Aryl Cations from Aromatic Halides. Photogeneration and Reactivity of 4-Hydroxy(methoxy)phenyl Cation

Article abstract of DOI:10.1021/jo049770+

The photochemistry of 4-chlorophenol (1) and 4-chloroanisole (2) has been examined in a range of solvents and found to lead mainly to reductive dehalogenation, through a homolytic path in cyclohexane and a heterolytic path in alcohols. Heterolysis of 1 and 2 in methanol and 2,2,2-trifluoroethanol offers a convenient access to triplet 4-hydroxy- and 4-methoxyphenyl cations. These add to π nucleophiles, viz., 2,3-dimethyl-2-butene, cyclohexene, and benzene, giving the arylated products in medium to good yields. Wagner-Meerwein hydride and alkyl migration are evidence for the cationic mechanism of the addition to alkenes. Arylation (with no rearrangement) was obtained to some extent also in nonprotic polar solvents such as MeCN and ethyl acetate, reasonably via an exciplex and with efficiency proportional to the nucleophilicity of the trap (2,3-dimethyl-2-butene > cyclohexene ? benzene).

Full text of DOI:10.1021/jo049770+

Ar yl Ca tion s fr om Ar om a tic Ha lid es. P h otogen er a tion a n d
Rea ctivity of 4-Hyd r oxy(m eth oxy)p h en yl Ca tion
Stefano Protti, Maurizio Fagnoni,* Mariella Mella, and Angelo Albini*
Dipartimento di Chimica Organica, Universita`, V. Taramelli 10, 27100 Pavia, Italy
fagnoni@unipv.it
Received February 10, 2004
The photochemistry of 4-chlorophenol (1) and 4-chloroanisole (2) has been examined in a range of
solvents and found to lead mainly to reductive dehalogenation, through a homolytic path in
cyclohexane and a heterolytic path in alcohols. Heterolysis of 1 and 2 in methanol and
2,2,2-trifluoroethanol offers a convenient access to triplet 4-hydroxy- and 4-methoxyphenyl cations.
These add to π nucleophiles, viz., 2,3-dimethyl-2-butene, cyclohexene, and benzene, giving the
arylated products in medium to good yields. Wagner-Meerwein hydride and alkyl migration are
evidence for the cationic mechanism of the addition to alkenes. Arylation (with no rearrangement)
was obtained to some extent also in nonprotic polar solvents such as MeCN and ethyl acetate,
reasonably via an exciplex and with efficiency proportional to the nucleophilicity of the trap (2,3-
dimethyl-2-butene > cyclohexene . benzene).
In tr od u ction
To widen the scope of the reaction, we considered
chlorophenols and chloroanisoles as further potential
phenyl cation precursors meeting the above require-
ments. Differently from chloroanilines, the photochem-
istry of these substrates has been previously investigated
in detail. Chlorophenols are largely used both in agri-
culture and in industry as antimicrobials, and their
release in the environment has required investigations
on their photostability, also with regard to the possible
photoformation of the highly toxic polychlorodioxins from
polyhalophenols.5 As a consequence, most of these studies
were carried out in water rather than in organic media.
In the case of 4-chlorophenol efficient dehalogenation was
reported to occur via two paths. Earlier studies in
aqueous alkali6 and in MeOH7 were interpreted as a
homolytic fission. More recently, the photoreactivity of
p-chlorophenol in water was explained by invoking a
carbene, 4-oxocyclohexa-2,5-dienylidene, as the key in-
termediate for the above reactions (see Scheme 1, path
a).8 This species could arise through deprotonation of
4-hydroxyphenyl cation, but no reaction was attributed
to the latter intermediate (path b).
The direct formation of an aryl-alkyl or an aryl-aryl
bond has received much attention in the past years in
view of its application to the synthesis of natural products
and organic conducting materials.1 This goal has been
achieved generally starting from aromatic halides by
means of organometal-mediated reactions. Among these,
palladium-catalyzed coupling is the most studied reaction
and has been applied in industrial processes for fine
chemical synthesis.1d Recently we have developed an
alternative photochemical path for the formation of an
aryl-carbon bond using aromatic monohalides (chlorides
or fluorides) as precursors. The method is based upon the
generation of aryl cations that readily form new C-C
bonds. The reaction occurs efficiently with 4-chloro- and
4-fluoroanilines and the corresponding N,N-dimethyl
derivatives.2 The aryl cation adds to π nucleophiles such
as alkenes,2a aromatics,2a and heteroaromatics.2c It ap-
pears that a photochemical SN1 path via aryl cation is
viable for aryl-carbon bond formation, provided that
some conditions are met: (a) The aromatic ring bears a
strong electron-donating substituent in order to make
photoheterolysis of the carbon-halogen bond permissible.
(b) ISC is efficient and the fragmentation occurs from the
triplet in order to generate the phenyl cation likewise as
a triplet.3 (c) A polar solvent such as MeCN, MeOH, or
2,2,2-trifluoroethanol (TFE) is used.
(3) Differently from the triplet, singlet aryl cations behave as
unselective electrophiles and reaction with the solvent is usually
observed also in the presence of π nucleophiles. See ref 4
(4) Milanesi, S.; Fagnoni, M.; Albini A. Chem. Commun. 2003, 216-
217.
(5) Skurlatov, Y. I. Ernestova, L. S.; Vichutinskaya, E. V.; Samsonov,
D. P.; Semenova, I. V.; Rod’ko, I. Y.; Shvidky, V. O.; Pervunina, R. I.;
Kemp, T. J .J . Photochem. Photobiol. A 1997, 107, 207-213.
(6) Omura, K.; Matsuura, T. Tetrahedron 1971, 27, 3101-3109.
(7) Pinhey, J . T.; Rigby, R. D. G. Tetrahedron Lett. 1969, 16, 1267-
1270.
(8) (a) Grabner, Richard, C.; Ko¨hler, G. J . Am. Chem. Soc. 1994,
116, 6, 11470-11480. (b) Durand, A. P.; Brown, R. G.; Worrall, D.;
Wilkinson, F. J . Chem. Soc., Perkin Trans. 2 1998, 365-370. (c)
Bonnichon, F.; Grabner, G.; Guyot, G.; Richard, C. J . Chem. Soc.,
Perkin Trans. 2 1999, 1203-1210. (d) Arnold, B. R.; Scaiano, J . C.;
Bucher, G. F.; Sander, W. J . Org. Chem. 1992, 57, 6469-6474.
(1) (a) Culkin, D. A.; Hartwig, J . F. Acc. Chem. Res. 2003, 36, 234-
245. (b) Prim, D.; Campagne, J .-M.; J oseph, D.; Andrioleti, B. Tetra-
hedron 2002, 58, 2041-2075. (c) Hassan, J .; Se´vignon, M.; Gozzi, C.;
Schulz, E.; Lemaire, M. Chem. Rev. 2002, 102, 1359-1469. (d) Beller,
M.; Zapf, A.; Maegerlein, W. Chem. Eng. Technol. 2001, 24, 575-582.
(2) (a) Mella, M.; Coppo, P.; Guizzardi, B.; Fagnoni, M.; Freccero,
M.; Albini, A. J . Org. Chem. 2001, 66, 6344-6352. (b) Guizzardi, B.;
Mella, M.; Fagnoni, M.; Freccero, M.; Albini, A. J . Org. Chem. 2001,
66, 6353-6363. (c) Guizzardi, B.; Mella, M.; Fagnoni, M.; Albini, A.
Tetrahedron 2000, 56, 9383-9389.
10.1021/jo049770+ CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/21/2004
J . Org. Chem. 2004, 69, 3465-3473
3465

Protti et al.
