Triquinanes from linear ketones via trimethylenemethane diyls

Article abstract of DOI:10.1016/j.tetlet.2006.12.099

Linear compounds containing a ketone and diene functional groups reacted with the anion of TMS-diazomethane to produce alkylidene carbenes that underwent intramolecular cyclopropanation followed by the formation of trimethylenemethane diyls which underwent [2+3] cycloaddition reaction to produce linearly fused triquinanes.

Full text of DOI:10.1016/j.tetlet.2006.12.099

Tetrahedron Letters 48 (2007) 1407–1410
Triquinanes from linear ketones via trimethylenemethane diyls
Hee-Yoon Lee,* Won-Yeob Kim and Sejin Lee
Center for Molecular Design and Synthesis, School of Molecular Science (BK21) and Department of Chemistry,
Korea Advanced Institute of Science and Technology (KAIST), Daejon 305-701, Republic of Korea
Received 22 November 2006; revised 14 December 2006; accepted 18 December 2006
Available online 20 December 2006
Abstract—Linear compounds containing a ketone and diene functional groups reacted with the anion of TMS–diazomethane to
produce alkylidene carbenes that underwent intramolecular cyclopropanation followed by the formation of trimethylenemethane
diyls which underwent [2+3] cycloaddition reaction to produce linearly fused triquinanes.
Ó 2006 Elsevier Ltd. All rights reserved.
Recently, we have reported a tandem cycloaddition
reaction of alkylidene carbenes of linear substrates into
triquinane compounds through the sequential formation
of alkylidene carbenes followed by trimethylenemethane
(TMM) diradical intermediates.1 As shown in Scheme 1,
alkylidene carbenes were generated from epoxyaziridi-
nyl imines that underwent intramolecular cyclopropana-
tion reaction to form highly strained intermediate 3.
Then methylenecyclopropane ring opened to trimethyl-
enemethane (TMM) diyl 4 that underwent [2+3] cyclo-
addition reaction to form linearly fused triquinanes
regio- and stereo-selectively.
Epoxyaziridinyl imine was chosen for the precursor of
alkylidene carbenes among several candidates because
the reaction condition was neutral and a relatively high
reaction temperature would guarantee the formation of
TMM diyls from cyclopropane rings.2 Since other meth-
ods of generating alkylidene carbenes3 have their own
advantages, we decided to examine another way to gen-
erate alkylidene carbene for the synthesis of triquinanes.
Since alkylidene carbenes can be generated from the
reaction of ketones with the anion of TMS–diazo-
methane,3d compounds with properly located olefins
and a ketone 6 can undergo tandem cycloaddition reac-
tion when reacted with the anion of TMS–diazomethane
(Scheme 2).
R
R
H
H
110 oC
A
A
H
O
toluene
N
Ph
N
R
OH
1
5
R
R
H
H
H
TMS
N2
A
A
O
A
A
R
R
A
A
6
10
H
OH
OH
HO
3
2
4
R
R
R
A
A
Scheme 1.
CN2
H
Keywords: Alkylidene carbene; Trimethylenemethane; Diyl;
Triquinane.
7
8
9
*
Corresponding author. Tel.: +82 42 869 2835; fax: +82 42 869
8370; e-mail: leehy@kaist.ac.kr
Scheme 2.
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.12.099

1408
H.-Y. Lee et al. / Tetrahedron Letters 48 (2007) 1407–1410
When compared to the aziridinylimine route,1 the cur-
rent route has advantages of being very high in atom
economy4 and having readily available substrates
through straightforward synthesis. However, the current
route might show possible complications due to the
basic nature of the reaction condition and instability
of the anion at a higher temperature than 0 °C.
of the substrates in method B. In method A, the gener-
ation of TMSCLiN2 before the addition of the ketone
minimizes the reaction of the ketone with BuLi but
might decompose before the desired reaction proceeds
in completion. In method B, it was hoped that
8
TMSCKN2 would react with ketones before decompo-
sition even at a reaction temperature higher than
À30 °C. Contrary to what we hoped for method B was
not any better than method A. It was presumed that
the lower concentration of the anion in method B than
in method A and different counter cations in two meth-
ods made method B less effective than we had expected.
