Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. The Study of Natural and Unnatural Peptides: Changing Medicinal & Structural Properties A thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Chemistry Suraj Patel Supervisor: Gareth J. Rowlands Co-Supervisor: Preet Singh 2021 V “When we hit our lowest point, we are open to the greatest change.” Avatar Aang VI VII Abstract Nature creates the foundations of life by linking together small, yet versatile, building blocks. Using this principle it can build information storage-molecules like DNA or functional catalyst like peptides and proteins. With the aid of non-covalent interactions these molecules form elegant 3D structures that allow them to have a range of functions and to complete a variety of tasks. However, the natural synthesis of the molecules are often limited by the biological resources available. As chemists, we have limitless choice in building blocks at our disposal, allowing us to make tweaks until we get the most efficient outcome. The purpose of this study is to see how chemists are able to improve on nature through the use of synthetic chemistry. This was attempted through two strands of research. The first strand tried to improve the efficiency of a naturally occurring peptide, opiorphin, by altering the chemical structure without inhibiting its analgesic potency. While the second project investigated a class of synthetic peptide, known as a foldamer, by using a unnatural amino acid building block. A viable design of a prodrug for opiorphin was proposed, which requires further research but has begun to show some promise. While the synthesis of a peptide was not achieved in the second project, issues seen during the synthesis of a dimer have been highlighted and alternative routes have been proposed. In addition, a novel substitution pattern for a cyclophane product has been reported. VIII IX Acknowledgements Firstly, I would like to thank my supervisor, Assoc. Prof. Gareth J. Rowlands. From the first lecture in 123.101, your enthusiasm for synthetic chemistry has been contagious and has motivated me to become a better chemist. The dedication you show your students hugely surpasses what is expected, and I feel incredibly fortunate to have had you as a mentor. Your insightful comments, suggestions, and knowledge have been greatly appreciated. In particular, your anal retentiveness for ChemDraw structures has been a blessing and curse, which I am grateful has been passed onto me. I must also thank my unofficial supervisor, Leonie Etheridge. You have taught me nearly everything I know about wet chemistry, lab maintenance and health and safety procedures. Your constant support, quiet words of wisdom and walks to the car park have been much appreciated over the years (and sorely missed in this last year). This work would not have been possible without some key staff members at Massey that have helped me through the years. Thank you to Graham Freeman for his constant sass and synthetic expertise. Dr Pat Edwards for his NMR guidance. David Lun for running high-res mass spectroscopy samples. Tyson Dais for collecting X-ray crystal structures and interpreting the data. Natisha Magan for turning me into a wine connoisseur and teaching me every wine is “a bit oaky at the back of the palate”. To the ‘Smart People + Suju’ gang, thank you for declining my invites to the pub every time I asked, this thesis would have never been completed without your heart-breaking rejections. Thank you to the Rowlands, Plieger and Filichev groups for their advice. I would particularly like to thank Tyson and Siddle (kinda Sam too, I guess) for their vast knowledge and proofreading skills. Oh, and Marryllyn for providing me with some form of a social life over the last two years. Lastly, I would like to thank my friends outside of university that have supported me through the last two years. To begin, Ditchell, for no longer living in Palmerston North and not wanting to see me, as again, I would not have been able to complete this research with you around. Secondly; Callum, Leila, Zara, Alex, Amy, and Dakota for supporting me through what have been a difficult set of years. The friendship and distractions you have provided me over the years has been thoroughly enjoyed, and I would not be here without your company. X Contents Abstract ......................................................................................................................... VII Acknowledgements ......................................................................................................... IX List of Abbreviations ................................................................................................... XIII Introduction ...................................................................................................................... 1 1.1 Natural vs Unnatural Amino Acids ......................................................................... 1 1.2 Natural Peptide - Opiorphin .................................................................................... 2 1.2.1 Blood-Brain Barrier .................................................................................................................... 