Experimental and theoretical characterization of the aromatization, epimerization, and fragmentation reactions of Bi-2H-azirin-2-yls prepared from 1,4-diazidobuta-1,3-dienes

Article abstract of DOI:10.1002/chem.201101220

1,4-Diazidobuta-1,3-dienes (Z,Z)-10, 17, and 21 were photolyzed and thermolyzed to yield the pyridazines 13, 20, and 23, respectively. To explain these aromatic final products, the generation of highly strained bi-2H-azirin-2-yls 12, 19, and 22 and their

Full text of DOI:10.1002/chem.201101220

FULL PAPER
DOI: 10.1002/chem.201101220
Experimental and Theoretical Characterization of the Aromatization,
Epimerization, and Fragmentation Reactions of Bi-2H-azirin-2-yls Prepared
from 1,4-Diazidobuta-1,3-dienes**
Klaus Banert,*[a] Frank Kçhler,[a] Antje Melzer,[a] Ingolf Scharf,[a] Gerd Rheinwald,[b]
Tobias Rꢀffer,[b] Heinrich Lang,[b] Rainer Herges,*[c] Kirsten Heß,[c]
Nugzar Ghavtadze,[d] and Ernst-Ulrich Wꢀrthwein*[d]
Abstract:
1,4-Diazidobuta-1,3-dienes
N
cyclic 2H-azirines 29 at a relatively low
temperature (758C). The fragmentation
of rac-22 to give alkyne 24 and two
molecules of acetonitrile was also stud-
ied by high-level quantum chemical
(Z,Z)-10, 17, and 21 were photolyzed
and thermolyzed to yield the pyrid-
ACHTUNGTRENNUNGazines 13, 20, and 23, respectively. To
explain these aromatic final products,
the generation of highly strained bi-
2H-azirin-2-yls 12, 19, and 22 and their
valence isomerization were postulated.
In the case of meso- and rac-22, nearly
quantitative formation from diazide 21,
isolation as stable solids, and complete
characterization were possible. On the
thermolysis of 22, aromatization to 23
was only a side reaction, whereas equil-
calculations. For
a
related model
system 30 (methyl instead of phenyl
groups), two transition states TS-30–31
of comparable energy with multicon
ACHTUNGTRENNUNGfig-
AHCTUNGTRENNuGUN rational electronic states could be lo-
calized on the energy hypersurface for
this one-step conversion. The symmet-
rical transition state complies with the
definition of a coarctate mechanism.
Introduction
The valence isomers of benzene, prismane (1), Dewar ben-
zene (2), benzvalene (3), and bicycloprop-2-enyl (4a), stimu-
lated the fantasy of chemists for nearly one and a half centu-
ries.[1] When Breslow et al. synthesized the first derivative of
this latter (CH)6 species, the highly substituted compound
4b, they discovered also the aromatization of this product to
give hexaphenylbenzene on heating, for example, in reflux-
ing xylene.[2] Less substituted starting materials, for instance,
meso-4c, rac-4c, and 4d, isomerize at 160–2008C in the gas
phase each to give mixtures of o-, m-, and p-xylenes.[3] Sev-
eral mechanisms were proposed to explain these aromatiza-
tion reactions,[3b,4] which are accompanied by Cope re-
A
Some attempts were made to generate the diaza ana-
logues of bicycloprop-2-enyls 4, namely, the bi-2H-azirin-2-
yls of type 7 (Scheme 1). However, neither the Neber re-
[a] Prof. Dr. K. Banert, Dr. F. Kçhler, Dr. A. Melzer, Dr. I. Scharf
Organic Chemistry, Chemnitz University of Technology
Strasse der Nationen 62, 09111 Chemnitz (Germany)
Fax : (+49)371-531-21229
AHCTUNGTREGaNNUN ction of hydrazonium salts or similar starting materials nor
E-mail: klaus.banert@chemie.tu-chemnitz.de
the coupling reaction of 2-chloro-2H-azirines with various
reagents led to the heterocycles 7.[5] Moreover, efforts to
prepare 1,4-diazidobuta-1,3-dienes of type 6 as potential
precursors of the target compounds 7 failed.[5a,6,7] But re-
cently, diazides 6 were synthesized by electrocyclic conrota-
tory ring opening of the diazidocyclobutenes 5b available by
nucleophilic substitution of the corresponding dihalides
5a.[8] The products 6 proved to be ideal starting compounds
to generate bi-2H-azirin-2-yls 7 in excellent yields by photol-
ysis or thermal reaction below ambient temperature. The
highly strained heterocycles 7 could not be isolated because
they isomerized nearly quantitatively to give pyridazines 8
[b] Dr. G. Rheinwald, Dr. T. Rꢀffer, Prof. Dr. H. Lang
Inorganic Chemistry, Chemnitz University of Technology
Strasse der Nationen 62, 09111 Chemnitz (Germany)
[c] Prof. Dr. R. Herges, K. Heß
Institute of Organic Chemistry, University of Kiel
Otto-Hahn-Platz 4, 24098 Kiel (Germany)
[d] Dr. N. Ghavtadze, Prof. Dr. E.-U. Wꢀrthwein
Institute of Organic Chemistry, University of Mꢀnster
Corrensstrasse 40, 48149 Mꢀnster (Germany)
[**] Reactions of Unsaturated Azides, Part 28. For Part 27, see refer-
ence [11].
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101220.
Chem. Eur. J. 2011, 17, 10071 – 10080
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10071

yield.[10] On the other hand, similar dimerizations to produce
pyrazine derivatives were reported several times.[9]
Unfortunately, the substitution pattern of 5 is limited to
R1, R2, R3, R4 =H and small alkyl or phenyl groups, and the
synthesis 5a!5b!6!7 cannot be performed on a large
scale. Moreover, the high reactivity of 7 did not allow isola-
tion of these substances and prevented experimental assign-
ment of the stereoisomers of 7. Very recently, however,
large-scale synthesis of novel 1,4-diazidobuta-1,3-dienes was
presented.[11] We describe here the complete characteriza-
tion of isolable stereoisomeric bi-2H-azirin-2-yls prepared
from such diazides. Thus, it was possible to investigate not
only aromatization but also epimerization and fragmenta-
tion processes of such heterocycles. Furthermore, we studied
these reactions by quantum chemical calculations.
Scheme 1. Synthesis and aromatization of known bi-2H-azirin-2-yls 7.
even at low temperatures, whereas also pyrimidines 9 were
formed by prolonged irradiation or in the presence of silver
salts. Thus, the reported[8] aromatizations of 7 indicate that
valence isomerizations to prepare pyridazines 8 are possible
although rearrangements to produce other diazabenzenes
were postulated in the literature.[5,6] The thermal transfor-
mation 7!8 showed some remarkable singularities.[8] Com-
pared to thermal reactions of simple 2H-azirines,[9] the low
temperature (ꢀ25 to +108C) at which the aromatization of
7 could be performed was surprising, especially when com-
pared to the temperatures that were necessary for analogous
valence isomerizations of bicycloprop-2-enyls 4.[3,4] The very
different reactivities of stereoisomeric meso- and rac-7 or
unlike- and like-7, respectively, are also noteworthy. Further-
more, the exclusive thermal transformation of 7 to yield py-
Results and Discussion
When a solution of diazide (Z,Z)-10[11] was irradiated by a
high-pressure mercury lamp at low temperature, only the
mono-azirine (Z)-11 (yield: 26%) but not the corresponding
biazirinyls 12 could be detected (Scheme 2). Prolonged pho-
tolysis gave only traces of the known[12] pyridazine 13,
whereas heating of (Z,Z)-10 in dichloromethane or chloro-
form resulted in the formation of 13 in 27% yield besides
small amounts of (Z)-11 (up to 2%). We assumed that
(Z,Z)-10 was transformed into 13 via the intermediates (Z)-
11 and the highly unstable biazirinyl 12. Thus, we irradiated
(Z,Z)-10 in the presence of cyclopentadiene to give the trap-
ping products meso-16 and rac-16 with excellent yield. The
assignment of the stereoisomers was confirmed by X-ray
single-crystal structure analysis of meso-16 (Figure 1). In the
case of incomplete photolysis, the aziridine (Z)-15 was also
isolated. If less reactive 2,3-dimethylbuta-1,3-diene was used
ACHTUNGTRENNUNGridazines 8 is in contrast to bicycloprop-2-enyls, such as 4c
or 4d, which led to mixtures of o-, m-, and p-xylenes.[3] The
formation of pyridazines from two molecules of simple 2H-
azirines was observed only in rare cases and in a very low
Scheme 2. Photolysis and thermolysis of diazide (Z,Z)-10.
