Nucleophilic Additions to Fused Bicyclic Five-Membered Ring Oxocarbenium Ions: Evidence for Preferential Attack on the Inside Face

Article abstract of DOI:10.1021/ja0375176

Evidence is provided that nucleophilic attack on five-membered ring oxocarbenium ions occurs from the inside face of the envelope. An eight-five fused-bicyclic system in which two substituents are constrained to pseudoequatorial positions underwent nucleophilic addition with selectivity that was comparable to an unconstrained monocyclic system. On the other hand, a bicyclic six-five analogue underwent reaction with low selectivity. This observation indicates that minimization of eclipsing interactions by attacking inside the envelope is not enough to control selectivity, but that the changes in the overall three-dimensional structure of the ring must be favorable as well. In the bicyclic six-five series, the six-membered ring is accommodated in the cation, but it destabilizes the transition state structure leading to the first-formed product of inside attack.

Full text of DOI:10.1021/ja0375176

Published on Web 10/24/2003
Nucleophilic Additions to Fused Bicyclic Five-Membered Ring
Oxocarbenium Ions: Evidence for Preferential Attack on the
Inside Face
Deborah M. Smith, Michelle B. Tran, and K. A. Woerpel*
Contribution from the Department of Chemistry, UniVersity of California,
IrVine, California 92697-2025
Received July 24, 2003; E-mail: kwoerpel@uci.edu
Abstract: Evidence is provided that nucleophilic attack on five-membered ring oxocarbenium ions occurs
from the inside face of the envelope. An eight-five fused-bicyclic system in which two substituents are
constrained to pseudoequatorial positions underwent nucleophilic addition with selectivity that was
comparable to an unconstrained monocyclic system. On the other hand, a bicyclic six-five analogue
underwent reaction with low selectivity. This observation indicates that minimization of eclipsing interactions
by attacking inside the envelope is not enough to control selectivity, but that the changes in the overall
three-dimensional structure of the ring must be favorable as well. In the bicyclic six-five series, the six-
membered ring is accommodated in the cation, but it destabilizes the transition state structure leading to
the first-formed product of inside attack.
Introduction
The two modes of attack provide the products in different
conformations. Upon nucleophilic attack, the oxocarbenium ion
The stereoselective addition of carbon nucleophiles to oxo-
carbenium ions has proven to be a useful method for the func-
tionalization of tetrahydrofuran rings.1-3 We recently reported
a model to explain the diastereoselective reactions of substituted
five-membered ring oxocarbenium ions.4 This model stipulates
that when a nucleophile attacks a five-membered ring oxocar-
benium ion in the envelope conformation, it attacks preferen-
tially from the inside face of the envelope (eq 1).5 Attack from
the outside face is disfavored, because eclipsing interactions
between the substituents at C-1 and C-2 develop in the transition
structure leading to the first-formed product 4 (eq 2). In
recognizing that staggered transition structures are lower in
energy than eclipsed ones, our model shares features with the
analysis reported by Reissig.6-8
carbon undergoes rehybridization from trigonal planar9 to
tetrahedral. This rehybridization significantly alters the entire
three-dimensional structure of the five-membered ring. The
consequence of this rehybridization upon the conformation of
the product is a fundamental tenet of the accepted model used
to understand nucleophilic additions to six-membered ring
iminium and oxocarbenium ions.10-12 In those cases, rehybrid-
ization and torsional angle changes are favored for a chairlike
transition structure instead of a twistlike structure.
In this paper, we provide further evidence that nucleophilic
attack on five-membered ring oxocarbenium ions occurs from
inside the envelope according to eq 1. In addition, these studies
indicate that the impetus to minimize eclipsing interactions is
not the only factor that governs selectivity, but that the overall
three-dimensional structure of the ring must be favorable as well.
Experiment Design
While the inside attack model was useful for analyzing a
number of reactions of five-membered ring oxocarbenium ions,4
observations such as the highly stereoselective reaction of
(4) Larsen, C. H.; Ridgway, B. H.; Shaw, J. T.; Woerpel, K. A. J. Am. Chem.
