Synthesis and Activity of Heterogeneous Lewis Acidic Sn Catalysts

Research Collection

Doctoral Thesis

Synthesis and Activity of Heterogeneous Lewis Acidic SnIVCatalysts

Author(s): Conrad, Sabrina

Publication Date: 2015

Permanent Link: https://doi.org/10.3929/ethz-a-010636811

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

https://doi.org/10.3929/ethz-a-010636811

http://rightsstatements.org/page/InC-NC/1.0/

https://www.research-collection.ethz.ch

https://www.research-collection.ethz.ch/terms-of-use

DISS. ETH NO. 23185

Synthesis and Activity ofHeterogeneous Lewis Acidic SnIV Catalysts

A thesis submitted to attain the degree of

DOCTOR OF SCIENCES of ETH ZURICH

(Dr. sc. ETH Zurich)

presented by

SABRINA CONRAD

Dipl. Chemist, Karlsruhe Institute of Technology, Germanyborn on 24.10.1986citizen of Germany

accepted on the recommendation of

Prof. Dr. C. Copéret, examinerProf. Dr. I. Hermans, co-examinerProf. Dr. C. Müller, co-examiner

2015

Life is what happens while you are busy making other plans.

- John Lennon

Acknowledgements

Firstly and mostly, I would like to thank Prof. Dr. Ive Hermans for giving me the opportunity

to conduct my doctoral thesis in a unique research environment. I would like to thank him in

particular for the fascinating insights into industrially related research, the exceptional facilities

and his impressive strive for excellence that has helped me to push things forward. I am also

very grateful for the opportunity to spend the second half of my Ph.D. at the University of

Wisconsin-Madison in the freezing but homey Midwest of the US. This invaluable experience

definitely broadened my mind in many different directions. Ive, thank you for always providing

me with space to try out new ideas, and at the same time for your scientific impulses; this

together greatly contributed to my personal and professional development throughout this

Ph.D. Thank you for letting me be part of an unmemorable trans-atlantic experience, and for

your warm welcome in West Lawn Avenue.

Secondly, I would like to thank Prof. Dr. Christophe Copéret for kindly taking over the

responsibility of being my ETH supervisor when the Hermans lab moved to the US. I am

also very thankful for the scientific discussions I had with him at ETH and UW and for the

opportunity to complete my thesis in his group at ETH.

Prof. Dr. Christoph Müller is kindly acknowledged for agreeing to act as co-examinator to

my thesis.

Special thanks go to the Hermans lab from ETH (Muppets Part I) for creating a very

positive and enjoyable working environment. I want to thank Ceri for his invaluable advice and

support, Philipp for introducing me into the secrets of grafting, and Camila for her tremendous

help in discussions about my research. I will also always remember the unforgettable moments

we shared outside of the lab, such as the fabulous cakes from Natascia, our very individual

group hit parade ("Ceri Ceri lady..."), and the special kind of Hermans humour that no one

can resist. I also want to thank the Hermans lab from Madison (Muppets Part II) for being

such great labmates and making the work in a new place very soon very enjoyable. I want

to thank Florian for his advice with my Sn project, Phil for his help with IR measurements

iv

whenever needed, and Alyssa for the amazing contributions from her Design Department. In

particular, I want to thank Patrick, my longest Ph.D.-mate and "727" roommate - for always

being a motivating and reliable labmate, for his Franconian way of cheering-up and for making

the move to Madison much easier than I thought. I also want to thank a list of people for their

support with logistical, technical and electronical questions throughout this Ph.D. As such, my

thanks go to René Verel, Max Wohlwend, Andreas Dutly, Erol Dedeoglu, Roland Walker and

Urs Krebs (Zurich), as well as Kat Myhre, Kristi Heming, Jeff Nielsen, Tracey Drier and Matt

Martin (Madison).

Lastly but truly, I want to thank my beloved family and friends. Nina, Janne, Franzi and

Amaia (ETH-Grazien) for turning my time at ETH into much more fun and for always being

there to chat. Hanna, for being my first and most sincerest consultant throughout this Ph.D.

My mom (Jutta), my dad (Reinhard) and Felix for their unconditional support and care, and

for just being there when I needed it most. Michael (Mr Mbughuni), who has stepped into

my life very recently, for genuinely sharing my academic ups-and-downs, for his thoughtful

encouragement, and for showing me day by day what is important.

Table of Contents

Acknowledgements iii

Table of Contents v

Summary vii

Zusammenfassung xi

1 Introduction 1

2 Simple and Scalable Preparation of Highly Active Lewis Acidic Snβ 15

3 Insights into the Baeyer-Villiger Oxidation of Cyclohexanone withH2O2 catalyzed by Snβ 25

4 Silica-Grafted SnIV Catalysts in Hydrogen-Transfer Reactions 39

5 Confinement Effects in Hydrogen-Transfer Reactions on Sn Sitesin Porous Silica Materials 57

6 Conclusions and Outlook 63

Bibliography 69

Appendix A Annexes 87

Appendix B List of Publications 99

Appendix C Presentations 101

Appendix D Cover Gallery 103

Appendix E Curriculum Vitae 109

Summary

The discovery of novel catalytic solutions for Lewis acid catalyzed reactions is and will be an

object in the focus of academical and industrial research efforts. This is primarily stimulated

by the broad array of catalytic applications of Lewis acids, and the need to re-conceptualize

existing homogeneous processes on the way to a more sustainable chemical industry. In

this regard, particularly promising results have lately been obtained by incorporating isolated

metal sites into framework positions of microporous zeolite structures. However, due to the

complexity of zeolitic frameworks, a fundamental understanding of the intrinsic reactivity,

and the subsequent tailored material engineering presents a significant challenge. In addition,

conventional synthesis procedures for such materials (i.e., hydrothermal syntheses) are in many

cases time- and skill-demanding, and may furthermore pose environmental issues, which thus

far restrict their application to laboratory scale.

Taking Sn-doped β-zeolite (Snβ) as a relevant example for the seminal class of Lewis

acid solids with appealing applications, the first part of this thesis (Chapter 2) addresses the

general interest to translate the preparation of this material to a commercial level. This is

approached by the development of a straightforward synthesis strategy, consisting of the post-

synthetic incorporation of Sn atoms into dealuminated, commercially available β-zeolite, which

overcomes several of the practical hurdles given by the conventional hydrothermal synthesis.

The obtained material possesses similar structural characteristics to conventional Snβ, and

demonstrated similar, or even higher activities and selectivities in the Baeyer-Villiger oxidation

of cyclohexanone with hydrogen peroxide and a triose sugar isomerization. This, in combination

with the increased amount of incorporated metal based on this novel procedure, is expected to

significantly facilitate the industrial utilization of this or a similar material.

The chemoselective interaction between the Lewis acidic SnIV-sites of Snβ and carbonyl

groups gives this catalyst unprecedented selectivity in reactions with carbonyl substrates.

At the same time, the chemoselectivity of the SnIV-sites exposes Snβ to potential catalyst

inhibition caused by the competitive adsorption of other Lewis basic molecules that may

viii

be present as solvent or product molecules in the reaction mixture to the active sites. In

Chapter 3 we demonstrate that such inhibition occurs for the Baeyer-Villiger oxidation of

cyclohexanone with hydrogen peroxide by measuring catalytic activities as a function of

lactone and water concentrations in the reaction solution. Aiming toward improving the

activity of Snβ in Baeyer-Villiger oxidations, we then modify our novel post-synthetic synthesis

method and prepare different Snβ catalysts with varying amounts of framework silanols (i.e.,

varying hydrophilicities), while monitoring changes in activity. We find that catalyst activity

goes through a maximum as a function of hydrophilicity, which indicates that a hydrophilic

framework aids the adsorption of the ketone substrate to the active SnIV-sites, and that this

effect is outweighed by competitive adsorption of solvent (water) and product (lactone) at higher

hydrophilicities. Hence, this work clearly illustrates that zeolite hydrophilicity influences the

activity of Snβ in Baeyer-Villiger oxidations with hydrogen peroxide, and that flexible synthesis

methods, such as our post-synthetic metal incorporation, allow optimizing activities through

targeted structural modifications.

In the second part of this thesis (Chapter 4) we study site-isolated silica-grafted SnIV-

sites as a catalytic model-system for SnIV/SiO2-based catalysts. The surface-anchored SnIV-

sites are stepwise functionalized, while implications of these modifications on catalyst structure

and activity are followed by in-depth material characterization and catalytic tests. With this

methodology, we distinguish different material properties that add to the activity of SnIV/SiO2-

based catalysts in the Meerwein-Ponndorf-Verley reduction of cyclohexanone with 2-butanol

(active site speciation, hydrophilicity, confinement effects). We furthermore demonstrate the

feasibility of model-systems as a tool to investigate which features contribute to the performance

of an active Lewis acid catalyst. Not least, the results presented in this chapter, indicate that

the grafting of our amine-containing Sn precursor leads to opening of siloxane bridges on the

surface of thermally pretreated silica; an observation that raises interesting questions for future

studies involving the grafting of various metal precursors and ligands.

The results, which were drawn from our catalytic model-study, indicate confinement effects

as one underlying reason for the difference in activity between benchmark Snβ and silica-grafted

SnIV-sites. With the aim of assessing this contribution in more detail, we extend our study by

ix

grafting SnIV-sites on to the surface of mesoporous MCM-41 (Chapter 5). In line with our

findings presented in Chapter 4, we find a strong dependence of activity on the pore size of

the support. We propose that this trend in reactivity can be attributed to an adsorption-

based confinement, which most likely consists of a suppressed reverse reaction in confined

environments.

Overall, this thesis describes decisive steps on the long path toward an improved

understanding and industrial realization of novel solid Lewis acid catalysts. We demonstrate

that innovative strategies, such as the utilization of model-systems or unconventional synthesis

approaches, can bring outstanding advancements. The results presented in this thesis also

expose the current challenges in the field, in particular, the complexity that zeolite catalysis

still represents, despite the insights gained to date.

Zusammenfassung

Die Entdeckung neuer katalytischer Lösungsansätze fur Lewissäure-katalysierte Reaktionen

ist und wird im Schwerpunkt akademischer und industrieller Forschung stehen. Dies liegt

vorwiegend am breitgefächerten Anwendungsbereich von Lewissäuren und dem dringenden

Bedarf existierende homogene Prozesse durch heterogene Konzepte zu ersetzen um den Weg

zu einer nachhaltigeren chemischen Industrie zu ebnen. In diesem Zusammenhang wurden

mit dem Einbau von isolierten Metallzentren in Gerüstpositionen von mikroporösen Zeolithen

vor kurzem äußerst vielversprechende Ergebnisse erzielt. Allerdings stellt ein fundamentales

Verständnis der intrinsischen Reaktivität dieser Materialien und die darauf folgende gezielte

Entwicklung neuer Materialen durch die Komplexität von zeolithischen Strukturen eine große

Herausforderung dar. Hinzu kommt, dass die konventionellen Syntheseverfahren derartiger

Materialien (d.h., hydrothermale Synthesen) in der Regel ein hohes Maß an synthetischem

Geschick sowie Zeit erfordern, und nicht selten belastend für die Umwelt sind. Dies hat die

Anwendung dieser Materialien bisher auf Synthesen im Labormaßstab beschränkt.

Vor dem Hintergrund eines außerordentlichen Vertreters der zukunftsträchtigen

Materialklasse von heterogenen Lewissäuren – Sn-dotiertem β-Zeolithen (Snβ) – widmet sich

der erste Teil dieser Dissertation (Kapitel 2) dem allgemeinen Interesse die Herstellung dieses

Materials auf kommerzieller Ebene zu ermöglichen. Dies geschieht durch die Entwicklung

einer simplen Synthesestrategie, bestehend aus der post-synthetischen Inkorporierung von

Sn-Atomen in dealuminierten, kommerziell erhältlichen β-Zeolithen, wodurch einige der

praktischen Schwierigkeiten, die mit der konventionellen hydrothermalen Synthese verbunden

sind, umgangen werden können. Das erhaltene Material weist ähnliche strukturelle

Eigenschaften zu konventionellem Snβ auf und demonstrierte ähnliche oder sogar höhere

Aktivitäten und Selektivitäten in der Baeyer-Villiger-Oxidation von Cyclohexanon mit

Wasserstoffperoxid sowie in einer Dreifachzucker-Isomerisierung. Mit Betonung auf der

gesteigerten Menge an inkorporiertem Metal ist zu erwarten, dass unsere neue Synthesestrategie

die industrielle Verwendung dieses oder eines ähnlichen Materials wesentlich erleichtern wird.

xii

Die chemoselektive Wechselwirkung zwischen den Lewissauren SnIV-Zentren in Snβ und

Carbonylgruppen verleiht diesem Katalysator seine beispiellose Selektivität in Reaktionen mit

Carbonylsubstraten. Gleichzeitig wird Snβ durch die Chemoselektivität seiner SnIV-Stellen

möglicher Katalysatorinhibierung ausgesetzt, da andere Lewisbasische Moleküle, die in der

Reaktionsmischung vorhanden sind (in Form von Lösungsmittel- oder Produktmolekülen),

um die Adsorption an den aktiven Stellen konkurrieren können. In Kapitel 3 zeigen

wir, dass solch eine Inhibierung für die Baeyer-Villiger-Oxidation von Cyclohexanon mit

Wasserstoffperoxid zu ε-Caprolakton auftritt, indem wir die katalytische Aktivität von Snβ in

Abhängigkeit von unterschiedlichen Lakton- undWasserkonzentrationen in der Reaktionslösung

aufzeichnen. Mit dem Ziel die Aktivität von Snβ in Baeyer-Villiger-Oxidationen zu verbessern,

modifizieren wir im Weiteren unsere neue post-synthetische Synthesestrategie und präparieren

eine Reihe von Snβ-Katalysatoren mit unterschiedlichen Mengen an Gerüstsilanolgruppen

(d.h., unterschiedlichen Hydrophilitäten). Wir stellen fest, dass die Katalysatoraktivität als

Funktion der Materialhydrophilität durch ein Maximum geht, wodurch angezeigt wird, dass ein

hydrophiles Framework die Adsorption des Ketonsubstrates unterstützt, und dass dieser Effekt

bei höheren Hydrophilitäten durch kompetitive Adsorption von Lösungsmittel (Wasser) und

Produckt (Lakton) überwogen wird. Diese Studie illustriert daher eindeutig, dass Hydrophilität

die Aktivität von Snβ beeinflussen kann, sowie dass flexible Synthesemethoden, wie zum

Beispiel unsere post-synthetische Metallinkorporierung, Aktivitätsoptimierung durch gezielte

strukturelle Abänderungen ermöglicht.

Im zweiten Teil dieser Dissertation (Kapitel 4) untersuchen wir isolierte SnIV-Stellen auf

einer Silikaoberfläche als katalytisches Modellsystem für SnIV/SiO2-basierte Katalysatoren.

Die SnIV-Stellen, die mittels einer Grafting-Technik auf die Silikaoberfläche aufgebracht

werden, werden schrittweise funktionalisiert. Diese Modifikationen und deren Auswirkungen

auf die Struktur des Katalysators werden mittels eingehender Charakterisierung und

katalytischen Tests verfolgt. Hierdurch können wir unterschiedliche Materialeigenschaften,

die zur Aktivität von SnIV/SiO2-basierten Katalysatoren in der Meerwein-Ponndorf-Verley-

Reaktion von Cyclohexanon mit 2-Butanol beitragen, unterscheiden (Struktur der aktiven

Stellen, Hydrophilität, Begrenzungeffekte). Darüber hinaus veranschaulicht unsere Studie die

xiii

Möglichkeit Modellsysteme als Hilfsmittel zu instrumentalisieren, um die unterschiedlichen

Eigenschaften, die zur Aktivität eines Lewissäurekatalysators beitragen, zu identifizieren. Nicht

zuletzt deuten die Ergebnisse diese Kapitels an, dass das Grafting von einem Amingruppen-

tragenden Sn-Ausgangsstoff zum Öffnen von Siloxanbrücken auf der Oberfläche des thermisch

vorbehandelten Silikaträgers führt. Diese Beobachtung wirft interessante Fragen für mögliche

Folgestudien auf, die das Grafting von unterschiedlichen Metalausgangsstoffen sowie -liganden

untersuchen.

Die Resultate, die wir aus unserer katalytischen Modellstudie gewonnen haben,

deuten darauf hin, dass Begrenzungseffekte eine zu Grunde liegende Ursache für den

Aktivitätsunterschied zwischen dem derzeitigen Benchmark-Material Snβ und SnIV-Stellen

auf einer Silikaoberfläche sind. Mit dem Ziel diesen Beitrag genauer bemessen zu

können, haben wir unsere Modellstudie in Kapitel 5 erweitert, indem wir SnIV-Stellen

auf mesoporösem MCM-41-Trägermaterial aufbringen. In Übereinstimmung mit unseren

Erkenntnissen aus Kapitel 4 stellen wir eine starke Abhängigkeit der Aktiviät von der

Porengröße des Silikaträgermaterials fest. Wir schlagen vor, dass dieser Aktivitätstrend auf

einem adsorptionsbasierten Begrenzungseffekt beruht, welcher höchstwahrscheinlich auf eine

unterdrückte Rückreaktion in stärker begrenzten Umgebungen zurükzuführen ist.

Gesamtheitlich beschreibt diese Dissertation entscheidende Entwicklungen auf dem Weg

zu einem verbesserten Verständnis sowie zur industriellen Implementierung von neuen festen

Lewissäurekatalysatoren. Wir demonstrieren, dass innovative Strategien, wie die Verwendung

von Modellsystemen und nicht-konventionelle Syntheseansätze bedeutende Fortschritte erzielen

können. Darüber hinaus legen die Ergebnisse dieser Dissertation bestehende Herausforderungen

des bearbeiteten Forschungsfeldes dar, insbesondere die Komplexität, die Zeolithkatalyse trotz

der bislang gewonnenen Erkenntnisse nach wie vor mit sich bringt.

Chapter 1

Introduction

In this chapter, a short introduction to heterogeneous Lewis acid catalysis is given. First of

all, the on-going shift from homogeneous to more sustainable heterogeneous catalytic systems

is illustrated and several potential industrial applications are presented. In the following, a

short overview of different approaches to synthesize solid Lewis acid catalysts is given. In the

end, the activity of solid Lewis acid catalysts, with the focus set on metal-containing zeolites,

is discussed by assessing the different material features contributing to their reactivity.

1.1. Sustainable Heterogeneous Lewis Acid Catalysis

Lewis acid catalyzed reactions constitute one of the most thoroughly investigated catalytic

systems, in which an organic molecule undergoes diverse chemical transformations with a

nucleophilic reagent in the presence of a Lewis acid catalyst.[1–3] The catalytic activity of Lewis

acids is rooted in the formation of an acid-base adduct between the catalyst and one of the

two reaction substrates in order to enhance their relative reactivity.[4] The process of activation

involves the transfer of electron-density from a substrate into the empty orbitals of the Lewis

acid, making it more prone to a nucleophilic attack.

Within the sector of industrial chemistry, Lewis acids have been found to accelerate an

array of pivotal chemical transformations,[5,6] including (but not limited to) olefin epoxidations

and isomerization reactions (Figure 1.1; A, C and D). Traditionally, homogeneous Lewis

acids, for instance those based on AlCl3 or ZnCl3 have been used for industrial applications

(e.g., for Friedel-Crafts alkylations).[7,8] However, for large-scale processes, their heterogeneous

counterparts are highly desirable both from an environmental and process point of view. Indeed,

the utilization of solid catalysts facilitates catalyst recovery, reuse and continuous production

technologies, and generally leads to reduced quantities of metal waste (i.e., inorganic salts

dissolved in wastewaters). Additionally, the emerging interest in the exploration of biomass

2 Chapter 1

(A) (B)

(D) (C)

(E) (F)

Figure 1.1. Examples of Lewis acid catalyzed reactions (with particular attention to SnIV catalyzedtransformations). (A) Epoxidation of propylene. (B) Baeyer-Villiger oxidation of cyclohexanone withhydrogen peroxide. (C) Isomerization of glucose to fructose. (D) Isomerization of the C-3 sugarsdihydroxyacetone (DHA) or glycerinaldehyde (GLA) to lactic acid in aqueous media. (E) Meerwein-Ponndorf-Verley reduction of cyclohexanone with 2-butanol. (F) Rearrangement of β-pinene oxideinto myrtanal.

as a sustainable carbon source,[9–13] and an environmentally driven shift toward non-toxic and

non-oxidizable solvents[14,15] has stimulated the search for well-performing Lewis acid catalysts,

which are stable in aqueous media.[16,17] The use of conventional homogeneous metal complexes

for such purposes is problematic due to prevailing deactivation caused by catalyst hydrolysis.

The pronounced hydrophobic character that has been found for several solid Lewis acids thus

offers unique opportunities for the design of novel water-tolerant catalytic systems.

Immobilized complexes of multiple main group and transition elements such as Al, Sn, Ti,

V, Cr and Fe, have proven to feature distinctive Lewis acidic properties.[5] Amongst these, Ti

is by far the most studied representative, given its unprecedented activity and selectivity in

oxidation reactions (e.g., epoxidations), which are the tool for the synthesis of huge quantities

of intermediates and monomers for the polymer industry. In this regard, the development of

TS-1 (a TiIV-doped MFI-type zeolite) is, for instance, considered as one of the most significant

Introduction 3

material innovations during the past decades.[18,19] Due to the high specificity of TS-1 toward

hydrogen peroxide, several peroxide-based reactions, such as the hydroxylation of phenol, the

ammoximation of cyclohexanone and cyclododecanone, and the epoxidation of propene, are

now industrial processes.[20] Propene epoxidation is an example of liquid phase oxidations that

have been the focus of studies striving for a more sustainable technology, which was realized

with the commercial implementation of the HPPO process, using aqueous hydrogen peroxide

as the oxidant for propene (largest HPPO plant: Antwerp/Belgium, BASF/Dow, 300 000 t

a-1).[20]

Besides Ti, Sn is a critical component that has recently received increasing attention due

to its widespread applicability.[21] Amongst other Lewis acids, it appears that Sn possesses the

unique capability to activate carbonyl functional groups without activation of the oxidant itself

(such as hydrogen peroxide), thus avoiding undesirable side-reactions and poor selectivities

in the overall reaction. One example is the Baeyer-Villiger oxidation, which converts

ketones (readily available building block compounds) into value-added esters or lactones for

the production of polymers (Figure 1.1 B).[22,23] Traditionally, peracids have been used as

stoichiometric oxidants for this reaction, resulting in one equivalent of acid by-product. The

heterogeneously catalyzed version of the Bayer-Villiger oxidation, utilizing hydrogen peroxide

as oxidant, is hence very appealing due to the simplification of the process conditions and the

elimination of stoichiometric waste.[24,25]

An additional emerging field for Lewis acid Sn catalysts is the conversion of renewable

carbohydrates to chemicals and fuels as alternate pathways to the "classical" chemical industry,

which is currently highly dependent on crude oil. Attractive routes are the isomerization

of glucose to fructose (Figure 1.1 C) for the production of high-fructose corn-syrups (HFCS,

8 x 106 tons/yr),[26–29] as well as the isomerization of triose sugars (dihydroxyacetone DHA,

glyceraldehyde GLA) in order to synthesize lactate-derivatives, such as lactic acid (Figure

1.1 D),[30–33] which are employed for the production of biodegradable polymers and solvents.

Moreover, furan derivatives such as 5-hydroxymethylfurfural (HMF) and furfural (FUR) can

be obtained from the selective dehydration of monosaccharides, which have been stated to be

relevant compounds for fuel, polymer and pharmaceutical industries.[34–36] As an example, the

4 Chapter 1

selective oxidation of HMF leads to 2,5-furandicarboxylic acid (FDCA), which is a potential

replacement for terephthalic acid that is widely used in synthetic polyesters such as those found

in soft drink bottles.[37,38]

Other applications are Meerwein-Ponndorf-Verley-Oppenauer redox reactions (Figure

1.1 E), which convert aldehydes and ketones to their corresponding alcohols,[39,40] various

rearrangements, such as the transformation of β-pinene oxide into myrtanal, which is used

as antiseptic (Figure 1.1 F),[41–43] and Carbon-Carbon Coupling Reactions.[44,45]

1.2. Synthesis of Solid Lewis Acid Catalysts

In general, there are two main approaches to synthesize solid Lewis acid catalysts as depicted

schematically in Figure 1.2: (A) The direct incorporation of the active species during the

synthesis of the material, e.g., via sol-gel chemistry, or (B) the immobilization of the active

species at the surface of pre-synthesized materials, e.g., via grafting.

The foremost process (Figure 1.2 A), also known as isomorphous substitution, is exemplified

by Lewis acid doped zeolites, in which a heteroatom substitutes the Si or Al in the framework of

a microporous crystalline solid (i.e., a zeolite). A pioneer in this field is the aforementioned TS-

1 (Ti-silicate-1), a siliceous material containing low amounts of Ti (< 2 wt%), isomorphously

substituted into the framework of MFI-zeolite. Stimulated by the successes achieved with TS-1,

interesting sites for catalytic applications were evolved when other metal atoms, such as Sn, were

embedded into framework positions of the large-pore β-zeolite (pore diameter ca. 0.7 nm).[46,47]

Firstly synthesized in 1998, this material soon turned out to be a chemoselective catalyst with

exceptional activities in carbonyl-based transformations (Figure 1.1 B-F).[26,30,39,48]

The preparation of metal-doped zeolites such as TS-1 or Snβ resembles the synthesis of

zeolites with the difference that an appropriate amount of metal precursor is added to the

synthesis mixture.[49,50] Usually, the syntheses are carried out in acidic or basic media from gels

containing a silica source, a metal precursor, a suitable template (e.g., tetraethylammonium

hydroxide, tetrapropylammonium hydroxide), fluoric acid and water. The materials are

subsequently crystallized at high-temperature (> 150 °C) and autogenic pressure over a period

of 24 – 240 hours. Depending on the utilized template and the exact synthesis conditions,

Introduction 5

SiO2

M+X$

(A)  direct incorporation e.g., via sol-gel chemistry

(B) immobilization on pre-synthesized supports e.g., via grafting on to silica

M$ M$ M$

M$

silicon'source'

M+X$

addi-ves'

Figure 1.2. Schematic outline of different synthesis approaches to synthesize solid Lewis Acidcatalysts.

a broad spectrum of materials with pore diameters from the micro- to the mesopore range

(micropores: < 2 nm, mesopores: 2 – 50 nm) with metal sites incorporated in the framework

may be prepared with this methodology.

