Stereochemistry in the reaction of 4-methyl-1,2,4-triazoline-3,5-dione (MTAD) with β,β-dimethyl-p-methoxystyrene. Are open biradicals the reaction intermediates?

Article abstract of DOI:10.1021/jo0005518

The reaction of 4-methyl-1,2,4-triazoline-3,5-dione (MTAD) with β,β-dimethyl-p-methoxystyrene (1) in chloroform affords four adducts: the ene, two stereoisomeric [4 + 2]/ene diadducts, and a minor product that is probably the double Diels-Alder diadduct. In methanol, only one regioisomeric methoxy adduct is formed. The stereochemistry of the reaction was examined by specific labeling of the anti methyl group of 1 as CD₃. In chloroform, the ene adduct is formed with >97% syn selectivity, while the [4 + 2]/ene diadducts are formed with 20% loss of stereochemistry at the methyl groups. In methanol, the methoxy adducts are formed with almost complete loss of stereochemistry. A mechanism involving open biradicals is inconsistent with the experimental results. It is likely that the reaction proceeds through the formation of an aziridinium imide and an open zwitterionic intermediate. The aziridinium imide leads to the formation of the ene adduct. The open zwitterion, which has sufficient lifetime to rotate around the C-C bond, leads to the formation of a [4 + 2] cycloadduct, which reacts with a second molecule of MTAD in an ene-type mode to afford two stereoisomeric [4 + 2]/ene diadducts. In methanol, solvent captures the zwitterionic intermediate and forms the methoxy adduct. The relative distribution of the products in chloroform depends on the reaction temperature. Lower temperatures favor the ene reaction (entropically favorable), whereas at higher temperatures the [4 + 2]/ene diadducts become the major products.

Full text of DOI:10.1021/jo0005518

3682
J . Org. Chem. 2001, 66, 3682-3687
Ster eoch em istr y in th e Rea ction of
4-Meth yl-1,2,4-tr ia zolin e-3,5-d ion e (MTAD) w ith
â,â-Dim eth yl-p-m eth oxystyr en e. Ar e Op en Bir a d ica ls th e Rea ction
In ter m ed ia tes?†
Manolis Stratakis,* Maria Hatzimarinaki, George E. Froudakis, and Michael Orfanopoulos
Department of Chemistry, University of Crete, 71409 Iraklion, Greece
stratakis@chemistry.uoc.gr
Received April 12, 2000
The reaction of 4-methyl-1,2,4-triazoline-3,5-dione (MTAD) with â,â-dimethyl-p-methoxystyrene
(1) in chloroform affords four adducts: the ene, two stereoisomeric [4 + 2]/ene diadducts, and a
minor product that is probably the double Diels-Alder diadduct. In methanol, only one regioisomeric
methoxy adduct is formed. The stereochemistry of the reaction was examined by specific labeling
of the anti methyl group of 1 as CD3. In chloroform, the ene adduct is formed with >97% syn
selectivity, while the [4 + 2]/ene diadducts are formed with 20% loss of stereochemistry at the
methyl groups. In methanol, the methoxy adducts are formed with almost complete loss of
stereochemistry. A mechanism involving open biradicals is inconsistent with the experimental
results. It is likely that the reaction proceeds through the formation of an aziridinium imide and
an open zwitterionic intermediate. The aziridinium imide leads to the formation of the ene adduct.
The open zwitterion, which has sufficient lifetime to rotate around the C-C bond, leads to the
formation of a [4 + 2] cycloadduct, which reacts with a second molecule of MTAD in an ene-type
mode to afford two stereoisomeric [4 + 2]/ene diadducts. In methanol, solvent captures the
zwitterionic intermediate and forms the methoxy adduct. The relative distribution of the products
in chloroform depends on the reaction temperature. Lower temperatures favor the ene reaction
(entropically favorable), whereas at higher temperatures the [4 + 2]/ene diadducts become the major
products.
In tr od u ction
azolinedione ene reactions were generally considered to
proceed through the formation of an aziridinium imide
(AI) as the intermediate.
