1

Highly Dispersed Supported Metal Catalysts Prepared via In-Situ Self-Assembled Core-Shell

Precursor Route

Liuye Mo, a Eng Toon Saw, a Yonghua Du, b Armando Borgna, b Ming Li Ang, a Yasotha Kathiraser, a

Ziwei Li, a Warintorn Thitsartarn, c Ming Lin, c Sibudjing Kawi a*

a Department of Chemical and Biomolecular Engineering, National University of Singapore, Singapore

117576

b Institute of Chemical & Engineering Sciences (ICES), A*STAR, 1 Pesek Road, Jurong Island,

Singapore 628733

c Institute of Materials Research and Engineering, A*STAR, 3 Research Link, Singapore 117602

Corresponding author:

Tel: (65) 6516 6312

Fax: (65) 6779 1936

E-mail: chekawis@nus.edu.sg

2

ABSTRACT

Preparation of highly dispersed supported metal catalysts, which has high activity and selectivty, is

still one of the main challenges for catalysis science. Highly dispersed supported metal catalysts via

an economical incipient wetness impregnation method by impregnation of a small amount of oleic

acid mixed with metal nitrate on catalyst support is reported herein. XRD, TEM(STEM), SEM, XPS,

FTIR and EXAFS results revealed that the small amount of oleic acid and metal nitrate solutions

would In-situ self-assemble to form core-shell precursors (metal nitrate species as core and metal

oleate as shell), and metal agglomeration during heat treatment was drastically prevented by the steric

hindrance of the shell structure. Therefore, the novel and facile method was named as in-situ self-

assembled core-shell precursor route. The effectiveness of this method was attested by copper and

nickel on silica catalysts. Compared to the conventional preparation method, the as-prepared highly

dispersed metal catalysts of Cu/SiO2 and Ni/SiO2 exhibited excellent catalytic activities for water gas

shift reaction and carbon dioxide reforming of methane at high temperatures.

Keywords: self-assembled core-shell, oleic acid, highly dispersed supported metal catalyst, water gas

shift reaction, dry reforming of methane

3

1. Introduction

Catalysis is an innovative and powerful technology to solve many of today’s energy and

environmental issues. Particularly, supported metal catalysts hold great importance in catalysis

researches and industrial applications [1-4]. Highly dispersed supported metal catalysts are greatly

sought after since a decrease in the supported particle size is accompanied by an increase in the surface

area of active phase and a consequent enhancement in the catalytic activity. In the past decades,

tremendous advances have been achieved in the preparation of highly-dispersed metal catalysts using

different methodologies, such as sol-gel method [5-11], co-precipitation or deposition precipitation

method [12-19] and ion exchange or strong electrostatic adsorption (SEA) [20, 21]. Although all of the

above-mentioned methods have been successfully adopted in some applications, numerous

disadvantages such as high cost and difficulty in scaling up hindered application of these methods in the

industrial setting.

Impregnation method, in particular incipient wetness impregnation, is highly desirable due to its

practical simplicity and its minimal generation of waste streams. In a bid to improve the metal

dispersion on support, various metal precursors and organic metal compounds (acetate [22], citrate [23,

24] and acetylacetonate [25, 26]), have been utilized. Recently, Ni(1,5-cyclooctadiene)2 was reacted

with octylsilane to form nickel-silicide colloids [NixSi−C8H17] in the presence of H2, and was used as a

precursor to prepare highly dispersed supported nickel catalysts[27]. Besides, a versatile method

involving a surface organometallic chemistry (SOMC) route that uses mono(η3-allyl)nickel complex as

precursors to prepare supported small size nickel nanoparticles (ca. 1–3 nm) with narrow size

distribution was reported [28]. However, the organometallic compounds mentioned above are not only

costly but also have low solubility in aqueous solution and/or highly sensitive to moisture or air,

rendering them difficult to be applied in industry. Due to their inherent advantages, such as low cost,

high solubility in water and ease of anion removal, metal nitrates are widely utilized as precursors for

the preparation of supported metal catalysts. However, the drawback of using metal nitrate during the

preparation of supported metal catalyst lies in the wide distribution of metal particle size which falls

4

between 1 to 100 nm. The ease of agglomeration of metal precursors during the drying and calcination

steps has been attributed for the cause of this phenomenon [29]. Interestingly, the de Jong group found

that the metal dispersion could be enhanced by using a NO/He mixture atmosphere during catalyst

calcination [29, 30]. It was revealed that NO could promote nearly complete hydrolysis of metal nitrate

to hydroxynitrates, which possessed a reduced mobility to agglomerate upon decomposition that

occurred between 100 oC and 150 oC [29]. In addition, Stucky et al. also reported another noteworthy

method to prepare oxide-supported metal nanoparticle in which the weak interaction between

presynthesized hydrophobic metal particles and oxides supports was utilized to assemble the

nanoparticles onto the oxides surface [31]. However, this method cannot be used in protic solvents and

large scale production. Hence, there remains a tremendous barrier to design a more general and

simplistic method for preparation of highly dispersed, supported metal catalysts. Precedently, we have

reported a method to prepare highly-dispersed and stable Cu/SiO2 and Ni-La2O3/SiO2 catalysts using the

incipient wetness impregnation (IWI) in the presence of a small amount of fatty acids and metal nitrate

[32, 33]. Herein, XRD, TEM, SEM, XPS, FT-IR and EXAFS techniques are applied to elucidate the

promoting mechanism of oleic acid on the dispersion of Ni/SiO2 and Cu/SiO2 catalysts. Catalytic

activities of the prepared Ni/SiO2 and Cu/SiO2 catalysts were measured using carbon dioxide reforming

of methane and water gas shift reaction, respectively.