The result was summarized in Table 1.
Substrates for the tandem reaction were prepared from
commercially
available
6-methyl-5-hepten-2-one
(Scheme 3). After protection of the ketone as the corre-
sponding acetal, isopropenyl group was replaced with
the allylic bromide in a four step sequence. Ozonolysis
followed by Wittig olefination with Ph3PCHCOOEt
produced unsaturated ester (I). DIBAL–H reduction
of the ester to the alcohol followed by bromination reac-
tion produced corresponding allylic bromide. This
allylic bromide was reacted with malonate anion
followed by subsequent allylation to afford ketodienes
(11, 13) after the hydrolysis of acetals. Phenyl group
was introduced to the terminal olefin to examine the
stereoselectivity and reactivity of the TMM diyl [2+3]
cycloaddition reaction.5 Compounds with reduced esters
(12, 14) were also prepared to minimize possible side
reactions due to the nucleophilic nature of the reaction
conditions for alkylidene carbene generation.
The reactions produced single major products9 whose
structures were determined unambiguously by NMR
through comparison to the previous reports.1,10 As ex-
pected, substrates without the electrophilic substituents
(12, 14) yielded a better result than carbonyl containing
ones (11, 13) and phenyl groups attached to the terminal
olefins (13, 14) provided better result than the unsubsti-
tuted ones (11, 12). The effect of the substituents on the
[2+3] cycloaddition reaction was bigger than the effect
observed in the previous report1 probably due to a lower
reaction temperature for a better selectivity. The basic
nature of the reaction and the instability of the anion
of TMSCHN2 probably was the reason for the low yield
of the products.
These substrates were subjected to a modified Shioiri’s
reaction conditions6 to generate alkylidene carbene from
ketones and TMSCLiN2.7 To ensure the transformation
of methylenecyclopropane ring of the intermediate to
the TMM intermediate and to minimize the decomposi-
tion of TMSCLiN2, the reaction was carried out at
À30 °C in two different ways. While TMSCLiN2 was
generated before the addition of the ketone substrates
in method A, TMSCKN2 was generated in the presence
The current synthetic strategy allowed us to examine the
effect of the methyl substituted TMM diyl to the tandem
cycloaddition reaction as the methyl group provided a
more steric interaction during the [2+3] cycloaddition
Table 1.
Substrate
Major product
Yield (%)
Method A Method B
a-c
O
O
O
H
H
EtOOC
EtOOC
I
11
21
38
34
53
25
21
27
31
EtOOC
H
d-f
15
16
g
COOEt
H
H
EtOOC COOEt
O
O
O
O
H
R
R
EtOOC
MeO
MeO
12
13
14
III
II
h-j
j
MeO OMe
Ph
O
EtOOC COOEt
H
H
O
H
R
EtOOC
EtOOC
11 (R=H)
13 (R=Ph)
12 (R=H)
14 (R=Ph)
17
Scheme 3. Reagents and conditions: (a) ethylene glycol, TsOH/PhH,
94%; (b) O3/CH2Cl2, À78 °C; PPh3; (c) Ph3P@CHCOOEt/PhH, reflux,
61% for two steps; (d) DIBAL–H/CH2Cl2, rt, 74%; (e) Ph3P–NBS/
CH2Cl2, À30 °C, 74%; (f) diethyl malonate, NaH/THF, rt, 80%; (g)
NaH/THF; RCH@CHCH2Br, rt, 84% (R = Ph), 91% (R = H); (h)
LAH/ Et2O, 0 °C, 72% (R = Ph), 77% (R = H); (i) NaH, MeI/THF, rt,
84% (R = Ph), 87% (R = H) and (j) 10% HCl (aq)/THF, rt, 82% (11),
71% (14), 84% (12, 13).