10072
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10071 – 10080

Bi-2H-azirin-2-yls Prepared from 1,4-Diazidobuta-1,3-dienes
FULL PAPER
Scheme 3. Photolysis and thermolysis of diazides 17.
Figure 1. Molecular structure of meso-16 determined by single-crystal X-
ray diffraction analysis.
In refluxing chloroform, the starting material (E,E)-21 or
the diastereomeric 1,4-diazidobuta-1,3-dienes of type 21
were transformed into a 1:2 mixture of meso- and rac-22
with 92% yield (Scheme 4). The same products were
formed in a 1:1 ratio on irradiation of 21 in 90% yield. The
bi-2H-azirin-2-yls 22 are stable solids, which were separated
easily by simple recrystallization. Since 21 is conveniently
available on a large scale,[11] multigram amounts of meso-
and rac-22 could be prepared without effort. Structure as-
signment was performed with the help of single-crystal X-
ray diffraction analyses (Figure 2). The relatively high ther-
mal stability of 22 is based upon steric shielding due to the
bulky substituents. In the case of 22, electron-withdrawing
substituents are bonded at C-2 but not at C-3 of the hetero-
cycles. Thus, significant electronic destabilization of the azir-
ine units, as is plausible for 12 and 19, is avoided in 22.
for such interception of azirines by [4+2]-cycloaddition,
only the product (Z)-14 could be characterized.
It is well-known from the work of Gilchrist et al.[13] and
others[14] that 2H-azirines with an electron-withdrawing
group in the 3-position are unstable but good dienophiles in
stereoselective Diels–Alder reactions. In the case of (Z,Z)-
10, the ester groups made it possible to prepare the diazide
by simple nucleophilic substitution of the corresponding di-
chloride,[11] but these electron-withdrawing groups increase
the electron deficiency at the C=N carbon atoms of the het-
erocycles 12, which are strongly destabilized. Thus, we
looked for other types of 1,4-diazidobuta-1,3-dienes to gen-
erate bi-2H-azirin-2-yls.
When we photolyzed a solution of the diazide (E,Z)-17,[11]
a mixture of two diastereomeric, highly unstable 2H-azirines
18, formed with 60 and 26% yield, was detected by
1H NMR spectroscopy, whereas the analogous treatment of
a symmetrical isomer of 17[11] gave only a single stereoiso-
mer of 18 in 67% yield
On heating the pure biazirinyls meso- or rac-22 in DMSO
at 75–1358C, pyridazine 23 was generated only in a side re-
action with 3–10% yield. However, a fragmentation afford-
ing two molecules of acetonitrile and the known[16] alkyne
24 (yield: 92%) was mainly observed. These thermal re-
(Scheme 3). Even irradiation
at ꢀ858C did not bring about
signals of the desired biazirin-
yls 19. However, we assume
that these elusive intermedi-
ates are involved in the ther-
molysis of 17 (DMSO, 828C),
which led to pyridazine 20 in
low yields (16–37% based on
1H NMR spectroscopy). The
surprising and disappointing
instability of 18 and 19 is most
probably also caused by ac-
ceptor properties of the sub-
stituents at C-3 of the hetero-
cyclic rings although the
CH2SO2Ph unit is a weak elec-
tron-withdrawing group.[15]
Scheme 4. Synthesis, thermolysis, and photolysis of biazirinyls 22.
Chem. Eur. J. 2011, 17, 10071 – 10080
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10073

K. Banert, R. Herges, E.-U. Wꢀrthwein et al.
Figure 3. Thermolyses of rac- and meso-22 in [D6]DMSO at 758C moni-
tored by 1H NMR spectroscopy. x=rac-22; +=meso-22; & =23; =24.
~
ed: The equilibration alone should lead to a ratio of rac-22/
meso-22 3:1, and nearly no products 23 and 24 are formed
directly from meso-22 (Scheme 5). An analogous balance
was previously found in the case of bicycloprop-2-enyls rac-
and meso-4c, which gave very similar distributions of o-, m-,
and p-xylene on gas-phase pyrolysis at 160–2008C.[3a] On the
other hand, fragmentation related to rac-22!24+2MeCN
was observed as a side reaction in the thermolysis (>2008C)
or in the photolysis of hydrocarbon 4e furnishing di-tert-bu-
Figure 2. Molecular structures of rac- (top) and meso-22 (bottom) deter-
mined by single-crystal X-ray diffraction analysis. In the case of rac-22,
an oxygen-bonded water molecule was omitted for reasons of clarity.
Analogously, an enclosed solvent molecule of benzene is not shown in
the case of meso-22.
ACHTUNGTRENNUNGactions were accompanied by an equilibration of meso- and
1
rac-22. A more detailed investigation supported by H NMR
spectroscopic monitoring of the thermolyses showed that
formation of 23 and 24 started exclusively from rac-22.
Thus, meso-22 had to be epimerized into rac-22 before 23
and 24 could be produced (Figure 3). This outcome is com-
patible with recent quantum chemical calculations, which
analyzed the valence isomerization of biazirinyls of type 7
to give pyridazines 8 and postulated a significantly higher
reactivity of rac-7 than that of diastereomeric meso-7.[8b]
When the corresponding rate constants for the balance of
rac- and meso-22, and the irreversible generation of 23 and
24 from rac-22 and possibly from meso-22 were estimated
from 1H NMR spectroscopic analysis of the processes and
with the help of GEPASI fitting,[17] two results were indicat-
Scheme 5. Rate constants for fragmentation, epimerization, and aromati-
zation of rac- and meso-22 based on 1H NMR spectroscopic data
(Figure 3) and first-order kinetics for each of the reactions. The thermol-
yses were performed in [D6]DMSO at 758C, and the rate constants were
calculated by using GEPASI fitting.
10074
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10071 – 10080

Bi-2H-azirin-2-yls Prepared from 1,4-Diazidobuta-1,3-dienes
FULL PAPER
tylacetylene without any products of aromatization or va-
lence isomerization processes.[18]
When anhydrous solutions of the single stereoisomers
meso-22 and rac-22 were irradiated in a prolonged way, nei-
ther epimerization of the biazirinyls nor formation of pyrid-
azine 23 could be detected. The photolyses of both sub-
strates proceeded similarly and led mainly to pyrimidine 25
besides small amounts (ca. 3% yield) of the fragmentation
products 24 and acetonitrile. However, the yields of 25 were
limited to approximately 22% because this aromatic com-
pound was not photostable.
