scieee Science in your language
[en] (orig)
Pure and acid-functionalized ordered
mesoporous silicas:
Hosts for metallo-supramolecular coordination
polymers
vorgelegt von
Diplom-Chemikerin
Dilek Akçakayıran
geboren in Istanbul-Türkei
Von der Fakultät II – Mathematik und Naturwissenschaften –
der Technischen Universität Berlin
zur Erlangung des akademischen Grades
Doktor der Naturwissenschaften
– Dr. rer. nat. –
genehmigte Dissertation
Promotionsausschuss:
Vorsitzende: Prof. Dr. R. von Klitzing
Gutachter: Prof. Dr. G. H. Findenegg
Gutachter: Prof. Dr. R. Schomäcker
Tag der wissenschaftlichen Aussprache: 11. 11. 2008
Berlin 2009
D 83
Für Kai
Erfolg hat drei Buchstaben: TUN!
Goethe
Zusammenfassung
Native und funktionalisierte Silika Materialien, die periodisch und mesoporös sind, wurden syn-
thetisiert und mittels N2-adsorption, SAXD, SEM/TEM, CHNS-Analyse, NMR, TGA/DTA, XPS,
FT-IR und potentiometrischer Titration charakterisiert. Die Funktionalisierung wurde nach den
Grafting und Co-Kondensation Methoden durchgeführt. Die funktionalisierten Silikas wurden (i)
für die Untersuchung der Oberflächeneigenschaften von MCM-41 und SBA-15, (ii) für Acidi-
tätsmessungen in wasserfreien Poren und (iii) als Wirtsysteme für MEPE verwendet.
Die Verteilung der Silanol Gruppen an den Porenwänden wurde durch Grafting von di- und tripo-
dalen funktionellen Reagenzien F((CH3)2Si(OCH3)2, CH3Si(OCH)3) an die Oberfläche unter-
sucht. Die Anzahl der kovalenten Bindungen zwischen der Silikaoberfläche und Fwurde durch
29Si CPMAS NMR Spektroskopie bestimmt. Für hoch geordnete MCM-41 können di- und tripo-
dale Fnur eine kovalente Bindung zur Oberfläche bilden. Im Fall von SBA-15 ist auch die Bildung
von zwei kovalenten Bindungen zur Oberfläche möglich, was auf einen kleineren Abstand zwis-
chen den Silanolgruppen hinweist.
Die Co-Kondensation Methode wurde verwendet, um SBA-15 mit Carboxlysäure (CA), Phos-
phonsäure (PA) und Sulfonsäuregruppen (SA) zu funktionalisieren. Die Reaktionsbedingungen
für die Maximierung des Funktionalisierungsgrades von SBA-15 unter Beibehaltung seiner struk-
turellen Ordnung werden wesentlich durch die Kinetik der Self-assembly bestimmt. Für ein tief-
eres Verständnis des Effektes der Hydrolyse und Kondensation der Silikaprecursors (TEOS oder
F) sowie deren Wechselwirkung mit Templatmolekülen wurden verschiedene Parameter wie der
Anteil des funktionellen Silans φ, der prehydrolysierte Silikaprecursor (TEOS oder F) und die
Zeit der Prehydrolyse variiert. Die Erhöhung von φsenkt den Anteil des Templats in der funk-
tionalisierten Silika und bewirkt dabei eine bessere Entfernung des Polymers durch die H2SO4
Behandlung. Die höchsten Oberflächenbedeckungen von SBA-15 mit CA, PA und SA sind jeweils
50%, 40% und 30% mit einer ähnlichen Abnahme der Ausbeute der Funktionalisierungsreaktio-
nen. Aciditätsmessungen der funktionalisierten Materialien wurden mit der 15N NMR Methode
durchgeführt. Die erhaltenen chemischen Verschiebungen der Pyridin Molekülen zeigen, dass
SA, PA and CA Gruppen durch Pyridin Moleküle deprotoniert werden können, was auf ein ho-
hes Protondonorvermögen von diesen festen Säuren hinweist. FT-IR Messungen von CA-SBA-15
beweisen, dass die CA Gruppen im wässrigen Medium bei pH 8 deprotoniert werden.
Der Effekt der Säurefunktionalisierung auf die Adsorptionseigenschafen von SBA-15 wurde
für Fe-MEPE untersucht, welches ein metallo-supramolekulares Koordinationspolymer ist und
durch die Komplexierung von Fe(II)-Acetat mit einem neutralen Bis-Terpyridin Liganden entsteht.
Die Adsorptionsaffinität und Kinetik der MEPE-Ketten wird stark erhöht, wenn (i) die Poren mit
CA Gruppen funktionalisiert sind oder (ii) der pH Wert der Lösung oder die Temperatur steigt. Das
Adsorptionsgleichgewicht von MEPE lässt sich gut durch die Langmuir Isotherme beschreiben.
Die große Adsorptionskonstante weist dabei auf eine starke Wechselwirkung zwischen Fe-MEPE
und CA dekorierten Porenwänden hin. Die MEPE Aufnahme in die Poren kann als eine Über-
lagerung eines langsamen und eines schnellen Prozesses erster Ordnung dargestellt werden. Der
schnelle Prozess ist mit einer Erniedrigung des pH Werts verbunden, die durch einen Ionenaus-
tauschprozess verursacht wird. Der langsame Prozess dauert über mehrere hundert Stunden.
MEPE-Aufnahme ist ein diffusionskontrollierter Prozess, und die Aufnahme ist ihrerseits kon-
trolliert durch Oberflächenschicht-Resistenz. Die Stöchiometrie von MEPE in den Poren (XPS)
ist unabhängig von der Beladung. Die Länge der 15N markierten MEPE-Ketten wurde vor und
nach der Einlagerung in die CA-SBA-15 durch 15N CPMAS NMR untersucht. Demnach bewirkt
die Einlagerung eine Abnahme der Kettenlänge, wenn der Komplex in die Poren eingelagert wird.
Abstract
Pure and functionalized periodic mesoporous silicas were synthesized and characterized by N2
adsorption, SAXD, SEM, TEM, CHNS analysis, NMR, TGA/DTA, FT-IR and potentiometric
titration. Functionalization was performed by grafting and by the co-condensation route. The
functionalized silicas were used (i) for examination of the surface properties of MCM-41 and
SBA-15 silicas, (ii) for acidity measurements in waterless confined geometry, and (iii) as hosts for
MEPE.
The arrangement of the silanol groups at the pore walls was studied by grafting di- and tripod-
ical functional reagents F((CH3)2Si(OCH3)2, CH3Si(OCH)3) to the surface. The number of
covalent bonds to the surface formed by Fwas studied by pyridine adsorption using 15N CPMAS
NMR spectroscopy. For high-quality MCM-41, di- and tripodical Fcan form only one covalent
bond to the surface. In case of SBA-15 the formation of two covalent bonds to the surface is
possible as well, indicating a closer mean distance between the silanol groups at the surface.
The co-condensation route was used to functionalize of SBA-15 with carboxylic acid (CA),
phosphonic acid (PA) and sulfonic acid (SA) groups. Reaction conditions for maximizing the
degree of functionalization of SBA-15 without losing structural order are strongly affected by the
kinetics of self-assembly. To better understand the influence of the hydrolysis and condensation
steps of the silica sources and their interaction with template, the percentage of the functional
silane φthe prehydrolyzed silica source (TEOS or F), and the prehydrolysis time were varied.
Increasing φdecreases the amount of template in the functionalized SBA-15 and results in a better
polymer removal by H2SO4treatment. The highest surface coverages of SBA-15 by CA, PA,
and SA groups are 50%, 40% and 30%, respectively, with a similar decrease in the yields of the
corresponding functionalization reactions. Acidity measurements of the functionalized SBA-15
materials performed via 15N NMR. Chemical shifts measurements of pyridine indicate that SA,
PA as well as CA groups are deprotonated by pyridine, suggesting a high proton donor ability of
all three solid acids. FT-IR measurements of CA functionalized SBA-15 show that the carboxylic
acid groups are deprotonated in aqueous media at pH 8.
The effect of surface functionalization on the adsorption properties of SBA-15 was investi-
gated for Fe-MEPE, a metallo-supramolecular coordination polyelectrolyte formed by complex-
ation of Fe(II)-acetate with an uncharged ditopic bis-terpyridine ligand. The adsorption affinity
and kinetics of the MEPE chains is strongly enhanced when the pore walls are doped with CA,
and when the pH of the aqueous medium or temperature is increased. The adsorption equilib-
rium of MEPE in the pores conforms to the Langmuir isotherm with a high adsorption constant,
indicating a strong interaction between Fe-MEPE and the CA-decorated pore walls. The uptake
of MEPE into the pores can be represented by a superposition of a slow and a fast first-order
process. The fast process is connected with a decrease of pH of the aqueous solution, indicating
an ion-exchange mechanism. The slow process extends over hundreds of hours. MEPE uptake
is a diffusion controlled process, and uptake is controlled by the surface layer resistance. The
stoichiometry of MEPE in the pores (determined by XPS) is independent of the loading and sim-
ilar to that of the starting material. The mean chain length of MEPE before and after embedding
in the CA-SBA-15 was studied by solid-state 15N NMR using 15N-labeled MEPE. The average
chain-length is reduced when the complex is incorporated in the pores.
Contents
1 Introduction 5
1.1 Outline of this thesis ............................ 9
2 Experimental Background 11
2.1 UV-vis spectra of metal-ligand complexes ................. 11
2.2 Atom absorption spectroscopy (AAS) ................... 12
2.3 X-ray photoelectron spectroscopy (XPS) .................. 14
2.4 Solid-state NMR spectroscopy ....................... 16
2.4.1 Magic angle spinning (MAS) .................... 18
2.4.2 Cross polarization (CP) ....................... 19
2.4.3 Analysis of 29Si MAS NMR spectra of functionalized silicas . . . 20
2.5 Small-angle X-ray diffraction (SAXD) ................... 21
2.6 Nitrogen adsorption ............................. 24
2.6.1 Specific surface area ........................ 25
2.6.2 Analysis of microporosity ..................... 26
2.6.3 Specific pore volume ........................ 28
2.6.4 Pore size .............................. 28
2.7 Thermogravimetric analysis (TGA/DTA) .................. 31
I Ordered Mesoporous Silicas: Self-assembly and Functionaliza-
tion 33
3 State of Knowledge 35
4 Materials and Methods 43
4.1 Materials .................................. 43
4.1.1 Pure SBA-15 ............................ 44
1
4.1.2 Pure MCM-41 ........................... 44
4.1.3 Functionalization by grafting .................... 44
4.1.3.1 HMDS and DCDMS ................... 44
4.1.3.2 DMDMS and MTMS .................. 45
4.1.3.3 Condensation of DMDMS and MTMS without silica . 45
4.1.4 Acid-functionalized SBA-15 by co-condensation route ...... 46
4.1.4.1 Carboxylic acid doped SBA-15 (CA-SBA-15) ..... 47
4.1.4.2 Phosphonic acid doped SBA-15 (PA-SBA-15) ..... 48
4.1.4.3 Sulfonic acid doped SBA-15 (SA-SBA-15) ....... 48
4.2 Methods ................................... 48
4.2.1 Nitrogen adsorption ........................ 48
4.2.2 SAXD ................................ 49
4.2.3 TGA ................................ 49
4.2.4 13C, 29Si and 15N solid-state NMR ................ 49
4.2.5 FT-IR ................................ 50
4.2.6 Potentiometric titration ....................... 51
4.2.7 Scanning electron microscopy (SEM) ............... 51
5 Functionalization by Grafting 53
5.1 HMDS and DCDMS grafted samples .................... 54
5.2 DMDMS and MTMS grafted samples ................... 56
5.3 Arrangement of the surface silanol groups ................. 57
5.4 Conclusions ................................. 60
6 Acid-Functionalization by Co-condensation 61
6.1 Template removal: calcination vs. acid treatment ............. 63
6.2 Carboxylic acid functionalization ...................... 67
6.2.1 Morphology, structure and porosity ................ 67
6.2.2 Polymer removal and microporosity ................ 69
6.2.3 Reaction yield and the degree of the functionalization ....... 72
6.2.4 Water stability of CA-SBA-15 materials .............. 73
6.3 Sulfonic acid functionalization ....................... 74
6.3.1 Study of template content ..................... 75
6.3.2 Morphology, structure and porosity ................ 76
6.3.3 Reaction yield and the degree of functionalization ......... 79
6.3.4 Polymer removal and microporosity ................ 83
6.3.5 The mechanism of the self-assembly ................ 85
6.4 Phosphonic acid functionalization ..................... 87
6.4.1 Morphology, structure and porosity ................ 88
6.4.2 Reaction yield and degree of the functionalization ......... 93
6.4.3 Polymer removal and microporosity ................ 95
6.4.4 The mechanism of the self-assembly ................ 99
6.5 Acidity measurements ...........................101
6.5.1 Measurements in non-aqueous environment by 15N NMR . . . . . 101
6.5.2 Measurements in aqueous environment by FT-IR .........103
6.6 Conclusions .................................105
II Functionalized SBA-15 Silica as a Host 107
7 Background 109
7.1 Supramolecular polymers ..........................109
7.1.1 Fe-MEPE ..............................110
7.2 Lifetime of a supramolecular bond .....................112
7.2.1 Lifetime of a coordinative bond ..................112
7.2.2 Living polymers ..........................114
7.3 Sorption ...................................115
7.3.1 Adsorption .............................115
7.3.2 Electrostatic sorption (ion exchange) ................116
7.4 Sorption equilibrium ............................117
7.4.1 General description .........................117
7.4.2 Isotherm equations .........................118
7.4.2.1 Henry adsorption isotherm ...............118
7.4.2.2 Freundlich adsorption isotherm .............119
7.4.2.3 Langmuir adsorption isotherm .............119
7.5 Sorption kinetics ..............................120
8 Materials and Methods 123
8.1 Materials ..................................123
8.1.1 Fe-MEPE ..............................123
8.1.2 Fe-Terpy ..............................123
8.1.3 Uptake isotherms ..........................124
4
8.1.3.1 UV-vis calibration curves ................124
8.1.3.2 Uptake in the silica ...................125
8.1.3.3 Settling of Fe-MEPE during centrifugation .......125
8.2 Methods ...................................127
8.2.1 XPS .................................127
8.2.2 AAS ................................127
8.2.3 UV-Vis ...............................127
8.2.4 15N solid-state NMR ........................127
9 Fe-MEPE in CA Doped SBA-15 Silica 129
9.1 Introduction .................................130
9.2 Results ....................................132
9.2.1 Characterization of silica hosts ...................132
9.2.2 Characterization of coordination compounds by NMR ......132
9.2.3 Effect of CA functionalization on the Fe-MEPE uptake ......135
9.2.4 Characterization of Fe-MEPE in CA-SBA-15 ...........138
9.2.5 Characterization by solid-state 15N NMR .............144
9.3 Discussion ..................................146
9.3.1 MEPE uptake by ion exchange ...................146
9.3.1.1 Estimation of pH change ................147
9.3.2 Effect of electrostatic interactions .................148
9.3.3 Adsorption affinity and chain packing ...............149
9.3.4 Stoichiometry and chain length of MEPE in the pores .......150
9.4 Conclusions .................................151
10 Kinetic Study of Fe-MEPE Uptake 153
10.1 Fe-Terpy-uptake vs. Fe-MEPE-uptake ...................154
10.2 Adsorption kinetics of Fe-MEPE ......................156
10.3 Adsorption equilibrium of MEPE ......................161
10.4 Conclusions .................................163
11 Summary and Outlook 165
Chapter 1
Introduction
Porous materials play an important role in nature. For example, porosity of soils is crucial
for water permeability and also for water and fossil oil and gas storage. In a different
direction, the heat insulation of building materials is mainly effected by their porosity.
Porosity is defined by the ratio of the volume of void-space and the total volume of
the material, including the solid and void components. According to the IUPAC definition
porous materials are categorized in three groups depending on their pore dimensions.
Materials having pore diameter less than 2 nm are denoted by microporous. The diameter
of mesoporous materials is between 2 and 50 nm, whereas macroporous materials have
diameters bigger than 50 nm.
In material science microporous materials, also known as molecular sieves, make up
an important class of materials. First there are zeolites belonging to the class of alu-
minosilicates which are crystalline structures with micropore channels. Besides about 50
naturally occurring zeolites more than 150 synthetic zeolite types have been manufactured
and are used in various fields including adsorption (water purification and production
of medical-grade oxygen), ion exchange (water softening, detergents, soil treatment in
agriculture), size selective molecular separation (chromatography) and in catalysis (fluid
catalytic cracking and hydro-cracking in the petrochemical industry)[1]. More recently,
metallo-organic framework (MOF) compounds have been gained interest for various ap-
plications including hydrogen gas storages [2]. However, for many other applications
microporous materials are not suitable due to their small pore size.
These constrains were overcome by the discovery of periodic mesoporous materials
such as MCM-41 and SBA-15 [3,4]. These materials constitute a 2D hexagonal arrange-
ment of cylindrical pores with a diameter tunable in a range from 2 to 12 nm. Due to their
wide pore openings, narrow pore size distribution and large internal surface area, these
5
6CHAPTER 1. INTRODUCTION
materials have a high potential for applications in many fields. Although the larger size
of the pores is a prerequisite for many applications, it is by itself often not sufficient when
specific surface properties are desired. Specific surface sites for metal-ion binding, catal-
ysis, or controlled sorption can be introduced through surface functionalizations. Surface
functionalization can be performed by different routes. Two of them, surface grafting and
co-condensation methods, are employed in this Thesis.
It is well known that high-quality MCM-41 silica constitutes quasi-ideal arrays of
uniform-size pores with thin pore walls and a smooth surface, while SBA-15 silica has
thicker pore walls with framework and surface defects (roughness) [5]. However, more
detailed information about the structure of the pore walls and inner surfaces of these ma-
terials is lacking. Such details are relevant for applications of these materials as catalysts
and host materials. The clear need to gain further insight into the surface characteristics
of these silica materials has been one of the motivations for this thesis. To this end, the
surface of MCM-41 and SBA-15 silicas were functionalized by the grafting route using
(CH3)2Si(OCH3)2and (CH3)Si(OCH3)3as grafting agents. The arrangement of surface
silanol groups at the silica surfaces before and after the functionalization was then studied
by solid-state NMR methods.
Acid-functionalized SBA-15 materials are promising solid-state catalysts for various
applications in the field of heterogenous catalysis [6]. Our aim was to study the acidity
of the functional groups inside the mesopore channels of SBA-15 silica via solid-state
NMR, in order to find out whether the proton-donor ability is effected by the confined
geometry. Pyridine was used as a probe molecule to measure the acidity of the func-
tional groups. This approach exploits the specific dependence of the 15N chemical shift of
hydrogen-bonded pyridine on the length of the corresponding N···H distance [5,7]. For
such experiments, high-quality SBA-15 materials with a large number of acid groups are
needed. Our synthesis of choice was the co-condensation route because the high surface
coverage of silica by acid groups could not be achieved by the grafting method. SBA-
15 materials containing carboxylic acid, phosphonic acid and sulfonic acid groups, with
increased acidy in the liquid phase, were synthesized by the co-condensation route (see
Figure 1.1). In this method, a fraction of the silica precursor tetraethoxysilane (TEOS)
is replaced by the functional silane. In order to obtain a high degree of functionalization
relatively large proportions of functional silane have to be used. However, increasing the
proportion of the functional silane beyond a certain extent leads to the formation of dis-
ordered materials or to materials with morphologies different from the typical worm-like
structures of SBA-15. Therefore, during the synthesis of functionalized materials various
7
Si OEt
OEt
OEt
OSi
O
O
O
OH
CN
pKa=4.7
Si OEt
OEt
OEt
P
O
EtO
EtO OSi PO
OH
OH
O
OpKa=2.1
O
O
OSi SOH
O
O
Si OH
OH
OH
S
O
O
OH pKa= - 2.6
in bulk:
carboxylic acid
phosphonic acid
sulfonic acid
CTES
2-cyanoethyl
triethoxysilane
PTES
Diethoxyphosphoryl
ethyltriethoxysilane
3-(trihydroxysilyl)-1-
propane-sulfonic acid
STHS
Figure 1.1: Functional silanes used in the preparation of acidic SBA-15 materials by co-
condensation route.
parameters had to be changed stepwise in order to understand the formation mechanism
and to optimize the reaction conditions for the desired functionalization.
The modified and functionalized porous silica materials have potential applications
in the field of heterogenous catalysis [8,6] and photocatalysis involving bulky grafted
catalysts, for example enzymes (enzyme immobilization) and/or the conversion of large
substances [9,10]. Other potential applications include ion exchange and separation,
removal of heavy metals, chromatography, stabilization of quantum wires, stabilization
of dyes and polymer composites [11]. With an eye on such applications we studied the
incorporation of Metallo-supramolecular coordination polyelectrolyte (MEPE) such as
Fe-MEPE (2) in the pores of SBA-15 (see Figure 1.2).
Fe-MEPE is a poly-cationic, water soluble organometallic complex assembled from
Fe(II)-acetate and 1,4-bis-(2,2´:6´2´´-terpyridine-4´-yl)-benzene (1) by the coordinative
bonds [12]. The quasi-rigid chain molecules with their stereochemically well-defined
octahedral coordination geometry can be regarded as cylinders of diameter 1.5 nm. More-
8CHAPTER 1. INTRODUCTION
COOH COOH COOH
COOH COOHCOOH
OH OH OH
OH OHOH
SBA-15 CA-SBA-15
COOH COOH COOH
COOH COOH COOH
COOHCOOH COOHCOOH COOHCOOH
COOH COOHCOOH COOHCOOH COOHCOOHCOOHCOOH
OH OH OH
OH OH OH
OHOH OHOH OHOH
OH OHOH OHOH OHOHOHOH
SBA-15 CA-SBA-15
(a)
(b)
(c)
Figure 1.2: a) Metallo-supramolecular coordination polyelectrolyte (MEPE) formed by
self-assembly of Fe(II) ions with the ditopic ligand 1,4-bis-(2,2´:6´2´´-terpyridine-4´-yl)-
benzene (1). (b) Monotopic Fe-Terpy (4) formed by complexation of Fe(II) ions and
2,2´:6´2´´-terpyridine (3). (c) Cartoon of the cylindrical mesopores of SBA-15 and CA-
SBA-15, in which the pore walls are decorated by propionic acid groups, here indicated
by -COOH.
over, Fe-MEPE is a supramolecular polymer, meaning that the bonds between the monomers
are not covalent but based on weaker, reversible interactions. For example, some proteins
in biological cells such as actin, fibrinogen and tubulin polymerize into long filaments
by supramolecular interactions. Such a polymerization is important for cell rigidity, cell
motility, and intercellular transport [13]. In the case of Fe-MEPE, the reversible interac-
tions based on the coordinative bonds between the ligand and the metal ion are caused the
living-polymer character of Fe-MEPE. This means that the chains brake and recombine
on the experimental timescale. As a consequence, the chain length distribution may vary
depending on the conditions, and it is possible to tune the properties of these polymers
by changing the concentration, solvent, pH, temperature, or other parameters. In addition
to dynamic properties coming from labile metal ion ligand interactions, Fe-MEPE ex-
1.1. OUTLINE OF THIS THESIS 9
hibits other potentially useful properties such as electro-chromic behavior and molecular
magnetism [14].
In this Thesis the uptake of Fe-MEPE into the mesopores of pure and acid-functionali-
zed SBA-15 was studied as an example for the adsorption behavior of charged bulky
supramolecular molecules into the silica pores. The adsorption was performed under
varying conditions of pH, temperature and the chemical properties of the pore wall. The
latter was accomplished through the use of SBA-15 functionalized by carboxylic acid
groups (CA-SBA-15, see Fig. 1.2c) which are similar to the acetate ions forming the
counter ions of Fe-MEPE in solution. We studied in detail the adsorption kinetic and
the adsorption equilibrium in water. In addition, the main chain length of Fe-MEPE
before and after the adsorption into the pores was examined by means of 15N solid-state
NMR. In this experiment the monotopic Fe-Terpy complex (4) formed by complexation
of Fe(II) ions and 2,2´:6´2´´-terpyridine (3) (see Fig. 1.2b) is used for the assignment
of the different 15N resonances of Fe-MEPE. The solid-state NMR proved to be a useful
method for estimating the chain lengths.
1.1 Outline of this thesis
This Thesis is divided into two parts. The first part concentrates on the preparation and
characterization of pure and functionalized MCM-41 and SBA-15 silica materials, while
the second part focuses on the uptake of the supramolecular coordination polymer Fe-
MEPE into the pores of SBA-15 silicas.
Part I deals with synthesis and characterization of functionalized silica materials syn-
thesized both by the co-condensation and grafting routes. The main results of Part I are
presented in Chapter 5 and Chapter 6. In Chapter 5, MCM-41 and SBA-15 silicas with
hydrophobized pore walls prepared by the grafting route using (CH3)2Si(OCH3)2and
(CH3)Si(OCH3)3reagents are presented. The 29Si NMR method is applied to obtain the
ratio of numbers of single-, double-, and triple surface-bonded species for each reagent.
These results in turn provide qualitative information about the arrangement of the surface
silanol groups, from which the surface properties of MCM-41 and SBA-15 silicas can be
assessed. Chapter 6 concentrates on the characterization of SBA-15 materials contain-
ing different acid groups (carboxylic, phosphonic, and sulfonic acid) tethered to the pore
walls, which were synthesized by co-condensation route in acidic medium. The effect
of several synthesis parameters, such as the relative amount of the functional silane, the
sequence in which the silica precursors (TEOS or functional silane) were added to the
10 CHAPTER 1. INTRODUCTION
synthesis solution, and the time allowed for prehydrolysis of that component on the mor-
phology and on the acidity of SBA-15 were investigated by various methods. After the
characterization of the products, the proton donor ability of the different acid functionali-
ties inside the pores was studied by 15N CPMAS NMR after pyridine adsorption into the
pores.
Part II deals with the investigation of the Fe-MEPE uptake into the pores of SBA-15.
The main results of Part II are presented in Chapter 9 and Chapter 10. In Chapter 9 it
is shown that the uptake can be increased considerably at higher temperatures, at higher
pH values and also by using carboxylic acid functionalized SBA-15 as adsorbent. It turns
out that the ion exchange process plays a role on the adsorption. In addition, the chain
lengths of Fe-MEPE in the pores estimated by 15N CPMAS NMR are shorter than those
in solution. From this we concluded that Fe-MEPE chains break into smaller entities in
the pores. Chapter 10 outlines the sorption kinetic and sorption equilibrium of MEPE in
the pores of carboxylic acid functionalized SBA-15. The Fe-MEPE uptake curves can be
represented by a sum of two exponential decay functions, for the fast initial process and a
subsequent slow uptake process which leads to the adsorption equilibrium. The diffusion
is controlled by the surface layer resistance and the solid-liquid equilibrium of Fe-MEPE
can be described by the Langmuir adsorption isotherm satisfactorily.
The work presented here arose from a cooperation with two other research groups in
the framework of Sonderforschungsbereich 448 Mesoskopisch struktruierte Verbundsys-
teme. All solid-state NMR measurements were performed in the group of Prof. Limbach
and Dr. Shenderovich of FU-Berlin, whereas the supramolecular coordination polymer
Fe-MEPE was provided by the group Prof. Möhwald and Dr. Kurth of MPI of Colloids
and Interfaces in Potsdam-Golm.
Chapter 2
Experimental Background
2.1 UV-vis spectra of metal-ligand complexes
The different types of electron transitions in an octahedral coordinated metal complex
ML6with πacceptor ligands by means of the molecule orbital scheme are shown in a
simplified form in 2.1a [15].
d
s
p
10 Dq
p*
s
p
MLCT IL
LMCT
LF
metal orbitals
molecule orbitals ligand orbitals
300 400 500 600 700 800
0,0
0,2
0,4
0,6
0,8
1,0
absorbance
l/nm
Fe-MEPE
(a) (b)
Figure 2.1: (a) MO scheme for a simple octahedral coordinated metal complex ML6; (b)
the UV-vis absorption spectrum of 0.2mM Fe-MEPE in water.
The molecule orbitals of the complex are formed by the combination of metal orbitals (s,
p and d) and ligand orbitals (σ,πand π∗). The following transitions can be observed:
1. Ligand field (LF) transitions take place between orbitals which basically have metal
character (d→dtransition)
2. Charge transfer (CT) transitions take place between metal and ligand orbitals. If
the transition occurs from a filled metal orbital to a empty ligand orbital, it is called
11
12 CHAPTER 2. EXPERIMENTAL BACKGROUND
Metal-to-Ligand-Charge-Transfer (MLCT) transition. In the opposite case, if the
transition occurs from a filled ligand orbital to a empty metal orbital, this is denoted
as Ligand-to-Metal-Charge-Transfer (LMCT) transition.
3. Intra ligand (IL) transitions take place between orbitals essentially having ligand
character (π→π∗transition)
These transitions can be detected by UV-vis spectroscopy. In Figure 2.1b a UV-vis ab-
sorption spectrum of 0.2mM Fe-MEPE in water is shown. The bands centered at 288
nm and 319 nm are assigned to IL transitions. The peak at 585 nm is attributed to the
MLCT transition. The small peak at 368 nm which appears as a shoulder comes from
a LF transition and is assigned to the transition of Fe2+ electrons between the d-orbitals
splitted in the octahedral ligand field.
The absorption of light follows the Lambert-Beer law. Accordingly, the intensity
of transmitted light decreases exponentially with increasing molar concentration of the
absorbing species cand the thickness dof the sample . The equation has the following
form
log I
I0
=−ε·c·d(2.1)
where I0is the incident intensity, Iis the intensity after passage through the sample, and ε
is the molar absorption coefficient of the species at the frequency of the incident radiation.
It depends on the frequency of the radiation, and its dimension (length.concentration)−1
is usually expressed in L mol−1cm2. The dimensionless quantity A=ε·c·dis called
absorbance (formerly extinction or optical density) of the sample, and the ratio I/I0is
called the transmittance T. These two quantities are connected by the equation
logT =−A. (2.2)
2.2 Atom absorption spectroscopy (AAS)
The Atom Absorption Spectroscopy (AAS) is a proven and fast method for qualitative
and quantitative analysis of many elements in (mostly) aqueous solutions [16]. The basis
of AAS is the ability of the atoms to absorb electromagnetic radiation of a certain wave-
length in their initial state and to go thereby into an excited state. For analytical use of
this principle in AAS the atoms are often generated by thermic splitting of suitable com-
pounds. In this work an atomizer-burner system with air/acetylene flame (T≈1800◦C)
is used for the atomization. This process is called flame AAS method (FAAS). During the
passageway through the flame various processes occur consecutively:
2.2. ATOM ABSORPTION SPECTROSCOPY (AAS) 13
•The solvent (mostly water or aqueous systems such as dilute acids) evaporates from
the aerosol (i.e., the droplets of sample solution)
•The resulting solid particles evaporate
•The molecules dissociate into atoms
•A very small number of atoms is excited thermally or ionized.
In the Figure 2.2 the schematic construction of a flame AAS spectrometer is shown. For
AAS it is essential to maximize the proportion of atoms converted into gas state and mini-
mize the number of excited or ionized atoms. Because the elements differ from each other
in their electronic structure they absorb radiation at strictly characteristic wavelengths.
For this reason the resulting "atom cloud" in the flame is illuminated with light character-
istic for the element to be determined. To this end light with the intensity I0is radiated
Figure 2.2: Schematic construction of a flame AAS spectrometer.
into the flame in the atomizer. Subsequently, owing to the resonance process, light with
the same wave length is emitted. However the emission occurs in all directions of the
room so that a decrease of the incident intensity to the value Iis detected. This decrease
is proportional to the concentration of the element in the sample and can be calculated by
the Beer-Lambert law (Equation 2.1).
AAS is a relative method. This means that for the determination of the element con-
centration it is necessary to measure the extinction of calibration standards of known
concentration. In this work the standard addition method [17] is used for the calibration.
Because of the short time span during which the atoms are exposed to radiation in the
AA spectrometer the sensitivity of the flame technique is not very high. This technique
has a detection limit in the range from a few mg/L up to a few µg/L, depending on the
element.
14 CHAPTER 2. EXPERIMENTAL BACKGROUND
2.3 X-ray photoelectron spectroscopy (XPS)
Photoelectron spectroscopy (PES) is the examination of energy levels of molecules by
determining the kinetic energies of electrons ejected by absorption of high-frequency
monochromatic radiation. If the incident radiation is in the X-ray region, then the tech-
nique is called XPS or ESCA (electron spectroscopy for chemical analysis). In XPS, the
sample is irradiated with low-energy( 1.5keV) X-rays, in order to provoke the photoelec-
tric effect. The kinetic energy of photoejected electron Ekin is approximately equal to the
difference between the energy of X-rays hν and the binding energy of the electron Eb:
Ekin =hν −Eb.(2.3)
Therefore, a determination of Ekin will give the value of Eband hence, via Koopmans’
theorem, the energy of the orbital from which the electron was ejected. Because the core
electrons are largely (but not entirely) independent of the state of the bonding atom, the
binding energies are characteristic of specific electron orbital in specific atoms. This can
be used to identify the elements present in the sample.
The following processes can occur after the photo-ionization [18]:
•Generation of Auger-electrons, which takes place after the escape of the photoelec-
tron. An electron from a higher energetic level drops into the resulting gap. The
corresponding energy causes the release of a third electron from the atom. This pro-
cess predominates in case of light elements. The kinetic energy of Auger electrons
is independent on the energy of X-rays. Therefore they can be distinguished from
photoelectron lines by measurements at two different X-ray excitation energies.
•Absorption of X-ray radiation followed by emission of characteristic X-rays (X-ray
fluorescence). This process predominates in case of heavy elements.
•Auto ionization.
Both XPS-electrons and Auger-electrons are analyzed with regard to their kinetic energy.
From these data it can be established, after comparison with tabulated values, which el-
ement is ionized by the X-ray radiation. Hydrogen and helium are missing from such
tables, because they are essentially impossible to detect by XPS. Helium does readily
form solid compounds and its 1s orbital has a tiny cross-section for photo emission. Hy-
drogen also has a tiny cross-section and suffers from having to share its only electron
in forming compounds, which then resides in a valence-like orbital with varying energy
2.3. X-RAY PHOTOELECTRON SPECTROSCOPY (XPS) 15
from compound to compound.
Chemical Shift
In XPS, the term chemical shift is used to describe the shift in binding energy of elec-
trons dependent on the chemical surrounding such as state of binding, valency and type of
surrounding ligands. In compounds of elements with different electronegativity (e.g. met-
als with oxygen) the valence electrons are shifted the more electronegative partner. This
leads to an increase of effective core charge, so that the core electrons of metal atoms stay
in a stronger Coulomb field and thus are stronger bonded. These electrons leave the sam-
ple surface with lower kinetic energy, so that the signal in the spectrum is shifted towards
higher binding energies.
The comparison of the N1s spectrum of embedded Fe-MEPE with that of bulk Fe-
398400402
Binding Energy (eV)
0
0.01
0.02
0.03
Intensity
a
b
c
N1s
399.5400.1
Figure 2.3: X-ray photoelectron N1s spectra of 20-CA loaded with Fe-MEPE (a) 0.4 mM
(b) 0.6 mM as well as (c) bulk Fe-MEPE normalized in intensity to spectrum b.
MEPE shows an intensity shift to higher binding energies. For pyridine a binding energy
of 399.5 eV was reported [19]. This upshift may therefore indicate the increased number
of nitrogen atoms participating in Fe coordination.
16 CHAPTER 2. EXPERIMENTAL BACKGROUND
Escape depth of photo electrons
X-rays can easily penetrate into materials to a depth of some µm. The use of XPS as
a surface investigation method results from the limited escape depth of photo electrons.
This escape length was determined for different materials. Thereby it was shown, that the
mean free path of the electrons λmis dependent on their kinetic energy Ekin. Seah and
Dench [20] give the following empirical equations describing the mean free path
λm=538
E2
kin
+ 0.41paEkin (2.4)
λm=2170
E2
kin
+ 0.72paEkin (2.5)
where λmis expressed in nm and Ekin is expressed in eV. The factor ain the root rep-
resents the layer thickness in nm. Equation 2.4 applies to elements and Equation 2.5 to
anorganic substances. These relations are used to plot the mean free path of an electron
against its kinetic energy. In a range from 10 eV to 1100 eV the values of λmare in
a range of approximately 1to 3nm, e.g., photoelectrons with this energy mainly come
from the top layers of the sample. Therefore XPS is very surface sensitive. For this reason
the sample analysis is conducted in a vacuum chamber under the best vacuum conditions
achievable, typically ca. 10−10 torr. This facilitates the transmission of the photoelectrons
to the analyzer and, more importantly, minimizes contamination of the sample.
2.4 Solid-state NMR spectroscopy
The elementary particles (neutrons and protons) composing an atomic nucleus, have the
intrinsic quantum mechanical property of spin. The overall spin of the nucleus is deter-
mined by the spin quantum number I. If the number of both the protons and neutrons in
a given isotope are odd then I6= 0, i.e. there is an overall spin. This non-zero spin, I, is
associated with a non-zero magnetic moment, µ, via
µ=γ·I(2.6)
where the proportionality constant, γ, is the gyromagnetic ratio. It is this magnetic mo-
ment that is exploited in NMR. The spin, and hence the magnetic moment, of the nucleus
may lie in 2I+ 1 different orientations relative to an arbitrary axis (usually z-axis). In the
stationary state, all orientations have equivalent energies. However, in an external mag-
netic field B0these 2I+ 1 orientations of the nucleus have different energies. The energy
2.4. SOLID-STATE NMR SPECTROSCOPY 17
difference ∆Ebetween the affected energy levels is proportional to the applied magnetic
field, and the resonance is achieved by excitation of the sample with electromagnetic ra-
diation in the radio frequency region [21].
In solution NMR, spectra consist of a series of sharp lines due to averaging of anisotropic
interactions by rapid random tumbling of the molecules. By contrast, solid-state NMR
spectra are very broad (up to several kHz), as the full effects of anisotropic or orientation-
dependent interactions are observed in the spectrum. Thus in solid state the chemical
shielding becomes anisotropic due to the different orientation of molecules relative to the
external magnetic field. Therefore the chemical shielding is characterized by a shield-
ing tensor. Molecular orbitals and crystallographic symmetry dictate the orientation and
magnitude of chemical shielding tensors [22]. In the solid state the interactions between
the magnetic active nuclei become anisotropic as well. These can be dipolar interactions
resulting from interaction of the nuclear spin with a magnetic field generated by another
nuclear spin, and vice versa. This is a direct through space interaction depending on the
magnetogyric ratio γof each nucleus as well as on the distance between the nuclei. If the
sample contains nuclei with a quadrupol moment (I≥1), the anisotropic interactions of
quadrupol with the electric field gradient affect the line form and the number of resonance
lines.
150 100 50 0ppm
Solution 13C NMR
Solid State 13C NMR
Figure 2.4: Comparison of 13CNMR spectra in solution and solid state [22].
18 CHAPTER 2. EXPERIMENTAL BACKGROUND
In a NMR experiment a directed macroscopic magnetization is induced. Its time de-
pendent decay is described by two independent time constants T1and T2. The relax-
ation time T1is called as longitudinal relaxation time. It gives information about re-
establishment of Boltzmann-distribution of the energy levels of a nucleus. In this process
an energy exchange of the nucleus with its surroundings takes place. ("spin-lattice" relax-
ation). The relaxation time T2, the so-called "transversal" relaxation time, characterizes
the decay of the intensity of the induced magnetization, and it is an entropic process. Thus
the constant T2describes the line width of the resonance signals whereas T1determines
repetition rate of the NMR experiment.
The potential of NMR measurements in solid state is affected in three different ways,
which do not pose serious problems in liquid state:
•Chemical shielding anisotropy
•Strong anisotropic dipolar interactions, i.e., short transversal relaxation time T2
•Long longitudinal relaxation time T1.
The first two lead to the distribution of the signal intensities over a wide frequency range
of several kHz. On the other hand, a rapid accumulation of several spectra is not possible
because T1is long. High resolution solid state NMR spectra can provide the same type of
information that is available from corresponding solution NMR spectra, but special tech-
niques and equipment are needed, including magic-angle spinning and cross polarization.
2.4.1 Magic angle spinning (MAS)
Each magnetic active nucleus produces a local magnetic field B0∝1
R3(1 −3cos2θ),
where R is the distance from the nucleus and θis the angle between the applied field B0
and the principal axis of the molecule. The anisotropy of the chemical shift also varies
with the angle θas (1 −3cos2θ). For this reason the line widths of solid samples can be
reduced by spinning the sample with high speed at the magic angle θ= 54.74◦at which
the factor (1 −3cos2θ) = 0. At adequate rapid spinning in time average all vectors in the
sample take the magic angle to the external magnetic field.
In order for the MAS method to be successful, spinning has to occur at a rate equal
or greater than the one corresponding to the dipolar line width. In our measurements the
rotation frequency was 6−8kHz, corresponding to a rotation rate of up to 8000 rps. In
this case the anisotropic interactions between the nuclei among themselves and between
the nuclei and the external magnetic field will be ineffective. Thus by MAS the problem
2.4. SOLID-STATE NMR SPECTROSCOPY 19
q
q=54.74°
Figure 2.5: The principle of Magic Angle Spinning NMR spectroscopy (MAS-NMR) and
spectra before and after MAS [22].
of anisotropy of chemical shielding is solved and at the same time the weak heteronuclear
and strong homonuclear dipolar interactions are eliminated or reduced.
2.4.2 Cross polarization (CP)
Cross polarization is a very important technique in solid state NMR. In this technique,
polarization from abundant spins such as 1Hor 19F is transferred to dilute spins such as
13C,29Si or 15N. The overall effect is to enhance the signal-to-noise ratio:
1. Cross polarization enhances the signal from dilute spins potentially by a factor of
γI/γS, where γIand γSare the magnetogyric ratios of the abundant and the dilute
spin. For instance, by transfer of the spin polarization from 1Hto 13Cthe signal of
the latter can be enhanced by the factor 4.
2. Since abundant spins are strongly dipolar coupled, they are subject to large fluctuat-
ing magnetic fields resulting from motion. This induces rapid spin-lattice relaxation
at the abundant nuclei. As a result, the repetition rate for the accumulation of the
spectra is no longer dependent on the T1of the dilute spin but on that of the abun-
dant spin. For instance, T1(1H)is in the region of seconds, and thus it is in general
significantly shorter than T1(S)for many nuclei S.
A major problem in the application of the CP technique is that, owing to the strong mag-
netic moment of I nuclei the anisotropic interactions between S and I are strong as well,
20 CHAPTER 2. EXPERIMENTAL BACKGROUND
and this leads to a widening of the resonance signals of S nuclei. However, this problem
can be solved by the so-called "spin locking" method, by which the Inuclei are decoupled
from Snuclei.
Cross polarization requires that nuclei are coupled to one another by dipolar inter-
actions, and it even works while samples are being spun at the magic angle. Hence the
acronym CPMAS NMR (Cross Polarization Magic-Angle Spinning NMR). By CPMAS
the time of the measurement can be reduced drastically. The disadvantage of this method
is that information about the concentration of the respective species is lost.
2.4.3 Analysis of 29Si MAS NMR spectra of functionalized silicas
When applying the MAS NMR technique to mesoporous silicas such as SBA-15 and
MCM-41, three peaks can be observed in the spectrum which may be assigned to three
kind of silicon atoms. Silicon atoms with 4 O−Si neighbors are denoted as Q4. These
atoms are located inside the silica framework. Analogously, Q3and Q2silicon atoms have
3 and 2 O−Si neighbors (and 1 or 2 OH neighbors), respectively. These silicon atoms
are situated at the pore walls, as shown in Figure 2.6a.
COOH
Si
Si
O
Si
O
Si
O
OH
OO
Si
O
O
Si
O
Si
OO
OH
Si
Si
O
O
Si
O
OH
OH
Si
O
O
O
Si
COOH
T2
Q4
Q2Q3
T3
Figure 2.6: Cartoon of a silica framework (a) and a 29Si MAS NMR spectrum of car-
boxylic acid doped SBA-15, 20-CA (b).
Silicon atoms carrying an alkyl chain are denoted by T. Again the exponent indicates
the number of O-Si neighbors of a silicon atom. The different Q and T silicon atoms have
different chemical shifts in the 29Si spectra as can seen in the small table in Figure 2.6b.
Also shown in Fig. 2.6b is the silicon NMR spectrum of a SBA-15 doped with carboxylic
2.5. SMALL-ANGLE X-RAY DIFFRACTION (SAXD) 21
acid (20-CA). The three peaks of this spectrum are assigned to Q4,Q3and T3silicon
atoms. The number of silicon atoms directly connected to functional groups F(propionic
acid) is proportional to the integral intensities of the peaks T3+T2, while the number of
silicon atoms bearing −OH groups is proportional to the peak intensities Q3+2Q2+T2.
Accordingly, the mole fraction of functional groups in the mixture of surface groups (here
−OH and −COOH) at the pore walls is given by:
NF
NOH +NF
=T2+T3
2Q2+Q3+ 2T2+T3(2.7)
To determine the mole fraction of F groups, the spectrum is fitted by a deconvolution
method. The data for the individual peaks are normalized to the Q4peak and adjusted to
the ratios of the integrals of the T and Q regions. Since in some cases it was not possible
to separate the peaks T3and T2, or Q3and Q2in the spectra, the fraction xof surface
silica atoms carrying a functional group is used as an approximate measure of the degree
of the surface coverage by functional groups at the pore walls:
xF=T2+T3
Q2+Q3+T2+T3(2.8)
Moreover, the yield of a synthesis by co-condensation route can be estimated from a
comparison of the ratio T/Q with the ratio of functional groups and TEOS in the synthesis
mixture, where T/Q is given by
T
Q=T2+T3
Q2+Q3+Q4(2.9)
2.5 Small-angle X-ray diffraction (SAXD)
Periodic mesoporous silica materials exhibits a periodic arrangement of mesopores while
the silica matrix is amorphous at the atomic scale. The periodic arrangement of the meso-
pores produces characteristic Bragg reflexes in the small-angle diffraction, as shown in
the SAXD spectrum in Figure 2.7a. Specifically, MCM-41 and SBA-15 comprise a two-
dimensional hexagonal pore lattice (space group p6mm) as indicated in Figure 2.7b.
The walls of the mesoporous materials investigated in this work are amorphous in the
atomic level. Moreover, as seen in Figure 2.7a, while examining these materials with
SAXD we observe Bragg-reflexes which suggest an order of the materials in the meso-
scopic scale. The reason for the appearance of Bragg-reflexes is that the mesopores in the
materials are well ordered. The peaks are to be attributed to the 2D hexagonal unit cell
22 CHAPTER 2. EXPERIMENTAL BACKGROUND
with the lattice constant a0. For this reason a crystallographic formulation can be used for
the small angle X-ray scattering on these materials.
0,4 0,8 1,2 1,6 2,0
30
21
20
11
Intensity/a.u.
q/nm-1
10 SBA-15
(a)
a0
dhk
q
D
C
A
B
qq
(b)
Figure 2.7: SAXD spectrum of a SBA-15 (a) and derivation of Bragg’s law for X-ray
diffraction on hexagonal ordered materials (b).
