Characterization of free radical spin adducts of 5-diisopropyloxy- phosphoryl5-methyl-1-pyrroline-A/-oxide using mass spectrometry and 31P nuclear magnetic resonance

Article abstract of DOI:10.1255/ejms.1062

5-Diisopropyloxy-phosphoryl-5-methyl-1-pyrroline-N-oxide;(DIPPMPO) was used to trap a variety of free radicals and the stable compounds generated by the natural decomposition of the initially formed spin adducts were characterized by ³¹P nuclear magnetic resonance (NMR) and mass spectrometry. Initially, the starting spin trap DIPPMPO was completely characterized using GC-MS and its fragmentation pathway was studied in detail. Then, DIPPMPO was used to trap an oxygen-centered free radical (the hydroxyl radical 'OH) and two carbon-centered free radicals (methyl CH₃ and 1-phenyl-ethanol-1-yl CCH₃(OH)Ph radicals). The ³¹P NMR signals were thus assigned and the structures of adducts were studied and confirmed by mass spectrometry. Overall, the fragmentation pathways of the radical adducts proceed mainly via the loss of the diisopropyloxy(oxido)phosphoranyl radical. For the specific case of trapping-OH radicals, it is possible to visualize the rearrangement of the nitroxide radical adduct to its nitrone form as invoked in the literature. This spin trapping technique, coupled with ³¹P NMR and MS, provides a tool for the identification of short-lived and low molecular weight free radicals present in a variety of processes. IM Publications LLP 2010 All rights reserved.

Full text of DOI:10.1255/ejms.1062

L. Zoia and D.S. Argyropoulos, Eur. J. Mass Spectrom. 16, 175–185 (2010)
175
received: 31 July 2009
n
revised: 5 December 2009 n Accepted: 9 December 2009 n Publication: 11 January 2010
EuroPEAN
JourNAL
of
MASS
SPEctroMEtry
Characterization of free radical spin
adducts of 5-diisopropyloxy-phosphoryl-
5-methyl-1-pyrroline-N-oxide using mass
spectrometry and 31P nuclear magnetic
resonance
Luca Zoiaa and Dimitris S. Argyropoulosb,c*
aDipartimento di Scienze dell’Ambiente e del territorio, università degli Studi di Milano-Bicocca, Piazza della Scienza 1, 20126, Milano, Italy.
E-mail: luca.zoia@unimib.it
borganic chemistry of Wood components Laboratory, Department of forest Biomaterials, North carolina State university, Nc 123321, uSA.
E-mail: dsargyro@ncsu.edu
cLaboratory of organic chemistry, Department of chemistry, university of Helsinki, Po Box 55, 00014 Helsinki, finland
5-Diisopropyloxy-phosphoryl-5-methyl-1-pyrroline-N-oxide (DIPPMPO) was used to trap a variety of free radicals and the stable
compounds generated by the natural decomposition of the initially formed spin adducts were characterized by 31P nuclear magnetic
resonance (NMR) and mass spectrometry. Initially, the starting spin trap DIPPMPO was completely characterized using GC-MS and its
fragmentation pathway was studied in detail. Then, DIPPMPO was used to trap an oxygen-centered free radical (the hydroxyl radical
•OH) and two carbon-centered free radicals (methyl •CH3 and 1-phenyl-ethanol-1-yl •CCH3(OH)Ph radicals). The 31P NMR signals were
thus assigned and the structures of adducts were studied and confirmed by mass spectrometry. Overall, the fragmentation pathways
of the radical adducts proceed mainly via the loss of the diisopropyloxy(oxido)phosphoranyl radical. For the specific case of trapping∙OH
radicals, it is possible to visualize the rearrangement of the nitroxide radical adduct to its nitrone form as invoked in the literature. This
spin trapping technique, coupled with 31P NMR and MS, provides a tool for the identification of short-lived and low molecular weight free
radicals present in a variety of processes.
Keywords: free radicals, spin trapping, DIPPMPO, mass spectrometry, GC-MS, 31P NMR
Introduction
for many years, spin traps have been used to decrease the
reactivity of free radicals in order for them to be identified
and detected by electron paramagnetic resonance (EPr)
spectrometry, thereby allowing the acquisition of abundant
bond of a diamagnetic compound (the spin trap) to form a
more stable radical (the spin adduct) which can be detected by
EPr spectrometry.5–9 Nitrones are the most frequently used
compounds as spin trapping agents, and the spin adducts
information on the production of such species in biological, being, in this case, a nitroxide. recent accounts10 have
biochemical and chemical systems.1–4 During this technique, demonstrated that phosphorus-containing spin traps give
a highly reactive free radical typically reacts with a double
rise to radical adducts that have longer half-lives compared
ISSN: 1469-0667
doi: 10.1255/ejms.1062
© IM Publications LLP 2010
All rights reserved

176
Characterization of Free Radical Spin Adducts of DIPPMPO
to other spin traps. this fact can be used to expand the capa- or nitrones obtained by oxido-reductive processes have been
bility of EPr spectrometry. these radical adducts eventu- shown to be readily detected.15–21 recently, a novel nitrone spin
ally degrade with time, becoming diamagnetic and, therefore, trap containing a phosphorous atom, namely 5-diisopropoxy-
EPr-undetectable. Moreover, a given set of EPr parameters
may not always clearly characterize a particular spin adduct
and the EPr-spin trap techniques allow the determination of
the general type of radical trapped. However, the presence of
phosphoryl-5-methyl-1-pyrroline-N-oxide (DIPPMPo) was
investigated.30 figure 1 shows the DIPPMPo spin trap system
and its operational chemistry.