3 1.2.1 Prodrugs ...................................................................................................................................... 5 1.3 Unnatural Amino Acids – Foldamers ..................................................................... 8 1.3.1 Non-Covalent Interactions .......................................................................................................... 9 1.3.2 Recent Examples of Foldamers ................................................................................................ 11 1.3.3 Aliphatic Foldamers .................................................................................................................. 15 1.3.4 Aromatic Foldamers ................................................................................................................. 16 1.4 [2.2]Paracyclophane .............................................................................................. 20 1.4.1 [2.2]Paracyclophanes as Asymmetric Catalysts ....................................................................... 23 1.4.2 [2.2]Paracyclophanes as a Scaffold for New Materials ............................................................ 24 1.4.3 [2.2]Paracyclophane as Amino Acids ....................................................................................... 27 1.5 Aims of this Research ........................................................................................... 29 1.5.1 Aim of Natural Amino Acid Research Strand .......................................................................... 29 1.5.2 Aim of Unnatural Amino Acid Research Strand ...................................................................... 31 Results and Discussion .................................................................................................. 33 2.1 Section 1 – Natural amino acids ........................................................................... 33 2.1.1 Synthesis of Hydrate Derivatives ............................................................................................. 33 2.1.2 Synthesis of Protected Arginine ............................................................................................... 38 2.1.3 Test Coupling Reactions ........................................................................................................... 42 2.2 Section 2 – Unnatural Amino acid ........................................................................ 47 2.2.1 Part 1 – Synthesis of the Amino Acid Precursor ...................................................................... 47 2.2.2 Part 2 – Synthesis of a Dimer ................................................................................................... 55 2.3 Section 3 – Learning from our First Attempts to Synthesise a Foldamer ............. 61 2.3.1 Resolution ................................................................................................................................. 63 XI 2.3.2 Aromatic Linker ....................................................................................................................... 67 2.4 Section 4 – Synthesis of [2.2]Metaparacyclophane Derivatives ........................... 79 Future perspective and conclusion ................................................................................ 96 3.1 Natural Amino Acid – Opiorphin ......................................................................... 96 3.2 Unnatural Amino Acid – Foldamer ....................................................................... 97 3.3 [2.2]Metaparacyclophane Derivatives .................................................................. 99 Experimental Methods ................................................................................................. 102 4.1 Natural Amino Acid ............................................................................................ 103 4.2 Unnatural Amino Acid ........................................................................................ 109 4.3 [2.2]Metaparacyclophane Derivatives ................................................................ 134 References ..................................................................................................................... 139 Appendix ....................................................................................................................... 149 XII XIII List of Abbreviations 22pc [2.2]Paracyclophane HOMO Highest occupied molecular orbital 2D Two-dimensional HPLC High-performance liquid chromatography AAP Alanine aminopeptidase HRMS High Resolution Mass Spectrometry Ac Acetyl Hz Hertz Arg Arginine IR Infrared aq Aqueous J Coupling constant BBB Blood-brain barrier LUMO Lowest unoccupied molecular orbital BINAP 2,2’-bis(Diphenylphosphino)- 1,1’-binaphthyl Me Methyl Boc tert-Butoxycarbonyl Mp Melting point bs Broad singlet m/z Mass to charge ratio CIP Cahn-Ingold-Prelog NDI 1,4,5,8- naphthalenetetracarboxylic acid diimide COSY Correlation Spectroscopy NEP Neprilysin d doublet NHS N-Hydroxysuccinimide DAN 1,5-dialkoxynaphthalene NMR Nuclear Magnetic resonance DCC N,N’- dicyclohexylcarbodiimide NOESY Nuclear Overhasuser Effect