2. Experimental Section

2.1 10%Cu/SiO2 catalyst preparation and water gas shift reaction conditions

Cu/SiO2 catalysts promoted with oleic acid (OA) were prepared with incipient wetness impregnation,

with a copper loading of 10wt.%. Firstly, 2.1 g copper nitrate (from Merck) was dissolved in D.I. water,

and then different amount of oleic acid (from Sigma-Aldrich) was added into the solution. At last, 5 g

spherical silica (from Kanto Chemical Co.: specific surface area = 753 m2/g, mean pore size = 7.5 nm)

was introduced into the above solution. The samples were subsequently impregnated at room

temperature for > 6 hours and then dried overnight at 100 oC. The above dried samples were calcined at

5

450 oC for 4 hours. The amount of OA was calculated as molar ratio of X = nOA/nCu, and X was adopted

as 0, 0.05, 0.25, 0.5, and 1.0, respectively. And the 10%Cu/SiO2 prepared with different X was

designated as 10%Cu/SiO2/X. To investigate the catalytic performance of the prepared catalysts at 300

oC, the Water Gas Shift Reaction (WGSR) was used as a probe reaction. The feed composition, with a

total flow rate of 50 ml/min, was as follows: 5 mol.% CO, 25 mol.% H2O, and He balance. The effluent

gases were analyzed by an HP-GC equipped with a Hayesep D column. All catalysts were reduced at

300 oC for one hour prior to the onset of WGSR.

2.2 5%Ni/SiO2 catalyst preparation and dry CO2 reforming of CH4 reaction conditions

The 5%Ni/SiO2/X catalysts were prepared similar to those of 10%Cu/SiO2. The X = nOA/nNi ratios

adopted were 0, 0.05, 0.1, 0.2, 0.5, 1.0, 1.5, 2.0, 3.0 and 4.0, respectively. Calcination was conducted at

700 oC for 4 hours in air atmosphere.

The dry CO2 reforming of CH4 (DRM) reactions were carried out in a quartz reactor. The catalyst

under study was loaded and packed in the middle of the quartz-tube using quartz wool which was inert

in the reaction. Under atmospheric pressure, the catalyst was heated from the initial room temperature to

a reaction temperature of 700 oC at a rate of 20.0 oC/min in a horizontal electric furnace. The Gas-

Hourly-Space-Velocity (GHSV) was 72000 ml.hr-1.g (cat.)-1. Additionally, high purity feed gases

(99.99%) were controlled by mass flow rate controllers and a molar ratio of CH4/CO2/N2 =1/1/1 was

adopted. Prior to the catalytic reaction, the catalysts were reduced in-situ in a H2 atmosphere (purity =

99.99%) at 700 oC for 1 hour and 0.05 g of catalyst was used in each catalytic test. To remove any

moisture from the effluent gas stream, a cold trap was employed to condense the moisture prior to gas

analysis by an on-line gas chromatograph (GC). Effluent gas from the reactor was analyzed by an

Agilent HP-GC equipped with two columns of 5Å and Porapak Q. The nitrogen in the reactants acted as

an internal standard.

2.3 Copper (nickel) oleate synthesis

6

The copper oleate synthesis follows the method reported in literature [34]. Typically, 0.85 g (5mM)

CuCl2.2H2O (Fisher Chemical, >99%) was dissolved in a mixture solution containing 6 ml of water, 8

ml of ethanol and 14 ml of hexane. The above solution was added to 3.04 g (10 mM) sodium oleate

(TCI, >97%) under vigorous stirring for 4 h at 70 oC. The organic phase was separated in a separatory

funnel and washed three times with 3 ml of water. Following that, the solvent of hexane and ethanol

were distilled. Finally, the dark blue solid obtained was dried in vacuum. Nickel oleate synthesis

procedures were the same as copper oleate.

2.4 Catalyst characterization

Powder X-Ray Diffraction (XRD) characterizations for the fresh and reduced catalysts were

performed on Shimazu XRD-6000 using a Cu target Kα-ray (40 kV and 30 mA) as the X-ray source.

The XRD analysis was promptly conducted after the 10%Cu/SiO2 and 5%Ni/SiO2 catalysts had

undergone reduction at 300 oC and 700 oC for one hour and the transfer of sample to the XRD chamber

was carried out with minimal exposure to air. Thereafter, the crystalline sizes of metal oxides and metals

were calculated by Scherrer equation.

Transmission Electron Microscopy (TEM) images were taken using a JEOL JEM-2100F. Prior to the

TEM analysis, the catalyst was reduced at 300 oC for copper catalysts and 700 oC for nickel catalysts in

purified H2 for one hour. The samples were dispersed in ethanol solution and ultrasonicated for 30 min.

The above solutions were then pipetted onto the copper grid for TEM observation

Scanning Transmission Electron Microscopy (STEM) was used to capture the catalyst images both

before and after catalytic reaction. The images were acquired by a FEI Titan 80-300 S/TEM which was

operated at 200 KV. In order to minimize contamination during the scanning process, the specimen was

placed in a high-vacuum chamber overnight and treated with oxygen plasma for 1-2 seconds before

observation under electron beams.

The morphology and surface elemental composition on the catalyst was analyzed using a Scanning

Electron Microscope coupled to an Energy Dispersive Spectroscopy (SEM-EDS, Jeol JSM-6700). The

7

samples were degassed under vacuum and coated with platinum (about 10 nm thickness) at 20 mA for

120 seconds prior to analysis. The 10%Cu/SiO2/X catalysts were reduced under hydrogen atmosphere at

300 oC for 1 hour before SEM-EDS analysis.

H2-Temperature-Programmed Reduction (H2-TPR) experiments were performed on CHEMBET-3000.

Prior to H2-TPR, the samples were pretreated in He atmosphere at 300 oC for 1 hour. After the sample

was cooled down to room temperature, 5%H2/N2 mixture gas with a flow rate of 30 ml/min was

introduced to the 50 mg catalyst fixed with quartz wool in a U quartz tube (I.D.: 1/4 inch). Upon

stabilization of the baseline, the temperature was increased with a ramping rate of 10 oC/min and the

effluent gas was analyzed by TCD.

X-ray Photoelectron spectroscopy (XPS) analyses were performed on KRATOS AXIS HIS 165.

