Amine-functionalized siloxane oligomer facilitated synthesis of subnanometer colloidal Au particles

Article abstract of DOI:10.1039/c4ta05954a

Amine-functionalized siloxane oligomers were synthesized and used successfully to prepare colloidal Au particles smaller than 1 nm. Using NMR to follow the interaction of Au(THT)Cl with the functionalized siloxane, it was determined that the amine ligands displaced the THT ligand effectively. By comparison with other functionalized siloxane oligomers/compound, parameters such as density of ligating groups, oligomer steric barrier and reduction rates were found to be essential for the formation and stability of subnanometer Au particles. Without further treatment, the formed Au particles were active catalysts for the reduction of p-nitrobenzaldehyde by triethylsilane, forming an imine as the major coupling product. Deposition of the Au colloids onto silica, followed by thermal treatment to remove the organic groups resulted in subnanometer Au on silica, indicating this to be a promising method of fabricating subnanometer supported Au catalyst. This journal is

Full text of DOI:10.1039/c4ta05954a

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
Amine-functionalized siloxane oligomer facilitated
synthesis of subnanometer colloidal Au particles†
Cite this: J. Mater. Chem. A, 2015, 3,
Zhen Wang,ab Evgeny V. Beletskiy,a Sungsik Lee,c Xianliang Hou,ab Yuyang Wu,d
1743
b
a
a
*
*
Tiehu Li, Mayfair C. Kung and Harold H. Kung
Amine-functionalized siloxane oligomers were synthesized and used successfully to prepare colloidal Au
particles smaller than 1 nm. Using NMR to follow the interaction of Au(THT)Cl with the functionalized
siloxane, it was determined that the amine ligands displaced the THT ligand effectively. By comparison
with other functionalized siloxane oligomers/compound, parameters such as density of ligating groups,
oligomer steric barrier and reduction rates were found to be essential for the formation and stability of
subnanometer Au particles. Without further treatment, the formed Au particles were active catalysts for
the reduction of p-nitrobenzaldehyde by triethylsilane, forming an imine as the major coupling product.
Deposition of the Au colloids onto silica, followed by thermal treatment to remove the organic groups
resulted in subnanometer Au on silica, indicating this to be a promising method of fabricating
subnanometer supported Au catalyst.
Received 4th November 2014
Accepted 25th November 2014
DOI: 10.1039/c4ta05954a
www.rsc.org/MaterialsA
reduced in the presence of a stabilizer. The choice of stabilizer
is essential, since large Au particles would form if the stabilizer
A Introduction
interacts too weakly with Au, whereas stabilizers that interact
too strongly could alter the properties of the Au particles
signicantly. The propensity of Au to adsorb sulfur compounds
makes thiols popular stabilizers for the preparation of small Au
colloidal particles, and clusters with 1–4 nm average diameter
can be prepared regularly.19–23 In addition to thiols and phos-
phines,24,25 bifunctional stabilizers have been successfully
employed also, and these include hydroxy mercapto glycol,26
hydroxy thiolate,27 tetraalkylammonium bromide,28 mercapto-
succinic acid,29 phosphonated mercapto compounds,30 amino-
phosphonate,31 and ethylenediamine tetraacetate.32 In these
bifunctional stabilizers, one functional group ligates the Au
particle, and the other renders a charged layer around the
particle, thereby stabilizing the particles from agglomeration by
charge–charge repulsion. Recently, preparations of atomically
monodisperse thiolate-stabilized clusters have been reported
for Au11, Au25, Au38, Au102, Au144, etc.3,33–36 A rapid small-scale
synthetic method that produces thiolate-coated gold nano-
clusters (d < 2 nm) in 2 min has also been reported.37
While effective in stabilizing nanosize Au particles, the
strong chemical interaction of sulfur or phosphorus with Au
could affect the chemical properties of the colloidal Au, espe-
cially those that depend on surface Au atoms. Removal of the
stabilizer typically entails rst the deposition of Au clusters on a
support followed by oxidative thermal treatments that are oen
harsh enough to result in signicant coarsening of the Au
particles. Residual S or P, in the form of stable sulfate and
phosphate, could poison the Au particles when used as cata-
lysts. Thus, there is a strong interest to identify alternate
Colloidal gold particles with diameters smaller than around 2
nm exhibit many interesting properties. When the size of a Au
cluster is comparable to the Fermi wavelength, the plasmon
resonance greatly diminishes, while size-dependent uores-
cence and molecule-like HOMO–LUMO transition emerges.1–7
Strong size-dependent catalytic properties have been observed
in this size range.8–12 For example, large Au particles are
signicantly less active in catalytic oxidation than small parti-
cles11,13 and in water–gas shi than atomic Au (or Au cations),14
and there is a change in selectivity from hydrogenation to
epoxidation in the reaction of propene in a mixture of H2 and O2
in the range of 1 to 3 nm.9 Small Au particles have also found
applications in DNA or drug delivery, as the efficiency in cell
membrane penetration increases with decreasing Au particle
size.15–17 In view of these, there have been intensive effort to
develop methods to synthesize small Au particles,18 particularly
those 2 nm and smaller.
Among the methods explored, colloidal preparation is
attractive because it is easily scalable for commercial applica-
tions. In colloidal preparation, a Au salt precursor in solution is
aChemical and Biological Engineering Department, Northwestern University, Evanston,
IL 60208, USA. E-mail: m-kung@northwestern.edu; hkung@northwestern.edu
bSchool of Materials Science and Engineering, Northwestern Polytechnical University,
Xi'an, Shaanxi 710072, P. R. China
cX-ray Science Division, Argonne National Laboratory, Argonne, IL 60439, USA
dChemistry Department, Northwestern University, Evanston, IL 60208, USA
† Electronic supplementary information (ESI) available: NMR, TEM, DLS, XANES,
SAX and absorption wavelengths of Au clusters calculation based on jellium
model. See DOI: 10.1039/c4ta05954a
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 1743–1751 | 1743

View Article Online
Journal of Materials Chemistry A
Paper
stabilizers that interact more weakly with Au and can be
removed by oxidation more easily, such as ligands with amino
or carboxylate groups, since the strength of interaction follows
the order (SH > P > N > O).
