Enol radical cations in solution. 13. First example of a radical dication rearrangement. One-electron oxidation of dihydrobenzofuranyl cations leads to drastic rate enhancement in the oxidative benzofuran formation from enols

Article abstract of DOI:10.1021/jo980859n

The synthesis and electrochemical investigation of six stable, simple enols El-EG are reported that are characterized by electron-releasing substituents in the a-position. Oxidative benzofuran formation from these enols is unusually slow because a key intermediate in the reaction, the dihydrobenzofuranyl cation X⁺, is substantially stabilized vs rearrangement by the attached electronreleasing substituents. The persistent cations X⁺ were characterized by 1H NMR and cyclic voltammetry, and the kinetics of their rearrangement was followed by UV/vis. Notably, upon oneelectron oxidation of X⁺ to the radical dication, the formation of the benzofurans B was markedly accelerated by a factor of > 10⁶.

Full text of DOI:10.1021/jo980859n

7328
J . Org. Chem. 1998, 63, 7328-7337
En ol Ra d ica l Ca tion s in Solu tion . 13. F ir st Exa m p le of a Ra d ica l
Dica tion Rea r r a n gem en t. On e-Electr on Oxid a tion of
Dih yd r oben zofu r a n yl Ca tion s Lea d s to Dr a stic Ra te En h a n cem en t
in th e Oxid a tive Ben zofu r a n F or m a tion fr om En ols
Michael Schmittel* and Anja Langels
Institut fu¨r Organische Chemie der Universita¨t Wu¨rzburg, Am Hubland,
D-97074 Wu¨rzburg, Federal Republic of Germany
Received May 6, 1998
The synthesis and electrochemical investigation of six stable, simple enols E1-E6 are reported
that are characterized by electron-releasing substituents in the R-position. Oxidative benzofuran
formation from these enols is unusually slow because a key intermediate in the reaction, the
dihydrobenzofuranyl cation X+, is substantially stabilized vs rearrangement by the attached electron-
releasing substituents. The persistent cations X+ were characterized by 1H NMR and cyclic
voltammetry, and the kinetics of their rearrangement was followed by UV/vis. Notably, upon one-
electron oxidation of X+ to the radical dication, the formation of the benzofurans B was markedly
accelerated by a factor of >106.
In tr od u ction
with antimony pentafluoride and to characterize it. The
redox potentials of some tris(dialkylamino)-substituted
cyclopropenyl cation/cyclopropenyl radical dication couples
were only recently determined by Taylor et al.5d
In the context of our ongoing investigations on the
chemistry of enol radical cations in solution1,2 we have
discovered the rare event of a selective activation of a
cationic intermediate via the corresponding radical di-
cation, which shall be described here in detail.3 Since
radical dications are usually formed through one-electron
oxidation of electron-deficient cationic precursors, this
class of mostly highly reactive intermediates is difficult
to access. Indeed, only a few reports have been concerned
with the reactivity4 and characterization5 of radical
dications.
A second class of radical dications was investigated by
Lenhard et al. in 1987.4a Electrochemical or chemical
one-electron oxidation of cationic dicarbocyanine dyes
generated the corresponding radical dications that were
found to undergo reversible dimerization at uneven
carbon atoms of the methine chain in acetonitrile at room
temperature.4a
The first investigation on a radical dication has been
reported by Gerson et al. in 1971.5a After electrochemical
oxidation of the stable tris(dimethylamino)cyclopropenyl
cation the corresponding deep red radical dication could
be identified by its EPR signal.
Another class of radical dications is derived from the
one-electron oxidation of stable heterocyclic cations such
as pyridinium, quinolinium, isoquinolinium, pyrazolium,
imidazolium, pyrylium or chalcogenopyrylium cations.
According to cyclic voltammetry investigations,5e,f in
many cases relatively stable radical dications can be
detected.
Our earlier investigations6 on â,â-dimesityl enols have
established that upon oxidation of E with 2 equiv of an
appropriate one-electron oxidant (e.g., tris(1,10-phenan-
throline)iron(III) hexafluorophosphate (F ep h en ), triaryl-
aminium salts), 4,6,7-trimethylbenzofurans B were af-
forded as stable products. The proposed mechanism is
shown in Scheme 1).7
Some years later Weiss et al.5b were able to isolate
(yields > 90%) the same stable radical dication by the
reaction of the tris(dialkylamino)cyclopropenyl cation
(1) Part 12: Schmittel, M.; Langels, A. J . Chem. Soc., Perkin Trans.
2 1998, 565-572.
(2) Schmittel, M.; Gescheidt, G.; Ro¨ck, M. Angew. Chem., Int. Ed.
Engl. 1994, 33, 1961-1963.
(3) Schmittel, M.; Langels, A. Angew. Chem., Int. Ed. Engl. 1997,
36, 392-395.
Kinetic isotope effect studies1 reveal that OH depro-
tonation to the corresponding R-carbonyl radicals R•
constitutes the primary reaction of the enol radical
(4) (a) Lenhard, J . R.; Parton, R. L. J . Am. Chem. Soc. 1987, 109,
5808-5813. (b) Weiss, R.; Grimmeiss, A. M. H. Z. Naturforsch. 1991,
46b, 104-110.
(5) (a) Gerson, F.; Plattner, G.; Yoshida, Z. Mol. Phys. 1971, 21,
1027-1032. (b) Weiss, R.; Schloter, K. Tetrahedron Lett. 1975, 3491-
3494. (c) Weiss, R.; Grimmeiss, A. M. H. Z. Naturforsch. 1989, 44b,
1447-1450. (d) Surman, P. W. J .; Anderson, R. F.; Packer, J . E.; Taylor,
M. J . J . Phys. Chem. A 1997, 101, 2732-2734. (e) Rudenko, A. P.;
Pragst, F. J . Prakt. Chem. 1983, 325, 325-341. (f) Farcasiu, D.;
Balaban, A. T.; Bologa, U. L. Heterocycles 1994, 37, 1165-1192.
cations E•+
. Since the reactions are performed under
oxidative conditions, the R-carbonyl radicals R•, which
(6) Schmittel, M.; Ro¨ck, M. J . prakt. Chem. 1994, 336, 325-330.
(7) Schmittel, M.; Ro¨ck, M. Chem. Ber. 1992, 125, 1611-1620.
S0022-3263(98)00859-7 CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/01/1998

First Example of a Radical Dication Rearrangement
J . Org. Chem., Vol. 63, No. 21, 1998 7329
Ta ble 1. Rea ction Tim es, On e-Electr on Oxid a n ts, a n d
Yield s of th e En ol Oxid a tion s
Sch em e 1. Mech a n ism of th e En ol Oxid a tion
reaction
time
% of
reaction
% of
oxidanta
E1 TTA
(min) benzofuran B time (h) benzofuran B
5
10
2
1
1
71
47
12
31
22
E2 F ep h en
E3 Cu (OTf)2
E4 Cu (OTf)2
E5 Cu (OTf)2
3
3.5
3.5
80
52
68
a
TTA ) tritolylaminium hexafluoroantimonate, Fephen
)
tris[1,10-phenanthroline]iron(III) hexafluorophosphate, Cu(OTf)2
) copper(II) trifluoromethanesulfonate.
tautomerically pure form after column chromatography
and recrystallization from ethanol.
exhibit lower oxidation potentials than the enols,8 are
readily oxidized. The resultant R-carbonyl cations C+
undergo a rapid Nazarov-type cyclization to the dihy-
drobenzofuranyl cations X+, which after a [1,2]-methyl
shift and deprotonation furnish the benzofurans B.