Anode HT with a voltage of 15 kV, a current of 10 mA and a pass energy of 80 eV was used. C1s (284.5

eV) was used as an internal standard to eliminate the charge effect.

Cu K-edge Extended X-ray absorption fine structure (EXAFS) spectra were recorded at room

temperature in a transition mode at Singapore Synchrotron Light Source (SSLS), XAFCA facility. A Si

(III) double crystal was used to monochromatize the X-rays from the 700Mev electron storage ring. In a

typical experiment, the sample was prepared as a compressed pellet and loaded into a cell. Each sample

was measured 6 times to improve the signal to noise ratio of data and to ensure the data. EXAFS data

were analyzed using the EXAFS program, Winxas. Fourier transformation of k3-weighted EXAFS data

were performed over the range k = 2 -12 Å-1. The bond length/peak position was adopted without phase

correction.

Differential Thermal Analysis and Thermal Gravimetry Analysis (DTA-TGA) was carried out using a

Shimadzu DTG-60 thermoanalyzer. The spent catalysts were heated in an atmosphere of air from room

temperature to 800 oC at a heating rate of 10.0 oC/min.

3. Results and discussion

3.1 XRD results

8

Fig. S1 (in Supporting Information) displays the XRD patterns of fresh 10%Cu/SiO2/X catalysts. Fig.

S1 shows that the CuO peaks were dramatically diminished when a small amount of OA (X = 0.05) was

introduced, indicating high dispersion of CuO on SiO2. This comes as a surprise since a small amount of

OA could promote the dispersion of CuO despite the insolubility of OA in water. However, at higher

values of X above 0.25, the intensity of CuO peaks increased. The crystalline sizes of CuO are 38.0,

11.9, 6.3, 9.3 and 10.6 nm for X = 0, 0.05, 0.25, 0.5, and 1.0, respectively. The optimal X is 0.25. The

copper crystalline size (dCu) on silica after reduction at 300 oC for one hour is depicted in Fig. 11. The

dCu exhibited significant dependence on X. With the addition of a small amount of OA, dCu decreased

tremendously from 38.3 nm to 11.0 nm as X increased from 0 to 0.05. However, upon a further increase

in X from 0.25 to 1.0, dCu increased from 6.0 nm to 17.6 nm.

Interestingly, the XRD peaks of NiO on 5%Ni/SiO2/X (the 5%Ni/SiO2 catalysts with different X were

abbreviated as 5%Ni/SiO2/X, and X was a molar ratio of nOA/nNi) were almost invisible as X varied in a

wide range of 0.05-1.5 (shown in Fig. S2), displaying a great disparity from the copper catalysts. The

peaks of NiO were evidently observed when the X was > 1.5. Accordingly, the crystalline sizes of

nickel (dNi) on reduced samples were around 3.0 nm for X between 0.05 and 1.5. The dNi increased

slightly from around 3.0 nm to 4.0 nm with the increment in X from 1.5 to 3.0. Comparatively, the

crystalline size of nickel on reduced 5%Ni/SiO2/0 was significantly larger at around 12.0 nm (Fig. S3).

In order to track the evolution in the catalyst composition, the 10%Cu/SiO2/X samples dried at 100

oC were characterized by XRD (shown in Fig. S4). From XRD profiles, it can be seen evidently that

copper hydroxynitrate phase could be detected in the sample without OA. However, no copper

compound phases could be seen for 10%Cu/SiO2/X samples with X ≧0.25. Copper nitrate would

partially hydrolyze to form copper hydroxynitrate and agglomerate into big particles during drying at

100 oC. Subsequently, they would redistribute and form big particles of CuO under high calcination

temperature conditions [29]. XRD results display that OA plays a vital role in preventing copper nitrate

from forming bigger particles on 10%Cu/SiO2/X during the thermal process, thereby achieving high

dispersion of CuO after high temperature calcination.

9

3.2 TEM (STEM) results

Fig. 1 shows the TEM image characterizing the Cu particle size distribution of 10%Cu/SiO2/0.25

catalyst after 300 oC reduction. It can be observed that the Cu particles were homogenous and had a

narrow distribution range between 2.0-6.0 nm with a mean size of 3.8 nm. However, the Cu particles of

10%Cu/SiO2/0 catalyst ranged from 2.0 to 200.0 nm, with a mean size of 38.0 nm (Fig. S5).

Fig. 1. TEM image (bright field) and particle size histogram of 10%Cu/SiO2/0.25 catalyst after 300 oC

reduction.

Fig. 2 exhibits STEM images of 5%Ni/SiO2 catalyst reduced at 700 oC for one hour. The nickel

particles dispersed well and homogeneously on the silica support with a narrow distribution between 1.0

and 6.0 nm, and the mean particle was 2.9 nm. The result is in good agreement with the crystalline size

measured with XRD (Fig. S3). In contrast, in the absence of OA, the nickel particles on silica derived a

wide range between 3.0 and 50.0 nm and the mean size was 19.0 nm (Fig. S6).

10

Fig. 2. STEM image (Dark field) and particle size distribution histogram of 5%Ni/SiO2/0.5 catalyst

after 700 oC reduction.

3.3 SEM-EDS results

Fig. 3. SEM-EDS images of 10%Cu/SiO2/0

11

Fig. 4. SEM-EDS images of 10%Cu/SiO2/0

SEM was conducted to observe the morphologies of the catalysts prepared in presence/absence of

OA. From SEM images, it is evident that the size of micro-spherical silica is between 20-60 m. The

10%Cu/SiO2/0 catalyst exhibited large rod-like particles that agglomerated on the silica outer surface

(Fig. 3). The molar ratio of Si to Cu (nCu/nSi) measured by EDS on the big particle surface was 1.61 (Fig.

3 Left), while it was 0.05 on the smooth surface (Fig. 3 right). The theoretical value calculated was 0.11.

Therefore, the big particles on the silica surface are most likely copper particles although the EDS result

showed that the big particles were not purely copper and had interferences with the silica support. On

the contrary, the surfaces of 10%Cu/SiO2/0.25 catalyst (Fig. 4) were smoother than that of

10%Cu/SiO2/0. The rough surface was analyzed by EDS and showed the nCu/nSi was as low as 0.04 (Fig.

