
■
S
(11) Cheng, C. T.; Chan, M. N.; Wilson, K. R. The Role of Alkoxy
Radicals in the Heterogeneous Reaction of Two Structural Isomers of
Dimethylsuccinic Acid. Phys. Chem. Chem. Phys. 2015, 17, 25309−
25321.
(12) Russell, G. A. Deuterium-Isotope Effects in the Autoxidation of
Aralkyl Hydrocarbons. Mechanism of the Interaction of Peroxy
Radicals. J. Am. Chem. Soc. 1957, 79, 3871−3877.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jpca.6b05289.
Measurement of OH exposure using hexane measure-
ment, ionization mechanism of the DART technique
(negative ionization mode) (PDF)
(13) Bennett, J. E.; Summers, R. Product Studies of the Mutual
Termination Reactions of Sec-Alkylperoxy Radicals: Evidence for
Non-Cyclic Termination. Can. J. Chem. 1974, 52, 1377−1379.
(14) Chacon-Madrid, H. J.; Donahue, N. M. Fragmentation Vs.
Functionalization: Chemical Aging and Organic Aerosol Formation.
Atmos. Chem. Phys. 2011, 11, 10553−10563.
(15) Kessler, S. H.; Nah, T.; Daumit, K. E.; Smith, J. D.; Leone, S. R.;
Kolb, C. E.; Worsnop, D. R.; Wilson, K. R.; Kroll, J. H. OH-Initiated
Heterogeneous Aging of Highly Oxidized Organic Aerosol. J. Phys.
Chem. A 2012, 116, 6358−6365.
AUTHOR INFORMATION
■
Corresponding Authors
*Phone: (852) 3943-9863. E-mail: mnchan@cuhk.edu.hk.
(M.N.C.)
*Phone: (1) 510-495-2474. E-mail: krwilson@lbl.gov. (K.R.W.)
Notes
The authors declare no competing financial interest.
(16) Peng, C.; Chan, M. N.; Chan, C. K. The Hygroscopic Properties
of Dicarboxylic and Multifunctional Acids: Measurements and
UNIFAC Predictions. Environ. Sci. Technol. 2001, 35, 4495−4501.
(17) Zuend, A.; Marcolli, C.; Luo, B. P.; Peter, T. A Thermodynamic
Model of Mixed Organic-Inorganic Aerosols to Predict Activity
Coefficients. Atmos. Chem. Phys. 2008, 8, 4559−4593.
ACKNOWLEDGMENTS
■
C.T.C. and M.N.C. are supported by a Direct Grant for
Research (4053089) and One-Time Funding Allocation of
Direct Grant (3132765), The Chinese University of Hong
Kong. K.R.W. and the experimental research facilities are
supported by the Department of Energy, Office of Science
Early Career Award, and the Director, Office of Energy
Research, Office of Basic Energy Sciences, Chemical Sciences,
Geosciences, and Biosciences Division of the U.S. Department
of Energy under Contract No. DE-AC02-05CH11231.
(18) Smith, J. D.; Kroll, J. H.; Cappa, C. D.; Che, D. L.; Liu, C. L.;
Ahmed, M.; Leone, S. R.; Worsnop, D. R.; Wilson, K. R. The
Heterogeneous Reaction of Hydroxyl Radicals with Sub-Micron
Squalane Particles: A Model System for Understanding the Oxidative
Aging of Ambient Aerosols. Atmos. Chem. Phys. 2009, 9, 3209−3222.
(19) Cody, R. B.; Laramee, J. A.; Durst, H. D. Versatile New Ion
Source for the Analysis of Materials in Open Air under Ambient
Conditions. Anal. Chem. 2005, 77, 2297−2302.
REFERENCES
■
(20) Cody, R. B. Observation of Molecular Ions and Analysis of
Nonpolar Compounds with the Direct Analysis in Real Time Ion
Source. Anal. Chem. 2009, 81, 1101−1107.
(1) Hallquist, M.; Wenger, J. C.; Baltensperger, U.; Rudich, Y.;
Simpson, D.; Claeys, M.; Dommen, J.; Donahue, N. M.; George, C.;
Goldstein, A. H.; et al. The Formation, Properties and Impact of
I
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
(21) Chan, M. N.; Nah, T.; Wilson, K. R. Real Time in situ Chemical
Characterization of Sub-Micron Organic Aerosols Using Direct
Analysis in Real Time Mass Spectrometry (DART-MS): The Effect
of Aerosol Size and Volatility. Analyst 2013, 138, 3749−3757.