B Experimental
B.1 Materials
Poly(methylhydro)siloxanes (PMHS I) with trimethylsiloxy
termination (1700–3200 g molꢀ1 average molecular weight, 12–
45 cSt viscosity, average degree of oligomerization estimated by
NMR spectroscopy 38 siloxane units), palladium hydroxide on
carbon (20 wt% loading), tetrahydrofuran (anhydrous, 99.9%),
hexanes (95%), triethylsilane (99%), Au(III) chloride trihydrate
($99.9% trace metals basis) and Au(III) chloride ($99.99% trace
metals basis) were purchased from Sigma-Aldrich. Celite 545
(lter aid, powder) and sodium sulfate (anhydrous, $99.0%)
were from EMD Chemicals. Ethyl ether (anhydrous, $99.0%)
was from Mallinckrodt Chemicals. Calcium sulfate (CaSO4)
(Drierite, without indicator, 20–40 mesh) was from Acros
Organics. N-Methyl-aza-2,2,4-trimethylsilacyclo-pentane was
purchased from Gelest. All solvents (anhydrous), reagents and
catalysts were used without further purication.
We have been exploring different approaches to synthesize
nanosize metal particles.38,39 In one approach that was facili-
tated by amine and titanyl modied oligomeric hydrosiloxanes,
Au nanoparticles between 1.75 and 2.25 nm and with very tight
particle size distributions were obtained.39 This method has an
advantage in that oxidative removal of the organic groups
attached to the siloxane oligomer would result in the metal
particles dispersed in a mixed Ti–Si oxide. This is also a high
yield process because all the added metal could be incorporated
into the resulting colloid, a fact that could be important for
noble metals. Thus, it appears that this is a promising approach
that deserves further development and understanding for the
synthesis of subnanometer Au particles.
We report here the results of using a modied version of this
method by using only amine functionalized hydrosiloxane to
successfully prepare subnanometer Au particles with narrow
size distributions – a feat that has been known difficult to
accomplish without using a strong stabilizer. In following the
synthesis process with solution NMR and with use of model
compounds, we gained important insights into the parameters
that control Au particle sizes. We evaluated the requirement for
proximal placement of the amine ligand and the reductant and
established the importance of the steric hindrance offered by
the oligomer. We collected evidence of binding of the amine
ligand to the Au cation precursor and to the Au particles.
Finally, we demonstrated that these Au nanoparticles are cata-
lytically active.
B.2 Synthesis of siloxanes oilgomers/compounds
B.2.1 Oligomers based on PMHS: N–PMHS IV, N–SiH–PMHS
V, and Me–SiH–PMHS VI
N–PMHS IV. Synthesis of IV is shown in Scheme 1. 0.64 mL
PMHS I possessing 10 mmol of SiH was added to 10 mL THF in
a Schlenk ask. Excess H2O (17 mmol) was used to completely
oxidize SiH to silanols (SiOHs) in the presence of 30 mg Pearl-
man's catalyst (20 wt% Pd(OH)2/C) to form II.40 This reaction
was followed by monitoring the hydrogen gas evolved. At the
cessation of H2 evolution, the reaction mixture was dried thor-
oughly using Na2SO4. The Pd catalyst was removed using
column chromatography. The solution containing II was
Scheme 1 Synthesis of N–PMHS IV and N–SiH–PMHS V, and structures of VI, VII, and VIII.
1744 | J. Mater. Chem. A, 2015, 3, 1743–1751
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
layered under N2 ow onto a column composing of a layer of Fig. S4.† 1H NMR (400 MHz, CDCl3) d 0.12 (–Si(CH3)2CH2); 0.19
CaSO4 sandwiched between layers of Celite using diethyl ether (–SiH(CH3)2); 0.44 (–SiCH2); 0.71 (–SiCH2); 0.90 (CH3CH–); 1.77
as the eluent.39 The puried II was kept as a dilute solution to (CH3CH–); overlapped 2.33 (–NHCH2, –NHCH3); 4.84 (Si–H);
minimize self-condensation until use. The desired oligomer IV 7.32–7.39 and 7.56–7.58 (–C6H5); 7.26 (CDCl3). 13C NMR (100
was formed by reacting the SiOH groups of II quantitatively with MHz, CDCl3) d 0.73 (–Si(CH3)2CH2); 1.34 (–SiH(CH3)2); 20.97
N-methyl-aza-2,2,4-trimethylsilacyclopentane in a Schlenk ask. (CH3CH–); 24.40 (–SiCH2); 28.93 (CH3CH–); 36.47 (–NHCH3);
The mixture was stirred for 20 min at room temperature fol- 61.52 (–NHCH2); 127.6, 129.76, 134.10, 136.09 (–C6H5); 77.00
lowed by solvent removal at 50 mtorr pressure to generate IV, (CDCl3). 29Si NMR (80 MHz, CDCl3) d ꢀ46.03 (–Si(C6H5)2); –4.76
1
which was a viscous clear liquid. The H, 13C, 29Si and HSQC (–SiH(CH3)2); 10.08 (–Si(CH3)2CH2).