Ph
H
H
H
MeO
MeO
18

H.-Y. Lee et al. / Tetrahedron Letters 48 (2007) 1407–1410
1409
E
reaction. The precursor for the cycloaddition reaction
was readily prepared from 6-methyl-5-hepten-2-one
(Scheme 4). After the selective hydroxylation of the
allylic methyl group, the allylic alcohol was converted
into the bromide. From this allylic bromide, substrate
19 was prepared in the same way as the compounds in
Scheme 3.
E
H
E
H
H
E
Ph
H
Ph
H
R1
R1
Ph
R2
R2
Ph
E
E
E
E
A
B
20
21
When 19 was reacted with TMSCLiN2 using method A,
two major products 20 and 21 were obtained in a 54%
yield with the ratio of 3:1 (Scheme 5).
E
E
E
R2
H
R2
H
E
R1
The structures of 20 and 21 were also confirmed through
NMR spectroscopy. Product 20 has the usual cis-anti-cis
triquinane structure and 21 has the cis-syn-cis triquinane
structure. The relative stereochemistry was assigned
unambiguously by coupling constants of tertiary hydro-
gens of 21 and through the NOE experiment on the
ozonolysis product 22 (Scheme 6).
Ph
R1
Ph
C
D
Scheme 7.
has been somewhat anticipated, the relative stereochem-
istry of 21 was not obvious as the transition state for the
formation of 21 could be either the transition state B or
D in Scheme 7. For the compounds with R1 = H, inter-
action between axially oriented hydrogen of the tether
and cyclopentene ring showed a clear preference of
cycloaddition reaction product from A. That explained
why the product from transition state C or D was not
observed. For the compounds with R1 = R2 = Me, it
looked apparent that the steric interaction between
two methyl groups in transition state A was bigger than
the one in transition state B or D and it was thought that
transition state B or D might have a similar energy to
that of transition state A. Even with the added steric
interaction, however, the product from transition state
A was the major product and only the cis-syn-cis prod-
uct 21 was produced as the minor product with no other
possible isomers detected. Since 21 was the product from
transition state B rather than the transition state D, it
became clear that the general preference for the transi-
tion states to form cis-anti-cis isomers could have been
overcome by the bigger steric interaction between axial
hydrogens in the tether and the methyl group of the
cyclopentene ring in transition state D than the steric
interaction between two methyl groups in the transition
state B.
This result allowed us to rationalize the selectivity of
[2+3] diyl cycloaddition reaction among isomeric prod-
ucts through comparison of four conformations of tran-
sition states (Scheme 7). Although the formation of 21
a-c
O
O
O
Br
d, e
COOEt
COOEt
f
O
EtOOC
EtOOC
O
O
Ph
Ph
19
Scheme 4. Reagents and conditions: (a) ethylene glycol, TsOH/PhH,
94%; (b) SeO2, t-BuOOH/CH2Cl2, 0 °C; 53%; (c) Ph3P–NBS/CH2Cl2,
À30 °C, 83%; (d) diethyl malonate, NaH/THF, rt; 88%; (e) NaH/THF;
PhCH@CHCH2Br, rt, 95% and (f) 10% HCl (aq)/THF, rt, 93%.
O
H
Ph
H
Ph
Ph
TMS
-30 oC - rt
N2
H
+
In summary, we have demonstrated that the alkylidene
carbenes generated from readily available ketones and
TMSCLiN2 were suitable for the TMM diyl [2+3] cyclo-
addition reaction and offered a supplementary synthetic
strategy to the previously reported epoxyaziridinyl imine
route for the construction of triquinanes. The current re-
sult explains the selectivity of the cycloaddition reaction
of unsymmetrical TMMs and provides the valuable
information for the future synthetic design and expected
outcome for compounds with various substitution
patterns.
EtOOC
EtOOC
COOEt
EtOOC
COOEt
20
COOEt
19
21
Scheme 5.
