Dissecting competitive mechanisms: Thionation vs. cycloaddition in the reaction of thioisomuenchnones with isothiocyanates under microwave irradiation

Article abstract of DOI:10.1021/jo900960a

(Chemical Equation Presented) This paper documents in detail the reaction of 1,3-thiazolium-4-olates (thioisomuenchnones) with aryl isothiocyanates. Having demonstrated with a chiral model that thionation occurs under these conditions to provide 1,3-thiaz

Full text of DOI:10.1021/jo900960a

pubs.acs.org/joc
Dissecting Competitive Mechanisms: Thionation vs. Cycloaddition in the
€
Reaction of Thioisomunchnones with Isothiocyanates under Microwave
Irradiation†
‡
‡
‡
‡
David Cantillo,*,‡ Martı
´
ꢀ
ꢀ ꢀ
n Avalos, Reyes Babiano, Pedro Cintas, Jose L. Jimenez,
Mark E. Light,§ and Juan C. Palacios‡
‡
´
ꢀ
ꢀ
Departamento de Quımica Organica e Inorganica, QUOREX Research Group, Facultad de Ciencias,
Universidad de Extremadura, E-06071 Badajoz, Spain, and Department of Chemistry, University of
§
Southampton, Highfield, Southampton SO17 1BJ, United Kingdom
dcannie@unex.es
Received May 8, 2009
€
This paper documents in detail the reaction of 1,3-thiazolium-4-olates (thioisomunchnones) with aryl
isothiocyanates. Having demonstrated with a chiral model that thionation occurs under these
conditions to provide 1,3-thiazolium-4-thiolates and that this process is actually a stepwise domino
€
reaction (J. Org. Chem. 2009, 74, 3698-3705), we extend this study to monocyclic thioisomunch-
nones. Herein, competition between thionation and 1,3-dipolar cycloaddition takes place. The
process is synthetically disappointing at room temperature requiring prolonged reaction times for
completion. The protocol has been subsequently investigated by using both microwave dielectric
heating and conventional thermal heating (oil bath) in DMF at 100 °C with an accurate internal
reaction temperature measurement. Although a slight acceleration was observed for reactions
conducted under microwave irradiation, for most cases the observed yields and chemoselectivities
were quite similar. Thus one can conclude that, within experimental errors, the reactivity is not
related to nonthermal effects in agreement with recent reassessments on this subject, particularly by
Kappe and associates (J. Org. Chem. 2008, 73, 36; J. Org. Chem. 2009, 74, 6157). The whole reaction
system, which includes numerous heavy atoms, can be computationally modeled with a hybrid
ONIOM[B3LYP/6-31G(d):PM3] level. This reproduces well experimental results and suggests a
sequential mechanism. To further corroborate the nonconcertedness, the potential energy surface
(PES) has been constructed for simplified models, locating the corresponding stationary points. In
doing so, we introduce for the first time a useful and convenient mathematical protocol to locate the
stationary points along a reaction path. The protocol is quite simple and should convince many
organic chemists that certain daunting theoretical treatments can be made easy.
Introduction
polarizabilities, the relative ease of their synthesis, and the
possibility of tailoring the associated optical properties by
manipulation of their chemical structure.2 However, the
required stability of NLO materials limits the use of mesoionic
systems, unless it can be enhanced by a suitable exchange of
the electron-withdrawing exocyclic atoms, namely incorpora-
tion of a heavier sulfur atom. Several synthetic procedures,
including tandem cycloaddition-retrocycloaddition pro-
cesses,3 alkylation with a trialkyloxonium salt followed by
the replacement of the resulting alkoxy group by sulfide or
During the past few decades there has been a growing
interest in synthetic strategies based on the 1,3-dipolar reac-
tivity of mesoionic heterocycles, leading to novel heterocyclic
scaffolds as well as natural products or their analogues.1 Also,
in recent years these fascinating heterocycles have been con-
sidered for nonlinear optics on the basis of their large hyper-
† Dedicated to Prof. Luis Castedo on celebrating his 70th birthday.
7644 J. Org. Chem. 2009, 74, 7644–7650
Published on Web 09/23/2009
DOI: 10.1021/jo900960a
r
2009 American Chemical Society

Cantillo et al.
JOCArticle
hydrogen sulfide anion,4 or the treatment with Lawesson’s
reagent,5 have been so far described for the conversion of
mesoionic olates into mesoionic thiolates. In this context
we have recently reported the conversion of imidazo-
[2,1-b]thiazolium-3-olate systems (1 and 3) into the cor-
responding imidazo[2,1-b]thiazolium-3-thiolate derivatives
(2 and 4) by reaction with aryl isothiocyanates, and proving
unequivocally that this thionation does actually proceed via a
domino mechanism.6
Results and Discussion
Syntheses and Reactivity. The reaction of monocyclic
€
3-aryl-2-dialkylaminothioisomunchnones with aryl isothiocya-
nates leads to 3-aryl-2-dialkylamino-1,3-thiazolium-4-thiolates.