Spectroscopy DCE 1,2-Dichloroethane Phe Phenylalanine DEPT Distortionless Enhancement by Polarisation Transfer ppm Parts per million dd Doublet of doublets q Quartet ddd Doublet of doublets of doublets quint Quintet L-DOPA L-3,4-dihydroxyphenylalanine s Singlet DMF N,N-Dimethylformamide Ser Serine DMS Dimethyl sulfide Sxt Sextet DPP3 Dipeptidyl peptidase 3 Rf Retention factor dt Doublet of triplets RT Room temperature ee Enantiomeric excess t Triplet eq Equivalents td Triplet of doublets ESI-MS Electrospray Ionization Mass Spectrometry Tf trifluoromethanesulfonate Et Ethyl THF Tetrahydrofuran Gln Glutamine TLC Thin Layer chromatography HMBC Heteronuclear Multiple Bond Correlation Ts p-toluenesulfonyl HMQC Heteronuclear Multiple Quantum Coherence UV Ultraviolet 1 Chapter 1 Introduction 1.1 Natural vs Unnatural Amino Acids The term amino acid describes a diverse range of molecules with many functions. Most people know that amino acids are abundant in nature, but they are also found in synthetic molecules including chiral ligands,1 organocatalysts,2 and synthetic polymers.3 By definition, an amino acid is any molecule that contains an amine and carboxylic acid. The term amino acid is most commonly limited to the 20 proteinogenic a-amino acids; those amino acids coded by our genes used in the synthesis of proteins.4 However, nature has synthesised more than 300 amino acids, all of which play critical roles in biology. An example of this is the natural amino acid, folate vitamin B9 1, a glutamic acid derivative (Figure 1). Figure 1 – Formation of dipeptide through the formation of a peptide/amide bond (left); the natural amino acid glutamic acid derivative folate vitamin B9 (centre); two examples of unnatural amino acids. Unnatural amino acids are synthetic derivatives containing both amine and carboxylic functionalities, which aren’t seen in nature. Proteins, formed from natural amino acids, invariably form secondary structures comprising of a-helices or b-sheets. Chemists, H2N R1 H2N N H R1 H2N O R2 R2 Peptide/amide bond H2O Dipeptide OH O OH O OH O Amino acid 1 Amino acid 2 HN N N N O NH NHO OH O O OH NH2 NH2 O OH Natural amino acid folate vitamine B9 Unnatural amino acids H2N R1 O OH 3 2 1 Proteinogenic α-amino acids used in protein synthesis 2 using natural amino acids, should have access to different structures and can incorporate an unlimited range of functionality. However, with unnatural amino acids, there is a limitless choice of building blocks allowing chemists to improve on nature and create new structures with alternative functions. Numerous studies on unnatural amino acid have been completed, including b-amino acids 2, and have been incorporated into interesting scaffolds like 3. A diverse assortment of natural and unnatural amino acids have been seen in applications including synthetic polymers,5 C—H bond activation,6 organometallics,7 and the synthesis of foldamers.8 This thesis investigates two key areas, natural and unnatural amino acids. However, both areas look to answer one question: are we able to improve on nature? 1.2 Natural Peptide - Opiorphin Opiorphin 4 is a naturally occurring peptide, consisting of the five amino acid sequence; Gln-Arg-Phe-Ser-Arg (Figure 2).9 It can be isolated from saliva, and has been reported to have an analgesic effect five times stronger than morphine.10 Opiorphin extends the duration of enkephalin activity, another peptide responsible for regulating pain, which is released in response to painful stimuli. Opiorphin induces a pain-killing effect by inhibiting three proteases, neprilysin (NEP), alanine aminopeptidase (AAP), and dipeptidyl peptidase (DPP3), which are responsible for the breakdown of enkephalin.10–12 Inhibiting these three proteases results in an increased duration of enkephalin activity, allowing it to bind to the opioid receptors, thus regulating the sensory nervous system reaction to pain. Most administered narcotics result in adverse effects, while 4 has only been shown to have an anti-depressive effect, which makes it promising candidate for future pain relief.11 3 Figure 2 – Molecular Structure of Opiorphin 4 (shown in it’s non-zwitterionic form). A study by Dr. Preet Singh from Massey University in 2019 showed 4 is less effective in vivo than morphine.13 This study looked at the changes in electroencephalogram (EEG) and results of a hot-plate and tail-flick test from female Sprague-Dawley rats when exposed to noxious stimuli.13 The study compared the analgesic effect of morphine (1 mg/kg), opiorphin (2 mg/kg), and saline (0.5 mL) after five minutes from receiving an intravenous injection. The results show opiorphin can pass through the cerebral spinal fluid to affect the spinal cord but is unable to target the brain effectively. Currently, the low efficacy could be due to two reasons; firstly, opiorphin is unable to pass through the blood-brain barrier (BBB) effectively; secondly, due to the short half-life (15 minutes) it is unable to effectively inhibit the proteases.12 Both of these issues can be resolved by altering the structures of opiorphin. Although altering the half-life and permeability of opiorphin may be possible at the same time, this thesis