VOT 74284

Design and Application of Chiral Solid Catalysts Synthesized by Molecular Imprinting Method with Polyaminoacid as Chiral

Promoter for Producing Pharmaceutical Products

HADI NUR

Ibnu Sina Institute for Fundamental Science StudiesUniversiti Teknologi Malaysia

2008

ii

UNIVERSITI TEKNOLOGI MALAYSIAUTM/RMC/F/0024 (1998)

BORANG PENGESAHANLAPORAN AKHIR PENYELIDIKAN

TAJUK PROJEK : DESIGN AND APPLICATION OF CHIRAL SOLID CATALYSTS SYNTHESIZED BY MOLECULAR IMPRINTING METHOD WITH POLYAMINOACID AS CHIRAL PROMOTER FOR PRODUCING PHARMACEUTICAL PRODUCTS

Saya HADI NUR (HURUF BESAR)

Mengaku membenarkan Laporan Akhir Penyelidikan ini disimpan di Perpustakaan Universiti Teknologi Malaysia dengan syarat-syarat kegunaan seperti berikut :

1. Laporan Akhir Penyelidikan ini adalah hakmilik Universiti Teknologi Malaysia.

2. Perpustakaan Universiti Teknologi Malaysia dibenarkan membuat salinan untuktujuan rujukan sahaja.

3. Perpustakaan dibenarkan membuat penjualan salinan Laporan AkhirPenyelidikan ini bagi kategori TIDAK TERHAD.

4. * Sila tandakan ( / )

5.

SULIT (Mengandungi maklumat yang berdarjah keselamatan atauKepentingan Malaysia seperti yang termaktub di dalamAKTA RAHSIA RASMI 1972).

TERHAD (Mengandungi maklumat TERHAD yang telah ditentukan oleh Organisasi/badan di mana penyelidikan dijalankan).

TIDAKTERHAD

TANDATANGAN KETUA PENYELIDIK

Nama & Cop Ketua Penyelidik

Tarikh : _________________

v

CATATAN : * Jika Laporan Akhir Penyelidikan ini SULIT atau TERHAD, sila lampirkan surat daripada pihak berkuasa/organisasi berkenaan dengan menyatakan sekali sebab dan tempoh laporan ini perlu dikelaskan

Lampiran 20

10 October 2008

Hadi Nur

i

ACKNOWLEDGEMENT

These acknowledgements must begin with the Ministry of Science,

Technology and Innovation (MOSTI) for allowing grant to this project through IRPA

funding No. 09-02-06-10041-EAR (Vot. 74284). I am particularly grateful to the

Research Management Center (RMC), UTM for infrastructure, facilities and technical

support.

Innumerable thanks go to Prof. Dr. Halimaton Hamdan, Associate Prof. Dr.

Zainab Ramli, Associate Prof. Mohd. Nazlan Mohd. Muhid, Assoc. Prof. Dr. Salasiah

Endud, and Mr. Lim Kheng Wei for their valuable contributions. The responsiveness

and generosity of such a distinguished group of researchers transformed the arduous

task of bringing this project together into a particularly gratifying one. I wish to

express my sincerest appreciation to Ibnu Sina Institute for Fundamental Science

Studies for research facilities.

Assoc. Prof. Dr. Hadi NurProject Leader

ii

ABSTRACT

Heterogeneous asymmetric catalysis remains an exciting research field in chiral

catalysis since the heterogeneous catalyst can be separated easily from the reaction

mixture compared to conventional homogeneous catalyst. The aim of this research is

to develop a novel heterogeneous asymmetric catalyst using amino acid as chiral

promoter. The catalysts were synthesized by attachment of amino acids such as L-

glutamic acid, L-leucine and L-phenylalanine onto the hydrophilic part of hydrolyzed

octadecyltrichlorosilane (OTS). The short-range order structure of silicon and organic

groups in the catalysts has been characterized by solid-state 29Si and 13C cross-

polarization magic-angle-spinning (MAS) NMR spectroscopies, respectively. The

solid state MAS NMR results showed that the amino acids interacted with hydrolyzed

OTS. This phenomenon was supported by 13C NMR spectra which the peaks of L-

glutamic acid, L-leucine and L-phenylalanine were shifted towards a higher magnetic

field. It was confirmed by 29Si NMR spectra which showed the peaks of cross-linked

–(OH)Si(R)-O-(OH)Si(R)- (R=octadecyl group) whereas those of R–Si=(OSiR)3 were

not present in amino acid-hydrolyzed OTS. This result suggested that the amino acid

was attached via cross-linked –(OH)Si(R)-O-(OH)Si(R)- of the hydrolyzed OTS. The

amino acid-hydrolyzed OTS materials were used as catalysts for the asymmetric

hydration of epoxycyclohexene to yield two diastereoisomers, namely (1R,2R)-trans-

1,2-cyclohexanediol and (1S,2S)-trans-1,2-cyclohexenediol as well cis-1,2-

cyclohexenediol. The catalysts exhibit 12-18% enantiomeric excess (ee) for the

asymmetric hydration of epoxycyclohexene to yield (1R,2R)-trans-1,2-

cyclohexanediol and (1S,2S)-trans-1,2-cyclohexanediol.

iii

KEY RESEARCHERS

Assoc. Prof. Dr. Hadi Nur

Assoc. Prof. Dr. Salasiah Endud

Assoc. Prof. Dr. Zainab Ramli

Mr. Lim Kheng Wei

Email : hadi@ibnusina.utm.my

Tel. No. : 07-5536077

Vot No. : 74284

iv

ABSTRAK

Pemangkinan asimetri heterogen kekal sebagai satu bidang penyelidikan yang

mengujakan dalam pemankinan kiral memandangkan mangkin heterogen mudah

dipisahkan dari campuran tindak balas berbanding mangkin homogen yang

konventional. Tujuan penyelidikan ini adalah untuk menghasilkan satu mangkin

asimetri heterogen yang unggul menggunakan asid amino sebagai pemula kiral.

Mangkin disintesis dengan menjalinkan asid-asid amino seperti asid L-glutamik, L-

leusin dan L-fenilalanina pada bahagian hidrofilik oktadesiltriklorosilana (OTS)

terhidrolisis. Struktur jarak pendek silikon dan kumpulan-kumpulan organic pada

mangkin telah dicirikan menggunakan 29Si dan 13C polarisasi bersilang putaran sudut

ajaib (MAS) NMR spektroskopi keadaan pepejal masing-masing. Keputusan MAS

NMR keadaan pepejal menunjukkan interaksi di antara asid-asid amino dan OTS

terhidrolisis. Fenomena ini disokong oleh spektra by 13C NMR, di mana puncak-

puncak asid L-glutamik, L-leusin dan L-fenilanalina telah beralih ke medan magnet

yang lebih tinggi. Ini dipastikan dengan spektra 29Si NMR yang menunjukkan

puncak-puncak untuk –(OH)Si(R)-O-(OH)Si(R)- (R=kumpulan oktadesil) terjalin

silang tetapi puncak-puncak untuk R–Si=(OSiR)3 tidak kelihatan pada OTS yang

dihidrolisis oleh asid amino. Keputusan ini mencadangkan bahawa asid amino

dijalinkan secara –(OH)Si(R)-O-(OH)Si(R)- terjalin silang pada OTS terhidrolisis.

OTS yang dihidrolisis oleh asid amino telah digunakan sebagai mangkin untuk

hidrasi asimetri epoksisikloheksena untuk mendapatkan dua diastereoisomer, iaitu

(1R,2R)-trans-1,2-sikloheksanadiol dan (1S,2S)-trans-1,2-sikloheksenadiol serta cis-

1,2- sikloheksenadiol. Mangkin-mangkin ini telah mempamerkan 12-18% lebihan

enantiomerik (ee) untuk hidrasi asimetri epoksisikloheksen unutuk menghasilkan

(1R,2R)-trans-1,2-sikloheksanadiol and (1S,2S)-trans-1,2-sikloheksanadiol.

v

PENYELIDIK UTAMA

Prof. Madya Dr. Hadi Nur

Prof. Madya Dr. Salasiah Endud

Prof. Madya Dr. Zainab Ramli

En. Lim Kheng Wei

Email : hadi@ibnusina.utm.my

Tel. No. : 07-5536077

Vot No. : 74284

vi

TABLE OF CONTENTS

CHAPTER TITLE PAGE NO.

TITLE PAGE

ACKNOWLEDGEMENT ii

ABSTRACT iii

KEY RESEARCHERS iv

ABSTRAK v

PENYELIDIK UTAMA vi

TABLE OF CONTENTS vii

1 INTRODUCTION 1

1.1 Problem Identification 1

1.2 Homogeneous Asymmetric Catalyst 3

1.3 Heterogeneous Asymmetric Catalyst 4

1.3.1 Modification of Metal/Solid Surfaces with

Chiral Molecules 4

1.3.2 Attaching Chiral Auxiliaries to Reactant 5

1.3.3 Intrinsically Chiral Solid/Surface Catalyst 6

1.4 Research Strategy 6

1.5 Scope and Objectives of the Study 8

vii

2 EXPERIMENTAL 11

2.1 Synthesis of Heterogeneous Asymmetric Catalyst 11

2.2 Characterizations of Catalyst 12

2.2.1 Dispersibility of Catalysts in Immiscible Organic

and Aqueous Phases 12

2.2.2 Solid State MAS NMR 12

2.3 Catalytic Performance 13

3 RESULTS AND DISCUSSION 17

3.1 Hydrophilicity-Hydrophobicity of Catalysts 17

3. 2 Solid State MAS NMR 18

3.3 Enantioselective Hydration of

Epoxycyclohexene 19

4 CONCLUSION 27

REFERENCES 28

PUBLICATIONS

1

CHAPTER 1

INTRODUCTION

1.1 Problem Identification

The control of enantioselectivity by heterogeneous asymmetric catalysis

remains a big challenge today. The main drive has been to find new, exciting routes to

chiral molecules while achieving high enantiomer selectivity. The demand for chiral

compounds, often as single enantiomers, has escalated sharply in recent years, driven

particularly by the demands of the pharmaceutical industry, but also by other

applications, including agricultural chemicals, flavors, fragrances, and materials. Two-

thirds of prescription drugs are chiral, with the majority of new chiral drugs being

single enantiomers. Although the most obvious applications are bio-related, materials

science also relies on the properties imparted by chirality, notably in chiral polymers

and liquid crystals. This widespread demand for chiral compounds has stimulated

intensive research to develop improved methods for synthesizing such compounds.

Many of the compounds associated with living organisms are chiral, for

example DNA, enzymes, antibodies and hormones. Therefore enantiomers of

compounds may have distinctly different biological activity. For example, the

enantiomers of limonene, both formed naturally, smell differently. One of the

2

enantiomers (S)-limonene smells of lemons, while the mirror image compound (R)-

limonene smells of oranges (Figure 1).

(R)-limonene (S)-limonene

Mirror Plane

Figure 1: (R)-Limonene smells of oranges and (S)-limonene smells of lemons.

We can distinguish between the two enantiomers because our nasal receptors

are also made up of chiral molecules that recognise the difference. Insects use chiral

chemical messengers (pheromones) as sex attractants and chemists have discovered

that one of the enantiomers of the insect pheromone, olean, attracts male fruit flies,

while its mirror image operates on the female of the species. Thus biology is very

sensitive to chirality and the activity of drugs also depends on which enantiomer is

used.

Most drugs consist of chiral molecules. Since a drug must match the receptor

in the cell, it is often that only one of the enantiomers is of interest. In certain cases

the other enantiomer may be harmful, even in the early 1960s; the drug thalidomide

was prescribed to alleviate morning sickness in pregnant woman. Tragically, the drug

3

also caused deformities in the limbs of children born by these women. It seemed that

one enantiomer of thalidomide (D-thalidomide) was beneficial while the other (L-

thalidomide) caused the birth defects.

As a result, pharmaceutical companies nowadays have to make sure that both

enantiomers of a drug are tested for their biological activity and toxicity before they

are marketed. Obviously, the purity of these molecules for their use in medical or

veterinary applications is paramount. Hence, there is always a need to find synthetic

strategies that can give both high yields and enantioselectivity for these important

molecules, so that the subsequent purification processes can be minimized.

1.2 Homogeneous Asymmetric Catalyst

Homogeneous catalysts have been developed for numerous C–H, C–C, C–O

and C–N bond formation reactions, which give high enantioselectivity. The design

has been predicated on the synthesis of chiral ligands for active catalyst centers that

ensures that the desired chiral transition state is readily accessed. However, these

ligands are often difficult to recover and reuse and, for these reasons, chiral

homogeneous catalysts have not had the significant commercial input that researchers

initially hoped for [2]. Additionally, attention has been focused on the design of

immobilized homogeneous chiral catalysts or heterogeneous chiral catalysts because

they can be readily recovered by filtration or a slurry reactor, or can be used in the

numerous fixed bed reactor options available. This introduces a different aspect of

4

catalyst usability, namely the stability of the heterogeneous catalyst, because leaching

of active components can represent a real problem concerning commercialization [3].

1.3 Heterogeneous Asymmetric Catalyst

To date there have been numerous approaches to the design of heterogeneous

asymmetric catalysts, since Schwab and coworkers first demonstrated that Cu and Ni

could be supported on chiral silica surfaces [4,5] and that the resulting catalysts could

give low enantioselection in the dehydration of butan-2-ol. Figure 2 shows three

models of heterogeneous asymmetric catalysis, i.e. chirally modified solid surfaces,

attaching chiral auxiliaries to reactant and intrinsically chiral solid or surface.

1.3.1 Modification of Metal/Solid Surfaces with Chiral Molecules

Heterogeneous-enantioselective hydrogenation of prochiral ketones has been

extensively studied using cinchona-modified-supported Pt catalysts [6-9] and tartrate-

modified-supported Ni catalysts [10–12]. However, the hydrogenation of pyruvate

esters using cinchonidine-modified Pt/SiO2 and Pt/? -Al2O3 represents the most-well-

investigated reactions, and these are often studied as a model system. For this

reaction, the enantioselectivity can be high and enantioselective excess (e.e.) in excess

of 90% has been reported in many studies [6-9, 13–15].

5

Chirally modified solid surfaces

Attaching chiral auxiliaries to reactant

Intrinsically chiral solid or surface“bare”

or as supportfor active centres

chiral modifierreactant

Figure 2. Models of heterogeneous asymmetric catalysis.