Initially, a given radical specie (r•) reacts with the double
phosphorus within these systems allows for the use of phos- bond of the nitrone spin trap DIPPMPo a to form the paramag-
phorus nuclear magnetic resonance (31P NMr) spectrometry
to investigate the detailed chemistry of radical reactions. this
technique was termed “NMr spin trapping” by Khramtsov
et al.11 31P NMr could be exploited to perform quantitative
analyses in the presence of a suitable internal standard and
qualitative analyses since the chemical shift of the 31P atom
was found to be strongly dependent on the nature of the
netic adduct b which is EPr detectable. the nitroxide para-
magnetic species b decays with time via different reactions
(unimolecular or/and bimolecular) such as oxidation, reduc-
tion, dismutation and rearrangement. In many case the prod-
ucts of the decomposition reactions are the corresponding
hydroxylamine c and nitrone d. these species (c and d) are
actually 31P NMr detectable and the chemical shift of the
adducts.11–14 to further confirm the identity of the spin adducts, phosphorous atom has been shown to be related to the nature
complementary analytical techniques that provide additional
information about the structure of such adducts need to be
of the radical being trapped.12–14 Moreover, compounds c and
d could be characterized by gas chromatography-mass spec-
used. Mass spectrometry (MS) has been shown to be a suit- trometry (Gc-MS) thus elucidating and confirming their struc-
able tool to identify and verify spin adducts. these techniques
were applied to the characterization of radical adducts of the
tures. During this work, we determined the chemical structure
of the DIPPMPo radical adducts after the decay reactions by
nitrones benzylidene(tert-butyl)azane oxide (PBN)15,16, 4-{[tert- subjecting the diamagnetic species to gas chromatography-
butyl(oxido)imino]-methyl}pyridine 1-oxide (PoBN),17–20 2,2
mass spectrometry techniques and we studied their fragmen-
dimethyl-3,4-dihydropyrroline N-oxide (DMPo),21–27 N-aryl- tation pathways. this prompted us to examine the possibility
c,c-dimethoxycarbonylnitrones28 and (2-methyl-1-oxido- of identifying various DIPPMPo adducts by Gc-MS.
3,4-dihydro-2H-pyrrol-2-yl)phosphonate (DEPMPo).29 the
mass spectrometric analyses were performed on the isolated
radical spin adducts using high-performance liquid chroma-
tography (HPLc) or with on-line chromatographic system such
Experimental
as HPLc or Gc.15–21 recently, the structural determination of
free radical spin adducts from a complex mixture has been
reported without any preliminary separation.22–29 Each method
presents different advantages and disadvantages. Isolation
with chromatographic systems is a long procedure and does
not permit the analysis of the initial paramagnetic spin adducts.
Materials
the spin trap DIPPMPo was synthesized and purified according
to the literature.30 All the other chemicals were purchased from
Sigma–Aldrich (St Louis, Mo, uSA) and used as received.
Spin trapping
Alternatively, MS analyses that do not need preliminary sepa- Hydroxyl radicals were generated in doubly distilled water
ration overcome these problems and could be used to perform
tandem mass spectrometry (MS/MS) structure elucidation
of the various free radical adducts in complex mixtures. the
on-line separation systems approach used in this work can
lead to a total, or partial, decomposition during the elution of
the initially formed nitroxide spin adducts. However, systems
such as Gc-MS are easily available and the hydroxylamine
by a standard fenton system. A 2mL solution of DIPPMPo
(30mmoldm–3) and the fenton system (8mmoldm–3 of feSo4
and H2o2 at a ratio of 1000:1 with respect to the fe+) containing
DtPA (diethylenetriaminepentaacetic acid, 25 mmol dm–3)
was left to react in the dark for 15 min under magnetic stir-
ring at room temperature. Methyl radicals were generated
by photolysis of H2o2 in the presence of dimethylsulfoxide
R
R
R
H C
3
H C
3
H C
3
H C
3
+
+
R
+
+
N
N
N
N
(Is-PrO) PO
2
-
-
(Is-PrO) PO
2
(Is-PrO) PO
2
(Is-PrO) PO
2
O
O
OH
O
a
b
c
d
R = -CH3 (methyl radical)
R = - C(OH)CH3Ph (1-phenyl-ethanol-1-yl radical)
Figure 1. The general trapping chemistry of the DIPPMPO spin trap system.

L. Zoia and D.S. Argyropoulos, Eur. J. Mass Spectrom. 16, 175–185 (2010)
177
(DMSo). A 2mL aqueous solution of DIPPMPo (68mmoldm–3)
containing 1% of H2o2 and 5% of DMSo was transferred to a
quartz cuvette and irradiated with a 450 W medium pres-
sure mercury–vapor lamp for 1 h. then an aliquot (500 μL)
of the aqueous reaction solution was diluted in a NMr tube
with 250 μL of D2o containing 30–35 mmol dm–3 of chro-
mium chloride and 30mmoldm–3 of trimethylphosphate and
and elaboration were done using an HPchem Station software
(Hewlett-Packard, uSA).
