Nitrile imines and nitrile ylides: Rearrangements of benzonitrile N-methylimine and benzonitrile dimethylmethylide to azabutadienes, carbodiimides, and ketenimines. chemical activation in thermolysis of azirenes, tetrazoles, oxazolones, isoxazolones, and oxadiazolones

Article abstract of DOI:10.1021/jo402667y

Flash vacuum thermolysis (FVT) of 1-methyl-5-phenyltetrazole (5b), 2-methyl-5-phenyltetrazole (1b), and 3-methyl-5-phenyl-1,3,4-oxadiazol-2(3H)-one (3b) affords the nitrile imine (2b), which rearranges in part to N-methyl-N′-phenylcarbodiimide (7b). Another part of 2b undergoes a 1,4-H shift to the diazabutadiene (13). 13 undergoes two chemically activated decompositions, to benzonitrile and CH₂=NH and to styrene and N ₂. FVT of 2,2-dimethyl-4-phenyl-oxazol-5(2H)-one (16) at 400 C yields 3-methyl-1-phenyl-2-azabutadiene (18) in high yield. In contrast, FVT of 3,3-dimethyl-2-phenyl-1-azirene (21) at 600 C or 4,4-dimethyl-3-phenyl- isoxazolone (20) at 600 C affords only a low yield of azabutadiene (18) due to chemically activated decomposition of 18 to styrene and acetonitrile. There are two reaction paths from azirene (21): one (path a) leading to nitrile ylide (17) and the major products styrene and acetonitrile and the other (path b) leading to the vinylnitrene (22) and ketenimine (23). The nitrile ylide PhC ^-=N⁺=C(CH₃)₂ (17) is implicated as the immediate precursor of azabutadiene (18). FVT of either 3-phenylisoxazol- 5(4H)one (25) or 2-phenylazirene (26) at 600 C affords N-phenylketenimine (28). The nitrile ylide PhC^-=N⁺=CH₂ (30) is postulated as a reversibly formed intermediate. N-Phenylketenimine (28) undergoes chemically activated free radical rearrangement to benzyl cyanide. The mechanistic interpretations are supported by calculations of the energies of key intermediates and transition states.

Full text of DOI:10.1021/jo402667y

Article
pubs.acs.org/joc
Nitrile Imines and Nitrile Ylides: Rearrangements of Benzonitrile
N‑Methylimine and Benzonitrile Dimethylmethylide to
Azabutadienes, Carbodiimides, and Ketenimines. Chemical
Activation in Thermolysis of Azirenes, Tetrazoles, Oxazolones,
Isoxazolones, and Oxadiazolones
Didier Begue,† Alain Dargelos,† Hans M. Berstermann,‡ Klaus P. Netsch,‡ Pawel Bednarek,§
́
́
and Curt Wentrup*,§
†Institut des Sciences Analytiques et de Physico-Chimie pour l’Environnement et les Mater
Universite de Pau et des Pays de l′Adour, 64000 Pau, France
́
iaux, Equipe Chimie Physique, UMR 5254,
́
‡Fachbereich Chemie der Philipps−Universitat Marburg, 35037 Marburg, Germany
̈
§School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, Queensland 4072, Australia
S
* Supporting Information
ABSTRACT: Flash vacuum thermolysis (FVT) of 1-methyl-5-phenyl-
tetrazole (5b), 2-methyl-5-phenyltetrazole (1b), and 3-methyl-5-phenyl-
1,3,4-oxadiazol-2(3H)-one (3b) affords the nitrile imine (2b), which
rearranges in part to N-methyl-N′-phenylcarbodiimide (7b). Another part
of 2b undergoes a 1,4-H shift to the diazabutadiene (13). 13 undergoes
two chemically activated decompositions, to benzonitrile and CH2NH
and to styrene and N2. FVT of 2,2-dimethyl-4-phenyl-oxazol-5(2H)-one
(16) at 400 °C yields 3-methyl-1-phenyl-2-azabutadiene (18) in high yield. In contrast, FVT of 3,3-dimethyl-2-phenyl-1-azirene
(21) at 600 °C or 4,4-dimethyl-3-phenyl-isoxazolone (20) at 600 °C affords only a low yield of azabutadiene (18) due to
chemically activated decomposition of 18 to styrene and acetonitrile. There are two reaction paths from azirene (21): one (path
a) leading to nitrile ylide (17) and the major products styrene and acetonitrile and the other (path b) leading to the vinylnitrene
(22) and ketenimine (23). The nitrile ylide PhC−N+C(CH3)2 (17) is implicated as the immediate precursor of
azabutadiene (18). FVT of either 3-phenylisoxazol-5(4H)one (25) or 2-phenylazirene (26) at 600 °C affords N-
phenylketenimine (28). The nitrile ylide PhC−N+CH2 (30) is postulated as a reversibly formed intermediate. N-
Phenylketenimine (28) undergoes chemically activated free radical rearrangement to benzyl cyanide. The mechanistic
interpretations are supported by calculations of the energies of key intermediates and transition states.
propargylic form PhCN+N−H is a transition state. The
INTRODUCTION
■
nitrile imine rearranges to N-phenylcarbodiimide (7a) on
Nitrile imines (nitrilimines) and nitrile ylides are important
synthetic intermediates,1 and their structures and reactivities are
of ongoing, fundamental interest.2 Recently, we described the
formation and direct spectroscopic observation of several nitrile
imines, including 2a−d, formed by matrix photolysis and/or
flash vacuum thermolysis (FVT) of 2,5-disubstituted tetrazoles
(1a−d) as well as the oxadiazolone (3a) (Scheme 1).3
Moreover, both thermal and photochemical rearrangements
of nitrile imines to carbodiimides (7) were observed,2b,3−5 and
this is believed to take place via the 1H-diazirene (4) and the
imidoylnitrene (6). 1H-Diazirenes are little-known compounds,
but Fausto and co-workers have reported the direct observation
of a few such species in low-temperature matrixes.6
Benzonitrile imine (2a) has been generated in an Ar matrix
by photolysis of 5-phenyltetrazole (1a) at 254 nm.3 The
fundamental, antisymmetric IR stretching vibration of the
cumulene moiety at 2073 cm−1 is in accord with the allenic
structure of 2a shown in Scheme 1, whereas the corresponding
further photolysis as well as on FVT at 700−800 °C.