1.3.2 Attaching Chiral Auxiliaries to Reactant

At present, attaching chiral auxiliaries to reactant is one of the most popular

approaches to the design of highly efficient heterogeneous asymmetric catalysts. The

strategy employed tends to depend on the reaction being catalyzed, but overall this

method has the potential for designing generic heterogeneous asymmetric catalysts

applicable to many reaction types.

6

1.3.3 Intrinsically Chiral Solid/Surface Catalyst

Compare to the above two methods (section 1.3.1 & 1.3.2), there are very few

reports on intrinsically chiral solid/surface catalyst. The only reported “intrinsically”

heterogeneous catalytic system to synthesize chiral/enantiomer compound is by BEA

type zeolite structure synthesized with a chiral template molecule. However, only 5%

enantioselectivity could be achieved [17, 18]. Therefore, the greatest challenge is to

synthesize a highly heterogeneous enantioselective catalyst. Heterogeneous

enantioselective catalyst or so called chiral solid catalyst is of paid the major interest

due to their advantages, not only ease to recover, recycling and environmental

friendliness, but it is also amenable for continuous processing. Based on the above

considerations, a strategy to obtain highly active catalyst in the enantioselective

hydration of epoxyclohexene is proposed.

1.4 Research Strategy

The research strategy is based on the ideas that the chiral reactions could be

induced by chiral amino acids and the use of heterogeneous micellar catalysis for

synthetic purposes will overcome practical separation problems. In order to realize

these ideas, chiral amino acid needs to be attached to the hydrophilic part of

hydrolyzed octadecyltrichlorosilane (OTS). Amino acids such as L-glutamic acid, L-

leucine and L-phenylalanine have been chosen because of their water-soluble

properties; hence they can be easily removed by treatment with water. It is expected

that the attachment of amino acid would result in a chiral solid catalyst with bimodal

hydrophobic-hydrophilic character (see Figure 3). In addition, due to OTS has three

7

chloro functional groups, one expects that the OTS has high tendency to polymerized

(see Figure 4).

One considers that a catalyst which possesses both hydrophobic and

hydrophilic components exhibits amphiphilic character. The flexibility of the

hydrophobic octadecyl groups therefore allows the formation of micellar aggregates

in the system containing immiscible organic and aqueous phases. Schematic

representation of heterogeneous micellar catalysis is depicted in Figure 5. As shown

in Figure 5, it is expected that the hydrophilic microdomains in micellar aggregates

will act as “chiral pool” for acid chiral reaction.

amino acid

amino acid-hydrolyzed OTS

octadecyltrichlorosilane (OTS)

hydrophobichydrophilic

Si

Cl

Cl

Cl

Wash with water in order to remove “free” amino acid

Figure 3. Attachment of chiral amino acid onto the hydrophilic part of hydrolyzed

OTS.

8

1.5 Scope and Objecives of the Study

The goal this research is to develop a novel heterogeneous asymmetric catalyst

by attachment of chiral amino acid onto the hydrophilic part of hydrolyzed

octadecyltrichlorosilane (OTS). Namely, this research will lead to the synthesis of

heterogeneous asymmetric catalyst. The new catalysts are expected to possess tunable

activity and selectivity for enantioselective hydration of epoxyclohexene. Structural

and mechanistic aspects as well as key to the catalytic performance will be

determined by physical and chemical characterization methods.

The objectives of the study are:

(i) To develop a new approach in preparation of heterogeneous asymmetric

catalyst.

(ii) To synthesize and characterize heterogeneous asymmetric catalyst using

amino acid as chiral promoter.

(iii) To study and evaluate the performance of synthesized heterogeneous

asymmetric catalyst in enantioselective hydration of epoxyclohexene.

9

Hydrophobic

hydrogencarbonsiliconoxygen

Hydrophilic

L- Glutamic AcidL- LeucineL- Phenylalanine

Amino acid

Amino acid

Figure 4. The interaction of amino acid with cross-linked –(R)Si(OH)-O-

(R)Si(OH)- of hydrolyzed OTS (R = octadecyl group).

10

O

OH

OH

HO

HO

+

Epoxycyclohexene (1R,2R) (1S,2S)

H2O + [H+]

Catalyst

watercontaining acid

H+

Expected to be “chiral pool” for acid chiral reaction

Figure 5. Amphiphilic chiral solid catalyst as heterogeneous micellar catalyst in

enantioselective hydration of epoxyclohexene.

11

CHAPTER 2

EXPERIMENTAL

2.1 Synthesis of Heterogeneous Asymmetric Catalyst

Three types of heterogeneous asymmetric catalyst were prepared using three

kinds of amino acid as chiral promoter i.e. L-Glutamic acid, L-Leucine and L-

Phenylalanine. The amino acids are attached onto the hydrophilic part of hydrolyzed

octadecyltrichlorosilane (OTS) during sol-gel synthesis.

Ten mmol of octadecyl trichlorosilane (OTS) was added into 20 mmol of L-

Glutamic acid. Ten mL of toluene was added into the mixture which was then stirred

at ambient temperature under open atmosphere in order to hydrolyze the OTS. The

resulting solids were then washed thoroughly using boiling distilled water to remove

free L-Glutamic acid which is not attached to hydrolyzed OTS and finally dried in an

oven at 50oC. The as-prepared catalyst was denoted as hydrolyzed OTS-Glu. A

similar procedure was also carried out to prepare L-Leucine attached to hydrolyzed

OTS (denoted as hydrolyzed OTS-Leu) and L-Phenylalanine attached to hydrolyzed

OTS (denoted as hydrolyzed OTS-PheAla).

12

2.2 Characterizations of Catalyst

2.2.1 Dispersibility of Catalysts in Immiscible Organic and Aqueous Phases

In order to establish the hydrophilicity-hydrodphobity property of the

catalysts, each of the hydrolyzed OTS-Glu, hydrolyzed OTS-Leu and hydrolyzed

OTS-PheAla samples (0.06 g) was added separately into an immiscible mixture of

toluene (3 ml) and water (1 ml). Then, the mixture was stirred vigorously at 1000 rpm

for 2 h and kept under static condition for 2 h. The capability of the particles to

stabilize the liquid–liquid system to form an emulsion was evaluated as a qualitative

measure of the hydrophobic-hydrophilic character and hence their surface properties.

2.2.2 Solid State MAS NMR

The structural information about the chemically modified material can be

obtained by means of Solid-State MAS NMR spectroscopy. In solid state samples,

due to the limited motion, strong dipolar-dipolar and chemical shift anisotropy

interactions occur. The line broadening effects can be cancelled using magic angle

spinning ( M AS ) technique.

The MAS NMR experiments were performed using Bruker Avance 400 MHz

9.4T spectrometer. The 29Si MAS NMR spectra were recorded at 79.44 MHz using 4

µs radio frequency pulses, a recycle delay of 9 s and spinning rate of 7.0 kHz using a

13

4 mm zirconia sample rotor. The 13C MAS NMR spectra were collected by a cross

polarization (CP) MAS method at 100.62 MHz with a 3000 µs 13C pulse, 5 s recycle

delay and spinning rate of 7.0 kHz using a 4 mm zirconia sample rotor. Both 29Si and

13C NMR chemical shifts were referred to external TMS at 0 ppm.

In this research, it is of interest to characterize the short range order structure

of amino acid attached on the hydrolyzed OTS by 29Si and 13C CP/MAS NMR. The

nomenclature and 29Si NMR chemical shifts of some silyl spesies are sown in Figure

6.

2.3 Catalytic Performance

Catalytic performance of the synthesized heterogeneous asymmetric catalyst is

tested using hydration of epoxycyclohexene as model reaction. Hydration of

epoxycyclohexene to cyclohexanediol, in principle, yields three stereoisomers: the

cis-1,2-cyclohexanediol and a pair of enantiomers, namely (1S,2S)-trans-1,2-

cyclohexanediol and (1R,2R)-trans-1,2-cyclohexanediol. The hydrolysis of

epoxycyclohexene to 1,2-cyclohexanediol was performed as follows. The catalyst

particles (50 mg) were placed in a glass tube and then 10 µmol of 0.1 M H2SO4 was

added to the catalysts. Then 50 mmol of epoxycyclohexene (Merck, >99%) was

added to the solid catalyst wetted with H2SO4 and reacted for 24 h at room

temperature under stirring condition. The products of the reaction were analyzed by

Agilent Model 7890N gas chromatography using Chiraldex B-DM capillary column

containing 2,3-di-omethyl-6-t-butyl silyl derivative of β-cyclodextrin. 1 µl of the

14

sample was injected into the GC inlet with the split ratio 100:1. The oven condition as

follows, initial temperature 50 oC, ramp at 4.3 oC min-1 to reach 120 oC, hold for 9

min, and finally ramp at 5 oC min-1 to reach 160 oC with a constant flow of 1.7 ml

min-1. Authentic samples i.e. cis-1,2-cyclohexanediol (Aldrich, >99%), (1S,2S)-trans-

1,2-cyclohexanediol (Fluka, >99%) and (1R,2R)-trans- 1,2-cyclohexanediol (Fluka,

>99%) were used to determine the products.

Si

OH

OH

H2C R Si

OH

O

H2C R

Si

O

OH

H2C R

SiH2C R

Si

OH

OH Si

Si

OH

OH

SiO

O

O

O

T1 -48 ppm

T2 -56 ppm

T3 -65 ppm

T2 -56 ppm

Q1 -92 ppm Q2 -101 ppm Q3 -110 ppm

Figure 6: Nomenclature and 29Si NMR chemical shifts of some silyl spesies[19].

15

Figure 7. Hydration of epoxycyclohexene to 1,2-cyclohexenediol

The hydrolysis of epoxycyclohexene to 1,2-cyclohexenediol was performed as

follows. The catalyst particles (50 mg) were placed in a glass tube and then 10 µmol

of 0.1 M H2SO4 was added to the catalysts. Then 50 mmol of epoxycyclohexene

(Merck, >99%) was added to the solid catalyst wetted with H2SO4 and reacted for 24

h at room temperature under stirring condition. The products of the reaction were

analyzed by Agilent Model 6890N gas chromatography using Chiraldex G-TA

capillary column containing γ-cyclodextrin trifluoroacetyl with length and internal

diameter 40 m x 0.25 mm. 1 µL of the sample was injected into the GC inlet with the

split ratio 100:1. The oven conditions are as follows: initial temperature of 40oC, held

for 10 minutes, continuous ramping at 2oC/minute to reach 130oC, again held for 8

minutes, and finally ramping at 5oC/minute to reach 160oC.

Authentic samples i.e. cis-1,2-cyclohexenediol (Aldrich, >99%), (1S,2S)-

trans-1,2-cyclohexenediol (Fluka, >99%) and (1R,2R)-trans-1,2-cyclohexenediol

(Fluka, >99%) were used to determine the products. The enantiopurity or

enantiomeric excess (e.e.) of the 1,2-cyclohexenediol was determined by Equation 1

O

OH

OH

HO

HO

+ +

OH

OH

Epoxycyclohexene (1R,2R)-trans-1,2-cyclohexenediol (1S,2S)-trans-1,2-cyclohexenediol Cis-1,2-cyclohexenediol

H2O + [H+]

Catalyst

16

in which % e.e. is percentage of enantiomeric excess, while Major is mmol of the

major product and Minor is mmol of the lower product.

%100..% ×+−=

MinorMajor

MinorMajoree (Equation 1)

17

CHAPTER 3

RESULTS AND DISCUSSION

3.1 Hydrophilicity-Hydrophobicity of Catalysts

As shown in Figure 8, when the hydrolyzed OTS-Glu particles were added to

a mixture of 3 ml toluene and 1 ml water, they were feasibly located at the organic

phase. In order to rationalize the hydrophilicity-hydrophobicity of catalyst one should

consider the formation of an emulsion, because in this form the specific interfacial

interactions between the solid catalyst surface and the two immiscible liquid phases

increase the surface contact (wettability) of the catalyst with the reactants. For

maximum efficiency, the catalyst should be wetted preferentially by the two liquid

phases. If the solid particles are too strongly wetted by either of the two liquid phases

the required stabilizing action will not result. Based on these considerations, the

formation of the emulsion in the presence of the solid particles was examined. It was

observed that an emulsion was formed in the system containing hydrolyzed OTS-Glu

particles under stirring condition. As is shown in Figure 8, it is clearly demonstrated

that an emulsion has been formed, resulting in an abrupt visual homogenization. This

suggests that the flexibility of the hydrophobic octadecyl groups allows the formation

18

of micellar aggregates in the system containing immiscible organic and aqueous

phases. Two hours after the stirring stopped, it was found that the OTS-amino acid

particles are settled into the organic phase (see Figure 8).

3. 2 Solid State MAS NMR

Figs. 9 and 10 show the 13C CP/MAS NMR spectra of hydrolyzed OTS-Glu and

hydrolyzed OTS-PheAla. The NMR spectrum of hydrolyzed OTS-Glu shows two

major signals at 177.0 and 174.1 ppm which provide strong evidence for the

prevalence of –COOH group of L-glutamic acid. The strong signal at 29.1 ppm

corresponds to C3 of alkylsilyl group of hydrolyzed OTS. It is clearly observed that

the peaks of C1 and C5 of –COOH groups of L-Gultamic acid attached to hydrolyzed

OTS are shifted towards a higher magnetic field in comparison to those of pure L-

Glutamic acid (Fig. 2). The shifting of the C2 peak toward a lower magnetic field of –

COOH was observed for hydrolyzed OTS-PheAla (Fig. 3). The shifting of the 13C

NMR signals can be explained by the interaction of the free electron pairs of the

oxygen atoms of carboxyl functional group of the amino acids with hydrolyzed OTS.

