Stereoselectivity in Diels-Alder reactions of diene-substituted N-alkoxycarbonyl-1,2-dihydropyridines

Article abstract of DOI:10.1021/jo0700575

(Chemical Equation Presented) Diene substituent effects on the regiochemical and stereochemical outcomes of uncatalyzed Diels-Alder reactions of N-alkoxycarbonyl-1,2-dihydropyridines with both styrene and methyl vinyl ketone (MVK) were studied. Alkyl subs

Full text of DOI:10.1021/jo0700575

Stereoselectivity in Diels-Alder Reactions of Diene-Substituted
N-Alkoxycarbonyl-1,2-dihydropyridines
Grant R. Krow,*,† Qiuli Huang,† Steven W. Szczepanski,† Fredrick H. Hausheer,‡ and
Patrick J. Carroll§
Department of Chemistry, Temple UniVersity, Philadelphia, PennsylVania 19122, BioNumerik
Pharmaceuticals, Inc., San Antonio, Texas 78229, and Department of Chemistry, UniVersity of
PennsylVania, Philadelphia, PennsylVania 19104
grantkrow@aol.com
ReceiVed January 10, 2007
Diene substituent effects on the regiochemical and stereochemical outcomes of uncatalyzed Diels-Alder
reactions of N-alkoxycarbonyl-1,2-dihydropyridines with both styrene and methyl vinyl ketone (MVK)
were studied. Alkyl substitution on the diene in all cases examined resulted in a kinetic preference for
7-endo isomers (7-phenyl 51-96% exo and 7-acetyl 54-96% exo). For both dienophiles, the highest
stereoselectivities (g89% endo) were observed with 5-methyl or 6-methyl substituents in the dihydro-
pyridine. Theoretical calculations of the energies of gas phase endo and exo transition states at the RHF/
3-21G(*) predict that total entropy, ∆Stotal, considerations favor endo cycloadducts for both dienophiles
with DHP, while total energy considerations, ∆Eq , favor endo cycloadducts for styrene and exo
o
cycloadducts for MVK. At this level, favored endo-phenyl isomers are correctly predicted for styrene
reactions, but the calculation of 7-acetyl exo or endo isomer dominance is diene-substituent-dependent
for MVK reactions. The preference for endo addition of MVK to the parent, 5-methyl, and 6-methyl-
DHPs was successfully predicted by calculations at the B3LYP/6-31G* theory level.
Introduction
2-azabicyclo[2.2.2]oct-5-enes. These structures have proven to
be useful in the synthesis of a range of interesting and useful
structures.1-7 The regiochemistry for this thermal cycloaddition,
at least for alkene-unsubstituted 1,2-dihydropyridines 1 reacting
with monosubstituted dienophiles 2, has the nitrogen atom near
The ease of synthesis of N-alkoxycarbonyl-1,2-dihydropy-
ridines1 (DHPs) and their ability to react with a variety of
dienophilic reagents have led to their wide use as precursors of
† Temple University.
‡ BioNumerik Pharmaceuticals, Inc.
(3) Tetrahydroisoquinoline alkaloids: (a) MaGee, D. I.; Lee, M. L.
Tetrahedron Lett. 2001, 42, 7177-7180. (b) MaGee, D. I.; Lee, M. L. Synlett
1997, 7, 786-788. (c) Wender, P.; Schaus, J.; Torney, D. Tetrahedron Lett.
1979, 27, 2485-2488. (d) Wender, P. A.; Schaus, J. M.; White, A. W. J.
Am. Chem. Soc. 1980, 102, 6157-6159. (e) Mariano, P. S.; Dunaway-
Mariano, D.; Huesmann, P. L. J. Org. Chem. 1979, 44, 124-133. (f)
Mariano, P. S.; Dunaway-Mariano, D.; Huesmann, P. L.; Beamer, R. L.
Tetrahedron Lett. 1977, 18, 4299-4302. (g) Baxter, E. W.; Labaree, D.;
Chao, S.; Mariano, P. S. J. Org. Chem. 1989, 54, 2893-2904. (h) Comins,
D. L.; Al-war, R. S. J. Org. Chem. 1992, 57, 4098-4103. (i) Comins, D.
L.; Brooks, C. A.; Al-awar, R. S.; Goehring, R. R. Org. Lett. 1999, 1, 229-
231.
(4) Homoepibatidines: Krow, G. R.; Cheung, O. H.; Hu, Z.; Huang,
Q.; Hutchinson, J.; Liu, N.; Nguyen, K. T.; Ulrich, S.; Yuan, J.; Xiao, Y.;
Wypij, D. M.; Zuo, F.; Carroll, P. J. Tetrahedron 1999, 55, 7747-7756.
(5) Piperidines: (a) Sales, M.; Charette, A. B. Org. Lett. 2005, 7, 5773-
5776. (b) Arakawa, Y.; Murakami, T.; Ozawa, F.; Arakawa, Y.; Yoshifuji,
S. Tetrahedron 2003, 59, 7555-7563.
§ University of Pennsylvania.
(1) (a) Weinstein, B.; Lin, L. C.; Fowler, F. W. J. Org. Chem. 1980, 45,
1657-1661. (b) Fowler, F. W. J. Org. Chem. 1972, 37, 1321-1323. (c)
Matsumura, Y.; Nakamura, Y.; Maki, T.; Onomura, O. Tetrahedron Lett.
2000, 41, 7685-7689. For reviews, see: (d) Eisner, U.; Kuthan, J. Chem.
ReV. 1972, 72, 1-42. (e) Stout, D.; Meyers, A. I. Chem. ReV. 1982, 82,
223-243. (f) Lavilla, R. J. Chem. Soc., Perkin Trans. 1 2002, 1141-1156.
(2) Iboga alkaloids: (a) Raucher, S.; Bray, B. L. J. Org. Chem. 1985,
50, 3236-3237. (b) Raucher, S.; Lawrence, R. F. Tetrahedron Lett. 1983,
24, 2927-2930. (c) Sundberg, R. J.; Bloom, J. D. J. Org. Chem. 1981, 46,
4836-4842. (d) Sundberg, R. J.; Cherney, R. J. J. Org. Chem. 1990, 55,
6028-6037. (e) Sundberg, R. J.; Amat, M.; Fernando, A. M. J. Org. Chem.
1987, 52, 3151-3159. (f) Krow, G. R.; Shaw, D. A.; Lynch, B.; Lester,
W.; Szczepanski, S. W.; Raghavachari, R.; Derome, A. E. J. Org. Chem.
1988, 53, 2258-2262; See also: Hodgson, D. M.; Galano, J.-M. Org. Lett.
2005, 7, 2221-2224. (g) Marazano, C.; Fourrey, J. L.; Das, B. C. J. Chem.
Soc., Chem. Commun. 1981, 37-39.
