Low-energy collision-induced fragmentation of negative ions derived from diesters of aliphatic dicarboxylic acids made with hydroxybenzoic acids

Article abstract of DOI:10.1002/jms.1426

Diesters of ortho-hydroxybenzoic acid (salicylic acid) made with glutaric, adipic, and pimelic acids are the monomers of some potential drug candidates for aspirin patches. Collision-induced dissociation (CID) spectra of negative ion derived from these compounds show a 120-Da 'neutral loss' specific to the ortho isomers. In contrast, the anions derived from diesters of meta- and para-hydroxybenzoic acids show a 138-Da loss for an elimination of elements of hydroxybenzoic acid by a charge-remote mechanism. Deuterium labeling studies confirmed that the hydrogen atom transferred for hydroxybenzoic acid loss originates specifically from the α position of the dicarboxylic acid moiety. Although all spectra showed a peak at m/z 137, a charge-mediated process specific for the ortho compounds renders it the most prominent peak in the spectra of ortho compounds. Appropriate deuterium labeling experiments demonstrated that the hydrogen atom transferred for the formation of the m/z 137 ion in ortho compounds is specifically derived from the α position of the dicarboxylic acid moiety. Copyright

Full text of DOI:10.1002/jms.1426

JOURNAL OF MASS SPECTROMETRY
J. Mass Spectrom. 2008; 43: 1502–1511
Published online 02 June 2008 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/jms.1426
Low-energy collision-induced fragmentation
of negative ions derived from diesters of aliphatic
dicarboxylic acids made with hydroxybenzoic acids
Upul Nishshanka and Athula B. Attygalle∗
Center for Mass Spectrometry, Department of Chemistry and Chemical Biology, Stevens Institute of Technology, Hoboken, NJ 07030, USA
Received 14 February 2008; Accepted 17 April 2008
Diesters of ortho-hydroxybenzoic acid (salicylic acid) made with glutaric, adipic, and pimelic acids are the
monomers of some potential drug candidates for aspirin patches. Collision-induced dissociation (CID)
spectra of negative ion derived from these compounds show a 120-Da ‘neutral loss’ specific to the ortho
isomers. In contrast, the anions derived from diesters of meta- and para-hydroxybenzoic acids show
a 138-Da loss for an elimination of elements of hydroxybenzoic acid by a charge-remote mechanism.
Deuterium labeling studies confirmed that the hydrogen atom transferred for hydroxybenzoic acid loss
originates specifically from the a position of the dicarboxylic acid moiety. Although all spectra showed a
peak at m/z 137, a charge-mediated process specific for the ortho compounds renders it the most prominent
peak in the spectra of ortho compounds. Appropriate deuterium labeling experiments demonstrated that
the hydrogen atom transferred for the formation of the m/z 137 ion in ortho compounds is specifically
derived from the a position of the dicarboxylic acid moiety. Copyright  2008 John Wiley & Sons, Ltd.
KEYWORDS: aspirin; transdermal drug delivery; CID; negative ions; fragmentation; even-electron ions
by negative-ion electrospray ionization mass spectrometry
(ESI-MS). However, before developing validated LC/MS
procedures, such as those required for characterizing impu-
rities in drug formulations and metabolites in biological
fluids, the underlying mass spectrometric fragmentation
mechanisms of the compounds of interest should be prop-
erly rationalized. Some of the putative impurities that could
be present in the monomeric compounds derived from sal-
icylic acid and short-chain dicarboxylic acids are the esters
of other isomers of hydroxybenzoic acid. In our pursuit
for a deeper understanding of collision-induced fragmen-
tation of low-molecular-weight anions,4 we noted that the
spectra recorded from dicarboxylic acid esters of the meta-
and para-hydroxybenzoic acids show unique peaks that are
absent in those obtained from the ortho isomers. The occur-
rence of such position-specific fragment ions enables the
development of validated liquid-chromatographic tandem-
mass-spectrometric methods for the determination of the
purity of these therapeutically significant compounds.
