Efficient and Spatially Controlled Functionalization of SBA-15 Von der Fakultät Chemie der Universität Stuttgart zur Erlangung der Würde eines Doktors der Naturwissenschaften (Dr. rer. nat.) genehmigte Abhandlung Vorgelegt von Ann-Katrin Beurer aus Bietigheim-Bissingen Hauptberichter: Apl.-Prof. Dr. Yvonne Traa Mitberichter: Apl.-Prof. Dr. Thomas Sottmann Tag der mündlichen Prüfung: 11.10.2023 Institut für Technische Chemie der Universität Stuttgart 2023 Erklärung über die Eigenständigkeit der Dissertation Ich versichere, dass ich die vorliegende Arbeit mit dem Titel Efficient and spatially controlled functionalization of SBA-15 selbständig verfasst und keine anderen als die angegebenen Quellen und Hilfsmittel benutzt habe; aus fremden Quellen entnommene Passagen und Gedanken sind als solche kenntlich gemacht. Declaration of authorship I hereby certify that the dissertation entitled Efficient and spatially controlled functionalization of SBA-15 is entirely my own work except where otherwise indicated. Passages and ideas from other sources have been clearly indicated. Ludwigsburg, 11.10.2023 Ann-Katrin Beurer Acknowledgments I would like to thank Apl.-Prof. Dr. Yvonne Traa for assigning me the A4-2 project of the Collab- orative Research Center (CRC) 1333. I thank her for the kind and constructive support during my PhD period with all its ups and downs. Furthermore, I want to thank Prof. Dr.-Ing. Elias Klemm for the opportunity to carry out my work at the Institute of Technical Chemistry (ITC). I would like to thank Apl.-Prof. Dr. Thomas Sottmann for taking over the second opinion for this thesis and Prof. Dr. Thomas Schleid for taking over the chair. I want to thank the entire ITC - my colleagues for their friendly and helpful cooperation, Elisa Favaro for her patience and support in all organizational matters, Ines Lauerwald for her help, Ingo Nägele and Andreas Stieber for the technical support, Barbara Gehring for performing el- ement analytical investigations and logging TGA measurements, and Heike Fingerle for doing ICP-OES measurements, which were never-ending in the meantime. Special thanks to my office colleagues Dr. Dennis Beierlein for the time during the Summer School in France, Dr. Robin Himmelmann for the support during my first steps in the PhD life and his friendship, Jan Floren- ski for scientific discussions in everyday life as well as proofreading of this work and Faeze Tari as good spirit in our office. I would like to thank Dr. Johanna R. Bruckner for her support throughout this work. Furthermore, I also want to thank Elif Kaya and Nils Hübener for their scientific work during their research internship and master thesis and the research assistants Sabrina Schätzle, Jacqueline Gebhardt, Jessica Bauhof, Nadine Schnabel and Fabian Bölzle for their help in the lab. I would like to thank the German Research Foundation for funding the CRC 1333 and the CRC members for their cross-project work and support - especially Dr. Elisabeth Rüthlein for her help in organizational matters, Helena Solodenko for all TEM images, Dr. Petia Atanasova for the SEM images, and Dr. Michael Dyballa for performing the NMR measurements. Last but not least, I thank my parents from the bottom of my heart for their support and encour- agement - without them, none of this would have been possible. Thanks to my brother, who has always given me back my faith in myself. And finally, thank you to Max Bachmann for his love and for wiping away my doubts when things did not go my way again. Publications List of publications directly relevant to my cumulative dissertation: A.-K. Beurer, S. Dieterich, H. Solodenko, E. Kaya, N. Merdanoǧlu, G. Schmitz, Y. Traa, and J. R. Bruckner, "Comparative Study of Lattice Parameter and Pore Size of Ordered Mesoporous Sil- ica Materials Using Physisorption, SAXS Measurements and Transmission Electron Microscopy", Microporous Mesoporous Materials 2023, 354, 1387-1811, DOI 10.1016/j.micromeso.2023.112508. A.-K. Beurer, J. R. Bruckner, and Y Traa, "Influence of the template removal method on the me- chanical stability of SBA-15", Chemistry Open 2021, 10, 1123-1128, DOI 10.1002/open.202100225. A.-K. Beurer, M. Kirchhof, J. R. Bruckner, W. Frey, A. Baro, M. Dyballa, F. Giesselmann, S. Laschat, and Y. Traa, "Efficient and spatially controlled functionalization of SBA-15 and initial results in asymmetric Rh-catalyzed 1,2-additions under confinement", ChemCatChem 2021, 13, 2407-2419, DOI 10.1002/cctc.202100229. In addition, the results of this work contributed to publishing the following publications: M. Kirchhof, K. Gugeler, A.-K. Beurer, F. R. Fischer, D. Batman, S. Bauch, S. Kolin, E. Nicholas, R. Schoch, C. Vogler, S. R. Kousik, A. Zens, B. Plietker, P. Atanasova, S. Naumann, M. Bauer, J. R. Bruckner, Y. Traa, J. Kästner and S. Laschat, "Tethering chiral Rh diene complexes inside meso- porous solids: experimental and theoretical study of substituent, pore and linker effects on asym- metric catalysis", Catalysis Science & Technology 2023, 13, 3709-3724, DOI 10.1039/d3cy00381g. C. Rieg, M. Kirchhof, K. Gugeler, A.-K. Beurer, L. Stein, K. Dirnberger, W. Frey, J. R. Bruck- ner, Y. Traa, J. Kästner, S. Ludwigs, S. Laschat, and M. Dyballa, "Determination of accessibility and spatial distribution of chiral Rh diene complexes immobilized on SBA-15 via phosphine- based solid-state NMR probe molecules", Catalysis Science & Technology 2023, 2, 410-425, DOI 10.1039/D2CY01578A. E. L. Goldstein, F. Ziegler, A.-K. Beurer, Y. Traa, J. R. Bruckner, and M. R. Buchmeiser, "Cationic molybdenum imido alkylidene N-heterocyclic carbene complexes confined in mesoporous silica: Tuning transition states towards Z-selective ring-opening cross-metathesis", ChemCatChem 2022, 14, DOI 10.1002/cctc.202201008. Z. Li, C. Rieg, A.-K. Beurer, M. Benz, J. Bender, C. Schneck, Y. Traa, M. Dyballa, and M. Hunger, "Effect of aluminum and sodium on the sorption of water and methanol in microporous MFI-type zeolites and mesoporous SBA-15 materials", Adsorption 2021, 27, 49-68, DOI 10.1007/s10450- 020-00275-8. Z. Li, M. Benz, C. Rieg, D. Dittmann, A.-K. Beurer, D. Häussermann, B. Arstad, and M. Dyballa, "The alumination mechanism of porous silica materials and properties of derived ion exchangers and acid catalysts", Materials Chemistry Frontiers 2021, 5, 4254-4271, DOI 10.1039/D1QM00282A. J. R. Bruckner, J. Bauhof, J. Gebhardt, A.-K. Beurer, Y. Traa, and F. Giesselmann, "Mechanisms and intermediates in the true liquid crystal templating synthesis of mesoporous silica materials", The Journal of Physical Chemistry B 2021, 125, 3197-3207, DOI 10.1021/acs.jpcb.0c11005. VIII Table of Contents List of Abbreviations and Symbols XI 1 Introduction 9 1.1 Porous Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.2 Mesoporous Silica Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.2.1 Mobil Composition of Matter Materials . . . . . . . . . . . . . . . . . . 10 1.2.2 Santa Barbara Amorphous Materials . . . . . . . . . . . . . . . . . . . . 10 1.2.3 Plugged Hexagonal Templated Silica Material . . . . . . . . . . . . . . . 13 1.2.4 Mesostructured Cellular Foam . . . . . . . . . . . . . . . . . . . . . . . 15 1.3 Process for the Preparation of Mesoporous Silica Materials . . . . . . . . . . . . 15 1.3.1 Synthesis of Mesoporous Silica Materials . . . . . . . . . . . . . . . . . 15 1.3.2 Template Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.4 Properties of Mesoporous Silica Materials . . . . . . . . . . . . . . . . . . . . . 22 1.4.1 Silanol Groups on Surfaces of Mesoporous Silica Materials . . . . . . . 22 1.4.2 Characterization of the Pore Size of Mesoporous Silica Materials . . . . 26 1.4.3 Stability of Mesoporous Silica Materials . . . . . . . . . . . . . . . . . . 30 1.5 Functionalization of Mesoporous Silica Materials . . . . . . . . . . . . . . . . . 32 1.5.1 Catalytically Active Species . . . . . . . . . . . . . . . . . . . . . . . . 32 1.5.2 Introduction of Functional Groups . . . . . . . . . . . . . . . . . . . . . 33 1.5.3 Selective Functionalization . . . . . . . . . . . . . . . . . . . . . . . . . 35 2 Motivation and Objectives 37 3 Results and Discussion 39 3.1 Mesoporous Silica Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.1.1 Characterization of Mesoporous Silica with Common Methods . . . . . . 39 3.1.2 Alternative Characterization Methods . . . . . . . . . . . . . . . . . . . 42 Table of Contents 3.1.3 Discussion of Common and Alternative Characterization Methods . . . . 46 3.1.4 Differences of Calcined and Extracted SBA-15 . . . . . . . . . . . . . . 47 3.2 Efficient and Spatially Controlled Functionalization of SBA-15 . . . . . . . . . . 49 3.2.1 Procedure of the Efficient and Spatially Controlled Functionalization . . 50 3.2.2 Functionalization of the Particle Surface and the Pore Entrances . . . . . 50 3.2.3 Control Step . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 3.2.4 Template Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 3.2.5 Thermal Treatment in Nitrogen . . . . . . . . . . . . . . . . . . . . . . 54 3.2.6 Functionalization of the Pore Walls . . . . . . . . . . . . . . . . . . . . 56 3.3 Influence of the Template Removal Method on the Mechanical Stability of SBA-15 57 3.3.1 Influence of the Template Removal Method on the Mechanical Stability of SBA-15 against a Pressure of 156 MPa . . . . . . . . . . . . . . . . . 58 3.3.2 Influence of the Template Removal Method on the Mechanical Stability of SBA-15 against a Pressure of 39 MPa . . . . . . . . . . . . . . . . . . 61 4 Summary and Outlook 63 Bibliography 67 Appendix 95 List of Figures 145 List of Tables 147 X List of Abbreviations and Symbols Abbreviations as as-synthesized Au-NP gold nanoparticle AzPTES 3-azidopropyltriethoxysilane BET Brunnauer-Emmett-Teller BJH Barrett, Joyner, and Halenda calc calcined CP cross polarization DFT density functional theory DRIFT diffuse infrared reflectance Fourier transform e.g. Latin exempli gratia; English for example E Soxhlet extraction with ethanol EO ethylene oxide ex external FFT fast Fourier transformation FTIR Fourier-transform infrared spectroscopy HMDS 1,1,1-trimethyl-N-(trimethylsilyl)silanamine List of Abbreviations and Symbols HPLC high-performance liquid chromatography i.e. Latin id est; meaning that is to say I precursor molecule in pore walls IR infrared ISEC inverse size exclusion chromatography IUPAC International Union of Pure and Applied Chemistry MAS magic angle spinning MC molecular simulation MCF mesostructured cellular foam MCM mobile composition of matter MOF metal-organic framework MS mass spectroscopy NLDFT non localized density functional theory NMR nuclear magnetic resonance OMS ordered mesoporous silica p thermal treatment PHTS plugged hexagonal templated silica PO propylene oxide re refilled S surfactant SAXS small angle X-ray scattering SBA Santa Barbara amorphous SEM scanning electron microscopy SiOH silanol groups XII List of Abbreviations and Symbols STP standard temperature and pressure TEM transmission electron microscopy TLCT true liquid crystal templating TMOS tetramethyl orthosilicate X counter ion XRD X-ray diffractometry Latin Letters a nm lattice parameter d nm a) distance; b) diameter g dimensionless packing parameter q nm scattering vector S m2 surface area V m3 volume Indices - superscript + positively charged - negatively charged 0 nonionic Indices - subscript 100/110/200 Miller hkl indices ads adsorption branch des desorption branch m repeat units of EO in block copolymers EOmPOnEOm meso mesoporous micro microporous n repeat units of PO in block copolymers EOmPOnEOm tot total XIII Abstract Mimicking enzymes is one of the biggest goals of current catalysis research. They have defined micro environments that enable high enantioselectivities and thus can produce targeted products. With organometallic catalysts, especially when immobilized on solid supports, often low enantios- electivities are achieved compared to enzymes. In addition, the costs of producing and recovering dissolved catalysts are often high. Considering environmental issues and resource scarcity, it is important to develop catalysts for enantioselective and efficient chemical reactions. The combina- tion of the advantages of homogeneous catalysis, with similarly high enantioselectivities as well as productivities as enzymes show, with the positive aspects of heterogeneous catalysis