Photo-induced aromatic assembly of benzocycloalka[1,2-b] furan and spiro[furan-2(3H),1′-benzocycloalkane] derivatives

Article abstract of DOI:10.1016/j.tetlet.2005.08.134

The photo-induced benzannulation of benzocycloalka[1,2-b]furans produced hydrohelicene-type compounds in good yields. A similar photoreaction of the spiro[furan-2(3H),1′-benzocycloalkane]s afforded dihydrophenalene derivatives in moderate yields. Benzo[kl]xanthene was also formed by a similar photoreaction of spiro[furan-2(3H),9′-xanthene]. The reaction pathway was elucidated by the ¹H NMR spectrum of the reaction mixture and the alternative synthesis of the intermediate.

Full text of DOI:10.1016/j.tetlet.2005.08.134

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 7303–7306
Photo-induced aromatic assembly of benzocycloalka[1,2-b]furan
and spiro[furan-2(3H),10-benzocycloalkane] derivatives
Reika Fujino,a Shougo Kajikawaa and Hiroshi Nishinob,*
aDepartment of Materials and Life Science, Graduate School of Science and Technology, Kumamoto University,
Kurokami 2-39-1, Kumamoto 860-8555, Japan
bDepartment of Chemistry, Faculty of Science, Kumamoto University, Kurokami 2-39-1, Kumamoto 860-8555, Japan
Received 10 August 2005; revised 24 August 2005; accepted 26 August 2005
Available online 13 September 2005
Abstract—The photo-induced benzannulation of benzocycloalka[1,2-b]furans produced hydrohelicene-type compounds in good
yields. A similar photoreaction of the spiro[furan-2(3H),10-benzocycloalkane]s afforded dihydrophenalene derivatives in moderate
yields. Benzo[kl]xanthene was also formed by a similar photoreaction of spiro[furan-2(3H),90-xanthene]. The reaction pathway was
1
elucidated by the H NMR spectrum of the reaction mixture and the alternative synthesis of the intermediate.
ꢀ 2005 Elsevier Ltd. All rights reserved.
There are many reports of the photochemical reaction of
furan derivatives.1 Recently, we reported that the irradia-
tion of 3-acetyl-5,5-diaryl-2-methyl-4,5-dihydrofurans
using a high-pressure mercury lamp gave naphthalene
derivatives in quantitative yields (Scheme 1).2 It could
be construed that this photoreaction presumably pro-
ceeded through the photoactivation of the a,b-unsatu-
rated carbonyl group, followed by an intramolecular
cyclization, and successive aromatization. In addition,
the photo-induced benzannulation was accelerated by
an acid catalyst, such as hydrochloric acid. In light of
the importance of the aromatic assembly in organic syn-
thesis, we applied the reaction to benzocycloalka[1,2-b]-
furans and spirofurans, with the aim of constructing the
polycyclic aromatic framework. Moreover, the other
objective was to clarify the reaction pathway of the pho-
tochemical reaction by the isolation of the intermedi-
ates. Herein, we briefly present these results.
Benzocyclopenta[1,2-b]furan (1a)3 (0.2 mmol) was irra-
diated for 6 h in acetonitrile (9 mL) containing 2 M
hydrochloric acid (1 mL) under an argon atmosphere
at room temperature using a 100 W high-pressure mer-
cury lamp, giving 6-acetyl-5-methyl-7H-benzo[c]fluorene
(2a) in good yield (Scheme 2 and Table 1, entry 1).4,5
Me
Me
Ac
O
h
ν
Ac
MeCN
2M HCl
argon
n
n
Me
R
O
a
: n = 1
1a-c
2a-c
h
ν
b : n = 2
c : n = 3
O
Me
Me
Me
R
MeCN
2M HCl
3 h
Me
n
O
h
ν
Me
Ac
(97%)
O
Ac
MeCN
2M
HCl
argon
Scheme 1.
n
Keywords: Photoreaction; Benzannulation; Aromatic assembly; Hydro-
helicenes; Dihydrophenalenes.
: n = 1
: n = 2
d
e
1d,e
2d,e
*
Corresponding author. Tel./fax: +81 96 342 3374; e-mail: nishino@
sci.kumamoto-u.ac.jp
Scheme 2.