OHC
1.4%
CH3
Ph
9.6%
Ph
H
O
O3; PPh3
H
H3C
13%
COOEt
EtOOC
COOEt
EtOOC
Acknowledgments
22
21
Scheme 6. NOE experiment.
We greatly acknowledge financial support from KOSEF
through the center of molecular design and synthesis,

1410
H.-Y. Lee et al. / Tetrahedron Letters 48 (2007) 1407–1410
(m, 1H), 1.52 (s, 3H), 1.40–1.52 (m, 1H), 1.54–1.73 (m,
2H), 1.87–1.97 (m, 1H), 2.11–2.24 (m, 1H), 2.26–2.38 (m,
1H), 2.41–2.54 (m, 2H), 2.70–2.83 (m, 2H), 3.01 (br, 1H),
3.90–4.06 (m, 4H). Compound 16: 1H NMR (200 Hz,
CDCl3); d 0.88–1.07 (m, 3H), 1.15–1.32 (m, 1H), 1.41–1.56
(m, 1H), 1.52 (s, 3H), 1.64–1.81 (m, 2H), 1.86–1.99 (m,
1H), 2.11–2.23 (m, 1H), 2.42–2.50 (m, 1H), 2.67–2.80 (m,
2H), 2.92–3.06 (m, 1H), 3.00 (s, 2H), 3.05 (s, 2H), 3.12 (s,
3H), 3.15 (s, 3H). Compound 17: 1H NMR (200 Hz,
CDCl3); d 0.76–0.96 (m, 2H), 1.12–1.39 (m, 6H), 1.62 (s,
3H), 1.60–1.72 (m, 1H), 1.84–2.09 (m, 2H), 2.34–2.52 (m,
1H), 2.53–2.79 (m, 2H), 2.80–3.03 (m, 2H), 3.05–3.19 (m,
1H), 3.57 (br, 1H), 4.10–4.22 (m, 4H), 6.85–7.24 (m, 5H).
Compound 18: 1H NMR (300 Hz, CDCl3); d 0.75–0.83
(m, 1H), 1.07–1.25 (m, 3H), 1.47–1.57 (m, 1H), 1.52 (s,
3H), 1.82–1.94 (m, 3H), 2.36–2.43 (m, 1H), 2.65–2.68 (d,
J = 7.91 Hz, 1H), 2.84–2.93 (m, 2H), 2.95–3.17 (m, 10H),
3.44 (br, 1H), 6.68–7.04 (m, 5H); 13C NMR (75 Hz, C6D6);
d 14.60, 24.47, 38.90, 39.04, 39.96, 42.46, 50.91, 51.97,
53.34, 55.89, 59.22, 59.27, 74.36, 77.51, 125.75, 127.12,
and from MarineBio21, Ministry of Maritime Affairs
and Fisheries, Korea.
References and notes
1. Lee, H.-Y.; Kim, Y. J. Am. Chem. Soc. 2003, 125, 10156.
2. (a) Platz, M. S.; Berson, J. A. J. Am. Chem. Soc. 1976, 98,
6743; (b) Duncan, C. D.; Corwin, J. A.; Davis, J. H.;
Berson, J. A. J. Am. Chem. Soc. 1980, 102, 2350.
3. (a) Kirmse, W. Angew. Chem., Int. Ed. 1997, 36, 1164; (b)
Wolinsky, J.; Clark, G. W.; Thorstenson, P. C. J. Org.
Chem. 1976, 41, 745; (c) Seyferth, O.; Marmor, R. M.;
Hilbert, P. H. J. Org. Chem. 1971, 36, 1379; (d) Shioiri, T.;
Aoyama, J. J. Synth. Org. Chem. Jpn. 1996, 54, 918; (e)
Kim, S.; Cho, C. M. Tetrahedron Lett. 1994, 35, 8405; (f)
Stang, P. J. Angew. Chem., Int. Ed. Engl. 1992, 31, 274.
