Photoextrusion of molecular nitrogen from annulated 5-alkylidene-4,5-dihydro-1H-tetrazoles: Annulated iminoaziridines and the first triplet diazatrimethylenemethane

Article abstract of DOI:10.1002/(SICI)1099-0690(199802)1998:2<317::AID-EJOC317>3.0.CO;2-A

Deprotonation of the annulated tetrazolium salts 4, 6, 8, 10, and 12 with sodium or potassium hydride yields the alkyh'denedihydrotetrazoles 5, 7, 9, 11, and 13, respectively. While 5a and b are unstable, even in solution at low temperatures, 7, 9, 11, and 13 form yellow oils that are distilled under high vacuum. - Irradiation of solutions of 7, 9, and 11 in [D₈]toluene at -60°C yields, besides molecular nitrogen, annulated iminoaziridines that have an exocyclic CN double bond, i.e. 14, 16, and 18, respectively. In addition, an equal amount of the isomer 19 with the endocyclic CN double bond is formed from 11. On thermolysis, 14, 16, and 18 undergo [2 + 1] cycloreversion into methyl isocyanide and the cyclic imines 15, 17, and 20, respectively. By contrast, 19 rearranges thermally to yield 18. While the doubly bridged alkylidenedihydrotetrazole 13a affords only unidentified decomposition products on photolysis, its methyl homologue 13b is converted into the hexahydronaphthyridine 22 which is also formed on thermolysis. - Irradiation of 13b in a 2-methyltetrahydrofuran or butyronitrile matrix at 77 K yields a triplet diradical showing a four-line EPR spectrum centred at 3362 G and a half-field transition (at 1669 G) with a hyperfine structure. The zero-field splitting parameters |D/hc| = 0.031 cm^-1 and |E/hc| = 0.0014 cm^-1 are obtained by simulation of the EPR spectrum. The signal-carrier is assigned the diazatrimethylenemethane structure 23 on the basis of the close similarity between its EPR spectrum and those of trimethylenemethane (28) and tris(N-methylimino)methane (29). - Structural features are discussed that are responsible for the observed differences between the photochemical pathways.

Full text of DOI:10.1002/(SICI)1099-0690(199802)1998:2<317::AID-EJOC317>3.0.CO;2-A

FULL PAPER
Photochemical Formation of Heteromethylenecyclopropanes, 28[᭛]
Photoextrusion of Molecular Nitrogen from Annulated 5-Alkylidene-4,5-
dihydro-1H-tetrazoles: Annulated Iminoaziridines and the First Triplet
DiazatrimethylenemethaneȠ
Helmut Quast*, Andreas Fuss, and Wolfgang Nüdling
Institut für Organische Chemie der Universität Würzburg,
Am Hubland, D-97074 Würzburg, Germany
Received July 31, 1997
Keywords: Aziridines / Cleavage reactions / Diradicals / EPR spectroscopy / Photolysis / Rearrangements / Ring contrac-
tions / Small ring systems / Strained molecules / Tetrazoles / Valence isomerization
Deprotonation of the annulated tetrazolium salts 4, 6, 8, 10,
and 12 with sodium or potassium hydride yields the alkylid-
enedihydrotetrazoles 5, 7, 9, 11, and 13, respectively. While
5a and b are unstable, even in solution at low temperatures,
7, 9, 11, and 13 form yellow oils that are distilled under high
vacuum. Ϫ Irradiation of solutions of 7, 9, and 11 in [D8]tol-
photolysis, its methyl homologue 13b is converted into the
hexahydronaphthyridine 22 which is also formed on ther-
molysis. Ϫ Irradiation of 13b in a 2-methyltetrahydrofuran or
butyronitrile matrix at 77 K yields a triplet diradical showing
a four-line EPR spectrum centred at 3362 G and a half-field
transition (at 1669 G) with a hyperfine structure. The zero-
uene at Ϫ60°C yields, besides molecular nitrogen, annulated field splitting parameters ͉D/hc͉ = 0.031 cmϪ1 and ͉E/hc͉ =
iminoaziridines that have an exocyclic CN double bond, i.e. 0.0014 cmϪ1 are obtained by simulation of the EPR spectrum.
14, 16, and 18, respectively. In addition, an equal amount of
The signal-carrier is assigned the diazatrimethylenemethane
the isomer 19 with the endocyclic CN double bond is formed structure 23 on the basis of the close similarity between its
from 11. On thermolysis, 14, 16, and 18 undergo [2 + 1] cyclo- EPR spectrum and those of trimethylenemethane (28) and
reversion into methyl isocyanide and the cyclic imines 15, 17,
and 20, respectively. By contrast, 19 rearranges thermally to
tris(N-methylimino)methane (29). Ϫ Structural features are
discussed that are responsible for the observed differences
yield 18. While the doubly bridged alkylidenedihydrotetra- between the photochemical pathways.
zole 13a affords only unidentified decomposition products on
Deprotonation of 1,4,5-substituted tetrazolium salts 1 with important role in organic synthesis as demonstrated by
sodium hydride yields the yellow 5-alkylidenedihydrotetra- many diyl trapping reactions.[10] We disclose here the detec-
zoles 2.[2][3][4] Irradiation of 2 with UV light results in a tion by EPR spectroscopy of a novel nitrogen analogue of
clean extrusion of molecular nitrogen and the quantitative, trimethylenemethane, i.e. the triplet diazatrimethylenemeth-
diastereoselective formation of (E)-iminoaziridines 3 which ane 23. Furthermore, the results of this work broaden the
undergo (E) v (Z) equilibration at room temperature.[2][5][6] basis for the definition of scope and limitation of the photo-
Iminoaziridines with unsaturated substituents, i.e. 3dϪf, chemical synthesis of iminoaziridines from 5-alkylidenedi-
can be observed only at temperatures below Ϫ40°C because hydrotetrazoles.
subsequent reactions are initiated by intramolecular hydro-
gen shifts at higher temperatures.[6]
Recently, we have disclosed the synthesis of a number of
annulated tetrazolium salts, viz. 4, 6, 8, 10, and 12.[1] We
now report on their deprotonation and the irradiation of
the resulting annulated 5-alkylidenedihydrotetrazoles. This
study was stimulated by the hope that the additional strain
imposed by the second rings might render the formation of
iminoaziridines difficult, if not impossible, and thus help to
observe triplet diradicals generated by extrusion of nitro-
gen, e.g. in low-temperature matrices. This concept orig-
inates from the seminal work of Berson and coworkers[7][8]
on bridged trimethylenemethane diradicals and has recently
been extended to bridged oxyallyl diradicals.[9] It plays an
[᭛]
Part 27: Ref.[1]
.