Figure 2.7b illustrates the principle of diffraction of monochromatic X-rays of wave-
length λby the lattice planes of a 2D hexagonal crystal lattice. If the lattice planes are
considered as semipermeable mirrors a portion of incident X-rays is reflected and the rest
is transmitted. A Bragg-reflex originates if X-rays reflected at numerous parallel lattice
planes interfere constructively. This happens when the pathlengths of the X-rays reflected
at two neighboring lattice planes differ by an integer multiple of λ, i.e. ADC =n·λ.
This is described by the Bragg equation. For our 2D ordered pore system where the lattice
planes can be labeled with Miller indices hk one has
n·λ= 2 ·dhk ·sin(θhk).(2.10)
The Bragg equation is the basic equation of diffractometry. It ties together the wavelength
λ, the lattice plane spacing dhk and the glancing angle θhk between the incident wave and
the lattice plane dhk;nis an integer and its physical interpretation is the interference
order. Thus at a given wavelength first-order reflection of X-rays by lattice planes occurs
only at a certain angle θhk. If the wavelength is known, the interplanar spacing dhk can
be calculated by measuring of the glacing angle θhk. For its determination in this work
diffraction in small angle region is applied. In small angle scattering the scattering angle
2θlies in range of 0.05◦<2θ < 5◦.
X-ray scattering at ordered crystal lattices represents a special case of the scattering at
an array of scattering centers. The origin of X-ray scattering is illustrated in Figure 2.8.
2.5. SMALL-ANGLE X-RAY DIFFRACTION (SAXD) 23
l tice ane
at pl
q
2q: scattering angle
q: glancing angle
kf
q
2q
ki
Figure 2.8: The origin of X-ray scattering and the relation between glacing angle and
scattering angle. ki: incident wave, kf: final wave, q: scattering vector.
X-ray scattering takes place if the electrons of the scattering particle oscillate in res-
onance with the frequency of the incident X-ray and after this interaction a coherent sec-
ondary X-ray wave is emitted. The wave vector of a wave with the wavelength λis defined
by
|−→
k|=2π
λ(2.11)
Elastic scattering represents an interaction between X-rays and the sample in which no
energy transfer occurs but an impulse exchange of ∆−→
p=~(−→
kf−−→
ki)takes place. The
term −→
q= (−→
kf−−→
ki)is referred to as scattering vector (see Fig. 2.8). In elastic scattering
the normal of the wave vector remains unchanged after the interaction process, i. e.:
|−→
kf|=|−→
ki|=2π
λ(2.12)
Combination of this relation with Equation 2.10 results in the following important rela-
tions between the wave vector, glancing angle and interplanar spacing dhk.
qhk =4π
λ·sinθhk =2π
dhk
(2.13)
In MCM-41 and SBA-15, the individual crystal domains contain approximately 100 and
1000 pores, respectively. The Bragg-reflexes are determined by the type of crystalline
system and the lattice parameter a0of the unit cell. For a 2D hexagonal lattice, the lattice
parameter is obtained by the following equation from the interplanar spacing
a0=2
√3·dhk ·√h2+hk +k2.(2.14)
Information about the pore diameter D and the nature of the pore walls can be determined
by SAXD structure modeling. This can be achieved by applying the so called-continuous
24 CHAPTER 2. EXPERIMENTAL BACKGROUND
density function (CDF) technique [23] in combination with the derivative difference min-
imization (DDM) method [24]. Alternatively, the intensities of the Bragg-reflections may
be fitted by modeling the form factor of the cylindrical pores in an appropriate way
[25,26]. Both of these techniques yield quantitative information about the pore diam-
eter Dand the wall thickness w along the pore center-to-center line, which are related to
the lattice parameter a0by a0=D+w.
2.6 Nitrogen adsorption
The physisorption of gases at solid surfaces and in porous solids represents a classical
method for characterizing the surface properties and porosity of the materials [27]. It is
based on weak interactions of the gasses with the material which in simple cases leads to
reversible adsorption. Due to the weak interaction, substantial adsorption takes place only
at relatively low temperatures, typically at or below the normal boiling temperature Tbof
the fluid. Nitrogen (Tb= 77.3K) is used as the adsorptive gas in many routine studies.
Typically, an adsorption isotherm is measured by increasing the pressure in small steps up
to the saturation pressure. Subsequently a desorption isotherm is measured by gradually
decreasing the pressure in the system.
Analysis of an adsorption isotherm
The form of adsorption isotherms strongly depends on the structure of the adsorbent
and on the interaction between adsorbent and adsorptive. Adsorption isotherms are clas-
sified into six different types by IUPAC. Mesoporous samples including SBA-15 and
MCM-41 often show an isotherm of type IV [28]. In this type of isotherms pore con-
densation of the fluid occurs. In some cases, such as SBA-15, this is accompanied by a
hysteresis, i.e. pore evaporation takes place at a lower pressure than pore condensation.
Three important quantities can be obtained by analysis of the adsorption isotherm of a
porous material:
1. Pore size distribution and mean pore size
2. Specific surface area (BET area)
3. Specific pore volume.
In Figure 2.9 a typical adsorption isotherm of nitrogen in a SBA-15 silica is shown. Also
indicated in the figure are the pressure ranges from which the different quantities are
2.6. NITROGEN ADSORPTION 25
0 , 0 0 , 2 0 , 4 0 , 6 0 , 8 1 , 0
100
200
300
400
500
600
700
a d s o r b e d v o l. / c m 3g- 1 ( S T P )
r e l a t i v e p r e s s u r e
a d s o r p t i o n
d e s o r p t i o n
S B A - 1 5
p o r e s i z e
t o t a l p o r e
v o l u m e
B E T s u r f a c e a r e a
Figure 2.9: Analysis of a typical nitrogen adsorption isotherm for a SBA-15 silica.
derived. The calculation of these quantities from the experimental isotherms is explained
in detail in the following sections.
2.6.1 Specific surface area
An important quantity, which can be obtained from nitrogen adsorption isotherms is the
specific surface area asin m2g−1of the porous solid. It is calculated according to the
theory of Brunauer, Emmet and Teller (BET). They developed an equation from which
the amount of the gas necessary for the formation of a complete monolayer at the surface
can be determined. In gas-volumetric measurements this amount is often expressed by the
respective gas volume Vmono under standard conditions (STP). By measuring the amount
of adsorbed gas Vads as a function of relative pressure p/p0the typical adsorption curve
is obtained which can be described by the BET equation. After linearization the BET
equation has the form
p/p0
VAds(1 −p/p0)=1
Vmono ·C+C−1
C·Vmono ·p
p0
.(2.15)
Here Cis the scalar BET constant which is dependent on the isotherm shape. It character-
izes the strength of the interaction between the gas molecules and the solid surface. Large
values of C(C>100) indicate relatively strong adsorption energy of molecules in the first
adsorption layer with regard to the condensation energy of the adsorptive. According to
the BET equation a linear relation is obtained if p/p0
VAds(1−p/p0)is plotted against p
p0(BET
plot). In this manner it is possible to obtain the parameters Vmono and C. The range of
26 CHAPTER 2. EXPERIMENTAL BACKGROUND
linearity of the BET plot is always restricted to a limited part of the isotherm usually to
the region 0.01 < p/p0<0.3.
The BET specific surface area is calculated from Vmono according to following formula
as=Vmono ·σ(N2)·NA.(2.16)
Here NAis the Avogadro constant, and σ(N2)is the average area (molecular cross-
sectional area in m2/molecule) occupied by each adsorbed molecule in a complete mono-
layer. It is usually assumed that BET nitrogen monolayer is closed-packed, giving 0.162
nm2at 77 K. This value of σis used in the present work for the calculations of specific
surface area of the silica materials. However, it must be kept in mind that a constant value
of σ(N2)is unlikely and that caution needs to be exercised in dealing with surfaces which
give rise to either especially strong or weak adsorbent-adsorbate interactions that are able
to influence the packing.
In case of porous solids the BET method yields the whole surface of the solid, i.e.,
the external plus the internal surface. For determining the mesopore surface area it is also
necessary to take into account the presence of micropores. In this case the micropore
surface area can be determined, for example, by the t−plotmethod [29,30] and can be
subtracted from the whole surface area. Details of the calculation of microporosity by
t-plot method will be presented in the following section.
2.6.2 Analysis of microporosity
Microporosity can be assessed by tracing comparison plots with reference isotherms, usu-
ally called t-plots. This method of isotherm data analysis was introduced by de Boer. It
assumes that in a certain isotherm region, the micropores are already filled whereas the ad-
sorption in larger pores occurs according to some simple equation characteristic for a large
class of solids. This equation should approximate adsorption in mesopores, macropores
and on a flat surface in a narrow pressure range just above complete filling of micropores,
but below critical vapor pore condensation pressure in mesopores. The adsorption within
this pressure region may be described by a simple linear dependence
V(p/p0) = Vmicro +k·aext ·t(p/p0).(2.17)
Here V(p/p0)is the adsorbed volume, Vmicro is the maximum adsorption in micropores,
kis a coefficient which depends on the units used for the values of adsorption and aext is
the external surface area, i.e., the surface area of pores larger than micropores. t(p/p0)is
2.6. NITROGEN ADSORPTION 27
the statistical thickness of adsorbed layer in meso- and macropores which is estimated by
the Harkins and Jura (HJ) [31] or by the Frenkel-Halsey-Hill (FHH) [32] equations. The
t-layer required for the analysis of microporosity is calculated by the HJ-equation in the
following form
t(p/p0)/nm = 0.1s13.99
0.034 −log(p/p0).(2.18)
This equation is commonly applied to a range of relative pressure from 0.1 to 0.2, cor-
responding to tvalues from 0.354 to 0.5 nm. Thus, according to the Equation 2.17 the
t-plot analysis consists of plotting the adsorbed volume V as a function of the calculated
film thickness tas determined by Equation 2.18. In the absence of micropores, when a
multilayer of adsorbate is formed unhindered on the solid surface, the adsorbed volume V
is a linear function of tpassing through the origin. The total surface area is then given by
the slope of the t-plot. In the presence of micropores, the graph of Vvs. twill be linear
at higher pressures (higher t) and an extrapolation of this linear region to t= 0 will give
a positive intercept corresponding to the micropore volume. In this case, negative devia-
tions from linearity of V(t)will occur at lower t. On the other hand, positive deviations
from linearity will occur at higher relative pressure (higher t-values) in case of capillary
condensation in mesopores. In this case the y-intercept gives the micropore volume and
the slope gives the ‘external’ area which is equal to mesopore +macropore area. Here
the micropore area corresponds to the difference of the whole specific surface area and
external area.
0 , 0 0 , 2 0 , 4 0 , 6 0 , 8
0
100
200
300
M C M - 4 1
a d s o r b e d v o l . / c m 3g- 1 ( S T P )
t / n m
S B A - 1 5
Figure 2.10: Comparison of t-plots of MCM-41 and SBA-15 silica.
In Figure 2.10 t-plots for MCM-41 and SBA-15 materials are displayed. In both cases
28 CHAPTER 2. EXPERIMENTAL BACKGROUND
the graph shows a positiv y-intercept indicating microporosity of the samples. However,
the intercept for SBA-15 (37 cm3g−1) is nearly 7 times higher than that for the MCM-41
(5.6 cm3g−1) sample. These values correspond to micropore volumes of 0.057 cm3g−1
and 0.009 cm3g−1for SBA-15 and MCM-41, respectively. It is well known that SBA-
15 materials have a higher microporosity than MCM-41. The micropores in SBA-15
arise from comparatively strong interactions of EO blocks of block copolymer with silica
precursor TEOS. During the synthesis the EO blocks penetrate into the silica matrix and
remain partially in it even after the aging. These leave small holes (micropores) in the
silica wall after burning out the organic material by calcination at 550◦C.
2.6.3 Specific pore volume
In contrast to adsorption on flat surfaces, the amount of the fluid adsorbed in pores is
limited by the volume of the pore space. When neglecting adsorption at the outer surface
of the porous particles, the amount of adsorbed gas in the plateau region of the isotherm
represents the total pore volume of the sample. In the absence of micropores this is also
called the mesopore volume. But the adsorption isotherm often exhibits a further increase
above the pore condensation pressure step. In this case, the specific pore volume can be
roughly estimated from the amount of adsorbed gas directly after the pore condensation
step.
In nitrogen adsorption measurements by the gas volumetric method the amount of
adsorbed nitrogen is indicated as gas volume under standard conditions (STP) in cm3
g−1. STP stands for Standard Temperature and Pressure (273.15 K and 1 bar). Under
these conditions nitrogen is a nearly ideal gas. The adsorbed gas volume of the nitrogen
can be converted into the volume of liquid at the normal boiling temperature (77.3 K) by
using the following equation:
Vl=VSTP ·M·P
ρ·R·T=VSTP ·1.547 ·10−3cm3g−1.(2.19)
Here Pand Tdenote the standard pressure and temperature, respectively, Ris the gas
constant, Mis the molar mass of a nitrogen molecule, and ρis the density of liquid
nitrogen equal to 808.6 kg m−3.
2.6.4 Pore size
A typical phenomenon for the adsorption of fluids in mesoporous materials is pore con-
densation. It is characterized by a condensation in the pore at a relative pressure p/p0<1
2.6. NITROGEN ADSORPTION 29
which is characteristic of the pore size. Pore condensation in pores of uniform size and
shape is quantitatively described by the Kelvin equation. This equation applies only for
mesopores, i.e., pores with a pore diameter bigger than 2 nm. For cylindrical pores the
Kelvin equation has the form
RT ln(p/p0)pc =−2γVMcos θ
rp
.(2.20)
This equation relates the pore condensation pressure (p/p0)pc for a certain pore radius
rpwith macroscopic properties of the fluid, namely the surface tension γ, molar volume
of liquid adsorptive VMand the and contact angle θof the fluid with the pore wall. For
nitrogen at its boiling temperature and standard pressure, γ=8.85 mN m−1and VM=34.71
cm3mol−1. According to Equation 2.20 the pressure at which the pore condensation oc-
curs is a function of pore radius. The smaller the radius rpthe smaller is relative pressure
(p/p0)pc at which the pore condensation takes place. For complete wetting of the wall the
so called Kelvin radius can be determined by the following equation:
rK/nm =−0.415
log(p/p0)pc
.(2.21)
In the case of complete wetting, at the onset of pore condensation the pore walls are
covered with an adsorbed film of thickness t, so that the Kelvin radius is smaller than the
real pore radius. Such a correction is considered in the modified Kelvin equation, viz.
r(p/p0) = rK+t(p/p0).(2.22)
In order to determine the effective pore size the thickness tof this adsorbed film must
be added to the Kelvin radius. This so-called statistical film thickness is dependent on the
pressure and is usually determined by gas adsorption measurements on chemically similar
but flat (nonporous) surfaces. Several empirical approaches for determining the statistical
film thickness have been proposed [33]. Nevertheless all of them are tainted with uncer-
tainty and none of them can be used universally for any adsorbent. For this reason no
singular method exists for determining the correct pore radius. When indicating a pore
diameter it is necessary to specify the method through which the tlayer was calculated.
For ordered masoporous solids like MCM-41 the pore diameter can be estimated from
the lattice parameter a0and the specific pore volume vpby the relation
D=s2√3
πa0rvpρs
1 + vpρs
(2.23)
30 CHAPTER 2. EXPERIMENTAL BACKGROUND
where ρsis the matrix density of silica ρs= 2.2g cm−3. Based on values of D obtained by
this relation and the respective pore condensation pressures (p/p0)pc for a set of MCM-41
materials, Kruk et al. [34] obtained an empirical correlation known as the KJS relation. It
has the form
D=−a
log(p/p0)pc
+ 2t+c. (2.24)
where a= 2 ·rKand cis the constant that fits Equation 2.24 to the experimental depen-
dence of Dversus (p/p0)pc obtained for the MCM-41 samples studied; tis the statistical
film thickness adopted from Harkins-Jura equation [31] which used in following form:
t(p/p0)/nm = 0.1 ( 60.65
0.03071 −log(p/p0))0.3968.(2.25)
which quite accurately represents the film thickness of adsorbed nitrogen on flat surfaces
in a pressures range from 0.1 to 0.95. Later, Jaroniec and Solovyov (2006) used X-ray
diffraction modeling (see Section 2.5) to determine the pore diameter of SBA-15 materials
independently of the adsorption measurements. As in the work of KJS [35] they obtained
a set of data D(p/p0)pc and fitted these data by a relation
D=−A
log(Bp/p0)pc
+ 2t+C(2.26)
where A, B, C were obtained by data fitting (best fit parameters: A= 1.15,B= 0.875
and C= 0.27) and tis given by Equation 2.25.
Hence the imp. KJS equation has following form:
D/nm =−1.15
log(0.875p/p0)pc
+ 0.2 ( 60.65
0.03 −log(p/p0)pc
)0.397 + 0.27.(2.27)
With the aid of this expression the pore size of hexagonal ordered mesoporous materials
with pore condensation pressure between 0.1 and 0.8 can be described very well.
2.7. THERMOGRAVIMETRIC ANALYSIS (TGA/DTA) 31
2.7 Thermogravimetric analysis (TGA/DTA)
Thermogravimetric analysis, (TGA) is an analytical technique used to determine a mate-
rial’s thermal stability and the fraction of volatile components in the material by monitor-
ing the weight change that occurs as a sample is heated [36]. In this method the weight
of the sample is recorded as a function of increasing temperature. For the determination
of thermal decomposition (pyrolysis) of a sample the measurement is carried out in an
inert atmosphere such as helium, argon or nitrogen, whereas for examination of thermo-
oxidative degradation (oxidation) of a sample the measurement is performed in oxygen or
air atmosphere.
-1,0
-0,8
-0,6
-0,4
-0,2
0,0
200 400 600 800 1000
80
85
90
95
100
M3
M2
TP3
TP2
TG(%)
T/°C
DTG
TP1
M1
Figure 2.11: Analysis of a typical TG curve (full line) exhibiting three overlapping steps
by means of the DTG signal (dashed line). M1,M2and M3are weight losses during
the steps 1, 2 and 3; TP1,TP2and TP3are the peak temperatures of DTG curve for the
respective processes.
Many thermal changes in materials (e.g. phase transitions) do not involve a change
of mass. For this reason, in addition to weight changes, some instruments also record
transient temperature differences between the sample and one or more reference pans as a
function of temperature (differential thermal analysis, DTA). When the sample undergoes
a physical or chemical change, the temperature increase differs between the inert reference
and the sample, and a peak or a dip is detected in the DTA signal.
Typical heating rates employed in TGA measurements with silica materials are in the
5-10 K min−1range. Owing to closely spaced or overlapping the weight changes multiple
step TGA curves often show no regions of a constant weight. In this case the differenti-
32 CHAPTER 2. EXPERIMENTAL BACKGROUND
ated signal dm/dt (where m is the sample mass) delivers secondary information. Figure
2.11 shows a TGA curve with a three-step weight decrease. Here the border between
two successive steps on the TG curve can be determined by the minimum value of the
DTG curve between these two steps. In this case the silica materials were studied up to
1000◦C, and the TGA curve is analyzed up to 700◦C. Above this temperature, due to the
condensation of isolated silanol groups, the amount of the siloxane bridges (Si −O−Si)
is increased leading to a further small loss of mass.
Part I
Ordered Mesoporous Silicas:
Self-assembly and Functionalization
33
Chapter 3
State of Knowledge
Ordered mesoporous silica materials of the type MCM-41 (Mobil Composition of Matter)
[3] and SBA-15 (University of California at Santa Barbara) [4] are synthesized by a coop-
erative assembly of surfactant or block copolymer micelles and associated silica species
in aqueous media. The products are disordered on the atomic scale but ordered on the
mesoscopic scale. Due to their well-defined mesopore structure and high surface area
these materials have gained much interest.
During the formation of such materials the matching of charge density at the surfactant
(S)/inorganic (I) interfaces governs the assembly process. In the case of the 2D hexagonal
formed MCM-41 the assembly process is controlled by electrostatic complementarity be-
tween the anionic inorganic ions in solution and the positively charged surfactants (S+I−).
The isoelectric point of silica is at pH 2. Basic conditions are therefore necessary to form
negatively charged silicate particles and to promote the hydrolysis of silica precursor
tetraethylorthosilicate (TEOS). The use of tetramethylorthosilicate (TMOS) is possible
as well. It can even be more beneficial in some advance preparations such as function-
alization of MCM-41 using organoalkoxysilanes. Namely, TMOS hydrolyzes faster than
TEOS does, due to the steric hindrance at ethoxide moieties and reduced solvation of re-
sulting ethanol [37]. The anionic surfactant cetyltrimethylammoniumbromid (CTAB) is a
quaternary ammonium compound with a long chain alkyl group. The pore size of MCM-
41 materials can be varied by changing the alkyl chain length of this cationic surfactant.
MCM-41 with 2D hexagonally ordered pores can be obtained in a chain length range
between C10 and C18. On the other hand, non-ionic amphiphilic polyethylene oxide-
polypropylene oxide-polyethylene oxide (PEO-PPO-PEO) triblock copolymers such as
P123, P103, etc. are used as structure-directing agents for the synthesis of SBA-15. The
formation of SBA-15 in acidic media, above pH=2, occurs by a pathway involving hy-
35
36 CHAPTER 3. STATE OF KNOWLEDGE
drogen bonding interactions among block copolymers and protonated silica species. The
main interaction for the self assembly of the polymer micelle and silica precursor should
be the (S0H+X−I+) type interaction, where X−is the counter ion. Under acidic condi-
tions, the PPO block is more hydrophobic than the PEO block upon heating from 35◦C to
80◦C. Accordingly, hydrophilic silicate species interact preferentially with the more hy-
drophilic PEO block. In the absence of sufficiently strong hydrogen-bonding interactions
at pH values 2-7 no precipitation occurs and amorphous or otherwise disordered silica are
formed. However, Kim et al. succeeded in preparing SBA-15 over a wide range of pH
(0-9) by controlling the relative rates of hydrolysis and condensation of silica species us-
ing fluoride and TMOS [38]. The authors suggest that fully hydrolyzed monomeric silica
species are needed for an attractive interaction with the block copolymer template for ob-
taining ordered mesostructures. This is due to a competition between the condensation of
partially hydrolyzed silica species and the hydrolysis of alkoxysilane moieties associated
with the silica precursors. Above pH 4, silica oligomers may contain the organic moieties
from incomplete hydrolysis due to relatively rapid condensation compared to the rate of
hydrolysis. The presence of such residual organic moieties leads to weaker interactions
between hydrophilic block polymer and silica oligomers, resulting in poorly organized
mesocomposites. So the reaction mixtures above pH 4 without fluoride yield gel-like
precipitation and disordered or amorphous silica structures. On the other hand, addition
of fluoride results in the formation of white precipitates and well ordered mesostructures
up to pH 9. This indicates that hydrolysis of TMOS can be completed before significant
condensation of the silica species occurs, which is consistent with the catalytic activity of
fluoride for hydrolysis. The more rapid hydrolysis of TMOS thus makes it a preferable sil-
ica precursor to TEOS for producing ordered mesostructures. Nevertheless, such ordered
mesostructures can be produced using TEOS, provided the silica sol is pre-hydrolyzed
near pH 2 where the hydrolysis rate is fast and the condensation rate is slowest. It has
also been reported that the addition of inorganic salts lead to a better order in mesostruc-
tured systems [39,40,41]. This is attributed to specific interactions between non-ionic
amphiphiles and metal ions [42]. In this manner, Li et al. were for example able to
synthesize highly ordered Fe-SBA-15 materials under weak acidic conditions [41].
SBA-15 materials have thicker pore walls than MCM-41 materials making them hy-
drothermally more stable. Higher temperatures or longer reaction times result in larger
pore sizes and thinner pore walls. This may be caused by protonation or temperature-
dependent hydrophilicity of the PEO block under the acidic synthesis conditions, or a
combination of both effects. Resulting moieties are expected to interact more strongly
37
extracted
or calcine
Si OR
RO
RO
OR
(OR)3Si
template
(OR)3Si
template extract
grafting route
co-condensation route
Figure 3.1: Schematic representation of the co-condensation and grafting routes for func-
tionalization of ordered mesoporous silica materials [6].
with the silica species and thus are more closely associated with the inorganic wall than
the more hydrophobic PPO block. At higher temperatures however, the PEO blocks be-
come more hydrophobic, resulting in increased hydrophobic domain volumes, smaller
lengths of PEO segments associated with the silica wall, and increased pore size.
While the structural properties of MCM-41 and SBA-15 are interesting for many po-
tential applications involving host-guest interactions, it is often necessary to modify the
surface composition of these materials, e.g., by attaching inorganic and organic compo-
nents to the surface. Organic functionalization can be achieved either by covalently graft-
ing of various organic species to surface silanol groups, or by incorporating functionalities
directly during the preparation.
The first approach, the grafting process, has been widely employed to anchor specific
organic groups onto surface silanols of a variety of silica supports. Typically, hydrolyz-
able moieties such as RSiX3, where X=Cl, OCH3or OCH2CH3, are used as silica precur-
sors for the surface functionalization. Moreover, different disilazane reagents NH(SiR2)
can also be used for grafting in a gas phase reaction, as done in this work. Due to the
relatively week N-Si bonds, these reagents are unstable at higher temperature and decom-
pose releasing two functional group and NH3. However, the grafting method has several
drawbacks:
38 CHAPTER 3. STATE OF KNOWLEDGE
1. It is difficult to control the extent of the functionalization and the position of the
anchored functional groups in an adequate way. The grafting rates depend on the
reactivity of the precursors, diffusion limitations, steric factors and accessibility of
surface silanol groups. For these reasons, variable and often rather low loadings are
obtained. In this context, Brunel at al. reported that calcination of micelle templated
silicas produces hydrophilic and hydrophobic zones on the surface and that grafting
occurs only on the hydrophobic zones, leading to clustering of the functional silane
even at relatively low loadings [43].
2. Post-synthesis grafting is time-consuming, as two steps are needed to accomplish
the goal. In particular, the steps required to obtain the functionalized product are
extensive, often requiring thorough drying of both the silica and the reaction sol-
vent before the grafting reaction in order to avoid the formation of unwanted by-
products.
3. To reach a high degree of modification it is often necessary to employ a large excess
of functional silane.
4. Post-synthesis modification may lead to reduced pore size due to the attachment of
a layer of the functional moiety on the surface.
These drawbacks my be overcome by a direct single-step process, i.e. incorporation of or-
ganic groups during the synthesis of the mesoporous silica by co-condensing siloxane and
organosiloxane precursors in a templating environment. It is believed that this approach
enables a higher and more homogeneous surface coverage of organosilane functionali-
ties [44], although little direct evidence is available for the latter point. Evidence based
on the kinetics of bromination of grafted and co-condensated vinyl-MCM-41 reported by
Lim et al. indicates that vinyl groups are located predominantly on external surfaces and
around the pore openings of grafted materials, but more homogenously distributed in ma-
terials prepared by co-condensation. Moreover, Fiorilli et. al studied the reactivity of the
carboxylic acid (CA) groups of CA-SBA-15 synthesized by co-condensation method in
reactions with ammonia in gas phase by following changes in the IR-spectra [45]. They
concluded that the acid groups are well-dispersed, not clustered and behave independently
of one another. In that work the materials were synthesized with a F/(F+TEOS) mole ratio
of 10%. But the synthesis of well-ordered SBA-15 materials with 20 mole % is possible
as well. However, to the best of our knowledge, similar investigations for silica samples
which are loaded with functional silane higher than 10 mole % in the initial synthesis
mixture are not existent yet.
39
On the other hand, Yokoi et al. provided evidence suggesting that some functional
silanes may in fact be contained within the pore walls of MCM-41 materials prepared
using the co-condensation method [46]. The authors found that materials synthesized
using 3-aminopropyltrimethoxysilane, 1-(2-aminoethyl)-3-aminopropyl trimethoxysilane
and 1-[3-(trimethoxysilyl)propyl]diethylenetriamine as functional silanes contained sim-
ilar amounts of nitrogen, but that the accessibility of the amine centers to metal cations in
solution decreased as the length of the organic chain increased. By comparing the results
from the elemental analyses and the argentometric titrations of the surfactant-extracted
samples, they concluded that not all amino moieties incorporated were present on the
surface, but some of them were located in the walls of the hexagonal channels.
Another significant difference between grafted functionalities and functionalities gen-
erated by the co-condensation route is their stability. Materials obtained by post-synthesis
modification are less stable than those prepared by co-condensation. This comes from the
fact that the of Si-O-C bonds formed in the grafting process are readily cleaved at elevated
temperatures. By contrast, in hybrid materials prepared by co-condensation method the
functional group is attached to the matrix by Si-C bonds. So co-condensation results in
very strong unhydrolyzable Si-C bonds which are substantially more stable than Si-O-C
bonds.
The co-condensation method was first reported by two research groups in 1996 [44,
47] for the functionalization of MCM-41 materials. A variety of functional groups have
been incorporated into mesoporous materials such as aliphatic hydrocarbons [44,48,49,
50], thiol groups [51,52,53], vinyl groups [54,48,55], phenyl groups [44,49,48], amine
groups [56,57,50], etc. Most of these works have been conducted with M41S designated
materials under basic synthesis conditions. These functionalized materials have potential
for applications in different areas such as adsorption and separation processes, enzyme
immobilization and catalysis.
An immobilized enzyme attached to an inert, insoluble material, provides increased
resistance to changes in conditions such as pH or temperature. Immobilization also allows
enzymes to be held in place throughout the reaction. After completion they are easily sep-
arated from the products and may be used again. It is a far more efficient process and so
is widely used in industry for enzyme catalyzed reactions. For example, the amine, vinyl,
thiol and carboxylic acid moieties are important functionalities for enzyme immobiliza-
tion on porous supports. Chong et al. have found that vinyl functionalized SBA-15 is a
candidate as a support material for penicillin acylase because of its high enzyme uptake
and high initial activity of the biocatalyst [58]. Similarly, studies of Yiu et al. have shown
40 CHAPTER 3. STATE OF KNOWLEDGE
that trypsin immobilized on SBA-15 functionalized by thiol and carboxylic acid moieties
displayed very high specific activities [10,59].
On the other hand, there has been increasing interest in the development of heteroge-
neous solid acid catalysts to avoid the use of traditional homogeneous acid catalytic sys-
tems (H2SO4, HF, AlCl3, BF3, etc.) which present serious drawbacks including hazards
handling, production of toxic waste, corrosiveness, and difficulties in separation. It has
been shown that the type of support material used is a critical factor in the performance of
the resulting supported catalyst. Two main factors must be considered when employing
a material as a support. First, the material needs to be stable, both thermally and chem-
ically, during the reaction process. Secondly, the structure of support needs to be such
that the active sides are well dispersed on its surface in order that these sites are easily
accessible. Aluminium substituted MCM-41 has been tested in acidic catalysis, though it
shows week acidity in comparison to conventional acid zeolites. As an alternative, the co-
valent attachment of alkyl sulfonic acid groups to the surface of MCM-41 and other HMS
(hexagonal mesoporous silica) molecular sieves has been proposed by several groups and
successfully tested in several acid catalyzed reactions, including esterification, and con-
densation [52,60,61]. In post-synthesis grafting, 3-mercaptopropyltrimethoxysilane has
been used as the key precursor, and the thiol groups were treated with H2O2to obtain
sulfonic groups. Margolese et al. used the same method to synthesize propyl sulfonic
acid functionalized SBA-15 by the co-condensation route. Subsequently, Melero et al.
reported the synthesis of arenesulfonic acid containing SBA-15 having increased acidity
compared to propyl sulfonic acid functionalized SBA-15 [62]. These materials proved to
be good catalysts for a variety of reactions [8] such as alcohol coupling to ethers [60],
Friedel-Crafts-acylation of aromatic compounds [63], Fries rearrangement of phenyl ac-
etate [64], Beckmann rearrangement of cyclohexanone oxime to -caprolactam [65] and
etherification of benzyl alcohol [66]. The catalytic activities of HMS materials function-
alized with phosphonic acid have been tested as well. Kawi et al. found that H3PO4-
MCM-41 synthesized by impregnation of H3PO4in MCM-41 through simply mixing is
a promising selective catalyst for specific reactions that need only Brønsted acidity such
as dehydration of isopropanol to propylene [67]. Le et al. immobilized tungsten species
onto the surface of SBA-15 functionalized by diethylphosphatoethyltriethoxysilane. This
catalyst shows high conversion and selectivity in the selective oxidation of cyclopentene
to produce glutaraldehyde using H2O2as the oxidant, and can preserve high activity after
several cycles of reaction [68].
41
In the present work, functionalized silica materials were synthesized both by co-
condensation and by the grafting route. In Chapter 5, the preparation of MCM-41 and
SBA-15 materials with hydrophobized pore walls by the grafting route is described using
(CH3)2Si(OCH3)2and (CH3)Si(OCH3)3reagents. These materials were used for investi-
gating the surface characteristics of MCM-41 and SBA-15 silicas. For this purpose solid-
state 29Si NMR methods were employed to obtain the ratio of numbers of single-, double-,
and triple surface-bound species for each reagent. These results ratio provide qualitative
information about the arrangement of the surface silanol groups at the pore walls from
which the surface properties of MCM-41 and SBA-15 silicas can be gathered. Chapter
6 deals with the synthesis of SBA-15 materials containing different acid groups inside
the pores prepared by the co-condensation route in acidic medium. SBA-15 functional-
ized with carboxylic acid, phosphonic acid and sulfonic acid groups were prepared by
this route. Several parameters, such as amount of the functional silane and the sequence
of prehydrolyzed component (TEOS or functional silane), were varied and the resulting
material were characterized by various methods. The acidity of the acids in the aqueous
solutions increases in the order carboxylic acid, phosphonic acid and sulfonic acid. Our
aim was to examine of the acidity inside the mesopore channels of SBA-15 materials con-
taining the aforementioned acid groups in its walls. The proton donor ability of these acid
sides were studied by 15N CPMAS NMR after pyridine adsorption into the pores.
42 CHAPTER 3. STATE OF KNOWLEDGE
Chapter 4
Materials and Methods
4.1 Materials
The synthesis of ordered mesoporous silica materials of the SBA-15 and MCM-41 type is
accomplished with organic structure-directing agents (templates) and proceeds in the way
illustrated in Figure 4.1. Template molecules of different kind are used in the synthesis
of the two silicas. In the case of SBA-15, nonionic amphiphilic block copolymers are
employed. Due to the stronger interactions between silica precursor TEOS and nonionic
block copolymer at a pH below the isoelectronic point of silica (pH∼2) the synthesis
is performed in acidic medium with pH<2. In case of MCM-41, CnTAB type cationic
surfactants are used, where n is the number of carbon atoms in the hydrophobic chain.
In this case the reaction is performed in basic media, above the isoelectronic point of
silica, because of the cationic character of template molecules. In the following section
the synthesis of these silica materials is described in detail.
silica precursor
(TEOS)
micellar solution
of template
hydrothermal
treatment
„aging“
calcination
as-synthesized
silica
SBA-15 or
MCM-41
550°C
silica precursor
(TEOS)
silica precursor
(TEOS)
micellar solution
of template
hydrothermal
treatment
„aging“
calcination
as-synthesized
silica
SBA-15 or
MCM-41
550°C
micellar solution
of template
hydrothermal
treatment
„aging“
calcination
as-synthesized
silica
SBA-15 or
MCM-41
550°C
Figure 4.1: General route for SBA-15 and MCM-41 synthesis.
43
44 CHAPTER 4. MATERIALS AND METHODS
4.1.1 Pure SBA-15
SBA-15 was prepared by the method reported in ref. [4] using technical grade poly(ethyl-
ene oxide)-poly(propyleneoxide)- poly(ethylene oxide) triblock copolymer (Pluronic P123,
BASF USA, Mount Olive, NJ) as the structure-directing agent. The molar composition
of the reaction mixture was 1 TEOS : 5.9 H2SO4: 323 H2O : 0.017 P123. In a typical
synthesis 13 ml of H2SO4(97%) was added to a solution of 4 g of P123 in 240 ml milli-Q
water to adjust pH <2. This solution was transferred to a glas bottle with SL screw cap
(1000 ml) and 9.2 ml tetraethoxysilane (TEOS, 99%, ABCR) was added under vigorous
stirring at 40◦C. The polymer-silica composite, formed as a fine precipitate, was kept in
the reaction solution at 40◦C for 20 h under constant stirring and then transferred to an
autoclave for aging (20 h at 105◦C). The product was filtered, washed with milli-Q water,
dried at 60◦C and finally calcined in air at 550◦C.
4.1.2 Pure MCM-41
MCM-41 was synthesized as described in ref.[69], using hexadecyltrimethylammonium
bromide (C16TAB) as structure-directing agent (template). The molar composition of the
reaction mixture was 1 TEOS : 5.6 NH3: 248 H2O : 0.122 C16TAB. In a typical synthesis
47 ml of aqueous ammonia (25 wt.%) was added to a solution of 5 g C16TAB (Fluka,
purity 99 %) in 500 ml Milli-Q water. 25 ml TEOS (ABCR, purity 97%) was added to
this solution under vigorous stirring at 40◦C. The surfactant-silica composite was formed
as a fine precipitate after a few minutes and was kept in the reaction solution for 4 h at
40◦C under constant stirring and then transferred into an autoclave for 48 h at 105◦C.
The material was filtered, washed with Milli-Q-H2O and dried at 60◦C for 6 h and finally
calcined in air at 550◦C.
4.1.3 Functionalization by grafting
4.1.3.1 HMDS and DCDMS
Hydrophobization of the pore walls of SBA-15 was performed with hexamethyldisilazane
(HN(Si(CH3)3)2, HMDS) or dichlorodimethylsilane ((CH3)2SiCl2, DCDMS). In both
cases, the modification was carried out by gas phase reaction. 1 g of SBA-15 was weighed
into an open tube which was then placed in the glass flask of the device shown in Figure
4.2. The sample was heated to 120◦C and then evacuated to a pressure of 10−2mbar
for several hours to remove adsorbed water from the pores. Afterwards the liquid reagent
4.1. MATERIALS 45
Figure 4.2: Equipment for the gas phase modification of the silicas [70]: 1small sample
tube, 2oven, 3N2-cryo trap, 4glass tap, 5dosing valve, 6blocking valve, 7oil pumpe, 8
manometer.
(HMDS or DCDMS, both from Fluka) was introduced dropwise into the glass flask where
it evaporates and the vapor reacts with the surface silanol groups of mesoporous silica ma-
terial. The sample was kept in the reaction chamber for 24 h at room temperature (HMDS)
or at 100◦C (DCDMS). The non-reacted excess amount of the reagent was then removed
by heating to 120◦C and pumping at 10−2mbar via a N2-cryo trap.
4.1.3.2 DMDMS and MTMS
Modification of MCM-41 and SBA-15 with dimethyldimethoxysilane ((CH3)2Si(OCH3)2,
DMDMS) and methyltrimethoxysilane (CH3)Si(OCH3)3, MTMS) was performed by liq-
uid phase reactions. 0.5 g of dry MCM-41 or SBA-15 were suspended in dry toluene in
a N2-atmosphere. An excess of the reagents DMDMS or MTMS (from Sigma Aldrich)
was then added slowly through a dropping funnel under constant stirring. The suspension
was then heated to 80◦C under reflux for 24 h. After the reaction the solid material was
vacuum-filtrated and washed each with 25 ml toluene and chloroform. Afterwards the
solid was dried in a drying closet at 80◦C.
4.1.3.3 Condensation of DMDMS and MTMS without silica
Hydrolysis and condensation of DMDMS and MTMS in the absence of silica was per-
formed in water in acidic conditions, using an excess of 2 mole water per mole of methoxy-
groups. The compositions of the reaction system were: 2.0 ml DMDMS (14.6 mmol)
with 1.0 ml (58.4 mmol) of 0.1 M HCl, and 2.0 ml MTMS (14.6 mmol) with 1.5 ml (84.6
46 CHAPTER 4. MATERIALS AND METHODS
mmol) 0.1 M HCl. After 2 days of stirring at room temperature the condensation resulted
in a clear, slightly viscous liquid in case of DMDMS, and a cloudy gel-like product in the
case of MTMS.
4.1.4 Acid-functionalized SBA-15 by co-condensation route
Acid functionalized ordered mesoporous silicas were prepared by a one-pot synthesis
method. In this method, a part of the silica precursor TEOS is replaced by a functional
silane reagent. The molar composition of the systems used in this work was [71]
(1-x) TEOS (T) : x functional silane (F) : 5.8 HCl : 193 H2O : 0.017 P123
where x was chosen between 0.1 and 0.3. The synthesis scheme is illustrated in Figure
4.3. Starting from a micellar hydrochloric acid solution of block copolymer (P123), the
Micellar solution
of P123 in HCl(aq)
functional
silane: F
40°C
TEOS: T
aging (24h)
as-synthesized
functionalized SBA-15
O
O
O
Si
acid
n48% H2SO4
template removal
Prehydrolysis
time
20h (stirring)
40°C
90-100°C
acid doped SBA-15
x F: (1-x): T
)
F
-15
O
O
O
Si
n
O
O
O
Si
2 4 2 4
°
°C
Figure 4.3: One pot synthesis functionalization route of SBA-15.
predetermined amount of functional silane is added, and a certain prehydrolysis time of
that component is allowed at 40◦C under stirring before adding the respective amount of
TEOS. In some cases, TEOS was added first and the functional silane second, in order to
asses the effect of the prehydrolysis. Aging occurred at 90-100◦C for 24 h. After filtering
and washing of the white precipitate as-synthesized functionalized SBA-15 materials were
obtained. They were treated with 48% sulfuric acid solution to remove the template and
produce the final acid containing SBA-15. Freshly prepared diluted H2SO4was used for
the template removal, by slowly diluting in an ice bath under stirring. The use of freshly
4.1. MATERIALS 47
prepared 48% H2SO4was found to be important for effective removal of the sample.
Typically, 1 g of as-synthesized SBA-15 was suspended in 200 ml of 48% H2SO4solution
and refluxed at 95◦C for 24 h under stirring. The treated powders were filtered and washed
extensively, first with acetone and ethanol and later with water until the pH of the eluent
was neutral. Finally the product was dried at 80◦C.
Acid-doping of SBA-15 with carboxylic acid (CA), phosphonic acid (PA) and sulfonic
acid (SA) was performed by this synthesis route. Only three parameters were varied in
the synthesis procedure. These were:
1. Pre-hydrolyzed component of silica precursor mixture, i. e. either the functional
silane (F) or TEOS (T) is prehydrolyzed with the acidic P123 solution.
2. y: Prehydrolysis time of the pre-mixed component expressed in minutes,
3. φ: Molar percentage of functional silane (F) in the silica precursor mixture.
φ=F
F+T·100
Throughout this work the acid-functionalized materials are denoted as
φ-CA-y-F,φ-PA-y-Fand φ-SA-y-F.
The as-synthesized materials are denoted as
φ-CN-y-F,φ-P-y-Fand φ-S-y-F
when the functional silane (F) was the prehydrolized component of the silica precursor.
In materials in which TEOS was used as the prehydrolized component, the Fis replaced
by T. However, in several cases abbreviated notations will be used as explained below.
4.1.4.1 Carboxylic acid doped SBA-15 (CA-SBA-15)
Materials were prepared by the synthesis route described above. Here the optimized syn-
thesis conditions were taken from the work of Yang et. al [71]. The functional silane
2-cyanoethyltriethoxysilane, CTES, (98%, Aldrich) was used in the synthesis of CA-SBA-
15. Molar composition of (1-x) T : x F with x = 0.1 and 0.2 were used (φ= x.100). In
the synthesis of these materials, the functional silane (F) was always used as the prehy-
drolyzed component, and a prehydrolysis time of 60 min was used in all cases. For this
reason an abbreviated nomenclature is used for these materials: The as-synthesized prod-
ucts are called 10-CN and 20-CN, and the respective end products are denoted by 10-CA
and 20-CA. The end products are formed by the hydrolysis of CN-groups through the
H2SO4treatment during the template removal.
48 CHAPTER 4. MATERIALS AND METHODS
4.1.4.2 Phosphonic acid doped SBA-15 (PA-SBA-15)
For the synthesis of PA-SBA-15 materials by the co-condensation route diethoxyphospho-
rylethyltriethoxysilane, PTES, (95%, ABCR) was chosen as the functional silane. Mate-
rials with molar compositions (1-x) T : x F with x = 0.1, 0.15, and 0.2 were prepared.
Either TEOS (T) or the functional silane (F) was used as the prehydrolyzed component.
The prehydrolysis time was varied between 0 and 95 min. The as-synthesized products
are called φ-P-y-F/T. The products containing phosphonate ester after polymer removal
by treatment with H2SO4are denoted as φ-POEt-y-F/T. Phosphonic acid functionalities
were formed by dealkylation of the phosphonate ester groups. To complete the cleavage
of phosphonate ester the samples were fluxed in concentrated HCl. For this purpose, ca. 1
g of the H2SO4treated material was suspended in 300 ml of 37% HCl and refluxed under
stirring at 95◦C for 24 h. After filtering the solid was extensively washed with water until
the eluent was pH neutral. Throughout this work samples containing free phosphonic acid
groups are referred to as φ-PA-y-F/T.
4.1.4.3 Sulfonic acid doped SBA-15 (SA-SBA-15)
For the synthesis of SA-SBA-15 materials the functional silane 3-(trihydroxysilyl)-1-
propane-sulfonic acid, STHS, (30-35% in water, ABCR) was used. Molar compositions
(1-x) : x F with x = 0.1 , 0.2, and 0.3 were used, and both TEOS and F was used as the
prehydrolyzed component. Notice that in case of STHS, the term ‘hydrolysis’ is not really
appropriate because it already exists in the hydrolyzed form. Rather we are dealing with
‘precondensation’ of the functional silane. For the sake of a uniform notation the term
will nevertheless be used. The prehydrolysis time was varied between 0 and 1480 min.
The as-synthesized products are called φ-S-y-F/T, and the products after treatment with
H2SO4are denoted as φ-SA-y-F/T.
4.2 Methods
4.2.1 Nitrogen adsorption
The pore structure of the silica materials was characterized by N2adsorption and small-
angle X-ray diffraction. Adsorption isotherms of nitrogen at 77 K were measured by
gas volumetry using a Gemini 2375 volumetric gas adsorption analyzer (Micromeritics).
The samples were dried and outgassed at 120◦C for 45 min at a pressure below 0.1 mbar
4.2. METHODS 49
and reweighed before the gas adsorption measurement, to determine the net mass of the
sample. For the measurements the standard Micromeritics sample tubes were used. For
thermal isolation of Dewar vessel in which the sample tube is immersed in liquid nitro-
gen, the original top of Dewar vessel is replaced by a styrofoam top. In this way it was
guaranteed that evaporation of the nitrogen was slow enough to maintain the temperature
of the sample constant for the full time of 12 h measurement. Because the barometric
pressure was measured only at the beginning of the adsorption ran subsequent variations
of air pressure and thus the actual saturation pressure p0of nitrogen could not be taken
into account.
4.2.2 SAXD
Small-angle X-ray diffraction (SAXD) measurements were made in the Max-Planck Insti-
tut für Kolloid- und Grenzflächenforschung in Potsdam/Golm. The SAXD profiles were
recorded in a range of the scattering vector q from 0.3 to 2 nm−1by a Bruker SAXS
Nanostar machine, using a 2D HI Star area detector and CuKαradiation (λ=1.54 Å). The
calibration of q scale was performed using silver behenate as reference. For the measure-
ments the mesoporous silica materials are filled in small glass tubes (Mark Röhrchen) of
1 mm diameter and 80 mm length, and a wall thickness of approximately 0.01 mm. After
loading with the powders the tubes were sealed in a flame.