However, on the basis of experimental results and
theoretical calculations, Singleton recently proposed a
reasonable mechanism according to which, apart from
AIs, open biradicals are the key intermediates.9 Open
zwitterionic intermediates have been proposed to par-
ticipate only in the reactions with electron-rich alkenes10
and dienes.11,12
Recently, we reported that the addition of 4-methyl-
1,2,4-triazoline-3,5-dione (MTAD) to â,â-dimethylsty-
rene13 in CH2Cl2 or in methanol affords exclusively the
ene product with >97% syn selectivity. It was proposed
that, in the transition state of the aziridinium imide
formation, there is a stabilizing interaction between the
negatively charged nitrogen of MTAD and the phenyl
ring that directs the ene reaction to take place from the
syn allylic methyl group. For a further study of this
Triazolinediones (RTADs)1 react smoothly with olefins
bearing allylic hydrogens in aprotic solvents to give ene
adducts. However, in protic solvents such as methanol,
RTAD-methoxy adducts are formed. With trisubstituted
alkenes, the new C-N bond in the ene adduct is formed
at the less substituted olefinic carbon2 (Markovnikov
selectivity). For trisubstituted alkenes also, when the
reaction is carried out in polar protic solvents such as
methanol, the solvent traps the intermediate regio-
specifically, forming a methoxy adduct at the more
substituted olefinic carbon. The methanol trapping reac-
tion with di- and trisubstituted alkylalkenes is stereo-
specific. Only one diastereomeric methoxy adduct is
formed.3 As a result of stereoisotopic studies,4,5 low-
temperature NMR,6,7 and theoretical calculations,8 tri-
† Taken in part from the undergraduate research thesis of M.
Hatzimarinaki.
(1) Radl, S. Adv. Heterocycl. Chem. 1997, 67, 119-204.
(2) Cheng, C. C.; Seymour, C. A.; Petti, M. A.; Greene, F. D.; Blount,
J . F. J J . Org. Chem. 1984, 49, 2910-2916.
(3) Smonou, I.; Khan, S.; Foote, C. S.; Elemes, Y.; Mavridis, I. M.;
Pantidou, A.; Orfanopoulos, M. J . Am. Chem. Soc. 1995, 117, 7081-
7087.
(8) Chen, J . S.; Houk, K. N.; Foote, C. S. J . Am. Chem. Soc. 1997,
119, 9852-9855.
(9) Singleton, D. A.; Hang, C. J . Am. Chem. Soc. 1999, 121, 11885-
11893.
(4) Seymour, C. A.; Greene, F. D. J . Am. Chem. Soc. 1980, 102,
6384-6385.
(10) Smonou, I.; Orfanopoulos, M.; Foote, C. S. Tetrahedron Lett.
1988, 29, 2769-2772.
(5) Orfanopoulos, M.; Smonou, I.; Foote, C. S. J . Am. Chem. Soc.
1990, 112, 3607-3614.
(11) J ensen, F.; Foote, C. S. J . Am. Chem. Soc. 1987, 109, 6376-
6381.
(6) Squillacote, M.; Mooney, M.; De Felippis, J . J . Am. Chem. Soc.
1990, 112, 5364-5365.
(12) Clennan, E. L.; Earlywine, A. D. J . Am. Chem. Soc. 1987, 109,
7104-7108.
(13) Stratakis, M.; Orfanopoulos, M.; Foote, C. S. J . Org. Chem.
1998, 63, 1315-1318.
(7) Poon, T. H. W.; Park, S. H.; Elemes, Y.; Foote, C. S. J . Am. Chem.
Soc. 1995, 117, 10468-10473.
10.1021/jo0005518 CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/05/2001

Reaction of MTAD with â,â-Dimethyl-p-methoxystyrene
J . Org. Chem., Vol. 66, No. 11, 2001 3683
Sch em e 1. P r od u cts in th e Rea ction of 1 w ith
MTAD in Ch lor ofor m a n d Meth a n ol a t 20 °C
methanol, however, only the methoxy trapping adduct
1e is formed, in which the solvent has captured regio-
specifically the secondary benzylic carbon. The stereo-
chemistry was determined by NOE experiments. Upon
irradiation of the methoxy group at 3.81 ppm, a positive
enhancement was measured on the benzylic hydrogen.
None of the diastereotopic geminal methyls showed
enhancement.
The formation of two stereoisomeric diadducts, 1b or
1c, is quite surprising and unprecedented. They could
be either conformational isomers due to the hindered
rotation of the urazole ring at the benzylic position or
diastereomers due to slow pyramidal inversion of the two
adjacent nitrogens in the six-membered ring. Consider-
ing, however, a planar urazole ring, there should not be
two such isomers. We attempted several times to obtain
crystals of 1b or 1c suitable for X-ray analysis, without
success. However, the crystal structure of the adduct 215b
derived from the reaction of 2,5-dichlorostyrene with
MTAD indicates that neither the six-membered ring nor
the five-membered urazole ring are planar.
Consistent with the crystallographic data are our
theoretical calculations. The geometry optimization of the
[4 + 2]/ene diadduct gives the two structures 1b and 1c
(Figure 1) as local minima, in which the urazole ring is
not completely planar. The calculations were performed
with the GAUSSIAN 94 program package.17 The theo-
retical treatment of the structures involved Density
Functional Theory (DFT) with the three-parameter hy-
brid functional of Becke and the use of Lee-Yang-Parr
correlation functional (B3LYP).18 The level of the calcula-
exceptional stereoselectivity, we studied the stereochem-
istry in the reaction of MTAD with â,â-dimethyl-p-
methoxystyrene (1).
