Heterogeneous reaction of NO2 on fresh and coated soot surfaces

Article abstract of DOI:10.1021/jp1021938

The heterogeneous reaction of nitrogen dioxide (NO₂) on fresh and coated soot surfaces has been investigated to assess its role in night-time formation of nitrous acid (HONO) in the atmosphere. Soot surfaces were prepared by incomplete combustion of propane and kerosene fuels under lean and rich flame conditions and then processed by heating to evaporate semivolatile species or by coating with pyrene, sulfuric acid, or glutaric acid. Uptake kinetics and HONO yield measurements were performed in a low-pressure fast-flow reactor coupled to a chemical ionization mass spectrometer (CIMS), using atmospheric-level NO₂ concentrations. The uptake coefficient and the HONO yield upon interaction of NO₂ with nascent soot depend on the type of fuel and combustion regime and are the highest for samples prepared using fuel rich flame. Heating the nascent soot samples before exposure to NO₂ removes the organic material from the soot backbone, leading to a significant increase in NO₂ uptake coefficient and HONO yield. Continuous exposure to NO₂ reduces the reactivity of soot because of irreversible deactivation of the surface sites. Our results support the oxidation-reduction mechanism involving adsorptive and reactive centers on soot surface where NO₂ is converted to HONO and other products. Coating of the soot surface by different materials to simulate atmospheric aging has a strong impact on its reactivity toward NO₂ and the resulting HONO production. Coating of pyrene has little effect on either reaction rate or HONO yield. Sulfuric acid coating does not alter the uptake coefficient, but significantly reduces the amount of HONO formed. Coating of glutaric acid significantly increases NO₂ uptake coefficient and HONO yield. The results of our study indicate that the reactivity and HONO generating capacity of internally mixed soot aerosol will depend on the chemical composition of the coating material and hence will vary considerably in different polluted environments.

Full text of DOI:10.1021/jp1021938

7516
J. Phys. Chem. A 2010, 114, 7516–7524
Heterogeneous Reaction of NO2 on Fresh and Coated Soot Surfaces
Alexei F. Khalizov, Miguel Cruz-Quinones, and Renyi Zhang*
Department of Atmospheric Sciences and Department of Chemistry, Texas A&M UniVersity, College Station,
Texas 77843
ReceiVed: March 10, 2010; ReVised Manuscript ReceiVed: May 10, 2010
The heterogeneous reaction of nitrogen dioxide (NO2) on fresh and coated soot surfaces has been investigated
to assess its role in night-time formation of nitrous acid (HONO) in the atmosphere. Soot surfaces were
prepared by incomplete combustion of propane and kerosene fuels under lean and rich flame conditions and
then processed by heating to evaporate semivolatile species or by coating with pyrene, sulfuric acid, or glutaric
acid. Uptake kinetics and HONO yield measurements were performed in a low-pressure fast-flow reactor
coupled to a chemical ionization mass spectrometer (CIMS), using atmospheric-level NO2 concentrations.
The uptake coefficient and the HONO yield upon interaction of NO2 with nascent soot depend on the type
of fuel and combustion regime and are the highest for samples prepared using fuel rich flame. Heating the
nascent soot samples before exposure to NO2 removes the organic material from the soot backbone, leading
to a significant increase in NO2 uptake coefficient and HONO yield. Continuous exposure to NO2 reduces the
reactivity of soot because of irreversible deactivation of the surface sites. Our results support the
oxidation-reduction mechanism involving adsorptive and reactive centers on soot surface where NO2 is
converted to HONO and other products. Coating of the soot surface by different materials to simulate
atmospheric aging has a strong impact on its reactivity toward NO2 and the resulting HONO production.
Coating of pyrene has little effect on either reaction rate or HONO yield. Sulfuric acid coating does not alter
the uptake coefficient, but significantly reduces the amount of HONO formed. Coating of glutaric acid
significantly increases NO2 uptake coefficient and HONO yield. The results of our study indicate that the
reactivity and HONO generating capacity of internally mixed soot aerosol will depend on the chemical
composition of the coating material and hence will vary considerably in different polluted environments.
1. Introduction
accelerate the O3 production in the morning and lead to a notable
increase in the daytime O3 concentration.8 This study, however,
assumed a quantitative conversion of NO2 to HONO on soot
aerosol with a time-independent reaction probability that might
have overestimated the contribution from heterogeneously
produced HONO to the ozone budget.
Nitrous acid (HONO), an important night-time reservoir of
hydroxyl radical (OH), can be readily photolyzed after dawn,
releasing OH into the planetary boundary layer.1,2 Mixing ratios
of HONO up to several parts per billion (ppb) have been
observed in urban areas,3 implicating that HONO plays an
important role in O3 production, particularly shortly after sunrise.
A number of mechanisms have been proposed4 to explain the
high night-time concentrations of HONO observed in field
measurements.1 However, model calculations including all
known sources, such as gas-phase chemistry, heterogeneous
aqueous chemistry of NOx, and direct emission from vehicle
exhaust could not explain the observed high levels of nitrous
acid.
The heterogeneous reaction of NO2 on soot aerosol leading
to the formation of HONO has attracted attention as a possible
night-time source of nitrous acid.5 Soot aerosol, which consists
primarily of elemental carbon with variable fraction of organic
materials, originates from incomplete combustion of fossil fuels
and biomass, with an average emission rate of 8-24 Tg of
carbon per year6 and mass loadings 1.5-20 µg m-3 in the urban
boundary layer.7 Although carbon soot accounts for less than
10-15% of the total atmospheric aerosol mass, soot particles
may provide a significant fraction of available aerosol reaction
surface due to their fractal morphology. A recent modeling study
has shown that heterogeneous conversion of NO2 to HONO on
soot aerosol surfaces in polluted urban environment can
Because of its atmospheric relevance, heterogeneous conver-
sion of NO2 to HONO has been extensively studied in the
laboratory for a broad range of NO2 concentrations on a variety
of carbonaceous materials, including commercial black carbon,
hydrocarbon flame soot, spark discharge soot, diesel soot, power
plant soot, etc.9-16 Despite the large number of studies, the
heterogeneous reaction mechanism leading to HONO formation
is not well understood, and significant discrepancies exist in
the measured NO2 kinetic uptake coefficients and HONO yields.
