Effect of experimental conditions on secondary organic aerosol formation in an oxidation flow reactor

Abstract

As global air pollution aggravates, it urges us to investigate the evolution of atmospheric related emissions. Recently oxidation flow reactor (OFR) has been widely used to simulate the aging of atmospheric related emissions as its excellent performance. In this work, we systematically characterized the SOA formation from α-pinene using a custom-built OFR under different conditions. The particle loss of the OFR was evaluated by particle transmission efficiency, then the effect of O3 concentration, relative humidity (RH), precursor amounts, OH exposure level, and acidic seed aerosol on SOA formation was investigated. The particle losses of our OFR for particles above 50 nm were very small, which was comparable to or even better than those of previous traditional flow reactors. The formation of SOA particles nearly achieved stability after UV radiation for 15 min. When OH exposure concentration at approximately 0.6 × 10¹² molec cm⁻³ s, the SOA yield reached the maximum yield of 0.51 and 0.39 in the presence and absence of acidic seed aerosol respectively. The addition of acidic seed aerosol increased both particle number concentration and particle size, resulting in an increase of SOA yield of 1.2–1.5 times at OH exposure concentration ranged from 0.1 × 10¹² to 1.8 × 10¹² molec cm⁻³ s. The results provided significant guidance for studying the aging of atmospheric related emission under different working conditions by using an OFR.

Full text

Atmospheric Pollution Research 12 (2021) 392–400
Available online 15 January 2021
1309-1042/© 2021 Turkish National Committee for Air Pollution Research and Control. Production and hosting by Elsevier B.V. All rights reserved.
Effect of experimental conditions on secondary organic aerosol formation in
an oxidation ow reactor
Ranran Zhao
a
, Qixing Zhang
a
,
*
, Xuezhe Xu
b
,
**
, Weixiong Zhao
b
, Hui Yu
b
,
c
, Wenjia Wang
a
,
Yongming Zhang
a
, Weijun Zhang
b
,
c
a
State Key Laboratory of Fire Science, University of Science and Technology of China, Hefei, 230026, Anhui, China
b
Laboratory of Atmospheric Physico-Chemistry, Anhui Institute of Optics and Fine Mechanics, HFIPS, Chinese Academy of Sciences, Hefei, 230031, Anhui, China
c
University of Science and Technology of China, Hefei, 230026, Anhui, China
ARTICLE INFO
Keywords:
OFR
SOA formation
O
3
OH
Seed aerosol
ABSTRACT
As global air pollution aggravates, it urges us to investigate the evolution of atmospheric related emissions.
Recently oxidation ow reactor (OFR) has been widely used to simulate the aging of atmospheric related
emissions as its excellent performance. In this work, we systematically characterized the SOA formation from
α
-pinene using a custom-built OFR under different conditions. The particle loss of the OFR was evaluated by
particle transmission efciency, then the effect of O
3
concentration, relative humidity (RH), precursor amounts,
OH exposure level, and acidic seed aerosol on SOA formation was investigated. The particle losses of our OFR for
particles above 50 nm were very small, which was comparable to or even better than those of previous traditional
ow reactors. The formation of SOA particles nearly achieved stability after UV radiation for 15 min. When OH
exposure concentration at approximately 0.6 ×10
12
molec cm
−3
s, the SOA yield reached the maximum yield of
0.51 and 0.39 in the presence and absence of acidic seed aerosol respectively. The addition of acidic seed aerosol
increased both particle number concentration and particle size, resulting in an increase of SOA yield of 1.2–1.5
times at OH exposure concentration ranged from 0.1 ×10
12
to 1.8 ×10
12
molec cm
−3
s. The results provided
signicant guidance for studying the aging of atmospheric related emission under different working conditions
by using an OFR.
Credit author statement
Ranran Zhao: Conceptualization, Methodology, Software, Valida-
tion, Investigation, Data curation, Writing – original draft, Writing –
review & editing. Qixing Zhang: Conceptualization, Methodology,
Investigation, Resources, Data curation, Writing – review & editing,
Supervision, Funding acquisition, Project administration. Xuezhe Xu:
Methodology, Software, Validation, Resources, Data curation, Writing –
original draft, Writing – review & editing, Data curation, Supervision,
Funding acquisition. Weixiong Zhao: Validation, Resources, Supervi-
sion, Funding acquisition. Hui Yu: Methodology, Software, Validation,
Data curation. Wenjia Wang: Writing – original draft, Software, Vali-
dation, Writing – review & editing. Yongming Zhang: Resources, Su-
pervision, Funding acquisition, Project administration. Weijun Zhang:
Resources, Supervision, Funding acquisition, Project administration.
1. Introduction
Secondary organic aerosol (SOA), which accounts for an important
fraction of global atmospheric polluted particles, plays a signicant role
in regional air quality and global climate change (Hallquist et al., 2009;
Jimenez et al., 2009; Wu et al., 2018; Zhao et al., 2017). However, the
understanding of SOA formation is still limited due to the complex
exposure conditions (e.g. oxidant types and the presence of inorganic
aerosol) in the atmosphere (Ziemann and Atkinson, 2012).
To investigate the SOA formation, environmental chamber (EC)
(Eddingsaas et al., 2012; Wang et al., 2014, 2020; Zhang et al., 2020;
Zhao et al., 2015) and oxidation ow reactor (OFR) (Kang et al., 2007;
Lambe et al., 2011; Li et al., 2019) have been designed. Traditional large
EC reactor provides an oxidizing environment similar to that of the at-
mospheric condition, allowing simulating atmospheric oxidation
Peer review under responsibility of Turkish National Committee for Air Pollution Research and Control.
* Corresponding author.
** Corresponding author.
E-mail addresses: qixing@ustc.edu.cn (Q. Zhang), xzxu@aiofm.ac.cn (X. Xu).
Contents lists available at ScienceDirect
Atmospheric Pollution Research
journal homepage: www.elsevier.com/locate/apr
https://doi.org/10.1016/j.apr.2021.01.011
Received 26 September 2020; Received in revised form 12 January 2021; Accepted 13 January 2021

Atmospheric Pollution Research 12 (2021) 392–400
393
processes that range from few hours to days (Eddingsaas et al., 2012;
Zhao et al., 2015). However, even for the most state-of-the-art EC
reactor, the gas and particle wall loss rates are more than 1.3 ×10
−4
min
−1
and 2.8 ×10
−3
min
−1
respectively (Wang et al., 2014). The sig-
nicant chamber wall losses can underestimate the aerosol yields up to 2
times (La et al., 2016). Besides, because of the limited oxidant exposures,
these chambers are not available to simulate the formation of substantial
highly aged organic aerosol that characterizes atmospheric SOA (Ng
et al., 2010). Previous studies have shown that the chemical composi-
tions of SOA were consistent in OFR and EC reactors at similar OH
exposure levels, indicating that the SOA formation is mainly governed
by gaseous homogeneous oxidation in OFR reactor and that agrees well
with those of EC chamber (Bruns et al., 2015; Lambe et al., 2015). OFR
has been regarded as an effective alternative tool to EC in aging organic
aerosol recently (Li et al., 2015). The prototype of OFR was rstly
introduced by Kang et al. (2007), who designated their ow reactor as
Potential Aerosol Mass (PAM) reactor. Such an OFR/PAM reactor has
minor wall losses, and it can provide a highly oxidizing environment,
allowing simulating atmospheric aging processes on time scale ranging
from a few days to weeks in a relatively short residence time (seconds to
minutes) (Lambe et al., 2011; Palm et al., 2016). Due to its simple
structure and strong oxidation capacity, the OFR/PAM reactor has been
rapidly developed ever since it came out (Feng et al., 2019; Lambe et al.,
2011; Li et al., 2019). Extensive OFR oxidation experiments were sub-
sequently conducted to simulate the aging of atmospheric aerosol.
