Detection of nitrated and oxygenated polycyclic aromatic hydrocarbons using atmospheric pressure chemical ionization high resolution mass spectrometry

Article abstract of DOI:10.1016/j.ijms.2016.01.001

Polycyclic aromatic hydrocarbons (PAHs) are common pollutants in the atmosphere and have long been recognized to be toxic to humans. These PAHs can be oxidized into more toxic products in both the gas and condensed (on the surface of suspended particulate matter) phases. In this work, we report fragmentation patterns observed using atmospheric pressure chemical ionization with high resolution mass spectrometry (APCI-HRMS) of PAH oxidation products. A representative group of 18 PAH derivatives containing carbonyl (oxo-PAHs), hydroxyl (hydroxy-PAHs), carboxyl (carboxy-PAHs), and/or nitro (nitro-PAHs) groups were studied. Ionization of carboxy-PAHs in negative mode yielded common fragments of [M-H]^-, [M-H-CO]^-, and [M-H-CO₂]^-. Oxo-PAHs provided common fragments of [M+H]⁺, [M+H-CO]⁺ and [M+H-2CO]⁺ in positive mode and [M+e^-]^- in negative mode. Mass spectra of aldehydes exhibited common fragments of [M+H]⁺, [M+H-CO]⁺, and [M+H-O]⁺ in positive mode and [M+e^-]^- and [M-H+O]^- in negative mode. For all nitro-PAHs, [M+H]⁺, [M+H-O]⁺ and [M+3H-O₂]⁺ ions were observed in positive mode. On the basis of the APCI-HRMS analysis of standards, eight reaction products of pyrene oxidation under heterogeneous conditions were characterized. HRMS data, specific fragments and common ions such as that of 205 m/z, characteristic for carbonyl phenanthrene, enabled the identification of the oxidation products.

Full text of DOI:10.1016/j.ijms.2016.01.001

G Model
MASPEC155491–12
ARTICLE IN PRESS
International Journal of Mass Spectrometry xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Mass Spectrometry
journal homepage: www.elsevier.com/locate/ijms
1
2
3
Detection of nitrated and oxygenated polycyclic aromatic
hydrocarbons using atmospheric pressure chemical ionization high
resolution mass spectrometry
Q1
1
∗
4
Richard E. Cochran , Irina P. Smoliakova, Alena Kubátová
5
6
University of North Dakota, Department of Chemistry, 151 Cornell St., Stop 9024, Grand Forks, ND 58202, USA
7
a r t i c l e i n f o
a b s t r a c t
8
9
1
1
1
1
1
Article history:
Polycyclic aromatic hydrocarbons (PAHs) are common pollutants in the atmosphere and have long been
recognized to be toxic to humans. These PAHs can be oxidized into more toxic products in both the
gas and condensed (on the surface of suspended particulate matter) phases. In this work, we report
fragmentation patterns observed atmospheric pressure chemical ionization with high resolution mass
spectrometry (APCI–HRMS) for PAH oxidation products. A representative group of 18 PAH derivatives
containing carbonyl (oxo-PAHs), hydroxyl (hydroxy-PAHs), carboxyl (carboxy-PAHs), and/or nitro (nitro-
PAHs) groups were studied. Ionization of carboxy-PAHs in negative mode yielded common fragments of
0
1
2
3
4
Received 13 August 2015
Received in revised form
29 December 2015
Accepted 14 January 2016
Available online xxx
1
1
1
1
1
2
2
5
6
7
8
9
0
1
Keywords:
Polycyclic aromatic hydrocarbon
derivatives
High resolution mass spectrometry
Ionization/fragmentation patterns
Identification of unknowns
−
−
−
+
+
[
M−H] , [M−H−CO] , and [M−H−CO2] . Oxo-PAHs provided common fragments of [M+H] , [M+H−CO]
+
− −
and [M+H−2CO] in positive mode and [M+e ] in negative mode. Mass spectra of aldehydes exhibited
+
+
+
−
−
−
common fragments of [M+H] , [M+H−CO] , and [M+H−O] in positive mode and [M+e ] and [M−H+O]
+ + +
in negative mode. For all nitro-PAHs, [M+H] , [M+H−O] and [M+3H−O2] ions were observed in positive
mode. On the basis of the APCI–HRMS analysis of standards, eight reaction products of pyrene oxida-
tion under heterogeneous conditions were characterized. HRMS data, specific fragments and common
ions such as that of 205 m/z, characteristic for carbonyl phenanthrene, enabled the identification of the
oxidation products.
©
2016 Published by Elsevier B.V.
2
3
1. Introduction
measure the contribution of primary and secondary emissions [23].
For example, 1-nitropyrene and 3-nitrofluoranthene are formed
through incomplete combustion of diesel fuel, while 2-nitropyrene
and 2-nitrofluoranthene are produced during gas-phase reactions
in the atmosphere [9,21,24–26].
In addition to identifying unknown PAH oxidation products,
the determination of their concentrations in the atmosphere is
also of importance. In order to obtain reliable concentration mea-
surements, versatile and robust analytical methods enabling their
effective separation and identification are required. Owing to its
high resolution, gas chromatography–mass spectrometry (GC–MS)
often turns out to be the method of choice. However, higher molec-
ular weight PAHs with five or more rings typically exhibit poor
GC sensitivity due to their low volatility [27]. Similarly, species
that are either thermally labile or contain highly polar functional
groups (i.e., hydroxy- and carboxy-PAHs) are not easily detected.
In addition, polar analytes require derivatization prior to analysis
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
2
2
2
2
2
2
3
3
3
3
3
3
3
4Q2
Polycyclic aromatic hydrocarbons (PAHs) are currently recog-
nized for their adverse effect on human health [1]. PAH derivatives
produced during incomplete combustion or as a result of atmo-
spheric reactions were previously shown to be even more toxic
than their native PAHs [2]. The major PAH derivatives previously
reported are those functionalized with carbonyl (oxo-PAHs), nitro
(nitro-PAHs), hydroxyl (hydroxy-PAHs), and carboxyl (carboxy-
PAHs) groups [3–16]. Nevertheless, due to the difficulty in their
analysis, PAH derivatives have been reported to a lesser extent than
their PAH precursors, with studies targeting either selected classes
of compounds or specific chemicals [12,17–22]. The identification
of specific compounds, particularly isomers, may provide further
insights into the mechanisms of atmospheric processes and help
5
6
7
8
9
0
1
2
3
4
5
6
[
28,29].
For these compounds, high performance liquid chromatography
HPLC) coupled to MS has been successfully applied as an alterna-
∗ Corresponding author. Tel.: +1 701 777 0348; fax: +1 701 777 2331.
E-mail address: alena.kubatova@UND.edu (A. Kubátová).
Current address: E331 Chemistry Building, University of Iowa, Iowa City, IA
(
1
tive method with optimized HPLC conditions for known standards
5
2242, USA.
http://dx.doi.org/10.1016/j.ijms.2016.01.001
387-3806/© 2016 Published by Elsevier B.V.
1
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
2
5
5
6
6
6
6
6
6
6
6
6
6
7
7
7
7
7
7
7
7
7
7
8
8
8
8
8
8
8
8
8
8
9
9
9
9
9
9
9
9
9
9
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
8
9
0
1
2
3
4
5
6
7
8
9
0
[17–20,30,31]. From the two commonly used ionization sources,
atmospheric pressure chemical ionization (APCI) and electrospray
ionization (ESI), the former was demonstrated to be more sensitive
for the analysis of PAH derivatives [18]. Major MS ions of several
PAH derivatives analyzed in the present work were described in
previous HPLC–APCI–MS studies [18–20]; however, fragmentation
pathways were reported for only a few of these species [11,30]. Fur-
thermore, these studies did not focus on confirming if the observed
ions were either molecular ions or fragments.
identities of these species using high resolution mass spectrometry
(HRMS) are yet to be confirmed.
124
125
126
127
128
129
130
131
132
133
134
135
136
137
In the present study, ionization and fragmentations that occur
during APCI of common atmospheric oxidation products of PAHs,
namely derivatives containing either nitro, amino, carbonyl, car-
boxyl, or hydroxyl groups, or a combination thereof, were evaluated
using HRMS. APCI was coupled with HPLC, which has been reported
to be more suitable for some classes of PAH derivatives than GC.
For each class of PAH derivatives, fragmentation pathways were
proposed based on the ions and fragments detected. The ioniza-
tion and fragmentation trends observed for standards were then
employed to identify unknown pyrene oxidation products formed
in flow reactor experiments involving the ozonation and nitration
of PAHs.
Using low resolution MS methods, limited data on the fragmen-
tation of PAH derivatives with HPLC–APCI have been reported for
only a small number of targeted species. For hydroxy-PAHs, the
•
−
main ions observed in negative mode are the molecular anion [M]
−
1Q3 and the deprotonated molecule [M−H] [29]. In positive mode, pro-
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
tonation and the subsequent loss of H O were the main ionization
2
and fragmentation processes previously observed [29]. Addition-
2. Materials and methods
138
139
+
ally, [M+15] adducts have been detected and tentatively identified
+
as [M+CH ] .
2.1. Chemicals
3
Letzel et al. analyzed two carboxy-PAH species, 2-naphthoic
acid and 2-anthraquinone acid, using HPLC–APCI with low resolu-
tion MS detection [31]. In negative APCI mode, the most common
The PAH derivatives including 9-nitroanthracene, 1-
nitropyrene, 1,6-dinitropyrene, anthrone, 9,10-anthracenedione,
140
141
142
143
144
145
146
147
148
149
150
151
152
−
ions resulted from deprotonation ([M−H] ions) followed by decar-
9,10-phenanthrenedione,
pyrene-4,5-dione, 9-phenanthrenecarboxaldehyde,
1,4-phenanthrenedione,
and 4-
−
boxylation ([M−H−CO ] anions) [31]. In addition, a fragment
2
−
[M−H−28−28] was observed and proposed to form by either
carboxy-5-phenanthrenecarboxaldehyde were obtained from
Sigma Aldrich (Atlanta, GA, USA). 9-Nitrophenanthrene, 3-
nitrophenanthrene and 9,10-dinitroanthracene were obtained
from Accustandard Inc. (New Haven, CT, USA). See Fig. 1 for the
structures of all PAH derivatives investigated in this work. LC–MS
Optima grade methanol (MeOH) and LC–MS Optima grade ace-
tonitrile were purchased from Fisher Scientific (Chicago, IL, USA).