Results and discussion
Fragmentation of DIPPMPO
submitted to 31P NMr analysis. the rest of the reaction solu- Since no MS study of DIPPMPo has ever been published, the
tion was extracted with chloroform and submitted to Gc-MS
analysis after suitable dilution. the 1-phenyl-ethanol-1-yl
radicals were generated via photochemical reaction. A
solution of acetophenone (50 mmol dm–3), 1-phenylethanol
(50mmoldm–3) and DIPPMPo (10mmoldm–3) in benzene was
added to Argon purged Pyrex tubes and degassed by several
freeze–thaw cycles. the sealed tube was then irradiated with
a 450W medium pressure mercury–vapor lamp for 2h. then
an aliquot (500μL) of the reaction solution was diluted in an
NMr tube with 250μL of cDcl3 containing 30–35mmoldm–3
fragmentation pathway of the spin trap was first examined
by Gc-MS, in order to make the structural elucidation of its
adducts easier. After separation, it was possible to detect a
single chromatographic peak with a retention time of 25.4min.
the mass spectrum of the compound under investigation
showed a peak at m/z 263 and other major fragments at m/z
221, m/z 179, m/z 162, m/z 144, m/z 98, m/z 82 and m/z 80. the
base peak was observed at m/z 98 (figure 2).
on the basis of this spectrum, a partial fragmentation
pathway was proposed (figure 3, r=–H, 1u). the peak at m/z
of chromium acetylacetonate and 30mmoldm–3 of trimethyl- 263 could be attributed to the parent radical ion [M•+], while
phosphate and submitted to 31P NMr analysis. An aliquot of
20μL was then diluted in 2mL of chloroform and submitted
to Gc-MS analysis.
the peaks at m/z 221 and m/z 179 were relative to the loss
of propene (42u) occurring via a McLafferty rearrangement
from the two isopropyl groups. the base peak at m/z 98 was
related to the loss of the diisopropyloxy(oxido)phosphoranyl
31P NMR
•
radical P(o)(o – c3H7)2 of 165 u. the peak at m/z 98 could
31P NMr spectra were acquired on a Bruker-300 spectro- also be generated by loss of the radical P(o)(oH)(o – c3H7)
•
meter (operating at 121.49 MHz). the chemical shifts
reported are relative to external orthophosphoric acid (85%). the P(o)(oH)2 of 81 u from the 179 m/z ion fragment. the
All spectra were acquired with proton decoupling. the total
number of scans for all experiments was 256–1024 with an
acquisition time of 1.60s. trimethylphosphate was used as
the internal standard. the relaxation time (t1) of the internal
of 123 u from the m/z 221 ion fragment and by the loss of
•
charge retention on the radical ion •+P(o)(o – c3H7)2 could
generate the ion fragments at m/z 165, m/z 123 and m/z 81
(after losses of propene 42u).
the mass spectrum obtained and the relative fragmenta-
standard was measured and was determined to be approxi- tion patterns are very similar to the data published by tuccio
mately 13.5 s. In order to decrease the relaxation time, a
relaxation agent [chromium chloride crcl3 or chromium
acetylacetonate cr(acac)3] was added to the mixture. With
the addition of the relaxation agent (30–35mmoldm–3) to the
samples prior to NMr measurement, the relaxation time of
the phosphorus nuclei was decreased to 200ms. 5 ×t1 was
used for the pulse delay.
et al.,29 where the researches have studied the fragmenta-
tion pathways of DEPMPo (5-diethoxyphosphoryl-5-methyl-
1-pyrroline-N-oxide) by electrospray ionization tandem mass
spectrometry (ESI-MS/MS). the similar b-phosphorylated
nitrone DEPMPo (where the two isopropyl group of the
DIPPMPo are replaced by the two ethyl groups in the phos-
phonate moiety) showed a fragmentation that proceeds mainly
via the loss of the diethoxy(oxido)phosphoranyl radical. the MS
spectra of DEPMPo and DIPPMPo were similar and exhibited
identical fragments ions at m/z 162, m/z 144, m/z 98, m/z 82
and m/z 80. the major differences observed were related to
Gas chromatography/mass spectrometry
(GC-MS)
Structural analyses were carried out by injecting 2μL of the
sample in a Hewlett Packard 5890-A gas chromatograph inter- the relative abundance of the fragments which are higher
faced to a Hewlett Packard 5972 mass spectrometer (EI 70eV). under electron ionization (EI) conditions than those observed
the mass range scanned was 40–500 m/z with a resolution of
1Da. chromatographic separation was performed on a DB-5
(30M×0.25mm and 0.25μm of thickness) fused silica capil-
lary column (J. and W. Scientific Agilent technologies, uSA).
in the ESI MS/MS experiments.
Fragmentation of radical adducts of
DIPPMPO
the injector temperature was 260°c. chromatographic condi- the spin trapping reaction between DIPPMPo and a free
tions: initial temperature 60°c, 2min isothermal, 10°cmin–1
up to 200°c, 6°cmin–1 up to 280°c, 20min isothermal. carrier
gas: He (purity 99.995%), constant flow 1.0mL min–1. When
radical leads to the formation of a nitroxide spin adduct. EPr
studies using DIPPMPo for trapping carbon- and oxygen-
centered radicals, under conditions similar to the ones used in
derivatization was needed, 20 μL of N,O-bis(trimethylsilyl)- this paper, are apparent in the literature.30 After formation, the
trifluoracetamide were add to the sample and the solution
was left to react 1h at 50°c before the injection. Acquisition
spin adduct undergoes decay to form different diamagnetic
species. consequently, our work is focused on the 31P NMr

178
Characterization of Free Radical Spin Adducts of DIPPMPO
1.2E+05
1.0E+05
8.0E+04
6.0E+04
4.0E+04
2.0E+04
0.0E+00
98
H3C
N+
H3C
O
P
O-
O
O
CH3
CH3
H3C
179
80
82
221
162
144
263
50
76
102
127
153
amu, m/z
181
208
237
271
Figure 2. EI mass spectrum of DIPPMPO.
