A synthetic alternative to the type-II intramolecular 4 + 3 cycloaddition reaction

Article abstract of DOI:10.1021/jo034356f

Oxyallyl cations generated in situ from dihaloketones have been found to undergo the Favorskii rearrangement in preference to an intramolecular Type-II 4 + 3 cycloaddition reaction in trifluoroethanol and hexafluoropropan-2-ol solvents, generating acrylat

Full text of DOI:10.1021/jo034356f

A Syn th etic Alter n a tive to th e Typ e-II
In tr a m olecu la r 4 + 3 Cycloa d d ition
Rea ction
Richard S. Grainger,* Richard B. Owoare,
Patrizia Tisselli, and J onathan W. Steed†
Department of Chemistry, King’s College London,
Strand, London WC2R 2LS
richard.grainger@kcl.ac.uk
Received March 18, 2003
Abstr a ct: Oxyallyl cations generated in situ from diha-
loketones have been found to undergo the Favorskii rear-
rangement in preference to an intramolecular Type-II 4 +
3 cycloaddition reaction in trifluoroethanol and hexafluoro-
propan-2-ol solvents, generating acrylate esters with high
cis selectivity. A synthetic alternative to the intramolecular
Type-II 4 + 3 cycloaddition has been developed on the basis
of intramolecular enolate alkylation to close a 7-membered
ring with an exocyclic double bond.
F IGURE 1. Intramolecular 4 + 3 cycloaddition reactions of
allylic cations.
SCHEME 1
The 4 + 3 cycloaddition reaction between a diene and
an allylic cation is a powerful method for the synthesis
of 7-membered rings.1 In the intramolecular series,
Harmata has delineated six possible ways of linking the
diene with the allylic cation (Figure 1).2 To date, only
examples of the Type-I system have been reported in the
literature.
As part of a synthetic approach to the diterpene
ingenol, we have investigated the Type-II intramolecular
4 + 3 cycloaddition reaction of an oxyallyl cation tethered
to the 3-position of a furan as a means to construct the
[4.4.1] bicyclic skeleton (BC ring system) of the natural
product (Scheme 1).3 We envisage that stereoselective
reduction of the bridgehead double bond system 1 offers
a means to access the thermodynamically unfavorable
trans ring junction (in-out stereochemistry)4 deemed
essential for the biological activity of ingenol.5,6 Here, we
report our attempts to achieve such a transformation, the
problems associated with such an approach, and an
alternative, more successful method for the construction
of the desired ring system 1 based on intramolecular
enolate alkylation.
zation are shown in Scheme 2. Homologation of 3-furfuryl
alcohol 27 was carried out via a known three-step
sequence.8 Conversion of alcohol 3 to the corresponding
bromide followed by displacement with dimethylmalonate
and Krapcho decarboxylation gave monoester 4, which
could be converted into the dichloroketone 5 or dibro-
moketone 6 using Barluenga’s methodology.9
(5) Total syntheses of ingenol: (a) Winkler, J . D.; Rouse, M. B.;
Greaney, M. F.; Harrison, S. J .; J eon, Y. T. J . Am. Chem. Soc. 2002,
124, 9726. (b) Tanino, K.; Onuki, K.; Asano, K.; Miyashita, M.;
Nakamura, T.; Takahashi, Y.; Kuwajima, I. J . Am.Chem. Soc. 2003,
125, 1498. For a review of approaches to the synthesis of ingenol, see:
Kim, S.; Winkler, J . D. Chem. Soc. Rev. 1997, 26, 387. For more recent
work, see: (c) Nakamura, T.; Matsui, T.; Tanino, K.; Kuwajima, I. J .
Org. Chem. 1997, 62, 3032. (d) Rigby, J . H.; Hu, J .; Heeg, M. J .
Tetrahedron Lett. 1998, 39, 2265. (e) Winkler, J . D.; Kim, S.; Harrison,
S.; Lewin, N. E.; Blumberg, P. M. J . Am. Chem. Soc. 1999, 121, 296.
