Aqueous phase reactivity of nitrate radicals (NO3) toward dicarboxylic acids

Article abstract of DOI:10.1524/zpch.2010.6150

Laser photolysis technique was used to study the reactivity of nitrate radical towards four dicarboxylic acids. The temperature dependence of the reactions was investigated in the range from 278K to 318K. The effect of the acid-base equilibrium was examined by measuring the activation parameters of the dissociated and undissociated acids at different pH. For droplets or aerosols under moderate acidic conditions, the charge exchange reaction at the carboxylate group might compete with the oxidation by the hydroxyl radical. On liquid particles enriched in nitrate (such as a deliquescent ammonium nitrate), the lifetime of carboxylic groups towards the NO₃ radical may be as short as few hours (or less). by Oldenbourg Wissenschaftsverlag.

Full text of DOI:10.1524/zpch.2010.6150

Z. Phys. Chem. 224 (2010) 1247–1260 . DOI 10.1524.zpch.2010.6150
© by Oldenbourg Wissenschaftsverlag, München
Aqueous Phase Reactivity of Nitrate Radicals
(NO3) Toward Dicarboxylic Acids
*
By Ph. G. de Sémainville, B. D’Anna, and Ch. George
Université de Lyon 1, CNRS, UMR5256, IRCELYON, Institut de recherches sur la catalyse
et l’environnement de Lyon, Villeurbanne, F-69626, France
Dedicated to Prof. Dr. Reinhard Zellner on the occasion of his 65th birthday
(Received December 9, 2009; accepted March 21, 2010)
Nitrate Radical . Kinetics . Dicarboxylic Acids
Laser photolysis technique was used to study the reactivity of nitrate radical towards four
dicarboxylic acids. The temperature dependence of the reactions was investigated in the range
from 278K to 318K. The effect of the acid-base equilibrium was examined by measuring the
activation parameters of the dissociated and undissociated acids at different pH. For droplets
or aerosols under moderate acidic conditions, the charge exchange reaction at the carboxylate
group might compete with the oxidation by the hydroxyl radical. On liquid particles enriched
in nitrate (such as a deliquescent ammonium nitrate), the lifetime of carboxylic groups towards
the NO3 radical may be as short as few hours (or less).
1. Introduction
Radicals are key compounds in a variety of chemical processes including the
natural degradation of pollutants. This is often observed in the atmosphere which
can be regarded as a giant chemical reactor. As an example, the oxidation capac-
ity of the atmosphere is mainly governed by the presence of radicals such as OH
or NO3 which will react will almost all organic pollutants. Such oxidation cycles
will lead, among other, to the photochemical production of ozone which is cur-
rently fairly well understood.
Furthermore, our society is facing a changing atmosphere as a response to a
man-made anthropogenic pressure and these changes are far from being well
understood. Over the last decades, aerosols and clouds have been considered as
the “least predictable” object in our understanding of the atmospheric and climate
*
Corresponding author. E-mail: christian.george@ircelyon.univ-lyon1.fr
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

