Gold nanocluster-catalyzed semihydrogenation: A unique activation pathway for terminal alkynes

Article abstract of DOI:10.1021/ja503724j

We report high catalytic activity of ultrasmall spherical Au ₂₅(SC₂H₄Ph)₁₈ and rod-shaped Au ₂₅(PPh₃)₁₀(C≡CPh)₅X ₂ (X = Br, Cl) nanoclusters supported on oxides for the semihydrogenation of terminal alkynes into alkenes with >99% conversion of alkynes and ～100% selectivity for alkenes. In contrast, internal alkynes cannot be catalyzed by such "ligand-on" Au₂₅ catalysts; however, with "ligand-off" Au₂₅ catalysts the internal alkynes can undergo semihydrogenation to yield Z-alkenes, similar to conventional gold nanoparticle catalysts. On the basis of the results, a unique activation pathway of terminal alkynes by "ligand-on" gold nanoclusters is identified, which should follow a deprotonation activation pathway via a R′-C≡C-[Au_nL_m] (where L represents the protecting ligands on the cluster), in contrast with the activation mechanism on conventional gold nanocatalysts. This new activation mode is supported by observing the incorporation of deprotonated -C≡CPh as ligands on rod-shaped Au₂₅(PPh₃)₁₀(C≡ CPh)₅X₂ nanoclusters under conditions similar to the catalytic reaction and by detecting the R′-C≡C-[Au _n(SC₂H₄Ph)_m] via FT-IR spectroscopy.

Full text of DOI:10.1021/ja503724j

Article
pubs.acs.org/JACS
Gold Nanocluster-Catalyzed Semihydrogenation: A Unique
Activation Pathway for Terminal Alkynes
Gao Li and Rongchao Jin*
Department of Chemistry, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, United States
*
S Supporting Information
ABSTRACT: We report high catalytic activity of ultrasmall spherical
Au (SC H Ph) and rod-shaped Au (PPh ) (CCPh) X (X = Br,
25
2
4
18
25
3 10
5 2
Cl) nanoclusters supported on oxides for the semihydrogenation of
terminal alkynes into alkenes with >99% conversion of alkynes and
∼
100% selectivity for alkenes. In contrast, internal alkynes cannot be
catalyzed by such “ligand-on” Au catalysts; however, with “ligand-off”
2
5
Au25 catalysts the internal alkynes can undergo semihydrogenation to
yield Z-alkenes, similar to conventional gold nanoparticle catalysts. On
the basis of the results, a unique activation pathway of terminal alkynes
by “ligand-on” gold nanoclusters is identified, which should follow a
deprotonation activation pathway via a R′CC[Au L ] (where L
n
m
represents the protecting ligands on the cluster), in contrast with the activation mechanism on conventional gold nanocatalysts.
This new activation mode is supported by observing the incorporation of deprotonated CCPh as ligands on rod-shaped
Au (PPh ) (CCPh) X nanoclusters under conditions similar to the catalytic reaction and by detecting the R′C
25
3 10
5 2
C[Au (SC H Ph) ] via FT-IR spectroscopy.
n
2
4
m
■
. INTRODUCTION
Au (SG) (SG = glutathione) nanoclusters were effective in
25 18
2
2,23
the reduction of 4-nitrophenol by NaBH4.
In very recent
The olefin (CC) is typically synthesized by semi-
work, we have investigated the catalytic performance of the free
hydrogenation of the alkyne functionality as a synthetic
1
−3
and oxide-supported Au (SPh) nanoclusters (where n = 25,
n
m
precursor, which is a common approach in organic synthesis.
In the previous catalytic approach, selective semihydrogenation
of alkynes was catalyzed by palladium,
iridium, iron, and nickel
recent research, nanogold catalysts have been reported to
exhibit excellent catalytic performance in hydrogenation
3
6, and 99, m = 18, 24, and 42, respectively) in the
4
−8
9−11
chemoselective hydrogenation of reactants with nitro and
aldehyde functional groups (e.g., nitrobenzaldehyde) into
ruthenium,
1
2
13
14−17
complexes or nanoparticles. In
alcohols with the nitro group unaffected using H gas as
2
24
hydrogen resource; indeed, the gold nanoclusters were found
to be an excellent catalyst in the hydrogenation reactions.
The available crystal structure of the ultrasmall Au (SR)
18
1
8−20
processes.
For the relatively large gold nanoparticles (>2
2
5
nm), Yan et al. reported the selective semihydrogenation of
alkynes catalyzed by nanoporous gold using organosilanes with
water as the hydrogen source, and the metallic Au(0) species
was suggested to be the catalytic active sites in the
hydrogenation reaction. Ren et al. reported that TiO2-
supported gold nanoparticles (average size: ∼3.5 nm) catalyzed
selective hydrogenation of quinoline with H as the reducing
agent.
Recently, atomically precise gold nanoclusters stimulated
great interest in the catalytic hydrogenation.
Zhu et al. investigated the Au (SR) (SR = thiolate)
nanoclusters in both free and oxide-supported forms as catalysts
for the chemoselective hydrogenation of α,β-unsaturated
aldehydes and ketones in solution phase using H gas as the
reducing agent. Although the conversion of unsaturated
aldehydes and ketones was not high enough (e.g., ∼30%),
the Au (SR) catalyst remained intact after the hydrogenation
25
nanocluster (1.27 nm metal core) permits correlation of its
reactivity with the atomic structure. The open facet of the
Au (SR) nanocluster exhibits a unique type of surface
25
18
19
structure, that is, triangular gold sites (Au ), for adsorbing and
3
activating reactant molecules. For example, phenylacetylene
(
PhCCH) molecules could be adsorbed on the triangular
2
2
0
Au site on the open facet of the Au (SR) nanocluster and
then underwent the Sonogashira cross-coupling reaction.
