ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2021

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 2018

Toward functional Metal-OrganicFrameworks: molecular doping,depth profiling, and surface growth

TIMOFEY LISEEV

ISSN 1651-6214ISBN 978-91-513-1142-5urn:nbn:se:uu:diva-435108

Dissertation presented at Uppsala University to be publicly examined in Polhemsalen,Ångströmlaboratoriet, Lägerhyddsvägen 1, Uppsala, Friday, 9 April 2021 at 09:15 for thedegree of Doctor of Philosophy. The examination will be conducted in English. Facultyexaminer: Professor Rob Ameloot (KU Leuven).

AbstractLiseev, T. 2021. Toward functional Metal-Organic Frameworks: molecular doping, depthprofiling, and surface growth. Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 2018. 91 pp. Uppsala: Acta UniversitatisUpsaliensis. ISBN 978-91-513-1142-5.

In this thesis, selected synthetic aspects of functional Metal-Organic Frameworks (MOFs) areconsidered. More specifically, the work is focused on MOFs doped with functional moleculesin the role of structural linkers. In the first part, preparation of such MOF/molecular catalysthybrid materials of two topologies UiO (UiO = Universitet i Oslo) and NU-1000 (NU =Northwestern University) – is investigated and three different synthetic pathways to thesematerials are studied. In the second part, a novel MOF depth profiling technique – RutherfordBackscattering Spectrometry – is introduced and then used to elucidate depth distribution ofmolecules post-synthetically introduced into a MOF. In the last part, efficient interfacing ofMOFs with electrodes is considered and a methodology for growing surface-mounted layer-pillar-type M2L2P MOFs (SURMOFs, M = metal, L = layer linker, P = pillar linker) with highdegree of orientation and thickness control on bare Si substrates is developed.

Keywords: MOF, water oxidation catalyst, RBS, PSE, depth profiling, SURMOF

Timofey Liseev, Department of Chemistry - Ångström, Synthetic Molecular Chemistry, 523,Uppsala University, SE-751 20 Uppsala, Sweden.

© Timofey Liseev 2021

ISSN 1651-6214ISBN 978-91-513-1142-5urn:nbn:se:uu:diva-435108 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-435108)

List of Papers

This thesis is based on the following papers and manuscripts, which are re-ferred to in the text by their Roman numerals. 1. Liseev, T., Howe, A., Hoque, M., Gimbert-Suriñach, C., Llobet, A., Ott, S.

(2020) Synthetic strategies to incorporate Ru-terpyridyl water oxidation catalysts into MOFs: direct synthesis vs. post-synthetic approach. Dalton Transactions, 49: 3753-13759 Author’s contribution: led the project, planned and conducted all experi-ments, synthesized the Ru complex, performed MOF syntheses and char-acterization, wrote and edited the manuscript.

2. Howe, A., Liseev, T., Gimbert-Suriñach, C., Llobet, A., Ott, S. (2021) Mixed linker MOF based on NU-1000 with an integrated ruthenium-based metallo-linker. Manuscript to-be submitted. Author’s contribution: was involved in planning of the project and the experiments, performed SEM analyses, contributed to manuscript writing and editing.

3. Liseev, T., Sortica, M., Paneta, V., McCarthy, B., Ott, S., Primetzhofer, D. (2021) Depth-profiling of intact metal-organic framework single crystals after post-synthetic modification. Manuscript submitted. Author’s contribution: led the project, performed all synthetic work, per-formed all analyses except RBS and non-bulk XRD, wrote and edited the manuscript.

4. McCarthy, B., Liseev, T., Beiler, A., Materna, K., Ott, S. (2019) Facile Orientational Control of M2L2P SURMOFs on 100 Silicon Substrates and Growth Mechanism Insights for Defective MOFs. ACS Applied Mate-rials & Interfaces, 11(41): 38294–38302 Author’s contribution: was involved in planning of the experiments, per-formed SEM analyses, contributed to manuscript writing and editing.

Reprints were made with permission from the respective publishers.

Contents

1. Introduction ................................................................................................. 9 1.1. Motivation ........................................................................................... 9 1.2. Approach ........................................................................................... 10 1.3. Scope of the thesis ............................................................................. 12 

2. Synthesis of functional molecule-MOF hybrids ....................................... 13 2.1. MOF synthesis: an overview ............................................................. 13 2.2. MOF defects ...................................................................................... 14 2.3. Incorporation of molecular species into MOFs ................................. 15 

2.3.1. Direct incorporation ................................................................... 15 2.3.2. Post-synthetic incorporation ...................................................... 15 2.3.3. Post-synthetic exchange and post-synthetic defect healing ....... 16 

2.4. Depth distribution of functional linkers in MOFs ............................. 17 

3. Analytical methods ................................................................................... 19 3.1. Powder X-Ray Diffraction (PXRD) .................................................. 19 3.2. Scanning Electron Microscopy (SEM) .............................................. 20 3.3. Brunauer-Emmett-Teller (BET) internal surface area measurements ........................................................................................... 21 3.4. Cyclic voltammetry (CV) .................................................................. 21 3.5. Rutherford Backscattering Spectrometry (RBS) ............................... 23 

3.5.1. History ....................................................................................... 23 3.5.2. Principle ..................................................................................... 24 3.5.3. Depth profiling using RBS ........................................................ 25 

3.6. Bulk analysis of digested MOFs ....................................................... 27 

4. Synthetic strategies for the incorporation of Ru-terpyridyl water oxidation catalysts into MOFs (Papers I and II) ........................................... 28 

4.1. Introduction ....................................................................................... 28 4.2. The Ru(tda)(Py)2 catalyst and derivative linkers ............................... 29 4.3. Synthesis of Ru(tda)(PyR)2 ............................................................... 31 4.4. Hybrid MOF/catalyst synthesis strategies ......................................... 32 4.4.1. UiO-type target topology (Paper I) ................................................ 34 

Direct solvothermal synthesis and co-synthesis .................................. 34 Indirect preparation by post-synthetic exchange ................................. 37 

4.4.2. NU-1000-type target topology (Paper II) ....................................... 42 4.5. Conclusions & outlook ...................................................................... 46 

5. Depth profiling of MOF crystals after post-synthetic modification using microbeam RBS (Paper III) .......................................................................... 48 

5.1. Introduction ....................................................................................... 48 5.2. Microbeam RBS method ................................................................... 48 5.3. Target MOF preparation .................................................................... 50 5.4. Post-synthetic exchange .................................................................... 50 5.5. Beam damage and probing mode ...................................................... 52 5.6. Results & Discussion ........................................................................ 53 

5.6.1. Depth distribution of ita and iba ............................................... 53 5.6.2. Framework defects and PSE yield ............................................. 56 

5.7. Conclusions & Outlook ..................................................................... 58 

6. Toward MOF electrodes: growing M2L2P SURMOFs on silicon with high degree of orientational control (Paper IV) ............................................ 59 

6.1. Introduction ....................................................................................... 59 6.2. Layer-by-layer SURMOF growth ..................................................... 60 6.3. Epitaxial growth of Cu2(bdc)2(dabco) ............................................... 60 

6.3.1. Growth on oxidized Si surface vs. quartz .................................. 61 6.3.2. On Si modified with anchoring SAMs ...................................... 62 6.3.3. On minimally treated Si ............................................................. 63 6.3.4. Method refinement: surface coverage and orientational control .................................................................................................. 65 6.3.5. Film thickness ............................................................................ 67 

6.4. Method extension: TiO2 substrate and isoreticular SURMOFs ......... 70 6.5. Conclusions & Outlook ..................................................................... 72 

7. Summary and Outlook .............................................................................. 73 

Acknowledgements ....................................................................................... 76 

Populärvetenskaplig sammanfattning på Svenska ........................................ 78 

Научно-популярное резюме по-русски ..................................................... 79 

References ..................................................................................................... 80 

Abbreviations

bdc benzenedicarboxylic acid bipy 4,4′-bipyridine BET Brunauer-Emmett-Teller CNT carbon nanotubes CV cyclic voltammetry/voltammogram dabco diazabicyclooctane DEF diethylformamide DMA dimethylacetamide DMF dimethylformamide DMSO dimethylsulfoxide edba 4,4'-ethynedibenzoic acid EDX energy-dispersive X-ray spectroscopy fa formic acid FWHM full width at half maximum GC glassy carbon electrode GIXRD grazing angle X-ray diffraction HER hydrogen evolution reaction iba iodobenzoic acid ICP-OES inductively coupled plasma-optical emission spectroscopy ina iso-nicotinic acid IPCC Intergovernmental Panel on Climate Change ita iodoterephthalic acid LbL layer-by-layer MOF metal-organic framework MWCNT multi-walled carbon nanotubes ndc 1,4-naphthalene dicarboxylate NDI naphthalene diimide NHE normal hydrogen electrode NMR nuclear magnetic resonance NU Northwestern University PSD post-synthetic deprotection PSDH post-synthetic defects healing

PSE post-synthetic exchange PSI post-synthetic insertion PSM post-synthetic metalation/modification PXRD powder X-ray diffraction Py pyridine PyR pyridine derivative RBS Rutherford Backscattering Spectrometry SALE solvent-assisted linker exchange SALI solvent-assisted ligand insertion SAM self-assembled monolayer SBU secondary building unit SEM scanning electron microscopy SLI sequential linker installation SURMOF surface-mounted metal-organic framework ta terephthalic acid H4TBAPy (1,3,6,8-pyrenetetrayl)tetrakis-benzoic acid tda 2,2’:6’,2’’-terpyridine-6,6’’-dicarboxylic acid UiO Universitet i Oslo UNECE United Nations Economic Commission for Europe WOC water oxidation catalyst XPS X-ray photoelectron spectroscopy XRD X-ray diffraction

9

1. Introduction

1.1. Motivation The steady, exponential development of the human civilization has brought about what is now colloquially known as Anthropocene – a geological epoch wherein all major geological and biological systems of the planet are severely affected by human activity. As early as 10,000 years BC, the rise of agriculture and the Neolithic Revolution entailed significant changes to biodiversity in human-populated regions.1 In the following centuries, ever-growing replace-ment of forests with farmlands, most often with the help of intentionally set fires, caused the first measurable human impact on the atmosphere.2 The in-vention of the first steam engine in the early 18th century then kicked off the Industrial Revolution, the effect of which on the atmosphere and the climate has been so profound that the Intergovernmental Panel on Climate Change (IPCC) has chosen the pre-industrial era (more accurately, year 1750) as the baseline when defining climate forcing.3 While there are other sides to anthro-pogenic impact on Earth, such as widespread radioactive contamination orig-inating from nuclear disasters and atomic weapons testing, by far the most significant impact to-date are atmospheric and hence climate changes origi-nating from burning of fossil fuels.4

For continuous existence, human society needs energy sources. While the gradual movement towards a Kardashev type I civilization5 may be inevitable, the crucial question is whether humanity will perish midway due to environ-mental collapse, or get creative about our energy sources in time to reverse the current “Hothouse Earth” trajectory.6 To tackle this problem, a variety of al-ternative energy sources are being researched and implemented, including wind, solar, geothermal, hydroelectric, and biomass-derived energy carriers.7 The chemistry community’s approach has been to develop catalysts for trans-formation of clean and renewable natural substrates (e.g., water and CO2) into fuels or useful chemicals with the help of sustainable energy sources such as sunlight – in essence, striving to develop artificial photosynthetic systems. The approach is underlined by the projected energy use scenarios published by United Nations Economic Commission for Europe (UNECE),8 which show continuing reliance on burning fuels until the end of the century. This stresses the importance of active CO2 capture to mitigate projected continued fossil fuel consumption.

10

Answering this call, many highly efficient catalysts for artificial photosyn-thesis have been and are being developed. These fall roughly into two catego-ries: material catalysts and molecular catalysts. The latter hold promise for exceptional activity per metal center9,10 while their modular character allows well-defined modifications to control product selectivity.11,12 Yet, molecular catalysts are still not being broadly implemented for producing hydrogen from water, or fuels and chemicals from carbon dioxide. At the moment, molecular catalysts are not suitable for large-scale, continuous industrial application: while extremely active, they tend to suffer from low stability combined with high production cost (due to precious metals and/or complicated ligands in their structure). To overcome this issue, a concerted effort is being made to develop cheaper molecular catalysts based on abundant metals, but there is another approach: combining molecular catalysts with bulk materials for im-proved stability and recyclability. This is the conceptual vision behind this thesis.

1.2. Approach By combining the high potency of a molecular catalyst with the bulk stability of a solid material, catalytic performance can benefit from both worlds, and the resulting heterogenization also facilitates handling and recycling – features appreciated by industry. The next question is: which bulk material to choose as the solid support for molecular catalysts?

Our group’s answer to this question is Metal-Organic Frameworks (MOFs), a class of porous crystalline materials praised for their versatility, high internal surface area and chemical tunability. MOFs are composed of two structural components: organic linkers acting as “struts,” and inorganic Secondary Building Units (SBUs) acting as framework “nodes” (Figure 1). Each linker, possessing ligating groups at opposite ends, connects two or more SBUs in an ordered and repeating fashion, giving rise to a porous crystalline lattice. A particular class of hydrolytically stable MOFs that is often targeted for catalyst incorporation are “Zr-MOFs”, featuring zirconium oxide SBUs and carbox-ylate linkers.13

11

Figure 1. Schematic representation of a MOF constructed from organic linkers and secondary building units (SBUs).14

The ability to choose chemical nature, size, and topology of both the linker and the SBU means almost unlimited versatility of the MOFs as a class of materials. For our cause, this means the scaffold can be uniquely tailored to host a particular molecular catalyst, together with the possibility of fine-tuning its environment. Unmatched internal surface area of MOFs promises high po-tential loading of the catalyst while also providing channels for transporting substrates and products to and from the active sites. In fact, the MOF-molec-ular catalyst composites are sometimes referred to as biomimetic systems, since they essentially emulate an enzyme’s architecture: an active site embed-ded in, protected, and modulated by an external scaffold.15–17 Indeed, it has been demonstrated that catalyst stability and efficiency can be improved by incorporating it into a MOF,18,19 and its substrate selectivity can be modulated by fine-tuning such incorporation.20

Naturally, these benefits come at a cost. Synthesis of MOF/functional mol-ecule hybrid materials is a challenge that sometimes needs to be solved in a case-by-case fashion. Should the molecule be introduced into a pre-made MOF, or integrated during the MOF synthesis? If it is inserted post-syntheti-cally, does the molecule penetrate throughout the MOF crystal, or is restricted to the outer layers only? How can this be determined? For practical application of MOF-catalyst hybrids, another set of challenges arises. MOF particle size needs to be optimized in order to balance the diffusion of substrate and product with turnover kinetics.21 If the MOF is intended to be electrocatalytic, an effi-cient MOF-electrode interface has to be developed, and a separate issue of low intrinsic electrical conductivity of MOFs has to be solved.22,23

12

1.3. Scope of the thesis Addressing all of the issues listed above is beyond the scope of a single PhD thesis. Yet, tangible progress was made herein towards solving some of these issues. Synthetic strategies for the incorporation of a challenging Ru(tda)(Py)2 (tda = 2,2’:6’,2’’-terpyridine-6,6’’-dicarboxylic acid) class of water oxidation catalysts into MOFs were studied and developed (Papers I and II). A novel analytical method for accurate and non-destructive MOF depth profiling was developed and then tested by determining spatial distribution of model mole-cules post-synthetically introduced into MOF crystals (Paper III). Finally, as a more practically oriented step toward MOF electrodes, we developed a set of methods for the growth of M2L2P surface MOFs (SURMOFs, M = metal, L = linker, P = pillared linker) films with high degree of orientational control directly on Si substrate (Paper IV).

13

2. Synthesis of functional molecule-MOF hybrids

This section will consider general synthetic aspects of MOF synthesis, focus-ing on the preparation of hybrid MOF materials that carry additional function-ality.

2.1. MOF synthesis: an overview A typical MOF synthesis involves three chemical components: the linker, the SBU or its precursor, and the solvent. In this process, the solvent (usually an amide) is an active participant: its gradual thermal decomposition produces amine species which deprotonate the linkers, thereby “activating” them for SBU coordination.24 The deprotonated linkers then bind to the SBUs (metals or metal clusters) and, ideally, form a crystalline framework. In reality, how-ever, the process as described tends to produce materials which are often highly defective and non-crystalline, especially in the case of Zr-based MOFs.25,26

In order to improve crystallinity and to have a degree of particle size con-trol, a modulator is used during the synthesis. Typically a smaller, monotopic molecule with lower pKa compared to the linker, the modulator slows down crystal growth by competing with the linker for SBU ligation.27–29 The for-mation of the node-ligand bond is an equilibrium-controlled reaction; by add-ing the modulator, another equilibrium is introduced, which, however, does not lead to structural expansion. This allows in situ correction of emerging erroneous structural motifs and so improves overall crystallinity. There are examples of high quality Zr-MOFs synthesized in the absence of organic mod-ulators, but still requiring HCl as a modulator surrogate.30–32 A completely modulator-free MOF synthesis using very low oversaturation conditions has been reported, but this approach remains an exception.33

14

2.2. MOF defects As outlined above, the use of modulators in MOF synthesis has clear benefits. However, interfering with node coordination, the modulator may also contrib-ute to the appearance of framework defects. During crystal growth, modula-tors may remain attached to the SBUs in places where linkers should be, thus creating defective, vacant sites. Alternatively, an entire SBU together with ad-jacent linkers might be missing from the structure. These scenarios describe the two types of MOF defects: missing-linker and missing-cluster (Figure 2).

Figure 2. Structural representation of framework defects on the example of UiO-67 MOF (UiO = Universitet i Oslo): (A) a missing-linker defect compensated for by two formate modulators; (B) a missing-cluster defect compensated for by acetate modulator ligands at adjacent SBUs.34

Historically, missing linkers replaced by pairs of modulators or solvent mole-cules were thought to be the predominant defect type in Zr-MOFs.35,36 Later, missing-cluster defects were identified in UiO-66, and even proposed as the major vacancy type.37 Eventually, both types of defects were found to coexist in UiO-66.38 An entire sub-field in MOF research is emerging that is dedicated to functional defect engineering for modulating MOF properties such as sta-bility,31 porosity,37 absorptivity,39 reactivity,34,40 and catalysis.41 In this thesis, framework defects are an important part of every chapter’s discourse.

15

2.3. Incorporation of molecular species into MOFs In order to create a resilient hybrid material, molecules that carry out a partic-ular function should be attached to the MOF either covalently or through strong coordination bonds. The alternative – simply trapping functional mol-ecules within MOFs’ pores in a “ship-in-a-bottle” approach – can result in gradual leaching and loss of functionality. The incorporation of functional molecules into MOFs fall roughly into two categories: direct and post-syn-thetic modification strategies.

2.3.1. Direct incorporation In the direct approach, functional molecules are introduced already at the stage of the MOF synthesis – acting as a sole linker, a co-linker, or sometimes as modulator. Using the functional molecule as the only linker, an obvious ben-efit of the direct solvothermal approach is that the highest possible loading is achieved. In case where the functional linker is accompanied by another linker in the solvothermal synthesis, the method most likely results in a statistical distribution of the functional moiety throughout the crystal, thereby avoiding for example core-shell type structures. Additionally, the direct approach pro-vides the product directly, without the need for subjecting the MOF to any further treatments. Conversely, conventional MOF syntheses normally in-clude prolonged incubation of the reaction mixture at high temperatures (often > 100 °C) – conditions that might exclude this approach for temperature-sen-sitive molecular species, such as molecular catalysts.18

2.3.2. Post-synthetic incorporation In the post-synthetic approach, the new functionality is introduced into a pre-made MOF, thus avoiding the harsh solvothermal conditions described above. The tool box of post-synthetic methods has been expanded considerably over the last decade, and many different techniques are now available (Figure 3).42,43 Post-Synthetic Insertion (PSI), also termed Sequential Linker Installa-tion (SLI), describes the incorporation of new linkers at unsaturated SBU sites in addition to existing linkers, changing the topology of the MOF in the pro-cess.44 Post-Synthetic Deprotection (PSD), as the name suggests, removes protecting groups from linker-borne functional moieties incorporated during the MOF synthesis stage.45 In Post-Synthetic Metalation (PSM), linker-borne functional metal complexes are assembled inside the MOF through ligation to specially designed linkers.46 In a way, PSD and PSM are hybrids of direct and post-synthetic methods, since they rely on incorporation of functional precur-sors already during the MOF synthesis.

16

Figure 3. Schematic representation of major MOF post-synthetic modification meth-ods: Post-Synthetic Installation (PSI), Deprotection (PSD), Metalation (PSM), Ex-change (PSE) and Defect Healing (PSDH).