Fig. 11 shows the 29Si MAS NMR spectra of hydrolyzed OTS and hydrolyzed

OTS-Glu. The intense peak at –110 ppm is from Si(OSi-)3 in hydrolyzed OTS. In

Fig. 4, it is observed that the three additional peaks from –50 to –80 ppm correspond

to three different environments of the siloxane groups in the hydrolyzed OTS-Glu

[8,9]: (i) isolated groups that are not bound to any neighbouring siloxanes (ii) terminal

groups that are only bound to one neighbouring siloxane, and (iii) cross-linked groups

that are bound to two neighbouring siloxane. However the peak at –110 ppm from −

19

Si(OSi-)3 is not observed in hydrolyzed OTS-PheAla. Based on the above results, it is

suggested that the presence of amino acid during sol-gel synthesis of the hydrolyzed

OTS-Glu inhibits the formation of Si(OSi-)3 bonding in the sample. This result

suggested that the amino acid was attached by cross-linked –(OH)Si(R)-O-

(OH)Si(R)- of the hydrolyzed OTS.

3.3 Enantioselective Hydration of Epoxycyclohexene

The prepared catalysts were tested for the asymmetric hydration of

epoxycyclohexene. All the catalysts are active for this reaction (Fig. 5). (1R,2R)-

trans-1,2-cyclohexanediol, (1S,2S)-trans-1,2-cyclohexanediol and cis-1,2-

cyclohexanediol were the products detected. The catalysts show promising

enantioselectivity with 12-19% ee (S) for R and S of trans-1,2-cyclohexanediol.

Considering that amino acids are the integral part of the catalysts, the lower

enantioselectivity of hydrolyzed OTS-PheAla compared to that of hydrolyzed OTS-

Glu may be due to the different structure of amino acid existing in both samples since

their concentrations are almost the same (ca. 50%). One suggests that the enhanced

catalytic activity of the hydrolyzed OTS-amino acids is mainly caused by the specific

adsorption and physical properties of the catalysts with amino acid as a chiral

promoter. The alternative explanation for the higher the different enantioselectivity of

the two tested amino acids is due to L-glutamic acid contains an additional carboxyl

group available for hydrogen bonding which can definitely contribute to the larger

nantioselectivity in the non-covalent enantioselective catalysis.

During the reaction, under stirring condition, an abrupt visual homogenization

20

was observed suggesting the formation of the emulsion in the presence of the solid

particles. One considers that a catalyst which possesses both hydrophobic and

hydrophilic components exhibits amphiphilic character. The flexibility of the

hydrophobic octadecyl groups therefore allows the formation of micellar aggregates

in the system containing immiscible organic and aqueous phases. Schematic

representation of heterogeneous micellar catalysis is depicted in Fig. 6. As shown in

Fig. 6, it is expected that the hydrophilic microdomains in micellar aggregates will act

as “chiral pool” for acid chiral reaction. The enantioselectivity, albeit not yet high

enough, demonstrates the possibility for synthesizing a new kind of chiral solid

catalysts for potential applications in asymmetric reactions.

21

Static for 2 h

Stir for 2 h

toluene

water

water

toluene

water

toluene

= hydrolyzed OTS

amino acid

Figure 9. Dispersibility of hydrolyzed OTS-Glu in a mixture of toluene and water

under stirring and static conditions.

22

L-GlutamicAcid

hydrolyzed OTS

240 220 200 180 160 140 120 100 80 60 40 20 0 ppm

Si CH22

Cl

Cl

Cl

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

CH31

C1

C2

C3

C2C1 C3

C4C5

hydrolyzed OTS-Glu

HO C5

O

H2C4

CH2

3CH2

C1

NH2

OH

O

octadecyltrichlorosilane (OTS)

The shifting of the C1and C5 peaks of –COOH

Figure 9. 13C CP/MAS NMR spectra of hydrolyzed OTS, L-glutamic acid and

hydrolyzed OTS-Glu.

23

hydrolyzed OTS-PheAla

hydrolyzed OTS

240 220 200 180 160 140 120 100 80 60 40 20 0 ppm

Si CH22

Cl

Cl

Cl

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

CH31

C1

C2

C3

C2C1 C3

C4

C5 – C9

H2N CH1

C2

CH23

OH

O

4

9

8

7

6

5

L-Phenylalanine

octadecyltrichlorosilane (OTS)

The shifting of the C2peak of –COOH

Fig. 10. 13C CP/MAS NMR spectra of hydrolyzed OTS, L-phenylalanine and

hydrolyzed OTS-PheAla.

24

SiOSiHO

Si SiO SiO OHO Si

Cross-linkedTerminal

O O

O

SiSi OH

O

Si Q3

Si Q1

OH

Terminal

(a)

Cross-linked

-150-100-50 ppm

(b)

Isolated

Si Q3

Figure 11. 29Si NMR of (a) hydrolyzed OTS and (b) hydrolyzed OTS-Glu.

25

0

5

10

15

20

Enan

tiose

lect

ivity

/ %

(S)

no catalyst hydrolyzed OTS-PheAla

hydrolyzed OTS-Glu

Figure 12. Enantioselectivity of catalysts for R and S of trans-1,2-cyclohexanediol in

the asymmetric hydration of epoxycyclohexene.

26

watercontaining acid

H+

Expected to be “chiral pool” for acid chiral reaction

Figure 13. Amphiphilic chiral solid catalyst as heterogeneous micellar catalyst in

enantioselective hydration of epoxyclohexene.

27

CHAPTER 4

CONCLUSION

Heterogeneous asymmetric catalyst using amino acid as chiral promoter has

been successfully synthesized by attachment of amino acid such as L-glutamic acid,

L-leucine and L-phenylalanine onto the hydrophilic part of hydrolyzed

octadecyltrichlorosilane (OTS). The short-range order structure of silicon and organic

groups in the catalysts has been confirmed by solid-state 29Si and 13C cross-

polarization magic-angle-spinning (MAS) NMR spectroscopies, respectively. The

13C CP/MAS NMR results showed the occurrence of the interaction of amino acid

with hydrolyzed OTS. This phenomenon was supported by the fact that the 13C

signals of the peaks of L-glutamic acid, L-leucine and L-phenylalanine were shifted

towards higher magnetic fields. The 29Si MAS NMR spectra confirmed that the peaks

due to cross-linked –(OH)Si(R)-O-(OH)Si(R)- (R=octadecyl group) were mainly

observed whereas R–Si=(OSiR)3 corresponding to Q3 silyl spesies was not observed

in amino acid-hydrolyzed OTS. This result suggested that the amino acid was

attached by cross-linked –(OH)Si(R)-O-(OH)Si(R)- of hydrolyzed OTS. The catalysts

exhibit 12-18% enantiomeric excess (ee) for the asymmetric hydration of

epoxycyclohexene to yield (1R,2R)-trans-1,2-cyclohexanediol and (1S,2S)-trans-1,2-

cyclohexanediol. The enantioselectivity, albeit not yet high enough, demonstrates the

possibility for synthesizing a new kind of chiral solid catalysts for potential

applications in asymmetric reactions.

28

REFERENCES

1. Catalytic asymmetric synthesis, Advanced information on the Nobel Prize in

Chemistry 2001. http://nobelprize.org/chemistry/laureates/2001/adv.

2. H.U. Blaser. (2003). “Enantioselective catalysis in fine chemicals production”.

Chem.Commun. 3:293–96

3. R.A. Sheldon, M. Wallau, I.W.C.E. Arends, U. Schuchart. (1998).

“Heterogeneous catalysts for liquid-phase oxidations: Philosophers’ stones or

Trojan horses?” Acc. Chem.Res. 31:485–93

4. P. McMorn, G.J. Hutchings. (2003). “Heterogeneous enantioselective catalysts

strategies for the immobilization of homogeneous catalysts”. Chem. Soc. Rev.

33:108–22

5. G.M. Schwab, L. Rudolph. (1932). “Catalytic cleavage of racemates by d- and l-

quartz”. Naturwisserschaft 20:363–64

6. A. Baiker. (2000). “Transition state analogues—a guide for the rational design

of enantioselective heterogeneous hydrogenation catalysts”. J. Mol. Catal. A

163:205–20

7. H.U. Blaser, J.P. Jalett, M. M¨uller, M. Studer. (1997). “Enantioselective

hydrogenation of a-ketoesters using cinchona modified platinum catalysts and

related systems: a review”. Catal. Today 37:441–63

8. P.B. Wells, A.G. Wilkinson. (1998). “Platinum group metals as heterogeneous

enantioselective catalysts”. Top. Catal. 5:39–50

29

9. M. Von Arx, T. Mallat, A. Baiker. (2002). “Asymmetric hydrogenation of

activated ketones on platinum: relevant and spectator species”. Top. Catal.

19:75–87

10. P. Kukula, L. Cerveny. (2002). “Effects of reaction variables on

enantioselectivity of modified Raney nickel catalyst”. J. Mol. Catal. A 185:195–

202

11. T. Sugimura, S. Nakayama, A. Tai. (2002). “Over 98% optical yield achieved by

a heterogeneous catalysis. Substrate design and analysis of enantio-

differentiating factors of tartaric acid-modified Raney nickel hydrogenation”.

Bull. Chem. Soc. Jpn. 75:355–63

12. T. Osawa, S. Sakai, T. Harada, O. Takayashu. (2001). “Highly durable enantio-

differentiating nickel catalyst for repeated use for the hydrogenation of methyl

acetoacetate”. Chem. Lett. 392–99

13. M. Schurch, N. Künzle, T. Mallat, A. Baiker. (1998). “Design of newmodifiers

for the enantioselective hydrogenation of ethyl pyruvate”. J. Catal. 176:569–71

14. G.J. Hutchings. (1999). “New approaches to rate enhancement in heterogeneous

catalysis”. Chem. Commun. 301–6

15. X. Zuo, H. Liu, M. Liu. (1998). “Asymmetric hydrogenation of a-ketoesters

over finely dispersed polymer-stabilized platinum clusters”. Tetrahedron Lett.

39:1941–44

16. G.M. Schwab, F. Rost, L. Rudolph. (1934). “Optically asymmetric catalysis on

quartz crystals”. Kolloid Z. 68:157–59

30

17. M. E. Davis and R. F. Lobo. (1992). “Zeolite and Molecular-Sieve Synthesis”,

Chemistry of Materials, 4, 756-768

18. R. F. Lobo, S. I. Zones, M. E. Davis. (1994). “Effect of the Stacking Probability

on the Properties of the Molecular Sieve CIT-1, SSZ-26 and SSZ-33”, Stud.

Surf. Sci. Catal., 84, 461

19. V. Coman, R. Grecu, J. Wegmann, S. Bachmann, K. Albert. (2001). “Solid State

NMR and FTIR Characterization of 3-Mercaptopropyl Silica Gel R”. 2nd

Conference on Isotopic and Molecular Processes PIM, http://www.itim-

cj.ro/PIM/2003/2001/FullText.html

20. X. Feng, G.E. Fryxell, L.Q. Wang, A.Y. Kim, J. Liu. (1997). “Organic

Monolayers on Ordered Mesoporous Supports”. Science, 276, 923-926.

21. J. Liu, X. Feng, G.E. Fryxell, L.Q. Wang, A.Y. Kim, M. Gong. (1998) “Hybrid

Mesoporous Materials with Functionalized Monolayers”. Advanced Materials,

10, 161-165.

1

2 Hydrolyzed octadecyltrichlorosilane functionalized

3 with amino acids as heterogeneous enantioselective

4 catalysts

5 Hadi Nur Lim Kheng Wei Salasiah Endud

6 Received: 15 January 2009 / Accepted: 2 May 20097 � Akademiai Kiado, Budapest, Hungary 2009

8 Abstract Heterogeneous asymmetric catalysts using amino acids as chiral pro-

9 moter were synthesized by attachment of amino acids such as L-glutamic acid and

10 L-phenylalanine onto the hydrophilic part of hydrolyzed octadecyltrichlorosilane

11 (OTS). The catalysts exhibited 12–18% enantiomeric excess (ee) for the asymmetric

12 hydration of epoxycyclohexene to yield (1R,2R)-trans-1,2-cyclohexanediol and

13 (1S,2S)-trans-1,2-cyclohexanediol.

14 Keywords Heterogeneous asymmetric catalysts � Amino acid as chiral promoter �

15 Asymmetric hydration of epoxycyclohexene

16

17 Introduction

18 The control of enantioselectivity by heterogeneous asymmetric catalysis remains a

19 big challenge today. The main drive has been to find new, exciting routes to chiral

20 molecules while achieving high enantiomer selectivity. To date, there have been

21 numerous approaches to the design of heterogeneous asymmetric catalysts [1–7]. In

22 this paper, a new strategy to obtain active catalyst in the enantioselective hydration

23 of epoxyclohexene is proposed. The research strategy is based on the ideas that the

24 enantioselective reactions could be induced by chiral amino acids and the use of

25 heterogeneous catalysis for synthetic purposes will overcome practical separation

26 problems. In order to realize these ideas, a chiral amino acid needs to be attached to

27 the hydrophilic part of hydrolyzed OTS. Amino acids such as L-glutamic acid and

28 L-phenylalanine have been chosen because of their water-soluble properties; hence

29 they can be easily removed by treatment with water. It is expected that the

A1 H. Nur (&) � L. K. Wei � S. Endud

A2 Ibnu Sina Institute for Fundamental Science Studies, Universiti Teknologi Malaysia (UTM),

A3 81310 Skudai, Johor, Malaysia

A4 e-mail: hadi@kimia.fs.utm.my

123

React Kinet Catal Lett

DOI 10.1007/s11144-009-0056-7

UNCORR

30 attachment of the amino acid would result in a chiral solid catalyst with bimodal

31 hydrophobic-hydrophilic character (see Fig. 1). In addition, due to OTS has three

32 chloro functional groups, one expects that the OTS has high tendency to polymerize.

33 Experimental

34 Two types of heterogeneous asymmetric catalyst were prepared using two kinds of

35 amino acids as a chiral promoter i.e. L-glutamic acid and L-phenylalanine. The

36 amino acids are attached onto the hydrophilic part of hydrolyzed OTS during

37 sol–gel synthesis. 10 mmol of OTS was added into 20 mmol of L-glutamic acid. Ten

38 mL of toluene was added into the mixture which was then stirred at ambient

39 temperature under open atmosphere in order to hydrolyze the OTS. The resulting

40 solids were then washed thoroughly using boiling distilled water to remove free

41 L-glutamic acid which is not attached to hydrolyzed OTS and finally dried in an

42 oven at 50 �C. The as-prepared catalyst was denoted as hydrolyzed OTS-Glu. A

43 similar procedure was also carried out to prepare L-Phenylalanine attached to

44 hydrolyzed OTS (denoted as hydrolyzed OTS-PheAla).