TABLE 1. Ir r a d ia tion of 4-Ch lor op h en ol (1) in Nea t
Solven t
SCHEME 1
1 (0.05 M)
1 + Cs2CO3 (0.025 M)
consumptiona products consumptiona products
solvent
(%)
(% yielda)
(%)
(% yielda)
cyclohexane
ethyl acetate
acetonitrile
methanol
TFE
38
11
23
100
52
3, 23
3, 9
3, <1
3, 96
3, 8
30
51
24
100
80
3, 30
3, 30
3, 23
3, 94
3, 56 (66)
4, 5 (6)
In the case of 4-chloroanisole, irradiation led to two
processes, photoreduction to yield anisole and photosub-
stitution of the chlorine atom to give aryl ethers in
alcohols. Several mechanistic hypotheses have been
presented. For example, Pinhey supposed that the first
process involved photohomolysis and the latter resulted
from the attack by the solvent onto an excited state of
anisole.9 Photosubstitution was also explained by the
attack of the solvent onto the radical cation of chloro-
anisole resulting from photostimulated emission of an
electron.10 On the contrary, the radical anion of 4-chlo-
roanisole was invoked by Siegman et al. as the interme-
diate in the formation of aryl ethers. The radical anion
resulted from photoinduced electron transfer from the
solvent (MeOH).11 In aqueous media or MeCN/water
solutions the formation of an excimer was proposed, and
both the radical cation and the radical anion of the
haloanisole could be formed.12 In addition, the presence
of strong electron-donating alkenes such as 1,1-diphen-
ylethylene in the reaction medium caused assisted ho-
molytic photodehalogenation.13
4, tr
a
The consumption and the products yield after 14 h of irradia-
tion were determined by GC. In parentheses are the isolated yields
after total consumption of the starting halide.
SCHEME 2
Thus, the matter has not been fully clarified, and we
deemed it worthwhile to reconsider the photochemistry
of 4-chlorophenol (1) and 4-chloroanisole (2) in different
solvents and in the presence of π nucleophiles to assess
whether the appealing triplet aryl cation chemistry could
be obtained also from these compounds, even if there was
no indication of it in the literature.
TABLE 2. Ir r a d ia tion of 4-ch lor oa n isole (2) in Nea t
Solven t
2 (0.05 M)
2 + Cs2CO3 (0.025 M)
consumptiona products consumptiona products
solvent
(%)
(% yielda)
(%)
(% yielda)
Resu lts
cyclohexane
ethyl acetate
acetonitrile
methanol
88
11
5
5, 80
5, 4
5, 3
5, 30
6, tr
7, 14
5, 9
6, 23
8, 10
93
9
17
60
5, 87
5, 5
5, 2
5, 42
6, tr
7, 17
5, 21 (32)
6, 13 (18)
8, 10 (12)
Ir r a d ia tion in Nea t Solven t The photoreactivity of
1 and 2 was first examined in 0.05 M solutions in five
solvents, viz., cyclohexane, ethyl acetate, MeCN, metha-
nol, and TFE by irradiation at 310 nm (see Table 1 and
Scheme 2). Two sets of experiments were carried out, the
first by irradiating for a fixed time (14 h) and the latter
by irradiating up to total consumption of the starting
halide for preparative experiments. In some cases it was
found that the material balance was unsatisfactory.
Apparently, this was due to thermal reactions of the
phenols induced by hydrochloric acid liberated in the
photoreaction, since the balance was better when the
experiments were carried out in the presence of an
equimolar amount of cesium carbonate as a buffering
agent.14 The process occurring was reductive photo-
dechlorination yielding phenol 3, though with different
47
TFE
48
66
a
Conditions, see Table 1.
efficiency (maximal in MeOH). A low yield of a further
photoproduct was isolated in TFE, viz., 2,4′-hydroxy-5-
chlorobiphenyl 4, whereas bicyclohexyl was found in
cyclohexane.
Irradiation of 2 caused likewise an efficient photore-
duction leading to anisole (5). The fastest conversion was
in this case obtained in cyclohexane (where some bicy-
clohexyl was formed), whereas the reaction in polar
aprotic solvents was slow (Table 2 and Scheme 2). 2,4′-
Dimethoxy-5-chlorobiphenyl (6, the analogue of 4 from
1) was formed in a modest yield in TFE (20%) and only
in traces in MeOH. Differently from the case of 1, small
(9) Pinhey, J . T.; Rigby, R. D. G. Tetrahedron Lett. 1969, 16, 1271-
1274.
(10) Heijer, J . D.; Shadid, O. B.; Cornelisse J .; Havinga, E. Tetra-
hedron 1977, 33, 779-786.
(11) Siegman, J . R.; Houser, J . J . J . Org. Chem. 1982, 47, 2773-
2779.
(12) Lemmetynen, H.; Konijnenberg, J .; Cornelisse, J .; Varma, C.
A. G. O. J . Photochem. 1985, 30, 315-338.
(13) Mangion, D.; Arnold, D. R. Can. J . Chem. 1999, 77, 1655-1670.
(14) The salt is completely soluble in MeOH or TFE but slightly
soluble in the other solvents.
3466 J . Org. Chem., Vol. 69, No. 10, 2004

Aryl Cations from Aromatic Halides
TABLE 3. Ir r a d ia tion of 1 a n d 2 in th e P r esen ce of 0.5
M 2,3-Dim eth yl-2-bu ten e
SCHEME 4
1 (0.05 M)
2 (0.05 M)
consumptiona products consumptiona products
solvent
(%)
(% yielda)
(%)
(% yielda)
cyclohexane
54b
3, 13
9, 10
84
5, 1 (2)
12, 28 (34)
13, 3 (4)
5, 4 (7)
12, 27 (54)
5, <1 (2)
12, 15 (49)
5, 18
ethyl acetate
acetonitrile
methanol
77
44
3, 15 (20)
9, 53 (69)
3, 8 (9)
9, 27 (67)
3, (61)
50
31
100
100
9, (22)
12, 21
13, 18
14, 10
methanol +
100
91
3, (45)
10, (26)
88
78
5, 20 (22)
12, 14 (16)
13, 15 (17)
14, 14 (17)
5, 16
Cs2CO3
TFE
3, 22
9, 15
11, 18
12, 10
13, tr
15, 7
16, 11
5, (12)
12, (21)
13, (14)
15, (17)
16, (8)
phenyl)-1-butene (9), obtained in a satisfactory yield (ca.
70%) in ethyl acetate or acetonitrile. In alcohols, the
results were somewhat different, both because of the
greater importance of reduction to 3 in MeOH and
because compound 9 was accompanied or substituted by
the corresponding â-alkoxyalkyl derivatives 10 (in MeOH,
only in the presence of Cs2CO3) and 11 (in TFE).
TFE +
100
3, (19)
9, (43)
11, (12)
100
Cs2CO3
a
b
Conditions, see Table 1. No further consumption after pro-
longed irradiation.
With 4-chloroanisole, reduction to 5 was generally less
important (<20%); addition of the alkene was quite
efficient but was varied with regard to the structure of
the final product (Table 3 and Scheme 4). Thus, aryl-
butene 12 was the exclusive or largely dominant product
in aprotic solvents, whereas it was accompanied by
arylbutane 13 and by the 3-alkoxy-2-arylbutanes 14 and
15 in alcohols. (Actually in TFE there was also a further
adduct, 2-alkoxy-2-arylbutane 16.) The presence of the
base increased somewhat the yield of the photoproducts.
Ir r a d ia tion of 1 a n d 2 in th e P r esen ce of Cyclo-
h exen e. The results of the irradiations in the presence
of 0.5 M cyclohexene are reported in Table 4 and Schemes
5 and 6. Reduction was again more important with
chlorophenol 1 (62% in MeOH, 20-30% with other
solvents) than with chloroanisole 2 (15-25%). Photolysis
of 1 in nonprotic solvents gave trans 1-(4-hydroxyphenyl)-
2-chlorocyclohexane (17) as the only adduct in moderate
yield. In MeOH a small fraction of adducts was obtained,
and because their separation was difficult, the raw
photolisate was treated with methyl iodide. Separation
then afforded three photoproducts, viz., 1-(4-methoxy-
phenyl)-1-methoxycyclohexane (18), 1-(1-(4-methoxyphen-
yl)-1-methoxy)methylcyclopentane (19), and trans 1-(4-
methoxyphenyl)-2-chlorocyclohexane (20) in a ca. 10%
overall yield. The reaction in TFE yielded again three
adducts, viz., the chlorocyclohexane 17 and the two ethers
1-(1-(4-hydroxyphenyl)-1-(2,2,2-trifluoroethoxy)methylcy-
clopentane (21) and 1-(p-hydroxyphenyl)-1-(2,2,2-trifluo-
roethoxy)cyclohexane (22). In TFE the yield of adducts
to cyclohexene considerably increased in the presence of
cesium carbonate.