and Gc-MS characterization of these diamagnetic compounds. the signals at 32.2ppm and 30.8ppm were related to the two
Assuming that nitroxides have totally decayed and are no stereoisomers of the hydroxylamine adducts, while the signal at
longer present in the reaction medium by the time of analyses, 24.5ppm was related to the nitrone form. the adduct observed
the data presented here mainly refers to the hydroxylamine
and nitrone chemical species produced by oxidation and/or
reduction reactions.12
in our experiments has a chemical shift of 23.1ppm.
the structure of the carbon-centered adduct has been eluci-
dated and confirmed by Gc-MS analyses of the reaction trap-
ping products. the chromatogram of the Gc-MS analyses
showed the presence of the starting spin trap at 25.4min, the
Carbon-centered radicals
Inaccordancewithourearlierwork,12 acarbon-centeredradical •oH radical adducts with DIMMPo at 27.5min and a compound
(methyl •cH3) was generated and trapped with DIPPMPo. the
(26.1min) whose mass spectrum is reported in figure 4.
the spectrum showed abundant fragments at m/z 277, m/z
235, m/z 193, m/z 176, m/z 112 and m/z 96. the ion at m/z
277 could be related to the nitrone d generated by oxida-
•
methyl radicals were generated using the oH-dependent
oxidation of the dimethylsulphoxide (DMSo) in dilute hydrogen
peroxide as described by Equations (1) and (2):
•
tion of the cH3 spin adduct (figure 1, r = –cH3). the other
•
•oH+(cH3)2So® cH3 +cH3So2– +H+
(1)
common potential product of the decay of the spin trap, the
hydroxylamine c, was not detected using Gc-MS. the 31P
NMr and Gc-MS data confirm that the trapping reaction of
methyl radical with DIPPMPo lead to radical adducts that
decay mainly to the nitrone form. Moreover, the 31P NMr data
for both DIPPMPo and DEPMPo are consistent: the nitrone
product of the methyl radical adduct of DEPMPo was shifted
0.9 ppm downfield from the parent spin trap (23.67 ppm to
24.54 ppm).11 In the present study, the nitrone formed by
trapping the methyl radical with DIPPMPo also appears at a
chemical shift difference of 0.9ppm downfield from the orig-
•cH3 +Spin trap®St(cH3)
(2)
the 31P NMr spectrum of the reaction solution showed the
presence of several components: at 22.2ppm due to the original
spin trap; at 25.2ppm due to the signal related to the •oH radical
adduct reflecting that the oH formation from the fenton system
was also present; a new peak at 23.1ppm that was related to the
•cH3 adducts.12 these results differ from those of Khramstov et
al.,11 where the authors studied the trapping of a methyl radical
with DEPMPo. the nitroxide methyl radical adduct after decay
was detected by 31P NMr at 32.3ppm, 30.8ppm and 24.5ppm. inal spin trap (22.2ppm to 23.1ppm). the different reactivity

L. Zoia and D.S. Argyropoulos, Eur. J. Mass Spectrom. 16, 175–185 (2010)
179
between DIPPMPo and DEPMPo was also confirmed by
comparing this work with the data obtained by tuccio and
co-workers, where the authors used an Lc-ESI-MS/MS
system. the nitroxide, the nitrone and the hydroxylamine
165u. It should be noted that the nitrone, generated from the
decomposition of the nitroxide spin adducts DIPPMPo/•cH3,
showed an EI mass spectrum similar to the original spin trap,
with a shift of the m/z value rationalized on the basis of the
were all detected for the methyl radical trapped by DEPMPo. cH3 radical addition of 15u on the DIPPMPo values (figure 3,
our results are in agreement with the trapping of methyl
radicals with DMPo31 where a similar adduct was detected
by Gc-MS.
r=–cH3, 15u).
furthermore, 1-phenyl-ethanol-1-yl radicals (•c(oH)cH3Ph)
were generated and trapped with DIPPMPo.14 the possibility
of generating a similar ketyl radical (2-propanol-2-yl radical
on the basis of these data, a possible fragmentation
pathway could be proposed (figure 3, r=–cH3, 15u). the peak •ccH3(oH)cH3) by photochemical reaction and trap them by
at m/z 277 is the molecular radical cation [M•+], and the peaks
at m/z 235 and m/z 193 were related to the loss of the two
isopropyl groups, in a manner similar to that of the original
DIPPMPo. the base peak at m/z 112 was related to the loss
DMPo has been explored using EPr techniques.32 the radical
1-phenyl-ethanol-1-yl could be produced from a mixture of
acetophenone and 1-phenylethanol under uV irradiation and
trapped by DIPPMPo. other products of this reaction were the
corresponding pinacols, namely the 2,3-diphenyl-2,3-butanol
•
of diisopropyl(oxido)phosphoranyl radical P(o)(o – c3H7)2 of
+
+
+
H C
R
3
H C
R
3
+
H C
R
3
+
H C
O
N
3
HO
N
+
- 42 u
- 42 u
P
HO
N
-
H C
3
P
-
O
O
P
O
O
-
O
CH
O
O
O
3
HO
CH
3
H C
3
H C
3
m/z 178 + R
- 17 u
m/z 220 + R
m/z 262 + R
- 165 u
-123 u
- 81 u
+
R
H C
3
R
+
+
N
+
H C
3
N
P
-
HO
O
-
O
O
m/z 97 + R
- H2O
m/z 161 + R
- 17 u
O
- H2
+
O
O
CH
2
H C
2
P
+
R
N
R
N
H C
2
H C
2
+
N
R
m/z 80 + R
m/z 79 + R
m/z 143 + R
R = -H (1 u)
DIPPMPO
methyl/DIPPMPO (nitrone)
R = -CH3 (15 u)
R = -C(OH)CH3Ph (121 u) ketyl/DIPPMPO (nitrone)
Figure 3. General EI MS of the nitrone form of the R• radical adduct with DIPPMPO.

