Photosensitized (electron-transfer) deconjugation of 1-arylcyclohexenes

Article abstract of DOI:10.1021/ol006755s

(equation presented) A series of 1-arylcyclohexenes have been deconjugated to the corresponding 3-arylcyclohexenes via a photosensitized electron-transfer reaction. The introduction of substituents on the aryl group has provided insight into the underlyin

Full text of DOI:10.1021/ol006755s

ORGANIC
LETTERS
2
001
Vol. 3, No. 1
5-48
Photosensitized (Electron-Transfer)
Deconjugation of 1-Arylcyclohexenes
4
Dino Mangion, Jamie Kendall, and Donald R. Arnold*
Department of Chemistry, Dalhousie UniVersity,
Halifax, NoVa Scotia, Canada B3H 4J3
arnold@ac.dal.ca
Received October 19, 2000
ABSTRACT
A series of 1-arylcyclohexenes have been deconjugated to the corresponding 3-arylcyclohexenes via a photosensitized electron-transfer reaction.
The introduction of substituents on the aryl group has provided insight into the underlying mechanism and has defined the scope and
limitations of the reaction.
Earlier studies from our laboratory demonstrate that it is
possible to deconjugate 1-phenyl- and 1,1-diphenylalkenes
by means of a photochemically induced electron-transfer
reaction.1 The method exhibits promising synthetic potential,
resulting in high yields of the 3-phenyl- and 3,3-diphenyl-
alkenes from the readily accessible conjugated tautomers.
For example, a mixture of 1,4-dicyanobenzene (1), biphenyl
redox potentials of the electron-transfer couple and the singlet
excited-state energy of the light absorber, as defined by the
Weller equation (steps 1-3).
(b) The alkene radical cation deprotonates. This step
depends on the acidity of the radical cation and can be
promoted by the addition of a nonnucleophilic base, such as
collidine (3) (step 4).
4
,2
(
2), collidine (2,4,6-trimethylpyridine, 3) and 2-methyl-1-
phenylpropene (4a) in acetonitrile, irradiated by means of a
kW medium-pressure mercury lamp, yielded 2-methyl-3-
(
c) The resulting allylic radical is reduced to the corre-
sponding anion via electron transfer from the acceptor radical
anion. At this stage, the reduction potentials of the allylic
radical and of the cyanoarene play an important role (step
1
2
phenylpropene (4b) in 90% yield.
A photoinduced electron-transfer mechanism from the
electron donor (alkene) to the electron acceptor (cyanoarene)
excited singlet state was proposed (Scheme 1). This mech-
anism includes several key steps:
5
).
(
d) The allylic anion is reprotonated at either terminus to
regenerate the starting material 4a or yield the deconjugated
tautomer 4b (step 6).
(a) The cyanoarene absorbs light, resulting in its excitation
The aims of the present study are two-fold: we are
interested in defining the scope and limitations of this
deconjugation reaction for synthetic purposes, and we also
want to elucidate the mechanism proposed in our earlier
work. The introduction of substituents on the phenyl ring of
the alkene can serve as a useful probe for the various steps
in the mechanism, which are expected to respond differently
to the first excited singlet state. This is followed by electron
transfer from the alkene to the cyanoarene excited state. The
yield of radical ions can be enhanced by the addition of a
3
co-donor such as biphenyl (2). This process relies on the
(1) Arnold, D. R.; Mines, S. A. Can. J. Chem. 1987, 65, 2312.
(2) Arnold, D. R.; Mines, S. A. Can. J. Chem. 1989, 67, 689.
(3) (a) Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem. Soc.
1
990, 112, 4290. (b) Lewis, F. D.; Bedell, A. M.; Dykstra, R. E.; Elbert, J.
E.; Gould, I. R.; Farid, S. J. Am. Chem. Soc. 1990, 112, 8055.
(4) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 59.