SCHEME 3
amounts of aryl ethers 1,4-dimethoxybenzene (7, 17%)
and 4-(2,2,2-trifluoroethoxy)anisole (8, 10%) were ob-
tained from 2 in the two alcohols tested.
Ir r a d ia tion in th e P r esen ce of π Nu cleop h iles. We
next explored the reactions of 1 and 2 when irradiated
in the above solvents in the presence of π nucleophiles.
2,3-Dimethyl-2-butene, cyclohexene, and benzene were
chosen for this purpose.
Since the acidity liberated may affect the results (e.g.,
by inducing polymerization of the alkene), a second set
of experiments in the presence of equimolar Cs2CO3 was
carried out. The results of these experiments are reported
only when differing from what observed in the absence
of the salt. Also in this case, the results from the
irradiation for a fixed time (14 h) and at total consump-
tion are reported.
Ir r a d ia tion of 1 a n d 2 in th e P r esen ce of 2,3-
Dim eth yl-2-bu ten e. In the presence of 0.5 M 2,3-
dimethyl-2-butene, 4-chlorophenol was converted to a
somewhat larger degree than in neat solvents (see Table
3 and Schemes 3 and 4). The products were phenol (3)
and an alkylated product, 2,3-dimethyl-3-(4-hydroxy-
As for chloroanisole 2, arylchlorocyclohexane 20 was
formed in aprotic solvents. In alcohols the results were
more varied, and two ethers (18 and 19) were obtained
J . Org. Chem, Vol. 69, No. 10, 2004 3467

Protti et al.
TABLE 4. Ir r a d ia tion of 1 a n d 2 in th e P r esen ce of 0.5
M Cycloh exen e
SCHEME 6
1 (0.05 M)
2 (0.05 M)
consumptiona products consumptiona products
solvent
(%)
(% yielda)
(%)
(% yielda)
cyclohexane
48b
3, 21
17, tr
66
5, 11 (16)
20, 25 (38)
5, 8 (15)
20, 18 (32)
5, 5 (26)
20, 7 (20)
5, 17 (18)
18, 31 (32)
19, 11 (15)
ethyl acetate
acetonitrile
methanol
56
25
3, 18 (24)
17, 5 (20)
3, 7 (18)
17, tr (12)
5, (62)
50
12
83
100c
18, (tr)
19, (8)
20, (4)
TFE
79
3, 27 (30)
17, 6 (8)
21, 5 (6)
22, (4)
80
5, 9 (11)
20, 24 (29)
23, 37 (45)
TFE +
Cs2CO3
100
3, (44)
100
5, (10)
17, (12)
21, (13)
22, (tr)
20, (9)
24, (20)
25, (21)
TABLE 5. Ir r a d ia tion of 1 a n d 2 in th e P r esen ce of 1 M
Ben zen e
1 (0.05 M)
2 (0.05 M)
a
b
Conditions, see Table 1. No further conversion obtained by
prolonged irradiation. c The raw photolisate was methylated with
CH3I before separation by column chromatography (see Experi-
mental Section).
consumptiona products consumptiona products
solvent
(%)
(% yielda)
(%)
(% yielda)
cyclohexane
36
3, 21
85
5, 75
27, tr
5, 6
27, 8
5, 4
27, 4
5, 12
27, 5
5, <1 (6)
27, 20 (70)
ethyl acetate
acetonitrile
methanol
TFE
23
10
3, 9
26, tr
3, 7
26, tr
25
12
40
54
SCHEME 5
100
63
3, 88
26, 10
3, 11 (16)
26, 50 (66)
a
Conditions, see Table 1.
SCHEME 7
points worth noting are the increased efficiency in the
reduction of 2 in the presence of benzene in cyclohexane
and the fact that a large amount of additive was required
for affecting the photoreactions when using benzene
(1 M) than when using the previous alkenes (0.5 M).
in methanol, and in neat TFE chloro derivative 20 was
formed along with arylcyclohexane 23. On the other
hand, although the presence of Cs2CO3 did not change
the result in MeOH, in TFE this led to the formation of
two trifluoroethyl ethers, the 1-(4-methoxyphenyl)cyclo-
hexyl derivative 24 and the 1-(4-methoxyphenyl)-1-cy-
clopentylmethyl derivative 25, along with compound 20.
As in the reaction with 2,3-dimethyl-2-butene com-
pounds 7 and 8 were not detected in any of the irradia-
tions in the presence of alkenes.
Ir r a d ia tion of 1 a n d 2 in th e P r esen ce of Ben zen e.
The results of the irradiations in the presence of 1 M
benzene are shown in Table 5 and Scheme 7. With both
compounds 1 and 2, reduction to 3 (or 5) predominated,
particularly with the first one (88% yield of phenol in
MeOH). However, 4-phenylphenol (26) and 4-phenylani-
sole (27) were isolated in ca. 70% yield in TFE. Further
Discu ssion
The above results show that, as hoped, the photoreac-
tivity of 4-chlorophenol (1) and 4-chloroanisole (2) can
be exploited for the arylation of π nucleophiles such as
alkenes and aromatics, similarly to what was previously
observed with haloanilines.2 The pattern of reactivity
observed is quite varied, however, and depends on the
solvent chosen.
P h ot or ea ct ion of 4-Ch lor op h en ol. Intersystem-
crossing (ISC) is known to be close to unitary with
chlorophenol 1 both in nonpolar (n-hexane) and polar
solvents (water),8a and thus the reaction is attributed to
the triplet state. Laser flash photolysis studies allowed
detection of 13* only in nonpolar solvents, and in polar
3468 J . Org. Chem., Vol. 69, No. 10, 2004

Aryl Cations from Aromatic Halides
SCHEME 8
or protic solvents this transient became too short-lived
to be detected. Thus, the photoreaction in cyclohexane
can be attributed to homolytic fragmentation and hydro-
gen abstraction from the solvent by the aryl radical, as
supported by the formation of bicyclohexyl.
In ethyl acetate and acetonitrile, 4-chlorophenol is
markedly less photoreactive, reasonably because of the
shorter lifetime of the triplet in this case.15a The low
reactivity in acetonitrile (Φ ≈ 0.0028a) contrasts with the
rather efficient fragmentation of 4-chloroaniline in this
solvent (Φ ) 0.448b). Phenol 1, however, does give
arylation reactions in these solvents and indeed also in
cyclohexane. The efficiency of such process depends on
the nucleophilicity (or electron donor ability) of the trap
used, 2,3-dimethyl-2-butene > cyclohexene . benzene.
We suggest that this involves the formation of an exciplex
between 13* and the nucleophile as the reaction inter-
mediate (compare below the case of 4-chloroanisole).
In protic solvents, the triplet undergoes heterolytic
fragmentation8a and a triplet aryl cation is formed. This
gives phenol 3 as the main product. The process is
thought to involve hydrogen abstraction from the solvent
to yield a radical cation intermediate and subsequent
reduction by the solvent (or the nucleophile used) to
closed-shell, neutral 3. Alternatively, the aryl cation is
trapped by the starting phenol to yield a small amount
of dihydroxybiphenyl 4. Addition of a π nucleophile traps
the intermediate. Since aryl cations are very short-lived
species,2 a relatively high amount of the nucleophile (0.5
M alkenes, 1 M benzene) is required in order to make
this reaction efficient. Even with this large excess, in
MeOH 4-hydroxyphenyl cation is effectively reduced
through a different mechanism. This path involves
deprotonation (Scheme 1) and hydrogen abstraction by
the carbene, as demonstrated by Grabner,8a and remains
predominant (>60%) in this solvent, limiting the aryl-
carbon bond formation from 1. TFE stabilizes the cation
and is the solvent of choice for arylation reactions. In this
solvent, arylation consistently predominates with both
1 and 2 and, contrary to the case of aprotic solvents, does
not depend on the nucleophilicity of the trap used.
Arylation of benzene to yield 26 (Scheme 8) requires no
further comment. As for the addition onto alkenes, this
is envisaged as involving a phenonium ion as an inter-
mediate16 as illustrated in the reaction with 2,3-dimethyl-
2-butene in Scheme 8, analogously to the case of 4-ami-
nophenyl cation.17
SCHEME 9
deprotonation that increases (9 grows from 16% to 43%),
the yield of trifluoroethoxy ether 11 not changing ap-
preciably.