2.3.3. Post-synthetic exchange and post-synthetic defect healing Perhaps the most widely known and used post-synthetic method is Post-Syn-thetic Exchange (PSE),47 also known as Solvent-Assisted Linker Exchange (SALE).48 In this method, the pre-made MOF is incubated in a solution of the molecule of interest possessing the same molecular size and configuration of ligating groups as the native MOF linker. This process results in the substitu-tion of the native linker by the functional one. Conceptually, in PSE, every newly introduced linker should replace one parent linker. In practice, how-ever, this “pure” mechanism tends to occur only in completely defect-free MOFs – at least in the UiO MOF family.40 In reality, common UiO MOFs feature missing-linker defects, and incoming linkers preferably substitute SBU-capping modulators, resulting in Post-Synthetic Defect Healing (PSDH).34,40,49 Only after the complete substitution of modulators is achieved does the “pure” PSE mechanism begin to take place to a significant extent.40,50 It should be noted that, unlike PSI, both PSE and PSDH proceed without al-tering the topology of the parent framework.

The solvent used in PSE plays a rather complex role that is not only limited to solvation and transport of the exchanging species through the pores. Tran-siently coordinating to the nodes, the solvent promotes linker replacement by facilitating the formation of “dangling”, i.e. partially detached linkers.30 Es-sentially, the role of the solvent is in creating temporary framework defects, which are more prone for PSE. Consequently, the nature of the solvent has a great effect on PSE outcome. In UiO-66 derivatives, the exchange ratio was found to correlate strongly with the solvent polarity and changed in the fol-lowing fashion: H2O>DMF>MeOH>CHCl3.43,47 More recently, a set of 30 commercially available solvents and some of their mixtures have been screened for their ability to modulate kinetics of PSE and linker diffusion.51

17

In Chapter 4 of this thesis, we investigate direct vs PSE approaches for the incorporation of a highly potent water oxidation catalyst, Ru(tda)(Py)2 (tda = [2,2′:6′,2″-terpyridine]-6,6″-dicarboxylate, Py = pyridine), into a MOF.

2.4. Depth distribution of functional linkers in MOFs When talking about MOF/functional molecule hybrids, an important question to consider is spatial distribution of the inserted species within the MOF crys-tal. Depending on how deep within the crystal the functional molecules re-sides, their interaction with the environment can change. Conceptually, two limiting situations can be imagined: the uniform distribution of the functional linker in the crystal, or the core-shell distribution, with intermediary architec-tures in between featuring concentration gradients. More sophisticated multi-layered “Matryoshka” structures are also possible.52 In general, linker-borne functionalities can be exposed near the crystal surface to facilitate interaction with the environment or, conversely, buried in the core of the crystal for better shielding. For example, by confining catalytic units to the core of the MOF, substrate selectivity can be modulated due to the size-exclusion provided by the outer shell.20

Depth distribution of functional molecules inside the MOF is thus an addi-tional degree of freedom in functional MOF engineering. To utilize its poten-tial, one needs to control distribution type and, if the core-shell architecture arises, shell parameters such as thickness and density. If the molecule is intro-duced using PSE (and not, for example, sequential growth),52 variables such as crystal size, PSE duration, solvent, linker nature, etc., dictate the eventual outcome – on a phenomenological level.53,54 On a more fundamental level, the shell formation process is governed by the ratio of two kinetic parameters: the rate of diffusion of the exchanging linker into the MOF and the rate of linker exchange, within relevant timeframes.51,54 If diffusion is fast compared to the rate of linker exchange, the incoming species may traverse and bind through-out the MOF crystal in a uniform fashion. If, on the other hand, diffusion is sluggish and the linker exchange is relatively fast, the incoming linker initially forms a “shell” in the outer regions of the crystal.

To understand and learn to control the spatial arrangement of molecules inside the MOF, a robust analytical technique is required for MOF depth pro-filing. Sometimes, the crystals in question are large and possess flat, platelet morphology, which enables direct microspectroscopic examination of dis-tinctly coloured or fluorescent layers.52 When a photoactive linker is involved, even non-flat crystals have been depth-profiled using confocal emission mi-croscopy, however, a relatively large crystal size is still required, and the ac-curacy is limited.55 Finally, the most common MOF depth profiling method in literature involves growing large (>100 µm) MOF crystals and embedding

18

them in epoxy resin to be physically sectioned (sliced), followed by micro-spectroscopic examination of the slices.20,54,56 While all of the abovemen-tioned methods can be useful, the microspectroscopic approach is still limited in scope (requiring large crystals and photoactive linkers), sensitivity and depth resolution. In Chapter 5, we introduce a novel method for MOF depth profiling based on Rutherford Backscattering Spectrometry (RBS), featuring small (~10 µm) operable crystal size and unprecedented sensitivity and reso-lution, while being non-destructive in nature. We demonstrate the power of this method by investigating the depth distribution of two ligand molecules that have been introduced into UiO-66 post-synthetically.

19

3. Analytical methods

This section will give a general overview of analytical techniques used in this thesis with particular emphasis on Rutherford Backscattering Spectrometry (RBS).

3.1. Powder X-Ray Diffraction (PXRD) Conventionally, when a fresh MOF batch is prepared, the first quality test is an examination of its PXRD pattern. This technique allows for a quick confir-mation of the MOF structure by matching the measured reflections with ex-pected ones obtained from literature or simulations. Strong, sharp reflections normally indicate a highly crystalline, high quality material, while weak, broadened peaks point at small crystallite sizes or a degree of amorphousness. In fact, this property is utilized to estimate average crystallite sizes by the Scherrer equation, using the full width at half maximum (FWHM) of a specific XRD peak.57,58 In case of highly oriented surface-mounted MOF (SURMOF) films, this analysis may directly yield the film thickness. If amorphous phases are present in the sample, they tend to appear as a “halo” at low angles (Figure 4).59 A missing peak compared to reference patterns may indicate an ordered array of defects, but can also originate from non-random orientation of the crystals, such as often the case with SURMOFs. Randomly distributed defects cannot be detected by this method.

20

Figure 4. Illustration of the “halo” phenomenon indicative of the presence of amor-phous phases.59,60 Reprinted with permission from John Wiley and Sons, Inc.

3.2. Scanning Electron Microscopy (SEM) For direct observation of MOF crystals, scanning electron microscopy (SEM) is a classical method. In SEM, the sharpness of the image correlates with the accelerating voltage with higher voltages yielding better-quality images. However, as a class of materials that is partly organic, MOFs are easily dam-aged by high electron currents. As a compromise, 3-5 kV is an acceptable voltage range that is tolerated (at least for a limited time) by most MOFs. For additional protection, as well as to avoid particle charging, sputter-coating of the MOF sample with a thin layer of highly conductive alloy (e.g., Pd/Au) is often used. The charging phenomenon (often observed with large MOF parti-cles) occurs due to the inherently insulating nature of the MOF and its inability to dissipate probing electrons efficiently. As a result, accumulated electrons appear as featureless lit-up areas and may distort the image around them, or even lead to matter ejection (Figure 5).

21

Figure 5. UiO-66 crystals imaged by SEM before (left, showing the charging phe-nomenon) and after (right) sputter-coating.

3.3. Brunauer-Emmett-Teller (BET) internal surface area measurements Nanoscale internal porosity is one of the key features of metal-organic frame-works. A family of analytical methods based on gas adsorption/desorption is used to assess various porosity parameters. The most common method and the one used in this thesis is the BET technique, which is used to quantify total internal surface area of the adsorbent, in this case a MOF. Typical surface areas for MOFs tend to range between 200 and 3000 m2/g;61 and the compar-ison of a measured value to the reported surface area of the same MOF may yield insights about framework anomalies. A moderate amount of bulk defects is known to result in an increase of the internal surface area (likely due to decreased material density),41 while even higher defect concentrations beyond a certain threshold (e.g., 15% for UiO-66) cause the internal surface area to drop.39 Pore collapse (either full or partial) is always accompanied by a dra-matic decrease in the internal surface area.59

3.4. Cyclic voltammetry (CV) Electrochemical measurements are crucial for the characterisation of redox-active MOFs. For such electro-active MOF/functional molecule hybrids, elec-trochemical accessibility, MOF and linker stability, and the performance of the active molecular species can be evaluated. The electrochemical analysis of redox-active MOFs has recently been reviewed in-depth by our group,59 so this section will focus on some practical aspects relevant to this thesis.

Cyclic voltammetry (CV) is one the most common electrochemical meth-ods in use. In essence, this technique produces and measures redox events in

22

substances of interest using the working electrode as electron donor or accep-tor instead of a chemical redox partner.62 In a typical 3-electrode setup, the potential (U) between the working and the reference electrodes is cyclically scanned within a limited range, while the induced current (I) is recorded be-tween the working and the counter electrodes. Plotting I against U results in a cyclic voltammogram, where occurring redox events are registered as “waves” when the scanning bias crosses the reduction potential of a redox couple (Fig-ure 6).

Figure 6. Example cyclic voltammograms of an NDI-derived linker (NDI = naphtha-lene diimide) and a corresponding Zr-based MOF (black). Both graphs display two distinct redox events pertaining to NDI.

Comparing electrochemical responses of a redox-active linker in solution and inside a MOF (Figure 6) is an effective diagnostic of the linker’s integrity and electronic environment.59 The redox signature observed in homogeneous phase should be qualitatively retained upon successful incorporation of the linker into a MOF. The appearance of new waves, missing waves or the loss of reversibility reveals distortions in electrochemical landscape resulting from, e.g., side reactions, linker decomposition, or irregular MOF structure. On the other hand, the absence of discernible signals from a MOF loaded with redox linkers indicates their electrochemical inaccessibility, which may be due to blocked charge hopping or a poor MOF/electrode interface (vide infra).

Establishing a robust and efficient interface between the MOF and the working electrode is one of the major challenges in MOF electrochemistry. Due to low intrinsic conductivity of MOFs, the quality of electrochemical data directly depends on the area of electrical contact between the electrode and

23

the MOF particles. Another issue is the stability of the interface, in other words, its resilience during electrochemical measurements against, for exam-ple, delamination. Two general approaches for interfacing MOFs exist: attach-ing the MOF powder to the electrode surface, or growing the MOF directly on the electrode. If a pre-made (i.e., powder) MOF is to be measured, it can be immobilized on the electrode using either electrophoretic deposition, or, more commonly, dropcasted in the form of a suspension.59 To increase the electrical contact area, inert conductive particles (e.g., mesoporous carbon or carbon nanotubes, CNT) are mixed into the suspension, and often a binder (e.g., Nafion) is added to prevent delamination of the resulting film.59 This approach is discussed in Chapter 4. On the other hand, growing the MOF directly on the electrode surface in principle can provide the best possible electrical contact – this approach is the focus of Chapter 6.

3.5. Rutherford Backscattering Spectrometry (RBS) 3.5.1. History At the end of the 19th century, the dominant, although disputed, model of the atom was the “plum pudding” model, which envisioned the atom as a posi-tively charged sphere with dot-like electrons distributed throughout it. To test an alternative model proposed by Rutherford, Hans Geiger and Ernest Marsden devised an experiment where α-particles were bombarding a thin gold foil surrounded by a fluorescent screen (Figure 7).63,64 If the “plum pud-ding” model was correct, all α-particles would pass through the foil and hit the screen behind it. In reality, the physicists observed the particles scattering in all directions, including back towards the source (backscattering). This led to the discovery of the atomic nucleus and the adoption of Rutherford’s plan-etary model of the atom. Half a century later, the potential to apply this type of experiment in analytical chemistry was first elaborated.65

Figure 7. Schematic representation of the Geiger-Marsden experiment.

24

3.5.2. Principle The basic principle used in Rutherford Backscattering Spectrometry is the fact that nuclei of different elements backscatter the probing ions differently. In a typical setup, the target is bombarded by a stream of accelerated ions (typi-cally α-particles or protons), some of which recoil back toward the energy detector located at a certain angle. Judging by the energy of these backscat-tered probe ions, the identity of the nuclei they recoiled from can be estab-lished (Figure 8). Under the conditions of elastic scattering (under which RBS operates), the colliding nuclei of the probing beam and the target are treated as solid balls with perfect conservation of momentum upon collision. There-fore, knowing the mass and velocity of probing ions before and after the col-lision (the kinematic factor), the mass of the scattering partner (a nucleus in the target) can be calculated. The incoming backscattered ions are counted, providing quantitative in addition to qualitative analysis of the atomic compo-sition. To be detectable, target nuclei should be heavier than the probing ions in order for recoil to occur, which makes hydrogen undetectable by this method.

25

Figure 8. Basic operation of the RBS method (recoil from the surface of the target). The target is bombarded by a probe beam of ions of approximately the same kinetic energy. After the recoil from different nuclei (depicted are Zr and I), probing ions with different amounts of residual kinetic energy reach the detector. The detector measures the energy of each incoming ion, which is then plotted against the number of collisions (inset RBS spectrum).

3.5.3. Depth profiling using RBS The feature that makes RBS particularly interesting to material chemists is its depth perception capability. While recoiling can happen on the surface of the target, probing ions may also penetrate deeper into the bulk and backscatter from nuclei at various depths. Traveling through the matter to and from the collision site, the ions interact with electron density and lose kinetic energy in a continuous fashion (stopping power). As a result, upon striking the detector, these ions possess lower kinetic energy compared to those that are scattered from the same element on the surface of the target. The amount of energy loss is proportional to the depth at which the collision occurred. Experimentally, this appears as a shift of the characteristic element signal toward lower ener-

26

gies (Figure 9). If the element is found throughout the target material, its cor-responding RBS signal forms a continuous “plateau” with the width propor-tional to the thickness of the target. If, on the other hand, the element is re-stricted to the surface, the corresponding signal will be narrow. The height of the signal is proportional to the concentration distribution.

Figure 9. Illustration of depth profiling by RBS

Another valuable feature of the RBS method is its non-destructive nature. Un-der properly tuned conditions, the target integrity is perfectly conserved dur-ing the measurement, which is why RBS is a popular technique for the analysis of works of art and historical artefacts.66–68

Combined, these features make RBS a nearly perfect method for studying MOFs: it can elucidate the constituting elements the MOF, their relative amounts and how deep in the crystal they reside. No MOF pre-treatment such as embedding in glue and slicing is required, and the material is not destroyed in the measurement. We see RBS as a sadly underrepresented method in the MOF community and seek to demonstrate its usefulness in Chapter 5 with the help of a particle accelerator at the Ångström Tandem Laboratory.

27

3.6. Bulk analysis of digested MOFs The classic and most straightforward way for compositional analysis of bulk MOF powders is to digest the material and to analyse the resulting solution. However, MOFs contain both organic and inorganic building blocks, which means that more than one method is usually needed for the full MOF compo-sition analysis. For measuring the organic part (e.g., the linkers), the simplest and most common method is NMR (Nuclear Magnetic Resonance) spectros-copy. For the inorganic part, a number of options are available, and the one we tend to employ is ICP-OES (Inductively Coupled Plasma – Optical Emis-sion Spectrometry), chosen for simplicity and high sensitivity to zirconium, which is a prime component of the SBU (Zr6O4(OH)4) in the UiO family of MOFs. Thereby, NMR and ICP-OES form a mutually complementary pair of methods.

Before the measurements, the material should be fully digested. The diges-tion conditions depend on the MOF type and sometimes need to be optimised to reach full dissolution. For example, UiO-66 MOF is digested by simply suspending the powder in DMSO (DMSO-d6 for 1HNMR) and adding small amount of concentrated HF. Other MOFs may require harsher conditions, such as high-temperature incubation in piranha solution or aqua regia. For a typical 1H-NMR measurement, around 5 mg of MOF is enough to produce a good sample signal. For a 13CNMR measurement, naturally, the amount of MOF needed is much higher.

ICP-OES is a spectroscopic method based on ionization of the injected sample by superheated plasma, normally generated from argon and sustained by radio-frequency solenoids.69 Interacting with the plasma at temperatures around 10000 K, sample molecules and clusters break down into ions that emit light of element-characteristic wavelengths. The method is extremely sensi-tive, and only minute amounts of the sample solution are used to perform the measurement. In practice, as little MOF powder as can be weighted with de-cent accuracy is sufficient.

One practical pitfall to be mindful of while measuring digested MOFs by any method is the quality of the digested powder. Often, especially in new, non-optimized MOF syntheses, the product MOF is contaminated by various amounts of inorganic chaff, mostly oxides of the metals that are used in the synthesis (e.g., various forms of ZrO2). If the analysis relies on absolute quan-tification of organic and inorganic parts of the MOF, this chaff, once digested, contributes to overestimation of the inorganic part and consequent underesti-mation of the organic part. Therefore, the amount of the chaff should be checked beforehand (e.g., by SEM), and, if need be, removed (for example, using centrifugal sequential precipitation).

28

4. Synthetic strategies for the incorporation of Ru-terpyridyl water oxidation catalysts into MOFs (Papers I and II)

In this chapter, different routes for the synthesis of hybrid MOF/molecular catalyst materials are investigated in which a molecular catalyst is incorpo-rated as a structural linker. Two metal complexes of the general formula Ru(tda)(PyR)2 (tda = 2,2’:6’,2’’-terpyridine-6,6’’-dicarboxylic acid; PyR = derivatives of pyridine) – a class of highly potent water oxidation catalysts (WOCs) – were used as model compounds in the study.

4.1. Introduction In context of global ecological and energy challenges, catalysts relevant to artificial photosynthesis have become an important part of MOF re-search.19,70,71 A key goal for the chemistry community is to achieve efficient (photo)electrocatalytic water splitting by combining two half-reactions: water reduction to produce hydrogen and water oxidation to produce oxygen.72 Since the two parts are to be coupled by the same stream of electrons, the overall kinetics will be limited by the slowest reaction. Water oxidation involves the transport of four electrons per oxygen molecule released, making it much more kinetically challenging to catalyze compared to the hydrogen evolution reaction (HER). In recent years, improving the performance of molecular WOCs by incorporation into MOFs has seen increased activity.16,73,82–87,74–81

As mentioned in the Introduction, the incorporation of molecular catalysts into MOFs improves catalyst stability, while further fine-tuning of the cata-lytic properties is possible through pore interior design.16,18,84 It is desirable to incorporate the molecular catalyst as the linker, an integral part of the frame-work, not only to secure its attachment to the MOF, but also to minimize clut-tering of the internal pore space, which is needed for the transport of ions and substrates/products.88 To ensure hydrolytic stability, we chose the family of MOFs based on a Zr6O4(OH)4 SBU and carboxylate linkers. With a ditopic linker, this combination results in a UiO-type MOF (UiO = Universitet i Oslo). A tetratopic linker produces, for example, NU-1000 MOF topology (NU =

29

Northwestern University). Both have been investigated as targets for the in-corporation of the Ru(tda)(PyR)2 catalysts: UiO type in Paper I and NU-1000 type in Paper II.

4.2. The Ru(tda)(Py)2 catalyst and derivative linkers A ruthenium-based metal complex featuring an equatorial terpyridyl dicar-boxylate (tda) ligand and two axial pyridine (Py) ligands (Figure 10A) was selected for these studies. In fact, this complex is a precursor to the actual catalyst that is formed in situ during the electrochemical “activation” stage, i.e., when the ruthenium core coordinates a water molecule to yield the active catalyst. Once activated, the complex is counted among the most potent mo-lecular water oxidation catalysts to-date.9 Its catalytic cycle, driven by step-wise oxidation of the Ru center from +II to +V, involves multiple coordination rearrangements in and around the tda ligand, while the axial pyridine ligands stay permanently coordinated (Figure 11).89

Figure 10. (A) The original Ru(tda)(Py)2 catalyst; (B) the ditopic Ru(tda)(iso-nico-tinic acid)2 modification used for integration into a UiO MOF; and (C) the tetratopic Ru(tda)(Py(PhCOOH)2)2 complex used for making a NU-1000 MOF.

30

Figure 11. Proposed catalytic cycle of water oxidation by Ru(tda)(Py)2. Reprinted with permission from ref.89 Copyright 2015 American Chemical Society.

In order to be incorporated into the MOF as a structural linker, the Ru complex must be modified to include SBU-ligating groups. In view of the mechanism briefly described above, synthetic manipulations of the tda ligand would likely disrupt the catalytic performance, so modification of the strongly-bound axial pyridine ligands was targeted instead. For incorporation into a UiO-type MOF, a ditopic analogue with iso-nicotinic acid (ina) in the axial positions was syn-thesized (Figure 10B). To create a NU-1000-type linker, a tetratopic version with Py(PhCOOH)2 ligands at the axial sites was prepared (Figure 10C). Both syntheses are described in more detail in the next section.

It is worth noting that the Ru(tda)(PyR)2 complexes in Figure 10B and 10C represent a type of linker that is particularly challenging to integrate into a MOF for the following reasons: Unlike conventional linkers, the dihedral angle between axial carboxylic

groups in Ru(tda)(PyR)2 is not fixed due to free rotation of the pyridyl ligands around the main axis of the complex. This may hinder MOF growth by obstructing pre-organisation of the anchoring groups.

In the (default) low oxidation states, the ruthenium centre is six-coordi-nate, leaving one of the equatorial carboxylic acid groups uncoordinated. This dangling carboxylate may potentially compete for SBU ligation and thereby disrupt ordered MOF formation.

The equatorial tda ligand is sterically demanding. Taking into considera-tion the fact that every (non-defective) SBU in a UiO-type MOF is coor-dinated by 12 linkers, this may introduce a significant steric stress that leads to highly defective or disordered materials.

