Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Atmospheric Environment 213 (2019) 231–238

Contents lists available at ScienceDirect

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

Kinetics and products of the aqueous-phase oxidation of β-caryophyllonic T


acid by hydroxyl radicals
Bartłomiej Witkowski∗, Mohammed Al-sharafi, Tomasz Gierczak
University of Warsaw, Faculty of Chemistry, al. Żwirki i Wigury 101, 02-089 Warsaw, Poland

ARTICLE INFO ABSTRACT

Keywords: Kinetics and mechanism of β-caryophyllonic acid (BCA) oxidation by hydroxyl radicals (OH) in the aqueous-
β-Caryophyllene phase were investigated at 298 ± 2 K using liquid chromatography (LC) coupled to the electrospray ionization
Secondary organic aerosol (ESI) mass spectrometry (MS). Triple quadrupole mass spectrometer operating in the multiple reaction mon-
Aqueous-phase itoring (MRM) was used to follow concentrations of organic compounds during the photooxidation at cloud-
Relative rates
relevant concentrations (several μM). The rate coefficients obtained for BCA + OH reaction were:
Hydroxyl radicals
4.2 ± 0.5 × 1010 M−1s−1 at pH = 2 and 6.5 ± 0.7 × 109 M−1s−1 at pH = 8. The kinetic data obtained in this
study was used to estimate atmospheric lifetimes of BCA in the aqueous-phase: between 24 min and 12 h de-
pending on pH. The results indicated that aqueous-phase oxidation is a relevant removal pathway for this
compound in clouds, fogs and wet aerosols. The products of BCA + OH reaction were studied with the high
resolution tandem MS. The two major, first-generation products identified were keto-BCA and hydroxyl-hy-
droperoxy BCA; these two molecules were formed by OH addition to the terminal C=C bond of the precursor.
The rest of the products were identified as oxygenated BCA derivatives with O:C ratios higher than the precursor
although some fragmentation of the original carbon backbone was also observed. Thus, the data acquired
provided insights into the mechanism of BCA + OH reaction.

1. Introduction Winterhalter et al., 2009). β-caryophyllene is the most abundant ses-


quiterpene in the atmosphere (Duhl et al., 2008; Hellén et al., 2018;
Global, non-methane hydrocarbons emission is dominated by vola- Helmig et al., 2007). β-caryophyllene is also an efficient precursor of
tile organic compounds (VOC) which are mostly emitted by terrestrial SOA as confirmed by the laboratory studies (Chan et al., 2011; Tasoglou
vegetation (Messina et al., 2016; Sindelarova et al., 2014). Isoprene, and Pandis, 2015) and field observations (Ding et al., 2014; Faiola
mono and sesquiterpenes are the most abundant non-methane VOCs in et al., 2018; Hellén et al., 2018; Jaoui et al., 2007; van Eijck et al.,
the atmosphere. These biogenic VOCs (BVOCs) undergo oxidation to 2013).
form secondary organic aerosols (SOAs) (Fuzzi et al., 2015; Zhang et al., Currently, a lot of attention is directed towards oxidation of semi-
2018). Consequently, SOAs formed from BVOCs oxidation significantly volatile organic compounds inside and on the surface of aqueous par-
impact the climate as well as human health (Carslaw et al., 2013; Fuzzi ticles in order to improve the understanding of SOA formation me-
et al., 2015; Shrivastava et al., 2017; Tsigaridis and Kanakidou, 2018). chanisms (Faust et al., 2017; McNeill, 2015; Sareen et al., 2017). SOA
Emissions of sesquiterpenes are lower than emissions of mono- formed in the aqueous-phase (aqSOA) is considered as a significant
terpenes (Messina et al., 2016; Sindelarova et al., 2014). However, source of highly-oxygenated, low-volatility organic compounds in the
sesquiterpenes are characterized by higher reactivity towards hydroxyl atmosphere (Ervens et al., 2011; Herrmann et al., 2015; Krechmer et al.,
radicals (OH) and ozone (O3) (Atkinson and Arey, 2003; Lee et al., 2015; Lim et al., 2010). Moreover, previously published results de-
2006; Richters et al., 2015; Shu and Atkinson, 1995) than mono- monstrated a high potential for aqSOA formation following oxidation of
terpenes (Lee et al., 2006; Leungsakul et al., 2005). Moreover, oxidation terpenoic acids (Aljawhary et al., 2016; Enami and Sakamoto, 2016a;
products of sesquiterpenes are high molecular weight (HMW) and low Otto et al., 2018; Witkowski et al., 2018a, 2018b). However, to-date,
volatility compounds (Bonn and Moortgat, 2003). For these reasons, water-soluble oxidation products of sesquiterpenes have not been in-
oxidation of sesquiterpenes significantly contributes to atmospheric vestigated as the potential aqSOA precursors.
SOA (Jaoui et al., 2013; Lee et al., 2006; Tasoglou and Pandis, 2015; β-caryophyllene possess both endo and exocyclic carbon-carbon


Corresponding author.
E-mail address: bwitk@chem.uw.edu.pl (B. Witkowski).

