Atmospheric Environment 79 (2013) 147e154

Contents lists available at SciVerse ScienceDirect

Atmospheric Environment
journal homepage: www.elsevier.com/locate/atmosenv

Effect of OH radical scavengers on secondary organic aerosol

formation from reactions of isoprene with ozone
Kei Sato a, *, Satoshi Inomata a, Jia-Hua Xing a, Takashi Imamura a, Risa Uchida b,
Sayaka Fukuda b, Kazumichi Nakagawa b, Jun Hirokawa c, Motonori Okumura d,
Susumu Tohno d
a
National Institute for Environmental Studies, Onogawa, Tsukuba, Ibaraki 305-8506, Japan
b
Kobe University, Nada-ku, Kobe, Hyogo 657-8501, Japan
c
Hokkaido University, Kita-ku, Sapporo, Hokkaido 060-0808, Japan
d
Kyoto University, Yoshida-Honmachi, Sakyo-ku, Kyoto-shi, Kyoto 606-8501, Japan

h i g h l i g h t s

Aerosol yields from isoprene ozonolysis were measured.

Cyclohexane, CO, n-hexane, and diethyl ether were used as OH radical scavengers.

Aerosol yield determined from runs with CO was close to or lower than others.

Runs with CO simulated atmospheric conditions with high H2O/RO2 ratios.

a r t i c l e i n f o a b s t r a c t

Article history: In order to understand the effect of OH radical scavengers on secondary organic aerosol formation,
Received 27 February 2013 aerosol yields from the isoprene ozonolysis were measured in the presence of sufficient amounts of OH
Received in revised form radical scavengers. Cyclohexane, CO, n-hexane, and diethyl ether were used as the OH radical scavengers.
3 June 2013
The aerosol yield was determined to be 0.002e0.023 for experiments without OH radical scavengers in
Accepted 15 June 2013
the aerosol mass range 2e120 mg m3. Similar aerosol yields were observed in experiments using
cyclohexane. The aerosol yield observed with n-hexane was close to that observed without scavengers at
Keywords:
120 mg m3, but this aerosol yield was slightly lower than those observed in reactions without scavengers
Environmental chamber
Biogenic volatile organic compound
in the range 3e83 mg m3. The offline aerosol samples obtained in experiments with cyclohexane or n-
Ozoneealkene reaction hexane contained oxygenated hydrocarbons with six or more carbon atoms. Aerosol formation in ex-
Stabilized Criegee intermediate periments that used cyclohexane or n-hexane as the scavenger was enhanced. This was caused by the
Radical chemistry oxidation products of the OH radical scavengers, although the increase in the yield could not be quan-
tified. The aerosol yields were 0.002e0.014 for experiments with CO and diethyl ether in the aerosol
mass range 4e120 mg m3. The reaction of CO with OH radicals forms HO2 radicals, whereas the reactions
of cyclohexane, n-hexane, and diethyl ether, respectively, with OH radicals form organic peroxy (RO2)
radicals. Present results show that the aerosol yield is independent of the HO2/RO2 ratio or that it de-
creases with increasing HO2/RO2 ratio. Since the HO2 concentration is much higher than the RO2 con-
centration in the atmosphere, the results obtained using CO in this study will be a good approximation of
the aerosol yield from the ozonolysis of isoprene in the atmosphere.
Ó 2013 Elsevier Ltd. All rights reserved.

Abbreviations: cHX, cyclohexane; CIs, Criegee intermediates; Et2O, diethyl ether; FT-IR, Fourier-transform infrared spectrometer; LC-TOF, liquid chromatography/time-of-
flight; LC/TOF-MS, liquid chromatography/time-of-flight mass spectrometry; GC/MS, gas chromatography/MS; nHX, n-hexane; SCIs, stabilized Criegee intermediates; SMPS,
scanning mobility particle sizer; SOA, secondary organic aerosol.
* Corresponding author. Tel.: þ81 29 850 2414; fax: þ81 29 850 2579.
E-mail addresses: kei@nies.go.jp (K. Sato), ino@nies.go.jp (S. Inomata), j2xing@jst.go.jp (J.-H. Xing), imamura@nies.go.jp (T. Imamura), uchida.risa@nies.go.jp (R. Uchida), s-
fukuda@chlorine-eng.co.jp (S. Fukuda), nakagawa@kobe-u.ac.jp (K. Nakagawa), hirokawa@ees.hokudai.ac.jp (J. Hirokawa), m.okumura@fy7.ecs.kyoto-u.ac.jp (M. Okumura),
tohno@energy.kyoto-u.ac.jp (S. Tohno).

1352-2310/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.atmosenv.2013.06.036
148 K. Sato et al. / Atmospheric Environment 79 (2013) 147e154