Carbodiimide (7a) is also formed by matrix photolysis of 3-
phenylsydnone, and the energetics involved in the intercon-
versions 2a ⇄ 4a ⇄ 6a ⇄ 7a have been evaluated at the
B3LYP/6-31+G** and QCISD(T) levels of theory.7 The
carbodiimide lies ca. 48 kcal/mol below the nitrile imine, the
barrier for cyclization of the nitrile imine (2a) to diazirene (4a)
is also ca. 48 kcal/mol, and the onward barrier to carbodimide
(7) is ca. 9 kcal/mol. Barriers of this magnitude are readily
accessible under FVT conditions.
Benzonitrile N-methylimine (2b) has been generated in a
similar manner by photolysis of 2-methyl-5-phenyltetrazole
(1b) in Ar matrix.3 The IR stretching vibration at 2032 cm−1
again indicates that 2b exists in the allenic form shown in
Received: December 2, 2013
Published: January 7, 2014
© 2014 American Chemical Society
1247
dx.doi.org/10.1021/jo402667y | J. Org. Chem. 2014, 79, 1247−1253

The Journal of Organic Chemistry
Article
corresponding phenyl(silyl) nitrile imine (12b) was obtained
readily on both FVT and photolysis of 10b (Scheme 3).3
Nevertheless, as we will show, the presumptive mechanism
presented in Scheme 2 is not responsible for the formation of
benzonitrile from 1b.
Scheme 1. Thermal and Photochemical Formation of Nitrile
Imines
In this paper, we wish to report a detailed study of the
thermal rearrangements of nitrile imines and nitrile ylides.
RESULTS AND DISCUSSION
■
1. C-Phenyl-N-methyl Nitrile Imine (2b) and Methyl-
(phenyl)carbodiimide (7b). Preparative FVT of 1-methyl-5-
phenyltetrazole (5b) at 600 °C (5 × 10−4 hPa) afforded an 85%
yield of methyl(phenyl)carbodiimide (7b), which has been
completely characterized spectroscopically;4 it is amenable to
gas chromatographic assay but polymerizes rapidly at room
temperature.
Scheme 1, and the corresponding propargylic form PhC
N+N−Me is a transition state. The nitrile imine rearranges
rapidly to N-methyl-N′-phenylcarbodiimide (7b) on photolysis
at 254 nm. FVT of either 1b (420−600 °C) or 1-methyl-5-
phenyltetrazole (5b) at 430−600 °C with Ar matrix isolation of
the product at 12 K also produced the carbodiimide (7b).3
The benzonitrile N-methylimine (2b) has also been
generated in solution by thermolysis of 1b and found to
undergo 1,3-dipolar cycloaddition, inter alia with nitriles to
furnish 1,2,4-triazoles.8 Surprisingly, thermolysis of 1b at 150
°C without any added benzonitrile still generated the 1,2,4-
triazole (8). Accordingly, it was postulated that the tetrazole
undergoes a competing 1,3-dipolar cycloreversion to benzoni-
trile (9) and methyl azide, CH3−N3. The benzonitrile so-
formed then undergoes a 1,3-dipolar cycloaddition with the
nitrile imine (2b) (Scheme 2).
Preparative FVT of 2-methyl-5-phenyltetrazole (1b) at 550−
960 °C/∼10−4 hPa afforded a much lower yield of the
carbodiimide (7b) (up to 9% yield). Instead, benzonitrile (9)
(up to 50%), styrene (15) (up to 10%), HCN (up to 33%), and
NH3 (up to 14%) were formed as determined by GC, IR, and
1H NMR spectroscopy (Scheme 4).
Scheme 4. C-Phenyl-N-methyl Nitrile Imine (2b) and Its
a
Isomers
Scheme 2. Postulated Mechanism of Formation of Triazole
(8) from Tetrazole (1b)8
The idea seemed reasonable: mild FVT of 5-phenyltetrazole
(1a) gives rise to small amounts of HN3 and benzonitrile (9)
formed by 1,3-dipolar cycloreversion.3 An attempt to generate
the phenyl(stannyl) nitrile imine (12a) by FVT of the tetrazole
(10a) led to quantitative cycloreversion to the stannyl azide
(11a) and benzonitrile (9) (Scheme 3),9 whereas the
a
Numbers in normal font are energies calculated at the B3LYP/6-
31G* level in kcal/mol relative to 2b.
Scheme 3. Formation of C-Phenyl N-Trimethylsilyl Nitrile
Imine (12b), Formation of 11a, and Absence of Formation
of N-Trimethylstannyl Nitrile Imine (12a) on FVT of 10
FVT of the oxadiazolone (3b) gave essentially the same
products as described for 1b except that an equivalent of CO2
was also formed. By analogy with the chemistry of 3a and 1b, it
is assumed nitrile imine (2b) was formed and rearranged to the
diazabutadiene (13) (Scheme 4).
While benzonitrile and methyl azide cannot, of course, be
formed by cycloreversion from 3b, benzonitrile and methylni-
trene CH3N could still be formed by cleavage of the N−N
bond in nitrile imine (2b). There is strong evidence that nitrile
1248
dx.doi.org/10.1021/jo402667y | J. Org. Chem. 2014, 79, 1247−1253

The Journal of Organic Chemistry
Article
imines can cleave to nitriles and nitrenes photochemically3,10
and also in the mass spectrometer.11 The formation of
benzonitrile was observed in the matrix photolysis of 2b.3 If
methylnitrene CH3N were formed, then methanimine (meth-
ylenimine) CH2NH would necessarily be a thermal
product,12 and as we have described in detail recently, this in
turn would undergo rapid decomposition to HCN and H2
under FVT conditions, accelerated by chemical activation
stemming from the exothermicity of the reaction (Scheme 5).12
expected) but also, in small amount, from the nitrile imine
precursors 1b and 3b. The calculated energy barrier for the
rearrangement of PhCNNCH3 (2b) to PhNCN
CH3 (7b) is 48 kcal/mol (Scheme 4). The same barrier (48
kcal/mol) was obtained for the rearrangement of N-phenyl-
nitrile imine to N-phenylcarbodiimide, PhNNCH → PhN
CNH.7 While this is a substantial barrier, it is readily
accessible under the FVT conditions employed here.