The thermal epimerization of meso- and rac-22, observed
already at 758C, raises the question whether the change of
configuration at such a low temperature is limited to biazir-
inyls or can be found also in the case of simple 2H-azirines.
Optically active derivatives of the latter heterocycles with a
stereogenic center at C-2 were used repeatedly as building
blocks in organic synthesis.[19] Moreover, some examples of
naturally occurring 2H-azirines are known.[20] These antibi-
otic compounds are all optically active 2H-azirines bearing a
center of chirality at C-2. However, an unwanted racemiza-
tion on heating seems to be possible.
Figure 4. Molecular structure of trans-29 determined by single-crystal X-
ray diffraction analysis. Only one of two crystallographically nonequiva-
lent molecules of trans-29 is shown (see the Supporting Information).
mal load of optically active 2H-azirnes, for example, during
distillation, should be avoided to exclude undesirable race-
mization, which occurs most probably via vinyl nitrenes as
short-lived intermediates.[22]
To clarify the configuration stability of simple 2H-azirines,
we prepared a mixture of cis- and trans-29 by a Neber re-
Quantum chemical calculations: To investigate the mecha-
nism of the twofold elimination reaction, by starting from
biazirinyl rac-22 and leading to the alkyne 24 and two mole-
cules of acetonitrile, quantum chemical calculations were
performed. At first, DFT-calculations by using the
(U)B3LYP/6-31+G(d) method as implemented in the
GAUSSIAN 09 program package[23] were performed to lo-
calize the appropriate structures on the energy hypersurface.
All stationary points were checked by frequency analyses,
and zero-point corrections at this level are included in all
relative energies. By utilizing the restricted DFT method,
we localized a symmetrical transition state TS-sym-22–24
(C2-symmetry), which perfectly fulfils all geometrical re-
ACHTUNGTRENNUNGaction of the hydrazonium iodide 28, which was accessible
from ketone 26[21] via hydrazone 27 (Scheme 6). After sepa-
quirements
of
a
coarctate
reaction
mechanism
(Scheme 7).[24] The calculated barrier (37.2 kcalmolꢀ1) fits
reasonably well to the reaction conditions (thermolysis in
DMSO at 75–1358C), and the reaction is predicted to be
exothermic by ꢀ22.0 kcalmolꢀ1. By using UB3LYP, we were
able to localize another transition with asymmetrical struc-
ture and significant diradical character TS-asym-22–24 (S2
expectation value of 0.819) with a slightly lower relative
energy of 32.2 kcalmolꢀ1. The pyridazine 23, which is ob-
tained in a small yield as a byproduct, is calculated to be
lower in energy by ꢀ36.7 kcalmolꢀ1 relative to rac-22, the
pyrimidine 25, obtained by irradiation, is lower by
ꢀ60.8 kcalmolꢀ1. The formation of the pyridazine may be
explained by multistep diradical pathways, which were cal-
culationally investigated in one of our earlier studies.[8b] Due
to the size of the system studied, high-level calculations to
appropriately treat both open-shell and closed-shell species
were not possible. Therefore, we continued our studies by
using the simpler model system 30, in which the phenyl
groups of the experimentally studied system were replaced
by methyl groups (Scheme 8 and Figure 5). Here, the
Scheme 6. Synthesis and equilibration of azirines 29.
ration of cis- and trans-29 by simple chromatography, assign-
ment of the diastereomers was performed with the help of
NMR spectroscopy, especially NOE experiments, and con-
firmed by single-crystal X-ray diffraction analysis (Figure 4).
Flash vacuum pyrolysis (FVP) of both spirocyclic com-
pounds led nearly quantitatively (93–95%) to 1:1 mixtures
of the stereoisomeric heterocycles. Equilibration was also
observed on heating cis- or trans-29 in toluene at 758C. But
in these cases, the reactions were accompanied by decompo-
sition. On irradiation of cis- or trans-29 in chloroform at
ꢀ408C, no equilibration could be detected. Thus, the ther-
Chem. Eur. J. 2011, 17, 10071 – 10080
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10075

K. Banert, R. Herges, E.-U. Wꢀrthwein et al.
Fock (HF) determinant, see
HF values reported below).
Thus, to approximately treat
both open-shell and closed-
shell species CCSD(T) on an
equal footing, single-point cal-
culations on these two transi-
tion-state geometries by using
the restricted (CCSD(T)) and
the unrestricted UCCSD(T)-
method were performed. The
corresponding results are ex-
pected to have larger errors
than typical for CCSD(T) be-
cause of the partial multirefer-
ence (MR) character involved
in the electronic structures.
Again, the unrestricted method
gives
a
lower barrier of
36.8 kcalmolꢀ1 in favor of the
unsymmetrical open-shell tran-
sition structure TS-asym-30–31
with significant open-shell/MR
character (S2 =1.616), whereas
for the symmetrical transition
state TS-sym-30–31, which also
has significant open-shell/MR
character (S2 =1.399), a slightly
higher barrier of 39.3 kcal
molꢀ1 is calculated. Note, that
a perfectly localized diradical
would yield an S2 value of
unity with Hartree–Fock. The
formation of the final products
31 plus two molecules of aceto-
Scheme 7. Calculated reaction pathways for the conversion of rac-22 into 24 involving two possible transition
states TS-sym-22–24 and TS-asym-22–24. The pyridazine 23 and pyrimidine 25 are also included. Relative en-
ergies at the (U)B3LYP/6-31+G(d) level [kcalmolꢀ1].
nitrile
is
exothermic
by
ꢀ9.3 kcalmolꢀ1. The computed
barriers are in reasonable
agreement with those from the
corresponding B3LYP treat-
ments.
Scheme 8. Calculated reaction pathways for the conversion of model compound rac-30 into 31 involving two
possible transition states TS-sym-30–31 and TS-asym-30–31. Geometries at the (U)B3LYP/6-31+G(d) level of
theory, for relative energies at various levels see text.
In
summary,
both
(U)CCSD(T) and (U)B3LYP
methods predict two possible
B3LYP-activation barriers were calculated 34.5 (TS-sym-30–
31, closed shell) and 30.1 kcalmolꢀ1 (TS-asym-30–31, open
shell with S2 =0.783) with an exothermicity to form 31 and
two molecules of acetonitrile of ꢀ20.7 kcalmolꢀ1. These
data indicate that the phenyl groups have only a minor in-
fluence on the reaction mechanism and its thermodynamic
data. Both transition states interconnect the starting materi-
al with the products without any other intermediates as
demonstrated by IRC calculations.
competing transition states with symmetrical (coarctate) and
asymmetrical geometry, both showing significant open-shell/
MR character. From the computational point of view, we
have reached here the present limit of the theoretical meth-
ods applicable, considering the size of the system studied.
Reasonable CASSCF or MRCI calculations, for example,
are not feasible for such a system. The coarctate transition
presented here offers a theoretically interesting alternative
to the asymmetrical transition state.
However, it seems as if the electronic states of these tran-
sition structures are very complicated and not even two-con-
figurational but truly multiconfigurational, which is in agree-
ment with the large S2 values (>1 in the case of a Hartree–
10076
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10071 – 10080

Bi-2H-azirin-2-yls Prepared from 1,4-Diazidobuta-1,3-dienes
FULL PAPER
at 30.4 MHz by utilizing MeNO2 as an external standard with (d=0). MS
and HRMS (ESI) spectra were recorded with an Applied Biosystems
Mariner 5229 mass spectrometer or Bruker micrOTOF-QII spectrometer.
