Detection of transient radical cations in electron transfer-initiated Diels-Alder reactions by electrospray ionization mass spectrometry

Article abstract of DOI:10.1021/ja046157z

The coupling of a simple microreactor to an atmospheric pressure ion source, such as electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI), allows the investigation of reactions in solution by mass spectrometry. The tris(p-bromo

Full text of DOI:10.1021/ja046157z

Published on Web 10/19/2004
Detection of Transient Radical Cations in Electron
Transfer-Initiated Diels-Alder Reactions by Electrospray
Ionization Mass Spectrometry
Sven Fu¨rmeier and Ju¨rgen O. Metzger*
Contribution from the Institute of Pure and Applied Chemistry, Carl Von Ossietzky UniVersity,
Carl-Von-Ossietzky-Strasse 9-11, 26111 Oldenburg, Germany
Received June 29, 2004; E-mail: juergen.metzger@uni-oldenburg.de
Abstract: The coupling of a simple microreactor to an atmospheric pressure ion source, such as electrospray
ionization (ESI) or atmospheric pressure chemical ionization (APCI), allows the investigation of reactions
in solution by mass spectrometry. The tris(p-bromophenyl)aminium hexachloroantimonate (1•+SbCl6-)-
initiated reactions of phenylvinylsulfide (2) and cyclopentadiene (3) and of trans-anethole (5) and isoprene
(6) and the dimerization of 1,3-cyclohexadiene (8) to give the respective Diels-Alder products were studied.
These preparatively interesting reactions proceed as radical cation chain reactions via the transient radical
cations of the respective dienophiles and of the respective Diels-Alder addition products. These radical
cations could be detected directly and characterized unambiguously in the reacting solution by ESI-MS-
MS. The identity was confirmed by comparison with MS-MS spectra of the authentic radical cations obtained
by APCI-MS and by CID experiments of the corresponding molecular ions generated by EI-MS. In addition,
substrates and products could be monitored easily in the reacting solution by APCI-MS.
and Eberlin et al.12 examined the Heck olefination of aryl
diazonium tetrafluoroborates with a palladium catalyst by ESI-
Introduction
The recent developments of mass spectrometric ionization
methods at atmospheric pressure (API), such as the atmospheric
pressure chemical ionization (APCI)1,2 and the electrospray
ionization mass spectrometry (ESI-MS),3 enable the investiga-
tion of liquid solutions by mass spectrometry. These ionization
methods opened up the access to the direct investigation of
chemical reactions in solution via mass spectrometry. In 1986,
the first real-time mass spectrometric investigation in electro-
chemical reactions4 was reported. The API methods were used
to detect and examine intermediates of reactions in solution.
The investigations were mostly performed off-line, for example,
with the oxidation of tetrahydropterins to radical cations,5 the
homogeneously catalyzed reactions, such as the Suzuki reaction,6
and the palladium-catalyzed oxidative self-coupling of arenebo-
ronic acids,7 as well as the Mitsunobu8 and Wittig reactions.9
Recently, Eberlin et al.10 detected a cyclic oxatitanium inter-
mediate of the Petasis olefination by APCI-MS. Roglans et al.11
MS and detected aryl palladium intermediates. In the palladium
dichloride-catalyzed coupling of vinylic tellurides with alkynes,
cationic Pd-Te intermediates were, for the first time, intercepted
and characterized by ESI-MS-MS.13
The group of Arakawa used an on-line ESI-MS system in
which a flowthrough photoreaction cell was attached to an
electrospray interface to detect intermediates of photochemical
reactions of some transition-metal complexes, with a lifetime
of a few minutes.14 Via direct irradiation of samples in an
optically transparent tip of an ESI source, Amster et al.
succeeded the direct detection of short-lived iron-cyclopenta-
diene complexes as key intermediates of cationic photoinduced
polymerization of epoxides.15
Chen et al. used ESI-MS to generate and isolate the active
species of homogeneously catalyzed reactions, such as Ziegler-
Natta polymerization16 or Olefin metathesis,17 and studied the
(1) Harrison, A. G. Chemical Ionization Mass Spectrometry, 2nd ed.; CRC:
Boca Raton, FL, 1992; pp 53-55.
(10) Meurer, E. C.; Santos, L. S.; Pilli, R. A.; Eberlin, M. N. Org. Lett. 2003,
5, 1391-1394.
(2) Lehmann, W. D. Massenspektrometrie in der Biochemie; Spektrum:
Heidelberg, Germany, 1996; pp 48-52.
(11) Masllorens, J.; Moreno-Man˜as, M.; Pla-Quintana, A.; Roglans, A. Org.
Lett. 2003, 5, 1559-1561.
(3) Cole, R. B. Electrospray Ionization Mass Spectrometry: Fundamentals,
Instrumentation and Applications; Wiley: New York, 1997.
(4) Hambitzer, G.; Heitbaum, J. Anal. Chem. 1986, 58, 1067-1070.
(5) (a) Scha¨fer, A.; Fischer, B.; Paul, H.; Bosshard, R.; Hesse, M.; Viscontini,
M. HelV. Chim. Acta 1992, 75, 1955-1964. (b) Scha¨fer, A.; Paul, H.;
Fischer, B.; Hesse, M.; Viscontini, M. HelV. Chim. Acta 1995, 78, 1763-
1776.
(6) Aliprantis, A.; Canary, J. J. Am. Chem. Soc. 1994, 116, 6985-6986.
(7) Aramend´ıa, M. A.; Lafont, F.; Moreno-Man˜as, M.; Pleixats, R.; Roglans,
A. J. Org. Chem. 1999, 64, 3592-3594.
(8) Wilson, S. R.; Perez, J.; Pasternak, A. J. Am. Chem. Soc. 1993, 115, 1994-
1997.
(12) Sabino, A. A.; Machado, A. H. L.; Correia, C. R. D.; Eberlin, M. N. Angew.