1248
Ph. G. de Sémainville et al.
systems. On the global scale, atmospheric aerosols are recognized as having a
major impact on the Earth’s climate through direct and indirect radiative forcing.
The organic component of aerosols may affect climate in multiple ways:
organic compounds can act as surfactants and by their limited solubility modulate
aerosol interactions with water vapour, and hence, hygroscopic growth and cloud
condensation nuclei (CCN) activity. A major effort has recently been devoted to
the understanding of organic aerosol formation, composition and physico-chemi-
cal properties. Aerosols have an atmospheric lifetime that can reach one or two
weeks. It is therefore necessary to investigate their chemical transformations in
the atmosphere. Chemical processes may lead to changes in both climate and
health effects of the aerosols due to variation in their composition, optical proper-
ties and interaction with water.
For instance, Saxena and Hildeman [1] showed that freshly emitted aerosols
are made of relatively non-oxidized compounds with hydrophobic properties,
while aged aerosols becomes more hydrophilic and contain more oxygenated
species.functionalities. Among the latter, dicarboxylic acids can made up to 50%
of the aerosol mass (with respect the soluble fraction). Carboxylic and especially
dicarboxylic acids are therefore key compounds that may affect aerosol lifetimes
and also cloud formation potential (both of them are intimately linked to the
changing climate).
Such carboxylic and dicarboxylic acids may be produced in the atmosphere
in a variety of gas phase reactions including ozonolysis of long chain fatty acids,
aromatics, etc… But the atmosphere is far from being a purely homogeneous
media and several processes may occur in clouds, which are ubiquitous. Each
individual cloud droplet acts a small aqueous chemical reactor in which soluble
pollutants will be processed. Ervens and Kreidenweis [2] showed in a modelling
study that short chain dicarboxylic acids (mainly oxalic acid) will be produced
in cloud.
Nevertheless, the atmospheric fate of such acids is still unclear as many
fundamental parameters have not been studied so far. In fact, kinetic and mecha-
nistic data on the reactivity of dicarboxylic acids are not well established. Similar
to the gas phase, chemical conversion within a cloud droplet is driven by radicals
(again such as OH or NO3) [3,4] but the possible reaction pathways are manifold.
As in the gas phase, H-abstraction is a possible reaction pathway but in the
aqueous phase, organic acids may also undergo acid-base dissociation and the
resulting ions may in addition initiate charge exchange reactions. Therefore, the
knowledge gained from gas phase chemistry should not be straightforwardly
transferred to the atmospheric aqueous phase, as reactivity and mechanisms may
be pH dependent. This latter parameter may be especially important in the marine
environment were the pH of deliquescent aerosol span over a quite wide dynamic
range influenced by the uptake or release of acids [5,6].
Due to the importance of the nitrate radical for chemical transformation in
clouds, we studied the reactivity of the NO3 towards four different dicarboxylic
acids i.e., oxalic, malonic, mesoxalic and succinic acids as a function of pH and
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

Aqueous Phase Reactivity of Nitrate Radicals (NO3)
1249
temperature by means of a laser photolysis technique. In the following sections
we will describe experimental layout, followed by the results and the discussion,
highlighting some features of the reactivity of the selected dicarboxylic acids.
2. Experimental section
Experimental details have been already published elsewhere [7–10], therefore
only a brief description is given below.
2.1 Nitrate radical generation
The generation of the nitrate radical was achieved using two different methods.
In the first method (A), the NO3 radicals were produced by the laser photolysis
at 266 nm of an aqueous nitrate solution at pH = 0.5 according to:
NOK3 +H++hv/NO•2+OH•
(R1)
HNO3+OH•/NO3• +H2O
(R2)
The method requires relatively high acidic conditions in order to produce
enough molecular nitric acid for reaction (R2) to occur at a reasonable rate. On
the other side the advantage is due to the limited number of short lived radicals
as precursors (i.e., OH). Accordingly the reaction sequence (R1–R2) is a rela-
tively clean source of radicals but limited to acidic conditions. Typically the
radicals were produced at pH = 0.5 with NaNO3 concentrations of 0.5 mol LK1
.
In the second method (B), the NO3 radicals were produced through laser-
flash photolysis of peroxodisulfate in presence of nitrate anion at 355 nm accord-
ing to:
S2O28K+hv/2SOK4 •
(R3)
SO4K•+NOK3 /SO42K+NO3•
(R4)
Using reactions (R3) and (R4) implies a delay longer than in method A to
achieve complete formation of the nitrate radicals. This time gap depends on
Laser fluence and reactant concentrations. Therefore, depending on the experi-
mental conditions, the decay of the nitrate radical may not necessarily be of first
order during the few microseconds following the laser discharge. Thus care must
be taken during NO3 radical analysis ensuring that its decay occurs.is evaluated
after complete production of NO3. This was achieved under following conditions:
[NaNO3] = 0.1 – 0.5 mol LK1 and [Na2S2O8] = 0.1 mol LK1
.
2.2 LASER photolysis set-up
A Continuum Surelite II operated at 266 or 355 nm was used as the photolysis
source. Under optimum conditions, this laser produces a pulse half-width of 3 ns
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