45 The structural information about the chemically modified material is obtained by

46 means of solid-state MAS NMR spectroscopy. The MAS NMR experiments were

amino acid

amino acid-hydrolyzed OTS

octadecyltrichlorosilane (OTS)

hydrophobichydrophilic

Si

Cl

Cl

Cl

Wash with water in order to remove “free” amino acid

Fig. 1 Attachment of chiral amino acid onto the hydrophilic part of hydrolyzed OTS

H. Nur et al.

123

47 performed using Bruker Avance 400 MHz 9.4T spectrometer. The 29Si MAS NMR

48 spectra were recorded at 79.44 MHz using 4 ls radio frequency pulses, a recycle

49 delay of 9 s and spinning rate of 7.0 kHz using a 4 mm zirconia sample rotor. The

50 13C MAS NMR spectra were collected by a cross polarization (CP) MAS method at

51 100.62 MHz with a 3,000 ls 13C pulse, 5 s recycle delay and spinning rate of

52 7.0 kHz using a 4 mm zirconia sample rotor. Both 29Si and 13C NMR chemical

53 shifts were referred to external TMS at 0 ppm.

54 Catalytic performance of the synthesized heterogeneous asymmetric catalyst was

55 tested using hydration of epoxycyclohexene as model reaction. Hydration of

56 epoxycyclohexene to cyclohexanediol, in principle, yields three stereoisomers: the

57 cis-1,2-cyclohexanediol and a pair of enantiomers, namely (1S,2S)-trans-1,2-

58 cyclohexanediol and (1R,2R)-trans-1,2-cyclohexanediol. The hydrolysis of epoxy-

59 cyclohexene to 1,2-cyclohexanediol was performed as follows. The catalyst

60 particles (50 mg) were placed in a glass tube and then 10 lmol of 0.1 M H2SO4

61 was added to the catalysts. Then 50 mmol of epoxycyclohexene (Merck,[99%)

62 was added to the solid catalyst wetted with H2SO4 and reacted for 24 h at room

63 temperature under stirring condition. The products of the reaction were analyzed by

64 Agilent Model 7890 N gas chromatography using Chiraldex B-DM capillary

65 column containing 2,3-di-omethyl-6-t-butyl silyl derivative of (–cyclodextrin. 1 ll

66 of the sample was injected into the GC inlet with the split ratio 100:1. The oven

67 condition were as follows: initial temperature 50 �C, ramp at 4.3 �C min-1 to reach

68 120 �C, hold for 9 min, and finally ramp at 5 �C min-1 to reach 160 �C with a

69 constant flow of 1.7 mL min-1. Authentic samples i.e. cis-1,2-cyclohexanediol

70 (Aldrich,[99%) (1S,2S)-trans-1,2-cyclohexanediol (Fluka,[99%) and (1R,2R)-

71 trans-1,2-cyclohexanediol (Fluka,[99%) were used to determine the products.

72 Results and discussion

73 Figures 2 and 3 show the 13C CP/MAS NMR spectra of hydrolyzed OTS-Glu and

74 hydrolyzed OTS-PheAla. The NMR spectrum of hydrolyzed OTS-Glu shows two

75 major signals at 177.0 and 174.1 ppm which provide strong evidence for the

76 prevalence of –COOH group of L-glutamic acid. The strong signal at 29.1 ppm

77 corresponds to C3 of alkylsilyl group of hydrolyzed OTS. It is clearly observed that

78 the peaks of C1 and C5 of –COOH groups of L-Gultamic acid attached to

79 hydrolyzed OTS are shifted towards a higher magnetic field in comparison to those

80 of pure L-Glutamic acid (Fig. 2). The shifting of the C2 peak toward a lower

81 magnetic field of –COOH was observed for hydrolyzed OTS-PheAla (Fig. 3). The

82 shifting of the 13C NMR signals can be explained by the interaction of the free

83 electron pairs of the oxygen atoms of carboxyl functional group of the amino acids

84 with hydrolyzed OTS.

85 Figure 4 shows the 29Si MAS NMR spectra of hydrolyzed OTS and hydrolyzed

86 OTS-Glu. The intense peak at –110 ppm is from Si(OSi–)3 in hydrolyzed OTS. In

87 Fig. 4, it is observed that the three additional peaks from –50 to –80 ppm

88 correspond to three different environments of the siloxane groups in the hydrolyzed

89 OTS-Glu [8, 9]: (i) isolated groups that are not bound to any neighbouring siloxanes

Hydrolyzed OTS functionalized with amino acids

123

ECTEDPROOF

90 (ii) terminal groups that are only bound to one neighbouring siloxane, and (iii)

91 cross-linked groups that are bound to two neighbouring siloxane. However the peak

92 at –110 ppm from—Si(OSi–)3 is not observed in hydrolyzed OTS-PheAla. Based on

93 the above results, it is suggested that the presence of amino acid during sol–gel

94 synthesis of the hydrolyzed OTS-Glu inhibits the formation of Si(OSi–)3 bonding in

95 the sample. This result suggested that the amino acid was attached by cross-linked

96 –(OH)Si(R)–O–(OH)Si(R)– of the hydrolyzed OTS.

97 The prepared catalysts were tested for the asymmetric hydration of epoxycyclo-

98 hexene. All the catalysts are active for this reaction (Fig. 5). (1R,2R)-trans-1,2-

99 cyclohexanediol (1S,2S)-trans-1,2-cyclohexanediol and cis-1,2-cyclohexanediol

L-GlutamicAcid

hydrolyzed OTS

240 220 200 180 160 140 120 100 80 60 40 20 0 ppm

CH22

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

CH3

1

C1

C2

C3

C2

C1 C3C4

C5

hydrolyzed OTS-Glu

HO C5

O

H2C4

CH2

3CH

2

C1

NH2

OH

O

octadecyltrichlorosilane (OTS)

The shifting of the C1and C5 peaks of –COOH

Fig. 2 13C CP/MAS NMR spectra of hydrolyzed OTS, L-glutamic acid and hydrolyzed OTS-Glu

H. Nur et al.

123

ECTEDPROOF

100 were the products detected. The catalysts show promising enantioselectivity with

101 12–19% ee (S) for R and S of trans-1,2-cyclohexanediol. Considering that amino

102 acids are the integral part of the catalysts, the lower enantioselectivity of hydrolyzed

103 OTS-PheAla compared to that of hydrolyzed OTS-Glu may be due to the different

104 structure of amino acid existing in both samples since their concentrations are

105 almost the same (ca. 50%). it may be suggested that the enhanced catalytic activity

106 of the hydrolyzed OTS-amino acids is mainly caused by the specific adsorption and

107 physical properties of the catalysts with amino acid as a chiral promoter. The

108 alternative explanation for the different enantioselectivity of the two tested amino

hydrolyzed OTS-PheAla

hydrolyzed OTS

240 220 200 180 160 140 120 100 80 60 40 20 0 ppm

CH22

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

H2C3

CH2

3

CH3

1

C1

C2

C3

C2

C1 C3 C4

C5 – C9

H2N CH1

C2

CH23

OH

O

4

9

8

7

6

5

L-Phenylalanine

octadecyltrichlorosilane (OTS)

The shifting of the C2 peak of –COOH

Fig. 3 13C CP/MAS NMR spectra of hydrolyzed OTS, L-phenylalanine and hydrolyzed OTS-PheAla

Hydrolyzed OTS functionalized with amino acids

123

ECTEDPROOF

109 acids is due to L-glutamic acid contains an additional carboxyl group available for

110 hydrogen bonding which can definitely contribute to the larger enantioselectivity in

111 the non-covalent catalysis.

112 During the reaction under stirring, an abrupt visual homogenization was observed

113 suggesting the formation of the emulsion in the presence of the solid particles. One

114 considers that a catalyst which possesses both hydrophobic and hydrophilic

115 components exhibits amphiphilic character. The flexibility of the hydrophobic

116 octadecyl groups therefore allows the formation of micellar aggregates in the system

117 containing immiscible organic and aqueous phases. Schematic representation of

118 heterogeneous micellar catalysis is depicted in Fig. 6. As shown in Fig. 6, it is

119 expected that the hydrophilic microdomains in micellar aggregates will act as

120 ‘‘chiral pool’’ for acid chiral reaction. The enantioselectivity, albeit not yet high

121 enough, demonstrates the possibility for synthesizing a new kind of chiral solid

122 catalysts for potential applications in asymmetric reactions.

Si OSiHO

Si SiO Si O O HO Si

Cross-linked Terminal

O O

O

SiSi OH

O

Si Q3

Si Q1

OH

Terminal

(a)

Cross-linked

-150 -100 -50 ppm

(b)

Isolated

Si Q3

Fig. 4 29Si NMR of (a) hydrolyzed OTS and (b) hydrolyzed OTS-Glu

H. Nur et al.

123

ECTED

123124 References

125 1. Blaser HU (2003) Chem Commun 3:293126 2. Sheldon RA, Wallau M, Arends IWCE, Schuchart U (1998) Acc Chem Res 31:485127 3. McMorn P, Hutchings GJ (2003) Chem Soc Rev 33:108128 4. Baiker A (2000) J Mol Catal A 163:205

0

5

10

15

20

Enantiosele

ctivity / %

(S

)

no catalyst hydrolyzed OTS

-PheAla

hydrolyzed OTS

-Glu

Fig. 5 Enantioselectivity of catalysts for R and S of trans-1,2-cyclohexanediol in the asymmetric

hydration of epoxycyclohexene

water containing acid

H+

Expected to be “chiral pool” for acid chiral reaction

Fig. 6 Amphiphilic chiral solid catalyst as heterogeneous micellar catalyst in enantioselective hydration

of epoxyclohexene

Hydrolyzed OTS functionalized with amino acids

123

129 5. Blaser HU, Jalett JP, Muller M, Studer M (1997) Catal Today 37:441130 6. Wells PB, Wilkinson AG (1998) Top Catal 5:39131 7. Von Arx M, Mallat T, Baiker A (2002) Top Catal 19:75132 8. Feng X, Fryxell GE, Wang LQ, Kim AY, Liu J (1997) Science 276:923133 9. Liu J, Feng X, Fryxell GE, Wang LQ, Kim AY, Gong M (1998) Adv Mater 10:161–165

134

H. Nur et al.

123

Materials Science and Engineering B 133 (2006) 49–54

Modification of titanium surface species of titania byattachment of silica nanoparticles

Hadi Nur ∗Ibnu Sina Institute for Fundamental Science Studies, Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia

Received 4 April 2006; received in revised form 22 April 2006; accepted 6 May 2006

Abstract

Silica nanoparticles-attached titania (TiO2@SiO2) was prepared by modification of the surface of titania (TiO2) with octadecyltrichlosilane(OTS) as a source of silica. The attachment of silica was achieved through repeated deposition–hydrolysis–calcination of OTS on the surface oftitania particles. For comparison, the silica–titania composite (TiO2–SiO2) containing titania and silica agglomerated particles was also prepared.The physicochemical characteristics of TiO2@SiO2 and TiO2–SiO2 particles are experimentally studied by X-ray diffraction (XRD), UV–visdiffuse reflectance (UV–vis DR), scanning electron microscopy (SEM), energy dispersive X-ray analysis (EDAX) and 29Si solid-state magic anglespinning nuclear magnetic resonance (MAS NMR) techniques. The effect of silica attaching on activity of titanium surface species has been testedin the liquid phase epoxidation of 1-octene to 1,2-epoxyoctane with aqueous hydrogen peroxide. It is demonstrated that the surface attachmentof titania with silica nanoparticles enhances the epoxidation activity of titania. UV–vis DR spectra of the solid particles showed the increase inintensity of the tetrahedral titanium in TiO2@SiO2. This indicates the occurrence of the transformation of some of the octahedral titanium to thetetrahedral structure during silica attaching of TiO2@SiO2 in which titanium tends to bond with –Si(OH)2(OSi)2 to make it stable in tetrahedralform. Considering that the tetrahedral titanium was considered the most active species in epoxidation of alkene, it can be concluded that a highepoxidation activity of TiO2@SiO2 particle was related to the modification of the local environment of titanium surface species by attachment ofsilica nanoparticles.© 2006 Elsevier B.V. All rights reserved.

Keywords: Silica nanoparticles-attached titania; Titanium surface species; Surface modification

1. Introduction

Titanium dioxide or titania is one of the most utilized partic-ulate materials in the world. Although it was discovered morethan 200 years ago and has been commercially processed for85 years, it is still being actively researched. Titania is the mostwidely used as pigment in cosmetic and paint products, andphotocatalyst. The photochemical activity of titania is modifiedby promoting or suppressing the recombination of electron andhole pairs which are formed by UV light excitation [1]. Highphotochemical activity is required when titania is used as pho-tocatalyst, whereas low photochemical activity is preferred forpaint or cosmetics application. One of methods to improve thephotocatalytic activity of titania is by attachment of silica. It

∗ Tel.: +60 7 5536077; fax: +60 7 5536080.E-mail address: hadi@kimia.fs.utm.my.URL: http://www.ibnusina.utm.my/∼hadi.

has been reported that the photoactivity of silica-incorporatedtitania was three times higher than that of the Degussa P25 [2].However, the reason why the photoactivity of silica-incorporatedtitania gave a higher activity compared to that of bare titania havenot yet been clearly understood.

Recently room temperature sol–gel attachments of silica havebeen successfully adopted in order to coat titania with silica [3].The attachment of titania with stable oxide layers has been alsodone by spraying of a metal oxide precursor solution on the sur-face of titania nanoparticles [4–6], acid hydrolysis of polysilicate[7] and hydrolysis of tetraethylorthosilicate (TEOS) using chlo-roform catalyst [8]. However, attachment of silica on the surfaceof titania by using octadecytrichlorosilane (OTS) as a source ofsilica had not been investigated yet. OTS was selected because itis well-known compound used in surface-modification reactionswith surface hydroxyl groups.

