Synthesis of mesoporous MCM-48 with nanodispersed metal and

metal oxide particles inside the pore system

Dissertation

zur Erlangung des Grades eines

Doktors der Naturwissenschaften

der Fakultät für Geowissenschaften der

Ruhr-Universität Bochum

vorgelegt von

Mahuya Bandyopadhyay

aus Sonamukhi

Bochum, Oktober 2004

Abstract

Mesoporous MCM-48 was synthesized using hydrothermal as well as microwave

heating technique.

Dip wet impregnation of calcined MCM-48 with metal salt solution and subsequent

calcination led to the incorporation of metal oxides inside the mesopore system. Different

characterization techniques showed that the oxides were inside. Aqueous route led to

formation of CuO particles but ZnO seemed to have coated the channel wall. However, ZnO-

MCM-48 prepared by the organometallic route led to the formation of ZnO particles.

The structure and size of the titania particles were characterized with UV-vis, XANES

and EXAFS spectroscopies. The titania particles (~2 nm) after successive loading, showed

more than one coordination environment, including resemblance with the structure of bulk

rutile-TiO2.

Au nanoparticles with ~3 nm crystal size were introduced inside TiO2 loaded MCM-

48 support. Au/TiO2-MCM-48 material showed catalytic activity in CO oxidation.

Keywords: MCM-48, Microwave synthesis, Metal salts, TEM, EXAFS, UV, CO oxidation

Zusammenfassung

Mesoporöses MCM-48 wurde konventionell hydrothermal und mikrowellen-hydrothermal

hergestellt.

Durch Tauchimprägnierung in Lösungen organischer Metallsalze und nachfolgender

Kalzinierung wurden Metalloxid-beladene mesoporöse Silicas erhalten. Mit verschiedenen

Charakterisierungstechniken wurde nachgewiesen, dass die Metalloxide in Porensystem

eingebaut vorliegen, jedoch in unterschiedlicher Struktur. Aus wässrigen Lösungen wurden

CuO-Partikel erhalten, während ZnO offensichtlich zu einer Reaktion mit der Wand der

Matrix führte. Auf der metallorganischen Route gelang es auch ZnO partikulär ins

Porensystem einzubauen.

Die Struktur der eingebauten TiO2-Partikel wurde mit Hilfe der UV-vis-Spektroskopie, der

EXAFS- und XANES-Spektroskopie analysiert. Die TiO2-Nanopartikel weisen mehr als eine

Koordinationssphäre auf und ähneln fehlgeordneten Rutil-Kristalliten.

Gold-Nanopartikel von kleiner 3 nm Größe wurden in TiO2- MCM-48 eingebaut. Das

Composit-Material zeigt katalytische Aktivität bei der CO-Oxidation.

Schlagworte: MCM-48, Mikrowellen-Synthese, Metallsalze, TEM, EXAFS, UV, CO-

Oxidation

Contents

Page

1. General introduction 1.1 General background 1

1.2 Application of M41S materials 4

1.3 Synthesis mechanism of M41S materials 6

1.4 Synthesis procedure 12

1.5 Physico-chemical characterization 13

1.5.1 X-ray diffraction 13

1.5.2 Adsorption measurement 14

1.5.3 Transmission electron microscopy (TEM) 15

1.5.4 X-ray photoelectron spectroscopy (XPS) 16

1.5.5 Extended X-Ray absorption spectroscopy (EXAFS) 16

1.5.6 Infrared spectroscopy (IR) 17

1.5.7 UV-visible spectroscopy 18

1.5.8 Nuclear magnetic resonance (NMR) 18

1.5.9 Scanning electron microscopy (SEM) 19

1.5.10 Inductively coupled plasma atomic emission spectroscopy (ICP AES) 19

1.6 Scope of the thesis 19

1.7 Objectives of the present work 21

1.8 References 22

2. Synthesis and characterization of mesoporous MCM-48 2.1 Introduction 28

2.2 Experimental 30

2.2.1 Synthesis of MCM-48 by hydrothermal heating technique 30

2.2.2 Synthesis of MCM-48 with different pore sizes 30

2.2.3 Microwave synthesis of MCM-48 31

2.3 Results and discussion 32

2.3.1 Synthesis of MCM-48 by hydrothermal heating technique 32

2.3.2 Study of kinetics during crystallization of MCM-48 40

2.3.3 Synthesis of MCM-48 by microwave heating technique 42

2.4 Conclusion 48

2.5 References 50

3. Post-synthetic impregnation of MCM-48 by CuO, ZnO and Cu/ZnO 3.1 Introduction 52

3.2 Experimental 54

3.2.1 Synthesis of Cu/ZnO-MCM-48 54

3.2.2 Synthesis of pure ZnO-MCM-48 and CuO-MCM-48 54

3.3 Results and discussion 55

3.4 Conclusion 74

3.5 References 75

4. Synthesis and characterization of TiO2 loaded MCM-48 4.1 Introduction 78

4.2 Experimental 80

4.2.1 Synthesis of TiO2 impregnated MCM-48 80

4.3 Results and discussion 80

4.4 Conclusion 98

4.5 References 99

5. Deposition of Au nanoparticles inside TiO2-MCM-48 and the

catalytic activity on CO oxidation 5.1 Introduction 102

5.2 Experimental 103

5.2.1 Synthesis of Au/TiO2-MCM-48 103

5.2.2 Catalytic activity measurements 104

5.3 Results and discussion 104

5.3.1 Synthesis 104

5.3.2 Catalytic evaluation on CO oxidation reaction 110

5.4 Conclusion 113

5.5 References 115

6. Summary 117

1. General introduction

1.1 General background

Porous molecular sieves are extensively used as adsorbents, catalysts and catalyst

supports owing to their high surface areas. According to IUPAC definition1 porous materials

are divided into three types based on their pore dimensions:

Type Pore Size (Å)

Microporous : < 20

Mesoporous : 20-500

Macroporous : > 500

Well-known members of the microporous materials are Zeolites2 and aluminophosphate

molecular sieves3. Zeolites are hydrated, crystalline, microporous aluminosilicates,

constructed from TO4 tetrahedra (T = tetrahedral atom, e.g., Si, Al) connected by oxygen

atoms. These materials possess uniform channels associated by rings of a definite number of

T atoms. The characteristics of their structure, thermal stability, different acid sites and

acidity, ion exchange properties, shape and size-selective pores and channels of zeolites have

been well established now. Modification of the framework and incorporation of different

species make these materials catalytically active. Till 1990, heterogeneous catalysis over

zeolites was restricted to the materials with pore sizes less than 20 Å. Therefore, pore

enlargement was one of the main aspects in zeolite chemistry. Considerable synthetic effort

has been devoted to develop frameworks within the mesoporous range. In 1982 Wilson and

co-workers4 synthesized a novel crystalline, microporous aluminophosphate (AlPO4)

material with a pore diameter of about 8 Å. In 1988, Davis et al.5 reported the synthesis of

VPI-5, an other AlPO4 molecular sieve possessing 18 membered ring channels with 12 Å

pore size. Unidirectional pore zeolite (e.g.UTD-1), possessing 14 membered ring (MR) was

synthesized using a cobalt organometallic complex as the template.6 AlPO4 molecular sieves

of 20 membered ring pore opening (e.g. JDF framework structure) was also reported.7

1

Table 1.1 Classification of Zeolites and molecular sieves based on pore size (< 20 Å)4,8

Group No. of tetrahedra

in pore opening

Pore size(Å)

Example Framework

type code

Small 8 3 AlPO4-20 SOD

Medium 10 6 AlPO4-11 AEL

Large 12 8 AlPO4-5 AFI

Extra large 18 12 VPI-5 VFI

14 7.9 × 8.7 AlPO4-8 AET

Different types of mesoporous materials have been reported in the literature, such as

silicas,9 pillared clays and other silicates.10,11 The synthesis of amorphous silica-alumina in

the presence of tetra alkyl ammonium cations12 has been reported, where the average pore

diameter was related to the size of the tetra alkyl ammonium cations. Although, these

materials were found to be active for acid catalyzed organic reactions, they were not

thermally stable at high temperatures.

In 1992, scientists in Mobil Oil Corporation research group discovered the new family

of mesoporous siliceous materials designated as M41S with exceptionally large and uniform

pores.13-19 These materials possess well-defined pores, whose diameters can be varied in the

range of approximately 15-100Å. At the beginning, three members of the M41S family of

materials were introduced, namely MCM-41, MCM-48 and MCM-50. The individual pores

of MCM-41 were considered to be obtained from silicate condensation about separate

cylindrical micelles, whereas the ordering of structure results from the hexagonal

arrangement of the silica-encased micellar array. The material has one dimensional channel

system,14,19 resembling a honeycomb network. The structure of MCM-48 is more complex

than the straightforward case of the hexagonal one-dimensional MCM-41. It is to be

2

analogous to that of liquid crystal like material with the cubic Ia3d symmetry. There are

several proposals about the mechanism of formation of the cubic liquid crystal phase such as

independent, mutually intertwined arrangements of surfactant rods towards a complex,

infinite, periodic minimal energy surface structure.20,21 The structure has cubic symmetry and

contains two intersecting unique three-dimensional channel system. The structure of the

stabilized lamellar MCM-50 belongs to the same proposed structure of a lamellar liquid

crystal phase. The phase can be represented as the arrangement of the sheets or bilayers of

surfactant molecules with the hydrophilic end, directed toward the silicate-water interface and

the hydrophobic ends of the surfactant molecules face one another. Any silicate structure

obtained from the lamellar phase could be similar to that of two-dimensional layered silicates

such as Magadiite or Kenyaite.22 The classification of the family of mesoporous materials

according to their different symmetry is given in Table 1.2 and Fig. 1.1.

Table 1.2 The classification of mesoporous material23

Crystal system Example Pore system Space group

Hexagonal MCM-41 1-D p6mm

Cubic MCM-48 3-D Ia3d

Stabilized lamellar MCM-50 2-D

Cubic SBA-1 Pm3n

Hexagonal SBA-3 2-D p6mm

3

a

b

c

Fig. 1.1: (a) Model of the three dimensional structure of cubic MCM-48, (b) one dimensionalhexagonal MCM-41, (c) Stabilized lamellar structure of MCM-50

(Ref. M. Kaneda, J. Phys. Chem. B, 106, 2002)

1.2 Applications of M41S materials

Defect free siliceous mesoporous molecular sieve with composition of SiO2 are

electrically neutral and this leads to a lack of strong intrinsic surface acidity. Numerous

attempts were made to incorporate transition metal atoms as well as main group elements into

mesoporous silica. Substitution of trivalent cations such as B+3,Ga+3, Al+3 and Fe+3 24-30 for

silicon in the wall of the mesoporous silica, results a negative framework charge, which can

be compensated by protons providing catalytically active acid sites. The number of acid sites

and strength is related to the amount and nature of the incorporated metal. These materials are

used in acid catalyzed reaction and the main applications are in petroleum refining

processes.31,32 The Ti, V and Cr containing MCM-48 molecular sieves are active catalysts for

oxidative double bond cleavage of methyl methacrylate pyruvate and benzaldehyde as the

dominant respective products.33-35 Zr-MCM-41 has been found to be active toward the

dehydration of isopropyl alcohol.36 A few reports describe the synthesis and characterization

4

of mesoporous silica modified by metals like Mn,37 or Mo,38 which have been found to be

catalytically active in the hydroxylation of phenol, 1-napthol and oxidation of aniline with

aqueous H2O2.

Besides variable pore diameters and large surface areas, mesoporous materials contain

different types of silanol groups, i.e. external surface, internal surface and interstitial silanol

group. The interstitial silanol groups are not accessible in chemistry, but the other two groups

can be functionalized by introducing functional organic groups. The silanol groups present at

the surface of the walls are suitable for chemical bonding of organic ligands or anchoring

inorganic species.39,40 This can be achieved through attachment of silane-coupling agents to

the mesoporous walls of previously synthesized and calcined materials.41 In one way

functionality is directly introduced via reaction of silanol groups,42,43 and, in another way, a

transition metal, metal oxide or a bimetal complex is grafted on the wall without the use of

intermediate silane complexing agent. Then the surface hydroxyl groups can directly react

with the metal species. For example Ti, V, Zr or Mo-grafted MCM-48 were used as a catalyst

for the oxidation of bulky 2,6-di-tert-butyl phenol (2,6-DTBP) using hydrogen peroxide.44

Mesoporous materials anchored with AlCl3, SnCl2, Zn(O2CMe)2 or Mn(O2CMe)2 have high

stability and catalytic activity along with ion-exchange capacity.45

It has been reported that Co and Mo incorporated Al-MCM-41 shows higher

hydrogenation and hydrocraking activities than Co-Mo/Al2O3 catalysts.46 Au/Ti-MCM-48

has been used for selective vapor-phase epoxidation of propylene in presence of oxygen and

hydrogen.47 A number of publications on catalytic reactions studied using mesoporous

materials as catalysts are listed in Table 1.3.

5

Table 1.3 Examples of different reactions over metal substituted MCM-48

Catalyst Reaction Ref.

Ti, V, Zr, Mo-MCM-48 Oxidation of bulky 2,6-di-tert-butyl phenol with H2O2 44

Ti-MCM-48 Oxidation of Cyclohexene, hexane, pent-2-en-1-ol with

H2O2

48

Ti-MCM-48, Cr-MCM-48 Peroxidative oxidation of methyl metacrylate and

styrene

49

Ti-MCM-48 Photocatalytic reduction of CO2 and H2O 50

Ti-MCM-48 Peroxidative halogenation of bulky, organic dyes at

neutral pH

51

Al, Ga, Fe-MCM-48 Cumene cracking 52

Au/Ti-MCM-48 Vapor-phase epoxidation of propene using H2 and O2 53

MCM-48/VOx Oxidation of methanol to formaldehyde 54

1.3 Synthesis mechanism of M41S materials

The most outstanding feature of the preparation of the M41S materials is the role of the

templating agents during the synthesis. The template molecules used are not single solvated

organic cations as used in zeolite synthesis, but the surfactants built molecular arrays, which

form complex micelles around which the main mesostructure is built up. Surfactants are large

organic molecules having a long hydrophobic tail of variable length (e.g.

alkyltrimethylammonium cations with the general formula CnH2n+1(CH3)3N+, where n > 8 )

and a hydrophilic head. The schematic diagram of the formation of microporous and

mesoporous material is given in Fig. 1.2. Similar to zeolite synthesis where organic

molecules play an essential role as structure directing agents (SDA), surfactants act as

6

templates forming an ordered organic inorganic composite material.55 The idea of formation

of the organic-inorganic composites is based on the electrostatic interactions between the

positively charged surfactants and the negatively charged silicate species. During synthesis

surfactants aggregate with their hydrophobic tails exposing their polar heads to the aqueous

solution to form complex micelles. When silica is introduced, there is a charge balance

between the surfactant and silicate ion pairs in which the silicate material formed inorganic

walls between ordered surfactant micelles.

(a)

(b)

Fig. 1.2: The schematic diagram for the formation of (a) microporous material usingindivitual short alkyl chain quanternary ammonium compounds, (b) mesoporous materialsusing long chain alkyl quarternary ammonium compounds 18

According to the liquid crystal templating (LCT) mechanism two different pathways18,56 were

described for the formation of the mesoporous systems with CTAB

(cetyltrimethylammoniumbromide) used as SDA. The two pathways can be described as

follows:

7

1. The liquid crystal phase is intact before the silicate species was added. First there is the

formation of the hexagonal liquid-crystal phase around which the growth of the inorganic

materials occurs. The surfactant micelles arrange together to form hexagonal arrays of

rods. Silicate anions in the reaction mixture interact with the surfactant head group at the

micelle water interface. Silica oligomers are formed by either acid or base catalyzed

hydrolysis of the silicate species, which then condense into a silica structure maintaining

the topology of the liquid crystal precursor.

2. Addition of silicate resulted in the ordering of the subsequent silicate-encased surfactant

micelles. To elaborate this effect, the randomly ordered rodlike micelles react with the

silicate species by coulombic interactions to produce two or three monolayers of silicate

at the external surfaces of the micelles. This composite species pack in such a manner to

form a energetically favored hexagonal mesoporous arrangement, accompanied by

silicate condensation. Hydrothermal treatment leads to the inorganic wall to condense.

Fig. 1.3: Possible pathways for the formation of MCM-4114

8

In both of the cases the resultant silicate/surfactant structure mimics structures known from

liquid crystal phases. But due to the low initial surfactant concentration, the second

mechanism one is more vulnerable, proceeding through a co-precipitation of the organic

template with the silica species out of the high water content synthesis gel. A large number of

publications11,57-62 dealing with formation mechanism of mesophases are available (Fig. 1.3).

Besides, the syntheses based on ionic interactions, the liquid-crystal theory has been

extended to different pathways showing organic-inorganic pathways other than ionic. Huo et

al.63 proposed a generalized approach to the synthesis of periodic mesophases of metal oxides

and cationic or anionic surfactants under a range of pH condition. Four pathways that lead to

mesoporous materials, are proposed and schematically depicted in Fig. 1.4.

Fig. 1.4: A general scheme for the self assembly reaction of different surfactant and inorganic species 63

9

In the first route it is considered that the direct co-condensation of anionic inorganic species

with a cationic surfactant take place (S+I-). Path 2 involves cooperative condensation of a

cationic inorganic species with an anionic surfactant (S-I+). On the other hand route 3 and 4

direct towards the condensation of ionic inorganic species in the presence of similarly

charged surfactant molecules. These routes are mediated by counter-ions of opposite charge

to that of the surfactant head group (S+X-I+ where X- = Cl-, Br- or S-M+I- where M+ = Na+,

K+).

MCM-48, MCM-41, lamellar phase were synthesized by the self-assembly of anionic silicate

and cationic surfactant molecules (S+I-).14,56,62 A Similar trend was observed using charge-

reversed situation, where an anionic surfactant was used to direct the condensation of cationic

oxide species. In this case C16H33SO3H was used in the synthesis of iron and lead oxides,

giving a hexagonal phase and different lamellar phase. The anionic polar head group can be

the part of the inorganic framework of these materials. For the synthesis of organized Zn-

containing phosphate mesostructured material S+X-I+ mechanism was applied. All of the

above routes are related with the charge matching between ionic surfactants and ionic

inorganic reagents, which correspond to strong interaction between template and the charged

framework and make it very difficult to recover the template. Tanev et al.64 further extended

the liquid crystal approach to neutral conditions where mesostructures are formed by using

neutral (S0) or neutral and inorganic precursors (I0) or non-ionic surfactants (N0).65 These

neutral templating route produces mesoporous materials with larger wall thickness, small

scattering domain sizes. These pathways allow the recovery of the template by simple solvent

extraction. In this approach hydrogen bonding is considered to be the driving force for the

formation of the mesophases. Different pathways for the synthesis of mesoporous materials

are given in Table1.4.

10

Table 1.4 Different pathways for the synthesis of mesoporous molecular sieves

Mechanism pH Symmetry Surfactant Reference

S+I- 10-13 Hexagonal, cubic

and lamellar

Cetyltrimethylammonium

ion + silicate species

14

S+X- I+ <2 Hexagonal Cetyltrimethylammonium

ion + silicate species

63

S0I0 <7 Hexagonal C12H25NH2 + (C2H5O)4Si 64

S = Surfactant, I = Inorganic species (Si) X = Cl, Br or OH

The research activities, for example synthesis, modification and catalytic application on

MCM-41 have been graphically depicted in Fig. 1.5.

Fig. 1.5: The schematic diagram of the research activities involving the use of MCM-41

11

1.4 Synthesis procedure

Mesoporous materials are hydrothermally synthesized by mixing organic molecules,

silica and/or silica-alumina source at a temperature between 80-150 º C for a selected period

of time. Like zeolite syntheses, the organic surfactant molecules act as templates forming an

ordered organic-inorganic composite material.55 The surfactant is removed through

calcination, leaving the porous silicate network. The mesophases can be altered by varying

the surfactant/SiO2 ratio.18 The primary difference between the hexagonal and cubic phases

was in the surfactant/SiO2 ratio, where a Surf./Si ≤ 1 gave MCM-41 and Surf./Si > 1 gave

MCM-48. The synthesis condition of cubic phases was first given by Mobil group.35

Cetyltrimethylammonium ion was used as a surfactant. Huo et al.23 developed the synthesis

route by choosing the surfactants which preferentially lead to the formation of MCM-48. His

strategy was based on the surfactant packing parameter within an amphiphilic liquid-crystal

array.66 For example using benzylalkonium surfactant (CBDAC),

[C16H33(CH3)2N+(CH2)(C6H5)]Cl- MCM-48 could be produced over a wide range of

surfactant to silica molar ratio as low as 0.1. Huo et al.23 described that synthesis procedure

as follows, collecting the product by filtration after a room temperature synthesis,

resuspending the solid in deionized water (pH 7-10) and treating it hydrothermally (100 °C)

for an additional amount of time. The most interesting approach towards the synthesis is the

discovery of gemini surfactant 67 {[CmHm+1N+(Me)2(CH2)12N+(Me)2CmHm+1]2Br-, where m =

16, 18, 22, }. The use of gemini surfactant first enabled the synthesis of MCM-48 at room

temperature, after aging for 24h. Gallis and Landry68 explored a temperature induced

hexagonal to cubic phase transition to carefully control and optimize the formation of cubic

MCM-48. A well ordered cubic MCM-48 with CTAB surfactant could be synthesized by

stirring the synthesis mixture for 2-3 h and then heating 3-4 h at 150 °C.

One of the most unique and useful feature of M41S family of materials is the ability to

tailor the pore diameter (15-100 Å). This can be achieved in different ways; 1. By varying the

12

chain length of alkyl group (from 8 to 22 atoms) in surfactant molecules,13,18 2. By adding

auxiliary chemicals such as 1,3,5 trimethylbenzene,13,18,69 which dissolve in the hydrophobic

region of the micelles, thus increasing their size or 3. By aging a sample prepared at a low

temperature in its mother liquor at a higher temperature for different periods of time.70

Besides, the pore diameter of mesoporous materials also depend on other factors such as

temperature, pH and crystallization time.71

1.5 Physico-chemical characterization

For the characterization of mesoporous molecular sieves mainly X-ray diffraction,

transmission electron microscopy (TEM) and adsorption measurements were used. Besides

these techniques, infrared spectroscopy, ultraviolet spectroscopy, nuclear magnetic resonance

and extended X-ray adsorption spectroscopy were also applied to obtain additional structural

information about mesoporous molecular sieves.

1.5.1 X-ray diffraction

X-ray diffraction is one of the most important and versatile techniques among the

different analytical tools available for characterization of mesoporous materials. Several

important structural features are obtained from XRD data such as: phase purity, degree of

crystallinity of the sample. It also helps in the study of the kinetics of crystallization of

molecular sieves. XRD patterns of mesoporous phase exhibit peaks in the low angle region.

The samples synthesized and used during the course of present work were analyzed by

powder X-ray diffraction (XRD) for qualitative and quantitative phase identification.