However, for OFR oxidation experiments, it is difcult to describe the
characteristics of specic oxidation products and quantify the SOA
levels in OFR without accurately controlling the experimental factors
that affect SOA formation (Peng et al., 2019).
Previous studies have performed many OFR/PAM oxidation experi-
ments under different experimental conditions to investigate the for-
mation of SOA over the past decade. Kang et al. (2007) studied the
production mechanism and yields of SOA under several experimental
conditions using an OFR/PAM reactor, but they neglected the particle
wall losses. Lambe et al. (2011) indicated that the OFR/PAM reactor
design and operation conditions signicantly affect SOA yields, in which
wall losses have the strongest effect on the estimated SOA yields. Lambe
et al. (2015) corrected the SOA yields from OFR/PAM reactor using
particle losses. They compared the SOA yields from OFR/PAM reactor to
EC under several experimental factors, whereas comparisons with other
OFR/PAM studies were lacking. While at the same OH exposure levels,
signicantly different SOA yields were observed from recent ow
reactor studies. For example, when the OH exposure concentration
higher than 0.4 ×10
12
molec cm
−3
s, some studies observed a smaller
SOA yield (Chen et al., 2013; Lambe et al., 2015; Sbai and Farida, 2019;
Simonen et al., 2017), while others observed a larger yield (Li et al.,
2019). Therefore, these discrepancies between ow reactor results are
likely to make the application of OFR/PAM reactor to age aerosol
difcult. Besides, Inorganic acidic seed aerosol usually plays an impor-
tant role in the formation of SOA (Ge et al., 2017). The inuence of
acidic seed aerosol on SOA formation has been extensively studied in EC
(Northcross and Jang, 2007), but the study in OFR is limited. In this
work, we developed an OFR based on the design of previous OFR/PAM
reactors. We mainly showed the performance characteristics of our OFR
and used it to study SOA formation from
α
-pinene under a series of
experimental conditions. We studied the particle wall loss of our ow
reactor and investigated the effect of O
3
concentration, relative hu-
midity (RH), precursor amounts, OH exposure level, and acidic seed
particles on
α
-pinene SOA formation. Comparisons of SOA yields to
previous OFR/PAM and EC results were also shown.
2. Material and methods
2.1. Description of OFR
Fig. 1 shows the schematic diagram of OFR experimental setup. The
OFR was developed based on previous OFR/PAM reactors (Chu et al.,
2016; Kang et al., 2007; Lambe et al., 2011). The reactor consists of a
mixing tube of 12 cm length and 6 cm diameter, a double stainless steel
cylinder of 49 cm length and 22 cm inner diameter and 25 cm outer
diameter, and four Teon-coated tubes of 55 cm length and 2.8 cm
diameter and 0.15 cm thickness. Inner wall surface of the PAM reactor
was coated with Teon FEP lm (0.5 mm thickness). This Teon coating
has excellent chemical stability and anti-aging property, which is often
used as the anti-stick coating to reduce wall reactivity (Chu et al., 2016).
The mixing tube was used to mix the inlet sample gases at the front of the
reactor. A center sampling port was used to minimize the inuence of
turbulent at the end of the reactor (Li et al., 2019). The volume from the
inlet mixing tube to the sampling port is 17.8 L. The total ow rate for
the OFR experiments is 6 ±0.1 L min
−1
, resulting in a residence time of
~178 s. Four UV lamps (emission spectrum peaks are 254 nm, the
output power is 18 W) located in the Teon-coated quartz tubes were
mounted on the wall inside the reactor. The UV lamps can be controlled
independently. A large ow of zero air swept away the heat and O
3
produced by lamps. To better control the operating temperature of OFR,
circulating water continually owed in the jacket of the double stainless
steel cylinder. The operating temperature of the reactor was controlled
at 21 ±1 ◦C. Reaction gases were divided into four paths separately
controlled by four mass ow controllers (MFC), including dilution air,
O
3
, gas precursor, and water vapor. One of the paths was humidied
with a Naon Membrane humidier to provide water vapor
(FC125-240-5 MP, Perma pure LIC, USA). RH of the reaction gases was
controlled by adjusting the ow ratio through the humidier. The RH
deviation was less than 2% at a given RH. A temperature and humidity
sensor (T & RH) was installed near the reactor outlet for continuous
measurement of the temperature and RH inside the ow reactor.
Before the experiment, all mass ow controllers were calibrated with
a primary ow calibrator system (combined with Sensidyne Gilibrator™
2 and Bubble Generator ranged 2–30 lpm, Gilian, USA). After each
experiment, the OFR was cleaned by the large ow of zero air purging
and UV irradiation. Zero air used in the experiments was generated with
a zero air generator (model 737 series, Aadco Instruments Inc., USA).
The interior of OFR is not considered clean until the concentration of
volatile organic compounds detected by the Gas Chromatography-Flame
Ionization Detector (GC-FID, model 7920 A, Agilent, USA) is zero and
the particle number concentration detected by the Scanning Mobility
Particle Sizer (SMPS) is less than 10 particles cm
−3
.
Fig. 1. Schematic diagram of the OFR experimental setup.
R. Zhao et al.

Atmospheric Pollution Research 12 (2021) 392–400
394
2.2. Generation of OH radicals
The OH radicals were generated via OFR254 mode, which was the
most universally applicable method (Peng et al., 2015, 2016). At 254 nm
irradiation, O
3
photolysis produced the excited oxygen atoms O (
1
D),
and then these O (
1
D) reacted with water vapor to produce OH radicals
(Atkinson and Arey, 2003; Li et al., 2015). The O
3
was generated from an
external O
3
generator (model COM-AD-01-OEM, Anshan Anseros Envi-
ronmental Protection CO. Lid, China) by plasma-discharge. As there is
no metal in the O3 generator electrodes, the concentration of NOx
produced is negligible. The O
3
concentration was continually measured
with the O
3
monitor (model 49i O
3
analyzer, Thermo Scientic Inc.,
USA) downstream of the ow reactor. The OH exposure concentration
was determined by the decay of SO
2
before and after UV lamps on
(Lambe et al., 2011). The details can be found in the supplement
(Figure S1). SO
2
concentration was measured with an SO
2
monitor
(model 43i-TLE SO
2
analyzer trace level enhanced, Thermo Scientic
Inc., USA) downstream of the ow reactor. The OH exposure concen-
tration was adjusted by changing the RH of reaction gases, and the
amounts of UV lamp (assumed the UV lamp has the same intensity), and
the O
3
concentration.
