Stereoselectivity control by oxaspiro rings during Diels-Alder cycloadditions to cross-conjugated cyclohexadienones: The syn oxygen phenomenon

Article abstract of DOI:10.1021/ja047027t

The diastereofacial selectivity operating in Diels-Alder additions involving spirocyclic cross-conjugated cyclohexadienones with dienes of varying reactivity has been investigated. The study has included the ether series 1a-c as well as the lactone/ketone pair 2a/2b. In all cases, the preferred [4+2] cycloaddition pathway consisted of bonding from that π-surface syn to the oxygen atom. 4-Substituted-4-methyl-2,5-cyclohexadienones (monocyclic systems) were also examined and found to undergo bond formation preferentially from the face bearing the more electron-withdrawing of the two groups at the 4 position. Kinetic parameters were determined for the cycloaddition of 1a and 2a to cyclopentadiene. The rate acceleration profile of solvents was in the order CF₃CH₂OH ? CH₃CN～CH₂Cl ₂ for the production of 9a from 1a and CF₃CH₂OH ? CH₂Cl₂ > CH₃CN for the production of 21a from 2a, respectively. This spread in polarity had no major impact on product distribution, a phenomenon also reflected in the behavior of 4-substituted-4-methyl-2,5-cyclohexadienones under comparable conditions. Theoretical assessment of these experimental facts was undertaken at the HF/6-31G* level. The facial selectivity is understandable in terms of the secondary interaction between the HOMO of the diene and LUMO of the dienophile as well as the effective hyperconjugation between the newly forming bond and the 4-anti-C-C σ-orbital due to the more electron-donating bond, as defined by the Cieplak model.

Full text of DOI:10.1021/ja047027t

Published on Web 12/03/2004
Stereoselectivity Control by Oxaspiro Rings during
Diels-Alder Cycloadditions to Cross-Conjugated
Cyclohexadienones: The Syn Oxygen Phenomenon
Katsuo Ohkata,*,† Yukiko Tamura,† Brandon B. Shetuni,‡ Ryukichi Takagi,†
Wataru Miyanaga,† Satoshi Kojima,† and Leo A. Paquette*,‡
Contribution from the Department of Chemistry, Graduate School of Science,
Hiroshima UniVersity, 1-3-1 Kagamiyama, Higashi-Hiroshima 739-8526, Japan and
The Ohio State UniVersity, Columbus, Ohio 43210
Received May 20, 2004; E-mail: paquette.1@osu.edu
Abstract: The diastereofacial selectivity operating in Diels-Alder additions involving spirocyclic cross-
conjugated cyclohexadienones with dienes of varying reactivity has been investigated. The study has
included the ether series 1a-c as well as the lactone/ketone pair 2a/2b. In all cases, the preferred [4+2]
cycloaddition pathway consisted of bonding from that π-surface syn to the oxygen atom. 4-Substituted-4-
methyl-2,5-cyclohexadienones (monocyclic systems) were also examined and found to undergo bond
formation preferentially from the face bearing the more electron-withdrawing of the two groups at the 4
position. Kinetic parameters were determined for the cycloaddition of 1a and 2a to cyclopentadiene. The
rate acceleration profile of solvents was in the order CF3CH2OH . CH3CN∼CH2Cl2 for the production of
9a from 1a and CF3CH2OH . CH2Cl2 > CH3CN for the production of 21a from 2a, respectively. This spread
in polarity had no major impact on product distribution, a phenomenon also reflected in the behavior of
4-substituted-4-methyl-2,5-cyclohexadienones under comparable conditions. Theoretical assessment of
these experimental facts was undertaken at the HF/6-31G* level. The facial selectivity is understandable
in terms of the secondary interaction between the HOMO of the diene and LUMO of the dienophile as well
as the effective hyperconjugation between the newly forming bond and the 4-anti-C-C σ-orbital due to the
more electron-donating bond, as defined by the Cieplak model.
The capacity for generating four contiguous stereogenic
centers in a single laboratory operation is obviously responsible
for the widespread interest in the synthetic utility of the Diels-
Alder reaction.1 The stereochemical features of such cycload-
ditions have more recently been recognized to be capable of
modulation by a neighboring heteroatomic substituent. Desym-
metrization in this manner with continued operation of high
diastereoselectivity offers additional stereochemical advantages.2
Previous studies involving functionalized facially unsymmetric
semicyclic dienes,3 cyclopentadienes,4 and 1,3-cyclohexadienes5
have given evidence that high facial selectivity can be attained
during [4+2] cycloaddition to symmetric dienophiles. Relatively
rigid cyclic dienophiles have been the least studied.6
observed facial selectivities have been accounted for in terms
of hyperconjugative,8 electrostatic,9 torsional,10 and orbital
interactions.11-15 A rationale consistent with all of the experi-
mental observations continues to be sought.
Our own interest in this area stems from two divergent
directions. The Ohio State group has had a long-standing interest
in the chemistry of spirotetrahydrofurans,16 and ultimately
became interested in the stereochemical issues surrounding the
participation of cyclohexadienone 1a in cycloadditions to
appropriately reactive dienes.17 Pursuit by the Hiroshima team
of a total synthesis of scyphostatin necessitated that appreciation
be gained of the π-facial selectivity capable of being exhibited
by 2a.18 These preliminary considerations have since been
In select cases, steric effects control selectivity.7 When
structural inequities are brought more into balance, increasingly
subtle effects gain prominence. Under these circumstances, the
(3) Exemplary functionalized semi-cyclic dienes; (a) Fisher, M. J.; Hehre, W.
J.; Kahn, S. D.; Overman, L. E. J. Am. Chem Soc. 1988, 110, 4625. (b)
Datta, S. C.; Franck, R. W.; Tripathy, R.; Quigley, G. J.; Huang, L.; Chen,
S.; Sihaed, A. J. Am. Chem. Soc. 1990, 112, 8472. (c) Kawamata, T.;
Harimaya, K.; Iitaka, Y.; Inayama, S. Chem. Pharm. Bull. 1989, 37, 2307.
(d) Giuliano, R. M.; Buzby, J. H.; Marcopulos, N.; Springer, J. P. J. Org.
Chem. 1990, 55, 3555. (e) Burnouf, C.; Lopez, J. C.; Calvo-Flores, F. G.;
Laborde, M. A.; Olesker, A.; Lukacs, G. J. Chem. Soc., Chem. Commun.
1990, 823. (f) Grabley, S.; Kluge, H.; Hoppe, H.-U. Angew. Chem., Int.
Ed. Engl. 1987, 26, 664.
(4) Exemplary functionalized cyclopentadienes; (a) Breslow, R.; Hoffman, J.
M. J. Am. Chem. Soc. 1972, 94, 2110. (b) Franck-Neumann, M.; Sedrati,
M. Tetrahedron Lett. 1983, 24, 1391. (c) Corey, E. J.; Weinshenker, N.
M.; Schaaf, T. K.; Huber, W. J. Am. Chem. Soc. 1969, 91, 5675. (d)
Paquette, L. A.; Vanucci, C.; Rogers, R. D. J. Am. Chem. Soc., 1989, 111,
5792. (e) Ishida, M.; Beniya, Y.; Inagaki, S.; Kato, S. J. Am. Chem. Soc.