INTRODUCTION
Aspirin is a widely used drug with many pharmacological
properties. It reduces pain and swelling by inhibiting
an enzyme called cyclooxygenase-2 (COX-2).1 However,
aspirin irritates the stomach lining and sometimes causes
bleeding.2 The pharmacologically active ingredient of aspirin
is salicylic acid. Delivering the active drug transdermally
avoids undesired complications observed with aspirin. Many
novel biodegradable polymers such as poly(anhydride-
esters) have been synthesized from salicylic acid, and their
efficacy has been evaluated as prodrugs for patch systems.3
Slow hydrolytic cleavage of the prodrug releases the active
ingredient, salicylic acid, to the blood stream (Scheme 1).
Typical monomeric candidates for the synthesis of the
poly(anhydride-esters) are the diesters derived from short-
chain dicarboxylic acids and salicylic acid.
OH
O
O
OH
O
O
(CH2)n
EXPERIMENTAL
O
O
n = 3, 4, 5....
Chemicals
Glutaric, adipic, and pimelic acids, o-, m-, and
p–hydroxybenzoic acids, and all other reagents were pur-
chased from Aldrich Chemical Co. (St Louis, MO). All
synthesized products were characterized by GC/MS on an
HP 5989B quadrupole mass spectrometer linked to a Hewlett
Packard 5890 Series II gas chromatograph, and ESI/MS on a
Micromass Quattro I triple quadrupole mass spectrometer.
Facile ionization afforded by the carboxylic group renders
these monomeric compounds ideal candidates for analysis
ŁCorrespondence to: Athula B. Attygalle, Center for Mass
Spectrometry, Department of Chemistry and Chemical Biology,
Stevens Institute of Technology, Hoboken, NJ 07030, USA.
E-mail: athula.attygalle@stevens.edu
Copyright  2008 John Wiley & Sons, Ltd.

CID of hydroxybenzoic acid diester anions
1503
O
O
O
O
OH
OH
O
O
O
O
(CH2)n
H2O
2m
+
m
HO
OH
(CH2)n
O
O
m
n = 3, 4, 5....
Scheme 1. Hydrolytic degradation of poly(anhydride-esters).
Synthesis of [2,2,4,4-2H4]butanedioic acid
Mass spectrometry
Collision-induced dissociation (CID) mass spectra were
recorded on a Micromass Quattro I spectrometer equipped
with an ESI source. Samples were introduced to the source
as acetonitrile–water–ammonia (v/v, 8 : 2 : 0.01) solutions
for negative-ion MS at a flow rate of 320 µl hꢀ1. The source
Hydrochloric acid (1.5 ml, 17.96 mmol) was added to a
[2,2,4,4-2H4]butanedinitrile (350 µl, 3.00 mmol) and the mix-
°
ture was kept at 100 C for 6 h. The product was isolated
under reduced pressure, and the white solid formed was
used for the next step.9
°
temperature was held at 80 C. The argon gas pressure in
Synthesis of 1,6-bis(m-carboxyphenoxy)-[2ꢀ,2ꢀ,5ꢀ,5ꢀ-
2H4]hexanoate
the collision cell was adjusted to attenuate precursor ion
transmission by 50%. Laboratory-frame collision energy was
optimized for each experiment.
1,6-bis(m-Carboxyphenoxy)-[20,20,50,50-2H4]hexanoate was
made by using [2,2,4,4-2H4]butanedinitrile by a procedure
similar to that described for the nonlabeled ester.
Chemical synthesis
Each mono- or dicarboxylic acid (100 mg) was dissolved in
Synthesis of octanoyl ester of 2-hydroxybenzoic
acid
thionyl chloride (2 ml) and two drops of dimethylformamide
Octanoyl chloride (125 µl, 0.730 mmol) in THF (0.5 ml) was
added slowly to a mixture of 2-hydroxybenzoic acid (100 mg,
0.725 mmol) in THF (2 ml) and pyridine (70 µl, 0.84 mmol).