Chem. 2004, 116, 2568-2572; Angew. Chem., Int. Ed. 2004, 43, 2514-
2518.
(13) Raminelli, C.; Prechtl, M. H. G.; Santos, L. S.; Eberlin, M. N.; Comasseto,
J. V. Organometallics 2004, 23, 3990-3996.
(14) (a) Arakawa, R.; Tachiyashiki, S.; Matsuo, T. Anal. Chem. 1995, 67, 4133-
4138. (b) Arakawa, R.; Lu, J.; Mizuno, K.; Inoue, H.; Doe, H.; Matsuo, T.
Int. J. Mass Spectrom. Ion Processes 1997, 160, 371-376. (c) Arakawa,
R.; Matsuda, F.; Matsubayashi, G.-e.; Matsuo, T. J. Am. Soc. Mass
Spectrom. 1997, 8, 713-717 and references therein.
(15) Ding, W.; Johnson, K. A.; Kutal, C.; Amster, J. Anal. Chem. 2003, 75,
4624-4630.
(9) Wang, C.-H.; Huang, M.-W.; Lee, C.-Y.; Chei, H.-L.; Huang, J.-P.; Shiea,
J. J. Am. Soc. Mass Spectrom. 1998, 9, 1168-1174.
(16) Feichtinger, D.; Plattner, D. A.; Chen, P. J. Am. Chem. Soc. 1998, 120,
7125-7126.
9
10.1021/ja046157z CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 14485-14492
14485

A R T I C L E S
Fu¨rmeier and Metzger
Scheme 1. Tris(p-bromophenyl)aminium Hexachloroantimonate
(1•+SbCl6-)-Initiated Radical Cation Chain Reaction of
Phenylvinylsulfide (2) and Cyclopentadiene (3) To Give the
Diels-Alder Product 5-(Phenylthio)norbornene (4) via the Reactive
Intermediates, 2•+ and 4•+
reaction with the substrate in the gas phase. C-H activation of
cationic iridium(III) complexes18 and platinum(II) complexes,19
as well as the high-valent oxorhenium complex-catalyzed
aldehyde olefination,20 was investigated. Feichtinger and Plattner
used the same technique to investigate the formation and
reactions of the active species of epoxidation catalysts (i.e.,
oxomanganese-salen complexes).21
The reaction mechanism is the detailed, step-by-step descrip-
tion of a chemical reaction. Most chemical reactions take place
through a complex sequence of steps via reactive intermediates.
Simple methods for their direct detection, if possible under the
conditions of the preparative reaction, are essential for elucidat-
ing and understanding the mechanisms of synthetically important
reactions in solution. We introduced a novel method to
investigate, directly, transient C radicals in preparatively
important radical chain reactions in solution by ESI-MS, coupled
to a microreactor system. The respective radicals were detected
unambiguously and characterized by ESI-MS-MS.22
Recently, we reported the first investigation of an electron
transfer-initiated radical cation chain reaction in solution using
the microreactor coupled ESI-MS system. The transient radical
cations of the tris(p-bromophenyl)aminium hexachloroanti-
monate (1•+SbCl6-)-mediated [2 + 2] cycloaddition of trans-
anethole were detected and characterized directly and unam-
biguously in the reacting solution by ESI-MS-MS.23
ally been accepted (Scheme 1). The stable radical cation 1•+
oxidizes dienophile 2 to radical cation 2•+ which adds to diene
3, resulting in radical cation 4•+, which is reduced by substrate
2 to give the Diels-Alder cycloaddition product 4. The
occurring reactive intermediates were indirectly established by
kinetic investigations. However, both transient radical cations
2•+ and 4•+ have not been directly detected in the reacting
solution (e.g., by ESR or UV spectroscopy). Preparatively, this
reaction experiment was carried out by the addition of a 30
Tris(p-bromophenyl)aminium hexachloroantimonate (1•+Sb-
Cl6-) is a well-known, commercially available compound. It is
a deep-blue salt consisting of a tris(p-bromophenyl)aminium
radical cation (1•+) and a hexachloroantimonate anion (SbCl6-).
The redox behavior of bromo-substituted triarylamines, in
general, was studied systematically by Schmidt and Steckhan
by cyclic voltammetry,24 thus the positive oxidation potential
was shown to be 1.30 V against NHE (normal hydrogen
electrode), indicating that 1•+SbCl6- is a one-electron oxidant.
For this reason, it was often used as an electron-transfer initiator
in organic synthesis, for example, in electron-transfer-initiated
Diels-Alder reactions and other cycloadditions pioneered by
Bauld and co-workers, which have been recently summarized.25
The reaction of phenylvinylsulfide (2) and cyclopentadiene
(3) to give 5-(phenylthio)norbornene (4), mediated by aminium
salt 1•+SbCl6-, is a very interesting example of an electron-
transfer-initiated Diels-Alder reaction which proceeds via a
radical cation chain mechanism, as was shown by kinetic
investigations.25,26 The following reaction mechanism has gener-
mol % solution of 1•+SbCl6 in dichloromethane to a solution
-
of 2 with 14 equiv of 3 at 0 °C; the Diels-Alder cycloaddition
product, 5-(phenylthio)norbornene (4), was obtained in 30%
yield with an endo/exo ratio of 3:1 after a few minutes.27
In chemical textbooks, electron-transfer (ET) processes are
classified as outer sphere ETs, involving no significant covalent
interactions between donor and acceptor, or inner sphere ETs,
involving significant weak to strong covalent interactions
between donor and acceptor.28 Some years ago, Bauld postulated
a further mechanism for the electron transfer of the tris(p-
bromophenyl)aminium radical cation to phenylvinylsulfide.