1990, 112, 8980.
† Hiroshima University.
‡ The Ohio State University.
(1) (a) Oppolzer, W. ComprehensiVe Organic Synthesis, Vol, 5; Trost, B. M.,
Fleming, I., Paquette, L. A., Eds.; Pergamon: Oxford, 1991; pp 315-399.
(b) Carruthers, W. Cycloaddition Reactions in Organic Synthesis; Perga-
mon: Oxford, 1990. (c) Fringuelli, F.; Taticchi, A. The Diels-Alder
Reaction-Selected Practical Methods; John Wiley and Sons: New York,
2002.
(2) (a) Fallis, A. G.; Lu, Y. F. In AdVances in Cycloaddition; Curran, D. P.,
Ed.; JAI Press Inc.: Greenwich, Ct 1993; Vol.1; pp 1-66. (b) Ohwada, T.
Chem. ReV. 1999, 99, 1337. (c) Mehta G.; Uma, R. Acc. Chem. Res. 2000,
33, 278.
9
10.1021/ja047027t CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 16783-16792
16783

A R T I C L E S
Ohkata et al.
Scheme 1
expanded to include spirooxetane 1b and spirooxirane 1c as
well as the spirocyclic diketone 2b. The discoveries made in
the context of these investigations have been amalgamated in
the present joint contribution.
The series 1a-c was probed in the expectation that alterations
in the size of the heterocyclic ring would be accompanied by
changes in the basicity of the ethereal oxygen center.19 Would
a ground-state parameter of this nature exercise any influence
on kinetic selectivity? Do the steric changes that accompany
spiro ring contraction facilitate attack from the same or opposite
face? And what can be expected of differences in hyperconju-
gation? The chemical nuances separating 2a from 2b are more
disparate. Are the prospects for recognizing an electrostatic
control mechanism thereby enhanced? These cross-conjugated
dienones appeared well suited for critical analysis of the source
of stereocontrol in their Diels-Alder reactions.
Results and Discussion
(5) Exemplary functionalized 1,3-cyclohexadienes; (a) Auksi, H.; Yates, P. Can.
J. Chem. 1981, 59, 2510. (b) Holmberg, K.; Kirudd, H.; Westin, G. Acta
Chem, Scand., Ser. B 1974, 28, 913. (c) Gillard, J. R.; Burnell, D. J. J.
Chem. Soc., Chem. Commun. 1989, 1439. (e) Gillard, J. R.; Newlands, M.
J.; Bridson, J. N.; Burnell, D. J. Can. J. Chem. 1991, 69, 1337.
(6) (a) Liotta, D.; Saindane, M.; Barnum, C. J. Am. Chem. Soc. 1981, 103,
3224. (b) Traˆn-Huu-Daˆn, M.-E.; Wartchow, R.; Winterfeldt, E.; Wong,
Y.-S. Chem. Eur. J. 2001, 7, 2349. (c) Jeroncic, L. O.; Cabal, M.-O.;
Danishefsky, S. J.; Shulte, G. M. J. Org. Chem. 1991, 56, 387. (d) Dauben,
W. G.; Kowalczyk, B. A.; Lichtenthaler, F. W. J. Org. Chem. 1990, 55,
2391. (e) Goldenstein, K.; Fendert, T.; Proksch, P.; Winterfeldt, E.
Tetrahedron 2000, 56, 4173. (f) Winterfeldt, E.; Bo¨hm, C.; Nerenz, F.
AdVances in Asymmetric Synthesis; JAI Press Inc.: Greenwich, CT 1997;
Vol 2, p 1.
Synthesis. Substrates 1a and 2a were prepared by oxidative
cyclization of 3-(4-hydroxyphenyl)propanol (3a) and its car-
boxylic acid derivative 3b with iodobenzene diacetate.20 The
rather modest yield associated with the production of 1a via
this protocol prompted alternate advance on this dienone by
way of the pathway given in Scheme 1. Addition of the Normant
reagent to 4
(7) (a) Paquette, L. A. In Asymmetric Synthesis; Morrison, J. D. Ed.; Academic
Press: New York, 1984; Vol.3, and references therein. (b) Silvero, G.;
Lucero, M. J.; Winterfeldt, E.; Houk, K. N. Tetrahedron 1998, 54, 7293.
(8) (a) Macaulay, J. B.; Fallis, A. G. J. Am. Chem. Soc. 1990, 112, 1136. (b)
Coxon, J. M.; McDonald, D. Q. Tetrahedron Lett. 1992, 33, 651. (c) Poirier,
R. A.; Pye, C. C.; Xidos, J. D.; Burnell, D. J. J. Org. Chem. 1995, 60,
2328.
(9) (a) Kahn, S. D.; Hehre, W. J. J. Am. Chem. Soc. 1987, 109, 663. (b)
Paquette, L. A.; Branan, B. M.; Rogers, R. D.; Bond, A. H.; Lange, H.;
Gleiter, R. J. Am. Chem. Soc. 1995, 117, 5992. (c) Wipf, P.; Kim, Y. J.
Am. Chem. Soc. 1994, 116, 11678. (d) Wipf, P.; Jung, J.-K. Chem. ReV.
1999, 99, 1469. (e) Wipf, P.; Kim, Y. J. Org. Chem. 1993, 58, 1649. (f)
Solomon, M.; Jamison, W. C. L.; McCormick, M.; Liotta, D.; Cherry, D.
A.; Mills, J. E.; Shah, R. D.; Rodgers, J. D.; Maryanoff, C. A. J. Am. Chem.
Soc. 1988, 110, 3702. (g) Swiss, K. A.; Hinkely, W.; Maryanoff, C. A.;
Liotta, D. C. Synthesis 1992, 127.
(10) (a) Brown, F. K.; Houk, K. N. J. Am. Chem. Soc. 1985, 107, 1971. (b)
Houk, K. N.; Gonzolez, J.; Li, Y. Acc. Chem. Res. 1995, 28, 81.
(11) (a) Inagaki, S.; Fukui, K. Chem. Lett. 1974, 509. (b) Inagaki, S.; Fujimoto,
H.; Fukui, K. J. Am. Chem. Soc. 1976, 98, 4054.
(12) (a) Ishida, M.; Kobayashi, H.; Tomohiro, S.; Inagaki, S. J. Chem. Soc.,
Perkin Trans. 2 2000, 1625. (b) Ishida, M.; Beniya, Y.; Inagaki, S.; Kato,
S. J. Am. Chem. Soc. 1990, 112, 8980. (c) Ishida, M.; Aoyama, T.; Beniya,
Y.; Yamabe, S.; Kato, S.; Inagaki, S. Bull. Chem. Soc. Jpn. 1993, 66, 3430.