1. Introduction Table 1
Initial experimental conditions and results of secondary organic aerosol (SOA) mass
concentration (M0) and SOA yield (Y) for experiments without scavengers and ex-
Secondary organic aerosol (SOA), produced by the atmospheric periments with cyclohexane and CO.
oxidation of volatile organic compounds or intermediate-
volatility organic compounds, is believed to affect the climate, Run [O3]0 [HC]0 [Scavengers]0 M0 Y Commentb,c
(ppm) (ppm) (ppm) (mg m3)a
human health, and visibility (Hallquist et al., 2009). The atmo-
spheric oxidation of isoprene makes a significant contribution to 12021401 2.483 0.989 None 32.5 0.0144
12021501 3.914 1.937 None 98.0 0.0200 LC/TOF-MS,
SOA formation at the regional and global scales. For example,
GC/MS
Henze and Seinfeld (2006) reported that the inclusion of isoprene 12021702 3.776 1.970 None 115.0 0.0227 LC/TOF-MS,
as an SOA source in a global model increases the global burden of GC/MS
SOA from all sources by more than a factor of two. Liao et al. 12062501 0.726 0.495 None 2.1 0.0029
12062601 1.971 0.815 None 11.6 0.0067
(2007) estimated that SOA formed from isoprene oxidation
12062701 2.952 1.490 None 64.3 0.0183 LC/TOF-MS
contributed 10e20% of PM2.5 mass in the southeastern United 12021301 2.561 1.014 156 (cyclohexane) 28.2 0.0119
States, as much as 38% in the northwest, and about 5e15% in other 12021601 4.745 1.978 312 (cyclohexane) 119.3 0.0221 LC/TOF-MS,
regions. The daytime sinks of isoprene are reactions with OH GC/MS
radicals and O3, whereas the nighttime sinks are reactions with 12022101 0.643 1.951 312 (cyclohexane) 7.1 0.0041
12070201 3.252 1.551 234 (cyclohexane) 86.3 0.0208 LC/TOF-MS
NO3 radicals and O3. The isoprene lifetimes associated with day-
12072502 3.003 1.317 234 (cyclohexane) 53.8 0.0182
time OH reactions ([OH] ¼ 2  106 molecule cm3), daytime O3 12021701 3.719 1.978 12,942 (CO) 19.6 0.0037 LC/TOF-MS,
reactions ([O3] ¼ 30 ppb), and nighttime NO3 reactions GC/MS
([NO3] ¼ 20 ppt) are 0.95, 17, and 1.3 h, respectively (Yokouchi, 12062801 5.145 3.887 26,219 (CO) 119.0 0.0138 LC/TOF-MS
1994; Atkinson and Arey, 2003). SOA yields from the reaction of 12072601 2.918 2.930 19,350 (CO) 62.6 0.0100
12072602 3.900 3.399 22,650 (CO) 75.7 0.0103
isoprene with OH radicals have been reported in a number of 12072701 2.528 1.016 6970 (CO) 3.9 0.0016
papers (Pandis et al., 1991; Edney et al., 2005; Kroll et al., 2005, a
The SOA volume concentration observed by using a scanning mobility particle
2006; Dommen et al., 2006; Kleindienst et al., 2006; Sato, 2008;
sizer (SMPS) was corrected taking into account SOA wall-deposition loss using a
Carlton et al., 2009; Sato et al., 2011). In contrast, information wall-loss rate of 3  105 s1 (Sato, 2008); the SMPS volume concentration was then
on SOA yields from reactions of isoprene with O3 and NO3 radicals converted to the SOA mass concentration using a density of 1.4 g cm3 (Kroll et al.,
is limited. To the best of our knowledge, the aerosol yields for 2005; Dommen et al., 2006).
b
LC/TOF-MS: Liquid chromatography/time-of-flight mass spectrometry.
reactions with O3 and reactions with NO3 radicals have only been c
GC/MS: Gas chromatography/mass spectrometry.
reported by Kleindienst et al. (2007) and Ng et al. (2008),
respectively.
The reaction of isoprene with O3 results in formation of stabi- reactions occur during isoprene ozonolysis in the presence of OH
lized Criegee intermediates (SCIs). The SCI molecules undergo radical scavengers:
dissociation to form OH radicals or undergo other oxidation pro-
cesses. The total yield of SCIs, including CH2OO and higher frag- O3 þ isoprene/OH þ other products; k1 ; fOH (R1)
ments, is 0.26 (Hasson et al., 2001), and the yield of OH radicals is
0.25e0.27 (Atkinson et al., 1992; Paulson et al., 1998). During lab-
oratory alkene ozonolysis experiments, the reaction of alkenes with OH þ isoprene/products; k2 (R2)
secondary OH radicals can be suppressed if a sufficient amount of
OH radical scavengers is injected into the reactor. However, HO2 OH þ scavengers/products; k3 (R3)
and/or RO2 (R is an organic group) radicals produced by the reac-
tion of OH radical scavengers affect the resultant SOA yield. For where k1, k2, and k3 are rate coefficients and fOH represents the
example, Docherty and Ziemann (2003) reported that the SOA yield yield of OH radicals. The initial OH radical scavenger concentration
from the b-pinene ozonolysis decreased with increasing HO2/RO2 was determined so as to satisfy the relationship k3[scavengers]0/
ratio, whereas Keywood et al. (2004) reported that the SOA yield
from the cyclohexene ozonolysis increased with increasing HO2/
Table 2
RO2 ratio.
Initial experimental conditions and results of secondary organic aerosol (SOA) mass
In this study, we investigated the effect of OH radical scavengers concentration (M0) and SOA yield (Y) for experiments with n-hexane and diethyl
on the SOA yield from the ozonolysis of isoprene. The key ether.
semi-volatile products formed from the separate ozonolysis of Run [O3]0 [HC]0 [Scavengers]0 M0 Y Commentb
cyclohexene and b-pinene are both dicarboxylic acids (Docherty (ppm) (ppm) (ppm) (mg m3)a
and Ziemann, 2003; Keywood et al., 2004), whereas those of the
12070501 2.509 1.020 184 (n-hexane) 18.9 0.0075
ozonolysis of isoprene are oligomers produced by condensed-phase 12070601 4.309 2.028 368 (n-hexane) 120.4 0.0218
reactions of oxygenated organic compounds (Nguyen et al., 2010). 12071901 2.673 1.515 276 (n-hexane) 23.2 0.0072
The aims of this study are to determine the SOA yields using various 12071902 2.955 2.028 368 (n-hexane) 83.0 0.0166 LC/TOF-MS
scavenger compounds, and to study the radical chemistry during 12072001 2.586 0.508 92 (n-hexane) 3.0 0.0022
12072501 3.210 1.826 368 (n-hexane) 48.8 0.0115
SOA formation from isoprene ozonolysis. 12072702 4.481 1.967 233 (diethyl ether) 37.3 0.0069
12073001 4.102 2.910 350 (diethyl ether) 56.2 0.0082
2. Materials and methods 12073101 3.785 3.804 466 (diethyl ether) 74.4 0.0095
12080201 5.311 3.797 466 (diethyl ether) 119.6 0.0120 LC/TOF-MS
12080202 4.726 1.007 117 (diethyl ether) 6.6 0.0024
Experiments were conducted using a 6 m3 evacuable chamber
a
(Akimoto et al., 1979). The initial O3 and isoprene concentrations The SOA volume concentration observed by using a scanning mobility particle
were set at 0.5e5 ppm and 0.5e4 ppm, respectively (Tables 1 and sizer (SMPS) was corrected taking into account SOA wall-deposition loss using a
wall-loss rate of 3  105 s1 (Sato, 2008); the SMPS volume concentration was then
2). The ozonolysis of isoprene was studied in the absence of OH converted to the SOA mass concentration using a density of 1.4 g cm3 (Kroll et al.,
radical scavengers or using cyclohexane (cHX), CO, n-hexane (nHX), 2005; Dommen et al., 2006).
b
or diethyl ether (Et2O) as OH radical scavengers. The following LC/TOF-MS: Liquid chromatography/time-of-flight mass spectrometry.
 K. Sato et al. / Atmospheric Environment 79 (2013) 147e154 149