The [1,2]-methyl shift constitutes the slowest step in
the oxidative benzofuran formation and is slower the
more electron-releasing the substituent R. The rate
constant of the 1,2-methyl shift in the phenyl-substituted
dihydrobenzofuranyl cation X+ (R ) Ph) was determined
to be k ) 1.8 × 10-4 s-1 at -20 °C,9 whereas when
stabilization is exerted by a p-C6H4NMe2 group, the rate
constant at room temperature is k ) 2.5 × 10-3 s-1 3
.
The structures of the enols were assigned on the basis
of their characteristic IR, 1H NMR, and 13C NMR
spectroscopic data as well as by elemental analysis or
high-resolution mass spectrometry since no crystals
suitable for X-ray analysis have been obtained so far. The
IR spectra of the enols (recorded in KBr) exhibited strong
and sharp O-H absorptions around 3500 cm-1 and
furthermore strong CdC absorptions around 1600 cm-1
We now wish to report on the synthesis and the
electroanalytical investigations10 of several novel â,â-
dimesityl enols linked to electron-releasing substituents
in the R-position. After oxidation, these enols give rise
to the formation of persistent dihydrobenzofuranyl cat-
ions X+, the rearrangement of which to the benzofurans
can be markedly accelerated via radical dication inter-
mediates.
1
typical for conjugated systems. The H NMR spectra of
the enols exhibited only a weak coalescence at room
temperature, with sharp singlets being obtained for the
OH groups at roughly 5 ppm. The sharp OH absorption
in the IR and 1H NMR spectra, the lack of any CdO
absorption in the IR and 13C NMR spectra, and a
comparison of the spectroscopic data with those of other
enols14 are all in line with their structural assignment
as stable simple enols.
Oxid a tion of th e En ols to Ben zofu r a n s. Oxidation
of enols E usually constitutes a fast reaction15 (reaction
time about 1-20 s) that furnishes the anticipated ben-
zofurans B in good yields (cf. Scheme 1).6
Resu lts
En ols. To stabilize the dihydrobenzofuranyl cation X+
we have chosen to prepare the â,â-dimesityl enols E1-
E6 with either electron-rich aryl groups or heterocycles11
in the R-position. The synthesis of enols E1-E6 was
straightforward. They were prepared from dimesityl
ketene, which was allowed to react with the appropriate
lithium reagents, an approach that was originally intro-
duced by Fuson et al.12 and has been used more recently
by Rappoport et al.13 The pure enols were obtained in
The oxidations of the enols E1-E6 were performed in
acetonitrile using 200 mol-% of the appropriate one-
electron oxidant, i.e., tri(p-tolyl)aminium hexafluoroan-
timonate (TTA, E1/2 ) 0.39 VFc) or copper(II) triflate
(Cu (OTf)2, E1/2 ) 0.67 VFc). The reactions were quenched
by adding aqueous NaHCO3 (for reaction times see Table
1), and the pure benzofurans B were obtained after
column chromatography.
(8) Schmittel, M.; Ro¨ck, M. J . Chem. Soc., Chem. Commun. 1993,
1739-1741.
(9) Schmittel, M.; Gescheidt, G.; Eberson, L.; Trenkle, H., J . Chem.
Soc., Perkin Trans. 2 1997, 2145-2150.
(10) (a) Heinze, J . Angew. Chem., Int. Ed. Engl. 1984, 23, 831-847.
(b) Wipf, D. O.; Wightman, R. M. Anal. Chem. 1988, 60, 2460-2464.
(c) Andrieux, C. P.; Hapiot, P.; Saveant, J .-M. Chem. Rev. 1990, 90,
723-738. (d) Rongfeng, Z.; Evans, D. H. J . Electroanal. Chem.
Interfacial Electrochem. 1995, 385, 201-207. (e) Hu, K.; Evans, D. H.
J . Phys. Chem. 1996, 100, 3030-3036.
(11) Details of the synthesis of E5 and E6 are published in the
context of the preparation of other stable, simple enols carrying
heterocycles. Schmittel, M.; Langels, A. Unpublished results.
(12) Fuson, R. C.; Rowland, S. P. J . Am. Chem. Soc. 1943, 65, 992-
993.
Notably, in the oxidation of enols E3-E5 the amount
of formed benzofuran increased with longer reactions
(14) Rappoport, Z.; Nadler, E. B. J . Am. Chem. Soc. 1987, 109,
(13) (a) Nugiel, D. A.; Rappoport, Z. J . Am. Chem. Soc. 1985, 107,
3669-3676. (b) Rappoport, Z.; Biali, S. E. Acc. Chem. Res. 1988, 21,
442-449.
2112-2127.
(15) Schmittel, M.; Baumann, U. Angew. Chem., Int. Ed. Engl. 1994,
33, 1961-1963.

7330 J . Org. Chem., Vol. 63, No. 21, 1998
Schmittel and Langels
F igu r e 1. Yield of benzofuran B3 after different reaction
times.
F igu r e 2. AM1-calculated minimum structure of the dihy-
drobenzofuranyl cation X1+.
times, as illustrated in more detail with E3. To monitor
the time dependence, a larger amount of enol E3 was
oxidized with copper(II) triflate and samples were with-
drawn and hydrolyzed at discrete reaction times. The
yields were determined by 1H NMR after adding m-
nitroacetophenone as internal standard. The yield steadily
increases until after 80 min finally 80% of the benzofuran
B3 is afforded (Figure 1).
It is noteworthy that the increased yields of B3 with
longer reaction times do not result from the presence of
acid. This was shown by a control experiment in which
the oxidation was carried out in the presence of trifluo-
roacetic acid. The yield of B3 obtained in this experiment
was in the same range (E3: 17%, 1 min) as in the
experiments without acid. The oxidation of the furyl-
substituted enol E6 proved to be much more complex and
will be described elsewhere.11
as individual singlets, indicative of the fact that the
rotation of the remaining mesityl group is hindered. In
agreement with these 1H NMR data, the AM1-calcu-
lated17 minimum structure is characterized by the mesi-
tyl group placed almost perpendicular to the N,N-
dimethylaminophenyl unit (Figure 2).
Similarly, enol E3 in acetonitrile-d3 was oxidized with
the appropriate amount of copper(II) triflate in acetoni-
trile-d3 directly in the NMR tube at room temperature.
The spectrum exhibits seven signals in the aromatic
and vinylic region and several different methyl groups
(Figure 3). Unfortunately, two methyl groups coincide
with the signal of the solvent. The signals at 6.44 and
6.81 ppm belong to the two vinylic protons of the
dihydrobenzofuranyl cation. The remaining mesityl ring
is no longer able to rotate since two signals are obtained
for the two mesityl protons (7.14 and 7.19 ppm). The
most indicative signals in the spectra are the singlet at
1.64 ppm and the doublet at 7.79 ppm. The signal at
1.64 ppm can be assigned analogously as in X1+ to the
methyl group at the 7a-position of the dihydrobenzofura-
nyl ring, whereas the doublet (J ) 9.1 Hz) at 7.79 ppm
belongs to 6′-H of the 3,4-dimethoxyphenyl ring. The
different shifts of 2′-H (appears as doublet at 7.27 ppm)
and 6′-H are apparently due to fact that 2′-H resides in
the shielding area of the mesityl ring which is placed
perpendicular to the 3,4-dimethoxyphenyl group.
An additional, albeit indirect, structural proof for the
occurrence of dihydrobenzofuranyl cations X+ in the
oxidation of the corresponding enols E1-E6 could be
obtained by trapping the cation X3+ with water as its
OH adduct. When the oxidation of E3 with 2 equiv of
copper(II) triflate in acetonitrile was immediately
quenched by adding aqueous NaHCO3 solution, product
X3-OH could be intercepted. It was isolated in 90%
yield after column chromatography and characterized by
1H NMR, 13C NMR, IR, and MS.