This O/S exchange occurs through two competitive reaction
pathways, a thionation process similar to that described for the
synthesis of 2 and 4, which retains the aryl group on N-3, and a
domino cycloaddition-retrocycloaddition process that adds
the aryl group from isothiocyanate to the new mesoionic
structure. Obviously, only one thiolate system is obtained when
the aryl groups of both reaction partners are identical.
Thus, mesoionic heterocycles 5-77 react with phenyl,
4-methoxyphenyl, and 4-nitrophenyl isothiocyanates (8-10) to
afford the corresponding mixtures of 3-aryl-2-(N-methylbenzyl-
amino)-5-phenyl-1,3-thiazolium-4-thiolate systems 11-13. Initi-
ally, such reactions were carried out in CH2Cl2 at room tem-
€
perature by using a 1:5 ratio of thioisomunchnone-isothio-
cyanate, and compounds 11-13 could be isolated by column
chromatography by using benzene-acetonitrile gradients after
no more than 48 h. The low yields thus obtained were increased
when DMF was used as solvent (Table 1). A further improve-
ment was achieved by conducting these transformations in DMF
at 100 °C (oil bath) leading to completion in less than 40 min.
To delineate the scope of this methodology, it is conve-
nient to move it from fused and polycyclic systems to the
more readily available monocyclic 1,3-thiazolium-4-thio-
lates. Furthermore, this study should allow us to disentangle
the fate of two competing processes, i.e., thionation and the
expected cycloaddition of the masked dipole with the hetero-
cumulenic partner. In the course of this theoretical research,
we came across a simple and intuitive protocol for the
location of all stationary points along the potential surface.
This methodology is described in detail and we judge many
synthetic chemists will find it both useful and computation-
ally inexpensive, thus overcoming the fears in moving from
the bench to virtual chemistry.
(1) (a) Potts, K. T. In 1,3-Dipolar Cycloaddition Chemistry; Padwa, A.;
Wiley: New York, 1984; Vol. 2, pp 1-82. (b) Ollis, W. D.; Stanforth, S. P.;
Ramsden, C. A. Tetrahedron 1985, 41, 2239. (c) Osterhout, M. H.; Nadler, W. R.;
Padwa, A. Synthesis 1994, 123–141. (d) Padwa, A. Top. Curr. Chem. 1997, 189,
121–158. (e) Gribble, G. W. In Synthetic Applications of 1,3-Dipolar Cycload-
dition Chemistry Toward Heterocycles and Natural Products; Padwa, A.,
Pearson, W. H., Eds.; Wiley: New York, 2003; pp 681-753. (f) Gribble, G. W.
In The Chemistry of Heterocyclic Compounds; Vol. 60. Oxazoles: Synthesis
Reactions and Spectroscopy, Part A; Palmer, D. A., Ed.; Wiley: New York, 2003;
Chapter 4. (g) Dhawan, R.; Dghaym, R. D.; Arndtsen, B. A. J. Am. Chem. Soc.
To find the optimal conditions for the syntheses of 11-13, the
above-mentioned reactions were alternatively attempted under
microwave irradiation, using a multimode microwave instru-
ment (maximum power: 300 W; temperature: 100 °C measured
with use of an internal fiber-optic probe) in DMF as solvent, a
polar medium suitable for dielectric heating.8 Under these con-
ditions, transformations are complete within 5 min. This accel-
eration causes energy savings but it does not substantially affect
the chemoselectivity of either of the global yields with respect to
the conventional heating for the same time.9 Although the
conventionally heated reaction, conducted in the identical reac-
tion vessel at 100 °C, usually requires 10-40 min for completion,
we deliberately stopped these protocols after 5 min and evaluated
both overall yields and selectivities (Table 1), which show a close
similarity to those completed after 5 min of irradiation.
ꢀ
ꢀ
2003, 125, 1474–1475. (h) Avalos, M.; Babiano, R.; Cintas, P.; Jimenez, J. L.;
Palacios, J. C. Acc. Chem. Res. 2005, 38, 460–468. (i) Lu, Y.; Arndtsen, B. A.
Angew. Chem., Int. Ed. 2008, 47, 5430–5433.
(2) (a) Morley, J. O. J. Phys. Chem. 1995, 99, 1923–1927. (b) Moura, G. L.