Eur. J. Org. Chem. 1998, 317Ϫ327
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0202Ϫ0317 $ 17.50ϩ.50/0
317

H. Quast, A. Fuss, W. Nüdling
FULL PAPER
Annulated 5-Alkylidene-4,5-dihydro-1H-tetrazoles
resemble those of monocyclic alkylidenedihydrotetra-
zoles,[3][4] of course, and are indicative of highly polarised,
electron-rich “double” bonds that characterise ylides.[12][13]
Deprotonation of the annulated tetrazolium salts was
performed with sodium or potassium hydride in ether sol-
vents, sometimes in the presence of 18-crown-6 or the com-
plex from 18-crown-6 and potassium cyanide,[11] handling
of which is more convenient. The progress of the depro-
tonation reaction was indicated by the evolution of hydro-
gen and the development of a yellow colour which is
characteristic for alkylidenedihydrotetrazoles.[2][3][4] A num-
ber of runs, performed with 6a, served to optimise the ex-
perimental protocol. Deprotonation with sodium or potas-
sium hydride in diethyl ether or tert-butyl methyl ether was
very slow in the absence of 18-crown-6. Traces of the sol-
vent inevitably appeared in the distilled alkylidenedihydro-
tetrazoles. Therefore, we preferred tert-butyl methyl ether
as solvent, whose traces could be employed as standard in
photolysis experiments that were monitored by NMR spec-
troscopy. On the other hand, tetrahydrofuran and 2-methyl-
tetrahydrofuran as solvents allowed the very rapid and
quantitative deprotonation of tetrazolium salts even with
sodium hydride and in the absence of 18-crown-6, and were
hence most convenient for the preparation of solutions of
known concentration which were needed in spectroscopic
studies.
Attempts at the isolation of the products 5 of the trimeth-
ylenetetrazolium hexafluorophosphates 4 met with failure.
Deprotonation of 4a at room temperature yielded orange-
coloured, very viscous, non-volatile decomposition prod-
ucts of 5a. Experiments in [D8]tetrahydrofuran solution at
low temperatures showed the onset of deprotonation at a
temperature of Ϫ50°C. Even at that temperature, slow de-
composition occurred which only permitted characterisa-
tion of 5a by NMR spectroscopy (Tables 2 and 3). Sub-
sequent irradiation at Ϫ60°C (λ Ն 320 nm) afforded a com-
plex mixture of unidentified products. Likewise, depro-
tonation of 4b in [D8]tetrahydrofuran commenced at
Ϫ45°C. The resulting yellow solution of 5b, which was pure
according to proton and carbon-13 spectra (Tables 2 and
3), also furnished an unidentified mixture, however, on ir-
radiation at Ϫ60°C (λ Ն 320 nm).
Unlike 5a and b, which are extremely prone to decompo-
sition, all other annulated alkylidenedihydrotetrazoles, viz.
7, 9, 11, and 13, could be prepared by deprotonation at 0°C
(13a) or even room temperature and isolated by distillation
under high vacuum as deeply yellow oils or as a low-melting
solid (13b). Except 13a, which decomposes at room tem-
perature during one day, the distilled, neat alkylidenedihyd-
rotetrazoles 7, 9, 11, and 13b, are reasonably stable at room
temperature, as long as light, air, and moisture are excluded.
During five days, only about 1% of 13b lost molecular ni-
trogen to afford 22.
Photolysis in Solution
Dilute, degassed solutions of the annulated alkylidene-
dihydrotetrazoles, contained in sealed NMR sample tubes,
were irradiated with the filtered (λ Ն 305 or 320 nm), fo-
cussed light of a high-pressure mercury lamp. In a number
of experiments, low temperature (Ϫ60°C) was maintained
during irradiation and recording of NMR spectra, and pre-
cluded thermal (E) v (Z) equilibration of the photolysis
Besides the long-wavelength absorption in the UV/Vis products of 7, 9, and 11. Fading of the original yellow col-
spectra (Figure 2), the most characteristic spectroscopic fea- our during irradiation indicated the disappearance of the
tures of the annulated alkylidenedihydrotetrazoles are the alkylidenedihydrotetrazoles, which occurred slower at low
high-field NMR signals of the methine groups (δH
ϭ
temperature than at 20°C. The course of the photolyses was
3Ϫ3.4, δC ϭ 51Ϫ66) and the analogous quaternary vinylic also monitored by NMR spectroscopy, which provided the
carbon atoms (δ ϭ 63Ϫ73) (Tables 2 and 3). These data clue to the structures of the photolysis products (Tables 4
318
Eur. J. Org. Chem. 1998, 317Ϫ327

Photochemical Formation of Heteromethylenecyclopropanes, 28
FULL PAPER
and 5). In all cases except 7a, the conversion and the yields, high diastereoselectivity. Apparently, the geometry of the
which were determined by a comparison with the internal moiety R1R2CϭCϪNMe of the precursors 2 is con-
standard tert-butyl methyl ether, were better than 95%, and served.[5][6] In accord with these results, irradiation of 7a in
no byproducts could be detected. Only 7a formed substan- [D8]toluene solution at Ϫ60°C afforded exclusively (E)-14a,
tial amounts of unidentified byproducts, and, therefore, the while a mixture of (E)- and (Z)-14a (66 : 34) was obtained
yield of 14a dropped to 30Ϫ50%.
at 20°C, obviously by way of (E) v (Z) diastereomerisation
In most cases, mixtures of isomers were formed, the ratio of primarily formed (E)-14a, which eventually led to an
of which changed slowly at room temperature. This process equilibrium between almost equal amounts of both dia-
led eventually to a constant ratio and is doubtless the ther- stereomers. Replacement of the vinyl proton of 7a with a
mal (E) v (Z) equilibration, long-known from other imino- methyl group, resulting in 7b, changed the (E/Z) ratio ob-
aziridines.[5][6][14][15][16][17] It allowed us to assign NMR sig- tained on photolysis. Besides (E)-14b, significant amounts
nals to individual compounds. The iminoaziridine struc- of (Z)-14b arose even at Ϫ60°C [(E)-/(Z)-14b ϭ 57:43]. In
tures 14, 16, 18, and 19 of the photoproducts were not only the equilibrium, (Z)-14b is slightly favoured over (E)-14b,
based on NMR evidence but also on the results of their probably because of steric repulsion between the two methyl
thermal [2 ϩ 1] cycloreversion into cyclic imines (15, 17, groups in the (E) diastereomer. Thermal [2 ϩ 1] cyclorever-
20) and methyl isocyanide (see below). In addition, the IR sion of the annulated iminoaziridines 14 occurred smoothly
spectrum of the photoproduct of 7a showed a CϭN ab- at 95°C and gave 1:1 mixtures of methyl isocyanide and the
sorption at a very high frequency (1795 cmϪ1) which is dihydropyrroles 15 (15a, 67%; 15b, 93%) of which 15a was
characteristic for iminoaziridines.[14][15][16][17] The configu- identified by a comparison of its NMR spectra (Tables 6
ration was elucidated with the help of correlations that had and 7) with those of the authentic compound.[18][19]
been derived from the NMR spectra of monocyclic imino-
aziridines. Accordingly, the methyl carbon atoms at the ex-
ocyclic nitrogen atom of (E) diastereomers resonate at lower
field than those of (Z) diastereomers.[5][15][16][17] Further-
more, the extent of the long-range coupling between the
proton at the aziridine ring and the protons of the N-meth-
ylimino group is larger for (Z) diastereomers than for (E)
5
diastereomers (5Jtransoid > Jcisoid).[5][6][14][15][16][17]
Experimental conditions and results of the irradiation
experiments and the thermal (E) v (Z) equilibration of the
annulated iminoaziridines 14, 16, and 18 are listed in
Table 1.
Table 1. Conditions and results of irradiation experiments and the
subsequent thermal E v Z equilibration of the photolysis products.