Formic acid (LC–MS grade) was obtained from Fluka (Atlanta, GA,
USA). MilliQ water (Millipore) was used during HPLC experiments.
two CO (28 Da) losses or consecutive losses of CO and C H
2
4
(28 Da). When using higher fragmentor voltages (prompting col-
lision induced dissociation and focusing the ion beam into the MS
−
−
analyzer), fragments [M−H−44−18] and [M−H−44−28] were
also observed. Their formation was explained as the losses of CO
2
followed by H O and CO, respectively. As for hydroxy-PAHs, the
2
identities of these fragments have not been confirmed using high
resolution data. Fragmentations of carboxy-PAHs in positive mode
were not observed, mainly due to their low proton affinities and,
therefore, a limited degree of protonation [31].
2.2. HPLC–APCI–MS analysis
153
Many research groups utilized HPLC–APCI–MS for detecting
oxo-PAHs [17–20,30]. However, in the majority of these studies,
only the most abundant ion was used during method optimization,
with common fragments reported to a limited extent. In positive
HPLC–MS analyses were carried out with an Agilent 1100 HPLC
system coupled to a high resolution Agilent 1969 Time-of-flight
MS (ToF-MS) equipped with an APCI source (Agilent Technologies,
Santa Clara, CA, USA).
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
+
+
mode, the losses of CO ([M+H−CO] ) and C O ([M+H−CO−CO] )
2
2
7Q4 from the protonated molecule were reported [20]. In negative
8
All HPLC separations were performed using a Restek C18
200 mm × 3.2 mm reverse phase HPLC column with 5 ␮m particle
size (Restek, Bellefonte, PA, USA). A binary solvent system consist-
ing of water (A) and methanol (B) was used. A gradient program
mode, C O loss from the deprotonated molecule (giving the
2 2
−
− −
9
[M−H−CO−CO] ion) and electron capture (giving the [M+e ]
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
ion) were the most common pathways [18,19]. In these stud-
ies, only single isomers of oxo-PAH derivatives were subjected to
MS analysis; thus fragmentation processes specific for each con-
stitutional isomer have yet to be identified. Another important
information reported by Delhomme et al. was increased sensitivity
for diketones in negative mode and for monoketones in positive
mode [18].
−
1
at a flow rate of 0.2 mL min
started with 20% B for 5 min, fol-
lowed with a linear increase to 90% B at 20 min, and hold at 90%
until 27 min, and then was linearly decreased to 20% at 30 min and
held at 20% B for 5 min to allow for equilibration. The column oven
◦
temperature was set at 30 C and injection volume was 50 ␮L.
APCI was performed in both positive and negative modes with
HPLC–APCI–MS methods for analysis of nitro-PAHs were uti-
lized in several studies; however, as with oxo-PAHs, most of the
reports were focused only on the quantification using the major
ions. The most frequently observed pattern for nitro-PAHs in spec-
tra obtained in positive APCI mode was the neutral loss of 30 Da.
The identity of this neutral fragment has been up for debate, being
attributed to either the loss of NO[32] or nitro group reduction
(i.e., losing 2O and gaining 2H, giving a net loss of 30 Da) [33].
Using deuterium oxide in place of water for the eluent solvent mix-
each sample containing 5 mM formic acid. Drying gas (N ) was set
2
◦
−1
at 300 C at a flow of 3 L min . For all experiments the capillary
voltage was set to 4500 V. In order to minimize the contribution
of post-source fragmentation, the fragmentor voltage was set to
120 V for all experiments. All HPLC–APCI–ToF-MS analyses were
performed with the corona discharge current set at 10 ␮A. For
experiments evaluating the contribution of the corona current to
gas-phase ion fragmentation, the corona discharge current was var-
ied within the range of 4–25 ␮A.
+
ture, Karancsi and Slegel showed that the fragment of [M−30] was
formed for singly substituted one- or two-ring containing nitroaro-
matic species as a result of nitro group reduction [33]. For three-ring
2.3. Reaction experiments
177
+
nitro-PAHs, the identity of the [M−30] fragment has not been
The flow reactor used for the ozonation of pyrene consisted of
three main parts: a gas injection/dilution system, mixing chamber,
and reaction chamber (Fig. S1). The gas injection/dilution system
delivered breathing quality air to a mixing chamber composed
178
179
180
181
182
confirmed. In negative mode, the most common ionization path-
way was also the loss of 30 Da proposed as the loss of NO based
−
on low resolution data [11,30,34,35]. Also, the [M−H+16] adduct
ꢀ
ꢀ
was observed and interpreted as the gain of oxygen [11,35]. The
entirely of Teflon (31.5 cm length × 9 cm I.D.) through ¼ stainless
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
3
Fig. 1. Structures of PAH derivatives studied in this work.
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
steel tubing. Two-way stainless steel valves were used for select-
ing the gases used in each experiment. All gas flows through the
reactor system were regulated with mass flow controllers (Alicat
Scientific, Tucson, AZ, USA) to achieve the desired dilution of gases.
After passing the mixing chamber, the gas flow was split using
Upon the completion of each reaction experiment, the filter was Q5 211
extracted using sonication with 10 mL of DCM for 30 min and the
resulting extract was collected. The vial and filter were washed
twice with 10 mL of DCM each time. Ultimately all DCM extract
fractions obtained were combined, filtered over deactivated glass
wool, and evaporated under a gentle stream of nitrogen to ∼0.5 mL.
The samples were then submitted to HPLC–APCI–ToF-MS analy-
sis.
Identification was based on a match of retention times of the
overlayed extracted ion chromatogram (EIC) peaks (± 0.03 Da), on
the relative intensities as well as high resolution MS data.
212
213
214
215
216
217
218
219
220
221
stainless steel tees, leading to the reaction chamber, an O gas ana-
3
lyzer, and an exhaust vent. During ozonation experiments, ozone
levels at the outlet of the mixing chamber were measured using
a photometric O3 gas analyzer (Model 400E, Teledyne, Thousand
Oaks, CA, USA). The gas mixtures from the mixing chamber were
supplied to the Teflon reaction chamber (43 cm in length × 9 cm
I.D.) through an inlet located on top of the chamber. A quartz tis-
sue filter (90 mm I.D., PALL Corporation, Port Washington, NY) was
spiked with pyrene solution and placed onto a Teflon-coated alu-
minum mesh at the bottom of the reaction chamber. The outlet of
the reactor chamber was located after the filter support. The total
flow through the reaction chamber was controlled using an oil-less
pump with a mass flow controller positioned between the reaction
chamber and the pump. The flow drawn by the pump was always
kept lower than the gas flow exiting the mixing chamber in order
to ensure that ambient air was not drawn through the flow splitter
(installed to prevent over-pressurization of the reaction chamber).
3. Results and discussion
222
223
3.1. Carboxy- & hydroxy-PAHs
The four carboxy-PAHs and one hydroxy-PAH investigated in
this study are presented in Table 1 and Fig. 1, top row. Three
of them, 9-phenathrene-, 4-phenathrene-, 4-anthracenecarboxylic
acids, are constitutional isomers that differ either in the ring
arrangement or the location of the carboxyl group. The fourth com-
pound, 4-carboxy-5-phenathrenecarboxaldehyde, contains both
carboxyl and carbonyl groups.
As previously reported [31], the APCI in negative mode was more
sensitive and resulted in more uniform fragmentation patterns
(Table 1; spectra shown in Fig. S2). The fragmentation proposed
in Scheme 1 corresponds to neutral losses previously observed by
224
225
226
227
228
229
230
231
232
233
234
An allotted amount of
a PAH stock solution (40 ␮L) at
−
1
100 ␮g mL per compound was spiked over the quartz tissue filter
and allowed to evaporate at room temperature for 60 s (resulting
in a final amount of 4 ␮g on the filter). All quartz filters were pre-
◦
baked at 500 C for 12 h, then placed in the flow reactor and exposed
to O3 at 1.25 ppm for 90 min.
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

Table 1
Pseudomolecular ions and fragments observed for hydroxy- and carboxy-PAHs with HPLC–APCI–HRMS in both positive and negative modes.
Compound
tR a
Positive polarity
Negative polarity
(min)
Measured massb (Da)
Rel.
abund.
Fragmentc
Calculated
empirical
formula
Calculated
mass (Da)
Mass
error
(ppm)
Measured massb (Da)
Rel.
abund.