Soc. 1999, 121, 12208-12209.
(5) Five-membered ring iminium ions are believed to undergo nucleophilic
attack by transition structures similar to those for oxocarbenium ions: Bur,
S. K.; Martin, S. F. Org. Lett. 2000, 2, 3445-3447.
The outside attack product 4 (eq 2) is not only differentiated
from the inside attack product 2 (eq 1) by eclipsing interactions.
(6) Schmitt, A.; Reissig, H.-U. Synlett 1990, 40-42.
(7) Schmitt, A.; Reissig, H.-U. Chem. Ber. 1995, 128, 871-876.
(8) Schmitt, A.; Reissig, H.-U. Eur. J. Org. Chem. 2000, 3893-3901.
(9) Dudley, T. J.; Smoliakova, I. P.; Hoffmann, M. R. J. Org. Chem. 1999,
64, 1247-1253.
(10) Stevens, R. V.; Lee, A. W. M. J. Am. Chem. Soc. 1979, 101, 7032-7035.
(11) Stevens, R. V. Acc. Chem. Res. 1984, 17, 289-296.
(12) Deslongchamps, P. Stereoelectronic Effects in Organic Chemistry; Perga-
mon: New York, 1983; pp 209-221.
(1) For a review that includes nucleophilic substitution on five-membered ring
acetals, see: Harmange, J.-C.; Figade`re, B. Tetrahedron: Asymmetry 1993,
4, 1711-1754.
(2) Levy, D. E.; Tang, C. The Chemistry of C-Glycosides; Pergamon:
Tarrytown, NY, 1995; Vol. 13.
(3) Postema, M. H. D. C-Glycoside Synthesis; CRC Press: Boca Raton, FL,
1995.
9
10.1021/ja0375176 CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 14149-14152
14149

A R T I C L E S
Smith et al.
disubstituted acetal 5 (eq 3)4,13,14 could be reinterpreted. We
proposed that the major product was formed by inside attack
on the diequatorial oxocarbenium ion 7.4 Alternatively, the major
product could arise from outside attack on the higher energy
diaxial conformer 8.15 Because both conformers would be
accessible, the conformer undergoing reaction (7 or 8) could
not be unambiguously determined. While reasonable assump-
tions could be made about which conformer predominated, it
was possible that the more reactive conformer of the oxocar-
benium ion may not be the more populated one, in accord with
the Curtin-Hammett principle.16,17 If the cation could be locked
into only one conformation, the diastereoselectivity of the nu-
cleophilic substitution would reveal the inherent stereoelectronic
preference for inside versus outside attack.
from which the diastereomerically pure trans isomer 14 could
be obtained by crystallization. Cyclization under acidic condi-
tions25 provided the trans lactone, which upon reductive
acylation21,22 yielded the required acetate 15 (eq 5).
Nucleophilic Substitutions
With the acetal substrates in hand, attention turned to the
nucleophilic substitution reactions, which likely proceed via
oxocarbenium ion intermediates.26,27 The nucleophile allyltri-
methylsilane was employed because attack by this nucleophile
onto a carbocation should be irreversible,28,29 and because this
small nucleophile is relatively insensitive to steric effects upon
approach to the cation.30
The nucleophilic substitution reactions of the six-five ring
system are shown in eq 6 and Table 1. Regardless of the solvent
A bicyclic acetal related to acetal 5 would provide the
appropriately constrained analogue.18 If the two substituents at
C-3 and C-4 were tethered together with a chain of a relatively
short length, only the diequatorial oxocarbenium ion 9 would
be accessible. The diaxial oxocarbenium ion 10 would be
severely strained, because this conformer requires a 180°
dihedral angle within the fused ring. With only a diequatorial
conformer available, the inherent stereoelectronic preference for
inside attack versus outside attack would be revealed because
neither trajectory for approach is disfavored for steric reasons.