k2[isoprene]0 > 10, for sufficient OH radical scavenging. The buffer

gas was purified air. The relative humidity of the purified air was
less than 0.01%. The total pressure of the mixture gas was about
1 atm. The temperature of the chamber wall was controlled to be
298  1 K. The concentrations of gaseous species were monitored
using a Fourier-transform infrared spectrometer (FT-IR; Nexus 670,
Thermo Fisher Scientific Inc., Waltham, MA, USA) combined with a
multi-reflection mirror system with a 221.5-m optical path. The
concentrations of O3 and isoprene were measured using the ab-
sorptions at 1055 and 893 cm1, respectively. The aerosol size
distribution was monitored using a scanning mobility particle sizer
(SMPS; Model 3934, TSI Inc., Shoreview, MN, USA). SOA particles
were collected in the SMPS through a 60-cm long stainless-steel
tube of outer diameter 6.35-mm at a flow rate of 0.3 L min1. The
SMPS volume concentration was corrected by taking into account
SOA wall deposition loss with a decay rate of 3  105 s1 (Sato,
2008). The SOA volume concentration was then converted to the
mass concentration by assuming a density of 1.4 g cm3 (Kroll et al.,
2005; Dommen et al., 2006). The FT-IR and SMPS data were
recorded every 6 min.
Offline aerosol samples were analyzed using liquid
chromatography/time-of-flight mass spectrometry (LC/TOF-MS;
LC-TOF, Agilent Technologies, Palo Alto, CA, USA). After the isoprene
ozonolysis was completed, an aerosol sample for LC/TOF-MS anal-
ysis was collected on a Teflon membrane filter (Fluoropore, diam-
eter 47 mm, pore size 1 mm, Sumitomo Electric, Osaka, Japan) at a
flow rate of 16.7 L min1 for 30 min in each experiment. An aerosol
sample for gas chromatography/MS (GC/MS; GCMS-QP2010plus,
Shimadzu Corporation, Kyoto, Japan) analysis was then collected
on another filter for 30 min. The aerosol sample filters were stored
at 256 K until filter pretreatment. Blank filters were prepared by
sampling purified air in the chamber through Teflon filters. No
impurities were identified by LC/TOF-MS or GC/MS analysis of the
blank filters.
Each aerosol sample filter for LC/TOF-MS analysis was sonicated
in 5 mL of methanol for 30 min. The extract was placed in a 5-mL Fig. 1. Time series of isoprene, ozone, and aerosol concentrations measured in ex-
periments (a) without OH radical scavengers (run 12021501), (b) with cyclohexane
amber vial and concentrated to near dryness under a gentle (run 12021601), and (c) with CO (run 12021701).
stream of nitrogen gas. The concentrated extract was then dis-
solved in 300 mL of formic acid/methanol/water (v/v/v ¼ 0.05/100/
99.95) solution. A 10-mL aliquot of the sample solution was loop-
injected into the LC/TOF-MS instrument. An octadecyl silica gel scavengers (run 12021501), (b) with cHX (run 12021601), and (c)
column (Model ODS-3, inner diameter 3.0 mm, length 150 mm, with CO (run 12021701). In these three experiments, the initial
particle size 5 mm, GL Science Inc., Tokyo, Japan) was used for ozone and isoprene concentrations were set at about 4 and 2 ppm,
sample separation. The mobile phases were 0.05% formic acid respectively. The maximum SOA concentration measured during
aqueous solution and methanol. The total flow rate of the mobile the experiment with cHX was close to that measured during the
phases was 0.4 mL min1. The temperature of the column was experiment in the absence of OH radical scavengers. In contrast, the
maintained at 298 K. The methanol concentration was initially 5%, maximum SOA concentration measured during the experiment
increased to 90% for 30 min, maintained at 90% for 15 min, and with CO was lower than that measured during the experiment in
maintained at 5% for 15 min. The analytes were ionized using the absence of OH radical scavengers. In the experiment with cHX,
electrospray ionization in negative polarity mode. The ions formed no apparent effect of OH radical scavengers on SOA formation was
from the analytes were analyzed using a TOF-MS with a mass ac- observed.
curacy of <3 ppm; this enabled us to predict the chemical formulas
of analyte ions.
2-Methyltetrols (2-methylthreitol and 2-methylerythritol) pre- 3.2. Kinetic measurements
sent in the SOA particles were quantified by GC/MS analysis, using
the methods reported by Claeys et al. (2004). The detailed pro- In order to check whether or not OH radicals were sufficiently
cedures of present GC/MS analysis are described in the scavenged in the experiment with cHX, the rate coefficient for the
Supplementary material. reaction of isoprene with O3 was determined using the concen-
trations measured for gas species in the present experiments. The
3. Results and discussion rate equation for isoprene is obtained assuming a steady-state
concentration of OH radicals during the reaction sequence of Eqs.
3.1. Time series of reactants and SOA (R1), (R2), and (R3):

