Polymer-incarcerated gold-palladium nanoclusters with boron on carbon: A mild and efficient catalyst for the sequential aerobic oxidation-Michael addition of 1,3-dicarbonyl compounds to allylic alcohols

Article abstract of DOI:10.1021/ja110142y

We have developed a polymer-incarcerated bimetallic Au-Pd nanocluster and boron as a catalyst for the sequential oxidation-addition reaction of 1,3-dicarbonyl compounds with allylic alcohols. The desired tandem reaction products were obtained in good to excellent yields under mild conditions with broad substrate scope. In the course of our studies, we discovered that the excess reducing agent, sodium borohydride, reacts with the polymer backbone to generate an immobilized tetravalent boron catalyst for the Michael reaction. In addition, we found bimetallic Au-Pd nanoclusters to be particularly effective for the aerobic oxidation of allylic alcohols under base- and water-free conditions. The ability to conduct the reaction under relatively neutral and anhydrous conditions proved to be key in maintaining good catalyst activity during recovery and reuse of the catalyst. Structural characterization (STEM, EDS, SEM, and N₂ absorption/desorption isotherm) of the newly prepared PI/CB-Au/Pd/B was performed and compared to PI/CB-Au/Pd. We found that while boron was important for the Michael addition reaction, it was found to alter the structural profile of the polymer-carbon black composite material to negatively affect the allylic oxidation reaction.

Full text of DOI:10.1021/ja110142y

ARTICLE
pubs.acs.org/JACS
Polymer-Incarcerated Gold-Palladium Nanoclusters with Boron on
Carbon: A Mild and Efficient Catalyst for the Sequential Aerobic
Oxidation-Michael Addition of 1,3-Dicarbonyl Compounds to Allylic
Alcohols
Woo-Jin Yoo, Hiroyuki Miyamura, and Shu Kobayashi*
Department of Chemistry, School of Science and Graduate School of Pharmaceutical Sciences, Hongo, Bunkyo-ku,
Tokyo 113-0033, Japan
S Supporting Information
b
ABSTRACT: We have developed a polymer-incarcerated bi-
metallic Au-Pd nanocluster and boron as a catalyst for
the sequential oxidation-addition reaction of 1,3-dicarbonyl
compounds with allylic alcohols. The desired tandem reac-
tion products were obtained in good to excellent yields under
mild conditions with broad substrate scope. In the course of
our studies, we discovered that the excess reducing agent, sodium borohydride, reacts with the polymer backbone to generate
an immobilized tetravalent boron catalyst for the Michael reaction. In addition, we found bimetallic Au-Pd nanoclusters to
be particularly effective for the aerobic oxidation of allylic alcohols under base- and water-free conditions. The ability to
conduct the reaction under relatively neutral and anhydrous conditions proved to be key in maintaining good catalyst activity
during recovery and reuse of the catalyst. Structural characterization (STEM, EDS, SEM, and N2 absorption/desorption
isotherm) of the newly prepared PI/CB-Au/Pd/B was performed and compared to PI/CB-Au/Pd. We found that while boron
was important for the Michael addition reaction, it was found to alter the structural profile of the polymer-carbon black
composite material to negatively affect the allylic oxidation reaction.
’ INTRODUCTION
with allylic alcohols catalyzed by PI-Au nanocluster/Lewis acid
(eq 1).
In nature, complex organic molecules are constructed ele-
gantly in confined environments through multistep cascade
reactions. In order to mimic the catalytic efficiencies enjoyed
by nature, synthetic chemists have shown great interest in the
preparation and application of self-assembled nanoreactors for
catalysis.1 In addition, the development of one-pot, multistep
homogeneous catalysis has been growing due to the inherent
efficiency caused by the elimination of isolation and purification
steps of intermediates generated from traditional iterative syn-
thetic methods.2
In our laboratory, we have a long-standing interest in the
immobilization of metal catalysts in self-assembled microstruc-
tures with the dual purpose of facilitating recovery and reuse and
serving as a reactive environment for efficient catalysis.3 Recently,
we have reported the synthesis and use of styrene-based polymer-
incarcerated (PI) transition metals and Lewis acids as robust and
highly active heterogeneous catalysts. As a natural extension of
this chemistry, we considered the incarceration of two different
metal catalysts that can perform two distinct organic transforma-
tions on our polymer support as an efficient heterogeneous
catalyst for tandem reactions (Scheme 1).4
’ RESULTS AND DISCUSSION
Discovery of an Immobilized Boron Catalyst for the
Michael Reaction. We began our study by examining the
compatibility of the mixed PI catalysts for the Michael addition
of β-ketoester 1a to methyl vinyl ketone in DCM at room
temperature (Table 1).