C.; Simas, A. M.; Miller, J. Chem. Phys. Lett. 1996, 257, 639–646. (c) Bezerra,
A. G. Jr.; Gomes, A. S. L.; Athayde-Filho, P. F.; Rocha, G. B. da; Miller, J.;
Simas, A. M. Chem. Phys. Lett. 1999, 309, 421–426. (d) Menezes, L. de S.;
ꢀ
Araujo, C. B. de; Alencar, M. A. R. C.; Athayde-Filho, P. F.; Miller, J.;
Simas, A. M. Chem. Phys. Lett. 2001, 347, 163–166. (e) Pilla, V.; Araujo, C.
ꢀ
B. de; Lira, B. F.; Simas, A. M.; Miller, J.; Athayde-Filho, P. F. Opt.
Commun. 2006, 264, 225–228. (f) Bosco, C. A. C.; Maciel, G. S.; Rakov,
ꢀ
N.; Araujo, C. B. de; Acioli, L. H.; Simas, A. M.; Athayde-Filho, P. F.;
Miller, J. Chem. Phys. Lett. 2007, 449, 101–106.
(3) (a) Huisgen, R.; Funke, E.; Schaefer, F. C.; Gotthardt, H.; Brunn, E.
Tetrahedron Lett. 1967, 1809–1814. (b) Lira, B. F.; Athayde-Filho, P. F.;
Miller, J.; Simas, A. M.; Dias, A. F.; Vieira, M. J. Molecules 2002, 7, 791–800.
(4) (a) Potts, K. T.; Houghton, E.; Husain, S. J. Chem. Soc., Chem.
Commun. 1970, 1025–1026. (b) Hanley, R. N.; Ollis, W. D.; Ramsden, C. A.
J. Chem. Soc., Perkin Trans. I 1979, 732–735. (c) Hanley, R. N.; Ollis, W. D.;
Ramsden, C. A. J. Chem. Soc., Perkin Trans. I 1979, 741-743. (d) Masuda, K.;
Adachi, J.; Nomura, K. J. Chem. Soc., Perkin Trans. I 1979, 956–959.
(e) Adachi, J.; Takahata, H.; Nomura, K.; Masuda, K. Chem. Pharm. Bull.
1983, 31, 1746–1750.
Microwave irradiation causes various thermal effects,
which are clearly distinctive from those from conventional
(7) Avalos, M.; Babiano, R.; Cabanillas, A.; Cintas, P.; Higes, F. J.;
ꢀ
Jimenez, J. L.; Palacios, J. C. J. Org. Chem. 1996, 61, 3738–3748.
(8) (a) For a recent tutorial on microwave heating and its applications in
organic synthesis see: Kappe, C. O. Chem. Soc. Rev. 2008, 37, 1127–1139.
(b) Perreux, L.; Loupy, A. In Microwaves in Organic Synthesis; Loupy, A., Ed.;
Wiley-VCH: Weinheim, Germany, 2006; Chapter 4, pp 134-218. (c) For a recent
and authoritative discussion on microwave apparatuses and protocols see:
Practical Microwave Synthesis for Organic Chemists. Strategies, Instruments,
and Protocols; Kappe, C. O., Dallinger, D., Murphree, S., Eds.; Wiley-VCH:
Weinheim, Germany, 2008. (d) Herrero, M. A.; Kremsner, J. M.; Kappe, C. O.
J. Org. Chem. 2008, 73, 36–47.
(5) Araki, S.; Goto, T.; Butsugan, Y. Bull. Chem. Soc. Jpn. 1988, 61,
2977–2978.
ꢀ
(6) (a) Cantillo, D.; Avalos, M.; Babiano, R.; Cintas, P.; Jimenez, J. L.;
ꢀ
Light, M. E.; Palacios, J. C. Org. Lett. 2008, 10, 1079–1082. (b) Cantillo, D.;
ꢀ
ꢀ
Avalos, M.; Babiano, R.; Cintas, P.; Jimenez, J. L.; Light, M. E.; Palacios, J. C.
J. Org. Chem. 2009, 74, 3698-3705.
(9) Razzaq, T.; Kappe, C. O. ChemSusChem 2008, 1, 123–132.
J. Org. Chem. Vol. 74, No. 20, 2009 7645

Cantillo et al.