focused on modifying the permeability through the BBB. 1.2.1 Blood-Brain Barrier The BBB prevents damage to the brain and is a vital physiological barrier comprised of a complex network of micro-vessels, which protect the organ from external factors while providing it with the required nutrients.14 However, this also makes it an obstacle for chemists when designing drugs that target the central nervous system. In comparison to peripheral capillaries where there is relatively free exchange between tissue and blood, H N O N H O H2N HN H2N NH N H H N OH H2N O O OH O O NH NH2HN 4 4 the capillaries found in the brain are the least permeable in the body.14 This is due to the addition of tight junctions; these are protein complexes that stitch adjacent cells together preventing the leakage of substances between cells (Figure 3).14 Due to the physiological architecture of the BBB, it is typically considered the rate-limiting factor for bioavailability, as it prevents the movement of 98% of small molecular drugs through and around the cells.15 Figure 3 – Cartoon representation of the blood-brain barrier. Potential drugs need to be absorbed from the blood to the endothelial cells. A pre-requisite for this movement is lipid solubility, a property which can be quantified by log P measurements. The partition coefficient, P, is the ratio of solute concentration between two solvents. When one phase is H2O, and the other is a non-polar solvent like n-octanol, log P is a measurement of lipophilicity. Opiorphin has a log P value (partition coefficient) of -3.42 and studies show a permeability of ~3% through the BBB, which is considered very low for an analgesic drug.10,16 The optimal log P for passive permeability through the cell membrane has been thoroughly investigated and is between -0.2 and 1.3.17 Opiorphin’s glutamine residue can be converted to pyroglutamate in physiological conditions, and this transformation preserves the analgesic effect while increasing the stability and lipophilicity as observed as an increase in log P from -3.42 to -2.49.10 Previous studies have shown the pyroglutamate-opiorphin derivative has increased bioavailability when administered to mice.10 An established study showed permeability to the BBB is based on surface activity, which takes into account hydrophobic and charged residues, and the molecular weight of the molecule.18 The study showed molecules that are highly charged or comprised of larger molecule weight have lower 5 permeability, whereas amphipathic molecules show optimal permeability. This highlights the primary reason for the low permeability of opiorphin; under acidic conditions a positive charge on both arginine residues is stabilised by the delocalised electron lone pair of the adjacent nitrogen groups, resulting in a highly positively charged molecule, obstructing its permeability through the BBB. The chemical modification of 4 has not been thoroughly investigated, with only one group reporting any variation of 4.19 A structure-activity relationship (SAR) study of 4 conducted by Rosa et al. demonstrated the importance of the aromatic ring on opiorphin's bioactivity with the protease NEP.20 Further studies have shown the glutamine residues interact with DPP3, limiting the functional groups of 4 which can be modified exclusively to the arginine and serine residue.21,22 However, as mentioned the arginine groups play the largest role in the low permeability of 4, masking them with a physiologically labile group is advantageous and will be the greatest tool for this stream of research. 1.2.1 Prodrugs With only 10% of all newly approved drugs being prodrugs, chemist have only recently started to realise their potential.23 A prodrug consists of an active parent drug attached to a non-toxic moiety masking or altering its physiological properties (Figure 5). The moiety allows the drug to be dormant during its administration, and typically requires minimal chemical or enzymic transformation to reveal the active component.24 Ideally, a prodrug delivers the active drug with high recovery. The first and most notable example of a viable prodrug that is still currently administered is L-3,4-dihydroxyphenylalanine (L-DOPA) 5.25 Initially used in 1961 to relieve symptoms of Parkinson’s disease, L-DOPA is a metabolic precursor to dopamine 6 and can efficiently pass through the BBB. While dopamine has no permeability, L-DOPA is broken down by DOPA carboxylase into dopamine and CO2 7.26 6 Figure 4 – Comparison of the BBB permeability mechanism for A) unmodified drug vs B) prodrugs. As mentioned earlier, SAR studies have shown both arginine residues of opiorphin have no direct effect on the analgesic properties of the compound. They are also problematic for the BBB, therefore altering these groups would be the most beneficial as it could solve bioavailability without comprising activity. As the arginine residues are the core reason for the molecule’s poor permeability through the BBB focusing on altering these groups would be the most beneficial route to increase permeability. Although the literature has shown the arginine side chain can be protected through the use of common protecting groups like the tosyl