180
Characterization of Free Radical Spin Adducts of DIPPMPO
1.2E+04
1.0E+04
8.0E+03
6.0E+03
4.0E+03
2.0E+03
0.0E+00
H3C
O
CH3
N+
O-
H3C
P
112
O
O
CH3
CH3
H3C
96
193
55
176
277
275
70
235
79
148
150
131
50
75
100
125
175
200
225
250
300
m/z , amu
Figure 4. EI mass spectrum of the methyl/DIPPMPO adduct.
in a racemic mixture (d, l and meso forms) from the radical
coupling reaction.
showed a peak related to the starting spin trap at 25.4 min
and different peaks related to the acetophenone, 1-pheyl
ethanol and to the d, l and meso form of the pinacol gener-
ated by the coupling of the ketyl radicals. In addition, two other
minor peaks were detected at 33.2min and 35.6min: the mass
spectra are reported in figures 5(a) and 5(b), respectively.
the mass spectra were assigned, on the basis of their m/z
value, to the nitrone and the hydroxylamine, respectively,
formed by oxidation, reduction and/or dismutation from the
initial nitroxide spin adduct.
PhcocH3 +hv→[PhcocH3]*
[PhcocH3]*+PhcH(oH)cH3 →Phc•(oH)cH3
2Phc•(oH)cH3 →meso+d,l–[Phc(oH)cH3]2
Phc•(oH)cH3 +DIPPMPo→Adducts
(3)
(4)
(5)
(6)
therefore, the mass spectrum reported in figure 5(a)
showed relative high abundance peaks at m/z 383, m/z 365,
m/z 341, m/z 323, m/z 299, m/z 281, m/z 264, m/z 256, m/z
238, m/z 218, m/z 200, m/z 182, m/z 158, m/z 105 and m/z 77.
During uV irradiation, ground-state ketones are converted to
exited singlets which, via intersystem crossing, become the
corresponding triplets, Equation (3). the excited triplets then
abstract a hydrogen atom from the corresponding alcohol, the base peak was observed at m/z 200. on the basis of the
Equation (4), to give ketyl radicals. the ketyl radicals react
via pinacol coupling, Equation (5), or in the presence of the
mechanism of trapping reported in figure 1, for r =–c(oH)
cH3Ph, this mass spectrum could be related to the nitrone
spin trap, complete the additional step to form the adducts, form d. the molecular radical cation is assigned at m/z 383.
Equation (6). the •c(oH)cH3Ph adduct showed a single
signal at 28.2ppm in 31P NMr analysis. No other peaks were
detected.
When a benzene organic solution, (after suitable dilution in
chloroform) was injected into the Gc-MS, the chromatograms
the peaks at m/z 365, m/z 341, m/z 323, m/z 299 and m/z 281
could easily be related to the alternate loss of water (18 u)
from the tertiary alcohol and the loss of the isopropyl groups
(42u) from the phosphonate unit. the base peak at m/z 200
is assigned as the fragment arising after loss of water and

L. Zoia and D.S. Argyropoulos, Eur. J. Mass Spectrom. 16, 175–185 (2010)
181
1.2E+04
200
(a)
CH
3
H C
3
+
1.0E+04
8.0E+03
6.0E+03
4.0E+03
2.0E+03
0.0E+00
N
OH
(Is-PrO) PO
2
-
O
182
183
264
105
107
158
157
256
281
365
367
218
341
238
238
299
306
77
79
323
383
50
132
212
264
300
337
m/z, amu
1.2E+04
1.0E+04
8.0E+03
6.0E+03
4.0E+03
2.0E+03
0.0E+00
(b)
202
CH
3
N
H C
3
(Is-PrO) PO
2
OH
184
82
123
105
158
96
77
367
174
175
50
75
100
125
150
200
225
250
275
300
325
350
375
400
m/z, amu
Figure 5. Mass spectra of the decomposition products of the ketyl radical/DIPPMPO spin adducts.
•
the loss of the radical •P(o)(o–c3H7)2. other peaks at m/z 105
and m/z 77 could be assigned as fragments containing the
aromatic ring. the mass spectrum could be rationalized on
mass of ketyl radical c(oH)cH3Ph, respectively before and
after loss of water from the tertiary alcohol.