3
25
18
26
21−24
In this work, we report the oxide-supported spherical
For example,
2
1
Au (SR) and rod-shaped Au (PPh ) (CCPh) X (X =
25
18
25
3 10
5 2
2
5
18
Cl, Br) nanocluster catalysts for the semihydrogenation of
terminal alkynes to alkenes using H as the hydrogen source
2
under relatively mild conditions (100 °C, 20 bar H ). To our
2
2
knowledge, there has not been any report in the literature on
the use of ultrasmall gold nanoclusters (1−2 nm) with
nonmetallic electronic properties as catalysts for the semi-
25
18
process as the UV−vis spectrum of the Au (SR) catalyst did
2
5
18
not show noticeable changes after the hydrogenation
Received: April 14, 2014
Published: July 30, 2014
2
1
reaction. It was also reported that intact Au (SR) and
2
5
18
©
2014 American Chemical Society
11347
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354

Journal of the American Chemical Society
Article
hydrogenation process of alkynes, albeit ultrasmall gold
nanocluster catalysts (e.g., Au (SR) nanoclusters) have
ionization mass spectrometry (MALDI−MS), and proton
1
nuclear magnetic resonance ( H NMR) spectroscopy. As
2
5
18
been reported for catalyzing other liquid-phase reaction
shown in Figure 2A, the similar optical spectra before/after the
2
1−24,26−29
process.
Importantly, a new alkyne-activation path-
way is identified in the gold nanocluster-catalyzed semi-
hydrogenation reaction.
■
. RESULTS AND DISCUSSION
Characterization of the Au (SR) /TiO Catalyst. The
2
5
18
2
−
synthesis of [Au (SR) ] nanoclusters (counterion: tetraoc-
tylammonium TOA , hereafter R = CH CH Ph) followed a
previously reported method. The oxide-supported nano-
cluster catalyst was made by impregnation of oxide powders
25
18
+
2
2
25
(
e.g., TiO , CeO , SiO , and Al O ) in dichloromethane
2 2 2 2 3
solution of Au (SR) nanoclusters, followed by drying and
25
18
annealing at 150 °C for 1 h in a vacuum oven; note that the free
(
i.e., unsupported) Au (SR) nanoclusters are stable during
2
5
18
2
6
the annealing treatment (even in air atmosphere). We chose
the Au (SR) /TiO catalyst as an example for detailed
25
18
2
characterization. The diffuse reflectance optical spectrum
(
400−1100 nm) of the Au (SR) /TiO catalyst is shown in
2
5
18
2
Figure 1A (black profile). The shorter-wavelength portion of
−
Figure 2. (A) UV−vis absorption spectra of the initial [Au (SR) ]
2
5
18
(
in dichloromethane) and the redissolved nanoclusters after thermal
treatment of powders at 150 °C in a vacuum oven for 1 h. (B) Positive
mode MALDI mass spectrum of Au (SR) after thermal treatment,
25
18
the peak marked with an asterisk (*) is a fragment caused by MALDI
1
rather than an impurity. (C) H NMR spectra (2.4 to 4.2 ppm range)
of the Au (SR) nanoclusters before and after the thermal treatment
25
18
(
powders dissolved in CDCl for NMR measurements); the full range
3
NMR spectra are shown in Figure S1 in the Supporting Information.
(
D) Thermal stability of Au (SR) in a closed cuvette (dissolved in
25 18
toluene, at 80 °C for 24 h).
treatment indicate that the Au (SR) nanoclusters remain
Figure 1. (A) Diffuse reflectance optical spectra of TiO2 and
25
18
intact after the thermal treatment; of note, some of the
Au (SR) /TiO , and solution-phase spectrum of free Au (SR)
25
18
2
25
18
−
0
nanoclusters (in dichloromethane). (B) STEM image of the
Au (SR) /TiO catalyst. Scale bar in panel B: 5 nm.
[Au (SR) ] nanoclusters were transformed to [Au (SR) ] ,
25 18 25 18
causing a slight change to the optical spectrum. Furthermore,
MALDI−MS analysis (Figure 2B) of the thermally treated
sample shows clean signals at m/z = 7394 (assigned to the
molecular ion peak of Au (SR) , theoretical molecular weight:
25
18
2
the spectrum is dominated by the band gap absorption signal of
TiO ; of note, the band gap of TiO is 3.0 eV (∼410 nm). At
2
2
25
18
longer wavelengths, a broad band from ∼600 to 900 nm is seen
in the diffuse reflectance spectrum of the catalyst, which is
consistent with the solution optical spectrum of free Au (SR)
m/z = 7394); note that a fragment of Au (SR) caused by
25 18
MALDI was observed at 6057 (assigned to Au (SR) , after
2
1
14
30
losing an Au (SR) unit from the Au (SR) nanocluster).
25
18
4
4
25
18
nanoclusters (Figure 1A, red profile). Z-contrast scanning
transmission electron microscopy (STEM) analysis of the
Au (SR) /TiO catalyst shows that the size of gold
This fragment is inevitable in MALDI analysis and also exists in
the analysis of the initial, pure Au (SR) nanoclusters and,
25
18
hence, does not indicate an impurity or decomposed product.
No other peaks were found in wider range mass spectrometric
analysis (e.g., from m/z 2000 up to 500k (k = 1000)), which
means that no smaller or larger size gold nanoclusters were
resulted after the thermal treatment of the Au (SR)
2
5
18
2
nanoclusters (Figure 1B) remains to be 1.3 nm, consistent
with the original size of monodisperse Au (SR) nano-
2
5
18
3
0
clusters. Furthermore, thermogravimetric analysis (TGA) of
the Au (SR) /TiO catalyst also suggests that the protecting
ligands are present on the gold nanoclusters. All these results
2
5
18
2
25
18
26
nanoclusters. Finally, the sample after thermal treatment was
analyzed by NMR (Figure 2C). Compared to the NMR
spectrum of the initial Au (SR) (monoanionic, counterion =
confirm that the Au (SR) nanoclusters loaded on TiO2
2
5
18
support did not decompose or grow to larger particles during
the annealing process.
2
5
18
+
TOA ) nanoclusters (Figure 2C, bottom profile), the thermally
Thermal stability of the Free Au (SR) Nanoclusters.
treated Au (SR) nanoclusters show the same four proton
25
18
25
18
To further confirm that the 150 °C annealing step does not
peaks in the range from 2.4 to 4.2 ppm (Figure 2C, full-range
spectra in Supporting Information Figure S1), which are from
the four sets of methylene protons (i.e., inner and exterior types
of CH CH Ph chemical environments in Au (SR) ) as
alter the Au (SR) nanoclusters, we performed a test of 150
25
18
°
C thermal treatment of the unsupported Au (SR) nano-
25 18
clusters (in powders) in a vacuum oven for 1 h (i.e., the same as
the supported Au (SR) /TiO catalyst). Of note, the thiolate
2
2
25
18
3
1,32
reported before.