https://doi.org/10.1016/j.atmosenv.2019.06.016
Received 27 February 2019; Received in revised form 4 June 2019; Accepted 6 June 2019
Available online 06 June 2019
1352-2310/ © 2019 Elsevier Ltd. All rights reserved.
B. Witkowski, et al. Atmospheric Environment 213 (2019) 231–238

double bonds (Nguyen et al., 2009; Winterhalter et al., 2009). Conse-


quently, the first generation oxidation products of β-caryophyllene are
unsaturated because they retain the less reactive, exocyclic bond
(Kundu et al., 2017; van Eijck et al., 2013). β-caryophyllonic acid
(C15H24O3, BCA - reaction (1)) is a major, first-generation oxidation
product of β–caryophyllene as demonstrated by both laboratory and
field studies (Gray Bé et al., 2017; Jaoui et al., 2007; Kundu et al., 2017;
van Eijck et al., 2013). BCA is an unsaturated, semi-volatile carboxylic
acid with a large carbon backbone (van Eijck et al., 2013) and it is also
soluble in water (Gray Bé et al., 2017). Moreover, the estimated Henrys-
law constant (H) of BCA is 3.1 × 105 M atm−1 (Grieshop et al., 2007).
Terpenoic acids with such a high H values will reside mostly in the
aqueous phase in clouds and fogs, as previously concluded (Lignell
et al., 2013; Otto et al., 2018; Witkowski et al., 2018a, 2018b).
Previously, it was demonstrated that oxidation of terpenoic acids in
the aqueous-phase resulted in the formation of highly-oxidized, low-
volatility products (Enami and Sakamoto, 2016a; Witkowski and
Gierczak, 2017a; Witkowski et al., 2018b). Even when some fragmen-
tation of the original carbon backbone occurs, the resulting products
were formed mostly by addition of the oxygenated functional groups to
the backbone of the precursor (Aljawhary et al., 2016; Enami and
Sakamoto, 2016a; Witkowski and Gierczak, 2017a; Witkowski et al.,
2018b). Therefore, aqueous-phase oxidation of BCA by hydroxyl radi-
cals (OH) can lead to aqSOA under realistic atmospheric conditions.
In this work, the reaction of BCA with OH - reaction (1) - was in- Fig. 1. Aqueous photoreactor.
vestigated in the newly constructed aqueous photoreactor:
The speed of fan rotation is controlled the temperature sensor inside the
lamp chamber thereby maintaining the temperature at ca. 298 ± 2 K.
The magnetic stirrer is used to mix the reaction solution during the
photooxidation. A Pyrex glass bottle (100 ml) was used as a reaction
vessel in all experiments (Witkowski et al., 2018a, 2018b; Witkowski
and Gierczak, 2017a).

Kinetics and mechanism of reaction (1) were studied using liquid 2.3. Kinetics of β-caryophyllonic acid + OH reaction
chromatography coupled to the electrospray ionization mass spectro-
metry (LC-ESI/MS). Kinetics of reaction (1) were measured using the The reaction mixture consisted of phosphoric acid (pH = 2) or so-
relative rate method. Moreover, products formed were analyzed with dium phosphate buffer (pH = 8) solution in deionized (DI) water. To
LC coupled to the high resolution (HR) MS. BCA was synthesized from that buffered reaction solution, BCA and the reference compounds were
β–caryophyllene and purified using a semi-preparative LC. added (concentration of each organic compound was ca. 7 μM). The
concentration of H2O2, an OH precursor, was ca. 3 mM. Reaction so-
2. Experimental section lution pH was checked with a glass electrode before and after each
experiment. During the photooxidation, 100 μl aliquots of the reaction
The list of reagents is provided in section S1 of the Supplementary solution were taken from the reactor every 10 min and immediately
Information (SI). quenched with a buffered catalase solution (ca. 0.05 mg/ml). Catalase
was dissolved in pH = 7 or pH = 5, 50 μM ammonium acetate buffer
2.1. Synthesis of β-caryophyllonic acid for the acidic or basic reaction solution, respectively. After decom-
posing the leftover H2O2 (ca. 5 min), 20 μl of acetonitrile (ACN) was
BCA (≥95%) was synthesized from β-caryophyllene as described in added and the sample was filtered through the 0.22 μm filter to remove
section S2. Initially, β-caryophyllene aldehyde was prepared directly any catalase that might have been denatured by ACN. The samples were
from β-caryophyllene using pyridine as an organocatalyst for the re- then analyzed by LC/MS (section 2.5.1).
ductive ozonolysis (Willand-Charnley et al., 2012). Then, β-car- Relative rate method was used to measure rate coefficients of re-
yophyllene aldehyde was oxidized to BCA with sodium chlorite action (1) using the reference compounds listed in Table 1.
(NaClO2) (Dalcanale and Montanari, 1986). Finally, BCA was purified
with a semi-preparative HPLC. The data available in the literature was Table 1
Bimolecular reaction rate coefficients for the aqueous-phase oxidation of the
used to confirm the structure of the synthesized standard using 1H NMR
reference compounds by OH.
(van Eijck et al., 2013).
Reference compounds k2 for oxidation by OH (M−1 s−1) × 10−9a

2.2. Aqueous photoreactor pH = 2 pH=8

The photoreactor (Fig. 1) designed in our laboratory was used for Caffeine 6.9 –
Limononic acid (LA) 13 ± 0.1 5.4 ± 0.1
the reaction (1) in the aqueous-phase.
Phenylalanine 5.7 6.7 ± 0.3
The photoreactor shown in Fig. 1 is equipped with four UVB lamps p-Toluic acid – 8.0
(Philips, PL-S 9W/01/2P 1CT). The circular reaction chamber is air-
cooled by a fan located underneath the plate of the magnetic stirrer. a
Uncertainties are provided if listed in the original study.

