Preparation and stabilization of gold nanoparticles formedby in situ reduction of aqueous chloroaurate ionswithin surface-modified mesoporous silica

Anirban Ghosh a, Chitta Ranjan Patra a, Priyabrata Mukherjee a,Murali Sastry b,*, Rajiv Kumar a,*

a Catalysis Division, National Chemical Laboratory, Pune 411008, Indiab Materials Chemistry Division, National Chemical Laboratory, Pune 411008, India

Received 1 July 2002; received in revised form 14 October 2002; accepted 18 November 2002

Abstract

A convenient method for the synthesis and simultaneous stabilization of gold nanoparticles in organo-functionalized

MCM-41 is described. The silanol groups present on the surface of the mesoporous matrix reduce aqueous chloroaurate

ions, resulting in the formation of gold nanoparticles, which are subsequently anchored to the host matrix via organic

functionality. The method, in which propylamine-MCM-41 or propylthiol-MCM-41 is treated with a dilute aqueous

solution of HAuCl4 at ambient conditions, is quite simple and environmentally benign as there is no need to add any

external reducing agent. The nano-Au-MCM-41 hybrid materials were characterized by XRD, UV–Vis, TEM, XPS,

EDX and N2 adsorption measurements.

� 2002 Elsevier Science Inc. All rights reserved.

Keywords: Auto-reduction; Gold nanoparticle; MCM-41; Organo-functionalized mesoporous silica; Organo-inorganic nanocomposite

1. Introduction

The discovery of mesoporous materials such asMCM-41 [1,2] with pore sizes ranging from 2 to 10

nm has opened up research not only in the field of

catalysis [3] but in materials science [4,5] as well.

Surface-modified MCM-41 materials [6], in which

the surface of MCM-41 contains organic func-

tionality, such as propylamine, propylthiol, vinyl,

etc., are emerging as important hosts for metallic

and semiconductor nanoparticles to form noveladvanced hybrid materials [7–9].

Generally, surfactant molecules containing

polar head groups like –SH, NH2, etc. are used

to stabilize nanoparticles and to prevent their ag-

glomeration to get the quantum dots. Propyl-

amine- and propylthiol-functionalized MCM-41

materials are particularly important in the sense

that they have the polar head group generally re-quired to stabilize the nanoparticles. Further, they

provide a solid support of well-defined pores to

the nanoclusters. Therefore, these nanocomposites

*Corresponding authors. Fax: +91-20-589-3952/3761.

E-mail addresses: sastry@ems.ncl.res.in (M. Sastry), rajiv@

cata.ncl.res.in (R. Kumar).

1387-1811/02/$ - see front matter � 2002 Elsevier Science Inc. All rights reserved.doi:10.1016/S1387-1811(02)00626-1

Microporous and Mesoporous Materials 58 (2003) 201–211

www.elsevier.com/locate/micromeso

can be visualized as a three-layered entity, where

the outer sphere is the inorganic support, the

second sphere is the organic moiety attached to

the surface of MCM-41 and protruding inside the

pores, and the third sphere consists of the nano-

particles hooked to the polar head groups of theorganic moiety present inside the pores. These

three-layered nanocomposites may lead to some

interesting non-linear optical properties for future

opto-electronic applications [10], and they are

excellent materials for studying host–guest inter-

actions [11]. The stabilization of nanoparticles

inside the well-defined pores of MCM-41 materials

may lead to the formation of nanowires, which arethe building units for future nano-electronic cir-

cuits.

However, in most reports on the formation

and stabilization of metallic nanoparticles, except

those of Esumi et al. [12] using sugar balls and

Mukherjee et al. [13–15] using microorganisms, the

hosts are usually passive and do not actively par-

ticipate in the reduction of metal ions to formnanoparticles followed by their entrapment in the

host matrix. Some external methods, either physi-

cal (vapor deposition, lithography or radiolysis) or

chemical (reduction by borohydride, citrate etc.)

have to be employed for the formation of nano-

particles prior to their stabilization.

Realizing the immense importance of nano-

composites in various applications, we have re-cently developed a novel method for the synthesis

of gold nanoparticles by in situ reduction of

aqueous chloroaurate ions (AuCl�4 ) via silanol

groups present on the surface of amorphous silica

[16] or on the surfaces of organo-functionalized

MCM-41 [17]. Employing this method, we can

avoid the use of any external-reducing environ-

ment other than the host matrix. Now, we reportour detailed studies on the synthesis of gold

nanoparticles of controlled and uniform size

and their simultaneous stabilization using propyl-

amine- and propylthiol-functionalized MCM-41

materials in a single step. The organic functional

groups act as anchors for the nanoparticles

by binding them through covalent interactions,

thereby avoiding the requirement of external cap-ping agents like alkylamines or alkylthiols for

particle size control [16,18].