Depending on the measurement technique, type of soot, assumed
surface area, and initial concentration of NO2, the reported
uptake coefficients vary over 7 orders of magnitude, from 10-1
to 10-8, and HONO yields vary from a few percent to about
100%. As discussed by Aubin and Abbatt,16 large uptake
coefficients reported in early studies were caused by referencing
the kinetic measurements to geometric rather than internal
surface areas of samples, whereas low HONO yields might have
originated from factors such as different flame combustion
conditions and incomplete desorption of HONO from the soot
surface in low temperature experiments.
Furthermore, the reactivity of soot can be significantly
affected by atmospheric aging through surface oxidation or
coating. Most previous studies reported that oxidative aging of
the soot surface by O3 and NO2 reduces its reactivity and HONO
* To whom correspondence should be addressed. Phone: 979-845-7656.
Fax: 979-862-4466. E-mail: Zhang@ariel.met.tamu.edu.
10.1021/jp1021938  2010 American Chemical Society
Published on Web 06/24/2010

Heterogeneous Reaction of NO2 on Soot
J. Phys. Chem. A, Vol. 114, No. 28, 2010 7517
formation capacity irreversibly.12,14,16 In a few cases, however,
reactivation of soot has been observed after heating10 or in the
presence of water vapor.17 Treating the soot surface with gaseous
nitric acid showed no significant impact on the NO2 heteroge-
neous reaction probability or HONO yield for atmospherically
relevant HNO3 concentration levels.15 For soot coated by sulfuric
acid the NO2 uptake rate was almost independent of the amount
of adsorbed H2SO4, but the HONO yield was reduced to zero
for a surface coverage greater than 1014 H2SO4 cm-2.11 To date,
the effect of surface processing on the heterogeneous reaction
of soot with NO2 has been investigated only for a limited number
of inorganic species, and either a decrease or no change in the
HONO production has been reported. Atmospheric soot is
commonly found to be thickly coated by inorganic species18
and functionalized organic matter,19 formed from photo-oxida-
tion of hydrocarbons.20-22 Particularly, low-molecular weight
dicarboxylic organic acids represent a significant component
of fine particulate matter in the troposphere.23 However, the
effect of organic coatings on soot reactivity toward NO2 has
not yet been investigated.
In the present study, the heterogeneous reaction of hydro-
carbon flame soot with atmospheric-level concentration of NO2
was investigated in a low-pressure, fast-flow reactor coupled
to a chemical ionization mass spectrometer (CIMS). Both fresh
and coated soot samples were examined to assess the role of
soot aerosol internal mixing in the atmosphere on its reactivity
and HONO generation capacity. The coated materials included
pyrene, sulfuric acid, and glutaric acid, which readily adsorb
on soot surface,24-26 forming coatings that can alter the
morphology, hygroscopicity, and optical properties of fractal
soot particles.27-30
injector tube to distribute the vapor of the coating material
uniformly throughout the soot surface. The coating process was
carried out for 30 min for each soot sample. Selected soot
samples were exposed to 100% relative humidity (RH) by
storing them in a desiccator above pure water. In a few cases,
fresh soot samples were treated by iso-propanol and toluene to
extract polar and nonpolar soluble organic materials, respec-
tively. A soot-coated glass tube was carefully immersed in a
graduated cylinder with corresponding solvent and left for
several hours. After extraction, adsorbed solvent traces were
removed by pumping on the sample in vacuum at room
temperature for several hours until no apparent degassing has
been observed. Although we did not verify the extraction
efficiency directly, NO2 uptake experiments indicated that the
treatment by iso-propanol removed enough polar organic
material to significantly increase soot reactivity. Neither visual
appearance nor integrity of the sample was affected by iso-
propanol extraction. On the contrary, extraction by toluene,
while having no effect on reactivity, often led to partial
disintegration of the soot film and its dislodgement from the
glass tube surface, indicating that toluene removed nonpolar
(e.g., PAH) organic material cementing soot to the glass
surface.
Fresh and coated soot samples were examined for the
presence of different functional groups and the coating
material by attenuated total reflection s fourier transmission
spectroscopy (ATR-FTIR). For this purpose, soot samples
were collected on aluminum foil that was placed inside the
Pyrex sample tubes. After sample collection, the foil was
removed from the Pyrex tube and cut into 1 cm wide strips.
Each strip was placed, with the soot side downward, onto
the surface of a ZnSe crystal (Pike Technologies) of the ATR
accessory, and an infrared spectrum was recorded using a
Nicolet Magna 560 spectrometer equipped with a liquid
nitrogen cooled MCT detector. Typically, 60 scans at a
resolution of 4 cm-1 over a range from 4000 to 600 cm-1
were averaged to obtain a spectrum. The ATR crystal was
cleaned between different samples to avoid cross contamina-
tion. A background spectrum with no sample on the face of
the crystal was collected before each sample spectrum. An
ATR correction was routinely applied to all reported spectra,
using an algorithm implemented in the FTIR instrument
software.
2. Experimental Section
2.1. Soot Sample Preparation and Characterization. Soot
was produced by incomplete combustion of propane and
kerosene fuels in a diffusion flame formed by a commercial
torch and a kerosene lantern with a glass chimney, respectively.
A flame was established by altering the amount of fuel with a
valve (propane torch) or by adjusting the length of a wick
(kerosene lantern). Soot films were deposited on the inner
surface of Pyrex glass tubes (10 cm length and 1.6 cm inner
diameter) under rich and lean flame conditions, corresponding
to type A and type B samples, respectively.25 Type A soot films
were prepared by allowing the reducing part of flame to touch
the soot surface and had a grayish appearance due to the
presence of organic matter. It has been reported that “activation”
of the soot surface by flame changes its adsorptive behavior
toward nitric acid31 and organic acids25 and may also be
responsible for variability in chemical reactivity of otherwise
identical soot samples.32 During preparation of type B films,
the tube was held high above the oxidizing flame tip, resulting
in a pitch-black soot surface composed mostly of elemental
carbon. Visually uniform soot films were obtained by rotating
the sample tube during the coating process, and typical soot
sample mass ranged from 1 to 30 mg.
To study the effect of aging on heterogeneous reactivity and
HONO yield, soot films were exposed to pyrene, glutaric
(pentanedioic) acid, and sulfuric acid vapors in an evacuated
glass container with a sample of the coating material placed at
the bottom. The bottom part of the container was heated to
85-100 °C, depending on the material, whereas the upper part
with the soot-coated tube was kept at room temperature. A 20
sccm (standard cubic centimeters per minute) helium carrier flow
was supplied to the bottom of the container through a glass
Internal surface areas of soot samples were characterized from
Brunauer, Emmett, and Teller (BET) isotherms33 measured by
Kr adsorption at 77 K, as described in our previous study.25 An
improved constant volume adsorption cell with a vacuum jacket
at the top34,35 was used to hold the samples. Soot samples were
degassed at room temperature before BET measurements, unless
stated otherwise.