Table 1. Stereoselectivity of Nucleophilic Substitution of Six-Five
Bicyclic Acetate 12 (eq 6)
entry
Lewis acid
solvent
16:17a
yieldb (%)
1
2
3
4
5
6
7
8
9
BF3‚OEt2
SnBr4
Me3SiOTf
Me2AlCl
MeAlCl2
TiCl4
BF3‚OEt2
BF3‚OEt2
BF3‚OEt2
BF3‚OEt2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
C6H5CH3
Et2O
73:27
69:31
71:29
65:35
68:32
76
82
74
67
76
c
c
-
-
Synthesis of Substrates
73:27
56:44
73:27
73:27
76
50
55
61
The required acetate substrates 12 and 15 were prepared as
shown in eqs 4 and 5. Nitrile 11, prepared in one step from
cyclohexene oxide,19 was cyclized to the lactone20 which was
reductively acylated21,22 to afford acetate 12 (eq 4).23 A mixture
of cis- and trans-lactones 13, which was prepared in one step
from cyclooctene,23 was converted to a mixture of anilides,24
CHCl3
CH3CN
10
a Diastereoselectivity determined by gas chromatography and confirmed
1
by H NMR spectroscopy. b Yield of product after flash chromatography.
c No product was isolated, but starting material was consumed.
and the Lewis acid, the selectivities were low.31 The configura-
tions of the products were determined by analysis of NOE data.
(13) Ishihara, J.; Miyakawa, J.; Tsujimoto, T.; Murai, A. Synlett 1997, 1417-
1419.
(14) Yang, J.; Cohn, S. T.; Romo, D. Org. Lett. 2000, 2, 763-766.
(15) Similarly, the minor product could arise from either outside attack on the
diequatorial oxocarbenium ion 7 or inside attack on the diaxial cation 8.
(16) Seeman, J. I. Chem. ReV. 1983, 83, 83-134.
(25) Askin, D.; Volante, R. P.; Ryan, K. M. Tetrahedron Lett. 1988, 29, 4245-
4248.
(26) See, for example: (a) Sammakia, T.; Smith, R. S. J. Am. Chem. Soc. 1994,
116, 7915-7916. (b) Matsutani, H.; Ichikawa, S.; Yaruva, J.; Kusumoto,
T.; Hiyama, T. J. Am. Chem. Soc. 1997, 119, 4541-4542. Other reactions
of acetals do not appear to involve free cations: (c) Denmark, S. E.;
Almstead, N. G. J. Am. Chem. Soc. 1991, 113, 8089-8110.
(27) In our experiments with both tetrahydrofuran and tetrahydropyran acetals,
the stereochemistry of nucleophilic substitution was independent of the
anomer ratios of the starting acetates: (a) ref 4. (b) Romero, J. A. C.;
Tabacco, S. A.; Woerpel, K. A. J. Am. Chem. Soc. 2000, 122, 168-169.
(28) Burfeindt, J.; Patz, M.; Mu¨ller, M.; Mayr, H. J. Am. Chem. Soc. 1998,
120, 3629-3634.
(17) Seeman, J. I. J. Chem. Educ. 1986, 63, 42-48.
(18) For a recent example that compares the stereoselectivity of constrained
versus unconstrained five-membered ring oxocarbenium ions, see: Crich,
D.; Hao, X. J. Org. Chem. 1999, 64, 4016-4024.
(19) Bartlett, P. A.; Ting, P. C. J. Org. Chem. 1986, 51, 2230-2240.
(20) Pirkle, W. H.; Adams, P. E. J. Org. Chem. 1980, 45, 4111-4117.
(21) Dahanukar, V. H.; Rychnovsky, S. D. J. Org. Chem. 1996, 61, 8317-
8320.
(22) Kopecky, D. J.; Rychnovsky, S. D. J. Org. Chem. 2000, 65, 191-198.
(23) Fristad, W. E.; Peterson, J. R. J. Org. Chem. 1985, 50, 10-18.
(24) Ooi, T.; Tayama, E.; Yamada, M.; Maruoka, K. Synlett 1999, 729-730.
