Organometallics
Article
However, there was still the possibility that the active species
was not the desired palladacycle, but could have been soluble-
metal-particle heterogeneous catalysis. Indeed, using similar
lations that do not require either cosolvent or dry environment
increase the interest in this research. When compared to the
literature, our poisoning experiment results raise a question on
the intrinsic parameters in the catalyst’s structure that affect the
leading mechanistic pathway of this reaction in water. As an
early hypothesis, we suggest that while water-soluble NHC-
palladacycles could act as reservoirs of palladium nanoparticles,
their insoluble counterparts seem to be the active species
themselves. Work is currently in progress to use the same
catalyst for different carbon−carbon coupling reactions such as
Heck coupling, in water.
́
catalysts under similar conditions, SanMartin, Dominguez, et al.
31
have stated that palladium−NHC complexes may sometimes
serve as mere Pd(0) reservoirs, leaking palladium atoms that
31
self-assemble into nanoparticles. However, the poisoning
results or our experiments state otherwise in the case of this
heterogeneous catalyst. In order to shed light on the nature of
the active species, a comparison test using pyridine and poly(4-
vinylpyridine) polymer has been carried in order to detect the
formation of palladium nanoparticles while using water-soluble
36
pyridine and water-insoluble poly(4-vinylpyridine). As a
matter of fact, both nitrogen ligands should bind easily to
palladium, but the insoluble polymer would additionally
remove metallic particles from the reaction medium, further
deactivating the catalyst, if it were homogeneous, by preventing
its contact with the reagents. Therefore, if one identifies a large
difference between these two experiments, the active species are
Pd(0) nanoparticles. If the catalyst is the starting palladacycle,
then no major difference should be observed between both
experiments, as the binding mechanism is the same (e.g.,
replacement of the bromide ligand). As shown in Table 6, the
results obtained for catalyst 1 with pyridine (25%) and poly(4-
vinylpyridine) (13%) are very similar, indicating a palladacycle-
driven mechanism.
EXPERIMENTAL SECTION
■
1
NMR spectra were recorded on 400 and 300 MHz spectrometers. H
13
and C chemical shifts (δ) are given in ppm (residual peak of
deuterated solvents was used as reference). The single crystals suitable
for X-ray analysis were sealed into a glass capillary, and the intensity
data of the single crystal were collected. High-resolution mass spectra
(HRMS) were recorded on an LC-MSD-Tof spectrometer in positive
electrospray mode.
Synthesis of the Palladium Complex. α,α′-Bis(benzimidazole)-
o-xylene. To 2 mL of an aqueous solution of KOH (17.8 M, 2 g of dry
KOH) were added benzimidazole (1.1000 g, 9.3100 mmol, 2 equiv)
and tetrabutylammonium bromide (TBAB) (120 mg, 0.3249 mmol,
0
.07 equiv). Then 10 mL of toluene was added to the mixture, which
was then vigorously stirred for 5 min. α,α′-Dibromo-o-xylene (1.2289
g, 4.6555 mmol, 1 equiv) was added to the solution, which was kept
under vigorous stirring overnight. The toluene was evaporated under
reduced pressure, and the product was recovered and rinsed with 100
mL of distilled water. The white powder was recuperated and dried
Table 6. Poisoning Experiments of Catalyst 1
under vacuum to give 1.5440 g of α,α′-bis(benzimidazole)-o-xylene
1
(
yield = 98%). H NMR (CDCl , 300 MHz): δ 7.87 (m, 1H), 7.85 (m,
3
1
H), 7.79 (s, 2H), 7.36 (m, 2H), 7.33 (d, 1H, J < 3Hz), 7.30 (d, 1H, J
3 Hz), 7.26 (td, 2H, J = 9 Hz, J < 3 Hz), 7.16 (m, 1H), 7.13 (m,
<
1
2
13
1
H), 7.09 (m, 2H), 5.32 (s, 4H). C NMR (75 MHz, CDCl ): δ
3
1
43.9, 142.8, 133.8, 133.0, 129.1, 123.5, 122.7, 120.8, 109.7, 46.3.
+
+
HRMS (ESI): calcd for C H N [M + H] 339.160 97, found
22
19
4
3
39.161 70.