Recently, Park and Kang [9] also successfully preparedsilica-coated titanium dioxide by using TEOS as a source ofsilica. In such system, silica is evenly coated on the surface of

0921-5107/$ – see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.mseb.2006.05.003

50 H. Nur / Materials Science and Engineering B 133 (2006) 49–54

Fig. 1. IR spectra of (a) TiO2, (b) TiO2@SiO2, and (c) TiO2–SiO2.

titania with thickness in the range of 10–15 nm. However, in thisstudy, the strategy is different from those previously reported[2–9] because the aim is not to cover titania with silica but toattach silica nanoparticles on the surface of titania. One expectsthat the local environment of titanium surface species of titaniaare modified by attachment of silica. Evenly-coated silica on thesurface of titania could not be achieved by using this method,since the presence of long chain alkyl groups (C18) of OTShinder the formation of silica layer during the surface modifica-tion. To examine the activity Ti surface species, the epoxidationof 1-octene by using aqueous H2O2 was used as the probereaction.

In contrast to well-documented techniques to attach the silicaon the surface of titania, to date, there is very limited under-standing on the correlation between preparation and the natureof the titanium surface species after attachment of silica. It hasbeen generally accepted that functional attachment of silica onthe surface of titania must be controllable, reproducible and tai-lored surface structures. The tailoring of the surface propertiesis especially important for functional attachment. In this paper,attachment of titania particles with silica nanoparticles involves

Table 1Atomic percentage of silicon and titanium of TiO2, TiO2@SiO2, andTiO2–SiO2

a

Element Percentage of titanium and silicon (%)

Ti Si

TiO2 100 0.0TiO2@SiO2 91.3 8.7TiO2–SiO2 97.6 2.5

a Chemical analysis using energy dispersive X-ray analysis (EDAX).

a simple deposition–hydrolysis–calcination of OTS on the sur-face of titania particles. It is demonstrated that the attachmentof silica provides the modification of the local environment oftitanium surface species of titania.

2. Experimental

2.1. Preparation of silica nanoparticles-attached titania

Silica nanoparticles-attached titania catalyst was preparedby modification of the surface of anatase TiO2 (Riedel-deHaen) with octadecyltrichlorosilane (OTS) (Sigma–Aldrich)as a source of silica. The TiO2 powder (6 g) was immersed intoluene (20 ml) containing OTS (3000 �mol) and the suspensionwas stirred for 20 min at room temperature. The suspension wascentrifuged to remove unreacted OTS and wash with toluene(60 ml) and precipitated were dried at 383 K overnight. Then, thesample was ground to obtain a fine powder. In the next step, thesample was calcined at 550 ◦C for 4 h. Now, the silica-attachedtitania was achieved through deposition–hydrolysis–calcinationprocesses. All the steps were repeated nine times, starting fromthe beginning until the accumulation of silica on the surface oftitania was achieved. The silica-attached titania obtained waslabeled as TiO2@SiO2. A smilar procedure was also carriedout to prepare the titania–silica composite (TiO2–SiO2) without

Fig. 2. X-Ray diffractograms pattern of (a) TiO2, (b) TiO2@SiO2, and (c)TiO2–SiO2.

H. Nur / Materials Science and Engineering B 133 (2006) 49–54 51

Fig. 3. SEM photographs of (a) TiO2, (b) TiO2@SiO2, and (c) TiO2–SiO2.

washing with toluene in the above procedure for preparationof TiO2@SiO2. In a typical experiment, OTS (3000 �mol) wasdissolved in toluene (20 ml). A dried titania powder (6 g) wasadded to the solution and the mixture was stirred in the fumecupboard until the solvent was completely dry. Then, the samplewas ground to obtain a fine powder and calcined at 550 ◦Cfor 4 h.

2.2. Characterizations

The TiO2, TiO2@SiO2 and TiO2–SiO2 were characterizedby UV–vis diffuse reflectance (UV–vis DR), X-ray diffraction(XRD), Fourier transform infrared (FTIR), scanning electronmicroscopy (SEM), energy dispersive X-ray analysis (EDAX)and 29Si solid-state magic angle spinning nuclear magnetic reso-nance (MAS NMR) techniques. UV–vis spectra were recordedusing Perkin-Elmer Lambda 900 spectrometer. XRD patternswere acquired on a Bruker D8 Advance instrument using Cu K�radiation (λ = 1.5418 A, kV = 40, mA = 40). SEM photographsand elemental analysis of silica and titanium elements in the

solid particles were obtained by using Philips XL 40 instrument.The 29Si MAS NMR experiments were performed using BrukerAvance 400 MHz 9.4T spectrometer. The spectra were recordedat 79.44 MHz using 4 �s radio frequency pulses, a recycle delayof 60 s and spinning rate of 7.0 kHz using a 4 mm zirconia samplerotor. 29Si MAS NMR chemical shifts were referred to externalTMS at 0 ppm.

2.3. Activity of Ti surface species

To examine the activity Ti surface species, the epoxidationof 1-octene by using aqueous H2O2 was used as the probe reac-tion. It is generally known that silica is not active in this reaction[10–12]. In the epoxidation reaction, 1-octene (4 ml, Aldrich),30% aqueous H2O2 (1 ml) and solid powder (50 mg) were placedin a glass tube, and the reaction was performed with stirringfor 24 h at ambient temperature. The resulting product werewithdrawn and analyzed with gas chromatograph (GC). Gaschromatograph–mass spectrometer (GC–MS) was also used toverify the resulting product.

52 H. Nur / Materials Science and Engineering B 133 (2006) 49–54

3. Results and discussion

3.1. Physical properties

Fig. 1 shows the FTIR spectra of TiO2, TiO2@SiO2 andTiO2–SiO2 particles, respectively. The peaks at 678 and505 cm−1 are attributed to Ti–O bond and the peak at 678 cm−1

refers to symmetric O–Ti–O stretch while 505 cm−1 is due tothe vibration of Ti–O bond. A new distinct and strong peakat 1081 cm−1 can be observed in TiO2@SiO2 which assignedto the asymmetric Si–O–Si stretching vibration indicating thepresence of silica in the TiO2@SiO2. The band of silica at1075 cm−1 can be observed in TiO2–SiO2 but it is less intensecompared to the peak of silica of TiO2@SiO2. An absorptionpeak of Ti–O–Si was observed at 950 cm−1 in TiO2@SiO2particles indicating the Ti–O–Si bond formation [9]. The C–Hpeaks were not observed in the FTIR spectra of TiO2@SiO2 andTiO2–SiO2 due to alkyl groups of OTS have been removed aftercalcinations.

Table 1 shows the percentage of Ti and Si in TiO2,TiO2@SiO2 and TiO2–SiO2 samples. The atomic percentageof Si in TiO2, TiO2@SiO2 and TiO2–SiO2 are 0.0%, 8.7%and 2.5%, respectively. Fig. 2 shows the XRD pattern of TiO2,TiO2@SiO2 and TiO2–SiO2. The major peaks appear at 25.4◦corresponds to the monoclinic phase of titania anatase. TheX-ray diffractograms show that TiO2@SiO2 and TiO2–SiO2possess a similar XRD pattern with bare titania indicating theanatase structure of titania remain unchange after the attachmentof silica. As shown in Fig. 2, a broad hump with maxima at about12◦ which is attributed to amorphous silica is clearly observedin TiO2–SiO2 diffractogram. However, there is no a broad humpabout 12◦ can be observed for TiO2@SiO2 although the amountof silica in TiO2@SiO2 is higher than that of TiO2–SiO2 (seeTable 1). This phenomenon was also observed in the FTIRspectra of TiO2@SiO2 and TiO2–SiO2. The band of silica at1075 cm−1 can be found in TiO2–SiO2 but it is less intensecompared to the silica peak of TiO2@SiO2 (see Fig. 1). Thissuggests that the silica particles attached on the surface of TiO2in TiO2@SiO2 is very well dispersed with very small crystallitesize which cannot be observed by XRD. This observation wassupported by SEM photographs shown in Fig. 3.

Fig. 3 shows the SEM photographs of TiO2, TiO2@SiO2and TiO2–SiO2. It reveals the difference in surface morphologyamong the samples. It is clearly observed the existence of silicananoparticles, with the size in the range of 20–50 nm, attachedon the surface of TiO2@SiO2. However, TiO2 and SiO2 particlesare very difficult to distinguish in TiO2–SiO2 since their size andshape are almost similar.

UV–vis DR spectra of TiO2, TiO2@SiO2 and TiO2–SiO2are shown in Fig. 4. As shown in spectra, the band in the rangeof 230–280 nm is attributed to a charge transfer of the tetrahe-dral titanium sites between O2− and the central Ti (IV) atom,while octahedral Ti was reported appear at around 260–330 nm[13,14]. It shows that the increase in intensity in the range of230–280 nm indicates the occurrence of the transformation ofsome of the octahedral titanium framework to the tetrahedralstructure during the attachment of silica.

Fig. 4. UV–vis spectra of (a) TiO2, (b) TiO2@SiO2, and (c) TiO2–SiO2.

The short-range order structure of the TiO2@SiO2 andTiO2–SiO2 was characterized by 29Si MAS NMR measure-ments (see Fig. 5). As a measure of cross-linkage, the ratio ofQ4 to Q3 sites are attributed to the degree of cross-linking. Ithas been calculated that the ratio Q4 to Q3 sites of TiO2@SiO2

Fig. 5. 29Si MAS NMR spectra of (a) TiO2@SiO2 and (b) TiO2–SiO2.

H. Nur / Materials Science and Engineering B 133 (2006) 49–54 53

and TiO2–SiO2 is almost similar (ca. 1.7). Q denominatesthe presence of the corresponding Si sites in the spectra:Q4 = Si(OSi)4, Q3 = SiOH(OSi)3, and Q2 = Si(OH)2(OSi)2 [15].Interestingly, as shown in Fig. 5, attachment of silica on thesurface of TiO2 in TiO2@SiO2 led to the formation of Q2

site. However, the amount of Q2 type of bonding is only ca.6% due to a partial condensation of Q2 to Q3 and Q4 sitesduring repeated deposition–hydrolysis–calcination of octade-cytrichlorosilane (OTS) processes. Based on the above results,it is suggested that the deposition–hydrolysis–calcination ofOTS to produce silica nanoparticles led to the formation ofSi(OH)2(OSi)2 bonding in the sample, although this type bond-ing is considerably a little amount in TiO2@SiO2.

3.2. Activity of Ti surface species

As can be seen in Fig. 6, all of solid particles showed activityfor epoxidation of 1-octene to give 1,2-epoxyoctane. In particu-lar, attachment of silica nanoparticles (TiO2@SiO2) led to betteractivity than TiO2–SiO2 composite. This phenomenon could beexplained in terms of the local environment of Ti active site. Asdescribed previously, the difference in the UV–vis DR spectraof TiO2, TiO2–SiO2 and TiO2@SiO2 samples is due mainly tocoordination of titanium on the surface of titania. It is generallyaccepted that isolated Ti (IV) in tetrahedral form are consideredthe most active species in epoxidation reaction [10–12]. Basedon these facts, the effect of attachment of silica on increasingthe epoxidation activity of TiO2@SiO2 can be explained by thepresence of isolated Ti (IV) in tetrahedral form. However, theapparent rate of epoxidation over TiO2@SiO2 was much lowerthan those previously reported [16]. One of the reasons for thelow epoxidation activity in TiO2@SiO2 particles is the presenceof few four-coordinate Ti species which are considered to bethe most active species in olefin epoxidation [17]. On the basisof these results, a model of the titanium surface species after

Fig. 6. Epoxidation activity of (a) TiO2, (b) TiO2@SiO2, and (b) TiO2–SiO2.

Fig. 7. The transformation of some of the octahedral titanium to the tetrahedralstructure during the attachment of silica.

attachment of silica nanoparticles is proposed (see Fig. 7). Oneconsiders that titanium tends to bond with Si(OH)2(OSi)2 tomake it stable in tetrahedral form.

4. Conclusions

The silica nanoparticles has been successfully attached onthe surface of titania by deposition–hydrolysis–calcination ofoctadecyltrichlorosilane (OTS). It is demonstrated that the sur-face attachment of titania with silica nanoparticles enhancesthe epoxidation activity of titania. UV–vis DR spectra of thesolid particles showed the increase in intensity of the tetrahedraltitanium in TiO2@SiO2. This indicates the occurrence of thetransformation of some of the octahedral titanium to the tetra-hedral structure during silica attaching of TiO2@SiO2 in whichtitanium tends to bond with Si(OH)2(OSi)2 to make it stable intetrahedral form. Considering that the tetrahedral titanium wasconsidered the most active species in epoxidation of alkene, itcan be concluded that a high epoxidation activity of TiO2@SiO2particle was related to the modification of the local environmentof titanium surface species by attachment of silica nanoparticles.

Acknowledgments

This research was supported by the Ministry of Science Tech-nology and Innovation Malaysia (MOSTI), under IRPA grantno. 09-02-06-10041-EAR and The Academy of Sciences for theDeveloping World, under TWAS Grants in Basic Sciences no.04-462 RG/CHE/AS.

References

[1] M.R. Hoffmann, S.T. Martin, W.Y. Choi, D.W. Bahnemann, Chem. Rev.95 (1995) 69.

[2] C. Anderson, A.J. Bard, J. Phys. Chem. 99 (1995) 9882.

54 H. Nur / Materials Science and Engineering B 133 (2006) 49–54

[3] A.M. El-Toni, S. Yin, T. Sato, Colloids Surf. A: Physicochem. Eng. Aspects274 (2006) 229.

[4] Y. Atou, H. Suzuki, Y. Kimura, T. Sato, T. Tanigaki, Y. Saito, C. Kaito,Physica E 16 (2003) 179.

[5] Q.H. Powell, G.P. Fotou, T.T. Kodas, B.M. Anderson, Y. Guo, J. Mater.Res. 12 (1997) 552.

[6] E. Matijevic, Q. Zhong, R.E. Partch, Aerosol Sci. Technol. 22 (1995) 162.[7] Y.L. Lin, T.J. Wang, Y Jin, Powder Technol. 123 (2002) 194.[8] European Patent Application: 91,301,183.9.[9] O.K. Park, Y.S. Kang, Colloids Surf. A: Physicochem. Eng. Aspects

257–258 (2005) 261.

[10] H. Nur, S. Ikeda, B. Ohtani, J. Catal. 204 (2001) 402.[11] H. Nur, S. Ikeda, B. Ohtani, Chem. Commun. (2000) 2235.[12] H. Nur, D. Prasetyoko, Z. Ramli, S. Endud, Catal. Commun. 5 (2004) 725.[13] A. Zecchina, S. Bordiga, C. Lamberti, G. Ricchiardi, C. Lamberti, G. Ric-

chiardi, D. Scarano, G. Petrini, G. Leofanti, M. Mantegazza, Catal. Today32 (1996) 97.