X-ray power diffraction experiments were carried out using a Siemens D5000

diffractometer with Cu-Kα1 radiation (λ = 1.54059 Å) in Debye-Scherrer transmission

geometry and Philips diffractometer with Cu-Kα1 radiation (λ = 1.5418 Å) in Bragg-

Brentano reflection geometry. The sample was loaded in a glass capillary for the Siemens and

on flat plate sample holders for the Philips diffractometer respectively. In order to check for

bulk oxide signals, powder XRD-diffractograms have been recorded up to 2θ = 50°. For long

13

exposure times, a Huber guinier imaging plate camera G 670 was also used with Cu-Kα1

radiation. The sample was loaded in a glass capillary.

1.5.2 Adsorption measurement

Molecular sieves possess uniform pores and channels with large void volume.

Molecular sieves also have high surface area, which is accessible to molecules of comparable

size to diffuse through the pores. Sorption capacities for probe molecules such as nitrogen,

benzene, n-hexane, water etc. provide information about the hydrophilicity/hydrophobicity,

pore dimensions and pore volume of the molecular sieves.

The Brunauer-Emmett-Teller (BET) gas adsorption method has become the most

widely used standard procedure for the determination of the surface area of finely-divided

porous materials.72 The relation between the amount adsorbed and the equilibrium pressure of

the gas at constant temperature is defined by the adsorption isotherm.1 N2 adsorption-

desorption isotherm of MCM-48 is of type IV isotherm. Characteristic features of the type IV

isotherm are its hysteresis loop, which is associated with capillary condensation taking place

in the mesopores, and the limiting uptake over a range of high p/p0. The initial part is

attributed to monolayer-multilayer adsorption of the surface. The sharpness and the height of

the step reflect the uniformity of the pore size and the pore volume respectively. The N2

adsorption measurement has been extensively used for the present study to confirm the type

IV isotherm for MCM-48 and to obtain values for surface area and porosity data.

Surface area and porosity measurements of the calcined samples by means of N2

physisorption were carried out on an automatic Quantachrome NOVA 2000 analyzer. The

BET surface area was calculated using the conventional BET equation.72 The mesopore size

distribution was estimated from the desorption branch of the N2-isotherm utilising the BJH

method.73 This experiment was done in collaboration with Technical Chemistry Department,

Ruhr University Bochum. For the titania loaded samples the N2 physisorption measurements

were performed on a Micromeritics ASAP 2010 unit. Data analysis was done by using the

14

non-local density functional theory (NLDFT) method and the data were processed using the

Autosorb package provided by Quantachrome. The pore size distributions were derived by

using the NLDFT equilibrium method for nitrogen adsorption on oxidic surfaces.74 The above

experiment was done in collaboration with Max Plank Institute für Kohlenforschung,

Mülheim, Germany.

1.5.3 Transmission electron microscopy (TEM)

TEM was used to evaluate the microstructure and pore structure of mesoporous

molecular sieves.13,18,55,75,76 It gives topographic information of materials at near atomic

resolution. However, the exact analysis of pore sizes and thickness of the pore walls is not

straightforward and not possible without additional simulations because of the “focus”

problem. High resolution transmission electron microscopy (HRTEM) is successfully used to

examine the micro-structural feature of mesoporous molecular sieves.75,76 In addition to

structural characterization, it is also used to detect the location of metal clusters and heavy

cations in the framework.76

A Hitachi H-8100 scanning and transmission electron microscope operating at

accelerating voltages up to 200 kV with a single crystal LaB6 filament was used for the TEM

studies. The specimens were prepared by placing a drop of the dilute suspension of the

calcined powder samples in ethanol on a carbon-coated copper grid. The samples were

allowed to dry at room temperature. Energy dispersive X-ray analysis or EDX analysis

method was used for identification of elemental composition of the material. The EDX

spectra not only identify the element corresponding to each of its peak but the type of X-ray

to which it corresponds as well. For example, a peak corresponding to the amount of energy

possessed by X-rays emitted by an electron in the L-shell going down to the K-shell is

identified as a K-α peak, whereas a peak corresponding to X-rays emitted by M-shell electron

going to the K-shell is identified as a K-β peak. From the peak intensities in the EDX

15

spectrum the ratio of the element present in the material can also be estimated. The above

experiment was done in cooperation with Physical Chemistry 1, Ruhr University Bochum.

1.5.4 X-ray photoelectron spectroscopy (XPS)

XPS is among the most frequently used techniques for surface characterization in

catalysis. It gives information on the elemental composition of surface and the binding energy

of elements is measured by this experiment. It is an important tool for the characterization of

the surface of zeolites and related materials.77 Depth selective information can be obtained by

varying the angle between the sample surface and analyzer. Mesoporous molecular sieves

containing different heteroatoms such as Al, Ti, V, Nb and Zr have been analyzed by XPS

technique to obtain information about the dispersion of the heteroatom in the structure.78-81

X-ray photoelectron and X-ray-induced Auger spectra (XPS, XAES) were obtained

with a Leybold LH 10 spectrometer equipped with an EA 10/100 multichannel detector

(Specs). The spectra were recorded with Mg Kα excitation (10 kV * 20 mA). Binding

energies (BE) were referenced to the Si (2p) line of the MCM-48 material, which was set to

103.2 eV. Elemental ratios in the XPS sampling region were evaluated from line intensity

ratios using the Scofield photoionization cross sections82 together with an empirically derived

function describing the energy dependence of analyzer transmittance of the instrument used.

These measurement and the calculation were done in collaboration with Technical Chemistry

Department.

1.5.5 Extended X-Ray absorption spectroscopy (EXAFS)

EXAFS spectra are a plot of the value of the absorption co-efficient of a material

against energy over a 50-1000 eV range. Through careful analysis of the oscillating part of

the spectrum after the edge, information relating to the coordination environment of a central

excited atom is obtained from this spectra.

X-ray absorption spectra (Cu K-edge at 8.979 keV and Zn K-edge at 9.659 keV) were

measured at Hasylab E4 station (Hamburg) using a Si (111) double crystal monochromator

16

detuned to 70 % maximum intensity to exclude higher harmonics present in the X-ray beam.

The spectra µ(k) were measured in transmission mode using ionization chambers, with the

sample at liquid nitrogen temperature. A copper metal foil was measured at the same time

(between the second and a third ionization chamber) for energy calibration. Samples were

pressed into self-supporting discs of suitable thickness and stored in the ambient atmosphere.

Data treatment was carried out using the software package VIPER.83 For background

subtraction a Victoreen polynomial was fitted to the pre-edge region. A smooth atomic

background, µ0(k), was evaluated using smoothed cubic splines. The radial distribution

function FT[k2 χ(k)] was obtained by Fourier transformation of the k2-weighted experimental

function χ(k) = (µ(k)-µ 0 (k)) / µ0 (k) multiplied by a Bessel window. For the determination of

structural parameters, theoretical references calculated by the FEFF8.10 code84 were used. To

minimize the number of free parameters, equal backscatters were fitted with the same E0-shift

wherever possible (variations of ± 1 eV admitted). Since the Zn K-edge is not far from the

Cu K-edge, the Zn spectra may be influenced by oscillations originating from the Cu edge.

Such influence can be removed from the data by modelling the Cu-edge spectrum first and

correcting the experimental data at the Zn-edge by the oscillations exhibited by the model in

the corresponding energy region. For the gold containing sample a gold metal foil (between

the second and the third ionisation chamber) was measured at the same time for energy

calibration purposes. The above experiments were done in collaboration with Technical

Chemistry Department of Ruhr University Bochum.

1.5.6 Infrared Spectroscopy (IR)

IR-spectroscopy is used to confirm the acidic nature and isomorphous substitution in

molecular sieves. The IR spectrum in the range 200-1300 cm-1 is used to characterize and

also to differentiate framework structures of different molecular sieves. For example a band

at around 960 cm-1 is observed for Ti or V substituted molecular sieves.

17

The IR spectra were used to confirm the presence of Si-OH group in mesoporous

material as well as Ti-O-Si bond. The IR spectra were recorded using a Perkin Elmer 882 IR

spectrometer. The samples were dried at 50 °C and pressed to tablets.

1.5.7 UV-visible spectroscopy

This technique measures the scattered light reflected from the surface of samples in

the UV-visible range (200-800 nm). For most of the isomorphously substituted molecular

sieves, transitions in the UV region (200-400 nm) are of prime interest. This spectroscopic

technique is used to determine the local structure e.g. the coordination state of transition

metal ions substituted in the matrix of the molecular sieve. The UV-visible spectra were used

in the present study to investigate the quantum size effect of the metal oxide particles and the

particle size was determined from the energy blue shift of the metal oxide impregnated

MCM-48 samples.

The UV measurement was done in a Perkin Elmer Lamda 9 UV/Vis/NIR equipped

with a RSA-PE-19 biconical optical bench for diffuse reflectance measurements. All samples

were measured with BaSO4 reference and in the mode of absorbance and reflection (%). The

UV measurement was done in cooperation with Inorganic Chemistry Department 1, Ruhr

University Bochum.

1.5.7 Nuclear magnetic resonance (NMR)

Lippmaa et al.85 first used the solid state 29Si MAS-NMR for explaining the

coordination state and polyhedra connectivity of silicates. This technique is used in

understanding the structural and physico-chemical properties of molecular sieves. 29Si and

27Al MAS-NMR spectra provide information on Si/Al ordering,86 crystallographically

equivalent and non-equivalent Si and Al atoms in framework lattice, 87,88 spectral correlation

with Si-O-T bond angles89 and Si-O bond lengths.90 29Si MAS NMR has been extensively

used here for the characterization of the local structure of M41S silicates.91 The broad 29Si

NMR spectra of mesoporous samples closely resemble those of amorphous silica.

18

29Si MAS-NMR was recorded with a Bruker ASX 400. The relaxation delay was

given as 300s and 60s for as made and the calcined samples respectively. Spinning speed was

set to about 3.5 kHz.

1.5.9 Scanning electron microscopy (SEM)

SEM experiment was used to see the shape and size of the crystals prepared in the

present study. The shape and morphology of the crystals were monitored by SEM using a

LEO-1530 Gemini microscope.

1.5.10 Inductively coupled plasma atomic emission spectroscopy (ICP AES)

The chemical composition of the samples was determined by ICP AES. The samples

were fused with Na2O2, dissolved in water and neutralized with HNO3. The analyses were

performed using ICP-AES spectrometer (PU 7000, Philips).

1.6 Scope of the thesis

Catalysts based on copper-zinc mixed oxides are of great importance for industrial

scale catalytic processes like low-pressure methanol formation from synthesis gas and steam

forming of methanol yielding H2 and CO2. The present work was a part of an effort to

encapsulate nanosized Cu/ZnO methanol synthesis catalysts in a mesoporous matrix. In

particular nanosized metal oxide particles seems to be more effective for methanol synthesis

reaction.92 For this purpose several aqueous and non-aqueous routes were employed to

prepare the composite catalyst. After the material preparation by co-impregnation of Cu and

Zn into MCM-48, the nature of oxide clusters inside porous cage was also explored by means

of several physico-chemical techniques. Not only CuO and ZnO oxide together but Cu-oxide,

Zn-oxide was also separately loaded in MCM-48 framework by wet impregnation process

and characterized by different methods. The great aspects of the present thesis are the

nanoencapsulation of metal oxide particles inside the porous host and their thorough

19

structural characterization. To our knowledge no detail study regarding this has so far been

reported.

The incorporation of quantumsized titania particles into MCM-48 has been extensively

studied. The deposition of Ti salts within the pore system of cubic MCM-48 and their

subsequent decomposition to Ti oxide cluster has been investigated in the present study. The

final product was thoroughly characterized by different physico-chemical methods to have a

complete idea about the size, position, coordination state of metal oxide inside the porous

matrix. As shown in this study, only the combination of various bulk and surface analytical

techniques provides a most complete picture of the nature of the composite material. Not

many reports are available on incorporation of quantum sized titania particles inside

mesoporous MCM-48. The present study concentrates on the detail study of the above

material.

The use of MCM-48 with quantum sized titania particles inside porous cages as a

support for nanosized Au particles was also one of the main interest of the present thesis. In

recent years, proton exchange membrane (PEM) H2 fuel cells have been used for power

generation in a variety of applications but this cells are very sensitive towards low level of

CO. A promising method to remove trace amount of CO from H2 is preferential oxidation of

CO in presence of excess H2.93-95 The catalytic evaluation of Au/TiO2-MCM-48 catalyst was

explored on CO oxidation reaction.

It is clear from the literature survey on mesoporous molecular sieves, that these

substances have opened a new vista in the field of catalysis. Incorporation of trivalent and

tetravalent elements in the framework make these materials potential catalysts in acid or

redox reactions. Metal species are usually highly dispersed in the silica framework making

the otherwise chemically inert framework catalytically active. The present thesis deals with

the synthesis of MCM-48, the post-synthesis modification with CuO, ZnO, TiO2, Au metal,

and the characterization with different physico-chemical methods.

20

1.7 Objectives of the present investigation

1. To synthesize MCM-48 and to optimize the synthesis condition.

2. Synthesize of MCM-48 by microwave heating technique.

3. Investigation on the kinetics of the synthesis.

4. To synthesize CuO, ZnO, Cu/ZnO, TiO2 and Au/TiO2 incorporated MCM-48

5. To characterize the synthesized samples in detail using various techniques such as, XRD,

NMR, ICP, IR, TEM, N2-physisorption isotherm, and XPS, EXAFS, UV-visible, methods

in collaboration with Physical Chemistry 1, Technical Chemistry, Inorganic Chemistry

Department, Ruhr University Bochum

6. The catalytic evaluation was done on Au/TiO2-MCM-48 in collaboration with MPI für

Kohlenforschung, Mülheim, Germany.

21

1.8 References

1. K.S.W. Sing, D.H. Everett, R.H.W. Haul, L. Moscou, R.A. Oierotti, J. Rouquerol, T.

Siemieniewska, Pure & Appl. Chem. 57 (1985) 603.

2. D.W. Breck in: Zeolites Molecular Sieves, John Wiley and Sons, New York, 1974, pp.1.

3. R. Szostak in: Molecular Sieves, Principles of synthesis and Identification, Van Nostrand

Reinfold, New York, 1989.

4. S.T. Wilson. B.M. Lok, E.M. Flanigen, U.S. Patent, 4, 310, 440, (1982); S.T. Wilson.

B.M. Lok, C.A. Messina,. T.R. Cannon, E.M. Flanigen, J. Am. Chem. Soc. 104 (1982)

1146.

5. M.E. Davis, C. Saldarriaga, C. Montes, J.M. Garces, J.A. Crowder, Nature 331 (1988)

698.

6. C.C. Freyhardt, M. Tsapatsis, R.F. Lobo, K.J. Balkus Jr, M.E. Davis, Nature 381 (1995)

369.

7. S. Oliver, A. Kuperman, A. Lough, A. Ozin, J.M. Garces, M.M Olken, P. Rudolf, Stud.

Sur. Sci. Catal. 84 (1994) 219.

8. R. Szotak, in: Molecular Sieves, Principles of synthesis and Identification, Second

Edition, Blackie Academic & Professional, 1998, 1.

9. R.K. Iler, in: The Chemistry of Silica, John Wiley & Sons, 1979.

10. T.J. Pinnavaia, Science 220 (1983) 365.

11. T. Yanagisawa, T. Schimizu, Kuroda, C. Kato, Bull. Chem. Soc. Japan. 63 (1990) 988.

12. A. Corma, J. Pérez-Pariente, V. Fornés, F. Rey, D. Rawlence, Appl. Catal. 63 (1990) 145.

13. C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, U.S. Patent 5, 098, 684 (1992).

14. C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710.

15. J.S. Beck, C.T. Chu, I.D. Johnson, C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli,

U.S. Patent 5, 108, 725 (1992).

22

16. J.S. Beck, D.C. Calabro, S.B. McCullen, B.P. Pelrine, K.D. Schmitt J.C. Vartuli, U.S.

Patent 5, 145, 816 (1992).

17. J.S. Beck, C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, U.S. Patent 5, 264, 203

(1993).

18. J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T-W

Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, J. Am.

Chem. Soc. 114 (1992) 10834.

19. J.C. Vartuli, C.T. Kresge, M.E. Leonowicz, A.S. Chu,., S.B. McCullen, I.D. Johnson,

E.W. Sheppard, Chem. Mater. 6 (1994) 2070.

20. K. Fontell, Colloid Interface Sci. 268 (1990) 164.

21. V. Luzzati, A. Tardieu, T.G. Krzywicki, E. Rivas, F.R. Husson, Nature 220 (1968) 485.

22. J.C. Vartuli, C.T. Kresge, W.J. Roth, S.B. McCullen, J.S. Beck, K.D. Schmitt, M.E.

Leonowicz, J.D. Lutner, E.W. Sheppard, in: Designed Synthesis of Mesoporous

Molecular Sieve Systems Using Surfactant-Directing Agents, Chapter 1, Advanced

Catalysts, 1996, pp. 1.

23. Q.S. Huo, D.I. Margolese, G.D. Stucky, Chem. Mater. 8 (1996) 1147.

24. A. Tuel, S. Gontier, Chem. Mater. 8 (1996) 114.

25. A. Sayari, I. Moudrakovaski, C. Danumah, C.I. Ratcliffe, J.A. Ripmeester, K.F. Preston,

J. Phys. Chem. 99 (1995) 16373.

26. U. Oberhagemann, I. Topalovic, B. Marler, H. Gies, Stud. Surf. Sci. Catal. 98 (1995) 17.

27. S. Liu, H. He, Z. Luan, J. Klinowski, J. Chem. Soc., Faraday Trans. 92 (1996) 2011.

28. C.F. Cheng, J. Klinowski, J. Chem. Soc., Faraday Trans. 92 (1996) 289.

29. C.F. Cheng, H. He,W. Zhou, J. Klinowski, L.F. Sousa Goncalves, L.F. Gladden, J. Phys.

Chem. 100 (1996) 390.

30. Z.Y. Yuan, S.H. Liu, T.H. Chen, J.Z. Wang, H.X. Li, J. Chem. Soc., Chem. Commun.

(1995) 973.

23

31. J.L. Casci, Stud. Surf. Sci. Catal. 85 (1994) 329.

32. A. Sayari, Chem. Mater. 8 (1996) 1840.

33. W. Zhang, T.J. Pinnavia, Catal. Lett. 38 (1996) 261.

34. C.T. Kresge, M.E. Leonowicz, W.J. Roth, K.D. Schmitt, J.C. Vartuli, U.S. Patent 5, 198,

282 (1993).

35. Z. Luan, J. Xu, H. He, J. Klinowski, L. Kevan, J. Phys. Chem. 100 (1996) 19595.

36. D.J. Jones, J. Jiménez-jiménez, A. Jiménez-Lopez, P. Maireles-Torres, P. Olivera-pastor,

E. Rodriguez-castellón, J. Roziére, Chem. Commun. (1997) 431.

37. D. Zhao, D. Goldfarb, J. Chem. Soc., Chem. Commun. (1995) 875.

38. W. Zhang, J. Wang, P.T. Tanev, T.J. Pinnavaia, Chem. Commun. (1996) 979.

39. T. Maschmeyer, R.D. Oldroyd, G. Sankar, J.M. Thomas, I.J. Shannon, J.A. Klapetko,

A.F. Masters, J.K. Beattie, C.R.A. Catlow, Angew.Chem. 109 (1997) 1713.

40. D.J. Macquarrie, D.B. Jackson, Chem. Commun. (1997) 1781.

41. D.H. Olson, G.D. Stucky, J.C. Vartuli, U.S. Patent 5, 364, 797 (1994).

42. L. Mercier, T.J. Pinnavaia, Adv. Mater. 9 (1997) 500.

43. A. Cauvel, G. Renard, D. Brunel, J. Org. Chem. 62 (1997) 749.

44. K.K. Kang, W.S. Ahn, J. Mol. Catal. A: Chem. 159 (2000) 403.

45. R. Ryoo, S. Jun, J.M. Kim, M.J. Kim, Chem. Commun. (1997) 2225.

46. S. Song, K.M. Reddy, Appl. Catal. A: General 176 (1999) 1.

47. T. Hayashi, K. Tanaka, M. Haruta, J. Catal. 178 (1998) 566.

48. T. Tatsumi, K.A. Koyano, N. Igarasi, Chem. Commun. (1998) 325.

49. W. Zhang, T.J. Pinnavaia, Catal. Lett. 38 (1996) 261.

50. S.G. Zhang, Y. Fuzii, H. Yamashita, K. Koyano, M. Anpo, Chem. Lett. (1997) 659.

51. J.V. Walker, M. Morey, H. Carlsson, A. Davidson, G.D. Stuky, A. Butler, J. Am. Chem.

Soc. 119 (1997) 6931.

24

52. H. Landmesser, H. Kosslick, U. Kürschner, R. Fricke, J. Chem. Soc., Faraday Trans. 94

(1998) 7, 971.

53. B.S. Uphade, T. Akita, T. Nakamura, M. Haruta, J. Catal. 209 (2002) 331.

54. M. Baltes, K. Cassiers, P.Vander Der Voort, B.M. Weckhuysen, R.A. Schoonheydt, E. F.

Vansant, J. Catal. 197 (2001) 1, 160.

55. J.S. Beck, J.C. Vartuli, G.J. Kennedy, W.J. Roth, S.E. Schramm, Chem. Mater. 6 (1994)

1816.

56. M.E. Raimondi, J.M. Seddon, Liquid Crystals 26 (1999) 3, 305.

57. A. Monnier, F. Schüth, Q. Huo, D. Kumar, D.I. Margolese, R.S. Maxwell, G.D. Stucky,

M. Krishnamurthy, P. Petroff, A. Firouzi, M. Janicke, B.F. Chmelka, Science 261 (1993)

1299.

58. A. Firouzi, F. Atef, A.G. Oertli, G.D. Stucky, B.F. Chmelka, J. Am. Chem. Soc. 119

(1997) 3596.

59. A. Steel, S.W. Carr, M.W. Anderson, J. Chem. Soc. Chem. Commun. (1994) 157.

60. C.Y. Chen, S.L. Burkett, H-X. Li, M.E. Davis, Microporous Mater. 2 (1993) 27.

61. G.S. Attard, J.C. Glyde, C.G. Göltner, Nature 378 (1995) 366.

62. S. Inagaki, Y. Fukushima, K. Kuroda, J. Chem. Soc. Chem. Commun. (1993) 680.

63. Q. Huo, D.I. Margolese, U. Ciesla, P. Feng, T.E. Gier, P. Sieger, R. Leon, P.M. Petroff, F.

Schüth, G.D. Stucky, Nature 368 (1994) 317.

64. P.T. Tanev, T.J. Pinnavaia, Science 267 (1995) 865.

65. S.A. Bagshaw, E. Prouzet, T.J. Pinnavaia, Science 269 (1995) 1242.

66. J.N. Israelachvili, D.J. Mitchell, B.W. Ninham, J. Chem. Soc. Faraday Trans. 2 (1976) 72,

1525.

67. R.Zana, M. Benrraou, R. Rueff, Langmuir 7 (1991) 1072.

68. K.W. Gallis, C.C. Landry, Chem. Mater. 9 (1997) 2035.

25

69. Q. Huo, D.I. Margolese, U. Ciesla, D.G. Demuth, P. Feng, T.E. Gier, P. Sieger, A.

Firouzi, B.F. Chmelka, F. Schüth, G.D. Stucky, Chem. Mater. 6 (1994) 1176.