Taken together, the Au (SR) nano-
2
5
18
2
25 18
30
desorption temperature is above 200 °C. We chose the
clusters are thermally stable during the 150 °C annealing
process. Furthermore, we tested the stability of the free
Au (SR) nanoclusters in toluene (solvent) at 80 °C (one of
unsupported Au (SR) nanoclusters as they can be readily
2
5
18
characterized by UV−vis, matrix-assisted laser desorption
25
18
1
1348
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354

Journal of the American Chemical Society
Table 1. Optimization of Reaction Conditions for Semihydrogenation of 4-Phenyl-1-butyne Catalyzed by Au (SR) /TiO
2
Article
2
5
18
a
Catalyst
b
b
entry
T (°C)
solvent
base
conversion (%)
selectivity (%)
1
2
3
4
5
6
7
8
9
100
100
100
100
100
100
100
100
100
100
80
DMF
DMF
DMF
DMF
DMF
DMF
DMF
PhMe
MeCN
EtOH
EtOH
EtOH
NH ·H O
Et NH
2
17.8
21.6
3.9
100
100
100
100
3
2
Et N
3
c
Pyridine
aniline
61.7/63.8
d
∼100
n.r.
26.4
K CO3
2
none
1.9
100
100
100
100
100
Pyridine
Pyridine
Pyridine
Pyridine
Pyridine
15.2
10.1
∼100
36.0
n.r.
10
11
12
60
a
Reaction conditions: 100 mg Au (SR) /TiO catalyst (1 wt % loading of Au (SR) ), 0.1 mmol 4-phenyl-1-butyne, 0.2 mmol base, 1 mL
25
18
2
25
18
b
1
c
solvent:H O (10:1, V/V), 100 °C, 20 bar H , 20 h. The conversion and selectivity were determined by H NMR. Isolated yield of 4-phenyl-1-
butene. The other product is 4-phenyl-2-butone and the selectivity is 73.6%. n.r. = no reaction.
2
2
d
the reaction temperatures, listed in Table 1) in a closed cuvette
for 24 h. The process was monitored by recording the spectra
at 8 h intervals; no sign of degradation was observed according
to the superimposable spectra (Figure 2D).
further to 60 °C, the conversion was decreased from ∼100% to
36% and to almost zero (Table 1, entry 10 vs entries 11 and
12). On the basis of the above optimizations, we chose
ethanol/H O as the mixed solvent, pyridine as the base and
2
Optimization of the Catalytic Semihydrogenation of
reaction temperature of 100 °C as the optimized conditions for
the semihydrogenation reaction.
Terminal Alkyne. With the Au (SR) /TiO catalyst, we
2
5
18
2
investigated the catalytic performance in semihydrogenation
reaction and optimized the reaction conditions. The reaction
was carried out under the following conditions: 0.1 mmol 4-
phenyl-1-butyne, 0.2 mmol base, 1 mL solvent, and 100 mg
TiO -supported Au (SR) catalyst (∼1 wt % loading of
Scope of the Catalytic Reaction: Terminal Alkynes.
The catalytic activity of Au (SR) /TiO was further examined
25
18
2
with various terminal alkynes under the optimized reaction
conditions. As shown in Table 2, the terminal chain alkynes (1-
hexyne, 1-octyne, and 1-decyne) and terminal aromatic alkyne
(phenylacetylene) all gave high conversion (>99%) and high
selectivity for the product of alkene (∼100%).
2
25
18
Au (SR) ), 20 bar H , and 20 h (see details in the
2
5
18
2
Experimental Section). The reaction product was analyzed by
NMR spectroscopy (Figure S2 in the Supporting Information).
Various combinations of solvent, base, and reaction temper-
ature were tested for the semihydrogenation process. The initial
result was quite promising; using ammonia as the base and
Support Effect. It is well known that the oxide supports of
nanogold catalysts may have large influences on the catalytic
Table 2. Semihydrogenation of Terminal Alkynes at 100%
a
DMF/H O mixture (10:1 V/V) as the solvent and temperature
2
Selectivity on Au (SR) /TiO Catalyst
2
5
18
2
at 100 °C, 17.8% conversion (referring to alkyne, the same
hereafter) was obtained (Table 1, entry 1). Other amines
(secondary amine: diethylamine, tertiary amine: triethylamine
and pyridine) were further examined (Table 1, entry 2−4);
note that amine additives can largely promote the semi-
20
hydrogenation reaction. Among these amines, pyridine as the
base gave the best catalytic result, that is, 61.7% conversion
(
Table 1, entry 4). There was no reaction when the base was
changed to inorganic salt (K CO , Table 1, entry 6), and when
2
3
no amine was present, only a 1.9% conversion was obtained
Table 1, entry 7). Primary amine (e.g., aniline) as the base has
(
also been investigated (Table 1, entry 5), in which ∼100%
conversion of 4-phenyl-1-butyne was obtained, but the
33
competing hydroamination process occurred and the ketone
product (i.e., 4-phenyl-2-butone) was formed in the crude
products (selectivity for 4-phenyl-2-butone: 73.6%).
Further, other solvents (including protic solvent ethanol and
aprotic solvents toluene and acetonitrile) were examined. The
protic solvent (e.g., ethanol) gave an excellent catalytic result
(
∼100% yield), which is much better than the aprotic solvents
a
(Table 1, entry 10 vs entries 8 and 9). Finally, we investigated
Reaction conditions: 100 mg Au (SR) (1 wt %)/TiO catalyst, 0.1
25
18
2
the temperature effect on the semihydrogenation. When the
mmol alkynes, 0.2 mmol pyridine, 1 mL EtOH:H O (10:1, V/V), 100
2
reaction temperature was decreased from 100 to 80 °C and
°C, 20 bar H , 20 h.
2
1
1349
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354

Journal of the American Chemical Society
Article
3
4,35
performance (e.g., conversion and selectivity).
