Reactivity of N-phenyl-1-aza-2-cyano-1,3-butadienes in the Diels-Alder reaction

Article abstract of DOI:10.1021/jo9520607

It is found that N-phenyl-2-cyano-1-azadiene 4, prepared via a two-step, one-pot, sequence from acrylanilide, undergoes efficient [4 + 2] cycloaddition with a complete range of electron rich, electron poor, and neutral dienophiles under remarkably mild thermal conditions (90-120 °C for 20-48 h). Regiospecific formation of the α-cycloadduct wherein the dienophile substituent is a to nitrogen is observed for vinyl ethers and styrene, whereas the Diels-Alder reactions with methyl acrylate and methyl vinyl ketone (MVK) produce α/β mixtures in which the α-cycloadduct is the major regioisomer (approximately 4-5:1). An essentially identical reaction pattern was observed in the Diels-Alder reaction of N-(p-methoxyphenyl)-2-cyano-1-azadiene 18 and the 4-methyl-substituted azadiene 27. For compound 19 derived from cycloaddition of 18 with ethyl vinyl ether, facile conversion to the dihydropyridine 21 through loss of EtOH on brief acid treatment was also noted. The 2,4-cis-disubstitution pattern confirmed by X-ray diffraction for the major cycloadduct 29 isolated from the reaction of 27 with styrene provides evidence for the endo mode of cycloaddition in the Diels-Alder reaction of N-phenyl(aryl)-2-cyano-1-azadienes. Calculation of the frontier orbital energies and coefficients, as well as the transition state geometries for the [4 + 2] cycloaddition of N-phenyl-2-cyano-1-azadiene 4 with methyl vinyl ether, styrene, and MVK were carried out at the RHF AM1 level (MOPAC, Version 5.0). The FMO treatment indicates that the reaction of 4 with methyl vinyl ether occurs under LUMO_diene control, whereas in contrast, the corresponding cycloaddition with MVK occurs preferably under HOMO_diene control. A high degree of asynchronicity is observed in the calculated transition states for reaction of 4 with the three representative dienophiles. In all cases the transition states leading to the α-cycloadducts are lower in energy than those giving the β-products. However, at the AM1 level the exo cycloaddition mode is found to be the preferred, this result contrasting with experimental results for azadiene 27.

Full text of DOI:10.1021/jo9520607

J . Org. Chem. 1996, 61, 3715-3728
3715
Rea ctivity of N-P h en yl-1-Aza -2-Cya n o-1,3-Bu ta d ien es in th e
Diels-Ald er Rea ction
Nicholas J . Sisti, Irina A. Motorina, Marie-Elise Tran Huu Dau, Claude Riche,
Frank W. Fowler,*,† and David S. Grierson*,‡
Institut de Chimie des Substances Naturelles, C. N. R. S., 91198 Gif-sur Yvette, France, and Department
of Chemistry, State University of New York at Stony Brook, Stony Brook, New York 11794
Received November 22, 1995X
It is found that N-phenyl-2-cyano-1-azadiene 4, prepared via a two-step, one-pot, sequence from
acrylanilide, undergoes efficient [4 + 2] cycloaddition with a complete range of electron rich, electron
poor, and neutral dienophiles under remarkably mild thermal conditions (90-120 °C for 20-48 h).
Regiospecific formation of the R-cycloadduct wherein the dienophile substituent is R to nitrogen is
observed for vinyl ethers and styrene, whereas the Diels-Alder reactions with methyl acrylate
and methyl vinyl ketone (MVK) produce R/â mixtures in which the R-cycloadduct is the major
regioisomer (approximately 4-5:1). An essentially identical reaction pattern was observed in the
Diels-Alder reaction of N-(p-methoxyphenyl)-2-cyano-1-azadiene 18 and the 4-methyl-substituted
azadiene 27. For compound 19 derived from cycloaddition of 18 with ethyl vinyl ether, facile
conversion to the dihydropyridine 21 through loss of EtOH on brief acid treatment was also noted.
The 2,4-cis-disubstitution pattern confirmed by X-ray diffraction for the major cycloadduct 29
isolated from the reaction of 27 with styrene provides evidence for the endo mode of cycloaddition
in the Diels-Alder reaction of N-phenyl(aryl)-2-cyano-1-azadienes. Calculation of the frontier orbital
energies and coefficients, as well as the transition state geometries for the [4 + 2] cycloaddition of
N-phenyl-2-cyano-1-azadiene 4 with methyl vinyl ether, styrene, and MVK were carried out at the
RHF AM1 level (MOPAC, Version 5.0). The FMO treatment indicates that the reaction of 4 with
methyl vinyl ether occurs under LUMOdiene control, whereas in contrast, the corresponding
cycloaddition with MVK occurs preferably under HOMOdiene control. A high degree of asynchronicity
is observed in the calculated transition states for reaction of 4 with the three representative
dienophiles. In all cases the transition states leading to the R-cycloadducts are lower in energy
than those giving the â-products. However, at the AM1 level the exo cycloaddition mode is found
to be the preferred, this result contrasting with experimental results for azadiene 27.
Sch em e 1
The Diels-Alder reaction of 1-azadienes provides a
potentially powerful method for the construction of
structurally defined polyfunctionalized six-membered
heterocyles, as a result of regio- and stereocontrol during
both the cycloadditon step and subsequent reaction of the
enamine double bond (Scheme 1). However, the optimi-
zation of this hetero Diels-Alder reaction as a general
synthetic method has required solving a number of
problems, principal among which are the lack of reactiv-
ity of simple 1-azadiene systems in [4 + 2] cycloadditions
and the inherent instability of the endocyclic ∆2-piperi-
deine enamines produced.1 Indeed, due to the electro-
philic nature of nitrogen, 1-azadienes are characteristi-
cally poor dienes in the normal [HOMOdiene controlled]
Diels-Alder reaction with electron deficient dienophiles.
Important contributions by several groups have shown
that this problem can be circumvented by introducing
either electron-donating or withdrawing substituents
onto the nitrogen atom. In the first instance, Ghosez et
al. demonstrated that N,N-dimethylamino substitution
alters the normal inverse electron demand character of
1-azadienes promoting their cycloaddition with electron
deficient dienophiles.2 However, considerable difficulties
have been encountered in the extension of this approach
to R,â-unsaturated oximes.3,5d,6d Approaching the prob-
lem from the opposite direction, Fowler4 and Boger5 have
shown that substitution of the azadiene nitrogen by
strong electron-withdrawing (N-acyl, N-sulfonyl) groups
augments considerably the inverse demand Diels-Alder
reactivity of the derived azadienes and effectively modu-
lates the reactivity of the enamine system in the ∆2-
piperideine products.
In more recent work from our laboratories further
attention has been drawn to the influence of a C-2 cyano
group together wifh different nitrogen substituents (COR,
phenyl, alkyl, OCH3) on the Diels-Alder reactivity of the
azadiene system.6 Particularly interesting is the com-
(2) (a) Serckx-Poncin, B.; Hesbain-Frisque, A.-M.; Ghosez, L. Tet-
rahedron Lett. 1982, 23, 3261. (b) Ghosez, L.; Serckx-Poncin, B.; Rivera,
M.; Bayard, P.; Sainte, F.; Demoulin, A.; Hesbain-Frisque, A.-M.;
Mockel, A.; Munoz, L.; Bernard-Henriet, C. J . Heterocycl. Chem. 1985,
22 Suppl. Issue (Lect. Heterocycl. Chem., 8, 69).
(3) (a) Wade, P. A.; Amin, N. B.; Yen, H.-K.; Price, D. T.; Huhn, G.
F. J . Org. Chem. 1984, 49, 4595. (b) Kusurkar, R. S.; Bhosale, D. K.
Tetrahedron Lett. 1991, 32, 3199. (c) Boger, D. L.; Zhu, Y. Tetrahedron
Lett. 1991, 32, 7643.
(4) (a) Hwang, Y. C.; Fowler, F. W. J . Org. Chem. 1985, 50, 2719.
(b) Cheng, T.-S.; Lupo, A. T.; Fowler, F. W. J . Am. Chem. Soc. 1983,
105, 7696.
(5) (a) Boger, D. L.; Kasper, A. M. J . Am. Chem. Soc. 1989, 111,
1517. (b) Boger, D. L.; Corbett, W. L.; Wiggins, J . M. J . Org. Chem.
1990, 55, 2999. (c) Boger, D. L.; Curran, T. T. J . Org. Chem. 1990, 55,
5439. (d) Boger, D. L.; Corbett, W. L.; Curran, T. T.; Kasper, A. M. J .
Am. Chem. Soc. 1991, 113, 1713.
† State University of New York at Stony Brook.
‡ Institut de Chimie des Substances Naturelles.
X Abstract published in Advance ACS Abstracts, May 1, 1996.
(1) For a general discussion of the Diels-Alder reaction of 1-aza-
dienes; see: Boger, D. L.; Weinreb, S. M. Hetero Diels-Alder Methodol-
ogy in Organic Synthesis; Academic Press: San Diego, 1987; Chapter
9.
S0022-3263(95)02060-3 CCC: $12.00 © 1996 American Chemical Society

3716 J . Org. Chem., Vol. 61, No. 11, 1996
Ta ble 1. Rea ction of N-p h en yl-2-cya n o-1-a za d ien es w ith Selected Dien op h iles
Sisti et al.
azadiene
4
dienophile
X ) OEt
2,3-dihydrofuran
3,4-dihydro-2H-pyran
X ) CO2Me
conditions
R
product (ratio)
â
yield, %
90 °C, 36 h
90 °C, 22 h
120 °C, 46 h
90 °C, 40 h
90 °C, 20 h
90 °C, 23 h
120 °C, 48 h
90 °C, 20 h
90 °C, 51 h
90 °C, 30 h
90 °C, 40 h
75 °C, 24 h
120 °C, 30 h
90 °C, 22 h
90 °C, 44 h
120 °C, 34 h
120 °C, 40 h
6
7
8
9
11
13
14
15
16
19
22
24
-
-
-
91
87
25
71
65
76
35
75
39
67
83
63
77
60
68
88
36
4:1
4.4:1
10
12
-
X ) COMe
X ) Ph
X ) (CH2)3CH3
trans-â-methylstyrene
cis-â-methylstyrene
X ) OEt
-
-
-
18
27
-
X ) CO2Me
4:1
8:1
4:1
2.8/1
2.8/1
2/1
23
25
30
-
X ) COMe
X ) Ph
29
X ) m-O2NC6H4
X ) p-BrC6H4
X ) OEt
31/32
33/34
35/36
37/38
-
-
39/40
X ) CO2Me
2.6/2:5/1
Sch em e 2
bined presence of the N-phenyl and C-2 cyano groups in
azadiene 4, as this compound is found to be unique in
its capacity to undergo Diels-Alder reactions with elec-
tron rich, neutral, and electron deficient dienophiles.6c,d
Following from these preliminary observations, a full
account of the cycloaddition chemistry of N-phenyl-2-
cyano-1-azadienes is presented in this paper, accompa-
nied by the results of theoretical calculations carried out
to gain insight into the electronic and structural proper-
ties of these heterodienes which govern their Diels-Alder
reactivity.
Resu lts a n d Discu ssion
Ch em istr y. N-Phenyl-1-aza-2-cyano-1,3-butadiene (4)
was prepared by treatment of a solution of acrylanilide
(1) in CH2Cl2 with trifluoromethanesulfonic anhydride
at -60 °C, followed by reaction of the in situ generated
O-triflyl imidate 2 with a suspension of lithium cyanide
and 12-crown-4 in THF (Scheme 2).7-9 Under these
conditions competing N-triflation of amide 1 to give 3 as
a byproduct occurs to only a very small extent.10 Aza-
diene 4 was isolated in 70% yield after chromatographic
purification (silica gel; heptane/EtOAc 10/1). Character-
istic in the 13C NMR of this compound were peaks at δ
148 and 109 for the C-2 and the cyano group of the
imidoyl cyanide unit, and absorptions at 2221 and 1623
cm-1 in the IR spectrum. Although stable over relatively
long periods if frozen in benzene at -20 °C, neat azadiene
4 was observed to decompose slowly at 0 °C to a more
polar material (13% after 1 week). Isolation and struc-
ture elucidation of this crystalline product revealed that
it corresponded to compound 5, resulting from a Diels-
Alder type dimerization of 4 in which one molecule reacts
as the dienophile component (Scheme 2). More efficient
conversion of 4 to 5 was subsequently achieved (60%) by
heating in benzene (0.6 M solution) for 36 h. At higher
concentrations of 4 (1.2 M) a similar yield of product was
obtained after heating for only 15 h. However, under
these conditions the azadiene was not completely con-
sumed. This result provided the first indication that, in
addition to the anticipated inverse electron demand
reactivity of 4, it would also react with electron poor
dienophiles. To evaluate the scope of reactivity of this
1-azadiene a series of experiments were conducted, the
results of which are presented in Table 1 and Schemes 3
and 4.
(6) (a) Teng, M.; Fowler, F. W. Tetrahedron Lett. 1989, 30, 2481.
(b) Teng, M.; Fowler, F. W. J . Org. Chem. 1990, 55, 5646. (c) Sisti, N.
J .; Fowler, F. W.; Grierson, D. S. Synlett 1991, 816. (d) Trione, C.;
Toledo, L. M.; Kuduk, S.; Fowler, F. W.; Grierson, D. S. J . Org. Chem.
1993, 58, 2075.
(7) Krutosikova, A.; Dandarova, M.; Konecny, V.; Kovac, J . Chem.
Papers 1990, 44, 119.
(8) For the corresponding vinyl triflate reaction, see: Piers, E.;
Fleming, F. F. J . Chem. Soc., Chem. Commun. 1989, 756.
(9) For the preparation of LiCN, see: Livinghouse, T. Org. Synth.
1981, 60, 126.
(10) In general, <3-5% of the N-triflyl product is formed under the
experimental conditions employed. However, it is important to separate
completely this contaminant from the product 2-cyano-1-azadienes, as
in instances where trace quantities of this material remain extensive
decomposition is observed in cycloaddition reactions with ethyl vinyl
ether and methyl acrylate: Parly, F.; Motorina, I.; Fowler, F. W.;
Grierson, D. S., manuscript in preparation.
To begin with, azadiene 4 was heated in ethyl vinyl
ether (40 equiv) at 90 °C in a closed tube for 36 h.11 This
led to very clean conversion to compound 6, isolated as a

