Synthesis and Characterization of Novel Imine-Linked Covalent Organic Frameworks

Home , Imine

THESIS

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in

the Graduate School of The Ohio State University

Toni Beirl

Graduate Program in Chemistry

The Ohio State University

2015

Master's Examination

Master's Examination Committee:

Professor Psaras McGrier, Advisor

Professor Jovica D. Badjic

Copyrighted by

Toni M.Beirl

2015

Abstract

Covalent organic frameworks (COFs) are a class of porous crystalline materials composed of light elements (such as H, B, C, N, and O) that are linked by covalent bonds. The modular nature of COFs permits the integration of various π-conjugated molecular building blocks into highly ordered polymeric structures with low densities and high thermal stabilities making them suitable for applications related to energy storage and conversion, catalysis, and gas storage. Since a majority of the early examples of COFs contained boroxine or boronate esters linkages, many of these materials were often susceptible to hydrolysis when exposed to aqueous conditions resulting in decomposition of the framework. The recent discovery of imine-linked COFs has sparked the creation of COFs with superior chemical stability on account of an intramolecular hydrogen bond between the hydroxyl and imine functional groups, which enhances their stability in aqueous and acidic environments.

Utilizing this feature, this thesis examines the synthesis and gas adsorption properties of novel imine-linked COFs that contain 1,3,5-tris(styryl)benzene and 1,3,5– tris(arylethynyl)benzene π-conjugated units. By creating analogs which were fluorescent in both solution and solid-state, studies were conducted to determine their ability to serve as chemical sensors for explosives. These studies highlight the potential benefits of utilizing both monomers for the development of COFs for gas storage and sensory applications.

Dedication:

This is dedicated to all my friends, family, and mentors who have all helped me

throughout the years

iii

Acknowledgements

I would like to thank Professor Psaras McGrier for accepting me into his group. His mentorship and guidance taught me to love chemistry and to become a better problem solver. I have learned a lot about myself and about my chemistry skills from working with him.

I would also like to thank my fellow group members. Luke Baldwin and Jon Crowe were always willing to help with any problem and listen if I was having issues with an experiment. They were very helpful with learning the new instrumentation I was unfamiliar with when entering the group. Grace Eder was accepted into the group as a fellow first year with me and was always there to listen to any problems going on.

My family has helped keep me grounded and focused through the time I was in graduate school. I appreciated their ability to listen to whatever I was going through. All of my friends, outside the department and within, were instrumental in my work here. They helped remind me to enjoy my time and learn as much as I could.

I would like to thank all the faculty whom helped me at The Ohio State University.

Dr. Chris Callam and Dr. Noel Paul allowed me to teach for them, whether it be in the teaching labs or being a head TA for them, something I learned much from. I would like to thank Tanya Young for running my solid state data for me, Cameron Begg for taking my SEMs, and Steven Bright for taking my TGAs. I would also like to thank Dr. Jovica

Badjic for being on my committee, as well as being an incredibly helpful professor in my time here, especially when taking classes.

Vita

May 2009……………………………………Ashland High School, Ashland, Wisconsin

May 2013……………………………………B.S. in Chemistry, University of Louisville

August 2013 to Present………………………Graduate Teaching Associate, Department of Chemistry and Biochemistry, The Ohio State University

Fields of Study:

Chemistry

Table of Contents

Abstract………………………………………………………………………………....ii

Dedication……………………………………………………………………………....iii

Acknowledgements……………………………………………………………………..iv

Vita……………………………………………………………………………………...vi

Table of Contents……………………………………………………………………….vii

List of Tables…………………………………………………………………………....x

List of Figures…………………………………………………………………………..xi

List of Schemes…………………………………………………………………………xiv

List of Abbreviations…………………………………………………………………...xv

Chapter 1: Synthesis and Characterization Imine-Linked COFs....……...... 1

1.1 Abstract……………………………………………………………………..1

1.2 Introduction…………………………………………………………………2

1.3 Results and Discussion……………………………………………………...7

1.4 Future Plans…………………………………………………………………26

vii

1.5 Conclusions……………………………………………………………….....27

Chapter 2: Gas Adsorption Studies of Imine-Linked COFs...... ………...... 29

2.1 Abstract……………………………………………………………………...29

2.2 Introduction……………………………………………………………….....29

2.3 Results and Discussion…………………………………………………...... 35

2.4 Future Plans………………………………………………………………….45

2.5 Conclusions………………………………………………………………….45

Chapter 3: Analog Synthesis and Fluorescence Quenching…...... 47

3.1 Abstract…………………………………………………………………...... 47

3.2 Introduction………………………………………………………………….47

3.3 Results and Discussion………………………………………………………52

3.4 Future Plans……………………………………………………………….…60

3.5 Conclusions………………………………………………………………….60

Chapter 4: Experimental Data……………………………………………….………...... 62

4.1 General Methods………………………………………………………….…62

4.2 Experimental Data for Linker, Vertex A, and Vertex B………………….…64

4.3 Synthesis of Alkene and Alkyne Analog and DhaTas and DhaTae………...74

viii

References and Notes……………………………………………………………...... …78

Appendix A: 1H and 13C NMR Spectra for Synthesized Compounds…………………..81

List of Tables

Table 1: DhaTas Attempts to Increase Surface Area...... 12

Table 2: DhaTae Attempts to Increase Surface Area...... 17

Table 3: Comparison of Surface Area and Distributional Pore Volume for DhaTab,

DhaTas, and DhaTae...... 44

Table 4: Percentage Fluorescence Quenching...... 59

List of Figures

Figure 1: Crystallinity of DhaTph after Water and Acid Treatment...... 6

Figure 2: N2 Isotherm of DhaTas...... 13

Figure 3: Pore Distribution of DhaTas...... 13

Figure 4: PXRD of Trial 4 of DhaTas...... 14

Figure 5: PXRD for DhaTas Comparing Reaction Times...... 14

Figure 6: N2 Isotherm of DhaTae...... 18

Figure 7: Pore Distribution of DhaTae...... 18

Figure 8: PXRD of DhaTae...... 19

Figure 9: FTIR of DhaTas...... 20

Figure 10: FTIR of DhaTae...... 21

Figure 11: Solid State 13C NMR of a) DhaTas b) DhaTae...... 23

Figure 12: TGA of a) DhaTas b) DhaTae...... 24

Figure 13: SEMs of a) DhaTas b) DhaTae...... 25

Figure 14: COF-102 and COF-108...... 31

Figure 15: a) CO2 uptake of DhaTas at 273 K and 298 K b) H2 uptake of DhaTas at 77 K and 87 K...... 36

Figure 16: a) Heat of Adsorption for CO2 of the DhaTas b) Heat of Adsorption for H2 of the DhaTas...... 37

Figure 17: a) CO2 uptake of DhaTae at 273 K and 298 K b) H2 uptake of DhaTae at 77 K and 87 K...... 38

Figure 18: a) Heat of Adsorption for CO2 of the DhaTae b) Heat of Adsorption for H2 of the DhaTae...... 39

Figure 19: a) CO2 Uptake for all COFs b) H2 Uptake for all COFs...... 43

Figure 20: Common Nitroaromatics...... 49

Figure 21: Electron-transfer Fluorescence Quenching Mechanism...... 50

Figure 22: a) COF Py-Azine b) Fluorescence Quenching using PA of COF Py-Azine..51

Figure 23: Images of Fluorescent a) Alkene and b) Alkyne Analogs...... 54

Figure 24: a) Absorbance of Alkene Analog in DMSO b) Emission of Alkene Analog in

DMSO c) Emission of Alkene Analog in Solid State...... 54

Figure 25: a) Absorbance of Alkyne Analog in DMSO b) Emission of Alkyne Analog in

DMSO c) Emission of Alkyne Analog in Solid State...... 55

Figure 26: a) Fluorescence Quenching of Alkene Analog with 4-Nitrophenol b)

Fluorescence Quenching of Alkene Analog with 2,4-Dinitrophenol...... 57

xii

Figure 27: a) Fluorescence Quenching of Alkyne Analog with 4-Nitrophenol b)

Fluorescence Quenching of Alkyne Analog with 2,4-Dinitrophenol...... 58

xiii

List of Schemes

Scheme 1: Synthesis of COF-5...... 2

Scheme 2: Imine-bond COF Formation...... 4

Scheme 3: Hydrazone-bond COF Formation...... 4

Scheme 4: Synthesis of COF DhaTph...... 6

Scheme 5: Synthesis of Vertex A...... 7

Scheme 6: Synthesis of Vertex B...... 8

Scheme 7: Synthesis of a) DhaTas b) DhaTae...... 9

Scheme 8: Synthesis of [HO2C]X%-H2P-COF from [HO]X%-H2P-COF...... 32

Scheme 9: Synthesis of CTC-COF...... 34

Scheme 10: Synthesis of DhaTab...... 42

Scheme 11: Synthesis of 2,2'-[[5'-[4-[[(2-hydroxyphenyl) methylene]amino] phenyl]-

[1,1':3',1''-terphenyl]-4,4''-diyl]bis (nitrilomethylidyne)]...... 48

Scheme 12: Synthesis of Alkene and Alkyne Analogs...... 53

xiv

List of Abbreviations oC degrees Celsius

α alpha

Å angstrom

β beta

γ gamma

∆ heat (reflux)

δ chemical shift in parts per million

μ micro

1H NMR proton nuclear magnetic resonance

13C NMR carbon 13 nuclear magnetic resonance

Ac acetyl

AcOH acetic acid aq aqueous atm atmosphere(s)

ATR attenuated total reflectance

BET Brunauer–Emmett–Teller

Boc tert-Butyloxycarbonyl br broad nBu normal-butyl sBu sec-butyl tBu tert-butyl xv c concentration calcd calculated cat catalytic

(CD3)2CO deuterated acetone

CD3CN deuterated acetonitrile

CDCl3 deuterated chloroform cm centimeter

CO2 carbon dioxide

COFs covalent organic frameworks

CuI copper iodide d day(s); doublet

DCM dichloromethane dd doublet of doublets

DMF N,N-dimethylformamide

DMSO dimethylsulfoxide dt doublet of triplets equiv equivalent(s)

ESI electrospray ionization

Et ethyl

FTIR Fourier transform infrared spectroscopy g gram(s)

GC gas chromatography

xvi

H2 hydrogen gas hr hour(s)

HBr hydrogen bromide

HCl hydrogen chloride

HRMS high resolution mass spectrometry

HWE Horner-Wadsworth-Emmons

Hz hertz

IR infrared iPr iso-propyl

J coupling constant in Hertz kJ kilojoules

L liter(s) m meta m milli; multiplet

M molarity

Me methyl

MeCN acetonitrile

MeOH methanol min minute(s) mmHg millimeters of mercury mmol millimole

MOFs metal organic frameworks

xvii mol mole(s)

MS mass spectrometry; molecular sieves

N normality: equivalents/liter

N2 nitrogen gas

NaH sodium hydride

NaHCO3 sodium bicarbonate

NaOH sodium hydroxide

NH4Cl ammonium chloride nm nanometer

NOESY nuclear Overhauser effect spectroscopy

NR no reaction o ortho p para p pentet

PdCl2(PPh3)2 Bis(triphenylphosphine)palladium(II) dichloride pdt product

Ph phenyl

POFs porous organic frameworks

POPs porous organic polymers ppm parts per million p-TSA para-toluenesulfonic acid