The mixture was stirred for 2 h at RT. After adding water
(2 ml), the mixture was acidified with 1% HCl. The product
was extracted into the ethyl acetate.
3
°
(DMF). The mixture was kept at 60 C for 4 h, and excess
thionyl chloride was removed by a stream of N2. Each ester
was prepared from 1-, 2-, or 3-hydroxybenzoic acid and an
appropriate diacid chloride in the presence of pyridine.5 For
example, ester 1 (n D 1) was synthesized by adding glutaryl
chloride (40 µl, 0.3 mmol) in tetrahydrofuran (THF) (0.5 ml)
slowly to a mixture of salicylic acid (85 mg, 0.6 mmol) in THF
(2 ml) and pyridine (100 µl, 1.2 mmol). After the mixture was
stirred for 4 h, water (2 ml) was added and the mixture was
acidified with dil HCl. The white solid formed was isolated
by vacuum filtration and characterized by ESI-MS. Yields
were 40–50%.
RESULTS AND DISCUSSION
A comparison of CID spectra of negative ions derived
from esters made from ortho, meta, and para isomers of
hydroxybenzoic acid with dicarboxylic acids such as glutaric,
adipic, and pimelic acids showed that the spectra of ortho
isomers can be readily distinguished from those of meta and
para isomers. For example, the spectra of disalicyl glutarate
(1), adipate (2), and pimelate (3) show peaks at m/z 251,
265, and 279, respectively, for a unique 120-Da ‘neutral loss’
from the precursor anion (Figs 1(A), 2(A), and 3(A)). The
intensity of this ortho-specific peak increases as the size of
the dicarboxylic acid residue increases (Although the exact
reason is not clear, it appears that the molecule achieves
the conformation necessary for the fragmentation readily
when the number of methylene groups in the dicarboxylic
acid moiety increases). A charge-remote10–13 (Scheme 2) or
a charge-mediated (Scheme 3) mechanism can be postulated
to rationalize this 120-Da loss, which appears to represent an
elimination of elements of benzyne and CO2.
Synthesis of dibromobutane
THF (2 ml, 0.025 mol) was mixed with PBr3 (2 ml, 0.02 mol)
and the mixture was kept at 50 C for 2 h. The mixture was
washed several times with 1% NaOH and the organic layer
was extracted into diethyl ether.6
°
Synthesis of butanedinitrile
Dibromobutane (600 µl, 5.00 mmol) in 95% ethanol was
slowly added to a mixture of KCN (700 mg, 10.75 mmol)
in water. The reaction mixture was kept overnight at 70 C,
and the product was extracted into diethyl ether.7
°
In order to evaluate which mechanism predominates, we
recorded the spectrum of salicyloctanoate. Since a peak for
the octanoate anion is observed at m/z 143 in the spectrum
of salicyloctanoate (Fig. 4), it is clear that the 120-Da loss can
occur even without the participation of the charge-remote
pathway (Scheme 4).