Id en tifica tion of Dih yd r oben zofu r a n yl Ca tion s a s
P er sisten t In ter m ed ia tes. To characterize the dihy-
1
drobenzofuranyl cations by H NMR, enols E1 and E3
were directly oxidized in the NMR tube. Oxidation of
enol E1 or the enolate A1 with tris(p-bromophenyl)-
aminium hexachloroantimonate (E1/2 ) 0.70 VFc) or
copper(II) triflate in CDCl3 or acetonitrile-d3 led to the
quantitative formation of a species that could be identi-
fied as the dihydrobenzofuranyl cation X1+. The spectra
recorded at -15 °C exhibited some interesting features.
Characteristically, the signals of the NMe2 group no
longer appear as one singlet but as two singlets (δ ) 3.35
and 3.43 ppm). This hints at high barriers for Cbf,Car
and Car,N bond rotation, the double-bond character of
which can be rationalized because of the significant
stabilization of the cation through the p-NMe2 group.16
Furthermore, a doublet (J ) 9.8 Hz) for one hydrogen
atom showed up downfield (δ ) 8.27 ppm), which can be
assigned to Hb. In contrast, Hb′ resides in the shielding
region of the mesityl ring and appears together with Ha
and Ha′ at δ ) 7.04-7.18 ppm (multiplet). The singlets
at 6.46 and 6.83 ppm can be assigned to the vinylic
protons at the former mesityl ring. Likewise, a new
aliphatic methyl group showed up as a singlet at δ ) 1.64
ppm, which can be assigned to the methyl group at the
7a-position of the dihydrobenzofuran ring. The other
methyl groups (two methyl groups from the dihydroben-
zofuranyl cation and three from the mesityl ring) appear
(16) Analogously in structurally related ammonium ions hindered
rotation about the C,N bond was reported. Pindur, U. Arch. Pharm.
(Weinheim, Ger.) 1980, 313, 301-306.
A second, but low-yield trapping product could be
characterized only by 1H NMR spectroscopy. In the

First Example of a Radical Dication Rearrangement
J . Org. Chem., Vol. 63, No. 21, 1998 7331
F igu r e 3. 1H NMR spectra of the dihydrobenzofuranyl cation X3+.
F igu r e 4. UV/vis spectra of (a) cation X3+ and (b) cation X6+ in acetonitrile at different times.
second isomer, the hydroxyl group is placed in the 3a-
Ta ble 2. Deter m in a tion of th e Ra te Con sta n t k a n d th e
position of the dihydrobenzofuranyl ring. Cyclic volta-
mmetry investigations revealed for X3-OH an irrevers-
ible peak potential Epa ) 0.92 VFc in acetonitrile at v )
100 mV s-1. Notably, X3-OH could be transformed in
84% yield to the corresponding benzofuran B3 by addition
of excess trifluoroacetic acid (2 h at room temperature).
P er sisten cy of th e Dih yd r oben zofu r a n yl Ca tion s
X+. The disappearance of the dihydrobenzofuranyl cat-
ions X+ was kinetically followed by UV/vis spectroscopy.
Cations X+ were prepared directly in the UV cell by
adding an appropriate amount of the dissolved oxidant
to the enol solution. In general, the dihydrobenzofuranyl
cations X+ exhibited two absorption bands around 500
and 370 nm in acetonitrile, which decreased continuously
with time. Hence, the determination of their half-life t1/2
was accomplished by following the decay of their absorp-
tion at a fixed wavelength (Table 2). During the kinetic
investigations also the formation of the corresponding
benzofurans B, could be monitored, since the emerging
absorption bands at 320-330 nm were unambiguously
assigned to B as proven by their independent UV/vis
characterization (Figure 4).
Ha lf-Life t1/2 of Ca tion s X1+, X3+, X5+, a n d X6+
wave length isosbestic point
cation
k (min-1
)
t1/2 (min)
(nm)
(nm)
X1+
X3+
X5+
X6+
1.50 × 10-1
2.28 × 10-2
3.78 × 10-2
4.00 × 10-2
4.6a
30.4b
18.3b
17.3b
512
512
509
340
350
282, 334
a
b
Determined by 1H NMR kinetics. Determined by UV/vis
kinetics.
Interestingly, the dihydrobenzofuranyl cation X6+
provided similar UV/vis spectra, although the anticipated
benzofuran B6 could not be detected in preparative
oxidation experiments with enol E6.11
To prepare X1+, enolate A1 was oxidized in order to
circumvent problems resulting from protonation at the
nitrogen that arise when starting from E1. However,
oxidation of enolate A1 did not lead to the clean formation
of benzofuran B1. Since some of the products exhibited
absorption bands almost identical to the dihydrobenzo-
furanyl cation X1+, we could not use the UV/vis technique
here. Instead the half-life of dihydrobenzofuranyl cation
X1+ was determined by H NMR kinetic investigations
1
(17) Dewar, M. J . S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J . J . P.
J . Am. Chem. Soc. 1985, 107, 3902-3909.
to t1/2 ) 4.6 min at room temperature.

7332 J . Org. Chem., Vol. 63, No. 21, 1998
Schmittel and Langels
Ta ble 3. Oxid a tion P oten tia ls of Ben zofu r a n s B a t ν )
100 m V s-1 in Aceton itr ile a n d Dich lor om eth a n e/TF AN
CH3CN
1/2(B)
Epa(B•+
(VFc (VFc
CH2Cl2/TFAN
E1/2(B) )
Epa(B•+
E
)
benzofuran
)
)
(VFc
)
(VFc)
B1
B2
B3
B4
B5
0.23
0.69c
0.65
0.71
0.68
0.72a
b
1.02
1.07
d
b
b
0.56
0.63
b
b
b
1.08
1.08
b
a
b
Partially reversible, no follow-up products detectable. Not
determined. c Only partially reversible. The radical cation of the
d
benzofuran B5 dimerizes to the bithiophene derivative. Schmittel,
M.; Langels, A. Unpublished Results.
F igu r e 5. Cyclic voltammogram of B3 in dichloromethane/
trifluoroacetic anhydride (ν ) 100 mV s-1, c ) 1 mM, working
electrode ) 1 mm platinum disk, counter electrode ) platinum
wire, reference electrode ) silver wire).
F igu r e 6. Cyclic voltammograms of (a) E1 and (b) E3 (ν )
100 mV s-1, c ) 1 mM in acetonitrile, working electrode ) 1
mm platinum disk, counter electrode ) platinum wire, refer-
ence electrode ) silver wire).
Cyclic Volta m m etr y of th e Ben zofu r a n s. The
cyclic voltammetry investigations of the electron-rich
benzofurans exhibited several new aspects.6,8 Usually,
only a single, reversible oxidation wave is obtained in
the CV for the benzofurans at about 0.81-0.99 VFc.8 The
benzofurans B1-B5 are reversibly oxidized at distinc-
tively lower potentials due to the electron-releasing
character of their substituents, but most notably, the first
oxidation wave of the benzofurans is followed by a second
one. The second wave constitutes the oxidation of the
benzofuran radical cations B•+ to the corresponding
dications B2+ (Table 3).