N-Phenyl-2-cyano-1-azadienes in the Diels-Alder Reaction
J . Org. Chem., Vol. 61, No. 11, 1996 3717
Sch em e 3
Sch em e 4
conditions, considerable decomposition of the dienophile
and cycloadduct was observed, contributing to the prob-
lems encountered in purifying compound 8, isolated in
only 25-30% yield. The small J 2,7 ) 0.8 Hz coupling for
the H-2 proton doublet signal (δ 4.70) confirmed the cis-
ring fusion in this product, and subsequent comparison
of the 1J 13C-H coupling constant for the C-2 methine
system of 6 (158 Hz) and 8 (153 Hz) enabled us to deduce
that it exists as an equilibrium mixture of conformers.12
As the formation of the self condensation product 5
suggests, azadiene 4 reacted with methyl acrylate at 90
°C for 40 h to give an approximately 4:1 mixture of the
regioisomeric products 9 and 10 (71%) (Scheme 3). The
positioning of the carbomethoxy substituent axial and R
to nitrogen in the major isomer 9 was determined from
the NMR data [δ 4.44 (dd, J ) 2.8, 4.5 Hz, H-2); δ 61.3
C-2]. Similarily, the reaction of 4 with methyl vinyl
ketone produced a 4.4:1 mixture of compounds 11 and
12 (65%). In this case the X-ray crystal structure of the
major product 11 was obtained, definitively showing that
the C-2 acetyl group was axial.13
A more quantitative idea as to the relative ease with
which azadiene 4 reacts with ethyl vinyl ether and
methyl vinyl ketone was obtained from a competition
experiment performed using 20 equiv of each dienophile
(90 °C, 24 h). By 1H NMR it was ascertained that
compounds 6 and 11/12 were produced in an approximate
7:5 ratio, and by heating compound 6 at the same
temperature and duration in the presence of excess
methyl vinyl ketone it was further verified that this ratio
reflects a kinetic product distribution.
pale yellow solid, in 91% yield. From the 1H and 13C
NMR spectra for this product it was clear that the ethoxy
group is present on the carbon center R to nitrogen and
that it adopts a preferred axial position [δ 88.3 (C-2); δ
4.82 (t, J ) 2.5 Hz, H-2)]. Further examination of the
spectral data revealed that the alternate regiosisomeric
cycloaddition product was not present in the crude
product mixture; a minor component (approximately
2-6%) being identified as the dimer 5.
The corresponding reactions using 2,3-dihydrofuran
and 3,4-dihydro[2H]pyran as representative Z-enol ether
type dienophiles gave very contrasting results. For the
five-membered ring dienophile efficient conversion to the
N-phenyl substituted bicyclic compound 7 was observed
(90 °C, 22 h; 87%) (Scheme 4, eq 1). On the other hand,
to induce appreciable reaction of 4 with dihydropyran
required heating at 120 °C for 46 h. This was presum-
ably necessary to overcome unfavorable steric interac-
tions incurred in the kinetically preferred endo cycload-
dition mode (vide infra). Under these somewhat harsher
(12) Measurement of 1J 13C-H coupling constants for the “anomeric”
methine center in carbohydrates and piperidine systems has proven
to be a sensitive probe of both axial/equatorial stereochemistry and
conformational mobility. In these systems 1J 13C-Heq is larger than
1J 13C-Hax by 10 ( 1 Hz. An only 5 Hz difference compared to a
compound of known stereochemistry or the near identity of this
coupling for compounds believed to be of opposite stereochemistry
strongly indicates that such systems are not conformationally stable.
For relevant applications of this technique, see: (a) Boch, K.; Pederson,
C. J . Chem. Soc., Perkin Trans. 2 1974, 293. (b) Takeuchi, Y.; Chivers,
P. J .; Crabb, T. A. J . Chem. Soc., Chem. Commun. 1974, 210. (c) Bonin,
M.; Chiaroni, A.; Riche, C.; Beloeil, J .-C.; Grierson, D. S. J . Org. Chem.
1987, 52, 382. (d) Reference 5d.
(11) As previously reported,6c good yields (75%) of cycloadduct 6 were
obtained on heating azadiene 4 with ethyl vinyl ether for shorter
periods (26 h), but under these conditions the azadiene 4 was not totally
consumed.

3718 J . Org. Chem., Vol. 61, No. 11, 1996
Sisti et al.
The Diels-Alder reaction of 4 with unactivated dieno-
philes such as styrene, 1-hexene, and cis- and trans-â-
methylstyrenes were subsequently examined. In the
styrene reaction (neat, 90 °C, 23 h) a single regioisomeric
adduct 13 was produced, isolated in 76% yield as a pale
yellow oil after chromatographic purification. In the
reaction of 4 with 1-hexene prolonged heating (neat, 120
°C, 48 h) was required, and consequently there was
formation of dimer 5 at the detriment of the C-2 alkyl
substituted product 14 (35% yield).
As anticipated for a concerted Diels-Alder process, the
stereochemistry of the dienophile was transferred to the
cycloadduct formed in the reaction of 4 with trans- and
cis-â-methylstyrenes. Furthermore, for the trans dieno-
phile facile conversion (90 °C, 20 h) to give 15 (75% yield)
was observed, whereas, for cis-â-methylstyrene, com-
pound 16 was formed under more stringent conditions
(90 °C, 51 h) and in characteristically lower yield (39%).
In the proton NMR spectrum of compound 15 a diaxial
arrangement of the phenyl and methyl substituents was
discerned from the observed small J 2eq,3eq ) 2.8 Hz
coupling constant for H-2 (d, δ 4.59). Similarily, the cis-
1,2 relative geometry in 16 was deduced from the larger
J 2,3 ) 4.1 Hz constant. However, the observation that
the 1J 13C-H coupling constant for the C-2 phenyl-
substituted center are essentially identical (139 ( 1 Hz)
for compounds 16, 13 (axial Ph attribution; J 2,3 ) 3.4 Hz),
and in particular compound 29 (equatorial Ph from X-ray
diffraction structure; J 2,3 ) 4.1 and 7.7 Hz) suggests that
these 2-phenyl-substituted ∆5,6-piperideines are, in fact,
conformationally mobile.
To determine the influence that an electron-donating
group on the aromatic ring will have on the relative rates
of these azadiene Diels-Alder reactions, the N-(p-meth-
oxyphenyl)-2-cyano-1-azadiene 18 was prepared from
amide 17. From the X-ray structure of this compound it
was determined that the azadiene component exists in
the extended s-trans conformation, and that the N-phenyl
group is twisted 46° out of plane so as to minimize steric
interactions with the C-6 cyano group and to optimize
the interaction of the aromatic ring π-electrons with both
the conjugated imine system and the nitrogen lone pair
of electrons.13 Similar attempts to prepare azadienes
bearing electron-withdrawing p-CN and p-CO2Me sub-
stitutents from the corresponding amides were unsuc-
cessful, due primarily to the insolubility of the starting
amides in the reaction medium.
deduced that trace quantities of acidic impurities in
CDCl3 catalyze this process. This was initially confirmed
by reacting 19 with dry HCl in CH2Cl2 at room temper-
ature for 1 h and isolation of 21 in 20% yield by
preparative layer chromatography. Subsequently, treat-
ment of 19 with acidic Al2O3 at 20 °C for 10 min was
observed to produce dihydropyridine 21 in up to 94%
yield.
The reaction of azadiene 18 with methyl acrylate (90
°C, 40 h) produced compounds 22 and 23, obtained as
an inseparable 4:1 mixture in high yield (83%). The
major regioisomer 22 in this mixture once again cor-
responded to the product bearing the carbomethoxy group
at the C-2 position, i.e. R to nitrogen. Similarily, diene
18 reacted with methyl vinyl ketone to give an ap-
proximately 8:1 mixture of cycloadducts 24 and 25. In
this case, silica column separation of the two regioiso-
meric products permitted their total characterization.
Taking into consideration that the time required for
complete consumption of azadienes 4 and 18 in their
respective Diels-Alder reactions with methyl acrylate
and methyl vinyl ketone are essentially the same, it is
apparent that the presence of the p-methoxy substituent
in 18 does not provide a large accelerating effect.
With the prospect in mind of employing more highly
functionalized N-phenyl-2-cyano-1-azadienes in several
synthetic projects, the [4 + 2] cycloaddition chemistry of
compound 27 was next examined. Indeed, an important
aspect of the reactivity of azadienes such as 27, bearing
a hydrogen on the C-4 substituent, is their propensity to
equilibrate to the more highly conjugated enamine tau-
tomer 28 (Scheme 4, eq 3). This rearrangement both
destroys the azadiene system and opens up a competing
Diels-Alder reaction pathway in which the derived
enamine reacts as the diene in a normal HOMOdiene
controlled process.14,15 Furthermore, the presence of two
chiral centers in the products of the Diels-Alder reaction
of 27 makes it a potentially easy matter to determine
the endo/exo selectivity in these reactions. Compound
27 was prepared from crotonamide 26 and isolated as a
low melting solid in 63% yield after flash column puri-
1
fication. In the H NMR spectrum of this product two
sets of signals (approximately 6:1 ratio) were present for
the methyl group protons (δ 1.91 and 2.05; J ) 1.6 and
6.8 Hz) and H-4 (δ 6.21 and 6.52; J ) 1.5; 3.0 and 15.9
Hz). Several lines of evidence led us to conclude that
compound 27 exists as an equilibrium mixture of s-cis
and s-trans diene conformers. First of all, identical
coupling constants were observed for the vinylic H-4
proton pair of signals, corresponding to a trans substitu-
tion pattern about the C3-C4 double bond. Secondly,
contrary to what would be expected if 27 was comprized
of a mixture of cis and trans 4-methyl isomers, the
proportions of the two azadiene components remained
unaltered during its reaction with selected dienophiles.16
Although the reaction of 18 with ethyl vinyl ether (90
°C, 30 h) proved somewhat less efficient than the corre-
sponding reaction of 4, cycloadduct 19 was still isolated
in good yield (67%) along with varying but small amounts
of dimer 20. In a subsequent competition experiment
equimolar amounts of azadienes 4 and 18 were reacted
in neat ethyl vinyl ether. After 5 h the ratio of com-
pounds 6 and 19 was approximately 3:1. However, after
1
continued heating for an additional 7 h, H NMR inte-
(14) Snyder, H. R.; Robinson, J . C. J r. J . Am. Chem. Soc. 1941, 63,
3279.
(15) (a) De Kimpe, N.; Stevens, C. J . Org. Chem. 1993, 58, 2904. (b)
De Kimpe, N.; Verhe´, R.; De Buyck, L.; Hasma, H.; Schamp, N.
Tetrahedron 1976, 32, 3063.
(16) In an attempt to provide further evidence in favor of a
conformational equilibrium for 27, its proton spectrum was recorded
in different solvents (DMSO-d6, C6D6, CD3OD). However, no ap-
preciable variations in signal heights (integration) were detected. On
heating the NMR probe to 100 °C the H-4 absorption for the minor
component in 27 was observed to broaden, and additional signals
appeared both in the olefinic region (δ 5.23-5.56 and 6.36) and for a
third methyl group (δ 1.89). However, assignment of these signals to
enamine 28 and to the cis 4-methyl isomer of 27 can only be considered
tentatively, since treatment of the mixture with acid did not result in
the expected dissapearance of the new olefinic signals (28 f 27).
gration of the peaks for the C-5 vinyl hydrogen revealed
that the two products were present in the mixture in
nearly equal proportions.
Interestingly, on several occasions compound 19 slowly
transformed to a new product, whose structure was
assigned to dihydropyridine 21 by proton NMR [δ 4.49
(ddd, H-3) and δ 5.99 (dt, H-2)] (Scheme 4, eq 2). It was
(13) The authors have deposited atomic coordinates for the struc-
tures 11, 18, and 29 with the Cambridge Crystallographic Data Centre.
The coordinates can be obtained, on request, from the Director,
Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge,
CB2 1EZ, UK.