Evidence for such a mechanism was found by kinetic investiga-
tions combined with substituent effect studies, which support
an electron-transfer process involving an electrophilic attack by
the aminium salt to the neutral phenylvinylsulfide, thus forming
a distonic radical cation intermediate followed by homolysis
of the newly formed covalent bond to give the amine and the
phenylvinylsulfide radical cation.27,29 Furthermore, Bauld dis-
cussed the position of the electrophilic attack and differentiated
between the attack at the unsubstituted vinyl carbon and the
nucleophilic sulfur atom (Scheme 2). Stereochemical studies
(17) (a) Adlhart, C.; Hinderling, C.; Baumann, H.; Chen, P. J. Am. Chem. Soc.
2000, 122, 8204-8214. (b) Adlhart, C.; Hofmann, P.; Chen, P. HelV. Chim.
Acta 2000, 83, 3306-3311. (c) Volland, M.; Kiener, C.; Chen, P.; Hofmann,
P. Chem.sEur. J. 2001, 7, 4621-4632.
(18) (a) Hinderling, C.; Plattner, D. A.; Chen, P. Angew. Chem. 1997, 109, 272-
274. (b) Angew. Chem., Int. Ed. 1997, 36, 243-245. (c) Hinderling, C.;
Feichtinger, D.; Plattner, D. A.; Chen, P. J. Am. Chem. Soc. 1997, 119,
10793-10804.
(19) Gerdes, G.; Chen, P. Organometallics 2003, 22, 2217-2225.
(20) (a) Chen, X.; Zhang, X.; Chen, P. Angew. Chem. 2003, 115, 3928-3931.
(b) Angew. Chem., Int. Ed. 2003, 42, 3798-3801. (c) Zhang, X.; Chen, P.
Chem.sEur. J. 2003, 9, 1852-1859.
(26) (a) Bauld, N. L.; Bellville, D. J.; Pabon, R.; Chelsky, R.; Green, G. J. Am.
Chem. Soc. 1983, 105, 2378-2382. (b) Lorenz, K. T.; Bauld, N. L. J. Am.
Chem. Soc. 1987, 109, 1157-1160.
(27) Bauld, N. L.; Aplin, J. T.; Yueh, W.; Loinaz, A. J. Am. Chem. Soc. 1997,
119, 11381-11389.
(28) Huheey, J.; Keiter, E.; Keiter, R. Inorganic Chemistry: Principles of
Structure and ReactiVity, 4th ed.; HarperCollins: New York, 1993.
(29) (a) Aplin, J. T.; Bauld, N. L. J. Chem. Soc., Perkin Trans. 2 1997, 853-
855. (b) Bauld, N. L.; Aplin, J. T.; Yueh, W.; Endo, S.; Loving, A. J.
Phys. Org. Chem. 1998, 11, 15-24. (c) Bauld, N. L.; Aplin, J. T.; Yueh,
W.; Loving, A.; Endo, S. J. Chem. Soc., Perkin Trans. 2 1998, 2773-
2776.
(21) Feichtinger, D.; Plattner, D. Chem.sEur. J. 2001, 7, 591-599.
(22) Griep-Raming, J.; Meyer, S.; Bruhn, T.; Metzger, J. O. Angew. Chem. 2002,
114, 2863-2866; Angew. Chem., Int. Ed. 2002, 41, 2738-2742.
(23) (a) Meyer, S.; Koch, R.; Metzger, J. O. Angew. Chem. 2003, 115, 4848-
4851. (b) Angew. Chem., Int. Ed. 2003, 42, 4700-4703. (c) Meyer, S.;
Metzger, J. O. Anal. Bioanal. Chem. 2003, 377, 1108-1114.
(24) Schmidt, W.; Steckhan, E. Chem. Ber. 1980, 113, 577-585.
(25) Bauld, N. L.; Gao, D. The Electron-Transfer Chemistry of Carbon-Carbon
Multiple Bonds. In Electron Transfer in Chemistry; Balzani, V., Ed.; Wiley-
VCH: Weinheim, Germany, 2001; Vol. II, Part 1; Organic Molecules;
Mattay, J., Ed.; Wiley-VCH: Weinheim, 2001; pp 133-205.
9
14486 J. AM. CHEM. SOC. VOL. 126, NO. 44, 2004

Detection of Radical Cations by ESI-MS
A R T I C L E S
Scheme 2. Possible Structures of the Distonic Radical Cation
Intermediate in the Two-Stage Polar Electron Transfer from
Tris(p-bromophenyl)aminium Radical Cation (1•+) to
Phenylvinylsulfide (2) (refs 27 and 29)
Figure 1. The microreactor allows for the effective mixing of the reactants
in solution (e.g., of 2 and 3 with 1•+SbCl6-) and the transfer to the mass
spectrometer.
of deuterated compounds suggested the attack was at the sulfur
atom.
Many more Diels-Alder reactions of electron-rich alkenes,
such as dienophile, with electron-rich 1,3-dienes initiated by
we can cover reaction times from 0.7 to 28 s. Longer reaction times
can easily be realized using a fused silica transfer capillary of variable
length between the microreactor and the spray capillary. Via a fused
silica transfer capillary, the microreactor was connected to the LCQ-
APCI source. APCI mass spectra of the reacting solution can be
acquired by this method in a reaction time range of approximately 1 s
to a few minutes.
the aminium salt 1•+SbCl6 have been studied, namely, the
-
reaction of anethole (5) and isoprene (6) (22% yield),30 as well
as the dimerization of 1,3-cyclohexadiene (8) (70% yield).31
Mechanistic diagnosis methods for the operation of radical
cation mechanisms have been developed.25 However, up to now,
it has not been possible to detect directly in the reacting solution
the transient radical cations of this important type of reaction.