232
B. Witkowski, et al. Atmospheric Environment 213 (2019) 231–238

As listed in Table 1, kinetics of the OH oxidation reaction were in- 2.6. Uncertainty and control experiments
vestigated for neutral (pH = 2) and deprotonated (pH = 8) BCA using
the rate coefficients listed in the literature (Buxton et al., 1988; Control experiments were carried out to confirm that the dis-
Witkowski et al., 2018a, 2018b). Eq. (I) was used to calculate the bi- appearance of the analytes was only due to reaction with OH, within
molecular reaction rate coefficients using the experimental data ac- the time-scale of the experiments. 35,36,42 Photooxidation was carried
quired: out under acidic and basic conditions without adding H2O2 to the re-
action mixture to check if the compounds listed in Table 1 would un-
[BCA]0 k1 [Ref]0 dergo direct photolysis. Another set of control experiments was also
Ln = Ln
[BCA]t k2 [Ref]t (I) carried for both acidic and basic solutions without turning on the lamp
to check if the analytes didn't reacted with H2O2. Finally, the relative
[BCA] and [Ref] are the initial (0) and intermediate (t) concentra- kinetic plots obtained for the reference compounds were compared with
tions of BCA and the reference compound and k1 and k2 are the bi- the literature data to check if the aqueous matrix (phosphate buffer)
molecular reaction rate coefficients for the aqueous-phase oxidation of had any impact on the kinetic results obtained. Results of these control
BCA and of the reference compounds by OH, respectively. experiments (see section S4) confirmed that the procedure used to
measure k1 values was reliable.
2.4. Mechanism of β-caryophyllonic acid + OH reaction Uncertainties values of the k1 values measured were between 10 and
15%. The combined uncertainty was calculated by taking into account:
The reaction solution consisted of BCA (concentration was between precision of the linear regression fits to the experimental data (eq. I),
20 and 30 μM) and H2O2 without any buffering agents. Aliquots of uncertainties of the kref values listed in the literature (Table 1) and the
200 μl were sampled every 10 min and subjected to the LC/MS analyses. uncertainty of the LC/MS measurements (chromatographic peak area
Temporal profiles of the products of reaction (1) were obtained with integration).
triple-quadrupole instrument (section 2.5.1). Moreover, the sample
after 30 min of the irradiation was analyzed with LC coupled with the 3. Results and discussion
ESI-Orbitrap mass spectrometer – section 2.5.2.
3.1. Rate coefficients (k1) measurements for reaction of OH radicals with β-
caryophyllonic acid
2.5. LC/MS analyses
Relative kinetic plots obtained in this work are shown in Fig. 2;
All analyses were carried out with reverse phase (RP) Luna
samples were prepared as described in section 2.3 and analyzed as
(Phenomenex) C18 column (100 mm × 2.1 mm, 3 μm, 100 Å) with a
described in section 2.5.1.
2 mm C18 pre-column. Flow rate for the mobile phase was 0.2 ml/min.
As shown in Fig. 2, the relative kinetic plots obtained were linear
Eluent A was 0.03% formic acid solution in water (pH = 2.8) and
(R2 > 0.99). Therefore, the slopes obtained were used to calculate k1
eluent B was ACN. Gradient elution program is provided in section S3.
values with eq. I. k1 values measured for reaction (1) were consistent
for all of the reference compounds as indicated by the results listed in
2.5.1. Analyses with LC-ESI-triple quadrupole MS Table 2. kOH values measured in this work as well as literature data for
Analyses were carried out with LC20A liquid chromatograph the other terpenoic acids are listed in Table 2.
(Shimadzu) coupled with the QTRAP 3200 (AB Sciex) triple quadrupole As listed in Table 2, kOH value obtained for neutral (protonated) BCA is
mass spectrometer equipped with the electrospray (ESI) ion source and very high as compared to the other unsaturated terpenoic acids. k1 values
the ionization conditions were: spray voltage: 4.5 kV (negative mode) measured also exceed the diffusion limit for the aqueous-phase oxidation
and 5.5 kV (positive mode), N2 was curtain (3 × 105 Pa), auxiliary of organic compounds by OH (1.3–1.5 × 1010 M−1 s−1) imposed by the
(3 × 105 Pa) and collision gas, source temperature was 450 °C. The Smoluchowski equation (Otto et al., 2017; Schöne et al., 2014). However,
triple-quadrupole MS was operating in the multiple reaction monitoring there are a number molecules for whom the experimentally obtained kOH
(MRM) and scan (50–700 m/z) modes. MRM analysis conditions sample exceed this diffusion limit by a factor of 2–4 (see section S5) (Buxton et al.,
chromatogram of the analytes listed in Table 1 are presented in section 1988). Clearly, further investigation of the diffusion-limitation of the
S3.

