Effect of composition on the catalytic properties of mixed-ligand-coated gold nanoparticles

Article abstract of DOI:10.1002/anie.201101821

Striped catalysts: The effect of composition and structure on the catalytic efficiency of gold nanoparticles protected by a monolayer composed of two types of ligands differing in length (see picture) is reported. By diluting catalytically active ligand molecules with simple catalytically inactive molecules the catalytic efficiency of the particles is enhanced.

Full text of DOI:10.1002/anie.201101821

Communications
DOI: 10.1002/anie.201101821
Heterogeneous Catalysis
Effect of Composition on the Catalytic Properties of Mixed-Ligand-
Coated Gold Nanoparticles**
Anirban Ghosh, Soubir Basak, Benjamin H. Wunsch, Rajiv Kumar, and Francesco Stellacci*
Ligand-coated nanoparticles (NPs) are hybrid materials with
novel properties.[1,2] Monolayer-protected gold NPs are
emerging as attractive candidates for catalysis,[3,4] in particular
as artificial enzymes (nanozymes).[5,6] There are many reasons
for their success ranging from the ease of forming dense
monolayers of complex catalytic molecules on their ligand
shells to their facile separation from the reaction products by
precipitation. Herein, we present a study on the effect of
composition on the catalytic efficiency of NPs with mixed-
monolayer ligand shells. We show that by using simple
catalytically inactive molecules it is possible to enhance the
catalytic efficiency of NPs.
Recently, we discovered that when two dislike thiolated
ligands are co-assembled on the ligand shell of gold NPs, the
separation of the two ligands generates stripelike domains
that are only a few molecules thick.[7–10] This phenomenon is
observed because of a balance between the enthalpy of phase
separation and the conformational entropy that longer (or
bulkier) ligands gain when surrounded by shorter ones.[11,12]
The difference in the chain length of the two ligands renders
the stripelike domains molecular grooves with the shorter
ligands being embedded in the surface; the bigger the
difference in length the more defined the stripes are supposed
to be.[11,12] As the ligand ratio changes the morphology of the
ligand shell changes; this in turn generates structure-depen-
dent variations in the solubility and interfacial energy of the
particles.[13,14] Herein, we study the effect of the composition
(and hence the morphology) of the ligand shell on the
catalytic activity.
We explore sulfonic-acid-functionalized gold NPs as acid
catalysts. For our experiments, 3-mercapto-1-propanesulfonic
acid (MPSA) was chosen as the precursor for the sulfonic acid
ligands, which are the shorter ligands. The longer ligands are
represented by two different types of nonpolar alkanethiols:
1) linear alkanethiols, namely, butanethiol (BT), hexanethiol
(HT), octanethiol (OT), decanethiol (DT), and dodecanethiol
(DDT) and 2) phenyl-group-terminated alkanethiols, namely,
6-phenylhexanethiol (6-Ph-HT), 8-phenyloctanethiol (8-Ph-
OT), and 10-phenyldecanethiol (10-Ph-DT). The choice of
MPSA was made because of two reasons. First, we focused on
a commercially available inexpensive molecule; second, we
found that most of the structure-driven effects take place
when the active molecule is inside the supramolecular
grooves, and hence the catalytically active ligand needed to
be a short molecule.[13,14] Furthermore we chose to test aryl-
terminated longer ligands to improve the interaction of the
“groove” walls with the reactants. The ligands and the target
nanoparticle size were chosen such that the probability of
formation of stripelike domains is maximized.[13–16]
The syntheses of phenyl-terminated alkanethiol ligands
with alkyl chain lengths 6—10 methylene units long were
performed in a two-step process, as depicted in Scheme 1. A
Scheme 1. Two-step synthesis of phenyl-terminated alkanethiol ligands.
[*] Dr. A. Ghosh, Dr. B. H. Wunsch, Prof. F. Stellacci
Department of Materials Science and Engineering
Massachusetts Institute of Technology
dibromo alkane of desired alkyl chain length was treated with
phenyl lithium under inert atmosphere to form a bromo-
phenyl alkane. The bromide group of the bromophenyl
alkane was substituted by a thiol group by treatment with
hexamethyldisilathane (HMDT) in the presence of tetrabu-
tylammonium fluoride (TBAF).