Fig. 1 shows the time series of ozone, isoprene, and aerosol

d½isoprene=dt ¼  k1 f1 þ fOH =ð1 þ k3 ½scavenger=
(1)
concentrations measured in experiments (a) without OH radical k2 ½isopreneÞg½O3 ½isoprene
150 K. Sato et al. / Atmospheric Environment 79 (2013) 147e154
Eq. (1) is integrated at the following specific conditions to
provide the relationship between the isoprene and O3 concen-
trations:
Z


ln ½isoprene0 =½isoprene ¼ k1 ð1 þ fOH Þ ½O3 dt ðk3 ½scavenger < < k2 ½isopreneÞ
Z
¼ k1 ½O3 dt ðk3 ½scavenger > > k2 ½isopreneÞ
Fig. 2 shows a plot of the ln([isoprene]0/[isoprene]) data as a
function of the ![O3]dt data for the selected experiments (runs
12021501, 12021601, and 12021701). According to Eq. (2b), the
slopes of the plots obtained from the experiments conducted in the
presence of OH radical scavengers correspond to the rate coefficient
for the O3 þ isoprene reaction.
From linear least-squares regression analysis, the slope
determined in the absence of OH radical scavengers was
(1.47  0.04)  1017 cm3 molecule1 s1. The value determined in
the absence of OH radical scavengers was higher than the literature
value (1.29  1017 cm3 molecule1 s1; Grosjean and Grosjean,
1996) because the decay rate measured for isoprene in the absence
of OH radical scavengers was enhanced as a result of the reaction of
isoprene with secondary OH radicals [Eq. (R2)]. The slopes obtained
in the experiments with cHX and CO were determined to be
(1.22  0.07)  1017 and (1.24  0.04)  1017 cm3 molecule1 s1,
respectively. The rate coefficients obtained in the experiments with
cHX and CO were close to the literature value of the rate coefficient
for the O3 þ isoprene reaction. These results indicate that secondary
OH radicals were sufficiently scavenged in the experiments with cHX
and CO. Similar results were obtained in the experiments with nHX
[k1 ¼ (1.28  0.03)  1017 cm3 molecule1 s1] and Et2O
[k1 ¼ (1.23  0.09)  1017 cm3 molecule1 s1].
3.3. SOA chemical composition
2-Methyltetrols are present in the SOA particles produced by the
OH þ isoprene reaction but are barely present in the SOA particles
produced by the O3 þ isoprene reaction (Kleindienst et al., 2007). If
OH radicals are sufficiently scavenged, the formation of 2-
Fig. 2. Plot of ln{[isoprene]0/[isoprene]} as a function of ![O3]dt for selected experi-
ments (runs 12021501, 12021601, and 12021701); the slope measured in the presence
of sufficient OH radical scavengers represents the rate coefficient for the O3 þ isoprene
reaction; the rate coefficients measured for runs 12021501, 12021601, and 12021701
are 1.462  1017, 1.314  1017, and 1.236  1017 cm3 molecule1 s1, respectively.
methyltetrols in the aerosol is suppressed. In order to confirm
that OH radicals are sufficiently scavenged in the present experi-
ments, 2-methyltetrols present in SOA particles were analyzed
using GC/MS. The ratio of the mass of 2-methyltetorols to the total
(2a)
(2b)
SOA mass was 0.82e1.59% in the absence of OH radical scavengers
(runs 12021501 and 12021702). The ratios of 2-methyltetrol mass to
total SOA mass measured in the experiments with cHX (run
12021601) and CO (run 12021701) were 0.13% and 0.18%, respec-
tively. The ratios of 2-methyltetrol mass to total SOA mass
measured in the presence of OH radical scavengers were lower than
those measured in the absence of OH radical scavengers. These
results also indicate that secondary OH radicals were sufficiently
scavenged in the experiments with cHX and CO.
Fig. 3 shows the LC/TOF-MS total ion chromatograms observed
for the offline samples obtained in the experiment without OH
radical scavengers (run12062701; Fig. 3a), the experiment with cHX
(run 12070201; Fig. 3b), the experiment with CO (run 12062801;
Fig. 3c), and the experiment with nHX (run 12071902; Fig. 3d).
Chromatographic peak 1 was detected in all the samples obtained in
these three runs. Peak 1 was identified as 2-methylglyceric acid,
which is an isoprene oxidation product (Kleindienst et al., 2007). A
chromatographic standard of 2-methylglyceric acid, which was
prepared using a method described by Sato (2008), was used for the
identifications. In contrast, the chromatographic peaks 2e5 were
only detected in the sample from the experiment with cHX. Peaks 2,
3, 4, and 5 were identified as a C6H12O4, C7H12O5, hexanedioic acid
(C6H10O4), and C8H14O5, respectively. A chromatographic standard
was used for the identification of hexanedioic acid. These products
with six or more carbons are attributed to monomeric or oligomeric
products formed by the oxidation of cHX or oligomeric products
resulting from the oxidation of the cHXeisoprene mixture.
The chromatograph observed in the experiment with nHX
(Fig. 3d) contained not only peak 1 but also peaks from C9H16O4
(peak 6), C10H18O4 (peak 7), C10H20O3 (peak 8), and C11H20O4 (peak
9). These oxygenated hydrocarbon products with 9e11 carbon
atoms are attributed to oligomeric products resulting from oxida-
tion of nHX or oxidation of the nHXeisoprene mixture. The chro-
matograph observed in the experiment with Et2O (run 12080201)
was very similar to that observed in the experiment with CO.
The respective aerosol yields from the reaction of cyclohexane
with OH radicals and the reaction of n-hexane with OH radicals are
known to be very low (Docherty and Ziemann, 2003; Kleindienst
et al., 2007). However, in the O3/isoprene/cyclohexane and O3/
isoprene/n-hexane systems, the gaseous, oxygenated organic
products from the reaction of the OH radical scavengers with OH
radicals are absorbed on existing particles and undergo oligomer-
ization with oxygenated organic products from isoprene. Detailed
mechanisms for the former reactions are discussed in the
Supplementary material.
3.4. SOA yield curves
Experiments under various initial conditions were conducted to
obtain SOA yield curves in the aerosol mass region 2e120 mg m3.
Fig. 4 shows plots of the SOA yield (Y) as a function of SOA mass
 K. Sato et al. / Atmospheric Environment 79 (2013) 147e154
Fig. 3. Total ion chromatograms measured in liquid chromatography/time-of-flight
mass spectrometry analysis of offline aerosol samples obtained in the experiment
without OH radical scavengers (run 12062701), the experiment with cyclohexane (run
12070201), the experiment with CO (run 12062801), and the experiment with n-
hexane (run 12071902).
loading. The data for the experiments without OH radical scaven-
gers and the experiments with cHX, CO, nHX, and Et2O are plotted
in Fig. 4. The SOA yield is defined as the ratio of the mass concen-
tration of aerosol to the mass concentration of isoprene reacted.
The uncertainty of the aerosol yield was calculated as 19% from the
uncertainty of the aerosol volume concentration (15%), that of the
aerosol density (10%), and that of the hydrocarbon concentration
(5%). The experimental data were fitted with a two-product gas/
particle absorptive partitioning model function (Odum et al., 1997),
using non-linear least-squares regression analysis. The fitting pa-
rameters, a1, K1 (m3 mg1), a2, and K2 (m3 mg1), for the experiments
without OH radical scavengers were determined to be 0.0220,
0.0414, 0.0395, and 0.000940; those for the experiments with cHX
were 0.0159, 0.0456, 0.0241, and 0.00552; those for the experi-
ments with nHX were 0.00476, 0.218, 0.155, and 0.00103; those for
151
the experiments with CO were 0.00752, 0.0206, 0.0253, and
0.00419; and those for the experiments with Et2O were 0.0641,
0.0697, 0.0422, and 0.00145. The Supplemental material summa-
rizes the results of the four-product volatility basis set model fit-
tings (Lane et al., 2008).
The aerosol yields determined for the experiments with CO
were lower than those determined for experiments without OH
radical scavengers in the aerosol mass region less than 120 mg m3.
If lower aerosol yields are observed as a result of OH radical scav-
enging by CO, the SOA yield from isoprene ozonolysis should be
lower than that from the OH þ isoprene reaction. However, there
could be another reason for the lower aerosol yields observed in the
experiments with CO. The SCIs produced by isoprene ozonolysis
react with CO (Hatakeyama and Akimoto, 1994; Mohammed et al.,
2011). If the reactions of SCI with CO suppress SOA formation, this
reaction may also explain the decrease in SOA formation. SCI
molecules are known to react not only with CO but also with car-
bonyls, alcohols, organic acids, and water (Tobias and Ziemann,
2001). However, to the best of our knowledge, reactions of SCIs
with ethers, esters, and hydrocarbons are very slow (Hatakeyama
and Akimoto, 1994; Tobias and Ziemann, 2001). The SOA yields
observed in the present experiments with a hydrocarbon (nHX)
and ether (Et2O) were both lower than those observed in experi-
ments without OH radical scavengers in the aerosol mass region
less than 120 mg m3. These results indicate that lower SOA yields
are observed as a result of OH radical scavenging rather than SCI
scavenging, that is, these results indicate that the SOA yield from
the O3 þ isoprene reaction is lower than that from the
OH þ isoprene reaction.
HO2 radicals are produced by the reaction of CO with OH radicals
(Baulch et al., 1992):
OH þ CO/H þ CO2 (R4)
H þ O2 þ M/HO2 þ M (R5)
where M represents a third body. RO2 radicals are mainly produced
by the reaction of Et2O with OH radicals (Wallington and Japar,
1991):