70. D. Khushalani, A. Kuperman,., G.. Ozin, K. Tanaka, J. Garcés, M.M. Olken, N. Coombs,

Adv. Mater. 7 (1995) 842.

71. X. Chen, L. Huang, Q. Li, J. Phys. Chem. B 101 (1997) 8460.

72. S. Brunauer, P. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938) 309.

73. E.P. Barrett, L.G. Joyner, P.P. Halenda, J. Am. Chem. Soc. 73 (1951) 373.

74. U. Ciesla, F. Schüth, Microporous Mesoporous Mater. 27 (1999) 131.

75. V. Alfredsson, M. Keung, A. Monnier, G.D. Stucky, K.K. Unger, F. Schüth, J. Chem.

Soc. Chem. Commun. (1994) 921.

76. A. Chenite, Y.L. Page A. Sayari, Chem. Mater. 7 (1995) 1015.

77. M. Stocker, Microporous Mater. 6 (1996) 235.

78. R. Mokaya, W. Jones, J. Mater. Chem. 9 (1999) 555.

79. R. Mokaya, J. Catal. 186 (1999) 4.

80. J.S. Reddy, A. Dicko, A. Sayari, Chem. Ind. 69 (1997) 405.

81. A. Sayari, K.M. Reddy, I. Moudrakovski, Stud. Surf. Sci. Catal. 98 (1995) 19.

82. J. H. Scofield, J. Electron Spectrosc. Relat. Phenom. 8 (1976) 129.

83. K.V. Klementiev, VIPER for windows (Visual Processing in EXAFS Researches),

freeware, Available from www.desy.de/~klmn/vipr.html.

84. A. L. Ankudinov, B. Ravel, J. J. Rehr, S. D. Conradson, Phys. Rev. B 58 (1998) 7565.

85. E. Lippmaa, M. Magi, A. Samosen, G. Engelhardt, A.R. Gummer, J. Am. Chem. Soc. 102

(1980) 4889.

86. E. Lippmaa, M. Magi, A. Samosen, M. Tarmak, G. Engelhardt, J. Am. Chem. Soc. 103

(1981) 4992.

87. C.A. Fyfe, G.C. Gobbi, J. Klinowski, J.M. Thomas, S. Ramdas, Nature 296 (1982) 530.

88. C.A. Fyfe, G.C. Gobbi, J. Klinowski, J.M. Thomas, J. Phys. Chem. 86 (1982) 1247.

26

89. J.M. Thomas, C.A. Fyfe, S. Ramdas, J. Klinowski, G.C. Gobbi, J. Phys. Chem. 86 (1982)

3061

90. S. Ramdas, J. Klinowski, Nature 308 (1984) 521.

91. A. Steel, S.W. Carr, M.W. Anderson, Chem. Mater. 7 (1995) 1829.

92. J.B. Hansen, in : G. Ertl, H. Knözinger , J. Weitkamp (Eds.), Handbook of Heterogeneous

Catalysis, Vol. 3, Second Edition, Weinheim, Wiley, 1997, pp.1856.

93. T.V. Choudhary, D.W. Goodman, Catal. Today 77 (2002) 65.

94. R.R. Davda, J.A. Dumesic, Angew. Chem. Int. Ed. 42 (2003) 4068.

95. O. Korotkikh, R. Farrauto, Catal. Today 62 (2000) 249.

27

2. Synthesis and characterization of mesoporous MCM-48

2.1 Introduction

The new family of mesoporous molecular sieves designated as M41S, was first

synthesized by C.T. Kresge and co-workers1. MCM-41, one of the members of this extensive

family of mesoporous sieves, possesses a hexagonal array of uniform mesopores (Fig. 1.1 b).

Since 1992 much attention has been paid on potential application of MCM-41. The synthesis

procedures of MCM-48 and MCM-41 are almost identical and MCM-48 is as reproducible as

MCM-41.1,2 Comparisons between MCM-48 and MCM-41 are as follows3:

• MCM-48 can be as easily synthesized as MCM-41

• Porosity of MCM-48 is similar to that of MCM-41

• Three dimensional channel system in MCM-48 is much more desirable than a one-

dimensional MCM-41 in catalytic point of view

• MCM-48 and MCM-41 have similar chemical and thermal stability.

Narrow pore size distribution and its regular cubic pore structure makes MCM-48 as a

matrix to immobilize catalytically active species.4-7 Therefore, it is necessary to find a

flexible and fast synthesis route that allows one to control the shape and porosity of MCM-

48. The present work has been devoted to develop a reliable synthesis route for the

preparation of highly crystalline and reproducible MCM-48. The ultimate goal of the present

thesis is the use of the good quality MCM-48 material as a host for different metal and metal

oxide particles inside mesoporous matrix. So initially more attention was paid to make high

quality and highly reproducible MCM-48 by a reliable synthesis route.

Mesoporous materials are meta-stable forms of metal oxides (e.g. Siliceous MCM-41

is thermodynamically a great deal less stable than quartz)8 and are generally synthesized at

low temperatures (25-100 °C), so that the condensation reactions are kinetically controlled.

After hydrothermal heating for several hours a composite silica/organic structured phase is

formed.

28

Unlike microporous materials, which have rigid structures, the pore size of mesoporous

materials is somewhat flexible and could be engineered to have a desired size at the time of

synthesis using organic additives, mainly long chain amines, of various sizes. This flexibility

of pore size is interesting for the preparation of materials with applications in the synthesis of

organic molecules.9,10 Mesoporous materials with variable pore sizes are also important for

separation and sorption techniques.11,12

To find faster and economical synthesis route for inorganic and organic materials has

always been a challenging task, and the use of microwave heating for this purpose is now a

subject of growing interest. The microwave heating process works in different way compared

to conventional heating.13 Microwave energy can heat the entire object, depending upon the

penetration depth of the microwave energy, to crystallization temperature rapidly and

uniformly resulting in homogeneous nucleation and shorter crystallization times compared to

the hydrothermal heating technique.14 Moreover, this method is energy-efficient and

economical. Microwave synthesis has been applied successfully for the synthesis of many

zeolites and molecular sieves such as A, Y, ZSM-5, Beta, MCM-41, TS-1,

aluminophosphates such as AlPO4-5, CoAPO-5, MgAPO-5, AlPO4-11 and large pore

gallophosphate cloverite and GaPO4 .15-22 Conventionally MCM-48 can be synthesized at 90

ºC by heating the synthesis gel in a oven at autogenous pressure for 4 days. Microwave

synthesis has reduced the time of synthesis of standard preparation by almost 2 orders of

magnitude.

This chapter contains

♦ Reliable synthesis of pure and well crystalline MCM-48 by conventional hydrothermal

heating technique

♦ Microwave synthesis of MCM-48

♦ MCM-48 with variable pore sizes was synthesized using different surfactants

♦ Investigation of crystallization kinetics

29

♦ Characterization of as synthesized and calcined samples by various physico-chemical

techniques

2.2 Experimental

2.2.1 Synthesis of MCM-48 by hydrothermal heating technique

MCM-48 was synthesized hydrothermally from 1TEOS : 0.70 CTACl : 0.5 NaOH :

64 H2O system. The material was synthesized under ambient pressure in polypropylene

bottles. First 1 M NaOH (Baker) solution was taken in a bottle, and CTACl (Cetyltrimethyl

ammoniumchloride, Fluka, 25 % solution in water) was added slowly under continuos

stirring. After 15 min of stirring, tetraethoxysilane, TEOS (Merck), was added drop wise and

the stirring was continued until the gel became homogeneous. The whole synthesis was done

in a water bath at 40-50 ºC. The sealed bottle was then heated at 90 ºC for 4 days.

Afterwards, the bottle was removed from the oven and the product formed was washed

thoroughly with water and dried overnight at room temperature. The dried sample was

washed again with a mixture of water, ethanol and hydrochloric acid (molar ratio 90 : 5 : 10)

and dried at room temperature. The crystalline sample thus obtained was calcined at 540 ºC

for 5 hr to decompose and remove the organic surfactant.

Synthesis was also carried out by using CTAB (cetyltrimethylammoniumbromide,

Aldrich) as surfactant. The synthesis procedure and synthesis condition was same like before.

During the synthesis CTAB (solid powder) was added after addition of NaOH with drop wise

addition of water under vigorous stirring.

2.2.2 Synthesis of MCM-48 with different pore sizes

The direct synthesis of MCM-48 with different amines was carried out using the same

molar composition. Mainly four different types of amines were used

• Octadecyltrimethylammonium ({ODTMA}+)

• Hexadecyltrimethylammonium ({HDTMA}+)

• Tetradecyltrimethylammonium ({TDTMA}+)

30

• Dodecyltrimethylammonium ({DDTMA}+)

For the synthesis with C14 amine and C12 amine ethanol was used as an additive for the

mesophase control. The synthesis mixture was heated at 90 ºC for 4-5 days. The synthesis

was done with both n-alkyltrimethylammonium bromide and chloride. The above

experiments deal with lowering the pore sizes of MCM-48 materials. MCM-48 with larger

pore sizes was also investigated. MCM-48 was synthesized using C18 (octadecyltrimethyl)

amine as surfactant.

2.2.3. Microwave synthesis of MCM-48

CTACl (cetyltrimethylammoniumchloride) was first used as surfactant for the

synthesis of MCM-48 from TEOS : 0.70 CTACl : 0.5 NaOH : 64 H2O synthesis mixture. In

a typical synthesis procedure 1molar (Baker) NaOH solution was taken in a polypropylene

bottle and surfactant was added with drop-wise addition of water under continuos stirring

over a water bath at 40-80 ºC. Afterwards, tetraethoxysilane, TEOS (Merck) was added and

the stirring was continued until the gel was homogeneous. For some cases the final gel was

aged for 3 hr at 80 ºC. After several attempts for the successful synthesis of MCM-48, it has

been found that from the final molar gel ratio TEOS : 0.15 CTABr : 0.5 NaOH : 80 H2O

(where CTABr = cetyltrimethylammoniumbromide, Aldrich), the desired result can be

obtained. At the end the final gel was divided in two parts and poured in Teflon autoclaves

for microwave heating. The microwave oven used for this synthesis was a MDS-2000 system

with a temperature controller as well as adjustable power output (maximum 650W). The gel

was allowed to heat at 100 ºC for 1h and 2h. The solid white products obtained were

separated by filtration, washed with a H2O, Ethanol and HCl mixture and then dried at room

temperature. The surfactant inside the as-synthesized material was removed by calcination at

100 ºC for 12 h then at 540 ºC for 5h.

31

2.3 Results and discussion

2.3.1 Synthesis of MCM-48 by hydrothermal heating technique

The gel composition, synthesis conditions and products obtained are listed in Table

2.1. MCM-48 materials with excellent periodicity were obtained from conventional

hydrothermal heating technique (Fig. 2.1). As an unambiguous indicator for the structure of

MCM-48, the typical XRD diagrams show a ~30% decrease in unit cell volume caused by the

condensation of silanol groups after calcination indicated by the shift of the (211)-peak from

~2.30° to ~2.56° 2θ, d (211) = 38.41Å and 34.24Å respectively (Fig. 2.1).23

a

b

Fig. 2.1: XRD patterns of (a) as made and (b) calcined MCM-48

Calcination leads to increased diffraction peak intensities (Fig. 2.1) due to the increased

diffraction contrast between channel pore and wall after the removal of the surfactant

template.24

32

Table 2.1 Gel composition, synthesis condition and product for MCM-48 synthesis

Sample name TEOS CTACl NaOH H2O Temp. (º C) Time (d) Product

MB02B 1 0.55x 0.5 62 90 4 MCM-41

MB05 1 0.70 0.5 64 90 4 MCM-48

MB08A 1 0.70 0.5 64 90 4 MCM-48

MB08B 1 0.70 0.5 64 90 4 MCM-48

MB 121 1 0.66x 0.5 103 90 4 MCM-48

MB 122 1 0.70x 0.5 103 90 4 MCM-48 xCTABr ( cetyltrimethylammoniumbromide )

Following typical synthesis conditions as described in Table 2.1, 130 batches of MCM-48

were prepared, which were highly crystalline and reproducible.

The 29Si MAS NMR spectrum also changes in a characteristic manner (Fig 2.2).

a b

Fig. 2.2: 29Si NMR of (a) as made (b) calcined MCM-48 sample

The spectra of MCM-48 before and after calcination show that the ratio of Q3 to Q4 [SiO4]

units decreased from 2.3 to 0.09 indicating the condensation of the silanol-group rich silicate

33

network in its as synthesized form to a more stable highly 4-connected tetrahedral silicate

network after calcination (Fig. 2.2). The 29Si MAS NMR spectra and the characteristic

crystallographic data for the as synthesized and the calcined MCM-48 materials are

summarized in Table 2.2 and 2.3.

Table 2.2 Summary of the 29Si NMR measurements.

MCM48 as syn. MCM48 calc.

Q2 Si -89 ppm -91 ppm

Q3 Si -98 ppm -102 ppm

Q4 Si -107 ppm -111 ppm

Ratio Q2:Q3:Q4 2:78:34 1:10:112

Table 2.3 Summery of crystallographic data on MCM-48

Space group symmetry Ia3d

Lattice parameter, as syn. 94.1 Å

Unit cell volume, as syn. 832,880 ų

Lattice parameter, calc. 83.9 Å

Unit cell volume, calc. 590,168 ų

List of reflexions, MCM-48, calc. hkl d-value (Å) °2θ (Cu-Kα) 1 1 2 34.24 2.58 0 2 2 29.66 2.98 0 1 3 26.52 3.33 2 2 2 24.21 3.65 1 2 3 22.42 3.94 0 0 4 20.97 4.21 0 3 3 19.77 4.47 1 1 4 19.77 4.47 0 2 4 18.76 4.71 2 3 3 17.88 4.94 2 2 4 17.12 5.16

34

Fig. 2.3: N2- physisorption isotherm of pure calcined MCM-48

Blue: adsorptionRed: desorption

0. 0 0.2 0.4 0.6 0.8 1.0

0

100

200

300

400

500

600

700

80

0

Relative pressure (p/p0)

VN

2 (ccg

-1)

Fig. 2.3 reports the N2- physisorption isotherm of calcined MCM-48. The isotherm is of type

IV, typical for mesoporous solids. The isotherm of MCM-48 are as known from the

literature,11 with a steep increase around p/p0 = 0.3, which indicates a narrow pore size

distribution and indicative of the filling of the mesopores. The p/p0 coordinate of the

inflection point depends on the pore size. The sharpness in this step suggests a uniform size

pore system. All these characteristic features for example X-ray diffraction diagrams, NMR

spectrum, N2-physisorption isotherms resemble with the typical MCM-48 reference material

reported in the literature.11 From the SEM image (Fig.2.4) the average size of the particles

was estimated as 0.4-0.5µm, The shape of the particles was found to be spherical.

35

Fig. 2.4: SEM picture of pure calcined MCM-48

MCM-48 with low and high pore sizes was also synthesized. The gel compositions and other

conditions are listed in Table 2.4.

Table 2.4 Experiments for MCM-48 with different pore sizes

Sample name

TEOS CTACl NaOH H2O Temp. (º C) Time (d) Product

MB105 1 0.65a 0.5 40 90 5 MCM-48

MB108 1 0.65b 0.5 100 90 5 MCM-48

MB 109e 1 0.65b 0.5 100 90 5 MCM-48

Mb110e 1 0.65c 0.5 100 90 5 MCM-48

MB181 1 0.70d 0.5 80 90 4 MCM-48 aDodecyltrimethylammoniumchloride, btetradecylammoniumbromide, cn-decyltrimethylammonium, doctadecyltrimethylammoniumbromide, eEtOH (5 M) was used as additive

36

MCM-48 with C18 (otcadecylammonium), C14 (tetradecylammonium) and C12 (Dodecyl

ammonium) surfactant was synthesized. The shift of the 1st (211) peak towards right side

indicates the lowering of pore sizes than the usual MCM-48 (Fig. 2.5).

Fig. 2.5: XRD patterns of as made (a) C18, (b) C16, (c) C14, and (d) C12 MCM-48

a

b

c

d

Lower the 2θ value higher the pore size, and this feature is clearly reflected in the XRD

patterns of the different MCM-48 samples (Fig. 2.5). All the samples were well crystalline.

The BET surface area, pore sizes and pore diameters are listed in Table 2.5 and Fig. 2.6.

Table 2.5 N2-Physisorption Results of MCM-48 synthesized with different surfactants

C18-MCM-48 C16-MCM-48 C14-MCM-48 C12-MCM-48

Pore diameter 26 Å 23 Å 21.5 Å 16.5 Å

Surface area (internal)

1100 m2/g 1257 m2/g 1240 m2/g 1142 m2/g

Pore volume 1.54 cm3/g 1.35 cm3/g 1.3 cm3/g 1.0 cm3/g

37

The pore diameter increased gradually from C12 to C18 MCM-48. Similar trend was

observed for pore volume also. The porosity result was comparable with the pattern obtained

from the powder XRD measurements.

10 20 30 40

Fig. 2.6: BJH pore size distribution curves of calcined (a) C12, (b) C14, (c) C16 and (d) C18 MCM-48

0.0

0.05

0

.1

0

.15

0.2

0.

25

0.3

Dv

(d),

cm3

Å-1

g-1

Pore diameter (Å)

a

b

c

d

38

The IR spectra of calcined MCM-48 showed a peak at around 960 cm-1, which is related to

Si-OH (silanol) stretching vibration (Fig. 2.7).

400 800 1200 1600 2000 2400 2800 3200 3600 4000

wave number (cm-1)

Tran

smitt

ance

Fig. 2.7: IR spectra of pure calcined MCM-48

Although radiation damage is a serious problem in TEM experiments of silicas in general,

images of the mesoporous framework have been obtained using low dose techniques.11,25 The

periodic arrangement of walls and channels in mesoporous materials leads to periodic

imaging contrast in the TEM experiment. TEM experiments on mesoporous powders also

show that single particles often don't scatter coherently. In images parts of a mesoporous

particle are well aligned with respect to the electron beam and show the periodicity of the

silica framework whereas other parts are not in scattering conditions and seem to be

amorphous. Slight changes in particle alignment, however, lead to regular and periodic

contrast variations and show that these parts of the crystallite are also ordered periodically. A

typical TEM image of MCM-48 after calcination is shown in Fig. 2.8.

39

Fig. 2.8: TEM image of an ordered area of a calcined MCM-48 sample

The lattice fringes in the center of the particle indicate the well aligned part of the crystallite

with d-spacing of d ~ 58Å corresponding to (110) planes. For all imaged MCM-48 particles

no damage of the periodic structure of the silicate framework was observed.

2.3.2 Study of kinetics during crystallization of MCM-48

The study of the crystallization kinetics of the MCM-48 was performed by collecting

the as made samples at different crystallization time and investigating the relative

crystallinity of the samples obtained with time (Fig. 2.9). In the XRD patterns the relative

crystallinity of the samples was measured from the sum of the intensities under the major

peaks (considering crystallinity of the sample with highest intensity as 100 %). After thermal

treatment for 4h, MCM-41 was obtained instead of MCM-48 and the intensity was low due to

low crystallinity of the sample. The sample removed after 8 hr and 12 hr showed

comparatively higher intensities but it was MCM-41 not MCM-48. At 24 hr the 2nd peak,

which is the indication of MCM-48 appears but it was with very low intensities.

40

Time / h

Rel

ativ

e cr

ysta

llini

ty /

%

0 8 16 24 32 40 48 56 64 72 80 88 96

Fig. 2.9: Crystallization kinetics of MCM-48

0

2

0

40

6080

100

Fig. 2.10: XRD patterns of as made samples removed after (a) 4h (b) 24 h (c) 4 days

a

b

c

41

With time the intensity of this peak and other peaks increases and finally the crystallinity

reached its maximum at about 4 days (Fig. 2.10).

2.3.3 Synthesis of MCM-48 by microwave heating technique

The synthesis conditions and products obtained are listed in Table 2.6.

Table 2.6 Gel composition and other synthesis conditions

Sample Molar gel composition Temperature Time

TEOS CTACl NaOH H2O ( ºC ) ( hour)

1 1 0.70 0.5 64 100 1

2 1 0.70 0.5 64 100 2

3a 1 0.70 0.5 64 100 1

4a 1 0.70 0.5 64 100 2

5b 1 0.70 0.5 64 100 1

6b 1 0.70 0.5 64 100 2

7c 1 0.15 0.5 80 100 1 a The gel was prepared at 80-90 ºC and the Teflon bomb was heated before the gel was poured inside it, b the final gel was aged for 3 hr without stirring and then poured in preheated Teflon bomb, cCTABr was used as surfactant.

The powder X-ray diffraction patterns of MCM-48 prepared by microwave heating technique

is shown in Fig. 2.11. The patterns of samples prepared by microwave technique resembled

to those prepared by conventional hydrothermal heating technique. However, the half width

of the peaks in the XRD patterns was wider in the microwave synthesis, which may be due to

low periodicity of the mesoporous MCM-48 product. Usually the increase of peak half width

in the diffraction pattern of the mesoporous material results from either sample

inhomogeneity or increased wall thickness. However, XRD experiments cannot reveal the

reason for unambiguity. In fact, because of the limited penetration depth of the microwave

during the heating there is significant temperature difference between the liquid inside and

the Teflon liner, which leads to crystallization inhomogeneity.

42

b

Fig. 2.11: XRD patterns of as made (a) microwave (b)hydrothermally synthesized MCM-48

a

The calcination procedure was carried out in a way different than the conventional

hydrothermal process. First the sample was heated at 100 ºC for 12h then increased the

temperature up to 540 ºC. After calcination the intensity of the main peak increased as a

result of increased diffraction contrast between channel pore and wall (Fig 2.12),24 As typical

for MCM-48, the first peak shifts towards higher 2θ value due to the decrease in unit cell

volume caused by the formation of the three dimensional silicate framework through

condensation of the silanol groups as expected for MCM-48. There was not much difference

in powder XRD patterns of the sample heated for 1hr and 2 hr, which means 1 hr heating was

sufficient for successful synthesis.

43

Fig. 2.12: XRD patterns of (a) as made and (b) calcined MCM-48samples prepared by microwave technique

a

b

0

1

2

3

4

5

0 2 4 6 8energy / keV

cps

SiK

α

CuK

α

CuK

10

β

OK

αC

uLα

CK

α

Fig. 2.13: TEM images (a, b) of ordered area and (c) worm-hole like structure of a sample along with the EDX analysis

a

b

c

44

TEM images (Fig. 2.13) reveal that parts of a mesoporous particles are well aligned with

respect to the electron beam and show the periodicity of the silica framework, however some

parts of the images show periodic disorder and seem to be worm hole like structure, which

may be one of the reasons for the peak broadening, reflected in the X-ray diffraction diagram.

For improving the sample quality the as made samples were kept at 50 °C for 4 days then

calcined as described before. After that TEM was recorded. Not much structural difference or

improvement was reflected in the TEM images.