Resu lts
The reaction of 1 with MTAD proceeds instantly in
CDCl3 at 20 °C (Scheme 1) and affords a mixture of
products that were fractionated by flash column chro-
matography (hexane/ethyl acetate ) 1/2). The yields of
(16) The minor product 1d has the following characteristic absorp-
tions in the 1H NMR spectrum of the crude reaction mixture: 6.17
(br. s, 1H), 5.83 (d, J ) 2.0 Hz, 1H), 5.25 (dd, J 1 ) 6.8 Hz, J 2 ) 2.0 Hz,
1H), 5.23 (d, J ) 6.8 Hz, 1H), 4.05 (br. s, 1H), 3.65 (s, 3H); the N-methyl
absorption(s) are buried in the region of 3-3.3 ppm, 1.70 (s, 3H), 1.34
(s, 3H). Decoupling experiments indicated that there is a correlation
between these six absorptions. We propose that product 1d is a single
diastereomer and one of the diadducts A or B (double Diels-Alder)
and C or D (Diels-Alder/[2 + 2]). More likely it is one of A or B. Similar
double Diels-Alder diadducts are well-known in the reaction of
styrenes with triazolinediones.1
1
the products were measured from the H NMR spectrum
of the crude reaction mixture. The first eluted product
was the ene adduct 1a (36%), followed by two slightly
separable products 1b (27%) and 1c (27%), with very
similar 1H and 13C NMR spectra. Both had the pattern
of a 1,2,4-trisubstituted benzene and were characterized
as stereoisomers of the diadduct shown in Scheme 1.
HRMS in combination with 2D-NMR heteronuclear (1H-
13C HMQC) correlation experiments allowed us to assign
the signals of all protons and carbons and to establish
the possible structures of these compounds. Formation
of diadducts with the structure of 1b or 1c is well-known
in the literature, in the reaction of PTAD with styrene14
and some substituted styrenes.15 Finally, the minor
adduct 1d (10%) of the crude reaction mixture is unstable
under the chromatographic conditions, and despite sev-
eral efforts we were unable to isolate and characterize it
properly.16 Therefore, we will not refer to it in the
accompanying discussion section of this manuscript. In
(14) Oshikawa, T.; Yamashita, M. Bull. Chem. Soc. J pn. 1983, 56,
2857-2858.
(15) (a) Pasto, D. J .; Chen, A. F-T. Tetrahedron Lett. 1973, 713-
716. (b) Lai, Y.-C.; Mallakpour, S. E.; Butler, G. B.; Palenik, G. J . J .
Org. Chem. 1985, 50, 4378-4381.

3684 J . Org. Chem., Vol. 66, No. 11, 2001
Stratakis et al.
Sch em e 2
Ta ble 1. Rela tive % Distr ibu tion of th e Ad d u cts a t
Differ en t Tem p er a tu r es
temp (°C)
1a
1b
1c
1d
-30
5
20
45
77
39
36
33
10
27
27
29
7
21
27
29
6
13
10
9
midal inversion simultaneously. The two nitrogens have
electron pairs in a trans arrangement. In the transition
state of the interconversion the urazole ring becomes
coplanar to the phenyl ring. Thus, electronic repulsions
occur between the electron pairs, which are now parallel.
Furthermore, significant steric strain develops between
the geminal methyls and the substituents on the benzylic
position (urazole and H), since they adopt eclipsing
conformations to each other.
F igu r e 1. Calculated structures of 1b and 1c.
tions used herein was B3LYP/6-31G*. Many different
starting conformations were used, but after the optimiza-
tions we found the structures 1b and 1c as local minima.
According to the calculations, the structure 1c is more
stable compared to 1b by 4.0 kcal/mol. For 1b, the N next
to the phenyl ring is almost sp2-hybridized (dihedral
angle 161°). The adjacent N atom is sp3-hybridized to a
significant extent (dihedral angle 136°). Similarly, for 1c,
the dihedral angles of the two adjacent nitrogens in the
six-membered ring are 164° and 140°, respectively. On
heating a mixture of 1b and 1c in refluxing CDCl3 one
of the two stereoisomers, probably 1b, which is the less
stable, is converted slowly to the more stable (1c). After
It is reasonable to assume that the diadducts 1b and
1c have as a common intermediate, the monoadduct
resulting from Diels-Alder cycloaddition of MTAD to 1
(Scheme 2). This highly reactive [4 + 2] monoadduct can
be observed only at low conversions of 1 and immediately
after the addition of MTAD (see Supporting Information).