To perform comparative studies we tried to synthesize
relatively monodisperse NPs. We followed the method
introduced by Stucky and co-workers, which involves the
tert-butylamine borane complex as reducing agent.[17] We
observed that this method does not work well with mixed
ligand systems, especially, when one or both of the ligands are
polar or charged. Therefore, we replaced the usual benzene
and chloroform solvent with N,N-dimethylformamide
(DMF), which dissolves sulfonate and alkanethiol ligands
simultaneously.
77 Massachusetts Avenue, Cambridge, MA 02139 (USA)
Prof. F. Stellacci
Institute of Materials
ꢀcole Polytechnique Fꢁdꢁrale de Lausanne
Lausanne, 1015 (Switzerland)
E-mail: francesco.stellacci@epfl.ch
Dr. S. Basak
Department of Chemical Engineering
Massachusetts Institute of Technology (USA)
Dr. R. Kumar
Advanced Materials and Green Chemistry Division
Tata Chemicals Limited, Innovation Centre (India)
[**] Financial support from Tata Chemicals Ltd. is gratefully acknowl-
edged. This work made use of the facilities at the Center for
Materials Science and Engineering, MIT and Department of
Chemistry Instrumentation Facility, MIT.
The esterification of carboxylic acids is a fundamental
organic transformation. The conversion in esterification
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201101821.
7900
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 7900 –7905

reactions is limited by slow reaction rates and by the
reversibility of the reactions. Acid catalysts are widely
employed in liquid-phase esterification to enhance the
reaction rate. The use of homogeneous catalysts such as
sulfuric acid and p-toluenesulfonic acid suffers from the
problems of side reactions, corrosion of the equipment,
environmentally hazardous acid waste, and difficulty in
separation of the catalyst from the reaction mixture.[18] Use
of MPSA, or other equivalent sulfonic-acid-terminated mol-
ecules as homogeneous catalysts will also pose similar
problems. However, sulfonic acid w-functionalized gold NPs
(both the striped and homoligand particles)[15] are noncorro-
sive, and can be easily recovered by centrifugation or
filtration and reused. We note that the Au NPs tested here
are insoluble in the reaction mixture before and after the
reaction, and can be separated easily by centrifugation. The
recycled catalyst also has similar activity compared to the
parent one (results not shown). Therefore, the sulfonic-acid-
protected Au NPs act truly as heterogeneous catalysts. We
chose acetylation of benzyl alcohol (BzOH) as a test reaction,
because the product benzyl acetate (BzAc) is a widely used
plasticizer, solvent, odorant, flavor chemical, and also a
precursor for a range of pharmaceuticals, agrochemicals, and
other fine chemicals.[19] Currently, zeolites with a large
number of acidic sites such as Faujasite (USY) are used to
produce BzAc through a heterogeneous catalytic process.[20]
Here, we show that sulfonate Au NPs—mainly when coated
with binary mixtures of ligands—relative to the zeolite
catalyze the reaction with a faster reaction rate, and a desired
product selectivity close to 100% (as compared to around
80% for the zeolite catalyst).
Figure 1. TEM images of Au NPs functionalized with a mixture of
MPSA and OT at 2:1 molar ratio (top left), and pure MPSA (top right),
with their corresponding size distribution.
The average size distribution for the NPs was calculated
from TEM images (Figure 1, summary in Table 1). The
chosen synthetic procedure delivers NPs (5–6 nm) of similar
core sizes with a relatively low standard deviation (around
20%). The particle sizes are within the size limits for which
stripelike domains have been observed.[16] Figure 2 shows
representative Au 4f and S 2p core-level spectra for an
MPSA–Au nanocomposite sample. The S 2p spectrum has
two components at binding energies of 162.6 and 168 eV,
which can be attributed to two different oxidation states of
sulfur (À2 and + 5, assigned to the thiolate and sulfonic acid
sulfur groups, respectively) present in the sample. The
quantification of the S 2p and O 1s spectra of all the samples
are summarized in Table 1. The observed weight ratio of S
We chose the esterification of carboxylic acids through
acid catalysis because its mechanism excludes cooperative
interactions between the acid groups. Scrimin and co-workers
have shown that NPs can be ideal for catalysis when
cooperative effects occur between catalysts constrained on
the surface of nanoparticles.[21] Here, we study morphology
effects that go beyond the clustering of catalytic ligands to
complement the studies by Scrimin and co-workers.
