On the [2+2] cycloaddition of 2-aminothiazoles and dimethyl acetylenedicarboxylate. Experimental and computational evidence of a thermal disrotatory ring opening of fused cyclobutenes

Article abstract of DOI:10.1021/jo060664c

The reaction of 2-(phenylamino)- and 2-(dimethylamino)thiazoles with dimethyl acetylenedicarboxylate led unexpectedly to dimethyl 6-(phenylamino)- and 6-(dimethylamino)-3,4-pyridinedicarboxylates. Those compounds reasonably result from a sequence of reactions initiated by a [2 + 2] cycloaddition of the alkyne to the formal C=C of the thiazole ring. These pyridines were obtained in nearly all the cases assayed as the exclusive reaction products under rather mild conditions and in fair to good yields. In contrast, the regioisomeric 2-amino-3,4-pyridinedicarboxylates, which would result from a [4 + 2] cycloaddition followed by sulfur extrusion, were only obtained in one particular case. The two reaction paths leading alternatively to both regioisomers were investigated computationally. The respective [2 + 2] and [4 + 2] cycloadducts were found to be formed stepwise from a common dipolar intermediate. Notably, the step following the [2 + 2] cycloaddition (i.e., the ring opening of the fused cyclobutene intermediate to give an all-cis 1,3-thiazepine) was found to take place in a disrotatory mode. Although geometric constraints and electronic factors may reduce the energy for the disrotation, the implication of the fused five-membered ring in the electronic reorganization leading to the 1,3-thiazepine is determinant. In this sense, this step could be regarded also as a thermally allowed six-electron five-center disrotatory electrocyclic ring opening. The proposed mechanism was experimentally supported by the isolation of several intermediates and other experimental facts.

Full text of DOI:10.1021/jo060664c

On the [2+2] Cycloaddition of 2-Aminothiazoles and Dimethyl
Acetylenedicarboxylate. Experimental and Computational Evidence
of a Thermal Disrotatory Ring Opening of Fused Cyclobutenes
Mateo Alajar´ın,*,† Jose´ Cabrera,† Aurelia Pastor,† Pilar Sa´nchez-Andrada,† and
Delia Bautista‡
Departamento de Qu´ımica Orga´nica, Facultad de Qu´ımica, UniVersidad de Murcia, Campus de
Espinardo, 30100 Murcia, Spain, and SerVicio de Instrumentacio´n Cient´ıfica, UniVersidad de Murcia,
Campus de Espinardo, 30100 Murcia, Spain
alajarin@um.es
ReceiVed March 28, 2006
The reaction of 2-(phenylamino)- and 2-(dimethylamino)thiazoles with dimethyl acetylenedicarboxylate
led unexpectedly to dimethyl 6-(phenylamino)- and 6-(dimethylamino)-3,4-pyridinedicarboxylates. Those
compounds reasonably result from a sequence of reactions initiated by a [2 + 2] cycloaddition of the
alkyne to the formal CdC of the thiazole ring. These pyridines were obtained in nearly all the cases
assayed as the exclusiVe reaction products under rather mild conditions and in fair to good yields. In
contrast, the regioisomeric 2-amino-3,4-pyridinedicarboxylates, which would result from a [4 + 2]
cycloaddition followed by sulfur extrusion, were only obtained in one particular case. The two reaction
paths leading alternatively to both regioisomers were investigated computationally. The respective [2 +
2] and [4 + 2] cycloadducts were found to be formed stepwise from a common dipolar intermediate.
Notably, the step following the [2 + 2] cycloaddition (i.e., the ring opening of the fused cyclobutene
intermediate to give an all-cis 1,3-thiazepine) was found to take place in a disrotatory mode. Although
geometric constraints and electronic factors may reduce the energy for the disrotation, the implication of
the fused five-membered ring in the electronic reorganization leading to the 1,3-thiazepine is determinant.
In this sense, this step could be regarded also as a thermally allowed six-electron five-center disrotatory
electrocyclic ring opening. The proposed mechanism was experimentally supported by the isolation of
several intermediates and other experimental facts.
Introduction
hetero Diels-Alder reaction of 4-methyl-5-ethoxythiazoles with
dimethyl fumarate at 200 °C to give pyridines after sulfur
extrusion. Later on, Weiss and co-workers reported that thiazole
rings reacted intramolecularly with acetylenes, leading to fused
thiophenes by an intramolecular Diels-Alder followed by
elimination of 1 mol of nitrile.3 Recently, Wong and Ye
described the intermolecular version of this process reacting
4-methyl and 4-phenylthiazole with bis(trimethylsilyl)acetylene
and diphenylacetylene.4
There is considerable interest for both synthetic and mecha-
nistic reasons in cycloadditions where five-membered ring
heterocycles participate, but only a few examples dealing with
thiazoles as starting materials have been reported. This is
probably due to the low reactivity of these heterocycles caused
by their considerable aromatic stabilization.1 The first cycload-
dition involving thiazoles was reported in 19652 describing the
The reaction of thiazoles with dimethyl acetylenedicarboxy-
late (DMAD) deserves special interest. The controversial nature
† Departamento de Qu´ımica Orga´nica.
‡ Servicio de Instrumentacio´n Cient´ıfica.
(1) Thiazole and its DeriVatiVes; Metzger, J. V., Ed.; Wiley & Sons:
(3) (a) Jacobi, P. A.; Weiss, K. T.; Egberson, M. Heterocycles 1984,
22, 281-286. (b) Jacobi, P. A.; Egbertson, M.; Frechette, R. F.; Miso, C.
K.; Weiss, K. T. Tetrahedron 1988, 44, 3327-3338.
New York, 1979.
(2) Takeda Chem. Ind., Ltd. Fr. Patent 1,400,843, 1965; Chem. Abstr.
1965, 63, 9922.
(4) Ye, X.-S.; Wong, H. N. C. J. Org. Chem. 1997, 62, 1940-1954.
10.1021/jo060664c CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/16/2006
5328
J. Org. Chem. 2006, 71, 5328-5339