Elemental analyses were performed with a Vario EL elemental analyzer
from Elementar Analysensysteme GmbH Hanau or with a Vario Micro
Cube from Elementar. Elemental analyses of explosive azides and highly
unstable azirines could not be performed. TLC was performed with Ma-
cherey–Nagel Polygram SIL G/UV254 polyester sheets. Flash column
chromatography was performed with 32–63 mm silica gel. HPLC pump 64
(Knauer) and a UV/VIS photometer (l=254 nm, Knauer) were utilized
for isolation of compounds by HPLC.
General procedure for photolysis: Irradiation was conducted by using a
high-pressure mercury lamp (TQ150, Heraeus GmbH) supplied with
glass equipment and an ethanol cryostat (ꢀ50 to ꢀ808C). Most of these
photolyses were monitored by NMR spectroscopic analysis by utilizing
anhydrous CDCl3 (ꢀ208C to ꢀ608C) or CD2Cl2 (ꢀ858C). A solution of
the appropriate starting material (typical concentration: 0.1 to
0.01 molLꢀ1) was irradiated in an NMR tube. To exclude oxygen, the so-
lution was flushed with argon in an ultrasonic bath prior to irradiation.
Dioxane or DMSO were used as standard to determine yields based on
1H NMR spectroscopic data.
Single-crystal X-ray diffraction analysis: The data collection for meso-16,
trans-29, and rac-22 were performed at 173 K on a Bruker Smart CCD 1k
diffractometer and for meso-22 at 293 K on an Oxford Gemini S diffrac-
tometer. All calculations, including structural solutions and refinement
by full-matrix least-squares on F2, were performed by using the
SHELXTL program.[25] CCDC-810980 (trans-29), 810981 (rac-22),
810982 (meso-16), and 810983 (meso-22) contain the supplementary
crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
Figure 5. Calculated structures of the two transition states TS-sym-30–31
(top) and TS-asym-30–31 (bottom; (U)B3LYP/6-31+G(d)).
Conclusion
Methyl 2-[(Z)-2-azido-2-(methoxycarbonyl)ethenyl]-2H-azirine-3-carbox-
ylate [(Z)-11]: Photolysis of (Z,Z)-10[11] in CDCl3 was performed at
ꢀ608C to give a maximum yield of (Z)-11 of 26% (1H NMR spectrosco-
py). 1H NMR (CDCl3): d=3.29 (d, 3J=8.7 Hz, 1H; 2-H), 3.82 (s, 3H;
OMe), 4.02 (s, 3H; OMe), 5.58 ppm (d, 3J=8.7 Hz, 1H; 1’-H); 13C NMR
(CDCl3): d=34.35 (d; C-2), 53.19 (q; OMe), 54.27 (q; OMe), 126.13 (d;
C-1’), 130.37 (s; C-2’), 157.30 (s), 162.24 (s), 165.89 ppm (s).
Clearly, the substitution pattern plays a dominant role when
1,4-diazidobuta-1,3-dienes lose dinitrogen to generate the
corresponding 2H-azirines. The nearly quantitative forma-
tion of bi-2H-azirin-2-yls 22 from diazides 21 is a felicitous
case, which allows us to study aromatization, epimerization,
and fragmentation reactions in detail. The reaction mecha-
nisms postulated for these transformations are based on
quantum chemical calculations and include diradical inter-
mediates[8b] and one-step processes with competing transi-
tion states with open-shell/multireference (MR) character
possessing symmetrical (coarctate) and asymmetrical geom-
etry.
Thermolysis of (Z,Z)-10 in CDCl3 at 608C led to a maximum proportion
of (Z)-11 of 2%, whereas the pyridazine 13 was formed in 27% yield.
The latter compound was prepared through a known[12] procedure for
comparison.
Photolysis of (Z,Z)-10 in the presence of cyclopentadiene or 2,3-dime-
thylbuta-1,3-diene: A solution of (Z,Z)-10 in a 4:1 mixture of CDCl3 and
cyclopentadiene was irradiated in an NMR tube at ꢀ208C until liberation
of nitrogen was complete. The products meso-16 and rac-16 were formed
quantitatively in a ratio 4:5. When a solution of (Z,Z)-10 (500 mg,
2.0 mmol) in a mixture of CHCl3 and cyclopentadiene was treated simi-
larly but without photolyzing to completeness, a mixture of (Z)-15, meso-
16, and rac-16 was generated. Separation was achieved by chromatogra-
phy using silica gel and Et2O (for (Z)-15), Et2O/acetone 3:1 (for rac-16),
and then acetone (for meso-16). If 2,3-dimethylbuta-1,3-diene was uti-
lized instead of cyclopentadiene in the analogous photolysis of (E,E)-10,
compound (Z)-14 was isolated with 35% yield as the only product, which
could be detected.
Experimental Section
Caution! Care must be taken in handling azides, which are explosive. Es-
pecially, neat azides can lead to large explosions on friction, impact, or
heating.
Methyl
7-[(Z)-2-azido-2-(methoxycarbonyl)ethenyl]-3,4-dimethyl-1-
Instrumentation and measurement: Melting points were determined with
a Pentakon Dresden Boetius apparatus. FTIR spectra were recorded
with a Bruker IFS 28 FTIR spectrophotometer. IR measurements were
made on solutions in KBr cuvettes. UV/Vis spectra were acquired with
Specord UV VIS from Carl Zeiss Jena. 1H NMR spectra were recorded
with Varian Gemini 2000 or Unity Inova 400 spectrometers operating at
300 and 400 MHz, respectively. By using the same spectrometers,
13C NMR spectroscopic data were recorded at 75 and 100 MHz. NMR
signals were referenced to TMS (d=0) or solvent signals and recalculat-
ed relative to TMS. The multiplicities of 13C NMR spectroscopic signals
were determined with the aid of DEPT135 experiments. 31P NMR spec-
troscopic data were recorded at 121.5 MHz by using 85% H3PO4 as an
external standard with (d=0), whereas 15N NMR spectra were measured
azabicycloACTHNUTRGNE[UNG 4.1.0]hept-3-ene-6-carboxylate ((Z)-14): Yellow liquid;
1H NMR (CDCl3): d=1.52 (s, 3H; Me), 1.66 (s, 3H; Me), 2.53 (d, 2J=
18.2 Hz, 1H; H-5), 2.69 (d, 2J=18.2 Hz, 1H; H-5), 3.00 (d, 3J=8.6 Hz,
2
2
1H; H-7), 3.27 (d, J=17.5 Hz, 1H; H-2), 3.64 (d, J=17.5 Hz, 1H; H-2),
3.71 (s, 3H; OMe), 3.77 (s, 3H; OMe), 6.06 ppm (d, 3J=8.6 Hz, 1H; H-
1’); 13C NMR (CDCl3): d=16.27 ppm (q; Me), 18.66 (q; Me), 28.42 (t; C-
5), 39.33 (d; C-7), 47.35 (s; C-6), 52.36 (t; C-2), 52.39 (q; OMe), 52.69 (q;
OMe), 119.23 (s), 120.34 (s), 126.38 (d; C-1’), 129.94 (s; C-2’), 162.44 (s;
C-3’), 170.99 ppm (s; C=O).