4 left), which suggested that the rough surface was not copper particle and was lower in copper

concentration compared to the smooth surface value (nCu/nSi =0.14). The SEM-EDS results showed that

OA could effectively prevent the copper species from agglomerating to form big particles on the silica

outer surface and promoted the copper species to homogeneously distribute on the silica support.

3.4 H2-TPR results

H2-TPR is a useful technique in characterizing the reduction behaviour of a catalyst, which affects its

catalytic performance. The TPR profiles of 10%Cu/SiO2/X catalysts are illustrated in Fig. 5. The CuO

on 10%Cu/SiO2/0 showed two reduction peaks at 268 and 337 oC. However, on 10%Cu/SiO2/0.05,

there was a single peak centred at 268 oC. With increasing X from 0 to 0.25, the reduction peak shifted

12

further to lower temperature at 252 oC together with the appearance of a small shoulder peak at 268 oC.

However, increment in X from 0.25 to 1.5 decreased intensity of the low temperature reduction peak

and increased the intensity of the high reduction peak at 288 oC. It could be observed that catalysts with

bigger CuO crystalline size had higher reduction temperature and these TPR results are well correlated

to XRD results [35, 36]. Furthermore, 10%Cu/SiO2/0 showed two peaks at low temperature and high

temperature which could be ascribed to the reduction of small CuO particles and bigger CuO particles,

respectively. Hence, it can be inferred from the TPR results that OA used in the catalyst preparation

promoted the dispersion and homogeneous distribution of CuO on the silica support whereby the

reduction peaks appeared as single or very near shoulder peaks.

Fig. 6 exhibits H2-TPR curves of 5%Ni/SiO2/0 and 5%Ni/SiO2/0.5. The peaks at low temperature

(417 and 499 oC) and high temperature at > 600 oC could be assigned to the reduction of nickel oxide

species and nickel silicate which had weak and strong interaction with the silica support, respectively

[24]. Low temperature reduction peaks between 417 and 499 oC were evident from the H2-TPR curve of

5%Ni/SiO2/0.5, indicating that there were nickel oxide species interacting weakly with silica or

unsupported free NiO that had no interaction with the support.

50 100 150 200 250 300 350 400 450 500

X=1.5

X=0.05

288 oC268 oC

Hyd

roge

n co

nsum

ptio

n (

a. u

. )

Temperature ( oC )

X=1.0

X=0.5

X=0.25

without OA

CuO

252 oC337 oC

Fig. 5. H2-TPR profiles of 10%Cu/SiO2/X

13

0 100 200 300 400 500 600 700 800 900 1000

5%Ni/SiO2 /0.5

609 oC

Hyd

roge

n c

ons

um

ptio

n (

a.u

.)

Temperature ( oC )

417 oC499 oC

5%Ni/SiO2 /0

Fig. 6. H2-TPR profiles of 5%Ni/SiO2/X

3.5 FT-IR results

2000 1800 1600 1400 1200 1000 800 600

1.5

1.0

0.5

0.25

0.05

Ads

orpt

ion(

a.u.

)

Wavenumber(cm-1)

1460 cm-1

1719

0

X=nOA

/nCu

1565

Fig. 7. FT-IR spectra obtained from uncalcined samples of 10%Cu/SiO2/X dried at 100 oC.

14

In order to further elucidate the promoting effect of OA, the 10%Cu/SiO2/X samples dried at 100 oC

were measured by FT-IR. From Fig. 7, the peak at 1719 cm-1 which can be assigned C=O stretch of free

oleic acid appeared for X > 0.5. The intensities of characteristic carboxylate peaks at 1565 and 1460 cm-

1, ascribed to antisymmetric stretch of COO- and symmetric stretch of COO- [34, 37], increased with X.

The carboxylate group may combine with a metal cation in either a monodentate, bidentate (chelating),

or bridging mode. The coordination type can be interpreted based on the wavenumber separation (△)

between the antisymmetric, υas(COO−), and the symmetric, υs(COO−), stretching bands. For △>200

cm−1, a monodentate ligand is expected, and for △<110 cm−1, a bidentate. For a bridging ligand, △ lies

somewhere between 140 and 200 cm−1 [38]. In this case, △=1565-1460=105, is smaller than 110 cm-1

indicating that oleic acid is bound in a bidentate mode, with both oxygen atoms symmetrically

coordinated to the surface.

3.6 XPS results

Fig. 8. Cu 2p3/2 XPS spectra of uncalcined 10%Cu/SiO2/X and copper oleate samples.

15

XPS is a sensitive and effective technique to characterize surface species of a solid. Fig. 8 displays

binding energy of Cu2p3/2 of uncalcined 10%Cu/SiO2/X and copper oleate samples. The sample without

OA showed two Cu 2p3/2 peaks at 935.3 eV and 933.0 eV. The peak at 935.3 eV could be ascribed to the

Cu2+ chemical shift in copper nitrate [39]. The peak at 933.0 may be attributed to the chemical shift of

Cu2+ in copper hydroxyl nitrate, which was formed during drying at 100 oC (Fig. S4). With increasing X

from 0 to 0.25, two peaks merged into one peak centered at 934.7 eV, which is 0.2 eV higher than that

of copper oleate. Further increment of X to 1.0 resulted in the decrease of binding energy to 934.5,

which is the binding energy of copper oleate. The XPS results postulated that the particles of copper

nitrate species were covered by a layer of copper oleate compounds at X≧0.25 and this corresponds

well with the FT-IR results. The XPS results of uncalcined 5%Ni/SiO2 samples also drew a similar

conclusion with that of uncalcined 10%Cu/SiO2 samples (Fig. S7).