1250
Ph. G. de Sémainville et al.
and a beam cross-sectional diameter of 8 mm in conjunction of an output of
100 mJ per pulse at 266 nm and 200 mJ at 355 nm. This beam was then expanded
to irradiate uniformly the photolysis cell (described in more details below) over
a rectangle of ca. 1!3 cm2 (height!width).
Transient species produced by the pulsed laser beam were monitored by
means of time resolved absorption spectroscopy. A He-Ne laser (LASOS,
LGK7654–8), with an output wavelength of 633 nm and a power of 25 mW was
used as the analyzing light. The light escaping the photolysis cell (described
below) was fed, through an interference filter at 633 nm, onto the active surface
of an amplified photodiode (FEMTO, HCA-S-200M-SI). The rise time (5 to
95%) of the amplified photodiode was below 2 ns. The transient optical absorp-
tion signal produced by the laser beam was subsequently collected by a Tektronix
oscilloscope (type 2430A). Up to 16 single shots were averaged before transmit-
ting to a computer and the photolysis cell content was renewed between each
laser pulse. We ensured that single and averaged signals lead to the same kinetic
determination.
2.3 Teflon AF photolysis cell
The centrepiece of the experimental set-up is a Liquid Core Waveguide (LCW)
made of Teflon AF 2400 (BioGeneral, San Diego, CA). Such type of material
has been shown to exhibit excellent optical properties such as high optical clarity
(at λ ≤ 200 nm more than 80% of light is transmitted through a 220 μm thick
film of the polymer described above) and very low refractive index (i.e., n =
1.29 for Teflon AF 2400 grades). As a result, such type of tubing once filled
with water will conduct light. The Teflon AF 2400 guide had an inner and outer
diameter of 0.8 mm and 1.0 mm, respectively, a length of 50 cm with an inner
volume of 0.25 mL. This very small liquid volume represents the major advan-
tage of the waveguides compared to standard White cells.
The highly flexible Teflon AF 2400 tubing was loosely coiled (less than
3 cm diameter) and placed in the laser beam path. The NO3 radicals were gener-
ated within the LCW following the laser flash of the precursors as discussed
above. As the waveguide was used as a coil, no concentration gradient could
build up along the waveguide length. However, non-uniform nitrate radical con-
centration can be produced if the laser fluence on the coil is not uniform. We
prevent shielding of the front and back side of the waveguide by placing it inside
a small box (1!3!3 cm3 height!width!depth) with inner walls coated by
aluminium foil. This ensured a better irradiation of the waveguide even if the
fluence could not be measured. Consequently we made sure that all kinetics
studied were unimolecular (after completion of NO3 generation as stated before).
The solution content of the Teflon photolysis cell was probed by focusing
the output of the He-Ne laser on the entry of an unpolished 500-μm diameter
fused silica optical fibre. Then, the light was allowed to escape the solid optical
fibre and collected in the liquid core waveguide. Again, the Teflon AF tubing
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

Aqueous Phase Reactivity of Nitrate Radicals (NO3)
1251
conducted the light up to its end where another fused silica optical fibre (located
in the liquid) collected most of this transmitted light to the amplified photodiode.
In this way, the measured signal is integrated (or averaged) over the full length
of the waveguide. As the observed decays were first order, this averaging al-
lowed the determination of the rate constants even in the unlikely situation where
the NO3 concentration was not uniform along the waveguide length.
2.4 Reagents
Solutions were prepared from following chemicals (used without further purifica-
tion): Na2S2O8 (Aldrich, 98%), NaNO3 (Aldrich, 99%), nitric acid (Aldrich
1 mol LK1), perchloric acid (Aldrich 70%), oxalic acid (Riedle de Haën, 99%),
malonic acid (Fluka, 99%), succinic acid (Fluka, 99%) and sodium mesoxalate,
monohydrate (Fluka, 98%). Water was taken from an 18MΩ purification system
(Millipore). In some test experiments, we used oxygen free solutions, which led
to the same results as non-degassed water.
3. Results
3.1 Kinetic data
The methodology described above has been applied to study the NO3 radical
reactions with oxalic (pKa1 = 1.23; pKa2 = 4.19), malonic (pKa1 = 2.83; pKa2 =
5.69), mesoxalic (pKa1 = 1.60; pKa2 = 3.90) and succinic (pKa1 = 4.19; pKa2 =
5.48) acids as a function of temperature and pH. Figures 1 and 2 show raw data
of mesoxalic acid demonstrating the measured kinetics are first order (Figure 2).
Bimolecular plots from which the second order rate reactions were derived are
given in Figure 3. The kinetic data, as well as the available literature values, are
summarized in Table 1 at a 2σ error level.
The rate constants were measured as a function of temperature in the range
from 278K to 318K for malonic (pH = 4.26), mesoxalic and succinic acid, from
278K to 298K for oxalic acid (because of thermal degradation of the oxalate)
and between 278K and 308K for malonic acid at pH = 8 (adjusted with NaOH).
Figure 4 shows typical Arrhenius plots for mesoxalic acid at pH = 0. Table 1
also lists Arrhenius parameters for all the reactions studied here i.e., pre-expo-
nential factor (A) and activation energies (Ea) along with the parameters derived
from the Eyring equation i.e.,
≠
kBT
h
k =
eKΔG ≠.RT
(E1)
namely following activation parameters:
kBT
h
ΔS≠ = R[ln(A)Kln( )−1]
(E2)
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