4. Trost, B. M. Science 1991, 254, 1471.
5. Little, R. D.; Ott, M. M. In Studies in Natural Products
Chemistry; Atta-ur-Rahman, Ed.; Elsevier Health Science:
Amsterdam, 2000; Vol. 22, p 195.
1
128.10, 128.69, 144.95, 147.50. Compound 20: H NMR
(500 Hz, Benzene-d6); d 0.85 (t, J = 7.10 Hz, 3H), 0.94 (t,
J = 7.05 Hz, 3H), 1.13 (s, 3H), 1.49 (m, 1H), 1.58 (s, 3H),
1.84 (m, 1H), 2.18 (dd, J = 5.7, 10.1 Hz, 1H), 2.42 (m,
1H), 2.43 (d, J = 16.5), 2.45 (dd, J = 16.9, 9.5 Hz, 1H),
2.58 (dd, J = 5.7, 10.1, 1H), 2.72 (m, 1H), 2.85 (dd,
J = 5.7, 12.0, 1H), 2.88 (d, J = 16.5 Hz, 1H), 3.13 (m, 1H),
3.91 (m, 2H), 4.02 (q, J = 7.10 Hz, 2H), 7.06–7.29 (m, 5H);
13C NMR (125 Hz, Benzene-d6); d 14.60, 14.73, 15.78,
23.15, 30.61, 32.33, 33.63, 41.91, 42.50, 48.96, 49.41, 61.94,
62.05, 63.86, 64.60, 67.73, 127.10, 127.93, 129.49, 129.76,
6. Sasaki, A.; Aoyama, T.; Shioiri, T. Tetrahedron 1999, 55,
3687.
7. Method A: To a THF solution (0.02 M) of TMSCHN2
(2 equiv to the substrate) was added n-BuLi (2 equiv) at
À78 °C under argon atmosphere. The mixture was stirred
for 10 min at À78 °C and then was warmed to À30 °C
before a solution of the substrate in THF was added over
20 min period. The reaction mixture was stirred for 10 min
at À30 °C and then was allowed to warm to room
temperature for 1 h. Brine was added to the reaction
mixture and products were extracted with diethyl ether.
The ether solution was dried over MgSO4 and was
concentrated. The crude product was purified by flash
chromatography. Method B: To a solution of TMSCHN2
(2 equiv) and the substrate in THF (0.01 M) was added
KN(TMS)2 (2 equiv, 0.5 M/toluene) over 20 min at
À30 °C under argon atmosphere. Then the rest of the
reaction followed the procedure in method A.
1
129.88, 143.53, 148.86, 172.92, 173.55. Compound 21: H
NMR (300 Hz, CDCl3); d 0.80 (s, 3H), 1.15–1.23 (m, 6H),
1.25 (s, 3H), 1.58–1.67 (m, 1H), 1.90–1.99 (m, 2H), 2.27–
2.32 (m, 1H), 2.33–2.43 (m, 1H), 2.48–2.64 (m, 1H), 2.66–
2.76 (m, 3H), 3.02–3.11 (m, 1H), 4.07–4.20 (m, 4H), 5.23–
5.26 (m, 1H), 7.15–7.29 (m, 5H); 13C NMR (75 Hz,
CDCl3); d 13.97, 14.01, 19.15, 28.93, 35.05, 38.20, 41.42,
47.76, 47.96, 58.52, 61.39, 61.48, 61.78, 61.81, 63.36,
116.62, 126.01, 127.29, 128.06, 128.22, 140.14, 165.66,
172.24, 172.65.
8. TMSCKN2 instead of TMSCLiN2 gave the better result
for the tandem reaction sequence.
9. Spectral data of 16–21. Compound 15: 1H NMR (200 Hz,
CDCl3); d 0.92–1.08 (m, 1H), 1.04–1.15 (m, 6H), 1.16–1.29
10. Little, R. D.; Muller, G. W. J. Am. Chem. Soc. 1981, 103,
2744.