As shown in entry 1, we confirmed that Sc(OTf)3 is an
excellent catalyst to provide the desired Michael adduct 3a in
excellent yield. We also confirmed that PI-Au is a poor catalyst for
the Michael reaction (entry 2).5 However, to our disappoint-
ment, we found PI-Au/Sc to be an ineffective catalyst as well
(entry 3). We realized that the loss in catalytic activity was due to
the deactivation of Sc(OTf)3 as a Lewis acid during the prepara-
tion step caused by the reducing agent required for the Au
In particular, we were interested in the incorporation of our
recently reported aerobic oxidation reaction catalyzed by Au
nanoclusters, and as such we decided to examine the sequential
aerobic oxidation-Michael reaction of 1,3-dicarbonyl compounds
Received: November 11, 2010
Published: February 08, 2011
r
2011 American Chemical Society
3095
dx.doi.org/10.1021/ja110142y J. Am. Chem. Soc. 2011, 133, 3095–3103
|

Journal of the American Chemical Society
ARTICLE
nanocluster formation.6 While only a trace amount of the desired
Michael adduct 3a was observed, we believed it was possible to
improve the reaction by increasing the concentration of the
reaction mixture. As shown in entries 4 and 5, a 10-fold increase
in concentration resulted in the formation of 3a with an excellent
yield. With this result in hand, we performed control studies in
order to confirm that the Au nanoclusters were responsible for
the catalytic activity. We established that the background reac-
tion (entry 6) and the cross-linked polymer support (entry 7) did
not play a significant role for the formation of 3a. However, when
we examined the cross-linked polymer with just the reducing
agent (NaBH4) as a catalyst for the Michael reaction, we were
surprised to find that it was just as effective as PI-Au (entry 8).7,8
We re-examined the metal content of PI-Au by inductively
coupled plasma-atomic emission spectroscopy (ICP) and found
significant amounts of boron (1.469 mmol/g). Furthermore,
when we prepared PI-Au without boron, we could not observe
any catalytic activity for the Michael reaction (entry 9).9
Scheme 1. Sequential Catalysis in a Confined Environment
Although we prepared (Scheme 2) and demonstrated that PI-B
is an active catalyst for the Michael reaction, there were some
technical difficulties associated with its synthesis. When NaBH4
was added to the dissolved polymer solution, over time (within 1 h),
the polymer began to cross-link and stirring became near
impossible. We overcame this issue by the introduction of carbon
black (CB) as a dispersion and support material. When NaBH4
was added to the heterogeneous mixture of the polymer-coated
CB, slight thickening of the composite material was observed, but
this did not prevent stirring. The newly prepared polymer-
incarcerated boron on carbon black (PI/CB-B) was applied to
the Michael reaction, and the desired Michael adduct 3a was
obtained in good yield (entry 10). Further optimization by
solvent screening (MeCN, hexane, THF, toluene) revealed that
both toluene and THF provides the desired product 3a in
quantitative yields (entry 11).
Table 1. Optimization of Reaction Conditions for the
Michael Reactiona
entry
catalyst
yield (%)b
1c
2c
3c
4
Sc(OTf)3 (2 mol %)
PI-Au (2 mol %)
PI-Au/Sc (2 mol %)
PI-Au/Sc (2 mol %)
PI-Au (2 mol %)
-
>95
<5
Chart 1. List of Michael Donors and Acceptors
<5
>95
>95
<5
5
6
7
cross-linked polymer
PI-B (5 mol %)
<5
8
>95
<5
9
PI-Au (boron-free) (5 mol %)
PI/CB-B (5 mol %)
PI/CB-B (5 mol %)
10
11d
90
>95
a β-Ketoester 1a (0.26 mmol), methyl vinyl ketone (0.39 mmol), and
catalyst in DCM (C = 1.0 M) at room temperature. b Yield based on
1a and determined by 1H NMR using mesitylene as an internal standard.
c C = 0.1 M. d Using toluene or THF as a solvent.
Scheme 2. Preparation Methods for PI-B and PI/CB-B
3096
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103

Journal of the American Chemical Society
ARTICLE
Table 2. Scope of the PI/CB-B-Catalyzed Michael Reactiona
Table 3. Recovery and Reuse of PI/CB-B Catalysta
run
yield (%)b
run
yield (%)b
1
2
3
4
5
>95
>95
>95
>95
>95
6
>95
>95
>95
93
7
8
9
10
93
a β-Ketoester 1a (1.0 equiv), vinyl ketone 2a (1.5 equiv), and PI/CB-B
(5.0 mol %) in THF (C = 1.0 M) at room temperature. b Yield based on
1a and determined by 1H NMR using mesitylene as an internal standard.
As expected, when β-ketoesters 1a-c were utilized as nucleo-
philes, the desired 1,4-addition products were obtained with
excellent yields when coupled with 2a (entries 1-3). Further-
more, 1,3-diketones 1d,e were found to be good Michael donors
(entries 4 and 5). However, when dimethyl malonate (1f) was
used as a substrate, the yield was slightly diminished. Under the
assumption that malonate 1f was less reactive due to being less
acidic than β-ketoesters and 1,3-diketones, we decided to use
chlorinated malonate 1g (entry 7). However, we did not observe
the expected improvement in the yield for the reaction. Next, we
examined the Michael acceptors. We found that reactive and
unstable acrolein (2b) was a viable electrophile when utilized at a
lower temperature (-20 °C) and furnished the desired product
3h with a modest yield (entry 8). In cases where nonvolatile
Michael acceptors such as oxazolidone 2c and vinyl sulfone 2d
were used, the amounts required for the electrophile could be
reduced and still provide the desired products in excellent yields
(entries 9 and 10).
Next, we examined the viability of recycling the PI/CB-B
catalyst (Table 3).
We found that PI/CB-B could be recovered and reused up to 7
times without noticeable loss of activity. Furthermore, pretreat-
ment of PI/CB-B was not necessary in order to reactivate the
catalyst. However, we found it was necessary to purify 2a and
remove the hydroquinone stabilizer prior to use in order to
achieve good recyclability.10
Sequential Aerobic Oxidation-Michael Addition of 1,3-
Dicarbonyl Compounds to Allylic Alcohols. With our seren-
dipitous discovery of an immobilized boron catalyst from the
excess NaBH4 required for the reduction of metal nanoclusters,
we decided to examine the viability of the sequential aerobic
oxidation-Michael addition of 1,3-dicarbonyl compounds to
allylic alcohols (Table 4). We decided to utilize bimetallic
nanoclusters on a polymer-carbon black (PI/CB-Au/M/B,
M = second metal) composite material as catalysts since we
anticipate these heterogeneous catalysts to facilitate the desired
sequential reaction under base-free11 conditions in relatively high
concentrations.12 The preparation of the PI/CB-Au/M/B com-
posite materials followed our previously reported synthesis of
PI/CB-Au/Pd with the minor exception of removing the aqu-
eous washing sequence in order to retain the boron required for
the Michael reaction.13
a 1,3-Dicarbonyl compounds 1a-g (0.26 mmol), Michael acceptors
2a-d (0.39 mmol), and PI/CB-B (5.0 mol %) in THF (C = 1.0 M) at
room temperature. b Yield based on 1a-g and determined by weight of
c
the isolated product. Reaction performed at -20 °C for 12 h. d 1.1
equiv of the Michael acceptor used.