group, it is likely these methods would result in a permanent change. Such a derivative would not be a prodrug. Due to this, we searched for a protecting group that has been shown to detach form the arginine side chain under mild conditions. Numerous 1-(4-hydroxyphenyl)ethan-1-one derivates have been shown to mask the guanidino groups of arginine and other molecules in a reversible fashion.27–36 Currently, the reaction between hydrate derivatives with the arginine side chain has been examined in various biological-based studies. These include selective labelling of arginine residues engaged in sulfated glycosaminoglycans binding, experiments to improve mass spectroscopy analysis of protein structures by binding cross-linkers to arginine, or various Promoiety Blood Brain Barrier Opiorphin OH O NH2 HO HO L-DOPA 5 OH O NH2 HO HO L-DOPA 5 C OO Carbon dioxide 7 NH2HO HO Dopamine 6 Chemical or Enzymatic Transformation PromoietyOpiorphinPromoietyOpiorphin A) B) B) Drug Opiorphin 7 other peptide based applications.27,30,31 Most reports in the literature have used methylglyoxal to cap the guanidine moiety to form advanced glycation end-products (AGEs).37 This has been highly investigated as AGEs are bio-markers that are believed to play a role in ageing and many degenerative diseases like diabetes, chronic kidney disease and Alzheimer’s disease.37,38 This indicates that a hydrate and arginine side chain reaction is possible and has been thoroughly investigated through a biological perspective. However, information that is critical to a chemist has been ignored throughout a lot of these studies. As these investigations are completed in a physiological setting, they require micro concentrations and are analysed through highly sensitive techniques. Due to this, isolation of the product has not been completed and thus yields cannot be stated. In the rare examples where yields have been stated they have come with other chemical inconsistency, which leads us to question the chemical methodology of the investigation.31 Nevertheless, the research provides a foundation of our work to begin. The reaction conditions have been reported, and by working on a macro scale we should be able to optimise the reaction between a hydrate and the arginine side chain to produce a viable prodrug. 8 1.3 Unnatural Amino Acids – Foldamers A foldamer is a synthetic oligomer, a molecule comprising of a monomeric units that self- organise and fold into periodic secondary structure in solution.39 As such, foldamers mimic biopolymers, like proteins and polysaccharides. The goal of foldamer research is not to recreate natural biopolymers, nature has already perfected these structures, but to be inspired by the same principles and form new structures with properties not observed in nature. The remarkable diversity of both large and compact structures reported in the literature demonstrates that foldamers have the ability to form architectures outside of the typical a-helices and b-sheets found in nature.39 As the selection of viable building blocks significantly increases, so too does the diversity of structures; allowing them to possess unique properties and increasing the applicability of the sequences. This has allowed foldamers to be utilised in various fields including cell penetration,40 interactions with biomolecules,41 catalysis,42 medicine,43 and more. The structures produced are direct result of the non-covalent interactions, which govern the formation of the various architectures depicted in Figure 5.39 The intramolecular interactions driving foldamer formation are largely predesigned within the backbone sequence of the molecule in the form of aliphatic or aromatic building blocks.39 Aliphatic foldamers have been a key focus in the majority of the literature, while aromatic foldamers are relatively unexplored and will be the key area investigated in this thesis. This section will first outline the interactions that control secondary structure formation before looking at some of the recent examples. Although these examples highlight what the field is capable of doing now, it is only natural this is followed by looking at the roots of the field and look at the decade’s worth of research required to catch up with nature. 9 Figure 5 – Various foldamer structures reported in literature. Adapted from reference.39 1.3.1 Non-Covalent Interactions 1.3.1.1 Hydrogen Bonding As with the secondary structure of proteins produced by natural amino acids, hydrogen bonding between repeating units is key to the formation of artificial architectures.44 In supramolecular chemistry, hydrogen bonding is one of the most commonly exploited driving forces in the formation of highly ordered structures.44 A variety of foldamers have been reported in the literature, highlighting the flexibility of hydrogen bonding in the primary structure of foldamers.44,45 Double-strand and turn Single-strand receptor Double-strand receptor Multi-strand templated sheet Double-strand helix Strand templated by helixSingle-strand ribbon Single-strand helix Single-strand metal coordinationHelix templated by strand Double-strand metal coordination 10 1.3.1.2 p-p Interactions In aromatic foldamers, the electrostatic