Moreover, tandem mass spectrometric experiments con-
ducted with the precursor anions derived from disali-
cylate derivative of dicarboxylic acids demonstrated that
Synthesis of [2,2,4,4-2H4]butanedinitrile
Butanedinitrile (500 µl, 4.44 mmol) was added to a mixture
of D2O/dioxane (7 : 2) and diazabicyclo[5.4.0]undec-7-ene
(DBU, 0.5 equiv.). The mixture was heated at 100 C for 2 h
and the solvent was removed under reduced pressure. After
two repetitions of the deuterium-exchange procedure, the
residue was washed with aqueous HCl and the product was
extracted into diethyl ether.8
°
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

1504
U. Nishshanka and A. B. Attygalle
137
O
O
100
(A)
O
O
O
O
O
O
O
O
O
HO
O
O
120 Da
HO
O
%
HO
371
251**
257
0
m/z
400
40
80
120
160
200
240
280
320
360
371
100
(B)
O
O
O
O
138 Da
O
O
O
O
O
%
C
O
233
O
O
O
HO
HO
O
137*
HO
93
0
m/z
40
80
120
160
200
240
280
320
360
371
400
100
(C)
O
O
O
O
O
O
O
O
O
O
138 Da
%
O
O
C
O
HO
OH
OH
O
93
137*
233
240
m/z
0
40
80
120
160
200
280
320
360
400
Figure 1. Product ion spectra of m/z 371 [M ꢀ H]ꢀ ion derived from 1,5-bis(o-carboxyphenoxy)pentanoate (A),
1,5-bis(m-carboxyphenoxy)pentanoate (B), and 1,5-bis(p-carboxyphenoxy)pentanoate (C). [All parameters including cone voltage
(20 V), collision gas pressure, and collision energy (15 eV) were kept identical] (depending on the mechanistic pathway of its origin,
the negative charge on the m/z 137Ł ion could be on either the carboxyl or the hydroxyl group; the charge on the m/z 251ŁŁ ion
could be on any one of the carboxyl groups).
after the initial 120-Da loss, the resulting product ion
undergoes a further120-Da loss. For example, the frag-
ment ion spectrum of m/z 265 ion (5 or 8) derived
from 1,6-bis(o-carboxyphenoxy)hexanoate (2) shows a peak
at m/z 145 (Fig. 5(B)). Moreover, a spectrum recorded
from the m/z 386 ion derived from a sample of 6-bis(o-
carboxyphenoxy)hexanoate (2) dissolved in D2O showed
a corresponding peak at m/z 266, also for a 120 Da loss
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

CID of hydroxybenzoic acid diester anions
1505
137
100
(A)
O
O
120 Da
O
O
O
O
O
HO
O
O
O
O
%
OH
265**
385
O
O
OH
O
257
m/z
0
40
80
120
160
200
240
280
320
360
360
360
400
385
100
O
(B)
O
O
O
O
O
O
O
O
O
138 Da
HO 137*
%
C
O
O
O
247
OH
O
HO
93
80
m/z
0
40
120
160
200
240
280
320
400
385
100
O
(C)
O
O
O
O
O
O
O
138 Da
O
O
%
O
O
C
HO
O
137*
HO
O
247
240
HO
93
80
0
m/z
40
120
160
200
280
320
400
Figure 2. Product ion spectrum of m/z 385 [M ꢀ H]ꢀ ion derived from 1,6-bis(o-carboxyphenoxy)hexanoate (A),
1,6-bis(m-carboxyphenoxy)hexanoate (B), and 1,6-bis(p-carboxyphenoxy)hexanoate (C). [All parameters including cone voltage
(20 V), collision gas pressure, and collision energy (15 eV) were kept identical.] (Depending on the mechanistic pathway of its origin,
the negative charge on the m/z 137Ł ion could be on either the carboxyl or the hydroxyl group; the charge on the m/z 265ŁŁ ion
could be on any one of the carboxyl groups).
(Fig. 5(A)). However, when the m/z 266 ion was generated
by in-source fragmentation and subjected to CID, a peak
was observed at m/z 146 confirming that the second 120-
Da loss may originate from a charge-remote (Scheme 5)
or a charge-mediated mechanism (Scheme 6) (Fig. 5(C)).