PXRD powder x-ray diffraction

xviii pyr pyridine q quartet rt room temperature s sec/secondary s singlet sat. saturated

SEM scanning electron microscope sep septet sext sextet

SM starting material t/tert tertiary t triplet

TEA trimethylamine

TGA thermogravimetric analysis

THF tetrahydrofuran

TFA trifluoroacetic acid

TLC thin layer chromatography

TMS trimethylsilyl

UV ultraviolet

xix

Chapter 1: Synthesis and Characterization of Imine-Linked COFs

1.1 Abstract

Two π-conjugated monomers 4,4',4''-((1E,1'E,1''E)-benzene-1,3,5-triyltris(ethene-

2,1-diyl))trianiline (Vertex A) and 4,4',4''-(benzene-1,3,5-triyltris(ethyne-2,1- diyl))trianiline (Vertex B) were synthesized using a combination of Horner reactions and palladium catalyzed cross coupling reactions. Both monomers were combined with 2,5- dihydroxyterephthalaldehyde (Dha) to form two new imine-linked COFs, DhaTas and

DhaTae with the overall goal of creating highly porous luminescent materials. Although a variety of different reactions conditions were explored to form the COFs, only amorphous porous polymers with modest BET surface areas and pore volumes were obtained. Utilizing FT-IR and cross polarization magic angle spinning (CP-MAS) 13C

NMR spectroscopies, the formation of the imine-linkages and the connectivity of the

COF could be confirmed. Both COFs exhibit high thermal stabilities and unique morphologies as indicated by the SEM micrographs

1.2 Introduction

Over the past decade, COFs have emerged as promising of porous organic materials for applications related to energy storage and conversion, catalysis, and gas storage.

Utilizing the concept of reticular chemistry, metal-organic frameworks (MOFs), which consist of organic linkers coordinated to inorganic metal clusters to give well-defined structures, were synthesized.1 In 2005, Yaghi and coworkers were able to apply this same concept to synthesize boronate-ester linked COF-5 (Scheme 1), which was the first example of a crystalline porous organic material.2

Scheme 1: Synthesis of COF-5

When synthesizing polymers, many of which contained irreversible bond forming reactions, only amorphous porous organic polymers (POPs) were obtained. When monomers were subjected to reversible bond forming reactions using dynamic covalent chemistry (DCC), crystalline COFs with high surface areas could be formed. The reversibility allows the bonds to be formed, broken, and reformed to yield the most thermodynamic product.3 This typically occurs through simultaneous bond formation and crystallization, which is reminiscent of a “self-healing” process permitting the formation of a highly porous crystalline material.4

COFs are a class of porous crystalline materials composed of light elements (such as H, B, C, N, and O) that are linked by covalent bonds. The modular nature of COFs permits the integration of various π-conjugated molecular building blocks into highly ordered polymeric structures with low densities and high thermal stabilities. COFs can be categorized into either two- (2D) or three-dimensional (3D) frameworks, varying in dimension by the linkers used in formation of the framework. 2D COFs typically form 2D sheets that stack to form a layered eclipsed structure. The ordered columns in 2D COFs could facilitate charge carrier transport in the stacking direction, which leads to the potential to form highly ordered organic materials for photovoltaic and optoelectronic applications. 3D COFs have an extended three dimensional framework containing an sp3 carbon or silane atom which gives high surface areas, many open sites in the framework, and low densities, which make these 3D COFs ideal for gas storage. 4

Initially, many of the COFs in the literature contained B-O linkages, as shown in work done by Yaghi and coworkers5. However these boronate ester and boroxine linkages, while thermally stable, were unable to maintain crystallinity when exposed to aqueous 3 conditions or moisture. This feature limited the practical applications to which this linkage could be applied. In 2009, Yaghi and coworkers synthesized the first example of an imine- linked COF. 6 Imine-bond formation is considered an ideal type of reaction as it is reversible, cost-effect, and often high yielding. These imine-based COFs can be categorized into two separate groups; the Schiff-base condensation of an aldehyde with an amine6 (Scheme 2) or the imine-bond formation between condensation of an aldehyde and a hydrazide7 (Scheme 3). Imine-based COFs are comparable to boron-containing COFs in terms of crystallinity, while also retaining stability in organic solvents.

Scheme 2: Imine-bond COF Formation

Scheme 3: Hydrazone-bond COF Formation

However, imine-based COFs, still exhibited partial decomposition upon exposure to aqueous conditions on account of the reversibility of the bond. In an effort to overcome the issue of degradation with boronate ester-linked COFs, COF-10 was doped with minimal

4 amounts of pyridine. Using the trend that nitrogen coordinates to boron in forming Lewis

Acid/Base relationships, this can stabilize the boronate ester bond and help to prevent the degradation in humidity and water9. However, this pyridine doping experiments led to a decrease in gas adsorption properties, even with the enhanced water stability. This did not lead to the most optimal route for stabilization and this pyridine doping would not have been applicable for imine-based COFs.

In 2013, Banerjee and coworkers found a way of not only enhancing crystallinity of the imine-based COFs, but also increasing the stability of the COF when exposed to not only water, but acidic and basic conditions10. By incorporating -OH functionalities adjacent to the Schiff base centers in the COF, an O-H …N=C intramolecular hydrogen bond system could be formed (Scheme 4)10. The presence of this hydrogen bond can act as a shield for the imine, protecting the basic imine nitrogen from acidic conditions and water. The formed COF DhaTph can retain crystallinity in water and 3N HCl for more than one week (Figure 1).10 The introduction of the hydrogen bond also allows for enhanced crystallinity and porosity. The hydrogen bond interaction allows for the suppression of the torsion of the edge units and locks the 2D sheets into a planar conformation. The planarization increases the interactions between the 2D layers and, upon AA stacking, has a positive effect on the properties of the material, including enhanced crystallinity and porosity. 11

Scheme 4: Synthesis of DhaTph

Figure 1: Crystallinity of DhaTph after Water and Acid Treatment

With this project, our aim was to synthesize novel imine-based COFs, supported by the intramolecular hydrogen bonding motif for enhanced chemical stability, that contained

π-conjugated monomers and test their gas adsorption properties. By synthesizing two new 6 vertices with amino groups, we envisioned performing Schiff-base chemistry to construct not only highly porous and crystalline imine-linked COFs with for gas storage applications, but also highly luminescent materials for chemical sensing applications.

1.3 Results and Discussion

The research project began with the synthesis of the two novel vertices; Vertex A

(Scheme 5) and Vertex B (Scheme 6). Vertex A was synthesized using the purchased

1,3,5-Tris(bromomethyl)benzene and performing an Arbuzov reaction to form 1,3,5- tris(diethoxyphosphorylmethyl)benzene (2). Taking this product and 4-nitrobenzaldehyde under Horner Wadsworth Emmons coupling conditions gave the precursor 3. Reduction of the nitro groups led to the formation of the desired product Vertex A, an orange solid which exhibits bright blue emission in solution and in the solid state.

Scheme 5: Synthesis of Vertex A

Vertex B was synthesized by taking 4-iodoaniline and performing an amine protection using a tert-butyloxycarbonyl (4). Once the amine was protected, a Sonogashira reaction was performed, followed by deprotecting the terminal alkyne to yield product 5.

5 and 1,3,5-tribromobenzene was reacted under Sonogashira conditions to give the precursor 6, which allowed access to Vertex B by an acid deprotection. This vertex was a brown solid which exhibits bright blue emission in solution and in the solid state.

Scheme 6: Synthesis of Vertex B

Both Vertex A and Vertex B were confirmed by 1H and 13C NMR as well as high resolution mass spectrometry (Appendix A).

Taking these new vertices and the known 2,5-dihydroxyterephthalaldehyde linker unit12, two new COF frameworks were explored (Scheme 7). It would be expected that these COFs would form hexagonal shaped 2D sheets, which would have a high crystallinity and porosity.

Scheme 7: a) Synthesis of DhaTas COF b) DhaTae 9

Vertex B was the easier of the two vertices to synthesize in the beginning, and many of the early COF formation trials were performed using that vertex. Once the nitro reduction was optimized for Vertex A, this vertex could be synthesized on gram scales and was used for many of the trials of the COF formation. We had believed that both COFs would have similar conditions to lead to their formation and trials were undergone with both Vertex A (Table 1) and, eventually Vertex B, which yielded some usable results.

Many of the early trials for DhaTae were before the washes and suspensions of the COF were determined. We also had difficulties determining an accurate mass of sample for the early experiments and, since the surface areas can be affected by faulty masses, many of the results before Trial 5 (Table 2) are not the most accurate data and would have to be re- run to confirm the results. However, as additional experiments were run, it appeared as though the 1:1 ratio of linker to vertex would not lead to the most optimized COF.

The first study performed with DhaTas was to vary the ratio of the linker to vertex in the system. It had been seen with many different literature papers that either 2:1, 3:2, or

3:1 were the most common and three trials were set up using these different ratios (Trial 2,

3, and 4; Table 1). The 1:1 which had been attempted earlier had not given a useful surface area. All three of these trials were sealed in the ampoule in the same concentration of mesitylene, dioxane, and 6M acetic acid. All three were placed in the oven at 120 oC for 3 days. After three days, the solids were filtered off, washing with DCM, acetone, and THF, followed by suspension overnight in DCM before suspension in three separate acetone washes over the next two days. The solids were dried and placed under vacuum at 150 oC

10 for 12 hours. The BET surface areas increased with each trial; the 2:1 gave the lowest surface area with 212 m2/g1, while the 3:2 gave 262 m2/g1, and the 3:1 gave the best result with a surface area of 622 m2/g1. This was the first trial that gave positive results and a promising surface area. Varying the solvent amounts (Trial 11, 12, and 13; Table 1) as well as changing up the solvent system (Trial 9 and 10; Table 1) gave differing results, but none were able to match the success of Trial 4’s conditions. Using the Trial 4 sample, a full nitrogen isotherm (Figure 2) was taken as well as the pore size distribution for DhaTas

(Figure 3). A PXRD was taken on the remainder of the sample (Figure 4), but didn’t show as crystalline a material as we expected. While there was one small peak at 2.5, we expected more concise peaks. To attempt to enhance the crystallinity, DhaTas were set up using the same conditions as Trial 4, but left in the oven for five days and seven days (Trial

7 and 8; Table 1). These trials did show similar surface areas, but the crystallinities did not appear to improve over the extended reaction time (Figure 5). When setting up the COFs for many of these trials, a solid would appear to already crash out before the reaction vessel was placed in the oven. Because it appeared that the reaction was occurring as soon as the acetic acid was added, the extended reaction time didn’t help enhance crystallinity of the framework because it was forming too rapidly. Once the framework began to form and crash out of solution, the reversibility of the reaction could no longer apply. Without that reversibility of the imine formation, the COF appeared to form in an amorphous powder.