The two pairs of geminal methyl groups at the six-mem-
The conversions and the yields were better than 95%
bered ring of 9 influenced the stereochemical result of the
photolysis in a surprising way. Not only the same product
ratio was formed at Ϫ60°C and 20°C, and in different sol-
vents, but the (Z)-iminoaziridine (Z)-16 predominated over
its diastereomer [(E)-16/(Z)-16 ϭ 2:3]. Thermal equili-
bration changed this ratio into 3:1. Thermolysis of 16 oc-
curred somewhat slower than that of 14. Only after 27 h at
95°C, 89% of 16 had decomposed into methyl isocyanide
and dihydropyrrol 17 (yield 86%).
While the nature of the thermolysis products left no
doubt of the iminoaziridine structures 16, the unexpected
diastereomeric ratio obtained on photolysis of 9 stimulated
the application of a third criterion for the assignment of
configurations, viz. the asymmetric solvent-induced shift.
The asymmetric solvation of the CϭNCH3 moiety of im-
ines by [D6]benzene molecules causes high-field shifts that
are larger for protons cis to the imino methyl group than
[a]
(E/Z) ratio after equilibration during the specified time at the
for trans protons.[20] Comparison of the proton spectra re-
[b]
given temperature.
Ϫ
Solvent A: [D8]tetrahydrofuran, B:
[c]
[D6]benzene, T: [D8]toluene. Ϫ
The yield was only 30%. Ϫ
corded for [D8]tetrahydrofuran and [D6]benzene solutions
shows (Table 4) that the signal of the proton at the aziridine
ring of the predominant photoproduct 16 experiences a
[d]
The yield was only 50%.
On photolysis at low temperatures, monocyclic alkylidene- high-field shift of only 0.06 ppm while that of the minor
dihydrotetrazoles 2 yield (E) iminoaziridines (E)-3 with isomer is shifted to higher field by no less than 0.27 ppm.
Eur. J. Org. Chem. 1998, 317Ϫ327
319

H. Quast, A. Fuss, W. Nüdling
FULL PAPER
This result confirms the assignment of configuration based
on the extent of long-range proton-proton coupling and on
chemical shifts in carbon-13 spectra.
Irradiation at low temperature of a [D8]toluene solution
of the alkylidenedihydrotetrazole 11, annulated with a
seven-membered ring, afforded two major products in simi-
lar amounts besides traces of a third product. The latter
increased at the expense of one of the former when the
experiment was performed at 20°C. This process continued
after photolysis until a constant ratio was reached (2:1).
Clearly, the photoproducts that interconvert at 20°C are
(E)- and (Z)-18. The carbon-13 shift criterion (vide supra)
allowed us to assign the (E) configuration to the predomi-
nating isomer. The structure 19 of the second major photo-
product is based on NMR evidence (Tables 4 and 5).
At 20°C, the ratio of the isomers [(E)-18 ϩ (Z)-18] / 19
remained almost unchanged during extended periods of
time. As before, increase of the temperature up to 95°C
eventually led to thermal decomposition into methyl iso-
cyanide and tetrahydropyridine 20, which was identified by
comparison of its proton and carbon-13 spectra (Tables 5
and 6) with those of the authentic compound.[18][19] The
thermolysis was monitored by proton spectroscopy. The re-
sulting conversion vs. time diagramme (Figure 1) shows a
simultaneous decrease of 18 and 19 and a deterioration of
the material balance, which is indicative of the formation
of unidentified decomposition products. No hint about the
presence of the hypothetical [2 ϩ 1] cycloreversion product
of 19, viz. 21, could be derived from the proton and carbon-
13 spectra, however. Interestingly, the yield of methyl iso-
cyanide formed on thermolysis (61%) significantly exceeded
the amount of 18 present at the beginning (44%). We
rationalise this result in terms of the thermal isomerisation
of 19 into 18 for which precedent exists in the series of
monocyclic iminoaziridines[16] and in the equilibration of
the analogous bicyclic alkenes.[21] This interpretation also
explains the lack of 21.
Figure 1. Conversion vs. time diagram for the thermolysis of a mix-
ture of 18 and 19 in [D8]toluene solution at 95°C
radiation of a solution of 13a in [D8]toluene at Ϫ60°C
yielded a mixture of unidentified compounds. Under the
same conditions, the tetramethyl homologue 13b furnished,
by contrast, only a single photoproduct. This compound
was also formed quantitatively by thermal denitrogenation
of 13b, which occurred slowly at 60, with a reasonable rate
The photolysis of the doubly bridged alkylidenedihydro- at 90°C. NMR and IR evidence left no doubt that photoly-
tetrazoles 13 appeared particularly intriguing because ex- sis and thermolysis had led to the hexahydronaphthyridine
tremely strained closed-shell products such as 24 or 25, or, 22. Apparently, 22 arises by a hydrogen shift from an elusive
at low temperatures, persistent diradicals like 23 might re- intermediate which is generated on photolysis of 13b. Pos-
sult from the photoextrusion of molecular nitrogen. Ir- sible candidates for the intermediate(s) are 23Ϫ25.
320
Eur. J. Org. Chem. 1998, 317Ϫ327

Photochemical Formation of Heteromethylenecyclopropanes, 28
FULL PAPER
and, so far, only nitrogen analogues of trimethylenemethane
(28).[22] Therefore, the smooth photolysis of alkylidenedi-
hydrotetrazoles stimulated the hope to generate triplet di-
azatrimethylenemethanes, which might be detected by EPR
spectroscopy at low temperatures. Irradiation with an argon
ion laser of the monocyclic alkylidenedihydrotetrazoles 2g,
h, and 26, and the annulated alkylidenedihydrotetrazoles 5b,
7b, 9, and 13a in 2-methyltetrahydrofuran matrices at 77 K
for a few minutes followed by recording the EPR spectra
at that temperature produced only signals of monoradicals,
however. By contrast, photolysis of 13b under the same
conditions furnished a centrosymmetric four-line spectrum,
centred at 3362 ± 10 G, which closely resembled the EPR
spectra of tris(N-methylimino)methane (29)[22] and trime-
thylenemethane (28)[23] (Figure 3, lower trace). The spec-
trum persisted for hours at 77 K. The half-field transition
(∆mS ϭ 2) which is a criterion of triplets is observed at 1669
G. It exhibits a hyperfine structure of at least 12 equidistant
lines separated by 8.5 ± 1 G. The absorption near the centre
of the four-line high-field spectrum stems from monorad-
icals generated during the irradiation.
With the view of observing labile intermediates, the co-
urse of photolysis of 13b was monitored by UV/Vis spec-
troscopy (Figure 2). Two isosbestic points at 260 and 320
nm, observed in irradiation experiments performed at room
temperature, were indicative of the absence of detectable in-
termediates on the way to the hexahydronaphthyridine 22,
however.
Figure 2. UV/Vis spectrum of 13b (tetrahydrofuran solution), UV/
Vis spectra recorded after irradiation of a 1.00·10Ϫ2  solution of
13b in tetrahydrofuran for 30, 60, and 90 min with a 100-W day-
light lamp, and UV spectrum of 22 obtained by irradiation of a
1.00·10Ϫ2  solution of 13b in tetrahydrofuran with a high-pres-
sure mercury lamp (5 min, λ Ն 320 nm)
At the first glance, the apparent four-line spectrum can
be assigned to the z and x,y signals of randomly oriented
triplet molecules having an axis of threefold or higher sym-
metry, i.e. E/hc ഠ 0 cm .