1
0.1021/ol006755s CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/07/2000

Scheme 1
Table 1. Photochemical (Electron-Transfer) Deconjugation of
1
-Arylcyclohexenes 5a-10aa
starting alkene
(% consumption)b
irradiation
time (h)
product alkene
(% yield)b
entry
1
2
3
4
5
6
7
8
5a (86)
6a (87)
7a (88)
8a (91)
8a (89)c
8a (57)d
9a (85)
10a (80)
2.5
2.5
2.5
2.5
2.5
2.5
5.0
5.0
5b (26), 5c (40)
6b (44)
7b (61)
8b (72)
8b (48)
8b (17)
9b (54)
10b (16)
a
Reaction mixture composition: alkene 5a-10a (0.05 M), 1 (0.02 M),
2
(0.05 M), 3 (0.05 M) in acetonitrile. Reaction conditions: 1 kW medium-
b
pressure mercury lamp, 10 °C, Pyrex filter. Consumption and yield
percentages were determined by calibrated GC-FID using 10 mL reaction
volumes in 1 cm i.d. Pyrex tubes. Products were isolated from large scale
reactions under identical conditions. c No biphenyl (2) in this reaction. No
biphenyl (2) or collidine (3) in this reaction.
d
by a competing side reaction that involves reaction with the
alkenes. This alternative mode of reactivity becomes the
6
dominant pathway in the absence of base. It is commonly
observed in the photochemistry of aliphatic alkenes and
7
to the substituents. For this purpose we investigated the
cyanoarenes in nonnucleophilic, polar media. This compet-
photochemical deconjugation of six 1-arylcyclohexenes, 5a-
ing reaction was largely responsible for the nonquantitative
product yields.
10a, bearing different substituents ranging from methoxy to
cyano, in the 4-position of the aryl group (Scheme 2).
1,4-Dicyanobenzene (1) exhibits two absorption maxima,
at 281 nm (ꢀ ) 1650 dm mol cm ) and 290 nm (ꢀ )
650). The other reaction constituents generally do not absorb
3
-1
-1
1
appreciably beyond the 280 nm absorption cutoff of Pyrex,
thus rendering 1 the major light absorber. There are some
exceptions: alkenes 9a and 10a do absorb competitively,
which might be partly responsible for their slower reactivity.
Electronic excitation of 1 enhances its oxidizing properties
such that the first excited state will undergo electron transfer
Scheme 2
with any substrate having an oxidation potential less than
8
2
.4 V. As shown in Table 2, photoinduced electron transfer
is predicted to be diffusion-controlled with almost all the
conjugated alkenes as well as with some of the deconjugated
isomers.
The introduction of biphenyl (2) as co-donor showed a
marked enhancement in all of the reactions investigated (see
Experimental details and product yields are shown in Table
1. Products 5b-10b were isolated on silica gel by flash
chromatography and/or preparative, centrifugally accelerated,
radial, thin-layer chromatography using a Chromatotron and
entries 4 and 5 in Table 1 as examples). Biphenyl (2) has a
ox
half-wave oxidation potential, E1/2 , of 1.85 V vs SCE (CH
3
-
5
were fully characterized by spectroscopic methods. Typi-
9
CN). This implies a free energy for photoinduced electron
cally, the co-donor 2 and the base 3 were quantitatively
recovered. The acceptor 1 was generally partially consumed
-
1
transfer, ∆GPET, of -75 kJ mol for step 3a, Scheme 1.
NaCl) ν (cm-1) 831 (s), 1176 (w), 1299 (w), 1326 (w), 1413 (w), 1446
1
(
5) 9b: oil; H NMR (250.13 MHz, CDCl3, TMS) δ 1.45-1.75 (m, 3H),
(w), 1503 (m), 1606 (s), 2227 (s), 2859 (m), 2931 (s), 3021 (w); MS (EI)
+
•
1
(
2
.98-2.19 (m, 3H), 3.40-3.50 (m, 1H), 5.66-5.70 (m, 1H), 5.90-5.98
m/z 183 (M , 100%), 168 (82), 154 (89), 140 (58), 129 (55), 115 (86);
1
3
+•
m, 1H), 7.30-7.60 (m, 4H, AA′BB′); C NMR (62.9 MHz, CDCl3) δ
M
, 183.1058, C13H13N requires M, 183.1048.
1
1.0 (t), 24.9 (t), 32.4 (t), 41.7 (d), 124.4 (s, q, JC-F ) 272 Hz), 125.2 (d,
(6) (a) Kojima, M.; Kakehi, A.; Ishida, A.; Takamuku, S. J. Am. Chem.
Soc. 1996, 118, 2612. (b) Kojima, M.; Ishida, A.; Kuriyama, Y.; Oishi, S.
Bull. Chem. Soc. Jpn. 2000, 73, 1557.
(7) (a) Arnold, D. R.; Wong, P. C.; Maroulis, A. J.; Cameron, T. S. Pure
Appl. Chem. 1980, 52, 2609. (b) Borg, R. M.; Arnold, D. R.; Cameron, T.