In our
activated by gold nanoparticles via two possible pathways: (1)
the whole CC bond of the alkyne interacts with the surface
of gold nanoparticle and becomes activated as discussed in
system, blank experiments using gold-free TiO2 gave no
conversion under the same reaction conditions; thus, it clearly
shows that the catalytic sites are associated with the Au
nanocluster. To investigate the potential effect of the oxide
supports, we compared four different oxides (i.e., CeO , SiO ,
18−20
previous work;
(2) the terminal of the CC bond (H
CCR) interacts with gold and becomes activated. We are
inclined to the second pathway, and the clue comes from the
fact that terminal alkyne can ligate with the gold nanocluster
2
2
TiO , and Al O ) as supports for the Au (SR) nanocluster.
2
2
3
25
18
36−38
For all the investigated Au (SR) /oxide catalysts (Figure S3
surface as reported by Tsukuda and co-workers.
In their
2
5
18
in the Supporting Information), no obvious influence was
observed in the catalytic performance of the semihydrogenation
reaction of 4-phenyl-1-butyne, that is, all the catalysts gave high
conversion and selectivity around 99%.
work, a series of alkyne-protected Au nanoclusters, such as
Au (CCPh) , Au (CCPh) , Au (EPT) , and
3
4
16
54
26
30
13
Au (EPT) (EPT = 9-ethynyl-phenanthrene), have been
35
18
synthesized; the protecting ligand (i.e., terminal alkynes) on the
Recyclability of the Au (SR) /TiO Catalyst. The
gold surface was exclusively deprotonated (i.e., RCC
2
5
18
2
39
catalytic performance of the recycled Au (SR) /TiO catalyst
Au), which was also supported by DFT simulations. Thus, the
second pathway (i.e., deprotonation-induced activation) is most
probable in the semihydrogenation catalyzed by Au (SR) /
2
5
18
2
was further investigated. In this test, the Au (SR) /TiO
2
5
18
2
catalyst was recovered simply by centrifugation after the
semihydrogenation reaction; then, a new catalytic reaction
was run under identical conditions except using the recycled
catalyst. After five cycles, we only found a slight decrease in the
conversion (within ∼1%) (Figure S4 in the Supporting
Information); thus, the catalyst has an excellent recyclability.
Pathway of the Terminal Alkyne Adsorption and
Activation on Au25 Catalyst. The catalytic performance of
Au (SR) catalyst can be correlated with its atomic structure.
2
5
18
TiO . To gain further evidence, we added phenylacetylene to
2
the Au (SR) nanoclusters (in toluene solution) in the
25
18
presence of excess pyridine under N atmosphere at 80 °C for 4
2
h, similar to the reaction conditions in this work. Fourier-
transform infrared (FT-IR) spectroscopic analysis identified the
−1
CC stretching peak at 2105 cm (Figure 4), consistent with
2
5
18
25
The X-ray crystal structure of the Au (SR) nanocluster
2
5
18
shows a 13-atom icosahedral Au13 core protected by six “
S(R)AuS(R)AuS(R)” dimeric staple motifs (Fig-
ure 3A). On the two open facets (Figure 3B), three external
Figure 4. FT-IR spectrum of the PhCC···[Au (SR) ]. The peak
2
5
18
−1
at ∼2105 cm belongs to CC stretching frequency.
−1
40,41
the reported CC stretching frequency (∼2124 cm ).
Further, there was no peak from 3400 to 3200 cm− (the range
for CCH stretch of free phenylacetylene), indicating that
the semihydrogenation reaction catalyzed by Au (SR)
1
2
5
Figure 3. (A) The crystal structure of Au (SR) nanoclusters. (B)
25
18
Scheme of the terminal alkyne molecule’s initial adsorption on the
surface of the Au (SC H Ph) nanocluster. The gold cluster is
shown as the space-filling model, and phenylacetylene is used to
represent the terminal alkynes. Color code: Au, green and cyan; S,
yellow; C, gray; H, white.
26
25
18
25
2
4
18
nanocluster proceeds via the second pathway, that is, the
deprotonation-induced activation.
Catalytic Performance of the Au25 Catalyst in the
Semihydrogenation of Internal Alkynes. To confirm the
new activation pathway of terminal alkynes on Au (SR) /
2
5
18
gold atoms from the respective three adjacent staples
TiO , we chose internal alkynes to examine the catalytic activity
2
constitutes a triangular Au site, which is well exposed with
of Au (SR) /TiO catalyst under the aforementioned reaction
3
25
18
2
less steric hindrance for reactant access. Previously, we
theoretically studied the adsorption of phenylacetylene on the
conditions. It was found that the Au (SR) /TiO catalyst
25 18 2
showed no activity (i.e., all below 1%) in the reactions (Table 3,
c.f. entry 1 vs 2−4); this agrees with the deprotonation
activation mechanism of terminal alkynes because an internal
alkyne has no terminal hydrogen and, thus, cannot be activated
by the ultrasmall Au (SR) nanoclusters according to the
26
surface of Au (SR) nanocluster. Density functional theory
25
18
(
DFT) simulations indicated that the phenylacetylene molecule
is adsorbed on the open Au site with an adsorption energy of
3
2
6
−
0.40 eV. In the initial adsorption configuration, the phenyl
25
18
ring of phenylacetylene facing an external gold atom of the Au3
above deprotonation activation pathway.
site and the CCH group of phenylacetylene points toward
Furthermore, we tested the ligand-off catalyst (Au /TiO ),
2
5
2
a second Au atom of the Au site (Figure 3B); note that the
which was prepared by thermally treating the Au (SR) /TiO
3
25 18 2
3
0
third Au atom of the Au site is unoccupied and thus can be
powder at 300 °C (note: thiolate desorption onset: 200 °C )
3
used for hydrogen adsorption and activation facilitated by
pyridine in the present semihydrogenation process.
With respect to the alkyne activation step in the semi-
hydrogenation, the terminal alkyne (HCCR) can be
for 1 h. With the bare Au /TiO catalyst, the semi-
2
5
2
hydrogenation reactions of internal alkynes (including 1-
phenyl-1-propyne, diphenylacetylene, and methyl 2-nonynoate)
were performed under the same reaction conditions (see notes
1
1350
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354

Journal of the American Chemical Society
Article
Table 3. Comparison of Conversion in Semihydrogenation
a
Catalyzed by Ligand-On and Ligand-Off Catalysts
entry
1
catalyst
R1
R2
conversion selectivity
>99
b
ligand-on Au25
sphere
PhC H
H
2
4
Figure 5. (A) Optical absorption spectra of Au (PPh ) (C
2
5
3 10
2
3
4
5
6
7
8
9
Ph
CH3
Ph
<1
CPh) X and Au (PPh ) (SC H Ph) X nanoclusters. The spectra
5
2
25
3
10
2
4
5 2
Ph
<1
are up-shifted for the ease of comparison. (B) Positive mode ESI-MS
n-C H
CO CH
<1
6
13
2
3
3
of the Au (PPh ) (CCPh) X (X = Cl/Br) nanoclusters.