In 29Si MAS NMR (Fig. 2.14) the as made samples exhibited two broad resonances at –

100 ppm for the Q3 environment and at –110 ppm for the Q4 environment, respectively, The

ratio of Q3 to Q4 units decreased after calcination indicating the condensation of the silanol

groups to form the more stable highly 4-connected tetrahedral silicate network. The nature of

the 29Si NMR of the as made sample is somewhat different than the MCM-48 prepared by

hydrothermal heating technique (Fig. 2.2 a). The ratio of Q3 to Q4 is almost 1:1 for the sample

prepared by microwave technique where as for conventionally prepared MCM-48 sample the

ratio was 2:1. Although the exact reason is not known but as some of the parts of the sample

exhibit worm-hole type structure (revealed by TEM) or the pore wall may be thick enough or

the irregular structure of the MCM-48 (from XRD) may the reasons for this type of deviation.

As shown in SEM (Fig. 2.15), the average size of the particles prepared by microwave

technique was about 1µm, which was bigger than that for those samples prepared my

conventional hydrothermal heating technique (0.4-0.5µm). The shape of the particles was

spherical and similar in both the methods.

45

a b

Fig. 2.14: 29Si NMR of (a) as made (b) calcined MCM-48 sample prepared by microwave technique

Fig. 2.15: SEM picture of a sample prepared by microwave technique

46

The pore size distribution curve of the sample prepared by microwave hydrothermal heating

technique is given in Fig. 2.16 and the relevant data are summarized in Table 2.7.

10 20 30 40 50 60 70 80 90 100 110

Fig. 2.16: Pore size distribution curve of MCM-48 (prepared by microwave technique)according to BJH formalism

0.0

0.0

05 0

.01

0

.015

0.

02

0.0

25

0.03

0

.035

0.

04

Pore diameter (Å)

Dv

(d),

cm3 Å

-1g-1

Table 2.7 Result of the N2-physisorption experiment

Sample BET-surface area

(m2g-1)

Total pore volume (BJH)

(cm3g-1)

Mean pore diameter (BJH)

(Å)

MCM-48h 1257 0.93 25

MCM-48m 834 0.70 20

hsample prepared by conventional hydrothermal heating technique, msample made by

microwave hydrothermal heating technique.

The sample shows similar trend in comparison with the sample prepared by conventional

hydrothermal heating technique. The surface area and pore diameter are as expected and the

47

range typical for MCM-48 type material.5 The surface area and pore diameter are lower than

the hydrothermal preparations. This can be explained considering the XRD, 29Si MAS NMR

and the TEM images obtained for the sample. The peak in the powder XRD pattern of the

sample made by microwave heating technique showed wide FWHM i.e. the pore wall might

be thicker than the sample made using hydrothermal heating technique. This is in agreement

with the 29Si NMR spectrum with a higher content of Q4 linked silicate tetrahedra, which

should be present in thicker pore walls. The increased pore wall thickness leads to less

surface area and pore volume in the material. As far as worm-hole structure is considered the

pore diameter and the wall structure should be comparable with the major phase MCM-48

and should not take influence on the values derived from N2 adsorption measurements. From

the lower specific pore volume it might be concluded that, in agreement with the XRD and

the TEM results, at least a significant fraction of the samples prepared via microwave

treatment is not homogeneous or ordered. Nevertheless, after several attempts, the synthesis

conditions were improved and MCM-48 was synthesized for the first time via microwave

hydrothermal heating techniques. This is the development, achieved in the present study.

2.4 Conclusion

The M41S family of materials are both unique and diverse. These mesoporous

materials are thermally stable and having high sorption capacity. MCM-48 is unique among

the mesoporous materials. The channel system in MCM-48 is three dimensional, allowing

easy diffusion of guest species through the channel voids. MCM-48 was successfully

synthesized by conventional hydrothermal heating technique. The synthesis procedure was

very reliable and it was the fast access to high quality of MCM-48 materials. The samples

were well crystalline as revealed by powder XRD patterns. One of the most beautiful aspects

of mesoporous materials is the flexibility of the pore size. Different surfactants were used to

make MCM-48 with different pore sizes. The change of pore sizes was identified by XRD

patterns and N2 physisorption measurements. The periodicity of the sample was investigated

48

by TEM measurement. TEM images showed that the sample prepared by hydrothermal

heating technique was very homogeneous and exhibited periodicity of the silicate framework.

The shape of the particles was spherical as revealed by SEM photographs. So the MCM-48,

which was synthesized was well characterized by different techniques and then was used for

different metal and metal oxide loading, which is described in next chapter.

The present study was also an approach of rapid synthesis of MCM-48 by microwave heating

method. Much shorter time of crystallisation had been achieved by this method compared to

conventional synthesis. Although the crystallinity and the homogeneity of the samples made

by this method was not as high as the one of samples made by the conventional method, but

mesoporous material with intersecting channel system has been obtained. Some of the

synthesis experiments yielded better and homogeneous products. For the first time MCM-48

was synthesized using microwave heating technique and the synthesis conditions were

developed after several attempts. The microwave heating method could be imagined as a time

and energy efficient route for obtaining MCM-48 type mesoporous materials with

intersecting channel system.

49

2.5 References

1. C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710.

2. J.M. Kim, S.K. Kim, R. Ryoo, Chem. Commun. 2 (1998) 259.

3. M.W. Anderson, Zeolites 19 (1997) 220.

4. R. Köhn, M. Fröba, Catal. Today 68 (2001) 227.

5. W.Z. Zhang, T.J. Pinnavaia, Catal. Lett. 38 (1996) 261.

6. A. Sayri, Chem. Mater. 8 (1996) 1840.

7. M.S. Morey, A. Davidson, G.D. Stucky, J. Porous. Mater. 6 (1998) 195.

8. J.C. Vartuli, C.T. Kresge, W.J. Roth, S.B. McCullen, J.S. Beck, K.D. Schmitt, M.E.

Leonowicz, J.D. Lutner, E.W. Sheppard, in: Designed Synthesis of Mesoporous

Molecular Sieve Systems Using Surfactant-Directing Agents, Chapter 1, Advanced

Catalysts (1996) pp. 1.

9. P.T. Tanev, T. Pinnavaia, Science 267 (1995) 865.

10. M.G. Clerici, P. Ingallina, J. Catal. 240 (1993) 71.

11. J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T-W

Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.L. Schlenker, J. Am.

Chem. Soc. 114 (1992) 10834.

12. G.A. Ozin, H. Yang, I. Sokolor, N, Coombs, Adv. Mater. 9 (1997) 662.

13. C. Gabriel, S. Gabriel, E.H. Grant, B.S.J. Halstead, D.M.P. Mingos, Chem. Soc. Rev. 27

(1998) 213.

14. C.S. Cundy, Collect. Czech. Chem. Commun. (1998) 63.

15. J.C. Jansen, A. Arafat, A.K. Barakat , H. van Bekkum, in: M.L. Occelli, H. Robson

(Eds.), Synthesis Of Microporous Mater. Vol. 1, Chap. 33, , Van Nostrand Reinhold,

New York, 1992, pp. 507.

16. A. Arafat, J.C. Jansen, A.R. Ebaid, H. van Bekkum, Zeolites 13 (1993) 162.

17. J.P. Zhao, C. Cundy, J. Dwyer, Stud. Surf. Sci. Catal. 105 (1997) 181.

50

18. C.G. Wu, T. Bein, Chem. Commun. (1996) 925.

19. S.E. Park, D.S. Kim, J.S. Chang, W.Y. Kim, Stud. Surf. Sci. Catal. 117 (1998) 265.

20. H. Chon, S.K. Ihm, Y.S. Uh, Stud. Surf. Sci. Catal. 105 (1997) 181.

21. G. Lischke, B. Parlitz, U. Lohse, E. Schreier, R. Fricke, Appl. Catal. A 166 (1998) 351.

22. M. Park, S. Kormarneni, Microporous Mesoporous Mater. 20 (1998) 39.

23. A. Steel, S.W. Carr, M.W. Anderson, Chem. Mater. 7 (1995) 1829.

24. B. Marler, U. Oberhagemann, S. Vortmann, H. Gies, Microporous Mater. 6 (1996) 375.

25. A. Chenite, Y. Lepage, A. Sayary, Chem. Mater. 7, 1995, 1015.

51

3. Post-synthetic impregnation of MCM-48 by CuO, ZnO and Cu/ZnO

3.1 Introduction

The structure of MCM-48 with a set of two independent three dimensional channel

systems, provides a large surface area and high accessibility of pore space within the porous

network structure. Due to its three dimensional pore structure, and the superior accessibility,

the mesoporous MCM-48 phase is very interesting as a support for catalytically active metal

species such as Al1, B2, Ti3, Fe4, Cu5 onto or within the silica walls.

Another area of application-oriented interest for OMMs is their use as carrier for

functional molecules or nanoparticles.6-9 Of particular interest are the silica based OMMs

because of their transparency, high thermal and chemical stability, and their mechanical

robustness. The advantage of the encapsulation is the protection of the functional molecules

or nanoparticle through the OMM silica-framework from oxidation, decomposition or

dissolution in the course of the reaction with solvent or atmosphere, and the size confinement

through the mesopores tailoring the particles in the process of formation and growth.10, 11

There is a report on nanoparticles of (Fe, Co, and Ni oxide) in MCM-48 host system.12

However, the protection by the silica sheath in aqueous solutions under hydrothermal

conditions is rather limited.13 Only in recent years, there is continuous progress being made

by improving the resistance of the silicate framework against hydrolysis by post-synthetic

modifications of the chemical and physical properties of the OMM carrier. A number of

reports have shown that synthesis strategies or post synthesis treatment can improve the

hydrothermal stability of the silicate OMM considerably.14-17 There are also reports on the

improvement of the mechanical stability of the material through post-synthesis treatment18

and increased wall thickness through modified syntheses.19

Cu-modified zeolites have attracted much attention due to their catalytic performance in

redox reactions, e.g. the decomposition of NO,20 the selective reduction of NO by

hydrocarbons21 or by ammonia,22 the cyclodimerization of butadiene.23 Cu-Zn mixed oxides

52

are of great importance for industrial scale catalytic processes like low pressure methanol

formation from synthesis gas and steam reforming of methanol yielding H2 and CO2.24

There are two different ways to introduce nanoparticles into the mesoporous MCM-48

system.

• Adding the metal ions to the synthesis gel prior to the hydrothermal treatment

• Post-synthetic incorporation

The lowering of long-range order of the final product as a result of distortion of the liquid

crystalline template by the metal ions during synthesis is one of the disadvantages of the first

method.

Starting with a well crystalline highly ordered material of high inner surface is the main

advantage of the post-synthetic incorporation method. The post synthetic modification can be

divided into two parts

♦ Grafting of precursor species from the gas phase or dry vapor phase

♦ Wet impregnation technique

Gas phase introduction leads to inefficient low loading without affecting the host

material, whereas the wet impregnation technique allows high loading but needs a careful

preparation of the material. There are chances of destruction of the host structure.

The characterization of the product, mainly the structural state and the particular

properties of chemical state of the nanostructured sorbate are difficult and require the

combination of various techniques. The location of the metal species and their state remain

often uncertain despite the employment of different comprehensive characterization

techniques, such as X-ray diffraction, high resolution transmission electron microscopy,

nuclear magnetic resonance, N2-physisorption, XPS25, XANES, and EXAFS26,27 etc.

This chapter contains the

♦ Synthesis of metal/metal oxide impregnated MCM-48 with species such as Cu, CuO,

Cu2O, ZnO in MCM-48

53

♦ Characterization of the as synthesized, calcined samples and also the post synthetic

modified samples by different physico-chemical techniques such as XRD, TEM, NMR

N2-physisorption, EXAFS etc.

3.2 Experimental

3.2.1 Synthesis of Cu/ZnO-MCM-48

In a typical wet impregnation procedure first a solution mixture of Cu-acetate and Zn-

acetate was prepared by stirring for 15 minutes. After that, the calcined MCM-48 sample was

added to the solution mixture and the stirring was continued for another 30 minute. The upper

liquid was decanted and the procedure was repeated twice. The stirring was continued for 2 h.

Finally, the solid was filtered off and carefully washed with water and then dried at 110 °C.

The thermal decomposition of the acetates was carried out in a programmed calcination oven

in a flow of air at 300 °C for 5 h. For getting different molar concentrations of Cu and Zn in

MCM-48, different amount of Cu and Zn were employed at the time of impregnation. This

impregnation process was repeated (for maximum 4 times) to load higher amounts of metal

acetates in the samples.

3.2.2 Synthesis of pure ZnO-MCM-48 and CuO-MCM-48

The synthesis procedure for the introduction of Zn and Cu inside MCM-48 with

respective acetate solutions was the same as described for Cu/ZnO-MCM-48. 1 M Zn-acetate

solution was prepared my adding 21.95 g zinc acetate in 100 ml water with stirring at room

temperature. The following procedure was same as described for Cu/ZnO-MCM-48. Zn

modified MCM-48 was prepared via an organometallic route also. In this process first MCM-

48 was dried in a vacuum oven at 180 °C for 17 h. After appropriate drying, the sample was

transferred in a three-mouth flask. During this transformation the sample could have been

exposed to air for maximum 1 minute. Finally 0.47 M Zn(C2H5)2 was added into the flask and

toluene was used as solvent. The mixture was stirred at room temperature for 6h, filtered off

and washed with toluene. The solid was dried over night at room temperature before being

54

calcined at 540 °C in air for 5 h (heating ramp 1 °C/min). The ZnO-MCM-48 materials

prepared via aqueous and organometallic routes will be differentiated by the labels (aq) and

(om), respectively.

In case of copper samples first saturated Cu-acetate solution was prepared at room

temperature by mixing 6 g Cu-acetate in 100 ml water with stirring. After that MCM-48 was

added to the solution. The impregnation process, filtration, drying and calcination were same

as Cu/ZnO-MCM-48 sample. The molar concentration of Cu in the CuO-MCM-48 sample

was varied by adding different amounts (low to high) of Cu-acetate at the time of

impregnation. This impregnation process was repeated (for maximum 4 times) to load higher

amounts of Cu in the silica support.

3.3 Results and discussion

MCM-48 samples impregnated with ZnO, CuO, CuO/ZnO are listed in Table 3.1 with

elemental analysis. From the Table 3.1 it is clear that the Cu, and Zn both metal content

increased with successive impregnation by metal acetates. But the metal uptake is different

for Zn samples prepared by aqueous route and organometallic route. The metal uptake is

much higher in the organometallic route (17 wt %) than in the aqueous route (4 wt% after 1st

impregnation). Interesting findings are observed working with Cu/ZnO system. For making

1:1 Cu : Zn more than one experiments were carried out. The input ratio (experimental) and

the output ratio (revealed by ICP) did not coincide for the Cu/ZnO-MCM-48 samples. When

0.15 (M) Cu-acetate and 0.5 (M) Zn-acetate solution was mixed for making Cu/ZnO-MCM-

48 the output ratio comes as 6.2 wt % Cu and 3.8 wt % Zn. The uptake of Cu was much more

(just dubble) than Zn. However, the ICP output was 3.14 : 5.84 when the ratio of Cu : Zn was

0.05 M : 0.5 M. Finally for making 1 : 1 Cu : Zn molar solution the intial molar ratio was

taken in between the above two ratio, i.e. Cu : Zn = 0.1 : 0.5 and after three times

impregnation, from the elemental analysis the wt % of Cu and Zn was obtained as 8.92 and

55

8.77. The uptake of Cu is higher in the presence of Zn and the ratio of Cu : Zn as 1: 5 gives

the final ratio of Cu : Zn as 1 : 1.

Table 3.1 Experimental and ICP results of ZnO, CuO and Cu/ZnO loaded MCM-48

Experimental Elemental analysis by ICP

Zn impregnated MCM-48 Wt % Si Wt % Zn

1st impregnated 1 M zinc acetate in H2O 27.33 3.99

2nd impregnated 1 M zinc acetate in H2O 26.51 6.03

3rd impregnated 1 M zinc acetate in H2O 27.01 6.96

Zn-MCM-48 (om) 0.47 M diethyl zinc in toluene 17

Cu impregnated MCM-48 Wt % Si Wt % Cu

1st impregnated Saturated solution of Cu-acetate 28.17 3.11

2nd impregnated Saturated solution of Cu-acetate 28.40 4.73

3rd impregnated Saturated solution of Cu-acetate 28.08 6.52

4 times impregnated Saturated solution of Cu-acetate 7.45

1st impregnated 0.05 M Cu-acetate in H2O 1.17

1st impregnated 0.0016 M Cu-acetate in H2O 0.15

1st impregnated 0.006 M Cu-acetate in H2O 0.36

1st impregnated 0.003 M Cu-acetate in H2O 0.18

Cu and Zn impregnated MCM-48 Wt % Cu Wt % Zn

1st loaded Cu (0.15M) : Zn (0.5M) in H2O 6.2 3.8

1st loaded Cu (0.05M) : Zn (0.5M) in H2O 3.14 5.84

1st loaded Cu (0.1M) : Zn (0.5M) in H2O 5.6 4.3

2nd loaded Cu (0.1M) : Zn (0.5M) in H2O 7.8 6.9

3rd loaded Cu (0.1M) : Zn (0.5M) in H2O 8.92 8.77

56

In systematic studies of MCM-41/sorbate system it was shown that the uptake of

molecules in the pores of mesoporous materials decreases the peak intensity in the diffraction

diagram.28 The higher the electron density of a sorbate molecule the lower the peak intensity

in the powder diagram is. However, the desorption of the molecules restores the scattering

contrast leading to diffraction properties of the starting material again showing that the

mesoporous silicate framework remains unchanged during the sorption/desorption process.

Calibration of the MCM-41/sorbate system even would allow to determine quantitatively the

amount of sorbate inside the channel volume. Similar diffraction properties can be measured

for MCM-48 making it simple to decide qualitatively on the uptake of sorbate species.29 X-

ray powder diffraction experiments of the impregnated product still containing hydrated

water showed only weaker intensity maxima in the diffraction pattern (Fig. 3.1).

a

b

Fig. 3.1: XRD patterns of as made 1st (a) and (b) 3rd Cu/ZnO loaded MCM-48

In order to study the systematic changes induced by the metal salt uptake, the impregnation of

MCM-48 was repeated for several samples up to 3 times. To remove the solvent and to

57

segregate the metal acetate sorbate most likely in the channel intersections, the sample was

first dried gently at 110 °C in an oven for 24h. In a second step, the dry sample was calcined

at 300 °C again in order to decompose the acetate anion. The loss of scattering intensity

remained unchanged by the thermal treatment. In Fig. 3.1 the diffraction diagram of MCM-48

after 1st and 3rd sorption treatments and drying is shown. The scattering contrast and the

resolution in the X-ray powder diagram was reduced with each impregnation step, however,

the mesopore structure of the silicate framework is maintained. As expected from the

experiments cited above, the scattering contrast between pore and pore wall in MCM-48

decreases considerably after the impregnation of MCM-48 with the metal acetate solution.

The successive loading also shows that the mesopores are still open after impregnation and

drying. This systematic change in scattering properties of mesoporous materials proves the

uptake of the metal acetate inside the channel pores and its thermal decomposition to the

corresponding oxide. However, powder XRD cannot rule out degradation of the mesoporous

framework. A significant change in the X-ray powder diagram, however, is observed for the

first time after the calcination of the materials after the 3rd impregnation (Fig. 3.2). The first

peak with index (211) at 2θ = 2.56° almost vanished and the rather weak signal of the (220)

reflexion at 2θ = 2.98° remains as the main maximum. This feature is not yet understood. In

order to check for bulk oxide signals, powder XRD-diagrams of the composite have been

recorded up to 50° 2θ. No additional diffraction peak was observed. This is in agreement with

the expected line broadening of particles of less than 50Å diameter with high disorder.

However, in cases where the segregation and precipitation of metal oxide particles on the

outer surface was observed in the TEM, bulk signals showed up in the XRD-diagram.

Instead, the presence of XRD-signals of bulk oxides was considered as indicator of externally

precipitated sorbate.

58

Fig. 3.2: XRD patterns of as made (a) and calcined (b) 3rd Cu/ZnO loaded MCM-48

a

b

The 29Si NMR spectra also showed no changes with respect to the corresponding calcined

sample. However, the T1 relaxation time was significantly shorter after impregnation

indicating a different relaxation mechanism caused by the metal oxide salt (Fig.3.3). The

characteristic data from the 29Si NMR spectra for the as synthesized, calcined and metal

loaded MCM-48 samples are given in Table 3.2.

Table 3.2 Summary of the 29Si NMR measurements

MCM-48 as syn. MCM48 calc. MCM-48 loaded

Q2 Si -89 ppm -91 ppm -91 ppm

Q3 Si -98 ppm -102 ppm -102 ppm

Q4 Si -107 ppm -111 ppm -111 ppm

Ratio Q2:Q3:Q4 2:78:34 1:10:112 2:8:63

59

(ppm)-170-160-150-140-130-120-110-100-90-80-70-60

0.0e+000

5.0e+007

1.0e+008

1.5e+008

2.0e+008

2.5e+008

3.0e+008

3.5e+008

4.0e+008

4.5e+008*** Current Data Parameters ***

NAME : cuzn3b~1

EXPNO : 1

PROCNO : 1

*** Acquisition Parameters ***

D[1] : 20.0000000 sec

D[3] : 0.0000500 sec

DATE_t : 08:36:07

DBPNAM2 :

DE : 71.4 usec

NS : 2800

NUCLEUS : 29Si

O1 : -4500.00 Hz

P[1] : 6.5 usec

PAPS : CP

PARMODE : 1D

PULPROG : hpdec

RG : 4096.0000000

SFO1 : 79.4855000 MHz

SW : 125.8091 ppm

TD : 1024

*** Processing Parameters ***

LB : 0.00 Hz

TDeff : 1024

TDoff : 0

*** 1D NMR Plot Parameters ***

NUCLEUS : 29Si

CuZn3, 29 Si, 0.18 mol Cu : 0.99 mol Zn

a

b

Fig. 3.3: 29Si NMR of (a) calcined empty and (b) Cu/ZnO loaded MCM-48

Although radiation damage is a serious problem in TEM experiments of silicas in general,

images of the mesoporous framework have been obtained using low dose techniques.30,31

The periodic arrangement of walls and channels in mesoporous materials leads to periodic

imaging contrast in the TEM experiment. Carefully selected and isolated particles were

analyzed for metal oxide particles in SAED (selected-area electron diffraction) experiments.

In Fig. 3.4 a particle is shown together with the EDX analysis. The regular contrast variation

in the image shows the intact MCM-48 silicate framework. In addition, few dark spots with

higher scattering contrast are visible as part of the regular framework. They are scattered

irregularly over the particle without translational coherence. This seems to indicate that

clustering of the metal oxide has occurred as consequence of the thermal treatment and that

the oxide particles segregate in specific locations inside the pores of the mesoporous

framework, most likely in the channel intersections. Since the sample was mounted on a Cu

grid, Cu analysis, therefore, is meaningless. However, there is a clear Zn signal at 8.7 keV.

60

These results indicate that Zn and Cu are incorporated inside MCM-48. The mesoporous

structure is maintained and the pore space is still accessible for further sorbate molecules.