In the OFR experiments, O
3
concentration ranged from 1.2 ppm to
18.5 ppm. The experiments included two conditions: dry and wet, where
the RH deviation was less than 2% at a given RH. The OH exposure
concentration ranged from 0.1 ×10
12
to 1.8 ×10
12
molec cm
−3
s. Ac-
cording to the average atmospheric OH concentration of 1.5 ×10
6
molec cm
−3
, it equals to the photochemical age of 0.7–14 days (Mao
et al., 2009).
2.3. Injection of organic precursor
The schematic diagram of the injection of organic precursors is
presented in Fig. 2S.
α
-pinene is one of the representative biological
organic precursors, and it is often used to simulate SOA formation (Kang
et al., 2007; Lambe et al., 2011; Li et al., 2019). A micro-syringe pump
(model WK-101P, Nanjing Wanke Precision Electromechanical Co. LID,
China) continually injected
α
-pinene (Aladdin, ≥99%) into a
custom-made glass three-ended round bottom ask. The ask was
wrapped with aluminum foil and was heated by a heating reactor at
40 ◦C. The organic liquid vaporized at the tip of a syringe and then the
organic vapor was introduced into the ow reactor by puried air. The
gas precursor concentration was continually measured with the GC-FID
downstream of the ow reactor. When the concentration of sample gases
was stable, switching on O
3
generator and UV lamps, as well as the water
and gas cooling. It should be noted that the precursor concentration by
injection was signicantly inuenced by multiple factors, including sy-
ringe size, the heating temperature of the ask, ow rate of carrier gas,
location of syringe tip, and operating prociency. These factors were
considered comprehensively when using a micro-syringe pump in these
experiments. In the OFR experiments, the concentration of
α
-pinene
ranged from 28 to 170 ppb, within the average absolute uncertainty of
~8%.
2.4. Particles monitoring and analysis
Although the wall loss of the ow reactor is much lower than that of
the EC reactor, it is still a key factor affecting SOA formation in the OFR
oxidation experiments. The wall loss of gas-phase species is very small
and that is generally negligible, whereas the particle loss is large and
that needs to be accounted for in aerosol production (Lambe et al.,
2011). To decrease the wall effects, this ow reactor was designed with a
larger radial/axial dimension ratio (2.25) and a smaller
surface-to-volume ratio (SA/V) (0.22 cm
−1
) compared to previous
OFR/PAM reactors (Huang et al., 2017; Lambe et al., 2011).
Figure S3 shows the schematic diagram of an experimental setup for
the determination of particle transmission efciency. In the particle loss
experiments, a constant output atomizer (model TSI 3076) atomized the
aqueous solution of ammonium sulfate (AS) into AS particles. Mono-
disperse AS particles were size-selected by a Differential Mobility
Analyzer (DMA, model TSI 3080) with the particle electric mobility
diameter (Dp) ranging from 50 to 200 nm. The particle transmission
efciency for the ow reactor was estimated by measuring particle
Fig. 2. Particle transmission efciency for our OFR
and previous OFR/PAM reactors. The particles in the
above picture included inorganic (AS and silver)
particles, organic particles (BES (bis (2-ethylhexyl)
sebacate) and DOS (dioctyl sebacate)) and vehicle
exhaust particles. The “PAM steel” referred to the
ow reactor that was made of stainless steel, and the
“PAM glass” referred to the ow reactor that was
made of quartz glass. Error bars indicated the stan-
dard deviation of at least three replicate experiments.
R. Zhao et al.

Atmospheric Pollution Research 12 (2021) 392–400
395
number concentration before and after the reactor. Total particle num-
ber concentrations from two CPC (model TSI 3775 and 3776) in-
struments before and after the ow reactor agree within ±6% when
sampling at the same ow. As a result, the aerosol mass was corrected
from the AS particle transmission efciency, with a correction of 24 ±
4%.
The particle number size distribution between 14.6 and 661.2 nm
was measured with SMPS (consisting of DMA and CPC). An activated
charcoal denuder was used to remove gas-phase species before the
sample ow entering into the SMPS. The aerosol mass concentration was
estimated by SMPS and assuming that the particles are spherical and
their density is 1.2 g cm
−3
(Keller and Burtscher, 2012; Zhang et al.,
2015). The SOA yield (Y) was calculated from the mass concentration of
aerosol (ΔM,
μ
g/m
3
) and reacted gaseous parent hydrocarbons (ΔHC,
μ
g/m
3
), where Y =ΔM/ΔHC (Seinfeld et al., 2001).
In the seeded experiments, a range of SO
2
from 21 ppb to 34 ppb was
introduced to produce sulfuric acid seed aerosol. The seeded SOA mass
concentration was referred to the total particle volume concentration
(after subtracting the volume concentration of sulfuric acid particles) by
multiplying the assumed density.
3. Results and discussion
3.1. Particle losses
Fig. 2 shows the particle transmission efciency for our OFR and
other OFR/PAM reactors. In this work, as the aerosol number concen-
tration increases from 5 ×10
2
to 5 ×10
3
particles cm
−3
, the loss rate for
the particles increases slightly (with deviations of 1%–6%). This is likely
due to the collision among particles and the collision of particles to the
wall. Overall, the average transmission efciency of AS particles is
~76% at 50 nm and ~88% at 100 nm. Furthermore, when the particle
size at 150 nm and larger, the average transmission efciency increases
to ~100%. The wall losses of our OFR mainly occur in small particles
(Dp <50 nm) with approximately a quarter of small particles, while the
losses for large particles (Dp >100 nm) are very small. This result was
similar to that of PAM steel in Karjalainen et al. (2016). In their work,
primary particles from vehicle exhaust were used to determine the
particle losses. This similarity in particle transmission efciency was
likely due to that the two reactors used the identical material tubing and
both were designed with a larger radial/axial size ratio (2.25 and 2.09
for this work and their work respectively) and a smaller SA/V (0.22 and
0.23 cm
−1
for this work and their work respectively). However, the
particle losses of our OFR disagreed with those of some previous ow
reactors (Fig. 2). For example, the particle transmission efciency of
TSAR (TUT Secondary Aerosol Reactor) glass, PEAR (Photochemical
Emission Aging Flow Tube Reactor) steel, and ECCC-OFR (Climate
Change Canada OFR) glass is higher than that of our OFR reactor (Iha-
lainen et al., 2019; Li et al., 2019; Simonen et al., 2017). This is because
TSAR is a small-volume OFR (3.3 L) with a shorter residence time (~40
s), but it is likely to limit the condensation of oxidized compounds
(Simonen et al., 2017). For PEAR (Ihalainen et al., 2019) and ECCC-OFR
(Li et al., 2019), they used a conical inlet used to minimize the estab-
lishment of jetting and recirculation, and side ows used to reduce the
losses of particle wall losses. However, PEAR is a high-volume OFR (139
L) than many previous ow reactors with similar residence times, which
is difcult to simulate the aging of mobile emission sources. Also, there
are some OFR studies whose particle losses larger than those of our ow
reactor, such as PAM glass (Lambe et al., 2011) and CPOT glass (Caltech
Photoxidation Flow Tube) (Huang et al., 2017). In their ow reactors,
the particle wall losses were more than 40% for particles below 50 nm.