1
We were able to identify in the H NMR spectrum of the
reaction mixture at less than 10% conversion of 1,
characteristic absorptions at 6.19 (1H ?, possibly a
doublet buried under the absorption of the olefinic
hydrogen of 1), 5.87 (br. s, 1H), 5.83 (d, J ) 8.9 Hz, 1H),
5.43 (d, J ) 3.8 Hz, 1H), 5.31 (br. s, 1H), 3.64 (s, 3H),
1.70 (s, 3H) and 1.52 ppm (s, 3H). However, during the
accumulation of the decoupling experiments the signals
disappeared. At such very low conversions (∼10%), the
relative ratio of the ene adduct versus the [4 + 2]/ene
diadducts is 56/44, whereas after complete reaction of 1
with MTAD, the ratio becomes 40/60. This is reasonable
because the initially formed [4 + 2] monoadduct (which
is present in the reaction mixture) reacts with a second
molecule of MTAD in an ene-type mode to give 1b and
1c.
The relative distribution of the reaction adducts in
chloroform depends significantly on the reaction tempera-
ture (Table 1). At -30 °C, the ene adduct is formed in
77% relative yield, while by increasing the reaction
temperature, the amounts of the adducts 1b and 1c
increase substantially.
To study the stereochemistry of the electrophilic ad-
dition and shed some light to reaction mechanism, the
trans methyl group with respect to the phenyl ring of 1
was selectively labeled as CD3 via sequential reduc-
1
24 h, only 1c appears in the H NMR spectrum, without
formation of other byproducts.19 On the other hand, under
the same conditions (heating at 80 °C for 12 h) the pure
[4 + 2]/ene diadduct isolated from the column (1c) did
not transform to 1b. It is reasonable that 1b, which was
calculated to be less stable than 1c, transforms quanti-
tatively to 1c since their free energy difference is very
high (4 kcal/mol). We postulate that the diadducts 1b and
1c are diastereomeric as a result of the high energy
barrier for the pyramidal inversion of the adjacent
nitrogens in the six-membered ring. In the literature
there are several examples of slow pyramidal inversion
of nitrogens when they are connected to another hetero-
atom20 with lone pair(s) of electrons, such as oxygen,
nitrogen, or halide. Even in cases when between two
adjacent nitrogens one is connected to an acyl substitu-
ent, the slow pyramidal inversion still exists.21 In the
present case, during the conversion of 1b to 1c, the two
nitrogens in the six-membered ring must undergo pyra-
(17) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Robb, M. A.; Cheeseman, J . R.; Keith, T.; Petersson,
G. A.; Montgomery, J . A.; Raghavachari, K.; Al-Laham, M. A.;
Zakrewski, V. G.; Ortiz, J . V.; Foresman, J . B.; Cioslowski, J .; Stefanov,
B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.;
Chen, W.; Wong, M. W.; Andres, J . L.; Replogle, R.; Gomperts, R.;
Martin, R. L.; Fox, D. L.; Binkley, J . S.; Defrees, D. J .; Baker, J .;
Stewart, J . P.; Head-Gordon, M.; Gonzalez, C.; Pople, J . A. Gaussian
94, Revision D.4; Gaussian, Inc.: Pittsburgh, PA, 1995.
(18) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785-789.
(19) The formation of two [4 + 2]/ene stereoisomers such as 1b and
1c appears generally in the reaction of MTAD with styrene derivatives
that possess an electron-donating group at the para position, such as
trans-anethole, p-methoxystyrene, or p-methylstyrene. Details of these
studies as well as the kinetics of the interconversion between these
stereoisomers will be reported elsewhere.
(20) (a) Brois, S. J . J . Am. Chem. Soc. 1968, 90, 506-508; 508-
509. (b) Mu¨ller, K.; Eschenmoser, A. Helv. Chim. Acta 1969, 52, 1823-
1830. (c) Dobler, M.; Dunitz, J . D.; Hawley, D. M. Helv. Chim. Acta
1969, 52, 1831-1833. (d) Kostyanovsky, R. G.; Rudchenko, V. F.;
Shtamburg, V. G.; Chervin, I. I.; Nasibov, S. S. Tetrahedron 1981, 37,
4245-4254. (e) Shustov, G. V.; Kadorkina, G. K.; Kostyanovsky, R.
G.; Rauk, A. J . Am. Chem. Soc. 1988, 110, 1719-1726. (f) Kosty-
anovsky, R. G.; Kadorkina, G. K.; Kostyanovsky, V. R.; Schuring, V.;
Trapp, O. Angew. Chem., Int. Ed. 2000, 39, 2938-2940.
(21) Hilpert, H.; Hoesch, L.; Dreiding, A. S. Helv. Chim. Acta 1985,
68, 1691-1697.