Similarly, the mass spectrum reported in figure 5(b) can
the basis of the fragmentation pathway reported in figure 3, be explained on the basis of the hydroxylamine structure. It
using for r the mass value of 121u or 103u, relative to the should be noted that the hydroxylamine form, with respect

182
Characterization of Free Radical Spin Adducts of DIPPMPO
to the nitrone, has a molecular weight higher by 2u (figure
similar radical adduct was proposed by Khramstov et al.,11
1). the theoretical molecular ion at m/z 385 was not detected. when using DEPMPo as the spin trap. once the concentra-
the ion observed at m/z 367, however, can be assigned due to
loss of water (18u) from the hydroxylamine. this is a typical
reaction of a tertiary alcohol in solution. the base peak at
m/z 202 was assigned as being due to the loss of the radical
•P(o)(o–c3H7)2. Moreover the mass spectrum showed peaks
of high abundance at m/z 184, m/z 158, m/z 123, m/z 121, m/z
105, m/z 96, m/z 82 and m/z 77. the peak at m/z 184 could be
assigned due to the release of water form the base peak. the
peaks at m/z 121, m/z 105 and m/z 77 could be assigned to the
release of the ketyl radical from the adduct. the peaks at m/z
123, m/z 96 and m/z 82 could be assigned to the spin trap.
the Gc-MS results are not in agreement with the 31P NMr
data. While we were able to detect both decay products of
the nitroxide spin adduct (nitrone and the hydroxylamine) by
tion of the radical adducts increases, disproportionation and
rearrangement reactions occur, with the formation of a new
nitrone and re-formation of a molecule of the original spin
trap by water elimination (figure 6).
the radical adducts were characterized by Gc-MS. the
aqueous solution containing the spin trap and the fenton
system was extracted with chloroform and the organic phase
was submitted to Gc-MS analyses. the chromatogram showed
the presence of the initial spin trap DIPPMPo and one other
compound with retention time at 27.5min, the mass spectrum
of which is shown in figure 7.
the mass spectrum of this compound exhibits a molecular
radical ion at m/z 279 with abundant fragmentation ions at
m/z 237, m/z 95, m/z 156, m/z 123, m/z 114 and m/z 98. the
Gc-MS, using 31P NMr we were only able to detect one peak, parent radical ion was observed at m/z 279 [M•+] and ions at
despite the fact that the two forms usually have a different
chemical shift.11 We can explain these results by water elimi- losses of the isopropyl groups (42u). the base peak at m/z
nation from the hydroxylamine, leading to the formation of the 114 could be assigned as being due to the loss of the radical
m/z 237 and m/z 195 could be assigned as being due to the
corresponding alkene that present a chemical shift similar to •P(o)(o-c3H7)2.
that of the nitrone form.14
the data from Gc-MS experiments confirmed the rear-
Again, we would like to emphasize that the mass spectra
of the nitrone and hydroxylamine could easily be rationalized
on the basis of the original DIPPMPo mass with the addition
of the mass of the radical being trapped, in this case the ketyl
radical (121u and 103u).
rangement of the adducts to the new nitrone as previously
reported. No other chemical species were detected during
these analyses. In order to further confirm this rearrangement,
the Gc-MS analyses were run after derivatization with BStfA.
from the mass spectrum (data not shown) it was possible to
confirm a single derivatization reaction with no side products.
the base peak at m/z 186 was assigned as being due to the
Oxygen-centered radical
•
As previously reported,12 the oH hydroxyl radical was gener- trimethylsilyl derivative of the nitrone (114+72u) after loss of
ated by a traditional fenton system and trapped with DIPPMPo. radical •P(o)(o–c3H7)2 (165u) from the parent ion peak [M+] at
the radical adducts were then detected by 31P NMr. the spec- 351 m/z (279+72u). Moreover, the structure of the compound is
trum showed a peak at 22.2ppm related to the original spin
trap (DIPPMPo) and a peak at 25.2 ppm that we related to
the •oH adducts. As reported in the same work, once hydroxyl
radicals are produced in the presence of DIPPMPo, the spin
trapping reaction between DIPPMPo and hydroxyl radical
(oH•) produces the nitroxide spin adduct DIPPMPo/•oH. A
consistent with the fact that its chemical shift in 31P NMr is pH
dependent: this peak was located at 25.2ppm for both chloro-
form and in water at neutral or acidic pHs. However, the same
peak was located at 28.0ppm at pH9. this behavior could be
related to the presence of the labile proton on the N–oH group
which ionizes to different extents at different pHs.
H
H
OH
H3C
H3C
(Is-PrO)2PO
OH
N+
O-
OH
N+
O-
H3C
H3C
+
+
HO
N
N
(Is-PrO)2PO
OH
(Is-PrO)2PO
(Is-PrO)2PO
O
Nitroxide (spin adduct)
Nitrone
Hydroxylamine
H O
-
2
O
H3C
H3C
N+
O-
N
(Is-PrO)2PO
(Is-PrO)2PO
OH
Figure 6. Mechanism for the spin trapping of hydroxyl radicals using DIPPMPO.