(29) Hagen, G.; Mayr, H. J. Am. Chem. Soc. 1991, 113, 4954-4961.
9
14150 J. AM. CHEM. SOC. VOL. 125, NO. 46, 2003

Reactions of Bicyclic Oxocarbenium Ions
A R T I C L E S
>90% selectivity arose from nucleophilic attack on the cation
from the inside face of the envelope (eq 9). Because the outside
approach of the nucleophile does not appear to be sterically
disfavored, the preference for inside attack must be a result of
steric interactions that form within the ring during outside attack
(minimizing eclipsing interactions, vide supra). Because the ster-
eoselectivity observed for the bicyclic system matches the selec-
tivity observed for the unconstrained acetal 5 (eq 3), the selec-
tivity of the unconstrained system most likely results from inside
attack on the diequatorially substituted cation 7 (vide supra).
Table 2. Stereoselectivity of Nucleophilic Substitution of
Eight-Five Bicyclic Acetate 15 (eq 7)
entry
Lewis acid
solvent
18:19a
yieldb (%)
1
2
3
4
5
6
7
8
9
BF3‚OEt2
SnBr4
Me3SiOTf
Me2AlCl
MeAlCl2
TiCl4
SnBr4
SnBr4
SnBr4
SnBr4
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
C6H5CH3
Et2O
93:7
93:7
93:7
92:8
92:8
83
82
85
64
81
c
c
-
-
96:4
91:9
92:8
91:9
80
82
81
74
CHCl3
CH3CN
The contrast between the results with the eight-five and the
six-five systems (cf., Tables 1 and 2) is meaningful. The low
selectivity observed for the six-five fused acetal 12 (eq 6, Table
1) indicates that the preference for inside attack of the nucleo-
phile in this system is diminished. If the minimization of eclip-
sing interactions were all that were important to define stereo-
selectivity, then the two fused systems must behave similarly.
Therefore, any analysis that considers only eclipsing interactions
in transition structures cannot accommodate these results.
The inside attack model accounts for the low selectivity of
nucleophilic attack exhibited by the six-five fused ring oxocar-
benium ion. The five-membered ring of oxocarbenium ion 23
adopts an envelope conformation,35 and the dihedral angles with-
in this ring accommodate the optimal chair conformer of the
six-membered ring. According to the inside attack model, the
envelope would undergo an overall ring flip upon nucleophilic
attack from the inside face, leading to the first-formed product
25 (eq 10). This overall change in the conformation of the five-
membered ring would modify the conformation of the six-
membered ring, because the dihedral angles around the C3-C4
bond are different in product 25 than in oxocarbenium ion 23.35
The six-membered ring of product 25 would be distorted from
the ideal chair structure, so the transition state leading to product
25 would be destabilized.35 On the other hand, attack on the
outside of the envelope would proceed by a transition structure
that would maintain the chair conformation of the six-membered
ring. As product 24 is formed, however, unfavorable eclipsing
interactions would develop in the five-membered ring. Conse-
quently, the two pathways for reaction of oxocarbenium ion 23
(inside and outside attack) each involve destabilized transition
structures, resulting in a reaction with low stereoselectivity.
10
a Diastereoselectivity determined by gas chromatography and confirmed
1
by H NMR spectroscopy. b Yield of product after flash chromatography.
c No product was isolated, but starting material was consumed.
The sense of stereoselectivity corresponds to that exhibited by
the unconstrained acetate 5 (eq 3), although the unconstrained
system showed markedly higher selectivity (about 10-fold
higher).
In contrast to the observations of the six-five ring system
(eq 6 and Table 1), nucleophilic substitution reactions with the
bicyclic eight-five ring system proceeded with high selectivity.