We decided to study some important examples of this reaction
by microreactor-coupled API-MS, focusing on the direct detec-
tion and mass spectrometric characterization of the transient
radical cations involved. Because the ESI method, in contrast
to APCI, normally releases ions preformed in solution, it can
be expected that the transient radical cations should be detectable
by ESI-MS-MS in the reacting solution under quasi-stationary
conditions, despite the presence of other species as was shown
in the example of the [2 + 2] cycloaddition of anethole.23 First,
we decided to study the APCI-MS spectra of substrates and
products to obtain MS-MS spectra of the authentic radical
cations we wanted to detect and to investigate the respective
reactions by APCI-MS to measure a reacting solution. For
comparison, the MS-MS spectra of the radical cations formed
by electron ionization (EI) should be studied, as well. The
intriguing results obtained are presented in this paper.
ESI operation conditions of a spray voltage of +3 kV and a heated
capillary temperature of 150 °C were utilized, whereas in the APCI
mode, a vaporizer temperature set to 300 °C, a corona discharge of 3
µA, and a capillary temperature of 150 °C were employed. Collision-
induced dissociation (CID, collision gas helium) was performed in the
ion trap region. The effective isolation width for CID experiments was
set at 1.0 u, with a collection time up to 500 ms. Data acquisition and
analysis were done with the Xcalibur (version 1.2, Thermoquest
Finnigan) software package. Additionally, ESI-MS experiments have
been carried out on the quadrupole time-of-flight instruments (Ultima
Q-ToF, Micromass) and Q-Star (Applied Biosystems).
The electron ionization experiments were performed using Finnigan
MAT 95 (Thermo Finnigan MAT) with an ion source temperature of
220 °C, an emission of 0.1 mA, and an electron energy of 70 eV. The
CID spectra in the first free-field region were taken using a B/E-linked
scan.
Reaction of Phenylvinylsulfide and Cyclopentadiene. A solution
of tris(p-bromophenyl)aminium hexachloroantimonate (1•+SbCl6-), 1
× 10-4 mol L-1, and a solution of phenylvinylsulfide (2), 1 × 10-3
mol L-1, with cyclopentadiene (3), 5 × 10-3 mol L-1, both in
dichloromethane, were mixed using a dual syringe pump feeding the
microreactor that was coupled directly to the ion source of the mass
spectrometer (Figure 1). The reacting solution was fed continuously
into the MS and investigated by APCI-MS and ESI-MS, as described
above.
Experimental Section
Tris(p-bromophenyl)aminium hexachloroantimonate (1•+SbCl6-),
trans-anethole (5), isoprene (6), and 1,3-cyclohexadiene (8) were
obtained from Aldrich (Steinheim, Germany). Phenylvinylsulfide (2),
dicyclopentadiene, and dichloromethane were purchased from Fluka
(Deisenhofen, Germany). Cyclopentadiene (3) was freshly distilled from
dicyclopentadiene. Dichloromethane was distilled after refluxing over
calcium hydride, and cyclohexadiene was distilled over sodium. The
other chemicals were used without further purification. endo/exo-5-
(Phenylthio)norbornene (4) and trans-1,5-dimethyl-4-(4′-methoxyphe-
nyl)cyclohexene (7) were synthesized according to published proce-
Reaction of trans-Anethole and Isoprene. The solution of 1•+SbCl6-
,
1 × 10-4 mol L-1, and a solution of anethole (5), 1 × 10-3 mol L-1
,
with isoprene (6), 5 × 10-3 mol L-1, in dichloromethane, were mixed
in the microreactor and fed continuously into the MS system.
Reaction of 1,3-Cyclohexadiene. Analogously, the solution of
1•+SbCl6-, 1 × 10-4 mol L-1, was mixed in the microreactor with a
solution of 1,3-cyclohexadiene (8), 5 × 10-3 mol L-1 in dichlo-
romethane, and the reacting solutions were examined by MS.
1
dures27,30 and characterized by H and 13C NMR.
API-MS experiments were performed using a Finnigan LCQ
(Thermo Finnigan, San Jose´, CA) quadrupole ion trap mass spectrom-
eter equipped with a standard ESI and APCI ion source. The standard
ESI source of the LCQ was used with a stainless steel metal capillary
(110 µm i.d., 240 µm o.d., 120.5 mm of length, Metal Needle Kit,
Thermo Finnigan), with a volume of 1.14 µL. Sample solutions were
infused using the dual syringe pump of the LCQ by varying the flow
rate in a range of 2.5-100 µL/min. By connecting a microreactor
(ALLTECH, PEEK mixing tee; see Figure 1) to the ESI spray capillary,
Control Experiments. Solutions of substrates 2, 2 with 3, 5, 5 with
6, and 8, as well as products 4 and 7, each in dichloromethane, however
without 1•+SbCl6-, were investigated in the same way by ESI-MS-
MS. Radical cations 2•+, 4•+, 5•+, 7•+, 8•+, and 9•+ could not be observed
in these experiments. Solutions of products 4, 6, and 9, in the presence
of 1•+SbCl6-, were studied in the same way. Radical cations 4•+, 7•+
,
and 9•+ could not be detected. A solution of substrate 2 and product 4
-
was mixed in the microreactor with a solution of 1•+SbCl6 and was
(30) Reynolds, D. W.; Bauld, N. L. Tetrahedron 1986, 42, 6189-6194.
(31) Bellville, D. J.; Wirth, D. D.; Bauld, N. L. J. Am. Chem. Soc. 1981, 103,
718-720.
investigated by ESI-MS-MS. Cation 2•+ was detected unambiguously;
4•+could not be detected.
9
J. AM. CHEM. SOC. VOL. 126, NO. 44, 2004 14487

A R T I C L E S
Fu¨rmeier and Metzger
Figure 2. Positive APCI mass spectrum of the reacting solution of
phenylvinylsulfide (2), cyclopentadiene (3), and tris(p-bromophenyl)-
aminium hexachloroantimonate (1•+SbCl6-) in dichloromethane after a
reaction time of approximately 20 s.
