Lanthanide Luminescent Metal-Organic Frameworks with Linear Dicarboxylate Ligands: Synthesis, Structure and Sensing Properties

by

Yu Li

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy (Ph.D.) Department of Chemistry University of Toronto

© Copyright by Yu Li 2013

Lanthanide Luminescent Metal-Organic Frameworks with Linear Dicarboxylate Ligands: Synthesis, Structure and Sensing Properties

Yu Li

Doctor of Philosophy

Department of Chemistry University of Toronto

2013 Abstract

The object of this thesis is to explore functionalized linear dicarboxylate ligands for constructing luminescent metal-organic framework (MOF) sensors. The first series of ligands developed is based on functionalized (E)-4,4’-(ethene-1,2-diyl)dibenzoic acids. Luminescent MOFs have been synthesized from these ligands and their porosity, thermal stability, luminescent properties have been discussed. The second series of ligands is based on functionalized [1,1':4',1''-terphenyl]- 4,4''-dicarboxylic acid. Ligands bearing methyl, methoxyl, thioether, aldehyde, , quaternary ammonium side chains have been synthesized and constructed into luminescent MOFs. Their structures, thermal stability, luminescent properties, solvent-dependent luminescence and the underlying mechanisms have been discussed. Vapor sensing experiments have been conducted with MOFs bearing ether and thioether side chains and strong luminescence turn-on is triggered by certain solvents. The MOF with ether side chains has been further tested as sensor slides to evaluate its response rate and recyclability.

ii

Acknowledgments

First and foremost, I am deeply indebted to my supervisor Prof. Datong Song for his vision, guidance and encouragement throughout my Ph.D. study. As a fresh graduate from a material background, my understanding of organic synthesis is limited to simple text book reactions. It is Datong who helped me develop my synthetic skills from scratch, starting from using science finder, setting up air-free reactions and running columns. Without his guidance, I could not have completed the systematic study of our ligand system. I am also impressed by his persistence and creativity, which have made our way through the two rejections and eventually led to our publication on Angew. Chem. Int. Ed.

Besides my supervisor, I would like to thank the rest of my committee members, Prof. Doug Stephan, Prof. Bob Morris and Prof. Ulrich Fekl, for their helpful suggestions

The past and present members of Song group (Prof. Yang Li, Dr. Shaolong Gong, Dr. Yunshan Sun, Dr. Tao Bai, Dr. Xiaofei Li, Dr. Runyu Tan, Dr. Alen Hadzovic, Ali Nazemi, Elzbieta Stepowska, Vincent Annibale, Trevor Janes, Charlie Kivi, Rhys Batup, Tara Cho, Yanxin Yang) also deserves my sincere thanks, for all their help and advice which have contributed greatly to my personal and professional development.

I would also like to thank Dr. Srebri Petrov, Dr. George Kretschmann for the help and discussion on PXRD experiments, Anna Liza Villavelez for all the administrative paper work, Jennifer Lofgreen for gas sorption experiments.

Finally, I want to thank my parents, other family members and Wei Feng. Without your support and input, I could not have completed my Ph.D. program.

iii

Table of Contents

Acknowledgments ...... iii

Table of Contents ...... iv

List of Tables ...... x

List of Schemes ...... xii

List of Figures ...... xiii

List of Symbols and Abbreviations ...... xxii

Chapter 1 Introduction ...... 1

1.1 Definition of MOF ...... 1

1.2 MOF synthetic methods ...... 2

1.2.1 Hydrothermal/Solvothermal synthesis ...... 2

1.2.2 Other synthetic methods ...... 3

1.2.3 Post-synthetic modification (PSM) ...... 5

1.3 MOF Structure and iso-reticular chemistry ...... 6

1.4 MOF functionalities and applications ...... 9

1.4.1 Gas storage and separation ...... 10

1.4.2 MOF catalysis ...... 10

1.4.3 MOF sensors ...... 11

1.5 Luminescent sensors – homogeneous systems and MOFs ...... 12

1.5.1 Homogeneous luminescent sensors ...... 12

1.5.2 Luminescent MOF sensors ...... 14

1.6 Lanthanide luminescence MOF sensors ...... 15

1.7 Scope of the thesis ...... 22

References ...... 24

iv

Chapter 2 Lanthanide Metal-Organic Framework of Functionalized (E)-4,4’-(ethene-1,2- diyl)dibenzoic acids ...... 28

2.1 Abstract ...... 28

2.2 Introduction ...... 28

2.3 Experimental section ...... 29

2.3.1 Materials and methods ...... 29

2.3.2 Synthesis of dimethyl 4,4'-(but-2-ene-2,3-diyl)dibenzoate (trans and cis LM ester) ...... 30

2.3.3 Synthesis of dimethyl 4,4’-(1,4-dibromobut-2-ene-2,3-diyl)dibenzoate (trans and cis LB ester) ...... 30

2.3.4 Synthesis of (E)-dimethyl 4,4’-(1,4-bis(methylthio)but-2-ene-2,3- diyl)dibenzoate (LS ester) ...... 31

2.3.5 Synthesis of (E)-4,4’-(1,4-bis(methylthio)but-2-ene-2,3-diyl)dibenzoic acid (H2LS) ...... 31

2.3.6 Synthesis of (E)-dimethyl 4,4'-(1,4-bis(phenylthio)but-2-ene-2,3- diyl)dibenzoate (LSPh ester) ...... 31

2.3.7 Synthesis of (E)-4,4'-(1,4-bis(phenylthio)but-2-ene-2,3-diyl)dibenzoic acid (LSPh acid) ...... 32

2.3.8 Synthesis of [Tb2(LH)3(DMSO)4] (TbLH) ...... 32

2.3.9 Synthesis of [Eu2(LH)3(DMSO)4] (EuLH) ...... 32

2.3.10 Synthesis of [Eu2(LS)3(DMF)2(H2O)2] (EuLS) ...... 33

2.3.11 Synthesis of [Tb2(LS)3(DMF)2(H2O)2] (TbLS) ...... 33

2.3.12 X-ray Crystallographic Analysis...... 33

2.4 Results and discussion ...... 34

2.4.1 Ligand synthesis ...... 34

2.4.2 MOF preparation and structure description ...... 36

2.4.3 Solvent inclusion and porosity ...... 42

2.4.4 Luminescence properties ...... 47

2.5 Summary ...... 47

v

References ...... 58

Chapter 3 Lanthanide Metal-Organic Framework of (2',5'-bis(methoxymethyl)-[1,1':4',1''- terphenyl]-4,4''-dicarboxylic acid and its Potential as Solvent Vapor Sensor ...... 59

3.1 Abstract ...... 59

3.2 Introduction ...... 59

3.3 Experimental section ...... 60

3.3.1 Materials and method ...... 60

3.3.2 Synthesis of 2',5'-bis(methoxymethyl)-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid (H2sLOM) ...... 61

3.3.3 Synthesis of [Eu2(sLOM)3(H2O)4] (EusLOM-1) ...... 61

3.3.4 Preparation of water-exchanged MOF (EusLOM-2) ...... 62

3.3.5 Preparation of water-exchanged MOF for lifetime measurements (EusLOM-3- H2O/D2O)...... 62

3.3.6 X-ray crystallographic analysis...... 62

3.3.7 Measurement for maximum luminescence change ...... 62

3.3.8 Preparation of recycled MOF sample and the recycling experiments ...... 63

3.3.9 Quantification of solvent uptake in MOF samples ...... 63

3.3.10 Preparation of water-exchanged MOF (EusLOM-3-H2O/D2O) and the lifetime study...... 64

3.3.11 Prototype sensor experimental setup ...... 64

3.4 Results and Discussion ...... 65

3.4.1 MOF synthesis and Structural description ...... 65

3.4.2 Thermogravimetric analysis ...... 69

3.4.3 Luminescence properties ...... 69

3.4.4 Solvent-dependent luminescence and sensing experiments ...... 70

3.4.5 Luminescence lifetime study and sensing mechanism ...... 73

3.4.6 Prototype sensing experiments and sensor response rate ...... 77

3.5 Summary ...... 78 vi

References ...... 93

Chapter 4 Properties of Lanthanide Metal-Organic Frameworks of Functionalized [1,1':4',1''- terphenyl]-4,4''-dicarboxylic acid...... 94

4.1 Abstract ...... 94

4.2 Introduction ...... 94

4.3 Experimental section ...... 95

4.3.1 Materials and methods ...... 95

4.3.2 Synthesis of 2,5-dibromo-1,4-benzenedicarboxaldehyde ...... 96

4.3.3 Synthesis of dimethyl 2',5'-diformyl-[1,1':4',1''-terphenyl]-4,4''-dicarboxylate (sLCHO ester) ...... 96

4.3.4 Synthesis of 2',5'-diformyl-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid (H2sLCHO) ...... 97

4.3.5 Synthesis of (3E,3'E)-4,4'-(2,5-dibromo-1,4-phenylene)bis(but-3-en-2-one) ...... 97

4.3.6 Synthesis of dimethyl 2',5'-bis((E)-3-oxobut-1-en-1-yl)-[1,1':4',1''-terphenyl]- 4,4''-dicarboxylate (sLCO ester) ...... 97

4.3.7 Synthesis of 2',5'-bis((E)-3-oxobut-1-en-1-yl)-[1,1':4',1''-terphenyl]-4,4''- dicarboxylic acid (H2sLCO) ...... 98

4.3.8 Synthesis of dimethyl 2',5'-bis(methylthio)-[1,1':4',1''-terphenyl]-4,4''- dicarboxylate (sLDS ester) ...... 98

4.3.9 Synthesis of 2',5'-bis(methylthio)-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid (H2sLDS) ...... 99

4.3.10 Synthesis of [Eu2(sLM)3(H2O)(DMF)2] (EusLM) ...... 99

4.3.11 Synthesis of [Tb2(sLM)3(H2O)2(DMF)2] (TbsLM) ...... 99

4.3.12 Synthesis of [Tb2(sLOM)3(H2O)4] (TbsLOM) ...... 100

4.3.13 Synthesis of [Eu2(sLCHO)3 (H2O)4] (EusLCHO) ...... 100

4.3.14 Synthesis of [Eu2(sLCO)3(H2O)4(DMF) (DMSO)]. (EusLCO) ...... 100

4.3.15 Synthesis of [Eu2(sLDS)3(H2O)4(DMF)2] (EusLDS) ...... 100

4.3.16 Synthesis of [Tb2(sLDS)3(H2O)4(DMF)2] (TbsLDS) ...... 101

4.3.17 Preparation of water-soaked EusLDS sample for sensing experiments ...... 101

vii

4.3.18 X-ray crystallographic analysis...... 101

4.3.19 Determination of solvent composition of different MOF samples ...... 101

4.3.20 1H NMR experiments procedures for MOF samples ...... 102

4.3.21 Luminescence measurements ...... 102

4.3.22 Vapor sensing experiments of EusLDS...... 103

4.4 Results and Discussion ...... 104

4.4.1 Ligand Synthesis, MOF Preparation and Structures ...... 104

4.4.2 Thermogravimetric analysis ...... 116

4.4.3 Luminescence properties ...... 120

4.4.4 Solvent dependent luminescence and sensing experiments ...... 127

4.5 Summary ...... 138

References ...... 159

Chapter 5 Lanthanide MOFs Synthesized from Ligands with Cationic Groups ...... 160

5.1 Abstract ...... 160

5.2 Introduction ...... 160

5.3 Experimental section ...... 163

5.3.1 Materials and methods ...... 163

5.3.2 Synthesis of N,N'-((4,4''-bis(methoxycarbonyl)-[1,1':4',1''-terphenyl]-2',5'- diyl)bis(methylene))bis(N,N-diethylethanaminium) bromide (sLN ester) ...... 163

5.3.3 Synthesis of 4''-carboxy-2',5'-bis((triethylammonio)methyl)-[1,1':4',1''- terphenyl]-4-carboxylate chloride (sLN-HCl) ...... 163

5.3.4 Synthesis of N,N'-((4,4''-dicarboxy-[1,1':4',1''-terphenyl]-2',5'- diyl)bis(methylene))bis(N,N-diethylethanaminium) nitrate (sLN-HNO3) ...... 164

5.3.5 N,N'-((4,4''-dicarboxy-[1,1':4',1''-terphenyl]-2',5'-diyl)bis(methylene))bis(N,N- diethylethanaminium) trifluoromethanesulfonate (sLN-HSO3CF3) ...... 164

5.3.6 Synthesis of N,N'-((4,4''-dicarboxy-[1,1':4',1''-terphenyl]-2',5'- diyl)bis(methylene))bis(N,N-diethylethanaminium) perchlorate (sLN-HClO4) . 165

5.3.7 Synthesis of [Eu2(sLN)3Cl6] (EusLN-Cl) ...... 165

viii

5.3.8 Synthesis of [Eu2(sLN)3(NO3)6] (EusLN-NO3) ...... 165

5.3.9 X-ray crystallographic analysis ...... 166

5.3.10 Determination of solvent composition of different MOF samples ...... 166

5.3.11 1H NMR experiments procedures for MOF samples ...... 166

5.3.12 Luminescence measurements ...... 167

5.3.13 Anion exchange experiment ...... 167

5.4 Results and discussion ...... 167

5.4.1 Ligand synthesis ...... 167

5.4.2 MOF synthesis and structure ...... 169

5.4.3 Thermogravimetric analysis ...... 175

5.4.4 Luminescent properties ...... 177

5.4.5 Anion exchange experiments ...... 177

5.5 Summary ...... 179

References ...... 186

Chapter 6 Summary and Outlook ...... 187

6.1 Summary of the thesis ...... 187

6.2 Future work ...... 189

6.3 Final remark ...... 190

ix

List of Tables

Table 2.1. Crystallographic data of trans/cis LM ester and trans/cis LB ester...... 49

Table 2.2. Crystallographic data of LS ester, LSPh ester, and H2LSPh...... 50

Table 2.3. Crystallographic data EuLH, TbLH, EuLS, and TbLS...... 51

Table 2.4. Selected bond lengths (Å) and angles (deg) of TbLH...... 52

Table 2.5. Selected bond lengths (Å) and angles (deg) of EuLH...... 53

Table 2.6. Selected bond lengths (Å) and angles (deg) of EuLS...... 54

Table 2.7. Selected bond lengths (Å) and angles (deg) of TbLS...... 55

Table 3.1. Selective bond length in crystal structure of EusLOM-1 ...... 66

Table 3.2. Solvent content after incubated with various solvent or drying reagent MgSO4 ...... 73

Table 3.3. Emission lifetime (τ) and intensity (Int) measured with ligand-based excitation and direct Eu excitation...... 75

Table 3.4. Crystallographic data of EusLOM ...... 79

Table 4.1. Summary of ligand singlet/triplet state energy levels and the respective energy gaps...... 124

Table 4.2. Luminescence lifetime and intensity of EusLCHO suspension in various solvent. .... 129

Table 4.3. Luminescence lifetime EusLCHO suspension in various solvent...... 130

Table 4.4. Calculation after H-D exchange is estimated with linear approximation ...... 132

Table 4.5. Luminescence lifetime and intensity of EuLOM in various solvents...... 132

Table 4.6. Luminescence lifetime of water-soaked EusLDS in various solvents...... 136

Table 4.7. Crystallographic data of EusLM, TbsLM, TbsLOM and EusLCHO ...... 139

x

Table 4.8. Crystallographic data of EusLCO, EusLDS, TbsLDS...... 140

Table 4.9. Selected bond lengths (Å) and angles (deg) for EusLM ...... 141

Table 4.10. Selected bond lengths (Å) and angles (deg) for TbsLM ...... 142

Table 4.11. Selected bond lengths (Å) and angles (deg) for TbsLOM ...... 144

Table 4.12. Selected bond lengths (Å) and angles (deg) for EusLCHO ...... 145

Table 4.13. Selected bond lengths (Å) and angles (deg) for EusLCO ...... 146

Table 4.14. Selected bond lengths (Å) and angles (deg) for EusLDS ...... 147

Table 4.15. Selected bond lengths (Å) and angles (deg) for TbsLDS ...... 148

Table 5.1. Crystallographic data of sLN ester, sLN-HNO3, EusLN-NO3, EusLN-Cl(a), and EusLN- Cl(b)...... 180

Table 5.2. Selected bond lengths (Å) and angles (deg) for EusLN-NO3 ...... 181

Table 5.3. Selected bond lengths (Å) and angles (deg) for EusLN-Cl(a) ...... 182

Table 5.4. Selected bond lengths (Å) and angles (deg) for EusLN-Cl(b) ...... 183

xi

List of Schemes

Scheme 2.1. MOF design in Chapter 2 ...... 29

Scheme 2.2. Synthesis of H2LS and H2LSPh...... 34

Scheme 2.3. Different coordination modes of carboxylate ligands...... 40

Scheme 3.1. General strategy for H2sLX ligand synthesis...... 60

Scheme 4.1. Synthesis of H2sLCHO and H2sLCO...... 108

Scheme 4.2. Synthesis of sLCO ester via Suzuki coupling reaction...... 109

Scheme 4.3. Synthesis of H2sLDS...... 112

xii

List of Figures

Figure 1.1. MOF-5 crystal structure ...... 2

Figure 1.2. MOF-5 type structure: (a) octahedral node; (c) examples of ligands used...... 7

Figure 1.3. Pillared paddle-wheel type structure: (a) octahedral node; (c) examples of dicarboxylate ligands used; (d) examples of bipyridine type ligands used...... 8

Figure 1.4. HKUST-1 type structure: (a) paddle-wheel metal node; (b) channel view; (c) big cage; (d) small octahedral cage; (e)-(g) examples of symmetrical and asymmetrical ligands...... 9

Figure 1.5. Examples of PET type sensor (a) and ICT type sensor (b)...... 13

Figure 1.6. Ligand-to-metal energy transfer mechanism in lanthanide complexes.[54] ...... 16

Figure 1.7. Examples of lanthanide complexes sensors for cationic (a, b) and anionic (c, d) analytes...... 18

Figure 1.8. Ligand mentioned in LnMOF sensor examples...... 20

Figure 1.9. An early MOF developed in our lab ...... 22

Figure 1.10. Ligand systems covered in this thesis...... 22

Figure 2.1. Crystal structures of (a) trans LM ester; (b) cis LM ester; (c) trans LB ester; (d) cis LB ester; (e) LS ester; (f) LSPh ester; (g) H2LSPh...... 36

Figure 2.2. ORTEP drawings of TbLH: (a) the coordination sphere of a Tb(III) center showing the tricapped trigonal prismatic coordination geometry; (b) the structure of a dinuclear node

[Tb2(LH)3(DMSO)4] with the stilbene linkers reduced to the ipso carbons. The thermal ellipsoids are plotted at 50% probabilities. Extended structures of TbLH: (c) a drawing of one set of 3D framework, projection along the b axis; (d) a space filling drawing showing the two interpenetrating frameworks, one in red and the other in blue. Hydrogen atoms and lattice solvent molecules are omitted for clarity...... 38

xiii

Figure 2.3. ORTEP drawings of EuLH: (a) the coordination sphere of a Eu(III) center showing the tricapped trigonal prismatic coordination geometry; (b) the structure of a dinuclear node

[Eu2(LH)3(DMSO)4]. The thermal ellipsoids are plotted at 50% probabilities. Extended structures of EuLH: (c) a drawing of one set of 3D framework, projection along the c axis; (d) a space filling drawing showing the two interpenetrating frameworks, one in red and the other in blue. Hydrogen atoms and lattice solvent molecules are omitted for clarity...... 39

Figure 2.4. ORTEP drawings: (a) a di-Eu node in EuLS; (b) a di-Tb node in TbLS. Thermal 2- ellipsoids are plotted at 50% probabilities. All LS ligands are reduced to the carboxylic groups and an ipso carbon from the phenyl ring and all hydrogen atoms are omitted for clarity. Extended structures of EuLS (or TbLS): (c) one set of framework; (d) spacefilling drawing of three interpenetrating frameworks coated with different colors...... 41

Figure 2.5. Triangular channel along the body diagonal direction of the ‘cube’ (left), the helical arrangement of the ligands defining a triangular channel in a single-fold framework with possible channel perpendicular to the triangular channel (middle) and in a 3-fold interpenetrating framework with no possible channel perpendicular to the triangular channel (right)...... 43

Figure 2.6.Diagrams showing the triangular channels: (a) TbLH, spacefilling model, projection down the a+b direction; lattice solvent molecules are omitted for clarity. (b) EuLS, spacefilling model for the skeleton and stick model for the thioether functional groups, projection down the a direction...... 44

Figure 2.7. TGA of (a) EuLH and TbLH; (b) EuLS and TbLS; (c) As-synthesized, de-solvated and re-solvated EuLS; (d) NaLH salt...... 44

Figure 2.8. IR spectra of EuLS sample: (a) as-synthesized; (b) de-solvated; (c) re-solvated...... 45

Figure 2.9. Powder X-ray diffraction of EuLS: (a) pattern predicted from crystal structure; (b) as- synthesized; (c) de-solvated; (d) re-solvated...... 45

Figure 2.10. Photoluminescence spectra of TbLH, EuLH, EuLS,TbLS, H2LH and H2LS (a-f, respectively). (a) and (b): Excitation and emission were measured at 450 and 404 nm, respectively. (c): Excitation, emission (i) and emission (ii) were measured at 615, 393, and 350 nm, respectively. (d) Excitation, emission (i) and emission (ii) were measured at 515, 410, and

xiv

487 nm, respectively; (e) Excitation and emission were measured at 464, 382 nm respectively, with H2LH solid; (f) Excitation and emission were measured at 479, 396 nm respectively, with

H2LS solid...... 46

Figure 2.11. PXRD pattern of as-synthesized TbLH ...... 56

Figure 2.12. PXRD pattern of as-synthesized EuLH ...... 56

Figure 2.13. PXRD pattern of as-synthesized EuLS ...... 57

Figure 2.14. PXRD pattern of as-synthesized TbLS ...... 57

Figure 3.1. Diagram of the experimental setup for incubating 2 under different solvent vapors. 63

Figure 3.2. Experimental setup for the prototype sensor experiment...... 65

Figure 3.3. (a) Coordination sphere of Eu; (b) The asymmetric unit of EusLOM-1 with thermal ellipsoids plotted at 50% probabilities. All hydrogen atoms and cocrystallized DMF solvent molecules are omitted for clarity...... 66

Figure 3.4. Crystal structure of EusLOM: (a) 1D chain with bridging carboxylates; (b) extended structure showing connectivity; (c) perspective view of 1D chains and open channels. All hydrogen atoms and channel solvent molecules are omitted for clarity...... 67

Figure 3.5. FTIR of as-synthesized MOF (EusLOM-1) and water-soaked MOF (EusLOM-2)...... 68

Figure 3.6. PXRD of EusLOM-1 and EusLOM-2. “Predicted” is the calculated pattern from crystal structure of EusLOM-1...... 68

Figure 3.7. Thermogravimetric analysis of EusLOM-1...... 69

Figure 3.8. Luminescence spectra of EusLOM-1 and EusLOM-2. Excitation and emission are measured at 616 and 323 nm respective. The peaks at 308 nm in excitation spectra and 646 nm in emission spectra are due to the scattering...... 70

xv

Figure 3.9. Increase in Eu emission intensity of EusLOM-2 (Iafter/Ibefore-1) after 24 h incubation under various solvent vapors and anhydrous MgSO4. The intensity is measured at 616 nm. Error bars indicate the standard deviations of three or four parallel experiments...... 71

Figure 3.10. Emission spectra of 3.2 before and after exposure to DMF vapor (excited at 323 nm). The broad peak around 640 nm is scattering peak...... 71

Figure 3.11. Photo of the sensor at on and off stages under a portable UV lamp irradiation. Note: the 'on' stage luminescence is so bright that it over-saturates the camera...... 72

Figure 3.12. Luminescence responses of the bulk regenerated sensor (by soaking in water) to selected solvent vapors upon 24 h incubation. The four bars for each solvent stand for fresh sample, 1st cycle, 2nd cycle, 3rd cycle, respectively...... 72

Figure 3.13. DMF molecules (shown with spacefilling model) that hinder the phenyl ring rotation. Top: DMF molecule is sandwiched between the central rings of two parallel ligands. Bottom: DMF molecules is located parallel to the side phenyl ring and the carboxylate...... 76

Figure 3.14. PXRD pattern of as-synthesized sample (EusLOM-1), water-soaked sample

(EusLOM-2), sample EusLOM-2 after incubated with various solvent/reagent for 24 h, sample regenerated from DMF incubated sample by soaking in water for 3 days (regenerated), and the calculated pattern from the crystal structure of EusLOM-1...... 76

Figure 3.15. On-off cycles of the sensor with alternating DMF-water vapor treatment. The intensity is measured at 616 nm with the excitation wavelength of 323 nm...... 77

1 Figure 3.16a. H NMR of EusLOM-3-D2O suspended in acetonitrile-d3...... 80

1 Figure 3.17a. H NMR of 9.9 mg of EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene...... 82

Figure 3.18. Luminescence intensity versus time plots of typical sensing experiments showing the reproducibility of the fast turn-on and -off...... 90

Figure 3.19a. Emission spectra of EusLOM-2 before and after exposure to various solvent vapors and drying reagent. (Cont.) ...... 91

xvi

Figure 4.1. Ligand series with different functional groups discussed in this chapter...... 94

Figure 4.2. Diagram of the experimental setup for incubating water-soaked EusLDS under different solvent vapors...... 104

Figure 4.3. Crystal structure of EusLM: (a) eight-coordinate Eu center; (b) 1D chain with bridging carboxylate and coordinated DMF; (c) perspective view of 1D chains and channels along c-axis, α ≈ 35°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity...... 105

Figure 4.4. Crystal structure of TbsLM: (a) seven-coordinate Tb1 center; (b) eight-coordinate Tb2 center; (c) 1D chain with alternating Tb1 and Tb2 centers bridged by carboxylate; (d) perspective view of 1D chains and channels, α ≈ 30°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity...... 106

Figure 4.5. Crystal structure of TbsLOM: (a) eight-coordinate Tb center; (b) 1D chain with bridging carboxylates; (c) perspective view of 1D chains and channels, α ≈ 40°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity...... 107

Figure 4.6. Crystal structure of EusLCHO: (a) nine-coordinate Eu1 center; (b) seven-coordinate Eu2 center; (c) 1D chain with alternating Eu1 and Eu2 centers bridged by carboxylates; (d) perspective view of 1D chains and channels, α ≈ 35°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity...... 110

Figure 4.7. Crystal structure of EuLCO: (a) eight-coordinate Eu center; (b) coordination environment of Eu center; (c) 1D chain with bridging by carboxylate; (d) layers of 2D sheets; (e) perspective view of 1D chains and channels. All hydrogen atoms and non-coordinate solvent molecules are omitted for clarity...... 111

Figure 4.8. Crystal structure of EusLDS: (a) eight-coordinate Eu center; (b) coordination environment of Eu; (c) 1D metal chain and the hydrogen bonding among coordinated solvents and the carboxylate oxygen; (d) channel view along the 1D metal chain, α ≈ 35°; (e) perspective view of the channel. All hydrogen atoms and non-coordinate solvent molecules are omitted for clarity...... 113

xvii

Figure 4.9. Crystal structure of TbsLDS: (a) eight-coordinate Tb center; (b) coordination environment of Tb; (c) 1D metal chain and the hydrogen bonding among coordinated solvents and the carboxylate oxygen; (d) channel view along the 1D metal chain, α ≈ 35°; (e) perspective view of the channel. All hydrogen atoms and non-coordinate solvent molecules are omitted for clarity...... 114

Figure 4.10. (a) The free DMF hydrogen bonded with the coordinated water in TbsLM. (b) The free DMF molecule sandwiched between the two central phenyl ring of two adjacent sLOM ligands. (c) π-π interaction in EusLCHO between adjacent sLCHO ligands. (d) π-π stacking in

EusLCO between adjacent sLCO ligands...... 115

Figure 4.11. (a) A typical ligand conformation in EusLDS where the methyl group is slightly tilted off the plane; (b) Ligand arrangement between adjacent metal chains...... 116

Figure 4.12. Thermogravimetric analysis of H2sLM, EusLM and TbsLM...... 117

Figure 4.13. Thermogravimetric analysis of H2LOM and TbsLOM...... 118

Figure 4.14. Thermogravimetric analysis of H2sLCHO and EusLCHO...... 118

Figure 4.15. Thermogravimetric analysis of H2sLCO and EusLCO...... 119

Figure 4.16. Thermogravimetric analysis of H2sLDS, EusLDS and TbsLDS...... 119

Figure 4.17. Luminescence spectrum of EusLM (solid) and TbsLM (solid). The broad peak at 307 nm on EusLM excitation, the broad peaks at 640 nm on EusLM and TbsLM emission are due to scattering...... 121

Figure 4.18. Luminescence spectrum of TbsLOM (solid). The broad peaks at 275 nm and 645 nm on TbsLOM excitation and emission respectively, are due to scattering...... 121

Figure 4.19. Luminescence spectrum of EusLCHO (solid). The peak at 308 nm is due to scattering...... 122

Figure 4.20. Luminescence spectrum of EusLDS (solid) and TbsLDS (solid). The peaks at 308 and

271 nm on EusLDS and TbsLDS excitation spectra respective, are due to scattering...... 122

xviii

Figure 4.21a. Room temperature fluorescence (RT FL, black), low temperature fluorescence (LT

FL, red) and low temperature phosphorescence (LT PH, blue) of the ammonium salt of H2LM. All emissions are measured with excitation at 320 nm...... 124

Figure 4.22. Intensity of Eu emission @ 614 nm in different suspension of EusLCHO...... 128

Figure 4.23. Excitation and emission spectrum of EuLCO suspension of various solvents...... 133

Figure 4.24. Emission intensity of EuLCO suspension in various solvents. Excitation at 400 nm was used and the average emission intensity at 481-485 nm was plotted as the y-axis...... 134

Figure 4.25. Emission spectra of water-soaked EusLDS suspended in various solvents...... 135

Figure 4.26. Maximum emission intensity of water-soaked EusLDS suspended in various solvents...... 136

Figure 4.27. Luminescence enhancement of water-soaked EusLDS after incubation with various solvent for 24h. “DMSO max” is the maximum enhancement achieved for incubation of DMSO

(after 72h). Luminescence enhancement is calculated as Iafter/Ibefore – 1 (Ibefore is the average intensity before incubation, Iafter is the intensity after incubation.)...... 137

1 Figure 4.28. H NMR of as-synthesized EusLCO (7.9 mg) in CD3OD...... 149

1 Figure 4.29. Control experiments for H NMR of EusLCO (only mesitylene in CD3OD)...... 150

1 Figure 4.30. H NMR of as-synthesized EusLDS (8.8 mg) in CD3OD ...... 151

1 Figure 4.31. H NMR of as-synthesized TbsLDS (8.1 mg) in CD3OD...... 152

1 Figure 4.32. Control experiments for H NMR of EusLDS and TbsLDS (only mesitylene in

CD3OD)...... 153

1 Figure 4.33. H NMR of water-soaked EusLDS...... 154

Figure 4.34. PXRD of as-synthesized EusLM and the simulated pattern from crystal structure...... 155

xix

Figure 4.35. PXRD of as-synthesized TbsLM and the simulated pattern from crystal structure...... 155

Figure 4.36. PXRD of as-synthesized TbsLOM and the simulated pattern from crystal structure...... 156

Figure 4.37. PXRD of as-synthesized EusLCHO and the simulated pattern from crystal structure...... 156

Figure 4.38. PXRD of as-synthesized EuLCO and the simulated pattern from crystal structure.157

Figure 4.39. PXRD of as-synthesized EusLDS and the simulated pattern from crystal structure...... 157

Figure 4.40. PXRD of as-synthesized TbsLDS and the simulated pattern from crystal structure...... 158

Figure 5.1. Ligand discussed in this chapter...... 160

Figure 5.2. Ligand (a) used in Ghosh’s study and the MOF structure: (b) Zn node; (c) 1D chains; (d) Channels and the nitrates located in them...... 161

Figure 5.3. Ligand (top) and the 2D grids in the MOF (bottom)...... 162

Figure 5.4. Crystal structure of sLN ester...... 168

Figure 5.5. Crystal structure of sLN-HNO3...... 169

Figure 5.6. Crystal structure of EusLN-NO3: (a) & (b) Eight-coordinate Eu center adopting bicapped trigonal prismatic geometry; (c) Bimetallic node bridged by four carboxylate; (d) Twisted tetrahedral geometry of each node, connected to four other nodes (ammonium side chain omitted for clarity); (e) Diamond shaped 1D channels along the c-axis; (f) Two-fold interpenetration in the framework...... 170

Figure 5.7. Crystal structure of EusLN-Cl(a): (a) & (b) Eight-coordinate Eu center adopting bicapped trigonal prismatic geometry; (c) Bimetallic node bridged by four carboxylate; (d) Twisted tetrahedral geometry of each node, connected to four other nodes (ammonium side chain

xx omitted for clarity); (e) Diamond shaped 1D channels along the c-axis; (f) Two-fold interpenetration in the framework...... 171

Figure 5.8. Crystal structure of EusLN-Cl(b): (a) & (b) Eight-coordinate Eu center adopting square anti-prismatic geometry; (c) Bimetallic node bridged by four carboxylate; (d) 2D grid in the MOF; (e) 2D grids stack on top of each other with an off set of half grid; (f) Two-fold interpenetration in the framework...... 172

Figure 5.9. PXRD of EusLN-NO3 ...... 174

Figure 5.10. PXRD of EusLN-Cl ...... 175

Figure 5.11. TGA of sLN-HNO3 and EusLN-NO3...... 176

Figure 5.12. Luminescent spectrum of EusLN-NO3. The peaks at 308 and 660 nm on the excitation and emission spectrum, respectively, are due to scattering...... 176

Figure 5.13. FTIR of EusLN-NO3 before and after KClO4 soaking...... 178

Figure 5.14. Excitation (ex) and emission (em) spectra of perchlorate exchanged EusLN-NO3 washed by EtOH or DMF...... 178

1 Figure 5.15. H NMR of EusLN-NO3 (7.5 mg, in acetone-d6) exchanged with KClO4 DMF solution...... 184

1 Figure 5.16. H NMR of the control experiment for EusLN-NO3 exchanged with KClO4 DMF solution ...... 185

Figure 6.1. Ligand systems discussed in this thesis...... 187

Figure 6.2. Other ligand candidates for MOF sensors...... 189

Figure 6.3. Other alternative ligand systems...... 190

xxi

List of Symbols and Abbreviations

MOF metal-organic framework

LnMOF lanthanide metal-organic framework

PSM post-synthetic modification

LMET ligand-to-metal energy transfer

PET photoinduced electron transfer

ICT internal charge transfer

EET electronic energy transfer

TGA themogravimetric analysis

PXRD powder X-ray diffraction

EA elemental analysis

THF tetrahydrofuran

DMF N,N’-dimethylformamide

DMSO dimethylsulfoxide

DCM dichloromethane

EtOH ethanol

MeOH methanol

MeCN acetonitrile

EtOAc ethyl acetate

xxii 1

Chapter 1 Introduction 1 1.1 Definition of MOF

Metal-organic frameworks (MOFs) are crystalline compounds constructed from metal ions/clusters and organic ligands. Coordination bonds are the major force that hold metal ions/clusters and organic ligands together, to form the infinite polymer-type structure. As a result, MOFs are sometimes termed coordination polymers (CPs). While some may argue that a framework needs to be either a 2D or 3D structure, 1D structures are also included in any discussions about MOFs in this thesis.

A well-known example of MOF (MOF-5)[1] by Yaghi group is shown in Figure 1.1, which is

synthesized from zinc nitrate and 1,4-benzenedicarboxylic acid (H2BDC) in DMF/chlorobenzene

mixture. The topologically octahedral nodes of MOF-5 framework are the Zn4O13 clusters,

consisting of four ZnO4 tetrahedrons sharing the central oxygen atom. These nodes are interconnected by the linear ditopic BDC ligands to form a 3D framework. It can be seen from the figure that the framework itself occupies only a limited amount of space; the rest is filled with free solvent molecules. Removal of the solvent can potentially leads to a very porous structure. In fact, according to the gas/liquid vapor sorption experiments, up to 61% of the volume in the desolvated MOF-5 can be occupied by guest molecules, demonstrating the highly porous nature of this material.

As one would imagine, the syntheses of MOFs are not just limited to MOF-5. Many metal ions have been used in constructing MOFs, including the most commonly used late transition metals Zn, Cu, Co, Cd, Ni, Fe;[2] rare earth metals Sc, Y and lanthanides;[3] early transition metal Zr;[4] and even main group metals Mg and Al[5] which typically have much lower coordination numbers and weaker coordination bond strength. Even more ligands have been used: bidentate, tridentate or tetradentate ligands with various shapes and coordinating groups, including carboxylate, pyridine nitrogen, phosphate, sulfate,[6] etc. Besides the coordinating groups on the ligand, which are used to connect with metal nodes, various functional groups can also be

2 incorporated on the ligand as well, to tune the properties of the resulting materials. In short, a potentially unlimited amount of combination can be chosen from to build MOFs.

Figure 1.1. MOF-5 crystal structure 1.2 MOF synthetic methods

1.2.1 Hydrothermal/Solvothermal synthesis

The synthetic methods used for the universe of MOFs are surprisingly limited. As MOFs are usually polymers insoluble in most solvents, characterization methods (X-ray crystallography) call for products in their crystalline form as well as requirement on the minimum size of the crystal, which makes it necessary to complete the formation of the product and crystallization process in one step. The most common methods have been the hydrothermal/solvothermal route: metal starting materials and ligand precursors are dissolved/suspended in a mixture of solvents in a sealed system and heated for typically a few days to yield the crystalline products.

These one-pot synthetic methods are straightforward and easy to apply; however, they are hard to monitor or study. As a result, little is understood about the process and researchers have to explore a variety of conditions to find the optimal synthetic procedure, including solvent

3 composition, temperature, concentration, reaction time, heating/cooling rate, etc. Since the major uncertainty in MOF development process lies in the time-consuming crystallization process, researchers tend to avoid ligands that requires substantial amount of synthetic efforts, despite the freedom in metal/ligand choices.

It is speculated that the high temperature and pressure in sealed systems greatly improve the solubility of reagents and reaction intermediates, promoting the formation of the thermodynamically favorable crystalline products rather than the amorphous precipitate; high solubility also leads to a reversible process and an overall slow product formation rate that is beneficial for obtaining large crystals suitable for X-ray crystallography. A detailed study on MOF-5 synthesis[7] has also shed some light on the reaction mechanism: the slow decomposition of the formamide solvent generates amine species, which lowers both proton activity and water concentration to favor the formation of MOF-5. Although the control of water concentration is somewhat unique to MOF-5 type systems containing the octahedral Zn4O13 nodes, we believe that the slow decomposition of formamide solvent is a common mechanism for MOF formation, especially in our systems featuring carboxylic acid ligands and formamide solvents as discussed in later chapters. Based on the aforementioned speculations and experimental observations, as well as the our own experience, we adopt the following general guideline for solvothermal synthesis to find initial crystallization conditions is – adjust solvent composition to improve solubility of starting materials and decrease reaction temperature when amorphous precipitates are obtained under certain conditions; and do the opposite if a clear solution is obtained.

1.2.2 Other synthetic methods

Besides hydro/solvothermal methods, there are also other alternatives to obtain crystalline MOF materials suitable for X-ray crystallography. While traditional hydro/solvothermal methods aim at creating homogeneous media for crystal growth, an alternative method intentionally utilizes two immiscible solvents to grow crystal at the interface.[8] The limited solubility serves as the control over the rate of reaction and crystallization. Besides this biphasic method, other synthetic approach also try to control the reaction rate, e.g., by slow vapor diffusion of base into the reaction mixture.[9]

With the development of new characterization techniques, preparation of the MOF is no longer limited to single crystals of a certain size. Instead, alternative methods that are rapid, cost-

4 effective and/or environmentally friendly have been developed, including mechanosynthesis,[10] microwave synthesis[11] and sonochemistry.[12] In liquid-assist mechanical grinding (LAG) method, [10] catalytic amount of solvent is used to improve the mobility of reactants, and the reaction system is comparable to nano-scale solvothermal synthesis.[13] Friščić and co-workers applied LAG method with small quantity of added solvents in MOF synthesis and successfully obtained various porous Zn MOFs.[14] All syntheses were completed quantitatively in 30 min at room temperature, much faster than typical solution-based methods which can take days or even weeks. They also showed that ZnO can be used in MOF synthesis, which is less expensive comparing to the more commonly used starting material Zn(NO3)2. Later studies by the same group further developed the method to use catalytic amount of salts to template the synthesis of [15] MOF [Zn2(ta)2(dabco)] within 20 min at room temperature, as opposed to the original method which requires 120 °C and two days.[16] Sonochemical methods apply ultrasonication instead of traditional heating to assist the synthesis of MOFs. The rapid creation, expansion and implosion of bubbles during the sonication process generate high local temperature and pressure which lead to rapid synthesis. For example, Ahn group applied sonochemical approach in the synthesis of MOF-5 which was previously obtained with solvothermal[17] or slow diffusion methods[1]. Reaction time of the synthesis is substantially reduced without sacrificing the crystallinity or sorption property.[12] Microwave irradiation has also been applied in MOF synthesis to replace traditional heating. Masel and coworkers successfully applied such methods in the synthesis of IRMOF1, IRMOF2 and IRMOF3. [17] They discovered that the synthesis can be completed within a few minutes to yield microcrystals with narrow size distribution, and the size can be further controlled by varying reagent concentration and changing reaction time accordingly.

