I. FUNCTIONALIZATION OF CRYTPOHANE CAGES FOR XENON MRI II. VANADIUM CATALYZED OXIDATIVE COUPLING OF SP3 C â•fi H BONDS TO HETEROARENES

The lack of resolution and selectivity in current imagining techniques such as xray, optical, and magnetic resonance imaging (MRI) has excelled the development of new biosensor technologies. Cryptophane A, a molecular cage composed of two cyclotriveratrylene (CTV) units connected by alkoxy bonds, can be turned into biosensors by attaching a target moiety capable of binding to a particular analyte. Cryptophanes can encapsulate xenon making it an attractive biosensor candidate for detection by hyperpolarized xenon-129 (HP-Xe) MRI. Further detection enhancement is achieved by using a technique called hyperpolarized chemical exchange saturation transfer (HYPER-CEST). One of the key challenges in developing Xebiosensors is the need for water soluble cryptophanes and their attachment to biomolecules that specifically bind physiological targets. The first manuscript entitled “Functionalization of Cryptophane cages for Xenon MRI” discusses the synthesis of cryptophane cages and their potential to be further functionalized. The manuscript centers on synthesizing cryptophanes that are water soluble functionalized with gold nanoparticles, which can eventually be further modified for imaging molecular events in vivo. The second manuscript, “Vanadium Catalyzed Oxidative Coupling of sp C-H Bonds to Heteroarenes” discusses oxidative aminomethyaltion of imdizolpyridines. This manuscript proposes a vanadium catalyzed oxidative coupling of imidazopyridines with N-methylmorpholine oxide which serves as both sp hybridized coupling partner and the oxidant. The reaction was optimized and performed with a variety of substrates to yield on a library of aminomethylated products. We investigated the mechanism and propose that a Mannich-type mechanism is responsible for the formation of the product.


INTRODUCTION
The ability to make a definitive medical diagnosis and follow the onset of certain diseases is vital in treating patients. Current imaging techniques, such as CT, PET, and MRI, possess different limitations that lower their overall effectiveness ( Table 1). 1 Computed tomography, a non-invasive imaging method, generates an image by using X-rays that are absorbed by tissues. 2 Though inexpensive, CT is incapable of molecular imaging and exposure to hazardous X-ray radiation renders CT a less attractive method for imaging. Positron emission tomography (PET) and single photon emission computed tomography (SPECT) are capable of anatomical and molecular imaging. Magnetic resonance imaging (MRI) is widely used for scanning deep tissue in the diagnosis of human diseases. Though this technique allows for high spatial resolution, only a small percentage of hydrogen nuclei align with the magnetic field, rendering this technique inherently insensitive. To increase signal intensity, contrast agents, such as gadolinium or iron oxide-based particles are often employed. These contrast agents work by changing the T1 and T2 relaxation times of water, thereby enhancing signal intensity. 3 However, recent findings suggest that development of nephrogenic systemic fibrosis (NSF) in patients can be triggered by the administration of gadolinium agents. 4 Other MRI contrast agents such as manganese and lanthanides, are toxic even at low concentrations and interfere with neuronal functions by blocking ion-channels. 5 As a result, the need for non-toxic and non-proton MRI contrast agents needs to be explored. These issues may be circumvented by incorporating the inert noble gas xenon-129 ( 129 Xe) as a molecular probe, since it is not present in the body, it is nontoxic, and it can be hyperpolarized to enhance its sensitivity. Herein, the use of xenon-129 in MRI, its ability to be hyperpolarized, Hyper-CEST techniques that allow for better detection, and the utilization of water soluble cryptophanes as hosts for Xe will be discussed.