is a promis- ing concept. Behind this approach is the immobilization of highly selective complex catalysts on porous support materials like mesoporous silica materials. By confining the reaction space in the pore and channel structure of the support material, shape selectivity can be achieved in desired product configurations. To enable shape selectivity through confinement, it is necessary to ensure that the immobilized metal complexes are located exclusively in the mesopores and channels of the support material. Therefore, the efficient and spatially controlled functionalization of mesoporous silica materials as support materials for catalytically active metal complexes is very important and can significantly advance heterogeneous catalysis. Precise knowledge of properties of the support material is required to ensure efficient and spatially controlled functionalization to mimic the functioning of enzymes. Various methods are available for determining characteristic properties such as the lattice parameter and pore size of mesoporous silica materials. In this work, four different mesoporous silica materials with an average pore size of 3 to 11 nm were used to compare commonly used characterization methods with alternative methods. Determination of lattice parameters by small angle X-ray (SAXS) measurements and from images from the transmission electron microscope yields similar values. Considering that errors during the evaluation of transmission electron microscopy (TEM) images may occur due to pore orientation, the determination of lattice parameters from TEM images is an alternative method that provides equivalent results. The pore size of various mesoporous silica materials was Abstract determined using physisorption measurements and SAXS studies. In addition, a new method was presented in which electron density maps were generated from SAXS data and used to determine pore sizes. Since it is difficult to determine the transition between the pore wall and the pore when viewing small pores in TEM images due to the different gray levels, the tractability of the open pore network was investigated in proof of principle experiments using gold nanoparticles (Au-NPs) as probe particles. The values for the pore size determined by different characterization methods agree well, although it should be noted that Au-NPs can only determine a range of pore sizes and not an exact value. Nevertheless, the comparison shows that the two new characterization methods - generation of electron density maps and use of Au-NPs as probe particles to determine the pore size - have the potential to become alternatives to the established methods. [1] The characterization of possible carrier materials is followed by functionalization, the most im- portant step for mimicking enzymes. In order to ensure that catalytically active species are located exclusively in the mesopores, all freely accessible silanol groups on the external surface of the support material must be inertized before the functionalization of the pore walls. Various meth- ods for controlled functionalization of porous silica have been reported in literature, but do not unequivocally demonstrate that selective functionalization of pore walls was achieved while the external surface was left inert. The procedure used in this work for selective functionalization of SBA-15 develop the selective functionalization procedures described in literature further. The in- troduction of a separate and independent control step ensures that the functionalization procedure results in a material in which the organic groups - in this case azide groups - are located exclusively on the pore walls. Click chemistry can then be used to attach linkers and noble metal complexes to the azide groups and provide the actual catalysts. Before the control step, the external sur- face of SBA-15 was previously functionalized with 1,1,1-trimethyl-N-(trimethylsilyl)silanamine (HMDS). In the following independent and separate control step, this material is treated with 3- azidopropyltriethoxysilane (AzPTES). When silanol groups on the external surface are accessible to the AzPTES molecules on the putatively fully functionalized external surface of SBA-15, they react with them. The azide groups on the external surface are then detectable by IR spectroscopy and elemental analysis after the control step. If all freely accessible silanol groups are indeed inert after functionalization of the external surface with HMDS, treatment with AzPTES has no effect. Both the infrared (IR) spectrum and the elemental analysis results show that SBA-15 functional- ized with HMDS has a completely inert external surface. This ensures that only the silanol groups on the pore walls of SBA-15 are functionalized with AzPTES. [2] Another important aspect in the preparation of selectively functionalized catalysts is their prepa- ration and use in reactors. Since mesoporous silica materials are often produced as powders, they must be shaped e.g. by tableting or extrusion. In addition, pressures act on catalysts when reactors are loaded and reactions take place. For example, pressures of up to 40 MPa prevail in laboratory-scale reactors and ultra-fast high-performance liquid chromatography (HPLC) systems. With this in mind, calcined SBA-15 and SBA-15, whose pores were opened by Soxhlet extraction with ethanol and subsequent heating in nitrogen were subjected to a pressure of 39 MPa for 10 min. 2 Abstract Results of different surface areas and pore volumes obtained from nitrogen physisorption measure- ments showed that both materials were stable up to this pressure. At higher pressure of 156 MPa, both samples lose parts of their porosity, as shown by the nitrogen physisorption measurements. The broadening of the characteristic SAXS reflections as well as the decrease in intensity confirm a partial destruction of the nanostructure of the samples pressed at 156 MPa. [3] Thus, the three publications of this cumulative dissertation are the basis for an efficient use of mesoporous silica materials as support for the immobilization complex catalysts and an unam- biguous determination of confinement effects. 3 Zusammenfassung Die Nachbildung von Enzymen ist eines der wichtigsten Ziele der aktuellen Katalyseforschung. Denn Enzyme verfügen über definierte Mikroumgebungen, die hohe Enantioselektivitäten er- möglichen. Somit ist es ihnen möglich gezielt Produkte herzustellen. Mit metallorganischen Katalysatoren, insbesondere wenn sie auf festen Trägern immobilisiert sind, werden im Vergle- ich zu Enzymen oft geringe Enantioselektivitäten erreicht. Darüber hinaus sind die Kosten für die Herstellung und Rückgewinnung gelöster Katalysatoren hoch. In Anbetracht von Umweltprob- lemen und Ressourcenknappheit ist es wichtig, Katalysatoren für enantioselektive und effiziente chemische Reaktionen zu entwickeln. Die Kombination der Vorteile der homogenen Katalyse, mit ähnlich hohen Enantioselektivitäten sowie Produktivitäten, wie sie Enzyme aufweisen, mit den positiven Aspekten der heterogenen Katalyse ist ein vielversprechendes Konzept, um gezielt enantiomerenreine Produkte kostengünstiger und umweltschonender herzustellen. Hinter diesem Ansatz steht unter anderem die Immobilisierung von hochselektiven Katalysatoren auf porösen Trägermaterialien wie mesoporösem Siliziumdioxid. Durch die räumliche Eingrenzung der Reak- tion in der Poren- und Kanalstruktur des Trägermaterials kann eine formselektive Reaktion zu gewünschten Produktkonfigurationen erreicht werden. Um Formselektivität zu ermöglichen, muss sichergestellt sein, dass sich die immobilisierten Katalysatoren wie beispielsweise Metallkom- plexe ausschließlich in den Mesoporen und Kanälen des Trägermaterials befinden. Daher ist die selektive und räumlich kontrollierte Funktionalisierung von mesoporösem Siliziumdioxid als Trägermaterial für katalytisch aktive Metallkomplexe sehr wichtig und kann die heterogene Katal- yse erheblich voranbringen. Um eine selektive und räumlich kontrollierte Funktionalisierung zur Nachbildung der Funktion- sweise von Enzymen zu gewährleisten, ist die genaue Kenntnis der Eigenschaften des Trägerma- terials erforderlich. Für die Bestimmung charakteristischer Materialeigenschaften wie des Git- terparameters und der Porengröße mesoporöser Siliziumdioxide stehen verschiedene Methoden zur Verfügung. In dieser Arbeit wurden vier mesoporöse Siliziumdioxide mit durchschnittlichen Porengrößen von 3 bis 11 nm verwendet, um gängige Charakterisierungsmethoden mit alternativen Zusammenfassung Methoden zu vergleichen. Die Bestimmung der Gitterparameter durch Röntgenkleinwinkelmes- sungen (SAXS) und aus Bildern des Transmissionselektronenmikroskops ergab ähnliche Werte. Unter Berücksichtigung der Tatsache, dass bei der Auswertung von Aufnahmen mit dem Trans- missionselektronenmikroskop (TEM) durch die Porenorientierung Fehler auftreten können, liefert die Charakterisierungsmethode gleichwertige Ergebnisse. Die Porengröße verschiedener meso- poröser Siliziumdioxide wurde klassisch mit Hilfe von Physisorptionsmessungen und aus den Daten von SAXS-Messungen bestimmt. Darüber hinaus wurde eine neue Methode vorgestellt, bei der Elektronendichtekarten aus SAXS-Daten erstellt und zur Bestimmung der Porengrößen verwendet werden. Da es bei der Betrachtung kleiner Poren in TEM-Aufnahmen aufgrund der unterschiedlichen Graustufen schwierig ist, den Übergang zwischen Porenwand und Pore zu bes- timmen, wurde in Proof-of-Principle-Experimenten unter Verwendung von Gold-Nanopartikeln (Au-NPs) als Sondenpartikel untersucht, ob die Siliziumdioxide als offene Porennetzwerke vor- liegen. Die mit den verschiedenen Charakterisierungsmethoden ermittelten Werte für die Poren- größe stimmen gut überein, obwohl zu beachten ist, dass Au-NPs nur einen Bereich von Poren- größen und keinen exakten Wert bestimmen können. Dennoch zeigt der Vergleich, dass die beiden neuen Charakterisierungsmethoden - die Erstellung von Elektronendichtekarten und die Verwen- dung von Au-NPs als Sondenpartikel zur Bestimmung der Porengröße - das Potenzial haben, Al- ternativen zu den etablierten Methoden zu werden. [1] Nach der Charakterisierung der möglichen Trägermaterialien folgt deren selektive und räumlich kontrollierte Funktionalisierung, der wichtigste Schritt bei der Nachbildung von Enzymen. Um sicherzustellen, dass sich die katalytisch aktive Spezies ausschließlich in den Mesoporen befindet, müssen vor deren Aufbringen auf das Trägermaterial alle frei zugänglichen Silanolgruppen auf der Partikeloberfläche des Trägermaterials inertisiert werden. Verschiedene Methoden zur slektiven und räumlich kontrollierten Funktionalisierung von porösem Siliziumdioxid sind in der Literatur beschrieben, zeigen aber nicht eindeutig, dass eine selektive und räumlich kontrollierte Funktion- alisierung der Porenwände erreicht wurde, während die äußere Oberfläche inert blieb. Das in dieser Arbeit verwendete Verfahren zur selektiven und räumlich kontrollierten Funktionalisierung von SBA-15 als Trägermaterial entwickelt die in der Literatur beschriebenen Verfahren weiter. Es unterscheidet sich durch die Einführung eines separaten und unabhängigen Kontrollschritts, der sicherstellt, dass das Funktionalisierungsverfahren zu einem Material führt bei dem sich die or- ganischen Gruppen - in diesem Fall Azidgruppen - ausschließlich an den Porenwänden befinden. Mittels Klick-Chemie können im Anschluss Linker und Edelmetallkomplexe an die Azidgrup- pen gebunden werden, um eine spezifische katalytisch aktive Spezies bereitzustellen. Vor dem Kontrollschritt wurde die äußere Oberfläche, bestehend aus Partikeloberfläche und den Porene- ingängen von SBA-15 zuvor mit 1,1,1-Trimethyl-N-(trimethylsilyl)silanamin (HMDS) funktion- alisiert. In dem folgenden, unabhängigen und separaten Kontrollschritt wird dieses Material mit 3-Azidopropyltriethoxysilan (AzPTES) behandelt. Sind noch Silanolgruppen auf der ver- meintlich voll funktionalisierten äußeren Oberfläche von SBA-15 zugänglich, reagieren sie mit dem AzPTES. Die Azidgruppen auf der äußeren Oberfläche sind nach dem Kontrollschritt durch 6 Zusammenfassung IR-Spektroskopie (IR) und Elementanalyse nachweisbar. Wenn alle frei zugänglichen Silanol- gruppen nach der Funktionalisierung der äußeren Oberfläche mit HMDS tatsächlich inert sind, hat die Behandlung mit AzPTES keine Auswirkungen. Sowohl das