NMR spectra of N–PMHS IV are shown in Fig. S1;† 1H NMR (400
B.2.3 Synthesis of amine functionalized Au colloid. Chlor-
MHz, CDCl3) d 0.1–0.2 ((–Si(CH3)3, –Si(CH3)2, –SiCH3); 0.38 o(tetrahydrothiophene)gold(I) (Au(THT)Cl) was synthesized
(–SiCH2); 0.63 (–SiCH2); 0.80 (NH); 0.89 (CH3CH–); 1.74 (CH3CH–); similar to the work of Uson et al.42 Aliquots of 6 mM Au(THT)Cl
overlapped 2.35 (–NHCH2, –NHCH3); 7.26 (CDCl3). 13C NMR solution in THF in a N2 lled Schlenk ask was added to ice-
(100 MHz, CDCl3) d ꢀ2.65 (–Si(CH3)3); 1.07 (–SiCH3); 1.34 cooled vials containing either oligomer IV, V, or VI, or
(–Si(CH3)2); 20.90 (CH3CH–)); 24.49 (–SiCH2); 29.02 (CH3CH–); compound VIII in THF solution (Fig. 1). The amount of olig-
36.53 (–NHCH3); 61.62 (–NHCH2); 77.00 (CDCl3). 29Si NMR (80 omer used was calculated to result in a functional group (amine
MHz, CDCl3)
(–Si(CH3)2).
d
ꢀ67.90 (–SiCH3); 6.99 (–Si(CH3)3); 8.49 or –Me2BuSi) to Au molar ratio of 10. When SiH reductant was
not an integral part of the oligomer, triethylsilane diluted in
N–SiH–PMHS V and Me–SiH–PMHS VI. Using the same THF was added as the reducing agent 1 h aer mixing. Two
procedure to form II, but with the added H2O limited to 5 mmol, different ratios 3 : 1 and 10 : 1 of SiH : Au were examined. Then
III was formed with only 50% of SiH oxidized to the corre- THF was added to result in a total solution volume of 10 mL and
sponding SiOH groups. Aer the SiOH groups in III was reacted a Au concentration of 3 mM. The mixture was stirred in an ice
quantitatively
with
N-methyl-aza-2,2,4-trimethyl-silacyclo- bath for another hour. Au/SiO2 was prepared by depositing Au
pentane, N–SiH–PMHS V was formed. V was characterized with colloid solution in THF onto Cabosil-L90 and calcined in a
NMR and the spectra were shown in Fig. S2.† 1H NMR (400 stream of O3 at 150 ꢁC, similar to the work of Mashayekhi et al.39
MHz, CDCl3) d 0–0.013 (–Si(CH3)3, –Si(CH3)2, –SiHCH3, –SiCH3);
0.36 (–SiCH2); 0.61 (–SiCH2); 0.82 (NH); 0.9 (CH3CH–); 1.79
(CH3CH–); overlapped 2.42 (–NHCH2, –NHCH3); 4.7 (Si–H); 7.26
B.3 Characterization
(CDCl3).13C NMR (100 MHz, CDCl3) d ꢀ2.61 (–Si(CH3)3); 1.05
1H, 13C and 29Si NMR spectra were recorded with an Agilent
(–SiCH3); 1.33 (–Si(CH3)2); 20.92 (CH3CH–); 24.51 (–SiCH2);
1
DD2 spectrometer, operated at 400 MHz for H, 100 MHz for
29.04 (CH3CH–); 36.56 (–NHCH3); 61.64 (–NHCH2); 77.00
13C, and 79 MHz for 29Si. NMR spectra of the moisture sensitive
(CDCl3). 29Si NMR (80 MHz, CDCl3) d ꢀ66.82 (–SiCH3); –36.51
compound were recorded in sealed J. Young NMR tubes. The 1H
(–SiHCH3); 6.28 (–Si(CH3)3); 7.90 (–Si(CH3)2).
and 13C NMR chemical shis, reported relative to SiMe4, were
III was reacted for 2 h with butyl(chloro)dimethylsilane in
determined by secondary references to the residual 1H and 13
C
the presence of diethylamine as the chloride scavenger to form
Me–SiH–PMHS VI. Aer the reaction, the white solid diethyl-
amine hydrochloride was removed by ltration and VI was
puried by vacuum evaporation to 50 mtorr. The clear viscous
liquid VI was characterized using NMR (Fig. S3†). 1H NMR (400
MHz, CDCl3) d 0.12–0.37 (–Si(CH3)3, –Si(CH3)2, –SiHCH3,
–SiCH3); 0.48 (–SiCH2); 0.84 (CH3CH2–); 1.17 (–SiCH2CH2CH2);
1.60 (CH3CH2–); 5.01 (Si–H); 7.26 (CDCl3). 13C NMR (100 MHz,
CDCl3) d ꢀ3.17 (–Si(CH3)3); –0.44 (–Si(CH3)2); 0.5 (–SiCH3); 1.21
(–SiHCH3); 13.34 (CH3CH2–); 17.37 (–SiCH2); 24.95 (CH3CH2–);
25.95 (–SiCH2CH2CH2); 77.00 (CDCl3). 29Si NMR (80 MHz,
CDCl3) d ꢀ66.47 (–SiCH3); –34.85 (–SiHCH3); 7.23 (–Si(CH3)3);
8.66 (–Si(CH3)2).
peak of the deuterium solvent resonances. 29Si NMR chemical
shis were referenced to TMS at 0.0 ppm. The chemical shis
reported are in ppm and coupling constants in Hz. All the NMR
data were processed and analysed by MNova 9.0 soware.
B.2.2 Trisiloxanes: HSiMe2OSiPh2OSiMe2C4H8NHCH3 VIII.
Disiloxane HSiMe2OSiPh2OH VII was prepared following the
procedure described in ref. 41 by coupling diphenylsilane diol
with dimethylchlorsilane. VII (5 mmol) was dissolved in 10 mL
dry THF in a Schlenk ask in a N2 glove box. While stirring, 5
mmol N-methyl-aza-2,2,4-trimethylsilacyclopentane was added
to react with the SiOH groups of VII quantitatively. The mixture
was stirred for 20 min and the solvent was removed by evacu-
ation to 50 mtorr to yield a viscous colorless liquid of the
Fig. 1 Scheme for synthesis of amine-stabilized Au colloid. Filled black
circles denote Si atoms, open circles denote amine functionality, and
gold circles are Au atoms. For clarity, only a section of two function-
alized PMHS oligomers are shown (rearrange so it goes with the rest of
desired product VIII. NMR characterization are shown in the figure).