Table 1: Compositional and catalytic-activity analysis for Au NPs.
Ligand 1 Ligand 2 Ratio c(ligand 1) c(ligand 2) Particle size[a] S:O
O 1s[b] S 2p[b] S:O
SO3H[c]
Conv.[d] Sel. [%] TON[e]
[%]
[mmol]
[mmol]
[nm]
(theo) [wt%] [wt%] (obs) [mmolgÀ1
]
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
MPSA
None
OT
OT
OT
OT
OT
BT
HT
DT
–
2.00
1.67
1.33
1.00
0.67
0.33
1.33
1.33
1.33
1.33
1.33
1.33
1.33
0.00
0.33
0.67
1.00
1.33
1.67
0.67
0.67
0.67
0.67
0.67
0.67
0.67
5.76Æ1.48
5.77Æ1.57
5.88Æ1.52
5.44Æ1.58
5.55Æ1.49
5.39Æ1.45
5.65Æ1.47
5.42Æ1.55
5.35Æ1.38
5.28Æ1.36
5.76Æ1.46
5.82Æ1.39
5.64Æ1.51
0.67
0.80
1.00
1.33
2.00
4.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
7.64
5.97
5.03
3.53
3.31
2.11
3.24
3.06
5.30
5.28
6.07
5.75
5.19
3.71
3.60
3.42
3.45
3.56
4.71
2.12
1.91
3.56
3.41
4.46
3.81
3.63
0.48
0.60
0.68
0.98
1.07
2.22
0.65
0.62
0.67
0.64
0.73
0.66
0.69
0.58
0.56
0.53
0.26
0.25
0.28
0.53
0.50
0.55
0.53
0.67
0.59
0.56
80
75
85
11
10
8
32
35
81
83
86
87
87
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
>99
692
671
802
212
198
143
301
349
734
784
638
737
775
5:1
2:1
1:1
1:2
1:5
2:1
2:1
2:1
2:1
DDT
6-Ph-HT 2:1
8-Ph-OT 2:1
10-Ph-DT 2:1
[a] From analysis of TEM images. [b] From quantification of XPS core-level spectra; for quantification of the S 2p spectra the total amounts were
considered. [c] From integration of the S5+ component of the S 2p spectra. [d] Reaction conditions: the ratio of benzyl alcohol to acetic acid was 1:1;
2 wt% catalyst; temperature 808C; reaction time 6 h. [e] The turnover number (TON) is given as (conversion of BzOHꢂamount of BzOH in mol)/
(amount of -SO3H groups in molꢂ100). The TONs are average values from three experiments with a maximum standard deviation of Æ10.
Angew. Chem. Int. Ed. 2011, 50, 7900 –7905
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7901

Communications
compositions of 5:1 and the homoligand, all other composi-
tions seem to be poorly active (but still better than the
zeolite). These results indicate that the high density of
sulfonate groups has a great effect on the catalytic activity
but seems to rule out (together with the kinetics data) a
simple cooperative effect. At the 2:1 composition at which we
expect to see stripelike domains[7,15] we observe the highest
activity; all other compositions seem to suggest a dilution
effect. If we just compare the TON of the MPSA–OT (ratio of
2:1) nanoparticles to the MPSA homoligand NPs we observe
an increase of around 16%. To make sure that this effect was
predominantly because of the composition of the ligand shell
and structure we performed a similar study on a series of Pt
NPs with a similar ligand-shell composition. Interestingly,
these series of MPSA–OT-coated Pt NPs revealed a similar
trend (see Figures S2, S3, and S4 in the Supporting Informa-
tion) in the TON with a ligand composition of 2:1 being the
most catalytically active. We believe that this result is a strong
indication on the role of the ligand shell in the catalytic
activity of these particles. We do not have enough data to
determine the role of the core material in this catalytic
reaction and we believe that its study is beyond the scope of
this work. A simple comparison with the Pt particles shows
that these particles are more active relative to their Au
counterparts. Hence, the core has an effect, yet the trend in
the composition of the ligand shell remains mostly identical.