[2 + 2] Cycloaddition of 2-Aminothiazoles and DMAD
SCHEME 1. Previously Reported Reactions between
Thiazoles and DMAD
CHART 1. General Structure of 2-Aminothiazoles 5
TABLE 1. 2-Amino-, 2-(Phenylamino)-, and
2-(Dimethylamino)thiazoles 5a-i
R1
R2
R3
R4
thiazole
H
H
H
Ph
(E)-4-MeC6H4CHdCH
(E)-4-MeC6H4CHdCH
(E)-4-MeC6H4CHdCH
Ph
4-ClC6H4
4-BrC6H4
4-MeC6H4
4-MeOC6H4
Ph
H
H
H
H
H
H
H
H
Me
5a
5b
5c
5d
5e
5f
5g
5h
5i
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
TABLE 2. Reaction Conditions and Yields of Pyridines 6 and 7
Obtained from 2-Aminothiazoles 5a-i and DMADa
entry 2-aminothiazole T (°C)
t (h)
product 6 (%)
7 (%)
of the reaction products between thiazole, 2-, 4-, or 5-meth-
ylthiazole, and 2,5-dimethylthiazole and DMAD5 was finally
established by Acheson and co-workers, who assigned their
structures, on the basis of NMR spectroscopy and X-ray
analysis, to the tetramethyl pyrido[2,1-b]thiazole-6,7,8,8a-tet-
racarboxylates 1 (Scheme 1).6 These species are not the expected
hetero Diels-Alder adducts but the result of two successive
additions of two molecules of DMAD across the CdN bond of
the heterocycle, followed by cyclization and further rearrange-
ment of the resulting product.6 Surprisingly, electron-donating
substituents at the heterocycle change completely the course of
the reaction as it was later described. Thus, the reaction of the
thiazoles 2 with DMAD gives the pyridines 3 through an hetero
Diels-Alder reaction and further sulfur extrusion,7 although in
this case extremely harsh conditions were required (Scheme 1).
The activating effect of the amino functionality in the reaction
of 2-(dimethylamino)thiazole toward electron-poor reagents such
as diethyl azodicarboxylate, tosyl isocyanate, or ketenes has been
also demonstrated.8 Nevertheless, no cycloadducts were formed
in these reactions, but only Michael-type products resulting from
the functionalization at the 5-position of the thiazole ring. This
activating effect was evident when 2-amino-4-methylthiazole
was allowed to react with DMAD. From this reaction, conducted
at room temperature, a new adduct was isolated in 42% yield
whose structure was assigned to the 7-thia-2-azabicyclo[2.2.1]-
hepta-2,5-diene 4 (Scheme 1).9 The formation of 4 was
explained as the result of a [4 + 2] cycloaddition in which the
heterocycle acts as heterodiene.
1
2
3
4
5
6
7
8
9
5a
5b
5c
5d
5e
5f
5g
5h
5i
25
25
25
25
25
25
25
25
24
12
12
12
a
b
c
d
e
f
g
h
i
0b
36
70
74
86 (60)
80
85
91
10
0b
0
0
0
6 d (72c)
6 d
0 (20)
0
0
0
12
12
reflux 12
15
a The reactions were run with an excess (3 equiv) of DMAD. b A complex
mixture of products was obtained. c DMAD (6 equiv) was used.
Herein, we disclose our findings in the reaction of 2-ami-
nothiazoles 5 with DMAD. The main reaction products,
6-(phenylamino)- and 6-(dimethylamino)pyridines, are formed
under rather soft conditions and in good yields. Computational
data and the isolation of some intermediates support that the
formation of these final products is the result of an unexpected
[2 + 2] cycloaddition of the DMAD to the heterocycle across
the formal CdC of the thiazole ring in the first step. Subsequent
ring opening, 6π-electrocyclization, and sulfur extrusion would
lead to the final products. Computational and experimental data
also demonstrate that the ring-opening of the fused cyclobutene
intermediate takes place in a disrotatory mode. Chart 1 shows
the general structure of 2-aminothiazoles 5.
Results and Discussion
Experimental Results. The 2-amino-, 2-(phenylamino)-, and
2-(dimethylamino)thiazoles 5a-i (Table 1) were obtained from
the corresponding R-haloketones and thioureas (Hantzsch
synthesis) in good yields (see Supporting Information). The
thiazoles 5a-c were allowed to react with DMAD (3 equiv) in
acetonitrile at room temperature (eq 1, Table 2, entries 1-3).
Under these conditions, 5b,c gave the pyridines 6b,c as the only
reaction products. Pyridines 6 are regioisomeric of 7, the initially
expected reaction products resulting from a [4 + 2] cycload-
dition of the DMAD to the thiazole ring acting as heterodiene,
and further sulfur extrusion. The best result coming from the
4-alkenyl-substituted thiazoles 5a-c was obtained by using the
2-(dimethylamino)thiazole 5c (entry 3). In contrast, the reaction
(5) (a) Reid, D. H.; Skelton, F. S. Tetrahedron Lett. 1964, 1797-1802.
(b) Acheson, R. M.; Foxton, M. W.; Miller, G. R. J. Chem. Soc. 1965,
3200-3206.
(6) (a) Abbott, P. J.; Acheson, R. M.; Eisner, U.; Watkin, D. J.;
Carruthers, J. R. J. Chem. Soc., Chem. Commun. 1975, 155-156. (b) Abbott,
P. J.; Acheson, R. M.; Eisner, U.; Watkin, D. J.; Carruthers, J. R. J. Chem.
Soc., Perkin Trans. 1 1976, 1269-1278.
(7) Kopka, I. E. Tetrahedron Lett. 1988, 29, 3765-3768.
(8) (a) Dondoni, A.; Medici, A.; Venturoli, C. J. Org. Chem. 1980, 45,
621-626. (b) Medici, A.; Pedrini, P.; Venturoli, C.; Dondoni, A. J. Org.
Chem. 1981, 46, 2790-2793.
(9) The adduct 4 was isolated along with other secondary products from
the reaction of the amino functionality with the acetylenic ester. See: Crank,
G.; Khan, H. R. J. Heterocycl. Chem. 1985, 22, 1281-1284.
J. Org. Chem, Vol. 71, No. 14, 2006 5329

Alajar´ın et al.
between the thiazole 5a, with a free amino group at the
2-position, and DMAD led to a complex mixture of products
(entry 1).
The formation of the pyridines 6 as the exclusive reaction
products is rather general. Thus, the reaction of the 2-(di-
methylamino)-4-arylthiazoles 5d-h with DMAD (3 equiv)
under the same reaction conditions (acetonitrile, 25 °C) led to
6 as single products in fair to good yields (entries 4-8). With
5e, an increase of the number of equivalents of DMAD (from
3 to 6) and decrease in the reaction time (from 6 d to 72 h) led
to a mixture of the regioisomeric pyridines 6e and 7e in a 3:1
ratio (entry 5, parenthesis). Notably, aromatic groups bearing
electron-donating substituents at the 4-position of the thiazole
ring accelerate the reactions (entries 5 and 6 versus entries 7
and 8). The reaction of the 2-(dimethylamino)-5-methyl-4-
phenylthiazole (5i) gave a mixture of the regioisomeric 6i and
7i both in low yields (entry 9). In this last case, the reaction
temperature needed to be increased from 25 to 83 °C (refluxing
acetonitrile).
1
FIGURE 1. Contacts from the H,1H-NOESY spectra of 6c, 6i, 7e,
and 7i.
The preparation of the dimethyl 6-(dimethylamino)-2-phenyl-
(methyl)-3,4-pyridinedicarboxylates 6d and 6j (Scheme 2) has
been previously reported in the literature, although they were
obtained in low yields or as mixture of products. Thus, treatment
of the 2-azavinamidinium perchlorate 8 with NaH led to the
2-azabutadiene 9, which was reacted in situ with DMAD to
give 6d in 28% yield (Scheme 2).10 Pyridine 6j was prepared
from the 2-azadiene 10 and DMAD, and it was isolated along
with the pyridine 6k.11 The simultaneous formation of pyridines
6j (32%) and 6k (40%) was rationalized in terms of a [1,3]-
shift of the Me2N group in the initial [4 + 2] adduct 11, followed
by elimination of Me2NH or MeSH from the intermediate 12.
We attempted the synthesis of the pyridine 6j by using the
methodology described in this article, reacting the 2-(dimethyl-
amino)-4-methylthiazole 5j with DMAD (Scheme 3). However,
under our standard reaction conditions (acetonitrile, 25 °C, 12
h, excess of DMAD), that reaction led to a complex mixture of
products. When the reaction was run with only 1 equiv of
DMAD, the 2-thia-4-azabicyclo[3.2.0]hepta-3,6-diene 13j could
be isolated in 52% yield. Under these reactions conditions, the
pyridine 6j was also formed in small amounts (10%). The
structural characterization of 13j was accomplished on the basis
of its spectroscopic data (1H- and 13C NMR, and IR), mass
spectrum, and elemental analysis. Its IR spectrum presents a
band at 727 cm-1 typical from a cis-double bond in a strained
ring.12 Notably, the 1H NMR spectrum of 13j exhibits a signal
characteristic of a bridgehead proton at δ 4.33 ppm. The
stereochemistry of 13j must be cis since a trans fusion of a
cyclobutene ring to a five- or six-membered ring can be ruled
The structural characterization of the regioisomeric pyridines
6 and 7 was done on the basis of their spectroscopic data (1H-
and 13C NMR, and IR), mass spectra, and elemental analyses.
The location of all substituents in 6 and 7 (except 6i and 7i,
see below) was determined on the basis of the 1H,1H-NOESY
spectra of 6c and 7e. The other tetrasubstituted pyridines 6b
1
and 6d-h were identified by comparison of their H- and 13C
NMR data with that of 6c. In these terms, the proton located at
the 5-position in the pyridine ring was the most informative
signal, appearing at a typical chemical shift (6.70-6.84 ppm).
This proton resonates at higher frequency (7.34 ppm) in the
pyridine 7e. In addition, the structure of 6c was unambiguously
determined by single-crystal X-ray diffraction analysis (see
Supporting Information). The most informative NOE effects in
compounds 6c and 7e are depicted in Figure 1. The main feature
of the NOESY spectrum of 6c was the strong NOE observed
between the proton directly attached to the pyridine ring and
those of the dimethylamino group, which was absent in the
spectrum of 7e. On the contrary, the NOESY spectrum of 7e
showed a cross-peak relating the proton located at the pyridine
nucleus and the ortho protons of the aromatic substituent at the
6-position.
The differentiation between the regioisomeric pyridines 6i
1
and 7i was done again on the basis of their H,1H-NOESY
spectra. Analogously to 6c, the spectrum of 6i showed cross-
peaks relating the methyl group at the 5-position and the protons
of the dimethylamino substituent at the 6-position in the pyridine
ring. On the contrary, for 7i a strong NOE effect was observed
between the methyl group located at the 5-position in the
pyridine ring and the ortho protons of the phenyl group (Figure
1).
(10) Gompper, R.; Heinemann, U. Angew. Chem., Int. Ed. Engl. 1980,
19, 217.
(11) Morel, G.; Marchand, E.; Prade`re, J.-P.; Toupet, L.; Sinbandhit, S.
Tetrahedron 1996, 52, 10095-10112.
(12) Paquette, L. A.; Barret, J. H. J. Am. Chem. Soc. 1966, 88, 1718-
1722.
5330 J. Org. Chem., Vol. 71, No. 14, 2006

