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,
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