Unexpected valence bond isomerization of [1,2,4]triazolo[3,4-c][1,2,4] benzotriazines under flash vacuum pyrolytic (fvp) conditions

Article abstract of DOI:10.1016/j.tet.2005.05.047

Flash vacuum pyrolysis (fvp) of some substituted [1,2,4]triazolo[3,4-c][1, 2,4]benzotriazine derivatives (1a-d) has been studied between 450 and 600°C. The only transformation observed up to 525°C was the unexpected valence bond isomerization of the angularly fused starting compounds to the isomeric linearly fused [1,2,4]triazolo[4,3-b][1,2,4]benzotriazine derivatives (9a-d), whereas at higher temperatures fragmentation products such as aromatic nitriles were also formed. Kinetic measurements revealed negative entropies of activation in the isomerization process, which suggest a concerted ring closure reaction to an intermediate antiaromatic diazirine. Reversibiblity of the title isomerization reaction was also proved by FVP experiments.

Full text of DOI:10.1016/j.tet.2005.05.047

Tetrahedron 61 (2005) 7489–7498
Unexpected valence bond isomerization of [1,2,4]triazolo[3,4-c]
[
1,2,4]benzotriazines under flash vacuum pyrolytic
fvp) conditions
(
a
Walter J. Pel a´ ez, Gloria I. Yranzo, Csilla Gr o´ f, Zsuzsanna Riedl and Gy o¨ rgy Haj o´ s
a
b
b
b,
*
a
INFIQC-Depto. de Qu ´ı mica Org a´ nica, Facultad de Ciencias Qu ´ı micas, U.N.C. C o´ rdoba, Argentina
Chemical Research Center, Institute of Biomolecular Chemistry, Hungarian Academy of Sciences, Pusztaszeri u´ t 59-67,
b
H-1025 Budapest, Hungary
Received 21 February 2005; revised 5 May 2005; accepted 19 May 2005
Available online 15 June 2005
Abstract—Flash vacuum pyrolysis (fvp) of some substituted [1,2,4]triazolo[3,4-c][1,2,4]benzotriazine derivatives (1a–d) has been studied
between 450 and 600 8C. The only transformation observed up to 525 8C was the unexpected valence bond isomerization of the angularly
fused starting compounds to the isomeric linearly fused [1,2,4]triazolo[4,3-b][1,2,4]benzotriazine derivatives (9a–d), whereas at higher
temperatures fragmentation products such as aromatic nitriles were also formed. Kinetic measurements revealed negative entropies of
activation in the isomerization process, which suggest a concerted ring closure reaction to an intermediate antiaromatic diazirine.
Reversibiblity of the title isomerization reaction was also proved by FVP experiments.
q 2005 Elsevier Ltd. All rights reserved.
1. Introduction
Flash vacuum pyrolysis (fvp) of nitrogen containing
heterocycles has been the focus of our studies for many
1
years as reflected by a recently published review. In the
course of these investigations, heteroaromatic rings
involving the N–N moiety in the ring seemed of special
interest due to the possibility of nitrogen extrusion and
Scheme 1.
1
,2
formation of diradicals or vinylcarbenes. From this point
of view, interesting results have been obtained by
comparison of the behavior of [1,2,3]- and [1,2,4]-
benzotriazines: while most of [1,2,3]-benzotriazines
the special structural feature that they contain two N–N
parts in the ring and, in principle, two kinds of nitrogen
elimination could occur under fvp conditions.
3
Five-membered compounds with an N]N moiety in the
ring can undergo nitrogen elimination under thermal and
photochemical conditions. Thus, 1-arylbenzotriazoles are
convenient precursors of carbazoles by thermal extrusion of
nitrogen, which is known as the Graebe-Ullman synthesis.
The reaction mechanism was confirmed to involve
cyclisation of a diradical or iminocarbene to 4H-carbazole
which isomerizes to substituted aromatic carbazoles by a
afforded reaction products via benzazetes, no evidence
for these intermediates was found in fvp of 1-deoxy or 1-oxy
derivatives of [1,2,4]-benzotriazines and, instead, benzo-
nitriles and biphenylene were found as final products.
Furthermore, in the case of 3-methylsulfanyl derivatives of
these compounds, a competing radical reaction was also
4
found in fvp due to the lability of the C–S bond.
5
hydrogen shift.
As a continuation of these studies we report here, on flash
vacuum pyrolysis of some [1,2,4]triazolo[3,4-c]benzo-
triazines (1a–d, Scheme 1). These model compounds have
6
7
Some [1,2,4]-triazoles with alkyl or acyl groups attached
to a ring atom afforded migration of these substituents
followed by further transformations. In the former case,
isomeric 1,2,4-triazoles were formed, whereas in the second
case nitrogen extrusion took place to result in a ring
transformation of the starting compound (Scheme 2).
Keywords: Flash vacuum pyrolysis; Valence bond isomerisation; Sigma-
tropic rearrangement; Enthropy of activation; Azirine intermediate.
*
Corresponding author. Tel.: C361 325 7550; fax: C361 325 7863;
e-mail: ghajos@chemres.hu
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.05.047

7490
W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
Scheme 2.
1
and 9a calculated from the H NMR spectra are summarized
in Table 1.
Thus, 1-acetyl-1H-1,2,4-triazole (2) underwent an N–C acyl
migration, and the resulting intermediate (4) lost one
molecule of nitrogen to yield 5-methyloxazole (5). A
similar transformation was observed with 1-benzoyl-1H-
pyrazole (3), where formation of 2-phenylfurane (7) via
Compound 10 may be formed from 1a or from 9a by
nitrogen extrusion or by fragmentation to benzeneiso-
cyanide (8) and subsequent isomerization to 10 via a well
8
intermediate 6 was reported. Unfortunately, no references
on reactions of fused 1,2,4-triazoles were found.
9
known pathway.
2.2. Fvp of 7-methyl-[1,2,4]triazolo[3,4-c][1,2,4]benzo-
triazine (1b)
2
. Results and discussion
.1. Fvp of [1,2,4]triazolo[3,4-c][1,2,4]benzotriazine (1a)
Reactions of 1a were carried out between 450 and 600 8C,
2
Reactions of 1b were carried out between 450 and 625 8C,
K2
K2
with pressures of w10 Torr and contact times of w10 s.