31

4.3. Synthesis of Ru(tda)(PyR)2 The tda ligand was synthesized according to the literature procedure by oxi-dation of dimethylterpyridine, which, in turn, was obtained from 2-bromo-6-methylpyridine and 2,6-dibromopyridine by a Stille coupling (Figure 12).89,90

Figure 12. Synthesis of the tda ligand.

The Py(PhCOOH)2 ligand was synthesized by a Suzuki cross coupling be-tween 3,5-dibromopyridine and (4-(methoxycarbonyl)phenyl)boronic acid followed by deprotection of the resulting methyl ester (Figure 13).

Figure 13. Synthesis of the Py(PhCOOH)2 ligand.

The precursor complex Ru(tda)(dmso)(H2O) was then prepared from RuCl2(dmso)4 by coordinating the tda ligand concomitant with exchange of one coordinated dmso molecule for water (Figure 14). From the precursor, the two variants of the catalytic complex were then prepared by coordinating ei-ther the monotopic ina or the ditopic Py(PhCOOH)2 at the axial positions (Fig-ure 14). The purity of the final products was confirmed using NMR spectros-copy and cyclic voltammetry (vide infra).

32

Figure 14. Synthesis of the precursor complex Ru(tda)(dmso)(H2O) and the linker derivatives Ru(tda)(ina)2 and Ru(tda)(Py(PhCOOH)2)2.

4.4. Hybrid MOF/catalyst synthesis strategies Conceptually, we envisioned three different strategies for the making of MOF/Ru(tda)(PyR)2 hybrids that include the catalytic complex derivatives as a linker (Figure 15): 1. Direct solvothermal synthesis using Ru(tda)(PyR)2 as the sole linker. This

is the most straightforward approach that, in theory, should result in a MOF with the highest possible catalyst loading. As a consequence, the distance between the linkers will be minimized, and so facilitate redox hopping charge transfer.59,82

2. Direct solvothermal synthesis with an inert co-linker. In this approach, lower catalyst loading is traded for spacing out the bulky catalytic linkers with inert and sterically less demanding co-linkers.

3. Post-synthetic incorporation of the Ru(tda)(PyR)2 complex into a pre-made parent MOF. While more steps are involved, this approach avoids subjecting the catalytic complex to the harsh MOF synthesis conditions, which may be critical, for example, in case of temperature-sensitive cata-lysts.18 Also, as the overall MOF topology is set by the parent MOF, post-

33

synthetic incorporation may be compatible with the equatorial carbox-ylates in the Ru(tda)(PyR)2 complexes that otherwise may compete for SBU coordination.

Figure 15. Synthetic strategies for the incorporation of Ru(tda)(PyR)2 analogues into MOFs, illustrated with Ru(tda)(ina)2 as an example. Reprinted with permission from The Royal Society of Chemistry.91

The direct approach 1 for the incorporation of WOCs has been successfully used before, mostly employing rigid porphyrin- or phthalocyanine-based cat-alytic linkers.73,81,84 The co-synthetic strategy 2 has also been reported.82,83,86 Our group has mainly used the post-synthetic approach 3 to produce catalyst-doped MOFs.16,18,85 It should be noted that for approaches 2 and 3, the catalyst loading needs to be measured after synthesis, e.g., by analysing digested bulk

34

MOFs (see Section 3.6). Furthermore, in the case of post-synthetic approach 3, the spatial distribution of the inserted catalyst needs to be considered. This, in itself, is a significant challenge that is addressed in Chapter 5 of this thesis. In the next two sections, we explore the synthetic strategies outlined above for incorporating derivatives of the Ru(tda) catalyst into MOFs of two target to-pologies: UiO and NU-1000.

4.4.1. UiO-type target topology (Paper I) To construct a UiO-type MOF that has Ru(tda)(Py)2 incorporated as a ditopic linker, the Ru complex was augmented with anchoring carboxylic groups at axial positions (vide supra).

Direct solvothermal synthesis and co-synthesis The direct solvothermal synthetic method using Ru(tda)(ina)2 as the sole linker to afford a UiO-type MOF was attempted first. Although the conven-tional solvent for MOF synthesis is DMF, dimethylacetamide (DMA) was used instead due to the suspected ability of DMF to deactivate catalytic activ-ity by coordination to the ruthenium centre.91 The reaction was set-up using a 1:1 molar mixture of Ru(tda)(ina)2, ZrOCl2(H2O)8, and a 10-fold molar excess of a modulator (either formic, acetic, or benzoic acid). The resulting mixtures were then sealed, ultrasonicated to dissolve the components, and incubated at a fixed temperature between 80 °C and 120 °C for two days. Unfortunately, after washing and drying, the obtained material 1 was found to be amorphous by PXRD and SEM (Figure 16), regardless of the conditions used. We hy-pothesized that the UiO structure was not achievable due to the high steric demand caused by the equatorial tda ligands at the Ru(tda)(ina)2.

Figure 16. PXRD pattern (left) and SEM micrograph (right) of material 1 demon-strating its amorphous nature.

35

Next, we attempted to dilute the fraction of the bulky Ru-terpyridyl linkers by introducing a less sterically demanding co-linker to the solvothermal synthe-sis. Ethynedibenzoic acid (edba) was chosen as a suitable co-linker as it matches the ruthenium complex in length, while also being structurally rigid. The explored MOF syntheses conditions were analogous to the ones in the previous section, with the addition of edba. Three different molar ratios of Ru(tda)(ina)2 : edba were tested (1:1, 1:2 and 1:8). Unfortunately, all attempts resulted in an amorphous material 2 as demonstrated by SEM (analogous to material 1) and PXRD (Figure 17).

Figure 17. PXRD patterns of the products of direct synthesis with varying ratios of co-linkers, all showing an amorphous phase. Reprinted with permission from The Royal Society of Chemistry.91

Thus, both the direct synthetic and the co-synthetic solvothermal approaches using the ditopic Ru(tda)(ina)2 linker failed to produce crystalline materials under all conditions tested. This finding indicates that the steric strain between the catalytic linkers was, in fact, not the major factor obstructing the MOF growth.

To better understand the problem, materials 1 and 2 were studied by cyclic voltammetry (CV). All measurements were performed in pH 7.0 phosphate buffer of ionic strength 0,1 M, using a Ag/AgCl reference electrode, a GC counter electrode, and a scan rate ν = 0,1 V/s. First, the homogeneous Ru(tda)(ina)2 complex was measured to obtain its characteristic redox features for reference purposes. The CV of the molecular reference complex features two well-separated reversible waves that correspond to Ru(II)/Ru(III) and Ru(III)/Ru(IV) redox transitions (Figure 18). If the complex is successfully incorporated into a MOF, we expect these features to be qualitatively con-served.

36

Figure 18. Cyclic voltammogram of homogeneous Ru(tda)(ina)2 complex in solution of pH 7.0 phosphate buffer of ionic strength 0,1 M, using a GC working electrode, a Ag/AgCl reference electrode, and a GC counter electrode, ν = 0,1 V/s.

In order to study heterogeneous particles such as MOFs electrochemically, they have to be interfaced with a working electrode. Initially, a conventional dropcasting method was used in which the solid MOF sample was suspended with mesoporous carbon in ethanolic Nafion solutions, dropcasted onto glassy carbon electrodes, and dried. Unfortunately, these conditions led to partial degradation and linker leaching from the materials. The alternative approach, which was subsequently used throughout this chapter, featured dropcasting ethanolic suspensions of the investigated materials mixed with highly conduc-tive multi-walled carbon nanotubes (MWCNT) without the use of Nafion or other adhesives. Analysis of the centrifuged supernatant of these suspensions revealed that no complex linkers had been leached under these conditions.

In contrast to the reference homogeneous complex, both 1 and 2 displayed two sets of signals, each consisting of multiple sub-waves (Figure 19). Com-paring the locations of the dominant redox features to those of the reference homogeneous complex Ru(tda)(ina)2 suggest that they arise from the same Ru complex. However, the well-defined sub-waves suggest several distinct elec-tronic environments around the discrete Ru centres giving rise to the observed multiple redox events. Taking into consideration the structure of the catalyst, we hypothesize that these different environments originate from lateral SBU coordination by the equatorial tda ligand in various combinations with the ax-ial coordination by the ina ligands. Indeed, this explains both the electrochem-ical behaviour and the disorganized structure of the solids 1 and 2.

37

Figure 19. Cyclic voltammograms of (A) material 1, product of the direct solvother-mal synthesis; (B) material 2, product of the co-synthesis with edba. Gray dotted lines represent the reference CV of the homogeneous complex Ru(tda)(ina)2 for comparison.

Indirect preparation by post-synthetic exchange Introducing the catalytic linker into a pre-made MOF where SBU spacing fits the linker’s axial dimension should help to avoid the issue of tda coordination to the SBUs. We chose to prepare such MOFs using the edba linker introduced above.

The synthesis of the “parent” MOF was performed by mixing ZrOCl2·8H2O with edba in a 1:1 molar ratio with 10 eq. formic acid modulator in DMA, followed by incubation of the mixture at 135 °C for 2 days. The resulting edba-MOF appeared crystalline by SEM, and its PXRD pattern closely matched the published pattern for the edba-based MOF of UiO topol-ogy (Figure 20).92–94

38

Figure 20. PXRD pattern of edba-MOF in comparison to the published pattern.94

Curiously, upon microscopic examination of the edba-MOF obtained in dif-ferent batches, several distinct crystal morphologies were discovered. Each batch tended to be dominated by one of two major morphologies: octahedral (material 3, Figure 21A) or interlaced (material 4, Figure 21B), even though both types produced essentially the same PXRD pattern (Figure 21C).

Figure 21. Characterisation of edba-MOF. SEM micrographs of (A) octahedral crys-tal morphology of material 3. (B) interlaced crystal morphology of material 4. (C) PXRD patterns of octahedral material 3 and interlaced material 4 in comparison to the published pattern.94

Intrigued by the fact that both crystal types result from the same preparation protocol, we hypothesized that the morphology type was governed by small differences in the ratio between linker and SBU precursor during the sol-vothermal synthesis. Such differences could originate from inconsistency dur-ing weighing out of the hygroscopic ZrOCl2·8H2O powder: on a humid day, the massed out zirconyl chloride absorbs more water and eventually generates less SBU clusters in the reaction mixture compared to an arid day. To test this,

39

we reproduced the edba-MOF synthesis using 10-50% molar excess of either edba linker or ZrOCl2·8H2O. In agreement with our hypothesis, the syntheses selectively produced either the octahedral edba-MOF 3 (excess linker) or the interlaced edba-MOF 4 (excess ZrOCl2) (Figure 22). The formation of crystals of similar interlaced morphology has been observed by Cohen and co-workers and, in their case, has been attributed to polymer-induced inhibition of octa-hedral crystal growth.95,96 Applying the same logic here, it appears that the shortage of edba linker inhibits the normal crystal growth, while the shortage of SBU is not critical.

Figure 22. SEM micrographs of selectively prepared material 3 (left) and material 4 (right). Reprinted with permission from The Royal Society of Chemistry.91

After activation in vacuo, both materials 3 and 4 were subjected to PSE with Ru(tda)(ina)2, resulting in the octahedral Ru-edba-MOF 5 and the interlaced Ru-edba-MOF 6. Structural integrity of the MOFs was confirmed by post-PSE PXRD patterns that closely match the original edba-MOF patterns (Figure 23).

Figure 23. PXRD patterns of post-PSE Ru-edba-MOFs 5 and 6 (red) in comparison to original MOFs 3 and 4 (black). (A) MOFs of the octahedral morphology. (B) MOFs of the interlaced morphology.

40

Given the massive difference in external surface area between the two mor-phologies, it was interesting to compare the PSE incorporation yields. To this end, Ru-edba-MOF 5 and 6 were digested and studied by two complementary techniques: ICP-OES to provide the Zr vs. Ru atomic ratio and 1H-NMR spec-troscopy to determine the edba : Ru(tda)(ina)2 ratio. Thereby, one technique measures the Ru-terpyridyl content relative to the SBUs, while the other re-lates the Ru-terpyridyl content to the number of native edba linkers. For Ru-edba-MOF 5 (octahedral morphology), NMR yields 3.2% PSE extent, while ICP gives 4.2%. For Ru-edba-MOF 6 (interlaced morphology), the numbers are somewhat lower with 2.1% by NMR and 2.2% by ICP. The lower incor-poration yield for the interlaced morphology was somewhat counterintuitive, considering that the high external surface area of this morphology should fa-cilitate linker penetration during the PSE. However, the same property may promote leaching of the incorporated linker from the MOF. Consistent with this reasoning, the interlaced Ru-edba-MOF 6 produced, qualitatively, more intensely coloured supernatants during the washing steps, compared to those of the octahedral Ru-edba-MOF 5.

In order to assess the internal surface areas of edba-MOFs 3 and 4 and the extent to which they are altered by metallolinker incorporation in 5 and 6, N2 absorption/desorption isotherms were recorded (Figure 24). From these iso-therms, the internal surface areas were derived by BET analysis. The octahe-dral edba-MOF 3 yielded 1300 m2/g, while the interlaced edba-MOF 4 had only 390 m2/g. This large discrepancy can be explained by a larger abundance of missing-linker defects in edba-MOF 4 compared to that in MOF 3. In fact, the hysteresis in the edba-MOF 4 isotherm hints at the presence of mesopores which, in turn, may result from high density of defects.97 As mentioned in Chapter 2, the BET surface area of UiO-66 is known to grow with increasing defects up to a certain threshold amount, beyond which it begins to drop.39 The number of defects in both edba-MOFs appears to be above this threshold, and the BET surface areas for both morphologies increase upon incorporation of the metallolinker. The effect is more pronounced in case of the interlaced morphology for which the BET surface areas increases from 390 m2/g (4) to 800 m2/g (6), while that of Ru-edba-MOF 5 is increased to 1390 m2/g com-pared to the 1300 m2/g for 3. This finding also indicates that the dominating mechanism of incorporation is likely PSDH rather than PSE.

41

Figure 24. N2 absorption/desorption isotherms of MOFs 3-6. Reprinted with permis-sion from The Royal Society of Chemistry.91

Finally, the electrochemical behaviour of Ru-edba-MOF 5 and 6 was investi-gated by CV. In contrast to materials 1 and 2, the CVs of both post-syntheti-cally produced Ru-MOFs display two well-defined and reversible waves at potentials that are similar to those observed in the CV of the homogeneous Ru-terpyridyl complex (Figure 25). The fact that each of the redox events ap-pears as a discrete single wave indicates that all of the Ru(tda)(ina)2 linkers experience the same electronic environment. From this, we infer that nearly all of the Ru(tda)(ina)2 linkers are incorporated through the same coordination mode and not via alternative SBU coordination modes, for example, through the equatorial tda carboxylates. Taking into consideration that Ru-edba-MOFs 5 and 6 exhibit no leaching of the Ru linker (which can be expected from monotopic coordination), the metallolinker is most likely incorporated in a ditopic fashion, i.e., as a structural linker that is bridging between the SBUs.

42

Figure 25. Cyclic voltammograms of (A) octahedral Ru-edba-MOF 5 and (B) inter-laced Ru-edba-MOF 6. Gray dotted lines represent the reference CV of the homoge-neous complex for comparison.

4.4.2. NU-1000-type target topology (Paper II) In a follow-up study, a tetratopic version of the Ru(tda)(Py)2 complex bearing Py(PhCOOH)2 ligands at the axial positions (Figure 10C) was considered. The new linker was designed to explore the possibility of its incorporation into NU-1000. The native linker of that MOF is H4TBAPy (1,3,6,8-pyrenetetrayl)tetrakis-benzoic acid) that is similar to the tetratopic Ru(tda)(Py(PhCOOH)2)2 complex in terms of geometry and dimensions, both of which are factors that should enable facile linker incorporation (Figure 26). In addition, the pyrene core of the TBAPy linker is known to be oxidized at ca. 1.1 V and 1.4 V vs. Ag/AgCl in CH2Cl298 and CH3CN,99 respectively, i.e. at a potential that is roughly similar to that at which activated Ru(tda) complexes catalyse electrochemical water oxidation.89 Although potentials in aqueous and non-aqueous media are not directly compara-ble, there is thus a possibility that the combination of the two linkers may be syn-ergistic. In an ideal scenario, the pyrene linkers may assist turnover by providing a hole transport pathway to catalytic linkers that are situated throughout the MOF. For this reason, most of our synthetic efforts focused on direct solvothermal ap-proaches to MOF construction, as the tetracarboxylic Ru(tda)(Py(PhCOOH)2)2 linker was expected to exhibit very slow intra-MOF diffusivity, and subsequently be restricted to outer MOF layers during any PSE process (this issue is discussed in detail in Chapter 5).54 Of the two direct solvothermal methods, the attempted synthesis involving the Ru complex as the sole linker resulted in amorphous ma-terials analogous to the situation with the ditopic Ru(tda)(ina)2. The alternative co-synthesis approach is the focus of the following section and Paper II.

43

Figure 26. Structures of (A) H4TBAPy – the native NU-1000 linker, (B) Ru(tda)(Py(PhCOOH)2)2, including its single-crystal X-ray structure.

The solvothermal mixed-linker synthesis of NU-1000 was performed by mix-ing 1 eq. ZrOCl2·8H2O, 0.35 eq. H4TBAPy, and Ru(tda)(Py(PhCOOH)2)2 with 73 eq. benzoic acid modulator in DMF followed by incubation at 100 °C for 3 days. The amount of the Ru linker in the synthesis was varied from 5 to 30 mol.% relative to the ZrOCl2 precursor, in 5% increments. In contrast to sim-ilar efforts with the ditopic Ru(tda)(ina)2, all of the co-synthesis conditions tested herein resulted in crystalline materials 7-12 with major PXRD reflec-tions characteristic of the NU-1000 topology (Figure 27).

Figure 27. PXRD patterns of the mixed-linker materials 7-12 obtained by solvother-mal co-synthesis with varying Ru(tda)(Py(PhCOOH)2)2 loadings, in comparison to the reference pattern of pristine NU-1000.

44

Comparing PXRD patterns of MOFs 7-12 revealed that increasing the amount of the Ru linker in the solvothermal synthesis produces frameworks with wider XRD reflections, while Ru loading of 75% and higher (not shown here) resulted in amorphous materials. This progressive loss of crystallinity can also be correlated with decreasing crystallite sizes. The latter is clearly seen in SEM images of these materials (Figure 28), which show average crystal size decreasing from ~4 µm down to sub-micrometer dimensions with increasing levels of Ru doping. Microscopic EDX (Energy-Dispersive X-ray spectros-copy) imaging confirmed the presence of Ru throughout the crystals in all of the materials.

Figure 28. SEM micrographs of materials 7-12 demonstrating progressively decreas-ing crystallite size with increasing loading of the Ru(tda) linker (indicated on the mi-crographs in mol. % relative to ZrOCl2).

45

The amount of incorporated Ru(tda)(Py(PhCOOH)2)2 relative to Zr was quan-tified by ICP-OES of samples of MOFs 7-12 that had been digested in con-centrated H2O2/HNO3. Curiously, the analysis showed that the systematic in-crease of the Ru linker during solvothermal syntheses resulted in a non-linear increase in Ru incorporation. The investigated MOFs formed two groups based on the measured Ru vs. Zr content: materials 7-9 with lower incorpora-tion yields (1.5% Ru), and materials 10-12 with higher incorporation yields (4% Ru). The situation when the final MOF composition does not directly mirror the ratio between co-linkers in the synthesis mixture is common in the literature and is usually explained by different propensities of the linkers to form coordination bonds with the SBU.100 However, even in such cases, a lin-ear correlation between linker ratios in the synthesis and in the final frame-work is usually observed. The non-linear increase in Ru content that we ob-served in the present study indicates the presence of additional factors that affect MOF growth, such as a vast difference in solubility between the linkers or a slight mismatch in dimension.

Next, the mixed-linker materials were investigated by cyclic voltammetry. For reference, also the CV of the homogeneous Ru(tda)(Py(PhCOOH)2)2 was measured (Figure 29A). In addition, due to the redox-active nature of the na-tive H4TBAPy linker, a reference CV of the pristine NU-1000 MOF in water was obtained to identify redox features of the pyrene linker (black CV in Fig-ure 29B). Indeed, similarly to measurements in organic solvents, pristine NU-1000 displays a pyrene-based oxidation close to 1.2 V vs. NHE, which lies far anodic of the characteristic Ru(tda)-based oxidations (as visible in Figure 29A). Unfortunately, no such Ru(tda)-based features could be observed in the CVs of any of the mixed-linker materials, as illustrated representatively by the CV of material 12 (30% Ru loading, Figure 29B), and the CV of 12 is essen-tially the same as that of pristine NU-1000.

Figure 29. Cyclic voltammograms of (A) reference homogeneous complex Ru(tda)(Py(PhCOOH)2)2, and (B) material 12 (red) in comparison to pristine NU-1000 (black). Conditions: GC working electrode, 0,1 M pH 7.0 phosphate buffer so-lution, Ag/AgCl reference electrode, GC counter electrode, ν = 0,1 V/s.