Precursor Salt. α,α′-Bis(benzimidazole)-o-xylene (1.5440 g, 4.5625
mmol, 1 equiv) and 4-(phenylethynyl)benzyl) bromide (2.5980 g,
.5813 mmol, 2.1 equiv) were placed in 100 mL of acetonitrile. The
entry
poisoning additive
Hg
yield (%)
a
1
0
0
9
b
2
CS2
mixture was heated to 85 °C and kept at this temperature under
agitation for 48 h. The precipitate was filtered and rinsed with 2 × 30
mL of acetonitrile. The yellow-white powder was dried under vacuum
c
3
PPh3
5
d
4
pyridine
25
13
e
1
5
poly(4-vinylpyridine)
to give 2.9333 g of the precursor salt (yield = 73%). H NMR
a
b
(
DMSO-d , 300 MHz): δ 9.96 (s, 2H), 7.97 (m, 4H), 7.69 (m, 4H),
One drop of Hg. 0.5 equiv of CS (per metal atom), reaction ran at
6
2
c
d
7
5
1
1
3
.61 (s, 8H), 7.53 (m, 4H), 7.44 (m, 8H), 7.26 (m, 2H), 6.07 (s, 4H),
4
(
5 °C. 0.3 equiv of PPh (per metal atom). 150 equiv of pyridine
3
e
.85 (s, 4H). 13C NMR (CDCl , 100 MHz): δ 143.6, 134.9, 132.4,
per metal atom). 150 equiv of poly(4-vinylpyridine) (per metal
3
31.9, 131.6, 129.9, 129.5, 129.3, 129.2, 127.5, 127.4, 123.1, 122.4,
atom), quantity calculated using the monomer’s molecular weight.
2+
2+
14.5, 90.6, 89.1, 65.4. HRMS (ESI): calcd for C H N [M + H]
52
40
4
60.162 10, found 360.163 41.
NHC-Pd Complex 1. Palladium acetate (0.2114 g, 0.2400 mmol, 1
equiv) was dissolved in 50 mL of dimethyl sulfoxide. The precursor
salt (0.0539 g, 0.2400 mmol, 1 equiv) was dissolved in an additional 50
mL of dimethyl sulfoxide. Both solutions were mixed and then stirred
at room temperature for 90 min. The mixture was then heated to 135
°C for an additional 60 min. The solvent was evaporated under
reduced pressure at high temperature. The resulting powder was then
triturated with 20 mL of methanol, rinsed with 2 × 20 mL of
methanol, and rinsed again with 2 × 20 mL of acetonitrile. After drying
under vacuum, 0.2057 g of a whitish-yellow powder was obtained
CONCLUSION
■
We have successfully demonstrated that a heterogeneous NHC-
based palladium species is a good alternative catalyst to what is
currently used for Suzuki−Miyaura couplings in water. Its high
stability and activity led to convincing results, with quantitative
activity toward a wide variety of substrates using catalyst
loadings as low as 10 mol %. To the best of our knowledge,
the recyclability of this process is yet unmatched by any other
phosphine-free, heterogeneous, and support-free NHC-Pd
complex performing Suzuki−Miyaura coupling in water. The
high recyclability and the low catalyst loadings used contribute
to the overall atom economy that makes this process greener
than many other previously reported processes. Additionally,
the mild conditions, low-cost reagents, and simple manipu-
−3
(
1
yield = 87%). Anal. Calcd for C H Br N Pd: C, 63.40; H, 3.89; Br,
6.22; N, 5.69; Pd, 10.8. Found: C, 59.54; H, 3.92; N, 5.33; S, 1.36. H
52 38 2 4
1
NMR (CDCl , 400 MHz): δ 8.05 (br s, 2H), 7.88 (d, 2H, J = 8 Hz),
3
7
6
.60 (br s, 2H), 7.53 (s, 4H), 7.36 (m, 15H), 7.08 (t, 5H, J = 6 Hz),
.83 (d, 2H, J = 8 Hz), 6.19 (d, 2H, J = 16 Hz), 5.53 (d, 2H, J = 12
13
Hz), 5.38 (d, 2H, J = 16 Hz). C NMR (CDCl , 100 MHz): δ 177.11,
3
134.82, 134.49, 134.31, 134.24, 134.01, 132.04, 131.61, 129.91, 128.48,
E
dx.doi.org/10.1021/om500878f | Organometallics XXXX, XXX, XXX−XXX