4.2.3 TGA
Weight loss curves due to thermal decomposition in air of functionalized SBA-15 materi-
als was measured on a TG/DTA Netzsch STA 409 simultaneous thermal analysis instru-
ment in the laboratory of inorganic chemistry at TU Berlin. Samples were heated at a rate
of 10 K min−1in a stream of synthetic air.
4.2.4 13C, 29Si and 15N solid-state NMR
13C, 29Si and 15N NMR measurements were performed in cooperation with the group of
Prof. H.-H Limbach at the Chemistry department of FU Berlin using a Bruker MSL-300
instrument operating at 7 Tesla, and equipped with a Chemagnetics-Varian 6 mm pencil
CPMAS probe. All samples were spun at 6-8 kHz under magic angle spinning (MAS)
conditions.
50 CHAPTER 4. MATERIALS AND METHODS
The 29Si MAS spectra were recorded employing a π/12 pulse-sequence and a recycle
delay of 180 s. The {1H}-29Si CPMAS spectra were recorded using a relatively long cross
polarization (CP) contact time (8 ms), which ensures sufficient polarization transfer to all
different Si species of silica, i.e. Q2,Q3and Q4; further parameters were a recycle delay
of 5 s. The 29Si chemical shifts of spectra shown in Chapter 5are referenced to liquid
TMS (Tetramethylsilane), whereas those presented in Chapter 6are referenced to solid
TSP (3-(Trimethylsilyl)propionic acid sodium salt).
The 13C NMR measurements were performed employing the {1H}-13C CPMAS tech-
nique with a CP contact time of 2 ms and a recycle delay of 5 s. The typical 90◦-pulse
length in the CP experiments was 3.5 sec for 1H. The 13C chemical shift values are refer-
enced to solid TSP.
The 15N MAS spectra were recorded employing a π/2 pulse-sequence, with a 90◦-
pulse length of 4.5 µsec. The room- temperature and low-temperature {1H}-15N CPMAS
spectra were recorded using a cross polarization contact time of 5 ms. The typical 90◦-
pulse length was 3.5 µsec for 1H. All 15N chemical shift values are referenced to solid
15NH4Cl. In this scale, the 15N nucleus of pyridine resonates at 275 ppm and that of
protonated pyridinium below 170 ppm.
4.2.5 FT-IR
FT-IR measurements of CA-SBA-15 samples were recorded using a Bruker Tensor 27
spectrometer in the Max-Volmer laboratory of TU-Berlin. For this purpose a thin silicone
spacer (10 mm diameter, 10 µm thick) was placed on a CaF2slide and an appropriate
amount of the silica powder was put into the ring. A small volume of D2O or a phosphate
buffer solution in D2O was then added by a microliter pipette in argon atmosphere. The
suspension was homogenized with a spatula and the excess solution was removed by
waiting in argon atmosphere. Subsequently the cell was closed under argon atmosphere.
4.2. METHODS 51
4.2.6 Potentiometric titration
Potentiometric titrations were performed by a Metrohm Titrando 836 automatic poten-
tiometer. Aliquots of ca. 50-70 mg of acid functionalized SBA-15 samples were added
to ca. 15 g of 2 M NaCl solution and equilibrated for 45 min to assist the proton release
via a cation exchange process. The solution obtained in this way was potentiometrically
titrated against 0.02 M NaOH (in case of SA-SBA-15) or 0.01 M (in case of CA-SBA-15)
solutions. The first derivative curve was used for computation of acidic capacity. Each
titration was performed twice at least. Typical titration curves obtained for the CA-SBA-
15 and SA-SBA-15 materials are shown in the Figure 4.4.
pH=5.08
pH=7.33
CA-SBA-15 SA-SBA-15
Figure 4.4: Titration plots of CA-SBA-15 and SA-SBA-15 (pH vs. VNaOH ).
4.2.7 Scanning electron microscopy (SEM)
SEM images were made in ZELMI at TU Berlin using a Hitachi S-4000 microscope. The
dry silica samples were directly attached to carbon-coated Cu grids by dispersing small
amounts of the powder using a spatula, to avoid contact to any solvents.
52 CHAPTER 4. MATERIALS AND METHODS
Chapter 5
Functionalization by Grafting∗
Abstract
MCM-41 and SBA-15 silicas were studied by 29Si and 15N solid state NMR in the pres-
ence of 15N-pyridine with the aim to formulate generic structural parameters which may
be used as a checklist for atomic-scale structural models of this class of ordered meso-
porous materials. High-quality MCM-41 silica constitutes quasi-ideal arrays of uniform-
size pores with thin pore walls, while SBA-15 silica has thicker pore walls with framework
and surface defects. The numbers of silanol (Q3) and silicate (Q4) groups were found to
be in the ratio of about 1:3 for MCM-41 and about 1:4 for our SBA-15 materials. Com-
bined with the earlier finding that the density of surface silanol groups is about 3 per
nm2in MCM-41 [5] this allows to discriminate between different atomic-scale models of
these materials. The arrangement of Q3groups at the silica surfaces was analyzed using
postsynthesis surface functionalization. It was found that the number of covalent bonds
to the surface formed by the functional reagents is affected by the surface morphology.
It is concluded that for high-quality MCM-41 silicas the distance between neighboring
surface silanol groups is greater than 0.5 nm. As a result, di- and tripodical reagents like
(CH3)2Si(OCH3)2(DMDMS) and (CH3)Si(OCH3)3(MTMS) can form only one covalent
bond to the surface. The residual hydroxyl groups of surface bonded functional reagents
either remain free or interact with other reagent molecules. Accordingly, the number of
surface silanol groups in a given MCM-41 or SBA-15 silica may not decrease but increase
after treatment with CH3Si(OH)3reagent. On the other hand, nearly all surface silanol
groups could be functionalized when HN(Si(CH3)3)2(HMDS) was used.
∗In modified form published as: Shenderovich, I. G.; Mauder, D.; Akcakayiran, D.; Buntkowsky, G.;
Limbach, H.-H.; Findenegg, G. H. J. Phys. Chem. B. 2007,111, 12088
53
54 CHAPTER 5. FUNCTIONALIZATION BY GRAFTING
5.1 HMDS and DCDMS grafted samples
Many potential applications of ordered mesoporous silica result from the surface silanol
groups at the pore walls. The proton donor ability of the surface hydroxyl groups is similar
to that of acids exhibiting a pKa of about 4 in water [5]. Thus, the surface of pure silica
is slightly acidic. The chemical activity of the surface can be selectively modified via the
grafting method leading to novel hybrid materials. The simplest way to convert the weakly
acidic surface of pure silica into a hydrophobic one is to coat it with hexamethyldisilazane
(HMDS) by a simply gas phase reaction in room temperature described in Section 4.1. In
this case the functionalization reagent contains the only reactive group which can interact
with the surface silanols to form Si-O-Si covalent bond as illustrated schematically in
Figure 5.1.
Grafting of dichlorodimethylsilane (CH3)2SiCl2, (DCDMS) onto the surface of SBA-15
+ HN(Si(CH3)3)2
Si OH
Si OH
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
29Si CPMAS NMR, 300K
15N CPMAS NMR, 130K
+ (CH3)2-Si-Cl2
Si OH
Si OH
+ 2 HCl
Si OH
+ H2O
Si OH
Si OH
a
b
cSi O-Si-OH
(CH3)2
+ HN(Si(CH3)3)2
Si OH
Si OH
+ HN(Si(CH3)3)2
Si OH
Si OH
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
29Si CPMAS NMR, 300K
15N CPMAS NMR, 130K
+ (CH3)2-Si-Cl2
Si OH
Si OH
+ (CH3)2-Si-Cl2
Si OH
Si OH
+ 2 HCl
Si OH
+ H2O
Si OH
Si OH
Si OH
Si OH
a
b
cSi O-Si-OH
(CH3)2
Si O-Si-OH
(CH3)2
Figure 5.1: Schematic illustration of the chemical structures of the surfaces before and
after the functionalization.
was also performed by a gas phase reaction at elevated temperatures (Section 4.1). Here
the functional reagent carries two reactive groups. Accordingly, it can form bonds to
either one or two silanol groups.
The functionalization and its efficiency were investigated by 29Si and 15N CPMAS
NMR spectroscopy. Experimental spectra obtained for the pure and modified SBA-15
are presented in Figure 5.2. After modification a line assigned to the functional silane
(QR) appears in the 29Si NMR spectra displayed in Fig. 5.2b and Fig. 5.2c. The level
5.1. HMDS AND DCDMS GRAFTED SAMPLES 55
of modification could be roughly estimated from the 29Si spectrum of the modified silica
by analyzing the reduction of the corresponding Q3peak. For both modified SBA-15
samples the ratio of the Q3:Q4intensities are smaller than for the pure SBA-15 (cf. Fig.
5.2a, Fig. 5.2b, and Fig. 5.2c on the left)
15N CPMAS NMR spectrum of HMDS modified SBA-15 shows a strong reduction
of the amount of pyridine bound to the surface hydroxyl groups via hydrogen bonds, (cf.
Fig. 5.2a and 5.2b right). Therefore, it appears that HMDS causes a nearly quantitative
+ HN(Si(CH3)3)2
Si OH
Si OH
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
29Si CPMAS NMR, 300K
15N CPMAS NMR, 130K
+ (CH3)2-Si-Cl2
Si OH
Si OH
+ 2 HCl
Si OH
+ H2O
Si OH
Si OH
a
b
cSi O-Si-OH
(CH3)2
+ HN(Si(CH3)3)2
Si OH
Si OH
+ HN(Si(CH3)3)2
Si OH
Si OH
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
+ NH3
Si O-Si-(CH3)3
Si O-Si-(CH3)3
29Si CPMAS NMR, 300K
15N CPMAS NMR, 130K
+ (CH3)2-Si-Cl2
Si OH
Si OH
+ (CH3)2-Si-Cl2
Si OH
Si OH
+ 2 HCl
Si OH
+ H2O
Si OH
Si OH
Si OH
Si OH
a
b
cSi O-Si-OH
(CH3)2
Si O-Si-OH
(CH3)2
Figure 5.2: Left: 29Si CPMAS NMR of pure (a) and functionalized (b, c) SBA-15 silica,
Right: 15N CPMAS NMR at 130 K of the same samples loaded with an equal amount
of 15N-pyridine. The functionalization reagents were HN(Si(CH3)3)2, HMDS (b), and
(CH3)2SiCl2, DCDMS (c).
functionalization of the surface. In contrast, the 15N spectrum of DCDMS modified SBA-
15 did not display a reduction of the available hydroxyl groups, (cf. Fig. 5.2a and Fig.
5.2c right). This finding indicates that some of the reactive groups of such reagents do
not form covalent bonds to the surface but are only hydrolyzed. Since the hydrolysis is
the step preceding surface functionalization, we conclude that not all hydrolyzed groups
were able to meet surface hydroxyl groups to form covalent bonds. This fact has been
employed to analyze the arrangement of the surface hydroxyl groups.
56 CHAPTER 5. FUNCTIONALIZATION BY GRAFTING
Table 5.1: Abbreviations of samples used and relative intensities of NMR signals corre-
sponding to surface functional groups of different podality.
sample abbreviation D1: D2T1: T2: T3
DMDMS+HCl DM 5 : 100
MTMS+HCl TM 22 : 100 : 354
MCM-41-DMDMS M-DM 223 : 100
SBA-15-DMDMS S-DM 91 : 100
MCM-41-MTMS M-TM 62 : 100 : -
SBA-15-MTMS S-TM 7 : 100 : 38
S: SBA-15; M: MCM-41, DM: DMDMS; TM: MTMS
5.2 DMDMS and MTMS grafted samples
In order to simplify the comparison between HMDS and reagents having two or three
reactive groups we studied silica samples modified by the simplest possible reagents,
namely dimethyldimethoxysilane (DMDMS) and methyltrimethoxysilane (MTMS). The
preparation of these samples is described in Section 4.1.3. Figure 5.3 shows 29Si NMR
spectra of these reagents after hydrolysis and condensation as well as spectra of function-
alized silica materials (samples: DM, M-DM, S-DM, and TM, M-TM, S-TM). Abbre-
viations of the modified samples used in this chapter are summarized in Table 5.1. The
spectra of the functionalized materials were obtained using the cross-polarization transfer
technique. Thus the relative intensities of different peaks of the same spectrum may not
precisely coincide with the relative concentration of the corresponding species. However,
a long contact time was used and each silicon nucleus carried at least one methyl group
that minimized the difference. For these reasons, it is justified to compare the relative in-
tensities of these peaks in order to estimate the relative concentrations of the different Si
species. The hydrolysis of pure DMDMS in aqueous 0.1 M HCl solution (DM) resulted
mostly in the condensation. The residual amount of silicon atoms carrying one hydroxyl
group (D1 species) was small (see Fig. 5.3a). In contrast, the D1:D2 ratio was close to 2
when this reagent was used to modify MCM-41 silica (M-DM) and about 1 for SBA-15
silica (S-DM) as shown in Fig. 5.3b and Fig. 5.3c. The condensation of pure MTMS in
0.1 M HCl solution (TM) was less efficient. About 20% of the silicon atoms were still
carrying one hydroxyl group (T2species, Fig. 5.3d). At the pore walls of MCM-41, only
about 60% of the reagent molecules were forming two covalent bonds to the surface and
5.3. ARRANGEMENT OF THE SURFACE SILANOL GROUPS 57
Figure 5.3: Left: 29Si NMR of hydrolysed dimethyldimethoxysilane (DMDMS), sample
DM (a) and 29Si CPMAS NMR of MCM-41, sample M-DM (b) and SBA-15, sample
S-DM (c) silica samples functionalized using DMDMS as reagent. Right: 29Si NMR of
hydrolyzed methyltrimethoxysilane (MTMS), sample TM (d) and 29Si CPMAS NMR of
MCM-41, sample M-TM (e) and SBA-15, sample S-TM (f) silica samples functionalized
using MTMS as reagent.
almost no molecules forming three bonds (M-TM, Fig. 5.3e). The T1:T2:T3 ratio on the
surface of SBA-15 was about 1:10:4 (S-TM, Fig. 5.3f). The resulting T1:T2:T3 ratios are
summarized in Table 5.1.
5.3 Arrangement of the surface silanol groups
We showed that silica functionalization using HMDS as reagent allows to obtain practi-
cally hydroxyl-free surfaces. This means that groups such as Si(CH3)3being immobilized
on the surface do not prevent functionalization of neighboring silanol groups. The diam-
eter of a Si(CH3)3group at the surface is about 0.5 nm. The mean surface density of
the silanol groups is about 3 per nm2, corresponding to a mean distance of about 0.58
nm between the neighboring silanol groups. However, this estimation does not allow to
58 CHAPTER 5. FUNCTIONALIZATION BY GRAFTING
conclude whether or not the silanol group at the surface of ordered mesoporous silica
are distributed uniformly. This question might be answered by analyzing the rotational
dynamics of the immobilized functional groups. They predict only the lower limit for
the distance between the neighboring silanol groups on the idealized MCM-41-like sur-
face. The latter indicates that any functionalization reagent of the (CH3)4−n-Si-(OH)n
type could form only one covalent bond to the surface and the residual chemically active
groups could be used only to form a new silica layer.
A closer look at spectra of functionalized silica in Fig. 5.3 shows that the half-widths
of the NMR peaks corresponding to D1and T1species is only about one half of the width
of D2, T2and T3species. In contrast, the half-width of the NMR peaks corresponding
to T2and T3species formed in aqueous solution after CH3Si(OMe)3hydrolyzation was
approximately the same as that of the D1and T1species on the silica surface. We ascribe
this effect to inhomogeneous broadening due to some variations in the geometry of differ-
ent D2, T2and T3species bonded to the silica surface. These deviations could be caused
either by mutual bonding between the functional groups or their multiple bonding to the
surface. Which of these two reasons plays the dominant role may be different for MCM-
41 and SBA-15 silica. When the surface was functionalized using CH3Si(OMe)3as the
reagent, we did not observe T3bonded species on the MCM-41 surface. This finding is
at variance with data reported in other studies [72,73]. However, in these studies the de-
gree of functionalization was higher than for our samples. Accordingly, the formation of
covalent bonds to further reagent molecules cannot be excluded. Indeed, the fact that the
number of T2species on the MCM-41 surface was much higher than the number of D2
species, when CH3Si(OMe)3reagent was used instead of (CH3)2Si(OMe)2, could hardly
be explained exclusively in terms of covalent bonding to the surface. The hydroxyl group
of the surface-bonded (CH3)2Si(OH) species should be directed closer to the surface than
one of the hydroxyl groups of the surface-bonded (CH3)Si(OH)2species, and this pro-
vides a better access of another reagent to the latter hydroxyl group. Thus, the surface of
MCM-41 silica did not contain silanol groups which would be able to react with the same
molecule of the (CH3)4−n-Si-(OH)ntype reagent. If the formation of mutually bonded
functional groups could be suppressed, such a surface would contain exclusively single
surface bonded species. On the other hand, each surface silanol group could carry the
functional group.
The amount of D2, T2and T3species strongly increases when the rough surface of
SBA-15 was used instead of the quasi-ideal surface of MCM-41. It is not clear whether T3
species observed on SBA-15 silica could be formed exclusively by the covalent bonding
5.3. ARRANGEMENT OF THE SURFACE SILANOL GROUPS 59
to the surface or by bonding to other functional groups as well. We assume that the latter
case is more realistic. In Figure 5.4 the situation is summarized in a pictorial way. For
simplicity let us assume a uniform arrangement of the silanol groups on the surface (Fig.
5.4a). When the (CH3)2Si(OMe)2reagent is used to functionalize such a surface, it may
Si
OO
O
H
Si
OO
O
Si
O
Si
OO
O
H
Si
O
Si
Me
Me
O
Si
Me
Me
OH
Si
OO
O
H
Si
OO
O
Si
O
Si
OO
O
Si
O
Si
OO
Si
OH
Me
Si
OH
Me
Me
OH
Si
OO
O
H
Si
OO
O
H
Si
O
Si
OO
O
H
Si
O
Si
OO
O
H
SiSi
OO
O
H
Si
OO
O
H
SiSi
OO
O
H
Si
O
Si
O
SiSi
O
Si
OO
O
H
SiSi
OO
O
H
Si
O
Si
O
SiSi
O
a
b
c
d
e
Si
OO
O
H
Si
OO
O
Si
O
Si
OO
O
H
Si
O
Si
OH
MeMe
Si
OO
O
H
Si
OO
O
Si
O
Si
OO
O
Si
O
Si
O
Si
OH
M
Me
OH
H
OH
Figure 5.4: Cartoon of the surface structures of a silica before (a) and after its treatment
with a small (b,d) and large (c,e) amount of (CH3)2Si(OCH3)2and (CH3)Si(OCH3)3
reagents.
form only one covalent bond and the second active group is just converted to hydroxyl
(Fig. 5.4b). It is hard to imagine how this group could form a second covalent bond to the
surface. Potentially, it might react with another reagent molecule (see Fig. 5.4c), but the
60 CHAPTER 5. FUNCTIONALIZATION BY GRAFTING
experimental spectra indicate that such a reaction was not very effective, as shown in Fig.
5.3b. When the CH3Si(OMe)3reagent is used, it may also form only one covalent bond to
the surface, but at least one of the two residual free hydroxyl groups can easily be attacked
by another reagent (Fig. 5.4d). If no further reaction would take place, the T1:T2ratio
would be about 1:1. In contrast, the experimental data for this reagent on MCM-41 silica
reveal a ratio of about 2:3 (Fig. 5.3e). This may indicate that a third reagent molecule
becomes involved, leading to the double surface bonded species (Fig. 5.4e). The presence
of surface defects will allow more complex structures with reduced numbers of unreacted
hydroxyl groups, as was indeed observed for SBA-15 silica.
5.4 Conclusions
The main results of the experiments described in this chapter can be summarized as fol-
lows. The arrangement of Q3silicon atoms on the silica surfaces was analyzed using sur-
face functionalization by grafting route. In particular, it was shown that Si(CH3)3groups
attached to the surface do not prevent reactions with the neighboring surface silanol
groups. This fact indirectly suggests that the distance between neighboring surface silanol
groups is greater than 0.4-0.5 nm. 29Si CPMAS NMR measurements of functionalized sil-
icas indicated that the number of covalent bonds formed by the reagents (CH3)2Si(OH)2
and CH3Si(OH)3depends on the surface roughness. The experiments show that these
reagents can form only one covalent bond to the surface of high-quality MCM-41 silica.
The residual hydroxyl groups of the surface-bonded reagents either remain free or interact
with other reagent molecules. Hence, the surface of MCM-41 did not contain neighbor-
ing silanol groups which would be able to form covalent bonds with the same molecule of
(CH3)2Si(OH)2or CH3Si(OH)3type. On the other hand, it appears that multiple covalent
binding to the surface is possible to some extent on the rough pore walls of SBA-15 silica,
at least for the CH3Si(OH)3reagent. In other words, the number of covalent bonds formed
by the functional reagents to the silica surface is affected by the surface morphology. The
pore walls can be almost fully functionalized when HMDS is used as the functionalization
agent. In contrast, at the same surface the number of silanol groups may even increase
by the functionalization under certain conditions when CH3Si(OH)3is used as grafting
agent.
Chapter 6
Acid-Functionalization by
Co-condensation
Abstract
Functionalized SBA-15 silica materials with acidic groups of different pKa values in the
liquid phase such as carboxylic acid, phosphonic acid and sulfonic acid are synthesized by
the co-condensation route in order to study the effect of confined geometry on acidity. The
functional silanes used are 2-cyanoethyltriethoxysilane (CTES) for carboxylic acid func-
tionalization of SBA-15 (CA-SBA-15), diethoxyphosphorylethyltriethoxysilane (PTES)
for phosphonic acid functionalization of SBA-15 (PA-SBA-15), and 3-(trihydroxysilyl)-
1-propane-sulfonic acid (STHS) for sulfonic acid functionalization of SBA-15 (SA-SBA-
15). The structure of the functionalized materials is affected by the amount of the func-
tional silane (F), by the prehydrolyzed silica source (TEOS or F) and by the prehydroly-
sis time. Reaction conditions for a high-degree of functionalization under preservation of
mesoscopic well-ordered structure of silica changes with the functional silane used. The
results suggest that the structural order of the materials are strongly affected by assem-
bly kinetics which involve the hydrolysis and the condensation of silica species and their
interaction with template. Nitrogen adsorption isotherms indicate that the functionalized
SBA-15 materials have pore diameters larger than 7 nm. Moreover, the microporosity of
acid containing SBA-15 samples is lower than that of pure SBA-15. The 13C CPMAS
NMR and TGA/DTA measurements show that the higher the percentage of the functional
silane φin the synthesis mixture, the lower the amount of polymer template in the as-
synthesized functionalized SBA-15, and the higher the degree of the polymer removal
by sulfuric acid treatment. In SBA-15 materials synthesized with a molar percentage of
functional reagent Fof φ= 20 the polymer template can be removed completely. The
61
62 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
highest surface coverages of SBA-15 by carboxylic acid, phosphonic acid, and sulfonic
acid groups are 50%, 40% and 30%, respectively. Also the yields of the functional-
ization reactions correspond to that decreasing sequence. The acidity measurements of
functionalized SBA-15 materials were performed via 15N CPMAS solid-state NMR after
adsorption of pyridine into the pores. The pKavalues corresponding to the 15N chemical
shifts for the protonated pyridine indicate a high proton donor ability of all three solid
acids. The dependence of the proton-donor ability of the 10-CA-SBA-15 (10-CA) and
20-CA-SBA-15 (20-CA) silicas on the pH of the solution by FT-IR measurements show
the deprotonation of COOH groups at pH 8. The results suggest that in the 10-CA ma-
terial with 20% surface coverage nearly all COOH groups are separated and accessible,
while in the 20-CA material with 50% surface coverage only a part of the COOH groups
is accessible.
6.1. TEMPLATE REMOVAL: CALCINATION VS. ACID TREATMENT 63
6.1 Template removal: calcination vs. acid treatment
The effectiveness of template removal by treatment with 48% H2SO4was studied with
pure (non-functionalized) SBA-15 by measuring the weight loss in TGA. The TGA curves
of SBA-15 as synthesized (as-S), SBA-15 after H2SO4treatment (S-acid) and after calci-
nation at 550◦C (S-cal) are displayed in Figure 6.1. The TGA curves indicate that as-S
100 200 300 400 500 600 700 800
5 0
6 0
7 0
8 0
9 0
100
S-cal
S-acid
TG (%)
T/°C
240°C
205°C
as-S
Figure 6.1: TGA curves of as-SBA-15 (as-S), SBA-15 after H2SO4treatment (S-acid) and
after calcination at 550◦C (S-cal).
contains approx. 47% of polymer template. After template removal with H2SO4the sam-
ple S-acid still contains 18% polymer, whereas the mass loss of the sample calcined at
550◦C (S-cal) is only 3%. Hence template removal from as-S by acid treatment is not as
effective as calcination. A possible explanation for this behavior is that the polymer rests
are occluded in the silica matrix and are not accessible to the acid. Figure 6.1 also shows
that the decomposition temperature of P123 in acid treated SBA-15, S-acid, (240◦C) is
slightly higher than that in as-SBA-15, S-as, (205◦C). Another finding obtained from the
TGA curves is that the content of physisorbed water in S-cal is ca. 3%, but less than 1%
in the sample containing a rest of the polymer template. The latter effect can be explained
by the hydrophobicity of the polymer groups.
To completely remove the polymer, the acid-treated sample was calcined at 250◦C
for 3 h in air. The maximum calcination temperature used was 250◦C in order to avoid
shrinkage during the high-temperature treatment. To check the degree of polymer removal
64 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
in the acid treated sample, CHN elemental analysis was performed at BAM (Berlin). In
this process the carbon content in the samples S-cal, S-acid, and the acid treated sample
after calcination at 250◦C for 3 h (S-acid-cal) were analyzed. The outcome of this analysis
is presented in Table 6.1. According to these results the acid-treated SBA-15 has the
highest carbon content and thus the highest polymer content, but after calcination at 250◦C
for 3 h the sample loses nearly 97% of the carbon content. However, comparing the carbon
contains in the samples calcined at 250◦C and 550◦C shows that the latter still contains
less carbon than the further. Hence, for polymer removal calcination at 550◦C is more
effective than acid treatment and subsequent calcination at 250◦C.
Table 6.1: Characterization of SBA-15 silicas by N2adsorption, SAXD and CHN ele-
mental analysis.
sample as/m2g−1vp/cm3g−1vm/cm3g−1(p/p0)pc D/nm a0/nm C wt%
S-cal 947 1.18 0.21 0.743 8.5 11.0 0.04
S-acid-cal 807 1.35 0.12 0.807 10.2 12.5 0.31
S-acid 483 1.0 <00.794 9.8 12.6 9.25
as: specific surface area, vp: pore volume, vm: micropore volume from t-plots according to the Harkins-
Jura equation, (p/p0)pc: pore condensation pressure of nitrogen (77 K), D: pore diameter (imp. KJS, Eq.
2.27, a0: lattice parameter, C wt%: carbon weight percent.
Nitrogen adsorption measurements corroborate these results. Figure 6.2a shows ni-
trogen adsorption isotherms in S-cal, S-acid, and , S-acid-cal. A t-plot analysis of these
isotherms for determining the micropore volume of the three samples is shown in Fig.
6.2b. The results of this analysis are presented in Table 6.1. It can be seen that the sample
calcined at 550◦C has the smallest pore diameter (D), which can be attributed to a shrink-
age of the matrix during the high-temperature treatment. The pore diameter of this sample
is about 1 nm smaller than that of acid treated SBA-15. By contrast, the pore volume (vp)
as well as the specific surface area (as) of acid treated SBA-15 sample are much smaller,
than the calcined sample because of the polymer content (18%) in the sample. Moreover,
calcination of the acid-treated SBA-15 at 250◦C for 3 h (S-acid-cal) causes an increase of
the pore diameter, the specific area and the pore volume. The pore volume of this sample
is even bigger than that of SBA-15 calcined at 550◦C (S-cal). The lattice parameter a0
obtained from SAXD measurements is 11.0 for S-cal, and 12.6 nm for S-acid. From these
values and the pore diameter the wall thickness (w=a0-D) can be calculated as 2.5 nm and
2.8 nm, respectively. Thus the wall thickness of S-cal is slightly less than that of S-acid.
6.1. TEMPLATE REMOVAL: CALCINATION VS. ACID TREATMENT 65
0,0 0,2 0,4 0,6 0,8 1,0
0
100
200
300
400
500
S-acid
S-acid-cal
adsorbedvol./cm3.g-1
t/nm
S-cal
(b)
0,0 0,2 0,4 0,6 0,8 1,0
0
150
300
450
600
750
900
S-cal
S-acid-cal
S-acid
adsorbedvol./cm3.g-1
relative pressure
(a)
Figure 6.2: Nitrogen adsorption isotherms in three SBA-15 materials: S-cal, S-acid and
S-acid-cal (a); t-plots of these three isotherms (b).
The results from CHN elemental analysis could be confirmed by the determination
of microporosity by the t-plot method. The data are also included in Table 6.2. In the
micropore region, the t-plot for the sample S-cal gives a straight line for values between
0.55 and 0.75 nm, for S-acid-cal between 0.4 and 0.8 nm, and for S-acid between 0.5 and
0.85 nm. As can be seen in Fig. 6.2b, the y-intercept and thus the micropore volume
increases when increasing the calcination temperature. This indicates that the micropore
volume is related to the amount of polymer left in the materials. Namely, the higher the
polymer content the smaller is the micropore volume. This trend can be attributed to the
fact that the microporosity of calcined samples is due to the removal of polymer from
narrow pores in the matrix, so that these pores become available for gas adsorption.
Figure 6.3 shows SAXD spectra of four samples: as-S, S-cal, S-acid and S-acid-cal.
Differences in the relative peak intensities are observed in the four samples. For example,
in polymer-filled sample (as-S) the (20) peak is higher than the (11) peak, and the (30)
peak is hardly observable. After acid treatment (S-acid) and after its calcination at 250◦C
(S-acid-cal) the (11) peak is higher than (20) peak. The relative peak intensities of these
two samples are similar except for the (30) peak. Its intensity in sample S-acid-cal is
weaker than in the sample only treated with acid (S-acid). The peak intensities of the
sample S-cal appear to be higher than those of the other samples, presumably because of
its lower Porod scattering. For this sample the intensity of the (20) and (30) peaks starts
to increase again relative to those of the acid-treated materials.
In conclusion, the template removal studies with ‘pure’ SBA-15 materials show that
template removal by treatment with 48 wt.% H2SO4followed by a mild calcination at
250◦C is an alternative method to direct calcination at 550◦C. SBA-15 materials obtained
66 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
0 , 5 1 , 0 1 , 5 2 , 0
S - c a l
S - a c i d - c a l
S - a s
S - a c i d
( 3 0 )
( 2 0 )
( 2 1 )
( 1 1 )
I n t e n s i t y / a . u .
q / n m - 1
( 1 0 )
Figure 6.3: SAXD spectra of four samples: as-S, S-cal, S-acid and S-acid-cal. The spectra
are shifted against each other.
in this way have larger pore sizes and pore volumes than the respective SBA-15 calcined
at 550◦C. Moreover, a small fraction of the polymer template remains which reduces the
micropore volume and thus the specific surface area of porous sample. With regard to
catalytic applications of the materials this is an undesirable situation. On the other hand,
a polymer coated silica surface can have some advantages too. For example, nitrogen
adsorption measurements of SBA-15 samples kept in air have shown that samples con-
taining some polymer are less affected by aging than samples calcined at 550◦C. The
changes of relative peak intensities in the SAXD spectra are most probably related to the
polymer content in the pore channels. The different amounts of polymer can be consid-
ered as organic films of different thickness at the pore walls. A determination of the film
thickness from the SAXD profiles is possible on the basis of appropriate models of the
electron density in the unit cell [74].
6.2. CARBOXYLIC ACID FUNCTIONALIZATION 67
6.2 Carboxylic acid functionalization
SBA-15 functionalized with carboxylic (propionic) acid materials were synthesized ac-
cording to a prescription by Yang et al. [71]. The polymer filled CN-SBA-15 materials
were examined by TGA measurements and compared with as-synthesized SBA-15 (see
Figure 6.4). The TGA curves of the CN-SBA-15 materials show two exothermic de-
composition steps at 185◦C and ca. 330 ◦C which are attributed to the block copolymer
template P123 and cyanoethyl groups, respectively. Hence the decomposition tempera-
ture of P123 in as-SBA-15 is slightly higher (about 20 ◦C) than in CN-SBA-15. As seen
in Fig. 6.4 the 20-CN material exhibits a weight loss of about 40%, whereas the content of
the organic material in the 10-CN and as-SBA-15 leads to a weight loss of approx. 50%.
Hence, the amount of P123 in 20-CN is lower than in the two latter silica materials.
- 8
- 6
- 4
- 2
0
300 600 900
4 0
6 0
8 0
100
10-CN
20-CN
TG (%)
T/°C
as-SBA-15
330°C
DTG of 20-CN
185°C
Figure 6.4: TGA analysis of as-synthesized SBA-15, 10-CN and 20-CN. The DTG curve
of 20-CN is shown by the dashed-dotted curve.
6.2.1 Morphology, structure and porosity
The structure and porosity of the carboxylic acid-containing SBA-15 materials was char-
acterized by SEM, SAXD, and N2adsorption measurements.
The SEM images (Figure 6.5a and Fig. 6.5b) reveal that the acid-containing SBA-15
material consists of worm-like bent particles of hexagonal cross-section which seem to
stick together. Moreover, the images indicate that the outer surface of the particles is not
68 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
smooth but exhibits of hairy structures which make the contour of the particles diffuse.
In image (c) the arrangement of individual pores which run parallel each other is visible.
The direction of the view is perpendicular to the pore axes.
2mm400nm 100nm
(a) (b) (c)
Figure 6.5: SEM images of 20-CA in different magnifications. At the magnification of
100 nm the parallel ordered mesopores appear.
Nitrogen adsorption isotherms and SAXD profiles of 10-CA and 20-CA are shown in
Figure 6.6 together with the data of pure SBA-15.
0,0 0,2 0,4 0,6 0,8 1,0
0
300
600
900
1200
1500
20-CA
10-CA
adsorbedvol./cm3.g-1
relative pressure
SBA-15
(a)
0,4 0,8 1,2 1,6 2,0
SBA-15
(21)
20-CA
(30)
(20)
(11)
(10)
Intensity/a.u.
q/nm-1
10-CA
(b)
Figure 6.6: Nitrogen adsorption isotherms (a) and SAXS spectra (b) of SBA-15 (S-acid-
cal), 10-CA calcined at 250◦C for 3h, and of 20-CA. Isotherms and SAXD spectra are
shifted against each other for clarity.
The measured typ IV adsorption isotherms of CA-SBA-15 materials show sharp pore
condensation steps with H1 type hysteresis loops corresponding to the filling of the or-
dered mesopores. The pore size distribution is calculated from the adsorption branch
according to the imp. KJS equation (Eq. 2.27 in Section 2.6).
6.2. CARBOXYLIC ACID FUNCTIONALIZATION 69
Table 6.2: Properties of pure and carboxylic acid functionalized SBA-15.
sample as/m2g−1vp/cm3g−1vm/cm3g−1(p/p0)pc D/nm a0/nm w/nm
SBA-15†807 1.35 0.10 0.807 10.2 12.5 2.3
10-CA‡544 0.98 0.02 0.778 9.4 12.1 2.7
20-CA 533 0.82 0.01 0.716 8.0 11.4 3.4
†S-acid-cal, template removal by sulfuric-acid treatment followed by the calcination at 250◦C for 3 h;
‡sample is calcined at 250◦C for 3 h; as: specific surface area, vp: pore volume, vm: micropore volume,
(p/p0)pc: pore condensation pressure of nitrogen, D: pore diameter, a0: lattice parameter.
Physicochemical properties of pure and carboxylic acid-functionalized silica are sum-
marized in Table 6.2. It can be seen that the average pore diameter, the pore volume and
the pore area all decrease when the molar percentage (φ) of functional reagent (CTES) in
the synthesis mixture is gradually increased. Also SAXD spectra show peaks which can
be assigned to a 2D hexagonal ordered pore system. The cell parameter is also decreases
when increasing the percentage φof functional reagent, whereas the wall thickness in-
creases. The smaller pore diameter and the lower amounts of P123 used in the synthesis
of 20-CA materials suggest a smaller aggregation number of P123 in the formation of
20-CN compared to that for pure as-SBA-15. In the SAXD spectra of the three samples
(Fig. 6.6b) the relative intensities of the Bragg peaks are different. We think that, this can
be attributed to the different amount of organic material present inside the pore channels.
6.2.2 Polymer removal and microporosity
The removal of the template P123 by treatment with sulfuric acid was monitored by 13C
NMR and TGA measurements. The 13C CPMAS NMR spectra are shown in Figure 6.7.
The spectra of materials before acid treatment contain lines attributed to P123 (71, 74,
76 and 18 ppm) and the cyanoethyl groups (122 and 18 ppm). The appearance of the
small peaks at 122 and 39 ppm indicates that some of the -CN groups have already been
hydrolyzed to COOH groups before the acid treatment. In the 13C CPMAS NMR spectra
of acid treated samples three resonance lines assigned to the (CH2)2-COOH groups (at
179, 27 and 5 ppm) are dominating. Whereas the spectrum of 20-CA indicates complete
removal of the polymer template, the spectrum of 10-CA exhibits small lines stemming
from remaining EO chains of P123 despite calcination at 250◦C. As discussed in Section
6.1, due to the low accessibility of the EO chains in the silica matrix for sulfuric acid,
acid-treated SBA-15 still contains ca. 20% of P123 template. Thus the 13C NMR results
70 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
200 150 100 50 0
ppm
250 50
-
O
Si
O
O CH2CH2C N
123
O
Si
O
OCH2CH2C
OH
O
1' 2' 3'
3‘
1‘
P123
1
1‘
p123
2‘
2+3, 3‘
10-CA
20-CN
20-CA
2‘
Figure 6.7: 13C CPMAS NMR of the samples 20-CN, 10-CA (after calcination at 250◦C
for 3 h), and 20-CA.
suggest that by increasing the proportion of functional reagent φin the synthesis mixture
the accessibility of EO chains of the polymer for sulfonic acid is increased. Presumably
in 20-CN materials the EO chains are either not incorporated or weakly bonded in the
silica matrix.
TGA/DTA measurements confirm the results of the 13C CPMAS NMR experiments:
TGA curves of 10-CA and 20-CA materials both show that the CH2-CH2-COOH surface
functional groups decompose at about 350◦C and 450◦(Figure 6.8). However, the peak
at 350◦C is not always observable and is rather obscured by the broad peak at 420-450◦C.
In the case of 10-CA a further peak at ca. 235◦C appears which can be assigned to the
remaining polymer in the silica matrix. This behavior is consistent with the microporosity
analysis of the materials by the t-plot method. t-Plots of the nitrogen adsorption in SBA-
15, 10-CA and 20-CA are displayed in Figure 6.9 and the resulting data are included in
Table 6.2. From this analysis one concludes that only mesopores and nearly no microp-
ores exists in all the CA-SBA-15 samples after acid treatment. The t-plots were obtained
6.2. CARBOXYLIC ACID FUNCTIONALIZATION 71
200 400 600 800 1000
76
80
84
88
92
96
100
-16
-12
-8
-4
0
4
8
TG(%)
T/°C
DTA
20-CA
420°C
200 400 600 800 1000
88
90
92
94
96
98
100
-9
-6
-3
0
3
6
9
235°C
350°C
DTA
TG(%)
T/°C
451°C
10-CA
Figure 6.8: TGA and DTA measurements of the samples 10-CA and 20-CA.
0 , 0 0 , 2 0 , 4 0 , 6 0 , 8 1 , 0
0
100
200
300
400
S B A - 1 5 1 0 - C A
a d s o r b e d v o l . / c m 3 . g- 1
t / n m
2 0 - C A
Figure 6.9: Nitrogen t-plots of samples SBA-15-cal250, 10-CA-cal250 and 20-CA.
in the micropore region from t= 0.36 nm to t=0.5 nm. As seen in Fig. 6.9, in spite of
small polymer content still remaining in the matrix, the y-intercept of SBA-15 sample
has the highest value, indicating that the sample has the highest microporosity. The mi-
cropore volume of 10-CA is three times higher than that of 20-CA. This shows that the
micropore volume is related to the amount of polymer left in the silica materials after the
acid treatment, which is in turn affected by the molar ratio of functional group φpresent
in the synthesis mixture.
72 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
6.2.3 Reaction yield and the degree of the functionalization
The degree of acid functionalization of SBA-15 materials synthesized by co-condensation
method was studied by TGA/DTA and 29Si NMR measurements, and potentiometric titra-
tions. Under the synthesis conditions the sample 10-CA (with φ=10) released about 1.38
mmol functional silane (F) per g SiO2while a loss of 1.66 mmol of F/g SiO2is expected.
For the sample 20-CA (with φ=20) a weight loss of 2.79 mmol F/g SiO2was measured,
while a loss of 3.66 mmol F/g is expected. Thus, both values indicate a high degree (ca.
80%) of carboxylic acid groups incorporated in these samples (see Table 6.3).
The extent of functional silane incorporation in the mesoporous materials was also
monitored by 29Si NMR. Figure 6.10 shows 29Si MAS NMR spectra of materials contain-
ing different amount of functional groups. The analysis of such spectra of functionalized
silica materials is explained in Section 2.4.
T2/T3
Q3
Q4
T3
T2
Q2
50 -50 -100 -150 -200
ppm
0
10-CA
20-CA
T2/T3
Q3
Q4
T3
T2
Q2
50 -50 -100 -150 -200
ppm
0
50 -50 -100 -150 -200
ppm
0
10-CA
20-CA
Figure 6.10: 29Si MAS NMR spectra of 10-CA and 20-CA.
The spectra of acid-treated materials in Fig. 6.10 indicate that an increase in the
concentration of functional silane (φ) in the initial mixture yields increasing the amount
of (CH2)2-COOH groups in the silica wall. The relative integrated intensities of siloxane
(Qn) and organosiloxane (Tm) NMR signals (Tm/Qn) (Equation 2.9) allow a quantitative
assessment of the degree of functionalization of the silica. Table 6.3 shows the data for
the individual peaks normalized to the Q4peak and adjusted to the ratios of the integrals
of the T and Q regions.
6.2. CARBOXYLIC ACID FUNCTIONALIZATION 73
Table 6.3: Expected quantities and the corresponding experimental values obtained from
29Si NMR, TGA/DTA and potentiometric titration.
expected NMR TGA titration
sample (T/Q) n x (T/Q) Y n Y n Y
10-CA 0.11 1.66 20% 0.08 75% 1.38 82%
20-CA 0.25 3.32 42% 0.21 83% 2.79 84% 2.53 76%
(T/Q): Expected and experimental ratio of organosiloxane (T) and siloxane (Q) groups; n: Expected and
experimental amount of functional silane per unit mass of SiO2(in mmol g−1); x: Degree of surface
functionalization; Y: Reaction yield.
Hence in the case of 10-CA 20% of the surface groups (-OH and (CH2)2-COOH) are
carboxylic acid groups, whereas in the case of of 20-CA the respective percentage is 42%.
Moreover the reaction yield of the incorporation of carboxylic acid groups for 10-CA and
20-CA is 75% and 83%, respectively. Hence, within the error margin these values are in
good agreement with those determined from solid-state NMR.
The functionalization degree of the 20-CA sample is also determined by potentiomet-
ric titration (for experimental detail see Section 4.2). Although the resulting amount of
functional silane in the sample and reaction yield are lower (see Table 6.6), within the
error margin these values are consistent with those obtained from NMR and TGA/TDA
measurements.
6.2.4 Water stability of CA-SBA-15 materials
Water stability is an important issue in many potential applications of porous materials.
For testing water stability the 20-CA-SBA-15 materials were exposed to water at ambient
temperature for 4 h, 6 h and 30 d. The effect of the water exposure on the structural
ordering of the material was monitored by nitrogen adsorption and SAXD measurements
(see Figure 6.11). No changes in the SAXD spectra before and after water treatment were
observed, indicating that no degradation of 20-CA occurs during this long-term exposure.
As shown in the inlet of Fig. 6.11a, no observable shrinkage of material occurs. In Figure
6.11b nitrogen adsorption isotherms of the same materials are shown. The values of the
pore diameter, specific surface area and specific pore volume derived from these isotherms
are given in Table 6.4. The adsorption isotherms confirm the result that the pore structure
of the 20-CA silica materials is not affected by the treatment of the materials with water.
Also the pore diameter is not significantly changed, but the pore volume of the 20-CA
74 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
0,0 0,2 0,4 0,6 0,8 1,0
0
100
200
300
400
500
adsorbedvol.cm3g-1
relative pressure
20-CA
20-CA-4h-w
20-CA-6h-w
20-CA-20d-w
(b)
0,5 1,0 1,5 2,0
1,0 1,1 1,2 1,3 1,4
(12)
(11)
Intensity/a.u.
q/nm-1
20-CA
20-CA-4h-w
20-CA-2d-w
20-CA-30d-w
(12)
(11)
Intensity/a.u.
q/nm-1
(10)
(a)
Figure 6.11: (a) SAXD profiles and nitrogen adsorption isotherms (b) for 20-CA before
and after water exposure for different periods of time.
material increases with the exposure to water. This effect may be caused by the ability
of water to remove some constrictions such as residual polymer in the pore system so
that formerly blocked parts of the pore volume are becoming accessible. The increase of
the specific surface area after water exposure can also be explained satisfactorily via this
interpretation.
Table 6.4: Properties of 20-CA before and after the water exposure for different periods
of time derived from the nitrogen adsorption isotherms of Fig. 6.11b.
sample as/m2g−1vp/cm3g−1D/nm
20-CA 400 0.60 8.1
20-CA-4h-w 470 0.72 8.1
20-CA-6h-w 462 0.70 8.2
20-CA-30d-w 489 0.75 8.2
as: specific surface area, (p/p0)pc: pore condensation pressure of nitrogen D: pore diameter (imp. KJS).
6.3 Sulfonic acid functionalization
In this section a simple one-step procedure is presented for the synthesis of sulfonic acid
functionalized SBA-15 based on the co-condensation of TEOS and 3-(trihydroxysilyl)-1-
propane-sulfonic acid (STHS) in the presence of block copolymers under acidic condi-
6.3. SULFONIC ACID FUNCTIONALIZATION 75
tions. For optimizing the degree of the surface functionalization of the pore walls and the
yield of the synthesis, several parameter were varied, such as the functional silane/TEOS
molar ratio in the synthesis mixture (φ), delay time for prehydrolyzing (prehydrolysis
time) and prehydrolyzed component. Notice that in case of STHS, the term ‘hydroly-
sis’ is not really appropriate because STHS is already in hydrolyzed state and does not
posses alkoxy-moieties to hydrolyze. Rather, we are dealing with precondensation of the
functional silane. For the sake of a uniform notation the term will nevertheless be used.
6.3.1 Study of template content
As a first step of the characterization of sulfonic acid functionalized SBA-15 (SA-SBA-
15), the polymer content before acid removal was studied by TGA measurements. TGA
curves of polymer-containing samples synthesized with different amount of functional
group are displayed in Figure 6.12.