46

As the presence of the Ru linker in the NU-1000 crystals was confirmed by ICP and EDX, this lack of clearly visible redox features that can be assigned to the Ru(tda)(Py(PhCOOH)2)2 linkers indicates that these linkers are electro-chemically inaccessible. We attribute this to a combination of low Ru doping and inherently poor conductivity of NU-1000 at potentials cathodic to pyrene oxidation. In order to reach a Ru complex inside the MOF, electrons or holes need to propagate through the framework by a hopping mechanism.22 In order to reach redox-active units situated deep inside the crystal, continuous hop-ping “pathways” need to be present. The probability of electron transfer by hopping is inversely proportional to the distance between the hopping sites.101 At applied voltages below the pyrene’s oxidation potential, the only accessible hopping sites are the Ru linkers themselves. With the doping levels around 6% (as determined by ICP), the average distance between neighbouring Ru(tda) centres might be too long for efficient hole hopping – a situation ag-gravated by the rather large dimensions of both participating linkers.

4.5. Conclusions & outlook In this chapter, three synthetic approaches for the incorporation of derivatives of Ru(tda)(PyR)2 as linkers into MOFs were considered: the direct solvother-mal synthesis (a) without or (b) with an inert co-linker, as well as (c) post-synthetic introduction. In addition, two different MOF topologies were inves-tigated. To achieve a UiO-type hybrid MOF, the ditopic Ru(tda)(ina)2 linker was used. For the NU-1000-type, the tetratopic Ru(tda)(Py(PhCOOH)2)2 linker was introduced.

For the UiO-type topology, both direct methods yielded amorphous mate-rials that most likely result from SBU ligation by the equatorial tda carbox-ylate in a random fashion. However, the post-synthetic approach yielded crys-talline materials. During the preparation of the parent edba-MOF, we discov-ered that it forms two distinct morphologies controlled by small variations in the linker:SBU ratio during the synthesis. The Ru complex could be success-fully incorporated into both morphologies post-synthetically. Based on BET internal surface area measurements, both morphologies appear to contain high concentrations of defects, although the number of missing linkers seems to be especially high in the interlaced edba-MOF 4. Strong indications were ob-tained that Ru(tda)(ina)2 insertion occurs preferentially through filling of de-fects (replacing modulator and solvent molecules at the nodes) rather than substitution of edba struts.

The efforts to introduce Ru(tda)(Py(PhCOOH)2)2 into NU-1000 focused on direct solvothermal methods. Using the Ru complex as the sole linker, the solvothermal synthesis resulted in amorphous materials, mirroring the out-come for the ditopic Ru(tda)(ina)2. However, in contrast to experiments with the ditopic Ru complex, combining tetratopic Ru(tda)(Py(PhCOOH)2)2 with a

47

pyrene-based co-linker in solvothermal syntheses produced a series of crys-talline materials with variated Ru loading. The overall incorporation yields of the Ru linker were found by ICP-OES to be lower than expected based on synthesis ratios of the linkers. At the same time, attempts to enhance the in-corporation yield by increasing the Ru(tda)(Py(PhCOOH)2)2 content during the synthesis resulted in MOFs of smaller crystal size and poorer quality, with Ru ratios above a certain threshold producing only amorphous materials. Cy-clic voltammograms of the NU-1000-Ru MOFs exhibited no characteristic Ru(tda)-based features, which indicates that the density of the Ru centres is below the percolation threshold for redox hopping – a phenomenon with prec-edent in the literature.22,82,101

Overall, this chapter describes two successfully solved synthetic problems. Suitable conditions were found to incorporate Ru(tda)(Py)2 derivatives into MOFs of two distinct topologies. For the future development of this project, steps should be taken toward catalyst activation by bulk electrolysis of the MOF. Preliminary experiments show that this is challenging; for the doped NU-1000, this is further complicated by the discovered inaccessibility of the Ru centres within the solvent potential window. A promising approach would be to investigate the electrochemical behaviour of this system at potentials sufficiently positive to involve pyrene oxidation: even at low incorporation yields, the Ru(tda)(Py(PhCOOH)2)2 linkers may yet be found accessible through hole hopping mediated by the pyrene linker, which may even permit catalyst activation toward water oxidation.

48

5. Depth profiling of MOF crystals after post-synthetic modification using microbeam RBS (Paper III)

In this chapter, a novel technique for accurate and non-destructive MOF depth profiling is introduced. Its utility is tested by studying depth distribution of molecules post-synthetically introduced into UiO-66 MOF.

5.1. Introduction In the previous chapter, PSE and PSDH were demonstrated to be effective methods for the incorporation of functional molecules into MOFs. As dis-cussed in Chapter 2, post-synthetic modifications can result in various spatial distributions of the inserted molecule throughout the crystalline material. If this variable can be controlled, it presents a powerful tool for modulating the material properties, as demonstrated, for example, by the application of core-shell MOF nanostructures in catalysis20 or drug delivery.102 In order to under-stand and exploit the mechanisms governing the PSE depth distribution, a ro-bust tool for MOF depth profiling is required. The currently existing MOF depth profiling techniques are lacking in universality, efficiency and accuracy, as they tend to rely on physical slicing of large MOF crystals and subsequent microspectroscopic analyses.

In this chapter, we demonstrate how a classical physical technique, Ruth-erford Backscattering Spectrometry, can be utilized for accurate and non-de-structive depth profiling of MOFs. As a testbed, we investigated the spatial distribution of molecules post-synthetically introduced into relatively large (10-15 µm) UiO-66 crystals.

5.2. Microbeam RBS method The standard RBS technique, as described in Chapter 3, relies on the wide-area probing of a bulk target with accelerated ions. While still extremely use-ful for bulk compositional analysis, conventional RBS suffers from low depth perception accuracy when applied to rough samples, such as MOF powders.

49

Uneven sample surfaces (such as randomly oriented crystals) also contribute to lower depth profiling accuracy (Figure 30).

Figure 30. Sample surface roughness compromises depth perception accuracy of conventional RBS. A feature nucleus located at the same depth will appear as lying more or less deep depending on the crystal’s orientation relative to the probing ion beam.

As a means of achieving high MOF depth profiling accuracy, a microbeam modification of the standard RBS method was chosen. In this set-up, the prob-ing ion beam is focused into a ~1 µm-wide spot on a single MOF crystal (Fig-ure 31). To provide accurate information about spatial distribution of nuclei, the crystal should be oriented with a face perpendicular to the probing beam. Also, the sample should consist of a population of single MOF crystals of similar size and orientation to allow for the accumulating of good statistics.

Figure 31. Illustration of the microbeam RBS modification.

50

5.3. Target MOF preparation A combination of two published methods was used for growing highly ori-ented individual UiO-66 on Si <100>,103 and of narrow size distribution.36 The optimal size of the crystals was found to be 10-15 µm, which is large enough to be suitable for microbeam RBS analysis but still sufficiently small to avoid detachment from a vertically-mounted surface during the measurements.

To obtain the desired MOF crystals, Si <100> slides that had been cleaned with piranha solution were submerged in a mixture of terephthalic acid (ta), ZrOCl2·8H2O, and formic acid (fa) modulator in diethylformamide (DEF) and then incubated at 135 °C for 2 days. The amount of formic acid, which dictates the size of the produced crystals, was varied until the desired crystal size was achieved. The resulting surface-anchored UiO-66 was washed, soaked in DMF and solvent exchanged in CH2Cl2 prior to further characterisation.

The obtained UiO-66@Si was characterized by SEM and PXRD. The mi-croscopic investigation revealed individual, mostly oriented crystals 10-20 µm wide and 10-15 µm thick. The PXRD pattern featured characteristic peaks, as well as the absence of peaks that are usually observed in randomly oriented mixtures, indicating an almost exclusive <111> orientation of the UiO-66 crystals (Figure 32).

Figure 32. SEM and XRD analysis of UiO-66@Si showing high degree of size- and orientational control.

5.4. Post-synthetic exchange In order to be distinguishable from native MOF components – organic linkers and zirconia clusters – by RBS, the post-synthetically introduced molecules

51

should ideally carry a heavy element marker, such as iodine. With this in mind, the pristine UiO-66@Si was subjected to PSE with two different molecules: iodoterephthalic acid (ita) and 3-iodobenzoic acid (iba) (Figure 33). Ita was chosen based on its structural similarity to the native terephthalic acid linker, as well as existing previous results from our group to refer to.53,104 Iba is a smaller analogue that lacks one carboxylate group and that can be expected to diffuse more easily into UiO-66 than ita.

Figure 33. Molecules incorporated into the model UiO-66@Si by PSE: iodotereph-thalic acid (ita, left) and 3-iodobenzoic acid (iba, right).

While optimising the PSE conditions, various temperatures, process durations, and concentrations of the inserted molecules were tested. Optimal conditions were found when the UiO-66@Si slides were incubated with a 250 mM solu-tion of ita or iba at 50 °C for 24 hours. Curiously, we found that the pre-evacuation (“activation”) procedure of the MOF is absolutely necessary to achieve a measurable PSE yield. Vacuum activation of the MOF is usually an implicit part of the routine preparation procedure, and has, to the best of our knowledge, not been explicitly recognized as a PSE enabler.42,105–107 Follow-ing the incubation, the UiO-66-ita@Si and UiO-66-iba@Si slides were washed and soaked over a prolonged period of time in order to remove mole-cules that are trapped in the pores, but not coordinated to the SBUs. The re-tention of structural integrity of the MOFs during the PSE was confirmed by SEM and XRD (Figure 34).

52

Figure 34. XRD patterns of UiO-66-ita@Si (left) and UiO-66-iba@Si (right) demonstrating conservation of MOF structural integrity during the PSE.

5.5. Beam damage and probing mode During early microbeam RBS attempts, the ion beam probe was spot-focused on an individual crystal for recoil data collection. Unfortunately, this approach led to irreversible local damage to the crystals, appearing as gradual deterio-ration of the RBS signal during the measurement as well as visible damage in post-mortem SEM images (Figure 35).

Figure 35. Example SEM micrographs of UiO-66@Si crystals damaged by the fo-cused ion microbeam.

Based on the hypothesis that the damage was caused by inefficient dissipation of highly localized heat, the probing beam was set to scanning mode over a wider area including one or several crystals of interest. The obtained data, en-coded with the position of the beam, was then integrated including only the areas corresponding to the crystals and discarding the background substrate

53

(Figure 36A). The obtained signal quality together with post-RBS SEM con-firmed that the target crystals were indeed able to withstand the thermal stress (Figure 36B) when the microbeam was not stationary, but scanned.

Figure 36. (A) Example map of Zr RBS signal corresponding to scattering of He+ ions from a single MOF crystal. The dashed lines indicate the approximate crystal location and orientation. (B) SEM micrograph of a target UiO-66@Si crystal after the scanning microbeam RBS measurement.

5.6. Results & Discussion 5.6.1. Depth distribution of ita and iba The RBS spectra obtained from UiO-66-ita@Si and UiO-66-iba@Si are pre-sented in Figure 37 (A and B, respectively). The RBS signal for an element evenly distributed in a thick target is expected to appear as a trapezoid plateau where the high-energy edge corresponds to surface recoil and provides quali-tative identification of the element, while the profile of the plateau corre-sponds to the element’s depth distribution. If the element is restricted to the surface, its signal profile will be narrow compared to other elements consti-tuting the target. In the two samples studied here, the signals for native MOF elements (C, O and Zr) are practically identical, while iodine signals that orig-inate from the introduced molecules display a dramatic difference.

54

Figure 37. RBS spectra of (A) UiO-66-ita@Si and (B) UiO-66-iba@Si overlaid with corresponding simulations.

Unlike the bulk MOF elements, iodine in the UiO-66-ita@Si MOF was found to have a narrow and sharp signal profile, which corresponds to highest ita concentration in the outermost 0,2 µm of the crystal and sharp, exponential decrease toward the centre. Although there is no clear boundary between io-dine-enriched and iodine-depleted regions, we can colloquially call this archi-tecture the “core-shell distribution.”54 On the other hand, UiO-66-iba@Si dis-plays a typical plateau-type iodine signal with no apparent fall-off at lower energies. This corresponds to uniform distribution of iba throughout the target crystal. For the sake of confidence, two theoretical fittings were considered (Figure 37B): assuming evenly distributed iodine (top), and gradient iodine concentration (bottom). The assumption of the uniformly distributed iodine produced a far superior fit to the experimental spectrum. The only way to ob-tain a decent fit assuming an iodine concentration gradient is to increase the Zr concentration accordingly when going deeper into the crystal – a scenario for which there is no physical reason. To summarize, we observe that PSE with 3-iodobenzoic acid produced a hybrid MOF with uniformly distributed iba, while PSE with iodoterephthalic acid results in a core-shell type of ar-rangement (Figure 38).

55

Figure 38. Schematic interpretation of the PSE outcomes as determined by RBS.

It is interesting to consider the result we obtained for UiO-66-ita@Si in con-text of previous experiments our group conducted on similar systems.53,104 In those early experiments, PSE of ita into sub-micrometer-sized UiO-66 crys-tals was studied and found to produce a uniform distribution of the exchanging molecule throughout the MOF. Seemingly in contradiction to the core-shell structure observed for ita here, it is, in fact, not surprising if the massive dif-ference in crystal size is taken into consideration. In the present study, using ~10 µm-thick crystals of the parent MOF, we essentially detected that the dif-fusion front of ita does not reach deeper than ~0.2 µm into the MOF over the duration of the experiment. In the earlier studies, on the other hand, the crys-tals themselves were approx. 0.2-0.5 µm in diameter, which, we believe, al-lowed the ita to diffuse throughout the material during the PSE. If the crystal is smaller or close in size to the distance travelled by the diffusion front of the exchanging molecule during PSE, then all areas in the crystal can be reached by overlapping diffusion fronts, resulting in a uniform distribution of the in-corporated molecule.

Moving on to UiO-66-iba, we see uniform distribution of the molecule throughout the entire 10 µm-thick crystal. Applying familiar logic, we ration-alize this in terms of intra-MOF diffusion, which in case of iba is fast enough to traverse the whole crystal over the duration of the PSE process. The intra-MOF diffusion rate depends on multiple factors, such as size, the amount of carboxylates (or other coordinating groups), and their pKa. As a general rule, ligands with higher pKa tend to form stronger bonds with the SBU,108,109 how-ever, they are less competitive for SBU ligation compared to ligands with lower pKa by having a smaller proportion in the deprotonated form.110 Exper-iments with Solvent-Assisted Ligand Insertion (SALI) showed that ligands with a pKa lower than that of the benzoic acid modulator quantitatively sub-stituted the SBU-bound benzoates, while ligands with higher pKa did not react at all.111 In the present case, the pKa values of ita (pKa1 = 2,45 ± 0.01)112 and iba (pKa = 3,87)113 are smaller and larger, respectively, than that of the native MOF species – terephthalic acid and formic acid (pKa1 = 3,54 and pKa = 3,75,

56

respectively),113 yet both successfully get introduced into the MOF. However, the higher pKa of iba means that it exists much less in the deprotonated state compared to ita, making it a less aggressive ligand, which, in turn, facilitates a less hindered diffusion into the crystal. Another major factor governing the intra-MOF diffusion rate is the number of carboxylic acid groups in the dif-fusing molecule. In fact, by comparing diffusion of mono- and dicarboxylic acids and their esters into large crystals of MOF-5, Matzger et al. demon-strated that the presence of free acid groups is the most critical factor in the formation of core-shell structures.54 In summary, we rationalize the fundamen-tally different post-PSE spatial distributions of ita and iba based on their dif-ferent intra-MOF diffusion rates, which, in turn, are determined by both the amount of COOH groups and their pKa values, with the former being the dom-inant factor.

5.6.2. Framework defects and PSE yield As mentioned in Chapter 2, PSE is known to preferentially occur first at defect sites (substituting modulator species), and the actual linker replacement takes place after exhausting most of the modulator-capped sites.40 With this in mind, we were interested in comparing the amount of inserted ita or iba to the amount of defects in the pristine UiO-66.

To establish the original defect population, a bulk UiO-66 sample was pre-pared by the same procedure as that used for growing the MOF on Si. After washing, soaking, and drying, the MOF was digested in a mixture of HF/dmso-d6 and the resulting solution measured by 1H-NMR to determine the ratio between terephthalic acid and formic acid. The obtained 1H-NMR spec-trum is presented in Figure 39. Integration of the peaks originating from for-mic acid and terephthalic acid yielded a molar ratio of 2:5, which corresponds exactly to the MOF formula Zr6O4(OH)4(ta)5(fa)2, where one linker per for-mula unit is missing and replaced by two modulator formates. With this infor-mation in hand, we then set out to determine the amount of ita and iba inserted by the PSE.

57

Figure 39. 1H-NMR spectrum of digested bulk pristine UiO-66 with intensity inte-gration of peaks corresponding to ta (at 7.99 ppm) and fa (at 8.07 ppm).

The RBS method allows accurate determination of the compositional I:Zr ra-tio in exchanged MOFs. In the case of UiO-66-ita@Si, the iodine concentra-tion is depth-dependent and is maximal in the outer shell with a I:Zr ratio of 0,14:1. This corresponds to 0,84 ita linkers for each Zr6O4(OH)4 SBU and can be interpreted in terms of two limiting linker incorporation scenarios. Assum-ing that all ita was incorporated in a ditopic manner, the ratio corresponds to 84% of all modulators in the shell being exchanged. On the other hand, if all ita was incorporated via a single carboxylate, only 42% of all modulators in the shell have been replaced.

In UiO-66-iba@Si, the iodine concentration is uniform throughout the crystal and is significantly higher compared to ita, with an I:Zr ratio of 0,44:1 as determined by RBS. This ratio corresponds to 2.6 iba molecules per SBU unit, which is, in fact, higher than expected for complete exchange of modu-lators (which would result in only 2 per SBU). To test whether partial replace-ment of structural linkers occurs, we reproduced the iba PSE experiment on a bulk UiO-66 sample, digested the resulting UiO-66-iba and measured the or-ganic content by NMR. Interestingly, the obtained iba:ta ratio of 0,4:1 was essentially identical to the original fa:ta ratio in the pristine UiO-66 MOF, indicating that iba incorporation indeed did not proceed beyond substitution of modulators, leaving the structural terephthalic acid linkers intact. This find-ing is in agreement with the existing PSE literature.40 Taking this into account, the high I:Zr ratio measured by RBS points to a significant amount of missing

58

cluster or point Zr defects, which are not detectable by NMR. As RBS quan-tifies I:Zr ratios, such missing Zr defects will lead to an overestimation of the incorporation yield of the iodine-containing molecule. This notion would res-onate with multiple reports that describe predominantly “missing cluster” character of defects in UiO-66.37,40,50,114 This finding also illustrates the power of RBS as it gives the very rare opportunity to determine linker:zirconium ratios which complement the linker:linker or linker:modulator ratios obtaina-ble by conventional methods such as NMR.

5.7. Conclusions & Outlook In this chapter, we explored Rutherford Backscattering Spectrometry (RBS) for applications in MOF depth profiling. To fully utilize this method, we de-veloped a surface-attached MOF platform with highly oriented individual UiO-66 crystals of controllable size. Benefitting from a microbeam probe in scanning mode, depth profiling could be performed on individual single crys-tals. The RBS technique and the MOF platform were used to answer a partic-ular question concerning the spatial distribution of two iodine-containing mol-ecules that were post-synthetically introduced into the surface-bound MOF. It was found that the monocarboxylic iba was incorporated evenly throughout the 10 µm-thick crystals, while the dicarboxylic ita exchanged predominantly in the outermost 0.2 µm layer of the crystals. In combination with digested bulk analysis by 1H-NMR, both ita and iba were found to react exclusively at defect sites, replacing formate modulators and not terephthalate linkers. Com-paring the iba PSE incorporation yields measured by RBS and bulk NMR re-vealed a deviation in structural Zr content, suggesting the presence of SBU defects.

Together, the microbeam RBS technique and the presented MOF platform form a powerful tool for addressing both practical and fundamental questions in MOF research, such as depth/concentration profiling, intra-MOF diffusion and framework defects. One such direction that is currently not addressed in the field involves the scrambling of linkers inside a MOF crystal.115 Given the equilibrium-controlled nature of the linker-SBU bond, the question arises whether linkers can migrate through the framework, a process that may, for example, be driven solvothermally. Such a process could, in principle, be de-tected as scrambling of an initial core-shell structure. Continuing the PSE depth distribution theme, the exploration of a broader selection of molecules such as esters and those with other coordinating groups, would help to shine light on intra-MOF diffusion kinetics. Such information is particularly scarce in liquid phase, but of high relevance in context of, for example, MOF elec-trochemistry and catalysis.

59

6. Toward MOF electrodes: growing M2L2P SURMOFs on silicon with high degree of orientational control (Paper IV)

In this chapter, a method development study for the construction of surface-mounted MOFs (SURMOFs) on bare silicon substrates with high degree of orientational control is presented. The study is an important step towards prac-tical implementation of MOFs in (photo)electrocatalysis.