Table 1: DhaTas Attempts to Increase Surface Area Using Various Conditions Trial Linker Vertex Ratio of Solvent Acetic Acid Days Surface (mmol) (mmol) Linker Conditions Area to (m2/g) Vertex 1 0.08 0.08 1:1 1 mL 0.2 mL 3 108 Mes/1mL (concentrated) Dioxane 2 0.08 0.04 2:1 1 mL Mes/ 0.2 mL (6M) 3 212 1 mL Dioxane 3 0.065 0.043 3:2 0.9 mL 0.18 mL (6M) 3 262 Mes/0.9 mL Dioxane 4 0.12 0.04 3:1 1.3 mL 0.27 mL (6M) 3 622 Mes/1.3 mL Dioxane 5 0.065 0.043 3:2 1.8 mL NONE 3 45 DMSO 6 0.12 0.04 3:1 1.3 mL 0.27 mL (6M) 3 652 Mes/1.3 mL Dioxane 7 0.12 0.04 3:1 1.3 mL 0.27 mL (6M) 5 565 Mes/1.3 mL Dioxane 8 0.12 0.04 3:1 1.3 mL 0.27 mL (6M) 7 587 Mes/1.3 mL Dioxane 9 0.12 0.04 3:1 1.3 mL 0.27 mL (6M) 3 375 Mes/1.3 mL Dioxane/0.5 mL Ethanol 10 0.12 0.04 3:1 1.3 mL 0.27 mL (6M) 3 430 DCB/1.3 mL Ethanol 11 0.12 0.04 3:1 1.73 mL 0.27 mL (6M) 3 95 Mes/0.87 mL Dioxane 12 0.12 0.04 3:1 0.87 mL 0.27 mL (6M) 3 235 Mes/1.73 mL Dioxane 13 0.13 0.09 1.44:1 1.7 mL 0.2 mL (8M) 3 36 Mes/0.3 mL Dioxane

Figure 2: N2 Isotherm of DhaTas

Figure 3: Pore Distribution of DhaTas

Figure 4: PXRD of DhaTas

Figure 5: PXRD for DhaTas Comparing Reaction Times

After obtaining these results for DhaTas, similar studies were performed with

Vertex B (Table 2). Varying the ratios of the linker to vertex (Trials 8, 9, and 10; Table 2) showed that there was an increase in surface area depending on the ratio, but the pattern differed from DhaTas. The 3:1 ratio, which gave the highest results for DhaTas, gave only a 124 m2/g1, while the 2:1 ratio was lower with a 72 m2/g1. The highest result for DhaTae was the 3:2 ratio which gave a surface area of 440 m2/g1. In one unique attempt, Trial 12, the reaction was not run in a sealed ampoule in an oven, but rather was performed in a hood. The aldehyde was heated to 50 oC in DMSO and a solution of the alkyne monomer in DMSO was added over a 30 minute period. Once the entire alkyne monomer was added, the reaction was refluxed under nitrogen for 24 hours. During this reflux period, a dark brown solid crashed out of the DMSO. This solid was filtered off, washed with DCM, acetone, and THF before being suspended in acetone for 2 days. However, this trial did not give a very high surface area and appeared unsuccessful in forming the desired COF.

DhaTae Trial 10 gave the highest surface area and was used to get the full nitrogen isotherm (Figure 6) and pore size distribution (Figure 7). A PXRD was taken, but showed no peaks present (Figure 8). While we were hoping there would be at least one peak present, as in the DhaTas PXRD, this one appeared to be completely amorphous. Because of the lack of crystallinity in both DhaTas and DhaTae, a few trials were run without any acid. These trials were run in hopes that without the acid, the reaction kinetics would slow down, allowing for longer reversibility of the formation of the framework and leading to enhanced crystallinity. Trial 5 from Table 1 and Trial 7 from Table 2 both contained no acid in the reaction. However, both these results gave low surface areas (40 m2/g-1) and

15 amorphous solids. The addition of the acid appears to be needed in the solvothermal formation of the COFs, but may not be necessary when trying to do the COF formation on the benchtop.

For both Trial 13 in Table 1 and Table 2, a specific set of conditions was attempted.

In a paper by Banerjee and coworkers13, a similar COF structure is synthesized using those specific conditions. We were hopeful that the results would be able to translate to our systems, but found that the results did not correlate. The difference in the systems, while small, appeared to make a significant difference and further optimization would be necessary for our systems.

Table 2: DhaTae Attempts to Increase Surface Area Using Various Conditions

Trial Linker Vertex Ratio of Solvent Acetic Acid Days Surface (mmols) (mmols) Linker Conditions Areas to (m2/g) Vertex 1 0.08 0.08 1:1 1 mL 0.2 mL 3 70 Mes/1 mL (concentrated) Dioxane 2 0.08 0.08 1:1 2 mL DMF 0.2 mL 3 18 (concentrated) 3 0.12 0.04 3:1 1 mL 0.2 mL 3 8 Mes/1 mL (concentrated) Dioxane 4 0.08 0.08 1:1 2 mL 0.2 mL 3 2 Mes/2 mL (concentrated) Dioxane 5 0.08 0.08 1:1 1 mL 0.2 mL 3 400 Mes/1 mL (concentrated) Dioxane 6 0.08 0.08 1:1 1 mL 0.2 mL 3 110 Mes/1 mL (concentrated) Dioxane 7 0.065 0.043 3:2 1.8 mL NONE 3 40 DMSO 8 0.12 0.04 3:1 1.3 mL 0.27 mL (6M) 3 124 Mes/1.3 mL Dioxane 9 0.08 0.04 2:1 1 mL 0.2 mL 3 72 Mes/1 mL (concentrated) Dioxane 10 0.065 0.043 3:2 0.9 mL 0.18 mL (6M) 3 440 Mes/0.9 mL Dioxane 11 0.065 0.043 3:2 1.8 mL 0.18 mL (6M) 3 340 Dioxane 12* 0.31 0.31 1:1 2 mL NONE 1 35 DMSO 13 0.13 0.9 1.44:1 1.7 mL 0.2 mL (8M) 3 16 Mes/0.3 mL Dioxane

Figure 6: N2 Isotherm of DhaTae

Figure 7: Pore Distribution of DhaTae

Figure 8: PXRD for DhaTae

Despite the lack of crystallinity of the COFs formed, the desired imine-bond formation could be proven using FTIR and solid state 13C NMR. The FTIR of both DhaTas and DhaTae could be compared to not only the monomers used to form the COF, but also the synthesized analogs (Chapter 3) as shown in Figure 9 and 10. DhaTas showed the imine-bond stretch at 1608 cm-1, which was comparable to the analog imine-bond stretch at 1615 cm-1 (Figure 10). The N-H stretch from the alkene monomer at 3353 cm-1 disappears in both DhaTas and the analog, agreeing with the formation of the imine-bonds in both species. DhaTae FTIR also showed the same results (Figure 11), having the imine- bond show at 1610 cm-1, which was highly comparable to the analog at 1613 cm-1. The alkyne bond stretch maintained in all three compounds, appearing at 2201 cm-1 in the COF,

2207 cm-1 for the analog, and 2197 cm-1 for the monomer; this shows that during the

19 formation of the COF with acetic acid, the alkyne bonds are not protonated or destroyed.

As shown in DhaTas COF, the N-H stretch at 3358 cm-1 for the alkyne monomer is gone in DhaTae. Both COFs appear to be forming the correct bonds, just forming an amorphous solid rather than an ordered system.

Figure 9: FTIR of DhaTas

Figure 10: FTIR of DhaTae

Solid-state CP-MAS 13C NMR spectra were taken for both DhaTas and DhaTae

(Figure 11). Both NMR spectra show the imine bond linkage at about 160 ppm for DhaTas and at 162 ppm for DhaTae. The spectral data for both appears to be what would be desired as compared to a publication from Banerjee and coworkers on a similar COF structure13.

In their 13C NMRs, they report the imine bond stretch at 162 ppm and the C-OH bond to be at 155 ppm; both of these are similar to the two COFs below. The rest of their spectrum included only 3 peaks, corresponding to the fact that many of the aryl carbons would have very similar stretching and would lead to more intense peaks, but may not show a distinct signal for each specific carbon. The one difference in the Banerjee spectrum for DhaTab to the DhaTas and DhaTae spectra would be one additional peak (134 ppm for DhaTas and

110 ppm for DhaTae). Both of these would be the expected alkene and alkyne carbons for the given system, as neither bond is altered in the formation of the framework. The DhaTae

COF does contain a significant amount of aldehyde monomer still trapped in the framework, as it shows two impurity peaks at about 200 ppm and 50 ppm. The DhaTas

COF contains only trace amounts of these impurities. The solid state 13C NMR does show the formation of the desired imine linkage of the COF systems and retains both the alkene and alkyne bonds in each system, as expected.

Figure 11: Solid State CP-MAS 13C NMR a) DhaTas b) DhaTae

Despite the formed polymer being amorphous and not having very high surface areas, both thermal stabilities of DhaTas and DhaTae were tested by TGA. DhaTae was more stable than DhaTas as shown in Figure 12; DhaTae showed only about a 60% weight at 800 oC, while DhaTas showed about a 40% weight at 800 oC. However, both DhaTas and DhaTae did not have any real drop off until about 400 oC; therefore, both appeared to have very high thermal stability to 400 oC which is very common for many COF systems.

Figure 12: a) TGA of DhaTas b) TGA of DhaTae

The final data we collected for DhaTas and DhaTae was the SEMs of the frameworks (Figure 13). While we were not expecting the most uniform images from the amorphous nature of the COFs, we were still curious as to the surface area images of these materials. DhaTae appeared to be more uniform in pattern than DhaTas, but both still appeared to be highly amorphous in nature.

Figure 13: SEM images of a) DhaTas b) DhaTae

1.4 Future Plans

While we were able to synthesize the new DhaTas and DhaTae systems, these were far from optimized structures. Both had relatively low surface areas and lacked high crystallinity. Since both COFs were amorphous, neither material exhibited luminescence in the solid-state. There appears to be two possible problems with the current synthesis attempts; 1) the reaction is occurring too quickly leading to rapid precipitation of the polymer, or 2) limited solubility of the monomers permits an inefficient nucleation process.

One attempt which could be used to slow the formation of the framework would be to form the COF on the benchtop. One such trial was attempted with DhaTae, but further exploration into this has not been attempted. The benchtop synthesis could be attempted by cooling the reaction down further, even placing the reaction in an ice bath, when the vertex solution is added to the linker solution, to help slow the initial reaction down. This could be allowed to stir at a lower temperature longer before refluxing to a higher temperature (180 oC or higher) for 1-3 days. While this may increase the amount of time required to form the framework, this could help lead to a more ordered system. Another possibility would be to use the solvothermal method to form the COF, but before placing the ampoule in an oven at 120 oC, the ampoule could be heated slightly at 50 oC for a few hours, or even overnight, to attempt a slower formation of the imine bond. After allowing to stir at a lower temperature first, the reaction vessel could then be placed in the oven and allowed to react at the higher temperature.

In an attempt to change the solubility of the monomers before subjecting them to heating for three days, the solvent system could be altered or the concentrations changed.

The concentrations of the solvent added never varied throughout the experiments, but changes in molarity has been shown to alter certain systems. Trying a more concentrated or more dilute solution could lead to a more soluble system. The solvent system would need more optimization as well. As shown with Banerjee and coworkers paper13 they found their COF to be optimized in 1.7 mL mesitylene and 0.3 mL dioxane with 0.2 mL

8M acetic acid. These conditions must have occurred after multiple solvent screenings, something of which has not been tried with either system currently. These solvent screenings should include trials with more dioxane for the DhaTae COF, as Trial 11 from

Table 2 showed only a slight decrease of surface area between that attempt and Trial 10 which was a 1:1 mesitylene/dioxane solvent system. Using these attempts to optimize the

COFs should be able to produce better results in the future.