Reaction of MTAD with â,â-Dimethyl-p-methoxystyrene
J . Org. Chem., Vol. 66, No. 11, 2001 3685
Sch em e 3. Isola ted P r od u cts in th e Rea ction of 3
w ith MTAD in Ch lor ofor m a
Sch em e 4. Rea ction of 3 w ith MTAD in Meth a n ol
Sch em e 5. Mech a n ism In volvin g Bir a d ica ls in th e
Rea ction of â,â-d im eth ylstyr en e w ith MTAD
corresponding to adducts 3e and 3f again in an 80/20
ratio.25 It is reasonable to assume that the major com-
ponents in both ratios are the adducts 3c and 3e, which
are expected to be the only products if the initial [4 + 2]
cycloaddition was suprafacial, according to the Wood-
ward-Hofmann rules. All product ratios were corrected
for 100% geometric purity of 3.
When the reaction was run in methanol-d4 (Scheme
4) where a single regioisomeric methoxy-trapping adduct
is formed, two almost equivalent methyl resonances
appear at 1.50 and 1.31 ppm, respectively. The product
1
was purified by flash column chromatography. In the H
NMR spectrum of the purified adduct (taken in CDCl3)
two methyl resonances appear at 1.46 and 1.44 ppm,
corresponding to the diastereomeric adducts 3g and 3h
in 55/45 ratio. It is notable that the tertiary proton on
the methoxy functionality has two slightly separable
resonances at 4.39 ppm differing by 0.4 Hz, attributable
to the diastereomeric adducts 3g and 3h .
a
Yields for the ene mode pathway correspond to the relative
ratio of 3a and 3b.
tion22,23 of (E)-2-methyl-3-(p-methoxyphenyl)prop-2-enoic
acid, which was prepared by a modified Horner-Wittig
reaction.24 The geometrical purity of the labeled alkene
3 was 96%.
Discu ssion
Recently, Singleton and Hang9 proposed that for the
ene reactions of triazolinediones with alkenes open
biradicals, in which the barrier to rotation around the
C-N bond is higher than the barrier to afford the ene
product, are the key intermediates. More specifically, to
explain the significant inverse â-secondary isotope effect
of kH/kD ) 0.955 per D atom, in the ene reaction of MTAD
with â,â-dimethylstyrene, the mechanism shown in
Scheme 5 was proposed, in contrast to the aziridinium
imide mechanism proposed earlier.13
The reaction of 3 with MTAD in CDCl3 at 20 °C
afforded a mixture of adducts, which were fractionated
by flash column chromatography. The isolated products
(Scheme 3) were again the ene and the two stereoisomeric
[4 + 2]/ene diadducts. The ene adduct (3a ) was formed
with >97% hydrogen abstraction from the methyl group
that is syn to phenyl ring (H-abstraction in the present
case), as has been found in the reaction of MTAD with
â,â-dimethylstyrene.13
Considering a mechanism involving open biradicals as
intermediates in the reaction of 3 with MTAD, it’s
difficult to rationalize the stereochemistry of the path-
ways leading to the ene and the [4 + 2]/ene products.
The first eluted diastereomeric [4 + 2]/ene diadduct
exhibits two methyl group resonances at 1.92 and 1.31
ppm corresponding to adducts 3c and 3d in a ratio 80/
20, respectively. Similarly, the more polar diadduct
exhibits two methyl resonances at 1.94 and 1.31 ppm
(25) Since both diastereomeric [4 + 2]/ene diadducts exhibit a methyl
group resonance at 1.31 ppm, and we were unable to isolate by column
chromatography the more polar diadduct in pure form, but rather as
a mixture with the less polar diadduct, the reaction of 1 with 4-phenyl-
1,2,4-triazoline-3,5-dione (PTAD) was performed in CDCl3 at 20 oC.
The two [4 + 2]/ene diadducts, which were formed in approximately
2/1 ratio, exhibit different absorptions for all four diastereomeric
methyls. The major [4 + 2]/ene diadduct has two methyl absorptions
at 1.94 and 1.35 ppm, while for the minor diadduct the methyl groups
resonate at 1.99 and 1.41 ppm, respectively. In the reaction of 3 with
PTAD, each diadduct exhibits two methyl absoprtions in approximately
80/20 ratio as found in the reaction of 3 with MTAD.
(22) Collinghton, E. W.; Meyers, A. I. J . Org. Chem. 1971, 36, 3044-
3045.
(23) Grdina, M. B.; Orfanopoulos, M.; Stephenson, L. M. J . Org.
Chem. 1979, 44, 2936-2938.
(24) Brittelli, D. R. J . Org. Chem. 1981, 46, 2514-2520.

3686 J . Org. Chem., Vol. 66, No. 11, 2001
Stratakis et al.