The only product up to 525 8C was 7-methyl-[1,2,4]triazo-
lo[4,3-b][1,2,4]benzotriazine (9b); over this temperature 10,
p-tolunitrile (11) and a mixture of dicyanobenzenes (12–14)
were also formed from decomposition of 1b and/or 9b
(Scheme 4, Table 1). Relative quantification of 1b and 9b
K2
K2
with pressures of w10 Torr and contact times of w10 s.
The reaction was clean affording one single product up to
25 8C identified as [1,2,4]triazolo[4,3-b][1,2,4] benzo-
triazine (9a) (Scheme 3). Over 550 8C small amounts of
benzonitrile (10) and a carbonaceous residue were formed
from decomposition (Scheme 3). The relative amounts of 1a
5
1
was carried out by the help of H NMR spectra. It is
important to note that 100% conversion could not be
achieved and over 600 8C, coke was also formed.
Compounds 10 and 12–14 can be formed by pyrolysis of
1
0
1
1 as already reported.
Scheme 3.
Table 1. Fvp reactions of 1a and 1b
T (8C)
% 1
% 9
Other
products
Scheme 4.
450
475
500
525
550
575
600
625
1a
95.3
94.2
90.6
90.4
82.5
82.4
74.8
71.8
72.7
69.5
72.7
62.5
79.2
64.9
—
4.7
5.8
9.4
9.6
—
—
—
—
—
—
—
—
1b
1a
1b
1a
17.5
17.6
25.2
28.2
27.3
30.5
27.3
37.5
20.8
35.1
—
1b
1a
1b
a
1a
10
11
a
1b
a
1a
10
9, 11
10
a
1b
a
1a
a
a
1b
10, 12–14
—
10, 12–14
1a
1b
76.2
23.8
a
Detected by GC/MS, not quantified.
Scheme 5.

W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
7491
2.3. Fvp of 7-methoxy [1,2,4]triazolo[3,4-c][1,2,4]benzo-
triazine (1c)
3-hydrazinobenzo[1,2,4]triazines (18) with ethyl ortho-
formiate—and carried out thorough analyses of the mother
liquors of these reaction mixtures to learn whether the
isomeric linearly fused compounds 9a–d had also been
formed in small amounts in these transformations
(Scheme 7).
Reactions of 1c were carried out between 440 and 540 8C,
K2
K2
with pressures of w10 Torr and contact times of w10 s.
Up to 520 8C the only product was 7-methoxy-[1,2,4]tri-
azolo[4,3-b][1,2,4]benzotriazine (9c), over this tempera-
ture, anisole (15), benzene (16) and coke were also formed
TLC of all four reaction mixtures indicated that besides the
intense spots for the main products (1a–d), a well defined
other spot is also present. Separation of these fractions by
column chromatography gave rise to the isolation of 9a–d
which proved to be entirely identical (mp, TLC, IR
NMR-spectra) with the samples obtained by FVP. It is
interesting to note that the 1/9 ratios in the reaction mixtures
obtained from 18a–d proved to be very different, the
(
Scheme 5 and Table 2). Relative quantification of 1c and 9c
was performed by H NMR.
1
Table 2. Fvp reactions of 1c and 1d
T (8C)
% 1
% 9
Other
products
440
460
480
500
520
540
560
1c
100
—
—
—
—
—
—
—
—
—
—
—
—
—
2
7
isolated yields are shown in Table 3 (Scheme 7).
1d
1c
97.7
85.6
78.7
80.5
69.2
74.7
58.8
69.7
61.1
64.4
—
12.3
14.4
21.3
19.5
30.8
25.3
41.2
30.3
38.9
35.6
—
1d
Table 3. Yields of 1a–d and 9a–d starting from 18a–d
1c
1
(%)
9 (%)
1d
1c
a
b
c
78.0
70.6
86.1
78.0
3.9
2.9
1.8
3.4
1d
1c
1d
d
a
1c
15, 16
17
a
1d
1c
—
10, 17
a
1d
64.9
35.1
a
Detected by GC/MS, not quantified.
2.4. Fvp of 7-chloro [1,2,4]triazolo[3,4-c][1,2,4]benzo-
triazine (1d)
Reactions of 1d were carried out between 460 and 560 8C
K2
K2
with pressures of w10 Torr and contact times of w10 s.
As in reactions of 1a–c, only one product was formed up to
5
20 8C, identified as 7-chloro[1,2,4]triazolo[4,3-b][1,2,4]-
benzotriazine (9d), at higher temperatures, 4-chlorobenzo-
nitrile (17) and benzonitrile (10) as well as coke were also
formed (Scheme 6, Table 2). Relative quantification of 1d
Scheme 7.
1
and 9d was performed by the help of H NMR spectra.
This finding that 9 can be formed from 1 under fvp
conditions raises the interesting question of a possible
mechanism for this transformation. Similar analoguous
valence bond isomerizations have been reported in the
literature. Four such cases for comparison should be
recalled here.
An example structurally related to the title compounds is the
behavior of bis[1,2,3]triazolopyrimidines where an equi-
librium between the angular (19) and the linear structure
(
21) is established through an open chain diazo form (20);
this reaction is described to take place in the presence of a
strong base. It is important to mention that the diazo form
was detected by spectral methods and the equilibrium
between the angular and the linear isomer is strongly
1
1
Scheme 6.
dependent on the substituents X and Y (Scheme 8). The
introduction of an additional nitrogen atom into the fused
azine ring facilitates triazole cleavage and potential
rearrangements. Thus, [1,2,3]triazolo[5,1-c][1,2,4]triazine
isomerized spontaneously to the thermodynamically more
Although, spectral analyses of the main products of these
FVP transformations suggested that derivatives of the the
linearly fused ring system [1,2,4]triazolo[3,4-c]benzo[1,2,4]
triazine (9a–d) were formed, preparative evidence for this
asignment seemed also desirable. To this end, we repeated
the ring closure reactions to 1a–d according to the described
1
2
stable 1,2,3-triazolo[1,5-b]triazine during its preparation.