[14] E. Astorino, J.B. Peri, R.J. Willey, G. Busca, J. Catal. 157 (1996) 482.[15] J.P. Rainho, J. Rocha, L.D. Carlos, R.M. Almeida, J. Mater. Res. 16 (2001)

2369.[16] T. Tatsumi, K.A. Koyano, N. Igarashi, Chem. Commun. (1998) 325.[17] B. Notari, Adv. Catal. 41 (1996) 253.

On the location of different titanium sites in Ti–OMS-2 andtheir catalytic role in oxidation of styrene

Hadi Nur *, Fitri Hayati, Halimaton Hamdan

Ibnu Sina Institute for Fundamental Science Studies, Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia

Received 20 December 2006; received in revised form 30 March 2007; accepted 2 April 2007Available online 10 April 2007

Abstract

Octahedral manganese oxide molecular sieves (OMS-2) modified by impregnation of TiO2 exhibit a higher catalytic activity for oxi-dation of styrene with tert-butylhydroperoxide in comparison to titanium-incorporated OMS-2, where the styrene conversions were ca.70% and 45–50%, respectively. The framework of titanium species has no effect on the enhancement of catalytic activity, while the non-framework of titanium species induces a synergetic effect that enhances the oxidation of styrene with tert-butylhydroperoxide.� 2007 Elsevier B.V. All rights reserved.

Keywords: Titanium-incorporated octahedral manganese oxide molecular sieves; Non-framework titanium; Oxidation of styrene

1. Introduction

Octahedral manganese oxide molecular sieves (OMS-2)are currently considered as one of the potential catalystsin oxidation of alcohols and olefins [1–3]. OMS-2 materialsare manganese oxide (MnOx; x = 1.85–2.00) with a frame-work structure consisting of 2 · 2 type tunnels, built up ofMnO6 octahedra with a pore diameter of 4.6 A [4].Although, recently, the oxidation of styrene has beenreported by using Me–OMS-2 (Me = Fe, Cu, Ni and Co)[5], it is desirable to study the catalytic performance ofOMS-2 material with the addition or incorporation of tita-nium. An understanding of the synthetic methods of cata-lyst preparation is needed for the precise control of thestructure and location of catalytically active sites onOMS-2. Here, we report on the study of the effect of thelocation of titanium sites, either in the framework ornon-framework of OMS-2, and their catalytic role in theoxidation of styrene with tert-butyl hydroperoxide (TBHP)as the oxidant. X-ray powder diffraction (XRD), photolu-

minescence, surface area and pyridine adsorption measure-ments were used to characterize these samples. Thecatalytic activity of Ti containing OMS-2 was also com-pared with TS-1 as a reference catalyst.

2. Experimental

2.1. Synthesis

OMS-2 was prepared by a precipitation method accord-ing to [6]. A 0.4 M solution of KMnO4 (13.3 g in 225 ml ofdeionized water) was added to a mixture of a 1.75 M solu-tion of MnSO4 Æ H2O (19.8 g in 67.5 ml deionized water)and 6.8 ml of concentrated HNO3. The resulting black pre-cipitate was stirred vigorously and refluxed at 373 K for24 h. The precipitate was filtered and washed with deion-ized water until neutral pH and dried at 393 K. This gaveOMS-2. Titanium incorporated OMS-2 (Ti–OMS-2) wasprepared by the stepwise addition of a KMnO4 solution(13.3 g in 225 ml of deionized water) to different amountsof Ti2SO4 (25–75 ml) (15% v/v in H2SO4) in order to pro-duce Ti–OMS-2 in the Ti/Mn ratio of 0.18, 0.43 and 0.67(as analyzed by using an atomic absorption spectrometer).Upon completion, the mixture was stirred, refluxed,

1566-7367/$ - see front matter � 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.catcom.2007.04.002

* Corresponding author. Tel.: +60 7 5536077; fax: +60 7 5536080.E-mail address: hadi@kimia.fs.utm.my (H. Nur).URL: http://www.ibnusina.utm.my/~hadi (H. Nur).

www.elsevier.com/locate/catcom

Available online at www.sciencedirect.com

Catalysis Communications 8 (2007) 2007–2011

filtered, washed, and dried following the above procedureand it was labeled as Ti–OMS-2 (0.18), Ti–OMS-2 (0.43)and Ti–OMS-2 (0.67), where the number in parentheses isthe molar ratio of Ti/Mn. Titanium(IV) tetra-2-propoxideðTiðOPri4Þ was impregnated from its toluene solution intoOMS-2 powder and calcined at 773 K for 3 h. Here, thismodified OMS-2 is called Ti–OMS-2 (imp). The molaramount of Ti calculated to give the molar ratio of Ti/Mnwas 0.18. For comparison, a mechanical mixture of TiO2

and OMS-2 was prepared by the addition of a calculatedamount of TiO2 (rutile) powder to OMS-2 with a Ti/Mnmolar ratio of 0.67. This catalyst was labeled as TiO2–OMS-2 (mix). Table 1 summarizes the preparation method,the chemical composition and labeling of titanium contain-ing OMS-2. In this study, TS-1 was also used as a referencecatalyst. TS-1 (2 mol% of titanium) was prepared accord-ing to a procedure described earlier [7,8].

2.2. Characterization

All samples were characterized by powder XRD for thecrystallinity and phase content of the solid materials usinga Shimadzu XRD 6000 diffractometer with the CuKak = 1.5405 A) radiation as the diffracted monochromaticbeam at 30 kV and 30 mA. Atomic absorption analysis(AAS) was employed for elemental composition analysesof manganese and titanium in the sample. A Perkin–Elmermodel Analyst 400 spectrophotometer was used to carryout the analyses. The acidity of the solids was characterizedby an absorbed base probe molecule. A wafer of the sample(10–12 mg) was locked in the cell equipped with CaF2 win-dows, and evacuated at 400 �C under vacuum condition for4 h. Pyridine as a probe molecule was introduced into theevacuated sample at room temperature. The IR spectraof the sample were monitored at room temperature afterdesorption of pyridine at 150 �C for 1 h. Photoluminescence

spectra were recorded in air at room temperature on a Per-kin–Elmer LS 55 spectrometer. The emission spectraobserved at an excitation wavelength was 430 nm.

2.3. Catalytic testing

Oxidation of styrene was carried out using the abovecatalysts. Styrene (5 mmol), 70% aqueous ter-butyl hydro-peroxide (TBHP) (10 mmol), catalyst (50 mg) and acetoni-trile (15 ml) as solvent were placed in a round-bottomedflask with a reflux condenser and the reaction was per-formed with stirring at 70 �C in an oil bath. The productswere collected in a period of time and analyzed by GCand GC–MS.

3. Results and discussion

XRD patterns of OMS-2, Ti–OMS-2 (0.18) and Ti–OMS-2 (0.43) show that the samples are pure and highlycrystalline and matched those of cryptomelane Q [9]; thenatural counterpart of OMS-2 material (see Fig. 1). Theresults confirmed that OMS-2, Ti–OMS-2 (0.18) and Ti–OMS-2 (0.43) materials consist of the cryptomelane struc-ture: 2 · 2 tunnels with a pore size of 4.6 A, composed ofdouble chains of edge sharing and corner sharing MnO6

octahedra [4]. The absence of other peaks in the XRD pat-terns except the cryptomelane peaks suggested that Ti wassuccessfully incorporated in the framework of Ti–OMS-2.In order to confirm the successful incorporation of tita-nium, the XRD patterns of Ti–OMS-2 were compared withthe mechanical mixture of TiO2 (rutile) and OMS-2 wherethe mixture (TiO2–OMS-2 (mix)) showed the presenceof the rutile phase (see Fig. 1 and Table 1). If Ti is success-fully incorporated in the framework of OMS-2, one expectsthat the bigger the substituted atom is, the bigger is the unitcell volume. Calculation of the unit cell volume of OMS-2

Table 1Chemical composition and physicochemical properties of OMS-2, Ti–OMS-2 and TiO2–OMS-2

Samples Molar ratioof Ti/Mna

Molaramount ofMnb

Molaramount ofTib

Methods ofintroduction of Tispecies

Location of Tispecies

Structure of non-framework Ti species

Surface area(m2 g�1)

OMS-2 0.00 552.2 0.0 – – – 155Ti–OMS-2

(0.18)0.18 412.5 74.4 Direct synthesis Framework – 152

Ti–OMS-2(0.43)

0.43 420.4 181.4 Direct synthesis Framework – 149

Ti–OMS-2(0.67)

0.67 342.0 228.2 Direct synthesis Framework andnon-framework

Rutile 152

TiO2–OMS-2 (imp)c

0.18 552.0 99.4 Impregnation Non-framework Amorphous N.d.e

TiO2–OMS-2 (mix)d

0.67 552.2 369.9 Mechanical mixing Non-framework Rutile N.d.e

a Analysis was carried out by atomic absorption spectrometer.b The amount of OMS-2, Ti–OMS-2 and TiO2–OMS-2 were 50 mg.c Titanium(IV) tetra-2-propoxide ðTiðOPri4Þ was impregnated from its toluene solution into OMS-2 powder and calcined at 773 K for 3 h.d Catalyst was prepared by addition of calculated amount of Ti from TiO2 powder to OMS-2.e Not determined.

2008 H. Nur et al. / Catalysis Communications 8 (2007) 2007–2011

(271 A3) and Ti–OMS-2 (0.43) (277 A3) shows that the unitcell volumes increase on incorporation of Ti in the frame-work of OMS-2. The lattice enlargement originates fromthe replacement of the smaller ionic Mn4+ (ionic radius is0.53 A) by the relatively larger ionic Ti4+ (ionic radius is0.61 A). This shows a strong evidence of isomorphous sub-stitution of Mn atoms by Ti into the framework ofTi–OMS-2. Moreover, it can be found that with the incor-poration of Ti, the full width at half maximum decreasesindicating an increase in the grain size. It is observed thatthe peaks of the rutile phase of TiO2 appear at a Ti/Mnratio higher than ca. 0.5 and are not observed at a Ti/Mnratio less than ca. 0.5 (see Fig. 1). By considering the upperlimit of the titanium that can be incorporated into theframework, one would expect non-framework titaniumspecies to be formed when the Ti/Mn ratio reached ca.

0.5. The amount of Ti located in the non-framework isca. 25% in Ti–OMS (0.67). This argument is supported by

the presence of the rutile phase of TiO2 in Ti–OMS-2(0.67) (see Fig. 1). However, no reflection for the rutilephase of TiO2 is observed in TiO2–OMS-2 (imp) wherethe catalyst is prepared by the impregnation method. Thisresult implies that the structure of TiO2 in TiO2–OMS-2(imp) is in the amorphous form.

The photoluminescence (PL) was also used to confirmthe absence of non-framework TiO2, because non-frame-work TiOx with a very small crystallite size cannot bedetected by XRD. The PL spectra are useful to disclosethe efficiency of charge carrier trapping, migration andtransfer, and to understand the nature of electron–holepairs in TiO2 semiconductor particles since PL emissionresults from the recombination of photo-excited free carri-ers [10]. In this study, the 430 nm excited PL spectra of allpressed-powder samples at room temperature were exam-ined in the range of 560–680 nm. The PL spectra ofOMS-2, Ti–OMS-2 (0.43) and TiO2–OMS-2 (mix) areshown in Fig. 2. The results indicated that the photolumi-nescence intensity of TiO2–OMS-2 (mix) (kem(max)�600 ± 10 nm; FWHM � 40 nm) was substantially higherthan that of OMS-2 and Ti–OMS-2 (0.43) (see Fig. 2). Therelative intensity of Ti–OMS-2 (0.43) and TiO2–OMS-2(mix) is similar suggesting that there are no TiOx particlesexist in Ti–OMS-2 (0.43) sample. This result reinforces oursuggestion that Ti is incorporated in the framework of Ti–OMS-2. The incorporation of Ti in the framework ofOMS-2 was also further supported by the surface areaanalysis (see Table 1). It is revealed that the surface areaof OMS-2 and Ti–OMS-2 is almost the same.

Reaction products of oxidation of styrene using TBHPas the oxidant catalyzed by TiO2 (rutile phase), OMS-2,TiO2–OMS-2 and Ti–OMS-2 were analyzed by GC. Themajor products in this reaction proved to be benzaldehyde,styrene oxide and phenylacetaldehyde. The selectivitiestowards benzaldehyde, styrene oxide and phenylacetalde-hyde as the reaction products are shown in Fig. 3. As

2θ / o30 402010 50 60 70

= TiO2 (rutile)**

** (g)

*

(e)

(d)

(c)

(b)

Rel

ativ

e in

tens

ity /

a.u.

(f)

(a)

* ** *

Fig. 1. X-ray diffractograms of (a) cryptomelane (JCPDS 29, 102), (b)OMS-2, (c) Ti–OMS-2 (0.18), (d) Ti–OMS-2 (0.43), (e) Ti–OMS-2 (0.67),(f) TiO2–OMS-2 (imp) and (g) Ti–OMS-2 (mix).

560 580 600 620 640 660 680

Wavelength / nm

Inte

nsity

/ a.

u.

TiO2-OMS-2 (mix)

Ti-OMS-2 (0.43)

OMS-2

Fig. 2. Photoluminescence spectra of OMS-2, Ti–OMS-2 (0.43) and TiO2–OMS-2 (mix). The excitation wavelength is 430 nm.