Hyperpolarized Xenon
Xe exists as two NMR active isotopes, a spin-1/2 nucleus 129 Xe and spin-3/2 nucleus 131 Xe. Only the former is capable of being hyperpolarized. 6 The ability to hyperpolarize 129 Xe is advantageous, as it allows for a 10,000-times signal enhancement as a result of the alignment of more nuclear spins with the magnetic field. Where typical proton MRI has only a 15 ppm chemical shift range, 129 Xe has range over 200 ppm, which allows for chemical shift imaging. 7 Solubility of xenon in lipid-rich tissue and blood allows for imaging of brain, lung and other regions. 8 Spin exchange optical pumping (SEOP) is a three-step process used to hyperpolarize 129 Xe. The first step involves the generation of circular polarized light by a Ti:sapphire laser (Figure 2). In the second step, a droplet of Rb is heated in a glass cell, while exposed to a magnetic field, to produce a vapor, which then absorbs circular polarized light, driving the selective excitation of the D1 transition state. The third step produces hyperpolarized 129 Xe. When polarized Rb collides with 129 Xe, it transfers the spin of its valence electron to the 129 Xe nucleus. As a result of this dipole interaction, xenon is now predominately in the state that is aligned with the external magnetic field. 9 Once hyperpolarized (HP) 129 Xe is achieved, it can be delivered in vivo to be imagined. The delivery of HP -129 Xe to the body can be achieved through inhalation or injection. Inhalation of hyperpolarized 129 Xe permeates through lung tissue and blood, allowing for exploration of certain high concentration of characteristics of lung function. 10 Studies have shown that inhalation of HP-129 Xe results in highly spatially resolved images (Figure 2  The polarization lifetime of 129 Xe is dependent on its molecular environment.
In oxygenated blood T1 = 13 s, deoxygenated blood T1 = 4 s, and T1 ≈ 100 s in deuterated saline solution. 6 While long polarization lifetimes are advantageous, they require longer time for imaging, which still does not result in optimal detection. Use of a host for 129 Xe allows for targeted imaging of molecular events. Such targeted molecular imaging via 129 Xe MRI is usually accomplished by employing molecular probes that contain cageshaped structures that are capable of encapsulating xenon atoms. The most common molecular cage that is used for this task is called cryptophane A.

Cryptophanes
Cryptophanes are molecular cages composed of two cyclotriveratrylenes (CTVs) that are connected by alkoxy linkers (Figure 3). 12 The lipophilic character of cryptophane's cavity allows for reversible binding of small, non-polar molecules, like xenon. The affinity and exchange rate of xenon depends on the cavity size of the cryptophane, which is varied by changing the number of carbons in the alkoxy linkers that join the two CTVs ( When 129 Xe is bound in the cavity of a cryptophane cage, it produces a unique chemical shift distinct from free 129 Xe in the 129 Xe NMR spectra. The typical chemical shift of Xe@cryptophane complexes is demonstrated in Figure 4, where the larger the cavity size, the more the peak corresponding to encapsulated 129 Xe is shifted up-field. 6 This trend is broken by Xe@cryptophane-1,1,1 with a chemical shift of 31.1 ppm, whereas cryptophane A is at 65 ppm. It has been postulated this anomalous result is derived from the absence of methoxy groups, which suggests that modifying cages with electron withdrawing or donating groups can alter the chemical shift of Xe@cryptophanes. 6 Functionalizing cryptophanes with water solublizing groups, which are necessary for use in vivo, results in a higher binding affinity of 129 Xe ( Table   2). 13,15  One of the most promising applications of cryptophanes that encapsulate xenon is their ability create biosensors. This can be achieved by conjugating cryptophanes to a targeting ligand, which is capable of binding to a specific biological target, such as receptors on tumors or sites of inflammation. For example, Dmoschowiski recently synthesized a cryptophane that was tethered to the carbonic anhydrase (CA)-specific ligand, benzenesulfonamide. 17 When the benzensulfanamide cryptophane binds to CA isozomes I or II, it produces a distinctive chemical shift from the biosensor when it is free in solution. 6 Despite the synthesis of new water-soluble and ligand-ligandfunctionalized cryptophanes, more sensitive techniques for detection of HP-129 Xe needs to be addressed.