Despite the promising developments in the preparation of solid Lewis acids through

isomorphous substitution, not a few of the obtained materials suffer from a number of

drawbacks, which complicate their industrial realization on a commercial level and necessitates

the exploration of alternative synthetic procedures. Many syntheses, for example, still rely on

the traditional fluoride method, which has the advantage of generally leading to higher product

yields and increasing the hydrophobicity of the obtained zeolite.[51] However, this route impedes

large-scale production due to the corrosive reaction conditions and the enormous environmental

impact of fluoride systems. Moreover, the incorporation of large Lewis acid centers, such as

SnIV, typically entails retardation of the zeolite nucleation, resulting in large crystals (approx.

0.5 – 2 µm), and therefore poor molecular diffusion throughout the zeolitic channels. Not

least, only restricted quantities of metal may commonly be incorporated into the framework

structures (< 2 wt%) via this route.

Owing to these difficulties and aiming at extending the usage of metal-doped zeolites

to reactions with bulkier molecules, the active metals were thereafter incorporated into the

framework of zeolites with larger porosity and into mesoporous structures. In this regard, Ti-

MWW exemplifies a successful attempt to overcome the pore restrictions of the medium-sized

6 Chapter 1

pores of TS-1 (approx. 0.55 nm).[52,53] The unique topology of the MWW-zeolite, with its 12

membered-ring "supercages" and pockets on the exterior has indeed proved to be more useful

for oxidations of larger molecules compared to TS-1.[52] Similar efforts have been undertaken

with SnIV as Lewis acid, leading to novel materials such as the mesoporous stannosilicates Sn-

MCM-41,[54–57] Sn-MFI[28,58] and Sn-SBA-15[59]. Despite their advantageous (more spacious)

pore architecture these catalysts can barely match the activities and selectivities of Snβ, which

cannot simply be explained by differences in diffusion properties but rather points to additional

changes in the material when shifting to a different type of framework structure (e.g., a different

local environment of the active sites or the amorphous structure of the silica walls).[59,60]

The immobilization of the active species at the surface of pre-synthesized materials (Figure

1.2 B) can be carried out by impregnation and grafting. To the best of our knowledge,

impregnation techniques are not thoroughly explored for the synthesis of solid Lewis acid

catalysts. One reason for this fairly poor amount of impregnation studies is the fact that

these techniques may lead to oligomeric species if a slight excess of precursor is added during

the impregnation step. Site-isolation (viz., well isolated species) is, however, a crucial trait to

obtain a highly active solid Lewis acid, since it provides that the individual metal centers on the

catalyst surface do not interact with each other.[61,62] There is one fairly well studied catalytic

material prepared by a wet impregnation technique. In this work, a commercial silica support

was co-impregnated with aqueous solutions of both a Pt and a Sn chlorine salt solution, to give

a Pt-Sn/SiO2 catalyst which was subsequently tested for the isomerization of β-pinene oxide

to naturanol (Figure 1.1 F).[41,42]

Grafting methods, on the other hand, are significantly better understood, and have

been investigated on several support materials (oxides, porous materials, clays).[63–69] In this

approach, metal precursors react with the surface functional groups of pre-synthesized supports,

which are in most of the cases SiO2- or Al2O3-based materials. In this regard, TiCl4-

grafted silica is an example for an industrially implemented catalyst prepared via grafting,

which is used for the epoxidation of propylene with ethylbenzene hydroperoxide (styrene

monomer/propylene oxide process, SMPO).[20] More grafting materials are currently researched

and can be found in the literature. As such, Corma and co-workers grafted different Sn-alkyl

Introduction 7

Figure 1.3. Atomplanting method as demonstrated by Wu and co-workers.[70] Sn atoms are post-synthetically incorporated into nanosized Snβ zeolites by exposing them with SnCl4 vapors at elevatedtemperatures.

precursors (RnSnIVX4-n) on to mesoporous MCM-41 and demonstrated that the precursors

with one or two alkyl substituents (e.g., BuSnCl3) reacted most efficiently with the silanol

groups of the support while avoiding the formation of inactive polymeric tin-oxo species.[64] A

comparison of the catalytic activity with SnIV-sites incorporated during the synthesis of MCM-

41 showed, however, that the catalysts prepared by direct synthesis gave some higher activity.

Another straightforward post-synthetic synthesis approach based on grafting was demonstrated

by Wu and co-workers, who treated highly dealuminated β-zeolite with SnCl4 vapor at elevated

temperatures (773 K) (Figure 1.3).[70] In this case, the Sn sites were introduced into the

framework via the reaction between silanols of the hydroxyl nests created during dealumination

and the SnCl4 molecules. The level of metal incorporation achieved with this synthesis (6.2 wt%

Sn) clearly exceeded the typical metal loading of Snβ synthesized via isomorphous substitution

(< 2 wt% Sn), indicating promising opportunities for industrial applications. Stimulated by this

work, many similar studies were undertaken, aiming toward the optimization of post-synthetic

methodologies for preparing heterogeneous Sn catalysts (see also Chapter 2).[71–75]

In studies involving the preparation of solid Lewis acids via grafting, the support material

is often carefully pretreated prior to contact with the metal precursor, which sets up unique

pathways that allow to control the exact structure of the immobilized metal sites and thus to

conduct rational structure-property studies. In this regard, silica presents an extensively studied

support for the preparation of isolated well-defined surface metal complexes by reaction with an

appropriate (organo-)metallic precursor.[76–85] In such a synthesis, a thermal treatment (under

vacuum) may initially be applied to reduce the amount of vicinal (hydrogen-bonded) silanol

8 Chapter 1

groups on the surface (Figure 1.4 A2) by condensation and elimination of water (Figure 1.4

B). At high temperatures (> 500 °C), this dehydration process results in low silanol densities

on the silica surface and an average distance between individual silanols that is sufficiently

enough to exclude interactions.[86,87] Upon reaction with a metal precursor ML4 that holds at

least on hydrolysable ligand isolated ≡Si-O-ML3 sites can then easily be introduced to the

surface (Figure 1.4 C).[88–96] Even though this strategy is often far from commercialization

due to the complex synthesis conditions and the costly metal precursors, it may lead to an

enhanced fundamental understanding given by the possibility to form uniform active sites,

which makes the establishment of structure-activity relations possible. This is exemplified by

several catalytic systems (e.g., alkane metathesis) that have been discovered and/or studied

using the conception and the techniques of surface organometallic chemistry.[97–100]

1.3. Activity of Solid Lewis Acid Catalysts

Having the superior performance of metal-doped zeolites in mind, it is not entirely surprising

that this class of materials has lately been the object of extensive computational and

experimental studies, striving for a better understanding of their catalytic behaviour. As

outcome from these works, researchers propose that the observed catalytic performance is

rooted in the combination of numerous material properties: (i) the exact local structure of the

active metal sites (i.e., site speciation),[21,88,89,101–105] (ii) the exact crystallographic site in the

lattice where the metal substitution with Si or Al during the synthesis has occurred,[106,107]

(iii) physical properties of the material, including material hydrophilicity,[16,17,108,109] pore size

distributions and crystallinity, (iv) confinement effects induced by the pores of the zeolite.[110–114]

Amongst these material properties, the identification of the nature of the active sites in solid

Lewis acids has recently been given highest priority. Nevertheless, the accurate determination

remains in many cases a topic of much debate. Indeed, two different types of active sites

have been proposed for Snβ.[101,111] Earlier data from 119Sn MAS NMR,[115] XPS,[47] TEM[47,106]

and EXAFS[47,106] suggested the presence of fully-encapsulated tetrahedral framework SnIV-

sites (Sn(–O–Si)4). More recent investigations, involving IR spectroscopy of adsorbed

acetonitrile,[101,111] pointed toward the existence of a second type of site, namely partially

Introduction 9

M"

M"M"M"

(A)

(B)

(C)

1

2

3

Figure 1.4. Schematic illustration of the immobilization of isolated metal sites on thermallypretreated silica. (A) Untreated silica surface consisting of isolated (1), vicinal (hydrogen-bonded)(2) and geminal (3) silanols. (B) Silica after thermal treatment at 700 °C with remaining isolated andgeminal silanols, as well as siloxane bridges. (C) Silica surface after reaction of (B) with an appropriatemetal precursor that has reacted with isolated and geminal silanols.

hydrolyzed Sn ((–Si–O)3Sn–OH), which was also reported in several NMR studies.[116–121]

Adsorption studies indirectly probe the active sites of a solid Lewis acid by studying shifts

in vibrational bands of adsorbed probe molecules, and correlate the extent in shift with the

Lewis acidity of the metal sites. In addition, comparison of experimentally observed shifts and

calculated ones are employed to draw conclusions about the active site structure.[122] In some

recent studies, individual bands are even used to quantify the amount of individual sites.[101]

However, not only the nature of the active site but also multiple adsorbed probe molecules might

cause broad IR shifts.[123] Moreover, these studies exclusively adsorb one probe compound and

neglect superimposed effects from additional molecules (solvent, product or other substrate

molecules) that may be present in the reaction mixture (i.e., under realistic conditions). Thus,

10 Chapter 1

there is a necessity to critically assess such studies by thorough consideration of potential effects

from co-adsorbed molecules, and to careful interpret the obtained experimental data.

Besides active site speciation, confinement effects have been demonstrated to impact

zeolite-catalyzed reactions.[110–114] These effects, which were captured as shape selectivity in

the early history of zeolite catalysis, are traced back to Van der Waals interactions between

reactants, products or intermediates and the walls of the zeolite cavities. This may result

in the stabilization of confined intermediates or transition states and significantly alter the

dynamics of elementary steps within a catalytic cycle. In addition, confinement effects may

influence the adsorption behavior of molecules in zeolites.[124,125] Indeed, it has been found that

the free enthalpies of adsorption of simple hydrocarbons in zeolites raise with an increase in

alkyl chain length, which was ascribed to an increase in the Van der Waals interactions of the

alkanes with every additional -CH2- group.[126] These findings have significant consequences for

the understanding of hydrocarbon-based zeolite-catalyzed reactions and ought to be carefully

considered.

Lastly, a key determinant of metal-doped zeolites is the encapsulation of the Lewis acid sites

inside a water-resistant framework (as found for highly siliceous zeolites), which prevents the

hydrolysis and subsequent deactivation of the Lewis acid, and thereby allows these promising

materials to be utilized for aqueous phase reactions.[16,17,109,127] This feature might make Lewis

acid doped zeolites become the material of choice for biomass conversions in water and biphasic

water-organic mixtures. Moreover, hydrophobic surroundings can weaken the binding of water

(or other hydrophilic solvent molecules such as alcohols) to the actives sites, which aids

in attenuating catalyst inhibition due to competitive adsorption.[17] Therefore, hydrophobic

environments as found in solid Lewis acid zeolites impart the material with exceptional stability

that generally leads to enhanced activities for such hydrophobic materials.

Introduction 11

1.4. Scope of the Thesis

The goal of this thesis is, on one hand, to enhance the large-scale applicability of Lewis acid

heterogeneous catalysts with tremendous catalytic potential. This necessitates the design of

a novel synthetic approach and targeted customization of the synthesis procedure depending

on the chosen application. In particular, this includes detailed material characterization and

catalytic testing in order to assure that the newly developed materials meet the desired material

properties and activities. Attention will be placed on industrially relevant test reactions and

commercially available materials. On the other hand, the goal of this thesis is to enhance the

understanding of activity of current benchmark Lewis acid heterogeneous catalysts, and hereby

aid the rational design of similar catalytic systems. With this in mind, a catalytic model-

system will be created, which enables the identification of explicit material properties that

influence activity by gradual modifications to the material. We expect that improved synthesis

and rationalization of activity will enable the optimized development of catalytic systems and

promote the introduction of sustainable industrial processes based on Lewis acid heterogeneous

catalysts.

1.5. Outline of the Thesis

The results of this thesis are presented in four chapters (Chapter 2 – 5) followed by the

conclusions and outlook of the challenges ahead (Chapter 6). The research presented in this

thesis has been carried out in the frame of an ETH Grant (ETH-38 12-1) titled "Synthesis and

Characterization of Sn-containing materials for sustainable Baeyer-Villiger oxidations".

Currently, Lewis acid doped zeolites are typically prepared in complicated hydrothermal

syntheses procedures, which have several hurdles that complicate or prevent their industrial

implementation. In Chapter 2 we address these limitations for the novel and state-of-the-art

heterogeneous Sn-catalyst Snβ, which to date has been restricted to the laboratory scale and is

prepared in gram quantities under precisely controlled conditions. Aiming toward the industrial

utilization of this or a similar material, we design a straightforward synthesis strategy, applying

a convenient post-synthetic route that consists of the incorporation of the metal into a pre-

synthesized zeolite. Extrapolation of the properties of the prepared material is undertaken by

12 Chapter 1

spectroscopic and catalytic comparison with hydrothermally synthesized Snβ.

The chemoselective interaction between Lewis acidic SnIV-sites and carbonyl groups gives

Snβ its unrivalled selectivity in a wide range of industrially relevant reactions, such as the

Baeyer-Villiger oxidation of cyclohexanone with hydrogen peroxide. At the same time, the

chemoselectivity of the SnIV-sites exposes Snβ to activity-restricting interactions with further

Lewis basic molecules that are present in the reaction mixture, such as the product lactone

and water in Baeyer-Villiger oxidations, as we show in Chapter 3. Based on the synthesis

knowledge gained in Chapter 2, we explore the opportunity to reduce catalyst inhibition

through co-adsorbed molecules to the active sites by modifying the surface properties of post-

synthetically prepared Snβ. This is attained by varying the amount of framework silanols

through modification of the synthesis protocol of the parent zeolite, subsequent quantifying of

the obtained surface hydrophilicities and monitoring of activity changes in the Baeyer-Villiger

oxidation of cyclohexanone with hydrogen peroxide. Furthermore, initial catalytic tests are

carried out in order to assure that all following investigations are undertaken in the kinetic

regime of the reaction.

To better understand and optimize the reactivity of current benchmark Lewis acid doped

zeolites, advanced characterization tools are used to build a detailed understanding of the

structural organization of the catalyst, and probe reactions are employed to link the obtained

structural information with the reactivity of the materials. However, the structural complexity

of zeolites complicates unequivocal distinction between the diverse material characteristics

that add to the performance of the catalyst and hence to rationally improve the catalyst.

Thus, in Chapter 4, the potential of a catalytic model-system is assessed in order to provide

complementary information to studies based on the direct investigation of the zeolitic materials.

This is approached by the preparation of silica-supported SnIV-sites, deploying chemical vapor

deposition as immobilization technique and step-wise thermal and chemical functionalizations.

A set of characterization techniques and catalytic tests are brought in to follow structural

modifications and alterations in activity, respectively. This in combination with the final

comparison to benchmark Snβ allows conclusions about the different contributions to reactivity

in SnIV/SiO2-based catalysts.

Introduction 13

In Chapter 4 we learn that confinement effects influence the activity of SnIV/SiO2-

based catalysts in the Meerwein-Ponndorf-Verley reaction of cyclohexanone with 2-butanol

significantly. Based on these findings, the next step encompasses targeted experimental

approaches in order to confirm the existence of this effect. Hence, in Chapter 5 we extend our

previous model-studies to SnIV-sites supported on mesoporous silica (MCM-41), representing a

pore size range that stands in between the micropores of β-zeolite and the external surface of

amorphous silica, which we used in Chapter 4.

Chapter 6 summarizes the key results of the research performed throughout this thesis

and identifies challenges ahead.

Chapter 2 – 4 of this thesis were written based on one publication and can be read

independently. Accordingly, some overlap between the chapter introductions occurs.

Chapter 2

Simple and Scalable Preparation of HighlyActive Lewis Acidic Snβ

In this chapter FT-IR and UV-Vis measurements, as well as the triose isomerization reactions

were performed by the author. Material synthesis, Raman measurements and the Baeyer-

Villiger reactions were performed by Ceri Hammond.

2.1. Introduction

Lewis acids are a versatile class of catalysts that exhibit remarkable activity for a number of

essential transformations, including oxidations and isomerizations.[16] Although homogeneous

analogues are well-established, heterogeneous catalysts offer several advantages for the

development of more sustainable technologies in terms of facile downstream processing and

process intensification. Of particular interest are Lewis acid-doped zeolites, some of which

exhibit remarkable activity, selectivity and lifetime for a number of processes.[39,52,128] The

development of TS-1 (a TiIV-doped MFI-type zeolite) is for instance viewed as one of the

greatest breakthroughs in sustainable chemistry over the recent decades, having resulted in a

"greener" process for the epoxidation of propylene,[20] amongst others. Promising results have

also been obtained in the development of SnIV-doped zeolite β, which has shown unparalleled

activity and selectivity for the isomerization of glucose to fructose and the Baeyer-Villiger

oxidation of ketones to lactones using H2O2 as green oxidant.[20,27,48,115]

Lewis acid-doped zeolites are typically obtained by direct framework incorporation during

hydrothermal synthesis.[20,27,39,48,52,115,128] It remains, however, a challenge to obtain a significant

amount of isolated sites within the structures without the undesirable formation of metal

oxide particles which are significantly less active. Moreover, even under optimized conditions,

the incorporation of large Lewis acidic centres, such as SnIV, typically leads to a significant

retardation of the zeolite nucleation, and hence long synthesis timescales (up to 40 days). This

16 Chapter 2

results in unfavourable large crystals. In order to facilitate the crystallisation of such materials,

additives such as HF are commonly added to the synthesis gel, posing additional practical

and environmental limitations. This, in combination with the limited amount of active metal

that can be incorporated into the structures (viz., < 2 wt%) currently limits the large scale

applicability of these otherwise promising materials.

With these limitations in mind, we aimed to develop a convenient post-synthetic route for

the incorporation of various Lewis acid centres into zeolitic frameworks. An attractive route

involves the incorporation of the desired transition metal ions into the vacant tetrahedral (T)-

sites of a pre-dealuminated zeolite. Not only does this avoid the long synthesis times associated

with the conventional hydrothermal synthesis routes, but it also allows for the synthesis of a

material with significantly smaller crystallite sizes than possible through direct synthesis. As

a proof of concept, this communication focusses on the preparation of Snβ. This material

has tremendous potential,[27,48,115] equivalent to TS-1, but the industrial implementation is

hampered by the tedious synthesis procedure.

2.2. Experimental

2.2.1. Material Synthesis

Commercial Zeolite H-β (ZeoChem) was dealuminated by treatment in HNO3 solution (13 M)

at 100 °C for 20 h (20 mL g−1zeolite). SSIE was performed by grinding the appropriate amount

of Sn(II)acetate with the required amount of dealuminated zeolite for 15 min. Samples were

calcined in an air flow at 550 °C.

Conventional Snβ was synthesized according to the original procedure of Corma et al.[48]

TEOS was added to a TEAOH solution under stirring. After a single phase was obtained, the

desired amount of Sn (SnCl4 · 5 H2O, 98 %, STREM) dissolved in H2O was added drop-wise.

The solution was stirred open to evaporate ethanol and water until a viscous gel was obtained.

Addition of hydrofluoric acid resulted in a solid gel with the molar composition 1.0 SiO2 : 0.01

SnCl4 : 0.55 TEAOH : 0.55 HF : 7.5 H2O. Lastly, a solution of dealuminated Beta seeds in water

was added and the mixture homogenized with a Teflon spatula. Crystallization was carried out

at 140 °C in 45 mL scale teflon-lined stainless-steel autoclaves, tumbled at 60 rpm for 14 days.

Simple and Scalable Preparation of Highly Active Lewis Acidic Snβ 17

After cooling down, the samples were filtered and subsequently washed with deionized water

and acetone before drying them in an oven at 110 °C over night. To remove the structure

directing agent, samples were calcined at 580 °C under a steady air flow for 6h.

2.2.2. Characterization Methods

The Sn-content was quantified with ICP-OES (Ultima 2 from Horiba Jobin Yvon) after digestion

of the samples with HF.

FT-IR spectroscopy was performed on a self-supporting wafer using a Bruker Alpha

spectrometer in transmission mode (resolution of 2 cm-1). Intensities were normalized to the

Si-O-Si overtones of the silica framework. Diffuse Reflectance UV-Vis spectra were recorded

with a Maya 200 spectrometer (Ocean Optics) equipped with a UV-Vis deuterium/halogen

light source (DH-2000-BAL from Mikropack) using BaSO4 as background. Both FT-IR and

UV-Vis analysis was carried out inside a glove box (< 1 ppm H2O and O2).

Porosimetry measurements were performed on a Micromimetics 2000 apparatus. The

samples were degassed prior to use (275 °C, 3 h). Adsorption isotherms were obtained at

77 K and analyzed using BET and t-plot methods.

2.2.3. Catalytic Experiments

Baeyer-Villiger oxidation of cyclohexanone was carried out in a 50 mL round bottomed flask

equipped with a reflux condenser. The vessel was charged with the reactant solution (5 mL,

0.33 M cyclohexanone in 1,4-dioxane) and the desired amount of catalyst (corresponding to

1 mol% Sn relative to ketone). The vessel was heated to the reaction temperature (90 °C)

for 15 min, prior to the addition of H2O2 at a final concentration of 0.5 M (H2O2/ketone =

1.5) and stirred vigorously for the required reaction period. Samples were taken periodically

and quantified by GC-FID against a biphenyl internal standard (30 m FFAP column). H2O2

concentrations were determined by titration against Ce4+. The conversion of dihydroxyacetone

(DHA) to ethyl lactate was performed in an autoclave reactor at 10 bar N2 pressure. The vessel

was charged with 25 mL of DHA solution (0.4 M in EtOH plus 0.3 mol% Sn, relative to DHA),

and the reaction performed at 100 °C. The reactant and product were quantified against a

biphenyl internal standard by means of GC-FID (30 m FFAP column) and HPLC (nucleodur

18 Chapter 2

100-5 NH2-RP column).

2.3. Incorporation of SnIV into Dealuminated β Zeolite

Preliminary work focussed on the efficient dealumination of a parent Al-β zeolite. Although

steaming is a well-known method for removing framework Al3+, it has the disadvantage

of leaving behind ill-defined extra-framework Lewis acidic Al3+ species, which could affect

the catalytic performance of the material. In view of this, an acidic pre-treatment with

HNO3 (13 M, 100 °C, 20 h, 20 mL g-1) was performed in order to extract and remove Al

quantitatively.[129] The results in Table 2.1 show indeed that nearly all Al can be removed

without destruction of the BEA framework, or significant alterations to its textural properties.

The removal of framework Al is exemplified by the loss of the Brønsted acidity, i.e., the sharp IR

signal at 3610 cm-1 (Figure 2.2 A). In its place, a broad absorbance around 3500 cm-1 appears,

confirming the successful formation of silanol nests and vacant T-sites for the incorporation of

SnIV (Figure 2.1 and 2.2 B).[130]

Subsequently, methods of introducing SnIV into the vacant T-sites were considered.

Although liquid-phase routes (e.g., impregnation) offer a convenient and scalable route for

the deposition of Sn, the solvation shell surrounding the Sn cations could potentially lead

to significant diffusion limitations within the zeolite micropores, negatively affecting the

incorporation of tin and resulting in poor dispersion. Furthermore, hydrolysis of the Sn-

precursor may also result in the partial formation of bulk oxide species in the final material.

Although traditional gas-solid deposition (e.g., chemical vapour deposition) offers exciting

possibilities of incorporating ‘naked’ metals into the T-sites, the low volatility of typical Sn-

Figure 2.1. Post-synthetic synthesis route to solid Lewis acidic Snβ zeolite. In a first stepcommercially available β-zeolite is dealuminated in an acidic treatment, in a second step SnIV isincorporated into the vacant silanols nests created during the dealumination by solid-state ion-exchange (SSIE).


Table 2.1. Physicochemical properties of the materials.

Entry Catalyst Treatment SBETa Vmicro

b SiO2/Al2O3c SiO2/SnO2

c

[m2 g−1] [cm3 g−1]

1 H-βd – 600 0.17 25 –2 deAl-βe H+ 620 0.18 > 1900 –3 Sn/deAl-βf H+/SSIE 610 0.17 > 1900 324 Sn/deAlβf H+/SSIE 600 0.17 > 1900 16a Brunauer-Emmett-Teller surface area.[131] b Micropore volume, t-plot method.[132]

c Molar ratio in solid, determined by AAS. d Commcercial H-β zeolite. e Dealuminatedβ zeolite. f Snβ zeolite.

precursors (e.g., SnCl4) results in difficulties grafting sufficient quantities of Sn into the final

material. Moreover, previous work has demonstrated that a multitude of Sn-species are formed

by SnCl4 grafting, the majority of which appear to be extra-framework species or bulk tin

oxides.[70] The catalytic activity of such materials is therefore rather poor.

Solid-state ion-exchange (SSIE) was therefore considered to be an interesting route for

incorporating the desired amount of metal into the structure, with both high dispersion and

homogeneity. The procedure simply involves mechanical grinding of dealuminated zeolite and

the appropriate precursor, in this case Sn(II)acetate, prior to calcination at 550 °C for removal

of the residual organic species.

Figure 2.2. IR spectra of (A) H-β, (B) dealuminated β (deAlβ), (C) 5 wt% Sn/deAlβ, and (D)10 wt% Sn/deAlβ (see Table 2.1).

20 Chapter 2

wavenumbers / cm-1

(1)

(2)

(3)

(4)

SnO2

Figure 2.3. UV-Vis spectra of (A) conventional Snβ (1.5 wt%), (B) 10 wt% Sn/deAlβ, (C) 10 wt%Sn/(O2)/deAlβ. The inset shows Raman spectra of (1) deAlβ, (2) 5 wt.% Sn/deAlβ, (3) 10 wt%Sn/deAlβ, (4) bulk SnO2.

2.4. Characterization of the Prepared Materials

A combination of spectroscopic techniques was used to determine the nature of the post-

synthetically prepared Sn-species. The incorporation of Sn into the dealuminated framework

can be observed through closure of the silanol nests by IR spectroscopy (Figure 2.2, plot C

and D). Diffuse reflectance spectroscopy also reveals a sharp UV-absorbance around 216 nm,

even for very high loadings (Figure 2.3 B); this signal is characteristic of isolated, tetrahedral

SnIV-species within the zeolite framework. No SnO2 could be detected with UV-Vis or Raman

spectroscopy (Figure 2.3). It should be noted that the UV-Vis spectra of the post-synthetically

prepared samples are almost identical to that of a reference Snβ sample prepared by direct

hydrothermal synthesis (Figure 2.3 A). This spectroscopic data indicates that very similar

SnIV-sites are obtained even at ca. eight-times the total loading.