JOCArticle
€
TABLE 1. Isolated Yields of Compounds 11-13 in the Reactions of Thioisomunchnones 5-7 with Aryl Isothiocyanates 8-10
reaction products (% yield)
procedure A:
CH2Cl2, rt,
24-48 ha
procedure B:
DMF, rt,
procedure D:
DMF, MW, 300 W,
100 °C, 5 min
procedure C: DMF,
100 °C, 10-40 mina
procedure E:
DMF, 100 °C, 5 min
entry
reactants
12-24 ha
1
2
3
4
5
6
7
8
9
5 þ 8
5 þ 9
5 þ 10
6 þ 8
6 þ 9
6 þ 10
7 þ 8
7 þ 9
7 þ 10
11 (9)
11 (56)
11 (68)
11 (58)
11 (62)
11 (2) þ 12 (2)
11 (44) þ 13 (1)
12 (3) þ 11 (2)
12 (2)
12 (38) þ 13 (1)
13 (7) þ 11 (27)
13 (3) þ 12 (7)
13 (61)
11 (21) þ 12 (39)
11 (99) þ 13 (1)
12 (54) þ 11 (8)
12 (32)
12 (99) þ 13 (1)
13 (11) þ 11 (57)
13 (8) þ 12 (59)
13 (>99)
11 (21) þ 12 (37)
11 (75) þ 13 (11)
12 (50) þ 11 (15)
12 (59)
12 (78) þ 13 (3)
13 (12) þ 11 (73)
13 (7) þ 12 (51)
13 (91)
11 (7) þ 12 (27)
11 (62) þ 13 (12)
12 (52) þ 11 (36)
12 (56)
12 (88) þ 13 (8)
13 (5) þ 11 (80)
13 (1) þ 12 (73)
13 (76)
11 (11) þ 12 (18)
11 (70) þ 13 (10)
12 (49) þ 11 (15)
12 (17)
12 (64) þ 13 (4)
13 (11) þ 11 (70)
13 (4) þ 12 (43)
13 (88)
aThe fastest reactions take place when the isothiocyanate bears an electron-withdrawing aryl substituent.
TABLE 2. Selected 1H and 13C Chemical Shifts (ppm) of Compounds 5-7 and 11-13
1H chemical shifts
13C chemical shifts
compd
N-CH2
N-CH3
O-CH3
N-CH2
N-CH3
O-CH3
C-2
C-4
C-5
5
6
7
11
12
13
4.23 s
4.28 s
4.12 s
4.28 s
4.33 s
4.30 s
2.74 s
2.79 s
2.64 s
2.75 s
2.80 s
2.80 s
57.7
57.6
58.1
59.1
59.0
59.5
40.0
40.0
40.5
41.5
41.5
42.0
156.2
156.5
155.5
166.6
166.8
166.8
161.2
161.3
161.4
156.8
156.9
156.1
79.8
79.9
79.8
108.0
108.1
108.6
3.80 s
57.6
Structure Characterization. Crystalline structures of com-
pounds 11, 12, and 13 could be satisfactorily elucidated by
X-ray diffraction analyses (Figure 1).
Molecular structures of 11, 12, and 13 show the perpendi-
cular and coplanar arrangements of the mesoionic heterocycle
with respect to the aryl groups linked to the N-3 and C-5 atoms,
respectively. In the case of 12, suitable crystals for X-ray
analysis were obtained from diethyl ether, while for compounds
11 and 13, diffraction analyses could only be accomplished after
recrystallization from benzene. A salient polymorphism was
observed depending on the solvent system and, remarkably, a
benzene molecule is encapsulated in the crystalline lattices of
compounds 11 and 13 (see the Supporting Information).
1H NMR data of compounds 11-13 are extremely simple
FIGURE 1. Structures of 11, 12, and13 obtained by X-ray diffraction.
heating. Nonthermal microwave effects appear now to be
flawed in the absence of experimental evidence. Leading
experts have clearly evidenced that performing meaningful
comparison experiments between microwaves and conven-
tional heating requires above all accurate internal tempera-
ture measurements.8 In agreement with such studies, the
present results reveal that both reactivity and chemoselec-
tivity should be considered essentially identical, within ex-
perimental errors, and not related to nonthermal effects.
It should be finally mentioned that these transformations were
also attempted in CH2Cl2 at 100 °C (sealed tube) under micro-
wave irradiation. Such protocols were unsuccessful as quick and
€
and similar to those of the parent thioisomunchnones 5-7
(Table 2). However, the O/S exchange causes a strong de-
shielding (Δδ ∼28 ppm) of the C-5 signal in 11-13 relative to
that of compounds 5-7, while the resonances of the C-2 and
C-4 carbon atoms are shifted downfield (Δδ ∼10.4 ppm) and
upfield (4.4-5.3 ppm), respectively. This spectroscopic beha-
vior is analogous to that previously found for compounds 1
and 2, whose equivalent carbon atoms resonate at 151.7,
155.0, 83.9 ppm, and 160.0, 149.0, 111.0 ppm, respectively.
Mechanistic Insights: A Theoretical Study. In agreement
with the mechanism proposed6b for the thionation of 1 and 3,
€
extensive decomposition of both the parent thioisomunchnone
€
the transformation of the thioisomunchnones 5-7 into
and the resulting 1,3-thiazolium-4-thiolate were observed.