IR-Spektrum als auch die Ergebnisse der Elementanalyse zeigen, dass das mit HMDS funktionalisierte SBA-15 eine völlig inerte äußere Oberfläche aufweist. Dies gewährleistet, dass in einem nachfolgenden Schritt nur die Silanolgruppen an den Porenwänden von SBA-15 funktionalisiert werden. Im Rahmen dieser Arbeit wurden die Silanolgruppen der Porenwände mit AzPTES funktionalisiert, um im Anschluss mittels Klick-Chemie katalytisch aktive Metallkomplexe in das Porensystem einzubringen. [2] Ein weiterer wichtiger Aspekt bei der Herstellung von selektiv funktionalisierten Katalysatoren ist die Herstellung und Verwendung in Reaktoren. Da mesoporöse Siliziumdioxide häufig als Pulver hergestellt werden, müssen sie beispielsweise durch Tablettierung oder Extrusion in Form gebracht werden. Außerdem wirken bei der Beladung der Reaktoren sowie bei der Durchführung von Reaktionen darin Drücke auf die Katalysatoren ein. So herrschen beispielsweise in Reak- toren im Labormaßstab und in ultraschnellen Hochleistungsflüssigkeitschromatographiesystemen (HPLC) Drücke von bis zu 40 MPa. Vor diesem Hintergrund wurden kalziniertes SBA-15 und SBA-15, deren Poren durch Soxhlet Extraktion mit Ethanol und anschließendem Erhitzen in Stick- stoff geöffnet wurden, einem Druck von 39 MPa für 10 min ausgesetzt. Die Ergebnisse der ver- schiedenen Oberflächen und Porenvolumina, die aus Stickstoffphysisorptionsmessungen ermittelt wurden, ergaben, dass beide Materialien bis zu diesem Druck stabil sind. Bei höherem Druck von 156 MPa verlieren beide Proben einen Teil ihrer Porosität, wie die Ergebnisse der Stickstoffph- ysisorptionsmessungen zeigen. Die Verbreiterung der charakteristischen SAXS-Reflexe sowie die Abnahme der Intensität bestätigen eine teilweise Zerstörung der Nanostruktur der bei 156 MPa gepressten Proben. [3] Somit bilden die drei Publikationen dieser kumulativen Dissertation die Grundlage für eine ef- fiziente Nutzung von mesoporösem Siliziumdioxid als Trägermaterial für die Immobilisierung komplexer Katalysatoren zur eindeutigen Durchführung formselektiver Reaktionen. 7 1 Introduction 1.1 Porous Materials Porous materials are solids with cavities or channels, which are deeper than wide. [4;5] Since their introduction, various porous materials have been developed, differing in pore size, properties and applications. In everyday life, porous materials can be found almost everywhere. For example, foams with pores in millimeter range are used for heat [6;7] and sound insulation [8]. The remarkable properties of GoreTex® membranes for clothing are based on micropores. [9] A further application for nanoporous silica gels as desiccants are the small pouches included with many electronic de- vices and clothing. [10] In addition, mesoporous silica materials are used in separation processes or as sensors, as well as in other electro-optical technologies. [11;12] Common materials with even smaller pore sizes are zeolites. Zeolites are used in detergents as ion exchangers. [13] The wide range of properties complicates the classification of porous materials. [10] According to the IUPAC technical report, porous materials are classified by their pore size. Porous materials are divided into microporous materials with a pore size <2 nm, mesoporous materials with a pore size be- tween 2 nm and 50 nm and macroporous materials with pore sizes >50 nm. [14] This work deals with mesoporous silica materials. 1.2 Mesoporous Silica Materials In 1971, literature described the synthesis of mesoporous silica for the first time. The genera- tion of mesoporous silica materials was achieved by using self-assembled molecules. [15;16] The lack of analytical capabilities for characterization meant that remarkable properties of these ma- terials initially remained undiscovered during this time. [17;18] Later on, with advanced analytical techniques, the essential properties were recognized. Today, it is possible to prepare a variety of mesoporous silica materials with the use of triblock copolymers as structure-directing template. In this context, mesoporous silica with pore sizes ranging from 5 nm to 30 nm are feasible. [19] Due to their large surface area, silica materials have a large number of surface-active centers. The Chapter 1 Introduction functionalization of these surface-active centers leads to a change and multiplication of physical and chemical properties. Consequently, porous silica materials are not only used as catalytically active materials, but their function as support materials is also of great interest. 1.2.1 Mobil Composition of Matter Materials The Mobil Oil Corporation described the first mesoporous silica material with an ordered pore structure in 1992 under the name M41S. [20;21] For the production of these materials, a self- assembling template was surrounded by silica precursor molecules. The silica precursor molecules react with each other via condensation reactions and build the structure of the mesoporous silica material. In general, M41S materials are prepared in an alkaline environment using a quater- nary ammonium salt or gemini surfactants. [18] Quaternary ammonium salts behave like a classical surfactant in solution. Predictions about the morphology of an M41S material prepared with a quaternary ammonium salt can therefore be made via the dimensionless packing parameter g. The packing parameter is calculated from the quotient of the effective volume of the hydrophobic chain, the surface area of the hydrophilic head group and the critical hydrophobic chain length. Spherical micelles are formed when g < 0.3, rod-shaped micelles when g = 0.5 and lamellar mi- celles when g= 1. [22] The surface area of the head groups of the ammonium salt can also influence the morphology. If the surface area is large, spherical structures are preferentially formed, while rod-like and lamellar structures are formed when the head groups can be densely packed and have a high aggregation number. Different reaction conditions can affect the packing parameters and thus the order of M41S materials. M41S materials can be synthesized as powders, thin film layers on various substrates or as monoliths. [18;23;24] The M41S family groups together all porous materi- als with uniformly defined and highly ordered pores. The different Mobile Composition of Matter (MCM) materials are assigned by a sequential number after their respective acronym. [21] The best known materials of the M41S family are MCM-41 [22;25–30], MCM-48 [30–34] and MCM-50 [35–37]. MCM-41, with the space group p6mm, is a well-known material and is used within this work. The pores of MCM-41 are arranged in a honeycomb shape and separated by amorphous pore walls, yielding to a large surface area and high pore volume. Figure 1.1 shows the characteristic particle shape and pore structure of MCM-41. The uniform pores can be tuned to a diameter be- tween 1.5 nm and 20 nm during the synthesis and exhibit a narrow pore size distribution. [18] To obtain large pores, swelling agents must be added during synthesis. The pore walls are 1 nm to 1.5 nm thick which is relatively thin compared to other mesoporous silica materials resulting in low chemical and hydrothermal stability. [18;38;39] 1.2.2 Santa Barbara Amorphous Materials Santa Barbara amorphous (SBA) materials are mesoporous silica materials with various pore sizes, which can be prepared by using the nonionic triblock copolymer consisting of blocks of poly(ethylene oxide) (EO) and poly(propylene oxide) (PO). Depending on the block sizes n 10 1.2 Mesoporous Silica Materials (a) (b) Figure 1.1: (a) Image from the scanning electron microscope of the MCM-41 particles as well as (b) TEM images of the honeycomb pore structure of MCM-41. [18] and m, different triblock copolymers poly(ethylene oxide)m-poly(propylene oxide)n-poly(ethylene oxide)m (EOmPOnEOm) are known. [40;41] These kind of silica materials find use as support for metals in nanowire shape [42–46], as template for the synthesis of (inverse) carbon replicas [47–51], for immobilization of enzymes [52;53] and in lithium batteries [54]. SBA materials are used as cat- alysts in reactions [17;55–64], for the controlled release of active ingredients or antioxidants from support materials [65;66] and for recovery of heavy metals [67]. In general, SBA materials are clas- sified as relatively neutral structures since the interactions between structure-directing template and silica lattice are very weak. [11] The different SBA materials are distinguished from each other by structural properties. These properties can be adjusted by varying the number of polypropy- lene oxide and polyethylene oxide blocks in the triblock copolymer. This allows the control of the morphology, the pore structure and size during the production process. Accordingly, the ac- tual structure in each SBA material is determined by the composition of the structure-directing template or more specifically by the quotient of the number of polyethylene oxide blocks and polypropylene oxide blocks in the triblock copolymer. This results in SBA materials with lamel- lar, hexagonal or cubic mesophase structure. [18] A well-known SBA material is SBA-15, whose mesopores are hexagonally arranged, indicating the space group p6mm. [18] The worm-like particles of SBA-15 can be seen in Figure 1.2 as well 11 Chapter 1 Introduction (a) (b) Figure 1.2: (a) Images of the worm-like particle structure of SBA-15 from the scanning electron microscopy (SEM) as well as (b) TEM images of the parallel channels within the particles and the hexagonal pore structure of SBA-15. [18] 12 1.2 Mesoporous Silica Materials as the typical pore and channel structure. The pore walls, which are 3 nm to 6 nm thick, con- sist of micropores, which are responsible for the hydrothermal stability. [68] Varying the number of polyethylene oxide blocks causes a change in the number and proportion of micropores in the pore walls, as well as the pore wall thickness. [51;69–73] The smaller the polyethylene block m, the smaller the micropores are. [69–71] By shortening or lengthening the polypropylene blocks n in the triblock copolymer, the size of the mesopores can be controlled by their hydrophobic properties. Concluding from this, it is possible to influence the size of mesopores as well as the microporos- ity of pore walls by varying the ratio of polyethylene and polypropylene blocks of the triblock copolymer. This is attributed to the fact that polypropylene blocks form agglomerates in solu- tion and represent the negative of the later mesoporous tunnel structure. [40;69;70;72;74] The larger the polypropylene oxide block, the larger the agglomerates in the aqueous phase leading to larger pores. [69;71] In addition to the template and the inorganic precursor molecules, reaction conditions such as temperature [68;69;75], pH value [76], and additives influence the material. [18;77] Solvents, swelling agents such as alkanes and organic compounds like mesitylene or electrolytes can be used as additives. [78–84] Depending on the additive, different functionalities of the additives are in- corporated into the material. In the case of short-chain alkanes, literature describes that molecules attach themselves between the tails of the template molecules. For longer-chain alkanes, such as decane, a core-shell structure is formed with molecules of the additive as the core and an enclosed layer of the template molecules as shell. [85;86] Another SBA material is SBA-16, which has hydrothermal properties similar to the ones of SBA- 15. The difference between SBA-15 and SBA-16 is the use of different templates during synthesis. For example, triblock copolymers with large polyethylene oxide blocks such as EO106PO70EO106 are used for synthesis of SBA-16. [87] SBA-16 has a three-dimensional cubic cage structure consist- ing of two non-intersecting three-dimensional channel systems with the space group Im3m. [87;88] The intersections of both channel systems form spherical cavities. [18] The nature of the hysteresis loop within the sorption isotherms suggests bottle-shaped pores. [88;89] In general, as with SBA-15, changes in the morphology of SBA-16 are controllable by the reaction conditions. [90–97] 1.2.3 Plugged Hexagonal Templated Silica Material An analog to SBA-15 is plugged hexagonal templated silica (PHTS) material. Compared to SBA- 15, the ratio of silicon source to structure-directing