Methyl
3-[(Z)-2-azido-2-(methoxycarbonyl)ethenyl]-2-azatricy-
clo[3.2.1.02,4]oct-6-ene-4-carboxylate ((Z)-15): Orange liquid; IR (CCl4):
n˜ =2128 cmꢀ1 (N3), 1725 (C=O); 1H NMR (CDCl3): d=1.65 (brd, 2J=
Chem. Eur. J. 2011, 17, 10071 – 10080
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10077

K. Banert, R. Herges, E.-U. Wꢀrthwein et al.
8.1 Hz, 1H; 8-H), 1.96 (dt, 2J=8.1, J=1.7 Hz, 1H; 8-H), 2.48 (d, 3J=
8.4 Hz, 1H; 3-H), 3.55 (m, 1H; 5-H), 3.69 (s, 3H; OMe), 3.72 (s, 3H;
OMe), 4.15 (m, 1H; 1-H), 5.71 (ddd, 3J=5.4, J=2.5, J=1.0 Hz, 1H; 7-
H), 6.11 (d, 3J=8.4 Hz, 1H; 1’-H), 6.20 ppm (ddd, 3J=5.4, J=3.3, J=
1.6 Hz, 1H; 6-H); 13C NMR CDCl3): d=46.96, 49.26, 50.35 (s; C-4),
52.42, 52.60, 59.30 (t; C-8), 67.08 (d; C-1), 124.56 (d; C-1’), 128.46 (d; C-
7), 129.00 (s; C-2’), 132.93 (d; C-6), 162.53 (s; C-3’), 171.05 ppm (s; C-1’’).
rac-22 was detected when a light-protected solution of (E,E)-21 was
stored some weeks at ambient temperature. After irradiation of a solu-
tion of (E,E)-21 (10 mg, 0.018 mmol) in anhydrous CDCl3 (ca. 0.7 mL)
for 5 min at ꢀ408C, the starting material was consumed completely, and
a 1:1 mixture of meso- and rac-22 was generated in 91% yield (1H NMR
spectroscopy).
Product rac-22: M.p. 144–1468C (MeOH/H2O), 147–1498C (CCl4/
meso-4,4’-Bis(methoxycarbonyl)bi-2-azatricyclo[3.2.1.02,4]oct-6-en-3-yl
(meso-16): Yellow solid; m.p. 1508C (Et2O); IR (CCl4): n˜ =1742 cmꢀ1
(C=O); 1H NMR (CDCl3): d=1.59 (brd, 2J=8.1 Hz, 2H; 8-H), 1.87 (dt,
2J=8.1, J=1.9 Hz, 2H; 8-H), 1.97 (s, 2H; 3-H), 3.47 (m, 2H; 5-H), 3.67
(s, 6H; OMe), 3.99 (m, 2H; 1-H), 5.57 (ddd, 3J=5.4, J=2.6, J=1.1 Hz,
ꢀ1
ꢀ
hexane); IR (CCl4): n˜ =3061, 1778 (C=N), 1438 (P Ph), 1195 cm (P=
O); UV/Vis (CH3CN): lmax (lg e)=212 (4.51), 225 nm (4.49); 1H NMR
(CDCl3): d=1.98 (s, 6H; Me), 7.32–7.55 (m, 12H; Ph), 7.61 (m, 4H; Ph),
7.92 ppm (m, 4H; Ph); 13C NMR (CDCl3, 62.89 MHz): d=13.75 (q, 1J-
A
R
ACHTUNGTRENNUNG
3
2H; 7-H), 6.06 ppm (ddd, J=5.4, J=3.4, J=1.7 Hz, 2H; 6-H); 13C NMR
A
E
ACHTUNGTRENNUNG
1
N
(31P, 13C)+4J
ACHUTGTNRENNUG CAHTUNGTRENNUNG
(CDCl3): d=46.67, 47.64 (s; C-4), 51.22, 52.21, 58.98 (t; C-8), 66.59 (d; C-
1), 128.39 (d; C-7), 132.75 (d; C-6), 171.80 ppm (s; C=O); elemental anal-
ysis calcd (%) for C18H20N2O4 (328.37): C 65.84, H 6.14, N 8.53; found: C
65.33, H 6.13, N 8.58.
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
rac-4,4’-Bis(methoxycarbonyl)bi-2-azatricyclo[3.2.1.02,4]oct-6-en-3-yl (rac-
16): Yellow liquid; IR (CCl4): n˜ =1727 cmꢀ1 (C=O); 1H NMR (CDCl3):
d=1.59 (brd, 2J=8.1 Hz, 2H; 8-H), 1.87 (dt, 2J=8.1, J=1.9 Hz, 2H; 8-
H), 1.97 (s, 2H; 3-H), 3.47 (m, 2H; 5-H), 3.67 (s, 6H; OMe), 3.99 (m,
AHCTUNGTRENNUNG
(13C,1 H)=9 Hz; C-3); the carbon atoms and both phosphorus atoms
form several AXXꢂ systems; another 13C NMR spectrum was measured
at 75.5 MHz to distinguish between close signals and 31P, 13C coupling;
31P NMR (CDCl3): d=32.59 ppm; 15N NMR (CDCl3): d=ꢀ105.89 ppm
3
(’t’, 2J(31P, 15N)+3J(31P, 15N)=5.4 Hz); MS (ESI): m/z: 509.2 [M++1]; ele-
ACTHNUTRGENNUG CAHTUNGTRENNUGN
2H; 1-H), 5.57 (ddd, J=5.4, J=2.6, J=1.1 Hz, 2H; 7-H), 6.06 ppm (ddd,
3J=5.4, J=3.4, J=1.7 Hz, 2H; 6-H); 13C NMR (CDCl3): d=46.45, 47.18
(s; C-4), 52.19, 52.61, 59.03 (t; C-8), 66.78 (d; C-1), 128.78 (d; C-7),
132.51 (d; C-6), 171.61 ppm (s; C=O); elemental analysis calcd (%) for
C18H20N2O4 (328.37): C 65.84, H 6.14, N 8.53; found: C 65.49, H 5.64, N
8.44.
mental analysis calcd (%) for C30H26N2O2P2 (508.49): C 70.86, H 5.15, N
5.51; found: C 70.20, H 5.05, N 5.29.
1
Product meso-22: M.p. 172–1748C (benzene); H NMR (CDCl3): d=2.01
(s, 6H; Me), 7.22–7.60 (m, 16H; Ph), 8.04 ppm (m, 4H; Ph); 13C NMR
(CDCl3): d=13.05 (Me), 37.46 (dd, 1J(31P, 13C)+2J(31P, 13C)=130.5 Hz;
ACTHNUTRGENNUG CAHTUNGTRENNUGN
Photolysis of diazides 17: A solution of 17 (symmetrical major isomer,[11]
11 mg, 0.025 mmol) in CDCl3 (0.75 mL) was irradiated at ꢀ508C for
13 min to give the corresponding highly unstable 2-[2-azido-3-(phenylsul-
fonyl)prop-1-enyl]-3-[(phenylsulfonyl)methyl]-2H-azirine (18) in 67%
yield (1H NMR spectroscopy). IR (CDCl3): n˜ =2126 (N3), 1324 (SO2),
1154 cmꢀ1 (SO2); 1H NMR (CDCl3/ꢀ508C): d=2.62 (d, 3J=8.4 Hz, 1H;
AXXꢂ system, C-2), 127.7–128.5 (several signals), 130.04, 130.63, 131.2–
132.2 (several signals), 166.63 ppm (C-3); 31P NMR (CDCl3): d=
31.66 ppm; MS (ESI): m/z: 509.2 [M++1]; elemental analysis calcd (%)
for C30H26N2O2P2·C6H6 (586.62): C 73.71, H 5.50, N 4.77; found: C 73.18,
H 5.13, N 4.64.