N-Phenyl-2-cyano-1-azadienes in the Diels-Alder Reaction
J . Org. Chem., Vol. 61, No. 11, 1996 3719
Sch em e 5
F igu r e 1. ORTEP drawing of cycloadduct 29.
remained possible to make regio- and stereochemical
attributions based upon coupling constant measurements
for the H-2 and H-5 absorptions in the proton spectrum.
The formation of the 2,4-cis cycloadducts 29 (31, 33),
35, and 37 as the major products with the three repre-
sentative dienophiles indicates that the principle Diels-
Alder reaction mode of azadiene 27 involves an endo
transition state. However, in light of the results of
temperature variable NMR experiments,16 a similar
conclusion that the pathway leading to the minor R-cy-
cloadducts in these reactions involves the alternative exo
transition state cannot be made at present. This latter
point will require further verification.
In summary, despite the reluctance of simple 1-aza-
dienes to participate in Diels-Alder reactions, the N-phen-
yl-2-cyano-1-azadienes examined in our work all undergo
efficient [4 + 2] cycloaddition with a complete range of
dienophiles under remarkably mild thermal conditions
(90-120 °C for 20-48 h). These findings are in agree-
ment with earlier observations that N-acyl-2-cyano-1-
azadienes 41 bearing phenyl substitution at C-4 also
undergo Diels-Alder reactions with electron rich, poor,
and neutral dienophiles (Scheme 5).6b,17 In this regard
the 2-cyano-1-azadiene system offers an advantage over
the N-(phenylsulfonyl)-1-azadienes 42 studied by Boger
et al., which, although more reactive than 4 and 41 with
respect to electron rich vinyl ethers, do not react with
methyl acrylate or simple alkenes under a variety of
conditions.5c,18 It is clear from our results, and from a
comparison of the relative rates of the intramolecular
cycloadditions of 1-azadienes, with and without 2-cyano
substitution,6b,19 that the cyano group provides a cycload-
dition rate accelerating capability, while at the same time
stabilizing both the 1-azadiene system and the cycload-
ducts formed. However, in view of the electron-with-
drawing nature of the CN group it is not intuitively
obvious why the presence of this substituent should
promote the inverse electron demand reaction with
dienophiles like methyl acrylate. This point leads us to
suggest that the cyano group together with the phenyl
In the reaction with styrene, it was comforting to
observe that azadiene 27 underwent the desired [4 + 2]
cycloaddition (120 °C, 30 h) to give selectively a 4:1
mixture of the C-2 phenyl-substituted products 29 and
30 in 77% combined yield. Similarily the reaction of 27
with m-nitro and p-bromostyrene led to formation of the
stereoisomeric R-cycloadducts 31/32 and 33/34, respec-
tively, isolated in 60-68% yields (2.8:1 ratio). Interpre-
1
tation of the H NMR spectrum of the major isomer 29
(as well as 31 and 33) was complicated, since the value
of the coupling constant for H-5 (δ 5.69, d, J 4,5 ) 3.2 Hz)
is intermediate between that used on previous occasions
to assign an axial (g5 Hz) or equatorial (2.5 Hz) orienta-
tion to the C-4 substituent,6b,d and J H2ax,H3ax ) 7.7 Hz is
smaller than the 10-12 Hz value which is generally
observed. Furthermore, as mentioned above, measure-
1
ment of the J 13C(2)-H coupling constant indicated that
the molecule possesses a conformationally mobile struc-
ture. Thus, even though the 2,4-cis-diequatorial struc-
ture of 29 was attributed on the basis of the 1H/13C NMR
data, its structure had to be confirmed by X-ray crystal-
lography. Indeed, the X-ray study shows that it pos-
sesses the predicted structure, the diequatorial confor-
mation presumably being perferred in the solid state in
order to diminish 1,3-diaxial type steric interactions at
the expense of an reinforced interaction between the N-
and C-2 phenyl groups [note that the N-phenyl substitu-
ent is almost perpendicular (99°) to the plane of the
piperideine ring] (Figure 1).
Although complete separation of the minor isomers 30,
32, and 34 from the major cycloadducts in the three
styrene cycloaddition reactions with 27 was not achieved,
the H-5 vinyl and H-2 signals were clearly visible in the
proton spectrum of each compound. This made it possible
to assign the 2,4-trans relative stereochemistry to these
products, with the Ph group oriented equatorially.
To complete our study, azadiene 27 was subsequently
reacted with the more polar dienophiles ethyl vinyl ether
and methyl acrylate. Surprisingly, in initial experiments
with ethyl vinyl ether, extensive formation of decomposi-
tion products was observed. However, when particular
care was taken to separate 27 from very minute amounts
of the N-triflamide contaminants,10 a clean reaction took
place producing a (2:1) mixture of the R-cycloadducts 35
and 36 in 88% yield (the conformer with the OEt group
axial being observed in each case). The reaction of 27
with methyl acrylate was complicated by the isolation of
all four possible R (37:38; 1:5) and â-cycloadducts (39:
40; 2.6:2) as an inseparable mixture (36%), but it
(17) For
a related papers on the propensity of N-aryl-3-cyano-
substituted 1-azadienes to react with both electron rich and electron
deficient dienophiles; see: (a) Sakamoto, M.; Nozaka, A.; Shimamoto,
M.; Ozaki, H.; Suzuki, Y.; Yoshioka, S.; Nagano, M.; Okamura, K.;
Date, T.; Tamura, O. Chem. Pharm. Bull. 1994, 42, 1367. (b) Sakamoto,
M.; Nozaka, A.; Shimamoto, M.; Ozaki, H.; Suzuki, Y.; Yoshioka, S.;
Nagano, M.; Okamura, K.; Date, T.; Tamura, O. J . Chem. Soc., Perkin
Trans. 1 1995, 1759.
(18) N-(Phenylsulfonyl)-1-azadienes have recently been shown to
react with unactivated dienophiles in intramolecular Diels-Alder
reactions: Boger, D. L.; Corbett, W. L. J . Org. Chem. 1993, 58, 2068.
(19) J ung, M. E.; Choi, Y. M. J . Org. Chem. 1991, 56, 6729.

3720 J . Org. Chem., Vol. 61, No. 11, 1996
Sisti et al.
Ta ble 2. Hea ts of F or m a tion (DH, k ca l/m ol), AM1 HOMO, LUMO En er gies (E, eV), a n d Coefficien ts for Eth ylen e, Meth yl
Vin yl Keton e, Styr en e, a n d Meth yl Vin yl Eth er a
ethylene
16.47
methyl vinyl ketone
styrene
38.72
methyl vinyl ether
DH:
-23.99
-25.62
HOMO
LUMO
HOMO
LUMO
HOMO
LUMO
HOMO
LUMO
E
C1′
C2′
-10.55
+0.71
+0.71
+1.44
+0.71
-0.71
-10.85
+0.65
+0.68
+0.05
+0.62
-0.42
-9.00
+0.47
+0.33
+0.02
+0.45
-0.30
-9.42
+0.67
+0.49
+1.46
+0.68
-0.70
a
The molecules lie on the XY plane. The coefficients refer to the pz atomic orbitals. The HOMO and LUMO extend to the pz orbitals
of the CO group, of the aromatic carbons, and of the oxygen of OMe group.
Ta ble 3. Hea ts of F or m a tion (DH, k ca l/m ol), AM1 HOMO, LUMO En er gies (E, eV), a n d Coefficien ts for Bu ta d ien e a n d
1-Aza d ien es 43 to 45, 18, a n d 4a
butadiene
30.69
43
44
18
4
45
DH:
35.25
68.90
63.49
101.26
133.25
HOMO
LUMO
HOMO
LUMO
HOMO
LUMO
HOMO
LUMO
HOMO
LUMO
HOMO
LUMO
E
-9.36
+0.56
+0.43
-0.43
-0.56
+0.46
+0.57
-0.42
-0.42
+0.57
-10.15
+0.46
+0.24
-0.59
-0.62
+0.40
+0.50
-0.45
-0.43
+0.60
-10.62
+0.45
+0.23
-0.59
-0.61
-0.43
+0.57
-0.49
-0.27
+0.46
-8.82
+0.19
+0.29
-0.14
-0.21
-0.88
+0.43
-0.47
-0.15
+0.33
-9.28
+0.23
+0.28
-0.17
-0.24
-0.86
+0.45
-0.48
-0.17
+0.36
-9.61
+0.22
+0.26
-0.18
-0.23
-1.30
+0.38
-0.45
-0.13
+0.31
C1(N1)
C2
C3
C4
a
The molecules lie on the XY plane. The coefficients refer to the pz atomic orbitals.The HOMO and LUMO extend to the pz orbitals
of the CN group and the aromatic carbons (including the oxygen of the OMe group in 18).
substituent, also present in the azadienes of type 4 and
41, determines their Diels-Alder reactivity. Indeed the
subtle interplay of the electronic effects brought to the
azadiene system by these electronically “opposite” func-
tionalities may be at the origin of the differences in the
R versus â-regiochemistry observed in the reactions of
the N-phenyl and N-acyl-4-phenyl azadienes with the
neutral dienophile styrene and methyl acrylate. Simi-
larily, it may be that the strong electron-withdrawing
nature of the N-phenylsulfonyl group in azadiene 42
overrides any contribution of the C-4 phenyl group in this
system. To provide answers to some of the questions
which issue from such speculation, we have undertaken
a more detailed study of the Diels-Alder reaction of
N-phenyl-2-cyano-1-azadiene 4 with ethyl vinyl ether,
styrene, and methyl acrylate both at the FMO level and
through calculation of the transition states for these
reactions.
Com p u ta tion s. The calculation of frontier molecular
orbitals, as well as the examination of orbital coefficients
and secondary overlap, was used with success by Boger
(AM1 and MNDO) to rationalize the observed endo
selectivity and the preferred regiochemical formation of
the R-cycloaddition products in reactions of N-(phenyl-
sulfonyl)-1-azadienes of type 42.1,5d Similarily, Nomura
interpreted the increase in reactivity of phenyl-substi-
tuted 2-azadienes with electron deficient dienophiles
using FMO arguments (CNDO/2),20 and Orsini et al. have
studied the electrophilic character of substituted 1,4-
diaza-1,3-butadienes in Diels-Alder cycloadditions (MN-
DO).21
azadienes 44 and 4 bearing a more bulky cyano substitu-
ent at C-2 there is a change in the relative energies of
the cis/trans conformers, the s-cis anti forms which react
in the Diels-Alder reaction being 0.96 and 0.99 kcal/mol
more stable, respectively. In addition, for the N-phenyl-
2-cyano-1-azadiene 4, the optimal orientation of the
phenyl ring with respect to the diene system was
determined to be 42.32°. This value coincides well with
the X-ray crystal data (46°) for azadiene 18.13
Considering next the ground state structures of the
dienophiles,25 there is a discrepancy for methyl vinyl
ketone (MVK). Ronayne et al. found that MVK exists as
an equilibrium mixture at room temperature with the
s-trans form predominating.26 Similarily, UV spectro-
scopic experiments in liquid acrolein indicate that the
s-trans conformer is the more stable by 2.1 kcal/mol.27
These values are reproduced only at the 6-31G*/3-21G
level [Ecis - Etrans ) 1.8 kcal/mol].28 Both lower level ab
initio calculations (3-21G) and semiempirical methods
give a preference for the s-cis conformer (0.94 kcal/mol;
AM1). However, this is once again inconsequential, as
the FMO energy levels and the corresponding coefficients
are nearly the same for the s-cis and s-trans conformers.
The frontier orbital energies and coefficients deter-
mined for butadiene, azadienes 4, 43, 44, and the p-OMe
and -CN-substituted azadienes 18 and 45 (not synthe-
sized) are presented along with the corresponding dieno-
phile values (ethylene, MVK, styrene, and methyl vinyl
ether) in Tables 2 and 3. It is seen, that by replacing
the carbon C-1 in butadiene by nitrogen, as in 43, the
As the first step in our study, the ground state
geometry of N-phenyl-2-cyano-1-azadiene 4 and the
model azadienes 43 and 44, were calculated at the RHF
AM1 level (MOPAC, Version 5.0).22,23 For the parent
1-azadiene 43 Bachrach has previously determined the
s-trans conformer to be more stable than the s-cis form
by between 2.55 kcal/mol (MP2/6-31G*) and 2.82 kcal/
mol (HF/6-31G*), and the configuration with NsH anti
to the CdN bond to be lower in energy than the
corresponding syn isomer by about 0.82-0.87 kcal/mol.24
The RHF AM1 calculations do not accurately reproduce
this trend (∆E ) 0.09 kcal/mol for the two anti forms of
43). However, as relative energies only are dealt with
this does not introduce an error into our analysis. For
(20) Nomura, Y.; Takeuchi, Y.; Tomoda, S.; Ito, M. M. Bull. Chem.
Soc. J pn. 1981, 54, 2779.
(21) Orsini, F.; Sala, G. Tetrahedron 1989, 45, 6531.
(22) Dewar, M. J . S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J . J . P.
J . Am. Chem. Soc. 1985, 107, 3902.
(23) Stewart, J . J . P. MOPAC5.0, QCPE program number 455, and
J . Comput. Chem. 1989, 10, 221.
(24) Bachrach, S. M.; Liu, M. J . Am. Chem. Soc. 1991, 113, 7929.
(25) To simplify the calculations ethyl vinyl ether was replaced by
the corresponding methyl derivative: methyl vinyl ether (MVE).
(26) (a) Ronayne, J .; Sargent, M. V.; Williams, D. H. J . Am. Chem.
Soc. 1969, 88, 5288; (b) Montaudo, G.; Librando, V.; Caccamese, S.;
Maravigna, P. J . Am. Chem. Soc. 1973, 95, 6365.
(27) DeGroot, M. S.; Lamb, J . Proc. R. Soc. London, Ser. A 1957,
242, 36.
(28) Loncharich, R. J .; Schwartz, T. R., Houk, K. N. J . Am. Chem.
Soc. 1987, 109, 14.