10.1021/jo0700575 CCC: $37.00 © 2007 American Chemical Society
Published on Web 03/30/2007
3458
J. Org. Chem. 2007, 72, 3458-3466

Diene-Substituted N-Alkoxycarbonyl-1,2-dihydropyridines
experimental details.6b,c Nevertheless, the cycloaddition results
with dihydropyridines 1 and alkyl acrylate 2a,b, acrolein (2c),
and MVK (2d) (entries 1-5) all indicate a greater kinetic
preference for endo cycloadducts 3a-e when compared to the
near 1:1 equilibrium ratio of 3/4 isomers having ester8 or
aldehyde3c substitution and the 3:2 preference for exo-acetyl
isomer 4e.3e Phenylvinylsulfone (2e) addition to 1,2-dihydro-
pyridine 1a forms a cycloadduct of undetermined stereochem-
istry (entry 6).9 During the development of synthetic approaches
to homoepibatidines, we discovered that a vinylpyridine 2f
undergoes cycloaddition with 1,2-dihydropyridine to give a
minor amount of exo cycloadduct 4g (entry 7).4
TABLE 1. Stereoselectivity for Thermal Cycloadditions of
N-Alkoxycarbonyl-1,2-dihydropyridines (DHP) 1 with
Monosubstituted Dienophiles 2 (CH2 ) CHZ)
yield
endo 3/exo 4 (%)
entry DHP alkene cycloadducts
R
Z
Thermal cycloadditions have been extended to include diene-
substituted 1,2-dihydropyridines 5 as well. In this regard,
substituted 1,2-dihydropyridine reactions with N-phenylmale-
imide11 and the intramolecular cycloadditions of 2-(3-butenyl)-
1,2-dihydropyridines have been carried out successfully with
substituents at all diene positions.7 Phenylvinylsulfone (2e) does
not afford cycloadducts with 6-methyl-1,2-dihydropyridine, but
gives 60-70% yields with 3-, 4-, or 5-methyldihydropyridines.12
Methyl vinyl ketone (2d) reacts with a 5-COOMe-1,2-
dihydropyridine to afford a 21:4 mixture of endo/exo cycload-
ducts (25% yield).7a For methyl acrylate (2a), the regiochemical
and stereochemical control exerted by diverse 3- and 5-substit-
uents on the cycloadditions of 1,2-dihydropyridines are shown
in Table 2.8,13 A 3-methyl (entry 1) has little effect upon
stereochemistry compared to the unsubstituted parent (Table 1,
entry 1); both are about 6:4. A mixture of 3-ethyl and 5-ethyl
dihydropyridines 5b/5c reacts with methyl acrylate to give
mixtures of stereoisomers of undetermined ratio (entry 2). The
preference for endo isomers is altered by a bulky 5-trimethylsilyl
group (entry 3), and there is now a 71:29 preference for exo-
carbomethoxy in the cycloadduct 7d. A 3-methoxy group on
the DHP (entry 4) causes a reversal of favored regiochemistry.
Since the methoxy group is a more powerful director than the
N-alkoxycarbonyl group of the DHP, the exo cycloadduct 9 is
now favored over the endo isomer 8.8 From these disparate
results, it was clear at the outset of the present study that DHP
substituents can have far reaching influences upon the regio-
chemistry and stereochemistry of cycloadditions.
When the present work was begun, there were no reports of
styrene cycloadditions with 1,2-dihydropyridines 1. We first
explored the possibility of addition of a styrene 2f to 1,2-
dihydropyridine 1a as a route to epibatidine homologues 6-aryl-
2-azabicyclo[2.2.2]oct-7-enes 3g/4g. When this cycloaddition
was shown to be successful (Table 1, entry 7), we wanted to
know more generally how alkyl groups on the DHP diene
carbons might influence exo/endo stereochemical outcomes for
reactions with styrenes. The results of this investigation are one
subject of this paper. During the course of our studies, the
investigation of substituent effects was extended to include DHP
cycloadditions of methyl vinyl ketone (2d) and substituted 1,2-
dihydropyridines. Attempts have also been made to assess our
experimental results by theoretical means.
1
2
3
4
5
6
7
1a
1a
1b
1a
1c
1a
1a
2a
2b
2a
2c
2d
2e
2f
3a and 4a
3b and 4b Me COOBn
3c and 4c t-Bu COOMe
3d and 4d Me CHO
Me COOMe
57:43
67:33
50:50
70:30
57:43
65a
39b
c
70d
89e
28f
44g
3e and 4e
3f (4f)
Et
COMe
Me SO2Ph
Me Ar
3g and 4g
79:21
a Xylene, reflux, 120 h, refs 5b,7c. b Xylene, reflux, 15 h, ref 5b.
c Toluene, reflux, no yield given, refs 6b,c. d Toluene, reflux, 28 h, ref 3c.
e Neat, 50 °C, 5 days, ref 3c. f Toluene, reflux, 60 h, ratio undetermined,
ref 9. g Ar ) 6-chloro-3-pyridyl, decalin, reflux, 12 h, ref 4.
the C7 dienophile substituent in the azabicycle. The stereo-
chemical outcome for uncatalyzed cycloadditions is to generally
give mixtures of products favoring endo isomers 3 in competi-
tion with exo isomers 4. The results of representative cycload-
ditions are shown in Table 1. Stereochemical outcomes for
methyl acrylate,8 acrolein,3c and methyl vinyl ketone3e cycload-
ditions with dihydropyridines have been shown to be of kinetic
origin.
Unlike the corresponding 1,3-cyclohexadiene that opposes a
methylene group as a steric impediment to an exo-oriented
dienophile substituent,10 it might seem reasonable to anticipate
that the lone pair of electrons on nitrogen and the N-acyl
substituent might play a favorable role in facilitating formation
of exo isomers. Indeed, for cycloadditions of N-methoxycar-
bonyl-1,2-dihydropyridine 1a with methyl acrylate 2a to give
a 57:43 endo/exo ratio of isomers 3a/4a in refluxing toluene
(entry 1), it was postulated that the nitrogen atom of the
dihydropyridine favorably interacts with the carbonyl of the ester
and that the carbonyl of the carbamoyl group might electrostati-
cally interact with the O atom of the ester carbonyl.5b This effect
appears to be balanced by a decreased preference for exo isomers
with an increase in size of the ester alkyl group (67:33 endo/
exo 3b/4b when R ) Bn, entry 2). Inconsistent with the
argument, an increased preference for exo adduct 4c containing
an N-Boc group has been reported, but without confirming
(6) Cyclohexanecarboxylic acids: (a) Arakawa, Y.; Tajima, M.; Arakawa,
Y.; Yoshifuji, S. Chem. Pharm. Bull. 2005, 53, 81-85. (b) Campbell, M.
M.; Mahon, M. F.; Sainsbury, M.; Searle, P. A.; Davies, G. M. Tetrahedron
Lett. 1991, 32, 951-954. (c) Campbell, M. M.; Sainsbury, M.; Searle, P.
A.; Davies, G. M. Tetrahedron Lett. 1992, 33, 3181-3184.
(7) Bridged azapolycycles: (a) Choi, Y.; White, J. D. J. Org. Chem.
2004, 69, 3758-3764. (b) Krow, G. R.; Lee, Y. B.; Raghavachari, R.;
Szczepanski, S. W.; Alston, P. V. Tetrahedron 1991, 47, 8499-8154.
(8) Sundberg, R. J.; Hamilton, G.; Trindle, C. J. Org. Chem. 1986, 51,
3672-3679.
(9) Krow, G. R.; Carey, J. T.; Cannon, K. C.; Henz, K. J. Tetrahedron
Lett. 1982, 23, 2527-2528.
(10) Equilibrium measurements for cyclohexa-1,3-diene cycloadducts
show endo adducts to be favored with methyl vinyl ketone [K(exo/endo)
) 0.41]: Ouellette, R. J.; Booth, G. E. J. Org. Chem. 1966, 31, 3065-
3067. Methyl acrylate [K(exo/endo) ) 0.46]: Ouellette, R. J.; Booth, G.
E. J. Org. Chem. 1965, 30, 423-425.
(11) Krow, G. R.; Cannon, K. C.; Carey, J. T.; Lee, Y. B.; Szczepanski,
S. W.; Ramjit, H. G. J. Heterocycl. Chem. 1985, 22, 131-135.
(12) Krow, G. R.; Alston, P. V.; Szczepanski, S. W.; Raghavachari, R.;
Cannon, K. C.; Carey, J. T. Synth. Commun. 1990, 20, 1949-1958.