(d) Ishida, M.; Tomohiro, S.; Shimizu, M.; Inagaki, S. Chem. Lett. 1995,
739. (e) Ishida, M.; Kobayashi, H.; Tomohiro, S.; Wasada, H.; Inagaki, S.
Chem. Lett. 1998, 41.
followed by monotosylation was met with spontaneous cycliza-
tion to give 6.21,22 Introduction of the pair of double bonds were
achieved by 2-fold bromination-dehydrobromination.23 Follow-
ing arrival at 7, the acetal functionality was hydrolyzed in dilute
acid to unmask the keto group.
The preparation of 1b was accomplished by irradiation of
benzene solutions of p-benzoquinone with a 125 W medium-
pressure mercury arc with concomitant bubbling of ethylene
through the medium.24 1-Oxaspiro[2.5]octa-4,7-dien-6-one (1c)
was formed by the oxidation of 8 with sodium bismuthate.25
Diketone 2b was conveniently available by several routes
reported earlier.26
(13) Mazzocchi, P. H.; Stahly, B.; Dodd, J.; Rondan, N. G.; Domelsmith, L.
N.; Rozeboom, M. D.; Caramella, P.; Houk, K. N. J. Am. Chem. Soc. 1980,
102, 6482.
(14) Paquette, L. A.; Bellamy, F.; Wells, G. J.; Bo¨hm, M. C.; Gleiter, R. J. Am.
Chem. Soc. 1981, 103, 7122.
(15) (a) Gleiter, R.; Paquette, L. A. Acc. Chem. Res. 1983, 16, 328. (b) Bo¨hm,
M. C.; Eiter, R. G. Tetrahedron 1980, 36, 3209. (c) Bo¨hm, M. C.; Carr, R.
V. C.; Gleiter, R.; Paquette, L. A. J. Am. Chem. Soc. 1980, 102, 7218. (d)
Paquette, L. A.; Carr, R. V. C.; Bo¨hm, M. C.; Gleiter, R. J. Am. Chem.
Soc. 1980, 102, 1186.
(16) (a) Paquette, L. A. In Chemistry for the 21st Century, Keinan, E., Schechter,
I., Eds.; Wiley-VCH: Weinheim, 2001; pp 37-53. (b) Paquette, L. A.
Recent Res. DeVel. In Chemical Sciences, 1997, 1. (c) Paquette, L. A. Aust.
J. Chem. 2004, 57, 7.
(17) Paquette, L. A.; Shetuni, B. B.; Gallucci, J. C. Org. Lett. 2003, 5, 2639.
(18) (a) Takagi, R.; Miyanaga, W.; Tamura, Y.; Ohkata, K. Chem. Commun.
2002, 2096. (b) Takagi, R.; Miyanaga, W.; Tamura, Y.; Kojima, S.; Ohkata,
K. Heterocycles 2003, 60, 785. (c) Takagi, R.; Tsuyumine, S.; Nishitani,
H.; Miyanaga, W.; Ohkata, K. Aust. J. Chem. 2004, 57, 439.
(19) Bordeje, M. C.; Mo, O.; Yanez, M.; Herreros, M.; Abboud, J.-L. M. J.
Am. Chem. Soc. 1993, 115, 7389.
(20) (a) McKillop, A.; McLaren, L.; Taylor, R. J. K. J. Chem. Soc., Perkin
Trans. 1 1994, 2047. (b) Pelter, A.; Elgendy, S. Tetrahedron Lett. 1988,
29, 677. (c) Tamura Y.; Yakura, T.; Haruta, J.; Kita, Y. J. Org. Chem.
1987, 52, 3927.
(21) Cahiez, G.; Alexakis, A.; Normant, J. F. Tetrahedron Lett. 1978, 19, 3013.
9
16784 J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004

Syn Oxygen Phenomenon
A R T I C L E S
Table 1. Stereoselectivities Determined for Diels-Alder Reactions
Involving 1a-c and CP or SH
tions that are featured. The consistent adoption of endo transition
states was further reflected in those coupling constants involving
the bridgehead protons (J ) 3.4-4.3 Hz).
run no.
dienone
solvent
reaction conditions
9: 10a
yield, %b
A. cyclopentadiene (CP):
Response of 1a-c to Other Dienes. Our interest in examin-
ing 1128 arises from its rigid conformational features, which
lock the exocyclic double bonds into the reactive cisoid
geometry. When 11 was heated with 1a-c in refluxing toluene
for several hours, a sole product was isolated in all three
examples. Definition of these adducts as the result of exclusive
syn capture as in 12 was validated by X-ray crystallographic
analysis of 12a and detailed NOESY studies involving 12b and
12c.17
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1b
1b
1c
1c
CH2Cl2
CH3CN
25 °C, 7 days
25 °C, 7 days
25 °C, 1 day
36 °C, 6 h
36 °C, 6 h
36 °C, 8 h
97:3
94:6
94:6
61
46
86
90
96
95
82
86
CF3CH2OH
CF3CH2OH
CF3CH2OH
CF3CH2OH
CF3CH2OH
CF3CH2OH
92:8
86:14
85:15
81:19
83:17
36 °C, 13 h
36 °C, 15 h
B. spiro[2,4]hepta-4,6-diene (SH):
9
10
11
12
13
1a
1b
1b
1c
1c
CF3CH2OH
CF3CH2OH
CF3CH2OH
CF3CH2OH
CF3CH2OH
55 °C, 15 h
55 °C, 18 h
55 °C, 21 h
55 °C, 11 h
55 °C, 17 h
77:23
67:33
64:36
65:35
66:34
86
83
84
67
59
1
a The ratios were determined by integration of the H NMR spectra of
unpurified field fragments. b Isolated yields after chromatographic purifica-
tion.
Cycloaddition of Cyclopentadiene (CP) and Spiro[2.4]-
hepta-4,6-diene (SH) to 1a-c. The first candidate reactions
involved CP and its somewhat more sterically shielded homo-
logue SH.27 These studies, which were initially performed in
the absence of Lewis acid catalysis to simplify mechanistic
considerations, are summarized in Table 1.
As for CP, endo adduct 9 and/or 10 can result in each
example depending on the level and direction of π-facial
selection that is operative. These reactions were found to proceed
very slowly in CH2Cl2 and CH3CN at room temperature or in
toluene at 90 °C. For this reason, a switch was made to the
more polar trifluoroethanol as solvent. As seen for runs 1-4,
diastereoselectivity was not hardly eroded in this medium even
as the temperature was increased to 36 °C.
The next candidate diene was the sterically more demanding
diphenylisobenzofuran 13.29 Heating this co-reactant with 1a
and 1b in benzene afforded uniquely the adducts 14a and 14b,
respectively. The relative configuration of these products was
assigned on the basis of NOESY data for 14a and X-ray
crystallographic data for 14b (Supporting Information). When
the same experiment was performed with 1c, adducts 14c and
15 were isolated in the proportion of 78:22. In this example,
the minor constituent of the product mixture was subjected to
three-dimensional crystallographic analysis.