H. Nur et al. / Catalysis Communications 8 (2007) 2007–2011 2009

indicated in Fig. 3, the reaction catalyzed by all the cata-lysts produced the highest yield of benzaldehyde and theirselectivities towards the formation of products are almostsimilar to each other. The high selectivity (52%) towardsbenzaldehyde over TS-1 was surprising since that phenyl-acetaldehyde and benzaldehyde were the major productsfrom the oxidation of styrene catalyzed by the TS-1 zeolite[10], while a speculated product, styrene oxide, was notdetected. Brønsted acid sites originating from frameworktitanium species catalyze the rearrangement of the interme-diate, leading to the formation of phenylacetaldehyde [11].This argument is in agreement with our oxidation of sty-rene over TS-1, since no Brønsted acid sites have beendetected in our TS-1 which has been analyzed by the pyri-dine adsorption method [7]. IR spectrum of acidity studyby pyridine adsorption after evacuation under a vacuumat 400 �C and 150 �C revealed in Fig. 4a shows that Lewisacid sites are formed in Ti–OMS-2 (0.43) as indicated bythe appearance of peaks at 1447 cm�1, 1489 cm�1 and1604 cm�1. In contrast, no peaks are observed for theOMS-2 sample in Fig. 4b. The absence of peaks at1540 cm�1 confirms that there are no Brønsted acid sitesin both samples. This is the possible reason why TiO2–OMS-2, Ti–OMS-2, TiO2 and OMS-2 catalysts are notselective towards phenylacetaldehyde. A high selectivitytowards benzaldehyde may be due to OMS-2 and TiO2

promoting the carbon–carbon bond cleavage, thus result-ing in the formation of benzaldehyde.

As shown in Fig. 3, a considerable increase in the con-version of styrene over Ti–OMS-2, OMS-2, TiO2–OMS-2,TiO2 and TS-1 after 3 h of the reactions is clearly observedwhen Ti–OMS-2 (0.67) and TiO2–OMS-2 (imp) are used ascatalysts. The increase in oxidation activity of Ti–OMS-2

(0.67) and TiO2–OMS-2 (imp) can be explained on thebasis of the presence of non-framework titanium species.The superior performance of Ti–OMS-2 (imp) and Ti–OMS-2 (0.67) strongly suggests the occurrence of a syner-getic effect of non-framework Ti with OMS-2, since Ti–OMS-2 (mix), a mechanical mixture of TiO2 and OMS-2,gives a relatively lower conversion of styrene (see Fig. 3).Leaching is a particular problem of solid catalysts in liquidphase reactions. The catalysts were recycled three times.The activities of the recovered and dried Ti–OMS-2,OMS-2 and TiO2–OMS-2 showed an insignificant change(ca. 3–5%) that corresponds to experimental observationswithin experimental error in their catalytic activity in the

0

10

20

30

40

50

60

70

Con

vers

ion

/ %

Phenylacetaldehyde

Styrene oxide

Benzaldehyde

Product selectivity (%):

125

83

168

763315

52

10

18

72

9

20

71

8

16

76

8

18

74

16

15

69

Blank TiO2 TS-1 OMS-2 Ti-OMS-2 TiO2 -OMS-2 TiO2 -OMS-2

Ti/Mn = 0.18

Ti/Mn = 0.43

Ti/Mn = 0.67 Ti/Mn = 0.18

Ti/Mn = 0.67

14

18

68

Fig. 3. The conversion and product selectivity of oxidation styrene with tert-butyl hydroperoxide (TBHP) using TiO2, TiO2–OMS-2, Ti–OMS-2, OMS-2and TS-1. All reactions were carried out at 70 �C with styrene (5 mmol), 70% aqueous TBHP (10 mmol), acetonitrile (15 ml) and catalyst (50 mg) withvigorous stirring.

1640 1600 1560 1520 1480 1440 1400

1604 1447

1489

Wavenumber / cm-1

Abs

orba

nce

/ a.u

.

(a)

(b)

Fig. 4. FTIR spectra of (a) Ti–OMS-2 (0.43) and (b) OMS-2 afterevacuation under vacuum at 400 �C for 4 h followed by pyridineadsorption at room temperature and evacuation at 150 �C for 1 h.

2010 H. Nur et al. / Catalysis Communications 8 (2007) 2007–2011

recycling test. This suggests a good regenerability of thecatalysts in the oxidation of styrene with tert-butylhydroperoxide.

The results of this investigation lead to the conclusionthat the framework of titanium species in Ti–OMS-2 hasno effect on the enhancement of catalytic activity, whilethe existence of Ti in the non-framework structure ofOMS-2 induces a synergetic effect that enhances the cata-lytic activity in the oxidation of styrene with tert-buty-lhydroperoxide. Ti–OMS-2 containing non-framework Tiis highly active (ca. 70% conversion of styrene) and selec-tive towards the oxidation of styrene to give benzaldehyde(68%) as the main product. Further detailed studies are,however, necessary to understand the interactions betweenTiO2 and OMS-2 support and the reaction mechanism.

Acknowledgements

This research was supported by the Ministry of ScienceTechnology and Innovation Malaysia (MOSTI), underIRPA grant and The Academy of Sciences for the Developing

World, under the TWAS Grants in Basic Sciences no. 04-462 RG/CHE/AS.

References

[1] V.V. Krishnan, S.L. Suib, J. Catal. 184 (1999) 305.[2] Y.-C. Son, V.D. Makwana, A.R. Howell, S.L. Suib, Angew. Chem.,

Int. Ed. 40 (2002) 4280.[3] R. Ghosh, Y.-C. Son, S.L. Suib, J. Catal. 224 (2004) 288.[4] C.L. O’Young, Y.F. Shen., R.N. DeGuzman, S.L. Suib., United

States Patent, 5,702,674 (1997).[5] R. Ghosh, X. Shen, J.C. Villegas, Y. Ding, K. Malinger, S.L. Suib, J.

Phys. Chem. B 110 (2006) 7592.[6] R.N. DeGuzman, Y.F. Shen, S.L. Suib, B.R. Shaw, C.L. O’Young,

Chem. Mater. 5 (1993) 1359.[7] H. Nur, D. Prasetyoko, Z. Ramli, S. Endud, Catal. Commun. 5

(2004) 725.[8] D. Prasetyoko, Z. Ramli, S. Endud, H. Nur, J. Mol. Catal. A 241

(2005) 118–125.[9] Database of Joint Committee for Powder Diffraction Studies

(JCPDS) PDF# 29, 102.[10] K. Fujihara, S. Izumi, T. Ohno, M. Matsumura, J. Photochem.

Photobiol. A 132 (2000) 99.[11] J. Zhuang, D. Maa, Z. Yan, X. Liu, X. Hana, X. Bao, Y. Zhang, X.

Guo, X. Wang, Appl. Catal. A 258 (2004) 1.

H. Nur et al. / Catalysis Communications 8 (2007) 2007–2011 2011

Hindawi Publishing CorporationInternational Journal of PhotoenergyVolume 2007, Article ID 98548, 6 pagesdoi:10.1155/2007/98548

Research ArticleStannic Oxide-Titanium Dioxide CoupledSemiconductor Photocatalyst Loaded with Polyanilinefor Enhanced Photocatalytic Oxidation of 1-Octene

Hadi Nur, Izan Izwan Misnon, and Lim Kheng Wei

Ibnu Sina Institute for Fundamental Science Studies, Universiti Teknologi Malaysia, Skudai 81310 UTM , Johor, Malaysia

Received 12 March 2007; Accepted 4 June 2007

Recommended by Hisao Hidaka

Stannic oxide-titanium dioxide (SnO2–TiO2) coupled semiconductor photocatalyst loaded with polyaniline (PANI), a conductingpolymer, possesses a high photocatalytic activity in oxidation of 1-octene to 1,2-epoxyoctane with aqueous hydrogen peroxide.The photocatalyst was prepared by impregnation of SnO2 and followed by attachment of PANI onto a TiO2 powder to give samplePANI-SnO2–TiO2. The electrical conductivity of the system becomes high in the presence of PANI. Enhanced photocatalyticactivity was observed in the case of PANI-SnO2–TiO2 compared to PANI-TiO2, SnO2–TiO2, and TiO2. A higher photocatalyticactivity in the oxidation of 1-octene on PANI-SnO2–TiO2 than SnO2–TiO2, PANI-TiO2, and TiO2 can be considered as an evidenceof enhanced charge separation of PANI-SnO2–TiO2 photocatalyst as confirmed by photoluminescence spectroscopy. It suggeststhat photoinjected electrons are tunneled from TiO2 to SnO2 and then to PANI in order to allow wider separation of excitedcarriers.

Copyright © 2007 Hadi Nur et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1. INTRODUCTION

Photocatalysis over titanium dioxide (TiO2) is initiated bythe absorption of a photon with energy equal to or greaterthan the band gap of TiO2 (3.2 eV), producing electron-hole(e−/h+) pairs [1]:

TiO2ecb

−−−−−−→ (TiO2

)+ hvb

+(TiO2). (1)

Consequently, following irradiation, the TiO2 particle can actas either an electron donor or acceptor for molecules in thesurrounding media. However, the photoinduced charge sep-aration in bare TiO2 particles has a very short lifetime be-cause of charge recombination. Therefore, it is importantto prevent electron-hole recombination before a designatedchemical reaction occurs on the TiO2 surface. TiO2 andhigh-recombination rate of the photogenerated electron-hole pairs hinders its further application in industry. Havingrecognized that charge separation is a major problem, nu-merous efforts have been attempted to improve its photocat-alytic activity by modifying the surface or bulk properties ofTiO2, such as deposition of metals, doping, surface chelation,and coupling of two semiconductors [2–4]. Among the cou-pled semiconductor photocatalysts, many efforts have been

devoted to the SnO2−TiO2 system [5, 6]. Here, SnO2−TiO2

coupled semiconductor photocatalyst loaded with PANI, aconducting polymer, has been studied as photocatalyst in theoxidation of 1-octene with aqueous hydrogen peroxide. Oneexpects that the attachment of polyaniline (PANI) on thesurface of SnO2−TiO2 composite will reduce the electron-hole recombination during the photocatalytic oxidation of1-octene due to PANI’s electrical conductive properties.

2. EXPERIMENTAL

2.1. Preparation of photocatalysts

2.1.1. PANI-TiO2

Preparation of PANI-TiO2 was done by using 360 µl of ani-line which is dissolved in 20 ml distilled water containing3 ml 30% hydrochloric acid. The solution was precooled at4◦C and 2 g of anatase TiO2 was added, and the suspen-sion was stirred for 30 minutes and allowed to stand forfurther period of 60 minutes. Then, 4 ml of potassium per-sulphate as initiator was added in the solution and stirredthoroughly until we can see green pale color. The resultant

2 International Journal of Photoenergy

product obtained was filtered, wash thoroughly with water,and dried till it shows constant weight at 60◦C. The polymerprepared in such manner is commonly termed as PANI-HCl(Emeraldine-salt). The amount of PANI in PANI-TiO2 wasca. 2 wt% as analyzed by thermogravimetry analyzer.

2.1.2. SnO2–TiO2

Here, 0.285 g stannous chloride was added to 10 mlmethanol. Then, 3 g of anatase TiO2 was mixed into the so-lution and stirred for 6 hours until the solvent is completelydry. The sample was calcined at 500◦C for 4 hours. The sam-ple was labeled as SnO2−TiO2. The molar amount of SnO2

in 1 g TiO2 was 500 µmol.

2.1.3. PANI-SnO2–TiO2

Stannic oxide-titanium dioxide (SnO2−TiO2) coupled semi-conductor photocatalyst loaded with polyaniline (PANI) wasprepared in two steps by a procedure similar to that prepara-tion of SnO2−TiO2 and PANI-TiO2. First, stannous chloridewas impregnated into TiO2 to give sample SnO2−TiO2 (seeSection 2.1.2). In the second step, PANI was attached on thesurface of SnO2−TiO2 particle (see Section 2.1.1). This mod-ified SnO2−TiO2 is called PANI-SnO2−TiO2.

2.2. Characterizations

The physical characteristics of TiO2, SnO2−TiO2, and PANI-SnO2−TiO2 particles are experimentally studied by X-raydiffraction (XRD), UV-Vis diffuse reflectance (UV-Vis DR),scanning electron microscopy (SEM), photoluminescencespectroscopy, and electrical conductivity analysis techniques.Electrical conductivity was carried by an impedance ana-lyzer. Pellets of 2–4 mm thick were prepared by placing suf-ficient amount of sample (50 mg) in a steel die measuring13 mm in diameter, and a pressure of 5 tons were appliedand held for 30 seconds. The electric properties of the pelletprepared were measured by AC impedance spectroscopy us-ing frequency response analyzer (Autolab POST AT 30) in thefrequency range from 0.01 Hz to 1 MHz with an applied volt-age of 10 mV. The impedance data were fitted by GPES soft-ware. Luminescence spectra were recorded on Perkin-ElmerLS 55 spectrometer. About 0.04 g of the sample was placedon a sample holder. After locating and locking sample holderin a proper place in the analyzer, samples were measured inthe emission λ (wavelength) scale of 200–900 nm at excita-tion λ = 333 nm.

2.3. Photocatalytic oxidation of 1-octene

Photocatalytic reactions were carried out in Pyrex glass tubes(φ = 10 mm) that contained photocatalyst powder (50 mg),1-octene (0.85 ml) and 30% hydrogen peroxide (1.5 ml). Inthis experiment, H2O2 was added in the photocatalytic re-action system because it has been reported that photocat-alytic oxidation of some compounds is activated by additionof H2O2 [7]. During the photocatalytic reaction, the pho-tocatalyst particles were suspended in the solution using a

PANI

34603230

TiO2

TiO2-PANI

SnO2-TiO2

SnO2-TiO2-PANI

1561 1296

1498

1173

673 505

4000 3000 2000 1500 1000 400

Wavenumber (cm−1)

Tran

smit

tan

ce(a

.u.)

Figure 1: IR spectra of photocatalysts.

magnetic stirrer. The suspension was photoirradiated using a450-Watt Hanovia high-pressure mercury-vapor lamp (AceGlass Co., Vineland, NJ). The light source was placed 8 cmaway from the reactor. The temperature of the reactor solu-tions was maintained at 4◦C throughout the experiments byusing a water circulation system equipped with chiller. Thephotocatalytic reactions were carried out for 20 hours. Theproducts were analyzed with a gas-liquid chromatograph(ThermoFinnigan Trace GC). The products were identifiedby coinjecting the corresponding authentic samples into thecolumns.