3.7 EXAFS results

0 1 2 3 4 5 6 7 8

X=1.0

1.2 1.3 1 .4 1 .5 1 .6 1 .7 1 .8 1 .9

R (a ngs trom )

1.53

1.52

1.50

1.43 Cu-O

FT

ma

gn

itud

e

R(Angstrom)

Cu-O

Cu-Si

2.72

X=0

X=0.25

Copper oleate

Fig. 9. EXAFS of uncalcined 10%Cu/SiO2/X and copper oleate samples.

16

Extended X-ray Absorption Fine Structure (EXAFS) was used to characterize the interaction between

copper nitrate species and silica support on the samples of 10%Cu/SiO2/X dried at 100 oC (Fig. 9). The

Cu-O distance shifted from 1.53 Å to 1.50 Å as X was increased from 0 to 1.0. In contrast, copper

oleate had a Cu-O bond distance of 1.42Å (Fig. 9 inset). The EXAFS results revealed that most of the

copper atoms on 10%Cu/SiO2/1.0 did not form copper oleate compound although copper oleate was the

only compound as detected by XPS (Fig. 8). Therefore, EXAFS and XPS results reveal that core-shell

particles would in situ self-assemble on 10%Cu/SiO2/X (X ≧ 0.25) with copper nitrate species as core

and copper oleate compounds as shell. Other important fact which could be derived from Fig. 9 is that

the peak intensity at 2.72 Å decreased when X was increased to 0.25 but completely vanished when X

was further increased to 1.0. The peak at 2.72 Å could be ascribed to Cu-Si distance, indicating that

Cu(Ⅱ) was bonded to silica surface [40]. Therefore, as X increased, the interaction between copper

nitrate species decreased.

Fig. 10. The proposed promotion mechanism of oleic acid on highly dispersed supported metal

catalysts preparation.

17

From XRD, TEM (STEM), H2-TPR, SEM-EDS, FT-IR, XPS and EXAFS results, the promotional

effect of OA was proposed in Scheme 1. In the case of X=0 (Fig. 10 (a)), although the particles interact

with the silica via hydrogen bond or coordination bond, the metal nitrate species form big particles and

distribute inhomogeneously. Druing thermal treatment, the remaining big metal nitrate species particles

would further agglomerate to form big metal oxide particles. If an appropriate amount of OA was added,

core-shell (metal oleate layer on metal nitrate species) particles would in situ self-assemble and interact

weakly with the support (Fig. 10 (b)). The agglomeration of metal nitrate species particles during

calcination or drying process was hindered by OA due to its long aliphatic carbon chain that poses a

strong steric hindrance – an important factor required for nanocrystal synthesis. However, too much OA

imposed a negative effect to the metal crystalline size as shown in XRD results for both copper and

nickel catalysts. As shown in Fig. 10 (c), in the presence of excess OA, OA would assemble on the

hydrophilic surface of silica via hydrogen bond with carboxylic group and form ester as the temperature

increased to > 100 oC [36]. Then, core-shell particles (metal oleate as the shell structure) would be

uplifted and floated on the hydrophobic layer of OA. As a consequence, the metal nitrate species would

distance away from the support surface, thereby decreasing the interaction with the support. This leads

to increased mobility of the particles on OA layer which would easily migrate and combine to form

bigger particles during calcination. However, another possible reason which could not be ruled out was

that excess OA could occupy the pores in the support, and hinder the diffusion of metal nitrate into the

pores, which would then agglomerate out of the pores. The 10%Cu/SiO2 catalysts were sensitive to X

but not for 5%Ni/SiO2 catalysts. This could be attributed to the strong interaction between nickel oxide

and silica compared to that of copper oxide and silica, leading to easier formation of more stable nickel

silicates in the former case [23, 24]. Therefore, high dispersion of nickel could be achieved in a wide

range of nOA/nNi ratio from 0.05 to 1.5 (Fig. S3).

18

Fig. 11. Relationship between X, crystalline size of Cu and catalytic activities for WGSR.

3.8 WGSR activities on 10%Cu/SiO2 catalysts

The water-gas shift reaction (WGSR, CO + H2O → H2 + CO2) is important in upgrading H2-rich fuel

gas streams for fuel cell and other applications [42-45]. Water gas shift reaction was used as a probe

reaction to test the catalytic performances of 10%Cu/SiO2 catalysts prepared in the presence/absence of

OA. Fig. 11 describes the relationships of catalytic activity (in terms of CO conversion), X and

crystalline size of copper (dCu). As X is increased from 0 to 0.05, CO conversion (XCO) improved

drastically from ~20% to ~37%. However, upon a further increase in X from 0.25 to 1.0, XCO decreased

from ~47% to ~38%. Thus, it can be observed that the catalytic activities of 10%Cu/SiO2/X catalysts

can be well correlated with dCu, i.e. smaller dCu gives higher WGSR activity.

3.9 DRM activities on 5%Ni/SiO2 catalysts and spent catalysts characterization

19

Fig. 12. The catalytic activities of 5%Ni/SiO2/0.5 (square symbol) and 5%Ni/SiO2/0 (triangle symbol)

for DRM reaction.

Dry reforming of methane (DRM) using carbon dioxide is considered to be an important reaction for

both the chemical industry and environment since it utilizes both greenhouse gases whilst producing the

valuable chemical precursor syngas [46-51]. Ni/SiO2 catalyst with high metal dispersion on silica

support can present itself as a good catalyst for DRM [24, 26, 52]. Therefore, the DRM reaction was

chosen as a probe reaction to test the Ni/SiO2 catalyst prepared with our methodology. Fig. 12 displays

the catalytic activities of 5%Ni/SiO2 catalysts with or without OA for at 700 oC with GHSV = 72000

ml.g(cat)-1.h-1. The catalyst of 5%Ni/SiO2/0.5 was relatively stable for 100 hours of reaction on stream,

while the conversions of CH4 and CO2 dropped slightly from ~65% and ~75% to ~60% and ~70%,

respectively. In contarast, rapid deactivation was observed with the 5%Ni/SiO2 catalyst without OA and

carbon deposition blocked the catalyst bed after one hour of reaction.