S. Can. J. Chem. 1984, 62, 1785.
3
2
q, JC-F ) 4 Hz), 128.0 (d), 128.6 (s, q, JC-F ) 33 Hz), 129.1 (d), 129.2
-
1
(
(
(
d), 150.7 (s); IR (film, NaCl) ν (cm ) 836 (m), 1018 (w), 1069 (s), 1124
s), 1164 (m), 1326 (s), 1418 (w), 1619 (w), 2835 (w), 2865 (w), 2934
+
•
+•
m), 3026 (w); MS (EI) m/z 226 (M , 24%), 211 (19), 129 (100); M
26.0991, C13H13F3 requires M, 266.0969. 10b: oil; H NMR (250.13 MHz,
,
1
2
CDCl3, TMS) δ 1.44-1.53 (m, 3H), 2.07-2.13 (m, 3H), 3.43-3.50 (m,
(8) It has been estimated that a photoinduced electron-transfer process
1
H), 5.63-5.67 (m, 1H), 5.92-6.00 (m, 1H), 7.30-7.60 (m, 4H, AA′BB′);
C NMR (62.9 MHz, CDCl3) δ 20.8 (t), 24.8 (t), 32.2 (t), 41.9 (d), 109.7
s), 119.1 (s), 128.4 (d), 128.5 (d), 129.6 (d), 132.1 (d), 152.2 (s); IR (film,
will proceed at a diffusion-controlled rate when the free energy is less than
13
-1 4
-20 kJ mol
.
(
(9) Arnold, D. R.; Du, X. Can. J. Chem. 1994, 72, 403.
46
Org. Lett., Vol. 3, No. 1, 2001

Table 2. Free Energies for the Photoinduced Electron Transfer between 1 and Alkenes 5-10 (∆GPET) and for the (Charge-Shift)
+
•
Electron Transfer between 2 and 5-10 (∆GET
)
E1/2ox
(V)a
∆GPET
(kJ mol-1)b
∆GET
(kJ mol-1)c
E1/2ox
(V)a
∆GPET
(kJ mol-1)b
∆GET
(kJ mol-1)c
alkene
alkene
5
6
7
8
9
a
a
a
a
a
1.55
1.87
1.87
2.00
2.40
2.42d
-104
-73
-73
-60
-22
-20
-29
+2
+2
+14
+53
+53
5b
6b
7b
8b
9b
1.70
2.08
2.40
2.44
2.54
2.44
-89
-52
-22
-18
-8
-14
+22
+53
+56
+67
+57
1
0a
10b
-18
a
Oxidation potentials were measured by cyclic voltammetry in a three-electrode electrochemical cell equipped with a spherical platinum working electrode,
a platinum coil counter electrode, and a saturated calomel reference electrode (SCE). Deoxygenated acetonitrile was used as solvent and 0.1 M
-
1
tetraethylammonium perchlorate was the supporting electrolyte. The voltammograms were acquired at a scan rate of 100 mV s and calibrated against the
ox
ferrocene/ferrocenium oxidation potential of 0.51 V. In all cases, the electron transfer was irreversible and E1/2 was taken as 0.03 V before the peak
9
b
4
ox
red
red
potential.
)
Determined by means of the Weller equation: ∆GPET ) F(E1/2 - E1/2 - e/ꢀR) - E0,0 where E1/2 (1) ) -1.66 V vs SCE (CH3CN), E0,0(1)
-
1
-1
7b
c
ox
red
408 kJ mol and the Coulombic attraction term Fe/ꢀR is estimated as 5.4 kJ mol in acetonitrile.
Determined using: ∆GET ) F(E1/2 - E1/2
)
red
+•
9
d
ox
ox
+
where E1/2 (2 ) ) 1.85 V vs SCE (CH3CN).
Could not be determined experimentally. E1/2 (10a) was estimated from a Hammett plot of E1/2 vs σ
+
2
ox
(F
) -10.4, R ) 0.995) using our experimentally determined E1/2 values.
ox
Although most of the conjugated alkenes have a E1/2 that
is higher than that of 2, suggesting an endergonic electron
transfer in step 3b, Scheme 1 (Table 2), an enhanced yield
of radical ions is still observed. This phenomenon is
derivative the reaction is slow and the selectivity for the
formation of the nonconjugated isomer is poor (entry 8, Table
1). Here, both tautomers 10a and 10b have exceedingly high
and very similar oxidation potentials. It has been argued that,
should the radical cation of the deconjugated tautomer form,
it would undergo rapid electron transfer from the ground-
state conjugated alkene. This would occur even in the
absence of 2, provided that the oxidation potential of the
conjugated isomer is sufficiently lower than that of its
9
,10
commonly noticed in co-donor mediated reactions.