(%)
Fragmentc
Calculated
empirical
formulad
Calculated
mass (Da)
Mass error
(ppm)e
d
e
(%)
+
−
4
-Carboxy-5-
18.55
235.07633 ± 0.00075
205.06540 ± 0.00084
100
95
92
85
75
65
50
[M+H−O]
C16H11O2
C15H9O
C16H11O3
C17H13O3
C15H11
235.07536
205.06479
251.07027
265.08590
191.08550
233.05971
177.06988
4.1 ± 3.2
3.0 ± 4.1
0.2 ± 3.4
4.7 ± 4.6
−1.4 ± 3.4
0.1 ± 3.9
−3.2 ± 3.8
177.07074 ± 0.00014
249.05509 ± 0.00023
205.06497 ± 0.00047
100
90
40
[M−H−C2O3]
C14H9
C16H9O3
C15H9O
177.07097
249.05572
205.06589
−1.3 ± 0.8
−2.5 ± 0.9
−4.5 ± 2.3
+
−
phenanthrenecarboxaldehyde
[M+H−CH2O2]
[M−H]
−
2
2
1
2
1
51.07032 ± 0.00087
65.08714 ± 0.00123
91.08524 ± 0.00065
33.05973 ± 0.00091
77.06931 ± 0.00067
[M+H]+
[M−H−CO2]
[M+H+CH2]+
+
[M+H−CO3]
+
[M+H−H2O]
C16H9O2
C14H9
+
[M+H−C2H2O3]
[M+H]+
C16H11O
C16H10O
C16H9O2
C16H10
219.08044
218.07262
233.05971
202.07770
13.9 ± 7.2
11.8 ± 5.6
7.3 ± 2.4
217.06663 ± 0.00098
232.05210 ± 0.00103
100
20
[M−H]
−
C16H9O
C16H8O2
217.06589
232.05188
3.4 ± 4.5
0.9 ± 4.4
1
9
4
-Hydroxypyrene
19.21
18.34
18.56
219.08350 ± 0.00158
100
85
75
2
2
2
2
18.07520 ± 0.00122
33.06141 ± 0.00055
02.07679 ± 0.00046
[M−e ]+
−
[M−2H+O]
−
+
[M+O−H]
—
+
[M−e −O]
−4.5 ± 2.3
+
-Phenanthrenecarboxylic acid
-Phenanthrenecarboxylic acid
209.05907 ± 0.00182
100
25
20
[M+H−CH2]
C14H9O2
C16H13O2
C15H11O2
C15H9O
209.05971
237.09101
223.07536
205.06479
−3.0 ± 8.7
−8.7 ± 2.4
11.1 ± 2.4
7.2 ± 4.2
221.06048 ± 0.00054
177.07080 ± 0.00083
193.06566 ± 0.00048
208.05256 ± 0.00044
100
25
5
[M−H]−
C15H9O2
C14H9
C14H9O
C14H8O2
221.05971
177.06988
193.06479
208.05188
3.5 ± 2.4
5.2 ± 4.7
4.5 ± 2.5
3.3 ± 2.1
+
−
2
2
2
37.08894 ± 0.00057
23.07783 ± 0.00054
05.06627 ± 0.00086
[M+H+CH2]
[M+H]+
[M−H−CO2]
−
[M−H−CO]
+
−
10
[M+H−H2O]
4
[M−CH2]
[M−H−2e ]+
−
C15H9O2
C15H9O
C16H13O2
C15H11O2
221.05971
205.06479
237.09101
223.07536
−2.3 ± 0.5
−0.1 ± 0.3
−1.7 ± 2.5
−1.6 ± 0.5
221.06036 ± 0.00049
177.07064 ± 0.00042
235.03951 ± 0.00045
193.06484 ± 0.00088
100
40
5
4
2
[M−H]
−
C15H9O2
C14H9
C15H7O3
C14H9O
C14H8O2
221.05971
177.06988
235.03897
193.06479
208.05188
3.0 ± 2.2
4.3 ± 2.4
2.3 ± 1.9
0.3 ± 4.6
10.0 ± 4.5
221.05919 ± 0.00011
100
80
50
+
−
2
2
2
05.06477 ± 0.00005
37.09060 ± 0.00060
23.07500 ± 0.00010
[M+H−H2O]
[M−H−CO2]
+
−
−
[M+H+CH2]
[M−3H+O]
45
[M+H]+
[M−H−CO]
−
208.05396 ± 0.00093
[M−CH2]
+
−
9
a
-Anthracenecarboxylic acid
18.94
209.09353 ± 0.00091
100
85
60
[M+3H−O]
C15H13O
C14H11O
C15H9O
209.09609
195.08044
205.06479
237.09101
−8.3 ± 4.4
−2.8 ± 2.4
3.3 ± 2.9
0.1 ± 3.9
−1
221.05990 ± 0.00042
177.07022 ± 0.00009
208.05184 ± 0.00100
193.06603 ± 0.00012
100
40
10
5
[M−H]
C15H9O2
C14H9
C14H8O2
C14H9O
221.05971
177.06988
208.05188
193.06479
0.9 ± 1.9
2.0 ± 0.5
−0.2 ± 4.8
6.4 ± 0.6
+
−
1
2
2
95.07990 ± 0.00047
05.06547 ± 0.00059
37.09103 ± 0.00093
[M+H−CO]
[M−H−CO2]
+
−
[M+H−H2O]
[M−CH2]
[M+H+CH2]+
C16H13O2
[M−H−CO]
−
30
Using a C18 200 mm × 3.2 mm column (Restek) with 5 ␮m particle size. A gradient program of A:H2O and B:MeOH was used with a 0.20 mL min flow rate: 20% B for 5 min, linear increase to 90% B at 20 min and held to 27 min, linear decrease to 20% B at
3
0 min and held to 35 min.
b
Measure masses are reported as averages and standard deviations of the masses observed at three different points along the chromatographic peak.
c
Fragments are shown to reflect changes in the empirical formula from the neutral molecule and do not necessarily reflect the ionization or fragmentation pathway.
d
Possible empirical formula matches for the observed mass were calculated with elemental ranges of C:8–18, H:6–16, O:0–4, and Na:0–1 for monomers and C:16–32, H:12–32, O:0–4, and Na:0–1 for suspected dimers.
e
The standard deviation calculated from the observed masses was converted to ppm error to show its influence on empirical formula estimations.

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
5
Scheme 1. Proposed fragmentation pathway for 9-phenanthrenecarboxylic acid
with APCI in negative mode. Steps shown with losses or additions only reflect
changes in the empirical formula between the parent and product ions, and do not
necessarily show the exact mechanisms of fragmentation.
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
35
36
37
38
39
40
others [31]. The spectra of four analyzed carboxy-PAH derivatives
−
exhibited the base peak of the deprotonated molecule [M−H] as
well as the signals corresponding to the loss of CO and CO from
2
the deprotonated molecule. In contrast to the previous work [31],
no ion corresponding to two subsequent losses of CO (giving the
−
[M−H−CO−CO] ion) was observed detected. For 4-carboxy-5-
4Q1 6 phenanthrenecarboxaldehyde, no CO loss from the deprotonated
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
molecule was observed, but rather there occurred successive losses
of CO from the aldehyde group and CO2 from the carboxyl group
following deprotonation (Scheme 2).
While the analysis of carboxy-PAHs with APCI in positive
mode resulted in lower response, the compounds studied still
exhibited common ionization/fragmentation trends but with sig-
nificantly varied ion abundances. The differences in abundance
for three isomeric compounds (9-anthracene-, 9-phenanthrene-
and 4-phenanthrenecarboxylic acids) may be essential for their
+
accurate identification. The protonated molecule [M+H] for
4- and 9-phenanthrenecarboxylic acids showed responses at
45% and 20% relative to the base peak, respectively, and was
not observed for 9-anthracenecarboxylic acid. For 4-carboxy-5-
phenanthrenecarboxaldehyde, a higher relative response of the
[M+H]+ ion was detected, most likely due to protonation of the
aldehyde group (see Scheme 2). The protonation followed by
Scheme
2. Proposed
fragmentation
pathway
for
4-carboxy-5-
phenanthrenecarboxaldehyde with APCI in positive mode. Steps shown with
losses or additions only reflect changes in the empirical formula between the
parent and product ions, and do not necessarily show the exact mechanisms of
fragmentation.
the loss of H O was observed for all carboxy-PAHs, although
2
3.2. Oxo-PAHs
279
in varied relative intensities. For 9-phenanthrenecarboxylic acid,
+
the [M+H−H O] species assigned to the carbonyl phenanthrene
2
To our knowledge, HRMS data of several major fragment ions
for seven oxo-PAHs (Table 2, structures are shown in Fig. 1, 2nd
and 3rd rows) are being reported for the first time in this work.
In positive mode, several common trends were observed among
the oxo-PAHs investigated. The major ionization pathway for
most of the mono- and diketones studied was protonation (spec-
tra shown in Fig. S3), mirroring observations made in previous
studies [18–20,30]. The loss of CO from the protonated molecule
was observed for each of these compounds but varying in abun-
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
+
ion [C
H
19
O] , was the least prevalent (10% relative inten-
1
5
sity to the base peak), while for the other three species this
ion of 205 m/z was much more abundant (60–80%). Three com-
pounds, 9-phenanthrenecarboxylic acid, 9-anthracenecarboxylic
acid and 4-carboxy-5-phenanthrenecarboxaldehyde, exhibited the
+
[M+H+CH ] ion at 25, 30, and 85% abundances relative to the base
2
peak, respectively. This ion is possibly a result of gas-phase esteri-
fication by methanol.
In this study, only one hydroxy-PAH, 1-hydroxypyrene, was
considered. This compound is one of the most abundant hydroxy-
PAHs in atmospheric samples [28]. In positive mode, ionization via
+
dance. Successive losses of CO giving the [M+H−C O ] ion were
2
2
observed for only two compounds, 9,10-phenanthrenedione and
pyrene-4,5-dione (Fig. S3). This seems to be a highly regioselective
process since spectra of the other two three-ring diketone PAHs
did not exhibit this fragment ion. We suggest that this fragmen-
protonation as well as electron abstraction [M+H]+ and [M−e ]
−
+
ions were observed in 100% and 85% relative abundance, respec-
+
tively. Additionally, an oxygen adduct [M−H+O] was detected in
tation is likely to occur via ring opening (Scheme 3a and b). For
75% relative abundance to the base peak. In negative mode, 1-
hydroxypyrene was mainly ionized via deprotonation, yielding the
+
9
,10-phenanthrenedione, the structure of the [M+H−C O ] ion
2
2
is similar to that of a deprotonated biphenyl cation (Scheme 3a).