Further use was made of the kinetic acceleration realized with
trifluoroethanol in the Diels-Alder reactions involving SH as
the diene. This less reactive 4π donor, like CP, gives rise to
chromatographically separable mixtures of endo adducts 9 and
10, with a very respectable kinetic preference for utilization of
the π surface syn to the ethereal oxygen. An interesting point
is that the formation of 9, which dominates significantly under
all circumstances (runs 1-8) involving reaction with CP, is met
with a drop-off in its relative percentage when the slower
reacting SH is involved.
Heating 16 in diglyme is recognized to allow for the in situ
generation of isobenzofuran (17).30 When this process was
brought about in the presence of 1a, three adducts 18-20 were
formed. These proved to be amenable to chromatographic
separation and independent characterization. The two less polar
bridged ethers 18 and 19 were readily identified as the end
products of endo addition, since both exhibited J values for
interaction of the bridgehead and fused-ring protons of the
(26) (a) Fivush A. M.; Strunk, S. R. Synth. Commun. 1996, 26, 1623. (b)
Swenton, J. S.; Callinan, A.; Wang, S. J. Org. Chem, 1992, 57, 78. (c)
Swenton, J. S.; Bradin, D.; Gates, B. D. J. Org. Chem. 1991, 56, 6156. (d)
Iwata, C.; Yamada, M.; Ida, Y.; Imao, T.; Miyagawa, H.; Miyashita, K.
Chem. Pharm. Bull. 1988, 36, 2864. (e) Iwata, C.; Miyashita, K.; Imao,
T.; Matsuda, K.; Kondo, N.; Uchida, S. Chem. Pharm. Bull. 1985, 33, 853.
(f) Iwata, C.; Yamada, M.; Shinoo, Y.; Kobayashi, K.; Okada, H. Chem.
Pharm. Bull. 1980, 28, 1932. (g) Iwata, C.; Yamada, M.; Shinoo, Y.;
Kobayashi, K.; Okada, H. J. Chem. Soc., Chem. Commun. 1977, 888.
(27) Wilcox, C. F.; Craig, R. R. J. Am. Chem. Soc. 1961, 83, 3866.
(28) (a) Steinmetz, M. G.; Sequin, K. J.; Udayakumar, B. S.; Behnke, J. S. J.
Am. Chem. Soc. 1990, 112, 6601. (b) Bennett, M. A.; Pelling, S.; Robertson,
G. B.; Wickramasinghe, W. A. Organometallics 1991, 10, 2166. (c) Shahlai,
K.; Hart, H.; Bashir-Hashemi, A. J. Org. Chem. 1991, 56, 6905. (d) Shahlai,
K.; Hart, H.; Bashir-Hashemi, A. J. Org. Chem. 1991, 56, 6912.
The relative configuration of adducts 9 and 10 was in all
instances assigned on the bases of 1H NMR NOE experiments
(Supporting Information). These compounds serve as represen-
1
tative examples of the characteristic long-range H/1H interac-
(22) (a) Gre´e, D.; Gre´e, R.; Lowinger, T. B.; Martelli, J.; Negri, J. T.; Paquette,
L. A. J. Am. Chem. Soc. 1992, 114, 8841. (b) Paquette, L. A.; Negri, J. T.;
Rogers, R. D. J. Org. Chem. 1992, 57, 3947.
(23) Garbisch, E. W., Jr. J. Org. Chem. 1965, 30, 2109.
(24) Bryce-Smith, D.; Evans, E. H.; Gilbert, A.; McNeill, H. S. J. Chem. Soc.,
Perkin Trans. 2 1991, 1587.
(29) Newman, M. S. J. Org. Chem. 1961, 26, 2630.
(25) Adler, E.; Holmberg, K.; Ryrfors, L.-O. Acta Chem. Scand. 1974, 88, 883.
(30) Fieser, L. F.; Haddain, M. J. J. Am. Chem. Soc. 1964, 86, 2081.
9
J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004 16785

A R T I C L E S
Ohkata et al.
Table 2. Stereoselectivities Determined for Diels-Alder Reactions
Involving 2a and 2b
Table 3. Pseudo-First-Order Rate Constants (kobs) and
Second-Order Rate Constants (k2) for the Production of 9a in the
Diels-Alder Reaction of 1a with CP
run no.
diene
solvent
reaction conditions
21: 22a
yield, %b
c
d
solvent
Ta
102[subst]b
[CP]b
106kobs
106k2
A. spirolactone 2a:
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH3CN
CH3CN
CH3CN
CH3CN
CF3CH2OH
CF3CH2OH
CF3CH2OH
CF3CH2OH
37.2
37.2
25
5.26
7.21
7.83
7.11
6.71
8.18
7.93
8.32
5.65
4.68
6.06
4.41
2.63
3.61
4.03
3.56
3.36
4.12
4.02
3.80
1.13
1.41
1.82
1.42
9.83
3.75
14
15
16
CP
CP
CP
CH2Cl2
CH3CN
CF3CH2OH
35 °C, 3 days
35 °C, 3 days
37 °C, 4 h
96:4
91:9
96:4
99
72
87
1.22 × 10
3.38
1.92
1.61
7.08
5.66
2.57
1.53
7.73
20
5.73
B. spiroketone 2b:
CH2Cl2
CH3CN
CF3CH2OH
37.2
37.2
30
20
11
0
-10
-15
2.38 × 10
2.33 × 10
1.03 × 10
5.80
17
18
19
CP
CP
CP
25 °C, 7 days
25 °C, 7 days
25 °C, 7 days
66:34
61:39
60:40
7
7
82
2.11 × 102
1.07 × 102
4.54 × 10
2.37 × 10
1.86 × 102
7.56 × 10
2.49 × 10
1.67 × 10
a The ratios were determined by integration of the H NMR spectra of
unpurified field fragments. b Isolated yields after chromatographic purifica-
tion.
1
proper magnitude (5.9, 4.8 Hz).31 The coupling constant between
the pair of fused-ring protons is 9.0 Hz. In contrast, no coupling
between the fused-ring and bridgehead protons is seen in the
1H NMR spectrum of 20, thereby defining it as an exo adduct.
The relative configuration of the spirotetrahydrofuran unit in
18 and 20 was established as before by X-ray and NOESY
a °C. b bmol/L. c 1/s. d l/mol‚s.
Table 4. Pseudo-First-Order Rate Constants (kobs) and
Second-Order Rate Constants (k2) for the Production of 21a in the
Diels-Alder Reaction of 2a with CP
c
d
solvent
Ta
102[subst]b
[CP]b
106kobs
106k2
17
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH3CN
CH3CN
CH3CN
CH3CN
CF3CH2OH
CF3CH2OH
CF3CH2OH
CF3CH2OH
43
40
34.1
30.2
30.2
10
43
37.2
19
19
52
4.21
4.87
1.95
2.37
6.24
1.33
4.19
4.41
14.6
8.18
0.45
2.02
3.67
3.55
0.34
0.97
0.39
0.47
1.25
2.65
0.32
0.32
2.92
1.64
0.03
0.23
0.73
0.71
2.05 × 10
4.90 × 10
1.18 × 10
1.36 × 10
3.08 × 10
1.40 × 10
7.26
6.02 × 10
5.03 × 10
3.04 × 10
2.87 × 10
2.48 × 10
5.28
2.24 × 10
1.29 × 10
3.43
methods, respectively.