0040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.08.134

7304
R. Fujino et al. / Tetrahedron Letters 46 (2005) 7303–7306
R2
R1
O
R2
Table 1. Acid-catalyzed photoreaction of benzocycloalka[1,2-b]furan
derivatives 1a–ea
R1
R3
h
ν
R3
n
Entry Benzocycloalkafuran Time
(h)
Recovery Product
(%)
(yield/%)b
MeCN
2M HCl
argon
1
2
3
4
5
1a (n = 1)
1b (n = 2)
1c (n = 3)
1d (n = 1)
1e (n = 2)
6
6
3
6
11
13
16
18
2a (76)
2b (61)
2c (45)
2d (52)
i.m.c
n
4a-e,d'
3a-h
1
2
3
a
b
c
d
e
f
g
h
d'
: R = Me, R = Ac, R = H, n = 2
: R = Me, R = Ac, R = Ph, n = 2
: R = Me, R = Ac, R = H, n = 3
: R = Me, R = CO2Me, R = H, n = 2
: R = Me, R = CO2Et, R = H, n = 2
: R = Me, R = Ac, R = H, n = 1
: R = Ph, R = Ac, R = H, n = 2
: R = Me, R = Bz, R = H, n = 2
1
2
3
5 days 48
1
2
3
a The photoreaction of benzocycloalkafuran 1 (0.2 mmol) was carried
out at 23 ꢁC in acetonitrile (9 mL) containing 2 M hydrochloric acid
(1 mL) using a 100 W high-pressure mercury lamp.
b The yield based on the benzocycloalkafuran 1 used.
c An intractable mixture was obtained without the recovery of 1e, and
the desired 2e was not isolated.
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
: R = OH, R = Ac, R = H, n = 2
R2
R1
R2
R1
Ph
Me
h
ν
Me
Ph
O
Ac
Ac
O
MeCN
2M HCl
argon
hν
O
O
MeCN
2M HCl
4 h
3i,j
4i,j,j'
i : R1 = Me, R2 = Ac
(94%)
j : R1 = Me, R2 = CO2Me
j': R1 = OH, R2 = Ac
1f
2f
Scheme 3.
Scheme 4.
Although the starting 1a was recovered after irradiation
(11% recovery), prolonged irradiation did not effectively
improve the yield. The photoreaction of other benzocy-
cloalka[1,2-b]furan derivatives 1b–e3 was next examined
under similar reaction conditions and the corresponding
benzannulated angular products 2b–d were obtained
except for the reaction of 1e (Scheme 2 and Table 1).
Surprisingly, the naphthalene-substituted tetrahydro-
naphtho[1,2-b]furan 1e resisted the photoreaction even
for a 5-day UV exposure and, ultimately, the desired
dihydro[5]helicene 2e was not isolated (entry 5), presum-
ably due to the steric hindrance between the naphtha-
lene ring and the dihydronaphthalene ring. However, a
similar reaction of structurally crowded dihydrofuran
such as 1f gave 4-(1-naphthyl)phenanthrene 2f in 94%
yield (Scheme 3).
Table 2. Acid-catalyzed photoreaction of spiro[furan-2(3H),10-ben-
zocycloalkane] derivatives 3a–ja
Entry Spirofuran Time Recovery Product
(h)
(%)
(yield/%)b
1
2
3
4
5
6
7
8
3a
3a
3a
3a
3b
3c
3d
3e
3f
1
2
3
6
3
20
7
4
1
5
4a (15)
4a (25)
4a (31)
4a (31)
4b (62)
4c (44)
4d (trace) 4d0 (4)
4e (trace) 4d0 (4)
i.m.c
4
0
18
12
6
14
12
4
9
10
11
12
13
3g
3h
3i
3
3
6
24
20
40
0
3h (36)
3g (21)
4i (12)
4j (trace) 4j0 (trace)
3j
0
We next examined a similar photoreaction of spiro-
furans 3a–j.3 Spiro[furan-2(3H),10-benzocyclohexane]
(3a) (0.2 mmol) was irradiated for 1–6 h in acetonitrile
(9 mL) containing 2 M hydrochloric acid (1 mL) under
an argon atmosphere at room temperature, giving 5-ace-
tyl-6-methyl-2,3-dihydrophenalene (4a) (Scheme 4 and
Table 2, entries 1–4).4,6 The spirofuran 3a was trans-
formed by the UV irradiation for approximately 3 h into
4a (entry 3). Prolonged irradiation did not change the
product distribution (entries 4) similar to the photoreac-
tion of 1a. Therefore, the product 4a might be stable
against UV light and 3 h should be enough to complete
the photoreaction of the spirofuran 3a. The use of a
triplet photosensitizer, such as benzophenone, did not
influence the photoreaction at all. Other spiro[furan-
2(3H),10-benzocycloalkane] derivatives 3b–j3 were next
examined under similar reaction conditions (Scheme 4
and Table 2). Although the irradiation of 3b gave the
a The photoreaction of spirofuran 3 (0.2 mmol) was carried out at
23 ꢁC in acetonitrile (9 mL) containing 2 M hydrochloric acid (1 mL)
using a 100 W high-pressure mercury lamp.
b The yield based on the spirofuran 3 used.
c An intractable mixture was obtained and the desired 4f was not
isolated.
corresponding dihydrophenalene 4b in good yield (entry
5), other spirofurans afforded the benzannulation prod-
ucts in moderate to poor yields (entries 6–8, 12, and 13).