A similar reaction was found in the pyrolysis of 2-(5-
tetrazolyl)pyridine (22), where the first nitrogen extrusion
2
7
literature procedure —that is, treatment of substituted

7492
W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
Scheme 8.
reaction afforded [1,2,3]triazolo[1,5-a]pyridine (23) in
equilibrium with 2-pyridyldiazomethane (24). At higher
temperatures [1,2,3]triazolo[1,5-a]pyridine (23) also
afforded nitrogen loss to form pyridylcarbene (25)
1
3
(
Scheme 9).
Scheme 11.
As formation of either 32 or 33 seemed rather unlikely,
more detailed considerations concerning the acting mecha-
nism involving reactive intermediates or concerted
reactions seemed straightforward as depicted in Scheme 13.
Route a involves diradicals with an imine group, isomeriza-
tion of these imines would lead to the final product by
cyclization, which is a known stereoisomeric isomerization
1
7
of imines. Pathway b is a concerted [1,5] sigmatropic shift
with rotation of a C–N bond partially doubly bonded and,
finally, route c is a [1,3] sigmatropic shift with a ring
contraction to form antiaromatic diazirine 34, too unstable
to be isolated, which by a concerted [1,3] shift (pathway d)
or by ring opening affording 32 and/or 33 would lead to the
final products.
Scheme 9.
The same kind of isomerization, through an open chain
compound, is described with several fused tetrazoles. Thus,
the existence of an equilibrium between two isomeric
tetrazolopyrimidones (26 and 28) through an azido
intermediate (27) as well as Arrhenius parameters has
Isomerizations of fused tetrazoles through azides afforded
positive entropies of activation (DS of C10.1 J deg
mol
1
4
#
K1
been reported (Scheme 10).
K1
#
and DH of C103.9 kJ mol ) (Scheme 10), with
K1
Angularly fused tetrazolo[5,1-c]benzo[1,2,4]triazines (29)
and the linearly fused isomers tetrazolo[1,5-b]benzo-
triazines (31) were reported to be in equilibrium with an
open chain azide (30) (Scheme 11) in DMSO, while 29 is
this reference for comparison, kinetic measurements of the
reactions described here would give valuable information
on the mechanism.
1
5
stable in the crystalline form. The relative amounts of 29,
0 and 31 are strongly dependent on solvent and on
substitution as was demonstrated in H NMR experiments.
2
.5. Kinetics
3
1
16
Reaction constants were measured at each temperature with
the relative concentrations of starting materials and products
determined by H NMR and averaged over at least three
1
It is important to note that in all these literature cases the
isomerisation of the angularly fused ring to the linearly
fused isomer (or the reverse reaction) proceeds via a well
defined and relatively stable intermediate (i.e., via a
heteroaromatic azide or diazomethane). An analysis of our
present results described here, however, seemed more
complicated because if the transformation of 1 to 9 proceeds
in a pathway analogous to these cited reactions, an N–C
bond cleavage would be anticipated to yield 32 and/or 33 as
possible intermediates (Scheme 12).
determinations. Reaction times were calculated as V /m
0
being V the volume of the reaction tube inside the hot zone
0
and m the carrier gas flow. Arrhenius parameters were
calculated by the classical equation (ln k vs 1/T) with data
for at least four different temperatures. To check the system,
kinetic parameters of ethylacetate pyrolysis were measured
in the fvp system and compared with those reported for a
static system, these results as well as a detailed description
1
8
of the methodology have already been described.
Scheme 10.

W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
7493
Scheme 12.
Scheme 13.
Reactions are first order, rate constants for reactions of
compounds 1a–d are depicted in Tables 4–7 and Arrhenius
parameters in Table 8.
Table 4. Kinetics of fvp of 1a
K2
t (10 ) s
K1
T (K)
C/C
0
k (s
)
a
a
b
b
723.15
748.15
773.15
798.15
0.953
0.906
0.825
0.748
2.2
2.1
2.0
2.0
2.2G0.2
4.7G0.2
9.2G0.2
14.7G0.6
From all this information, the most important data are the
negative entropies of activation, which preclude path a in
Scheme 13: this is a ring opening reaction that should have a
positive value and it is not expected for a ring closure
reaction of a diradical to be the rate determining step. Paths
b and c are concerted reactions where the better movement
of the atoms is in the supra-supra mode as they are all
bonded and a supra-antara situation should have steric
impediments. Path b is a concerted [1,5] sigmatropic shift,
although, this reaction is allowed by the Woodward-
a
Average of three determinations.
Average of four determinations.
b
Table 5. Kinetics of fvp of 1b
T (K) C/C
K2
t (10 ) s
K1
0
k (s
)
a
b
b
b
723.15
748.15
773.15
798.15
0.942
0.907
0.824
0.718
2.2
2.1
2.0
2.0
2.7G0.4
4.7G0.2
9.5G0.9
16.8G0.9
19
Hoffmann’s rules, it has strong steric impediments as the
C]N bond (Fig. 1) of triazole ring should rotate 908 to
allow a bonding interaction between all the atoms involved
in the transition state.
a
Average of three determinations.
Average of four determinations.
b
Table 6. Kinetics of fvp of 1c
T (K) C/C
The third possibility shown is Scheme 13 (path c) is a [1,3]
0
K2
t (10 ) s
K1
sigmatropic shift; according to Woodward–Hoffmann s
0
k (s
)
rules it should be a supra-antara process that has strong
steric impediment. But, it is possible to consider a supra-
supra transition state, with lower steric requeriments,
formally forbidden. This incompatibility can be over-
compensated if the Migrating Group (MG) and the
Migrating Framework (MF) differ in their electron
a
a
b
b
7
7
7
7
33.15
53.15
73.15
93.15
0.888
0.787
0.692
0.588
2.2
2.1
2.1
2.0
5.4G0.6
11.2G0.9
17.8G0.9
26.1G0.9
a
Averaged of three determinations.