2.2. Measurements of NO2 Uptake and HONO Produc-
tion. Heterogeneous interaction of NO2 with soot surfaces was
studied in a fast-flow coated-wall reactor coupled to a chemical
ionization mass spectrometer (CIMS)25 as shown in Figure 1.
A glass tube with the soot-coated inner surface was inserted
into the flow reactor of 30 cm length and 2.0 cm inner diameter,
through which a flow of 130 sccm helium carrier gas at a 1.5
Torr total pressure was established, resulting in a 530 cm s-1
laminar flow. Gaseous NO2 was introduced to the flow reactor
at 1-2 sccm through a movable injector from a glass bulb
containing 20-100 ppm NO2 in He at 800 Torr. The reaction
time was varied by changing the injector position relative to
that of the soot-coated tube. Uptake of NO2 and evolution of
-
HONO were monitored by following the signals of NO2 and

7518 J. Phys. Chem. A, Vol. 114, No. 28, 2010
Khalizov et al.
Figure 1. Schematic of a low-pressure fast-flow reactor coupled to a chemical ionization mass spectrometer.
HF·NO2- ions produced in ion-molecule reactions of NO2 (eq
- 36
1) and HONO (eq 2) with the reagent ion, SF6
:
NO2 + SF-6 f NO2- + SF6, k1
(1)
(2)
HONO + SF-6 f HF- · NO2 + SF5, k2
Reagent ions were generated by corona discharge in nitrogen
carrier gas in the presence of about 10 ppm SF6. Copper and
stainless steel tubing and a gas purifier (P-300-2, VICI) were
implemented in the CI system to establish an oxygen-free
environment (less than 50 ppb O2) in order to minimize NO2
background. Additionally, the CI region of the mass spectrom-
-
eter was modified to minimize the dissociation of the HF ·NO2
ion clusters through collisions with bath gas molecules in the
electrical field. In several selected experiments HF- ·NO3 ion
was monitored to detect formation of HNO3 in the reaction of
soot with NO2. Concentration of NO2 in the flow reactor was 5
× 109 to 5 × 1011 molecules cm-3, corresponding to mixing
ratios of about 0.2-20 ppb at 760 Torr pressure. Previously
-
measured rate coefficients for reaction of NO2 with SF6 and
for fluoride ion transfer from SF6 to HONO are 1.4 × 10-10
-
and 6 × 10-10 cm3 molecule-1 s-1, respectively.17 Since the
latter value is not known with adequate accuracy, additional
calibration measurements were performed using two different
HONO sources in order to calculate the HONO/NO2 concentra-
tion ratio from the ratio of their mass spectrometer signals.
In the first method, HONO was synthesized by slowly adding
5 mL of 0.1 M NaNO2 to 10 mL of 40 wt % H2SO4 that was
chilled to 273 K and vigorously stirred.36 During calibration, a
flow of about 100 sccm N2 was passed through the aqueous
HONO solution in a bubbler and then directed into a 11.0 cm
long quartz absorption cell located inside a UV/vis Perkin-Elmer
spectrophotometer, where optical absorbances at 250 and 450
nm due to HONO + NO2 and NO2, respectively were measured.
Immediately after the cell, the HONO flow was diluted into a
few standard liters per minute stream of He before being
introduced to the CIMS. The final mixing ratios of HONO
ranged between 0.1 and 10 ppb (at 760 Torr). The second
method involved the reaction between gaseous hydrogen
chloride and solid sodium nitrite, as previously described by
Aubin and Abbatt.16 HCl from a permeation tube (VICI
Metronics) maintained at 40 °C was carried by a flow of 200
sccm He into a PFA Teflon column packed with NaNO2, where
HONO was formed. The CIMS measurement showed that
HONO produced by this method was NO2-free, hence hetero-
geneous conversion of HONO to NO and NO2 on the NaNO2
and tubing wall surfaces was negligible.16 The concentration
of HONO was quantified by a NOx monitor (Thermo Scientific).
To determine the k2/k1 ratio, NO2 was added to the flow reactor
from a bulb along with HONO as described above. HONO
Figure 2. Uptake of NO2 and release of HONO on type B soot samples
prepared using propane flame: (a) 1.9 mg of fresh soot; (b) 1.9 mg of
preheated soot; (c) 1.2 mg of fresh soot extracted by iso-propanol.
calibrations produced by two independent methods agreed within
(15%, resulting in a k2/k1 ratio of 9.6, which was a factor of
2.2 higher than the value based on previous measurements.17
3. Results and Discussion
3.1. NO2 Uptake and HONO Formation on Different
Types of Soot. Figure 2a presents temporal profiles for NO2
uptake and HONO formation on fresh type B propane soot. The
initial drop of NO2 signal at 10 min corresponds to retracting
the injector by 8 cm, thus exposing an 8 cm length of the soot
film to NO2. The initial rapid uptake of NO2 is followed by a
slow asymptotic recovery, but complete saturation of the soot
sample does not occur, and even after long exposure the NO2
signal remains below its initial level. When NO2 partitions to
the surface, a simultaneous increase in the HONO signal is
observed. The concentration of HONO reaches a maximum and
then decreases exponentially and remains at a constant nonzero
level. When NO2 exposure is terminated by moving the injector
downstream the soot sample, the concentration of HONO drops
to zero almost instantaneously, whereas NO2 signal peaks and
then returns to the initial level. This intermittent rise corresponds

Heterogeneous Reaction of NO2 on Soot
J. Phys. Chem. A, Vol. 114, No. 28, 2010 7519
TABLE 1: Number of NO2 Molecules Taken and HONO
Molecules Released for Fresh, Pre-Heated, and Extracted
Type B Propane Flame Soot Samples
TABLE 2: HONO Formation from Heterogeneous Reaction
of NO2 on Different Types of Soota
HONO × 10-15, molecules mg-1
NO2 × 10-15
,
HONO × 10-15
,
type A
type B
soot sample
molecules mg-1
molecules mg-1
fuel
fresh
heated
fresh
heated
fresh
2.5 ( 0.9
9.0 ( 0.9
6.3 ( 0.9
2.0 ( 0.9
1.1 ( 0.2
7.2 ( 0.2
4.7 ( 0.2
0.8 ( 0.2
preheated to 300 °C
extracted by iso-propanol
extracted by toluene
propane
1.5 ( 0.2
(58 ( 21)
2.5 ( 0.2
(86 ( 19)
5.4 ( 0.2
(74 ( 11)
3.8 ( 0.2
(85 ( 10)
0.8 ( 0.2
(38 ( 12)
0.8 ( 0.4
(53 ( 21)
7.6 ( 0.2
(72 ( 9)
7.8 ( 0.2
(91 ( 8)
kerosene
to a release of NO2 that is physically adsorbed on the soot
surface. The observed variation in NO2 signal, showing a fast
reversible and a slow irreversible uptake, is in agreement with
the heterogeneous mechanism that involves physical adsorption
of NO2 followed by chemical conversion to HONO through the
reaction with reductive sites, for example, hydride, in an
otherwise mostly elemental carbon backbone.9-12 The fast
decrease in the NO2 uptake and HONO formation rates in the
first few minutes of exposure, as displayed in Figure 2, have
been previously explained by a temporary depletion of the
surface adsorption sites.12 Subsequent slow conversion of NO2
to HONO may occur for hours. We used the area between
recovery and steady state signals of NO2 and the area under
the signal of HONO to calculate the total number of NO2
molecules irreversibly reacted on soot and HONO molecules
produced, respectively. During the 70 min exposure period, 2.5
× 1015 NO2 molecules are consumed, whereas 1.1 × 1015
HONO molecules are formed per 1 mg of fresh soot (Table 1).