It is noteworthy the extent of the influence of aryl groups,
€
present in the thioisomunchnone and isothiocyanate partners, on
the chemoselectivity of this reaction. When the aryl substituent
11-13 might involve a stepwise mechanism (depicted in
Scheme 1 for the reaction of 5 and 8), which, in this case,
could also account for the formation of product mixtures
when different N-aryl groups are present in both reagents,
on the isothiocyanate is more electron withdrawing than the N-
€
€
the thioisomunchnone and the aryl isothiocyanate.
aryl group of the thioisomunchnone, the thionation becomes the
favorite pathway (entries 3 and 6, Table 1). On the other hand,
the 1,3-dipolar cycloaddition route is prevalent when the more
The overall process would be launched with the formation
of an open dipolar intermediate (14) by nucleophilic addition
of the C-5 heterocyclic carbon of 5 onto the thiocarbonylic
carbon of the aryl isothiocyanate. Then, 14 could evolve into
the azetidinic intermediate (15) by nucleophilic attack of the
€
electron withdrawing aryl group is linked to the thioisomunch-
none (entries 7 and 8, Table 1). A mechanistic rationale account-
ing for these experimental observations will be discussed later.
7646 J. Org. Chem. Vol. 74, No. 20, 2009

Cantillo et al.
JOCArticle
SCHEME 1. Stepwise Mechanisms for Both Thionation and
1,3-Dipolar Cycloaddition of the 1,3-Thiazolium-4-olate 5 with
Phenyl Isothiocyanate (8)a
FIGURE 2. ONIOM[B3LYP/6-31G(d):PM3]-optimized structures
€
for the thioisomunchnone 5 and phenyl isothiocyanate (8).
aBlue is used to show the fate of the phenyl isothiocyanate moiety in the
outcome of such processes.
FIGURE 3. Energetics of the competitive thionation and cyclo-
€
addition reactions of thioisomunchnone 5 and phenyl isothio-
cyanate (8).
nitrogen on the neighboring carbonyl group. The free rota-
tion around the C-S bonds of 15 enables the isothioureidic
nitrogen to be close to the thiocarbonylic carbon giving rise
to a new open dipolar intermediate (16), which would
ultimately lead to the 1,3-thiazolium-4-thiolate (11), the
latter retaining the N-aryl group of the parent 1,3-thiazo-
lium-4-olate system (5). Alternatively, 14 could be converted
into the cycloadduct 17 by attack of the nucleophilic nitrogen
on the C-2 carbon atom. The retrocycloaddition reaction
of phenyl isocyanate could also lead to a 1,3-thiazolium-
4-thiolate (11), but in this case, its N-phenyl group might
arise from phenyl isothiocyanate (Scheme 1). In addition, the
extrusion of sulfur from 17 could give rise to the betaine 19.
This substance has not been detected in these reactions,
though this type of heterocyclic system was formed during
the thionation of compounds 1 and 3.
isothiocyanate (8) using the ONIOM[B3LYP/6-31G(d):
PM3]10,11 method. Figure 2 shows both reactants in which
balls and sticks represent the partial system treated at the
B3LYP/6-31G(d) level and the wire-shaped structure corre-
sponds to the molecular fragment assessed at a semiempirical
level (PM3).
Figure 3 collects the relative electronic energies of all
stationary states involved in both the thionation and the
1,3-dipolar cycloaddition. Remarkably, a tetrahedral inter-
mediate (20) and two saddle points (TS15f20 and TS20f16
)
could be optimized as previous stationary points leading to
16. An analogous tetrahedral intermediate, however, could
not be obtained for the transformation of 14 into 15.
This mechanism agrees with the observed chemoselecti-
vity, which can further be rationalized in terms of the HSAB
principle. An electron-withdrawing group at the N-3 atom of a
zwitterion like 14 does increase the hardness of the C-2
position, thereby favoring the formation of cycloadduct 17,
which will be especially noticeable on increasing the anion
hardness with an electron-releasing group at N-6. In stark
contrast, electron-withdrawing groups at N-6 significantly
favor the thionation step because the anion hardness di-
minishes. Under such circumstances, addition to the vicinal,
softer carbonyl group becomes a more favorable pathway.
Moreover, we have carried out a computational study on
The relative electronic energy of reactants and products
points to a thermodynamically favored transformation
of 5 into 11 and, from a kinetic viewpoint, the rate-
determining step for both competitive processes (thiona-
tion as well as 1,3-dipolar cycloaddition) must involve
formation of the open dipolar intermediate 16. By estimat-
ing the magnitude of the two energy barriers affording 16
(17.8 and 15.4 kcal/mol) one concludes that the 1,3-dipolar
cycloaddition seems to be faster than the thionation; how-
ever, experimental results (Table 1) show that the major
product of each reaction depends largely on the reaction
conditions and the donor/acceptor character of the aryl
€
the global reactivity of the thioisomunchnone 5 with phenyl
ꢀ
(10) (a) Dapprich, S.; Komaroni, I.; Byun, K. S.; Morokuma, K.; Frish,
M. J. THEOCHEM 1999, 462, 1–21. (b) Svensson, M.; Humbel, S.; Froese,
R. D. J.; Matsubara, T.; Sieber, S.; Morokuma, K. J. Phys. Chem. 1996, 100,
19357–19363.