25
3
10
5 2
ligand-off Au25
PhC H
H
95.6
2
4
Ph
CH3
Ph
52.8
59.7
52.6
99.8
97
Information), indicating that the molecular ions (I−V) are all
Ph
>99
99
2
4
+ charged. The peak assignment is as follows: peak I at m/z =
n-C H
CO CH
2+
6
13
2
083.7 to [MBrCl] , (where M = Au (PPh ) (CCPh) ,
25
3 10
5
ligand-on Au25
rod
PhC H
H
2+
2
4
theoretical m/z: 4084.0), peak II at m/z 4096.6 ([MBrClNa] ,
theoretical m/z 4095.5), peak III at m/z 4106.7 ([MBr ] ,
theoretical m/z 4106.3), peak IV at m/z 4118.1 ([MBr Na] ,
2
+
10
11
12
Ph
CH3
Ph
<1
<1
<1
2
2
+
Ph
2
theoretical m/z 4117.3), and peak V at m/z 4129.7
n-C H
CO CH
6
13
2
3
2+
a
b
([MBr Na ] , theoretical m/z 4129.3), with the experimental
2 2
Reaction conditions: as noted in Table 2. Stereoselectivity for Z-
alkene (see Figure S6 in the Supporting Information for the NMR
spectra).
isotopic patterns matching well with simulations (Figure S5C
and D in the Supporting Information). These results indicate
that the alkyne can also ligate with the phosphine-protected
Au nanoclusters and the alkyne ligands are also deprotonated,
25
in Table 2). Interestingly, 52.6 to 59.7% conversions of internal
alkynes with >97% stereoselectivity for Z-alkene were obtained
consistent with the alkyne activation pathway on thiolated
Au (SR) nanoclusters. Additionally, the optical absorption
25
18
(
Table 3, entries 6−8), in striking contrast with the ligand-on
spectrum of Au (PPh ) (CCPh) X nanoclusters (Figure
2
5
3 10
5 2
Au (SR) /TiO catalyst, which showed no activity (<1%
conversion) in the reactions. Of note, the observed stereo-
selectivity over the ligand-off Au /TiO catalyst is similar to
5A) shows three stepwise peaks at 350, 407, and 626 nm,
similar to the peaks of the reported Au (PPh ) (SR) X
5 2
25
18
2
2
5
3 10
4
2,43
nanocluster
(Figure 5A). The similarities in both the
2
5
2
the reported literature work in which conventional gold
nanoparticles were used as the catalyst. We also note that,
formula and optical spectrum imply that the newly obtained
19
Au (PPh ) (CCPh) X nanocluster should possess a
2
5
3
10
5 2
similar to the ligand-on catalyst, the ligand-off Au /TiO can
biicosahedral structure (Figure S5F in the Supporting
Information) similar to the X-ray structure of the
Au (PPh ) (SR) X nanocluster (Figure S5E in the Support-
2
5
2
catalyze the semihydrogenation of terminal alkynes, for
example, 95.6% conversion of 4-phenyl-1-butyne (Table 3,
entry 5).
2
5
3 10
5 2
4
2,43
ing Information).
In the biicosahedral structure, the Au25
The above catalytic results imply that the activation process
of the terminal alkynes by Au (SR) nanocluster catalyst is
rod is composed of two icosahedral Au13 units by sharing one
common Au vertex, and ten phosphines coordinate to the two
2
5
18
different than the cases by bare Au25 nanocluster and
conventional gold nanoparticle catalysts. The activation of
terminal alkynes by Au (SR) nanocluster catalyst should
Au pentagonal rings on the two ends of the rod, and two
5
halides (Cl or Br) bind to the two apical Au atoms, and five
deprotonated alkyne ligands should bridge the two Au13
icosahedra, resembling the thiolate ligands (Figure S5E and F
in the Supporting Information).
2
5
18
occur through the terminal of the CC bond (HCC
R′), and the constraint exerted by the ligand-on triangular Au
3
catalytic site apparently dictates this unique activation mode
compared to the cases of bare nanoclusters and larger
nanoparticles.
Catalytic Property of Rod-Shaped Au25 Nanoclusters.
The catalytic performance of the Au (PPh ) (CCPh) X
25
3 10
5 2
nanocluster supported on TiO2 (∼1 wt % loading) was
investigated. Using 4-phenyl-1-butyne as the reactant under the
identical reaction conditions (Table 3, entry 9), a conversion of
99.8% with ∼100% selectivity for the alkene product was
obtained, but when using internal alkynes as the reactant, low
catalytic activity was seen (Table 3, entries 10−12). These
trends are similar to the case of the ligand-on Au (SR)
18
Comparison of Spherical and Rod-Shaped Au25
Nanoclusters. In order to confirm that the terminal alkynes
can also ligate with gold nanoclusters protected by other
ligands, rather than only the thiolate-protected nanoclusters, we
etched triphenylphosphine (PPh )-protected gold nanoclusters
3
in the presence of excess phenylacetylene (PhCCH) and
pyridine at 80 °C (similar to the catalytic reaction conditions in
this work). A new monodisperse product−gold nanocluster
2
5
catalyst.
Mechanistic Insight. Taken together, the above catalytic
results imply that the terminal alkyne activation should undergo
protected by both PPh and CCPh ligands, was obtained
3
(
Figure 5) and determined to be [Au (PPh ) (C
a similar process over both types of ligand-on Au nanocluster
2
5
3
10
25
2+
CPh) X ] (X = Br, Cl) by electrospray ionization mass
catalysts. On the basis of the experimental results and the
previous DFT calculations, we propose the following semi-
hydrogenation mechanism (Figure 6). In the case of the
spherical Au (SR) nanocluster (Figure 6, left panel), the
5
2
spectrometry (ESI−MS) analysis (Figure 5B). In the zoom-in
spectrum (Figure S5A in the Supporting Information),
prominent peaks (labeled I to V) are observed in the range
of m/z 4080 to 4140, and the spacing of the experimental
isotopic patterns is 0.5 (Figure S5C and D in the Supporting
25
18
reactants (i.e., hydrogen and terminal alkyne (R′CCH))
should be first absorbed on the open triangular Au facet of the
3
1
1351
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354

Journal of the American Chemical Society
Article
PPh -protected Au nanoclusters are etched by phenylacetylene
3
under similar conditions as in the catalytic reaction. The R′
CC[Au (SR) ] was observed by FT-IR spectroscopy with
n
m
−1
the AuCC stretching mode at 2105 cm . Compared to
conventional nanoparticles, the unique catalytic properties of
nanoclusters are remarkable and such nanoclusters hold great
promise in revealing the fundamental aspects of catalytic
reactions.