2.5. Catalytic Performance of Post-Synthetically Prepared Snβ

Zeolite

The catalytic efficiency of the prepared materials was evaluated for the Baeyer-Villiger oxidation

of cyclohexanone with H2O2 (Figure 1.1 B and Table 2.2). As can be seen, SSIE of SnIV into

dealuminated β (Table 2.2, Entries 1 and 2) results in the formation of an active catalyst,

comparable in both yield and TON to that previously reported for hydrothermally-synthesized


Table 2.2. Catalytic Activity of various Sn-containingsamples for the Baeyer-Villiger oxidation of cyclohexanonewith aqueous H2O2.a

Entry Catalyst Conversion Selectivity Yield[%] [%] [%]

1 5 wt% Sn/deAlβ 37 75 262 10 wt% Sn/deAlβ 42 93 383 10 wt% Sn(O2)/deAlβ 0 – 04 10 wt% Sn/β 0 – 0a Experimental conditions: cyclohexanone in 1,4-dioxane (0.33 M),H2O2 (30 wt% solution, H2O2/ketone = 1.5), Sn content (relativeto ketone) = 1 mol%, 90 °C, 4 h.

materials.[48] Interestingly, the 5 and 10 wt% Sn/deAlβ show comparable activity per SnIV-

site, but lead to a different lactone selectivity. We attribute this to the presence of un-closed

silanol nests in the 5 wt% sample, resulting in the presence of residual Brønsted acid sites.

Such sites are capable of catalyzing the formation of 6-hydroxycaproic acid via acid-catalyzed

hydrolysis, as confirmed by lactone stability studies. At higher Sn loadings, a larger fraction

of the silanol nests are closed, thus decreasing the number of available Brønsted acid sites,

resulting in increased selectivity.

To substantiate the hypothesis that isolated, tetrahedral SnIV-species are indeed the active

sites formed in the post-synthetic material, we verified that dispersion of SnO2 into/onto the

dealuminated framework (Table 2.2, Entry 3) leads to zero activity. Similarly, the dispersion of

Sn(II)acetate into/onto a non-dealuminated framework (Table 2.2, Entry 4) leads to an inactive

material; clearly, when the formation of framework SnIV-sites is prohibited (either due to an

inappropriate precursor or the lack of vacant framework sites), an inactive catalyst is obtained.

In order to confirm the heterogeneous nature of the catalytic reaction, a hot-filtration test

was performed,[133] demonstrating that removal of the catalyst after 1 hour (i.e., after ca. 15 %

conversion) leads to complete termination of the reaction (Figure 2.4 B).

Comparing the SSIE synthesized material with the hydrothermally-made material, it is

clear that along with proceeding at comparable levels of lactone and H2O2-based selectivity

(Table 2.3), the Sn/deAlβ is significantly more productive, given the ca. three-fold increase

in space-time-yield. This is predominantly due to the incorporation of such high quantities

of SnIV, without causing significant unwanted side effects such as agglomeration, increased

22 Chapter 2

Figure 2.4. (A) Time-dependent formation of ε-caprolactone in the presence of 10 wt% Sn/deAlβ,and (B) hot-filtration test for the catalyst (conditions, see Table 2.2).

crystallite sizes and leaching. This improved reaction efficiency, in combination with the facile

and scalable synthesis procedure is expected to catalyze the industrial applicability of Snβ.

To explore the general applicability of the post-synthetic material, we studied the conversion

of the triose sugar dihydroxyacetone to ethyl lactate (Figure 2.5), a convenient building

block towards biorenewable and biodegradable solvents and polymers.[30] Recently, it has been

reported that SnIV-containing compounds are amongst the most active and selective materials

for these reactions,[59] and that a combination of Lewis acidity and mild Brønsted acidity is

advantageous for this reaction.[32] As can be observed from the results in Table 2.4, along with

proceeding at exceptional levels of selectivity, the 10 wt% Sn/deAlβ catalyst is significantly

more productive than the benchmark materials in terms of turnover numbers and space-time-

yield.

2.6. Conclusions

In conclusion, a convenient route for the preparation of Lewis acidic Snβ has been developed.

In addition to exhibiting comparable or higher levels of catalytic activity and selectivity to the

state-of-the-art materials, significantly higher space-time-yields can be obtained through the

Figure 2.5. Additionally applied probe reaction to test the catalytic activities of the post-synthetically prepared Snβ catalysts: Conversion of dihydroxyacetone to ethyl lactate.


Table 2.3. Comparison of post-synthetic and hydrothermal routes.a

Entry Catalyst Selectivity(lactone)b Selectivity(H2O2)c STYd Ref.[%] [%] [g kg-1 h-1]

1 1.5 wt% Snβ > 98 > 95 374 [48]2 10 wt% Sn/deAlβ 93 > 95 1075 this worka Comparison under identical experimental conditions as previously described: cyclohexanone in1,4-dioxane (0.33 M), H2O2 (30 wt% solution, H2O2/ketone = 1.5), Sn content (relative to ketone)= 1 mol%, 90 °C, 4 h. b Calculated as mole ε-caprolactone formed per mole cyclohexanoneconverted. c Calculated as mole ε-caprolactone formed per mole H2O2 converted. d Calculatedas glactone kg−1

zeolite h-1.

Table 2.4. Conversion of dihydroxyacetone to ethyl lactate.a

Entry Catalyst Selectivity(lactate)b TONc STYd Ref.[%] [–] [g kg-1 h-1]

1 3.9 wt% Sn/MCM-41 98 30 195 [59]2 1.6 wt% Snβ > 99 120 70 [134]3 Sn-carbon-silica 100 350 200 [32]4 10 wt% Sn/deAlβ > 99 250 1050 this worka Experimental conditions: dihydroxyacetone in ethanol (0.4 M), Sn content (relative todihydroxyacetone) = 0.3 mol%, 100 °C, 10 bar N2, 24 h. b Calculated as mole ethyllactate formed per mole dihydroxyacetone converted. c Calcuated as mole ethyl lactateformed per mole of Sn. d Calculated as glactone kg−1

zeolite h-1.

preparation of a high metal-content material. Furthermore, the developed procedure requires

significantly less synthesis time and produces no toxic waste in comparison to the benchmarked

process. We expect such a straightforward approach to facilitate the utilization of the same or

similar materials on a large scale.

Chapter 3

Insights into the Baeyer-Villiger Oxidation ofCyclohexanone with H2O2 catalyzed by Snβ

In this chapter FT-IR, UV-Vis and BET measurements, as well as the catalytic tests to study

the influence of water and lactone concentration and the effect of hydrophilicity on Snβ were

performed by the author. Material synthesis and water adsorption studies were performed by

Patrick Wolf, the initial kinetic studies were performed by Hailey Orsted.

3.1. Introduction

The Baeyer-Villiger (BV) oxidation of ketones to esters or lactones is of high synthetic utility

in the pharmaceuticals and fine chemicals industry.[22,23,135,136] So far, mostly peracids such as

3-chloroperbenzoic acid (m-CPBA) have been used as oxidizing agent, complicating the general

applicability of this process through the formation of undesired by-product (carboxylic acid)

and the high explosiveness of the oxidant itself (Figure 3.1).[25,137] In an alternative approach,

hydrogen peroxide (H2O2) is utilized as oxidant, which imposes less environmental (water is

the only by-product) and less safety issues (Figure 3.1).[24,138–141]

Figure 3.1. Baeyer-Billiger oxidation of cyclohexanone to ε-caprolactone. The traditional route(top), using a percarboxylic acid as oxidizing agent, leads to the production of one equivalent oforganic acid as by-product. Alternative routes (bottom), using H2O2 as "green" oxidant, have wateras the only by-product.

Since H2O2 is kinetically inert, it requires catalytic activation in order to perform

as required. Initially, homogeneous transition metal complexes,[142–146] Brønsted acid

26 Chapter 3

catalysts[147–152] and Ti-silicalites[153] have been reported to catalyze BV oxidations by activating

H2O2. However, these catalysts showed low selectivities due to product hydrolysis and/or

preferred epoxidation reactions if additional functional groups are present. Corma and co-

workers showed that SnIV-sites incorporated into the framework of β-zeolite (Snβ) show hitherto

unrivalled catalytic activities and selectivities for the BV oxidation of various ketones and

aldehydes with H2O2.[48,115,122,154]

In order to explain the high chemoselectivity of Snβ, it was suggested that unlike previous

catalysts, which activate H2O2, Snβ activates the carbonyl substrate, hereby making it more

reactive toward an attack by the peroxide.[48] Indeed, IR adsorption studies have demonstrated

that the carbonyl oxygen of cyclohexanone interacts with the SnIV-centers incorporated in the

β-zeolite framework, based on an observed shift in the carbonyl IR stretch when cyclohexanone

is adsorbed on to Snβ.[48,108,122]

The selective interaction between Lewis acidic SnIV-centers and Lewis basic carbonyl groups

give Snβ its unprecedented chemoselectivity in BV oxidations with H2O2. However, it also

exposes Snβ to interfering interactions with other basic molecules that are present in the

reaction mixture. As such, there is the product carbonyl (a lactone or an ester), which can be

expected to compete with cyclohexanone for coordination to the SnIV-centers, thus restricting

the conversion of substrate.[122] In addition, the utilization of H2O2 as oxidant poses the difficulty

of water as co-solvent (and product). Water may, similar to the product carbonyl, hinder the

substrate to effectively interact with the active SnIV-sites and/or the peroxide, hereby restricting

the performance of the catalyst.[17,122,155]

Recently, it was shown that Snβ catalysts which are prepared following different synthesis

protocols, have different capacities to adsorb water (i.e., hydrophilicities), resulting from

different amounts of framework silanols.[156,157] The possibility to synthesize the parent Al-

β-zeolite in various media (fluoride or hydroxyl) prior to dealumination and post-synthetic Sn

incorporation indeed allows controlling the concentration of silanols in Snβ catalysts. It has

not yet been studied if such changes in hydrophilicity influence the catalytic performance of

Snβ in the BV oxidation with aqueous H2O2.

In this work, we look at the influence of the presence of water and ε-caprolactone on the

Insights into the Baeyer-Villiger Oxidation of Cyclohexanone with H2O2 catalyzed by Snβ 27

activity of Snβ in the BV oxidation of cyclohexanone with H2O2. We prepare a series of Snβ

catalysts with different hydrophilicities, both by traditional hydrothermal synthesis and by post-

synthetic incorporation of SnIV into dealuminated β-zeolites. We quantify the hydrophilicities

of the catalysts by recording water adsorption isotherms as well as by analyzing the metal

hydroxyl IR region of the materials. Finally, we test the obtained catalysts in the BV oxidation

of cyclohexanone with H2O2 in order to investigate the influence of catalyst hydrophilicity on

the efficiency of Snβ in this reaction.

3.2. Experimental


Al-β in fluoride media was prepared via hydrothermal synthesis according to a literature

procedure.[158] First Aluminum powder (99.99 %, Acros) was dissolved in an aqueous solution of

tetraethyl ammonium hydroxide (TEAOH; 35 %, SACHEM). After complete dissolution of the

aluminum the solution was added to a tetraethyl orthosilicate (TEOS; 98 %, Sigma-Aldrich)

plus TEAOH solution. The resulting mixture is stirred until complete evaporation of ethanol,

formed upon TEOS hydrolysis. To the resulting viscous gel, hydrofluoric acid (48 %, Sigma-

Aldrich) was added to result in a gel with the following composition: 1 SiO2 : x Al2O3 : (0.54

+ 2x) TEAOH : (0.54 + 2x) HF : (7 + 2x) H2O. Crystallization was carried out at 140 °C in

45 mL scale teflon-lined stainless-steel autoclaves, tumbled at 60 rpm for 7 days. After cooling

down, the samples were filtered and subsequently washed with deionized water and acetone

before drying them in an oven at 110 °C over night. To remove the structure directing agent,

samples were calcined at 580 °C under a steady air flow for 6 h.

Post-synthetic incorporation of Sn was performed with an improved synthesis method

originating from the procedure that we described in Chapter 2.[159] Dealumination of the

parent Al-β zeolites was done by acid leaching (13 M HNO3, 20 mL g-1, 100 °C, 20 h).

Sn was incorporated via solid-solid ion-exchange by grinding the dealuminated β with the

appropriate amount of the Sn(II) acetate precursor (Sigma Aldrich) followed by subsequent

3 h heat treatments under N2 and air at 550 °C. 10SnβOH25, 1SnβOH25 and 1SnβOH300

represent the post-synthetic materials starting from commercial Al-β zeolite (SiO2/Al2O3 =

28 Chapter 3

25, Zeochem and SiO2/Al2O3 = 300, Zeolyst) with 10 and 1 wt% Sn, respectively. 1SnβF30,

1SnβF200 and 1SnβF400 (SiO2/Al2O3 = 30, 200, 400) are synthesized from Al-β synthesized

in fluoride media vide supra.

Snβ zeolite via direct incorporation of Sn during hydrothermal synthesis was synthesized

as described in Chapter 2.


The Sn-content was quantified with ICP-OES (Perkin Elmer Optima 2000) after digestion of

the samples with HF (48 %, Sigma-Aldrich).

FT-IR and UV-Vis analysis was performed as described in Chapter 2.

Water sorption experiments were performed on a Micromeritics 3Flex instrument at 298 K.

Water was purified by three freeze and thaw cycles. Microporous water uptake was determined

at the relative pressure p/p0 that showed complete filling of the micropores in the N2 adsorption.

N2 sorption measurements were performed on a Micromeritics 3Flex apparatus at 77 K. Samples

were degassed under vacuum at 350 °C for 3 h prior to every sorption analysis. The surface

area was calculated using the Brunauer-Emmett-Teller (BET) theory.


The experiments were carried out in 10 mL thick wall tube reactors capped with a

PTFE/silicone seal, capable of holding 15 bar over pressure. For reactions under standard

conditions, the vessel was charged with the reactant solution (4 mL, 0.33 M cyclohexanone

in 1,4-dioxane) and the desired amount of catalyst. The vessel was heated to the required

reaction temperature for 10 min prior to the addition of H2O2 (H2O2/ketone = 1.5) and stirred

vigorously for the required reaction period. Aliquots were taken periodically and quantified

against the internal standard decane by GC-FID (30 m FFAP column).


3.3. Results and Discussion

3.3.1. Initial Kinetic Studies

Initial catalytic tests focused on the determination of the kinetic regime for the BV oxidation of

cyclohexanone with H2O2. For this we used a 10 wt% Snβ catalyst prepared via post-synthetic

incorporation of SnIV into dealuminated β-zeolite with a SiO2 to Al2O2 ratio of 25 (material

denoted as 10Snβ-OH25). We performed the reactions under similar conditions to those that

we previously employed (see Chapter 2), using 50 instead of 30 wt% aqueous H2O2 in order to

minimize the concentration of water in the reaction mixture. We studied initial reaction rates

as a function of temperature and catalyst concentration to test for diffusion limitations.

kinetic regime EA = 13.9 kcal/mol

diffusion limitations EA = 5.2 kcal/mol

65°C

45°C

90°C

Figure 3.2. Arrhenius plot of the BV oxidation of cyclohexanone with aqueous H2O2 in thetemperature range between 45 and 90 °C. Experimental conditions: cyclohexanone in 1,4-dioxane(0.33 M), H2O2 (50 wt% aq. solution, H2O2/ketone = 1.5), 10Snβ-OH25.

Figure 3.2 shows initial reaction rates as a function of temperature (Arrhenius plot) in

the range from 45 to 90 °C. The plot shows a curved shape, pointing toward mass-transfer

limitations at higher reaction temperature. An activation energy of ca. 14 kcal mol-1 proofs that

the system is under chemical control between 45 and 65 °C. This value is in good agreement with

computational predictions by Corma and co-workers, which resulted in an activation energy of

14.8 kcal mol-1 for hydrothermally synthesized Snβ.[122] Increasing the temperature to about

90 °C, leads to a drop in activation energy to ca. 5 kcal mol-1, similar to the experimentally

determined activation energy of 7.2 kcal mol-1 by Sels and co-workers for post-synthetically

30 Chapter 3

synthesized Snβ in a similar temperature range.[104] Although Sels and co-workers exclude

mass-transfer limitations, our data clearly suggests the presence of diffusion limitations at

temperatures above 65 °C and a Sn loading of 1.0 mol% (relative to cyclohexanone).

As can be seen in Figure 3.3 (80 °C), the oxidation reaction is first order in Snβ up to

a Sn loading of around 0.6 mol% (relative to ketone), after which the reaction rate levels off.

This observation indicates mass transfer limitations at higher catalyst loadings. In order to

exclude that the observed leveling-off is caused by a change in reaction order, going from low

to high catalyst loadings, we performed a similar study at a lower temperature. Figure 3.3

(60 °C) shows that the leveling off of the initial reaction rates is less pronounced than at 80 °C

suggesting that at Sn loadings higher than 0.6 mol% diffusion limitations are present. Despite

of the clear indication for diffusion limitations, it is not possible to distinguish between extra-

and intragranular mass transfer limitations based on our observations.

mass-transfer limitations

first order in Snβ

Figure 3.3. Initial reaction rates as a function of the employed Sn loading (relative to ketone) forthe BV oxidation of cyclohexanone with aqueous H2O2 at 80 and 60 °C. Experimental conditions seeFigure 3.2.

Diffusion limitations are typical for heterogeneously catalyzed reactions.[160,161] However,

their occurrence has barely been considered when trying to optimize the BV oxidation with

Snβ. This lack of attention is surprising since it is well established that mass-transfer limitations

may cause a lowering of the selectivity for exothermic reactions. In order to avoid diffusion

limitations in our following tests, we hence employed a Sn loading of 0.5 mol% at a reaction

temperature of 80 °C.


3.3.2. Influence of Water and Lactone Concentration

To study the impact of water on the performance of Snβ, we performed reactions with different

initial water concentrations in the reaction solution. For this purpose, different amounts of

concentrated aqueous H2O2 (10, 30 and 50 wt% H2O2) were added to the reaction solution,

while keeping the final peroxide concentration constant. Additional data points were obtained

by adding pure water to the standard amount of 50 wt% aqueous H2O2.

As depicted in Figure 3.4, the performance of Snβ decreases with increasing amounts of

water in the reaction mixture, as evidenced by a gradual decline in lactone formation (red

arrow indicates increasing water content). Initial reaction rates plotted as a function of the

water concentration (inset in Figure 3.4) show an exponential decrease with increasing water

concentration. Interestingly, selectivity values at same levels of ketone conversion are similar

for the different reactions, which indicates that the loss in activity does not arise from enhanced

hydrolysis of the lactone (to hydroxycaproic acid), but rather from catalyst inhibition through

active site blocking.

The best performance is reached if 50 wt% aqueous H2O2 is used, given by the lowest amount

of water introduced to the system, which minimizes catalyst inhibition. This observation is of

practical relevance since it suggests the usage of the highest possible concentration of aqueous

H2O2 for BV oxidations with Lewis acid catalysts.

Figure 3.4. ε-Caprolcatone yield over time for different initial water concentrations in the BVoxidation of cyclohexanone with aqueous H2O2 at 80 °C. Experimental conditions see Figure 3.2.

32 Chapter 3

Figure 3.5. Initial reaction rates as a function of the initial lactone and water concentration in thereaction solution for the BV oxidation of cyclohexanone with aqueous H2O2 at 80 °C. Experimentalconditions see Figure 3.2.

Analogous reactions with different initial lactone concentrations were performed by adding

different amounts of ε-caprolactone shortly before initiating the reaction through H2O2

addition. A decline in the formation of lactone was observed (Figure A.1). In addition, similarly

to the influence of water, an exponential decrease of the initial reaction rate with increasing

lactone concentration was noticed (Figure 3.5, top curve). However, the drop in activity based

on the presence of lactone compared to water is not as strong (Figure 3.5), which indicates that

catalyst inhibition through the lactone is less detrimental than through water.

3.3.3. Effect of Hydrophilicity on the Catalytic Efficiency of Snβ

From our preliminary catalytic tests, we learned that an increase in both water and lactone

concentration in the reaction mixture negatively impacts the performance of Snβ. An open

question in this context is if the catalyst hydrophilicity (i.e., the silanol density) may influence

the catalyst efficiency in BV oxidations with H2O2. A less hydrophilic zeolite surface might

indeed reduce the affinity of water and product carbonyl to the zeolite, and hereby lower the

local concentration of water and lactone at the active SnIV-sites. In general, we expect a

preferential adsorption of our substrate carbonyl (cyclohexanone) compared to our product

carbonyl (ε-caprolactone) due to the higher polarity of ketone- compared to ester-compounds.

To investigate the impact of surface polarity on the activity of Snβ in BV oxidations more


closely, we prepared a series of Snβ catalysts with different hydrophilicities by following various

synthesis protocols. We quantified the hydrophilicities of the obtained materials by recording

water adsorption isotherms. We also analyzed the metal hydroxyl IR regions of the catalysts

and utilized the peak integrals of the SiO–H signal to quantify the silanol groups of the zeolite.

Lastly, we tested their catalytic activities in the BV oxidation of cyclohexanone with H2O2,

using our optimized experimental conditions (0.5 mol% Sn; 50 wt% aq. H2O2).

The Snβ catalysts with a Sn loading of 1.0 ± 0.2 wt% were prepared by both post-synthetic

incorporation of the metal into dealuminated β-zeolites (five materials) and by traditional

hydrothermal synthesis (one material, denoted as 1Snβ-HT). The post-synthetic samples were

prepared from commercial Al-β zeolites that were synthesized in hydroxide media (SiO2/Al2O2

ratios of 25 and 300; materials denoted as 1Snβ-OH25 and 1Snβ-OH300) and from Al-β zeolites

that were synthesized in fluoride media (SiO2/Al2O2 ratios of 30, 200 and 400; materials denoted

as 1Snβ-F30, 1Snβ-F200 and 1Snβ-F400).

In a next step, we recorded water adsorption isotherms in order to quantify the

hydrophilicities of the Snβ catalysts. Table 3.1 summarizes the total uptake of water per surface

area for the different catalysts. As some of us reported recently, the uptake of water is biggest

for the catalysts prepared from Al-β zeolites synthesized in hydroxide media (1Snβ-OH25

and 1Snβ-OH300), compared to those prepared from zeolites synthesized in fluoride media

(1Snβ-F30, 1Snβ-F200, 1Snβ-F400) and the Snβ from direct hydrothermal Sn incorporation

(1Snβ-HT).[157] This is not surprising, since it is known that zeolites prepared in hydroxide

media have more defect sites, leading to a higher intrinsic hydrophilicity.[158]

Another observation we make is that the hydrophilicity decreases with increasing SiO2

to Al2O3 ratio for the materials prepared from Al-β zeolites in fluoride media. This is most

likely caused by different amounts of residual silanol nests, which are not being filled with Sn

during metal incorporation. For 1Snβ-F400, for instance, almost all the vacant T-sites ought

to be filled with SnIV, resulting in a low hydrophilicity. For 1Snβ-F30, however, only around

10 % of the silanol nests created during dealumination are being occupied, leading to a higher

hydrophilicity compared to 1Snβ-F400. The very similar water adsorption for 1Snβ-OH25 and

1Snβ-OH300 indicates that the defect sites of these materials constitute the majority of silanols.

34 Chapter 3

Table 3.1. Water adsorption results and catalytic activities ofdifferent Snβ catalysts in the BV oxidation of cyclohexanone withaqueous H2O2.a

Entry Catalyst H2Ob,cads TOFd,e Conversionf Selectivityf

[µmol m-2] [h-1] [%] [%]

1 1Snβ-F30 11.0 141 ± 21 50 852 1Snβ-F200 3.98 112 ± 17 52 703 1Snβ-F400 4.15 106 ± 16 48 754 1Snβ-HT 2.73 18 ± 3 13 855 1Snβ-OH25 19.5 128 ± 19 41 856 1Snβ-OH300 15.9 99 ± 15 42 84a Experimental conditions: cyclohexanone in 1,4-dioxane (0.33 M), H2O2

(50 wt% aq. solution, H2O2/ketone = 1.5), Sn content (relative to ketone)= 0.5 mol%, 80 °C, 4 h. b Quantity of water adsorbed at p/p0 = 0.82,normalized by the BET surface areas of the catalysts.[162] c See Table A.1for BET surface areas. d Defined as mole ε-caprolactone produced per moleSn per hour at the initial stage of the reaction. e Error based on ICP-OESdetermined Sn-loading. f After 2.5 h of reaction.

In addition to our water adsorption measurements, we determined the peak areas of the

silanol IR signals of the different catalysts in order to quantify their hydrophilicities. Figure 3.6

B shows the metal hydroxyl IR region of all our prepared Snβ catalysts (see Figure A.2 for

full spectra). The broad absorbance around 3500 cm-1 arises from unclosed silanol nests, the

small signal at ca. 3660 cm-1 from Sn hydroxyl groups, and the sharp peak at ca. 3750 cm-1

from isolated silanols. The Sn hydroxyl vibration is only detected for 1HT-Snβ, 1Snβ-F400

and 1Snβ-F200, while it is overlapped by more intense silanol signals for the other catalysts.

As can be seen in Figure 3.6 A, the quantities of adsorbed water, as determined by our water

adsorption measurements, show a linear correlation with the peak areas of the silanol IR signals.

This indicates that the hydrophilicity of Snβ correlates with the amount of framework silanols.

To link the results from our water adsorption studies and IR analysis with the activity of

the catalysts, we performed BV oxidations of cyclohexanone with H2O2. As can be seen in

Table 3.1, the different Snβ catalysts show different activities. The highest turnover frequency

(TOF) is obtained with 1Snβ-F30, closely followed by 1Snβ-OH25. The lowest activities are

observed for 1Snβ-OH300, which has still a ca. four times higher activity compared to 1Snβ-

HT.