0 1 2 3 4 5 6 7 8 9 100

4

8

12

energy / keV

cps

CuL

a, Z

nLa

OK

a

SiK

a

CuK

aZ

nKa

CuK

ßZ

nKß

Fig. 3.4: An isolated Cu/ZnO loaded MCM-48 particle along with EDX analysis

Using a calcined, empty MCM-48 sample as reference, N2 sorption measurements show that

the pore diameter decreases after impregnation by 8Å whereas the surface area decreased

only by ~25% to 870 m2/g (Table 3.3).

Table 3.3 Result of the N2-physisorption of calcined MCM-48 and Cu/ZnO MCM-48

Reference MCM-48 Cu/ZnO loaded MCM-48

Pore diameter 25 Å 17 Å

Surface area (internal) 1150 m2/g 870 m2/g

Pore volume 0.93 cm3/g 0.43 cm3/g

61

As metaloxide containing MCM-48 silicas still show mesoporosity, it can be concluded that

the metaloxide particles formed (or coating of the inner surface, which has discussed later in

EXAFS part) inside the pore system. Nevertheless, no complete filling of the pores could be

observed in any case.

Additional support for the introduction of the transition metal oxides into the pore

system is provided by surface analysis of Cu/ZnO-MCM-48 compared with reference

compounds. The metal content was analyzed with XPS binding energies and XAES energies

giving total Me/Si atomic ratios of 0.051 for Cu and 0.046 for Zn (Table 3.4) The atomic

ratios in the surface region are not larger than the bulk atomic ratios, they are even somewhat

smaller. Hence, there is no enrichment of Cu or Zn in the external surface region, which

indicates that both elements are well dispersed over the pore system.

Table 3.4 XPS binding energies, XAES kinetic energies and surface atomic ratios of Cu/ZnO-MCM-48

BE (2p3/2), eV KE(Auger), eV � eVa Me / Si atomic ratios

XPS total

Cu/ZnO-MCM-48 - Cu

Reference: CuO32

Cu2O32

933.2

933.8 932.3

913.6

917.6 916.6

1846.8

1851.4 1848.9

0.030 0.051

Cu/ZnO-MCM-48 - Zn

Reference: ZnO33

1022.2

1022.1

987.3

987.7

2009.5

2009.8

0.016 0.046

aAuger parameter αCu = BE (Cu(2p3/2)) + KE (Cu L3M45M45) αZn = BE (Zn(2p3/2)) + KE (Zn L3M45M45)

In the case of copper, additional support for this conclusion can be taken from the

kinetic energy of the Cu L3M45M45 Auger line, which is significantly below the values typical

for the bulk oxides. It has been shown earlier that such deviation indicates a highly dispersed

or disordered state of the Cu oxide species.34,35 In the presence of a pore system, these should

be located in the pores since extra-pore material should aggregate to bulk-like oxide particles

upon calcination. To exclude possible influences of surface charging on the XPS/XAES data,

62

comparison is usually performed with the Auger parameter, which is the sum of the XPS

binding energy and the Auger kinetic energy and does not depend on referencing of the BE

scale. The Cu Auger parameter in Cu/ZnO-MCM-48 is clearly below that of Cu2O and CuO

(Table 3.4), indicating intrapore location of the copper. Unfortunately, the Zn Auger

parameter does not possess this diagnostic potential. On the other hand, the observation that

the Zn Auger parameter of Cu/ZnO-MCM-48 coincides with that of ZnO is not unexpected

and permits no conclusion about the location of the Zn species.

The EXAFS spectra of the Cu/ZnO-MCM-48 material are shown in Fig. 3.5. The k-

space spectra are given in panels c and d and in panels a and b, the absolute part of the

Fourier-transformed spectra (weighted by k2) are compared with those of reference oxides

(CuO and ZnO). The results of the spectral analysis are given in Table 3.5. Although there is

a second shell in the Cu K-edge spectra, it is obvious that there is no order at higher distances,

i.e., the particles formed are very small. In the Cu-edge spectrum, the distance of the second

scattering event corresponds to that in CuO. In the ZnK-edge spectrum, there are two very

weak scattering events beyond the O shell next to Zn, but their distance deviates from that of

the major Zn shell in ZnO. Spectra taken with the same material reduced at different

temperatures confirm this observation.36 In the quantitative analysis (Table 3.5), almost 4 O

atoms are found around the Cu atoms as in CuO. However, the first Cu shell has a smaller

coordination number than that found in the bulk oxide (NCu = 4). From this it can be

concluded that the oxide clusters formed are small and/or disordered at higher distances.

63

0

1

2

3

-1

0

1

2 4 6 8 10 12-2

-1

0

1

2d

c

b

a

Cu-Zn-MCM-48

Cu-Zn-MCM-48

CuO

FT(χ

k2 )

0 2 4 6 80

2

4

o

ZnO

o

FT(χ

k2 )

r, A Zn edge

Cu edge

χ k2

k, A-1

χ k2

0 2 4-2

-1

0

1

2

O S i

O

o

FT(

χ k2 )

r , A

0 2 4-2

-1

0

1

2

O , CuO

o

FT(χ

k2 )

r , A

Fig. 3.5 The EXAFS spectra of Cu/ZnO MCM-48 which is presented in the Fourier-transformation

Table 3.5 EXAFS parameter obtained from the analysis of the spectra shown in Fig. 3.5

O O Cu edge model R, Å N R, Å N R, Å N

Cu 1.93 (± 0.02) 3.35 (± 0.45)

2.78 (± 0.38) 0.45 (± 1.05) 2.94 (± 0.08) 1.6 (± 1.6)

O R, Å N R, Å N R, Å N

Zn 1

2

3

1.9(± 0.003)

1.9(± 0.003)

1.9(± 0.005)

4.2(± 0.1)

4.2(± 0.1)

4.4(± 0.1)

Si: 3.167 (± 0.006)

Si: 3.13 (± 0.02)

O: 2.93 (± 0.02)

2.5 (± 0.2)

1.0 (± 0.2)

2.8 (± 0.6)

Si: 3.52 (± 0.01)

Zn1: 3.40 (± 0.03)

Si: 3.49 (± 0.03)

4.2 (± 0.3)

3.7 (± 0.7)

2.6 (± 0.7)

Debye-Waller factors (σ2) below 10-2Å2 if not indicated otherwise1σ2 = 2.5 * 10-2Å2, Crystallographic data for CuO: Cu-O: R = 1.95-1.96 Å, N = 4, Cu-Cu: R = 2.90 Å, N = 4,

64

The analysis of the Zn K-edge spectra is more complicated. The first shell can be fitted

with 4 oxygen atoms as in ZnO (Table 3.5). The scatters beyond this shell can be O, Si, or

Zn. The analysis of these combinations allowed to exclude Zn as the next neighbour on the

basis of significantly larger deviations between the model and experiment than with Si or O

at this position. For the same reason, a model with 3 oxygen shells was abandoned. The best

models are compared in Table 3.5, and the least accurate of them is exemplified in Figure

3.5d. Model 1 is unlikely because of the high resulting number of Si neighbours. Model 3

implies a monoatomic dispersion of Zn on the silicate walls, or a complete randomness of the

distance of possible Zn neighbours. In model 2, there are some Zn (or Cu) neighbours around

the Zn atom, but at a distance different from those in ZnO (3.21 Å, 3.25 Å), and with a high

statical disorder (high Debye-Waller factor at liquid nitrogen temperature). The latter would

indicate the formation of a highly dispersed and disordered amorphous structure on the

silicate walls, where the neighbours might be Zn or Cu from the Cu oxide clusters. XPS,

EXAFS, XANES, N2-adsorption results have been obtained in collaboration with Technical

Chemistry Department, Ruhr University Bochum. The results have been published in

different journals.36, 37 We duly acknowledge their work which is included in this present

thesis.

Similar observations are noticed with CuO-MCM-48 and ZnO-MCM-48. The intensity

in the XRD patterns decreased with higher loading of metal oxide (Fig. 3.6 and Fig 3.7a). The

XRD pattern of two different ZnO-MCM-48 samples prepared from organometallic and

aqueous route is given in Fig. 3.7b. From the XRD diagram it has been noticed that the peak

intensity of the sample prepared by organometallic route has decreased much more than the

three times impregnated aqueous route sample.

65

Fig. 3.6: XRD patterns of different Cu loaded calcined MCM-48, (a) 1st loaded (b) 2nd loaded and (c) 3rd loaded

a

b

c

a

b

c

Fig. 3.7a : XRD patterns of different Zn loaded ( Zn-acetate was used as Zn source)calcined MCM-48, (a) 1st impregnated (b) 2nd imp. and (c) 3rd impregnated

66

Fig. 3.7b : XRD patterns of 1st Zn loaded calcined MCM-48, (a) Zn-acetate was used as Zn source ((b) diethylzinc was used as Zn source (om).

a

b

The ZnO-MCM-48 (om) has 17 wt % Zn (revealed by ICP ) whereas the ZnO-MCM-48 (aq)

contains only 7 wt % Zn after 3rd loading. Higher the metal uptake lower is the peak

intensity and this feature has been reflected in the powder XRD pattern.

The porosity results are shown in Table 3.6.

Table 3.6 Porosity data for empty MCM-48 and CuO, ZnO and Cu/ZnO impregnated MCM-48

Sample BET Surface area

m2/g

Pore diameter

Å

Pore volume cm3/g

MCM-48

CuO-MCM-48 (1st imp.) aZnO-MCM-48 (1st imp.) aZnO-MCM-48 (3rd imp.) bZnO-MCM-48 (1st imp.)

1257

867

900

770

400

25

20

21

20

20

1.8

0.61

0.85

0.60

0.63

asample prepared by aqueous route, bsample prepared by organometallic route

67

The reduction of specific surface area was observed for all the samples. The intrapore

formation of the metal oxide nanoparticles within the mesopores reduces the mean pore

diameter. Although CuO formed nanosized particles but it was observed for Cu/ZnO-MCM-

48 that ZnO only coated the inner surface (discussed before). But for ZnO-MCM-48 prepared

by organometallic route the surface area decreased to 400 m2/g, may be due to the high

content of metal oxide. The reduced pore volume and pore diameter was almost same for

material prepared via aqueous and organometallic route. It is not so easy to say whether the

metal oxides are coated the inner surface or form nanoparticles considering N2-physisorption

results. For confirming the exact state of the metal oxides some other techniques such as

EXAFS is also needed. The N2-physisorption isotherms of CuO and ZnO-MCM-48 are

shown in Fig. 3.8a and Fig. 3.9a and pore size distribution curves are given in Fig. 3.8b and

3.9b.

0.0 0.2 0.4 0.6 0.8 1.0

VN

2 (ccg

-1)

0

1

00

2

00

30

0

400 Blue: adsorption

Red: desorption

Relative pressure (p/p0)

Fig. 3.8a : N2-physisorption isotherm of 1st Cu loaded MCM-48

68

0.0 0.2 0.4 0.6 0.8 1.0

Relative pressure (p/p0)

VN

2 (ccg

-1)

0

100

2

00

300

4

00

500

Blue: adsorptionRed: desorption

Fig. 3.9a : N2-physisorption isotherm of 1st Zn loaded MCM-48 (aq)

0 20 40 60 80 100

Dv

(d),

cm3 Å

-1 g

-1

0

0

.01

0

.02

0

.03

0

.04

0.05

0.06

0.07

0

.08

Pore diameter (Å)

Fig. 3.8b: BJH pore size distribution curve of 1st Cu loaded MCM-48

69

0 20 40 60 80 100

Pore diameter (Å)

0

0.0

2

0.04

0.

06

0.08

0.

10

0.1

2

Dv

(d),

cm3

Å-1

g -1

Fig. 3.9b: BJH pore size distribution curve of 1st Zn loaded MCM-48 (aq)

The isotherms as known from the literature, with a steep increase around p/p0 = 0.3, which

indicates a narrow pore size distribution. All the isotherms follow type IV, which is typical

for mesoporous materials.38 Regular mesopore structure is confirmed by a step increase of

capillary condensation in a partial pressure range of nitrogen, which is expected for

mesopores. There is a clear difference between the isotherm of pure empty MCM-48 (Fig. 2.3

in chapter 2) and metal oxide (CuO and ZnO) loaded MCM-48 material. The p/p0 coordinate

of the inflection point in the isotherm depends on the pore size. For pure MCM-48 material

the steep increase around p/p0 0.3 in the isotherm, which indicates a narrow pore size

distribution. The sharpness in this step, suggests a uniform size pore system. However, for

CuO and ZnO loaded MCM-48 samples the sharpness of the isotherm around p/p0 0.3 is

decreased indicating the pore filling by metal oxides.

70

The Fourier-transformed (k2-weighted, absolute values) EXAFS spectra of CuO

impregnated MCM-48 compared with reference bulk CuO is given in Fig. 3.10 and analysis

of the spectra is given in Table 3.7.

0 2 4 6 8

Cu-Sil

Cu-MCM-48

CuO 0.5

FT (χ

k2 )

r, Å

Fig. 3.10: EXAFS spectra of CuO-MCM-48 compared with bulk CuO

Table 3.7 Model parameters derived from Fig. 3.10: structure of the coordination spheres around the Cu atoms

Sample path R / Å N σ2/ 10-3 Å2 E0 / eV

Cu-MCM-48b

Cu-Oa

Cu-Cua

Cu-Cua

Cu-Cua

1.925 ±0.008

2.96

3.04

3.15

4.0 ±0.2

2.5

3.1

0.8

6.2 ±1.5

9.1

33

4.0

4.2 ±0.5

1.9

1.9

1.9 a Crystallographic data for CuO: Cu-O: R = 1.95-1.96 Å, N = 4, Cu-Cu: R = 2.90 Å, N = 4, Cu-Cu: r = 3.08 Å, N = 4, Cu-Cu: R = 3.17 Å, N = 2,39 badditional O shell at R = 2.80 Å (N = 0.3) and at R = 3.60 Å (N = 4.4), σ2 = Debye-Waller factors, E0 = Binding energy N = coordination no. and R = Bond distance d (Cu-x)

71

Beyond the first shell, scattering events of Cu species in MCM-48 are at the same distances

as in CuO, but exhibit strong amplitude decay. This indicates the formation of very small

copper oxide particles. The analysis of the Cu K-EXAFS spectra (Table 3.7) confirms this

showing the same neighbors at similar distances as in CuO, but with lower coordination

numbers.

The EXAFS spectra of ZnO-MCM-48 prepared by organometallic route (spectra b in

Fig 3.11) exhibit some differences than the Zn sample prepared by aqueous route.

Comparison of EXAFS spectra of Zn-silicalite from Bochum and Mülheim with that of ZnO

0 2 4 6 80

1

2

3

4

5

Zn-MCM-48 (Bochum) ZnO Zn-MCM-48 (Mülheim)

FT(χ

(k)·k

2 )

R, Å

Fig. 3.11: EXAFS spectra of Zn modified MCM-48 Fourier-transformed ZnK spectra, (a) ZnO as a reference (b) ZnO-MCM-48 (organometallic route) (c) organometallic route but prepared by different group ( MPI Mülheim)

a

bc

The scattering events beyond the first neighbour were the same in ZnO and in the Zn species

in siliceous matrix. In particular, in the region of the intense second shell in ZnO (Zn), there

is a clear scattering event for the sample prepared by organometallic route. In this material

the second shell coincides with the Zn shell in ZnO, but with small amplitude, indicating the

formation of either extremely small or disordered ZnO particles. If the MCM-48, which was

used for impregnation is not properly dried then there is a chance of interaction of diethyl

zinc with absorbed water to form ZnO aggregates. Nevertheless, there is a possibility of the

72

presence of the ZnO particles outside the porous matrix. In order to check for bulk oxide,

TEM experiment was done on it. Fig. 3.12 shows the TEM image of the sample. After

impregnation and careful calcination of the sample TEM images prove that the structure of

the mesoporous silicate framework is still maintained and show periodic order. Large ZnO

particles, which may be situated outside the silicate framework, were not found in the TEM

images. In the EDX analysis there is a clear Zn signal at 8.7 keV.

0

2

4

6

0 2 4 6 8energy / keV

cps

CuK

◊

CuK

◊

SiK

◊

CuL

◊

OK ◊

CK

◊

ZnK ◊

ZnK

◊

Fig. 3.12: TEM image of an isolated MCM-48 particle loaded with Zn (om)along with EDX analysis

10

The sample was mounted on a Cu grid and for that reason there is a Cu signal in the EDX

spectra. With the combination of EXAFS spectra and TEM image it is clear that there is a

formation of ZnO particles, which are situated inside the porous matrix.

73

3.4 Conclusion

Impregnation with Zn and Cu-acetates inside porous siliceous matrix led to the

incorporation of metal oxides inside the mesopore system of MCM-48. The structural

properties of the resulting materials were investigated by a combination of different physico-

chemical characterization techniques. XRD, TEM, NMR, N2-physisorption, XPS, and

EXAFS analyses showed that the oxides are inside, however with different structural

properties. MCM-48, which was used for impregnation, was well crystalline and exhibits

high surface area and pore volume. With successive impregnation the decrease in intensity

directly proved the uptake of the metal acetate inside the channel pore. The TEM images of

the metal/metal oxide impregnated MCM-48 showed periodicity of the structure. There was

no indication of structural damage or destruction after metal impregnation. In the EDX

spectra a clear Zn signal was observed for the Zn impregnated sample. The wet impregnation

process led to spread Zn-oxide most likely on the surface of the silicate channel wall, reveled

by EXAFS analysis. However, Cu-oxide showed a completely different picture. It was nano-

dispersed in organized particles. The Zn-MCM-48 prepared by organometallic route led to

the formation of ZnO particles inside the MCM-48 pores (< 3 nm), which might be a

promising approach towards the preparation of effective catalyst for methanol synthesis. As

shown in this study, only the combination of various bulk and surface analytical techniques

provides a most complete picture of the nature of the composite material.

74

3.5 References 1. Z. Luan, C.F. Cheng, W. Zhou, J. Klinowski, J. Phys. Chem. 99 (1995) 1018.

2. U. Oberhagemann, I. Kinski, I. Dierdorf, B. Marler, H. Gies, J. Noncryst. Solids 197

(1996) 145.

3. W. Zhang, M. Förba, J. wang, P.T. Tanev, J. Wong, T.J.J. Pinnavaia, J. Am. Chem. Soc.

118 (1996) 9164.

4. B. Echchahed, A. Moen, D. Nicholson, L. Bonneviot, Chem. Mater. 9 (1997) 1716.

5. M. Hartmann, S. Racouchot, C. Bishof, Chem. Commun. (1997) 2367.

6. C.G. Wu, T. Bein, Science 264 (1994) 264.

7. I. Kinski, H. Gies, F. Marlow, Zeolites 19 (1997) 375.

8. R. Hoppe, A. Ortlam, J. Rathousky, G. Schulz-Ekloff, A. Zukal, Microporous Mater. 8

(1997) 267.

9. G. Schulz-Ekloff, D. Wöhrle, B. Van Duffel, R.A. Schoonheydt, Microporous

Mesoporous Mater. 51 (2002) 91.

10. F. Schüth, A. Wingen, J. Sauer, Microporous Mesoporous Mater. 44-45 (2001) 465.

11. M. Fröba, R. Kohn, G. Bouffaud, O. Richard, G. van Tendeloo, Chem. Mater. 11 (1999)

858.

12. R. Köhn, M. Fröba, Catal. Today 68 (2001) 227.

13. J.M. Kim, R. Ryoo, Bull. Kor. Chem. Soc. 17 (1996) 66.

14. R. Ryoo, S. Jun, J. Phys. Chem. B 101 (1997) 317.

15. D. Das, C.M. Tsai, S.F. Cheng, Chem. Commun. 5 (1999) 473.

16. R. Mokaya, J. Phys. Chem. B 103 (1999) 10204.

17. L. Chen, T. Horiuchi, T. Mori, K. Maeda, J. Phys. Chem. B 103 (1999)1216.

18. T. Tatsumi, K.A. Koyano, Y. Tanaka, S. Nakata, J. Porous Mater. 6 (1999) 13.

19. D.Y. Zhao, J.L. Feng, Q.S. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D.

Stucky, Science 279 (1998) 548.

75

20. M. Iwamoto, H. Yahiro, Catal. Today 22 (1995) 5.

21. M. Shelef, Chem. Rev. 95 (1995) 209.

22. A. Z. Ma, M. Muhler, W. Grünert, Appl. Catal. B 27 (2000) 37.

23. T.V. Voskoboinikov, V. Coq, F. Fajula, R. Brown, G. McDougall, J.L. Couturier,

Microporous Mesoporous Mater. 24 (1988) 89.

24. M. Hartmann, S. Racouchot, C. Bishof, Microporous Mesoporous Mater. 27 (1999) 309.

25. J. H. Scofield, J. Electron Spectrosc. Relat. Phenom. 8 (1976) 129.

26. K.V. Klementiev, VIPER for windows (Visual Processing in EXAFS Researches),

freeware, Available from www.desy.de/~klmn/vipr.html.

27. A. L. Ankudinov, B. Ravel, J. J. Rehr, S. D. Conradson, Phys. Rev. B 58 (1998) 7565.

28. B. Marler, U. Oberhagemann, S. Vortmann, H. Gies, Microporous Mater. 6 (1996) 375.

29. U. Wertmann, Dissertation, Ruhr-Universität Bochum, 1996, 85.

30. J.S. Breck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.-W.

Chu, D.H. Olson, E.W. Sheppard, S.B. McCullen, J.B. Higgins, J.B. Schlenker, J.Am.

Chem. Soc. 114 (1992) 10834.

31. A. Chenite, Y. Lepage, A. Sayary, Chem. Mater. 7 (1995) 1015.

32. W. Grünert, N.W. Hayes, R.W. Joyner, E.S. Shpiro, M.R.H. Siddiqui, G.N. Baeva. J.

Phys. Chem. 98 (1994) 10832.

33. C.D. Wagner, in: D. Briggs, M.P. Seah (Eds.), Practical Surface Analysis, second ed.,

Wiley, New York, 1990 (Appendix 5).

34. W. Grünert, N. W. Hayes, R. W. Joyner, E. S. Shpiro, M. R. H. Siddiqui, G. N. Baeva, J.

Phys. Chem. 98 (1994) 10832.

35. B. A. Sexton, T. D. Smith, J. V. Sanders, J. Electron Spectrosc. Relat. Phenom. 35

(1985) 27.

76

36. O.P. Tkachenko, K.V. Klementiev, E. Löffler, I. Ritzkopf, F. Schüth, M. Bandyopadhyay,

S. Grabowski, H. Gies, V. Hagen, M. Muhler, Lianhai Lu, R.A. Fischer, W. Grünert,

Phys. Chem. Chem. Phys. 5 (2003) 4325.

37. H. Gies, S. Grabowski, M. Bandyopadhyay, W. Grünert, O.P. Tkachenko, K.V.

Klementiev, A. Birkner, Microporous and Mesoporous Materials 60 (2003) 31.

38. K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T.

Siemieniewska, Pure & Appl. Chem. 57 (1985) 603.

