Photoreaction of halomethyl substituted benzocyclic ketones with amines: Radical cyclization and ring expansion reactions promoted through photoinduced electron transfer processes

Article abstract of DOI:10.1039/a703974c

Photoreaction of ethyl 2-bromomethyl-1-oxo-1,2,3,4-tetrahydronaphthalene-2-carboxylate 1a or other related compounds 1b-d with Me₃SiCH₂NEt₂ in aqueous MeCN afforded ethyl 5-oxo-6,7,8,9-tetrahydrobenzocycloheptene-7-carboxylate 2a or corresponding ring expansion products 2a-d respectively.

Full text of DOI:10.1039/a703974c

View Article Online / Journal Homepage / Table of Contents for this issue
Photoreaction of halomethyl substituted benzocyclic ketones with amines:
radical cyclization and ring expansion reactions promoted through
photoinduced electron transfer processes
Eietsu Hasegawa,* Yukinobu Tamura and Emi Tosaka
Department of Chemistry, Faculty of Science, Niigata University, Ikarashi, Niigata 950-21, Japan
Photoreaction of ethyl 2-bromomethyl-1-oxo-1,2,3,4-
It was also found that water significantly affects the product
tetrahydronaphthalene-2-carboxylate 1a or other related
ratio when the reactions were conducted for 1 h in MeCN
containing various amounts of water: 61% of 2a, 10% of 3a,
20% of 4a and 24% of 1a in 5% aq. MeCN; 68, 7, 16 and 25%
in 10% aq. MeCN; 75, 6, 10 and 25% in 30% aq. MeCN. While
addition of water increased the yield of 2a and decreased the
yields of 3a and 4a, the total yield of the products remained
almost constant at about 90%. This would suggest that the
observed products are formed through common intermediates.
In order to obtain information on the nature of the intermediates,
a deuterium labeling experiment was performed. Thus, when
3 2 2
compounds 1b–d with Me SiCH NEt in aqueous MeCN
afforded ethyl 5-oxo-6,7,8,9-tetrahydrobenzocycloheptene-
7
-carboxylate 2a or corresponding ring expansion products
2
a–d respectively.
Irradiation of carbonyl compounds in the presence of electron
donors such as amines produces carbonyl anion radicals and
1
donor cation radicals. Apparently proton transfer from donor
cation radicals to carbonyl anion radicals is a feasible process.
On the other hand, it is also reasonable to expect that carbonyl
anion radicals must act as single electron donors. Therefore,
generation of carbonyl anion radicals would irreversibly reduce
distant carbon–halogen bonds in the same molecules through
irradiation of 1a for 2 h were conducted in 5% D
2
O as well as
5% H O containing MeCN solutions, similar conversions (90
2
and 90%) as well as product yields (2a: 60 and 59%; 3a: 7 and
7%) were observed, respectively. NMR analyses of the isolated
products revealed that ca. 82% of hydrogen at C in 2a was
3
2
intramolecular electron transfer. Although photoreactions of a
few carbonyl compounds categorized in this group have been
explored in the presence of electron donors, the substrates
studied were limited to enones and dienones. In this context,
substituted by deuterium and no deuterium was incorporated in
3a. Therefore, 2a is mainly formed through an ionic route in
which its precursor is protonated. On the other hand, 3a is
formed through hydrogen abstraction by the radical inter-
mediate.
On the basis of the above results, the reaction mechanism in
Scheme 2 is proposed. Single electron transfer to the photo-
3
photoinduced electron transfer (PET) reaction of haloalkylated
benzocycloalkanones such as 1 must be interesting since
reductive carbon–halogen bond cleavage should generate
carbon radicals followed by radical cyclization and ring
4
expansion pathways. Moreover, the literature observation that
excited state of 1a from Me
radical of 1a and the cation radical of Me
3
SiCH
2
NEt
2
produces the anion
SiCH NEt .‡ The
the yields of ring expansion products from 1 (X = Br) were
relatively low under ordinary free radical conditions5 also
prompted us to explore photoreaction of 1 with electron donors.
We report here the first example of PET promoted radical
cyclization and ring expansion reactions of ethyl 2-bromo-
methyl-1-oxo-1,2,3,4-tetrahydronaphthalene-2-carboxylate 1a
and other related compounds 1b–d (Scheme 1).
3
2
2
resulting anion radical of 1a rapidly reacts,§ undergoing
carbon–bromine bond cleavage through intramolecular electron
transfer to give the carbon radical 5a. The radical 5a abstracts
hydrogen to produce 3a, which is consistent with the deuterium
labeling experiment. On the other hand, 5a cyclizes to give the
oxy radical 6a followed by the ring opening to the ethoxy
When a N
l = 295 nm, e 820; 0.64 mmol) and Me
was irradiated for 1 h with a 500 W Xe-Hg lamp through Pyrex,
0% of the desired benzocycloheptanone 2a, 10% of the
reduced product 3a and 21% of unexpected dimeric compound
a (as a mixture of two diastereomers 4a :4a = 55:45) was
isolated together with 35% of unreacted 1a after chromato-
graphic separation (the product yields are reported on the basis
of the conversion of 1a).† Stirring of a similar solution for 2 h
in the dark resulted in quantitative recovery of 1a. Upon
2
pre-purged MeCN solution (10 ml) of 1a
(
3
SiCH NEt
2
2
(5 equiv.)6
O
O –
•
O
Me
E
5
•
Br
H •
E
E
4
1
2
Br –
3a
O•
1a– •
5a
E: CO2Et
O
irradiation of 1a for 2 h in the absence of Me
3
2
SiCH NEt
2
, 92%
E
of 1a was recovered without formation of 2a. These control
experiments clearly suggest that 2a is produced through the
•
E
6a
interaction between the photoexcited 1a and Me
3 2 2
SiCH NEt .
e –
O
7a
O
7a
O
O
X
H •
E
2
hn, Me3SiCH2NEt2, N2
–
E
CO2Et
H2O, MeCN
CO2Et
O
H +
8a
4a
n
n
1
a X = Br, n = 1
b X = I, n = 1
c X = Br, n = 2
d X = Br, n = 0
2a n = 1
c n = 2
d n = 0
E
2a
Scheme 1
Scheme 2
Chem. Commun., 1997
1895