With the optimized conditions in hand, we decided to examine
the scope of the Michael addition of 1,3-dicarbonyl compounds
(1a-g) with various electron-deficient alkenes (2a-d) (Chart 1)
catalyzed by PI/CB-B (Table 2).
On the basis of our previous studies with aerobic oxidation of
alcohols using bimetallic nanoclusters, we initially examined PI/
CB-Au/Pt/B as a catalyst in the absence (entry 1) and presence
3097
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103

Journal of the American Chemical Society
ARTICLE
Table 4. Optimization of Reaction Conditions for the Se-
quential Aerobic Oxidation of β-Ketoester 1a to Allylic
Alcohol 5a a
Chart 2. Allylic Alcohols 5a-i
entry
M
metal source
Na2PtCl6
solvent
yield (%)b
1c
2
Pt
DCM
DCM
BTF
N.R.
65
59
54
69
70
>95
12
22
7
Pt
Na2PtCl6
3
Pt
Na2PtCl6
4
Pt
Na2PtCl6
MeCN
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
THF
5
Pt
Na2PtCl6
6
Ru
Pd
Cu
Ni
Co
Rh
Fe
Pt
[RuCl2(p-cymene)]2
Pd(OAc)2
Cu(OAc)2
(PPh3)2NiCl2
CoCl2
7
Next, we examined a variety of different metals and found that Pd
(entry 7) provides the desired Michael adduct 3a in excellent
yield. We also found that the yield of 3a could be improved when
the catalyst loading is slightly increased for bimetallic nanoclus-
ters containing Pt and Ru (entries 13 and 14). As a control, we
examined the tandem reaction with PI/CB-Au/B (entry 15) and
PI/CB-Pd/B (entry 16) to clearly show the beneficial effect of
utilizing a bimetallic system.
8
9
10
11
12
13d
14d
15e
16f
[RhCl(COD)]2
FeCl2
22
8
Na2PtCl6
>95
>95
25
48
Ru
-
[RuCl2(p-cymene)]2
-
Next, we examined the viability of recycling the PI/CB-Au/
M/B catalysts (Table 5).
Pd
Pd(OAc)2
a β-Ketoester 1a (0.50 mmol), allylic alcohol 5a (0.75 mmol), and PI/
CB-Au/M/B (1:1 Au:M, 1 mol % wrt Au) in solvent:H2O (C = 1.0 M) at
30 °C in O2 (1 atm). b Yield based on 1a and determined by 1H NMR
using mesitylene as an internal standard. c In the absence of H2O. d Using
1.5 mol % of PI/CB-Au/M/B. e Using 2.0 mol % of PI/CB-Au/B. f Using
2.0 mol % of PI/CB-Pd/B.
We found that in both cases with PI/CB-Au/Ru/B and PI/
CB-Au/Pd/B, the catalytic activity was diminished after each
subsequent recovery and reuse. Examination by ICP analysis of
the resulting filtrate of the reaction mixture revealed significant
leaching of both boron and palladium. Although we believed that
water was responsible for the metal leaching, our previous results
suggested that water plays an important role in the aerobic
oxidation of the allylic alcohol. However, to our delight, we found
that PI/CB-Au/Pd/B could catalyze the sequential aerobic
oxidation-Michael addition reaction in the absence of added
water. The recovered Au-Pd bimetallic nanoclusters were
shown to catalyze the desired tandem reaction with excellent
yields without noticeable loss of the desired Michael adduct 3a
up to the third run.
Table 5. Recovery and Reuse Studies of PI/CB Bimetallic
Catalystsa,b
With the optimized reaction in hand, we examined the scope
of the PI/CB-Au/Pd/B catalyzed sequential aerobic oxidation-
Michael addition of 1,3-dicarbonyl compounds 1a-f with allylic
alcohols 5a-i (Chart 2) (Table 6).
1.5 mol % PI/CB-Au/Ru/B
1 mol % PI/CB-Au/Pd/B
THF:H2O
run THF:H2O (9:1) (%) THF (%)
(9:1) (%)
THFd (%)
As expected, when β-ketoesters 1a-c were utilized as nucleo-
philes, the desired 1,4-addition products were obtained in
excellent yields with allylic alcohol 5a (entries 1-3). Further-
more, 1,3-diketones 1d,e were found to be good Michael donors
(entries 4 and 5). When dimethyl malonate (1f) was used as a
substrate, higher reaction temperature and slight excess of 5a was
required in order to obtain a high yield of 3f (entry 6). Next, we
examined allylic alcohols 5b-i (entries 7-14). We found that
the reaction temperature must be slightly increased in order to
fully convert some of the allylic alcohols to the corresponding
R,β-unsaturated ketones required for the Michael reaction with
β-ketoester 1a. Fortunately, we did not observe any significant
decomposition or isomerization of these substrates at elevated
temperatures and were able to decrease the equivalence required
for these relatively nonvolatile Michael acceptors without sacrifi-
cing the excellent yields enjoyed for the sequential aerobic
oxidation-Michael addition reaction.