■
. EXPERIMENTAL SECTION
Figure 6. Proposed mechanism for the ligand-on Au25 nanocluster
Chemicals. Tetrachloroauric(III) acid (HAuCl ·3H O,
4
2
catalyzed semihydrogenation of terminal alkynes to alkenes using H
2
99.99% metal basis, Aldrich), tetraoctylammonium bromide
(TOAB, 98%, Fluka), sodium borohydride (99.99% trace
metals basis, Aldrich), phenylethanethiol (98%, Sigma-Aldrich),
triphenylphosphine (99%, Sigma-Aldrich), toluene (HPLC
grade, 99.9%, Aldrich), methanol (absolute, 200 proof,
Pharmco), methylene chloride (HPLC grade, 99.9%, Sigma-
Aldrich), acetone (HPLC grade, 99.9%, Sigma-Aldrich),
toluene (HPLC grade, 99.9%, Sigma-Aldrich), phenylacetylene
2
5
(
20 bar). Left panel: Au (SR) nanocluster. Right panel:
25 18
Au (PPh ) (CCPh) X (X = Cl, Br) nanocluster was drawn
according to the crystal structure of the Au (PPh ) (SR) X
5 2
25
3
10
5 2
2
5
3
10
4
2,43
nanocluster.
Color code: Au, green; S, yellow; C, gray; P, pink;
X, cyan. Hydrogen atoms are omitted for clarity. The areas marked in
orange are the Au active sites (left panel) and the waist active sites
3
(
right panel) in the alkyne semihydrogenation reaction, respectively.
(
99%, Sigma-Aldrich), pyridine (99%, Sigma-Aldrich). All
Au (SR) nanocluster, and then the HH bond and C
chemicals were used as received. Nanopure water (resistivity
18.2 MΩ·cm) was purified with a Barnstead NANOpure
Diwater system. All glassware was thoroughly cleaned with
aqua regia (HCl:HNO3 = 3:1 v/v), rinsed with copious
nanopure water, and then dried in an oven prior to use.
Thermal Stability Test of the Free Au (SR) Nano-
clusters. In this test, 4 mg of powders of Au (SC H Ph)
nanoclusters (monoanionic, counterion = tetraoctylammonium
2
5
18
CH triple bond would be activated in the presence of a base
e.g., pyridine). Finally, the alkene product should be yielded
after the addition process under the cooperation of three gold
TM
(
atoms of the Au active site (Figure 6, left panel). In the case of
3
the Au (PPh ) (CCPh) X rod cluster (Figure 6, right
25
3 10
5 2
2
5
18
panel), the activation process of alkyne and the catalytic site of
the semihydrogenation should be associated with the “waist” of
the Au25 rod since the waist Au atoms should be bonded with
the deprotonated alkynes (Figure S5F in the Supporting
Information and Figure 6 right panel). The deprotonation
activation of terminal alkyne (R′CCH) and bridging
adsorption of R′CC on the catalytic site should be similar
in both Au25 sphere and rod nanoclusters. This pathway is in
contrast with the adsorption of alkynes on the “bare” Ru
colloid, in which a type of Ru−vinylidene interfacial bond was
25
2
4
18
+
ion, TOA ) was placed in a 5 mL tube and then annealed in a
vacuum oven at 150 °C for ∼1 h. After that, the sample
(
dissolved in solution) was characterized by UV−vis, MALDI−
1
MS, and H NMR spectroscopy. UV−vis spectra of the clusters
before/after annealing (dissolved in dicholomethane) were
acquired on a Hewlett-Packard (HP) Agilent 8453 diode array
spectrophotometer at room temperature. MALDI−MS was
performed with a PerSeptiveBiosystems Voyager DE super-
STR time-of-flight (TOF) mass spectrometer. Trans-2-[3-(4-
tert-butylphenyl)-2-methyl-2-propenyldidene] malononitrile
4
4
formed.
■
. CONCLUSION
(
DCTB) was used as the matrix in MALDI analysis. The
In summary, the catalytic performance of the spherical
nanoclusters and the matrix were disolved in dicholomethane
and then spoted onto the steel plate and dried in vacumm prior
to MALDI analysis. For NMR analysis, the Au (SC H Ph)
18
Au (SR) and rod-shaped Au (PPh ) (CCPh) X nano-
25
18
25
3 10
5 2
clusters (supported on oxides) have been investigated for the
semihydrogenation of terminal alkynes to alkenes. High
conversion of a range of terminal alkynes (up to ∼100%)
with excellent selectivity (∼100%) for alkene products was
25
2
4
nanoclusters were dissolved in 0.5 mL CDCl in a NMR tube
3
1
and H NMR was run on a Bruker 300 MHz spectrometer.
I d e n t i fi c a t i o n o f R ′  C  C  [ A u ( S R ) ] .
Au (SC H Ph) nanoclusters (1 mg) were dissolved in 1
25 2 4 18
n
m
achieved using ethanol/H O as the mixed solvent and pyridine
2
as the base at 100 °C. No distinct effect of the oxide supports
mL of toluene, and then 100 μL of phenylacetylene and 90 μL
of pyridine were added. The solution was heated to 80 °C
under vigorous stirring and maintained at this temperature for 4
was observed. The nanocluster catalyst showed excellent
recyclability. The open triangular Au facet on the spherical
3
Au (SR) nanocluster and the waist sites of the rod-shaped
h (under a N atmosphere). After that, the toluene was
25
18
2
Au (PPh ) (CCPh) X (X = Br/Cl) nanocluster are
removed by rotary evaporation, and the left liquid was washed
with hexane for six times to remove excess phenylacetylene and
pyridine. Then the product was extracted by acetone, followed
by centrifugation again to remove the insoluble components.