Although it is experimentally difficult to distinguish between
the two mechanisms, the charge-remote mechanism should
play a more dominant role for the second 120-Da loss if the
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

1506
U. Nishshanka and A. B. Attygalle
O
O
137
100
(A)
O
120 Da
O
O
O
O
O
O
HO
O
O
O
279**
HO
399
%
O
HO
O
257
m/z
0
40
80
120
160
200
240
280
320
360
360
360
400
399
100
O
(B)
O
O
O
O
O
O
138 Da
O
O
O
O
%
C
261
O
HO
137*
O
HO
O
93
HO
m/z
0
40
80
120
160
200
240
280
320
400
399
100
(C)
O
O
O
137*
O
O
O
O
O
O
O
O
138 Da
HO
%
C
O
O
261
O
HO
HO 93
m/z
0
40
80
120
160
200
240
280
320
400
Figure 3. Product ion spectra of m/z 399 [M ꢀ H]ꢀ ion derived from 1,7-bis(o-carboxyphenoxy)heptanoate (A),
1,7-bis(m-carboxyphenoxy)heptanoate (B), and 1,7-bis(p-carboxyphenoxy)heptanoate (C). [All parameters including cone voltage
(20 V), collision gas pressure, and collision energy (15 eV) were kept identical] (depending on the mechanistic pathway of its origin,
the negative charge on the m/z 137Ł ion could be on either the carboxyl or the hydroxyl group; the charge on the m/z 279ŁŁ ion
could be on any one of the carboxyl groups).
m/z 265 anion is generated more from the charge-mediated
mechanism as illustrated in Scheme 3. According to one
school of thought, labile protons are mobile within gaseous
ionic structures.14–16 Although it may be possible that the
labile proton on the aromatic carboxyl group in Structures
8 and 10 rearranges to those shown in Structures 5 and
13, respectively, we presume that the m/z 265 is better
represented as a composite of both Structures 5 and 8.
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

CID of hydroxybenzoic acid diester anions
1507
O
O
O
O
O
O
H
O
O
120 Da
HO
O
( )n
( )n
CO2
+
+
O
O
O
O
1, n = 1, m/z 371
2, n = 2, m/z 385
3, n = 3, m/z 399
4, n = 1, m/z 251
5, n = 2, m/z 265
6, n = 3, m/z 279
Scheme 2. Proposed charge-remote mechanism for the 120-Da loss from [M ꢀ H]ꢀ of ortho compounds.
O
OH
O
O
O
OH
O
O
O
120 Da
O
( )n
( )n
CO2
+
+
O
O
O
O
1, n = 1, m/z 371
2, n = 2, m/z 385
3, n = 3, m/z 399
7, n = 1, m/z 251
8, n = 2, m/z 265
9, n = 3, m/z 279
Scheme 3. Proposed charge-mediated mechanism for the 120-Da loss from [M ꢀ H]ꢀ of ortho compounds.
hydrogen atom from the ˛ position was proven by recording
a spectrum 2-carboxyphenyl [1,1-2H2]decanoate that showed
its most intense peak exclusively at m/z 138.
137
HO
100
O
O
120 Da
Although the intensity is lower than that observed for the
ortho compounds, the presence of a significant peak at m/z
137 in the spectra of meta and para compounds indicate that
it could originate from charge-remote mechanism as well
(Figs 1(B) and (C), 2(B) and (C), and 3(B) and (C)). On the
other hand, the intensity of the m/z 137 peak in the spectra
of meta- and para-carboxyphenoxy octanoate is either very
minute or absent (data not shown), which indicated that at
least for the monosubstituted compounds the charge-remote
elimination of a ketene unit plays little role in the formation
of the m/z 137 ion.
The most striking feature in the spectra of negative
ions derived from para and meta isomers is the absence
of a peak for the 120-Da loss. Instead, both meta and para
compounds show peaks for an expulsion of a 138-Da unit
from the corresponding precursor anion (Figs 1, 2, and 3).
To eliminate a 138-Da molecule of hydroxybenzoic acid, the
ester bond should be cleaved and a hydrogen atom must
be transferred to the oxygen atom attached to the benzene
ring. To rationalize the elimination of a hydroxybenzoic
acid, the following charge-remote mechanism shown in
Scheme 8 was postulated. In this mechanism, an ˛-hydrogen
atom is specifically transferred to the phenolic oxygen atom
(Scheme 8).