1
2
3
4
5
6
>95
66
29
-
<5
-
-
-
-
-
>95
78
64
-
>95
>95
>95
91
-
-
93
-
-
88
a β-Ketoester 1a (0.50 mmol) and allylic alcohol 5a (0.75 mmol) in
solvent (C = 1.0 M) at 30 °C in O2 (1 atm). b Yield based on 1a and
determined by 1H NMR using mesitylene as an internal standard.
c Average metal leaching determined by ICP analysis of the reaction
filtrate (Au, N.D.; Pd, 3%; B, 8%). d Average metal leaching determined
by ICP analysis of the reaction filtrate (Au, N.D.; Pd, 0.8%; B, 0.2%).
of water (entry 2) to learn that water is a critical component to
provide 3a with a modest yield. Screening of solvents revealed
that THF provides a slight improvement in yield (entries 2-5).
3098
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103

Journal of the American Chemical Society
ARTICLE
Table 6. Scope of the PI/CB-Au/Pd/B-Catalyzed One-Pot Sequential Reactiona
a 1,3-Dicarbonyl 1a-f (0.50 mmol), allylic alcohol 5a,c (0.75 mmol), and PI/CB-Au/Pd/B (1:1 Au:Pd, 1 mol % wrt Au) in THF (C = 1.0 M) at 30 °C in
O2 (1 atm). b Yield based on 1a-f and determined by weight of the isolated products 3a-f, 3h, 3k-q. c 60 °C. d 2.0 equiv of 5a. e 1.1 equiv of allylic
alcohol 5b,d-i.
Structural Characterization of PI Catalysts. With the un-
expected immobilization of boron and its use as a catalyst for the
Michael reaction, we wanted to understand the nature of the
boron immobilization process. From our previous efforts in
developing highly active immobilized metal catalysts, the im-
mobilization process is believed to occur through the encapsula-
tion of the metal species via nonbonding interaction between the
π-electrons of the aromatic rings of the polystyrene backbone
and the vacant orbitals of the metal species.3b However, in the
case of boron, we believe that the boron immobilization occurs
through the formation of a tetravalent boron complex through
the esterification of NaBH4 with the alcohol functional groups
present on the polymer backbone (Scheme 3).
Scheme 3. Immobilization of Boron
This hypothesis is quite reasonable since Echavarren14 and
Sundararajan15 both reported the preparation of tetraalkoxybo-
rates from NaBH4 and alcohols and these anionic species were
shown to be excellent catalysts for the Michael reaction. Evidence
to support the existence of the tetravalent boron species was
found through 11B CP-MAS (cross-polarization magic angle
spinning) NMR spectrum of PI/CB-B. While a mixture of broad
boron peaks were observed, the major peak of the spectrum was
at 0.46 ppm, which is indicative of a tetravalent alkoxy boron
species.16
In addition, we conducted scanning transmission electron
microscopy (STEM) and energy dispersive spectroscopy
(EDS) analyses to characterize and understand the newly pre-
pared bifunctional composite catalyst (Figure 1).
From the EDS line analysis, a single Gaussian distribution of
X-rays of Au and Pd across the nanoparticle (3.6 ( 1.3 nm) was
obtained and is indicative of the random arrangement of Au and
Pd atoms within a Au/Pd alloy. EDS analysis also revealed a
higher ratio of Au:Pd (∼1.3:1 to 2.5:1 of Au:Pd, Figure S-3,
Supporting Information) within the metal nanoclusters while
3099
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103

Journal of the American Chemical Society
ARTICLE
Figure 1. STEM and EDS analysis of PI/CB-Au/Pd/B: (A) Typical STEM image and size distribution of PI/CB-Au/Pd/B; (B) EDS line analysis of a
cluster of PI/CB-Au/Pd/B (red = Au, green = Pd).
(to wash away the boron), the small, uniformed spheres were
restored. We believed that this difference in the structure
between PI/CB-Au/Pd/B and PI/CB-Au/Pd should affect its
physical properties such as surface area. This hypothesis was
confirmed by the N2 adsorption/desorption analysis of the PI/
CB catalysts (Figure 3).
On the basis of the adsorption/desorption isotherm, the
Bruanauer-Emmett-Teller (BET) surface area was calculated
to be 19.1 m2/g for PI/CB-Au/Pd/B and 56.4 m2/g for PI/CB-
Au/Pd.17 Examination of the isotherms also revealed these PI/
CB catalysts to be essentially macroporous since the hysteresis
loops were only noticeable at relatively high pressures and that
the adsorption isotherms increased sharply near the saturation
vapor pressures.18
Comparison Studies. We performed some control studies to
understand our catalyst system and to demonstrate advantages of
performing the tandem aerobic oxidation-Michael addition
reaction with the bifunctional PI/CB-Au/Pd/B. We first exam-
ined the Michael addition reaction between β-ketoester 1a and
methyl vinyl ketone (MVK) (Table 7).
Figure 2. Typical SEM image of (A) CB; (B) CB þ polymer 4; (C) PI/
CB-Au/Pd/B; (D) PI/CB-Au/Pd.
From the data shown in Table 7, the order of reactivity of PI/
CB-M catalysts for the Michael reaction is PI/CB-B > PI/CB-Au >
PI/CB-Pd > PI/CB-Au/Pd. Although there exists the possibi-
lity that the metal nanoparticles may interact with the boron
species to alter its reactivity, we believe this trend simply reflects
the amount of tetravalent boron that exists in the various PI/
CB-M catalysts. Since 1 equiv of NaBH4 is required for the
reduction of M(I) to M(0), PI/CB-Au/Pd/B should possess
the least amount of the active tetravalent boron Michael
addition catalysts.
Next, we performed comparison studies between our compo-
site bifunctional PI/CB-Au/Pd/B catalyst and a mixed catalyst
system composed of PI/CB-Au/Pd and PI/CB-B to see if an
advantage in the rate of reaction is gained from the bifunctional
catalysts (Table 8).19
ICP analysis of the PI/CB-Au/Pd/B determined the Au:Pd ratio
to be essentially 1:1. These results suggest that Pd exists in the
metal nanoclusters and is also dispersed throughout the poly-
mer/CB composite material.