6.1. Introduction Practical applications of electrocatalytic MOFs require their efficient interfac-ing with the electrode. Given the inherently poor electrical conductivity of most MOFs, the interface requires maximizing the area of contact between the MOF and the electrode for optimal performance. This call is answered by the concept of growing MOFs as films directly on the electrode – in contrast to depositing pre-made bulk MOFs by dropcasting116,117 or electrophore-sis.98,118,119 The thickness of the resulting film is an important variable in po-tential applications as it affects the kinetic interplay between charge and coun-terion transport, diffusion of substrates/products, and catalytic turnover – all of which occur inside the MOF lattice.88 That said, the most common tech-nique for growing MOF films in situ – solvothermal synthesis over the elec-trode substrate82,120 – suffers from a poor level of control over thickness, cov-erage, and roughness of the produced film.121

An alternative approach is the use of surface-mounted MOFs (SUR-MOFs).122,123 Grown epitaxially in a layer-by layer (LbL) fashion, SURMOFs provide excellent substrate coverage and control over film thickness, as well as single orientation of the crystalline film – in principle. Conventionally, SURMOFs are prepared on Au substrates modified with thiol-based self-as-sembled monolayers (SAMs).57,124 While this method is efficient, a more earth-abundant material would likely be needed for large-scale industrial ap-plications. Silicon is a good alternative substrate for MOF growth due to its flatness, high availability, well-developed surface chemistry, and even high

60

conductivity with the correct dopants. Additionally, being a photoelectric ma-terial, Si is a potential platform for the future coupling of MOF electrocatalysis to light harvesting.125

In this chapter, the development of methodology for epitaxial growth of layer-pillar type SURMOFs on polished Si <100> substrates is presented, us-ing MOFs of the general formula M2L2P (M = metal, L = layer linker, P = pillar linker) as a model system.

6.2. Layer-by-layer SURMOF growth Conceptually, SURMOF preparation involves exposing a flat solid substrate (often modified with a SAM) separately and sequentially to dilute solutions of metal salt (SBU) and linkers in an alternating fashion, normally with washing steps in between. Ideally, each cycle discretely extends the film by one layer, allowing complete control of the film thickness by the number of repeating growth cycles.122

Since SURMOF growth involves many repetitions, the method is very te-dious if performed manually. Instead, a number of technical solutions exist for automating the process, such as pump systems, dipping robot arms, and spray-ing. In this work, we utilized an automated peristaltic pump system to move solutions of metal salt and linkers as well as rinsing solvent to and from the sample-containing chamber. The resulting SURMOF films were routinely characterized by out-of-plane grazing incidence X-ray diffraction (GIXRD) and SEM.

6.3. Epitaxial growth of Cu2(bdc)2(dabco) SURMOF growth on Si was investigated on a classic paddlewheel Cu2(bdc)2(dabco) SURMOF (bdc = 1,4-benzene dicarboxylate, dabco = 1,4-diazabicyclo[2.2.2]octane, Figure 40). This MOF has previously been inves-tigated as a SURMOF on Au/SAM substrates.126–129

61

Figure 40. Structure of Cu2(bdc)2(dabco) SURMOF and its principal crystallo-graphic planes (parts of the structure are omitted for clarity). Reprinted with permis-sion from ACS.130

In this MOF, the Cu2 paddlewheel SBUs are connected into “layers” by ter-ephthalate linkers along the [100] crystallographic axis; dabco “pillars,” in turn, connect the layers along the perpendicular [001] dimension. Accord-ingly, priming the substrate with pyridyl- or carboxylate-terminated SAMs has been used to guide oriented crystal growth along the [001] and [100] di-mensions, respectively.126,131

6.3.1. Growth on oxidized Si surface vs. quartz Our initial proof-of-concept attempt at growing Cu2(bdc)2(dabco) on silicon was inspired by an existing report of growing isoreticular Cu2(D/L-camphoric acid)2(dabco) SURMOF on quartz.132 Hypothesizing that the native SiO2 layer of a Si wafer should mimic the quartz surface, we subjected piranha-cleaned Si slides to 40 cycles of LbL growth, with each cycle consisting of consecutive soaking in ethanolic solutions of Cu(OAc)2, terephthalic acid and dabco for 15 minutes, with rinsing steps in between, in a chamber that is kept at 50 °C. For comparison, the original procedure with quartz was reproduced using the same piranha treatment and soaking sequence. Investigation of the resulting SURMOF films on Si and quartz with GIXRD revealed opposite results for the two substrates: while quartz produced a film with the preferred [001] ori-entation according to literature (Figure 41 A),132 the MOF on oxidized Si grew mostly along the [100] orientation (Figure 41 B). SEM analysis revealed Cu2(bdc)2(dabco)@Si to mostly consist of isolated vertical platelets with very poor surface coverage (Figure 41 C). In the literature, analogous SURMOF Cu2(F4bdc)2(dabco) (F4bdc = tetrafluorobenzene-1,4-dicarboxylate) grown on SAM-modified Au substrates showed similar vertical platelets with poor cov-erage in case of [100] orientation, but smooth and homogeneous films when

62

the [001] orientation was obtained.131 Inspired by this precedent, from this point forward we sought to achieve the [001] orientation of our layer-pillar MOF in order to optimize surface coverage and homogeneity.

Figure 41. Out-of-plane GIXRD patterns of Cu2(bdc)2(dabco) grown on (A) quartz substrate and (B) Si substrate in comparison to simulated PXRD pattern of analo-gous Cu2(F4bdc)2(dabco) including all crystallographic orientations.133 (C) SEM mi-crograph of Cu2(bdc)2(dabco) grown on Si.130

6.3.2. On Si modified with anchoring SAMs In our next step, we sought to imitate the conventional approach used for gold substrates: priming the final SURMOF orientation using substrates pre-coated with SAMs presenting the desired headgroups (e.g., pyridyl, hydroxyl or car-boxylate) to the solution. In order to template Cu2(bdc)2(dabco) growth with pillars perpendicular to the surface ([001] orientation), the substrate should be decorated with nitrogen-based monodentate coordinating groups. With that in mind, three anchoring molecules were tested featuring either carboxylate (ina) or silatrane (sil1 and sil2) anchoring groups for binding to the Si surface and either pyridine or amine headgroups to prime the [001] SURMOF growth (Figure 42).

Figure 42. Surface anchors used for templating SURMOF growth.

63

After cleaning and modifying the surface, the anchor-decorated Si slides were subjected to the same soaking sequence. Pre-cleaning and sensitization meth-ods were varied, but still the majority of obtained SURMOF films displayed a mixture of [100] and [001] orientations upon GIXRD investigation, or even less growth overall (Figure 43). We interpret this inability of SAMs on Si to prime exclusively oriented Cu2(bdc)2(dabco) growth as resulting from incom-plete SAM surface coverage.

Figure 43. GIXRD patterns of Cu2(bdc)2(dabco) SURMOF grown on SAM-modi-fied Si substrates under various cleaning and treatment conditions in comparison to simulated PXRD pattern of analogous Cu2(F4bdc)2(dabco).130

6.3.3. On minimally treated Si Surprisingly, among the multitude of surface pre-treatment conditions tested, the best results were achieved either by simply washing the Si slides with eth-anol or by washing followed by mild sensitization with 1 mM ina. The GIXRD patterns recorded with films that received this treatment revealed SURMOFs almost exclusively oriented along the [001] direction (Figure 44).

64

Figure 44. GIXRD patterns of Cu2(bdc)2(dabco) SURMOF grown on mildly treated Si substrates in comparison to simulated PXRD pattern of analogous Cu2(F4bdc)2(dabco).130

To establish which of the two Si pre-treatment methods – sole washing with EtOH or that followed by ina sensitization – yielded the best-quality SUR-MOF, SEM micrographs of the resulting films were recorded and compared (Figure 45). To our surprise, although both SURMOFs indeed displayed a high degree of orientational preference, the film grown on Si that was simply washed with ethanol without any sensitization displayed much superior cov-erage and homogeneity as compared to the SURMOF grown on ina-sensitized Si slide. Increasing the number of cycles from 50 to 100 led to a significant increase in the undesired [100] orientation, which clearly indicates the need for further optimisation of the methodology in order to achieve both thickness and orientational control over the produced SURMOFs.

Figure 45. SEM micrographs of Cu2(bdc)2(dabco) SURMOF grown on Si (A) washed with ethanol and sensitized with ina; (B) washed only with ethanol. Re-printed with permission from ACS.130

65

At a first glance, the fact that Cu2(bdc)2(dabco) not only grows on untreated Si surface, but also achieves better coverage compared to surface-sensitized Si, may be surprising given the absence of Cu(II)-ligating molecules on the substrate. There are, however, reports of metal ion SBUs forming a donor-acceptor bond with oxygen lone pairs on the surface of quartz and metal ox-ides.134–136 We believe nucleation of the Cu2(bdc)2(dabco) on untreated Si sur-face happens by the same mechanism involving a native SiO2 layer.

6.3.4. Method refinement: surface coverage and orientational control While very promising, the methods described above still produced rough films with many pinholes and a significant amount of undesirable [100]-oriented crystals, especially when the number of cycles was increased to improve sur-face coverage or increase film thickness. Accordingly, our next goal was to refine the methodology to obtain smoother, more homogeneous films and fur-ther suppress growth of the [100] orientation. Reviewing the literature on M2L2P SURMOF growth, we found examples of a modified LbL procedure where the rinsing step after the very first SBU treatment is omitted,127,137 in-cluding one Cu2(bdc)2(dabco) SURMOF example.138 Hypothesizing that re-sidual Cu(II) ions might improve initial layer coverage, we reproduced our earlier established technique but without the ethanol soaking between the very first SBU and linker treatments. Comparing SEM images of the films resulting from 20 cycles of LbL Cu2(bdc)2(dabco) growth with and without the initial ethanol rinse indeed revealed a significantly decreased amount of [100]-ori-ented crystals (Figure 46), which appear as brighter thin vertical platelets.

Figure 46. SEM micrographs of Cu2(bdc)2(dabco) SURMOF after 20 cycles of growth on ethanol-washed Si (A) with the conventional double ethanol rinse be-tween each soaking step, and (B) omitting the very first ethanol rinsing step. Re-printed with permission from ACS.130

In principle, initial templating should unequivocally define the eventual crys-tal orientation of the film, as demonstrated by early SURMOF work.122,126,139

66

With that in mind, the persistent appearance of misaligned [100] crystals in M2L2P SURMOFs may indicate a general in situ crystal twinning mechanism, as proposed by Terfort et al.131 In their analysis, it was envisioned that miss-ing-linker defects might allow a 90° flipping of the Cu2 paddlewheel SBU around the bdc ligand axis, essentially locally priming the new [100] orienta-tion (Figure 47).

Figure 47. Missing linkers allow local 90° rotation of the Cu2 SBU, which may act as a primer for a new crystal orientation via crystal twinning. Inspired by work of Terfort et al.131 Reprinted with permission from ACS.130

Based on this analysis, we hypothesized that increasing the concentration of dabco or H2bdc linkers during the synthesis should limit the formation of miss-ing-linker defects and thus suppress crystal twinning. The validity of this hy-pothesis was investigated in two sets of SURMOFs that were prepared using the refined soaking sequence (omitting the very first EtOH rinse) and featuring altered linker concentrations: one with a fivefold increased dabco concentra-tion and another with a doubled H2bdc concentration. Although both proce-dures produced SURMOFs with high preference for [001] orientation as evi-denced by GIXRD, SEM analysis found massive differences in film quality. While increased dabco concentration resulted in rough and non-homogeneous surfaces with still many [100]-oriented crystals (Figure 48A), the increase in H2bdc concentration almost completely suppressed crystal twinning and re-sulted in a smooth, homogeneous film with nearly perfect coverage (Figure

67

48B). Encouraged by this result, we next sought to refine the method for grow-ing films of different thicknesses.

Figure 48. SEM micrographs of Cu2(bdc)2(dabco) SURMOF after 20 cycles of growth on ethanol-washed Si using (A) 5 times increased dabco concentration or (B) 2 times increased H2bdc concentration. Reprinted with permission from ACS.130

6.3.5. Film thickness As mentioned in the Introduction, film thickness will play an important role in the eventual performance of electrocatalytic SURMOFs. Therefore, we next attempted to control the thickness of the M2L2P MOF layer on Si. To this end, Cu2(bdc)2(dabco) was grown on Si using 10, 20, 35, 50, and 65 cycles, and the crystallinity was measured by GIXRD (Figure 49A). The full width at half maximum (fwhm) of corresponding [001] peaks were then extracted from the patterns and used to estimate film thickness using the Scherrer equation (see Chapter 2: Analytical methods).

68

Figure 49. (A) GIXRD patterns of Cu2(bdc)2(dabco) films grown with varying num-ber of cycles focused on the [001] peak region and (B) corresponding film thick-nesses calculated using the Scherrer equation. Reprinted with permission from ACS.130

In contrast to the expected linear expansion of the films with increasing num-ber of growth cycles, film thicknesses determined by the Scherrer equation reached a plateau around 28 nm after 35 growth cycles (Figure 49B). SEM micrographs of the resulting films showed that SURMOF growth up to 20 cycles completed surface coverage by filling pinholes (Figure 50A, B), but further growth up to 50 cycles resulted in increased numbers of [100]-oriented crystals (Figure 50C). Cross-sectional SEM analysis of the SURMOF grown by 50 cycles (Figure 50D) revealed average thickness of 70-100 nm – seem-ingly in contradiction to the plateauing film thickness derived from Scherrer analysis. However, close examination of the SEM images revealed that all films appeared to consist of discrete interlocking disc-shaped crystals with a limited maximum size, which explains why the thicknesses determined by the Scherrer analysis reached a plateau. Although the films might grow beyond 28 nm, further growth seems to occur by stacking the discrete domains instead of extending them.

69

Figure 50. SEM micrographs of Cu2(bdc)2(dabco) SURMOF films grown using the optimized procedure with (A) 10 cycles, showing incomplete surface coverage, (B) 20 cycles, showing complete surface coverage and interlocking disk-shaped crystal domains, and (C) 50 cycles, showing crystal domains and increased number of [100]-oriented crystals. (D) cross-sectional image of the latter SURMOF. Reprinted with permission from ACS.130

The abovementioned results indicate an apparent preference of this SURMOF to grow in stacks of small interlocked crystallites instead of forming larger crystals, which resonates with the fact that no single-crystal structure for Cu2(bdc)2(dabco) exists to-date. Considering the natural inclination toward crystal twinning, we believe this may be caused by the high tendency for de-fect formation in this SURMOF. In fact, X-ray Photoelectron Spectroscopy (XPS) analysis of Cu2(bdc)2(dabco) films grown both with and without the first ethanol rinse revealed the Cu:N ratio to be 1,8 – almost twice as high as the expected value of 1:1. This discrepancy suggests missing dabco linkers and/or trapped copper in the SURMOF.

Assuming an ideal LbL SURMOF growth mechanism, each cycle of Cu2(bdc)2(dabco) growth should extend the film by one layer (approx. 1 nm). However, cross-sectional SEM of the film grown using 50 cycles (vide supra) revealed a significantly thicker profile. We interpret this as resulting from ex-tra reactants being “stored” in the SURMOF pores even through the rinsing steps – a phenomenon with precedent.58,140 Such stored reagents will lead to uncontrolled growth beyond the one-layer-at-a-time growth that is intended by the LbL strategy.

70

6.4. Method extension: TiO2 substrate and isoreticular SURMOFs Having established the protocol for growing highly oriented Cu2(bdc)2(dabco) SURMOF on Si, we were interested to attempt the same synthesis using an-other common solid support, namely TiO2@Si. TiO2 is an important substrate to explore due to its utility for Si surface protection141,142 and blocking back electron transfer143 in device engineering. Thin TiO2 layers can conveniently be deposited on the underlying Si substrate by atomic layer deposition. The optimized growth procedure was tested on TiO2@Si substrates, some of which had been sensitized with ina. The resulting films exhibit GIXRD patterns that indicate successful [001] Cu2(bdc)2(dabco) film formation independent of whether the TiO2@Si substrates had been ina-sensitized or not (Figure 51).

Figure 51. GIXRD patterns of Cu2(bdc)2(dabco) grown with 50 cycles on treated and untreated TiO2@Si in comparison to simulated pattern of the analogous Cu2(F4bdc)2(dabco) SURMOF for orientation reference. Reprinted with permission from ACS.130

The universality of the developed methodology was further tested by applying the optimised growth protocol to two isoreticular MOFs with expanded unit cells along both the [100] and [001] directions: Cu2(1,4-ndc)2(dabco) (1,4-ndc = 1,4-naphthalene dicarboxylate) and Cu2(bdc)2(bipy) (bipy = 4,4′-bipyri-dine). We envision SURMOFs with linkers of these dimensions to be the ma-jor targets for incorporation of molecular catalysts.

71

Cu2(1,4-ndc)2(dabco) is a well-known SURMOF usually grown on SAM/Au substrates.144–146 Using the technique developed here, we were able to successfully produce a highly crystalline and [001]-oriented film on bare washed Si substrate (Figure 52A).

Figure 52. GIXRD patterns of isoreticular M2L2P SURMOFs grown using the devel-oped procedure: (A) Cu2(1,4-ndc)2(dabco), and (B) Cu2(bdc)2(bipy). Reprinted with permission from ACS.130

Cu2(bdc)2(bipy) is an interesting target not only by virtue of carrying the lig-and directly capable of coordinating metal ions, but also as a testbed for sup-pressing interpenetrated SURMOF growth, which is a known issue with link-ers of this size.147,148 Indeed, applying our developed growth technique, we were able to obtain a crystalline film of Cu2(bdc)2(bipy) (Figure 52B). Com-paring the obtained GIXRD pattern to simulated reference patterns, the reflec-tions closely match the [001] orientation of the EtOH-solvated crystal form, which was attributed to the non-interpenetrated topology.148 Furthermore, the [002] reflection appears slightly enhanced relative to the [001] peak, com-pared to intensities in the simulated pattern. For an analogous Zn2(bdc)2(bipy) SURMOF on SAM-coated Au substrate, this phenomenon was attributed to complete suppression of the interpenetrated phase growth.128 Together, these observations indicate the successful formation of non-interpenetrated Cu2(bdc)2(bipy) SURMOF film.

72

6.5. Conclusions & Outlook Surface-mounted MOFs is a class of materials that presents an excellent plat-form for practical implementation of MOF-based electrocatalysis. Canoni-cally, SURMOFs are grown almost exclusively on SAM-modified Au sur-faces, which enable precise orientational control and full surface coverage. In this chapter, we present the first systematic study of orientational control of a M2L2P SURMOF grown on bare Si substrate. Conditions producing selec-tively the [100] or the desired [001] orientation of the crystalline film with respect to the surface were investigated, and the synthetic protocol was refined to maximize surface coverage and uniformity of the [001] SURMOF film. Additionally, we explored the degree to which the film thickness can be con-trolled using the developed methodology. Several important mechanistic in-sights have been reached: 1. The Cu2(bdc)2(dabco) SURMOF can be primed for epitaxial [001] growth

on bare Si by coordinating the first Cu layer to oxygen lone pairs in the native SiO2 layer.

2. Crystal twinning resulting from Cu2 cluster flipping can be suppressed by minimizing missing bdc linker defects.

3. The Cu2(bdc)2(dabco) film grows as disk-shaped interlocked crystals lim-ited in size around 30 nm, likely indicative of high tendency for defect formation.

4. During the synthesis, part of the reactants is stored inside the framework pores, which results in thicker film than predicted by the number of growth cycles.

Although we briefly investigated the scope of our technique by successfully applying it to TiO2 substrate and two isoreticular SURMOFs with extended unit cells, more work is needed to fully assess its universality. Also, better control over film thickness needs to be developed before detailed kinetic anal-yses of electrocatalytic reactions can be performed.

73

7. Summary and Outlook

This thesis summarizes efforts toward solving various synthetic difficulties associated with the development of functional MOF platforms. More specifi-cally, the work is focused on the design of MOFs into which functional mol-ecules have been introduced as constituting linkers. The development of ex-perimental methods to characterize the spatial distribution of the linkers after their introduction has been addressed, and surface growth of MOFs on (photo)electrodes has be re-visited.

The preparation of molecularly doped MOFs can be synthetically challeng-ing and requires particular attention with respect to the functional molecule of interest, the target topology, and other factors. Different linker motifs may require different synthetic strategies for their incorporation into crystalline MOF materials. This issue was investigated in detail in Chapter 4 of this the-sis, where the incorporation of Ru(tda)(Py)2 water oxidation catalyst deriva-tives was investigated using various synthetic strategies. During the course of that work, post-synthetic modifications were found to be the most effective method for introducing a Ru(tda)-based ditopic linker into a UiO-type MOF. The tetratopic linker based on the same metal core could be successfully in-corporated using direct solvothermal synthesis with an inert co-linker to pro-duce a mixed linker version of NU-1000. In the UiO part of this work, an unexpected polymorphism in edba-MOFs was discovered, and the two domi-nating morphologies were thoroughly characterized, including the determina-tion of the factors that favour one morphology over the other. An investigation of the final UiO-based Ru-MOF products indicated that linker incorporation had proceeded mostly by filling missing-linker defects instead of exchanging structural linkers. Overall, the project has successfully solved a challenging synthetic task: the incorporation of two different variants of the catalytic Ru(tda)(Py)2 complex into different MOFs, and by different synthetic strate-gies. Going forward, the next steps in this project will include the activation of the complex for catalysis inside the MOF pores. Preliminary work done in this direction indicates that the task is challenging and may at least partially be complicated by the low intrinsic conductivity of the MOFs. Therefore, combining electrocatalytic linkers with linkers promoting electron hopping is a promising direction in which the project can develop, as well as tuning dif-fusion of water and counter-ions through the pores. In addition, for future ap-plications of electrocatalytic MOF platforms, an efficient MOF-electrode in-terface is needed – this issue is addressed in Chapter 6 of this thesis.