3. RESULTS AND DISCUSSION

3.1. Physical properties

The XRD patterns of TiO2, PANI-TiO2, and SnO2−TiO2

(data not shown) demonstrated that all the modified cat-alysts possess a similar crystalline structure which is corre-sponded to anatase phase of TiO2. Figure 1 shows the FTIRspectra of PANI, TiO2, TiO2-PANI, SnO2−TiO2, and PANI-SnO2−TiO2. The broad peaks around 500–680 cm−1 of TiO2,TiO2-PANI, SnO2−TiO2, and PANI-SnO2−TiO2 are due tothe Ti−O bond in the TiO2. The peak at 673 cm−1 refers tosymmetric O−Ti−O stretch while 505 cm−1 is due to the vi-bration of Ti−O bend. As shown in Figure 1, the main char-acteristic peaks of PANI are assigned as follows: the band at3460 and 3230 cm−1 can be attributed to the free (nonhy-drogen bonded) N−H stretching vibration and hydrogen-bonded N−H bond between amine and imine sites, C=N

Hadi Nur et al. 3

339

219

319

TiO2

SnO2-TiO2

190 250 300 350 400 450 500 550 600

Wavelength (nm)

K-M

(a.u

.)

Figure 2: UV-Vis DR spectra of photocatalysts in the range of wave-length from 190 to 600 nm.

and C=C stretching modes for the quinonoid and benzenoidunits occur at 1561 and 1498 cm−1, the band at 1296 cm−1

has been attributed to C−N stretching mode for benzenoidunit, while the strong band at 1173 cm−1 was considered tobe a measure of the degree of electron delocalization and thusit is a characteristic peak of PANI conductivity [8]. The peaksat 1173 and 1498 cm−1 were observed in TiO2-PANI andPANI-SnO2−TiO2 samples thereby endorsing the presenceof PANI in the TiO2-PANI and PANI-SnO2−TiO2. How-ever, the others low-intense PANI’s peaks were not observedin PANI-TiO2 and PANI-SnO2−TiO2 since the amount ofPANI was only ca. 2 wt% in these samples as analyzed bythermogravimetry analyzer.

The UV-Vis spectra of TiO2, PANI-TiO2, and SnO2−TiO2

are shown in Figures 2 and 3. As revealed in Figure 2, theband in the range of 200–240 nm is attributed to a charge-transfer of the tetrahedral titanium sites between O2− andthe central Ti(IV) atoms, while octahedral Ti was reportedappear at around 260–330 nm [9]. The UV-Vis spectrumof SnO2−TiO2 (see Figure 3), on the other hand, showsabsorption at higher wavelengths (339 nm) due to the pres-ence of hexa-coordinated Sn species or polymeric Sn−O−Snunits [10].

It is generally known that the peak of PANI could be ob-served in the range of wavelength from 400 to 1400 nm. Asshown in Figure 3, PANI-TiO2 and PANI-SnO2−TiO2 sam-ples show two characteristic absorptions peaks at 428 and

794

428

PANI-TiO2

PANI-SnO2-TiO2

400 600 800 1000 1200 1400

Wavelength (nm)K

-M(a

.u.)

Figure 3: UV-Vis DR spectra of photocatalysts in the range of wave-length from 400 to 1400 nm.

TiO2 SnO2-TiO2 PANI-SnO2-TiO2

3.5

3

2.5

2

1.5

1

0.5

0

Ele

ctri

calc

ondu

ctiv

ity

(Sm−1

)

×10−6

Figure 4: Electrical conductivity of photocatalysts.

794 nm. These peaks originate from the charged cationicspecies known as polarons [11, 12].

Considering that PANI is a conducting polymer, itis of interest to check the electrical conductivity ofPANI-SnO2−TiO2. The electrical conductivity of TiO2 andSnO2−TiO2 were increased significantly due to the presenceof PANI in the composite (see Figure 4). Electrical transfer isobviously associated with the PANI attached on the surfaceof PANI-SnO2−TiO2. However, the electrical conductivity ofPANI-SnO2−TiO2 is much lower compared than the resultobserved by Bian and Xue for PANI attached on the surfaceof TiO2 [13]. They observed that the electrical conductiv-ity was 0.5 S cm−1. Our observations are different because

4 International Journal of Photoenergy

TiO2

(a)

TiO2

SnO2

(b)

TiO2

SnO2

PANI

(c)

Figure 5: SEM photographs of (a) TiO2, (b) SnO2−TiO2, and (c)PANI-SnO2−TiO2.

the amount of PANI attached on the TiO2 particles differsfrom those used by Bian and Xue. The amount of PANI re-ported by Bian and Xue was 60 wt%. Instead, we decreasedthe PANI content thirty fold. Clearly these changes influencethe electrical conductivity of PANI-SnO2−TiO2. One expectsthat by using a large amount PANI, the surface of TiO2 par-ticle is covered by PANI. In this study, the amount of PANIin PANI-SnO2−TiO2 was only ca. 2 wt% as analyzed by ther-mogravimetry analyzer, so that TiO2 particle is only coveredby PANI with a lower coverage compared to those previouslyreported [13]. Since the photocatalytic reaction is occurredon the surface TiO2, it is necessary to use a small amount ofPANI to avoid coating TiO2 with PANI.

Figure 5 shows the SEM photographs of TiO2,SnO2−TiO2, and PANI-SnO2−TiO2. It reveals the dif-ference in surface morphology among the samples. It isclearly observed the existence of SnO2 nanoparticles at-tached on the surface of SnO2−TiO2. SEM photographs ofSnO2−TiO2 (see Figure 5(b)) and PANI-SnO2−TiO2 (seeFigure 5(c)) shows that the amount of particles stickingon the surface of PANI-SnO2−TiO2 is higher than those

TiO2 TiO2 PANI-TiO2 SnO2-TiO2 PANI-SnO2-TiO2

140

120

100

80

60

40

20

0

1,2-

epox

yoct

ane

(µm

ol)

Darkcondition

Under UV light

Figure 6: The photocatalytic epoxidation of 1-octene over PANI-SnO2−TiO2, SnO2−TiO2, PANI-TiO2, and TiO2. All reactions werecarried out at 4◦C under dark and UV light irradiation conditions.

of SnO2−TiO2 confirming the existence of PANI in PANI-SnO2−TiO2. However, SnO2 and PANI particles are verydifficult to distinguish in PANI-SnO2−TiO2 since their sizeand shape are almost similar. Based on the above results, itcan be concluded that PANI is successfully attached on thesurface of TiO2 and SnO2−TiO2.

3.2. Photocatalytic activity and radiativerecombination process

Gas chromatography analyses indicated that 1,2-epoxyoc-tane was the sole product, and other expected by-products,such as 2-octanone, 1-octanol, 2-octanol, or 1,2-octanediol,were not detected. As shown in Figure 6, a considerable in-crease in 1,2-epoxyoctane yield was observed when the re-action was carried under UV irradiation, suggesting thatthe photocatalytic action was occurred on the surface ofTiO2. A similar trend was also observed when the surface ofTiO2 was attached with SnO2, and PANI. It is observed thatthe photocatalytic activity in decreasing order of the yieldof 1,2-epoxyoctane as the following: PANI-SnO2−TiO2 >SnO2−TiO2 > PANI-TiO2 > TiO2. The most active photocat-alyst, PANI-SnO2−TiO2, combined three materials, PANI,SnO2 and TiO2 which complemented each other in photo-catalytic reaction as described below.

Semiconductors such as TiO2 and SnO2 can be excited byphotons with appropriate energy to produce the photogener-ated electron or hole pairs [1]. It was reported that the highphotocatalytic activity arose from the increase in the reactionefficiency due to the electron transfer from TiO2 to SnO2.The valence band (VB) edge of SnO2 (+3.67 V) is much lowerthan that of TiO2 (+2.87 V) [1, 6]. Since holes move in theopposite direction from the electrons, they can be trapped inthe TiO2. As such, charge separation is increased so recom-bination is reduced. In order to confirm this argument, thephotoluminescence (PL) experiments were carried out.

The PL emission spectra have been widely used to in-vestigate the efficiency of charge carrier trapping, immigra-tion, and transfer, and to understand the fate of e−/h+ pairsin semiconductor particles [14]. As depicted in Figure 7, itis observed that the maximum of the PL emission peaks of

Hadi Nur et al. 5

TiO2

SnO2-TiO2

PANI-SnO2-TiO2

350 400 450 500 550

Wavelength (nm)

Inte

nsi

ty(a

.u.)

Figure 7: Photoluminescence (PL) spectra of photocatalysts.

CB −CB − e−

PANI

OH−

OH+

+

TiO2 hνSnO2 hν

VB

+

VB+

O

Figure 8: The proposed mechanism of photocatalytic epoxidationof 1-octene over PANI-SnO2−TiO2.

the pure TiO2 and SnO2−TiO2 occurred at 385 nm, whilethe maximal emission for the PANI-SnO2−TiO2 appeared at380 nm. One suggests that the interaction between the con-jugated polymer chains and the surface of the TiO2 particlesplayed a key role in the resulting blue shift [15]. Comparedwith the spectrum of TiO2, it is found that the intensity of thePL spectra of the PANI-SnO2−TiO2 and SnO2−TiO2 in thesame range of wavelength are much lower than that of pureTiO2. This indicates the decrease of the radiative recombina-tion process in these samples. The intensity of the PL spectrain decreasing order as the following: TiO2 > SnO2−TiO2 >PANI-SnO2−TiO2 (see Figure 7). On the basis of these re-sults, a model of the mechanism of photocatalytic epoxida-tion of 1-octene using PANI-SnO2−TiO2 system is proposed(see Figure 8). By attaching SnO2 and PANI on the surfaceof TiO2, it is possible to drive the photogenerated electronsfarther away from the TiO2, thereby achieving more efficientcharge separation in these semiconductor particles. As a re-sult, a more active photocatalyst is obtained.

4. CONCLUSION

An enhancement of the photocatalytic activity of theSnO2−TiO2 coupled semiconductor photocatalyst loadedwith polyaniline (PANI), a conducting polymer, has been

confirmed in the oxidation of 1-octene to 1,2-epoxyoctanewith aqueous hydrogen peroxide. A higher photocatalyticactivity in the oxidation of 1-octene on PANI-SnO2−TiO2

then PANI-TiO2, SnO2−TiO2, and TiO2 can be consideredas an evidence of enhanced charge separation of PANI-SnO2−TiO2 photocatalyst as confirmed by photolumines-cence spectroscopy. It suggests that photoinjected electronsare tunneled from TiO2 to SnO2 and then to PANI in orderto allow wider separation of excited carriers.

ACKNOWLEDGMENTS

This research was supported by the Ministry of Science,Technology and Innovation (MOSTI), Malaysia under IRPAGrant no. 09-02-06-10041-EAR and The Academy of Sci-ences for the Developing World, under TWAS Grants in BasicSciences no. 04-462 RG/CHE/AS.

REFERENCES

[1] A. L. Linsebigler, G. Lu, and J. T. Yates Jr., “Photocatalysis onTiO2 surfaces: principles, mechanisms, and selected results,”Chemical Reviews, vol. 95, no. 3, pp. 735–758, 1995.

[2] M.-W. Xu, S.-J. Bao, and X.-G. Zhang, “Enhanced photocat-alytic activity of magnetic TiO2 photocatalyst by silver deposi-tion,” Materials Letters, vol. 59, no. 17, pp. 2194–2198, 2005.

[3] D. Chatterjee and A. Mahata, “Visible light induced pho-todegradation of organic pollutants on dye adsorbed TiO2 sur-face,” Journal of Photochemistry and Photobiology A, vol. 153,no. 1-3, pp. 199–204, 2002.

[4] H. Tada, Y. Kubo, M. Akazawa, and S. Ito, “Promoting effectof SiOx monolayer coverage of TiO2 on the photoinduced ox-idation of cationic surfactants,” Langmuir, vol. 14, no. 11, pp.2936–2939, 1998.

[5] S. Pilkenton and D. Raftery, “Solid-state NMR studies of theadsorption and photooxidation of ethanol on mixed TiO2-SnO2 photocatalysts,” Solid State Nuclear Magnetic Resonance,vol. 24, no. 4, pp. 236–253, 2003.

[6] P. V. Kamat and K. Vinodgopal, “Environmental photo-chemistry with semiconductor nanoparticles,” Molecular andSupramolecular Photochemistry, vol. 2, pp. 307–350, 1998.

[7] J. Fernandez, J. Kiwi, J. Baeza, J. Freer, C. Lizama, and H. D.Mansilla, “Orange II photocatalysis on immobilised TiO2: ef-fect of the pH and H2O2,” Applied Catalysis B, vol. 48, no. 3,pp. 205–211, 2004.

[8] S. Quillard, G. Louarn, S. Lefrant, and A. G. McDiarmid, “Vi-brational analysis of polyaniline: a comparative study of leu-coemeraldine, emeraldine, and pernigraniline bases,” PhysicalReview B, vol. 50, no. 17, pp. 12496–12508, 1994.

[9] H. Nur, N. Y. Hau, I. I. Misnon, H. Hamdan, and M. N.M. Muhid, “Hydrophobic fluorinated TiO2-ZrO2 as catalystin epoxidation of 1-octene with aqueous hydrogen peroxide,”Materials Letters, vol. 60, no. 17-18, pp. 2274–2277, 2006.

[10] N. K. Mal, A. Bhaumik, R. Kumar, and A. V. Ramaswamy,“Sn-ZSM-12, a new, large pore MTW type tin-silicate molec-ular sieve: synthesis, characterization and catalytic propertiesin oxidation reactions,” Catalysis Letters, vol. 33, no. 3-4, pp.387–394, 1995.

[11] L. Zhang and M. Wan, “Polyaniline/TiO2 composite nan-otubes,” Journal of Physical Chemistry B, vol. 107, no. 28, pp.6748–6753, 2003.

6 International Journal of Photoenergy

[12] H. Xia and Q. Wang, “Preparation of conductive polyani-line/nanosilica particle composites through ultrasonic irradi-ation,” Journal of Applied Polymer Science, vol. 87, no. 11, pp.1811–1817, 2003.

[13] C. Bian and G. Xue, “Nanocomposites based on rutile-TiO2

and polyaniline,” Materials Letters, vol. 61, no. 6, pp. 1299–1302, 2007.

[14] H. Yamashita, Y. Ichihashi, S. G. Zhang, et al., “Photocatalyticdecomposition of NO at 275 K on titanium oxide catalysts an-chored within zeolite cavities and framework,” Applied SurfaceScience, vol. 121-122, pp. 305–309, 1997.

[15] J. B. Lagowski, “Ab initio investigation of conformational andexcitation energies of phenylene vinylene oligomers,” Journalof Molecular Structure: THEOCHEM, vol. 589-590, pp. 125–137, 2002.