100 200 300 400 500 600 700 800 900
5 0
6 0
7 0
8 0
9 0
100
TG (%)
T/°C
7-S
20-S
30-S
Figure 6.12: TGA curves of three sulfonic-acid-containing samples before polymer re-
moval: 30-S: 30-S-1480-F, 20-F: 20-S-00, 7-S: 7-S-60-F.
The graph shows that the content of organic component decreases when increasing
the amount of functional silane φin the synthesis. The weight loss of 48% for 7-S which
is similar to the polymer content in as-SBA-15 (see Section 6.1), decreases to 39% for
20-S and to 32% for 30-SA. Moreover, for the samples with φ= 20 and 30 a second step
in the TGA curve appears, which gets higher when raising the percentage φof functional
reagent. According to the height of this second step, the sample 30-S contains approx-
imately 20% of sulfonic acid, and the sample 20-S contains ca. 17% of sulfonic acid
76 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
groups. This observation shows that the P123 content decreases when increasing φin the
synthesis mixture. In fact, 20-S-60-F contains ca. 30 % P123 whereas 30-S-1480-F con-
tains only 12% P123. This in turn suggests, as in case of CN-SBA-15, a smaller number
of monomers in the micelles of P123.
6.3.2 Morphology, structure and porosity
The mesoscopic order of the materials was studied by nitrogen adsorption, SAXD and
scanning electron microscopy (SEM). Figure 6.13 shows SEM images of different sul-
fonic acid functionalized SBA-15 materials prepared by premixing the functional reagent.
The images reveal that the silica particles have long bent worm-like character which is
typical for SBA-15 materials in which φis small. When φis increased the particles be-
come shorter and more curved. The SEM image of the 30-SA-1480-F sample reveals that
this sample has almost no order which is most likely due to the high percentage of func-
tional silane φin the initial synthesis mixture. These results are consistent with nitrogen
3-SA-60-F 7-SA-60-F
20-SA-60-F 30-SA-1480-F
1mm
2mm
6mm
2mm
Figure 6.13: SEM images of sulfonic acid functionalized SBA-15 prepared by prehydrol-
ysis of the functional reagent.
adsorption isotherms and SAXD profiles of these samples, which are displayed in Figure
6.14. They show that neither the order of the materials nor their pore diameter is changed
significantly by increasing φup to 20. Only when φis increased to 30, the order of the
material deteriorates. Here the pore condensation is smeared out and the SAXD curve
does not show any higher peaks. The adsorption isotherms indicate that the increase of
6.3. SULFONIC ACID FUNCTIONALIZATION 77
0,0 0,2 0,4 0,6 0,8 1,0
0
300
600
900
1200
1500
1800
30-SA-1480-F
20-SA-60-F
7-SA-60-F
adsorbedvol.cm3g-1
relative presure
3-SA-60-F
(a)
0,4 0,8 1,2 1,6 2,0
20-SA-60-F
3-SA-60-F
(30)
(21)
(20)
(11)
7-SA-60-F
Intensity/a.u.
q/nm-1
30-SA-1480-F
(10) (b)
Figure 6.14: Nitrogen adsorption isotherms (a) and SAXD spectra (b) of sulfonic acid
functionalized SBA-15 materials prepared by premixing the functional reagent. The
isotherms and the SAXD profiles are shifted against each other.
φcauses a decrease of the steepness of the isotherms which in turn indicates a broader
pore size distribution and thus less order. The SAXD measurements indicate that the acid
treatment causes no shrinkage of the matrix. In some cases even a slight increase of the
lattice constant is observed after acid treatment. Thus acid treatment can cause an expan-
sion of the silica matrix whereas calcination at temperatures higher than 250◦C induce
a contraction. The results derived from nitrogen adsorption and SAXD measurements
are summarized in Table 6.5. The results indicate an increase of the lattice constant and
wall thickness when increasing the molar percentage of functional reagent φto 20 and 30.
This is very different from CA-SBA-15 materials, where the lattice constant decreases
when increasing φ. The relative intensities of the individual Bragg reflexes are changing
with the content of functional silane, as it was found for the CA-SBA-15 samples. In
particular the (11), (20) and (21) peaks undergo strong changes. These can be explained
when the results of Si-NMR are considered, which will occur in Section 6.3.3. In order to
check whether the properties of the materials or the yield of the synthesis depend on the
prehydrolysis of one of the silica sources two further samples were prepared. 20-SA-00
synthesized with no prehydrolysis time, i.e., TEOS and functional silane were added at the
same time, and 20-SA-55-T where TEOS was the prehydrolyzed component. The SEM
78 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
Table 6.5: Properties of pure and sulfonic acid functionalized SBA-15 silica.
sample as/m2g−1vp/cm3g−1vm/cm3g−1(p/p0)pc D/nm a0/nm w/nm
3-SA-60-F 579 1.26 0.02 0.799 10.0 12.6 2.6
7-SA-60-F 572 0.92 0.05 0.800 10.0 12.6 2.6
20-SA-60-F 577 1.01 0.02 0.784 9.6 12.9 3.3
30-SA-1480-F 672 0.86 0 0.765 9.1 13.4 4.3
20-SA-00 564 0.93 0.02 0.791 9.7 13.0 3.3
20-SA-55-T 419 0.72 0.03 0.797 9.9 12.8 2.9
SBA-15cal550 561 0.89 0.10 0.757 8.9 11.5 2.6
SBA-15cal250 807 1.35 0.11 0.807 10.2 12.8 2.6
as: specific surface area, vp: pore volume, vm: micropore volume, (p/p0)pc: pore condensation pressure of
nitrogen, D: pore diameter, a0: lattice parameter.
images of the samples 20-SA-00 and 20-SA-55-T in Figure 6.15 show the typical worm-
like habitus of SBA-15 materials. The samples are hexagonally ordered, and the length
of worms is short in comparison to pure SBA-15. The nitrogen adsorption isotherms and
20-SA-00 20-SA-55-S
4mm4mm
Figure 6.15: SEM images of sulfonic acid functionalized SBA-15 prepared by premixing
the functional reagent.
SAXD profiles of these two samples are displayed in Figure 6.16 together with 20-SA-
60-F, and the resulting data are included in Table 6.5. All isotherms are of type IV with
an H1 type hysteresis loop, as is typical for well-ordered SBA-15 silicas. The isotherms
reveal materials with similar pore sizes and pore size distributions independent from pre-
hydrolyzed silica source (STHS or TEOS). The isotherms of 20-SA-60-F (prehydrolyzed
component is functional silane (F)) and 20-SA-00 (no prehydrolyzed component, TEOS
and F added simultaneously) indicate similar pore volume and specific surface area, but
6.3. SULFONIC ACID FUNCTIONALIZATION 79
these parameter are larger than for 20-SA-55-T (prehydrolyzed component is TEOS, (T))
(Table 6.5). SAXD profiles corroborate these behavior: The three samples have similar
0,4 0,8 1,2 1,6 2,0
20-SA-60-F
(30)
(21)
(20)
(11)
20-SA-55-T
20-SA-00
Intensity/a.u.
q/nm-1
(10) (b)
0,0 0,2 0,4 0,6 0,8 1,0
0
100
200
300
400
500
600
adsorbedvol./cm3g-1
relative presure
20-SA-00
20-SA-60-F
20-SA-55-T
(a)
Figure 6.16: Nitrogen adsorption isotherms (a) and SAXD spectra (b) of sulfonic acid
functionalized SBA-15 prepared by variation the prehydrolyzed component: 20-SA-55-T
(TEOS prehydrolyzed), 20-SA-60-F (STHS prehydrolyzed), 20-SA-00 (STHS and TEOS
added simultaneously); SAXD spectra are shifted against each other.
pore diameters, but the lattice constant and the wall thickness of the samples 20-SA-60-F
and 20-SA-00 are somewhat larger than that of 20-SA-55-T. The relative Bragg peak in-
tensities are also different in three samples. Closer inspection shows that the intensities
of the (11) and (21) peak decrease in the order 20-SA-55-T >20-SA-60-F >20-SA-00
whereas the intensity of the (20) peak increase in that sequence. These changes of the
relative intensities of the Bragg peaks correlate with the degree of surface coverage of
SBA-15 by functional groups as determined by 29 Si MAS NMR measurements. This is
explained in the following section.
6.3.3 Reaction yield and the degree of functionalization
Quantitative determinations of the content of functional groups after removal of the block
copolymer were performed with TGA/DTA and 29Si MAS NMR measurements. 29Si
MAS NMR spectra of samples synthesized by ‘precondensation’ of the functional silane
(STHS) are shown in Figure 6.17a. They indicate that an increase of functional reagent
concentration φin the initial mixture yields increasing amounts of organo-functionalized
moieties in the silica wall. A gradual increase in the intensities of organosiloxane Tmis
readily observed. The relative integrated intensities of the siloxane (Qn) and organosilox-
80 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
150 300 450 600 750 900
75
80
85
90
95
100
30-SA-1480-F
20-SA-00
20-SA-60-F
TG(%)
T/°C
20-SA-55-T
-93.5-93.5
-103.4 -112.4- -112.4
Q4
TT2
TT2TT3
TT2TT3
0-50 -100
50 ppm -150 -200
7-SA-60-F
20-SA-60-F
30-SA-1480-F
3-SA-60-F
-93.5
-103.4 -112.4
-93.5
- -112.4
Q2
-103.4 -112.4- -112.4
-93.5-93.5 Q3
(a)
(b)
Figure 6.17: (a) 29Si MAS NMR spectra of four SA-SBA-15 materials with increasing
molar ratios of functional reagent φ; (b) TGA curves of four different SA-SBA-15 sam-
ples.
ane (Tm) NMR signals (Tm/Qn) allow a quantitative assessment of the incorporation de-
gree of the organic moiety. The calculated yields of the synthesis presented in Table 6.6
show that less than 50% functional silane applied in the initial synthesis mixture was
incorporated into the silica. Quantitative integration of the Tmsignals was not possible
for the samples 3-SA-60-F and 7-SA-60-F for which only small proportion of functional
reagent were used in the synthesis. A comparison of the reaction yields of 30-SA-1480-F
(37%) and its pre-stage 30-S-1480-F (33%) shown in Table 6.6 reveals that this low yield
is not caused by washing out of the functional groups during the acid treatment, but has
already occurred before acid treatment. According to the 29Si MAS NMR results, the
degree of surface coverage increases when increasing the portion of functional reagent in
the synthesis, where it is used as prehydrolyzed component. However, the order of the
materials is lost when applying 30 mol% functional silane.
The extent of incorporation of functional silane into the silica matrix was also mon-
6.3. SULFONIC ACID FUNCTIONALIZATION 81
Table 6.6: Expected quantities and the corresponding experimental values obtained from
29Si NMR, TGA/DTA and potentiometric titration.
expected NMR TGA titration
sample (T/Q) n x (T/Q) Y n Y n Y
20-SA-60-F 0.25 3.32 25% 0.09 37% 0.95 28% 0.69 21%
30-S-1480-F 0.43 4.98 33% 0.14 33%
30-SA-1480-F 0.43 4.98 37% 0.16 37% 1.18 24%
20-SA-00 0.25 3.32 31% 0.12 46% 1.10 33% 1.13 34%
20-SA-55-T 0.25 3.32 16% 0.06 25% 0.42 13%
(T/Q): Expected and experimental ratio of organosiloxane (T) and siloxane (Q) groups; n: Expected and
experimental amount of functional silane per unit mass of SiO2(in mmol g−1); x: Degree of surface
functionalization; Y: Reaction yield.
itored by TGA/DTA measurements. The amount of sulfonic acid contained in different
samples was calculated from the TGA curves displayed in Fig. 6.17b. The results in-
cluding the resulting reaction yields are summarized in Table 6.6. More details about the
TGA/DTA measurements will be discussed in Section 6.3.4. As can bee seen in Table
6.6, for all samples TGA/TDA gives smaller yields than 29Si-MAS-NMR measurements.
A further method used for the determination of the extent of incorporation of func-
tional silane into the silica matrix was potentiometric titration. The samples 20-SA-00
and 20-SA-60-F were titrated against diluted NaOH solution. For experimental details
see Section 4.2. The experimental amounts of functional silane and reaction yields de-
termined by potentiometric titration are rather low and similar to those obtained from
TGA/TDA measurements (see Table 6.6).
To summarize, for synthesis of ordered SA-SBA-15 materials the highest reaction
yields obtained from NMR, TGA and potentiometric titration measurements are 30-40%,
30%, and 20-30%, respectively. Thus, the good agreement of the yields of CA-SBA-
15 samples determined from NMR and TGA cannot be observed here. We investigated
whether the small yield of the TGA experiments is caused by the sulfur remnants in the
matrix after the heating to 1000◦C. The presence of sulfur remnants in the matrix would
increase its unit mass and decrease the computed yield from TGA. Elemental analysis
of the sample 20-SA-00 after heating to 1000◦C has shown that no sulfur is present in
the silica matrix. Thus, a corruption of TGA measurements by the presence of sulfur
residues in the silica matrix can be ruled out. Nevertheless, among the methods applied
82 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
the solid-state NMR is the more exact one. The relative error caused by the integration of
the resonance peaks is about 5-10%. In case of TGA and potentiometric titration methods
repeated measurements have shown that the relative error is about 20%.
The results in Table 6.6 also reveal the effect of the prehydrolysis. In fact the yield is
highest when the functional silane (STHS) and TEOS are added simultaneously, and it is
lowest when TEOS is prehydrolyzed. Another interesting feature of SA-SBA-15 samples
revealed by the TGA measurements is their high water content. According to TGA, the
samples release about 10% water which is three times higher than the water content in
CA-SBA-15 samples.
By combining the results of Si-NMR presented in Table 6.6 with those from SAXD
profiles and nitrogen adsorption isotherms the following results are obtained:
1. It appears that the sulfonic acid reagent (STHS) is not incorporated in the silica ma-
trix in amounts as high as in CA-SBA-15 materials. Even if the percentage of functional
silane is increased to φ=30 the loading remains low (1.18 mmol g−1SiO2) and the silica
matrix loses its order at this proportion of the functional silane.
2. Samples containing high amounts of sulfonic acid, i.e. 20-SA-60-F, 30-SA-60-F
and 20-SA-00, have a larger lattice parameter than samples whose surface coverage are
less than 15%. On the other hand, the pore size is not significantly changed when the
percentage of functional silane increases. Accordingly, the wall thickness of the samples
containing high amounts of sulfonic acid is larger than that of the latter (see Table 6.5).
In contrast, Margolese et al. [53] found a gradual decrease of the d−spacing for SBA-
15 samples synthesized with higher concentrations of functional reagent [53]. However,
in their study mercaptopropyltrimethoxysilane was used as functional silane, and it was
possible to incorporate higher amounts of functional silane in the pore walls.
3. The relative intensities of the (11) and (21) Bragg peaks in SAXD profiles (Fig.
6.14b and Fig. 6.16b) decrease and that of the (20) peak increases when surface coverage
of SBA-15 by sulfonic acid groups increases. Similar changes of relative peak intensities
in SAXD profiles have been observed in in-situ SAXD gas adsorption measurements in
SBA-15 [74], where relative peak intensities change as thickness of the adsorbed film
increases in each gas adsorption step. Since the pore diameters of the sulfonic acid func-
tionalized SBA-15 materials are similar, the differences in the relative intensities of the
Bragg peaks of the present samples must be related to the thickness of the film formed by
the propyl-sulfonic acid on the pore walls.
6.3. SULFONIC ACID FUNCTIONALIZATION 83
6.3.4 Polymer removal and microporosity
13C CPMAS NMR spectra of the samples 7-S-60-F, 20-SA-55-T, 20-SA-00 and 30-SA-
1480-F are displayed in Figure 6.18a. The polymer containing sample 7-S-60-F shows
carbon peaks assigned to P123 at 76, 74, 71 and 18 ppm, and in addition peaks at 12
ppm, 18 ppm and 56 ppm which come from carbon chains of the functional group. How-
ever, the peak at 18 ppm (C2) overlaps with the carbon peak stemming from P123. In the
spectra of 20-SA-55-T, 20-SA-00 very small peaks assigned to P123 are detectable. In
the spectrum of 30-SA-1480-F, however, the peaks attributed to P123 are missing. This
resembles the case of CA-SBA-15 materials, namely the higher the molar percentage of
the functional silane φin the sample the better the access of sulfuric acid to block copoly-
mer to remove it. Figure 6.18b shows DTG curves of different sulfonic acid containing
150 300 450 600 750 900
-0,2
0,0
0,2
0,4
0,6
0,8 250°C
P123
75°C
165°C
CH2-CH2-CH2-SOOOH
P123
20-SA-1480-F
20-S-1480-F
20-SA-60-F
20-SA-55-T
20-SA-00
435°C
H2O
DTG
T°C
20-S-00
3
P123+2
P123
{
SiO
O
O
CH2CH2CH2S
O
O
OH
123
1
30-SA-1480-F
0 -5050
100150 ppm
20-SA-00
7-S-60-F
3
20-SA-55-T
2
(a)
(b)
Figure 6.18: (a): 13C CPMAS NMR spectra of the samples 7-S-60-F, 20-SA-55-T, 20-
SA-00 and 30-SA-1480-F; (b): DTG curves of different sulfonic acid containing SBA-15
samples.
84 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
SBA-15 samples. Here four different peaks are observed: The peak at ca. 75◦C results
from physically adsorbed water. The peaks at 165◦C and 250◦C are assigned to P123
and appear only for the samples 20-S-00 and 30-S-1480-F, i.e., samples without H2SO4
treatment. The remaining samples which are all treated with sulfuric acid show only the
peak at about 435◦C attributed to the functional groups. Hence, here one cannot detect
the small amount of template after sulfuric acid treatment that was detected by 13C NMR
in samples having 20% functional reagent in the synthesis mixture. Therefore, in SA-
SBA-15 samples with φ≥20 the block copolymer template is almost fully accessible to
sulfonic acid, independently of which component (TEOS or STHS) was prehydrolyzed
first. This easy access may be due to a weak penetration of the EO chains in the silica
matrix. It is well known that during the hydrothermal treatment at higher temperature (ca.
100◦C) the EO chains of P123 become more and more hydrophobic. The sulfonic acid
head group is a stronger acid and more hydrophilic than carboxylic acid. This difference
may lead to a weaker interaction between silica matrix and EO chains of block copolymer.
If that is the case it should lead to a low microporosity. To check this conjecture, a t-plot
analysis of all sulfonic acid functionalized samples was performed (see Figure 6.19). In
the micropore region between t=0.4 and t=0.64 nm, these plots give a straight line.
0 , 0 0 , 3 0 , 6 0 , 9
0
100
200
300
400
a d s o r b e d v o l . c m 3g- 1
t / n m
3 - S A - 6 0 - F
7 - S A - 6 0 - F
2 0 - S A - 6 0 - F
3 0 - S A - 1 4 8 0 - F
2 0 - S A - 0 0
2 0 - S A - 5 5 - T
S B A - 1 5 - c a l 5 5 0
Figure 6.19: t-plot analysis of nitrogen adsorption isotherms for different sulfonic acid
functionalized SBA-15 in comparison with native SBA-15 calcined at 550◦C.
Figure 6.19 shows that the y-intercepts of all SA-SBA-15 samples are smaller than
those of pure SBA-15 calcined at 550◦C. This indicates that, similar to CA-SBA-15 ma-
6.3. SULFONIC ACID FUNCTIONALIZATION 85
terials, all SA-SBA-15 samples have indeed a much lower microporosity than pure SBA-
15. The weakly ordered sample 30-SA-1480-F even has no micropores. The calculated
micropore volumes are included in Table 6.5.
In Figure 6.20 the TGA and DTA curves of the samples 20-S-00 (a) and 20-SA-00
(b) are displayed which are typical for all sulfonic acid functionalized SBA-15 samples.
For 20-S-00 the TGA curve shows two steps. The corresponding DTA curve indicates an
150 300 450 600 750 900
76
80
84
88
92
96
100
-12
-9
-6
-3
0
3
6
108°C
400°C
DTA
TG(%)
T/°C
490°C
20-SA-00 (b)
150 300 450 600 750 900
50
60
70
80
90
100
-16
-12
-8
-4
0
485°C
86°C
400°C
DTA
TG(%)
T/°C
315°C
20-S-00
(a)
Figure 6.20: Typical TGA-DTA curves of sulfonic acid functionalized SBA-15 before
and after acid treatment. The samples presented here are 20-S-00 and 20-SA-00.
exothermic peak at ca. 86◦C assigned to the physically adsorbed water inside the pores,
followed by three endothermic peaks at 315◦C, 400◦C and 485◦C which can be attributed
to the decomposition of organic components (block copolymer and sulfonic acid groups).
In contrast, the TGA curve of 20-SA-00 sample (Fig. 6.20b) shows only one step. In
the DTA curve the peak at 315◦C disappears. Hence, this peak is to be attributed to the
degradation of P123, whereas the two residual peaks are to be assigned to the sulfonic
acid groups.
6.3.5 The mechanism of the self-assembly
The functional reagent 3-(trihydroxysilyl)-1-propane-sulfonic acid (STHS) has a hydropho-
bic propyl chain with a polar sulfonic acid head group. However, the propyl chain is
probably not sufficiently hydrophobic enough to make the functional silane act as a co-
surfactant. A cosurfactant would be preferentially adsorbed close to the hydrophobic PPO
portion of the P123 micelles and therefore cause a decrease in the interfacial curvature and
thus lead to an increase in the pore diameter or even cause a transformation to a different
structure as observed in other systems [55,75]. However, under the synthesis condi-
tions applied in this work such phenomenon has never been observed but always either
86 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
hexagonal ordered, or non-ordered materials were obtained. The pore size of synthesized
SA-SBA-15 materials was never larger than that of pure SBA-15 but usually smaller.
The reaction yield achieved in the functionalization of SBA-15 with sulfonic acid
(30-40%) is much lower than the reaction yield in the carboxyl acid functionalization of
SBA-15 (ca. 80%). In addition, the surface coverage with sulfonic acid is lower than in
the case of carboxylic acid: It is between 40 and 50% for 20-CA, but only approx. 30%
for 20-SA. By increasing the mol percentage of functional silane φin the synthesis, the
surface coverage can be increased up to ca. 40%, but in this case the hexagonal order
of SBA-15 gets lost. So 30-S-1480-S/SA does not show a 2D hexagonal order. Most
probably the high proportion of functional group in the synthesis mixture decreases the
interfacial curvature of P123 micelles leading to a less well-ordered structure. On the
O
Si
O
Si
O
OR
Si
Si
O
Si
O
R
Si
EtO
EtO
O
O R
Si
O
Si R
EtO
R
R
R
R
EtO
O
EtO
EtO
EtO
-
-
-
O
Si
O
Si
O
OEt
R
Si
Si
O
Si
O
R
Si
OEt
OEt
O
OEt
R
Si
O
SiR
O
O
O
OEt
R
R
R
R
OEt
OEt
O
Si
O
Si
O
OR
Si
Si
O
Si
O
R
Si
EtO
EtO
O
O R
Si
O
Si R
EtO
R
R
R
R
EtO
O
EtO
EtO
EtO
-
-
-
O
Si
O
Si
O
OEt
R
Si
Si
O
Si
O
R
Si
OEt
OEt
O
OEt
R
Si
O
SiR
O
O
O
OEt
R
R
R
R
OEt
OEt
O
Si
O
Si
O
OR
Si
Si
O
Si
O
R
Si
EtO
EtO
O
O R
Si
O
Si R
EtO
R
R
R
R
EtO
O
EtO
EtO
EtO
-
-
-
O
Si
O
Si
O
OEt
R
Si
Si
O
Si
O
R
Si
OEt
OEt
O
OEt
R
Si
O
SiR
O
O
O
OEt
R
R
R
R
OEt
OEt
EO EOPO
EO EOPO
EO EOPO
SOO
Si
SOO
Si
Si
O
H
SOOH
Si
O
O
SOOH
Si
O
SOO
Si
O
Si
O
Si
O
SOO
Si
O
O
SOO
-
Si
OH
O
SOO
Si
O
SOO
Si
O
SOO
Si
O
SOO
H
Si
O
SOO
Si
O
O
SOO
Si
S
t
Condensaion
O
H
O
H
H
O
H
O
H
O
H
O
H
SOH
O
Si
Si
OH
O
O
H
O
H
O
OH
HO
OH
-
-
-
HH
HH
H
H
-
O
H
O
H
O
H
O
H
O-
-
-
-
Figure 6.21: Proposed arrangement of functional silane in the periphery of P123 micelles
during the formation of sulfonic acid functionalized SBA-15; R=propyl sulfonic acid.
basis of these results, the following formation mechanism is proposed (see Figure 6.21).
6.4. PHOSPHONIC ACID FUNCTIONALIZATION 87
As mentioned before, the sulfonic acid reagent we used is rather hydrophilic. Therefore,
due to the attractive interaction between the polar propyl-SOOH groups and hydrophilic
PEO chains of the block copolymer the trihydroxysilyl groups can uniformly organize
on the surface of the micelles and their increased length will assist in the rapid cross-
linking/condensation between the micelle-oriented trihydroxysilyl groups, as depicted in
Fig. 6.21. The resulting ‘side-on’ packing of the silicate coated cylindrical micelles will
give rise to long rod-like nanoparticles. A similar mechanism was proposed by Cai et
al. and Huh et al. in the functionalization of MCM-41 by different functional groups via
co-condensation method [76,77,50].
In the synthesis of CA-SBA-15, 2D-hexagonal order of SBA-15 can only be obtained
if the functional silane is prehydrolyzed. Prehydrolysis of TEOS or simultaneous addi-
tion of TEOS and functional silane (CTES) leads to disordered silica materials [71]. On
the other hand, the functional silane (STHS) used in the synthesis of SA-SBA-15 has no
hydrolytic sensitivity and forms stabile aqueous solutions in the hydrolyzed form. Thus it
does not need a prehydrolysis time. For this reason we always obtained 2D-ordered mate-
rials with this functional silane, independent of which silica source (TEOS or STHS) was
added first. Moreover the highest reaction yield is obtained when sulfonic acid reagent
and TEOS are added at the same time. It can probably be explained by the slow conden-
sation rate of the hydroxyl groups of the highly acidic functional silane. Very likely the
functional silane organizes uniformly at the surface of P123 micelles due to the attrac-
tive interactions, and its condensation is enhanced by the hydrolysis and condensation of
TEOS.
6.4 Phosphonic acid functionalization
In this section the synthesis of phosphonic acid functionalized SBA-15 silica materials
by the co-condensation method is presented. Diethoxyphosphorylethyltriethoxysilane
(PTES) was used as the functional reagent. The synthesis of such materials was reported
already by Corriu et al. [78], but these authors used different reaction conditions than pre-
sented here. Specifically, they used NaF as an additional reaction component. Fluoride
ions are a well-known catalysts for the hydrolysis and polymerization of silica species
and have been employed in the synthesis of mesoporous silica materials under various
conditions in order to improve structural order [38]. Corriu et al. prepared phosphonic
acid functionalized materials with a constant mol percentage of φ= F/(F+T) = 9.9 in the
starting mixture. In the present work, phosphonate ester containing SBA-15 materials
88 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
with different molar composition of φwere synthesized. Free phosphonic acid function-
alities could then be formed by dealkylation of these phosphonate ester groups through
refluxing the samples in concentrated HCl. Additionally, the effect of prehydrolysis was
investigated, when either functional silane or TEOS was added first, or both components
were added simultaneously. Results of this study are presented in this section.
6.4.1 Morphology, structure and porosity
As with the CA-SBA-15 and SA-SBA-15 materials, the mesoscopic order and morphol-
ogy of phosphonic acid containing SBA-15 samples were investigated by SEM, SAXD
and nitrogen adsorption. SEM images of the samples 10-P-60-F and 10-POEt-60-F (phos-
10-P-60-F10-P- -F
10-POEt-60-F10-POEt-60-F
3mm
500nm
300nm
2mm
Figure 6.22: SEM images of phosphonate ester functionalized SBA-15 samples prepared
by prehydrolysis of the functional reagent (φ=10) before and after polymer removal.
phonic acid ester contained SBA-15 before and after template removal with sulfuric acid)
are displayed in Figure 6.22. These samples were synthesized with φ=10% of functional
silane and a prehydrolysis time of functional silane of 60 minutes. The images clearly
show the hexagonal habitus and elongated worm-like character of the particles before and
after polymer removal, which is typical for highly ordered SBA-15 materials. From this
it follows that the treatment with sulfuric acid does not cause any morphological changes
of the particles.
6.4. PHOSPHONIC ACID FUNCTIONALIZATION 89
However, as seen in the SEM images displayed in Figure 6.23, the morphology of the
materials changes when increasing the percentage of functional silane. For example if the
percentage of functional silane is increased to 15% (samples 15-P-95-F and 15-POEt-95-
F) much shorter particles result (Fig. 6.23a). Bent worm-like structures can only be seen
partially, and they seems to lump together forming cluster-like structures. But the hexag-
onal cross-sections of individuell particles are still recognizable in the images. When
20-P-60-F
20-POEt-60-F
15-P-60-F
15-POEt-60-F
TEM
100nm
3mm
4mm1mm
2mm7mm1mm
1mm
4mm
(c)
(a) (b)
Figure 6.23: SEM images of two phosphonate ester functionalized SBA-15 prepared by
prehydrolysis of functional silane, before and after polymer removal: (a) 15-P-60-F and
15-POEt-60-F; (b) 20-P-60-F and 20-POEt-60-F; (c) TEM image of 20-P-60-F exhibiting
pores running parallel to each other.
increasing the percentage of functional silane to φ=20, particles become more shorter. As
seen in the SEM images in Fig. 6.23b the samples 20-P-60-F and 20-POEt-60-F do not
form elongated rod-like particles, but rather broad hexagonal disks which tend to stick
90 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
together. In any case the disks constitute arrays of ordered pores as seen in the TEM
micrograph of the sample 20-P-60-F displayed in Fig. 6.23c.
Figure 6.24 shows SEM images of the samples 15-P-40-T and 15-POEt-40-T in which
TEOS was the prehydrolyzed silica precursor. The images show exclusively elongated
worm-like structures. However, the length of the worm-like particles is only about 1 µm,
i.e., are shorter than in pure SBA-15 samples that have particle lengths of several µm.
SEM images show that a simultaneously addition of TEOS and functional silane (φ=15)
15-P-40-T
15-POEt-40-T
4mm1mm
4mm1mm
Figure 6.24: SEM images of a phosphonate ester functionalized SBA-15 material pre-
pared by prehydrolysis of TEOS before polymer removal (15-P-40-T) and after polymer
removal (15-POEt-40-T).
leads to completely disordered silica materials. Further characterizations of this sample
were therefore not performed.
Nitrogen adsorption isotherms and SAXD profiles of phosphonate ester functional-
ized SBA-15 samples are displayed in Figure 6.25, and the results of the data analysis
are summarized in Table 6.7. The results are consistent with the conclusions drawn from
the SEM images. The type IV nitrogen adsorption isotherm of sample 10-POEt-60-F
resembles that of pure SBA-15. They exhibit a steep pore condensation step indicating
a narrow pore size distribution accompanied by a H1 type hysteresis loop. The mate-
rial has a high pore volume and large pore diameter. The SAXD profile of this sample
shows four resolved peaks that can be assigned to the 2D hexagonal unit cell of ordered
cylindrical pores. The nitrogen adsorption isotherms of the samples 15-POEt-95-F and
6.4. PHOSPHONIC ACID FUNCTIONALIZATION 91
0,50 0,75 1,00 1,25 1,50 1,75 2,00
(21)
(30)
(20)
(11)
15-POEt-95-F
10-POEt-60-F
20-POEt-60-F
15-POEt-40-T
Intensity/a.u.
(10)
q/nm-1
(b)
0,0 0,2 0,4 0,6 0,8 1,0
0
100
200
300
400
500
600
700
800
900
1000
1100
15-POEt-40-T
20-POEt-60-F
15-POEt-95-F
adsorbedvol.cm3g-1
relative pressure
10-POEt-60-F
(a)
Figure 6.25: Nitrogen adsorption isotherms (a) and SAXD profiles (b) of phosphonate
ester functionalized SBA-15 materials by prehydrolysis of the functional reagent. Ad-
sorption isotherms and SAXD spectra are shifted against each other for clarity.
20-POEt-60-F show features which are not typical for pure SBA-15. Their low-pressure
region corresponds to type IV isotherms with a not very steep pore condensation step ac-
companied by H1 type hysteresis loop. However, after this step the adsorption continues
to increase and is accompanied by a secondary hysteresis of type H3 which is typical for
agglomerates of disk-like particles. The sample 15-POEt-40-T, on the contrary, exhibits
a typical type IV isotherm with a H1 type hysteresis, but no significant further increase of
adsorption or a secondary hysteresis is observed. Hence the morphological characteris-
tics found in the SEM images are also seen in the nitrogen adsorption measurements. The
Table 6.7 shows that the specific surface area, the mesopore volume, and pore diameter
of the latter three samples are considerably smaller than for 10-POEt-60-F, whereas the
lattice parameter a0is approx. 12 nm for all phosphonate ester functionalized samples.
This implies that the pore walls of these samples are thicker than that of 10-POEt-60-F
and pure SBA-15. However, when increasing the percentage of functional silane φfrom
15 and 20 no pronounced changes in pore volume and diameter are observed, i.e., the
expected decrease in pore diameter and pore volume does not occur. Concerning these
parameters there are no pronounced differences between these samples. The larger wall
thickness of the PA-SBA-15 samples can be attributed either to a wider SiO2layer be-
92 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
Table 6.7: Properties of phosphonate ester-functionalized and pure SBA-15 silica.
sample as/m2g−1vp/cm3g−1vm/cm3g−1(p/p0)pc D/nm a0/nm w/nm
10-POEt-60-F 571 0.94 0.05 0.753 8.8 12.0 3.2
10-PA-60-F 595 0.95 0.05 0.758 8.8 12.0 3.2
15-POEt-95-F 530 0.79 0.03 0.676 7.2 12.3 5.1
20-POEt-60-F 520 0.73 0.06 0.694 7.6 12.1 4.5
15-POEt-00 disordered
15-POEt-40-T 502 0.64 0.04 0.691 7.5 12.0 4.5
15-PA-40-T 531 0.66 0.05 0.693 7.5 12.0 4.5
SBA-15cal550 561 0.89 0.10 0.757 8.9 11.5 2.6
SBA-15cal250 807 1.35 0.11 0.807 10.2 12.8 2.6
as: specific surface area, vp: pore volume, vm: micropore volume, (p/p0)pc: pore condensation pressure of
nitrogen, D: pore diameter, a0: lattice parameter.
tween two neighboring pores, or to the organic film of the functional groups at the pore
walls. Probably both effects play a role. A closer examination of SAXD profiles of the
samples displayed in Fig.6.25b reveals that the relative intensities of the leading peaks are
quite similar in the samples 15-POEt-95-F, 20-POEt-60-F and 15-POEt-40-T, but differ-
ent from the relative peak intensities in the SAXD spectrum of 10-POEt-60-F. Most likely,
this is due to the constant value of the lattice parameter a0but different pore sizes of these
two groups of samples. The lattice parameter determines the peak positions in the SAXD
spectrum. The pore diameter of the cylindrical pores has a direct influence on the form
factor which for cylindrical pores is given by Bessel functions. The latter exhibit sharp
maxima and minima depending on pore diameter. The intensity of the Bragg reflexes is
modulated by these maxima and minima of the form factor [74].
In the SAXS profiles of the samples 15-POEt-95-F, 20-POEt-60-F and 15-POEt-40-
T with similar pore sizes, the (11) and (30) Bragg-reflexes are purged because of the
minimum of Bessel function at this positions, whereas in the profile of 10-POEt-60-F
the (21) reflex is purged for the same reason. Thus the similarity of the SAXD profiles
of the samples 20-POEt-60-F and 15-POEt-40-T can be explained by their similar pore
sizes and lattice parameters. Hence, the relative intensities of the Bragg-peaks in a SAXD
spectrum are not only effected by long-range order of the samples but also by the relation
of lattice parameter and pore size (and film thickness).
6.4. PHOSPHONIC ACID FUNCTIONALIZATION 93
6.4.2 Reaction yield and degree of the functionalization
In this section the results of examination of the functional group content in the different
samples are presented. 29Si MAS NMR and TGA/DTA methods were used for the quan-
titative determination of the functional group content in the silica samples after template
removal. 29Si MAS NMR spectra and TGA/DTA curves of the phosphonate ester contain-
ing samples are displayed in Figure 6.26. Results of the NMR and TGA are summarized
in the Table 6.8.
150 300 450 600 750 900
80
85
90
95
100
15-POEt-60-F
20-POEt-60-F
15-POEt-40-T
10-POEt-60-F
TG(%)
T/°C
T33
T2
T2
10-POEt-60-F
20-POEt-60-F
15-POEt-60-F
T33
T2
T2
T33
T
2
-112.4-112.4112.4
Q4
Q3
Q2
-150-150 -200-200
-100-100
-50-50
00
5050 ppm
T33
T
2
15-POEt-40-T
(T/Q)=0.07
(T/Q)=0.15
(T/Q)=0.17
(T/Q)=0.19
(a) (b)
Figure 6.26: (a) 29Si MAS NMR spectra and (b) TGA curves of phosphonate ester con-
taining SBA-15 samples with increasing content of functional reagent φin the initial
mixture.
The gradual increase of the (T/Q) ratio in the spectra of samples synthesized by pre-
hydrolysis of the functional silane shows that an increase of the percentage of functional
reagent φin the synthesis mixture yields increasing numbers of functional groups at the
94 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
silica walls. This is vindicated by TGA measurements (see Table 6.8). Closer examina-
tion of the experimental (T/Q) and n values reveals that an increase of φfrom 10 to 15
effects an 100% increase of the number functional groups at the wall, while a further in-
crease of φfrom 15 to 20 causes an increase of functional groups by about 7% (NMR) or
about 20% (TGA) indicating that an upper limit of the functionalization degree has been
reached. Moreover, both methods show that when TEOS is the prehydrolyzed component,
as is the case for the sample 15-POEt-40-T, a slightly higher (T/Q) value is observed than
for the sample 15-POEt-60-F.
Table 6.8: Expected quantities and the corresponding experimental values obtained from
29Si NMR, TGA/DTA and potentiometric titration.
expected NMR TGA
sample (T/Q) n x (T/Q) Y n Y
10-P-60-F 0.11 1.66 0.07 16% 63%
10-POEt-60-F 0.11 1.66 0.065 17% 59% 0.63 38%
10-PA-60-F 0.11 1.66 18% 0.30
15-P-95-F 0.18 2.49 0.15 34% 85%
15-POEt-95-F 0.18 2.49 0.15 37% 84% 1.13 45%
20-POEt-60-F 0.25 3.32 0.17 31% 66% 1.39 42%
15-POEt-40-T 0.18 2.49 0.16 32% 87% 1.18 47%
15-PA-40-T 0.18 2.49 0.11 38% 63% 0.91 36%
(T/Q): Expected and experimental ratio of organosiloxane (T) and siloxane (Q) groups; n: Expected and
experimental amount of functional silane per unit mass of SiO2(in mmol g−1); x: Degree of surface
functionalization; Y: Reaction yield.
Though the NMR (Fig. 6.26a) and TGA (Fig. 6.26b) experiments show the same
tendencies the reaction yields calculated from these methods (Table 6.8) differ from each
other. Namely, NMR gives higher yields than the TGA. Reactions yields calculated from
NMR spectra range from 60% to 90% whereas those calculated from the TGA curves
range from 40% to 50%.
Also here we investigated whether the small yield of the TGA experiments is caused
by the formation of a crystalline Si-P mixed-oxide phase in the matrix during the heat-
ing. The presence of phosphor remnants in the matrix would increase its unit mass and
decrease the computed yield from TGA. The WAXS (wide angle X-ray scattering) spec-
tra measured after heating the samples to 1000◦C showed that the amorphous structure
6.4. PHOSPHONIC ACID FUNCTIONALIZATION 95
of the PA-SBA-15 indeed becomes partially crystalline. The reflexes of the crystalline
phase, however, could only be associated with ‘cristobalite’, the high-temperature modi-
fication of SiO2(see spectra in Appendix). In the crystalline phase they are no structures
containing phosphor. But this does not rule out the presence of phosphor in the amor-
phous phase. Here the presence of phosphor could be examined, e.g., through phosphor
elemental analysis. This, however, was beyond the scope of this thesis.
The reactions yields for phosphonic acid are higher than the corresponding values of
sulfonic acid functionalization but smaller than the reaction yields obtained for carboxylic
acid functionalization. For the sample 10-POEt-60-F the degree of the surface function-
alization, x, is 16% whereas for samples prepared with a higher percentage of functional
silane xranges from 30 to 40%. The latter values are similar to the values for surface
coverage xof sulfonic acid functionalized materials.
6.4.3 Polymer removal and microporosity
TGA curves of the polymer-containing samples 15-P-40-T, 20-P-60-F, 10-P-60-F are
shown in Figure 6.27. The contents of organic components in these samples are 28%,
36%, and 51%, respectively. The fact that the main part of the organic component forms
the block copolymer template indicates that the polymer content in the samples decreases
as the degree of surface functionalization increases. The same trend was also observed in
0 150 300 450 600 750 900
4 8
5 4
6 0
6 6
7 2
7 8
8 4
9 0
9 6
102
10-P-60-F
20-P-60-F
TG (%)
T/°C
15-P-40-T
Figure 6.27: TGA curves of polymer containing 15-P-40-T, 20-P-60-F, 10-P-60-F sam-
ples.
96 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
CA-SBA-15 and SA-SBA-15 materials. Hence, it appears that the number of the block
copolymer micelles decreases as the number of functional groups is increased.
Figure 6.28 compares the DTA curves of the polymer-containing phosphonate ester
functionalized sample (15-P-40-T), after polymer removal by sulfuric acid (15-POEt-40-
T), and after phosphonate ester cleavage by HCl treatment (15-PA-40-T). The exotherme
peak at ca. 100◦C is attributed to physically adsorbed water. The pronounced endother-
mic peak at 185◦C seen in the polymer-filled sample has disappeared after sulfuric acid
treatment. Therefore this peak is attributed to the block copolymer template. The peak
at ca. 300◦C has disappeared after HCl treatment, indicating that this peak is caused
by phosphonic acid ester groups. Finally, the DTA curve of the sample containing free
phosphonic acid groups (15-PA-40-T) shows only two peaks at 400◦C and 500◦C.
150 300 450 600 750 900
- 3 0
- 2 5
- 2 0
- 1 5
- 1 0
- 5
0
5
185°C
300°C
400°C
15-POEt-40-T
15-PA-40-T
DTA
T/°C
15-P-40-T
500°C
Figure 6.28: DTA curves of 15-P-40-T, 15-POEt-40-T and 15-PA-40-T.
Figure 6.29a shows the 13C CPMAS NMR spectra of 10-P-60-F, 10-POEt-60-F and
15-POEt-95-F. The spectrum of the polymer containing sample 10-P-60-F displays reso-
nances at 76, 74, 71, and 17 ppm which are attributed to the block copolymer template.
Two additional carbon signals at the positions 5 ppm and 63 ppm exists that are assigned
6.4. PHOSPHONIC ACID FUNCTIONALIZATION 97
to the functional group. After polymer removal with sulfuric acid no resonance signals
from block copolymer can be seen anymore, indicating the complete template removal.
150 300 450 600 750 900
-0,3
0,0
0,3
0,6
CH2-CH2-PO(OEt)2
P123
500°C
300°C
180°C
15-P-40-T
15-POEt-40-T
20-POEt-60-F
15-POEt-95-F
DTG
T/°C
10-POEt-60-F
75°C
H2O
15-POEt-95-F
3
P123
10-P-60-F
3,3’
1
P123+2+4,4’
1
ppm
150 100 50 0-50
10-POEt-60-F
2, 4
Si CH2
O
O
OCH2P
O
CH2
O
OH
CH3
12
3
4
Si CH2
O
O
OCH2
CH3
P
O
CH2
CH3
O
CH2
O
12
3'
4'
3
4
(a) (b)
Figure 6.29: (a) 13C CPMAS NMR spectra of the samples 10-P-60-F, 10-POEt-60-F and
15-POEt-95-F; (b) DTG curves of different phosphonate ester and phosphonic acid con-
tained SBA-15 samples.
The 13C CPMAS NMR spectra of 10-POEt-60-F and 15-POEt-95-F are very similar
representing three resonances at the positions 5 ppm, 17 ppm and 63 ppm. In Figure 6.29a
these were assigned to the different carbon atoms of the functional group. According to
these spectra, the samples after polymer removal still contain phosphonic acid ester. Thus
48% sulfuric acid is not strong enough to hydrolyze the phosphonic acid ester groups
completely. In Figure 6.29b DTG curves of different samples are displayed. The curves
show four peaks. The peak at 75◦C comes from physically adsorbed water. The peak
at 180◦C is attributed to the decomposition of block copolymer and the remaining two
peaks at 300◦C and 500◦C are assigned to the degradation of functional group. No peak
from polymer appears after sulfuric acid treatment. Thus the results from TGA are in
full agreement with those from 13C CPMAS NMR. As mentioned earlier, cleavage of
phosphonic acid ester is performed by refluxing the samples in concentrated HCl for 24h.
98 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
The 13C CPMAS NMR spectra of 15-POEt-40-T and 15-PA-40-T are displayed in
Figure 6.30a. As can be seen the peak at 63 ppm attributed to phosphonic acid ester
2, 4
3
100 50 0 -50
ppm
150
2
1
---
1
Si CH2
O
O
OCH2P
OH
O
OH
12
Si CH2
O
O
OCH2P
O
CH2
O
OH
CH3
12
3
4
T3
T2
Q2
Q3Q4
T3
T3
5050
T2
T2
00 -50-50 -100-100 -150-150 - 200-
ppm
15-POEt-40-T
15-PA-40-T
(T/Q)=0.19
(T/Q)=0.11
Figure 6.30: (a) 13C CPMAS NMR spectra and (b) 29Si MAS NMR spectra of the samples
15-POEt-40-T (before HCl treatment) and 15-PA-40-T (after HCl treatment).
groups disappears completely after HCl treatment for 24 h. Thus, SBA-15 containing
free phosphonic acid groups at the pore walls can be produced in this way. Figure 6.30b
shows the 29Si MAS NMR spectra of the samples. The (T/Q) ratio becomes smaller after
the treatment with HCl. The amount of functional groups in the pore walls decreases from
1.18 mmol g−1to 0.91 mmol g−1after HCl treatment. The respective data are included
in Table 6.8. According to these results, ca. 20% of functional groups are cleaved after
HCl treatment. Moreover, the number of Q2and Q3Si atoms relative to Q4decrease
after HCl treatment indicating further condensation of silicon species. For this reason
despite the lost of functional groups, the degree of surface functionalization x, given
by x=T2+T3
Q2+Q3+T2+T3, increases from 32% to 38% in the sample 15-PA-40-T. On the
other hand, nitrogen adsorption and SAXD data of 15-POEt-40-T and 15-PA-40-T are
very similar (see Table 6.7). The data show that the lattice spacing and pore size remain
constant whereas pore volume and specific surface area slightly increase in the case of 15-
PA-40-T. The micropore volume of the samples was determined by t-plot analysis (see
Figure 6.31). The values are ranging from 0.03 to 0.06 cm3g−1. Hence, though they are
bigger than the micropore volumes of CA-SBA-15 and SA-SBA-15 materials, they are
6.4. PHOSPHONIC ACID FUNCTIONALIZATION 99
0 , 0 0 , 2 0 , 4 0 , 6 0 , 8 1 , 0
0
5 0
100
150
200
250
300
a d s o r b e d v o l . c m 3g- 1
t ( H & J ) / n m
1 0 - P O E t - 6 0 - F
1 5 - P O E t - 9 5 - F
2 0 - P O E t - 6 0 - F
1 5 - P O E t - 4 0 - T
1 5 - P A - 4 0 - T
Figure 6.31: t-plot analysis of nitrogen adsorption isotherms for different phosphonic acid
ester functionalized SBA-15 samples and for a SBA-15 sample containing free phospho-
nic acid groups (15-PA-40-T).
still smaller than those of native SBA-15.