4.13
1.00 × 10
4.74
2.90
4.46 × 102
2.84 × 102
6.63 × 10
4.67 × 10
1.48 × 104
1.20 × 103
9.05 × 10
6.58 × 10
0
-20
-30
a °C. b bmol/L. c 1/s. d l/mol‚s.
Information). Therefore, there exists no doubt that 2a and 2b,
like 1a-c, possess an inherent structural feature that causes all
of these spiro dienones to capture cyclopentadiene with a kinetic
preference for that face syn to oxygen.
π-Facial Selectivity Characteristics Exhibited by 2a and
2b. The spiro lactone system 2a was found to exhibit appreciable
diastereoselectivity upon reaction with CP (Table 2). The levels
of selectivity favoring 21a exceeded the 90% level irrespective
of the reaction medium (runs 14-16). In contrast, the [4+2]
cycloaddition of CP to spiro ketone 2b occurs with the lowest
level of discrimination (60:40) observed thus far in this
investigation (runs 17-19). The facial selectivity involving 2b
Kinetic Parameters for the Cycloaddition of 1a and 2a to
Cyclopentadiene. To evaluate relative reactivities in different
organic solvents (CH2Cl2, CH3CN, CF3CH2OH), rate measure-
ments of the Diels-Alder reactions of 1a and 2a with
cyclopentadiene were carried out and kinetic parameters were
calculated. Substrates 1a and 2a were treated with an excess of
cyclopentadiene at a thermostated temperature and the appear-
ance of major products 9a and 21a, respectively, was monitored
by HPLC. The pseudo-first-order (kobs) and second-order rate
constants (k2) for these processes are summarized in Tables 3
and 4.
The kinetic parameters in Table 5 were evaluated on the basis
of Eyring plots (Figures S1 and S2: Supporting Information).
The Gibbs free energies of activation (∆G‡) followed the order
CH3CN∼CH2Cl2 . CF3CH2OH for 1a and CH3CN > CH2Cl2
. CF3CH2OH for 2a, respectively. The activation free energies
for 2a were smaller than those associated with 1a when
compared in the same solvent. As anticipated from the bimo-
lecularity of the Diels-Alder reaction, all of the activation
entropies (∆S‡) were significantly negative. The (∆S‡) value
for the reaction of 1a in CH2Cl2 is particularly noteworthy. The
smallest free energy of activation (∆G‡), which is observed for
is not generated by the presence of an ethereal-type oxygen atom
positioned homoconjugatively to the dienone part structure.
Greater spatial separation between the reaction centers is also
at play. Detailed NOE data were secured for 21b and 22b. These
measurements compared remarkably closely with the long-range
interactions observed for 21a and 22a (Supporting Information).
The highly crystalline nature of 21a provided the opportunity
to derive added structural proof by X-ray methods (Supporting
(31) Marchand, A. P. Stereochemical Applications of NMR Studies in Rigid
Bicyclic Systems; Verlag Chimie International: Deerfield Beach, FL, 1982.
9
16786 J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004

Syn Oxygen Phenomenon
A R T I C L E S
Table 5. Kinetic Parameters for the Production of 9a and 21a in the Diels-Alder Reactions of 1a and 2a with CP, Respectively
1a
f
9a
2a
f
21a
a
a
solvent
∆
G‡
∆
H‡a
∆
S‡b
∆
G‡
∆
H‡a
∆
S‡b
298
298
CH2Cl2
CH3CN
CF3CH2OH
25.2 ( 1.52
25.2 ( 5.8
21.8 ( 1.2
7.98 ( 0.8
14.6 ( 2.9
13.3 ( 0.6
-57.8 ( 2.5
-35.4 ( 9.6
-28.6 ( 2.1
23.9 ( 1.1
24.7 ( 1.7
20.8 ( 2.8
12.5 ( 0.5
14.1 ( 0.9
10.3 ( 1.3
-38.5 ( 1.8
-35.5 ( 2.9
-35.0 ( 4.9
a kcal/mol. b e.u.
the reaction of 2a in trifluoroethanol, may be attributed to
hydrogen bonding at the transition state. In particular, the
presence of a carbonyl group in the lactone ring of 2a may
enhance stabilization more effectively than that available to 1a
by hydrogen bonding. This effect would give rise to a small
activation enthalpy (∆H‡).
Consequences of Lewis Acid Catalysis. Reaction of spiro
lactone 2a with CP in the presence of 0.7 equiv of anhydrous
zinc chloride in CH2Cl2 at 25 °C for 2 days afforded exclusively
the endo cycloadduct 21a (>99%) in 96% yield. The sluggish-
ness exhibited by 1,3-cyclohexadiene (CH) and 2,3-dimethyl-
butadiene (DB) toward spirolactone 2a prompted us to probe
the consequences, if any, of Lewis acid catalysis on these [4+2]
cycloadditions. As expected, the co-addition of 0.6-0.7 equiv
of anhydrous zinc chloride in CH2Cl2 led to the consumption
of these dienes at room temperature, although at very different
rates. For CH, the conversion to 23 and 24 was essentially
complete after 5 days. For DB, a comparable level of completion
required 5 days. In both examples, the overwhelmingly major
diastereomer was that resulting from π-facial stereoselectivity
syn to oxygen. This selectivity was opposite that of Liotta’s
cyclohexadienone 4-tertiary carbinol system in the presence of
SnCl4, which favored facial attack opposite to the oxygen atom.6a
These workers rationalized the observed diastereoselectivity
under Lewis acid conditions to steric hindrance introduced upon
complexation of SnCl4 to the hydroxyl group. In the present
system, ZnCl2 is a considerably weaker Lewis acid than SnCl4
and furthermore is generally known to prefer coordination to a
carbonyl moiety rather than to an ether group.
Figure 1. Stereoselectivities observed for the 1,2-addition of CH3MgBr
to 1a, 2a, and 28a. Note that R attack is favored for nucleophilic capture
at C-1.
configuration of 23 was established by X-ray crystallographic
analysis (Supporting Information). The stereochemistry of 25
was likewise secured by crystallographic examination of its
epoxide 27. The extent to which 2a adopts the indicated reaction
pathway is impressive not only on its own merits, but also
because it reflects an enhancement of the stereoselectivity pattern
that is operative in the absence of a Lewis acid promoter.