Averaged of four determinations.
b
2
0
releasing–electron withdrawing ability. The difference
I KA (I Zionization potential of the donor partner; A Z
Table 7. Kinetics of fvp of 1d
T (K) C/C
D
A
D
A
K2
t (10 ) s
K1
electron affinity of the acceptor partner), affords an index of
the activation energy of a sigmatropic shift with the reaction
0
k (s
)
a
b
b
b
b
733.15
753.15
773.15
793.15
813.15
0.856
0.805
0.747
0.697
0.644
2.2
2.2
2.1
2.0
2.0
7.1G0.9
10.1G0.8
13.9G0.9
17.9G0.8
22.1G0.9
21
barrier becoming lower as I KAA decreases. These
arguments, supported by molecular orbital calculations,
D
22
were used in fvp isomerizations of isoxazoles, which have
similar entropies of activation values to the ones reported in
this article suggesting a common reaction pathway.
Recently, some theoretical calculations carried out by
a
Averaged of three determinations.
Averaged of four determinations.
b

7494
W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
Table 8. Arrhenius parameters of fvp of 1a–d and of some isoxazoles*
K1
k (s ) (500 8C)
K1
K1
#
DS (eu)
#
DH (kJ mol
K1
#
DG (kJ mol
K1
E
a
(kJ mol
)
log A (s
)
)
)
r
1
1
1
1
3
3
a
b
c
9.2G0.2
9.5G0.9
17.8G0.9
13.9G0.9
123G3
118G3
125G4
71G2
108G1
109G1
9.2G0.5
9.0G0.3
9.7G0.8
5.9G0.2
9.00
K18.2G0.4
K19.4G0.3
K16.0G0.8
K33.5G0.2
116G3
112G3
119G4
64G2
176G4
176G4
172G8
172G4
K0.997
K0.998
K0.992
K0.997
d
a
5
6
b
9.00
a
From Ref. 17.
From Ref. 22.
b
diazirine ring opens, as a concerted reaction (way d) or
through an open chain isomer (way e).
The other point that was checked is the reversibility of the
reaction: if the angular isomer is also formed from the linear
one under the same fvp conditions. In order to study this
possibility, fvp reactions of 9b were carried out at 480 8C.
As expected, 1b (26.3%) was found to be the reaction
product as well as unreacted 9b (73.7%), confirming the
reversibility of the reaction. These results also show that the
system is not in equilibrium, probably because this is a flow
system with short contact times, since the yields starting
from 9a and from 9b are different. These results may bring
clarity to the whole mechanism taking into account the
microscopic reversibility principle. Thus, it was demon-
strated that the angular compounds form the diazirine with
negative entropies of activation. If the linear compounds
also afford isomerization to the angular one, the reaction
should proceed through the diazirine. If this is so, the open
chain intermediates (32 and 33) postulated in path c–e
(Scheme 13) cannot be present, as the angular compounds
can be formed directly from these intermediates without
cyclization to the diazirine. With this in mind, path c–d
seems more possible (Scheme 15, Fig. 2) as the actual
mechanism in accordance with the microscopic reversibility
principle. As the linear isomer has an unfavorable ortho
quinoid arrangement it is expected to be less stable than the
linear isomer (as naphthalene and phenanthrene), so the
qualitative energy profile of the reaction in Figure 2 could
be a suitable representation of these reactions.
Scheme 14.
Figure 1. Steric representation of [1,5]sigmatropic shift of A via transition
state B.
2
Davico supported the mechanism proposed by some of us
2
1
8,23
2
0 years ago on a kinetic basis.
The author demon-
strated that the rate limiting step is a concerted ring closure
to the isomeric azirine in a similar reaction as described
here. Arrhenius parameters for fvp isomerization reactions
of 5-amino-3,4-dimethylisoxazole (35) and 5-amino-4-
methylisoxazole (36) (Scheme 14) were compared with
the ones of 1a–d and reported in Table 8.
1
8
2
3
It should be mentioned that some of the azirines obtained
from isoxazoles are stable enough to be isolated or detected
by spectroscopy, which is not the case of the diazirines
proposed as intermediates in the reactions here reported.
These diazirines are antiaromatic and this is the reason why
they are not detected nor isolated, similar antiaromatic
compounds, such as the 1H-azirines were proposed as
intermediates in reactions of N-substituted 1,2,3-triazoles
The kinetic expression for this reaction (assuming stationary
state for B) is:
k Z k CðkK1kK2=k2Þ
(1)
1
2
4
and benzotriazoles and in reactions of 1-phthalimido-
2
5
1
,2,3-triazoles.
An analysis of the relative expected values of the different
rate constants show that k should be larger than k as the
2
K1
With the kinetic evidence reported here (large negative
entropies of activation), it is possible to propose path c of
Scheme 13 as the one taking place in these reactions. As
formation of the diazirine is the rate determining step, with
these results it is not possible to distinguish how the
reaction goes to products; kK2 should have a smaller value
than k but larger than k as this reaction starts from the
thermodynamically less stable isomer. These assumptions
make the second term of the rate constant expression (Eq. 1)
2
1
negligible, so the measured rate constant kobswk , which is
1
Scheme 15.

W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
7495
at higher temperatures. These results are not surprising and
were also found in reactions of 2-substituted 1,2,4-
benzotriazines described in a previous paper. It is common
4
for aromatic hydrocarbons to lose a substituent by radical
reactions, so products from radical combination are also
formed as well as coke.
The striking point is that in reactions of 1b and 1c the cyano
group is maintained while it is lost in reactions of 1c. In
order to rationalise this finding, some calculations on bond
length have been carried out by using Hypercheme
programme [A semi-empirical method (AM1) with algor-
ithm Polak-Ribiere (Conjugate Gradient)]. Results showed
˚
that the C–CN bond is of 1.42 A for 10, 11, 17, and
p-methoxybenzonitrile while the C–CH in 11 bond is of
3
˚ ˚
.48 A, the C–Cl bond in 17 is of 1.70 A and the C–OCH in
3
1
Figure 2. Qualitative representation of reaction coordinate for reactions of
a–d. A: initial angularly fused triazole 1, B: azirine intermediate, C:
linearly fused triazole product.