In contrast, for pyrene-4,5-dione, an analogous ring-opening pro-
cess could result in a structure that resembles a deprotonated
−
[M−H] ion. In the spectra obtained using positive mode, the signal
−
of the oxygen adduct [M−2H+O] was again found.
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

Table 2
Pseudomolecular ions and fragments observed for oxo-PAHs with HPLC-APCI-HRMS in both positive and negative modes.
Compound
tR a
Positive polarity
Negative polarity
(min)
Measured massb (Da)
Rel.
abund.
Fragmentc
Calculated
empirical
formula
Calculated
mass (Da)
Mass
error
(ppm)
Measured massb (Da)
Rel.
abund.
(%)
Fragmentc
Calculated
empirical
formulad
Calculated
mass (Da)
Mass error
(ppm)e
d
e
(%)
[M+H]+
C14H11O
C13H9
C28H17O2
C15H11O2
195.08040
165.06990
385.12231
223.07530
−5.5 ± 2.2
−0.8 ± 4.3
−4.0 ± 3.2
−2.1 ± 2.1
193.06546 ± 0.00015
208.05230 ± 0.00057
385.11995 ± 0.00827
100
30
10
[M−H]
−
C14H9O
C14H8O2
C28H17O2
193.06589
208.05290
385.12340
−2.2 ± 0.8
−2.9 ± 2.7
−8.9 ± 2.1
Anthrone
21.45
195.07934 ± 0.00043
100
10
8
+
−
−
1
3
2
65.06976 ± 0.00070
85.12075 ± 0.00125
23.07483 ± 0.00047
[M+H−CH2O]
[M−2H+O]
+
[M+M−3H]
[M+M−3H]
4
[M+H+CO]+
[M+H]+
−
−
−
9
9
,10-Anthracenedione
23.37
18.92
209.06013 ± 0.00040
100
30
10
C14H9O2
C14H9O
C28H17O2
209.05971
193.06479
385.12230
2.0 ± 1.9
0.1 ± 1.7
0.3 ± 3.9
208.05187 ± 0.00031
208.05205 ± 0.00073
100
100
[M+e
[M+e
]
]
C14H8O2
C14H8O2
208.05298
208.05298
−5.3 ± 1.5
−4.5 ± 3.5
+
1
3
93.06482 ± 0.00033
85.12241 ± 0.00151
[M+H−O]
+
[M+M+H−2O]
+
−
,10-Phenanthrenedione
181.06486 ± 0.00021
100
90
60
50
50
[M+H−CO]
C13H9O
C14H9O2
C12H9
C14H12NO2
C15H11O2
181.06479
209.05971
153.06988
226.08620
223.07530
0.4 ± 1.1
9.3 ± 1.6
209.06165 ± 0.00033
153.06918 ± 0.00009
226.08584 ± 0.00011
223.07472 ± 0.00030
[M+H]+
+
[M+H−C2O2]
−4.6 ± 0.6
−1.6 ± 0.5
−2.6 ± 1.4
+
[M+NH4]
[M+CH3]+
−
−
−
1
,4-Phenanthrenedione
21.65
21.06
209.05992 ± 0.00043
100
50
15
[M+H]+
C14H9O2
C15H11O3
C13H9O
209.05971
239.07020
181.06479
1.0 ± 2.1
0.5 ± 0.7
3.8 ± 4.8
208.05282 ± 0.00074
195.04460 ± 0.00082
223.03901 ± 0.00075
100
45
20
[M+e
]
C14H8O2
C13H7O2
C14H7O3
208.05298
195.04510
223.04007
−0.8 ± 3.6
−2.6 ± 4.2
−4.8 ± 3.4
−
2
1
39.07032 ± 0.00018
81.06548 ± 0.00087
[M+CH3O]+
[M+e −CH]
+
−
[M+H−CO]
[M+H]+
[M−H+O]
−
−
Pyrene-4,5-dione
233.05827 ± 0.00029
100
85
70
C16H9O2
C15H9O
C14H9
233.05970
205.06470
177.06980
250.08620
−6.1 ± 1.2
3.6 ± 2.7
232.05187 ± 0.00110
100
[M+e
]
C16H8O2
232.05298
−4.8 ± 4.7
+
205.06545 ± 0.00056
177.06969 ± 0.00025
250.08569 ± 0.00036
[M+H−CO]
+
[M+H−C2O2]
−0.6 ± 2.5
−2.0 ± 1.5
60
[M+NH4]+
C16H12NO2
Phenanthrene-9-
carboxaldehyde
23.47
23.47
207.07998 ± 0.00017
221.09549 ± 0.00019
100
90
60
40
10
5
[M+H]+
C15H11O
C16H13O
C15H11
C14H11
C16H9O2
C11H10Na
207.08040
221.09600
191.08550
179.08553
233.05970
165.06747
−2.0 ± 0.8
−2.3 ± 0.9
−2.6 ± 1.9
−3.7 ± 1.4
2.7 ± 2.2
[M+CH3]+
+
1
1
2
1
91.08501 ± 0.00036
79.08487 ± 0.00024
33.06034 ± 0.00051
65.06739 ± 0.00007
[M+H−O]
+
[M+H−CO]
+
[M+CO−H]
−0.5 ± 0.4
[M+H]+
C15H11O
C14H11
C14H10
C15H11
207.08040
179.08553
178.07770
191.08550
−3.8 ± 2.0
6.3 ± 3.6
1.9 ± 2.0
−1.6 ± 1.1
−1
206.07313 ± 0.00042
221.06031 ± 0.00120
100
50
[M+e
−
]
−
C15H10O
C15H9O2
206.07371
221.06080
−2.8 ± 2.1
−2.2 ± 5.4
Anthracene-9-
carboxaldehyde
207.07961 ± 0.00042
179.08665 ± 0.00064
100
50
20
8
+
−
[M+H−CO]
[M−H+O]
+
1
1
78.07803 ± 0.00036
[M+H−CHO]
+
91.08519 ± 0.00022
[M+H−O]
a
Using a C18 200 mm x 3.2 mm column (Restek) with 5 ␮m particle size. A gradient program of A:H2O and B:MeOH was used with a 0.20 mL min flow rate: 20% B for 5 min, linear increase to 90% B at 20 min and held to 27 min, linear decrease to 20% B at
3
0 min and held to 35 min.
b
Measure masses are reported as averages of the masses observed at three different points along the chromatographic peak.
c
Fragments are shown to reflect changes in the empirical formula from the neutral molecule and do not necessarily reflect the ionization or fragmentation pathway.
d
Possible empirical formula matches for the observed mass were calculated with elemental ranges of C:8–18, H:6–16, O:0–4, and Na:0–1 for monomers and C:16–32, H:12–32, O:0–4, and Na:0–1 for suspected dimers.
e
The standard deviation calculated from the observed masses was converted to ppm error to show its influence on empirical formula estimations.

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
7
Scheme 3. (a) Proposed fragmentation pathway for 9,10-phenanthrenedione with APCI in positive mode. Steps shown with losses or additions only reflect changes in the
empirical formula between the parent and product ions, and do not necessarily show the exact mechanisms of fragmentation. (b) Proposed fragmentation pathway for
pyrene-4,5-dione with APCI in positive mode. Steps shown with losses or additions only reflect changes in the empirical formula between the parent and product ions, and
do not necessarily show the exact mechanisms of fragmentation.
−
2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
99
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
phenanthrene molecule. The ring-opening pathway would explain
pathway was deprotonation forming the [M−H] anion while 9-
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
+
+
the absence of both the [M+H−CO] and [M+H−C O ] ions for
phenanthrenealdehyde was not ionized at all.
2
2
9,10-anthracenedione, whose specific ring fusion pattern makes
this process unlikely. Similarly, these fragments were not detected
in the spectrum of 1,4-phenanthrenedione, a compound with two
carbonyl groups not adjacent to each other.
Additional fragmentation/ionization processes in APCI neg-
ative mode were observed for 1,4-phenanthrenedione and
9-anthracenecarboxaldehyde (Fig. S4). We identified an oxygen
−
adduct of the deprotonated molecule, [M−H+O] , and confirmed its
Dimerization was observed only for two studied compounds,
anthrone and 9,10-anthracenedione. Anthrone exhibited a small
degree of dimerization, most likely into a structure similar to a pro-
tonated bianthrone ion observed at m/z of 385 (Fig. S3). The same
ion (m/z 385) was also discernible for 9,10-anthracenedione, sug-
gesting the loss of CO prior to the dimerization. For anthrone, a
composition with by high mass accuracy (−2 ± 5 ppm; Table 2). This
−
species was previously proposed to be the [M+CH ] ion [31]. The
3
−
occurrence of the [M+CH3] adduct was ruled out based on a high
mass error of −166 ± 5 ppm from the measured value. The proposed
pathway for the formation of the oxygen adducted using 1,4-
phenanthrenedione as a model compound is shown in Scheme 5. A
deprotonated methanol molecule [CH O] may undergo a nucleo-
philic addition to the carbonyl group to create the [M+OCH ]
fragment that loses CH4 to give the [M−H+O] ion.
−
fragment of 165 m/z was assigned to C H , formed possibly by the
1
3
9
3
+
−
loss of CH O from the [M+H] ion; however, the mechanism of this
2
3
−
process is unclear.
Using APCI in positive mode, two monocarboxaldehydes
showed similar ionization pathways as the mono- and diketone
PAHs. The major ionization pathway was protonation (Fig. S3). The
loss of CO from the protonated molecule was also observed for both
aldehydes. It appears that the loss of CO for these aldehydes and
diketones proceeds via different pathways. Instead of the expected
ion at m/z 177 as a result of the protonation of the aldehydic oxygen
3.3. Nitro-PAHs
355
The APCI ionization/fragmentation of three- and four-ring PAH
derivatives with either one or two nitro groups was investigated
to evaluate the ionization processes and confirm the identities
356
357
358
followed by the loss of formaldehyde CH O, the ion at m/z 179 was
2
observed for both compounds (Fig. S3). The latter ion was likely
formed after the loss of CO.