1252
Ph. G. de Sémainville et al.
Fig. 1. Experimental decay, absorbance versus time, for mesoxalic acid at pH = 0 (method A
at 266nm).
Fig. 2. First order kinetic plot for mesoxalic acid at pH = 0 (method A at 266nm).
ΔH≠ = EaKRT
(E3)
(E4)
ΔG≠ = ΔH≠KTΔS≠
Table 1 also shows the pH dependence of the measured kinetic parameters.
Obviously dicarboxylic acids have two pKas therefore, depending on the pH, the
chemistry of the neutral (or molecular), the singly and doubly charged anions of
the acid are investigated. And as can be clearly seen in Table 1, the protonation.
deprotonation degree strongly influences the reactivity of the acid.
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

Aqueous Phase Reactivity of Nitrate Radicals (NO3)
1253
Fig. 3. Bimolecular plots for mesoxalic acid at pH = 0 (method A at 266nm) at different
temperatures.
Table 1. Summary of kinetic and thermodynamic data for the investigated NO3 reactions in
aqueous solution measured in this study.
Compound nH* k2nd.298K
[MK1sK1
A
EA
ΔG‡
ΔH‡
ΔS‡
BDE**
Method
]
[MK1sK1
]
[kJ.mol] [kJ.mol] [kJ.mol] [J.mol·K] [kJ.mol]
Malonic acid
2
2
(5.1 0.5) · 104
40024
B pH 0.3
Malonate
monoanion
(5.6 2.1) · 106 (5.0 0.6) · 1011 28 7
(2.3 0.5) · 107 (6.3 0.8) · 1011 25 8
(1.7 0.5) · 106 (5.1 0.4) · 108 13 4
(2.3 0.4) · 107
34 8
31 9
37 6
26 7
23 8
11 4
K(29 3)
K(27 3)
K(86 6)
–
–
–
–
–
B pH 4.26
Malonate di-
anion
2
0
0
0
B pH 8
A pH 0
B pH 2.7
B pH 8
Mesoxalic
acid
Mesoxalate
monoanion
Mesoxalate
dianion
(4.9 1.5) · 107 (1.4 0.1) · 1012 26 4
28 5
25 7
32 4
23 4
20 6
23 3
K(21 1)
K(17 2)
K(27 3)
Oxalic acid
0
0
<5 · 104
(4.4 0.2) · 107
–
–
Oxalate mo-
noanion
B pH 2.7
B pH 8
Oxalatedian-
ion
0
(2.2 0.8) · 108 (2.2 0.2) · 1012 23 6
–
Succinic acid
4
4
(5 5) · 103
(1.1 0.2) · 107
39824
B pH 0.3
B pH 4.2
Succinate
monoanion
–
Succinate di-
anion
4
(1.8 0.2) · 107 (6.2 0.6) · 1011 26 3
–
B pH 8
* nH = number of the identically most easily abstractable H-atoms in the molecule;
** for C-H bonds.
In aqueous solutions, these acids will react either through an H-abstraction
channel or by a charge exchange reaction.
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