[2 + 2] Cycloaddition of 2-Aminothiazoles and DMAD
SCHEME 2. Previously Reported Methodologies for the Synthesis of 6d and 6j
SCHEME 3. Sequential Conversion of Thiazole 5j into
Bicycle 13j and Pyridine 6j
dents related with this transformation, we propose the mecha-
nism depicted in Scheme 4 for explaining the formation of the
pyridines 6 and 7 in the reaction of 2-aminothiazoles 5 with
DMAD. A [2 + 2] cycloaddition involving the CdC bond of
the thiazole ring and the CtC bond of the DMAD should lead
to the cycloadduct 13. Since the [2s + 2s] cycloaddition is not
a thermally allowed process according to the Woodward-
Hoffmann rules and the allowed [2s + 2a] requires a geo-
metrically distorted transition state that should be unfavorable
except under unusual circumstances,14 many chemists have
reasoned that [2 + 2] cycloadditions may occur in a stepwise
manner involving diradical or dipolar intermediates.15 There is
experimental evidence of thermal [2 + 2] cycloadditions
proceeding via two-step mechanisms through zwitterionic
intermediates. Among these processes are the reactions of
tetracyanoethylene with electron-rich alkenes such as enol
ethers,16 tetralkoxyethylenes,17 and thioenolethers.18 Dipolar
intermediates have been also invoked as intermediates for the
[2 + 2] cycloaddition reactions of electron-deficient acetylenes
and alkenes.19 Reinhoudt and others demonstrated that 3-amino-
4,5-dihydrothiophenes,20 3-aminothiophenes,21 and related het-
erocycles12,22 react with DMAD similarly to normal enamines23
giving the corresponding [2 + 2] cycloadducts which usually
out on the basis of the steric strain involved.13 This stereo-
chemical assignment was further confirmed by an 1H,1H-
NOESY spectrum (see Supporting Information). Thus, a contact
was observed relating both substituents, the proton and the
methyl group, located at the bridgehead carbon atoms. The
bicycle 13j cleanly transformed into the pyridine 6j in 90% yield
when a solution in acetonitrile was heated at reflux temperature
(Scheme 3).
The conversion of 13j into 6j in toluene-d8, CDCl3, and CD3-
CN was followed by 1H NMR spectroscopy, and their respective
rates were compared. As shown in Figure 2, the reaction was
faster in the more polar CD3CN. On the contrary, there was no
significant difference by using CDCl3 or toluene-d8. This
reaction was also conducted in the presence of 1 equiv of CF3-
COOH by using CDCl3 as solvent. Under these conditions, no
conversion of 13j was observed (see below).
(13) Reinhoudt, D. N.; Verboom, W.; Visser, G. W.; Trompenaars, W.
P.; Harkema, S.; van Hummel, G. J. J. Am. Chem. Soc. 1984, 106, 1341-
1350.
(14) Houk, K. N.; Li, Y.; Evanseck, J. D. Angew. Chem., Int. Ed. Engl.
1992, 31, 682-708.
(15) Epiotis, N. D.; Yates, R. L.; Carlberg, D.; Bernardi, F. J. Am. Chem.
Soc. 1976, 98, 453-461.
(16) Huisgen, R. Acc. Chem. Res. 1977, 10, 117-124.
(17) Hoffmann, R. W.; Bressel, U.; Gehlhaus, J.; Ha¨user, H. Chem. Ber.
1971, 104, 873-885.
With all these experimental results in our hands and on the
basis of the chemical knowledge and the bibliographic prece-
(18) (a) Graf, H.; Huisgen, R. J. Org. Chem. 1979, 44, 2594-2595. (b)
Huisgen, R.; Graf, H. J. Org. Chem. 1979, 44, 2595-2596.
(19) (a) Huebner, C. F.; Dorfman, L.; Robison, M. M.; Donoghue, E.;
Pierson, W. G.; Strachan, P. J. Org. Chem. 1963, 28, 3134-3140. (b) Sasaki,
T.; Kanematsu, K.; Uchide, M. Tetrahedron Lett. 1971, 4855-4858. (c)
Snider, B. B.; Rodini, D. J.; Conn, R. S. E.; Sealfon, S. J. Am. Chem. Soc.
1979, 101, 5283-5293.
(20) Reinhoudt, D. N.; Leliveld, C. G. Tetrahedron Lett. 1972, 3119-
3120.
(21) (a) Reinhoudt, D. N.; Kouwenhoven, C. G. J. Chem. Soc., Chem.
Commun. 1972, 1233-1234. (b) Reinhoudt, D. N.; Kouwenhoven, C. G.
Tetrahedron 1974, 30, 2093-2098. (c) Reinhoudt, D. N.; Trompenaars,
W. P.; Geevers, J. Tetrahedron Lett. 1976, 4777-4780. (d) Verboom, W.;
Visser, G. W.; Trompenaars, W. P.; Reinhoudt, D. N. Tetrahedron 1981,
37, 3525-3533. (e) Reinhoudt, D. N.; Geevers, J.; Trompenaars, W. P. J.
Org. Chem. 1981, 46, 424-434.
(22) (a) Paquette, L. A.; Begland, R. W. J. Am. Chem. Soc. 1966, 88,
4685-4692. (b) Reinhoudt, D. N.; Kouwenhoven, C. G. Tetrahedron Lett.
1972, 5203-5204. (c) Reinhoudt, D. N.; Kouwenhoven, C. G. J. Chem.
Soc., Chem. Commun. 1972, 1232-1233. (d) Temciuc, E.; Ho¨rnfeldt, A.-
B.; Gronowitz, S.; Stålhandske, C. Tetrahedron 1995, 51, 13185-13196.
FIGURE 2. Plots representing the percentage of 13j in the reaction
mixture (13j + 6j) versus the reaction time (days) in CD3CN (red, 9),
CDCl3 (orange, [), and toluene-d8 (4).
J. Org. Chem, Vol. 71, No. 14, 2006 5331

Alajar´ın et al.
SCHEME 4. Proposed Mechanism for Rationalizing the Formation of Pyridines 6 and 7
evolve forming more stable species. Contrary to expectation,
these processes take place with high levels of stereoselectivity
which have been rationalized on the basis of the formation of
1,4-dipolar intermediates showing restricted rotation of the two
poles caused by electrostatic interaction (tied ion pair).21d,24
In our case, the 5-position of the thiazole ring in 5 is activated
by the presence of the amino group at carbon 2. This fact has
precedents in the participation of 2-(dimethylamino)thiazoles
in nucleophilic substitution reactions8 and was, in our hands,
evidenced by the apparition at low frequency of the proton at
this position in the NMR spectra of thiazoles 5a-h and 5j.25
We propose that the nucleophilic attack of the thiazole, at its
5-position, to DMAD would give the zwitterionic intermediate
14 whose cyclization would lead to the bicycle 13 (Scheme 4).
A methyl substituent at the reactive position (5) as in 5i (R4 )
Me) would make the approach of the electrophile more difficult
due to steric reasons, and, consequently, higher temperature
would be required as it was observed (see above). Independent
of the mechanism, concerted or stepwise, the cis-stereochemistry
observed for the cycloadduct 13j is imposed by steric constraints
as mentioned before.13
was unambiguously determined by single-crystal X-ray diffrac-
tion analysis (see Supporting Information). The isolation of (E)-
15 and (Z)-15 was taken as proof of the participation of the
dipolar intermediate 14 in the formation of the bicycle 13j and
thus of the stepwise nature of the mechanism in discussion also
supported by computational data (see below).
Turning back to the proposed mechanistic pathway described
in Scheme 4, the electrocyclic ring opening of the cyclobutene
ring in 13, or in other words, the expansion of the bicyclic
skeleton, would lead to the seven-membered ring, the 1,3-
thiazepine 17. Woodward and Hoffmann28 and Longuet-Higgins
and Abrahamson29 discussed the two possible modes of
concerted thermal ring opening of cyclobutene systems, namely
conrotary and disrotatory. In principle, the symmetry-allowed
conrotatory opening might lead to the formation of an energeti-
cally unfavorable seven-membered ring with a trans double
bond. However, it has been proposed that in compounds in
which the cyclobutene ring is cis-annulated to another one
possessing less than eight atoms, a conrotatory process is
unlikely, and the disrotatory opening is observed.22a,30 Through
this mechanism, the more stable all-cis 1,3-thiazepine 17 would
be formed.
We sought evidence for the implication of the zwitterionic
intermediate 14 by trapping experiments.26 Thus, the thiazole
5g was reacted with 1 equiv of DMAD by using methanol as
solvent. Under these conditions, a mixture of the diastereomeric
thiazoles (E)- and (Z)-15 in a 24:76 ratio was formed (90%
overall yield) (eq 2).27 The structure of the diastereomer (Z)-15
(23) Brannock, K. C.; Burpitt, R. D.; Goodlett, V. W.; Thweatt, J. G. J.
Org. Chem. 1963, 28, 1464-1468.
(24) Epiotis, N. D.; Shaik, S. J. Am. Chem. Soc. 1978, 100, 9-17.
(25) These protons appear in the range of 5.95-6.68 ppm and, thus, at
lower frequencies compared to the chemical shift of this proton in the
thiazole ring (7.41 ppm).
(26) The participation of dipolar intermediates in some of these reactions
is supported by the high yields of fumarates obtained when the reaction is
conducted in methanol, where a proton is available to quench the zwitterion.
See: Acheson, R. M.; Brisdon, J. N.; Cameron, T. S. J. Chem. Soc., Perkin
Trans. 1 1972, 968-975.
The 1,3-thiazepines are little known compounds. Most of their
bibliographic antecedents deal with the relevant biological
5332 J. Org. Chem., Vol. 71, No. 14, 2006