The
underlying mechanism is not fully understood. The formation
of alkene radical cations, despite an endergonic electron-
transfer process, is generally attributed to the existence of
an equilibrium between 2 and the alkene radical cation.
Rapid consumption of the alkene radical cation drives the
reaction forward. Alternatively, this may be viewed as the
formation of a complex between 2 and the alkene. This
complex imparts substantial positive charge onto the alkene.
It should be kept in mind that the second electron-transfer
process (steps 3b and 3c, Scheme 1) is a charge-shift electron
transfer that may behave differently from the more common
and better understood charge-separation process (step 3a,
Scheme 1). The overall enhancement in the co-donor
mediated reaction is a consequence of the slower rate of back
+
•
2
tautomer. In fact, in the current study, an enrichment of the
nonconjugated tautomer is generally observed in the absence
of 2, although the selectivity is lower than in the co-donor
mediated reaction (entry 5, Table 1).
+
•
Despite a highly exergonic electron transfer, the methoxy
derivative 5a does not undergo an efficient deconjugation
reaction (entry 1, Table 1). The radical cation of 5a is
expected to exhibit the lowest thermodynamic acidity among
the alkenes studied. It has been well established that the
radical cations of 1-arylalkene systems bearing strongly
-
•
+•
electron transfer (BET) for the 1 /2 pair as opposed to
that of the 1 /alkene radical cation pair. This is attributed
electron-donating groups are relatively inert toward any type
-
•
3
6b,11,12
of reaction that involves the cationic site.
The structur-
+•
to a smaller solvent reorganization energy for 2 going back
to 2.
One of the crucial steps in the reaction is the selective
consumption of the conjugated alkene over its nonconjugated
tautomer. Clearly, free energy considerations in the initial
photoinduced electron transfer between 1 and the alkene (step
ally similar 1-phenylpropene radical cation was reported to
2
+•
have a pK
a
13
of -12; the pK
a
of 5a can thus be estimated
as ca. -4, which, however, is still acidic enough to be
deprotonated by collidine (pK ) 16.8).
a
The acidity of the alkene radical ions increases with
oxidation potential, with the cyano derivative 10a having
2
, Scheme 1) cannot account for the selectivity. In most
an estimated pK of approximately -22. The introduction
a
cases, the electron transfer is expected to be diffusion-
controlled for both tautomers. In the biphenyl-mediated
mechanism, however, the situation is different. The second
electron transfer (steps 3b and 3c, Scheme 1) is not diffusion-
controlled. It will thus be sensitive to the redox potentials
of the constituents, favoring electron transfer from the alkene
of lower oxidation potential. In the extreme case of the cyano
of collidine (3) is crucial for a successful deconjugation
reaction, even though most of the alkene radical ions in this
study are acidic enough to be deprotonated by acetonitrile
(pK
a
) ca. -11). However, in the absence of 3, the
deconjugation reaction is slow and other reactions dominate
(11) (a) Mattes, S. L.; Farid, S. J. Am. Chem. Soc. 1986, 108, 7356. (b)
Tojo, S.; Morishima, K.; Ishida, A.; Majima, T.; Tokamuku, S. J. Org.
Chem. 1995, 60, 4684. (c) Arnold, D. R.; Du, X.; Chen, J. Can. J. Chem.
1995, 73, 307.
(12) Arnold, D. R.; Du, X.; Henseleit, K. M. Can. J. Chem. 1991, 69,
839.
(
10) (a) Tamai, T.; Mizuno, K.; Hashida, I.; Otsuji, Y. J. Org. Chem.
1
992, 57, 5338. (b) Yamashita, T.; Yasuda, M.; Isami, T.; Nakano, S.;
Tanabe, K.; Shima, K. Tetrahedron Lett. 1993, 34, 5131. (c) Mangion, D.;
Arnold, D. R.; Cameron, T. S.; Robertson, K. N. J. Chem. Soc., Perkin
Trans. 2, accepted for publication.
(13) Nicholas, A. M. P.; Arnold, D. R. Can. J. Chem. 1982, 60, 2165.