6.4.4 The mechanism of the self-assembly
The diethoxyphosphorylethyltriethoxysilane (PTES) has two rather hydrophobic phos-
phone ester groups attached to the triethoxysilane via an ethyl group (see Fig. 1.1). Due
to the strong P-O bond the phosphonic acid ester groups are stable in the initial synthesis
conditions and thus the dealkylation does not take place. Hence, one expects that they
are adsorbed close to the hydrophobic PPO part of the block copolymer. However, under
the same reaction conditions the organosiloxane groups are hydrolyzed partially. There-
fore they are hydrophilic and can interact with the hydrophilic PEO moieties. Because
of the similar solubilities of PTES in PEO and PPO blocks of the block copolymer these
molecules may cause reduced microphase separation of the two blocks and thus a de-
creased mesoporous ordering. Such a behavior was proposed by Margolese et al. in the
preparation of propyl-thionyl functionalized SBA-15 using mercaptopropyltrimethoxysi-
lane [53]. This functional silane is rather hydrophobic and thus resembles the functional
silane PTES used in the present work.
Our investigations showed that when PTES is the prehydrolyzed silica source the pres-
ence of a molar percentage of functional silane φgreater than 10 in the co-condensation
100 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
perturbs self-assembly of block copolymer aggregates since the worm-like structure of
SBA-15 disappears and stuck-together particles are generated. When TEOS is allowed
to prehydrolyze the initial stage in nanoparticle formation involves the formation of in-
dividual silica coated P123 aggregates and materials with improved mesoscopic order
are formed. In this way it is possible to increase φup to 15 by retaining the worm-like
mesostructure of SBA-15.
6.5. ACIDITY MEASUREMENTS 101
6.5 Acidity measurements
6.5.1 Measurements in non-aqueous environment by 15N NMR
Acid-base interactions in aqueous solutions can be described in terms of conventional
chemical reaction equilibria. This situation is quite different from that of acid-base inter-
actions in ‘dry’ systems, when the acid and base are in direct contact via hydrogen bonds.
Solid state NMR is a method which can be applied to investigate such interactions. Proton
acceptors such as pyridine can be used to probe the proton donor ability of Brønsted acids.
For instance, Shenderovich et al. have characterized the interactions of 15N-labeled pyri-
dine, a simple proton acceptor, with silica hosts using solid state NMR spectroscopy [5].
Earlier 15N MAS solid-state NMR experiments showed that the isotropic 15N chemical
shift of pyridine is modulated when the molecule is incorporated at different sites of acid
zeolites [79] or amorphous alumina [80]. In particular, a strong high-field shift occurs
when pyridine is protonated.
Low-temperature NMR studies of the influence of hydrogen bonding and protonation
of pyridine on NMR parameters have shown that for 1:1 hydrogen bonded complexes
of an acid (AH) with pyridine (B) an increase of the acidity of AH leads to a gradual
transformation of the molecular hydrogen bonded complex A-H···B via a proton-sharing
complex Aδ−···H···Bδ+to a zwitterionic complex A−···H-B+, where Aδ−···H···Bδ+
exhibits the shortest distance A···B. In order to understand the influence of a varia-
tion of the relative acidity on the acid-base hydrogen bonds, Lorente et al. carried out
low-temperature measurements on a series of hydrogen-bonded crystallized 1:1 acid-base
complexes of 15N-labeled collidine (2,4,6-trimethylpyridine) with a number of different
carboxylic acids and their deuterated analogues [81]. They observed a correlation of the
15N isotropic and anisotropic chemical shifts with the dipolar 1H-15N couplings and hence
the nitrogen-hydrogen distances and pKavalues. A graph showing this correlation, taken
from the work of Lorente et al., is displayed in Figure 6.32b. Using this correlation, the
O-H···N distances in the pyridine and surface hydroxyl pairs of MCM-41 and SBA-15
silicas were determined [5]. The O-H···N distance found in the solid state corresponds to
acids exhibiting in water a pKaof about 4. Thus, the surface hydroxyl groups of MCM-41
and SBA-15 materials are weakly acidic.
In Figure 6.32a the 15N NMR spectra of our carboxylic acid, phosphonic acid and sul-
fonic acid functionalized SBA-15 materials obtained after 15N-pyridine adsorption into
the pores are shown. The peak at ca. 0 ppm is assigned to an excess of free (not bonded)
pyridine in the pores. The peak at ca. -25 ppm is assigned to pyridine hydrogen bonded
102 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
10-PA
0 -50 -100 -150 ppm
50
20-CA
20-SA
(a) (b)
Figure 6.32: (a) 15N MAS NMR spectra of 20-CA, 10-PA (10-POEt-60-F) and 20-SA
(20-SA-60-F) after pyridine adsorption in the pores; (b) pKavalues of different acids AH
as a function of 15N chemical shift in acid-collidine solid-state complexes. The graph is
taken from the work of Lorente et al. [81].
to the surface SiOH groups and in case of CA-SBA-15 also to the single COOH groups
(Aδ−···H···Bδ+state). The peak at -110 ppm is attributed to pyridine completely pro-
tonated from acid groups (A−···H-B+state). Figure 6.32a shows that this peak is most
pronounced in the case of 20-SA, in agreement with the expected highest acidity of the
sulfonic acid groups, and weakest but still clearly recognizable in the case of CA-SBA-15.
The existence of totally protonated pyridine is unexpected in the case of carboxylic-acid
functionalized SBA-15 which is a weak acid in aqueous solution (acid dissociation con-
stant pKa=4.7). The surface silanol groups have a similar acidity as carboxylic acids
in water but are not protonated by pyridine in the dry state at the pore wall. The high
proton-donor ability of carboxylic acid functional groups at the pore walls may be due
to the formation of hydrogen-bonds between several carboxylic acid groups at the sur-
face. It is known that in solutions the formation of hydrogen-bonded chains of carboxylic
acid molecules leads to an enhanced acidity of the terminal proton [7]. The high local
concentration of carboxylic acid groups in the CA-functionalized SBA-15 material (more
than 40% of the surface, see Section 6.2) may lead to a similar conjugation effect in the
pores. Additional 15N NMR measurements are being performed to check this conjecture.
6.5. ACIDITY MEASUREMENTS 103
According to the present state of our study all three solid acids can be deprotonated by
pyridine. The pKavalues corresponding to the 15N chemical shifts for the protonated
pyridine indicate a high proton donor ability of all three solid acids. The acidity measure-
ments of the acid-functionalized SBA-15 silicas reported in this section were performed
at Free University Berlin as a part of the PhD thesis of Daniel Mauder. Accordingly, the
results will be discussed more detailed in his PhD thesis [82].
6.5.2 Measurements in aqueous environment by FT-IR
The proton donor ability of the COOH groups of 10-CA and 20-CA materials in aqueous
solutions was studied by means of FT-IR measurements. Experimental details of these
measurements were given in Section 4.2. FT-IR spectra of 10-CA and 20-CA samples in
the dry state are shown in Figure 6.33. Drying was performed by exposure of the samples
to dry air in the sample cell. The drying time for 20-CA and 10-CA was one week and 2
days, respectively. Because of the shorter drying time a peak at ca. 1633 cm−1appears in
the FT-IR spectrum of 10-CA, which is attributed to the O-H bending mode (δ(O-H)) of
H2O.
Figure 6.33 shows two spectral ranges of particular interest, namely the C-H stretching
4000 3600 3200 1800 1650 1500 1350
ν(C -O -O )a s
ν( C = O )
1 0 - C A
d r y
ν/ c m - 1
absorbance
2 0 - C A
Figure 6.33: FT-IR spectra of 10-CA and 20-CA samples in the dry state.
region from 2900 to 3800 cm−1, and the finger-print region from 1800 to 1250 cm−1. The
weak band at 3740 cm−1is due to quasi isolated SiOH groups, whereas the broad band
104 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
centered at 3319 cm−1is due to hydrogen-bonded silanols. These interactions can come
from hydrogen bonding of SiOH groups with each other, or with the COOH groups at the
surface, or with residual H2O physisorbed at the surface. The weak band at 2979 cm−1,
assigned to the CH3stretching mode, indicates the presence of residual amounts of block
copolymer template. These residues contribute also to the bands at 2939 and 2916 cm−1.
The latter are attributed to the CH2stretching mode and come mostly from ethyl chains
of the functional groups. The existence of small amounts of residual template in these
samples was also seen by means of 13C CPMAS NMR measurements (Section 6.2).
In the finger-print region, a strong band at 1718 cm−1is assigned to the C=O stretching
mode, ν(C=O), of the COOH head group. Concomitant with this band, a band at 1413
cm−1appears, which is assigned to the C-O stretching mode of the carboxyl group, that
coupled to the O-H bending mode.
The dependence of the proton-donor ability of the 10-CA and 20-CA silicas on the
pH of the solution was investigated by exposing the samples to pure D2O of pH 5.5 and
to a phosphate buffer in D2O of pH 8. The resulting spectra in the finger-print region are
shown in Figure 6.34 in comparison with the spectra of the dry materials. The fact that
1800 1700 1600 1500 1400 1300
n(COO-)s
n(C=O)
10-CA
pH5.5
dry
pH8
absorbance
n/cm-1
n(COO-)as
1800 1700 1600 1500 1400 1300
n(C=O)
n(COO-)as n(COO-)s
pH8
dry
pH5.5
n/cm-1
absorbance
20-CA
(a) (b)
Figure 6.34: FT-IR spectra of (a) 20-CA and (b) 10-CA samples in dry state, in D2O with
pH 5.5, and in phosphate buffer (K3PO4K2HPO4) in D2O with pH 8.
spectra show nearly no difference between the dry state and in D2O at pH 5.5, indicates
that exposure to D2O at pH 5.5 does not cause significant deprotonation of the COOH
groups. After exposure to the D2O solution of pH 8, two new bands at 1557 and 1387
cm−1appear which are attributed to the asymmetric (νas) and symmetric (νs) stretching
vibration of carboxylate groups (COO−), respectively. At the same time the intensity
6.6. CONCLUSIONS 105
of the band at 1718 cm−1decreases. These changes indicate that a deprotonation of
COOH groups has occurred. However, the decrease of the intensity of the band at 1718
cm−1does not occur to the same extent in the 10-CA and 20-CA materials. In 10-CA the
signal nearly disappears indicating that most of the COOH groups have been deprotonated
whereas the signal intensity is reduced only to about one half in case of 20-CA. This
suggests that in the 10-CA material with 20% surface coverage nearly all COOH groups
are separated and accessible, while in the 20-CA material with 50% surface coverage only
part of the COOH groups are accessible.
6.6 Conclusions
Co-condensation route was used for the functionalization of SBA-15 with carboxylic acid
(CA), phosphonic acid (PA) and sulfonic acid (SA) groups in order to study the acidity in
the confined geometry of the nanopores. Reaction conditions for maximizing functional-
ization of SBA-15 under retention of the structural order of the silica are strongly affected
by assembly kinetics. To understand the involved hydrolysis and the condensation of sil-
ica sources and their interaction with template, the percentage of the functional silane φ,
the prehydrolyzed silica source (TEOS or F) and the prehydrolysis time are varied.
In case of carboxylic acid functionalization using CTES hexagonally ordered mate-
rials are obtained only if CTES is prehydrolyzed. The increase of the functional silane
up to 20 mole% is possible without disordering the SBA-15. The yield of the function-
alization is approx. 80% indicating a substantial incorporation of the functional silane.
The situation is different in the phosphonic acid functionalization. Applying functional
silane PTES as the prehydrolyzed component, ordered SBA-15 materials with worm-like
morphology can be obtained up to φ= 10. Increasing φto 15 and 20 leads to a stepwise
shortening of the worm-like particles up to the formation of disk-like structures under
preservation of the 2D hexagonally order. The amount of PTES in the synthesis mixture
can be increased to φ=15 mole% under retention of the elongated worm-like morphology
of SBA-15 by prehydrolyzing of TEOS. The simultaneous addition of PTES and TEOS in
the synthesis mixture, however, leads to the formation of disordered materials. The sur-
face coverage of SBA-15 by phosphonic acid groups ranges from 30 to 40%. The yields
of the functionalization reactions obtained from 29Si-MAS NMR measurements are high
(60-90%). Nevertheless, the yields decrease nearly 50% after the additional treatment
of the products (SBA-POEt) with concentrated HCl to cleave the phosphate-ester groups
for producing free phosphonic acid species. In the synthesis of sulfonic acid functional-
106 CHAPTER 6. ACID-FUNCTIONALIZATION BY CO-CONDENSATION
ized SBA-15 the prehydrolysis of STHS does not play a role because it already exists in
the hydrolyzed form. As a hydrophilic functional reagent because of the sulphonic acid
groups it preferentially interact with the hydrophilic PEO part of the block copolymer al-
lowing a good microphase separation. Accordingly, an amount of STHS in the synthesis
mixture up to φ=20 mole% yields SBA-15 materials with the typical well-known order
and morphology, and this independently of the silica source (TEOS or STHS) added first.
Nevertheless, the degree of the incorporation of the sulfonic acid reagent using STHS is
the lowest among other functionalization reactions. Increasing the mole fraction of func-
tional silane φto 30% results in formation of disordered silica. The surface coverage of
SBA-15 by sulfonic acid groups ranges from 20 to 30%. The yields of the functional-
ization reactions obtained from 29Si-MAS NMR measurements are high (30-50%). The
lower degree of the incorporation of the STHS compared to the other functionalization
reactions can probably be explained by the slow condensation rate of the hydroxyl groups
of the highly acidic functional silane.
The acidity measurements of functionalized SBA-15 materials in water free condi-
tions were performed via 15N CPMAS NMR after adsorption of pyridine into the pores.
The pKavalues corresponding to the 15N chemical shifts for the protonated pyridine in-
dicate a high proton donor ability of all three solid acids. The proton donor ability of the
COOH groups of 10-CA and 20-CA materials in aqueous solutions was studied by means
of FT-IR measurements. These show the deprotonation of COOH groups in 10-CA and
20-CA materials at pH 8. The results suggest that in the 10-CA material with 20% sur-
face coverage nearly all COOH groups are separated and accessible, while in the 20-CA
material with 50% surface coverage only part of the COOH groups are accessible.
Part II
Functionalized SBA-15 Silica as a Host
107
Chapter 7
Background
7.1 Supramolecular polymers
Loosely defined, supramolecular polymers are polymers in which the monomers are held
together by weak noncovalent reversible interactions. Noncovalent interactions with little
specificity or directionality are present in all condensed molecular materials. However,
when highly directional forces dominate the interaction between neighboring molecules,
long chains or networks of concatenated molecules can be formed, resulting in many
of the (mechanical) properties that have made polymeric materials so successful. Long
chains, which lead to polymer-like behavior, are only formed when the interactions be-
tween the monomeric units are strong enough. The presence of linear chains, which
persist when a material is heated or dissolved, is what guarantees a successful design of
strong and directionally interacting functionalities. In a fundamental research context,
where the goal is to understand the relation between molecular structure and macroscopic
properties, strength and directionality are of prime importance. For supramolecular poly-
mers there are many examples in nature. For example, globular proteins in biological
cells, such as tubulin, actin, and fibrinogen, are known to polymerize into long filaments
by means of non-covalent interactions. This polymerization plays an important role in the
cell rigidity and motility, and in intercellular transport.
Another well-known example of supramolecular assembly is that of amphiphilic sur-
factant molecules or polymers. These molecules can form a variety of aggregate structures
such as isotropic micellar and vesicle phases, or hexagonal and lamellar mesophases. As
we have seen in Part I such systems can be used as structure directing agents for the
synthesis of ordered mesoporous materials. In order to create a linear supramolecular
polymer, bifunctional monomers are needed possessing two binding sites that have a spe-
109
110 CHAPTER 7. BACKGROUND
cific interaction with other binding sites. Most design of supramolecular polymers are
based on hydrogen bonding [83]. Due to their specific and highly directional character,
hydrogen bonds are ideally suited for the assembly of supramolecular polymers. The
strength of the interactions correlates with the number hydrogen bonds of a bonding site.
An important further class of supramolecular polymers is formed by coordination
polymers. In a coordination polymer the bonds between monomers are based on metal
ligand interactions as shown in Figure 7.1 in the following Section.
7.1.1 Fe-MEPE
The supramolecular coordination polymer Fe-MEPE studied in this work is an organometal-
lic complex assembled from Fe(II)-acetate and 1,4-bis-(2,2´:6´2´´-terpyridine-4´-yl)-benzene
by coordinative bonds (see Figure 7.1) which results in a stereochemically well-defined
octahedral coordination with D2dsymmetry [84]. Because of the ditopic character of
the ligand, Fe-MEPE forms a linear rigid-rod-like structure (1.2 nm diameter) in which
the Fe(II) ions are periodically positioned along the backbone at a distance of 1.55 nm.
The rigidity results from the short length of the spacer, in our case phenyl, by which the
terpyridine ligands are connected. More flexible coordination polymers are obtained by
using longer spacer chain lengths.
(1)
Fe-MEPE (2)
(1)
Fe-MEPE (2)
Figure 7.1: Metallo-supramolecular coordination polyelectrolyte (Fe-MEPE) (2) formed
by self-assembly of Fe(II) ions with the ditopic ligand 1,4-bis-(2,2´:6´2´´-terpyridine-4´-
yl)-benzene (1).
Fe2+ having the electron configuration [Ar]3d6has 24 electrons. In an octahedral
ligand field the 6 valence electrons in the 3d orbitals split to form two sets of degenerated
orbitals. These are the t2gorbitals dxz, dxy, dyz, and the egorbitals dz2and dx2−y2as
illustrated in Figure 7.2. The energy difference between the t2gand egorbitals depends on
7.1. SUPRAMOLECULAR POLYMERS 111
E
dxy, dxy, dyz
dz2, dx2-dy2
spherical field octahedral field
t2g
eg
+0.6Do
-0.4Do
Do
Figure 7.2: Splitting of the d-orbitals of Fe2+ in an octahedral ligand field.
the strength of the ligand field and is denoted by ∆O, the ligand field splitting. Relative to
the spherical ligand field, t2gorbitals are lowered by 0.4∆O, and the egorbitals increased
by 0.6∆O. Accordingly, for an octahedral complex with electronic configuration tm
2gen
gthe
ligand field stabilization energy (LFSE) is given by
LFSEokt = (−0.4m+ 0.6n)∆O.(7.1)
In its ground state, Fe2+ forms a low-spin complex with electronic configuration t6
2ge0
g.
Thus the LFSE of Fe(II)-MEPE is −2.4∆O.
The transition metal complexes are most stable when they fulfill the 18-electron rule.
according to this rule, those configurations are stable in which the sum of the metal va-
lence electrons and the ligand electrons offered in dative bonds exactly add up to the
valence electron number of next following noble gas atom. In Fe-MEPE each Fe(II) is co-
ordinated by 6 nitrogen atoms of two terpyridine ligands. Because Fe(II) has 6 d-electrons
and each nitrogen atom contributes 2 electrons into the coordinative bond, Fe(II)-MEPE
fulfil the 18 electron rule, making it an especially stable complex.
For the octahedral coordination of the Fe(II) ion, contorting of the two terpyridine
molecules by 90◦against each other is necessary. This is illustrated in Figure 7.3. In this
coordination not all Fe-N bond distances are equal: The bond distances of the two N-
atoms in the central pyridine rings is 1.89 Å, while the bond distance of the four N-atoms
of the terminal pyridine rings is 1.99 Å[86]. These values correspond to an octahedral co-
ordination geometry with two N-atoms being in the central and four being in the terminal
positions of the octahedron. In Fe-MEPE films obtained by drying of MEPE-solutions at
solid surfaces like quartz the octahedral coordination geometry becomes distorted, lead-
ing to a partial paramagnetic high-spin complex of Fe(II)-MEPE. Magnetic susceptibility
measurements of such Fe(II)-MEPE films at room temperature has shown that the 15-30
112 CHAPTER 7. BACKGROUND
N
N
N
N
N
N
N
N
N
N
N
N
Fe2+
(a) (b)
Figure 7.3: (a) Coordination of Fe2+ with ditopic ligand 1,4-bis-(2,2´:6´2´´-terpyridine-
4´-yl)-benzene in an octahedral ligand field; (b) real space illustration of Fe-MEPE in
which the two terpyridine molecules are contorted by 90◦against each other [85].
% of Fe-complex is transformed to the paramagnetic high-spin state [87].
7.2 Lifetime of a supramolecular bond
Supramolecular polymers consist of bifunctional monomers that each have two functional
groups F. If a functional group of one monomer is linked to that of another monomer a
bond Bbetween these two monomers is formed [83].
···F+F··· ka
GGGGGGB
FGGGGGG
kd···B···
where kaand kdare the rate constants of bond formation and dissociation, respectively.
The dynamic properties of supramolecular polymers are mainly determined by the life-
time of a bond
τB=1
kd
.(7.2)
τBis practically infinite for covalent polymers. For supramolecular polymers however, it
is usually much smaller. It may vary from less than a nanosecond up to thousand of years,
depending on the type of supramolecular bond.
7.2.1 Lifetime of a coordinative bond
As explained above, in coordination polymers each metal ion (M) is coordinated by two
ligand groups (L) in form of a complex:
7.2. LIFETIME OF A SUPRAMOLECULAR BOND 113
M+Lka1
GGGGGGB
FGGGGGG
kd1
ML
ML +Lka2
GGGGGGB
FGGGGGG
kd2
ML2
with the stability constants
K1=[ML]
[M][L]and K2=[ML2]
[ML][L](7.3)
The ML complexes and free ligand groups L form the chain end whereas the ML2com-
plexes constitute the bonds between monomers. The dissociation rate of ML2complexes
kd2determines the lifetime of a bond. It is given by kd2=ka2/K2where K2is the stability
constant of an ML2complex.
In most cases for ML complexes in aqueous solutions the binding of a ligand to a
metal ion involves the rapid formation of an intermediate outer sphere complex, which is
an ion pair between the ligand and hydrated metal, and the expulsion of a water molecule.
Because of the slowness of the latter the formation of ML depends on the water-exchange
rate of the metal ion [88]. Accordingly, this is the rate limiting step and the rate constant
of complex formation can be written as
ka1=Kos
k−w
(7.4)
where k−wis the rate constant of dehydration and Kos is the equilibrium constant of the
intermediate outer sphere complex. The dehydration rate is characteristic for a given
metal ion at a given temperature. Table 7.1 presents the approximate dehydration rates
of some ions, taken from Morel and Hering [89]. The equilibrium constant Kos primarily
Table 7.1: Approximate rate constants of dehydration for different metal ions at room
temperature [89].
metal ion k−w/s−1metal ion k−w/s−1
Pb2+ 7.109Fe2+ 4.106
Cu2+ 1.109Ni2+ 3.104
Cd2+ 3.108Fe3+ 2.102
Zn2+ 7.107Cr3+ 5.10−7
depends on the charges of the ligand and the metal ion and on the ionic strength. It can be
114 CHAPTER 7. BACKGROUND
calculated from statistical considerations [89]. For divalent metal ions and neutral ligands
such as Fe2+ and 1,4-bis-(2,2´:6´2´´-terpyridine-4´-yl)-benzene used for the formation of
Fe-MEPE in this work, Kos is approximately 3.10−1M−1[83]. Thus, the rate constant
ka1is primarily determined by the metal ion and not by the ligand. For an ML2complex,
the association rate constant ka2is determined by removal of a water molecule from an
ML complex. This is generally faster than the dehydration of a free metal ion, so that
ka2> ka1. Thus, Equation 7.4 gives a lower limit for ka2. A rough estimate yielding
a upper limit for the life time of a coordinative bond can be obtained as τB= 1/kd2≈
K2/(k−wKos). As can be seen from the values of k−wpresented in Table 7.1 the kinetics
of dehydration varies strongly from one metal ion to another. Because of these large
differences the properties of coordination polymers depend strongly on the metal ion used.
The dynamics also changes drastically if the valence state of the ion changes, as seen in
the Table by comparison of Fe2+ and Fe3+.
In the literature the stability constants K1and K2for the Fe(II)-MEPE complex are
given as 107M−1and 1014 M−1, respectively. These values were determined in 1966
by Holyer et al. for the monotopic ligand Fe-Terpy (Fig. 1.2b) by means of stopped-
flow measurements [88] and are adopted also for Fe-MEPE. We note that although the
association rate constant ka2> ka1for the reasons explained above, for most coordination
complexes the stability constant K2< K1because the first ligand group causes some
repulsions for the second ligand group. The value for K2of 1014 M−1for Fe(II)-Terpy
given by Holyer et al. is unusually high and new measurements are needed for verifying
this value. On the other hand, these values were commonly used in the literature for
calculations of the chain length of Fe-MEPE in solution [14,90]. A lower value of K2
would result in a lower estimate of the chain length of Fe-MEPE.
7.2.2 Living polymers
The behavior of supramolecular polymers depends strongly on the lifetime of a bond τB
as compared to the time scale of interest t. For τBt, polymers behave as ordinary,
covalent polymers with a fixed length distribution. For τBt, however, every functional
group changes frequently between the free and bonded state within a time t, so that a
state of dynamic equilibrium is reached. If this is the case, the supramolecular polymer
represents a living polymer (or equilibrium polymer).
In many cases, living polymers have the characteristics of ordinary polymers such as
low osmotic pressure, high viscosity and viscoelasticity at higher concentrations. On the
other hand, the possibility of the chains to break and recombine on experimental time
7.3. SORPTION 115
scale also introduces new features. For instance, the chain length distribution is not fixed
but changes with variable conditions. Because of this, tuning the properties of these poly-
mers is possible by varying the concentration, the solvent, the pH and the temperature.
Whereas living polymers exhibit typical polymer properties at low temperature, such as
a high viscosity and elasticity, the chains break into shorter units at elevated temperature
and behave more liquid-like. Another difference with ordinary polymers is the thixotropic
behavior of living polymers: When a solution of living polymers is sheared, some bonds
may break, leading to a lower viscosity. When the shear is stopped, the bonds are restored
and the viscosity reaches its initial state again after some time.
Fe-MEPE used in this work is a water-soluble living polymer consisting of coordi-
native bonds. It shows all characteristics mentioned above. These are examined in the
group of Kurth and Möhwald in the Max-Planck Institut für Kolloid- und Grenzflächen-
forschung in Potsdam/Golm. In the present work the uptake of Fe-MEPE in the mesopore
channels of SBA-15 by adsorption from solution has been studied. Basic aspects of the
adsorption from solution are introduced in the next section.
7.3 Sorption
Liquids and solids can bind molecules from their environment. This process is referred
to as sorption. The substances can penetrate into the interior of the condensed phase
(sorbent) or can accumulate at its surface. The former process called absorption, the
latter process adsorption. In that case, a gas or component of liquid mixture forms a
molecular or atomic film at the surface. Thus, sorption is a generic name for adsorption
and absorption [91]. Following Falbe [92] the term sorption can be extended to a general
concept encompassing adsorption, absorption, ion exchange, and precipitation. In all
these processes, a substance is taken up (immobilized) by another substance that is in
contact with it. The general term sorption is used if the immobilization is based on more
than one of these processes or is not precisely known. Below, some forms of sorption that
are relevant in this work will be introduced.
7.3.1 Adsorption
Adsorption is the most important process among all forms of sorption and represents an
accumulation of substances at the surface of a solid or liquid. Let us first introduce the
key definitions. The material in the adsorbed state is called adsorbate and the substance to
116 CHAPTER 7. BACKGROUND
be adsorbed is called the adsorptive. The substance, onto which adsorption takes place, is
the adsorbent. In general adsorption is an exothermic process (∆H < 0). Depending on
the magnitude of the adsorption enthalpy and on the type of bonding involved, adsorption
can be classified as follows:
1. Physical sorption or physisorption. Here the adsorbate is held to the surface by
relatively weak van der Waals forces and multiple layers may be formed with nearly
the same heat of adsorption. The heat of physisorption of gases and vapors is of the
same order of magnitude as the heat of condensation of the vapor. Therefore this
type of adsorption is stable only at temperatures below 150◦C.
2. Chemical sorption or chemisorption. This involves an exchange of electrons be-
tween specific surface sites and adsorbed molecules and, as a result, a chemical
bond is formed. Chemisorption is characterized by interaction energies between the
surface and adsorbate comparable to the strength of chemical bonds and is therefore
much stronger and stable up to higher temperatures than physisorption.
7.3.2 Electrostatic sorption (ion exchange)
At a fist glance, adsorption and ion exchange are two different phenomena with diverse
characteristics. Nevertheless ion exchange is similar to adsorption, since mass transfer
from a fluid to a solid phase is common in both process, i.e. they are both diffusion pro-
cesses. Ion exchange is a term reserved for Coulomb attractive forces between ions and
charged functional groups. The exchange of ions of the same charge sign takes place
between a solution and an insoluble solid in contact with it, or between two immisci-
ble solvents one of which contains a soluble material with immobilized ionic groups. In
most cases the term is used for purification, separation, and decontamination processes
of aqueous and other ion-containing solutions with solid polymeric or mineralic ‘ion ex-
changers’. Commonly, the latter are solid materials (stationary phase) capable of taking
up charged ions from a solution and releasing an equivalent amount of other ions into
the solution (mobile phase). The ability to exchange ions is due to the properties of the
structure of the materials. The exchanger consists of a so-called matrix with positive or
negative excess charge. This excess charge is immobilized in specific locations on the
stationary phase and is compensated by the counter ions, which can move within the free
space of the matrix and can be replaced by other ions of equal charge sign [93]. Typical
ion exchangers are ion exchange resins (functionalized porous or gel polymers), zeolites,
7.4. SORPTION EQUILIBRIUM 117
clays, and soil humus. Normally, an exchanger has a large surface area and pores of vari-
able size and shape. Few inorganic exchangers contain pores of uniform cross section.
An example for this kind of exchangers are zeolites. Ordered mesoporous silicas such as
MCM-41 or SBA-15 containing aluminum in the framework also have good application
potentials as an exchanger in separation processes. Ion exchangers can be unselective or
have binding preferences for certain ions or classes of ions depending on their chemical
structure.
The characteristic difference between adsorption and ion-exchange is that the latter
is a stoichiometric process, i.e., for every ion that is removed, an ion of the same sign is
released into the solution. Adsorption from solution can also be considered as a displace-
ment process at the surface, but this is not strictly stoichiometric as in ion exchange.
Ion exchange can be seen as a reversible reaction involving chemically equivalent
quantities [94]. A typical ion exchange is represented in the following equation
R−A++B+=R−B++A+(7.5)
In Equation 7.5 R−, represents the matrix. The ion exchange takes place if the matrix
has a selectivity for B+ions. The cation A+, which was originally balancing the negative
charge on the matrix, must not be too strongly held by the matrix as otherwise no exchange
will occur. The process of Eq. 7.5 form the basis of cation exchange chromatography. If
the matrix contains cationic sites, it is capable of exchanging anions, and the process is
applied in anion exchange chromatography. The characterization of an ion exchange as a
chemical process is rather misleading. Ion exchange is in principle a redistribution of ions
between two phases by diffusion, and chemical factors are less significant or even absent.
For this reason the heat evolved in the course of ion exchange is usually very small to
negligible, often less than 2 kcal mol−1[93]. Only when an ion exchange is accompanied
by a reaction such as neutralization can the the process be characterized as ‘chemical’.
7.4 Sorption equilibrium
7.4.1 General description
Adsorption from solution onto inert surfaces represents a displacement process at the sur-
face and follows the general principles of thermodynamic reaction equilibria. If there is
only one adsorbable component in the solution, as is the case in this work, the equilibrium
can be characterized by three factors. These are the equilibrium concentration in the solu-
tion, the surface concentration at equilibrium, and temperature. Accordingly, at constant
118 CHAPTER 7. BACKGROUND
temperature one has the equilibrium condition
aeq =f(ceq)T=const (7.6)
where aeq is some measure of the surface concentration of the sorptive at equilibrium, and
ceq is the equilibrium concentration of sorptive in the solution. After achieving equilib-
rium at a given temperature, a plot of the surface concentration of sorbent vs. the equi-
librium concentration of solution yields the so-called sorption isotherm. In an adsorption
process, if the adsorbent represents a powder of well-defined specific surface area, the
equilibrium amount of adsorbent is commonly expressed as the amount per unit mass of
adsorbent (specific adsorption) and is obtained from
aeq =V
m(c0−ceq),(7.7)
where c0is the initial concentration in solution, Vis the volume of the solution, and mis
the mass of the solid (adsorbent).
7.4.2 Isotherm equations
Experimental adsorption isotherms may be represented by a mathematical isotherm equa-
tion, but there is no single equation universally describing all experimental isotherms.
Instead, a series of different equations exits and the applicability of them is to be checked
on a case-by-case basis. Three simple isotherm equations are commonly used in adsorp-
tion from solution: Henry equation, Freundlich equation, and Langmuir equation. These
isotherms are sketched in Figure 7.4.
7.4.2.1 Henry adsorption isotherm
The Henry equation assumes a linear relation between the loading of the adsorbent and
the equilibrium concentration in the solution. Accordingly, the adsorbed amount of a sub-
stance per unit of adsorbent (aeq) is directly proportional to the equilibrium concentration
of the solution (aeq) and depends only on the so-called Henry coefficient (KH).
aeq =KH·ceq.(7.8)
Thus the Henry coefficient is the ratio of the loading of the adsorbent and the equilibrium
concentration of the solution and is constant for an achieved equilibrated state. Henry
isotherm applies to any adsorption equilibrium as a limiting law for low concentrations,
i.e,
lim
c→0(aeq/ceq) = KH(T).
7.4. SORPTION EQUILIBRIUM 119
ce q
ae q
H e n r y F r e u n d l i c h
Langm uir
Figure 7.4: Schematic plot of three most common adsorption isotherms; aeq: The loading
of adsorbent at the equilibrium; ceq: The concentration of the solution at the equilibrium.
7.4.2.2 Freundlich adsorption isotherm
This isotherm equation is an empirical relation developed by Freundlich and Küster to
describe a non-linear dependency of the loading on the equilibrium concentration in the
solution (see Fig. 7.4). It is mathematically expressed as
aeq =KF·c1/n
eq (7.9)
where KFis the Freundlich constant and 1/n is an arbitrary constant (n > 1) for a given
adsorbate and adsorbent at a particular temperature. In a double-logarithmic plot Equation
7.9 leads to a straight line in form lnaeq =lnKF+ 1/n ·lnceq. The parameter KFand
ncan be determined by a linear regression. The Freundlich equation, though commonly
used, is thermodynamically inconsistent as it does not give Henry’s law for c→0. This
is evident from the fact that in the plot of ln aeq vs. ln ceq the slope does not approach 1as
ceq is decreased.
7.4.2.3 Langmuir adsorption isotherm
The Langmuir equation relates the surface concentration of the adsorptiv to equilibrium
concentration in the solution at a fixed temperature. It was first derived by Irving Lang-
muir in 1916 on the following assumptions.
1. The molecules are adsorbed into discrete binding sites at the surface of the adsor-
bent.
120 CHAPTER 7. BACKGROUND
2. Each site can bind only one molecule.
3. The binding energy is independent of the presence of the other bound molecules.
The Langmuir equation has following form:
aeq =am
KL·ceq
1 + KLceq
(7.10)
where amis the maximum loading of sorbent with sorptive (monolayer capacity). KL
is the Langmuir adsorption coefficient which increases with the adsorption strength and
with a decrease in temperature.
Equation 7.10 reduces to Henry’s law in the limit of low concentrations. At high con-
centrations it approaches a maximum of the sorption am(see Fig.7.4) and thus accounts
for the finiteness of the sorption space, which is not the case in the Henry and Freundlich
equations.
7.5 Sorption kinetics
The time dependent devolution of the sorption process up to achieving the equilibrium
is referred to as sorption kinetics. It is determined by the character and magnitude of
transport resistances which counteract the mass transfer of sorptive from the solution up
to the sorption centers at the outer or inner surfaces of the adsorbent. Commonly, the
following steps are considered [95].
1. Film diffusion: diffusion of the sorptive through a boundary layer or film adjacent
to the external surface of the sorbent (interphase diffusion).
2. Pore diffusion: diffusion of the sorptive through the porous interior of the sorbent
where the local sorption centers are located (intraparticle diffusion)
3. Sorption: accumulation of sorptive on the sorption centers of the inner sorbent
surface.
The sorption is normally a fast processes. Therefore dependent on the hydrodynamic con-
ditions, either the film diffusion and/or the pore diffusion is the rate determining step. The
driving force for film diffusion and pore diffusion is the resulting concentration gradient
of the adsorptive.
For quantitative studies of the sorption kinetics, a solution of initial concentration c0
and volume Vis brought into contact with the sorbent powder of mass m, and the residual
7.5. SORPTION KINETICS 121
concentration in the supernatant is measured as a function of time t. For each time tthere
is a mass balance analogous to equilibrium (Eq. 7.7).
at=V
m(c0−ct)(7.11)
where atis the loading of sorbent at time tand ctis the sorptive concentration in solution
until the point in time t.
∞
t1
a
at 1
t
a
a = f ( t )
Figure 7.5: Schematic plot of a kinetic curve.
A kinetic curve can be expressed in the form a=f(t)(see Figure 7.5). At the point
in time tan average loading of the sorbent at1is then calculated. The loading a∞is only
reached for sufficiently long sorption times.
In this work the uptake kinetics of Fe-MEPE in SBA-15 is analyzed according to
standard kinetic models for diffusion controlled- systems proposed by J. Crank [96]. For
a system in which the sorption is controlled by the diffusion through the surface layer
(controlling resistance) the uptake curve following a step change in concentration is given
by
at−a∗
a∞−a∗
= 1 −e−αkt (7.12)
where a∞is the final loading at equilibrium with the solution, atis the sorbate loading
time t,a∗is the initial loading before the step change in concentration, αis the mass
transfer rate coefficient and kis the ratio of the external area to volume. After its conver-
sion and linearization Equation 7.12 take the form ln(1 −(at−a∗)/(a∞−a∗)) = −αkt.
122 CHAPTER 7. BACKGROUND
Hence, the plot ln(1−(at−a∗)/(a∞−a∗)) vs. time (t) should give a straight line through
the origin with slope (-αk) allowing straightforward extraction of the mass transfer coef-
ficient αif the sorption is controlled by the diffusion through the surface layer. Equation
7.12 can be applied regardless of the dimensionality of the adsorbent pore system. If
the sorption rate is controlled by internal diffusion (diffusivity D), the following kinetic
equations apply:
for sphere (diffusion in 3 dimensions)
at−a∗
a∞−a∗
= 1 −6
π2
∞
X
n=1
1
n2exp(−n2π2Dt/R2),(7.13)
for slab (one-dimensional diffusion)
at−a∗
a∞−a∗
= 1 −8
π2
∞
X
n=0
exp[−(2n+ 1)2π2Dt/4l2]
(2n+ 1)2(7.14)
where Ris the radius of the sphere and lthe half thickness of the parallel sided slab. Up
to about 50% uptake the Equations 7.13 and 7.14 are well approximated by the limiting
form for the short times
at−a∗
a∞−a∗
≈2krDt
π.(7.15)
If the sorption process is controlled by the internal diffusion process the plot (at−
a∗)/(a∞−a∗)vs. √tshould give a straight line through the origin with slope 2kpD/π.
It is worth noting that a sorption rate measurement can yield only the diffusional time
constant R2/D or l2/D. The estimation of diffusivity requires either knowledge or an as-
sumption about the dimensionality and extension of the pore system. For our cylindrical
pore systems these quantities can be estimated from nitrogen adsorption measurements.
Chapter 8
Materials and Methods
8.1 Materials
8.1.1 Fe-MEPE
The ligand 1,4-bis(2,2´:6´2´´-terpyridine-4´-yl) benzene (1) and Fe-MEPE (2) (see Fig-
ure 1.2a) were synthesized as described elsewhere [97,98,99]. Terephthaldialdehyde,
acetylpyridine, ammonia solution (25%), and ethanol were purchased from Sigma-Aldrich
and used without further purification. Throughout this work Fe-MEPE was used in the
form of the acetate salt. The assembly of ligand (1) (Fig. 1.2a) and Fe(OAc)2(Sigma-
Aldrich, 99.995%) was carried out in 75% acetic acid. In view of the oxidation sensitivity
of Fe(OAc)2after contact with air, the Fe-MEPE material used in most of this work (ma-
terial I) was prepared with a small excess of the iron salt. The 15N NMR measurements
aimed at determining the average chain length of Fe-MEPE before and after uptake in the
pores were performed with a Fe-MEPE sample (material II) assembled under inert condi-
tions from freshly prepared Fe(OAc)2and 15N labeled ligand at an initial metal-to-ligand
stoichiometry of 0.8:1. The NMR measurements revealed that the stoichiometry of the
product was close to 0.9:1. Complementary 15N NMR measurements to characterize the
Fe-MEPE in the solid state were made with two unlabeled Fe-MEPE samples (material
III) and (material IV) of 1:1 and 0.8:1 metal-to-ligand stoichiometry, respectively.
8.1.2 Fe-Terpy
The ligand 2,2´:6´2´´-terpyridine (Terpy) (3), (see Figure 1.2b) was purchased from Sigma-
Aldrich (98%) and used without further purification. Fe-Terpy (4),Fe(Terpy)2(OAc)2(Fig.
123
124 CHAPTER 8. MATERIALS AND METHODS
1.2b), was prepared according to standard procedures published by Margan and Burstall
using freshly prepared Fe(OAc)2and Terpy in 75% acetic acid [100].
8.1.3 Uptake isotherms
8.1.3.1 UV-vis calibration curves
For determining the adsorbed amount of Fe-MEPE and Fe-Terpy from solution in the
silica by UV-vis measurements data were collected first to draw UV-vis calibration curves.
For this purpose different concentrations of Fe-MEPE (in H2O and in KOAc) and Fe-
Terpy (in H2O) were prepared and the absorbance of the MLCT band, which is very
specific for the complexes, was plotted against the concentrations as shown in Figure
8.1. The equations obtained through linear regression of the calibration curves (see Table
0,0 0,2 0,4 0,6 0,8
0,0
0,3
0,6
0,9
1,2
1,5
Fe-MEPE(KOAc)
Fe-MEPE(H2O)
absorbance
c/mM
Fe-Terpy(H2O)
(b)
400 450 500 550 600 650 700 750
0,0
0,2
0,4
0,6
0,8
1,0
1,2
Fe-MEPE in H2O
absorbace
l/nm
0.025mM
0.03mM
0.05mM
0.075mM
0.15mM
0.1mM
0.25mM
0.2mM
0.3mM
0.4mM
(a)
Figure 8.1: (a) Absorbance of the MLCT band of Fe-MEPE solved in H2O (milli-Q) in
different concentrations; (b) calibration curves of Fe-MEPE in H2O and in KOAc and of
Fe-Terpy in H2O.
8.1) were used for the determination of the remaining concentration of the complexes in
solution after adsorption in silica. The main information obtained from the calibration
curves is listed in Table 8.1. Thus the frequency of UV transition bands of Fe-MEPE
depends on the pH of solution. So the frequency of MLCT band of Fe-MEPE in milli-
Q water (pH 5.5) shifts by 15 nm if Fe-MEPE is dissolved in 0.1M KOAc solution
(pH 7.25)
8.1. MATERIALS 125
Table 8.1: Information obtained from the calibration curves
sample λMLCT /nm ε/L mol−1cm2equation
Fe-MEPE in H2O 586 26100 y= 2.83x−0.011
Fe-MEPE in KOAc 601 27200 y= 3.11x−0.002
Fe-Terpy in H2O 552 9490 y= 0.889x+ 0.003
λ: wave length of MLCT transition, ε: molar absorption coefficient
8.1.3.2 Uptake in the silica
The uptake of the MEPE in the silica matrix was studied by bringing a mass msof silica
into contact with a volume Vlof aqueous MEPE solution of an initial concentration c0.
The mass mtof the MEPE complex included in the matrix after a time twas calculated
from the corresponding solution concentration ctby the relation
at=mt
ms
=MVl(c0−ct)
ms
(8.1)
where Mis the molar mass of the entity embedded in the pores (M= 714.6g mol−1for
Fe-MEPE acetate). The concentration ctwas obtained from UV-vis measurements of the
MLCT band of Fe-MEPE in the supernatant solution using the equations listed in Table
8.1. Experiments were made at initial concentrations c0of MEPE in a range from 0.4
to 1.2 mM and with different MEPE-to-silica silica mass ratios µ=mMEP E/mswith
mMEP E =MVlc0by dissolving appropriate amounts of Fe-MEPE in the volume Vl(50
ml) of milli-Q-water (pH=5.5) or 0.1 M potassium acetate solution (KOAc, pH= 7.25).
For studying the time dependence of the uptake, the suspensions were slowly rotated in
a test tube rotator at the chosen temperature and the concentration was measured after
centrifugation (4000 rpm for 15 min) and taking small aliquots of the supernatant for UV-
vis analysis. At the end of the uptake time, when mt/mshad reached a steady value (a∞),
the solid product, having an intense violet color, was washed two times with 50 ml H2O
and dried at 60◦C before further analysis.
8.1.3.3 Settling of Fe-MEPE during centrifugation
Blank measurements in the absence of the silica (but otherwise under identical conditions
as in the determination of the uptake isotherms) were made in order to see if centrifugation
is causing gravitational settling of Fe-MEPE, which would cause systematic errors in the
126 CHAPTER 8. MATERIALS AND METHODS
determination of the uptake. Results for Fe-MEPE in milli-Q water at 50◦C are presented
in Figure 8.2, where the depletion of the solution (in percent of its original concentration)
due to gravitational settling is plotted as a function of time telapsed since the beginning of
the experiment (start of rotation of the sample at the chosen temperature). Open symbols
070 140 210 280 350
-2
0
2
4
6
8 centrifuged
shaked
wt.%settledMEPE
time/h
Figure 8.2: Blank measurement showing the effect of gravitational settling induced by
centrifugation of a 0.4 mM Fe-MEPE solution in water at 50◦C. The data give the w %
of settled MEPE after 15 min centrifugation (open circles) and after centrifugation and
subsequent shaking of the samples (full symbols) as a function of the time elapsed since
sample preparation and rotation at 50◦C.
give the depletion measured immediately after centrifugation, full symbols the respective
depletion after centrifugation and subsequent shaking of the sample. Figure 8.2 shows that
the depletion effect caused by settling of MEPE is increasing with the time over which the
samples had been rotated at the elevated temperature. This effect is attributed tentatively
to a gradual increase of the mean molar mass of Fe-MEPE in the aqueous medium. In
quantitative terms, this effect is rather small (less than 3%) for times up to 150 h, i.e., for
all data in the fast-uptake regime. However, progressively greater settling effects were
observed at longer times (nearly 20% after 800 h). The uptake data for the slow-uptake
regime presented in Chapter 9(samples CA-MEPE-7 and CA-MEPE-8) extend to uptake
times of or up to 300 h. They were corrected for the effect of gravitational settling.