Thermolysis of rac- and meso-22: Solutions of 5–10 mg of rac- or meso-
22 in [D6]DMSO (ca. 0.7 mL) were heated in sealed NMR tubes at 758C
CH N), 3.81 (d, 2J=14.3 Hz, 1H; CH2), 4.06 (d, 2J=14.3 Hz, 1H; CH2),
4.64 (s, 2H; CH2), 4.76 (d, J=8.4 Hz, 1H; CH=), 7.56–7.76 (m, 6H; Ph),
ꢀ
3
1
to monitor the thermolyses by H NMR spectroscopy (see Figure 3). The
7.88–8.00 ppm (m, 4H; Ph); the same product was formed in 26% yield
maximum yield of 23 was 10%. After heating a solution of rac-22 (1.50 g,
2.95 mmol) in DMSO (20 mL) at 1208C for 3 h, the solvent including the
formed MeCN was removed in vacuo (0.001 Torr) at ambient tempera-
ture. The product mixture (1.35 g) was treated with Et2O, which led to a
precipitate of the known[16] 24 as a beige solid. Filtration with suction
gave nearly pure 24 (1.16 g, 92%), which contained only traces of 23.
The filtrate was purified by chromatography (silica gel, Et2O/MeOH
10:1, yellow fraction). After removal of the solvent, addition of Et2O
gave 23 (48 mg, 3%) as a yellow solid.
when (E,Z)-17[11] was photolyzed similarly. In this case, the other diaste-
reomer of 18 was generated with 60% yield. 1H NMR (CDCl3): d=2.65
3
(d, J=8.1 Hz, 1H; CH N), 3.90 (s, 2H; CH2), 4.22 (d, 3J=8.1 Hz, 1H;
ꢀ
CH=), 4.55 (s, 2H; CH2), 7.56–7.76 (m, 6H; Ph), 7.88–8.00 ppm (m, 4H;
Ph).
3,6-Bis[(phenylsulfonyl)methyl]pyridazine (20): A solution of (E,Z)-17
(2.0 mg, 0.0045 mmol) in [D6]DMSO (0.78 g) was heated at 81.58C for
2.5 h to achieve the formation of 20 in a maximum yield of 37%
(1H NMR spectroscopy). When 17 (symmetrical major isomer) was treat-
ed similarly, the product 20 was generated in a maximum yield of 16%
after 30 min. After heating 17 (mixture of diastereomers)[11] in boiling
CH2Cl2 for 5 h, the product 20 could be isolated as a white solid. M.p.
232–2358C (CH2Cl2, decomp.); IR (CDCl3): n˜ =1323 (SO2), 1155 cmꢀ1
(SO2); 1H NMR (CDCl3): d=4.72 (s, 4H; CH2), 7.43–7.50 (m, 4H; Ph),
7.58–7.65 (m, 6H; Ph), 7.82 ppm (s, 2H; CH); 1H NMR ([D6]DMSO):
d=5.08 (s, 4H; CH2), 7.55–7.77 ppm (m, 12H; Ar, CH); 13C NMR
([D6]DMSO): d=60.46 (t), 127.91 (d), 129.24 (d), 129.27 (d), 134.13 (d),
138.05 (s, i-Ph), 152.29 ppm (s); HRMS (ESI): m/z: calcd for
C18H17N2O4S2: 389.0635 [M+H]+; found: 389.0624.
4,5-Bis(diphenylphosphinoyl)-3,6-dimethylpyridazine (23): M.p. 258–
2608C (CH2Cl2/hexane); 1H NMR (CDCl3): d=2.53 (s, 6H; Me), 7.16–
7.56 ppm (m, 20H; Ph); 13C NMR (CDCl3): d=24.68, 128.05 (’t’, 3J-
(31P, 13C)+6J(31P, 13C)=12.6 Hz; m-Ph), 131.94 (d, 1J
AHCTUNGTREUNNGN ACHTUNGTRENNUNG ACTUHNGTRENNUGN
Ph), 131.9–132.4 (several signals, o- and p-Ph), 138.48 (d, 1J
ACHTUNGTRENNUNG
91 Hz; C-4/C-5), 159.27 (t, 2J(31P, 13C)+3J
ACTHNUGTRENNUGN ACHTUNGTRENNNUG
al AXX’ systems; 31P NMR (CDCl3): d=34.36 ppm; HRMS (ESI): m/z:
calcd for C30H26N2O2P2: 509.1542 [M+H]+; found: 509.1558; elemental
analysis calcd (%) for C30H26N2O2P2·1= H2O (517.51): C 69.63, H 5.26, N
2
5.41; found: C 69.41, H 4.78, N 5.27.
Photolysis of rac- and meso-22: Irradiation of solutions of rac- or meso-
22 in CDCl3 (ca. 0.02 molLꢀ1) was performed at ꢀ408C and monitored
by 1H and 31P NMR spectroscopy (see the Supporting Information). The
maximum yield of 25 was ca. 22% (30–50% unreacted starting material)
beside up to 3% 24 and MeCN. Very strict exclusion of moisture was
necessary, otherwise completely different main products were formed
(see the Supporting Information). After irradiating a solution of meso-22
(500 mg, 0.983 mmol) in anhydrous CHCl3 (100 mL, degassed and flushed
with dry argon) at ꢀ408C for 35 min, the solvent was removed in vacuo,
and the residue was purified by flash chromatography by using silica gel
and Et2O/MeOH 10:1. The fraction containing 25 was liberated from the
solvent and treated with Et2O/hexane (stirring for 1 h). This led to a pre-
cipitate of 25 (30 mg, 6%) as a yellow solid.
2,2’-Bis(diphenylphosphinoyl)-3,3’-dimethylbi-2H-azirin-2-yl (22): A solu-
tion of (E,E)-21[11] (0.50 g, 0.89 mmol) in CHCl3 (20 mL) was heated to
reflux (bath temperature: 808C) for 16 h. After removal of the solvent,
the residue was recrystallized from MeOH/H2O and dried over P2O5 to
give rac-22 (278 mg, 62%). The mother liquor was extracted with CHCl3,
and the organic layer was dried with MgSO4. After removal of the sol-
vent, nearly pure meso-22 (137 mg, 30%) was isolated as a viscous
orange oil, which crystallized after addition of Et2O. Thermolysis of
(E,E)-21 or a mixture of diastereomers 21[11] could be performed on a
10 g scale, and workup was possible also by recrystallization from ben-
zene first to obtain meso-22 followed by concentration of the mother
liquor and purification of rac-22 by recrystallization from CH2Cl2/hexane.
Very slow formation of the corresponding mono-azirine and meso- and
10078
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10071 – 10080

Bi-2H-azirin-2-yls Prepared from 1,4-Diazidobuta-1,3-dienes
FULL PAPER
4,5-Bis(diphenylphosphinoyl)-2,6-dimethylpyrimidine (25): M.p. 221–
2238C (CH2Cl2/hexane); 1H NMR (CDCl3): d=2.67 (s, 3H; Me), 2.76 (s,
3H; Me), 7.2–7.6 (m, 16H; Ph), 7.80 ppm (m, 4H; Ph); 13C NMR
previously.[27] Before use, the tube was dried for 2 h at 4508C and
10ꢀ5 Torr. Starting with cis-29 (45 mg, 0.19 mmol), led to a 1:1 mixture of
cis- and trans-29 (43 mg, 95%). By using trans-29 (44 mg, 0.18 mmol), the
same procedure gave also a 1:1 mixture of cis- and trans-29 (41 mg,
93%).