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 1743–1751 | 1745

View Article Online
Journal of Materials Chemistry A
Paper
High angle annular dark eld (HAADF) STEM images were
obtained using a JEOL 2100-F transmission electron microscope
with a eld emission gun at 200 kV, which has a spatial reso-
lution of 0.3 nm. Particle size analyses were performed using
the program Image J. A Brookhaven BI9000 high speed corre-
lator. Dynamic Light Scattering (DLS) instrument was used to
measure the hydrodynamic diameters of the oligomer-coated
Au nanoparticles. The incident light source was a 3 W Ar-ion
laser operating at 514 nm. The sample concentration was
1 mg mLꢀ1 and the cuvette volume was 300 mL. UV-vis
absorption spectra of the Au colloid solutions were acquired
using a HP (Agilent) 8452 diode array spectrophotometer. In a
typical procedure, a freshly prepared Au colloid sample was
diluted to 0.5 mM with cold THF and kept at ice temperature
Fig. 2 1H NMR of various sample after addition of Au(THT)Cl to
N–PMHS IV or SiH–N–PMHS V after 1 or 5 h. (Left) region of spectra
that showed resonances of unbound THT (marked by dotted line), NH
(blue ovals), and CH3 + CH2 at d ꢂ 2.3 to ꢂ2.4 ppm. (Right) expanded
and away from light until analysis. The procedure was the same view of the resonances at d ꢂ 2.3 to ꢂ2.4 ppm.
for aged samples that had been stored in the refrigerator at
3 mM concentration.
Small angle X-ray scattering and X-ray absorption measure-
ments of 3 mM Au colloids in THF was measured at the 12-ID-B
and 12-BM-B line of the Advanced Photon Source of Argonne
National Laboratory. The SAXS data were tted with Shulz–
Zimmerman distribution function.
B.4 Catalytic reduction of p-nitrobenzaldehyde with
triethylsilane
Triethylsilane (0.05 mL, 0.3 mmol) was added to an ethanol
suspension (1.5 mL) of p-nitrobenzaldehyde (43 mg, 0.28 mmol)
and Au–N–SiH–PMHS (1.5 mL, 3 mM Au in THF). The resulting
mixture was stirred in the dark under N2 for 16 h, aer which
the solvent was evaporated and the products were analyzed by
1H NMR aer dissolving in CDCl3. The resonances of the
aldehyde group of the starting material (10.17 ppm) and the
imine product (10.12 ppm) and the CH2 signal of the alcohol
product (4.84 ppm) were used for quantication.
Fig. 3 1H NMR of N–PMHS IV and Au–IV.
–SiH groups to reduce Au(I) to Au(0). This contrasted with the
mixture of Au and N–PMHS where there was ꢂ1 residual –SiH to
60 –RNH in the oligomer, which was equivalent to 1 –SiH to 6Au
(determined from NMR peak area of Si–H).
C Results and discussion
C.1 Interaction of Au(THT)Cl with N–PMHS IV and
N–SiH–PMHS V
C.1.1 d 1.9 and 2.8 ppm resonances. Aer mixing Au(THT)
Cl and N–PMHS IV or N–SiH–PMHS V, the appearance of
resonances at d 1.9 and 2.8 ppm in the 1H NMR due to unbound
THT and the disappearance of the broad resonances of bound
THT (d 1.9–2.4 and 3.1–3.6 ppm) were indicative of displace-
ment THT originally coordinated to Au (Fig. S5†). A similar
displacement of the THT ligand from Au(I) was also observed in
a control experiment in which diethylamine (Et2NH) was used
instead of N–PMHS (Fig. S6†). These observations were consis-
tent with eqn (1) and with previous reported literature.43 Using
an added standard for quantication, it was determined
that the displacement of THT was complete within 1 h for both
Au–IV and Au–V. The strength of the interaction of amine with
AuCl was purported to derive from hydrogen bonding between
N–H and Cl and aurophilic interactions.44
The interaction of Au(THT)Cl with the amino groups on
N–PMHS IV and N–SiH–PMHS V were examined as a function of
time using 1H and 13C NMR. Exposure of IV or V to Au(THT)Cl
resulted in three sets of changes in the spectra (Fig. 2, 3, and
S5†): (1) appearance of sharp resonances at d 1.9 and 2.8 ppm
due to unbound THT. (2) Changes in the resonance associated
with the majority of the –NH protons (Fig. 2, highlighted by
ovals) and the protons of N–CH3 and N–CH2– groups at ꢂd 2.3–
2.4 ppm (Fig. 2). (3) Appearance of a broad resonance at d ꢂ 2.92
ppm (Fig. 3) assigned to the N–CH3 of the amino groups of IV
coordinated to Au(I), and at d ꢂ 2.60 ppm assigned to the
N–CH2– of the same group. These changes will be discussed in
the ensuing paragraphs. It should be noted that with an amine
to Au ratio of 10 in these samples, direct observation of amine
coordination to Au would be difficult because of its limited
concentrations. The large excess of amine also implied that in
the mixture of Au and N–SiH–PMHS, there was a large excess of
–NHR + Au(THT)Cl ¼ THT + Au(NHR)Cl
(1)
1746 | J. Mater. Chem. A, 2015, 3, 1743–1751
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
protons (Fig. S8†). Nonetheless, the assignment was supported
C.1.2 –NH, N–CH3 and N–CH2– protons. The resonances by the fact this correlation in HSQC was more obvious when the
of these protons are shown in Fig. 2. The two protons of the Au concentration was increased to an amine/Au ratio of 2.5
N–CH2– are diastereotopic and appeared as the broad multiplet (Fig. S9†).
at the same general position as the N–CH3 resonance (ꢂd 2.3–
The ratio of ligated amine to Au was 1.6 and 1 for the mixture
2.4 ppm, right panel). Within ꢃ20%, the integrated areas of of IV and Au(THT)Cl when the amine/Au ratios were 10 and 2.5,
–NH, N–CH2– and N–CH3 accounted for all, or at least a large respectively. In contrast, the ratio was 2 using the model
majority, of the protons associated with the amines in the compound Et2NH. This lower ratio for Au–IV could be due to the
samples.
fact that in IV, there might not be two amino groups in the same
The presence of SiH in V caused rapid reduction of Au(I) aer vicinity and the right conguration to coordinate to Au(I)
mixing and large changes in the resonances of –NH, and simultaneously.