Signals of the fragments corresponding to the loss of O were also
discernable for both monocarboxaldehydes, with a higher abun-
dance observed for 9-phenanthrenecarboxaldehyde (Table 2, Fig.
+
S3). The latter compound provided the abundant [M+15] ion, the
signal of which was not detected in the spectra of other oxo-PAHs
except 9,10-phenanthrenedione. This ion was attributed to the for-
mation of the [M+CH3]+ adduct. Scheme 4 shows the proposed
pathway for its formation through the addition of methanol to the
carbonyl group followed by the loss of water.
The ionization of oxo-PAHs with APCI in negative mode resulted
in a very limited, if any, fragmentation compared to that in
positive mode, with observed fragments having lower relative
abundances (Table 2, Fig. 3S). Similar to previous investigations
[18,30], spectra of all studied oxo-PAHs with the exception of
anthrone and 9-phenanthrenealdehyde featured the major peak
Scheme 4. Proposed fragmentation pathway for 9-phenanthrenecarboxaldehyde
with APCI in positive mode. Steps shown with losses or additions only reflect changes
in the empirical formula between the parent and product ions, and do not necessarily
show the exact mechanisms of fragmentation.
−
−
assigned to the [M+e ] anion. For anthrone, the main ionization
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
8
and 9-nitrophenanthrene as the signal of the [M+H−16]+ ion was
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
the most prevalent for both compounds. However, this species
was not observed for 3-nitrophenanthrene (Fig. S5). Spectra of
+
mononitro-PAHs exhibited the [M+H−30] fragment. Our HRMS
data (<10 ppm) support the previously proposed elemental com-
+
position of this ion as [M+3H−2O] [33]. This fragment is formed
upon reduction of the NO group to NH3+. The relative abundance
2
of the [M+H]+ ion was low for all mononitro-PAHs studied (ran-
ging from 5 to 50% of the base peak response). This may be due
to a competition between protonation of the neutral molecule and
gas-phase reduction.
Dinitro-PAHs, 9,10-dinitroanthracene and 1,6-dinitropyrene,
provided rather complex spectra in APCI positive mode (Table 3,
Fig. S5). The spectra appear to have two competitive fragmenta-
tion pathways involving each of the two nitro groups. As expected,
the nitro group reduction to the protonated amino moiety was the
major ionization pathway for both compounds. However, the most
abundant ion registered in the mass spectra of 1,6-dinitropyrene
+
corresponded to the [M+H−30] species with only one nitro group
reduced, while the base peak for 9,10-dinitroanthracene belonged
Scheme 5. Proposed fragmentation pathway for 1,4-phenanthrenedione with APCI
in negative mode. Steps shown with losses or additions only reflect changes in the
empirical formula between the parent and product ions, and do not necessarily show
the exact mechanisms of fragmentation.
+
to the [M+H−62] ion formed upon reduction of both nitro groups.
+
Spectra of both compounds showed peaks for [M+H−O] fragments
with different relative abundances.
APCI-MS spectra of both mono- and dinitro-PAHs in negative
mode displayed a significantly smaller number of fragments com-
pared to that in positive mode (Fig. S5; see also Table 3). For all
species studied, the most abundant ion was produced via asso-
3
3
3
3
3
3
3
3
3
3
3
3
59
60
61
62
63
64
65
66
67
68
69
70
of fragments with HRMS data. To our knowledge, the ioniza-
tion/fragmentation processes for dinitro-PAHs in APCI have not
been previously reported. The studied species were chosen to rep-
resent differences in the location of the nitro groups and the ring
fusion pattern of the aromatic backbone, allowing the identification
of structure-specific processes.
−
−
ciative electron capture by the neutral molecule, giving [M+e ] .
This corroborates the observation reported in previous studies per-
formed using low resolution MS detection [11,32]. The associative
−
addition of oxygen following deprotonation to give the [M−H+O]
In positive APCI mode, all nitro compounds underwent similar
ionizations and fragmentations with varied relative abundances of
major ions (see Table 3, Fig. S5). All species consistently yielded the
ion was common for mononitro-PAHs. The oxygen adduct has been
reported previously [11,35]. To evaluate the eluent solvents as pos-
−
+
sible oxygen contributors to the [M−H+O] ions, direct infusion
[M+H−16] ion, which was likely formed due to the loss of oxy-
+
of 1-nitropyrene at 1 ppm was performed in various solvents; the
results are shown in Fig. 2. When 1-nitropyrene was dissolved
gen from the protonated ion ([M+H−O] ) as shown in Scheme 6.
This ionization process was highly specific for 9-nitroanthracene
−
− −
in acetonitrile, the response ratio of [M−H+O] to [M+e ] (m/z
2
62:m/z 247) was minimal; this was in contrast to the spectra
produced when oxygen-containing solvents were used. It should
be noted that dinitro-PAHs did not undergo this oxygen addition.
Thus, the presence of the second nitro group might play a key role
−
Fig. 2. The contribution of various solvents to formation of [M−H+O] ions for 1-
Scheme 6. Proposed fragmentation pathway for 3-nitrophenanthrene with APCI
in positive mode. Steps shown with losses or additions only reflect changes in the
empirical formula between the parent and product ions, and do not necessarily show
the exact mechanisms of fragmentation.
nitropyrene in APCI with negative polarity. Spectra were recorded and averaged
during the direct infusion of 1-nitropyrene (at 1 ppm with 5 mM formic acid) in
−
1
three different solvents at a flow rate of 0.2 mL min . The height of each mass peak
was used as the response (intensity, counts per second).
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

Table 3
Pseudomolecular ions and fragments observed for nitro-PAHs with HPLC-APCI-HRMS in both positive and negative modes.
Compound
tR a
Positive polarity
Negative polarity
(min)
Measured massb (Da)
Rel.
abund.
Fragmentc
Calculated
empirical
formula
Calculated
mass (Da)
Mass error
(ppm)
Measured massb (Da)
Rel.
abund.
(%)
Fragmentc
Calculated
empirical
formulad
Calculated
mass (Da)
Mass error
(ppm)e
e
d
(%)
+
− −
]
9
-Nitroanthracene
23.74
208.07546 ± 0.00024
100
17
17
10
3
[M+H−O]
C14H10NO
C14H12N
C14H10O
C14H10NO2
C14H10
208.07569
194.09643
194.07262
224.07060
178.07770
−1.1 ± 1.2
−2.0 ± 4.2
121 ± 4.2
2.4 ± 2.7
223.06308 ± 0.00062
238.05027 ± 0.00102
193.06543 ± 0.00030
208.05150 ± 0.00095
256.06066 ± 0.00090
100
50
20
3
[M+e
C14H9NO2
C14H9NO3
C14H9O
C14H8O2
C14H10NO4
223.06388
238.05097
193.06589
208.05243
256.06153
−3.6 ± 2.8
−2.9 ± 4.3
−2.4 ± 1.6
−4.5 ± 4.6
−3.4 ± 3.5
+
+
−
1
1
2
1
94.09603 ± 0.00082
94.09603 ± 0.00082
24.07114 ± 0.00060
78.07725 ± 0.00003
[M+3H−2O]
[M−H+O]
+
−
−
[M+H−NO]
[M+e −NO]
[M+H]+
[M−H+O−NO]
−
−
−
[M+H−NO2]
−2.6 ± 0.1
3
[M+e +HO2]
+
− −
]
3
9
1
9
-Nitrophenanthrene
-Nitrophenanthrene
-Nitropyrene
24.65
24.12
26.46
24.16
194.09626 ± 0.00031
100
80
15
6
[M+3H−2O]
C14H12N
C14H10NO
C14H10
194.09643
208.07569
178.07770
224.07060
−0.9 ± 1.6
−0.8 ± 0.6
−4.0 ± 0.8
0.9 ± 2.1
223.06250 ± 0.00049
238.04966 ± 0.00107
209.05960 ± 0.00109
193.06676 ± 0.00043
100
30
3
[M+e
C14H9NO2
C14H8NO3
C14H9O2
C14H9O
223.06388
238.05097
209.06080
193.06589
−6.2 ± 2.2
−5.5 ± 4.5
−5.7 ± 5.2