The larger particle losses may attribute to the non-centerline sampling,
tubing materials, or particle types (Huang et al., 2017; Lambe et al.,
2011). Although the particle losses of our OFR are slightly larger than
those of some specially designed ow reactors and that may contribute
to underestimating the SOA yield, this deviation of wall loss is thought to
be acceptable as long as appropriate wall loss corrections were made.
The loss corrections were of 24 ±4% for SOA yield in our work.
Wall loss is an important parameter conguration for OFR to form
SOA. Overall, the particle transmission efciency of our OFR is com-
parable to or even better than that of the traditional OFR/PAM reactors.
3.2. SOA formation dominated by O
3
oxidation
As an atmospheric pollutant, O
3
reacts with organic precursors to
form a large number of aerosol particles, which severely exacerbate air
pollution (Mozaffar et al., 2020).
α
-pinene, as one of the representative
gaseous air pollutants, reacts rapidly with O
3
to form SOA. Although
studies on the SOA formation by ozonolysis
α
-pinene have been exten-
sively reported in previous literature, detailed characterizations on the
ozonolysis of
α
-pinene by using OFR are still limited. Figure S4 shows
that the
α
-pinene SOA yield is dependent on O
3
concentration at dark
dried conditions. SOA yield initially increases as the increase of O
3
concentration and then nearly reaches a stable level when O
3
concen-
tration larger than ~10 ppm. The results were compared to those of
Kang et al. (2007), in which they conducted the O
3
oxidation experi-
ments with the same precursor amounts at dark dried conditions.
Although similar oxidation trends were observed, our yields were
slightly higher than their results. This is likely due to the material tubing
of the OFR/PAM reactor used in their work was Teon FEP lm, while
stainless steel was used in our work that enables to avoid the losses of
charged particles. Electrostatic charge from the particles can be released
through the stainless steel material at the inlet and outlet of the reactor.
With the comparable oxidation condition, the design of the OFR has a
signicant inuence on the estimated SOA yields. The higher SOA yield
demonstrated that our OFR has better design characteristics and can
produce more SOA.
Besides the effect of O
3
concentration on SOA yields, the effect of
precursor amounts and SOA mass on SOA yields was also investigated.
Although many previous studies have conducted OFR experiments to
investigate the dependence between SOA yield and precursor concen-
tration, the range of precursor concentration for the SOA formation used
in most studies is relatively limited. We studied the SOA yield over a
wide range of precursor concentrations (45–170 ppb), which can pro-
vide experimental data support for the modeling system. To facilitate
comparison with the results in previous studies, we carried out OFR
oxidation experiments under similar experimental conditions, where O
3
was constant at 5 ppm. As the concentration of gas precursors increased
from 45 to 170 ppb, SOA yield increased from 0.04 to 0.29 (Fig. 3a), SOA
mass increased from 6.9 to 275.9
μ
g/m
3
(Fig. 3b). Meanwhile, the SOA
yield showed a signicant dependence on SOA mass (Fig. 3c). The SOA
mass and yield were similar to the results from previous studies (Chen
and Torres, 2009). In their studies, the SOA mass increased from 8 to
249
μ
g/m
3
and the SOA yield increased from 0.1 to 0.46 were observed
for
α
-pinene ozonolysis at O
3
concentration of 4.5 ppm. These results can
be explained by gas-particle partitioning. In the gas-phase O
3
oxidation,
the increase of precursor concentration increases the condensation of
oxidation products. The presence of more condensable products can
facilitate the partitioning of semi-volatile products into the aerosol
phase, increasing particles in the aerosol phase and further an increase
in SOA yield (Loza et al., 2014; Odum et al., 1996; Sbai and Farida,
2019). In addition, as the mass of aerosol particles signicantly depends
on their particle number concentration and size distribution (Seinfeld
and Pankow, 2003), these results can also be explained by the particle
number size distribution of aerosol (Fig. 4). As the concentration of gas
precursor increased from 45 to 170 ppb, the total particle number
concentration gradually increased from 0.2 ×10
6
to 7.5 ×10
6
particles
cm
−3
(Figure S5a) and the particle medium diameter increased from 68
to 122 nm (Figure S5b). The results demonstrated that the increase of
precursor amounts promoted the condensation and growth of particles
and further contributed to an increase in SOA mass and yield.
Fig. 3 also shows the effect of RH on
α
-pinene SOA mass and yield.
R. Zhao et al.

Atmospheric Pollution Research 12 (2021) 392–400
396
Although no noticeable changes in SOA mass and yield were observed as
the RH increased to 36%, the particle number size distribution showed
some differences among different RH (Fig. 4). For the ozonolysis of the
same precursor amounts, the number concentration of aerosol showed a
slight decrease and the particle medium diameter showed a slight in-
crease as the RH increased (Figure S5). The differences were gradually
evident when the SOA mass is signicant that consuming precursor
amounts more than 90 ppb. This phenomenon may be related to the
water uptake of aerosol particles, resulting in the coagulation of small
particles, and further an increase in particle diameter and a decrease in
particle number concentration. Despite these differences, the RH did not
cause signicant changes in SOA mass and yield (Fig. 3), with the largest
changes of 30%. The results agreed well with those of Kang et al. (2007),
where they indicated that RH has little effect on SOA yield as the RH
increased to 60%. Making this comprehensive consideration, when the
RH increased to 36%, although there is no signicant change in SOA
mass and yield, there are signicant differences in the particle number
size distribution of SOA under different RH. The low RH may have little
effect on the particle number size distribution when the precursor con-
centration and SOA mass are low. However, when the precursor con-
centration is high enough (above 90 ppb), the trend of decreasing
particle number concentration and increasing the particle size of SOA is
signicant.
3.3. SOA formation dominated by OH oxidation
In the atmosphere, the oxidation initiated by OH radicals is a more
important oxidation pathway compared to O
3
oxidation (Kang et al.,
2007, 2011). The
α
-pinene OH photo-oxidation is as well investigated as
the
α
-pinene ozonolysis is. Figure S6 shows that SOA yield depended on
RH when UV irradiation, where SOA yield (from 0.16 to 0.27) increased
with RH (from 7% to 48%). Similar results were also observed in Kang
et al. (2007). These results were due to the formation of OH that resulted
from reaction with O
3
by the increased water vapor. Table 1 shows that
the SOA mass (from 19.1 to 311.9
μ
g/m
3
) and yield (from 0.12 to 0.56)
increased with the increasing of precursor amounts (from 28 to 100 ppb)
at comparable OH concentration conditions (1.2 ×10
12
–1.4 ×10
12
molec cm
−3
s). The results showed that both SOA mass and yield were
signicantly dependent on precursor amounts. Similar trends were also
observed in previous studies (Chen and Torres, 2009; Kang et al., 2011).