Additionally, in order to meet the demand of practical applications, procedures to prepare MOF thin films are also being developed.[18] These procedures are usually based on the relative well- developed MOF systems, including MOF-5, HKUST-1, [Zn2(bdc)2(dabco)], and the methods used include modified solvothermal synthesis on different substrates, stepwise chemical deposition,[19] and electrochemical deposition.[20] Although the possibility of preparing MOF films have been demonstrated, the control over structural orientation is still difficult and requires further efforts before being applied to separation, catalysis or sensing applications.

5

1.2.3 Post-synthetic modification (PSM)

In most MOF synthetic strategies, crystallization step is the last step after ligands have been properly functionalized, while another approach is to crystallize the MOF first and then perform post-synthetic modification on the resulting MOFs. PSM refers to chemical modifications on the ligands in MOFs after they have been assembled. It is an alternative approach to obtaining MOFs with specific functionalities that are otherwise not accessible with tradition methods. Cohen has reviewed the topic of PSM and only a few selective examples will be discussed here.[21] Sada group utilized the convenient click chemistry in the post-synthetic modification process.[22] Carboxylic acid ligands bearing azide groups are first used to assemble MOFs with Zn ion, and the azide groups are subsequently reacted with substituted alkynes successfully to introduce ester group, hydroxyl group and alkyl chain. Rieger and co-workers developed a post-synthetic modification strategy based on terminal olefin groups.[23] A 4,4’-bipyridine derivative containing terminal olefin groups is first assembled into a MOF with Zn ion and 9,10-triptycenedicarboxylic acid ligand. The olefin moiety is then oxidized to epoxide by dimethyldioxirane, and further functionalized through ring opening reaction with ethyl mercaptan to incorporate hydroxyl and thioether functionalities. Photolabile protecting groups have also been used in PSM by Telfer group.[24] Carboxylic acid ligands containing 2-nitrobenzyl ether groups are used with Zn ion in the synthesis of MOF. After the synthesis, the ether groups can be cleaved through UV irradiation to generate hydroxyl groups.

The importance of these PSM discoveries relies not only in the fact that some of the aforementioned functionalities are difficult to synthesize (ester group that co-exist with carboxylic acid groups on the ligand) or may interfere with MOF crystallization process (acidic phenol hydroxyl group), PSM also offers access to MOFs with different functional groups on the exact same structural skeleton, as even a small modification on the ligands could lead to structural distortion due to sterics or interpenetration if the crystallization were to be done last. PSM however has its own limits. First, the constrained space in MOF pores and the fixed conformation of the ligands will prevent certain reactions with spacial requirements from happening. Secondly, the limited size of the MOF channels will slow down the diffusion of reagents/by-products and exclude reagents over certain sizes. Last but not least, synthetic methods used in PSM are also limited by the stability of MOFs, especially when reaction time is generally long due to the slow diffusion. As most MOFs are unstable in both acidic and basic

6 condition, many widely used protection/deprotection methods in organic chemistry are not applicable under MOF context.

1.3 MOF Structure and iso-reticular chemistry

So far, the field of MOF has yielded a large number of different structures resulted from the combination of metals and ligands. Despite the seemingly random structures, common features are observed among MOFs synthesized from similar ligands. The need to summarize such relations between metal node or ligand topology and the resulting MOF structures has led to the development of iso-reticular chemistry.[25] Iso-reticular chemistry studies the relationship between MOF structure and ligand/metal node topology. The goal of iso-reticular chemistry is to be able to obtain desired MOF structure through the careful selection of metal and ligand. Three of the most well-known systems will be discussed here. Although more iso-reticular systems are known,[26] many of the structures are complicated and thus hard to visualize without looking at the 3D structure.

One of the most widely applied iso-reticular systems is the MOF-5 derivative system (Figure

1.2). As discussed in early Section 1.1, the structure contains the octahedral Zn4O13 clusters (Figure 1.2a) as nodes, which are internected through the linear BDC ligands to form the cubic type structure. Yaghi and co-workers later discovered that the BDC ligands can be replaced with many different linear dicarboxylic acid ligand derivatives (Figure 1.2c), while still maintaining [17] the cubic-type structure. Extending the linear ligands leads to higher porosity, and the ability to introduce functional groups in the MOF without disturbing the original topology offers versatility on MOF properties.

7

Figure 1.2. MOF-5 type structure: (a) octahedral node; (c) examples of ligands used.

Another widely used system is the pillared paddle-wheel structures synthesized from Zn ion (Figure 1.3), linear dicarboxylate and bipyridine-type ligands.[27,28] Shown in Figure 1.3a is the node and the ligand connection around it. Each node contains two Zn2+, briged by four carboxylate groups from the dicarboxylate ligands, extending the structure to layers of infinite 2D grids. The axial coordination sites of the Zn2+ are occupied by N donors from the bipyridine ligands, which serve as pillars between each layer of the 2D grinds. Because metal nodes look like paddle-wheels, this type of structure is termed “pillared paddle-wheel structure”. Both the dicarboxylate ligands and the bipyridine ligands can be replaced by ligand with similar topology (Figure 1.3c,d), as suggested by the studies of the Kim group.[27] Application of this MOF structural system has yielded functional MOFs by introducing functionalized ligands. Hupp and coworkers demonstrated that, by introducing a catalytic functionality into the ligand as shown in Figure 1.3d (far right), the MOF can be used to catalyze certain reactions. This example will be discussed more in MOF catalysis in Section 1.4.2.

8

Figure 1.3. Pillared paddle-wheel type structure: (a) octahedral node; (c) examples of dicarboxylate ligands used; (d) examples of bipyridine type ligands used.

Besides linear ligands, structural systems containing trigonal ligands have also been studied and applized in MOF synthesis. HKUST-1 was first synthesized by Williams and co-workers, with copper nitrate and benzene-1,3,5-tricarboxylic acid (BTC). The structural node in HKUST-1 is the aforementioned “paddle-wheel” di-copper node but without coordinated water occupying the two axial sites rather than the N donors from pillar ligands (Figure 1.4a). Each node connects to eight more nodes through four of the trigonal ligands, yielding the 3D structure shown in Figure 1.4b. Two types of cage exist in the 3D structure. The bigger opening represents the big cage (Figure 1.4c), surrounded by 12 metal nodes (as the vertexes) interconnected by 8 ligands (as 8 of the 16 triangular facets). The more “filled” area next to the big cage is the small cage, surrounded by 6 metal nodes (as vertexes) interconnected by 4 ligands (as 4 of the 8 triangular facets). Because of these pores, HKUST-1 possesses high surface area. This structure was later modified by Yaghi and co-workers, by replacing the BTC ligands with 1,3,5-benzenetribenzoate (BTBC) ligands (yielding MOF-177), greatly increasing the pore size as well as the surface area.[29b] The Matzger further expand this structural series with a “vertex desymmetrization” strategy. They modified the trigonal ligands to break the trigonal symmetry and obtained MOFs with each of the ligand shown in Figure 1.4f & g. The nodes and connectivity are still the same as HKUST-1 and MOF-177, but with distorted shape overall. Such desymmetrization strategy may be applied to reduce self-interpenetration and improve the framework porosity.

9

Figure 1.4. HKUST-1 type structure: (a) paddle-wheel metal node; (b) channel view; (c) big cage; (d) small octahedral cage; (e)-(g) examples of symmetrical and asymmetrical ligands.

Although those iso-reticular MOF systems have been well studied and some are already applied in the design of functionalized MOF materials as will be seen in later discussion, they are among the limited examples of iso-reticular systems. Exploration of other iso-reticular systems is still an on-going field.

1.4 MOF functionalities and applications

Because of the flexible choices of metal and organic ligands used in MOF syntheses, the resulting MOFs may inherit the magnetic, optical, electrical properties and ligand functionalities from their building blocks. MOFs with wide variety of properties have been synthesized and their applications have been demonstrated in the field of gas storage, separation, catalysis, and sensing. New applications of MOF materials are also emerging as researchers explore new

10 metal-ligand combinations, such as MOF thermometer,[30] proton conducting MOFs for fuel cell application,[31] ferromagnetic MOF materials[32] and MOFs as biomedicine.[33]

1.4.1 Gas storage and separation

MOFs are attractive materials for gas storage and separation application because of its high surface area and flexible ligand choices which allows for fine-tuning of the MOF-guest interaction. As a result, gas storage and separation are arguably the most well-studied fields [34] [35] [36] [37] among all the MOF applications. Storage of H2, CO2, methane, acetylene in MOFs have been extensively studied and some have been summarized in a few review articles. Gas separation, which is closely related to gas adsorption (storage), has also been extensively studied.

Topic gas includes CO2, O2, H2, olefin, H2S, SO2, Cl2 and etc., and these studies have been covered by Zhou in a few review articles.[38]

1.4.2 MOF catalysis

MOF catalysts possess the same advantages as other heterogeneous catalysts, such as easy product separation and catalyst recovery. However, they also have their own distinct features. MOF catalysts do suffer from limited thermal and chemical stability compared with other inorganic porous materials (eg. zeolites), but they instead offers greater chemical variety than zeolite, as well as tunable and ordered porous structures. As a result, studies in MOF catalysis have been focusing on developing the size-, shape- and even enantio-selective MOF catalysts for the synthesis of small molecules under mild conditions.

Early examples of MOF catalysts are discovered with serendipity. Fujita and co-workers have found that the 2D MOF synthesized with Cd(NO3)2 and 4,4’-bipyridine can be used in Lewis acid catalyzed cyanosilylation of aldehydes, where the catalysis is proposed to occur via substrate displacement of axially coordinated nitrate ions.[39] Kitagawa group reported a MOF catalyst containing basic catalytic sites on MOF Cd(4-btapa)2(NO3)2 (4-btapa = 1,3,5-benzene tricarboxylic acid tris[N-(4-pyridyl)amide] ).[40] This Cd MOF can catalyze Knoevenagel condensation reaction of benzaldehyde with different nitriles, displaying some size-selectivity on substrates nitriles.

Recent advances in the field of MOF catalysts are achieved through heterogenization of homogeneous catalysts, taking advantages of the well-studied homogeneous catalytic systems.

11

Pioneering work by Lin and co-workers features a MOF constructed with Cu2+ and series of tetra-carboxylate ligands bearing chiral 1,1’-bi-2-naphthol (BINOL) moiety. PSM of the framework with Ti(OiPr)4 created chiral Ti sites, which are able to catalyze various diethylzinc/alkynylzinc addition to aromatic aldehydes with moderate to high enantiomeric excess.[41] Another well-developed homogeneous catalysis system that has been heterogenized is the chiral Mn(salen) system. Two different strategies have been developed independently by two groups. Hupp group heterogenized the Mn(salen) system by applying the pillared paddle-wheel structure.[28] The Mn(salen) core was derivatized with two pyridyl groups to a linear bipyridine type ligand, which was combined with Zn2+ and biphenyldicarboxylate to form the pillared paddle-wheel structure. Not only did the MOF catalyze asymmetric epoxidation reaction with moderate enantiomeric excess (e.e.), it also demonstrated improved stability compared with homogeneous system consisting of the free Mn(salen) ligand, possibly due to the elimination of intermolecular deactivation process achieved by the immobilization of catalytic centers. Lin group incorporated the Mn(salen) catalyst core into a MOF-5 type structure, where the Mn(salen) core was derivatized with carboxylate and varying spacer groups.[42] Different spacer and solvent template were used in crystallization process to tune the pore size, and moderate to good e.e. can be achieved in the epoxidation reaction.

The heterogenization strategy might seems straightforward to apply given the well-studied iso- reticular MOF systems; however, the use of metalloligands introduced selectivity issue between the two metal species (catalytic site and MOF node) in the crystallization process and considerable efforts are required when applying the heterogenization strategy to other homogeneous catalysis systems.

1.4.3 MOF sensors

As heterogeneous sensor materials, MOF sensors not only share with other heterogeneous sensors the same good recyclability and ease of device fabrication, they also offer the potential of many possible types of sensor-analyte interaction used in the homogeneous systems because of the flexible ligand/metal choices. In addition, the periodic porous structure could lead to extra size-selectivity and enhance the sensor-analyte interaction. The concept of MOF sensor is fairly straightforward: analyte interact with MOF to cause some measurable property change of the MOF, and such property change is converted back to provide information about the existence or

12 exact amount of the analyte. Various types of properties have been used as the sensing signal, including visible color, luminescence emission after irradiated, refractive index, conductivity (or resistance) and mechanical properties. MOF sensors utilizing these properties have been covered in a review article by Hupp et al.[43] Luminescence MOF sensors are directly related with this thesis and will be discussed in details in Section 1.4 and 1.5.

1.5 Luminescent sensors – homogeneous systems and MOFs

1.5.1 Homogeneous luminescent sensors

Sensors with visible luminescent signal change upon exposure to analytes continue to attract a lot of research interest, despite the large amount of sensing systems that have been developed, some of which have already made their way to real life applications. Those homogeneous luminescent sensor systems can be summarized into a few categories according to the sensing mechanism.

Photoinduced electron transfer (PET) is one type of mechanisms that is involved in many sensor systems. A typical PET type sensor consists of a fluorophore connected to a receptor without orbital overlaps. In an ideal PET type sensor, luminescence is not observed when the analyte is absent as the receptor can transfer an electron to the excited fluorophore to quench it. When analyte binds with the receptor, the redox potential of the receptor changes, rendering the electron transferring process energetically infeasible and therefore emission from the fluorophore is observed as it deactivates radiatively. The opposite behavior (analyte triggering turn-off) is still useful although not optimal. In PET type sensors, fluorophores are usually π-conjugated organic moieties such as naphthalene and anthracene, and the receptors are those that are known to bind strongly with the target analytes, e.g. amine/aniline for proton, crown ether or other macrocycles and their derivatives for alkali/alkaline earth cations, multidentate ligands with O/N donors for transition metal ions, and multiple hydrogen-bond donor for anions. Figure 1.5a shows one example of a PET type sensor, which displays fluorescence enhancement triggered by the binding of Li+.[44] Note that the receptor crown moiety is separated from the anthracene by the ethylene bridge.

13

Figure 1.5. Examples of PET type sensor (a) and ICT type sensor (b).

A second type of sensing mechanism involves internal charge transfer (ICT). In contrast with PET system, an ICT system consists of a fluorophore and a receptor that are integrated together to allow for orbital overlaps. Excitation of the molecule results in the redistribution of electron density. Such redistribution is affected upon binding of analytes, leading to a shift in the maximum of emission from the respective excited states. ICT systems and PET systems share the similar fluorophores and receptors designs, differs only in the connection in between. Shown in Figure 1.5b is a sensor example featuring ICT mechanism.[45] Binding of Zn2+ quenches the fluorescence in acetonitrile. Note that the terpyridine receptor is conjugated with the pyrene fluorophore unit.

A third type of sensing systems includes electronic energy transfer system (EET). An EET system consists of a pair of fluorophores. The donor fluorophore in its excited states can transfer energy to an acceptor fluorophore (quencher) via dipole-dipole coupling, the efficiency of which is sensitive to the separation of the two fluorophore. Any interaction between analyte and the system that can alter the distance can lead to the change in observed emission. A good example of EET system is reported by Kramer and coworkers,[46] where the fluorophore-quencher pair is held together initially through the intramolecular interaction of the nucleotide that it is attached to, resulting in no fluorescence. Once the nucleotide strand hybridize with the target DNA/RNA segment, the pair is separated, leading to the turn-on of the fluorescence.

Other sensor systems include excimer/exciplex type systems, where target analyte is involved in the formation or interruption of excimer/exciplex interaction accompanied with fluorescence change,[47] transition metal complex systems where the metal-to-ligand charge transfer (MLCT)

14 excited states is involved, and lanthanide complex systems featuring ligand-to-metal energy transfer (LMET).

1.5.2 Luminescent MOF sensors

The concept of heterogenization can also been seen in the design of MOF sensors, to take advantage of the knowledge gained in the homogeneous sensor studies. However, only those systems with relative simple ligands have been reported, presumably due to the difficulties in crystallization process.

The most common luminescent MOF sensor mechanism involves analyte-ligand interactions that influence ligand-based emission signal, and the metal ions/clusters in the MOF do not participate in the sensing process but act as pure structure node. Rosseinsky and co-workers designed a tetra-carboxylate ligand featuring a pyrene core and assembled the MOF along with In ions.[48] A fine-structure emission was observed for the as-synthesized MOF, as channel DMF molecules separate the pyrene ligands, minimizing self-quenching phenomena. Removal of DMF increases the structural disorder and triggers red-shifting of the emission, while the introduction of toluene guest reduces the emission intensity due to the strong π-stacking effect, as suggested by diffraction studies. Lin group reported a chiral MOF for sensing chiral amino .[49] The optically pure 1,1’-bi-2-naphthol (BINOL), which has been reported as enantioselective fluorescence sensor for various chiral compounds,[50] is incorporated into the ligand design. In the MOF constructed from BINOL derivatives and Cd ions, the two hydroxyl groups can selectively hydrogen bond with amino alcohols, leading to the decrease in emission intensity to varying extend depending on the nature of the analytes. Kitagawa group demonstrated a naked- eye detectable MOF sensor for aromatic compounds, using Zn2+, benzenedicarboxylic acid ligands and a bipyridine type ligands containing 1,4,5,8- naphthalenediimide (NDI) moiety. NDI is known to interact with aromatic compounds to trigger exciplex emission of the charge transfer type.[51] The reported MOF can interact with various aromatic compounds and display emission with varying colors, depending on the electronic properties of the analytes. Thiol functionalized BDC ligands have been designed by Xu and co-workers in an effort to construct MOF sensors for heavy metal ions.[52] The MOF constructed from thiol BDC and Zr4+ uptake Hg2+/Hg preferably as the thiol group reacts with Hg2+ or Hg, accompanying with the quenching of ligand-based luminescence which can be used as a sensing signal.

15

In the second type of MOF sensor materials, metal ions are involved in the sensing process as well.[53] Lin and co-workers synthesized a novel bimetallic MOF containing both Zn and Ir. The synthesis is divided to two steps: IrCl3 ∙ 3H2O is first reacted with methyl 3-(2-pyridyl)benzoate followed by hydrolysis to form the Ir complex; the carboxylic acid groups of this complex is later used to construct MOF with Zn(NO3)2 ∙ 6H2O under solvothermal condition. The resulting

MOF is luminescent, and its emission was found to be quenched by O2 in a reversible fashion, similar to other homogeneous cyclometalated Ir complexes. Although one could consider such mechanism to involve only the metalloligand (cyclometalated Ir complex with carboxylate groups) but not the structural node (Zn cluster), it still represents one type of MOF sensor unique to the previous examples.

Other MOF sensors systems where metal ions are involved are lanthanide MOFs (LnMOFs). Features of LnMOF sensors will be the topic of the next section.

1.6 Lanthanide luminescence MOF sensors

Lanthanide MOFs are fairly different from the well-studied transition metal MOFs in both their coordination chemistry and luminescence properties. Because of the high charge and small ion radii, lanthanide coordination bonds are mostly ionic in nature. Lanthanide ions have poorly defined coordination geometry and variable coordination number ranging from 3 to 12,[55] depending on the steric of the ligands, with 7 to 9 being the mostly common. As a result, MOF nodes built with lanthanide ions usually do not share the same composition or geometry, leading to a large variety of metal nodes observed so far. Such coordination chemistry of lanthanide ions presents additional challenges in MOF sensor designs. No iso-reticular LnMOF systems have been developed so far, as the flexibility in coordination number and geometry makes it almost impossible to maintain the same MOF topology with different functional ligands. As a result, systematic studies on LnMOFs with ligand series bearing different functionality are limited.

Many lanthanide ions emit visible light if properly excited, but their luminescent properties are distinct from other materials. On one hand, their 4f electrons are shielded by the filled 5s25p6 sub-shell, energy level of different f-electron configurations are hardly affected by the coordination environment, resulting in sharp and characteristic emission from the respective f-f transitions. On the other hand, direct f-f transitions are symmetry forbidden; ligand-to-metal energy transfer (LMET) is required for efficient excitation of lanthanide ions. The LMET

16 process in a typical lanthanide complex is shown in Figure 1.6. The ligands serve as chromophores: they absorb light and are excited into singlet state, which then undergoes intersystem crossing into ligand triplet states. The triplet states can then transfer energy to lanthanide ions, which subsequently decay through radiative pathway to emit light. In LMET process, ligand singlet-triplet state gap and the ligand triplet state energy level have great impact on the LMET efficiency. Optimized energy level for efficient LMET process has been well- studied in homogeneous systems:[54] a minimum singlet-triplet state gap of 5000 cm-1 is necessary for efficient intersystem crossing, and the optimal ligand triplet state energy level lies between 2500 and 3500/4000 cm-1 above the respective Eu/Tb excited states (17374 and 20455 cm-1 respectively). However, application of these numbers in designing luminescent LnMOFs is not straightforward. In homogeneous systems, the measurements of ligand energy levels are to be conducted with analogous Gd complexes that closely resemble the conformation of the Eu/Tb complexes, as the conformation of ligands affects their energy levels. If this methodology is to be applied to MOF systems, Gd MOFs have to be made first, which undermines the predicting power as similar conditions usually yield Eu and Tb MOFs as well (where the luminescence properties can be directly measured). Given the time-consuming crystallization process, predictions based on the properties of free ligands will be ideal, even at the cost of accuracy.

Figure 1.6. Ligand-to-metal energy transfer mechanism in lanthanide complexes.[55]

Due to the unique luminescent properties, Lanthanide MOFs are attractive candidates for sensing application. Their unique LMET mechanism allows the emission modulation to be triggered by

17 both analyte-metal and analyte-ligand interaction, as opposed to only the latter interaction seen in their transition counterparts. Because LMET efficiency is sensitive to changes in ligand energy level, weak analyte-ligand interaction may be translated to significant change in lanthanide emission intensity. In addition, the characteristic emission peaks simplify the detecting device: only a single parameter - the intensity at a fixed wavelength - needs to be measured, comparing to the transition metal counterparts where the full spectrum needs to be collected to identify the shift in emission maximum. The use of a single parameter also facilitates the quantitative analysis.

Before moving on to LnMOF sensor examples, it is worth noting that many homogeneous lanthanide complexes have been developed for sensing application,[56] including cation sensors and anion sensors. The sensing mechanism is centered on the LMET process and a few instances of Ln complex used are shown in Figure 1.7. Cationic sensor usually features two multi-dentate binding pockets, one for Ln3+ and one for cationic analyte. Binding of the analyte will affect the LMET process which enhances or quenches the emission. Anionic sensor either takes advantage of the direct binding of the anion to the Ln3+ metal which adds/removes quencher molecules around the metal, affecting the emission intensity and lifetime,[57] or incorporate antennas whose single/triplet state can be quenched through the interaction with analyte.[58]

18

Figure 1.7. Examples of lanthanide complexes sensors for cationic (a, b) and anionic (c, d) analytes.

LnMOF-based sensors are relatively unexplored compared to their transition metal counterparts and only a limited number of sensors have been reported so far. These sensors can be divided to two groups based on the sensing mechanism. One type of mechanism involves changes on the metal coordination environment that affects possible non-radiative transition pathways. O-H, N- H and C-H bonds are known to quench Eu/Tb emission when present in the first coordination sphere, as their vibrational modes can accept energy from Eu/Tb excited states.

Chen has reported an Eu MOF sensor which displays varying emission intensity in different [59] solvent suspensions. The subject MOF, Eu(BTC)(H2O) ∙ 1.5 H2O (BTC = benzene-1,3,5- tricarboxylate, Figure 1.8), is synthesized from solvothermal reaction between Eu(NO3)3 ∙ 6H2O and H3BTC in the mixture of DMF, ethanol and water. The resulting framework features 1D channels filled with coordinated and lattice water. The water content can be removed at elevated temperature and the de-solvated MOF displays characteristic Eu emission with varying intensity when suspended in different solvents, with DMF suspension giving the strongest emission while acetone almost quenches the emission completely. As proposed by the author, the removal of

19 coordinated water on Eu generates open metal sites that can interact with different solvent molecules to trigger changes in emission intensity.

Another MOF sensor developed by the Chen group is Tb(BTC), synthesized from solvothermal [60] reaction of Tb(NO3)3 ∙ 6H2O and H3BTC in the mixture of DMF, ethanol and water. Tb(BTC) displays an almost identical structure to the aforementioned Eu(BTC) (H2O) ∙ 1.5 H2O, with similar 1D channels filled with solvents. When the MOF is suspended in methanol solution of different sodium salts, Tb emission is enhanced to a varying degree, with NaF triggering the most turn-on. Such luminescence change is also quite sensitive that NaF concentration as low as 10-5 M can trigger significant changes. The author speculated that fluoride can hydrogen bond with the proton of the coordinated methanol, which reduces the quenching ability of methanol O- H bond.

The Rocha group reported an Eu MOF (ITQMOF-1) as ethanol vapor sensor.[61] ITQMOF-1 is synthesized from EuCl3 and 4,4’-(1,1,1,3,3,3-hexafluoropropane-2,2-diyl)dibenzoic acid (HFIPBB, Figure 1.8) under solvothermal condition. Although the crystal structure cannot be solved due to strong twinning, porosity of desolvated sample is demonstrated through gas sorption experiments. The author found that ITQMOF-1 can respond to ethanol vapor rapidly, resulting in turn-off of the Eu emission. Such turn-off can be reverted quickly as ethanol is removed from the air stream, and the on-off cycle can be repeated for a few times. The sensing mechanism is believed to involve the quenching ability of ethanol -OH group.

The second type of mechanism involves analyte-ligand interactions that affect the LMET efficiency. Incorporation of interaction in MOFs seems straightforward at a glance, as different binding groups may be built into the organic ligands used to construct the MOFs. However, complex ligands add to the difficulties of crystallization, and most known examples are achieved with simple ligands.

Wong and co-workers constructed a Tb MOF using mucic acid and TbCl3 in a liquid phase slow diffusion method.[62] The MOF contains 2D sheets hydrogen bonded together via free water, with square-shaped channels running perpendicular to the 2D sheets. While the desolvation of the framework leads to the emission enhancement due to the removal of quencher water, further increase in luminescent intensity is observed with addition of aqueous salt solutions to the

20 desolvated sample. Such turn-on effect is ascribed to the hydrogen bonding interaction between the anions and the hydroxyl groups on the ligand.

Lu reported the synthesis and sensing properties of K5[Tb5(IDC)4(ox)4] ∙ 20H2O (IDC = imidazole-4,5-dicarboyxlate, and ox = oxalate, Figure 1.8).[63] This MOF is synthesized from hydrothermal reaction between Tb(NO3)3 ∙ 6H2O, H3IDC, H2ox and KOH. The framework is anionic, and K+ serves as the counter ion located in the channels. It is discovered in the cation exchange experiments that the Tb emission intensity is greatly enhanced when Ca2+ is present. The presence of 3 equivalent of Ca2+ has led to 100% increase in luminescent emission, which is ascribed by the author to the strong interaction between Ca2+ and oxalate which improves the LMET efficiency by reducing vibration-induced deactivation from oxalate.

Figure 1.8. Ligand mentioned in LnMOF sensor examples.

Zhang and co-workers designed an Eu MOF, incorporating a carboxylate ligand containing [64] photoactive viologen moiety. [Eu(BA)(Bpybc)1.5(H2O)] ∙ 2NO3 ∙ 5H2O (HBA = benzoic acid,

H2BpybcCl2 = 1,1’-bis(4-carboxybenzyl)-4,4’-bipyridinium dichloride, Figure 1.8) is synthesized from slow evaporation method from the precursor solution containing H2BpybcCl2, sodium benzoate and EuNO3. The photoactive viologen moiety was found to affect the MOF emission property: with UV irradiation, the resulting MOF can be converted to its reduced form which shows weak emission; upon subsequent exposure to atmospheric dioxygen, the MOF will return to its neutral form which gives enhanced emission.

21

Among all the reported LnMOF sensors, detailed mechanistic studies have been lacking. Major efforts in the field have been focusing on the exploration of new sensing materials, and their mechanisms are usually proposed according to similar homogeneous systems, as the experimental methods to directly observe weak analyte-MOF interaction are limited. In this thesis, luminescence lifetime experiments were introduced into the MOF systems to assist the mechanistic study on analyte-metal interaction for the first time. As mentioned in earlier, O-H bond is a good quencher for Eu emission. According to an early study,[65] the luminescence quenching rate contributed by O-H bonds is proportional to the number of coordinating water molecules in the first coordination sphere of Eu and this quenching rate displays such a strong isotope effect that O-D bond essentially has no quenching effect. The average number of coordinating water molecules per Eu center can be calculated using equation 1.1, where τH2O and

τD2O are the lifetimes of Eu emission in ms, for H2O and D2O containing samples, respectively. When O-H bonds are also present, the equation 1.2 is used to account for the contribution of coordinated water and alcohol, which is based on the fact that the quenching ability of an alcohol O-H bond is approximately 45% of water.[66,67] By monitoring lifetime of Eu emission, changes in the number of coordinated water/alcohols can be determined at various stage of the sensing experiments. Also, the luminescent signal change contributed from factors other than the quenching on the metal centers can be isolated when deuterated solvents are used in the experiments, allowing the evaluation of analyte-ligand interaction.

−1 -1 Eq 1.1. nH2O = 1.05 (τH2O −τD2O )

−1 -1 Eq 1.2. nH2O + 0.45 nROH= 1.05 (τH2O −τD2O )

22

1.7 Scope of the thesis

Figure 1.9. An early MOF developed in our lab

The project described in this thesis is inspired from the early work in our lab.[9] The MOF is constructed with Eu3+ and the L ligand with acetylene spacer under slow vapor diffusion method. The structure of the MOF features the triangular channels shown in Figure 1.9 and the Eu emission is found to be quenched by I2. As the central acetylene did not display any sensing interaction, we decided to replace the acetylene spacer with C=C double bond (Figure 1.10) which will allow us to test out more functional groups. Chapter 2 described our initial discoveries on this ligand series. This work has been published on CrystEngComm. [68]

Figure 1.10. Ligand systems covered in this thesis.

23

After we had done some exploration on the LX series, we discovered the inherent disadvantages regarding the low efficiency of synthesis as well as the low energy level which leads to the low

LMET efficiency. Such disadvantages have prompted us to move to the sLX ligand series.

Chapter 3 describes EusLOM, one of the MOFs made within this ligand series. This MOF was discovered to display strong solvent dependent luminescence and we were able to evaluate its sensing ability and unravel the underlying sensing mechanism. This work has been published in Angewante Chemie International Edition.[69]

Given the success in the EusLOM system, we decided to fully explore this ligand system. Chapter 4 includes the synthesis and characterization of various functionalized ligands and the respective

MOFs. Among the ligands studied, there are the plain version of the ligands with X = -CH3, and other functional groups that are π-conjugated with the ligand backbone phenyl ring systems, as our strategy to enhance the luminescence response interaction. With this family of MOFs, we were able to uncover some of the relationship between ligand and MOF properties. Different solvent sensing behaviors were also observed. The discoveries will be useful in guiding the future work of MOF sensors in our lab. The manuscript of this chapter has been submitted for publication.

In Chapter 5, a positively charged ligand and the respective MOFs are discussed. The introduction of the ammonium groups represents another strategy to enhance the analyte-ligand interaction, through the electrostatic interaction. Despite the inherent problem related to the solubility of the ligand, we were able to obtain a few MOFs with different counter anion and evaluate their properties preliminarily. All the results and discovery of the project are summarized in the last chapter, with some future directions given.

24

References

1. H. Li, M. Eddaoudi, M. O’Keeffe, O. M. Yaghi Nature, 1999, 402, 276-279

2. (a) N. L. Rosi, J. Kim, M. Eddaoudi, B. Chen, M. O'Keeffe, O. M. Yaghi, J. Am. Chem. Soc. 2005, 127, 1504-1518; (b) B. Chen, M. Eddaoudi, S. T. Hyde, M. O'Keeffe, O. M. Yaghi, Science, 2001, 291, 1021-1023; (c) J. An, S. J. Geib, N. L. Rosi, J. Am. Chem. Soc. 2010, 132, 38-39; (d) Y.-Q. Huang, B. Ding, H.-B. Song, B. Zhao, P. Ren, P. Cheng, H.- G. Wang, D.-Z. Liao, S.-P. Yan, Chem. Commun. 2006, 47, 4906-4908; P. Maniam, N. Stock, Inorg. Chem. 2011, 50, 5085-5097; (e) S. B. Choi, M. J. Seo, M. Cho, Y. Kim, M. K. Jin, D.-Y. Jung, J.-S. Choi, W.-S. Ahn, J. L. C. Rowsell, Cryst. Growth Des. 2007, 7, 2290-2293. 3. (a) J. Perles, N. Snejko, M. Iglesias, M. A. Monge, J. Mater. Chem. 2009, 19, 6504-6511; (b) K. A. White, D. A. Chengelis, K. A. Gogick, J. Stehman, N. L. Rosi, S. Petoud, J. Am. Chem. Soc. 2009, 131, 18069-18071. 4. A. Schaate, P. Roy, T. Preusse, S. J. Lohmeier, A. Godt, P. Behrens, Chem. Eur. J. 2011, 17, 9320-9325. 5. (a) Z. Guo, G. Li, L. Zhou, S. Su, Y. Lei, S. Dang, H. Zhang, Inorg. Chem. 2009, 48, 8069-8071; (b) H. Reinsch, M. Kruger, J. Wack, J. Senker, F. Salles, G. Maurin, N. Stock, Microporous Mesoporous Mater. 2012, 157, 50-55. 6. G. K. H. Shimizu, R. Vaidhyanathan, J. M. Taylor, Chem. Soc. Rev. 2009, 38, 1430-1449. 7. S. Hausdorf, F. Baitalow, J. Seidel, F. O. R. L. Mertens, J. Phys. Chem. A 2007, 111, 4259-4266. 8. P. M. Forster, P. M. Thomas, A. K. Cheetham, Chem. Mater. 2002, 14, 17-20. 9. B. T. Nguyen Pham, L. M. Lund, D. Song, Inorg. Chem. 2008, 47, 6329-6335 10. T. Friščić, J. Mater. Chem. 2010, 20, 7599-7605. 11. (a) Z. Ni, R. I. Masel, J. Am. Chem. Soc. 2006, 128, 12394-12395; (b) S. H. Jhung, J.-H. Lee, J. W. Yoon, C. Serre, G. Férey, J.-S. Chang, Adv. Mater. 2007, 19, 121-124. 12. W.-J. Son, J. Kim, J. Kim, J., W. –S. Ahn, Chem. Commun. 2008, 6336-6338. 13. V. V. Boldyrev, Powder Technol. 2002, 122, 247. 14. T. Friščić, L. Fábián, CrystEngComm 2009, 11, 743745. 15. D. N. Dybtsev, H. Chun, K. Kim, Angew. Chem. Int. Ed. 2004, 43, 5033-5036. 16. T. Friščić, D. G. Reid, I. Halasz, R. S. Stein, R. E. Dinnebier, M. J. Duer, Angew. Chem. Int. Ed. 2010, 49, 712-715. 17. M. Eddaoudi, J. Kim, N. Rosi, D. Vodak, J. Wachte, M. O'Keeffe, O. M. Yaghi, Science 2002, 295, 469-472. 18. D. Zacher, O. Shekhah, C. Wöll, R. A. Fisher, Chem. Soc. Rev. 2009, 38, 1418-1429. 19. O. Shekhah, H. Wang, S. Kowarik, F. Schreiber, M. Paulus, M. Tolan, C. Sternemann, F. Evers, D. Zacher, R. A. Fischer, C. Wöll, J. Am. Chem. Soc. 2007, 129, 15118-15119.

25

20. M. Li, M. Dincă, J. Am. Chem. Soc. 2011, 133, 12926-12929. 21. (a) Z. Wang, S. M. Cohen, Chem. Soc. Rev. 2009, 38, 1315-1329; (b) S. M. Cohen, Chem. Rev. 2012, 112, 970-1000. 22. Y. Goto, H. Sato, S. Shinkai, K. Sada, J. Am. Chem. Soc. 2008, 130, 14354. 23. K. Hindelang, S. I. Vagin, C. Anger, B. Rieger, Chem. Commun. 2012, 48, 2888-2890. 24. R. K. Deshpande, G. I. N. Waterhouse, G. B. Jameson, S. G. Telfer, Chem. Commun. 2012, 48, 1574-1576. 25. (a) M. O'Keeffe, M. A. Peskov, S. J. Ramsden, O. M. Yaghi, Acc. Chem. Res. 2008, 41, 1782-1789; (b) N. W. Ockwig, O. Delgado-Friedrichs, M. O'Keeffe, O. M. Yaghi, Acc. Chem. Res. 2005, 38, 176-182; (c) M. O'Keeffe, Chem. Soc. Rev. 2009, 38, 1215-1217; (d) D. J. Tranchemontagne, J. L. Mendoza-Cortés, M. O'Keeffe, O. M. Yaghi, Chem. Soc. Rev. 2009, 38, 1257-1283; (e) M. O'Keeffe, O. M. Yaghi, Chem. Rev. 2012, 112, 675-702. 26. H. Furukawa, K. E. Cordova, M. O’Keeffe, O. M. Yaghi, Science 2013, 341, 974-985 27. (a) B. Chen, S. Xiang, G. Qian, Acc. Chem. Res. 2010, 43, 1115-1124; (b) H. Chun, D. N. Dybtsev, H. Kim, K. Kim, Chem. Eur. J. 2005, 11, 3521-3529. 28. S. Cho, B. Ma, S. T. Nguyen, J. T. Hupp, T. E. Albrecht-Schmit, Chem. Commun. 2006, 2563-2565. 29. (a) S. S.-Y. Chui,S. M.-F. Lo, J. P. H. Charmant, A. G. Orpen, I. D. Williams, Science, 1999, 283, 1148-1150; (b) H. K. Chae, D. Y. Siberio-Perez, J. Kim, Y. Go, M. Eddaoudi, A. J. Matzger, M. O’Keeff, O. M. Yaghi, Nature, 2004, 427, 523-527; (c) J. K. Schnobrich, O. Lebel, K. A. Cychoze, A. Dailly, A. G. Wong-Foy, A. J. Matzger, J. Am. Soc. Chem. 2010, 132, 13941-13948. 30. Y. Cui, H. Xu, Y. Yue, Z. Guo, J. Yu, Z. Chen, J. Gao, Y. Yang, G. Qian, B. Chen, J. Am. Chem. Soc. 2012, 134, 3979-3982. 31. (a) J. M. Taylor, K. W. Dawson, G. K. H. Shimizu, J. Am. Chem. Soc. 2013, 135, 1193- 1196; (b) S. Kim, K. W. Dawson, B. S. Gelfand, J. M. Taylor, G. K. H. Shimizu, J. Am. Chem. Soc. 2013, 135, 963-966; (c) J. M. Taylor, R. K. Mah, C. I. Ratcliffe, I. Moudrakovski, R. Vaidhyanathan, G. K. H. Shimizu, J. Am. Chem. Soc. 2010, 132, 14055-14057; (d) J. A. Hurd, R. Vaidhyanathan, V. Thangadurai, C. I. Ratcliffe, I. L. Moudrakovski, G. K. H. Shimizu, Nature Chem. 2009, 1, 705-710. 32. W. Zhang, R.-G. Xiong, Chem. Rev. 2012, 112, 1163-1195. 33. P. Horcajada, R. Gref, T. Baati, P. K. Allan, G. Maurin, P. Couveru, G. Férey, R. E. Morris, C. Serre, Chem. Rev. 2012, 112, 1232-1268. 34. L. J. Murray, M. Dincă, J. R. Long Chem. Soc. Rev. 2009, 38, 1294-1314. 35. K. Sumida, D. L. Rogow, J. A. Mason, T. M. McDonald, E. D. Bloch, Z. R. Herm, T.-M. Bae, J. R. Long, Chem. Rev. 2012, 112, 724-781. 36. (a) M. Fernandez, T. K. Woo, C. E. Wilmer, R. Q. Snurr, J. Phys. Chem. C 2013, 117, 7681-7689; (b) M. Eddaoudi, J. Kim, N. Rosi, D. Vodak, J. Wachter, M. O’Keeffe, O. M. Yaghi, Science 2002, 295, 469-472; (c) H. Wu, W. Zhou, T. Yildirim, J. Am. Chem. Soc. 2009, 131, 4995-5000; (d) H. Wu, J. M. Simmons, Y. Liu, C. M. Brown, X.-S. Wang, S. Ma, V. K. Peterson, P. D. Southon, C. J. Kepert, H.-C. Zhou, T. Yildirim, W. Zhou,

26

Chem. Eur. J. 2010, 16, 5205-5214; (e) A. W. Thornton, K. M. Nairn, J. M. Hill, A. J. Hill, M. R. Hill, J. Am. Chem. Soc. 2009, 131, 10662-10669; (f) M. Gallo, D. Glossman- Mitnik, J. Phys. Chem. C 2009, 113, 6634-6642. 37. (a) S. Xiang, W. Zhou, J. M. Gallegos, Y. Liu, B. Chen, J. Am. Chem. Soc. 2009, 131, 12415-12419; (b) Y. Hu, S. Xiang, W. Zhang, Z. Zhang, L. Wang, J. Bai, B. Chen, Chem. Commun. 2009, 48, 7551-7553; (c) Y.-S. Xue, Y. He, S.-B. Ren, Y. Yue, L. Zhou, Y.-Z. Li, H.-B. Du, X.-Z.You, B. Chen, J. Mater. Chem. 2012, 22, 10195-10199. 38. (a) J.-R. Li, J. Sculley, H.-C. Zhou, Chem. Rev. 2012, 112, 869-932; (b) J.-R. Li, R. J. Kuppler, H.-C. Zhou, Chem. Soc. Rev. 2009, 38, 1477-1504. 39. M. Fujita, Y. J. Kwon, S. Washizu, K. Ogura J. Am. Chem. Soc. 1994, 116, 1151-1152. 40. S. Hasegawa, S. Horike, R. Matsuda, S. Furukawa, K. Mochizuki, Y. Kinoshita, S. Kitagawa, J. Am. Chem. Soc. 2007, 129, 2607-2614. 41. L. Ma, J. M. Falkowski, C. Abney, W. Lin, Nat. Chem. 2010, 2, 838-846. 42. F. Song, C. Wang, J. M. Falkowski, L. Ma, W. Lin J. Am. Chem. Soc. 2010, 132, 15390. 43. L. E. Kreno, K. Leong, O. K. Farha, M. Allendorf, R. P. Van Duyne, J. T. Hupp, Chem. Rev. 2012, 112, 1105. 44. J. P. Geue, N. J. Head, D. Ward, S. F. Lincoln, Dalton Trans. 2003, 521-526. 45. A. C. Benniston, A. Harriman, D. J. Lawrie, A. Mayeux, K. Rafferty, O. D. Russel, Dalton Trans. 2003, 4762-4769. 46. S. Tyagi, F. R. Kramer, Nature Biotechnol. 1996, 14, 303-308. 47. Y. Kanekiyo, R. Naganawa, H. Tao, Chem. Commun. 2004, 1006-1007. 48. K. C. Stylianou, R. Heck, S. Y. Chong, J. Bacsa, J. T. A. Jones, Y. Z. Khimyak, D. Bradshaw, M. J. Rosseinsky, J. Am. Chem. Soc. 2010, 132, 4119-4130. 49. M. M. Wanderley, C. Wang, C.-D. Wu, W. Lin, J. Am. Chem. Soc. 2012, 134, 9050-9053 50. L. Pu, Acc. Chem. Res. 2012, 45, 150. 51. Y. Takashima, V. M. Martínez, S. Furukawa, M. Kondo, S. Shimomura, H. Uehara, M. Nakahama, K. Sugimoto, S. Kitagawa, Nat. Commun. 2011, 2, 168. 52. K.-K. Yee, N. Reimer, J. Liu, S.-Y. Cheng, S.-M. Yiu, J. Weber, N. Stock, Z. Xu, J. Am. Chem. Soc. 2013, 135, 7795-7798. 53. Z. Xie, L. Ma, K. E. deKrafft, A. Jin, W. Lin, J. Am. Chem. Soc. 2010, 132, 922-923. 54. G. Liu, B. Jacquier, eds. Spectroscopic properties of Rare Earth in Optical Materials, 2005, Springer. 55. S. Cotton. Lanthanide and Actinide Chemistry, 2006, John Wiley & Sons. 56. C. M. G. dos Santos, A. J. Harte, S. J. Quinn, T. Gunnlaugsson, Coord. Chem. Rev. 2008, 252, 2512-2527. 57. Y. Bretonnière, M. J. Cann, D. Parker, R. Slater, Org. Biomol. Chem. 2004, 2, 1624. 58. R. S. Dickins, T. Gunnlaugsson, D. Parker, R. D. Peacock, Chem. Commun. 1998, 1777- 1778.