[2 + 2] Cycloaddition of 2-Aminothiazoles and DMAD
activities of their polyhydroderivatives.31 Although little is
known about the chemical behavior of 1,3-thiazepines,32 it
should be analogous to that of its isomeric 1,4-thiazepines, which
show a pronounced instability caused by their ability to undergo
sulfur extrusion.33 Accordingly, we propose that the 1,3-
thiazepine 17 would evolve rapidly through a symmetry-allowed
disrotatory 6π-electrocyclic ring closure to give its valence
isomer, the 7-thia-2-azanorcaradiene 18. Finally, desulfurization
of 18 would give rise to the pyridine 6.
SCHEME 5. Reaction between Acetylenedicarboxylic Acid
(20) and 2-Aminothiazole (21) Leading to
6-Amino-3,4-pyridinedicarboxylic Acid (24) and
2-Amino-3,4-pyridinedicarboxylic Acid (25)
Alternatively, the zwitterionic intermediate 14 could lead to
the 7-thia-2-azabicyclo[2.2.1]hepta-2,5-diene 19, the formal [4
+ 2] cycloadduct, through the attack of its carbanionic center
to the 2-position of the heterocycle (Scheme 4). Further sulfur
extrusion in 19 would give the regioisomeric pyridine 7. This
reaction path proved to be operative, although in low extension,
only in two cases (entries 5 and 9 in Table 2).
Computational Study
As mentioned earlier, [2 + 2] cycloadditions between
electron-deficient acetylenes and alkenes may occur in a
stepwise manner since the [2s + 2s] cycloaddition is not a
thermally allowed process according to the Woodward-
Hoffmann rules, and the allowed [2s + 2a] requires an
unfavorable geometrically distorted transition state.14,15 Conse-
quently, zwitterions have been proposed as intermediates in these
types of reactions.19-23 The mechanism of the [2 + 2]
cycloaddition of electron-deficient acetylenes and π-excedent
five-membered heterocycles has not yet been scrutinized com-
putationally. In contrast, the mechanisms of the Diels-Alder
reaction between DMAD and 1-methylpyrrole34 and acetylene-
dicarboxylic acid and 2-methylfuran35 have been approached
by ab initio and DFT calculations showing that the reactions
take place in a stepwise manner with the initial formation of
zwitterionic intermediates. PM3 semiempirical methods have
also showed that the reaction of 6-aminopyrimidin-4(3H)-ones
with DMAD lead to the corresponding [4 + 2] cycloadducts
on a concerted mode.36
For simplicity, to model the reaction of DMAD with the
thiazoles 5 presented in the experimental part of this work, we
selected as reactants the more simple structures acetylenedi-
carboxylic acid (20) and 2-aminothiazole (21). We have carried
out an intensive exploration of the potential energy surface
associated with the reaction of 20 with 21 leading to the
thiazanorcaradiene 22 (coming from successive transformations
of the [2 + 2] cycloadduct initially formed) and 7-thia-2-
azabicyclo[2.2.1]hepta-2,5-diene (23) (the corresponding [4 +
2] cycloadduct) (Scheme 5). The sulfur extrusions giving
6-amino-3,4-pyridinedicarboxylic acid (24) and 2-amino-3,4-
pyridinedicarboxylic acid (25) as final products have not been
included in this theoretical approach as it has been well-
documented that the sulfur extrusion in related heterocycles takes
place in several steps without appreciable energetic cost.37
We have found the stepwise mechanisms outlined in Figure
3 for the formation of products 22 and 23. Both the [2 + 2]
(path A, blue color) and the [4 + 2] (path B, red color)
cycloaddition reactions, leading respectively to the cycloadducts
26 and 23, take place in a stepwise manner, involving the initial
formation of a common polar intermediate INT. No transition
states corresponding to concerted [4 + 2] or [2 + 2] cycload-
dition reactions between 20 and 21 could be located.
(27) The reaction pathway to give the diastereomeric compounds 15 may
involve the formation of the bent dipole 14g, which can easily equilibrate
with 14g′′ through the cumulenic 14g′. The (E)- and (Z)-15 adducts should
be formed by migration of the hydrogen of the former C5 of the thiazole
in 14g and 14g′′ or, alternatively, from the allenic structure 16 by a
tautomeric equilibrium. See: (a) Caramella, P.; Houk, K. N. Tetrahedron
Lett. 1981, 22, 819-822. (b) Dickstein, J. I.; Miller, S. I. In The Chemistry
of the Carbon-Carbon Triple Bond; Patai, S., Ed.; Wiley: New York, 1978;
pp 819-826.
(28) Woodward, R. B.; Hoffmann, R. J. Am. Chem. Soc. 1965, 87, 395-
397.
(29) Longuet-Higgins, H. C.; Abrahamson, E. W. J. Am. Chem. Soc.
1965, 87, 2045-2046.
(30) (a) Epiotis, N. D. Angew. Chem., Int. Ed. Engl. 1974, 13, 751-
780. (b) Seebach, D. In Houben-Weyl, Methoden der organischen Chemie;
Mu¨ller, E., Ed.; G. Thieme Verlag: Stuttgart, 1971; Vol. IV/4, pp 418-
419.
Figure 3 displays the qualitative reaction profile at the
B3LYP/6-31+G* theoretical level and the location of the
stationary points for the reaction between acetylenedicarboxylic
(31) See, for example: (a) Di Cesare, M. A.; Campiani, G.; Butini, S.
Patent WO2005097797, 2005; Chem. Abstr. 2005, 143, 405933. (b) Patel,
R. N. Biomol. Eng. 2001, 17, 167-182.
(32) (a) Mente, P. G.; Heine, H. W. J. Org. Chem. 1971, 36, 3076-
3078. (b) Schmidt, R. R.; Berger, G. Synthesis 1974, 187-189.
(33) (a) Yamamoto, K.; Yamazaki, S.; Osedo, H.; Murata, I. Angew.
Chem., Int. Ed. Engl. 1986, 25, 635-637. (b) Masquelin, T.; Obrecht, D.
Tetrahedron 1997, 53, 641-646. (c) Cabarrocas, G.; Ventura, M.; Maestro,
M.; Mah´ıa, J.; Villalgordo, J. M. Tetrahedron: Asymmetry 2001, 12, 1851-
1863.
(34) Domingo, L. R.; Picher, M. T.; Zaragoza, R. J. J. Org. Chem. 1998,
63, 9183-9189.
(35) (a) Domingo, L. R.; Picher, M. T.; Aurell, M. J. J. Phys. Chem. A
1999, 103, 11425-11430. (b) Domingo, L. R.; Picher, M. T.; Andres, J. J.
Org. Chem. 2000, 65, 3473-3477.
(36) Cobo, J.; Melguizo, M.; Nogueras, M.; Sa´nchez, A.; Dobado, J.
A.; Nonella, M. Tetrahedron 1996, 52, 13721-13732.
(37) (a) Gleiter, R.; Krennich, G.; Cremer, D.; Yamamoto, K.; Murata,
I. J. Am. Chem. Soc. 1985, 107, 6874-6879. (b) Miller, K. J.; Moschner,
K. F.; Potts, K. T. J. Am. Chem. Soc. 1983, 105, 1705-1712.
J. Org. Chem, Vol. 71, No. 14, 2006 5333

