2
Tetrahedron Letters
this view, we wanted to study the effect of replacing the CO2
appropriate substituted aniline (Scheme 2). This reaction gave
high yields (70-99%) depending on the substrate. Diimines 4a-d
were synthesized according to literature procedures.24,25 For the
synthesis of diimine 4e, the aforementioned conditions26 were not
as effective; therefore, we modified the reaction conditions by
heating 1 equivalent of 2,3-butanedione and 2 equivalents of
aniline in methanol (instead of ethanol), in the presence of 98%
formic acid at 50 °C for 16 h. By using methanol, we succeeded
in the quantitative precipitation of 4e, since its solubility in
methanol is lower than ethanol.
protecting group by the C6HF5 group, thus overcoming the
difficulties of the previously reported cyclization procedures.22
According to the literature, the C6HF5 group can be satisfactorily
removed with heating to give the free carbene.23 The most widely
used approach for the synthesis of imidazolylidene and
imidazolinylidene NHC precursors is the introduction of the
precarbenic carbon in the cyclization step.
For the synthesis of the CO2-adducts,22 initial formation of the
corresponding salt is required. Then, the free carbene is released
using a base and finally the carbenic carbon is protected with the
CO2-group using carbon dioxide gas. On the contrary, in our
approach, the 2,3,4,5,6-pentafluorobenzaldehyde condenses with
the corresponding diamine to afford the desired protected
carbene, thus saving two extra steps and the selection of an
appropriate base. The C6HF5-based NHC adducts are air- and
moisture-stable solids, which represent convenient alternative
precursors of free carbenes.
The next step involved reduction of the diimines to the
corresponding diamines with excess NaBH4 at r.t. for 3 h. For
diamines 5a-d the reaction proceeded normally, however this
was not the case with diamine 5e. The use of excess NaBH4 or
LiBH4 at r.t. or at reflux did not produce diamine 5e in a
satisfactory yield. However, we were able to isolate the mixture
of the two isomers of diamine 5e in quantitative yield by using 2
equivalents of LiAlH4 in dry THF at 50 °C. The two pure
diastereoisomers were collected after separation by column
chromatography, in a cis:trans ratio of 1:2.
R
R
R
H
i
O
ii
Ar
N
Ar
N
Ar NH2
O
N
Ar
N
Ar
H
R
R
R
The final cyclization step for the synthesis of the NHC
precursors consisted of the condensation of 2,3,4,5,6-
pentafluorobenzaldehyde with the corresponding diamines in
glacial acetic acid. After a few hours, the catalysts precipitated as
white or light-yellow solids; they were then filtered off, rinsed
with cold methanol and dried under high vacuum. Precatalyst 6a
was fully characterized and the spectroscopic data were in full
agreement with the literature values,23 while precatalysts 6c and
6d are novel protected NHCs and their NMR spectra were
comparable to those of precatalyst 6a. The trans-catalyst 6e is
also a novel NHC precursor. During the final condensation step,
we also tried to obtain cis-6e under the same conditions, but it
did not precipitate. Different attempts at the synthesis of cis-6e
were made, but according to TLC monitoring of the reaction,
there was unsatisfactory conversion, presumably due to its higher
steric hindrance (for details, see the ESI). Despite our efforts,
only the trans isomer of 6e was isolated in a good yield.
2a-d
3a-b
4a-e
5a-e
iii
2
a
3
a
4 6
R
Ar
-
a
H
F
F
F
F
F
F
H
H
b
c
a
a
b
c
F
F
F
F
H
H
N
Ar
Ar
Ar
Ar
N
N
N
H3C
CH3
6a-d
6e
H
d
a
a
b
d
e
CH3
The synthetic protocol used for NHC adducts 6a and 6c-e, as
well as the reported literature procedure27 for precatalyst 6b, were
unsuccessful. Obviously, the large steric hindrance of the
isopropyl groups at the 2,6-positions of the phenyl ring blocks
the 2,3,4,5,6-pentafluorobenzaldehyde from effectively leading to
ring closure. For details on the various attempts for the synthesis
of 6b that failed, see the ESI. When a catalytic amount of 10-
camphorsulfonic acid was used as an acidic additive, precatalyst
6b successfully precipitated and was purified by recrystallization
from methanol/diethyl ether. These reaction conditions were also
applied in a final attempt to prepare cis-6e, however no solid
precipitated.
Scheme 2. General synthesis of NHC-adducts 6a-e. Reagents and conditions:
i) HCOOH (a few drops), o.n., (67-99%) for 4a-c: MeOH (0.6 M), r.t., for
4d: iPrOH/H2O 2:1 (0.6 M), r.t., for 4e: MeOH (0.6 M), 50 °C; ii) for 5a-d:
NaBH4 (8 equiv.), dry THF/abs. MeOH 6:4 (0.2 M), 0 °C to r.t., 3 h (87-
99%), for 5e: LiAlH4 (1 M in THF, 2 equiv.), dry THF (0.2 M), 0 °C to r.t., 3
h (99%); cis/trans : 1:2; iii) for 6a,c-e: 2,3,4,5,6-pentafluorobenzaldehyde (1
equiv.), glacial CH3COOH (0.1 M), r.t., o.n. (50-70%), for 6b: 2,3,4,5,6-
pentafluorobenzaldehyde (1 equiv.), 10-camphorsulfonic acid (0.1 equiv.),
dry CH2Cl2 (0.2 M), (80%).
Precatalyst SiMes·CO2 (1) was used as a reference for the
design of the NHCs. We initially decided to study the effect of
the C6HF5 group when compared to the CO2 group; therefore, we
synthesized precatalysts 6a and 6b for comparative purposes.
Furthermore, we synthesized the novel precatalysts 6c and 6d to
study the effect of the absence of p-CH3 and one o-CH3 group on
the catalytic activity. Finally, we increased the steric hindrance
on the distal carbons of the five-membered ring, while
maintaining the 2,4,6-trimethylphenyl group in NHC precursor
6e. During the experimental procedures, several difficulties arose
that will be discussed in detail.
Precatalysts 6a-e were then tested in the catalytic
hydroxymethylation of aldehydes. Heptanal and benzaldehyde
were chosen as representative aliphatic and aromatic aldehydes
and the experimental results are listed in Table 1. The reactions
were carried out using the microwave-assisted procedure
previously optimized in our laboratory.21 Thus, one equivalent of
an aldehyde was reacted with 3 equivalents of paraformaldehyde
in the presence of an NHC precursor (10 mol%) in dry
tetrahydrofuran for 1 h at 100 °C in a microwave reactor (50W
maximum power).
According to the synthetic procedure, the first step was the
condensation of glyoxal 3a or 2,3-butanedione 3b with the
Table 1. Microwave-assisted hydroxymethylation of heptanal and benzaldehyde using catalyst precursors 1 and 6a-e. Reagents and conditions: (CH2O)n (3
equiv.), NHC precursor (10 mol%), dry THF (3.3 M), MW at 50 W, 100 ºC, 1 h.