2. Experimental

Two different procedures were adopted for

the synthesis of siliceous MCM-41 (Si-MCM-41),

propylamine-modified MCM-41 (NH2-MCM-41)and propylthiol-modified MCM-41 (SH-MCM-

41) materials, as described in detail elsewhere [19,

20]. Tetraethyl orthosilicate (TEOS, Aldrich), 3-

aminopropyltrimethoxy silane (APTS, Aldrich), 3-

mercaptopropyltrimethoxy silane (MPTS, Aldrich)

and cetyltrimethylammomium bromide (CTABr,

SD Fine Chem, India) were used as received. In a

typical synthesis of organo-functionalized MCM-41 (NH2-MCM-41 or SH-MCM-41), the gel

composition was 1.0TEOS–yXPTS–0.21CTABr–0.32NaOH–75H2O–16MeOH [XPTS ¼ APTS orMPTS, y ¼ 0:17–0.4]. However, the samples pre-pared with y ¼ 0:4 were used in the majority of theexperiments. Methanol was used in the initial gel

mixture to reduce and control the fast hydrolysis

of XPTS. The reaction mixture was first stirred atroom temperature for 12 h, transferred to a 250

cm3 polypropylene bottle, and heated at 100 �C for36 h under static conditions and autogenous

pressure.

For the synthesis of siliceous MCM-41 (Si-

MCM-41) [19] two types of silica sources were

used, viz. sodium metasilicate (Na2SiO3, SD Fine

Chem, India) and fumed silica (SiO2, Aldrich),where the Na2SiO3:SiO2 ratio was 0.124:1. The

silica sources were suspended in a 25% aque-

ous solution of tetramethylammonium hydroxide

(TMAOH, Aldrich). This suspension was then

combined with an aqueous solution of CTABr.

The initial gel composition was 1.0SiO2–0.22

NaOH–0.1TMAOH–0.21CTABr–125H2O. The

synthesis gel was transferred to a 250 cm3 poly-propylene bottle and kept at 100 �C for 36 h. Allthe as-synthesized mesoporous samples were

washed with distilled water and acetone, and dried

at 95 �C for 4 h. The surfactant from the NH2-MCM-41 and SH-MCM-41 (1 g each) materials

was removed by solvent extraction with a mixture

of solvents containing 85 g of methanol and 2 g of

HCl (35.5%) under reflux conditions for 24 h [20].The as-prepared Si-MCM-41 material was calc-

ined in air at 540 �C for 12 h for complete removalof the template.

202 A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211

The Si-MCM-41, NH2-MCM-41 and SH-

MCM-41 materials were characterized by low-

angle X-ray diffraction (XRD) recorded on a

Rigaku D Max III VC instrument with Ni filtered

CuKa radiation (k ¼ 1:5404 �AA) in the 2h range of1.5–10� at a scan rate of 1�/min. The specific sur-face areas of the samples were determined by the

BET method through N2 adsorption at 77 K using

an Omnisorb CX-100 Coulter instrument. Prior to

the adsorption experiments, the samples were ac-

tivated at 150 �C for 6 h at 1:3� 10�2 Pa. The poresize distributions of the samples were computed by

the BJH model. Chemical analyses of the NH2-

MCM-41 and SH-MCM-41 materials were doneon a Carlo Erba EA1108 elemental analyzer.

Nano-Au-MCM-41 hybrid materials were

formed by in situ reduction of AuCl�4 ions as

follows: In a typical experiment, 1.0 g of the Si-

MCM-41, NH2-MCM-41 or SH-MCM-41 mate-

rial were separately treated with 100 cm3 of a 10�4

M HAuCl4 solution for 96 h. After the specified

duration, the samples were filtered, washed thor-oughly with copious amounts of water and ace-

tone and finally dried under vacuum at room

temperature. Thereafter, 0.5 g of the nano-Au–Si-

MCM-41, nano-Au–NH2-MCM-41 or nano-Au–

SH-MCM-41 material were separately stirred with

50 cm3 of distilled water for 12 h. The materials

thus obtained after aqueous treatment (designated

as Au–Si-MCM-41-w, Au–NH2-MCM-41-w andAu–SH-MCM-41-w, respectively) were filtered,

washed with water and dried under vacuum. All

the nano-Au-MCM-41 hybrid materials were

characterized by XRD, UV–Vis spectroscopy,

transmission electron microscopy (TEM), X-ray

photoelectron spectroscopy (XPS), energy-disper-

sive X-ray (EDX) analysis and N2 adsorption

measurements. Further, UV–Vis spectra of all thefiltrates obtained in the above experiments were

taken.