As shown in eq 7 and Table 2, the diastereoselectivities of these
reactions matched the selectivities observed for the uncon-
strained system, and these selectivities were independent of the
solvent and the Lewis acid.27,31,32 The configuration of the major
product was assigned by conversion of the alkene to the
carbamate 20 (eq 8), whose structure was determined by X-ray
crystallography.33
Discussion
The highly stereoselective reactions of the fused eight-five
bicyclic system 15 (eq 7 and Table 2) provide evidence that a
five-membered ring oxocarbenium ion undergoes nucleophilic
attack with a stereoelectronic preference for inside attack.
Because of the constraints imposed by the trans ring fusion and
the short tether, only a diequatorially substituted conformer of
the five-membered ring should be accessible (i.e., 21).34 Under
all conditions examined, the product that was obtained with
(30) Bear, T. J.; Shaw, J. T.; Woerpel, K. A. J. Org. Chem. 2002, 67, 2056-
2064.
(31) With each Lewis acid, the counterion for the oxocarbenium ion is different,
but the selectivity is independent of this counterion. In addition, because
the selectivities are insensitive to the solvent, ion pairing is not evidently
important.
(34) While other conformations of the eight-membered ring may be relatively
close in energy, it is the conformation of the five-membered ring that is
important in controlling stereoselectivity. Oxocarbenium ion 21 and the
product 22 are shown with the eight-membered ring in a boat-chair
conformation, because the eight-membered rings of trans-fused eight-five
ring systems adopt boat-chair forms: Umehara, M.; Hosomi, H.; Ohba, S.
Acta Crystallogr., Sect. C 1999, 55, 1721-1725.
(32) The diastereoselectivity of this nucleophilic substitution is independent of
the ratio of anomeric acetates, consistent with the intermediacy of
oxocarbenium ions (details of these experiments are provided as Supporting
Information).
(33) Additional details are provided as Supporting Information.
9
J. AM. CHEM. SOC. VOL. 125, NO. 46, 2003 14151

A R T I C L E S
Smith et al.
1.15-1.35 (m, 4H), 1.07 (m, 1H). 13C NMR (125 MHz, CDCl3): δ
Conclusion
135.5, 117.3, 84.2, 77.2, 44.7, 41.7, 35.9, 31.8, 29.6, 26.4, 24.8.
These experiments with fused-ring acetals provide additional
support that the reactions of five-membered ring oxocarbenium
ions with nucleophiles are strongly influenced by stereoelec-
tronic effects. The preference for attack from inside the envelope
was observed for the eight-five ring system, in which the five-
membered ring is constrained to an envelope conformer with
the eight-membered ring spanning two equatorial positions. The
different behavior of the six-five and eight-five ring systems
illuminates that, upon nucleophilic attack, the five-membered
ring undergoes a significant change in its overall three-
dimensional structure, and that this change must not engender
too much strain elsewhere if high selectivity is to be observed.
1
Minor Isomer 17. H NMR (500 MHz, CDCl3, distinctive peaks):
δ 3.15 (td, J ) 10.3, 3.7, 1H), 1.42 (m, 1H). 13C NMR (125 MHz,
CDCl3, distinctive peaks): δ 82.9, 78.2, 46.7, 41.6, 40.0, 31.9, 29.5,
26.2, 24.8.
Eight-Five Bicyclic Alkene (18). The standard allylation procedure
was followed with acetate 15 (0.050 g, 0.24 mmol) and BF3‚OEt2 (0.04
mL, 0.3 mmol). GC and 1H NMR spectroscopic analysis of the
unpurified product showed a pair of diastereomers in a 93:7 ratio of
1,3-trans:cis diastereomers. Purification by flash chromatography (2:
98 EtOAc/hexanes) provided the product as a colorless oil (0.038 g,
1
83%). H NMR (500 MHz, CDCl3): δ 5.81 (ddt, J ) 17.2, 10.2, 7.0,
1H), 5.05 (m, 2H), 3.82 (quintet, J ) 6.7, 1H), 3.66 (td, J ) 8.5, 4.0,
1H), 2.33 (m, 1H), 2.17 (m, 1H), 2.04 (m, 2H), 1.83 (m, 2H), 1.69 (m,
4H), 1.60 (m, 1H), 1.33-1.46 (m, 6H). 13C NMR (125 MHz, CDCl3):
δ 135.6, 117.0, 85.4, 76.5, 42.9, 41.3, 40.7, 35.5, 34.9, 28.2, 27.4, 25.5,
23.6. IR (thin film): 3075, 2922, 2853, 1641 cm-1. HRMS (EI) m/z
calcd for C13H23O (M + H)+ 195.1749, found 195.1748. Anal. Calcd
for C13H22O: C, 80.36; H, 11.41. Found: C, 80.53; H, 11.55.