2.5.2. Analyses with LC-ESI-Orbitrap MS


Analyses were carried out with Agilent series 1200 liquid chroma-
tograph coupled with the Orbitrap mass spectrometer (Thermo Fisher
Scientific) equipped with the ESI probe. The ionization conditions were
as follows: spray voltage - 2.5 kV (negative mode) and 3.5 kV (positive
mode), ion transfer tube temperature 325 °C and vaporizer temperature
275 °C. N2 was a sheath gas (35 arbitrary units, arb), auxiliary (10 arb)
and collision gas. Orbitrap mass resolution was 60000 and it was op-
erating in scan (50–1200 m/z) and in the automated MS/MS modes. In
the automated MS/MS mode the precursor ion with the intensity well
above the background is automatically selected and subjected to the
collision induced dissociation (CID) using low (15, arb) and high (55,
arb) collision energies.
The data acquisition was carried out only in the negative ionization
mode due to very low ionization efficiency of the analytes in the po-
sitive ionization mode (results not shown). Elemental formulas were
assigned to the detected [M-H]- pseudomolecular ions. The [M-H]- ions
detected were assumed to be singly charged and contain C, H and O
atoms. Fig. 2. Relative kinetic plots for reaction (1).

233
B. Witkowski, et al. Atmospheric Environment 213 (2019) 231–238

Table 2
kOH of the terpenoic acids measured at 298 K under acidic and basic conditions.
Name Structure kOH (M−1 s−1) × 10−10 Ref.


Neutral (AH) Dissociated (A )

β-caryophyllonic acid 4.2 ± 0.5 0.65 ± 0.07 This work

Limononic acid 1.3 ± 0.1 0.54 ± 0.1 (Witkowski et al., 2018a, 2018b)

Limonic acida,b 1.3 ± 0.2 0.39 ± 0.06 30

Dicarbonyl derivative of limononic acid 1.3 ± 0.2 0.50 ± 0.10

7-hydroxy limononic acid 1.2 ± 0.3 0.8 ± 0.1

cis-Pinonic acid 0.36 ± 0.03 0.30 ± 0.03 Witkowski and Gierczak (2017a)
0.33 ± 0.05 – Aljawhary et al. (2016)
0.27 ± 0.01 0.25 ± 0.01 Otto et al. (2018)
Camphoric acidc 0.20 ± 0.01 0.26 ± 0.03 Otto et al. (2018)

a
Average values measured in pH = 7 and 8.
b
As indicated by the estimated pKa values the limonic acid exists as dianion when pH > 7.
c
Not an atmospherically-abundant terpenoic acid, a surrogate for dicarboxylic terpenoic acids; the same kOH values were obtained for mono and dianion of
camphoric acid (Otto et al., 2018).

aqueous-phase reactions between organic compounds and inorganic ra-


dicals is needed but such a study was beyond the scope of this work.
The data listed in Table 2 indicates that the reactivity of protonated
(AH) form of the terpenoic acids towards OH is generally higher than
the reactivity of the corresponding carboxylate anions (A−) (Witkowski
et al., 2018a, 2018b). On the other hand, no pH dependence of the kOH
was observed for two saturated carboxylic acids: CPA and camphoric
acid (Otto et al., 2017; Witkowski and Gierczak, 2017a). Further studies
are needed before a more general conclusions about reactivity of ter-
penoic acids towards OH can be presented.
The k1 values measured were used to estimate lifetimes of BCA in
the atmosphere - section 4.

3.2. Mechanism of β-caryophyllonic acid oxidation by hydroxyl radicals

Chromatogram of the reaction solution after 30 min of irradiation is


shown in Fig. 3. Samples were prepared as described in section 2.4 and
analyzed by LC-HR/MS (section 2.5.2).
As it could be seen from Fig. 3, several products of reaction (1) were Fig. 3. LC-HR/MS analysis of reaction (1) products, negative ionization mode.
detected. Two chromatographic peaks were detected for compound III
and the data acquired indicated that several isomers of compounds II
information about products of reaction (1). Compound I (the most
and V/VI were present in the reaction mixture – Fig. S9.
abundant product detected) was identified as keto-BCA as indicated by
Elemental formulas were assigned to the [M-H]- ions detected and
the fragmentation spectrum shown in Fig. 4 (Alfarra et al., 2012). Two
used to calculate the aromaticity indexes (AI), double bonds equivalents
isomeric compounds III (C15H26O6), were the second most abundant
(DBE), oxygen to carbon ratio (O:C) and hydrogen to carbon (H:C) ratio
products (Fig. 3). The [M-H]- ions of compounds III readily eliminated
shown in Table 3 (Bateman et al., 2009).
CH4O2 (most likely simultaneous neutral loss of CH2 and H2O2) fol-
As expected, the AI values (not listed in Table 3) calculated
lowing CID yielding fragment ion with the nominal m/z 253
were < 0.5 thereby confirming absence of aromatic products (Bateman
(C14H21O4).
et al., 2009). DBE values and O:C ratios indicate that products of re-
Based on the data acquired, the mechanism of reaction (1) was
action (1) are more oxidized than BCA. Products detected were also
proposed and it is presented in Fig. 5.
eluted from the RP C18 column before BCA thereby confirming the
As presented in Fig. 5 reaction (1) most likely proceeds mainly via
formation of more oxidized compounds (Fig. 3).
OH addition to the terminal C=C bond with some contribution from
HR-MS/MS spectra of compound I and compounds III are shown in
the abstraction pathway (Renard et al., 2013; Schöne et al., 2014).
Fig. 4.
As previously proposed, keto-BCA (compound I) is most likely
As shown in Fig. 4, the HR-MS/MS spectra provided structural

234
B. Witkowski, et al. Atmospheric Environment 213 (2019) 231–238

Table 3
LC/HR-MS analysis of the products of reaction (1) detected in the negative ionization mode.
Compound Retention time (min)a Elemental composition Nominal mass (Da) b
Δ (mmu) DBE O:C H:C Major neutral lossesc