8.2. METHODS 127
8.2 Methods
8.2.1 XPS
XPS measurements were carried out in Fritz-Haber Institut using a modified LHS/SPECS
EA200 MCD system equipped with a Mg Kαsource (1253.6 eV, 168 W). The binding
energy scale of the system was calibrated using Au4f7/2= 84.0eV and Cu2p3/2= 932.67
eV from foil samples. The powder samples were placed in a stainless steel sample holder
with a 0.6 mm deep rectangular well, covering an area of (12 ×8) mm2. During the XPS
experiments the pressure in the UHV analysis chamber increased to about 1×10−9mbar.
The binding energies of the SBA-15 based samples were referred to the Si2p signal of
silica at 103.6 eV to correct for charging. Data reduction included satellite deconvolution,
and subtraction of a Shirley background. Quantitative data analysis was performed on the
basis of peak areas by fitting with 30/70 Gauss-Lorentz product functions. Atomic ratios
were calculated using empirical cross sections [101].
8.2.2 AAS
Atomic absorption spectroscopy (AAS) measurements of Fe in the silica matrix were
made in the analytic chemistry department of TU-Berlin by a 1100B Perkin-Elmer Flame
AAS spectrometer after dissolving the silica matrix in 25 ml of 0.1 M NaOH. To dissolve
the Fe-MEPE which precipitates in this high pH region, 2 ml of each sample solution
was acidified with concentrated HCl and filled up to 10 ml with milli-Q water. After this
pretreatment flame atomization was performed in air-acetylen gas mixture. Calibration
was done by 0.1 M FeCl3standard (titrisol) by means of the standard addition method.
8.2.3 UV-Vis
UV-vis absorption spectra were recorded on a Carry 50 UV-vis spectrometer by Varian
using quartz glass cells of 1 mm thickness.
8.2.4 15N solid-state NMR
The solid-state 15N NMR measurements were performed in the chemistry department
of FU-Berlin on a Bruker MSL-300 instrument operating at 7 Tesla, equipped with a
Chemagnetics-Varian 6 mm pencil CPMAS probe. The samples were spun at 6-8 kHz
under magic angle spinning (MAS) conditions. The 15N MAS spectrum was recorded
128 CHAPTER 8. MATERIALS AND METHODS
employing π/2 pulse-sequence, with a 90◦-pulse length of 4.5 sec. The {1H}-15N CPMAS
spectra were recorded using a cross polarization contact time of 5 ms, the typical 1H-90◦-
pulse lengths were 3.5-4.5 sec. All 15N chemical shift values are referenced to solid
15NH4Cl.
Chapter 9
CA doped SBA-15 as a host for MEPE∗
Abstract
The adsorption of a metallo-supramolecular coordination polymer (Fe-MEPE) in the cylin-
drical pores of SBA-15 silica with pure and carboxylic acid (CA) carrying pore walls has
been studied. Fe-MEPE is an intrinsically stiff polycation formed by complexation of
Fe(II)-acetate with an uncharged ditopic bis-terpyridine ligand. The adsorption affinity
and kinetics of the Fe-MEPE chains is strongly enhanced when the pore walls are doped
with CA, and when the pH of the aqueous medium or temperature is increased. The ini-
tial fast uptake is connected with a decrease of pH of the aqueous solution, indicating an
ion-exchange mechanism. It is followed by a slower (presumably diffusion-controlled)
further uptake. The maximum adsorbed amount of Fe-MEPE in the CA doped mate-
rial corresponds to a monolayer of Fe-MEPE chains disposed side-by-side along the pore
walls. The stoichiometry of Fe-MEPE in the pores (determined by XPS) was found to be
independent of the loading and similar to that of the starting material. The mean chain
length of Fe-MEPE before and after embedding in the CA doped matrix was studied by
solid-state 15N NMR using partially 15N-labeled Fe-MEPE. It is shown that the average
chain-length of Fe-MEPE is reduced when the complex is incorporated in the pores.
∗In modified form published as: Akcakayiran, D.; Mauder, D.; Hess, C.; Sievers, T. K.; Kurth, D. G.;
Shenderovich, I. G.; Limbach, H.-H.; Findenegg, G. H. J. Phys. Chem. B.,2008,112, 14637
129
130 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
9.1 Introduction
A new era in inclusion chemistry began with the discovery of periodic mesoporous materi-
als such as SBA-15 [4] which constitutes a 2D-hexagonal arrangement of cylindrical pores
with diameters in a range 6-12 nm. Because of their wide pore openings, narrow pore size
distribution and their large internal surface area, these materials have a high potential as
catalyst supports [102], [103], adsorbents and host materials for organic guest molecules.
Incorporation of functional molecular components or metal ions into such a mesoporous
oxide matrix can be carried out by different ways, namely by a one-pot synthesis [104],
ion-exchange in as-synthesized material [105,106,107], covalent grafting followed by
ion exchange [108], solid state grinding [109] or by direct loading sorption [110,111].
Among the incorporated functional molecular components dyes and related molecules
represents an important class. For instance, Fe-MEPE [110], methylen blue, rhodamin,
thionine dyes [112], coumarine derivate dyes [113], fluorescein [114], direct blue 71 dye
[115] and chromophores [116] have been incorporated into mesoporous silica matrices.
In many cases, the interaction between the dye molecules and the host system causes a
shift of UV-vis absorption bands. Due to this interaction with the host, a higher degree
of organization and increased diffusion stability of the guest can be achieved, which is
of importance for potential applications in lasers, optical sensors, photochromic materials
and photocatalysts [117,116,118].
This Chapter diels with the uptake of a metallo supramolecular coordination poly-
electrolytes (MEPE) in SBA-15 type silica matrices. MEPE materials are prepared by a
metal-ion induced self-assembly of two-valent transition metal ions with ditopic terpyri-
dine ligands. The resulting materials exhibit several potentially useful properties such as
electro-chromic behavior [119,120,121], molecular magnetism [122], as well as dynamic
properties due to the labile metal ion ligand interaction [14]. The properties of MEPE are
readily manipulated by simply changing the ligands or metal ions. The substance studied
in this work, Iron-(II)-MEPE (2) (see Fig. 7.1 in Section 7.1) is assembled from Fe-
(II)-acetate and 1,4-bis (2,2´:6´2´´-terpyridine-4´-yl) benzene (1), which results in stereo-
chemically well-defined octahedral coordination geometry with D2dsymmetry [84]. At an
exact 1:1 stoichiometry of Fe-(II) to ligand, the average length of the Fe-MEPE chains in
aqueous solution can grow to very high values, depending on concentration, temperature,
and pH. Small deviations from 1:1 stoichiometry are causing a reduction of chain-length
[14]. Hence, the metal-to-ligand stoichiometry is another important parameter affecting
the properties of the resulting assemblies and the amount of MEPE adsorbed in the pores.
9.1. INTRODUCTION 131
A prominent feature of Fe-MEPE is the positive charge along the backbone. This
implies that screened electrostatics plays an important role in the interaction of MEPE
with the silica pore wall, which at ambient pH is negatively charged due to the weakly
acidic surface silanol groups. It is of interest to see if this charge effect can be boosted
by decorating the pore walls with acidic groups. For this purpose, a chemically modified
SBA-15 material with short-chain carboxylic acid groups grafted to the pore walls (CA-
SBA-15) was prepared (see Fig. 1.2c). The carboxylic acid (CA) functionality was chosen
because the ionized form of these groups is similar to the acetate ions which form the
counter ions of MEPE in solution. The transfer of the MEPE polycations from solution
into the pores may represent an ion exchange process, involving either the ionization of
carboxylic acid groups at the pore wall and a release of protons or the release of the
counterions of MEPE polycations. In either case, the ion exchange represents an entropic
driving force for the adsorption of MEPE in the pores. To test this possible mechanism
we made a comparative study of the uptake of Fe-MEPE in pure (‘native’) and CA-doped
SBA-15 materials. We also studied the effect of temperature and pH on the uptake of
MEPE, as both can affect the sorption process.
Another interesting question concerns the mechanism of the MEPE uptake and the
mean chain-length of MEPE in the pores. It is known that in aqueous media, due to the
labile nature of the metal ion ligand interaction, MEPE represent dynamic equilibrium
polymers, i.e., assemblies of different length coexist in solution [123]. It is feasible that
short chains adsorb more easily into the pores, or that longer chains break up into shorter
chain segments when entering the pore. These processes would lead to a shorter chain
length of MEPE in the pores as compared to the solution. Alternatively, short chain
segments may merge into longer ones in the pores if the pore walls stabilize the chains.
Two ways for studying the chain length of MEPE in the pores are used in this work: (I)
Indirect information is derived from the number ratio of metal ions to carbon or nitrogen
atoms in the pore, which can be determined by X-ray photoelectron spectroscopy (XPS).
(II) A more direct way is to measure the ratio of terminal and internal terpyridine groups
of the Fe-MEPE using solid state NMR spectroscopy. This latter method requires 15N
NMR measurements, which made it necessary to synthesize a 15N-labeled ligand.
A combination of experimental techniques was used to characterize pure and CA
doped SBA-15 materials and the state of Fe-MEPE in the pores. The silica samples be-
fore MEPE incorporation were characterized by nitrogen adsorption, small angle X-ray
diffraction (SAXD), solid state 29Si and 13C NMR, whereas Fe-Terpy, Fe-MEPE and Fe-
MEPE incorporated in the silica materials were studied by UV-vis-spectroscopy, atomic
132 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
absorption spectroscopy (AAS), X-ray photoelectron spectroscopy (XPS) and 15N solid-
state NMR. The materials and methods used in this study have been already introduced
in Chapter 2and in Chapter 8, and the results are presented in Section 9.2. The Discus-
sion (Section 9.3) is focused on the driving forces for the uptake of MEPE in the pores,
the maximum uptake, and the stoichiometry and average chain length of Fe-MEPE in the
pores of pure and CA doped SBA-15. Finally, Section 9.4 winds up the main findings of
the work.
9.2 Results
9.2.1 Characterization of silica hosts
The synthesis of SBA-15, 10-CA and 20-CA silica materials used as hosts for Fe-MEPE
were described in Section 4.1. The physical properties of silica materials used are sum-
marized in Table 9.1.
Table 9.1: Characterization of pure and-COOH-functionalized SBA-15 by nitrogen ad-
sorption SAXD and 29Si NMR
sample as/m2g−1vp/cm3g−1D/nm a0/nm xCOOH
SBA-15 650 0.89 8.9 11.50 −
10-CA 513 0.97 9.8 12.13 0.19
20-CA 400 0.60 8.0 11.61 0.42
as: BET specific surface area, vp: specific pore volume D: pore diameter (imp. KJS algorithm),
a0: lattice constant, xCOOH : mole fraction of carboxylic acid groups on the surface
9.2.2 Characterization of coordination compounds by NMR
The synthesis of Fe-MEPE and Fe-Terpy, the record of UV-adsorption isotherms and the
methods used were explored in Chapter 8. The free terpyridine (Terpy) ligand and the
Fe-Terpy complex were investigated by 15NCPMAS NMR spectroscopy to determine
the chemical shift of the different nitrogen atoms, which is needed for peak assignment of
the Fe-MEPE complex. For free Terpy in its energetically preferred trans-configuration
two signals at chemical shifts of 252 ppm and 267 ppm appear, resulting from nitrogen
atoms on the central ring () and the outer rings (dashed ), respectively (Figure 9.1a).
9.2. RESULTS 133
Figure 9.1: Solid-state 15N CPMAS NMR spectra of (a) free ligand Terpy and (b)
Fe(Terpy)2OAc2. The signals are assigned to atoms by symbols as follows: free Terpy;
♦complexed Terpy. Solid symbols indicate N atoms of the central ring, dashed symbols
N atoms of the outer rings.
Figure 9.1b shows the 15NCPMAS NMR spectrum of Fe-Terpy, which is taken as a
simple model of a Fe-MEPE unit. For complexation with Fe(II) ions, Terpy molecules
have to change their configuration from trans to cis to allow octahedral coordination of the
ligands to the center. The presence of the Fe(II) ion and the different ligand configuration
have an effect on the chemical shift for the different nitrogen species in the Fe-Terpy
complex. The signal at 242 ppm is assigned to the central nitrogen atom of Fe-Terpy,
slightly shifted to high field, and the signal at 207 ppm is assigned to the outer nitrogen
atoms of Fe-Terpy, shifted 60 ppm to high field, compared to the free ligand.
15NCPMAS NMR spectra of solid Fe-MEPE materials III and IV, having different
stoichiometric Fe:ligand ratios (see Section 8.1), are displayed in Figure 9.2. By com-
parison with the 15NCPMAS NMR spectra of free Terpy and Fe-Terpy (Fig. 9.1) we
can assign each peak to the either of the two types of nitrogen atoms. In the Fe-MEPE
material with a Fe:ligand ratio of 1:1 only signals from Fe(II)-coordinated ligands (238
and 205 ppm) are detected (Figure 9.2a). This implies that nearly all ligands are coor-
134 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
Figure 9.2: Solid state 15NCPMAS NMR spectra of (a) Fe-MEPE material III with an
Fe:ligand ratio 1:1; (b) Fe-MEPE material IV with an Fe:ligand ratio of 0.8:1. Fe-MEPE
chains of 1:1 stoichiometry (spectrum a) are long and thus end groups are detected. In
the material having a ligand excess (spectrum b), uncoordinated Natoms are present at
an appreciable concentration.
dinated to two Fe(II) ions, so that the number of non-coordinated nitrogen atoms at the
chain ends is too small to be detected by solid state 15NNMR. From this, we conclude
that Fe-MEPE with an Fe:ligand ratio of 1:1 consists of long chains. However, as such
long chains are hardly soluble in aqueous media, they are unsuitable for uptake studies
into the porous matrix. Accordingly, a Fe-MEPE sample with an Fe:ligand stoichiom-
etry of approximately 0.8:1 (material IV) is used because an excess of ligand results in
shorter Fe-MEPE chains [14]. The 15NNMR spectrum of this Fe-MEPE sample (Fig.
9.2b) shows two additional signals corresponding to N-atoms in free ligands at chain ends
(269 and 259 ppm) besides the signals from coordinated ligands (242 and 206 ppm). This
result illustrates that 15Nsolid-state NMR can be used to estimate the chain-length of
9.2. RESULTS 135
Fe-MEPE by analyzing the signal intensities of coordinated and free Natoms of the lig-
and. An estimate of the mean chain length of Fe-MEPE before and after inclusion into
the pores by signal deconvolution of the respective signals is presented in Section 9.2.5
9.2.3 Effect of CA functionalization on the Fe-MEPE uptake
The uptake of Fe-MEPE (material I) in the silica matrix as a function of time was deter-
mined by UV-vis spectroscopy as explained in Section 8.1. The strong MLCT band is
ideally suited to quantify the concentration of Fe-MEPE. The uptake of Fe-MEPE into
the cylindrical pores of pure and CA doped SBA-15 was studied at different mass ratios
of Fe-MEPE and silica matrix µin a range from 0.05 to 0.3.
0100 200 300 400 500 600
0
20
40
60
80
100
mgMEPE/gsilica
time/h
m=0.1
m=0.05
m=0.05
(b)
400 500 600 700
0,0
0,4
0,8
1,2
absorbance
l/nm
before mixing
2h
22h
120h
168h
264h
312h
456h
576h
625h
t
(a)
Figure 9.3: Example of the uptake of Fe-MEPE from 0.1 M KOAc into pure SBA-15 and
the CA doped material 20-CA: (a) Decrease of the MLCT band of the supernatant solution
during an uptake experiment with pure SBA-15; (b) corresponding uptake isotherms into
pure SBA-15 () and 20-CA (•) at a MEPE/silica mass ratio µ= 0.05, and into 20-CA at
µ= 0.1 (◦). Curves are drawn to guide the eye.
Figure 9.3 shows a typical example for the decreasing intensity of the MLCT band in
the supernatant solution as a function of time (a) , and the respective uptake isotherms (b)
for Fe-MEPE in pure SBA-15 and 20-CA. For both samples, we observe a fast initial up-
take of Fe-MEPE. This fast process is most pronounced for 20-CA: At a MEPE-to-silica
mass ratio µ=0.05, it leads to nearly quantitative uptake of MEPE within 2 h, indicated by
a change of the initially deep blue color of the Fe-MEPE solution to completely colorless,
while the initially colorless silica turned violet. For pure SBA-15 at the same MEPE-to-
silica mass ratio the fast process leads to a lower uptake, but is followed by a much slower
136 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
second uptake process that leads to nearly quantitative uptake of MEPE after about 600 h.
This slow process also takes place with Fe-MEPE in 20-CA at a higher MEPE-to-silica
mass ratio (µ=0.10), where it leads to an increase in the MEPE uptake from 50% to more
than 80% of the overall amount after about 200 h (Fig. 9.3b).
020 40 60 80 100 120
0
10
20
30
40
50
mgMEPE/gsilica
time/h
10-CA
SBA-15
(b)
020 40 60 80 100 120
4.2
4.5
4.8
5.1
5.4
5.7
6.0
pH
time/h
SBA-15
10-CA
(a)
Figure 9.4: Uptake of Fe-MEPE in pure SBA-15 and 10-CA (20◦C): (a) Evolution of
pH in the aqueous solution before (t=0) and after addition of the silica (t>0); (b) uptake
isotherms in the two materials (MEPE-to-silica mass ratio µ=0.10, room temperature).
Dashed curves are drawn as a guide to the eye.
It is well-known that SBA-15 materials with carboxylate groups have cation-exchange
properties [124]. In order to check whether ion exchange is a driving force in the uptake of
Fe-MEPE in the present materials, time dependent measurements of pH were performed
with freshly prepared suspensions of the silica materials in aqueous Fe-MEPE solutions in
the absence of KOAc. Results for 10-CA and pure SBA-15, both at µ=0.10, are displayed
in Figure 9.4a, and the respective uptake isotherms are shown in Figure 9.4b. The initial
pH (5.6) represents the value of the MEPE solution in milli-Q water. Addition of the silica
causes a decrease in pH, from 5.6 to 5.2 for pure SBA-15 and from 5.6 to 4.5 for 10-CA,
during the first 1.5-2 h, that is, in the period of the fast uptake process. Later, during the
slow uptake process, the pH of the suspensions remains constant. The finding that the
pH change is linked with the fast uptake process of MEPE indicates that ion exchange
is indeed involved in this step. Moreover, the larger pH change induced by 10-CA, as
compared to the pure SBA-15, is in line with the larger initial uptake of MEPE in 10-CA
(Fig. 9.4b).
To assess the influence of pH, and thus the degree of dissociation of the CA groups at
the pore wall, a comparison of the uptake of Fe-MEPE from pure water (initial pH 5.5)
and from 0.1 M KOAc solution (pH 7.3) was made. Uptake isotherms for MEPE at an
9.2. RESULTS 137
020 40 60 80 100
0
10
20
30
40
50
60
20-CA, KOAc
10-CA, KOAc
10-CA, H2O
mgMEPE/gCA-SBA-15
time/h
(a)
050 100 150 200 250 300 350
0
10
20
30
40
mgMEPE/gSBA-15
time/h
m=0.05, KOAc
m=0.09, KOAc
m=0.09, H2O
(b)
Figure 9.5: (a) Effect of pH on the uptake isotherms of Fe-MEPE in carboxylic-acid
doped SBA-15: Fe-MEPE uptake in •10-CA at pH 5.5; N10-CA at pH 7.3; 20-CA at
pH 7.3; (b): Effect of pH and µon the uptake isotherms of Fe-MEPE in pure SBA-15:
at pH 7.3 and µ=0.05, at pH 7.3 and µ=0.09, and Fat pH 5.5 and µ=0.09; dashed lines
are drawn as a guide to the eye.
initial concentration c0=0.4 mM, and a MEPE-to-silica mass ratio µ=0.05 are shown
in Figure 9.5a. It can be seen that in 10-CA both the fast and the slow uptake process
are more effective at the higher pH, at which complete uptake is accomplished within
about 50 h. At the lower pH, at which one expects a lower degree of ionization of the
CA groups at the pore wall, the fast process leads to a lower initial uptake and the slow
process proceeds at a significantly lower rate than at higher pH. Figure 9.5a also indicates
that at a given pH (7.3) the uptake of MEPE in 10-CA is less effective than in 20-CA,
for which complete uptake is accomplished within 2 h at the given MEPE-to-silica mass
ratio. Figure 9.5b shows the uptake isotherms for MEPE in pure SBA-15 at pH 5.5 and
pH 7.3. As in the case of CA-SBA-15, at a given MEPE-to-silica mass ratio (µ=0.09)
an increase of pH leads to an increase of the adsorbed amount of Fe-MEPE. However, in
the case of pure SBA-15 the effect is much smaller than observed in case of CA-SBA-
15. Figure 9.5b also shows that the adsorbed amount of Fe-MEPE (nearly 40 mg/g) does
not increase when increasing µ, but the uptake at µ=0.05 is rather higher than the uptake
at µ=0.09. This is distinctly different from the MEPE uptake in CA-SBA-15, where an
increase of µleads to an increase in the adsorbed amount of MEPE. This shows clearly
that the adsorption affinity of MEPE in pure SBA-15 is much less than in CA-SBA-15.
We also studied the effect of temperature on the rate of uptake of Fe-MEPE into the
pores of pure and CA doped SBA-15. Results for Fe-MEPE in 20-CA at 20◦C and 45◦C
are shown in Figure 9.6a. In this example (µ=0.09) at 45◦C, nearly quantitative uptake
138 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
0,0 0,1 0,2 0,3
0
50
100
150
200
250
mgMEPE/gsilica
ceq/mM
SBA-15
20-CA
(b)
050 100 150 200 250 300 350
0
25
50
75
100
20°C
45°C
mgMEPE/g20CA
time/h
(a)
Figure 9.6: (a) Effect of temperature on the uptake rate in 20-CA (µ=0.1); the vertical bar
indicate complete uptake; (b) Adsorption isotherms of Fe-MEPE from aqueous solution
(without added salt) in 20-CA (•) and in pure SBA-15 (); results for different MEPE-to-
silica ratios µat 50◦C. The solid line represents a fit of the 20-CA data by the Langmuir
equation; the dashed-dotted line for the data in SBA-15 is drawn to guide the eye.
was reached after 150 h while at 20◦C uptake was not complete after 350 h. This behavior
indicates a significant kinetic resistance for the slow process. For this reason, measure-
ments of the equilibrium adsorption isotherm were performed at an elevated temperature.
Figure 9.6b shows data for the uptake of Fe-MEPE in 20-CA up to 300 h as a function
of the concentration in the supernatant solution. The individual data points correspond
to different initial MEPE-to-silica ratios µ. Also shown in this figure are two points for
the adsorption of Fe-MEPE in pure SBA-15, to indicate the large difference in adsorp-
tion affinity between pure and carboxylic-acid doped SBA-15. A detailed account of the
adsorption kinetics of Fe-MEPE in 20-CA will be presented in Chapter 10.
9.2.4 Characterization of Fe-MEPE in CA-SBA-15
To characterize the state of Fe-MEPE in the pores of pure and CA doped SBA-15 materi-
als, some of the composite samples were studied by a combination of different techniques.
The effect of the MEPE uptake on the properties of the matrix was determined by nitrogen
adsorption isotherms and SAXD measurements, while the Fe-MEPE content was deter-
mined by atomic absorption spectroscopy (AAS) and X-ray photoelectron spectroscopy
(XPS). Composite samples were prepared by exposing the silica to Fe-MEPE solutions
(50 mL) of different initial concentrations c0and different MEPE-to-silica mass ratios µ,
and centrifugation after 300 h. (The terms bare silica and composite sample are used for
the silica samples without and with embedded Fe-MEPE, respectively.) Table 9.2 gives
9.2. RESULTS 139
the Fe-MEPE content of these samples, expressed as adsorbed mass of Fe-MEPE per unit
mass of silica, and as the mean number of MEPE chains accommodated side-by side in a
pore. Both quantities were derived from the UV-vis depletion measurements. The sam-
ples specified as CA-MEPE-4 to CA-MEPE-6 come from the high-affinity regime of the
MEPE uptake by the silica. The samples CA-MEPE-7, CA-MEPE-8 and CA-15N-MEPE
result from the highest initial MEPE concentrations c0and highest MEPE-to-silica mass
ratios µ, at which the uptake was not complete after 300 h. Values of the Fe-MEPE con-
tent of these samples were corrected for the effect of sedimentation of Fe-MEPE during
the centrifugation step (see Section 8.1).
Table 9.2: Fe-MEPE (material I) in SBA-15 (SBA-MEPE) and in 20-CA (CA-MEPE-4-8)
prepared for comparative studies.
sample c0/mM µmFe−MEPE/SiO2S
SBA-MEPE 0.37 0.094 0.033 3
CA-MEPE-4 0.37 0.094 0.091 10
CA-MEPE-5 0.48 0.112 0.119 13
CA-MEPE-6 0.62 0.159 0.156 17
CA-MEPE-7 0.81 0.207 0.195 21
CA-MEPE-8 1.20 0.305 0.223 24
CA-15N-MEPE†1.127 0.225 0.219 18
c0: initial MEPE concentration; µ: MEPE to silica mass ratio; mF e−MEP E /SiO2: specific uptake of Fe-
MEPE acetate; S: Number of MEPE chains accommodated side-by-side in the pores. †prepared from a
different 20-CA charge using Fe-MEPE material II.
The effect of embedding Fe-MEPE on the pore structure and pore volume pure and
CA-doped SBA-15 samples was studied by small-angle X-ray diffraction (SAXD) and
nitrogen adsorption. In this way it can be checked if the silica matrix was affected by
the exposure to aqueous solutions and elevated temperatures during the embedding of
Fe-MEPE. The SAXD measurements were also made in order to see if a layer of Fe-
MEPE adsorbed at the pore walls causes characteristic changes in the scattering curves.
Nitrogen adsorption isotherms for a CA-MEPE composite and the respective bare CA-
doped silica sample are presented in Figure 9.7a. Also shown is the isotherm of the bare
CA-doped sample after exposure to water under similar conditions as during the MEPE
uptake procedure.
The pore diameters, specific surface area and specific pore volume derived from these
isotherms are given in Table 9.3. It can be seen that the pore diameter is not significantly
140 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
0,5 1,0 1,5 2,0
1,5 1,6 1,7 1,8 1,9
q/nm-1
21
Intensity/a.u.
q/nm-1
20-CA
CA-MEPE-4
CA-MEPE-8
10
11 20
21
(b)
0,0 0,2 0,4 0,6 0,8 1,0
100
200
300
400
500
20-CA
20-CA-w
CA-MEPE-6
adsorbedvol./cm3.g-1
relative pressure
(a)
Figure 9.7: (a) Nitrogen adsorption isotherms for sample 20-CA in its original state (◦),
after exposure to water for 30 d (20-CA-w) (•), and after uptake of Fe-MEPE (sample
CA-MEPE-6) (N); for all samples the amount of adsorbed nitrogen is referenced to unit
mass of the 20-CA matrix; (b) SAXD profiles for 20-CA and the composite samples CA-
MEPE-4 (µ=0.09) and CA-MEPE-8 (µ=0.3). The profile for the sample with µ=0.3
is moved up by a factor 1.4 in order to have the same intensity at the position of the 10
reflection for all samples.
changed by the exposure to water nor by the incorporation of Fe-MEPE. On the other
hand, the pore volume of the bare 20-CA material is increased by the exposure to water.
Hence, the similar pore volumes of the samples CA-MEPE and bare 20-CA before water
exposure can be attributed to a compensation of two effects. (i) an increase of vpdue to
the exposure to water and (ii) a reduction of vpdue to the incorporation of Fe-MEPE. The
former effect may be caused by the ability of water to remove some constrictions in the
pore system and thus to improve the accessibility of the pores of 20-CA. The decrease of
the pore volume by ca. 0.14 cm3g−1in step (ii) for the uptake of 0.16 g MEPE/g of silica
is of the expected order of magnitude assuming a density of about 1 g/cm3for Fe-MEPE.
Systematic measurements on the effect of water exposure on the pore volume and specific
area of CA-doped SBA-15 materials corroborating this effect were presented in Section
6.2.4. As was shown there, the pore volume of CA-SBA-15 silica increased by exposure
to water without a significant change of the pore diameter.
The SAXD profiles of two composite samples, CA-MEPE-4 (µ=0.09) and CA-MEPE-
8 (µ=0.3), and of the respective bare 20-CA (µ=0) are shown in Figure 9.7b . One finds
that the leading Bragg reflections (10, 11, 20) are shifted to larger q with increasing load-
ing, indicating some shrinkage of the lattice spacing with the incorporation of Fe-MEPE.
Presumably this effect is caused by immersing the silica sample into water during the
9.2. RESULTS 141
Table 9.3: Specific surface area as, specific pore volume vp, and mean pore diameter D as
derived from the nitrogen adsorption isotherms of Figure 9.7a.
sample as/m2g−1vp/cm3g−1D/nm
20-CA 400 0.60 8.1
20-CA-w 489 0.75 8.2
CA-MEPE-6 331 0.61 8.0
sample preparation, which took 6 days for the sample with µ=0.09 and 35 days for the
sample with µ=0.3. The relative intensities of the three leading Bragg peaks (10, 11 and
20) do not change significantly as the amount of MEPE in the matrix is increased. This
indicates that the MEPE chains are not forming a layer of uniform thickness and electron
density at the pore wall. On the other hand, one observes a significant lowering of the dif-
fuse scattering at large q (above the 20 Bragg peak) as the content of MEPE is increased.
Due to the decreased diffuse scattering the 21 Bragg reflection becomes detectable for the
sample with µ=0.3, although this peak is not detectable at lower loadings of the pores.
This point is discussed in Section 9.3.3 of this Chapter.
The samples CA-MEPE-4, CA-MEPE-6 and CA-MEPE-8 were characterized by tak-
ing XPS spectra of carbon, nitrogen and iron. Figure 9.8A shows C1s XPS spectra of bare
20-CA (a) and of the samples CA-MEPE-4 (b) and CA-MEPE-6 (c). The peaks at 285.1
and 289.6 eV of bare 20-CA are assigned to the carbon atoms of the alkyl and carboxyl
groups, respectively, of the propionic acid groups at the pore wall. Upon loading with
Fe-MEPE, a shift of the 285.1 eV peak to 285.4 eV is observed. To compare the spectral
features of embedded Fe-MEPE with bulk Fe-MEPE, difference spectra were calculated
by subtraction of spectrum (a) from the spectra (b) and (c). The resulting difference spec-
tra are displayed in Fig. 9.8B as spectra (a) and (b). Regarding the peak position of the
signal at 285.4 eV, the C1s spectra of embedded Fe-MEPE closely resemble the respective
spectrum of bulk Fe-MEPE which is shown as spectrum (c) in Fig. 9.8B. The observed
binding energy of 285.4 eV is in good agreement with the literature value of 285.5 eV for
pyridine [19]. Figure 9.8B also shows that the intensity of the C1s signal of the embedded
Fe-MEPE increases from CA-MEPE-4 to CA-MEPE-6.
Figure 9.8C shows X-ray photoelectron Fe2p3/2spectra of Fe-MEPE in the samples
CA-MEPE-4 (a) and CA-MEPE-6 (b); the respective Fe2p3/2spectrum of bulk Fe-MEPE
normalized in intensity to spectrum (b) is shown as spectrum (c). All spectra are char-
acterized by an asymmetric peak with an intensity maximum at 709.3 eV. This value of
142 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
282284286288290292
Binding Energy (eV)
0
0.05
0.1
Intensity(Si2pnorm.)
a
b
c
C1s
289.6 285.4 285.1
282284286288290292
Binding Energy (eV)
0
0.02
0.04
0.06
0.08
0.1
Intensity
a
b
c
C1s
285.4
706708710712714716718
Binding Energy (eV)
0
0.005
0.01
0.015
0.02
0.025
Intensity
a
b
c
Fe2p3/2
709.3
AB
C
Figure 9.8: (A) Si2p normalized X-ray photoelectron C1s spectra of bare 20-CA (a) and
20-CA loaded with Fe-MEPE at µ=0.10 (b) and µ=0.16 (c); (B) X-ray photoelectron
C1s difference spectra of 20-CA loaded with Fe-MEPE at µ=0.10 (a) and µ=0.16 (b)
as well as bulk Fe-MEPE (c) normalized in intensity to spectrum (b); (C) X-ray photo-
electron Fe2p3/2spectra of 20-CA loaded with Fe-MEPE (a) µ=0.10 and (b) µ=0.16
as well as (c) bulk Fe-MEPE normalized in intensity to spectrum b. For details see text.
Spectra in each graphic are shifted for clarity.
the binding energy as well as the shape of the asymmetric tail are consistent with re-
sults reported for other Fe(II) compounds such as Fe(phthalocyanine) [125] or FeO [126].
However, due to the complexity of the X-ray photoelectron Fe2p spectra, a contribu-
tion of Fe(III) ions cannot be ruled out. X-ray photoelectron N1s spectra of the samples
CA-MEPE-4 and CA-MEPE-6 and bulk Fe-MEPE were also determined. From the nor-
malized intensities of the N1s and Fe2p3/2signals the atom ratio N/Fe of the MEPE
complexes in the pores was determined (Table 9.4).
Table 9.4 compares the iron content of the 20-CA-MEPE composite samples as de-
termined by three methods, viz. depletion of the supernatant solution (UV-vis), AAS and
XPS. The AAS and UV-vis data are in reasonably good agreement (deviation of AAS
from UV-vis no greater than 12%), except for the sample with the highest loading (CA-
9.2. RESULTS 143
Table 9.4: Characterization of CA-MEPE samples specified in Table 9.2: Fe content and
N/Fe atomic ratio.
wt % Fe∗XPS
sample UV-vis AAS XPS at.%Fe at.%N N/Fe
CA-MEPE-4 0.65 −0.60 0.30 1.90 5.3
CA-MEPE-5 0.83 0.93 −
CA-MEPE-6 1.06 0.96 1.36 0.43 2.30 5.4
CA-MEPE-7 1.27 1.14 −
CA-MEPE-8 1.42 1.12 1.96 0.61 3.30 5.4
wt % Fe∗of the Fe-MEPE loaded samples as determined by UV-vis, AAS, and XPS
MEPE-8) for which the AAS values are significantly lower than that obtained by the
depletion method. For the sample of lowest Fe-MEPE content (CA-MEPE-4), the iron
content obtained by XPS also agrees with that obtained by the depletion method. At
higher loadings, however, XPS yields significantly higher values of the iron content than
the two other methods. Higher Fe concentrations obtained by XPS than by AAS may
indicate higher concentrations of Fe-MEPE near the surface than in the core of the silica
grains, because XPS is a surface-sensitive technique while the other methods are integrat-
ing over the whole sample. Specifically, the very high value for CA-MEPE-8 obtained by
XPS may indicate that part of the Fe-MEPE was not included in the pores but formed a
multilayer film at the outside of the silica grains. Table 9.4 also shows that the number
ratio of nitrogen and Fe atoms has the same value (5.4 ±0.3) for all composite samples
studied by XPS. This indicates that the composition of Fe-MEPE was the same in these
three samples, which is remarkable in view of their very different loadings. For octahe-
dral coordination of the metal centers and infinitely long Fe-MEPE chains one expects
an atom number ratio N/Fe = 6. The somewhat lower experimental value indicates an
excess of iron. This result is consistent with the fact that the Fe-MEPE used in these
experiments (material I) was prepared with a small metal excess. It thus appears that the
stoichiometry of the Fe-MEPE in the pores is similar to that of the original material, in
line with the fact that most of the Fe-MEPE is adsorbed in the pores in the high-affinity
region of the uptake isotherm.
144 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
9.2.5 Characterization by solid-state 15N NMR
Figure 9.9: (a) 15N MAS NMR and (b) 15N CPMAS NMR of bulk solid 15N Fe-MEPE
and (c) 15N CPMAS NMR of 15N Fe-MEPE (material II) embedded in the pores of 20-CA
silica.
15N solid-state NMR was used to estimate the mean chain-length of Fe-MEPE (mate-
rial II) in bulk and in the pores of 20-CA. This analysis is based on signal deconvolution
and observing the change in signal intensities of the Fe-coordinated and free nitrogen
atoms (Ncoor/Nfree) in the ligand molecules before and after uptake of Fe-MEPE in the
host material. Because of the relatively low amount of Fe-MEPE in these samples (com-
pared to the bulk Fe-MEPE materials; cf. Figure 9.2) it was necessary to use 15N-labeled
material to enhance the 15N NMR sensitivity. Only the nitrogen atom in the central ring
of the ligand was accessible for 15N labeling, although labeling of the nitrogen atoms in
the outer rings would be advantageous because they exhibit a bigger difference in chem-
9.2. RESULTS 145
Table 9.5: Number ratio of coordinated and free 15N atoms, Ncoor/Nfree, and mean chain
length,n, of Fe-MEPE (material II) in the bulk state and in the pores of 20-CA.
sample NMR method Ncoor/Nfree n
bulk MEPE MAS 15 ±1)/27-8
bulk MEPE CP MAS 15 ±1)/27-8
CA-MEPE CP MAS 7±1)/23-4
ical shift between free and coordinated ligand. However, the difference in chemical shift
of the central 15N nuclei between free and coordinated ligand is still sufficient to clearly
distinguish these two signals. Figure 9.9 shows the 15N MAS NMR and 15N CPMAS
NMR spectra of the 15N-labeled bulk Fe-MEPE, and the 15N CPMAS NMR spectrum of
a 20-CA-MEPE composite.
The spectrum of 15N-labeled Fe-MEPE in the bulk state was recorded by 15N MAS
NMR (Fig. 9.9a). This technique directly gives the ratio of coordinated and free nitrogen
atoms, Ncoor/Nfree, from the relative intensities of the corresponding signals. However,
this method could not be used for Fe-MEPE embedded in the silica matrix, because of the
relatively low concentration of the Fe-MEPE in the samples and the need to use a long re-
cycle delay time (10 min), which would lead to excessive measurement times. Hence, the
cross-polarisation transfer technique (15N CPMAS NMR) was used to study Fe-MEPE
both in bulk and embedded in the pores (Figure 9.9b and 9.9c). A disadvantage of this
method is that the relative intensities of the peaks in a spectrum may not precisely reflect
the relative concentration of the individual species. This problem was reduced by applying
a sufficiently long contact time (5 ms) in the pulse sequence. Hence, the relative intensi-
ties of the peaks could be mutually compared, and the relative concentrations of the two
nitrogen species were estimated from the respective peak intensities of the CPMAS NMR
spectra. Assuming that the MEPE chains are terminated on both sides by ligand (as to
be expected in the case of ligand excess), then Ncoor/Nfree = 2n/2, where nis the mean
chain length. Results for Ncoor/Nfree and nare summarized in Table 9.5. For bulk solid
Fe-MEPE (Fig. 9.9a and Fig. 9.9b) signal deconvolution yields Ncoor/Nfree=(15±1)/2
for both methods, corresponding to an average chain length n=7−8. The fact that the
MAS and CPMAS techniques give concordant result for bulk Fe-MEPE indicates that the
CPMAS method may indeed be used for estimating the concentration of the two nitrogen
species. On this basis, signal deconvolution of the CPMAS spectrum for Fe-MEPE em-
bedded in silica (Figure 9.9c) yields Ncoor/Nfree=(7±1)/2, corresponding to an average
146 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
chain length n= 3 −4. Hence, the results of Table 9.5 indicate that the mean length of
Fe-MEPE chains in the pores of 20-CA is smaller than in the bulk state. The significant
broadening of the signal corresponding to the Fe-coordinated nitrogen species in Fig. 9.9c
can be attributed to the confined pore geometry, which reduces the possible orientations
of the Fe-MEPE-chains inside the pores of 20-CA.
9.3 Discussion
The present study shows that the uptake of Fe-MEPE into the pores of SBA-15 is strongly
enhanced when the pore walls are decorated with a layer of CA. This enhancement can
be attributed to an ion exchange process and to electrostatic interactions of the Fe-MEPE
polycations with the negatively charged pore walls. These two driving forces of the fast
uptake process, and the relation between high-affinity adsorption and the arrangement
of the MEPE chains in the pores, are discussed below. Finally, the relation between
stoichiometry and mean chain length of Fe-MEPE before and after uptake in the pores
will be taken up briefly.
9.3.1 MEPE uptake by ion exchange
Figure 9.10: Cartoon of the ion exchange transfer of MEPE polycations (indicated as
−(−M2+ −L−)n−)from the solution into the CA functionalized pores. The acetate
counterions are partly neutralized by the released protons.
From the pH change accompanying the initial fast uptake of MEPE from aqueous
MEPE it can be concluded that this process involves an ion exchange in which protons
of the silanol or CA groups at the pore walls of the CA-SBA-15 are replaced with Fe-
MEPE polycations. In the proposed ion exchange process of Figure 9.10, the uptake of
a MEPE polycation of nchain segments is accompanied by a release of nprotons from
9.3. DISCUSSION 147
undissociated or dissociated carboxylic acid groups at the pore wall. These protons leave
the pore space and cause a shift of pH according to the pKavalue of acetic acid. Similarly,
in the uptake of MEPE from KOAc solutions, when a fraction αof the carboxylic acid
groups at the pore wall exists in deprotonated form, nα potassium ions and n(1−α)
protons are released on the uptake of a single polycation of nsegments. In either case, the
release of a larger number of small ions causes an entropic driving force for the uptake
and binding of the polycations in the pore. This offers an explanation for the high rate
of the initial uptake. The finding that the ion exchange process with 10-CA causes a
larger pH change than with pure SBA-15 (Fig. 9.4) indicates that the CA doped material
has a higher proton donor ability than pure SBA-15, because the overall proton exchange
capacities of the two materials are similar. Tentatively, this higher proton donor ability
is attributed to steric effects caused by the flexibility of the propionic acid chains, which
allows the COOH groups to get in closer contact with the metal ions of the Fe-MEPE
chains than the silanol groups of the surface. The overall proton exchange capacity of
10-CA is estimated as 3.2 mmol g−1, about 20% of which are carboxylic acid groups
(see Table 9.1). In the following section a simple estimate is presented, which shows that
the number of protons released from the 10-CA matrix during the fast uptake step of Fe-
MEPE from a salt-free solution (Fig. 9.4) amounts to only a few percent of the proton
exchange capacity of this material.
9.3.1.1 Estimation of pH change
In the uptake measurement of Fe-MEPE from aqueous solutions into 10-CA shown in
Fig. 9.4, 50 mL of a Fe-MEPE solution of initial concentration c0=0.4 mM were brought
in contact with 145 mg of 10-CA silica, corresponding to a MEPE-to-silica mass ratio
µ=0.1. Hence, the number of CA groups in this sample was about 1.10−4mol. The fast
uptake step was connected with a pH change from 5.6 to 4.5, corresponding to a change
in the proton concentration of ∆cH+= 2.9×10−5M. Further protons are consumed by
the protonation of acetate counterions of Fe-MEPE in the solution. From the Henderson-
Hasselbalch equation it follows
log [HA]
[A−]=pKa−pHf= 4.75 −4.50 = 0.25
where HA and A−represent undissociated acetic acid and acetate ions, pKa=4.75 is the
acid dissociation constant of acetic acid and pHf=4.5 is the pH of the solution after ion
exchange. Since [A−] = c0−[HA], where c0=0.8 mM is the overall acetate concentration
(counter ions of Fe-MEPE), we find [HA] = 0.64c0= 5.2×10−5Mfor the concentration
148 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
of undissociated acetic after ion exchange. Hence, the overall increase in the number of
protons released is given by
∆nH+= (0.05L)(2.9mM + 5.2mM)=4×10−6mol.
This result implies that only about 4% of the CA groups at the pore wall are contributing
to ion exchange in the case of pure aqueous MEPE acetate, since for 10-CA the amount
of CA-groups at the pore wall is estimated as 0.6 mmol g−1. This estimate is based on
the assumption that the overall density of surface groups (silanol + carboxylic acid) in the
CA doped materials is equal to the density of the silanol groups in SBA-15 (3.7 nm−2)[5].
From the values of the specific surface area (as) and mol fraction of CA groups (xCOOH)
in Table 9.1 one than obtains the amount of CA groups at the pore wall as 0.6 mmol g−1
(10-CA) and 1.0 mmol g−1(20-CA). Somewhat higher values are obtained by titration
with NaOH by the method described by Bruzzoniti et al. [127].
Accordingly, the ion exchange mechanism process appears to play only a relatively
small role in the uptake of the Fe-MEPE polycations at the weakly acidic pH of the salt-
free solution.
9.3.2 Effect of electrostatic interactions
Beyond ion exchange, electrostatic interactions will play an important role in the uptake
of the Fe-MEPE polycations in the pores. In the aqueous medium, the pore walls of both
pure and carboxylic-acid functionalized SBA-15 are negatively charged, but the acid-
functionalized silicas have a higher surface charge density than the pure SBA-15 samples
in the pH range of the present study. In particular, whereas the surface charge density of
pure silica is only weakly dependent on pH between 4 and 8 [128], a more pronounced pH
dependence of the surface charge density is found for the carboxylic-acid functionalized
SBA-15 materials. This is concluded from FT-IR measurements (see Section 6.5.2) which
revealed that the ratio of the COO−/COOH peak intensities of 10-CA and 20-CA clearly
increase with pH from pH 5.5 (D2O) to pH 8 (phosphate buffer in D2O). On this basis, the
difference in MEPE uptake in 10-CA from milli-Q water (pH 5.5) and 0.1 M KOAc (pH
7.25) shown in Fig. 9.5a may be explained by the interplay of two effects: (i) attractive
electrostatic interactions between a MEPE polycation and the pore wall; and (ii) repulsive
electrostatic interactions between the adsorbed MEPE polycations. Initial adsorption is
dominated by the attractive interaction of individual MEPE molecules with the pore wall,
which is higher at higher pH due to the higher surface charge density. Electrostatic re-
pulsion between adsorbed MEPE chains becomes significant at higher surface coverage,
9.3. DISCUSSION 149
where it tends to limit the maximum adsorption. However, the range of these repulsive
interactions is reduced as the ionic strength of the system is increased by the addition of
electrolyte. Accordingly, the repulsive interactions between the adsorbed MEPE chains
are expected to be weaker in KOAc than in milli-Q water, leading to a higher adsorption
in the presence of KOAc. Hence, both effects are causing a higher adsorption in the pres-
ence of KOAc. These conclusions are in line with observations by other authors. Bruzotti
et. al. [127] have found that the retention of Fe(III) ions in CA-SBA-15 increases with in-
creasing pH, and Katiyar et al. [129] observed that the adsorption of a positively charged
protein (lysoczyme) in SBA-15 is increased by a factor 2 from pH 5 to 8 in appropriate
buffers.
