Angewandte
Communications
Chemie
Table 1: Ruthenium-catalyzed synthesis of 4-substituted oxazolidin-2-ones (3a–3g)
from ureas or carbamates (1, 4, 5) and terminal diols (2a–2h).[a]
under the applied reaction conditions. Obviously,
a high degree of selectivity is required owing to the
nucleophilic and electrophilic character of both
substrates, which could lead to different heterocyclic
compounds and oligomers or polymers. In this
context, we herein describe the first highly selective
ruthenium-catalyzed synthesis of oxazolidinones
from inexpensive urea and vicinal diols by hydrogen
autotransfer.
Our investigations began with the optimization
of the reaction conditions for the model reaction of
urea (1, 1 mmol) with 1-phenylethane-1,2-diol (2a,
1 mmol), a substrate with two different hydroxy
Entry
XR1
R2
Product
Yield [%][b]
1
2
3
4
NH2 (1)
NH2 (1)
NH2 (1)
NH2 (1)
phenyl (2a)
(3a)
(3b)
(3c)
(3d)
76
71
65
73
p-ClC6H5 (2b)
p-F3CC6H5 (2c)
p-MeOC6H5 (2d)
5
6
NH2 (1)
NH2 (1)
benzyl (2e)
n-butyl (2 f)
(3e)
(3 f)
64
69
groups
(see
the
Supporting
Information,
Scheme S1). Based on previous studies by our
group,[10] we initially used [Ru3(CO)12] as the catalyst
in combination with different phosphines. Only
traces of the desired product were observed under
ligand-free conditions, whereas the use of different
mono- and bidentate phosphines afforded 4-phenyl-
oxazolidin-2-one (3a) selectively in moderate yields
(20–52%). Based on these first results, we performed
a more detailed optimization of the best ligand
7
8
NH2 (1)
NH2 (1)
n-hexyl (2g)
5-hexenyl (2h)
(3g)
(3g)
66
65[c]
9[d]
NHMe (4a)
NHEt (4b)
NHPh (4c)
NHBn (4d)
OMe (5a)
OtBu (5b)
OPh (5c)
phenyl (2a)
phenyl (2a)
phenyl (2a)
phenyl (2a)
phenyl (2a)
phenyl (2a)
phenyl (2a)
phenyl (2a)
(3a)
(3a)
(3a)
(3a)
(3a)
(3a)
(3a)
(3a)
56
54
49
45
62
71
69
72
10[d]
11[d]
12[d]
13[d]
14[d]
15[d]
16[d]
system
(dppf = 1,1’-bis(diphenylphosphino)ferro-
cene). As shown in the Supporting Information,
Table S1, we studied different ruthenium precursors,
catalyst loadings, ratios of 1/2a, and additives and
varied other critical parameters, such as the solvent,
temperature, and reaction time. Finally, we found
that the reaction of urea (1, 1 mmol) and 2a
(0.5 mmol) was most efficient in the presence of
[Ru3(CO)12] (0.01 mmol) and dppf (0.03 mmol) in
1,4-dioxane at 1508C for 22 h, which enabled the
isolation of 3a in 76% yield (Table 1, entry 1).
Considering that at least four steps are involved in
this reaction sequence, the optimized yield indicates
OBn (5d)
[a] Unless otherwise specified, all reactions were carried out with 1 (1.0 mmol), 2
(0.5 mmol), Ru3(CO)12 (0.01 mmol), and dppf (0.03 mmol) in 1,4-dioxane (1 mL) at
1508C for 22 h. [b] Yields of isolated products. [c] The maximum amount of product
would be 0.25 mmol. In addition, the formation of unsaturated isomers in 10% yield
was observed. [d] Urea or carbamate (4 or 5, 1.5 mmol).
that each individual transformation proceeds with high
efficiency, while the formation of water and ammonia as the
only byproducts renders this process atom-efficient and
environmentally benign. Gratifyingly, the reaction proceeded
with chemoselectivity as 3a was almost exclusively obtained
with only traces of the cyclic carbonate (8% yield, see
Scheme 1). Furthermore, the 4-substituted product was
formed as the only regioisomer, which suggests that nucleo-
philic substitution preferably occurs with the terminal alcohol
and intramolecular amination with the internal one.
1,2-diol (2 f and 2g) under the initially established conditions
led to the desired products in similar yields (69 and 66%;
entries 6 and 7). Finally, we were interested in studying the
behavior of unsaturated oct-7-ene-1,2-diol (2h). In this case,
=
reduction of the terminal C C bond was also observed,
providing the saturated oxazolidin-2-one 3g in 65% yield
(entry 8). The well-known ability of ruthenium complexes to
promote the isomerization or reduction of multiple bonds[11]
implies that the hydrogen extracted in the first step is used in
=
both imine and C C bond reduction.
To expand the scope of this method, we next evaluated the
reactivity of different terminal diols. Electron-deficient
styrene diol derivatives, such as compounds 2b and 2c, were
effectively converted into the corresponding oxazolidin-2-
ones 3b and 3c (71 and 65% yield, respectively; Table 1,
entries 2 and 3). Analogously, this reaction can also be
extended to aryl diols with electron-donating groups such as
2d, which afforded 3d in 73% yield (entry 4). On the other
hand, the reaction of benzyl derivative 2e also took place
regioselectively to give 4-benzyloxazolidin-2-one (3e) in
a good yield of 64% (entry 5). Aside from aromatic
substrates, this reaction also worked well with aliphatic
terminal diols. Thus, the reactions of hexane- and octane-
At this point, our catalytic system was applied to the
reaction of 1-phenylethane-1,2-diol (2a) with different sub-
stituted ureas and carbamates (Table 1). Here, our initial
objective was the synthesis of N-protected oxazolidin-2-ones
by using N-monosubstituted ureas as the starting materials.
Unfortunately, the ruthenium-catalyzed reaction of 4a–4d
using the conditions previously developed provided the free
(NH) oxazolidin-2-one 3a in moderate yields (45–56%;
entries 9–12). These results suggest that the primary amine
is a better leaving group than ammonia in the initial
nucleophilic substitution step, and that the amination of the
secondary alcohol takes place preferably with the free NH2
group. The presence of the primary amine in the reaction
Angew. Chem. Int. Ed. 2016, 55, 7826 –7830
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7827