39. R.W.G. Wyckoff, in: Crystal Structures, John Wiley and Sons, New York, Vol. 1,2 1963.

77

4. Synthesis and characterization of TiO2 loaded MCM-48

4.1 Introduction

The mesoporous materials with narrow pore size distribution and the composition

related to zeolites, were first thought to be the geometric extension of the microporous

crystalline zeolites into the mesopore regime. Using the silica-based materials, research was

focused in the beginning on the exploration of the chemical and physical properties of the

mesopore system related to zeolitic properties such as sorption, separation and catalytic

processes. Two of the several reviews about these activities are given here.1,2

In addition to the zeolitic applications, the M41S materials are thought to be

particularly useful as carriers or matrices for functional molecules or nanoparticles because of

their high thermal and chemical stability. Their mesoporous structure could be explored as a

host or matrix to immobilize catalytically active species on or providing nanosize

confinement inside the pore system.3,4-10 In addition, physical, electronic and sensing

properties of other matrix isolated nanoparticles might be of interest.

The inactive property of the siliceous framework of M41S materials leads to the

substitution of silicon by many metals such as Al, Ti, V, Ga to introduce active acid sites.

Substitution of Ti is particularly interesting because of the success of the Ti-substituted

zeolites in oxidation of various organic molecules. Microporous titanosilicates such as TS-1

with MFI structure has proven to be active catalyst for oxidation of phenol to catechol 11 and

TS-2 with MEL has found to be active catalyst for oxidation of phenol to hydroquinone.12

Ti-β was shown to be an active catalyst for hydrogen peroxide oxidation reaction under mild

condition.13

There are several ways to introduce Ti in mesoporous silicas. The conventional way to

introduce Ti in mesoporous silicas is usually achieved by a direct synthesis, i.e. addition of

transition metal cations during hydrothermal synthesis of mesoporous material. However, the

incorporation of this transition metal ion in the silicate framework is not always easy and

78

usually limited to less than 2-3 wt%.14 One of the reasons is that these cations (Ti+4) have a

larger radius than Si+4, and leads to distort the silicate framework. Therefore, the silicate

framework can only be stabilized by accommodation of small metal ion concentration with

lowering the long-range order of the final material.3,15 Post-synthetic incorporation of Ti in

mesoporous silicates is much less investigated, some literature is available, for example on

the use of titanium alkoxides,16 or metallocenes17 for grafting processes. One report is

available which describes the Molecular Designed Dispersion method, by which transition

metal acetylacetonate complexes have been used for the preparation of supported oxides.18

The dispersion of the complex occurs via either a hydrogen bonding or a ligand exchange

path on the silica support. In a number of papers various metal oxides such as TiO2, CdS,

ZnO, and SnO2 have been deposited inside mesoporous MCM-41 in order to study the

physical properties of the nanoparticles inside the mesoporous silica matrix.19-23 Less

attention has been paid to the cubic MCM-48 type materials.24 This is also true for SBA-

type,25,26 STA-type,27,28 and other mesoporous silicate based materials,29,30 which have been

synthesized meanwhile. But so far the pure empty mesoporous MCM-48 has not been used

for the support for nanodispersed titania particles inside the porous matrix.

In the previous chapter the use of MCM-48 as a matrix for Cu and Zn oxides10,21 has

been discussed. In the present study the post synthetic wet impregnation method has been

applied for the incorporation of Ti salts in siliceous MCM-48. So far Ti has been incorporated

in MCM-48 either by grafting technique or adding the metal ion during synthesis. The post-

synthetic wet impregnation process has a great advantage and that is, the opportunity to start

with a highly ordered mesoporous material with a very high inner surface and pore volume.

79

This chapter contains

• The deposition of Ti salts by the post synthetic wet impregnation process within the

pore system of cubic MCM-48 and their subsequent decomposition to Ti-oxides

clusters

• The characterization of the final product by means of X-ray powder diffraction, ICP

AES analysis N2-physisorption, IR-spectroscopy, UV-visible, EXAFS, XANES

spectroscopy techniques along with TEM and EDX.

4.2 Experimental 4.2.1 Synthesis of TiO2 impregnated MCM-48

Two different Ti sources, titanylacetylacetone (TiO(acac)2, Merck) and

tetrabutylorthotitanate (Merck), were used for the impregnation. During the impregnation

process, 0.01 M solution of TiO(acac)2 in water was prepared. The impregnation of the

calcined and dried MCM-48 (0.6 g) was carried out by stirring the sample for 2 h with the

aqueous TiO(acac)2 solution (200 ml). The mixture was then filtered and washed thoroughly

with water and dried at room temperature for overnight. The thermal decomposition of the

metal acetylacetonate was carried out in a calcination oven at 300 °C for 5 h in presence of

air.

On the other hand, for the impregnation with tetrabutylorthotitanate, the standard

MCM-48, which was dried overnight in vacuum at 180 °C, was reacted with a solution of this

Ti source in dry acetone (0.05 M). The mixture was stirred for 6 h at room temperature. As

tetrabutylorthotitanate is moisture sensitive, the whole procedure was carried out in dry N2

atmosphere. Finally, the solid was filtered off with acetone, dried at room temperature and

then calcined at 300 °C for 5 h. The impregnation procedure was repeated for three times.

4.3 Results and discussion

MCM-48, which was synthesized by the hydrothermal heating technique was well

periodically ordered. This ordered MCM-48 was used for titania loading. The powder XRD

80

patterns for the samples loaded with increasing amounts of titania are given in Fig. 4.1. It is

noteworthy that the (220) peak at around 2.8° 2θ was present for all materials even after 2nd

and 3rd loading with titania. This confirms that after the impregnation procedure the structure

of the silica matrix is well maintained.

a

b

c

Fig. 4.1: Powder XRD patterns of (a) empty (b) 1st loaded (c) 3rd loaded TiO2-MCM-48 (tetrabutylorthotitanate was used as Ti source)

From systematic studies of the influence of the sorbate on the intensity of the diffraction

peaks, it is known that the uptake of molecules inside the pores reduces the scattering contrast

between pore wall and pore thus leading to a decrease in peak intensity.31 The higher the

electron density of the sorbate molecule the lower is the residual peak intensity in the powder

diffraction diagram. This is clearly reflected in the diffraction diagrams of the titania loaded

samples in Fig. 4.1. However, the decrease in peak intensity with successive loading of metal

salt may also be due to loss of sample integrity. In order to rule out the degradation of the

sample and to confirm the interpretation of the XRD experiments, complementary TEM

studies were carried out (Fig. 4.2). The integrity of every sample was checked with TEM

81

together with its qualitative titania content using EDX. The results of the analyses always

showed the well ordered and well-maintained silica host structure, no external titania

particles, and a well resolved EDX titanium signal from material deposited inside the pores.

In order to check for titania deposited on the external surface of the silica matrix,

powder XRD with Cu-Kα of all the materials were recorded up to 2θ of 50°. No additional

diffraction peak was observed at higher angles in regular step scan mode and step counting

times of 5 s. In addition, diffraction experiments were carried out using a position sensitive

detector and 36 h exposure time in order to detect minor amounts of titania nanoparticles in

the sample after 3rd loading (Fig. 4.1a).

Fig. 4.1a : (a)Wide angle XRD pattern (36 h exposure time) of 3rd Tiloaded MCM-48 material and (b) XRD patterns of rutile, anatase and brookite

a

b

In the range from 20 to 35° 2θ, in which the major peaks of the three crystalline polymorphs

rutile, anatase and brookite are found, no additional diffraction intensities were recorded

which could be unambiguously attributed to crystalline titania (Fig. 4.1a). These experiments

support the conclusion that no titania has been deposited on the outer surface with particle

82

size 3 nm and that the size confinement of the MCM-48 matrix has led to the formation of

titania particles of size less than the MCM-48 pore diameter (~ 3 nm) inside the pores.

The Ti content of the impregnated samples was determined from EXAFS

measurements and ICP AES analyses (Table 4.1). Within the experimental uncertainty, the

results of the analyses are in agreement with approximately 14 wt % Ti after the third

loading.

Table 4.1 Chemical analysis by ICP and from XANES spectra of Ti loaded MCM-48

Sample Chemical Analysis

(Ti-content)

EXAFS Analysis

(Ti-content)

TiO2-MCM-48 (1) 13 wt % 7 wt %

TiO2-MCM-48 (2) 14 wt % 10 wt %

TiO2-MCM-48 (3) 16 wt % 12 wt %

TiO2-MCM-48 (4) 11 wt % 5 wt %

(1) 1st loaded, (2) 2nd loaded and (3) 3rd loaded Ti-MCM-48 (tetrabutylorthotitanate) (4) 1st loaded, titanylacetylacetonate was used as Ti source,

In Fig. 4.2 and 4.2a TEM images of a well-aligned MCM-48 particle after Ti-

impregnation using titanylacetylacetonate as Ti source and 1st impregnated Ti-MCM-48

using tetrabutylorthotitanate as Ti source together with the EDX analysis are shown. There

are no differences in the TEM images of the TiO2-MCM-48 materials using acetylacetonate

or tetrabutylorthotitanate as Ti source. The lattice fringes of this well-ordered crystallite have

a d-spacing of d ~ 3.2 nm corresponding to (110) planes. The regular contrast variation in the

TEM image shows the intact MCM-48 silicate framework. For all imaged MCM-48 particles

no damage of the periodic structure of the silicate framework was observed.

83

Fig. 4.2: TEM image of an ordered area of TiO2-MCM-48 (1st loaded), titanylacetylacetonate was used as Ti source

0

3

6

9

12

15

0 2 4 6 8 1energy / keV

cps

TiK

aTi

K

CuK

CuK

0

SiK

CuL

OK

CK

Fig. 4.2a : TEM image of an ordered area of TiO2-MCM-48 (1st loaded), tetrabutylorthotitanate was used as Ti source

Sikα

Tikβ

Cuk

α

βCuL

α

α

Tikα

84

It is concluded that no structural damage has been caused by the impregnation procedure and

the consecutive thermal treatment. In the course of the TEM measurements of a large number

of isolated MCM-48 particles as well as aggregates, no large TiOx particles were observed

which may be present outside of the silicate framework and might be responsible for the Ti

signals in the EDX spectrum. The EDX spectrum in Fig. 4.2a shows the O-Kα and Si-Kα

signals from the silicate framework as well as the Ti-Kα and Ti-Kβ signals resulting from the

Ti-impregnation of the calcined MCM-48 sample. The C-Kα signal in the EDX spectrum

results from sample contamination during the TEM and EDX measurements and the Cu

signals in this spectrum are caused by the copper grid for sample preparation. The TEM and

EDX results indicate that Ti is incorporated inside MCM-48 using titanylacetylacetonate or

tetrabutyltitanite as Ti source.

The IR spectrum of Ti-MCM-48 is shown in Fig. 4.3.

400 800 1200 1600 2000 2400 2800 3200 3600 4000

wave number (cm-1)

trans

mitt

ance

Fig. 4.3: IR spectrum of calcined 1st loaded Ti (titanylacetylacetonate was used as Ti source) impregnated MCM-48

IR spectra were used as a standard technique for the characterization of Ti in porous silicates.

85

A sharp band at 960 cm-1 was observed. This band has been assigned to the Si-O stretching

vibration of Si-O-Ti groups in titanosilicates. However, the assignment is not unambiguous

because the signal might as well be due to silanol groups. The IR results are thus not

conclusive and further analyses are needed to really identify the incorporation of titania

inside mesoporous matrix.

Adsorption isotherms are used as a macroscopic average measurement for exploring the

surface area, the pore diameter and volume of the samples. Table 4.2 summarizes the

measurements of the different MCM-48 samples with and without titania. In the BJH method

if the step for N2 at 77 K lies below p/p0 = 0.42, the analysis becomes fairly weak because

this is considered to be the stability limit of the meniscus. The thermodynamically based BJH

method over-estimate the relative pressure at desorption and therefore underestimate the

calculated pore diameters by ca. 1.0 nm. In NLDFT method,32 (non local density functional

theory, DFT) the adsorption and desorption isotherms in pores are calculated based on the

intermolecular potentials of fluid-fluid and solid-fluid interaction and this theoretical model

based calculation seems to give more accurate results. Based on the NLDFT method the

surface area of calcined MCM-48 was determined to 813 m²/g with 3.3 nm pore diameter and

0.77 cm³/g pore volume. This is in agreement with data published in the literature33 and

confirms the good quality of the MCM-48 sample. The results have been obtained in

collaboration with MPI für Kohlenforschung, Mülheim, Germany.

Table 4.2 Result of the N2 physisorption experiment

Sample Pore diameter

(Å)

Pore volume

(cm3/g)

Surface area

(m2/g)

MCM-48 33 0.77 813

TiO2-MCM-48 (2) 32 0.66 880

TiO2-MCM-48 (3) 28 0.45 618

Tetrabutylorthotitanate was used as Ti source

86

The successive loading of titania leads to a stepwise decrease of the pore volume (0.45 cm³/g)

and a decrease of the average pore diameter (28 Å). However, in addition to the mesopores

also micropores have been observed with ~ 10 Å pore diameter in the titania loaded samples.

This might indicate that the pore wall degrades and becomes more and more porous or that

the formation of titania particles inside the MCM-48 channels leads to micropores between

the silica wall and the titania, or between titania particles. The fact that the micropore

diameter decreases only little, supports the formation of particles, since the creation of pores

in the silica wall should lead to further widening of the micropore diameter at successive

loading as the corrosion process proceeds. Surprisingly, the surface area even after the second

impregnation is almost unchanged. Only after the third impregnation, a significant decrease

in surface area to 618 m²/g has been measured. Normalized to 1g silica MCM-48, this relates

to 760 m²/g. This finding has been confirmed by several measurements, also using the BJH

approach for the interpretation of the isotherms. The results of the surface measurements in

combination with the development of the pore volume are in agreement with the formation of

particles inside the pores which reduce the pore volume continuously, however, creating

additional surface area for low loadings, e.g. through surface roughening. Only after

considerable loading of titania, the coverage and blockage of surface area of the matrix

outweighs the contribution of the newly created surface by the titania particles. Similar

observations have already been reported by Wark et al.34 for titania in MCM-41, however

without detailed discussion.

UV-Vis spectra of Ti-containing mesoporous materials are usually characterized by

broad absorption bands appearing at 220 and 260-270 nm.35,36 The band at 220 nm could be

associated with isolated Ti in the framework, which is similar in character to that in TS-137

and Ti-Beta.13 This band has been assigned to the Ti atoms in octahedral coordination, in

which two water molecules form a part of the metal coordination sphere. The band at 270 nm

could be due to the presence of partially polymerized hexa-coordinated Ti species,37

87

however, Ti-O-Ti clusters can also exist along with the isolated Ti sites to some extent. The

UV-Vis spectra of calcined TiOx/MCM-48 compared with MCM-48 and one of the pure

rutile-TiO2 is given in Fig. 4.4.

abso

rban

ce(a

. u )

200 300 400 500 600 700 800

Fig. 4.4: UV-visible spectra of (b) Ti impregnated MCM-48 (titanylacetylacetonate, once loaded) with (a) pure MCM-48 and (c) TiO2

Wave length (nm)

a

b

c

In the Fig. 4.4 the absence of a broad absorption band at around 330 nm confirms the absence

of bulk TiO2 in TiOx/MCM-48 material which would be related to titania particles deposited

on the external surface. A clear shift of the absorption by ~ 20 nm in the TiOx/MCM-48

sample after first loading towards lower wavelength is observed. This is an indication for a

particle size effects on the absorption energy. For nanoparticles in the quantum dot range, the

frequency of the absorbed light shifts towards lower wavelength with the decrease of the

particle diameter. This can be proved considering the following equation. For a spherical

quantum dot, the energy expression can be derived as ∆ E = h2/8R2(1/me+1/mh). Where ∆ E =

energy change in band gap energy due to quantumsize effect, h = plank constant, R = radius

of the quantum dot, me = effective mass of electron and mh = effective mass of hole. E = hν =

88

h.c / λ also, where c = velocity of light and λ = wavelength of the light absorbed. From the

above two equations it is clear that absorption energy of the quantum dot will shift to higher

frequency or lower wavelength with decreasing diameter of the dots with a dependence of

1/R2. The pores of MCM-48 provide a size confinement for titania with a maximum particle

size of ~3 nm for which a blue shift in the absorption of titania is expected. This feature is

clearly reflected here. The UV spectra of successively loaded TiOx/MCM-48 samples are

shown in Fig. 4.5. The energy shifts towards higher wavelength with increasing metal content

indicates a particle size increase.

0 100 200 300 400 500 600 700 800 900Wave length (nm)

Ref

lect

ion

(%)

a

bc

Fig. 4.5: UV-visible spectra of 1st (a) , 2nd (b) and 3rd (c) loaded TiO2-MCM-48,tetrabutylorthotitanate was used as Ti source

Considering the energy shift of approx 1700 cm-1 (Fig. 4.4) and using the formalism

described by Kormann et al.38 A particle size of ~ 2 nm in diameter is calculated which is in

perfect agreement with the pore confinement by the MCM-48 matrix. The calculation is as

follows:

89

According to the Plank`s Quantum theory if an oscillator emitting a frequency ν can

only radiate in units or quanta of the magnitude hν, where h is the fundamental constant of

nature, it’s called Plank constant and the value is 6.626 × 10–34 J.sec, where J = Joul.

Einstein’s law of photochemistry says, if the light absorbed has a frequency (ν) per sec or

wave length λ in cm, then its energy is hν ergs.

E = hν = h.c / λ where c is the velocity of light and its value is 300,000 km/sec.

The value of ∆ λ, which is obtained from the blue shift of Ti MCM-48 ( λ1 = 330 nm approx )

than pure titania ( λ2 = 350 nm ).

∆ E = h [c / λ1 − c / λ2] h = 6.626 × 10–34 J.sec

= h × c [1/330 − 1/350] c = 300,000 km/sec, 1nm = 10-7cm

= h (J.ses) × c (km.sec-1) [20/115500] (nm-1)

= 6.626 × 10–34 J.sec × 3 × 1010cm sec-1× 1.73 ×10-4 × 107 cm-1

= 6.626 × 3 × 1.73 × 10-21 J 1 e.V = 1.60 × 10-19 J

= 34.38 × 10-21 J 1 e.V = 8065.5 cm-1

∆ E = 34.38 × 10-21/1.60× 10-19 e.V

= 21.48 × 10-2 e.V

= 8065.5 × 21.48 × 10-2 cm-1

= 1732.47 cm-1

According to Kormann et al38 the shift of 1209.75 cm –1 wave numbers correspond to 1.2 nm

particle size of TiO2. So the 1732.47 cm –1 wave number corresponds to formation of titania

particles of 2 nm of diameter approximately.

Based on a particle size of 2 nm, the crystallographic data of MCM-48 and the density of

calcined MCM-48 of ~ 0.97 g/cm³, the distribution of nanoparticles inside MCM-48 has been

calculated. Now if it is considered that titania particles are most likely situated in the

90

intersection of the MCM-48 silicate framework one can calculate the maximum occupancy of

the titania particle inside the unit cell. The calculation is as follows:

The number of intersections / unit cell for MCM-48 is 16.

The molecular weight of titania is 79.98 g/mol, the densities of Anatase, Brookite and Rutile

are 3.84, 4.17 and 4.26, respectively. The average density of titania can be taken as 4.0 g/cm3.

The volume of each titania particle would be 79.98 / 4 = 19.98 cm3/mol. The surface area of a

spherical particle of 2 nm diameter would be πr2 = 3.14 nm2 and the volume 4/3πr3 = 4.18

nm3 = 4.18 × 10–21 cm3. The mass of the each titania particle would be v.d = 4.18 × 10-21 × 4

= 1.67 × 10-20g (where v is the volume and d is the density of titania particle). Now if it is

considered that each of the 16 intersections per unit cell of MCM-48 is occupied by one

titania particle then there will be 16 titania particles and the total mass of the 16 titania

particles would be 16 × 1.67 × 10-20g. Weight fraction of Ti in TiO2 is 47.9 / 79.9 = 0.599

where the molecular weight of Ti is 47.9 g/mol. The total mass of Ti is 2.67 × 0.59 × 10-19 =

1.59 × 10-19g. Now for MCM-48 density is 0.97 g/cm3 and unit cell volume is 590 nm3. So

the mass of unit cell is 590 × 10-21 × 0.97 = 572.2 × 10-21 g. So the total mass of the MCM-48

filled with 16 titania particles in every intersections is 5.72 × 10-19 + 2.67 × 10-19 = 8.39 × 10-

19 g. From the total mass of 8.39 × 10-19 g Ti is occupies only 1.59 × 10-19 g and the

percentage would be only 18.95 %. This finding can make an assumption that in the unit cell

of MCM-48, if every intersection is filled with spherical TiO2 particles of 1nm radius it

should cover ~ 19 % of total mass available. Assuming that the deposition occurs

preferentially at the channel intersection for the three times loaded sample with 16 wt% Ti, ~

80% occupancy has been calculated. That means still there is free space for three dimensional

diffusion. Since the void space at the intersection is at least 3 nm in diameter as Terasaki39 has

shown experimentally from electron diffraction structure analysis for MCM-48, micropores

are created, still allowing for the three dimensional diffusion without pore blockage.

91

The local environment around Ti was studied through XAS spectra at the Ti K-edge

(4966 eV). In Fig. 4.6, the XANES spectra of Ti loaded samples are compared with those of

the reference oxides rutile and anatase. It can be seen that the scattering features above the

edge are much less distinct in the TiOx/MCM-48 materials than in the reference oxides,

which indicates a less ordered structure. Closer inspection shows that the sequence of

scattering features is completely different from that in anatase while it resembles with rutile.

This suggests that the short-range order of the TiOx entities is closer to rutile than to anatase.

4960 4980 5000 5020 E, eV

Nor

mal

ized

abs

orpt

ion

Fig. 4.6: XANES spectra of rutile (a), titanyacetylacetonate loaded MCM-48 (b) 1st (c) 2nd (d) 3rd (e) loaded TiO2-MCM-48 (Ti source tetrabutylorthotitanate) and (f) anatase

a

b

c

d

e

f

The discussion of Ti XAS and XANES spectra is often focused on the distinctive pre-

edge peak,40-43 the normalised peak height of which is ~0.25 in case of the TiOx/MCM-48

samples under investigation. The energy position of the pre-edge peak lies around 4970.3 eV

(Table 4.3). Farges et al.40 have conducted an extensive study of position and height of this

92

pre-edge peak using titanium-containing reference compounds. They constructed a profile of

normalised peak height versus its absolute position and defined three regions corresponding

to 4, 5 and 6-fold coordinated Ti, which have been summarised in Fig. 4.7. From this, the

intra-porous titania entities should contain Ti atoms in both 4 and 6 coordinated

environments, the latter being increasingly predominant with increasing Ti loading.