Hyper-CEST
Despite the high affinity of cryptophanes to bind 129 Xe, long acquisition times are necessary to achieve well-resolved spectra of bound 129 Xe. This issue is circumvented by incorporating a method called chemical exchange saturation transfer (CEST) to enhance detection of hyperpolarized 129 Xe, when used in conjunction with hyperpolarized 129 Xe this technique is called Hyper-CEST. Xenon is an ideal candidate for this method due to its long relaxation times and large chemical shift difference between bound and free 129 Xe. 14 Instead of directly detecting the 129 Xe that is bound in a cryptophane, HyperCEST detects the depletion of free 129 Xe. In order to do this, an off-and onresonance spectra is acquired. The off-resonance spectrum is a reference spectra obtained by applying a continuous wave saturation (Figure 5 A). The on-resonance spectrum is obtained by applying a radio frequency pulse tuned to 129 Xe@cryptopane signal (Figure 5 B). 18 The difference of the on and off resonance spectra results in signal only arising from the depletion of the free 129 Xe peak. 19 The exchange of 129 Xe from inside to outside the cryptophanes results in a reduction of the free 129 Xe peak.

General Introduction
While conventional magnetic resonance imaging (MRI) allows for anatomical imaging, it is inherently insensitive due to the minimal amount of protons capable of aligning with the instrument's magnetic field. Minimization of the signal to noise issues that are inherent in 1 H MRI, can be achieved by using 129 Xe as a molecular probe, as it is not present in the body and has the ability to be hyperpolarized, thereby giving a 10,000-times stronger signal. 1 When 129 Xe is encapsulated by a porous cage-shaped molecule, such as a cryptophane, a new signal, which is well resolved from the chemical shift of free 129 Xe, is produced in the 129 Xe NMR spectrum. In principle, magnetic resonance imaging techniques can translate this unique signal into a high-resolution 3D image. If these cryptophanes are attached to target-specific ligands, the cryptophane could serve as a biosensor, where its unique signal would be localized at a specific site within the body. This cryptophane-ligand biosensor will allow for the specific detection of molecular processes or receptors.
Using hyperpolarized chemical exchange saturation transfer (HyperCEST) provides another source of signal amplification allowing for detection of submicromolar concentrations of physiological targets. By selectively irradiating the 129 Xecryptophane signal, the loss of magnetization of the larger bulk Xe peak is easily detected. This indirect detection method allows for very low detection limits, indicating that HyperCEST imaging could be used for true molecular imaging. 1  Gold nanoparticles (AuNPs) are capable of being functionalized with a myriad of organic or biological ligands that bind selectively to small molecules or biological targets. This is achieved by varying the type of capping ligand used and fast or slow addition of NaBH4. 5 The use of gold nanoparticles in imaging has grown over the years due to their unique optical properties. Along with the use of gold nanoparticles for various modes of imaging, they can be also used for drug delivery. 7 For instance, Kim et al., developed multifunctional gold nanoparticles capable of anticancer therapy and computed tomography (CT) imaging of cancer cells. 8 Further benefits of using gold nanoparticles is that they are nontoxic and non-immunogenic making them ideal drug delivery scaffolds. 9 Therefore, tethering both cryptophanes and biological ligands to gold nanoparticles would allow for targeted imagining via HP-129 Xe MRI.

Scheme 1: Click reaction of cryptophane A to AuNPs (where n = # carbons)
The use of gold nanoparticles to facilitate drug delivery and imaging when tethered to a cryptophane is unprecedented. The proposed study will focus on efforts towards synthesizing a biosensor consisting of a cryptophane tethered to gold nanoparticles via a thiol linker. The addition of protected thiol linkers to cryptophanes will be accomplished via the copper-catalyzed Huisgen cycloaddition, the so-called azide-alkyne click reaction (Scheme 1). After deprotection of the thiol group, gold nanoparticles capped with thiol ligands will be added to AuNPs that have been previously capped with labile thiol ligands. These thiols will undergo ligand exchange with the thiols on the cryptophanes to yield the desired nanoparticles decorated with cryptophanes. The amount of cryptophane loaded onto the nanoparticle surface will be quantified by infrared spectroscopy.