4.5 ± 2.2
+
−
2
1
2
08.07553 ± 0.00012
78.07699 ± 0.00015
24.07080 ± 0.00048
[M+H−O]
[M−H+O]
+
−
−
−
[M+H−NO2]
[M+e +O−NO]
[M+H]+
C14H10NO2
2
[M+e −NO]
−
−
+
− −
208.07512 ± 0.00031
100
70
10
5
[M+H−O]
C14H10NO
C14H12N
C14H10
208.07569
194.09643
178.07770
224.07060
−2.7 ± 1.5
−5.8 ± 0.6
−7.0 ± 2.5
5.4 ± 2.8
223.06354 ± 0.00086
238.05060 ± 0.00114
209.05954 ± 0.00054
193.06529 ± 0.00057
100
8
4
[M+e
]
C14H9NO2
C14H8NO3
C14H9O2
C14H9O
223.06388
238.05097
209.06080
193.06589
−1.5 ± 3.9
−1.5 ± 4.8
−6.0 ± 2.6
−3.1 ± 3.0
+
+
−
1
1
2
94.09530 ± 0.00012
78.07645 ± 0.00044
24.07182 ± 0.00064
[M+3H−2O]
[M−H+O]
−
[M+H−NO2]
[M+e +O−NO]
[M+H]+
C14H10NO2
4
[M+e −NO]
−
−
+
−
−
−
218.09619 ± 0.00162
100
50
40
[M+3H−2O]
C16H12N
C16H10NO2
C16H10NO
C16H10
218.09643
248.07060
232.07569
202.07770
−1.1 ± 7.5
−2.7 ± 4.4
2.4 ± 4.3
2.3 ± 1.7
247.06206 ± 0.00079
217.06424 ± 0.00096
262.05055 ± 0.00105
233.05968 ± 0.00149
100
30
27
5
[M+e
]
C16H9NO2
C16H9O
C16H8NO3
C16H9O2
247.06388
217.06589
262.05097
233.06080
−7.4 ± 3.2
−7.6 ± 4.4
−1.6 ± 4.0
−4.8 ± 6.4
[M+H]+
[M+H−O]
[M+e −NO]
−
2
2
2
48.06993 ± 0.00109
32.07624 ± 0.00100
02.07816 ± 0.00035
+
−
[M−H+O]
+
−
−
−
20
[M+H−NO2]
[M+e +O−NO]
+
−
−
−
,10-Dinitroanthracene
207.09072 ± 0.00041
100
90
60
25
22
8
8
4
3
[M+3H−4O]
C14H11N2
C14H10NO
C14H10O
C14H10NO2
C14H11N2O
C14H11N2O2
C13H10N
207.09160
208.07570
194.07260
224.07060
223.08650
239.08150
180.08078
178.07770
253.06077
−4.2 ± 2.0
−3.3 ± 2.2
9.4 ± 2.6
−1.9 ± 3.1
9.4 ± 3.1
8.6 ± 6.0
268.04708 ± 0.00089
238.04936 ± 0.00070
254.04414 ± 0.00035
223.06213 ± 0.00069
208.05118 ± 0.00074
193.06499 ± 0.00042
100
60
10
4
4
2
[M+e
]
C14H8N2O4
C14H8NO3
C14H8NO4
C14H9NO2
C14H8O2
268.04896
238.05097
254.04588
223.06388
208.05298
193.06589
−7.0 ± 3.3
−6.7 ± 3.0
−6.9 ± 1.4
−7.8 ± 3.1
−8.6 ± 3.5
−4.7 ± 2.2
+
−
208.07501 ± 0.00045
194.07443 ± 0.00050
224.07017 ± 0.00069
223.08860 ± 0.00070
239.08356 ± 0.00144
180.08218 ± 0.00074
178.07527 ± 0.00096
253.06031 ± 0.00062
[M+2H−NO3]
[M+e −NO]
+
−
[M+2H−N2O3]
[M+e +O−NO]
+
−
−
[M+H−NO2]
[M+e +H−NO2]
+
−
−
[M+3H−3O]
[M+e −N2O2]
+
−
−
[M+3H−2O]
[M+e +H−N2O3]
C14H9O
+
[M+2H−CNO4]
7.8 ± 4.1
−8.0 ± 5.4
−1.8 ± 2.4
+
[M+2H−2NO2]
C14H10
C14H9N2O3
+
[M+H−O]
+
−
−
−
]
1
,6-Dinitropyrene
25.17
263.08100 ± 0.00014
100
60
30
25
25
25
20
15
10
[M+3H−2O]
C16H11N2O2
C16H9N2O3
C16H11N2O
C16H10NO
C16H9NO
C16H10N2O
C16H9N2O2
C16H12N
263.08151
277.06070
247.08660
232.07570
231.06786
246.07870
261.06590
218.09640
216.08080
−2.0 ± 0.5
0.1 ± 0.5
292.04801 ± 0.00131
262.05035 ± 0.00115
276.05313 ± 0.00148
100
20
10
[M+e
C16H8N2O4
C16H8NO3
C16H8N2O3
292.04896
262.05097
276.05404
−3.2 ± 4.5
−2.4 ± 4.4
−3.3 ± 5.4
+
−
277.06074 ± 0.00014
247.08933 ± 0.00050
232.07572 ± 0.00044
231.07159 ± 0.00177
246.07829 ± 0.00021
261.06604 ± 0.00015
218.09593 ± 0.00020
216.08061 ± 0.00026
[M+H−O]
[M+e −NO]
+
−
−
[M+3H−3O]
−1.6 ± 2.7
0.1 ± 1.9
7.5 ± 4.9
−1.7 ± 0.8
0.5 ± 0.6
[M+e −O]
+
[M+2H−NO3]
+
[M+H−NO3]
+
[M+2H−3O]
+
[M+H−2O]
+
[M+4H−NO4]
−2.1 ± 0.9
+
[M+2H−NO4]
C16H10N
−0.9 ± 1.2
a
−1
Using a C18 200 mm × 3.2 mm column (Restek) with 5 ␮m particle size. A gradient program of A:H2O and B:MeOH was used with a 0.20 mL min flow rate: 20% B for 5 min, linear increase to 90% B at 20 min and held to 27 min, linear decrease to 20% B at
3
0 min and held to 35 min.
b
Measure masses are reported as averages and standard deviations of the masses observed at three different points along the chromatographic peak.
c
Fragments are shown to reflect changes in the empirical formula from the neutral molecule and do not necessarily reflect the ionization or fragmentation pathway.
d
Possible empirical formula matches for the observed mass were calculated with elemental ranges of C:8–18, H:6–16, O:0–4, and Na:0–1 for monomers and C:16–32, H:12–32, O:0–4, and Na:0–1 for suspected dimers.
e
The standard deviation calculated from the observed masses was converted to ppm error to show its influence on empirical formula estimations.

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
1
0
4
4
4
4
4
4
12
13
14
15
16
178
in deactivating the deprotonated molecule toward oxygen addi-
From the mass spectra of each chromatographic peak, specific
trends in common ions were observed (Fig. 3). Several compounds
showed a significant fragment of 205 m/z, tentatively assigned
to an ionized phenanthrene carbonyl ion. The common ions
of 205, 235 and 191 m/z enabled the tentative identification of
three compounds: 4-carboxy-5-phenanthrenecarboxaldehyde,
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
−
tion. As with positive APCI mode, abundance of the [M−H+O] ion
varied for each constitutional isomer. 9-Nitrophenanthrene exhib-
−
−
ited a low relative abundance (∼8% of [M+e ] base peak) of the
+
[M−H+O] ion, while 3-nitrophenanthrene and 9-nitroanthracene
yielded abundances of 30% and 50% of the analogous ion relative to
the base peak.
phenanthrene-4,5-dicarboxaldehyde
and
hydroxymethyl-
phenanthrenecarboxylic acid, assuming that their formation
occurred via the ring-opening oxidation of pyrene (Figs.
4
19
3.4. Products from the heterogeneous ozonation of PAHs
3
and S6). The first compound confirmed by a standard
seemed to coelute with 4-phenanthrenecarboxylic acid (also
confirmed by using the corresponding standard). 4-Carboxy-
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
The interpretation of APCI ionization/fragmentation pathways
using HRMS data demonstrated above enabled the identification
of unknown products formed upon exposure of pyrene embed-
ded on a quartz filter to ozone (see Fig. 3) [36,37]. We were able
to confirm the identity of two compounds using standards and
tentatively identify five additional compounds based on the EICs
(obtained with high resolution of ± 0.03 m/z) for the major fragment
ions observed with the standards (Table 4; mass spectra shown in
Figs. S6 & S7).
Hydroxy-PAHs, carboxy-PAHs and oxo-PAHs identified were the
compounds formed through the ring-opening pathway as opposed
to a direct oxidation [7,36,37]. Although most of these products
have been previously observed both in aqueous media and on
model particles [7,36,38], the present study is the first one to report
HRMS data to confirm their elemental composition.
5
-phenanthrenecarboxaldehyde was previously detected upon
the ozonation of pyrene coated on azelaic acid particles [7]. The
dicarboxaldehyde was identified based on the concurrent ions of
2
05 m/z (from the neutral loss of CH O) and 191 m/z (loss of CO ),
2
2
+
both formed from the protonated molecule [M+H] of 235 m/z.
This product was previously observed during the homogeneous
and heterogeneous ozonation of pyrene [5,7,38]. One of the
possible constitutional isomers of phenanthrenecarboxylic acid
+
+
was identified through the observed [M+H] , [M+H−H O] and
2
+
[
M+H−CH O ] ions. While this compound was previously identi-
2
2
fied as a product of pyrene ozonation in aqueous solution [5,38],
to our knowledge, it has never been reported in heterogeneous
ozonation reactions. We were not able to unequivocally identify
Table 4
Pseudomolecular ions and fragments observed for products from the heterogeneous ozonation of pyrene in a small-scale flow reactor using HPLC-APCI-HRMS in positive
mode.
Peak #
Proposed compound
IDa
tR b
Positive polarity
(
min)
Measured massc (m/z Da)
Rel.
abund.