Table 4.3 Peak data for the pre-edge feature found in the XAS spectra of TiOx/MCM-48

Sample Corrected pre-edge

positiona (eV)

Shift relative to Ti foilb

(eV) Normalised pre-edge heightc

acac 4970.0 3.2 0.26

1st impregnation 4970.3 3.5 0.25

2nd impregnation 4970.3 3.5 0.24

3rd impregnation 4970.5 3.7 0.24

rutile 4971.6 4.8 0.17

rutile40 4971.6 not given 0.22

aCorrected with respect to the position of the pre-edge feature of the Ti reference foil and to the literature value for rutile, bshift relative to the position of the pre-edge feature in the Ti reference foil, cnormalised to the absorption edge height, acac is once titania loaded MCM-48 material (source titanylacetylacetonate), 1st impregnation, 2nd impregnation and 3rd impregnation are titania loaded materials (source tetrabutylorthotitanate)

93

Fig. 4.7: Relation between absolute position and the height of the Ti pre-edge feature for theTi containing standard compounds, areas are showing coordination of the central Ti atom;lines indicate mixture of compounds, where the arrows indicate a 1:1 ratio betweencompounds. Symbols represent like � 1st titania loaded MCM-48 ( acetylacetonate source),O 1st titania loaded (tetrabutylorthotitanate) ∆ 2nd loaded, 3rd loaded, and rutile

In Fig. 4.8, the TiK-EXAFS spectra of the Ti-loaded MCM-48 samples are compared with

those of the rutile and anatase reference oxides. The figure confirms that the titania entities

formed are highly disordered and most likely very small. The scattering features are much

less intense than those in the reference oxides, and no scattering features can be discerned

above 3.5 Å (uncorrected). The radial distributions of the scattered intensity resemble that of

rutile in particular after the second and third impregnation, where significant scattering

intensity is found between the features at 1.6 and 2.6 Å (uncorrected) unlike the spectrum of

anatase. This agrees with observations made in the XANES region. Closer inspection shows,

however that the feature near 3 Å (uncorrected) appears rather at the position expected for

anatase than for rutile (indicated by the broken line), in fact in some of the spectra this feature

appears to be split into two weak maxima.

94

Fig. 4.8: Fourier transform Ti-K edge EXAFS spectra of calcined TiOx-MCM-48, (a)1st loded TiO2-MCM-48 ( acetylacetonate source), (b) 1st, (c) 2nd and (d) 3rd loadedTiO2-MCM-48 (tetrabutylorthotitanate source)

a

b

c

d

This indicates that the short-range order is not identical to that of rutile or that more than one

type of coordination spheres is present.

In order to extract more detailed information about the coordination environment in the

samples, representative EXAFS spectra were fitted to structural models. Results of this fitting

procedure (bond lengths, coordination numbers, Debye-Waller factors and edge-shift

corrections) are reported in Tables 4.4 and 4.5. To validate the data reduction procedures

adopted and the FEFF parameters used, the EXAFS spectrum of rutile was fitted first. It

should be noted that a satisfactory fit of the rutile spectrum beyond the first oxygen sphere as

given in Fig. 4.9a and Table 4.4 was possible only when the k range used was cut at the rather

high value of ca. 5 Å-1 (actual range to 5.1 < k < 15.8 Å-1). This is obviously due to intense

multi-electron excitations near the Ti K edge,40 which distort the signal at low k values also

in the EXAFS region, and is therefore a general problem of EXAFS at the Ti K edge although

it does not seem to have received much attention in earlier model calculations.44-46

95

Table 4.4 Parameters of the coordination spheres of rutile (comparison of EXAFS model fit (Fig. 4.9a with crystallographic data)

No. Element R, Å

Model Crystallographic

C.Na 103σ2, Å-2 E0, eV

1 O 1.935 ± 0.013 1.948 4 7.3 ± 1.8 6.7

2 O 1.983 ± 0.020 1.980 2 4.6 ± 2.5 6.7

3 Ti 2.9783 ± 0.015 2.959 2 3.6 ± 1.5 4.5

4 Ti-Ob 3.48 ± 0.1 3.428 8 4 ± 15 4.5

5 O 3.521 ± 0.035 3.487 4 2.6 ± 3.7 6.2

afixed to crystallographic values, bmultiple scattering path, R = Bond distance d (Ti-x), C.N. = Coordination no., σ2 = Debye-Waller factors, E0 = Binding energy

Table 4.5 Model parameters for the best EXAFS fits for TiOx/MCM-48 (first and third impregnation, from Figure 4.9b and c)

No. Element R, Å C. N. 103 σ2, Å-2 E0, eV

First impregnation

1 O 4.1 ± 0.2 8.7 ± 0.7 7.0

2 O 1.0 ± 0.2 4.5 ± 1.7 6.4

3 Ti

1.876 ± 0.005

2.052 ± 0.014

3.110 ± 0.015 3.4 ± 0.6 17.5 ± 2.4 14.2

3rd impregnation

1 O 4.1 ± 0.2 11.0 ± 0.6 5.0

2 O 1.2 ± 0.3 14.2 ± 3.3 5.0

3 Ti

1.880 ± 0.004

2.006 ± 0.021

3.051 ± 0.008 3.5 ± 0.4 14.7 ± 1.2 6.5

R = Bond distance d (Ti-x), C.N. = Coordination no., σ2 = Debye-Waller factors, E0 = Binding energy

The ab initio XAFS code FEFF considers multi-electron excitations as an integral amplitude-

damping effect, without particular energy dependence, which in some cases leads to

96

inadequate EXAFS amplitude at small k values. Considering the limitation of the k range

mentioned above the EXAFS spectrum of rutile correlate with the rutile crystal structure. The

spectra of the TiOx/MCM-48 sample attain a shape that permitted an interpretation consistent

with the observations in the XANES region.

The spectra of the TiOx-MCM-48 samples could be modelled with two oxygen and one

titanium shell as in the rutile structure but at different distances (Table 4.5, Fig. 4.9 b, c). For

the sample after the third impregnation, alternative models were tested, e.g. with a third

oxygen shell instead of a titanium one, but they reproduced the spectrum less accurately. This

supports that the TiOx entities have, indeed, formed clustered structures (though disordered

as shown by the sometimes very high Debye-Waller factors) instead of two-dimensional

coating as found after introduction of ZnOx into MCM-48.21

Fig. 4.9: Fourier spectra for TiO2-MCM-48 sample in k2 space (left) and in r-space(right) top: rutile middle: once impregnated and bottom : 3rd impregnated

97

Despite some differences in the shape of the Fourier-transformed spectrum (Fig. 4.9), the

models for the samples after the first and the third impregnation are rather similar. When

more titania is deposited, some shift in the shell distances which tend towards the values of

rutile can be seen. Analogously, the sum of coordination numbers of the two oxygen shells

increases slightly. It has to be admitted that this increase is hardly significant with the given

error limits of the method, however it agrees with a similar conclusions drawn from the

XANES region (Fig. 4.6). The above data and calculation have been obtained in collaboration

with the Technical Chemistry Department, Ruhr University Bochum.

4.4 Conclusion

A simple and effective wet impregnation method was applied for the incorporation of

Ti in siliceous MCM-48, which yielded highly ordered mesoporous material with high

surface area. The decrease in intensity, reflected in the X-ray diffraction diagram, with

successive loading by metal oxide directly proves the uptake of titania inside the pore system.

The regular contrast variation in TEM image showed the intactness of MCM-48 silicate

framework and EDX spectra indicated Ti incorporation inside MCM-48. There were no

differences in the TEM images of Ti-MCM-48 materials prepared by titanylacetylacetonate

or tetrabutylorthotitanate. Pore volume and pore diameter decreased with titanium content,

however, surface area increased for low loading of titania due to additional surface of the

nanoparticles deposited inside the pores. The blue shift of the absorption of TiOx in UV

visible spectra indicated the formation of quantum-sized titania particles. The transition shift

towards higher wavelength from 1st to 3rd impregnation was one of the main evidences of

the increase of cluster size with successive loading by metal oxide. XANES spectra yielded 5

to 6 fold coordination for Ti in TiOx/MCM-48, and shows resemblance of the spectra with the

one of rutile. From EXAFS it could be concluded that the nanoparticles are highly disordered

after the first loading, however, successive uptake of titania lead to a local order with more

than one type of coordination sphere including a resemblance to rutile.

98

4.5 References

1. A. Sayari, Chem. Com. (1996) 1840.

2. A. Corma, Chem. Rev. 97 (1997) 2373.

3. R. Köhn, M. Fröba, Catal. Today 68 (2001) 227.

4. M. Morey, A. Davidson, G.D. Stucky, Microporous Mater. 6 (1996) 99.

5. K.A. Koyano, T. Tatsumi, Chem. Commun. (1996) 145.

6. A. Corma, Q. Kan, F. Rey, Chem. Commun. (1998) 579.

7. J.V. Walker, M. Morey, H. Carlsson, A. Davidson, G.D. Stucky, A. Butler, J. Am. Chem.

Soc. 119 (1997) 6921.

8. M. Morey, A. Davidson, H. Eckert, G.D. Stucky, Chem. Mater. 8 (1996) 486.

9. S. Kawi, M. Te, Catal. Today 44 (1998) 101.

10. H. Gies, S. Grabowski, M. Bandyopadhyay, W. Grünert, O.P. Tkachenko, K.V.

Klementiev, A. Birkner, Microporous Mesoporous Mater. 60 (2003) 31.

11. J.S Reddy, R. Kumar, P. Ratnasamy, Appl. Catal. (1990) 58.

12. A. Thangaraj, R. Kumar, S.P. Mirajkar, P. Ratnaswamy, J. Catal. 130 (1991) 1.

13. M.A. Camblor, A. Corma, A. Martinez, J. Chem. Soc. Chem. Commun. (1992) 589.

14. A. Corma, M.T. Navarro, J. Perez-Pariente, J. Chem. Soc. Chem. Commun. (1994) 147.

15. Y.W. Chen, H.Y. Lin, J. Porous Mater. 9 (2002) 175.

16. W.S. Ahn, D.H. Lee, T.J. Kim, G. Seo, R. Ryoo, Appl. Catal. A: General 181 (1999) 39.

17. T. Maschmeyer, F. Rey, G. Sankar, J.M. Thomas, Nature 378 (1995) 159.

18. K. Schrijnemakers, E.F. Vansant, J. Porous Mater. 8 (2001) 83.

19. D. Trong On, D. Desplantier-Giscard, C. Danumah, S. Kaliaguine, Appl. Catal. A 222

(2001) 299.

20. H. Wellmann, J. Rathousky, M. Wark, A. Zukal, G. Schulz-Ekloff, Microporous

Mesoporous. Mater. 44 (2001) 419.

99

21. O.P. Tkachenko, K.V. Klementiev, E. Löffler, I. Ritzkopf, F. Schüth, M. Bandyopadhyay,

S. Grabowski, H. Gies, V. Hagen, M. Muhler, L. Lianhai, R.A. Fischer, W. Grünert,

Phys. Chem. Chem. Phys. 5 (2003) 4325.

22. M. Wark, Y. Rohlfing, Y. Altindag, H. Wellmann, Phys. Chem. Chem. Phys. 5 (2003)

5188.

23. B.S. Uphade, T. Akita, T. Nakamura, M. Haruta, J. Catal. 209 (2002) 331.

24. C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710.

25. Q. Huo, R. Leon, P.M. Petrov, G.D. Stucky, Science 268 (1995) 1324.

26. Q.S. Huo, D.I. Margolese, G.D. Stucky, Chem. Mater. 8 (1996) 1147.

27. G.W. Noble, P.A. Wright, P. Lightfoot, R.E. Morris, K.J. Hudson, A. Kvick, H.

Graafsma, Angew. Chem. Int. Ed. Engl. 36 (1997) 81.

28. R. Garcia, E.F. Philp, A.M.Z. Slawin, P.A. Wright, J. Mater. Chem. 11 (2001) 1421.

29. P.T. Tanev, T.J. Pinnavaia, Science 279 (1995) 548.

30. S. Inagaki, Y. Fukushima, K. Kuroda, in: J. Weitkamp, H.G. Karge, H. Pfeifer, W.

Hölderich (Eds.), Zeolites and Related Micropor. Materials: State of the Art 1994, Studies

in Surface Science and Catalysis, Elsevier, amsterdam, 84, Part A, 1994, pp. 125.

31. B. Marler, U. Oberhagemann, S. Vortmann, H. Gies, Microporous Mater. 6 (1996) 375.

32. P.I. Ravikovitch, S.C. O’Domhnaill, A.V. Neimark, F. Schüth, K.K. Unger, Langmuir, 11

(1995) 4765.

33. P.I. Ravikovitch, D. Wei, W.T. Chueh, G.L. Haller, A.V. Neimark, J. Phys. Chem. B 101

(1997) 3671.

34. P. Prochnow, G. Schulz-Ekloff, M. Wark, J.K. Thomas, A. Zukal, J. Rathousky, in: A.

Galareau, F. di Renzo, F. Fajula, J. Vedrine, (Eds.), Proceedings of the 13th International

Zeolite Conference, Studies in Surface Science and Catalysis Vol.135, Elsevier,

Montpellier, France, 2001, pp. 359.

35. W.S. Ahn, N.K. Kim, S.Y. Jeong, Catal. Today 68 (2001) 83.

100

36. W. Zhang, M. Fröba, J. Wang, P.T. Tanev, J. Wong, T.J. Pinnavaia, J. Am. Chem. Soc.

118 (1996) 9164.

37. G. Petrini, A. Cesana, G.De Alberti, F. Genoni, G. Leofanti, M. Padovan, G. Paparatto, P.

Rofia, Stud. Surf. Sci. Catal. 68 (1991) 761.

38. C. Kormann, D.W. Bahnmann, M.R. Hoffmann, J. Phys. Chem. 92 (1988) 5196.

39. M. Kaneda, T. Tsubakiyama, A. Carlsson, Y. Sakamoto, T. Ohsuna, O. Terasaki, J. Phys.

Chem. B 106 (2002) 1256.

40. F. Farges, G.E. Brown, J.J. Rehr, Phys. Rev. B 56 (1997) 1809.

41. R.J. Davis, Z. Liu, J.E. Tabora, W.S. Wieland, Catal. Lett. 34 (1995) 101.

42. G. Sankar, J.M. Thomas, C. Richard, A. Catlow, Top. Catal. 10 (2000) 255.

43. H. Yoshitake, T. Sugihara, T. Tatsumi, Phys. Chem. Chem. Phys. 5 (2003) 767.

44. G. Grubert, M. Stockenhuber, O. P. Tkachenko, M. Wark, Chem. Mater. 16 (2002) 2458.

45. D. Gleeson, S. Sankar, C.R.A. Catlow, J.M. Thomas, G. Spanó, S. Bordiga, A. Zecchina,

C. Lamberti, Phys. Chem. Chem. Phys. 2 (2000) 4812.

46. A. Zecchina, S. Bordiga, G. Spoto, A. Damin, G. Berlier, F. Bonino, C. Prestipino, C.

Lambertim, Top. Catal. 21 (2002) 67.

101

5. Deposition of Au nanoparticles inside TiO2-MCM-48 and the catalytic activity on CO

oxidation reaction

5.1 Introduction

Before 1980s, very little attention has been paid to gold metal as catalyst, which was

considered to be inert owing to the electronic configuration of noble metal and this

corresponds to usually very low activities.1 However, after the pioneering work by Haruta et

al.,2-4 gold as catalyst attracted extensive attention. Gold can exhibit considerably catalytic

activity when it is highly dispersed on metal oxides supports.5-9 It is really important that the

small gold particles should be highly dispersed on the oxide support.7 Au/oxides catalysts

have been widely applied to many important chemical reactions such as CO oxidation2,

hydrogenation of unsaturated hydrocarbon10, Reduction of NOX11, epoxidation of C3H6

12,

selective CO oxidation in a hydrogen rich steam13, combustion of methane14, etc. Recently

Hua et al.15 have used Au/iron oxide catalyst for water gas shift reaction.

The catalytic performances of different gold-supported systems are influenced by the

preparation method, the synthesis parameters, Au particle size, and details.5,16,9 The

deposition-precipitation (DP) process is considered to be the best method to synthesize highly

active gold particles17 because of the control of the particle size of gold by adjustment of pH

during catalyst preparation. In addition, some other methods like chemical vapor deposition18

(CVD) or co-sputtering19 are also widely used for catalyst formation. The extensive study on

Au/Al2O3 catalyst prepared by both DP and CVD methods reveal that finely dispersed

Au/Al2O3 with Au particle size < 5nm can be easily prepared by CVD method, whereas

traditional DP method leads to the formation of larger gold (> 7nm) particles on Al2O3

support.20 A series of active Au/TiO2, Au/Co3O4, Au/ZrO2 catalysts were synthesized by

Yang. et. al.9 and it is clear that with similar gold particle sizes, Au/TiO2 is more active than

Au/Al2O3. Many researches have then focused on Au/titania system.12,21,22 There is one report

on supported gold catalyst derived from interaction of a Au-phosphine complex with

102

conventional titanium oxide and as-precipitated titanium hydroxide.23 A novel method of

preparation of a nanosized gold catalyst supported TiO2 by the thermal relaxation technique

has also been reported.24 In addition, very active gold catalyst supported on TiO2 have been

formed using nanosized gold particle/nylon-11 oligomer composites.25

The work on the deposition of Ti salts within the pore system of cubic MCM-48 and

their subsequent decomposition to Ti-oxides clusters26 has been discussed in the previous

chapter (Chapter 4). The use of MCM-48 with quantumsized titania particles inside the

porous matrix, as a support for Au particles is the main interest of the present study.

Quantumsized titania particles are formed by post-synthetic wet impregnation process. The

Deposition-precipitation technique was used for the deposition of gold particles. The final

product underwent thorough characterization by means of X-ray powder diffraction, EXAFS,

XANES-spectroscopy along with TEM. The reactivity of Au/TiO2-MCM-48 catalyst was

verified using the CO oxidation reaction.

This chapter deals with

♦ Deposition of Au nanoparticles inside TiO2-MCM-48

♦ Characterization of the final product

♦ Catalytic application of Au/TiO2-MCM-48 with the CO oxidation reaction

♦ Post catalytic characterization of the material 5.2 Experimental

5.2.1 Synthesis of Au/TiO2-MCM-48

Ti-MCM-48 was prepared by post-synthetic wet impregnation method and the

synthesis procedure was described in the previous chapter (chapter 4) in detail. Three times

Ti loaded MCM-48 (tetrabutylorthotitanate, Merck) was used for Au loading. Nanoparticles

of Au were deposited following deposition-precipitation method, described by Haruta et. al.21

During catalyst preparation 100 ml of an aqueous solution of HAuCl4. 3H2O (Aldrich, 8 wt %

Au with respect to support) was heated at 70 °C. The initial pH was around 2.5. The pH was

103

adjusted to neutral value 7 by drop-wise addition of dilute NaOH. After that 0.1 g of Ti-

MCM-48 was dispersed in the solution. The pH of the solution dropped to 5~6 and the pH

was then readjusted to 7 again by addition of NaOH. The suspension was stirred for 1h at

same temperature. Finally the solid was filtered with water, freeze-dried overnight in

vacuum. The whole procedure was carried out in the absence of light.

The wet impregnation method was also used for the Au loading. During the synthesis

procedure aqueous hydrogentetrachloroauratetrihydrate (HAuCl4. 3H2O) solution in water

(0.002 M) was prepared. After that, dried MCM-48 was added to the solution with stirring.

Finally the solid was filtered off, washed with water and dried at room temperature for

overnight. However, for comparison some samples were prepared in presence of light.

5.2.2 Catalytic activity measurements

For the CO oxidation reaction a plug-flow reactor with inner diameter of 4.5 mm was

used. The temperature ramp was 1 °C/min. The reaction gas contained 1 % CO, 20 % O2 and

rest N2. During the reaction 1 % CO was passed over 50mg of the catalyst (Au/TiO2-MCM-

48) with a flow rate of 67 ml min-1.

5.3 Result and discussion

5.3.1 Synthesis

The wide angle XRD pattern is given in Fig. 5.1. There are clear signals of gold at

around 38.3 °, 64.7 °, 77.96 ° 2θ value. According to the Scherrer equation, considering the

full width of half maxima (FWHM) of a particular 2θ value, the approximate particle size can

be calculated and the average particle size of gold comes as 28 Å. The Scherrer equation is

written as t = kλ / βcosθ, where t = crystal size in Å, λ = (wave length of Cu-Kα radiation)

1.54059 Å, β = breadth of the beam at full width of half maxima (FWHM, in radian), k =

Scherrer costant, depending upon particle shape k is in range between 0.5 to 1.1.

104

a

b

Fig. 5.1: Wide angle powder XRD patterns of (a) three times Ti impregnated and (b) Au/TiO2 loaded MCM-48

When 2θ = 64.70 °, the full width of half maxima was 3.352 ° and this value should be

changed in radian value and the value of β becomes 0.058. The value of k is 0.9 when the

particle is considered as spherical, and cosθ = cos 64.70 /2 = 0.841. Putting all the values in

the above equation

t = 0.9 × 1.54059 /0.058 × 0.841 Å

= 1.386 / 0.049 Å

= 28.29 Å. Using same equation but 2θ value as 77.96 ° the FWHM comes as 3.7344 ° or

0.065 radian and cosθ = 0.77, then the crystal size appears as 27. 72 Å.

The particle size is small enough to fit inside the porous matrix of Ti-MCM-48 as discussed

in the previous chapter (chapter 4)26 that the pore diameter of Ti-MCM-48 after 3rd loading is

28-29 Å. To confirm the interpretation of the XRD experiments, complementary TEM studies

were carried out. The integrity of every sample was checked with TEM together with its

qualitative titania and gold content using EDX. The results of the analyses always showed the

105

well ordered and well maintained silica host structure. The TEM image (Fig. 5.2) shows

uniform and highly dispersed gold particles inside the channel. Low angle powder XRD

patterns of Au loaded TiO2 (prepared by DP method) show that the parent structure is

maintained after successive loading by titania then with Au, and co-relates with the

observation revealed from TEM images.

Fig. 5.2: The TEM image of an area of Au/TiO2- MCM-48

The EDX spectra is given in Fig. 5.2a. In the EDX spectra there was clear signal of AuMα

and AuMβ around 2.1, which indicated the incorporation of Au nanoparticles inside porous

matrix. The sample was mounted on Cu grid and for that reason Cu signal was obtained in the

EDX spectra. However, the samples prepared by wet impregnation method did not show

nanosized Au particle inside the channels (Fig. 5.3).

106

0,0

1,5

3,0

4,5

0 2 4 6 8 1

energy / keV

cps

C

0

Kα

OK

α

CuL

α

SiK α

TiK α

TiK β

CuK

α

AuL

α

CuK

β

AuM

αA

uMβ

Fig. 5.2a : The TEM image of an area of Au/TiO2- MCM-48 along with the EDX Analysis

Fig. 5.3: The TEM image of an area of Au/TiO2- MCM-48 prepared by wet impregation method

107

The wet impregnation method could not be used as the proper method to prepare small Au

particles inside porous matrix. From the TEM image it is clear that Au particles are big and

were not inside the silicate framework. The local environment of gold was studied by XAS

spectroscopy at the Au LIII-edge. In Fig. 5.4 (left-hand side) the XANES and EXAFS

(measured at liquid nitrogen temperature, right-hand side) of Au/TiO2-MCM-48 ( not

exposed to light) are compared to gold metal foil.