p-p interactions between two aromatic rings play a large role in the formation and stabilisation of secondary structures.46 The p-cloud of an aromatic ring results in a region of high electron density sitting above and below the plane of the benzene ring, while the C—H bonds are polarised towards the carbon atom, as shown in Figure 6.47 These electronic properties play a crucial role in determining the conformation two aromatic rings will adopt.46,47 Three main possible motifs can be used in constructing aromatic foldamers these are face-to-face, offset face-to-face, or edge-to- face (Figure 6).47 The face-to-face orientation results in a high energy arrangement due to the repulsive electrostatic forces caused by the p electron clouds being in close proximity to one another. The offset face-to-face orientation is favoured as this allows attractive interactions.47 The edge-to-face conformation is the most favourable conformation of two aromatic rings due to the dipole-dipole interaction of a hydrogen atom interacting with the electron cloud of the other aromatic ring, creating a C−H···p interaction.46,47 However, to the best of our knowledge, the edge-to-face orientation has not been utilised for the construction of foldamer architectures. Figure 6 – Electron distribution of a benzene ring (left) and the various interactions between two benzene rings. 1.3.1.3 Solvophobic Effects Solvophobic effects are responsible for the association of poorly solvated molecules or sections of molecules.48 However, unlike hydrogen bonding and other specific interactions, solvophobic effects rely on the collective interactions between solvent, H H H H HH H H H H HH H H H H HH H H H H HH H H H H H H H H H H HH Face-to-face (unfavourable) Offset face-to-face (favourable) Edge-to-face C-H-p interactions (most favourable) H H H H HH (Top View) Quadrupole Hδ+ δ- HH H H H 11 solutes, and non-specific functional groups.48 Solvent effects can be either direct or indirect.49 Direct solvent effects compete for the supramolecular reactive sites and impede the folding, in other words they disrupt hydrogen bonding.49 This is why hydrogen bonded supramolecular structures breakdown in H2O or DMSO.49 Indirect solvent effects are typified by hydrophobic effects, where the forces driving association of non-polar groups in aqueous solution is from the unique properties of H2O (size and hydrogen bonding), and not the dispersive interactions among solute molecules.48–50 1.3.2 Recent Examples of Foldamers Research into foldamers commenced over 60 years ago, and covers a range of topics.39 There are a number of good reviews 39,51–55 so in the following section only recent, interesting examples will be discussed below. 1.3.2.1 Macrocyclisation-Induced by Catalytic Foldamer One goal of foldamer chemistry is to synthesise structures that have a function not observed in natural systems. A means of achieving this is by combining two or more types of unnatural amino acids; one gives the molecule structure while the other adds functionality.56 This extends the scope of available transformations in comparison to those seen in nature. Gellman et al. have produced a heptapeptide foldamer that catalyses the macrocyclisation of various aldehydes and has led to the synthesis of two natural products; robustol 9 and nostocyclyne A 10 (Figure 7).57 The formation of macrocycles is an entropically disfavoured process and is considered a challenge for chemists.57 A sequence of a- and b- amino acids was used to create a specific secondary structure, a helix.56 This created a stable three-dimensional framework, from which reactive functionality could be suspended; positioning it so that selective intramolecular aldol condensation reactions to occur. Gellman et al. analysed multiple derivatives, altering the position of the two reactive functional groups, which influenced the efficiency of cyclisation.57 The optimal sequence, 8, is shown in Figure 7. This sequence was then compared to small-molecular catalysts which served as analogues of the active site of the foldamer but without the positioning effect. These failed to furnish the desired macrocycle under the same conditions.57 This demonstrates how the three-dimensional structure formed by the foldamer plays a crucial role in macrocyclisation. The foldamer was 12 capable of synthesising macrocycles as large as 22 atoms, proving foldamers can control the selectivity and reactivity that previously were characteristics exclusive to enzymes and proteins.57 This recent example shows how the development of a new foldamer catalyst opens up the possibility of alternative transformations beyond the aldol reaction, enriching the general synthetic repertoire currently available to chemists. Figure 7 – (A) Molecular structure of the most effective foldamer sequence synthesised for H H O O N H H2N OH O H H O H O O H HO HO HO 7 Robustol 9 Nostocyclyne A 10 NH HN NH HN NH NH N H HN O O O O O H2N O O O NH OH (A) Foldamer Sequence O O NH NH2 O O NH NH2 O O N N H (B) Small-molecule catalyst vs foldamer (C) Molecular Strcutures (D) Macrocyclisation mechanism 8 OH OH 3 10 mol % catalyst 2 eq. Et3N 2 eq. ProO2H 96:4 IProOH:H2O 10 mM, 37 °C, 24 h 