Figure 3.7 shows the TOFs of the different Snβ catalysts as a function of their


(A) (B)

Figure 3.6. (A) Quantities of adsorbed water obtained from water adsorption measurements plottedagainst the peak areas of the SiO–H IR signal between 3750 and 3000 cm-1 (both values are normalizedby the BET surface areas of the catalysts). (B) Metal hydroxyl region taken from FT-IR spectra ofthe different Snβ catalysts (see Figure A.2 for full IR spectra).

hydrothermal

fluoride media post-synthetic hydroxide media

post-synthetic

Figure 3.7. Catalytic activities of the different Snβ catalysts as a function of their hydrophilicities,as determined by the peak areas of the SiOH IR signals between 3750 and 3000 cm-1 (see Table 3.1and Figure 3.6 B).

hydrophilicities (as given by the peak areas from their metal hydroxyl IR regions). It can

be seen, that the activities of the Snβ catalysts increase with increasing hydrophilicity up to a

certain level, after which they decrease again. For the catalysts prepared from parent zeolites

synthesized in fluoride media, we observe an increase in activity with increasing hydrophilicity.

For the catalysts prepared from zeolites synthesized in hydroxide media we observe, however,

36 Chapter 3

a small drop in activity with increasing hydrophilicity. The Snβ prepared via traditional

hydrothermal synthesis is, on the other hand, about four times less active than 1Snβ-F400,

although the hydrophilicities of these two materials are similar.

These observations demonstrate that the activity of Snβ in the BV oxidation of

cyclohexanone with H2O2 is influenced by the hydrophilicity of the zeolite framework, and that

higher hydrophilicity (i.e., higher silanol density) has a positive effect on the reaction rate up to

a certain level, beyond which it leads to a lowering of the reaction rate (c.f., Sabatier principle).

This behavior most likely shows that a certain amount of silanols aids the performance of the

catalyst by providing an optimal surface character for the ketone to approach and adsorb to

the active SnIV-sites. At higher silanol amounts this effect is offset by increasingly strong

interactions between the silanols and water, substrate or product molecules, which leads to a

reduced local concentration of substrate at the active SnIV-sites and a lowering of activity.

The data presented here indicates that the activity of Snβ in the Baeyer-Villiger oxidation of

cyclohexanone with H2O2 is highly dependent on the hydrophilicity of the material. However, to

better understand the observed trend in reactivity, also other physical and chemical properties of

the Snβ catalysts have to be considered. As such, BET surface areas between 545 and 718 m2 g-1

were obtained from N2 adsorption measurements at 77 K (see Table A.1), which do not show a

clear correlation with our measured activities, even though the high surface area of 1Snβ-F30 is

possibly related with the high activity of this best-performing catalyst. Furthermore, we know

from a previous study that 1Snβ-F400 and 1Snβ-HT show significantly lower external surface

areas compared to the other Snβ catalysts as a result of their prolonged crystallization process

in fluoride media.[157] However, this does not fall in line with the catalytic activities we observe,

since 1Snβ-F400 and 1Snβ-HT exhibited significantly different activities in our catalytic tests.

We also learned in a previous study that the materials with a more hydrophobic framework

have higher crystallinities, even though no direct correlation between crystallinities and our

observed catalytic activities in the Baeyer-Villiger oxidation of cyclohexanone with H2O2 can

be established.[157]

Lastly, also differences in the nature of the active sites between the various Snβ catalysts

might contribute to their different activities. Indeed, based on the data presented here, we


cannot exclude that active site speciation has an influence on the reactivity of our different Snβ

catalysts in the BV oxidation of cyclohexanone with H2O2. In a previous combined experimental

and computational study some of us investigated the influence of the active site structure

of a series of Snβ catalysts on their activities in the aqueous phase isomerization of glucose

to fructose.[157] Their experiments suggest that post-synthetically prepared Snβ catalysts are

mostly represented by closed (T8) and open (T5/T9) sites incorporated into the hydrophilic

framework. Hydrothermally synthesized Snβ, on the other hand, mainly consists of closed (T6,

T5, T7) sites, which can be hydrolyzed under reaction conditions. Based on our data, open

sites (i.e., 1Snβ-OH25, 1Snβ-OH100, 1Snβ-F30), as well as closed sites (i.e. 1Snβ-F200, 1Snβ-

F400) are active in the Baeyer-Villiger oxidation of cyclohexanone with H2O2, which does not

indicate that site structure has a prevailing influence on activity in this reaction system.

Interestingly, Snβ samples obtained from post-synthetic modification (i.e., 1Snβ-OH25,

1Snβ-OH300, 1Snβ-F30, 1Snβ-F200, 1Snβ-F400) showed a ca. 3-5 times lower activity per Sn

in the isomerization of glucose to fructose (in terms of TOF) compared to hydrothermally

synthesized Snβ catalyst (i.e., 1Snβ-HT),[157] which is the opposite trend in reactivity to

what we observe for the BV oxidation of cyclohexanone with H2O2. Indeed, hydrothermally

synthesized Snβ zeolite shows lowest activity in our study. This indicates that the relation

between catalyst properties and catalytic activities is strongly dependent on the individual

reaction system. The disclosure of such contrary trends, may, however, strongly contribute

to the understanding of structure-activity-relationships for a material in different catalytic

systems.

3.4. Conclusions

In this work, we determine the kinetic regime of the BV oxidation of cyclohexanone with

aqueous H2O2 for post-synthetically prepared Snβ zeolite. For this we study initial reaction

rates as a function of metal loading and temperature and define reaction conditions that exclude

diffusion limitations for following investigations.

We also study the impact of the water and lactone concentration in the reaction solution

on the catalytic activity of Snβ and find that both lower the performance of the catalyst,

38 Chapter 3

which we assign to active-site inhibition through these compounds. We furthermore prepare a

series of Snβ catalysts with different hydrophilicities by following different synthesis protocols.

The hydrophilicities of the prepared materials are quantified with water adsorption studies and

by analysis of the silanol IR region, which results in a good correlation between these two

quantities.

The catalytic performance of the different Snβ catalysts in the BV oxidation of

cyclohexanone with aqueous H2O2 is different and is influenced by the zeolite hydrophilicities.

We observe that an optimum in hydrophilicity exists, which indicates that a hydrophilic

framework aids the adsorption of the ketone substrate to the active SnIV-sites, and that this

effect is outweighed by active site inhibition through solvent (water) and product molecules

(lactone) at higher hydrophilicities.

Our work demonstrates that hydrophilicity may influence the catalytic performance of

Snβ, and that flexible synthesis methods, such as our post-synthetic metal incorporation, allow

optimizing activities through targeted structural modifications.

Chapter 4

Silica-Grafted SnIV Catalysts inHydrogen-Transfer Reactions

In this chapter all catalytic experiments, material synthesis and characterization was conducted

by the author. NMR measurements were performed with the assistance of René Verel. The

synthesis of the reference material Snβ was done by Patrick Wolf.

4.1. Introduction

A wide range of important chemical transformations, including isomerizations and oxidations,

are catalyzed by heterogeneous catalysts with isolated Lewis acid sites.[5,16,61] Most notable

amongst them are metal-doped zeolites, some of which have demonstrated outstanding activity

and selectivity, as well as high stability and water tolerance.[20,52,128,163] A contemporary example

is SnIV-containing β-zeolite (Snβ).[46,164] This catalyst excels in activating carbonyl groups and

has been explored for various reactions, such as Baeyer-Villiger oxidations,[48,115,154] Meerwein-

Ponndorf-Verley (MPV) hydrogen-transfer reactions,[39,108] sugar isomerizations[26,29,30] and C-C

bond formations.[45,165]

It is expected that various material properties contribute to the overall performance of

Snβ: (i) the molecular environment of the active sites,[21,101] (ii) confinement effects induced

by the micropores of the zeolite,[110,166] and (iii) the hydrophobicity of the framework (as found

for highly siliceous zeolites).[16,17,108,109] In particular, the exact site structure has recently

received significant attention and initiated characterization work using 119Sn NMR,[116–121]

EXAFS,[47,106] TEM[47,106] and XPS.[47] These studies yielded valuable structural information

and have been complemented by IR probe studies, which are more closely related to the

actual reaction conditions. In these investigations, interaction-induced shifts in vibrational

bands of adsorbed probe molecules can indirectly provide information regarding the catalytic

transformation.[60,64,101,111,115,123,127]

40 Chapter 4

One important probe molecule in this context is cyclohexanone,[64,115,127] a widely used

substrate in Snβ catalyzed reactions.[26,29,30,39,45,48,108,115,154,165] It is assumed that the observed

shifts of the carbonyl band correlate with the Lewis acidity of the SnIV-sites,[64,115,127] and

is hence a good measure for the catalytic performance of the site in question. Additionally,

the comparison of experimental and computed shifts allowed researchers to draw structural

conclusions about the active sites.[122] Some studies also correlated measured catalytic activities

with the individual activities of different types of sites in Snβ.[111] Therefore, the intensities of

the different bands are used to estimate the amounts of individual sites present in the catalyst.

However, not only the nature of the active site, but also multiple adsorption to one site might

induce similar shifts of the IR bands.[123] This leads to difficulties for an unambiguous assignment

of the active site(s), based solely upon IR bands. Moreover, exclusively cyclohexanone is

adsorbed in these studies, while a mixture of molecules is present under reaction conditions.

It is not entirely clear whether these molecules (co-reactants, solvent or product molecules)

competitively adsorb to certain types of active centers.

Besides the structure of the active site, also the nature of the zeolite (pore architecture and

polarity) might contribute to the outstanding performance of Snβ. It is therefore challenging

to distinguish between the different contributions in a zeolite system. An appealing approach

to separate the individual factors is to use a model-system.

One possibility to form isolated active sites on a silica surface (where confinement effects

are absent) is grafting.[67,76–79] In this approach, a metal precursor[88–90,93,167] reacts, after

careful pretreatment of the support, with surface silanol groups,[86] which leads to a uniform

distribution of highly dispersed sites. By applying post-grafting treatments, the coordination

of the active centers can be modified.[91,92,102,168] For instance, the metal may form additional

bonds to the surface during a thermal treatment (i.e., multipodal anchoring).[88,89] All these

synthetic steps can be monitored by various characterization techniques (FT-IR, UV-Vis, solid-

state NMR) and often change the activity of the catalyst.[88,89,92,102,168]

In this work (Figure 4.1), we form isolated SnIV-sites on amorphous silica upon grafting (I)

and apply a thermal and a chemical post-treatment (II), leading to a set of three Sn-model-

catalysts. We characterize the active sites with different spectroscopic techniques and test

Silica-Grafted SnIV Catalysts in Hydrogen-Transfer Reactions 41

Figure 4.1. To investigate the effect of the active site structure and surface hydrophobicity onthe activity of SnIV-based catalysts, we form isolated SnIV-species on amorphous silica via chemicalvapour deposition of Sn(NMe2)4 (I). Subsequently we perform post-synthetic treatments to changethe bonding environment of the sites (II), and test the catalytic activities of the materials in MPVreductions (III). We finally relate this to results obtained from IR adsorption studies (IV).

their performance in the MPV reduction of cyclohexanone with 2-butanol (III), a hydrogen-

transfer-reaction. We study the adsorption of cyclohexanone with FT-IR (IV) and relate this

to their catalytic activities. Comparing our data with Snβ then allows to draw conclusions

about different contributions to the activity of SnIV/SiO2 catalysts.

4.2. Experimental


Silica powder (Aerosil 200® from Degussa, specific surface area: 200 m2 g-1) was impregnated

with water and dried overnight at 100 °C in a vacuum oven. Afterwards, the material was

dehydrated at 700 °C (16 h) at 15 µbar dynamic vacuum and stored inside of a glove box

(< 1 ppm O2 and H2O) in order to avoid re-adsorption of water. Sn(NMe2)4 (Sigma-Aldrich,

two times distilled before use, colorless) was then deposited onto 300 mg of dried silica at ca.

20 µbar dynamic vacuum (5 equivalents based on initial silanol content). The transfer phase

(45 min) was followed by a reaction phase at RT (30 min) and a mild thermal post-treatment at

50 °C and 15 µbar dynamic vacuum (1 h) in order to evaporate excess of transferred Sn(NMe2)4

(material denoted as Sn/SiO2(700)).

Samples were calcined in an air flow at 550 °C for 3 h (material denoted as Sn/SiO2(700)-

O2) and dehydrated for further use (300 °C, 3 h, 15 µbar dynamic vacuum). N,N -

bis(trimethylsilyl)methylamine (Sigma Aldrich, two times distilled before use) was used for

the silylation of the dried silica (material denoted as TMS-SiO2(700)). The silylating agent was

42 Chapter 4

contacted with the material in the same way as Sn(NMe2)4 and thermally post-treated at 250 °C

(5 equivalents based on initial silanol content). Afterwards, Sn(NMe2)4 was deposited on the

silylated material to obtain Sn/TMS-SiO2(700) (2 equivalents based on initial silanol content).

N,N -bis(trimethylsilyl)methylamine was also used for the silylation of the dehydrated

calcined material Sn/SiO2(700)-O2 (material denoted as Sn/SiO2(700)-O2-TMS) using the same

deposition procedure (5 equivalents based on initial silanol content).


ICP-OES, FT-IR and UV-Vis analysis was performed as described in Chapter 2.

The nitrogen content was obtained with a thermal conductivity detector. The combustion

products (CO2, H2O) were quantitatively analyzed by infrared spectroscopy. The nitrogen

content was obtained with a thermal conductivity detector.

The DRIFT spectra were recorded by averaging 32 scans with a resolution of 8 cm-1. The

DiffusIR accessory (PIKE Technologies) was flushed with synthetic air at a flow rate of ca.

20 mL min-1 and heated at a heating rate of 20 °C min-1 up to a temperature of 550 °C.

Solid-state 13C-NMR spectra were acquired on an Avance NMR spectrometer (Bruker,

Karlsruhe, Germany) operating at a 1H Larmor frequency of 700 MHz. The samples were spun

around the Magic Angle with a rate of 10 kHz at room temperature using a double resonance

4 mm probe (using ca. 40 mg sample). The two channels of the probehead were tuned to the

resonance frequencies of 1H (700.13 MHz) and 13C (176.06 MHz). The ppm scale of the spectra

was calibrated using the 13C signal of adamantane as an external secondary reference. The 13C

spectrum was acquired using a Cross Polarization (CP) sequence with higher power TPPM

decoupling of the protons during detection of the 13C signal.

Thermogravimetric analysis of the calcination step was performed using a TGA-DSC

Thermobalance (Mettler Toledo) in combination with an OmniStarTM massspectrometer

(Pfeiffer Vacuum).

4.2.3. Adsorption Studies

The samples were initially heated to 300 °C at ca. 15 µbar dynamic vacuum (3 h) in order

to remove any water from the surface of the material. After cooling the samples to room


temperature, cyclohexanone vapors were exposed to the samples using a closed evacuated

system (ca. 20 µbar static vacuum) consisting of a sample holder (10 mL tube reactor)

connected to a capillary reactor filled with cyclohexanone. The exposure time was complete

when the filling level of the ketone in the capillary reactor was stable. After recording an

FT-IR spectrum of the material inside a glove box (< 1 ppm O2 and H2O) the sample holders

were evacuated several times in order to gradually re-adsorb the probe molecule. After each

desorption, an FT-IR spectrum was measured.


The Meerwein-Ponndorf-Verley reactions of cyclohexanone were carried out in a 20 mL round

bottomed-flask fitted with a reflux condenser, using 2-butanol as the hydrogen transfer agent.

In a typical reaction, 1 mmol of cyclohexanone and 60 mmol of 2-butanol (corresponds to 5.5 mL

of a 0.18 M reaction solution of cyclohexanone in 2-butanol) were added to the appropriate

amount of catalyst (corresponding to 0.5 mol% Sn relative to the ketone). The vessel was

heated at a constant temperature (90 °C) for 6 h under stirring (500 rpm). The reactant

and product were quantified against the internal standard biphenyl by GC-FID analysis (30 m

FFAP column).

4.3. Grafting of Sn(NMe2)4 to Thermally Treated Silica

The first step in our synthesis procedure is the pretreatment of the silica support (Aerosil 200®)

at 700 °C under vacuum in order to desorb physisorbed water and to condensate hydrogen-

bonded vicinal silanols (material denoted as SiO2(700)). Figure 4.2 shows the FT-IR spectrum of

SiO2(700) (B), together with a spectrum of silica treated at 200 °C (SiO2(200), (A)). Both spectra

exhibit Si-O-Si overtones of the silica framework (ca. 1500 – 2100 cm−1), as well as a signal at

higher wavenumbers in the IR region characteristic of SiO-H vibrations. While the O-H band

of SiO2(200) features a low-energy shoulder, stemming from residual hydrogen-bonded silanols

(blue area), the O-H band of SiO2(700) is sharp and symmetrical (red area).1 This indicates

that the surface of SiO2(700) consists predominantly of non-interacting silanols.[86]

To anchor SnIV to these isolated silanols, we then transferred Sn(NMe2)4 onto Sn/SiO2(700)

1Under these conditions, the presence of minor amounts (≤ 5 %) of geminal diols cannot be excluded.[86]

44 Chapter 4

SiO2(200)

SiO2(700)

Sn(NMe2)4

Sn/SiO2(700)

(A)

(B)

(C)

(D)

Figure 4.2. FT-IR spectra of silica dehydrated at (A) 200 and (B) 700 °C, (C) Sn(NMe2)4 and (D)Sn/SiO2(700). The thermal treatment induced the condensation of hydrogen-bonded silanols (blue)and left isolated surface silanols behind (red). Grafting of Sn(NMe2)4 led to the conversion of allisolated silanols and the appearance of C-H precursor stretches (green).

in the gas phase. After evacuating under mild heating, we obtained the grafted material

Sn/SiO2(700). The IR spectrum in Figure 4.2 D indicates that all silanols were quantitatively

consumed during grafting. In addition, intense IR bands could be observed in the C-H stretching

(2800 – 3100 cm−1) and bending (1400 – 1470 cm−1) range (green area). Comparison of the IR

bands of Sn/SiO2(700) with those of the Sn(NMe2)4 precursor (Figure 4.2 C) suggests that only

minor changes in the precursor structure took place upon grafting. We believe that all these

observations indicate a surface reaction between silanols and Sn(NMe2)4 to yield HNMe2 and

≡SiOSn(NMe2)3 species. This hypothesis was supported by the unambiguous identification of

HNMe2 in the liquid-nitrogen trapped effluent of the synthesis reactor with 1H- and 13C-NMR

after Sn(NMe2)4 deposition (Figure A.4). In addition, we performed ICP- and CHN-analyses

to obtain molar ratios (C/Sn, H/Sn, N/Sn). These ratios are in good agreement with the

proposed monopodal SnIV-sites (≡SiOSn(NMe2)3) (Table A.2).

Surprisingly, the Sn-loading, that we obtained from bulk analysis (4.0 ± 0.1 wt%,

0.33 mmol Sn g−1cat, corresponds to a site density of 1 Sn site per nm2. This exceeds the

site density of silanols on SiO2(700)[78] by around 25 %. Therefore, we suspect that silanols are


Figure 4.3. Possible pathways for the reaction of surface siloxanes with Sn(NMe2)4. Indirect route:(A) opening of siloxanes through HNMe2 (liberated during the grafting); (B) reaction of the formedsilanols with Sn(NMe2)4. The direct route (C) immediately yields a 1:1-mixture of Sn and aminespecies through cleavage of siloxanes by Sn(NMe2)4.

not the only surface functional group that can react with Sn(NMe2)4.

A possible explanation for this is the participation of (strained) siloxane bridges (≡SiOSi≡),

formed during the dehydration. A fraction of these siloxanes is part of strained ring-structures

((SiO)n, n=2).[86,87,169] Earlier studies demonstrated the reaction of these strained sites with

molecules containing O-H or N-H bonds.[170–172] If water and ammonia are used, such a surface

reaction results in adjacent ≡SiOH groups or a 1:1-mixture of ≡SiOH and ≡SiNH2 groups,

respectively.[170–172] We expect that the leaving group of our grafting reaction (HNMe2) and/or

the Sn-precursor itself (Sn(NMe2)4) can react with the strained siloxanes of SiO2(700) in a similar

manner, which would explain the unexpected high Sn-loading.

Based on this hypothesis, we propose two possible pathways for the reaction of the surface

siloxanes of Sn(NMe2)4 with HNMe2 and/or Sn(NMe2)4 under grafting conditions (Figure 4.3).

One pathway ((A)-(B) in Figure 4.3) consists of two consecutive steps. In a first step, siloxane

bridges are opened by HNMe2, yielding a 1:1-mixture of new silanols and N-functionalized

silicon atoms (≡SiNMe2). The formed silanols can then, similar to isolated silanols, react with

Sn(NMe2)4 to form SnIV-sites and HNMe2 (which leaves the surface or reacts with additional

siloxanes). The second pathway ((C) in Figure 4.3) consists of the direct cleavage of siloxanes

by Sn(NMe2)4. This immediately yields a 1:1-mixture of Sn- and amine-sites.

To verify if siloxanes can directly react with Sn(NMe2)4 (via step (C) in Figure 4.3) we

protected the silanols of SiO2(700) via silylation with N,N -bis(trimethylsilyl)amine (viz., ≡SiOH

46 Chapter 4

δ / ppm

Sn/SiO2(700)

Sn/TMS-SiO2(700)

(A)

(B)

Figure 4.4. 13C-NMR spectra of Sn(NMe2), grafted on to (A) SiO2(700) and (B) TMS-SiO2(700). Twosignals in the region of amine-based methyl-groups were observed (green and orange area), which givesevidence for the proposed pathways in Figure 4.3. The silylated material also showed an expectedTMS-related peak at 0 ppm (pink area).

→ ≡SiOSi(CH3)3; material denoted as TMS-SiO2(700)), prior to bringing it into contact with

Sn(NMe2)4. The resulting material contained 0.9 ± 0.02 wt% Sn, which exactly corresponds

to the amount of Sn exceeding the exclusive reaction with silanols during the synthesis of

Sn/SiO2(700) (Sn-content of 4.0 ± 0.1 wt%). This confirms the participation of siloxanes in the

grafting reaction. We further performed solid-state 13C-NMR measurements (Figure 4.4) to

investigate the nature of the organic groups we observed with FT-IR. The 13C-NMR spectrum

of Sn/SiO2(700) (Figure 4.4 A) shows two signals in the chemical shift region of amines (green

and orange area). The 13C spectrum of Sn/TMS-SiO2(700) (Figure 4.4 B) showed signals at the

exact same shifts (albeit lower intensity due to the smaller relative site concentration), together

with an expected TMS-related peak at 0 ppm (pink area). This experiment indicates that there

are two distinct amine species on Sn/SiO2(700) and Sn/TMS-SiO2(700), which are most likely of

the type ≡SiOSn(NMe2)3 (42 ppm) and ≡SiNMe2 (36 ppm). We based our assignment on the

higher expected amount of Sn-coordinated amine groups compared to Si-bound amine groups.

To obtain additional evidence for the existence of these two sites, we analyzed the C-H

stretch region in the IR spectrum of Sn/TMS-SiO2(700) (Figure A.5 and Figure A.6). Since the

theoretical ratio between our hypothesized amine (≡SiNMe2) and SnIV-sites (≡SiOSn(NMe2)3)

is 1:1 in Sn/TMS-SiO2(700), but 1:5 in Sn/SiO2(700), we expected that signals arising from

the ≡SiNMe2 sites should be (relatively) more pronounced in Sn/TMS-SiO2(700). Compared


to Sn/SiO2(700), Sn/TMS-SiO2(700) revealed similar signals, along with two additional bands,

which might be attributed to ≡SiNMe2 sites. We emphasize that our silylating agent (N,N -

bis(trimethylsilyl)amine) has been reported to also open a small fraction of siloxanes to form

≡SiN(TMS)Me sites.[89] Computationally predicted methyl C-H vibrational frequencies for

≡SiNMe2 and ≡SiN(TMS)Me (Table A.3) allow us to assign the higher values to the TMS-

holding site, in line with our experimental observations. We believe this demonstrates the

presence of grafted amine species, which are likely amine sites of the type ≡SiNMe2.

To verify route (A) in Figure 4.3, we exposed TMS-SiO2(700) also to HNMe2. We obtained a

material with a small fraction of nitrogen, indicating that the direct reaction between siloxanes

and pure HNMe2 is negligible. In order to test if the siloxane-oxygen atom first needs to be

activated by a Lewis acid (i.e., Sn(NMe2)4) to react with amines, we simultaneously deposited

HNEt2 and Sn(NMe2)4 on to SiO2(700). By using diethyl- instead of dimethylamine, it was

possible to differentiate Sn(NMe2)4- and HNEt2-related organic fragments. This experiment

yielded a material with significantly more intense signals in the C-H IR stretch region compared

to Sn/SiO2(700) (Figure A.7). Since the material revealed the exact same Sn-loading as

Sn/SiO2(700), this increase in intensity did not arise from additional SnIV-sites but from grafted

diethylamine groups and their extra secondary C-H groups. By comparing this to the negligible

reaction between HNMe2 and TMS-SiO2(700), we consider this as clear indication of the Lewis

acid-activation of siloxane-oxygens through Sn(NMe2)4, which then allows HNEt2 (and therefore

most likely also HNMe2) to react with them. Accordingly, we believe that both pathways (A)-

(B) and (C) in Figure 4.3 are possible.

4.4. Post-Synthetic Functionalizations

At this stage, the monopodal SnIV-species are coordinated by three –N(CH3)2 ligands as

illustrated by bulk analysis, NMR and FT-IR. We continued the synthesis by subjecting

Sn/SiO2(700) to a calcination in an air flow in order to obtain SnIV-centers that are exclusively

coordinated by oxygen atoms (material denoted as Sn/SiO2(700)-O2). This is expected to

increase the Lewis acidity of the SnIV-species by avoiding electron-donating amine ligands.

We monitored the calcination process in situ with DRIFT spectroscopy (Figure 4.5). A

48 Chapter 4

gradual decrease of all C-H stretches (region 1: 2800 – 3100 cm−1) confirmed the complete

removal of all organic groups. In parallel, a signal in the IR region typical of N-H stretches

(signal 2: 3290 cm−1) increased in intensity, before it decreased again at around 150 °C and

disappeared above 350 °C. We assume that this signal arises from an intermediate ammonium

species that is formed on the silica surface during the calcination. Furthermore, we observed

two peaks appearing in the IR region characteristic of metal hydroxyl O-H stretches (signal

3: 3664 cm−1; signal 4: 3740 cm−1), which we attribute to Sn hydroxyls and isolated silanol

groups, respectively. The build-up of a SnOH signal suggests that the initial Sn-NMe2 bonds

have been replaced by Sn-OH groups, while the appearance of a silanol band (4 in Figure 4.5)

shows that isolated silanols are generated during the calcination.