2. Benzonitrile Methylide (17) and Ketenimine (23).
Analogous chemistry is observed for the benzonitrile methylide
(17) (Scheme 6). 3-Methyl-1-phenyl-2-azabutadiene (18) is
Scheme 5. Decomposition of Methyl Azide to Methylnitrene,
Methanimine, and HCN
a
Scheme 6. Benzonitrile Methylide (17) and Its Isomers
The ammonia observed as a product from both 1b and 3b is
likely to arise from the oligomerization of CH2NH to
hexamethylenetetramine, which takes place above −80 °C with
evolution of ammonia.13
However, the most likely source of benzonitrile (9) and
CH2NH in the present reaction is the cleavage of the
diazabutadiene (13), itself formed by rearrangement of the
nitrile imine (2b) (Scheme 4). In fact, FVT of 13 itself yielded
substantial amounts of benzonitrile (up to 60% isolated yield)
and HCN (up to 48% isolated yield) together with styrene
(10−16%) and ca. 1% of carbodiimide (7b) at temperatures of
840 °C and above, but not at 550 °C. Azines are known to form
nitriles on FVT via homolysis of the NN bond.14 Therefore,
a free radical route to PhCN + CH2NH is possible. However,
we also found an intramolecular route from 13 (s-E) to 9 +
CH2NH with an activation barrier of 63.8 kcal/mol (see
calculated energies at the B3LYP/6-31G* level in Scheme 4).
This is virtually identical with the NN bond dissociation
energy in hydrazine (63
3 kcal/mol).15 Such barriers are
achievable under FVT at 840 °C. Moreover, as detailed below,
when compound 13 is generated from 2b, it will carry
substantial chemical activation, thereby facilitating its fragmen-
tation.
a
Numbers in normal font are energies calculated at the B3LYP/6-
31G* level in kcal/mol relative to 17.
Formation of the diazabutadiene (13) from the nitrile imine
(2b) can take place via a 1,4-H shift with a modest calculated
barrier of 33.5 kcal/mol (Scheme 4). Compound 13 will be
chemically activated by the sum of the activation energy plus
the exothermicity of its formation, by as much as 73 kcal/mol
when formed from 5b and ∼59 kcal/mol when formed from 2b
(Scheme 4). The lowest energy minimum of 13 is the s-E
conformer (13 s-E); the s-Z conformer has one imaginary
frequency and is a rotational transition state lying 14.8 kcal/mol
above 13 s-E. The chemically activated 13 will thus be able to
cross the activation barrier toward 4-π electrocyclization to
diazetine (14), which lies 30.8 kcal/mol above 13 s-Z. This is
another example of a reaction with two consecutive transition
states,16 the first one being a rotational TS.17 The diazetine
(14) can now fragment to N2 and styrene (15). 1,2-Diazetines
are known to eliminate nitrogen with activation energies of ca.
32 kcal/mol,18 and indeed, we calculate a barrier of 31.8 kcal/
mol. Importantly, the diazabutadiene (13) was isolable when
the FVT temperature was kept low (10% yield at 500 °C; 1.5%
at 800 °C). Thus, the diazabutadiene (13) emerges as the
immediate precursor of benzonitrile, styrene, and methanimine
CH2NH, the latter decomposing to HCN, H2, and NH3
(Scheme 4).
formed in high yield on FVT of 2,2-dimethyl-4-phenyl-oxazol-
5(2H)-one (16) at 400 °C and above.19 Compound 18 is also
formed on FVT of 3,3-dimethyl-2-phenyl-1-azirene (21) but in
much lower yieldonly traces at 600−650 °Cdue to
decomposition as described below.
FVT of the isoxazolone (20) at 600 °C and above results in
elimination of CO2 and formation of azirene (21) together with
the products of thermolysis of 21. The thermolysis of 20 is
initiated by N−O bond breakage,20 and the calculated barrier
20 → 21 is 54 kcal/mol (Scheme 6). Oxazolones like 16
eliminate CO2 with significantly lower activation barriers, so
that these reactions take place already in solution.18 Compound
16 is thermodynamically much more stable than 20, but its
activation barrier toward cycloreversion to CO2 and nitrile ylide
(17) is lower, calculated as only 38 kcal/mol (Scheme 6).
There is essentially no barrier for the reverse reaction, the
cycloaddition of CO2 to 17 to form 16.
The differences between the outcomes of FVT of the three
precursors 16, 20, and 21 can be understood in terms of
chemical activation of azabutadiene (18) arising from the
activation energies for its formation and the exothermicities of
the reactions. Azabutadiene (18) can be chemically activated by
as much as 64−75 kcal/mol when generated from 21 or 20 but
only by ca. 55 kcal/mol when generated from 16 (see
computational data in Scheme 6). The resulting “hot”
It is interesting to note that the carbodiimide (7b) is
obtained not only from the imidoylnitrene precursor (5b) (as
1249
dx.doi.org/10.1021/jo402667y | J. Org. Chem. 2014, 79, 1247−1253

The Journal of Organic Chemistry
Article
azabutadiene decomposes to styrene (15) and acetonitrile via
ring closure to 19 and subsequent [2 + 2] cycloreversion. This
becomes the near-exclusive reaction at 700 °C. The chemical
activation can be removed in part by collisional deactivation by
using N2 as a carrier gas at 1 hPa, which leads to the isolation of
18 in 10% yield from azirene (21); at the same time, the
ketenimine (23), previously only detectable in trace amounts,
benzyl cyanide. The formation of benzyl cyanide (29) is not
necessarily due to a 1,3-phenyl shift in ketenimine (28) but
rather a free-radical cleavage21 of 28 to Ph• and •CH2CN
radicals, because benzene and acetonitrile are also formed. This
is discussed further below. The vinylnitrene (27) can also in
principle rearrange to 29 by means of a 1,2-phenyl shift.27
Although the nitrile ylide (30) has not been observed directly
under these reaction conditions, analogy with the processes
depicted in Scheme 6 suggests that 30 is formed reversibly as
an intermediate (Scheme 7). 30 is observed on matrix
1
becomes clearly observable by H NMR and IR spectroscopy.