(13) Acrylonitrile adds to N-methoxycarbonyl-4-hydroxyethyl-1,2-dihy-
dropyridine or its 3-methyl-4-hydroxyethyl counterpart to give 2.1:1 and
1.8:1 mixtures of exo/endo cycloadducts. The same substrates with methyl
acrylate, acrolein, or methyl vinyl ketone gave mixtures described as ranging
from 1:2 to 1:1 exo/endo cycloadducts without further experimental detail;
see ref 3a.
J. Org. Chem, Vol. 72, No. 9, 2007 3459

Krow et al.
TABLE 2. Stereochemical and Regiochemical Results from Cycloadditions of Substituted Dihydropyridines 5 and Methyl Acrylate (2a)a
endo/exo
ratio
endo/exo
equilibrium ratio
yield
(%)
entry
1,2-DHP
Rb
adduct(s)
R
1
2
5a
3-Me
3-Et
5-Et
5-SiMe3
3-OMe
6a/7a
6b/7b
6c/7b
6d/7d
8/9
4-Me
4-Et
6-Et
60:40
c
c
29:71
33:67
1:1
d
d
d
d
70
65
5b +
5cc
5d
3
4
6-SiMe3
60
25
5e
4-OMee
a Sealed tube, 120 °C, 48 h. See ref 8. b Other 3-isomers (R ) SMe, Cl, SiMe3, SnMe3, SnBu3, and COOMe) successfully reacted in 56-74% yields, but
regiochemical and stereochemical ratios of cycloadducts 6/7 were not determined. c A mixture of unreported ratio. d Not determined. e Ester is at C8.
Results and Discussion
Synthesis and Reaction of Substituted 1,2-Dihydropy-
ridines. The requisite 1,2-dihydropyridines for the present study
were prepared as described. N-Carbomethoxy-1,2-dihydropy-
ridine (1a),1b N-benzyloxycarbonyl-1,2-dihydropyridine (1b),2c
4-methyl-N-carbomethoxy-1,2-dihydropyridine (10d),14 and
4-phenyl-N-carbomethoxy-1,2-dihydropyridine15 (10g) have been
prepared previously. Other N-alkoxycarbonyl-1,2-dihydropy-
ridines were prepared according to the reported procedure by
best carried out on neat mixtures under argon containing 2.5-
fold excess of MVK at 50-60 °C for 5-6 days. In order to
obtain pure samples, the reaction mixtures were purified using
a combination of techniques including repeated silica gel column
chromatography, preparative TLC, medium-pressure reverse-
phase column chromatography, and HPLC. The endo/exo
stereoisomeric ratios were determined by 1H NMR proton
integrations, HPLC analyses, and/or column separations. Results
from all three methods usually were consistent. For example,
the ratio of 12a and 13a was 84:16 from 1H NMR integrations
of H1 protons, 82:18 from HPLC analysis, and 84:16 from
column separation.
Fowler,1b except for 5-methyl-N-carbomethoxy-1,2-dihydropy-
ridine (10c).2b The reduction of pyridine and its analogues with
sodium borohydride in the presence of methyl chloroformate
normally gives a mixture of N-substituted 1,2- and 1,4-
dihydropyridines. If the reaction temperature is controlled well
below -70 °C, the formation of 1,4-DHP isomers is minimized
and they can be removed by flash column chromatography. If
the C4 position of the pyridine ring is substituted, no 1,4-
dihydropyridine derivatives were detected or isolated. The
2-picoline did not react readily with methyl chloroformate and
sodium borohydride in methanol to give 6-methyl-1,2-DHP (10i)
when the temperature was below -70 °C, but if the reaction
temperature was raised to -50 °C, the reaction afforded 10i in
14% yield along with unreacted 2-picoline. All 1,2-dihydropy-
ridines were purified by the silica gel flash column chroma-
tography under argon prior to use.
Diels-Alder Cycloadditions of DHPs with Styrenes 11a-c
or Methyl Vinyl Ketone (2d). Identification and Analysis of
Product Mixtures. The cycloadditions of DHPs with styrenes
11 were carried out with 3-5 equiv of the dienophile in
refluxing decalin, boiling point 189-191 °C, under argon for
17-72 h. Decalin can be removed by passing the crude reaction
mixtures through a silica gel flash column and then rinsing with
an adequate amount of nonpolar solvent, such as cyclohexane.
The cycloadditions of methyl vinyl ketone (2d) to DHPs were
The structure elucidations and peak assignments of 7-acetyl-
and 7-aryl-2-carbomethoxy-2-azabicyclo[2.2.2]oct-5-enes were
derived from combinations of 1H NMR, 13C NMR, 1H-1H 2D
1
COSY, H-13C 2D COSY, and APT 1D NMR. The chemical
shift of each proton of the 2-azabicyclo[2.2.2]oct-5-ene skeleton
was assigned using decoupling experiments. The interconversion
of two conformational isomers of the N-carbamate was restricted
at ambient temperature, and two sets of peaks corresponding
1
to two different conformers appeared in the H NMR and 13C
NMR spectra. NMR experiments at high temperature (100 °C)
were used in order to unambiguously identify the coupling
patterns for the protons for the endo and exo stereoisomers.
1
It was interesting to note that the H NMR spectra of the
7-endo-aryl isomers did not clearly demonstrate two sets of
signals after flash column chromatography purifications and
therefore most peaks were broad, but if D2O was added, the
splitting pattern representing the two conformers was restored.
(14) Natsume, M.; Ogawa, M. Heterocycles 1980, 14, 615-618.
(15) Kurita, J.; Iwata, K.; Sakai, H.; Takashi, T. Chem. Pharm. Bull.
1985, 33, 4572-4580.
3460 J. Org. Chem., Vol. 72, No. 9, 2007

Diene-Substituted N-Alkoxycarbonyl-1,2-dihydropyridines
TABLE 3. Stereoselectivity for Cycloadditions of
SCHEME 1. N-Alkoxycarbonyl-1,2-dihydropyridine
Cycloadditions with Styrenes
N-Alkoxycarbonyl-1,2-dihydropyridines with Styrenesa
Substitutions
product ratio
yield
(%)b
entry
DHP
R
X
endo/exo
1
2
3
4
5
1a
1b
1a
1a
1a
Me
Bn
Me
Me
Me
Ph
Ph
84(12a)/16(13a)
86(12b)/14(13b)
80(12c)/20(13c)
79(12d)/21(13d)
79(3g)/21(4g)
32
50
29
30
44
Ph-p-OMe
Ph-m-NO2
6-Cl-3-pyridylc
a Reactions were run in refluxing decalin. b Isolated yields. c See ref 4.
The conformational isomers of 7-exo-aryl compounds always
presented two distinct sets of peaks in the H NMR spectra.
SCHEME 2. Substituted N-Alkoxycarbonyl-1,2-dihydro-
pyridine/Styrene Additions
1
The structure of 7-anti-phenyl isomer 12a, as determined by
high-temperature 1H NMR decoupling and 1H-13C 2D COSY
experiments, is indicative of the method used for stereochemical
assignments. There is a long-range W-plan coupling between
H8a and H3n protons in a 2-azabicyclo[2.2.2]oct-5-ene system.16
1
In the H NMR spectrum of 12a, the protons at δ 1.68 and
3.03 ppm are W-plan coupled with a coupling constant of 2.0
Hz. The multiplet at δ 1.68 ppm can thus be assigned to H8a,
and the multiplet at δ 3.03 ppm can be assigned to the H3n
proton. Further, there are coupling constants of 5.4 Hz for H8a
and the H7 proton, and 9.6 Hz for H8s and the H7 proton.
Therefore, the H7 proton has the cis configuration to the H8s
proton (larger coupling) and the trans configuration to H8a. Thus,
the 7-phenyl group of 12a has the endo configuration toward
the 5,6-alkene.