1.5 Conclusions

We synthesized two new vertices, Vertex A and Vertex B using various reactions conditions. Using both of these new vertices, DhaTas and DhaTae COFs could be synthesized using the Dha linker. Many trials were run in an attempt to produce highly porous and crystalline materials, but only moderate results were obtained. Despite the low surface areas and lack of crystallinity, the formation of the imine-linkages was confirmed

27 by FT-IR and solid-state CP-MAS 13C NMR spectroscopies. TGA analysis also revealed high thermal stability for both COFs. These COFs, once optimized, could lead to exciting materials with unique luminescent properties.

Chapter 2: Gas Adsorption Studies of Imine-Linked COFs

2.1 Abstract

Gas adsorption studies are very prominent in scientific research currently. As CO2 levels rise and H2 attempts to become a viable alternative fuel source, gas adsorption of polymers, especially using COFs, has grown. Taking the newly developed DhaTas and

DhaTae systems, CO2 and H2 gas adsorption studies were performed. CO2 adsorptions were run on both systems at 273 K and 295 K to obtain the isoteric heats of adsorption. H2 studies were run at 77 K and 87 K to obtain the heats of adsorption for both systems.

2.2 Introduction

As the burning of fossil fuels and combustion energy has increased, the greenhouse gas emissions from these fuels have increased as well. These greenhouse gases building up in the atmosphere could be leading to a change in Earth’s climate at a dramatic rate, a rise in the ocean levels, and an increasing acidity of the oceans4. One of the most dangerous greenhouse gases, CO2, has become increasingly important in scientific research. CO2 is 29 the most common gas released from the burning of fossil fuels as it accounts for 84% of

14 all greenhouse gases emitted by humans . CO2 emissions have increased by 80% between

1970 and 200415.

Because of the dramatic increase in CO2 levels, research in carbon capture has been pushed to the forefront of scientific research. One of the main concerns with the chemical system which will be employed to uptake CO2 is that the chemical source must be reusable.

The energy input for the regeneration is one of the factors for cost and efficiency of these

CO2 uptake.

CO2 gases often have to be separated from other gases in the atmosphere or in a fuel cell and therefore can lead to separation issues. Three issues which hold the greatest promise for reducing CO2 emissions are the separation from power plant flue streams, separation from natural gas wells, and separation from fuel gas16. Each separation leads to its own challenges, meaning that the chemical source must have a variety of properties for optimal separation. In post combustion capture from flue gas wells, there is low pressure and low concentration of CO2 and must be separated from high concentrations of N2 gas.

Natural gas reserves of CH4 are contaminated with over 40% CO2 and N2 and the use of these gas fields is only acceptable if the additional CO2 is sequestered at the source of production. CO2 separation from fuel gas occurs at high pressures and temperatures (250-

o 16 450 C) for pre combustion separation of CO2/H2 . The key challenge for separation is that the differences in the properties of the gases are small. However there are some small differences in quadrupolar moment and polarizability; CO2 has a large quadrupole moment

17 compared to the other gases and CH4 adsorbs over N2 due to higher polarizability .

As CO2 uptake and storage has been studied in the polymer field, COFs began to approach the topic. Many of the necessary properties of chemical materials for CO2 uptake are present in COFs; COFs can be re-usable, they are stable in high temperatures, they have the ability to uptake CO2 at both low and high pressures, and the systems can be controlled at the molecular level to take advantage of the chemical differences between the gases8.

CO2 uptake has been found to be directly related to the COF system’s total pore volume.

COF-102 (Figure 14) shows higher adsorption at lower pressures due to its compact atomic packing, while at high pressures COF-108 shows higher saturation storage capacities due to its larger pore volumes (Figure 14)4.

Figure 14: COF-102 and COF-108

In 2015, Jiang and coworkers published a new COF which used the hydroxyl group located on the linker as a hinge to insert functional groups into the pore of the COF18. The synthesis of [HO2C]X%-H2P-COF from [OH]X%-H2P-COF is shown in Scheme 8. This simple insertion of functionality into the pore of the COF led to a dramatic uptake in CO2 gas. In the [HO]50%-H2P-COF, the CO2 uptake in mg/g at 273 K was only 46 with a heat of adsorption of 29.4 kJ/mol. However, when the COF was transformed to form

[HO2C]100%-H2P-COF, the CO2 uptake in mg/g at 273 K was 174, with a heat of adsorption of 43.5 kJ/mol18. These numbers are much more competitive and show the ability to tune

COF frameworks to increase the capacity for CO2 uptake.

Scheme 8: Synthesis of [HO2C]X%-H2P-COF from [HO]X%-H2P-COF

Hydrogen gas storage has also become a high source of research and experiments as it can be used as an alternative fuel source. H2 has clean combustion and high chemical energy density, three times as great as gasoline, which leads to an ideal substitute for fossil fuels19. However, hydrogen gas requires an efficient and safe way to be stored, as hydrogen gas is highly flammable. Many studies have been undergone in recent years on hydrogen storage using porous materials.

A general trend that has appeared is as the surface areas increase, the uptake of H2 increases as well8. Computational studies have been performed and predict the binding energies between the hydrogen molecules and the COF framework has almost no role in

20 the H2 uptake, but rather the H2 uptake is determined by surface area of the framework .

However, there have been some studies performed in changing the COF framework to increase the H2 uptake. In a study performed by Zheng and coworkers, they changed the framework of COF-5 by adding an undulated macrocyclic cyclotriacetechylene (CTC) and

21 found an increase in H2 uptake at lower pressure (Scheme 9) . The CTC-COF has higher low pressure H2 uptake than COF-5 and COF-10 with planar 2D structures and comes close to the uptake of 3D COF materials. This study shows that it is possible to tune the COF framework to potentially increase the uptake of H2.

Scheme 9: Synthesis of CTC-COF

The biggest problem that seems to appear with H2 uptake for COFs gas adsorption typically occurs at cryogenic temperatures (<100 K). The US Department of Energy (DOE) had set a target for hydrogen storage of 5.5 wt% at an operating temperature of 273-298 K and 100 atm maximum pressure8. Unfortunately for materials which rely on physisorption at low temperatures, only porous materials with very high surface areas (greater than 3000 m2/g) and pore sizes between 0.7 and 1.2 nm have the potential to meet the DOE target4.

However, there have been theoretical studies performed which indicate that hydrogen storage at room temperatures would be practically possible for COFs22. There is a possibility that COFs could transition into great materials for H2 storage and transport with more studies.

We decided to take our two new COFs synthesized in Chapter 1, DhaTas and

DhaTae, and test their gas adsorption properties. Both CO2 and H2 uptake were tested and the isoteric heats of adsorption found to see if our COFs could be useful in gas adsorption.

2.3 Results and Discussion

We decided to investigate the gas adsorption properties of the newly formed

DhaTas and DhaTae. Despite lacking crystallinity and high porosity, it was expected that there could still be gas adsorption for both CO2 and H2. The CO2 adsorption experiments were run at 273 K as well as 295 K. For the H2 adsorption experiments, both were run at

77 K and 87 K. Using the data found from both COFs, the heats of adsorption for CO2 and

H2 could be determined. The data results of the experiments are shown below in Figure 15 and 16 for DhaTas, while the results for DhaTae are shown in Figure 17 and 18. All this data was collected using Trial 10 (Table 1 and 2; Chapter 1) for both DhaTas and DhaTae.

Because both of these COFs had similar surface area (430 m2/g and 440 m2/g) we determined that the adsorption experiments would yield data more comparable than if using a higher surface area system such as Trial 4 for the DhaTas COF.

Figure 15: a) CO2 uptake of DhaTas at 273 K and 295 K b) H2 uptake of DhaTas at 77 K and 87 K

Figure 16: a) Heat of Adsorption for CO2 of the DhaTas b) Heat of Adsorption for H2 of the DhaTas

Figure 17: a) CO2 uptake of DhaTae at 273 K and 298 K b) H2 uptake of DhaTae at 77 K and 87 K

Figure 18: a) Heat of Adsorption for CO2 of the DhaTae b) Heat of Adsorption for H2 of the DhaTae

DhaTas for both the CO2 and H2 adsorptions had a higher value than DhaTae.

3 3 DhaTas, adsorbing a high of 32 cm /g at 273 K and 20 cm /g at 295 K for CO2 uptake.

DhaTae was significantly lower with only a high of 21 cm3/g at 273 K and 16 cm3/g at 295

K for the CO2 uptake. This is an opposite trend of what we imagined would happen. This

39 idea will be discussed later in this chapter. The H2 adsorptions show the same pattern as

3 3 the CO2 graphs. The high values for DhaTas were 50 cm /g at 77 K and 39 cm /g at 87 K.

DhaTae showed high values of 41 cm3/g at 77 K and 30 cm3/g at 87 K.

The heat of adsorption is an indicator of the strength of the interactions between the adsorbate and the solid adsorbent. There can be two different adsorptions; physical adsorption (physisorption) and chemical adsorption (chemisorption). Physisorption exists if the force of attraction between adsorbate and absorbent are Van der Waals forces. These forces of attraction are very weak and can easily be reversed by heating or decreasing the pressure in the system. Chemisorption exists if the force between the absorbate and the absorbent are almost the same strength as chemical bonds. These forces of attraction are very strong and are not reversible. The desired heat of adsorption for CO2 between the absorbate and the absorbent is about 20-50 kJ/mol, whereas for H2 is a lower range of about

9-15 kJ/mol. Both DhaTas and the DhaTae have heats of adsorption falling in the desired range for the CO2 plots, as both have a high of about 25 kJ/mol. This would mean that the only interaction occurring with the CO2 adsorption the interaction of the gas with the pore itself, rather than any of the π-systems of the vertices. However, the H2 values do not both fall in this range, as DhaTas is only about 4.5 kJ/mol. DhaTae is 9 kJ/mol and falls on the lower spectrum of the desired range; this could indicate there is a stronger interaction for

H2 with the extra π-system of the alkyne bond.

Because we were having trouble forming a porous, crystalline system, another COF system using only the basic single bond linkage between the center phenyl ring and the aniline rings was proposed. At the time, we synthesized the known vertex and tried one

40 attempt at forming the COF structure (Scheme 10). The first trial gave a surface area of

400 m2/g, a similar surface area to that which we ran the gas adsorption experiments of both DhaTas and DhaTae. However, a recently published work by Banerjee and coworkers13 showed the synthesis of DhaTab with optimized conditions. Under their conditions, the surface area for the COF was 1500 m2/g, a value which was much higher than any of the results found for both DhaTas and DhaTae. Because of this dramatically large surface area difference, which can directly be correlated to gas adsorption, the data used in this chapter only includes our one attempted trial at forming DhaTab (surface area of 400 m2/g). However, seeing the positive results that the Banerjee group was able to achieve with the triaryl linker gave hope that, with further optimization, the surface areas and crystallinity of DhaTas and DhaTae could be increased. The same gas adsorption experiments run for DhaTas and DhaTae were run on DhaTab. These experiments yielded interesting results, as our DhaTab results were higher for both the CO2 and H2 experiments

(Figure 19).