We emphasize that no Sn is released from the surface during the calcination (as confirmed

by bulk analysis). To test if water impurities in the air flow could open siloxanes under our

calcination conditions and could, thus, be responsible for the formation of new silanols, we

exposed SiO2(700) to a humid air stream. By monitoring the hydroxyl region with DRIFT

spectroscopy while heating, we observed a signal of hydrogen-bonded silanols (Figure A.8). This

indicates the presence of adjacent ≡SiOH groups, due to opening of siloxanes by water. Since

the silanol band of Sn/SiO2(700)-O2 (1 in Figure 4.5) possesses no such low-energy shoulder, we

conclude that it is not water that induces the release of silanols during the thermal treatment.

Another possibility is that the Sn hydroxyl groups, which are formed during the calcination

4 3

2 1

1 2

3 4

Figure 4.5. In situ DRIFTS study of the air calcination of Sn/SiO2(700). Initially, the C-H bandsof the amine ligands (1) were visible. At rising temperatures an intermediate ammonium species (2),SnO-H (3) and SiO-H (4) stretches appeared.


Figure 4.6. Thermal reconstruction of (A) ≡SiO-Sn(OH)3 and (B) (≡SiO)3Ti-Cl species.[88]

(3 in Figure 4.5) react with nearby siloxane bridges to form bipodal Sn-sites ((≡SiO)2Sn(OH)2)

plus silanols. This would provide the Sn with additional (stabilizing) anchoring points to

the silica surface and, at the same time, explain the generation of new isolated silanols (4 in

Figure 4.5). To gain evidence for this site-restructuring we estimated the amount of silanols

that were released during the calcination by comparing the silanol peak areas in the IR spectra

of Sn/SiO2(700)-O2 and SiO2(700). It turned out that relative to the Sn-loading of Sn/SiO2(700)-

O2, 50 % (± 5) new silanols are formed. This is consistent with the quantification of the SnOH

IR signal, which resulted in a SnOH-to-Sn ratio of 2.5 ± 0.3 (Figure A.9). We thus assume

that about half of the monopodal Sn species (≡SiO-Sn(OH)2) react with nearby siloxanes

and restructure to bipodal Sn species ((≡SiO)2Sn(OH)2), while generating isolated silanols

(Figure 4.6 A). Conclusively, our calcination step yields a mixture of mono- and bipodal

hydrolyzed SnIV-sites, which have been suggested to be more reactive than closed SnIV-sites

(i.e., fully anchored to a material framework (≡SiO)4Sn).[45,60]

Conversely, earlier work within our group demonstrated that similar site-restructuring of

monopodal Ti-Cl-species (≡SiO-TiCl3) resulted in quantitative conversion to tripodal sites

((≡SiO)3Ti-Cl; Figure 4.6 B), whereas no tripodal species were observed on our materials

based on two quantification methods. More work is necessary to confirm and understand the

molecular reasons behind this behavior.[88] In this respect, DQ/TQ 1H MAS NMR at high

spinning rates is one envisaged technique in order to differentiate and quantify various types of

Sn hydroxyl species.

50 Chapter 4

wavelength / nm

Kub

elka

Mun

k

(A)

(B)

(C)

(D)

(E)

(F)

wavelength / nm

Sn-β

SnO2 Sn/SiO2(700)-O2

Sn/SiO2(700)

Sn(NMe2)4 Sn/SiO2-O2(700)-TMS

Figure 4.7. Comparing the UV-Vis spectra of (A) Sn(NMe2)4, (B) Sn/SiO2(700), (C) Sn/SiO2(700)-O2

and (D) Sn/SiO2(700)-O2-TMS with (E) Snβ and (F) SnO2 indicated the absence of SnOx-species in ourmaterials (for the UV/Vis spectra of Sn/TMS-SiO2(700) and Sn/TMS-SiO2(700)-O2, see Figure A.11).

Having the hydrophobic character of the highly siliceous cavities of Snβ in mind, our

second functionalization consisted of the capping of all hydroxyl groups on the surface (SiOH

and SnOH) with trimethylsilyl-units using N,N -bis(trimethylsilyl)methylamine. The infrared

spectrum of the obtained material (denoted as Sn/SiO2(700)-O2-TMS) confirms the conversion

of both types of hydroxyls together with the introduction of C-H stretches from the organic

substituents of the silylating agent (3100 – 2700 cm−1 and 1500 – 1400 cm−1; Figure A.10).

In order to show that no SnOx species are formed during the post-synthetic treatments

we performed Diffuse Reflectance UV-Vis measurements. According to the obtained spectra

(Figure 4.7), there is no indication of SnOx species, given by the lack of absorbance above

280 nm. As expected, we also observed a blue-shift, going from our Sn-precursor (Sn(NMe2)4:

233 nm) along the different synthesis steps (Sn/SiO2(700)-O2-TMS: 226 nm), since four nitrogen

atoms are gradually replaced by four oxygen atoms. Compared to the absorption band of Snβ

with a maximum at 216 nm (Figure 4.7 E), which is indicative of tetrahedral isolated SnIV-

sites in a zeolite framework, the bands of our materials appear at slightly higher wavelengths

(i.e., Sn/SiO2(700): 228 nm; Sn/SiO2(700)-O2: 222 nm; Sn/SiO2(700)-O2-TMS: 226 nm). One

possible explanation for this is the lower coordinative strain on the more flexible silica support,

compared to the rigid crystalline zeolite framework.


We also performed deuterated acetonitrile adsorption studies with our Sn materials and

reference Snβ (Figure A.13). Both Sn/SiO2(700)-O2 and Sn/SiO2(700)-O2-TMS show signals at

around 2310± 2 cm−1, which is characteristic of Lewis acid sites, as shown before for Snβ.[101,111]

This observation points towards the similarities between our silica-supported SnIV-sites and the

framework SnIV-sites in Snβ.

4.5. Catalytic Reactivity of the Supported Species

To test the catalytic activities of our prepared materials we performed Meerwein-Ponndorf-

Verley (MPV) reductions of cyclohexanone with 2-butanol (Figure 4.8).[48,64,173,174] Bulk

analyses of the Sn-loading before and after the reactions confirmed that no leaching of the

active metal takes place.

Figure 4.8. Catalytic reaction to probe the activities of our materials: Meerwein-Ponndorf-Verleyreaction of cyclohexanone with 2-butanol (0.18 M cyclohexanone, 90 °C, 0.5 mol% Sn). For moredetails see Experimental Part.

The catalytic performances of our materials follow the trend Sn/SiO2(700) � Sn/SiO2(700)-

O2 < Sn/SiO2(700)-O2-TMS, with turnover frequencies (TOF) for Sn/SiO2(700), SnO2/SiO2(700)

and SiO2(700) being close to zero (see Table 4.1 and Figure 4.9).

Figure 4.9. Catalytic activities of various Sn-containing materials and SiO2(700) in the MPV reductionof cyclohexanone with 2-butanol (reaction conditions see Figure 4.8 and Experimental Part).

52 Chapter 4

Table 4.1. Catalytic activities in the performed MPV reduction and amounts ofadsorbed cyclohexanone (after exposure to 2-butanol) from IR adsorption studiesfor our materials and Snβ.a

Entry Material TOFinitb,c TON6h

d Areac

[h-1] [–] [abs. cm-1 mol−1Sn ]

1 Sn(NMe2)4 < 1 < 1 N/Af

2 Sn/SiO2(700) < 1 < 1 36 ± 4.53 Sn/SiO2(700)-O2 6 ± 0.2 10 60 ± 7.04 Sn/SiO2(700)-O2-TMS 8 ± 0.2 28 85 ± 9.45 Snβ zeolitef 140 ± 4 181 35 ± 0.83a Reaction conditions see Figure 4.8; b Defined as mole product generated per mole Snper hour extrapolated to 0 % conversion; c Errors are estimated using the Sn loadingdetermined with ICP; d Defined as the mole product generated per mole Sn after 6hreaction time; e The area of chemisorbed cyclohexanone is estimated with the peak areasof shifted C=O signals (marked in grey in Figure 4.10), normalized to the Sn-loadings.All deconvoluted spectra are given in Figure A.12. Errors correspond to the standarderrors of the peak fit given by OriginPro 8.5.1.; f No adsorption studies were performedwith the Sn precursor, which is a liquid at RT; g Snβ zeolite with a Sn-loading of 10 wt%was prepared in a two-step post-synthetic method.[164]

To rationalize the observed trend in reactivity we performed an IR adsorption study of

cyclohexanone, a well-established technique to probe interactions between metal sites and probe

molecules.[60,64,101,111,115,123,127]

For all three catalysts, we found significant IR signals in the carbonyl region (between

ca. 1600 and 1730 cm−1) after the adsorption of the ketone to the catalysts (red curves

in Figure 4.10). However, the three samples show very different signals. While the

signal of chemisorbed cyclohexanone (around 1650 cm−1) is most dominant for Sn/SiO2(700),

Sn/SiO2(700)-O2 has a main feature in the region of physisorbed carbonyl (around 1700 cm−1).

In Sn/SiO2(700)-O2-TMS both features are present with similar intensities.

In the literature, the presence and extent of shifted bands (relative to the band of

physisorbed cyclohexanone) has been linked to the activity of the SnIV-catalysts.[64,115,127]

However, in this study we do not find such a correlation, since the catalyst with the highest

intensity in this region (viz., Sn/SiO2(700)) does not show any catalytic activity.

One possible reason is that not only cyclohexanone, but also 2-butanol is present under

reaction conditions. Therefore, we also studied the adsorption of cyclohexanone after the

catalyst was exposed to the alcohol co-reagent. This pretreatment induced significant changes in

the observed spectra (blue curves in Figure 4.10). Almost all the intensity in the chemisorption


wavenumbers / cm-1

Sn/SiO2(700) Sn/SiO2(700)-O2 Sn/SiO2(700)-O2-TMS

Figure 4.10. FT-IR spectra of cyclohexanone adsorbed on to our three SnIV-catalysts. Red curvesrepresent the adsorption of only cyclohexanone, blue curves the adsorption of ketone after exposureto 2-butanol. (Spectra were background corrected.)

region disappeared for Sn/SiO2(700) and a distinct physisorption feature appeared. This

spectrum was very similar to the spectrum of Sn/SiO2(700)-O2 without exposure to the alcohol,

which indicates a ligand exchange at the SnIV-center. CHN analysis confirmed this hypothesis

(Table A.2). For Sn/SiO2(700)-O2 the shape of the spectrum remains unchanged and only

the overall intensity decreased, while the spectra with or without alcohol are very similar for

Sn/SiO2(700)-O2-TMS.

In a next step we deconvoluted these spectra in order to link the data with the catalytic

activity results (Figure A.12). As shown in Table 4.1 (Entry 3 and 4), it is possible to correlate

the peak areas of the activated carbonyl features with the catalytic activity. This allows a

qualitative understanding of the performance for Sn/SiO2(700)-O2 and Sn/SiO2(700)-O2-TMS,

but it does not explain the inactivity of Sn/SiO2(700).

Although our data does not unambiguously allow identifying the underlying reasons of the

observed catalytic behavior, we believe that possible explanations include differences in the

structure of the SnIV-sites and in surface hydrophobicity.

4.6. Discussion

Our observations in the IR study of cyclohexanone adsorption after exposure to alcohol clearly

indicate that it is the ability of the ketone to adsorb to the SnIV-centers that semi-quantitatively

describes the activity of the catalyst. However, we could not establish a similar correlation for

the ketone adsorption in the absence of alcohol. This is not entirely surprising, since the

presence of alcohol mimics realistic conditions in the MPV reduction.

54 Chapter 4

For MPV reductions on Snβ, the current benchmark material for Sn-catalyzed

heterogeneous reactions,[26,29,30,39,45,48,108,115,154,165] several computational studies suggested a

reaction pathway.[175,176] These mechanisms propose that both the ketone and the alcohol are

initially adsorbed to the SnIV-sites (Figure 4.11). Our results confirm this, since they imply

that under reaction conditions both the alcohol and the ketone simultaneously interact with

the SnIV-centers.

Figure 4.11. Proposed way of adsorption of cyclohexanone and 2-butanol on a partially hydrolysedSnIV-site in Snβ.[175]

In order to put the absolute activities of our materials in perspective, we also measured

the activity of Snβ in the MPV reduction (Figure A.14). It turned out that Snβ has a TOF

that is ca. 18 times higher compared to the TOF of our silylated material. This is remarkable

since the peak area of the perturbed carbonyl stretch in the IR spectrum of Snβ is significantly

smaller compared to the silylated material with co-adsorption of alcohol (Table 4.1, Entries 4

and 5; for spectra of Snβ see Figure A.12). This adsorption behavior does not explain the high

activity of Snβ and strongly suggests that we need to take additional parameters into account

such as (i) different active site structures and (ii) confinement effects in the pores of Snβ.

An indication for structural differences is given by the differences found in the UV-Vis

spectra of our grafted materials compared to Snβ, pointing towards more strain (i.e., higher

reactivity) for the zeolite-based SnIV/SiO2 catalyst. Interestingly, our results show that open

(SiO)xSn(OH)4-x sites are not necessarily more active than closed sites as is the current

hypothesis for Snβ. However, based on our data, we cannot exclude that SiOH groups adjacent

to open sites in Snβ could play a major role in activating the substrate(s). Nevertheless, it is

remarkable, that our calcined material (with Sn hydroxyls) is able to activate two times more


ketone but is 23 times less active compared to Snβ.

In addition to the nature of the active sites, confinement effects have been shown to

be important for various zeolite-catalyzed reactions. Most often they are traced back to a

greater stabilization of the transition state, relative to the reactants by weaker van der Waals

forces.[110,166] Most likely, this effect is complemented by an adsorption-based confinement

effect. Adsorption to active sites inside zeolite pores is composed of two components:[124,125]

(i) chemisorption of the molecule to the active site and (ii) van der Waals interactions with

the cavity. Clearly, this confinement effect increases the local concentration of the reactants.

Additionally, it needs to be considered that both reactants and products in our reaction

consist of an alcohol and a ketone. Based on the arguments above, the adsorption of ketone

to the active site is possible in the presence of an alcohol. We expect the chemisorption

interactions for the reactant and product ketone to be of similar magnitude. However, the van

der Waals interactions increase with the number of C-atoms in the molecule.[124–126] Since the

product ketone (2-butanone) contains less C-atoms than the reactant ketone (cyclohexanone),

the concentration of butanone at the SnIV-sites will be far lower than the concentration of

cyclohexanone compared to the silica materials. Therefore, the active sites are less likely to

be covered by the product ketone and the reverse reaction is less likely in Snβ. Both factors

should increase the net forward reaction rate significantly.

4.7. Conclusions

In this work we formed isolated SnIV-sites on the surface of amorphous silica via a simple

grafting procedure using Sn(NMe2)4 as volatile Sn-precursor. We investigated the structure of

the prepared materials with UV-Vis and FT-IR spectroscopy, as well as with solid-state 13C-

NMR experiments. Our results indicate that opening of strained siloxane bridges is initiated

by the amine-containing Sn-precursor. After a thermal treatment in air, the resulting Sn-OH

and Si-OH groups were modified with trimethylsilyl groups in order to increase the surface

hydrophobicity.

We furthermore measured the activity of our materials in the Meerwein-Ponndorf-Verley

reduction of cyclohexanone with 2-butanol. Only the calcined and silylated material showed

56 Chapter 4

activity, which we attribute to changes in the structure of the SnIV-sites and hydrophobicity,

respectively. Additionally, we established a semi-quantitative relationship between the IR

spectra of adsorbed cyclohexanone in the presence of 2-butanol. This clearly indicates that the

activity of the catalyst is related to the ability to chemisorb ketone under reaction conditions.

When comparing the activity of our silica-based materials to Snβ, the state-of-the-art

material for this reaction, we find a significantly lower activity for our materials. The underlying

reasons for this behavior are not yet unambiguously established. Our work demonstrates that

the development of model-catalysts is a convenient and feasible route to investigate which

features contribute to the performance of an active Lewis acid catalyst. Based upon this work

we suggest that confinement effects might significantly influence catalytic activity in the studied

system. Even though it is not possible to quantitatively asses this contribution to activity, more

targeted experimental approaches will help to further elucidate this effect in the future.

Chapter 5

Confinement Effects in Hydrogen-TransferReactions on Sn Sites in Porous Silica Materials

All experiments in this chapter were conducted by the author.

5.1. Introduction

In zeolite catalysis reactions take place inside micropores and the confining environment plays

a key role in understanding the activity as well as the selectivity of the catalyst for a series of

reactions.[113,177] While early ideas revolve around steric constraints of the pore, in recent years

a more detailed picture of these effects has emerged.[110,114,178–182] It involves the adsorption

of the molecule into the pore, where it is initially stabilized by van der Waals interactions,

its chemisorption to the active site and finally the reaction itself. Thermodynamically these

processes are governed by their respective free energies and depending on the molecules,

catalysts and the studied reactions confinement can affect reaction rates differently.

A reaction that has been demonstrated to be catalyzed by zeolite-based materials

is the Meerwein-Pondorff-Verley (MPV) reduction (see Figure 4.8), which is a promising

transformation for several biomass-relevant reaction schemes, and a frequently used tool in

organic synthesis.[39,108,175,176,183,184] In Chapter 4 we separated the contributions to the activity

of benchmark Sn-doped β-zeolite (Snβ) in the MPV reduction of cyclohexanone with 2-butanol

by grafting SnIV-sites, similar to those encountered in the zeolite framework, to silica.[185]

However, the obtained sites showed activities of more than an order of magnitude lower

compared to Snβ even after targeted thermal and chemical post-treatments, which led to slight

activity increases. Besides possible differences in the local structure of the active sites and

differences in hydrophobicity of the materials, we suggested that confinement effects in the

zeolite pores are responsible for the significantly higher activity of Snβ in this reaction. To

confirm this contribution, we extend our activity studies to SnIV-sites grafted to a mesoporous

58 Chapter 5

support (MCM-41), which represents a pore system that stands in between microporous β

zeolite and amorphous silica (i.e., a surface support where confinement effects can be excluded).

We prepare two MCM-41 supports with different pore sizes and expect that the resulting Sn

catalysts show activities that are higher than our Sn-silica-system, but lower than Snβ.

5.2. Experimental


Post-synthetic incorporation of Sn was performed in two steps as described in Chapter 3.[159]

MCM-41 was synthesized according to a procedure described elsewhere.[186] n-

Alkyltrimethylammonium bromides of different alkyl chain lengths from C12 and C16 were

used as template. The template was dissolved in 120 g of deionized water to yield a 0.055 M

solution, and 8.2 g of aqueous ammonia (28 wt%, 0.14 mol) was added to the solution. While

stirring, 10 g of tetraethoxysilane (0.05 mol) was added slowly to the surfactant solution over

a period of 15 min resulting in a gel with the following molar composition: 1 TEOS : 0.152 n-

alkyltrimethylammonium bromide : 2.8 NH2 : 141.2 H2O. The mixture was stirred for one

hour, then the white precipitate was filtered and washed with 100 mL of deionized water. After

drying at 363 K for 12 h, the sample was heated to 823 K (rate: 1 K min-1) in air and kept at

this temperature for 5 h to remove the template.

The grafting of Sn to MCM-41 was performed as reported in Chapter 4. The support was

dehydrated at 700 °C (16 h) at 15 µbar dynamic vacuum and stored inside of a glove box

(< 1 ppm O2 and H2O) in order to avoid re-adsorption of water. Sn(NMe2)4 (Sigma-Aldrich,

two times distilled before use, colorless) was then deposited onto 300 mg of dried support at ca.

20 µbar dynamic vacuum. The transfer phase (45 min) was followed by a reaction phase at RT

(30 min) and a mild thermal post-treatment at 50 °C and 15 µbar dynamic vacuum (1 h) in

order to evaporate excess of transferred Sn(NMe2)4 (material denoted as Sn/MCM-41). Samples

were calcined in an air flow at 550 °C for 5 h (material denoted as Sn/MCM41-O2).

Confinement Effects in Hydrogen-Transfer Reactions on Sn Sites in Porous Silica 59


The Sn-content was quantified with ICP-OES (Perkin Elmer Optima 2000) after digestion of

the samples with HF (48 %, Sigma-Aldrich). FT-IR and UV-Vis analysis was performed as

described in Chapter 2.

Powder diffraction patterns were recorded on a Bruker D8 advance diffractometer using

Cu-Kα1 radiation and a Lynxeye detector.

N2 sorption measurements were performed on a Micromeritics Gemini VII (Version 2.00) at

77 K. Samples were degassed under vacuum at 350 °C for 3 h prior to every sorption analysis.

The surface area was calculated using the Brunauer-Emmett-Teller (BET) theory.


The Meerwein-Ponndorf-Verley reductions of cyclohexanone were performed in 10 mL thick

wall tube reactors capped with a PTFE/silicon seal capable of holding 15 bar over-pressure. In

a typical reaction, 1 mmol of cyclohexanone and 60 mmol of alcohol (ethanol, 2-butanol, iso-

propanol and 2-pentanol) were added to the appropriate amount of catalyst (corresponding to

0.5 mol% Sn relative to the ketone). The vessel was heated at a constant temperature (80 °C)

for 6 h under vigorous stirring (500 rpm). Aliquots were taken periodically and the reactant

and product were quantified against the internal standard biphenyl by GC-FID analysis (30 m

FFAP column).

5.3. Results and Discussion

Snβ was prepared via post-synthetic incorporation of SnIV into dealuminated β-zeolite (see

Chapter 3). For the MCM41-grafted catalysts, we first synthesized two mesoporous MCM-41

materials with different pore sizes (2.5 and 3.3 nm; Table A.4), applying a procedure by Grün

and co-workers. In order to form SnIV-sites on these materials, we then followed the preparative

steps that we reported for the synthesis of silica-grafted SnIV-sites (see Chapter 4). We

dehydrated the materials at high temperatures (700 °C) to provide a surface that predominantly

consists of non-interacting isolated silanols, prior to contacting it with an appropriate amount

of Sn-precursor (materials obtained denoted as Sn/MCM41-25 and Sn/MCM41-33). We then

60 Chapter 5

subjected the materials to a calcination in air in order to remove the organic parts of the

precursor and obtain SnIV-sites that are exclusively coordinated by oxygen atoms. Opposed

to previous work, we omitted the silylation step to avoid blocking of the pores. Comparing

the UV-Vis spectra of Sn/MCM41-25 and Sn/MCM41-33 with SnO2 and Snβ shows that

the MCM41-grafted Sn catalysts predominantly contain tetrahedrally coordinated isolated Sn

centers (Figure 5.1).

Figure 5.1. UV-Vis spectra of different Sn-silica materials. (A) Snβ, (B) SnO2, (C) Sn/MCM41-25,(D) Sn/MCM41-36, (E) Sn/SiO2.

Figure 5.2. Catalytic activity measurements of different Sn-silica materials in the Meerwein-Ponndorf-Verley reduction of cyclohexanone with 2-butanol. Experimental conditions: 1 mmol ofcyclohexanone, 60 mmol of 2-butanol, 0.5 mol% Sn (relative to ketone), 80 °C, 6 h reaction time.

To probe the catalytic activities of the prepared materials we then performed MPV-

Confinement Effects in Hydrogen-Transfer Reactions on Sn Sites in Porous Silica 61

reductions of cyclohexanone with 2-butanol. We observe that for 2-butanol as reducing agent,

the activity of Sn/MCM41-33 is very similar to silica-grafted SnIV-sites, i.e., a surface catalyst

(Figure 5.2). However, for the same alcohol the activity more than doubles when the pore

diameter decreases to 2.5 nm (i.e., for Sn/MCM41-25) and is more than an order of magnitude

higher for microporous Snβ (pore diameter approx. 0.7 nm).

This observation clearly confirms that the pore size of the catalyst, (i.e., confinement)

influences the reaction rate of SnIV/SiO2-based catalysts in the Meerwein-Ponndorf-Verley

reduction of cyclohexanone with 2-butanol. In line with the arguments in our previous work

with silica-grafted SnIV-sites, we propose that an enhanced reverse reaction at active sites in

less-confined environments is one possible explanation for the observed smaller reaction rates for

MCM41- and silica-grafted SnIV-sites compared to Snβ. The underlying reason are stabilizing

Van der Waals forces in confining environments, which stabilize the larger substrate ketone

(cyclohexanone, C6) more strongly than the product ketone (2-butanone, C4), which would

lead to a higher coverage of the active sites with the reactant than with the product, and

hereby to an acceleration of the forward reaction and a slowdown of the reverse reaction. In

addition, other confinement effects could play a role. As such, we can not exclude an impact

of confinement on other thermodynamic contributions, such as the reaction entropy or reaction

enthalpy. Lastly, differences in the nature of the active sites or changes in the hydrophilicity of

the support material might also cause the observed trend in reactivity.

5.4. Conclusions

The results presented in this work indicate that confinement effects are one determinant

that contributes to the high activity of Snβ in the Meerwein-Ponndorf-Verley reduction of

cyclohexanone with 2-butanol. This is demonstrated by an increase in activity with decreasing

pore diameters for SnIV/SiO2-based catalysts (Sn/SiO2, Sn/MCM-41, Snβ). We attribute the

observed confinement effect with a possible change of the adsorption behaviour of substrate

and product molecules in confining environments. In the future we want to arrive at a solid

theoretical framework to describe these interactions quantitatively. We expect similar effects to

be present in different zeolite catalyzed reactions, and that more targeted approaches stimulated

62 Chapter 5

by the ideas presented here might help to rationally improve operating conditions for various

catalytic system

Chapter 6

Conclusions and Outlook

Accompanied by the statement of the principles of green or sustainable chemistry, a logical

transition from homogeneous to heterogeneous Lewis acid catalysts has been initiated, providing

catalysis researchers with new scientific challenges to overcome. As such, the exploration of

novel material solutions and an enhanced understanding of the obtained catalysts is necessary

and will sustain a long-lasting stream of academic research looking into basic aspects of

heterogeneous catalyst preparation, characterization, and testing.

Prior to the commencement of this thesis, the commercial realization of the state-of-the-art

solid Lewis acid Sn catalyst – Snβ zeolite – was hampered by its tedious and lengthy synthesis

procedure, and the identification of the activity characteristics of Snβ was complicated by its

complex structure. In order to fully utilize the commercial potential of advanced materials such

as Snβ, a few critical tasks need to be addressed. First, is the design of simple and scalable

synthesis procedures resulting in materials with comparable or higher activity, and second, is

the need for an in-depth understanding of the material reactivity in order to provide a rationale

for the design of materials with improved activity.