In the case of benzofuran B1, the second wave is
partially reversible in acetonitrile, whereas for the ben-
zofurans B3 and B4 the second oxidation wave is
irreversible. Small reduction waves could only be moni-
tored at 0 °C in acetonitrile or at higher scan rates (ν )
1 or 2 V s-1). However, when performing the CV
experiments in dichloromethane or in dichloromethane
with added trifluoroacetic anhydride (TFAN), reversibil-
ity was attained for both redox steps (Figure 5), allowing
the determination of the thermodynamic half-wave po-
tentials for B•+ to B2+ even at low scan rates (ν g 50 mV
s-1).
ible oxidation waves emerged. Once more for all systems
E1-E6, the potential difference between the two oxida-
tion waves is 0.9-1.1 V (Figure 6).
Among the model compounds, enol E1 occupies a
special place since the cyclic voltammogram (CV) of enol
E1 exhibited two additional oxidation waves besides the
first oxidation wave. The second wave at Epa ) 0.76 VFc
(wave Ib in Figure 6a) in the CV of E1 could be readily
assigned to the oxidation of the protonated enol E1-H+
by control experiments in the presence of added base (4-
N,N-(dimethylamino)pyridine) and acid (trifluoromethane-
sulfonic acid). The oxidation chemistry of E1-H+ has
already been extensively described and shall not be
repeated here.3 Wave II in the CV of enol E1 seems to
correspond to the second waves in the CVs of the other
enols as judged by its anodic shift vs E1 (Table 4).
Cyclic Volta m m etr y of th e En ols. The enols E1-
E6 were submitted to cyclic voltammetry experiments at
different scan rates. In standard cyclic voltammetry
investigations irreversible oxidation waves were moni-
tored for all enols at scan rates between 20 and 500 mV
s-1, indicative of fast follow-up reactions (deprotonation,
see Scheme 1) of the generated enol radical cations.
Noticeably, all enols displayed an additional irreversible
oxidation wave at higher anodic potential, which is
uniformly placed 0.9-1.2 V above the first oxidation
wave. In dichloromethane as solvent again two irrevers-
Since in earlier investigations the dihydrobenzofuranyl
cations X+ of two systems (R ) Ph, p-C6H4NMe2)2,9 had
proved to be rather long-lived, we supposed that the
additional wave in the cyclic voltammograms of E1-E6
was due to the oxidation of X1+-X6+. When enols E1,
E3, and E4 were submitted to multisweep cyclic volta-
(18) When scanning beyond the second wave, fast coating of the
electrode hampered the multisweep cyclic investigations of E5 and E6.

First Example of a Radical Dication Rearrangement
J . Org. Chem., Vol. 63, No. 21, 1998 7333
Ta ble 4. Oxid a tion P oten tia ls of En ols E1-E6 a n d
Dih yd r oben zofu r a n yl Ca tion s X1+-X6+ a t ν ) 100 m V s-1
F igu r e 8. Multisweep cyclic voltammograms of E3: (top)
digital simulation; (bottom) experiment in dichloromethane/
trifluoroacetic anhydride.
to complex protonation/deprotonation equilibria.3 But in
the CV of enol E319 two reversible redox waves emerged
belonging to the oxidation (reduction) of the formed
benzofuran B3 (see Figure 7b). This was also the case
when investigating E4 in dichloromethane.
Digital simulation of the multisweep cyclic voltammo-
gram of enol E3 using the implicit Crank-Nicholson
technique20 showed good agreement with the experimen-
tal curve based on an ECECEC mechanism (see Schemes
1 and 2). A comparison of the experimental curve
(disregarding the extra wave caused by enol protonation
at Epa ) 0.76 VFc) and the simulated curve indicated that
especially the current intensities of the formed benzofu-
ran B3 and the corresponding radical cation B3•+ were
reproduced very well (Figure 8).
The oxidation potentials of X1+ and X3+ could also be
determined in independent cyclic voltammetry investiga-
tions. When the corresponding enols E1 and E3 were
oxidized with 2 equiv of a one-electron oxidant and when
the CVs of the resultant solution were immediately
recorded in acetonitrile, an irreversible wave appeared
at 1.12 VFc for X1+, whereas for X3+ an irreversible wave
at 1.55 VFc could be monitored.
F igu r e 7. Multisweep cyclic voltammograms of (a) E1 (ν )
500 mV s-1, c ) 1 mM in acetonitrile) and (b) E3 (ν ) 500 mV
s-1, c ) 1 mM in dichloromethane/trifluoroacetic anhydride);
working electrode ) 1 mm platinum disk, counter electrode )
platinum wire, reference electrode ) silver wire.
mmogram investigations at the anodic potential of the
second wave,18 the formation of two new waves could be
detected (Figure 7).
The new waves were unambiguously assigned to the
corresponding benzofurans (vide supra). For enol E1 only
the reduction waves could be detected (see Figure 7a) due
Discu ssion
En ol Oxid a tion . At first glance oxidation of electron-
rich enols E1-E6 seems to follow the well-established
mechanistic pattern outlined in Scheme 1, although
reaction times proved to be much longer than observed
earlier. In contrast to less electron-rich â,â-dimesityl
enols,15 however, none of the cyclic voltammograms
exhibited any benzofuran formation on the time scale of
the cyclic voltammetry experiment but instead displayed
a new oxidation wave. Notably, the second wave exhibits
the same anodic peak current as the enol oxidation wave
(see Table 5). Since the enol wave is characterized by
an ECE process (two electrons consumed), the second
(19) Unfortunately the enol oxidation wave in dichloromethane splits
up into two oxidation waves after some time. Supposedly a protonated
species emerges since the wave disappears in the presence of base.
Another severe problem in this MS CV experiment was the coating of
the electrode after scanning the second wave.
(20) (a) Lasia, A. J . Electroanal. Chem. Interfacial Electrochem.
1983, 146, 397-411. (b) Heinze, J .; Sto¨rzbach, M.; Mortensen, M. J .
Electroanal. Chem. Interfacial Electrochem. 1984, 165, 61-70. (c)
Heinze, J .; Sto¨rzbach, M. J . Electroanal. Chem. Interfacial Electrochem.
1993, 346, 1-27.

7334 J . Org. Chem., Vol. 63, No. 21, 1998
Schmittel and Langels
Ta ble 5. Com p a r ison of th e An od ic P ea k Cu r r en ts of
Wa ve I a n d Wa ve II in th e CVs of th e En ols E1 a n d
E3-E6
stable cation (t1/2 ) 30.4 min) in this series, whereas the
cation X1+ exhibited the shortest half-life (t1/2 ) 4.6 min).
This difference in half-life compared to X3+ is not easy
to rationalize at first since a p-Me2N group in X1+ should
be a much better electron donor than the methoxy groups.
But in this case protonation/deprotonation equilibria in
X1+ complicate the situation.3 Protons that are released
in the oxidation process lead to protonation of the NMe2
group and thus change the electronic nature of the
substituent (from electron releasing to electron with-
drawing). In the protonated X1-H+ the [1,2]-methyl
migration is very rapid.
Ipa(wave I)/Ipa(wave II) of
scan rate
(mV s-1
)
E3
E4
E5
E6
50
100
200
500
1.0:0.9
1.0:0.9
1.0:0.9
1.0:0.9
1.0:1.0
1.0:1.0
1.0:1.0
1.0:1.0
1.0:0.8
1.0:0.8
1.0:0.8
1.0:0.9
1.0:0.9
1.0:0.8
1.0:0.8
1.0:0.8
The half-lives of the thienyl-substituted cation X5+ and
the furyl-substituted X6+ are almost identical. The fact
that thiophene stabilizes the cationic charge a bit better
than its oxygen analogue is expected in the light of their
electronic nature as jugded by the difference of the
oxidation potentials of thiophene and furan.21
Altogether, the investigations demonstrate that the
dihydrobenzofuranyl cations X1+-X6+ are well stabilized
by the electron-releasing substituents, with the conse-
quence that the [1,2]-methyl shift only occurs very slowly.