20

Fig. 13. TEM images of the spent 5%Ni/SiO2/0 (left) and 5%Ni/SiO2/0.5 (right) catalysts used for

DRM reaction.

The spent catalyst of 5%Ni/SiO2/0 and 5%Ni/SiO2/0.5 were analyzed by DTA-TGA. The carbon

deposition rate was as high as 76.9 mg/g(cat.).h on the spent 5%Ni/SiO2/0 catalyst and the carbon

combustion temperature was ~660 oC (Fig. S8). The carbon morphology on the spent 5%Ni/SiO2/0

catalyst was characterized by TEM (Fig. 13, left). Carbon nanotubes were formed on the spent

5%Ni/SiO2/0 catalyst. Therefore, the deactivation of catalyst of 5%Ni/SiO2/0 was clearly due to its

severe carbon deposition. On the other hand, negligible carbon was detected on the spent

5%Ni/SiO2/0.5 catalyst as measured by DTA-TGA (Fig. S9) and TEM (Fig. 13 (right)) despite

observation of several long carbon nanotubes on the spent catalyst by STEM (Fig. S10). These

nanotubes could have been produced by the unsupported nickel particles or by those particles which had

less interaction with the support. The spent catalyst had a mean nickel particle size of 3.8 nm (Fig. S10),

which was marginally bigger than that of the freshly reduced sample. The spent 5%Ni/SiO2/0.5 catalyst

was analyzed by XRD and the crystalline size of nickel increased slightly from 2.7 nm to 3.9 nm, which

showed that the nickel particles were highly stable in catalytic activity during the 100 hours of reaction.

21

Highly nickel dispersion of 5%Ni/SiO2/0.5 catalyst should be the major attribute for low carbon

formation on the catalyst [24, 26, 53].

4. Conclusions

A novel, simple and large-scale impregnation method has been successfully devised to prepare highly

dispersed supported metal catalysts via in-situ self-assembled core-shell precursor route for sustainable

hydrogen production. The core-shell precursors formed easily through in-situ self-assembly during the

impregnation of metal precursor with a small amount of OA (oleic acid). The carboxylic group of OA

chelated to metal ion as bidentate type oleate outside of metal nitrate species particles (core-shell)

prevented the particles of metal nitrate species and metal oxide from agglomeration during thermal

treatment at high temperature. High catalytic activities and stabilities were exhibited by the catalysts

prepared in the presence of OA as compared to the catalysts prepared with conventional impregnation.

The novel method reported in this communication can also be used effectively to prepare metal catalysts

besides copper and nickel metals (Fig. S11, S12). Besides, it can also be used to prepare mesoporous

material (i.e. SBA-15 (Fig. S13)) supported metal catalysts, supported bi-metallic catalysts or metal

catalysts doped with metal oxides. Utilization of other metal oxides, like ceria, alumina etc, to support

metals are currently explored.

Acknowledgements

The authors gratefully thank the National University of Singapore and NEA (NEA-ETRP Grant

No.1002114, RP No. 279-000-333-490) for generously supporting this work.

22

Reference

[1] C. Aydin, J. Lu, N.D. Browning, B.C. Gates, A “Smart” Catalyst: Sinter-Resistant Supported

Iridium Clusters Visualized with Electron Microscopy, Angew. Chem. Int. Ed., 51 (2012) 5929-5934.

[2] B.C. Gates, Supported Metal Clusters: Synthesis, Structure, and Catalysis, Chem. Rev., 95 (1995)

511-522.

[3] X. Wu, S. Kawi, Steam reforming of ethanol to H2 over Rh/Y2O3: crucial roles of Y2O3 oxidizing

ability, space velocity, and H2/C, Energy Environ. Sci., 3 (2010) 334-342.

[4] Y. Dai, Y. Wang, B. Liu, Y. Yang, Metallic Nanocatalysis: An Accelerating Seamless Integration

with Nanotechnology, Small, 11 (2015) 268-289.

[5] J. Liu, S. Zou, S. Li, X. Liao, Y. Hong, L. Xiao, J. Fan, A general synthesis of mesoporous metal

oxides with well-dispersed metal nanoparticles via a versatile sol-gel process, J. Mater. Chem. A, 1

(2013) 4038-4047.

[6] W.S. Xia, Y.H. Hou, G. Chang, W.Z. Weng, G.B. Han, H.L. Wan, Partial oxidation of methane into

syngas (H2 + CO) over effective high-dispersed Ni/SiO2 catalysts synthesized by a sol–gel method, Int.

J. Hydrogen Energ, 37 (2012) 8343-8353.

[7] U. Oemar, M.L. Ang, W.F. Hee, K. Hidajat, S. Kawi, Perovskite LaxM1−xNi0.8Fe0.2O3 catalyst for

steam reforming of toluene: Crucial role of alkaline earth metal at low steam condition, Appl. Catal. B,

148–149 (2014) 231-242.

[8] U. Oemar, P.S. Ang, K. Hidajat, S. Kawi, Promotional effect of Fe on perovskite LaNixFe1−xO3

catalyst for hydrogen production via steam reforming of toluene, Int. J. Hydrogen Energ, 38 (2013)

5525-5534.

[9] U. Oemar, A. Ming Li, K. Hidajat, S. Kawi, Mechanism and kinetic modeling for steam reforming

of toluene on La0.8Sr0.2Ni0.8Fe0.2O3 catalyst, AIChE J., 60 (2014) 4190-4198.

[10] S. Park, Y. Bang, S. J. Han, J. Yoo, J. H. Song, J. C. Song, J. Lee, I. K. Song, Hydrogen production

by steam reforming of liquefied natural gas (LNG) over mesoporous nickel–iron–alumina catalyst, Inter.

J. Hydrogen Energ., 40( 2015) 5869–5877.

23

[11] Y. Zhang, H. Yang, J. Zhou, Z. Wang, J. Liu, K. Cen, Catalytic decomposition of sulfuric acid over

CuO/CeO2 in the sulfur–iodine cycle for hydrogen production, Inter. J. Hydrogen Energ., 40(2015)

2099–2106.[12] K.Y. Koo, U.H. Jung, W.L. Yoon, A highly dispersed Pt/γ-Al2O3 catalyst prepared via

deposition–precipitation method for preferential CO oxidation, Int. J. Hydrogen Energ, 39 (2014) 5696-

5703.