Cryptophanes and Gold Nanoparticles
Over recent years, AuNPs have become increasingly popular in imaging due to their unique optical properties and ability to functionalize the surface with a myriad of organic compounds or biological ligands that selectively bind small molecules or biological targets. 5 Additionally, gold nanoparticles are can be used as drug delivery vehicles, as gold nanoparticles have been shown to be non-toxic. Browen et al., achieved this by tethering the anti-cancer drug oxaliplatin to gold nanoparticles resulting in direct site drug delivery. 6 Tethering gadolinium (Gd) chelates to AuNPs has the capability to enhance the contrast signal in 1 H MRI images. This enhancement is a result of the large number of Gd chelates on each AuNP, allowing for a more pronounced enhancement in contrast near the site of the AuNps. 7 However, as previously discussed, the toxicity of Gd chelates renders them as less attractive modes of imaging. The capability to achieve signal enhancement utilizing gold nanoparticles initiated our investigation into developing AuNP-based 129 Xe biosensors using a cryptophane.
As originally reported by Dmochowski, the tripropargyl cryptophane-2,2,2 (2 refers to the number of carbons in the alkoxy bridge between cage top and cage bottom), derivative can be accomplished in ten steps. 4 The top of the cage, 1.3, was synthesized in three steps (Scheme 2). Vanily alcohol is protected with an allyl group in order to avoid polymerization in the cyclization step (Scheme 2). Cyclization was achieved through a Friedel-Crafts cyclization using HClO4 and MeOH. Deprotection of the allyl group required excess amounts of Pd/C and TsOH.

Scheme 2: Synthesis of cage top
Though the number of carbons in the propargyl linker can be varied, incorporating ethylene bridges results in a cryptophane cavity size that allows for optimal binding and exchange of 129 Xe. Consequently, we set out to synthesize a cryptophane cage containing ethylene bridges between the two cyclotriveratrylenes.
The linker containing the propargyl group was synthesized in five steps (Scheme 3). resulting in analytically pure product. Since there is more of a possibility for side products during the first reaction, flash chromatography was kept as a purification procedure for the first step. When using recrystallization for the second step, a 77% yield was obtained, while using flash chromatography for the same step gave a 75% yield. Recrystallization for the reduction step resulted in comparable yields to when flash chromatography was used. Though similar yields were obtained, recrystallization is less time consuming and requires less solvent, so it became our preferred purification method.

Scheme 3: Synthesis of propargyl linker
Once the linker was synthesized, it was added in excess to the cage top and heated for two days under nitrogen at 55 ºC in the presence of the base, cesium carbonate (Scheme 4, 1.9). The final step in the synthesis of the conjugatable cryptophane (1.10) was the cyclization of the cage bottom, through a second Friedel-Crafts reaction, mediated by perchloric acid (Scheme 4). The overall yield of this synthesis is rather low due unwanted polymerization.

Scheme 4: Synthesis of tripropargyl cryptophane A
The conversion of tri-propargyl Cryptophane A to a biosensor can be achieved via copper(I)-catalyzed Hugsien cycloaddition "click" of a targeting moiety. 2 Our aim was to decorate gold nanoparticles (AuNPs) with cryptophane cages, which potentially could increase the cryptophanes' water solubility improving their applicability for invivo imaging. Using gold nanoparticles also acts as a second scaffold to which biological ligands could be tethered.
There are multitudes of ways to synthesize gold nanoparticles. For our purposes, the Brust-Schiffron method was the most attractive, as the resulting tiol capped AuNPs are easily re-dispersed in organic solvents, which we envisioned to be an advantage for subsequent modification reactions. Additionally, the dodecanethiol capping ligand used in the AuNP synthesis is known to readily undergo ligand exchange of a variety of thiols. 5 To carry out the click chemistry with the alkynes on cryptophanes, an azide-thiol linker needed to be synthesized. We attempted the synthesis of two azide-thiol linkers (Scheme 5). In order to avoid poisoning of the copper catalyst and unwanted side products during the click reaction, the thiol group needed to be protected. This was achieved using trityl chloride to protect 2-aminoethanelthiol (1.11, Scheme 5) and acetate to protect mercaptopropionic acid (1.14). The synthesis of the azide portion of the two possible linkers was carried out by refluxing 3-bromopropionic acid or 2bromoethylamine with sodium azide. Both reactions required extended reaction times and, unfortunately, were low yielding. The synthesis of both azide-thiol linkers was then achieved using peptide-coupling conditions. To achieve optimal yields when carrying out the click reaction with cryptophanes, the reaction was first optimized using the cryptophane linker, 1. These optimal conditions were modified to achieve the click reaction at all three alkynes on the cryptophane. This was achieved by using 40 mol% Cu(I)Br and 40% tren Another pathway to functionalize nanoparticles was attempted by performing ligand exchange between dodecanethiol capping ligand and 11-bromoundecylthiol, which was converted into an azide via an SN2 reaction with sodium azide. The formation of the azide-decorated AuNPs was verified via NMR, allowing for click chemistry to be carried out with tri-propargyl cryptophane using Cu(I)Br and tren ligand. While the click reaction seemed to be successful, the AuNPs aggregated.