Sch em e 6. Mech a n ism in th e Rea ction of 3 w ith
MTAD In volvin g Bir a d ica ls a s In ter m ed ia tes
Sch em e 7. P r op osed Mech a n ism in th e Rea ction
of MTAD w ith 1
Let’s assume biradical BRI (Scheme 6) as a common
intermediate for the reaction pathways. The loss of
stereochemistry of the [4 + 2]/ene diadducts indicates
that partial rotation around the C-C bond must occur
in BRI before forming the [4 + 2] monoadduct. However,
if rearrangement of the partially rotated BRI occurs to
the biradical BR II that leads to the ene product, both
geminal methyl groups would be reactive. This is in
contrast to the >97% syn methyl selectivity for the ene
pathway found in 3.
We propose that in chloroform two intermediates are
partitioning, the aziridinium imide and an open zwitter-
ion, while in methanol the open zwitterion prevails. The
stereochemistry of the products in the reaction of 3 with
MTAD can be explained on the basis of these two
intermediates, as follows:
(a ) En e P a th w a y. The high preference for syn ene
addition observed for 3 (>97%) is identical to that
reported in the addition of MTAD to â,â-dimethylsty-
rene.13 This exceptional specificity had been rationalized
in terms of the favorable formation of the aziridinium
imide intermediate in the more substituted side of the
alkene, where the negatively polarized nitrogen interacts
with the phenyl ring. We propose that the same reason
is responsible for the high syn ene preference in 3. It is
remarkable however that, although reactions of RTADs
with trisubstituted alkenes follow the Markovnikov rule2
and the secondary p-methoxybenzylic carbocation is far
more stable than the tertiary alkyl, the ene pathway is
still favorable and even the major one when the reaction
takes place at temperatures below 0 °C. The driving force
of AI formation is the strong interaction of the MTAD to
the electron-rich aryl ring.
(b) [4 + 2]/en e P a th w a y. Concerning the pathway of
the formation of the [4 + 2]/ene diadducts, the reaction
is not stereospecific. If the initial [4 + 2] reaction was
concerted (suprafacial mode), only the diastereomers 3c
and 3e should be formed. We propose that an open
zwitterionic intermediate is formed that has enough
lifetime for partial rotation and then closes to form the
[4 + 2] adducts in a non-suprafacial manner. Subse-
quently, the monoadduct undergoes a second ene-type
addition with MTAD, which leads to the [4 + 2]/ene
diadducts. The loss of stereochemistry in the diadducts
derives from the initial step during the formation of the
[4 + 2] adduct. Since the Diels-Alder monoadduct is the
common precursor for the diastereomeric [4 + 2]/ene
diadducts, identical stereochemical outcome is expected
for both diastereomers, as occurs indeed.
[4 + 2] pathway in chloroform can be attributed to the
higher dielectric constant of methanol, which increases
the lifetime of the zwitterion and eventually its free
rotation around the C-C or the C-N bonds.
The temperature dependence of the products distribu-
tion in the reaction of 1 with MTAD in chloroform (Table
1) is also consonant with the participation of the two
intermediates. The aziridinium imide leading to the ene
product is more structurally organized and is expected
to be entropically favored at lower temperatures. On the
other hand, the open zwitterion leading to the [4 + 2]
adduct is less organized as a result of the free rotation
around the C-C or the C-N bonds and thus more
favorable at higher temperatures.
According to these results we propose that MTAD adds
to â,â-dimethyl-p-methoxystyrene and forms two inter-
mediates, an aziridinium imide and an open zwitterion
(Scheme 7). Their relative distribution depends on the
reaction temperatute. Low temperatures favor formation
of the AI (entropically favorable), while at higher tem-
peratures the open zwitterion prevails. The possibility
that initially only the aziridinium imide is formed and
subsequently it rearranges to an open zwitterion cannot
be excluded. The aziridinium imide in which MTAD is
placed syn to the aryl ring leads to the ene products. The
open zwitterion has enough lifetime to rotate before
forming the cycloadducts or being trapped by methanol.
The two pathways are competitive in chloroform, but in
the more polar methanol the zwitterionic pathway pre-
vails.
In conclusion, the stereochemistry of the products in
the addition of MTAD to â,â-dimethyl-p-methoxystyrene-
d3 (3) indicates that the reaction proceeds through the
formation of two intermediates: (i) an open zwitterion,
and (ii) an aziridinium imide where the MTAD is directed
toward the more substituted side of the alkene. At least
for styrene-type substrates such as 1, a mechanism
involving biradicals as proposed recently9 is less likely
to occur.
Exp er im en ta l Section
Nuclear magnetic resonance spectra were obtained on a 500
MHz instrument. Isomeric purities were determined by 1H
NMR and by GC on an HP-5 capillary column. All spectra
reported herein were taken in CDCl3.