˚
1
p-methoxybenzonitrile is of 1.38 A. These data show that
C–CN is the weaker bond in p-methoxybenzonitrile and this
may be the reason why anisole is the product formed by the
radical scission reaction.
in accordance with the negative entropy of activation for a
ring contraction reaction.
Valuable conclusions can be drawn by analysis of the effect
of substituents reflected in the Arrhenius parameters. In
NMR studies of the equilibrium between the angular and the
linear structures of tetrazolo-benzo-as-triazines through the
isomeric azide (Scheme 11) the authors demonstrated that
substituents attached to C7 have a strong influence on the
3
. Conclusions
Reactions described here support that the isomerization
equilibrium between [1,2,4]triazolo[3,4-c][1,2,4]benzo-
triazines (1) and [1,2,4]triazolo[4,3-b][1,2,4]benzotriazines
(
9) goes through antiaromatic diazirines. Formation of these
1
6,26
position of the ternary equilibrium.
Looking at the data
intermediates is the rate-limiting step with large negative
entropies of activation. It was proved that electron-donating
substituents diminish the concerted character reflected in a
less negative entropy of activation and larger reaction rate
constant. All the studied compounds have almost the same
free energy of activation values, which suggests that all of
them react by the same mechanism. Reactions described
here show that the mechanism of isomerization is different
from those described for similar compounds with 1,2,3-
triazoles and tetrazoles where open chain intermediates
were isolated or detected by spectroscopy.
reported in Table 8, it is clear that all the free energies of
activation are almost the same, but, the entropic and
enthalpic data are quite different, as well as reaction
constants measured at the same temperature. The more
electron donating the substituent (methoxy in this case) the
faster the reaction. Chlorine has a lower rate constant than
methoxy but faster than hydrogen and methyl. Hydrogen
and methyl have almost the same values within the
experimental errors. Thus, the rate constant values follow
the electronic effect of the substituents while the entropies
of activation have a different sequence. The large negative
value for chlorine can be interpreted as a more concerted
character of the reaction probably due to a combination of
inductive (electron withdrawing) and electronic (electron
donor) effects or an isokinetic relation, DH* is very low, as
all compounds have similar DG* confirming that they react
by the same mechanism.
4. Experimental
4
.1. General
Flash vacuum pyrolysis reactions were carried out in a
vycor glass reactor using a Thermolyne 21100 tube furnace
with a temperature controller device. Oxygen-free dry
nitrogen was used as carrier gas. Samples to be pyrolyzed
Finally, with the above described results it can be concluded
that these reactions could be included in types I or III AX of
0
K2
Epioti s classification. Theoretical calculations should be
were of w40 mg. Contact times were around 10 s and
made to determine, which are the donor and the acceptor
partners with the lower energy difference for HOMO–
LUMO interaction.
pressures of 0.02 Torr were used. Products were trapped at
the liquid air temperature, extracted with solvent and
submitted to different analysis or separation techniques.
Gas chromatography/mass spectrometry (GC/MS) analysis
were performed with a SE-30 column, using helium as
eluent at a flow rate of 1 mL/min, the heating rate was
different for each compound or mixture of compounds.
Mass spectra were obtained in the electron impact mode
Some additional remarks on formation of aromatic
compunds at higher temperatures should be made concern-
ing the analysis of these reactions. As it was stated above, 1a
afforded benzonitrile (10) as well as 9a over 550 8C; 1b
afforded p-methylbenzonitrile (11), dicyanobenzenes
1
13
(EI) using 70 eV as ionization energy. H NMR and
NMR spectra were carried out in CDCl with a Bruker 200
C
(
12–14) and 9b over 550 8C; 1c afforded anisole (15),
3
benzene (16) and 9c over 520 8C and 1d afforded
p-chlorobenzonitrile (17), benzonitrile (10) and 9d. In all
of these reactions also, a carbonaceous residue was obtained
FT spectrometer (at 200 MHz) and in DMSO with Varian
UNITY INOVA spectrometer (200 and 400 MHz for H and
1
1
3
100 MHz for C). Chemical shifts are reported in ppm

7496
W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
downfield from TMS. The IR spectra were recorded with a
Thermo Nicolet AVATAR 320 FT-IR spectrophotometer.
Column and thin layer chromatographies were performed
on silica gel. Solvents were analytical grade. Recovery of
material was O90% in all fvp experiments. Melting Points
were determined by a B u¨ chi apparatus and are uncorrected.
all cases which was separated, evaporated, and the residue
recrystallized from the given solvent.
4.3.1. [1,2,4]Triazolo[4,3-b]benzo[1,2,4]triazine (9a).
Synthesis of 1a
2
7
was carried out starting from
2
6
3-hydrazinobenzo[1,2,4]triazine (18a, 0.806 g, 5 mmol).
Work up of the reaction mixture (column chromatography
with eluent hexane/ethyl acetate 3:7) yielded, besides 1a, 9a
as brown crystals (0.033 g, 3.86%): mp 207–209 8C; IR
(KBr) nmax 3091, 1505, 1381, 1290, 1140, 1023, 932, 783,
4
.2. Synthesis of starting materials
Compounds 1a and 1d were prepared by a previously
described procedure.
2
7
K1
1
760, 729, 708, 607, 421 cm
00 MHz) d 7.74–7.86 (m, 2H, H-7, H-8), 7.86 (d, 1H,
JZ9.5 Hz, H-9), 7.92 (d, 1H, JZ9.5 Hz, H-6), 10.08 (s, 1H,
;
H NMR (DMSO-d6
4
4
(
.2.1. 7-Methyl[1,2,4]triazolo[3,4-c]benzo[1,2,4]triazine
1b). This compound was prepared by the same procedure
as described for 1a and 1d starting from 3-hydrazino-7-
1
3
H-3); C NMR (DMSO-d
) d 127.4 (C-9), 129.6 (C-8),
133.7 (C-7), 136.0 (C-6), 137.3 (C-3), 142.0 (C-5a), 145.2
6
2
6
methyl[1,2,4]benzotriazine (1.0 g, 5.7 mmol). It was
mixed with triethyl orthoformate (10.0 g, 11.2 mL,
(C-9a), 147.4 (C-10a); HRMS calcd for C
found 171.0544. Anal. Calcd for C
8
5
H
5
N
(171.1625): C,
5
171.0545,
H
N
8
5
6
1
7.5 mmol) and refluxed for 10 h (oil bath temperature
30–140 8C). The reaction mixture was cooled down to
56.14; H, 2.94; N, 40.92. Found C, 56.50; H, 2.76; N, 40.62.
room temperature and the precipitated solid was filtered off.