(f) Kigoshi, H.; Suzuki, Y.; Aoki, K.; Uemura, D. Tetrahedron Lett. 2000,
41, 3927. (g) Tang, H.; Yusuff, N.; Wood, J . L. Org. Lett. 2001, 3, 1563.
(h) Rigby, J . H.; Bazin, B.; Meyer, J . H.; Mohammadi, F. Org. Lett.
2002, 4, 799.
In general, oxyallyl cations are too unstable to be
isolated and are typically generated in situ. Although a
variety of precursors can be used to generate oxyallyl
cations, R-haloketones have found most application in the
chemical literature.1,2 Construction of R-dihaloketones 5
and 6 required to test the proposed Type-II cycli-
† To whom correspondence regarding the X-ray crystal structure
should be addressed.
(1) (a) Noyori, R.; Hayakawa, Y. Org. React. (N.Y.) 1983, 29, 163-
344. (b) Hoffmann, H. M. R. Angew. Chem., Int. Ed. Engl. 1984, 23, 1.
(c) Mann, J . Tetrahedron 1986, 42, 4611.
(2) (a) Harmata, M. Tetrahedron 1997, 53, 6235. (b) Harmata, M.
In Advances in Cycloaddition; Lautens, M., Ed.; J AI: Greenwich, 1997;
Vol. 4, pp 41-86.
(3) Harmata has shown the intramolecular Type-I cycloaddition
reaction to be an excellent method for the rapid assembly of the
isoingenol skeleton (cis ring junction between the B- and C-rings):
Harmata, M.; Elahmad, S.; Barnes, C. L. Tetrahedron Lett. 1995, 36,
1397.
(6) Isomerization of a cis- to trans-ring system via formation and
stereoselective reduction of a bridgehead double bond is the key feature
in Rigby’s elegant approach to ingenol: Rigby, J . H.; de Sainte Claire,
V.; Cuisiat, S. V.; Heeg, M. J . J . Org. Chem. 1996, 61, 7992. See also
ref 5d,h.
(7) Prepared by reduction of commercially available 3-furoic acid
(lithium aluminium hydride, THF, rt, 83%). Wang, E. S.; Choy, Y. M.;
Wong, H. N. C. Tetrahedron 1996, 52, 12137.
(8) New, D. G.; Tesfai, Z.; Moeller, K. D. J . Org. Chem. 1996, 61,
1578.
(9) Barluenga, J .; Llavona, L.; Concello´n, J . M.; Yus, M. J . Chem.
Soc., Perkin Trans. 1 1991, 297.
(4) Alder, R. W.; East, S. P. Chem. Rev. 1996, 96, 2097.
10.1021/jo034356f CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/09/2003
J . Org. Chem. 2003, 68, 7899-7902
7899

SCHEME 2a
ing of the alternative C-C bond in the cyclopropane
intermediate 15 or via Favorskii rearrangement of a 1,3-
dichloroketone formed via 1,3-Cl migration in 5.
Products derived from Favorskii rearrangements have
been observed as minor components in 4 + 3 cycloaddi-
tion reactions of oxyallyl cations,13 but their formation
is generally minimized by the use of the nonnucleophilic
solvent, trifluoroethanol. Recourse to the even less nu-
cleophilic solvent hexafluoropropan-2-ol suppressed the
Favorskii rearrangement of dichloroketone 5 and pro-
moted formation of the desired cycloadduct 9 (albeit in
low yield), thus establishing for the first time that the
Type-II 4 + 3 cyclization mode is possible (Table 1, entries
4 and 5). Unfortunately, application of these conditions
to the dibromoketone 6 only gave trace amounts of the
desired cycloadduct 11, with cis-alkene 10 now predomi-
nating (Table 1, entry 6).
a
Reagents and conditions: (a) PBr3, pyridine, Et2O; (b) al-
lylMgCl, THF; (c) Sia2BH, then NaOH, H2O2, 45% over three steps;
(d) CBr4, Ph3P, CH2Cl2, rt, 3 h, 89%; (e) dimethyl malonate, NaH,
DMF, THF, reflux, 24 h, 70%; (f) NaCl, DMSO, H2O, reflux, 3 h,
90%; (g) Cy2NLi, CH2Cl2, THF, -78 °C, 2 h, 64% 5; (h) LDA,
CH2Br2, THF, -78 °C, 2 h, 63% 6.