OH þ CH3 CH2 OCH2 CH3 /H2 O þ CH3 CH
ÞOCH2 CH3 (R6)



CH3 CH
ÞOCH2 CH3 þ O2 þ M/CH3 CH OO
ÞOCH2 CH3 þ M
(R7)
Similarly, RO2 radicals are mainly produced by the reactions of
cHX and nHX with OH radicals. However, the SOA yields observed
in experiments with cHX or nHX were higher than those from
experiments with Et2O in the aerosol mass region less than
120 mg m3. The aerosol offline samples obtained in the experi-
ments with cHX and nHX contained oxygenated hydrocarbons with
six or more carbon atoms, which were not present in the samples
obtained in the experiments with Et2O. The SOA yields observed in
the experiments with cHX and nHX could be overestimated as a
result of aerosol formation from the reactions of the OH radical
scavengers. However, we cannot quantify how the aerosol yields
are overestimated since products detected by LC/TOF-MS were not
quantified. Therefore, we concluded that the aerosol yield is inde-
pendent of the HO2/RO2 ratio or decreases with increasing HO2/RO2
ratio.
The predominant semi-volatile products present in aerosol from
isoprene are oligomers formed from the condensed-phase re-
actions of low-molecular-weight products, and the predominant
low-molecular-weight products present in this aerosol are organic
acids (Nguyen et al., 2010). The gas-phase radical chemistry of the
152 K. Sato et al. / Atmospheric Environment 79 (2013) 147e154

Fig. 4. Secondary organic aerosol yield curves measured for O3 þ isoprene reaction plotted together with fitted curves of gas/particle absorptive partitioning model; as a result of
the fitting, the parameters of the model, a1, K1 (m3 mg1), a2, and K2 (m3 mg1), for experiments without OH radical scavengers were determined to be 0.0220, 0.0414, 0.0395, and
0.000940; those for experiments with cyclohexane were 0.0159, 0.0456, 0.0241, and 0.00552; those for experiments with n-hexane were 0.00476, 0.218, 0.155, and 0.00103; those
for experiments with CO were 0.00752, 0.0206, 0.0253, and 0.00419; and those for experiments with diethyl ether were 0.0641, 0.0697, 0.0422, and 0.00145.

isoprene ozonolysis will affect the formation of monomer acids,
although it will not affect subsequent oligomer formation in the
condensed phase. Jenkin (2004) suggested that the total yield of
semi-volatile products (which are mainly carboxylic acids) from the
ozonolysis of cyclic alkenes is sensitive to the HO2/RO2 ratio, if both
tertiary RO2 and HO2 are formed from Criegee intermediates (CIs).
Three types of CIs are formed by the isoprene ozonolysis:
O3 þ CH2 ]CHCðCH3 Þ]CH2 /CH2 ]CHCðCH3 ÞOO þ CH2 O
/CH2 ]CHCOCH3 þ CH2 OO
/CH2 O þ CH2 ]CðCH3 ÞCHOO
/CH2 OO þ CH2 ]CðCH3 ÞCHO
The reactions of CH2OO* from Eqs. (R8b) and (R8d) lead to the
formation of HO2:
tertiary RO2 will not be produced during the ozonolysis of isoprene.
According to the hypothesis by Jenkin (2004), the organic acid yield
from the isoprene ozonolysis depends weakly on the HO2/RO2 ratio.
This is consistent with the fact that the aerosol yield from the
isoprene ozonolysis is independent of the HO2/RO2 ratio or de-
creases with increasing HO2/RO2 ratio.
3.5. Comparisons with literature data
(R8a) The yield curves measured using CO, Et2O, and nHX in this study
are compared with the experimental data obtained in a previous
(R8b) study using nHX as the OH radical scavengers (Fig. 5). According to
the previous study by Kleindienst et al. (2007), the SOA yields at
aerosol masses of 13, 15, and 30 mg m3 are 0.0071, 0.0077, and
(R8c)
0.0157, respectively. When the present fitting function from ex-
periments with nHX is used, the SOA yields determined at aerosol
(R8d) masses of 13, 15, and 30 mg m3 are 0.0057, 0.0059, and 0.0087,
respectively. Kleindienst et al. (2007) reported that the uncertainty