The crude product was recrystallized from DMF to give the
product (0.746 g, 70.6%): mp 289–290 8C; IR (KBr) nmax
4.3.2. 7-Methyl[1,2,4]triazolo[4,3-b]benzo[1,2,4]triazine
(9b). Synthesis of 1b was carried out starting from
2
6
7-methyl-3-hydrazinobenzo[1,2,4]triazine (18b, 0.876 g,
5 mmol). Work up of the reaction mixture yielded, besides
1b, 9b as ocher crystals (0.026 g, 2.89%): mp 240–241 8C;
UV l (nm) log 3 in acetonitrile: 352 (3.645), 444 (3.379); IR
(KBr) nmax 3109, 1540, 1373, 1296, 1150, 1029, 1009, 939,
3
107, 1579, 1524, 1477, 1360, 1316, 1216, 1227, 1191,
K1
1124, 1030, 957, 825, 772, 662, 636, 600, 552, 424 cm
H NMR (DMSO-d 400 MHz) d 2.6 (s, 3H, H–CH ), 7.95
;
1
6
3
(
8
dd, 1H, JZ2.0, 8.4 Hz, H-8), 8.34 (d, 1H, JZ8.4 Hz, H-9),
.48 (d, 1H, JZ2.0 Hz, H-6), 10.5 (s, 1H, H-1); C NMR
1
3
K1
1
814, 728, 709, 659, 567, 424 cm
200 MHz) d 2.58 (s, 3H, H–CH ), 7.60 (s, 1H, H-6), 7.59 (d,
; H NMR (CDCl3
(
DMSO-d ) d 21.3 (C–CH ), 117.0 (C-9), 120.8 (C-9a),
6
3
3
1
(
31.4 (C-6), 135.7 (C-1), 138.0 (C-8), 138.4 (C-5a), 139.2
C-7), 153.5 (C-3a). Anal. Calcd for C H N (185.19): C,
1H, JZ9.5 Hz, H-8), 7.84 (d, 1H, JZ9.5 Hz, H-9), 9.43 (s,
3 3
1
1H, H-3); C NMR (CDCl ) d 22.52 (C–CH ), 123.7 (C-6),
3
9
7
5
5
8.37; H, 3.81; N, 37.82. Found: C, 58.36; H, 3.67; N, 37.52.
129.2 (C-9), 135.9 (C-8), 138.5 (C-7), 138.9 (C-3), 142.0
C-9a), 144.3 (C-5a), 147.2 (C-10a); HRMS calcd for
185.0701, found 185.0696. Anal. Calcd for C H N
9 7 5
(
C
4
.2.2. 7-Methoxy[1,2,4]triazolo[3,4-c]benzo[1,2,4]tri-
9
H
7
N
5
azine (1c). This compound was prepared by the same
procedure as described for 1a and 1d starting from
(185.19): C, 58.37; H, 3.81; N, 37.82. Found C, 58.26; H,
3.75; N, 37.48.
2
6
3
5
-hydrazino-7-methoxy[1,2,4]benzotriazine
.2 mmol). It was mixed with triethyl orthoformate
(1.0 g,
4.3.3. 7-Methoxy[1,2,4]triazolo[4,3-b]benzo[1,2,4]tri-
azine (9c). Synthesis of 1c was carried out starting from
(
10.0 g, 11.2 mL, 67.5 mmol) and refluxed for 10 h (oil
2
6
bath temperature 130–140 8C). The reaction mixture was
cooled down to room temperature and the precipitated solid
was filtered off. The crude product was recrystallized from
DMF to give the product (0.905 g, 86.1%): mp 298–299 8C;
IR (KBr) nmax 3111, 1616, 1582, 1520, 1458, 1389, 1364,
7-methoxy-3-hydrazinobenzo[1,2,4]triazine
(18c,
0.956 g, 5 mmol). Work up of the reaction mixture yielded,
besides 1c, 9c as yellow crystals (0.0179 g, 1.77%): mp
216–218 8C; UV l (nm) log 3 in acetonitrile: 368 (3.564),
439 (2.992); IR (KBr) nmax 3094, 1626, 1548, 1535, 1450,
1428, 1383, 1300, 1249, 1213, 1180, 1132, 1034, 998, 944,
1
6
327, 1253, 1156, 1116, 1038, 1009, 945, 847, 770, 650,
31, 529 cm
K1
1
K1 1
;
H NMR (DMSO-d6 400 MHz) d 4.0
828, 738, 709, 660, 622, 557, 406 cm ; H NMR (CDCl
3
(
(
(
(
s, 3H, H–OCH ), 7.77 (dd, 1H, JZ9.0, 2.8 Hz, H-8), 8.18
400 MHz) d 4.00 (s, 3H, H–OCH ), 6.90 (s, 1H, H-6), 7.44
3
3
d, 1H, JZ2.8 Hz, H-6), 8.42 (d, 1H, JZ9.0 Hz, H-9), 10.06
(d, 1H, JZ9.9 Hz, H-8), 7.80 (d, 1H, JZ9.9 Hz, H-9), 9.34
1
s, 1H, H-1); C NMR (DMSO-d ) d 57.1 (C–MeO), 111.9
3
13
(s, 1H, H-3); C NMR (CDCl ) d 56.6 (C–OCH ), 99.2 (C-6),
6
3
3
C-6), 117.7 (C-9a), 118 (C-9), 126.2 (C-8), 136.2 (C-1),
39.3 (C-5a), 153.7 (C-3a), 159.6 (C-7). Anal. Calcd for
130.78 (C-9), 132.8 (C-8), 135.9 (C-3), 143.4 (C-9a), 143.5
(C-5a), 147.2 (C-7), 162.6 (C-10a); HRMS calcd for
C H N O 201.065, found 201.0642. Anal. Calcd for
1
C H N O (201.18): C, 53.73; H, 3.51; N, 34.81. Found: C,
9
7
5
9
7 5
5
3.69; H, 3.27; N, 34.41.