74

Post-synthetic introduction of functionality, as used in this thesis as an in-dispensable synthetic strategy, still poses many fundamental questions, such as: what is the depth distribution of the inserted molecules inside the final MOF and what are the determining factors? Do the exchanged species reside only near the crystal surface, throughout the framework, or concentrated in the core? Assessing this distribution is paramount for understanding and en-gineering the properties of functional hybrid MOFs, in particular those that engage in catalytic reactions.88 In Chapter 5 of this thesis, the microbeam RBS technique is introduced for non-destructive compositional depth profile anal-ysis inside single MOF crystals with unprecedented accuracy. The power of this technique is illustrated by investigating depth distribution of two model molecules that are post-synthetically introduced into a UiO-66 MOF. For this purpose, the method for growing individual surface-borne MOF crystals of controlled size and orientation was developed. The RBS measurements con-firmed previous findings by us and others, arriving at a picture where mono-topic ligands exchange throughout the crystal bulk, while the ditopic linker exchanged preferentially in the outer layers of the MOF crystal due to limiting diffusion. RBS results combined with bulk analysis of the digested MOFs doped with both molecules indicated that linker introduction occurs predomi-nantly by filling of defect sites, i.e., by the PSDH mechanism, in agreement with findings in Chapter 4. Looking into the future, this part of the thesis in-troduces an analytical tool that enables a variety of further MOF studies. Ac-curate and non-destructive depth profiling of large MOF crystals can be used for addressing intricate aspects of MOF chemistry such as framework defects and intra-crystal diffusion, as well as application-related practical issues such as depth distribution of functional molecules, leaching or linker scrambling.

Finally, as a step in a more practical-oriented direction, interfacing of MOFs with (photo)electrodes was considered, in particular, by growing the MOFs directly on electrode surfaces in an epitaxial fashion. In Chapter 6 of this thesis, Si surfaces are explored as an alternative and more sustainable sub-strate to the conventional SAM-coated Au surfaces for the growth of an M2L2P-type SURMOF. In the course of this work, a methodology for layer-by-layer growth of highly oriented, high-quality M2L2P SURMOF films has been developed and tested on several variations of the substrate and the MOF, including frameworks with extended unit cell. In summary, the project has resulted in a robust technique for growing layer-pillar-type SURMOFs on Si surfaces. The project can be developed further by elaboration of methods for controlling film thickness, as well as by using the resulting SURMOF@Si films in fundamental charge transport studies. Another promising avenue would be to grow layered SURMOFs in which different layers contain differ-ent linkers with unique functionalities. Incidentally, the RBS method intro-duced in Chapter 5 would be a perfect method for the characterization of such layered structures.

75

To summarize, the work presented in this thesis furthers our understanding of selected aspects of functional MOF platforms. It addresses different syn-thetic strategies to dope MOFs with functional molecules, the development of analytical methods to determine the spatial distribution of post-synthetically introduced molecules, and the controlled growth of MOFs on Si-based sub-strates. The author hopes that the results may contribute to the development of electrocatalytic MOFs for energy applications, or otherwise assist in the global effort to solve the energy and climate crises.

76

Acknowledgements

This PhD thesis would not have been possible without the help and support of my colleagues, friends, family, and many other people. Here, I would like to extend my gratitude to the following individuals:

Prof. Sascha Ott for his guidance, support, and genuine care. For the phe-nomenal group he has built. Put simply, I cannot think of a better supervisor for my PhD studies.

My co-supervisor Prof. Reiner Lomoth for help and guidance in the early stages of my PhD.

Dr. Brian McCarthy for patiently helping me construct this thesis and other manuscripts. For inviting me to participate in the SURMOF project, which led to Paper IV. For being a great colleague and a great friend.

Prof. Antoni Llobet for setting up a collaboration between our groups, which eventually led to Papers I and II.

Dr. Carolina Gimbert-Suriñach for help in preparing manuscript for Paper I.

Andrew Howe for sharing a joint project with me, which resulted in Papers I and II.

Prof. Daniel Primetzhofer for his enthusiasm and drive to build a bridge be-tween the departments of physics and chemistry. For his help in our joint pro-ject and preparing the manuscript for Paper III.

Dr. Mauricio Sortica for his help with the RBS measurements and their in-terpretation that created the basis for Paper III.

Dr. Valentina Paneta for the help in early steps of the RBS project, adapta-tion of the ion beam system to our MOF platform, and help in preparing the manuscript for Paper III.

77

Dr. Victoria Stenhagen for introducing me to the cleanroom facilties and teaching me the operation of SEM.

Sven Johansson for making sure the House 7 labs are operating smoothly.

Prof. Leif Hammarström for his enthusiasm in leading the CAP as well as departmental fikas, great discussions, and being a role model for students.

My friend and colleague Dr. Kjell Jorner for correcting the Swedish sum-mary and for the amazing discussions and informal lessons in Swedish lan-guage we had during our joint lunches.

My friend and colleague Dr. Brigitta Nemeth for the support and all the bril-liant nerdy discussions we held during lunchtime.

My friend and colleague Kateryna Kukil for being a good comrade and great lunch buddy.

The Otters! Dr. Anna Arkhypchuk, Dr. Anna Beiler, Ashleigh Castner, Dr. Ben Johnson, Dr. Hemlata Agarwala, Dr. Maxime Laurans, Dr. Ni-colas D’Imperio, Dr. Juri Mai, Dr. Michele Bedin, and others – for the most valuable inputs and discussions, and for being awesome colleagues!

My brother Mikhail Liseev for shaping me into the person I am. For under-standing me like no one else can.

My parents Nadezhda and Yuri for their infinite support and care.

Dr. Tatiana Alexandrovna Kotelnikova for lighting the interest to chemistry in my heart when I was 15.

Anna Soloveva for her help, support, and honest criticism.

My daughter Larisa for being an eternal source of love and happiness in my life.

78

Populärvetenskaplig sammanfattning på Svenska

Den moderna världen står inför en av historiens största utmaningar. Männi-skans aktivitet förändrar planetens klimat, och inom nästa århundrade kan dessa förändringar bli oåterkalleliga. Den främsta orsaken är den växande an-vändningen av fossila energikällor. Sökandet efter alternativa energikällor är en av de centrala frågorna inom den moderna vetenskapen. Ett av de mest lovande bränslena är vätgas. Produktion av vätgas genom spjälkning av vatten är ett särskilt lovande tillvägagångssätt. Eftersom reaktionen under vanliga förhållanden är extremt långsam kräver denna process effektiva katalysatorer. Katalysatorer kan grovt delas in i två kategorier: (a) materialbaserade och (b) molekylära. De sistnämnda är ofta billiga och mer aktiva, men har tyvärr lägre stabilitet.

För att öka stabiliteten kan molekylära katalysatorer kombineras med fasta material. Denna avhandling beskriver ett sådant tillvägagångssätt, där materi-alen är så kallade metall-organiska ramverk (metal-organic frameworks, MOF). Dessa ramverk är kristaller med extremt högt porositet som består av oorganiska "knutar" kopplade av organiska "stag" på ungefär samma sätt som byggnadsställningar. Katalysatorerna installeras i ramverket som "stag" vilket ger hybrida MOFs.

I den första delen av avhandlingen undersöks olika syntetiska metoder för att skapa sådana hybrida MOFs som innehåller en ruteniumbaserad katalysa-tor för vattenoxidation. Detta sker bl.a. genom en process som kallas postsyn-tetiskt utbyte (post-synthetic exchange, PSE) där stag som innehåller kataly-satorn ersätter de stag eller andra ämnen som fanns i den ursprungliga MOF:en. I den andra delen undersöks hur djupt i MOF-kristallen de molekyler som introduceras genom PSE processen kan installeras. För detta ändamål an-vänds en analytisk teknik baserad på jon-spridning med hög energi (Ruther-ford backscattering spectrometry, RBS). I den sista delen av avhandlingen ut-vecklas en metod för lager-på-lager syntes av MOFs med en hög grad av kon-troll över ramverkets orientering och tjocklek på kiselytor, vilket ger en pris-värd och perspektiv arbetselektrod.

79

Научно-популярное резюме по-русски

Современный мир сталкивается с одним из величайших испытаний в ис-тории: человеческая деятельность изменяет климат планеты, и уже в те-чение ближайшего столетия эти изменения могут приобрести необрати-мый характер. Главная причина – растущие темпы потребления ископа-емых источников энергии. Поиск альтернативных источников энергии есть один из центральных вопросов современной науки. Один из самых перспективных видов топлива – водород. Получение водорода посред-ством расщепления воды составляет горячайший вопрос в современной химии. Будучи чрезвычайно кинетически затруднённым, этот процесс требует эффективных катализаторов. Катализаторы можно грубо разде-лить на две категории: а) на основе материалов и б) молекулярные. По-следние зачастую более активны и доступны, однако страдают от низкой стабильности.

Чтобы повысить срок службы, молекулярные катализаторы можно совместить с твёрдым материалом. В данной работе описывается такой подход, где материалом служат т.н. металлоорганические каркасы – чрезвычайно высокопористые материалы, состоящие из неорганических «узлов», связанных органическими «распорками» подобно строитель-ным лесам. Катализаторы устанавливаются в качестве структурных «распорок».

В первой части данной работы исследуются различные синтетические подходы к созданию таких гибридных каркасов, несущих в себе катали-затор окисления воды на основе рутения. Во второй части работы иссле-дуется, насколько глубоко в кристалл каркаса могут проникать вноси-мые пост-синтетически молекулы, для чего вводится новая для этого направления аналитическая техника – спектроскопия резерфордовского обратного рассеяния. В заключающей части работы исследуется послой-ный эпитаксиальный синтез каркасов на поверхности кремния – доступ-ного и перспективного рабочего электрода – с высокой степенью кон-троля над кристаллической ориентацией и толщиной конечной плёнки каркаса.

80

References

(1) Ceballos, G.; Ehrlich, P. R.; Barnosky, A. D.; García, A.; Pringle, R. M.; Palmer, T. M. Accelerated Modern Human-Induced Species Losses: Entering the Sixth Mass Extinction. Sci. Adv. 2015, 1 (5), 9–13. https://doi.org/10.1126/sciadv.1400253.

(2) Ruddiman, W. F. The Anthropogenic Greenhouse Era Began Thousands of Years Ago. Clim. Change 2003, 61 (3), 261–293. https://doi.org/10.1023/B:CLIM.0000004577.17928.fa.

(3) NOAA Earth System Research Laboratory, R/GMD, 325 Broadway, Boulder, C. 80305-3328. THE NOAA ANNUAL GREENHOUSE GAS INDEX (AGGI) https://www.esrl.noaa.gov/gmd/aggi/aggi.html.

(4) Dziba, L.; Erpul, G.; Fazel, A.; Fischer, M.; Hernández, A. M. Summary for Policymakers of the Global Assessment Report on Biodiversity and Ecosystem Services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Service; 2019.

(5) Kardashev, N. S. Transmission of Information by Extraterrestrial Civilizations. Astron. Zhurnal 1964, 41 (2), 282–278.

(6) Steffen, W.; Rockström, J.; Richardson, K.; Lenton, T. M.; Folke, C.; Liverman, D.; Summerhayes, C. P.; Barnosky, A. D.; Cornell, S. E.; Crucifix, M.; Donges, J. F.; Fetzer, I.; Lade, S. J.; Scheffer, M.; Winkelmann, R.; Schellnhuber, H. J. Trajectories of the Earth System in the Anthropocene. Proceedings of the National Academy of Sciences of the United States of America. National Academy of Sciences August 14, 2018, pp 8252–8259. https://doi.org/10.1073/pnas.1810141115.

(7) Owusu, P. A.; Asumadu-Sarkodie, S. A Review of Renewable Energy Sources, Sustainability Issues and Climate Change Mitigation. Cogent Eng. 2016, 3 (1), 1–14. https://doi.org/10.1080/23311916.2016.1167990.

(8) (UNECE), U. N. E. C. for E. Pathways to Sustainable Energy – Accelerating Energy Transition in the UNECE Region; 2017; Vol. 41.

(9) Matheu, R.; Garrido-Barros, P.; Gil-Sepulcre, M.; Ertem, M. Z.; Sala, X.; Gimbert-Suriñach, C.; Llobet, A. The Development of Molecular Water Oxidation Catalysts. Nat. Rev. Chem. 2019, 3 (5), 331–341. https://doi.org/10.1038/s41570-019-0096-0.

(10) Zhang, B.; Sun, L. Artificial Photosynthesis: Opportunities and Challenges of Molecular Catalysts. Chemical Society Reviews. Royal Society of Chemistry 2019, pp 2216–2264. https://doi.org/10.1039/c8cs00897c.

(11) Andreiadis, E. S.; Chavarot-Kerlidou, M.; Fontecave, M.; Artero, V. Artificial Photosynthesis: From Molecular Catalysts for Light-Driven Water Splitting to Photoelectrochemical Cells. Photochem. Photobiol. 2011, 87 (5), 946–964. https://doi.org/10.1111/j.1751-1097.2011.00966.x.

81

(12) Berardi, S.; Drouet, S.; Francàs, L.; Gimbert-Suriñach, C.; Guttentag, M.; Richmond, C.; Stoll, T.; Llobet, A. Molecular Artificial Photosynthesis. Chem. Soc. Rev. 2014, 43 (22), 7501–7519. https://doi.org/10.1039/C3CS60405E.

(13) Bai, Y.; Dou, Y.; Xie, L. H.; Rutledge, W.; Li, J. R.; Zhou, H. C. Zr-Based Metal-Organic Frameworks: Design, Synthesis, Structure, and Applications. Chem. Soc. Rev. 2016, 45 (8), 2327–2367. https://doi.org/10.1039/c5cs00837a.

(14) Sumby, C. MOF the chart: why a record-breaking surface area matters https://theconversation.com/mof-the-chart-why-a-record-breaking-surface-area-matters-9915.

(15) Liu, J.; Chen, L.; Cui, H.; Zhang, J.; Zhang, L.; Su, C. Y. Applications of Metal-Organic Frameworks in Heterogeneous Supramolecular Catalysis. Chem. Soc. Rev. 2014, 43 (16), 6011–6061. https://doi.org/10.1039/c4cs00094c.

(16) Bhunia, A.; Johnson, B. A.; Czapla-Masztafiak, J.; Sá, J.; Ott, S. Formal Water Oxidation Turnover Frequencies from MIL-101(Cr) Anchored Ru(Bda) Depend on Oxidant Concentration. Chem. Commun. 2018, 54 (56), 7770–7773. https://doi.org/10.1039/C8CC02300J.

(17) Wu, C.-D.; Zhao, M. Incorporation of Molecular Catalysts in Metal-Organic Frameworks for Highly Efficient Heterogeneous Catalysis. Adv. Mater. 2017, 29 (14), 1605446. https://doi.org/10.1002/adma.201605446.

(18) Pullen, S.; Fei, H.; Orthaber, A.; Cohen, S. M.; Ott, S. Enhanced Photochemical Hydrogen Production by a Molecular Diiron Catalyst Incorporated into a Metal–Organic Framework. J. Am. Chem. Soc. 2013, 135 (45), 16997–17003. https://doi.org/10.1021/ja407176p.

(19) Downes, C. A.; Marinescu, S. C. Electrocatalytic Metal-Organic Frameworks for Energy Applications. ChemSusChem 2017, 10 (22), 4374–4392. https://doi.org/10.1002/cssc.201701420.

(20) Kim, S.; Lee, J.; Jeoung, S.; Moon, H. R.; Kim, M. Surface-Deactivated Core–Shell Metal–Organic Framework by Simple Ligand Exchange for Enhanced Size Discrimination in Aerobic Oxidation of Alcohols. Chem. - A Eur. J. 2020, 26 (34), 7568–7572. https://doi.org/10.1002/chem.202000933.

(21) Johnson, B. A.; Beiler, A. M.; McCarthy, B. D.; Ott, S. Transport Phenomena: Challenges and Opportunities for Molecular Catalysis in Metal-Organic Frameworks. J. Am. Chem. Soc. 2020, 142 (28), 11941–11956. https://doi.org/10.1021/jacs.0c02899.

(22) Lin, S.; Usov, P. M.; Morris, A. J. The Role of Redox Hopping in Metal–Organic Framework Electrocatalysis. 2018. https://doi.org/10.1039/C8CC01664J.

(23) Kung, C.; Goswami, S.; Hod, I.; Wang, T. C.; Duan, J.; Farha, O. K.; Hupp, J. T. Charge Transport in Zirconium-Based Metal − Organic Frameworks. 2020. https://doi.org/10.1021/acs.accounts.0c00106.

(24) Hausdorf, S.; Baitalow, F.; Seidel, J.; Mertens, F. O. R. L. Gaseous Species as Reaction Tracers in the Solvothermal Synthesis of the Zinc Oxide Terephthalate MOF-5. J. Phys. Chem. A 2007, 111 (20), 4259–4266. https://doi.org/10.1021/jp0708291.

(25) Schaate, A.; Roy, P.; Godt, A.; Lippke, J.; Waltz, F.; Wiebcke, M.; Behrens, P. Modulated Synthesis of Zr-Based Metal-Organic Frameworks: From Nano to Single Crystals. Chem. - A Eur. J. 2011, 17 (24), 6643–6651. https://doi.org/10.1002/chem.201003211.

82

(26) Ren, J.; Langmi, H. W.; North, B. C.; Mathe, M.; Bessarabov, D. Modulated Synthesis of Zirconium-Metal Organic Framework (Zr-MOF) for Hydrogen Storage Applications. Int. J. Hydrogen Energy 2014, 39 (2), 890–895. https://doi.org/10.1016/j.ijhydene.2013.10.087.

(27) Tsuruoka, T.; Furukawa, S.; Takashima, Y.; Yoshida, K.; Isoda, S.; Kitagawa, S. Nanoporous Nanorods Fabricated by Coordination Modulation and Oriented Attachment Growth. Angew. Chemie Int. Ed. 2009, 48 (26), 4739–4743. https://doi.org/10.1002/anie.200901177.

(28) Diring, S.; Furukawa, S.; Takashima, Y.; Tsuruoka, T.; Kitagawa, S. Controlled Multiscale Synthesis of Porous Coordination Polymer in Nano/Micro Regimes. Chem. Mater. 2010, 22 (16), 4531–4538. https://doi.org/10.1021/cm101778g.

(29) Kirchon, A.; Feng, L.; Drake, H. F.; Joseph, E. A.; Zhou, H.-C. From Fundamentals to Applications: A Toolbox for Robust and Multifunctional MOF Materials. Chem. Soc. Rev. 2018, 47 (23), 8611–8638. https://doi.org/10.1039/C8CS00688A.

(30) Marreiros, J.; Caratelli, C.; Hajek, J.; Krajnc, A.; Fleury, G.; Bueken, B.; De Vos, D. E.; Mali, G.; Roeffaers, M. B. J.; Van Speybroeck, V.; Ameloot, R. Active Role of Methanol in Post-Synthetic Linker Exchange in the Metal–Organic Framework UiO-66. Chem. Mater. 2019, 31 (4), 1359–1369. https://doi.org/10.1021/acs.chemmater.8b04734.

(31) Shearer, G. C.; Chavan, S.; Ethiraj, J.; Vitillo, J. G.; Svelle, S.; Olsbye, U.; Lamberti, C.; Bordiga, S.; Lillerud, K. P. Tuned to Perfection: Ironing Out the Defects in Metal–Organic Framework UiO-66. Chem. Mater. 2014, 26 (14), 4068–4071. https://doi.org/10.1021/cm501859p.

(32) Bárcia, P. S.; Guimarães, D.; Mendes, P. A. P.; Silva, J. A. C.; Guillerm, V.; Chevreau, H.; Serre, C.; Rodrigues, A. E. Reverse Shape Selectivity in the Adsorption of Hexane and Xylene Isomers in MOF UiO-66. Microporous Mesoporous Mater. 2011, 139 (1–3), 67–73. https://doi.org/10.1016/j.micromeso.2010.10.019.

(33) Cubillas, P.; Anderson, M. W.; Attfield, M. P. Crystal Growth Mechanisms and Morphological Control of the Prototypical Metal-Organic Framework MOF-5 Revealed by Atomic Force Microscopy. Chem. - A Eur. J. 2012, 18 (48), 15406–15415. https://doi.org/10.1002/chem.201202261.