27

59. B. Chen, Y. Yang, F. Zapata, G. Lin, G. Qian, E. B. Lobkovsky, Adv. Mater. 2007, 19, 1693-1696 60. B. Chen, L. Wang, F. Zapata, G. Qian, E. B. Lobkovsky, J. Am. Chem. Soc. 2008, 130, 6718-6719 61. B. V. Harbuaru, A. Corma, F. Rey, P. Atienzar, J. L. Jordá, H. García, D. Ananias, L. D. Carlos, J. Rocha, Angew. Chem. Int. Ed. 2008, 47, 1080-1083. 62. K.-L. Wong, G.-L. Law, Y.-Y. Yang, W.-T. Wong, Adv. Mater. 2006, 18, 1051-1054. 63. W.-G. Lu, L. Jiang, X.-L. Feng, T.-B. Lu, Inorg. Chem. 2009, 48, 6997-6999. 64. J.-K. Sun, L.-X. Cai, Y.-J. Chen, Z.-H. Li, J. Zhang, Chem. Commun. 2011, 47, 6870- 6872. 65. W. D. Horrocks Jr., D. R. Sudnick, J. Am. Chem. Soc. 1979, 101, 334-340. 66. M. Steppert, I. Císařová, T. Fanghänel, A. Geist, P. Lindqvist-Reis, P. Panak, P. Štĕpnička, S. Trumm and C. Walther, Inorg. Chem., 2012, 51, 591-600. 67. R. M. Supkowski, W. D. Horrocks, Jr., Inorg. Chim. Acta. 2002, 340, 44-48. 68. Y. Li, D. Song, CrystEngComm. 2011, 13, 1821-1830. 69. Y. Li, S. Zhang, D. Song, Angew. Chem. Int. Ed. 2013, 52, 710-713.

28

Chapter 2 Lanthanide Metal-Organic Framework of Functionalized (E)-4,4’- (ethene-1,2-diyl)dibenzoic acids 2 2.1 Abstract

The functionalized ligand (E)-4,4’-(1,4-bis(methylthio)but-2-ene-2,3-diyl)dibenzoic acid (H2LS) 3+ and four metal-organic frameworks, [M2(LH)3(DMSO)4]•xDMSO (TbLH and EuLH for M = Tb , 3+ Eu respectively; LH = (E)-4,4’-(ethene-1,2-diyl)dibenzoate) and [M2(LS)3(DMF)2(H2O)2] (EuLS 3+ 3+ and TbLS for M = Eu , Tb ) have been synthesized and characterized. All four MOFs adopt the anticipated connectivity, but the fold of interpenetration in each MOF crystal varies as a result of different solvent inclusion and ligand functionality. Triangular channels are formed, with the

central CC double bond of the ligand exposed inside. EuLH and TbLH contain large amount of

solvent molecules and tend to lose solvent in air. EuLS and TbLS are stable crystals without lattice solvent molecules in the crystals, as the thioether side chains completely block the channels; the coordinating solvent molecules can be reversibly removed and replenished with the

skeletons intact. All four MOFs show mainly the ligand-based luminescence, while EuLS also displays metal-based luminescence.

2.2 Introduction

Our first functionalized MOFs system designed is based on two known functionalized examples: [1] [Tb2(ADB)3(DMSO)4]•16DMSO (where ADB = 4,4’-azodibenzoate) by Yaghi and co-workers

and [Eu2L3(DMSO)2(CH3OH)2]•2DMSO (where L = 4,4’-ethyne-1,2-diyldibenzoate) by our group.[2] By replacing the N=N and C≡C from the ligands in the known MOFs with a C=C linkage, different functional groups can be implemented on the olefin spacer of the linear ligand. Although the stilbene dicarboxylate ligands have been reported in MOF synthesis with transition metals[3], the lanthanide counterparts have not been explored. We envision that a similar skeleton could be anticipated when a functionalized ligand is used, i.e., each bimetallic node is connected with 6 other nodes to form 3D porous frameworks with triangular channels along the body diagonal and functional groups dangling inside the channels (Scheme 2.1). In this chapter, ligand

29 synthesis, MOF preparation and characterization based on the functionalized stilbene dicarboxylic acid ligands will be discussed.

Scheme 2.1. MOF design in this chapter 2.3 Experimental section

2.3.1 Materials and methods

Elemental analyses were performed in our Chemistry Department on a PE 2400 C/H/N/S analyzer. Thermogravimetric analyses (TGA) were performed on a TA Instruments SDT Q600 instrument under a dinitrogen atmosphere with a heating rate of 10 °C per minute. NMR spectra were recorded on a Varian 400, or a Bruker Avance 400 spectrometer. Both 1H and 13C NMR spectra were referenced and reported relative to the solvent’s residual signals. Photoluminescence spectra were measured using a SPEX Fluorolog-3 spectrofluorometer (Jobin Yvon/SPEX, Edison, New Jersey). The powder XRD experiments were performed on an automated Siemens/Bruker D5000 diffractometer equipped with a high power line focus Cu-Ka source operating at 50 kV/35 mA. Unless otherwise stated, all manipulations were performed in air and all reagents were purchased from commercial sources and used without further purification. McMurry coupling reaction was performed under an Ar atmosphere, but the work- up was carried out under ambient conditions. The THF solvent for McMurry coupling was purified by John Morris Scientific IT PureSolv PS-MD-6 system; Zn-Cu couple[4] and methyl 4- acetylbenzoate[5] for the coupling reaction was prepared according to the literature procedures.

30

2.3.2 Synthesis of dimethyl 4,4'-(but-2-ene-2,3-diyl)dibenzoate (trans and cis LM ester)

With vigorous stirring TiCl4 (2.7 mL, 0.025 mol) was added slowly to a cold (0 ºC) suspension of Zn-Cu couple (3.32 g) in THF (50 mL). The mixture was then refluxed for 1 h, cooled to ambient temperature, and followed by the addition of a solution of methyl 4-acetylbenzoate (2.0 g, 0.011 mol) in THF (10 mL). The reaction mixture was further refluxed for 12 h. After cooling to ambient temperature, the reaction mixture was quenched with 10 mL of saturated K2CO3 (aq) and extracted with Et2O. The organic layer was washed with brine, and dried over MgSO4. After

MgSO4 was filtered off, the solvents were removed to give the crude product, which was purified by silica gel column chromatography (eluted with CH2Cl2) to give a trans/cis mixture of LM ester (1.0 g, 55% yield) as a light yellow oil. The two isomers can be separated through chromatography and isolated as white or light yellow solids, although the separation is not 1 necessary for the next step. cis LM ester: H NMR (CDCl3, 400 MHz, 25°C)  7.74 (d, J = 8.4 Hz, 13 4H), 7.00 (d, J = 8.4 Hz, 4H), 3.85 (s, 6H), 2.18(s, 6H); C NMR (CDCl3, 400 MHz, 25°C) 

167.14, 149.33, 133.83, 129.31, 127.82, 52.16, 21.46. Anal.calcd for C20H20O4: C, 74.06; H, 6.21. 1 Found: C, 74.13; H, 6.24. trans LM ester: H NMR (CDCl3, 400 MHz, 25°C)  8.06 (d, J = 8.0 13 Hz, 4H), 7.34 (d, J = 8.0 Hz, 4H), 3.93 (s, 6H), 1.88(s, 6H); C NMR (CDCl3, 100 MHz, 25°C)  167.21, 149.18, 133.47, 129.88, 128.48, 52.30, 22.41.

2.3.3 Synthesis of dimethyl 4,4’-(1,4-dibromobut-2-ene-2,3- diyl)dibenzoate (trans and cis LB ester)

A mixture of LM ester (648 mg, 2 mmol) as obtained above, N-bromosuccinimide (712 mg, 4 mmol) and benzoyl peroxide (97 mg, 0.4 mmol) were refluxed in cyclohexane (30 mL) for 5 h. The reaction mixture was cooled to ambient temperature and filtered. After the removal of solvent from the filtrate, the residual crude product was purified using silica gel column chromatography (eluted with 1:7 mixture of EtOAc:hexanes), affording colorless crystals of 1 trans LB ester (460 mg, 48% yield) and cis LB ester (141 mg, 15% yield). trans LB ester: H

NMR (CDCl3, 400 MHz, 25°C)  8.15 (d, J = 8.4 Hz, 4H), 7.54 (d, J= 8.4 Hz, 4H), 4.00 (s, 4H), 13 3.96(s, 6H); C NMR (CDCl3, 100 MHz, 25°C)  166.81, 142.62, 139.13, 130.44, 130.19,

128.67, 52.51, 34.57. Anal.calcd for C20H18O4Br2: C, 49.82; H, 3.76. Found: C, 49.93; H, 3.95. 1 cis LB ester: H NMR (CDCl3, 400 MHz, 25°C)  7.92 (d, J = 8.4 Hz, 4H), 7.14 (d, J= 8.4 Hz,

31

13 4H), 4.49 (s, 4H), 3.87(s, 6H); C NMR (CDCl3, 100 MHz, 25°C)  166.51, 143.77, 140.06,

129.87, 129.47, 129.23, 52.16, 31.14. Anal.calcd. for C20H18O4Br2 • 0.75CH2Cl2: C, 45.66; H, 3.60. Found: C, 45.68; H, 3.52.

2.3.4 Synthesis of (E)-dimethyl 4,4’-(1,4-bis(methylthio)but-2-ene-2,3- diyl)dibenzoate (LS ester)

Sodium methanethiolate (108 mg, 1.54 mmol) and trans LB ester (300 mg, 0.62 mmol) were refluxed in methanol (30 mL) under N2 for 3 h. After the solvent was removed in vacuo, distilled water was added to dissolve the residual solid. The aqueous solution was acidified with excess 1

M HCl solution and extracted with EtOAc. The organic phase was then dried over MgSO4. After filtration the solution was concentrated to dryness, affording LS ester as a light yellow powder 1 (231 mg, 95% yield). H NMR (CDCl3, 400 MHz, 25°C)  8.10 (d, J = 8.4 Hz, 4H), 7.44 (d, J = 13 8.4 Hz, 4H) , 3.94 (s, 6H), 3.29 (s, 4H), 1.83(s, 6H); C NMR (CDCl3, 100 MHz, 25°C) 

166.85, 144.83, 136.99, 129.66, 129.38, 129.12, 52.21, 38.84, 16.02. Anal.calcd for C22H24O4S2: C, 63.43; H, 5.81. Found: C, 62.89; H, 5.83.

2.3.5 Synthesis of (E)-4,4’-(1,4-bis(methylthio)but-2-ene-2,3- diyl)dibenzoic acid (H2LS)

Potassium hydroxide (1.1 g, 19 mmol) and LS ester (500 mg, 1.2 mmol) were refluxed in methanol (60 mL) for 2 h. After removal of solvents in vacuo, distilled water was added to dissolve the residual solid. The aqueous solution was acidified with excess 1 M HCl to give a white suspension. The suspension was directly extracted with EtOAc. The organic phase was dried over MgSO4 and concentrated to give 466 mg white powder of H2LS in quantitative yield. 1 H NMR (CD3OD, 400 MHz, 25°C)  8.11 (d, J = 8.4 Hz, 4H), 7.58 (d, J = 8.4 Hz, 4H), 3.40 (s, 13 4H), 1.84(s, 6H); C NMR (DMSO-d6, 100 MHz, 25°C)  166.94, 144.17, 136.11, 129.59,

129.09, 128.98, 37.56, 14.6. Anal.calcd for C20H20O4S2: C, 61.83; H, 5.19. Found: C, 61.70; H, 5.34.

2.3.6 Synthesis of (E)-dimethyl 4,4'-(1,4-bis(phenylthio)but-2-ene-2,3- diyl)dibenzoate (LSPh ester)

Sodium hydride (60% in mineral oil, 175 mg) was suspended in dried THF (20 mL) and thiophenol (490 μL) was added slowly at 0 °C. The mixture was then stirred at 0 °C for 30 min

32

before filtered and transferred to trans LB ester solution (1.0 g LB ester, 2.1 mmol, in 20 mL dry THF). The resulting mixture was refluxed for 4 h. After cooled to ambient temperature, water was added and the mixture was extracted with ethyl acetate. The organic phase was washed with brine, dried over MgSO4 and then concentrated to give the desired product as a light yellow solid 1 (1.06 g, 95% yield). H NMR (CDCl3, 400 MHz, 25°C)  7.93 (d, J = 8.0 Hz, 4H), d 7.18 (d, J = 13 8.0 Hz, 4H), 7.09 (m, 6H), 6.98 (m, 4H), 3.87 (s, 6H), 3.59 (s, 4H). C NMR (CDCl3, 100 MHz, 25°C)  166.44, 143.80, 136.53, 135.02, 130.58, 129.15, 128.93, 128.65, 128.44, 126.39, 51.80,

39.85. Anal.calcd for C32H28O4S2: C, 71.08; H, 5.22. Found: C, 70.64; H, 5.66.

2.3.7 Synthesis of (E)-4,4'-(1,4-bis(phenylthio)but-2-ene-2,3- diyl)dibenzoic acid (LSPh acid)

LSPh ester (1.85 g, mmol), NaOH (4.8 g, mmol) were suspended in a mixture of THF (48 mL) and methanol (72 mL), and stirred at R.T. for 5 h. After the reaction, the organic solvent was removed under vacuum and the remaining solid was dissolved in water. 1M HCl was added till pH=1 and the precipitate was collected by filtration, washed with methanol to give LSPh acid as a 1 light yellow solid (1.46 g, 83 % yield). H NMR (DMSO-d6, 400 MHz, 25°C)  12.98 (s, 2H), d 7.92 (d, J = 8.0 Hz, 4H), 7.34 (d, J = 8.0 Hz, 4H), 7.21 (m, 6H), 7.06 (m, 4H), 3.76 (s, 4H). 13C

NMR (DMSO-d6, 100 MHz, 25°C)  158.00, 134.59, 127.30, 126.31, 120.91, 120.45, 120.11, 119.94, 117.36. Anal.calcd for C30H24O4S2: C, 70.29; H, 4.72. Found: C, 69.91; H, 5.03.

2.3.8 Synthesis of [Tb2(LH)3(DMSO)4] (TbLH)

A small vial containing a 2 mL DMSO solution of TbCl3•6H2O (8 mM) and H2LH (12 mM) was inserted into a large vial containing DMSO (8 mL) and triethylamine (0.024 mL, 0.17 mmol). The large vial was then sealed. The double-vial system was allowed to stand for 14 d at room temperature to afford TbLH as colorless X-ray quality crystals. The crystals were collected by filtration and dried under vacuum for 1 h (7.6 mg, 46% yield). The synthesis of TbLH has poor reproducibility. Anal. Calcd. for C56H54O16S4Tb2 • 8C2H6SO: C, 42.10; H, 5.00. Found: C, 42.44; H, 4.90.

2.3.9 Synthesis of [Eu2(LH)3(DMSO)4] (EuLH)

A small vial containing a 2 mL DMSO solution of EuCl3•6H2O (8 mM) and H2LH (12 mM) was inserted into a large vial containing DMSO (8 mL) and triethylamine (0.024 mL, 0.17 mmol).

33

The large vial was then sealed. The double-vial system was allowed to stand for 14 d at room temperature to afford EuLH as colorless X-ray quality crystals. The crystals were collected by filtration and dried under vacuum for 1 h (8.2 mg, 41% yield). Anal. calcd. for C56H54O16S4Eu2 •

14C2H6SO: C, 40.21; H, 5.54. Found: C, 39.92; H, 5.17.

2.3.10 Synthesis of [Eu2(LS)3(DMF)2(H2O)2] (EuLS)

Eu(NO3)3•6H2O (17.8 mg, 0.040 mmol), H2LS (23.3 mg, 0.060 mmol), N,N’-dimethylformamide (4 mL) and distilled water (6 mL) were mixed and sealed in 23 mL PTFE lined autoclave, then heated at 120 °C for 3 d. After cooling down, yellow needle-shaped crystals of EuLS were collected by filtration, washed with ethanol, and dried in air (26.5 mg, 80% yield). Anal. Calcd. for C66H72Eu2N2O16S6: C, 48.17; H, 4.41; N, 1.70. Found: C, 47.93; H, 4.17; N, 2.12.

2.3.11 Synthesis of [Tb2(LS)3(DMF)2(H2O)2] (TbLS)

The synthesis of TbLS is similar to that of EuLS, but with TbCl3•6H2O (14.9 mg, 0.040 mmol).

Yellow needle-shaped crystals of TbLS were collected by filtration, washed with ethanol, and dried in air (22.6 mg, 68% yield). Anal.calcd for C66H72N2O16S6Tb2: C, 47.77; H, 4.37. Found: C, 47.87; H, 4.47.

2.3.12 X-ray Crystallographic Analysis.

X-ray quality single crystals of EuLH, TbLH, EuLS, TbLS were obtained as described in the sections above; those of trans/cis LM ester, trans/cis LB ester, LS ester, LSPh ester and H2LSPh were obtained from slow evaporation of the solvent from the corresponding CH2Cl2 solutions. All crystals were mounted on the tip of a MiTeGen MicroMount. All single-crystal X-ray diffraction data were collected on a Bruker Kappa Apex II diffractometer with graphite-monochromated Mo K radiation (λ = 0.71073 Å) operating at 50 kV and 30 mA, at 120 K or 150 K controlled by an Oxford Cryostream 700 series low temperature system. The data integration and absorption correction were performed with the Bruker Apex 2 software package.[6] All structures were solved by the direct methods except for TbLS which was solved by Patterson method. All structure solution and refinements were performed using SHELXTL V6.14.[7] A disordered phenyl ring of the ligand in EuLH and the thioether side chains in EuLS and TbLS have been modeled successfully. The residual diffuse electron density of unidentified solvent molecules in [8] the lattices of EuLH and TbLH was removed with the SQUEEZE function of PLATON program

34 and their contributions were not included in the formula, although elemental analysis results suggest that the removed electron density may originate at least partially from several DMSO molecules. All non-hydrogen atoms except for the atoms involved in the disordered portions were refined anisotropically. The positions of the hydrogen atoms were calculated using the riding model. Selected crystallographic data of the MOFs and synthesis intermediates are summarized in Table 2.1 - Table 2.3, while selected bond lengths and angles of MOFs are listed in Table 2.4 – Table 2.7.

2.4 Results and discussion

2.4.1 Ligand synthesis

O O O O MeOH,CO Zn/Cu,TiCl4 O I O Pd/C,Et N O THF 3 O

LM ester

O Br NBS,(PhCOO) O S 2 O NaSMe O O C6H12 MeOH O Br O S O

LB ester LS ester O S (1)KOH,MeOH O HO (2)HCl S OH

H2LS

O Br O SPh PhSH,NaH O O O THF O Br O PhS O

LSPh ester

O SPh (1)NaOH,THF/MeOH O (2)HCl HO PhS OH

H2LSPh

Scheme 2.2. Synthesis of H2LS and H2LSPh.

35

As shown in Scheme 2.2, ligand H2LS can be synthesized in 5 steps in a 19% overall yield. Methyl 4-acetylbenzoate was synthesized from carboxylation of 4-iodo-acetophenone under CO in 80% yield.[9] McMurry coupling[10] reaction of methyl 4-acetylbenzoate produces a mixture of 1 trans and cis isomers of LM ester, in about 1:4 ratio according to the H NMR spectrum of the crude product. Although cis LM ester is the major product, the subsequent bromination of LM ester using NBS causes the isomerization of the double bond to give trans LB ester as the major product. Therefore, the mixture of trans/cis LM ester resulting from the McMurry coupling step can be used directly for the bromination reaction and the trans/cis LB ester can be separated after the bromination step using column chromatography. Recrystallization has been attempted but none of the conditions were successful in isolating the trans/cis isomer. The treatment of the trans LB ester with NaSCH3 affords trans LS ester, which can be subsequently hydrolyzed to produce H2LS. LSPh ester can be prepared by deprotonating thiophenol with sodium hydride and subsequently reacting with trans LB ester. The corresponding acid H2LSPh was obtained by careful hydrolysis at room temperature. The geometry at the C=C linkage of trans/cis LM ester, trans/cis LB ester, LS ester, LSPh ester, H2LSPh has been confirmed by X-ray crystallography (Figure 2.1). According to the crystal structures, the two phenyl rings on the C=C linkage of trans LM ester, trans LB ester, LS ester, LSPh ester and H2LSPh are no longer co-planar due to the steric effect imposed by the substituents.

36

Figure 2.1. Crystal structures of (a) trans LM ester; (b) cis LM ester; (c) trans LB ester; (d) cis LB ester; (e) LS ester; (f) LSPh ester; (g) H2LSPh. 2.4.2 MOF preparation and structure description

The ligand (E)-4,4’-(ethene-1,2-diyl)dibenzoic acid, H2LH, is commercially available. The MOFs

TbLH and EuLH can be prepared in crystalline form in moderate yields by slow diffusion of triethylamine vapor into the mixed solution of H2LH and MCl3•6H2O (M = Tb or Eu) at ambient temperature. Some unidentified amorphous precipitates were also observed during the preparation. These amorphous precipitates have lower density compared to the crystalline EuLH and TbLH and therefore, can be flushed away with DMSO. The MOF crystals collapse into

37 powders quickly upon contact with acetone and chloroform; they also lose solvents and become opaque rapidly upon standing in air. They are insoluble in common solvents, suggesting the polymeric nature of their structures.

2- Because of the structural similarity between LH and 4,4’-azodibenzoate (ADB) used by [11] Yaghi, the corresponding Tb(III) MOFs, TbLH and [Tb2(ADB)3(DMSO)4]•xDMSO, show almost identical structures, e.g., same space group, similar unit cell parameters and topologies.

TbLH crystallized in the monoclinic space group C2/c. As shown in Figure 2.2a, each Tb(III) center adopts distorted tricapped trigonal prismatic coordination geometry with seven oxygen 2- donor atoms from the carboxylic groups of LH ligands and two oxygen donor atoms from terminal DMSO ligands occupying the nine coordination sites. The trigonal prism is defined by O2, O3, O4, O5, O6, and O8; the three rectangular faces are capped by O1, O4A, and O7, respectively. The three capping atoms and Tb1 are roughly coplanar, as indicated by that the sum of the three relevant bond angles being 355.4(1)º, close to 360°. As shown in Figure 2.2b, two adjacent Tb(III) centers are bridged by four carboxylate groups to form a bimetallic node with the Tb-Tb distance of 4.0866(4) Å. There is a crystallographically imposed center of inversion at the centroid of each di-Tb node. Because of unsymmetrical bridging, the Tb1-O4A bond length is 2.851(3) Å, while all the remaining Tb-O bond lengths are within the range of 2.293(3)–2.480(4) Å. As shown in Figure 2.2c, each di-Tb node is connected to six neighboring nodes via stilbene linkers to form a 3D infinite framework. Similar to Yaghi’s

[Tb2(ADB)3(DMSO)4]•xDMSO, a 2-fold interpenetration has been observed in the crystal lattice of TbLH (Figure 2.2d), as a result of long linker group employed in the organic ligand.

38

Figure 2.2. ORTEP drawings of TbLH: (a) the coordination sphere of a Tb(III) center showing the tricapped trigonal prismatic coordination geometry; (b) the structure of a dinuclear node [Tb2(LH)3(DMSO)4] with the stilbene linkers reduced to the ipso carbons. The thermal ellipsoids are plotted at 50% probabilities. Extended structures of TbLH: (c) a drawing of one set of 3D framework, projection along the b axis; (d) a space filling drawing showing the two interpenetrating frameworks, one in red and the other in blue. Hydrogen atoms and lattice solvent molecules are omitted for clarity.

39

Figure 2.3. ORTEP drawings of EuLH: (a) the coordination sphere of a Eu(III) center showing the tricapped trigonal prismatic coordination geometry; (b) the structure of a dinuclear node [Eu2(LH)3(DMSO)4]. The thermal ellipsoids are plotted at 50% probabilities. Extended structures of EuLH: (c) a drawing of one set of 3D framework, projection along the c axis; (d) a space filling drawing showing the two interpenetrating frameworks, one in red and the other in blue. Hydrogen atoms and lattice solvent molecules are omitted for clarity.

Although switching from Tb(III) to Eu(III) usually produces isostructural analogues, the Eu(III) 2- MOF of LH , EuLH, crystallized in a different space group (P212121) compared to TbLH. Similar to the Tb(III) centers in TbLH, each Eu(III) center adopts distorted tricapped trigonal prismatic 2- coordination geometry with seven oxygen donor atoms from the carboxylic groups of LH ligands and two oxygen donor atoms from terminal DMSO ligands occupying the nine coordination sites (Figure 2.3a). The trigonal prism is defined by O1, O4, O6, O8, O9, and O13; the three rectangular faces are capped by O7, O10, and O14, respectively. The three capping atoms and Eu1 are roughly coplanar, as indicated by that the sum of the three relevant bond

40 angles is 355.6(2)º, close to 360°. As shown in Figure 2.3b, two adjacent Eu(III) centers are bridged by four carboxylate groups to form a bimetallic node with the Eu-Eu distance of [13] 4.0743(6) Å, comparable to the literature values. In contrast to the di-Tb nodes in TbLH, there is no crystallographically imposed center of inversion at the centroid of the di-Eu nodes in EuLH, as a result of the chiral space group in which EuLH crystallized. Because of unsymmetrical bridging by O7 and O4, Eu1-O7 and Eu2-O4 bond lengths are 2.802(6) and 2.813(6) Å, respectively, while all the remaining Eu-O bond lengths are within the range of 2.346(6)–2.511(3) Å. As shown in Figure 2.3c, each bimetallic node is connected to six neighboring nodes via stilbene linkers to form a 3D infinite framework. Because of the large spacing generated by the long linker, 2-fold interpenetration has been observed in the crystal lattice of 7 (Figure 2.3d).

Similar to the carboxylates in [Eu2L3(DMSO)2(CH3OH)2] •2DMSO (L = 4,4’-ethyne-1,2- diyldibenzoate), three coordination modes have been observed for the carboxylates in the lattices of EuLH and TbLH (Scheme 2.3): chelation mode, bridging mode, and chelate-bridging mode. In 2- the lattice of [Eu2L3(DMSO)2(CH3OH)2]•2DMSO, the two carboxylates within each L ligand can adopt different coordination modes. In contrast, the two carboxylate ligands within each 2- LH ligand always adopt the same coordination mode in the lattices of EuLH and TbLH.

Scheme 2.3. Different coordination modes of carboxylate ligands.

41

Figure 2.4. ORTEP drawings: (a) a di-Eu node in EuLS; (b) a di-Tb node in TbLS. Thermal ellipsoids are plotted at 2- 50% probabilities. All LS ligands are reduced to the carboxylic groups and an ipso carbon from the phenyl ring and all hydrogen atoms are omitted for clarity. Extended structures of EuLS (or TbLS): (c) one set of framework; (d) spacefilling drawing of three interpenetrating frameworks coated with different colors.

EuLS and TbLS can be prepared as light yellow needle-shaped crystals in good yields from

M(NO3)3•6H2O (M = Eu or Tb) and H2LS solution under solvothermal conditions in DMF/H2O mixture. These two isostructural 3D coordination polymers both crystallized in the triclinic space group P-1. EuLS will be used as an example for structure descriptions below. Each Eu(III) 2- center adopts a highly irregular coordination geometry with seven oxygen donor atoms from LS ligands, one oxygen donor atom from a DMF molecule, and one oxygen donor atom from a H2O molecule occupying the nine coordination sites. Each Eu(III) center is linked with an adjacent Eu(III) center by two bridging oxygen atoms from two carboxylates, forming a di-Eu node as shown in Figure 2.4a. A similar di-Tb node found in TbLS is shown in Figure 2.4b. There is a crystallographically imposed center of inversion at the centroid of each bimetallic node. Unlike the carboxylates in EuLH and TbLH which show three types of coordination modes, the carboxylates in EuLS and TbLS only show two types of coordination modes: the chelating mode

42

2- 2- (for 2/3 of the LS ligands) and the chelate-bridging mode (for 1/3 of the LS ligands). The bridging mode is not observed. Most Eu-O and Tb-O bond lengths are in the ranges of 2.390(5)– 2.480(5) Å and 2.371(5)–2.463(5) Å, respectively, except for those of Eu1-O3 and Tb1-O3 which are 2.648(5) and 2.653(5) Å, respectively, due to the unsymmetrical bridging of O3 in each case. In the crystal lattice, each bimetallic node is connected with six adjacent nodes via six 2- organic linkers to form an infinite 3D framework (Figure 2.4c). Compared to the LH ligand, 2- the LS ligand has bulkier thioether substituents on the central CC double bond. Because EuLH 2- and TbLH assembled with LH ligands display a 2-fold interpenetration, we expect no more than

2-fold of interpenetration in EuLS and TbLS. To our surprise, 3-fold interpenetrating frameworks without any non-coordinating guest molecules were observed in both (Figure 2.4d). Calculation via PLATON program23 shows no solvent accessible space, indicating that the addition framework has taken up all the space that is occupied by solvent molecules in EuLH and TbLH.

Despite various attempts by varying temperature, reaction time, solvent system, metal/ligand ratio and concentration, MOF synthesis with H2LSPh has been unsuccessful in obtaining crystals suitable for single-crystal X-ray diffraction.

2.4.3 Solvent inclusion and porosity

In all four MOFs reported here, the organic spacers and the bimetallic nodes assemble into approximately orthogonal skeletons. Each set of framework can also be viewed as an infinite assembly of ‘cubes’ defined by twelve organic spacers and eight bimetallic nodes. If the frameworks are viewed along the body diagonal direction of the ‘cubes’, triangular shaped channels can be seen (Figure 2.5). TbLH and EuLH have similar triangular shaped channels in which the lattice solvent molecules reside. Gas adsorption/desorption experiments on neither of the two show any porosity, probably because the removal of solvents causes the structure to collapse. As shown in Figure 2.6b, EuLS also has triangular channels in which the thioether side chains are dangling. However, the thioether side chains block the channels completely. In addition, the 3-fold interpenetration blocks other possible channels that are perpendicular to the triangular ones, leading to the non-porous structures.

43

Figure 2.5. Triangular channel along the body diagonal direction of the ‘cube’ (left), the helical arrangement of the ligands defining a triangular channel in a single-fold framework with possible channel perpendicular to the triangular channel (middle) and in a 3-fold interpenetrating framework with no possible channel perpendicular to the triangular channel (right).

Although there is no solvent accessible space in EuLS, we envision that the removal of the coordinating solvents may create solvent accessible void space. The solvent removal can be achieved by heating the as-synthesized EuLS at 140 °C for 4 h. The resulting de-solvated sample displays little weight loss at 130-165 °C (Figure 2.7b) compared to as-synthesized (Figure 2.7c), suggesting the removal of a significant amount of solvents. Accordingly, the IR spectrum -1 of the de-solvated EuLS shows a significant decrease in intensities of the signals at 1661 cm [12] -1 (coordinating DMF molecules) and 3300 cm (broad peak from H2O), compared to the as- synthesized EuLS (Figure 2.8). When the de-solvated EuLS is soaked in the mixture of

DMF/H2O (v:v =1:1) overnight and then washed with ethanol and air-dried, the resulting re- solvated EuLS shows a weight loss at 120-165 °C in TGA (Figure 2.7c), similar to the as- synthesized sample. Correspondingly, the IR spectrum of the re-solvated sample shows the recovery of the intensities of the signals at 1661 and 3300 cm-1. Powder XRD experiments show that EuLS remains crystalline after the solvent removal and re-entry processes (Figure 2.9). Gas adsorption experiments of the de-solvated sample still yields no porosity, possible due to the pore-blocking caused by prolonged degas process.

44

Figure 2.6.Diagrams showing the triangular channels: (a) TbLH, spacefilling model, projection down the a+b direction; lattice solvent molecules are omitted for clarity. (b) EuLS, spacefilling model for the skeleton and stick model for the thioether functional groups, projection down the a direction.

(a) (b) 100 TbL 100 H EuL S EuL 90 H TbL S 80 80

70

60 60

50

40

Weight (%) 40

Weight(%) 30

20 20

10

0 0 0 200 400 600 800 0 200 400 600 800 o Temperature ( C) o Temperature ( C)

(c) 100 (d) 100

90 90

80 80

70 70

60 Weight (%) As-synthesized EuL Weight (%) 60 S 50 De-solvated EuL S 50 Re-solvated EuL 40 S

30 40 0 200 400 600 800 0 200 400 600 Temperature (oC) Temperature (oC)

Figure 2.7. TGA of (a) EuLH and TbLH; (b) EuLS and TbLS; (c) As-synthesized, de-solvated and re-solvated EuLS; (d) NaLH salt.

45

c

b

a

3500 3000 1600 1400 1200 1000 800 600 400 Wavenumber (cm-1)

Figure 2.8. IR spectra of EuLS sample: (a) as-synthesized; (b) de-solvated; (c) re-solvated.

d

c

b

a

10 20 30

2 (deg)

Figure 2.9. Powder X-ray diffraction of EuLS: (a) pattern predicted from crystal structure; (b) as-synthesized; (c) de- solvated; (d) re-solvated.

46

1.0 Excitation of EuL (a) Excitation of TbL 1.0 H H b. Emission of EuL Emission of TbL H H

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

NormalizedIntensity (A.U.) NormalizedIntensity (A.U.)

0.0 0.0 250 300 350 400 450 500 550 600 650 700 250 300 350 400 450 500 550 600 650 700 Wavelength (nm) Wavelength (nm)

d. Excitation of TbL (C)1.4 S Excitation of EuL 1.2 S Emission(i) of TbL S Emission(i) of EuL 1.2 S Emission(ii) of TbL S Emission(ii) of EuL 1.0 S 1.0 0.8 0.8 0.6 0.6

0.4 0.4

0.2

0.2

NormalizedIntensity (A.U.) NormalizedIntensity (A.U.)

0.0 0.0

300 350 400 450 500 550 600 650 700 250 300 350 400 450 500 550 600 650 700 Wavelength (nm) Wavelength (nm)

(e) 1.0 Excitation H L (f) 2 H 1.0 Emission H L Excitation of H L 2 H 2 S Emission of H L 2 S 0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

NormalizedIntensity (A.U.) NormalizedIntensity (A.U.)

0.0 0.0 250 300 350 400 450 500 550 600 650 700 250 300 350 400 450 500 550 600 650 700 Wavelength (nm) Wavelength (nm)

Figure 2.10. Photoluminescence spectra of TbLH, EuLH, EuLS,TbLS, H2LH and H2LS (a-f, respectively). (a) and (b): Excitation and emission were measured at 450 and 404 nm, respectively. (c): Excitation, emission (i) and emission (ii) were measured at 615, 393, and 350 nm, respectively. (d) Excitation, emission (i) and emission (ii) were measured at 515, 410, and 487 nm, respectively; (e) Excitation and emission were measured at 464, 382 nm respectively, with H2LH solid; (f) Excitation and emission were measured at 479, 396 nm respectively, with H2LS solid.

47

2.4.4 Luminescence properties

The photoluminescence excitation and emission spectra of the four MOFs are shown in Figure

2.10. Both TbLH and EuLH show broad ligand-based emission between 420 and 490 nm with the emission maximum at 447 nm, which are similar to the luminescence spectrum of the free ligand

H2LH.

The emission of EuLS consists of both the ligand-based and metal-based luminescence. The sharp lines at 393, 464 and 535 nm in the excitation spectrum can be assigned to the metal- 7 5 7 5 7 5 centered transitions F0,1 → L6, F0,1 → D2, F0,1 → D1, respectively. The weak broad band centered at ~370 nm in the excitation spectrum can be attributed to the ligand-based excitation. The sharp emission peaks at 579, 591, 615, and 696 nm in the emission spectrum are 5 7 5 7 5 7 5 7 characteristic Eu-centered transitions D0 → F0, D0 → F1, D0 → F2, and D0 → F4 respectively. When 393 nm is used as the excitation wavelength, at which an intense metal- centered excitation and a weak ligand-centered excitation co-exist, the ligand-based emission can be observed as a weak broad band centered at around 490 nm, along with intense Eu-based emissions. To examine whether LMET contributes to the Eu-based emissions, a second emission spectrum was recorded with 350 nm excitation wavelength, at which only ligand-based absorption occurs. The resulting emission spectrum shows not only the ligand-based emission, but also the characteristic Eu emissions with relatively low intensity compared to the ligand- based emission, indicating the occurrence of LMET at low efficiency. TbLS shows broad ligand- based excitation and emission at 350 ~ 450 nm, 475 ~ 600 nm respectively, while the direct f-f 3+ 7 5 transition of Tb happens under 487 nm excitation ( F6 → D4) and the emission peaks at 542

5 7 5 7 5 7 nm, 584 nm, 620 nm on emission (ii) can be assigned to D4 → F5, D4 → F4, D4 → F3, respectively. Preliminary luminescent response tests were carried out by measuring the luminescence spectrum of desolvated EuLS soaked in HgCl2 solution in THF and CdCl2 aqueous solution. No significant luminescence signal change has been observed, possibly because the pore size of de-solvated EuLS is smaller than the solvated heavy metal species.