Scheme 10: Synthesis of DhaTab

Figure 19: a) CO2 Uptake for all COFs at 273 K b) H2 Uptake for all COFs at 77 K

Because the CO2 could have had interactions with the π-bonds present in both vertices, it would be expected that DhaTas and DhaTae would have higher adsorption values for CO2. However, when looking at the pore distributions for all three systems, the largest pore volume found would be in the DhaTab, followed by DhaTas, and finally

DhaTae (Table 3). In other COF systems studied, it has been found that the larger the pore

8 volume of the system, the higher the amount of CO2 adsorbed . This must mean that the larger pore volume is more influential on CO2 uptake than the vertex affinity for binding

CO2. Because both DhaTas and DhaTae are amorphous in nature and far from optimized conditions, the pore volume would be expected to increase as a more ordered system is produced. The pore volume of the more amorphous COF is not as valid as it does not contain one distinct pore, but rather many smaller pores located on the surface of the COF.

It would be expected that DhaTae would have a much larger pore volume than presented currently, as well as DhaTas increasing once the system was fully optimized. Using

Banerjee and coworkers DhaTab example, their CO2 and H2 uptakes were much higher than the results we got with our one DhaTab trial13. Seeing these results with the higher surface area and more crystalline material would lead us to believe that the CO2 and H2 uptakes could both be increased when the COFs become optimized further.

Table 3: Comparison of Surface Area and Distributional Pore Volume for DhaTab, DhaTas, and DhaTae

COF CO2 Uptake H2 Uptake at Pore Pore Width Surface at 273 K 77 K Volume (nm) Area (m2/g) (cm3/g) (cm3/g) (cm3/g) DhaTab 35 56 1.37 2.09 400

DhaTas 32 50 1.33 3.10 430

DhaTae 21 41 0.90 2.13 440

2.4 Future Plans

Future work with this project would be similar to the future plans of Chapter 1; optimizing the conditions of both DhaTas and DhaTae. Once the conditions were optimized, the CO2 and H2 adsorption experiments would be able to be run again. It would be expected that both of these systems would see an increase in adsorption. These systems would also be interesting to see if they could be used for separation experiments by getting adsorptions of both N2 and CH4 gases. These studies could be easily done on the optimized

COF systems.

2.5 Conclusions

Using the DhaTas and DhaTae formed in Chapter 1, the gas adsorption properties of each could be studied. It was found that DhaTas had higher CO2 and H2 adsorption than

DhaTae. However, when comparing gas uptake in these systems, higher CO2 uptake occurs when there is a greater maximum pore volume of the COF; DhaTas has a higher maximum pore volume at this point in time than DhaTae. Because of the difference in pore volumes, DhaTae would be expected to have lower uptake, despite having a higher theoretical affinity for CO2 because of its additional π-bond. This would indicate that CO2 uptake is often determined more by pore volume than by the vertices present in the COF system. It would be expected that, as the COFs were both optimized, that the CO2 and H2 45 uptakes would increase as shown by the published DhaTab13. For both DhaTas and

DhaTae, the heats of adsorption for the CO2 fall within the desired range at about 25 kJ/mol; meaning that the strength of the CO2 interaction with the solid COF does not depend on the interactions of the π-systems, but rather interactions with the pore of the

COFs. The H2 heats of adsorption, however, show that there is stronger interactions between the additional π-system of DhaTae as the heat of adsorption is twice as high as that of DhaTas.

Chapter 3: Analog Synthesis and Fluorescence Quenching

3.1 Abstract

After obtaining COFs that were amorphous and not luminescent in the solid-state, we decided to examine the photophysical properties of the imine-linked 1,3,5- tris(styryl)benzene and 1,3,5–tris(arylethynyl)benzene π-conjugated analogs. Refluxing

Vertex A or B with salicylaldehyde in ethyl acetate produced the imine-linked alkene and alkyne analogs. Using these analogs, the photophysical properties and chemical sensing studies upon the addition of nitroaromatic compounds were examined in DMSO. Both the alkene and alkyne analogs exhibited fluorescence quenching in the presence of 4- nitrophenol and 2,4-dinitrophenol.

3.2 Introduction

When developing new COF systems, a basic analog can be synthesized. Analogs can allow for many different kinds of studies from verifying bond formation to crystal structures. Analogs are synthesized using whichever vertex is desired for the two- dimensional COF and reacting with only a one-dimensional linker (Scheme 11)13. These

47 analogs can be tuned, such as adding longer alkyl chains to the system, to be soluble in common organic solvents. This solubility can allow for 1H and 13C NMRs, as well as high resolution mass spectrometry, crystal structure growth, and fluorescent and UV studies.

Scheme 11: Synthesis of 2,2'-[[5'-[4-[[(2-hydroxyphenyl) methylene]amino] phenyl]- [1,1':3',1''-terphenyl]-4,4''-diyl]bis (nitrilomethylidyne)]

When synthesizing conjugated polymer networks, chemical sensing can be investigated. Detection of chemical explosives is necessary for military operations, homeland security, and cleaning of the environment. Sensing of chemical explosives, especially nitroaromatics, has become a widely researched topic. Different methods are already in place for sensing, such as gas chromatography, Raman spectroscopy, ion- mobility spectroscopy, and fluorescence spectroscopy23. Each has advantages and disadvantages when detecting different explosive compounds. Nitroaromatics, 1,3,5- trinitrotoluene (TNT), 1,3,5-trinitrophenol (Picric Acid), 2,4-dinitrotoluene (DNT), and

2,4-dinitrophenol (DNP) (Figure 20) just to name a few, can be challenging compounds for vapor sensing due to low vapor pressure24. The electron-poor nitroaromatic compounds 48 can form good π-stacking interactions with electron-rich conjugated polymers. This fluorescence quenching mechanism shown in Figure 2124 is achieved through an electron- transfer donor-acceptor process. The electron withdrawing nitro groups placed on the aromatic ring lowers the empty π* orbital, increasing the electron accepting component of the nitroaromatics. The conjugated polymers electron donor abilities are enhanced when in their delocalized π* excited states. This excited state delocalization allows the exciton migration to increase the frequency of interaction with the desired quencher24. There are two main classes used in these organic sensing: organic polymers, which bind the electron- poor explosives to the electron-rich polymer backbone, and inorganic polymers which use a combination of Lewis acid/Lewis base and electron-rich/poor interactions25.

Figure 20: Common Nitroaromatics

Figure 21: Electron-transfer Fluorescence Quenching Mechanism

As the research into fluorescence quenching studies has advanced, the field has ventured into the porous polymer area, including MOFs and porous organic frameworks

(POFs). Porous materials have an advantage as their frameworks contain flexible and adjustable pores which allow for guest molecules to freely interact with pore walls26. These studies have not yielded high sensitivity or selectivity of detection due to either poor analyte-host interaction in the POFs or chemical instability of the MOFs. Using these studies, research has shifted towards using COFs as COFs have been shown to contain high stability, crystallinity, and porosity.

The first example of using COFs as chemical sensors was performed in 2013 by

Jiang and coworkers, creating an azine-linked COF called Py-Azine (Figure 22)27. Using this COF’s high chemical stability, fluorescence quenching studies were performed by adding ppm of nitroaromatics in THF. The results showed high selectivity for TNT sensing

50 as the fluorescence was decreased dramatically (Figure 22)27. This started the research into using COFs as nitroaromatic sensors. Banerjee and coworkers recently published work synthesizing imide-linked COFs, which would be expected to be applicable to chemical sensing26. However, the actual results are poor due to aggregated π-stacked layers, poor electron mobility, and ineffective interaction with analytes. In their case, they used Liquid

Phase Exfoliation (LPE) technique to produce covalent organic nanosheets (2D CONs) which were found to be selective towards picric acid and TNT26. These two examples show that luminescent COFs can be used in a variety of ways as chemical sensors.

Figure 22: a) COF Py-Azine b) Fluorescence Quenching using PA of COF Py-Azine

In our lab, we envisioned the ordered synthesis of our DhaTas and DhaTae would lead to luminescent solids which could be used as chemical sensors. However, the lack of success in producing an ordered and optimized system for either COF caused us to take a step back and form the analog version of Vertex A and Vertex B. With the proof of the 51 desired bond formation, as well as an insight into whether the solids would maintain their fluorescent properties, we could advance the studies towards the 2D structures.

3.3 Results and Discussion

Because of the inability to form luminescent COFs with our attempted conditions, we decided to form the analogs of the COF systems. These analogs would allow for proof of the imine bond formation between salicylaldehyde and the amine-based vertices and would allow for full characterization. Both analogs were synthesized under similar reaction conditions (Scheme 12) by refluxing the aldehyde and amine in ethyl acetate overnight. Both solids would crash out during the reaction, as they were both highly insoluble in most organic solvents. Using ethanol and acetone as washes removed any unreacted salicylaldehyde present from the reaction. For the alkene analog, both 1H and

13C NMRs (Appendix A) were taken in DMSO, as well as a high resolution MALDI. Due to solubility issues with the alkyne analog, only a 1H NMR (Appendix A) and high resolution MALDI were used for characterization. By comparing all of the spectral data from both analogs to the published analog shown in Scheme 1113 from the Introduction, it was determined that both analogs had been successfully synthesized.

Scheme 12: Synthesis of Alkene and Alkyne Analogs

Due to the highly conjugated system, it was expected that the analogs would both be luminescent in solution and solid state. This property was strongly desired to eventually translate to the COFs and use them as chemical sensors. Placing both the alkene and alkyne analogs under long wave UV light showed a bright yellow fluorescence in both solution and solid state (Figure 23). Absorbance and emission studies were performed on both analogs in a 7 μM solution of DMSO as well as gathering the emission data of both analogs in the solid state (Figure 24 and 25). The maximum absorbance of the alkene analog was

375 nm, while the alkyne analog showed a slight blue shift and absorbed at a maximum wavelength of 360 nm. This trend followed in the emission data as well, as the alkene analog had a maximum intensity at 543 nm in solution while the alkyne analog had a maximum intensity at 540 nm in solution.

Figure 23: Images of Fluorescent a) Alkene and b) Alkyne Analogs

Figure 24: a) Absorbance of Alkene Analog in DMSO b) Emission of Alkene Analog in DMSO c) Emission in Solid State of Alkene Analog 54

Figure 25: a) Absorbance of Alkyne Analog in DMSO b) Emission of Alkyne Analog in DMSO c) Emission in Solid State of Alkyne Analog

The fluorescent data made us excited about the potentials of the COFs when they could synthesized, as the luminescent properties would be expected to transfer to the two-

55 dimensional structure as well. Because both 4-nitrophenol and 2,4-dinitrophenol were available in chemical recycling, the fluorescence quenching abilities of both analogs could be tested. If any quenching, especially significant quenching, was possible for these two nitrophenols, it would be expected that the much more reactive species (TNT and picric acid) would have significant quenching abilities in the systems. A 7 μM solution of both the alkene and alkyne analogs in DMSO was prepared. A 0.075 mM solution of either 4- nitrophenol or 2,4-dinitrophenol in DMSO was titrated into the cuvette at varying amounts, starting with 0.01 mL and increasing the amount until 0.2 mL per trial was added. Both analogs observed fluorescence quenching (Figure 26 and 27) at various degrees of success.

Using Equation 1, the percentage of fluorescence quenched could be determined where I0 was the initial intensity and I was the intensity after 16 equivalents of nitroaromatic was added.