BCA 24.9 C15H24O3 252 −0.168 4 0.2 1.6 CO2


I 13.8 C14H22O4 254 −0.402 4 0.3 1.6 H2O, CO2, H2O and CO2
II 12.3 C15H24O5 284 −0.263 4 0.3 1.6 H2O, 2H2O, 3 H2O, CO2
III 10.3 C15H26O6 302 −0.102 4 0.3 1.6 H2O2 and CH2, H2O, CO2, H2O and CO2
IV 9.4 C15H26O5 286 −0.747 3 0.3 1.7 H2O, 2H2O, CO2
V 8.5 C14H22O6 286 −0.062 4 0.4 1.6 H2O, 2H2O, 3H2O, CO2
VI 7.8 C14H22O6 286 −0.662 5 0.4 1.4 H2O, 2H2O, 3H2O, CO2
VII 7.3 C9H14O4 184 0.118 3 0.4 1.6 H2O, CO2, C3H6O, C4H6O
VIII 2.2 C4H6O4 118 -.632 2 1 1.5 H2O
IX 1.9 C5H8O4 132 −0.582 2 0.8 1.6 H2O, CO2

a
Retention time of the major isomer (largest peak) is listed for the unseparated and/or multiple peaks.
b
The difference between m/z value measured and the molecular weight from the assigned formula.
c
Neutral molecules eliminated from the parent ion and/or from the major fragment ions.

formed via the addition pathway involving formation of tetroxide by- also formed via the abstraction pathway as previously proposed for
products (Renard et al., 2013; Schöne et al., 2014; Witkowski et al., LA + OH reaction (Witkowski et al., 2018b). Only one possible
2018b; Zhang et al., 2010). Compounds III were tentatively identified pathway for the abstraction mechanism is presented in Fig. 5 for clarity.
as hydroxy hydroperoxy derivatives of BCA based on the HR-MS/MS Note that the gas-phase SAR parameters indicate that the H atom ab-
data acquired (Fig. 4). Formation of the analogous hydroxy hydroper- straction from BCA by OH is most likely to occur in positions 3 and 5
oxides was also reported for the aqueous-phase oxidation of methyl- due to formation of the tertiary alkyl radicals (Renard et al., 2013;
vinyl ketone by OH (Renard et al., 2013). Two isomers of compound III Schöne et al., 2014). After hydrogen atom abstraction from BCA there
can be formed according to the mechanism proposed what is in a good are two possible pathways for cyclobutyl ring opening, leading to the
agreement with the results presented in Fig. S9 (two chromatographic formation of the two isomers of compound II. Analogous mechanism of
peaks were observed for compound III). the cyclobutyl ring opening following hydrogen atom abstraction by
Compounds II and IV are most likely formed after H2O2 elimination OH was recently proposed for cis-pinonic acid (Enami and Sakamoto,
(compound II) and following O2 elimination/disproportionation (com- 2016a; Müller et al., 2012). Consequently, several isomers of compound
pounds II and VI) from the tetroxide by-products (Schöne et al., 2014; II can be formed via the abstraction pathway and this conclusion is also
Zhang et al., 2010). confirmed by the LC/MS analysis as several unseparated chromato-
Temporal profiles obtained for compounds I, II (the major isomer – graphic peaks were detected for this product – see Fig. S9. Subse-
see Fig. S10), III – VI indicated that concentrations of all of the major quently, these unsaturated isomers of compound II can undergo rapid
products detected increased right after the onset of the reaction thereby oxidation by OH leading to the formation of compounds VI and V
indicating that they were first-generation products. (Witkowski et al., 2018b). The mechanism proposed is also confirmed
As presented in Fig. 5, unsaturated isomers of compound II can be by the detection of multiple chromatographic peaks for compound II as

Fig. 4. HR-MS/MS fragmentation spectra of compounds I and III (the major products detected) only centroid profiles are shown for clarity.

235
B. Witkowski, et al. Atmospheric Environment 213 (2019) 231–238

Fig. 5. Tentatively proposed structures and formation mechanisms of reaction (1) products.

well as for compounds V and VI as shown in Fig. S9. 4. Conclusions and atmospheric implications
RP-LC/MS is more sensitive towards higher-MW compounds that
are retained by the C18 column used but some weakly retained, lower- Atmospheric lifetimes of BCA due to reaction with OH were calcu-
MW products were also detected (see Table 3). These minor products lated assuming [OH]aq between 1 × 10−14 and 3.5 × 10−15 M,
were identified as highly-oxygenated molecules with a shorter carbon (Herrmann et al., 2015; Otto et al., 2018) and [OH]gas between 2 and
backbone-compounds VII, VIII and IX. It is also reasonable to assume 4 × 106 molecule × cm−3 (Percival and McGillen, 2008). Bimolecular
that fragmentation will eventually dominate at the later stages of oxi- rate coefficient for the gas-phase oxidation of BCA by OH was estimated
dation, thereby resulting in a fragmentation of the original carbon with SAR as 1 × 1010 cm3 × molecule−1 s−1 (Atkinson, 1987; Percival
skeleton and formation of lower-MW molecules, such as oxalic acid and McGillen, 2008). By using the k1 values measured and the esti-
(Ervens et al., 2011; Myriokefalitakis et al., 2011). mated pKa value it was possible to calculate the k1 of BCA under in-
Oxygenated derivatives of BCA were formed following reaction (1) termediate pH conditions (section S5) (Aljawhary et al., 2016;
while production of higher-MW products (oligomers/dimers) was not Witkowski et al., 2018a, 2018b, Marvin).
observed under the experimental conditions used here. Results pre- The gas-phase lifetimes of BCA due to reaction with OH were esti-
sented here as well as the previously published data indicate that mated as 24 min - 1.5 h. In the aqueous-phase, the lifetimes calculated
higher-MW products are not formed following terpenoic acids + OH were between 24 min and 12 h, since the reaction (1) slowed down
reactions and production of oxygenated derivatives of the precursor is under neutral and basic pH-conditions (section 3.1). Taking both un-
favored (Enami and Sakamoto, 2016a; Witkowski and Gierczak, 2017a; certainly and variability of atmospheric OH concentration into account,
Witkowski et al., 2018b; Zhao et al., 2017). (Herrmann et al., 2015; Otto et al., 2018; Percival and McGillen, 2008)
the results obtained indicate that, if there are aqueous particles present,
aqueous-phase oxidation of BCA and similar compounds should be