The UV–Vis spectra of the nano-Au-MCM-41

hybrid materials were recorded on a Shimadzu

UV-2101PC spectrophotometer operating in the

reflection mode at a resolution of 2 nm using bar-

ium sulphate as a standard for background cor-

rection. Further, UV–Vis spectra of all the filtratesobtained in the above experiments were taken on

the same instrument using distilled water as a

standard for background correction. In order to

estimate the size of the gold nanoparticles formed

by spontaneous reduction of chloroaurate ions,

XRD measurements of the nano-Au-MCM-41

hybrid materials were carried out on a Philips PW

1830 instrument operating at 40 kV voltage and acurrent of 30 mA with CuKa radiation betweenthe 2h range 30� and 45� at a scan rate of 1�/min.The nano-Au-MCM-41 samples were dispersed by

acetone on Holey carbon grids and TEM images

were scanned on a Jeol Model 1200 EX instrument

operated at an accelerating voltage of 100 kV. The

XPS measurements of the nano-Au-MCM-41 hy-

brid materials were carried out on a VG Micro-Tech ESCA 3000 instrument at a pressure better

than 1:3� 10�7 Pa. The general scan and C 1s,S 2p, N 1s, Si 2p and Au 4f core level spectra were

recorded with non-monochromatized MgKa ra-diation (photon energy ¼ 1253:6 eV) at a passenergy of 50 eV and an electron takeoff angle

(angle between electron emission direction and

surface plane) of 60�. The overall resolution of themeasurements was thus approximately 1 eV for the

XPS measurements. The core level spectra were

background-corrected using the Shirley algorithm

[21] and the chemically distinct species resolved

using a non-linear least squares procedure. The

core level binding energies (BE) were aligned with

respect to the Au 4f7=2 binding energy of 84 eV.

EDX analyses of the materials were carried out ona Kevex equipment attached to a Jeol JSM-5200

scanning microscope.

Catalytic hydrogenation reactions of different

olefinic substrates on the nano-Au-X-MCM-41

materials were performed in a 100 cm3 high-pres-

sure autoclave at a temperature of 100 �C and ahydrogen pressure of 4 MPa for 6 h at 400 rpm.

The reaction mixtures were analyzed by a Shima-dzu 17A series gas chromatograph containing a

capillary column (10% permethylated b-cyclodex-trin, 30 m� 0:32 mm� 0:25 lm film thickness)and a flame ionization detector.

3. Results and discussion

The porosity of all mesoporous samples was

evaluated via N2 adsorption. Fig. 1A and B show

A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211 203

typical N2 adsorption–desorption isotherms and

the corresponding pore size distribution curves

(insets) for the Si-MCM-41 and nano-Au–Si-

MCM-41 samples, respectively. All the samples

showed isotherms of type IV with inflection points

around P=P0 ¼ 0:3–0:45, which is characteristicof M41S type ordered mesoporous materials.

The samples exhibit complementary textural and

framework-confined mesoporosity, as evidenced

by the presence of two separate, well-defined hys-

teresis loops. One is in the P=P0 ¼ 0:3–0:45 regionand indicates framework-confined mesopores, and

the other one occurs at P=P0P 0:8 correspondingto capillary condensation in the inter-particle

pores [22]. The position of the inflection point inthe P=P0 ¼ 0:3–0:45 region depends on the dia-meter of the mesopores, and its sharpness indicates

the uniformity of the narrow pore size distribu-

tion. The specific BET surface areas, pore volumes

and average pore diameters calculated from the N2adsorption isotherms using the BJH model are

summarized in Table 1. After in situ reduction of

the AuCl�4 ions by the mesoporous samples, thenature of the N2 adsorption isotherms remained

the same, but a decrease of approximately 9%,

20% and 22% in the surface areas of nano-Au–Si-

MCM-41, nano-Au–NH2-MCM-41 and nano-

Au–SH-MCM-41, respectively, was observed. The

pore volumes of the materials also decreased by

approximately 8%, 18% and 22%, respectively,

after treatment with AuCl�4 . However, the averagepore diameters of the samples after entrapment of

Au nanoparticles did not change considerably,

indicating filling of part of the mesopores by gold

nanoparticles keeping the mesoporous structure

intact.

Results of elemental analyses of different NH2-

MCM-41 and SH-MCM-41 materials synthesized

with different TEOS: XPTS ratios, and EDX an-alyses of the corresponding nano-Au–NH2-MCM-

41 and nano-Au–SH-MCM-41 samples are given

in Table 2. From the table, it is evident that the

concentration of Au in the silicate matrices does

not significantly depend upon the population of

pendant –NH2 or –SH groups in the inner surfaces

of MCM-41.

After treatment of the Si-MCM-41, NH2-MCM-41 and SH-MCM-41 materials with AuCl�4ions for 96 h, it was observed that all materials had

attained a deep pink color, evidently due to the

presence of Au nanoparticles in the cavities of the

silicate matrix. Au nanoparticles, due to their

surface plasmon vibrations, have a characteristic

absorption band in the visible region of the elec-

tromagnetic spectrum at around 520–550 nm,which is responsible for the striking violet to pink

range of colors of the nanoparticles depending

Fig. 1. N2 adsorption–desorption isotherms and corresponding

pore size distribution curves (insets) for (A) Si-MCM-41 and

(B) nano-Au–Si-MCM-41 samples.