Results
1. Reaction of Phenylvinylsulfide (2) and Cyclopentadiene
(3): APCI-MS Measurement. The separate measurements of
substrates 2 and 3 and product 4 by APCI-MS were the starting
point of the investigations. In the APCI mass spectrum of a
solution of 2 in dichloromethane, an intensive ion at m/z 137,
due to the protonated molecule [2 + H]+, an ion at m/z 136,
due to the radical cation 2•+, and an ion at m/z 185, which can
be identified via the characteristic isotopic pattern to become a
cationized molecule [2 + CH2Cl]+, are detected. This chlo-
romethylen adduct ion was found as a typical ion formed in
the APCI spray process using dichloromethane as the solvent.
Product 4 was measured analogously. The spectrum shows the
protonated molecule [4 + H]+ (m/z 203), the 5-(phenylthio)-
norbornene radical cation (4•+) at m/z 202, and, in addition, the
chloromethylen adduct ion [4 + CH2Cl]+ at m/z 251. In both
cases, the protonated molecules and the radical cations are
generated in the APCI process. Cylopentadiene 3 was examined
analoguously. Remarkably, there were no characteristic signals
of 3 to observe directly in the APCI-MS experiment. Thus, 3
was not ionized in the APCI process using dichloromethane as
the solvent.
Figure 3. (a) APCI-MS-MS spectrum of the molecular ion of phenylvi-
nylsulfide 2•+ (m/z 136). (b) APCI-MS-MS spectrum of the molecular
ion of 5-(phenylthio)norbornene 4•+ (m/z 202).
To investigate the Diels-Alder reaction, a solution of
1•+SbCl6- and a solution of 2 with 3, both in dichloromethane,
were efficiently mixed in the microreaction system and fed
continuously into the APCI-MS. In Figure 2, the mass spectrum
of the reacting solution after a reaction time of approximately
20 s is presented. At m/z 479-486, signals derived from
-
1•+SbCl6 are observed. The signals of substrate 2 and of
product 4 can be recognized, as well, thus pointing out that a
solution with an ongoing reaction is being investigated. Some
additional ions were observed in the APCI mass spectrum of
the reacting solution, giving evidence for the formation of minor
byproducts.
APCI-MS-MS of Radical Cations 2•+ and 4•+. Radical
cations 2•+ and 4•+ are formed in the APCI process, allowing
the characterization of authentic 2•+ and 4•+ by MS-MS (Figure
3a,b).
EI-MS-MS of the Molecular Ions of Phenylvinylsulfide
(2) and 5-(Phenylthio)norbornene (4). Radical cations 2•+ and
4•+ can also be generated by the EI ionization of substrate 2
and product 4, respectively. The high-energy CID mass spectrum
of the parent ion of phenylvinylsulfide 2 (Figure 4a) shows the
same fragmentations as the APCI-MS-MS spectrum (Figure 3a).
The CID mass spectrum of the 5-(phenylthio)norbornene
Figure 4. (a) CID mass spectrum of EI-generated parent ion of phenylvi-
nylsulfide 2•+ (m/z 136). Zoomed into (25-fold) the mass range of m/z 80-
130. (b) CID mass spectrum of EI-generated parent ion of 5-(phenylthio)-
norbornene 4•+ (m/z 202). Zoomed into (20-fold) the mass range of m/z
115-195.
molecular ion 4•+ (Figure 4b) shows, in comparison to that of
APCI-MS-MS (Figure 3b), an intensive retro-Diels-Alder
fragmentation to give the product ion m/z 136. It seems to be
remarkable that the latter fragmentation cannot be observed in
the low-energy CID in the ion trap.
9
14488 J. AM. CHEM. SOC. VOL. 126, NO. 44, 2004

Detection of Radical Cations by ESI-MS
A R T I C L E S
Figure 5. Positive ESI mass spectrum of the reacting solution of
phenylvinylsulfide (2), cyclopentadiene (3), and tris(p-bromophenyl)-
aminium hexachloroantimonate (1•+SbCl6-) in dichloromethane after a
reaction time of approximately 14 s.
ESI-MS Investigation of the Diels-Alder Reaction of
Phenylvinylsulfide (2) and Cyclopentadiene (3). The reacting
solution of the 1•+SbCl6--initiated Diels-Alder reaction of 2
and 3 was examined by ESI-MS. The mass spectrum, after a
reaction time of 14 s, is depicted in Figure 5, showing the
intensive signal of radical cation 1•+ at m/z 479-485, with the
characteristic isotopic cluster of three bromine atoms. Substrate
2 and product 4 are not ionized in the ESI process. The transient
radical cations 2•+ (m/z 136) and 4•+ (m/z 202) cannot be
unambiguously detected in the spectrum directly. Zooming into
the chemical underground, we observe recognizable very weak
signals at the expected values of the radical cations m/z 136
and 202, respectively. Remarkably, signals of similar intensity
were observed at m/z 135 and 137, as well as at m/z 201 and
203.
Figure 6. (a) ESI-MS-MS spectrum of ion m/z 136 of the reacting solution
of phenylvinylsulfide (2), cyclopentadiene (3), and tris(p-bromophenyl)-
aminium hexachloroantimonate (1•+SbCl6-) in dichloromethane, showing
identical fragmentations compared to those in the APCI-MS-MS spectrum
of the authentic radical cation 2•+ (Figure 3a). (b) ESI-MS-MS spectrum
of ion m/z 202 of the same reacting solution, showing identical fragmenta-
tions compared to those in the APCI-MS-MS spectrum of the authentic
radical cation 4•+ (Figure 3b).