(CDCl3): d=25.51, 27.29, 127.76 (d, 3J
(d, 3J(31P, 13C)=12.6 Hz; m-Ph), 128.92 (dd, 1J
ACHTUNGTRENNUNG
A
E
ACHTUNGTRENNUNG
4
2
18.3 Hz; C-5), 131.49 (d, J
(31P, 13C)=2.9 Hz; p-Ph), 131.59 (d, J
E
A
4
1
9.2 Hz; o-Ph), 131.87 (d, J
(31P, 13C)=
ACHTUNGTRENNUNG
106 Hz; i-Ph), 132.49 (d, 2J
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
E
ACHTUNGTRENNUNG
(31P, 13C)=10.3 Hz; C-4), 173.97 ppm (m; C-2);
Acknowledgements
31P NMR (CDCl3): d=22.77, 31.77 ppm; HRMS (ESI): m/z: calcd for
C30H26N2O2P2: 509.1542 [M+H]+; found 509.1543.
Financial support of this work by the Deutsche Forschungsgemeinschaft
(BA 903/10/1-3) is gratefully acknowledged. We thank also Dr. J. Leh-
mann and Dr. N. Ramezanian for assistance with kinetic analysis and
with the manuscript. We are grateful to Prof. S. Grimme for valuable
hints and supporting discussions.
4-tert-Butylcyclohexyl phenyl ketone N,N-dimethylhydrazone (27): By
starting with ketone 26[21] (9.00 g, 36.8 mmol), N,N-dimethylhydrazine
(3.00 g, 49.9 mmol), sodium acetate (0.52 g, 6.3 mmol), and three drops of
acetic acid, the product was prepared by using an analogous procedure
from reference [26]. After recondensation at 1308C/0.001 Torr, hydra-
zone 27 (6.70 g, 64%) was isolated as a yellow oil, which was not purified
but used directly for the next step. 1H NMR (CDCl3): d=0.78 (s, 9H;
tBu), 0.8–2.0 (m), 2.32 (s, 6H; Me), 7.10–7.60 ppm (m, 5H; Ph);
13C NMR (CDCl3): d=27.05 (t), 27.52 (q), 31.23 (t), 32.39 (s), 47.21 (d),
47.31 (q), 47.45 (d), 126.81 (d), 127.36 (d), 127.82 (d), 138.00 (s),
169.07 ppm (s).
[1] Reviews: a) E. E. van Tamelen, Angew. Chem. 1965, 77, 759–767;
Angew. Chem. Int. Ed. Engl. 1965, 4, 738–745; b) E. E. van Tame-
len, Acc. Chem. Res. 1972, 5, 186–192; c) A. H. Schmidt, Chem.
Unserer Zeit 1977, 11, 118–128.
[2] a) R. Breslow, P. Gal, J. Am. Chem. Soc. 1959, 81, 4747–4748; b) R.
Breslow, P. Gal, H. W. Chang, L. J. Altman, J. Am. Chem. Soc. 1965,
87, 5139–5144.
[3] a) J. H. Davis, K. J. Shea, R. G. Bergman, Angew. Chem. 1976, 88,
254–255; Angew. Chem. Int. Ed. Engl. 1976, 15, 232–234; b) J. H.
Davis, K. J. Shea, R. G. Bergman, J. Am. Chem. Soc. 1977, 99, 1499–
1507 and references therein; c) W. H. de Wolf, I. J. Landheer, F.
Bickelhaupt, Tetrahedron Lett. 1975, 16, 179–182.
[4] For summaries, see: a) A. Padwa, S. I. Goldstein, R. J. Rosenthal, J.
Org. Chem. 1987, 52, 3278–3285; b) F. Bickelhaupt, W. H. de Wolf,
Recl. Trav. Chim. Pays-Bas 1988, 107, 459–478.
4-tert-Butylcyclohexyl phenyl ketone N,N,N-trimethylhydrazonium
iodide (28): By starting with hydrazone 27 (13.40 g, 46.78 mmol) and
methyl iodide (12.50 g, 88.07 mmol) in anhydrous Et2O (15 mL), the
product was prepared by using an analogous procedure from refer-
ence [26]. The hydrazonium iodide 28 (16.90 g, 84%) was isolated as a
yellow solid, which was not purified but utilized directly for the next
step. M.p. 134–1378C; 1H NMR (CDCl3): d=0.80 (s, 9H; Me), 0.90–1.10
(m, 2H), 1.20–1.30 (m, 2H), 1.80–2.00 (m, 5H), 2.42 (m, 1H), 3.59 (s,
9H; Me), 7.20 (m, 2H; Ph), 7.50 ppm (m, 3H; Ph); 13C NMR (CDCl3):
d=26.39 (t), 27.22 (q), 29.89 (t), 32.14 (s), 46.86 (d), 50.22 (d), 57.89 (q),
126.15 (d), 128.88 (d), 130.10 (d), 131.66 (s), 181.39 ppm (s).
[5] a) A. Padwa, T. J. Blacklock, P. H. J. Carlsen, M. Pulwer, J. Org.
Chem. 1979, 44, 3281–3287; b) T. C. Gallagher, R. C. Storr, Tetrahe-
dron Lett. 1981, 22, 2905–2908.
[6] A. Padwa, J. Smolanoff, A. Tremper, J. Org. Chem. 1976, 41, 543–
549.
6-tert-Butyl-2-phenyl-1-azaspiroACTHNUTRGNEUGN[2,5]oct-1-ene (29): After reacting
sodium (0.19 g, 8.3 mmol) with 2-propanol (25 mL), the resulting solution
was used to treat 28 (3.80 g, 8.87 mmol) as described in an analogous pro-
cedure in reference [26]. Workup by recondensation (1158C/0.001 Torr)
gave crude products (1.90 g), which were purified and separated by flash
chromatography (silica gel, Et2O/hexane 1:4, elution order: cis-29, then
trans-29) to yield both cis-29 (418 mg, 20%) and trans-29 (532 mg, 25%)
as colorless crystals.
[7] A. Hassner, J. Keogh, Tetrahedron Lett. 1975, 16, 1575–1578.
[8] a) K. Banert, F. Kçhler, Angew. Chem. 2001, 113, 173–176; Angew.
Chem. Int. Ed. 2001, 40, 174–177; b) K. Banert, S. Grimme, R.
Herges, K. Heß, F. Kçhler, C. Mꢀck-Lichtenfeld, E.-U. Wꢀrthwein,
Chem. Eur. J. 2006, 12, 7467–7481.
[9] Reviews on 2H-azirines: a) J. Backes in Houben–Weyl, Vol. E 16c
(Ed.: D. Klamann), Thieme, Stuttgart, 1992, pp. 321–369; b) V. Nair
in The Chemistry of Heterocyclic Compounds, Small-Ring Heterocy-
cles, Vol. 42, Part 1 (Ed.: A. Hassner), Wiley, New York, 1983,
pp. 215–332; c) W. H. Pearson, B. W. Lian, S. C. Bergmeier in Com-
prehensive Heterocyclic Chemistry II, Vol. 1A (Ed.: A. Padwa), Per-
gamon, New York, 1996, pp. 1–60; d) F. Palacios, A. N. Ochoa de R-
etana, E. Martinez de Marigorta, J. M. de Los Santos, Eur. J. Org.
Chem. 2001, 2401–2414; e) T. L. Gilchrist, Aldrichimica Acta 2001,
34, 51–55; f) K. M. L. Rai, A. Hassner in Advances in Strained and
Interesting Organic Molecules, Vol. 8 (Ed.: B. Halton), Jai Press,
Greenwich, 2000, pp. 187–257.
[10] J. H. Boyer, W. E. Krueger, R. Modler, Tetrahedron Lett. 1968, 9,
5979–5982.