template is higher for PHTS. [51;98–100] The difference between both materials is the fact that amorphous nanoparticles are in the mesoporous channels of PHTS. The nanoparticles in the pore channels have a positive effect on the stability of PHTS through their support function. [101;102] Despite the nanoparticles in the channel system, the hexagonal pore structure of PHTS as well as the pore size remains similar to SBA-15 (Figure 1.3). Just like SBA-15, the pore walls are interspersed with micropores. The amount of grafted pores can be adjusted by varying the synthesis conditions from only open to only grafted pores. [98;103;104] Moreover, by changing reaction conditions, size and stability of the nanoparticles located in the 13 Chapter 1 Introduction (a) (b) Figure 1.3: (a) SEM images of the worm-like particle structure of PHTS as well as (b) a TEM im- age of the channels within the particles and the hexagonal pore structure of PHTS. [18] pore channels can be varied. [102;105] The pore size and the size of the nanoparticles can be ad- justed, for example, via the synthesis temperature, means the higher the temperature, the larger the pore size and nanoparticles. Another difference of PHTS compared to SBA-15 is the larger mi- cropore volume. This is because the micropores in the nanoparticles increase the total micropore volume of PHTS. [99] Furthermore, the morphology of PHTS can be influenced by the proportion of the silica precursor and by different reaction temperatures. Possible morphologies are smooth fibers, rough fibers, and spherical particles. At low temperatures and low silica precursor concen- trations, smooth PHTS fibers are formed, while high temperatures form spherical particles. The relationship between reaction temperature and morphology can be found in the cloud point of the template. The cloud point is the temperature at which a clear liquid becomes turbid by crystalliza- tion under defined conditions. Other factors why there could be different morphologies for PHTS 14 1.3 Process for the Preparation of Mesoporous Silica Materials are the degree of polymerization of the silica matrix and the amount of mesoporous silica material already formed. [100] 1.2.4 Mesostructured Cellular Foam Another mesoporous silica material is the mesostructured cellular foam (MCF). This material has a sponge-like structure (Figure 1.4) with thick pore walls responsible for its hydrothermal stability. [18] The pores of MCF are bottle-shaped and are accessible through 5 nm to 20 nm wide openings. [18] By adding ammonium fluoride during the synthesis, the openings can be selectively enlarged by 50 % to 80 %. [106;107] As a result of the large pores, rapid kinetic mass transport is possible to transport large molecules such as polymers or enzymes. [108–111] Other structural properties can be influenced during the production of MCF by changing the reaction conditions such as temperature, reaction time, adjusted pH value, as well as additives and swelling agents such as mesitylene. [18;106–108;112–116] The swelling agent causes the micelles to expand, creating larger micelle radii in the sponge-like foam structure. MCF can be produced as a powder or monolith. [117–119] The material can be modified to find use in biocatalysis, catalysis, sorption, and controlled release of sorbed substances. [83;109–111;120–124] (a) (b) Figure 1.4: (a) SEM and (b) TEM images of the sponge-like structure of MCF. [18] 1.3 Process for the Preparation of Mesoporous Silica Materials 1.3.1 Synthesis of Mesoporous Silica Materials In general, preparation of most inorganic mesoporous materials is based on the use of organic template molecules. These are brought into solution and form micelles due to their composition. The subsequently added inorganic precursor molecules enclose the template micelles. During hy- drothermal treatment, condensation reactions take place to build a lattice around the micelles of the structure-directing template (Figure 1.5). [18] In the literature, simple oligomers or liquid crys- 15 Chapter 1 Introduction Figure 1.5: Schematic representation of the synthesis route of mesoporous silica materials ac- cording to the cooperative self-assembly mechanism. talline phases are used as templates for the preparation of mesoporous silica materials. Using such liquid crystalline phases consisting of neutral, cationic and anionic diblock copolymers, meso- porous silica materials with large spherical pores [125], hexagonally arranged cylindrical pores [126] and lamellar vesicular pores [127] are successfully prepared in acidic environment. Besides diblock copolymers, triblock copolymers are suitable for producing mesoporous silica materials in acidic environment. [41;87;128–130] One of the most useful groups of surfactants are triblock copolymers of poly(ethylene oxide)m-poly(propylene oxide)n-poly(ethylene oxide)m (EOmPOnEOm). These triblock copolymers consisting of polyethylene oxide and polypropylene oxide blocks arrange mi- celles in aqueous reaction solutions and have the ability to form liquid crystal structures. In the case of micelle formation, the interior of the micelles consist of polypropylene oxide blocks, while the shell of the micelles consist of polyethylene oxide blocks. [17;18] Depending on the template, the structural properties of mesoporous silica materials can be influenced by the number of re- peat units of individual polymer blocks. [18] When the structure of the formed mesoporous silica materials is sufficiently built up by condensation reactions of the precursor molecules, the pores are filled with structure-directing template (Figure 1.5, as-synthesized). The structure-directing template molecules are no longer needed and can be removed in a further step. As a result, the pores are open and a porous structure is built (Figure 1.5, mesoporous silica material). [11;131] The mechanistic consideration of the pore structure formation of mesoporous materials bases on dif- ferent pathways. In the following, the mechanism of cooperative self-assembly, true liquid crystal templating (TLCT) and nanocoasting are considered. [131] In the mechanism of cooperative self-assembly, the lyotropic liquid crystalline phase is already present but at low surfactant concentrations. This is possible if there is cooperative self-assembly of the structure-directing template and the already added inorganic precursor. As with the TLCT mechanism, liquid crystalline phases with hexagonal, cubic or lamellar arrangement can be formed in the cooperative self-assembly process. [132;133] The cooperative self-assembly formation mecha- nism was first observed under basic reaction conditions, but can also occur in acidic environments as in the synthesis of SBA-15. Regardless of the pH value during preparation of mesoporous sil- ica materials, this formation mechanism requires the occurrence of attractive interactions between molecules of the structure-directing template and the silica precursor molecules. Inclusion of the structure-directing template is only guaranteed without phase separation in this case. [131] The pos- 16 1.3 Process for the Preparation of Mesoporous Silica Materials sible interactions between inorganic precursor molecule (I) and structure-directing template, the surfactant (S), are shown in Figure 1.6. [131;134;135] The direct co-condensation of mesoporous materials takes place in an alkaline environment. The quaternary ammonium surfactant molecules used as structure-directing template are present as cations (S+), while the silica precursor molecules are in their anionic form (I– ). This formation mechanism is referred to as the S+I– pathway. [131;135] The formation mechanism according to the S+I– pathway can be carried out in acidic reaction solutions. This means that the pH value is below the isoelectric point of the silanol groups of the silica precursor. In this case, the silica precursor molecules are positively charged (I+). Since the structure-directing template molecules are positively charged by the acidic environment (S+), negatively charged counter ions (X– ) are required to ensure interactions between surfactant and inorganic species. This formation mechanism is referred to the S+X– I+ pathway. [131;135] Another possible formation mechanism is the S– I+ pathway. In this case, the formation of meso- porous materials takes place in acidic environment, which is why the structure-directing template molecules are negatively charged (S– ). In contrast, the silanol groups of the silica precursor are positively charged (I+). A counterion is not necessary in this formation mechanism. [131;135] In the S– X+I– pathway, both the structure-directing template and the silica precursor molecules are present in their anionic form in an alkaline environment ((S– ) and (I– )). Thus, a positively charged counter ion (X+) is required during this formation mechanism. [131;135] In addition to the ionic formation mechanisms, non-ionic mechanisms are also known. In this case, the interactions are hydrogen bond interactions instead of electrostatic interactions between the structure-directing template and the silica precursor molecules, as it is the case with ionic for- mation mechanisms. Hydrogen bond interactions occur when nonionic surfactants (S0) are used. The silanol groups of the silica precursor molecules are neutral (I0), so this formation mechanism is called the S0I0pathway. If the silanol groups of the silica precursor molecules are positively charged, charge equalization must occur via a free electron pair. As a result, an ion pair is present and the formation mechanism is called the S0(XI)0 pathway. [131;135] Another process for the preparation of mesoporous silica materials is the true liquid crystal tem- plating (TLCT) process. In the TLCT mechanism, the template concentration is so high that an ordered lyotropic liquid crystal phase is formed as a function of temperature and pH value. During the formation process, the structure-directing template molecules form spherical micelles initially. Afterwards, the micelles agglomerate into rod-shaped micelles. In the next step, template rods aggregate to form a lyotropic liquid crystal phase, which determines the structure of the meso- porous silica material. The precursor molecules for the inorganic silica material are added to the surface-active liquid crystalline phase and condensate to the silica matrix during the hydrothermal treatment. The formation process following the TLCT mechanism was proposed in the early 1990s as a possible mechanism for the formation of MCM-41. [20] Later, the mechanism was confirmed by the hydrolysis of tetramethyl orthosilicate (TMOS) within a liquid crystalline phase. [136] It was shown that the lyotropic liquid crystal phase was destroyed when methanol was formed as side 17 Chapter 1 Introduction Figure 1.6: Schem atic representation ofthe possible interactions betw een the inorganic precursorm olecule (I)and the structure-directing tem plate,the surfactant(S)during the synthesis ofm esoporous m aterials via the self-assem bly form ation m echanism . In tw o cases,counter ions (X ) are required to ensure interactions betw een surfactantand inorganic species. T he schem a w as created according to the literature. [131;134;135] 18 1.3 Process for the Preparation of Mesoporous Silica Materials product. After the evaporation of alcohol, the lyotropic liquid crystal phase was reformed. There- fore, it is difficult to make a sharp distinction from the mechanism of cooperative self-assembly. [17] In contrast to the TLCT mechanism, the cooperative self-assembly mechanism described above relies on cooperative interactions between the surfactant and the inorganic species, yielding an organic-inorganic mesostructured composite. [131] Overall, various synthetic routes for mesoporous silica materials are known and it is quite pos- sible that a material can be produced via several synthetic routes. However, independent of the production process of mesoporous silica materials, the cavities and channels must be opened by removing the template (Chapter 1.3.2). 