Using the MS-MS technique,3 which allows the separation
of the ions of interest from all other ions and, by CID, their
mass spectrometric characterization, we could detect and identify
both radical cations 2•+ (Figure 6a) and 4•+ (Figure 6b)
unambiguously in the reacting solution, using the ion trap as
well as the Q-Tof instrument. This is confirmed by comparison
with the APCI-MS-MS spectra of the authentic ions of 2•+
(Figure 3a) and 4•+ (Figure 3b). The obtained spectra are nearly
identical. A cyclopentadiene radical cation could not be
observed. The ions at m/z 135, 137, 201, and 203 were
investigated by ESI-MS-MS and identified by comparison of
the those of the spectra obtained by APCI-MS-MS as [2 -
H]+, [2 + H]+, [4 - H]+, and [4 + H]+, respectively.
To rule out that 2•+ and 4•+ are generated in small amounts
by the ESI process itself, control experiments in the absence of
Scheme 3. Tris(p-bromophenyl)aminium Hexachloroantimonate
(1•+SbCl6-)-Initiated Diels-Alder Reaction of trans-Anethole (5)
and Isoprene (6) To Give
trans-1,5-Dimethyl-4-(4′-methoxyphenyl)cyclohexene (7) via the
Reactive Intermediates, 5•+ and 7•+ (ref 30)
-
1•+SbCl6 were performed. Solutions of 2, 3, and 4 in
dichloromethane were examined by ESI-MS-MS. In the
appropriate MS-MS experiments, 2•+ and 4•+ could not be
observed, giving clear evidence that the ESI process neither
oxidizes substrate 2 to 2•+ nor product 4 to 4•+. Thus, radical
cations 2•+ and 4•+ are not formed during the mass spectrometric
ionization process in the ESI source, pointing out that transient
radical cations 2•+ and 4•+ (Figure 6b,c) were detected directly
in the reacting solution. Finally, to rule out that 4•+ is formed
by oxidation of product 4 under reaction conditions, or in the
m/z 615-624. Experiments to detect this intermediate directly
by MS-MS, as described for radical cations 2•+ and 4•+, were
not successful using the ion trap as well as the Q-ToF
instruments.
2. Reaction of trans-Anethole (5) and Isoprene (6). The
Diels-Alder reaction of trans-anethole (5) and isoprene (6) to
give trans-1,5-dimethyl-4-(4′-methoxyphenyl)cyclohexene (7)
initiated by aminium salt 1•+SbCl6- (Scheme 3) was investigated
analogously.30
ion source, solutions of product 4 and 1•+SbCl6 were mixed
-
in the microreactor and investigated as described above for the
reacting solution. Cation 4•+ could not be detected.
The postulated distonic radical cation of the two-stage polar
mechanism of electron transfer (Scheme 2) is to be expected at
Substrate 5, product 7, and the respective reacting solution
of the Diels-Alder reaction were studied by APCI-MS. The
protonated molecule and the radical cation of substrate 523 and
9
J. AM. CHEM. SOC. VOL. 126, NO. 44, 2004 14489

A R T I C L E S
Fu¨rmeier and Metzger
Figure 7. ESI-MS-MS spectrum of ion m/z 216: radical cation 7•+ of
the reacting solution of trans-anethole (5), isoprene (6), and tris(p-
bromophenyl)aminium hexachloroantimonate (1•+SbCl6-) in dichlo-
romethane.
Scheme 4. Tris(p-bromophenyl)aminium Hexachloroantimonate
(1•+SbCl6-)-Initiated Dimerization of Cyclohexadiene (8) via
Radical Cations 8•+ and 9•+ (ref 31)
Figure 8. (a) ESI-MS-MS spectrum of ion m/z 80: radical cation 8•+ of
-
the reacting solution of cyclohexadiene (8) and 1•+SbCl6 in dichlo-
romethane. (b) ESI-MS-MS spectrum of ion m/z 160: radical cation 9•+
in the same reacting solution.
solution of 8 in dichloromethane was examined by APCI-MS.
Remarkably, 8 could be ionized; however, it did show a
relatively low signal-to-noise ratio, giving evidence that the
ionization is not very efficient. The APCI mass spectrum shows
the protonated molecule [8 + H]+ at m/z 81, the radical cation
8•+ at m/z 80, the protonated molecule [9 + H]+ at m/z 161,
and the 9•+ at m/z 160. The APCI investigation of the chemical
reaction by mixing the substrate solution with a solution of
aminium salt 1•+SbCl6- shows the protonated product [9 + H]+
with a higher intensity and substrate [8 + H]+ with a lower
intensity compared to the measurement of the solution of 8.
Authentic radical cations 8•+ and 9•+ could be characterized in
the reacting solution as well as in the solution of 8 by APCI-
MS-MS.
The ESI mass spectrum of the same reacting solution showed
the intensive ion of radical cation 1•+ at m/z 479-485. Zooming
into the chemical underground, we observed and investigated
ions of m/z 80 and 160 by MS-MS. The obtained ESI-MS-
MS spectra of m/z 80 and 160 (Figure 8) are nearly identical to
those in the APCI-MS-MS of the authentic radical cations of
8•+ and 9•+, respectively, giving unambiguous evidence that
transients 8•+ and 9•+ are present in the reacting solution.
1,5-dimethyl-4-(4′-methoxyphenyl)cyclohexene (7), respectively,
could be observed in the APCI mass spectra, allowing the
characterization of authentic radical cation 7•+ by APCI-MS-
MS. Isoprene (6) was not ionized and thus not observed in the
mass spectrum of the reacting solution. Remarkably, the minor
formation of 1,2-bis(4-methoxyphenyl)-3,4-dimethylcyclobu-
tane, the product of the [2 + 2] cycloaddition of anethole (5),
was observed in the reacting solution as a protonated molecule
at m/z 297 and a radical cation at m/z 296.23 The high-energy
CID mass spectrum of the molecular ion of Diels-Alder product
7 shows the RDA fragmentation (m/z 148) in addition to the
fragmentations observed in the APCI-MS-MS spectrum.