Essentially the same reactions are observed in the
presence of cyclohexene (Scheme 9). Arylation of the
alkene is here less efficient, and the evolution of the
intermediate phenonium ion in part different. Deprotona-
tion is not observed, apparently because of the unfavor-
able alignment of the C-H cyclopropane bond with the
positively charged π-system, and the expected 1,2-addi-
tion derivatives are formed only when chloride is the
nucleophile and then in a regioselective (trans) fashion
(Scheme 9, path a). This led to chlorinated 17, formed in
a higher amount in nonalcoholic solvents and absent in
MeOH, where reduction overcomes any other photopro-
cess.
In non-nucleophilic solvents, deprotonation of the
adduct cation affords allylphenol 9 in a satisfactory yield
(path a). In alcohols, the phenonium ion intermediate
undergoes competitive nucleophilic attack by the solvent,
giving the corresponding ethers 10 and 11 (path b).
Cesium carbonate has an effect here. In MeOH, addition
of this salt favors methoxide addition to form â-alkoxy-
derivative 10, whereas in TFE it is the base-induced
(15) (a) Laser flash photolysis experiments on 4-chlorophenol in
acetonitrile did not reveal any triplet signal. See ref 8a. (b) Evidence
for a similar mechanism in the reduction of the analoguous 4-ami-
nophenyl cation has been presented; see ref 2b. A study on the
mechanism of the photoreaction of 1 in alcoholic solvents is underway.
(16) We have no positive evidence for the phenonium structure of
the adduct cation, but this is calculated to be the lowest-lying adduct
with an alkene; see ref 2a.
Noteworthy, the most abundant products in TFE and
a significant proportion in MeOH contain a rearranged
alkyl chain. The processes occurring are classical car-
bocation rearrangement at the phenonium ion level.
(17) Guizzardi, B.; Mella, M.; Fagnoni, M.; Albini, A. J . Org. Chem.
2003, 68, 1067-1074.
J . Org. Chem, Vol. 69, No. 10, 2004 3469

Protti et al.
Wagner-Meerwein hydride shift (path b) and solvent
addition lead to benzyl ethers 18 and 22, while alkyl
rearrangement causes ring contraction to yield cyclopen-
tylmethyl ethers (path c). The occurrance of these rear-
rangements further supports the cationic course of the
reaction and is reminescent of the formation of cyclopen-
tylmethyl derivatives in the photoreaction of 4-halo-
anilines in the presence of cyclohexene and amines.17
P h otor ea ction of 4-Ch lor oa n isole. Most of the
above considerations on 4-chlorophenol can be extended
to 4-chloroanisole. The fact that the photoreaction is most
efficient in cyclohexane fits with previous report,18 and
the formation of anisole 5 and bicyclohexyl support that
homolysis from the triplet is involved. The triplet has
been identified as a relatively long-lived species (τ 900
ns) in perfluorohexane.12 The reaction is quite inefficient
in polar, aprotic solvents. Indeed, in acetonitrile a broad
band at 420-500 nm has been observed by laser flash
photolysis and attributed to the ion pair (2+•/ 2-•) from
the triplet state.12 This complex was reported to decay
unproductively to the ground state.12 In alcohols, photo-
heterolysis is the main process, giving 4-methoxyphenyl
cation. Obviously, whereas 4-hydroxyphenyl cation easily
deprotonates in these solvents, 4-methoxyphenyl cation
can undergo no such cleavage and the triplet carbene
pathway is precluded. Therefore, although reduction to
anisole remains efficient in MeOH, trapping has more
chance. Dimethoxybiphenyl 6 is formed in trace amount
in MeOH but in >20% yield in TFE. Furthermore, aryl
ethers 7 and 8 are obtained in both alcohols (10-17%).
Because these compounds are not formed in irradiations
with alkenes or benzene, it is reasonable that they arise
from the same intermediate. Triplet aryl cations are not
expect to add to σ nucleophiles. These species are not
localized cations and rather have a diradical nature, with
a carbene character at the divalent carbon. This is
supported by previous calculations and experimental
studies on the analoguous 4-aminophenyl cation, where
likewise no solvent addition takes place,2a,4 and well
explains both reduction by hydrogen-donating solvents
and addition to nucleophiles observed also in the present
case. Unlike the triplet, singlet aryl cations add efficiently
even to relatively poor σ nucleophiles, e.g., to acetoni-
trile.3,4 Thus, formation of the ethers may involve ISC
from the initially formed triplet to the low-lying singlet
cation in solution (for an analoguous proposal, see ref 4).19
Paths not involving photoheterolysis, such as electron
transfer from the solvent11 or photostimulated ionization
of the aromatic compound (important in water),10,12 have
been suggested in related cases but appear unlikely in
the present reaction in alcohols. Likewise, alkene-as-
sisted photodehalogenation has been invoked by Arnold
for the reaction of some aryl halides with arylolefins13
but does not apply to weak donors such as the present
aliphatic traps.
12 is more important under all conditions (33% in
cyclohexane). A specific path is formation of some (<20%)
arylbutane 13 under some conditions. This may result
from the stabilization through intramolecular electron
transfer in the phenonium ion (path d, Scheme 8), leading
to a diradical structure in the intermediate and facilitat-
ing hydrogen abstraction. A similar mechanism has been
invoked for the reaction of the 4-N,N-dimethylamino-
phenyl cation with norbornadiene.20 In alcohols, methyl
and trifluoroethyl ethers were formed, including Wag-
ner-Meerwein rearranged product 16 (path c, Scheme
8), more abundant in TFE than unrearranged 15.
With cyclohexene, trans 1-aryl-2-chlorocyclohexane 20
is the main product in nonprotic solvent, whether polar
or apolar solvents. In MeOH, rearranged ethers (both H
and alkyl shift, paths b, c, Scheme 9) were formed in a
better yield than from phenol 1, as a result of the lower
proportion of photoreduction (18% rather than 62%).
Interestingly, in neat TFE only arylchlorocyclohexane 20
and arylcyclohexane 23 not incorporating the solvent
were formed (paths a, d). In the presence of carbonate
the enhanced nucleophilicity of the solvent led to ethers
24 and 25, characterized by the rearrangement of the
chain (paths b, c).
Con clu sion
Summing up, the efficiency of the photofragmentation
of 4-chlorophenol and 4-chloroanisole shows a U-shaped
dependence on the solvent polarity. Homolysis predomi-
nates in cyclohexane, heterolysis in alcohols. In ethyl
acetate and acetonitrile the reaction is less efficient,
apparently because the lifetime of the triplet state is
shortened and makes homolysis less efficient, while
heterolysis becomes significant only with the assistance
by alcohols. This is consistent with sparse reports in
previous literature.
The target of the investigation, however, was to
determine whether photocleavage could be made useful
for the arylation of alkenes and aromatics. This is indeed
the case, and the products obtained are essentially the
same from the phenol and from the anisole, thus dis-
counting the hypothesis that 4-oxocylohexadienylidene
(Scheme 1) has a role in this addition.21 More precisely,
two paths can be followed. In aprotic solvents an exciplex
is formed and leads to arylation via an intimate radical
ion pair. The adduct cation undergoes either deprotona-
tion, giving allylbenzenes, or addition of the original
nucleofuge, giving â-chloroalkylbenzenes. The proportion
of the exciplex path is proportional to the nucleophilicity
of the trap.
In contrast, alcohols as the solvent make unimolecular
fragmentation of the triplet feasible22 and give the
(20) Guizzardi, B.; Mella, M.; Fagnoni, M.; Albini A. Chem. Eur. J .
2003, 9, 1549-1555.
(21) At any rate, the carbene would be expected to give isolable spiro-
cyclopropylcyclohexadienones rather than alkylated phenols; see:
Ershov, V. V.; Nikiforov, G. A.; de J onge, C. R. Quinone Diazides;
Elsevier: Amsterdam, 1981. Field, K. W.; Schuster, G. B. J . Org. Chem.