CH2 OO /OH þ CHO (R9)

CHO þ O2 /CO þ HO2 (R10)

The reactions of CH2]CHC(CH3)OO* from Eq. (R8a) lead to the
formation of primary RO2:


CH2 ]CHCðCH3 ÞOO /CH2 ]CHC OOH ]CH2 (R11)


CH2 ]CHC OOH ]CH2 /CH2 ]CHCOCH2
þ OH (R12)

CH2 ]CHCOCH2
þ O2 þ M/CH2 ]CHCOCH2 OO
þ M (R13)

The reactions of CH2]C(CH3)CHOO* from Eq. (R8c) barely lead Fig. 5. Aerosol yield curves measured for isoprene ozonolysis in the presence of CO,
to the formation of RO2. CIs from Eq. (R8c) are stabilized, and they diethyl ether, and n-hexane compared with that measured in a previous study using n-
then react with SCI scavengers to form stable products. Therefore, hexane as the OH radical scavengers (Kleindienst et al., 2007).
 K. Sato et al. / Atmospheric Environment 79 (2013) 147e154
of aerosol yield is a factor of two. The yields from the present work
and those of Kleindienst et al. (2007) are consistent. The aerosol
yields determined using CO and Et2O as the OH radical scavengers
are lower than those from Kleindienst et al. (2007).
Carlton et al. (2009) reported that the SOA yield from the re-
action of isoprene with OH radicals was distributed between
0.004 and 0.042 at an SOA mass of 10 mg m3. The previous yield
data for the reaction of isoprene with OH radicals are scattered
because the aerosol yield from the OH þ isoprene reaction de-
pends strongly on the initial NOx concentration. The aerosol yield
increases with decreasing initial NOx concentration. The SOA yield
at an SOA mass of 10 mg m3 was determined to be 0.0023,
0.0032, 0.0048, and 0.0062 using the present fitted curves ob-
tained from the experiments with CO, Et2O, nHX, and cHX,
respectively. The SOA yield from the O3 þ isoprene reaction
(0.0023e0.0062) was lower than the literature values of the SOA
yield from the OH þ isoprene reaction in the absence of NOx
(0.042). Similarly, the SOA yield from the O3 þ isoprene reaction
was lower than the literature value of the SOA yield from the
OH þ isoprene reaction in the absence of NOx in the aerosol mass
region less than 50 mg m3.
3.6. Atmospheric implications
Since the HO2 concentration is much higher than the RO2 con-
centration in the atmosphere, the experiments using CO as the OH
radical scavengers would be better simulations of the ambient at-
mosphere than the experiments using cHX, nHX, and Et2O will be.
Isoprene is mainly oxidized by reaction with OH radicals in the
ambient atmosphere. According to the present results, the SOA yield
from the O3 þ isoprene reaction under high HO2/RO2 conditions is
lower than that from the OH þ isoprene reaction. SOA formed from
isoprene ozonolysis therefore makes a minor contribution
compared with that from SOA produced by the reaction of isoprene
with OH radicals. The relative amount of SOA formed from the re-
action of isoprene with an oxidant X (SOAx, X ¼ OH, O3, or NO3) can
be estimated if we use the background concentrations of OH radicals
(0.1 ppt), O3 (30 ppb), and NO3 radicals (10 ppt), the rate coefficients
of isoprene with OH radicals (1.00  1010 cm3 molecule1 s1;
Atkinson, 1986), O3 (1.29  1017 cm3 molecule1 s1; Grosjean and
Grosjean, 1996), and NO3 radicals (6.77  1013 cm3 molecule1 s1;
Atkinson, 1991), and the SOA yields at a mass loading of 10 mg m3
from the reactions with OH radicals (0.042; low NOx conditions;
Kroll et al., 2006), O3 (0.0023; this work), and NO3 radicals (0.12; Ng
et al., 2008). Using these data, the ratio of SOAOH:SOAO3:SOANO3 is
determined to be 1.0:0.0021:1.9. This estimate clearly shows that
SOA formation from ozonolysis is limited compared with SOA for-
mation from the reaction of isoprene with OH radicals or NO3 rad-
icals. However, accurate SOA yields from isoprene ozonolysis would
be useful for a better understanding of the effect of anthropogenic
activities on global SOA formation. The background ozone level has
increased as a result of anthropogenic activities in the last few de-
cades (Parrish et al., 2009; Tanimoto, 2009). The present results
show that the SOA yield from the O3 þ isoprene reaction is lower
than that from the OH þ isoprene reaction. This suggests that the
amount of SOA formed from the reaction of isoprene decreases with
increasing background ozone level.
The aerosol yield from the isoprene ozonolysis might be affected
by the relative humidity because SCIs from isoprene reacts with
water vapor. To the best of our knowledge, effect of the relative
humidity on the aerosol yield from the isoprene ozonolysis has
never been studied. The yield curves measured for the ozonolysis of
a- and b-pinene have very weak dependences on the relative hu-
midity (Pathak et al., 2007; von Hessberg et al., 2009). For the
discussion presented in this section, we assumed that the aerosol
153
yield from the isoprene ozonolysis is independent of the relative
humidity.
4. Conclusions
In order to understand the effect of OH radical scavengers on
secondary organic aerosol formation, the SOA yield from isoprene
ozonolysis was determined in the absence or presence of OH radical
scavengers in this study. In these experiments, cHX, nHX, CO, and
Et2O were used as the OH radical scavengers. The aerosol yield was
determined to be 0.002e0.023 for experiments without OH radical