Finally, the product was deposited onto a NaCl window by
spotting a CH Cl solution of the product, followed by FT-IR
2
5
3
10
5 2
rationalized to be the active sites in the semihydrogenation
process. In contrast with terminal alkynes, the internal alkynes
cannot be catalyzed by the ligand-on nanoclusters, whereas the
ligand-off nanoclusters are found to be capable of catalyzing the
semihydrogenation of internal alkynes. Thus, a unique
activation pathway is identified for terminal alkynes over
ligand-on gold nanoclusters, and this mode is in contrast with
the case of conventional gold nanoparticle catalysts. Specifically,
the terminal alkynes should be activated via a R′C
C[Au L ] (where L represents protecting ligands on the
2
2
measurements on a Thermo/ATI/Mattson 60AR instrument
−
1
−1
(resolution, 1 cm ; scans, 16; range, 1000−4000 cm ).
Preparation of Au (PPh ) (CCPh) X (X = Cl, Br)
2
5
3 10
5 2
Nanoclusters. HAuCl ·3H O (78.8 mg, 0.2 mmol) and
4
2
tetraoctylammonium bromide (TOAB, 131.3 mg, 0.24 mmol)
were dissolved in 15 mL of acetone in a trineck round-bottom
flask. After the solution was vigorously stirred for 10 min, PPh3
n
m
cluster) with clues from the incorporation of CCPh in the
Au (PPh ) (CCPh) X (X = Br/Cl) nanoclusters when
25
3 10
5 2
1
1352
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354

Journal of the American Chemical Society
78.6 mg, 0.3 mmol) was added. The solution became colorless
Article
(
REFERENCES
■
after 20 min. Then, NaBH (22.8 mg, 0.6 mmol, dissolved in 5
4
(
1
1) Oger, C.; Balas, L.; Durand, T.; Galano, J. Chem. Rev. 2013, 113,
313−1350.
(2) Furstner, A.; Davies, P. W. Chem. Commun. 2005, 41, 2307−
2320.
3) Michaelides, I. N.; Dixon, D. J. Angew. Chem., Int. Ed. 2013, 52,
06−808.
4) Mitsudome, T.; Takahashi, Y.; Ichikawa, S.; Mizugaki, T.;
Jitsukawa, K.; Kaneda, K. Angew. Chem., Int. Ed. 2013, 52, 1481−1485.
5) Hauwert, P.; Maestri, G.; Sprengers, J. W.; Catellani, M.; Elsevier,
C. J. Angew. Chem., Int. Ed. 2008, 47, 3223−3226.
6) Shen, R.; Chen, T.; Zhao, Y.; Qiu, R.; Zhou, Y.; Yin, S.; Wang, X.;
mL ethanol) was added all at once. The solution immediately
turned dark. The reaction was allowed to proceed for 2 h under
air environment at room temperature. The crude black product
̈
(
8
(
(
containing PPh -protected Au nanoparticles) was collected
3
after rotary evaporation of the solvent. The black product was
washed several times with hexane to remove excess PPh and
3
TOAB. The precipitated product was extracted by acetone,
followed by centrifugation again to remove insoluble
components. The left solution of Au nanoparticles was dried
by rotary evaporation, and the solids were dissolved in 0.2 mL
of phenylacetylene and 0.2 of mL pyridine. The solution was
vigorously stirred at 80 °C for 4 h. The products were obtained
after drying the solution by rotary evaporation and washing
with hexane and finally extraction with methanol.
(
(
Goto, M.; Han, L. J. Am. Chem. Soc. 2011, 133, 17037−17044.
(7) Shirakawa, E.; Otsuka, H.; Hayashi, T. Chem. Commun. 2005, 41,
5
(
(
1
885−5886.
8) Li, J.; Hua, R.; Liu, T. J. Org. Chem. 2010, 75, 2966−2970.
9) Belger, C.; Neisius, N. M.; Plietker, B. Chem.Eur. J. 2010, 16,
2214−12220.
10) Trost, B. M.; Ball, Z. T.; Jo
7922−7923.
Preparation of Au (SR) /Oxide and Ligand-off Au /
25
18
25
(
̈
ge, T. J. Am. Chem. Soc. 2002, 124,
TiO Catalysts. Typically, 1 mg of Au (SR) nanoclusters
2
25
18
were dissolved in 5 mL of dicholomethane, and 100 mg of
(11) Li, J.; Hua, R. Chem.Eur. J. 2011, 17, 8462−8465.
(12) Tani, K.; Iseki, A.; Yamagata, T. Chem. Commun. 1999, 35,
1821−1822.
13) Belger, C.; Plietker, B. Chem. Commun. 2012, 48, 5419−5421.
14) Barrios-Francisco, R.; García, J. J. Appl. Catal. A: Gen. 2010, 385,
oxide (e.g., CeO , TiO , Al O , or SiO ) was added. After
2
2
2
3
2
stirring for 12 h at rt, the Au (SR) /oxide catalysts were
2
5
18
(
(
1
(
collected by centrifugation and dried in a vacuum. The catalysts
were then annealed at 150 °C for ∼1 h in a vacuum oven.
When the annealing temperature was increased to 300 °C, the
thiolate ligands were desorbed and ligand-off Au /TiO
08−113.
15) Carenco, S.; Leyva-Per
Mezailles, N.; Sanchez, C.; Corma, A. Nano Today 2012, 7, 21−28.
16) Witte, P. T.; Boland, S.; Kirby, F.; van Maanen, R.; Bleeker, B.
́ ́ ̀
ez, A.; Concepcion, P.; Boissiere, C.;
2
5
2
́
catalyst was obtained. No apparent particle size growth was
observed in STEM analysis.
(
F.; de Winter, D. A. M.; Post, J. A.; Geus, J. W.; Berben, P. H.
Typical Procedure for the Semihydrogenation Re-
action. In a typical semihydrogenation reaction, alkyne (0.1
mmol), base (0.2 mmol), Au (SR) /oxide (100 mg, 1 wt %
ChemCatChem 2013, 5, 582−587.
(
17) Chung, J.; Kim, C.; Jeong, H.; Yu, T.; Binh, D. H.; Jang, J.; Lee,
J.; Kim, B. M.; Lim, B. Chem.Asian J. 2013, 8, 919−925.