The persistency of the cations X3+-X5+ is also the reason
for the longer reaction times (2 h instead of several
minutes) needed in the preparative one-electron oxida-
tions of the enols E3-E5.
A special situation is given with E6. The UV/vis
spectroscopic investigations indicate that the benzofuran
B6 is indeed initially formed, but in the preparative
oxidation the benzofuran could not be detected. This is
most likely due to follow-up reactions, presumably caused
by protons that are released in the reaction course. One
could envisage that the furan ring opens in the presence
of acid (enol ether cleavage) leading to polymerization.11
Ra d ica l Dica tion s a s In ter m ed ia tes in th e Ben -
zofu r a n F or m a tion . We have observed in all cyclic
voltammetry investigations on enols E1-E6 a new
oxidation wave that can now be assigned to the oxidation
of X+.
F igu r e 9. AM1-calculated charge density in the cyclohexa-
dienyl cation X3+ and the corresponding radical dication X3•2+
.
wave should equally correspond to an overall two-
electron-transfer process.
For all model systems the second wave is shifted 0.9-
1.2 V anodically vs the enol wave (Table 4). This
proposes that an intermediate on the way to the benzo-
furan B is rather long-lived and its oxidation wave is
therefore observable in the cyclic voltammetry investiga-
tions.
Dih yd r oben zofu r a n yl Ca tion s. The long-lived in-
termediate in the oxidative transformation could be
unambiguously assigned as the dihydrobenzofuranyl
cation X1+-X6+. The structural assignment is based (1)
on direct NMR evidence for dihydrobenzofuranyl cations
X1+and X3+ and (2) on trapping X3+ with water provid-
ing the OH adducts X3-OH and its isomer in basically
quantitative yield.
AM1 calculations17 on X3+ (Figure 9) indeed reveal that
the positive charge is placed next to the oxygen atom in
the furan ring (2-position), which is the preferred position
for attack of water as a nucleophile leading to X3-OH.
The half-life of X1+, X3+, X5+, and X6+ has been
measured either by UV/vis or by 1H NMR (Table 2). The
dihydrobenzofuranyl cation X3+ constitutes the most
This assignment is based on several observations.
(1) Dihydrobenzofuranyl cations obtained in prepara-
tive oxidations and identified by 1H NMR exhibited
oxidation waves at potentials identical to waves II in the
oxidations of the corresponding enols.
(2) The oxidation potentials of X3+-X6+ proved to be
identical in acetonitrile and dichloromethane (see Table
4). This observation excludes the possibility of strong
acetonitrile complexation to X+ (in acetonitrile as solvent)
and the occurrence of nitrilium adducts.
(3) The oxidation potential of X3-OH (Epa ) 0.91 VFc)
differed markedly from that of wave II (Epa ) 1.54 VFc)
in the enol E3 oxidation. This demonstrates that alcohols
X-OH, possibly formed by spurious traces of water in
either acetonitrile or dichloromethane, cannot be respon-
sible for the formation of waves II.
(21) Mizuno, K.; Ishii, M.; Otsuji, Y. J . Am. Chem. Soc. 1981, 103,
5570-5572.

First Example of a Radical Dication Rearrangement
J . Org. Chem., Vol. 63, No. 21, 1998 7335
Sch em e 2. Rea ction P a th w a ys of X+ a n d X•2+
stabilized vs the [1,2]-methyl shift. On the other hand,
it might be possible that the increased reactivity is due
to an inherently higher reactivity of the radical dications
X•2+ compared to the cations X+ as a consequence of their
odd-electron nature. Such an inherently higher reactivity
is characteristic for radical cations when comparing them
with their neutral counterparts.22 It is noteworthy that
no common-sense route to benzofuran B is conceivable
by deprotonation of X•2+, although the radical dication
should be highly acidic.
Besides the high reactivity the radical dications X•2+
also exhibit a high selectivity. This could be demon-
strated with the help of a digital simulation. For the
simulation of the multisweep experiments of enol E3 a
set of parameters was selected that reflect a very fast
and quantitative transformation of the radical dication
X3•2+ to a persistent benzofuran dication B32+. Since the
simulated and experimental CVs show very good agree-
ment as to what concerns the peak currents of the
benzofuran B32+ and B3•+ (see Figure 8), it can be
concluded that indeed the radical dication X3•2+ reacts
to benzofuran B3 with a high yield.
(4) Benzofuran formation could be monitored in cyclic
voltammetry experiments starting with X3+. For that
purpose enol E3 was oxidized directly in the cyclic
voltammetry cell with 2 equiv of NOSbF 6. After adding
the oxidant the resultant deep red solution exhibited an
irreversible oxidation wave at Epa ) 1.55 VFc, which
decreased slowly (t1/2 ) several minutes). With this wave
decreasing two new waves emerged which belonged to
the benzofuran B3 and also the color of the solutions
slowly turned brighter.
The multisweep cyclic voltammetry investigations
clearly indicate that upon scanning up to wave II in the
enol oxidation, that is after anodic oxidation of X1+, X3+,
and X4+, the corresponding benzofurans are formed
rapidly. The assignment of the new waves to the
benzofurans B is straightforward and basically double-
checked, because it is based on two emerging reduction
waves, one for B2+ and one for B•+ reduction.
Su m m a r y
Six persistent dihydrobenzofuranyl cations have been
investigated by 1H NMR and UV/vis spectroscopy as well
as by cyclic voltammetry. For the first time a selective
and high-yield rearrangement via radical dications has
been unambiguously established. As this reaction was
shown to proceed much more rapidly (>106) than the
corresponding rearrangement of the cations, it is cer-
tainly fair to ask whether other stable or persistent
cations can likewise be rearranged after one-electron
oxidation. Theoretical and further experimental work
about these rare intermediates is hopefully stimulated
by our results.
The cyclic voltammetry and preparative oxidation
investigations now indicate unambiguously that both
species (X+ and X•2+) yield the same product although
with significantly different reaction times. Hence, the
formation of benzofurans B starting from X+ can be
accelerated by selectively activating the stable cations
X+ by one-electron oxidation to their corresponding
radical dications X•2+. The very reactive radical dication
X•2+ undergoes a fast [1,2]-methyl shift followed by
deprotonation to provide the benzofuran radical cation
Exp er im en ta l Section
Ma ter ia ls. Commercial reagents were purchased from
standard chemical suppliers and were used without further
purification. Dimesityl ketene was prepared according to ref.23
Acetonitrile for the cyclic voltammetry investigations and the
one-electron oxidation reactions was of HPLC quality (Riedel-
de Haen) and was distilled from CaH2.
Ap p a r a tu s. Infrared Spectra: Perkin-Elmer 1605 FT-IR
infrared spectrophotometer. UV/vis spectra: Varian Cary 1E
UV/vis spectrophotometer. For all other instrumentations see
ref 1.
Cyclic Volta m m etr y. In a glovebox tetra(n-butyl)ammo-
nium hexafluorophosphate (232 mg, 600 µmol) and the elec-
troactive species (6 µmol) were placed in a thoroughly dried
cyclic voltammetry cell. Outside of the box acetonitrile (6.0
mL) was added through a gastight syringe to the cell hooked
up to a high-purity argon line, and then a 1 mm platinum disk
electrode as working electrode, a Pt wire counter electrode,
and a Ag reference electrode were placed into the solution. The
cyclic voltammograms were recorded at various scan rates
B
•+, which is immediately oxidized at the applied poten-
tial to the dication B2+. The latter is rather persistent
(k ) 3 × 10-2 s-1 for B32+ and 1 s-1 for B42+ in
dichloromethane/TFAN) and can be reduced twice to the
neutral benzofuran B via the benzofuran radical cation
B•+ (Scheme 2).
