6186
K. Banert et al. / Tetrahedron Letters 54 (2013) 6185–6188
Me
N
S
N
Ph
H
N
K2CO3
(1.5 equ.)
Me
Ph
Me
N
6
H
N
Me
Ph
NCS
Br
Br
DMI
120 °C
H
N
S
5
N
Ph
C
Me
Me
4
S
Ph
N
NCS
0 °C, 24 h
32%
Me
7
8
S
Me
Me
Me
K2CO3
(1.5 equ.)
H2N
N
Ph
Me
N
N
N
S
N
N
Me
S
Ph
Br
S
+
Br
S
Ph
Ph
Br
Ph
N
Me
CHCl3
rt, 5 h
75%
DMI
120 °C
1.5 h
N
NH HBr
Me
Me
10
59%
12
8%
9
11
Scheme 2. Cylization products of thiourea 5 and isothiourea 10.
Thus, the high energy content of the molecule should exclude the
isolation of 15.
Results and discussion
When we repeated the reaction of 13 with 2-(cyclohex-1-
enyl)ethylamine as described,20 we obtained two substances,
which indicated 1H NMR and 13C NMR spectroscopic data that
were identical with those reported for 14 and 15. However, some
additional NMR experiments showed that the structure of 15 has
to be excluded. The 1H NMR spectrum of this minor product exhib-
its a vicinal coupling within the CH2–NH moiety, which allows to
distinguish between both CH2–N units. But another vicinal cou-
pling between NH and the ring proton 5-H cannot be found even
when the measurement is performed in d6-DMSO instead of CDCl3.
Moreover, the minor product shows three quarternary 13C NMR
signals at low field (d = 172.5, 167.3, 149.0), which indicated
long-range correlations to CH2–N protons that are incompatible
with the reported structure of 15. Our HMBC experiments, opti-
mized for J(13C,1H) = 8 Hz, detected coupling between the protons
of the CH2–N unit and the carbon atoms with d = 167.3 and 172.5,
as well as correlation between the CH2 protons of the CH2–NH moi-
ety and the carbon atom with d = 149.0. These and other results
from 2D NMR spectroscopy23 led to the idea that the assigned
structure of 15 has to be corrected by the isomeric 3-aminomalei-
mide structure 18. Furthermore, we noticed that the UV and IR
spectroscopic data, which were reported for 15,20 are similar to
those published29a for compound 16. The later heterocycle is easily
accessible by heating the alkyne 17 with an excess of cyclohexyl-
amine.29 Thus, we measured the 13C NMR spectrum of 16, and
especially the d values assigned to the four carbon signals of the
maleimide ring (d = 172.8, 167.6, 147.7, 83.3) bear resemblance
to the data, which were reported for the butenolide unit of 15. Fi-
nally, we treated 17 with 2-(cyclohex-1-enyl)ethylamine to syn-
thesize the heterocycle 18. In every respect,23 this compound
was identical with the product claimed to possess the structure
of 15.
Recently, Narasaka and co-workers described the formation of
1H-azirine 6 in 28% yield, besides trace amounts of thiazole 7, by
treating thiourea 5 with potassium carbonate in 1,3-dimethyl-2-
imidazolidinone (DMI) at 120 °C (Scheme 2).19 Unfortunately, no
information on the spectroscopic data or the properties of the
antiaromatic heterocycle 6 was given. Moreover, hints to the syn-
thesis of the new compound 5 are also missing completely. Thus,
we subjected the known22 isothiocyanate 4 to N-benzylmethyl-
amine, which led to the desired thiourea 5 in good yield.23 After
treating 5 with potassium carbonate in DMI at 120 °C, we obtained
the aromatic heterocycle 7 with 27.5% yield. But no other side
products that might have a 1H-azirine structure like 6 were de-
tected. The thiazole 7 could be conveniently prepared for compar-
ison when allenyl isothiocyanate (8)24 was reacted with N-
benzylmethylamine. Alkylation of known25 N-benzyl-N-meth-
ylthiourea with the help of dibromide 9 did not produce 5, how-
ever, the hydrogen bromide salt of isothiourea 10 was formed
instead. This outcome is not surprising because S-alkylation of
thioureas is well known.26 Treatment of 10 with potassium car-
bonate in DMI at 120 °C led to thiazoles 11 and 12. The formation
of the side product 12 might tentatively be explained by the decay
of DMI and the condensation of the resulting formaldehyde with
11.27 For comparison, the known28 thiazole 11 was also prepared
by Hantzsch synthesis from chloroacetone and N-benzyl-N-meth-
ylthiourea in 92% yield.23
We assume that Narasaka and co-workers19 erroneously as-
signed the structure of the antiaromatic compound 6 to the iso-
meric substance 7 (see the comparable yields) or perhaps to the
thiazole 11.
Quite recently, Chen, Wang, and coworkers reported on the
reaction of the 2(5H)-furanone 13 with 2-(cyclohex-1-enyl)ethyla-
mine in the presence of potassium fluoride, which afforded the
simple substitution product 14 and the 1H-azirine 15 after separa-
tion by chromatography (Scheme 3).20 Both products were charac-
terized by melting points, elemental analysis, and UV, IR, MS, 1H
NMR, and 13C NMR spectroscopic data. In the case of 15, a mecha-
nism was offered to explain the formation of this unique nitrogen
heterocycle.
Subjection of halo-substituted butenolides of type 13 to amines
is well known to give not only trivial mono-substitution30
compounds, such as 14, but also several other products,30d,31 for
example, maleic acid monoamide monoesters32 or isomaleimides.33
To the best of our knowledge, however, maleimides like 18 were
never reported as a result of these transformations.34 A simple
mechanism to explain the genesis of 18 might include substitution
of the methoxy group in 13 by the attack of the primary amine at the
carbonyl group and formation of a cis-3-aminocarbonyl-acrolein
We suppose that the bicyclic structure of 15 combines unfavor-
able antiaromatic electronic properties with extraordinary ring
strain, which significantly exceeds that of monocyclic azirines.