In addition, the scanning electron microscopy (SEM) image of
PI/CB-Au/Pd/B was obtained and significant differences in the
structural morphologies of the composite material with and
without boron were revealed (Figure 2).
From the SEM image of CB, relatively uniform spheres (≈45
nm) were observed and the polymer-CB composite material
also possessed similar structural profile with slightly larger
spheres (≈55 nm). The increase observed size of the micro-
spheres is most likely due to the polymer absorbed on the
surface of the carbon black particles. Surprisingly, the SEM
image of PI/CB-Au/Pd/B revealed dramatic structural
changes with uneven and large (>100 nm) spheres. These
large spheres may represent clustered polymer-CBs held
together by cross-linked borate ester bonds. Indeed, when
PI/CB-Au/Pd/B was washed thoroughly with H2O and THF
As shown in Table 8, the aerobic oxidation-Michael reac-
tion was slightly faster when the mixed catalyst system was
employed. One possible explanation for this result is the
decrease in the rate of the allylic oxidation using PI/CB-Au/
Pd/B due to the reduced surface area of the polymer-CB
3100
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103

Journal of the American Chemical Society
ARTICLE
Figure 3. Nitrogen adsorption/desorption isotherm of (A) PI/CB-Au/Pd/B; (B) PI/CB-Au/Pd (þ = adsorption, O = desorption).
Table 7. Comparison of Various PI/CB-M Catalysts for the
Michael Reactiona
Table 8. Comparison between Bifunctional Composite Cat-
alyst and Mixed Catalyst System for the Sequential Oxida-
tion-Michael Reactiona
PI/CB-M yield 3a (%)b
yield 3l (%)b
entry
t (h)
Au/Pd/Bc
Pd/Bd
Au/Be
Bf
entry
t (h)
PI/CB-Au/Pd/Bc
PI/CB-Au/Pd þ PI/CB-Bd
1
2
0.5
1.5
40
73
60
93
76
>95
>95
>95
1
2
3
4
5
0.5
1.5
4
N.D.
<5
N.D.
6
a β-Ketoester 1a (0.50 mmol), alkene 2a (0.75 mmol), and mesity-
lene (0.17 mmol) in THF-d8 (C = 1.0 M) at 30 °C. b Yield based on 1a
and determined by 1H NMR using mesitylene as an internal standard.
c Metal loading of PI/CB-Au/Pd/B: 0.1962 mmol/g Au, 0.1847
mmol/g Pd, 1.778 mmol/g B. d Metal loading of PI/CB-Pd/B:
0.1783 mmol/g Pd, 2.467 mmol/g B. e Metal loading of PI/CB-Au/
B: 0.2041 mmol/g Au, 2.371 mmol/g B. f Metal loading of PI/CB- B:
2.173 mmol/g B.
13
18
8
35
49
22
>95
>95
a β-Ketoester 1a (0.50 mmol), allylic alcohol 5d (0.56 mmol), and
mesitylene (0.17 mmol) in THF-d8 (C = 1.0 M) at 60 °C in O2 (1 atm).
b Yield based on 1a and determined by 1H NMR using mesitylene as an
internal standard. c Catalyst loading of PI/CB-Au/Pd/B: 1.0 mol % Au,
1.2 mol % Pd, 16.2 mol % B. d Catalyst loading of PI/CB-Au/Pd þ PI/
CB-B: 1.0 mol % Au, 1.1 mol % Pd, 13.2 mol % B.
composite material. Indeed, when the allylic oxidation of 5d
was examined with PI/CB-Au/Pd/B and PI/CB-Au/Pd, we
observed an increase in the rate of the oxidation reaction with
the boron-free catalyst (Table 9).
reaction occurred smoothly under our optimized reaction con-
ditions (eq 2).
These results also demonstrate the synthetic utility of per-
forming the tandem allylic oxidation-Michael addition reaction
since the R,β-unsaturated ketone 2e quickly reacts to form the
Michael adduct 3l instead of undergoing side reactions. In fact,
while monitoring the sequential oxidation-Michael reaction
shown in Table 8, the expected intermediate, R,β-unsaturated
1
ketone 2e, was not detected by H NMR. Interestingly, by
’ CONCLUSION
comparing the results from Tables 8 and 9, we observe a decrease
in the rate of the allylic oxidation reaction in the presence of β-
ketoester 1a. This may suggest that coordinating ligands can
affect the rate of the oxidation reaction catalyzed by the bimetallic
nanoclusters.
In summary, we have developed a novel sequential aerobic
oxidation-Michael addition reaction of 1,3-dicarbonyl com-
pounds to allylic alcohols using a Au-Pd bimetallic nanocluster
with boron immobilized on a polymer-CB composite material.
The PI/CB-Au/Pd/B nanoclusters proved to be an excellent
catalyst that delivered the desired Michael adducts in good to
Finally, we examined the sequential reaction using air (under
balloon pressure) as the terminal oxidant and found that the
3101
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103

Journal of the American Chemical Society
ARTICLE
Table 9. Comparison between PI/CB-Au/Pd/B and PI/
measurements at 77 K were performed on a ASAP 2010 volumetric
adsorption analyzer surface area and porosity analyzer equipped with
pressure transducers covering the 1, 10, and 1000 Torr ranges
(Micromeritics Instrument Corp.). The instrument was equipped with
a turbo vacuum pump capable of reducing the pressure to below 10-7
Torr. The initial pressure at the start of any run was below 10-5 Torr.
Reagents. Unless stated otherwise, commercial reagents were used
as received with the exception of NaBH4, which was recrystallized from
diglyme according to the literature.21 The carbon black source used in
this study was ketjen black EC 300J purchased from Lion Corporation.