Thus, in summary, there are two reaction paths from azirene
(21): one (path a) leading to nitrile ylide (17) and the major
products styrene (15) and acetonitrile and the other (path b)
leading to the vinylnitrene (22) and ketenimine (23). The
ketenimine disappears on FVT at 700 °C, not by rearrange-
ment to the dimethylbenzyl cyanide (24) (α-methylphenyla-
cetonitrile) but rather by free radical decomposition leading to
benzene and 2-methylpropionitrile. There has been much
discussion about the mechanism of rearrangement of
ketenimines to nitriles, but it is clear from product studies
that a free-radical mechanism is common.21 This is discussed
further below.
Scheme 7. Rearrangements of 2-Phenylazirene (26) and N-
Phenylketenimine (28)
Path a has been observed previously by Wendling and
Bergmann,22−24 but path b (formation of ketenimine (23)) has
not. The conditions used by these authors were not FVT;
instead, the reactions were carried out at atmospheric pressure
using helium as a carrier gas and residence times of ca. 10 s.
This causes efficient deactivation of any “hot” molecules, and
the phenomenon of chemical activation of the products was not
observed. Instead, the long contact time and high pressure
cause collisional activation.
The nitrile ylide (17) is not observed directly in any of these
FVT reactions, but several nitrile ylides generated by FVT have
been observed by matrix isolation.25 Like nitrile imines,2,3,5
nitrile ylides can exist in either allenic or propargylic forms.
Consequently, the cumulenic IR absorptions of nitrile ylides
span a very wide range, from the 1900s to near 2300
cm−1.1,2b,26 Calculations demonstrate that compound 17 has
the allenic structure indicated in Scheme 6 with a harmonic
vibrational frequency of 1968 cm−1; the corresponding
propargylic form PhCN+C−(CH3)2 is a transition
state. The calculated energy of 17 is ca. 25 kcal/mol above
the ketenimine (23), and the barrier for the 1,4-H shift to 18 is
30 kcal/mol. The barrier for cyclization of 18 to azetine (19) is
only 12 kcal/mol relative to 17. The final fragmentation to
styrene (15) and acetonitrile requires an overall barrier of 39
kcal/mol. These reactions will be very facile under FVT
conditions.
photolysis of 26.26b,c There is no obvious escape route for 30
under FVT conditionsthe thermal cleavage to benzonitrile
and methylene is not likely. Instead, the high temperature and
chemical activation will ensure that the most stable products,
28 and 29, are formed.
The lowest electronic singlet state of vinylnitrenes has an
open-shell structure (OSS).28,29 Therefore, multiconfigurational
calculations are necessary in order to obtain reliable energies of
the associated transition states. We carried out CASPT2
calculations for the relevant reactions of 1-azirene itself with the
results shown in Scheme 8.
Scheme 8. CASPT2 Calculations Based on CASSCF(6,5) 2σ
a
+ 3π Geometries
3. 2-Phenylazirene (26) and N-Phenylketenimine (28).
FVT of either 3-phenylisoxazol-5(4H)-one (25) or 2-phenyl-
azirene (26) at or above 600 °C gave rise to N-phenyl-
ketenimine (28). Like other isoxazolones, 25 is expected to
undergo elimination of CO2 with an activation barrier of the
order of 55 kcal/mol20 on FVT. The azirene itself was observed
as a product of FVT of 25 at 600 °C. Both 26 and 28 were
isolated and characterized by comparison of IR and NMR data
with data for authentic samples. However, ketenimine (28) was
isolated in only 7% yield on FVT of 26 at 700 °C, and it
disappeared from the thermolyzate on FVT above 850 °C. The
low yield of ketenimine (28) can be ascribed to a rearrange-
ment to benzyl cyanide (29); the ratio of yields 28:29 was 1:3
at 700 °C (24% yield of 29). The easy formation of 29 can
again be ascribed to chemical activation of the ketenimine (see
below), because FVT of the pure, isolated ketenimine at the
same temperature (700 °C) led to only 8% rearrangement to
a
Energies in kcal/mol relative to 1-azirene (31). The values in
parentheses are derived from thermochemical kinetics (see text).
The thermal activation energy for formation of the nitrile
ylide (33) is higher than that for formation of the nitrene (32),
but the ylide has a lower ground state energy. The closed-shell
singlet (CSS) state of the vinylnitrene S2 is 25.3 kcal/mol above
the open-shell (OSS) S1.
1250
dx.doi.org/10.1021/jo402667y | J. Org. Chem. 2014, 79, 1247−1253

The Journal of Organic Chemistry
Article
The nitrile ylide (33) can be chemically activated by ca. 40
kcal/mol when formed by FVT of the azirene (the activation
energy minus the energy of the product). In the reactions of the
substituted azirene (21) in Scheme 6, this excess energy was
carried over in the subsequent exothermic rearrangement to the
azabutadiene (18), thereby facilitating its cyclization to 19 and
cleavage to 15.
an activation barrier of only 30 kcal/mol. Azabutadiene (18) is
also chemically activated, by 64−75 kcal/mol when generated
from 20 or 21, and consequently undergoes cyclization to
azetine (19) and finally fragmentation to styrene and
acetonitrile with an overall barrier of 64 kcal/mol. A lower
heat of formation of oxazolone (16) and a lower barrier for its
fragmentation to nitrile ylide (17) and CO2 mean that the
nitrile ylide (17) generated this way carries much less chemical
activation (38 kcal/mol). The ylide (17) still isomerizes to
azabutadiene (18), but this compound is now isolable in much
better yield.
A second thermal pathway from oxazolone (20) involves
ring-opening of azirene (21) to vinylnitrene (22); the latter
isomerizes to N-phenylketenimine (23).
FVT of the isoxazolone (25) causes loss of CO2 and
formation of 2-phenylazirene (26). Further thermolysis of 26
yields N-phenylketenimine (28), probably via vinylnitrene
(27). The N-phenylketenimine (28) is again chemically
activated by ca. 78 kcal/mol and therefore rearranges to
phenylacetonitrile, most probably via a free radical fragmenta-
tion and recombination.