9.3.3 Adsorption affinity and chain packing
The adsorption isotherm of Fe-MEPE in 20-CA (Fig. 9.6b) indicates that the high-affinity
regime extends to a loading of about 150 mg Fe-MEPE acetate per 1 g of the matrix. In
the study of the uptake kinetics, a similar value was found for the upper limit of the fast-
uptake regime. It is of interest to see how this borderline value of the adsorbed amount is
related to possible adsorption geometries of the MEPE chains in the pores. The number
of adsorbed MEPE units per unit mass of the silica matrix is the respective number for
single occupancy of the pores, σ, multiplied with S, the number of chains accommodated
side-by-side in each pore [110]. The former number is given by σ= 4vp/lD2π, where vp
is the specific pore volume, Dthe pore diameter, and `=1.5 nm is the length of a MEPE
unit (distance between Fe2+ ions along the chain). Accordingly, the number of chains
arranged side-by-side can be calculated from the experimental data as
S=NA
Mσ(mFe−MEPE/msilica)(9.1)
where mFe−MEPE/msilica is the mass of Fe-MEPE per unit mass of silica matrix, Mis the
molar mass of Fe-MEPE, and NAis the Avogadro’s constant. The maximum number of
MEPE chains, Smax, that can be accommodated side-by-side in a close-packed monolayer
at the pore wall can be estimated from the diameter of the MEPE chain (d=1.2 nm) and
the pore diameter D. For 20-CA (D=8.0 nm), this yields Smax =18. Table 9.2 indicates
that this value is attained at loadings between sample CA-MEPE-6 and CA-MEPE-7,
that is, near the upper limit of the high-affinity adsorption regime of Fe-MEPE in the
silica matrix. This supports the conjecture that high-affinity adsorption occurs as long as
the Fe-MEPE chains are accommodated in the monolayer in direct contact with the pore
150 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
wall. Values of S>18, as they are obtained for the samples CA-MEPE-7 and CA-MEPE-
8, then indicate the formation of a second layer of Fe-MEPE in the pores or at the outside
of the silica grains, or both.
As mentioned earlier, some of the CA-MEPE composite samples of Table 9.2 were
also studied by SAXD. No significant changes in the intensities of leading Bragg reflec-
tions (10, 11, 20; cf. Fig. 9.7b) were observed, as they were found for physisorbed films
of vapors at the pore wall of SBA-15 [130,26]. However, the MEPE-loaded samples ex-
hibit a significantly lower level of diffuse scattering than the empty material in the region
of higher q, so that the 21 Bragg reflection that was masked by diffuse scattering in the
empty material is detectable in the material of highest loading with MEPE. Because dif-
fuse scattering is partly due to the roughness of the pore wall, we may conclude that the
formation of a layer of MEPE chains reduces the effective roughness of the pore walls. On
the other hand, the fact that no significant modulation of the Bragg intensities is observed
for the CA-MEPE materials indicates that the layer of MEPE chains is less uniform than
a physisorbed fluid film at the pore walls.
9.3.4 Stoichiometry and chain length of MEPE in the pores
Information about the stoichiometry and chain length of Fe-MEPE in the pores was de-
rived from the number ratio of nitrogen and iron atoms (N/Fe) as determined by XPS, and
from the number ratio of coordinated and free nitrogen atoms of the ligand (Ncoor/Nfree)
as determined by 15N solid-state NMR. The XPS measurements, which were made with
Fe-MEPE material I, give the same N/Fe atom ratio for the three CA-MEPE composites
studied. This indicates that the stoichiometry of Fe-MEPE in the pores is independent of
the degree of loading which differs strongly for the three samples (Table 9.4). The experi-
mental atom ratio N/Fe of 5.4 ±0.3 instead of 6 indicates an excess of metal over ligand
in the complex, as to be expected for Fe-MEPE material I (see Section 8.1). Because the
number of segments, n, of a chain terminated by metal at both sides (Mn+1Ln) is related to
the atom number ratio as N/Fe = 6n/(n+1), the XPS data indicate a mean chain length
n≈9 or greater for Fe-MEPE material I in the CA-MEPE composite materials. However,
this value must be taken with caution, as it relies on the assumption that the MEPE chains
are terminated by metal on both ends. Lower values of nwould result on the assumption
that a fraction of the chains are of type MnLninstead of Mn+1Ln.
The 15N MAS NMR measurements were performed with a 15N-labeled Fe-MEPE
(material II) having a ligand excess (see Section 8.1), for which one expects that the
chains are terminated by ligand on both sides (MnLn+1). From the measured ratio of
9.4. CONCLUSIONS 151
coordinated and free nitrogen atoms, Ncoor/Nfree, it was concluded that this Fe-MEPE
material has a mean chain length n=7-8 in the bulk state, and a value of about n=3-
4 in the pores of 20-CA. Because the measurements were made with the same MEPE
material, this result implies that the chains are breaking into two or more fragments when
the material is dissolved in the aqueous phase or during the uptake into the pores. For
instance, if a MEPE chain MnLn+1 breaks into two fragments,
MnLn+1 −→ MmLm+1 +(ML)n−m,
one of the fragments will not be terminated by ligand on both sides, and thus the relation
Ncoor/Nfree =nis not strictly applicable for estimating the chain length in the pores.
Instead, for the mixture of the fragments, one finds, for any m<n,Ncoor/Nfree =
(2n−1)/3, which is smaller than n. Hence, the important finding that Ncoor/Nfree in the
pores is smaller than in the bulk state clearly implies that the MEPE chains are breaking
into smaller entities when transferred into the pores. A direct complexation of Fe ions by
the surface acid groups might also be envisaged in view of the relatively strong binding
of Fe ions to these groups [127]. However, since direct complexation involves a release
of ligand L, this alternative seems rather unlikely in view of the high complex formation
constants of Fe(II) with the Terpy-type ligands [88].
9.4 Conclusions
This work has shown that intrinsically stiff metallo-supramolecular polyelectrolytes like
Fe-MEPE are adsorbed into the cylindrical nanopores of SBA-15 and that the adsorption
affinity is strongly enhanced when the pore walls are functionalized with carboxylic acid
groups. Two uptake mechanisms have been identified: a fast process and a slower sub-
sequent process. The fast process is connected with a lowering of pH in the supernatant
solution, indicating an ion exchange mechanism. It is favored by increasing the pH which
causes stronger dissociation of the surface functional groups and thus stronger electro-
static interactions of the polycations with the negatively charged pore walls. The nature
and rate-determining resistance of the slow process is not fully understood. Generally, an
increase of temperature from 20 to 50◦C causes an increase of the uptake rate and leads
to higher levels of uptake of Fe-MEPE in the pores. For the material 20-CA, in which
nearly one-half of the surface silanol is replaced by propionic acid groups, the maximum
adsorbed amount corresponds to more than a densely packed layer of Fe-MEPE-chains
aligned side-by-side along the pores.
152 CHAPTER 9. FE-MEPE IN CA DOPED SBA-15 SILICA
Except for the high uptake regime, the results of the uptake measurements by UV-vis
analysis of the supernatant are consistent within the error limits with AAS determinations
of the iron content in the dry samples. However, XPS measurements gave significantly
higher values at high Fe-MEPE loadings. Because XPS is a surface sensitive probe, this
may indicate that at high loadings Fe-MEPE is accumulating near or at the surface of the
grains. The metal-to-ligand stoichiometry and average chain length of Fe-MEPE in the
matrix was characterized by XPS and 15N solid-state NMR spectroscopy. XPS measure-
ments of the atomic ratio of nitrogen to iron showed that the stoichiometry of the Fe-
MEPE complex was independent of the level of uptake in the matrix. This result suggests
that the Fe-MEPE chains are transferred into the pores as they exist in the solution. 15N
solid-state NMR measurements of the ratio of coordinated and free nitrogen (Ncoor/Nfree)
were made with a 15N-labeled Fe-MEPE sample with ligand excess. Assuming that in the
Fe-MEPE material obtained under these conditions the chains are terminated by ligand at
both ends, the mean chain length was taken as Ncoor/Nfree =n. For the bulk material, a
value n= 7 −8was obtained. For the same Fe-MEPE material in the pores of 20-CA,
Ncoor/Nfree =n= 3 −4was found. From the increase of the relative number of free
(noncoordinated) nitrogen atoms of the ligand it was concluded that the Fe-MEPE chains
break into smaller entities in the pores, because fragmentation leads to a higher proportion
of uncoordinated nitrogen.
Chapter 10
Kinetic Study of Fe-MEPE Uptake
Abstract
In this section a quantitative analysis of the adsorption kinetics and adsorption equilib-
rium of Fe-MEPE in the pores of carboxylic-acid functionalized SBA-15 (CA-SBA-15)
is presented. The starting point is a comparison of the uptake kinetics of Fe-Terpy and
Fe-MEPE. As in the case of Fe-MEPE, the affinity of Fe-Terpy for CA-SBA-15 is greater
than for pure SBA-15. The uptake of Fe-Terpy is much faster than the uptake of Fe-
MEPE and equilibrium is reached within a few hours, whereas equilibrium adsorption
of Fe-MEPE is attained only after several hundred hours. The equilibrium adsorption
isotherm of Fe-MEPE can be described satisfactorily by the Langmuir model. A fit of
the equilibrium data yields a high adsorption constant and a high monolayer capacity of
230 mg MEPE /g silica which is in fact higher than the value estimated for a monolayer
of Fe-MEPE chains at the pore wall. The uptake curves of Fe-MEPE can be represented
by a sum of two exponential decay functions for the fast initial process, and a subsequent
slow uptake process leading to the adsorption equilibrium. The analysis suggests that the
uptake of Fe-MEPE into the pores is controlled by surface layer resistance and not by
internal diffusion inside the pore channels. The estimated mass transfer coefficient αof
10−9-10−10 m s−1indicates a slow entering process of Fe-MEPE chains into the pores.
153
154 CHAPTER 10. KINETIC STUDY OF FE-MEPE UPTAKE
10.1 Fe-Terpy-uptake vs. Fe-MEPE-uptake
We studied the uptake of Fe-Terpy, a simpler system than Fe-MEPE because of its mono-
meric character, to confirm the higher affinity of carboxylic acid functionalized SBA-15
for these kind of complexes compared with pure SBA-15. Figure 10.1 shows uptake
isotherms of Fe-Terpy into pure SBA-15 and 20-CA under the same experimental condi-
tions. As can be seen, the amount of adsorbed Fe-Terpy in 20-CA is grater by a factor
four than the adsorbed amount in pure SBA-15. Since the surface area of 20-CA is not
greater (but in fact smaller) than that of pure SBA-15, this result is a clear evidence for the
higher adsorption affinity of 20-CA for the Fe-Terpy- complexes. In both cases the uptake
took place in pure water at pH 5.5. Under these conditions the surface of pure SBA-15
is negatively charged because the isoelectronic point of SBA-15 is near pH 2. Still, its
affinity for the positively charged Fe-Terpy complex is much less than 20-CA. This ob-
servation could be ascribed to (i) the higher negative surface charge of 20-CA material in
comparison to pure SBA-15 because of its higher proton donor ability in solution, and (ii)
the higher entropy during the adsorption in 20-CA due to the release of acetate counter
ions of Fe-Terpy in solution as a consequence of their replacement by wall carboxylic
acid groups after the adsorption.
0 120 240 360 480 600
0
2 0
4 0
6 0
8 0
100
2 0 - C A - M E P E
2 0 - C A - T e r p y
m g c o m p l e x / g s i l i c a
t i m e / h
S B A - 1 5 - T e r p y
Figure 10.1: Uptake isotherms of Fe-Terpy in pure SBA-15 (•), in 20-CA (◦) and of
Fe-MEPE in CA-20 (N). The preparations were performed in milli-Q-water and with a
complex to silica ratio of µ=0.1. Curves are drawn to guide the eye.
Also shown in Figure 10.1 is a comparison of the uptake of Fe-Terpy and Fe-MEPE
10.1. FE-TERPY-UPTAKE VS. FE-MEPE-UPTAKE 155
in 20-CA silica material. A pronounced difference in the adsorption kinetics of Fe-Terpy
and Fe-MEPE into the pores of 20-CA can be seen. Although in the final equilibrium state
after 300 h the amount of adsorbed MEPE and Terpy is nearly the same, the uptake kinet-
ics of the two complexes is grossly different: The uptake curve of Fe-Terpy shows a fast
initial increase followed by a plateau, indicating a rapid achievement of the equilibrium.
The uptake curve of Fe-MEPE also exhibits a rapid initial increase but to a much lower
extend than for Fe-Terpy, and this fast process is followed by a much slower uptake ex-
tending over several hundred hours. We believe that this difference in the uptake behavior
is caused by the the polymeric character of Fe-MEPE. The grossly different equilibration
200 300 400 500 600 700 800 900
6 5
7 0
7 5
8 0
8 5
9 0
9 5
mg complex / g silica
time/h
50°C
3h
50°C
4.5h
MEPE
Terpy
Figure 10.2: Uptake isotherms of Fe-Terpy (N) and Fe-MEPE () in 20-CA at 20◦C. The
vertical lines show the times, at which the temperature was raised from 20◦C to 50◦C for
a period of 3h and for a period of 4.5 h, respectively. Curves are drawn to guide the eye.
times for the adsorption of Fe-Terpy and Fe-MEPE is also revealed by temperature jump
experiments shown in Figure 10.2. In these experiments the temperature of the systems
was taken from 20◦C to 50◦C and back to 20◦C after 3-4 h. The adsorbed amounts deter-
mined by UV-vis measurement of the supernatant solution immediately after the heating
step indicated that temperature increase leads to a decrease of Terpy adsorption and to an
increase of MEPE adsorption. After cooling to 20◦C, the adsorbed amount of Fe-MEPE
remains at the higher level of adsorption while the adsorption of Fe-Terpy returns to the
higher adsorption level attained before the temperature jump. Figure 10.2 shows that this
temperature jump experiment could be repeated at a later time, when the adsorption of
Fe-MEPE in the matrix had attained a higher level while the adsorption of Fe-Terpy had
156 CHAPTER 10. KINETIC STUDY OF FE-MEPE UPTAKE
remained at a nearly constant value.
The reversible changes in the adsorbed amount observed in the case of Terpy/20-CA
system suggest that these changes are induced by a temperature-induced shift of the ad-
sorption equilibrium to a lower level of adsorption at higher temperature, as to be expected
for an exothermic adsorption process. Accordingly, the adsorbed Fe-Terpy molecules dif-
fuse out of the pores to establish the new equilibrium at the higher temperature. After
cooling the system back to 20◦C , the system returns to the original equilibrium state.
In the case of Fe-MEPE adsorption, equilibrium has not yet been attained at he time of
the temperature jump. In this case an increase of the temperature increases the adsorp-
tion rate, and therefore, the adsorption remains at the higher level after return to 20◦C.
Possibly, this behavior comes from the supramolecular coordination polymer character of
Fe-MEPE (see section 7.1). An increase of the temperature, i.e. kinetic energy supply,
leads to shorter chain lengths making the entering and diffusion of the chains in the pore
channels easier.
10.2 Adsorption kinetics of Fe-MEPE
0 200 400 600 800
0
5 0
100
150
200
250
0 . 3 0
0 . 2 1
0 . 1 6
0 . 1 2
m g M E P E / g 2 0 - C A
t i m e / h
µ
Figure 10.3: Effect of the initial Fe-MEPE concentration (MEPE to 20-CA mass ratio µ)
on the adsorption rate. All preparations were performed in 50 ml pure water. The solid
lines are fits. The short vertical lines indicate the complete adsorption of Fe-MEPE by
20-CA.
In order to clarify the mechanism of Fe-MEPE adsorption into the pores of CA-SBA-
10.2. ADSORPTION KINETICS OF FE-MEPE 157
Table 10.1: Parameters from fitting the MEPE adsorption curves to the Equation 10.1
µc0/mM A1/mg g−1t1/h A2/mg g−1t2/h R
0.12 0.5 25±2 0.004±0 94±1 18±1 0.99932
0.16 0.6 52±2 1.3±0.1 106±2 54±2 0.99968
0.21 0.8 70±2 1.8±0.2 131±2 122±4 0.99924
0.30 1.2 80±6 2.1±0.3 143±4 178±12 0.99875
µ: Fe-MEPE to silica mass ratio; c0: initial Fe-MEPE concentration
15 silica systematic studies were performed by measuring the uptake rate with different
initial concentrations of Fe-MEPE, that is, at different silica to MEPE mass ratios µ(see
Figure 10.3). In order to attain acceptable equilibration times for the uptake of Fe-MEPE,
these adsorption measurements were conducted at 50◦C because we found that the ad-
sorption rate of Fe-MEPE increases by increasing the temperature (see Section 9.2.3).
20-CA was used as adsorbent because of its higher MEPE adsorption capacity. In view
of the high stability of Fe-MEPE in salt free solution the adsorption studies were made
in pure water. The uptake curves displayed in Figure 10.3 can be fitted by a sum of two
exponential decay functions
y=y0−A1e−t/t1−A2e−t/t2(10.1)
where yis the uptake at time tand y0is the limiting uptake at t→ ∞. Hence, the uptake
curves may be described as the combination of two first-order processes with different
relaxation times. The uptake time t1is short in comparison to the uptake time t2. The
parameters obtained from the fit of the MEPE uptake curves to the Equation 10.1 are
summarized in Table 10.1. Table 10.1 reveals that the time constant t1is of the order of
1 h or less, while the time constant t2is of the order of 100 h. Hence the time scales of
the two processes are separated by two orders of magnitude. Both time constants increase
with increasing MEPE-to-silica mass ratio µ.
In order to interpret these experimental data by means of the kinetic standard models
for diffusion-controlled systems discussed in Section 7.5 viz. diffusion control through
the surface layer (Eq. 7.12) and internal diffusion control (Eq. 7.15) the parameters a∞
and a∗of Eq. 7.12 have to be extracted from Eq. 10.1. We consider this relation for three
different times
(i) at t=0 : y=0 and Eq. 10.1 yields y0=A1+A2
158 CHAPTER 10. KINETIC STUDY OF FE-MEPE UPTAKE
(ii) at t=t∗so that t1t∗t2: y= a∗
a∗=y0−A2
(iii) at t t∗: y=a∞
a∞=y0
Accordingly, we find a∞=A1+A2and a∗=A1. By substituting a∞and a∗into Eq.
10.1 and changing yto atone obtains
at=a∞−a∗e−t/t1−(a∞−a∗)e−t/t2.(10.2)
Since t=t∗and t1t∗in case (ii), e−t/t1≈0. Thus
at=a∞−(a∞−a∗)e−t/t2(10.3)
at−a∗
a∞−a∗
= 1 −e−t/t2.(10.4)
This analysis is based on the fact that the time scale of the two processes are well-
separated so that the quantity a∗introduced by (ii) is well-defined.
0246810 12
0,0
0,2
0,4
0,6
0,8
1,0
(at-a*)/(m¥-a*)
Öt/h1/2
m = 0.12
036912 15 18
0,0
0,2
0,4
0,6
0,8
1,0
(at-a*)/(m¥-a*)
Öt/h1/2
m = 0.16
0510 15 20 25 30
-0,3
0,0
0,3
0,6
0,9
1,2
(at-a*)/(m¥-a*)
Öt/h1/2
m = 0.21
0510 15 20 25 30
0,0
0,2
0,4
0,6
0,8
1,0
(at-a*)/(m¥-a*)
m = 0.30
Öt/h1/2
Figure 10.4: Fe-MEPE uptake curves fitted with internal diffusion model.
10.2. ADSORPTION KINETICS OF FE-MEPE 159
If the Fe-MEPE sorption is controlled by the internal diffusion process (see Eq. 7.15),
the plot (at−a∗)/(m∞−m∗)vs. √tshould give a straight line through the origin with
slope 2kpD/π. Figure 10.4 shows that the data points cannot be fitted well by Equation
7.15 indicating that the adsorption process is not controlled by the internal diffusion.
After its linearization Equation 10.4 takes the form ln(1 −(at−a∗)/(a∞−a∗)) =
−(1/t2)t. Hence, if the adsorption is controlled by the diffusion through the surface layer,
the plot of ln(1 −(at−a∗)/(a∞−a∗)) vs. time (t) should give a straight line through
the origin with slope (-1/t2) which is equal to (-αk), where αis the mass transfer rate
coefficient and kis the ratio of the external area to volume (see Section 7.5). Plots for all
four MEPE-to-silica mass ratios µare displayed in Figure 10.5. As can be seen, the data
020 40 60 80 100
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
ln(1-(at-a*)/(a¥-a*))
time/h
m = 0.16
040 80 120 160 200
-2,0
-1,5
-1,0
-0,5
0,0
ln(1-(at-a*)/(a¥-a*))
time/h
m = 0.21
010 20 30 40 50 60
-4
-3
-2
-1
0m = 0.12
time/h
ln(1-(at-a*)/(a¥-a*))
030 60 90 120 150 180
-1,2
-0,9
-0,6
-0,3
0,0
ln(1-(at-a*)/(a¥-a*))
time/h
m = 0.30
Figure 10.5: Fe-MEPE uptake curves fitted with surface-layer resistance model.
can be represented well by the surface-layer resistance model. The parameters obtained
from the fits are summarized in Table 10.2.
To estimate the external surface area (Aext) we assume a hexagonal cylindrical shape
of SBA-15 particles as indicated in Figure 10.6. The ratio of external area to volume k≈
2.06×107m−1is calculated using geometrical considerations (cf. Equation 10.8 below).
In order to determine the mass transfer rate coefficient α, we need to know the ratio of
the external surface area to total pore volume k=Aext/Vp. For this purpose we assume
160 CHAPTER 10. KINETIC STUDY OF FE-MEPE UPTAKE
Table 10.2: Parameters from fitting the MEPE uptake curves to the Equation 10.1
µslope R kα/s−1α/m s−1
0.12 -0.0595±0.0009 0.998425 17 2.3×10−10
0.16 -0.0188±0.0003 0.997573 53 7.2×10−10
0.21 -0.0082±0.0002 0.989572 122 1.6×10−9
0.30 -0.0058±0.0001 0.997821 172 2.3×10−9
µ: Fe-MEPE to silica mass ratio; k: external area to volume ratio ≈2.06×107m−1;α: mass transfer rate
coefficient
D
c
a0
Figure 10.6: left: Scheme of a hexagonal cylindrical SBA-15 particle, consisting of
cylindrical mesopores, for estimation of the external surface area; right: Unit cell of 2D
hexagonally ordered pores with lattice parameter a0.
that the SBA-15 particles have an ideal hexagonal cross section and a high aspect ratio,
i.e., the length cof the cylinder is large in comparison to the cross sectional diameter D
of the cylinder. In this case Aext can be approximated as
Aext ≈6cD
√3= 2√3c·D(10.5)
where the diameter Dof the cross sectional area of the cylinder is estimated as 450 nm
from the SEM micrographs (see Figure 10.6 left). The total pore volume Vpof the particle
is given by
Vp=πd2
4N·c(10.6)
where dis the diameter of the mesopores and Nis the number of pores in the cylindrical
particle, which is approximately given by
N=AQ
AU
=√3/2D2
√3/2a2
0
=D2
a2
0
(10.7)
Here AQdenotes the cross sectional area of the cylinder and AUdenotes the area of a unit
10.3. ADSORPTION EQUILIBRIUM OF MEPE 161
cell of the 2D hexagonally arranged pore lattice (lattice parameter a0). By substituting
Equation 10.7 into Equation 10.6 and using Equation 10.5 one gets
k=Aext
Vp≈8√3a2
0
π d2D≈0.02 nm−1(10.8)
where we have used a0=11.6 nm (from SAXD), d=8 nm (from nitrogen adsorption) and
D=450 nm (from SEM micrographs of CA-SBA-15 particles). This relation applies to
long worm-like particles for which the front and the back cross-sectional areas of the
hexagonal cylinder are disregarded in Eq. 10.8. In this case the unknown quantity c
which appears as a factor both in the approximation of Aext and Vpcancels out. For
shorter particles the external area-to-volume ratio of the adsorbent will be larger, when
the cross section of particles is taken into the account. Thus we expect that kwill be in a
range 0.02 < k < 0.05 nm−1. With the known value of kthe mass transfer coefficient α
can be extracted directly. Values of α(based on k= 0.02nm−1) for each initial Fe-MEPE
concentration are included in Table 10.2.
As can be seen the mass transfer rate coefficient increases as the initial Fe-MEPE
concentration is increased. A comparison with the mass transfer rate of methanol vapor
in Ferrite, for which α= 4.2×10−8m s−1) [131], shows that the mass transfer of Fe-
MEPE in CA-SBA-15 is a rather slow process.
10.3 Adsorption equilibrium of MEPE
The Langmuir isotherm (Eq. 7.10) was used to represent the adsorption equilibrium re-
lationship (see Figure 10.7). The data for aeq (equilibrium loading of Fe-MEPE at the
surface of CA-SBA-15) and ceq (equilibrium concentration of Fe-MEPE in solution) are
taken from the uptake curves of Fig. 10.3 for equilibration times of 300-350 h.
After linearization the Langmuir equation takes the form
1
aeq
=1
am
+1
amKL
1
ceq
.(10.9)
The plot of 1/aeq vs. 1/ceq gave a straight line. From the slope and y-intercept one finds
am=230 mg g−1, and KL=107.5 lmmol−1which corresponds to 150.5 lg−1. Figure 10.7
shows that the data for aeq are well represented by the Langmuir equation. The high value
of KLindicates a high adsorption affinity of Fe-MEPE in 20-CA. High adsorption affinity
is often observed for polymer or protein adsorption from solution [132]. In this type of
adsorption the molecules bind so strongly that no remnant can be detected in solution,
162 CHAPTER 10. KINETIC STUDY OF FE-MEPE UPTAKE
0 , 0 0 , 1 0 , 2 0 , 3
0
5 0
100
150
200
250
ae q
ceq/ m M
2 0 - C A
Figure 10.7: Langmuir isotherm of samples prepared with different µat 50◦C.
which is exactly what we observed for the adsorption of Fe-MEPE in the carboxylic acid
functionalized SBA-15.
20-CA has a pore diameter of 8 nm and a pore volume of 0.60 cm3g−1. On the other
hand, Fe-MEPE can be considered as a cylinder of diameter 1.2 nm. A schematic illus-
tration true to scale shows that a maximum of 18 Fe-MEPE chains can be accommodated
side-by-side in a 8 nm pore (see Figure 10.8). On the basis of the model discussed in
Section 9.3.3, 18 Fe-MEPE chains correspond to an adsorbed amount of 183 mg MEPE/g
silica, which on the basis of the geometrical model should represent the monolayer ca-
pacity. However, this value is smaller than the value am=230 mg g−1derived on the basis
17-18
Figure 10.8: The schematic illustration of Fe-MEPE in a 8 nm pore. According to the
illustration a maximum of 18 Fe-MEPE chains can be accommodated side-by-side in a 8
nm pore. The illustration is true to scale.
of the Langmuir adsorption isotherm which would correspond to ca. 25 Fe-MEPE chains
10.4. CONCLUSIONS 163
side-by-side in the pores. Values higher than 18 may indicate the formation of a second
layer of Fe-MEPE in the pores, or at the outside of the silica particles, or both. An alterna-
tive explanation of the high value of amderived from the Langmuir isotherm is suggested
by nitrogen adsorption measurements of pure 20-CA silica, which indicated an enlarge-
ment of the pore volume of silica after exposure to water (see Section 6.2.4). From the
fact that the pore diameter remained constant while the pore volume was increased, it was
suggested that the exposure to water makes part of the pores which were formerly blocked
accessible to adsorption. This would imply that the silica can adsorb more Fe-MEPE than
estimated on the basis of the nitrogen adsorption data obtained prior to the exposure to
water.
10.4 Conclusions
In this chapter, comparative adsorption measurements of Fe-Terpy and Fe-MEPE in pure
and carboxylic-acid functionalized SBA-15 (20-CA) were performed. These showed (i)
the adsorption affinity of Fe-Terpy in 20-CA is much higher than in pure SBA-15; (ii) un-
der comparable conditions the adsorption of Fe-Terpy is much faster than the adsorption
of Fe-MEPE. Result (i) confirms the outcomes of the uptake study of Fe-MEPE in the
pure and carboxylic acid doped silica materials (Chapter 9). Most likely this can be as-
cribed to the higher negative surface charge of 20-CA in comparison to pure SBA-15 and
to a higher entropic driving force connected with ion exchange in 20-CA. The difference
in the uptake kinetics of Fe-MEPE and Fe-Terpy are attributed to the polymeric character
of Fe-MEPE. The uptake curve of Fe-Terpy shows a fast initial increase leading directly
to a plateau, indicating a rapid achievement of the adsorption equilibrium. The uptake
curve of Fe-MEPE also exhibits a fast initial uptake. But this is followed by a much
slower second uptake process extending for many days. The uptake curves of Fe-MEPE
could be modelled by a combination of two exponential decay functions, the first with a
short relaxation time t1(corresponding to the fast adsorption process), the second with a
much longer relaxation time t2, accounting for the slow uptake process that leads to the
adsorption equilibrium. An analysis of the kinetic data on the basis of models for surface-
layer-resistance or internal diffusion as rate determining process showed that the diffusion
during the uptake of Fe-MEPE into the pores is controlled by surface layer resistance and
not by internal diffusion inside the pore channels. This means that the rate-limiting step
of adsorption during the uptake is the entering of the Fe-MEPE chains into the pores and
not the diffusion of the MEPE chains inside of the pore channels. The estimated mass
164 CHAPTER 10. KINETIC STUDY OF FE-MEPE UPTAKE
transfer coefficient αof 10−9-10−10 m s−1is small, indicating a slow entering process of
Fe-MEPE chains into the pores.
The different equilibration times for the adsorption of Fe-Terpy and Fe-MEPE are also
revealed by repeated temperature jump experiments. A temperature increase (50◦C) leads
to a decrease of Terpy adsorption and to an increase of MEPE adsorption. After cool-
ing (20◦C), the adsorbed amount of Fe-MEPE remains at the higher level of adsorption
while the adsorption of Fe-Terpy returns to the higher adsorption level attained before the
temperature jump. The reversible changes in the adsorbed amount observed in the case of
Terpy uptake suggests that these are induced by a temperature-induced shift of the adsorp-
tion equilibrium to a lower level of adsorption at higher temperature, as to be expected
for an exothermic adsorption process. In the case of Fe-MEPE adsorption equilibrium
has not yet been attained at the time of the temperature jump. In this case an increase of
the temperature increases the adsorption rate, and therefore, the adsorption remains at the
higher level after returning to 20◦C. Moreover, the Langmuir adsorption isotherm turned
out to describe the solid-liquid equilibrium of Fe-MEPE satisfactorily. The high value of
the Langmuir constant KLindicates a high adsorption affinity of Fe-MEPE in 20-CA.
Chapter 11
Summary and Outlook
In this Thesis, pure and functionalized 2D-hexagonally ordered mesoporous silica mate-
rials have been synthesized and characterized by a large variety of methods. The func-
tionalization was performed by two different routes:
1. post-modification by surface grafting
2. functionalization by co-condensation (one-pot synthesis)
The functionalized materials were used (i) for examination of the surface properties of
mesoporous MCM-41 and SBA-15 silicas; (ii) for acidity measurements in water-free
confined geometry; (iii) as hosts for metallo-supramolecular coordination polymers (MEPE).
Grafting route was used for modifying the surface of MCM-41 and SBA-15 silicas
using (CH3)2-Si(OCH3)2and (CH3)-Si(OCH3)3as grafting reagents. In the next step the
arrangement of surface silanol groups at the silica surfaces was studied by solid-state
15N NMR method after pyridine adsorption into the pores. We found that the number of
covalent bonds to the surface formed by the functional reagents is affected by the surface
morphology. It was concluded that for high-quality MCM-41 silicas the distance between
neighboring surface silanol groups at the pore walls is greater than 0.5 nm. As a result,
the di- and tripodical functional reagents can form only one covalent bond to the surface.
In the case of SBA-15, however, the formation of two covalent bonds to the surface is
possible as well. The residual hydroxyl groups of surface-bonded functional reagents
either remain free or interact with other reagent molecules. Accordingly, the number of
surface silanol groups at a given MCM-41 or SBA-15 silica needs not decrease but may
increase after grafting with CH3Si(OH)3reagent.
Co-condensation route was used for functionalization of the SBA-15 silica with acidic
groups of different pKavalues in the liquid phase. These were carboxylic acid, phospho-
165
166 CHAPTER 11. SUMMARY AND OUTLOOK
nic acid and sulfonic acid. This synthesis route was chosen because the high surface
coverage of silica by acid groups which is needed for the acidity measurements can-
not be reached by the grafting method. The following features are common to all acid-
functionalized SBA-15 materials prepared by the co-condensation route:
1. The higher the proportion of the functional silane φin the silica precursor, the
lower is the amount of polymer template in the as-synthesized functionalized SBA-
15 and the higher is the degree of the polymer removal achieved by the sulfuric acid
treatment.
2. The microporosity of acid-functionalized SBA-15 samples is lower than that of pure
SBA-15.
3. Well-ordered SBA-15 materials are obtained with a maximum molar percentage of
functional silane 20 < φ< 30.
The maximum surface coverage of the pore walls and the yield of the functionalization
reaction reached by the co-condensation method are different for the three acids. The
highest surface coverages of SBA-15 by carboxylic acid, phosphonic acid, and sulfonic
acid groups are 50%, 40% and 30%, respectively. The reaction yields of the functional-
ization reactions show a similar tendency. The optimum reaction conditions for producing
ordered SBA-15 type materials with a high functionalization degree are different, and they
strongly depend on the (hydrophilic or hydrophobic) nature of the functional silane used.
An ideal functional silane should be able to interact attractively both with the hydrophilic
silica precursor TEOS and with the hydrophilic PEO block of the blockcopolymer (P123)
template. However, it must not interact equally well with the hydrophobic PPO and hy-
drophilic PEO blocks of the copolymer as this would affect the microphase separation
of the two blocks which is needed for the formation of ordered materials. Our studies
showed that these requirements become increasingly important at higher mole percent-
ages φof the functional silane used in the reaction mixture, and thus for the synthesis
of SBA-15 with a high degree of functionalization. Moreover, the hydrolysis and con-
densation rate of the functional silane relative to that of TEOS plays a crucial role for
the ordering of silica material obtained. Ideally, the hydrolysis and condensation rates of
TEOS and functional silane should be similar. Furthermore, at the chosen reaction condi-
tions the rate of condensation of the silica sources must not be faster than their hydrolysis
rates. Presumably, the hydrophilicity of the functional silanes decrease in the order STHS
> CTES > PTES. In the case of carboxylic acid functionalization, CTES is expected to be-
come more hydrophilic during the synthesis due to the partial hydrolysis of the CN head
167
group, whereas the phosphonic acid silane PTES with its not-cleavable phosphonic acid
ester groups remains rather hydrophobic. Thus, for CTES and the sulfonic acid silane
STHS an attractive hydrophilic interaction with PEO and TEOS is suggested during the
synthesis. However, simultaneous addition of TEOS and CTES leads to the formation
of disordered silica materials, indicating different hydrolysis/condensation rates of these
silica sources. Two probable explanations for this behavior are (i) the hydrolysis rate of
the CTES is slower than that of TEOS, and (ii) hydrolysis of CN-groups to COOH-groups
increases the hydrophilicity of the functional reagent and yields a better interaction with
both the hydrophilic POE block of copolymer and TEOS and a weak interaction with the
hydrophobic PPO block of copolymer which effects a better microphase separation. This
suggests that a prehydrolysis of CTES before the addition of TEOS will improve the for-
mation of functionalized products, which indeed leads to ordered composite materials up
to φ= 20 of CTES in the silica precursor mixture. In the case of sulfonic acid functional-
ization using STHS, prehydrolysis does not play a role, because STHS already exists in
the hydrolyzed form. Simultaneous addition of TEOS and STHS yields ordered function-
alized SBA-15 material. However, the degree of incorporation of the sulfonic acid reagent
is low compared to the other functionalization reactions. Probably this can be explained
by the slow condensation rate of the hydroxyl groups of the highly acidic functional silane.
In contrast, the interactions of hydrophobic phosphonic acid silane PTES both with the
block copolymer and TEOS are unfavorably. Presumably PTES interacts both with PEO
and PPO blocks of the block copolymer, and thus reduces the microphase separation nec-
essary for the formation of ordered silicas. Nevertheless, the system seems to tolerate
this up to PTES contents of φ= 10. Increasing φto 15 and 20 causes a shortening of the
worm-like structures and leads to the forming of disk-like stuck-together structures under
preservation of the 2D-hexagonally order. The amount of PTES in the precursor mixture
can be increased to φ=15 without destruction of the worm-like morphology of SBA-15
by prehydrolysis of TEOS. This probably indicates that the hydrolysis of TEOS is slow
in comparison to that of PTES.
Acidity measurements of the functional groups at the pore wall were performed by
15N solid-state NMR after adsorbing pyridine into the silica materials. These measure-
ments performed in cooperation with the NMR Group in Freie Universität Berlin in the
framework of SFB 448 showed that all three acids can be deprotoned pyridine. This is
unexpected in the case of carboxylic-acid functionalized SBA-15 which is a weak acid in
aqueous solution (acid dissociation constant pKa=4.7). The surface silanol groups have a
similar acidity as carboxylic acids in water but are not protonated by pyridine in the dry
168 CHAPTER 11. SUMMARY AND OUTLOOK
state at the pore wall. The high proton-donor ability of carboxylic acid functional groups
at the pore walls may be due to the formation of hydrogen-bonds between several car-
boxylic acid groups at the surface. It is known that in solutions the formation of hydrogen-
bonded chains of carboxylic acid molecules lead to an enhanced acidity of the terminal
proton. The high local concentration of carboxylic acid groups in the CA-functionalized
SBA-15 material may lead to a similar conjugation effect in the pores. Further measure-
ments within the PhD thesis of Daniel Mauder (FU-Berlin) are in progress to check this
conjecture. In addition to these experiments, comparative pH dependent proton donor
ability measurements of 10-CA-SBA-15 (10-CA) and 20-CA-SBA-15 (20-CA) samples
were performed by FT-IR spectroscopy. An increase of the pH from 5.5 to 8.0 leads to
the appearance of the asymmetric and symmetric stretching bands of carboxylate anions,
while the intensity of the C=O stretching band of COOH group decreases. The FT-IR
measurements reveal an interesting difference between 10-CA and 20-CA: Whereas for
10-CA the C=O stretching has almost disappeared at pH 8, it is only weakened in the case
of 20-CA. This suggests, that in 10-CA, where the surface is covered by COOH groups
to 20%, all COOH groups are separated and accessible while in 20-CA with 50% surface
coverage only a part of COOH groups are accessible.
The adsorption behavior of metallo-supramolecular polyelectrolytes (MEPE) in the
pores of pure and CA-functionalized SBA-15 was studied for the case of Fe-MEPE. The
adsorption of Fe-MEPE acetate from aqueous solutions was studied under different con-
ditions of pH, ionic strength, temperature and chemical properties of the pore wall. The
uptake kinetics and the adsorption equilibrium in water was studied in detail. Electro-
static interactions play an important role in the interaction of Fe-MEPE with the silica
pore wall, because Fe-MEPE has positive charges along the backbone, whereas the walls
of the silica at ambient pH are negatively charged due to the weakly acidic surface silanol
groups. This charge effect can be boosted by using carboxylic-acid functionalized SBA-
15. The amount of adsorbed Fe-MEPE at the pore walls of CA-functionalized SBA-15
corresponds to a close-packed monolayer of Fe-MEPE chains disposed parallel to each
other at the pore walls. The increase of the amount adsorbed in CA-doped silica relative to
pure SBA-15 can be verified by using the monotopic Fe-Terpy complex formed by com-
plexation of Fe(II) ions with 2,2´:6´2´´-terpyridine. The equilibrium of Fe-Terpy uptake
in silica occurs in less than one hour. By contrast, the uptake of Fe-MEPE extends over
a long period of time and can be represented by a sum of two first order processes, a fast
initial process with relaxation time t1and a second much slower process with relaxation
time t2. The decrease of pH during the initial-fast uptake implies an ion exchange be-
169
tween the protons on the wall-carboxyl groups and Fe-MEPE. The slow process extends
over hundred of hours until the adsorption equilibrium is reached. The adsorption equi-
librium of Fe-MEPE can be described by a Langmuir isotherm with a large adsorption
constant KLindicating strong interactions between the carboxylic acid decorated silica
wall and Fe-MEPE. The kinetic study indicates that Fe-MEPE adsorption in the pores is a
diffusion-controlled process, where diffusion is controlled by the surface-layer resistance.
Presumably, this resistance is due to reduction of configurational freedom of the chains
on entering the pores. Solid-state NMR method has proved to be a useful method for
estimating the chain length of Fe-MEPE before and after the adsorption into the pores.
Comparative measurements showed that Fe-MEPE chains incorporated into the pores are
shorter than those in solution, indicating that the chains break into smaller entities in the
pores.
In conclusion, the present work has shown that the interactions between the polymer
template, functional silane and TEOS during the synthesis are crucial for the formation
of well-ordered functionalized SBA-15 silicas. However, a detailed examination of these
complex interactions requires more systematic studies in the future. For example, deter-
mining the relative hydrolysis and condensation rates of both functional silane and TEOS
in diluted aqueous systems will cast new light on the formation mechanism. The inves-
tigation of these entities by spectroscopic methods such as UV-Vis or IR spectroscopy
provides a challenge for future work.
Surface charge titrations of the pure SBA-15 and CA-SBA-15 at well-defined pH val-
ues may be very useful for resolving the question of the difference in surface charge
between pure SBA-15 and CA-SBA-15. In this context it would also be interesting to
study the adsorption of Fe-Terpy and Fe-MEPE in the stronger acidic PA-SBA-15 and
SA-SBA-15 samples that are expected to have higher surface charge. If the acids act as
oxidizing agents, this would destroy the complexes which in turn would be detectable by
the color change of the complex.
The time and concentration dependent precipitation of Fe-MEPE chains in solution as
well as during the adsorption in SBA-15 also calls for further investigation. Specially, it
should be clarified to what extent the slow uptake process of Fe-MEPE is affected by this
precipitation.
170 CHAPTER 11. SUMMARY AND OUTLOOK
Bibliography
[1] MUMPTON, F. A.: La roca magica : Uses of natural zeolites in agriculture and
industry. Proc. Natl. Acad. Sci. USA, 96:3463, 1999.
[2] MUELLER, U.; SCHUBERT, M.; TEICH F.; PUETTER H.; S.-ARNDT K.; PAS-
TRÉ J.: Metal-organic frameworks- prospective industrial applications. J. Mater.
Chem., 16:626, 2005.
[3] KRESGE, C. T.; LEONOWICZ; M. E.; ROTH, W. J.; VARTULI J. C.; BECK
J. S.: Ordered mesoporous molecular sieves synthesized by aliquid-cyristal tem-
plate mechanism. Nature, 359:710, 1992.
[4] ZHAO, D.; FENG, J.; HUO Q.; MELOSH N.; FREDRICKSON G. H.; CHMELKA
B. F.; STUCKY G. D.: Triblock copolymer syntheses of mesoporous silica with
periodic 50 to 300 Angstrom pores. Science, 279:1998, 1998.
[5] SHENDEROVICH, I. G.; BUNTKOWSKY, G.; SCHREIBER A.; GEDAT E.; SHARIF
S.; ALBRECHT J.; GOLUBEV N. S.; FINDENEGG G. H.; LIMBACH H.-H.:
Pyridine-15N- A mobile NMR sensor for surface acidity and surface defects of
mesoporous silica. J. Phys. Chem. B, 107:11924, 2003.
[6] CLARK, J. H.; MACQUARRIE, D. J.; TAVENER S. J.: The application of modified
mesoporous silicas in liquid phase catalysis. Dalton Trans., 4297, 2006.
[7] SMIRNOV, S. N.; GOLUBEV, N. S.; DENISOV G. S.; BENEDICT H.; S.-
MOHAMMEDI P.; LIMBACH H.-H.: Hydrogen/Deuterium isotope effects on the
NMR chemical shifts and geometries of intermolecular low-barrier hydrogen-
bonded complexes. J. Am. Chem. Soc., 118:4094, 1996.
[8] MELERO, J.A.; VAN GRIEKEN, R.; MORALES G.: Advances in the synthesis and
catalytic applications of organosulfonic-functionalized mesostructured materials.
Chem. Rew., 106:3790, 2006.
171
172 BIBLIOGRAPHY
[9] SCOTT, B. J.; WIRNSBERGER, G.; STUCKY G. D.: Mesoporous and mesostruc-
tured materials for optical Applications. Chem. Mater., 13:3140, 2001.
[10] YIU, H. P.; WRIGHT, P. A.; BOTTING N. P.: Enzyme immobilisation using SBA-
15 mesoporous molecular sieves with functionalised surfaces. J. Mol. Catal. B,
15:81, 2001.
[11] MOLLER, K.; BEIN, T.: Inclusion chemistry in periodic mesoporous hosts. Chem.
Mater., 10:2950, 1998.
[12] SCHÜTTE, M.; KURTH, D. K.; LINFORDT M. R.; CÖLFEN H.; MÖHWALD H.:
Herstellung und Charakterisieung von metallosupramolekularen dünnen Polyelek-
trolytfilmen. Angew. Chem., 110:3058, 1998.
[13] BRAY, D.: Cell movements: from molecules to motility. Garland Publishing, New
York, 2001.
[14] KURTH, D. G.; HIGUCHI, M.: Transition metal ions: Weak links for strong poly-
mers. Soft Matter, 2:915, 2006.
[15] GADE, L. H.: Koordinationschemie. Wiley-VCH, Weinheim, 1998.
[16] WELZ, B.; SPERLING, M.: Atomabsorptionsspektrometrie. Wiley-VCH, Wein-
heim, 1999.
[17] ZELLMER, D. L.: http://zimmer.csufresno.edu/ davidz/Chem106/StdAddn/StdAddn.html.
website, 1998. California State University, Fresno.
[18] BRIGGS, D.; GRANT, J. T.: Surface Analysis by Auger and X-ray Photoelectron
Spectroscopy. IM Publications, Chichester, 2003.
[19] BABER, M.; CONNOR, J. A.; GUEST M. F.; HILLIER I. H.; SCHWARZ-M.;
STACEY M.: Bonding in some donor acceptor complexes involving boron trifluo-
ride. Study by means of ESCA and molecular orbital calculations. J. Chem. Soc.
Faraday Trans. II, 69:551, 1973.
[20] SEAH, M. P.; DENCH, W. A.: Quantitative electron spectroscopy of surfaces: A
standard data base for electron inelastic mean free paths in solids. Surf. Interface
Anal., 1:2, 1979.
[21] HORNAK, J. P.: Basics of NMR. website. http://www.cis.rit.edu/htbooks/nmr/.
BIBLIOGRAPHY 173
[22] SCHURKO, R.: http://mutuslab.cs.uwindsor.ca/schurko/ssnmr/ssnmr_schurko.pdf.
[23] SOLOVYOV, L. A.; KIRIK, S. D.; SHMAKOV A. N.; ROMANNIKOV V. N.: X-
ray structural modeling of silicate mesoporous mesophase material. Microporous
Mesoporous Mater., 44:17, 2001.
[24] SOLOVYOV, L. A.; MOLOKEEV, M. S.; VASILEV A. D.: Powder diffraction
crystal structure analysis using derivative difference minimization: example of the
potassium salt of 1-(tetrazol-5-yl)-2-nitroguanidine. Acta Crystallogr. B, 61:435,
2005.
[25] IMPÈROR-CLERC, M.; DAVIDSON, P.; DAVIDSON A.: Existence of a microp-
orous corona around the mesopores of silica-based SBA-15 materials templated by
triblock copolymers. J. Am. Chem. Soc., 122,:11925, 2000.
[26] ZICKLER, G.A.; JÄHNERT, S.; WAGERMAIER W.; FUNARI S.S.; FINDENEGG
G.H.; PARIS O.: Physisorbed films in periodic mesoporous silica studied by in
situ synchrotron small-angle diffraction. Phys. Rev. B, 73:184109, 2006.
[27] SING, K. S. W.: Characterization of porous materials: past, present and future.
Colloids and surfaces A, 241:3, 2004.