1988, 53, 4000-4006. Ohno, T.; Martin, N.; Knight, B.; Wudl, F.;
Suzuki, T.; Yu, H. J . Org. Chem. 1996, 61, 1306-1309. Sander, W.;
Ko¨tting, C.; Hu¨bert, R. J . Phys. Org. Chem. 2000, 13, 561-568. Becker,
H. D.; Elebring, T. J . Org. Chem. 1985, 50, 1319-1322.
The reaction with 2,3-dimethyl-2-butene follows quali-
tatively the same pattern as with 1. Deprotonation of the
phenonium ion (path a) to yield allylanisole derivative
(18) For example, the quantum yield of photodegradation in MeOH
is reported to be Φ ) 0.1, whereas in cyclohexane it is somewhat higher
at Φ ) 0.35. See ref 11.
(19) Singlet aryl cation are to our knowledge obtained only from
the photolysis of diazonium salt (see Gasper, S.; Devadoss, C.; Schuster,
G. B. J . Am. Chem. Soc. 1995, 117, 5206-5211 and ref 4)
(22) The role of the formation of the hydrogen-halide bond in
facilitating the heterolysis of the triplet has been demonstrated for
4-fluoroanilines. See: Freccero, M.; Fagnoni, M.; Albini, A. J . Am.
Chem. Soc. 2003, 125, 13182-13190.
3470 J . Org. Chem., Vol. 69, No. 10, 2004

Aryl Cations from Aromatic Halides
solvent chosen (cyclohexane, ethyl acetate, MeCN, MeOH, or
TFE) for 14 h; the same experiments were repeated in the
presence of 0.025 M Cs2CO3, and product distribution was
analyzed by GC. Phenol and anisole were determined on the
basis of calibration curves. Preparative experiments were
carried out only with TFE as the solvent (see below).
Ir r a d ia tion of 1 in TF E. A solution of 95 mg of 1 (0.75
mmol) and 122 mg of Cs2CO3 (0.0375 mmol) in 15 mL of TFE
was irradiated for 26 h. Purification by column chromatogra-
phy (cyclohexane/ethyl acetate 8:2) gave 43 mg of 3 (66% yield)
and 9 mg of 2,4′-hydroxy-4-chlorobiphenyl 430 (6% yield,
viscous oil, lit.6 mp 118-119 °C). Data for 4: 1H NMR (CDCl3)
δ 7.30 (m, 1H), 7.25 (m, 3H), 6.95 (m, 3H), 5.1 (bs, 1H), 5.0
(bs, 1H); IR (neat) ν/cm-1 3360 (OH), 2950, 1688, 1591, 1507,
1231, 1496, 808. Anal. Calcd for C12H9ClO2: C 65.32, H 4.11.
Found: C 65.24, H 4.16.
Ir r a d ia tion of 2 in TF E. A solution of 105 mg of 2 (0.75
mmol) and 122 mg of Cs2CO3 (0.0375 mmol) in 15 mL of TFE
was irradiated for 24 h (90% consumption of 2). Purification
by column chromatography (cyclohexane/ethyl acetate 99:1)
gave 29 mg of 5 (32% yield) and a mixture containing 37 mg
of 2,4′-dimethoxy-4-chlorobiphenyl 631 (18% yield) and 20 mg
of 4-(2,2,2-trifluoroethoxy)anisole 832 (12% yield). A small
amount of pure compound 8 was isolated by further chroma-
tography separation. Data for 6: 1H NMR (CDCl3, from the
mixture) δ 7.5 (m, 1H), 7.45 (m, 1H), 7.25 (m, 1H), 6.65 (m,
1H), 6.90 (d, 2H, J ) 8.5 Hz), 6.85 (m, 1H), 3.90 (s, 6H); IR
(neat, from the mixture) ν/cm-1 2935.6, 2338, 1713, 1610, 1246,
1225, 1029, 972, 831; MS (m/z) 248 (M+, 90), 198 (100). Data
for 8: 1H NMR (CDCl3) δ 6.80-6.90 (m, 4H), 4.30 (q, 2H, J )
8 Hz), 3.80 (s, 3H); 13C NMR (CDCl3) δ 155.0, 151.5, 124.0 (q,
J ) 275 Hz, CF3), 116.2 (CH), 114.7 (CH), 67.0 (q, J ) 35 Hz),
55.6 (CH3); IR (neat) ν/cm-1 2954, 1509, 1228, 1164, 826, 754.
Anal. Calcd for C9H9F3O2: C 52.43, H 4.40. Found: C 52.40,
H 4.33. MS (m/z) 206 (M+, 60), 123 (100), 95 (50).
Ir r a d ia tion of 1 a n d 2 in th e P r esen ce of π Nu cleo-
p h iles. Gen er a l P r oced u r e. Experiments were carried out
as described above starting from 190 mg of 1 or 210 mg of 2
(1.5 mmol, 0.05 M) in 30 mL of the chosen solvent after adding
the π nucleophile (0.5 M of 2,3-dimethyl-2-butene and cyclo-
hexene or 1 M of benzene). Two sets of experiments were
carried out, the first one by irradiating for the same time as
in neat solvent (14 h) and the latter by irradiating up to total
consumption of the starting halide for the preparative experi-
ments (see below). Reactions in the presence of equimolecular
Cs2CO3 were also tested, and the results are reported for the
cases where an increase of the overall yield or a different
photoproducts distribution were observed.
solvated phenyl cation. This is trapped by different
π-nucleophiles with the same efficiency and gives the
likewise solvated phenonium ion. As one may expect, the
latter intermediate reacts with the solvent rather than
with chloride and lives long enough to undergo typical
hydride and alkyl rearrangements, particularly in ion
stabilizing TFE.23 In this poorly nucleophilic solvent the
last step of the reaction is affected by the presence of a
base, which increases formation of rearranged ethers.24
However, cesium carbonate in some cases increases
reduction in neat solvents.
It can be concluded that the photochemical arylation
of alkenes and aromatics starting from chloroaromatics
is not limited to chloroanilines and has been extended
to 4-chlorophenol and 4-chloroanisole. The detailed course
of the reaction and the products formed depend on
conditions. In alcohols 4-hydroxy and 4-methoxyphenyl
cations are involved, whereas in other solvents an exci-
plex/ion pair path is more appropriate. However, in both
ways this novel arylation reaction appears to have
synthetic significance as a mild, metal-free alternative
to current methods26 that may be carried out directly on
phenols, skipping protection of the OH group.29
Exp er im en ta l Section
NMR spectra were recorded on a 300 MHz spectrometer.
The attributions were made on the basis of 1H and 13C NMR,
as well as DEPT-135 experiments; chemical shifts are reported
in ppm downfield from TMS. Cyclohexene, 2,3-dimethyl-2-
butene, benzene, and compounds 1 and 2 were freshly distilled
before use. The solvent was used as received. No previous
anhydrification was carried out, since this did not affect the
results.
The photochemical reactions were performed by using
nitrogen-purged solutions in quartz tubes and a multilamp
reactor fitted with six 15-W phosphor-coated lamps (maximum
of emission 310 nm) for the irradiation. The reaction course
was followed by TLC (cyclohexanes-ethyl acetate) and GC by
comparison with authentic samples. Workup of the photolytes
involved concentration in vacuo and chromatographic separa-
tion.
Ir r a d ia tion of 1 a n d 2 in Nea t Solven t. Gen er a l
P r oced u r e. A 0.05 M solution of 1 or 2 was irradiated in the
P h otoch em ica l Rea ction s of 1 in th e P r esen ce of 2,3-
Dim eth yl-2-bu ten e. In Cycloh exa n e. After 14 h of irradia-
tion (54% of consumption), purification (cyclohexane/ethyl
acetate 8:2 as the eluant) of the raw photolisate afforded 18.3
mg of 3 (24% yield) and 25.6 mg of 2,3-dimethyl-3-(4-hydroxy-
phenyl)-1-butene (9, 18% yield, see below).