L. Zoia and D.S. Argyropoulos, Eur. J. Mass Spectrom. 16, 175–185 (2010)
183
1.2E+06
O
H3C
O
114
N
1.0E+06
OH
P
H3C
O
O
CH3
8.0E+05
6.0E+05
4.0E+05
CH3
H3C
195
2.0E+05
156
98
123
237
56
86
279
219
0.0E+00
50
75
100
125
150
175
200
225
250
275
300
m/z, amu
Figure 7. Mass spectrum of DIPPMPO/•OH adduct
Similar results have been reported by castro et al.27 these
authors reported the Gc-MS analyses of trapping reactions
with DMPo for radicals generated in a fenton system in
adducts by the formation of a base peak in the MS that corre-
sponds to the mass of 5-methyl-1-pyrroline-N-oxide (98–1u)
plus the mass of the trapped radical. for the case of •oH trap-
ethanol. During the Gc-MS analyses, only one adduct (gener- ping by DIPPMPo, a rearrangement reaction was observed.
ated by trapping the •oH radical with DMPo) has been detected. overall, using DIPPMPo as a spin trap provides a powerful
related to castro’s results, we did not observe (in the presence
of DIPPMPo as the spin trap) a product arising from DMPo
nitroxide dimerization. We also did not detect the formation
of the di-tMS derivative as reported by castro. these data
evidently displayed the different trapping chemistries and the
different processes operating during the decomposition of the
spin radical adducts of DMPo and DIPPMPo.
tool, complementary to EPr detection, for the identification of
short-lived and difficult to elucidate free radical species.
References
1. E. finkelstein, G.M. rosen, E. rauckman and J. Paxton,
“Spin trapping of superoxide”, Mol. Pharmacol. 16, 676,
(1979).
2. E.G. Janzen, D.E. Nutter, E.r. Davis, B.J. Blackburn, J.L.
Poyer and P.B. Mccay, “on spin trapping hydroxyl and
hydroperoxyl radicals”, Can. J. Chem. 56, 2237 (1978).
doi: 10.1139/v78-368
3. G.r. Buettner and r.P. Mason, “Spin-trapping methods
for detecting superoxide and hydroxyl free radicals in
vitro and in vivo”, Methods Enzymol. 186, 127 (1990). doi:
10.1016/0076-6879(90)86101-Z
Conclusions
the accumulated data demonstrates the effectiveness of gas
chromatography-mass spectrometry for the identification of
low molecular weight spin adducts. the DIPPMPo spin trap
offers a convenient tool that allows a variety of radical species
to be trapped and determined with 31P NMr spectrometry. the
structures of the adducts present in the resulting mixtures
were elucidated by Gc-MS equipped with a conventional elec- 4. S.I. Dikalov and r.P. Mason, “reassignment of organic
tron ionization source. the adducts showed a fragmentation
pattern that involved the loss of diisopropiloxy(oxido)phospo-
ranyl radical •P(o)(o-c3H7)2 causing a decrease of 165u. this
behavior allowed the unambiguous identification of radical
peroxyl radical adducts”, Free Radicals Biol. Med. 27, 864
(1999). doi: 10.1016/S0891-5849(99)00134-3
5. E.G. Janzen, “Spin trapping”, Accounts Chem. Res. 4, 31
(1971). doi: 10.1021/ar50037a005

184
Characterization of Free Radical Spin Adducts of DIPPMPO
6. c.A. Evans and c. Anderson, “Spin trapping”, Aldrichim.
Acta 12, 23 (1979).
7. G.r. Buettner, “Spin trapping: ESr parameters of
spin adducts”, Free Radical Biol. Med. 3, 259 (1987).
doi:10.1016/S0891-5849(87)80033-3. doi: 10.1016/S0891-
5849(87)80033-3
8. c. Mottley and r.P. Mason, in: Biological Magnetic
Resonance, Ed by L.J. Berliner and J. reuben. Plenum
Publishers, New york, uSA, p. 489 (1989).
lipid peroxidations of omega-6 polyunsaturated fatty
acids via Lc/ESr, Lc/MS, and tandem MS”, Free
Radical Biol. Med. 34, 1017 (2003). doi: 10.1016/S0891-
5849(03)00031-5
20. H. Iwahashi, “Identification of the several new
radicals formed in the reaction mixture of oxidized
linoleic acid with ferrous ions using HPLc-ESr and
HPLc-ESr-MS”, Free Radic. Res. 37, 939 (2003). doi:
10.1080/1071576031000150463
9. P. tordo, “Spin-trapping: recent developments and
applications”, Elect. Paramagn. Reson. 16, 116 (1998). doi:
10.1039/9781847553539-00116
21. A.G. Krainev, t.D. Williams and D.J. Bigelow, “oxygen-
centered spin adducts of 5,5-dimethyl-1-pyrroline
N-oxide (DMPo) and 2H-imidazole 1-oxides”, J. Magn.
Reson. Ser. B 111, 272 (1996). doi: 10.1006/jmrb.1996.0093
22. r.u.r. Wahl, L.S. Zeng, S.A. Madison, r.L. De Pinto and
B.J. Shay, “Mechanistic studies on the decomposition
of water soluble azo-radical-initiators”, J. Chem. Soc.
Perkin Trans. 2, 2009 (1998).
23. P. Domingues, M.r.M. Domingues, f.M.L. Amado and
A.J. ferrer-correia, “Detection and characterization of
hydroxyl radical adducts by mass spectrometry”, J. Am.
Soc. Mass Spectrom. 12, 1214 (2001). doi: 10.1016/S1044-
0305(01)00310-5
24. M.r.M. Domingues, P. Domingues, A. reis, c. fonseca,
f.M.L. Amado and A.J.V. ferrer-correia, “Identification
of oxidation products and free radicals of tryptophan by
mass spectrometry”, J. Am. Soc. Mass Spectrom. 14, 406
(2003). doi: 10.1016/S1044-0305(03)00127-2
25. A. reis, M.r.M. Domingues, f.M.L. Amado, A.J.V.
ferrer-correia and P. Domingues, “Detection and
characterization by mass spectrometry of radical
adducts produced by linoleic acid oxidation”, J. Am. Soc.