The Diels-Alder reactions of 1,2-DHPs and styrene have
been performed at a very high temperature of 190 °C. The ratio
of endo/exo products might result from either a kinetically
controlled or thermodynamically controlled reaction. A kinetic
study of the reaction of 1,2-DHP 1b and styrene (entry 2) has
shown that the product ratio remained between 90:10 (3 h) and
86:14 (24 h) endo/exo by HPLC during the 24 h course of the
cycloaddition. Additionally, when endo-Ph 12b (53 mg) was
heated at reflux in decalin under argon for 48 h, 40 mg of pure
endo-Ph 12b was recovered after chromatography. The exo-Ph
stereoisomer 13b was not observed by 1H NMR or HPLC
analysis. Similarly, when exo-Ph isomer 13b (55 mg) was heated
at reflux in decalin for 24 h, 34 mg of 13b was recovered after
chromatographic purification. The endo-Ph stereoisomer 3a was
not detected. The endo/exo products are therefore derived from
kinetically controlled Diels-Alder reactions. These results show
that, while there may be some decomposition of cycloadducts
over time, the stereoisomeric ratios are relatively invariant over
time and the endo and exo stereoisomers are not interconverting.
We next turned our attention to the impact of alkyl substit-
uents upon the endo/exo stereoselectivity, as shown in Scheme
2. The results are given in Table 4. The Diels-Alder cycload-
ditions of styrene with the parent (entry 1), the 3-methyl-1,2-
DHP (entry 2), and 3,5-dimethyl-1,2-DHP (entry 3) afford
similar endo/exo ratios of stereoisomers 12a-c/13a-c; how-
ever, minor amounts of the 8-endo-phenyl regioisomers 14e and
14f were also isolated from reactions of the 3-methyl DHPs
10a,b. This effect upon regiochemistry by the electron-releasing
methyl group at C3 is less significant than that observed for a
3-methoxyl group that was controlling for C8 isomers with
methyl acrylate/DHP cycloaddition (Table 2, entry 4).8
Stereochemical Results for DHP/Styrene Cycloadditions.
Styrene (11a) cycloadds to cyclohexa-1,3-diene under conditions
of photoinitiation in the presence of an electron transfer catalyst
to give a 10:1 endo/exo mixture in 13% yield, accompanied by
large amounts of styrene polymer.17a The same reaction occurs
electrochemically in 13-33% yield (19-30:1 endo/exo).17b
Despite apparent difficulties, our goal was to see if styrenes
could react with 1,2-dihydropyridines under thermal conditions
(Scheme 1). The results for unsubstituted 1,2-dihydropyridines
are in Table 3.
There were no significant differences observed in terms of
regioselectivity and stereoselectivity between 12a/13a and 12b/
13b as a result of changing the carbamate alkyl group from
methyl to the potentially bulkier benzyl group (entries 1 and
2). The phenyl of the benzyl group can apparently avoid steric
interaction with the styrene aryl group, and the stereochemistry
did not change significantly when electron-rich 4-methoxysty-
rene (entry 3) or electron-poor 3-nitrostyrene (entry 4) was used
in place of styrene as the dienophile to give 12c/13c and 12d/
13d. All reactions gave a similar ratio of about 80:20 for the
endo and exo products in similar yields (40 ( 11%).
(16) (a) Krow, G. R.; Rodenbaugh, R. Org. Magn. Reson. 1973, 5, 73-
75. (b) Krow, G. R.; Rodenbaugh, R.; Carmosin, R.; Figures, W.; Pannella,
H.; Devicaris, G.; Grippi, M. J. Am. Chem. Soc. 1973, 95, 5273-5280.
(17) (a) Photochemical: Mlcoch, J.; Steckhan, E. Angew. Chem. 1985,
97, 429-431. (b) Electrochemical: Mlcoch, J.; Steckhan, E. Tetrahedron
Lett. 1987, 28, 1081-1084.
The lone 5-methyl substituent of DHP 10c results in enhanced
89% endo selectivity in cycloadduct 12g (entry 4). This result
and that of the 3,5-dimethyl isomer (entry 3) are contrary to
J. Org. Chem, Vol. 72, No. 9, 2007 3461

Krow et al.
TABLE 4. Stereoselectivity for Cycloadditions of
TABLE 5. Stereoselectivity in the Cycloadditions of
Alkyl-N-Methoxycarbonyl-1,2-dihydropyridines with Styrenea
Alkyl-Substituted N-Alkoxycarbonyl-1,2-dihydropyridines with
Methyl Vinyl Ketone (2d)a
products
endo/exo/regioisomer
DHP Substitution
yield
(%)b
entry DHP
3
4
5
6
ratios
1
2
3
4
5
6
7
8
9
1a
H
H
H
H
H
Me
Et
t-Bu
Ph
Bn
H
H
H
Me
Me
H
H
H
H
H
H
84(12a)/16(13a)
75(12e)/21(13e)/4(14e)
78(12f)/8(13f)/14(14f)
89(12g)/11(13g)
55(12h)/45(13h)
62(12i)/38(13i)
32
36
52
38
47
53
22
30
29
45
10a Me
10b Me
H
H
H
H
H
H
H
H
10c
10d
10e
10f
10g
10h
10i
H
H
H
H
H
H
H
products
51(12j)/49(13j)c
54(12k)/46(13k)d
55(12l)/45(13l)d
product
substituent
endo/exo isomer
ratios
yield
(%)b
entry
DHP
R
1
2
3
4
5
6
7
8
9
1c
1a
1b
10a
10b
10c
10d
10e
10i
Et
parent
parent
parent
4-Me
4,6-di-Me
6-Me
5-Me
5-Et
1-Me
57(3e)/43(4e)
89c
63
73
51
27
22
81
69g
45
10
H
Me 96(12m)/4(13m)
Me
Bn
Me
Me
Me
Me
Me
Me
74(15a)/26(16a)
80(15b)/20(16b)d
68(15c)/32(16c)e
53(15d)/47(16d)f
90(15e)/10(16e)
60(15f)/40(16f)
54(15g)/46(16g)
96(15h)/4(16h)
a Reactions were run in refluxing decalin. b Isolated yields. c By NMR
proton integration endo/exo ) 55:45. d By NMR proton integration.
SCHEME 3. Cycloadditions of N-Alkoxycarbonyl-1,2-
dihydropyridines with Methyl Vinyl Ketone
a Reactions were run neat at 55 °C, 5 days. b Isolated yields. c See Table
1, entry 5 and ref 3e, 50 °C, 6 days, neat. d NMR proton integration and
HPLC endo/exo 70:30. e By NMR proton integration. f By NMR proton
integration endo/exo 54:46. g Reaction was run neat at 50 °C, 6 days, neat.
reaction centers of the TSs so as not to be the determinative
factor in controlling endo or exo product orientations.
The cycloadditions of MVK with 3-methyl-DHP 10a (entry
4) and 3,5-dimethyl-1,2-DHP 10b (entry 5) resulted in only
modest preferences for endo isomers 15c and 15d. However,
the 5-methyl-1,2-DHP (entry 6) showed a significant 90:10
preference for endo adduct 15e versus exo adduct 16e. The
cycloaddition yields from the 5-methyl-DHPs 10b and 10c are
quite low, however.
The 4-substituted 1,2-DHPs 10d and 10e (entries 7 and 8)
did give higher yields of 7-exo-acetyl products; however,
7-endo-acetyl stereoisomers 15f and 15g are the stereochemi-
cally favored products by only a slight margin. The Diels-
Alder cycloaddition of 6-methyl-1,2-DHP 10i and MVK
afforded 96:4 endo/exo products favoring 15h (entry 9). This
high preference for endo cycloadduct from the 6-methyl-DHP
10i previously was noted for its cycloaddition with styrene to
give cycloadduct 12m (see Table 4, entry 10).