236
B. Witkowski, et al. Atmospheric Environment 213 (2019) 231–238

relevant in the atmosphere. 1735–1751.


A number of recently published studies, (Aljawhary et al., 2016; Alfarra, M.R., et al., 2012. The effect of photochemical ageing and initial precursor
concentration on the composition and hygroscopic properties of β-caryophyllene
Lignell et al., 2013; Otto et al., 2018) including results reported by our secondary organic aerosol. Atmos. Chem. Phys. 12, 6417–6436.
group, (Witkowski et al., 2018a, 2018b; Witkowski and Gierczak, Aljawhary, D., et al., 2016. Kinetics, mechanism, and secondary organic aerosol yield of
2017a) strongly indicates that aqueous-phase processing of terpenoic aqueous phase photo-oxidation of α-pinene oxidation products. J. Phys. Chem. A 120,
1395–1407.
acids will occur inside the aqueous particles. Moreover, such reactions Atkinson, R., 1987. A structure-activity relationship for the estimation of rate constants
can lead to aqSOA since higher-MW because formation of highly-oxy- for the gas-phase reactions of OH radicals with organic compounds. Int. J. Chem.
genated products was confirmed in bulk reactors (Witkowski et al., Kinet. 19, 799–828.
Atkinson, R., Arey, J., 2003. Gas-phase tropospheric chemistry of biogenic volatile or-
2018a; Witkowski and Gierczak, 2017a) as well as on the gas-liquid ganic compounds: a review. Atmos. Environ. 37, 197–219.
interface (Enami and Sakamoto, 2016a). Bateman, A.P., et al., 2009. Time-resolved molecular characterization of limonene/ozone
Consequently, in the atmosphere, the aqueous-phase processing of aerosol using high-resolution electrospray ionization mass spectrometry. Phys. Chem.
Chem. Phys. 11, 7931–7942.
terpene SOA can result in formation of additional particles that contain
Bonn, B., Moortgat, G.K., 2003. Sesquiterpene ozonolysis: origin of atmospheric new
more oxygenated organic compounds. Forming highly-oxidized SOA particle formation from biogenic hydrocarbons. Geophys. Res. Lett. 30, 1585.
due to aqueous-phase reactions can also result in increased toxicity of Buxton, G.V., et al., 1988. Critical Review of rate constants for reactions of hydrated
the atmospheric particles, as recently observed for the biomass burning electrons, hydrogen atoms and hydroxyl radicals (⋅OH/⋅O− in Aqueous Solution. J.
Phys. Chem. Ref. Data 17, 513–886.
aerosol (Decesari et al., 2017). Carslaw, K.S., et al., 2013. Large contribution of natural aerosols to uncertainty in indirect
However, very little information is currently available about the forcing. Nature 503, 67.
Dalcanale, E., Montanari, F., 1986. Selective oxidation of aldehydes to carboxylic acids
aqSOA formation from terpenoic acids oxidation as recently pointed out
with sodium chlorite-hydrogen peroxide. J. Org. Chem. 51, 567–569.
by Otto et al. (2018) A more quantitative approach is needed in the Decesari, S., et al., 2017. Enhanced toxicity of aerosol in fog conditions in the Po Valley,
future to better describe formation of aqSOA following oxidation of Italy. Atmos. Chem. Phys. 17, 7721–7731.
terpenoic acids under realistic atmospheric conditions. Due to lack of Ding, X., et al., 2014. Spatial distributions of secondary organic aerosols from isoprene,
monoterpenes, β-caryophyllene, and aromatics over China during summer. J.
standards, measuring the yields for every product formed following Geophys. Res.: Atmosphere 119 (11), 811–891 877.
terpenoic acids oxidation is challenging (Enami and Sakamoto, 2016b; Duhl, T.R., et al., 2008. Sesquiterpene emissions from vegetation: a review.
Witkowski and Gierczak, 2017b; Witkowski et al., 2018c). One of the Biogeosciences 5, 761–777.
Enami, S., Sakamoto, Y., 2016a. OH-radical oxidation of surface-active cis-pinonic acid at
possible approaches recently developed involves nebulizing the reac- the air–water interface. J. Phys. Chem. A 120, 3578–3587.
tion mixture followed by the analysis of the particle size-distribution Enami, S., Sakamoto, Y., 2016b. OH-radical oxidation of surface-active cis-pinonic acid at
and their chemical composition (Aljawhary et al., 2016). A simpler, the air–water interface. J. Phys. Chem. A 120, 3578–3587.
Ervens, B., et al., 2011. Secondary organic aerosol formation in cloud droplets and aqu-
alternative that might be consider in the future would be analysis of the
eous particles (aqSOA): a review of laboratory, field and model studies. Atmos. Chem.
total organic carbon (TOC) or drying aliquots of the reaction mixture to Phys. 11, 11069–11102.
simulate evaporation of the water droplets in the atmosphere. Faiola, C.L., et al., 2018. Terpene composition complexity controls secondary organic
aerosol yields from Scots pine volatile emissions. Sci. Rep. 8, 3053.