Longer reaction times were required to consume the spi-
rofuran esters 3d and 3e. However, a small amount of
the unusual 6-hydroxydihydrophenalene 4d0 was only
isolated along with a trace amount of the desired dihydro-
phenalene esters 4d and 4e (entries 7 and 8). The struc-
ture of 4d0 was elucidated by a spectroscopic method
and finally X-ray crystallography.7 Presumably, the

R. Fujino et al. / Tetrahedron Letters 46 (2005) 7303–7306
7305
photodegradation of the ester group might complicate
Me
Ph
O
O
the reaction. Spiro[furan-2(3H),10-(20,30-dihydroindene)]
3f was decomposed under the photoreaction conditions,
and no products were isolated (entry 9). Very interest-
ingly, the irradiation of 3g for 3 h gave 3h in 36% yield
along with the unchanged 3g (20% recovered) (entry 10).
On the other hand, a similar irradiation of 3h afforded
3g in 21% yield together with the unchanged 3h (40%
recovered) (entry 11).8 These results indicated that the
spirofuran was opened once by the UV irradiation,
and recyclized to give the corresponding isomeric spiro-
furan. That is to say, it supported the assumption that
the intermediate formed from 3g and 3h should be
the same compound. In addition, it appears that the
spirofuran 3h is more stable than 3g under the same
conditions, therefore, the equilibrium of the photo-
isomerization lies with the spirofuran 3h (3g:3h =
1:1.8–1.9). In fact, the result was also supported by the
PM3 calculation.9 2-Acetyl-3-methyl-1-phenyl-7-oxa-
7H-benz[de]anthracene (4i) was also obtained by the
photoreaction of 4-acetyl-5-methyl-3-phenylspiro[furan-
2(3H),90-xanthene] (3i) (entry 12). However, the yield of
4i was lower than that of the other photoreaction prod-
ucts 4a–c. Presumably, it might be difficult to cyclize the
bulky intermediate. The single crystals of 3i and 4i were
measured by X-ray diffraction and the exact structures
were determined.7 In comparison with these data,
it was apparent that the UV irradiation caused a
rearrangement of the dihydrofuran ring of 3i into the
new aromatic ring of 4i. The result of the reaction of
the spiro[furan,xanthene] ester 3j was similar to that of
esters 3d and 3e (entry 13).
O
O
Ph
Me
Ph
Ph
h
ν
Ph
Ph
7
MeCN
2M HCl
argon
7'
H
O
O
H
Me
Ph
Ph
C
MeCN
2M HCl
argon
h
ν
Ph
O
Me
Ph
O
+
Me
Ph
Ph
8
8'
From 7 : 8 (28%) + 8' (70%)
From 7': 8 (24%) + 8' (73%)
Scheme 6.
(E)-A is ca. 3 kcal/mol more stable than (Z)-B based on
the PM3 calculation, the intermediate (Z)-B would be
cyclized by the acid-catalyzed photoreaction, followed
by aromatization accompanying dehydration to give
4a. On the contrary, in the case of 3b (R1 = Me,
R2 = Ac, R3 = Ph), the intermediate corresponding to
B (R3 = Ph) should be favorable for the intermediate,
such as A (R3 = Ph), due to the steric repulsion between
the R3 phenyl group and the tetrahydronaphthalene
ring,10 which allowed 4b to be preferentially produced.