1.3.2 Template Removal Organic template molecules are usually used as structure-directing template to produce meso- porous silica materials (Chapter 1.3.1). They are often dissolved in an acidic solution before in- organic precursor molecules are added. Inorganic precursor molecules arrange themselves around the micelles, formed by the structure-directing template. During hydrothermal treatment, conden- sation reactions of the precursor molecules take place and build a negative of the micelles. [18] After formation of the inorganic matrix of the mesoporous silica material, the structure-directing tem- plate is no longer needed and must be removed to make the pores accessible. [11] Template removal from the inorganic matrix is not trivial, because it is possible that characteristic properties of the mesoporous silica material will change significantly as a result of the template removal method. [11] Various methods are available in the literature describing how to remove the structure-directing template from pores and channels. [19;103;137;138] Depending on the interactions between template molecules and the embedded matrix and the adaptation of the template to the matrix, different approaches are suitable for an efficient and gentle removal of the template molecules. Ideally, the organic template molecules are removed without changing the properties of the host structure. This means that the inorganic precursor remains as a negative of the template micelles structure and the surface properties of the mesoporous silica material are still preserved. It is important to ensure the removal of the entire structure-directing template. [137] As described in the literature, the amount of removed template as well as the change of the mesoporous silica structure depends on the cho- sen method. [11;18;139] It should be noted that the removal of the structure-directing template is not always the same owing to the structural properties of different mesoporous silica materials. [11] The most commonly used method to open the pores of as-synthesized mesoporous silica materials is calcination. In this process, the as-synthesized mesoporous silica materials are heated up to temperatures of 550 ◦C in a stream of oxygen or air, and the temperature is maintained for 4 h to 8 h. [11] At these temperatures, organic structure-directing template molecules burn and make the pores of the mesoporous silica material accessible. However, if the combustion is not complete, carbon species can be formed. These can block pores and channels. The multistage decomposition process of the structure-directing template of as-synthesized SBA-15 can be illustrated by thermo- 19 Chapter 1 Introduction analytical investigations. At temperatures below 150 ◦C, the physisorbed water molecules firstly desorb in an endothermic process from the surface of SBA-15. Between 150 ◦C and 280 ◦C, ther- mal decomposition of the template molecules occurs in an exothermic step. All fragments detected by the mass spectrometer during the exothermic process can be assigned to the structure-directing template. [11;140] The mass loss of this decomposition step is more than 40 % and is assigned to carbon dioxide and water. At temperatures above 280 ◦C, a further mass loss of 10 % is observ- able, which is attributable to the removal of water and residues of the carbon species. This leads to a total mass loss of more than 50 %. [11;19;68;87] Temperature-dependent X-ray analytical studies show structural changes of SBA-15 during calci- nation. When as-synthesized SBA-15 is heated up to 200 ◦C, an increase in the intensity of the scattering reflections can be observed. This means that no changes in the hexagonal structure are observed up to this temperature. Between 200 ◦C and 250 ◦C, the intensities of the scattering reflections are initially weaker than at temperatures below 200 ◦C. The intensities again become stronger when the temperature increases above 250 ◦C. At temperatures above 250 ◦C, addition- ally a shift of the scattering reflections to smaller angles can be observed. Shifts of the scattering reflections to smaller angles are equivalent to the shrinkage of the silica lattice compared to the as- synthesized SBA-15. [11;141–143] Accordingly, calcined SBA-15 shrinks by about 13 % after cooling to room temperature. [11] Shrinkage means the decrease of the lattice parameter of the mesoporous silica material. The ef- fect can be explained by the restructuring of the silica lattice during the thermal treatment. The rearrangement of the silica lattice is possible as not all possible condensation reactions take place during the synthesis of mesoporous silica materials, causing structural defects. The high tem- perature during the calcination enables restructuring and quenching of defects within the silica lattice. [141–143] Since the structure shrinkage differs between various mesoporous silica materials such as SBA and MCM materials, it is assumed that silica materials catalyze the thermal decom- position and oxidation of the template molecules in presence of the oxygen, even at low temper- atures. [11;68] Carbon species and water formed in these reactions are removed from the pores and channels only at higher temperatures. In this process, condensation reactions are still possible, which shrink the structure further. [11] Another template removal method is extraction of the structure-directing template with an extract- ing agent. In the literature, solvents such as ethanol or propanol are used for this purpose. The use of sulfuric acid or hydrochloric ethanol solutions and the use of neutral salts in ethanol are also described to remove the structure-directing template from the pores and channels of meso- porous silica materials. [139;144–146] Regardless of the extraction agent, mesoporous silica materials are treated under reflux or by Soxhlet extraction to make the pores and channels accessible. [11] When as-synthesized mesoporous silica materials are treated with up to 60 wt% sulfuric acid in ethanol, polypropylene oxide blocks are removed. Due to the treatment with sulfuric acid, ether groups are cleaved. The built fragments are removed by diffusion from the channel and pore structure, resulting in opened mesopores whereas the micropores remain closed. [147] The influ- 20 1.3 Process for the Preparation of Mesoporous Silica Materials ence of acid concentration during the extraction process as well as the extraction time were in- vestigated. [139] The X-ray analytical studies show that the hexagonal structure shrinks at a con- centration of 60 wt% sulfuric acid solution in the same way as during the calcination process. At the same time, it should be noted that high concentrations are not necessary, as a concentration of 48 wt% sulfuric acid solution is sufficient to remove the polypropylene oxide blocks completely. The micropores sealed after the treatment with sulfuric acid can be accessed by a subsequent tem- perature treatment. [147] Compared to the calcined samples, the samples treated with sulfuric acid have larger cell constants and larger pore sizes, but thinner pore walls. Reaction time studies showed that both the amount of removed template and the structural properties of SBA-15 are affected. Regardless of the con- centration and reaction time, the micropores remain closed and can only be opened by thermal treatment. The micropore volume of mesoporous silica materials treated with sulfuric acid and then calcined at 200 ◦C is about twice that of silica materials, which are calcined at 540 ◦C. Even when sulfuric acid treated mesoporous silica materials are additionally calcined at 540 ◦C, the micropore volume is significantly larger than that of a mesoporous silica material calcined exclu- sively at 540 ◦C. The gradual removal of the structure-directing template is confirmed by nitrogen physisorption measurements of the silica material treated with 60 wt% sulfuric acid solution. [139] Another alternative for removing the structure-directing template from the pores of silica materials is microwave digestion. Microwave digestion has the same effect as calcination, but shortens the reaction time for template removal significantly compared to the calcination process. [148] Major advantages of this method are the smaller structural shrinkage and the retention of a higher degree of freely accessible silanol groups. Furthermore, higher surface areas and pore volumes are ob- tained by removing the structure-directing template with microwave digestion. This is clearly demonstrated by nitrogen physisorption measurements. At the same time, elemental analysis shows that more than 99.5 % of the template is removed by microwave digestion. [137] In summary, it is shown that the calcination process removes the structure-directing template com- pletely but at the expense of the mesopore structure. Extraction of structure-directing template with sulfuric acid showed that the sulfuric acid concentration and the treatment time influence the degradation of the structure-directing template and thus the structural properties of SBA-15. The comparison of silica materials that are exclusively calcined with those, whose structure-directing template is removed by extraction in a first step and then thermally treated, shows that template removal via extraction is a good alternative. Further advantages of extraction arise from an ecolog- ical and economic point of view when the triblock copolymer used as a template and the solvents used for extraction are recycled. A big disadvantage of extraction of the template molecules is that during the extraction often not the entire structure-directing template from the channel-pore system is removed. [18] 21 Chapter 1 Introduction 1.4 Properties of Mesoporous Silica Materials 1.4.1 Silanol Groups on Surfaces of Mesoporous Silica Materials When the structure-directing template is removed from the as-synthesized mesoporous silica ma- terial, silanol groups (SiOH) are on the whole surface. These are either isolated silanol groups or silanol groups interacting via hydrogen bridges with each other or geminal silanol groups (Figure 1.7). [140] Type and number of accessible silanol groups on the surface of mesoporous silica ma- terials are of enormous importance for various fields of applications. For example, the number of accessible silanol groups decide how many catalyst complexes can be anchored to the surface. Figure 1.7: Schematic representation of the single, geminal and hydrogen bridged silanol groups on silica surfaces [140] and the assignment of Q2, Q3 and Q4 sites in 29Si MAS NMR spectra of MCM and SBA materials [149]. Acidic properties introduced via silanol groups depend, among other things, on the nature of the accessible silanol groups on the silica surface. [150;151] Isolated or geminal silanol groups have higher acid strength compared to silanol groups connected by hydrogen bridges, which exhibit a non-acidic character. [150;152] For this reason, only isolated and geminal silanol groups can be functionalized, whereas the non-acidic silanol groups do not react with the molecules of the func- tionalization reagents. [140] Therefore, knowledge about the nature of accessible silanol groups on the surface of silica materials is important depending on the application. Various analytical meth- ods are known to investigate the nature of silanol groups on the silica surface. [92;149–151;153] Among other methods, the 29Si magic angle spinning (MAS) nuclear magnetic resonance (NMR) measurement is one of them. [149;153] The 29Si MAS NMR measurement is a bulk method. This means that the 29Si MAS NMR measurement examines not only the silanol groups on the surface of the silica materials, but also silicon atoms of the SiO4 tetrahedra in the lattice structure of the silica material. [149] The 29Si MAS NMR spectrum shows signals for isolated and geminal silanol groups as well as for silanol groups, which are connected by hydrogen bridges. Accordingly, the spectrum of a silica material exhibits the signal of Q4 groups at −110 ppm. The signal is charac- 22 1.4 Properties of Mesoporous Silica Materials teristic for four bonded SiO4 tetrahedra in the lattice structure of silica materials (Figure 1.7). The silanol groups on the silica surface are observed at −101 ppm and −92 ppm. They are assigned to the silicon atoms to which one silanol group SiOH (Q3 groups) or two isolated silanol groups Si(OH)2 (Q2 groups) are bound. [149;151;153] In the 29Si MAS NMR spectrum of as-synthesized MCM-41, the signal from Q3 groups has the highest intensity compared to the signals for Q2 and Q4 groups. Calcination, which removes