1-(p-Meth oxyp h en yl)-2-m eth ylp r op -1-en e (1). The com-
pound was prepared in 65% yield by Wittig coupling of
(c) Meth a n ol Tr a p p in g P a th w a y. Methanol traps
the intermediate zwitterion, which has aldready rotated,
and results in a 55/45 ratio of the two diastereomeric
methoxy adducts 3g and 3h . The higher loss of stereo-
specificity in the trapping reaction compared to the

Reaction of MTAD with â,â-Dimethyl-p-methoxystyrene
J . Org. Chem., Vol. 66, No. 11, 2001 3687
isopropyltriphenylphosphorane with p-methoxybenzaldehyde.
Purification was accomplished by vacuum distillation. 1H
NMR: 7.17 (d, J ) 8.7 Hz, 2H), 6.88 (d, J ) 8.7 Hz, 2H), 6.23
(s, 1H), 3.83 (s, 3H), 1.91 (s, 3H), 1.87 (s, 3H). 13C NMR:
157.62, 133.90, 131.38, 129.77, 124.53, 113.48, 55.21, 27.74,
19.28.
Subsequently, 4.8 mL of p-methoxybenzaldehyde was syringed
over a period of 5 min. The reaction mixture was stirred for 1
h at room temperature and then quenched with 5 mL of
ethanol. The crude organic layer was poured into a beaker
containing 500 mL of water. The aqueous layer was washed
with ether, acidified and then extracted with ether. The
geometrical purity of the isolated acid was >97% E. 1H NMR:
7.78 (s, 1H), 7.43 (d, J ) 8.5 Hz, 2H), 6.94 (d, J ) 8.5 Hz, 2H),
3.85 (s, 3H), 2.15 (s, 3H).
Rea ction of 1 w ith MTAD in Ch lor ofor m . The reaction
mixture consists of four adducts. On attempted flash column
purification, only the ene and two diastereomeric [4 + 2]/ene
diadducts, which are the major products (∼90%), were isolated.
1H NMR of th e en e a d d u ct 1a : 7.60 (br. s, 1H, N-H
exchangeable with D2O), 7.25 (d, J ) 8.6 Hz, 2H), 6.90 (d, J )
8.5 Hz, 2H), 5.67 (s, 1H), 5.17 (s, 1H), 4.93 (s, 1H), 3.84 (s,
3H), 3.06 (s, 3H), 1.75 (s, 3H). 13C NMR: 159.82, 154.69,
153.57, 141.33, 130.05, 126.92, 114.43, 114.25, 63.90, 55.28,
25.19, 21.15. 1H NMR of t h e less p ola r [4 + 2]/en e
d ia d d u ct: 8.22 (br. s, 1H, N-H exchangeable with D2O), 8.10
(d, J ) 1.7 Hz, 1H), 7.13 (d, J ) 8.5 Hz, 1H), 6.68 (dd, J 1 ) 8.5
Hz, J 2 ) 1.7 Hz, 1H), 5.04 (s, 1H), 3.85 (s, 3H), 3.15 (s, 3H),
3.05 (s, 3H), 1.92 (s, 3H), 1.31 (s, 3H). 13C NMR: 161.13,
154.44, 151.83, 147.95, 133.90, 130.52, 111.71, 108.88, 100.37,
60.99, 60.45, 55.53, 25.31, 25.17, 24.11, 23.06. HR-MS (MALDI-
FTMS): (M + Na)+, calculated for C17H20O5N6 411.1387;
experimental 411.1397. 1H NMR of th e m or e p ola r [4 +
2]/en e d ia d d u ct: 9.25 (br. s, 1H, N-H exchangeable with
D2O), 8.11 (d, J ) 1.8 Hz, 1H), 7.52 (d, J ) 8.6 Hz, 1H), 6.69
(dd, J 1 ) 8.6 Hz, J 2 ) 1.8 Hz, 1H), 5.57 (s, 1H), 3.87 (s, 3H),
3.19 (s, 3H), 3.05 (s, 3H), 1.94 (s, 3H), 1.31 (s, 3H). 13C NMR:
161.63, 154.15, 151.46, 150.98, 148.07, 133.42, 132.28, 110.86,
109.85, 100.57, 78.93, 59.88, 55.49, 25.88, 25.12, 23.19, 22.84.
Rea ction of 1 w ith MTAD in Meth a n ol. Only one product
was formed resulting from methanol addition to the inter-
mediate(s). The crude adduct was purified by flash column
(E)-Meth yl 2-Meth yl-3-(p-m eth oxyph en yl)pr op-2-en oate.
The above crude acid was dissolved in 100 mL of methanol,
and a catalytic amount of p-toluenesulfonic acid was added.