N–CH2– and N–CH3 (Fig. 2). The –NH resonance was shied
Addition of the reductant Et3SiH to the solution of IV bound
from d 0.8 to 2.1–2.4 ppm. The overlapping N–CH2– and N–CH3 to Au(I) (amine/Au ¼ 2.5) caused the d 2.90 ppm peak to
peaks now appeared as two peaks, a broader one at d ꢂ 2.38 disappear within 10 min (Fig. S10†), thus further conrming
ppm and a narrower one at d ꢂ 2.42–3.43 ppm. These extensive that this peak was associated with amines coordinated to Au(I).
spectroscopic changes indicated structural changes in the This was also in line with the observation that this peak
oligomer. We hypothesized that the reduction of Au(I) by silane decreased in intensity 5 h aer mixing Au(THT)Cl and IV, which
resulted in the formation of –SiCl (eqn (2)). The formed –SiCl could be explained by the small amount of residual SiH groups
could readily react with a nearby free –RNH in the manner in IV, and its absence in the Au–V samples that had little
shown in eqn (3) (ref. 45) to form a new silazane bond and the unreduced Au.
concomitant protonation of another amine. These structural
modications caused conformational changes of the oligomer
and the observed spectroscopic changes.
C.2 Determination of particle size and stability of Au
colloidal particles
2NHRAuICl + 2R3SiH ¼ 2NHRAu0 + 2R3SiCl + H2
R3SiCl + 2NHR2 ¼ R3Si–NR2 + R2NH2+Clꢀ
(2)
(3)
The importance of the role of oligomer and the proximity of
ligating amine and silane reductant on the Au particle size was
examined by using four different combination of silane reduc-
tant and amine ligands:
(1) amine and silane on the same oligomer (N–SiH–PMHS V).
(2) Amine on oligomer and silane added independently
(Et3SiH + N–PMHS IV).
(3) Oligomer possessing silane but no amine (Me–SiH–
PMHS VI).
C.1.3 d ꢂ 2.92 and ꢂ2.60 ppm resonances. The third
spectroscopic change aer mixing Au(THT)Cl with IV was the
appearance of a new broad resonance at d ꢂ 2.90 ppm (Fig. 3).
This new resonance was observable only for this Au–IV mixture
and not for the Au–V mixture, or in samples without Au. Thus, it
was due to the interaction of Au(I) with IV. We assigned this
peak to the protons of N–CH3 that was directly ligated to Au(I)Cl.
The assignment was supported by HSQC data (Fig. S8†).
The assignment of the very broad, low intensity peak around
d 2.60 ppm to N–CH2– of the amine ligated to Au(I) was also
based on HSQC results, but was less denitive due to the low
intensities of the resonance peaks consequent of the protons
being diastereotopic and being split by multiple neighboring
(4) Trisiloxane molecule containing both silane and amine
(HSiMe2OSiPh2OSiMe2C4H8NHCH3 VIII).
In all these experiments, the Au/amine ratio was kept the
same at 10. Additional experiments were conducted to examine
the effect of preparation temperature and Au/SiH ratio. Table 1
shows the result of these experiments where the average Au
particle sizes were determined from STEM and representative
STEM images are shown in Fig. 4. Not shown in the table is the
preparation using an oligomer without amine ligand (Me–SiH–
Table 1 Average Au particle size in different preparations
Synthesis solution conditionsa
Amine and silane carrier [Olig]b or [trisiloxane] mM [Amine]c mM Au/SiH T (ꢁC) Au colloid diameter (nm) Au/SiO2 diameterf (nm)
a
b
c
d
e
N–SiH–PMHS V
1.6
0.8
0.8
30g
0.8
30
30
30
30
30
1 : 10
1 : 3
1 : 10
1 : 10
1 : 3
0
0
0
0
25
0.69 ꢃ 0.13d
0.89 ꢃ 0.24d
0.94 ꢃ 0.32d
3.05 ꢃ 2.51e
2.55 ꢃ 1.94e
0.92 ꢃ 0.70d
1.95 ꢃ 0.67d
1.87 ꢃ 0.72d
4.03 ꢃ 3.09e
2.99 ꢃ 1.89e
Et3SiH + N–PMHS IV
Et3SiH + N–PMHS IV
Trisiloxane VIII
Et3SiH + N–PMHS IV
a
b
c
N/Au atomic ratio was 10. Based on an average of 38 siloxane units per oligomer. Calculated using the extent of conversion of SiH to amino
units and assuming an average of 38 siloxane units per oligomer. d Based on 200 particles. e Based on 500 particle. f Aer deposition of colloidal
particles onto SiO2 and thermal treatment in an O3 atmosphere at 150 ꢁC. g Same concentration of SiO unit as that present in the preparation of
sample a.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 1743–1751 | 1747

View Article Online
Journal of Materials Chemistry A
Paper
PMHS VI) which produced large nonuniform Au agglomerates examine the effect of preparation temperature. With all other
>10 nm (STEM image Fig. S11†). Thus, not surprisingly, stabi- synthesis parameters identical, the Au particles prepared at ice
lization by amine ligands is essential for the formation of small temperature were smaller and more uniform than those
Au particles.