2
5
18
(18) Arcadi, A. Chem. Rev. 2008, 108, 3266−3325.
(19) Yan, M.; Jin, T.; Ishikawa, Y.; Minato, T.; Fujita, T.; Chen, L.;
loading of Au (SR) ) and 1 mL of solvent were added to a
25
18
reactor (Parr Instrument Company, 22 mL capacity, series
700) and H was introduced until 20 bar. The reaction was
Bao, M.; Asao, N.; Chen, M.; Yamamoto, Y. J. Am. Chem. Soc. 2012,
134, 17536−17542.
4
2
kept at 100 °C for 20 h as indicated in Table 1. The product
(20) Ren, D.; He, L.; Yu, L.; Ding, R.; Liu, Y.; Cao, Y.; He, H.; Fan,
was obtained after removal of solvent. Both the conversion of
alkyne and the selectivity for alkene were determined by H
NMR analysis.
K. J. Am. Chem. Soc. 2012, 134, 17592−17598.
1
(21) Zhu, Y.; Qian, H.; Drake, B. A.; Jin, R. Angew. Chem., Int. Ed.
2
(
010, 49, 1295−1298.
22) Shivhare, A.; Ambrose, S. J.; Zhang, H.; Purves, R. W.; Scot, R.
W. J. Chem. Commun. 2013, 49, 276−278.
ASSOCIATED CONTENT
■
(23) Yamamoto, H.; Yano, H.; Kouchi, H.; Obora, Y.; Arakawa, R.;
Kawasaki, H. Nanoscale 2012, 4, 4148−4154.
*
S
Supporting Information
Full NMR analysis of the free Au (SC H Ph) nanoclusters
before and after annealing, testing of the support effects and
(
(
24) Li, G.; Zeng, C.; Jin, R. J. Am. Chem. Soc. 2014, 136, 3673−3679.
25) Zhu, M.; Aikens, C. M.; Hollander, F. J.; Schatz, G. C.; Jin, R. J.
2
5
2
4
18
Am. Chem. Soc. 2008, 130, 5883−5885.
recyclability of Au (SC H Ph) nanocluster catalysts, exper-
(26) Li, G.; Jiang, D.; Liu, C.; Yu, C.; Jin, R. J. Catal. 2013, 306, 177−
2
5
2
4
18
imental and simulated isotopic distribution patterns of
183.
2
+
(
27) Li, G.; Jin, R. Acc. Chem. Res. 2013, 46, 1749−1758.
[
Au (PPh ) (CCPh) X ] (X = Cl, Br) nanoclusters,
25
3 10
5 2
1
(28) Liu, J.; Krishna, K. S.; Losovyj, Y. B.; Chattopadhyay, S.; Lozova,
and data of H NMR. This material is available free of charge
via the Internet at http://pubs.acs.org.
N.; Miller, J. T.; Spivey, J. J.; Kumar, C. Chem.Eur. J. 2013, 19,
1
(
12007.
0201−10208.
29) Li, G.; Liu, C.; Lei, Y.; Jin, R. Chem. Commun. 2012, 48, 12005−
AUTHOR INFORMATION
■
(30) Qian, H.; Zhu, M.; Wu, Z.; Jin, R. Acc. Chem. Res. 2012, 45,
Corresponding Author
rongchao@andrew.cmu.edu.
1
(
470−1479.
31) Liu, Z.; Zhu, M.; Meng, X.; Xu, G.; Jin, R. J. Phys. Chem. Lett.
2
(
2
(
011, 2, 2104−2109.
32) Qian, H.; Zhu, M.; Gayathri, C.; Gil, R. R.; Jin, R. ACS Nano
011, 5, 8935−8942.
33) Corma, A.; Concepcion, P.; Dominguez, I.; Fornes, V.; Sabater,
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
M. J. J. Catal. 2007, 251, 39−47.
■
(
(
34) Haruta, M. CATTECH 2002, 6, 102−115.
35) Makosch, M.; Sa, J.; Kartusch, C.; Richner, G.; van Bokhoven, J.
This work is financially supported by U. S. Department of
Energy, Office of Basic Energy Sciences (Grant DE-FG02-
2ER16354). We thank Prof Yu Lei for assistance with STEM
́
̈
A.; Hungerbuhler, K. ChemCatChem 2012, 4, 59−63.
1
(36) Maity, P.; Tsunoyama, H.; Yamauchi, M.; Xie, S.; Tsukuda, T. J.
Am. Chem. Soc. 2011, 133, 20123−20125.
imaging.
1
1353
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354

Journal of the American Chemical Society
Article
(
37) Maity, P.; Wakabayashi, T.; Ichikuni, N.; Tsunoyama, H.; Xie,
S.; Yamauchi, M.; Tsukuda, T. Chem. Commun. 2012, 48, 6085−6087.
38) Maity, P.; Takano, S.; Yamazoe, S.; Wakabayashi, T.; Tsukuda,
T. J. Am. Chem. Soc. 2013, 135, 9450−9457.
39) Nykanen, L.; Hakkinen, H.; Honkala, K. Carbon 2012, 50,
752−2763.
40) Kilpin, K. J.; Horvath, R.; Jameson, G. B.; Telfer, S. G.; Gordon,
K. C.; Crowley, J. D. Organometallics 2010, 29, 6186−6195.
41) Chen, W.; Zuckerman, N. B.; Kang, X.; Ghosh, D.; Konopelski,
J. P.; Chen, S. J. Phys. Chem. C 2010, 114, 18146−18152.
42) Shichibu, Y.; Negishi, Y.; Watanabe, T.; Chaki, N. K.;
Kawaguchi, H.; Tsukuda, T. J. Phys. Chem. C 2007, 111, 7845−7847.
43) Qian, H.; Eckenhoff, W. T.; Bier, M. E.; Pintauer, T.; Jin, R.
Inorg. Chem. 2011, 50, 10735−10739.
44) Kang, X.; Zuckerman, N. B.; Konopelski, J. P.; Chen, S. J. Am.
Chem. Soc. 2012, 134, 1412−1415.
(
(
̈
̈
2
(
(
(
(
(
1
1354
dx.doi.org/10.1021/ja503724j | J. Am. Chem. Soc. 2014, 136, 11347−11354