4-Substituted-4-methyl-2,5-cyclohexadienones as Dieno-
philes. The spatial orientation of the polar substituents in 28a-
c20c,32-34 can be regarded to be closely similar to the situation
in 1 and 2. Methoxy derivative 28a and ester 28c are able to
populate orientations of their polar substituent that are not
available to the spiro analogues. This somewhat increased steric
demand is recognized for 28a
to result in a deceleration of nucleophilic attack. The level of
π-facial stereoselection observed for approach of the methyl
Grignard reagent (note that π-attack on the opposite side with
respect to the oxygen atom substituent is favored here) is also
eroded as reflected in Figure 1. To what extent and in what
direction will the release of those cyclic constrains embedded
in spirocyclic cyclohexadienones impact on dienophilic re-
sponse?
Table 6 summarizes the results of reacting 28a-c with CP in
CH2Cl2 as well as trifluoroethanol. As anticipated, all three
dienophiles underwent cycloaddition in the more polar solvent
after 2 days at room temperature. When CH2Cl2 was the reaction
medium (25 °C), methoxy ketone 28a failed to capture CP (run
The tentative structural assignments to 23 and 25 were
validated by means of NOE methods. Finally, the relative
9
J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004 16787

A R T I C L E S
Ohkata et al.
Table 6. Stereoselectivities Determined for [4+2] Cycloaddition of
CP to 28a-c
two exo transition state options were found as expected to be
less stable than their endo counterparts. The most stable
alternative involved endo addition syn to the heteroatom. This
isomer, which features positioning of the oxygen atom in the
interior of the developing adduct in a notably asynchronous
manner offers positive electrostatic attraction rather than
untoward repulsions. These authors concluded that the observed
high levels of stereoselectivity arise from the lower steric
demands of oxygen relative to a methylene group.
run no.
solvent
reaction conditions
29: 30a
yield, %b
A. methoxy ketone 28a:
20
21
CH2Cl2
CF3CH2OH
25 °C, 3 days
25 °C, 2 days
0
70
91:9
B. cyano ketone 28b:
25 °C, 3 days
22
23
CH2Cl2
CF3CH2OH
97:3
96:4
61
78
25 °C, 2 days
C. keto ester 28c:
25 °C, 3 days
To investigate this matter further, MO calculations on spiro
compounds 1, 2, and their related congeners 28 were performed
at the HF/6-31 G* level.36 The LUMO orbitals of dienophiles
1a-c, 2a,b, and 28a-c are illustrated in Figure 2. The size of
the lobes of the LUMOs at C3 (and C5) appear to be similar in
size with respect to the plane of symmetry (cyclohexadienone
ring) and thus cannot directly account for the high selectivity.
However, there is another feature to take into consideration. A
common feature of the LUMOs is that the orbitals of the σ bonds
involving the C4 carbon are all out of phase in relation to the
neighboring C3 (or C5) π lobe. Furthermore, the shape of the
orbital is also important. The LUMO orbitals associated with
C4-CH2 or C4-CH3 are more extended in the direction being
attacked by the diene than that of the C-X. (X ) O, CN, CO2-
CH3). For dienophiles with high facial selectivity, the difference
in size and shape between the two out-of-phase orbitals at C4
happens to be large with the larger of the two out-of-phase
orbitals residing opposite to the face that is preferentially
attacked. Thus, one explanation for the facial selectivity could
be that attack occurs upon the face which has less undesirable
secondary orbital interactions. For the series 1a-1c, the size
differences in the lobes at C4 do not vary significantly. However,
the unfavorable C4 interaction appears to be alleviated as the
size of the ring becomes smaller. Although the ineffectiveness
of FMO theory as a tool for explaining this phenomenon is now
widely accepted, the facial selectivity in these [4+2] cycload-
dition can be explained in large part by considering the
secondary orbital interaction of the FMO.37
The total electron densities of 1a-c, 2a, 2b, and 28a-c in
their respective ground states were evaluated by HF/6-31G*
calculations (Figure 3, Supporting Information). For 1a and 2a,
the electron cloud over CH2 of the spiro ring extends slightly
more to the direction of the reaction site than that of the oxygen
atom. Thus, the electron repulsion buildup between oxaspiro
compounds 1a and 2a and π-electrons of the diene during the
reaction may be a contributing factor to the π-facial selectivity
in the Diels-Alder reaction as described by Houk et al.10b For
1b-c, 2b, and 28a-c, extension of the electron cloud over both
sides of the reaction site is universally comparable. The low
reactivity of 1b-c, 2b, and 28a-c can also be explained in
terms of van der Waals repulsion.
24
25
CH2Cl2
CF3CH2OH
100: 0
97:3
trace
100
25 °C, 2 days
1
a The ratios were determined by integration of the H NMR spectra of
b
unpurified products. Isolated yields after chromatographic purification.
20) and keto ester 28c gave rise to only a trace of adduct 29c
(run 24).
The results derived from runs 21-25 reflect a pronounced
π-facial bias favoring bonding from the surface syn to X. The
relative configurations of 29 and 30 were readily deduced by
NOE methods (Supporting Information). Although the Diels-
Alder reactions were again accelerated in trifluoroethanol, little
or no change in the proportion of 29 to 30 was seen. Dienone
28a proved to be the least selective in its discriminating
capability.
Theoretical Assessment of the π-Facial Stereoselectivity.
Notwithstanding considerable variation in the substitution
pattern, including that of the monocyclic systems 28a-c, no
switch away from the adoption of facial selectivity favoring
predominant â attack was seen in any example. Therefore,
predominant dienophile approach syn to the heteroatomic
substituent constitutes a general reactivity trend for 4,4-
disubstituted cyclohexadienones whether they be spirocyclic or
monocyclic. This remarkable kinetic bias operates irrespective
of whether an alkoxy, acyloxy, cyano, or carbomethoxy sub-
stituent resides at C-4. The downward progression in size from
tetrahydrofuran to oxetane and finally to oxirane in the spiro
oxygenated heterocycle causes no deviation in an otherwise
consistent stereochemical outcome. The ineffectiveness of FMO
theory as a tool for explaining this phenomenon is now widely
accepted.35 Kahn and Hehre have suggested more recently that
electrostatic interactions likely play a major role in dictating
the diasterofacial selectivity of Diels-Alder reactions.9a
Houk and co-workers have calculated the energies of all four
possible transition states associated with [4+2] cycloadditions
to 4,4-disubstituted cyclohexadienones at the PM3 level.7b The
Atomic distances between the proton at C10 of the attacking
diene and the closest atom within the X or Y moiety in the
transition state should be related to the van der Waals repulsion
(Table S2, Supporting Information). The atomic distances in
the transition states are shorter than the sum of van der Waals
(32) (a) Schultz, A. G.; Macielag, M. J. Org. Chem. 1986, 51, 4983. (b) Schultz,
A. G.; Lavieri, F. P.; Macielag, M.; Plummer, M. J. Am. Chem. Soc. 1987,
109, 3991. (c) Schultz, A. G.; Harrington, R. E.; Macielag, M.; Mehta, P.
G.; Taveras, A. G. J. Org. Chem. 1987, 52, 5482.
radii of the corresponding atoms, except for those of TS1c-9c
,
TS1c-10c, TS28b-29b, and TS28b-30b (beyond the van der Waals
(33) Marx, J. N.; Zuerker, J.; Hahn, Y.-S. P. Tetrahedron Lett. 1991, 32, 1921.