Experimental Section33
General Procedure for Allylation of γ-Lactol Acetates. A solution
of acetate in CH2Cl2 (0.10 M) was treated with allyltrimethylsilane (4
equiv) and then cooled to -78 °C. After treatment with Lewis acid
(1.1 equiv), the reaction mixture was warmed to 22 °C over 2 h. The
reaction mixture was treated with saturated aqueous Na2HPO4 (1 mL
per mmol of acetate). The aqueous layer was then extracted three times
with CH2Cl2 (1 mL per mmol of acetate), and the organic phases were
dried (Na2SO4) and concentrated in vacuo.
Acknowledgment. This research was supported by the
National Institutes of Health, National Institute of General
Medical Science (GM-61066), and by Johnson & Johnson.
Acknowledgment is also made to the donors of the Petroleum
Research Fund, administered by the American Chemical Society,
for partial support of this research. D.M.S. thanks Johnson &
Johnson for a graduate research fellowship. K.A.W. thanks
Merck Research Laboratories, Johnson & Johnson, and the Sloan
Foundation for awards to support research. We thank Ms.
Catharine Larsen for carrying out some preliminary experiments,
Dr. Brian Ridgway for helpful discussions, Dr. John Greaves
and Dr. John Mudd for mass spectrometric data, and Dr. Joe
Ziller for the X-ray crystallographic analysis.
Six-Five Bicyclic Alkene (16 and 17). The standard allylation
procedure was followed with acetate 12 (0.051 g, 0.28 mmol) and
BF3‚OEt2 (0.4 mL, 0.3 mmol). GC and 1H NMR spectroscopic analysis
of the unpurified product showed a pair of diastereomers in a 73:27
ratio of 1,3-trans:cis diastereomers. Purification by flash chromatog-
raphy (2:98 Et2O/pentane) provided the product as a clear oil (0.035 g,
76%). The major isomer 16 was isolated as a pure sample, while the
minor isomer 17 was isolated as a mixture of 16 and 17. IR and mass
spectrometry data were obtained for 16 and 17 as a mixture of
diastereomers. IR (thin film): 2932, 2857, 1641, 1072 cm-1. HRMS
(EI) m/z calcd for C8H13O (M - C3H5)+ 125.0966, found 125.0969.
1
Major Isomer 16. H NMR (500 MHz, CDCl3): δ 5.81 (ddt, J )
Supporting Information Available: Complete experimental
procedures, product characterization, stereochemical proofs,
X-ray crystallographic data for carbamate 20, and GC and
spectral data for selected compounds (PDF and CIF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
17.2, 10.3, 6.9, 1H), 5.09 (m, 1H), 5.06 (m, 1H), 4.12 (m, 1H), 3.08
(td, J ) 10.1, 3.7, 1H), 2.40 (m, 1H), 2.29 (m, 1H), 2.16 (m, 1H), 1.90
(m, 1H), 1.81 (m, 1H), 1.71 (m, 2H), 1.60 (td, J ) 11.7, 9.1, 1H),
(35) The dihedral angle between C3 and C4 within the five-membered ring of
structure 25 is about 25° smaller than that for the chair conformation of
cyclohexane. Consequently, the six-membered ring of 25 should be distorted
from a chair conformation by a comparable angle. For a review of five-
membered ring conformational analysis, see: Fuchs, B. Top. Stereochem.
1978, 10, 1-94.
JA0375176
9
14152 J. AM. CHEM. SOC. VOL. 125, NO. 46, 2003