The ketenimine (34) can be chemically activated by as much
as 78 kcal/mol due to the high barrier for its formation and the
very exothermic reaction. This excess energy facilitates the
observed rearrangement to acetonitrile. However, this is
probably not a direct 1,3-shift of hydrogen, which will have a
high activation barrier. The chemically activated ketenimine
may return to the vinylnitrene (32), which can now undergo a
1,2-H shift with a low barrier of only ca. 7 kcal/mol.
Furthermore, given such a high level of chemical activation
(78 kcal/mol), the rearrangement of ketenimines R2CC
NR′ to nitriles R′R2CCN via cleavage to free radicals
R2CCN• + •R′ followed by recombination becomes
perfectly feasible. The occurrence of free radical decomposition
of ketenimines under FVT conditions has been demonstrated
previously.21a The average CN bond dissociation energy is
only 73 kcal/mol, and the experimental activation energy for
the thermal gas-phase rearrangement of ketenimine CH2C
NH to acetonitrile CH3CN is 70 3 kcal/mol in a shock
tube extrapolated to the high-pressure limit.30 Using this value
in Scheme 8 gives a transition state energy of 53 kcal/mol for
the H2CCN• + •H reaction, and this is the easiest route
for isomerization of CH2CNH to CH3CN. Moreover, the
reverse reaction, CH3CN → H2CCN• + •H → CH2
CNH is equally possible as a source of the ketenimine (34)
in Scheme 8.
CASPT2 calculations on the unsubstituted azirene (31)
indicate that the two thermal ring-opening reactions, to
vinylnitrene (32) and nitrile ylide (33), have activation barriers
of 33 and 48 kcal/mol, respectively. Thus, under ordinary
reaction conditions, vinylnitrene formation will be energetically
preferred, but both pathways become possible under high-
temperature FVT conditions.
COMPUTATIONAL METHODS
■
Structures of ground and transition states were investigated by DFT
theory at the B3LYP/6-31G* level using the Gaussian 09 suite of
programs.33 Energies of open-shell singlet nitrenes were estimated
using the Cramer−Ziegler method.34 For the reactions in Scheme 8,
the energies were also calculated at the CASPT2 level based on
CASSCF(6,5) 2σ + 3π geometries. Where applicable, calculated
wavenumbers were scaled by a factor of 0.9613.35
The best experimental heat of formation of the CH2CN
radical is 60.4 kcal/mol.31 Adding the heat of formation of a
hydrogen atom (52.1 kcal/mol32) gives ΔfH°(H2CCN• + •H)
= 112.5 kcal/mol, or 94.8 kcal/mol above CH3CN (17.7 kcal/
mol32). Thus, ΔfH°(H2CCN• + •H) = 45.5 kcal/mol on the
scale of Scheme 8.
EXPERIMENTAL SECTION
■
We did not locate a transition state for the direct 1,3-shift 34
→ 35 at the CASSCF level. However, a barrier of 67.4 kcal/mol
for the homolysis 34 → H2CCN• + •H (36) was calculated at
ab initio levels up to MP4(SDTQ)/6-31G*//MP2(FU)/6-
FVT reactions were carried out in unpacked quartz tubes (20 × 2 cm
I.D.), usually in dynamic vacua of 10−4−10−3 hPa maintained by using
high capacity oil diffusion pumps, except when N2 was used as a carrier
gas in order to achieve collisional deactivation, usually at 1 hPa under
continuous pumping. Products were collected in liquid-N2-cooled
traps or on liquid-N2-cooled cold fingers. Further details have been
reported.12 After the end of each FVT experiment, the HCN/NH3
mixture was volatilized by warming the cold finger/cold trap to ca.
−20 °C, collected in an evacuated IR gas cell, and assayed by IR
spectroscopy by comparison with gas mixtures of known composition.
Higher boiling liquids and solids were collected after equalizing the
system pressure. Product mixtures were assayed by gas chromatog-
raphy on SE30, SE50, or Carbowax columns as well as 1H NMR
spectroscopy, and known compunds were identified by coinjection and
by comparison of their spectroscopic data with those of authentic
materials.
31G* in good agreement with the shock tube value of 70
3
kcal/mol. Thus, by this calculation, the energy of the radical
pair is 50 kcal/mol relative to azirene (31) on the scale of
Scheme 8 in reasonably good agreement with the experimental
value of 45.5 kcal/mol derived above.
CONCLUSION
■
C-Phenyl-N-methylnitrile imine (2b) is generated by FVT of
tetrazole (1b) and 1,3,4-oxadiazolone (3b). 2b undergoes a 1,4-
H shift to the azabutadiene (13) with a calculated activation
barrier of 33.5 kcal/mol. Furthermore, compound 13
decomposes to styrene, N2, benzonitrile, and CH2NH due
to chemical activation under the low-pressure conditions of its
formation. B3LYP calculations indicate that the decomposition
to styrene and N2 takes place via 4-electron electrocyclization
to diazetine (14) with an overall activation barrier of 38 kcal/
mol.
MATERIALS
■
2-Methyl-5-phenyltetrazole3,36 (mp 50 °C), 3-methyl-5-phenyl-1,3,4-
oxadiazol-2(3H)-one37 (mp 105 °C), benzaldehyde methylenehydra-
zone (2,3-diaza-1-phenylbuta-1,3-diene)38 (mp (decomp.) 146 °C),
2,2-dimethyl-4-phenyloxazol-5(2H)-one39 (mp 36 °C, bp 90−100 °C/
10−2 hPa), 3-phenylisoxazol-5-(4H)-one40 (mp 148 °C), 2-phenyl-1-
azirene41 (bp 64 °C/5 hPa), 3,3-dimethyl-2-phenyl-1-azirene42 (bp 94
°C/15 hPa), 3-ethyl-1-phenyl-2-azabuta-1,3-diene,13 N-methyl-N′-
phenylcarbodiimide (by FVT of 1-methyl-5-phenyltetrazole at 600
°C),3 N-phenylketenimine (by FVT of 4-(N-anilinomethylidene)-3-
Isoxazolone (20), azirene (21), and oxazolone (16) all afford
benzonitrile dimethylmethylide (17) on FVT. However, 17 is
chemicaly activated by as much as 51 kcal/mol when generated
from 20. This causes the 1,4-H shift to azabutadiene (18) with
1251
dx.doi.org/10.1021/jo402667y | J. Org. Chem. 2014, 79, 1247−1253

The Journal of Organic Chemistry
Article
methylisoxazol-5(4H)-one at 650 °C),43 and N-phenyl-3,3-dimethyl-
ketenimin44 were synthesized according to the cited literature
procedures.