13 macrocyclisation.57 (B) Reaction conditions for macrocyclisation comparing a small-molecular catalyst to a synthesised foldamer. (C) Molecular structure of the two natural products synthesised using foldamer sequence. (D) Macrocyclisation mechanism utilising the foldamer developed by Gellman et al.57 1.3.2.2 Foldamers for Biological Cures Huc et al. have developed aromatic foldamers containing phosphonate and carbonyl groups that mimic B-DNA in both the charged surface and solid-state structure, as determined by X-ray crystallography (Figure 8).58 This type of foldamer has been reported to be a competitive inhibitor of DNA topoisomerase 1 (Top1) and human immunodeficiency virus integrase (HIV-1 IN).59 The group synthesised a foldamer using QPho, mQPho, and Q5Pho (11-13, respectively) building blocks. A alternating sequence of mQPho and QPho or Q5Pho was shown to develop a helical structure with a ~35° minor groove twist. The helical structure is a result of electrostatic repulsion and offsets face-to-face orientation of the aromatic building blocks stabilised by hydrogen bonding similar to the helical structure adopted by DNA, as shown in Figure 8B. Although they were initially designed to resemble DNA, the most remarkable properties are due to their differences. These alterations result in the formation of a more versatile sequence, while maintaining stability. A homogenous sequence of (QPho)8 showed remarkable stability of the secondary structure in protic solvents. The helical structure was maintained even at 120 °C in DMSO, whereas DNA begins to denature at 60 °C.58 The foldamers were found to be able to bind and inhibit several DNA-binding enzymes at a higher affinity than DNA itself. The sequence of (mQPhoQPho)8 was shown to inhibit HIV-IN, while in comparison (mQPhoQ5Pho)8 was shown to have a lower binding affinity.59 Camptothecin and raltegravir are considered the best inhibitors for Top1 and HIV-1 IN, respectively, and were shown to have a lower affinity than the designed foldamer, demonstrating the versatility of these compounds and efficiency over natural biomolecules.59 14 Figure 8 – (A) building blocks used to mimic DNA synthesised by Huc et al. (B) Foldamer heterogenic sequence synthesised for HIV-IN and Top-1 inhibition. (C) X-ray structures of the two foldamer sequences synthesised. (D) Molecular structure of Camptothecin and Raltegravir. Although foldamers have only recently begun to be used in biological and chemical applications, the architectures being developed today are a result of decades of research N O P HO OHO ONH N O P HO OHO O NH N P HO OHO NH O QPho 11 mQPho 12 Q5Pho 13 N O P HO O O O N N O P OH O O O N mQPhoQPho N O P OH O O O N N N O P O O HO mQPhoQ5Pho Minor groove Minor groove Major groove Major groove Major groove Major groove (A) (B) (C) N N O O O Camptothecin 14 N N H N O F H N O O NN OH O Raltegravir 15 H H H H (D) Boc(mQQ5)4OBn Boc(mQQ4)8OTMSE 15 looking at various techniques and building blocks to induce folding. The following section will briefly look at the two classes of foldamers, aliphatic or aromatic, and highlight related examples and principals to this thesis. 1.3.3 Aliphatic Foldamers The vast majority of foldamer research has focused on aliphatic foldamers, which are sequences comprised of saturated building blocks relying on hydrogen bonding and solvophobic effects to induce folding.39,43,60,61 Peptidomimetic foldamers are a sub-class that have been the initial driving force of foldamer research. Peptidomimetic foldamers mimic natural proteins but are comprised of unnatural amino acids. Most of these are b-, g-, or d- amino acids. These differ from standard a-amino acids by addition of extra methylene (-CH2-) groups between the amine and carboxylic acid. These backbone sequences can be functionalised to produce a variety of foldamers and are summarised in Figure 9.39 Although these foldamers were the foundation of the field, the research completed in this thesis revolves around the development of an aromatic foldamer. For this reason, the details of aliphatic foldamers will not be discussed further. Figure 9 – Various categories of aliphatic foldamers seen in the literature. Adapted from reference.39 α-peptides β-peptides γ-peptides δ-peptides 1,2-diamino- ethanes N-permethylated α-peptdies peptoids azapeptides azatides oxazolidin-2-ones pyrrolinones cyclic β-peptides α-aminoxy acids sulfonamides sulfinamides sulfoximines vinylogous sulfonyl peptides vinylogous peptides ureas cyclic ureas hydrazino peptides carbamates phosphodiesters carbopeptoids amide-linked sialooligomers anthranil- amides 16 1.3.4 Aromatic Foldamers Aromatic building blocks offer an alternative route to the synthesis of novel architectures. Amide linkages are used to promote hydrogen bonding, but other non-covalent interactions also play a critical role in developing higher order architectures. Since the 1990s, a range of aromatic foldamers with unique architectures have been constructed.41,46 One may intuitively presume rigid aromatic rings hamper the formation of secondary structures like helices as they prevent conformational freedom. However, the rigid backbone allows for a predictable arrangement of monomers. This means isomeric structures can readily be synthesised by changing the substitution pattern of the aromatic ring, resulting in different secondary structures. Alternatively, different aromatic rings (benzene, naphthalene, anthracene, pyridine etc.) are capable of producing unique structures when utilised as building blocks. 