204 A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211

upon the particle size [23,24]. Fig. 2A shows theUV–Vis spectra recorded on the (a) parent Si-

MCM-41, (b) nano-Au–Si-MCM-41, and (c) Au–

Si-MCM-41-w materials. Fig. 2B presents the

UV–Vis spectra of (a) parent NH2-MCM-41,

(b) nano-Au–NH2-MCM-41, and (c) Au–NH2-

MCM-41-w. Fig. 2C shows the UV–Vis spectra of

the (a) parent SH-MCM-41, (b) nano-Au–SH-

MCM-41, and (c) Au–SH-MCM-41-w. A strongabsorption at approxmately 540 nm is observed

for all the mesoporous materials after treatment

with HAuCl4 solution and is a clear indication of

reduction of the AuCl�4 ions by all the three ma-

terials, viz. Si-MCM-41 (Fig. 2A, curve b), NH2-

MCM-41 (Fig. 2B, curve b), and SH-MCM-41

(Fig. 2C, curve b). This resonance is absent in the

parent Si-MCM-41 (Fig. 2A, curve a), NH2-MCM-41 (Fig. 2B, curve a) and SH-MCM-41

(Fig. 2C, curve a) materials as expected. An in-

teresting observation is the presence of an addi-

tional resonance at approximately 725 and 710 nm

in the case of the nano-Au–NH2-MCM-41 (Fig.

2B, curve b) and nano-Au–SH-MCM-41 (Fig. 2C,

curve b) materials, respectively. This feature arisesdue to excitation of longitudinal surface plasmon

vibrations due to close packing of the gold nano-

particles [25,26] within the cavities of these mate-

rials, which shows from anchoring of the gold

nanoparticles to the inner surface of the siliceous

matrices by the amine or thiol groups. It is known

that primary amines and thiols covalently bind to

gold nanoparticles [18], and may be the bindingmode in our study also.

To assess the stability of the gold nanoparticles

formed by in situ reduction within the MCM-41

matrix, the nano-Au–Si-MCM-41, nano-Au–NH2-

MCM-41 and nano-Au–SH-MCM-41 materials

were separately treated with distilled water for 12

h, filtered and dried. The UV–Vis spectra of the

samples after aqueous treatment, namely Au–Si-MCM-41-w (curve c), Au–NH2-MCM-41-w

(curve c) and Au–SH-MCM-41-w (curve c) are

shown in Fig. 2A–C, respectively. The character-

istic absorption at approximately 540 nm of gold

nanoparticles happens to be missing in Au–Si-

MCM-41-w (Fig. 2A, curve c), clearly indicating

Table 1

Pore diameters, pore volumes, BET surface areas, d1 0 0 spacings, unit cell parameters (a0) and framework thicknesses (FWT) of theMCM-41 samples

Sample Pore diameter (�AA) Pore volume (cm3 g�1) BET SA (m2 g�1) d1 0 0 (�AA) a0a (�AA) FWTb (�AA)

Si-MCM-41 36.61 1.14 1022 42.64 49.24 12.63

Nano-Au–Si-MCM-41 36.55 1.05 927 42.43 48.99 12.44

NH2-MCM-41 35.77 0.96 747 42.03 48.53 12.76

Nano-Au–NH2-MCM-41 35.69 0.79 596 41.83 48.30 12.61

SH-MCM-41 33.79 0.91 683 40.11 46.31 12.52

Nano-Au–SH-MCM-41 33.65 0.71 535 39.94 46.12 12.47

a a0 ¼ 2d1 0 0=ffiffiffi

3p.

b FWT ¼ a0 � pore diameter.

Table 2

Results of elemental analyses, and mean diameters of Au nanoparticles in different nano-Au-X-MCM-41 samples

Sample XPTSa:TEOS N (mmol g�1) S (mmol g�1) Au (wt.%) Mean diameter of Au nanoparticlesb (nm)

NH2-MCM-41 0.17 1.41 – 3.1 3.4� 0.50.25 2.06 – 3.2 3.4� 0.50.4 3.24 – 3.1 3.4� 0.5

SH-MCM-41 0.17 – 1.37 3.4 3.2� 0.50.25 – 2.01 3.5 3.2� 0.50.4 – 3.21 3.2 3.2� 0.5

aXPTS ¼ APTS or MPTS.bCalculated from the Debye–Scherrer equation.