2.5 Summary

In this chapter, we have designed and synthesized a few ligands with different side chains, and 2- 2- successfully assembled both non-functionalized LH and thioether functionalized LS ligand into

48

luminescent 3D MOFs with lanthanide metal ions. EuLH, TbLH and TbLS display ligand-based luminescence, while EuLS showed both ligand-based and Eu-centered luminescence. The structures of EuLH and TbLH closely resemble Yaghi’s [Tb2(ADB)3(DMSO)4]•16DMSO (where

ADB = 4,4’-azodibenzoate). The skeletons of EuLS and TbLS define similar triangular channel as seen in EuLH and TbLH. However, the thioether functional groups block these channels completely and the 3-fold interpenetration blocks other possible channels that are perpendicular to the triangular ones, making as-synthesized EuLS and TbLS non-porous. Small pores can be generated upon removal of the coordinating solvents in EuLS and TbLS; such solvent removal can be reversed by soaking the desolvated crystals in a DMF and H2O mixture. The pores created by solvent removal are likely too small for solvated heavy metal species to get in. By changing the size or number of the functional groups on the ligands, we may potentially create some extra space in the resulting MOFs. Finally, the installation of thioether side chains enables the LMET and thus enhances Eu-based luminescence. This effect could be attributed at least partly to the steric bulk provided by the side chains, which prevents the ligand from being coplanar and in turn, alters the ligand energy levels.

49

Table 2.1. Crystallographic data of trans/cis LM ester and trans/cis LB ester.

trans LM ester cis LM ester trans LB ester cis LB ester · CH2Cl2

formula C20H20O4 C20H20O4 C20H18Br2O4 C21H20Br2O4Cl2 F.W. 324.36 324.36 482.16 567.09 T (K) 150(2) 150(2) 150(2) 150(2)

space group C2/c C2/c P-1 P21/n a (Å) 12.4169(6) 12.0232(12) 5.6686(2) 10.7778(5) b (Å) 9.8259(6) 8.3543(12) 7.1886(3) 11.6120(5) c (Å) 14.0212(6) 17.166(2) 12.3615(5) 18.5198(9)  (deg) 90 90 75.833(2) 90  (deg) 91.585(5) 98.840(5) 80.6490(10) 97.184(3)  (deg) 90 90 79.1370(10) 90 V (Å3) 1710.03(17) 1703.8(4) 476.06(3) 2299.59(18) Z 4 4 1 4 −3 Dc (g·cm ) 1.260 1.264 1.682 1.638  (mm−1) 0.087 0.087 4.279 3.782 no. reflns collcd 20362 9076 8066 27090 no. indept reflns 2945 3331 2179 7017 GOF on F2 0.995 1.012 1.055 1.006

R [I > 2σ(I)] R1 = 0.0531 R1 = 0.0535 R1 = 0.0204 R1 = 0.0473

wR2 = 0.1088 wR2 = 0.1353 wR2 = 0.0515 wR2 = 0.0853

R (all data) R1 = 0.1353 R1 = 0.1105 R1 = 0.0223 R1 = 0.1208

wR2 = 0.1370 wR2 = 0.1646 wR2 = 0.0523 wR2 = 0.1027

50

Table 2.2. Crystallographic data of LS ester, LSPh ester, and H2LSPh.

LS ester LSPh ester H2LSPh

formula C22H24O4S2 C32H28O4S2 C34H36O6S4 F.W. 416.53 540.66 668.87 T (K) 150(2) 150(2) 150(2) space group P-1 P21/c P-1 a (Å) 7.7087(5) 14.9728(9) 5.9151(7) b (Å) 8.0642(3) 8.1687(5) 11.4139(15) c (Å) 9.4794(6) 12.0536(7) 12.6124(17)  (deg) 85.956(2) 90 97.264(5)  (deg) 88.701(2) 107.634(3) 95.174(6)  (deg) 63.240(2) 90 97.555(5) V (Å3) 524.83(5) 1405.49(15) 832.40(19) Z 1 2 1 −3 Dc (g·cm ) 1.318 1.278 1.334  (mm−1) 0.279 0.225 0.329 no. reflns collcd 8691 12071 12404 no. indept reflns 2369 3243 3612 GOF on F2 1.095 1.112 0.996

R [I > 2σ(I)] R1 = 0.0393 R1 = 0.0703 R1 = 0.0605

wR2 = 0.1061 wR2 = 0.1788 wR2 = 0.1339

R (all data) R1 = 0.0433 R1 = 0.1140 R1 = 0.1312

wR2 = 0.1092 wR2 = 0.1943 wR2 = 0.1633

51

Table 2.3. Crystallographic data EuLH, TbLH, EuLS, and TbLS.

TbLH EuLH EuLS TbLS

formula C24H15O6Tb · C48H30O12Eu2 · C33H36NO8S3Eu C33H36NO8S3Tb 7 C2H6OS 17 C2H6OS F.W. 1105.18 2430.82 822.77 829.73 T (K) 120(2) 150(2) 150(2) 150(2)

space group C2/c P212121 P-1 P-1 a (Å) 27.6859(5) 16.5736(7) 8.7097(3) 8.6455(12) b (Å) 16.9435(4) 21.7897(8) 14.0738(5) 14.142(2) c (Å) 28.0346(6) 35.1329(14) 16.2065(6) 16.283(3)  (deg) 90.00 90 114.085(2) 113.240(9)  (deg) 101.6030(10) 90 91.028(2) 91.363(9)  (deg) 90.00 90 94.463(2) 94.361(10) V (Å3) 12882.2(5) 12687.4(9) 1805.42(11) 1806.8(5) Z 8 4 2 2

−3 Dc (g·cm ) 1.140 1.273 1.513 1.525  (mm−1) 1.367 1.318 1.958 2.178 no. reflns collcd 13994 51717 31341 30192 no. indept reflns 13994 22077 6985 6242 GOF on F2 1.046 1.035 1.029 0.972

R [I > 2σ(I)] R1 = 0.0547 R1 = 0.0610 R1 = 0.0589 R1 = 0.0523

wR2 = 0.1528 wR2 = 0.1343 wR2 = 0.1475 wR2 = 0.1373

R (all data) R1 = 0.0790 R1 = 0.0859 R1 = 0.0774 R1 = 0.0659

wR2 = 0.1671 wR2 = 0.1430 wR2 = 0.1621 wR2 = 0.1448

52

Table 2.4. Selected bond lengths (Å) and angles (deg) of TbLH.

TbLH Tb(1)-O(4) 2.293(3) O(6)-Tb(1)-O(7) 74.96(14) Tb(1)-O(5) 2.307(3) O(3)-Tb(1)-O(7) 136.09(13) Tb(1)-O(6) 2.322(3) O(8)-Tb(1)-O(7) 69.29(15) Tb(1)-O(3) 2.395(4) O(4)-Tb(1)-O(1) 146.46(13) Tb(1)-O(4) 2.293(3) O(5)-Tb(1)-O(1) 79.48(13) Tb(1)-O(5) 2.307(3) O(6)-Tb(1)-O(1) 136.28(13) Tb(1)-O(6) 2.322(3) O(3)-Tb(1)-O(1) 73.28(12) Tb(1)-O(3) 2.395(4) O(8)-Tb(1)-O(1) 72.37(14) Tb(1)-O(8) 2.416(4) O(7)-Tb(1)-O(1) 100.09(14) Tb(1)-O(7) 2.433(4) O(4)-Tb(1)-O(2) 154.10(12) Tb(1)-O(1) 2.450(4) O(5)-Tb(1)-O(2) 129.15(13) Tb(1)-O(2) 2.480(4) O(6)-Tb(1)-O(2) 85.95(13) Tb(1)-O(4A) 2.851(3) O(3)-Tb(1)-O(2) 71.77(12) O(4)-Tb(1)-O(5) 76.48(12) O(8)-Tb(1)-O(2) 101.42(14) O(4)-Tb(1)-O(6) 77.23(12) O(7)-Tb(1)-O(2) 70.41(13) O(5)-Tb(1)-O(6) 129.70(13) O(1)-Tb(1)-O(2) 52.84(13) O(4)-Tb(1)-O(3) 123.63(12) O(4)-Tb(1)-O(4A) 75.32(13) O(5)-Tb(1)-O(3) 79.12(13) O(5)-Tb(1)-O(4A) 64.90(12) O(6)-Tb(1)-O(3) 80.96(14) O(6)-Tb(1)-O(4A) 67.33(12) O(4)-Tb(1)-O(8) 79.39(13) O(3)-Tb(1)-O(4A) 48.31(10) O(5)-Tb(1)-O(8) 76.19(15) O(8)-Tb(1)-O(4A) 137.35(12) O(6)-Tb(1)-O(8) 138.19(15) O(7)-Tb(1)-O(4A) 140.68(12) O(3)-Tb(1)-O(8) 140.61(13) O(1)-Tb(1)-O(4A) 114.58(12) O(4)-Tb(1)-O(7) 86.13(13) O(2)-Tb(1)-O(4A) 116.19(12) O(5)-Tb(1)-O(7) 143.64(15) O(6)-Tb(1)-O(7) 74.96(14)

53

Table 2.5. Selected bond lengths (Å) and angles (deg) of EuLH.

EuLH Eu(1)-O(13) 2.399(7) O(4)-Eu(1)-O(14) 78.01(19) O(7)-Eu(2)-O(12) 146.4(2) Eu(1)-O(8) 2.424(6) O(6)-Eu(1)-O(14) 74.8(2) O(5)-Eu(2)-O(12) 133.8(2) Eu(1)-O(9) 2.455(6) O(1)-Eu(1)-O(14) 136.6(2) O(2)-Eu(2)-O(12) 82.57(19) Eu(1)-O(14) 2.475(6) O(13)-Eu(1)-O(14) 68.6(2) O(3)-Eu(2)-O(12) 72.5(2) Eu(1)-O(10) 2.491(6) O(8)-Eu(1)-O(14) 139.5(2) O(15)-Eu(2)-O(12) 97.9(2) Eu(1)-O(7) 2.802(6) O(9)-Eu(1)-O(14) 72.2(2) O(7)-Eu(2)-O(16) 79.1(2) Eu(2)-O(7) 2.346(6) O(4)-Eu(1)-O(10) 155.5(2) O(5)-Eu(2)-O(16) 138.3(2) Eu(2)-O(5) 2.353(5) O(6)-Eu(1)-O(10) 129.4(2) O(2)-Eu(2)-O(16) 75.1(2) Eu(2)-O(2) 2.355(6) O(1)-Eu(1)-O(10) 86.24(19) O(3)-Eu(2)-O(16) 136.79(19) Eu(2)-O(3) 2.397(6) O(13)-Eu(1)-O(10) 70.4(2) O(15)-Eu(2)-O(16) 69.5(2) Eu(2)-O(15) 2.419(7) O(8)-Eu(1)-O(10) 70.9(2) O(12)-Eu(2)-O(16) 71.0(2) Eu(2)-O(12) 2.439(6) O(9)-Eu(1)-O(10) 53.2(2) O(7)-Eu(2)-O(11) 151.9(2) Eu(2)-O(16) 2.447(6) O(14)-Eu(1)-O(10) 101.9(2) O(5)-Eu(2)-O(11) 82.21(18) Eu(2)-O(11) 2.511(6) O(4)-Eu(1)-O(7) 75.79(19) O(2)-Eu(2)-O(11) 132.7(2) Eu(2)-O(4) 2.813(6) O(6)-Eu(1)-O(7) 67.75(19) O(3)-Eu(2)-O(11) 73.4(2) O(4)-Eu(1)-O(6) 74.6(2) O(1)-Eu(1)-O(7) 66.6(2) O(15)-Eu(2)-O(11) 69.1(2) O(4)-Eu(1)-O(1) 78.0(2) O(13)-Eu(1)-O(7) 140.1(2) O(12)-Eu(2)-O(11) 53.30(19) O(6)-Eu(1)-O(1) 131.1(2) O(8)-Eu(1)-O(7) 49.23(18) O(16)-Eu(2)-O(11) 101.7(2) O(4)-Eu(1)-O(13) 87.3(2) O(9)-Eu(1)-O(7) 115.6(2) O(7)-Eu(2)-O(4) 75.13(19) O(6)-Eu(1)-O(13) 141.9(2) O(14)-Eu(1)-O(7) 138.69(17) O(5)-Eu(2)-O(4) 67.98(19) O(1)-Eu(1)-O(13) 74.6(2) O(10)-Eu(1)-O(7) 115.0(2) O(2)-Eu(2)-O(4) 66.11(18) O(4)-Eu(1)-O(8) 124.8(2) O(7)-Eu(2)-O(2) 75.1(2) O(3)-Eu(2)-O(4) 50.18(17) O(6)-Eu(1)-O(8) 80.0(2) O(5)-Eu(2)-O(2) 131.59(19) O(15)-Eu(2)-O(4) 139.4(2) O(1)-Eu(1)-O(8) 83.5(2) O(7)-Eu(2)-O(3) 125.0(2) O(12)-Eu(2)-O(4) 118.0(2) O(13)-Eu(1)-O(8) 136.4(2) O(5)-Eu(2)-O(3) 84.4(2) O(16)-Eu(2)-O(4) 137.63(18) O(4)-Eu(1)-O(9) 144.4(2) O(2)-Eu(2)-O(3) 78.0(2) O(11)-Eu(2)-O(4) 116.8(2) O(6)-Eu(1)-O(9) 79.2(2) O(7)-Eu(2)-O(15) 85.4(2) O(1)-Eu(1)-O(9) 137.5(2) O(5)-Eu(2)-O(15) 73.7(2) O(13)-Eu(1)-O(9) 99.2(2) O(2)-Eu(2)-O(15) 142.1(2) O(8)-Eu(1)-O(9) 72.4(2) O(3)-Eu(2)-O(15) 138.5(2)

54

Table 2.6. Selected bond lengths (Å) and angles (deg) of EuLS.

EuLS Eu(1)-O(7) 2.390(5) O(4)-Eu(1)-O(6) 77.78(19) Eu(1)-O(4) 2.418(5) O(3A)-Eu(1)-O(6) 146.26(17) Eu(1)-O(3A) 2.424(5) O(8)-Eu(1)-O(6) 72.80(18) Eu(1)-O(8) 2.424(5) O(7)-Eu(1)-O(1) 125.31(19) Eu(1)-O(6) 2.441(5) O(4)-Eu(1)-O(1) 81.54(19) Eu(1)-O(1) 2.448(5) O(3A)-Eu(1)-O(1) 77.36(17) Eu(1)-O(2) 2.480(5) O(8)-Eu(1)-O(1) 143.66(17) Eu(1)-O(5) 2.480(5) O(6)-Eu(1)-O(1) 136.12(17) Eu(1)-O(3) 2.648(5) O(7)-Eu(1)-O(2) 75.8(2) O(7)-Eu(1)-O(4) 151.7(2) O(4)-Eu(1)-O(2) 124.11(19) O(7)-Eu(1)-O(3A) 79.51(19) O(3A)-Eu(1)-O(2) 84.44(17) O(4)-Eu(1)-O(3A) 119.10(17) O(8)-Eu(1)-O(2) 149.73(18) O(7)-Eu(1)-O(8) 76.26(19) O(6)-Eu(1)-O(2) 111.11(19) O(4)-Eu(1)-O(8) 86.15(19) O(1)-Eu(1)-O(2) 53.21(18) O(3A)-Eu(1)-O(8) 79.17(17) O(7)-Eu(1)-O(5) 99.3(2) O(7)-Eu(1)-O(6) 75.9(2) O(4)-Eu(1)-O(5) 72.50(18)

55

Table 2.7. Selected bond lengths (Å) and angles (deg) of TbLS.

TbLS Tb(1)-O(7) 2.371(5) O(6)-Tb(1)-O(1) 135.62(17) Tb(1)-O(8) 2.391(5) O(7)-Tb(1)-O(5) 100.62(19) Tb(1)-O(3A) 2.395(5) O(8)-Tb(1)-O(5) 125.32(16) Tb(1)-O(4) 2.405(5) O(3A)-Tb(1)-O(5) 155.20(16) Tb(1)-O(6) 2.421(5) O(4)-Tb(1)-O(5) 72.64(18) Tb(1)-O(1) 2.421(5) O(6)-Tb(1)-O(5) 53.42(17) Tb(1)-O(5) 2.452(5) O(1)-Tb(1)-O(5) 83.02(16) Tb(1)-O(2) 2.463(5) O(7)-Tb(1)-O(2) 75.59(18) Tb(1)-O(3) 2.653(5) O(8)-Tb(1)-O(2) 149.32(18) O(7)-Tb(1)-O(8) 76.12(18) O(3A)-Tb(1)-O(2) 84.20(16) O(7)-Tb(1)-O(3A) 79.03(18) O(4)-Tb(1)-O(2) 125.05(18) O(8)-Tb(1)-O(3A) 78.99(16) O(6)-Tb(1)-O(2) 110.64(18) O(7)-Tb(1)-O(4) 151.84(19) O(1)-Tb(1)-O(2) 53.90(17) O(8)-Tb(1)-O(4) 85.61(18) O(5)-Tb(1)-O(2) 71.88(17) O(3A)-Tb(1)-O(4) 118.66(16) O(7)-Tb(1)-O(3) 135.55(17) O(7)-Tb(1)-O(6) 76.30(19) O(8)-Tb(1)-O(3) 69.09(15) O(8)-Tb(1)-O(6) 73.52(17) O(3A)-Tb(1)-O(3) 68.07(17) O(3A)-Tb(1)-O(6) 146.60(16) O(4)-Tb(1)-O(3) 51.04(16) O(4)-Tb(1)-O(6) 78.08(18) O(6)-Tb(1)-O(3) 117.25(17) O(7)-Tb(1)-O(1) 125.68(18) O(1)-Tb(1)-O(3) 76.05(15) O(8)-Tb(1)-O(1) 143.23(16) O(5)-Tb(1)-O(3) 121.80(16) O(3A)-Tb(1)-O(1) 77.49(16) O(2)-Tb(1)-O(3) 127.17(16) O(4)-Tb(1)-O(1) 81.39(18) O(6)-Tb(1)-O(1) 135.62(17)

56

As-synthesized

Calculated

10 20 30 2 (degree)

Figure 2.11. PXRD pattern of as-synthesized TbLH

As-synthesized

Calculated

10 20 30 2 (degree)

Figure 2.12. PXRD pattern of as-synthesized EuLH

57

As-synthesized

Calculated

10 20 30 2 (degree)

Figure 2.13. PXRD pattern of as-synthesized EuLS

As-synthesized

Calculated

10 20 30 2 (degree)

Figure 2.14. PXRD pattern of as-synthesized TbLS

58

References

1. T. M. Reineke, M. Eddaoudi, D. Moler, M. O'Keeffe, O. M. Yaghi, J. Am. Chem. Soc. 2000, 122, 4843-4844. 2. B. T. Nguyen Pham, L. M. Lund, D. Song, Inorg. Chem. 2008, 47, 6329-6335. 3. (a) C. A. Bauer, T. V. Timofeeva, T. B. Settersten, B. D. Patterson, V. H. Liu, B. A. Simmons, M. D. Allendorf, J. Am. Chem. Soc. 2007, 129, 7136-7144. (b) J. Yang, J. Ma, S. R. Batten, Z. Su, Chem. Commun. 2008, 2233-2235. (c) F. P. Doty, C. A. Bauer, A. J. Skulan, P. G. Grant, M. D. Allenforf, Adv. Mater. 2009, 21, 95-101. (d) J. Yang, J. Ma, Y. Liu, S. R. Batten, CrystEngComm 2009, 11, 151-159. (e) L. Zhang, Y. Yao, Y. Che, J. Zheng, Cryst. Growth Des 2010, 10, 528-533 4. J. E. McMurry, M. P. Fleming, K. L. Kees, L. R. Krepski, J. Org. Chem. 1978, 43, 3255- 3266. 5. J. Liu, J. Chen, C. Xia, J. Catal. 2008, 253, 50-56. 6. Apex 2 Software Package; Bruker AXS inc., 2008 7. G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr. 2008, 64, 112-122. 8. (a) A. L. Spek, Acta Crystallogr. 1990, A46, C34. (b) A. L. Spek, PLATON, A Multipurpose Crystallographic Tool; Utrecht University: Utrecht, The Netherlands, 2000. 9. L. Ma, C. Abney, W. Lin, Chem. Soc. Rev. 2009, 38, 1248-1256. 10. J. E. McMurry, M. P. Fleming, K. L. Kees, L. R. Krepski, J. Org. Chem. 1978, 43, 3255- 3266. 11. T. M. Reineke, M. Eddaoudi, D. Moler, M. O'Keeffe, O. M. Yaghi, J. Am. Chem. Soc. 2000, 122, 4843-4844. 12. (a) B. T. Nguyen Pham, L. M. Lund, D. Song, Inorg. Chem. 2008, 47, 6329-6335; (b) S. Viswanathan, A. de Bettencourt-Dias, Inorg. Chem. 2006, 45, 10138-10146. 13. D. C. Wilson, S. Liu, X. Chen, E. A. Meyers, X. Bao, A. V. Prosvirin, K. R. Dunbar, C. M. Hadad, S. G. Shore, Inorg. Chem. 2009, 48, 5725-5735.

59

Chapter 3 Lanthanide Metal-Organic Framework of (2',5'- bis(methoxymethyl)-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid and its Potential as Solvent Vapor Sensor 3 3.1 Abstract

A new luminescent lanthanide MOFs synthesized from ether functionalized ligand (2',5'-

bis(methoxymethyl)-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid, H2sLOM), displays efficient turn-on triggered by solvent vapors, with excellent selectivity for DMF vapor. Sensor response rate and recyclability have been tested under our prototype sensor setup, showing fast response rate and reversible signal change. The sensing mechanism is studied through luminescence

lifetime measurements of samples prepared with H2O/D2O and the results indicate that, while solvent-metal interaction contributes to the turn-on, strong solvent-ligand interaction is the major contributor to the luminescence response. Potential solvent-ligand interaction has been proposed based on crystal structure of as-synthesized MOF.

3.2 Introduction

Our early exploration of luminescent lanthanide MOF sensors based on functionalized (E)-4,4’- (ethane-1,2-diyl)dibenzoic acids encountered major difficulties. First, the use of an olefin spacer

group introduced trans/cis isomer, which reduced the efficiency of ligand synthesis. Although LB ester can be easily converted to different functionalized ligands, its synthesis involves a rather

difficult and time-consuming column chromatography to isolate the desired trans LB ester from the cis isomer, which limits the reaction scale to a maximum of about 1 g. Given that the crystal growth is a try-and-error process, it is ideal to have an efficient ligand synthesis so more efforts

can be spared towards growing MOF crystals. Secondly, MOFs built with H2LH and H2LS do not

show the characteristic lanthanide emission except for EuLS, which only displayed limited LMET efficiency. These observations suggest that this ligand system is far from optimal for the sensing application. Therefore, we decided to look at other alternative ligand systems. One of the ligand systems we decided to try is based on a phenyl spacer (Scheme 3.1), which can be synthesized from the Suzuki coupling between different phenyl dibromide and phenyl boronic

60 acid with carboxylate group. Functionalization can be done either before or after the Suzuki coupling, depending on the functional group compatibility. Methyl ether group was picked first because of the straightforward synthesis.

Interesting solvent-dependent luminescence properties was first observed on the MOF built from this ligand, which later leads to the development of the turn-on MOF sensor for DMF vapor. While luminescence turn-on upon exposure to liquid solvents have been demonstrated for MOFs,[1] to the best of our knowledge, there is no LnMOF sensor and only one transition metal counterpart displaying luminescence turn-on towards gaseous analytes in the literature.[2] Recent studies have suggested serious health concerns associated with DMF exposure both through skin contact and inhalation in various industries, such as the production of synthetic fibers, film, coating, and leather tanning. Studies of exposed workers[3] and animal experiments[4] have confirmed the hepatotoxicity of DMF; other toxic effects have been suggested including embryotoxicity[5] and carcinogenesis.[6] Because of the large population at risk and high production rate, DMF has been prioritized for field studies. The standard sampling and detection method[7] for DMF vapor in air requires lengthy procedures involving physical absorption/desorption of air samples and the use of gas chromatography. A reusable, selective, and easy-to-use sensor for DMF vapor is desirable.

X X O O O B(OH)2 + Br Br O O O X X

Scheme 3.1. General strategy for H2sLX ligand synthesis.

3.3 Experimental section

3.3.1 Materials and method

Elemental analyses were performed in our Chemistry Department on a PE 2400 C/H/N/S analyzer. Thermogravimetric analyses (TGA) were performed on a TA Instruments SDT Q600 instrument under dinitrogen atmosphere with a heating rate of 10 °C per minute. NMR spectra were recorded on a Varian 400, or a Bruker Avance 400 spectrometer. Both 1H and 13C NMR spectra were referenced and reported relative to the solvent residual signals. Photoluminescence

61 spectra were measured using QuantaMaster 30 system from Photon Technology International (Canada) Inc. equipped with a 75W high intensity Xenon lamp (UXL-75XE), a R928 PMT detector, and a solid sample holder; the incident and detector slit widths are 1 and 0.4 nm, respectively. The powder XRD experiments were performed on PANalytical machine with generator PW1830, control unit PW3710 and the proportional gas detector PW3011, equipped with a Cu sealed tube (λ = 1.54178 Å). Unless otherwise stated, all manipulations were performed in air and all reagents were purchased from commercial sources and used without further purification. Compressed air cylinder was purchased from Linde Canada. 2',5'-

Bis(bromomethyl)-[1,1':4',1''-terphenyl]-4,4''-dicarboxylate (sLBr ester) was synthesized following the literature method.[8]

3.3.2 Synthesis of 2',5'-bis(methoxymethyl)-[1,1':4',1''-terphenyl]-4,4''- dicarboxylic acid (H2sLOM)

The mixture of sLBr ester (3.0 g, 5.64 mmol) and sodium hydroxide (4.0 g, 71.4 mmol) was suspended in methanol (100 mL) and refluxed for 24 h. After cooling to ambient temperature, methanol was removed in vacuo and the resulting solid was dissolved in H2O. The aqueous solution was then acidified with 1M HCl and the white precipitate of H2sLOM acid was collected 1 via filtration and dried at 60 °C under vacuum (2.16 g, 94% yield). H NMR (DMSO-d6, 400 MHz, 25°C)  8.04 (d, J = 8.4 Hz, 4H), 7.57 (d, J = 8.4 Hz, 4H), 7.46 (s, 2H), 4.35 (s, 4H), 3.24 13 (s, 6H). C NMR (DMSO-d6, 100 MHz, 25°C) d 167.04, 143.99, 139.51, 134.72, 130.46, 129.67,

129.21, 129.06, 71.12, 57.59. Anal.calcd for C24H22O6: C, 70.92; H, 5.46. Found: C, 70.55; H, 5.50.

3.3.3 Synthesis of [Eu2(sLOM)3(H2O)4] (EusLOM-1)

Eu(NO3)3∙6H2O (17.8 mg, 0.04 mmol) and H2sLOM (24.4 mg, 0.06 mmol) were dissolved in a mixture of DMF (8 mL) and H2O (2 mL) in a 20 mL scintillation vial and sealed. The vial was then heated in oil bath at 80°C for 3d. Colorless crystals of EusLOM-1 were collected via vacuum filtration (33.8 mg, 86% yield). Anal.calcd for C78H82Eu2N2O24 ∙ 3C3H7NO: C, 53.79; H, 4.96; N, 2.32. Found: C, 53.71; H, 5.06; N, 2.89. (TGA solvent weight loss Calcd 16.1%, Found 15.7%).

62

3.3.4 Preparation of water-exchanged MOF (EusLOM-2)

Crystals of EusLOM-1 were soaked in distilled water (~1 mL per 10 mg of EusLOM-1) for three days, collected by vacuum filtration, washed with water and ethanol, and then air dried.

3.3.5 Preparation of water-exchanged MOF for lifetime measurements (EusLOM-3-H2O/D2O).

Crystals of EusLOM-1 was soaked in distilled H2O or D2O (~1 mL per 10 mg of EusLOM-1) for three days, then the sample was lyophilized to remove free H2O/D2O. The residual solid was then transferred into a dry N2-filled glovebox (to avoid H-D exchange between D2O in the sample and moisture in air) for lifetime measurements.

3.3.6 X-ray crystallographic analysis.

X-ray quality single crystals of EusLOM (EusLOM-1) were obtained as described in the sections above. All crystals were mounted on the tip of a MiTeGen MicroMount. The single-crystal X-ray diffraction data were collected on a Bruker Kappa Apex II CCD diffractometer with Mo Kα radiation ( = 0.71073 Å) operating at 50 kV and 30 mA, at 150 K controlled by an Oxford Cryostream 700 series low temperature system. The data integration and absorption correction were performed with the Bruker Apex 2 software package. The structure of EusLOM-1 was solved by direct methods and refined using SHELXTL V6.14.[9] Disordered DMF molecules in both structure are modeled successfully. All non-hydrogen atoms except for the atoms involved in the disordered portions were refined anisotropically. The positions of the hydrogen atoms were calculated using the riding model. Crystallographic table of is provided in Table 3.4 at the end of this chapter.

3.3.7 Measurement for maximum luminescence change

Powder samples of EusLOM-2 were used for sensing experiments. For each experiment, 20-25 mg of EusLOM-2 was placed into a quartz cuvette and the luminescence spectrum was measured in a solid sample holder before exposure to various solvent vapors. The powder and the cuvette were then placed into a sealed container (usually ~100 mL bottle) which contains ~5 mL of a given solvent (Figure 3.1) for 24 h. Subsequently the cuvette was taken out of the container, quickly sealed and the emission spectrum was taken again in the solid sample holder. In order to

63 reproduce the intensity, the front face of the cuvette exposed to the incidental beam has to be covered by the powder sample completely.

Figure 3.1. Diagram of the experimental setup for incubating 2 under different solvent vapors.

The Eu emission intensity at 616 nm was used for comparison among all the samples after exposure to solvent vapors. For selectivity test (Figure 3.9), luminescence enhancement for each solvent was measured four times with 4 different batches of samples. The average of Iafter/Ibefore-1

(Iafter, Ibefore stand for emission intensity at 616 nm after and before exposure to solvent vapor, respectively) was used as the height of the column and the standard deviations were drawn as the error bar.

3.3.8 Preparation of recycled MOF sample and the recycling experiments

Samples of EusLOM-2 previously used in sensing experiments were combined together and regenerated by soaking in distilled water for 3 days. After that, the solid was collected via filtration, washed with water and ethanol then air dried. The resulting powder was used to measure the maximum luminescence change after incubation in various solvent vapors using the same procedure described in the above section.

3.3.9 Quantification of solvent uptake in MOF samples

Powder samples of EusLOM-2 after incubated with various solvent vapors or drying reagent

MgSO4, were weighted (6 - 10 mg) and then suspended in 0.7 mL acetone-d6 in NMR tubes. 2.5 μL of mesitylene was added to the suspension as the internal standard. A control group was set 1 up by adding 2.5 μL of mesitylene to 0.7 mL acetone-d6. H NMR spectra of the suspensions

64 were taken using 25 s of relaxation delay to ensure the accuracy of the integration. The amount of water and organic solvent were determined according to the NMR experiments; water content was calculated from both H2O and HDO signals, and further adjusted by subtracting the amount of water in the control group to exclude the water introduced from acetone-d6, moisture or mesitylene. Assumption was made that four coordinated water molecules per Eu2(sLOM)3 units were not exchanged. NMR spectra with integration are attached in Figure 3.17 at the end of this chapter.

3.3.10 Preparation of water-exchanged MOF (EusLOM-3-H2O/D2O) and the lifetime study.

1 The completeness of D2O exchange for EusLOM-3-D2O was confirmed by taking H NMR of

EusLOM-3-H2O/D2O sample suspension: about 10 mg of EusLOM-3-H2O and EusLOM-3-D2O were weighted, suspended in acetonitrile-d3 in two NMR tubes and 5.0 μL of toluene was added to each sample as internal standard. The H2O signal at 2.13 pm is well observed in NMR spectra of EusLOM-3–H2O (Figure 3.16a), but is completely absent in that of EusLOM-3-D2O (Figure

3.16b), which confirms the complete exchange of D2O.

In a typical experiment, 20-40 mg of EusLOM-3-H2O/D2O was placed into a quartz cuvette in a dry N2-filled glovebox, sealed and then taken out for luminescence measurements. The sample was then shipped back into the glovebox, placed into a sealed container (usually ~100 ml bottle) to incubate for 24 h with solvent without direct contact between solvent and the powder sample (Figure 3.1). After that, the cuvette was sealed and taken out of the glovebox for measurements. The phosphorescence lifetime was calculated using PTI Flex32 software.

3.3.11 Prototype sensor experimental setup

The whole experimental setup for the prototype sensor experiments are shown in Figure 3.2. The sensor slides were prepared by sticking powder of EusLOM-2 (less than 1 mg) onto a glass slide through a double-sided tape. The slide was then attached to a piece of plastic foam and placed into a cuvette on the solid sample holder. Compressed air from cylinder was bubbling through DMF or water at a rate of 250 mL/min and then directed into the cuvette containing the sensor slide. During the experiments, DMF and water vapor was introduced in alternating manner; when switching solvent, the gas flow was stopped for 30 s but the data collection wasn’t stopped.

65

Small disturb to the intensity was sometimes observed during the switch. The emission intensity @ 616 nm was monitored throughout the experiments with excitation @ 323 nm.

Figure 3.2. Experimental setup for the prototype sensor experiment.

We found in our early sensing experiments that the response rate of the sensor slide is very sensitive to the experimental setup. DMF in the vial has to be pre-bubbled for 30 min before use to make sure DMF is saturated in the atmosphere inside the vial, and the tubes in the experiment setup need to be purged for 30 min before use. The spacial arrangement of the two needles in the cuvette also has a big impact on the response rate and it should not be changed between experiments in order to reproduce a similar response rate.

3.4 Results and Discussion

3.4.1 MOF synthesis and Structural description

Needle-shaped crystals of [Eu2(sLOM)3(H2O)4]∙3DMF (EusLOM-1) can be obtained in high yield by heating a solution of Eu(NO3)3 and H2sLOM in DMF/H2O mixture (v:v=4:1) in a sealed scintillation vial at 80 °C for 3 days. The asymmetric unit of EusLOM-1 is shown in Figure 3.3 and Figure 3.4. Each Eu center is eight-coordinate, adopting bi-capped trigonal prismatic geometry with six oxygen donors from carboxylates and two from coordinated water molecules occupying the eight coordination sites. The trigonal prism is defined by O1, O8, O10, O7, O2

66 and O11; O5 and O6 caps two of the three rectangular faces. Only the bridging coordination mode is adopted by the carboxylate groups: adjacent pairs of Eu centers are doublly or quadruply bridged by carboxylates in alternating order to form 1-D chains along the a-axis (Figure 3.4a).

Each 1-D chain is linked with 4 adjacent chains via sLOM ligands, extending the structure into a 3-D framework with solvent channels along the a-axis, where DMF molecules are located. The channel DMF molecules can be replaced by soaking the crystals of EusLOM-1 in distilled water -1 for 3 days, resulting in the water-exchanged framework (EusLOM-2). The DMF νc=o at 1660 cm disappeared in FTIR of EusLOM-2 (Figure 3.5) and N element is no longer detected in elemental analysis (Anal.calcd for [Eu2(sLOM)3(H2O)4]∙13H2O: C, 47.43; H, 5.20; N, 0. Found: C, 47.26; H, 5.11; N, 0.), suggesting complete removal of channel DMF molecules. PXRD shows that

EusLOM-2 is still crystalline (Figure 3.6).

Figure 3.3. (a) Coordination sphere of Eu; (b) The asymmetric unit of EusLOM-1 with thermal ellipsoids plotted at 50% probabilities. All hydrogen atoms and cocrystallized DMF solvent molecules are omitted for clarity.

Table 3.1. Selective bond length in crystal structure of EusLOM-1

Bond Bond length Bond Bond Length

Eu(1)-O(2)#1a, b 2.345(4) Eu(1)-O(1) 2.436(4)

Eu(1)-O(8) 2.354(4) Eu(1)-O(5)#3 2.398(4)

Eu(1)-O(6)#2 2.364(4) Eu(1)-O(11) 2.478(5)

Eu(1)-O(7)#1 2.395(4) Eu(1)-O(10) 2.565(4)

[a]. Symmetry transformations used to generate equivalent atoms: #1 -x-2,-y+2,-z; #2 -x-2,-y+2,-z-1; #3 x-1,y,z+1. [b] O(11) and O(12) are oxygen atoms from coordinated water, and the rest of oxygen atoms are from carboxylate groups. All Eu-O bond length are within literature values except Eu(1)-O(10), likely due to the weak interaction between Eu and coordinated water.

67

Figure 3.4. Crystal structure of EusLOM: (a) 1D chain with bridging carboxylates; (b) extended structure showing connectivity; (c) perspective view of 1D chains and open channels. All hydrogen atoms and channel solvent molecules are omitted for clarity.

68

3.1 (As-synthesized)

-1 DMF  =1660 cm c=o

3.2 (Water-exchanged)

4000 3500 3000 2500 2000 1500 1000 500

-1 Wavenumber (cm )

Figure 3.5. FTIR of as-synthesized MOF (EusLOM-1) and water-soaked MOF (EusLOM-2).

Water-soaked

As-synthesized

Predicted

5 10 15 20 25 30 2 (deg)

Figure 3.6. PXRD of As-synthesized sample (EusLOM-1) and water-soaked sample (EusLOM-2). “Predicted” is the calculated pattern from crystal structure of EusLOM-1.

69

3.4.2 Thermogravimetric analysis

TGA (Figure 3.7) shows that EusLOM-1 starts losing solvent molecules at around 75 °C and become solvent free at 210 °C. Thereafter the solvent free framework is stable up to 360 °C after which the framework starts to decompose. The weight loss of solvent was found to be 15.7%, which matched well with 16.1%, calculated value from elemental analysis.

TGA of 3.1 100

90

80

70 Weight (%)Weight

60

50

40 0 100 200 300 400 500 600 700 800 Temperature (oC)

Figure 3.7. Thermogravimetric analysis of EusLOM-1. 3.4.3 Luminescence properties

MOF EusLOM-1 displays characteristic Eu emissions at 578, 590, 616, 698 nm in the emission 5 7 5 7 5 7 5 7 spectrum (Figure 3.8), corresponding to D0 – F0, D0 – F1, D0 – F2, D0 – F4 transitions of Eu, respectively. No significant ligand-based emission was observed, thus the LMET process is efficient. In the excitation spectrum, only a broad band of ligand-based excitation is observed around 341 nm, indicating ligand-based excitation through LMET process. The water-exchanged framework EusLOM-2 displays a much weaker luminescence under UV irradiation (Figure 3.8).

This finding prompted us to examine the solvent-dependent luminescence of EusLOM-2.

70

Excitation of 3.1 20 Emission of 3.1 18 Emission of 3.2

16

14

12

10

8

6 Intensity (A.U.) Intensity

4

2

0

-2 250 300 350 400 450 500 550 600 650 700 750 Wavelength (nm)

Figure 3.8. Luminescence spectra of EusLOM-1 and EusLOM-2. Excitation and emission are measured at 616 and 323 nm respective. The peaks at 308 nm in excitation spectra and 646 nm in emission spectra are due to the scattering. 3.4.4 Solvent-dependent luminescence and sensing experiments

Luminescence responses of EusLOM-2 upon 24 h incubation under various solvent vapors have been measured and the results are plotted in Figure 3.9. Original emission spectra before and after the incubation are attached in Figure 3.19 at the end of the chapter. All incubated samples except the one with formamide reached maximum in 24 h; emission intensity of formamide incubated sample continued to increase slowly to 1.3-fold enhancement after 96 h, likely due to its low vapour pressure (11 Pa at 20 °C). The histogram clearly shows that while most solvents give around 1-fold enhancement, DMF displays a superior turn-on (Figure 3.9 – Figure 3.11), i.e., >8-fold of luminescence enhancement. The used powder of EusLOM-2 can be regenerated by soaking in water and reused in aforementioned experiments without notable loss of the selectivity (Figure 3.12). EusLOM-1 also displayed similar turn-on response but less folds of turn-on, due to the higher initial emission intensity. It is worth noting that, because of the simplicity of solvent composition, EusLOM-1 would be a better candidate for sensing application.

71

Regeneration of EusLOM-1 would require careful control of DMF/H2O ratio in order to reproduce the same initial emission intensity.

Figure 3.9. Increase in Eu emission intensity of EusLOM-2 (Iafter/Ibefore-1) after 24 h incubation under various solvent vapors and anhydrous MgSO4. The intensity is measured at 616 nm. Error bars indicate the standard deviations of three or four parallel experiments.

Figure 3.10. Emission spectra of 3.2 before and after exposure to DMF vapor (excited at 323 nm). The broad peak around 640 nm is scattering peak.

72

Figure 3.11. Photo of the sensor at on and off stages under a portable UV lamp irradiation. Note: the 'on' stage luminescence is so bright that it over-saturates the camera.

8

6

4

2 Luminescenceenhancement (fold) 0 DMF Methanol Hexanes Ethyl acetate

Figure 3.12. Luminescence responses of the bulk regenerated sensor (by soaking in water) to selected solvent vapors upon 24 h incubation. The four bars for each solvent stand for fresh sample, 1st cycle, 2nd cycle, 3rd cycle, respectively.