(22) Nah, T.; Chan, M.; Leone, S. R.; Wilson, K. R. Real Time in Situ
Chemical Characterization of Submicrometer Organic Particles Using
Direct Analysis in Real Time-Mass Spectrometry. Anal. Chem. 2013,
85, 2087−2095.
Reaction Barrier Heights. Phys. Chem. Chem. Phys. 2009, 11, 9062−
9074.
(41) Monod, A.; Doussin, J. F. Structure-Activity Relationship for the
Estimation of OH-Oxidation Rate Constants of Aliphatic Organic
Compounds in the Aqueous Phase: Alkanes, Alcohols, Organic Acids
and Bases. Atmos. Environ. 2008, 42, 7611−7622.
(42) Wiegel, A. A.; Wilson, K. R.; Hinsberg, W. D.; Houle, F. A.
Stochastic Methods for Aerosol Chemistry: A Compact Molecular
Description of Functionalization and Fragmentation in the Heteroge-
neous Oxidation of Squalane Aerosol by OH Radicals. Phys. Chem.
Chem. Phys. 2015, 17, 4398−4411.
(43) Hayyan, M.; Hashim, M. A.; AlNashef, I. M. Superoxide Ion:
Generation and Chemical Implications. Chem. Rev. 2016, 116, 3029−
3085.
(44) Pankow, J. F.; Asher, W. E. SIMPOL.1: A Simple Group
Contribution Method for Predicting Vapor Pressures and Enthalpies
of Vaporization of Multifunctional Organic Compounds. Atmos. Chem.
Phys. 2008, 8, 2773−2796.
(45) Huisman, A. J.; Krieger, U. K.; Zuend, A.; Marcolli, C.; Peter, T.
Vapor Pressures of Substituted Polycarboxylic Acids Are Much Lower
Than Previously Reported. Atmos. Chem. Phys. 2013, 13, 6647−6662.
(46) Yang, L.; Ray, M. B.; Yu, L. E. Photooxidation of Dicarboxylic
AcidsPart II: Kinetics, Intermediates and Field Observations. Atmos.
Environ. 2008, 42, 868−880.
(47) Kroll, J. H.; Donahue, N. M.; Jimenez, J. L.; Kessler, S. H.;
Canagaratna, M. R.; Wilson, K. R.; Altieri, K. E.; Mazzoleni, L. R.;
Wozniak, A. S.; Bluhm, H.; et al. Carbon Oxidation State as a Metric
for Describing the Chemistry of Atmospheric Organic Aerosol. Nat.
Chem. 2011, 3, 133−139.
(48) Enami, S.; Hoffmann, M. R.; Colussi, A. J. Stepwise Oxidation of
Aqueous Dicarboxylic Acids by Gas-Phase OH Radicals. J. Phys. Chem.
Lett. 2015, 6, 527−534.
(23) Davies, J. F.; Wilson, K. R. Nanoscale Interfacial Gradients
Formed by the Reactive Uptake of OH Radicals onto Viscous Aerosol
Surfaces. Chem. Sci. 2015, 6, 7020−7027.
(24) Houle, F. A.; Hinsberg, W. D.; Wilson, K. R. Oxidation of a
Model Alkane Aerosol by OH Radical: The Emergent Nature of
Reactive Uptake. Phys. Chem. Chem. Phys. 2015, 17, 4412−4423.
(25) Slade, J. H.; Knopf, D. A. Multiphase OH Oxidation Kinetics of
Organic Aerosol: The Role of Particle Phase State and Relative
Humidity. Geophys. Res. Lett. 2014, 41, 5297−5306.
(26) Berkemeier, T.; Huisman, A. J.; Ammann, M.; Shiraiwa, M.;
Koop, T.; Poschl, U. Kinetic Regimes and Limiting Cases of Gas
̈
Uptake and Heterogeneous Reactions in Atmospheric Aerosols and
Clouds: A General Classification Scheme. Atmos. Chem. Phys. 2013,
13, 6663−6686.