Taking into account previous results we suggest that the Au
cores have no effect whereas Pt has some effect.[22]
The conversion of benzyl alcohol with increasing time for
1) the homoligand MPSA–Au NPs, 2) MPSA/alkanethiol–Au
NPs, 3) MPSA/phenyl-terminated alkanethiol–Au NPs, and
4) the commercial zeolite USY catalyst is shown in Figure 3.
The error bars in Figure 3 and in all the subsequent Figures
show the standard deviation from the average value of three
experiments. The maximum conversion for the nanoparticle
catalysts is around 86% at 808C, whereas for the USY
catalyst it is around 51%. Because the reaction is reversible, it
is not possible to obtain a conversion of 100% in any case. For
the zeolite catalyst, the selectivity toward benzyl acetate was
always around 75–80% during the entire course of the
reaction. However, for the NP catalysts, the selectivity toward
benzyl acetate was always > 99% for all the experiments.
Therefore, the selectivity was not shown in Figure 3 or in any
of the Figures hereafter.
Figure 2. Representative X-ray photoelectron spectra of Au 4f (left
side), and S 2p for MPSA–Au NPs (right side). The core-level binding
energies (BE) were aligned with respect to the Au 4f7/2 BE of 84 eV.
The S 2p spectrum has two components attributed to S5+ and S2À
oxidation states, respectively.
and O is lower than the theoretical value, but follows the same
increasing trend with decrease in the initial molar ratio of
MPSA in the synthetic mixture. The linear correlation
between the theoretical and observed S:O ratios is shown in
Figure S1 (see the Supporting Information). For different
alkanethiols at a fixed molar ratio of MPSA (2:1), the
observed S:O weight ratio remains almost constant (around
0.65–0.7). The quantification of the S5+ component from the
S 2p spectra is necessary to estimate the concentration of
catalytically active sites, which in turn was used to calculate
the turnover numbers (TONs) of the acetylation reaction
given in Table 1. Quantification from XPS might not be fully
accurate, because it only shows the surface composition.
Table 1 also summarizes the catalytic activity of all the Au
nanoparticle catalysts in terms of conversion of benzyl
alcohol, selectivity toward benzyl acetate, and TONs of the
acetylation reaction. The respective values for a commercial
zeolite USY catalyst are a conversion rate of 51%, selectivity
of 76%, and TONs of 84. It should be noted that the
conversions represented here are the absolute values
observed during the reaction with a fixed amount of NP
catalyst (2 wt%).
The TON values represent an authentic comparison of the
catalytic efficiency, since the values are normalized with
concentration of sulfonic acid groups present on the catalyst
surface. All the experiments were performed thrice to check
the reproducibility and to estimate the error bars. The
MPSA–OT(2:1)–Au, MPSA–DT(2:1)–Au, and MPSA–DDT-
(2:1)–Au NPs have much better activity (TON) than MPSA–
BT(2:1)–Au and MPSA–HT(2:1)–Au NPs. At present it is
hard to draw definitive conclusions on the true reasons for
this behavior but we can speculate that a minimum “groove”
depth is required to create a favorable effect for the catalysis.
This is somewhat also supported by the fact that the MPSA–
HT nanoparticles have lower efficiency when compared to
the MPSA–6-Ph-HT particles; such a difference disappears
when MPSA–OT to MPSA–8-Ph-OT particles are compared.
Additionally, as the chain length difference decreases so does
the driving force towards the formation of highly defined
stripelike domains.[11] Yet, it is not immediately clear how this
would affect the catalytic efficiency. In general our complete
study on the MPSA–OT system indicates a complex effect of
ligand composition and morphology. The highest TON is
observed for the 2:1 ratio, followed by the MPSA-rich
The MPSA–8-Ph-OT(2:1)–Au NPs have better efficiency
in the initial phases of the reaction than the other two
nanoparticle catalysts, probably because of the favorable
wetting with the reactants. Again, the MPSA–OT(2:1)–Au
NPs have slightly better efficiency than the MPSA–Au NPs.
The better efficiency of the MPSA/phenyl-terminated alka-
nethiol—Au NPs at the initial phase of the reaction prompted
us to conduct kinetic experiments at a 1:1 molar ratio of
benzyl alcohol to acetic acid.