Conclusion
Future work involves using different deprotection strategies and/or protecting groups. Additional routes for AuNP functionalization could be accomplished by synthesizing a different cryptophane starting material by substituting one of the propargyl linkers in the cryptophane for a linker that is functionalized with thiol group.

Compound 1.7 [3 propargyloxy-4-(2-iodoethoxy)phenyl]methanol:
In a round bottom flask, sodium iodide (4.86 g, 32.46 mmol) and 1.6 (2.30 g, 8.11 mmol) were dissolved in acetone and refluxed overnight. Solvent was removed under vacuum and solid was dissolved in DCM and subsequently washed with sodium thiosulfate (2 x 50 mL), water (50 mL) and brine (2 x 50 mL). The organic layer was dried over MgSO4, filtered and solvent was removed under vacuum to yield pure 1.7 (2.44 g, 98%) as a white solid. All spectral information matched literature values. 1

Compound 1.15: 2-azidoethylamine
Bromoethylamine (2.5 g, 12.20 mmol) was dissolved in 10 mL of water, followed by the addition of NaN3 (2.38 g, 36.36 mmol) and was refluxed for 21 hrs. Reaction was then cooled to 0°C and KOH (4g) was slowly added. The reaction mixture was then extracted using diethyl ether (3 x 10 mL) and organic layers were collected, dried over MgSO4 and concentrated until 10 mL was left (0.41g, 40%). All spectral information matched literature values. 5

Compound 1.16
To round bottom containing 1.15 (0.67 g, 4.5mmol) and 1.14 (0.43 g, 5.00 mmol) were dissolved in DMF. EDC (2.96 g, 15.4 mmol) and HOBT (1.25 g, 8.19 mmol) were then added and the reaction was stirred at room temperature for two days.
Reaction was diluted with water and extracted with DCM (3 x 50 mL

Compound 1.18
To a vial equipped with a stir bar, 1.10 (0.11 g, 0.11mmol) and 1.13 (0.14 g, 0.33 mmol) were dissolved in THF (5 mL  and concentrated down to ~5 mL, which was then mixed with 200 mL of EtOH and kept at -18° C for 4 hours. The dark brown precipitate was then filtered off and redissolved in 5 mL of toluene, which was then precipitated again using 100 mL EtOH.
The suspension was then filtered and washed with copious amounts of EtOH to remove any free dodecane thiol, product was dried under vacuum overnight.
Transmission electron miscopy validated the presence of AuNPs.

Compound 1.21: Azide-AuNPs
The 11-bomoundecylthioacetate (0.10 g, 0.32 mmol) was dissolved in 1 mL of ethanol. NaOH was added (0.1 mL) and reaction mixture was refluxed for 2 hrs. The reaction was then neutralized with 6 mL of 2 M HCl. The mixture was transferred to separatory funnel and then 10 mL of diethyl ether was added followed by the addition of with 5 mL of H2O. There are numerous methods to oxidatively couple two sp 2 hybridized carbons in high yields. 2 The coupling of an sp 3 to an sp 2 hybridized carbon is much less common due to the less reactive nature of sp 3 hybridized C-H bonds compared to that of sp 2 ¬C-H bonds. 4 In order to achieve sp 3 C-H bond activation, reactive directing groups are often needed. 3 Another difficulty is the potential for β-hydride elimination after C-H activation of sp 3 bonds. 4 The oxidative coupling of a sp 3 C-H bond to a sp 2 C-H bond has received little attention, 4,5,6,7 so there is still a great need to invent new ways of oxidatevly coupling sp 3 -sp 2 carbon bonds in an efficient and green way.
Pharmaceuticals such as necopidem, saripidem, and zolpidem contain a substituted imidazo[1,2-a]pyridines back bone (Figure 1). 8  The synthesis of the starting material is accomplished in one step via an imminium formation followed by a nucleophlic cyclization (Scheme 2). 9 The nature of the substituents on pyridine or bromoacetophenone did not greatly affect the starting material yields 1.