Alajar´ın et al.
FIGURE 3. Qualitative reaction profiles at the B3LYP/6-31G+* level of reaction between acetylenedicarboxylic acid (20) and 2-aminothiazole
(21) leading to the thiazanorcaradiene 22 and the 7-thia-2-azabicyclo[2.2.1]hepta-2,5-diene 23.
acid and 2-aminothiazole: reactants 20 + 21; products 22 and
23, intermediates INT, 26 and 27, and the transition structures
TS1, TS2, TS3, TS4, and TS5. Figures 4 and 5 show the
geometries of these stationary points, including the most relevant
bond distances, while Table 3 includes the relative energies of
the stationary points at the B3LYP/6-31+G*//B3LYP/6-31+G*
and B3LYP-PCM/6-31+G*//B3LYP/6-31+G* theoretical levels
also with the energy barriers. We will comment only the results
obtained at the B3LYP/6-1+G* theoretical level, unless oth-
erwise stated.
Pathway A: Mechanism of the Formation of the Thia-
zanorcaradiene 22. In this transformation, the first step consists
of the formation of the C5-C11 σ-bond (see Figure 3 for
numeration) through the transition state TS1 leading to the
dipolar intermediate INT, 17.8 kcal‚mol-1 higher in energy than
that of the reactants. The calculated barrier for this step was
18.3 kcal‚mol-1. As mentioned earlier, the presence of the amino
group at the 2-position of the thiazole ring activates the C5
center for acting as a nucleophile attacking the conjugated
acetylenic system of the other reactant. As a consequence of
this attack, partial or total positive and negative charges are
generated at the C4 and C12 carbon atoms, respectively, of both
TS1 and INT. Thus, the natural charge at C4 changes from
0.184 in 21 to 0.346 in TS1 and 0.400 in INT, whereas the
natural charge at C12 changes from -0.025 in 20 to -0.040 in
TS1 and -0.026 in INT (see Supporting Information). The
presence of the carboxyl and the amino groups allows for an
effective delocalization of the negative and positive partial
charges, respectively.
Figure 4). This spatial arrangement, also present in INT, allows
for the effective delocalization of the charge being formed at
C12. In both TS1 and INT the thiazole ring unexpectedly
deviates from planarity entailing an arrangement that places the
plane containing the carboxyl group at C12 (negative charge)
nearly parallel to the plane containing the C4-N3-C2-N6
fragment (charged positively). This probably allows for attractive
electrostatic interactions and also the formation of a weak
hydrogen bond between an oxygen atom of the carboxyl group
and one of the hydrogen atoms at N6. On the other hand, it is
worth pointing out that the configuration of the vinyl anionic
moiety being formed as result of the change of hybridization
(from sp to sp2) at C11 and C12 is E in both TS1 and INT.
Notwithstanding, the values of the C11-C12-C13 bond angles,
159.9° and 153.4° in TS1 and INT, respectively, are higher
than those expected for an sp2-hybridized C12 carbon atom.
This fact indicates that the sp2 lobe at the C12 center, filled by
the electron density transferred from the thiazole to the
acetylenedicarboxylic acid, is delocalized in some extension into
the carbonyl oxygen of the adjacent carboxyl group. Conse-
quently, the C12 carbon atom is cumulenic in a certain degree.
The dipolar intermediate INT can evolve by the two
alternative pathways, A and B (see Figure 3). The formation of
the σ-bond C4-C12 occurs via the transition state TS2, 6.0
kcal‚mol-1 above INT, leading to the [2 + 2] cycloadduct 26.
This step is highly exothermic, by 39.4 kcal‚mol-1. The
animation of the imaginary frequency in TS2 shows that the
nuclear motions are similar to that expected for the inversion
of configuration (E to Z) at the C12 carbon atom, but going
further until the formation of the C4-C12 bond. The cycload-
duct 26 thus formed features a cis fusion at the bridgehead atoms
(see Figure 4).
In TS1, the plane containing the carboxyl group at C11 is in
the plane containing the C14-C11-C12-C13 atoms, whereas
the carboxyl group at C12 is perpendicular to that plane (see
5334 J. Org. Chem., Vol. 71, No. 14, 2006

[2 + 2] Cycloaddition of 2-Aminothiazoles and DMAD
FIGURE 4. B3LYP/6-31+G*-optimized geometries of the stationary points found in the reaction of acetylenedicarboxylic acid (20) with
2-aminothiazole (21) leading to the thiazanorcaradiene 22.
According to the cis-fusion of 26, the opening of the
cyclobutene ring of that compound by an electrocyclic, sym-
metry-allowed conrotatory process should lead to the corre-
sponding cis,trans,cis-2-amino-1,3-thiazepine or cis,cis,trans-
2-amino-1,3-thiazepine, depending on the clockwise or the
counterclockwise conrotation along the C4-C12 and C5-C11
bonds, respectively. However, we have found the transition state
TS3 whose geometry resembles a disrotatory electrocyclic ring
opening of 26. IRC calculations confirmed that TS3 connects
with 26 and the cis,cis,cis-2-amino-1,3-thiazepine 27. The
computed barrier for this transformation is 24.0 kcal‚mol-1, and
mode, demonstrating that the large energy of concert in the
electrocyclic ring opening of cyclobutenes can be eroded by
geometric constraints. On the other hand, for compounds in
which the cyclobutene ring is cis-annulated to another system
that possesses less than eight atoms, it is generally accepted
that the ring opening must occur by way of the symmetry-
forbidden disrotatory mode.22a,30 Along with the geometric
constraints, the electronic characteristic of the substituents can
reduce the activation energy for the disrotation, in a way that
Epiotis qualified as polar disrotatory ring opening.30a Several
examples support that these predictions are correct.22b,23,39 Thus,
Reinhoudt et al.21a,b,e,22c,40 showed that, in cis-fused [2 + 2]
cycloadducts of thiophenes with acetylenes that isomerize at
low temperature to the corresponding cis,cis,cis-thiepines, the
the process is exothermic by 13.8 kcal‚mol-1
.
In the literature, there are several precedents of disrotatory
ring openings of cyclobutenes. Houk and co-workers have
recently confirmed38 Roth’s hypothesis that a cyclobutene forced
to planarity might have a smaller preference for the conrotatory
(39) (a) Chapman, O. L.; Pasto, D. J.; Borden, G. W.; Griswold, A. A.
J. Am. Chem. Soc. 1962, 84, 1220-1224. (b) Breslow, R.; Napierski, J.;
Schmidt, A. H. J. Am. Chem. Soc. 1972, 94, 5906-5907.
(40) Reinhoudt, D. N.; Volger, H. C.; Kouwenhoven, C. G. Tetrahedron
Lett. 1972, 5269-5272. See also ref 22d.
(38) Lee, P. S.; Sakai, S.; Ho¨rsteman, P.; Roth, W. R.; Kallel, E. A.;
Houk, K. N. J. Am. Chem. Soc. 2003, 125, 5839-5848.
J. Org. Chem, Vol. 71, No. 14, 2006 5335