Fig. 5 shows the dependence of SOA yield on aerosol mass concen-
tration at high OH exposure levels. The initial SOA yield increased
rapidly and then slowly with the increase of SOA mass. This phenome-
non was attributed to the gas-particle partitioning, where the increase of
reaction precursor concentration increases the condensed phase prod-
ucts and further facilitates the partitioning of gas-particle compounds to
particle phase, then gradually tending to the gas-particle partitioning
equilibrium (Donahue et al., 2006; Odum et al., 1996). The trend was
similar to that of previous OFR/PAM studies (Fig. 5) (Ahlberg et al.,
2017; Chen and Torres, 2009; Kang et al., 2011), where the wide range
of SOA yields was associated with the different OH exposure levels and
precursor concentrations among studies (Table 1). For example, the
higher SOA yield in the study of Ahlberg et al. (2017) was attributed to
the relative lower OH exposure concentration (~0.82 ×10
12
molec
cm
−3
s) in their work compared to that in our experiments (1.2 ×
10
12
–1.4 ×10
12
molec cm
−3
s). Besides, the SOA yield from lower
precursor concentration in this work was slightly lower than those of
Kang et al. (2011) (Table 1 and Fig. 5), in which their SOA yields showed
with obviously larger error bars. This may attribute to the derived SOA
mass from lower precursor concentration that has large uncertainties.
On the other hand, this may be related to the working temperature of the
OFR and the measurement methods of SOA mass. The working tem-
perature is 25 ±1 ◦C and 21 ±1 ◦C in their work and our work
respectively. Besides, SMPS measures the volume concentration of
particles. It converts into mass concentration by an assumed particle
density. The assumed density of particles in our work is 1.2 g cm
−3
,
which is lower than the assumed density (1.4 g cm
−3
) in their work.
3.4. Effect of seed aerosol on SOA formation
Fig. 6 shows the particle number concentration and size distribution
of aerosol from
α
-pinene photo-oxidation in the absence and presence of
acidic seed particles. The OH exposure concentration was 0.8 ×10
12
molec cm
−3
s and
α
-pinene was ~40 ppb. Before UV lamps on, the
background concentrations of particle mass and number were lower
Fig. 3. The dependence between precursor amounts and SOA yield and mass.
The oxidation was conducted at xed O
3
concentration that of 5 ppm. (a) SOA
yield as a function of
α
-pinene amounts. (b) SOA mass as a function of
α
-pinene
amounts. (c) SOA yield as a function of SOA mass. Legend in (a) also applies to
(b) and (c). Error bars indicated the standard deviation of at least three repli-
cate experiments.
R. Zhao et al.

Atmospheric Pollution Research 12 (2021) 392–400
397
than 0.01
μ
g/m
3
and 10 particles cm
−3
respectively. When turning on
the UV lamps, the particle number size distribution showed that particle
number concentration increased rapidly and then reached a stable level
(Fig. 6). Fig. 6a shows that the particle number concentration reached at
the maximum of 8.6 ×10
6
particles cm
−3
after UV radiation for ~3min,
and then that plateaued into the stable of ~8.2 ×10
5
particles cm
−3
after UV radiation for ~15 min. The initial increase of particle number
concentration indicated that there were rapid oxidation and nucleation
of aerosol (Bruns et al., 2015). The stable particle number concentration
and size distribution indicated that the continually introducing of gas
samples into the reactor resulted in the stable formation of aerosol
particles (Pereira et al., 2019). Then the stable formation of aerosol
particles indicated that there was a stable oxidant exposure condition in
this ow reactor. In the seeded OH photo-oxidation experiments, SO
2
was introduced into the reactor, and the sulfuric acidic seed aerosol was
formed by photo-oxidation SO
2
with OH. Fig. 6b shows that the particle
number concentration reached a maximum of 9.2 ×10
6
particles cm
−3
and then that plateaued into the stable of ~9.0 ×10
5
particles cm
−3
. The
addition of acidic seed aerosol signicantly increased the total particle
number concentration of aerosol, especially for large size particles (Dp
>100 nm). The aerosol mass concentrations were 161.9
μ
g/m
3
and
70.5
μ
g/m
3
in the presence and absence of acidic seed particles
respectively. The results indicated that the addition of acidic seed
aerosol is favorable for catalyzing the heterogeneous reactions of low
volatile organic species in the particle-phase (Jang et al., 2002, 2003).
Then this will lead to an increase of particle number concentration and
particle size, further increasing SOA mass.
Fig. 7 shows the effect of OH exposure level on SOA yields in the
presence and absence of acidic seed aerosol. The OH concentration
ranged from 0.1 ×10
12
to 1.8 ×10
12
molec cm
−3
s, which equals to
atmospheric photochemical ages of 0.7–14 days. When OH exposure
concentration at approximately 0.6 ×10
12
molec cm
−3
s, the SOA yields
reached the maximum yield of 0.51 and 0.39 in the presence and
absence of acidic seed aerosol respectively. At different OH exposure
levels, the presence of acidic seed aerosol consistently increases SOA
yield by 1.2–1.5 times, which supports the enhancement of SOA mass. In
addition, the enhancement of yields was similar to that of Kang et al.
(2007), whose SOA yields increased by 1.4 times. Fig. 7 also shows that
both in the presence and absence of acidic seed particles, SOA yields
initially increased and then decreased with the increase of OH exposure
concentration. The increase of SOA yields was attributed to that the
Fig. 4. Number size distribution of aerosol as a function of precursor amounts under different RH conditions. (a) RH <2%, (b) RH =18%, (c) RH =36%.
Table 1
Comparison of the precursor amounts, OH exposure, SOA mass, and yields for
α
-pinene SOA experiments.
α
-pinene
(ppb)
OH exposure
(10
12
molec cm
−3
s)
SOA mass
(
μ
g/m
3
)
SOA yield Reference
28 ±2 1.4 ±0.2 19.1 ±5.3 0.13 ±
0.03
35 ±4 1.2 ±0.1 30.1 ±7.7 0.16 ±
0.04
50 ±4 1.3 ±0.2 90.1 ±15.3 0.33 ±
0.05
This work
65 ±5 1.4 ±0.2 192.9 ±19.4 0.55 ±
0.05
100 ±9 1.4 ±0.2 311.9 ±14.0 0.58 ±
0.05
15 8 0.10 ±
0.10
25 22 0.17 ±
0.09
50 NA
a
78 0.29 ±
0.07
(Chen and
Torres, 2009)
75 152 0.37 ±
0.06
100 249 0.46 ±
0.05
7 ±1 1.5 ±0.5 12 ±5 0.31 ±
0.14
19 ±3 1.4 ±0.5 22 ±5 0.22 ±
0.06
33 ±5 1.4 ±0.5 62 ±7 0.35 ±
0.06
39 ±6 1.3 ±0.5 94 ±9 0.45 ±
0.08
(Kang et al.,
2011)
48 ±8 1.3 ±0.5 83 ±8 0.32 ±
0.05
57 ±9 1.3 ±0.5 150 ±13 0.49 ±
0.08
79 ±13 1.2 ±0.4 220 ±18 0.51 ±
0.08
14–179 0.82 ±0.1 <0.1–95.1 0.01–0.53 (Ahlberg et al.,
2017)
a
OH photo-oxidation experiments were conducted at RH of 37%, resulting in
typical oxidant mixing ratios were 200 ppt for OH and 3 ppb for HO
2
.