[13] J. Ashok, Y. Kathiraser, M.L. Ang, S. Kawi, Bi-functional Hydrotalcite-Derived NiO-CaO-Al2O3

Catalysts for Steam Reforming of Biomass and/or Tar Model Compound at Low Steam-to-Carbon

Conditions, Appl. Catal. B., 172(2015) 116-128.

[14] J. Ashok, S. Kawi, Steam reforming of toluene as a biomass tar model compound over CeO2

promoted Ni/CaO–Al2O3 catalytic systems, Int. J. Hydrogen Energ, 38 (2013) 13938-13949.

[15] J. Ashok, S. Kawi, Steam reforming of biomass tar model compound at relatively low steam-to-

carbon condition over CaO-doped nickel–iron alloy supported over iron–alumina catalysts, Appl. Catal.

A, 490 (2015) 24-35.

[16] A.G. Bhavani, W.Y. Kim, J.W. Lee, J.S. Lee, Influence of Metal Particle Size on Oxidative CO2

Reforming of Methane over Supported Nickel Catalysts: Effects of Second-Metal Addition,

ChemCatChem, 7 (2015) 1445-1452.

[17] L. Li, D.H. Anjum, H. Zhu, Y. Saih, P.V. Laveille, L. D'Souza, J.-M. Basset, Synergetic Effects

Leading to Coke-Resistant NiCo Bimetallic Catalysts for Dry Reforming of Methane, ChemCatChem, 7

(2015) 427-433.

[18] A. Djaidja, H. Messaoudi, D. Kaddeche, A. Barama, Study of Ni–M/MgO and Ni–M–Mg/Al

(M=Fe or Cu) catalysts in the CH4–CO2 and CH4–H2O reforming, Inter. J. Hydrogen Energ., 40(2015),

4989–4995.

[19] K.Y. Koo, U.H. Jung, W.L. Yoon, A highly dispersed Pt/γ-Al2O3 catalyst prepared via deposition–

precipitation method for preferential CO oxidation, Inter. J. Hydrogen Energ., 39(2014), 5696–5703.[20]

L. Jiao, J. R. Regalbuto, The synthesis of highly dispersed noble and base metals on silica via strong

24

electrostatic adsorption: I. Amorphous silica, J. Catal. 260 (2008) 329–341.[21] S. Zhu, S. Wang, L.

Jiang, Z. Xia, H. Sun, G. Sun, High Pt utilization catalyst prepared by ion exchange method for direct

methanol fuel cells, Inter. J Hydrogen Energ., 37 (2012), 14543–14548.

[22] E. Marceau, M. Che, J. Čejka, A. Zukal, Nickel(II) Nitrate vs. Acetate: Influence of the Precursor

on the Structure and Reducibility of Ni/MCM-41 and Ni/Al-MCM-41 Catalysts, ChemCatChem, 2

(2010) 413-422.

[23] D.J. Lensveld, J. Gerbrand Mesu, A. Jos van Dillen, K.P. de Jong, Synthesis and characterisation

of MCM-41 supported nickel oxide catalysts, Microporous and Mesoporous Materials, 44–45 (2001)

401-407.

[24] S. He, H. Wu, W. Yu, L. Mo, H. Lou, X. Zheng, Combination of CO2 reforming and partial

oxidation of methane to produce syngas over Ni/SiO2 and Ni–Al2O3/SiO2 catalysts with different

precursors, Int. J. Hydrogen Energ, 34 (2009) 839-843.

[25] J.C. Kenvin, M.G. White, Supported catalysts prepared from mononuclear copper complexes:

Catalytic properties, J. Catal., 135 (1992) 81-91.

[26] X.-Y. Quek, D. Liu, W.N.E. Cheo, H. Wang, Y. Chen, Y. Yang, Nickel-grafted TUD-1

mesoporous catalysts for carbon dioxide reforming of methane, Appl. Catal. B, 95 (2010) 374-382.

[27] D. Baudouin, K.C. Szeto, P. Laurent, A.D. Mallmann, B. Fenet, L. Veyre, U. Rodemerck, C.

Coperet, C. Thieuleux, Nickel–Silicide Colloid Prepared under Mild Conditions as a Versatile Ni

Precursor for More Efficient CO2 Reforming of CH4 Catalysts, J. Am. Chem. Soc., 134 (2012) 20624-

20627.

[28] L. Li, E. Abou-Hamad, D.H. Anjum, L. Zhou, P.V. Laveille, L. Emsley, J.-M. Basset, Well-

defined mono([small eta]3-allyl)nickel complex [triple bond, length as m-dash]MONi([small eta]3-

C3H5) (M = Si or Al) grafted onto silica or alumina: a molecularly dispersed nickel precursor for

syntheses of supported small size nickel nanoparticles, Chem. Commun., 50 (2014) 7716-7719.

25

[29] P. Munnik, M. Wolters, A. Gabrielsson, S.D. Pollington, G. Headdock, J.H. Bitter, P.E.d. Jongh,

K.P.d. Jong, Copper Nitrate Redispersion To Arrive at Highly Active Silica-Supported Copper

Catalysts, J. Phys. Chem. C, 115 (2011) 14698-14706.

[30] J.R.A. Sietsma, J.D. Meeldijk, J.P.d. Breejen, M. Versluijs-Helder, A.J.v. Dillen, P.E.d. Jongh,

K.P.d. Jong, The Preparation of Supported NiO and Co3O4 Nanoparticles by the Nitric Oxide

Controlled Thermal Decomposition of Nitrates, Angew. Chem. Int. Ed., 46 (2007) 4547-4549.

[31] N. Zheng, G.D. Stucky, A General Synthetic Strategy for Oxide-Supported Metal Nanoparticle

Catalysts, J. Am. Chem. Soc., 128 (2006) 14278-14280.