the structure-directing template from the pores, alters the 29Si MAS NMR spectrum. Corresponding to this aspect, the 29Si MAS NMR spectrum no longer shows a splitting of the individual signals. Rather, a broad, fuzzy signal is observed exclusively at a chemi- cal shift of −107 ppm. [153] A possible reason for this is that free silanol groups in the pore walls of mesoporous silica materials of the M41S family combine by condensation reactions. [154] Broad- ening and shifting of signals occur from changes in Si-O bond length and Si-O-Si angle. When calcined mesoporous silica materials of the M41S family are treated hydrothermally, the signals of Q2 and Q3 groups reappear in the 29Si MAS NMR spectrum instead of the broadened signal. [153] The 29Si MAS NMR spectrum of calcined SBA-15 shows a different distribution of signal inten- sities of Q2, Q3, and Q4 groups compared to the 29Si MAS NMR spectrum of calcined MCM-41. The signal from Q4 groups exhibits the highest intensity. If the 29Si cross polarization (CP) MAS NMR spectra are considered, it can be seen that most of the silicon atoms are surrounded by O-Si units. This indicates that most of the silicon atoms are in the pore walls rather than on the surface of SBA-15. [155] The ratio of the signals from Q3 and Q4 groups is 1 : 3 for MCM-41 and 1 : 4 for SBA-15. Differences in the ratios can be attributed to various pore sizes and pore wall thicknesses. The Q3:Q4 ratio confirms that SBA-15 has thicker pore walls compared to MCM-41. [156] In addition to the investigations of silanol groups by 29Si MAS NMR measurements, it is possible to distinguish them by IR spectroscopy. [149;150;153] Overlaps of the individual bands sometimes make assignment difficult. This complicates the quantitative determination of different types of silanol groups. [157] The presence of isolated and geminal silanol groups can be evidenced by the band between 3760 cm−1 and 3735 cm−1. The band between 3620 cm−1 and 3200 cm−1 is as- signed to interacting silanol groups. This band often overlaps with bands of isolated silanol groups and Si-O bonds in the silica lattice. [157] It is described that silanol groups connected by hydrogen bridges appear as a shoulder of the band of isolated silanol groups at 3710 cm−1. [149] Besides the knowledge of the nature of silanol groups, it is important to characterize the number of accessible silanol groups on the silica surfaces in terms of the chemical properties of silica materials. Quantification is difficult because water from air is easily adsorbed onto the silanol groups. Accurate quantification by MAS NMR, Fourier-transform infrared spectroscopy (FTIR), selective chemisorption, deuterium exchange, or mass spectroscopy (MS) is usually hampered by the need of very controlled conditions for sample preparation and the measurement itself. [149] It is impossible to localize silanol groups on different surfaces using these methods. [151] Another problem is the comparability of data known from the literature about the silanol group density on silica surfaces. Table 1.1 lists exemplary silanol group densities from literature. The silanol 23 Chapter 1 Introduction Table 1.1: Silanol group densities on surfaces of MCM- and SBA-type materials reported in the literature. silanol group density of silanol group density of reference MCM-type materials SBA-type materials [SiOH/nm2] [SiOH/nm2] ZHAO ET AL. [140] 2.5 - 3 6 - 8 RAMÍREZ ET AL. [150] not investigated 3.4 - 6.5 KOZLOVA AND KIRIK [158] 3 - 4 5 - 6 group density on surfaces of MCM-41 reported by ZHAO ET AL. ranges from 2.5 SiOH/nm2 to 3 SiOH/nm2, while the silanol group density on surfaces of SBA-15 ranges from 6 SiOH/nm2 to 8 SiOH/nm2. [140] For comparison, KOZLOVA AND KIRIK indicated a silanol group density on surfaces of MCM-41 ranging from 3 SiOH/nm2 to 4 SiOH/nm2 and a silanol group density on the surface of SBA-15 ranging from 5 SiOH/nm2 to 6 SiOH/nm2. [158] Differences in the data can be explained by various reaction conditions and parameters used during the determination of the silanol group density. [149] For example, an influence on the quantification is the temperature dependency of silanol groups. With increasing temperature, a decrease of the silanol group con- densation is visible as result of random condensation reactions. Silanol groups, which interact via hydrogen bridges convert into free silanol groups. At temperatures above 600 ◦C, the number of free silanol groups decreases. This is due to the formation of siloxane bridges by further conden- sation reactions. The formation of these siloxane bridges can lead to the collapse of surface and pore structure. [92;159] Various approaches for determining silanol group densities on silica surfaces and their accessibility are discussed in the literature. One possibility is the determination based on hydrated and dehydrated silica and the investigation with IR spectroscopy. According to this, the bands of water in the IR spectrum occur at 1630 cm−1 and 5260 cm−1, while the ones for silanol groups occur in the ranges of 3800 cm−1 to 3000 cm−1 and of 4800 cm−1 to 4200 cm−1. [160] The quantification of silanol groups using diffuse infrared reflectance Fourier transform (DRIFT) spectroscopy is possible. DRIFT measurements using pyridine as probe molecules are suitable for this purpose. [150] The band appearing in the DRIFT spectrum at 1595 cm−1 can be attributed to pyridine molecules interacting with silanol groups via hydrogen bridges. In contrast, the band at 1446 cm−1 is characteristic for weaker interactions between isolated silanol groups and pyridine molecules. [92;161] The broad band at 3050 cm−1 in the DRIFT spectrum masks the vibrational bands of isolated and hydrogen bridged silanol groups interacting with pyridine molecules. [150] For determining the number of isolated silicon groups and via hydrogen bridges connected ones, the peak areas of pyridine bands, the surface area of the silica material from nitrogen physisorption measurements, and the amount of desorbed pyridine are needed. The amount of desorbed pyri- dine is determined by thermogravimetric measurements. In the literature, silanol group densities between 3.4 SiOH/nm2 and 6.5 SiOH/nm2 are indicated for SBA-3. [150] Another discussed way to determine silanol group densities is treating silica materials with benzylamine. Afterwards, 24 1.4 Properties of Mesoporous Silica Materials the amount of benzylamine adsorbed by the silica material is determined spectroscopically. As for nitrogen physisorption measurements, the adsorption isotherms show the typical behavior of forming a monolayer before reaching a plateau at higher benzylamine concentrations (1.7 µmol benzylamine per m2 for SBA-15). Thus, the estimated silanol group density is approximately equal to 1 SiOH/nm2. It should be noted that a considerable part of the surface determined by nitrogen physisorption measurements is inaccessible to benzylamine molecules. The microporous wall of SBA-15 is the reason for this. The micropores are accessible for nitrogen molecules, but not for benzylamine molecules, which is why no conclusions can be drawn about the proportions of various silanol groups. [151] It is also possible to determine silanol group densities by thermogravimetric measurements. As with IR spectroscopic investigations, the silica material is loaded with pyridine. Depending on the interaction of pyridine molecules with various silanol groups, thermogravimetric measurements yield expected peaks between 50 ◦C and 100 ◦C and between 120 ◦C and 170 ◦C. The desorp- tion features ranging between 50 ◦C and 100 ◦C are attributed to desorption of pyridine molecules interacting with hydrogen bridged silanol groups. [150] Pyridine desorbing from isolated silanol groups is classified within the peak between 120 ◦C and 170 ◦C. [162] However, thermogravimetric studies identify three peaks rather than just two. The first two peaks at 50 ◦C and 100 ◦C are also assigned to desorption processes of pyridine molecules interacting with hydrogen bridged and iso- lated silanol groups during the study. In contrast, the peak in the temperature range between 140 ◦C and 150 ◦C is assigned to desorption of pyridine molecules from both isolated silanol groups and silanol groups connected by hydrogen bridges in small cavities. Depending on how small the cavities or their entries are, the penetration of pyridine is hindered, giving rise to a broad band. Determining the number of silanol groups, one pyridine molecule is assumed to interact with each pair of silanol groups connected by hydrogen bridges. Using the findings of thermogravimetric measurements, silanol group densities between 3 SiOH/nm2 and 5 SiOH/nm2 are observed as a function of calcination temperature and duration. Most of the identified silanol groups are silanol groups linked by hydrogen bridges. [150] The comparison of silanol group densities determined from IR spectroscopic and thermogravi- metric investigations reveals that silanol group densities determined by IR spectroscopy are higher than these determined by thermogravimetric measurements. The pyridine loss of the third peak is assigned to free silanol groups and, therefore, only one pyridine molecule is counted per silanol group, which might be the reason for the observed discrepancy. [150] In summary, it can be stated that especially the quantification of the silanol group density of silica materials poses a challenge. Overlaps of the individual bands sometimes make assignment diffi- cult. Since the density of silanol groups depends on many factors, literature values can only be used as estimates. 25 Chapter 1 Introduction 1.4.2 Characterization of the Pore Size of Mesoporous Silica Materials After the synthesis of mesoporous silica materials, the structure-directing template must be re- moved. The hexagonal structure of mesoporous silica materials remains, whereby the thickness of the pore walls depends on the material itself and the conditions during the synthesis. The structure of mesoporous silica materials is of great interest in determining the most suitable material for various applications. Mesoporous silica materials consist of voids, i.e. pores or channels, sepa- rated by the silica matrix. Depending on the pore sizes and the associated pore walls of different thicknesses, the hydrothermal, mechanical, and chemical stabilities of the materials differ (Chapter 1.4.3). [163] Various methods are known for determining pore sizes as well as pore wall thicknesses of silica materials. Based on nitrogen physisorption measurements, the detection of micro- and mesopores is possible using various methods. [14] For sorption studies, gases such as nitrogen, carbon monoxide, and argon, or vapors like water vapor are used as adsorptives. [164] Depending on the pore size and the surface properties, the appropriate adsorptive must be selected. In physisorption measurements, the micropores are initially filled with the adsorptive at low relative pressures. The exact range of the relative pressure depends on the shape and size of the micropores as well as on the size of the adsorptive molecules and the interactions between the molecules of the adsorptive and between adsorbent and adsorptive. When filling, a distinction is made between ultra micropores, which are no more than two or three molecular diameters in size, depending on the pore geometry and larger micropores. After filling the ultra micropores, which is called "primary micropore filling", the larger micropores are filled in a secondary process at higher relative pressures. In this process, the interactions between adsorp- tive and adsorbent decrease, while the cooperative interactions between adsorptive and adsorbate increase in a confined space. Thus, the choice of adsorptive is important. Nitrogen is frequently used for the analysis of micropores and mesopores in the standard measurement at 77 K. However, due to the quadrupolar properties nitrogen is not optimal, since it leads to specific interactions with a variety of surface functional groups and exposed ions. This affects the orientation of the adsorbed nitrogen molecules on the surface of the adsorbent and strongly influences the filling pressure of the micropores, which is why extremely low relative pressures are required. In this pressure range, the diffusion rate is slow, which makes it difficult