73

3.4.5 Luminescence lifetime study and sensing mechanism

[10] The O-H bond of coordinating H2O is known as a good quencher for Eu luminescence. We reason that when the channel H2O molecules are partially replaced by other solvent molecules, the equilibrium between coordinating and channel H2O will shift, leaving fewer O-H bonds around the Eu centers and thus triggering enhancement of Eu emission. Such a hypothesis is consistent with the fact that anhydrous MgSO4, which may remotely remove water from

EusLOM-2, can enhance the emission as well. Our NMR experiments (Table 3.2, NMR spectra attached in Figure 3.17 at the end of this chapter) also confirmed the uptake of various solvents and the decrease of water content in EusLOM-2 after 24 h incubation. However, such a decrease in water content might only reflect the “exchangeable” water (or channel water). To investigate the behavior of the coordinated water during the turn-on event, we measured the lifetime of Eu emission, before and after exposure to DMF vapor, with MOF samples EusLOM-3-H2O and

EusLOM-3-D2O, which are prepared by soaking compound EusLOM-1 in H2O and D2O and subsequently lyophilized. The lifetime measurements were first attempted in air with D2O soaked sample EusLOM-2 (EusLOM-2-D2O), but the luminescence lifetime was found to be very close to that of EusLOM-2-H2O, presumably because of the fast exchange between D2O in EusLOM-2-D2O and the moisture in air. Therefore, we decided to move the experiments into the glovebox, which required us to lyophilize the samples. Consequently, channel/coordinated water in EusLOM-3-

H2O/D2O will be partially removed, resulting in lower water content compared to EusLOM-2 sample.

Table 3.2. Solvent content after incubated with various solvent or drying reagent MgSO4

Solvent incubated Amount of H2O (per Eu2(sLOM)3 unit) Amount of solvent (per Eu2(sLOM)3 unit)

None a 19.9 N/A

b MgSO4 10.8 N/A

MeCN 3.5 10.0

EtOAc 5.3 9.3

Formamide 1.7 1.2

DMF 0.9 6.6

Benzene 5.6 5.6

[a] Powder sample EusLOM-2 was used without incubation. [b] Anhydrous MgSO4 was used instead of any solvent.

74

The completeness of deuteration has been confirmed by NMR experiments (Figure 3.16). According to an early study,[11] the luminescence quenching rate contributed by O-H bonds is proportional to the number of coordinating water molecules in the first coordination sphere of Eu and this quenching rate displays such a strong isotope effect that O-D bond essentially has no quenching effect. The average number of coordinating water molecules per Eu center can be calculated using Eq.1:

−1 -1 nH2O = 1.05 (τH2O −τD2O ) (Eq. 3.1)

where τH2O and τD2O are the lifetimes of Eu emission in ms, for H2O and D2O containing samples, respectively. The numbers of coordinated water were calculated for EusLOM-3 before and after DMF exposure and results are shown in Table 3.3, together with intensity of Eu emission at 616 nm for comparison. Due to the presence of H2O in EusLOM-3-H2O, the lifetime of EusLOM-3-

H2O is much shorter than that of EusLOM-3-D2O, which calculates to be around 0.5 water molecule on each Eu. After exposure to DMF vapor, the difference in lifetime between EusLOM-

3-H2O and EusLOM-3-D2O is much smaller and almost all water molecules on Eu center have been removed. As a result, the Eu emission intensity is greater in EusLOM-3-D2O than in

EusLOM-3-H2O before exposure to DMF vapor, and both increase to the similar level after the exposure. These observations are consistent with our hypothesis of “water effect” that removal of coordinated water contributes to the turn-on.

However, the water effect does not account for all the luminescence turn-on of EusLOM-3 after exposure to DMF vapor. If we ascribe the difference in emission intensity between EusLOM-3-

H2O and EusLOM-3-D2O before exposure to DMF vapor (243 and 342, respectively, Table 3.3 top half) to water effect, the much greater luminescence enhancement on EusLOM-3-D2O after exposure to DMF vapor (1037, comparing to 342 before exposure) can only be explained as the solvent-specific effect, since no O-H bond is present in the first coordination sphere of Eu in

EusLOM-3-D2O. When the Eu emission intensity and lifetime was measured using the direct Eu- 2- based excitation at 466 nm to eliminate the involvement of ligand sLOM via LMET (Table 3.3, bottom half), the observed emission enhancement of EusLOM-3-D2O after exposure to DMF vapor is much weaker, suggesting that the major contributor of the strong turn-on effect is the more efficient LMET process rather than any local environment change at the metal sites. The selectivity for DMF could be attributed to the tailored solvent channels since the MOF channels

75 are built around DMF as the solvent template. In fact, in the powder XRD patterns of sample incubated in various solvent vapors (Figure 3.14), the breathing effect[12] triggered by solvent adsorption was observed, with DMF causing the most significant change among all solvent vapors tested. While other solvent may have similar size and shape, the MOF clearly highly favors DMF over others. In the crystal structure of EusLOM-1, three DMF molecules were located around the ligand, two of which may hinder the rotation of phenyl rings of the ligands (Figure 3.13). The constraint introduced by channel DMF, as well as the polarity of DMF, presumably influenced the ligand energy level, leading to a more efficient LMET process.

Table 3.3. Emission lifetime (τ) and intensity (Int) measured with ligand-based excitation and direct Eu excitation.

a τH2O @ 616 τD2O @ 616 No.of IntH2O @ 616 IntD2O @ 616 nm nm coordinating nm nm water (μs) (μs) (A.U.) (A.U.) molecules

Ligand-based excitation (@ 323 nm):

before DMFb 748 1134 0.48 243 342

after DMFb 1244 1310 0.04 972 1037

Direct Eu-based excitation (@ 466 nm):

before DMFb 688 1087 0.56 92 114

after DMFb 1090 1140 0.04 111 132

[a]. τH2O and τD2O are the luminescence lifetime for EusLOM-3-H2O and EusLOM-3-D2O respectively, with emission measured at

616 nm. IntH2O and IntD2O are the emission intensity at 616 nm for EusLOM-3-H2O and EusLOM-3-D2O respectively. [b]. “before DMF” and “after DMF” correspond to before and after exposure to DMF vapor respectively.

76

Figure 3.13. DMF molecules (shown with spacefilling model) that hinder the phenyl ring rotation. Top: DMF molecule is sandwiched between the central rings of two parallel ligands. Bottom: DMF molecules is located parallel to the side phenyl ring and the carboxylate.

Formamide

MeCN

EtOAc

Benzene

MgSO4

Regenerated

DMF

Water-soaked (3.2)

As-synthesed (3.1) Predicted

5 10 15 20 25 30 2 (deg)

Figure 3.14. PXRD pattern of as-synthesized sample (EusLOM-1), water-soaked sample (EusLOM-2), sample EusLOM- 2 after incubated with various solvent/reagent for 24 h, sample regenerated from DMF incubated sample by soaking in water for 3 days (regenerated), and the calculated pattern from the crystal structure of EusLOM-1.

77

3.4.6 Prototype sensing experiments and sensor response rate

Besides good selectivity, fast response is also an important criterion for a good sensor. In order to test the response rate of EusLOM-2, we designed a prototype sensor setup to monitor the real time luminescence response. The sensor slide was prepared by attaching powder of EusLOM-2 on a glass slide via double-sided tape. DMF and water vapors are introduced to the sensor slide in an alternating manner, and the luminescence intensity at 616 nm was monitored continuous with excitation at 323 nm. Luminescence intensity versus time is plotted in Figure 3.15, and more plots are attached in Figure 3.18 at the end of the chapter. Under our experimental conditions, the response rates to DMF vapor in different cycles or experiments are almost identical: in most cycles >95% of the luminescence enhancement can be achieved within the first minute; the turn- off caused by water vapor is even faster, >95% of the decrease within the first 10~20 seconds. In the first turn-on/off cycle, the luminescence enhancement ranges from 7 to 8 folds, and lower in later cycles. The decrease in the intensity of “on state” is not observed for recycled powder samples (Figure 3.12), suggesting that the slight intensity decrease over many on-off cycles shown in Figure 3.15 is likely due to the instability of the sensor slides or the incomplete removal of solvent through out the cycles.

Figure 3.15. On-off cycles of the sensor with alternating DMF-water vapor treatment. The intensity is measured at 616 nm with the excitation wavelength of 323 nm.

78

3.5 Summary

In summary, we have successfully synthesized a new luminescent lanthanide MOF which shows selective luminescence turn-on response to DMF vapor. The XRD, NMR, and luminescence lifetime studies suggest that the majority of the turn-on in response to DMF vapor is caused by DMF-ligand interaction that presumably alters the ligand excited state energy level and thus facilitates the LMET process. Such a turn-on mechanism opens up the possibility to build lanthanide-based turn-on luminescent MOF sensors by proper ligand design, targeting analytes through ligand-analyte interactions. We also found that the removal of water ligands from the Eu centers adds to the turn-on effect as a minor contributor. Our proto-type sensor showed fast responding rate.

79

Table 3.4. Crystallographic data of EusLOM

EusLOM-1

formula C72H68O22 Eu2 · 3.5 C3H7NO

F.W. 1845.02

T (K) 150(2)

space group P-1

a (Å) 9.3824(4)

b (Å) 13.6113(6)

c (Å) 18.2701(7)

 (deg) 74.7319(16)

 (deg) 86.3824(17)

 (deg) 88.5107(19)

V (Å3) 2246.27(16)

Z 1

Dc (g·cm−3) 1.364

 (mm−1) 1.456

no. reflns collcd 38308

no. indept reflns 10288

GOF on F2 1.028

R [I > 2σ(I)] R1 = 0.0597

wR2 = 0.1444

R (all data) R1 = 0.0876

wR2 = 0.1560

80

7.255 7.236 7.203 7.202 7.200 7.199 2.890 2.773 2.329 1.940

MeCN-d3

Toluene

Toluene DMF

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm (t1)

1 Figure 3.16a. H NMR of EusLOM-3-D2O suspended in acetonitrile-d3.

81

7.274 7.256 7.242 7.237 7.200 7.182 7.152 2.889 2.774 2.331 2.129 1.940

Toluene MeCN-d3

Toluene H2O

DMF

3.00 0.38

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 ppm (t1)

1 Figure 3.16b. H NMR of EusLOM-3-H2O suspended in acetonitrile-d3

82

6.769 2.829 2.223 2.050

2.829 2.795

10.98 (H2O) 0.85 (HDO)

10.98 0.85

3.00 2.90 2.80 2.70 2.60

ppm (t1)

2.90 10.98 0.85 9.00 2.22

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 ppm (t1)

1 Figure 3.17a. H NMR of 9.9 mg of EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene.

83

6.771 2.800 2.768 2.765 2.763 2.226 2.050

2.800 2.768 2.765 2.763

0.63 (H2O)

0.38 (HDO)

0.63 0.38

2.900 2.850 2.800 2.750 2.700 2.650

ppm (t1)

2.97 0.63 0.38 9.00 2.22

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 ppm (t1)

1 Figure 3.17b. H NMR of 2.5 μL mesitylene in 0.7 mL acetone-d6 (control group).

84

6.771 2.819 2.225 2.087 2.051

2.819 2.785

6.34 (H2O)

0.60 (HDO)

6.34 0.60

2.900 2.850 2.800 2.750 2.700 ppm (t1)

6.34 (H2O)

0.60 (HDO)

2.96 6.34 0.60 9.00 2.10

7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 3.17c. H NMR of 9.1 mg of MgSO4 incubated EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene.

85

6.771 2.805 2.225 2.048

2.805 2.773 2.771 2.225 2.048

2.69 (H2O)

0.42 (HDO)

2.69 0.42 9.00 11.62

2.80 2.70 2.60 2.50 2.40 2.30 2.20 2.10 2.00 ppm (t1)

2.69 (H2O)

0.42 (HDO)

2.93 2.69 0.42 9.00 11.62

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 ppm (t1)

1 Figure 3.17d. H NMR of 11.6 mg of MeCN incubated EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene.

86

8.151 8.120 8.118 6.771 3.694 3.692 2.807 2.225 2.087 2.051

2.807 2.773

1.79 (H2O)

0.45 (HDO)

1.79 0.45

3.00 2.90 2.80 2.70 2.60 ppm (t1) 1.79 (H2O)

0.45 (HDO)

0.48 3.04 0.53 1.79 0.45 9.00 2.29

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 ppm (t1)

1 Figure 3.17e. H NMR of 11.6 mg of formamide incubated EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene.

87

6.771 4.080 4.062 4.044 4.027 2.804 2.769 2.224 2.087 2.051 1.964 1.216 1.198 1.180

2.804 2.769

3.21 (H2O)

0.61 (HDO)

3.21 0.61

2.800 2.750 ppm (t1)

3.21 (H2O)

0.61 (HDO)

2.95 5.49 3.21 0.61 9.00 3.21 7.89 7.79

7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 ppm (t1)

1 Figure 3.17f. H NMR of 12.8 mg of EtOAc incubated EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene.

88

7.962 6.771 2.939 2.805 2.780 2.226 2.087 2.051

2.939 2.805 2.780

4.87 1.01 5.28

2.950 2.900 2.850 2.800 2.750 ppm (t1) 2.805 ppm - int 1.01 (H2O)

2.780 ppm - int 5.28 (HDO+DMF)

1.56 3.10 4.87 1.01 5.28 9.00 2.22

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm (t1)

1 Figure 3.17g. H NMR of 9.1 mg of DMF incubated EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene.

89

7.354 6.772 2.809 2.775 2.226 2.087 2.051

2.809 2.775

3.16 (H2O)

0.45 (HDO)

3.16 0.45

2.850 2.800 2.750

ppm (t1)

7.96 2.96 3.16 0.45 9.00 2.19

7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 3.17h. H NMR of 9.1 mg of benzene incubated EusLOM-2 suspended in 0.7 mL acetone-d6 and 2.5 μL mesitylene.

90

Start of DMF vapor flow 250 Start of water vapor flow Start of DMF vapor flow Stop of flow Start of water vapor flow 250 Stop of flow 200

200 150

150

100 Intensity (A.U.) Intensity 100

Intensity at 616(A.U.) nmat Intensity 50 50

0 0 0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 Time (s) Time (s)

Start of DMF vapor flow Start of DMF vapor flow 300 Start of water vapor flow Start of water vapor flow Stop of flow Stop of flow 300 250 250

200 200

150 150

100

100

Intensity at 616(A.U.) nmat Intensity 616(A.U.) nmat Intensity

50 50

0 0 0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 Time (s) Time (s)

Start of DMF vapor flow Start of water vapor flow Start of DMF vapor flow 300 250 Start of water vapor flow Stop of flow Stop of flow 250 200

200 150

150

100

100

Intensity at 616(A.U.) nmat Intensity Intensity at 616(A.U.) nmat Intensity 50 50

0 0 0 500 1000 1500 2000 2500 3000 3500 0 500 1000 1500 2000 2500 3000 Time (s) Time (s)

Figure 3.18. Luminescence intensity versus time plots of typical sensing experiments showing the reproducibility of the fast turn-on and -off.

91

260000 Before Acetone 240000 Before Benzene 240000 After Acetone 220000 After Benzene 220000 200000 200000 180000 180000 160000 160000 140000 140000 120000 120000 100000 100000 80000

80000

Intensity (A. U.) (A. Intensity Intensity (A. U.) (A. Intensity 60000 60000 40000 40000 20000 20000 0 0 -20000 -20000 540 560 580 600 620 640 660 680 700 720 740 540 560 580 600 620 640 660 680 700 720 740 Wavelength (nm) Wavelength (nm)

200000 Before DCM 280000 After DCM Before Et2O 180000 260000 After Et2O 240000 160000 220000 140000 200000

120000 180000 160000 100000 140000 80000 120000 100000 60000

Intensity (A. U.) (A. Intensity 80000 Intensity (A. U.) (A. Intensity 40000 60000 40000 20000 20000 0 0 -20000 -20000 540 560 580 600 620 640 660 680 700 720 740 540 560 580 600 620 640 660 680 700 720 740 Wavelength (nm) Wavelength (nm)

340000 320000 Before EtOH 280000 Before EtOAc 300000 After EtOH 260000 After EtOAc 280000 240000 260000 220000 240000 200000 220000 180000 200000 160000 180000 140000 160000 120000 140000 120000 100000 100000 80000 U.) (A. Intensity Intensity (A. U.) (A. Intensity 80000 60000 60000 40000 40000 20000 20000 0 0 -20000 -20000

540 560 580 600 620 640 660 680 700 720 740 540 560 580 600 620 640 660 680 700 720 740 Wavelength (nm) Wavelength (nm)

Figure 3.19a. Emission spectra of EusLOM-2 before and after exposure to various solvent vapors and drying reagent. (Cont.)

92

Before MeCN 220000 180000 Before Hexanes After MeCN 200000 After Hexanes 160000

180000 140000 160000 120000 140000 100000 120000

100000 80000

80000 60000 Intensity (A. U.) (A. Intensity Intensity (A. U.) (A. Intensity 60000 40000 40000 20000 20000 0 0

-20000 -20000 540 560 580 600 620 640 660 680 700 720 740 540 560 580 600 620 640 660 680 700 720 740 Wavelength (nm) Wavelength (nm)

240000 Before THF Before MeOH 220000 After THF 400000 After MeOH 200000 180000 160000 300000 140000 120000 200000 100000

80000 Intensity (A. U.) (A. Intensity

Intensity (A. U.) (A. Intensity 60000 100000 40000 20000 0 0 -20000 540 560 580 600 620 640 660 680 700 720 740 540 560 580 600 620 640 660 680 700 720 740 Wavelength (nm) Wavelength (nm)

260000 Before MgSO4 240000 140000 After MgSO4 220000 Before formamide 200000 120000 After formamide 180000 160000 100000 140000 120000 80000 100000

80000 60000 Intensity (A. U.) (A. Intensity 60000 Intensity (A.U.) Intensity 40000 40000 20000 20000 0 -20000 0 540 560 580 600 620 640 660 680 700 720 740 550 600 650 700 Wavelength (nm) Wavelength (nm)

Figure 3.19b. Emission spectra of EusLOM-2 before and after exposure to various solvent vapors and drying reagent. (Cont.)

93

References

1. (a) B. Chen, Y. Yang, F. Zapata, G. Lin, G. Qian, E. B. Lobkovsky, Adv. Mater. 2007, 19, 1693-1696; (b) Z.-Z. Lu, R. Zhang, Y.-Z. Li, Z.-J. Guo, H.-G. Zheng, J. Am. Chem. Soc. 2011, 133, 4172-4174; (c) S. Liu, Z. Xiang, Z. Hu, X. Zheng, D. Cao, J. Mater. Chem. 2011, 21, 6649-6653; d) Z. Guo, H. Xu, S. Su, J. Cai, . Dang, S. Xiang, G. Qian, H. Xhang, M. O’Keeffe, B. Chen, Chem. Commun. 2011, 47, 5551-5553. 2. Y. Takashima, V. M. Martínez, S. Furukawa, M. Kondo, S. Shimomura, H. Uehara, M. Nakahama, K. Sugimoto, S. Kitagawa, Nat. Commun. 2011, 2, 168. 3. (a) G. Long, M. E. Meek, J. Environ. Sci. Health., Part C Environ. Carcinog. Ecotoxicol. Rev. 2001, 19, 161-187; (b) M. Hamada, M. Abe, Y. Tokumoto, T. Miyake, H. Murakami, Y. Hiasa, B. Matsuura, K. Sato, M. Onji, Intern. Med. (Tokyo, Jpn.) 2009, 48, 1647-1650. 4. (a) H. Ohbayashi, K. Yamazaki, S. Aiso, K. Nagano, S. Fukushima, H. Ohta, J. Toxicol. Sci. 2008, 33, 327-338; (b) R. Ding, D. Chen, Y. Yang, Environ. Toxicol. Pharmacol. 2011, 31, 357-363. 5. (a) A. M. Saillenfait, J. P. Payan, D. Beydon, J. P. Fabry, I. Langonne, J. P. Sabate, F. Gallissot, Fundam. Appl. Toxicol. 1997, 39, 33-43; (b) J. J. Hellwig, J. Merkle, H. J. Klimisch, R. Jäckh, Food Chem. Toxicol. 1991, 29, 193-201. 6. (a) A. M. Ducatman, D. E. Conwill, J. Crawl, J. Urol. 1986, 136, 834-836. (b) S. M. Levin, D. B. Baker, P. J. Landrigan, S. V. Monaghan, E. Frumin, M. Braithwaite, W. Towne, Lancet, 1987, 8568, 1153-1153. (c) H. Senoh, S. Aiso, H. Arito, T. Nishizawa, K. Nagano, S. Yamamoto, T. Matsushima, J. Occup. Health, 2004, 46, 429-439. 7. (a) V. Rimatori, G. Carelli, Scand. J. Work. Environ. Health 1982, 8, 20-23; (b) E. W. March, L. S. Ettre, Chromatography Newsletter, 1977, 5, 7-8; (c) http://www.osha.gov/dts/sltc/methods/organic/org066/org066.html 8. Goto, Y.; Sato, H.; Shinkai, S.; Sada, K. J. Am. Chem. Soc. 2008, 130, 14354-14355 9. Apex 2 Software Package; Bruker AXS inc., 2008. 10. Y. Haas, G. Stein, J. Phys. Chem. 1971, 75, 3677-3681. 11. W. D. Horrocks Jr., D. R. Sudnick, J. Am. Chem. Soc. 1979, 101, 334-340. 12. G. Férey, C. Serre, Chem. Soc. Rev. 2009, 38, 1380-1399.

94

Chapter 4 Properties of Lanthanide Metal-Organic Frameworks of Functionalized [1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid 4 4.1 Abstract

In this chapter, a series of pseudo iso-reticular (similar rather than identical topology) metal- organic frameworks (MOFs) have been assembled with functionalized [1,1':4',1''-terphenyl]-4,4''- dicarboxylic acid and the relationship of the properties between the ligands and corresponding MOFs were studied, including porosity, thermal stability and luminescence properties. MOFs constructed from ligands bearing aldehyde, ketone and thioether groups were found to display solvent dependent luminescence properties, as a result of solvent-metal and solvent-ligand interaction.

4.2 Introduction

H2LM: R = -CH3 R H2LOM: R = -CH2OCH3 O OH

H2LCHO: R = -CHO HO O

R H2LCO: R = -CH=CHCOCH 3

H2LDS: R = -SCH3

Figure 4.1. Ligand series with different functional groups discussed in this chapter.

In our previous MOF sensor EusLOM, solvent-ligand interaction was found to be the major contributor for the selective and strong turn-on effect. Our success in this MOF has driven us to fully explore this ligand system (Figure 4.1), studying the ligand – structure relationship, impact of other functional groups on MOF properties, and strategies to enhance the analyte-ligand

interaction as well as the triggered luminescence modulation. Although EusLOM showed very sensitive response to DMF vapor, it did not respond to any metal cation in solution. Such lack of response may be attributed to the fact that the introduced functional group – ether group – does not interact with metal ions strongly, or perhaps the interaction is not able to cause significant

95 change on the ligand triplet state energy level to trigger meaningful luminescence change. In this chapter we will discuss several lanthanide MOFs constructed from ligands with different functional groups, namely methyl group, ether group, aldehyde group, alkene extended ketone chain and thioether group, compare them with each other as well as EusLOM/TbsLOM systems, to understand the impact of different functional groups on the MOF structures, stability, and 2- luminescence properties. sLM (2',5'-dimethyl-[1,1':4',1''-terphenyl]-4,4''-dicarboxylate) was 2- chosen as the simplest in the series to study its structural implication. sLCHO (2',5'-diformyl- 2- [1,1':4',1''-terphenyl]-4,4''-dicarboxylate) and sLCO (2',5'-bis((E)-3-oxobut-1-en-1-yl)-[1,1':4',1''- terphenyl]-4,4''-dicarboxylate) represent one type of our strategies to improve the sensitivity of MOFs – by introducing a stronger donor oxygen (as comparing to ether oxygen) which is 2- conjugated with the ligand π system. sLDS (2',5'-bis(methylthio)-[1,1':4',1''-terphenyl]-4,4''- dicarboxylate) follows a similar strategy, where the lone pair on the thioether atom is conjugated with ligand backbone π system. Aldehyde/ketone oxygen may interact with metal ions more strongly compared to ether oxygen, and the thioether sulfur may preferably interact with heavy metal ions (e.g. Hg2+ and Cd2+). The π-conjugation of these donor atoms may transmit electron density changes triggered by analyte-ligand interaction more readily to the ligand backbone to affect its energy levels.

4.3 Experimental section

4.3.1 Materials and methods

Elemental analyses were performed in our Chemistry Department on a PE 2400 C/H/N/S analyzer. Thermogravimetric analyses (TGA) were performed on a TA Instruments SDT Q600 instrument under dinitrogen atmosphere with a heating rate of 10 °C per minute. NMR spectra were recorded on a Varian 400, or a Bruker Avance 400 spectrometer. Both 1H and 13C NMR spectra were referenced and reported relative to the solvent residual signals. The powder XRD experiments were performed on either an automated Siemens/Bruker D5000 diffractometer equipped with a Cu sealed tube (λ = 1.54178 Å) or Bruker AXS Benchtop D2 Phaser diffractomerer operating with line-focus Cu-Kα1 (λ = 1.54059 Å). Unless otherwise stated, all manipulations were performed in air and all reagents were purchased from commercial sources and used without further purification. 2',5'-dimethyl-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid [1] [2] (H2sLM), 2',5'-bis(methoxymethyl)-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid (H2sLOM),

96

(triphenylphosphoranylidene)propan-2-one,[3] and 2,5-dibromo-1,4-bis(methylthio)benzene[4] were prepared through literature methods.

4.3.2 Synthesis of 2,5-dibromoterephthalaldehyde

[5] The following is a modified procedure based on a literature method. Concentrated H2SO4 (21 mL) was added slowly to a mixture of acetic anhydride (60 mL) and acetic acid (30 mL) at 0 °C, and grinded powder of 2,5-dibromoxylene (6.0 g, 22.7 mmol) was added. CrO3 (9.0 g, 90.0 mmol) was then added in proportions in 5 h and the mixture was vigorously stirred at 0 °C during the addition. The resulting mixture was poured into ice-water mixture and the (2,5- dibromo-1,4-phenylene)bis(methanetriyl) tetraacetate was collected by filtration. The crude tetraacetate was then suspended in a mixture of H2O (60 mL), ethanol (60 mL) and concentrated

H2SO4 (6 mL) and refluxed for 5 h. 120 mL of water was then added to the resulting mixture and 1 the product was collected by filtration as a light yellow solid. H NMR (CDCl3, 400 MHz, 25°C)  10.34 (s, 2H), 8.16 (s, 2H).

4.3.3 Synthesis of dimethyl 2',5'-diformyl-[1,1':4',1''-terphenyl]-4,4''- dicarboxylate (sLCHO ester)

4-methoxycarbonylphenyl boronic acid (6.17 g, 34.3 mmol), 2,5-dibromoterephthalaldehyde

(4.00 g, 13.7 mmol), Na2CO3 (18.2 g, 0.172 mol) and palladium acetate (0.385 g, 1.72 mmol, 5 mol%) was suspended in degassed N,N’-dimethylformamide water mixture (80 mL, VDMF:VH2O

= 2:1), and stirred for 24 h under N2 atmosphere. The reaction mixture was then acidified with 1

M HCl and then petitioned between H2O and CHCl3 and the aqueous phase was extracted several times with CHCl3. The organic phase was combined, washed with H2O and brine for several times and then dried over MgSO4. Pd black and MgSO4 was then removed via filtration through silica gel and a white solid was obtained after concentrating the organic phase, which was then washed with ethyl acetate (50 mL) to give the desired product sLCHO ester (1.16 g, 88% yield). 1 H NMR (CDCl3, 400 MHz, 25°C)  10.07 (s, 2H), 8.20 (d, J = 8.4 Hz, 4H), 8.13 (s, 2H), 7.53 (d,

J = 8.4 Hz, 4H), 3.99 (s, 6H). 13C NMR (CDCl3, 100 MHz, 25°C)  190.82, 166.42, 143.92,

140.76, 136.50, 130.57, 130.38, 130.04, 130.00, 52.41. Anal.calcd for C24H18O6: C, 71.64; H, 4.51. Found: C, 71.16; H, 4.48.

97

4.3.4 Synthesis of 2',5'-diformyl-[1,1':4',1''-terphenyl]-4,4''-dicarboxylic acid (H2sLCHO) sLCHO ester (2.0 g, 5.0 mmol), and NaOH (3.2 g, 80.0 mmol) was suspended in a mixture of methanol (80 mL) and H2O (80 mL) and the mixture was degassed by bubbling N2 through for

0.5 h. The suspension was then refluxed for 3 h under N2 atmosphere. After cooled to room temperature, the resulting solution was acidified with 1 M aqueous HCl to pH = 1 and the white 1 precipitate of H2sLCHO was collected by filtration (1.82 g, 98% yield). H NMR (DMSO-d6, 400 MHz, 25°C)  13.15 (s, 2H), 10.02 (s, 2H), 8.10 (d, J = 8.4 Hz, 4H), 8.02 (s, 2H), 7.69 (d, J = 8.4 13 Hz, 4H), 7.20 (s, 2H). C NMR (DMSO-d6, 100 MHz, 25°C)  182.41, 166.85, 142.64, 140.56,

136.08, 130.67, 130.59, 130.06, 129.45. Anal.calcd for C22H14O6∙ 0.4 DMF: C, 69.04; H, 4.20; N, 1.39. Found: C, 68.85; H, 4.46; N, 1.30.

4.3.5 Synthesis of (3E,3'E)-4,4'-(2,5-dibromo-1,4-phenylene)bis(but-3- en-2-one)

2,5-dibromo-1,4-benzenedicarboxaldehyde (4.0 g, 13.7 mmol) and (triphenylphosphoranylidene)propan-2-one (10.9 g, 34.2 mmol, 2.5 equiv.) were dissolved in 130 mL chloroform and stirred at R.T. for 4 h. Solvent was then removed in vacuo and the resulting solid was washed with methanol (50 mL) to yield the desired product as a light yellow solid 1 (3.48 g, 68 % yield). H NMR (CDCl3, 400 MHz, 25°C)  7.86 (s, 2H), 7.76 (d, J = 16.0 Hz, 2H), 13 6.66 (d, J = 16.4 Hz, 2H), 2.43 (s, 6H). C NMR (CDCl3, 100 MHz, 25°C)  197.61, 139.45,

137.42, 132.11, 131.11, 124.57, 27.90. Anal.calcd for C14H12Br2O2: C, 45.20; H, 3.25. Found: C, 45.01; H, 3.29.

4.3.6 Synthesis of dimethyl 2',5'-bis((E)-3-oxobut-1-en-1-yl)-[1,1':4',1''- terphenyl]-4,4''-dicarboxylate (sLCO ester)

(3E,3'E)-4,4'-(2,5-dibromo-1,4-phenylene)bis(but-3-en-2-one) (1.00 g, 2.7 mmol), 4- methoxycarbonylphenyl boronic acid (1.22 g, 6.8 mmol), Na2CO3 (2.91 g, 27.5 mmol), palladium acetate (61 mg, 0.27 mmol) were suspended in degassed DMF/H2O mixture (40 mL,

VDMF:VH2O = 2:1) and then heated under N2 atmosphere at 80 °C for 20 h. The resulting mixture was neutralized with 1 M aqueous HCl, extracted with chloroform and then washed with water and brine. The organic phase was then dried over MgSO4 and concentrated to give the crude product. The crude product was then washed with 40 mL of ethylacetate to yield the desired

98

1 product sLCO ester as a light yellow solid (0.99 g, 76% yield). H NMR (CDCl3, 400 MHz, 25°C)  8.17 (d, J = 8.4 Hz, 4H), 7.72 (s, 2H), 7.48 (d, J = 16.4 Hz, 2H), 7.46 (d, J = 8.4 Hz, 4H), 6.72 13 (d, J = 16.4 Hz, 2H), 3.98 (s, 6H), 2.24 (s, 6H). C NMR (CDCl3, 100 MHz, 25°C)  197.77, 166.60, 143.35, 141.74, 140.40, 134.27, 129.96, 129.84, 129.71, 129.67, 128.83, 52.32, 27.52.

Anal.calcd for C30H26O6 ∙ 0.1 CHCl3: C, 73.11; H, 5.32. Found: C, 73.46; H, 5.81.

4.3.7 Synthesis of 2',5'-bis((E)-3-oxobut-1-en-1-yl)-[1,1':4',1''-terphenyl]- 4,4''-dicarboxylic acid (H2sLCO)

(triphenylphosphoranylidene)propan-2-one (3.53 g, 11.1 mmol) and H2sLCHO (1.00 g, 2.7 mmol) were dissolved in DMF (100 mL) and the mixture was heated at 80 °C for 2 days. The resulting mixture was acidified with 1 M HCl to pH = 1 and 200 mL of water was added. The precipitate formed was collected via filtration and washed with methanol to yield the desired product 1 H2sLCO acid as light yellow solid (0.95 g, 78% yield). H NMR (DMSO-d6, 400 MHz, 25°C)  13.12 (s, 2H), 8.09 (d, J = 8.0 Hz, 4H), 7.97 (s, 2H), 7.59 (d, J = 8.4 Hz, 4H), 7.40 (d, J = 16.0 13 Hz, 2H), 7.04 (d, J = 16.4 Hz, 2H), 2.21 (s, 6H). C NMR (DMSO-d6, 100 MHz, 25°C)  197.46, 166.92, 142.75, 141.21, 138.81, 133.58, 130.17, 129.96 129.31, 128.78, 40.34, 28.21. Anal.calcd for C28H22O6 ∙ 1.55 DMSO: C, 64.90; H, 5.48. Found: C, 64.99; H, 5.67.

4.3.8 Synthesis of dimethyl 2',5'-bis(methylthio)-[1,1':4',1''-terphenyl]- 4,4''-dicarboxylate (sLDS ester)

4-methoxylcarbonylphenylboronic acid (2.75 g, 15.3 mmol, 2.5 equiv.), 2,5-dibromo-1,4- bis(methylthio)benzene (2.00 g, 6.1 mmol, 1 equiv.), palladium tetrakis triphenylphosphine (0.35 g, 0.31 mmol, 0.05 equiv.) and cesium carbonate (11.92 g, 36.6 mmol, 6 equiv.) were suspended in the degassed mixture of DMF (60 mL) and THF (60 mL), and then heated to 90 °C with stirring for 24 h under a N2 atmosphere. After cooled to room temperature, the precipicate was removed by filtration and the resulting mixture was acidified and then petitioned between chloroform and water. The aqueous phase was extracted with chloroform and all organic phase was combined, washed with H2O and brine, and then dried over MgSO4. After filtered off

MgSO4, all solvents were removed to give the crude product, which was then washed with 1 diethylether to give 2.26 g product of sLDS ester (85% yield). H NMR (CDCl3, 400 MHz, 25 °C )  8.13 (d, J = 8.8 Hz, 4H), 7.55 (d, J = 8.4 Hz, 4H), 7.15 (s, 2H), 3.96 (s, 6H), 2.34 (s, 6H). 13C

NMR (DMSO-d6, 100 MHz, 25 °C )  167.19, 144.87, 140.68, 134.12, 129.91, 129.88, 129.74,

99

127.96, 52.56, 16.80. Anal.calcd for C24H22O4S2 ∙ 0.1CH2Cl2: C, 64.75; H, 5.01. Found: C, 65.11; H, 5.06.

4.3.9 Synthesis of 2',5'-bis(methylthio)-[1,1':4',1''-terphenyl]-4,4''- dicarboxylic acid (H2sLDS) sLDS ester (1.0 g, mmol), sodium hydroxide (1.6 g, mmol) were suspended in methanol (40mL) and refluxed for 24 h. Methanol was then removed in vacuo and the remaining solid was dissolved in hot water, filtered through Celite, and then acidified with 1M HCl to pH <1. The precipitate was then collected via filtration and dried at 60 °C to give the desired product H2sLDS 1 (0.83 g, 88% yield). H NMR (DMSO-d6, 400 MHz, 25°C )  13.02 (s, 2H), 8.03 (d, J = 8.4 Hz, 13 4H), 7.60 (d, J = 8.8 Hz, 4H), 7.18 (s, 2H), 2.54 (s, 4H), 2.38 (s, 6H). C NMR (DMSO-d6, 100 MHz, 25°C )  167.04, 143.75, 139.38, 133.17, 130.06, 129.41, 129.18, 126.81, 15.34. Anal.calcd for C22H18O4S2∙ 0.6 DMSO: C, 60.85; H, 4.73. Found: C, 60.48; H, 4.90.

4.3.10 Synthesis of [Eu2(sLM)3(H2O)(DMF)2] (EusLM)

Eu(NO3)3∙6H2O (17.8 mg, 0.04 mmol), H2sLM (20.1 mg, 0.06 mmol) was dissolved in a mixture of DMF (5 mL) and H2O (5 mL) in a 22 mL PTFE liner in an acid digestion bomb and sealed.

The bomb was then heated at 140 °C for 3d. Colorless crystals of EusLM were collected via filtration (15.6 mg, 50% yield, assuming MW = 1555.74 according to EA/1H NMR). Anal.calcd for C72H64Eu2N2O15 ∙ 0.5C3H7NO ∙ H2O : C, 56.74; H, 4.50; N, 2.25. Found: C, 56.36; H, 4.63; N, 2.27. (TGA solvent weight loss Anal.calcd. 14.1%, Found 13.4%)

4.3.11 Synthesis of [Tb2(sLM)3(H2O)2(DMF)2] (TbsLM)

Tb(NO3)3∙6H2O (18.1 mg, 0.04 mmol), H2sLM (20.1 mg, 0.06 mmol) was dissolved in a mixture of DMF (4 mL) and H2O (6 mL) in a 22 mL PTFE liner in an acid digestion bomb and sealed.

The vial was then heated in oil bath at 150 °C for 3d. Colorless crystals of TbsLM were collected via filtration (18.8 mg, 59% yield, assuming MW = 1602.20 according to EA/1H NMR).

Anal.calcd for C72H66N2O16Tb2 ∙ C3H7NO: C, 56.08; H, 4.58; N, 2.62. Found: C, 55.74; H, 4.62; N, 2.55. (TGA solvent weight loss Anal.calcd. 15.9%, Found 14.6%)

100

4.3.12 Synthesis of [Tb2(sLOM)3(H2O)4] (TbsLOM)

Tb(NO3)3∙6H2O (18.1 mg, 0.04 mmol), H2sLOM (24.4 mg, 0.06 mmol) was dissolved in a mixture of DMF (8 mL) and H2O (2 mL) in a 20 mL scintillation vial and sealed. The vial was then heated at 80 °C for 3d. Colorless crystals of TbsLOM were collected via filtration (28.1 mg, 76% 1 yield, assuming MW = 1853.63 according to EA/ H NMR). Anal.calcd for C78H68O22Tb2 ∙

3.4C3H7NO: C, 53.32; H, 5.00; N, 2.57. Found: C, 53.11; H, 5.07; N, 3.03. (TGA solvent weight loss Anal.calcd. 17.3%, Found 16.6%)

4.3.13 Synthesis of [Eu2(sLCHO)3 (H2O)4] (EusLCHO)

Eu(NO3)3∙6H2O (17.8 mg, 0.04 mmol), H2sLCHO (22.5 mg, 0.06 mmol) was suspended in a mixture of DMF (3 mL), acetone (3 mL) and H2O (4 mL) in a 20 mL scintillation vial and sealed.

The vial was then heated in oil bath at 100 °C for 3 d. Yellow crystals of EusLCHO were collected via filtration (25.8 mg, 76% yield, assuming MW = 1689.57 according to EA/1H NMR).

Anal.calcd for C66H44Eu2O22 ∙ 2C3H7NO ∙ 4.8H2O: C, 51.18; H, 3.79; N, 1.66. Found: C, 51.64; H, 4.18; N, 1.56 (TGA solvent weight loss Anal.calcd. 15.9%, Found 14.9%)

4.3.14 Synthesis of [Eu2(sLCO)3(H2O)4(DMF) (DMSO)]. (EusLCO)

Eu(NO3)3∙6H2O (13.4 mg, 0.03 mmol), H2sLCO (13.6 mg, 0.03 mmol) was dissolved in DMSO (1.75 mL) in a 7 mL scintillation vial, and a mixture of DMF (1.25 mL), THF (0.75 mL) and

H2O (1.25 mL) was added. The vial was sealed and then heated in oil bath at 70 °C for 7 d.

Yellow crystals of EusLCO were collected via filtration (8.8 mg, 41% yield, assuming MW = 1 2137.57 according to EA/ H NMR). Anal.calcd for C89H81NO24SEu2 ∙ 0.7C3H7NO ∙ 6H2O ∙

1.2C2H6OS: C, 52.53; H, 4.96; N, 1.11. Found: C, 52.50; H, 4.96; N, 1.09 (TGA solvent weight loss Anal.calcd. 22.9%, Found 22.6%)

4.3.15 Synthesis of [Eu2(sLDS)3(H2O)4(DMF)2] (EusLDS)

Eu(NO3)3∙6H2O (8.9 mg, 0.02 mmol), sLDS acid (12.8 mg, 0.03 mmol) was suspended in a mixture of DMF (2.5 mL), DMSO (1.0 mL) and H2O (1.5 mL) in a 20 mL scintillation vial and sealed. The vial was then heated in oil bath at 80 °C for 3 d. Colorless crystals of EusLDS were collected via filtration and washed with ethanol (15.3 mg, 79% yield, assuming MW = 1925.88 1 according to EA/TGA/ H NMR). Anal.calcd for C69H59Eu2NO15S6 ∙ 0.7C2H6OS ∙ 1.3C3H7O ∙

101

4.8H2O: C, 47.64; H, 4.54; N, 2.35. Found: C, 47.95; H, 4.83; N, 2.02 (TGA solvent weight loss Anal. 20.6%, Found 20.7%)

4.3.16 Synthesis of [Tb2(sLDS)3(H2O)4(DMF)2] (TbsLDS)

TbsLDS was synthesized with the same method except that Tb(NO3)3∙6H2O (9.1 mg, 0.02 mmol) was used. Colorless crystals of TbsLDS were collected via filtration and washed with ethanol (15.0 mg, 79% yield, assuming MW = 1903.46 according to EA/TGA/1H NMR). Anal.calcd for

C69H59NO15S6Tb2 ∙ 0.7C2H6OS ∙ 1.7C3H7NO ∙ 4H2O: C, 47.64; H, 4.40; N, 1.99. Found: C, 47.37; H, 4.16; N, 1.95 (TGA solvent weight loss Anal. 18.3%, Found 18.7%).