R. Zhao et al.

Atmospheric Pollution Research 12 (2021) 392–400
398
oxidation was dominated by functionalization reaction at low OH levels,
whereas the decrease of SOA yields was attributed to that the oxidation
was dominated by fragmentation reaction at high OH levels (Kroll et al.,
2009, 2015). When at low OH levels (<0.4 ×10
12
molec cm
−3
s), with
the increasing of oxidant concentration, the OH photo-oxidation of gas
precursor generated substantial condensation products that have lower
volatility. This continual existence of condensation products increased
aerosol yields. The results showed similar trends with those of previous
ow reactor studies under both seeded and unseeded conditions, but
signicantly lower than the traditional chamber results (Bruns et al.,
2015; Eddingsaas et al., 2012). The high SOA yield in EC from the study
of Bruns et al. (2015) is attributed to the participation of more pre-
cursors (192–200 ppb). For the similar precursor concentration and OH
concentration, the slightly lower yield in the ow reactor may be
attributed to the losses of low-volatility compounds. Table S1 shows the
detailed experimental conditions and SOA yields in this work and
Fig. 5. Comparison of SOA yields to those of previous OFR studies at high OH exposure level. Error bars indicated the standard deviation of at least three replicate
experiments.
Fig. 6. Particle number concentration and size distribution of aerosol in OH photo-oxidation. (a) The
α
-pinene oxidation. (b) The
α
-pinene oxidation with the
addition of acid seed aerosol. Particle wall losses have not been taken into account.
R. Zhao et al.

Atmospheric Pollution Research 12 (2021) 392–400
399
previous works. While at high OH levels (>0.8 ×10
12
molec cm
−3
s),
fragmentation reaction becomes signicant with the continual
increasing OH exposure concentration (Simonen et al., 2017). It was
attributed to the carbon-carbon bond breaks in the gas phase, forming
small molecules with high volatility and the fragmented molecules are
unable to condense to form particles (Hunter et al., 2014; Kroll et al.,
2009). Even if the seed particles were added, it is difcult to reduce
completely the inuence of fragmentation reaction on the aerosol par-
ticle formation. These results show reasonable agreements with those of
OFR/PAM experiments in the study of Bruns et al. (2015), though
different concentrations of precursors were used to participate in
oxidation experiments. However, the results were different from those of
Li et al. (2019). In their work, the high OH exposure has little effect on
seeded SOA yields. Their consistent higher SOA yields may be attributed
to the lower precursor concentration (13.7 ppb and 41 ppb for previous
and current works respectively) for oxidation and lower losses of gas and
particle species.
4. Conclusion
In this work, we provided the basic characterization of the custom-
built OFR and used it to study the yields from
α
-pinene oxidation. Re-
sults from the experiments were compared to those of previous OFR/
PAM and EC studies. In our ow reactor, the temperature, RH, gas
precursor concentration, and oxidant exposure concentration can be
controlled independently, which enables to control independently
exposure oxidative conditions for simulating aerosol aging.
Particle losses for this OFR were very small, with 24% at 50 nm and
12% at 100 nm, and nearly 0% for particle size above 150 nm. The re-
sults indicated that the particle transmission efciency of our ow
reactor is comparable to or even better than that of the traditional OFR/
PAM reactors. In O
3
dominated oxidation, the results showed that SOA
yields were mainly dependent on O
3
concentration, precursor amounts,
and SOA mass, but less dependent on RH. Although the RH has little
effect on SOA yield, the number concentration of aerosol showed a slight
decrease and the particle medium diameter showed a slight increase as
the RH increased. In OH photo-oxidation, SOA yields were signicantly
dependent on RH, precursor amounts, SOA mass, OH exposure con-
centration, and sulfate seed particles. The formation of aerosol particles
nearly achieved stability after UV radiation for 15 min. In the seeded
experiments, heterogeneous acid-catalyzed reaction initiated by acidic
seed aerosol plays a signicant role in the enhancement of SOA mass.
The addition of acidic seed aerosol increased both particle number
concentration and particle size, resulting in an increase of SOA yield of
1.2–1.5 times at the photochemical age ranged from 0.7 to 14 days.
This work focused on the systematical characterization of OFR
oxidation experiments and the yields validation from
α
-pinene SOA
formation. The results highlight the importance of systematical char-
acterization of OFR oxidation experiments in different conditions, which
provided great guidance for the continued using the OFR to simulate the
aging of atmospheric related emissions.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgments
This work was nancially supported by the National Natural Science
Foundation of China (41675024) and the Natural Science Foundation of
Anhui Province (1908085QD157).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.apr.2021.01.011.
References
Ahlberg, E., Falk, J., Eriksson, A., Holst, T., Brune, W.H., Kristensson, A., Roldin, P.,
Svenningsson, B., 2017. Secondary organic aerosol from VOC mixtures in an
oxidation ow reactor. Atmos. Environ. 161, 210–220.
Atkinson, R., Arey, J., 2003. Gas-phase tropospheric chemistry of biogenic volatile
organic compounds: a review. Atmos. Environ. 37, 197–219.
Bruns, E.A., El Haddad, I., Keller, A., Klein, F., Kumar, N.K., Pieber, S.M., Corbin, J.C.,
Slowik, J.G., Brune, W.H., Baltensperger, U., et al., 2015. Inter-comparison of
laboratory smog chamber and ow reactor systems on organic aerosol yield and
composition. Atmos. Meas. Tech. 8, 2315–2332.
Fig. 7. Comparison of SOA yields as a function of OH exposure concentration in the presence and absence of sulfate aerosol. Error bars indicated the standard
deviation of at least three replicate experiments.
R. Zhao et al.

Atmospheric Pollution Research 12 (2021) 392–400
400
Chen, S., Brune, W.H., Lambe, A.T., Davidovits, P., Onasch, T.B., 2013. Modeling organic
aerosol from the oxidation of
α
-pinene in a Potential Aerosol Mass (PAM) chamber.
Atmos. Chem. Phys. 13, 5017–5031.
Chen, Z., Torres, O., 2009. An examination of oxidant amounts on secondary organic
aerosol formation and aging. Atmos. Environ. 43, 3579–3585.
Chu, B., Liu, Y., Ma, Q., Ma, J., He, H., Wang, G., Cheng, S., Wang, X., 2016. Distinct
potential aerosol masses under different scenarios of transport at a suburban site of
Beijing. J. Environ. Sci. 39, 52–61.