%
O
O
O
O
263
O
143
O
m/z
0
40
80
120
160
200
240
280
Figure 4. Product ion spectra of m/z 263 [M ꢀ H]ꢀ ion derived
from salicyloctanoate (cone voltage 20 V; collision energy
15 eV).
Spectra of all carbophenoxy esters examined in this study
showed a prominent peak at m/z 137. Many pathways
are possible for the formation of this ion, for example, a
mechanism analogous to that illustrated in Scheme 9. The
most favorable pathway however appears to be the direct
charge-mediated cleavage (Scheme 7). This is supported
by the observation that the base peak of the product ion
spectrum of m/z 263 from salicyl octanoate (Fig. 4), and
that of m/z 371 from 1,5-bis(o-carboxyphenoxy)pentanoate
(Fig. 1(A)) is observed at m/z 137. The specific transfer of a
O
O
O
O
CO2
+
+
O
O
m/z 143
m/z 263
Scheme 4. Proposed charge-mediated mechanism for the 120-Da loss from m/z 263 [M ꢀ H]ꢀ of salicyloctanoate.
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

1508
U. Nishshanka and A. B. Attygalle
138
100
(A)
OD
O
O
O
O
O
O
O
O
120 Da
O
O
386
O
%
OD
O
137
266*
O
O
O
OD
O
HO
m/z
m/z
m/z
0
40
80
120
137
160
200
240
280
320
360
400
100
(B)
O
O
O
O
O
HO
O
120 Da
%
O
OH
O
O
145
265*
O
HO
0
40
80
120
137
160
200
240
280
320
360
400
100
(C)
O
O
O
O
HO
O
O
120 Da
%
O
OD
O
O
O
OD
146
266*
O
138
O
DO
0
40
80
120
160
200
240
280
320
360
400
Figure 5. (A) Product ion spectra of m/z 386 [M ꢀ D]ꢀ derived from 1,6-bis(o-carboxyphenoxy)hexanoate sprayed as a D2O solution
(cone voltage 20 V; collision energy 15 eV). (B) Product ion spectrum of m/z 265 generated by in-source fragmentation of
1,6-bis(o-carboxyphenoxy)hexanoate (cone voltage 30 V; collision energy 20 eV). (C) Product ion spectrum of m/z 266 generated by
in-source fragmentation of 1,6-bis(o-carboxyphenoxy)hexanoate sprayed as a D2O solution (cone voltage 30 V; collision energy
20 eV). (ŁDepending on the mechanistic pathway of its origin, the negative charge could be on any one of the carboxyl groups).
To verify the proposed mechanism, 1,6-bis(m-
carboxyphenoxy)-[20,20,50,50-2H4]hexanoate was synthesized
and a product ion spectrum of its m/z 389 anion (19) was
recorded (Fig. 6). The spectrum showed a peak at m/z 250 for
a specific loss of a 139-Da unit confirming the transfer of an ˛-
hydrogen atom by the charge-remote mechanism (Scheme 8).
On the other hand, the presence of a peak at m/z 137 even
in the spectrum of the tetradeuterio compound indicated
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

CID of hydroxybenzoic acid diester anions
1509
O
O
R
O
O
120 Da
RO
O
( )2
( )2
CO2
+
+
O
O
O
O
11, R = H, m/z 145
12, R = D, m/z 146
8, R = H, m/z 265
10, R = D, m/z 266
Scheme 5. Proposed charge-remote mechanism for the loss of 120 Da from 8 and 10.
O
O
O
OR
120 Da
O
OR
( )2
( )2
CO2
+
+
O
O
O
O
14, R = H, m/z 145
15, R = D, m/z 146
5, R = H, m/z 265
13, R = D, m/z 266
Scheme 6. Proposed charge-mediated mechanism for the loss of 120 Da from 5 and 13.