The acetylation of benzyl alcohol is expected to follow
second-order kinetics, the reaction is of first order with
respect to the concentrations of both benzyl alcohol and
acetic acid. For such a reaction, when the initial concentra-
tions of the two reactants are equal, a plot of 1/[BzOH]t versus
the reaction time (t) will be linear at conversions smaller than
7902 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 7900 –7905

Figure 5. Plot of the reaction rate versus time at different temperatures
for MPSA–OT(2:1)–Au nanoparticle as catalyst (reaction conditions:
the ratio of benzyl alcohol to acetic acid was 1:1, 2 wt% catalyst). The
rate constants were determined from the slope of the respective linear
curves.
Figure 3. Conversion (solid circles) of benzyl alcohol and selectivity
(empty squares) toward benzyl acetate with increasing time (reaction
conditions: the ratio of benzyl alcohol to acetic acid was 1:1; 2 wt%
Au catalysts, 15 wt% USY; temperature 808C).
50%, where [BzOH]t represents the concentration of BzOH
at a given time “t”. Figure 4 displays the plot of 1/[BzOH]t
versus “t” at 808C for three representative catalysts: 1) the
homoligand MPSA–Au NPs, 2) the MPSA/alkanethiol–Au
NPs, and 3) the MPSA/phenyl-terminated alkanethiol–Au
NPs. The reaction rate increases linearly with time. The
MPSA–8-Ph-OT(2:1)–Au NPs have a much faster reaction
rate than the other two catalysts, which corroborates our
previous observation (Figure 3).
Figure 5 shows the plot of 1/[BzOH]t versus “t” for the
MPSA–OT(2:1)–Au NPs at different temperatures ranging
from 60–1008C. The reaction rate was low at temperatures of
60–708C, and at temperatures higher than 808C the reaction
rate is considerably enhanced. From the slope of the curves,
the rate constants (k) at different temperatures were eval-
uated. Figure 6 represents the Arrhenius plot of lnk versus
1/T for the three catalysts mentioned in Figure 4. Interest-
ingly, the plot is linear for the first three data points, and also
for the last three data points. That is, the curve has two
different gradients with increasing temperature. There are
two possible explanations of this nonlinearity of the Arrhe-
nius plot. The first one could be that there are two different
mechanisms operating at different temperatures. However,
we have observed that the difference between rate constants
at 60 and 708C, at 80 and 908C, and also at 90 and 1008C are
similar (Figure 5). That is, the reaction rate shows a similar
temperature dependence at low and high temperature
regimes. Therefore, we believe that two different mechanisms
are unlikely. Esterification reactions in the liquid phase follow
a standard acid-catalyzed mechanism (see Scheme S1 in the
Supporting Information). The alternative explanation is that
with increasing temperature, and specifically after a certain
Figure 4. Plot of the reaction rate versus time for MPSA–Au, MPSA–
OT(2:1)–Au and MPSA–8-Ph-OT(2:1)–Au nanoparticles as catalysts
(reaction conditions: the ratio of benzyl alcohol to acetic acid was 1:1;
2 wt% catalyst; temperature 808C).
Figure 6. Arrhenius plot for MPSA–Au, MPSA–OT(2:1)–Au and
MPSA–8-Ph-OT(2:1)–Au nanoparticles as catalysts.
Angew. Chem. Int. Ed. 2011, 50, 7900 –7905
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7903

Communications
1 h. The reaction was quenched by dropwise addition of 20 mL of
water, and the solution was then extracted with DCM. The organic
layer was dried over anhydrous MgSO4, and DCM was removed
under vacuum at room temperature. The light yellow-colored oil was
dissolved in a minimum volume of DCM, purified by chromatography
on a silica column, and eluted with hexane. The solvents were
removed under vacuum to yield light yellow-colored oils: 1-bromo-6-
phenylhexane 2.447 g (70%), 1-bromo-8-phenyloctane 2.913 g
(75%), and 1-bromo-10-phenyldecane 3.112 g (72%). 1H NMR
critical temperature is reached (higher than 708C in the
present case), there is an enhanced diffusion of the reactants
into the ligand shell of the particles. This diffusion factor
contributes to the temperature dependence of the pre-
exponential factor of the Arrhenius equation. The energy of
activation at these two different temperature regimes is
summarized for the three catalysts in Table 2.
À
À
À
À
À
À
(CDCl3): d = 1.3 (q, CH2 ), 1.6 (q, CH2 ), 1.8 (q, CH2 ), 2.6
Table 2: Energy of activation for the catalytic Au nanoparticles.