Mariano demonstrated that the cycloaddition between the
parent 1,2-DHP 1c and MVK gives a kinetically derived endo/
exo 3e/4e ratio of 4:3,3e but the equilibrium endo/exo ratio is
2:3. In order for us to show that alkyl-substituted DHPs also
give kinetically controlled isomer ratios, stereoisomers 15g and
16g independently were subjected to the original reaction
conditions and were heated to 55 °C for 1 week under argon.
They also were heated at 105 °C for 24 h. No epimerization of
the 7-acetyl group was detected by 1H NMR in either case, and
decomposition of adducts was minimal. When the temperature
was raised to 130 °C for 2 days, endo cycloadduct 15g was
converted to a 65:35 mixture of 15g/16g (35% recovery).
Independently, the exo cycloadduct 16g was epimerized to a
19:81 mixture of 15g/16g (43% recovery). When epimerizations
occur, they do so at higher temperatures than our 55 °C reaction
temperature. Thus, the stereoselectivities of Diels-Alder reac-
tions between 1,2-DHP and MVK are kinetically controlled.
expectations for a possible adverse transition state steric
interaction between a 5-methyl and the styrene phenyl.
The presence of a 4-alkyl or 4-aryl DHP substituent (entries
5-9) results in a surprising decrease in endo stereoisomer
preference (51-62% endo) compared to substituents at other
positions. Interestingly, the particular size of the substituents
at the C4 position of the 1,2-DHP 10d-h is irrelevant to the
endo/exo selectivity since all products from 4-substituted DHPs,
even with the bulkier tert-butyl and phenyl, are in a very close
range of preferences for endo adducts 12h-l (55-62%).
The Diels-Alder cycloaddition of 6-methyl-1,2-DHP 10i with
styrene (entry 10) affords the endo-Ph stereoisomer 12m as the
dominating product with the endo/exo 12m/13m ratio of 96:4.
The methyl group has somehow either hindered approach of
the styrene phenyl from the exo orientation or facilitated endo
cycloadduct. Theoretical calculations were undertaken to gain
insights on this issue (see below).
Stereochemical Results for DHP/MVK Cycloadditions.
With a number of alkyl-1,2-DHPs in hand, we desired to see if
methyl vinyl ketone cycloadditions would result in substituent
stereochemical preferences similar to those observed with
styrene. The reactions shown in Scheme 3 were carried out,
and the results are in Table 5. Mariano3e reported the cycload-
dition of N-ethoxycarbonyl-DHP 1c with MVK as a neat thermal
reaction at 55 °C. The result was a 4:3 endo/exo ratio of acetyl
regioisomers 3e/4e (entry 1). Upon base-catalyzed equilibration,
a 2:3 mixture of these endo/exo isomers was observed. Our
results at 50 °C by NMR analyses of mixtures prior to separation
with both N-methoxycarbonyl-DHP (entry 2) and N-benzoxy-
carbonyl-DHP (entry 3) found slightly higher 70-74% prefer-
ences for endo-acetyl isomers 15a and 15b. The alkyl groups
of the carbamates appear to be sufficiently distant from the
The stereochemical outcome for formation of cycloadducts
15a/16a from MVK and 1,2-DHP 1a was found to be solvent
independent. As shown in Table 6, the highest cycloaddition
3462 J. Org. Chem., Vol. 72, No. 9, 2007

Diene-Substituted N-Alkoxycarbonyl-1,2-dihydropyridines
TABLE 6. Solvent Effects upon endo/exo 15a/16a Ratios for
TABLE 7. Calculated Transition State endo/exo (12NC/13XT)
Energy Differences for Generation of
Reaction of MVK with N-Methoxycarbonyl-1,2-dihydropyridine (1a)
7-Phenyl-2-azabicyclo[2.2.2]oct-5-enes 12C and 13Ta,b
15a/16a
endo/exo
yield
(%)
entry
conditionsa
1
2
3
4
neat
76:24
74:26
79:21
79:21
63
45c
35c
22c
cyclohexaneb
dichloromethaned
acetonitrilee
a MVK/1a 5:1; 55 °C, 5 days. b Concentration of 1a ) 100 mg/mL. c The
reaction did not go to completion. d Concentration of 1a ) 26 mg/mL.
e Concentration of 1a ) 43 mg/mL.
endo/exo 12/13
c
d
c
entry products substituent ∆Eq
∆Stotal ∆Gq
exptl (calcd)e
o
T
1
2
3
4
5
6
7
8
9
12a/13a
12b/13b
12c/13c
12d/13d
12e/13e
12f/13f
12g/13g
12h/13h
12i/13i
parent
parentf
-1.67 1.58 -2.30 84:16 (92:8)
-2.55 1.31 -2.89 86:14 (96:4)
SCHEME 4. Major Cycloaddition Transition States for
Azabicycles 12 and 13
Ph-p-OMe -1.85 2.0
-2.67 80:20 (95:5)
Ph-m-NO2 -3.30 0.44 -3.37 85:15 (97:3)
4-Me
-1.63 2.25 -2.57 78:22g (94:6)
4,6-di-Me -1.49 1.6
-2.08 91:9h (91:9)
6-Me
5-Me
5-Et
-1.52 0.89 -1.76 96:4 (87:13)
-1.00 1.31 -1.48 55:45 (83:17)
-1.25 1.15 -1.67 62:38 (86:14)
-1.09 0.38 -1.10 55:45 (77:23)
-1.40 1.95 -2.22 54:46 (92:8)
-1.67 4.66 -3.83 55:45 (98:2)
-3.29 3.02 -4.65 96:4 (99:1)
10 12j/13j
11 12k/13k
12 12l/13l
5-t-Bu
5-Ph
5-Bn
13 12m/13m 1-Me
a Reaction of 1,2-DHPs and styrene. b Ab initio RHF/3-21G(*) ZPVE-
corrected energies, Eq , of the cycloaddition transition states and free energy,
o
Gq, values were calculated using Spartan’06 for Windows software. The
O-methyl group of the carbamate preferred to be syn to carbonyl oxygen
in all calculations. c In kcal/mol. d In cal/mol. e (∆Eq or ∆Gq) ) (Eq or
o
o
Gq) 12NC - (Eq or Gq) 13XT for all TSs leading to 12C and 13T. f N-
o
COOBn analogue. g For the regioisomer, ∆Eqo ) (Eq ) 12eC - (Eq ) 14eC
o
o
) -2.185; 12e/14e 95:5 (92:8). h For the regioisomer, ∆Eqo ) (Eq ) 12fC
o
- (Eq ) 14fC ) -1.850; 12f/14f 85:15 (88:12).
o
yield was under neat conditions. All other reactions did not go
to completion in 5 days.
calculation results in Table 7 are qualitatively consistent with
the experimentally determined endo isomer preferences and
justify the level of sophistication of theory utilized.
Calculations of Styrene-DHP Cycloadditions. To gain
insights into substituent influences upon the stereoselectivity
of the styrene/DHP cycloadditions, ab initio calculations were
performed. Four major possible transition states (TSs) were
evaluated for cycloaddition reactions leading to azabicycles 12
and 13 (Scheme 4). These TS energies depend on the endo/exo
(N/X) orientation of the styrene phenyl group and also on the
conformational orientation of the carbamate carbonyl group as
s-cis or s-trans (C/T) toward C1 in the developing product
azabicycle. In the general case, these TSs will be denoted as
12NC, 12NT, 13XT, and 13XC. The two cycloadducts and their
major amide conformations, 12C/12T and 13T/13C, derive
directly from these TSs by closing the partial bonds shown.