(11) Froese, R. D. J.; Morokuma, K. In The Encyclopedia of Computa-
tional Chemistry; Schleyer, P. v. R., Allinger, N. L., Clark, T., Gasteiger, J.,
Kollman, P. A., Schaefer, H. F., III, Schreiner, P. R., Eds.; John Wiley:
Chichester, UK, 1998; pp 1245-1257.
€
groups present in both reactants, thioisomunchnone and
J. Org. Chem. Vol. 74, No. 20, 2009 7647

Cantillo et al.
JOCArticle
TABLE 3. Energy Barrier Differences of the Rate-Determining Steps
(Thionation Minus Cycloaddition) As a Function of the Nature of the Aryl
Substituents of Both Reagents
ΔΔEq
(kcal/mol)a
€
thioisomunchnone
entry
aryl isothiocyanate
1
2
3
4
5
6
5
5
7
7
10
10
9
8
9
-4.99
-3.83
þ4.84
þ10.52
þ11.38
aCalculated using the ONIOM[B3LYP/6-31G(d):PM3] method.
SCHEME 2
FIGURE 4. PES corresponding to the formation of 23 by reaction
of 21 and 22, calculated at the B3LYP/6-31G(d) level with full
optimization of each structure.
isothiocyanate. Table 3 shows the variation in these energy
barriers when the electronic nature of such aryl groups
changes. Product ratios and yields reported in Table 1 are
consistent with the calculated ΔΔEq values, which consti-
tutes a sound support for the mechanism proposed
(Scheme 1). Likewise, this fact validates the theoretical
method at the level of theory employed, even though the
aryl substituents of the whole molecular system were
assessed at a semiempirical level.
At this stage, one could formulate an intriguing question,
even in the absence of evidence, i.e., Does an alternative
concerted mechanism for the 1,3-dipolar cycloaddition of
€
thioisomunchnones with aryl isothiocyanates exist? To
provide an unambiguous response, we decided to construct
the potential energy surface (PES) for this reaction. To do it
at a reasonable computational cost, we were forced to
modify the structures of 5, 8, and 17 to generate the
simplified models 21, 22, and 23, which preserve the essen-
tial characteristics of the original molecules (Scheme 2). A
total of 234 (13 ꢀ 18) full optimized [B3LYP/6-31G(d) level]
structures were required to create the PES (Figure 4),
stemming from the sequential lengthening (0.1 A) of both
the C-N (a) and C-C (b) bonds.
FIGURE 5. Three-dimensional representation of the Chebyshev
polynomial that fit to the PES data of Figure 4.
Location of Stationary Points Along the Reaction Path.
Although at first sight the PES suggests a stepwise mechan-
ism, a simple inspection does not allow us to establish the real
location of each stationary point for this process. To make
this work easier, we have come up with a suitable strategy to
locate stationary points such as optimal structures (local
minima) and saddle points in a reliable and unambiguous
fashion.
A fitting procedure of these PES data using a Chebyshev
polynomial12 produces a new hypersurface with a three-
dimensional representation essentially identical (r > 0.9999)
with the former (Figure 5), thereby evidencing that this poly-
nomial constitutes the best approximation to a continuous
function.
such stationary points are defined as points in the two-
dimensional space (surface), where
Dz
Dx
Dz
Dy
¼ 0 and
¼ 0
Calculations of first derivatives with respect to either x or y
coordinates and their graphical plots give rise to contour
maps of isovalue lines. Colored areas can be used to depict
zones where the derivative becomes close to zero (<0.001 or
<0.0001), the lesser the derivative values the thinner the area
obtained. Superposition of both plots affords intersection
points, which correspond to any stationary point along the
reaction path (Figure 6).
The contour maps readily identify three stationary points
corresponding to two saddle points (located in blue and
pink areas, which indicate C-N and C-C bond forming
processes, respectively, Figure 7), plus an intermediate
(in the white area, Figure 7), thus pointing to the stepwise
For local minima and saddle points (structures with zero
gradient), the function slope and therefore its derivative
equal zero. In mathematical language, it is well-known that
(12) Boyd, J. P. Chebyshev and Fourier Spectral Methods; Dover Publica-
tions, Inc.: Mineola, NY, 2001.