À
À
À
(m, CH2 ), 3.5 (t, CH2Br), 7.3 (d, aromatic), 7.4 ppm (d, aromatic).
In a second step phenyl alkane thiols are synthesized from the
respective bromophenyl alkanes. In a 100 mL round-bottomed flask
Catalyst
Ea1[a] [kJmolÀ1
]
Ea2[b] [kJmolÀ1
]
MPSA–8-Ph-OT(2:1)–Au
MPSA–OT(2:1)–Au
MPSA–Au
71
54
76
31
22
17
equipped with
a dropping funnel, 6.9 mmol of the respective
bromophenyl alkane in 14 mL of dry THF was placed. The flask
was shielded from light (to avoid photoinduced side reactions), and
cooled to À208C using dry ice and acetone. Then a solution of 1.48 g
of HMDT (8.3 mmol) and 1.98 g of TBAF (7.6 mmol) in 10 mL of
THF was added to the stirred solution. The resulting mixture was
allowed to warm to room temperature and stirring was continued for
12 h. After the disappearance of the bromide (monitored by TLC),
the reaction mixture was concentrated to remove THF, diluted with
DCM, and washed with a saturated aqueous solution of NH4Cl. The
organic layer was dried over anhydrous MgSO4, and the solvent was
removed under vacuum, yielding light yellow oil. The oil was
dissolved in a minimum volume of DCM, purified with a neutral
alumina column, and eluted with ethylacetate. The solvents were
removed under vacuum to yield colorless oils: 6-phenyl-hexanethiol
[a] At temperatures of 60–808C. [b] At temperatures of 80–1008C.
In conclusion, the acetylation reaction of benzyl alcohol
with acetic acid has been performed in the prescence of
nanoparticle catalysts functionalized with a mixture of MPSA
and alkanethiols. The activity was compared with that of a
conventional solid-acid zeolite USY catalyst. Formation of
benzyl acetate was observed for the nanoparticle catalysts
with a selectivity higher than 99%, whereas dibenzyl ether
was formed as a minor product with a selectivity of around
20% by using the zeolite catalyst USY. We find that the
homoligand MPSA NPs already outperform the commercial
zeolite catalyst. Additionally, when we study the effect of
dilution of the MPSA with OTwe find the highest activity for
a composition of 2:1. This finding excludes cooperative effects
and indicates the role of the morphology of the ligand shell.[15]
This work establishes sulfonic-acid-coated Au NPs as a new
generation of solid-acid catalysts and provides the basis for
studying the mixed-ligand effect of these catalysts.
1.225 g (91%), 8-phenyl-octanethiol 1.379 g (89%), and 10-phenyl-
decanethiol 1.401 g (81%). H NMR (CDCl3): d=1.3–1.4 (q, CH2 ),
1
À
À
À
À
À
À
À
1.5 (t, SH), 1.6 (q, CH2 ), 2.5–2.7 (m, CH2 ), 7.3 (d, aromatic),
7.4 ppm (d, aromatic).
À
To detect the presence of SH groups in the synthesized thiol
ligands, a solution of NaOH infused ethanol and sodium nitroprusside
was added to the solution of the ligand. The resulting color change to
purple indicated the presence of thiol groups.
Synthesis of composite nanoparticles: In a typical synthesis,
0.496 g of AuPPh3Cl (1 mmol) was dissolved in 50 mL of DMF. Then
to a solution of requisite amounts of MPSA the alkane thiol ligand
(Table 1) in 50 mL of DMF was added under stirring. The mixture was
allowed to equilibriate for 10 min at 508C. To the mixture a solution
of 0.87 g of the tert-BuNH2·BH3 complex (10 mmol) in 60 mL of DMF
was added dropwise. The stirring was continued at 508C for 1 h. After
the reaction mixture was cooled to room temperature, it was kept in a
freezer to precipitate the NPs overnight. The NPs were centrifuged at
3400 rpm for 5 min and washed with ethanol twice and dried under
vacuum.