O
O
O
OH
H
O
O
C
+
O
126 Da
O
m/z 263
m/z 137
Scheme 7. Proposed mechanism for the formation of m/z 137 ion from salicyl octanoate.
H
OH
O
O
O
( )n
( )n
C
O
+
O
O
O
138 Da
O
OH
O
OH
O
O
O
O
16, n = 1, m/z 233
17, n = 2, m/z 247
18, n = 3, m/z 261
1, n = 1, m/z 371
2, n = 2, m/z 385
3, n = 3, m/z 399
Scheme 8. Proposed charge-remote mechanism for a loss of 138 Da from m/z 385 [M ꢀ H]ꢀ of 1,6-bis(m-carboxyphenoxy)hexanoate.
that there are other pathways that generate the m/z 137
ion in diesters. To rationalize this observation, we propose
the charge-assisted mechanism illustrated in Scheme 9. This
mechanism can also rationalize the observation that the
intensity of the m/z 137 peak increases with the increase of
the number of methylene groups in the dicarboxylate moiety
(Figs 1, 2, and 3).
Moreover, the expulsion of a cyclic anhydride by the
mechanism illustrated in Scheme 9 appears to be common
to all the diesters used in this study. In fact, this mechanism
justifies the presence of the m/z 138 peak as the base peak
in the fragment ion spectrum of the m/z 386 anion derived
from 1,6-bis(o-carboxyphenoxy)hexanoate sprayed as a D2O
solution (Fig. 5(A)). The peak m/z 137, on the other hand,
originates from a charge-mediated hydrogen atom transfer
from the ˛ position (Scheme 7). Thus, we conclude that
the m/z 137 ions observed from meta and para diesters are
composites in which the negative charge is located either on
the carboxyl group or the hydroxyl group).
CONCLUSIONS
Under CID conditions, the benzoate anion and most of its
ring-substituted derivatives are known to undergo a facile
decarboxylation to produce a phenyl anion.17,18 Although
none of the anions derived from carboxylate diesters of this
study underwent a decarboxylation (Figs 1, 2, and 3), other
specific fragmentation these compounds endure provide
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

1510
U. Nishshanka and A. B. Attygalle
O
389
O
100
O
HO
O
O
DO
138
139 Da
137
O
O
O
O
O
D
D
%
D
O
D
O
O
D
D
D
O
250
93
HO
DO
OH
C
94
O
O
m/z
0
40
80
120
160
200
240
280
320
360
400
Figure 6. Product ion spectra of m/z 389 [M ꢀ H]ꢀ ion derived from 1,6-bis(m-carboxyphenoxy)-[20,20,50,50-2H4]hexanoate. (cone
voltage 20 V; collision energy (15 eV).
COOH
COOH
O
O
O
O
O
COOH
CD2
O
O
O
O
O
CD2
CD2
O
CD2
+
O
CD2
O
O
m/z 137
CD2
O
O
O
(20)
O
m/z 389 (19)
Scheme 9. Proposed charge-assisted mechanism for the formation of the m/z 137 ion. From the m/z 389 [M ꢀ H]ꢀ of
1,6-bis(m-carboxyphenoxy)-[20,20,50,50-2H4]hexanoate.
from ortho-, meta-, and para-hydroxyphenyl carbaldehydes,
ketones, and related compounds. Journal of Mass Spectrometry
2007; 42: 1207.
5. Prudencio A, Schmeltzer RC, Uhrich KE. Effect of linker
structure on salicylic acid-derived poly(anhydride-esters).
Macromolecules 2005; 38: 6895.
6. Ispiryan RM, Belen’kii LI, Gol’dfarb YL. Cleavage of the
tetrahydrofuran ring by acid halides and acid anhydrides.
Russian Chemical Bulletin 1967; 16: 2391.
7. Marvel CS, McColm EM. Trimethylene Cyanide. Organic
Synthesis Collective 1941; 1: 536.