Ϫ1 [24] Close scrutiny of EPR spectra,
recorded with a much higher signal-to-noise ratio, un-
covered a shoulder of both apparent x,y lines, however (Fig-
ure 3, middle spectrum). A very similar EPR spectrum ap-
peared in a butyronitrile matrix (Figure 3, upper trace). The
intensity of this shoulder indicated that it could not be the
result of a poor resolution of otherwise separated x and y
lines. This was unequivocally demonstrated by calculation
of the EPR spectrum with extensive variation of the zero-
field splitting parameters DЈ and EЈ. Eventually, the best
agreement between the experimental and the calculated
spectrum (Figure 4) was obtained with the zero-field split-
ting parameters DЈ ϭ 335 G (͉D/hc͉ ϭ 0.031 cmϪ1) and
EЈ ϭ 15 G (͉E/hc͉ ϭ 0.0014 cmϪ1) which differed slightly
from those that could be directly read-off from the EPR
spectrum, viz. ͉D/hc͉ ϭ 0.033 cmϪ1 and ͉E/hc͉ ഠ 0. None
EPR Detection of a Matrix-Isolated Diazatrimethylenemethane
Diradical
Photoextrusion of molecular nitrogen from iminodihyd- of the numerous calculated EPR spectra reproduced the
rotetrazoles in low-temperature matrices gives rise to the shoulder of the virtually collapsed x and y lines. Therefore,
formation of triplet tris(imino)methanes, e.g. 29, the first we conclude that it belongs to an additional, minor para-
Eur. J. Org. Chem. 1998, 317Ϫ327
321

H. Quast, A. Fuss, W. Nüdling
FULL PAPER
spectrum to those of tris(N-methylimino)methane (29)
Figure 3. First derivatives of the EPR absorption obtained after
irradiation of the alkylidenedihydrotetrazole 13b in 2-methyltetra-
hydrofuran matrices at 77 K (spectra below and middle) and in a
butyronitrile matrix at 77 K (above). The insert in the middle shows
the expanded signal of the half-field transition (∆mS ϭ 2)
(͉D/hc͉ ϭ 0.033 ± 0.001 cm )
Ϫ1 [22] and trimethylenemethane
(28) (͉D/hc͉ ϭ 0.025 cm Ϫ1),[23][25] and the structure of the
precursor 13b. In terms of a dipole approximation,[24] the
separation 2 DЈ between the low-field and high-field z lines
corresponds to an average separation r of the unpaired elec-
[26]
˚
trons of 3.5 A (Eq. 1),
in reasonable agreement with the
[22]
proposed structure 23 and the value found for 29.
DЈ ϭ D/gµB ϭ 3gµB/2r3 ϭ 1.39·104 g/r3 [G] (r in A)
(1)
˚
Perturbations of the D3h structure of 28, such as the five-
membered ring of BersonЈs bridged trimethylenemethanes,
give rise to measurable E values, e.g. ͉E/hc͉ ϭ 0.0055 cmϪ1
for 27, R ϭ H, 0.0034 cmϪ1 for 27, R ϭ CH3.[8][24b] Like-
wise, the disturbation in 23 of the threefold axial symmetry
lifts the degeneracy of the sublevels, which is observed for
28 and 29, thus yielding an E parameter of similar size.
Discussion
Not unexpected, strain plays the predominant role for the
stability of annulated alkylidenedihydrotetrazoles and for
the results of their photolysis as well. Both alkylidenedihyd-
rotetrazoles with the shortest bridge, viz. 5a and b, are
highly unstable and give rise to the formation of complex
mixtures of products. All other annulated alkylidenedihyd-
rotetrazoles are readily obtained by deprotonation of tetra-
zolium salts and form distillable yellow oils which show
moderate persistence at room temperature under inert con-
ditions.
On photoextrusion of nitrogen from 7a, b, and 9, ring
contraction occurs exclusively which leads to annulated im-
inoaziridines having an exocyclic CN double bond, i.e. 14a,
b, and 16, respectively. No trace of their hypothetical iso-
mers with an endocyclic CN double bond at the bridgehead
of a bicyclo[4.1.0] system could be detected, even at tem-
peratures as low as Ϫ60°C. By contrast, approximately
equal amounts of both types of iminoaziridines, i.e. 18 and
19, are formed on photolysis of 11, whose seven-membered
ring is obviously large enough to accommodate a CN
double bond at the bridgehead. Clearly, the direction of
cyclisation of the aziridine rings reflects the relative stabilit-
ies of the corresponding carbocyclic systems.[21][27] We note
in passing, that some other heteroanalogues of 6-methylene-
bicyclo[3.1.0]hexane[28] and 7-methylenebicyclo[4.1.0]hep-
tane[29] are already known.
magnetic species. The simulation of the EPR spectrum was
also indispensable for the determination of the EЈ value
(15 G) because it was significantly smaller than half of the
estimated linewidth (40 G) of the x and y lines. Otherwise,
the collapsing of the x and y lines would have misled into
the wrong assumption of a triplet molecule having an axis
of threefold symmetry.
Figure 4. Experimental and calculated (with the parameters given
in the text) high-field part of the EPR spectrum obtained after
irradiation of the alkylidenedihydrotetrazole 13b in a 2-methyltetra-
hydrofuran matrix at 77 K
An interpretation of the (E/Z) diastereoselectivity in the
formation of the exocyclic imines 14, 16, and 18 is less
straightforward. On photolysis at low temperatures, the
monocyclic alkylidenedihydrotetrazoles 2 form (E)-imino-
aziridines with high diastereoselectivity under kinetic con-
trol as evidenced by the subsequent (E/Z) equilibration at
elevated temperatures.[2][3][4] This behaviour is also shown
by 7a [Ǟ (E)-14a] and 11 [Ǟ (E)-18a] but not by 7b (Ǟ
14b) and 9 (Ǟ 16), which produce mixtures of (E) and (Z)
diastereomers. The compositions of these mixtures differ
The assignment of the diazatrimethylenemethane significantly from those at equilibrium and even donЈt de-
structure 23 to the species responsible for the predominant pend on the temperature of the photolysis experiments in
EPR spectrum is based on the close resemblance of the case of 9 (Table 1). Apparently, (E) diastereoselectivity, that
322
Eur. J. Org. Chem. 1998, 317Ϫ327

Photochemical Formation of Heteromethylenecyclopropanes, 28
FULL PAPER
is preservation of the geometric arrangement of the Cϭ Experimental Section
CϪNϪMe moiety of the alkylidenedihydrotetrazole, is ob-
Conditions and results of irradiation and (E) v (Z) equilibration
experiments: Table 1. Ϫ 1H NMR: Tables 2, 4, 6. Ϫ 13C NMR:
Tables 3, 5, 7. Ϫ 1H NMR: Bruker AC 200, AC 250 and DMX
served only, if these are not encumbered by a methyl group
at the CC double bond or geminal methyl groups at the
trimethylene bridge. Otherwise, intermediates are formed 600 (14a, 18Ϫ20). The assignments are based on 1H,1H COSY
spectra (14a, 18Ϫ20, 22), decoupling experiments (15b), or NOE
experiments (9, 17, 22). Ranges of chemical shifts refer to mul-
tiplets, single figures denote the chemical shifts of singlets unless
specified otherwise. Ϫ 13C NMR: Bruker AC 200, AC 250, and
DMX 600 (14a, 18Ϫ20). The signals were assigned on the basis of
DEPT, 13C,1H COSY (7, 9, 11, 13bϪ15b, 16, 17, 22) and 1H,13C
COSY (DMX 600; 14a, 18Ϫ20) spectra. Ϫ IR: Perkin-Elmer 1420.