7648 J. Org. Chem. Vol. 74, No. 20, 2009

Cantillo et al.
JOCArticle
FIGURE 6. (a) Superposition of first derivative plots. Red and blue areas correspond to zones where the derivative values with respect to x
(C-N bond) and y (C-C bond), respectively, are close to zero. (b) Superposition of first derivative plots over the Chebyshev polynomial
representation.
consistent with a mathematical algorithm and has proven to
be both reliable and accurate in locating stationary points in
numerous cycloaddition reactions.14
Concluding Remarks
In summary, this work provides an experimental and
theoretical study for the reaction of monocyclic 2-amino-
1,3-thiazolium-4-olates with aryl isothiocyanates. This
transformation is consistent with two competitive stepwise
pathways, namely the thionation and 1,3-dipolar cycloaddi-
tion processes, which strongly depend on the substituents of
the aryl groups of both reagents. Compounds 11-13 could
be gratifyingly obtained as crystalline materials suitable for
single-crystal diffraction analyses, after chromatographic
purification (benzene-acetonitrile gradients).
FIGURE 7. Superposition of first and second derivative plots. Pink
and blue zones indicate concave surface areas (maxima) with respect
to x and y, respectively.
While conventional reactions conducted under mild con-
ditions are sluggish, either microwave irradiation or oil-bath
heating at 100 °C in DMF represent valuable protocols from
a synthetic point of view. Global reactivities and selectivities
should be considered almost identical, and therefore not
directly associated to the presence of the microwave field.
A hybrid ONIOM-based theoretical methodology has
been performed to simulate the chemical reaction at the
B3LYP/6-31G(d):PM3 level, which is consistent with experi-
mental data. Finally, we introduce a novel and straightfor-
ward method to construct and interpret potential energy
surfaces (PES) to evaluate the reaction pathways. This
analysis on a simplified reaction model confirms the non-
concerted nature of these reactions.
FIGURE 8. Full optimized structures of saddle points 24 and 26
and the open dipolar intermediate 25.
Experimental Section
pathway. These stationary points were further confirmed
by conventional optimization and frequency analysis
(Figure 8).
€
General Procedures for the Reaction of Thioisomunchnones
5-7 with Aryl Isothiocyanates 8-10. Procedure A. To a solution
of thioisomunchnone (3 mmol) in CH2Cl2 (30 mL) was added
€
At this point, it should strongly be emphasized that the
above methodology is not intended to replace the analytical
construction of PESs, which represents a formidable task,
even for very discrete species.13 This rather naive protocol is
the corresponding aryl isothiocyanate (15 mmol). The reaction
mixture was stirred at room temperature and monitored by TLC
(benzene/acetonitrile 3:1) until the complete disappearance of
(14) We have tested the methodology in more than 30 typical cycloaddi-
tion reactions and, in all the cases, the stationary points thus obtained could
further be confirmed by optimization and frequency analyses (unpublished
results). The validity and extension to other organic reactions is underway.
(13) For a recent revision see: Albu, T. V.; Espinosa-Garcı
D. G. Chem. Rev. 2007, 107, 5101–5132.
´
a, J.; Truhlar,
J. Org. Chem. Vol. 74, No. 20, 2009 7649

Cantillo et al.
JOCArticle
the substrate (no more than 48 h). Then, the solvent was
evaporated and the resulting crude mixture was purified by
flash chromatography (benzene/acetonitrile gradient elution
from 10:1 to 1:3).
Procedure B. In a 10-mL round flask were placed the thio-
€
isomunchnone (0.8 mmol), the corresponding aryl isothiocyanate
(4 mmol), and DMF (5 mL). The mixture was stirred at room
temperature until the complete disappearance of the substrate (no
more than 24 h). Then, the mixture was purified by flash chroma-
tography (benzene/acetonitrile gradient elution from 10:1 to 1:3).
Procedure C. In a 10-mL round flask were placed the thioi-
(dd, 2H, J=0.8, 8.4 Hz), 7.46 (t, 3H, J=7.2 Hz), 7.34 (m, 7H),
7.18 (t, 1H, J=7.2 Hz), 7.01 (dd, 2H, J=2.4, 7.6 Hz), 4.28 (s,
2H), 2.75 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 166.6, 156.8,
138.4, 133.5, 129.9, 129.5, 129.3, 129.1, 128.5, 128.3, 128.2,
127.1, 126.7, 126.0, 108.0, 59.1, 41.5. Anal. Calcd for
C23H20N2S2 1/2C6H6: C, 73.03; H, 5.42; N, 6.55; S, 15.00.
3
Found: C, 72.75; H, 5.68; N, 6.77; S, 14.72.