The results in Table 1 clearly demonstrate that al-
though the Type-II cyclization mode is possible, it is never
going to be synthetically viable under these reaction
conditions. Although a range of alternative substrates for
oxyallyl cation formation can be envisaged which might
avoid the problem of competing Favorskii rearrange-
ment,14 we chose instead to investigate a complementary
approach to the requisite ring system 1, involving a novel
intramolecular enolate alkylation to close a 7-membered
ring (Scheme 1).15 Analysis of simple molecular models
suggested the conformation required for an intramolecu-
lar SN2 reaction to occur is only accessible on the face of
the enolate opposite the oxygen bridge. Furthermore,
cyclization of the desired enolate appeared to be clearly
favored over its regioisomer, thereby allowing us to
potentially exploit thermodynamic conditions for enolate
generation. This strategy was brought into practice as
follows (Scheme 4). Intermolecular 4 + 3 cycloaddition
of 3-substituted furan 3 with the oxyallyl cation gener-
ated in situ from 1,1,3-trichloropropan-2-one proceeded
readily to provide alcohol 16 after dehalogenation. A
variety of bases (KH, NaH, LDA, KHMDS, Cs2CO3, KO-
t-Bu) and conditions (solvent/temperature) have been
screened for the 7-membered ring closure on both tosylate
17 and bromide 18. Potassium tert-butoxide proved
greatly superior to other basessin refluxing THF under
high dilution conditions tosylate 17 can be closed to
ketone 1 in a synthetically useful 80% yield. Notably, this
reaction proceeds without cleavage of the strained ether
bridge in 1.16 The structure of 1 has been unambiguously
proven by X-ray crystallography.17
SCHEME 3
Intramolecular Type-II 4 + 3 cycloaddition reactions
of 5 and 6 were attempted under a variety of standard
conditions, including LiClO4/Et3N/Et2O, Et3N/trifluoro-
ethanol (TFE), and NaTFE/TFE.10 Reactions proceeded
slowly at room temperature to give complex mixtures of
products. Analysis of 1H NMR of the crude reaction
mixtures suggested retention of the monosubstituted
furan functionality under almost all conditions.11 In the
case of reactions in trifluoroethanol, slow conversion to
a mixture of trifluoroethyl acrylic esters 7 and 8 was
observed, the ratio of which varied with the base used to
generate the oxyallyl cation (Table 1, entries 1-3).
The formation of 7 and 8 can be ascribed to a competing
Favorskii rearrangement in preference to the desired 4
+ 3 cycloaddition reaction (Scheme 3). The cis double-
bond geometry in 7 was assigned on the basis of coupling
constants in the 1H NMR spectrum and is consistent with
a stereospecific disrotative ring closure of an oxyallyl
cation 13 to a cis-cyclopropane 14, followed by a ste-
reospecific SN2-type ring opening via 15 to 7.12 The
formation of the 2-methylene ester 8 can be ascribed to
elimination of an R-chloromethyl ester formed via open-
In conclusion, an investigation into a Type-II intramo-
lecular 4 + 3 cycloaddition has shown the Favorskii
(13) (a) Fo¨hlisch, B.; Gehrlach, E.; Henle, G.; Boberlin, U.; Gekeler,
M.; Geywitz, B.; Ruck, M.; Vogl, H. J . Chem. Res., Miniprint 1991,
1401. (b) Reference 11. (c) Fo¨hlisch, B.; Kreiselmeier, G. Tetrahedron
2001, 57, 10077.
(14) For recent examples, see: (a) Myers, A. G.; Barbay, J . K. Org.
Lett. 2001, 3, 425. (b) Cho, S. Y.; Lee, H. I.; Cha, J . K. Org. Lett. 2001,
3, 2891. (c) Aungst, R. A., J r.; Funk, R. L. Org. Lett. 2001, 3, 3553. (d)
Xiong, H.; Hsung, R. P.; Berry, C. R.; Rameshkumar, C. J . Am. Chem.