The instantaneous HONO yield, defined as [HONO]/(∆[NO2]),
remains fairly constant during reaction to the point where the
NO2 signal recovers to about 85% of its initial value. Note that
this HONO yield (about 44%) was among the smallest yields
for fresh soot samples observed in our study.
a HONO yields (%) are given in parentheses; sample mass is
1-4 mg.
TABLE 3: BET Specific Surface Areas of Type A and Type
B Propane and Kerosene Soot Samplesa,b
BET specific surface area, m2 g-1
fuel
type A
type B
propane
79
147
kerosene
67 (61)
106 (116)
a Determined in our previous study;25 before measurement,
samples were preheated to 200 °C for 1 h. Values in parentheses
are obtained in the present work. b Geometric surface area of the
sample tube is 50.3 cm2.
the enhancement in reactivity is not as pronounced as in the
case of preheated soot. On the contrary, extraction of soot by
toluene has no effect on the number of NO2 molecules reacted
or HONO molecules formed (Table 1). On the basis of the
results of heating experiments, we conclude that most of
adsorptive and reactive sites in our samples belong to the soot
backbone rather than condensed organic material. Also, extrac-
tion experiments provide tentative evidence that molecules
blocking these sites are of polar nature.
Preheating the soot sample to 300 °C in vacuum before the
uptake experiment drastically increases its reactivity, as indicated
by a larger drop in the NO2 signal and larger increase in the
HONO signal upon retracting the injector (Figure 2b). The
number of NO2 molecules reacted and HONO molecules formed
are higher than those of unheated soot, by a factor of 3.6 and
6.5, respectively (Table 1). Incomplete combustion of hydro-
carbon fuels produces a broad range of semivolatile organic
products, such as polycyclic aromatic hydrocarbons (PAH),
saturated and unsaturated hydrocarbons, and partially oxidized
organics (e.g., alcohols, carbonyls, organic acids), which can
condense on soot upon cooling.37,38 Apparently, the increased
reactivity of heated soot is caused by removal of incomplete
combustion products, rendering more adsorptive and/or reactive
surface sites. The presence of condensed material on the soot
surface is indirectly confirmed by observing a thin organic film
that builds up over time on the cool top part of the vacuum
container used to heat soot-coated Pyrex tubes. The film is the
thickest for kerosene soot and barely observable for propane
soot samples, indicating that kerosene soot contains a larger
organic fraction. Individual samples do not show a detectable
decrease in their mass after heating, which indicates that even
for kerosene soot the mass fraction of condensed organic
material is less than 2%. Also, heating caused no visible changes
in the infrared spectra of soot.
The amount of organic material on soot and its oxidation state
depend on the flame equivalence ratio37 (actual fuel to oxidizer
ratio over the fuel to oxidizer ratio at stoichiometric conditions)
and sample preparation method.12 Soot particles are chain-like
aggregates of elemental carbon spheres that are clearly identifi-
able in the soot produced under fuel-lean conditions,18 but they
appear blended and with less clear boundaries due to organic
material in the case of a fuel-rich diffusion flame.26 For the
diffusion flame used in the present study, the fraction of organic
carbon increases with the C/H ratio of the hydrocarbon fuel;
also, it is higher for type A soot samples that are prepared by
allowing the reducing part of the flame to touch the soot surface.
Table 2 shows that the total amount of HONO formed upon
heterogeneous interaction of NO2 with soot depends on the fuel
and sample preparation method. Among unheated samples, type
A soot produces the highest amount of HONO, particularly when
kerosene is used to generate soot films, whereas type B soot
produces the least amount of HONO. After heating, however,
the largest HONO production is observed for type B soot.
Although fresh type A soot is already exposed to heat from the
fame during preparation, this soot is not the same as preheated
type B soot, because of the differences in their visual appear-
ance, HONO formation capacity (Table 2), BET surface area
(Table 3), and reactivity (Table 4).
To examine the nature of condensed organic materials and
their role on soot reactivity, we have investigated soot samples
that were extracted by iso-propanol and toluene, which are
expected to remove polar and nonpolar organics from the soot
backbone, respectively. As shown in Figure 2c, soot samples
treated by iso-propanol consume more NO2 and form more
HONO than the fresh soot for a similar exposure time, although
A factor of 6.5-9.6 increase in the amount of HONO formed
on heated type B soot cannot be explained merely from
increased BET surface area, because the latter remains un-
changed after heating (Figure 3). Since soot is a highly porous
material, an important implication of the unchanged internal
surface area is that the condensed material does not block pore
throats. The internal surface area of soot samples calculated from

7520 J. Phys. Chem. A, Vol. 114, No. 28, 2010
Khalizov et al.
TABLE 4: Irreversible Uptake Coefficients of NO2 on Soot
Calculated Using BET Surface Areas of Samplesa
uptake coefficient, γBET × 105
type A
type B
fuel
fresh
heated
fresh
heated
propane
kerosene
4.9 ( 3.0
1.6 ( 0.5
23.0 ( 7.0
4.7 ( 1.4
1.0 ( 0.7
1.6 ( 0.6
4.3 ( 1.5
4.6 ( 1.7
a Sample mass is 1-7 mg.