Allylic alcohols 2b, 2d-i were prepared according to the literature.22
Preparation of PI/CB-B. To a solution of copolymer 4 (0.1563 g)
in diglyme (3.9 mL) at 40 °C was added ketjen black (0.1573 g). After
stirring the heterogeneous solution for 15 min, a solution of NaBH4 (17.7
mg, 0.469 mmol) in diglyme (1.6 mL) was added dropwise. After the
heterogeneous solution was stirred for 4 h at 40 °C, diethyl ether (40 mL)
was added dropwise. The boron-containing polymer-coated carbon black
waswashedseveraltimeswithdiethyletheranddriedat room temperature.
Next, the recovered solid powder was heated at 170 °C for 4 h without
solvent. After being cooledto room temperature, the cross-linked polymer-
coated carbon black was washed several times with DCM, filtered, ground
with a mortar and pestle, and dried under vacuum to provide the desired
black powder (PI/CB-B, 0.3312 g, B loading: 0.9624 mmol/g).
B-Au/Pd for the Allylic Oxidation of 5da
PI/CB-Au/Pd/Bc
PI/CB-Au/Pdd
entry t (h) yield 2e (%)b conv (%)b yield 2e (%)b conv (%)b
1
2
3
4
5
0.5
1.5
4
N.D.
<5
15
<5
14
26
46
48
25
20
44
13
33
64
8
25
60
91
22
59
>95
>95
a Allylic alcohol 5d (0.50 mmol), and mesitylene (0.17 mmol) in THF-
d8 (C = 0.86 M) at 60 °C in O2 (1 atm). b Yield and conversion is based
1
on 5d and determined by H NMR using mesitylene as an internal
standard. c Catalyst loading of PI/CB-Au/Pd/B: 1.0 mol % Au, 1.2 mol
% Pd, 16.2 mol % B. d Catalyst loading of PI/CB-Au/Pd: 1.0 mol % Au,
1.1 mol % Pd.
Typical Procedure for the PI/CB-B-Catalyzed Michael Ad-
dition of 1,3-Dicarbonyl Compounds to Electron-Deficient
Olefins. To a screw-cap glass vial were added PI/CB-B (0.0261 g,
0.0251 mmol, 5 mol %), β-ketoester 1a (87.0 μL, 0.502 mmol), and vinyl
ketone 2a (61.0 μL, 0.754 mmol). The mixture was dissolved in THF
(0.5 mL) and stirred for 20 h at room temperature. Then the reaction
mixture was filtered and washed with DCM, and the resulting filtrate was
concentrated under reduced pressure. The crude product was purified
by column chromatography (EtOAc:hexane = 1:4) to provide 3a
(0.1239 g, 0.472 mmol, 94%) as a clear colorless oil.
Ethyl 2-Benzoyl-5-oxohexanoate (3a) (Table 2, entry 1). Rf
0.22 (EtOAc:hexane = 1:4); 1H NMR (CDCl3, 500 MHz) δ 8.00-7.91
(m, 2H), 7.53-7.50 (m, 1H), 7.43-7.40 (m, 2H), 4.37 (dd,
J = 6.3, 6.2 Hz, 1H), 4.10-4.04 (m, 2H), 2.58-2.47 (m, 2H),
2.21-2.10 (m, 2H), 2.06 (s, 3H), 1.09 (t, J = 7.4 Hz, 3H); 13C NMR
(CDCl3, 125 MHz) δ 207.8, 195.2, 169.7, 135.8, 133.6, 128.7, 128.6,
61.3, 52.5, 40.4, 29.9, 22.6, 13.9. This is a known compound, and the
spectral data are identical to those reported in literature.23
Preparation of PI/CB-Au/Pd/B. To a solution of 4 (0.1448 g) in
diglyme (9.6 mL) was added ketjen black (0.1458 g). The resulting
suspension was cooled to 0 °C, and then a solution of NaBH4 (46.0 mg,
1.21 mmol) in diglyme (2.2 mL) was added dropwise. The reaction
mixture was stirred for approximately 15 min, and then a solution of
PPh3AuCl (40.2 mg, 0.0811 mmol) and Pd(OAc)2 (18.2 mg, 0.0811
mmol) in diglyme (5.8 mL) was added dropwise. The reaction mixture
was allowed to warm to roomtemperature and was stirred for 4 h. Diethyl
ether (80 mL) was added dropwise, and then the suspension was filtered
and washed with excess diethyl ether. The isolated polymer-coated
carbon black was allowed to dry, gently ground with a mortar and pestle,
and then heated at 170 °C for 4 h. The polymer-incarcerated catalyst was
then resuspended in DCM, filtered, washed with excess DCM, and
allowed to dry to provide PI/CB-Au/Pd/B (0.4035 g, Au loading: 0.2007
mmol/g, Pd loading: 0.2183 mmol/g, B loading: 2.456 mmol/g).
Typical Procedure for the PI/CB-Au/Pd/B-Catalyzed Se-
quential Aerobic Oxidation-Michael Addition of 1,3-Dicar-
bonyl Compounds to Allylic Alcohols. In a screw-cap glass vial
was added PI/CB-Au/Pd/B (0.0250 g, 5.02 μmol, 1 mol % with respect
to Au), β-ketoester 1a (87.0 μL, 0.502 mmol), and allylic alcohol 5d
(73.0 μL, 0.555 mmol) and dissolved in THF (0.5 mL). The reaction
mixture was allowed to stir for 20 h at 60 °C under a balloon of oxygen
gas. Then the reaction mixture was filtered and washed with DCM, and
excellent yields under mild conditions. We have uncovered the
dual role that NaBH4 plays as both the reducing agent required
for the generation of the bimetallic nanoclusters and as a catalyst
precursor for the Michael reaction. In addition, the Au-Pd
bimetallic nanoclusters were shown to be a remarkably effective
catalyst that was able to oxidize allylic alcohols without addition
of water as a cosolvent under neutral conditions. By eliminating
the water cosolvent, the leaching of boron was minimized to
maintain good catalyst activity during the recovery and reuse
cycles. Finally, we demonstrate the structural changes of the
polymer-CB composite material caused by boron and its
negative effects on the catalytic activity for the allylic oxidation
reaction.