The TS geometries were optimized at the RHF/3-21G(*) level,
and each transition structure was confirmed by carrying out a
frequency calculation to yield one and only one imaginary
frequency. The vibration associated with the imaginary fre-
quency was checked to correspond with a movement in the
direction of the reaction coordinate. The results for the favored
The Table 7 calculations do overestimate the preference for
the endo isomer in nearly all cases studied. The exception was
the cycloaddition of styrene with 6-methyl-1,2-DHP 10i to gave
endo/exo adducts 12m/13m (entry 7), for which there is a
slightly smaller endo prediction than the experimental value.
The largest deviations between experimental and observed
isomer ratios are for the 5-substituted isomers 12h-l/13h-l
(entries 8-12) for which calculations consistently overestimate
endo ratios by 22-43%. Calculations for the sterically crowded
1-Me derivative 12m/13m (entry 13) capture the high endo
preference. In light of the calculations for the MVK/DHP
cycloadditions that follow (Table 9), it is notable that both
transition state energy (∆Eq ) and entropy (∆Stotal) calculations
o
favor formation of endo isomers in all of the examples in Table
7.
The calculations of transition state bond distances demonstrate
that both endo and exo TSs for the cycloaddition of styrene
with DHPs are asynchronous (Table 8). The TSs for the
1-methyl-1,2-DHP 10i to give 12m/13m (entry 13) are the most
asynchronous of the TSs. In the favored 12NC TS for 12m
(Scheme 4), the newly forming C1-C7 bond (2.32 Å) is longer
than that calculated for any other endo-phenyl 12NC TS, and
the C4-C8 bond (2.10 Å) is the shortest such TS bond. For the
13XT TS for exo-phenyl isomer 13m, the C1-C7 bond (2.43
Å) is again the longest such bond of the structures, and the
C4-C8 bond (2.05 Å) is the shortest such bond. The 1-Me of
the 1,2-DHP 10i has hindered the approaching styrene but has
not altered the regiochemical outcome. It can be noted that the
endo and exo TSs are listed in Table 7. The calculated Eq
elec
energy order (kcal/mol) of the TSs is 12NC (0.00) < 12NT
(0.43) < 13XT (2.0) < 13XC (2.90). Single-point total energy
calculations of the TSs were carried out at the RHF/6-31G*
level. Both endo TSs are more stable than the exo TSs. The
calculated energy difference between 12NC, the more stable
endo TS, and 13XT, the more stable exo TS, is Eq ) 2.0
elec
kcal/mol (Eq ) 1.7 kcal/mol) in favor of the endo stereose-
o
lectivity. Thus, for further geometry optimizations and total
energy minimizations of substituted 1,2-DHPs and styrene or
its derivatives, the orientations of 12NC and 13XT were adopted
as the preferred endo and exo TS geometries. These lead to
12C and 13T as endo and exo product geometries. The
J. Org. Chem, Vol. 72, No. 9, 2007 3463

Krow et al.
TABLE 8. Calculated Bond Lengths (Å) between Reacting Partners in the 12NC and 13XT TSs for Formation of
7-Phenyl-2-azabicyclo[2.2.2]oct-5-enes 12C and 13Ta
12NCb
13XTb
entry
products
substituent
C1-C7
C4-C8
∆d (Å)
C1-C7
C4-C8
∆d (Å)
1
2
3
4
5
6
7
8
9
10
11
12
13
12a/13a
12b/13b
12c/13c
12d/13d
12e/13e
12f/13f
12g/13g
12h/13h
12i/13i
parent
parentc
Ph-p-OMe
Ph-m-NO2
4-Me
4,6-di-Me
6-Me
5-Me
5-Et
5-t-Bu
5-Ph
5-Bn
1-Me
2.268
2.268
2.264
2.285
2.259
2.272
2.279
2.273
2.271
2.273
2.296
2.285
2.322
2.130
2.132
2.133
2.118
2.130
2.128
2.117
2.135
2.133
2.155
2.118
2.139
2.095
0.138
0.136
0.131
0.167
0.129
0.144
0.162
0.138
0.138
0.118
0.178
0.146
0.227
2.325
2.332
2.318
2.344
2.316
2.300
2.312
2.349
2.338
2.347
2.365
2.340
2.433
2.115
2.113
2.116
2.102
2.116
2.121
2.127
2.101
2.105
2.128
2.089
2.114
2.054
0.210
0.219
0.202
0.242
0.200
0.179
0.185
0.248
0.233
0.219
0.276
0.226
0.379
12j/13j
12k/13k
12l/13l
12m/13m
a Reaction of 1,2-DHPs and styrene calculated at RHF/3-21G(*) level. b Calculated TS bond distances (d) in angstroms. c N-COOBn analogue.
a
TABLE 9. Calculated Transition State 15NCCk/16XCCk Energy Differences to Form 7-Acyl-2-azabicyclo[2.2.2]oct-5-enes 15CCk and 16CCk
endo/exo 15/16
exptl (calcd)
c
d
c
entry
products
substituent
∆Eq
∆Stotal
∆Gq
o
T
RHF/3-21G(*)b
2.05
1
2
3
4
5
6
7
8
15a/16a
15b/16be
15c/16c
15d/16d
15e/16e
15f/16f
parent
parente
4-Me
4,6-di-Me
6-Me
5-Me
5-Et
1-Me
3.02
3.34
6.39
3.21
2.40
5.38
4.16
5.28
0.00
0.40
-0.12
-0.08
0.05
-0.20
-0.82
0.39
74:26 (50:50)
70:30 (35:65)
71:29 (54:46)
53:47 (53:47)
90:10 (48:52)
60:40 (58:42)
54:46 (78:22)
96:4 (36:64)
1.58
2.10
0.97
0.84
1.62
0.59
2.16
15g/16g
15h/16h
B3LYP/6-31G*f
9
10
11
15a/16a
15e/16e
15h/16h
parent
6-Me
1-Me
2.47
0.13
1.37
4.97
5.26
6.80
-0.13
-1.69
-0.25
74:26 (55:45)
90:10 (93:7)
96:4 (60:40)
a Reaction of 1,2-DHPs and MVK. b Transition states were derived from RHF/3-21G(*). ZPVE-corrected energies Eq of the cycloaddition transition
o
states and Gq values were calculated using Spartan’06 for Windows software. (∆Eq , ∆Stotal, or ∆G*) ) (Eq , Stotal, or Gq) 15CCk - (Eq , Stotal, or Gq)
o
o
o
16CCk, respectively. Eq was converted to Hq
by addition of Htotal and RT (328.15 K). Gq ) Hq
- TStotal. The O-methyl group of the carbamate
total
o
total
preferred to be syn to carbonyl oxygen in all calculations. c In kcal/mol. d In cal/mol. e N-COOBn analogue. f The same calculations as footnote b using TS:
B3LYP/6-31G*-derived values.
12NC TSs leading to endo products 12C are less asynchronous
than the 13XT TSs leading to exo adducts 13T in all cases
studied.
level and single-point total energy calculations of the TSs were
carried out at the RHF/6-31G* level. ZPVE-corrected energy
differences were calculated for the lowest energy endo 15NCCk
and exo 16XCCk TSs and products 15CCk and 16CCk, formed
by closing the partial bonds shown in Scheme 5. At this level
of calculation, the minimum 16XCCk exo transition state, as
measured by ∆Eqelec, is 2.05 kcal/mol more stable than the
minimum 15NCCk endo transition state. Introduction of DHP
substituents (Table 9) did not alter the preference for ∆Eqelec of
16XCCk to be the minimum exo TS or for 15NCCk to be the
minimum endo TS. In all cases, ∆Eqelec of 16XCCk is the lowest
energy TS. However, the endo isomers 15b-h are experimen-
Calculations of MVK/DHP Cycloadditions. For the cy-
cloaddition of N-methoxycarbonyl DHP with MVK, there are
eight potentially important TSs. These depend on the approach
of the 1,2-DHP with respect to the acetyl group (endo/exo or
N/X), the orientation of the amide carbonyl (s-cis or s-trans
toward C1 of the cycloadduct, or C/T), and the conformational
bias of the MVK (carbonyl s-cis or s-trans toward the methylene
group of the cycloadduct, or Ck/Tk).