Equation 1: η= (I0-I)/I0 X 100%

The alkene analog had a higher percentage of quenching for both the 4-nitrophenol and

2,4-dinitrophenol (Table 4), while the alkyne analog had less of a decrease in percentage of quenching between the two nitroaromatics (only 7% decrease compared to 25%).

Figure 26: a) Fluorescence quenching of alkene analog with 4-nitrophenol b) fluorescence quenching of alkene analog with 2,4-dinitrophenol

Figure 27: a) Fluorescence quenching of alkyne analog with 4-nitrophenol b) fluorescence quenching of alkyne analog with 2,4-dinitrophenol

Table 4: Percentage of Fluorescence Quenched

Analog Molarity of Molarity of 4- Molarity of Equivalents Percentage Analog Nitrophenol 2,4- Added Quenched (DMSO) (DMSO) Dinitrophenol (DMSO) Alkene 7.0 X 10-6 M 7.5 X 10-5 M NONE 16 88%

Alkene 7.0 X 10-6 M NONE 7.5 X 10-5 M 16 63%

Alkyne 7.0 X 10-6 M 7.5 X 10-5 M NONE 16 65%

Alkyne 7.0 X 10-6 M NONE 7.5 X 10-5 M 16 58%

Both analogs showed that they were able to undergo fluorescent quenching as nitroaromatics were added to the solution. Using the data above, it would be assumed that both would have substantial quenching in picric acid or TNT, as the addition of another nitro group to the aromatic ring increases the reactivity of the compound dramatically. It would be interesting to see the drop in equivalents necessary for quenching and if either analog was more selective to certain explosives. Certain conjugated systems can be fluorescence quenched dramatically regardless of the nitroaromatic present in the system28, while others have selective quenching for picric acid29 or TNT27. Determining exact equivalents necessary for the more explosive picric acid or TNT and selectivity for certain nitroaromatics could occur in future reactions.

3.4 Future Plans

To continue to test the ability of the fluorescence quenching with the more reactive, and more relevant, nitroaromatics it would be required to purchase both TNT and picric acid and perform similar experiments as the ones described above. However, due to the more reactive nature of these nitroaromatics, it would be assumed that less equivalents would be needed. The percentage quenched of both the alkene and alkyne analogs could be determined. While this data would be interesting to have, the most exciting chemistry would be the ability to translate these fluorescent quenching experiments into the two- dimensional COFs.

3.5 Conclusions

The synthesized novel vertices Vertex A and Vertex B could be refluxed with salicylaldehyde in ethyl acetate to form both the alkene and alkyne analogs. These analogs displayed the characteristics we would expect of the two-dimensional polymer; both of the analogs were highly fluorescent in the solid state. Using these fluorescent properties, the idea of using these polymer systems to undergo chemical sensing was tested. Using solutions of both 4-nitrophenol and 2,4-dinitrophenol, quenching experiments were run, finding that both the alkene and alkyne analog could be quenched using the less reactive nitroaromatics. This would lead us to believe that when a highly ordered and porous two- 60 dimensional covalent framework would be synthesized, it would be luminescent and have interesting properties as a chemical sensor.

Chapter 4: Experimental Data

4.1 General Methods

Unless stated otherwise all reagents were purchased from commercial sources and used without further purification. Dioxane and mesitylene was freshly distilled over calcium hydride. Tetrahydrofuran and methylene chloride were obtained from a solvent purification system (activated alumina columns) and used without further drying. All reactions were performed under an air atmosphere, unless stated otherwise. Thin-layer chromatography (TLC) was conducted with SiliCycle glass backed 60 Å UV254 plates

(0.25 mm) and visualized with UV lamps or KMnO4 stain followed by heating. Flash chromatography was performed using normal phase Aldrich 40-63 μm 60 Å silica gel.

Infrared spectra were recorded on a Thermo Scientific Nicolet iS5 with a iD7 diamond ATR attachment and are uncorrected.

UV-Vis/NIR absorbance spectroscopy were recorded on a Cary 5000 spectrophotometer using an internal DRA with stock powder cell holder to record % reflectance spectra. Emission spectra were recorded on a Cary Eclipse Fluorescence spectrophotometer equipped with a xenon flash lamp.

X-ray diffraction patterns were recorded on a Bruker D8 Powder X-Ray

Diffractometer employing Cu K(α)1 line focused irradiation at 40kV, 50 mA power and equipped with a Ge (111) monochromater. Samples were mounted on a zero background sample holder by dropping powders from a vial and then flattening them by firmly pressing the sample with a wide-blade spatula. No sample grinding was used prior to analysis. The holder was then placed on the mounting apparatus in the diffractometer. Data was collected after a 12 minute delay time using a 0.015° 2θ step scan from 1-34° with an exposure time of 2-3 sec per step.

Thermogravimetric analyses (TGA) were carried using a Perkin-Elmer thermal gravimetric analyzer 7 by heating samples in a platinum pan from 35 °C to 800 °C under nitrogen atmosphere at a heating rate of 10 °C min1 without an equilibration delay.

Scanning electron microscopy (SEM) was performed on a FEI Sirion FE-SEM.

Materials were deposited onto a film of wet colloidal silver paint on an aluminum sample stub and dried in a vacuum oven at 40 °C. The samples were coated with gold in a Leica

EM ACE600 coater, using rotation, to a depth of approximately 20nm. After coating the samples were imaged in the SEM at 5keV, without tilting, using both the secondary electron (SE) detector and the through lens detector (TLD).

Mass Spectra were obtained on a Bruker maXis Electrospray in conjuction with funding P30 CA016058 and NSF award 1040302 and on a Bruker ultrafleXtreme MALDI-

TOF/TOF MS with Proteineer fc II from the mass spectrometry facility at The Ohio State

University.

Surface area measurements were conducted on a Micromeritics ASAP 2020

Surface Area and Porosity Analyzer using ca. 20 mg samples. Nitrogen isotherms were generated by incremental exposure to ultra high purity nitrogen up to ca. 1 atm in a liquid nitrogen (77K) bath. Surface parameters were determined using BET adsorption models in the instrument software. Pore sizes distributions were determined using NLDFT model

(cylinder pore, N2-cylindrical pores-oxide surface with high regularization) in the instrument software (Micromeritics ASAP 2020 V4. 02).

1HNMR spectra were recorded in deuterated solvents on a Bruker Avance DPX 400

(400 MHz). Chemical shifts are reported in parts per million (ppm, δ) using the solvent as internal standard. 13CNMR spectra were recorded on a Bruker Avance DPX 400 (100

MHz) using the solvent as an internal standard. Solid state 13C NMR spectra were recorded on a Bruker Avance III HD Ascend 800 MHz.

4.2 Experimental Data for Linker, Vertex A, and Vertex B

2,5-Bis(bromomethyl)-1,4-dimethoxybenzene: This was synthesized based off a known procedure30. 1,4-dimethoxybenzene (10.00 g, 72.37 mmol) and paraformaldehyde (4.27 g,

144.75 mmol) was suspended in glacial acetic acid (50 mL) and HBr (30 mL) and heated at 50 oC for one hour. After cooling, the product was hydrolyzed in water (200 mL). The

64 white solid was collected by filtration, suspended in DCM (50 mL), and refluxed for 10 minutes. Once cooled, the white solid was filtered off and washed with water to give the final product (11.7 g, 36 mmols, 45% yield). FTIR (ATR, cm-1): 3045, 2932, 1506, 1402,

1 1203, 1037, 662; H NMR (400 MHz, CDCl3): δ 6.87 (s, 2H), 4.53 (s, 4H), 3.87 (s, 6H);

13 C NMR (100 MHz, CDCl3): δ 151.3, 127.5, 113.9, 56.3, 28.5.

Spectroscopic data are in accordance with those described in the literature.

2,5-dimethoxyterephthalaldehyde: Following a modified procedure12, 2,5-

Bis(bromomethyl)-1,4-dimethoxybenzene (2.22g, 6.85 mmols), and hexamethylenetetramine (1.93 g, 13.8 mmols, 2 equivs) was placed in DCM (21 mL) and refluxed for 3 hours. After cooling to room temperature, the product was further cooled to

0 oC for 10 minutes. The white solid was collected by filtration, suspended in water (25 mL), and refluxed for 2 hours. Once cooled, 3M HCl (5 mL) was added to the reaction and the yellow solid filtered off to give the pure product (0.53 g, 2.74 mmols, 40% yield).

-1 1 FTIR (ATR, cm ): 3127, 2871, 1671, 1394, 1209, 1028, 659; H NMR (400 MHz, CDCl3):

13 δ 10.51 (s, 2H), 7.46 (s, 2H), 3.95 (s, 6H); C NMR (100 MHz, CDCl3): δ 189.2, 155.8,

129.2, 111.0, 56.2.

Spectroscopic data are in accordance with those described in the literature.

2,5-dihydroxyterephthalaldehyde: Following a known procedure12, a mixture of 2,5- dimethoxyterephthalaldehyde (0.34 g, 1.75 mmols), HBr (48%, 14 mL, 8 mL/mmol), and acetic acid (99.5%, 16.6 mL, 9.5 mL/mmol) was refluxed for 14 hours. After cooling to room temperature, the mixture was poured into a 1:1 mixture of DCM/water (50 mL).

Using DCM (5 X 50 mL), the organic product was extracted off. The organic layers were combined and washed with water (2 X 30 mL), dried using sodium sulfate, and evaporated down to give a yellow solid (0.026 g, 0.16 mmols, 9% yield). FTIR (ATR, cm-1): 3270,

1664, 1475, 1280, 1124, 795, 670, 509; 1H NMR (400 MHz, DMSO): δ 10.31 (s, 2H),

10.28 (s, 2H), 7.22 (s, 2H); 13C (100 MHz, DMSO): δ 190.7, 153.2, 128.2, 115.7.

Spectroscopic data are in accordance with those described in the literature.

1,3,5-Tris(diethoxyphosphorylmethyl)benzene: Following a general procedure31, 1,3,5-

Tris(bromomethyl)benzene (1.00 g, 2.80 mmols) and triethylphosphite (1.33 mL, 7.76 mmols, 2.77 equivs) was refluxed for 4 hours. Once cooled, excess triethylphosphite was

66 evaporated off before being placed under vacuum. The final product (1.41 g, 2.66 mmols,

95% yield) was found to be a pale yellow oil pure enough to be taken on to the next step.

-1 1 FTIR (ATR, cm ): 2980, 1247, 1017, 943, 775, 520; H NMR (400 MHz, CDCl3): δ 7.12

(m, 3H), 3.98 (m, 12H), 3.08 (d, J=22 Hz, 6H), 1.23 (t, J=7 Hz, 18H); 13C NMR (100 MHz,

CDCl3): δ 132.3, 129.8, 62.1, 34.2, 16.4.

Spectroscopic data are in accordance with those described in the literature.

1,3,5-tris((E)-4-nitrostyryl)benzene: The Horner Wadsworth Emmons product was prepared with a modification of a known literature procedure32. To a flame dried flask under nitrogen, NaH (0.20 g, 8.33 mmols, 9 equivs) was weighed out and cooled to 0 oC.