From the fact that in fast scan cyclic voltammetry
investigations of enol E3 in acetonitrile no reduction
wave is obtained (up to a scan rate of 200 V s-1) for the
oxidation of X3+ to the corresponding radical dication
X3•2+(wave II) it can be concluded that the follow-up
reaction of X3•2+ is very fast (k > 200 s-1). A comparison
(22) (a) Eberson, L. Electron-Transfer Reactions in Organic Chem-
istry; Springer: Berlin, 1987. (b) Roth, H. D. Acc. Chem. Res. 1987,
20, 343-350. (c) Bauld, N. L. Tetrahedron 1989, 45, 5307-5363. (d)
Amatore, C.; Kochi, J . K. Adv. Electron-Transfer Chem. 1991, 1, 55-
148. (e) Mizuno, K.; Otsuji, Y. Top. Curr. Chem. 1994, 169, 301-346.
(f) J ohnston, L. F.; Schepp, N. P. Pure Appl. Chem. 1995, 67, 71-78.
(g) Moeller, K. D. Top. Curr. Chem. 1997, 185, 49-86. (h) Schmittel,
M.; Burghart, A. Angew. Chem., Int. Ed. Engl. 1997, 36, 2550-2589.
(23) Fuson, R. C.; Rowland, S. P.; Armstrong, C. J .; Chadwick, D.
H.; Kneisley, J . W.; Shenk, W. J .; Soper, Q. F. J . Am. Chem. Soc. 1945,
67, 386-393.
with the rate constant for X3+ (k ) 3.8 × 10-4 s-1
)
determined by UV/vis shows that the formation of the
benzofuran B3 is at least accelerated by a factor of 106.
The reason for this acceleration can be seen in the
dramatic change of the electronic situation. Since the
electron-releasing substituent is transformed to an elec-
tron-withdrawing substituent R•+, the system is no longer

7336 J . Org. Chem., Vol. 63, No. 21, 1998
Schmittel and Langels
using a manifold of starting and reversal potentials. For
determination of the oxidation potentials ferrocene (E1/2 ) 0.39
VSCE) was added as internal standard. For fast scan cyclic
voltammetry investigations, 385 mg (1.00 mmol) of supporting
electrolyte, 50 µmol of substrate, and 4.0 mL of solvent were
used. Fast scan cyclic voltammograms were carried out at 25
µm Au ultramicro electrodes, a Pt wire serving as counter
electrode, and a Ag wire as reference electrode. Cyclic volta-
mmograms were recorded using a Princeton Applied Research
model 362 potentiostat with an Philips model PM 8271 XYt-
H), 2.91 (s, 6 H), 5.05 (s, 1 H), 6.49 (d, J ) 9.0 Hz, 2 H), 6.67
(s, 2 H), 6.87 (s, 2 H), 7.20 (d, J ) 9.0 Hz, 2 H). 13C NMR:
(CDCl3, 50 MHz): δ 20.69, 20.81, 21.24, 21.33, 40.25, 108.36,
111.32, 129.18, 129.52, 129.73, 129.85, 130.09, 130.24, 133.49,
135.22, 136.40, 138.22, 150.26, 150.80. HRMS: M+ (C28H33
NO) calcd 399.2562, found 399.2560.
-
2,2-Dim esityl-1-(3,4-d im eth oxyp h en yl)eth en ol (E3). A
solution of n-butyllithium (20.0 mmol, 12.5 mL of a 1.6 M
solution in n-hexane) was added dropwise to an ice-cooled
solution of 4-bromoveratrole (4.34 g, 20.0 mmol) and 20 mL of
dry Et2O. After stirring for 1 h at 0 °C a solution of dimesityl
ketene (5.06 g, 20.0 mmol) in dry Et2O (30 mL) was added,
and the whole reaction mixture was stirred for 3 h at this
temperature and for another 3 h at room temperature. Then
a half-saturated aqueous NH4Cl (30 mL) solution was added,
the aqueous layer was extracted with Et2O (3 × 50 mL), and
the combined organic layers were dried (NaSO4). After
removal of the solvent in vacuo the remaining brown oil was
chromatographed (cyclohexane/ethyl acetate, 3:1) and recrys-
tallized from EtOH to yield the desired compound as a yellow
solid (4.67 g, 11.2 mmol, 56%). E3: mp 166-168 °C. IR
(KBr): ν ) 3517 cm-1 (s, O-H), 2998 (m, C-H), 1609 (s, Cd
C), 1575 (m), 1509 (s), 1320 (m), 1236 (s), 1072 (s), 817 (m).
1H NMR (CDCl3, 200 MHz): δ 1.93, 2.19 and 2.28 (3s,
coalescence, 18 H), 3.48 (s, 3 H), 3.84 (s, 3 H), 5.19 (s, 1 H),
6.71 (m, 4 H), 6.90 (bs, 2 H), 7.06 (dd, J ) 8.5 Hz, J ) 2.1 Hz,
1 H). 13C NMR: (CDCl3, 50 MHz): δ 20.69, 20.81, 55.18, 55.68,
106.70, 108.75, 110.09, 110.13, 110.43, 111.98, 128.67, 129.21,
132.78, 135.63, 135.73, 136.74, 138.20, 147.55, 148.90, 149.93.
Anal. Calcd C 80.73, H 7.74; found C 80.83 H 7.77.
2,2-Dim esityl-1-(6-m eth oxyn a p h th -2-yl)eth en ol (E4).
A solution of 2-bromo-6-methoxynaphthalene (1.44 g, 6.07
mmol) in 15 mL of dry Et2O was cooled to 0 °C and treated
with n-butyllithium (6.07 mmol, 3.80 mL of a 1.6 M solution
in n-hexane). After stirring for 30 min at this temperature a
solution of dimesityl ketene (1.69 g, 6.07 mmol) in 25 mL of
dry Et2O was added. The reaction mixture was stirred for
another 1 h at 0 °C and for 4 h at room temperature before it
was hydrolyzed with half-saturated aqueous NH4Cl (30 mL).
The aqueous phase was extracted with Et2O (4 × 30 mL), the
combined organic layers were dried (Na2SO4), and the solvent
was evaporated. Column chromatography (cyclohexane/ethyl
acetate, 10:1) of the remaining oil yielded the product as a
gray-white solid (572 mg, 130 mmol, 21%). E4: mp 100-101
°C. IR (KBr): ν ) 3504 cm-1 (s, O-H), 3490 (s, O-H), 2916
(m, C-H), 1627 (s), 1607 (s, CdC), 1593 (s), 1481 (s), 1390
(m), 1235 (s), 1059 (m), 850 (s). 1H NMR (CDCl3, 200 MHz):
δ 1.92, 2.19 and 2.30 (3 s, coalescence, 18 H), 3.89 (s, 3 H),
5.26 (s, 1 H), 6.67 (s, 2 H), 6.92 (s, 2 H), 7.03 (d, J ) 2.4 Hz, 1
H) 7.08 (dd, J ) 8.8 Hz, J ) 2.4 Hz, 1 H), 7.24 (dd, J ) 8.6 Hz,
J ) 1.5 Hz, 1 H), 7.43 (d, J ) 8.6 Hz, 1 H), 7.62 (d, J ) 8.8 Hz,
1 H), 7.90 (d, J ) 1.5 Hz, 1 H). 13C NMR: (CDCl3, 63 MHz):28
δ 20.70, 20.79, 20.84, 55.21, 105.48, 107.87, 111.46, 118.56,
125.64, 126.82, 127.86, 128.40, 129.99, 130.39, 132.89, 134.33,
135.40, 135.77, 136.13, 136.83, 137.44, 138.12, 150.30, 158.06.