1254
Ph. G. de Sémainville et al.
Fig. 4. Arrhenius plots for the reactions with B succinate dianion, ▲ malonate dianion,▼
mesoxalate dianion, oxalate dianion.
3.2 Reactivity of the acid in its neutral form
Under acidic conditions (i.e., typically one unit below the first pKa of the acid)
only the molecular form of the acid should be present in solution (with maybe
traces of the singly charged anion). Under such condition the possible reaction
pathway is an H-abstraction either from a C-H or an O-H bond.
From Table 1, it can be seen that the reaction coefficients of the protonated
acids span several orders of magnitude (from below 104 up to 106 L molK1 sK1
)
highlighting a quite different reactivity of the various acids. For oxalic acid and
at pH below 1, the only abstractable H atom is from the carboxylic group. Yang
et al. [11] measured a rate constant of (2.4 0,2) 104 L molK1 sK1 by pulse
radiolysis in 2 mol LK1 nitric acid. If the reaction occurs by an H-abstraction,
we do not expect the reactivity to be strongly pH dependent. But as shown in
Figure 5, a very strong pH dependence of the measured rate constants is observed
which points towards an important influence of traces of singly charged species
especially around pH = 1. If we do assume that the measured reactivity is due
to both the neutral (AH) and singly charged (AK) forms of the acid then the
following equation may apply:
K
ktot([AH]+[AK]) = kAH[AH]+kA [A ].
(E5)
K
where ktot is the measured overall rate constant i.e., a combination of kAH and
kAK which are the rate constants for the molecular form and its monoanion.
Assuming that the reactivity of the neutral acid AH can be neglected and using
the relationship between pH and pKa, one can obtain the following relationship:
log(ktot(1+10pHKpKa)) = log(kA )KpKa+pH
(E6)
K
As shown by Figure 5, this linear relationship is indeed observed, validating
the hypothesis that the neutral acid is quite unreactive toward the nitrate radical.
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

Aqueous Phase Reactivity of Nitrate Radicals (NO3)
1255
Fig. 5. Plot of log(kIIx(1+10pHKpKa)) as a function of pH for oxalic acid at 298K (according to
equation E6).
The useful pH range is limited here by the pKa of the acid (i.e., to have only
traces of the dissociated acid). We therefore concluded that H-abstraction from
the carboxylic function can be neglected in aqueous solutions which is in agree-
ment with the quite high O-H bond dissociation energy measured (BDE =
468 kJ molK1 [12]). In conclusion, for oxalic acid, the reactivity of the mono
anion dominates even at pH below 1, despite being present at very low concen-
tration, explaining also the measurements by Yang et al. [11]. Therefore there is
no H abstraction channel for oxalic acid. This is in agreement with Neta et al.
[13,14] who studied the reactivity of H atoms toward a series of carboxylic acids
and also showed that there is no H abstraction from oxalic acid.
Malonic acid has two C-H bonds between two carboxylic functions, accord-
ingly H abstraction is possible and even favoured by the proximity of the -COOH
groups. Therefore, the neutral form of the malonic acid is more reactive than the
oxalic acid (where no reaction is observed at all) and has a rate constant of
(5.1 2.0) 104 L molK1 sK1 comparable to that measured for a similar compound
as dimethylmalonate [8].
Succinic acid differs from malonic acid just by one additional CH2 group
and the measured rate constant was (5 5) 103 L molK1 K1, an order of magni-
s
tude lower than for malonic acid. This can be speculatively explained by a less
efficient inductive effect of the –COOH group.
Interestingly, under acidic conditions, the most reactive acid is the mesoxalic
one, with a measured rate constant of (2.27 0.50) 106 L molK1 sK1 at pH = 0.
As for oxalic acid, this value is influenced by the chemistry initiated by traces
of mesoxalic monoanion (calculated to be less than the molecular form by a
factor of 40 at pH = 0). Taking advantage of the kinetics measured at intermedi-
ate pH (see below and Table 1), we can subtract the influence of the singly
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