Ϫ UV/Vis: Hitachi U 3200; 1.00·10Ϫ2  solutions in 1-mm quartz
cells. Ϫ EPR: Bruker ESP 300, klystron frequency 9.4 GHz. A total
of 12Ϫ36 spectra was accumulated with the help of the Bruker 1620
data system. The EPR spectrum of Figure 3 (middle) and Figure
4 was recorded with a sweep width of 960 G and 2048 data points.
Calculation of the EPR spectra was performed with the Bruker
programme WIN EPR SimFonia, version 1.25 (1996). The best
agreement between experimental and calculated spectrum was ob-
tained with the zero-field splitting parameters given in the text, the
following Hamilton parameters of the electrons g(x) ϭ g(y) ϭ
2.006, g(z) ϭ 2.003, a linewidth of the x and y direction of 40 G
and the z direction of 55 G.
that undergo (E) Ǟ (Z) diastereomerisation before closing
the aziridine ring.
Thermal [2 ϩ 1] cycloreversion of the photochemically
formed annulated iminoaziridines into cyclic imines and
methyl isocyanide served as a convenient method to con-
firm the structure. Compound 19 with the endocyclic CN
double bond does not follow this path, which would lead
to the acyclic ω-iminoisocyanide 21, but rearranges to the
isomer 18 which decomposes in the usual way. Perhaps, 19
cannot undergo cheletropic one-step cleavage because it is
locked in the (E) configuration which is less favourable for
a non-linear cheletropic pathway than the (Z) configura-
tion.[30]
The most intriguing aspect of the present work is the
identification of the triplet diazatrimethylenemethane 23 by
EPR spectroscopy. Non-Kekule´ molecules derived concep-
tually by heteroatom-for-carbon substitution in trimeth-
ylenemethane (28) are rare species. While singlet molecules
of this type have been and still are invoked as intermediates
on innumerable occasions, a few heterotrimethyleneme-
thanes have been studied by high-level calculations more
recently, viz. oxyallyl,[9a][9b][9c][31] thioxyallyl,[32] and diaza-
trimethylenemethane.[16] Accordingly, the unsubstituted
oxyallyl shows a slight preference for a triplet ground state
(by 1Ϫ2 kcal molϪ1) while a pair of methyl groups already
suffices to confer a singlet ground state on this diradical.[31]
All three planar diastereomers [(E,E), (E,Z), and (Z,Z)] of
the parent diazatrimethylenemethane (30) are calculated on
the CCSD(T)/UHF level to possess triplet ground states.[16]
While we have not yet probed the preference of the diradical
23 for a triplet or singlet ground state with the help of Cur-
ieЈs law, the computational results support the assumption
of a triplet ground state for 23 which is suggested by the
persistence of its EPR signal in low-temperature matrices.
The failure to observe triplet EPR signals in irradiated low-
temperature matrices of the other alkylidenedihydrotetra-
zoles investigated in this way reveals the structural features
that favour formation of triplet diazatrimethylenemethanes:
(i) Cyclisation of intermediates in the photolysis of alkylid-
enedihydrotetrazoles must be prohibited by strain, (ii) the
comparison of 13a with 13b shows that kinetic stabilisation
by methyl groups at the bridges is necessary to avoid reac-
tions that furnish monoradicals or EPR-silent products.
These structural requirements are much more stringent
than in the case of tris(imino)methanes, the only other trip-
let heterotrimethylenemethanes that have been identified ex-
perimentally as yet.[22]
tert-Butyl methyl ether was distilled from KH and kept under
Ar. [D8]Tetrahydrofuran, [D6]benzene, and [D8]toluene were dried
with NaH. 2-Methyltetrahydrofuran (99%, Acros Organics) was
distilled from NaH, carefully degassed and saturated with Ar. Bu-
tyronitrile (Ն 99%) was purchased from Fluka, carefully degassed
and saturated with Ar. The annulated tetrazolium salts 4, 6, 8, 10,
and 12a were prepared as described.[1]
All experiments with alkylidenedihydrotetrazoles were carried
out under purified Ar (99.998%) with strict exclusion of air and
moisture in rigorously dried glassware. The deprotonation of tetra-
zolium salts was carried out with exclusion of daylight. Solutions
of alkylidenedihydrotetrazoles in sealed tubes were stored in the
dark at 77 K.
4,4,9,9-Tetramethyl-5,6,6a,7,8,9-hexahydro-4H-1,2,3,10-tetra-
azaacenaphthylenium Hexafluorophosphate (12b): A solution of
5-[4-chloro-1-(3-chloro-3-methylbutyl)-4-methylpentyl]-1H-1,2,3,4-
tetrazole (46 g, 0.21 mol)[1] in CHCl3 (90 ml) was placed into 3
thick-walled glass tubes (30 ϫ 4 cm). Dry HCl was introduced for
5 h into the cooled (0Ϫ5°C) solutions. The glass tubes were sealed
with a flame and heated for 15 h at 110°C. Distillation of the sol-
vent from the combined mixtures afforded a viscous, brown oil (81
g) which was dissolved in ethanol (100 ml). A solution of NH4PF6
(53 g, 0.33 mol) in water (0.1 l) was added dropwise to the stirred
solution. After 1 d, the precipitate was collected by filtration to
afford a colourless solid (75 g, 89%), m.p. 216Ϫ222°C. Recrystallis-
ation from ethanol/acetonitrile (2:1) yielded colourless crystals (52
g, 68%), m.p. 230Ϫ233°C (dec.) [ref.[1]: 50% (in two steps), m.p.
232°C (dec.)].
1-Methyl-5,6-dihydro-1H-pyrrolo[1,2-d][1,2,3,4]tetrazole (5a): A
suspension of 4a (54 mg, 0.2 mmol), KH (28 mg, 0.4 mmol) and
18-crown-6 (3 mg, 10 µmol) in [D8]tetrahydrofuran (0.6 ml) was
magnetically stirred at Ϫ78°C in an NMR sample tube attached
to a bubble counter. The temperature was gradually raised until a
yellow colour indicated the formation of an alkylidenedihydrotetra-
zole (Ϫ50°C). The suspension was stirred for 1 h while the colour
changed to brown. After removal of the magnetic stirring-bar, the
suspension was degassed by several freeze-thaw cycles (Ϫ196 Ǟ
Ϫ60°C/10Ϫ2 Torr), and the NMR sample tube was sealed with a
We thank Mrs. E. Ruckdeschel and Dr. D. Scheutzow for re-
cording NMR spectra and Dr. H. Harrer for recording some of
the EPR spectra. Financial support by the Fonds der Chemischen
Industrie, Frankfurt am Main, is gratefully acknowledged.
Eur. J. Org. Chem. 1998, 317Ϫ327
323

H. Quast, A. Fuss, W. Nüdling
FULL PAPER
flame. The solid material was allowed to settle during 1 h at Ϫ60°C
before NMR spectra were taken at Ϫ40°C.
Table 3. Chemical shifts (δ values) in carbon-13 spectra of annu-
lated alkylidenedihydrotetrazoles. Chemical shifts that are printed
in italics and may be exchanged
1,7-Dimethyl-5,6-dihydro-1H-pyrrolo[1,2-d][1,2,3,4]tetrazole
(5b): As described in the preceding experiment from 4b (85 mg, 0.3
mmol), KH (28 mg, 0.4 mmol), and 18-crown-6 (3 mg, 10 µmol)
in [D8]tetrahydrofuran (0.6 ml). The suspension was stirred at
Ϫ45°C for 3 h until the evolution of H2 had subsided and the
colour had changed to yellow.