N-Phenyl-2-cyano-1-azadienes in the Diels-Alder Reaction
J . Org. Chem., Vol. 61, No. 11, 1996 3721
Ta ble 4. Bon d Len gth s (Å), Bon d An gles (d eg) a n d Tor sion a l An gles (d eg) for th e Tr a n sition Sta tes of 1-Aza d ien e w ith
Eth ylen e a n d of 1-Meth yl-1-a za d ien e w ith P r op en e
1-azadiene + ethylene
STO3G (MP2/6-31G)
1-azadiene +
ethylene AM1
1-methyl-1-azadiene +
1-methyl-1-azadiene +
propene STO3G
propene AM1
C2′N1
C1′C4
C2′N1C2
C1′C4C3
C2′N1C2C3
C1′C4C3C2
2.108 (2.207)
2.188 (2.161)
107.88 (107.3)
103.51 (104.8)
-58.63
1.889
2.255
111.03
97.50
-54.94
55.52
2.195
2.122
105.64
105.91
-57.29
62.43
2.670
1.695
91.40
111.34
-46.27
74.88
60.46
HOMO energy (∆E ) 0.79 eV) is lowered to a much
greater extent than the LUMO energy (∆E ) 0.06 eV).
Azadiene 43 would thus be predicted to be more electro-
philic than butadiene and consequently display little
tendency to react in normal HOMOdiene-controlled Diels-
Alder reactions. These AM1 calculations agree well with
ab initio 6-31G results obtained by Bachrach [butadiene
HOMO (-0.3305 au)/1-azadiene HOMO (-0.3558 au); ∆
) 0.69 eV].29 In fact, essentially the only difference
between the two methods is that the decrease in the
LUMO level is more important at the ab initio level than
in the AM1 calculations [butadiene LUMO (0.1389 au)/
1-azadiene LUMO (0.1147 au); ∆ ) 0.66 eV).
The presence of the electron-withdrawing cyano sub-
stituent in azadiene 44 results in a further lowering of
the HOMO with respect to 43 (∆ ) 0.47 eV) and an even
larger drop in the LUMO energy level (∆ ) 0.83 eV). This
suggests that it may react efficiently (if not exclusively)
with electron rich and neutral dienophiles such as vinyl
ethers and styrene in the inverse electron demand Diels-
Alder mode.
4) f +0.46 (N-1), -0.62 (C-4)] and LUMO [+0.57, +0.57
f +0.50, +0.60] of 1-azadiene 43. The introduction of
the electron-withdrawing cyano group at C-2 inverses the
relative size of the N-1 and C-4 coefficients in the LUMO
of 44, and this trend is maintained in N-phenyl-2-cyano-
1-azadiene 4. Furthermore, the coefficients in 4 are all
reduced in value due to delocalization of the diene
π-electrons through the phenyl ring, and the coefficients
for N-1 and C-4 are nearly identical in both the HOMO
and LUMO. An interpretation of the regiochemistry of
the reaction of 4 with the three electronically different
dienophiles is thus difficult to make on the basis of the
relative magnitudes of the diene terminal atom coef-
ficients. However, as suggested from Boger’s work,5d the
regioselective formation of the R-product in the reaction
of 4 with ethyl vinyl ether and styrene can be rationalized
to result from secondary orbital interactions, since the
LUMO diene C-2 coefficient (-0.48) is significantly larger
than the coefficient for C-3 (-0.17). Unfortunately, a
similarily clear cut explanation for the selective (4 to 5:1)
formation of the R-cycloadduct in the HOMOdiene
controlled
Striking is the further modification of the frontier
orbital energies brought about by N-phenyl substitution
of the 2-cyano-1-azadiene system.20,30,31 This results in
a major shift in the position of the HOMO to higher
energy, and consequently, to a significant decrease in the
difference between the HOMO and LUMO levels (∆ )
1.77 eV). Indeed, compared to azadiene 44, the HOMO
energy in 4 is raised by 1.34 eV, whereas the LUMO level
continues to fall (∆ ) 0.43 eV). Measurements of
HOMO-LUMO energy differences show that the experi-
mentally observed reaction of 4 with methyl(ethyl) vinyl
ether [LUMOdiene control ) 8.56 eV; HOMOdiene control ) 10.74
reaction of 4 with MVK does not issue from secondary
orbital considerations (C-2/3 ) +0.28/-0.17). For this
reason, plus the fact that neither the Hij differential
overlap integrals or steric effects are included in the FMO
analysis, the transition states for the reactions of aza-
diene 4 were calculated.
The transition state for the prototype Diels-Alder
reaction of 1-azadiene 43 with ethylene has been calcu-
lated at several levels of theory (MP2/6-31G*, RHF/6-
31+G*, RHF/3-21G, RHF/STO3G). These calculations
(Table 4) display similar features, excepting that the
C2′-N1 distance (2.207 Å, 2.139 Å) is slightly longer than
the C1′-C4 distance (2.161 Å, 2.082 Å) at the MP2/6-
31G* and RHF/6-31+G* levels, respectively,29,32 and is
slightly shorter at RHF/3-21G and RHF/STO-3G (C2′-
N1: 2.103 Å/2.108 Å; C1′-C4: 2.158 Å/2.188 Å).33
Globally, in these studies the C2′-N1 and C1′-C4
distances differ by <0.11 Å, the other distances differ by
less than 0.02 Å, and the bond angles differ by no more
than 2°.
eV] and styrene [LUMOdiene
) 8.14 eV; HOMOdiene
control
control ) 9.30 eV] occurs under LUMOdiene control, and
alternatively, that the reaction of 4 with MVK can occur
under both LUMO and HOMO diene control, the latter
mode being the more favorable [E diene LUMO - E
dienophile HOMO ) 9.99 eV, and E diene HOMO - E
dienophile LUMO ) 9.33 eV].
The AM1 calculations also suggest that the reaction
of the p-cyanophenyl azadiene 45 with electron rich
dienophiles would be accelerated with respect to the
corresponding reaction of 4, due to a continued lowering
of the LUMO energy level (∆E LUMO ) 0.44 eV). In
contrast, the p-methoxyphenyl azadiene 18 should react
with even greater ease with methyl vinyl ketone (∆E
HOMO ) 0.46 eV). In fact, compared to the reaction of
4 with MVK, a small increase in reaction rate and yield
is observed.
It is immediately obvious from the results of the RHF
AM1 calculated transition state of the reaction of 43 with
ethylene (Table 4) that the degree of asymmetry is much
more pronounced: C2′-N1 ) 1.889 Å and C1′-C4 )
2.255 Å.34,36 As a better model to the Diels-Alder
reactions of 4, the endo approach of propene to N-methyl-
1-azadiene 46 was also studied. At the STO-3G level40
the distances (Table 4) between the reacting centers
Concerning the orbital coefficients, compared to buta-
diene there is a decrease in the relative magnitude of the
coefficient of the N-1 atom at the expense of the coef-
ficient of C-4 in both the HOMO [+0.56 (C-1), -0.56 (C-
(32) Tietze, L. F.; Fennen, J .; Geissler, H.; Schulz, G.; Anders, E.
Liebigs Ann. 1995, 1681.
(33) Tran Huu Dau, M. E.; Flament, J . P.; Lefour, J .-M.; Riche, C.;
Grierson, D. S. Tetrahedron Lett. 1992, 33, 2343.
(34) All the RHF AM1 transition structures were located using the
procedures in MOPAC (Version 5.0). All variables were optimized by
minimizing the sum of the squared scalar gradients (NLLSQ35a and
SIGMA35b). Final values of the gradient norms were <1 kcal/Å and
each transition structure had one negative eigenvalue in the Hessian
matrix as required.
(29) Bachrach, S. M.; Liu, M. J . Org. Chem. 1992, 57, 6736.
(30) Houk, K. N.; Strozier, R. H. J . Am. Chem. Soc. 1973, 95, 4072.
(31) Eisenstein, O.; Lefour, J . M.; Nguyen Trong Anh; Hudson, R.
F. Tetrahedron 1977, 33, 523.

3722 J . Org. Chem., Vol. 61, No. 11, 1996
Sisti et al.
F igu r e 2. AM1 optimized geometries of the transition structures of N-phenyl-2-cyano-1-azadiene + methyl vinyl ketone. The
distances are in angstroms.
Ta ble 5. Activa tion En er gies (AE, k ca l/m ol), Bon d Len gth s (Å), Bon d An gles (d eg), Tor sion a l An gles (d eg), Ch a r ges (q
on a tom or gr ou p of a tom s), Qu a n tities of Electr on Given by th e Dien op h ile to th e Dien e (Q, e), a n d Bon d Or d er s (b
(a tom -a tom )) for th e Tr a n sition Sta tes of 4 w ith Differ en t Dien op h iles
N-phenyl-2-cyano-1-azadiene 4
methyl vinyl ketone
styrene
ethylene
methyl vinyl ether
R
â
R
â
R
â
endo
exo
endo
exo
endo
exo
endo
exo
endo
exo
endo
exo
AE
C2′N1
C1′C4
C2′N1C2
C1′C4C3
C2′N1C2C3
C1′C4C3C2
C1′N1
34.72
2.655
1.720
96.31
109.91
33.65
2.596
1.724
96.49
107.98
36.77
36.70
32.04
2.825
1.753
91.98
111.46
31.69
2.817
1.742
94.37
109.55
39.43
38.74
33.87
2.502
1.771
97.26
108.37
27.04
2.830
1.797
93.87
111.06
25.73
2.810
1.793
95.94
109.73
39.54
39.43
-44.53 -46.41
74.35
-44.95 -43.12
79.42
-47.74 -41.32 -39.54
70.15
70.87
78.27
81.33
79.21
1.786
2.443
1.813
2.391
1.842
2.342
1.865
2.300
1.870
2.309
1.869
2.307
C2′C4
C1′N1C2
C2′C4C3
C1′N1C2C3
C2′C4C3C2
qN1
qC4
qC1′
qC2′
117.95
93.91
-48.39 -50.58
115.38
93.93
113.55
95.38
-53.55 -56.40
110.09
96.73
110.85
94.99
-57.31 -58.52
108.76
98.22
57.02
-0.11
-0.08
0.01
-0.39
+0.06
-0.14
-0.09
1.02
59.29
-0.11
-0.07
0.001 -0.20
-0.37
+0.06
-0.15
-0.06
1.01
57.31
-0.11
-0.10
-0.07
-0.22
+0.04
-0.15
0.03
57.85
-0.11
-0.09
-0.08
-0.20
+0.04
-0.15
0.04
57.74
-0.08
-0.17
-0.14
-0.09
+0.03
-0.16
0.11
55.71
-0.07
-0.17
-0.15
-0.07
+0.03
-0.16
0.12
-0.14
-0.10
-0.14
-0.19
+0.05
-0.16
0.03
-0.13
-0.10
-0.14
-0.18
+0.04
-0.16
0.05
-0.18
-0.06
-0.18
-0.06
-0.20
-0.02
+0.02
-0.17
0.19
-0.19
-0.05
-0.22
-0.09
+0.02
-0.17
0.18
-0.25
-0.01
-0.27
0.09
-0.01
-0.18
0.28
-0.26
-0.01
-0.27
0.10
-0.01
-0.18
0.30
-0.03
+0.02
-0.17
0.18
qPhenyl
qCyano
Q
b(C-O)(C-C)
1.01
0.99
1.12
1.11
1.05
1.04
-
1.16
1.16
1.03
1.03
corresponded closely to those found by Bachrach for the
prototype reaction of 43.29 In the AM1 treatment there
is similarily a trend toward a longer C2′-N1 distance
[C2′-N1 ) 2.670 Å, C1′-C4 ) 1.695 Å], albeit clearly
exaggerated. This is most probably due to an overesti-
mation of core-core repulsion at intermediate dis-
tances.37 In line with these results, Houk et al. found
C-N and C-C distances of 1.982 and 2.266 Å, respec-
tively, in the RHF/3-21G-calculated transition structure
for the reaction of butadiene with formaldimine.43 Houk
further found that the concerted exo transition structure
in the Diels-Alder dimerization of 1,3-butadiene is quite
asynchronous with two partial σ bonds equal to 2.07 and
2.39 Å, at the CASSCF/3-21G level.44
(35) (a) Bertels, R. H. Report CNA-44, University of Texas Center
for Numerical Analysis: Austin, TX, 1972. (b) McIver, J . W.; Komor-
nicki, A. J . Am. Chem. Soc. 1972, 94, 2625.
(36) As the energy minimization from the stationary point, initiated
by a step along the C-N or C-C bond in one sense leads to the
reactants and in the opposite sense to the product,35 we are confident
that the Diels-Alder reaction with azadiene 4 should be a single-step
reaction analogous to the Diels-Alder reaction of butadienes (note that
the experimental results obtained in the reaction of 4 with cis- and
trans-â-methylstyrenes are also in accord with a concerted mechanism).
The high degree of asymmetry is probably due in large part to the
overestimation of the core-core repulsion of the method at intermedi-
ate distances. However, one should be extremely careful in drawing
conclusions regarding reaction mechanisms from semi-empirical
calculations.37-39
The AM1 calculated ∆Eact for the cycloaddition of 43
(37) Lehd, M.; J ensen, F. J . Org. Chem. 1990, 55, 1034.
(38) Sodupe, M.; Oliva, A.; Bertran, J .; Dannenberg, J . J . J . Org.
Chem. 1989, 54, 2488.
(39) Tietze et al. (ref 32) calculated the energy surface for the
reaction of azadiene 43 with ethylene at AM1/CI. They located two
competing reaction channels, one corresponding to the concerted
process, and the second to a preferred lower energy two-step reaction
pathway.