Using Spartan ’06 for Windows software, the geometries of
the eight transition states were optimized at the RHF/3-21G(*)
3464 J. Org. Chem., Vol. 72, No. 9, 2007

Diene-Substituted N-Alkoxycarbonyl-1,2-dihydropyridines
TABLE 10. Bond Lengths (Å) between Reacting Partners in the 15NCCk and 16XCCk TSs to Form 7-Acyl-2-azabicyclo[2.2.2]oct-5-enes
a
15CCk and 16CCk
b
b
15NCCk
16XCCk
entry
products
substituent
C1-C7
C4-C8
∆d (Å)
C1-C7
C4-C8
∆d (Å)
1
2
3
4
5
6
7
8
15a/16a
15b/16b
15c/16c
15d/16d
15e/16e
15f/16f
parent
parentc
4-Me
4,6-di-Me
6-Me
5-Me
5-Et
1-Me
2.354
2.352
2.349
2.370
2.381
2.382
2.374
2.480
2.061
2.063
2.061
2.047
2.043
2.043
2.042
1.984
0.293
0.289
0.288
0.323
0.388
0.339
0.332
0.496
2.544
2.525
2.502
2.487
2.507
2.635
2.66
2.01
0.534
0.507
0.481
0.463
0.486
0.657
0.664
0.912
2.018
2.021
2.024
2.021
1.978
1.996
1.933
15g/16g
15h/16h
2.845
a Cycloadditions of 1,2-DHPs and MVK at the RHF/3-21G(*) level. b Calculated TS bond distances (d) in angstroms. c N-COOBn analogue.
SCHEME 5. Favored Cycloaddition Transition States for
Azabicycles 15 and 16
chronous than the 16XCCk TSs leading to exo adducts 16CCk
for all cases studied. The most asynchronous TSs are for MVK
reacting with 6-methyl DHP 10i to give the 1-methyl isomers
15h/16h (entry 8). Consistent with a steric effect, in the 15NCCk
TS for 15h, the newly forming C1-C7 bond (2.48 Å) is longer
than that calculated for any other endo 15NCCk structure, and
the C4-C8 bond (1.984 Å) is the shortest such TS bond. For
the 16XCCk TS for exo isomer 16h, the C1-C7 bond (2.845
Å) is again the longest such bond of the structures in Table 10,
and the C4-C8 bond (1.933 Å) is the shortest such bond. These
bond distances do not, however, explain the endo preference
for the 1-Me isomer 15h. The calculated thermodynamic
parameters at two levels in Table 9 (entries 8 and 11) indicate
that, although the 16XCCk TSs are favored by the activation
tally favored, so ∆Eqelec values are inadequate to reproduce the
observed endo preference for MVK/DHP cycloadditions.
To improve the analysis, the Spartan program was used to
obtain Htotal and Stotal, the sums of translational, rotational, and
vibrational enthalpies and entropies for the transition state
structures. These were used to obtain the ∆GqT values from the
∆Eqo values in Table 9. While the ∆GqT values do predict endo
isomers to be favored for the 4-Me (entry 3) and 5-alkyl isomers
(entries 4, 6, and 7), they still do not correctly predict endo
isomers for the parents 15a,b (entries 1 and 2), the 6-methyl
derivative 15e (entry 5), and the 1-methyl derivative 15h (entry
8).
energies, Eq , total entropy (Stotal) considerations favor the
o
15NCCk TSs.
Conclusion
A series of 3-, 4-, 5-, and 6-substituted 1,2-dihydropyridines
have been reacted with styrenes and methyl vinyl ketone to
prepare 7-substituted-2-azabicyclo[2.2.2]hex-5-enes. There is a
kinetic preference for 7-endo cycloadducts in all examples
studied. Transition state energy calculations indicate that for
styrene/DHP cycloadditions TS total energy ∆Eqo and entropy
∆Stotal factors both favor endo cycloadducts. Calculations of
We next turned to the B3LYP/6-31G* theory level for
transition state geometry optimization and energy calculations
for entries 1, 5, and 8. At this higher level of theory, the
calculated TSs leading to the parent 15a (entry 9), 6-methyl
15e (entry 10), and 1-methyl 15h (entry 11) each now favor
the 15NCCk TS leading to the observed endo cycloadduct
isomers 15, although only the 6-methyl isomer 15e is quanti-
tatively close to its observed endo preference. It is noteworthy
for these calculated MVK cycloadditions that, while the
transition state enthalpy contributions of 16XCCk favor forma-
tion of the exo isomers 16CCk, this factor is not product
MVK/DHP cycloaddition transition states indicate that ∆Eq
o
considerations favor exo cycloadducts, ∆Stotal considerations
favor endo cycloadducts, and the resultant free energy favors
endo cycloaddition. The total entropy advantage for these endo
cycloaddition transition states is relevant to the question of
whether secondary orbital interactions, which would reduce
rather than increase conformational freedom, are responsible
for endo/exo selectivity preferences.18
determinative: Gq ) Hq
- TStotal, and the greater total
total
Experimental Section
transition state entropy contribution of the 15NCCk transition
states is dominant and accounts for the preferences for endo
isomers 15CCk.
General Procedure for the Diels-Alder Cycloadditions
between 1,2-DHPs and Styrene Derivatives. To the 1,2-DHP in
a decalin solution was added 3-5 equiv of styrene. The mixture
The calculations of transition state bond distances (Table 10)
demonstrate that both endo and exo TSs for the cycloaddition
of MVK with DHPs in Scheme 4 are asynchronous. The
15NCCk TSs leading to endo products 15CCk are less asyn-
(18) Garcia, J. I.; Mayoral, J. A.; Salvatella, L. Acc. Chem. Res. 2000,
33, 658-664.
J. Org. Chem, Vol. 72, No. 9, 2007 3465

Krow et al.
was heated at reflux under argon for 24 h. After it was cooled to
room temperature, the reaction mixture was loaded on a silica gel
flash column and rinsed with cyclohexane to remove the high
boiling decalin. The products were then eluted by flash column
chromatography with a cyclohexane/EtOAc solution.
product mixture. For endo isomer 15a, Rf ) 0.50 (cyclohexane/
EtOAc 1:1): 1H NMR (CDCl3, 300 MHz) δ 1.68-1.85 (m, 2H,
H8s and H8a), 2.16 (m, 3H), 2.84 (br, 1H, H4), 2.95 (m, 1H, H3n),
3.11 (m, 1H, H7), 3.26 (d, J ) 9.9 Hz, 1H, H3x), 3.70 (s, 3H), 4.98
and 5.17 (br and br, 1H, H1), 6.29 (m, 1H, H6), 6.40 (m, 1H, H5);
13C NMR (CDCl3, 75.5 MHz) δ 24.7 and 25.5, 28.2 and 28.3, 30.6
and 30.9, 46.8 and 47.0, 47.0 and 47.2, 52.4, 52.4 and 52.8, 130.0,
135.1, 155.3 and 155.9, 206.2 and 206.5; HRMS m/z 232.0941,
calcd for C11H15NO3Na (M + Na) 232.0950. For the exo isomer
16a, Rf ) 0.54 (cyclohexane/EtOAc 1:1): 1H NMR (CDCl3, 300
MHz) δ 1.33 (m, 1H, H8a), 2.17 (m, 1H, H8s), 2.23 and 2.30 (s and
s, 3H), 2.67 (m, 1H, H7), 2.75 (m, 1H, H4), 3.93 (dt and dt, J )
9.9, 2.7 Hz, 1H, H3n), 3.26 (dd, J ) 9.9, 2.1 Hz, 1H, H3x), 3.61
and 3.64 (s and s, 3H), 4.94 and 5.13 (m and m, 1H, H1), 6.42-
6.53 (m, 2H, H5, H6); 13C NMR (CDCl3, 75.5 MHz) δ 23.3, 28.4
and 28.7, 30.1 and 30.3, 47.1 and 47.5, 47.6 and 47.9, 52.3, 52.5,
131.6 and 132.0, 135.5 and 135.6, 155.3 and 156.3, 206.5 and 207.2;
HRMS m/z 232.0940, calcd for C11H15NO3Na (M + Na) 232.0950.