(34) Danishefsky, S.; Yan, C.-F.; Singh R. K.; Gammill, R. B.; McCurry, P.
M.; Fritsch, N.; Clardy, J. J. Am. Chem. Soc. 1979, 101, 7001.
(35) For reviews, consult: (a) Anh, N. T. Tetrahedron 1973, 29, 3227. (b)
Reference 15a.
(36) The calculations were performed with PC Spartan Pro, Wavefunction, Inc.,
Irvine, CA 92715.
(37) Huang, X. L.; Dannenberg, J. J.; Duran, M.; Bertra´n, J. J. Am. Chem. Soc.
1993, 115, 4024.
9
16788 J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004

Syn Oxygen Phenomenon
A R T I C L E S
Figure 2. LUMO’s and HOMO’s of the dienophiles 1a-c, 2a, 2b, and 28a-c evaluated by HF/6-31G* calculations.
radial distance). In the transition states for the reaction of 2b to
give the adducts 21b and 22b; however, the distances are 2.30
Å for 21b and 2.33 Å for 22b, respectively, indicating their
values to be essentially identical. Thus, the selectivity in the
reactions (60/40 in CF3CH2OH) of 2b that cannot be disregarded
is better explained in terms of electronic effects rather than steric
effects (electron-donating ability: CH2-CH2 > CH2-CO).
Although the 60/40 selectivity was very small, the electron-
donating ability (CH2-CH2 > CH2-CO > CH2-O) at C4
position is qualitatively reflected in the selectivity. This explana-
tion is not inconsistent with the Cieplak model (Figure 4).38
These results show the major adducts 9a-c, 21a,b, 23, 25, and
29a-c to be favored both kinetically and thermodynamically.
The differences (δ∆Hcalcd) in the activation enthalpies between
the major and minor reactions for 1a, 1b, 21a, and 29a-c are
more than 4 kcal/mol, while those for 1c and 21b are 2.76 and
0.60 kcal/mol, respectively. The calculated sense of the Diels-
Alder reactions is clearly consistent with that of the observed
reactions, although the degree of difference itself (δ∆Hcalcd
)
varies to some extent.
Wipf et al. pointed out that the major product in the 1,2-
nucleophilic addition of 1a and 2a with Grignard reagents may
arise from favorable cancellations of dipole moment within the
reactants as shown in Figure 1.9c-e Although the dipole moments
of major adducts 9a-c, 21a,b, 23, and 25 and their correspond-
ing TS structures are smaller than those of the minor congeners
10a-c, 22a,b, 24, and 26 and the transition state structures
leading to the corresponding products (Table 7, entries 1-14),
the dipole moments of the transition states in the Diels-Alder
reactions of 28a-28c with cyclopentadiene are not necessarily
consistent with the observed high facial selectivities. Thus,
rationalization based upon dipole moment does not appear
Thermodynamic parameters (∆H° and ∆H*calcd) and dipole
moment values (µ) for 1a-c, 2a,b, and 28a-c were determined
by ab initio calculations at the HF/6-31G* basis set level (Table
7). The calculated enthalpy values (∆H°) are consistent with
the observed major adducts being thermodynamically more
stable than the minor counterparts. The activation enthalpies
(∆H*calcd) were calculated on the basis of the corresponding
heats of formation between the transition state and the reactant.
(38) (a) Cieplak, A. S. J. Am. Chem. Soc. 1981, 103, 4540. (b) Cieplak, A. S.;
Tait, B. D.; Johnson, C. R. J. Am. Chem. Soc. 1989, 111, 8447.
9
J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004 16789

A R T I C L E S
Ohkata et al.
Figure 4. Cieplak model for the Diels-Alder reaction.
(major products) at an antiperiplanar position. Selected dihedral
angels ( X-C4-C5-C10 and Y-C4-C5-C10) involving
atoms in anti and syn relationships were compared (Table S3,
Supporting Information). The angles for anti-positioned atoms
reside in the border region between anti-periplanar and anti-
clinal while those for syn-positioned atoms are between syn-
periplanar and syn-clinal. In each example, the dihedral angle
of the minor transition state arrived at by R-attack is larger than
that by â-attack.
The anti dihedral angles of the transition states for the major
Diels-Alder products of 1b and 1c (136.9° for 9b and 111.3°
for 9c, respectively) are smaller than those (142.1-151.2°) of
the other systems (1a, 2a, 28a, 28b, and 28c). Likewise, the
syn-periplanar dihedral angle of 1c (44.1° for 9c) is larger than
that (28.1-38.2°) of the other systems (1a, 1b, 2a, 28a, 28b,
and 28c). Since the most favorable dihedral angle for hyper-
conjugation is 180°, this structural feature may be one of the
factors underlying the somewhat lower stereoselectivity (86/
16-81/19) of 1b and 1c relative to that (92/8-96/4) of the other
systems.39 On the other hand, the dihedral angles of 29a,b
(150.6° and 148.6° for the major product and 153.3° and 153.0°
for the minor product, respectively) are similar to each other.
In the 29a-c subset, the electronic factor may perhaps be more
effective in stereoselectivity control relative to the steric factor
(anti-periplanarity).
Figure 3. Total electron density in the ground state of dienophile (1a-c,
2a, 2b, and 28a-c at the 6-31G* level.