Org. Lett., Vol. 3, No. 1, 2001
47

anion 1-•.16 Therefore, the influence of this step on the
reaction outcome is marginal when the substituents are
electron-withdrawing or weakly electron-donating. Conse-
quently, the formation of adducts similar to 5c is less
favorable.
Scheme 3
In summary, the photochemical electron-transfer decon-
jugation of conjugated aromatic alkenes has been extended
to include 1-arylcyclohexenes with a range of substituents
on the aromatic ring. This reaction constitutes a simple and
straightforward synthetic method for converting the readily
available conjugated alkenes to the corresponding 3-aryl-
cyclohexenes that are not as easily accessible. The observed
substituent effects on the reaction follow the predictions made
on the basis of the mechanism proposed. For a successful
reaction outcome there are several key steps that have to be
satisfied: formation of radical ions, deprotonation of the
alkene radical cation, reduction of the allylic radical, and
selective consumption of the conjugated alkene relative to
the nonconjugated isomer. The introduction of strongly
electron-donating substituents (e.g., methoxy) inhibits the
reaction as the result of an inefficient reduction of the alkene-
derived allylic radical to the corresponding anion. On the
other hand, strongly electron-withdrawing substituents (e.g.,
cyano) hinder the initial formation of radical ions and reduce
the selectivity toward the formation of the deconjugated
tautomer because of similar oxidation potentials for both
conjugated and nonconjugated isomers.
6
(
entries 5 and 6, Table 1). The optimal collidine (3)
concentration was found to be ca. 0.05 M. Higher concentra-
tions led to an increasingly inefficient reaction. We have
evidence that this is most likely due to quenching by 3 of
the 1,4-dicyanobenzene excited state, probably via an
electron-transfer mechanism.
The next step in the proposed mechanism, that is, the
reduction of the allylic radical to the corresponding anion
-
•
by 1 , is more likely to be the reason for the poor yield of
deconjugation product 5b. The reduction potential is expected
to be too high for the allylic radical derived from 5a as a
result of the electron-donating effect of the methoxy group
destabilizing the anion.14 The inability of the allylic radical
derived from 5a to be reduced is confirmed by the isolation
of 3-(4-cyanophenyl)-1-(4-methoxyphenyl)cyclohexene 5c in
good yield. This product is thought to arise via coupling of
the allylic radical with the radical anion of 1, followed by
expulsion of cyanide anion as driven by rearomatization
Acknowledgment. This work was supported financially
by the National Sciences and Engineering Research Council
of Canada. D. Mangion is grateful for a scholarship from
the Izaak Walton Killam Memorial Foundation.
(
Scheme 3). The identity of product 5c has been firmly
established by spectroscopic techniques.
Electron-withdrawing groups raise the reduction potential
of the radical, promoting reduction by the sensitizer radical
15
OL006755S
(
14) E1/2red for the 4-methoxybenzyl radical is -1.82 V vs SCE, which
-1 -•
implies a free energy of +15 kJ mol for its reduction by 1 . In
(s), 113.7 (d), 119.1 (s), 124.1 (d), 126.2 (d), 128.6 (d), 132.2 (d), 134.3
red
-1
comparison, E1/2 for the benzyl radical is -1.43V vs SCE, giving a free
(s), 138.3 (s), 152.4 (s), 158.9 (s); IR (film, NaCl) ν (cm ) 824 (m), 1036
-1
-• 16
energy of -22 kJ mol for its reduction by 1
.
(m), 1181 (m), 1249 (s), 1513 (s), 1607 (m), 2227 (m), 2832 (w), 2932
1
+•
+•
(
15) 5c: oil; H NMR (250.13 MHz, CDCl3, TMS) δ 1.52-2.51 (m,
(m), 3036 (w); MS (EI) m/z 289 (M , 100%), 261 (46), 121 (43); M
,
4
1
(
H), 2.47-2.51 (m, 2H), 3.59-3.66 (m, 1H), 3.81 (s, 3H), 5.99 (br. s,
289.1445, C20H19NO requires M, 289.1467.
(16) Sim, B. A.; Milne, P. H.; Griller, D.; Wayner, D. D. M. J. Am.
Chem. Soc. 1990, 112, 6635.
13
H), 6.86-7.39 (m, 4H, AA′BB′), 7.34-7.64 (m, 4H, AA′BB′); C NMR
62.9 MHz, CDCl3) δ 21.5 (t), 27.3 (t), 32.0 (t), 42.7 (d), 55.3 (q), 109.9
48
Org. Lett., Vol. 3, No. 1, 2001