(27) Arangio, A. M.; Slade, J. H.; Berkemeier, T.; Poschl, U.; Knopf,
̈
D. A.; Shiraiwa, M. Multiphase Chemical Kinetics of OH Radical
Uptake by Molecular Organic Markers of Biomass Burning Aerosols:
Humidity and Temperature Dependence, Surface Reaction, and Bulk
Diffusion. J. Phys. Chem. A 2015, 119, 4533−4544.
(28) Parmar, M. L.; Awasthi, R. K.; Guleria, M. K. A Study on
Viscosities of Citric Acid and Tartaric Acid in Water and Binary
Aqueous Mixtures of Ethanol at Different Temperatures. Indian J.
Chem. A 2004, 43, 41−44.
(29) Vel Leitner, N. K.; Dore, M. Hydroxyl Radical Induced
Decomposition of Aliphatic Acids in Oxygenated and Deoxygenated
Aqueous Solutions. J. Photochem. Photobiol., A 1996, 99, 137−143.
(30) Bothe, E.; Behrens, G.; Schulte-Frohlinde, D. Mechanism of the
First Order Decay of 2-Hydroxy-Propyl-2-Peroxyl Radicals and of O2·
Formation in Aqueous Solution. Z. Naturforsch., B: J. Chem. Sci. 1977,
32, 886−889.
(49) Lim, Y. B.; Ziemann, P. J. Effects of Molecular Structure on
Aerosol Yields from OH Radical-Initiated Reactions of Linear,
Branched, and Cyclic Alkanes in the Presence of NOx. Environ. Sci.
Technol. 2009, 43, 2328−2334.
(31) Bothe, E.; Schuchmann, M. N.; Schulte-Frohlinde, D.; von
Sonntag, C. HO2 Elimination from α-Hydroxyalkylperoxyl Radicals in
Aqueous Solution. Photochem. Photobiol. 1978, 28, 639−643.
(32) Bothe, E.; Schulte-Frohlinde, D.; von Sonntag, C. Radiation
Chemistry of Carbohydrates. Part 16. Kinetics of HO2 Elimination
from Peroxyl Radicals Derived from Glucose and Polyhydric Alcohols.
J. Chem. Soc., Perkin Trans. 2 1978, 416−420.
(33) Booth, A. M.; Montague, W. J.; Barley, M. H.; Topping, D. O.;
McFiggans, G.; Garforth, A.; Percival, C. J. Solid State and Sub-Cooled
Liquid Vapour Pressures of Cyclic Aliphatic Dicarboxylic Acids. Atmos.
Chem. Phys. 2011, 11, 655−665.
(34) Donahue, N.; Robinson, A.; Stanier, C.; Pandis, S. Coupled
Partitioning, Dilution, and Chemical Aging of Semivolatile Organics.
Environ. Sci. Technol. 2006, 40, 2635−2643.
(35) Charbouillot, T.; Gorini, S.; Voyard, G.; Parazols, M.; Brigante,
M.; Deguillaume, L.; Delort, A.-M.; Mailhot, G. Mechanism of
Carboxylic Acid Photooxidation in Atmospheric Aqueous Phase:
Formation, Fate and Reactivity. Atmos. Environ. 2012, 56, 1−8.
(36) Blanksby, S. J.; Ellison, G. B. Bond Dissociation Energies of
Organic Molecules. Acc. Chem. Res. 2003, 36, 255−263.
(37) Atkinson, R. Rate Constants for the Atmospheric Reactions of
Alkoxy Radicals: An Updated Estimation Method. Atmos. Environ.
2007, 41, 8468−8485.
(38) Denisov, E. T.; Afanas’ ev, I. B. Oxidation and Antioxidants in
Organic Chemistry and Biology; CRC Press: Boca Raton, FL, 2005.
(39) Peeters, J.; Fantechi, G.; Vereecken, L. A Generalized Structure-
Activity Relationship for the Decomposition of (Substituted) Alkoxy
Radicals. J. Atmos. Chem. 2004, 48, 59−80.
(40) Vereecken, L.; Peeters, J. Decomposition of Substituted Alkoxy
Radicals-Part I: A Generalized Structure-Activity Relationship for
J
DOI: 10.1021/acs.jpca.6b05289
J. Phys. Chem. A XXXX, XXX, XXX−XXX