Table 7. HF/6-31G* Calculations for the Diels-Alder Reaction of
1a-c, 2a-c, and 28a-c
substrate
transition state
product
b
b
entry
µa
diene
∆H*calcd
µa
product
∆H
°
µa
Selected bond lengths (or distances) around the reaction site
(C5-C10, C4-X, and C4-Y) in the transition states are
summarized in Figure 5 (Table S4, Supporting Information). A
comparison of these bond lengths reveals that the anti-bond
lengths for both major (C4-Y) and minor (C4-X) products
are 0.01-0.02 Å longer than the corresponding syn-bonds
lengths (C4-X and C4-Y, respectively), suggesting effective
orbital interaction of these anti-C-C bonds with the newly
forming C5-C10 bonds. The C5-C10 distances in the major
products are 0.03-0.07 Å shorter than those of the minor
products except for the case of TS2b-21b and TS2b-22b transition
states where selectivity was low. The shortening of the newly
forming bonds may partially contribute to the stabilization of
the transition state for the major product relative to that of the
minor product. From a stereoelectronic point of view, the
shortening of the C5-C10 distance peculiar to the transition
states for the major adducts may be attributed to the larger
electron-donating ability of the C-C bond relative to the C-O
bond (σC-C > σC-O).39 On the other hand, from a structural
point of view, the lengthening of the C5-C10 distance in the
minor products compared with the major products by 0.03-
0.07 Å may be attributed to the steric repulsion between the
dienophile and the diene in the transition state. In the case of
TS2b-21b and TS2b-22b, the steric environments around the
1
2
3
4
5
6
7
8
9
10
11
12
13
14
1a 5.45 CP TS1a-9a
TS1a-10a
1b 4.76 CP TS1b-9b
TS1b-10b
1c 3.72 CP TS1c-9c
TS1c-10c
2a 4.17 CP TS2a-21a
TS2a-22a
39.53 4.77 9a
43.53 5.90 10a
37.89 4.03 9b
41.86 5.83 10b
38.33 3.10 9c
41.09 5.14 10c
37.75 1.51 21a
42.05 7.37 22a
45.55 1.66 23
50.10 7.34 24
45.56 1.74 25
50.11 7.32 26
44.19 1.62 21b
44.79 5.11 22b
-13.29 4.53
-10.69 5.25
-13.49 3.86
-11.83 3.75
-13.81 3.08
-13.31 2.65
-14.51 2.45
-12.05 6.09
-20.00 1.89
-18.00 6.50
-37.42 3.53
-36.65 4.92
-8.99 1.41
-8.76 4.00
-8.73 4.17
-8.06 4.03
-12.53 1.84
-10.56 5.20
-9.61 3.37
-9.42 2.06
2a CH
TS2a-23
TS2a-24
TS2a-25
TS2a-26
2a DB
2b 3.29 CP TS2b-21b
TS2b-22b
15 28a 1.42 CP TS28a-29a 43.91 4.66 29a
16 TS28a-30a 47.45 4.55 30a
17 28b 3.63 CP TS28b-29b 39.79 1.42 29b
18 TS28b-30b 43.59 6.58 30b
19 28c 5.25 CP TS28c-29c
20 TS28c-30c
42.44 4.32 29c
46.35 3.99 30c
a dipole moment: D. b calculated from heat of formation: kcal/mol.
sufficient because the Diels-Alder reactions are not as polar
as the nucleophilic reactions.
Figure 5 illustrates the molecular structures of the Diels-
Alder transition states involving dienophiles 1a-c, 2a,b, and
28a-c as evaluated by HF/6-31G* calculations. All of the
structures show capability for hyperconjugative interaction
between the newly forming C-C bond (σ*-orbital) (C5-C10)
at C10 and the C4-X bond (minor products) or the C4-Y bond
(39) Li, H.; Silver, J. E.; Watson, W. H.; Kashyap, R. R.; le Noble, W. J. J.
Org. Chem. 1991, 56, 5932.
9
16790 J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004

Syn Oxygen Phenomenon
A R T I C L E S
Figure 5. Molecular structures of the Diels-Alder transition state evaluated by HF/6-31G* calculations.
9
J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004 16791

A R T I C L E S
Ohkata et al.
the rate acceleration observed in reactions involving CF3CH2-
OH can be rationalized in terms of the protic solvent behaving
as a Brønsted acid to activate the dienophile 1a and 2a or to
stabilize the transition state rather than ground state. This
explanation is not inconsistent with the activation parameters
of the reactions performed in CF3CH2OH.
Conclusion. The diastereofacial selectivity operating in the
Diels-Alder additions involving spirocyclic cross-conjugated
cyclohexadienones with dienes of varying reactivity has been
investigated. The study has included the ether series 1a-c as
well as the lactone/ketone pair 2a/2b. In all cases, the preferred
[4+2] cycloaddition pathway consisted of bonding from that
π-surface syn to the oxygen atom. 4-Substituted-4-methyl-2,5-
cyclohexadienones 28a9c (monocyclic systems) were also
examined and found to undergo bond formation preferentially
from the face bearing the more electron-withdrawing of the two
groups at the 4 position. The facial selectivity is understandable
in terms of seconary orbital interactions between the HOMO
of diene and the LUMO of dienophile and the orbital interactions
(effective hyperconjugation) between the newly forming bond
and the higher 4-anti-C-C4 σ-orbital (the more electron-
donating bond) relative to the counterpart 4-anti-substituent-
C4 σ-orbital in the transition state.
Figure 6. Solvent effects in Diels-Alder reaction of 1a, and 2a with CP.
reaction sites are virtually the same. Consequently, the observed
stereoselectivity can be considered to result from stereoelectronic
factors (difference between -CH2CO and -CH2CH2 at the anti-
position) rather than steric repulsion.
According to the calculated dipole moments compiled in
Table 7, solvent effects in the Diels-Alder reaction of 1a and
2a with cyclopentadiene can be explained in terms of the
differences in solvation between the ground and transition states
(the rate-determining and the product-determining steps) as
illustrated in Figure 6. The dipole moments (µ) of the stable
conformers of 1a and 2a in the ground state are 5.45 and 4.17
D, respectively, while those of the transition states (TS1a-9a and
TS2a-21a) leading to products 9a and 21a are 4.77 and 1.51 D,
respectively. From the dipole moment of TS2a-21a, especially,
it can be anticipated that cycloaddition will proceed less readily
in a more polar solvent such as CH3CN relative to CH2Cl2, since
CH3CN should stabilize the ground state of dienophile 2a by
solvation more than CH2Cl2. In fact, the rates of 2a in CH2Cl2
are faster than those in polar CH3CN. Furthermore, the higher
stereoselectivity in CH2Cl2 compared with CH3CN can also be
rationalized in terms of transition state dipole moments. The
transition state dipole moments are µ ) 4.77 D for TS1a-9a
(major product) and 5.91 D for TS1a-10a (minor product), and
µ ) 1.51 D for TS2a-21a (major product) and 7.37 D for TS2a-22a
(minor product). The adduct arising from the less polar TS
predominates and is expected to be favored more in the less
polar CH2Cl2 relative to CH3CN (Figure 6). On the other hand,
Acknowledgment. We are grateful to Dr. Yoshikazu Hiraga,
Hiroshima Prefectural Institute of Science and Technology, for
some of the 500 MHz NMR measurements and to F-TECH,
Inc. for a gift of trifluoroethanol. Partial financial support of
this work was provided through a Grants-in-Aid for Scientific
Research (No. 11740355 and No. 14740349) provided by the
Japan Society for the Promotion of Science and a partial
scholarship (to B.B.S.) from the Undergraduate Honors Program
at OSU.
Supporting Information Available: General experimental,
spectroscopic characterization of all adducts, crystallographic
data for 14b,c, 21a, 23, and 27 (including Table S1), Figures
S1 and S2, as well as relevant NOE of 9a, 10a, 21a,b, 22a,b,
23, 29a,b, and 30a,b, and last Tables S2, S3, and S4. This
material is available free of charge via the Internet at
http:pubs.acs.org.40
JA047027T
(40) Crystallographic data for structures in this paper have been deposited with
the Cambridge Crystallographic Data Center; 21a: CCDC-193077, 23:
CCDC-193078; 27: CCDC-193079. For a discussion of the structures of
14b and 14c, consult Gallucci, J. C.; Tamura, Y.; Paquette, L. A. Acta
Cryst. 2004, C60, 656.
9
16792 J. AM. CHEM. SOC. VOL. 126, NO. 51, 2004