Anal. Calcd C 85.28, H 7.39; found C 85.06 H 7.53.
recorder for scan rates < 1 V s-1
. For fast scan cyclic
voltammetry, a Hewlett-Packard (HP) model 3314A function
generator was used, connected to a three-electrode potentiostat
developed by C. Amatore.24 Data were recorded by a HP 54510
A digitizing oscilloscope linked to a 486DX33 computer using
the Hewlett-Packard data transfer program Scopelink.
UV/Vis Kin etics. A 1 mL portion of acetonitrile and 0.1
mL of the enol solution in acetonitrile (c ≈ 2.5 mmol L-1) were
placed in a UV/vis cell. After adding the appropriate amount
(2 equiv) of copper(II) triflate in acetontrile (c ≈ 5 mmol L-1
)
the spectra were recorded at room temperature. The deter-
mination of the half-life of the dihydrobenzofuranyl cations
X+ was accomplished by following the decay of their absorption
at a fixed wavelength at room temperature.
Digita l Sim u la tion of th e Mu ltisw eep Cyclic Volta m -
m ogr a m of En ol E3. The computer simulation of the redox
chemistry of enol E3 was carried out on P166+ computer using
the Crank-Nicholson technique20 and DigiSim.25 All chemical
reactions were assumed to be irreversible first-order processes
except the ET equilibria. With a standard diffusion constant
of D ) 10-5 cm2 s-1, the heterogeneous electron-transfer rate
constant of k0hetero and the rate constants of the chemical steps
were varied to achieve the best possible agreement with the
experimental curves. Only for the follow-up reaction of the
dihydrobenzofuranyl cation X3+ was the rate constant (kf )
2.28 × 10-2 min-1) as determined from UV/vis experiments
used.
Ca lcu la tion . The force-field program SYBYL26 was em-
ployed to minimize the energies. The AM1 parameter file was
used from VAMP 5.0.27
Syn t h esis. The synthesis of 2,2-dimesityl-1-(4-methoxy-
phenyl)ethenol (E2)14 and 2-(4-methoxyphenyl)-3-mesityl-4,6,7-
trimethylbenzofuran (B2)6 has already been described. The
synthesis of 2,2-dimesityl-1-(2-thienyl)ethenol (E5), 2,2-di-
mesityl-1-(2-furyl)ethenol (E6), and 2-(2-thienyl)-3-mesityl-
4,6,7-trimethylbenzofuran (B5) is described elsewhere.11
2,2-Dim esit yl-1-(N,N-d im et h yla m in op h en yl)et h en ol
(E1). To a solution of 4-bromo-N,N-dimethylaniline (3.30 g,
16.5 mmol) in dry diethyl ether (20 mL) was added dropwise
n-butyllithium (15.5 mmol, 6.60 mL of a 2.5 M solution in
hexane). The reaction mixture was refluxed for 17 h before
cooling it to 0 °C. At this temperature, a solution of dimesityl
ketene (3.28 g, 11.8 mmol) in dry diethyl ether (25 mL) was
added, and then the solution was slowly brought to reflux.
After 20 h the mixture was allowed to cool to room temperature
before hydrolyzing it with half-saturated NH4Cl (30 mL). The
aqueous layer was extracted with diethyl ether (3 × 35 mL),
and the combined organic layers were dried (MgSO4). The
solvent was removed in vacuo, and the remaining brown oil
purified by chromatography on silica gel (cyclohexane/ethyl
acetate 1:1). The pure product crystallized from ethanol
affording 480 mg (2.20 mmol, 10%), mp 205-208 °C. IR
(KBr): ν ) 3494 cm-1 (s, O-H), 2912 (m, C-H), 1607 (s, Cd
C), 1522 (m), 1362 (m), 1236 (m), 1062 (m), 812 (s). 1H NMR
(CDCl3, 200 MHz): δ 1.90, 2.20 and 2.26 (3s, coalescence, 18
2-(p -N,N-Dim et h yla m in op h en yl)-3-m esit yl-4,6,7-t r i-
m eth ylben zofu r a n (B1). In a glovebox tri(p-tolyl)aminium
hexafluoroantimonate (44.0 mg, 70.0 µmol) and the enol E1
(14.0 mg, 35.0 µmol) were placed into two separate test tubes
equipped with stirring rods. At a high-purity argon line 3 mL
of acetonitrile was added to each test tube to dissolve the
reactants. The solution of the one-electron oxidant was then
added through a syringe to the solution of the enol. After 5
min the mixture was quenched with 2 mL of saturated aqueous
NaHCO3 and diluted with 10 mL of CH2Cl2 and 10 mL of H2O.
The aqueous layer was extracted three times with CH2Cl2. The
combined organic layers were washed with saturated aqueous
NaCl and water and dried over Na2SO4. Removal of the
solvent afforded the crude product, which was purified by
chromatography on silica gel (cyclohexane/dichloromethane,
(24) Amatore, C.; Lefrou, C.; Pflu¨gler, F. J . Electroanal. Chem. 1989,
270, 43-59.
(25) (a) Rudolph, M. J . Electroanal. Chem. Interfacial Electrochem.
1992, 338, 85-98. (b) Rudolph, M.; Reddy, J . P.; Feldberg, S. W. Anal.
Chem. 1994, 66, 589 A-600A.
(26) SYBYL; Tripos Associates: 1699 St. Hanley Road, Suite 303,
St. Louis, MO, 63144.
(27) Rauhut, G.; Chandrasekhar, J .; Alex, A.; Beck, B.; Sauer, W.;
Clark, T. VAMP 5.0; Oxford Molecular Limited: The Magdalen Centre,
Oxford Science Park, Sandforde on Thames, Oxford 4GA, England.
(28) Due to the nature of the â,â-dimesityl unit the 13C signals often
coincide. In enol E4 the two signals for the p-Mes-CH3 groups coincide.

First Example of a Radical Dication Rearrangement
J . Org. Chem., Vol. 63, No. 21, 1998 7337
1:1 and 2:1) furnishing 9.8 mg (25 µmol, 71%) of benzofuran
B1. IR (KBr): ν ) 2917 cm -1 (s, C-H), 2848 (m, C-H), 1612
(s, CdC), 1522 (s), 1443 (m), 1355 (m), 1195 (m), 1115 (m),
1064 (w), 854 (w), 820 (m). 1H NMR (CDCl3, 200 MHz): δ )
1.85 (s, 3 H), 2.02 (s, 6 H), 2.35 and 2.37 (2 s, 6 H), 2.50 (s, 3
H), 2.93 (s, 6 H), 6.60 (d, J ) 9.2 Hz, 2 H), 6.71 (s, 1 H), 6.96
(s, 2 H), 7.39 (d, J ) 9.2 Hz, 2 H). 13C NMR (CDCl3, 50 MHz):
δ 11.41, 17.15, 19.00, 20.39, 21.24, 40.25, 112.14, 113.02,
120.09, 125.21, 125.85, 125.97, 126.21, 127.82, 128.15, 131.09,
131.58, 136.98, 137.79, 148.98, 149.68, 153.23. HRMS: M+
(C28H31NO) calcd 397.2406, found 397.2406.