Figure 3. Dependence of the measured BET internal surface area of
type A (6 mg) and type B (9 mg) kerosene flame soot on the sample
preheating temperature.
Figure 5. Uptake of NO2 and release of HONO on 16 mg type A
propane flame soot: (a) fresh soot sample; (b) preheated soot sample.
progressively lower asymptotic NO2 signal and less distinct
reversible part at longer reaction times. The uptake of NO2 on
heated samples consists of irreversible part only and leads to
higher HONO production (Figure 5b). A 50-70% drop in the
NO2 signal and a corresponding increase in the HONO signal
are observed upon retracting the injector by 2 cm; however,
little additional NO2 uptake or HONO formation occurs when
the exposed soot area is increased by further retracting the
injector. This behavior indicates that gas-phase diffusion of NO2
toward the sample surface and possibly inside the sample pores
limit the rate of heterogeneous uptake on preheated soot.
Furthermore, when the exposure is stopped by returning the
injector to its initial position, little NO2 is released by soot
because all physically adsorbed nitrogen dioxide molecules have
been converted to HONO or other species. Thus, whereas for
unheated soot the overall uptake rate is limited by the number
of vacant adsorptive and reactive sites, for heated samples
diffusion and physical adsorption correspond to the limiting
steps.
The asymptotic portions of the uptake curves at different
injector positions were used to calculate the first-order loss
constants, kobs, by plotting the logarithm of NO2 signal versus
the soot-NO2 interaction time (Figure 6). Note that kobs corre-
sponds to the irreversible reactive uptake and does not include
the fast initial physical uptake. For the majority of experiments
with preheated soot, only two points could be used in the
regression analysis because no additional NO2 loss was observed
after retracting the injector beyond 2 cm length of the soot film.
In such cases the observed loss constants and corresponding
uptake coefficients derived for preheated soot samples represent
a lower limit to actual values. The wall-loss rate constant, kw,
was calculated from the observed first-order loss by taking into
account the diffusion of NO2 in helium from the center of the
reactor to the wall (eqs 3 and 4):39-41
Figure 4. The integral number of NO2 molecules taken and HONO
molecules released by preheated type B propane soot samples of
different mass.
BET isotherms is 2 orders of magnitude larger than their
geometric surface (Table 3). To determine the actual surface
area involved in the reaction with NO2, a series of uptake
experiments were performed using soot samples of varied mass.
Figure 4 shows that the number of NO2 molecules consumed
and HONO molecules released increase linearly with the soot
sample mass initially, but level off for samples in excess of
about 4 mg, which corresponds to a 0.1 mg cm-2 critical surface
density, in close agreement with the values measured in previous
studies.10,13 Even for soot samples with mass below 4 mg, total
HONO production of (0.1-0.8) × 1016 molecule mg-1 is at
the lower range of previously reported data, (1-4) × 1016
molecule mg-1 10,12,14
This can be attributable to that NO2 is
.
unable to reach deeper layers of soot on the time scale of our
experiments and hence not all of internal surface areas of soot
films are available for the heterogeneous reaction because of
low NO2 concentrations used in our study.
3.2. Kinetics of NO2 Uptake. Measurements of the uptake
coefficient were performed by exposing varying lengths of the
soot film to NO2 while monitoring the change in NO2 and
HONO signals. Unheated soot samples display a combination
of reversible and irreversible uptakes (Figure 5a). Stepwise
retracting of the injector leads to an uptake pattern with
1
kobs
1
kd
1
kw
)
+
(3)

Heterogeneous Reaction of NO2 on Soot
J. Phys. Chem. A, Vol. 114, No. 28, 2010 7521
NO2 concentrations (ca. 5 × 1011 molecules cm-3) results in
gradual deactivation of soot samples on a time scale of a single
kinetic experiment. Also, the reactivity of soot samples that are
exposed to ambient air for several days is reduced by more than
1 order of magnitude.
NO2 uptake coefficients, γΒΕΤ, were calculated according to
eq 5 using internal surface areas of soot films, ABET, obtained
from BET measurements (Table 3):
DP,He
pHe
3.66
r2
kd )
(4)
where DP,He is the diffusion coefficient of NO2 in helium at
pressure PHe and r is the radius of the soot coated tube. Figure
7 shows that initially kw increases with the sample mass, but
levels off for samples in excess of 5-10 mg, in close agreement
with the mass dependence observed for the total number of NO2
molecules reacted on soot.
Large variability between measurements conducted for dif-
ferent soot samples may originate from variation in soot film
properties, such as different amount of semivolatile materials
generated by flame or nonuniformity in the soot film thickness
along the tube length. Control experiments show good reproduc-
ibility for consecutive measurements carried out using the same
soot sample. For each repetitive exposure, the NO2 loss rate
decreases by less than 5-10%. Hence, only a small fraction of
soot reactive sites are consumed in kinetic experiments carried
out over short time scale using low NO2 concentrations (1-2
× 1010 molecules cm-3). Use of longer exposure times or higher
2kw
2rkw Ageom
ω ABET
V
γBET
)
)
(5)
ω Ageom
where ω is the mean molecular speed of NO2 and Ageom is
geometric surface area of the sample. The results are sum-
marized in Table 4, which presents values of γΒΕΤ averaged
over several soot samples with masses in the range 1-7 mg.
Fresh soot has the lowest reactivity, γΒΕΤ ) (1-5) × 10-5
,
which increases by a factor of 3-5 after preheating of the
samples in vacuum to 300 °C. The effect of heating may explain
the variation in the reactivity and HONO production capacity
of soot samples. Our kinetic data are in good agreement with
previously reported NO2 uptake coefficients (γBET ) 10-6 to
10-4) that were derived using BET surface areas.16
3.3. Mechanism of NO2 Uptake and HONO Formation.
Observation of HONO yields in excess of 50% and absence of
HNO3 among the reaction products unambiguously rule out the
surface-catalyzed disproportionation of NO2 in water as a major
pathway for HONO formation (reaction 6):
2NO2 + H2O f HONO + HNO3
(6)
Pumping on the soot samples for extended period of time to
remove adsorbed water has no effect on the HONO yield,
indicating that water is not involved in the redox reaction of
NO2 with soot active sites (reaction 7),10
Figure 6. Kinetics of NO2 loss on different types of propane flame
soot (type A and B soot sample mass is 8 and 7 mg, respectively).