[11] K. Banert, F. Kçhler, A. Melzer, I. Scharf, G. Rheinwald, T. Rꢀffer,
H. Lang, Synthesis 2011,1561–1568.
[12] T. Hierstetter, B. Tischler, J. Sauer, Tetrahedron Lett. 1992, 33,
8019–8022.
[13] a) P. Bhullar, T. L. Gilchrist, P. Maddocks, Synthesis 1997, 271–272;
b) M. J. Alves, T. L. Gilchrist, J. Chem. Soc. Perkin Trans. 1 1998,
299–303; c) M. J. Alves, J. F. Bickley, T. L. Gilchrist, J. Chem. Soc.
Perkin Trans. 1 1999, 1399–1401; d) T. L. Gilchrist, M. J. Alves in
Organic Azides: Syntheses and Applications (Eds.: S. Brꢃse, K.
Banert), Wiley, New York, 2010, pp. 167–190.
Product cis-29: M.p. 79–818C (hexane); IR (CCl4): n˜ =2943, 1727 cmꢀ1
2
(C=N); 1H NMR (CDCl3): d=0.92 (s, 9H; Me), 1.14 (brd, J4eq,4ax
=
13.5 Hz, 2H; 4eq-H), 1.20 (tt, 3J6ax,5ax =12.0, 3J6ax,5eq =3.0 Hz, 1H; 6ax-H),
1.56 (qd, 2J5ax,5eq =3J5ax,6ax =3J5ax,4ax =12.5, 3J5ax,4eq =3.8 Hz, 2H; 5ax-H),
1.81 (m, 2H; 5eq-H), 2.10 (td, J4ax,4eq =3J4ax,5ax =12.5, J4ax,5eq =4.0 Hz, 2H;
4ax-H), 7.52 (m, 3H, Ph), 7.77 ppm (m, 2H, Ph); the assignment of cis
stereochemistry was supported by 1H NMR spectroscopic NOE experi-
ments that indicated an interaction between 4ax-H and ortho protons;
13C NMR (CDCl3): d=25.30 (t; C-5 und C-7), 27.65 (q; tBu), 32.69 (s),
35.05 (t; C-4 und C-8), 40.03 (s), 48.03 (d; C-6), 126.36 (s), 128.77 (d),
129.00 (d), 132.43 (d), 178.65 ppm (s; C-2); MS (ESI): m/z: 242.2 [M+
+1]; elemental analysis calcd (%) for C17H23N (241.38): C 84.59, H 9.60,
N 5.80; found: C 84.29, H 9.55, N 5.68.
2
3
Product trans-29: M.p. 90–928C (hexane); IR (CCl4): n˜ =2943, 1725 cmꢀ1
(C=N); 1H NMR (CDCl3): d=0.95 (s, 9H; Me), 1.10–1.40 (m, 5H; 6ax-
H, 5ax-H, 4ax-H), 1.90–2.20 (m, 4H; 5eq-H, 4eq-H), 7.57 (m, 3H; Ph),
7.84 ppm (m, 2H; Ph); the assignment of trans stereochemistry was sup-
ported by 1H NMR spectroscopic NOE experiments that indicated inter-
actions between 4eq-H/5ax-H and ortho protons (see also Figure 4);
13C NMR (CDCl3): d=27.73 (q; tBu), 28.79 (t; C-5, C-7), 32.43 (s), 35.71
(t; C-4, C-8), 40.88 (s), 47.61 (d; C-6), 125.81 (s), 129.06 (d), 129.11 (d),
132.55 (d), 179.46 ppm (s; C-2); MS (ESI): m/z: 242.2 [M++1]; elemental
analysis calcd (%) for C17H23N (241.38): C 84.59, H 9.60, N 5.80; found:
C 84.16, H 9.38, N 5.56.
Thermolysis of cis- and trans-29: Flash vacuum pyrolysis was performed
at 4008C and 10ꢀ5 Torr by utilizing a quartz glass apparatus described
[14] H. Hemetsberger, D. Knittel, Monatsh. Chem. 1972, 103, 205–209.
[15] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
Chem. Eur. J. 2011, 17, 10071 – 10080
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10079

K. Banert, R. Herges, E.-U. Wꢀrthwein et al.
[16] a) H. Hartmann, C. Beermann, H. Czempik, Z. Anorg. Allg. Chem.
1956, 287, 261–272; b) E. P. Kyba, S. P. Rines, P. W. Owens, S.-S. P.
Chou, Tetrahedron Lett. 1981, 22, 1875–1878.
[17] a) P. Mendes, Comput. Appl. Biosci. 1993, 9, 563–571; b) GEPASI
homepage, http://www.gepasi.org.
[23] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.
Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,
M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.
Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.
Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.
Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O.
Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox,
Gaussian, Inc., Wallingford CT, 2009.
[24] a) R. Herges, J. Chem. Inf. Comput. Sci. 1994, 34, 91–102; b) R.
Herges, Angew. Chem. 1994, 106, 261–283; Angew. Chem. Int. Ed.
Engl. 1994, 33, 255–276.
[25] SHELXTL Version 5.1, An Integrated System for Solving, Refining
and Displaying Crystal Structures from Diffraction Data, G. M.
Sheldrick, Siemens Analytical X-ray Instruments, Madison, 1990.
[26] S. Sato, Bull. Chem. Soc. Jpn. 1968, 41, 1440–1444.
[18] G. Maier, A. Schick, I. Bauer, R. Boese, M. Nussbaumer, Chem.
Ber. 1992, 125, 2111–2117.
[19] a) F. Palacios, A. M. Ochoa de Retana, J. I. Gil, J. M. Ezpeleta, J.
Org. Chem. 2000, 65, 3213–3217; b) I. P. Piskunova, A. V. Eremeev,
A. F. Mishnev, I. A. Vosekalna, Tetrahedron 1993, 49, 4671–4676;
c) L. Gentilucci, Y. Grijzen, L. Thijs, B. Zwanenburg, Tetrahedron
Lett. 1995, 36, 4665–4668; d) M. M. H. Verstappen, G. J. A. Ariaans,
B. Zwanenburg, J. Am. Chem. Soc. 1996, 118, 8491–8492; e) F. A.
Davis, C.-H. Liang, H. Liu, J. Org. Chem. 1997, 62, 3796–3797.
[20] a) E. O. Stapley, D. Hendlin, M. Jackson, A. K. Miller, S. Hernan-
dez, J. M. Mata, J. Antibiot. 1971, 24, 42–47; b) T. W. Miller, E. W.
Tristram, F. J. Wolf, J. Antibiot. 1971, 24, 48–50; c) T. F. Molinski,
C. M. Ireland, J. Org. Chem. 1988, 53, 2103–2105; d) F. A. Davis,
G. V. Reddy, H. Liu, J. Am. Chem. Soc. 1995, 117, 3651–3652;
e) C. E. Salomon, D. H. Williams, D. J. Faulkner, J. Nat. Prod. 1995,
58, 1463–1466.
[21] a) C. I. McIntosh, Can. J. Chem. 1967, 45, 2267–2270; b) J. P. Cole-
man, F. I. Shah-Malak, J. H. P. Utley, J. Chem. Soc. B 1970, 666–670.
[22] Our results are compatible with a previous article, in which treat-
ment of a 2H-azirine at 105–1208C in decalin is described: K. Iso-
mura, G.-I. Ayabe, S. Hatano, H. Taniguchi, J. Chem. Soc. Chem.
Commun. 1980, 1252–1253.
[27] K. Banert, S. Groth, H. Hꢀckstꢃdt, J. Lehmann, J. Schlott, K.
Vrobel, Synthesis 2002, 1423–1433.
Received: April 20, 2011
Published online: July 21, 2011
10080
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10071 – 10080