FVT of 2-Methyl-5-phenyltetrazole (1b). Portions of 150 mg
(0.9 mmol) were subjected to FVT at temperatures between 550 and
960 °C (5 × 10−4 hPa). The sublimation temperature was 45 °C.
More than 95% of the unchanged starting material was recoved at FVT
temperatures below 500 °C. The product compositions and yields,
determined by GC (IR spectroscopy for HCN and NH3), at various
temperatures are collected in Table 1.
(10%), acetonitrile (40%), styrene (40%), and benzonitrile (ca. 5%)
was obtained together with ca. 1% percent of 1-phenyl-3-methyl-2-
azabuta-1,3-diene (18). (c) At 700 °C, the starting material was almost
completely consumed. The product consisted of a 1:1 mixture of
acetonitrile and styrene, and the azabutadiene (18) was no longer
detectable. (d) The pyrolysis was repeated at 600 °C with N2 as carrier
gas at a pressure of 1 hPa under continuous pumping. The product
mixture consisted of starting material (4%), acetonitrile (35%), styrene
(51%), azabutadiene (18) (10%), and the ketenimine (23) (ca. 1%).
The trapping of the volatile acetonitrile may be incomplete under
these conditiions.
FVT of 4,4-Dimethyl-3-phenylisoxazol-5-(4H)-one (20). (a)
Preparative pyrolysis of this compound was carried out at 600 °C with
isolation of the products in a liquid-N2 cold trap to yield starting
material (20) (6%), azirene (21) (44%), styrene (15) (16%),
acetonitrile (14%), azabutadiene (18) (4%), and keteniminene (23)
(1%). (b) When the pyrolysis products were condensed on a KBr
target in a liquid-N2-cooled cryostat, the formation of acetonitrile
(2252 cm−1), N-phenyl-3,3-dimethylketenimine (23) (2016 cm−1),
and 3,3-dimethyl-2-phenylazirene (21) (1735 cm−1) was confirmed by
the IR absorptions.
FVT of 3-Phenylisoxazol-5-(4H)-one (25). Compound 25 (500
mg portions) was pyrolyzed at 600 and 700 °C with a sublimation
temperature of 120 °C. At 600 °C, a mixture of the unchanged starting
material, benzonitrile, and 2-phenyl-1-azirene (26) was obtained, the
latter two in a ca. 1:1 ratio. At 700 °C, the starting material was
completely consumed, and benzonitrile, 2-phenyl-1-azirene (26), N-
phenylketenimine (28), and benzyl cyanide (29) were obtained in
ratios of ca. 5:5:1:4.
Table 1. Yields of Products from the FVT of 2-Methyl-5-
phenyltetrazole (1b)
yield/FVT temp (°C)
styrene
550
9
800
6
900
0.5
32
0
960
1
benzonitrile
diazabutadiene (13)
carbodiimide (7b)
HCN
44
10
1.5
24
10
49
1.5
6.5
31
13
42
0
7
9
32
14
31
13
NH3
FVT of 3-Methyl-5-phenyl-1,3,4-oxadiazol-2(3H)-one (3b).
Portions of 200 mg (1.14 mmol) were subjected to FVT at
temperatures between 700 and 980 °C at a sublimation temperature
of 70 °C. The product composition and yields are given in Table 2.
The gas mixtures consisted of HCN/NH3 in the usual molar ratio of
70:30 together with substantial amounts of CO2.
FVT of 2-Phenyl-1-azirene (26). Portions of 500 mg were
pyrolyzed at 700−900 °C. At 700 °C, a large proportion of the starting
material remained unchanged. Benzonitrile, 2-phenyl-1-azirene (26),
N-phenylketenimine (28), and benzyl cyanide (29) were obtained in
ratios of ca. 2:5:1:2. At 800 °C, benzonitrile, 2-phenyl-1-azirene, N-
phenylketenimine, and benzyl cyanide were obatined in ratios of ca.
5:1:0:4. At 850 °C, no azirene remained in the product. Benzonitrile
and benzyl cyanide were obtained in a ratio of ca. 1:1. N-
Phenylketenimine (IR 2030 cm−1) was not detectable at 800 and
850 °C. Pyrolysis at 900 °C gave virtually the same result as that at 850
°C.
FVT of N-Phenylketenimine (28). Complete pyrolysis of this
compund was not achieved because a part of the compound invariably
polymerized in the sample flask prior to pyrolysis. At 700 °C, a
conversion to 8% benzyl cyanide was obtained. At 800 °C, ca. 34%
conversion to benzyl cyanide was achieved. The formation of benzene
and acetonitrile was confirmed by GC in each case.
Table 2. Yields of Products from the FVT of 3-Methyl-5-
phenyl-1,3,4-oxadiazol-2(3H)-one (3b)
yield/FVT temp (°C)
styrene
700
8
900
12
75
0
960
6
benzonitrile
diazabutadiene (13)
carbodiimide (7b)
HCN
71
0
63
0.2
9
1
1.5
34
14
18
8
NH3
FVT of 2,3-Diaza-1-phenylbuta-1,3-diene (Benzaldehyde
Methylenehydrazone) (13). A sample of 50 mg (0.38 mmol) was
subjected to FVT between 550 and 960 °C with a sublimation
temperature of 100−120 °C. The products were analyzed as described
above, and the results are collected in Table 3.
ASSOCIATED CONTENT
* Supporting Information
Computational details for the calculated species in Schemes 4,
6, and 8. This material is available free of charge via the Internet
at http://pubs.acs.org.