’ EXPERIMENTAL SECTION
General Information. Standard column chromatography was
performed on 70-230 mesh silica gel (obtained from Merck) using
flash column chromatography techniques.20 IR spectra were measured
on a JASCO FT/IR-610 spectrometer. Melting points were obtained
using a B€uchi Melting Point B-545 and were not corrected. 1H and 13
C
NMR spectra were recorded on a JEOL ECX-500 in CDCl3. Chemical
shifts were reported in parts per million (ppm) from tetramethylsilane
using the solvent resonance as the internal standard (chloroform: δ 7.26
ppm) for 1H NMR and (deuterochlorofom: δ 77.0 ppm) for 13C NMR.
11B CP-MAS NMR spectrum was recorded on a JEOL/JNM ECA-500
spectrometer with a 4 mm MAS probe at spectral frequency 160.47
MHz. The boron spectrum was referenced to H3BO3 as a standard. High
resolution mass spectrometry was carried out using a JEOL JMS-
T100TD (DART). The content of all metals in the final material was
determined by ICP analysis using a Shimadzu ICPS-7510 equipment:
10-20 mg of the final material was heated in a mixture of sulfuric acid
and nitric acid at 200 °C for 2 h. After cooling at room temperature, aqua
regia (HCl:HNO3 = 1:3 v/v) was added and the samples were diluted to
50 mL with distilled water. STEM images and EDS analyses were
obtained using a JEOL JEM-2100F instrument operated at 200 kV. All
TEM specimens were prepared by placing a drop of a suspension of the
sample in methanol on carbon-coated copper grids and allowed to dry
in air without staining. SEM images were obtained using a JEOL
JSM-6700F instrument operated at 5.0 KV. Nitrogen adsorption
3102
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103

Journal of the American Chemical Society
ARTICLE
the resulting filtrate was concentrated under reduced pressure. The
crude product was purified by column chromatography (EtOAc:hexane =
1:4) to provide 3l (0.1544 g, 0.476 mmol, 95%) as a clear colorless oil.
Ethyl 2-Benzoyl-5-oxo-5-phenylpentanoate (3l) (Table 6,
entry 9). Rf 0.36 (EtOAc:hexane = 1:4); IR (neat, NaCl): 3064 (m),
2981 (s), 2937 (s), 1974 (w), 1913 (w), 1734 (s), 1686 (s), 1596 (m),
1448 (s), 1221 (s), 974 (s) cm-1; 1H NMR (CDCl3, 500 MHz) δ 8.07
(d, J = 7.9 Hz, 2H), 7.95 (d, J = 7.4 Hz, 2H), 7.59-7.53 (m, 2H),
7.50-7.43 (m, 4H), 4.57 (dd, J = 7.9, 6.2 Hz, 1H), 4.19-4.10 (m, 2H),
3.22-3.16 (m, 1H), 3.12-3.06 (m, 1H), 2.49-2.36 (m, 2H), 1.15 (t, J =
7.4 Hz, 3H); 13C NMR (CDCl3, 125 MHz) δ 199.2, 195.3, 169.8, 136.6,
135.9, 133.6, 133.1, 128.71, 128.70, 128.6, 128.0, 61.4, 52.8, 35.6, 23.2,
13.9; ESI-HRMS (m/z) calcd for C20H20O4 [(M þ H)þ]: 325.14398,
found: 325.14334.
(8) NaBH4 is a common reducing agent used for the formation of
nanoclusters from metal salts. However, to the best of our knowledge,
the possibility of boron remaining and acting as a catalyst has yet been
explored.
(9) Boron-free PI-Au was prepared by the microencapsulation of
pre-prepared PPh3-stabilized Au nanoclusters with the styrene-based
polymer, followed by thermally induced cross-linking. Alternatively,
boron-free PI-Au can be prepared by stirring the boron-containing
immobilized Au catalyst in a THF/H2O solution overnight, followed by
filtration and drying. For the preparation of PPh3-stablized Au nanoclus-
ters, see: Shichibu, Y.; Negishi, Y.; Tsukuda, T.; Teranishi, T. J. Am.
Chem. Soc. 2005, 127, 13464.
(10) We found that mixing hydroquinone with NaB(OMe)4 in
DCM results in an unidentified blue solid.
(11) The Michael reaction between β-ketoester 1a and methyl vinyl
ketone with 3 equiv of K2CO3 in DCM (C = 1.0 M) provided 53% of the
Michael adduct 3a along with undesired side products. From our
previous studies, we have found that PI bimetallic Au-M nanoclusters
can facilitate the oxidation of alcohols under base-free conditions. See:
Miyamura, H.; Matsubara, R.; Kobayashi, S. Chem. Commun. 2008,
2031.
’ ASSOCIATED CONTENT
S
Supporting Information. Detailed experimental proce-
b
dures and characterization of all new compounds. This material is
free of charge via the Internet at http://pubs.acs.org.
(12) From our previous studies, relatively low levels of Au nanoclus-
ters could be loaded on the polymer support (0.07-0.08 mmol/g of Au)
to ensure small cluster size and high catalytic activity. We have found that
a polymer-carbon black composite material facilitates high Au loading
(0.25 mmol/g of Au) while maintaining small cluster size and high
catalytic activity. See: Lucchesi, C.; Inasaki, T.; Miyamura, H.; Matsu-
bara, R.; Kobayashi, S. Adv. Synth. Catal. 2008, 350, 1996.