prepared at room temperature (compare samples b and e). At a
A comparison of sample a, prepared using N–SiH–PMHS V higher temperature, both the reduction rate and mobility of Au
with both silane and amine present in the oligomer, and sample were expected to be higher, which would lead to larger Au
c, prepared with N–PMHS IV that did not possess silane, showed particles.
that both the Au particle size and size distribution (Table 1)
In addition to ST0045M, other techniques were also used to
were smaller for synthesis mediated with V. The contribution of characterize the Au–siloxane colloidal complexes. Dynamic light
the proximity of silane and amine to these differences may be scattering (DLS), which provides information about the hydrody-
minor as our NMR results established that Au(I) was bound to namic diameters of colloidal particles, showed that the mean
the amine upon mixing of Au(THT)Cl and IV. Thus, when Au(I) sizes of the colloidal particles were 1.5, 2.0, 2.7, and 5.6 nm,
was reduced by the externally added silane, a high concentra- respectively for samples a–d (Fig. S12†), consistent with the trend
tion of amines was in the proximity of Au, and the environment found with STEM. Since the DLS data reected the size of the
would be similar to the reduction of Au using V. On the other colloid particles, that is, Au particles embedded in a shell of
hand, the two synthesis were conducted with different organosiloxane oligomers, the particle sizes determined by this
concentrations of oligomers in order to maintain an identical method was expected to be slightly larger than the Au particle size
concentration of amine. Since V had half the amount of amine derived from the STEM data. For similar reasons, SAXS charac-
as IV, its concentration was twice as high. Thus, the Au density terization of sample a (Fig. S13†) also showed much larger parti-
per oligomer V was half of that per oligomer IV. The lower Au cles than those derived from STEM. XANES spectra of all these
density would result in a thicker layer of the viscous siloxane samples conrmed that all the Au were metallic (Fig. S15†).
surrounding every Au particle, which should be more effective
in retarding agglomeration by inter-oligomer diffusion of Au (diameter),6 the very narrow size distribution shown in the data
particles. indicated sample uniformity – a testimony that the method
Since the intensity of the DLS signal is proportional to
The importance of steric barrier rather than the proximity of reported here was indeed facile and versatile, as no purication
the reducing and ligating functionalities in controlling the Au or separation step was employed aer mixing the Au precursor
particle size was further underscored by the large and highly and the oligomer.
non-uniform Au particles present in sample d. This preparation
UV-visible spectroscopy was also used to characterize the Au
was conducted under identical conditions as sample a, except colloids. As shown in panel A of Fig. 5, only sample d showed
that a small molecule disiloxane VIII was used. Although VIII the characteristic surface plasmon resonance at 525 nm for Au
also possessed both SiH and amine functionalities, it was much nanoparticles >2 nm.1 Spectra of all other samples showed only
less effective in providing steric barrier than the long exible weak features: small peaks at 330 and 280 nm for sample a, the
oligomer V in deterring Au particle agglomeration.
barely detectable absorption at ꢂ560 and ꢂ410 nm, and more
When the silane Et3SiH was added externally, the prepara- intense absorption at 330 nm for sample b. For sample c the
tion using a lower concentration of reductant resulted in absorptions at 560, 410, and 330 became easily identiable.
smaller Au particles (compare samples b and c). Possibly, this
The small absorption features detected for samples a–c
was because the rate of reduction was slower at a lower reduc- suggested the presence of signicant concentrations of some
tant concentration, allowing more effectively amine coordina- well-dened species. It has been shown by matrix isolation that
tion before Au agglomeration could take place. That the rate of atomic to trimeric Au showed electronic transitions within 200–
reduction was important was consistent with the experiments to 400 nm due to transitions between discrete HOMO–LUMO
levels.46 For Au clusters encapsulated in well-dened poly(-
amidoamine) dendrimers, Zheng et al.47 reported that the
maxima of the excitation spectra for Au5, Au8 and Au13 clusters
were 332, 385 and 431 nm, respectively. Negishi et al. found that
the absorbance maximum of their glutathione capped clusters
ranged from 330 nm to 570 nm for clusters with Au core sizes
that ranged from 10–22.48 Thus literature results suggested that
these absorption features corresponded to very small Au clus-
ters (<1 nm) It is likely that many of these clusters were too
small to be observed with the STEM instrument we used (limit
ꢂ0.6 ꢃ 0.2 nm). Thus, in samples where UV-vis data suggested
their presence, the average particle size of these samples should
be smaller than those shown in Table 1. The variation of the
density of these smaller particles were consistent with the
trends determined by STEM.
These colloidal particles were quite stable on storage in the
absence of light in a refrigerator. Nonetheless there was
Fig. 4 STEM images of the Au colloid, (a)–(e) corresponded to
experiments 1 to 5 in Table 1, respectively. Scale bar is 20 nm.
1748 | J. Mater. Chem. A, 2015, 3, 1743–1751
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
organics and convert the siloxane into SiO2. Particle growth
occurred, but the Au particle sizes and uniformity for the
different samples maintained the same trend as
was observed for the parent colloidal samples (Table 1 and
Fig. 6). The average Au particle size of sample a remained
subnanometer.
C.3 Accessibility of Au colloids surface
Sample a was chosen to test for catalytic activity to see
whether Au clusters embedded in the oligomer was accessible
to reactants. The test reaction was p-nitrobenzaldehyde
reduction by triethylsilane (Et3SiH) using equimolar quanti-
ties of the two reactants (Scheme 2). In this reaction, the
product IX was formed by rst reduction of p-nitro-
benzaldehyde to p-aminobenzaldehyde, followed by the
subsequent domino reaction of p-aminobenzaldehyde
coupling with the nitrobenzaldehyde to form an imine bond.