■
Table 3. Yields of Products from the FVT of 2,3-Diaza-1-
phenylbuta-1,3-diene (13)
S
yield/FVT temp (°C)
styrene
550
9
840
16
59
0
900
10
55
0
980
12
60
0
benzonitrile
diazabutadiene (13)
carbodiimide (7b)
HCN
32
45
1
AUTHOR INFORMATION
Corresponding Author
*E-mail: wentrup@uq.edu.au
Notes
■
1
1
0.5
46
30
48
31
45
29
NH3
The authors declare no competing financial interest.
FVT of 2,2-Dimethyl-4-phenyloxazol-5(2H)-one (16). FVT of
samples of 250 mg (1.3 mmol) was carried out at 400 and 600 °C. At
400 °C, a ca. 1:1 mixture of the unchanged starting material and 1-
phenyl-3-methyl-2-azabuta-1,3-diene (18) was obtained. At 600 °C, 18
(170 mg) was obtained as the only discrete product in 87% yield.
FVT of 3,3-Dimethyl-2-phenyl-1-azirene (21). Portions of 500
mg were pyrolyzed. At 500 °C, the starting material was recovered
unchanged. (a) At 600 °C, a mixture of starting material (33%),
acetonitrile (30%), and styrene (35%) was obtained together with ca.
2% of azabutadiene (18). (b) At 650 °C, a mixture of starting material
ACKNOWLEDGMENTS
■
This work was supported by The Deutsche Forschungsge-
meinschaft, The University of Queensland, the National
Computing Infrastructure facility (NCMAS g01) supported
by the Australian Government, and the computing facilities
MCIA (Mes
Universite de Bordeaux and of the Universite
Pays de l’Adour.
́
ocentre de Calcul Intensif Aquitain) of the
́
́
de Pau et des
1252
dx.doi.org/10.1021/jo402667y | J. Org. Chem. 2014, 79, 1247−1253

The Journal of Organic Chemistry
Article
1,4-shifts in photolysis of azirenes, see: Padwa, A.; Carlsen, P. H. J.;
Tremper, A. J. Am. Chem. Soc. 1978, 100, 4481.
Early experimental contributions by Dr Stefan Fischer and
Dr Dieter Laqua are gratefully acknowledged.
(24) For examples of CO2 extrusion from oxazolones in substance or
in liquid solution with azabutadiene formation at 120−130 °C, see:
(a) Steglich, W.; Gruber, P.; Heininger, H.-U.; Kneidl, F. Chem Ber.
1971, 104, 3816. (b) Fischer, J.; Steglich, W. Angew. Chem., Int. Ed.
Engl. 1979, 18, 167.
REFERENCES
■
(1) (a) Sharp, J. T. Nitrile Ylides and Nitrile Imines. In The Chemistry
of Heterocyclic Compounds 59: Synthetic Applications of 1,3-Dipolar
Cycloaddition Chemistry Toward Heterocycles and Natural Products;
Padwa, A., Pearson, W. H., Eds.; John Wiley & Sons: New York, 2002.
(b) 1,3-Dipolar Cycloaddition Chemistry, Vols. I and II; Padwa, A., Ed.;
Wiley-Interscience: New York, 1984.
(25) Wentrup, C. Acc. Chem. Res. 2011, 44, 393.
(26) (a) Inui, H.; Murata, S. J. Am. Chem. Soc. 2005, 127, 2628.
(b) Orton, E.; Collins, S. T.; Pimentel, G. C. J. Phys. Chem. 1986, 90,
6139. (c) Chapman, O. L.; LeRoux, J.-P. J. Am. Chem. Soc. 1978, 100,
282.
(27) Hassner, A.; Wiegand, N. H.; Gottlieb, H. E. J. Org. Chem. 1986,
51, 3176.
(28) (a) Bornemann, C.; Klessinger, M. Chem. Phys. 2000, 259, 263.
(b) Klessinger, M.; Bornemann, C. J. Phys. Org. Chem. 2002, 15, 514.
(29) For calculations on other vinylnitrenes, see: (a) Rajam, S.;
Murthy, R. S.; Jadhav, A. V.; Li, Q.; Keller, C.; Carra, C.; Pace, T. C. S.;
Bohne, C.; Ault, B. S.; Gudmundsdottir, A. D. J. Org. Chem. 2011, 76,
9934. (b) Nunes, C. M.; Reva, I.; Pinho e Melo, M. V. D.; Fausto, R.;
Solomek, T.; Bally, T. J. Am. Chem. Soc. 2011, 133, 18911. (c) Gamage,
D. W.; Li, Q.; Upul Ranaweera, R. A. A.; Sarkar, S. K.; Weragoda, G.
K.; Carr, P. L.; Gudmundsdottir, A. D. J. Org. Chem. 2013, 78, 11349.
(30) Doughty, A.; Backsay, G. B.; Mackie, J. C. J. Phys. Chem. 1994,
98, 13546.
(2) (a) Caramella, P.; Houk, K. N. J. Am. Chem. Soc. 1976, 98, 6397.
(b) Bertrand, G.; Wentrup, C. Angew. Chem., Int. Ed. Engl. 1994, 33,
527. (c) Mawhinney, R. C.; Muchall, H. M.; Peslherbe, G. H. Chem.
Commun. 2004, 1862. (d) Mawhinney, R. C.; Muchall, H. M.;
Peslherbe, G. H. Can. J. Chem. 2005, 83, 1615. (e) Cargnoni, F.;
Molteni, G.; Cooper, D. L.; Raimondi, M.; Ponti, A. Chem. Commun.
2006, 1030. (f) Mawhinney, R. C.; Peslherbe, G. H.; Muchall, H. M. J.
Phys. Chem. B 2008, 112, 650. (g) Ess, D. H.; Houk, K. N. J. Am.
Chem. Soc. 2008, 130, 10187. (h) Braida, B.; Waiter, C.; Engels, B.;
Hiberty, P. C. J. Am. Chem. Soc. 2010, 132, 7631.
(3) Beg
5339.
́ ́
ue, D.; Qiao, G. G.; Wentrup, C. J. Am. Chem. Soc. 2012, 134,
(4) Fischer, S.; Wentrup, C. Chem. Commun. 1980, 502.