Results and Discussion
In order to achieve optimum conditions we screened solvent, catalys, catalyst loading, time, and equivalents of NMO. Ethanol, toluene, and tetrahydrofuran were initially tested as alternative solvents ( which still did not result in comparable yields to when VO(acac)2 was used under the same conditions (entry 14) . The best yield was achieved using 10% of (VO(acac)2), 5 equivalents of NMO for 6 hours in methylene chloride (entry 12).
However by changing the solvent from methylene chloride to 1,4 -dioxane and increasing the catalyst loading VO(acac)2, we were able to avoid any of 3 from forming and achieve an 80% yield. Performing a GC time study gave further insight into when the reaction would produce the least impurity (see Chart 1). The starting material was always present in small amounts as was 3. We stopped the reaction at 12 hours despite

61
the presence of starting material to avoid product decomposition. As such we decided to perform our substrate scope using entry 13 as our optimized conditions.
Performing this reaction using substrates with electron withdrawing groups on the para position of the phenyl substituent resulted in less favorable yields or did not react. While electron donating groups in the para position lead to higher yields. Looking  Previous studies of vanadium oxidative coupling of structures similar to 1 propose a radical mechanism, 10 while others theorize product is formed by way of a 63 Mannich-type reaction. 5 We propose a Mannich type mechanism is responsible for the formation of our product (Scheme 4). This was determined by running a Mannich reaction using 1, formaldehyde, and morpholine to make the imminium in situ resulting in yields up to 98% (scheme 3). Running our normal reaction conditions in the presence of TEMPO, a radical inhibitor, did not prevent product from forming further proving this reaction a Mannich-type mechanism is responsible.

Scheme 3-Pure Mannich Conditions
As such we believe the product is formed by the following mechanism (scheme 4). The N-methyl morpholine oxidizes the VO(acac)2 catalyst resulting in the formation of the imminium ion. The vanadium species extracts a proton from 1 which then attacks the immionium ion forming our product 2. The elimination of water from vanadium regenerates the catalyst. We believe the main impurity is formed when the imminiom reacts with one of the ligands from the vanadium catalyst.
64 Scheme 4-Proposed mechanism of product and impurity

Conclusions
The vanadium cataylzed oxidative coupling of substituted imidazole pyridines to nmethylmorpholene was achieved in yields up to 90%. Despite the ability to produce this product using true mannich conditions we believe this to useful method to demonstrate oxidative coupling.
2-amino pyridine (1.3 g 13.8 mmol), 1-bromoacetophenone (2.19 g, 11.1 mmol), sodium bicarbonate (1.45 g, 17.25 mmol) and ethanol (9 ml) were added to a flask. A stir bar was added and the reaction was stirred at 60°C for 4hrs. The reaction was monitored by TLC at a 50:50 ethyl acetate: hexanes solvent system. After reaction was completed, the ethanol was rotavapped off, and 50mL of water was added in portions to dissolve the solid. The water was then extracted with two 50mL portions of ethyl acetate and dried with sodium sulfate. The filtrate was rotavapped off. A column was ran (80 g column on a gradient) to yield 1.66 g (78%) of product. All spectral information matched literature values. 1 Representative Procedure for Synthesis of substituted imidazole pyridines with N-methylmorpholene Imidazo[1,2-a]pyridine (1 mmol), NMO (5 mmol), VO(acac)2 (20 mol%), were dissolved in 1,4-dioxane (4 mL), and refluxed for 12 hrs. Upon completion the reaction was evaporated and diluted with water (10mL) and extracted with EtOAc (2x10mL). Organic layers were dried of Na2SO4 and concentrated. Purified by column chromatography using 4/1 EtOAc:Hex.