At the end of reaction, there were ꢂ20% compound IX and
ꢂ8% p-nitrobenzyl alcohol in the product. Since the forma-
tion of each molecule IX consumes six silanes, the result
indicated complete consumption of silane. In the absence of
the Au colloid from Expt 1, the yield of IX and p-nitrobenzyl
alcohol were only 2% and 5%, respectively. Thus, the Au
colloid was an active catalyst, and the Au particles were
accessible for reaction.
Fig. 5 UV-visible spectra of samples a–d listed in Table 1. Panel A:
spectra taken 2 h after completion of synthesis and panel B: 4 weeks
later. The spectra for samples b–d in panel A were displaced verti-
cally from the baseline by 0.25 absorbance unit for clarity. HPLC
grade THF was the solvent for samples. a–c in panel A, whereas
research grade was used for sample d. The concentrations of Au
were the same for all samples.
evidence of agglomeration with time. As shown in Fig. 5 panel
B, aer 4 weeks of storage, with the exception of sample a the
other samples all developed the plasmon peak around 520 nm
that was characteristic of Au nanoparticles. Even the origi-
nally colorless solution of sample a turned brown, suggesting
some degree of agglomeration. Since the Au particles in these
samples were surrounded by siloxane oligomers, the initial
growth of Au clusters probably occurred within the same Au–
PMHS colloidal particle. With time, collisions between
colloidal particles occurred where interparticle exchanges of
Au particles took place that enabled their subsequent
agglomeration.
Fig. 6 STEM images of Au/SiO2; images (a)–(e) corresponded to
samples in Table 1 (last column). Scale bar is 20 nm.
Another measure of stability was to examine the degree of
particle agglomeration aer thermal treatment. We depos-
ited these colloidal particles onto silica and thermally treated
them in an O3 atmosphere at 150 ꢁC to remove all the
Scheme 2 Catalytic conversion of p-nitrobenzaldehyde catalysed by
sample a.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 1743–1751 | 1749

View Article Online
Journal of Materials Chemistry A
Paper
10 J. K. Edwards, B. Solsona, E. Ntainjua, A. F. Carley,
A. A. Herzing, C. J. Kiely and G. J. Hutchings, Science, 2009,
323, 1037–1041.
11 M. Turner, V. B. Golovko, O. P. Vaughan, P. Abdulkin,
A. Berenguer-Murcia, M. S. Tikhov, B. F. Johnson and
R. M. Lambert, Nature, 2008, 454, 981–983.
12 M. Boronat, A. Leyva-Perez and A. Corma, Acc. Chem. Res.,
2014, 47, 834–844.
13 A. Corma, P. Concepcion, M. Boronat, M. J. Sabater, J. Navas,
M. J. Yacaman, E. Larios, A. Posadas, M. A. Lopez-Quintela,
D. Buceta, E. Mendoza, G. Guilera and A. Mayoral, Nat.
Chem., 2013, 5, 775–781.
14 M. Flytzani-Stephanopoulos, Acc. Chem. Res., 2014, 47, 783–
792.
15 W. Jiang, B. Y. Kim, J. T. Rutka and W. C. Chan, Nat.
Nanotechnol., 2008, 3, 145–150.
16 C. K. Kim, P. Ghosh, C. Pagliuca, Z.-J. Zhu, S. Menichetti and
V. M. Rotello, J. Am. Chem. Soc., 2009, 131, 1360–1361.
17 A. Verma, O. Uzun, Y. Hu, Y. Hu, H.-S. Han, N. Watson,
S. Chen, D. J. Irvine and F. Stellacci, Nat. Mater., 2008, 7,
588–595.
D Conclusions
Using the preparation of Au particles, we demonstrated a
practical, convenient and facile method that does not rely on
strong stabilizers to prepare metal colloids. The synthesis is
mediated by siloxane oligomers containing both reducing (SiH)
and ligating (amine) functionalities and permits controlled
variation of metal particle size at subnanometer scale. Variation
of the ratios of these functionalities and using small model
molecules allowed us to identify relevant synthesis parameters,
such as the extent of steric and ligand stabilization and the rates
of reduction, that affect the properties of the resultant Au
particles. Monitoring the synthesis with NMR at different stages
of the synthesis showed the ready displacement of THT from Au
by amines in the oligomer with the concomitant appearance of
new NMR peaks associated with amines bound to Au(I). These
peaks subsequently disappeared with the reduction to metallic
Au.
The Au particles obtained by this method were quite stable,
even aer oxidation treatment to remove the organics. Since the
principles used here are general, this method should be appli-
cable to the synthesis of subnanometer particles of other sup-
ported late transition metals. Thus, it could have many
potential applications when such particles are called for, such
as in sensors and catalysis.
18 P. Zhao, N. Li and D. Astruc, Coord. Chem. Rev., 2013, 257,
638–665.
19 P. C. Ohara, D. V. Leff, J. R. Heath and W. M. Gelbart, Phys.
Rev. Lett., 1995, 75, 3466–3469.
20 R. H. Terrill, T. A. Postlethwaite, C.-h. Chen, C.-D. Poon,
A. Terzis, A. Chen, J. E. Hutchison, M. R. Clark and
G. Wignall, J. Am. Chem. Soc., 1995, 117, 12537–12548.
21 M. M. Alvarez, J. T. Khoury, T. G. Schaaff, M. Shagullin,
I. Vezmar and R. L. Whetten, Chem. Phys. Lett., 1997, 266,
91–98.
22 R. L. Whetten, J. T. Khoury, M. M. Alvarez, S. Murthy,
I. Vezmar, Z. L. Wang, P. W. Stephens, C. L. Cleveland,
W. D. Luedtke and U. Landman, Adv. Mater., 1996, 8(5),
428–433.
23 M. Brust, M. Walker, D. Bethell, D. J. Schiffrin and
R. Whyman, J. Chem. Soc., Chem. Commun., 1994, 801–802.
24 W. W. Weare, S. M. Reed, M. G. Warner and J. E. Hutchison,
J. Am. Chem. Soc., 2000, 122(51), 12890–12891.
25 N. de Silva, J.-M. Ha, A. Solovyov, M. M. Nigra, I. Ogino,
S. W. Yeh, K. A. Durkin and A. Katz, Nat. Chem., 2010, 2,
1062–1068.