N-Phenyl-2-cyano-1-azadienes in the Diels-Alder Reaction
J . Org. Chem., Vol. 61, No. 11, 1996 3723
F igu r e 3. AM1 optimized geometries of the transition structures of N-phenyl-2-cyano-1-azadiene + styrene. The distances are
in angstroms.
and ethylene (34.48 kcal/mol) agrees remarkably with the
ab initio value of 31.80 kcal/mol (MP4SDQ/6-31G*//MP2/
6-31G* with ZPE at HF/6-31G*//HF/6-31G*),29 and is
larger than the corresponding AM1 activation energy for
the butadiene + ethylene reaction (32.99 kcal/mol). All
factors considered, the AM1 transition states for the
model Diels-Alder reactions coincide sufficiently well
with ab initio treatments to provide the degree of
confidence necessary to proceed with the calculation of
the transition states for the observed [4 +2] cycloaddi-
tions of the more complex functionalized azadiene 4 with
styrene, ethyl vinyl ether, MVK, and ethylene.37,38 The
activation energies (obtained by difference from the
calculated enthalpies of the fully optimized reactants in
their ground states), interatom distances, and angles, as
well as other relevant parameters for the different
transition structures involving azadiene 4 (R, â, and endo,
exo orientations of the dienophiles) are presented in Table
5, and illustrated in Figures 2-5.
Looking first qualitatively at the calculated geometries
of these transition states one sees that, as for the simpler
model systems, there is a high degree of asymmetry.
However, it is interesting that there are only slight
differences between the exo and endo transition states
leading to the R products, despite the greater steric
interactions that would, a priori, be expected in the endo
reactions. For the transition states leading to the â
products the asymmetry is in the opposite sense, i.e.
N-1-dienophile distance is less than that between C-4
and the requisite dienophile carbon center. This differ-
ence in the sense of asymmetry may be construed as
reflecting mainly diminished steric interactions between
the dienophile and the crowded N-1 center. However,
this is an incomplete argument since the transition state
for the reaction of 4 with ethylene (a model for hex-1-
ene) has a geometry, which although somewhat less
asymmetric, is clearly R like with a longer N1-C distance.
The finding that the computed activation energies for
formation of the R cycloadducts are lower than the
corresponding energies for the â regioisomers correlates
nicely with the experimental results. Furthermore, this
agreement holds independently of whether the compari-
son is made between endo or exo modes for each of the
regioisomeric transition states. Indeed, the differences
in activation energies between the R and â reaction
pathways involving methyl vinyl ether and styrene are
7-13 kcal/mol. The difference in the activation energies
between the R and â reaction pathways is less important
in the cycloaddition of 4 with MVK (approximately 2 kcal/
mol) which is qualitatively consistent with the observed
formation of a 4.4:1 mixture of regioisomers.
It is also noteworthy that in the R transition structures
for the reactions of MVK and MVE, the forming bonds
(C1′-C4/C2′-N1 ) 1.724/2.596, 1.742/2.817, 1.771/2.502,
1.793/2.810 Å) are shorter in the MVK transition struc-
ture, suggesting that the transition state for this reaction
is later than the corresponding R-TS with MVE. As
observed by Lehd et al.,37 the values of activation energy
can be correlated with the lengths of the forming bonds:
the longer are the forming bonds the lower the activation
energy. Even though the relative differences in the AM1
calculated activation energies are too large, the observed
trend between ∆Eact and approach distance provides an
explanation why the â transition state energies are
higher than those leading to the R products [MVK: C1′-
C4/C2′-N1//C1′-N1/C2′-C4 ) 1.724/2.596//1.813/2.391
Å, styrene: 1.742/2.817//1.865/2.300, MVE: 1.793/2.810//
1.869/2.207 Å].
(40) The computation of the transition structure of the endo Diels-
Alder reaction of 1-methyl-1-azadiene with propene was carried out
using STO-3G basis set41 and gradient techniques with optimization
of all variables with the help of HONDO8 program.42 The stationary
point is characterized at the STO-3G restricted Hartree Fock level as
a true transition state having a single negative Hessian eigenvalue.
The convergence threshold on the maximum gradient component is
0.0006 Å.
(41) Hehre, W. J .; Stewart, R. F.; Pople, J . A. J . Chem. Phys. 1969,
51, 2657.
(42) Dupuis, M.; J ohnston, F.; Marquez, A. HONDO 8.1 from
CHEM-Station; IBM Corporation: Kingston, NY 12401.
(43) McCarrick, M. A.; Wu, Y.-D.; Houk, K. N. J . Org. Chem. 1993,
58, 3330.
(44) (a) Yi Li; Houk, K. N. J . Am. Chem. Soc. 1993, 115, 7478. (b)
Houk, K. N.; Loncharich, R. J .; Blake, J . F.; J orgensen, W. L. J . Am.
Chem. Soc. 1989, 111, 9172.
The agreement with experiment appears less satisfac-
tory, however, concerning the question of endo versus exo

3724 J . Org. Chem., Vol. 61, No. 11, 1996
Sisti et al.
F igu r e 4. AM1 optimized geometries of the transition structures of N-phenyl-2-cyano-1-azadiene + methyl vinyl ether. The
distances are in angstroms.
n f σ* interaction would be a reduction of the bond order
in the electron-acceptor bond. In our case, there is no
evidence for such stabilization, the bond orders of the
electron acceptor bond in the endo transition structures
being equivalent to the corresponding bond orders in the
exo transition structures.
To have at least a partial appreciation of the electronic
component which contributes, or more likely controls, the
regiochemical sense of the Diels-Alder reactions of 4, the
charge distributions were examined. Mulliken popula-
tion analysis shows that in all cases, except for the
transition state leading to the â regioisomer with methyl
vinyl ketone, the dienophile is globally donating a
quantity of electron “Q” to the diene (i.e. Q > 0). It was
F igu r e 5. AM1 optimized geometry of the transition structure
deduced from this calculation that the Diels-Alder
of N-phenyl-2-cyano-1-azadiene + ethylene. The distances are
reaction of 4 with MVE, ethylene, styrene, and methyl
vinyl ketone to give the R-product will occur under LUMO
diene-control. As can be expected, the extent of electron
donation increases with the electron-releasing property
of the dienophile (0.03) MVK < styrene < ethylene <
MVE (0.28). Examining the charge distributions on the
individual atoms the inductive effects produced by varia-
tion in the dienophile substituent can be measured: i.e.
the presence of an electron-withdrawing group at C2′ of
the dienophile is inclined to decrease the negative charge
on the adjacent atoms C1′ and N1 in the R transition
states, and on C1′ and C4 in the â TS’s, thereby
augmenting the negative charge on C2′. Inversely, the
presence of an electron-releasing group on C2′ is inclined
to increase the negative charges on the adjacent atoms
C1′ and N1 in the R TS’s and C1′ and C4 in the â TS’s,
resulting in a reduction of the negative charge on C2′.
For example, on going from MVK to MVE the negative
charge on C1′ and N1 in the R transition structures
increases from -0.14 and -0.13 to -0.27 and -0.26,
respectively, and the charge on C2′ becomes positive
(-0.18 to +0.10). Similarily, in the â transition struc-
tures for MVK and MVE the negative charge on C1′ and
C4 increases from 0.0 and -0.07 to -0.15 and -0.17,
respectively, and the negative charge on C2′ diminishes
(-0.37 to -0.07). Furthermore, with respect to the
in angstroms.
Diels-Alder reactivity. NMR and X-ray data show that
the endo product is formed in the Diels-Alder reaction
of the N-phenyl-2-cyano-4-methyl-1-azadiene 27 with
styrene, whereas the calculations for azadiene 4 indicate
that the exo transition states are slightly lower in energy.
Such preferences for the exo TS have been found in the
theoretical treatment of other Diels-Alder processes,
even at the RHF/3-21G and CASSCF levels. The study
by Apeloig and Matzner provides the first systematic and
quantitative evidence supporting the idea that secondary
orbital interaction control the stereochemistry in the
Diels-Alder reaction.45 According to other theoretical
studies, it is recognized that the calculations slightly
underestimate the stability of the endo TS relative to the
exo TS.44 A further factor which may diminish the
contribution of secondary orbital interaction in our
calculations is the large distance between the diene and
the C1′ center in the dienophile which bears the sub-
stituent.
Boger et al. have attributed the high endo selectivity
observed in the reaction of their N-phenylsulfonyl aza-
diene systems with vinyl ethers to the combined influence
of secondary overlap and anomeric type effects.5d Ac-
cording to Cieplak,46 the consequence of an anomeric type
(45) Apeloig, Y.; Matzner, E. J . Am. Chem. Soc. 1995, 117, 5375.
(46) Cieplak, A. S. J . Am. Chem. Soc. 1981, 103, 4540.

N-Phenyl-2-cyano-1-azadienes in the Diels-Alder Reaction
J . Org. Chem., Vol. 61, No. 11, 1996 3725
mp 106-107 °C; IR (neat) 2263, 2228, 1623, 1595, 1490 cm-1
1H NMR (CDCl3) δ 2.03 (m, 1H), 2.38 (m, 2H), 2.55 (m, 1H),
4.70 (t, J ) 4.2, 3.5 Hz, 1H), 5.96 (t, J ) 3.6, 4.1 Hz, 1H), 7.04
(d, J ) 7.6 Hz, 2H), 7.19-7.45 (m, 8H); 13C NMR (CDCl3) δ
19.8, 21.8, 64.5, 111.0, 115.4, 117.9, 120.1, 123.1, 123.5, 125.8,
127.8, 129.2, 129.5, 143.1, 145.1, 148.0; MS m/z 312 (M+), 220,
183, 155, 92, 77. Anal. Calcd for C20H16N4: C, 76.90; H, 5.16;
N, 17.94. Found: C, 76.55; H, 5.42, N, 17.71.
;
ground state charge on the phenyl ring (+0.06e) and the
cyano group (-0.16e), one sees that there is a build-up
of negative charge in the phenyl ring and that there is
very little charge variation in the cyano substituent in
the R transition state of the reaction of 4 with MVK (Ph:
+0.04; CN: -0.16) and MVE (Ph: -0.01; CN: -0.18).
However, in the â transition structures the charge on the
phenyl ring (+0.06 to +0.04e) and the cyano group (-0.15
to -0.16e) do not vary significantly. This movement of
electrons reveals that the charge given by the dienophile
to the diene is better stabilized by the phenyl ring on
nitrogen than by the cyano group at C2. This stabiliza-
tion effect of the N-phenyl group rationalizes on the one
hand the stability of the R transition states compared to
the â transition states, and on the other hand the R-like
orientation in the transition structure for the reaction of
4 with ethylene.
N -P h e n yl-1,2,3,4-t e t r a h yd r o-2-e t h oxy-6-cya n op yr i-
d in e (6). N-Phenyl-2-cyano-1-aza-1,3-butadiene (4) (78 mg,
0.5 mmol) and ethyl vinyl ether (1.2 mL, 20 mmol) were placed
in an argon-flushed 10-mL capacity thick glass-walled tube,
equipped with a Rotoflo tap and a magnetic stirring bar, and
heated under closed conditions at 90 °C (oil bath temperature)
for 36 h. After cooling, the excess dienophile was removed
under vacuum and the residue was flash column chromato-
graphed (silica, 6:1 heptane/EtOAc). Compound 6 was ob-
tained as a pale yellow solid (104 mg, 91%): mp 54-56°C
1
(heptane-EtOAc); IR (neat) 2228, 1630, 1595, 1490 cm-1; H
NMR (CDCl3) δ 1.26 (t, J ) 7.0 Hz, 3H), 1.61 (m, 1H), 1.97
(m, 1H), 2.19 (ddt, J ) 19.3, 1.3, 5.1 Hz, 1H), 2.39 (m, 1H),
3.67 (m, 1H), 3.90 (m, 1H), 4.82 (t, J ) 2.5 Hz, 1H), 6.02 (brt,
1H), 7.16 (m, 3H), 7.36 (t, 2H); 13C NMR (CDCl3) δ 15.2, 18.9,
24.0, 63.0, 88.3, 115.5, 116.3, 123.3, 124.7, 125.2, 129.5, 144.9;
MS m/z 228 (M+), 199, 196, 183, 155, 104, 77, 51. Anal. Calcd
for C14H16N2O: C, 73.65; H, 7.06; N,12.27. Found: C, 73.66;
H, 6.93; N, 12.35.
In view of the asymmetric nature of both the R and â
transition states for the reaction of azadiene 4 it is
tempting to try and deduce from the data something
concerning the asynchronicity/synchronicity (mechanism)
of these reactions.36 However, the answer to this perti-
nant question is beyond the reach of the semiempirical
approach which we have been constrained to employ
because of the complexity of the system under study.
Indeed, even at very high levels of theory this issue still
remains open for the simple prototype reaction of 1-aza-
diene 43 with ethylene. In spite of this limitation the
FMO analysis does permit an understanding of the
observed reactivity and regiochemistry in the Diels-
Alder reactions of N-phenyl-2-cyano-1-azadiene 4 with
such electronically different dienophiles as methyl vinyl
ketone, styrene, and ethyl(methyl)vinyl ether. Further
exploration of the Diels-Alder reaction of 2-cyano-1-
azadienes is in progress in our laboratories.
Diels-Ald er Ad d u ct 7. As described for 6, a mixture of
N-phenyl-2-cyano-1-aza-1,3-butadiene (4) (100 mg, 0.64 mmol)
in freshly distilled 2,3-dihydrofuran (1.94 mL, 25.6 mmol) was
heated under closed conditions at 90 °C for 22 h. Compound
7 was obtained as a yellow oil (126 mg, 87%) after flash column
chromatography (silica gel, 10:1 heptane/EtOAc): IR (neat)
1
2226, 1630, 1604, 1492, 1414 cm-1; H NMR (CDCl3) δ 1.86
(m, 1H), 2.22 (m, 2H), 2.35 (ddd, J ) 4.7, 6.3, 17.5 Hz, 1H),
2.63 (m, 1H), 3.80 (dt, J ) 8.3, 6.4 Hz, 1H), 5.23 (d, J ) 5.9
Hz, 1H), 5.95 (t, J ) 5.1 Hz, 1H), 7.12 (t, J ) 7.3, 1H), 7.22 (d,
J ) 7.6, 2H), 7.34 (t, J ) 7.9, 2H); 13C NMR (CDCl3) δ 24.0,
30.4, 37.7, 64.9, 89.9, 115.3, 121.6, 122.3, 124.5, 129.1, 143.9;
MS m/z 226 (M+), 225, 197, 195, 183, 181, 169, 155, 104, 77.
HRMS calcd for C14H14N2O 226.1106, found 226.1095. Anal.
Calcd for C14H14N2O: C, 74.31; H, 6.24; N,12.38. Found: C,
73.81; H, 6.38; N, 11.96.
Exp er im en ta l Section
Diels-Ald er Ad d u ct 8. As described for 6, a mixture of
N-phenyl-2-cyano-1-aza-1,3-butadiene (4) (70 mg, 0.45 mmol)
in freshly distilled 3,4-dihydro-2H-pyran (4.6 g, 54 mmol) was
heated under closed conditions at 120 °C for 46 h. Compound
8 was obtained as a yellow oil (27 mg, 25%) after flash column
chromatography (silica gel, 4:1 heptane/ether): IR (neat) 2225,
1625, 1600, 1500, 1413 cm-1; 1H NMR (CDCl3) δ 1.36 (m, 1H),
1.83 (m, 3H), 2.10 (m, 2H), 2.56 (m, 1H), 3.61 (m, 1H), 4.09
(dd, J ) 3.6, 11.6 Hz, 1H), 4.70 (d, J ) 0.8 Hz, 1H), 5.71 (dd,
J ) 3.0, 5.0 Hz, 1H), 7.16-7.38 (m, 5H); 13C NMR (CDCl3) δ
20.5, 22.7, 27.4, 30.7, 68.3, 88.4, 115.7, 116.3, 118.0, 123.9,
125.4, 129.2, 144.3; MS m/z 240 (M+), 239, 181, 155, 97, 84,
77; HRMS calcd for C15H16N2O 240.1262, found 240.1262.
N-P h en yl-1,2,3,4-t et r a h yd r o-2-ca r b om et h oxy-6-cya n -
op yr id in e (9) a n d N-P h en yl-1,2,3,4-t et r a h yd r o-3-ca r -
bom eth oxy-6-cya n op yr id in e (10). As described for 6, a
mixture of N-phenyl-2-cyano-1-aza-1,3-butadiene (4) (45 mg,
0.29 mmol) and methyl acrylate (1.04 mL, 11.5 mmol) was
heated under closed conditions at 90°C for 40 h. Compounds
9 and 10, an inseparable 4:1 mixture of regioisomers, were
obtained as a yellow oil (50 mg, 71%) after flash column
chromatography (silica gel, 8:1 heptane/ethyl acetate). These
regioisomers were separated by HPLC (silica gel; 93:7 heptane/
ethyl acetate).
N-P h en yl-2-cya n o-1-a za -1,3-bu ta d ien e (4). Triflic an-
hydride (4.4 g, 16.3 mmol) was added dropwise over 10 min to
a cold (-60 °C) solution of acrylanilide (1) (prepared in 82%
yield according to ref 7) (2.0 g, 13.6 mmol) and dry diisopro-
pylethylamine (2.6 g, 20.4 mmol) in 40 mL of anhydrous CH2-
Cl2, and the resulting mixture was stirred for 1 h (argon
atmosphere). A suspension of LiCN (0.6 g, 19.0 mmol; predried
for 2 h at 80 °C; 0.01 mmHg) in 40 mL of anhydrous THF
containing 12-crown-4 (0.27 g, 0.14 mmol) was then added
dropwise over a period of 10 min, and stirring was continued
at -60 °C for an additional 45 min. The reaction mixture was
subsequently warmed to -20 °C over a period of 15 min and
quenched with 50 mL of water. The organic layer was
removed and the aqueous phase washed with ether. The
combined organic layers were washed with water, dried over
sodium sulfate, and concentrated. N-phenyl-2-cyano-1-aza-
1,3-butadiene (4) was obtained as a yellow oil (1.5 g, 70%) after
flash column purification (silica gel, 10:1 heptane/EtOAc): IR
(neat) 2221, 1623, 1581 cm-1; 1H NMR (CDCl3) δ 6.09 (d, J )
10.5 Hz, 1H), 6.33 (d, J ) 17.5 Hz, 1H), 6.75 (dd, J ) 10.5, 7.0
Hz, 1H), 7.13 (dd, J ) 7.4, 1.5 Hz, 2H), 7.25-7.47 (m, 3H); 13
C
NMR (CDCl3) δ 109.8, 120.4, 127.7, 129.3, 129.6, 135.5, 140.5,
148.7; MS m/z 156 (M+), 155, 130, 77; HRMS Calcd for C10H8N2
156.0687, found 156.0703.
Compound 9: IR (neat) 2227, 1742, 1622, 1596, 1496 cm-1
;
N-ph en yl-2-cyan o-1-aza-1,3-bu tadien e dim er 5. N-Phen-
yl-2-cyano-1-aza-1,3-butadiene (4) (75 mg, 0.48 mmol) in 0.4
mL of anhydrous benzene was placed in an argon-flushed 10-
mL capacity thick glass-walled tube, equipped with a Rotoflo
tap and a magnetic stirring bar, and heated under closed
conditions at 90 °C (oil bath temperature) for 15 h. After
cooling, the solvent was removed under vacuum, and the
residue was flash column chromatographed (silica, 4:1 heptane
ether). Compound 5 was obtained after subsequent recrys-
tallization (pentane-ether) as a colorless solid (45 mg, 60%):
1H NMR (CDCl3) δ 1.93 (m, 1H), 2.20-2.39 (m, 3H), 3.78 (s,
3H), 4.44 (dd, J ) 2.8, 4.5 Hz, 1H), 5.77 (dt, J ) 1.1, 4.5 Hz,
1H), 7.12 (m, 3H), 7.34 (t, 2H); 13C NMR (CDCl3) δ 20.2, 22.8,
52.6, 61.3, 115.7, 117.3, 120.5, 122.8, 124.7, 129.3, 145.1, 145.4,
171.7; MS m/z 242 (M+), 184, 183, 181, 129, 104, 77, 51.
Compound 10: IR (neat) 2228, 1736, 1621, 1596, 1495 cm-1
;
1H NMR (CDCl3) δ 2.53 (m, 2H), 2.78 (m, 1H), 3.58 (s, 3H),
3.65 (m, 1H), 3.87 (ddd, J ) 0.7, 3.2, 12.6 Hz, 1H), 5.88 (t, J )
4.0 Hz, 1H), 7.07 (m, 3H), 7.34 (m, 2H); 13C NMR (CDCl3) δ