4.3.17 Preparation of water-soaked EusLDS sample for sensing experiments

As-synthesized EusLDS sample was soaked in water (~0.5 mL water per mg of EusLDS) for 4 days. The resulting solid was collected via filtration, washed with ethanol, dried in air, and then grinded to fine powders for sensing experiments.

4.3.18 X-ray crystallographic analysis.

X-ray quality single crystals were obtained as described in the sections above. All crystals were mounted on the tip of a MiTeGen MicroMount. The single-crystal X-ray diffraction data were collected on a Bruker Kappa Apex II CCD diffractometer with Mo Kα radiation (λ = 0.71073 Å) operating at 50 kV and 30 mA, at 150 K controlled by an Oxford Cryostream 700 series low temperature system. The data integration and absorption correction were performed with the Bruker Apex 2 software package. All the structures were solved by direct methods and refined using SHELXTL V6.14. All non-hydrogen atoms except for the atoms involved in the disordered portions were refined anisotropically. The positions of the hydrogen atoms were calculated using the riding model.

4.3.19 Determination of solvent composition of different MOF samples

The solvent compositions of MOFs are determined by a combination of different characterizations, usually including x-ray crystallography, elemental analysis (EA) and thermogravimetric analysis (TGA), all performed on the same batch of sample. For those MOFs that were synthesized from two-component solvent systems and whose coordinated/free solvents

102

can be readily located in their crystal structures, namely EusLM, TbsLM and TbsLOM, the aforementioned characterization are considered to be sufficient to determine the actual solvent compositions of the as-synthesized sample. For the MOFs synthesized from systems containing more than two solvents, or whose coordinated/free solvents are disordered in the crystal structures, 1H NMR experiments were also performed to assist solvent composition determination. The use of different deuterated solvent (methanol-d4 or acetone-d6) produces consistent results in most cases, except for EuLCO, where the amount of solvent calculated from using acetone-d6 is much smaller than that from using methanol-d4. Such differences are likely resulted from the difference in porosity of the MOFs, or the ability of the solvents to exchange the coordinated/channel solvents. Whenever the inconsistency occurs, we consider the results from the methanol system to be more accurate because of the smaller size and arguably stronger coordinating ability. EusLCHO solvent content is determined according to only EA, TGA and X- ray crystallography as the channel solvents cannot be completely exchanged with either acetone- d6 or methanol-d4.

4.3.20 1H NMR experiments procedures for MOF samples

In a typical 1H NMR experiment, 6~10 mg of MOF sample is weighted and suspended in deuterated solvent (0.7 mL of methanol-d4 or acetone-d6), and mesitylene (2.5 μL) is added as internal standard. A control group is set up under the same condition except that MOF sample is not added. The NMR sample is then sonicated for a few minutes and the NMR spectrum is taken with 25 s relaxation delay to ensure the accuracy of integration. The amount of water and organic solvent were determined based on the ratios of their integration to that of mesitylene; water content is calculated from H2O and HDO (when acetone-d6 is used) signals, further adjusted by subtracting the amount of water in the control group to exclude the water introduced from the deuterated solvent, moisture or mesitylene. The exchange between water and methanol-d4 was found to have no significant impact in estimating the water content according to our control experiment.

4.3.21 Luminescence measurements

Photoluminescence spectra were measured using QuantaMaster 30 system from Photon Technology International (Canada) Inc. equipped with a 75W high intensity Xenon lamp (for fluorescence), a Xenon short arc lamp (for phosphorescence), a R928 PMT detector, and a solid

103 sample holder; the incident and detector slit widths are 1 and 0.4 nm for fluorescence measurements, 8 and 8 nm for phosphorescence, respectively. Solvent dependent luminescence spectra of different MOFs were measured in suspensions in a standard quartz cuvette where 3.0 mg of the grinded MOF powder is suspended in 3.0 mL of different solvents. The suspension was stirred throughout the measurement and the emission intensity was found to be stable except for EusLCHO. The luminescent emission intensities of different batches of the same samples are found to be consistent. The measurements used to determine ligand singlet/triplet state energy are measured with approximately 10-4 M degassed ligand solution in 1:1 toluene and ethanol in regular J. Young tubes. Low temperature values are obtained from the aforementioned solution frozen by liquid N2. Phosphorescence is measured 100 μs after light pulse (or 200 μs delay time on the instrument). Luminescence lifetimes are calculated using flex32 or flex software based on the phosphorescence decay data.

4.3.22 Vapor sensing experiments of EusLDS.

Powder samples of water-soaked EusLDS were used for vapour sensing experiments. For each experiment, ~25 mg of water-soaked EusLDS was placed into a quartz cuvette and the luminescence spectrum was measured in a solid sample holder before exposure to various solvent vapors. The powder and the cuvette were then placed into a sealed container (usually ~100 mL bottle) which contains ~5 mL of a given solvent (See Figure 4.2) for 24 h. Subsequently the cuvette was taken out of the container, quickly sealed and the emission spectrum was taken again in the solid sample holder. In order to reproduce the intensity, the front face of the cuvette exposed to the incidental beam has to be covered by the powder sample completely.

The Eu emission intensity at 616 nm was used for comparison among all the samples after exposure to solvent vapors. For Figure 4.27 in the main text, luminescence enhancement for each solvent was measured three times with three different batches of samples. The average of

Iafter/Ibefore-1 (Iafter, Ibefore stand for emission intensity at 615 nm after and before exposure to solvent vapor, respectively) was used as the height of the column and the standard deviations were drawn as the error bar.

104

Figure 4.2. Diagram of the experimental setup for incubating water-soaked EusLDS under different solvent vapors. 4.4 Results and Discussion

4.4.1 Ligand Synthesis, MOF Preparation and Structures

EusLM and TbsLM can be synthesized via solvothermal method from Eu(NO3)3/Tb(NO3)3 and

H2sLM in moderate yield. EusLM crystallized in the monoclinic space group C2/c. As shown in Figure 4.3, each Eu center is eight-coordinate, adopting square anti-prismatic geometry with six oxygen donor from carboxylic group of sLM ligand, one oxygen donor atom from terminal DMF molecule and one oxygen donor atom from bridging water molecules. The two square planes are defined by O3, O4, O7, O8 and O1, O2, O5, O6 respectively. All Eu-O bond length are within the range of 2.342(5)-2.438(6) Å , except for Eu-O7 bond, which is 2.709(4) Å. Such long Eu-O bond length is likely due to the μ2 bridging mode of the coordinated water rather than hydroxide.[8] Along the c-axis, Eu ions are bridged by carboxylate groups and water molecules to form a 1-D chains (Figure 4.3b), which are interconnected to four adjacent parallel chains to form the 3-D framework. As a result, there are 1-D parallelogram-shaped channels along the c- axis (Figure 4.3c), which are occupied by lattice solvent molecules. Compared to the MOF constructed from a similar ligand (H2sLOM), the two structures are analogous in terms of the 1D metal chains and the connectivity between chains, but the α angle is much smaller, resulting in much flatter channels. Because coordinated DMF molecules occupy even more space, there is essentially no space left for free solvent – only approximately one lattice water molecule was found per Eu2(sLM)3 unit according to elemental analysis (EA) and thermogravimetric analysis (TGA)

105

Figure 4.3. Crystal structure of EusLM: (a) eight-coordinate Eu center; (b) 1D chain with bridging carboxylate and coordinated DMF; (c) perspective view of 1D chains and channels along c-axis, α ≈ 35°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity.

106

Figure 4.4. Crystal structure of TbsLM: (a) seven-coordinate Tb1 center; (b) eight-coordinate Tb2 center; (c) 1D chain with alternating Tb1 and Tb2 centers bridged by carboxylate; (d) perspective view of 1D chains and channels, α ≈ 30°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity.

TbsLM crystallizes in the monoclinic space group P2(1)/n. Two types of Tb center are present in the framework (Figure 4.4a and b). Tb1 center is seven-coordinate, adopting approximately pentagonal bipyramidal geometry with six oxygen donors from carboxylate and one from a

107 coordinated water molecules. Tb2 center is eight-coordinate, adopting bi-capped trigonal prismatic geometry with five oxygen atoms from carboxylates, two DMF oxygen atoms and one water oxygen atom occupying the eight coordination sites. All Tb-O bond length are within the range of 2.240(5) to 2.468(4) Å. One of the carboxylate groups coordinates through only one oxygen atom, leaving the remaining oxygen atom hydrogen bonded with the an adjacent coordinated water, while all the other carboxylates coordinate in bidentate fashion through both oxygen atoms. The two types of Tb centers are alternating along the 1D chain bridged by carboxylate groups (Figure 4.4c). Each chain is interconnected to two four other chains to yield a 3D framework (Figure 4.4d). The α angle is fairly small, leading to small solvent channels with only one lattice DMF per Tb2(sLM)3 unit.

Figure 4.5. Crystal structure of TbsLOM: (a) eight-coordinate Tb center; (b) 1D chain with bridging carboxylates; (c) perspective view of 1D chains and channels, α ≈ 40°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity.

[2] TbsLOM can be synthesized in a similar fashion with its Eu counterpart. It crystallized in the triclinic space group P-1. Each eight-coordinate Tb center adopts a distorted square anti-

108 prismatic geometry with O1, O5, O2, O6 and O3, O7, O4, O8 defining the two square planes (Figure 4.5a). Among the eight coordinated oxygen atoms, six are from carboxylate groups and the other two are from coordinated water molecules. 1D Tb chains are found along the a-axis with four or two carboxylate groups bridging the adjacent Tb centers in an alternating manner (Figure 4.5b).These 1D chains are interconnected to four adjacent chains through the ligand to form a 3D framework (Figure 4.5c) with solvent channels parallel with the 1D metal chains. Lattice DMF molecules are located in the solvent channels according to the crystal structure, which was also confirmed also by EA and TGA. This structure is almost identical with the previously reported EuLOM, as one would expect due to the similarity between Eu and Tb.

Scheme 4.1. Synthesis of H2sLCHO and H2sLCO.

As shown in Scheme 4.1, the aldehyde functionalized dibromide was synthesized according to a modified literature procedure, and then coupled with 4-methoxycarbonylphenyl boronic acid to yield dimethyl 2',5'-diformyl-[1,1':4',1''-terphenyl]-4,4''-dicarboxylate (sLCHO ester). The following hydrolysis was completed under N2 atmosphere as H2sLCHO is not stable and readily oxidized in air under basic condition to yield a yellow precipitate. Although dimethyl 2',5'-

109

bis((E)-3-oxobut-1-en-1-yl)-[1,1':4',1''-terphenyl]-4,4''-dicarboxylate (sLCO ester) can be readily synthesized with ease from the Suzuki coupling reaction between the respective boronic acid and functionalized dibromide (Scheme 4.2), the subsequent hydrolysis cannot be performed with reasonable yield due to the instability of the side chain under basic condition. Alternatively,

H2sLCO was synthesized from the reaction between H2sLCHO and the Wittig reagent, (triphenylphosphoranylidene)propan-2-one in DMF.

Scheme 4.2. Synthesis of sLCO ester via Suzuki coupling reaction.

H2sLCHO can be assembled into a 3D MOF EusLCHO with Eu(NO3)3 under solvothermal condition in the mixture of DMF, acetone and water. Although no acetone is found in the crystal structure of the resulting MOF, the use of acetone is necessary to obtain the crystals. We have screened various DMF/H2O and DMF/H2O/solvent systems and only found crystals when acetone is used. EuLCHO crystallized in the monoclinic space group P2(1)/m. Two types of Eu centers are present in the crystal structure as shown in Figure 4.6. Eu1 adopts tricapped trigonal prismatic geometry, with nine coordination sites occupied by six carboxylate oxygen and three water oxygen atoms, while Eu2 adpots a mono-capped trigonal prismatic geometry, with six carboxylate oxygen and one water oxygen atoms occupying the seven coordination sites. All Eu- O bond are within the range of 2.359(3) and 2.449(5), except for Eu(1)-O(2) bond length of 2.601(3), due to the μ2 bridging coordination mode of the oxygen atom. Unlike the coordination mode of most carboxylate, two of the carboxylate groups on Eu2 coordinate through only one of the two oxygen atoms, leaving the other oxygen atom hydrogen bonded with the coordinated water on the neighbouring Eu1 center. Eu1 and Eu2 centers are bridged by carboxylate groups in alternating order to form 1D metal chains, which are then interconnected by sLCHO ligands with four adjacent chains, yielding a 3D framework with small solvent channels parallel to the metal chains. All coordinated water and uncoordinated carboxylate oxygen atoms are exposed on the channel wall.

110

Figure 4.6. Crystal structure of EusLCHO: (a) nine-coordinate Eu1 center; (b) seven-coordinate Eu2 center; (c) 1D chain with alternating Eu1 and Eu2 centers bridged by carboxylates; (d) perspective view of 1D chains and channels, α ≈ 35°. All hydrogen atoms and coordinated solvent molecules are omitted for clarity.

111

Figure 4.7. Crystal structure of EuLCO: (a) eight-coordinate Eu center; (b) coordination environment of Eu center; (c) 1D chain with bridging by carboxylate; (d) layers of 2D sheets; (e) perspective view of 1D chains and channels. All hydrogen atoms and non-coordinate solvent molecules are omitted for clarity.

EusLCO was synthesized by reacting H2sLCO with Eu(NO3)3 in a four-solvent system, and crystallized in the triclinic space group P-1.Unlike the other MOFs which adopted similar topology featuring the interconnected 1D metal chains, EusLCO is a 2D MOF. Figure 4.7a,b show the square anti-prismatic geometry adopted by the eight-coordinate Eu center, with O3, O5, O2, O11 and O6, O10, O12, O1 defining the two planes respectively. The eight coordination sites are occupied by five oxygen atoms from the carboxylate groups and three oxygen atoms from coordinated solvents. All Eu-O bond length are within the range of 2.290(5) to 2.456(5) Å except for Eu(1)-O(12) bond length of 2.550(9) Å, suggesting a weakly coordinated DMF molecule. Among all five carboxylate groups that coordinate on the same Eu center, four of them are bridging between two neighbouring metal centers; the remaining one only coordinated through one oxygen atom, leaving the other oxygen atom hydrogen bonded with the neighbouring coordinated water on the same Eu center. Disordered solvents were found on the Eu centers, with DMF and DMSO occupying two of the sites with only 50% occupancy, likely with water occupying the remaining sites. The Eu centers are bridged by carboxylate groups to form 1D metal chains. Each chain is interconnected to two adjacent chains, extending the structure to 2D sheets. Because of the long ketone side chains is oriented perpendicular to the 2D

112 sheets, the 2D sheets stack with considerable space between layers (approximately 5 to 8 Å spacing along b-axis), yielding 1D solvent channels parallel with the 1D metal chains. Coordinated solvent and uncoordinated oxygen atoms from the carboxylate groups are exposed in the channel.

Scheme 4.3. Synthesis of H2sLDS.

2,5-dibromo-1,4-bis(methylthio)benzene, synthesized according to a literature method, can be coupled with 4-methoxylcarbonylphenylboronic acid to yield sLDS ester, which is subsequently hydrolyzed to generate H2sLDS in good yield (Scheme 4.3). H2sLDS can be assembled into MOF crystals with Eu(NO3)3∙6H2O/Tb(NO3)3∙6H2O in the solvent mixture of DMF, DMSO and H2O.

The structure of EusLDS and TbsLDS are almost identical and only EusLDS will be discussed here.

EusLDS crystallized in the monoclinic space group C2/c. Each Eu center is eight-coordinate, adopting a distorted square anti-prismatic geometry, similar to EusLM (Figure 4.8). Of the eight oxygen atoms, five are from carboxylate groups of the ligand, two are from coordinated water and the last is from coordinated DMF. All Eu-O bond length are within the range of 2.322(5) to 2.494(6) Å. Adjacent pairs of Eu3+ are bridged by two carboxylate groups to form a 1D chain, and there is one extra carboxylate on each Eu3+ that only coordinates with one oxygen, leaving the other oxygen hydrogen bonded with one water molecule coordinated on the same metal center. This water molecule is also hydrogen bonded with another coordinated water on the 3+ adjacent Eu . These 1D chains are interconnected with four other chains by sLDS ligands,

113 extending the structure to a 3D framework. The α angle is approximately 35°, similar to that in

EusLCHO, and therefore the channel size is also small. Thioether side chains are exposed on the channel wall disordered over two possible orientations, suggesting the availability of the space around the thioether group which is crucial for sensing. Coordinated DMF molecules on Eu3+ are modelled with 50% occupancy; they are exposed in the solvent channel, along with other coordinated water and the carboxylate oxygen. Free solvents in the channel are disordered and are not successfully modelled; instead they are calculated according to 1H NMR experiments.

Figure 4.8. Crystal structure of EusLDS: (a) eight-coordinate Eu center; (b) coordination environment of Eu; (c) 1D metal chain and the hydrogen bonding among coordinated solvents and the carboxylate oxygen; (d) channel view along the 1D metal chain, α ≈ 35°; (e) perspective view of the channel. All hydrogen atoms and non-coordinate solvent molecules are omitted for clarity.

114

Figure 4.9. Crystal structure of TbsLDS: (a) eight-coordinate Tb center; (b) coordination environment of Tb; (c) 1D metal chain and the hydrogen bonding among coordinated solvents and the carboxylate oxygen; (d) channel view along the 1D metal chain, α ≈ 35°; (e) perspective view of the channel. All hydrogen atoms and non-coordinate solvent molecules are omitted for clarity.

115

Figure 4.10. (a) The free DMF hydrogen bonded with the coordinated water in TbsLM. (b) The free DMF molecule sandwiched between the two central phenyl ring of two adjacent sLOM ligands. (c) π-π interaction in EusLCHO between adjacent sLCHO ligands. (d) π-π stacking in EusLCO between adjacent sLCO ligands.

As discussed above, all MOFs except EuLsCO demonstrate a topology similar to that of EusLOM, despite the differences in α angle which lead to various degree of porosity. TbsLOM structure is almost identical to EusLOM as a result of the similarity between Eu and Tb. In EusLM and TbsLM, the introduction of non-polar methyl group has lead to less porous structures compared with

EuLOM. In Eu/TbsLOM, one free DMF molecule (per Eu2(sLOM)3 unit) is found to be sandwiched between two central phenyl rings of two adjacent sLOM ligands indicating strong π-π interaction

(Figure 4.10b). Such interaction is absent in Eu/TbsLM: no free DMF was found in the channel of EusLM, and only one free DMF was found in TbsLM, hydrogen bonded with coordinated water (Figure 4.10a). The lack of strong solvent-ligand interaction makes the inclusion of free solvent in the crystal thermodynamically unfavorable and thus leads to less porous structures.

EusLCHO is still less porous than EusLOM despite the existence of polar aldehyde groups, which seems quite contradictory to what was observed in Eu/TbsLM and Eu/TbsLOM systems. However, a careful look at the structure will reveal that the aldehyde groups are now coplanar with the central phenyl ring due to the conjugation effect, which enhanced the π-π interaction between the central rings of adjacent ligands. As shown in the Figure 4.10c, the intra-ligand π-π interaction is

116 so strong that solvent inclusion is no longer favored and thus a less porous structure forms.

EusLCO has functional groups that have extended π systems, which form even bigger planes that strongly favor the stacking between ligands (Figure 4.10d). Such strong π-π interaction as well as the large size of the functional groups leads to the formation of the 2D structure.

Figure 4.11. (a) A typical ligand conformation in EusLDS where the methyl group is slightly tilted off the plane; (b) Ligand arrangement between adjacent metal chains.

2- In Eu/TbsLDS, sulfur atoms in sLDS are coplanar with the central phenyl ring (Figure 4.11a), due to the direct conjugation. However, the intra-ligand π-π interaction is not observed presumably due to the size of sulphur atom and the freely rotating methyl group. As shown in 2- Figure 4.11b, there is plenty of space around the central phenyl ring of sLDS ligand. The fact that we do not observe the disordering of the central ring in the form of rotation suggests the existence of significant amount of solvent molecules around the ligand. These solvents may be exchanged or removed to free up the rotation of the central ring, potentially affecting the ligand energy level and consequently LMET efficiency.

4.4.2 Thermogravimetric analysis

Thermogravimetric analysis is performed on all the ligand acid and their corresponding MOFs under nitrogen atmosphere, to determine their thermal stability and to study the effect of

117

introduction of functional groups. TGA of EusLM (Figure 4.12) shows a stepwise solvent loss at elevated temperatures. The loss of lattice DMF and water starts at 40 °C and completes at around 120 °C, which accounts for about 3.7% of total weight. The second weight loss starts at 160 °C and completes at 220 °C which corresponds to the removal of one coordinated DMF and some of the coordinated water molecules. The remaining coordinated solvents are removed at around 320 °C, and the desolvated framework is stable up to 500 °C before the ligand starts to decompose. TbsLM shows a similar three-step weight loss at 75, 160, and 310 °C respectively, followed by the ligand decomposition at around 530 °C.

H sL 100 2 M EusL M TbsL M 80

60 Weight (%) 40

20

0 100 200 300 400 500 600 700 800 Temperature (deg C)

Figure 4.12. Thermogravimetric analysis of H2sLM, EusLM and TbsLM.

TbsLOM (Figure 4.13) shows a 16.6% weight loss starting at ambient temperature and completing at 200 °C, corresponding to the removal of both coordinated and lattice solvents.

Ligand decomposition starts at around 400 °C, similar with EusLOM. Introduction of the ether group on the ligand negatively affected the thermal stability of the ligand: H2sLOM start to decompose at 280 °C as oppose to 310 °C for H2sLM. Such effect is further magnified to about 100 °C difference in the resulting MOF.

118

100 H sL 2 OM TbsL 90 OM

80

70

Weight (%) 60

50

40

0 100 200 300 400 500 600 700 800 Temperature (deg C)

Figure 4.13. Thermogravimetric analysis of H2LOM and TbsLOM.

100 H sL 2 CHO EusL CHO

90

80 Weight(%) 70

60

0 100 200 300 400 500 600 700 800 Temperature (deg C)

Figure 4.14. Thermogravimetric analysis of H2sLCHO and EusLCHO.

119

100 H sL 2 CO EusL 90 CO

80

70

Weight % 60

50

40 0 100 200 300 400 500 600 700 800 Temperature (deg C)

Figure 4.15. Thermogravimetric analysis of H2sLCO and EusLCO.

100 H sL 2 DS EusL 90 DS TbsL DS

80

70 Weight (%) 60

50

40 0 100 200 300 400 500 600 700 800 Temperature (deg C)

Figure 4.16. Thermogravimetric analysis of H2sLDS, EusLDS and TbsLDS.

EusLCHO (Figure 4.14) shows a two step weight loss of lattice and coordinated solvent, at ambient temperature and 160 °C respectively, as shown in Figure SA10. Decomposition of

120

ligands start at 400 °C, as suggested by the TGA of H2LCHO of which major decomposition starts at 290 °C. According to TGA of EusLCO (Figure 4.15), the first weight loss of 9.5% starts at room temperature and complete at 110 °C, which is followed by a second weight loss in the temperature range between 210 and 290 °C. The approximately 22% weight loss in these two steps corresponds well with the solvent composition of 22.3% as suggested by EA and 1H NMR experiments (Figure 4.38).

EusLDS and TbsLDS both display multi-step solvent removal starting at ambient temperature (Figure 4.16). Solvent is completely removed from the framework at 280 °C, which subsequently start to decompose at 400 °C. Comparing to H2sLDS, of which minor decomposition starts at 220 °C, the improvement of thermal stability of the MOF (~200 °C) is greatest among all the examples.

Overall, the thermal stability of MOFs is found to be superior to their corresponding ligand (acid), as one would normally expect for carboxylate acids and their salts. The introduction of polar functional groups have lead to significant deterioration of thermal stability of MOFs – initial decomposition temperature drops by at least 100 °C comparing to the methyl functionalized version. Furthermore, unsaturated side chains have led to ill-defined weight loss process, as suggested by the much larger range of temperature where the weight loss occurs. In general, the similarities between TGA of ligand and corresponding MOF suggest that TGA of the ligand may serve as a good forecast for the thermal stability of corresponding MOF. Besides the thermal decomposition under nitrogen atmosphere, unsaturated side chains as in EusLCHO and

EusLCO, or the thioether groups in Eu/TbsLDS, are also prone to oxidation at elevated temperature, although not reflect in the aforementioned TGA experiments.

4.4.3 Luminescence properties

EusLM, EusLCHO and EusLDS display strong Eu emission upon excitation at 323, 386 and 373 nm respectively, as show in Figure 4.17, Figure 4.19 and Figure 4.20. The emission peaks at 576, 5 7 7 7 7 591, 615, 699 nm correspond to transitions from D0 to F0, F1, F2, and F4, respectively. TbsLM,

TbsLOM, and TbsLDS all show strong Tb emission upon excitation at 323, 323 and 364 nm (Figure 4.17, Figure 4.18 and Figure 4.20), and the emission peaks at 489, 544, 585, and 621

121

5 7 7 7 7 nm correspond to transitions from D4 to F6, F5, F4, and F3 respectively. No significant emission is observed from EusLCO upon excitation between 250 and 500 nm.

EusL ex@612nm 1.4 M EusL em@323nm M 1.2 TbsL ex@545nm M TbsL em@323nm M 1.0

0.8

0.6

0.4 Intensity(A.U.)

0.2

0.0

300 400 500 600 700 Wavelength (nm)

Figure 4.17. Luminescence spectrum of EusLM (solid) and TbsLM (solid). The broad peak at 307 nm on EusLM excitation, the broad peaks at 640 nm on EusLM and TbsLM emission are due to scattering.

TbsL ex @543nm 1.2 OM TbsL em @323nm OM 1.0

0.8

0.6

0.4 Intensity(A.U.)

0.2

0.0

300 400 500 600 700 Wavelength (nm)

Figure 4.18. Luminescence spectrum of TbsLOM (solid). The broad peaks at 275 nm and 645 nm on TbsLOM excitation and emission respectively, are due to scattering.

122

1.4 EusL em @613nm CHO 1.2 EusL ex @386nm CHO

1.0

0.8

0.6

0.4 Intensity(A.U.)

0.2

0.0

200 300 400 500 600 700 Wavelength (nm)

Figure 4.19. Luminescence spectrum of EusLCHO (solid). The peak at 308 nm is due to scattering.

1.4 EusL ex @615nm DS EusL em @373nm DS 1.2 TbsL ex @542nm DS TbsL em @364nm DS 1.0

0.8

0.6

Intensity(A.U.) 0.4

0.2

0.0

300 400 500 600 700 Wavelength (nm)

Figure 4.20. Luminescence spectrum of EusLDS (solid) and TbsLDS (solid). The peaks at 308 and 271 nm on EusLDS and TbsLDS excitation spectra respectively, are due to scattering.

Because of the LMET process involved in lanthanide luminescence, both the lowest lying ligand triplet state energy level and the repsective singlet – triplet energy gap have great impact on lanthanide emission in lanthanide complexes.[6] Low temperature luminescence measurements were carried out using diluted frozen solution of ligand triethylamine salt at liquid N2 temperature, in order to determine the lowest lying ligand triplet state energy level and calculate the singlet – triplet energy gap. The spectra can be found in Figure 4.21 and the results are

123 summarized in Table 4.1.The emission peaks that are observed in both room temperature and low temperature spectra (with small shifts) are originated from ligand singlet states; the lowest lying triplet state is found to the low energy side of the lowest lying singlet state. For ligands other than sLDS, the emission peaks originated from the lowest lying triplet states are observed as a shoulder or a weak peak as indicated by an arrow in each low temperature fluorescence spectra, or a noisy broad peak in each low temperature phosphorescence spectra, due to their weak intensity. The emission from the sLDS triplet state is very strong and easily observed in both low temperature fluorescence and phosphorescence spectra. The singlet energy levels presented in Table 4.1 are all obtained from the low temperature spectra in order to be consistent with the triplet energy levels. The general rule for optimal Eu/Tb emission[6] requires ligand singlet – triplet gap of at least 5000 cm-1 and the triplet energy level at least 2500 cm-1 above the -1 [7] respective Eu/Tb energy level of 17374 and 20455 cm , respectively. sLM, sLOM, and sLCHO ligands can excite Eu while sLCO cannot, which matches the theory pretty well. However, the optimal energy requirement is not met by either sLM or sLOM, yet Tb emission is still observed. Such discrepancy may be caused by the different ligand conformation which affects its triplet state energy when assembled into the MOF. Although sLDS has small singlet - triplet energy gap, it display the strongest emitting triplet state, because the existence of heavy (sulphur) atom facilitates intersystem crossing. Such enhanced intersystem crossing process even overcomes the disadvantage of low triplet state energy: although sLDS has a lower triplet state energy level than sLCHO, yet Tb emission is still observed in TbsLDS. Overall, the data obtained from ligand salt solution possess considerable predicting power in the luminescence properties of the corresponding MOFs, especially for Eu emission.

124

Table 4.1. Summary of ligand singlet/triplet state energy levels and the respective energy gaps.

Ligand (salt) sLM sLOM sLCHO sLCO sLDS Triplet State (cm-1) 20747 21277 20534 18692 19455 Singlet State1(cm-1) 29762 29412 21834 25773 24331 T-S gap1(cm-1) 9015 8135 1300 7082 4876 Singlet State2(cm-1) 28571 28329 23041 24390 T-S gap2(cm-1) 7825 7052 2508 5699 Singlet State3(cm-1) 24038 T-S gap1(cm-1) 3505 Singlet State4(cm-1) 27248 T-S gap1(cm-1) 6714

(a) 1.0 2 RT FL 0.9 LT FL 0.8 LT PH

0.7 1 PH 0.6

Intensity (A.U.) Intensity

0.5

0.4 0

0.3 Intensity(A.U.)

0.2 FL 0.1 -1 0.0 489 nm -0.1 300 350 400 450 500 550 600 650 Wavelength (nm)

Figure 4.21a. Room temperature fluorescence (RT FL, black), low temperature fluorescence (LT FL, red) and low temperature phosphorescence (LT PH, blue) of the ammonium salt of H2LM. All emissions are measured with excitation at 320 nm.

125

(b) 1.2 RT FL LT FL 1.0 LT PH 1

PH PH

0.8

Intensity (A.U.) Intensity

0.6

0

0.4 Intensity(A.U.)

FL 0.2

0.0 470 nm -1 300 350 400 450 500 550 600 650 Wavelength (nm)

Figure 4.21b. Room temperature fluorescence (RT FL, black), low temperature fluorescence (LT FL, red) and low temperature phosphorescence (LT PH, blue) of the ammonium salt of H2LOM. All emissions are measured with excitation at 320 nm.

(c) 0.20 RT FL 0.18 LT FL 4 0.16 LT PH

0.14 PH

0.12 (A.U.) Intensity 2 0.10

0.08

0.06 Intensity(A.U.) 0

0.04 FL 0.02 487 nm

0.00

-0.02 -2 300 350 400 450 500 550 600 650 Wavelength (nm)

Figure 4.21c. Room temperature fluorescence (RT FL, black), low temperature fluorescence (LT FL, red) and low temperature phosphorescence (LT PH, blue) of the ammonium salt of H2LCHO. All emissions are measured with excitation at 320 nm.

126

(d) 0.08 0.8 RT FL 0.07 LT FL LT PH 0.6 0.06

PH PH

0.05 0.4

Intensity (A.U.) Intensity

0.04 0.2 0.03

Intensity(A.U.) 0.02

0.0 FL 0.01

-0.2 0.00 535 nm 350 400 450 500 550 600 650 Wavelength (nm)

Figure 4.21d. Room temperature fluorescence (RT FL, black), low temperature fluorescence (LT FL, red) and low temperature phosphorescence (LT PH, blue) of the ammonium salt of H2LCO. All emissions are measured with excitation at 320 nm.

(e) 8 RT FL 7 LT FL 60 LT PH 6

5 (A.U.) Intensity 40 4

3 514 nm 20

2 Intensity(A.U.) 1

0 0

-1 350 400 450 500 550 600 650 Wavelength (nm)

Figure 4.21e. Room temperature fluorescence (RT FL, black), low temperature fluorescence (LT FL, red) and low temperature phosphorescence (LT PH, blue) of the ammonium salt of H2LDS. All emissions are measured with excitation at 320 nm.

127

4.4.4 Solvent dependent luminescence and sensing experiments

4.4.4.1 Eu/TbsLM and TbsLOM solvent dependent luminescence

Solvent vapour sensing experiments have been conducted using EusLM and TbsLM but no luminescence change was observed, presumably due to the small channel. Given that the Eu counterpart of TbsLOM has been reported to be a good solvent vapor sensor functioning through solvent-ligand interaction, TbsLOM is expected display similar sensing ability except for being less sensitive to O-H quenching effect due to the larger energy gap between Tb excited states and ground states. However, TbsLOM does not display consistent luminescence intensity between different batches and therefore the sensing properties cannot be evaluated. Such difference may be due to its high sensitivity to solvent content as a result of the bigger energy gap.

4.4.4.2 EusLCHO solvent dependent luminescence and lifetime study

Solvent dependent luminescence was observed for EusLCHO (Figure 4.22). The suspension of

EusLCHO showed varying intensity of Eu emission depending on the identity of the solvent but the differences among all the non-deuterated solvents were not significant enough to be distinguished from the random error. When D2O was used in the suspension, the emission intensity was approximately 3 times of that in other solvent suspension. Such difference can be expected as D2O replaced the coordinated water which is a good quencher for Eu emission. However, despite the ability of DMF and other solvents to replace the coordinated water, no significant enhancement of emission was observed. To better understand this observation, we conducted emission lifetime study with water and methanol suspension and the results are shown in Table 4.2. It is worth noting that only one component of Eu emission decay is observed, despite the existence of two different Eu centers in the framework.

128

1200

1100

1000

900

800

700

600

500

400 Intensity (A.U.) Intensity

300

200

100

0 D2O H2O MeOH EtOH iPrOH MeCN DMF DMSO AcetoneEtOAc THF

Figure 4.22. Intensity of Eu emission @ 614 nm in different suspension of EusLCHO.

129

Table 4.2. Luminescence lifetime and intensity of EusLCHO suspension in various solvent.

H2O D2O Time Lifetime Int.@615nm Time Lifetime Int.@615nm n b (min) (μs) (A.U.)a (min) (μs) (A.U.) H2O 0 367 445 0 760 1133 1.48 10 361 427 10 912 1228 1.75 20 360 419 20 951 1334 1.82 30 356 442 30 936 1298 1.83 40 356 454 40 943 1277 1.84 60 350 414 60 933 1168 1.88 90 347 364 90 938 1215 1.91

CH3OH CD3OD Time Lifetime Int.@615nm Time Lifetime Int.@615nm n c (min) (μs) (A.U.) (min) (μs) (A.U.) ROH 0 455 369 0 557 421 0.94 10 489 437 10 622 574 1.01 20 493 462 20 661 647 1.20 30 499 516 30 709 678 1.39 40 498 532 40 728 728 N/A 60 505 510 60 753 849 N/A 90 504 545 90 795 985 N/A 1110 504 575 1170 961 1461 N/A

−1 -1 [9-11] [a] Emission intensity @ 615 nm, excited at 375 nm. [b] nH2O = 1050 (τH2O −τD2O ) (lifetime in μs unit). [c] calculated from −1 -1 “0.45nROH = 1050 (τCH3OH −τCD3OD ) “(lifetime in μs unit); value after 30 min are assumed to be the same according to the CH3OH run.

130

Table 4.3. Luminescence lifetime EusLCHO suspension in various solvent.

Time Lifetime (μs) (min) Ethanol DMF Acetone 0 457 403 440 10 492 409 469 20 499 414 473 30 494 414 475 40 498 416 478 50 501 415 479 60 501 417 N/A 70 N/A N/A 476

The difference in emission lifetime in H2O and D2O suspension slowly increased in 60 min, suggesting a slow exchange of coordinated water or its protons. At equilibrium, lifetimes in D2O and H2O suspensions are found to be around 938 and 347 μs respectively, calculated to be approximately 1.91 coordinated H2O molecules per Eu center replaced by D2O, matching the crystal structure. CH3OH suspension showed a similar slow increase in lifetime leading to equilibrium within 30 min, due to the replacement of coordinated H2O by CH3OH. CD3OD suspension however, displayed a two-stage process. The first stage is similar to CH3OH suspension, while the second stage required much longer to reach equilibrium. This second stage which is unique to CD3OD but not CH3OH suspension can only be explained by slow deuteration of coordinated H2O by H-D exchange between coordinated water and channel CD3OD. In other words, the initial lifetime increase in 30 min is due to the exchange of coordinated H2O by

CD3OD plus the deuteration of coordinated H2O; the further increase after 30 min is contributed only by the slow deuteration (H-D exchange) process which completed in 10 h. Because the deuteration is much slower (judging from the time to reach equilibrium) than the replacement of

H2O by CD3OD, we can calculate the number of coordinated H2O replaced by CD3OD in the first 30 min while ignoring the contribution from deuteration. Note that in the first 30 min, the difference between CH3OH and CD3OD sample is only the coordinated methanol (coordinated water in both samples are H2O), thus nH2O (which represent the difference in coordinated H2O) is zero and nROH can be calculated with Eq. 4.1 and the results are shown in Table 4.2.

−1 -1 Eq. 4.1: 0.45nROH = 1050 (τH −τD ) (lifetime in μs unit)

131

In the first 30 min after the initial mixing, the number of coordinated H2O replaced by methanol increased from 0.94 to 1.39 according to the change in luminescence lifetime. Further increase in lifetime after 30 min in CD3OD suspension was due to the deuteration of the remaining coordinated water. After 10 h, the lifetime stabilized at around 950 μs, suggesting complete deuteration of coordinated water and the resulting MOF contained only coordinated D2O and

CD3OD. Since neither D2O nor CD3OD quenches Eu emission, the deactivation rate and the luminescence lifetime of Eu should be the same in the CD3OD suspension and in D2O sample. In fact, emission lifetime of CD3OD suspension at equilibrium (~960 μs) is quite similar to that of

D2O suspension (~940 μs), where the difference may come from experimental error or weak solvent-ligand interaction.

st Because CD3OD is in large excess in suspension, the H-D exchange is like a 1 order reaction.

Difference between CD3OD and CH3OH suspension includes not only the coordinated methanol, but also small amount of coordinated water due to H-D exchange. Therefore we have:

0.45nCD3OD + nD2O = 1050 (1/τCH3OH – 1/τCD3OD) (3)

There should be a linear relationship between ln(nD2O) and time according to the reaction order.

However, nCD3OD can only be calculated when τD2O is known. To avoid the iteration process to solve for nD2O and nCD3OD, linear approximation of 0.45nCD3OD + nD2O can be used from 30 to 90 min to estimate the contribution of H-D exchange (change in nD2O) in the first 30 min and the slope of the fitted line is found to be 0.0026 per min, corresponding to the contribution from H-D exchange. H-D exchange effect can then be removed from the 0.45 nCD3OD + nD2O data obtained in the first 30 min by subtraction 0.0026 * time and the number of water replaced by methanol was found to be 0.96, 1.08 and 1.17 at 10, 20 and 30 min respectively (Table 4.4). These values, althtough smaller than that when H-D exchange is ignored, are still consistent with the conclusion drawn.

If we further assume that the solvent only interact with MOF through the Eu coordination site (given the weak solvent-ligand interaction shown in water and methanol suspension), we can −1 -1 calculate the number of H2O replaced by ethanol using 0.55 nEtOH = 1050 (τH2O −τEtOH ), as the net effect of replacing one H2O is the decrease of quenching rate constant by 0.55 (1-0.45), where τH2O and τsolvent are the lifetime measured from suspension in H2O and the solvent of interest and nEtOH is the number of EtOH that replaced coordinated H2O. Similarly for DMF and

132

−1 -1 acetone, nsolvent = 1050 (τH2O −τsolvent ) was used. Accordingly, ethanol was found to be able to replace 1.40 coordinated water (Table 4.3), similar to methanol, while DMF and acetone can only replace (or remove) 0.88 and 0.45 coordinated water respectively, due to their bigger size and weaker coordinating ability. It is worth noting that the 1.40 value calculated from ethanol suspension is no comparable with the methanol value obtained (1.17 at 30 min) when linear approximation of H-D exchange is used, due to the different assumptions used.