Alajar´ın et al.
SCHEME 6. Ring Opening of Bicycle 26 Leading to
1,3-Thiazepine 27 Showing Some Resonance Forms of the
Latter
We suppose that both the electronic features and the geo-
metrical constraints of the bicycle 26 clearly favor a polar
disrotatory ring opening. Thus, on the basis of the structure of
TS3 it can be seen that not only the cyclobutene ring of 26 but
also the fused five-membered ring are involved in the electronic
reorganization leading to 27. This fact can be inferred by
analyzing the bond distances, the natural charges, as well as
the bond orders on going from 26 to 27 through TS3. In relation
to the changes in natural charges, the most relevant is the
increase of the positive charge at C4 in TS3 (0.364) with regard
to 26 (0.203) and 27 (0.339), and the increase of the negative
charge at C5 (-0.162) with regard to 26 (-0.124) and 27
(-0.059). The positive charge at C4 is delocalized into C2, N6,
and S1, and the negative one into C12, the carbonyl oxygen
atom of the carboxyl group at this atom, and also into the vicinal
S1 atom (see the values of bond orders and natural charges in
the Supporting Information). These features are consistent with
the fact that the breaking of the C4-C5 bond is assisted by the
lone pair at the exocyclic nitrogen atom, by the intracyclic sulfur
atom, and by the carboxyl group at C12. Thus, the isomerization
of 26 to its valence tautomer 27 could be regarded also as a
six-electron five-center disrotatory electrocyclic ring opening,
thermally allowed, with simultaneous reorganization of the
resultant cyclic π-system.41 The participation of the thiazole ring
in this reorganization can be easily visualized by considering
the neutral resonance structure of 27 with a λ4 sulfur atom. In
fact, the values of the bond orders and the natural charges point
to the contribution of several polar resonance forms to the
structure of thiazepine 27 (Scheme 6).
FIGURE 5. B3LYP/6-31+G*-optimized geometries of the stationary
points found in the transformation of the dipolar intermediate INT into
23.
TABLE 3. Relative Energies with Zero-Point Vibrational Energy
Corrections (kcal‚mol-1) at the B3LYP/6-31+G*//B3LYP/6-31+G*
and the B3LYP-PCM/6-31+G*//B3LYP/6-31+G* Theoretical Levels,
Low or Imaginary Frequencies (cm-1) for the Stationary Points
Found in the Conversions 20 + 21 f 22 and 20 + 21 f 23
Calculated at the B3LYP/6-31+G* Theoretical Level Also with the
Calculated Energy Barriersa (kcal‚mol-1
)
relative energy
energy barriers
B3LYP-
B3LYP-
PCM
low
frequencies
structure B3LYP
B3LYP
PCM
20 + 21
0.0
0.00
36.0 (20)
249.9 (21)
-331.9
45.8
-103.9
37.3
∆Ea1
18.3
9.9
TS1
INT
TS2
26
TS3
27
TS4
22
TS5
23
18.3
17.8
23.7
9.9
6.9
5.0
∆Ea2
∆Ea3
∆Ea4
∆Erxn22 -20.0
∆Ea5
∆Erxn23
6.0
24.0
20.6
23.6
22.3
-21.6 -30.1
-26.4
These theoretical data are in line with those extracted from
the experimental results. Thus, the conversion of the bicycle
2.4 -6.4
-197.4
0.6
-6.2
3.6
-35.4 -42.5
-14.9 -20.2
-20.0 -26.4
53.3
-10.9
-444.4
30.8
-97.7
32.3
(41) Reinhoudt has also proposed (see ref 21b) that the ring opening of
thiabicyclo[3.2.0]hepta-3,6-dienes leading to the corresponding thiepines
can be regarded as an electrocyclic symmetry-allowed disrotatory reaction
of the dihydrothiophene ring analogous to that occurring in the isoelectronic
bicyclo[4.2.0]octatriene which isomerizes rapidly to cyclooctatetraene.
See: Glass, D. S.; Watthey, J. W. H.; Winstein, S. Tetrahedron Lett. 1965,
377-383.
18.4
-6.2 -10.9
10.5
a See Figure 3 for the notation of the energy barriers.
presence of an electron-donating group at the bridgehead carbon
atom enhanced the rate of isomerization, in the same way that
electron-withdrawing groups at the sp2 carbon atoms of the
cyclobutene fragment did.13
5336 J. Org. Chem., Vol. 71, No. 14, 2006

[2 + 2] Cycloaddition of 2-Aminothiazoles and DMAD
13j into the pyridine 6j was faster in the more polar CD3CN
compared to that in CDCl3 or toluene-d8, perhaps as result of
the highly polar character of the transition state of the rate-
determining step, the formation of the thiazepine ring. Other
experimental observations can be correctly interpreted on the
basis of the data obtained from the theoretical calculations; in
fact, bicyclic 13 could be isolated only when C4 bears a
nonaromatic group such as in 13j (R3 ) Me in Scheme 4). For
the rest of the cases in which an alkenylic or an aromatic group
is attached to this carbon atom (R3 ) CHdCHAr or Ar in
Scheme 4) only the corresponding pyridine 6 was detected in
the reaction mixture. Even more, it is a matter of the fact that
aromatic groups bearing electron-donating substituents at the
4-position in the thiazole ring accelerate the reaction (Table 2;
entries 5 and 6 versus entries 7 and 8). These observations may
be interpreted by considering that the developing positive charge
at C4 in the corresponding transition state TS3 may be stabilized
by resonance by the appropriate substituents, in special electron-
donating ones, leading to a decrease in the activation energy of
this step and thus favoring the evolution to the pyridine 6.
Finally, the conversion of 13j into 6j in the presence of 1 equiv
of CF3COOH did not occur at all (see the previous section).
This experimental fact shows the relevant role that the lone pair
at the exocyclic nitrogen atom N6 plays in assisting the cleavage
of the C4-C5 bond of the bicycles type 26.
As stated earlier, the chemical behavior of 1,3-thiazepines
should be analogous to that of their isomeric 1,4-thiazepines,
which show a pronounced instability caused by their ability to
undergo sulfur extrusion.33 The propensity to sulfur extrusion
has been also well-documented in the related seven-membered
ring thiepines.37 The instability of the latter compounds has been
rationalized as a consequence of its antiaromatic character42 and,
moreover, as due to the low activation energy required for the
desulfurization reaction which converts it into the benzene
nucleus, via the formation of a thianorcaradiene as intermedi-
ate.43
of the thiazanorcaradiene 22 to the desulfurization reaction
leading to the pyridine 24, the formation of 22 from the
thiazepine 27 can be considered as an irreversible process once
the thiazanorcaradiene 22 has converted into the pyridine 24,
despite 22 being more energetic than the thiazepine 27 and even
though the calculated barrier for the reversion of 22 to 27 is
only 5.2 kcal‚mol-1
.
Pathway B: Mechanism of the Formation of the 7-Thia-
2-azabicyclo[2.2.1]hepta-2,5-diene 23. The alternative path for
the evolution of intermediate INT consists of its transformation
into the formal [4 + 2] cycloadduct 23 involving the nucleo-
philic attack of the anionic C12 center on the C2 carbon atom
of the thiazole ring. At the HF/6-31G* theoretical level, we have
found that this process involves the initial inversion of the
configuration at the C12 carbon atom leading to the correspond-
ing intermediate of Z configuration [(Z)-INT], and subsequently,
the formation of the C2-C12 bond takes place through TS5′
(see Supporting Information) leading to 23. However, at the
B3LYP/6-31+G* level we have found only a transition
structure, TS5, that connects the polar intermediate INT with
the bicycle 23, whose structure resembles that expected for the
E to Z configuration inversion at the C12 carbon atom. Thus,
the value of the C11-C12-C13 bond angle is 176.2°,45 and
the C2-C12 bond formation is not much advanced: the bond
distance is 2.663 Å and the bond order is 0.17. The animation
of the imaginary frequency showed that the nuclear motions
correspond to the inversion at C12 and the simultaneous
approaching of the C2 and C12 atoms. IRC calculations
confirmed the connection of TS5 with INT and 23, notwith-
standing that the IRC curve toward the formation of 23 showed
an inflection point in a small and very flat region corresponding
to a structure similar to the transition state TS5′ found at the
HF level. The calculated barrier at the B3LYP/ 6-31+G* level
for the conversion of INT into 23 through TS5 was almost
imperceptible, only 0.6 kcal‚mol-1, the process being exothermic
by 23.9 kcal‚mol-1, whereas the overall transformation (i.e.,
20 + 21 f 23) is exothermic by only 6.2 kcal‚mol-1. It is
expected that the sulfur extrusion in the bicycle 23, leading to
2-amino-3,4-pyridinedicarboxylic acid (25), should occur easily
without appreciable energetic cost.
Analogous to the more stable conformation of the thiepine
ring,37,44 the optimized structure of the thiazepine 27 is boat-
shaped. This conformation seems to be appropriate to allow a
6π-electrocyclic disrotatory ring closure through the transition
structure TS4, leading finally to the thiazanorcaradiene 22. The
calculated energy barrier for this step is 20.6 kcal‚mol-1, and
By considering the calculated reaction profile for the trans-
formations 20 + 21 f 22 and 20 + 21 f 23 (Figure 3) and
the energy barriers associated with each step (Table 3), this
theoretical approach predicts that the intermediate 23 should
be the kinetically controlled product, whereas the thiazanor-
caradiene 22 seems to be the product of thermodynamic control.
Therefore, the nearly exclusive formation of pyridines 6 in the
experiments described earlier can only be rationalized as
resulting from the prevalence of the [2 + 2] cycloaddition
pathway under thermodynamic control. To this end, the rever-
sion of the [4 + 2] cycloadduct 23 into the dipolar intermediate
INT should occur, and the energy barrier of this process is 24.6
the process is endothermic by 15.4 kcal‚mol-1
.
The transformation of the thiazanorcaradiene 22 into the final
6-amino-3,4-pyridinedicarboxylic acid (24), as mentioned ear-
lier, has not been included in this theoretical study as previous
works dealing with the sulfur extrusion in thiepines and other
heterocycles have been already reported. These investigations
revealed that this process involves a sequence of reactions
entailing several intermediates in which the number of sulfur
atoms increases, ending with the formation of the corresponding
aromatic compound and stable forms of sulfur. The activation
energy for the steps leading to the sulfur extrusion is lower than
kcal‚mol-1
.
5 kcal‚mol-1 37
Therefore, if we consider a similar propensity
.
These results correspond to calculations in the gas phase but,
because several stationary points along the reaction coordinates
resulting in 22 and 23 are highly polarized (i.e., TS1, INT, TS3),
it is conceivable that the solvent may alter considerably the
(42) (a) Dewar, M. J. S.; Trinajstiæ, N. J. Am. Chem. Soc. 1970, 92,
1453-1459. (b) Hess, B. A., Jr.; Schaad, L. J. J. Am. Chem. Soc. 1973, 95,
3907-3912. (c) DasGupta, N. K.; Birss, F. W. Tetrahedron 1980, 36, 2711-
2720.
(43) (a) Scho¨nberg, A.; Fayez, M. B. E. J. Org. Chem. 1958, 23, 104-
105. (b) Loudon, J. D. In Organic Sulfur Compounds; Kharash, N., Ed.;
Pergamon Press: New York, 1961, pp 299-305.
(44) Yamamoto, K.; Yamakazi, S.; Kohashi, Y.; Murata, I. Tetrahedron
Lett. 1982, 23, 3195-3198.
(45) This value indicates that the C12 carbon atom presents an sp
hybridization where the charge located at this center occupies a p atomic
orbital. A similar value has been found in the transition state corresponding
to the configuration inversion connecting two polar intermediates coming
from the reaction of DMAD with 1-methylpyrrole (see ref 34).
J. Org. Chem, Vol. 71, No. 14, 2006 5337