Table 2. Chemical shifts (δ values) and coupling constants [Hz]
(absolute values, in italics) in proton spectra of annulated alkylid-
enedihydrotetrazoles. The shift ranges include multiplets which
stem from a pair of geminal protons unless specified otherwise
[a]
[b]
Solvent A: [D8]tetrahydrofuran, T: [D8]toluene. Ϫ
trum was recorded at Ϫ40°C.
The spec-
mmol), and 18-crown-6·KCN (2 mg, 5 µmol). The suspension was
stirred for 17.5 h. Ϫ UV/Vis (acetonitrile): λmax ϭ 349 nm, λmin
ϭ
311 nm.
1-Methyl-5,6,7,8-tetrahydro-1H-[1,2,3,4]tetrazolo[1,5-a]-
azepine(11): From 10 (0.3 g, 1 mmol), KH (0.2 g, 5 mmol), and 18-
crown-6·KCN (16 mg, 50 µmol). The suspension was stirred for
18 h.
5,6,8,9-Tetrahydro-4 H,7H-1,2,3,10-tetraazaacenaphthylene
(13a): From 12a (62 mg, 0.2 mmol), KH (70 mg, 1 mmol), and 18-
crown-6 (3 mg, 10 µmol). The suspension was stirred at 0°C for
23 h.
[a]
[b]
Solvent A: [D8]tetrahydrofuran, T: [D8]toluene. Ϫ
trum was recorded at Ϫ40°C.
The spec-
4,4,9,9-Tetramethyl-5,6,8,9-tetrahydro-4 H,7H-1,2,3,10-tetraaza-
acenaphthylene (13b): a) According to the General Procedure from
12b (73 mg, 0.2 mmol), KH (70 mg, 1 mmol), and 18-crown-6 (3
mg, 10 µmol). The suspension was stirred for 17 h.
Synthesis of the Alkylidenedihydrotetrazoles 7, 9, 11, and 13. Ϫ
General Procedure: A suspension of tetrazolium salt, KH, and 18-
crown-6·KCN in tert-butyl methyl ether (7 ml) was magnetically
stirred overnight in a small tube (12 ml) closed with a rubber sep-
tum. The solid material was allowed to settle with the help of a
centrifuge. The clear, yellow solution was transferred via syringe
into a 25-ml flask attached to the high-vacuum distillation ap-
paratus described previously[4] which had been filled with Ar. The
remaining grey solid was washed with tert-butyl methyl ether (2 ϫ
1.5 ml). Distillation of the solvent in vacuo (Ϫ50 Ǟ 20°C/10Ϫ2
Torr) afforded viscous, deeply yellow oils which were distilled at
20Ϫ40°C bath temp./10Ϫ5 Torr upon the cold finger (Ϫ196°C).
b) Preparative experiment: A suspension of 12b (0.73 g, 2 mmol),
KH (0.4 g, 10 mmol) and 18-crown-6 (5 mg, 0.02 mmol) in tert-
butyl methyl ether (7 ml) was stirred for 19 h in a small tube (12
ml) closed with a rubber septum. The solid material was allowed
to settle with the help of a centrifuge. The clear, yellow solution
was transferred via syringe into a 25-ml flask attached to the high-
vacuum distillation apparatus.[4] Distillation of the solvent afforded
a yellow solid (m.p. ca. 30°C) which was sublimed at 30Ϫ40°C
bath temp./10Ϫ5 Torr upon the cold finger (Ϫ40°C). The product
may contain some 22 (IR) formed by decomposition of 13b. Ϫ UV/
Vis (tetrahydrofuran, Figure 2): λmax [nm] (ε) ϭ 359 (780), λmin
[nm] (ε) ϭ 320 (500). Ϫ IR ([D8]tetrahydrofuran): ν˜ ϭ 1717 cmϪ1
(CϭC).
1-Methyl-1,5,6,7-tetrahydro[1,2,3,4]-tetrazolo[1,5-a]pyridine
(7a): From 6a (0.28 g, 1 mmol), KH (0.2 g, 5 mmol), and 18-crown-
6·KCN (16 mg, 50 µmol). The suspension was stirred for 15 h.
1,8-Dimethyl-1,5,6,7-tetrahydro[1,2,3,4]tetrazolo[1,5a]pyridine
(7b): From 6b (0.15 g, 0.5 mmol), KH (0.1 g, 2.5 mmol), and 18-
crown-6·KCN (16 mg, 50 µmol). The suspension was stirred for
17.5 h.
Solutions of Alkylidenedihydrotetrazoles. Ϫ a) For NMR Spec-
troscopy: The deuterated solvent (0.6 ml, Table 1), which had been
degassed by several freeze-thaw cycles (Ϫ196°C/10Ϫ5 Torr), was
condensed upon the yellow solid on the cold finger of the high-
1,5,5,7,7-Pentamethyl-1,5,6,7-tetrahydro[1,2,3,4]tetrazolo- vacuum distillation apparatus[4] (see General Procedure). After
[1,5-a]pyridine (9): From 8 (0.68 g, 0.2 mmol), KH (40 mg, 1 warming, the yellow solution dropped into the NMR sample tube
324
Eur. J. Org. Chem. 1998, 317Ϫ327

Photochemical Formation of Heteromethylenecyclopropanes, 28
FULL PAPER
Table 4. Chemical shifts (δ values) and coupling constants [Hz]
(absolute values, in italics) in proton spectra of annulated iminoazi-
Table 5. Chemical shifts (δ values) in carbon-13 spectra of annu-
lated iminoaziridines. The chemical shifts of the geminal methyl
ridines. The shift ranges include multiplets which stem from a single groups of 16 (printed in italics) may be exchanged
proton unless specified otherwise. The chemical shifts of the gemi-
nal methyl groups of 16 (printed in italics) may be exchanged
[a]
14a, 16 (First lines), 18, 19: solvent [D6]benzene; 14b: solvent
[D8]toluene; 16 (second lines): solvent [D8]tetrahydrofuran.
yellow solution was transferred via syringe into a 1-mm quartz
UV cell.
d) For IR Spectroscopy: A suspension of 12b (73 mg, 0.2 mmol)
and NaH (24 mg, 1 mmol) in [D8]tetrahydrofuran (2.5 ml) was
magnetically stirred for 5 min in a small tube (12 ml) closed with
a rubber septum. The solid material was allowed to settle with the
help of a centrifuge. The clear, yellow solution was transferred via
syringe into a 0.5-mm NaCl IR cell.
[a]
14a, 16 (First lines), 18, 19: solvent [D6]benzene; 14b: solvent
[D8]toluene; 16 (second lines): solvent [D8]tetrahydrofuran. Ϫ
[b]
The chemical shifts could not be determined because the signals
[c]
were hidden under those of the major isomers. Ϫ
(NMe).