The ESI mass spectrum of the reacting solution shows the
intensive ion of 1•+ and the very low intensity signals near the
chemical noise at m/z 148 and 216 analogously to those shown
in Figure 5. Transient radical cation 5•+ (m/z 148) could be
detected by MS-MS in the reacting solution. The MS-MS
spectrum is in perfect agreement with that of ref 23. The second
reactive intermediate, 7•+, of the radical cation chain reaction
could also be detected by MS-MS (Figure 7). Both transient
radical cations were detected using the ion trap as well as the
Q-Tof instrument. The obtained ESI-MS-MS spectrum of 7•+
is nearly identical to the APCI-MS-MS spectrum of the
authentic radical cation 7•+. A radical cation of isoprene (6•+)
Discussion
Steady-state conditions are necessary for the detection of the
transient radical cations in the radical cation chain reactions
(Schemes 1, 3, and 4) by ESI-MS. Therefore, when the solution
containing the substrates and the solution with the initiator
-
could not be detected by MS-MS in the reacting solution.
1•+SbCl6 are mixed in the efficient micromixer, the chain
3. Dimerization of 1,3-Cyclohexadiene (8). The 1•+SbCl6
-
reaction is started and steady-state conditions are established.
Thus, pumping the reacting solution continuously in the ESI
source, a quasi-stationary concentration of the radical cations,
the reactive intermediates of the chain reactions, will be
-
initiated dimerization of cyclohexadiene (8) via radical cations
8•+ and 9•+ to give the Diels-Alder dimerization product 9
(endo/exo, 5:1)31 was studied analogously (Scheme 4). First, a
9
14490 J. AM. CHEM. SOC. VOL. 126, NO. 44, 2004

Detection of Radical Cations by ESI-MS
A R T I C L E S
transferred in the mass spectrometer. Because the ESI method
normally releases ions preformed in solution, it can be expected
that these radical cations should be detectable by ESI-MS-
MS in the reacting solution, despite the presence of other
species. The quasi-stationary concentration is estimated to be
approximately 1 × 10-7 mol L-1, assuming a diffusion-
controlled termination reaction. This concentration is 3 orders
of magnitude lower than the concentration of radical cation 1•+.
Thus, the ions of the transient radical cations are expected to
disappear in the chemical noise and will not be unambiguously
detected in the normal ESI mass spectrum. However, their
concentration is high enough to allow the detection and
characterization by ESI-MS-MS, considering the fact that the
reacting solution is continuously fed in the mass spectrometer
and many spectra may be accumulated. Therefore, it is of
essential importance that the reaction is not completed at the
moment of the electrospray ionization to enable the detection
of the intermediates. To verify that a solution with an ongoing
chain reaction is being studied, it is also important to be able
to follow the reaction by measuring the decrease of the substrate
and the corresponding increase of the product. Unfortunately,
the neutral substrates are normally not present as preformed
ions and will not be ionized by the ESI process. However, this
was shown to be possible by using APCI-MS. In the APCI
mode, neutral molecules are transferred from the liquid phase
into the gas phase and are ionized at atmospheric pressure via
either protonation, deprotonation, or charge exchange.1,2 Thus,
the APCI spectrum of the reacting solution (Figure 2) showed
signals of substrate 2 and product 4, and the ESI spectrum
showed the intensive signal of the initiator 1•+, giving clear
evidence that the radical cation chain reaction was not yet
finished at the moment of ionization. At the moment of
ionization by ESI, as well as by APCI, the reaction in solution
is being stopped and the mass spectrum will give a picture of
the state of the reaction in solution because, in the gas phase,
bimolecular reactions can be excluded. However, low-energy
monomolecular reactions of the formed ions, such as rearrange-
ments, are possible and can be expected. Thus, we will observe
the thermodynamically most-stable isomer.
Radical cations 2•+, 4•+, 5•+, 7•+, 8•+, and 9•+ could not be
observed in the performed control experiments in the absence
of 1•+SbCl6-. These findings led to the conclusion that the ESI
process oxidizes neither substrates nor products to the respective
radical cations and pointed out that the transient radical cations
were detected directly in the reacting solution. Furthermore,
oxidation experiments by mixing products 4, 7, and 9 with
1•+SbCl6- showed no respective radical cations, thus excluding
the possibility of their formation in the in the gas phase during
ESI by the in-source reaction of products with 1•+. Additionally,
when substrate 2 and product 4 were mixed with a solution of
1•+SbCl6-, we detected radical cation 2•+ but not 4•+, giving
clear evidence that radical cation 4•+ was not formed by the
in-source reaction of 2•+ and product 4.
It may be mentioned that we observed the formation of the
[2 + 2] cycloaddition product of 5, in the reacting solution by
APCI, as a minor byproduct. This was also reported to be
formed in preparative reactions using a 5-fold excess of diene.30
Thus, the [2 + 2] cycloaddition of 5 seems to be faster than
the Diels-Alder addition to isoprene.
Bauld27,29 presented convincing evidence of a two-stage polar
mechanism of the oxidation of sulfide 2 by the aminium radical
cation 1•+, giving a distonic radical cation (Scheme 2). We could
not detect this ion directly in the reacting solution. This may
have different reasons. An explanation could be that the
dissociation step is very fast, and thus the lifetime will be too
short for the detection by the long collecting ion trap. However,
investigations using a Q-ToF-MS with a much shorter collec-
tion time failed, as well, to detect this intermediate. Thus, this
distonic radical cation may have a very short lifetime, possibly
being, more likely, a transition state.
The termination reaction is an important part of a chain
reaction. Bauld found a bimolecular chain-termination step by
kinetic investigations for the Diels-Alder dimerization of
cyclohexadiene (Scheme 4) and assumed a bimolecular coupling
of radical cation 8•+ to produce a dication.26b An alternative
bimolecular termination reaction could be a disproportion of
two radical cations to give two alkyl cations. The respective
ions were observed in the ESI mass spectrum and characterized
by MS-MS, which may possibly be an indication for such a
termination step (eq 1).