Fragmentd
Calculated
empirical
formula
Calculated
mass
(m/z Da)
Mass errorf
(ppm)
e
(
%)
+
1
4-Carboxy-5-
phenanthrenecarboxaldehyde
S
18.18
235.07392 ± 0.00040
251.06876 ± 0.00045
100
65
40
25
20
20
5
[M+H−O]
C16H11O2
C16H11O3
C15H9O
C15H11
C16H9O2
C17H13O3
C14H9
235.07536
251.07027
205.06479
191.08550
233.05971
265.08590
177.06988
−8.0 ± 1.7
−5.5 ± 1.8
−3.4 ± 2.0
−3.7 ± 2.7
−5.2 ± 2.7
−4.5 ± 2.0
−4.9 ± 2.2
+
[M+H]
+
2
1
2
2
1
05.06395 ± 0.00041
91.08474 ± 0.00052
33.05831 ± 0.00063
65.08434 ± 0.00052
77.06856 ± 0.00039
[M+H−CH2O2]
+
[M+H−CO3]
+
[M+H−H2O]
+
[M+H+CH2]
+
[M+H−C2H2O3]
+
2
3
4
5
6
7
Phenanthrenecarboxylic acid
4-Oxapyrene-5-one
T
T
T
T
T
T
18.28
19.43
19.54
19.84
19.93
21.45
205.06503 ± 0.00034
100
50
40
[M+H−H2O]
C15H9O
C15H9O2
C15H11O2
205.06479
221.06036
223.07536
−0.4 ± 1.7
−3.8 ± 4.9
4.0 ± 1.6
−
+
2
21.06055 ± 0.00107
23.07667 ± 0.00037
[M−H−2e
]
+
2
[M+H]
+
193.06518 ± 0.00024
100
50
25
[M+H−CO]
C14H9O
C15H9O2
C14H9
193.06479
221.05971
177.06988
3.4 ± 1.2
4.6 ± 1.7
8.6 ± 2.4
2
1
21.06031 ± 0.00037
77.07143 ± 0.00043
[M+H]+
+
[M+H−CO2]
+
Phenanthrenecarboxaldehyde
191.08438 ± 0.00043
100
40
5
[M+H−O]
C15H11
C15H11O
C14H11
191.08553
207.08044
179.08553
−3.4 ± 2.3
3.7 ± 0.7
1.9 ± 0.5
+
2
1
07.08104 ± 0.00015
79.08597 ± 0.00010
[M+H]
+
[M+H−CO]
235.07533 ± 0.00046
205.06556 ± 0.00062
100
50
5
[M+H]+
C16H11O2
C15H9O
C15H11
235.07536
205.06479
191.08553
−0.2 ± 2.0
3.2 ± 3.0
Phenanthrene-4,5-
dicarboxaldehyde
+
[M+H−CH2O]
+
191.08508 ± 0.00098
[M+H−CO2]
−7.3 ± 5.1
ꢀ
ꢀ
ꢀ
+
(1,1 -Biphenyl)-2,2 ,6,6 -
tetracarboxaldehyde
239.06893 ± 0.00054
251.07065 ± 0.00050
100
10
10
[M+H−CO]
C15H11O3
C16H11O3
C16H11O4
239.07027
251.07027
267.06518
−5.9 ± 2.3
3.7 ± 2.0
+
[M+H−O]
+
267.06482 ± 0.00052
[M+H]
−3.3 ± 1.9
+
Hydroxymethyl-
phenanthrenecarboxylic
acid
235.07457 ± 0.00068
205.06375 ± 0.00101
191.08645 ± 0.00041
100
60
20
5
[M+H−H2O]
C16H11O2
C15H9O
C15H11
235.07536
205.06479
191.08553
253.08592
−4.2 ± 2.9
0.6 ± 4.9
7.2 ± 2.2
−1.9 ± 1.6
+
+
[M+H−CH4O2]
[M+H−CH2O3]
+
253.08556 ± 0.00041
[M+H]
C16H13O3
a
Identification was either confirmed by comparison of MS spectra to that of a standard (S) or tentatively confirmed through interpretation of measure MS spectra (T).
Using a C18 200 mm × 3.2 mm column (Restek) with 5 ␮m particle size. A gradient program of A:H2O and B:MeOH was used with a 0.20 mL min flow rate: 20% B for
b
−1
5
min, linear increase to 90% B at 20 min and held to 27 min, linear decrease to 20% B at 30 min and held to 35 min.
c
Q9
Measure masses are reported as averages of the masses observed at three different points along the.
Fragments are shown to reflect changes in the empirical formula from the neutral molecule and do not necessarily reflect the ionization or fragmentation pathway.
Possible empirical formula matches for the observed mass were calculated with elemental ranges of C:8–18, H:6–16, O:0–4, and Na:0–1 for monomers and C:16–32,
d
e
H:12–32, O:0–4, and Na:0–1 for suspected dimers.
f
The standard deviation calculated from the observed masses was converted to ppm error to show its influence on empirical formula estimations.
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
11
Fig. 3. HPLC–APCI–HRMS smoothed TIC and EIC chromatograms obtained in positive mode of products formed during the heterogeneous ozonation of pyrene in a small-scale
flow reactor. Numbers shown correspond to those listed in Table 4 and Figs. S6 & S7.
4
4
4
4
4
4
4
4
4
4
4
4
4
60
61
62
63
64
65
66
67
68
69
70
71
72
the last of the m/z 205 peaks; however, this compound is likely to
be another isomer of phenanthrenecarboxylic acid.
APCI–HRMS data, a number of products of pyrene ozonation were
tentatively identified based on common ions of 205, 235 and
191 m/z as well as relative intensities of common fragments exhib-
ited by standard compounds. For the first time, HRMS data were
obtained to support the identities of these products. Formation of
some of the pyrene ozonation products under the heterogeneous
conditions is reported for the first time.
485
486
487
488
489
490
491
Two additional products proposed to result from a ring-opening
ꢀ
attack by ozone were tentatively identified as (1,1 -biphenyl)-
ꢀ
ꢀ
2,2 ,6,6 -tetracarboxaldehyde (peak 6 in Fig. 3) and 4-oxapyrene-
5-one (peak 3 in Fig. 3). The former was identified based on the
+
+
[M+H−O] and [M+H−CO] ion fragments observed with the alde-
hyde standards. For 4-oxapyrene-5-one, similar to the studied
+
diketone standards, the [M+H−CO] ion was observed. In addition,
Acknowledgments
492
493
the spectrum of this compound displayed the peak assigned to the
+
[M+H−CO ] ion, which was not detected in the spectra of other
2
We would like to thank Dr. Kozliak for editing this manuscript.
The authors would like to acknowledge the funding by the National Q7 494
Science Foundation (NSF) Grant No. ATM-0747349, which made
this study possible. Any opinions, findings, and conclusions or rec-
ommendations expressed in this material are those of the author(s)
and do not necessarily reflect the views of the NSF.
ketone standards. This could be explained by the absence of the
ester moiety on the pyrene scaffold in the available standards.
495
496
497
498
4
73
4. Conclusions
4
4
4
4
4
4
4
4
4
4
4
74
75
76
77
78
79
80
81
82
83
84
Derivatives of PAHs containing either one or combination of
two functional groups (i.e., hydroxyl, carboxyl, carbonyl, and nitro
groups) were analyzed using HPLC–APCI–HRMS. High resolution
data for common fragmentation patterns provided by three and
four-ring PAH derivatives with similar functional groups were
reported for the first time. The data described in this study are
expected to be useful for the identification of unknown species
when standards are not available. The fragments observed for dif-
ferent types of PAH derivatives with various functional groups
reported in this study may facilitate future investigations aimed
at determining the mechanisms behind their formation. Using
Appendix A. Supplementary data
499
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ijms.2016.01.001.
500
501
References
502
[
1] K. Ravindra, A.K. Mittal, R. Van Grieken, Health risk assessment of urban
suspended particulate matter with special reference to polycyclic aromatic
hydrocarbons: a review, Rev. Environ. Health 16 (2001) 169–189.
503
504
505
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001

G Model
MASPEC155491–12
ARTICLE IN PRESS
R.E. Cochran et al. / International Journal of Mass Spectrometry xxx (2016) xxx–xxx
1
2
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
2Q8 8
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
[2] A.S. Wilson, C.D. Davis, D.P. Williams, A.R. Buckpitt, M. Pirmohamed, B.K. Park,
Characterisation of the toxic metabolite(s) of naphthalene, Toxicology 114
(1996) 233–242.
[3] M.T. Galceran, E. Moyano, J. Poza, Pentafluorobenzyl derivatives for the gas
chromatographic determination of hydroxy-polycyclic aromatic hydrocarbons
in urban aerosols, J. Chromatogr. A 710 (1995) 139–147, http://dx.doi.org/10.
1016/0021-9673(95)00263-M.
[4] M.G. Nishioka, C.C. Howard, D.A. Contos, L.M. Bail, J. Lewtas, Detection of
hydroxylated nitro aromatic and hydroxylated nitro polycyclic aromatic com-
pounds in an ambient air particulate extract using, Environ. Sci. Technol. 22
(1988) 908–915.
[5] Y.-I. Choi, A. Hong, Ozonation of polycyclic aromatic hydrocarbon in hexane
and water: identification of intermediates and pathway, Korean J. Chem. Eng.
24 (100) (2008) 3–1008, http://dx.doi.org/10.1007/s11814-007-0111-x.
[6] H.A. Herner, J.E. Trosko, S.J. Masten, The epigenetic toxicity of pyrene and
related ozonation byproducts containing an aldehyde functional group, Envi-
ron. Sci. Technol. 35 (2001) 3576–3583.
[7] S. Gao, Y. Zhang, J. Meng, J. Shu, Online investigations on ozonation products of
pyrene and benz[a]anthracene particles with a vacuum ultraviolet photoion-
ization aerosol time-of-flight mass spectrometer, Atmos. Environ. 43 (2009)
3319–3325, http://dx.doi.org/10.1016/j.atmosenv.2009.04.021.
[8] J.G. Calvert, R. Atkinson, K.H. Becker, R.M. Kamens, J.H. Seinfeld, T.J. Wallington,
et al., The Mechanisms of Atmospheric Oxidation of Aromatic Hydrocarbons,
1st ed., Oxford Press, New York, NY, 2002.
[9] B.J. Finlayson-Pitts, J.N. Pitts Jr., Upper and Lower Atmosphere, 1st ed., Academic
Press, San Diego, CA, 2000.
[10] C. Walgraeve, K. Demeestere, J. Dewulf, R. Zimmermann, H. Van Langenhove,
Oxygenated polycyclic aromatic hydrocarbons in atmospheric particulate mat-
ter: molecular characterization and occurrence, Atmos. Environ. 44 (2010)
1831–1846.