Alajar´ın et al.
values of the energy barriers. This is why we studied the
influence of the solvent in the mechanism of these transforma-
tions.
Experimental Section
Synthesis of the Thiazoles 5a-j. Thiazoles 5d,47 5e,48 5f,48 5g,48
5h,47 and 5j47 were known compounds, and they were prepared
following methodologies previously described in the literature. The
general procedure for the synthesis of the thiazoles 5a-c and 5i
and their structural characterization have been included in the
Supporting Information.
General Procedure for the Synthesis of Pyridines 6b-i.
DMAD (0.35 g, 2.45 mmol) was added to a solution of the
corresponding thiazole 5 (0.82 mmol) in acetonitrile (10 mL), and
the reaction mixture was stirred at 25 °C or under reflux (temper-
atures and reaction times depicted in Table 2). The solvent was
evaporated to dryness, and the residue was purified by silica gel
column chromatography.
(1R*,5S*)-Dimethyl 3-(dimethylamino)-5-methyl-2-thia-4-
azabicyclo[3.2.0]hepta-3,6-diene-6,7-dicarboxylate (13j). DMAD
(0.30 g; 0.21 mol) was added to a solution of the thiazole 5j (0.30
g, 0.21 mol) in MeCN (15 mL), and the reaction mixture was stirred
at 25 °C for 6 d. The solvent was evaporated to dryness, and the
residue was purified by silica gel column chromatography eluting
with 1:1 AcOEt/hexane (Rf 0.16); yield 52%; mp 89.2-90.2 °C
(colorless prisms, CHCl3/Et2O); IR (Nujol) 1741, 1718, 1605, 1290,
1122 cm-1; 1H NMR (CDCl3) δ 1.71 (s, 3H), 2.96 (s, 6H), 3.82 (s,
3H), 3.83 (s, 3H), 4.33 (s, 1H); 13C NMR (CDCl3) δ 22.3 (q), 39.6
(2 × q), 52.1 (q), 52.2 (q), 57.1 (d), 86.3 (s), 138.4 (s), 148.2 (s),
160.9 (s), 161.2 (s), 162.7 (s); MS (EI, 70 eV) m/z (relative intensity)
284 (M+, 7), 252 (52), 237 (45), 223 (100), 221 (50). Anal. Calcd
for C12H16N2O4S (284.33): C, 50.69; H, 5.67; N, 9.85; S, 11.28.
Found: C, 50.40; H, 5.89; N, 9.66; S, 10.81.
Solvent Effects. We have carried out single-point calculations
by using the polarized continuum model (PCM) with acetonitrile
as solvent, because this is the one used in the experimental part
of this work. Table 3 reports the B3LYP/6-31+G* relative
energies with inclusion of solvent effects. As we expected, these
data show notorious changes in the energy barriers, and they
are especially significant when comparing the two competitive
paths, A and B, as now the transition state TS2 is less energetic
than the transition state TS5. Actually, with the inclusion of
the solvent, there is no barrier for the transformation of INT
into the [2 + 2] cycloadduct 26,46 whereas the calculated energy
barrier for the conversion of INT into the [4 + 2] cycloadduct
23 is now 3.6 kcal‚mol-1, and therefore the thiazanoracardiene
22 should be both the kinetically and thermodynamically
controlled product, thus explaining the exclusiVe formation of
the reaction products 6.
It is worth pointing out that, although all the stationary points
are stabilized by the inclusion of solvent in the calculations,
there are no notorious changes in the energy barriers of the
remaining steps in the transformation 20 + 21 f 22, except in
that of the first one leading to the dipolar intermediate INT,
whose value decreases to 9.9 kcal‚mol-1
.
Conclusions
Dimethyl 6-(dimethylamino)-2-methyl-3,4-pyridinedicarboxy-
late (6j). A solution of 13j (0.12 g, 0.42 mmol) in MeCN (15 mL)
was stirred under reflux for 2 h. The solvent was evaporated to
dryness, and the residue was purified by silica gel column
chromatography eluting with 1:2 AcOEt/hexane (Rf 0.43); yield
90%; mp 63.8-64.2 °C (lit. 64 °C)11 (colorless prisms, Et2O/
The reaction of the 2-(phenylamino) and 2-(dimethylamino)-
thiazoles 5 with DMAD leads to the unexpected pyridines 6 as
the exclusive reaction products, except under special circum-
stances in which the regioisomeric pyridines 7 are isolated in
small amounts. The two possible competitive pathways for the
reaction between the simplified reagents 2-aminothiazole and
acetylenedicarboxylic acid leading to the regioisomeric 6-amino
and 2-amino-3,4-pyridinedicarboxylic acids (24 and 25) (i.e.,
those resulting from a [2 + 2] or a [4 + 2] cycloaddition,
respectively) have been computationally scrutinized. The analy-
sis of the energy profiles for both pathways at the B3LYP-
PCM level of theory indicates that pyridines 24 are both the
kinetically and thermodynamically controlled products. Both
alternative cycloaddition processes were found to occur stepwise
through a common dipolar intermediate (INT). Notably, the step
following the [2 + 2] cycloaddition (i.e., the ring opening of
the fused cyclobutene intermediate to give the all-cis 1,3-
thiazepine) has been found to take place in a disrotatory mode.
Although geometric constraints and electronic factors may
reduce the energy barrier for the disrotation, the implication of
the fused five-membered ring in the electronic reorganization
leading to the 1,3-thiazepine seems to be determinant in this
sense, since this step could be regarded also as a thermally
allowed six-electron five-center disrotatory electrocyclic ring
opening. The proposed mechanism is experimentally supported
by the isolation of several intermediates and other experimental
facts such as the influence of the polarity of the solvent on the
rate of conversion of the fused cyclobutene 13j into the pyridine
6j and the lack of reaction of 13j in the presence of trifluoro-
acetic acid.
1
hexane); H NMR (CDCl3) δ 2.53 (s, 3H), 3.13 (s, 6H), 3.84 (s,
3H), 3.88 (s, 3H), 6.59 (s, 1H); 13C NMR (CDCl3) δ 23.8 (q), 37.7
(2 × q), 52.1 (q), 52.7 (q), 101.9 (d), 112.9 (s), 140.8 (s), 157.5
(s), 158.6 (s), 167.9 (s), 168.6 (s).
Reaction of the Thiazole 5g and DMAD in Methanol. DMAD
(0.03 g; 0.23 mmol) was added to a solution of the thiazole 5g
(0.05 g, 0.23 mol) in methanol (10 mL), and the reaction mixture
was stirred at 25 °C for 48 h. Then, the solvent was evaporated to
dryness, and the residue was purified by silica gel column
chromatography eluting with 1:3 AcOEt/hexane to give a mixture
of (Z)-15 and (E)-15.
Dimethyl (Z)-[2-(Dimethylamino)-4-(4-methylphenyl)thiazol-
5-yl]-2-butenedioate [(Z)-15]: (Rf 0.18); yield 68%; mp 102.8-
103.7 °C (yellow prisms, CHCl3/n-hexane); IR (Nujol) 1720, 1618,
1
1253, 1216 cm-1; H NMR (CDCl3) δ 2.33 (s, 3H), 3.16 (s, 6H),
3.40 (s, 3H), 3.67 (s, 3H), 6.65 (s, 1H), 7.13 (d, J ) 8.0 Hz, 2H),
7.42 (d, J ) 8.0 Hz, 2H); 13C NMR (CDCl3) δ 21.3 (q), 40.0 (2 ×
q), 51.8 (q), 52.7 (q), 110.2 (s), 126.1 (d), 128.5 (2 × d), 128.9 (2
× d), 132.9 (s), 137.5 (s), 138.0 (s), 155.2 (s), 165.5 (s), 167.2 (s),
170.5 (s); MS (EI, 70 eV) m/z (relative intensity) 360 (M+, 14),
301 (44), 300 (100), 271 (36), 269 (39), 242 (47). Anal. Calcd for
C18H20N2O4S (360.43): C, 59.98; H, 5.59; N, 7.77; S, 8.90.
Found: C, 59.68; H, 5.74; N, 7.74; S, 8.69.
Dimethyl (E)-[2-(Dimethylamino)-4-(4-methylphenyl)thiazol-
5-yl]-2-butenedioate [(E)-15]: (Rf 0.12); yield 22%; mp 84.1-
85.5 °C (yellow prisms, CHCl3/n-hexane); IR (Nujol) 1734, 1709,
1556, 1288, 1157 cm-1; 1H NMR (CDCl3) δ 2.36 (s, 3H), 3.15 (s,
6H), 3.32 (s, 3H), 3.68 (s, 3H), 5.85 (s, 1H), 7.16 (d, J ) 8.0 Hz,
2H), 7.34 (d, J ) 8.0 Hz, 2H); 13C NMR (CDCl3) δ 21.3 (q), 40.0
(46) In fact, INT is slightly higher in energy than TS2, probably as
consequence of using the approximation of frozen geometries. See: (a)
Bonaccorsi, R.; Cammi, R.; Tomasi, J. J. Comput. Chem. 1991, 12, 301-
309. (b) Tun˜o´n, I.; Silla, E.; Tomasi, J. J. Phys. Chem. 1992, 96, 9043-
9048.
(47) Mahon, C. M.; Meakins, G. D. J. Chem. Res., Synop. 1990, 290.
(48) Birkinshaw, T. N.; Meakins, G. D.; Plackett, S. J. J. Chem. Soc.,
Perkin Trans. 1 1988, 2209-2212.
5338 J. Org. Chem., Vol. 71, No. 14, 2006