1256
Ph. G. de Sémainville et al.
Scheme 1.
charge compound as being 2.3!107.40 or 5.8!105 L molK1 sK1. The corrected
rate constant is then (1.7 0.5) 106 L molK1 K1. This acid differs from the
s
previous ones as it undergoes gem-diol reactions according to Scheme 1.
At low pH, the gem-diol structure is strongly favoured depending also on
the temperature [15]. At pH = 1.7, Schuchmann et al. [15] showed that the ratio
of diol to the carbonyl structure is larger than 100. Therefore we will consider
that under acidic conditions only the diol is present in the aqueous solution. In
this case, there is no abstractable H atom (as there is no C-H bond and the O-H
bond is too strong to react with NO3).
Interestingly, a similar high reactivity has also been observed for pyruvic acid
which also undergoes gem diol equilibrium [7]. Therefore, one might wonder if
this feature is characteristic for a carbon linked to two OH groups, as hypothe-
sized by Gilbert et al.[16], leading an electron transfer to one OH function. From
the data reported here, we can only speculate about the existence of this pathway.
3.3 Reactivity at intermediate pH
At intermediate pH values i.e., at pH between the two pKas of the acids, the
mono acid dominates and both H abstraction and electron transfer are possible.
And as observed by Exner et al. [17], the reactivity increases drastically. Interest-
ingly, while at low pH the rate constants were spanning over quite a large range,
at intermediate pH all acids exhibit similar reactivity around 106–
107 mol LK1 sK1 (see Table 1), even for mesoxalic acid which was “an outlier”
at low pH.
For oxalic acid, rate constants at 298 K listed in Table 1 are in agreement
with those previously reported by Yang et al. [11] and Zellner et al. [18] who
reported the activation energies at pH = 4.
The two pKas of succinic acid are very close and it is consequently very
difficult to measure the reactivity of the corresponding monoanion. The experi-
ments were therefore conducted at pH = 4.2 close to the first pKa, to limit the
influence of the dianion. The derived rate coefficient was (6.6 1.8) ·
106 L molK1 K1, from which the contribution of the dianion (see below) i.e.,
s
1.8 · 107x10pHKpKa2 which equals 9 · 105 L molK1 K1 was subtracted. At pH
s
equal to the first pKa the monoanion concentration is half of the total acid con-
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

Aqueous Phase Reactivity of Nitrate Radicals (NO3)
1257
centration, this last correction leads to the reported rate constant of
(1.1 0.2) 107 L molK1 sK1
.
3.4 Reactivity at high pH
At high pH (i.e., above the second pKa), the acids are fully deprotonated and
the dianions dominate and under such condition the reactivity further increases,
as can be seen from Table 1.
The longest carbon chain is the one of the succinic acid, where the two
carboxylic functions should have a weaker interaction than for shorter dicarbox-
ylic acids. For succinic acid the reactivity of the dianion is roughly the double
(within the experimental errors) of the reactivity of the monoanion (see Table 1).
This can be easily explained by the doubling of the carboxylate group concentra-
tion. In other words, per –COOK function the reactivity at intermediate and high
pH are similar underlining certainly a similar reaction pathway i.e., a charge
exchange reaction from the carboxylate function.
However, the ratio between the measured rate constants for the monoanion
and dianion decreases for the other acids, and the oxalate dianion showed the
highest reactivity. This highlights the fact that the –COOK groups do stronger
interact leading to an increase in reactivity. The activation parameters for this
reaction with the oxalate dianion have been previously reported by Zellner et al.
[18] at pH = 9. The measured activation energies are in reasonable good agree-
ment with this previous study.
4. Atmospheric implications
As discussed above, the H-abstraction mechanism is a possible reaction channel,
while still minor in most cases, for the undissociated carboxylic acids (i.e., under
acidic conditions), leading to the formation of HNO3. Nevertheless electron
transfer dominates as soon as the pH reached the first pKa of the acid. These
reactions are then in competition with photolysis processes, gas phase chemistry
as well as the uptake into atmospheric aqueous droplets or particles [19–21].
From the measured rate constants, it becomes possible to evaluate the life-
time of an organic compound in solution with respect to a given radical, simply
using the following equation:
1
τ =
(E7)
k
2nd$[radical]
The calculated lifetimes are listed in Table 2 under two extreme scenario i.e.,
under acid (urban pollution) and near neutral conditions (rural or marine case)
applying the NO3 concentrations of atmospheric aqueous phase reported by Herr-
mann et al., 2005 [22]. It must be underlined that OH and NO3 do not have the
same diurnal cycle and therefore reactions with NO3 radical will be important
for the degradation of carbonyl compounds at very low actinic fluxes [23].
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