Soc. 2001, 123, 7174. (e) Harmata, M. Acc. Chem. Res. 2001, 34, 595
and references therein. (f) Handy, S. T.; Okello, M. Synlett 2002, 489.
(g) Harmata, M.; Ghosh, S. K.; Hong, X.; Wacharasindhu, S.; Kirch-
hoefer, P. J . Am. Chem. Soc. 2003, 125, 2058.
(10) Sendelbach, S.; Schwetzler-Raschke, R.; Radl, A.; Kaiser, R.;
Henle, G. H.; Korfant, H.; Reiner, S.; Fo¨hlisch, B. J . Org. Chem. 1999,
64, 3398.
(11) Disubstituted furans can potentially arise via an intramolecular
(15) For representative examples of 7-membered ring-closure via
enolate alkylation, see: (a) Conia, J .-M.; Rouessac, F. Bull. Soc. Chim.
Fr. 1963, 1930. (b) Casadei, M. A.; Galli, C.; Mandolini, L. J . Am. Chem.
Soc. 1984, 106, 1051. (c) Spreitzer, H.; Pichler, A.; Holzer, W.; Schlager,
C. Helv. Chim. Acta 1998, 81, 40.
aromatic substitution reaction. For
a recent example, see: Kre-
iselmeier, G.; Fo¨hlisch, B. Tetrahedron Lett. 2000, 41, 1375.
(12) Schamp, N.; De Kimpe, N.; Coppens, W. Tetrahedron 1975, 31,
2081.
7900 J . Org. Chem., Vol. 68, No. 20, 2003

TABLE 1. Attem p ted Typ e-II In tr a m olecu la r 4 + 3 Cycloa d d ition Rea ction s of Dih a lok eton es 5 a n d 6
SCHEME 4a
cycloadduct have necessitated the development of an
alternative approach to the same ring system based on
intramolecular enolate alkylation under thermodynamic
conditions. Further studies on the selective functional-
ization and ring opening of oxatricyclic systems such as
1 are in progress and will be reported in due course.
Exp er im en ta l Section
In ter m olecu la r 4 + 3 Cycloa d d ition of F u r a n (3). A
solution of NaOCH2CF3 (10.7 mL, 21.4 mmol) in CF3CH2OH
(10.7 mL) and 1,3,3-trichloroacetone (2.3 mL, 2.4 mmol) were
added separately via syringe to alcohol 3 (1.5 g, 10.7 mmol) over
1 h at 0 °C. The reaction was allowed to warm to room
temperature and stirred for 3 days. The reaction was quenched
with water and extracted with diethyl ether. The organic layer
was dried (MgSO4) and concentrated to give a crude product that
was taken up in methanol (10 mL) and added to a solution of
Zn (14.8 g, 227.3 mmol) and Cu(I)Br (8.7 g, 60.6 mmol) in
methanol (130 mL). The mixture was stirred at room temper-
ature for 2 days before filtering through 10 g of Celite and the
filtrate concentrated to give a residue the was dissolved in a 5%
solution of HCl. The aqueous layer was extracted with diethyl
a
Reagents and conditions: (a) 1,1,3-trichloropropan-2-one,
NaOCH2CF3, CF3CH2OH, rt, 3 days; (b) Zn, CuBr, MeOH, 2 days,
54% over two steps; (c) TsCl, Et3N, DMAP, CH2Cl2, 0 °C to rt, 4.5
h, 64%; (d) CBr4, Ph3P, CH2Cl2, 71%; (e) t-BuOK, THF, reflux, 80%
from 17, 60% from 18.
rearrangement to be the major reaction pathway for
oxyallyl cations generated from dihaloketones under
classical conditions.18 The low yields of the Type-II
(18) Although determination of the optimal chain length between
diene and oxyallyl cation was not the aim of this study, it may well be
that other systems undergo the Type-II 4 + 3 cycloaddition with greater
ease than 5 or 6. In this respect, it is perhaps notable that to the best
of our knowledge Type-I intramolecular 4 + 3 cycloadditions are only
known with three or four atoms in the tether to form five- or
six-membered rings, respectively.