[2 + 2] Cycloaddition of 2-Aminothiazoles and DMAD
(2 × q), 51.6 (q), 52.1 (q), 114.7 (d), 115.2 (s), 128.5 (2 × d),
129.4 (2 × d), 131.9 (s), 138.5 (s), 141.5 (s), 156.7 (s), 165.5 (s),
166.3 (s), 168.9 (s); MS (EI, 70 eV) m/z (relative intensity) 360
(M+, 9), 301 (40), 300 (100), 271 (33), 269 (30), 242 (38). Anal.
Calcd for C18H20N2O4S (360.43): C, 59.98; H, 5.59; N, 7.77; S,
8.90. Found: C, 59.62; H, 5.88; N, 7.88; S, 8.66.
The solvent effects have been considered by B3LYP/6-31+G*
single-point calculations using the self-consistency reaction field52
method, based on the polarized continuum model53 of Tomasi and
co-workers, in acetonitrile as solvent.
Natural charges and Wiberg bond indices were calculated with
the natural bond orbital (NBO) method.54
Computational Methods. All calculations were carried out with
the Gaussian0349 suite of programs. An intensive characterization
of the potential energy surface was done at the HF/6-31G*50
theoretical level and then with the B3LYP51 functional using the
6-31+G* basis set. Harmonic frequency calculations at each level
of theory verified the identity of each stationary point as a minimum
or a transition state, and they were used to provide an estimation
of the zero-point vibrational energies, which were not scaled. IRC
calculations were carried out to determine what minima each
transition structure connects.
Acknowledgment. We are grateful to the “Ministerio de
Educacio´n y Ciencia” of Spain and FEDER (Project CTQ2005-
02323/BQU) and the “Fundacio´n Se´neca-CARM” (Project PI-
1/00749/ FS/01) for funding. J.C. is grateful to the “Ministerio
de Educacio´n y Ciencia” of Spain for a fellowship. A.P. also
thanks the “Ministerio de Educacio´n y Ciencia” of Spain and
the University of Murcia for a contract in the frame of the
“Ramo´n y Cajal” program.
Supporting Information Available: Table S1 including the
chief electronic and energetic features for all stationary points
discussed in the text. Cartesian coordinates of local minima and
transition structures discussed in the text, bond orders, and atomic
natural charges from the NBO analysis. Detailed experimental
procedures and full characterization of the thiazoles 5a-c and 5i.
Analytical and spectroscopic data for pyridines 6b-i, 7e, and 7i.
1H,1H-NOESY spectra of the pyridines 6c, 6i, 7e, 7i, and the
bicycle 13j. Crystallographic information files (CIF) for 6c and
(Z)-15. This material is available free of charge via the Internet at
http://pubs.acs.org.
(49) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03,
revision B.03; Gaussian, Inc.: Pittsburgh, PA, 2003.
(50) Hehre, W. J.; Radom, L.; Schleyer, P. v. R.; Pople, J. A. In Ab
Initio Molecular Orbital Theory; Wiley: New York, 1986; pp 71-82 and
references therein.
(51) (a) Parr, R. G.; Yang, W. Density-Functional Theory of Atoms and
Molecules; Oxford University Press: New York, 1989. (b) Bartolotti, L.
J.; Fluchichk, K. In ReViews in Computational Chemistry; Lipkowitz, K.
B., Boyds, D. B., Eds.; VCH Publishers: New York, 1996; Vol. 7, pp 187-
216. (c) Kohn, W.; Becke, A. D.; Parr, R. G. J. Phys. Chem. 1996, 100,
12974-12980. (d) Ziegler, T. Chem. ReV. 1991, 91, 651-667.
JO060664C
(52) (a) Tomasi, J.; Persico, M. Chem. ReV. 1994, 94, 2027-2094. (b)
Simkin, B. Y.; Sheikhet, I. Quantum Chemical and Statistical Theory of
Solutions-A Computational Approach; Ellis Horwood: London, 1995; pp
78-101.
(53) (a) Miertus, S.; Tomasi, J. Chem. Phys. 1981, 55, 117-129. (b)
Cammi, R.; Tomasi, J. J. Chem. Phys. 1994, 100, 7495-7502. (c) Barone,
V.; Cossi, M.; Tomasi, J. J. Chem. Phys. 1997, 107, 3210-3221.
(54) (a) Redd, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys.
1985, 83, 735-746. (b) Reed, A. E.; Curtiss L. A.; Weinhold, F. Chem.
ReV. 1988, 88, 899-926. (c) Reed, A. E.; Schleyer, P. v. R. J. Am. Chem.
Soc. 1990, 112, 1434-1445.
J. Org. Chem, Vol. 71, No. 14, 2006 5339