C H N O (201.18): C, 53.73; H, 3.51; N, 34.81. Found C,
9 7 5
5
3.47; H, 3.35; N, 34.70.
4
[
.3. General procedure for isolation of [1,2,4]triazolo-
4,3-b]benzo[1,2,4]triazines (9a–d)
4.3.4. 7-Chloro[1,2,4]triazolo[4,3-b]benzo[1,2,4]triazine
9d). Synthesis of 1d was carried out starting from
(
2
6
Reaction mixtures obtained with the synthesis of 1a–d
starting from 3-hydrazinobenzo[1,2,4]triazines (18a–d,
7-chloro-3-hydrazinobenzo[1,2,4]triazine (18d, 0.978 g,
5 mmol). Work up of the reaction mixture yielded, besides
1d, 9d as yellow crystals (0.035 g, 3.40%): mp 295–299 8C;
UV l (nm) log 3 in acetonitrile: 354 (3.672), 444 (3.352); IR
(KBr) nmax 3122, 3066, 1611, 1505, 1435, 1296, 1248, 1149,
1056, 1022, 935, 877, 834, 799, 712, 615, 579, 522,
5
(
mmol) were worked up as follows. The main product
1) was removed by filtration, the mother liquor was
evaporated and the residue was subjected to column
chromatography on silica. Besides the main product a well
defined yellow fraction with a higher R value appeared in
K1
1
423 cm ; H NMR (CDCl 200 MHz) d 7.68 (d, 1H, JZ
3
f

W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
7497
1
9
H-9), 9.49 (s, 1H, H-3); C NMR (CDCl ) d 118.3 (C-6),
.7 Hz, H-8), 7.91 (s, 1H, H-6), 7.94 (d, 1H, JZ9.7 Hz,
H NMR. Data for at least three reactions were averaged for
the kinetic measurements.
1
3
3
1
24.4 (C-9), 131.1 (C-8), 131.7 (C-7), 136.2 (C-9a), 136.6
C-5a), 140.0 (C-3), 141.6 (C-10a); HRMS calcd for
(
C H N Cl ( Cl) 205.01552, found 205.01550. Anal.
Methoxybenzene (anisole) (15). Mass Spectrum: more than
90% match with NIST database. CAS No. 100-66-3.
3
5
8
4
5
Calcd for C H N Cl (205.60): C, 46.73; H, 1.96, N,
4 5
8
3
4.06. Found C, 46.76, H, 1.96, N, 33.76.
Benzene (16). Mass Spectrum: more than 90% match with
NIST database. CAS No. 71-43-2.
4
.4. Flash vacuum pyrolysis of 1a
4.7. Flash vacuum pyrolysis of 1d
Reactions of 1a were carried out between 450 and 600 8C,
with pressures of w10 Torr and contact times of w10 s.
The reaction was clean affording a single product up to
K2
K2
Reactions of 1d were carried out between 460 and 560 8C
with pressures of w10 Torr and contact times of w10 s.
As in reactions of 1a–c only one product was formed up to
–
2
K2
5
25 8C identified as [1,2,4]triazolo[4,3-b][1,2,4]benzo-
triazine (9a) (Scheme 6). Over 550 8C small amounts of
0 and a carbonaceous residue were formed from
520 8C, identified as 7-chloro[1,2,4]triazolo[4,3-b][1,2,4]-
benzotriazine (9d), at higher temperatures 4-chlorobenzo-
nitrile (17) and 10 as well as coke were also formed
1
decomposition. The relative amounts of 1a and 9a were
calculated by H NMR. The structure of 9a was confirmed
1
(Scheme 10). The structure of 9d was confirmed by
comparison with authentic sample prepared from 18d.
by comparison with authentic sample prepared from 18a.
Data for at least three reactions were averaged for the
kinetic measurements.
1
Relative quantification of 1d and 9d was performed by H
NMR. Data for at least three reactions were averaged for the
kinetic measurements.
Benzonitrile (10). Mass Spectrum: more than 90% match
with NIST database. CAS No. 100-47-0.
4-Chlorobenzonitrile (17). Mass Spectrum: more than 90%
match with NIST database. CAS No. 623-03-0.
4
.5. Flash vacuum pyrolysis of 1b
Acknowledgements
Reactions of 1b were carried out between 450 and 600 8C,
with pressures of w10 Torr and contact times of w10 s.
The reaction was clean affording a single product up to
K2
K2
Thanks are due to financial support provided by OTKA T
47317, NKFP MediChem2, COST B16 as well as the
5
25 8C identified as [1,2,4]triazolo[4,3-b][1,2,4]benzo-
triazine (9b) (Scheme 8). Over 550 8C small amounts of
0, p-methylbenzonitrile (11) and a mixture of dicyano-
benzenes (12–14) were also formed from decomposition of
b and/or 9b (Scheme 8). The structure of 9b was confirmed
QLK2-CT-2002-90436 project funded by the European
Union for Center of Excellence in Biomolecular Chemistry.
The Argentinean group thanks for financial support from
CONICET, FONCyT, SECyT-UNC Fundaci o´ n Antorchas
and TWAS.
1
1
by comparison with authentic sample prepared from 18b.
Relative quantification of 1b and 9b was performed by H
1
NMR. It is important to mention that 100% conversion
could not be achieved and that over 600 8C coke was also
formed. Compounds 10 and 12–14 are formed from
pyrolysis of 11. Data for at least three reactions were
averaged for the kinetic measurements.
References and notes
1
. Yranzo, G. I.; Moyano, E. L. Curr. Org. Chem. 2004, 8,
071–1088.
. See for example: (a) Seybold, G. Angew. Chem., Int. Ed. Engl.
977, 16, 365–373. (b) Brown, R. F. C. In Pyrolytic Methods in
1
p-Methylbenzonitrile (11). Mass Spectrum: more than 90%
match with NIST database. CAS No. 100-47-0.