In Eth yl Aceta te. After 23 h of irradiation, purification
(cyclohexane/ethyl acetate 8:2 as the eluant) of the raw
photolisate afforded 27.7 mg of 3 (20% yield) and 182 mg of 9
(colorless solid, mp 70-71 °C, 69% yield). Data for 9: 1H NMR
(CDCl3) δ 6.75-7.20 (AA′BB′, 4H), 4.85-5.00 (bs, 2H), 1.55
(s, 3H), 1.40 (s, 6H); IR (neat) ν/cm-1 3286 (OH), 2960, 2378,
1551, 1432, 1370, 1230, 1176, 894, 830. Anal. Calcd for
(23) The growing interest in the use of fluorinated alcohols in organic
synthesis, as a result of the clean and selective reactions obtained, is
apparent in a recent review: Be´gue´, J .-P.; Bonnet-Delpon, D.; Crousse,
B. Synlett 2004, 18-29.
(24) The effect of addition of cesium carbonate is complex. It
somewhat enhances the conversion of chlorophenol 1, reasonably
because some phenolate anion is then present, which absorbs a large
proportion of the light used for irradiation (lamps centered at 310 nm)
since its absorption is red-shifted by ca. 10 nm. This effect seems more
important for 1 in polar nonprotic solvents despite the slight solubiblity
of the base in this medium. In particular, in ethyl acetate the addition
of cesium carbonate increased the conversion of 1 from 11% to 50%,
with a concomitant increase of the photoreduction yield. Even if the
phenolate absorbs part of the light, the photoproducts distribution did
not change, in accord with previous reports.8a,25 In accord with this
rationalization, the reactivity of a nonacidic compound such as anisole
2 was affected to a much lower degree.
C
14H19F3O2: C 60.86, H 6.93. Found: C 60.89, H 7.01.
In Aceton itr ile. After 24 h of irradiation, purification
(25) Boule, P.; Guyon, C.; Lemaire, J . Chemosphere 1982, 11, 1179-
1188.
(cyclohexane/ethyl acetate 8:2 as the eluant) of the raw
photolisate afforded 12.6 mg of 3 (9% yield) and 177 mg of 9
(67% yield).
(26) The formation of â-alkoxyalkylarenes is in common with the
NOCAS reaction (see ref 27), but the scope is complementary (the
NOCAS reaction applies to arylnitriles and not to electron-donating
substituted aromatics) and the mechanism is opposite. Likewise, the
S
RN1 reaction28 applies to electrophilic but not nucleophilic alkenes.
(27) Mangion, D.; Arnold, D. R. Acc. Chem. Res. 2002, 35, 291-304.
(28) Rossi, R. A.; Pierini, A. B.; Santiago, A. N. Org. React. 1999,
(30) Sarakha, M.; Bolte, M.; Burrows, H. D. J . Phys. Chem. A 2000,
104, 3142-3149.
(31) Nakamura, M.; Sawasaki, K.; Okamoto, Y.; Takamuku, S. J .
Chem. Soc., Perkin Trans. 1 1994, 141-146.
(32) Camps, F.; Coll, J .; Messeger, A.; Pericas, M. A. Synthesis 1980,
727-728.
54, 1-271.
(29) J ackson, R. F. W.; Rilatt, I.; Murray, P. J . Org. Biomol. Chem.
2004, 2, 110-113.
J . Org. Chem, Vol. 69, No. 10, 2004 3471

Protti et al.
In Meth a n ol. After 14 h of irradiation, purification (cyclo-
hexane/ethyl acetate 9:1 as the eluant) of the raw photolisate
afforded 86 mg of 3 (61% yield) and 58 mg of 9 (22% yield).
The same reaction carried out in the presence of Cs2CO3
(244 mg, 0.075 mmol, 0.025 M) afforded 63.4 mg of 3 (45%
yield) and 80 mg of 2,3-dimethyl-2-(4-hydroxyphenyl)-3-meth-
oxybutane (10, oil, 26% yield). Data for 10: 1H NMR (CDCl3)
δ 6.75-7.35 (AA′BB′, 4H), 3.15 (s, 3H), 1.35 (s, 6H), 1.03 (s,
6H); 13C NMR (CDCl3) δ 153.1, 139.7, 129.6 (CH), 113.6 (CH),
78.5, 49.5 (CH3), 44.7, 24.4 (CH3), 19.8 (CH3); IR (neat) ν/cm-1
3365 (OH), 2981, 2365, 1611, 1512, 1371, 1264, 1152, 1064,
832, 739. Anal. Calcd for C13H20O2: C 74.96, H 9.68. Found:
C 75.01, H 9.77.
In TF E. After 14 h of irradiation of the solution buffered
with Cs2CO3, purification (cyclohexane/ethyl acetate 9:1 as the
eluant) of the raw photolisate yielded 27 mg of 3 (19% yield),
114 mg of 2,3-dimethyl-3-(4-hydroxyphenyl)-1-butene (9, oil,
43% yield), and 51 mg of 2,3-dimethyl-3-(4-hydroxyphenyl)-3-
(2,2,2-trifluoroethoxy)butane (11, oil, 12% yield). Data for 11:
1H NMR (CDCl3) δ 6.75-7.35 (AA′BB′, 4H), 3.65 (q, 2H, J )
8 Hz), 1.40 (s, 6H), 1.05 (s, 6H); 13C NMR (CDCl3) δ 153.3,
129.5 (CH), 127.2 (q, J ) 275 Hz, CF3), 113.7 (CH), 109.15,
80.4, 60.5 (q, J ) 35 Hz), 44.6, 24.2 (CH3), 20.4 (CH3); IR (neat)
ν/cm-1 3355 (OH), 2975, 1612, 1514, 1377, 1280, 1162, 972,
831. Anal. Calcd for C12H16O: C 81.77, H 9.15. Found: C 81.71,
H 9.17.
P h otoch em ica l Rea ction s of 2 in th e P r esen ce of 2,3-
Dim eth yl-2-bu ten e. In Cycloh exa n e. After 26 h of irradia-
tion, purification (eluant changing form neat cyclohexane to
cyclohexane/ethyl acetate 98:2 mixtures) of the raw photolisate
afforded 2.3 mg of 5 (2% yield) and a mixture containing 95.3
mg of 2,3-dimethyl-3-(4-methoxyphenyl)-1-butene 1233 (34%
yield, see below) and 10 mg of 2,3-dimethyl-2-(4-methoxyphen-
yl)butane (13, 4% yield). Data for 13: 1H NMR (CDCl3, from
the mixture) δ 6.80-7.30 (AA′BB′, 4H), 3.80 (s, 3H), 1.25 (s,
6H), 0.75 (d, 6H, J ) 9 Hz); MS (m/z) 192 (M+, 2), 149 (100),
121 (28); IR (neat, for the mixture) ν/cm-1 2929, 1611, 1514,
1448, 1247, 1178, 1095, 1038, 820.
tane (15, 17% yield) and 36 mg of 2-(4-methoxyphenyl)-2-(2,2,2-
trifluoroethoxy)-3,4-dimethylbutane (16, 8% yield). Data for
15: 1H NMR (CDCl3, from the mixture) δ 6.80-7.30 (AA′BB′,
4H), 3.8 (s, 3H), 3.65 (q, 2H, J ) 8 Hz), 1.4 (s, 6H), 1.05 (s,
6H); 13C NMR (CDCl3) δ 153.3, 138.9, 129.9, 124 (q, J ) 270
Hz, CF3), 112.7 (CH), 80.5, 60.2 (q, J ) 35 Hz), 55.2 (CH3),
44.6, 24.9 (CH3), 20.4 (CH3); MS (m/z) 191 (M-CF3CH2O, 19),
149 (16), 73 (100). Data for 16: 1H NMR (CDCl3, from the
mixture) δ 6.85-7.40 (AA′BB′, 4H), 3.85 (s, 3H), 3.65 (q, 2H,
J ) 8 Hz), 1.6 (s, 3H), 0.9 (s, 9H); 13C NMR (CDCl3) δ 158.3,
132.9, 129.3 (CH), 124 (q, J ) 270 Hz, CF3), 112.6 (CH), 84.3,
66.3 (q, J ) 35 Hz), 55.2 (CH3), 38.7, 25.4 (CH3), 18.9 (CH3).
IR, (neat, for the mixture) ν/cm-1 2974, 1611, 1513, 1281, 1248,
1162.
P h ot och em ica l R ea ct ion s of 1 in t h e P r esen ce of
Cycloh exen e. In Cycloh exa n e. After 14 h of irradiation
(48% of consumption), phenol 3 (21% yield) and traces of
compound 17 (see below) were detected by GC analysis.
Further irradiation did not increase the amount of the arylated
product.