1258
Ph. G. de Sémainville et al.
Table 2. Aqueous phase rate constants as well as tropospheric lifetimes in urban (acidic condi-
tions) and remote (neutral conditions) areas of the studied compounds. (a) from Ervens et al.
[2]; (b) from Electronic Supplementary Material from Herrmann et al.[22]; (c) calculated from
Ervens et al. [2];. (d) from Walling et al. [25];(e) from Logan [26]; (f) from Gligorovski et al
[27].
Rate constant [MK1 sK1
Aqueous phase
]
tropospheric lifetimes τ [h] – aqueous phase
rural
urban
kOH
kNO
[OH•]
[NO3•]
[OH•]
[NO3•]
3
1.5 · 10K13
M
1.0 · 10K14
M
1.0 · 10K14
M
1.7 · 10K13
M
Malonic acid 1.4 · 106(d)
5.1 · 104
1.4 · 107
2.3 · 106
–
–
2.0 · 104
3.2 · 104
Malonate(*)
1.7 · 108(d)
1.8 · 108 (f)
10.9
–
2.0 · 103
–
–
Mesoxalic
acid
–
1.5 · 102
7.1 · 102
Mesoxalate
1.2 · 108 (f)
2.4 · 107(c)
4.9 · 107
8 · 105
15.4
5.7 · 102
–
–
Oxalic acid
(**)
–
–
1.2 · 103
2.0 · 103
Oxalate
1.6 · 108(a)
2.2 · 108
5.0 · 103
1.4 · 107
11.6
–
1.3 · 102
–
2.0 · 103
–
–
Succinic acid 1.1 · 108 (a)
Succinate(*)
5.0 · 108 (a)
2.5 · 102
3.3 · 105
3.7
–
–
(*) Due to the high pKa values of these acids (malonic, pKa = 5.69: succinic, pKa = 5.48 respectivement), an av-
erage reactivity between the singly and doubly compounds has been calculated.
(**) For oxalic acid, as its reactivity is governed by the singly compound the measured rate constant at pH = 6
has been arbitrarily taken.
Table 2 clearly shows that the OH reaction dominates in the urban scenario
(acidic conditions), due to the higher reactivity of OH radical, which is partly
compensated by higher NO3 concentration. So in most cases the OH radical is the
major oxidant, but the nitrate radical gains importance near neutral conditions.
Interestingly, under neutral conditions, when electron transfer reaction dominates
the reaction between the NO3 radical and the dicarboxylic acids, the nitrate
chemistry will introduce a new chemical pathway in which decarboxylation may
occur. For example, in the marine environment, this may lead to organic com-
pounds escaping more easily the liquid phase i.e., the wet aerosols will act as a
VOC source.
5. Conclusions
This work addressed the oxidation of several dicarboxylic acids by the nitrate
radical using a newly developed experimental technique. The experimental ap-
proach described herein takes advantage of the novel Teflon AF 2400 material.
The physical nature of such tubing allows for the construction of photolysis
reaction cells with very low volumes and long optical path lengths. We have
successfully applied such photolysis cell to the measurement of the rate coeffi-
cients for the liquid phase reaction of the nitrate radical with oxalic, malonic,
mesoxalic and succinic acids as a function of pH and temperature.
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