Experimental Section
Materials: Chloro(triphenylphosphine)gold(I) (AuPPh3Cl), borane-
tert-butylamine complex (tBuNH2.BH3), 3-mercapto-1-propanesul-
fonic acid sodium salt (MPSA), 1-butanethiol (BT), 1-hexanethiol
(HT), 1-octanethiol (OT), 1-decanethiol (DT), 1-dodecanethiol
(DDT), 1,6-dibromohexane, 1,8-dibromooctane, 1,10-dibromode-
cane, phenyl lithium (PhLi), hexamethyldisilathane (HMDT), tetra-
butylammonium fluoride (TBAF), benzyl alcohol (BzOH), N,N-
dimethylformamide (DMF), dichloromethane (DCM), tetrahydro-
furan (THF), and ethanol (EtOH) were purchased from Sigma–
Aldrich. Concentrated H2SO4 and MgSO4 were purchased from EMD
Chemicals, acetic acid (AcOH), ethylacetate, hexane, and NH4Cl
were purchased from Mallinckrodt Chemicals. All chemicals were
reagent grade and were used as received. The protonated form of
ultrastable Y (USY) zeolite (surface area 780 m2 gÀ1) was obtained
from Zeolyst International. The zeolite was activated at 1508C before
use.
The terminal SO3À head groups were acidified by treating 100 mg
of the NPs with 5 mL of 1m H2SO4 for 4 h at room temperature. In the
absence of this step no catalytic reaction was observed. Then 10 mL of
ethanol was added to it, and kept in the freezer to precipitate the NPs
overnight. The NPs were centrifuged at 3400 rpm for 5 min and
washed with ethanol twice and dried under vacuum.
Catalytic testing of the composite nanoparticles: The acetylation
reaction was carried out in a 25 mL two-neck round-bottomed flask
fitted with a reflux condenser and containing a tiny magnetic stir bar.
In a typical experiment, 0.54 g of benzyl alcohol (5 mmol), 0.3 g of
acetic acid (5 mmol), and 10 mg of the nanoparticle catalyst were
taken directly into the reaction vessel (for the relative amount of
active sites refer to Table 1) and sonicated for 2 min to disperse the
nanoparticles uniformly in the reaction medium. The reaction vessel
was immersed in an oil bath and put on a IKAMAG RCT basic hot
plate with a magnetic stirrer. The temperature of the oil bath was
controlled by an IKA ETS-D5 electronic contact thermometer
connected to the hot plate. The reaction mixture was stirred at
808C and at 600 rpm for 6 h. For the experiments with the zeolite
USY catalyst, 75 mg of the catalyst was used.
Synthesis of phenyl-terminated alkanethiol ligands: In a first step
bromophenyl alkanes are synthesized from the respective dibromo
alkane precursor molecules. In a 100 mL round-bottomed flask
equipped with a dropping funnel, 1.9 mL of a 1.14m solution of PhLi
(2.2 mmol) in hexane/diethylether was placed under N2 atmosphere.
The flask was cooled to À148C using a freezing mixture, and
14.5 mmol of the respective dibromo alkane in 10 mL of dry THF was
added to the solution at once under stirring. The solution was allowed
to stir for 6 h at À148C, and was then stirred at room temperature for
7904 www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 7900 –7905

However, experiments were also performed by varying the
reaction temperature (60–1008C), and the reaction time (0–6 h) to
obtain various kinetic parameters. The reaction mixture was contin-
uously stirred during the reaction at a constant temperature. After the
desired reaction time, the catalyst was separated from the reaction
mixture by centrifugation at 14000 rpm for 5 min and washed with
ethanol.
For the kinetic experiments, 25 mL of the reaction mixtures was
sampled every 15 min and the samples were kept in a freezer
overnight to precipitate the NPs. After centrifugation, the reaction
mixtures were diluted with 1 mL ethanol and analyzed by gas
chromatography (GC). The main product of the reaction was found to
be benzyl acetate. The formation of dibenzyl ether was also observed
when USY was used as catalyst.
[3] M. Bartz, J. Kuether, R. Seshadri, W. Tremel, Angew. Chem.
1998, 110, 2646; Angew. Chem. Int. Ed. 1998, 37, 2466.
[4] C. Briggs, T. B. Norsten, V. M. Rotello, Chem. Commun. 2002,
1890.
[5] L. Pasquato, P. Pengo, P. Scrimin, Supramol. Chem. 2005, 17, 163.
[6] F. Manea, F. B. Houillon, L. Pasquato, P. Scrimin, Angew. Chem.