8. Grossert JS, Cook MC, White RL. The influence of
structural features on facile McLafferty-type, even-electron
rearrangements in tandem mass spectra of carboxylate anions.
Rapid Communications in Mass Spectrometry 2006; 20: 1511.
9. Prout FS, Aguilar VN, Girard FH, Lee DD, Shoffner JP. ˛-Ethyl-
˛-methylsuccinic acid. Organic Synthesis Collective 1973; 5: 572.
10. Jensen NJ, Tomer KB, Gross ML. Gas-phase ion decompositions
occurring remote to a charge site. Journal of the American Chemical
Society 1985; 107: 1863.
ample information to distinguish the ortho compounds
from those of the meta and para isomers. The spectral
interpretations described in this article will be useful in
purity evaluations and metabolic studies of poly(anhydride-
esters) made from salicylic acid, which are being evaluated
as prodrugs for aspirin patch systems.
Acknowledgements
We thank Kathryn E. Uhrich (Rutgers University) for drawing our
attention to this group of compounds. This research was supported
by funds provided by Stevens Institute of Technology, NJ.
REFERENCES
1. Schwab JM, Chiang N, Arita M, Serhan CN. Resolvin E1 and
protectin D1 activate inflammation-resolution programmes.
Nature 2007; 447: 689.
2. Okabe S, Takeuchi K, Honda K, Takagi K. Effects of
acetylsalicylic acid (ASA), ASA plus L Glutamine and
Glutamine on healing of chronic gastric ulcer in the rat. Digestion
1976; 14: 85.
L
11. Gross ML. Charge-remote fragmentation: an account of research
on mechanisms and applications. International Journal of Mass
Spectrometry 2000; 200: 611.
3. Erdmann I, Uhrich KE. Synthesis and degradation characteris-
tics of salicylic acid-derived poly(anhydride-esters). Biomaterials
2000; 21: 1941.
4. Attygalle AB, Ruzicka J, Varughese D, Bialecki JB, Jafri S. Low-
energy collision-induced fragmentation of negative ions derived
12. Tomer KB, Crow FW, Gross ML. Location of double bond
position in unsaturated fatty acids by negative ion MS/MS.
Journal of the American Chemical Society 1983; 105: 5487.
13. Chen G, Pramanik BN, Bartner PL, Saksena AK, Gross ML.
Multiple-stage mass spectrometric analysis of complex
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms

CID of hydroxybenzoic acid diester anions
1511
oligosaccharide antibiotics (Everninomicins) in a quadrupole
ion trap. Journal of the American Society for Mass Spectrometry
2002; 13: 1313.
16. Paizs B, Suhai S. Fragmentation pathways of protonated
peptides. Mass Spectrometry Reviews 2005; 24: 508.
17. Attygalle AB, Kharbatia N, Bialecki J, Ruzicka J, Svatosˇ A,
Stauber EJ. An unexpected ion-molecule adduct in negative-
ion collision-induced decomposition ion-trap mass spectra
of halogenated benzoic acids. Rapid Communications in Mass
Spectrometry 2006; 20: 2265.
18. Bandu ML, Watkins KR, Bretthauer ML, Moore CA, Desaire H.
Prediction of MS/MS data. 1. A focus on pharmaceuticals
containing carboxylic acids. Analytical Chemistry 2004; 76: 1746.
14. Wysocki VH, Tsaprailis G, Smith LL, Breci LA. Mobile and
localized protons: a framework for understanding peptide
dissociation. Journal of Mass Spectrometry 2000; 35: 1399.
15. Dongre´ AR, Jones JL, Somogyi A, Wysocki VH. Influence
of peptide composition, gas-phase basicity, and chemical
modification on fragmentation efficiency: Evidence for the
mobile proton model. Journal of the American Chemical Society
1996; 118: 8365.
Copyright  2008 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2008; 43: 1502–1511
DOI: 10.1002/jms