(13) Kaizuka, K.; Miyamura, H.; Kobayashi, S. J. Am. Chem. Soc.
2010, 132, 15096.
(14) (a) Gꢁomez-Bengoa, E.; Cuerva, J. M.; Mateo, C.; Echavarren,
A. M. J. Am. Chem. Soc. 1996, 118, 8553. (b) Campa~na, A. G.; Fuentes,
N.; Gꢁomez-Bengoa, E.; Mateo, C.; Oltra, J. E.; Echavarren, A. M.;
Cuerva, J. M. J. Org. Chem. 2007, 72, 8127.
(15) Abraham, S.; Sundararajan, G. Tetrahedron 2006, 62, 1474.
(16) It is most likely that the tetravalent boron is only the resting
state for the true catalyst for the Michael reaction. NaB(OMe)4 is in
equilibrium with B(OMe)3 and NaOH and may act as a bifunctional
acid-base catalyst. For further discussion of the catalytic activity of
NaB(OMe)4, see ref 14b.
’ AUTHOR INFORMATION
Corresponding Author
shu_kobayashi@chem.s.u-tokyo.ac.jp
’ ACKNOWLEDGMENT
This work was partially supported by a Grant-in-Aid for
Scientific Research from the Japan Society of the Promotion of
Science (JSPS), ERATO (JST), NEDO, and GCOE. W.-J.Y.
thanks JSPS for the JSPS Postdoctoral Fellowship for Foreign
Researchers. We also thank Mr. Noriaki Kuramitsu (The Uni-
versity of Tokyo) for STEM, EDS, and SEM analyses, and Mr.
Tomoyoshi Higashi, Mr. Yoshikuni Okumura, and Mr. Tetsuo
Nakajo (SDK) for N2 adsorption/desorption analysis.
’ REFERENCES
(17) Brunauer, S.; Emmett, P. H.; Teller, E. J. Am. Chem. Soc. 1938,
(1) (a) Vriezema, D. M.; Aragonꢀes, M. C.; Elemans, J. A. A. W.;
Cornelissen, J. J. L. M.; Rowan, A. E.; Nolte, R. J. M. Chem. Rev. 2005,
105, 1445. (b) Koblenz, T. S.; Wassenaar, J.; Reek, J. N. H. Chem. Soc.
Rev. 2008, 37, 247.
(2) (a) Lee, J. M.; Na, Y.; Han, H.; Chang, S. Chem. Soc. Rev. 2004,
33, 302. (b) Wasilke, J.-C.; Obrey, S. J.; Baker, R. T.; Bazan, G. C. Chem.
Rev. 2005, 105, 1001. (c) Chapman, C. J.; Frost, C. G. Synthesis 2007, 1.
(d) Shindoh, N.; Takemoto, Y.; Takasu, K. Chem.—Eur. J. 2009, 15,
12168. (e) Zhou, J. Chem. Asian J. 2010, 5, 422.
60, 309.
(18) Kruk, M.; Li, Z.; Jaroniec, M. Langmuir 1999, 15, 1435.
(19) Rate enhancement has been observed for a cascade reaction
catalyzed by two or more enzymes co-immobilized on the surface of
silica nanoparticles. See: Pescador, P.; Katakis, I.; Toca-Herrera, J. L.;
Donath, E. Langmuir 2008, 24, 14108.
(20) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923.
(21) Brown, H. C.; Mead, E. J.; Rao, B. C. S. J. Am. Chem. Soc. 1955,
77, 6209.
(3) (a) Kobayashi, S.; Manabe, K. Acc. Chem. Res. 2002, 35, 209.
(b) Akiyama, A.; Kobayashi, S. Chem. Rev. 2009, 109, 594.
(4) The incorporation of multiple metal catalysts on a single
polymer support is quite rare. See: Pittman, C. U., Jr.; Smith, L. R.
J. Am. Chem. Soc. 1975, 97, 1749.
(22) Briot, A.; Baehr, C.; Brouillard, R.; Wagner, A.; Mioskowski, C.
J. Org. Chem. 2004, 69, 1374.
(23) Takeuchi, M.; Akiyama, R.; Kobayashi, S. J. Am. Chem. Soc.
2005, 127, 13096.
(5) Gold nanoclusters have been previously suggested to possess
Lewis acidic properties under aerobic conditions. See: Kamiya, I.;
Tsunoyama, H.; Tsukuda, T.; Sakurai, H. Chem. Lett. 2007, 36, 646.
(6) The addition of Sc(OTf)3 to a solution of NaBH4 in diglyme
resulted in the evolution of a gas.
(7) The use of immobilized boron as a catalyst is relatively uncom-
mon. For examples, see: (a) Itsuno, S.; Kamahori, K.; Watanabe, K.;
Koizumi, T.; Ito, K. Tetrahedron: Asymmetry 1994, 5, 523. (b) Kamahori,
K.; Ito, K.; Itsuno, S. J. Org. Chem. 1996, 61, 8321. (c) Smith, K.; El-Hiti,
G. A.; Hou, D.; DeBoos, G. A. J. Chem. Soc., Perkin Trans. 1 1999, 2807.
(d) Roesler, R.; Har, B. J. N.; Piers, W. E. Organometallics 2002, 21, 4300.
(e) Itsuno, S.; Arima, S.; Haraguchi, N. Tetrahedron 2005, 61, 12074.
(f) Lakouraj, M. M.; Mokhtary, M. Monatsh. Chem. 2009, 140, 53.
’ NOTE ADDED AFTER ASAP PUBLICATION
After this paper was published ASAP February 8, 2011, the
description of the preparation of PI/CB-B was modified in the
Experimental Section. The corrected version was published
February 11, 2011.
3103
dx.doi.org/10.1021/ja110142y |J. Am. Chem. Soc. 2011, 133, 3095–3103