3726 J . Org. Chem., Vol. 61, No. 11, 1996
Sisti et al.
1
26.0, 37.0, 52.1, 64.6, 117.3, 120.1, 122.5, 124.4, 129.7, 145.2,
172.8; MS m/z 242 (M+), 183, 181, 104, 77. Anal. of mixture
of regioisomers: Calcd for C14H14N2O2: C, 69.40; H, 5.82; N,
11.56. Found: C, 69.79; H, 5.73; N, 11.77.
1495 cm-1; H NMR (CDCl3) δ 0.88 (d, J ) 6.8 Hz, 3H), 1.81
(ddd, J ) 2.8, 11.8, 18.8 Hz, 1H), 2.13-2.34 (m, 2H), 4.57 (d,
J ) 4.1 Hz, 1H), 5.77 (dd, J ) 2.8, 5.5 Hz, 1H), 7.05 (m, 3H),
7.29 (m, 7H); 13C NMR (CDCl3) δ 18.4, 27.4, 30.2, 68.5, 118.2,
118.8, 121.6, 122.7, 124.4, 125.8, 127.7, 127.8, 128.7, 129.0,
129.1, 139.8, 146.2; MS m/z 274 (M+) 259, 156, 103, 91, 77;
HRMS calcd for C19H18N2 274.1470, found 274.1463.
N -P h e n yl-1,2,3,4-t e t r a h yd r o-2-a ce t yl-6-cya n op yr i-
d in e (11) a n d N-P h en yl-1,2,3,4-tetr a h yd r o-3-a cetyl-6-cy-
a n op yr id in e (12). As described for 6, a mixture of N-phenyl-
2-cyano-1-aza-1,3-butadiene (4) (55 mg, 0.35 mmol) and methyl
vinyl ketone (1.0 g, 14 mmol) in 1.0 mL of anhydrous benzene
was heated under closed conditions at 90 °C for 20 h.
Compound 11 was obtained as a white solid (42 mg, 53%), and
12 was obtained as a yellow oil (10 mg, 12%) after flash column
chromatography (silica gel, 3:1 heptane/ether and then 2:1
heptane/ether).
N-(p-Meth oxyph en yl)-2-cyan o-1-aza-1,3-bu tadien e (18).
Following the procedure described for the preparation of
N-phenyl-2-cyano-1-aza-1,3-butadiene (4), p-methoxyacryla-
nilide (17) (1.0 g, 5.65 mmol) in CH2Cl2 (40 mL) containing
diisopropylethylamine (1.75 g, 13.56 mmol) was treated with
triflic anhydride (1.91 g, 1.14 mL, 6.78 mmol) at -73 °C, and
then with LiCN (0.56 g, 16.95 mmol) and 12-crown-4 (0.10 g,
0.57 mmol). N-(p-Methoxyphenyl)-2-cyano-1-aza-1,3-butadi-
ene (18) was obtained as a yellow low melting crystalline solid
(0.83 g, 79%) after flash column purification (silica gel, 5:1
heptane/EtOAc): mp 52-54 °C; IR (neat) 2220, 1609, 1503,
Compound 11: mp 101-103 °C; IR (neat) 2225, 1712, 1593,
1500 cm-1; 1H NMR (CDCl3) δ 1.74 (m, 1H), 2.17 (m, 2H), 2.36
(s, 3H), 2.37 (m, 1H), 4.24 (dd, J ) 3.2, 1.2 Hz, 1H), 5.92 (t, J
) 4.0 Hz, 1H), 7.12 (m, 3H), 7.36 (t, 2H); 13C NMR (CDCl3) δ
20.3, 20.9, 27.0, 68.7, 115.6, 117.3, 122.1, 124.2, 124.6, 129.5,
145.5, 208.7; MS m/z 226 (M+), 183, 182, 166, 155, 143, 129,
104, 77, 51. Anal. Calcd for C14H14N2O: C, 74.31; H, 6.23; N,
12.38; O, 7.07. Found: C, 74.15; H, 6.28; N, 12.53; O, 7,28.
Spectral data for 12: IR (neat) 2227, 1709, 1622, 1596, 1496
1
1250 cm-1; H NMR (CDCl3) δ 3.84 (s, 3H), 6.05 (d, J ) 10.5
Hz, 1H), 6.29 (d, J ) 17.5 Hz, 1H), 6.75 (dd, J ) 10.6, 17.5
Hz, 1H), 6.94 (m, 2H), 7.27 (m, 2H); 13C NMR (CDCl3) δ 55.6,
110.7, 114.6, 123.3, 128.2, 136.0, 137.5, 141.4, 159.1; MS m/z
186 (M+), 171, 160, 143, 85. Anal. Calcd for C11H10N2O: C,
70.95; H, 5.41; N, 15.04. Found: C, 70.65; H, 5.24; N, 14.76.
N-(p-Meth oxyp h en yl)-1,2,3,4-tetr a h yd r o-2-eth oxy-6-cy-
a n op yr id in e (19). As described for 6, a mixture of N-(p-
methoxyphenyl)-2-cyanoazadiene (18) (100 mg, 0.54 mmol) and
ethyl vinyl ether (1.6 g, 21.6 mmol) was heated under closed
conditions at 90 °C for 30 h. Compound 19 was obtained as a
pale yellow oil (93 mg, 67%) after flash column chromatogra-
phy (silica gel, 4:1 heptane/ether): IR (neat) 2227, 1622, 1510
1
cm-1; H NMR (CDCl3) δ 2.15 (s, 3H), 2.47 (m, 2H), 2.83 (m,
1H), 3.49 (m, 1H), 3.86 (dd, J ) 3.2 Hz, 1H), 5.89 (t, J ) 4.1
Hz, 1H), 7.09 (m, 3H), 7.35 (t, 2H); 13C NMR (CDCl3) δ 25.8,
28.7, 44.2, 51.7, 115.4, 118.3, 120.9, 122.9, 124.5, 129.4, 145.3,
207.4; MS m/z 226 (M+), 183, 155, 147, 105, 93, 77, 55; HRMS
calcd for C14H14N2O 226.1106, found 226.1104.
N -P h e n yl-1,2,3,4-t e t r a h yd r o-2-p h e n yl-6-cya n op yr i-
d in e (13). As described for 6, N-phenyl-2-cyano-1-aza-1,3-
butadiene (4) (64 mg, 0.41 mmol) and styrene (1.71 g, 16.4
mmol) were heated under closed conditions at 90 °C for 23 h.
Compound 13 was obtained as a pale yellow oil (81 mg, 76%)
after flash column chromatography (silica gel, heptane): IR
cm-1 1H NMR (CDCl3) δ 1.26 (t, J ) 7.0 Hz, 3H), 1.59 (m,
;
1H), 1.95 (m, 1H), 2.16 (ddt, J ) 19.3, 1.3, 5.1 Hz, 1H), 2.40
(m, 1H), 3.62 (m, 1H), 3.79 (s, 3H), 3.90 (m, 1H), 4.68 (t, J )
2.4 Hz, 1H), 5.90 (dd, J ) 3.0, 1.0 Hz, 1H), 6.89 (m, 2H), 7.08
(m, 2H); 13C NMR (CDCl3) δ 15.3, 18.6, 23.9, 55.6, 63.1, 88.4,
114.8 2, 116.5, 122.7, 125.7, 138.3, 157.4; MS m/z 259 (M+),
214, 187, 172; Anal. Calcd for C15H19N2O2: C, 69.47; N, 7.38;
N, 10.80. Found: C, 69.57; H, 7.08; N, 10.89.
1
(neat) 2227, 1622, 1596, 1496 cm-1; H NMR (CDCl3) δ 1.86
(m, 1H), 1.99-2.25 (m, 3H), 4.92 (t, J ) 3.4 Hz, 1H), 5.76 (dt,
J ) 1.6, 4.8 Hz, 1H), 7.09 (m, 3H), 7.22-7.41 (m, 7H); 13C NMR
(CDCl3) δ 19.5, 25.6, 62.5, 116.2, 117.0, 121.9, 122.1, 124.0,
125.9, 127.3, 128.9, 129.2, 140.8, 146.1; MS m/z 260 (M+), 196,
183, 155, 130, 104, 84, 77, 51. Anal. Calcd for C18H16N2: C,
83.04; H, 6.19; N, 10.76. Found: C, 82.66; H, 5.81, N, 10.43.
N-P h en yl-1,2,3,4-t et r a h yd r o-2-b u t yl-6-cya n op yr id in e
(14). As described for 6, a mixture of N-phenyl-2-cyano-1-aza-
1,3-butadiene (4) (50 mg, 0.3 mmol) and 1-hexene (3.4 g, 36
mmol) was heated under closed conditions at 120 °C for 48 h.
Compound 14 was obtained as a light red oil (25 mg, 35%)
after flash column chromatography (silica gel, 4:1 hep-
N -(p -Me t h oxyp h e n yl)-2-cya n o-1-a za -1,3-b u t a d ie n e
Dim er 20. N-(p-Methoxyphenyl)-2-cyanoazadiene (18) (206
mg, 1.10 mmol) in 3.0 mL of anhydrous benzene was placed
in an argon-flushed 10-mL capacity thick glass-walled tube,
equipped with a Rotoflo tap and a magnetic stirring bar, and
heated under closed conditions at 100 °C (oil bath temperature)
for 36 h. After cooling, the solvent was removed and the
residue was flash column chromatographed (silica; 6/1 hep-
tane/ether). Compound 20 was obtained after subsequent
recrystallization (pentane-ether) as a bright yellow solid (127
mg, 62%). mp ) 97-98 °C; IR (neat) 2227, 2221, 1622, 1510
1
tane/ether): IR (neat) 2227, 1625, 1595, 1492 cm-1; H NMR
(CDCl3) δ 0.95 (t, J ) 6.8 Hz, 3H), 1.50 (m, 4H), 1.69 (m, 2H),
2.21 (m, 2H), 3.68 (m, 1H), 5.90 (t, J ) 4.0 Hz, 1H), 7.04 (m,
3H), 7.32 (m, 2H); 13C NMR (CDCl3) δ 14.2, 19.8, 22.9, 28.9,
30.6, 60.2, 116.2, 116.7, 122.4, 123.6, 123.7, 129.2, 146.7; MS
m/z 240 (M+), 183, 155, 130, 104, 77; HRMS calcd for C16H20N2
240.1627, found 240.1629.
cm-1; H NMR (CDCl3) δ 2.02 (m, 1H), 2.34 (m, 2H), 2.47 (m,
1
1H), 3.78 (s, 3H), 3.81 (s, 3H), 4.53 (t, J ) 4.0 Hz, 1H), 5.82 (t,
J ) 4.0 Hz, 1H), 6.92 (m, 4H), 7.20 (m, 4H). 13C NMR (CDCl3)
δ 19.8, 22.2, 55.6, 65.1, 111.9, 114.6, 114.9, 115.6, 118.9, 120.5,
122.9, 126.3, 138.4, 140.0, 140.6, 158.0, 160.0; MS, m/z 373
(M+), 250, 213, 123; HRMS calcd for C22H20N4O2 372.1587,
found 372.1590. Anal. Calcd for C22H20N4O2: C, 70.95; H,
5.41; N, 15.04. Found: C, 70.85; H,5.65; N, 15.10.
N-P h en yl-1,2,3,4-tetr a h yd r o-tr a n s-2-p h en yl-3-m eth yl-
6-cya n op yr id in e (15). As described for 6, a mixture of
N-phenyl-2-cyano-1-aza-1,3-butadiene (4) (50 mg, 0.32 mmol)
and trans-â-methylstyrene (1.5 g, 12.8 mmol) was heated
under closed conditions at 90 °C for 20 h. Compound 15 was
obtained as a light red oil (66 mg, 75%) after flash column
chromatography (silica gel, 5:1 heptane/ether): IR (neat) 2227,
N-(p-Meth oxyph en yl)-2-cyan o-1,4-dih ydr opyr idin e (21).
Procedure 1: N-(p-Methoxyphenyl)-1,2,3,4-tetrahydro-2-ethoxy-
6-cyanopyridine (19) was treated with dry HCl in CH2Cl2 for
1 h at room temperature. The mixture was then concentrated
and separated by preparative TLC on SiO2 (heptane-EtOAc,
5:1). Compound 21 was obtained as a colorless oil (20%).
Procedure 2: N-(p-Methoxyphenyl)-1,2,3,4-tetrahydro-2-
ethoxy-6-cyanopyridine (19) (57 mg, 0.22 mmol) in CH2Cl2 (5
mL) was absorbed on 1.0 g of acidic Al2O3 (J anssen Chimica,
50-200 µm). The solvent was then evaporated under reduced
pressure, and the mixture was stirred under nitrogen at room
temperature for 10 min. The adsorbent was washed several
times with CH2Cl2, and the combined washings were evapo-
rated giving compound 21 (44 mg, 94%) (TLC and NMR
pure): IR (film) 2231, 1669, 1606, 1513 cm-1; 1H NMR (CDCl3)
δ 3.11 (m, 2H), 3.80 (s, 3H), 4.49 (ddd, J ) 8.1, 5.5, 3.0 Hz,
1H), 5.39 (m, 1H), 5.99 (dt, J ) 8.1, 1.2 Hz, 1H), 6.89 (d, J )
8.9 Hz, 2H), 7.14 (d, J ) 8.9 Hz, 2H); 13C NMR (CDCl3) δ 29.82,
55.62, 98.20, 114.72, 115.77, 118.57, 126.70, 128.35, 131.75,
1616, 1596, 1496 cm-1 1H NMR (CDCl3) δ 1.07 (d, J ) 6.8
;
Hz, 3H), 1.92 (m, 1H), 2.43 (m, 2H), 4.59 (d, J ) 2.8 Hz, 1H),
5.62 (dd, J ) 3.7 Hz, 1H), 7.04 (m, 3H), 7.30 (m, 7H); 13C NMR
(CDCl3) δ 19.3, 26.4, 32.3, 68.0, 115.9, 116.5, 118.4, 121.6,
123.9, 125.7, 127.3, 128.9, 129.1, 141.7, 146.2; MS m/z 274
(M+), 194, 179, 176, 161, 136, 121, 105, 86, 84, 77. Anal. Calcd
for C19H18N2: C, 83.18; H, 6.61; N, 10.21. Found: C, 83.17;
H, 6.66; N, 9.92.
N-P h en yl-1,2,3,4-t et r a h yd r o-cis-2-p h en yl-3-m et h yl-6-
cya n op yr id in e (16). As described for 6, N-phenyl-2-cyano-
1-aza-1,3-butadiene (4) (50 mg, 0.32 mmol) and cis-â-methyl-
styrene (1.5 g, 12.8 mmol) were heated under closed conditions
at 90°C for 51 h. Compound 16 was obtained as a light yellow
oil (34 mg, 39%) after flash column chromatography (silica gel,
10:1 heptane/ethyl acetate): IR (neat) 2225, 1724, 1616, 1597,