to measure equilibrium adsorption isotherms. It must be considered that nitrogen molecules can be adsorbed at the entrances of narrow micropores and close them. As a result, the pore filling pressure does not correlate uniquely with pore size and structure. Despite the potential difficulties, the literature mainly contains data on nitrogen physisorption measurements. Aside from nitrogen, argon is a relatively common adsorptive. It does not have quadrupolar prop- erties and therefore does not exhibit specific interactions with surface functional groups. When argon physisorption measurements are performed at a temperature of 77 K, the analysis of the isotherms is difficult, while the cryogenic temperature is 87 K. [14] One reason why the advantages 26 1.4 Properties of Mesoporous Silica Materials of argon in determining pore size and surface area do not yield the expected benefits is that argon physisorption measurements at 77 K make it impossible to characterize the pores in the microp- orous and mesoporous regions. At 77 K, the measurement temperature is about 6.5 K below the triple point temperature of bulk argon. Therefore, only pores with a pore size of less than 12 nm can be characterized. This can be avoided by performing argon physisorption measurements at 87 K. [165] Another reason why argon physisorption measurements at 77 K are not optimal is that although the argon adsorption isotherms at 77 K are shifted to higher relative pressures compared to those of nitrogen physisorption measurements, the shift is not large enough compared to argon physisorption measurements at 87 K. Accordingly, adsorption in the narrow micropores occurs at relative pressures lower than 10−5. At these pressures, the adsorption kinetics are very slow, so the supercooled state of argon in the micropores should be assumed. [166] For these reasons, argon physisorption measurements are performed usually at 87 K. [5;166;167] The higher tempera- ture allows the narrow micropores to be filled with argon at much higher relative pressures. This accelerates equilibration and enables the measurement of adsorption isotherms with high resolu- tion. Furthermore, it leads to measurements with a more direct correlation between the pore filling pressure and the confinement effect (despite the dependence on pore size and shape). For zeolitic materials, metal-organic frameworks (MOFs), and some oxides, as well as for activated carbon, this is particularly important. Another adsorptive that is used rather infrequently is carbon dioxide. Kinetic limitations at cryo- genic temperatures (87 K, 77 K) limit the adsorption of argon and nitrogen for the characterization of very narrow micropores. This can be circumvented by using carbon dioxide as adsorptive at 277 K. [14] With the correct choice of the adsorptive, the determination of the micropore volume from type I physisorption isotherms is relatively straightforward. This is caused by the horizontal plateau of the isotherm which corresponds to the limited uptake for the micropores. Thus, the capacity can be considered as the adsorption of the gas at the corresponding measurement temper- ature as the micropore volume. For the conversion of the capacity to the micropore volume, the pores are assumed to be filled with the condensed adsorptive in the normal liquid state (Gurvich rule). [5;167] The problem is that the plateau of the adsorption isotherm is rarely horizontal, most microporous materials having a large external surface area and additional mesopores. This means that the Gurvich rule for determining the micropore volume is not always applicable. Several pro- posed solutions are known to circumvent this problem. For example, an empirical comparison of an isotherm with a suitable standard is performed. The standard must be a non-porous reference material with a similar chemical composition. The t-plot method is a standard multilayer thickness curve and depends on the application of the Brunnauer-Emmett-Teller (BET) method, which cannot be always applied. This problem can be circumvented with the αs-plot method. It does not require evaluation of monolayer capacitance and is more adaptable compared to the t-plot method. [5;167] Effects of micropore size and shape on molecular packing are not considered in this method. 27 Chapter 1 Introduction In many cases, it is useful to apply methods related to molecular simulation (MC) and density functional theory (DFT) to determine the micropore volume. These methods describe the config- uration of the adsorbed phase at molecular level and are superior to empirical and semi-empirical methods. They offer a more reliable approach to pore size analysis over the entire nanopore range. The basic principle of these methods is to describe the distribution of adsorbed molecules in the pores at the molecular level. This provides detailed information about the local fluid structure near the adsorbent surface. A pore model depends on the interaction potential between liquid and solid, so various pore shape models (e.g., slot, cylindrical, spherical, and hybrid geometries) have been developed for different classes of materials such as carbons, silicas, and zeolites. For this reason, the application of methods relying on MC and DFT is only useful if a given nanoporous system is compatible with the chosen MC/DFT kernel. [14] Besides the micropores, the pore volume and pore size of mesopores can be derived from ph- ysisorption measurements. If the material is free of macropores, a type IV isotherm is obtained that is nearly horizontal at high relative pressures. The pore volume can then be derived from the amount of adsorbent at a relative pressure close to the relative pressure of one. The pores are assumed to be filled with the adsorbent in the liquid state. [5;167] The analysis of the mesopores is based on the modified Kelvin equation (Barrett, Joyner, and Halenda (BJH) and Broeckhoff and de Boer). It should be noted that these methods significantly underestimate the pore size for narrow mesopores. [14;168] By adding MC or DFT, these inaccuracies can be corrected, and a more reliable assessment of the pore size distribution over the entire domain is obtained. [14] With the pore sizes determined from physisorption measurements and the lattice parameters determined from X-ray analytical studies, the pore wall thickness of mesoporous silica materials can be asti- mated. [25;169;170] The standard method for characterizing macropores is mercury porosimetry. Mercury porosimetry can also be used to study small mesopores (dpore,DFT >3 nm). In addition, mercury porosimetry is used to obtain information on pore shape as well as network effects. [171;172] The main assumption of mercury porosimetry is a cylindrical pore shape of the porous materials. This assumption is used to describe the relationship between pressure difference at the curved mercury interface and the corresponding pore size using the surface tension of the mercury and the contact angle between solid and mercury. The relationship is known as the Washburn equation. If the pore shape of the material is not cylindrical, there may be large differences between the pore size determined from mercury porosimetry and the actual pore size. During the analysis process, mercury is pressed into the pores, which is why pressure is one of the most important measured variables. This is ex- plained by the fact that an incorrect pressure measurement automatically means an incorrect pore size. The capacitance measurement is almost always made between a metal shield on the outside of a measuring cell and the length of the mercury column inside, so the pore volume is another important parameter. This is because if the inside or the outside of the measuring cell is not uni- form or some other external factors affect the capacitance, this will lead to inaccurate values. [173] The unique feature of mercury porosimetry is the non-wetting characteristic. This means that the 28 1.4 Properties of Mesoporous Silica Materials contact angle between mercury and the sample surface is >90°. For this reason, the mercury must be forced into the pores with pressure. With increasing pressure, the mercury can penetrate pores with decreasing pore size. Accordingly, there is an inverse relationship between pressure and pore size, thus the pore size distribution can be determined. [4] Although the contact angle is a param- eter that significantly affects the results, a fixed value is usually chosen, regardless of the sample material. [173] As an alternative technique for determining pore size distributions, thermogravimetric analysis was described in the early 1990s. In this method, the porous materials are first loaded with an adsorbate. Then, thermogravimetric analysis is performed and the pore size is determined indi- rectly by mass loss. At the beginning of the measurement, the desorption of liquid outside the pores is detected. The second stage of mass loss corresponds to the capillary condensed liq- uid inside the pores together with the adsorbed film on the pore walls. This corresponds to the measurement of the total pore volume. The inflection point of the step in the desorption curve corresponds to the maximum of the pore size distribution. With the Kelvin equation, it is possible to convert the pore size distribution from sample mass loss as a function of temperature to the volume loss as a function of the pore radius. [174] Various publications discuss the optimization of the technique for determining the pore size distribution with respect to the conditions during the desorption process and the choice of adsorptive for different samples. [175–178] Benzenes [175;177], alcohols such as n-butanol [176;178], acetone, diethyl ketone [177] and aliphatic hydrocarbons such as n-heptane [177;178] were used as adsorptive. It was found that it made a difference whether the program was a heating program (constant heating rate) or a quasi-isothermal program (no constant heating rate). Depending on the adsorptive, the optimum desorption conditions must be selected. The optimum desorption conditions are determined by the heating rate where the shape of the distribution curve approximates those of the nitrogen physisorption measurement. Thereby, the changes in heating coefficients in the quasi-isothermal program only slightly affect the maximum of the distribution curves. [176;178] With the selection of the appropriate adsorbate, it is possible to determine the total pore volume as well as the pore size distributions from the thermogravimet- ric studies. [175;177] However, corrections must be made owing to surface film effects. The degree of interference depends on the choice of adsorptive. [177] Overall, the differences from the values determined from nitrogen physisorption measurements as well as from mercury porosimetry are large. [178] Another technique for determining the pore size distribution is nuclear magnetic resonance (NMR) spectroscopy. The 1H NMR measurement is proving particularly useful for in situ experiments. The reason is that characterization of porous materials with standard methods requires dry samples during the measurement. For low-field spin-lattice NMR relaxation measurements, dry samples are not required. [179] The relaxation time of water molecules is faster or slower depending on where they are located in the porous material. Enhanced interactions of water molecules at liquid- solid interface that disrupt molecular motion are one reason for the different behavior. Due to the externally applied high-frequency voltage, water molecules, in the pores relax faster than water 29 Chapter 1 Introduction molecules in the bulk phase. This leads to a decrease in the decay constant of the relaxation rate, which is directly related to the pore volume to surface area ratio. [180;181] An alternative to characterize pores by means of nitrogen physisorption measurements as well as X-ray analytical examinations is the measurement of the pore size by taking TEM images. Determination in this manner should be used with caution, as the measurement is error prone. Inaccuracies can occur in the measurement of pore size and wall thickness and during the acquisi- tion of the TEM images. This is because it cannot be guaranteed that the frontal view on the pore entrances of the porous silica particle is always available for measurement. A