(5) The photochemical rearrangement of the unsubstituted nitrile
imine HCNNH (1) to cyanamide has also been described: (a) Maier,
G.; Eckwert, J.; Bothur, A.; Reisenauer, H. P.; Schmidt, C. Liebigs Ann.
1996, 1041. (b) Duvernai, F.; Borget, F.; Aycard, J.-P. J. Phys. Chem. A
2005, 109, 603.
(6) (a) Gomez-Zavaglia, A.; Reva, I. D.; Frija, L.; Cristiano, M. L.;
Fausto, R. J. Phys. Chem. A 2005, 7967. (b) Gomez-Zavaglia, A.; Reva,
I. D.; Frija, L.; Cristiano, M. L. S.; Fausto, R. J. Photochem. Photobiol., A
2006, 180, 175. (c) Frija, L. M. T.; Reva, I. D.; Gomez-Zavaglia, A.;
Cristiano, M. L. S.; Fausto, R. J. Phys. Chem. A 2007, 111, 2879.
(7) Veedu, R. N.; Kvaskoff, D.; Wentrup, C. Aust. J. Chem. 2014,
DOI: 10.1071/CH13536.
(31) Lafleur, R. D.; Szatary, B.; Baer, T. J. Phys. Chem. A 2000, 104,
1450.
(32) NIST Webbook, webbook.nist.gov/cgi/cbook.cgi?ID=
C12385136&Mask=20, accessed Nov 24, 2013.
(33) Frisch, M. J.; et al. Gaussian 09, revision A.2; Gaussian, Inc.:
Pittsburgh, PA, 2009. The full reference is given in the Supporting
Information.
(34) Johnson, W. T. G.; Sullivan, M. B.; Cramer, J. C. Int. J. Quantum
Chem. 2001, 85, 492.
(35) (a) Wong, M. W. Chem. Phys. Lett. 1996, 256, 391. (b) Scott, A.
P.; Radom, L. J. Phys. Chem. 1996, 100, 16502.
(36) (a) Henry, R. A. J. Am. Chem. Soc. 1951, 73, 4470. (b) Fraser, R.
R.; Haque, K. E. Can. J. Chem. 1968, 46, 2855.
(37) Ainsworth, C. Can. J. Chem. 1965, 43, 1607.
(8) (a) Huisgen, R.; Seidel, M.; McFarland, J. W.; Wallbillich, G. J.
Org. Chem. 1959, 24, 892−893. (b) Huisgen, R.; Grashey, R.; Seidel,
M.; Wallbillich, G.; Knupfer, H.; Schmidt, R. Justus Liebigs Ann. Chem.
1962, 653, 105.
(38) Bestmann, H. J.; Fritzsche, H. Chem. Ber. 1961, 94, 2477.
(39) Padwa, A.; Wetmore, S. I. J. Am. Chem. Soc. 1974, 96, 2414.
(40) (a) Claisen, L. Ber. Dtsch. Chem. Ges. 1891, 24, 140.
(b) Hantzsch, A. Ber. Dtsch. Chem. Ges. 1891, 24, 495.
(41) Hortmann, A. G.; Robertson, D. A.; Gillard, B. K. J. Org. Chem.
1972, 37, 322.
(42) Leonard, L. J.; Zwanenburg, B. J. Am. Chem. Soc. 1967, 89, 4456.
(43) Winter, H.-W.; Wentrup, C. Angew. Chem., Int. Ed. Engl. 1980,
19, 720.
(9) Fischer, S. Nitriliumbetaine. Thermische Erzeugung und
spektroskopische Nachweise. Ph.D. Thesis, Philipps-Universitat
̈
Marburg, Germany, 1985.
(10) Toubro, H. H.; Holm, A. J. Am. Chem. Soc. 1980, 102, 2093.
(11) Wentrup, C.; Maquestiau, A.; Flammang, R. Org. Mass Spectrom.
1981, 16, 115.
(12) Wentrup, C. Aust. J. Chem. 2013, 66, 852.
(13) (a) Lin, C.; Smith, M. A. Icarus 2012, 220, 627. (b) Braillon, B.;
Lasne, M. C.; Ripoll, J. L.; Denis, J. M. New J. Chem. 1982, 6, 121.
(14) Bird, K. J.; Chan, A. W. K.; Crow, W. D. Aust. J. Chem. 1976, 29,
2281.
(44) Singer, L. A.; Davis, G. A. J. Am. Chem. Soc. 1967, 89, 598.
(15) Szwarc, M. Proc. R. Soc. London, Ser. A: Math. Phys. Sci. 1949,
198, 267.
(16) Birney, D. M. Curr. Org. Chem. 2010, 14, 1658.
(17) Finnerty, J. J.; Wentrup, C. J. Org. Chem. 2005, 70, 9735.
(18) Breton, G. W.; Shugart, J. H. J. Org. Chem. 2003, 68, 8643.
(19) Berstermann, H.-M.; Netsch, K.-P.; Wentrup, C. Chem.
Commun. 1980, 503.
(20) Rzepa, H. S.; Wentrup, C. J. Org. Chem. 2013, 78, 7565.
(21) (a) Reichen, W.; Wentrup, C. Helv. Chim. Acta 1976, 59, 2615.
(b) Lee, K.-W.; Horowitz, N.; Ware, J.; Singer, L. A. J. Am. Chem. Soc.
1977, 99, 2622. (c) Neuman, R. C.; Sylwester, A. P. J. Org. Chem.
1983, 48, 2285. (d) Clarke, L. F.; Hegarty, A. F.; O’Neill, P. J. Org.
Chem. 1992, 57, 262.
(22) Wendling, L. A.; Bergman, R. G. J. Org. Chem. 1976, 41, 831.
(23) (a) For other examples of azabutadiene formation by gas phase
pyrolysis of azirenes (80 cm pyrolysis tube filled with glass beads, 340
°C, 0.1 hPa), see: Demoulin, A.; Gorissen, H.; Heshain-Frisque, A.
M.; Ghosez, L. J. Am. Chem. Soc. 1975, 97, 4409. (b) For examples of
1253
dx.doi.org/10.1021/jo402667y | J. Org. Chem. 2014, 79, 1247−1253