N-Phenyl-2-cyano-1-azadienes in the Diels-Alder Reaction
J . Org. Chem., Vol. 61, No. 11, 1996 3727
136.28, 158.45; MS (CI, isobutane, 170 °C), m/z 269 (M + 57),
213 (M + H), 124, 105.
flash column chromatography (silica gel, heptane/EtOAc, 97:
3, then 95:5).
N -(p -Me t h o x y p h e n y l)-1,2,3,4-t e t r a h y d r o -2-c a r b o -
m eth oxy-6-cya n op yr id in e (22) a n d N-(p-Meth oxyp h en -
yl)-1,2,3,4-tetr a h yd r o-3-ca r bom eth oxy-6-cya n op yr id in e
(23). As described for 6, a mixture of N-(p-methoxyphenyl)-
2-cyanoazadiene (18) (110 mg, 0.59 mmol) and methyl acrylate
(2.03 g, 2.13 mL, 23.7 mmol) was heated under closed condi-
tions at 90 °C for 40 h. Compounds 22 and 23, an inseparable
7:1 mixture of regioisomers, were obtained as a yellow oil (134
mg, 83%) after flash column chromatography (silica gel, 4:1
heptane/ethyl acetate).
Compound 29 (major isomer): IR (neat) 2228, 1616, 1602,
1497, 1413 cm-1; 1H NMR (CDCl3) δ 0.85 (d, J ) 7.2 Hz, 3H),
2.05 (m, 1H), 2.25 (ddd, J ) 4.5, 5.8, 13.5 Hz, 1H), 2.61 (ddd,
J ) 3.4, 6.8, 13.4 Hz, 1H), 4.70 (dd, J ) 4.1, 7.7 Hz, 1H), 5.69
(d, J ) 3.2 Hz, 1H), 7.03 (m, 3H), 7.25 (m, 7H); 13C NMR
(CDCl3) δ 20.9, 28.3, 38.6, 63.2, 116.2, 119.2, 124.5, 125.2,
126.4, 127.1, 127.4, 128.6, 129.0, 141.2, 145.4; MS m/z 274
(M+), 259, 180, 168, 104, 91, 77. Anal. Calcd for C19H18N2:
C, 83.18; H, 6.61; N, 10.21. Found: C, 83.16; H, 6.85; N, 10.16.
Compound 30 (minor isomer): 1H NMR (CDCl3) δ 1.07 (d,
J ) 6.9 Hz, 3H), 1.77 (m, 1H), 1.96-2.28 (m, 2H), 4.88 (t, J )
Compound 22: 1H NMR (CDCl3) δ 1.95 (m, 1H), 2.17-2.34
(m, 3H), 3.75 (s, 3H), 3.78 (s, 3H), 4.33 (dd, J ) 3.1, 4.4 Hz,
1H), 5.64 (dt, J ) 1.1, 4.2 Hz, 1H), 6.87 (dt, J ) 2.7, 8.9 Hz,
2H), 7.10 (dt, J ) 2.7, 8.9 Hz, 2H); 13C NMR (CDCl3) δ 20.0,
22.7, 55.5, 61.7, 114.5, 115.8, 117.7, 125.6, 138.5, 157.4, 171.8.
Compound 23: 1H NMR (CDCl3) δ 2.50 (m, 2H), 2.77 (m,
1H), 3.54 (m, 2H), 3.62 (s, 3H), 3.78 (s, 3H), 5.78 (t, J ) 4.0,
1H), 6.87 (dt, J ) 2.1, 8.9 Hz, 2H), 7.01 (dt, J ) 2.1, 8.9 Hz,
2H); 13C NMR (CDCl3) δ 25.8, 36.7, 52.5, 52.7, 61.7, 115.8,
118.3, 118.9, 124.7, 138.9, 157.3, 173.0.
3.5 Hz, 1H), 5.61 (d, J ) 1.5 Hz, 1H), 6.87-7.38 (m, 10H); 13
C
NMR (CDCl3) δ 19.5, 25.3, 34.8, 63.0, 116.3, 119.2, 124.1-
129.2, 7 peaks, 141.4, 145.9.
N-P h en yl-1,2,3,4-t et r a h yd r o-cis-2-(3-n it r op h en yl)-4-
m eth yl-6-cya n op yr id in e (31) a n d Its Tr a n s Isom er 32.
As described for 6, N-phenyl-2-cyano-4-methyl-1-aza-1,3-buta-
diene (27) (37 mg, 0.22 mmol) and 3-nitrostyrene (1.3 g, 8.8
mmol) were heated under closed conditions at 90 °C for 22 h.
Compounds 31 (31 mg, 44%) and 32 (11 mg, 16%) were
obtained as yellow oils after two flash column chromatogra-
phies (silica gel, 50:1 heptane/EtOAc):
For the mixture of regioisomers: IR (neat) 2225, 1744, 1613,
1506, 1450 cm-1; MS m/z 272 (M+), 257, 213, 134, 92; HRMS
calcd for C15H16N2O3 272.1161, found 272.1151.
Compound 31 (major isomer): IR (neat) 2228, 1609, 1595,
1532, 1356 cm-1; 1H NMR (CDCl3) δ 0.94 (d, J ) 7.2 Hz, 3H),
2.02 (dt, J ) 8.4, 13.6 Hz, 1H), 2.26 (dddd, J ) 0.6, 3.8, 5.7,
13.7 Hz, 1H), 2.71 (m, 1H), 4.77 (dd, J ) 3.7, 8.7 Hz, 1H), 5.69
N-(p-Meth oxyp h en yl)-1,2,3,4-tetr a h yd r o-2-a cetyl-6-cy-
a n op yr id in e (24) a n d N-(p-Meth oxyp h en yl)-1,2,3,4-tet-
r a h yd r o-3-a cetyl-6-cya n op yr id in e (25). As described for
6, a mixture of N-(p-methoxyphenyl)-2-cyanoazadiene 18 (100
mg, 0.54 mmol) and methyl vinyl ketone (1.1 g, 21.6 mmol)
was heated under closed conditions at 75 °C for 24 h.
Compound 24 was obtained as a white solid (78 mg, 56%) and
25 as a yellow oil (10 mg, 7%) after flash column chromatog-
raphy (silica gel, 4:1 heptane/ether).
(dd, J ) 2.5, 3.7 Hz, 1H), 6.98-7.61 (m, 8H), 8.01 (m, 1H); 13
C
NMR (CDCl3) δ 20.9, 28.4, 39.0, 63.0, 115.7, 119.8, 122.3, 122.4,
122.7, 125.5, 125.8, 126.3, 129.4, 129.7, 133.2, 143.7, 144.8,
148.5; MS m/z 319 (M+), 304, 197, 169, 155, 93, 77. HRMS
calcd for C19H17N3O2 319.1321, found 319.1302.
Compound 32 (minor isomer): IR (neat) 2228, 1602, 1532,
1497, 1349 cm-1; 1H NMR (CDCl3) δ 1.13 (d, J ) 6.9 Hz, 3H),
1.84 (ddd, J ) 4.2, 11.4, 12.9, 1H), 2.02 (m, 1H), 2.26 (m, 1H),
4.96 (t, J ) 3.5 Hz, 1H), 5.68 (dd, J ) 1.4, 2.5 Hz, 1H), 7.04-
7.73 (m, 8H), 8.17 (m, 1H); 13C NMR (CDCl3) δ 19.5, 25.2, 34.3,
62.7, 115.8, 116.6, 121.0, 122.4, 122.6, 125.0, 127.5, 129.5,
130.0, 132.2, 143.7, 145.6, 148.9.
N-P h en yl-1,2,3,4-tetr a h yd r o-cis-2-(4-br om op h en yl)-4-
m eth yl-6-cya n op yr id in e (33) a n d Its Tr a n s Isom er 34.
As described for 6, N-phenyl-2-cyano-4-methyl-1-aza-1,3-buta-
diene (27) (116 mg, 0.68 mmol) and 4-bromostyrene (5.0 g, 27.2
mmol) were heated under closed conditions at 90 °C for 44 h.
Compounds 33 (120 mg, 50%) and 34 (43 mg, 18%) were
obtained as pale yellow oils after repetitive flash column
chromatography (silica gel, 97:3 heptane/EtOAc):
Compound 24: IR (neat) 2259, 2227, 1717, 1618, 1508 cm-1
;
1H NMR (CDCl3) δ 1.75 (m, 1H), 2.17 (m, 3H), 2.33 (s, 3H),
3.79 (s, 3H), 4.11 (t, J ) 4.0 Hz, 1H), 5.79 (t, J ) 4.0 Hz, 1H),
6.88 (m, 2H), 7.06 (m, 2H); 13C NMR (CDCl3) δ 20.2, 20.7, 27.1,
55.6, 114.8, 115.8, 118.3, 121.5, 124.7, 139.1, 157.3, 208.8; MS
m/z 256 (M+), 213, 134, 92. Anal. Calcd for C15H16N2O2: C,
70.28; H, 6.29; N, 10.93. Found: C, 69.88; H, 6.46; N, 10.78.
1
Compound 25: IR (neat) 2228, 1711, 1621, 1514 cm-1; H
NMR (CDCl3) δ 2.17 (s, 3H), 2.44 (m, 2H), 2.82 (m, 1H), 3.44
(m, 1H), 3.75 (dd, J ) 3.0, 10.0 Hz, 1H), 3.80 (s, 3H), 5.81 (t,
J ) 4.2 Hz, 1H), 6.89 (m, 2H), 7.01 (m, 2H); 13C NMR (CDCl3)
δ 25.4, 28.6, 44.1, 52.4, 55.6, 114.8, 118.7, 115.5, 119.3, 124.9,
139.0, 157.3, 207.6; MS m/z 256 (M+), 213, 134, 84; HRMS
calcd for C15H16N2O2 256.1212, found 256.1210.
Compound 33 (major isomer): IR (neat) 2228, 1623, 1595,
N-P h en yl-2-cya n o-4-m eth yl-1-a za -1,3-bu ta d ien e (27).
Following the procedure described for the preparation of
N-phenyl-2-cyano-1-aza-1,3-butadiene (4), crotanilide 26 (1.777
g, 11.04 mmol) in CH2Cl2 (60 mL) containing diisopropylethyl-
amine (3.418 g, 26.50 mmol) was treated with triflic anhydride
(3.735 g, 2.23 mL, 13.25 mmol) at -73 °C, and then with LiCN
(1.093 g, 33.12 mmol) and 12-crown-4 (0.194 g, 1.1 mmol).
N-Phenyl-2-cyano-4-methyl-1-aza-1,3-butadiene (27), a 5:1
mixture of isomers, was obtained as a yellow low melting
crystalline solid (1.181 g, 63%) after flash column purification
(silica gel, 15:1 heptane/EtOAc): mp 35 °C; IR (neat) 2228,
1
1497 cm-1; H NMR (CDCl3) δ 0.86 (d, J ) 7.2 Hz, 3H), 1.97
(m, 1H), 2.20 (ddd, J ) 4.1, 5.9, 13.6 Hz, 1H), 2.61 (ddd, J )
3.2, 7.3, 13.7 Hz, 1H), 4.64 (dd, J ) 3.9, 8.0 Hz, 1H), 5.67 (d,
J ) 3.1 Hz, 1H), 6.97 (d, J ) 7.4, 2H), 7.04 (t, J ) 7.4, 1H),
7.11 (d, J ) 8.4, 2H), 7.21 (t, J ) 7.7, 2H), 7.38 (d, J ) 8.4,
2H, 2H); 13C NMR (CDCl3) δ 20.9, 28.2, 38.5, 62.7, 116.0, 119.3,
121.2, 124.7, 125.5, 126.2, 128.8, 129.1, 131.7, 140.3, 145.1;
MS m/z 354, 352 (M+), 353, 351, 339, 337, 231, 182, 184, 169,
77. Anal. Calcd for C19H17N2Br: C, 64.60; H, 4.85; N, 7.93.
Found: C, 64.44; H, 5.01; N, 7.64.
1645, 1581, 1490, 1441, 1251 cm-1 1H NMR (CDCl3) major
;
Compound 34 (minor isomer): IR (neat) 2228, 1625, 1595,
1
1490 cm-1; H NMR (CDCl3) δ 1.09 (d, J ) 6.8 Hz, 3H), 1.77
isomer δ 2.05 (dd, J ) 1.6, 6.8 Hz, 3H), 6.52 (ddd, J ) 1.5, 3.0,
15.9 Hz, 1H), 6.96 (m, 1H), 7.08 (d, J ) 7.3, 2H), 7.27 (t, J )
7.4, 1H), 7.42 (t, J ) 7.3, 2H); minor isomer δ 1.91 (dd, J )
1.6, 6.9 Hz, 3H), 6.21 (ddd, J ) 1.5, 3.1, 15.6 Hz, 1H), 6.90 (m,
1H), 7.08 (d, J ) 7.3, 2H), 7.27 (t, J ) 7.4, 1H), 7.42 (t, J )
7.3, 2H); 13C NMR (CDCl3) δ 18.8, 110.3, 119.9, 120.4, 122.8,
125.9, 127.2, 129.3, 130.8, 140.4, 144.6, 146.1, 147.9, 149.1;
MS (CI) m/z 227 (M + 57+), 171 (M + H), 144. Anal. Calcd
for C11H10N2: C, 77.62; H, 5.92; N, 16.46. Found: C, 77.50;
H, 5.81; N, 16.41.
N-P h en yl-1,2,3,4-t et r a h yd r o-cis-2-p h en yl-4-m et h yl-6-
cya n op yr id in e (29) a n d Its Tr a n s Isom er 30. As described
for 6, N-phenyl-2-cyano-4-methyl-1-aza-1,3-butadiene (27) (141
mg, 0.83 mmol) and styrene (3.45 g, 33.2 mmol) were heated
under closed conditions at 120 °C for 30 h. Compound 29 was
obtained as a white solid (141 mg, 62%), and 30 (admixture
with 29) as a pale yellow oil (33 mg, 15% from NMR) after
(ddd, J ) 4.3, 11.0, 12.6, 1H), 2.06 (m, 1H), 2.19 (m, 1H), 4.83
(dd, J ) 3.3, 3.6 Hz, 1H), 5.62 (dd, J ) 1.6, 2.5 Hz, 1H), 7.02-
7.33 (m, 7H), 7.50 (d, J ) 8.5, 2H); 13C NMR (CDCl3) δ 19.5,
25.2, 34.6, 62.6, 116.0, 120.4, 121.2, 124.4, 127.4, 127.7, 129.3,
132.0, 140.5, 145.8; MS m/z 354, 352 (M+), 353, 351, 339, 337,
231, 195, 182, 169, 93, 77.
N-P h en yl-1,2,3,4-t et r a h yd r o-cis-2-et h oxy-4-m et h yl-6-
cya n op yr id in e (35) a n d Its Tr a n s Isom er 36. As described
for 6, N-phenyl-2-cyano-4-methyl-1-aza-1,3-butadiene (27) (60
mg, 0.35 mmol) and ethyl vinyl ether (1.02 g, 14.0 mmol) were
heated under closed conditions at 120 °C for 34 h. Compounds
35 and 36 (73 mg, 88%) were obtained as a 2:1 mixture of
isomers after flash column chromatography (silica gel, 20:1
heptane/EtOAc):
Compound 35 (major isomer): 1H NMR (CDCl3) δ 1.23 (d,
J ) 5.8 Hz, 3H), 1.26 (d, J ) 5.3 Hz, 3H), 1.95 (m, 1H), 2.42