2004, 116, 6291; Angew. Chem. Int. Ed. 2004, 43, 6165.
[7] A. M. Jackson, J. W. Myerson, F. Stellacci, Nat. Mater. 2004, 3,
330.
[8] A. M. Jackson, P. Jacob Silva, Y. Hu, F. Stellacci, J. Am. Chem.
Soc. 2006, 128, 11135.
[9] G. A. DeVries, M. Brunnbauer, Y. Hu, A. M. Jackson, B. Long,
B. T. Neltner, O. Uzun, B. H. Wunsch, F. Stellacci, Science 2007,
315, 358.
Analytic measurements: TEM experiments were carried out on a
JEOL 200CX instrument operated at 200 kV. The core sizes of the
NPs were determined using TEM images. The diameter of around
1000 NPs per sample was estimated and averaged using ImageJ
software.
[10] A. Centrone, Y. Hu, A. M. Jackson, G. Zerbi, F. Stellacci, Small
2007, 3, 814.
[11] C. Singh, P. K. Ghorai, M. A. Horsch, A. M. Jackson, R. G.
Larson, F. Stellacci, S. C. Glotzer, Phys. Rev. Lett. 2007, 99,
226106.
XPS experiments were performed on a Kratos Axis Ultra X-ray
photoelectron spectrometer employing a monochromatic Al Ka
source (1486.7 eV) and an electron takeoff angle of 908 relative to
the sample plane. The survey scan was conducted at a pass energy of
80 eV for binding energies (BEs) ranging from 0–1100 eV. The high-
resolution scans of O 1s, C 1s, S 2p, and Au 4f were conducted at a
pass energy of 10 eV. The core-level spectra were corrected for the
background using the Shirley algorithm and the core-level BE was
aligned with respect to the Au 4f7/2 BE of 84 eV.
[12] C. Singh, A. M. Jackson, F. Stellacci, S. C. Glotzer, J. Am. Chem.
Soc. 2009, 131, 16377.
[13] A. Centrone, E. Penzo, M. Sharma, J. W. Myerson, A. M.
Jackson, N. Marzari, F. Stellacci, Proc. Natl. Acad. Sci. USA
2008, 105, 9886.
[14] J. J. Kuna, K. Voitchovsky, C. Singh, H. Jiang, S. Mwenifumbo,
P. K. Ghorai, M. M. Stevens, S. C. Glotzer, F. Stellacci, Nat.
Mater. 2009, 8, 837.
[15] O. Uzun, Y. Hu, A. Verma, S. Chen, A. Centrone, F. Stellacci,
Chem. Commun. 2008, 196.
[16] R. P. Carney, G. A. DeVries, C. Dubois, H. Kim, J. Y. Kim, C.
Singh, P. K. Ghorai, J. B. Tracy, R. L. Stiles, R. W. Murray, S. C.
Glotzer, F. Stellacci, J. Am. Chem. Soc. 2008, 130, 798.
[17] N. Zheng, J. Fan, G. D. Stucky, J. Am. Chem. Soc. 2006, 128, 6550.
[18] W. T. Liu, C. S. Tan, Ind. Eng. Chem. Res. 2001, 40, 3281.
[19] A. Zaidi, J. L. Gainer, G. Carta, Biotechnol. Bioeng. 1995, 48,
601.
The reaction products were analyzed by a gas chromatograph
from Agilent containing a polyethylene glycol capillary column and a
flame ionization detector (FID).
Received: March 14, 2011
Revised: May 31, 2011
Published online: July 8, 2011
[20] S. R. Kirumakki, N. Nagaraju, S. Narayanan, Appl. Catal. A
2004, 273, 1.
[21] G. Zaupa, C. Mora, R. Bonomi, L. J. Prins, P. Scrimin, Chem.
Eur. J. 2011, 17, 4879.
[22] S. Kato, K. Nakagawa, N. Ikenaga, T. Suzuki, Catal. Lett. 2001,
73, 175.
Keywords: catalysts · gold · heterogeneous catalysis ·
nanoparticles
.
[1] M. C. Daniel, D. Astruc, Chem. Rev. 2004, 104, 293.
[2] A. P. Alivisatos, Science 1996, 271, 933.
Angew. Chem. Int. Ed. 2011, 50, 7900 –7905
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7905