Donahue, N.M., Robinson, A.L., Stanier, C.O., Pandis, S.N., 2006. Coupled partitioning,
dilution, and chemical aging of semivolatile organics. Environ. Sci. Technol. 40,
2635–2643.
Eddingsaas, N.C., Loza, C.L., Yee, L.D., Chan, M., Schilling, K.A., Chhabra, P.S.,
Seinfeld, J.H., Wennberg, P.O., 2012.
α
-pinene photooxidation under controlled
chemical conditions – Part 2: SOA yield and composition in low- and high-NOx
environments. Atmos. Chem. Phys. 12, 7413–7427.
Feng, Z., Huang, M., Cai, S., Xu, X., Yang, Z., Zhao, W., Hu, C., Gu, X., Zhang, W., 2019.
Characterization of single scattering albedo and chemical components of aged
toluene secondary organic aerosol. Atmos. Pollut. Res. 10, 1736–1744.
Ge, S., Xu, Y., Jia, L., 2017. Effects of inorganic seeds on secondary organic aerosol
formation from photochemical oxidation of acetone in a chamber. Atmos. Environ.
170, 205–215.
Hallquist, M., Wenger, J.C., Baltensperger, U., Rudich, Y., Simpson, D., Claeys, M.,
Dommen, J., Donahue, N.M., George, C., Goldstein, A.H., et al., 2009. The formation,
properties and impact of secondary organic aerosol: current and emerging issues.
Atmos. Chem. Phys. 9, 5155–5236.
Huang, Y., Coggon, M.M., Zhao, R., Lignell, H., Bauer, M.U., Flagan, R.C., Seinfeld, J.H.,
2017. The Caltech Photooxidation Flow Tube reactor: design, uid dynamics and
characterization. Atmos. Meas. Tech. 10, 839–867.
Hunter, J.F., Carrasquillo, A.J., Daumit, K.E., Kroll, J.H., 2014. Secondary organic
aerosol formation from acyclic, monocyclic, and polycyclic alkanes. Environ. Sci.
Technol. 48, 10227–10234.
Ihalainen, M., Tiitta, P., Czech, H., Yli-Piril¨
a, P., Hartikainen, A., Kortelainen, M.,
Tissari, J., Stengel, B., Sklorz, M., Suhonen, H., et al., 2019. A novel high-volume
Photochemical Emission Aging ow tube Reactor (PEAR). Aerosol Sci. Technol. 53,
276–294.
Jang, M., Carroll, B., Chandramouli, B., Kamens, R.M., 2003. Particle growth by acid-
catalyzed heterogeneous reactions of organic carbonyls on preexisting aerosols.
Environ. Sci. Technol. 37, 3828–3837.
Jang, M., Czoschke, N.M., Lee, S., Kamens, R.M., 2002. Heterogeneous atmospheric
aerosol production by acid-catalyzed particle-phase reactions. Science 298,
814–817.
Jimenez, J.L., Canagaratna, M.R., Donahue, N.M., Prevot, A.S., Zhang, Q., Kroll, J.H.,
DeCarlo, P.F., Allan, J.D., Coe, H., Ng, N.L., et al., 2009. Evolution of organic
aerosols in the atmosphere. Science 326, 1525–1529.
Kang, E., Root, M.J., Toohey, D.W., Brune, W.H., 2007. Introducing the concept of
potential aerosol mass (PAM). Atmos. Chem. Phys. 7, 5727–5744.
Kang, E., Toohey, D.W., Brune, W.H., 2011. Dependence of SOA oxidation on organic
aerosol mass concentration and OH exposure: experimental PAM chamber studies.
Atmos. Chem. Phys. 11, 1837–1852.
Karjalainen, P., Timonen, H., Saukko, E., Kuuluvainen, H., Saarikoski, S., Aakko-
Saksa, P., Murtonen, T., Bloss, M., Dal Maso, M., Simonen, P., et al., 2016. Time-
resolved characterization of primary particle emissions and secondary particle
formation from a modern gasoline passenger car. Atmos. Chem. Phys. 16,
8559–8570.
Keller, A., Burtscher, H., 2012. A continuous photo-oxidation ow reactor for a dened
measurement of the SOA formation potential of wood burning emissions. J. Aerosol
Sci. 49, 9–20.
Kroll, J.H., Lim, C.Y., Kessler, S.H., Wilson, K.R., 2015. Heterogeneous oxidation of
atmospheric organic aerosol: kinetics of changes to the amount and oxidation state
of particle-phase organic carbon. J. Phys. Chem. 119, 10767–10783.
Kroll, J.H., Smith, J.D., Che, D.L., Kessler, S.H., Worsnop, D.R., Wilson, K.R., 2009.
Measurement of fragmentation and functionalization pathways in the heterogeneous
oxidation of oxidized organic aerosol. Phys. Chem. Chem. Phys. 11, 8005–8014.
La, Y.S., Camredon, M., Ziemann, P.J., Valorso, R., Matsunaga, A., Lannuque, V., Lee-
Taylor, J., Hodzic, A., Madronich, S., Aumont, B., 2016. Impact of chamber wall loss
of gaseous organic compounds on secondary organic aerosol formation: explicit
modeling of SOA formation from alkane and alkene oxidation. Atmos. Chem. Phys.
16, 1417–1431.
Lambe, A.T., Ahern, A.T., Williams, L.R., Slowik, J.G., Wong, J.P.S., Abbatt, J.P.D.,
Brune, W.H., Ng, N.L., Wright, J.P., Croasdale, D.R., et al., 2011. Characterization of
aerosol photooxidation ow reactors: heterogeneous oxidation, secondary organic
aerosol formation and cloud condensation nuclei activity measurements. Atmos.
Meas. Tech. 4, 445–461.
Lambe, A.T., Chhabra, P.S., Onasch, T.B., Brune, W.H., Hunter, J.F., Kroll, J.H.,
Cummings, M.J., Brogan, J.F., Parmar, Y., Worsnop, D.R., et al., 2015. Effect of
oxidant concentration, exposure time, and seed particles on secondary organic
aerosol chemical composition and yield. Atmos. Chem. Phys. 15, 3063–3075.
Li, K., Liggio, J., Lee, P., Han, C., Liu, Q., Li, S.-M., 2019. Secondary organic aerosol
formation from
α
-pinene, alkanes, and oil-sands-related precursors in a new
oxidation ow reactor. Atmos. Chem. Phys. 19, 9715–9731.
Li, R., Palm, B.B., Ortega, A.M., Hlywiak, J., Hu, W., Peng, Z., Day, D.A., Knote, C.,
Brune, W.H., de Gouw, J.A., et al., 2015. Modeling the radical chemistry in an
oxidation ow reactor: radical formation and recycling, sensitivities, and the OH
exposure estimation equation. J. Phys. Chem. 119, 4418–4432.