(34) Gutov, O. V.; Hevia, M. G.; Escudero-Adán, E. C.; Shafir, A. Metal–Organic Framework (MOF) Defects under Control: Insights into the Missing Linker Sites and Their Implication in the Reactivity of Zirconium-Based Frameworks. Inorg. Chem. 2015, 54 (17), 8396–8400. https://doi.org/10.1021/acs.inorgchem.5b01053.

(35) Valenzano, L.; Civalleri, B.; Chavan, S.; Bordiga, S.; Nilsen, M. H.; Jakobsen, S.; Lillerud, K. P.; Lamberti, C. Disclosing the Complex Structure of UiO-66 Metal Organic Framework: A Synergic Combination of Experiment and Theory. Chem. Mater. 2011, 23 (7), 1700–1718. https://doi.org/10.1021/cm1022882.

(36) Trickett, C. A.; Gagnon, K. J.; Lee, S.; Gándara, F.; Bürgi, H.-B.; Yaghi, O. M. Definitive Molecular Level Characterization of Defects in UiO-66 Crystals. Angew. Chemie Int. Ed. 2015, 54 (38), 11162–11167. https://doi.org/10.1002/anie.201505461.

(37) Shearer, G. C.; Chavan, S.; Bordiga, S.; Svelle, S.; Olsbye, U.; Lillerud, K. P. Defect Engineering: Tuning the Porosity and Composition of the Metal-Organic Framework UiO-66 via Modulated Synthesis. Chem. Mater. 2016, 28 (11), 3749–3761. https://doi.org/10.1021/acs.chemmater.6b00602.

83

(38) Liu, L.; Chen, Z.; Wang, J.; Zhang, D.; Zhu, Y.; Ling, S.; Huang, K.-W.; Belmabkhout, Y.; Adil, K.; Zhang, Y.; Slater, B.; Eddaoudi, M.; Han, Y. Imaging Defects and Their Evolution in a Metal–Organic Framework at Sub-Unit-Cell Resolution. Nat. Chem. 2019, 1. https://doi.org/10.1038/s41557-019-0263-4.

(39) Wang, K.; Li, C.; Liang, Y.; Han, T.; Huang, H.; Yang, Q.; Liu, D.; Zhong, C. Rational Construction of Defects in a Metal-Organic Framework for Highly Efficient Adsorption and Separation of Dyes. Chem. Eng. J. 2016, 289, 486–493. https://doi.org/10.1016/j.cej.2016.01.019.

(40) Taddei, M.; Wakeham, R. J.; Koutsianos, A.; Andreoli, E.; Barron, A. R. Post-Synthetic Ligand Exchange in Zirconium-Based Metal-Organic Frameworks: Beware of The Defects! Angew. Chemie Int. Ed. 2018, 57 (36), 11706–11710. https://doi.org/10.1002/anie.201806910.

(41) Ren, J.; Ledwaba, M.; Musyoka, N. M.; Langmi, H. W.; Mathe, M.; Liao, S.; Pang, W. Structural Defects in Metal–Organic Frameworks (MOFs): Formation, Detection and Control towards Practices of Interests. Coord. Chem. Rev. 2017, 349, 169–197. https://doi.org/10.1016/j.ccr.2017.08.017.

(42) Cohen, S. M. The Postsynthetic Renaissance in Porous Solids. J. Am. Chem. Soc. 2017, 139 (8), 2855–2863. https://doi.org/10.1021/jacs.6b11259.

(43) Marshall, R. J.; Forgan, R. S. Postsynthetic Modification of Zirconium Metal-Organic Frameworks. Eur. J. Inorg. Chem. 2016, 2016 (27), 4310–4331. https://doi.org/10.1002/ejic.201600394.

(44) Yuan, S.; Lu, W.; Chen, Y.-P.; Zhang, Q.; Liu, T.-F.; Feng, D.; Wang, X.; Qin, J.; Zhou, H.-C. Sequential Linker Installation: Precise Placement of Functional Groups in Multivariate Metal–Organic Frameworks. J. Am. Chem. Soc. 2015, 137 (9), 3177–3180. https://doi.org/10.1021/ja512762r.

(45) Kusaka, S.; Matsuda, R.; Kitagawa, S. Generation of Thiyl Radicals in a Zinc( <scp>ii</Scp> ) Porous Coordination Polymer by Light-Induced Post-Synthetic Deprotection. Chem. Commun. 2018, 54 (38), 4782–4785. https://doi.org/10.1039/C8CC01837E.

(46) Wu, C.; Irshad, F.; Luo, M.; Zhao, Y.; Ma, X.; Wang, S. Ruthenium Complexes Immobilized on an Azolium Based Metal Organic Framework for Highly Efficient Conversion of CO 2 into Formic Acid. ChemCatChem 2019, 11 (4), 1256–1263. https://doi.org/10.1002/cctc.201801701.

(47) Kim, M.; Cahill, J. F.; Su, Y.; Prather, K. A.; Cohen, S. M. Postsynthetic Ligand Exchange as a Route to Functionalization of ‘Inert’ Metal–Organic Frameworks. Chem. Sci. 2012, 3 (1), 126–130. https://doi.org/10.1039/C1SC00394A.

(48) Takaishi, S.; Demarco, E. J.; Pellin, M. J.; Farha, O. K.; Hupp, J. T. Solvent-Assisted Linker Exchange (SALE) and Post-Assembly Metallation in Porphyrinic Metal-Organic Framework Materials. Chem. Sci. 2013, 4 (4), 1509–1513. https://doi.org/10.1039/c2sc21516k.

(49) Taddei, M. When Defects Turn into Virtues: The Curious Case of Zirconium-Based Metal-Organic Frameworks. Coord. Chem. Rev. 2017, 343 (April), 1–24. https://doi.org/10.1016/j.ccr.2017.04.010.

(50) Shearer, G. C.; Vitillo, J. G.; Bordiga, S.; Svelle, S.; Olsbye, U.; Lillerud, K. P. Functionalizing the Defects: Postsynthetic Ligand Exchange in the Metal Organic Framework UiO-66. Chem. Mater. 2016, 28 (20), 7190–7193. https://doi.org/10.1021/acs.chemmater.6b02749.

(51) Dodson, R. A.; Kalenak, A. P.; Matzger, A. J. Solvent Choice in Metal − Organic Framework Linker Exchange Permits Microstructural Control. 2020. https://doi.org/10.1021/jacs.0c10224.

84

(52) Koh, K.; Wong-Foy, A. G.; Matzger, A. J. MOF@MOF: Microporous Core–Shell Architectures. Chem. Commun. 2009, 0 (41), 6162. https://doi.org/10.1039/b904526k.

(53) Fluch, U.; Paneta, V.; Primetzhofer, D.; Ott, S. Uniform Distribution of Post-Synthetic Linker Exchange in Metal–Organic Frameworks Revealed by Rutherford Backscattering Spectrometry. Chem. Commun. 2017, 53 (48), 6516–6519. https://doi.org/10.1039/C7CC02631E.

(54) Boissonnault, J. A.; Wong-Foy, A. G.; Matzger, A. J. Core–Shell Structures Arise Naturally During Ligand Exchange in Metal–Organic Frameworks. J. Am. Chem. Soc. 2017, 139 (42), 14841–14844. https://doi.org/10.1021/jacs.7b08349.

(55) Maza, W. A.; Padilla, R.; Morris, A. J. Concentration Dependent Dimensionality of Resonance Energy Transfer in a Postsynthetically Doped Morphologically Homologous Analogue of UiO-67 MOF with a Ruthenium(II) Polypyridyl Complex. J. Am. Chem. Soc. 2015, 137 (25), 8161–8168. https://doi.org/10.1021/jacs.5b03071.

(56) Gamage, N. D. H.; McDonald, K. A.; Matzger, A. J. MOF-5-Polystyrene: Direct Production from Monomer, Improved Hydrolytic Stability, and Unique Guest Adsorption. Angew. Chemie - Int. Ed. 2016, 55 (39), 12099–12103. https://doi.org/10.1002/anie.201606926.

(57) Bétard, A.; Fischer, R. A. Metal–Organic Framework Thin Films: From Fundamentals to Applications. Chem. Rev. 2012, 112 (2), 1055–1083. https://doi.org/10.1021/cr200167v.

(58) Arslan, H. K.; Shekhah, O.; Wohlgemuth, J.; Franzreb, M.; Fischer, R. A.; Wöll, C. High-Throughput Fabrication of Uniform and Homogenous MOF Coatings. Adv. Funct. Mater. 2011, 21 (22), 4228–4231. https://doi.org/10.1002/adfm.201101592.

(59) McCarthy, B. D.; Beiler, A. M.; Johnson, B. A.; Liseev, T.; Castner, A. T.; Ott, S. Analysis of Electrocatalytic Metal-Organic Frameworks. Coord. Chem. Rev. 2020, 406, 213137. https://doi.org/10.1016/j.ccr.2019.213137.

(60) Kalidindi, S. B.; Nayak, S.; Briggs, M. E.; Jansat, S.; Katsoulidis, A. P.; Miller, G. J.; Warren, J. E.; Antypov, D.; Corà, F.; Slater, B.; Prestly, M. R.; Martí-Gastaldo, C.; Rosseinsky, M. J. Chemical and Structural Stability of Zirconium-Based Metal-Organic Frameworks with Large Three-Dimensional Pores by Linker Engineering. Angew. Chemie Int. Ed. 2015, 54 (1), 221–226. https://doi.org/10.1002/anie.201406501.

(61) Kumar, K.; Maddila, S.; Babu, S.; Jonnalagadda, S. B. Characteristics of MOF , MWCNT and Graphene Containing Materials for Hydrogen Storage : A Review. J. Energy Chem. 2019, 30, 132–144. https://doi.org/10.1016/j.jechem.2018.04.012.

(62) Elgrishi, N.; Rountree, K. J.; McCarthy, B. D.; Rountree, E. S.; Eisenhart, T. T.; Dempsey, J. L. A Practical Beginner’s Guide to Cyclic Voltammetry. J. Chem. Educ. 2018, 95 (2), 197–206. https://doi.org/10.1021/acs.jchemed.7b00361.

(63) Geiger, H.; Marsden, E. The Laws of Deflexion of α Particles through Large Angles. London, Edinburgh, Dublin Philos. Mag. J. Sci. 1913, 25 (148), 604–623. https://doi.org/10.1080/14786440408634197.

(64) Rutherford, E. The Scattering of α and β Particles by Matter and the Structure of the Atom. London, Edinburgh, Dublin Philos. Mag. J. Sci. 1911, 21 (125), 669–688. https://doi.org/10.1080/14786440508637080.

85

(65) Rubin, S.; Passell, T. O.; Bailey, L. E. Chemical Analysis of Surfaces by Nuclear Methods. Anal. Chem. 1957, 29 (5), 736–743. https://doi.org/10.1021/ac60125a001.

(66) Vadrucci, M.; Mazzinghi, A.; Sorrentino, B.; Falzone, S.; Gioia, C.; Gioia, P.; Loreti, E. M.; Chiari, M. Characterisation of Ancient Roman Wall‐painting Fragments Using Non‐destructive <scp>IBA</Scp> and <scp>MA‐XRF</Scp> Techniques. X-Ray Spectrom. 2020, 49 (6), 668–678. https://doi.org/10.1002/xrs.3178.

(67) Vilarigues, M.; Da Silva, R. C. Ion Beam and Infrared Analysis of Medieval Stained Glass. Appl. Phys. A Mater. Sci. Process. 2004, 79 (2), 373–378. https://doi.org/10.1007/s00339-004-2538-9.

(68) Santos, H. C.; Added, N.; Silva, T. F.; Rodrigues, C. L. External-RBS, PIXE and NRA Analysis for Ancient Swords. Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 2015, 345, 42–47. https://doi.org/10.1016/j.nimb.2014.12.076.

(69) Hou, X.; T. Jones, B. Inductively Coupled Plasma/Optical Emission Spectrometry. In Encyclopedia of Analytical Chemistry; Meyers, R. A., Ed.; John Wiley & Sons Ltd: Chichester, 2000; pp 9468–9485.

(70) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T. Metal-Organic Framework Materials as Catalysts. Chem. Soc. Rev. 2009, 38 (5), 1450–1459. https://doi.org/10.1039/b807080f.

(71) Zhang, T.; Lin, W. Metal-Organic Frameworks for Artificial Photosynthesis and Photocatalysis. Chemical Society Reviews. Royal Society of Chemistry August 21, 2014, pp 5982–5993. https://doi.org/10.1039/c4cs00103f.

(72) Bard, A.; Fox, M. A. Artificial Photosynthesis: Solar Splitting of Water to Hydrogen and Oxygen. Acc. Chem. Res. 1995, 28, 141–145. https://doi.org/10.1021/ar00051a007.

(73) Micheroni, D.; Lan, G.; Lin, W. Efficient Electrocatalytic Proton Reduction with Carbon Nanotube-Supported Metal–Organic Frameworks. J. Am. Chem. Soc. 2018, 140 (46), 15591–15595. https://doi.org/10.1021/jacs.8b09521.

(74) Roy, S.; Huang, Z.; Bhunia, A.; Castner, A.; Gupta, A. K.; Zou, X.; Ott, S. Electrocatalytic Hydrogen Evolution from a Cobaloxime-Based Metal–Organic Framework Thin Film. J. Am. Chem. Soc. 2019, 141 (40), 15942–15950. https://doi.org/10.1021/jacs.9b07084.

(75) Wang, C.; Xie, Z.; DeKrafft, K. E.; Lin, W. Doping Metal–Organic Frameworks for Water Oxidation, Carbon Dioxide Reduction, and Organic Photocatalysis. J. Am. Chem. Soc. 2011, 133 (34), 13445–13454. https://doi.org/10.1021/ja203564w.

(76) Wang, W.; Xu, X.; Zhou, W.; Shao, Z. Recent Progress in Metal-Organic Frameworks for Applications in Electrocatalytic and Photocatalytic Water Splitting. Adv. Sci. 2017, 4 (4), 1600371. https://doi.org/10.1002/advs.201600371.

(77) Kung, C. W.; Mondloch, J. E.; Wang, T. C.; Bury, W.; Hoffeditz, W.; Klahr, B. M.; Klet, R. C.; Pellin, M. J.; Farha, O. K.; Hupp, J. T. Metal-Organic Framework Thin Films as Platforms for Atomic Layer Deposition of Cobalt Ions to Enable Electrocatalytic Water Oxidation. ACS Appl. Mater. Interfaces 2015, 7 (51), 28223–28230. https://doi.org/10.1021/acsami.5b06901.

(78) Lu, X.-F.; Liao, P.-Q.; Wang, J.-W.; Wu, J.-X.; Chen, X.-W.; He, C.-T.; Zhang, J.-P.; Li, G.-R.; Chen, X.-M. An Alkaline-Stable, Metal Hydroxide Mimicking Metal–Organic Framework for Efficient Electrocatalytic Oxygen Evolution. J. Am. Chem. Soc. 2016, 138 (27), 8336–8339. https://doi.org/10.1021/jacs.6b03125.

86

(79) Shen, J.-Q.; Liao, P.-Q.; Zhou, D.-D.; He, C.-T.; Wu, J.-X.; Zhang, W.-X.; Zhang, J.-P.; Chen, X.-M. Modular and Stepwise Synthesis of a Hybrid Metal–Organic Framework for Efficient Electrocatalytic Oxygen Evolution. J. Am. Chem. Soc. 2017, 139 (5), 1778–1781. https://doi.org/10.1021/jacs.6b12353.

(80) Usov, P. M.; Ahrenholtz, S. R.; Maza, W. A.; Stratakes, B.; Epley, C. C.; Kessinger, M. C.; Zhu, J.; Morris, A. J. Cooperative Electrochemical Water Oxidation by Zr Nodes and Ni-Porphyrin Linkers of a PCN-224 MOF Thin Film. J. Mater. Chem. A 2016, 4 (43), 16818–16823. https://doi.org/10.1039/c6ta05877a.

(81) Usov, P. M.; Huffman, B.; Epley, C. C.; Kessinger, M. C.; Zhu, J.; Maza, W. A.; Morris, A. J. Study of Electrocatalytic Properties of Metal-Organic Framework PCN-223 for the Oxygen Reduction Reaction. ACS Appl. Mater. Interfaces 2017, 9 (39), 33539–33543. https://doi.org/10.1021/acsami.7b01547.

(82) Lin, S.; Pineda-Galvan, Y.; Maza, W. A.; Epley, C. C.; Zhu, J.; Kessinger, M. C.; Pushkar, Y.; Morris, A. J. Electrochemical Water Oxidation by a Catalyst-Modified Metal-Organic Framework Thin Film. ChemSusChem 2017, 10 (3), 514–522. https://doi.org/10.1002/cssc.201601181.

(83) Lin, S.; Ravari, A. K.; Zhu, J.; Usov, P. M.; Cai, M.; Ahrenholtz, S. R.; Pushkar, Y.; Morris, A. J. Insight into Metal-Organic Framework Reactivity: Chemical Water Oxidation Catalyzed by a [Ru(Tpy)(Dcbpy)(OH 2 )] 2+ -Modified UiO-67. ChemSusChem 2018, 11 (2), 464–471. https://doi.org/10.1002/cssc.201701644.

(84) Jia, H.; Yao, Y.; Zhao, J.; Gao, Y.; Luo, Z.; Du, P. A Novel Two-Dimensional Nickel Phthalocyanine-Based Metal–Organic Framework for Highly Efficient Water Oxidation Catalysis. J. Mater. Chem. A 2018, 6 (3), 1188–1195. https://doi.org/10.1039/C7TA07978H.

(85) Johnson, B. A.; Bhunia, A.; Ott, S. Electrocatalytic Water Oxidation by a Molecular Catalyst Incorporated into a Metal–Organic Framework Thin Film. Dalt. Trans. 2017, 46 (5), 1382–1388. https://doi.org/10.1039/C6DT03718F.

(86) Natarajan, K.; Gupta, A. K.; Ansari, S. N.; Saraf, M.; Mobin, S. M. Mixed-Ligand-Architected 2D Co(II)-MOF Expressing a Novel Topology for an Efficient Photoanode for Water Oxidation Using Visible Light. ACS Appl. Mater. Interfaces 2019, 11 (14), 13295–13303. https://doi.org/10.1021/acsami.9b01754.

(87) Johnson, B. A.; Bhunia, A.; Ott, S. Electrocatalytic Water Oxidation by a Molecular Catalyst Incorporated into a Metal–Organic Framework Thin Film. Dalt. Trans. 2017, 46 (5), 1382–1388. https://doi.org/10.1039/C6DT03718F.

(88) Johnson, B. A.; Beiler, A. M.; McCarthy, B. D.; Ott, S. Transport Phenomena: Challenges and Opportunities for Molecular Catalysis in Metal–Organic Frameworks. J. Am. Chem. Soc. 2020, 142 (28), 11941–11956. https://doi.org/10.1021/jacs.0c02899.

(89) Matheu, R.; Ertem, M. Z.; Benet-Buchholz, J.; Coronado, E.; Batista, V. S.; Sala, X.; Llobet, A. Intramolecular Proton Transfer Boosts Water Oxidation Catalyzed by a Ru Complex. J. Am. Chem. Soc. 2015, 137 (33), 10786–10795. https://doi.org/10.1021/jacs.5b06541.

87

(90) Galaup, C.; Couchet, J. M.; Bedel, S.; Tisnès, P.; Picard, C. Direct Access to Terpyridine-Containing Polyazamacrocycles as Photosensitizing Ligands for Eu(III) Luminescence in Aqueous Media. J. Org. Chem. 2005, 70 (6), 2274–2284. https://doi.org/10.1021/jo0480242.

(91) Liseev, T.; Howe, A.; Hoque, M. A.; Gimbert-Suriñach, C.; Llobet, A.; Ott, S. Synthetic Strategies to Incorporate Ru-Terpyridyl Water Oxidation Catalysts into MOFs: Direct Synthesis vs. Post-Synthetic Approach. Dalt. Trans. 2020, 13753–13759. https://doi.org/10.1039/d0dt01890b.

(92) Lv, X.; Tong, M.; Huang, H.; Wang, B.; Gan, L.; Yang, Q.; Zhong, C.; Li, J. A High Surface Area Zr ( IV ) -Based Metal – Organic Framework Showing Stepwise Gas Adsorption and Selective Dye Uptake. J. Solid State Chem. 2015, 223, 104–108. https://doi.org/10.1016/j.jssc.2014.07.001.

(93) Gadzikwa, T.; Zeng, B.-S.; Hupp, J. T.; Nguyen, S. T. Ligand-Elaboration as a Strategy for Engendering Structural Diversity in Porous Metal–Organic Framework Compounds. Chem. Commun. 2008, No. 31, 3672. https://doi.org/10.1039/b714160b.

(94) Marshall, R. J.; Griffin, S. L.; Wilson, C.; Forgan, R. S. Single-Crystal to Single-Crystal Mechanical Contraction of Metal–Organic Frameworks through Stereoselective Postsynthetic Bromination. J. Am. Chem. Soc. 2015, 137 (30), 9527–9530. https://doi.org/10.1021/jacs.5b05434.