scavengers in the aerosol mass range 2e120 mg m3. Similar aerosol
yields were observed in experiments using cyclohexane. The
aerosol yield observed with n-hexane was close to that observed
without scavengers at 120 mg m3 but was slightly lower than those
observed without scavengers in the range 3e83 mg m3. The
aerosol yields were determined to be 0.002e0.014 for experiments
with CO and diethyl ether in the aerosol mass range 4e120 mg m3.
The SOA yield was shown to depend on the OH radical scavenger. In
the experiments with cHX and nHX, SOA would be formed from the
reactions of the scavengers as well as the oxidation of isoprene.
Present results show that the aerosol yield is independent of the
HO2/RO2 ratio or it decreases with increasing HO2/RO2 ratio. Since
the HO2 concentration is much higher than the RO2 concentration
in the atmosphere, the results obtained using CO will be a good
approximation of the aerosol yield from the ozonolysis of isoprene
in the atmosphere.
Acknowledgments
This work was supported by a Grant-in-Aid for Scientific Research
from the Ministry of Education, Culture, Sports, Science and Tech-
nology of Japan (No. 200120003, FY 2008e2012; No. 20120005, FY
2008e2012; and No. 201200007, FY 2008e2012; No. 21221001, FY
2009e2013). K.S. thanks Drs. Yoshikatsu Takazawa, Mai Takagi, and
Kiyoshi Tanabe for technical advice on LC/TOF-MS analysis.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.atmosenv.2013.06.036.
References
Akimoto, H., Hoshino, M., Inoue, G., Sakamaki, F., Washida, N., Okuda, M., 1979.
Design and characterization of the evacuable and bakable photochemical smog
chamber. Environmental Science & Technology 13, 471e475.
Atkinson, R., 1986. Kinetics and mechanisms of the gas-phase reactions of the hy-
droxyl radical with organic compounds under atmospheric conditions. Chem-
ical Reviews 86, 69e201.
Atkinson, R., 1991. Kinetics and mechanisms of the gas-phase reactions of the NO3
radical with organic compounds. Journal of Physical and Chemical Reference
Data 20, 459e507.
Atkinson, R., Arey, J., 2003. Atmospheric degradation of volatile organic compounds.
Chemical Reviews 103, 4605e4638.
Atkinson, R., Aschmann, S.M., Arey, J., Shorees, B., 1992. Formation of OH radicals in
the gas phase reactions of O3 with a series of terpenes. Journal of Geophysical
Research 97, 6065e6073.
Baulch, D.L., Cobos, C.J., Cox, R.A., Esser, C., Frank, P., Just, T., Kerr, J.A., Pilling, M.J.,
Troe, J., Walker, R.W., Warnatz, J., 1992. Evaluated kinetic data for combustion
modeling. Journal of Physical and Chemical Reference Data 21, 411e429.
Carlton, A.G., Wiedinmyer, C., Kroll, J.H., 2009. A review of secondary organic
aerosol (SOA) formation from isoprene. Atmospheric Chemistry and Physics 9,
4987e5005.
Claeys, M., Wang, W., Ion, A.C., Kourtchev, I., Gelencsér, A., Maenhaut, W., 2004.
Formation of secondary organic aerosols from isoprene and its gas-phase
oxidation products through reaction with hydrogen peroxide. Atmospheric
Environment 38, 4093e4098.
Docherty, K.S., Ziemann, P.J., 2003. Effects of stabilized Creiegee intermediate and
OH radical scavengers on aerosol formation from reactions of b-pinene with O3.
Aerosol Science and Technology 37, 877e891.
154 K. Sato et al. / Atmospheric Environment 79 (2013) 147e154
Dommen, J., Metzger, A., Duplissy, J., Kalberer, M., Alfarra, M.R., Gascho, A.,
Weingartner, E., Prevot, A.S.H., Verheggen, B., Baltensperger, U., 2006. Labora-
tory observation of oligomers in the aerosol from isoprene/NOx photooxidation.
Geophysical Research Letters 33, L13805. http://dx.doi.org/10.1029/
2006GL026523.
Edney, E.O., Kleindienst, T.E., Jaoui, M., Lewandowski, M., Offenberg, J.H., Wang, W.,
Claeys, M., 2005. Formation of 2-methyltetrols and 2-methylglyceric acid in
secondary organic aerosol from laboratory irradiated isoprene/NOx/SO2/air
mixtures and their detection in ambient PM2.5 samples collected in the eastern
United States. Atmospheric Environment 39, 5281e5289.
Grosjean, E., Grosjean, D., 1996. Rate constants for the gas-phase reaction of ozone
with 1,1-disubstituted alkenes. International Journal of Chemical Kinetics 28,
911e918.
Hallquist, M., Wenger, J.C., Baltensperger, U., Rudich, Y., Simpson, D., Clayes, M.,
Dommen, J., Donahue, N.M., George, C., Goldstein, A.H., Hamilton, J.F.,
Herrmann, H., Hoffmann, T., Iinuma, Y., Jang, M., Jenkin, M.E., Jimenez, J.L.,
Kiendler-Scharr, A., Maenhaut, W., McFiggans, G., Mentel, Th F., Monod, A.,
Prevot, A.S.H., Seinfeld, J.H., Surratt, J.D., Szmigielski, R., Wildt, J., 2009. The
formation, properties and impact of secondary organic aerosol: current and
emerging issues. Atmospheric Chemistry and Physics 9, 5155e5236.
Hasson, S.H., Ho, A.W., Kuwata, K.T., Paulson, S.E., 2001. Production of stabilized Criegee
intermediates and peroxides in the gas phase ozonolysis of alkenes 2. Asymmetric
and biogenic alkenes. Journal of Geophysical Research 106, 34143e34153.
Hatakeyama, S., Akimoto, H., 1994. Reactions of Criegee intermediates in the gas
phase. Research on Chemical Intermediates 20, 503e524.
Henze, D.K., Seinfeld, J.H., 2006. Global secondary organic aerosol from isoprene