[28] GREGG, S. J.; SING, K. S. W.: Adsorption, Surface Area and Porosity. Academic
Press, London, 1982.
[29] GALARNEAU, A.; CAMBON, H.; DIRENZO F.; FAJULA F.: True microporosity
and surface area of mesoporous SBA-15 silicas as a function of synthesis tempera-
trure. Langmuir, 17:8328, 2001.
[30] MIYAZAWA, K.; INAGAKI, S.: Control of the microporosity within the pore walls
of ordered mesoporous silica SBA-15. Chem. Commun., 2121, 2000.
[31] HARKINS, W. D.; JURA, G.: Extension of the attractive energy of a solid into
an adjacent liquid film, the decrease of energy with distance, and the thickness of
films. J. Am. Chem. Soc., 66:919, 1944.
[32] DOLLIMORE, D.; HEAL, G. R.: An improved method for the calculation of pore
size distribution from adsorption data. J. Appl. Chem., 14:109, 1964.
174 BIBLIOGRAPHY
[33] AKCAKAYIRAN, D.: Funktionalisierung der Porenwaende und Einlagerung von
Metall-Koordinations-Polyelektrolyten in SBA-15 Silika. Diplom Arbeit, TU-
Berlin, 2004.
[34] KRUK, M.; JARONIEC, M.: Application of large pore MCM-41 molecular sieves
to improve pore size analysis using nitrogen adsorption measurements. Langmuir,
13:6267, 1997.
[35] JARONIEC, M; SOLOVYOV, L. A.: Improvement of the Kruk-Jaroniec-Sayari
method for pore size analysis of ordered silicas with cylindrical mesopores. Lang-
muir, 22:6757, 2006.
[36] EHRENSTEIN, G. W.; RIEDEL, G.; TRAWIEL P.: Thermische Analyse von Kunst-
stoffen. Carl Hanser Verlag, München, 2003.
[37] LI, Y., ZHANG W.; ZHANG L.; YANG Q.; WEI Z.; FENG Z.; LIC.: Direct
sytnhesis of Al-SBA-15 mesoporous materials via hydrolysis-controlled approach.
J. Phys. Chem. B, 108:9739, 2004.
[38] KIM, J. M.; HAN, Y.-J.; CHMELKA B. F.; STUCKY G. D.: One-step synthesis
of ordered mesocomposites with non-ionic amphiphilic block copolymers: impli-
cations of isoelectric point, hydrolysis rate and fluoride. Chem. Commun., 2437,
2000.
[39] YU, C.; TIAN, B.; FAN-J.; STUCKY G. D.; ZHAO D.: Salt effect in the synthesis
of mesoporous silica templated by non-ionic block copolymers. Chem. Commun.,
2726, 2001.
[40] WEI, Y. L.; WANG, Y. M.; ZHU J. H.; WUZ. Y.: In-situ coating of SBA-15 with
MgO: Direct synthesis of mesoporous solid bases from strong acidic systems. Adv.
Mater., 15(22):1943, 2003.
[41] LI, Y.; FENG, Z.; LIAN Y.; SUN K.; ZHANG L.; JIA G.; YANG Q.; LIC.:
Direct synthesis of highly ordered Fe-SBA-15 mesoporous materials under weak
acidic conditions. Micropor. Mesopor. Mater., 84:41, 2005.
[42] ZHANG, W.; GLOMSKI, B.; PAULY T. R.; PINNAVAIA T. J.: A new nonionic sur-
factant pathway to mesoporus molecular sieve silicas with long range framework
order. Chem. Commun., 1803, 1999.
BIBLIOGRAPHY 175
[43] BRUNEL, D.; CAUVEL, A.; DIRENZO F.; FAJULA F.; FUBINI B.; ONIDA
B.; GARRONE E.: Preferential grafting of alkoxysilane coupling agents on the
hydrophobic portion of the surface of micelle-templated silica. New. J. Chem.,
24:807, 2000.
[44] BURKETT, S.L.; SIMS, S. D.; MANN S.: Synthesis of hybrid iorganic-organic
mesoporous silica by co condensation of siloxane and organosiloxane precursors.
Chem. Commun., 1367, 1996.
[45] FIORILLI, S.; ONIDA, B.; BONELLI B.; GARRONE E.: In situ infrard study of
SBA-15 functionalized with carboxylic groups incorporated by a co-condensation
route. J. Phys. Chem. B, 109:16725, 2005.
[46] YOKOI, T.; YOSHITAKE, H.; TATSUMI T.: Synthesis of amino-functionalized
MCM-41 via direct co-condensation and post-synthesis grafting methods using
mono-, di- and tri-amino-organoalkoxysilanes. J. Mater. Chem., 14:951, 2004.
[47] MACQUARRIE, D. J.: Direct preparation of organically modified MCM-type ma-
terials. Preparation and characterization of aminopropyl-MCM and 2 cynoethyl-
MCM. Chem. Commun., 1996.
[48] HALL, S. R.; FOWLER, C. E.; LEBEAU B.; MANN S.: Template-directed synthe-
sis of bi-functionalized organo-MCM-41 and phenyl-MCM-48 silica mesophases.
Chem. Commun., 201, 1999.
[49] MERCIER, L.; PINNAVAIA, J.: Direct synthesis of hybrid organic-inorganic
nanoporous silica by a neutra amine assembly route: Structure-function control by
stoichiometric incorparation of organosiloxane molecules. Chem.Mater., 12:188,
2000.
[50] HUH, S.; WIENCH, J. W.; YOO J.-C.; PRUSKI M.; LIN V. S.-Y.: Organic func-
tionalization and morphologie control of mesoporous silicas via a co-condensation
synthesis method. Chem. Mater., 15:4247, 2003.
[51] FOWLER, C: E.; BURKETT, S. L.; MANN-S.: Synthesis and characterization
of ordered organo-silica-surfactant mesophases wih functionalized MCM-41-type
architecture. Chem. Commun., 1769, 1997.
176 BIBLIOGRAPHY
[52] LIM, M. H.; BLANFORD, C. F.; STEIN-A.: Synthesis of ordered microporous
silicates with organosulfur surface groups and their aplication as solid catalyst.
Chem. Mater., 10:467, 1998.
[53] MARGOLESE, D.; MELERO, J. A.; CHRISTIANSEN S. C.; CHEMELKA B. F.;
STUCKY G. D.: Direct synthesis of ordered SBA-15 mesoporous silica containing
sulfonic acid groups. Chem. Mater., 12:2448, 2000.
[54] LIM, M. H.; BLANFORD, C. F.; STEIN A.: Comparative Studies of grafting and
direct syntheses of inorganic-organic hybrid mesoporous materilas. Chem. Mater.,
11:3285, 1999.
[55] WANG, Y. Q.; YANG, C. M.; ZIBROWIUS B.; SPLIETHOFF B.; LINDÉN M.;
SCHÜTH F.: Directing the formation of Vinyl-functionalized silica to the hexagonal
SBA-15 or large-pore Ia3d structure. Chem. Mater., 15:5029, 2003.
[56] CHONG, A. S. M.; ZHAO, X. S.: Functionalization of SBA-15 with APTES and
chracterization of functional materials. J. Phys. Chem. B, 107:12650, 2003.
[57] RICHER, R.; MERCIER, L.: Direct synthesis of functionalized mesoporous silica
by non-ionic alkylpolyetyleneoxide surfactant assembly. Chem. Commun., 1775,
1998.
[58] CHONG, A.S.M.; ZHAO, X.S.: Functionalized nanoporous silicas for the immo-
bilization of penicillin acylase. Appl. Surf. Sci., 237:398, 2004.
[59] YIU, H. P.; BOTTING, C. H.; BOTTING N. P.; WRIGHT P. A.: Size selective
protein adsorption on thiol-functionalised SBA-15 mesoporous molecular sieve.
PCCP, 3:2983, 2001.
[60] VAN RHIJN, W. M.; DEVOS, D. E.; SELS B. F.; BOSSAERT W. D.; JACOBS
P. A.: Sulfonic acid functionalised ordered mesoporous materials as catalysts for
condensation and esterification reactions. Chem. Commun., 317, 1998.
[61] BOSSAERT, W. D.; DEVOS, D. E.; VAN RHJIN W. M.; BULLEN J.; GROBET P.
J.; JACOBS P. A.: Mesoporous sulfonic acids as selective heterogeneous catalysts
for the synthesis of monoglycerides. J. Catal., 182:156, 1999.
[62] MELERO, J. A.; STUCKY, G. D.; VAN GRIEKEN R.; MORALES G.: Direct
synthesis of ordered SBA-15 mesoporous materials containing arenesulfoic acid
groups. J. Mater. Chem., 12:1664, 2002.
BIBLIOGRAPHY 177
[63] MELERO, J.; VAN GRIEKEN, R.; MORALES G.; NUNO V.: Friedel Crafts acyla-
tion of aromatic compounds over arenesulfonic containing mesostructureed SBA-
15 materials. Catal. Commun., 5:131, 2004.
[64] GRIEKEN, R.; MELERO, J. A.; MORALES G. VAN:Fries rearrangement of
phenyl acetate over sulfonic modified mesostructured SBA-15 materials. Appl.
Catal., A, 289:143, 2005.
[65] WANG, X.; CHEN, C. C.; CHEN S. H.; MOU Y.; CHENG S.: Arenesulfonic acid
functionalized mesoporous silica as a novel acid catalyst for the liquid phase Beck-
mann rearrangement of cyclohexanone oxime to var epsilon-caprolactam. Appl.
Catal. A, 281:47, 2005.
[66] GRIEKEN, R.; MELERO, J. A.; MORALES G. VAN:Etherification of benzyl al-
cohols with 1-hexanol over organosulfonic acid mesostructured materials. J. Mol.
Catal. A: Chem., 256:29, 2006.
[67] KAWI, S.; SHEN, S. C.; CHEW P. L.: Generation of Bronsted acid sites on Si-
MCM-41 by grafting of phosphorus species. J. Mater. Chem., 12:1582, 2002.
[68] LE, Y.; YANG, X.; DAI W. L.; GAO R.; FAN K.: Unexpected mononuclear
W(VI) complexes containing phosphonate ligands anchored on mesoporous silica.
Another strategy for immobilization. Catal. Commun., 9:1838, 2008.
[69] GRUEN, M.; UNGER, K. K.; MATSUMOTO A.; TSUTSUMI K.: Novel pathways
for the preparation of mesoporous MCM-41 materials: control of porosity and
morphology. Microporous Mesoporous Mater., 27:207, 1999.
[70] SCHREIBER, A.: Phasenverhalten reiner Stoffe in mesoporösen Silika-
Materialien. PhD Thesis, TU-Berlin, 2002.
[71] YANG, C. M.; ZIBROWIUS, B.; SCHMIDT W.; SCHÜTH F.: Formation of
cyanide-functionalized SBA-15 and its transformation to carboxylate functional-
ized SBA-15. Phys. Chem. Chem. Phys., 6:2461, 2004.
[72] ZHANG, C.; ZHOU, W.; LIU S.: Synthesis and characterization of organofunc-
tionalized MCM-41 by the original stepped templated sol-gel technology. J. Phys.
Chem., 109:24319, 2005.
178 BIBLIOGRAPHY
[73] KUMAR, R.; CHEN, H. T.; ESCOTO J. L. V.; LIN V. S. Y.; PRUSKI M.: Template
Removal and Thermal Stability of Organically Functionalized Mesoporous Silica
Nanoparticles. Chem. Mater., 18:4319, 2006.
[74] JÄHNERT, S.: Untersuchung von reinen Stoffen in periodisch mesoporösen Silika-
materialien mit thermophysikalischen Methoden und in-situ Röntgenkleinwinkel-
streuung. PhD Thesis, TU-Berlin, 2008.
[75] KLEITZ, F.; CHOI, S. H.; RYOO R.: Cubic Ia3d large mesoporous silica: syn-
thesis and replication to platinium nanowires, carbon nanoroads and carbon nan-
otubes. Chem. Commun., 2136, 2003.
[76] CAI, Q.; LUO, Z.-S.; PANG W.-Q.; FAN Y.-W.; CHEN X.-H.; CUI F.-Z.: Dilute
solution routes to various controllable morphologies of MCM-41 silica with a basic
medium. Chem. Mater., 13:258, 2001.
[77] SAJANIKUMARI, S.; KHUSHALANI, D.; MANN S.: Synthesis and shape modifi-
cation of organo-functionalized silica nanoparticles with ordered mesostructured
interiors. Chem. Mater., 13:1023, 2003.
[78] CORRIU, R. J. P.; DATAS, L.; GUARI-Y.; MEHDI A.; REYÉA C.; THIEULEUXA-
C.: Ordered SBA-15 mesoporous silica containing phosphonic acid groups pre-
pared by a direct synthetic approach. Chem. Commun., 763, 2001.
[79] MACIEL, G. E.; HAW, J. F.; CHUANG-I-S.; HAWKINS B. L.; EARLY-T. A.;
MCKAY-D. R.; PETRAKIS L.: NMR studies of pyridine on silica-alumina. J. Am.
Chem. Soc., 105:5529, 1983.
[80] RIPMEESTER, J. A.: Surface acid site characterization by means of CP/MAS
nitrogen-15 NMR. J. Am. Chem. Soc., 105:2925, 1983.
[81] LORENTE, P.; SCHENDEROVICH I. G.; GOLUBEV, N. S.; DENISOV G. S.; BUN-
TKOWSKY G.; LIMBACH H.-H.: 1H/15N NMR chemical shift shielding, dipolar
15N,2H coupling and hydrogen bond geometry correlations in a novel series of
hydrogen-bonded acid-base complexes of collidine with carboxylic acids. Magn.
Reson. Chem., 39:18, 2001.
[82] MAUDER, D.: PhD Thesis in preparation. 2009.
BIBLIOGRAPHY 179
[83] GUCHT, J.; VAN:Equilibrium polymers in solution and at interfaces. phD Thesis,
Wageningen University, Holland, 2004.
[84] C., CONSTABLE E.: The coordination chemistry of 2,2’:6’,2”-terpyridine and
higher oligopyridines. Adv. Inorg. Chem. Radiochem., 30:69, 1987.
[85] CONSTABLE, E. C.; THOMPSON ALEXANDER, M. W. C.: Multinucleating
2,2´:6´, 2´´-terpyridine ligands as building blocks for the assembly of co-ordination
polymers and oligomers. J. Chem. Soc. Dalton Trans., 3467, 1992.
[86] BODENTHIN, Y.; PIETSCH, U.; MÖHWALD H.; KURTH D. G.: Inducing
spin crossover in metallo-supramolecular polyelectrolytes through an amphiphilic
phase transition. J. Am. Chem. Soc., 125:3110, 2005.
[87] SCHWARZ, G.: Charakterisierung von Eigenschaften und Struktur Metallo-
supramolekularer Funktionseinheiten. PhD Thesis, Potsdam Universitaet, Pots-
dam, 2008.
[88] HOLYER, R. H.; HUBBARD, C. D.; KETTLE S. F. A.; WILKINS R. G.: The ki-
netics of replacement Reactions of complexes of the transition metals with 2,2’,2”-
Terpyridine. Inorg. Chem., 5:622, 1966.
[89] MOREL, F. M. M.; HERING, J. G.: Principles and applications of aquatic chem-
istry. John Wiley, New York, 1993.
[90] SIEVERS, T.: Multischichten aus vernetzten Koordinationspolyelektrolyten.
Diplom Arbeit, MPI-Golm and FU-Berlin, 2005.
[91] M., MILLER. J.: Chromatography: Concepts and Contrast. John Wiley and Sons,
Inc., USA, New Jersey, 2nd , 2005.
[92] FALBE, J.; REGITZ M.: Roempp-Chemie-Lexikon, Band 5. Thieme, Stutgart,
1992.
[93] HELFFERICH, F. G.: Ion Exchange. Dover Publications, New York, 1995.
[94] INGLEZAKIS. V. J.; POULOPOULOS, S. G.: Adsorption, ion exchange and catal-
ysis: Design of operations and environmental applications. Elsevier, Amsterdam,
2006.
180 BIBLIOGRAPHY
[95] DITTMEYER, R.; EMIG, G.: Simultaneous heat and mass transfer and chemical
reaction., 8, 1727. Wiley-VCH, Weinheim, G. Ertl. et al. , 2008.
[96] CRANK, J.: The mathematics of diffusion. Clarendon Press, Oxford, second , 1975.
[97] KRÖHNKE, F.: The specific synthesis of pyridines and oligopyridines. Synthesis,
1, 1976.
[98] VADUVESCU, S.; POTVIN, P. G.: Synthesis and Characterization of a Novel
Linear Trinuclear Ruthenium(II) Complex: Variation of Photosensitization Abil-
ity with Chain Length in a Homologous Series. Inorg. Chem., 41:4081, 2002.
[99] VADUVESCU S.; POTVIN, P.G.: Linear Multinuclear Ru(II) Photosensitizers. Eur.
J. Inorg. Chem, 2004:1763, 2004.
[100] MARGAN, G., BURSTALL F.H.: 347. Researches on residual affinity and
co-ordination. Part XXXVII. Complex metallic salts containing 2 : 6-di-2-
pyridylpyridine (2 : 2 : 2-tripyridyl). J. Chem. Soc., 1649, 1937.
[101] BRIGGS, D.; SEAH, M. P.: Practical Surface Analysis. Wiley, Chichester, 1990.
[102] KIM, D. J.; DUNN, B. C.; HUGGINS F.; HUFFMAN G. P.; KANG M.; YIE J. E.;
EYRING E. M.: SBA-15-supported iron catalysts for Fische-Tropsch production of
diesel fuel. Energy and fuels, 20:2608, 2006.
[103] ERNST, S.; GLASER, R.; SELLE M.: Pogress in zeolite and microporous mate-
rials. H. CHON, S.-K. IHM, Y.S. UH(EDS) (): Stud. Surf. Sci. Catal., 105B,
1021, Amsterdam, 1997. Elsevier.
[104] HONMA, I.; SASABE, H.; ZHOU H. S.: Synthesis of self-assembled functional
molecules in mesoporous materials. KOMARNENI, S.; PARKER, J. C.; WOLLEN-
BERGER H. J. (EDS) (): Symposium V: Nanophase and Nanocomposite Materials
II, vol. 457, 525, USA, 1997. Mater. Res. Soc. Symp. Proc., MRS.
[105] ZHU, J.; KÓNYA, Z.; PUNTES V. F.; KIRICSI E.; MIAO C. X.; AGER J. W.;
ALIVISATOS A. P.; SOMORJAI G. A.: Encapsulation of Metal (Au, Ag, Pt)
Nanoparticles into the Mesoporous SBA-15 Structure. Langmuir, 19:4396, 2003.
[106] HOLLAND B. T.; WALKUP C.; STEIN, A.: Encapsulation, stabilization, and cat-
alytic properties of flexible metal porphyrin compolexes in MCM-41 with minimal
electronic perturbation by environment. J. Phys. Chem. B, 102:4301, 1998.
BIBLIOGRAPHY 181
[107] GANSCHOW, M.; WARK, M.; WÖHRLE D.; SCHULZ-EKLOFF G. Angew. Chem.,
112:167, 2000.
[108] HESS, C.; HOEFELMEYER, J. D.; TILLEY T. D.: Spectroscopic characterization
of highly dispersed vanadia supported on SBA-15. J. Phys. Chem. B, 108:9703,
2004.
[109] WANG, Y. M.; WU, Z. Y.; WANG H. J.; ZHU J. H.: Fabrication of metal oxides
occluded in mesoporous hosts via a solid-state grinding route: The influence of
host-guest interacions. Adv. .Func. Mater., 16:2374, 2006.
[110] AKCAKAYIRAN, D.; KURTH, D. G.; RÖHRS S.; RUPPRECHTER G.; FIND-
ENEGG G. H.: Self-assembly of a metallosupramolecular coordination polyelec-
trolyte in the pores of SBA-15 and MCM-41. Langmuir, 21:7501, 2005.
[111] ERNST, S.; SELLE, M. Micropor. Mesopor. Mater, 27:355, 1999.
[112] ZANJANCHI M. A.; EBRAHIMIAN, A.; ALIMOHAMMADI Z: A spectroscopich
study on the adsorption of cationic dyes into mesopoous AlMCM-41 materials.
Optical Materials, 29:794, 2007.
[113] LI, D. M.; ZHAO, W. J.; SUN X. D.; ZHANG J. L.; ANPO M.; ZHAO J.C.:
Photophysical properties of coumarin derivates incorporated in MCM-41. Dyes
and pigments, 68:33, 2006.
[114] YAO, Y. F.; ZHANG, M. S.; SHI J. X.; GONG M. L.; ZHANG H. J.; YANG
Y. S.: Encapsulation of fluorescein into MCM-41 mesoporous molecular sieve by
a sol-gel method. Materials Letters, 48:44, 2001.
[115] REN, T. Z.; YUAN, Z. Y.; SUB. Y.: Encapsulation of direct blue dye into meso-
poruos silca based materials. Colloids and Surfaces A, 300:79, 2007.
[116] S.-EKLOFF, G; WÖHRLE, D.; VAN DUFFEL B.; SCHOONHEYDT R. A.: Chro-
mophores in porus silicas and minerals: preparation and optical properties. Mi-
cropor. Mesopor. Mater., 51:91, 2002.
[117] REN, T. Z.; YUAN, Z. Y.; SUB. Y.: Direct blue dye encapsulatated mesostruc-
tured MCM-41 composites: Microwave asisted preparation and characterization.
Colloids and Surfaces A, 300:88, 2007.
182 BIBLIOGRAPHY
[118] ASEFA, T.; MACLACHAN, M. J.; COOMBS N.; OZIN G. A.: Periodic meso-
porous organosilicas with organic groups inside the hannel walls. Nature, 402:867,
1999.
[119] KURTH, D. G.; LÓPEZ, J. P.; DONG W. F.: A new Co(II)-metalloviologen-based
electrochromic material integrated in thin multilayer films. Chem. Commun.,
2119, 2005.
[120] HAN, F. S.; HIGUCHI, M.; KURTH D. G.: Metallosupramolecular polyelec-
trolytes self-assembled from various pyridine ring-substituted bisterpyridines and
metal Ions: photophysical, electrochemical, and electrochromic properties. J. Am.
Chem. Soc., 130:2073, 2008.
[121] HAN, F. S.; HIGUCHI, M.; KURTH D. G.: Metallo-supramolecular polymers
based on functionalized bis-terpyridines as novel electrochromic materials. Adv.
Mater., 19:3928, 2007.
[122] BODENTHIN, Y.; SCHWARZ, G.; GUTBERLET T; GEUE T.; STAHN J.; MÖH-
WALD H.; KURTH D.G.; PIETSCH U.: Molecular magnetism in thin metallo-
supramolecular films: A combined neutron and soft x-ray reflectometry study. Su-
perlattices and Microstructures, 41:138, 2007.
[123] VERMONDEN, T; VAN DER GUCHT J.; DE WAARD, P.; MARCELIS A. T. M.;
BESSELING N. A. M.; SUDHÖTER E. J. R.; FLEER G. J.; COHEN STUART
M. A.: Water-soluble reversible coordination polymers: chains and rings. Macro-
molecules, 36:7035, 2003.
[124] YANG, C. M.; ZIBROWIUS, B.; SCHMIDT W.; SCHÜTH F.: A novel synthetic
route for negatively charged ordered mesoporous silica SBA-15. Chem. Commun.,
1772, 2003.
[125] MAROIE, S.; SAVY, M.; VERBIST J.: ESCA and EPR studies of monomer, dimer,
and polymer iron phthalocyanines: involvements for the electrocatalysis of molec-
ular oxygen reduction. Inorg. Chem., 18:2560, 1979.
[126] JOSEPH, Y.; KETTELER, G.; KUHRS C.; RANKE W.; WEISS W.; SCHLÖGL R.:
On the preparation and composition of potassium promoted iron oxide model cat-
alyst films. Phys. Chem. Chem. Phys., 3:4141, 2001.
BIBLIOGRAPHY 183
[127] BRUZZONITI, M. C.; PRELLE, A.; SARZANINI C.; ONIDA B.; FIORILLI
S.; GARRONE:Retention of havy metal ions on SBA-15 mesoporous silica func-
tionalized with carboxylic groups. J. Sep. Sci., 30:2414, 2007.
[128] ILER, R. K.: The chemistry of Silica. Wiley, New York, 1978.
[129] KATIYAR, A.; JI, L.; SMIRNIOTIS P.; PINTO N. G.: Protein adsoption on the
mesoporous molecular sieve silicate SBA-15: effects of pH and pore size. Journal
of chromatography A, 1069:119, 2005.
[130] HOFFMANN, T.; WALLACHER, D.; HUBER P.; BIRRINGER R.; KNORR K.;
SCHREIBER A.; FINDENEGG. G.H.: Small-angle x-ray diffraction of Kr in meso-
porous silica: Effects of microporosity and surface roughness. Phys. Rev. B,
72:064122, 2005.
[131] KORTUNOV, P.; CHMELIK, C.; KÄRGER J.; RAKOCZY R. A.; RUTHVEN D. M.;
TRAA Y.; VASENKOV S.; WEITKAMP J.: Sorption kinetics and intracrystalline
diffusion of methanol in ferrite: An example of disguised kinetics. Adsorption,
11:235, 2005.
[132] BUTT, H. J.; GRAF, K.; KAPPL M.: Physics and Chemistry of Interfaces. Wiley-
VCH Verlag and Co. KGaA, Weinheim, 2003.
184 BIBLIOGRAPHY
Appendix
UV-Spectrum of Fe-Terpy
Figure A 1: UV-vis spectrum of Fe-Terpy.
The MLCT-band in the UV-vis spectrum at 552 nm is responsible for the pink-red colour
of the Fe-Terpy complex.
300 400 500 600 700
0.0
0.4
0.8
1.2
1.6
absorbance
λ/nm
MLCT
Fe-Terpy
d→d
π→π*
Wide Angle X-Ray diffractions of Acid-Functionalized Silicas
Figure A 1: XRD spectrum of 20-SA-00. The sharp reflex at 2θ=28° comes from the
silicon in the sample carrier. The broad reflex at approx. 2θ=22° is assigned to the
amorphous structure of the 20-SA-00 sample.
Figure A 2: XRD spectrum of 20-SA-00 heated to 1000°C with a heat rate of 10Kmin-1.
The broad reflex at approx. 2θ=22° is assigned to the amorphous structure of the 20-SA-
00 sample.
20-SA-00
20-SA-00-1000°C
STOE Powder Diffraction System 19-Sep-08
2Theta10.0 20.0 30.0 40.0 50.0 60.0 70.0
0
200
400
600
800
1000
1200
1400
Absolute Intensity
C:\stoe-winxpow\DATA\Dilek\3A-sto.raw / (Range 1)
Figure A 3: XRD spectrum of 15-POEt-40-T. The sharp reflex at 2θ=28° comes from
the silicon in the sample carrier. The broad reflex at approx. 2θ=22° is assigned to the
amorphous structure of the 15-POEt-40-T sample.
15-POEt-40-T
STOE Powder Diffraction System 19-Sep-08
2Theta10.0 20.0 30.0 40.0 50.0 60.0 70.0
0
4000
8000
12000
16000
20000
24000
Absolute Intensity
C:\stoe-winxpow\DATA\Dilek\3L-sto.raw / (Range 1)
Figure A 4: XRD spectrum of 15-POEt-40-T heated to 1000°C with a heat rate of
10Kmin-1. In this case some sharp reflexes appear in the spectrum coming from a
crystalline structure.
STOE Powder Diffraction System 19-Sep-08
2Theta10.0 20.0 30.0 40.0 50.0 60.0 70.0
0.0
20.0
40.0
60.0
80.0
100.0
Relative Intensity (%)
C:\stoe-winxpow\DATA\Dilek\3L-sto.rmb / (Range 1)
[71-785] Si O2 / Silicon Oxide / Cristobalite low (Range 1)
Figure A 5: The background subtracted spectrum in Fig. A 4. The reflexes can be
associated with cristobalite which is a high temperature modification of SiO2.
15-POEt-40-T-1000°C
15-POEt-40-T-1000°C
29Si Solid-State NMR Results
Carboxylic-Acid Functionalized SBA-15
sample new names T2 T
3 Q
2 Q
3 Q
4 (T/Q)exp (T/Q)NMR Y x
90-60-20-CN(1) 20-CN-60-F(1) 21.2 4.99 36.6 100 0.25 0.19 77% 42%
90-60-20-CA(1)-2 20-CA-60-F(1)-2 29.4 33.6 100 0.25 0.22 88% 50%
90-60-20-CA(1)-3 20-CA-60-F(1)-3 4.7 16.8 30.3 100 0.25 0.17 66% 42%
90-60-20-CA(1)-4 20-CA-60-F(1)-4 14.78 14.37 40.7 100 0.25 0.21 83% 42%
100-60-10-CA(1)-
2cal 10-CA-60-F(1)-2 10.6 1.1 40.6 100 0.11 0.08 75% 20%
Table A 1: Expected (T/Q)exp and experimental (T/Q)NMR ratio of organosiloxane (T) and
siloxane (Q) groups; Y: Reaction yield; x: Degree of surface functionalization. All
samples are discussed in the main text.
Phosphonic-Acid Functionalized SBA-15
sample new names T2 T
3 Q
2 Q
3 Q
4 (T/Q)exp (T/Q)NMR Y x
90-60-10-PEt 10-P-60-F 0.57 10.1 54.1 100 0.11 0.07 63% 16%
90-60-10-PA(1)-1 10-POEt-60-F(1)-1 9.4 2.2 42.8 100 0.11 0.065 59% 17%
90-60-10-PA(1)-1-
H2SO4-24h 10-PA-60-F(1)-1
90-60-10-PA(1)-2 10-POEt-60-F(1)-2 4.28 7.08 0.96 39.6 100 0.11 0.03 73% 22%
90-60-20-PEt(1) 20-P-60-F(1) 25.1 16.7 68.3 100 0.25 0.25 99% 38%
90-60-20-PA(1)-1 20-POEt-60-F(1)-1 5.1 46.2 100 0.25 0.035 14% 10%
90-60-20-PEt(2) 20-P-60-F(2) 19 23.1 73.8 100 0.25 0.24 97% 36%
90-60-20-PA(2)-1 20-POEt-60-F(2)-1 26.6 60.5 100 0.25 0.17 66% 31%
100-95-15-Pet 15-P-95-F(1) 7.3 14 41.2 100 0.18 0.15 85% 34%
100-95-15-PA 15-POEt-95-F(1)-1 6.5 13.3 33.6 100 0.18 0.15 84% 37%
100-95-15-PA-
EtOHsox-cal250 15-POEt-95-F(1)-2 15.3 6.3 90.8 100 0.18 0.11 64% 19%
100-95-15-PA-
EtOHkolb 15-POEt-95-F(1)-3 12.1 14.4 65.1 100 0.18 0.16 91% 29%
100-00-15-PEt 15-P-00 disordered!
100-(-40)-15-PA 15-POEt-40-T(1) 12.7 12.6 3.82 56.90 100 0.18 0.19 89% 29%
100-(-40)-15-PA-
HCl-24h 15-PA-40-T(1) 6.1 7.5 22 100 0.18 0.11 73% 38%
Table A 2: Expected (T/Q)exp and experimental (T/Q)NMR ratio of organosiloxane (T) and
siloxane (Q) groups; Y: Reaction yield; x: Degree of surface functionalization. Marked
samples are discussed in the main text.
Sulfonic-Acid Functionalized SBA-15
sample new names T2 T
3 Q
2 Q
3 Q
4 (T/Q)exp (T/Q)NMR Y x
90-60-20-SA 20-SA-60-F 4.6 8.7 43 100 0.25 0.093 37% 24%
90-1480-30-S 30-S-1480-F 20.4 42.3 100 0.43 0.143 33% 33%
90-1480-30-SA 30-SA-1480-F 14 7.4 35.6 100 0.43 0.158 37% 38%
100-00-20-SA 20-SA-00 5.87 8.73 0.71 31.72 100 0.25 0.116 46% 31%
100-(-55)-20-SA 20-SA-55-T 8.75 0.29 44.19 100 0.25 0.063 25% 16%
Table A 3: Expected (T/Q)exp and experimental (T/Q)NMR ratio of organosiloxane (T) and
siloxane (Q) groups; Y: Reaction yield; x: Degree of surface functionalization. All
samples are discussed in the main text.
TGA Results
sample new names w % F TGA w % SiO2 TGA nTGA n
exp Y
90-60-20-CA(1)-2 20-CA-60-F(1)-2 16.43 78.8 2.86 3.32 86%
90-60-20-CA(1)-3 20-CA-60-F(1)-3 15 83 2.48 3.32 75%
90-60-20-CA(1)-4 20-CA-60-F(1)-4 15.9 78.1 2.79 3.32 84%
100-60-10-CA(1)-1 10-CA-60-F(1)-1 8.4 86.5 1.33 1.66 80%
100-60-10-CA(1)-2 10-CA-60-F(1)-2 9.6 88.4 1.49 1.66 90%
100-60-10-CA(1)-2cal 10-CA-60-F(1)-2cal 8.7 86.3 1.38 1.66 83%
90-60-20-SA(2)-1 20-SA-60-F(1)-1a 10 85.7 0.95 3.32 29%
90-60-20-SA(2)-1a 20-SA-60-F(1)-1b 11 83.4 1.07 3.32 32%
90-60-20-SA(2)-1b 20-SA-60-F(1)-1c 8 80.3 0.81 3.32 24%
Average: 20-SA-60-F(1)-1 9.7 83.1 0.95 3.32 28%
100-(-55)-20-SA 20-SA-55-T 4.64 89.3 0.42 3.32 13%
100-00-20-SA 20-SA-00 10.3 75.8 1.10 3.32 33%
90-1480-30-SA 30-SA-1480-F 10.93 75 1.18 4.98 24%
90-60-10-POEt(1)a 10-POEt-60-F(1)-a 6 88.6 0.63 1.66 38%
90-60-10-POEt(1)b 10-POEt-60-F(1)-b 5.6 90 0.58 1.66 35%
90-60-10-PA(1)-H2SO4 10-PA-60-F 2.97 92.6 0.30 1.66 18%
90-60-20-POEt(1)-1 20-POEt-60-F(1)-1 7.2 89.6 0.74 3.32 22%
90-60-20-POEt(2)-1 20-POEt-60-F(2)-1 12.2 81 1.39 3.32 42%
100-95-15-POEt 15-POEt-95-F(1)-1 10.28 84 1.13 2.49 45%
100-(-40)-15-POEt 15-POEt-40-T 10.4 81.7 1.18 2.49 47%
100-(-40)-15-PA(HCl) 15-PA-40-T 8.48 86.4 0.91 2.49 36%
Table A 4: F: functional group; w % F TGA: weight percentages for F and for SiO2
obtained from TGA; expected nexp and experimental amount nTGA of F per unit mass of
SiO2 (in mmol g-1); Y: Reaction yield of functionalization. Marked samples are discussed
in the main text.
N2 Adsorption Data
Table A 1: (p/p0)pc: pore condensation pressure of nitrogen; D: pore diameter;
vp: pore volume; as: BET specific surface area; vm: micropore volume, C: BET constant.
Sample-old name new name (p/p0)pc D/nm vp/
cm3 g-1 as /
m2g-1 vm
/cm3g-1 C
90-SBA-15a S-acid 0.799 10.0 0.98 464 -0.015 75
90-SBA-15-b S-acid 0.799 10.0 0.96 449 -0.022 66
90-SBA-15-c-gr S-acid 0.797 9.9 0.96 459 -0.020 70
90-SBA-15-d-m S-acid 0.794 9.8 1.00 483 -0.033 55
90-SBA-15 cal 550 S-cal 0.7425 8.5 1.26 947 0.099 -978
90-SBA-15 cal 250.3h-b-m S-acid-cal 0.807 10.2 1.35 807 0.082 -1346
P123s05 SBA-15 0.757 8.9 0.89 648 0.05 1371
90-60-20-CA(1)-1-a 20-CA-60-F(1)-1 0.718 8.0 0.79 506 -0.001 102
90-60-20-CA(1)-1-b 20-CA-60-F(1)-1 0.72 8.1 0.79 471 44.405
90-60-20-CA(1)-2 20-CA-60-F(1)-2 0.72 8.1 0.82 526 -0.010 86
90-60-20-CA(1)-1/2 20-CA-60-F(1/2) 0.716 8.0 1.20
90-60-20-CA(1)-1/2-cal250 20-CA-60-
F(1/2)cal250 0.716 8.0 0.87 571 0.0093 129
90-60-20-CA(1)-3 20-CA-60-F(1)-3 0.718 8.0 0.60 389 0.006 130
90-60-20-CA(1)-4 20-CA-60-F(1)-4 0.716 8.0 0.82 533 0.026 206
100-60-20-CA(1)1 20-CA-20-F(2) 0.644 6.7 0.49 399 -0.001 104
100-90-20-CA(1)1 20-CA-20-F(3) 0.726 8.2 0.88 549 -0.003 98
100-60-10-CA(1)-1
cal250-2h 10-CA-60-F(1)cal250 0.778 9.4 0.98 544 0.018 166
100-60-10-CA(1)-2 10-CA-60-F(2) 0.793 9.8 0.98 490 0.029 248
100-60-10-CA(1)-2-
cal250-2h 10-CA-60-F(2)cal250 0.794 9.8 0.97 513 0.005 121
90_60_20_Pet(1) 20-P-60-F(1) 0.14 42 -0.008 26
90_60_20_Pet(1)
Cal(250 C- 5 h) 20-P-60-F(1)cal250 0.678 7.3 0.73 666 0.061 6449
90_60_20_PA(1)
[(1-a) 48% H2SO4] 20-POEt-60-F(1)-1 0.815 10.5 1.39 687 0.031 213
90_60_20_PA(1)-2a
( 40% H2SO4) 20-POEt-60-F(1)-2-
90_60_20_PA(1)-2-
cal250-3h 20-POEt-60-F(1)-2-
cal250 0.819 10.6 1.20 588 0.022 189
90_60_20_Pet(1)-
cal550°C 20-P-60-F(1)cal550 0.624 6.5 0.26 123 -0.001 89
Sample-old name new name (p/p0)pc D/nm vp/
cm3g-1 as /
m2g-1 vm
/cm3g-1 C
90_60_20_Pet(2) 20-P-60-F(2) 0.12 50 -0.004 46
90_60_20_Pet(2)-
cal550°C 20-P-60-F(2)cal550 0.601 6.2 0.31 220 0.003 130
90_60_20_PA(2)-1 20-POEt-60-F(2)-1 0.694 7.6 0.73 520 0.032 322
90_60_10_Pet(1) 10-P-60-F(1) 0.656 6.9 0.15 80 -0.021 20
90_60_10_PA(1)-1a
[48%H2SO4] 10-POEt-60-F(1)-1 0.753 8.8 0.94 571 0.035 319
90_60_10_PA(1)-1b
[48%H2SO4] 10-POEt-60-F(1)-1 0.758 8.9 0.94 559 0.031 292
90_60_10_PA(1)-1
cal250.3h 10-POEt-60-F(1)-1-
cal250 0.770 9.2 0.94 572 0.038 386
90-60-10-PA(1)-2 10-POEt-60-F(1)-2 0.740 8.5 0.94 617 0.026 209
90-60-15-PA 15-POEt-60-F(1)-1 0.683 7.4 0.80 553 0.022 201
90-60-15-Pet-550°C 15-POEt-60-F(1)-1-
cal550 0.592 6.0 0.31 219 0.004 139
100-95-15-PA(1) 15-POEt-95-F(2)-1 0.676 7.2 0.79 530 0.021 197
100-95-15-PA(1)-
EtOH-Soxlet 15-POEt-95-F(2)-2 0.656 6.9 0.73 503 0.040 583
100-95-15-PA(1)-
EtOH-Kolben 15-POEt-95-F(2)-3 0.646 6.8 0.55 350 -0.021 60
100-(40)-15-PA 15-POEt-40-T(1)-1 0.691 7.5 0.64 502 0.042 619
100-(40)-15-PA-HCl 15-PA-40-T(1)-1 0.693 7.5 0.66 531 0.025 234
100-(40)-15-POEt(1)-2 15-POEt-40-T(1)-2 0.709 7.8 0.36 244 0.010 203
90_60_20_S(1) 7-S-60-F(1) 0.735 8.4 0.18 87 -0.014 30
90_60_20_S(1)-
cal(250°C_5 h) 7-S-60-F(1)cal250 0.765 9.1 1.23 904 0.064 489
90_60_20_SA(1)-1
48% H2SO4 7-SA-60-F(1)-1 0.692 7.5 0.76 560 0.046 938
90_60_20_SA(1)-2
40% H2SO4 7-SA-60-F(1)-2 0.796 9.9 0.79 438 0.005 124
90_60_20_SA(1)-2
40% H2SO4cal250 7-SA-60-F(1)-2cal250 0.8 10.0 0.95 572 0.034 306
90_60_20_SA(2) 20-SA-60-F(2)-1 0.7935 9.8 0.94 557 126
100-00-20-SA 20-SA-00(1) 0.791 9.7 0.96 593 0.029 218
100-00-20-SA(1)-2 20-SA-00(1)-1 0.79 9.7 0.93 564 0.007 125
100-(-55)-20-SA 20-SA-55-T 0.798 10.0 0.77 443 0.022 205
90-1480-30-SA 30-SA-1480-F 0.765 9.1 0.86 672 -0.007 92
Sample-old name new name (p/p0)pc D/nm vp/
cm3g-1 as /
m2g-1 vm /
cm3g-1 C
90_60_10_S(1) 3-S-60-F(1) 0.721 8.1 0.19 81 27
90_60_10_S(1)
cal(250 C. 5 h) 3-S-60-Fcal250 0.7553 8.8 1.17 836 2170
90-60-10-SA(1)-1 3-SA-60-F(1)1 0.799 10.0 1.26 579 0.045 590
Acknowledgements
This Thesis was written at the Stranski Laboratory for Physical and Theoretical Chem-
istry at the Technical University Berlin, in the framework of Sonderforschungbereich 448
"Mesoskopisch struktruierte Verbundsysteme" as a part of the the subproject B1Phasen-
übergänge und Self-assembly in Mesoporen.
First, I would like to thank my supervisor, Professor Gerhard H. Findenegg, for the oppor-
tunity to work in his group, for many stimulating discussions, for all the time he devoted
to my thesis, and for his invaluable advice and guidance. My thanks also go to Professor
Reinhard Schomäcker for agreeing to be my second supervisor, and for critical comments
and encouragement. Moreover I thank Prof. Regine von Klitzing for agreeing to be the
chairlady in my PhD defence.
The experimental work was carried out in cooperation with several groups of the SFB-448
and the Fritz-Haber Institut, various inorganic chemistry departments of TU-Berlin, and
with BAM. I am deeply indebted to the following persons:
Stranski co-workers for such a convenient and harmonic working atmosphere.
Our cooperation partner and NMR specialists Dr. Ilja Shenderovich and Daniel Mauder
from FU-Berlin. Dear Ilja, I thank you for many useful suggestions during our long dis-
cussions. Dear Daniel I thank you for performing all the solid state NMR measurements,
and for such a close and productive cooperation. I think our cooperation within the SFB
will set a paradigm! Now it’s your turn. I wish you all the best in writing your thesis.
Dr. Dirk Kurth (MPI-Golm) for the opportunity to work with MEPE’s, and also for nu-
merous suggestions and discussions.
Torsten Sievers (MPI-Golm) for time-consuming synthesis of the MEPE’s and for his
cooperativeness and assistance.
Dr. Christian Hess (FHI-Berlin) for performing the XPS measurements and also for long
and valuable discussions.
Prof. Hildebrandt (TU-Berlin) for the opportunity to work in his group with different
devices. I thank Dr. Inez Weidinger for helping me with the FT-Raman measurements of
Fe-MEPE in the pores. The spectra were not very predicative due to the complexity of the
system. But that’s science! Special thanks for Dr. Ingo Zebger for his help by performing
the FT-IR measurements.
Dr. Oskar Paris (MPI-Golm) for the opportunity to perform the SAXD measurements
in his group, and Ingrid Zenke for carrying out the SAXD measurements always very
quickly.
Dr. Matthias Koch (BAM-Berlin) for giving me the opportunity to perform CHN ele-
mental analysis in his group.
Prof. Martin Lerch (TU-Berlin) for giving me the opportunity to measure in his group
and Brigitte Hahn for conducting the TGA/DTA measurements.
Prof. Thorsten Ressler (TU-Berlin) for the opportunity to measure in his group, and
Dr. Ingo Piotrowski for performing the Fe-AAS measurements.
Our indefatigable secretary Christiane Abu-Hani for her helpfulness, and always being
there with a contagious smile.
Thanks are also due to Dersy Lugo for her proof-reading and her companionship in our
sports sessions.
I would especially like to thank my friends and colleges in the office; Susanne Jähnert
for always radiating high sprits, and Stefan Wellert for his proof-reading and for lending
a sympathetic ear every time when I had something important to tell.
I owe a special debt of gratitude to Kasia Ciunel-Hänni and Tatjana Mauser for moti-
vating me and for helping me to stay in a good mood.
And many cordial thanks go to Kai, for his huge support during all the years, and for
the readiness to listen to me at any time, and for all the good advice.
Publications
Schemmel, S.; Akcakayiran, D.; Rother, G.; Brulet, A.; Farago, B.; Hellweg, Th.; Finde-
negg, G. H., Phase Separation of a Binary Liquid System in Controlled-Pore Glass, Mat.
Res. Soc. Symp. Proc., 2004,790, P7.2.1.
Akcakayiran, D.; Kurth, D.; Röhrs, S.; Rupprechter, G.; Findenegg, G. H., Self assembly
of a Metallo-Supramolecular Coordination Polyelectrolyte in the Pores of SBA-15 and
MCM-41 Silica, Langmuir 2005,21, 7501.
Shenderovich, I. G.; Mauder, D.; Akcakayiran, D.; Buntkowsky, G.; Limbach, H.-H.;
Findenegg, G. H., NMR Provides Checklist of Generic Properties for Atomic-Scale Mod-
els of Periodic Mesoporous Silica, J. Phys. Chem. B, 2007,111, 12088.
Akcakayiran, D.; Mauder, D.; Hess, C.; Sievers, T.; Kurth, D. G.; Shenderovich, I. G.;
Limbach, H.-H.; Findenegg, G. H., Carboxylic Acid-Doped SBA-15 Silica as a Host for
Metallo-Supramolecular Coordination Polymers, J. Phys. Chem. B, 2008,112, 14637.
Findenegg, G.H.; Jähnert, S.; Akcakayiran, D.; Schreiber, A., Freezing and Melting of
Water Confined in Silica Nanopores, ChemPhysChem, 2008,9, 2651.
Akcakayiran, D.; Findenegg, G. H., Adsorption Kinetics of Metal Coordination Poly-
mers into Acid-functionalized SBA-15, in preparation.
Akcakayiran, D.; Mauder, D.; Tabak, A.; Lerch, M. Shenderovich, I. G.; Kurth, D.
G.; Limbach, H.-H.; Findenegg, G. H., Properties of Acid- functionalized SBA-15. Part
I: Carboxylic, Phosphonic and Sulfonic Acid Containing Materials Synthesized by Co-
condensation, in preparation.
Mauder, D.; Akcakayiran, D.; M. Shenderovich,; D. G.; Findenegg, G. H.; Limbach,
H.-H., Properties of Acid- functionalized SBA-15. Part II: Acidity Measurements by 15N
Solid-State NMR, in preparation.