3-(4-Methoxyphenyl)-2-(N-methylbenzylamino)-5-phenyl-1,3-
thiazolium-4-thiolate (12): mp 156-157 °C; 1H NMR (400
MHz, CDCl3) δ 8.29 (dd, 2H, J=0.8, 8.4 Hz), 7.34 (m, 5H),
7.28 (d, 2H, J=9.2 Hz), 7.18 (t, 1H, J=7.2 Hz), 7.03 (dd, 2H, J=
2.4, 8.0 Hz), 6.96 (dd, 2H, J=2.4, 6.8 Hz), 4.33 (s, 2H), 3.81 (s,
3H), 2.80 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 166.8, 160.3,
133.7, 133.5, 130.8, 130.5, 129.1, 128.5, 128.2, 127.0, 126.7,
126.0, 114.5, 59.0, 55.5, 41.5. Anal. Calcd for C24H22-
€
somunchnone (0.8 mmol), the corresponding aryl isothiocya-
nate (4 mmol), and DMF (1 mL). The mixture was heated in an
oil bath at 100 °C until the complete disappearance of the
substrate (no more than 40 min). Then, the mixture was cooled
to room temperature and purified by flash chromatography
(benzene/acetonitrile gradient elution from 10:1 to 1:3).
N2OS2 1/2C6H6: C, 70.86; H, 5.51; N, 6.12; S, 14.01. Found:
3
C, 70.64; H, 5.37; N, 6.39; S, 14.22.
2-(N-Methylbenzylamino)-3-(4-nitrophenyl)-5-phenyl-1,3-thia-
Procedure D. In a 10-mL round flask were placed the thio-
€
1
isomunchnone (0.8 mmol), the corresponding aryl isothiocya-
zolium-4-thiolate (13): mp 145-146 °C; H NMR (400 MHz,
nate (4 mmol), and DMF (1 mL). The mixture was heated in an
open vessel under microwave radiation (maximum power:
300 W) with a multimode instrument (Ethos, Advanced MW
Lab-station) from Milestone, with the temperature ramping
from rt to 100 °C within 1 min (using a fiber-optic probe to
estimate the temperature profile). Then, the reaction was held at
this temperature for 4 min. The mixture was cooled to room
temperature and purified by flash chromatography (benzene/
acetonitrile gradient elution from 10:1 to 1:3).
CDCl3) δ 8.28 (dd, 2H, J=1.2, 8.4 Hz), 8.23 (d, 2H, J=8.8 Hz),
7.58 (d, 2H, J=8.8 Hz), 7.38 (t, 2H, J=7.6 Hz), 7.33 (t, 3H, J=
3.2 Hz), 7.22 (t, 1H, J =7.2 Hz), 6.58 (m, 2H), 4.30 (s, 2H),
2.80 (s, 3H); 13C NMR (100 MHz, CDCl3) δ 166.8, 156.1,
148.1, 143.5, 133.1, 132.9, 131.0, 129.3, 128.9, 128.4, 126.8,
126.6, 126.4, 124.3, 108.6, 59.5, 42.0. Anal. Calcd for
C23H19N3O2S2 2/3C6H6: C, 66.78; H, 4.77; N, 8.65; S, 13.21.
3
Found: C, 66.51; H, 5.03; N, 8.48; S, 12.89.
Procedure E. In a 10-mL round flask were placed the thio-
€
Acknowledgment. This work was supported by the Mini-
ꢀ
isomunchnone (0.8 mmol), the corresponding aryl isothiocya-
sterio de Educacion y Ciencia and FEDER (CTQ2007-
66641) and the Junta de Extremadura (PRI07A016 and
PRI08A032). One of us (D.C.) thanks the Spanish Ministerio
nate (4 mmol), and DMF (1 mL). The mixture was heated in an
oil bath at 100 °C for 5 min. Then, the mixture was cooled to
room temperature and purified by flash chromatography
(benzene/acetonitrile gradient elution from 10:1 to 1:3).
The complete set of conventional (36 reactions) and micro-
waved (9 reactions) transformations has been carried out fol-
lowing the above procedures A-D and E, respectively. Their
product distribution and yields are listed in Table 1. Analytical
data of compounds 11-13 are as follows.
ꢀ
de Educacion y Ciencia for a fellowship.
Supporting Information Available: 1H and 13C NMR spec-
tra for compounds 11-13, crystal data for compounds 11-13 in
CIF format, and computational data for all structures com-
puted at the ONIOM[B3LYP/6-31G(d):PM3] level. This ma-
terial is available free of charge via the Internet at http://pubs.
acs.org.
2-(N-Methylbenzylamino)-3,5-diphenyl-1,3-thiazolium-4-thio-
late (11): mp 158-159 °C; 1H NMR (400 MHz, CDCl3) δ 8.30
7650 J. Org. Chem. Vol. 74, No. 20, 2009