B. In Cyclohexane. A solution of MVK (1.0 g, 14.3 mmol) and
DHP 1a (0.401 g, 2.9 mmol) in cyclohexane (4 mL) was heated to
55 °C and stirred under argon for 5 days. The reaction was 87%
complete by 1H NMR. Purification afforded 271 mg of the mixture
Preparations of 7-endo-Phenyl-2-carbomethoxy-2-azabicyclo-
[2.2.2]oct-5-ene (12a) and 7-exo-Phenyl-2-carbomethoxy-2-
azabicyclo[2.2.2]oct-5-ene (13a). According to the general pro-
cedure, from DHP 1a (1.08 g, 7.77 mmol) and styrene (3.88 g,
37.3 mmol) in decalin (10 mL) after 17 h under argon there was
obtained a colorless oil (526 mg, 32%) of the mixed products 12a
and 13a. The mixture, Rf ) 0.40 (hexane/EtOAc 2:1), could not
1
be separated by normal phase chromatography. From H NMR
integrations of the H1 protons, the ratio of 12a and 13a was 84:16.
The ratio of 12a and 13a was 82:18 if analyzed by HPLC, eluting
with 35% acetonitrile aqueous solution. The retention times for 12a
and 13a were 10.7 and 8.08 min, respectively. A sample of 112
mg of the mixture was loaded to a medium-pressure (60 psi),
reversed-phase C18 column rinsed with 35% acetonitrile aqueous
solution to give 84 mg of 12a and 16 mg of 13a. The ratio of 12a
and 13a from separation was 84:16. For endo isomer 12a: 1H NMR
(CDCl3, 500 MHz, 75 °C) δ 1.68 (m, 1H, H8a), 2.17 (ddd, J )
13.5, 9.5, 2.5 Hz, 1H, H8s), 2.90 (br, 1H, H4), 3.03 (dt, J ) 10.0,
2.0 Hz, 1H, H3n), 3.36 (dd, J ) 10.0, 2.0 Hz, 1H, H3x), 3.42 (m,
1H, H7), 3.71 (s, 3H), 4.74 (br, 1H, H1), 6.28 (t, J ) 7.0 Hz, 1H,
H6), 6.55 (t, J ) 7.0 Hz, 1H, H5), 7.09-7.21 (m, 5H); 13C NMR
(CDCl3, 75 MHz) δ 31.3, 31.5, 44.4, 46.7, 50.7, 52.3, 126.3, 128.1,
130.9, 134.8, 143.4, 155.5; HRMS m/z 243.1270, calcd for C15H17-
NO2 243.1260. For exo isomer 13a: 1H NMR (CDCl3, 300 MHz)
δ 1.69 (m, 1H, H8s), 1.92 (m, 1H, H8a), 2.89 (m, 2H, H7, H4), 3.13
(dt and dt, J ) 10.5, 2.7 Hz, 1H, H3n), 3.34 and 3.60 (s and s, 3H),
3.41 and 3.51 (dd and dd, J ) 10.5, 2.4 Hz, 1H, H3x), 4.43, 4.45,
4.72 and 4.74 (br, 1H, H1), 6.40-6.64 (m, 2H, H5, H6), 7.18-7.34
(m, 5H); 13C NMR (CDCl3, 75 MHz) δ 29.2 and 30.0, 30.8 and
31.0, 44.6 and 45.4, 48.7 and 49.0, 50.6, 52.1 and 52.5, 126.5, 127.6
and 127.8, 128.5, 133.5, 134.1 and 134.3, 143.0 and 143.2, 156.3
and 156.6; HRMS m/z 266.1152, calcd for C15H17NO2 Na (M +
Na) 266.1156. DHP 1a (0.5 g, 3.6 mmol) and styrene (0.82 mL,
7.2 mmol) in xylene (50 mL) did not form cycloadduct 12a/13a
after heating to reflux for 2 h with 600 W microwave irradiation.
General Procedure for the Diels-Alder Cycloadditions
between 1,2-DHPs and Methyl Vinyl Ketone (MVK). Prepara-
tion of 7-endo-Acetyl-2-carbomethoxy-2-azabicyclo[2.2.2]oct-5-
ene (15a) and 7-exo-Acetyl-2-carbomethoxy-2-azabicyclo[2.2.2]-
oct-5-ene (16a). A. Neat Conditions. A mixture of MVK (0.908
g, 13.0 mmol) and DHP (1a) (0.361 g, 2.6 mmol) was heated to
55 °C and stirred under argon for 5 days. The reaction mixture
was purified repeatedly by flash column chromatography (cyclo-
hexane/EtOAc 5:1) to give a 74:26 ratio of 15a (251 mg) and 16a
(89 mg) with an overall yield of 63%. The isomer ratio was 76:24
1
of 15a/16a (45%) in a ratio of 74:26 by H NMR integrations of
H5 and H6 protons. C. In CH2Cl2. In a sealed vial purged with
argon, the solution of MVK (0.339 g, 4.8 mmol) and DHP 1a (132
mg, 0.95 mmol) in CH2Cl2 (5 mL) was heated to 55 °C and stirred
1
for 5 days. The reaction was only 55% complete by H NMR.
Purification gave 70 mg (35%) of a mixture of 15a/16a in a 79:21
1
ratio by from H NMR integrations of H5 and H6 protons. D. In
CH3CN. In a sealed vial purged with argon, the solution of MVK
(0.339 g, 4.8 mmol) and DHP 1a (132 mg, 0.95 mmol) in
acetonitrile (3 mL) was heated to 55 °C and stirred for 5 days. The
reaction was 58% complete by 1H NMR integrations. Purification
1
gave 29 mg (22%) of a 79:21 mixture of 15a/16a by H NMR
integrations of H5 and H6 protons.
Acknowledgment. Acknowledgment is made to BioNumerik
Pharmaceuticals, Inc. and Temple University for support of this
research. We thank Philip Sonnet for helpful insights.
Supporting Information Available: Experimental procedures
and spectroscopic data for 1,2-dihydropyridines 1a,b, 10a-c,e,f,h,i,
and cycloadducts 12b-m/13b-m, 14e,f, and 15b-j/16b-j, as well
1
as copies of H NMR and 13C NMR for these compounds, X-ray
data for 15d, Spartan-derived RHF/6-31G* energies for all cy-
cloadducts, RHF/3-21G(*) calculations for energies of all cycload-
dition transition states, and B3LYP/6-31G* calculations for tran-
sition states 12a/13a and 15a,e,h/16a,e,h. This material is available
free of charge via the Internet at http://pubs.acs.org.
1
from H NMR integrations of the H5 and H6 protons of the crude
JO0700575
3466 J. Org. Chem., Vol. 72, No. 9, 2007