Loza, C.L., Craven, J.S., Yee, L.D., Coggon, M.M., Schwantes, R.H., Shiraiwa, M.,
Zhang, X., Schilling, K.A., Ng, N.L., Canagaratna, M.R., et al., 2014. Secondary
organic aerosol yields of 12-carbon alkanes. Atmos. Chem. Phys. 14, 1423–1439.
Mao, J., Ren, X., Brune, W.H., Olson, J.R., Crawford, J.H., Fried, A., Huey, L.G.,
Cohen, R.C., Heikes, B., Singh, H.B., et al., 2009. Airborne measurement of OH
reactivity during INTEX-B. Atmos. Chem. Phys. 9, 163–173.
Mozaffar, A., Zhang, Y.-L., Fan, M., Cao, F., Lin, Y.-C., 2020. Characteristics of
summertime ambient VOCs and their contributions to O3 and SOA formation in a
suburban area of Nanjing, China. Atmos. Res. 240.
Ng, N.L., Canagaratna, M.R., Zhang, Q., Jimenez, J.L., Tian, J., Ulbrich, I.M., Kroll, J.H.,
Docherty, K.S., Chhabra, P.S., Bahreini, R., et al., 2010. Organic aerosol components
observed in northern hemispheric datasets from aerosol mass spectrometry. Atmos.
Chem. Phys. 10, 4625–4641.
Northcross, A.L., Jang, M., 2007. Heterogeneous SOA yield from ozonolysis of
monoterpenes in the presence of inorganic acid. Atmos. Environ. 41, 1483–1493.
Odum, J.R., Hoffmann, T., Bowman, F., Collins, D., Flagan, R.C., Seinfeld, J.H., 1996.
Gas/particle partitioning and secondary organic aerosol yields. Environ. Sci.
Technol. 30, 2580–2585.
Palm, B.B., Campuzano-Jost, P., Ortega, A.M., Day, D.A., Kaser, L., Jud, W., Karl, T.,
Hansel, A., Hunter, J.F., Cross, E.S., et al., 2016. In situ secondary organic aerosol
formation from ambient pine forest air using an oxidation ow reactor. Atmos.
Chem. Phys. 16, 2943–2970.
Peng, Z., Day, D.A., Ortega, A.M., Palm, B.B., Hu, W., Stark, H., Li, R., Tsigaridis, K.,
Brune, W.H., Jimenez, J.L., 2016. Non-OH chemistry in oxidation ow reactors for
the study of atmospheric chemistry systematically examined by modeling. Atmos.
Chem. Phys. 16, 4283–4305.
Peng, Z., Day, D.A., Stark, H., Li, R., Lee-Taylor, J., Palm, B.B., Brune, W.H., Jimenez, J.
L., 2015. HOx radical chemistry in oxidation ow reactors with low-pressure
mercury lamps systematically examined by modeling. Atmos. Meas. Tech. 8,
4863–4890.
Peng, Z., Lee-Taylor, J., Orlando, J.J., Tyndall, G.S., Jimenez, J.L., 2019. Organic peroxy
radical chemistry in oxidation ow reactors and environmental chambers and their
atmospheric relevance. Atmos. Chem. Phys. 19, 813–834.
Pereira, K.L., Rovelli, G., Song, Y.C., Mayhew, A.W., Reid, J.P., Hamilton, J.F., 2019.
A new aerosol ow reactor to study secondary organic aerosol. Atmos. Meas. Tech.
12, 4519–4541.
Sbai, S.E., Farida, B., 2019. Photochemical aging and secondary organic aerosols
generated from limonene in an oxidation ow reactor. Environ. Sci. Pollut. Res. Int.
26, 18411–18420.
Seinfeld, J.H., Erdakos, G.B., Asher, W.E., Pankow, J.F., 2001. Modeling the formation of
secondary organic aerosol (SOA). 2. The predicted effect of relative humidity on
aerosol formation in the
α
-pinene, β-pinene, sabinene, Δ
3
-carene, and
cyclohexeneozone systems. Environ. Sci. Technol. 35, 1806–1817.
Seinfeld, J.H., Pankow, J.F., 2003. Organic atmospheric particulate material. Annu. Rev.
Phys. Chem. 54, 121–140.
Simonen, P., Saukko, E., Karjalainen, P., Timonen, H., Bloss, M., Aakko-Saksa, P.,
R¨
onkk¨
o, T., Keskinen, J., Dal Maso, M., 2017. A new oxidation ow reactor for
measuring secondary aerosol formation of rapidly changing emission sources.
Atmos. Meas. Tech. 10, 1519–1537.
Wang, S., Du, L., Tsona, N.T., Jiang, X., You, B., Xu, L., Yang, Z., Wang, W., 2020. Effect
of NOx and SO2 on the photooxidation of methylglyoxal: implications in secondary
aerosol formation. J. Environ. Sci. 92, 151–162.
Wang, X., Liu, T., Bernard, F., Ding, X., Wen, S., Zhang, Y., Zhang, Z., He, Q., Lü, S.,
Chen, J., et al., 2014. Design and characterization of a smog chamber for studying
gas-phase chemical mechanisms and aerosol formation. Atmos. Meas. Tech. 7,
301–313.
Wu, X., Liu, J., Wu, Y., Wang, X., Yu, X., Shi, J., Bi, J., Huang, Z., Zhou, T., Zhang, R.,
2018. Aerosol optical absorption coefcients at a rural site in Northwest China: the
great contribution of dust particles. Atmos. Environ. 189, 145–152.
Zhang, W., Wang, W., Li, J., Peng, C., Li, K., Zhou, L., Shi, B., Chen, Y., Liu, M., Ge, M.,
2020. Effects of SO2 on optical properties of secondary organic aerosol generated
from photooxidation of toluene under different relative humidity conditions. Atmos.
Chem. Phys. 20, 4477–4492.
Zhang, Y., Sanchez, M.S., Douet, C., Wang, Y., Bateman, A.P., Gong, Z., Kuwata, M.,
Renbaum-Wolff, L., Sato, B.B., Liu, P.F., et al., 2015. Changing shapes and implied
viscosities of suspended submicron particles. Atmos. Chem. Phys. Discuss. 15,
6821–6850.
Zhao, D.F., Buchholz, A., Tillmann, R., Kleist, E., Wu, C., Rubach, F., Kiendler-Scharr, A.,
Rudich, Y., Wildt, J., Mentel, T.F., 2017. Environmental conditions regulate the
impact of plants on cloud formation. Nat. Commun. 8.
Zhao, D.F., Kaminski, M., Schlag, P., Fuchs, H., Acir, I.H., Bohn, B., H¨
aseler, R., Kiendler-
Scharr, A., Rohrer, F., Tillmann, R., et al., 2015. Secondary organic aerosol
formation from hydroxyl radical oxidation and ozonolysis of monoterpenes. Atmos.
Chem. Phys. 15, 991–1012.
Ziemann, P.J., Atkinson, R., 2012. Kinetics, products, and mechanisms of secondary
organic aerosol formation. Chem. Soc. Rev. 41, 6582–6605.
R. Zhao et al.