130.00, 130.73, 133.83, 136.20, 137.40, 137.61, 148.33, 153.50,
158.30. HRMS: M+ (C31H30O2) calcd 434.2245, found 434.2246.
2-(3,4-Dim et h oxyp h en yl)-3-m esit yl-4,6,7a -t r im et h yl-
2,7a -d ih yd r oben zo[b]fu r a n -2-ol. Enol E3 (137 mg, 329
µmol) and copper(II) triflate (238 mg, 658 µmol) were sepa-
rately weighed into different flasks. Acetonitrile was added
by means of gastight syringes to the enol (10 mL) and to the
oxidant (5 mL). When both compounds had dissolved, the one-
electron oxidant solution was given to the enol solution and 5
mL of saturated NaHCO3 was immediately added to the deep
red solution. The aqueous layer was extracted three times
with CH2Cl2 (10 mL). After washing with saturated NaCl (3
× 10 mL) the organic layers were dried (Na2SO4). Thereafter
the solvents were removed in vacuo and the remaining solid
was chromatographed (cyclohexane/dichloromethane, 3:1) yield-
ing X3-OH as yellow-red solid (128 mg, 29.6 µmol, 90%), mp
73-74 °C. IR (KBr): ν˜ ) 3474 (m, O-H), 2918 (m, C-H),
2862 (m, C-H), 1608 (w, CdC), 1514 (s), 1465 (m), 1263 (s),
1165 (m), 1137 (m), 1029 (s), 923 (m), 861 (m), 807 (m), 764
2-(3,4-Dim eth oxyp h en yl)-3-m esityl-4,6,7-tr im eth ylben -
zofu r a n (B3). E3 (33.0 mg, 79.2 µmol) and copper(II) triflate
(57.3 mg, 158 µmol) were separately weighed into different
test tubes. A 3 mL portion of acetonitrile was added by means
of gastight syringes to each test tube. When both compounds
had dissolved, the one-electron oxidant solution was given to
the solution of the enol, and after 1 min 1 mL of saturated
NaHCO3 was added. After removal of the solvents in vacuo
the remaining crude product was dissolved in diethyl ether
and chromatographed (cyclohexane/dichloromethane, 3:1) to
yield the pure benzofuran B3 as a white solid (26.3 mg, 63.5
µmol, 80%). B3: IR (KBr): ν ) 2915 cm-1 (m, C-H), 2861
(w, C-H), 1608 (m, CdC), 1513 (s), 1459 (m), 1266 (s), 1144
(m), 1121 (m), 1027 (s), 858 (s), 806 (s). 1H NMR (CDCl3, 200
MHz): δ ) 1.91 (s, 3 H), 2.03 (s, 6 H), 2.35 (s, 3 H), 2.37 (s, 3
H), 2.51 (s, 3 H), 3.55 (s, 3 H), 3.86 (s, 3 H), 6.75 (s, 1 H), 6.81
(d, J ) 8.5 Hz, 1 H), 6.89 (d, J ) 2.1 Hz, 1 H), 6.99 (s, 2 H),
7.25 (dd, J ) 8.5 Hz, J ) 2.1 Hz, 1 H). 13C NMR (CDCl3, 50
MHz): δ 11.48, 17.21, 19.06, 20.36, 21.15, 55.20, 55.84, 108.02,
111.20, 114.96, 116.93, 117.66, 124.70, 125.42, 126.12, 128.27,
129.76, 130.97, 132.37, 137.34, 137.76, 148.01, 148.53, 148.71,
153.47. HRMS: M+ (C28H30O3) calcd 414.2195, found 414.2196.
3-Mesityl-2-(6-m eth oxyn aph th -2-yl)-4,6,7-tr im eth ylben -
zofu r a n (B4). E4 (10.9 mg, 25.0 µmol) and Cutriflat (18.1
mg, 50.0 µmol) were separately weighed into different test
tubes. A 3 mL portion of acetonitrile was added by means of
gastight syringes. When both compounds had dissolved, the
one-electron oxidant solution was given to the solution of the
enol, and after 3.5 h 1 mL of saturated NaHCO3 was added.
After removal of the solvents in vacuo the remaining crude
product was dissolved in diethyl ether and chromatographed
(cyclohexane/dichloromethane, 3:1) furnishing the pure ben-
zofuran B4 as a light yellow solid (5.7 mg, 13 µmol, 52%). IR
(KBr): ν ) 2922 cm-1 (s, C-H), 2849 (m, C-H), 1607 (m, Cd
C), 1513 (s), 1463 (m), 1250 (m), 1199 (m), 1116 (m), 1033 (s),
855 (m). 1H NMR (CDCl3, 200 MHz): δ 1.90 (s, 3 H), 2.03 (s,
6 H), 2.39 and 2.41 (2 s, 6 H), 2.57 (s, 3 H), 3.90 (s, 3 H), 6.77
(s, 1 H), 7.00 (s, 2 H), 7.04 (d, J ) 2.4 Hz, 1 H), 7.10 (dd, J )
8.7 Hz, J ) 2.4 Hz, 1 H) 7.49 (dd, J ) 8.9 Hz, J ) 1.7 Hz, 1
H), 7.58 (d, J ) 8.9 Hz, 1 H), 7.63 (d, J ) 8.7 Hz, 1 H), 7.90
(m, 1 H). 13C NMR (CDCl3, 50 MHz): δ 11.55, 17.21, 19.06,
20.42, 21.30, 55.32, 105.68, 116.18, 117.02, 118.93, 123.33,
123.88, 125.45, 126.21, 126.91, 127.12, 128.33, 128.50, 128.85,
(m) cm-1 1H NMR (CDCl3, 200 MHz,): δ ) 1.09 (s, 3 H), 1.42
.
(s, 3 H), 1.69 (s, 3 H), 1.79 (d, J ) 1.5 Hz, 3 H), 2.22 (s, 3 H),
2.51 (s, 3 H), 2.72 (s, 1 H), 3.50 (s, 3 H), 3.83 (s, 3 H), 5.53 (mc,
1 H), 5.95 (mc, 1 H), 6.52 (d, J ) 2.1 Hz, 1 H), 6.54 (s, 1 H),
6.70 (d, J ) 8.4 Hz, 1H), 6.87 (s, 1 H), 6.98 (dd, J ) 8.4 Hz, J
) 2.1 Hz, 1H). 13C NMR (CDCl3, 63 MHz): δ ) 17.69, 18.21,
18.84, 20.94, 21.15, 29.15, 55.38 (OCH3), 55.80 (OCH3), 89.50
(C-2), 109.93, 110.44, 110.93, 118.30, 119.09, 127.27, 127.73,
127.79, 128.15, 128.64, 130.00, 130.31, 130.97, 135.31, 136.58,
137.49, 143.80, 147.80, 148.74. HRMS: M+ (C28H32O4) calcd
432.2301, found 434.2292.
In vestiga tion s in th e P r esen ce of Acid . To X3-OH (7.7
mg, 18 µmol) dissolved in 3 mL of acetonitrile was added 1
mL of trifluroacetic acid. After 2 h stirring at room temper-
ature the reaction mixture was hydrolyzed by adding saturated
NaHCO3 (1 mL). After removing the solvent in vacuo the pure
benzofuran B3 (6.2 mg, 15 µmol, 84%) was afforded.
Ack n ow led gm en t. For generous financial support
we are indebted to the DFG (SFB 347 “Selective
reactions of metal activated molecules”) and the Fonds
der Chemischen Industrie. We thank the companies
Degussa for electrode materials and Kodak (Rochester,
NY) for a generous gift of tritolylamin.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR data and
peak assignments for E1, E3, E4, B1, B3, B4, and X3-OH
and parameters used for the digital simulation (3 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
J O980859N