[11] R.P. Barreto, F.C. Albuquerque, A.D.P. Netto, Optimization of an improved ana-
lytical method for the determination of 1-nitropyrene in milligram diesel soot
samples by high-performance liquid chromatography–mass spectrometry, J.
Chromatogr. A 1163 (2007) 219–227, http://dx.doi.org/10.1016/j.chroma.2007.
06.040.
[12] D.Z. Bezabeh, H.A. Bamford, M.M. Schantz, S.A. Wise, Determination of nitrated
polycyclic aromatic hydrocarbons in diesel particulate-related standard refer-
ence materials by using gas chromatography/mass spectrometry with negative
ion chemical ionization, Anal. Bioanal. Chem. 375 (2003) 381–388, http://dx.
doi.org/10.1007/s00216-002-1698-8.
[13] H.A. Bamford, D.Z. Bezabeh, M.M. Schantz, S.A. Wise, J.E. Baker, Determination
and comparison of nitrated-polycyclic aromatic hydrocarbons measured in air
and diesel particulate reference materials, Chemosphere 50 (2003) 575–587.
[14] H.A. Bamford, A.L.W.D. Bezabeh, M. Lopez De Alda, D.L. Poster, L.C. Sander, M.M.
Schantz, P. Schubert, SRM 2975 diesel particulate matter certificate of analysis,
NIST 64 (32) (2004) 1–6, http://dx.doi.org/10.1080/00365510410006036.
[15] J.N. Pitts Jr., K.A. Van Cauwenberghe, D. Grosjean, P. Joachim, D.R. Fitz, W.L.
Belser, et al., Atmospheric reactions of polycyclic aromatic hydrocarbons: facile
formation of mutagenic nitro derivatives, Science 80 (202) (1978) 515–519.
[16] C. Walgraeve, S. Chantara, K. Sopajaree, P. De Wispelaere, K. Demeestere, H.
Van Langenhove, Quantification of PAHs and oxy-PAHs on airborne particu-
late matter in Chiang Mai, Thailand, using gas chromatography high resolution
mass spectrometry, Atmos. Environ. 107 (26) (2015) 2–272, http://dx.doi.org/
10.1016/j.atmosenv.2015.02.051.
[17] C. Adelhelm, R. Niessner, U. Pöschl, T. Letzel, Analysis of large oxygenated
and nitrated polycyclic aromatic hydrocarbons formed under simulated
diesel engine exhaust conditions (by compound fingerprints with SPE/LC-API-
MS), Anal. Bioanal. Chem. 391 (2008) 2599–2608, http://dx.doi.org/10.1007/
s00216-008-2175-9.
[18] O. Delhomme, M. Millet, P. Herckes, Determination of oxygenated poly-
cyclic aromatic hydrocarbons in atmospheric aerosol samples by liquid
chromatography–tandem mass spectrometry, Talanta 74 (2008) 703–710,
http://dx.doi.org/10.1016/j.talanta.2007.06.037.
[19] J. Lintelmann, K. Fischer, G. Matuschek, Determination of oxygenated polycyclic
aromatic hydrocarbons in particulate matter using a high-performance liquid
chromatography–tandem mass spectrometry, J. Chromatogr. A 1133 (2006)
241–247, http://dx.doi.org/10.1016/j.chroma.2006.08.038.
[20] J. Lintelmann, K. Fischer, E. Karg, A. Schröppel, Determination of selected
polycyclic aromatic hydrocarbons and oxygenated polycyclic aromatic hydro-
carbons in aerosol samples by high-performance liquid chromatography and
liquid chromatography–tandem mass spectrometry, Anal. Bioanal. Chem. 381
[21] B. Zielinska, S. Samy, Analysis of nitrated polycyclic aromatic hydrocarbons,
Anal. Bioanal. Chem. 386 (2006) 883–890, http://dx.doi.org/10.1007/s00216-
006-0521-3.
[22] M.M. Schantz, J.J. Nichols, S.A. Wise, Evaluation of pressurized fluid extraction
for the extraction of environmental matrix reference materials, Anal. Chem. 69
(1997) 4210–4219, http://dx.doi.org/10.1021/ac970299c.
[23] J.A. Sweetman, B. Zielinska, R. Atkinson, A possible formation pathway for the
2-nitrofluoranthene observed in ambient particulate organic matter, Atmos.
Environ. 20 (1986) 235–238.
[24] M. Tsapakis, E.G. Stephanou, Diurnal cycle of PAHs, nitro-PAHs, and oxy-PAHs
in a high oxidation capacity marine background atmosphere, Environ. Sci. Tech-
nol. 41 (2007) 8011–8017.
[25] J.N. Pitts, J.A. Sweetman, B. Zielinska, R. Atkinson, A.M. Winer, W.P. Harger, For-
mation of nitroarenes from the reaction of polycyclic aromatic hydrocarbons
with dinitrogen pentoxide, Environ. Sci. Technol. 19 (1985) 1115–1121, http://
dx.doi.org/10.1021/es00141a017.
[26] N. Jariyasopit, M. McIntosh, K. Zimmermann, J. Arey, R. Atkinson, P.H.-Y.
Cheong, et al., Novel nitro-PAH formation from heterogeneous reactions of
PAHs with NO2, NO3/N2O5, and OH radicals: prediction, laboratory studies,
and mutagenicity, Environ. Sci. Technol. 48 (2014) 412–419, http://dx.doi.org/
10.1021/es4043808.
[27] S.K. Pandey, K.-H. Kim, R.J.C. Brown, A review of techniques for the determi-
nation of polycyclic aromatic hydrocarbons in air, TrAC Trends Anal. Chem. 30
(171) (2011) 6–1739, http://dx.doi.org/10.1016/j.trac.2011.06.017.
[28] R.E. Cochran, N. Dongari, H. Jeong, J. Beránek, S. Haddadi, J. Shipp, et al.,
Determination of polycyclic aromatic hydrocarbons and their oxy-, nitro-, and
hydroxy-oxidation products, Anal. Chim. Acta 740 (9) (2012) 3–103, http://dx.
doi.org/10.1016/j.aca.2012.05.050.
[29] M.T. Galceran, E. Moyano, Determination of hydroxy polycyclic aromatic
hydrocarbons by liquid chromatography–mass spectrometry comparison of
atmospheric pressure chemical ionization and electrospray, J. Chromatogr. A
731 (1996) 75–84.
[30] G. Mirivel, V. Riffault, J.-C. Galloo, Simultaneous determination by ultra-
performance liquid chromatography–atmospheric pressure chemical ioniza-
tion time-of-flight mass spectrometry of nitrated and oxygenated PAHs found
in air and soot particles, Anal. Bioanal. Chem. 397 (2010) 243–256, http://dx.
doi.org/10.1007/s00216-009-3416-2.
[31] T. Letzel, U. Pöschl, E. Rosenberg, M. Grasserbauer, R. Niessner, In-
source fragmentation of partially oxidized mono- and polycyclic aro-
matic hydrocarbons in atmospheric pressure chemical ionization mass
spectrometry coupled to liquid chromatography, Rapid Commun. Mass
Spectrom. 13 (245) (1999) 6–2468, http://dx.doi.org/10.1002/(SICI)1097-
0231(19991230)13:24<2456::AID-RCM812>3.0.CO;2-2.
[32] C. Schauer, R. Niessner, U. Pöschl, Analysis of nitrated polycyclic aromatic
hydrocarbons by liquid chromatography with fluorescence and mass spec-
trometry detection: air particulate matter, soot, and reaction product studies,
Anal. Bioanal. Chem. 378 (2004) 725–736, http://dx.doi.org/10.1007/s00216-
003-2449-1.
[33] T. Karancsi, P. Slegel, Reliable molecular mass determination of aromatic nitro
compounds: elimination of gas-phase reduction occurring during atmospheric
pressure chemical ionization, J. Mass Spectrom. 34 (1999) 975–977.
[34] E.A. Straube, W. Dekant, W. Völkel, Comparison of electrospray ioniza-
tion, atmospheric pressure chemical ionization, and atmospheric pressure
photoionization for the analysis of dinitropyrene and aminonitropyrene
LC–MS/MS, J. Am. Soc. Mass Spectrom. 15 (2004) 1853–1862, http://dx.doi.
org/10.1016/j.jasms.2004.08.017.
[35] T.T. Williams, H. Perreault, Selective detection of nitrated polycyclic aromatic
hydrocarbons by electrospray ionization mass spectrometry and constant
neutral loss scanning, Rapid Commun. Mass Spectrom. 14 (2000) 1474–1481,
http://dx.doi.org/10.1002/1097-0231(20000830)14:16<1474::AID-RCM46>3.
0.CO;2-Z.
[36] K. Miet, K. Le Menach, P.M. Flaud, H. Budzinski, E. Villenave, Heterogeneous
reactions of ozone with pyrene, 1-hydroxypyrene and 1-nitropyrene adsorbed
on particles, Atmos. Environ. 43 (369) (2009) 9–3707, http://dx.doi.org/10.
1016/j.atmosenv.2009.04.032.
[37] Y. Zhang, B. Yang, J. Meng, S. Gao, X. Dong, J. Shu, Homogeneous and het-
erogeneous reactions of phenanthrene with ozone, Atmos. Environ. 44 (2010)
697–702, http://dx.doi.org/10.1016/j.atmosenv.2009.11.017.
[38] J.-J. Yao, Z.-H. Huang, S.J. Masten, The ozonation of pyrene: pathway and prod-
uct identification, Water Res. 32 (1998) 3001–3012.
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
(
2005) 508–519, http://dx.doi.org/10.1007/s00216-004-2883-8.
Please cite this article in press as: R.E. Cochran, et al., Int. J. Mass Spectrom. (2016), http://dx.doi.org/10.1016/j.ijms.2016.01.001