NO2(ad) + H2O(ad) + reduced (red) soot f HONO(ad)
+
oxidized (ox) soot (7)
Also, we observed no reactivation of the NO2-processed soot
samples after exposure to elevated relative humidity.
Our results support a mechanism in which the soot surface
itself provides a hydrogen atom for NO2 to form HONO.10,12,16
Initial uptake of NO2 occurs through physical adsorption
(reaction 8) and corresponds to the reversible part of curves
shown in Figures 2 and 5a (brackets denote adsorbed species
and surface sites).
NO2 + {S1} T {NO2 · S1}
(8)
{NO2 · S1} + {C-H}red f {S1} + {HONO} + {C}ox
(9)
{NO2 · S1} + {S2} f {S1} + {NO2-S2}
(10)
{HONO} f HONO
(11)
(12)
{HONO} f NO + other products
Figure 7. Dependence of the NO2 loss rate on the soot sample mass:
(a) propane flame soot; (b) kerosene flame soot. For all data the sample
mass corresponds to a 10 cm length of the soot coating.
HONO is formed through chemical interaction of the adsorbed
NO2 with reducing surface sites, {C-H}red, such as allylic

7522 J. Phys. Chem. A, Vol. 114, No. 28, 2010
Khalizov et al.
hydrogens (reaction 9).16,42 Soot acts as a consumable reactant
rather than a catalyst and cannot be recycled. Reaction 10 has
been previously invoked to explain formation of nitrogen-
containing species on the soot surface in experiments with high
(∼10 ppm) NO2 concentrations.11 However, its contribution is
expected to be negligible under low NO2 conditions (∼1 part
per billion) of our experiments.12,16
The fast initial decrease in the NO2 uptake rate on fresh soot
is caused by depletion of surface adsorption sites {S1}. The rate
of subsequent irreversible reaction involving adsorbed NO2 is
limited by the availability of {C-H}red reactive sites, a large
fraction of which are blocked by condensed organic material.
Heating the soot film removes the organics that blocks these
reactive sites, resulting in an increased irreversible uptake rate.
This may explain the apparent partial reactivation of NO2-
processed soot by heating reported previously.10 Heated and
extracted soot films that have been depleted of reactive sites
show no enhancement in the NO2 desorption peak relative to
that for fresh soot (Figure 2). The absence of increased physical
adsorption indicates that organic molecules from incomplete
combustion and NO2 occupy different adsorptive sites on the
soot surface. The same is true for adsorption of Kr atoms
because removal of organic material by heating the soot film
has little or no effect on the internal surface area derived from
BET adsorption isotherms measured using Kr (Figure 3).
Figure 8. ATR-FTIR extinction spectra of type B kerosene flame soot
before and after exposure to vapor of (a) pyrene, (b) sulfuric acid, and
(c) glutaric acid. Spectra of pure coating materials are given for
reference.
According to the redox mechanism outlined above, the
oxidation state of soot surface largely determines the reactivity
of soot toward NO2. Less oxidized (type A) and preheated soot
films react at a faster rate and form larger absolute amounts of
HONO at higher relative HONO yields, which approach 90%
(Table 2). Lower HONO yields on oxidized (type B) soot can
be caused by reaction of HONO with polar organic compounds
adsorbed on soot surface. This is in agreement with previous
studies by Rossi and co-workers who suggested that NO2 is
converted to HONO at a 100% yield on any soot, but some of
the HONO may react on the surface to form NO (eq 12).12,13
Future research is required to identify the nature of the polar
species converting HONO to NO. Possible candidates are
hydroxylated PAH,38,43 such as phenols and alkyl aryl alcohols,
which can be oxidized by HONO to benzoquinonoxims44 and
alkyl aryl carbonyls.45
3.4. NO2 Uptake and HONO Formation on Coated Soot.
Our study provides experimental evidence that condensed
materials have a significant impact on heterogeneous interaction
between NO2 and soot. To further investigate the effect of
coatings, we examined NO2 uptake on soot films that were
pretreated with vapors of pyrene, sulfuric acid, and glutaric acid.
Although no measurable increase in the sample mass was
observed after coating (i.e., less than 2% or 1017 molecules
mg-1), the presence of the coating materials on soot was
confirmed from ATR-FTIR spectra as shown in Figure 8.
Spectra taken at different locations along the tube (top, middle,
and bottom) show that a thicker coating is deposited on the
bottom part that faces the coating material whereas the top and
middle parts are coated uniformly. Therefore, uptake experi-
ments were performed by exposing only the top and middle
parts of the tubes to NO2. Control runs indicate that NO2 does
not react with pure coating materials in the absence of soot under
conditions of our experiments. Figure 9 shows that NO2 uptake
and HONO production vary greatly depending on the nature of
the coating material. Uptake coefficients, total numbers of NO2
molecules reacted and HONO molecules formed, and relative
HONO yields measured for coated soot samples are summarized
in Tables 5 and 6.
Figure 9. Uptake of NO2 and formation of HONO on type B propane
soot coated by different materials: (a) pyrene, soot mass 1.6 mg; (b)
sulfuric acid, soot mass 3.5 mg; (c) glutaric acid, soot mass 1.2 mg.
Experiments with soot coated by pyrene show no significant
change in the uptake rate (Table 5), HONO yield (Table 6), or
amount of NO2 desorbed after exposure (Figure 9b), indicating
that pyrene blocks neither adsorptive nor reactive surface sites.
Pyrene belongs to a class of polycyclic aromatic hydrocarbons

Heterogeneous Reaction of NO2 on Soot
J. Phys. Chem. A, Vol. 114, No. 28, 2010 7523
TABLE 5: Irreversible Uptake Coefficients of NO2 on
Coated Type B Soot and Corresponding HONO Yieldsa
sites accessible to NO2, but also reduces conversion of adsorbed
HONO to NO through reaction 12. Similarly to pyrene and
sulfuric acid, glutaric acid does not change the number of vacant
adsorption sites {S1} because reversible NO2 uptake and internal
BET surface area of soot remain unaffected by coating.
uptake coefficient, γBET × 105
coating
pyrene
propane
kerosene
1.5 ( 0.7
1.4 ( 0.4
4.2 ( 1.4
1.7 ( 0.3
2.0 ( 0.8
3.0 ( 1.3
sulfuric acid
glutaric acid
4. Conclusions
We have used a fast-flow reactor to study heterogeneous
reaction of nitrogen dioxide with hydrocarbon soot produced
under rich (type A) and lean (type B) combustion conditions.