non-frontal view creates an incorrect determination. Similar to the determination of the pore size, the lattice param- eter and the pore wall thickness can be detected with the help of TEM images. The procedure is not or only little explained in the literature. [103;170;182] In addition to the methods already described for determining pore size, inverse size exclusion chromatography (ISEC) is another option. [183] This method has been used to characterize various materials to determine pore size and pore size distribution. [184–187] In ISEC, the material to be characterized is often packed as a monolith into a column, which is then treated with solutes of different particle sizes. [188;189] It is important that the mobile phase and solvent do not interact with the stationary phase, i.e., the material under study. If the solute is chosen correctly, the corre- sponding chemical interactions between the stationary and mobile phases are suppressed, so that exclusion is based on a physical sieving process. Information about the pore size as well as its dis- tribution can then be obtained from the resulting retention time. [189;190] ISEC is usually performed using standard monodisperse solutions. However, there are also studies showing that the same information about pore structure can be obtained with polydisperse suspensions. In this case, care must be taken to minimize dispersive mass transfer effects. [191] 1.4.3 Stability of Mesoporous Silica Materials Mesoporous silica materials modified with different functionalities can be used in heterogeneous catalysis. [192;193] For the study of confinement effects in catalytic applications, spatially controlled functionalization of mesoporous silica materials is necessary. At the same time, it is important that the developed catalysts can be used in larger reactors and synthesized on a larger scale. Therefore, the catalysts, which are often prepared as powder, must be formed, e.g., by tableting or extrusion. For this reason, mechanical stability of the catalysts is of interest. The mechanical stability of calcined SBA-15 as a function of pressures between 16 MPa and 260 MPa was investigated. [194;195] The X-ray diffractograms showed a loss of intensity of the characteristic reflections d100, d110, and d200 for the calcined and pressed SBA-15. The decrease of the intensities was attributed to loss of long-range order within the mesoporous silica mate- rial. The lattice parameters and the main pore size of the SBA-15 treated with pressures between 16 MPa and 260 MPa did not change compared to the unpressed sample. However, the pore size distribution widened with increasing pressure. The widening of the pore size distribution is ex- 30 1.4 Properties of Mesoporous Silica Materials plained by a deformation of the pores under pressure. The change in the pores reduced surface area and mesopore volume. [194;195] Literature has dealt with the difference of various silica ma- terials with regard to their mechanical stability. The findings regarding the mechanical stability in air of MCM-41 and MCM-48 are similar to those of SBA-15. When the investigations were done in nitrogen, the samples are more stable against pressure having removed physisorbed water molecules prior to mechanical testing. This results in fewer Si-O-Si bonds being hydrolyzed dur- ing compression. 29Si MAS NMR spectroscopic measurements show that the unpressed materials have Q2, Q3 and Q4 species, while the Q2 and Q3 species increase at the expanse of the Q4 for the samples pressed in air (compare Chapter 1.4.1). This effect is not visible for the samples pressed in nitrogen. [163;196] Investigations done by GALARNEAU ET AL. [197] suggest, that mechanical and thermal degrada- tion follow different mechanisms. A decrease in surface area and pore volume is visible by both influences, but a change in intergranular porosity is observed only under pressure. Considering thermal stability of mesoporous silica materials, pore wall thickness has an important influence onto their mechanical stability. Results from nitrogen physisorption measurements suggest that the thicker the pore walls the more stable the mesoporous silica material is against high tempera- tures. [197] Studies on the stability against hydrothermal treatment show that surface areas and pore volume decrease while the pore size distribution for MCM-41 broadened. X-ray analysis shows a deteri- oration of the structure after the hydrothermal treatment. The reason for these observations is the partial hydrolysis of Si-O-Si bonds within the silica matrix caused by the hydrothermal treatment. Owing to the thicker pore walls of SBA-15 compared to MCM-41, the structural loss of SBA-15 is lower as more Si-O-Si bonds need to be hydrolyzed to destroy the porous structure. The higher the temperature during the hydrothermal treatment, the greater the destruction of the mesoporous silica structure. [163] Since mesoporous silica materials are stable in aqueous solutions with a pH value of seven or less, the chemical stability studies are conducted in an alkaline environment. [198] For the chemi- cal stability studies, the mesoporous silica materials were treated with 0.5 mol sodium hydroxide solution. This treatment completely hydrolyzed the silica lattice of MCM-41 and SBA-15 within one hour. Not even the change from a hydrophilic to a hydrophobic surface makes the mesoporous silica material more stable. After two hours, the sample has dissolved to such an extent that the particles are too small for characterization. [163] It is shown that various mesoporous silica materials are stable to different influences. Structural properties such as pore wall thickness or surface properties play an important role and enable the stability of the materials to be influenced during production and subsequent modification. 31 Chapter 1 Introduction 1.5 Functionalization of Mesoporous Silica Materials In general, surfaces of mesoporous silica materials are functionalized in many cases to multi- ply their catalytic properties. Functionalization of porous materials describes the modification of surface properties and thus their activity. This makes materials suitable for a wider range of ap- plications, because functionalization with specific groups improves the range of properties of the porous materials without deteriorating the already existing positive properties such as porosity or large surface areas. [12;199;200] Porous materials with a large surface area are particularly suitable as support materials. In principle, it is possible to introduce active sites directly during the synthesis of the support materials or to add them in a further step after preparation. [17;201–203] X-ray ana- lytical studies show that the insertion and anchoring of active sites do not significantly affect the structure of the material. [12;201] However, functionalization of porous materials leads to a reduc- tion in surface area. [201] Porous silica materials like SBA-15 are particularly suitable as support material for surface modifications thanks to their large pore sizes and thick walls resulting in a high hydrothermal stability. [204] 1.5.1 Catalytically Active Species Depending on the field of application, the surface silanol groups can be active species. But com- pared to zeolites, silanol groups of mesoporous silica materials as active centers are no that defined due to the relatively thick pore walls. However, with the functionalization of mesoporous silica materials, active centers are generally introduced into the lattice structure or on the surface of a support material. In many cases, metals in form of ions, particles or oxides are introduced into a support materi- als as active centers. [17] By introducing heteroatoms such as aluminum, boron, fluorine, titanium, gallium, or oxide groups, silicon atoms are replaced in the lattice structure of the support mate- rial. [17;205] When this creates a charge imbalance, cations must be introduced into the pores for charge balancing. Replaced ions, which are introduced, act as acidic or redox active centers on the pore walls of the support material. The concentration of heteroelements in the structure is significantly lower, when they have been introduced into an as-synthesized support material. [17] Rather, different centers are present in various local environments and are more similar to metal- substituted amorphous silica materials. When the lattice atoms are replaced during the synthesis, relatively homogeneous incorporation and uniform distribution of heteroatoms in the support ma- terial are achieved. [17;205] Furthermore, it is possible to exchange heteroatoms after the synthesis of support materials, but mainly the pore walls are then functionalized. [17] Another method for the introduction of catalytic active centers is the deposition of nanoparticles on surfaces of support materials. The introduction of active species in form of nanoparticles is possible in various ways. On the one hand, nanoparticles can already be introduced into the gel during the synthesis. [146;206–211] On the other hand, it is possible to apply active species to support 32 1.5 Functionalization of Mesoporous Silica Materials materials by impregnation or vacuum evaporation. [146;209;211–214] The advantage of the deposition of nanoparticles is the large catalytically active surface area, which is introduced by the high sur- faces of nanoparticles. [215] A disadvantage is the risk of pore blocking if, e.g., agglomerates of the nanoparticles are formed due to their high loading. [17;205] A third way to modify porous materials is to anchor molecular catalysts on surfaces of a sup- port material. [17] The simplest method is to physically adsorb the molecular catalyst on the sur- face. [216;217] However, the interactions between the catalytically active component and the support material are relatively weak. Leaching of the active species in the presence of solvents is thus possible. [17] An alternative to the adsorption of catalyst complexes on surface-active groups of the support material is the attachment by forming chemical bonds. The introduction of functional groups bound to the surface of the silica material via a chemical bond is possible in various ways. One of them is the possibility to introduce functional groups as active centers into a lattice struc- ture already during the synthesis of the mesoporous silica material used as a support material by adding an organosilane. [204] The advantage of this method is to achieve a relatively homogeneous distribution of the functional groups on the surface. [17] Alternatively, anchoring the functional groups on the surface of the support material after synthesis can also be realized. [17] In this case, the functional groups are subsequently applied to the surface via covalent bonds. [204] Afterwards, catalyst complexes can be attached to the functional groups, which are then called anchor groups, applied to the surface of the support material, regardless of the type of modification. In this way, molecular catalysts are immobilized onto the surface of the support material. [218] Literature-known organosilanes serve as anchor groups. [200;219] 1.5.2 Introduction of Functional Groups The introduction of functional groups into the structure of mesoporous silica materials is possi- ble already during the preparation. [220] In this approach, referred to co-condensation approach or one-pot synthesis, hydrolysis and co-condensation of tetraalkoxysilanes with one or more organoalkoxysilane take place in a single reaction mixture in presence of the structure-directing template. Organic residues introduced in this way are mainly located on the silica surface or protrude from the walls into the pores. There are two reasons for this. Firstly, incompatibil- ity of the hydrophobic organic residues with the polar silica lattice results in a large portion of the organosilane being enclosed in the pore walls. Secondly, lower hydrolysis and condensation rate of the organosilane component in a basic environment compared to pure mesoporous silica materials plays an important role. This leads to a slower polycondensation process. [221] By inte- grating functional groups during the synthesis of mesoporous silica materials, they are relatively homogeneously distributed without affecting the pore size. [17;221] A major disadvant