Unfortunately, the intermediates, such as A and B,
could not be isolated from the reaction mixture, but
the corresponding ring-opened dihydronaphthalene 5
was alternatively obtained in 73% yield by the acid-cat-
alyzed decomposition of 1b (Scheme 5).3,11 The photo-
reaction of 5 gave 2b in a better yield (81%) than that
of the direct irradiation of 1b (61%) and, furthermore,
the reaction time required only 3 h (compared with
Table 1, entry 2). Naturally, stirring 5 under similar
reaction conditions without UV irradiation did not give
any products. Therefore, it strongly supported the sup-
position that the furan ring-opened products, such as
A, B, or 5, must be the intermediate in the photoreac-
tion. In addition, the UV irradiation is essential for
the acceleration of both the ring-opening and recycliza-
tion step.2 In fact, the results of the photoreaction of 7
and 70 could also be explained by the supposition, that
is, both naphthalenes 8 and 80 were obtained in a ratio
of 3:7 during a similar ring-opened intermediate C
(Scheme 6).
1
In order to examine the photoreaction pathway, the H
NMR spectrum of the reaction mixture of 3a was taken
just after the 1 h irradiation. As a result, it showed the
presence of an enol hydrogen (d = 16.5 ppm) derived
from the corresponding furan ring-opened intermediates
(E)-A and (Z)-B (Scheme 5). Although the intermediate
O
Me
R3
O
H
Me
R3
H
O
h
ν
h
ν
O h
ν
3a
3b
4a
4b
Me
Me
MeCN
2M HCl
argon
MeCN
2M HCl
argon
a: R3 = H
(E)-A
-60.7 kcal/mol
(Z)-B
-57.4 kcal/mol
b: R3 = Ph
Me
Ph
O
H
O
h
ν
p-TsOH
1b
2b
MeCN
2M HCl
argon
3 h
dry benzene
reflux
Me
36 h
5
(73%)
(81%)
Me
Ac
DDQ
It was concluded that the photoreaction of the benzo-
cycloalka[1,2-b]furans 1 and spirofurans 3 in the
presence of acid resulted in the hydrohelicene 2 and
dihydrophenalene derivatives 4, respectively. The reac-
tion pathway was also explained by the photo-induced
cleavage of the carbon–oxygen bond of the furan ring
dry benzene
reflux, 3 d
(80%)
6
Scheme 5.

7306
R. Fujino et al. / Tetrahedron Letters 46 (2005) 7303–7306
mercury lamp (Eikohsha EHB-WI-100 lamp) for 6 h using
a merry-go-round apparatus. After the irradiation, the
mixture was extracted with dichloromethane and the
obtained products were separated by silica gel TLC
(Wakogel B-10) with diethyl ether–hexane (2:8 v/v) as
the developing solvent. The analytical sample was further
purified by recrystallization from ethanol.
assisted by acid, giving the intermediate, such as (Z)-B,
which would be intramolecularly cyclized by the photo-
excitation, accompanied by the dehydration to provide
the final aromatic aggregates. In this connection, the
dehydrogenation of 2b was carried out using DDQ in
dry benzene for 3 days to give the [4]helicene 6 having
an acetyl and a methyl functionality in 80% yield
(Scheme 5).12
5. 6-Acetyl-5-methyl-7H-benzo[c]fluorene (2a): Rf = 0.76
(7:3 diethyl ether–hexane); yellow prisms (from ethanol),
mp 104.0–105.0 ꢁC; IR (neat) m 1703 (C@O); 1H NMR
(CDCl3) d 8.61 (1H, d, J = 8.45 Hz, arom H), 8.19 (1H, d,
J = 7.89 Hz, arom H), 7.98 (1H, d, J = 8.45 Hz, arom H),
7.58–7.13 (5H, m, arom H), 3.73 (2H, s, CH2), 2.54 (3H, s,
CH3), 2.52 (3H, s, COCH3); 13C NMR (CDCl3) d 207.1
(C@O), 143.6, 142.0, 137.4, 136.8, 135.3, 132.3, 129.6,
128.0 (arom C), 126.9, 126.8, 125.9, 125.6, 125.2, 124.7,
124.2, 122.7 (arom CH), 36.4 (C-7), 32.3 (COCH3), 16.4
(CH3); Anal. Calcd for C20H16O: C, 88.20; H, 5.92.
Found: C, 88.40; H, 5.92.