1,3,5-Tris(diethoxyphosphorylmethyl)benzene (0.50 g, 0.95 mmols) in THF (3 mL) was added slowly to the NaH and stirred for 10 minutes. 4-nitrobenzaldehyde (0.86 g, 5.69 mmols, 6 equivs) in THF (12 mL) was added to the solution, causing a color change to deep purple. This solution was warmed to room temperature and stirred overnight, with the solution turning a deep brown color. After completion of the reaction, the reaction was quenched with water (100 mL) and 3M HCl added (100 mL), causing an orange solid to crash out of solution. This solid was filtered off, washed with acetone (400 mL), and dried.

The pure product (0.36 g, 0.70 mmols, 74% yield) was isolated as a yellow/orange solid 67 and carried on to the next step. FTIR (ATR, cm-1): 1589, 1504, 1334, 1108, 859, 692; 1H

NMR (400 MHz; DMSO): δ 8.29 (d, J=9 Hz, 6H), 7.98 (s, 3H), 7.91 (d, J=9, 6H), 7.61 (s,

6H).

4,4',4''-((1E,1'E,1''E)-benzene-1,3,5-triyltris(ethene-2,1-diyl))trianiline: Following a modified procedure33, 1,3,5-tris((E)-4-nitrostyryl)benzene (1.00 g, 1.93 mmols) was suspended in acetone (33 mL). A solution of NH4Cl (1.32 g, 24.6 mmols) in water (6 mL) was added to the suspension and heated to boiling. Once boiling, the heat was removed and zinc (2.64 g, 40.5 mmols) was added slowly in small portions. Once the reaction had subsided, another portion of zinc (1.32 g, 20.25 mmols) was added and the reaction refluxed overnight. The reaction was hot filtered to remove excess zinc and washed with ethyl acetate. This was evaporated down to an orange solid, basified using 3M NaOH, dissolved in ethyl acetate, and extracted (EtOAc; 5 X 30 mL) from water. The organic layers were combined, dried using sodium sulfate, and evaporated down to give a dark orange solid. After placing under vacuum for a day, the final product was found to be a light orange solid (0.79 g, 2 mmols, 96% yield). FTIR (ATR, cm-1): 3346, 3210, 2967,

1 1603, 1580, 1513, 1176, 961, 840; H NMR (400 MHz, (CD3)2CO): δ 7.52 (s, 3H), 7.34

(d, J=8 Hz, 6H), 7.18 (d, J=16 Hz, 3H), 6.96 (d, J=16 Hz, 3H), 6.68 (d, J=8 Hz, 6H), 4.81 68

13 (s, 6H); C NMR (100 MHz, (CD3)2CO): δ 149.4, 139.8, 130.2, 128.6, 127.2, 124.5, 122.9,

+ 115.3; HRMS (ESI): calculated for C30H27N3 [M+H] : 430.2277, found 430.2265.

tert-butyl (4-iodophenyl)carbamate: 4-iodoaniline (2.19 g, 10 mmols), di-tert-butyl dicarbonate (2.84 g, 13 mmols, 1.3 equivs) and NaHCO3 (1.27 g, 15 mmols, 1.5 equivs) was dissolved in a mixture of THF/water (1:1, 30 mL, 3 mL/mmol). This was stirred at room temperature for 16 hours. The mixture was poured into a 1:1 mixture of DCM/water and extracted with DCM (2 X 30 mL). The organic layers were combined, dried with sodium sulfate, and evaporated down. The crude product was purified using column chromatography on silica gel (9:1 hexanes/ethyl acetate) to give a white solid (2.81 g, 9 mmols, 88% yield). FTIR (ATR, cm-1): 3382, 2984, 1660, 1152, 1117, 1065, 1052, 818;

1 H NMR (400 MHz, CDCl3): δ 7.55 (d, J=9 Hz, 2H), 7.13 (d, J=9 Hz, 2H), 6.47 (s, 1H),

13 1.51 (s, 9H); C NMR (100 MHz, CDCl3): δ 152.6, 146.9, 138.0, 120.6, 85.8, 85.3, 81.1,

28.5, 27.6.

Spectroscopic data are in accordance with those described in the literature.

tert-butyl (4-ethynylphenyl)carbamate: Following a known procedure34, to an oven- dried flask containing a stir bar, tert-butyl (4-iodophenyl)carbamate (2.81 g, 8.80 mmols),

PdCl2(PPh3)2 (0.31 g, 0.44 mmols, 5 mol%), and CuI (0.17 g, 0.88 mmols, 10 mol%) was added under nitrogen. A co-solvent of THF and NEt3 (16 mL, 8 mL, 2:1) was added to the flask, followed by the trimethylsilylacetylene (1.04 g, 10.50 mmols, 1.50 mL, 1.2 equivs).

The reaction was then refluxed with stirring for four hours. Once cooled back to room temperature, the product was washed with DCM (100 mL) and ammonium chloride (50 mL). The product was extracted using DCM (3 X 40 mL), dried using sodium sulfate, and evaporated down to give a gray solid. The crude product was purified using column chromatography on silica gel (4:1 hexanes/ethyl acetate) giving a brown/yellow solid. This solid was dissolved in methanol (18 mL), K2CO3 added (2.98 g, 21.59 mmols, 3 equivs), and stirred at room temperature for 8 hours. This mixture was poured into water and extracted with ethyl acetate (3 X 20 mL), dried with sodium sulfate, and evaporated down to give a brown oil. The crude product was run through a short silica gel plug (DCM) to give a brown solid (1.45 g, 6.69 mmols, 76% yield). FTIR (ATR, cm-1): 3382, 3294, 2106,

1 1704, 1517, 1500, 1229, 1150, 1055, 833, 592, 534; H NMR (400 MHz, CDCl3): δ 7.44

(d, J=9 Hz, 2H), 7.35 (d, J=9 Hz, 2H), 6.55 (s, 1H), 3.05 (s, 1H), 1.55 (s, 9H); 13C NMR

(100 MHz, CDCl3): δ 152.3, 138.9, 133.0, 118.0, 116.4, 83.6, 80.9, 28.9.

Spectroscopic data are in accordance with those described in the literature.

tri-tert-butyl ((benzene-1,3,5-triyltris(ethyne-2,1-diyl))tris(benzene-4,1- diyl))tricarbamate: Following modifications of a known procedure35, to an oven-dried flask equipped with a stir bar, 1,3,5-tribromobenzene (0.53 g, 1.68 mmols), PdCl2(PPh3)2

(0.07 g, 0.10 mmols, 0.06 equivs), and CuI (0.38 g, 0.20 mmols, 0.12 equivs) was added under nitrogen. A co-solvent of NEt3 and THF (16.8 mL/8.4 mL, 2:1, 5 mL/mmol) was added to the flask along with a mixture of tert-butyl (4-ethynylphenyl)carbamate (1.45 g,

6.67 mmols, 4 equivs) in THF (8.4 mL, 5 mL/mmol). This mixture was refluxed under nitrogen for 12 hours. Once cooled to ambient temperature, the mixture was washed with ethyl acetate (50 mL) and ammonium chloride (50 mL). The product was extracted using ethyl acetate (3 X 30 mL), dried with sodium sulfate, and evaporated down to a dark brown solid. The crude product was purified using column chromatography on silica gel (9:1 hexanes/ethyl acetate; 4:1 hexanes/ethyl acetate) to give the pure product as a yellow solid

(0.68 g, 0.94 mmols, 58% yield). FTIR (ATR, cm-1): 2974, 1709, 1578, 1514, 1491, 1219,

1 1147, 831, 532; H NMR (400 MHz, CDCl3): δ 7.52 (s, 3H), 7.37 (d, J=9 Hz, 6H), 7.29 (d,

13 J=9 Hz, 6H), 6.47 (s, 3H), 1.46 (s, 27H); C NMR (100 MHz, CDCl3): δ 152.5, 138.9,

133.8, 132.7, 124.3, 118.3, 117.3, 90.5, 87.4, 81.1, 28.5.

4,4',4''-(benzene-1,3,5-triyltris(ethyne-2,1-diyl))trianiline: Tri-tert-butyl ((benzene-

1,3,5-triyltris(ethyne-2,1-diyl))tris(benzene-4,1-diyl))tricarbamate (0.20g, 0.28 mmols) was dissolved in DCM (1 mL) and cooled to 0 oC with stirring. TFA (1 mL) was added slowly and the purple solution allowed to stir at 0 oC for 10 minutes, followed by stirring at room temperature for 2 hours. The solvent was evaporated and the solid dissolved in ethyl acetate. The solution was basified with 3M NaOH and extracted using ethyl acetate

(5 X 40 mL), dried with sodium sulfate, and evaporated down to a brown/green solid. The solid was dissolved in DCM and 3M HCl added, causing a brown solid to crash out of solution. The solid was dissolved in water and the organic layer separated off. The DCM layer was washed with water (3 X 10 mL) and the aqueous layers combined. Using 3M

NaOH to basify the aqueous phase, the product was extracted using ethyl acetate (5 X 50 mL), dried with sodium sulfate, and evaporated down to give a black/green solid (0.098 g,

0.23 mmols, 84% yield). FTIR (ATR, cm-1): 3354, 3211, 2199, 1604, 1575, 1175, 824; 1H

NMR (400 MHz, CD3CN): δ 7.51 (s, 3H), 7.30 (d, J=9 Hz, 6H), 6.66 (d, J=9 Hz, 6H), 4.53

13 (s, 6H); C NMR (100 MHz, CD3CN): δ 149.7, 133.5, 132.8, 125.5, 114.8, 110.6, 92.3,

+ 86.0; HRMS (ESI): calculated for C30H21N3 [M+H] : 424.1808, found 424.1811.

1,3,5-Tris(4-aminophenyl)benzene: Following a known procedure36, 4- aminoacetophenone (0.49 g, 3.63 mmols) and p-toluenesulfonic acid (0.63 g, 3.63 mmols,

1 equiv) were stirred neat at 145oC for 16 hours. Once cooled, the product was dissolved in DCM and washed with NaHCO3. The product was extracted using DCM (4 X 35 mL), dried using sodium sulfate, and evaporated down. The crude product was purified using column chromatography on silica gel (1:1 hexanes/ethyl acetate; 3:2 ethyl acetate/hexanes) gave the light brown solid (0.13 g, 0.36 mmols, 10% yield). FTIR (ATR, cm-1): 3346,

1 3028, 2921, 1590, 1513, 1278, 1180, 820. H NMR (400 MHz, CDCl3): δ ppm 7.60 (s,

3H), 7.50 (d, J=9 Hz, 6H), 6.77 (d, J=9 Hz, 6H), 3.73 (br s, 6H); 13C NMR (100 MHz,

CDCl3): δ 144.9, 141.0, 131.0, 127.2, 121.9, 114.4.

Spectroscopic data are in accordance with those described in the literature.