View Article Online
O
O
OH
derived from the cation radical of DMPBI is readily oxidized to
a stable imidazolium ion, it would act as a stronger reductant
than that from the cation radical of Me SiCH NEt .
3 2 2
–
H •
•
As described above, we have discovered that a PET process
can promote the conversion of certain halomethyl substituted
benzocyclic ketones to ring expanded ketones. Further studies
focused on designing substrates as well as electron donors are
now in progress, with the aim of providing a thorough
mechanistic understanding and an expansion of the synthetic
versatility of this unique PET reaction.
This work was partly supported by a Grant-in-Aid for
Scientific Research (No. 08640676) from the Ministry of
Education, Science, Sports and Culture of Japan. We thank
Professors Yasutake Takahashi and Tsutomu Miyashi (Tohoku
University) for performing the pulsed laser photolyses and their
useful comments. We thank Professor Masaki Kamata (Faculty
of Education) for his assistance in measuring redox potentials.
We thank Professors Yoshiki Okamoto and Takaaki Horaguchi
(Faculty of Science) for their generous support.
E
E
E
11
10
9
e –
O
O
Me
N
H
H +
N
Me
Ph
E
E
DMPBI
12
2d
E = CO2Et
Scheme 3
carbonyl substituted carbon radical 7a. Apparently, 7a has a
long enough lifetime to dimerize, giving 4a. Cyclization of the
carbon radical on the carbonyl group and ring opening would be
7
a reversible process (5a ? 6a ? 7a). If 7a abstracts hydrogen,
2
a should be formed. However, the deuterium labeling
Footnotes and References
experiment suggests that this is a minor route, even though it
exists. Instead, 7a is reduced to its anion 8a followed by
protonation to give the formation of 2a.
In the proposed reaction mechanism, the most plausible
candidate for the reducing agent of 7a must be the a-amino
radical formed through desilylation of the cation radical of
* E-mail: ehase@sc.niigata-u.ac.jp
† Ring expansion products 2a,c and the reduced product 3a are known
compounds (ref. 5). The structure of unknown 2d was determined by direct
comparison of its spectral data with that of the compound independently
prepared by reaction of 1d with tributyltin hydride. Compounds 4a and 9
were satisfactorily characterized by their NMR and IR spectroscopic
data.
3 2 2
Me SiCH NEt . Thus, effective formation of the a-amino
radical is required for the formation of 2a. Indeed, when 1a was
irradiated with 1,4-diazabicyclo[2.2.2]octane (DABCO) for 5 h
in MeCN, no formation of 2a was observed. Although DABCO
is a strong enough electron donor,‡ its cation radical is known
to be stable, so that the efficient formation of the a-amino
‡
Free energy changes (DG) for single electron transfer between the triplet
excited state of 1a and amines were estimated (D. Rehm and A. Weller, Isr.
J. Chem., 1970, 8, 259) from the triplet energy of 1,2,3,4-tetra-
hydronaphthalen-1-one (72.7 kcal mol : S. L. Murov, Handbook of
Photochemistry, Marcel Dekker, New York, 1973) and the redox potentials
of 1a (21.85 V vs. SCE) and amines (219 kcal mol for Me SiCH NEt ,
2
§
2
1
3
21
8
radical by deprotonation cannot be expected. In the reaction
system of 1a and Me SiCH NEt , water would act as a silophile
to generate the a-amino radical by desilylation of the cation
3
2
2
2
1
21
15 kcal mol for DABCO, 211 kcal mol for NEt
3
).
3
2
2
Pulsed laser photolysis (355 nm) of 1a in nitrogen-saturated MeCN
resulted in a transient absorption around 370 nm which was assigned as the
T–T absorption. In the presence of Me SiCH NEt , the T–T absorption was
replaced by a new shoulder absorption ranging from 400 to 500 nm, which
9
3 2 2
radical of Me SiCH NEt . On the other hand, water would not
3
2
2
be a strong enough base for a-deprotonation of an amine cation
radical.10 This was confirmed by the observation that photo-