Faust, J.A., et al., 2017. Role of aerosol liquid water in secondary organic aerosol for-
Declaration of interest statement mation from volatile organic compounds. Environ. Sci. Technol. 51, 1405–1413.
Fuzzi, S., et al., 2015. Particulate matter, air quality and climate: lessons learned and
The authors have no competing interests to declare. future needs. Atmos. Chem. Phys. 15, 8217–8299.
Gray Bé, A., et al., 2017. Cloud activation potentials for atmospheric α-pinene and β-
caryophyllene ozonolysis products. ACS Cent. Sci. 3, 715–725.
Acknowledgments Grieshop, A.P., et al., 2007. Is the gas-particle partitioning in alpha-pinene secondary
organic aerosol reversible? Geophys. Res. Lett. 34.
Hellén, H., et al., 2018. Long-term measurements of volatile organic compounds highlight
This project was founded by the Polish National Science Centre: the importance of sesquiterpenes for the atmospheric chemistry of a boreal forest.
grant number 2014/13/B/ST4/04500. We thank the Structural Atmos. Chem. Phys. 18, 13839–13863.
Research Laboratory (SRL) at the Department of Chemistry of Helmig, D., et al., 2007. Sesquiterpene emissions from pine trees − identifications,
emission rates and flux estimates for the contiguous United States. Environ. Sci.
University of Warsaw for making LC/MS measurements possible. SRL
Technol. 41, 1545–1553.
was established with financial support from European Regional Herrmann, H., et al., 2015. Tropospheric aqueous-phase chemistry: kinetics, mechanisms,
Development Found, project no: WPK_1/1.4.3./1/2004/72/72/165/ and its coupling to a changing gas phase. Chem. Rev. 115, 4259–4334.
Jaoui, M., et al., 2007. β-caryophyllinic acid: an atmospheric tracer for β-caryophyllene
2005/U. The study was carried out at the Biological and Chemical
secondary organic aerosol. Geophys. Res. Lett. 34.
Research Centre, University of Warsaw, established within the project Jaoui, M., et al., 2013. Secondary organic aerosol formation from the oxidation of a series
co-financed by European Union from the European Regional of sesquiterpenes: α-cedrene, β-caryophyllene, α-humulene and α-farnesene with O3,
Development Fund under the Operational Programme Innovative OH and NO3 radicals. Environ. Chem. 10, 178–193.
Krechmer, J.E., et al., 2015. formation of low volatility organic compounds and secondary
Economy, 2007–2013. Bartlomiej Witkowski was also funded by the by organic aerosol from isoprene hydroxyhydroperoxide low-NO oxidation. Environ. Sci.
the Polish-Us Fulbright Commission, Senior Award Scholarship for the Technol. 49, 10330–10339.
academic year 2018/2019. We thank mgr Anna Nowak for helping us Kundu, S., et al., 2017. Molecular formula composition of β-caryophyllene ozonolysis
SOA formed in humid and dry conditions. Atmos. Environ. 154, 70–81.
with the synthesis of BCA and LA. We thank dr Dagmara Tymecka for Lee, A., et al., 2006. Gas-phase products and secondary aerosol yields from the ozonolysis
BCA and LA purification with the semi-preparative HPLC. We thank dr of ten different terpenes. J. Geophys. Res.: Atmosphere 111.
Marcin Wilczek for 1H NMR spectra of BCA and LA. We thank dr Julio Leungsakul, S., et al., 2005. Kinetic mechanism for predicting secondary organic aerosol
formation from the reaction of d-limonene with ozone. Environ. Sci. Technol. 39,
Torres for the LC/-HR/MS/MS analyses. We thank the anonymous re- 9583–9594.
viewers for very helpful, constructive and insightful comments and Lignell, H., et al., 2013. Experimental and theoretical study of aqueous cis-pinonic acid
suggestions. photolysis. J. Phys. Chem. A 117, 12930–12945.
Lim, Y.B., et al., 2010. Aqueous chemistry and its role in secondary organic aerosol (SOA)
formation. Atmos. Chem. Phys. 10, 10521–10539.
Appendix A. Supplementary data Marvin v17.28.0 Developed by ChemAxon Was Used to Calculate pKa of BCA, Website.
https://chemaxon.com/products/marvin.
McNeill, V.F., 2015. Aqueous organic chemistry in the atmosphere: sources and chemical
Supplementary data to this article can be found online at https://
processing of organic aerosols. Environ. Sci. Technol. 49, 1237–1244.
doi.org/10.1016/j.atmosenv.2019.06.016. Messina, P., et al., 2016. Global biogenic volatile organic compound emissions in the
ORCHIDEE and MEGAN models and sensitivity to key parameters. Atmos. Chem.
References Phys. 16, 14169–14202.
Müller, L., et al., 2012. Formation of 3-methyl-1,2,3-butanetricarboxylic acid via gas
phase oxidation of pinonic acid – a mass spectrometric study of SOA aging. Atmos.
Chan, M.N., et al., 2011. Influence of aerosol acidity on the chemical composition of Chem. Phys. 12, 1483–1496.
secondary organic aerosol from β-caryophyllene. Atmos. Chem. Phys. 11, Myriokefalitakis, S., et al., 2011. In-cloud oxalate formation in the global troposphere: a

