Date post: | 23-Feb-2018 |
Category: |
Documents |
Upload: | vuongxuyen |
View: | 213 times |
Download: | 0 times |
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