0 2 4 6 811880 11920 11960 12000

Au-Au

1.0

FT (χ

(k)·k

2 )

r (Å)

Au foil Au/TiO2-MCM-48

0.5

Nor

mal

ised

abs

orpt

ion

Photon energy (eV)

Fig. 5.4: XANES (left) and EXAFS (right) spectra of Au foil and Au/TiO2-MCM-48

The gold content in the Au/TiO2-MCM-48 sample is estimated to be ~2.4 wt % from the edge

step height. Due to this low metal content, the spectra obtained were a touch noisy. The shape

of the Au/TiO2-MCM-48 XANES closely resembles that of the gold metal foil. From this it is

concluded that the gold in the MCM-48 sample is in the metallic state, as the ionic form of

gold gives rise to a distinctly different XANES structure. Further, because the features in the

XANES of the reference and the sample are comparably pronounced, it would seem that the

gold particles in MCM-48 possess an ordered structure. In the EXAFS it can be seen that the

108

Au-Au contribution of Au/TiO2-MCM-48 (Fig. 5.4, right-hand side) is distinctly smaller

when compared to the foil, which can be seen as indicative for a small particle size. The

splitting of the Au-Au contribution that is found in the gold metal foil can also be seen in the

EXAFS of Au/TiO2-MCM-48. This splitting is prone to occur with heavier elements like

gold. From XAS studies it is clear that well ordered small metallic gold particles were found

in the MCM-48 material. Fig. 5.5 shows the XANES and EXAFS spectra of two different

samples, one was exposed to light (a) and the other was not (b) (same sample in Fig. 5.4).

0 2 4 6 80

1

2

3

4

5

6

Au-Au

Au foil

Au/TiO2-New

Au/TiO2-Old (double)

Au/TiO2-MCM-48

k2 ·F

T (χ

)

r (Å)

a

b

11880 11920 11960 120000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Au foil

Au/TiO2-New

Au/TiO2-Old (double)

Nor

mal

ised

abs

orpt

ion

(-)

Photon energy (eV)

Au/TiO2-MCM-48

Fig. 5.5: XANES (left) and EXAFS (right) spectra of Au foil compared with (a) Au/TiO2-MCM-48 which was exposed to light and (b) Au/TiO2-MCM-48 which was not exposed to light.

a

b

a

b

There are clearly some differences between two samples, which have been reflected in the

spectra. In the sample a, which was exposed to light, the first oscillation is much more

pronounced than the b sample, indicating that the oxidation state of Au is 3+, whereas, the b

sample is quite similar to the Au reference foil that is metallic gold. That means that in the

presence of light Au metal oxidizes to Au+3 ion. XANES is very sensitive to oxidation state.

There is no pre-edge peak observed.

109

The parent TiO2-MCM-48 had a BET surface area of 671 m2 g-1and a pore volume of 0.85

cm3 g-1. The gold loaded sample showed a decreased surface area of 400 m2 g-1 and a pore

volume of 0.5 cm3 g-1. The decrease of surface area and pore volume indicated the presence

of oxide particles inside porous matrix, which was further proved by TEM and EXAFS

measurements.

5.3.2 Catalytic evaluation on CO oxidation reaction

The results of the activity measurements of different samples in CO oxidation are

shown in Fig. 5.6, 5.7, 5.8 and 5.9. In Fig. 5.6 the results of the catalytic test of the sample,

which had big Au particles outside the channel is shown. The sample showed very poor

activity. Above 300 °C 40 % conversion was measured. The Au/TiO2-MCM-48, which was

prepared by DP method, but exposed to light during catalyst synthesis showed better result

(Fig. 5.7). This catalyst showed 50 % conversion of CO2 at 200 °C. This sample was better

than the previous one but still not good enough for CO oxidation reaction. Further

improvement was noticed when the catalyst was synthesized in absence of light and CO

oxidation reaction was carried out with this. From Fig. 5.8 it is clear that there is an

improvement on the catalyst behavior. The catalyst showed activity before 100 °C but after

that the activity decreased due to some unknown reason. 50 % CO2 yield was observed at 180

°C. For further improvement of the catalyst Au loading was performed twice then it was used

for CO oxidation reaction (Fig. 5.9). After one run the same used catalyst was again used for

the reaction. 50 % CO2 conversion was obtained at 175 °C but in the 2nd run there is further

improvement in the catalytic behavior. A CO2 yield of 50 % was observed at 140 °C. From

all of the experiments it is clear that Au/TiO2-MCM-48 is catalytically active and showed

gradual improvement in the oxidation reaction after step-by-step modification.

110

50 100 150 200 250 300

0,0

0,2

0,4

0,6

0,8

1,0

CO CO2

conv

eris

on

temperature / °C

Fig. 5.6: Catalytic activity of Au/TiO2-MCM-48 (prepared by wet impregnation method), conversion plotted against temperature

50 100 150 200 250

0,0

0,2

0,4

0,6

0,8

1,0

CO CO2

conv

ersi

on

tempterature / °C

Fig. 5.7: Catalytic activity of Au/TiO2-MCM-48 (prepared by DP method, but the sample was exposed to light), conversion plotted against temperature

111

50 100 150 200 250 300

0,0

0,2

0,4

0,6

0,8

1,0

CO CO2

Konz

entra

tion

Temperatur / °C

Fig. 5.8: Catalytic activity of Au/TiO2-MCM-48 (prepared by DP method, but the sample was not exposed to light), conversion plotted against temperature

0 50 100 150 200 250

0,0

0,2

0,4

0,6

0,8

1,0

1,2

2 1

CO2

CO

conv

ersi

on %

temp.oC

First cycle: T1/2 = 180oC

Second cycle: T1/2 = 150oC

Fig. 5.9: Catalytic activity of Au/TiO2-MCM-48 (two times Au loaded by DP method and not exposed to light), conversion plotted against temperature

112

The TEM image of the Au/TiO2-MCM-48 after CO oxidation reaction is shown in Fig. 5.10

and the image is compared with the image taken before the catalysis.

Fig. 5.10: The TEM image of an area of Au/TiO2- MCM-48 (a) before and (b) after CO oxidation reaction

(a) (b)

From the images it is clear that the particles look similar before and after catalysis and there

are no aggregates of Au particles. No big aggregate of particles out side the framework was

also observed. That means after catalysis also the particles are inside and still show catalytic

activity towards oxidation reaction. The development of the synthesis procedure and the

proper after treatment for making the material catalytically active was a great achievement in

this present work. The catalytic reactions were carried out in collaboration with MPI für

Kohlenforschung, Mülheim, Germany.

5.4 Conclusion

TiO2-MCM-48 was used as the support for the deposition of Au particles. The post-

synthetic wet impregnation method did not work out for the deposition of Au nanoparticles

inside mesoporous material. Very big Au particles, which were situated outside the silica

113

framework were formed during this preparation method. Using DP method for the synthesis,

Au/TiO2-MCM-48 was obtained with nanosized Au particles inside the channel of MCM-48.

From the powder XRD pattern the average size of the Au particles was calculated as 28 Å,

which was small enough to fit inside the titania loaded MCM-48. TEM experiments showed a

very ordered and periodic silica framework structure, and nanosized Au particles were

noticed in the images. There was a clear Au signal in the EDX spectra. The shape of the

Au/TiO2 loaded MCM-48 XANES spectra resembled with the metallic Au foil and this

indicated the formation of metallic Au particles inside MCM-48. However, the sample, which

was exposed to light showed the presence of Au+3 ion. This material also showed

considerable catalytic activity in CO oxidation reaction. After the development of the

synthesis procedure (from wet impregnation to DP method) and careful after treatment, for

example during the synthesis and after that there was no light exposure, there was a

noticeable improvement in the catalytic activity. First time when the material with big Au

particle was used for catalysis, the material showed 50 % CO2 conversion at 300 °C, however

at the end, during the 2nd run (after first cycle with the same sample), the material with small

Au particles and prepared in absence of light 50 % CO2 conversion at 140 °C. This step by

step improvement of the synthesis procedure and its reflection on the catalysis was one of the

main achievements of the present study.

114

5.5 References

1. B. Hammer, J.K. Norskov, Nature 376 (1995) 238.

2. M. Haruta, N. Yamada, T. Kobayashi, S. Iijima, J. Catal. 115 (1989) 301.

3. M. Haruta, N. Yamada, T. Kobayashi, H. Kageyama, M. Genet, B. Delmon, J. Catal. 144

(1993) 175.

4. M. Haruta, T. Kobayashi, H, Sano, N. Yamada, Chem. Lett. (1987) 405.

5. M. Haruta, Catal. Today 36 (1997) 153.

6. M. Valden, X. Lai, D.W. Goodman, Science 281 (1998) 1637.

7. W. Vogel, D.A.H. Cunningham, K. Tanaka, M. Haruta, Catal. Lett. 40 (1996) 175.

8. M.M. Schubert, S. Hackenberg, A.C. van Veen, M. Muhler, V. Plzak, R.J. Behm, J.

Catal. 197 (2001) 113.

9. A. Wolf, F. Schüth, Appl. Catal. A: general 226 (2002) 1.

10. F. Cosandey, T.E. Madey, Surf. Rev. Lett. 8 (2001) 73.

11. E. Seker, E. Gulari, R.H. Hammerle, C. Lambert, J. Leerat, S. Osuwan, Appl. Catal. A:

General 226 (2002) 183.

12. B.S. Uphade, Y. Yamada, T. Akita, T. Nakamura, M. Haruta, Appl. Catal. A: General

215 (2001) 137.

13. G.K. Bethke, H.H. Kung, Appl. Catal. A: General 194 (2000) 43.

14. R.J.H. Grisel, D.T. Kooyaman, B.E. Nieuwenhuys, J. Catal. 191 (2000) 430.

15. J. Hua, K. wie, Q. Zheng, X. Lin, Appl. Catal. A: General 259 (2004) 121.

16. M.M. Schubert, V. Plzak, J. Garche, R.J. Behm, Catal. Lett. 76 (2001) 143.

17. S. Tsubota, D.A.H. Cunningham, Y. Bando, M. Haruta, Stud. Surf. Sci. Catal. 91 (1995)

227.

18. M. Okumura, K. Tanaka, A. Ueda, M. Haruta, Solid State Ionics 95 (1997) 143.

19. T. Kobayashi, M. Haruta, S. Tsuota, H. Sano, Sens. Actuators B1 (1990) 222.

20. Y.J. Chen, C.T. Yeh, J. Catal. 200 (2001) 59.

115

21. A.K. Sinha, S. Seelan, T. Akita, S. Tsubota, M. Haruta, Appl. Catal. A: General 240

(2003) 243.

22. Y. Iizuka, H. Fujiki, N. Yamauchi, T. Chijiiwa, S. Arai, S. Tsubota, M. Haruta, Catal.

Today 36 (1997) 115.

23. Y.Z. Yuan, K. Asakura, A.P. Kozlova, H.L. Wan, K.R. Tsai, Y. Iwasawa, Catal. Today

44 (1998) 333.

24. K. Sayo, S. Deki, S. Hayashi, J. Colloid Inter. Sci. 212 (1999) 2 597.

25. K. Sayo, S. Deki, S. Hayashi, J. Mater. Chem. 9 (1999) 4 937.

26. M. Bandyopadhyay, A. Birkner, M W.E. van den Berg, K. V. Klementiev, W. Schmidt,

W. Grünert, H. Gies, Chem. Mater. (submitted).

116

6. Summary

After the discovery of mesoporous silica materials by Mobil research group, this field

has attracted so much interest that it has grown into a separate area of research. For the M41S

family of materials, the composition is restricted to silicate frameworks and there are three

main mesoporous silicate host structures, MCM-41 (with hexagonal honeycomb symmetry),

MCM-48 (with cubic Ia3d symmetry) and MCM-50 (a lamellar phase). Due to this elegant

three-dimensional pore system, MCM-48 received more interest concerning both its structure

elucidation and potential applications in catalysis and separation science. Three-dimensional

MCM-48 is much more desirable than one-dimensional channel of MCM-41 from diffusional

and catalytic point of view.

The present thesis is a thorough investigation of advances made with the cubic

mesoporous MCM-48 in its synthesis, modification and some application. A reliable

hydrothermal synthesis procedure was developed and MCM-48 was successfully synthesized

from the TEOS : 0.70 CTACl : 0.5 NaOH : 64 H2O system. The resulting material

resembled with the material reported in literature. The material was highly crystalline,

reproducible with high surface area and pore volume. Thorough investigations were carried

out by changing different parameters, for example surfactant, gel composition, temperature

etc. The material was characterized by different physico-chemical methods, such as powder

XRD, N2 physisorption, NMR, TEM, EXAFS and so on.

Finding faster and economical synthesis route for inorganic and organic materials has

always been a challenging task, and the use of microwave heating for this purpose is now a

subject of growing interest. The microwave heating technique was also used for synthesizing

MCM-48. The MCM-48 material prepared by the microwave heating method was compared

with the reference material prepared by conventional hydrothermal heating technique. The

TEM, sorption properties looked similar and the shape of the material particles also

resembled each other. Overall, microwave synthesis reduced the time of synthesis of standard

117

preparations by almost 2 orders of magnitude making the material readily available in bulk

quantities. This opened a new range of applications for their direct and indirect use.

The pore size of mesoporous materials is somewhat flexible and could be engineered to

a desired size at the time of synthesis using organic additives, mainly long chain amines of

various sizes. This flexibility of pore size is interesting for the preparation of catalysts with

applications in the synthesis of organic molecules. MCM-48 with pore diameter ranging

between 17-26 Å was successfully synthesized by using surfactant with variation of the alkyl

chain length. For this purpose C12 to C18 (from dodecyl to octadecyl) ammonium surfactant

was used.

The crystallization kinetics was also investigated and it was found that at the very

beginning MCM-41 was formed and slowly converted to MCM-48. With time the intensity of

the peaks including the 2nd peak (022), which appeared only after 24 hr and is considered to

be the indication for MCM-48 framework structure, increased and finally the crystallinity

reached its maximum within about 4 days. From this experiment the formation path of MCM-

48 was explored and the detailed synthesis procedure was described in Chapter 2.

The post-synthetic wet-impregnation/calcination method was applied for the loading of

metal and metal oxides within the mesoporous MCM-48. Highly ordered MCM-48 material

with very high inner surface area was used for accommodating nanosized metal and metal

oxides inside the pore.

The deposition of metal salts, their subsequent decomposition within the pore system of

the cubic MCM-48 material to the corresponding oxide, e.g. Cu-oxide, Zn-oxide and Cu/ZnO

oxides was performed and the resulting materials were thoroughly characterized. Dip-

impregnation of calcined MCM-48 with aqueous solutions of Cu-acetates and subsequent

calcination led to the incorporation of metal oxides inside the mesopore system of MCM-48.

The preservation of the mesoporous host structure remained intact even after calcination also

as supported by XRD, TEM, NMR, N2-physisorption, XPS, and EXAFS analyses. The above

118

experiments showed that the oxides were inside, however, with different structural properties.

EXAFS investigations allowed the determination of the local structure of the nanoparticles

formed within the mesoporous host. Considering Cu K-edge it was found that beyond the

first shell, scattering events of Cu species in the siliceous matrix were at the same distances

as in CuO, but exhibit a strong amplitude decay, indicated the formation of very small

particles. However, the Zn K-edge spectra of ZnO-MCM-48 was not as simple like Cu K-

edge and there was almost no order beyond the first neighbor. EXAFS proved that zinc did

not form aggregates in the silica framework. Cu-oxide was formed to be nanodispersed in

organized particles whereas ZnO most likely coated the surface of the silicate channel wall.

Wet impregnation process did not lead to the formation of ZnO particles inside porous

matrix, however, ZnO-MCM-48 prepared using the organometallic route led to the formation

of ZnO particles, reveled from TEM images and EXAFS spectra; hence this might be a

promising approach towards the preparation of active catalyst for methanol synthesis. The

thorough explanation had been given in chapter 3.

The deposition of Ti salts within the pore system of cubic MCM-48 and investigation

of their structure, properties, location, and coordination state of Ti in TiOx was also one of the

main interests of the present thesis. The regular contrast variation in TEM image showed the

intact MCM-48 silicate framework after impregnation and from the Ti signal in EDX spectra

Ti incorporation inside MCM-48 was confirmed. After careful investigation of UV spectra

and EXAFS measurement it could be concluded that quantum sized titania particles of ~2 nm

diameter were formed and were highly dispersed inside the porous system. Successive uptake

of titania leads to a local order with more than one type of coordination sphere (5 to 6

coordination), leading to a resemblance with rutile structure.

The final approach of the present study was to introduce Au nanoparticles inside TiO2

loaded MCM-48 support. For this purpose the deposition-precipitation process was found to

be the better method than the wet-impregnation method. Au nanoparticles with ~3 nm crystal

119

size were prepared and the particle size was calculated from the powder XRD pattern using

the Scherrer equation. The deposition of the Au nanoparticles inside the channel was

confirmed by TEM experiments and EDX analysis. TEM images showed that the particles

were inside and the host silicate framework maintained the periodicity. XANES and EXAFS

measurements provided information about the local structure of the nanoparticles within the

mesoporous host. From XAS studies it is concluded that well ordered small metallic gold

particles were found in the MCM-48 material. The Au/TiO2-MCM-48 material showed

catalytic activity in CO oxidation. The careful handling of the material for example, without

any light exposure during synthesis and after synthesis also allowed the material to be more

active. The step-by-step development of the synthesis procedure, which was reflected in the

oxidation reaction was depicted clearly in the chapter 5.

In short, the present thesis was devoted to explore the synthesis and post synthesis

modification of cubic mesoporous MCM-48. Detailed study was done on the incorporations

of CuO, ZnO, TiO2 and Au/TiO2 oxides inside the mesoporous host. Some experiments were

carried out with the collaboration of Technical Chemistry Department, Ruhr University

Bochum, and the CO oxidation reaction on Au/TiO2-MCM-48 was performed with the help

of MPI für Kohlenforschung, Mülheim, Germany.

120

List of publications

1. Synthesis and characterization of silica MCM-48 as carrier of size-confined

nanocrystalline metal oxides particles inside the pore system. H. Gies, S. Grabowski,

M. Bandyopadhyay, W. Grünert, O.P. Tkachenko, K.V. Klementiev, A. Birkner,

Microporous and Mesoporous Materials 60 (2003) 31. (Contribution from chapter 3).

2. The structure of zinc and copper oxide species hosted in porous siliceous matrices.

O.P. Tkachenko, K.V. Klementiev, E. Loffler, I. Ritzkopf, F. Schüth, M.

Bandyopadhyay, S. Grabowski, H. Gies, V. Hagen, M. Muhler, L. Lianhai, R.A.

Fischer, W. Grünert, Physical Chemistry Chemical Physics 5 (2003) 4325.

(Contribution from chapter 3).

3. The reduction of copper in porous matrix. O.P. Tkachenko, K.V. Klementiev, N. Koc,

X. Yu, M. Bandyopadhyay, S. Grabowski, H. Gies, W. Grünert, in: E. van Stehen et

al. (Eds.), proceeding 14th International Zeolite Conference, Cape Town, South

Africa, 2004, pp. 1670. (Contribution from chapter 3).

4. Synthesis and characterization of mesoporous MCM-48 containing TiO2

nanoparticles. M. Bandyopadhyay, A. Birkner, M.W.E van den Berg, K.V.

Klementiev, W. Schmidt, W. Grünert, H. Gies, Chemistry of Materials

(Communicated, contribution from chapter 4).

5. Synthesis of MCM-48 by microwave hydrothermal process. M. Bandyopadhyay, H.

Gies, Comptes Rendus (Communicated, contribution from chapter 2).

6. Deposition of nanosized gold particles inside the TiO2-MCM-48 pore system and their

catalytic activity on oxidation of CO. M. Bandyopadhyay, M.W.E van den Berg, W.

Grünert, A. Birkner, J. Demkowski, W. Schmidt, F. Schüth, H. Gies, Chemistry of

Materials (To be communicated, contribution from chapter 5).

Acknowledgements

First of all, let me express my sincere thanks and deep sense of gratitude to my

supervisor, Prof. Dr. Hermann Gies, for his constant support, sincere guidance and

invaluable discussions during my day-to-day research work and also during the preparation

of my Ph.D. dissertation.

I am also thankful to the senior scientific members of Institut für Mineralogie, Dr. B.

Marler, Prof. Dr. W. Schmahl, Prof. Dr. S. Chakraborty and Dr. H. Graetsch, Dr. M.

Fechtelkord for helping me in various ways during my study, and also to most of the

technical staff and other colleagues of the institute including Ms. S. Grabowski, Ms. A.

Michele, Ms. I. Wilke, Mr. H. Mammen and Mr. U. Trombach for lending their helping hands

during my experimental work. I am thankful to Dr. R. D. Neuser for doing SEM

measurement. I am also grateful to Ms. U. Antoinette for the assistance in official and paper

works that I availed from her from time to time.

During my research work I had to collaborate with many scientific persons from

other institutes inside and outside Ruhr Universität, and I acknowledge their prolific

assistance. In particular, I should not forget to mention the helps from Dr. A Birkner and J.

Demkowsky of Physikalische Chemie1 for providing me the facility of TEM measurement,

Prof. Dr. W. Grünert and M. W. E van den Berg of Technische Chemie for EXAFS, XPS

measurements and calculations, Prof. Dr. M. Muhler, S. Wiedemeyer for doing N2-

physisorption experiment. My special thanks to Miss L. Khodeir of Technische Chemie for

helping me using the freeze-drying machine. I would also like to thank Prof. Dr. R. Fischer

and Dr. R. Becker of Anorganische Chemie11 for providing me the facility of UV

measurement. I am also thankful to Dr. W. Schmidt for doing N2-physisorption

measurements, Prof. Dr. F. Schüth, Dr. M. Kalwei, Dr. L. Wen-cui of MPI für

Kohlenforschung, Mülheim for doing catalytic test.

Last but not least, I must appreciate my husband, Dr. Rajib Bandyopadhyay, for his

constant, active help and mental support, without which it could have been impossible for me

to continue my higher studies and fulfill my aspiration. Finally, I dedicate my thesis to my

parents who showed me the right path for higher education.

Curriculum Vitae

Name: Mahuya Bandyopadhyay

Marital status: Married

Nationality: Indian

Date of birth: 28.12.1975 in India

School: 1986-1991 Secondary School

1991 Secondary Exam.

1991-1993 Higher Secondary School

1993 Higher Secondary Exam.

University: 1994-1997

B.Sc. with Chemistry (Chemistry Honours)

Physics and Mathematics from Burdwan University,

W.B. India

1999-2001

M. Tech. in Applied Chemistry

from Gifu University, Japan

2001-2004

Scientific Coworker in Institut für

Geologie, Mineralogie und Geophysik,

Ruhr-Universität Bochum.

Supported by Deutsche Forschungsgemeinschaft

within the framework of the Sonderforschungsbereich SFB 558

“Metal-support interactions in heterogeneous catalysis”:

B3 “Cu/ZnO stabilisiert in mesoporösen geordneten Silica-Phasen

mit mehrdimensionalen Kanalsystemen”

I hereby declare that the work incorporated in the present thesis entitled “ Synthesis of

mesoporous MCM-48 with nanodispersed metal and metal oxide particles inside the

pore system,” is carried out by myself and without the use of unauthorized help. The work,

which has been done in collaboration with other departments is duly acknowledged. This

work has not been submitted to any other department or university in this or any other form.

Mahuya Bandyopadhyay