A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211 205

that almost all the gold nanoparticles were leached

out into the aqueous phase. This is further

supported by the UV–Vis spectra of the filtrate

obtained from the Si-MCM-41 material after

treatment with HAuCl4 for 96 h (Fig. 3A, curve a)

and that of the filtrate obtained from the nano-Au–Si-MCM-41 material after treatment with

distilled water (Fig. 3A, curve b). The character-

istic absorbance of Au nanoparticles at approxi-

mately 520 nm was observed for both filtrates.

Comparing the UV–Vis spectra of the nano-Au–

Si-MCM-41 material (Fig. 2A, curves b and c), it

can be inferred that although the entrapment of all

the gold nanoparticles in the cavities of Si-MCM-

41 was not complete, the leaching of the goldnanoparticles during washing was almost com-

Fig. 2. UV–Vis spectra recorded on (A) (a) Si-MCM-41, (b)

nano-Au–Si-MCM-41, and (c) Au–Si-MCM-41-w materials;

(B) (a) NH2-MCM-41, (b) nano-Au–NH2-MCM-41, and (c)

Au–NH2-MCM-41-w materials; and (C) (a) SH-MCM-41, (b)

nano-Au–SH-MCM-41, and (c) Au–SH-MCM-41-w materials.

Fig. 3. UV–Vis spectra of the filtrates obtained from (A) (a) Si-

MCM-41 after treatment with HAuCl4 solution, and (b)

nano-Au–Si-MCM-41 after treatment with water; (B) (a)

NH2-MCM-41 after treatment with HAuCl4 solution, and (b)

nano-Au–NH2-MCM-41 after treatment with water; and (C)

SH-MCM-41 after treatment with HAuCl4 solution, and (b)

nano-Au–SH-MCM-41 after treatment with water.

206 A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211

plete. Moreover, a weak absorbance at approxi-

mately 520 nm for the filtrate obtained after aque-

ous treatment of the nano-Au–Si-MCM-41material

(Fig. 3A, curve b) is further indicative of the

leaching of the gold nanoparticles in the aqueous

phase, which was previously inferred from the UV–Vis spectra of the Au–Si-MCM-41-w material (Fig.

2A, curve c). The UV–Vis spectra of the filtrates

obtained from the NH2-MCM-41 (Fig. 3B, curve a)

and SH-MCM-41 (Fig. 3C, curve a) materials after

treatment with HAuCl4, and those of the filtrates

obtained from the nano-Au–NH2-MCM-41 (Fig.

3B, curve b) and nano-Au–SH-MCM-41 (Fig. 3C,

curve b) materials after aqueous treatment did notshow any absorbance in this visible region, clearly

indicating that there is no leaching of Au nano-

particles in the case of NH2-MCM-41 and SH-

MCM-41 materials. Further, the UV–Vis spectra of

the Au–NH2-MCM-41-w (Fig. 2B, curve c) and

Au–SH-MCM-41-w (Fig. 2C, curve c) materials,

which are almost similar to those of the nano-

Au–NH2-MCM-41 (Fig. 2B, curve b) and nano-Au–SH-MCM-41 (Fig. 2C, curve b) materials,

respectively, including the additional absorbance at

approximately 725 and 710 nm, respectively, indi-

cate the stabilization of the close-packing of gold

nanoparticles in the open, string-like networks in-

side NH2-MCM-41 and SH-MCM-41.

Since all the organo-functionalized MCM-41

materials were synthesized by the ‘‘one-pot co-condensation’’ method, which leads to grafting of

all the organic functional groups on the inner

surfaces of the mesopores [27], quite obviously the

Au nanoparticles would be stabilized within the

mesopores only, and not on the outer surfaces, as

there is no binding moiety present on the outer

surfaces. So we can conclusively say that the Au

nanoparticles are present inside the pores of theorgano-functionalized MCM-41 materials, which

was previously inferred from the decrease in sur-

face area and pore volume of the materials.

Fig. 4 shows the low-angle XRD patterns of (a)

NH2-MCM-41, (b) nano-Au–NH2-MCM-41, (c)

SH-MCM-41, (d) nano-Au–SH-MCM-41, (e) Si-

MCM-41, and (f) nano-Au–Si-MCM-41. The

main (1 0 0) Bragg reflection is visible in all thematerials along with the weak (1 1 0), (2 0 0) and

(2 1 0) reflections, indicating a high degree of or-

dered hexagonal mesopores even after in situ for-

mation of gold nanoparticles within the cavities.

The d1 0 0 values along with the corresponding unitcell parameter (a0) of the different MCM-41 sam-ples are given in Table 1. The a0 values arecalculated by the equation a0 ¼ 2d1 0 0=

ffiffiffi

3p. To

evaluate the average size of the gold nanoparticles,

XRD measurements of the nano-Au–NH2-MCM-41 and nano-Au–SH-MCM-41 were carried out in

the 2h range from 30� to 45� with a scan rate of 1�/min. The diffraction patterns in the region of the

(1 1 1) Bragg reflection (2h ¼ 38:2�) are shown inthe inset of Fig. 4, the solid lines being the Lo-

rentzian fits to the respective (1 1 1) reflections.