After overnight refluxing it was neutralized with aqueous
NaHCO3 and then extracted with ether to produce 2.5 g of
1
pure E-ester. H NMR: 7.64 (s, 1H), 7.38 (d, J ) 8.7 Hz, 2H),
6.92 (d, J ) 8.7 Hz, 2H), 3.84 (s, 3H), 3.81 (s, 3H), 2.13 (s,
3H).
(E)-2-Met h yl-3-(p -m et h oxyp h en yl)p r op -2-en -1-ol-1,1-
d 2. The R,â-unsaturated ester (2.4 g) was reduced by reacting
with a slurry of LiAlD4 (0.38 g, 9 mmol) and AlCl3 (0.40 g, 3
mmol) in 30 mL of dry ether. The allylic alcohol was isolated
1
in 85% yield. H NMR: 7.25 (d, J ) 8.6 Hz, 2H), 6.88 (d, J )
8.6 Hz, 2H), 6.45 (s, 1H), 3.81 (s, 3H), 1.91 (s, 3H), 1.59 (br. s,
1H).
(E)-2-Meth yl-3-(p-m eth oxyph en yl)pr op-2-en -1-ch lor ide-
1,1-d 2. The allylic alcohol was transformed to the correspond-
ing chloride according to a literature procedure.22 The alcohol
(1.12 g) and 0.90 mL of 2,6-dimethylpyridine were placed in a
dry flask under nitrogen. Subsequently, 0.29 g of anhydrous
LiCl was added and 5 mL dry DMF, enough to dissolve the
reactants, followed by 0.6 mL of CH3SO2Cl at 0 °C. Stirring
was continued for 12 h, and then 100 mL of water and 50 mL
of ether were added. The ether layer was washed with a
saturated solution of Cu(NO3)2 to remove the pyridine. The
crude chloride was a mixture of E/Z ) 96/4. It was used directly
in the next step without purification to avoid decomposition.
1H NMR: 7.24 (d, J ) 8.7 Hz, 2H), 6.87 (d, J ) 8.7 Hz, 2H),
6.53 (s, 1H), 3.82 (s, 3H), 1.99 (s, 3H).
1
chromatography using 1/1 hexane/ethyl acetate as eluent. H
NMR: 8.70 (br. s, 1H), 7.18 (d, J ) 8.3 Hz, 2H), 6.87 (d, 2H,
J ) 8.3 Hz), 4.51 (s, 1H), 3.81 (s, 3H), 3.20 (s, 3H), 3.03 (s,
3H), 1.47 (s, 3H), 1.38 (s, 3H). A more than 10 times diluted
sample in CDCl3 exhibits very different absorptions for the
diastereotopic methyls and the tertiary benzylic hydrogen,
which can be attributed to the different degree of inter- or
intramolecular hydrogen bonding. 1H NMR spectrum of the
diluted sample: 7.65 (s, 1H), 7.17 (d, J ) 8.3 Hz, 2H), 6.91 (d,
2H, J ) 8.3 Hz), 4.36 (s, 1H), 3.84 (s, 3H), 3.23 (s, 3H), 3.06 (s,
3H), 1.47 (s, 3H), 1.45 (s, 3H). 13C NMR of the condensed
sample: 159.57, 154.38, 152.98, 129.38, 128.38, 113.51, 87.51,
64.43, 56.99, 55.15, 24.78, 22.94, 21.40. HR-MS (MALDI-
FTMS): (M + Na)+, calculated for C15H21O4N3 330.1424;
experimental 330.1433.
(E)-1-(p-Meth oxyph en yl)-2-m eth ylpr op-1-en e-3,3,3-d3 (3).
In a dry flask under inert atmosphere were placed 0.3 g of
LiAlD4 and 15 mL of dry THF.23 The crude allylic chloride was
added,and the reaction mixture was stirred for 1 day. It was
then quenched with water, and the aqueous layer was ex-
tracted with ether. Purification was accomplished by flash
column chromatography (petroleum ether as eluent). The
1
alkene was a mixture, E/Z ) 96/4. H NMR of the E-isomer:
7.18 (d, J ) 8.7 Hz, 2H), 6.88 (d, J ) 8.7 Hz, 2H), 6.23 (s, 1H),
3.83 (s, 3H), 1.87 (s, 3H). MS, m/z 165 (M+, 100).
(E)-2-Met h yl-3-(p -m et h oxyp h en yl)p r op -2-en oic Acid .
This compound was prepared according to a literature proce-
dure.24 In a two-necked dry flask were placed 4.85 g of NaH
(60% in oil) and 50 mL of dry DME followed by the slow
addition of diethyl phosphite (5.1 mL) at 0 °C. After 30 min,
3.5 mL of 2-bromopropionic acid was slowly added at 0 °C,
and stirring was continued until hydrogen evolution ceased.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
compounds 1 and 3 and for the adducts in their reactions with
MTAD. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O0005518