The irreversible uptake coefficient and HONO yield are the
highest for less oxidized type A soot. Heating of the soot
samples in vacuum removes condensed organic materials and
unblocks reactive sites, leading to increased uptake coefficients
and HONO yields. A comparable enhancement is also observed
after extracting the soot samples with iso-propanol, which
indicates that organic molecules blocking reactive sites have
polar nature. The impact of condensed materials is further
investigated using soot films pretreated with vapors of pyrene,
sulfuric acid, or glutaric acid. Coating of pyrene has little effect
on either reaction rate or HONO yield. Sulfuric acid coating
does not alter the uptake coefficient, but significantly reduces
the amount of HONO formed. Coating of glutaric acid
significantly increases NO2 uptake coefficient and HONO yield.
These results imply that the reactivity and HONO generating
capacity of internally mixed soot aerosol will be strongly
dependent on the chemical composition of the coating material
and hence will vary greatly in different polluted environments.
Irreversible uptake coefficients of NO2 on fresh soot calcu-
lated using BET surface areas of samples have the initial values
of (1-5) × 10-5. Hence, from a kinetic prospective (eq 13),
about 1 ppb of HONO can be produced during night for typical
NO2 concentrations and soot surface loadings (σ) in the urban
troposphere, assuming no deposition losses for HONO.9
a Uptake coefficients were calculated using BET surface areas of
soot samples; sample mass is 1-7 mg.
TABLE 6: Number of NO2 Molecules Taken, HONO
Molecules Released, and Corresponding HONO Yields for
Coated Type B Propane Flame Soot Samplesa
NO2 × 10-15
,
HONO × 10-15
,
HONO
yield %
coating
pyrene
sulfuric acid
glutaric acid
molecules mg-1
molecules mg-1
1.9 ( 0.9
0.8 ( 0.9
20.2 ( 0.9
1.3 ( 0.2
0.3 ( 0.2
17.3 ( 0.2
68 ( 18
38 ( 34
86 ( 5
a Sample mass is 1-7 mg.
(PAH) produced from incomplete combustion. Our measure-
ments imply that although PAH may contribute a significant
fraction of the organic soot mass, particularly for soot formed
in fuel-rich flames,46 their effect on the soot reactivity toward
NO2 is negligible.
The presence of sulfuric acid coating on soot has little effect
on the NO2 uptake coefficient, but significantly reduces the
absolute amount of released HONO. Since sulfuric acid is less
reactive toward NO2 than soot (γ < 10-6 has been reported for
surfaces of H2SO4 solutions47), a unchanged uptake coefficient
on the H2SO4-coated surface indicates that sulfuric acid does
not impede the interaction between NO2 and soot. The observed
low production of HONO can be explained by acid-catalyzed
conversion of {HONO} to NO. Our results are consistent with
those reported earlier for commercial carbon samples coated
by H2SO4, where the HONO yield decreased to nearly zero and
the yield of NO increased with increasing sulfuric acid surface
coverage.11 After cycling of the coated soot sample through
elevated relative humidity (i.e., humidification to 100% RH
followed by drying), the absolute amount of HONO formed in
the reaction decreases to zero, but the NO2 uptake coefficient
remains virtually unchanged. Upon humidification, sulfuric acid
absorbs water, increasing the coating volume by more than an
order of magnitude.48 When humidity is decreased, water
evaporates and leaves behind a thin film of sulfuric acid evenly
distributed on the soot surface, making H2SO4 readily available
for interaction with {HONO}. Preheating H2SO4-coated soot
samples to 200 °C partially restores the HONO production, but
does not change the uptake coefficient. Incomplete recovery of
HONO production can be caused by sulfuric acid remaining in
the soot pores or by sulfuric acid-induced oxidation of the soot
surface during heating.
γσω[NO2]
d[HONO]
)
(13)
dt
4
However, continuous exposure to NO2 reduces the reactivity
of soot because of deactivation of the surface sites. Furthermore,
soot particles will be coated with inorganic and organic species
during their atmospheric aging,18 which may enhance or reduce
their reactivity with NO2 and HONO formation.
Acknowledgment. This work was supported by the Houston
Advanced Research Center (H101), National Science Foundation
(AGS-0938352 and CBET-0932705), and the Robert A. Welch
Foundation (Grant A-1417). R. Z. acknowledges additional
support from the National Natural Science Foundation of China
Grant (40728006). We thank Sarah Brooks for the use of an
analytical balance.
Coating the soot films by glutaric acid increases the NO2
uptake rate and the amount of formed HONO by a factor of
2-4 and 20-25, respectively. The magnitude of this enhance-
ment is comparable to or even exceeds that observed for
preheated soot. It is notable that absolute and relative HONO
yields are both increased after coating (Table 6), which indicates
that either more HONO is produced for each consumed NO2
molecule or less HONO is converted to NO on the soot surface.
We hypothesize that glutaric acid forms strong hydrogen-bonded
complexes49 with polar organic molecules from the flame
occupying {C-H}red sites on the soot surface. Binding those
polar molecules to glutaric acid not only makes the reactive
References and Notes
(1) Febo, A.; Perrino, C.; Allegrini, I. Atmos. EnViron. 1996, 30, 3599.
(2) Winer, A. M.; Biermann, H. W. Res. Chem. Intermed. 1994, 20,
423.
(3) Stutz, J.; Alicke, B.; Ackermann, R.; Geyer, A.; Wang, S. H.; White,
A. B.; Williams, E. J.; Spicer, C. W.; Fast, J. D. J. Geophys. Res. 2004,
109, 14.
(4) Lammel, G.; Cape, J. N. Chem. Soc. ReV. 1996, 25, 361.
(5) Aumont, B.; Madronich, S.; Ammann, M.; Kalberer, M.; Balt-
ensperger, U.; Hauglustaine, D.; Brocheton, F. J. Geophys. Res. 1999, 104,
1729.
(6) Penner, J. E.; Eddleman, H.; Novakov, T. Atmos. EnViron., Part A
1993, 27A, 1277.

7524 J. Phys. Chem. A, Vol. 114, No. 28, 2010
Khalizov et al.
(7) IPCC IntergoVernmental Panel on Climate Change. Report. Ed
Solomon, S.; Cambridge Univ Press: Cambridge, UK, 2007; http://
www.ipcc.ch/ipccreports/ar4-wg1.htm.