δ ϭ 2.34
Irradiation Experiments: a) Solutions of the alkylidenedihydrote-
trazoles 5, 7, 9, 11, and 13 in sealed, degassed NMR sample tubes
were placed in a 1-cm quartz UV cell filled with methanol. The
UV cell was held in the middle of a copper block which was con-
nected to a low-temperature thermostat. The copper block was
placed in a quartz dewar equipped with suprasil quartz windows
and flushed with dry N2. A 500-W high-pressure mercury lamp
(Osram HBO 500 W/2) was employed which was focussed by
quartz optics. The light passed through a 10-cm water filter and
a 5-mm cut-off filter (WG 305 or 320 from Schott, Mainz). The
conversion was monitored with the help of proton spectra recorded
at the temperature that was maintained during the irradiation (inte-
gration of singlets of starting material, products, and the internal
standard, i.e. tert-butyl methyl ether). Results: Tables 1, 4, and 5.
which was finally sealed with a flame after cooling at Ϫ196°C and
evacuating at 10Ϫ5 Torr.
b) For EPR Spectroscopy: A suspension of powdered tetrazolium
salt, e.g. 4b, 6b, 8, or 12 (0.2 mmol), and NaH (24 mg, 1 mmol) in
2-methyltetrahydrofuran, saturated with Ar, (1.5 ml) was magneti-
cally stirred for 1 h or 5 min (4b) in a small tube (12 ml) closed
with a rubber septum. The solid material was allowed to settle with
the help of a centrifuge. The clear, yellow solution was transferred
via syringe into a quartz EPR sample tube. Ϫ Likewise, a solution
of 13b in butyronitrile was obtained from 12b (0.22 g, 0.6 mmol)
and NaH (12 mg, 0.5 mmol) in butyronitrile (4 ml) saturated
with Ar.
b) Solutions of the alkylidenedihydrotetrazoles 2g, h, 5b, 7b, 9,
13, and 26 in quartz EPR sample tubes were placed in a quartz
dewar filled with liquid N2. The solid glasses were irradiated for 2
c) For UV/Vis Spectroscopy: A suspension of powdered 12b (183
mg, 0.500 mmol) and NaH (60 mg, 2.5 mmol) in tetrahydrofuran min with the light of an argon ion laser (Coherent Innova 100, λ ϭ
(50 ml) was magnetically stirred for 1 h in a 50-ml volumetric flask. 333, 351, 364 nm, 1.5Ϫ2.0 W output, quartz optics). The EPR
The solid material was allowed to settle during 30 min. The clear, sample tubes were placed into the precooled cavity (77 K) of the
Eur. J. Org. Chem. 1998, 317Ϫ327
325

H. Quast, A. Fuss, W. Nüdling
FULL PAPER
(ε) ϭ 281 (2850), λmin [nm] (ε) ϭ 260 (2610). Ϫ IR (neat liquid):
Table 6. Chemical shifts (δ values) and coupling constants [Hz]
(absolute values, in italics) in proton spectra of cyclic imines
ν˜ ϭ 1605, 1666 (CϭC, CϭN), 3350 (NH) cmϪ1
.
Thermal Equilibration of Annulated Iminoaziridines. Colourless
solutions of the annulated iminoaziridines, obtained by irradiation
of 7a, b, 9, or 11, were kept at 20°C in the dark for several days. The
equilibration was monitored with the help of proton spectroscopy
(integration of NMe singlets). Results: Table 1.
Thermolysis of Annulated Iminoaziridines. The NMR sample
tubes containing solutions of 14a, 14b, 16, 18, and 19, prepared in
the preceding experiments were completely immersed in an oil bath
and heated at constant temperatures [14a: 1 h at 60, 75, and 90, 6
h at 95°C, 95% conversion, yield of (15a ϩ CNϪMe) ϭ 67%; 14b:
1 h at 80, 6 h at 95°C, 96% conversion, yield of (15b ϩ CNϪMe) ϭ
93%; 16: 1 h at 80, 27 h at 95°C, 89% conversion, yield of (17 ϩ
CNϪMe) ϭ 86%; mixture of 18 and 19 (44:56): 1 h at 55, 65, 75,
and 85, 7 h at 95°C (Figure 1)]. The conversions were monitored
with the help of proton spectra recorded at 20°C (integration of
singlets and the 1:1:1 triplet of methyl isocyanide, internal standard
tert-butyl methyl ether). The products were identified on the basis
of their proton (Table 6) and carbon-13 spectra (Table 7).
[a]
[b]
Solvent B: [D6]benzene, T: [D8]toluene. Ϫ
Spectra recorded
for [D]trichloromethane solutions: ref.[18] (80 MHz), ref.[19] (270
MHz). Ϫ
[c] 4
5
J
2-Me/3-H
ϭ 0.7 Hz, J2-Me/5-H ϭ 1.8 Hz.
Table 7. Chemical shifts (δ values) in carbon-13 spectra of cyclic
imines
Ƞ
Dedicated to Professor Dieter Seebach on the occasion of his
60th birthday.
[1]
H. Quast, J. Balthasar, A. Fuss, W. Nüdling, Liebigs Ann. 1997,
671Ϫ683. Ϫ The results are taken from the dissertations by A.
Fuss, 1981, and W. Nüdling, 1997, University of Würzburg.
[2]
H. Quast, L. Bieber, Angew. Chem. 1975, 87, 422Ϫ423; Angew.
Chem. Int. Ed. Engl. 1975, 14, 428Ϫ429.
[3]
H. Quast, L. Bieber, G. Meichsner, D. Regnat, Chem. Ber. 1988,
121, 1285Ϫ1290.
[4]
H. Quast, T. Hergenröther, Liebigs Ann. Chem. 1992, 581Ϫ590.
[5]
H. Quast, L. Bieber, G. Meichsner, Chem. Ber. 1988, 121,
2117Ϫ2120.
[a]
[6]
Solvent B: [D6]benzene, C: [D]trichloromethane, T: [D8]toluene.
H. Quast, T. Hergenröther, Chem. Ber. 1992, 125, 2095Ϫ2101.
[b]
Ϫ
20-MHz spectrum: ref.[18]
.
[7]
J. A. Berson, R. J. Bushby, J. M. McBride, M. Tremelling, J.
Am. Chem. Soc. 1971, 93, 1544Ϫ1546.
[8]
J. A. Berson, Acc. Chem. Res. 1978, 11, 446Ϫ453; in Diradicals
EPR spectrometer and EPR spectra were taken immediately. Evo-
lution of gas was observed on warming the EPR sample tubes to
20°C. Results with 13b: Figure 3.
c) A 1.00·10Ϫ2  solution of 13b in tetrahydrofuran contained
in a 1-mm quartz UV cell was irradiated with a 100-W daylight
lamp, placed at a distance of 5 cm. UV/Vis spectra were recorded
after time intervals of 30, 60, and 90 min. Results: Figure 2.
(Ed.: W. T. Borden), 1st ed., Wiley, New York, 1982, chapter 4;
in The Chemistry of Quinonoid Compounds (Eds.: S. Patai, Z.
Rappoport), 1st ed., vol. 2, Wiley, New York, 1988, p. 455Ϫ536.
[9] [9a]
A. S. Ichimura, P. M. Lahti, A. R. Matlin, J. Am. Chem.
[9b]
Soc. 1990, 112, 2868Ϫ2875. Ϫ
H. K. Powell, W. T. Borden,
[9c]
J. Org. Chem. 1995, 60, 2654Ϫ2655. Ϫ
D. A. Hrovat, A.
Rauk, T. S. Sorensen, H. K. Powell, W. T. Borden, J. Am. Chem.