6.1. Conclusions

Simple and Scalable Synthesis. In this thesis, a simpler and "greener" route for the preparation

of Lewis acidic Snβ zeolite is described, consisting of the post-synthetic incorporation of Sn

atoms into dealuminated β-zeolite via solid-state-ion-exchange (Chapter 2). The obtained

material demonstrates similar to higher activity and selectivity compared to Snβ prepared via

hydrothermal synthesis, and the increased amount of incorporated metal significantly raises

the achieved space-time-yield (defined as gproduct kg−1catalyst h-1). In addition, the developed

procedure requires less synthesis time and avoids the formation of undesired toxic waste. As

such, the work presented in this thesis starts addressing one of the key challenges in shifting

64 Chapter 6

from homogeneously to heterogeneously based Lewis acid catalysts, that is the design of simpler

and environmentally benign synthesis routes to active materials. Based on this work, we expect

that the industrial realization of Snβ or a similar material (e.g., with different types of metals

or zeolite frameworks) will be strongly promoted.

Hydrophilicity. The Baeyer-Villiger oxidation of cyclohexanone with hydrogen peroxide is

one out of many potential industrial applications of Snβ zeolite. It is known from literature that

the high activity and selectivity of Snβ in this reaction arises from the selective interaction of the

active SnIV-sites of the catalyst with the carbonyl groups of substrate molecules. However, this

chemoselectivity also causes catalyst inhibition, most likely as a result of competitive adsorption

of solvent (water) and product molecules (lactone) to the active SnIV-sites, as we demonstrate

in Chapter 3. With this background, we modify the synthesis protocol of our newly developed

post-synthetic synthesis route in order to alter the amount of framework silanols, and hereby the

hydrophilicity of Snβ, aiming toward improving the activity in Baeyer-Villiger oxidations. We

learn that an optimum in hydrophilicity exists, reflecting the need to allow strong coordination

of the substrate (cyclohexanone) to the active SnIV-sites but to avoid inhibition by solvent and

product molecules. As such, the work presented in this chapter shows that flexible synthesis

protocols, such as the one from our post-synthetic synthesis route, allow targeted modifications

to the catalyst that may lead to optimized overall activities. Furthermore, our work clearly

identifies the hydrophilicity of the β-framework as one determinant of the activity of Snβ, and

therefore gives fundamental insights on material reactivity.

Catalytic Model-System. In the second part of this thesis (Chapter 4), we choose the

utilization of a catalytic model-system for SnIV/SiO2-based catalysts as an alternate pathway

to enhance our understanding of the reactivity of complex catalytic systems such as Snβ zeolite.

This is approached by grafting isolated SnIV-sites on to silica and gradually modifying the

surface-anchored metal sites with a thermal and a chemical post-treatment. Implications of

these treatments on catalyst structure and activity are followed with various characterization

techniques and catalytic tests, respectively. With this methodology we are able to distinguish

different contributions to the reactivity of SnIV/SiO2-based catalysts (including active site

speciation, material hydrophilicity and confinement effects), and demonstrate that catalytic

Conclusions and Outlook 65

model-systems are a feasible tool to untangle the material properties that determine the

performance of a Lewis acid catalyst. Moreover, our results indicate that the grafting of

our amine-containing Sn precursor (Sn(NMe2)4) leads to opening of siloxane bridges on the

thermally pretreated silica surface, which presents an interesting observation in view of surface

organometallic synthesis concepts.

Confinement. The results from our catalytic model-study (Chapter 4) point towards

confinement effects as one reason for the difference in activity between benchmark Snβ and

silica-grafted SnIV-sites. In order to confirm this contribution we extend our model-study by

grafting SnIV-sites on to a mesoporous MCM-41-support (Chapter 5). In line with our previous

findings, we show that the activity of SnIV/SiO2-based catalysts in the Meerwein-Ponndorf-

Verley reduction of cyclohexanone with 2-butanol strongly depends on the pore size of the

employed SiO2-based-support. We attribute this trend in reactivity with an adsorption-based

confinement effect, which most probably results from a suppressed reverse reaction in more

confined environments.

6.2. Outlook

The work described in this thesis has led to achievements in the large-scale applicability of

state-of-the-art Snβ zeolite, and the distinction of material properties that govern the activity

of Lewis acid SnIV/SiO2-based catalysts. Nonetheless, there are several aspects that need

further dedication.

For example, the synthesis of a catalyst for industrial purposes is always done in the context

of the overall process optimization, giving consideration to the utilization of the catalyst in a

flow reactor in order to facilitate continuous operation and to reduce catalyst recuperation

cost. For systems, where consecutive reactions of the product cause a selectivity decrease

at increasing conversion, the overall performance of the catalyst can be expected to increase

under continuous flow conditions. Emphasizing the industrial importance of the Baeyer-Villiger

oxidation of cyclohexanone to ε-caprolactone (a momomer intermediate) and stimulated by

the promising results with our post-synthetically prepared Snβ, we believe that it is of general

interest to include aspects of reaction engineering into future work on this system. A preliminary

66 Chapter 6

experimental test investigating the Baeyer-Villiger oxidation with 10 wt% Snβ under continuous

operation shows stability of the catalyst over 4 days (Figure 6.1), although at lower conversion

and yield levels than those achieved under batch conditions, which requires further optimizing

studies. In addition, the possibility to utilize binders as catalyst carrier would be an interesting

aspect to include into future work, aiming toward higher mechanical stability.

Figure 6.1. Cyclohexanone conversion and ε-caprolactone yield in a continuous Baeyer-Villigerexperiment. Experimental conditions: cyclohexanone in 1,4-dioxane (0.33 M), 80 °C, H2O2 (50 wt%aq. solution, H2O2/ketone = 0.8), 10 wt% Snβ.

Another example is the remaining question if the different proposed types of active SnIV-sites

for Snβ (i.e., open or closed) correlate with different individual activities. Indeed, our catalytic

model-study (Chapter 4) suggests that open sites ((SiO)xSn(OH)4-x) are not necessarily more

active than closed sites as is the current hypothesis for Snβ. Our Baeyer-Villiger study

(Chapter 3), on the other hand, indicates that both open and closed sites are active, which

questions that site structure has a prevailing influence on activity in this reaction system. The

relation between active site structures and catalytic activities is, hence, strongly dependent on

the characteristics of individual reaction systems. We believe that this topic, which has already

been the object of several investigations, will continue interesting academic research in order to

improve the rational design of novel materials. One promising approach in this context is the

conjunction of experimental and computational studies based on different Snβ catalysts and

probe reactions. The comparison of structure-activity information gained for various reaction

systems will reveal contrary and/or similar trends, which will help to further elucidate the

catalytic behavior of Snβ in different types of reactions and improve our understanding of the

Conclusions and Outlook 67

impact of site structure on catalytic activities for this catalyst. Another intriguing approach in

this field would be to deploy a methodology introduced by Notestein and co-workers, involving

the "counting" of active sites by selective poisoning of TiIV-sites with hydroxyl groups, which

results in "true turn over frequencies" that are normalized on an active-site-basis.[187] With an

appropriate poisoning agent, this could be applied to SnIV-sites with hydroxyl groups (i.e., open

sites) in different Snβ catalysts with varying active site distributions and activities, respectively,

and allow the establishment of structure-activity-relationships.

Moreover, in terms of understanding the influence of catalyst hydrophilicity on catalyst

activity in the Baeyer-Villiger oxidation of cyclohexanone with hydrogen peroxide, this thesis

has only been exploring the implications on the observed performance, but not in view to

possible underlying mechanistic reasons. Thus, it could be envisaged to expand the scope of

future research considering this aspect. As such, it is conceivable that silanol groups do not only

steer the hydrophilicity, and hereby influence occuring catalyst inhibition but also participate

in the mechanistic cycle, possibly by activating the peroxide substrate.

Not least, our work demonstrates various pathways by which a metal precursor can react

with the functional groups found on silica surfaces (i.e., via reaction with silanol groups

and/or via reaction with siloxane bridges of the type ≡Si–O–Si≡). An important question

in this regard is which aspects of the metal precursor govern such surface anchoring and

subsequent restructuring mechanisms. We believe that targeted experimental studies with

various metal precursors (and ligands) will help to further elucidate this interesting material

synthesis question and result in a better understanding of catalyst preparation by means of

grafting and impregnation chemistry.

Bibliography

[1] H. Yamamoto, Lewis Acids in Organic Synthesis, vol. 1. Wiley-VCH Verlag GmbH, 2000.

[2] M. Santelli and J.-M. Pons, Lewis Acids and Selectivity in Organic Synthesis, vol. 2. CRCPress, 1995.

[3] D. Schinzer, Selectivities in Lewis Acid Promoted Reactions. Kluwer Academic PublishersDordrecht:, 1989.

[4] G. N. Lewis, Valence and the Structure of Atoms and Molecules. Chemical Catalog Com.;New York, 1923.

[5] A. Corma and H. Garcia, “Lewis Acids: From Conventional Homogeneous to GreenHeterogeneous Catalysis,” Chemical Reviews, vol. 103, no. 11, pp. 4307–4366, 2003.

[6] A. Corma, “Attempts to Fill the Gap between Enzymatic, Homogeneous, andHeterogeneous Catalysis,” Catalysis Reviews, vol. 46, no. 3-4, pp. 369–417, 2004.

[7] G. A. Olah, Friedel-Crafts Chemistry. World Scientific, 1973.

[8] C. Erre and C. Roussel, “Friedel-Crafts Tetraacylation of Isobutene Precursors: Scope andLimitations of a New and General Synthesis of 1,3,6,8-Tetraalkyl-2,7-Naphthyridines,”Bulletin de la Société Chimique de France, no. 11-12, pp. 449–453, 1984.

[9] T. Werpy, G. Petersen, A. Aden, J. Bozell, J. Holladay, J. White, A. Manheim, D. Eliot,L. Lasure, and S. Jones, “Top Value Added Chemicals from Biomass. Volume 1 – Resultsof Screening for Potential Candidates from Sugars and Synthesis Gas.” DTIC Document,2004.

[10] D. L. Klass, Biomass for Renwable Energy, Fuels, and Chemicals. Academic press, 1998.

[11] A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J. Cairney, C. A. Eckert,W. J. Frederick, J. P. Hallett, D. J. Leak, and C. L. Liotta, “The Path Forward forBiofuels and Biomaterials,” Science, vol. 311, no. 5760, pp. 484–489, 2006.

[12] J. N. Chheda, G. W. Huber, and J. A. Dumesic, “Liquid-Phase Catalytic Processingof Biomass-Derived Oxygenated Hydrocarbons to Fuels and Chemicals,” AngewandteChemie International Edition, vol. 46, no. 38, pp. 7164–7183, 2007.

[13] M. J. Climent, A. Corma, and S. Iborra, “Converting Carbohydrates to Bulk Chemicalsand Fine Chemicals Over Heterogeneous Catalysts,” Green Chemistry, vol. 13, no. 3,pp. 520–540, 2011.

70 Bibliography

[14] S. Kobayashi and K. Manabe, “Development of Novel Lewis Acid Catalysts for SelectiveOrganic Reactions in Aqueous Media,” Accounts of Chemical Research, vol. 35, no. 4,pp. 209–217, 2002.

[15] C.-J. Li, “Organic Reactions in Aqueous Media with a Focus on Carbon-Carbon BondFormations: A Decade Update,” Chemical Reviews, vol. 105, no. 8, pp. 3095–3166, 2005.

[16] Y. Román-Leshkov and M. E. Davis, “Activation of Carbonyl-Containing Molecules withSolid Lewis Acids in Aqueous Media,” ACS Catalysis, vol. 1, no. 11, pp. 1566–1580, 2011.

[17] R. Gounder, “Hydrophobic Microporous and Mesoporous Oxides as Brønsted and Lewisacid catalysts for Biomass Conversion in Liquid Water,” Catalysis Science & Technology,vol. 4, no. 9, pp. 2877–2886, 2014.

[18] M. Taramasso, G. Perego, and B. Notari, “Preparation of Porous Crystalline SyntheticMaterial Comprised of Silicon and Titanium Oxides,” Oct. 18 1983. US Patent 4,410,501.

[19] D. Huybrechts, L. De Bruycker, and P. Jacobs, “Oxyfunctionalization of Alkanes withHydrogen Peroxide on Titanium Silicalite,” Nature, vol. 345, no. 6272, pp. 240–242, 1990.

[20] F. Cavani and J. H. Teles, “Sustainability in Catalytic Oxidation: An AlternativeApproach or a Structural Evolution?,” ChemSusChem, vol. 2, no. 6, pp. 508–534, 2009.

[21] M. Boronat, P. Concepción, A. Corma, and M. Renz, “Peculiarities of Sn-Beta andPotential Industrial Applications,” Catalysis Today, vol. 121, no. 1, pp. 39–44, 2007.

[22] J. Ritz, H. Fuchs, H. Kieczka, and W. C. Moran, “Caprolactam,” Ullmann’s Encyclopediaof Industrial Chemistry, 2001.

[23] M. A. Woodruff and D. W. Hutmacher, “The Return of a Forgotten Polymer –Polycaprolactone in the 21st Century,” Progress in Polymer Science, vol. 35, no. 10,pp. 1217–1256, 2010.

[24] C. Jiménez-Sanchidrián and J. R. Ruiz, “The Baeyer-Villiger Reaction on HeterogeneousCatalysts,” Tetrahedron, vol. 64, no. 9, pp. 2011–2026, 2008.

[25] G.-J. Ten Brink, I. Arends, and R. Sheldon, “The Baeyer-Villiger Reaction: NewDevelopments Toward Greener Procedures,” Chemical Reviews, vol. 104, no. 9, pp. 4105–4124, 2004.

[26] R. Bermejo-Deval, R. S. Assary, E. Nikolla, M. Moliner, Y. Román-Leshkov, S.-J.Hwang, A. Palsdottir, D. Silverman, R. F. Lobo, and L. A. Curtiss, “Metalloenzyme-LikeCatalyzed Isomerizations of Sugars by Lewis Acid Zeolites,” Proceedings of the NationalAcademy of Sciences, vol. 109, no. 25, pp. 9727–9732, 2012.

Bibliography 71

[27] M. Moliner, Y. Román-Leshkov, and M. E. Davis, “Tin-Containing Zeolites are HighlyActive Catalysts for the Isomerization of Glucose in Water,” Proceedings of the NationalAcademy of Sciences, vol. 107, no. 14, pp. 6164–6168, 2010.

[28] C. M. Lew, N. Rajabbeigi, and M. Tsapatsis, “Tin-Containing Zeolite for theIsomerization of Cellulosic Sugars,” Microporous and Mesoporous Materials, vol. 153,pp. 55–58, 2012.

[29] W. R. Gunther, Y. Wang, Y. Ji, V. K. Michaelis, S. T. Hunt, R. G. Griffin, andY. Román-Leshkov, “Sn-Beta Zeolites with Borate Salts Catalyse the Epimerization ofCarbohydrates via an Intramolecular Carbon Shift,” Nature Communications, vol. 3,p. 1109, 2012.

[30] M. S. Holm, S. Saravanamurugan, and E. Taarning, “Conversion of Sugars to LacticAcid Derivatives Using Heterogeneous Zeotype Catalysts,” Science, vol. 328, no. 5978,pp. 602–605, 2010.

[31] Q. Guo, F. Fan, E. A. Pidko, W. N. van der Graaff, Z. Feng, C. Li, and E. J. Hensen,“Highly Active and Recyclable Sn-MWW Zeolite Catalyst for Sugar Conversion to MethylLactate and Lactic Acid,” ChemSusChem, vol. 6, no. 8, pp. 1352–1356, 2013.

[32] F. de Clippel, M. Dusselier, R. Van Rompaey, P. Vanelderen, J. Dijkmans, E. Makshina,L. Giebeler, S. Oswald, G. V. Baron, and J. F. Denayer, “Fast and Selective SugarConversion to Alkyl Lactate and Lactic Acid with Bifunctional Carbon-Silica Catalysts,”Journal of the American Chemical Society, vol. 134, no. 24, pp. 10089–10101, 2012.

[33] P. Wolf, C. Hammond, S. Conrad, and I. Hermans, “Post-Synthetic Preparation of Sn-,Ti-and Zr-Beta: A Facile Route to Water Tolerant, Highly Active Lewis Acidic Zeolites,”Dalton Transactions, vol. 43, no. 11, pp. 4514–4519, 2014.

[34] V. Choudhary, A. B. Pinar, S. I. Sandler, D. G. Vlachos, and R. F. Lobo, “XyloseIsomerization to Xylulose and its Dehydration to Furfural in Aqueous Media,” ACSCatalysis, vol. 1, no. 12, pp. 1724–1728, 2011.

[35] E. Nikolla, Y. Román-Leshkov, M. Moliner, and M. E. Davis, “"One-pot" Synthesis of5-(Hydroxymethyl)furfural from Carbohydrates Using Tin-Beta Zeolite,” ACS Catalysis,vol. 1, no. 4, pp. 408–410, 2011.

[36] C. Aellig and I. Hermans, “Continuous D-Fructose Dehydration to 5-Hydroxymethylfurfural under Mild Conditions,” ChemSusChem, vol. 5, no. 9,pp. 1737–1742, 2012.

72 Bibliography

[37] L. Hu, G. Zhao, W. Hao, X. Tang, Y. Sun, L. Lin, and S. Liu, “Catalytic Conversion ofBiomass-Derived Carbohydrates into Fuels and Chemicals via Furanic Aldehydes,” RSCAdvances, vol. 2, no. 30, pp. 11184–11206, 2012.

[38] X. Tong, Y. Ma, and Y. Li, “Biomass into Chemicals: Conversion of Sugars to FuranDerivatives by Catalytic Processes,” Applied Catalysis A: General, vol. 385, no. 1, pp. 1–13, 2010.

[39] A. Corma, M. E. Domine, L. Nemeth, and S. Valencia, “Al-Free Sn-Beta Zeolite as aCatalyst for the Selective Reduction of Carbonyl Compounds (Meerwein-Ponndorf-Verleyreaction),” Journal of the American Chemical Society, vol. 124, no. 13, pp. 3194–3195,2002.

[40] C. Jiménez-Sanchidrián and J. R. Ruiz, “Tin-Containing Hydrotalcite-Like Compoundsas Catalysts for the Meerwein-Ponndorf-Verley Reaction,” Applied Catalysis A: General,vol. 469, pp. 367–372, 2014.

[41] G. Neri, G. Rizzo, C. Crisafulli, L. De Luca, A. Donato, M. Musolino, R. Pietropaolo,et al., “One-Pot Synthesis of Naturanol from α-Pinene Oxide on Bifunctional Pt-Sn/SiO2

Heterogeneous Catalysts: Part I: The Catalytic System,” Applied Catalysis A: General,vol. 325, no. 1, pp. 15–24, 2007.

[42] G. Neri, G. Rizzo, A. Pistone, L. De Luca, A. Donato, M. Musolino, and R. Pietropaolo,“One-Pot Synthesis of Naturanol from α-Pinene Oxide on Bifunctional HeterogeneousCatalysts: Part II. The Reaction,” Applied Catalysis A: General, vol. 325, no. 1, pp. 25–33, 2007.

[43] O. de la Torre, M. Renz, and A. Corma, “Biomass to Chemicals: Rearrangement of β-Pinene Epoxide into Myrtanal with Well-Defined Single-Site Substituted Molecular Sievesas Reusable Solid Lewis-Acid Catalysts,” Applied Catalysis A: General, vol. 380, no. 1,pp. 165–171, 2010.

[44] A. Corma and M. Renz, “Sn-Beta Zeolite as Diastereoselective Water-ResistantHeterogeneous Lewis-Acid Catalyst for Carbon-Carbon Bond Formation in theIntramolecular Carbonyl-Ene Reaction,” Chemical Communications, no. 5, pp. 550–551,2004.

[45] S. Van de Vyver, C. Odermatt, K. Romero, T. Prasomsri, and Y. Román-Leshkov,“Solid Lewis Acids Catalyze the Carbon-Carbon Coupling Between Carbohydrates andFormaldehyde,” ACS Catalysis, vol. 5, no. 2, pp. 972–977, 2015.

[46] N. Mal and A. Ramaswamy, “Synthesis and Catalytic Properties of Large-Pore Sn-β andAl-Free Sn-β Molecular Sieves,” Chemical Communications, no. 5, pp. 425–426, 1997.

Bibliography 73

[47] L. Nemeth, J. Moscoso, N. Erdman, S. Bare, A. Oroskar, S. Kelly, A. Corma, S. Valencia,and M. Renz, “Synthesis and Characterization of Sn-Beta as a Selective OxidationCatalyst,” Studies in Surface Science and Catalysis, vol. 154, pp. 2626–2631, 2004.

[48] A. Corma, L. T. Nemeth, M. Renz, and S. Valencia, “Sn-Zeolite Beta as a HeterogeneousChemoselective Catalyst for Baeyer–Villiger Oxidations,” Nature, vol. 412, no. 6845,pp. 423–425, 2001.

[49] M. E. Davis and R. F. Lobo, “Zeolite and Molecular Sieve Synthesis,” Chemistry ofMaterials, vol. 4, no. 4, pp. 756–768, 1992.

[50] C. S. Cundy and P. A. Cox, “The Hydrothermal Synthesis of Zeolites: History andDevelopment from the Earliest Days to the Present Time,” Chemical Reviews, vol. 103,no. 3, pp. 663–702, 2003.

[51] S. V. Valencia and A. C. Canós, “Stannosilicate Molecular Sieves,” Oct. 23 2001. USPatent 6,306,364.

[52] P. Wu, T. Tatsumi, T. Komatsu, and T. Yashima, “A Novel Titanosilicate with MWWStructure. I. Hydrothermal Synthesis, Elimination of Extraframework Titanium, andCharacterizations,” The Journal of Physical Chemistry B, vol. 105, no. 15, pp. 2897–2905, 2001.

[53] T.-K. Kim, S.-T. Yang, D. R. Park, I. K. Song, K.-E. Jung, and W.-S. Ahn, “Ti-MWWSynthesis and Catalytic Applications in Partial Oxidation Reactions,” Topics in Catalysis,vol. 53, no. 7-10, pp. 470–478, 2010.

[54] A. Corma, M. T. Navarro, L. Nemeth, and M. Renz, “Sn-MCM-41 – a HeterogeneousSelective Catalyst for the Baeyer-Villiger Oxidation with Hydrogen Peroxide,” ChemicalCommunications, no. 21, pp. 2190–2191, 2001.

[55] A. Corma, S. Iborra, M. Mifsud, and M. Renz, “Mesoporous Molecular Sieve Sn-MCM-41 as Baeyer-Villiger Oxidation Catalyst for Sterically Demanding Aromatic and α, β-Unsaturated Aldehydes,” Arkivoc, vol. 9, pp. 124–132, 2005.

[56] K. Chaudhari, T. Das, P. Rajmohanan, K. Lazar, S. Sivasanker, and A. Chandwadkar,“Synthesis, Characterization, and Catalytic Properties of Mesoporous Tin-ContainingAnalogs of MCM-41,” Journal of Catalysis, vol. 183, no. 2, pp. 281–291, 1999.

[57] T. Gaydhankar, P. Joshi, P. Kalita, and R. Kumar, “Optimal Synthesis Parametersand Application of Sn-MCM-41 as an Efficient Heterogeneous Catalyst in Solvent-FreeMukaiyama-Type Aldol Condensation,” Journal of Molecular Catalysis A: Chemical,vol. 265, no. 1, pp. 306–315, 2007.

74 Bibliography

[58] H. J. Cho, P. Dornath, and W. Fan, “Synthesis of Hierarchical Sn-MFI as Lewis AcidCatalysts for Isomerization of Cellulosic Sugars,” ACS Catalysis, vol. 4, no. 6, pp. 2029–2037, 2014.

[59] L. Li, C. Stroobants, K. Lin, P. A. Jacobs, B. F. Sels, and P. P. Pescarmona, “SelectiveConversion of Trioses to Lactates over Lewis Acid Heterogeneous Catalysts,” GreenChemistry, vol. 13, no. 5, pp. 1175–1181, 2011.

[60] C. M. Osmundsen, M. S. Holm, S. Dahl, and E. Taarning, “Tin-Containing Silicates:Structure-Activity Relations,” in Proceedings of the Royal Society of London A:Mathematical, Physical and Engineering Sciences, vol. 468, pp. 2000–2016, The RoyalSociety, 2012.

[61] R. K. Grasselli, “Genesis of Site Isolation and Phase Cooperation in Selective OxidationCatalysis,” Topics in Catalysis, vol. 15, no. 2-4, pp. 93–101, 2001.

[62] J. M. Thomas, “The Societal Significance of Catalysis and the Growing PracticalImportance of Single-Site Heterogeneous Catalysts,” in Proceedings of the Royal Societyof London A: Mathematical, Physical and Engineering Sciences, p. rspa20120196, TheRoyal Society, 2012.

[63] T. Hara, M. Hatakeyama, A. Kim, N. Ichikuni, and S. Shimazu, “Preparation of Clay-Supported Sn Catalysts and Application to Baeyer-Villiger Oxidation,” Green Chemistry,vol. 14, no. 3, pp. 771–777, 2012.

[64] A. Corma, M. T. Navarro, and M. Renz, “Lewis acidic Sn(IV) centers – grafted ontoMCM-41 – as Catalytic Sites for the Baeyer-Villiger Oxidation with Hydrogen Peroxide,”Journal of Catalysis, vol. 219, no. 1, pp. 242–246, 2003.

[65] A. Tribalis, G. D. Panagiotou, G. Tsilomelekis, A. G. Kalampounias, K. Bourikas,C. Kordulis, S. Boghosian, and A. Lycourghiotis, “Temperature-Dependent Evolutionof the Molecular Configuration of Oxo-Tungsten(VI) Species Deposited on the Surface ofTitania,” The Journal of Physical Chemistry C, vol. 118, no. 21, pp. 11319–11332, 2014.

[66] J. Joubert, F. Delbecq, P. Sautet, E. L. Roux, M. Taoufik, C. Thieuleux, F. Blanc,C. Copéret, J. Thivolle-Cazat, and J.-M. Basset, “Molecular Understanding of AluminaSupported Single-Site Catalysts by a Combination of Experiment and Theory,” Journalof the American Chemical Society, vol. 128, no. 28, pp. 9157–9169, 2006.

[67] S. L. Wegener, T. J. Marks, and P. C. Stair, “Design Strategies for the Molecular LevelSynthesis of Supported Catalysts,” Accounts of Chemical Research, vol. 45, no. 2, pp. 206–214, 2011.