Aqueous Phase Reactivity of Nitrate Radicals (NO3)
1259
These acids, depending upon the pH, may react with the nitrate radical
through different pathways i.e., either H abstraction (at all pH at the C-H bond)
and charge exchange reaction (for pH above the first pKa). Clearly when compar-
ing the kinetics, the reaction is much faster under alkaline conditions strongly
suggesting that the charge exchange transfer is much faster than the H abstrac-
tion. Also when going from the singly and doubly charged species, the rate
constant roughly doubles following the change in the COOK concentration.
For not (too) acidic pH of droplets and.or aerosols the charge exchange
reaction at the carboxylate functional group might compete with the oxidation
by hydroxyl radical. Also, for liquid particles enriched in nitrate (such as a deli-
quescent ammonium nitrate), the lifetime associated to a carboxylic acid being
oxidised by NO3 in may be as short a few hours (or less). Accordingly, one could
speculate that any fatty acid (such as oleic acid) acting a surfactant deposited on
such a deliquescent particle may be oxidised in both phase i.e., by ozone from
the gas phase and by the nitrate radical within the particle. In addition, a charge
transfer reaction on the carboxylate function may initiate a decarboxylation proc-
ess and desorption from the particle of volatile hydrocarbons.
Acknowledgement
Support of this work by the Programme National de Chimie Atmosphérique
(PNCA) from the CNRS is gratefully acknowledged along the Ph-D grant for
PGS from the French research ministry. CG thanks Reinhard Zellner for having
stimulated the study of the nitrate radicals.
References
1. P. Saxena, L. M. Hildemann, J. Atmos. Chem. 24 (1996) 57.
2. B. Ervens, S. M. Kreidenweis, Environ. Sci. Technol. 41 (2007) 3904.
3. H. Herrmann, Chem. Rev. 103 (2003) 4691.
4. L. Deguillaume, M. Leriche, K. Desboeufs, G. Mailhot, C. George, N. Chaumerliac,
Chem. Rev. 105 (2005) 3388.
5. F. Schweitzer, P. Mirabel, C. George, J Phys. Chem. A 104 (2000) 72.
6. R. Sander, Surveys in Geophysics 20 (1999) 1.
7. P. G. de Semainville, D. Hoffmann, C. George, H. Herrmann, Phys. Chem. Chem.
Phys. 9 (2007) 958.
8. D. Rousse, C. George, Physical Chemistry Chemical Physics 6 (2004) 3408.
9. C. George, H. El Rassy, J. M. Chovelon, Internat J. Chem. Kin. 33 (2001) 539.
10. C. George, D. Rousse, E. Perraudin, R. Strekowski, Phys. Chem. Chem. Phys. 5
(2003) 1562.
11. X. K. Yang, J. Q. Wang, T. D. Wang, Chin. Chem. Lett. 15 (2004) 583.
12. J. Blanksby Stephen, G. B. Ellison, Acc.Chem. Res. 36 (2003) 255.
13. R. A. Witter, P. Neta, J. Org. Chem. 38 (1973) 484.
14. P. Neta, R. H. Schuler, J. Phys. Chem. 76 (1972) 2673.
15. M. N. Schuchmann, H. P. Schuchmann, M. Hess, C. Von Sonntag, J. Amer. Chem.
Soc. 113 (1991) 6934.
16. B. C. Gilbert, J. K. Stell, W. J. Peet, K. J. Radford, J. Chem. Soc., Faraday Transac-
tions 1: Physical Chemistry in Condensed Phases 84 (1988) 3319.
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM

1260
Ph. G. de Sémainville et al.
17. M. Exner, H. Herrmann, R. Zellner, Ber. Bunsen-Gesellschaft 96 (1992) 470.
18. R. Zellner, H. Hermann, M. Exner, H. W. Jacobi, G. Raabe, A. Reese, Transport
and Chemical Transformation of Pollutants in the Troposphere 2 (1996) 146.
19. J.-P. Le Crane, E. Villenave, M. D. Hurley, T. J. Wallington, J. C. Ball, J. Phys.
Chem. A 109 (2005) 11837.
20. B. Cabanas, S. Salgado, P. Martin, M. T. Baeza, E. Martinez, J.Phys. Chem. A 105
(2001) 4440.
21. A. Mellouki, G. Le Bras, H. Sidebottom, Chem. Rev. 103 (2003) 5077.
22. H. Herrmann, A. Tilgner, P. Barzaghi, Z. Majdik, S. Gligorovski, L. Poulain, A.
Monod, Atmos. Environ. 39 (2005) 4351.
23. A. Tilgner, Z. Majdik, A. M. Sehili, M. Simmel, R. Wolke, H. Herrmann, Atmos.
Environ. 39 (2005) 4389.
24. V. E. Tumanov, E. A. Kromkin, E. T. Denisov, Russian Chemical Bulletin (Transla-
tion of Izvestiya Akademii Nauk, Seriya Khimicheskaya) 51 (2002) 1641.
25. C. Walling, G. M. El-Taliawi, J. Amer. Chem. Soc. 95 (1973) 844.
26. S. R. Logan, J. Chem. Soc., Perkin Transactions 2: Physical Organic Chemistry
(1972–1999) (1989) 751.
27. S. Gligorovski, D. Rousse, C. H. George, H. Herrmann, Int. J. Chem. Kin. 41 (2009)
309.
Brought to you by | University of Florida
Authenticated
Download Date | 10/28/17 12:23 PM