237
B. Witkowski, et al. Atmospheric Environment 213 (2019) 231–238

3-D modeling study. Atmos. Chem. Phys. 11, 5761–5782. direct forcing on climate. Curr. Clim. Chang. Rep. 4, 84–98.
Nguyen, T.L., et al., 2009. The gas-phase ozonolysis of [small beta]-caryophyllene van Eijck, A., et al., 2013. New tracer compounds for secondary organic aerosol formation
(C15H24). Part II: a theoretical study. Phys. Chem. Chem. Phys. 11, 4173–4183. from β-caryophyllene oxidation. Atmos. Environ. 80, 122–130.
Otto, T., et al., 2017. Tropospheric aqueous-phase oxidation of isoprene-derived dihy- Willand-Charnley, R., et al., 2012. Pyridine is an organocatalyst for the reductive ozo-
droxycarbonyl compounds. J. Phys. Chem. A 121, 6460–6470. nolysis of alkenes. Org. Lett. 14, 2242–2245.
Otto, T., et al., 2018. Aqueous-phase oxidation of terpene-derived acids by atmo- Winterhalter, R., et al., 2009. The gas-phase ozonolysis of [small beta]-caryophyllene
spherically relevant radicals. J. Phys. Chem. A 122, 9233–9241. (C15H24). Part I: an experimental study. Phys. Chem. Chem. Phys. 11, 4152–4172.
Percival, C., McGillen, M., 2008. Overview of Structure-Activity Relationship Methods for Witkowski, B., Gierczak, T., 2017a. cis-Pinonic acid oxidation by hydroxyl radicals in the
Predicting Gas-Phase Rate Coefficients. Springer Netherlands, Dordrecht, pp. 47–59. aqueous phase under acidic and basic conditions: kinetics and mechanism. Environ.
Renard, P., et al., 2013. Radical mechanisms of methyl vinyl ketone oligomerization Sci. Technol. 51, 9765–9773.
through aqueous phase OH-oxidation: on the paradoxical role of dissolved molecular Witkowski, B., Gierczak, T., 2017b. cis-Pinonic acid oxidation by hydroxyl radicals in the
oxygen. Atmos. Chem. Phys. 13, 6473–6491. aqueous phase under acidic and basic conditions: kinetics and mechanism. Environ.
Richters, S., et al., 2015. Gas-phase rate coefficients of the reaction of ozone with four Sci. Technol. 51, 9765–9773.
sesquiterpenes at 295 ± 2 K. Phys. Chem. Chem. Phys. 17, 11658–11669. Witkowski, B., et al., 2018a. Kinetics of limonene secondary organic aerosol oxidation in
Sareen, N., et al., 2017. Potential of aerosol liquid water to facilitate organic aerosol the aqueous phase. Environ. Sci. Technol. 52, 11583–11590.
formation: assessing knowledge gaps about precursors and partitioning. Environ. Sci. Witkowski, B., et al., 2018b. Limononic acid oxidation by hydroxyl radicals and ozone in
Technol. 51, 3327–3335. the aqueous phase. Environ. Sci. Technol. 52, 3402–3411.
Schöne, L., et al., 2014. Atmospheric aqueous phase radical chemistry of the isoprene Witkowski, B., et al., 2018c. Limononic acid oxidation by hydroxyl radicals and ozone in
oxidation products methacrolein, methyl vinyl ketone, methacrylic acid and acrylic the aqueous phase. Environ. Sci. Technol. 52, 3402–3411.
acid – kinetics and product studies. Phys. Chem. Chem. Phys. 16, 6257–6272. Zhang, X., et al., 2010. Laboratory simulation for the aqueous OH-oxidation of methyl
Shrivastava, M., et al., 2017. Recent advances in understanding secondary organic vinyl ketone and methacrolein: significance to the in-cloud SOA production. Atmos.
aerosol: implications for global climate forcing. Rev. Geophys. 55, 509–559. Chem. Phys. 10, 9551–9561.
Shu, Y., Atkinson, R., 1995. Atmospheric Lifetimes and Fates of a Series of Sesquiterpenes. Zhang, H., et al., 2018. Monoterpenes are the largest source of summertime organic
Sindelarova, K., et al., 2014. Global data set of biogenic VOC emissions calculated by the aerosol in the southeastern United States. Proc. Natl. Acad. Sci. Unit. States Am. 115,
MEGAN model over the last 30 years. Atmos. Chem. Phys. 14, 9317–9341. 2038–2043.
Tasoglou, A., Pandis, S.N., 2015. Formation and chemical aging of secondary organic Zhao, R., et al., 2017. Rapid aqueous-phase photooxidation of dimers in the α-pinene
aerosol during the β-caryophyllene oxidation. Atmos. Chem. Phys. 15, 6035–6046. secondary organic aerosol. Environ. Sci. Technol. Lett. 4, 205–210.
Tsigaridis, K., Kanakidou, M., 2018. The present and future of secondary organic aerosol

238

You might also like