Besides the difference in lifetime caused by the change in coordinated solvents, it is also worth noting that, the emission intensity is highly correlated with the luminescent lifetime: when the lifetime is longer, the intensity is also stronger. The explanation of such relation is straightforward for EusLCHO system where solvent-ligand interaction is very weak: coordinated solvents that have O-H groups provide extra deactivation pathway for Eu excited states, reducing their lifetime and at the same time decrease the number of Eu that deactivate through the radiative transition, leading to lower emission intensity.

Table 4.4. Calculation after H-D exchange is estimated with linear approximation

Time CH3OH CD3OD 0.45 nCD3OD + nD2O Subtracted by n(CD3OD)

(min) Lifetime Lifetime (1050*(1/τCH3OH – 1/τCD3OD)) (0.0026 * Time) 0 455 557 0.42242 0.42242 0.94 10 489 622 0.45646 0.43046 0.96 20 493 661 0.53954 0.48754 1.08 30 504 709 0.60237 0.52437 1.17

Table 4.5. Luminescence lifetime and intensities of EuLOM in various solvents. Solvent Lifetime Int.@615nm Solvent Lifetime Int.@615nm n(water) (μs) (A.U.)a (μs) (A.U.)a

H2O 417 227 D2O 1427 737 1.78 Solvent Lifetime Int.@615nm Solvent Lifetime Int.@615nm n(methanol) (μs) (A.U.)a (μs) (A.U.)a

CH3OH 707 1197 CD3OD 1529 2247 1.69

In order to compare EusLCHO system with EusLOM system, lifetime of different EusLOM suspensions are measured and summarized in Table 4.5. The number of coordinated water/methanol in the respective suspension are 1.78 and 1.69, comparing to 1.91 and 1.39 in

133

EusLCHO case. Comparing to EusLOM, the overall difficulty to replace/remove coordinated water is likely resulting from the small channel size and hydrogen bonding interaction between water and carboxylate groups. For the same reasons, solvent molecules or metal ions did not trigger any significant luminescence change through the interaction with the aldehyde or carboxylate groups; the former is buried in the channel wall while the latter is occupied by coordinated water.

4.4.4.3 EusLCO solvent dependent luminescence

Solvent dependent luminescence is also observed on EusLCO (Figure 4.23, Figure 4.24), despite the absence of Eu emission. The as-synthesized MOF does not show any luminescence emission when measured in solid state; the emission was only observed in suspension. Strongest emission was observed in its water suspension, while the emission intensities are comparable among the all other solvent suspension. Although ketone and carboxylate oxygen atoms are exposed in the solvent channels, hydrogen bonding interaction is not dominant: emission intensity in alcohols and amines do not exceed that in other solvents significantly. The strong emission observed in water might be due to its high polarity and small size, while other solvent-ligand interaction has contributed to the minor differences among other solvents.

H2O Emission 70 H2O Excitation EtOH Emission EtOH Excitation 60 THF Emission THF Excitation 50 DMF Emission DMF Excitation 40

30 Intensity (A.U.) 20

10

0 250 300 350 400 450 500 550 600 650 700 Wavelength (nm)

Figure 4.23. Excitation and emission spectrum of EuLCO suspension of various solvents.

134

70

60

50

40

30

Intensity (A.U.) 20

10

0 H2O MeOHEtOH iPrOHEt2NHDMF THF EtOAcMeCNAcetoneEt2O

Figure 4.24. Emission intensity of EuLCO suspension in various solvents. Excitation at 400 nm was used and the average emission intensity at 481-485 nm was plotted as the y-axis.

4.4.4.4 EusLDS solvent dependent luminescence and vapor sensing

Both EusLDS and TbsLDS display varying luminescence intensity depending on the solvent.

However, TbsLDS seem to suffer from the similar problem as TbsLOM that the intensity values are not very reproducible, presumably due to the highly sensitive nature of Tb emission.

Because the as-synthesized EusLDS contains three different solvents, we decided to first exchange them with water before used for sensing experiments, in order to achieve the best turn- on contrast and recyclability. The complete removal of DMF and DMSO is confirmed with 1H

NMR experiment (Figure 4.33). The emission spectra of water-soaked EusLDS sample suspended in different solvents have been measured. Selectively spectra are provided in Figure 4.25 and their maximum emission intensity is plotted in the histogram in Figure 4.26. As shown in Figure 4.26, H2O suspension gives the weakest emission while DMSO suspension gives the strongest emission, followed by ethyl acetate and then acetone. It is surprising that the Eu intensity in D2O suspension is among the weakest. D2O suspension shows the strongest emission for EusLCHO, as the use of D2O completely eliminates the O-H quenching effect. Thus, the dominating factor in EusLDS system is no longer O-H quenching, but more likely, the solvent-

135

ligand interaction. Luminescence lifetimes of the EusLDS suspension were measured (Table 4.6) and the number of water and methanol coordinated on Eu in suspension were calculated to be 2.63 and 2.59 respective. These numbers are very close to each other, and match fairly well with the observed value of 3 in the crystal structure. Such observation also shows that methanol is able to replace all three coordinated water that are hydrogen bonded together and with the carboxylate oxygen.

Two types of possible solvent-ligand interactions exist in EusLDS system. First, after coordinated water molecules are removed, it is possible that other solvents cannot bind to the open metal sites as effectively as those two coordinated water molecules that are hydrogen bonded (not the case for methanol according to lifetime studies), due to their bigger size. Such postulation is consistent with the fact that DMSO, acetone, ethyl acetate and DMF are bigger than acetonitrile, water or methanol. It is also possible that even if these solvents can occupy the open metal sites that were occupied by water, they cannot effectively hydrogen bond with the carboxylate oxygen that was originally hydrogen bonded with coordinated water. The loss of hydrogen bonding on the carboxylate group presumably shifts the ligand energy states or improves the LMET process that leads to the enhancement of the emission.

800 D2O 700 H2O EtOAc 600 DMSO 500

400

300 Intensity(A.U.) 200

100

0

560 580 600 620 640 660 680 700 720 Wavelength (nm)

Figure 4.25. Emission spectra of water-soaked EusLDS suspended in various solvents.

136

775 800

700

600

500

400

300 196 200 137 141 119 MaximumIntensity (A.U.) 100 46 44 34 8 0 EtOAc D O H O CH OH CD OD DMF MeCN DMSO Acetone 2 2 3 3

Figure 4.26. Maximum emission intensity of water-soaked EusLDS suspended in various solvents.

Table 4.6. Luminescence lifetime of water-soaked EusLDS in various solvents.

H2O D2O CH3OH CD3OD

Luminescence lifetime (μs) 307 1334 548 1399

Another type of possible solvent-ligand interaction involves the conformation change of the ligands. As discussed earlier, there is plenty of space around the central phenyl rings of the ligands however disorder is not observed in the crystal structure as the space is occupied by disorder free solvent. It is possible that the exchange of solvent in the suspension lead to the change in the preferred conformation of the central ring, which will affect the energy levels of the ligands and therefore affect the emission intensity.

Given the strong enhancement observed in suspension, vapor sensing experiments were performed to test the sensitivity of EusLDS. Water-soaked EusLDS samples were incubated with various solvent of 24 h (up to 72 h for DMSO) and the luminescence enhancement is plotted in Figure 4.27. As shown in the histogram, ethyl acetate caused the strongest turn-on, followed by DMF and DMSO; and in contrast to most other solvents, MeCN triggered a moderate turn-off. The solvent ability to trigger luminescence turn-on in vapor experiments here not only is much

137 weaker, the trend is also different from that in the suspension, especially for DMSO, which triggered weaker turn-on than ethyl acetate in vapor experiments. The major difference between suspension and vapor is that, the concentration of the testing solvent in the MOF in suspension is much higher than that in vapor tests and therefore the turn-on contribution from the solvent- ligand interaction is lower. DMSO not only suffered from the lower concentration in general, its much lower vapor pressure also limited the rate of diffusion and amount that diffused into the MOF. In fact, the maximum turn-on triggered by DMSO after incubation is slightly higher than that by ethyl acetate, although a much longer time (72h) is required to achieve the maximum.

Although the target of the thioether group is heavy metal ions, sensing experiments performed with different metal salt in EusLDS suspension in DMSO did not show any significant luminescence signal change, as the concentration of metal salts are still much lower than solvents.

6

5

4

3

2

1

0 Luminescenceenhancement (fold)

DMF DMSO DMSO max EtOAc Et2O Acetone MeOH MeCN -1

Figure 4.27. Luminescence enhancement of water-soaked EusLDS after incubation with various solvent for 24h. “DMSO max” is the maximum enhancement achieved for incubation of DMSO (after 72h). Luminescence enhancement is calculated as Iafter/Ibefore – 1 (Ibefore is the average intensity before incubation, Iafter is the intensity after incubation.).

138

4.5 Summary

In this chapter, synthesis of a series of new luminescent MOFs with similar structure but different functionalized ligand have been described and their properties be discussed. By comparing their properties with the respective ligand properties, we can summarize the connection. Strong ligand-solvent interaction through the functional group is necessary to achieve a porous structure as concluded from the comparison between Eu/TbsLM and Eu/TbsLOM; however, strong ligand-ligand interaction caused by conjugated side-chain is detrimental to the formation of solvent channels. Thermal stability of the ligands decreased as more reactive groups were introduced, and such reduced stability is generally inherited in the resulting MOFs. Luminescent properties of the MOF can be predicted with some accuracy from the single/triplet energy levels of the ligands, following the well-established energy requirements. Although

EusLCHO and EusLCO did not show strong response to different solvent, they displayed some interesting MOF-solvent interaction that leads to luminescence property change, via metal- solvent and ligand-solvent interaction respectively. Conversely, EusLDS demonstrated very strong response to different solvents in suspension and vapor tests, triggered by ligand-solvent interaction. No significant sensing ability for metal ions/salts has been observed for the MOFs discussed in this chapter, suggesting that stronger interactions are needed for metal ions that are less concentrated (than solvents) in the testing environment. The knowledge obtained from these ligands/MOFs will further guide our design of luminescent LnMOFs for sensing application in the future.

139

Table 4.7. Crystallographic data of EusLM, TbsLM, TbsLOM and EusLCHO

EusLM TbsLM TbsLOM EuLCHO

formula C144H132N4O32Tb4 ∙ C36H34O11Tb ∙ C36H32NO7.5Eu C66H44O22Eu2 2C3H7NO 2.1C3H7NO ∙ 1.5H2O

F.W. 750.61 3212.41 1072.26 1492.97 T (K) 150(2) 150(2) 150(2) 150(2) space group C2/c P2(1)/n P-1 P2(1)/m a (Å) 36.1865(18) 18.9512(10) 9.3494(4) 8.8531(15) b (Å) 10.8094(6) 19.6780(12) 13.5299(6) 34.493(6) c (Å) 17.0355(8) 35.745(2) 18.2593(8) 11.045(2)  (deg) 90 90 74.738(2) 90  (deg) 90.300(2) 90.499 86.071(2) 107.839(8)  (deg) 90 90 88.207(2) 90 V (Å3) 6663.4(6) 13329.4(13) 2222.93(17) 3214.2(10) Z 8 4 2 2

−3 Dc (g·cm ) 1.494 1.601 1.467 1.534  (mm−1) 1.933 2.180 1.659 2.010 no. reflns collcd 52541 57291 37047 51122 no. indept reflns 7642 22963 10153 7441 GOF on F2 1.080 0.932 1.097 1.049 R [I > 2σ(I)] R1 = 0.0556 R1 = 0.0440 R1 = 0.0308 R1 = 0.0391 wR2 = 0.1228 wR2 = 0.0700 wR2 = 0.0808 wR2 = 0.0939 R (all data) R1 = 0.0958 R1 = 0.0971 R1 = 0.0349 R1 = 0.0496 wR2 = 0.1355 wR2 = 0.0803 wR2 = 0.0836 wR2 = 0.0976

140

Table 4.8. Crystallographic data of EusLCO, EusLDS, and TbsLDS.

EusLCO EusLDS TbsLDS

formula C44.5H40.5N0.5O12S0.5Eu ∙ C36H35NO9S3Eu C36H35NO9S3Tb 0.3C3H7NO ∙ 2.5H2O

F.W. 1009.24 873.82 880.78 T (K) 150(2) 150(2) 150(2) space group P-1 C2/c C2/c a (Å) 10.7331(13) 36.1052(7) 35.9719(13) b (Å) 13.0512(17) 11.6490(3) 11.6117(4) c (Å) 18.657(2) 19.7209(5) 19.6158(7)  (deg) 84.384(4) 90 90  (deg) 89.831(4) 96.8600(10) 96.8540(10)  (deg) 66.735(4) 90 90 V (Å3) 2387.6(5) 8235.0(3) 8134.4(5) Z 4 8 8

−3 Dc (g·cm ) 1.363 1.345 1.380  (mm−1) 1.397 1.720 1.938 no. reflns collcd 40000 36344 36594 no. indept reflns 10907 9456 9336 GOF on F2 1.062 1.111 1.019 R [I > 2σ(I)] R1 = 0.0592 R1 = 0.0673 R1 = 0.0701 wR2 = 0.1661 wR2 = 0.1809 wR2 = 0.1891 R (all data) R1 = 0.0761 R1 = 0.0927 R1 = 0.1076 wR2 = 0.1785 wR2 = 0.1968 wR2 = 0.2124

141

Table 4.9. Selected bond lengths (Å) and angles (deg) for EusLM

EusLM Eu(1)-O(4)#1 2.342(5) O(5)-Eu(1)-O(6)#2 126.41(17) Eu(1)-O(3) 2.352(5) O(4)#1-Eu(1)-O(8) 73.00(19) Eu(1)-O(1)#2 2.355(5) O(3)-Eu(1)-O(8) 73.62(19) Eu(1)-O(2) 2.373(5) O(1)#2-Eu(1)-O(8) 76.6(2) Eu(1)-O(5) 2.388(5) O(2)-Eu(1)-O(8) 141.61(19) Eu(1)-O(6)#2 2.417(5) O(5)-Eu(1)-O(8) 138.51(18) Eu(1)-O(8) 2.438(6) O(6)#2-Eu(1)-O(8) 76.17(19) Eu(1)-O(7) 2.709(4) O(4)#1-Eu(1)-O(7) 66.60(15) O(4)#1-Eu(1)-O(3) 93.50(17) O(3)-Eu(1)-O(7) 70.69(17) O(4)#1-Eu(1)-O(1)#2 87.87(17) O(1)#2-Eu(1)-O(7) 137.02(14) O(3)-Eu(1)-O(1)#2 148.33(18) O(2)-Eu(1)-O(7) 66.71(15) O(4)#1-Eu(1)-O(2) 132.93(18) O(5)-Eu(1)-O(7) 66.34(18) O(3)-Eu(1)-O(2) 76.61(17) O(6)#2-Eu(1)-O(7) 138.23(12) O(1)#2-Eu(1)-O(2) 123.94(18) O(8)-Eu(1)-O(7) 123.0(2) O(4)#1-Eu(1)-O(5) 76.63(18) O(4)#1-Eu(1)-O(1) 143.86(16) O(3)-Eu(1)-O(5) 136.35(17) O(3)-Eu(1)-O(1) 115.42(14) O(1)#2-Eu(1)-O(5) 74.67(18) O(1)#2-Eu(1)-O(1) 78.62(17) O(2)-Eu(1)-O(5) 79.81(17) O(2)-Eu(1)-O(1) 45.45(15) O(4)#1-Eu(1)-O(6)#2 148.76(19) O(5)-Eu(1)-O(1) 67.53(14) O(3)-Eu(1)-O(6)#2 82.55(16) O(6)#2-Eu(1)-O(1) 61.50(16) O(1)#2-Eu(1)-O(6)#2 80.14(16) O(8)-Eu(1)-O(1) 133.78(18) O(2)-Eu(1)-O(6)#2 76.44(17) O(7)-Eu(1)-O(1) 101.57(17)

142

Table 4.10. Selected bond lengths (Å) and angles (deg) for TbsLM

TbsLM Tb(1)-O(8) 2.249(4) O(1)-Tb(1)-O(3) 95.11(15) Tb(1)-O(1) 2.255(4) O(8)-Tb(1)-O(4) 95.76(17) Tb(1)-O(3) 2.305(4) O(1)-Tb(1)-O(4) 82.95(16) Tb(1)-O(4) 2.344(4) O(3)-Tb(1)-O(4) 142.42(13) Tb(1)-O(5) 2.348(4) O(8)-Tb(1)-O(5) 92.96(14) Tb(1)-O(2) 2.426(3) O(1)-Tb(1)-O(5) 84.42(14) Tb(1)-O(6) 2.430(3) O(3)-Tb(1)-O(5) 73.52(12) Tb(2)-O(9) 2.287(4) O(4)-Tb(1)-O(5) 142.77(13) Tb(2)-O(16) 2.293(4) O(8)-Tb(1)-O(2) 83.93(13) Tb(2)-O(11) 2.324(3) O(1)-Tb(1)-O(2) 101.49(14) Tb(2)-O(14) 2.355(3) O(3)-Tb(1)-O(2) 72.75(12) Tb(2)-O(15) 2.393(4) O(4)-Tb(1)-O(2) 70.94(13) Tb(2)-O(12) 2.432(4) O(5)-Tb(1)-O(2) 146.13(13) Tb(2)-O(10) 2.456(4) O(8)-Tb(1)-O(6) 83.83(14) Tb(2)-O(13) 2.460(4) O(1)-Tb(1)-O(6) 89.78(14) Tb(3)-O(22) 2.240(5) O(3)-Tb(1)-O(6) 143.39(13) Tb(3)-O(17) 2.242(4) O(4)-Tb(1)-O(6) 74.18(13) Tb(3)-O(19) 2.320(4) O(5)-Tb(1)-O(6) 70.89(12) Tb(3)-O(20) 2.354(5) O(2)-Tb(1)-O(6) 141.51(13) Tb(3)-O(21) 2.355(4) O(9)-Tb(2)-O(16) 144.51(14) Tb(3)-O(23) 2.374(3) O(9)-Tb(2)-O(11) 93.22(14) Tb(3)-O(18) 2.397(3) O(16)-Tb(2)-O(11) 92.09(14) Tb(4)-O(32) 2.301(4) O(9)-Tb(2)-O(14) 87.18(14) Tb(4)-O(27) 2.308(4) O(16)-Tb(2)-O(14) 105.91(14) Tb(4)-O(25) 2.349(3) O(11)-Tb(2)-O(14) 148.08(13) Tb(4)-O(29) 2.375(3) O(9)-Tb(2)-O(15) 140.91(14) Tb(4)-O(31) 2.390(4) O(16)-Tb(2)-O(15) 74.42(14) Tb(4)-O(26) 2.419(5) O(11)-Tb(2)-O(15) 77.98(14) Tb(4)-O(30) 2.461(4) O(14)-Tb(2)-O(15) 81.73(13) Tb(4)-O(28) 2.468(4) O(9)-Tb(2)-O(12) 71.55(15) O(8)-Tb(1)-O(1) 173.59(13) O(16)-Tb(2)-O(12) 143.61(15) O(8)-Tb(1)-O(3) 89.74(15) O(11)-Tb(2)-O(12) 77.26(14)

143

Table 4.10 (cont.). Selected bond lengths (Å) and angles (deg) for TbsLM

TbsLM O(14)-Tb(2)-O(12) 72.62(14) O(19)-Tb(3)-O(18) 73.42(12) O(15)-Tb(2)-O(12) 69.36(15) O(20)-Tb(3)-O(18) 69.22(13) O(9)-Tb(2)-O(10) 75.12(14) O(21)-Tb(3)-O(18) 146.39(13) O(16)-Tb(2)-O(10) 73.68(14) O(23)-Tb(3)-O(18) 140.63(13) O(11)-Tb(2)-O(10) 70.65(13) O(32)-Tb(4)-O(27) 145.33(14) O(14)-Tb(2)-O(10) 139.28(13) O(32)-Tb(4)-O(25) 93.46(14) O(15)-Tb(2)-O(10) 133.74(13) O(27)-Tb(4)-O(25) 91.47(14) O(12)-Tb(2)-O(10) 131.64(15) O(32)-Tb(4)-O(29) 105.83(13) O(9)-Tb(2)-O(13) 77.60(14) O(27)-Tb(4)-O(29) 86.69(13) O(16)-Tb(2)-O(13) 76.62(14) O(25)-Tb(4)-O(29) 148.36(13) O(11)-Tb(2)-O(13) 141.70(13) O(32)-Tb(4)-O(31) 72.73(14) O(14)-Tb(2)-O(13) 69.43(13) O(27)-Tb(4)-O(31) 141.69(14) O(15)-Tb(2)-O(13) 131.06(14) O(25)-Tb(4)-O(31) 78.36(13) O(12)-Tb(2)-O(13) 131.58(14) O(29)-Tb(4)-O(31) 83.65(13) O(10)-Tb(2)-O(13) 71.06(12) O(32)-Tb(4)-O(26) 142.69(16) O(22)-Tb(3)-O(17) 171.56(13) O(27)-Tb(4)-O(26) 71.63(16) O(22)-Tb(3)-O(19) 91.59(15) O(25)-Tb(4)-O(26) 76.27(15) O(17)-Tb(3)-O(19) 90.58(15) O(29)-Tb(4)-O(26) 73.19(15) O(22)-Tb(3)-O(20) 94.71(18) O(31)-Tb(4)-O(26) 70.08(16) O(17)-Tb(3)-O(20) 88.52(18) O(32)-Tb(4)-O(30) 74.08(14) O(19)-Tb(3)-O(20) 142.40(13) O(27)-Tb(4)-O(30) 80.98(14) O(22)-Tb(3)-O(21) 87.76(14) O(25)-Tb(4)-O(30) 141.81(13) O(17)-Tb(3)-O(21) 85.08(14) O(29)-Tb(4)-O(30) 69.00(13) O(19)-Tb(3)-O(21) 72.97(12) O(31)-Tb(4)-O(30) 128.50(14) O(20)-Tb(3)-O(21) 144.20(13) O(26)-Tb(4)-O(30) 134.06(16) O(22)-Tb(3)-O(23) 82.95(15) O(32)-Tb(4)-O(28) 75.87(15) O(17)-Tb(3)-O(23) 90.66(14) O(27)-Tb(4)-O(28) 73.54(15) O(19)-Tb(3)-O(23) 145.59(13) O(25)-Tb(4)-O(28) 70.59(13) O(20)-Tb(3)-O(23) 72.01(13) O(29)-Tb(4)-O(28) 137.89(13) O(21)-Tb(3)-O(23) 72.88(13) O(31)-Tb(4)-O(28) 133.80(14) O(22)-Tb(3)-O(18) 93.56(14) O(26)-Tb(4)-O(28) 130.49(15) O(17)-Tb(3)-O(18) 94.88(14) O(30)-Tb(4)-O(28) 71.37(12)

144

Table 4.11. Selected bond lengths (Å) and angles (deg) for TbsLOM

TbsLOM Tb(1)-O(2)#1 2.319(2) O(2)#1-Tb(1)-O(1) 122.48(7) Tb(1)-O(4)#2 2.321(2) O(4)#2-Tb(1)-O(1) 79.31(7) Tb(1)-O(5) 2.324(2) O(5)-Tb(1)-O(1) 80.63(8) Tb(1)-O(3) 2.365(2) O(3)-Tb(1)-O(1) 137.17(8) Tb(1)-O(6)#1 2.365(2) O(6)#1-Tb(1)-O(1) 76.39(8) Tb(1)-O(1) 2.404(2) O(2)#1-Tb(1)-O(7) 72.99(8) Tb(1)-O(7) 2.469(3) O(4)#2-Tb(1)-O(7) 74.47(8) Tb(1)-O(8) 2.547(2) O(5)-Tb(1)-O(7) 137.05(8) O(2)#1-Tb(1)-O(4)#2 146.27(8) O(3)-Tb(1)-O(7) 72.68(8) O(2)#1-Tb(1)-O(5) 72.41(8) O(6)#1-Tb(1)-O(7) 70.95(8) O(4)#2-Tb(1)-O(5) 140.58(8) O(1)-Tb(1)-O(7) 140.55(8) O(2)#1-Tb(1)-O(3) 87.82(7) O(2)#1-Tb(1)-O(8) 139.37(8) O(4)#2-Tb(1)-O(3) 90.90(7) O(4)#2-Tb(1)-O(8) 69.29(8) O(5)-Tb(1)-O(3) 81.33(7) O(5)-Tb(1)-O(8) 72.00(8) O(2)#1-Tb(1)-O(6)#1 79.99(7) O(3)-Tb(1)-O(8) 68.03(8) O(4)#2-Tb(1)-O(6)#1 81.26(7) O(6)#1-Tb(1)-O(8) 138.10(8) O(5)-Tb(1)-O(6)#1 125.88(7) O(1)-Tb(1)-O(8) 69.52(8) O(3)-Tb(1)-O(6)#1 143.59(8) O(7)-Tb(1)-O(8) 124.95(8)

145

Table 4.12. Selected bond lengths (Å) and angles (deg) for EusLCHO

EusLCHO Eu(1)-O(4)#1 2.363(3) O(4)#1-Eu(1)-O(2) 145.89(10) Eu(1)-O(4) 2.363(3) O(4)-Eu(1)-O(2) 98.30(10) Eu(1)-O(7) 2.385(5) O(7)-Eu(1)-O(2) 134.49(10) Eu(1)-O(5) 2.411(4) O(5)-Eu(1)-O(2) 73.35(12) Eu(1)-O(8) 2.449(5) O(8)-Eu(1)-O(2) 71.71(12) Eu(1)-O(1)#1 2.508(3) O(1)#1-Eu(1)-O(2) 114.56(9) Eu(1)-O(1) 2.508(3) O(1)-Eu(1)-O(2) 50.91(9) Eu(1)-O(2) 2.601(3) O(4)#1-Eu(1)-O(2)#1 98.30(10) Eu(1)-O(2)#1 2.601(3) O(4)-Eu(1)-O(2)#1 145.89(10) Eu(2)-O(9)#1 2.317(3) O(7)-Eu(1)-O(2)#1 134.49(10) Eu(2)-O(9) 2.317(3) O(5)-Eu(1)-O(2)#1 73.35(12) Eu(2)-O(3)#1 2.359(3) O(8)-Eu(1)-O(2)#1 71.71(12) Eu(2)-O(3) 2.359(3) O(1)#1-Eu(1)-O(2)#1 50.91(9) Eu(2)-O(2) 2.402(3) O(1)-Eu(1)-O(2)#1 114.56(9) Eu(2)-O(2)#1 2.402(3) O(2)-Eu(1)-O(2)#1 67.85(12) Eu(2)-O(6) 2.447(5) O(9)#1-Eu(2)-O(9) 78.60(16) O(4)#1-Eu(1)-O(4) 75.52(15) O(9)#1-Eu(2)-O(3)#1 80.96(11) O(4)#1-Eu(1)-O(7) 77.84(12) O(9)-Eu(2)-O(3)#1 131.22(11) O(4)-Eu(1)-O(7) 77.85(13) O(9)#1-Eu(2)-O(3) 131.22(11) O(4)#1-Eu(1)-O(5) 134.46(10) O(9)-Eu(2)-O(3) 80.96(11) O(4)-Eu(1)-O(5) 134.46(10) O(3)#1-Eu(2)-O(3) 80.22(14) O(7)-Eu(1)-O(5) 77.82(17) O(9)#1-Eu(2)-O(2) 125.38(11) O(4)#1-Eu(1)-O(8) 74.35(12) O(9)-Eu(2)-O(2) 79.25(11) O(4)-Eu(1)-O(8) 74.35(12) O(3)#1-Eu(2)-O(2) 146.13(11) O(7)-Eu(1)-O(8) 144.61(17) O(3)-Eu(2)-O(2) 92.97(10) O(5)-Eu(1)-O(8) 137.58(17) O(9)#1-Eu(2)-O(2)#1 79.25(11) O(4)#1-Eu(1)-O(1)#1 70.95(10) O(9)-Eu(2)-O(2)#1 125.38(11) O(4)-Eu(1)-O(1)#1 145.25(10) O(3)#1-Eu(2)-O(2)#1 92.97(10) O(7)-Eu(1)-O(1)#1 86.33(8) O(3)-Eu(2)-O(2)#1 146.13(11) O(5)-Eu(1)-O(1)#1 69.61(7) O(2)-Eu(2)-O(2)#1 74.35(13) O(8)-Eu(1)-O(1)#1 104.51(8) O(9)#1-Eu(2)-O(6) 140.56(8) O(4)#1-Eu(1)-O(1) 145.25(10) O(9)-Eu(2)-O(6) 140.56(8) O(4)-Eu(1)-O(1) 70.95(10) O(3)#1-Eu(2)-O(6) 73.74(12) O(7)-Eu(1)-O(1) 86.33(8) O(3)-Eu(2)-O(6) 73.74(12) O(5)-Eu(1)-O(1) 69.61(7) O(2)-Eu(2)-O(6) 72.51(12) O(8)-Eu(1)-O(1) 104.51(8) O(2)#1-Eu(2)-O(6) 72.51(12) O(1)#1-Eu(1)-O(1) 139.20(14)

146

Table 4.13. Selected bond lengths (Å) and angles (deg) for EusLCO

EusLCO Eu(1)-O(6)#1 2.290(5) O(6)#1-Eu(1)-O(11) 133.4(2) Eu(1)-O(5) 2.338(4) O(5)-Eu(1)-O(11) 71.95(19) Eu(1)-O(2) 2.368(4) O(2)-Eu(1)-O(11) 71.4(2) Eu(1)-O(1) 2.369(4) O(1)-Eu(1)-O(11) 76.4(2) Eu(1)-O(3) 2.373(4) O(3)-Eu(1)-O(11) 124.29(19) Eu(1)-O(11) 2.423(7) O(6)#1-Eu(1)-O(10) 77.98(17) Eu(1)-O(10) 2.456(4) O(5)-Eu(1)-O(10) 147.44(15) Eu(1)-O(12) 2.550(9) O(2)-Eu(1)-O(10) 76.79(16) O(6)#1-Eu(1)-O(5) 88.41(15) O(1)-Eu(1)-O(10) 72.60(15) O(6)#1-Eu(1)-O(2) 154.04(19) O(3)-Eu(1)-O(10) 73.94(14) O(5)-Eu(1)-O(2) 110.01(14) O(11)-Eu(1)-O(10) 137.48(19) O(6)#1-Eu(1)-O(1) 94.59(16) O(6)#1-Eu(1)-O(12) 66.3(2) O(5)-Eu(1)-O(1) 138.63(15) O(5)-Eu(1)-O(12) 76.1(2) O(2)-Eu(1)-O(1) 83.64(14) O(2)-Eu(1)-O(12) 134.5(2) O(6)#1-Eu(1)-O(3) 89.07(17) O(1)-Eu(1)-O(12) 67.7(2) O(5)-Eu(1)-O(3) 76.43(14) O(3)-Eu(1)-O(12) 143.3(2) O(2)-Eu(1)-O(3) 78.22(15) O(11)-Eu(1)-O(12) 68.1(3) O(1)-Eu(1)-O(3) 144.74(15) O(10)-Eu(1)-O(12) 122.8(2)

147

Table 4.14. Selected bond lengths (Å) and angles (deg) for EusLDS

EusLDS Eu(1)-O(2) 2.322(5) O(2)-Eu(1)-O(7) 76.4(2) Eu(1)-O(1) 2.333(5) O(1)-Eu(1)-O(7) 74.7(2) Eu(1)-O(5) 2.351(5) O(5)-Eu(1)-O(7) 141.8(2) Eu(1)-O(4) 2.370(5) O(4)-Eu(1)-O(7) 73.9(2) Eu(1)-O(3) 2.426(6) O(3)-Eu(1)-O(7) 127.95(18) Eu(1)-O(7) 2.486(6) O(2)-Eu(1)-O(6) 72.0(2) Eu(1)-O(6) 2.493(7) O(1)-Eu(1)-O(6) 143.55(19) Eu(1)-O(8) 2.494(6) O(5)-Eu(1)-O(6) 73.2(2) O(2)-Eu(1)-O(1) 143.7(2) O(4)-Eu(1)-O(6) 77.08(19) O(2)-Eu(1)-O(5) 102.02(19) O(3)-Eu(1)-O(6) 73.17(19) O(1)-Eu(1)-O(5) 87.7(2) O(7)-Eu(1)-O(6) 137.72(19) O(2)-Eu(1)-O(4) 87.14(19) O(2)-Eu(1)-O(8) 74.8(2) O(1)-Eu(1)-O(4) 105.28(18) O(1)-Eu(1)-O(8) 75.43(19) O(5)-Eu(1)-O(4) 144.2(2) O(5)-Eu(1)-O(8) 71.6(2) O(2)-Eu(1)-O(3) 144.3(2) O(4)-Eu(1)-O(8) 143.6(2) O(1)-Eu(1)-O(3) 71.9(2) O(3)-Eu(1)-O(8) 133.4(2) O(5)-Eu(1)-O(3) 74.82(19) O(7)-Eu(1)-O(8) 71.3(2) O(4)-Eu(1)-O(3) 77.80(19) O(6)-Eu(1)-O(8) 124.25(19)

148

Table 4.15. Selected bond lengths (Å) and angles (deg) for TbsLDS

TbsLDS Tb(1)-O(2) 2.292(6) O(2)-Tb(1)-O(6) 71.9(2) Tb(1)-O(1) 2.304(6) O(1)-Tb(1)-O(6) 144.2(2) Tb(1)-O(4) 2.326(6) O(4)-Tb(1)-O(6) 73.6(2) Tb(1)-O(5) 2.341(6) O(5)-Tb(1)-O(6) 76.5(2) Tb(1)-O(3) 2.409(6) O(3)-Tb(1)-O(6) 73.2(2) Tb(1)-O(6) 2.462(8) O(2)-Tb(1)-O(7) 76.1(2) Tb(1)-O(7) 2.463(7) O(1)-Tb(1)-O(7) 74.7(2) Tb(1)-O(8) 2.482(7) O(4)-Tb(1)-O(7) 142.1(3) O(2)-Tb(1)-O(1) 143.3(3) O(5)-Tb(1)-O(7) 74.0(2) O(2)-Tb(1)-O(4) 101.9(2) O(3)-Tb(1)-O(7) 128.1(2) O(1)-Tb(1)-O(4) 88.2(2) O(6)-Tb(1)-O(7) 136.9(2) O(2)-Tb(1)-O(5) 87.6(2) O(2)-Tb(1)-O(8) 74.6(2) O(1)-Tb(1)-O(5) 104.9(2) O(1)-Tb(1)-O(8) 75.4(2) O(4)-Tb(1)-O(5) 143.7(3) O(4)-Tb(1)-O(8) 71.8(3) O(2)-Tb(1)-O(3) 144.2(2) O(5)-Tb(1)-O(8) 143.9(2) O(1)-Tb(1)-O(3) 72.4(2) O(3)-Tb(1)-O(8) 133.9(2) O(4)-Tb(1)-O(3) 75.1(2) O(6)-Tb(1)-O(8) 124.5(2) O(5)-Tb(1)-O(3) 77.0(2) O(7)-Tb(1)-O(8) 71.3(2)

149

7.978 6.757 4.846 3.310 2.993 2.861 2.656 2.233

0.44 2.94 9.49 7.20 1.29 1.30 3.44 9.00

8.0 7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 4.28. H NMR of as-synthesized EusLCO (7.9 mg) in CD3OD.

150

6.758 4.847 3.310 2.234

2.97 4.28 6.76 9.00

7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 4.29. Control experiments for H NMR of EusLCO (only mesitylene in CD3OD).

151

7.971 6.751 4.840 3.304 2.985 2.854 2.649 2.226

0.88 3.01 5.24 4.14 2.48 2.44 1.29 9.00

8.0 7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 4.30. H NMR of as-synthesized EusLDS (8.8 mg) in CD3OD

152

7.971 6.750 4.840 3.301 2.986 2.854 2.650 2.226

0.79 2.98 5.86 4.08 2.47 2.44 1.49 9.00

8.0 7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 4.31. H NMR of as-synthesized TbsLDS (8.1 mg) in CD3OD.

153

6.750 4.841 3.304 2.226

3.02 1.70 3.42 9.00

7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 4.32. Control experiments for H NMR of EusLDS and TbsLDS (only mesitylene in CD3OD).

154

6.769 2.831 2.223 2.050

3.06 8.83 9.00 3.52

7.0 6.0 5.0 4.0 3.0 2.0 ppm (t1)

1 Figure 4.33. H NMR of water-soaked EusLDS.

155

EusL as-synthesized M

EusL calculated M

5 10 15 20 2 (deg)

Figure 4.34. PXRD of as-synthesized EusLM and the simulated pattern from crystal structure.

TbsL as-synthesized M

TbsL calculated M

5 10 15 20 2 (deg)

Figure 4.35. PXRD of as-synthesized TbsLM and the simulated pattern from crystal structure.

156

TbsL as-synthesized OM

TbsL calculated OM

5 10 15 20 25 30 2 (deg)

Figure 4.36. PXRD of as-synthesized TbsLOM and the simulated pattern from crystal structure.

EusL calculated CHO

EusL calculated CHO

5 10 15 20 25 30 2 (deg)

Figure 4.37. PXRD of as-synthesized EusLCHO and the simulated pattern from crystal structure.

157

EusL as-synthesized CO

EusL calculated CO

5 10 15 20 25 30 35 2 (deg)

Figure 4.38. PXRD of as-synthesized EuLCO and the simulated pattern from crystal structure.

EusL as-synthesized DS

EusL calculated DS

5 10 15 20 25 30 2 (deg)

Figure 4.39. PXRD of as-synthesized EusLDS and the simulated pattern from crystal structure.

158

TbsL as-synthesized DS

TbsL calculated DS

5 10 15 20 25 30

2 (deg)

Figure 4.40. PXRD of as-synthesized TbsLDS and the simulated pattern from crystal structure.

159

References

1. H.-L. Jiang, D. Feng, T.-F. Liu, J.-R. Li, H.-C. Zhou, J. Am. Chem. Soc. 2012, 134, 14690-14693. 2. Y. Li, S. Zhang, D. Song, Angew. Chem. Int. Ed. 2013, 52, 710-713. 3. P. F. Schuda, C. B. Ebner, S. J. Potlock, Synthesis 1987, 12, 1055-1057. 4. P. Gao, X. Feng, X. Yang, V. Enkelmann, M. Baumgarten, K. Müllen, J. Org. Chem. 2008, 73, 0297-0213. 5. Z. Peng, M. E. Galvin, Acta Polymer. 49, 244-247 6. G. Liu and B. Jacquier, eds. Spectroscopic properties of rare earths in optical materials, 2005, Springer. 7. G. S. Ofelt, J. Chem. Phys. 1963, 38, 2171. 8. (a) K. Suzuki, M. Sugawa, Y. Kikukawa, K. Kamata, K. Yamaguchi, N. Mizuno, Inorg. Chem. 2012, 51, 6953-6961; (b) J.-G. Kang, S.-K. Yoon, Y. Sohn, J.-G. Kim, Y.-D. Kim, I.-H. Suh, J. Chem. Soc., Dalton Trans. 1999, 1467-1473; (c) J. Zhou, L. An, X. Liu, H. Zou, F. Hu, C. Liu, Chem. Commun., 2012, 48, 2537-2539; (d) W.-J. Liu, N. Wang, Z.-Q. Wei, X. Zhao, L.-M. Chang, S.-T. Yue, Y.-L. Liu, Y.-P. Cai, Inorg. Chem. Commun. 2011, 14, 1807-1814; (e) Z.-P. Deng, W. Kang, L.-H. Huo, H. Zhao, S. Gao. Dalton Trans., 2010, 39, 6276-6284; (f) L. Natrajan, J. Pecaut, M. Mazzanti, C. LeBrun, Inorg. Chem. 2005, 44, 4756-4765. 9. W. D. Horrocks Jr., D. R. Sudnick, J. Am. Chem. Soc. 1979, 101, 334-340. 10. M. Steppert, I. Císařová, T. Fanghänel, A. Geist, P. Lindqvist-Reis, P. Panak, P. Štĕpnička, S. Trumm and C. Walther, Inorg. Chem. 2012, 51, 591-600. 11. R. M. Supkowski, W. D. Horrocks, Jr. Inorg. Chim. Acta. 2002, 340, 44-48.

160

Chapter 5 Lanthanide MOFs Synthesized from Ligands with Cationic Groups 5 5.1 Abstract

In this chapter, a ligand with charged quaternary ammonium groups (Figure 5.1) have been synthesized with different counter anions, and three different MOF structures have been obtained from the ligand with chloride and nitrate anion. The nitrate anion can be partially exchanged and significant luminescence property change was observed.

+ N Et3

O OH

HO O

+ Et3N

Figure 5.1. Ligand discussed in this chapter. 5.2 Introduction

There have been very few examples of MOF sensors utilizing charge interaction. Ghosh and co- workers have reported a 1D cationic Zn MOF synthesized from (NE,N’E)-4,4’-(ethane-1,2- diyl)bis(N-(pyridin-2-ylmethylene)aniline) and zinc nitrate (Figure 5.2).[1] The framework carries net positive charge due to the neutral ligand and nitrate anions are located in the channels. - - - - These nitrate can be exchanged with N3 , SCN , ClO4 and N(CN)2 to trigger turn-on or turn-off of the emission. The observed changes were attributed to a combination of various effects including electronic interactions of anions with the framework walls or metal centers.