(16) Reviews on ring opening: (a) Woo, S.; Keay, B. A. Synthesis
1996, 669. (b) Chiu, P.; Lautens, M. In Topics in Current Chemistry,
Vol. 190; Metz, P., Ed.; Springer-Verlag: New York, 1997; pp 1-85.
(c) Lautens, M.; Fagnou, K.; Hiebert, S. Acc. Chem. Res. 2003, 36, 48.
(17) The author has deposited X-ray coordinates with the Cambridge
Crystallographic Data Center (CCDC 202411).
J . Org. Chem, Vol. 68, No. 20, 2003 7901

ether and once with ethyl acetate. The combined organic layer
was dried (MgSO4) and concentrated a crude residue that was
purified by column chromatography (3:1 60-80 petroleum ether/
ethyl acetate) to give 16 (807 mg, 54% over two steps) as a light
yellow oil: Rf 0.23 (3:1 ethyl acetate/60-80 petroleum ether);
2360, 1700, 1050 cm-1; δH (360 MHz; CDCl3) 1.51 (m, 4 H), 1.99
(m, 2 H), 2.23 (m, 2 H), 2.39 (s, 3 H), 2.66 (m, 2 H), 3.96 (t, J )
6.1 Hz, 2 H), 4.67 (d, J ) 4.9 Hz, 1 H), 4.90 (d, J ) 4.5 Hz, 1 H),
5.71 (d, J ) 1.8 Hz, 1 H), 7.36 (d, J ) 8.1 Hz, 2 H), 7.30 (d, J )
8.3 Hz, 2 H); δC (90 MHz, CDCl3) 22.1 (CH3), 23.6 (CH2), 26.8
(CH2), 28.8 (CH2), 46.5 (CH2), 70.4 (CH2), 78.0 (CH), 79.4 (CH),
126.3 (C), 128.3 (CH), 130.3 (CH), 133.4 (C), 145 (C), 147.8 (C),
206.0 (C); m/z (EI) 350 (M+; 100), 178 (22), 135 (67), 121 (18);
HRMS calcd for C14H14O3F6S 350.1188, found 350.1172.
Oxa tr icyclic Keton e (1). To a stirred solution of KO-t-Bu
(50 mg, 0.45 mmol) in THF (38 mL) was added a solution of 17
(120 mg, 0.34 mmol) in THF (110 mL). The mixture was refluxed
for 1 h, and then a further 1.3 equiv of KOtBu (50 mg, 0.45 mmol)
was added in one portion. The resulting mixture was refluxed
for 30 min, by which time the reaction was seen to have gone to
completion (by TLC). The reaction was quenched with saturated
aqueous NH4Cl, followed by extraction of the aqueous layer with
diethyl ether. The ethereal layer was dried (MgSO4) and con-
centrated in vacuo to afford compound 1 (47 mg, 80%) as a yellow
solid: mp 72-73 °C; Rf 0.42 (3:1 60-80 petroleum ether/ethyl
acetate); νmax (neat) 2932, 2858, 1704, 1046 cm-1; δH (360 MHz;
CDCl3) 1.14 (m, 4 H), 1.50 (m, 2 H), 1.70 (m, 2 H), 2.15 (m, 1 H),
2.26 (m, 2 H), 2.56 (dd, J ) 15.8, 4.6 Hz, 1 H), 2.74 (m, 1 H),
4.72 (d, J ) 4.4 Hz, 1 H), 4.97 (t, J ) 1.6 Hz, 1 H), 5.64 (s, 1 H);
δC (90 MHz, CDCl3) 23.7 (CH2), 26.1 (CH2), 26.6 (CH2), 38.1
(CH2), 46.27 (CH2), 56.2 (CH), 78.9 (CH), 82.4 (CH), 126.9 (CH),
147.5 (CH), 207.6 (C); m/z (EI) 178 (M+; 100), 149 (24), 121 (41),
79 (27); HRMS calcd for C11H14O2 179.1072, found 179.1073.