[9d]
Soc. 1996, 118, 4159Ϫ4166. Ϫ
A. P. Masters, M. Parvez, T.
S. Sorensen, F. Sun, J. Am. Chem. Soc. 1994, 116, 2804Ϫ2811.
R. D. Little, Chem. Rev. 1996, 96, 93Ϫ114.
[10]
[11]
[12]
2,2,7,7-Tetramethyl-1,2,3,5,6,7-hexahydro[1,8]naphthyridine
(22): a) By irradiation of a solution of 13b in [D8]toluene at Ϫ60°C:
Table 1.
T. Livinghouse, Org. Synth. Coll. Vol. 1990, 7, 517Ϫ521.
H. Quast, M. Ach, M. K. Kindermann, P. Rademacher, M.
Schindler, Chem. Ber. 1993, 126, 503Ϫ516.
[13]
[14]
[15]
[16]
N. Kuhn, H. Bohnen, G. Henkel, J. Kreutzberg, Z. Natur-
forsch., Teil B 1996, 51, 1267Ϫ1278.
b) 22 was obtained quantitatively, when a solution of 13b in
[D8]toluene was heated for 3 h at 60°C and for 1 h at 90°C (1H
NMR).
H. Quast, E. Schmitt, Angew. Chem. 1970, 82, 395Ϫ396; Angew.
Chem. Int. Ed. Engl. 1970, 9, 381Ϫ382.
H. Quast, S. Aldenkortt, E. Heller, P. Schäfer, E. Schmitt,
Chem. Ber. 1994, 127, 1699Ϫ1706.
c) Preparative experiment: A stirred, orange-coloured solution of
12b (2 mmol) in tert-butyl methyl ether (7 ml) prepared as described
above was heated under reflux for 3 d. The pale orange-coloured
suspension was transferred via syringe into a 25-ml flask attached
to the high-vacuum distillation apparatus.[4] Distillation of the sol-
vent afforded a viscous oil which was distilled at 30Ϫ40°C bath
temp./10Ϫ5 Torr upon the cold finger (Ϫ40°C) to yield a colourless
H. Quast, S. Aldenkortt, P. Schäfer, E. Schmitt, E.-U.
Würthwein, Liebigs Ann. 1995, 2171Ϫ2188.
[17]
[18]
H. Quast, D. Regnat, Chem. Ber. 1990, 123, 2195Ϫ2202.
J.-C. Guillemin, J.-M. Denis, M.-C. Lasne, J.-L. Ripoll, Tetra-
hedron 1988, 44, 4447Ϫ4455.
[19]
[20]
H. Bock, R. Dammel, Chem. Ber. 1987, 120, 1971Ϫ1985.
G. J. Karabatsos, S. S. Lande, Tetrahedron 1968, 24, 3907Ϫ3922;
G. J. Karabatsos, R. A. Taller, Tetrahedron 1968, 24,
3923Ϫ3937; F. H. A. Rummens, R. H. Krystynak, J. Am. Chem.
Soc. 1972, 94, 6914Ϫ6921.
1
oil. Ϫ H NMR ([D8]toluene): δ ϭ 0.90 (br. s, 2-Me2), 1.25 (br. s,
3
7-Me2), 1.48 (dd, 6-H2, J6-H/5-H ϭ 6.2, 7.2 Hz), 1.87 (dt, 3-H2,
5
4
[21]
[22]
[23]
3J3-H/4-H ϭ 4.3, J3-H/5-H ϭ 2.2 Hz), 2.25 (ddtd, 5-H2, J5-H/4-H
ϭ
A. S. Kende, E. E. Riecke, J. Chem. Soc. Chem. Commun.
1974, 383Ϫ384.
2.0 Hz), 3.89 (br. s, NH), 5.41 (tt, 4-H). Ϫ 13C NMR ([D8]toluene,
Ϫ60°C): δ ϭ 25.0 (CH2-6), 26.8 (2-Me2), 28.6 (7-Me2), 35.3 (CH2-
5), 37.9 (CH2-3), 49.1, 50.1 (C-2, C-7), 123.9 (C-4a), 127.1 (CH-4),
152.2 (C-8a). Ϫ UV/Vis (tetrahydrofuran, Figure 2): λmax [nm]
H. Quast, L. Bieber, W. C. Danen, J. Am. Chem. Soc. 1978,
100, 1306Ϫ1307.
P. Dowd, K. Sachdev, J. Am. Chem. Soc. 1967, 89, 715Ϫ716.
[24] [24a]
E. Wasserman, R. S. Hutton, Acc. Chem. Res. 1977, 10,
326
Eur. J. Org. Chem. 1998, 317Ϫ327

Photochemical Formation of Heteromethylenecyclopropanes, 28
FULL PAPER
[24b]
27Ϫ32. Ϫ
D. A. Dougherty, Matrix Isolation EPR Spec-
in Reactive Intermediates (Ed.: R. A. Abramovitch), 1st ed., vol.
3, Plenum Press, New York, 1983, p. 326Ϫ328.
H. Quast, R. Frank, A. Heublein, E. Schmidt, Liebigs Ann.
Chem. 1980, 1814Ϫ1835; T. S. Sorensen, F. Sun, Can. J. Chem.
1996, 74, 79Ϫ87.
C. A. Renner, F. D. Greene, J. Org. Chem. 1976, 41, 2813Ϫ2819.
CCSD(T)/RHF calculations yield a difference of 12 kcal molϪ1
between the energy barriers involved in the cheletropic de-
composition of the parent (E)- and (Z)-iminoaziridines;[16] c.f.
M. T. Nguyen, A. Van Keer, L. G. Vanquickenborne, J. Chem.
Soc., Perkin Trans. 2 1996, 299Ϫ305; Chem. Ber. 1997, 130,
69Ϫ75.
troscopy of Biradicals, in Kinetics and Spectroscopy of Carbenes
and Biradicals (Ed.: M. S. Platz), 1st ed., Plenum Press, New
York and London, 1990, chapter 5.
[28]
[25]
[26]
[27]
Recent reviews on trimethylenemethane: W. T. Borden, in Reac-
tive Intermediates (Eds.: M. Jones, Jr., R. A. Moss), 1st ed., vol.
3, Wiley, New York, 1985, 151Ϫ188; G. Maier, H. P. Reisenauer,
T. Preiss, H. Pach, D. Jürgen, R. Tross, S. Senger, Pure Appl.
[29]
[30]
Chem. 1997, 69, 113Ϫ118; and ref.[8,24b]
.
G. R. Luckhurst, Biradicals as Spin Probes, in Spin Labeling:
Theory and Applications (Ed.: L. J. Berliner), 1st ed., Academic
Press, New York, 1976, chapter 4; S. S. Eaton, K. M. More, B.
M. Sawant, G. R. Eaton, J. Am. Chem. Soc. 1983, 105,
6560Ϫ6567.
[31]
[32]
M. B. Coolidge, K. Yamashita, K. Morokuma, W. T. Borden,
J. Am. Chem. Soc. 1990, 112, 1751Ϫ1754.
T. Furuhata, W. Ando, Tetrahedron Lett. 1986, 4035Ϫ4038.
G. Köbrich, Angew. Chem. 1973, 85, 494Ϫ503; Angew. Chem.
Int. Ed. Engl. 1973, 12, 464; M. Baumann, G. Köbrich, Tetra-
hedron Lett. 1974, 1217Ϫ1220; G. Szeimies, Bridgehead Olefins,
[97232]
Eur. J. Org. Chem. 1998, 317Ϫ327
327