reactions of 1a with triethylamine‡ in dry as well as wet MeCN
reveal no significant water effect on the product ratio. Another
notable observation was made when the reactions were
conducted for 1 h with addition of various quantities of
6
21
decayed with a rate constant of ca. 3 3 10 s . A plausible explanation for
these observations is that the triplet excited state of 1a is quenched by
3 2 2
Me SiCH NEt to generate the anion radical of 1a, which decays rapidly
owing to its high intramolecular reactivity. The cyclic voltammetric
behaviour of 1a is also consistent with the above explanation: the redox
2
1
process of 1a is totally irreversible (100 mV s in MeCN).
An alternative explanation for this phenomenon would be the occurrence
Me
of 4a and 39% of 1a with 1.5 equiv. of Me
6 and 25 with 5 equiv.; 73, 10, 12 and 27 with 10 equiv.
Increasing the quantity of Me SiCH NEt increased the yield of
a as well as 3a and decreased that of 4a, which would suggest
that excess Me SiCH NEt donates a hydrogen to the radical
3
2
SiCH NEt
2
in 10% aq. MeCN: 56% of 2a, 7% of 3a, 24%
¶
3
SiCH NEt ; 68, 7,
2
2
of direct hydrogen transfer from the cation radical of DMPBI or its neutral
form to the radical 11 to give 2d (Note added in proof).
1
3
2
2
2
1
Photoinduced Electron Transfer, Part C, ed. M. A. Fox and M. Chanon,
Elsevier, Amsterdam, 1988; U. C. Yoon, P. S. Mariano, R. S. Givens
and B. W. Atwater III, in Advances in Electron Transfer Chemistry, ed.
P. S. Mariano, JAI, Greenwich, 1994, vol. 4, pp. 117–205.
3
2
2
intermediates, leading to the formation of 2a and 3a, and
eventually decreasing the yield of 4a.
In order to determine the generality of this PET reaction,
photoreactions of other halo ketones 1b–d were briefly
conducted (product yields were not optimized). Irradiation of
the iodide 1b with Me SiCH NEt produced 46% of 2a. The
3 2 2
moderate yield of 2c (41%) from 1c is still encouraging since
thermal reaction of 1c with tributyltin hydride produced 2c only
as a minor product. More interestingly, it turned out that the
major product from 1d was ethyl 1-hydroxynaphthalene-
-carboxylate 9 (52%) instead of 2d, which was isolated in 14%
yield in the photoreaction of 1d with Me SiCH NEt . As shown
in Scheme 3, naphthol 9 is considered to be an enol form of
ketone 10, which is probably formed via hydrogen abstraction
from the ethoxy carbonyl substituted carbon radical 11. If so,
the relative yield of 2d vs. 9 should increase under the
conditions in which 11 is efficiently reduced to its anion 12.
2 N. Kimura and S. Takamuku, J. Am. Chem. Soc., 1994, 116, 4087.
3
R. S. Givens and B. W. Atwater III, J. Am. Chem. Soc., 1986, 108,
5
028; E. Hasegawa, Y. Tamura, T. Horaguchi, K. Isogai and T. Suzuki,
Tetrahedron Lett., 1994, 35, 8643.
4
5
6
P. Dowd and Z. Wei, Chem. Rev., 1993, 93, 2091.
W. R. Bowman and P. J. Westlake, Tetrahedron, 1992, 48, 4027.
E. Hasegawa, W. Xu, P. S. Mariano, U. C. Yoon and J. U. Kim, J. Am.
Chem. Soc., 1988, 110 8099.
5
7 D. P. Curran, in Comprehensive Organic Synthesis; ed. B. M. Trost and
I. Fleming, Pergamon, Oxford, 1991, vol. 4, pp. 779–831.
8 Y. L. Chow, W. C. Danen, S. F. Nelsen and D. H. Rosenblatt, Chem.
Rev., 1978, 78, 243.
3
3
2
2
9
X. Zhang, S. R. Ye, M. Freccero, A. Albini, D. E. Falvey and
P. S. Mariano, J. Am. Chem. Soc., 1994, 116 4211.
1
1
0 S. Das and C. von Sonntag, Z. Naturforsh., 1986, 416, 505.
1 E. Hasegawa, T. Kato, T. Kitazume, K. Yanagi, K. Hasegawa and
T. Horaguichi, Tetrahedron Lett., 1996, 37, 7079 and references cited
therein.
3 2 2
This was actually the case when Me SiCH NEt was replaced
11
by 1,3-dimethyl-2-phenylbenzimidazoline (DMPBI), which
gave 47% of 2d and 11% of 9.¶ Since the a-amino radical
Received in Cambridge, UK, 9th June 1997; 7/03974C
1896
Chem. Commun., 1997