ν
max (neat) 3430 br, 2940, 1708 cm-1; δH (360 MHz; CDCl3) 1.52
(m, 4 H), 1.82 (br, 1 H), 2.08 (br, 2 H), 2.27 (m, 2 H), 2.62 (m, 2
H), 3.57 (t, J ) 5.3 Hz, 2 H), 4.72 (d, J ) 4.9 Hz, 1 H), 4.92 (d,
J ) 4.5 Hz, 1 H), 5.76 (d, J ) 1.6 Hz, 1 H); δC (90 MHz, CDCl3)
23.9 (CH2), 27.3 (CH2), 32.6 (CH2), 46.5 (CH2), 46.6 (CH2), 62.9
(CH2), 78.0 (CH), 80.0 (CH), 126.1 (CH), 148.0 (C), 206.4 (C);
m/z (EI) 196 (M+; 97.3), 178 (43), 153 (92), 95 (100); HRMS calcd
for C11H16O3Na (M + Na+) 219.0992, found 219.0989.
Br om id e (18). To a solution of alcohol 16 (1.0 g, 5.1 mmol)
and carbon tetrabromide (2.1 g, 6.3 mmol) in CH2Cl2 (30 mL)
was added portionwise triphenylphosphine (2.1 g, 8.0 mmol) at
0 °C. The mixture was stirred for 2 h while warming to room
temperature. The reaction was quenched with water, followed
by extraction with CH2Cl2. The combined organic extracts were
washed with brine and dried over MgSO4. Concentration of the
organic layer in vacuo gave a residue that was purified by
column chromatography (2:1 petroleum ether/ethyl acetate) to
afford the title compound (940 mg, 71%) as a colorless oil: Rf
0.45 (3:2 petroleum ether/ethyl acetate); νmax (neat) 2939, 1711
cm-1; δH (360 MHz; CDCl3) 1.66 (m, 2 H), 1.89 (m, 2 H), 2.15
(m, 2 H), 2.35 (m, 2 H), 2.75 (m, 2 H), 3.42 (t, J ) 6.6 Hz, 2 H),
4.79 (d, J ) 4.9 Hz, 1 H), 5.00 (d, J ) 4.5 Hz, 1 H), 5.85 (d, J )
3.5 Hz, 1 H); δC (90 MHz, CDCl3) 26.09 (CH2), 26.6 (CH2), 32.5
(CH2), 33.8 (CH2), 46.5 (CH2), 78.0 (CH), 79.4 (CH), 126.3 (CH),
148.0 (C), 206.1 (C); m/z (EI) 258 (M+; 36.7), 217 (100), 279 (35),
137 (59); HRMS calcd for C11H15O2BrNa (M + Na+) 281.0148,
found 281.0111.
Tosyla te (17). To a suspension of alcohol 16 (200.0 mg, 1.02
mmol), 4-DMAP (11.0 mg, 0.09 mmol), and triethylamine (0.63
mL, 4.52 mmol) in CH2Cl2 (11 mL) was added p-toluenesulfonyl
chloride (328 mg, 1.72 mmol) in one portion with cooling in an
ice-water bath. The reaction was stirred for 4.5 h while warming
to room temperature and then diluted with diethyl ether (15
mL). The organic phase was washed with aqueous sodium
bicarbonate solution and dried (MgSO4). The crude reaction was
purified by column chromatography (2:1 petroleum ether 60-
80/ethyl acetate) to give 17 (227 mg, 64%): mp 42-43 °C; Rf
0.48 (2:1 60-80 petroleum ether/ethyl acetate); νmax (neat) 2930,
Ack n ow led gm en t. We thank EPSRC for financial
support (studentship to R.B.O.) and the SOCRATES
European student exchange program (P.T.). We also
thank Prof. Viresh Rawal (University of Chicago) for
useful suggestions.
Su p p or tin g In for m a tion Ava ila ble: General experimen-
tal procedures and representative procedures for 4 + 3
cycloaddition reactions/Favorskii rearrangements. Copies of
1H and 13C NMR spectra for all compounds used in this study.
X-ray data for compound 1 (CIF). This material is available
free of charge via the Internet at http://pubs.acs.org.
J O034356F
7902 J . Org. Chem., Vol. 68, No. 20, 2003