Dong group also reported a 2D cationic MOF sensor synthesized from copper nitrate and 4,4’- (9,9-dibutyl-9H-fluorene-2,7-diyl)dipyridine (Figure 5.3).[2] In the framework, Cu2+ are interconnected by the linear bipyridine-type ligand to form 2D grids, and the axial positions of Cu2+ are occupied by nitrate and water. Nitrate can be exchanged with other anions to trigger visible color change. However, no mechanism is given for the color change.

161

Figure 5.2. Ligand (a) used in Ghosh’s study and the MOF structure: (b) Zn node; (c) 1D chains; (d) Channels and the nitrates located in them.

162

Figure 5.3. Ligand (top) and the 2D grids in the MOF (bottom) of built from 4,4’-(9,9-dibutyl-9H-fluorene-2,7- diyl)dipyridine and copper ion.

In our previous studies, different π-conjugated side chains have been incorporated into the ligand as a strategy to enhance analyte-ligand interaction in sensing applications. While certain ligands have lead to strong solvent-ligand interaction, non-solvent species (e.g. metal ions) cannot trigger significant change in luminescence. Therefore, we decided to try out a different strategy to enhance the analyte-ligand interaction – through the use of charged ammonium groups on the ligands. The use of cationic framework not only offers the charge interaction between cationic groups and anionic guests, the flexible choices of counter anions may provide extra tunability of framework porosity through the templating effect. Moreover, by using different counter anions to construct the MOF, we can generate binding sites that are favorable for the specific counter anion. If these counter anions can be exchanged later, the resulting MOF may have good selectivity to uptake them, which is desirable for sensing applications.

163

5.3 Experimental section

5.3.1 Materials and methods

Elemental analyses were performed in our Chemistry Department on a PE 2400 C/H/N/S analyzer. Thermogravimetric analyses (TGA) were performed on a TA Instruments SDT Q600 instrument under dinitrogen atmosphere with a heating rate of 10 °C per minute. NMR spectra were recorded on a Varian 400, or a Bruker Avance 400 spectrometer. Both 1H and 13C NMR spectra were referenced and reported relative to the solvent residual signals. Photoluminescence spectra were measured using QuantaMaster 30 system from Photon Technology International (Canada) Inc. equipped with a 75W high intensity Xenon lamp (UXL-75XE), a R928 PMT detector, and a solid sample holder; the incident and detector slit widths are 1 and 0.4 nm, respectively. The powder XRD experiments were performed on PANalytical machine with generator PW1830, control unit PW3710 and the proportional gas detector PW3011, equipped with a Cu sealed tube (λ = 1.54178 Å). Unless otherwise stated, all manipulations were performed in air and all reagents were purchased from commercial sources and used without further purification.

5.3.2 Synthesis of N,N'-((4,4''-bis(methoxycarbonyl)-[1,1':4',1''- terphenyl]-2',5'-diyl)bis(methylene))bis(N,N-diethylethanaminium) bromide (sLN ester) sLB ester (4.0 g, 7.5 mmol) and triethylamine (4.4 mL, 31.5 mmol) was dissolved in CHCl3 (100 mL) and refluxed for 20 h. The solvent and excess triethylamine was removed under vacuum and the remaining solid was recrystallized in CHCl3 to give sLN ester ∙6CHCl3 (7.8 g, 70% yield) as a 1 white solid. H NMR (DMSO-d6, 400 MHz, 25°C )  8.15 (d, J = 8.8 Hz, 4H), 7.76 (s, 2H), 7.72 (d, J = 8.0 Hz, 4H), 4.63 (s, 4H), 3.92 (s, 6H), 3.07 (q, J = 7.2 Hz, 12H), 0.95 (t, J = 7.0 Hz, 13 18H). C NMR (DMSO-d6, 100 MHz, 25°C )  165.73, 143.25, 142.98, 137.36, 130.35, 129.82,

129.16, 127.18, 55.83, 52.32, 52.20, 7.31. Anal.calcd for C36H50Br2N2O4∙ 2.5CHCl3: C, 44.76; H, 5.12; N, 2.71. Found: C, 44.59; H, 5.36; N, 2.73.

5.3.3 Synthesis of 2',5'-bis((triethylammonio)methyl)-[1,1':4',1''- terphenyl]-4,4''-dicarboxylate hydrochloride (sLN-HCl)

NaOH (2.4 g, 60 mmol) was dissolved in H2O (40 mL) and sLN ester (6.0 g, 4.6 mmol) was added and stirred at room temperature for 2 h. The mixture was then neutralized with 1 M HCl

164 aqueous solution to pH = 7, and the white precipitate was collected by filtration and then washed 1 with small amount of acetone to give the desired product sLN-HCl ∙ 5H2O (2.2 g, yield 71%). H

NMR (Methanol-d4, 400 MHz, 25°C )  8.20 (d, J = 8.4 Hz, 4H), 7.77 (s, 2H), 7.60 (d, J = 8.4 Hz, 13 4H), 4.72 (s, 4H), 3.15 (q, J = 7.1 Hz, 12H), 1.08 (t, J = 7.2 Hz, 18H). C NMR (Methanol-d4, 100 MHz, 25°C )  171.58, 146.10, 143.13, 138.70, 136.22, 131.73, 130.99, 129.00, 57.62, 54.28, 1 8.03. Anal.calcd for C34H45ClN2O4∙ 5H2O (see H NMR): C, 60.84; H, 8.26; N, 4.17. Found: C, 60.38; H, 7.84; N, 4.14.

5.3.4 Synthesis of N,N'-((4,4''-dicarboxy-[1,1':4',1''-terphenyl]-2',5'- diyl)bis(methylene))bis(N,N-diethylethanaminium) nitrate (sLN- HNO3)

NaOH (2.8 g, 70 mmol) was dissolved in H2O (70 mL) and sLN ester (3.5 g, 2.7 mmol) was added and stirred at room temperature for 4 h. The mixture was then neutralized with 1 M HNO3 aqueous solution to pH = 1, and the white precipitate was collected by filtration and then washed with small amount of ethanol to give the desired product of sLN-HNO3 ∙ 3H2O (1.6 g, yield 81%). 1 H NMR (DMSO-d6, 400 MHz, 25°C )  13.25 (s, 2H), 8.13 (d, J = 8.0 Hz, 4H),7.75 (s, 2H), 7.67 (d, J = 8.0 Hz, 4H), 4.62 (s, 4H), 3.06 (q, J = 7.2 Hz, 12H), 0.95 (t, J = 6.8 Hz, 18H). 13C

NMR (DMSO-d6, 100 MHz, 25°C )  166.82, 143.10, 142.87, 137.36, 130.36, 130.17, 129.98,

127.17, 55.38, 52.18, 7.26. Anal.calcd for C33H44N4O10 ∙ 2.3H2O: C, 57.34; H, 7.16; N, 7.87. Found: C, 57.76; H, 7.50; N, 7.83.

5.3.5 N,N'-((4,4''-dicarboxy-[1,1':4',1''-terphenyl]-2',5'- diyl)bis(methylene))bis(N,N-diethylethanaminium) trifluoromethanesulfonate (sLN-HSO3CF3)

NaOH (0.4 g, 10 mmol) was dissolved in H2O (10 mL) and sLN ester (0.5 g, 4.6 mmol) was added and stirred at room temperature for 2 h. The mixture was then neutralized with 1 M aqueous triflic acid to pH = 7, and the white precipitate was collected by filtration and then washed with small amount of acetone to give the desired product sLN-HSO3CF3 ∙ 2H2O (0.20 g, 1 yield 66%). H NMR (DMSO-d6, 400 MHz, 25°C )  13.24 (s, 2H), 8.13 (d, J = 8.4 Hz, 4H), 7.76 (s, 2H), 7.67 (d, J = 8.4 Hz, 4H), 4.62 (s, 4H), 3.06 (q, J = 7.2 Hz, 12H), 0.95 (t, J = 7.2 Hz, 13 18H). C NMR (DMSO-d6, 100 MHz, 25°C )  166.90, 143.20, 142.90, 137.45, 130.50, 130.23,

165

130.07, 127.26, 55.95, 52.29, 7.35. Anal.calcd for C36H46F6N2O10S2 ∙ 2H2O: C, 49.08; H, 5.72; N, 3.18. Found: C, 49.58; H, 5.63; N, 3.19.

5.3.6 Synthesis of N,N'-((4,4''-dicarboxy-[1,1':4',1''-terphenyl]-2',5'- diyl)bis(methylene))bis(N,N-diethylethanaminium) perchlorate (sLN-HClO4)

NaOH (1.2 g, 30 mmol) was dissolved in H2O (20 mL) and sLN ester (1.5 g, 1.2 mmol) was added and stirred at room temperature for 4 h. The mixture was then neutralized with 1 M HCl aqueous solution to pH = 1, and excess of 1 M NaClO4 aqueous solution was added. The white precipitate of sLN-HClO4 ∙ 2H2O was collected by filtration and then dried at 60 °C to give the 1 desired product (0.75 g, yield 83%). H NMR (DMSO-d6, 400 MHz, 25°C )  13.24 (s, 2H), 8.13 (d, J = 8.4 Hz, 4H), 7.76 (s, 2H), 7.67 (d, J = 8.4 Hz, 4H), 4.62 (s, 4H), 3.06 (q, J = 7.2 Hz, 12H), 13 0.95 (t, J = 6.8 Hz, 18H). C NMR (DMSO-d6, 100 MHz, 25°C )  166.89, 143.19, 142.91,

137.44, 130.46, 130.24, 130.06, 127.25, 55.94, 52.28, 7.35. Anal.calcd for C34H52Cl2N2O15 ∙

2H2O: C, 52.24; H, 6.45; N, 3.58. Found: C, 52.14; H, 6.54; N, 3.40.

5.3.7 Synthesis of [Eu2(sLN)3Cl6] (EusLN-Cl)

Eu(NO3)3∙6H2O (8.9 mg, 0.02 mmol), sLN-HCl (20.2 mg, 0.03 mmol) was suspended in a mixture of DMF (3 mL), HCl (0.25 mL, 0.122 M) and THF (2 mL) in a 20 mL scintillation vial and sealed. The vial was then heated in oil bath at 60 °C for 5 d. Colorless crystals of EusLN-Cl (20.1 mg) were collected via filtration and washed with THF and air dried.

5.3.8 Synthesis of [Eu2(sLN)4(NO3)6] (EusLN-NO3)

Eu(NO3)3∙6H2O (8.9 mg, 0.02 mmol), sLN-HNO3 (21.4 mg, 0.03 mmol) was dissolved in a mixture of DMF (3 mL) and ethanol (2 mL) in a 20 mL scintillation vial and sealed. The vial was then heated in oil bath at 70 oC for 5d. Colorless crystals of EusLN were collected via filtration and washed with ethanol (8.1 mg, yield 33% – assuming MW 3305.94 according to EA/TGA).

Anal.calcd for C136Eu2H176O16N8 ∙ 14H2O ∙ 4C2H6O ∙ 0.2C3H7NO (TGA solvent weight loss Anal. 13.7 %, Found 12.4 %)

166

5.3.9 X-ray crystallographic analysis

X-ray quality single crystals were obtained as described in the sections above. All crystals were mounted on the tip of a MiTeGen MicroMount. The single-crystal X-ray diffraction data were collected on a Bruker Kappa Apex II CCD diffractometer with Mo Kα radiation (λ = 0.71073 Å) operating at 50 kV and 30 mA, at 150 K controlled by an Oxford Cryostream 700 series low temperature system. The data integration and absorption correction were performed with the Bruker Apex 2 software package. All the structures were solved by direct methods and refined using SHELXTL V6.14. All non-hydrogen atoms except for the atoms involved in the disordered portions were refined anisotropically. The positions of the hydrogen atoms were calculated using the riding model.

5.3.10 Determination of solvent composition of different MOF samples

The solvent compositions of MOFs are determined by a combination of different characterizations, including x-ray crystallography, elemental analysis (EA), thermogravimetric analysis (TGA) and 1H NMR, all performed on the same batch of sample. The use of different deuterated solvent (methanol-d4 or acetone-d6) produces consistent results in all cases

5.3.11 1H NMR experiments procedures for MOF samples

In a typical 1H NMR experiment, 6~10 mg of MOF sample is weighted and suspended in deuterated solvent (0.7 mL of methanol-d4 or acetone-d6), and mesitylene (2.5 μL) is added as internal standard. A control group is set up under the same condition except that MOF sample is not added. The NMR sample is then sonicated for a few minutes and the NMR spectra is taken with 25s relaxation delay to ensure the accuracy of integration. The amount of water and organic solvent were determined based on the ratios of their integration to that of mesitylene; water content is calculated from H2O and HDO (when acetone-d6 is used) signals, further adjusted by subtracting the amount of water in the control group to exclude the water introduced from the deuterated solvent, moisture or mesitylene. The exchange between water and methanol-d4 was found to have no significant impact in estimating the water content according to our control experiment. Because of the errors inherited in integration and the fact that the observed solvent signal may only reflect the “exchangeable” portion of the solvent in the MOF, such numbers are used as qualitative rather than quantitative results unless matched with other characterization methods (TGA, EA and etc.).

167

5.3.12 Luminescence measurements

Photoluminescence spectra were measured using QuantaMaster 30 system from Photon Technology International (Canada) Inc. equipped with a 75W high intensity Xenon lamp (for fluorescence), a Xenon short arc lamp (for phosphorescence), a R928 PMT detector, and a solid sample holder; the incident and detector slit widths are 1 and 0.4 nm for fluorescence measurements, 8 and 8 nm for phosphorescence, respectively. Solvent dependent luminescence was measured with the suspension of the powder MOF samples in a standard quartz cuvette containing 3.0 mg of the MOF powder and 3.0 mL of solvent. The suspension was stirred throughout the measurement and the emission intensity was found to be stable.

5.3.13 Anion exchange experiment

EusLN-NO3 (~50 mg) was soaked in KClO4 DMF solution (0.24 M, 5 mL) for 7 days and the solution was replaced with fresh solution once after 4 days. After 7 days, the MOF sample was recovered by filtration, washed with DMF (DMF washed sample) or with DMF and then ethanol (EtOH washed sample) and air dried.

5.4 Results and discussion

5.4.1 Ligand synthesis

The sLN ester can be synthesized with ease from the reaction between sLB ester (Chapter 3) and triethylamine in CHCl3. As CHCl3 are removed from the resulting mixture after the reaction, sLN ester crystallized out with four CHCl3 nicely as white crystals, which can be collected with filtration without further purification. Crystal structure of sLN ester was obtained via X-ray crystallography to confirm the existence of bromide anion (Figure 5.4), and the crystallographic table attached at the end of this chapter.

168

Figure 5.4. Crystal structure of sLN ester.

The sLN ester can be hydrolyzed with NaOH in aqueous solution in a few hours, yielding a clear solution of sLN salt. At this stage, different acid can be used to adjust the pH to yield the sLN ligand precursor with different counter anions. Ligand precursor with nitrate (sLN-HNO3) or perchlorate counter anion (sLN-HClO4) can be obtained by acidifying the basic solution to pH =

1; compound sLN-HNO3 and sLN-HClO4 crystallized out in fairly good yield. To obtain the ligand precursor with chloride anion, the basic sLN salt solution needs to be acidified slowly to about pH = 7. Excess of acid will dissolve the precipitate of sLN-HCl resulting in a low yield, which indicates that sLN-HCl is the mono-protonated version of the ligand. When triflic acid is used, the most precipitate is observed around pH = 7, the amount of which decreases as more acid is added, but the precipitate do not disappear completely even at pH = 1. Elementary analysis confirmed that the mono-protonated version forms at pH = 7 and the fully protonated version forms at lower pH. Due to the poor solubility of the mono-protonated version, only the fully protonated version is reported here. X-ray structure of sLN-HNO3 is obtained (Figure 5.5), from which the nitrate counter anions can be clearly identified (crystallographic table is attached at the end of this chapter). We were not able to obtain crystals of sLN-HCl, sLN-HClO4 or sLN-

HSO3CF3 with good enough quality for single crystal X-ray diffraction.

169

Figure 5.5. Crystal structure of sLN-HNO3. 5.4.2 MOF synthesis and structure

Three different MOFs have been synthesized with solvothermal method by using sLN ligand with chloride and nitrate anions; no crystals have been obtained with sLN-HClO4 or sLN-HSO3CF3. Due to the charge on the side chain, the zwitterionic form of the ligand is very soluble in polar solvent which renders it a bad ligand for MOF synthesis. We have attempted a lot of different solvent combinations before we realized that the amount of polar solvent used needs to be minimized in order for MOFs to form. All three MOFs are also soluble in water and some other polar solvents, despite two of them being 3D structures. Among the three structures: EusLN-NO3,

EusLN-Cl(a) and EusLN-Cl(b), the first two structures are poor in quality, partly due to the huge unit cell.

170

Figure 5.6. Crystal structure of EusLN-NO3: (a) & (b) Eight-coordinate Eu center adopting bicapped trigonal prismatic geometry; (c) Bimetallic node bridged by four carboxylate; (d) Twisted tetrahedral geometry of each node, connected to four other nodes (ammonium side chain omitted for clarity); (e) Diamond shaped 1D channels along the c-axis; (f) Two-fold interpenetration in the framework.

171

Figure 5.7. Crystal structure of EusLN-Cl(a): (a) & (b) Eight-coordinate Eu center adopting bicapped trigonal prismatic geometry; (c) Bimetallic node bridged by four carboxylate; (d) Twisted tetrahedral geometry of each node, connected to four other nodes (ammonium side chain omitted for clarity); (e) Diamond shaped 1D channels along the c-axis; (f) Two-fold interpenetration in the framework.

172

Figure 5.8. Crystal structure of EusLN-Cl(b): (a) & (b) Eight-coordinate Eu center adopting square anti-prismatic geometry; (c) Bimetallic node bridged by four carboxylate; (d) 2D grid in the MOF; (e) 2D grids stack on top of each other with an off set of half grid; (f) Two-fold interpenetration in the framework.

173

EusLN-NO3 crystallized in orthorhombic space group Fddd. Each eight-coordinate Eu center adopts bicapped trigonal prismatic geometry, with O1, O3, O4 and O5, O7 O8 defining the two trigonal plane. All oxygen donor atoms are from the carboxylate group of the ligands thus no coordinated solvent molecule was present, which is rarely seen in lanthanide MOFs constructed from carboxylate ligands. Unlike most other 3D MOFs reported in this series (in Chapter 3 and 4) where metal centers are bridged by carboxylate groups to form 1D chains, bimetallic nodes were found in the structure, with four bridging carboxylate groups between the two metal centers and two terminal bidentate carboxylate on each metal, respectively (Figure 5.6c). These bimetallic nodes are interconnected with four other nodes in a distorted tetrahedral geometry to form a 3D network. Two independent networks exist in the crystal structure(Figure 5.6f), interpenetrating each other while still yielding a highly porous structure. As shown in Figure 5.6e, diamond shaped channels are running along the a-axis with the quaternary ammonium side-chain exposed in the channels. The channel is formed by the helix of the two interpenetrating networks in alternating order. Some of the nitrate counter ions were found according to the electron density map, located in various positions in the channels with occupancy lower than 1. Disordered lattice ethanol, water and likely small amount of DMF molecules occupy the rest of the channel space as suggested by 1H NMR, elemental analysis and TGA. The purity of the bulk material is confirmed by PXRD (Figure 5.9).

Two different crystal structures (EusLN-Cl(a) and (b)) have been obtained from the crystals of

EusLN-Cl. EusLN-Cl(a) (Figure 5.7) displayed similar topology and connectivity as EusLN-NO3. It crystallized in orthorhombic space group Fddd. Each Eu center is eight-coordinate, adopting bicapped trigonal prismatic geometry. Same as EusLN-NO3, distorted tetrahedral bimetallic node is found in EusLN-Cl(a) and it is connected to four adjacent nodes through the linear ligands.

EusLN-Cl(a) also possesses the 1D channel along a-axis which is built from the helix of the two interpenetrating frameworks. Chloride and solvents likely occupy the channel space. However, they are severely disordered and cannot be modeled from the electron density map.

EusLN-Cl(b) crystallized in triclinic space group P-1. Each Eu center adopts the square anti- prismatic geometry with two square plane defined by O3, O4, O5, O6 and O1, O2, O7, O8 respectively. Among all eight coordinated oxygen, six are from carboxylate groups and two are

174 from coordinated DMF molecules. All Eu-O bond lengths are within the range of 2.324(3) to 2.488(3) Å. Two Eu ions are bridged by four carboxylate to form a bimetallic node, which was interconnected to four other nodes by linear linkers to form a 2D grid with a big hole in each grid unit (Figure 5.8d). The adjacent grids stack with an offset of a half grid in the crystal structure, blocking most of the opening of the grid. As a result, this structure is not as porous as the EusLN- Cl(a). Solvent molecules and chloride located in the channels are disordered and cannot be located crystallographically.

EusL -NO as-synthesized N 3

EusL -NO calculated N 3 2 4 6 8 10 12 14 16 18 20 2 (deg)

Figure 5.9. PXRD of EusLN-NO3

175

EusL -Cl(b) calculated N

EusL -Cl(a) as-synthesized N

EusL -Cl(a) calculated N

4 6 8 10 12 14 16 18 20 2 (deg)

Figure 5.10. PXRD of EusLN-Cl

PXRD of EusLN-Cl crystals shows the existence of both EusLN-Cl(a) and EusLN-Cl(b) phase in bulk sample (Figure 5.10), as opposed to the pure EusLN-NO3. Such difference might due to the anion templating effect: bigger nitrate ions can interact strongly with solvent molecules, stabilizing the porous structure, while such effect is weaker when chloride is the counter anion, leading to the formation of a second phase that is less porous.

5.4.3 Thermogravimetric analysis

TGA of EusLN-NO3 and its ligand sLN-HNO3 are shown in Figure 5.11. The thermal stability of both the ligand and MOF is dominated by the poor thermal stability of the ammonium side chain.

The first 5.5% weight loss of sLN-HNO3 starts at room temperature and completes at around 100 °C, corresponding to the loss of lattice water. The ligand is very stable thereafter until

200 °C when the ammonium side chain of the ligand started to decompose. EusLN-NO3 follows similar decomposition pattern: solvent loss starts at room temperature and completes around 200 °C, and ligand decomposition starts around 220 °C.

176

100 sL -HNO N 3 90 EusL -NO N 3

80

70

60 Weight (%) 50

40

30

0 100 200 300 400 500 600 700 800 Temperature (oC)

Figure 5.11. TGA of sLN-HNO3 and EusLN-NO3.

160 EusL -NO em @330nm N 3 EusL -NO ex @615nm 140 N 3

120

100

80

60

Intensity(A.U.) 40

20

0

250 300 350 400 450 500 550 600 650 700 Wavelength (nm)

Figure 5.12. Luminescent spectrum of EusLN-NO3. The peaks at 308 and 660 nm on the excitation and emission spectrum, respectively, are due to scattering.

177

5.4.4 Luminescent properties

EusLN-NO3 display strong Eu emission upon excitation at 330 nm, as shown in Figure 5.12. The 5 7 7 7 emission peaks at 577, 591, 615, and 699 nm correspond to transitions from D0 to F0, F1, F2, 7 and F4, respectively. The small peak at 395 nm in the excitation spectrum can be assigned to the 7 5 metal-centered transition from F0,1 to L6. The emission spectra of EusLN-NO3 suspension in different solvent show no significant solvent dependent luminescence.

5.4.5 Anion exchange experiments

Because of the inclusion of counter anion in EusLN-NO3, it is very interesting to see the effect of anion exchange on the MOF properties. As EusLN-NO3 dissolves in water and methanol, DMF solution of salts is used, which limits the anions that can be tested due to solubility. NaI was first tested for anion exchange but iodide did not effectively exchange nitrate out, as suggested by the - unchanged υ(NO3 ) intensity in FTIR experiments. Such observation may be due to the hydrogen bonding interaction between nitrate and framework or channel solvent. Therefore, we decided to test out perchlorate which has similar properties.

By soaking EusLN-NO3 in DMF solution of KClO4, significant amount of nitrate is exchanged - -1 by perchlorate, as indicated by FTIR (Figure 5.13): υ(NO3 ) at 1300 and 1400 cm is weakened - -1 [3,4] 1 and υ(ClO4 ) at 1100 cm is enhanced. H NMR and EA experiments were further conducted to check if nitrate is complete removed. For the KClO4 exchanged sample washed with ethanol (to remove possible DMF trace that affect N composition in EA), DMF is almost completely removed (<0.1 DMF per Eu; 1H NMR spectrum provided in Figure 5.15 at the end of the chapter). However, EA result (C, 47.44; H, 6.47;N, 4.23) shows much higher N content than that when nitrate is completely removed (C, 49.13; H, 6.16; N, 3.37; calculated according to the solvent composition indicated by 1H NMR). Such incomplete exchange demonstrated strong interaction between nitrate and the framework or channel solvents. Sulfate or phosphate might be better candidates to exchange nitrate; however they cannot be tested under current condition due to the poor solubility of their salt in DMF.

178

KClO exchanged 4

- (ClO ) 4

EusL -NO as-synthesized N 3 - (NO ) 3

4000 3500 3000 2500 2000 1500 1000 500 Wavelength (cm-1)

Figure 5.13. FTIR of EusLN-NO3 before and after KClO4 soaking.

240 Em-Before exchange 220 Em-After exchange (EtOH) Em-After exchange (DMF) 200 Ex-Before exchange 180 Ex-After exchange (EtOH) 160 Ex-After exchange (DMF)

140

120

100

80

Intensity(A.U.) 60

40

20

0

250 300 350 400 450 500 550 600 650 700 Wavelength (nm)

Figure 5.14. Excitation (ex) and emission (em) spectra of perchlorate exchanged EusLN-NO3 (solid) washed by EtOH or DMF.

179

Luminescence spectra of the anion exchanged sample have been obtained to evaluate the impact of perchlorate (Figure 5.14). Eu emission intensity of the EtOH washed sample showed 70% decrease, while that of DMF washed sample showed 30% increase. The suspension of the anion exchanged MOF in various solvent did not show solvent dependent luminescence however, suggesting that the solvent identity has no impact on the Eu emission. The difference in emission between the DMF/EtOH washed samples is likely due to the different amount of solvent uptake.

In fact, both samples were found to contain approximately 13 H2O per Eu2(sLN)3 unit, but only the DMF washed sample contains approximately 3 DMF.

5.5 Summary

We have successfully synthesized sLN ligands with different counter ions, and assembled these ligands into three MOF structures with chloride and nitrate as the counter ions. Nitrate ion provides the templating effect which makes the more porous framework favorable while chloride does not have such effect. All three MOFs obtained showed characteristic Eu emission, however no solvent dependent luminescence was observed. Nitrate ion in EusLN-NO3 can be partially exchanged with prolonged soaking in KClO4 DMF solution and the resulting MOF showed enhanced/weakened emission depending on the solvent used.

180

Table 5.1. Crystallographic data of sLN ester, sLN-HNO3, EusLN-NO3, EusLN-Cl(a), and EusLN- Cl(b).

sLN ester sLN-HNO3 EusLN-NO3 EusLN-Cl(a) EusLN-Cl(b)

formula C21H28BrCl9NO2 C34H50N4O12 C51H66N4O9Eu C68H88N4O8Eu C57H80N5O8Eu ∙ ∙ 1.65H2O 7H2O ∙ C3H7NO

F.W. 725.40 706.78 1060.76 969.05 648.65 T (K) 150(2) 150(2) 150(2) 150(2) 150(2) space group P-1 P2(1)/n Fddd Fddd P-1 a (Å) 9.6427(6) 9.0991(4) 30.880(2) 31.3583(16) 12.359(4) b (Å) 11.1943(8) 15.6841(8) 43.309(3) 43.200(2) 14.285(4) c (Å) 16.4296(13) 12.3651(5) 59.581(4) 59.445(3) 20.011(6)  (deg) 79.865(3) 90 90 90 82.329(7)  (deg) 73.083(3) 96.310(2) 90 90 75.676(7)  (deg) 66.595(2) 90 90 90 86.191(7) V (Å3) 1553.43(19) 1753.95(14) 79683(9) 80529(7) 3390.2(17) Z 2 4 32 32 2

−3 Dc (g·cm ) 1.551 1.338 0.865 0.813 1.271  (mm−1) 2.117 0.102 0.671 0.658 0.990 no. reflns 26010 12072 105978 85587 84559 collcd no. indept 7121 3073 22813 18688 28105 reflns GOF on F2 1.075 1.034 1.370 1.299 1.108

R [I > 2σ(I)] R1 = 0.0773 R1 = 0.0427 R1 = 0.1126 R1 = 0.1150 R1 = 0.0634

wR2 = 0.2067 wR2 = 0.1008 wR2 = 0.3543 wR2 = 0.3698 wR2 = 0.1932

R (all data) R1 = 0.0874 R1 = 0.0594 R1 = 0.1936 R1 = 0.1600 R1 = 00721

wR2 = 0.2163 wR2 = 0.1093 wR2 = 0.4089 wR2 = 0.3927 wR2 = 0.2030

181

Table 5.2. Selected bond lengths (Å) and angles (deg) for EusLN-NO3

EusLN-NO3 Eu(1)-O(5) 2.312(6) O(2)#1-Eu(1)-O(3) 76.5(2) Eu(1)-O(1) 2.320(7) O(5)-Eu(1)-O(7) 89.7(2) Eu(1)-O(6)#1 2.375(7) O(1)-Eu(1)-O(7) 147.4(2) Eu(1)-O(2)#1 2.388(6) O(6)#1-Eu(1)-O(7) 75.9(3) Eu(1)-O(3) 2.440(6) O(2)#1-Eu(1)-O(7) 134.1(2) Eu(1)-O(7) 2.435(7) O(3)-Eu(1)-O(7) 91.0(3) Eu(1)-O(8) 2.506(7) O(5)-Eu(1)-O(8) 75.7(3) Eu(1)-O(4) 2.517(6) O(1)-Eu(1)-O(8) 156.6(2) Eu(1)-C(63) 2.787(10) O(6)#1-Eu(1)-O(8) 121.6(2) Eu(1)-C(28) 2.814(9) O(2)#1-Eu(1)-O(8) 80.6(2) Eu(1)-Eu(1)#1 4.3362(10) O(3)-Eu(1)-O(8) 79.3(2) O(5)-Eu(1)-O(1) 107.3(3) O(7)-Eu(1)-O(8) 53.6(2) O(5)-Eu(1)-O(6)#1 76.7(3) O(5)-Eu(1)-O(4) 156.4(2) O(1)-Eu(1)-O(6)#1 81.2(2) O(1)-Eu(1)-O(4) 77.5(2) O(5)-Eu(1)-O(2)#1 80.4(2) O(6)#1-Eu(1)-O(4) 81.4(2) O(1)-Eu(1)-O(2)#1 77.2(2) O(2)#1-Eu(1)-O(4) 122.8(2) O(6)#1-Eu(1)-O(2)#1 142.0(3) O(3)-Eu(1)-O(4) 52.5(2) O(5)-Eu(1)-O(3) 148.3(3) O(7)-Eu(1)-O(4) 76.5(2) O(1)-Eu(1)-O(3) 88.4(2) O(8)-Eu(1)-O(4) 109.3(3) O(6)#1-Eu(1)-O(3) 133.9(2)

182

Table 5.3. Selected bond lengths (Å) and angles (deg) for EusLN-Cl(a)

EusLN-Cl-big Eu(1)-O(8) 2.326(7) O(8)-Eu(1)-O(1) 88.2(2) Eu(1)-O(6) 2.319(7) O(6)-Eu(1)-O(1) 149.2(2) Eu(1)-O(5) 2.377(6) O(5)-Eu(1)-O(1) 133.7(2) Eu(1)-O(7) 2.392(7) O(7)-Eu(1)-O(1) 76.2(3) Eu(1)-O(3) 2.439(7) O(3)-Eu(1)-O(1) 91.1(3) Eu(1)-O(1) 2.463(6) O(8)-Eu(1)-O(4) 154.6(3) Eu(1)-O(4) 2.514(6) O(6)-Eu(1)-O(4) 75.8(2) Eu(1)-O(2) 2.536(6) O(5)-Eu(1)-O(4) 123.9(2) O(8)-Eu(1)-O(6) 107.0(3) O(7)-Eu(1)-O(4) 79.2(2) O(8)-Eu(1)-O(5) 80.5(2) O(3)-Eu(1)-O(4) 53.9(2) O(6)-Eu(1)-O(5) 76.2(2) O(1)-Eu(1)-O(4) 79.4(2) O(8)-Eu(1)-O(7) 76.4(3) O(8)-Eu(1)-O(2) 78.7(2) O(6)-Eu(1)-O(7) 81.5(2) O(6)-Eu(1)-O(2) 155.1(2) O(5)-Eu(1)-O(7) 141.4(3) O(5)-Eu(1)-O(2) 81.0(2) O(8)-Eu(1)-O(3) 149.3(2) O(7)-Eu(1)-O(2) 123.2(2) O(6)-Eu(1)-O(3) 88.8(2) O(3)-Eu(1)-O(2) 76.6(2) O(5)-Eu(1)-O(3) 78.0(2) O(1)-Eu(1)-O(2) 52.7(2) O(7)-Eu(1)-O(3) 133.0(2) O(4)-Eu(1)-O(2) 109.8(3)

183

Table 5.4. Selected bond lengths (Å) and angles (deg) for EusLN-Cl(b) EusLN-Cl-small Eu(1)-O(4)#1 2.324(3) O(5)-Eu(1)-O(8) 73.02(11) Eu(1)-O(6)#1 2.338(3) O(4)#1-Eu(1)-O(7) 144.95(10) Eu(1)-O(3) 2.357(3) O(6)#1-Eu(1)-O(7) 139.39(11) Eu(1)-O(5) 2.380(3) O(3)-Eu(1)-O(7) 76.15(10) Eu(1)-O(8) 2.444(3) O(5)-Eu(1)-O(7) 76.81(10) Eu(1)-O(7) 2.448(3) O(8)-Eu(1)-O(7) 72.22(10) Eu(1)-O(1) 2.453(3) O(4)#1-Eu(1)-O(1) 82.60(10) Eu(1)-O(2) 2.488(3) O(6)#1-Eu(1)-O(1) 79.23(10) Eu(1)-C(1) 2.818(3) O(3)-Eu(1)-O(1) 134.61(11) Eu(1)-Eu(1)#1 4.4009(9) O(5)-Eu(1)-O(1) 146.53(11) O(4)#1-Eu(1)-O(6)#1 75.59(12) O(8)-Eu(1)-O(1) 75.17(10) O(4)#1-Eu(1)-O(3) 124.15(10) O(7)-Eu(1)-O(1) 103.09(10) O(6)#1-Eu(1)-O(3) 74.44(11) O(4)#1-Eu(1)-O(2) 131.31(10) O(4)#1-Eu(1)-O(5) 79.98(10) O(6)#1-Eu(1)-O(2) 77.25(12) O(6)#1-Eu(1)-O(5) 122.92(11) O(3)-Eu(1)-O(2) 85.19(11) O(3)-Eu(1)-O(5) 78.45(12) O(5)-Eu(1)-O(2) 148.27(10) O(4)#1-Eu(1)-O(8) 76.12(11) O(8)-Eu(1)-O(2) 105.93(11) O(6)#1-Eu(1)-O(8) 143.82(11) O(7)-Eu(1)-O(2) 72.94(10)

184

6.770 2.939 2.814 2.780 2.224 2.050

2.939 2.814 2.780

0.10 4.70

3.00 2.90 2.80 2.70

ppm (f1)

2.94 0.10 4.70 9.00

7.0 6.0 5.0 4.0 3.0 2.0 ppm (f1)

1 Figure 5.15. H NMR of EusLN-NO3 (7.5 mg, in acetone-d6) exchanged with KClO4 DMF solution.

185

6.769 2.801 2.767 2.225 2.223 2.050

2.801 2.767

0.19 0.18

2.950 2.900 2.850 2.800 2.750 2.700 2.650 2.600

ppm (f1)

2.99 0.19 0.18 9.00

7.0 6.0 5.0 4.0 3.0 2.0 ppm (f1)

1 Figure 5.16. H NMR of the control experiment for EusLN-NO3 exchanged with KClO4 DMF solution

186

References

1. B. Manna, A. K. Chaudhari, B. Joarder, A. Karmakar, S. K. Ghosh, Angew. Chem. Int. Ed. 2013, 52, 998. 2. J.-P. Ma, Y. Yu, Y.-B. Dong, Chem. Commun. 2012, 48, 2946-2948. 3. J. Degenhardt, A. J. McQuillan, Langmuir 1999, 15, 4595-4602. 4. S. K. Srinivasan, S. Ganguly, Catal. Lett. 1991, 10, 279-287.

187

Chapter 6 Summary and Outlook 6 6.1 Summary of the thesis

Figure 6.1. Ligand systems discussed in this thesis.

In this thesis, two ligand systems have been discussed (Figure 6.1). LX system has inherent disadvantages as discussed in Chapter 2, including low yielding synthesis and mis-matched

energy level which cannot effectively promote LMET. The alternative sLX system has proved successful. First, the synthetic route is much more efficient: yield of the synthesis can usually reach over 80%, without much column chromatography involved. Secondly, ligand energy level

matches well with the requirement, as all MOFs except for the one with H2sLCO display characteristic lanthanide emission.

Among the MOFs synthesized from the sLX ligand series, different solvent dependent

luminescent behaviors have been observed. EusLM showed no response to solvent, likely due to

the poor porosity. In EusLCHO, the analyte-metal interaction is the factor that dominates the

emission change: D2O and CD3OD suspensions emit the strongest while the emission in other solvent suspension does not differ much from each other. A slow solvent exchange process at the metal site is also observed through the lifetime studies, which may be due to the hydrogen

bonding of water or the limited pore size. Both EusLOM and EusLDS showed strong effect of solvent-ligand interaction on their emission intensity, even when the solvents are present at a low

188 concentration as in the vapor sensing experiments. The difference of their emission intensity is observed in vapor sensing experiments or solvent suspension experiments when H2O or D2O is present in the MOF samples. However, such solvent-metal interaction has only contributed to the emission change by a small factor; the major contribution comes from the solvent-ligand interaction as suggested by the experiments conducted with deuterated solvents. Besides the observation of different types of sensing behavior, the relation between the properties of MOFs and the corresponding ligands has been summarized through the studies of sLX ligand series. It has been found that the thermal stability and the singlet/triplet energy of the ligands can be used to predict the thermal stability and luminescent properties of the respective MOF. All these discoveries will facilitate our future MOF design to better meet the requirement for sensing applications.

Two contributions have also been made by this project in terms of methodology. First, luminescence lifetime study was first introduced into the MOF sensing systems to monitor the solvent-metal interaction. Although this method has been widely used in homogeneous systems or MOF systems to determine the number of coordinated water molecules, it has never been applied to the study of MOF sensing mechanism. This method has proved to be a powerful tool in the study of various solvent sensing systems in previous chapters, to distinguish the solvent- metal interaction from other type of interactions. Secondly, the experimental setup to test MOF powders for vapor sensing is developed. The setup used in testing EusLOM allows us to monitor the luminescence intensity change in real time, to determine the respond rate of the underlying materials. These two methods will help with the future development of MOF sensors in understanding the mechanism and in determining the sensor response rate.

189

6.2 Future work

Figure 6.2. Other ligand candidates for MOF sensors.

Given its convenient synthesis and desirable energy level, sLX system is worth further exploration. A few promising candidate is shown in Figure 6.2. Adding a second set of thioether group on the H2sLDS ligand (Figure 6.2a) may help enhance the interaction with heavy metal ions, as the ligand can now interact with metal ions in a bidentate fashion. Oxidizing the thioether to thionyl group (Figure 6.2b) may enhance the templating effect of solvent as the thionyl group interacts with polar solvents more strongly than thioether. Combination of these two modifications (tetra thionyl group) is also an option, although it will be synthetically more difficult.

The water solubility of EusLN-Cl and EusLN-NO3 is one of the major disadvantages, as aqueous media is the most convenient and common media used in sensing applications. To reduce the solubility of MOFs built from the ligands bearing ammonium groups, reducing the number of ammonium groups per ligand (Figure 6.2c) to one may be a simple option.

190

Figure 6.3. Other alternative ligand systems.

Besides those aforementioned ligands within sLX series, alternative ligand systems may also lead to desirable MOF sensor materials. As shown in Figure 6.3a, the functional groups can be moved to X1 or X2 position. In the typical structure adopted by EusLM, EusLOM, EusLCHO and

EusLDS, the functional groups on the central ring of the ligands are not very well exposed in the 1D channels; moving the functional groups to the terminal ring would solve this problem if the same connectivity is adopted. With the same logic, the central phenyl spacer can be replaced with acetylene spacer (Figure 6.3b). The energy level of a few model ligands within this series may be also calculated first to determine whether it is desirable for LMET process.

6.3 Final remark

The field of MOF sensing is still in its infancy – very limited systematic studies have been developed in this field. As is shared by all material oriented researches, the ultimate goal is to understand the underlying principle from material preparation to structures and from structures to properties. By developing and fully studying different ligand systems, we can contribute to the pool of MOF knowledge which will eventually form the stream of rational design that leads to the vast ocean of functional MOF materials.