Under the experimental conditions of the XRD

Fig. 4. XRD patterns of (a) NH2-MCM-41, (b) nano-Au–

NH2-MCM-41, (c) SH-MCM-41, (d) nano-Au–SH-MCM-41,

(e) Si-MCM-41, and (f) nano-Au–Si-MCM-41. Inset shows the

diffraction patterns of (a) nano-Au–SH-MCM-41 and (b) nano-

Au–NH2-MCM-41 in the region of the Au (1 1 1) Bragg re-

flection at 2h ¼ 38:2�, the solid lines being the Lorentzian fit tothe respective curves.

A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211 207

measurement (small slit width and monochroma-

tized CuKa radiation), it was felt that the XRDline-shape would be better approximated by a

Lorentzian than a Gaussian. Indeed, poorer chi-

square error values (which denote the goodness of

fit) were observed while fitting the (1 1 1) Braggreflection of gold to a Gaussian line-shape. A

similar pattern and peak width was obtained for

the nano-Au–Si-MCM-41 material (not shown).

The mean diameter of the gold nanoparticles was

estimated from the broadening of the (1 1 1) peak

using the Debye–Scherrer equation [28], which

yielded the values of approximately 3:4� 0:5 nmfor nano-Au–NH2-MCM-41 (inset of Fig. 4, curveb) and approximately 3:2� 0:5 nm for nano-Au–SH-MCM-41 (inset of Fig. 4, curve a), respec-

tively. These values are in good agreement with the

maximum possible particle size that can be ac-

commodated within the mesopores of the NH2-

MCM-41 and SH-MCM-41 materials, the mean

pore diameters of which, as we recollect, are 3.6

and 3.4 nm, respectively. The average diameter ofthe Au nanoparticles does not depend upon the

concentration of pendant –NH2 or –SH groups, as

evident from the samples prepared with different

TEOS: XPTS ratios (Table 2).

The TEM images of the nano-Au–NH2-MCM-

41 sample are given in Fig. 5A and B, the inset of

Fig. 5A showing a selected area electron diffrac-

tion pattern. The parallel fringes due to the side-onview of the long pores of MCM-41 are well visible

in both images. These equidistant parallel fringes

are characteristic features of separate layers, the

addition of which results in the formation of a

bunch of layers. The electron diffraction pattern

of the sample exhibits well-defined hexagonal

maxima, further confirming the periodicity of

the structure. The entrapped Au nanoparticles arevisible in Fig. 5B (highlighted by arrows in the

figure), which shows that the spatial distribution of

the nanoparticles in the perpendicular direction to

the channels is consistent with the pore spacing.

A chemical characterization of the nano-Au–

SH-MCM-41 and nano-Au–NH2-MCM-41 mate-

rials was carried out using XPS. Fig. 6 shows the

Au 4f and N1s core levels from the nano-Au–NH2-MCM-41 material (A and B) as well as the

Au 4f and S 2p core levels recorded from the nano-

Au–SH-MCM-41 sample (C and D). Strong sig-

nals from the gold nanoparticles entrapped in the

mesoporous silica matrix can be seen in both

samples (Fig. 6A and C). These core levels could

be fit to a single spin–orbit pair indicating no

chemical shifts arising due to complexation with–NH2 and –SH pendant groups in the MCM-41

matrix. The S 2p signal from the nano-Au–SH-

MCM-41 sample (Fig. 6D) could be satisfactorily

fit to a single spin–orbit pair separated by 1.14 eV.

It is interesting to note the absence of an addi-

tional high binding energy component in the S 2p

spectrum at approximately 168 eV that is often

observed in self-assembled monolayers of alkane-thiols on silver alloy surfaces [29]. The fact that

this component is not observed indicates that the

Fig. 5. (A,B) Representative TEM images of nano-Au–NH2-MCM-41. The arrows in (B) identify gold nanoparticles within the pores

of the NH2-MCM-41 material. The inset of (A) shows a selected area electron diffraction pattern recorded from the same sample.

208 A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211

thiol groups are not oxidized to a sulfonate species

and that complexation with gold nanoparticles

stabilizes them. The N 1s signal from the nano-

Au–NH2-MCM-41 sample (Fig. 6B) shows the

presence of two distinct chemical species with

binding energies of 399 and 403.4 eV, respectively.

The low binding energy component is assigned tonitrogen in the pendant amine groups directly

bound to the entrapped gold nanoparticles, while

the higher binding energy component arises due to

electron emission from nitrogen in protonated

amine pendant groups. Similar binding ener-

gies have been observed in earlier studies on

Langmuir–Blodgett films of octadecylamine in a

partially ionized state [30]. These protonated am-monium groups cannot bind to the gold nano-

particle surface.