In Eth yl Aceta te. After 24 h of irradiation, purification
(cyclohexane/ethyl acetate 9:1 as the eluant) of the raw
photolisate afforded 38 mg of 3 (24% yield) and 70 mg of trans
1-(p-hydroxyphenyl)-2-chlorocyclohexane (17, 20% yield, color-
less solid, mp 127-128 °C). Data for 17: 1H NMR (CDCl3) δ
6.9-7.1 (AA′BB′, 4H), 3.9 (dt, 1H, J ) 4 and 12 Hz), 2.6 (dt,
1H, J ) 4 and 12 Hz), 2.4 (m, 1H), 1.9 (m, 4H), 1.5 (m, 3H);
13C NMR (CDCl3) δ 154.0, 136.2, 128.4 (CH), 115.1 (CH), 64.7
(CH), 52.6 (CH), 37.1 (CH2), 35.6 (CH2), 26.5 (CH2), 25.7 (CH2);
IR (neat) ν/cm-1 3415 (OH), 2928, 1611, 1513, 1448, 1247,
1150, 826. Anal. Calcd for C12H15ClO: C 68.41, H 7.18.
Found: C 68.50, H 7.08. The assignment of the trans config-
uration was based on the high value (12 Hz) of the coupling
constant between H1 and H2 in the cyclohexane ring.
In Aceton itr ile. After 24 h of irradiation, purification
(cyclohexane/ethyl acetate 9:1 as the eluant) of the raw
photolisate afforded 25 mg of 3 (18% yield) and 38 mg of 17
(12% yield).
In Meth a n ol. A solution of 190 mg (1.5 mmol, 0.05 M) of
1, 1.6 mL (15 mmol, 0.5 M) of cyclohexene in 30 mL of MeOH
was irradiated for 14 h. Solvent was removed ,and the
resulting raw photolisate was dissolved in 10 mL of DMF and
treated with 300 mg of CaCO3 and 400 µL (6 mmol) of CH3I.
The resulting solution was stirred for 3 h, diluted with 20 mL
of water, and extracted with Et2O. The crude product obtained
was purified by column chromatography (cyclohexane/ethyl
acetate 99:1) affording 67 mg of 5 (62% yield) and a mixture
(resolved by NMR spectroscopy) containing 26 mg of 1-(1-(p-
methoxyphenyl)-1-methoxy)methylcyclopentane (19, 8% yield)
and 12 mg of trans 1-(p-methoxyphenyl)-2-chlorocyclohexane
(20,34 4% yield, see below). Data for 19: 1H NMR (CD3COCD3,
from the mixture) δ 6.9-7.4 (AA′BB′ 4H), 3.78 (d, 1H, J ) 9
Hz), 3.8 (s, 3H), 2.9 (s, 3H), 2.1 (m, 1H), 1.2-2.0 (m, 8H); 13C
NMR ((CD3COCD3) δ 158.5, 137.9, 127.0 (CH), 113.2 (CH), 87.6
(CH), 55.4 (CH3), 46.9 (CH3), 46.9 (CH), 35.2 (CH2), 29.5 (CH2),
24.9 (CH2), 21.7 (CH2). IR, (neat, for the mixture) ν/cm-1 2838,
1611, 1513, 1459, 1278, 1250, 1115, 1037, 970, 830.
In TF E. After 24 h of irradiation, purification (cyclohexane/
ethyl acetate 9:1 as the eluant) of the raw photolisate afforded
42 mg of 3 (30% yield) and mixtures containing 24 mg of 17
(8% yield), 24.5 mg of 1-(1-(p-hydroxyphenyl)-1-(2,2,2-trifluo-
roethoxy)methylcyclopentane (21, 6% yield), and 16.5 mg of
1-(p-hydroxyphenyl)-1-(2,2,2-trifluoroethoxy)cyclohexane (22,
4% yield). Data for 21: 1H NMR (CD3COCD3, from the
mixture) δ 6.8-7.4 (AA′BB′, 4H), 4.1 (m, 1H), 3.7 (q, 2H, J )
9 Hz), 2.2 (m, 1H), 1.1-1.8 (m, 8H); 13C NMR (CD3COCD3) δ
156.7, 135.1, 135.0 (q, CF3, J ) 275 Hz), 127.3 (CH), 114.8
(CH), 87.4 (CH), 64.7 (q, CH2, J ) 35 Hz), 46.5 (CH), 35.6
In Eth yl Aceta te. After 23 h of irradiation, purification
(eluant changing from neat cyclohexane to cyclohexane/ethyl
acetate 98:2 mixtures) of the raw photolisate afforded 11.3 mg
of 5 (7% yield) and 154 mg of 12 (54% yield). Data for 12: 1H
NMR (CDCl3) δ 6.80-7.30 (AA′BB′, 4H), 4.90 (bs, 1H), 5.00
(bs, 1H), 3.80 (s, 3H), 1.60 (s, 3H), 1.45 (s, 6H); 13C NMR
(CDCl3) δ 157.4, 152.8, 140.4, 127.0 (CH), 113.2 (CH), 109.2
(CH2), 55.1 (CH3), 43.1, 28.4 (CH3), 20.0 (CH3); IR (neat) ν/cm-1
2968, 1610, 1512, 1250, 831; MS (m/z) 190 (M+, 49), 175 (91),
149 (100), 121 (44). Anal. Calcd for C13H18O: C 82.06, H 9.53.
Found: C 81.99, H 9.51.
In Aceton itr ile. After 26 h of irradiation, purification
(cyclohexane/ethyl acetate 98:2 as the eluant) of the raw
photolisate afforded 4 mg of 5 (2% yield) and 139 mg of 12
(49% yield).
In Meth a n ol. After 16 h of irradiation of the solution
buffered with Cs2CO3, purification (cyclohexane/ethyl acetate
99:1 as the eluant) of the raw photolisate yielded 35.7 mg of 5
(22% yield) and two fractions containing 45 mg of 12 (16%
yield), 48 mg of 13 (17% yield) and 55 mg of 2,3-dimethyl-2-
(4-methoxyphenyl)-3-methoxybutane (14, 17% yield). Further
purification of the fractions containing product 14 afforded a
pure sample of such compound. Data for 14: 1H NMR
(CDCl3): δ 6.80-7.40 (AA′BB′, 4H), 3.8 (s, 3H), 3.15 (s, 3H),
1.4 (s, 6H), 1.05 (s, 6H); 13C NMR (CDCl3) δ 157.2, 139.5, 129.4
(CH), 112.9 (CH), 78.5, 55.0 (CH3), 49.4 (CH3), 32.4, 24.4 (CH3),
17.8 (CH3); IR (neat) ν/cm-1 2973, 1611, 1513, 1250, 832. Anal.
Calcd for C14H22O2: C 75.63, H 9.97. Found: C 75.53, H 10.01.
In TF E. After 14 h of irradiation of the solution buffered
with Cs2CO3, purification (cyclohexane/ethyl acetate 99:1 as
the eluant) of the raw photolisate yielded 19.5 mg of 5 (12%
yield) and two fractions: the first contained 60 mg of 12 (21%
yield) and 41 mg of 13 (14% yield), and the second 74 mg of
2,3-dimethyl-3-(4-methoxyphenyl)-3-(2,2,2-trifluoroethoxy)bu-
(33) Shirakawa, E.; Takahashi, G.; Tsuchimoto, T.; Kawakami, Y.
Chem. Commun. 2002, 2210-2211.
(34) Govindasamy, S.; Nishiyama, H. J . Am. Chem. Soc. 2001, 123,
3603-3604.
3472 J . Org. Chem., Vol. 69, No. 10, 2004

Aryl Cations from Aromatic Halides
(CH2), 29.6 (CH2), 24.5 (CH2), 21.5 (CH2). Data for 22: 1H NMR
(CD3COCD3, from the mixture) δ 6.8-7.1 (AA′BB′, 4H), 3.6
(q, 2H, J ) 9 Hz), 2.1 (m, 2H), 1.1-1.8 (m, 8H); 13C NMR (CD3-
COCD3) δ 157.2, 135.0 (q, CF3 J ) 275 Hz), 131.0, 128.1 (CH),
115.0 (CH), 59.9 (q, CH2, J ) 35 Hz), 35.0 (CH2), 28.9 (CH2),
24.9 (CH2), 24.8 (CH2), 21.5 (CH2). IR (neat, for the mixture
of 21 and 22) ν/cm-1 3356 (OH), 1638, 1612, 1513, 1438, 1364,
1236, 994, 915, 827.