Bibliography 75

[68] E. A. Pidko, S. M. Almutairi, B. Mezari, P. C. Magusin, and E. J. Hensen, “ChemicalVapor Deposition of Trimethylaluminum on Dealuminated Faujasite Zeolite,” ACSCatalysis, vol. 3, no. 7, pp. 1504–1517, 2013.

[69] H. Shintaku, K. Nakajima, M. Kitano, N. Ichikuni, and M. Hara, “Lewis Acid Catalysis ofTiO4 Tetrahedra on Mesoporous Silica in Water,” ACS Catalysis, vol. 4, no. 4, pp. 1198–1204, 2014.

[70] P. Li, G. Liu, H. Wu, Y. Liu, J.-g. Jiang, and P. Wu, “Postsynthesis and SelectiveOxidation Properties of Nanosized Sn-Beta Zeolite,” The Journal of Physical ChemistryC, vol. 115, no. 9, pp. 3663–3670, 2011.

[71] B. Tang, W. Dai, G. Wu, N. Guan, L. Li, and M. Hunger, “Improved PostsynthesisStrategy to Sn-Beta Zeolites as Lewis Acid Catalysts for the Ring-Opening Hydration ofEpoxides,” ACS Catalysis, vol. 4, no. 8, pp. 2801–2810, 2014.

[72] W. N. van der Graaff, G. Li, B. Mezari, E. A. Pidko, and E. J. Hensen, “Synthesis ofSn-Beta with Exclusive and High Framework Sn Content,” ChemCatChem, vol. 7, no. 7,pp. 1152–1160, 2015.

[73] J. Dijkmans, D. Gabriëls, M. Dusselier, F. de Clippel, P. Vanelderen, K. Houthoofd,A. Malfliet, Y. Pontikes, and B. F. Sels, “Productive Sugar Isomerization with HighlyActive Sn in Dealuminated β Zeolites,” Green Chemistry, vol. 15, no. 10, pp. 2777–2785,2013.

[74] C.-C. Chang, H. J. Cho, Z. Wang, X. Wang, and W. Fan, “Fluoride-Free Synthesis of aSn-BEA Catalyst by Dry Gel Conversion,” Green Chemistry, vol. 17, no. 5, pp. 2943–2951, 2015.

[75] P. Y. Dapsens, C. Mondelli, J. Jagielski, R. Hauert, and J. Pérez-Ramírez, “HierarchicalSn-MFI Zeolites Prepared by Facile Top-Down Methods for Sugar Isomerisation,”Catalysis Science & Technology, vol. 4, no. 8, pp. 2302–2311, 2014.

[76] C. Copéret, M. Chabanas, R. Petroff Saint-Arroman, and J.-M. Basset, “Homogeneousand Heterogeneous Catalysis: Bridging the Gap Through Surface OrganometallicChemistry,” Angewandte Chemie International Edition, vol. 42, no. 2, pp. 156–181, 2003.

[77] D. Gajan and C. Copéret, “Silica-Supported Single-Site Catalysts: To Be or Not To Be?A Conjecture on Silica Surfaces,” New Journal of Chemistry, vol. 35, no. 11, pp. 2403–2408, 2011.

[78] F. Rascón, R. Wischert, and C. Copéret, “Molecular Nature of Support Effects in Single-Site Heterogeneous Catalysts: Silica vs. Alumina,” Chemical Science, vol. 2, no. 8,pp. 1449–1456, 2011.

76 Bibliography

[79] T. J. Marks, “Surface-Bound Metal Hydrocarbyls. Organometallic Connections BetweenHeterogeneous and Homogeneous Catalysis,” Accounts of Chemical Research, vol. 25,no. 2, pp. 57–65, 1992.

[80] J. M. Thomas, R. Raja, and D. W. Lewis, “Single-Site Heterogeneous Catalysts,”Angewandte Chemie International Edition, vol. 44, no. 40, pp. 6456–6482, 2005.

[81] E. A. Quadrelli and J.-M. Basset, “On Silsesquioxanes’ Accuracy as Molecular Models forSilica-Grafted Complexes in Heterogeneous Catalysis,” Coordination Chemistry Reviews,vol. 254, no. 5, pp. 707–728, 2010.

[82] A. Rimola, D. Costa, M. Sodupe, J.-F. Lambert, and P. Ugliengo, “Silica SurfaceFeatures and Their Role in the Adsorption of Biomolecules: Computational Modelingand Experiments,” Chemical Reviews, vol. 113, no. 6, pp. 4216–4313, 2013.

[83] R. Anwander, “Immobilization of Molecular Catalysts,” Handbook of HeterogeneousCatalysis, 2008.

[84] F. Averseng, M. Vennat, and M. Che, “Grafting and Anchoring of Transition MetalComplexes to Inorganic Oxides,” Handbook of Heterogeneous Catalysis, 2008.

[85] J.-M. Basset, F. Lefebvre, and C. Santini, “Surface Organometallic Chemistry: SomeFundamental Features Including the Coordination Effects of the Support,” CoordinationChemistry Reviews, vol. 178, pp. 1703–1723, 1998.

[86] L. Zhuravlev, “The Surface Chemistry of Amorphous Silica. Zhuravlev Model,” Colloidsand Surfaces A: Physicochemical and Engineering Aspects, vol. 173, no. 1, pp. 1–38, 2000.

[87] V. Lygin and O. Chertihina, “The Structure and Properties of the Dehydroxylated SilicaSurface: Quantum-Chemical Studies,” Colloids and Surfaces, vol. 63, no. 1, pp. 93–95,1992.

[88] P. Mania, R. Verel, F. Jenny, C. Hammond, and I. Hermans, “Thermal Restructuring ofSilica-Grafted TiClx Species and Consequences for Epoxidation Catalysis,” Chemistry-AEuropean Journal, vol. 19, no. 30, pp. 9849–9858, 2013.

[89] P. Mania, S. Conrad, R. Verel, C. Hammond, and I. Hermans, “Thermal Restructuringof Silica-Grafted –CrO2Cl and –VOCl2 species,” Dalton Transactions, vol. 42, no. 35,pp. 12725–12732, 2013.

[90] M. P. Conley, A. J. Rossini, A. Comas-Vives, M. Valla, G. Casano, O. Ouari, P. Tordo,A. Lesage, L. Emsley, and C. Copéret, “Silica-surface Reorganization During OrganotinGrafting Evidenced by 119Sn DNP SENS: A Tandem Reaction of Gem-Silanols andStrained Siloxane Bridges,” Physical Chemistry Chemical Physics, vol. 16, no. 33,pp. 17822–17827, 2014.

Bibliography 77

[91] G. L. Rice and S. L. Scott, “Characterization of Silica-Supported Vanadium(V)Complexes Derived from Molecular Precursors and Their Ligand Exchange Reactions,”Langmuir, vol. 13, no. 6, pp. 1545–1551, 1997.

[92] A. Hamieh, Y. Chen, S. Abdel-Azeim, E. Abou-Hamad, S. Goh, M. Samantaray,R. Dey, L. Cavallo, and J. M. Basset, “Well-Defined Surface Species Prepared by DirectMethylation, a Catalyst for Cycloalkane Metathesis and Transformation of Ethylene toPropylene,” ACS Catalysis, vol. 5, no. 4, pp. 2164–2171, 2015.

[93] E. Le Roux, M. Chabanas, A. Baudouin, A. de Mallmann, C. Copéret, E. A. Quadrelli,J. Thivolle-Cazat, J.-M. Basset, W. Lukens, A. Lesage, et al., “Detailed StructuralInvestigation of the Grafting of [Ta(CH)tBu)(CH2

tBu)3] and [Cp*TaMe4] on SilicaPartially Dehydroxylated at 700 ◦ C and the Activity of the Grafted Complexes TowardAlkane Metathesis,” Journal of the American Chemical Society, vol. 126, no. 41,pp. 13391–13399, 2004.

[94] C. Alonso, A. Antiñolo, F. Carrillo-Hermosilla, P. Carrión, A. Otero, J. Sancho, andE. Villaseñor, “Modified Silicas as Supports for Single-Site Zirconocene Catalysts,”Journal of Molecular Catalysis A: Chemical, vol. 220, no. 2, pp. 286–296, 2004.

[95] C. Tiozzo, C. Bisio, F. Carniato, A. Gallo, S. L. Scott, R. Psaro, and M. Guidotti,“Niobium-Silica Catalysts for the Selective Epoxidation of Cyclic Alkenes: TheGeneration of the Active Site by Grafting Niobocene Dichloride,” Physical ChemistryChemical Physics, vol. 15, no. 32, pp. 13354–13362, 2013.

[96] V. Riollet, E. A. Quadrelli, C. Copéret, J.-M. Basset, R. A. Andersen, K. Köhler,R.-M. Böttcher, and E. Herdtweck, “Grafting of [Mn(CH2

tBu)2(tmeda)] on Silica andComparison with its Reaction with a Silsesquioxane,” Chemistry-A European Journal,vol. 11, no. 24, pp. 7358–7365, 2005.

[97] C. Copéret and B. Chaudret, Surface and Interfacial Organometallic Chemistry andCatalysis, vol. 16. Springer Science; Business Media, 2005.

[98] V. Vidal, A. Théolier, J. Thivolle-Cazat, and J.-M. Basset, “Metathesis of AlkanesCatalyzed by Silica-Supported Transition Metal Hydrides,” Science, vol. 276, no. 5309,pp. 99–102, 1997.

[99] F. Lefebvre and J.-M. Basset, “Recent Applications in Catalysis of Surface OrganometallicChemistry,” Journal of Molecular Catalysis A: Chemical, vol. 146, no. 1, pp. 3–12, 1999.

[100] V. Mougel and C. Copéret, “Isostructural Molecular and Surface Mimics of the ActiveSites of the Industrial WO3/SiO2 Metathesis Catalysts,” ACS Catalysis, vol. 5, no. 11,pp. 6436–6439, 2015.

78 Bibliography

[101] M. Boronat, P. Concepción, A. Corma, M. Renz, and S. Valencia, “Determination ofthe Catalytically Active Oxidation Lewis Acid Sites in Sn-Beta Zeolites, and TheirOptimisation by the Combination of Theoretical and Experimental Studies,” Journalof Catalysis, vol. 234, no. 1, pp. 111–118, 2005.

[102] P. J. Cordeiro and T. D. Tilley, “Enhancement of Epoxidation Efficiencies for Ta-SBA15 Catalysts. The Influence of Modification with –EMe3 (E = Si, Ge, Sn) Groups,”Langmuir, vol. 27, no. 10, pp. 6295–6304, 2011.

[103] M. Boronat, A. Corma, M. Renz, and P. M. Viruela, “Predicting the Activity of SingleIsolated Lewis Acid Sites in Solid Catalysts,” Chemistry-A European Journal, vol. 12,no. 27, pp. 7067–7077, 2006.

[104] J. Dijkmans, M. Dusselier, W. Janssens, M. Trekels, A. Vantomme, E. Breynaert,C. Kirschhock, and B. F. Sels, “An inner-/outer-sphere stabilized sn active site in β-zeolite: Spectroscopic evidence and kinetic consequences,” ACS Catalysis, vol. 6, pp. 31–46, 2015.

[105] D. G. Hanna, S. Shylesh, Y.-P. Li, S. Krishna, M. Head-Gordon, and A. T. Bell,“Experimental and Theoretical Study of n-Butanal Self-Condensation over Ti SpeciesSupported on Silica,” ACS Catalysis, vol. 4, no. 9, pp. 2908–2916, 2014.

[106] S. R. Bare, S. D. Kelly, W. Sinkler, J. J. Low, F. S. Modica, S. Valencia, A. Corma,and L. T. Nemeth, “Uniform Catalytic Site in Sn-β-Zeolite Determined Using X-RayAbsorption Fine Structure,” Journal of the American Chemical Society, vol. 127, no. 37,pp. 12924–12932, 2005.

[107] G. Sastre and A. Corma, “Relation Between Structure and Lewis Acidity of Ti-Beta andTS-1 Zeolites: A Quantum-Chemical Study,” Chemical Physics Letters, vol. 302, no. 5,pp. 447–453, 1999.

[108] A. Corma, M. E. Domine, and S. Valencia, “Water-Resistant Solid Lewis Acid Catalysts:Meerwein-Ponndorf-Verley and Oppenauer Reactions Catalyzed by Tin-Beta Zeolite,”Journal of Catalysis, vol. 215, no. 2, pp. 294–304, 2003.

[109] R. Gounder and M. E. Davis, “Monosaccharide and Disaccharide Isomerization overLewis Acid Sites in Hydrophobic and Hydrophilic Molecular Sieves,” Journal of Catalysis,vol. 308, pp. 176–188, 2013.

[110] R. Gounder and E. Iglesia, “The Catalytic Diversity of Zeolites: Confinement andSolvation Effects within Voids of Molecular Dimensions,” Chemical Communications,vol. 49, no. 34, pp. 3491–3509, 2013.

Bibliography 79

[111] M. Boronat, P. Concepción, A. Corma, M. T. Navarro, M. Renz, and S. Valencia,“Reactivity in the Confined Spaces of Zeolites: The Interplay between Spectroscopy andTheory to Develop Structure–Activity Relationships for Catalysis,” Physical ChemistryChemical Physics, vol. 11, no. 16, pp. 2876–2884, 2009.

[112] G. Sastre and A. Corma, “The Confinement Effect in Zeolites,” Journal of MolecularCatalysis A: Chemical, vol. 305, no. 1, pp. 3–7, 2009.

[113] M. Dusselier, P. Van Wouwe, A. Dewaele, P. A. Jacobs, and B. F. Sels, “Shape-selectiveZeolite Catalysis for Bioplastics Production,” Science, vol. 349, no. 6243, pp. 78–80, 2015.

[114] B. Smit and T. L. Maesen, “Towards a Molecular Understanding of Shape Selectivity,”Nature, vol. 451, no. 7179, pp. 671–678, 2008.

[115] M. Renz, T. Blasco, A. Corma, V. Fornés, R. Jensen, and L. Nemeth, “Selective andShape-Selective Baeyer-Villiger Oxidations of Aromatic Aldehydes and Cyclic Ketoneswith Sn-Beta Zeolites and H2O2,” Chemistry-A European Journal, vol. 8, no. 20, pp. 4708–4717, 2002.

[116] R. Bermejo-Deval, R. Gounder, and M. E. Davis, “Framework and Extraframework TinSites in Zeolite Beta React Glucose Differently,” ACS Catalysis, vol. 2, no. 12, pp. 2705–2713, 2012.

[117] P. Wolf, M. Valla, A. J. Rossini, A. Comas-Vives, F. Núñez Zarur, B. Malaman, A. Lesage,L. Emsley, C. Copéret, and I. Hermans, “NMR Signatures of the Active Sites in Sn-bZeolite,” Angewandte Chemie International Edition, vol. 53, pp. 1–6, 2014.

[118] R. Bermejo-Deval, M. Orazov, R. Gounder, S.-J. Hwang, and M. E. Davis, “Active Sitesin Sn-Beta for Glucose Isomerization to Fructose and Epimerization to Mannose,” ACSCatalysis, vol. 4, no. 7, pp. 2288–2297, 2014.

[119] J. Dijkmans, M. Dusselier, D. Gabriëls, K. Houthoofd, P. C. Magusin, S. Huang,Y. Pontikes, M. Trekels, A. Vantomme, and L. Giebeler, “Cooperative Catalysis forMultistep Biomass Conversion with Sn/Al Beta Zeolite,” ACS Catalysis, vol. 5, no. 2,pp. 928–940, 2015.

[120] S.-J. Hwang, R. Gounder, Y. Bhawe, M. Orazov, R. Bermejo-Deval, and M. E.Davis, “Solid State NMR Characterization of Sn-Beta Zeolites that Catalyze GlucoseIsomerization and Epimerization,” Topics in Catalysis, vol. 58, no. 7-9, pp. 435–440,2015.

[121] W. R. Gunther, V. K. Michaelis, M. A. Caporini, R. G. Griffin, and Y. Román-Leshkov,“Dynamic Nuclear Polarization NMR Enables the Analysis of Sn-Beta Zeolite Prepared

80 Bibliography

with Natural Abundance 119Sn Precursors,” Journal of the American Chemical Society,vol. 136, no. 17, pp. 6219–6222, 2014.

[122] M. Boronat, A. Corma, M. Renz, G. Sastre, and P. M. Viruela, “A Multisite MolecularMechanism for Baeyer-Villiger Oxidations on Solid Catalysts Using EnvironmentallyFriendly H2O2 as Oxidant,” Chemistry-A European Journal, vol. 11, no. 23, pp. 6905–6915, 2005.

[123] S. Roy, K. Bakhmutsky, E. Mahmoud, R. F. Lobo, and R. J. Gorte, “Probing Lewis AcidSites in Sn-Beta Zeolite,” ACS Catalysis, vol. 3, no. 4, pp. 573–580, 2013.

[124] F. Göltl and J. Hafner, “Modelling the Adsorption of Short Alkanes in ProtonatedChabazite: The Impact of Dispersion Forces and Temperature,” Microporous andMesoporous Materials, vol. 166, pp. 176–184, 2013.

[125] T. Jiang, F. Göltl, R. E. Bulo, and P. Sautet, “Effect of Temperature on the Adsorptionof Short Alkanes in the Zeolite SSZ-13 Adapting Adsorption Isotherms to MicroporousMaterials,” ACS Catalysis, vol. 4, no. 7, pp. 2351–2358, 2014.

[126] B. A. De Moor, M.-F. Reyniers, O. C. Gobin, J. A. Lercher, and G. B. Marin, “Adsorptionof C2-C8 n-Alkanes in Zeolites,” The Journal of Physical Chemistry C, vol. 115, no. 4,pp. 1204–1219, 2010.

[127] A. Corma, F. X. Llabrés i Xamena, C. Prestipino, M. Renz, and S. Valencia, “WaterResistant, Catalytically Active Nb and Ta Isolated Lewis Acid Sites, HomogeneouslyDistributed by Direct Synthesis in a Beta Zeolite,” The Journal of Physical ChemistryC, vol. 113, no. 26, pp. 11306–11315, 2009.

[128] C. Hammond, M. M. Forde, A. Rahim, M. Hasbi, A. Thetford, Q. He, R. L. Jenkins,N. Dimitratos, J. A. Lopez-Sanchez, and N. F. Dummer, “Direct Catalytic Conversionof Methane to Methanol in an Aqueous Medium by using Copper-Promoted Fe-ZSM-5,”Angewandte Chemie International Edition, vol. 51, no. 21, pp. 5129–5133, 2012.

[129] S. Dzwigaj, M. Peltre, P. Massiani, A. Davidson, M. Che, T. Sen, and S. Sivasanker,“Incorporation of Vanadium Species in a Dealuminated β Zeolite,” Chem. Commun.,no. 1, pp. 87–88, 1998.

[130] P. Wu, T. Komatsu, and T. Yashima, “IR and MAS NMR Studies on the Incorporation ofAluminum Atoms into Defect Sites of Dealuminated Mordenites,” The Journal of PhysicalChemistry, vol. 99, no. 27, pp. 10923–10931, 1995.

[131] S. Brunauer, P. H. Emmett, and E. Teller, “Adsorption of Gases in MultimolecularLayers,” Journal of the American Chemical Society, vol. 60, no. 2, pp. 309–319, 1938.

Bibliography 81

[132] B. Lippens and J. De Boer, “Studies on Pore Systems in Catalysts: V. The T method,”Journal of Catalysis, vol. 4, no. 3, pp. 319–323, 1965.

[133] R. A. Sheldon, M. Wallau, I. W. Arends, and U. Schuchardt, “Heterogeneous Catalysts forLiquid-Phase Oxidations: Philosophers’ Stones or Trojan Horses?,” Accounts of ChemicalResearch, vol. 31, no. 8, pp. 485–493, 1998.

[134] E. Taarning, S. Saravanamurugan, M. Spangsberg Holm, J. Xiong, R. M. West, and C. H.Christensen, “Zeolite-Catalyzed Isomerization of Triose Sugars,” ChemSusChem, vol. 2,no. 7, pp. 625–627, 2009.

[135] A. Baeyer and V. Villiger, “Einwirkung des Caroschen Reagens auf Ketone,” Berichteder deutschen chemischen Gesellschaft, vol. 32, no. 3, pp. 3625–3633, 1899.

[136] M. Renz and B. Meunier, “100 Years of Baeyer-Villiger Oxidations,” European Journalof Organic Chemistry, vol. 1999, no. 4, pp. 737–750, 1999.

[137] G. Strukul, “Transition Metal Catalysis in the Baeyer-Villiger Oxidation of Ketones,”Angewandte Chemie International Edition, vol. 37, no. 9, pp. 1198–1209, 1998.

[138] A. Berkessel, M. R. Andreae, H. Schmickler, and J. Lex, “Baeyer-Villiger Oxidationswith Hydrogen Peroxide in Fluorinated Alcohols: Lactone Formation by a NonclassicalMechanism,” Angewandte Chemie International Edition, vol. 41, no. 23, pp. 4481–4484,2002.

[139] K. Neimann and R. Neumann, “Electrophilic Activation of Hydrogen Peroxide: SelectiveOxidation Reactions in Perfluorinated Alcohol Solvents,” Organic Letters, vol. 2, no. 18,pp. 2861–2863, 2000.

[140] M. Uyanik and K. Ishihara, “Baeyer-Villiger Oxidation Using Hydrogen Peroxide,” ACSCatalysis, vol. 3, no. 4, pp. 513–520, 2013.

[141] A. Berkessel and M. R. Andreae, “Efficient Catalytic Methods for the Baeyer-VilligerOxidation and Epoxidation with Hydrogen Peroxide,” Tetrahedron Letters, vol. 42, no. 12,pp. 2293–2295, 2001.

[142] R. Gavagnin, M. Cataldo, F. Pinna, and G. Strukul, “Diphosphine-Palladium and-Platinum Complexes as Catalysts for the Baeyer-Villiger Oxidation of Ketones:Effect of the Diphosphine, Oxidation of Acyclic Ketones, and Mechanistic Studies,”Organometallics, vol. 17, no. 4, pp. 661–667, 1998.

[143] G. Strukul, A. Varagnolo, and F. Pinna, “New (Old) Hydroxo Complexes of Platinum(II)as Catalysts for the Baeyer-Villiger Oxidation of Ketones with Hydrogen Peroxide,”Journal of Molecular Catalysis A: Chemical, vol. 117, no. 1, pp. 413–423, 1997.

82 Bibliography

[144] A. M. F. Phillips and C. Romao, “Synthesis of γ-Butyrolactones by a Baeyer–VilligerOxidation with Hydrogen Peroxide, Catalysed by Methyltrioxorhenium,” EuropeanJournal of Organic Chemistry, vol. 1999, no. 8, pp. 1767–1770, 1999.

[145] G.-J. ten Brink, J.-M. Vis, I. W. C. A. Arends, and R. A. Sheldon, “Selenium-Catalyzed Oxidations with Aqueous Hydrogen Peroxide. 2. Baeyer-Villiger Reactions inHomogeneous Solution,” The Journal of Organic Chemistry, vol. 66, no. 7, pp. 2429–2433,2001.

[146] S. E. Jacobson, R. Tang, and F. Mares, “Oxidation of Cyclic Ketones by HydrogenPeroxide Catalysed by Group 6 Metal Peroxo Complexes,” Journal of the ChemicalSociety, Chemical Communications, no. 20, pp. 888–889, 1978.

[147] J. Fischer and W. F. Hölderich, “Baeyer-Villiger-oxidation of Cyclopentanone withAqueous Hydrogen Peroxide by Acid Heterogeneous Catalysis,” Applied Catalysis A:General, vol. 180, no. 1, pp. 435–443, 1999.

[148] M. Lenarda, M. Da Ros, M. Casagrande, L. Storaro, and R. Ganzerla, “Post-SyntheticThermal and Chemical Treatments of H-BEA Zeolite: Effects on the Catalytic Activity,”Inorganica Chimica Acta, vol. 349, pp. 195–202, 2003.

[149] R. A. Steffen, S. Teixeira, J. Sepulveda, R. Rinaldi, and U. Schuchardt, “Alumina-Catalyzed Baeyer-Villiger Oxidation of Cyclohexanone with Hydrogen Peroxide,” Journalof Molecular Catalysis A: Chemical, vol. 287, no. 1, pp. 41–44, 2008.

[150] L. Balbinot, U. Schuchardt, C. Vera, and J. Sepúlveda, “Oxidation of Cyclohexanolto Epsilon-Caprolactone with Aqueous Hdrogen Peroxide on H3PW12O40 andCs2.5H0.5PW12O40,” Catalysis Communications, vol. 9, no. 9, pp. 1878–1881, 2008.

[151] R. T. Taylor and L. A. Flood, “Polystyrene-Bound Phenylseleninic Acid: CatalyticOxidations of Olefins, Ketones, and Aromatic Systems,” The Journal of OrganicChemistry, vol. 48, no. 26, pp. 5160–5164, 1983.

[152] A. Lambert, D. J. Macquarrie, G. Carr, and J. H. Clark, “The Catalytic Oxidation ofCyclohexanone to Caprolactone Using Hexagonal Mesoporous Silica Supported SbF3,”New Journal of Chemistry, vol. 24, no. 7, pp. 485–488, 2000.

[153] A. Bhaumik, P. Kumar, and R. Kumar, “Baeyer-Villiger Rearrangement Catalysed byTitanium Silicate Molecular Sieve (TS-1)/H2O2 System,” Catalysis Letters, vol. 40, no. 1-2, pp. 47–50, 1996.

[154] A. Corma, V. Fornés, S. Iborra, M. Mifsud, and M. Renz, “One-Pot Synthesis of Phenolsfrom Aromatic Aldehydes by Baeyer-Villiger Oxidation with H2O2 Using Water-TolerantLewis Acids in Molecular Sieves,” Journal of Catalysis, vol. 221, no. 1, pp. 67–76, 2004.

Bibliography 83

[155] R. R. Sever and T. W. Root, “Comparison of Epoxidation and Baeyer-Villiger ReactionPathways for Ti(IV)-H2O2 and Sn(IV)-H2O2,” The Journal of Physical Chemistry B,vol. 107, no. 38, pp. 10521–10530, 2003.