In our previous work, we have demonstrated

that silanol groups present on the surface of fumed

silica could reduce aqueous chloroaurate ions to

form gold nanoparticles [16], where the concen-

tration of gold nanoparticles on the silica surface

increased with increasing density of surface hy-

droxyl groups. The fumed silica surface subse-

quently provided a support for the nanoparticles

thus formed. However, the average size of the

nanoparticles was in the order of approximately

17� 0:5 nm. We also observed that the particlesize can be controlled by addition of alkylaminemolecules present in solution, the size being de-

pendent on the concentration of the alkylamine

[16]. This presumably occurred by ex situ forma-

tion of self-assembled monolayers of the alkyl-

amine molecules on the gold nanoparticles during

the initial nucleation stage of the nanoparticles.

The minimum particle obtained using alkylamine

was in the order of 10� 0:5 nm. However, in thiswork, addition of alkylamine molecules is not at

all necessary, since the –NH2 or –SH functionality

residing deep inside the mesoporous channels

serves the same purpose by trapping the gold

nanoparticles in situ, after the reduction of the

AuCl�4 ions by the silanol groups of the inner wall

of the channel, the exact mechanism of which

Fig. 6. (A) Au 4f and (B) N1s core level spectra recorded from the nano-Au–NH2-MCM-41 material; (C) Au4f and (D) S 2p core level

spectra recorded from the nano-Au–SH-MCM-41 material.

A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211 209

requires further work to be understood. The meanparticle sizes obtained here are far less than those

obtained earlier by fumed silica and alkylamines.

It is pertinent to mention here that the size of the

gold nanoparticles is of utmost importance in in-

fluencing their catalytic activity. Haruta [31] and

Bond and Thompson [32] have demonstrated in

recent years, that gold nanoparticles of an average

diameter ranging between 2 and 4 nm exhibit ex-cellent catalytic activity in a variety of reactions

including decomposition of trace amounts of am-

ines and conversion of CO to CO2. Hence, our

approach for the preparation and stabilization of

gold nanoparticles in the cavities of the ordered

MCM-41 type materials in the size range of ap-

proximately 3–4 nm offers distinct advantages in

exploiting the catalytic potential of these hybridmaterials, as has been briefly demonstrated in the

hydrogenation of styrene to ethylbenzene in our

recent short communication [17]. To evaluate the

catalytic potential of the nano-Au-MCM-41

hybrid materials, the hydrogenation of differ-

ent olefinic substrates was carried out in a high-

pressure autoclave. The results are summarized

in Table 3. From these results, it is evident thatthe nano-Au–NH2-MCM-41 and nano-Au–SH-

MCM-41 materials have fairly comparable cata-

lytic activities.

4. Conclusions

Gold nanoparticles can be synthesized and en-trapped within the pores of propylamine- and

propylthiol-functionalized MCM-41 materials by

auto-reduction of aqueous chloroaurate ions,which occurs via silanol groups present on the

inner surface of the mesopores. The Au-nano-X-

MCM-41 (where X ¼ –NH2 or –SH) materials are

examples for inorgano–organic nanocomposites

where the mesoporous matrix actively participates

in the reduction of the metal ions and acts as a

support for the nanoparticles through covalent

linkage by the pendant organic functionality,thereby perfectly accommodating them inside the

mesopores. The dimension of the pores in the

mesoporous material is in accordance with the size

range of the gold nanoparticles where they show

excellent catalytic activity. This makes the Au-

MCM-41 nano-sized hybrid materials quite excit-

ing for future applications in catalysis.

Acknowledgements

AG and CRP gratefully acknowledge the

Council of Scientific and Industrial Research,

Government of India for granting research fel-

lowships. The authors thank Drs. A.B. Mandale

and R. Parischa for their helpful co-operation incarrying out XPS and TEM experiments.

References

[1] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S.

Beck, Nature 359 (1992) 710.

[2] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T.

Kresge, K.D. Schmitt, C.T.-W. Chu, D.H. Olson, E.W.

Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, J.

Am. Chem. Soc. 114 (1992) 10834.

[3] A. Corma, Chem. Rev. 97 (1997) 2373.

Table 3

Results of catalytic hydrogenation of different olefinic substratesa by nano-Au-X-MCM-41 materials

Catalyst Substrate Conversion (mol%) Selectivityb (%)

Nano-Au–NH2-MCM-41 Cyclohexene 8.9 100

1-Hexene 18.5 100

1-Octene 13.5 100

Nano-Au–SH-MCM-41 Cyclohexene 9.7 100

1-Hexene 18.7 100

1-Octene 12.9 100

aReaction conditions: substrate ¼ 10 mmol; catalyst ¼ 200 mg, solvent ¼ toluene, 30 cm3; temperature ¼ 100 �C; H2 pressure ¼ 4MPa; reaction time ¼ 6 h.b Selectivity for cyclohexane, n-hexane and n-octane, respectively.