Mass Spectrom. 14, 1250 (2003). doi: 10.1016/S1044-
0305(03)00538-5
26. A. reis, P. Domingues, A.J.V. ferrer-correia and
M.r.M. Domingues, ”Identification of free radicals of
glycerophosphatidylcholines containing omega-6 fatty
acids using spin trapping coupled with tandem mass
spectrometry”, Free Radical Res. 41, 432 (2007). doi:
10.1080/10715760601118353
27. J.A. castro and G.D. castro, in Methods in Molecular
Biology, Vol. 186: Oxidative Stress Biomarkers and
Antioxidant Protocols, Ed by D. Armstrong. Humana
Press Inc., totowa, NJ, uSA, p. 89 (2002).
10. c. frejaville, H. Karoui, B. tuccio, f. Le Moigne,
M. culcasi, S. Pietri, r. Lauricella and P.J. tordo,
“5-(Diethoxyphosphoryl)-5-methyl-1-pyrroline-oxide: a
new efficient phosphorylated nitrone for the in vitro and
in vivo spin trapping of oxygen-centered radicals”, J. Med.
Chem. 38, 258 (1995). doi: 10.1021/jm00002a007
11. V. Khramstov, L.J. Berliner and t.L. clanton, “NMr
spin trapping: detection of free radical reactions
using a phosphorus-containing nitrone spin trap”,
Magn. Res. Med. 42, 228 (1999). doi: 10.1002/(SIcI)1522-
2594(199908)42:2<228::AID-MrM3>3.0.co;2-t
12. D.S. Argyropoulos, L. Hongyang, A.r. Gaspar, K. Smith,
L.A. Lucia and o.J. rojas, “Quantitative 31P NMr detec-
tion of oxygen-centered and carbon-centered radi-
cal species”, Bioorg. Med. Chem. 14, 4017 (2006). doi:
10.1016/j.bmc.2006.02.009
13. L. Zoia and D.S. Argyropoulos, “Phenoxy radical detec-
tion using 31P NMr spin trapping”, J. Phys. Org. Chem.
22(11), 1070 (2009). doi: 10.1002/poc.1561
14. L. Zoia and D.S. Argyropoulos, “Detection of ketyl radi-
cals using 31P NMr spin trapping”, J. Phys. Org. Chem.
on-line doi: 10.1002/poc.1630 (2009).
15. L.r.c. Barclay, G.r. cromwell and J.W. Hilborn,
“Photochemistry of a model lignin compound. Spin trap-
ping of primary products and properties of an oligomer”,
Can. J. Chem. 72, 35 (1994).
16. H. Sang, E.G. Janzen, J.L. Poyer, P.B. Mccay, “the struc-
ture of free radical metabolites detected by EPr spin
trapping and mass spectroscopy from halocarbons in
rat liver microsomes”, Free Radicals. Biol. Med. 22, 843
(1997). doi: 10.1016/S0891-5849(96)00431-5
17. c.E. Parker, H. Iwahashi and K.B. tomer, “Spin-trapped
radicals: determination by Lc-tSP-MS and Lc-ESI-MS”,
J. Am. Soc. Mass Spectrom. 2, (5) 413 (1991). doi:
10.1016/1044-0305(91)85007-S
18. H. Iwahashi, c.E. Parker, r.P. Mason and K.B.
tomer, “combined liquid-chromatography electron-
paramagnetic resonance spectrometry electrospray
ionization mass-spectrometry for radical identifi-
cation”, Anal. Chem. 64, 2244 (1992). doi: 10.1021/
ac00043a011
28. I. El Hassan, L. charles, r. Lauricella and B. tuccio,
“Detection and identification of various carbon-centred
free radicals using N-arylketonitrones: a spin trap-
ping/EPr/MS study”, New J. Chem. 32, 680 (2008). doi:
10.1039/b713790g
29. B. tuccio, r. Lauricella and L. charles,
“characterisation of free radical spin adducts of the
cyclic b-phosphorylated nitrone DEPMPo using tan-
dem mass spectrometry”, Int. J. Mass Spectrom. 252, 47
(2006). doi: 10.1016/j.ijms.2006.02.009
19. S.y. Qian, G.H. yue, K.B. tomer and r.P. Mason,
“Identification of all classes of spin-trapped carbon-
centered radicals in soybean lipoxygenase-dependent
30. f. chalier and P.J. tordo, “5-Diisopropoxyphosphoryl-
5-methyl-1 pyrroline N-oxide, DIPPMPo, a crystalline
analog of the nitrone DEPMPo: synthesis and spin

L. Zoia and D.S. Argyropoulos, Eur. J. Mass Spectrom. 16, 175–185 (2010)
185
trapping properties”, J. Chem. Soc., Perkin Trans. 2, 2110
(2002). doi: 10.1039/b206909c
32. t.M. chiu, A. Siemiarczuk, S.K. Wong and J.r. Bolton,
“time resolution technique applied to a study of the
31. E.J. Locigno, J.L. Zweier and f.A. Villamena, “Nitric
oxide release from the unimolecular decomposition of
the superoxide radical anion adduct of cyclic nitrones
in aqueous medium”, Org. Biomol. Chem. 3, 3220 (2005).
doi: 10.1039/b507530k
absolute rate of reaction of ketyl radical with a spin
trap using flash photolysis electron paramagnetic
resonance”, J. Phys. Chem. 89, 3343 (1985). doi: 10.1021/
j100261a037