(8) Lei, W. F.; Zhang, R. Y.; Tie, X. X.; Hess, P. J. Geophys. Res.
2004, 109, D12301.
(9) Ammann, M.; Kalberer, M.; Jost, D. T.; Tobler, L.; Rossler, E.;
Piguet, D.; Gaggeler, H. W.; Baltensperger, U. Nature 1998, 395, 157.
(10) Gerecke, A.; Thielmann, A.; Gutzwiller, L.; Rossi, M. J. Geophys.
Res. Lett. 1998, 25, 2453.
(11) Kleffmann, J.; Becker, K. H.; Lackhoff, M.; Wiesen, P. Phys. Chem.
Chem. Phys. 1999, 24, 5443.
(12) Stadler, D.; Rossi, M. J. Phys. Chem. Chem. Phys. 2000, 2, 5420.
(13) Salgado, M. S.; Rossi, M. J. Int. J. Chem. Kinet. 2002, 34, 620.
(14) Lelievre, S.; Bedjanian, Y.; Laverdet, G.; Le Bras, G. J. Phys. Chem.
A 2004, 108, 10807.
(15) Kleffmann, J.; Wiesen, P. Atmos. Chem. Phys. 2005, 5, 77.
(16) Aubin, D. G.; Abbatt, J. P. D. J. Phys. Chem. A 2007, 111, 6263.
(17) Longfellow, C. A.; Ravishankara, A. R.; Hanson, D. R. J. Geophys.
Res. 1999, 104, 13833.
(30) Xue, H. X.; Khalizov, A. F.; Wang, L.; Zheng, J.; Zhang, R. Y.
Phys. Chem. Chem. Phys. 2009, 11, 7869.
(31) Aubin, D. G.; Abbatt, J. P. J. Phys. Chem. A 2003, 107, 11030.
(32) McCabe, J.; Abbatt, J. P. D. J. Phys. Chem. C 2009, 113, 2120.
(33) Brunauer, S.; Emmett, P. H.; Teller, E. J. Am. Chem. Soc. 1938,
60, 309.
(34) Yanazawa, H.; Ohshika, K.; Matsuzawa, T. AdsorptsJ. Int. Adsorpt.
Soc. 2000, 6, 73.
(35) Zhang, R.; Leu, M. T.; Molina, M. J. Geophys. Res. Lett. 1996,
23, 1669.
(36) Zhang, R. Y.; Leu, M. T.; Keyser, L. F. J. Phys. Chem. 1996, 100,
339.
(37) Slowik, J. G.; Cross, E. S.; Han, J. H.; Kolucki, J.; Davidovits, P.;
Williams, L. R.; Onasch, T. B.; Jayne, J. T.; Kolb, C. E.; Worsnop, D. R.
Aerosol Sci. Technol. 2007, 41, 734.
(38) Goldberg, E. D. Black Carbon in the EnVironment: Properties and
Distribution; J. Wiley: New York, 1985.
(39) Zhang, R.; Suh, I.; Lei, W.; Clinkenbeard, A. D.; North, S. W. J.
Geophys. Res. 2000, 105, 24627.
(18) Zhang, R.; Khalizov, A. F.; Pagels, J.; Zhang, D.; Xue, H.;
McMurry, P. H. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 10291.
(19) Shiraiwa, M.; Kondo, Y.; Moteki, N.; Takegawa, N.; Miyazaki,
Y.; Blake, D. R. Geophys. Res. Lett. 2007, 34, L16803.
(20) Fan, J.; Zhang, R. EnViron. Chem. 2004, 1, 140.
(21) Lei, W. F.; Zhang, R. Y.; McGivern, W. S.; Derecskei-Kovacs,
A.; North, S. W. Chem. Phys. Lett. 2000, 326, 109.
(22) Suh, I.; Lei, W. F.; Zhang, R. Y. J. Phys. Chem. A 2001, 105,
6471.
(23) Saxena, P.; Hildemann, L. M. J. Atmos. Chem. 1996, 24, 57.
(24) Aubin, D. G.; Abbatt, J. P. EnViron. Sci. Technol. 2006, 40, 179.
(25) Levitt, N. P.; Zhang, R. Y.; Xue, H. X.; Chen, J. M. J. Phys. Chem.
A 2007, 111, 4804.
(26) Zhang, D.; Zhang, R. EnViron. Sci. Technol. 2005, 39, 5722.
(27) Khalizov, A. F.; Zhang, R.; Zhang, D.; Xue, H.; Pagels, J.;
McMurry, P. H. J. Geophys. Res. 2009, 114, D05208.
(28) Khalizov, A. F.; Xue, H.; Wang, L.; Zheng, J.; Zhang, R. J. Phys.
Chem. A 2009, 113, 1066.
(29) Xue, H.; Khalizov, A. F.; Wang, L.; Zheng, J.; Zhang, R. EnViron.
Sci. Technol. 2009, 43, 2787.
(40) Molina, M. J.; Molina, L. T.; Zhang, R. Y.; Meads, R. F.; Spencer,
D. D. Geophys. Res. Lett. 1997, 24, 1619.
(41) Zhang, R. Y.; Leu, M. T.; Keyser, L. F. Geophys. Res. Lett. 1995,
22, 1493.
(42) Pryor, W. A.; Lightsey, J. W. Science 1981, 214, 435.
(43) Smith, D. M.; Akhter, M. S.; Jassim, J. A.; Sergides, C. A.; Welch,
W. F.; Chughtai, A. R. Aerosol Sci. Technol. 1989, 10, 311.
(44) Adachi, A.; Okano, T. Chemosphere 2003, 51, 441.
(45) Moodie, R. B.; Richards, S. N.; Thorne, M. P. J. Chem. Soc.-Chem.
Comm. 1987, 870.
(46) Slowik, J. G.; Stainken, K.; Davidovits, P.; Williams, L. R.; Jayne,
J. T.; Kolb, C. E.; Worsnop, D. R.; Rudich, Y.; DeCarlo, P. F.; Jimenez,
J. L. Aerosol Sci. Technol. 2004, 38, 1206.
(47) Kleffmann, J.; Becker, K. H.; Wiesen, P. Atmos. EnViron. 1998,
32, 2721.
(48) Tang, I. N. J. Geophys. Res. 1996, 101, 19245.
(49) Zhao, J.; Khalizov, A.; Zhang, R.; McGraw, R. J. Phys. Chem. A
2009, 113, 680.
JP1021938