Acknowledgements
This work was supported by the Department of Energy, Basic
Energy Sciences, grant no. DE-FG02-01ER15184. Z. W. and X. H.
acknowledge funding support from China Scholarship Council.
This work made use of the TEM and NMR in the EPIC facility of
NUANCE Center and IMSERC facility at Northwestern Univer-
sity. This research used resources of the Advanced Photon
Source, a U.S. Department of Energy (DOE) Office of Science
User Facility operated for the DOE Office of Science by Argonne
National Laboratory under Contract no. DE-AC02-06CH11357.
Notes and references
1 M. M. Alvarez, J. T. Khoury, T. G. Schaaff, M. N. Shagullin,
I. Vezmar and R. L. Whetten, J. Phys. Chem. B, 1997, 101,
3706–3712.
26 A. G. Kanaras, F. S. Kamounah, K. Schaumburg, C. J. Kiely
and M. Brust, Chem. Commun., 2002, 2294–2295.
2 J. F. Hicks, A. C. Templeton, S. Chen, K. M. Sheran, R. Jasti,
R. W. Murray, J. Debord, T. G. Schaaff and R. L. Whetten, 27 J. M. Abad, I. E. Sendroiu, M. Gass, A. Bleloch, A. J. Mills and
Anal. Chem., 1999, 71, 3703–3711.
3 R. Jin, Nanoscale, 2010, 2, 343–362.
4 Y. Negishi, K. Nobusada and T. Tsukuda, J. Am. Chem. Soc.,
2005, 127, 5261–5270.
D. J. Schiffrin, J. Am. Chem. Soc., 2007, 129, 12932–12933.
ˆ
ˆ
28 V. I. Parvulescu, V. Parvulescu, U. Endruschat, G. Filoti,
¨
F. E. Wagner, C. Christian Kubel and R. Richards, Chem.–
Eur. J., 2006, 12(8), 2343–2357.
5 M. Zhu, C. M. Aikens, F. J. Hollander, G. C. Schatz and R. Jin, 29 Y. Negishi and T. Tsukuda, J. Am. Chem. Soc., 2003, 125,
J. Am. Chem. Soc., 2008, 130, 5883–5885.
4046–4047.
6 M.-C. Daniel and D. Astruc, Chem. Rev., 2004, 104, 293–346. 30 M.-Y. Tsai and J.-C. Lin, J. Biomed. Mater. Res., 2001, 55, 554–
7 J. Zheng, C. Zhang and R. M. Dickson, Phys. Rev. Lett., 2004,
93, 077402.
8 M. Haruta, Gold Bull., 2004, 37, 27–36.
565.
31 F. Zhang, Y. Zhou, Y. Chen, Z. Shi, Y. Tang and T. Lu, J.
Colloid Interface Sci., 2010, 351(2), 421–426.
9 J. Huang, T. Akita, J. Faye, T. Fujitani, T. Takei and 32 D. Wang, Y. Liu, X. Zhou, J. Sun and T. You, Chem. Lett.,
M. Haruta, Angew. Chem., Int. Ed., 2009, 48, 7862–7866.
2007, 36(7), 924–925.
1750 | J. Mater. Chem. A, 2015, 3, 1743–1751
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
Journal of Materials Chemistry A
33 P. D. Jadzinsky, G. Calero, C. J. Ackerson, D. A. Bushnell and 41 W. Xue, M. C. Kung and H. H. Kung, Chem. Commun., 2005,
R. D. Kornberg, Science, 2007, 318, 430–433. 2164–2166.
34 M. W. Heaven, A. Dass, P. S. White, K. M. Holt and 42 R. Uson, A. Laguna and M. Laguna, Inorg. Synth., 1989, 26,
R. W. Murray, J. Am. Chem. Soc., 2008, 130, 3754–3755. 85–91.
35 M. Zhu, E. Lanni, N. Garg, M. E. Bier and R. Jin, J. Am. Chem. 43 S. Gomez, K. Philippot, V. Colliere, B. Chaudret, F. Senocq
Soc., 2008, 130, 1138–1139. and P. Lecante, Chem. Commun., 2000, 1945–1946.
36 J. F. Parker, J. E. Weaver, F. McCallum, C. A. Fields-Zinna 44 B. Ahrens, P. G. Jones and A. K. Fischer, Eur. J. Inorg. Chem.,
and R. W. Murray, Langmuir, 2010, 26, 13650–13654. 1999, 1103–1110.
37 M. N. Martin, D. Li, A. Dass and S. K. Eah, Nanoscale, 2012, 4, 45 H. Yoshida, T. Morishita, H. Fukushima, J. Ohshita and
4091–4094. A. Kunai, Org. Lett., 2007, 9, 3367–3370.
38 B. Fu, M. N. Missaghi, C. M. Downing, M. C. Kung, 46 W. Harbich, S. Fedrigo, J. Buttet and D. M. Lindsay, J. Chem.
H. H. Kung and G. Xiao, Chem. Mater., 2010, 22, 2181–2183. Phys., 1992, 96, 8104–8108.
39 N. A. Mashayekhi, Y. Y. Wu, M. C. Kung and H. H. Kung, 47 J. Zheng, C. Zhang and R. M. Dickson, Phys. Rev. Lett., 2004,
Chem. Commun., 2012, 48, 10096–10098. 93, 077402/077401–077402/077404.
40 Y. Li and Y. Kawakami, Macromolecules, 1999, 32, 3540–3542. 48 Y. Negishi, K. Nobusada and T. Tsukuda, J. Am. Chem. Soc.,
2005, 127, 5261–5270.
This journal is © The Royal Society of Chemistry 2015
J. Mater. Chem. A, 2015, 3, 1743–1751 | 1751