The same reaction carried out in the presence of Cs2CO3
and irradiated for 14 h afforded 61.5 mg of 3 (44% yield), a
mixture containing 37.5 mg of 17 (12% yield), 54 mg of 1-(1-
(p-hydroxyphenyl)-1-(2,2,2-trifluoroethoxy)methylcyclopen-
tane (21, 13% yield).
P h ot och em ica l R ea ct ion s of 2 in t h e P r esen ce of
Cycloh exen e. In Cycloh exa n e. After 26 h of irradiation,
purification (cyclohexane as the eluant) of the raw photolisate
gave 26 mg of 5 (16%) and 125.4 mg of 20 (38%, oil). Data for
20: 1H NMR (CDCl3) δ 6.8-7.2 (AA′BB′, 4H), 4.2 (dt, 1H, J )
4 and 11 Hz), 3.8 (s, 3H), 2.7 (dt, 1H, J ) 4 and 11 Hz), 2.3
(m, 1H), 2.1 (m, 1H), 1.8 (m, 3H), 1.6 (m, 3H); 13C NMR (CDCl3)
δ 158.3, 136.3, 128.5 (CH), 113.4 (CH), 64.5 (CH), 54.9 (CH3),
52.7 (CH), 37.9 (CH2), 35.6 (CH2), 26.1 (CH2), 25.6 (CH2); IR
(neat) ν/cm-1 2932, 1610, 1511, 1459, 1247, 1178, 1074, 1039,
829. Anal. Calcd for C13H17ClO: C 69.48, H 7.62. Found: C
69.34, H 7.52. MS (m/z) 224 (M+, 50), 189 (10), 147 (100), 121
(45), 91 (32). The assignment of the trans configuration was
based on the high value (11 Hz) of the coupling constant
between H1 and H2 in the cyclohexane ring.
2H), 1.20-1.80 (m, 8H); 13C NMR (CDCl3) δ 157.5, 140.3, 127.5
(CH), 113.6 (CH), 55.1 (CH3), 48.3 (CH3), 35.2 (CH2), 29.0
(CH2), 25.3 (CH2), 25.0 (CH2), 21.7 (CH2); IR (neat, for the
mixture) ν/cm-1 2928, 2855, 2363, 1610, 1512, 1247, 1178,
1037, 824; MS (m/z) 190 (M+, 60), 147 (100), 134 (30), 121 (56),
91 (45), 43 (45).
The same reaction carried out in the presence of Cs2CO3
and irradiated for 14 h yielded 26 mg of 5 (10% yield), 25 mg
of 20 (9% yield) and a mixture containing 84 mg of 1-(p-
methoxyphenyl)-1-(2,2,2-trifluoroethoxy)cyclohexane (24, 20%
yield) and 89.4 mg of 1-(1-(p-methoxyphenyl)-1-(2,2,2-trifluo-
roethoxy)methylcyclopentane (25, 21% yield). Data for 24: 1H
NMR (CDCl3, from the mixture) δ 6.8-7.2 (AA′BB′, 4H), 3.8
(s, 3H), 3.6 (m, 2H), 2.15 (m, 2H), 1.1-1.8 (m, 8H); 13C NMR
(CDCl3) δ 159.8, 135.8, 135.0 (q, CF3, J ) 275 Hz), 128.1 (CH),
113.7 (CH), 60.9 (q, CF3, J ) 35 Hz), 29.2 (CH2), 25.7 (CH2),
25.2 (CH2), 25.0 (CH2), 22.1 (CH2). Data for25: 1H NMR
(CDCl3, from the mixture) δ 6.8-7.3 (AA′BB′, 4H), 4.1 (d, 1H,
J ) 9 Hz), 3.8 (s, 3H), 3.6 (m, 2H), 2.1 (m, 1H), 1.1-1.8 (m,
8H); 13C NMR (CDCl3) δ 158.8, 136.3, 135.0 (q, CF3, J ) 275
Hz), 128.1 (CH), 113.6 (CH), 87.8 (CH), 65.6 (q, CH2, J ) 35
Hz), 46.7 (CH), 29.8 (CH2), 27.3 (CH2), 22.1 (CH2), 21.3 (CH2).
IR (neat, for the mixture) ν/cm-1 2969, 1638, 1612, 1512, 1438,
1364, 1236, 994, 915, 827.
P h ot och em ica l R ea ct ion s of 1 in t h e P r esen ce of
Ben zen e. In TF E. A solution of 190 mg (1.5 mmol, 0.05 M)
of 1, 2.7 mL (30 mmol, 1 M) of benzene in 30 mL of TFE was
irradiated for 24 h. Purification by column chromatography
(cyclohexane/ethyl acetate 8:2) afforded 23 mg of 3 (16% yield)
and 168 mg of 4-phenylphenol (26,36 66% yield) as a colorless
solid, mp 163-165 °C (lit. 165 °C Acros-organics), with
spectroscopic characteristics identical to those of an authentic
sample.
In Eth yl Aceta te. After 23 h of irradiation, purification
(cyclohexane as the eluant) of the raw photolisate gave 24 mg
of 5 (15%) and 115 mg of 20 (32%).
In Aceton itr ile. After 24 h of irradiation, purification
(cyclohexane as the eluant) of the raw photolisate gave 41 mg
of 5 (26%) and 73 mg of 20 (20%).
P h ot och em ica l R ea ct ion s of 2 in t h e P r esen ce of
Ben zen e. In TF E. A solution of 210 mg (1.5 mmol, 0.05 M)
of 2, 2.7 mL (30 mmol, 1 M) of benzene in 30 mL of TFE was
irradiated for 24 h (90% consumption). Purification by column
chromatography (cyclohexane/ethyl acetate 9:1) afforded 10 mg
of 5 (6%) and 193 mg of 4-phenylanisole (27, 70%). Crystal-
lization from hexane gave a colorless solid, mp 81-83 °C (lit.37
90-90.5 °C). Spectroscopic data were in accordance with the
literature.38
In Meth a n ol. After 24 h of irradiation, purification (cyclo-
hexane as the eluant) of the raw photolisate gave 26.5 mg of
5 (18% yield) and a mixture containing 104 mg of 18 (32%
yield) and 48 mg of 19 (15% yield). Data for 18: 1H NMR
(CDCl3, from the mixture) δ 6.90-7.20 (AA′BB′, 4H), 3.80 (s,
3H), 3.15 (s, 3H), 2.10-2.20 (m, 2H), 1.20-1.80 (m, 8H); 13C
NMR (CDCl3) δ 159.0, 134.0, 128.2 (CH), 113.3 (CH), 76.3, 54.4
(CH3), 48.3 (CH3), 35.2 (CH2), 29.0 (CH2), 25.3 (CH2), 25.0
(CH2), 21.7 (CH2); IR (neat, for the mixture) ν/cm-1 2932, 2856,
1609, 1512, 1463, 1248, 1178, 1038, 825, 737.
In TF E. After 26 h of irradiation (80% conversion), purifica-
tion (cyclohexane as the eluant) of the raw photolisate afforded
14 mg of 5 (11% yield) and a mixture containing 80 mg of 20
(29% yield) and 105 mg of p-methoxyphenylcyclohexane (23,35
45%). Data for 23: 1H NMR (CDCl3, from the mixture) δ 6.80-
7.15 (AA′BB′, 4H), 3.77 (s, 3H), 3.15 (s, 3H), 2.10-2.20 (m,
Ack n ow led gm en t. Partial support of this work by
Murst, Rome is gratefully acknowledged.
J O049770+
(36) Blass, B. E.; Harris, C. L.; Portlock, D. E. Tetrahedron Lett.
2001, 42, 1611-1614.
(37) Abramovitch, S. J . Chem. Soc. B 1966, 733-735.
(38) Li, G. Y. J . Organomet. Chem. 2002, 653, 63-68.
(35) Tateiwa, J .; Hayama, E.; Nishimura, T.; Uemura, S. J . Chem.
Soc., Perkin Trans. 1 1997, 1923-1928.
J . Org. Chem, Vol. 69, No. 10, 2004 3473