3728 J . Org. Chem., Vol. 61, No. 11, 1996
Sisti et al.
(m, 1H), 3.58 (m, 1H), 3.84 (m, 1H), 4.88 (t, J ) 2.9 Hz, 1H),
6.03 (d, J ) 4.2 Hz, 1H), 7.06-7.45 (m, 5H); 13C NMR (CDCl3)
δ 15.3, 21.6, 26.3, 31.9, 63.1, 88.8, 119.9, 120.4, 122.7, 124.4,
127.2, 129.3, 129.5, 130.8, 131.7, 144.5.
Compound 36 (minor isomer): 1H NMR (CDCl3) δ 1.14 (d,
J ) 7.2 Hz, 3H), 1.23 (d, J ) 6.7 Hz, 3H), 1.77-2.07 (m, 1H),
2.72 (m, 1H), 3.59 (m, 1H), 3.87 (m, 1H), 4.75 (t, J ) 2.6 Hz,
1H), 5.83 (dd, J ) 1.4, 2.5 Hz, 1H), 7.11 (m, 3H), 7.37 (m, 2H);
13C NMR (CDCl3) δ 18.9, 19.4, 24.8, 33.0, 63.2, 89.1, 116.4,
120.4, 123.6, 124.9, 129.6, 130.5, 131.8, 144.5.
X-r a y Str u ctu r e An a lysis of Com p ou n d 29. Crystal
data. C19H18N2, molecular weight 274.37; monoclinic system;
space group P21/n; Z ) 4; a ) 12.699(4), b ) 7.909(2), c )
16.149(5) Å, â ) 107.55(5)°; V ) 1546(1) Å3; dc ) 1.18 g cm-3
;
F(000) ) 584; l (Cu KR) ) 1.5418 Å; µ ) 0.50 mm-1 (absorption
ignored). Data were collected on CAD4 Enraf-Nonius
diffractometer using graphite monochromated Cu KR radiation
From the 3804 reflexions measured by the (θ-2θ) scan
a
.
technique up to θ ) 66, 2697 were independent (Rint ) 0.024)
and 1842 were considered as observed with I > 3.0σ(I), σ(I)
from counting statistics.
For the mixture of isomers: IR (neat) 2228, 1645, 1631,
1595, 1490 cm-1; MS (FAB) m/z 242 (M+). Anal. Calcd for
C15H18N2O: C, 74.35; H, 7.49; N, 11.56. Found: C, 74.29; H,
7.20; N, 11.78.
The structure was solved by direct methods using
SHELXS8647 and refined by full matrix least-squares with
SHELX76,48 minimizing the function Σw(Fo
-
Fc )2. The
hydrogen atoms, located in difference Fourier maps, were
introduced in theoretical position (d(C-H) ) 1.00 Å) and
assigned an isotropic thermal factor equivalent to that of the
bonded carbon atom, plus 10%. Convergence was reached at
N -P h e n yl-1,2,3,4-t e t r a h yd r o-cis-2-ca r b om e t h oxy-4-
m eth yl-6-cya n op yr id in e 37, Its Tr a n s Isom er 38, a n d
N-P h en yl-1,2,3,4-tetr a h yd r o-cis-3-ca r bom eth oxy-6-cya n -
op yr id in e (39) a n d Its Tr a n s Isom er 40. As described for
6, N-phenyl-2-cyano-4-methyl-1-aza-1,3-butadiene (27) (120
mg, 0.70 mmol) and methyl acrylate (2.43 g, 28.0 mmol) were
heated under closed conditions at 120 °C for 40 h. Compounds
37-40 (64 mg, 36%, mixture of isomers, 5:1:2.6:2) were
obtained as a yellow oil after flash column chromatography
(silica gel, 20:1 heptane/EtOAc):
For the mixture of isomers: IR (neat) 2231, 1738, 1619,
1606, 1500 cm-1; 1H NMR (CDCl3) δ 1.02 (d, J ) 7.3 Hz, 4-Me),
1.03 (d, J ) 6.6 Hz, 4-Me), 1.14 (d, J ) 6.9 Hz, 4-Me), 1.29
(dd, J ) 6.4, 13.3 Hz, H-4), 1.43 (d, J ) 6.7 Hz, H-3), 1.64
(ddd, J ) 4.7, 11.3, 13.1 Hz, H-3), 2.21 (t, J ) 5.1 Hz, H-3),
2.30-2.54 (m, H-5), 2.85 (m, H-2, H-3), 3.56 (dd, J ) 11.1, 12.8
Hz, H-2), 3.59 (s, Me), 3.68 (s, Me), 3.73 (s, Me), 3.77 (s, Me),
3.82 (dd, J ) 1.3, 2.7 Hz, H-2), 4.36 (t, J ) 5.0 Hz, H-2), 4.41
(dd, J ) 2.9, 4.6 Hz, H-2), 5.58 (dd, J ) 1.5, 2.5 Hz, H-5), 5.72
(d, J ) 4.1 Hz, H-5), 5.76 (d, J ) 3.8 Hz, H-5), 5.79 (d, J ) 4.5
Hz, H-5), 7.10 (m, ArH), 7.26 (m, ArH), 7.34 (m, ArH); 13C NMR
(CDCl3) δ 15.9, 17.4, 19.8, 21.2, 26.2, 26.6, 27.8, 30.0, 31.5,
31.7, 32.2, 41.5, 44.6, 47.4, 50.4, 51.6, 51.8, 52.1, 52.4, 52.7,
57.2, 59.8, 61.8, 115.5, 115.7, 116.6, 117.2, 117.7, 119.7, 119.9,
120.0, 122.2, 122.4, 122.6, 122.9, 123.0, 123.5, 123.9, 124.5,
124.9, 125.0, 125.5, 125.7, 125.8, 127.1, 127.3, 128.4, 129.3,
129.6, 129.8, 134.2, 144.8, 145.0, 145.4, 172.0; MS m/z 256
(M+), 241, 205, 197, 181, 169, 93, 77. Anal. Calcd for
C15H16N2O2: C, 70.29; H, 6.29; N, 10.93. Found: C, 70.49; H,
6.25; N, 10.99.
2
1/2
R ) 0.053 and Rw ) 0.090 (with Rw ) {Σw(Fo - Fc )2/ΣwFo
}
and w ) 1/[σ2(Fo) + 0.0058Fo2]. No residual was higher than
0.29 e Å-3 in the final difference map.49
Ack n ow led gm en t. Support of this work by NATO
(900305), NIH (GM3972903), and a Chateaubriand
scholarship for N.J .S. is gratefully acknowledged. We
also thank the Centre de Calcul (IDRIS) for calculation
time, the microanalytical department at the ICSN for
their comprehension and expertise, and Dr. Leticia
Toledo for her help with the X-ray diffraction studies.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
data with complete peak assignments for all compounds, and
an ORTEP drawing for 18 (12 pages). This material is
contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
J O9520607
(47) TEXAN (1985). TEXRAY Structure Analysis Package Molecular
Structure Corporation.
(48) Sheldrick, G. M. SHELXS86. Program for the solution of crystal
structures. Univ. of Go¨ttingen, Germany, 1986.
(49) Sheldrick, G. M. SHELX76. Program for crystal structure
determination. Univ. of Cambridge, England, 1976.