4.3 Synthesis of Alkene and Alkyne Analog and DhaTas and DhaTae COFs

Alkene Analog: In an oven-dried flask equipped with a stir bar, 4,4',4''-((1E,1'E,1''E)- benzene-1,3,5-triyltris(ethene-2,1-diyl))trianiline (0.20 g, 0.47 mmols) and salicylaldehyde (0.17 g, 1.40 mmols, 0.15 mL, 3 equivs) was refluxed in ethyl acetate (10 mL) overnight under nitrogen. Once cooled, the solid was filtered off and washed with acetone (20 mL) to give an orange solid. FTIR (ATR, cm-1): 1614, 1565, 1279, 1189, 1149,

956, 845, 748, 520 ; 1H NMR (400 MHz, DMSO): δ 13.15 (s, 3H), 9.06 (s, 3H), 7.83 (s,

3H), 7.76 (d, J=8 Hz, 6H), 7.68 (dd, J=8 Hz, 3H), 7.51 (d, J=8 Hz, 6H), 7.36 (m, 9H), 6.98

(m, 6H); 13C NMR (150 MHz, DMSO): δ 163.4, 160.8, 147.7, 138.4, 136.4, 134.0, 133.0,

128.9, 128.8, 128.1, 124.5, 122.5, 119.7, 117.1; HRMS (MALDI): calculated for

+ C51H39N3O3 [M+H] : 742.8816, found 742.289.

Alkyne Analog: In an oven-dried flask equipped with a stir bar, 4,4',4''-(benzene-1,3,5- triyltris(ethyne-2,1-diyl))trianiline (0.20 g, 0.47 mmols) and salicylaldehyde (0.17 g, 1.40 mmols, 0.15 mL, 3 equivs) was refluxed in ethyl acetate (10 mL) overnight under nitrogen.

Once cooled, the solid was filtered off and washed with acetone (100 mL) to give a yellow solid. FTIR (ATR, cm-1): 2209, 1614, 1564, 1278, 1149, 833, 743, 530. 1H NMR (400

MHz, DMSO): δ 12.84 (s, 3H), 9.03 (s, 3H), 7.82 (s, 3H), 7.70 (d, J=9 Hz, 9H), 7.51 (d,

J=9 Hz, 6H), 7.43 (td, J=9 Hz, 3H), 6.99 (m, 6H); HRMS (MALDI): calculated for

+ C51H33N3O3 [M+H] : 736.8340, found 736.175.

DhaTas COF: To a pyrex tube (10 mL), Dha (19.9 mg, 0.12 mmol) and Tas (17.2 mg,

0.04 mmol) was dissolved in 1.3 mL mesitylene and 1.3 mL dioxane and 0.27 mL of 6M acetic acid was added. The mixture was sonicated for 5 min before being flash frozen at

77 K (liquid N2 bath). The mixture was then degassed by three freeze-pump-thaw cycles.

The tube was sealed and heated at 120 oC for 3 days. After 3 days, the solid was filtered off, washing with DCM, acetone, and THF. The orange solid was suspended overnight in

DCM, washed again with acetone, and suspended in three separate acetone washes for two days. The solid was dried and heated at 150 oC under vacuum for 12 hours to yield an orange solid in 63% (23.3 mg) isolated yield. FTIR (ATR, cm-1): 1609, 1578, 1495, 1312,

1210, 1152, 958, 840, 680, 528.

DhaTae COF: The synthesis of DhaTae was carried out in a similar manner to that of

DhaTas. To a pyrex tube (10 mL), Dha (10.8 mg, 0.065 mmol) and Tae (18.2 mg, 0.045 mmol) was dissolved in 0.9 mL mesitylene and 0.9 mL dioxane and 0.18 mL of 6M acetic acid was added. The mixture was sonicated for 5 min before being flash frozen at 77 K

(liquid N2 bath). The mixture was then degassed by three freeze-pump-thaw cycles. The tube was sealed and heated at 120 oC for 3 days. After 3 days, the solid was filtered off, washing with DCM, acetone, and THF. The orange/brown solid was suspended overnight in DCM, washed again with acetone, and suspended in three separate acetone washes for two days. The solid was dried and heated at 150 oC under vacuum for 12 hours to yield an orange/brown solid in 78% (22.7 mg) isolated yield. FTIR (ATR, cm-1): 2201, 1610, 1564,

1278, 1149, 833, 743, 530

DhaTab COF (unpublished): The synthesis of DhaTab was carried out in a similar manner to that of DhaTas. To a pyrex tube (10 mL), Dha (14.9 mg, 0.12 mmol) and Tab

(10.5 mg, 0.04 mmol) was dissolved in 1.0 mL mesitylene and 1.0 mL dioxane and 0.2 mL of 6M acetic acid was added. The mixture was sonicated for 5 min before being flash

76 frozen at 77 K (liquid N2 bath). The mixture was then degassed by three freeze-pump- thaw cycles. The tube was sealed and heated at 120 oC for 3 days. After 3 days, the solid was filtered off, washing with DCM, acetone, and THF. The brown solid was suspended overnight in DCM, washed again with acetone, and suspended in three separate acetone washes for two days. The solid was dried and heated at 150 oC under vacuum for 12 hours to yield a brown solid in 67% (16.9 mg) isolated yield.

FTIR was the same as reported compound13

References and Notes

(1) Yaghi; O. M., O'Keeffe, M.; Ockwig, N. W.; Chae, H. K.; Eddaoudi, M.; Kim, J.;

Nature, 2003, 423, 705

(2) Cote, A. P.; Benin, A.I.; Ockwig, N. W.; O’Keeffe, M.; Matzger, A. J.; Yaghi, O.

M.; Science, 2005, 310, 1166

(3) Xiang, Z.; Cao, D.; J. Mater. Chem A, 2013, 1, 2691

(4) Feng, X.; Ding, X.; Jiang, D.; Chem. Soc. Rev., 2012, 41, 6010-6022

(5) Cote, A. P.; EI-Kaderi, H. M.; Furukawa, H.; Hunt, J. R.; Yaghi, O. M.; J. Am.

Chem. Soc., 2007, 129, 12914

(6) Uribe-Romo, F. J.; Hunt, J. R.; Furukawa, H.; Klock, C.; O’Keeffe, M.; Yaghi, O.

M.; J. Am. Chem. Soc., 2009, 131, 4570

(7) Uribe-Romo, F. J.; Doonan, C. J.; Furukawa, H.; Oisaki, K.; Yaghi, O. M.; J. Am.

Chem. Soc., 2011, 133, 11478

(8) Ding, S. Y.; Wang, W.; Chem. Soc. Rev., 2013, 42, 548

(9) Du, Y.; Mao, K.; Kamakoti, P.; Ravikovitch, P.; Paur, C.; Cundy, S.; Li, Q.;

Calabro, D.; Chem. Commun., 2012, 48, 4606-4608

(10) Kandambeth, S.; Shinde, D. B.; Panda, M. K.; Lukose, B.; Heine, T.; Banerjee,

R.; Angew. Chem. Int. Ed., 2013, 52, 13052 – 13056

(11) Chen, X.; Addicoat, M.; Jin, E.; Zhai, L.; Xu, H.; Huang, N.; Guo, Z.; Liu, L.; Irle,

S.; Jiang, D. J. Am. Chem. Soc., 2015, 137, 3241

(12) Kretz, T.; Bats, J. W.; Lerner, H. W.; Wagner, M.; Z. Naturforsh., 2007, 62b, 66-

(13) Kandambeth, S.; Venkatesh, V.; Shinde, D.B.; Kumari, S.; Halder, A.; Verma, S.;

Banerjee, R.; Nat. Commun., 2015, 6, DOI:10.1038/ncomms7786

(14) Metz, B.; Davidson, O.; de Coninck, H.; Loos, M.; Meyer, L.; Intergovernmental

Panel on Climate Change. Special Report on Carbon Dioxide Capture and Storage,

Cambridge University Press, Cambridge, 2005, http://www.ipcc.ch/

(15) Stern, N.; Stern Review on the Economics of Climate Change, Cambridge

University Press, Cambridge, 2006

(16) D’Alessandro, D. M.; Smit, B.; Long, J. R.; Angew. Chem. Int. Ed, 2010, 35, 6058-

6082

(17) Chowdhury, P.; Bikkina, C.; Gumma, S.; J. Phys. Chem. C, 2009, 113, 6616

(18) Huang, N.; Chen, X.; Krishna, R.; Jiang, D.; Angew. Chem. Int. Ed, 2015, 10,

2986-2990

(19) Schlapbach, L.; Zuttel, A.; Nature, 2001, 414, 353

(20) Srepusharawoot, P.; Scheicher, R. H.; Moyses Araujo, C.; Blomqvist, A.; Pinsook,

U.; Ahuja, R.; J. Phys. Chem. C, 2009, 113, 8498–8504

(21) Yu, J. –T.; Chen, Z.; Sun, J.; Huang, Z. –T.; Zheng, Q. –Y.; J. Mater. Chem., 2012,

22, 5369

(22) Tylianakis, E.; Klontzas, E.; Froudakis, G. E.; Nanoscale, 2011, 3, 856

(23) Thomas, S. W.; Joly, G. D.; Swager, T. M.; Chem. Rev., 2007, 107, 1339–1386 79

(24) Toal, S. J.; Trogler, W. C.; J. Mater. Chem., 2006, 16, 2871–2883

(25) Germain, M. E.; Knapp, M. J.; Chem. Soc. Rev., 2009, 38, 2543–2555

(26) Das, G.; Biswal, B. P.; Kandambeth, S.;Venkatesh, V.; Kaur, G.; Addicoat, M.;

Heine, T.; Verma, S.; Banerjee, R.; Chem. Sci., 2015, 6, 3931-3939

(27) Dalapati, S.; Jin, S.; Gao, J.; Xu, Y.; Nagai, A.; Jiang, D.; J. Am. Chem. Soc., 2013,

135, 17310

(28) Ko, J. H.; Moon, J. H.; Kang, N.; Park, J. H.; Shin, H. W.; Park, N.; Kang, S.; Lee,

S. M.; Kim, H. J.; Ahn, T. K.; Lee, J. Y.; Son, S. U.; Chem. Commun., 2015, 51,

8781-8784

(29) Acharyya, K.; Mukherjee, P. S.; Chem. Commun., 2014, 50, 15788-15791

(30) White, K. A.; Chengelis, D. A.; Zeller, M.; Geib, S. J.; Szakos, J.; Petoud, S.; Rosi,

N. L.; Chem. Commun., 2009, 4506-4508

(31) Mahieu, G.; Grandidier, B.; Stiẻvenard, D.; Krzeminski, C.; Delerue, C.;

Langmuir, 2003, 19, 3350-3356

(32) E. Dίez-Barra, E.; Garcίa-Martinez, J. C.; Merino, S.; del Ray, R.; Rodriquez-

Lόpez, J.; Sánchez-Verdŭ, P.; Tejeda, J.; J. Org. Chem., 2001, 66, 5664-5670

(33) Miljanić, S.; Frkanec, L.; Meić, Z.; Žinić, M.; Eur. J. Org. Chem., 2006, 1323-

1334

(34) Flatt, A. K.; Yau, Y.; Maya, F.; Tour, J. M.; J. Org. Chem., 2004, 69, 1752-1755

(35) Yamaguchi, Y.; Ochi, T.; Miyamura, S.; Tanaka, T.; Kobayashi, S.; Wakamiya,

T.; Matsubara, Y.; Yoshida, Zen-ichi; J. Am. Chem. Soc., 2006, 128, 4504-4505

(36) Gattuso, G.; Grasso, G.; Marino, N.; Notti, A.; Pappalardo, A.; Pappalardo, S.;

and Parisi, M. F.; Eur. J. Org. Chem., 2011, 5696-5703 80

Appendix A: 1H and 13C NMRs for Synthesized Compounds

100

101

102

103

104

105