Our experimental results give clear evidence that aminium
salt-initiated Diels-Alder reactions can be studied successfully
by microreactor-coupled API-MS. It may be mentioned that up
to now we did not observe any special effects in the microre-
actor. We used it as efficient micromixing system of two
solutions having a very low flow. This system allowed us to
detect and to characterize the transient radical cations of
dienophiles 3, 5, and 8 and adduct radical cations 4, 5, and 9 of
the radical cation chain reactions investigated (Schemes 1, 3,
and 4) directly and unambiguously by ESI-MS-MS. On the
other side, it could be demonstrated that radical cations of dienes
3 and 5 are not formed. This was to be expected because the
oxidation potential of cyclopentadiene (3) is approximately 0.4
V higher than the respective oxidation potential of phenylvi-
nylsulfide (2), which is oxidized by radical cation 1•+.27,32 In
the case of the Diels-Alder reaction of trans-anethole (5) and
isoprene (6), 5 has an oxidation potential ∼0.55 V lower than
that of 6, and therefore 5 was oxidized by radical cation 1•+.33
8
•+ + 8•+ f [8 - H]+ + [8 + H]+
(1)
On the basis of ab initio calculations, the Diels-Alder reaction
of the 1,3-butadiene radical cation with ethene to give the
cyclohexene radical cation C6H10•+ was described as a concerted,
nonsynchronous, activationless cycloaddition.34 Recent DFT and
high-correlated MO calculations of the Diels-Alder reaction
of the 1,3-butadiene radical cation with ethene found a stepwise
addition involving open-chain intermediates and leading to the
cyclohexene radical cation.35 Similar intermediates were found
in comparable quantum-mechanical calculations of the radical
cation Diels-Alder reactions of 1,3-cyclohexadiene (8) and
indole.36
(34) (a) Bauld, N. L.; Bellville, D. J.; Harirchian, B.; Lorenz, K. L.; Pabon, R.
A.; Reynolds, W. W.; Wirth, D. D.; Chiou, H.-S.; Marsh, B. K. Acc. Chem.
Res. 1987, 20, 371-378. (b) Bauld, N. L. Tetrahedron 1989, 45, 5307-
5363. (c) Bauld, N. L. J. Am. Chem. Soc. 1992, 114, 5800-5804.
(35) (a) Hofmann, M.; Scha¨fer, H. F., III. J. Am. Chem. Soc. 1999, 121, 6719-
6729. (b) Haberl, U.; Wiest, O.; Steckhan, E. J. Am. Chem. Soc. 1999,
121, 6730-6736. (c) Hofmann, M.; Scha¨fer, H. F., III. J. Phys. Chem. A
1999, 103, 8895-8905.
(32) Baltes, H.; Steckhan, E.; Scha¨fer, H. J. Chem. Ber. 1978, 111, 1294-1314.
(33) Bauld, N. L.; Harirchian, B.; Reynolds, D. W.; White, J. C. J. Am. Chem.
Soc. 1988, 110, 8111-8117.
9
J. AM. CHEM. SOC. VOL. 126, NO. 44, 2004 14491

A R T I C L E S
Fu¨rmeier and Metzger
Thus, in the radical cation chain reactions studied, radical
cations 4•+, 7•+, and 9•+ could be formed by a stepwise
mechanism, which proceeds via a distonic open-chain radical
cation followed by ring closure to give the Diels-Alder radical
cation. The MS-MS spectra of 4•+ (Figures 3b and 6b), 7•+
(Figure 7), and 9•+ (Figure 8b) give clear evidence that the
closed Diels-Alder radical cations were observed exclusively
under our experimental conditions. The distonic open-chain
radical cations can be excluded due to the missing strong
fragmentation of 4•+, 7•+, and 9•+ to m/z 136 (2•+), 148 (5•+),
and 80 (8•+), respectively. Obviously, the ring closure of the
possible initially formed distonic ion to give the cyclohexene
radical cations has to be much faster than our long collecting
ion trap as well as the Q-ToF.
reaction. The reaction process could be followed easily by APCI.
For the first time, the transients of these radical cation chain
reactions were detected and characterized unambiguously under
steady-state conditions in the reacting solution by ESI-MS-
MS. The detection was confirmed by comparison with APCI-
MS-MS and EI-CID MS spectra of the respective authentic
radical cations. The microreactor-coupled API mass spectrom-
etry is of importance since it can be generally applied to all
chemical reactions in solution. Transient intermediates can be
detected, presuming that the species of interest are ionic or can
be ionized.
Acknowledgment. We thank the German Research Associa-
tion (DFG) and the Heinz Neumu¨ller Foundation, Oldenburg,
for financial support. We thank Mrs. R. Fuchs (Micromass) and
Mr. A. Besa (Applied Biosystems) for performing the ESI-MS
experiments on the Q-ToF instruments.
Conclusion
In conclusion, three electron transfer-initiated Diels-Alder
reactions were studied by APCI and ESI-MS, coupled with a
microreactor under conditions very similar to the preparative
Supporting Information Available: Additional MS and MS-
MS spectra. This material is available free of charge via the
Internet at http://pubs.acs.org.
(36) (a) Wiest, O.; Steckhan, E.; Grein, F. J. Org. Chem. 1992, 57, 4034-
4037. (b) Haberl, U.; Steckhan, E.; Blechert, S.; Wiest, O. Chem.sEur. J.
1999, 5, 2859-2865. (c) Saettel, N. J.; Wiest, O.; Singleton, D. A.; Meyer,
M. P. J. Am. Chem. Soc. 2002, 124, 11552-11559.
JA046157Z
9
14492 J. AM. CHEM. SOC. VOL. 126, NO. 44, 2004