(95) Ayala, S.; Bentz, K. C.; Cohen, S. M. Block Co-PolyMOFs: Morphology Control of Polymer–MOF Hybrid Materials. Chem. Sci. 2019, 10 (6), 1746–1753. https://doi.org/10.1039/C8SC04250K.

(96) Ayala, S.; Zhang, Z.; Cohen, S. M. Hierarchical Structure and Porosity in UiO-66 PolyMOFs. Chem. Commun. 2017, 53 (21), 3058–3061. https://doi.org/10.1039/C6CC10225E.

(97) Thommes, M.; Kaneko, K.; Neimark, A. V.; Olivier, J. P.; Rodriguez-Reinoso, F.; Rouquerol, J.; Sing, K. S. W. Physisorption of Gases, with Special Reference to the Evaluation of Surface Area and Pore Size Distribution (IUPAC Technical Report). Pure Appl. Chem. 2015, 87 (9–10), 1051–1069. https://doi.org/10.1515/pac-2014-1117.

(98) Hod, I.; Bury, W.; Karlin, D. M.; Deria, P.; Kung, C.-W.; Katz, M. J.; So, M.; Klahr, B.; Jin, D.; Chung, Y.-W.; Odom, T. W.; Farha, O. K.; Hupp, J. T. Directed Growth of Electroactive Metal-Organic Framework Thin Films Using Electrophoretic Deposition. Adv. Mater. 2014, 26 (36), 6295–6300. https://doi.org/10.1002/adma.201401940.

(99) Hod, I.; Bury, W.; Gardner, D. M.; Deria, P.; Roznyatovskiy, V.; Wasielewski, M. R.; Farha, O. K.; Hupp, J. T. Bias-Switchable Permselectivity and Redox Catalytic Activity of a Ferrocene-Functionalized, Thin-Film Metal−Organic Framework Compound. J. Phys. Chem. Lett 2015, 6, 29. https://doi.org/10.1021/acs.jpclett.5b00019.

(100) Burrows, A. D.; Fisher, L. C.; Richardson, C.; Rigby, S. P. Selective Incorporation of Functional Dicarboxylates into Zinc Metal–Organic Frameworks. Chem. Commun. 2011, 47 (12), 3380. https://doi.org/10.1039/c1cc10143a.

(101) Mohammad-Pour, G. S.; Hatfield, K. O.; Fairchild, D. C.; Hernandez-Burgos, K.; Rodríguez-López, J.; Uribe-Romo, F. J. A Solid-Solution Approach for Redox Active Metal-Organic Frameworks with Tunable Redox Conductivity. J. Am. Chem. Soc. 2019, 141 (51), 19978–19982. https://doi.org/10.1021/jacs.9b10639.

88

(102) Wu, B.; Fu, J.; Zhou, Y.; Luo, S.; Zhao, Y.; Quan, G.; Pan, X.; Wu, C. Tailored Core‒shell Dual Metal–Organic Frameworks as a Versatile Nanomotor for Effective Synergistic Antitumor Therapy. Acta Pharm. Sin. B 2020, 10 (11), 2198–2211. https://doi.org/10.1016/j.apsb.2020.07.025.

(103) Hinterholzinger, F. M.; Wuttke, S.; Roy, P.; Preuße, T.; Schaate, A.; Behrens, P.; Godt, A.; Bein, T. Highly Oriented Surface-Growth and Covalent Dye Labeling of Mesoporous Metal-Organic Frameworks. Dalt. Trans. 2012, 41 (14), 3899–3901. https://doi.org/10.1039/c2dt12265k.

(104) Paneta, V.; Fluch, U.; Petersson, P.; Ott, S.; Primetzhofer, D. Characterization of Compositional Modifications in Metal-Organic Frameworks Using Carbon and Alpha Particle Microbeams. Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 2017, 404, 198–201. https://doi.org/https://doi.org/10.1016/j.nimb.2017.01.058.

(105) Fei, H.; Shin, J.; Meng, Y. S.; Adelhardt, M.; Sutter, J.; Meyer, K.; Cohen, S. M. Reusable Oxidation Catalysis Using Metal-Monocatecholato Species in a Robust Metal-Organic Framework. J. Am. Chem. Soc. 2014, 136 (13), 4965–4973. https://doi.org/10.1021/ja411627z.

(106) Kim, M.; Cahill, J. F.; Fei, H.; Prather, K. A.; Cohen, S. M. Postsynthetic Ligand and Cation Exchange in Robust Metal-Organic Frameworks. J. Am. Chem. Soc. 2012, 134 (43), 18082–18088. https://doi.org/10.1021/ja3079219.

(107) Pereira, C. F.; Howarth, A. J.; Vermeulen, N. A.; Almeida Paz, F. A.; Tomé, J. P. C.; Hupp, J. T.; Farha, O. K. Towards Hydroxamic Acid Linked Zirconium Metal-Organic Frameworks. Mater. Chem. Front. 2017, 1 (6), 1194–1199. https://doi.org/10.1039/c6qm00364h.

(108) Karagiaridi, O.; Bury, W.; Tylianakis, E.; Sarjeant, A. A.; Hupp, J. T.; Farha, O. K. Opening Metal-Organic Frameworks Vol. 2: Inserting Longer Pillars into Pillared-Paddlewheel Structures through Solvent-Assisted Linker Exchange. Chem. Mater. 2013, 25 (17), 3499–3503. https://doi.org/10.1021/cm401724v.

(109) Deria, P.; Mondloch, J. E.; Karagiaridi, O.; Bury, W.; Hupp, J. T.; Farha, O. K. Beyond Post-Synthesis Modification: Evolution of Metal-Organic Frameworks via Building Block Replacement. Chem. Soc. Rev. 2014, 43 (16), 5896–5912. https://doi.org/10.1039/c4cs00067f.

(110) Zhang, W.; Bu, A.; Ji, Q.; Min, L.; Zhao, S.; Wang, Y.; Chen, J. P Ka-Directed Incorporation of Phosphonates into Mof-808 via Ligand Exchange: Stability and Adsorption Properties for Uranium. ACS Appl. Mater. Interfaces 2019, 11 (37), 33931–33940. https://doi.org/10.1021/acsami.9b10920.

(111) Deria, P.; Bury, W.; Hupp, J. T.; Farha, O. K. Versatile Functionalization of the Nu-1000 Platform by Solvent-Assisted Ligand Incorporation. Chem. Commun. 2014, 50 (16), 1965–1968. https://doi.org/10.1039/c3cc48562e.

(112) Calculated Using Advanced Chemistry Development (ACD/Labs) Software V11.02 (© 1994-2020 ACD/Labs).

(113) CRC Handbook of Chemistry and Physics, 100th ed.; Rumble, J. R., Ed.; CRC Press, Taylor & Francis Group, 2019.

(114) Cliffe, M. J.; Wan, W.; Zou, X.; Chater, P. A.; Kleppe, A. K.; Tucker, M. G.; Wilhelm, H.; Funnell, N. P.; Coudert, F. X.; Goodwin, A. L. Correlated Defect Nanoregions in a Metal-Organic Framework. Nat. Commun. 2014, 5 (May), 1–8. https://doi.org/10.1038/ncomms5176.

89

(115) Bozal‐Ginesta, C.; Pullen, S.; Ott, S.; Hammarström, L. Self‐Recovery of Photochemical H 2 Evolution with a Molecular Diiron Catalyst Incorporated in a UiO‐66 Metal–Organic Framework. ChemPhotoChem 2020, 4 (4), 287–290. https://doi.org/10.1002/cptc.201900273.

(116) Sikdar, N.; Jayaramulu, K.; Kiran, V.; Rao, K. V.; Sampath, S.; George, S. J.; Maji, T. K. Redox-Active Metal-Organic Frameworks: Highly Stable Charge-Separated States through Strut/Guest-to-Strut Electron Transfer. Chem. - A Eur. J. 2015, 21 (33), 11701–11706. https://doi.org/10.1002/chem.201501614.

(117) Usov, P. M.; Simonov, A. N.; Bond, A. M.; Murphy, M. J.; D’Alessandro, D. M. Untangling Complex Redox Chemistry in Zeolitic Imidazolate Frameworks Using Fourier Transformed Alternating Current Voltammetry. Anal. Chem. 2017, 89 (19), 10181–10187. https://doi.org/10.1021/acs.analchem.7b01224.

(118) Hod, I.; Sampson, M. D.; Deria, P.; Kubiak, C. P.; Farha, O. K.; Hupp, J. T. Fe-Porphyrin-Based Metal-Organic Framework Films as High-Surface Concentration, Heterogeneous Catalysts for Electrochemical Reduction of CO2. ACS Catal. 2015, 5 (11), 6302–6309. https://doi.org/10.1021/acscatal.5b01767.

(119) Goswami, S.; Nelson, J. N.; Islamoglu, T.; Wu, Y.-L.; Farha, O. K.; Wasielewski, M. R. Photoexcited Naphthalene Diimide Radical Anion Linking the Nodes of a Metal–Organic Framework: A Heterogeneous Super-Reductant. Chem. Mater. 2018, 30 (8), 2488–2492. https://doi.org/10.1021/acs.chemmater.8b00720.

(120) Johnson, B. A.; Bhunia, A.; Fei, H.; Cohen, S. M.; Ott, S. Development of a UiO-Type Thin Film Electrocatalysis Platform with Redox-Active Linkers. J. Am. Chem. Soc. 2018, 140 (8), 2985–2994. https://doi.org/10.1021/jacs.7b13077.

(121) Munuera, C.; Shekhah, O.; Wang, H.; Wöll, C.; Ocal, C. The Controlled Growth of Oriented Metal–Organic Frameworks on Functionalized Surfaces as Followed by Scanning Force Microscopy. Phys. Chem. Chem. Phys. 2008, 10 (48), 7257. https://doi.org/10.1039/b811010g.

(122) Shekhah, O.; Wang, H.; Kowarik, S.; Schreiber, F.; Paulus, M.; Tolan, M.; Sternemann, C.; Evers, F.; Zacher, D.; Fischer, R. A.; Wöll, C. Step-by-Step Route for the Synthesis of Metal−Organic Frameworks. J. Am. Chem. Soc. 2007, 129 (49), 15118–15119. https://doi.org/10.1021/ja076210u.

(123) Shekhah, O.; Wang, H.; Strunskus, T.; Cyganik, P.; Zacher, D.; Fischer, R.; Wöll, C. Layer-by-Layer Growth of Oriented Metal Organic Polymers on a Functionalized Organic Surface. Langmuir 2007, 23 (14), 7440–7442. https://doi.org/10.1021/la701148z.

(124) Zhuang, J.-L.; Terfort, A.; Wöll, C. Formation of Oriented and Patterned Films of Metal–Organic Frameworks by Liquid Phase Epitaxy: A Review. Coord. Chem. Rev. 2016, 307, 391–424. https://doi.org/10.1016/j.ccr.2015.09.013.

(125) Beiler, A. M.; McCarthy, B. D.; Johnson, B. A.; Ott, S. Enhancing Photovoltages at P-Type Semiconductors through a Redox-Active Metal-Organic Framework Surface Coating. Nat. Commun. 2020, 11 (1), 5819. https://doi.org/10.1038/s41467-020-19483-5.

(126) Shekhah, O. Layer-by-Layer Method for the Synthesis and Growth of Surface Mounted Metal-Organic Frameworks (SURMOFs). Materials (Basel). 2010, 3 (2), 1302–1315. https://doi.org/10.3390/ma3021302.

90

(127) Zacher, D.; Yusenko, K.; Bétard, A.; Henke, S.; Molon, M.; Ladnorg, T.; Shekhah, O.; Schüpbach, B.; de los Arcos, T.; Krasnopolski, M.; Meilikhov, M.; Winter, J.; Terfort, A.; Wöll, C.; Fischer, R. A. Liquid-Phase Epitaxy of Multicomponent Layer-Based Porous Coordination Polymer Thin Films of [M(L)(P)0.5] Type: Importance of Deposition Sequence on the Oriented Growth. Chem. - A Eur. J. 2011, 17 (5), 1448–1455. https://doi.org/10.1002/chem.201002381.

(128) Shekhah, O.; Wang, H.; Paradinas, M.; Ocal, C.; Schüpbach, B.; Terfort, A.; Zacher, D.; Fischer, R. A.; Wöll, C. Controlling Interpenetration in Metal–Organic Frameworks by Liquid-Phase Epitaxy. Nat. Mater. 2009, 8 (6), 481–484. https://doi.org/10.1038/nmat2445.

(129) Shekhah, O.; Hirai, K.; Wang, H.; Uehara, H.; Kondo, M.; Diring, S.; Zacher, D.; Fischer, R. A.; Sakata, O.; Kitagawa, S.; Furukawa, S.; Wöll, C. MOF-on-MOF Heteroepitaxy: Perfectly Oriented [Zn2(Ndc)2(Dabco)]n Grown on [Cu2(Ndc)2(Dabco)]n Thin Films. Dalt. Trans. 2011, 40 (18), 4954. https://doi.org/10.1039/c0dt01818j.

(130) McCarthy, B. D.; Liseev, T.; Beiler, A. M.; Materna, K. L.; Ott, S. Facile Orientational Control of M2L2P SURMOFs on â &die;100»Silicon Substrates and Growth Mechanism Insights for Defective MOFs. ACS Appl. Mater. Interfaces 2019, 11 (41), 38294–38302. https://doi.org/10.1021/acsami.9b12407.

(131) Zhuang, J.-L.; Kind, M.; Grytz, C. M.; Farr, F.; Diefenbach, M.; Tussupbayev, S.; Holthausen, M. C.; Terfort, A. Insight into the Oriented Growth of Surface-Attached Metal–Organic Frameworks: Surface Functionality, Deposition Temperature, and First Layer Order. J. Am. Chem. Soc. 2015, 137 (25), 8237–8243. https://doi.org/10.1021/jacs.5b03948.

(132) Gu, Z.-G.; Bürck, J.; Bihlmeier, A.; Liu, J.; Shekhah, O.; Weidler, P. G.; Azucena, C.; Wang, Z.; Heissler, S.; Gliemann, H.; Klopper, W.; Ulrich, A. S.; Wöll, C. Oriented Circular Dichroism Analysis of Chiral Surface-Anchored Metal-Organic Frameworks Grown by Liquid-Phase Epitaxy and upon Loading with Chiral Guest Compounds. Chem. - A Eur. J. 2014, 20 (32), 9879–9882. https://doi.org/10.1002/chem.201403524.

(133) Kitaura, R.; Iwahori, F.; Matsuda, R.; Kitagawa, S.; Kubota, Y.; Takata, M.; Kobayashi, T. C. Rational Design and Crystal Structure Determination of a 3-D Metal−Organic Jungle-Gym-like Open Framework. Inorg. Chem. 2004, 43 (21), 6522–6524. https://doi.org/10.1021/ic049005d.

(134) Liu, J.; Zhou, W.; Liu, J.; Howard, I.; Kilibarda, G.; Schlabach, S.; Coupry, D.; Addicoat, M.; Yoneda, S.; Tsutsui, Y.; Sakurai, T.; Seki, S.; Wang, Z.; Lindemann, P.; Redel, E.; Heine, T.; Wöll, C. Photoinduced Charge-Carrier Generation in Epitaxial MOF Thin Films: High Efficiency as a Result of an Indirect Electronic Band Gap? Angew. Chemie - Int. Ed. 2015, 54 (25), 7441–7445. https://doi.org/10.1002/anie.201501862.

(135) Gu, Z. G.; Pfriem, A.; Hamsch, S.; Breitwieser, H.; Wohlgemuth, J.; Heinke, L.; Gliemann, H.; Wöll, C. Transparent Films of Metal-Organic Frameworks for Optical Applications. Microporous Mesoporous Mater. 2015, 211, 82–87. https://doi.org/10.1016/j.micromeso.2015.02.048.

(136) Bétard, A.; Bux, H.; Henke, S.; Zacher, D.; Caro, J.; Fischer, R. A. Fabrication of a CO2-Selective Membrane by Stepwise Liquid-Phase Deposition of an Alkylether Functionalized Pillared-Layered Metal-Organic Framework [Cu2L2P]n on a Macroporous Support. Microporous Mesoporous Mater. 2012, 150 (1), 76–82. https://doi.org/10.1016/j.micromeso.2011.09.006.

91

(137) Liu, B.; Shekhah, O.; Arslan, H. K.; Liu, J.; Wöll, C.; Fischer, R. A. Enantiopure Metal-Organic Framework Thin Films: Oriented SURMOF Growth and Enantioselective Adsorption. Angew. Chemie Int. Ed. 2012, 51 (3), 807–810. https://doi.org/10.1002/anie.201104240.

(138) Wang, Z.; Rodewald, K.; Medishetty, R.; Rieger, B.; Fischer, R. A. Control of Water Content for Enhancing the Quality of Copper Paddle-Wheel-Based Metal-Organic Framework Thin Films Grown by Layer-by-Layer Liquid-Phase Epitaxy. Cryst. Growth Des. 2018, 18 (12), 7451–7459. https://doi.org/10.1021/acs.cgd.8b01212.

(139) Shekhah, O.; Wang, H.; Zacher, D.; Fischer, R. A.; Wöll, C. Growth Mechanism of Metal-Organic Frameworks: Insights into the Nucleation by Employing a Step-by-Step Route. Angew. Chemie Int. Ed. 2009, 48 (27), 5038–5041. https://doi.org/10.1002/anie.200900378.

(140) Yusenko, K.; Meilikhov, M.; Zacher, D.; Wieland, F.; Sternemann, C.; Stammer, X.; Ladnorg, T.; Wöll, C.; Fischer, R. A. Step-by-Step Growth of Highly Oriented and Continuous Seeding Layers of [Cu2(Ndc)2(Dabco)] on Bare Oxide and Nitride Substrates. CrystEngComm 2010, 12 (7), 2086. https://doi.org/10.1039/b927212g.

(141) Chen, Y. W.; Prange, J. D.; Dühnen, S.; Park, Y.; Gunji, M.; Chidsey, C. E. D.; McIntyre, P. C. Atomic Layer-Deposited Tunnel Oxide Stabilizes Silicon Photoanodes for Water Oxidation. Nat. Mater. 2011, 10 (7), 539–544. https://doi.org/10.1038/nmat3047.

(142) Hu, S.; Shaner, M. R.; Beardslee, J. A.; Lichterman, M.; Brunschwig, B. S.; Lewis, N. S. Amorphous TiO2 Coatings Stabilize Si, GaAs, and GaP Photoanodes for Efficient Water Oxidation. Science (80-. ). 2014, 344 (6187), 1005–1009. https://doi.org/10.1126/science.1251428.

(143) Gish, M. K.; Lapides, A. M.; Brennaman, M. K.; Templeton, J. L.; Meyer, T. J.; Papanikolas, J. M. Ultrafast Recombination Dynamics in Dye-Sensitized SnO 2 /TiO 2 Core/Shell Films. J. Phys. Chem. Lett. 2016, 7 (24), 5297–5301. https://doi.org/10.1021/acs.jpclett.6b02388.

(144) Shekhah, O.; Hirai, K.; Wang, H.; Uehara, H.; Kondo, M.; Diring, S.; Zacher, D.; Fischer, R. A.; Sakata, O.; Kitagawa, S.; Furukawa, S.; Wöll, C. MOF-on-MOF Heteroepitaxy: Perfectly Oriented [Zn2(Ndc)2(Dabco)]n Grown on [Cu2(Ndc)2(Dabco)]n Thin Films. Dalt. Trans. 2011, 40 (18), 4954. https://doi.org/10.1039/c0dt01818j.

(145) Liu, J.; Wöll, C. Surface-Supported Metal–Organic Framework Thin Films: Fabrication Methods, Applications, and Challenges. Chem. Soc. Rev. 2017, 46 (19), 5730–5770. https://doi.org/10.1039/C7CS00315C.

(146) Heinke, L.; Wöll, C. Adsorption and Diffusion in Thin Films of Nanoporous Metal–Organic Frameworks: Ferrocene in SURMOF Cu2(Ndc)2(Dabco). Phys. Chem. Chem. Phys. 2013, 15 (23), 9295. https://doi.org/10.1039/c3cp50578b.

(147) Ma, B.-Q.; Mulfort, K. L.; Hupp, J. T. Microporous Pillared Paddle-Wheel Frameworks Based on Mixed-Ligand Coordination of Zinc Ions. Inorg. Chem. 2005, 44 (14), 4912–4914. https://doi.org/10.1021/ic050452i.

(148) Sakata, Y.; Furukawa, S.; Kondo, M.; Hirai, K.; Horike, N.; Takashima, Y.; Uehara, H.; Louvain, N.; Meilikhov, M.; Tsuruoka, T.; Isoda, S.; Kosaka, W.; Sakata, O.; Kitagawa, S. Shape-Memory Nanopores Induced in Coordination Frameworks by Crystal Downsizing. Science (80-. ). 2013, 339 (6116), 193–196. https://doi.org/10.1126/science.1231451.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 2018

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally throughthe series Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-435108

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2021