[32] L. Mo, S. Kawi, An in situ self-assembled core-shell precursor route to prepare ultrasmall copper

nanoparticles on silica catalysts, J. Mater. Chem. A, 2 (2014) 7837-7844.

[33] L. Mo, K.K.M. Leong, S. Kawi, A highly dispersed and anti-coking Ni-La2O3/SiO2 catalyst for

syngas production from dry carbon dioxide reforming of methane, Catal. Sci. Technol., 4 (2014) 2107-

2114.

[34] J. Park, K. An, Y. Hwang, J.-G. Park, H.-J. Noh, J.-Y. Kim, J.-H. Park, N.-M. Hwang, T. Hyeon,

Ultra-large-scale syntheses of monodisperse nanocrystals Nature Mater., 3 (2004) 891-895.

[35] A.J. Marchi, J.L.G. Fierro, J. Santamaría, A. Monzón, Dehydrogenation of isopropylic alcohol on a

Cu/SiO2 catalyst: a study of the activity evolution and reactivation of the catalyst, Appl. Catal. A, 142

(1996) 375-386.

[36] K.V.R. Chary, K.K. Seela, G.V. Sagar, B. Sreedhar, Characterization and Reactivity of Niobia

Supported Copper Oxide Catalysts, J. Phys. Chem. B, 108 (2004) 658-663.

[37] N. Wu, L. Fu, M. Su, M. Aslam, K.C. Wong, V.P. Dravid, Interaction of Fatty Acid Monolayers

with Cobalt Nanoparticles, Nano Lett., 4 (2004) 383-386.

[38] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, in, Wiley,

New York, 1997, pp. 59.

[39] V.I. Nefedov, É.K. Zhumadilov, T.Y. Kopytova, Comparison of the chemical shifts of X-ray

electron and Auger lines, J. Struct. Chem., 18 (1977) 549-553.

26

[40] S.F. Cheah, G.E. Brown, Jr. and G. A. Parks, XAFS Spectroscopy Study of Cu(II) Sorption on

Amorphous SiO2 and gamma-Al2O3: Effect of Substrate and Time on Sorption Complexes, J. Colloid

Inter. Sci., 208 (1998) 110.

[41] K. Marshall, C.H. Rochester, Infrared study of the adsorption of oleic and linolenic acids onto the

surface of silica immersed in carbon tetrachloride, J. Chem. Soc., Faraday Trans. 1, 71 (1975) 1754-

1761.

[42] M.L. Ang, U. Oemar, E.T. Saw, L. Mo, Y. Kathiraser, B.H. Chia, S. Kawi, Highly Active

Ni/xNa/CeO2 Catalyst for the Water–Gas Shift Reaction: Effect of Sodium on Methane Suppression,

ACS Catal., 4 (2014) 3237-3248.

[43] E.T. Saw, U. Oemar, X.R. Tan, Y. Du, A. Borgna, K. Hidajat, S. Kawi, Bimetallic Ni–Cu catalyst

supported on CeO2 for high-temperature water–gas shift reaction: Methane suppression via enhanced

CO adsorption, J. Catal., 314 (2014) 32-46.

[44] Z. Fu, J. Wang, N. Zhang, Y. An, Z. Yang, Effect of Cu doping on the catalytic activity of Fe3O4 in

water-gas shift reactions, Int. J. Hydrogen Energ, 40 (2015) 2193-2198.

[45] R. Knapp, S.A. Wyrzgol, A. Jentys, J.A. Lercher, Water–gas shift catalysts based on ionic liquid

mediated supported Cu nanoparticles, J. Catal., 276 (2010) 280-291.

[46] J. Guo, H. Lou, H. Zhao, D. Chai, X. Zheng, Dry reforming of methane over nickel catalysts

supported on magnesium aluminate spinels, Appl. Catal. A, 273 (2004) 75-82.

[47] M.-S. Fan, A.Z. Abdullah, S. Bhatia, Utilization of Greenhouse Gases through Dry Reforming:

Screening of Nickel-Based Bimetallic Catalysts and Kinetic Studies, ChemSusChem, 4 (2011) 1643-

1653.

[48] Z. Li, Y. Kathiraser, J. Ashok, U. Oemar, S. Kawi, Simultaneous Tuning Porosity and Basicity of

Nickel@Nickel–Magnesium Phyllosilicate Core–Shell Catalysts for CO2 Reforming of CH4, Langmuir,

30 (2014) 14694-14705.

27

[49] Z. Li, Y. Kathiraser, S. Kawi, Facile Synthesis of High Surface Area Yolk–Shell Ni@Ni

Embedded SiO2 via Ni Phyllosilicate with Enhanced Performance for CO2 Reforming of CH4,

ChemCatChem, 7 (2015) 160-168.

[50] Z. Li, L. Mo, Y. Kathiraser, S. Kawi, Yolk–Satellite–Shell Structured Ni–Yolk@Ni@SiO2

Nanocomposite: Superb Catalyst toward Methane CO2 Reforming Reaction, ACS Catal., 4 (2014)

1526-1536.

[51] K. Sutthiumporn, T. Maneerung, Y. Kathiraser, S. Kawi, CO2 dry-reforming of methane over

La0.8Sr0.2Ni0.8M0.2O3 perovskite (M = Bi, Co, Cr, Cu, Fe): Roles of lattice oxygen on C–H activation

and carbon suppression, Int. J. Hydrogen Energ, 37 (2012) 11195-11207.

[52] Y.-X. Pan, P. Kuai, Y. Liu, Q. Ge, C.-J. Liu, Promotion effects of Ga2O3 on CO2 adsorption and

conversion over a SiO2-supported Ni catalyst, Energy Environ. Sci., 3 (2010) 1322-1325.

[53] C.-J. Liu, J. Ye, J. Jiang, Y. Pan, Progresses in the Preparation of Coke Resistant Ni-based Catalyst

for Steam and CO2 Reforming of Methane, ChemCatChem, 3 (2011) 529-541.