oxidation. Geophysical Research Letters 33, L09812. http://dx.doi.org/10.1029/
2006GL025976.
Jenkin, M.E., 2004. Modelling the formation and composition of secondary organic
aerosol from a- and b-pinene ozonolysis using MCM v3. Atmospheric Chem-
istry and Physics 4, 1741e1757.
Keywood, M.D., Kroll, J.H., Varutbangkul, V., Bahreini, R., Flagan, R.C., Seinfeld, J.H.,
2004. Secondary organic aerosol formation from cyclohexane ozonolysis: effect
of OH scavenger and the role of radical chemistry. Environmental Science &
Technology 38, 3343e3350.
Kleindienst, T.E., Edney, E.O., Lewandowski, M., Offenberg, J.H., Jaoui, M., 2006.
Secondary organic carbon and aerosol yields from the irradiations of isoprene
and a-pinene in the presence of NOx and SO2. Environmental Science & Tech-
nology 40, 3807e3812.
Kleindienst, T.E., Lewandowski, M., Offenberg, J.H., Jaoui, M., Edney, E.O., 2007.
Ozone-isoprene reaction: re-examination of the formation of secondary organic
aerosol. Geophysical Research Letters 34, L01805. http://dx.doi.org/10.1029/
2006GL027485.
Kroll, J.H., Ng, N.L., Murphy, S.M., Flagan, R.C., Seinfeld, J.H., 2005. Secondary organic
aerosol formation from isoprene photooxidation under high-NOx conditions.
Geophysical Research Letters 32, L18808. http://dx.doi.org/10.1029/2005GL023637.
Kroll, J.H., Ng, N.L., Murphy, S.M., Flagan, R.C., Seinfeld, J.H., 2006. Secondary organic
aerosol formation from isoprene photooxidation. Environmental Science &
Technology 40, 1869e1877.
Lane, T.E., Donahue, N.M., Pandis, S.N., 2008. Simulating secondary organic aerosol
formation using the volatility basis-set approach in a chemical transport model.
Atmospheric Environment 42, 7439e7451.
Liao, H., Henze, D.K., Seinfeld, J.H., Wu, S.L., Mickley, L.J., 2007. Biogenic secondary
organic aerosol over the United States: comparison of climatological simula-
tions with observations. Journal of Geophysical Research: Atmospheres 112,
D06201. http://dx.doi.org/10/1029/2006JD007813.
Mohammed, S.A., Camredon, M., Rickard, A.R., Carr, T., Wyche, K.P., Hornsby, K.E.,
Monks, P.S., Bloss, W.J., 2011. Total radical yields from tropospheric ethane
ozonolysis. Physical Chemistry Chemical Physics 13, 11002e11015.
Ng, N.L., Kwan, A.J., Surratt, J.D., Chan, A.W.H., Chhabra, P.S., Sorooshian, A.,
Pye, H.O.T., Crounse, J.D., Wennberg, P.O., Flagan, R.C., Seinfeld, J.H., 2008. Sec-
ondary organic aerosol (SOA) formation from reaction of isoprene with nitrate
radicals (NO3). Atmospheric Chemistry and Physics 8, 4117e4140.
Nguyen, T.B., Bateman, A.P., Bones, D.L., Nizkorodov, S.A., Laskin, J., Laskin, A.,
2010. High-resolution mass spectrometry analysis of secondary organic
aerosol generated by ozonolysis of isoprene. Atmospheric Environment 44,
1032e1042.
Odum, J.R., Jungkamp, T.P.W., Griffin, R.J., Forstner, H.J.L., Flagan, R.C., Seinfeld, J.H.,
1997. Aromatics, reformulated gasoline, and atmospheric organic aerosol for-
mation. Environmental Science & Technology 31, 1890e1897.
Pandis, S.N., Paulson, S.E., Seinfeld, J.H., Flagan, R.C., 1991. Aerosol formation in the
photooxidation of isoprene and b-pinene. Atmospheric Environment 25, 997e
1008.
Parrish, D.D., Millet, D.B., Goldstein, A.H., 2009. Increasing ozone in marine
boundary layer inflow at the west coasts of North America and Europe. At-
mospheric Chemistry and Physics 9, 1303e1323.
Pathak, R.K., Presto, A.A., Lane, T.E., Stanier, C.O., Donahue, N.M., Pandis, S.N., 2007.
Ozonolysis of a-pinene: parameterization of secondary organic aerosol mass
fraction. Atmospheric Chemistry and Physics 7, 3811e3821.
Paulson, S.E., Chung, M., Sen, A.D., Orzechowska, G., 1998. Measurement of OH
radical formation from the reaction of ozone with several biogenic alkenes.
Journal of Geophysical Research 103, 25533e25539.
Sato, K., 2008. Detection of nitrooxypolyols in secondary organic aerosol formed
from the photooxidation of conjugated dienes under high-NOx conditions. At-
mospheric Environment 42, 6851e6861.
Sato, K., Nakao, S., Clark, C.H., Qi, L., Cocker III, D.R., 2011. Secondary organic aerosol
formation from the photooxidation of isoprene, 1,3-butadiene, and 2,3-
dimethyl-1,3-butadiene under high NOx conditions. Atmospheric Chemistry
and Physics 11, 7301e7317.
Tanimoto, H., 2009. Increase in springtime tropospheric ozone at a mountainous
site in Japan for the period 1998e2006. Atmospheric Environment 43, 1358e
1363.
Tobias, H.J., Ziemann, P.J., 2001. Kinetics of the gas-phase reactions of alcohols, al-
dehydes, carboxylic acids, and water with the C13 stabilized Criegee interme-
diate formed from ozonolysis of 1-tetradecene. Journal of Physical Chemistry A
105, 6129e6135.
von Hessberg, C., von Hessberg, P., Pöschl, U., Bilde, M., Nielsen, O.J., Moortgat, G.K.,
2009. Temperature and humidity dependence of secondary organic aerosol
yield from the ozonolysis of b-pinene. Atmospheric Chemistry and Physics 9,
3583e3599.
Wallington, T.J., Japar, S.M., 1991. Atmospheric chemistry of diethyl ether and ethyl
tert-butyl ether. Environmental Science & Technology 25, 410e415.
Yokouchi, Y., 1994. Seasonal and diurnal variation of isoprene and its reaction
products in a semi-rural area. Atmospheric Environment 28, 2651e2658.