6. 5-Acetyl-6-methyl-2,3-dihydrophenalene (4a): Rf = 0.37
(2:8 diethyl ether–hexane); colorless prisms (from etha-
nol), mp 55.5–56.0 ꢁC; IR (CHCl3) m 1685 (C@O); 1H
NMR (CDCl3) d 7.97 (1H, d, J = 8.64 Hz, arom H), 7.46
(1H, dd, J = 8.64, 6.98 Hz, arom H), 7.32–7.27 (2H, m,
arom H), 3.10 (4H, dd, J = 11.94, 5.88 Hz, CH2 · 2), 2.72
(3H, s, CH3), 2.63 (3H, s, COCH3), 2.12–2.02 (2H, m,
CH2); 13C NMR (CDCl3) d 204.8 (C@O), 136.8, 136.7,
134.4, 133.2, 130.9, 130.8 (arom C), 126.1, 125.4, 123.3,
121.8 (arom CH), 31.5 (C-3), 31.4 (C-1), 31.0 (COCH3),
23.1 (C-2), 15.8 (CH3); MS m/z (rel intensity), 224
(M+, 63), 210 (17), 209 (100), 181 (39), 165 (48), 152
(18), 76 (13), 43 (17). Anal. Calcd for C16H16O: C, 85.68;
H, 7.19. Found: C, 85.58; H, 7.30.
Acknowledgement
This research was supported by Grants-in-Aid for Scien-
tific Research, Nos. 13640539 and 15550039, from the
Japan Society for the Promotion of Science.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2005.08.134.
References and notes
1. (a) Lablache-Combier, A. In Photochemistry of Hetero-
cycloic Compounds; Buchardt, O., Ed.; Wiley: New York,
1976; pp 123–206; (b) Wasserman, H. H.; Lipshutz, B. H.
In Singlet Oxygen; Wasserman, H. H., Ed.; Academic:
New York, 1979; pp 429–509; (c) Couture, A.; Delevallee,
7. The X-ray crystallographic data of 3i, 4d0, and 4i have
been deposited as supplementary publication numbers
CCDC249342, CCDC251416, and CCDC251415, respec-
tively. Copies of the data can be obtained, free of charge,
on application to CCDC, 12 Union Road, Cambridge,
CB2 1EZ, UK [fax: +44(0) 1223 336033 or e-mail:
deposit@ccdc.cam.ac.uk].
8. The blank experiment of 3h was conducted in the same
reaction media for 36 h in the dark to afford 3g in only 9%
yield.
´
´
A.; Lablache-Combier, A.; Parkanyl, C. Tetrahedron 1975,
31, 785–796; (d) Hiraoka, H. Tetrahedron 1973, 29, 2955–
2961; (e) Takamatsu, K.; Ryang, H.-S.; Sakurai, H.
J. Org. Chem. 1976, 41, 541–543; (f) Ditto, S. R.; Davis,
P. D.; Neckers, D. C. Tetrahedron Lett. 1981, 22, 521–524;
(g) Kirby, N. V.; Reid, S. T. J. Chem. Soc., Chem.
Commun. 1980, 150–151; (h) Kuhn, H. J.; Gollnick, K.
Chem. Ber. 1973, 106, 674–696; (i) Schreiber, S. L.;
Desmaele, D.; Porco, J. A., Jr. Tetrahedron Lett. 1988, 29,
6689–6692; (j) Carless, H. A. J.; Halfhide, A. F. E. J.
Chem. Soc., Perkin Trans. 1 1992, 1081–1082; (k) Carless,
H. A. J.; Haywood, D. J. J. Chem. Soc., Chem. Commun.
1980, 1067–1068.
9. The heats of formations for 3g and 3h were revealed to be
ꢀ21.7 and ꢀ22.7 kcal/mol, respectively, based on the
PM3 calculations. The PM3 calculations were performed
by CAChe Work System version 4.9.3.
2. Nishino, H.; Kajikawa, S.; Hamada, Y.; Kurosawa, K.
Tetrahedron Lett. 1995, 36, 5753–5756.
10. The ring-opened intermediate corresponding to
B
(R3 = Ph) from 3b was ca. 12 kcal/mol more stable than
the intermediate, such as A (R3 = Ph), according to the
PM3 calculation.
3. Fujino, R.; Nishino, H. Synthesis 2005, 731–740.
4. The benzocyclopenta[1,2-b]furan (1a) or spirofuran 3a
(0.2 mmol) was added to a mixture of acetonitrile (9 mL)
and 2 M hydrochloric acid (1 mL) in a 10 mL quartz
irradiation tube previously purged with argon for a period
of 1 min. The tube was sealed with a glass stopper. The
solution was then irradiated by a 100 W high-pressure
11. Kajikawa, S.; Nishino, H.; Kurosawa, K. Tetrahedron
Lett. 2001, 42, 3351–3354.
12. (a) Carreno, M. C.; Garcˆıa-Cerrada, S.; Urbano, A. J.
˜
Am. Chem. Soc. 2001, 123, 7929–7930; (b) Gingras, M.;
Dubois, F. Tetrahedron Lett. 1999, 40, 1309–1312.