[156] J. W. Harris, M. J. Cordon, J. R. Di Iorio, J. C. Vega-Vila, F. H. Ribeiro, and R. Gounder,“Titration and quantification of open and closed lewis acid sites in sn-beta zeolites thatcatalyze glucose isomerization,” Journal of Catalysis, vol. 335, pp. 141–154, 2016.

[157] P. Wolf, M. Valla, F. Núñez Zarur, A. Comas-Vives, A. J. Rossini, C. Firth, H. Kallas,A. Lesage, L. Emsley, C. Copéret, and I. Hermans, “Correlating Synthetic Methods,Morphology, Atomic-Level Structure and Catalytic Activity of Snβ Catalysts.” to besubmitted.

[158] M. A. Camblor, A. Corma, and S. Valencia, “Synthesis in Fluoride Media andCharacterisation of Aluminosilicate Zeolite Beta,” Journal of Materials Chemistry, vol. 8,no. 9, pp. 2137–2145, 1998.

[159] A. Parvulescu, U. Müller, J. Teles, N. Vautravers, G. Uhl, I. Hermans, P. Wolf, andC. Hammond, “A Tin-Containing Zeolitic Material Having a BEA Framework Structure,”May 14 2015. WO Patent App. PCT/EP2014/073,803.

[160] G. Ertl, H. Knözinger, and J. Weitkamp, Handbook of Heterogeneous Catalysis. VCHWeinheim, 1997.

[161] J. R. Ross, Heterogeneous Catalysis: Fundamentals and Applications. Elsevier, 2011.

[162] M. Thommes, S. Mitchell, and J. Pérez-Ramírez, “Surface and Pore Structure Assessmentof Hierarchical MFI Zeolites by Advanced Water and Argon Sorption Studies,” TheJournal of Physical Chemistry C, vol. 116, no. 35, pp. 18816–18823, 2012.

[163] M. Moliner, “State of the Art of Lewis Acid-Containing Zeolites: Lessons from FineChemistry to New Biomass Transformation Processes,” Dalton Transactions, vol. 43,no. 11, pp. 4197–4208, 2014.

[164] C. Hammond, S. Conrad, and I. Hermans, “Simple and Scalable Preparation of HighlyActive Lewis Acidic Sn-β,” Angewandte Chemie International Edition, vol. 51, no. 47,pp. 11736–11739, 2012.

[165] J. J. Pacheco and M. E. Davis, “Synthesis of Terephthalic Acid via Diels-Alder Reactionswith Ethylene and Oxidized Variants of 5-Hydroxymethylfurfural,” Proceedings of theNational Academy of Sciences, vol. 111, no. 23, pp. 8363–8367, 2014.

[166] A. Corma, H. García, G. Sastre, and P. M. Viruela, “Activation of Molecules in ConfinedSpaces: an Approach to Zeolite-Guest Supramolecular Systems,” The Journal of PhysicalChemistry B, vol. 101, no. 23, pp. 4575–4582, 1997.

84 Bibliography

[167] R. Berthoud, N. Rendón, F. Blanc, X. Solans-Monfort, C. Copéret, and O. Eisenstein,“Metal Fragment Isomerisation Upon Grafting a d2 ML4perhydrocarbyl Os Complex ona Silica Surface: Origin and Consequence,” Dalton Transactions, no. 30, pp. 5879–5886,2009.

[168] P. J. Cordeiro and T. D. Tilley, “Enhancement of the Catalytic Activity ofTitanium-Based Terminal Olefin Epoxidation Catalysts via Surface Modification withFunctionalized Protic Molecules,” ACS Catalysis, vol. 1, no. 5, pp. 455–467, 2011.

[169] P. Hoffmann and E. Knözinger, “Novel Aspects of Mid and Far IR Fourier SpectroscopyApplied to Surface and Adsorption Studies on SiO2,” Surface Science, vol. 188, no. 1,pp. 181–198, 1987.

[170] B. Morrow and I. Cody, “Infrared Studies of Reactions on Oxide Surfaces. 5. Lewis AcidSites on Dehydroxylated Silica,” The Journal of Physical Chemistry, vol. 80, no. 18,pp. 1995–1998, 1976.

[171] B. Morrow and I. Cody, “Infrared Studies of Reactions on Oxide Surfaces. 6. Active Siteson Dehydroxylated Silica for the Chemisorption of Ammonia and Water,” The Journalof Physical Chemistry, vol. 80, no. 18, pp. 1998–2004, 1976.

[172] B. Morrow, I. Cody, and L. S. Lee, “Infrared Studies of Reactions on Oxide Surfaces. 7.Mechanism of the Adsorption of Water and Ammonia on Dehydroxylated Silica,” TheJournal of Physical Chemistry, vol. 80, no. 25, pp. 2761–2767, 1976.

[173] M. Limbourg, J. Denayer, G. V. Baron, V. Parvulescu, P. J. Grobet, D. E. De Vos, P. A.Jacobs, et al., “Mesoporous Zr and Hf Catalysts for Chemoselective MPV Reductions ofUnsaturated Ketones,” Applied Catalysis A: General, vol. 254, no. 2, pp. 189–201, 2003.

[174] Y. Zhu, S. Liu, S. Jaenicke, and G. Chuah, “Zirconia Catalysts in Meerwein-Ponndorf-Verley Reduction of Citral,” Catalysis Today, vol. 97, no. 4, pp. 249–255, 2004.

[175] M. Boronat, A. Corma, and M. Renz, “Mechanism of the Meerwein–Ponndorf–Verley–Oppenauer (MPVO) Redox Equilibrium on Sn-and Zr-beta Zeolite Catalysts,” TheJournal of Physical Chemistry B, vol. 110, no. 42, pp. 21168–21174, 2006.

[176] R. S. Assary, L. A. Curtiss, and J. A. Dumesic, “Exploring Meerwein–Ponndorf–VerleyReduction Chemistry for Biomass Catalysis Using a First-Principles Approach,” ACSCatalysis, vol. 3, no. 12, pp. 2694–2704, 2013.

[177] A. Corma, “Inorganic Solid Acids and Their Use in Acid-Catalyzed HydrocarbonReactions,” Chemical Reviews, vol. 95, no. 3, pp. 559–614, 1995.

Bibliography 85

[178] A. Bhan, R. Gounder, J. Macht, and E. Iglesia, “Entropy Considerations inMonomolecular Cracking of Alkanes on Acidic Zeolites,” Journal of Catalysis, vol. 253,no. 1, pp. 221–224, 2008.

[179] R. Gounder and E. Iglesia, “Catalytic Consequences of Spatial Constraints and Acid SiteLocation for Monomolecular Alkane Activation on Zeolites,” Journal of the AmericanChemical Society, vol. 131, no. 5, pp. 1958–1971, 2009.

[180] R. Gounder and E. Iglesia, “Effects of Partial Confinement on the Specificity ofMonomolecular Alkane Reactions for Acid Sites in Side Pockets of Mordenite,”Angewandte Chemie International Edition, vol. 49, no. 4, pp. 808–811, 2010.

[181] R. Gounder and E. Iglesia, “The Roles of Entropy and Enthalpy in Stabilizing Ion-Pairsat Transition States in Zeolite Acid Catalysis,” Accounts of Chemical Research, vol. 45,no. 2, pp. 229–238, 2011.

[182] A. J. Jones, S. I. Zones, and E. Iglesia, “Implications of Transition State Confinementwithin Small Voids for Acid Catalysis,” The Journal of Physical Chemistry C, vol. 118,no. 31, pp. 17787–17800, 2014.

[183] H. Y. Luo, D. F. Consoli, W. R. Gunther, and Y. Román-Leshkov, “Investigation of theReaction Kinetics of Isolated Lewis Acid Sites in Beta Zeolites for the Meerwein-Ponndorf-Verley Reduction of Methyl Levulinate to γ-Valerolactone,” Journal of Catalysis, vol. 320,pp. 198–207, 2014.

[184] P. P. Samuel, S. Shylesh, and A. Singh, “Catalytic Properties of Tin-ContainingMesoporous Molecular Sieves in the Selective Reduction of Carbonyl Compounds(Meerwein-Ponndorf-Verley (MPV) reaction),” Journal of Molecular Catalysis A:Chemical, vol. 266, no. 1, pp. 11–20, 2007.

[185] S. Conrad, R. Verel, C. Hammond, P. Wolf, F. Göltl, and I. Hermans, “Silica-Grafted SnIV

Catalysts in Hydrogen-Transfer Reactions,” ChemCatChem, vol. 7, no. 20, pp. 3270–3278,2015.

[186] M. Grün, K. K. Unger, A. Matsumoto, and K. Tsutsumi, “Novel Pathways for thePreparation of Mesoporous MCM-41 Materials: Control of Porosity and Morphology,”Microporous and Mesoporous Materials, vol. 27, no. 2, pp. 207–216, 1999.

[187] T. R. Eaton, A. M. Boston, A. B. Thompson, K. A. Gray, and J. M. Notestein, “ActiveSites on Titanium Oxide–Silica Catalysts for Hydrogen Peroxide Activation through InSitu Poisoning with Phenylphosphonic Acid,” ChemCatChem, vol. 6, no. 11, pp. 3215–3222, 2014.

Appendix A

Annexes

Chapter 3

Figure A.1. ε-Caprolactone yield over time for different initial lactone concentrations in the BVoxidation of cyclohexanone with H2O2. Experimental conditions: cyclohexanone in 1,4-dioxane(0.33 M), H2O2 (50 wt% aq. solution, H2O2/ketone = 1.5), 10Snβ-OH25, desired amount of ε-Caprolactone.

Table A.1. Physicochemical properties of differentSnβ catalysts.

Entry Catalyst SBETa Si/Snb H2O

cads

[m2 g−1] [–] [mmol g-1]

1 1Snβ-F30 718 200 7.902 1Snβ-F200 570 173 2.273 1Snβ-F400 545 192 2.264 1Snβ-HT 610 182 1.675 1Snβ-OH25 616 217 12.06 1Snβ-OH300 637 227 10.1a Brunauer-Emmett-Teller surface area.[131]

b Determined by ICP-OES. c Quantity of wateradsorbed at p/p0 = 0.82.

88 Chapter A

Figure A.2. FT-IR spectra of different Snβ catalysts, normalized to the Si-O-Si overtones of thesilica framework between 2200 and 1400 cm-1.

Figure A.3. Diffuse Reflectance UV-Vis measurements of different Snβ catalysts and SnO2 as areference.

Annexes 89

Chapter 4

1H NMR 13C NMR

Figure A.4. The formation of HNMe2 during the grafting was confirmed by analyzing the liquidnitrogen-trapped effluent of the synthesis reactor with 1H- and 13C-NMR. 1H NMR (500 MHz, CDCl3):δ 2.39 (6H, s); 13C NMR (500 MHz, CDCl3): δ 38.65 (2C, s), 77.01 (1C, t; solvent).

TMS νCH

Figure A.5. After grafting Sn(NMe2)4 on to TMS-SiO2(700) (i.e., silica with capped silanols of thetype ≡SiOSiMe3, –SiMe3 = TMS), we observed additional bands in the C-H IR region (see Sn/TMS-SiO2(700)). In combination with the measured Sn loading of Sn/TMS-SiO2(700), this confirms theparticipation of siloxanes in the grafting reaction. Additionally, by comparing the C-H bands ofSn/TMS-SiO2(700) (after subtraction of TMS-SiO2(700); see red curve) with Sn/SiO2(700), a strongoverlap could be noted. This suggests that similar species are formed when Sn(NMe2)4 reacts withsilanols or siloxanes, respectively.

90 Chapter A

Table

A.2.

Elementalanalysis

dataofthe

differentprepared

materials. a

EntryMaterial

SnH

CN

C/Sn

H/Sn

N/Sn

[wt%

][wt%

][wt%

][wt%

]exp.(calc.)

exp.(calc.)exp.(calc.)

1Sn/SiO

2(700) b,c4.0±

0.12.4±

0.10.7±

0.02d

1.6±

0.04d

6.0/5.6(6)

21.5(20.2)

3.4/3.2(3)

2TMS-SiO

2(700) dN/A

1.0±

0.030.2±

0.0050.05

±0.001

3.1(3)

10.3(9)

0.2(0)

3Sn/T

MS-SiO

2(700) e0.9±

0.021.8±

0.10.5±

0.010.5±

0.019.0

(8)42.6

(24)4.5

(4)4

Sn/SiO2(700) -O

24.0±

0.10.06

±0.002

0.04±

0.001(0.40)

0.05±

0.001–

––

5Sn/SiO

2(700)+

2-butanol f4.0±

0.11.0±

0.031.0±

0.030.5±

0.019.5

(12)26.4

(27)1.1

(0)aC

HN

andIC

P-OES

analyseswere

performed

inorder

todeterm

ineexperim

entalmolar

ratios((C

,H,N

)/Sn).These

valuesare

compared

with

calculatedratios

basedon

ourhypothesized

activesite

structures.bT

hecalculated

valuesare

basedon

thehypothesis

thatSn(N

Me2 )4

isbond

tothe

silicasurface

byone

singleSn–O

bondwhile

threeam

inegroups

(–NMe2 )

areattached

tothe

Snatom

.cT

he(C

,H,N

)/Snratios

aredeterm

inedunder

theassum

ptionthat

only≡Sn(N

M2 )3

speciesare

presentOR

thatalso≡SiN

Me2

speciesare

presentwith

amolar

amount

of0.071mmolg

-1(derived

fromthe

amount

ofSnthat

isgrafted

tosilylated

silica).dM

olarratios

correspondto

(C,H

,N)/≡

SiOH

ratios.e(C

,H,N

)/Snratios

arecalculated

aftersubtracting

themolar

amounts

ofC,H

,Nmeasured

onTMS-SiO

2(700) .fSn/SiO

2(700)after

theadsorption

of2-butanol.Calculated

ratiosare

basedon

theassum

ptionthat

amine

ligandsare

replacedby

2-butoxyligands.

Annexes 91

*

*

**

(A) (B)

(C)

Figure A.6. In addition to our 13C NMR measurements, the presence of grafted amine species≡SiNMe2 (see B) is indicated by two C-H bands in the IR spectrum of Sn/TMS-SiO2(700) (markedwith blue star in (A)). These sites are formed when Sn(NMe2)4 reacts with siloxane bridges. However,their C-H bands are not clearly visible in the IR spectrum of Sn/SiO2(700), due to the lower relativeconcentration of such sites in this material. More evidence for this was found in the IR spectrum ofTMS/SiO2(700). During the silylation step (C) mostly ≡SiOTMS, but also ≡SiN(TMS)Me speciesare formed in small amounts. Computationally predicted νCH frequencies for these sites appear athigher wavenumbers compared to ≡SiNMe2. This agrees with the position of C-H bands observedfor TMS-SiO2(700) (marked with red star in (A)).

Table A.3. Asymmetric and symmetric C–H stretch vibrations calculated withB3LYP/6-31 G8(d,p).a

Entry Structure Asymmetric stretch ν(C,H) Symmetric stretch ν(C,H)[cm-1] [cm-1]

1 ≡SiNMe2 2985 30402 ≡SiN(TMS)Me 3000 3069a For absolute values a scaling factor of 0.9614 has to be applied.

92 Chapter A

Figure A.7. More intense signals in the IR region of C-H stretches when HNEt2 is grafted togetherwith Sn(NMe2)4 (see red curve) indicates that HNEt2 reacts with siloxanes. The more intense bandsdo not arise from a higher amount of grafted precursor, since the Sn loading of the material is identicalto Sn/SiO2(700). This observation shows that both of our hypothesized pathways for the opening ofsiloxanes ((A)-(B) or (C)) are possible. However, pure HNMe2 does not react with SiO2(700). Thus,we believe that under our grafting conditions, amines such as HNR2 (R = Me, Et), only react withsiloxanes when they are Lewis acid activated by Sn(NMe2)4.

!!

(A) (B)

(C)

Figure A.8. DRIFT spectra from the thermal treatment of pure SiO2(700) with (black curves) andwithout (red curves) humidity in the air stream (gas stream was passed through a water trap, ca.3.2 % water in the stream) indicate opening of siloxanes by water, as shown by the shoulder at ca.3660 cm-1 (B). Additionally, a broad signal appears at around 3550 cm-1 (C) at lower temperatures,which signalizes the presence of physisorbed water on the surface that is released during heating.

Annexes 93

Figure A.9. Quantification of the SnOH IR signal of Sn/SiO2(700)-O2 (red) with the SnOH IRsignal of a reference Sn hydroxyl compound – tricyclohexyltinhydroxide (blue) (IR spectrum taken ofa physical (homogeneous) mixture with SiO2(700) to allow normalization to the Si-O-Si overtones ofthe silica framework).

Figure A.10. Our second functionalization step consist of the silylation of Sn/SiO2(700)-O2 with N,N -bis(trimethyl)silylamine (TMS2NMe). The IR spectrum of Sn/SiO2(700)-O2-TMS shows the conversionof both types of hydroxyls (SiOH, SnOH), together with the introduction of C-H stretches from theorganic substituents of the silylating agent.

94 Chapter A

Figure A.11. In the UV-Vis spectra of Sn/TMS-SiO2(700) and Sn/TMS-SiO2(700)-O2 there is no signof the formation of SnOx species.

Sn/SiO2(700) Sn/SiO2(700)-O2

Sn/SiO2(700)-O2-TMS

Sn-β zeolite

Figure A.12. Deconvolutions of the IR regions of adsorbed cyclohexanone after exposure to 2-butanol(performed with origin 8.5.1). The fitted curves are shown in black (individual and cumulative peakfits), the original curves are depicted in red. All y-axes are scaled equally. Areas from peaks between1600 and 1690 cm-1 were used for the analysis. Sn/SiO2(700), R2 = 0.99898; Sn/SiO2(700)-O2, R2 =0.99988; Sn/SiO2(700)-O2-TMS, R2 = 0.99796; Snβ zeolite, R2 = 0.99932.

Annexes 95

Figure A.13. FT-IR spectra of deuterated acetonitrile adsorbed on to our three SnIV-catalysts andSnβ (spectra are background corrected). Sn/SiO2(700)-O2 and Sn/SiO2(700)-O2-TMS, our catalyticallyactive materials (see Figure 4.9 and Table 4.1, Entries 3 and 4), show signals at around 2310 cm-1,which is characteristic of Lewis acidic metal sites as shown before for Snβ. This observation pointstowards strong similarities between our SnIV-sites and the sites in Snβ.[101,111]

Figure A.14. Catalytic activity measurements of all Sn-silica materials (A) – (C) and Snβ zeolite(D) for the MPV reduction of cyclohexanone with 2-butanol. The Turnover Number (TON) is definedas the moles of product generated per mole Sn.

Annexes 97

Chapter 5

Table A.4. Properties of MCM-41 samples prepared in heterogeneousmedium with n-alkyltrimethylammonium bromides as templates, followinga procedure from Grün and co-workers.[186]

Entry Template SBETa SBET

b Pore diameterc,e Pore diameterd,e

[m2 g−1] [m2 g−1] [nm] [nm]

1 C12TMABr 1549 1424 2.6 2.52 C16TMABr 1400 1269 3.6 3.3a Brunauer-Emmett-Teller surface area.[131] b Brunauer-Emmett-Teller surfacearea after a thermal treatment at 700 °C under vacuum (approx. 20 µbar).[131]

c Pore diameter. d Pore diameter after a thermal treatment at 700 °C undervacuum (approx. 20 µbar). e Determined from X-ray diffraction data, ascalculating the pore size from nitrogen sorption data leads to an underestimationof the pore sizes for MCM-41 materials.[186] Calculated using the d100-value andassuming a pore wall thickness of 1.0 nm: d = 2d100√

3 − 1.0 with d100 = 1.542sinθ .

Appendix B

List of Publications

Doctoral Publications

F. Göltl, S. Conrad, P. Wolf, P. Müller, J. Wheeler, R.J. Hamers, G. Kresse, I. HermansExperiment to Understand the Large Stoke Shift for Isolated Copper Centers in Zeolitesto be submitted

S. Conrad,* P. Wolf,* P. Müller, H. Orsted, C. Hammond, I. Hermans* these authors contributed equallyInsights into the Baeyer-Villiger Oxidation of Cyclohexanone with H2O2 catalyzed by Snβsubmitted (Chapter 3)

S. Conrad, R. Verel, C. Hammond, P. Wolf, F. Göltl, I. HermansSilica-Grafted SnIV-Catalysts in Hydrogen-Transfer ReactionsChemCatChem 2015, 7, 3188-3403 (front cover) (Chapter 4)

P. Wolf, C. Hammond, S. Conrad, I. HermansPost-Synthetic Preparation of Sn-, Ti-, Zr-β: A Facile Route to Water-Tolerant, Highly ActiveLewis Acidic ZeolitesDalton Trans. 2014, 43, 4515-4519

C. Aellig, D. Scholz, S. Conrad, I. HermansIntensification of TEMPO-Mediated Aerobic Alcohol Oxidations under Three-Phase-FlowConditionsGreeen Chem. 2013, 15, 1975-1980

P. Mania, S. Conrad, R. Verel, C. Hammond, I. HermansThermal Restructuring of Silica-Grafted –CrO2 and –VOCl2 speciesDalton Trans. 2013, 42, 12725-12732

C. Hammond, M. Schümperli, S. Conrad, I. HermansHydrogen Transfer Processes Mediated by Supported Iridium Oxide NanoparticlesChemCatChem 2013, 5, 2983-2990

100 Chapter B

C. Hammond, S. Conrad, I. HermansSimple and Scalable Preparation of Highly Lewis Acidic SnβAngew. Chem. Int. Ed. 2012, 51, 11736-11739 (back cover) (Chapter 2)

C. Hammond, S. Conrad, I. HermansOxidative Methane UpgradingChemSusChem 2012, 9, 1668-1686

Pre-Doctoral Publications

A. Gänzler, M. Casapu, A. Boubnov, O. Müller, S. Conrad, H. Lichtenberg, R. Frahm, J.-D.GrunwaldtOperando Spatially and Time-Resolved X-Ray Absorption Spectroscopy and InfraredThermography during Oscillatory CO OxidationJ. Catal. 2015, 328, 216-224

A. Boubnov, A. Gänzler, S. Conrad, M. Casapu, J.-D. GrunwaldtOscillatory CO Oxidation over Pt/Al2O3 Catalysts studied by In Situ XAS and DRIFTSTop. Catal. 2013, 56, 333-338

Appendix C

Presentations

International Symposium on Activation of Dioxygen and Homogeneous Oxidation CatalysisMadison (WI), United States, 2015Poster: "Silica-Grafted SnIV-Catalysts in Hydrogen-Transfer Reactions"S. Conrad, R. Verel, C. Hammond, P. Wolf, F. Göltl, I. Hermans

125th International Summer Course for Chemists and EngineersBASF-SE, Ludwigshafen, Germany, 2014Talk: "Toward a Better Understanding of the Activity of Site-Isolated Lewis Acid Catalysts"S. Conrad, C. Hammond, I. Hermans

Advisory Board Meeting - Chemistry DepartmentUniversity of Wisconsin-Madison, Madison (WI), United States, 2014Poster: "Influence of Surface Polarity and Cavity Effects on Hydrogen-Transfer-Reactions"S. Conrad, C. Hammond, I. Hermans

Catalysis Club of Chicago - Spring SymposiumBP Naperville Campus, Chicago (IL), United States, 2014Poster: "Influence of Surface Polarity and Cavity Effects on Hydrogen-Transfer-Reactions"S. Conrad, C. Hammond, I. Hermans

Appendix D

Cover Gallery

Based on the work in this thesis, two covers were published, which are presented on the followingpages. Each cover comes along with a short explanation to illustrate the scientific background.

104 Chapter D

This (back) cover relates to a communication published in the Angewandte Chemie, and isbased on Chapter 2. This publication is about the heterogeneous catalyst Snβ, which iscurrently prepared in a complicated hydrothermal synthesis, which has several hurdles thatprevent its industrial implementation. A new convenient preparation of Snβ by solid-state-ion-exchange is reported in this communication, which profits from less time and synthetic skillsdemanded. Additionally, the produced catalyst has more favorable catalytic properties, givingspace-time-yields over one order of magnitude higher than previously observed.

Cover Gallery 105

The heterogeneous catalyst Sn-b is currently prepared in a complicated hydro-thermal synthesis, which has several hurdles that prevent its industrial implemen-tation. In their Communication (10.1002/anie.201206193), Hermans and co-work-ers report a convenient preparation of Sn-b by solid-state ion exchange. It requiresmuch less time and synthetic skill, and the product has more favorable catalyticproperties, giving space-time-yields over one order-of-magnitude higher thanpreviously observed.

Dateiname: W208116e Pagina: 2Pfad: l:/daten/verlage/vch/ach/hefte/pool/ Seite: 1 te von 1Status Neusatz Umfang (Seiten): 1Datum: 42 KW., 12. Oktober 2012 (Freitag) Zeit: 7:14:41 Uhr

106 Chapter D

This (front) cover, based on results presented in Chapter 4, illustrates silica-supported SnIV

sites that are tuned to be active in the Meerwein-Ponndorf-Verley reaction of cyclohexanonewith 2-butanol. In the corresponding article published in ChemCatChem, the catalyticperformance of three silica-grafted SnIV model-catalysts is studied to differentiate betweenthe various parameters that contribute to to the activity of SnIV/SiO2-based catalytic systems.The differences in performance observed highlight that the activity of such systems significantlydepends on active site speciation and the hydrophobicity of the framework. Comparison tobenchmark Snβ further suggests that confinement effects inside zeolite cavities play a key role.

Cover Gallery 107

ISSN 1867-3880 · Vol. 7 · No. 20 · October, 2015

20/2015Front Cover:

I. Hermans et al.Silica-Grafted SnIV Catalysts in Hydrogen-Transfer Reactions

www.chemcatchem.org

A Journal of

Appendix E

Curriculum Vitae

Name Sabrina Conrad

Date of birth October 24th, 1986

Place of birth Stuttgart, Germany

Nationality German

Education

2014 – 2015 External continuation of ETH Doctoral Studies under the supervisionof Prof. Dr. Ive Hermans and Prof. Dr. Christophe Copéret at theUniversity of Wisconsin-Madison, USA

2012 – 2013 Ph.D. under the supervision of Prof. Dr. Ive Hermans at ETH Zurich,Switzerland

2007 – 2011 Diploma studies in Chemistry, Karlsruhe Institute of Technology,Germany

Date post:	07-Feb-2017
Category:	Documents
Upload:	vuongdat
View:	219 times
Download:	2 times

Synthesis and Activity of Heterogeneous Lewis Acidic Sn Catalysts

Documents