2
1
Organic Chemistry; Wasserman, H., Ed.; Academic: New
York, 1980; pp 141–148. (c) Banciu, M. D.; Popescu, A.;
Simion, A.; Draghici, C.; Mangra, C.; Mihaiescu, D.; Pocol,
M. J. Anal. Appl. Pyrol. 1999, 48, 129–146.
Dicyanobenzenes (12–14). Mass Spectrum: (1,2) More than
9
Spectrum: (1,3) More than 90% match with NIST database.
CAS No. 626-17-5. Mass Spectrum: (1,4) more than 90%
match with NIST database. CAS No. 623-26-7.
0% match with NIST database CAS No. 91-15-6. Mass
3
. (a) Adger, B. M.; Keating, M.; Rees, C.; Storr, R. C. J. Chem.
Soc., Perkin Trans. 1 1975, 41–45. (b) Adger, B. M.; Rees, C.;
Storr, R. C. J. Chem. Soc., Perkin Trans. 1 1975, 45–52.
(c) Westiuk, N. H.; Roy, C. D.; Ma, J. Can. J. Chem. 1995, 73,
4
.6. Flash vacuum pyrolysis of 1c
1
46–149.
Reactions of 1c were carried out between 440 and 540 8C,
4. Riedl, Z.; Haj o´ s, Gy.; Pel a´ ez, W. J.; Gafarova, I. T.; Moyano,
E. L.; Yranzo, G. I. Tetrahedron 2003, 59, 851–856.
5. Maquestiau, A.; Flammang-Barbieux, M.; Flammang, R.;
Chen, L. Z. Bull. Soc. Chim. Belg. 1988, 97, 245–254.
6. Carlsen, P. J.; Gautun, R. Acta Chem. Scand. 1992, 46,
469–473.
K2
K2
with pressures of w10 Torr and contact times of w10 s.
Up to 520 8C the only product was 7-methoxy-[1,2,4]tri-
azolo[4,3-b][1,2,4]benzotriazine (9c), over this temperature
anisole (15), benzene (18) and coke were also formed
(
Scheme 9). The structure of 9c was confirmed by
comparison with authentic sample prepared from 18c.
Relative quantification of 1c and 9c were performed by
7. Maquestiau, A.; Puk, E.; Flammang, R. Bull. Soc. Chim. Belg.
1987, 96, 181–186.

7498
W. J. Pel a´ ez et al. / Tetrahedron 61 (2005) 7489–7498
8
9
. P e´ rez, J. D.; Yranzo, G. I. J. Anal. Appl. Pyrol. 1989, 16,
65–172.
. (a) R u¨ chardt, C.; Meier, M.; Haaf, K.; Pakusch, J.; Wolber,
D. R.; Campbell, R. M. J. Chem. Soc., Perkin Trans. 2 1976,
1501–1506.
18. P e´ rez, J. D.; de D ´ı az, R. G.; Yranzo, G. I. J. Org. Chem. 1981,
46, 3505–3508.
1
E. K. A.; M u¨ ller, B. Angew. Chem., Int. Ed. Engl. 1991, 30,
893–901. (b) McNab, H. Contemp. Org. Synth. 1996, 3,
373–396.
19. Woodward, R. B.; Hoffmann, R. The Conservation of Orbital
Symmetry; Verlag Chemie: Weinhein, 1970.
1
1
0. (a) Boyer, J. H.; Selvarajan, R. J. Am. Chem. Soc. 1969, 91,
122–6126. (b) Wiersum, U. E. J. Royal Netherlands Chem.
Soc. 1982, 101–317.
20. (a) Epiotis, N. D. J. Am. Chem. Soc. 1972, 94, 1924–1934.
(b) Epiotis, N. D. J. Am. Chem. Soc. 1973, 95, 1191–1200.
21. Epiotis, N. D. Theory of Organic Reactions; Springer: New
York, 1978.
6
1. L’Abb e´ , G.; Godts, F.; Toppet, S. Bull. Soc. Chim. Belg. 1986,
9
5, 679–680.
22. Davico, G. E. J. Phys. Org. Chem. 2005, 18, 434–440.
23. P e´ rez, J. D.; Yranzo, G. I.; Wunderlin, D. A. J. Org. Chem.
1982, 47, 982–984.
1
1
1
2. L’Abb e´ , G. Bull. Soc. Chim. Belg. 1987, 99, 281–291.
3. Wentrup, C. Tetrahedron 1974, 30, 1301–1311.
4. Temple, C., Jr.; Coburn, W. C., Jr.; Torpe, M.; Montgomery,
J. A. J. Org. Chem. 1965, 30, 2395–2398.
5. Messmer, A.; Haj o´ s, Gy.; Tam a´ s, A.; Neszm e´ lyi, A. J. Org.
Chem. 1979, 44, 1823–1825.
24. (a) Gilchrist, T. L.; Gymer, G. E.; Rees, C. W. J. Chem. Soc.,
Perkin Trans. 1 1975, 1–8. (b) Gilchrist, T. L.; Rees, C. W.;
Thomas, C. J. Chem. Soc., Perkin Trans. 1 1975, 8–11.
25. Gilchrist, T. L.; Gymer, G. E.; Rees, C. W. J. Chem. Soc.,
Perkin Trans. 1 1973, 555–561.
26. Messmer, A.; Haj o´ s, Gy.; Benk o´ , P.; Pallos, L. Acta Chim.
Hung. 1980, 105, 123–133.
27. Messmer, A.; Haj o´ s, Gy.; Benk o´ , P.; Pallos, L. Acta Chim.
Hung. 1980, 105, 189–199.
1
1
1
6. Castill o´ n, S.; Mel e´ ndez, E.; Pascual, C.; Vilarrasa, J. J. Org.
Chem. 1982, 47, 3886–3890.
7. (a) Bjørgo, J.; Boyd, D. R.; Jennings, W. B.; Muckett,
P. M.; Warning, L. C. J. Org. Chem. 1977, 42,
3
700–3704. (b) Jennings, W. B.; Al-ShowimN, S.; Boyd,