210 A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211

[4] Y. Sakamoto, M. Kaneda, O. Terasaki, D. Zhao, J.M.

Kim, G.D. Stucky, H.J. Shin, R. Ryoo, Nature 408 (2000)

449.

[5] Q. Huo, D.I. Margolese, G.D. Stucky, Chem. Mater. 8

(1996) 1147.

[6] X. Feng, G.E. Fryxell, L.-Q. Wang, A.Y. Kim, J. Liu,

K.M. Kemner, Science 276 (1997) 923.

[7] P. Mukherjee, M. Sastry, R. Kumar, Phys. Chem. Comm.

3 (2000) 15.

[8] Y. Xu, C.H. Langford, J. Phys. Chem. B 101 (1997) 3115.

[9] T. Hirai, H. Okubo, I. Komasawa, J. Phys. Chem. B 103

(1999) 4228.

[10] D.H. Gracias, J. Tien, T.L. Breen, C. Hsu, G.M. White-

sides, Science 289 (2000) 1170.

[11] K. Moller, T. Bein, Chem. Mater. 10 (1998) 2950.

[12] K. Esumi, T. Hosoya, A. Suzuki, K. Torigoe, Langmuir 16

(2000) 2978.

[13] P. Mukherjee, A. Ahmad, D. Mandal, S. Senapati, S.R.

Sainkar, M.I. Khan, R. Ramani, R. Parischa, P.V.

Ajayakumar, M. Alam, M. Sastry, R. Kumar, Angew.

Chem. Int., Ed. Engl. 40 (2001) 3585.

[14] P. Mukherjee, A. Ahmad, D. Mandal, S. Senapati, S.R.

Sainkar, M.I. Khan, R. Parischa, P.V. Ajayakumar, M.

Alam, R. Kumar, M. Sastry, Nano Lett. 1 (2001) 515.

[15] P. Mukherjee, S. Senapati, D. Mandal, A. Ahmad, M.I.

Khan, R. Kumar, M. Sastry, Chem. Bio. Chem. 3 (2002)

461.

[16] P. Mukherjee, C.R. Patra, A. Ghosh, M. Sastry, R.

Kumar, Chem. Mater. 14 (2002) 1678.

[17] P. Mukherjee, C.R. Patra, R. Kumar, M. Sastry, Phys.

Chem. Comm. 4 (2001) 24.

[18] D.V. Leff, L. Brandt, J.R. Heath, Langmuir 12 (1996)

4723.

[19] P. Mukherjee, R. Kumar, U. Schuchardt, in: L. Bonneviot,

F. B�eeland, C. Danumah, S. Giasson, S. Kaliaguine (Eds.),

Mesoporous Molecular Sieves, Studies in Surface Science

and Catalysis, vol. 117, Elsevier, Amsterdam, 1998, p. 351.

[20] P. Mukherjee, S. Laha, D. Mandal, R. Kumar, in: A.

Sayari, M. Jaroniec, T.J. Pinnavaia (Eds.), Nanoporous

Materials II, Studies in Surface Science and Catalysis, vol.

129, Elsevier, Amsterdam, 2000, p. 283.

[21] D.A. Shirley, Phys. Rev. B. 5 (1972) 4709.

[22] X. Chen, L. Huang, Q. Li, J. Phys. Chem. B. 101 (1997)

8460.

[23] S. Underwood, P. Mulvaney, Langmuir 10 (1994) 3427.

[24] S. Link, M.A. El-Sayed, J. Phys. Chem. B. 103 (1999)

4212.

[25] C.G. Blatchford, J.R. Campbell, J.A. Creighton, Surf. Sci.

120 (1982) 435.

[26] K.S. Mayya, V. Patil, M. Sastry, Langmuir 13 (1997) 3944.

[27] S.L. Burkett, S.D. Sims, S. Mann, Chem. Commun. (1996)

1367.

[28] R.C. Rau, Advances in X-ray Analysis, vol. 5, Plenum

Press, Inc., New York, 1962, p. 104.

[29] K. Bandyopadhyay, K.S. Mayya, K. Vijayamohanan, M.

Sastry, J. Electron. Spectrosc. Relat. Phenom. 87 (1997)

101.

[30] P. Ganguly, D.V. Paranjape, M. Sastry, J. Am. Chem. Soc.

115 (1993) 793.

[31] M. Haruta, Catal. Today 36 (1997) 153.

[32] G.C. Bond, D.T. Thompson, Catal. Rev. Sci. Eng. 41

(1999) 319.

A. Ghosh et al. / Microporous and Mesoporous Materials 58 (2003) 201–211 211