1.3.4.1 Examples of Aromatic Foldamers While aromatic foldamers permit greater control and predictability of secondary structure this comes at a cost. The lipophilic nature of aromatic rings reduces water solubility and makes it more problematic to prepare biologically active assemblies. Chemists have risen to the challenge with a number of innovative solutions. Iverson et al. synthesised a water- soluble aromatic foldamer called an “aedamer”.63 This involved alternating 1,5- dialkoxynaphthalene (DAN) and 1,4,5,8-naphthalenetetracarboxylic acid diimide (NDI) linked by amino acids (Figure 10). The success of this system is due to the strong interactions between the electron rich DAN core and the electron deficient NDI core. The complementary groups allow for charge-transfer absorbance from the HOMO of the donor (DAN) to the LUMO of the acceptor (NDI), stabilising a face-centred geometry of the aromatic rings.63 This system is now an archetypal foldamer due to its p-p interactions and electrostatic complementarity. Many new foldamers are based on this design principle as it promotes secondary structure formation. Two oligo sequences were designed with aromatic units that are linked by aspartic acid residues, providing water solubility. The use of aspartic acid residues is chosen to mimic DNA, where the negative charges result in non-specific oligomer aggregation, driving the DAN-NDI interaction, 17 and showed a 1:1 binding stoichiometry with the ability to discriminate between the two strands.63 Figure 10 – Molecular structure and folded conformation of two DAN-NDI foldamers. O O R R N N OO O O R RO O R R N N OO O O R RO O R R N N OO O O R RO O R R N N OO O O R R O O R R N N OO O O R RO O R R N N OO O O R R O R O R N N OO O O R R = = O O NH HN O O O O O O N N OO O O HN O O O NHO O O O O R R N N OO O O R RO O R R N N OO O O R RO O R R N N OO O O R RO O R R N N OO O O R R O O R R N N OO O O R RO O R R N N OO O O R R + N N OO O O NH NHO O O O N H O O O O NH HN O O O 18 Moore et al. showed m-phenylene ethynylene will fold into a dynamic helical conformation when dissolved in polar solvents like MeCN.64 Solvophobic interactions cause the structure to unravel into a random conformation in less polar solvents, such as CHCl3.64 The foldamer is photoresponsive and irradiation causes the helix to switch directions, a characteristic which is rarely reported. Modelling shows the diameter of a foldamer of 12-units has an interval cavity of 8.7 Å.64 This cavity has been shown to host small solvent molecules that are readily displaced in a mixture of 40% water and MeCN. Hydrophobic molecules with binding energies in the range of 4-5 kcal/mol can bind within the cavity. Due to the larger size of an oligomer built from 22 monomers, it is able to constrain rod-like guest such as cis-(2S,5S)-2,5-dimethyl-N,N’-diphenylpiperazine, as shown in Figure 11.39,64 The host guest interaction is solvophobically driven by the burial of poorly solvated surfaces. The tightest binding is found when there are complementary interactions between host and guest. This occurs when the host and guest molecules are matching in terms of shape and size.64 The highest binding of cis-(2S,5S)-2,5-dimethyl- N,N’-diphenylpiperazine occurs when n = 20 and 22 as shown by circular dichroism measurements.64 Despite these aromatic oligomers solely relying on non-covalent interactions, the restricted rotation and strong geometric constraints are essential to the formation of secondary structures of these foldamers. 19 Figure 11 – (A) Chemical structure of and oligo(m-phenylene ethynylene)-based foldamer with X-ray crystal structure (with PEG R group omitted for clarity). (B) Chemical structure of rod-like guest with a cartoon representation of the helix templated by strand foldamer structure. Large strides in the formation of novel architectures have been made in the past three decades. From the common a-helices, first produced using b-amino acids in the 1990s, to the formation of rod-helix templates, catalyst and functional foldamers that can out preform small molecular catalysts. The successful synthesis of future foldamers will need to take into consideration two key design principles; first, the balance of attractive and repulsive non-covalent interactions to stabilise the desired structure. Secondly, the flexibility and rigidity of the backbone of these structures need to be taken into account, as these structures need to be able to readily form. There is still a significant amount of work to be completed within this field, with many more avenues to be explored. It is clear that nature is not alone anymore, and synthetic chemists now have the tools to be able to improve on the foundation it has provided us. O O O O O Si(CH3)3 N N n n = 10, 12, 14, 16, 18, 20, 22, 24 H N N - In MeCN - in CHCl3 cis-(2S,5S)-2,5- dimethyl-N,N’- diphenylpiperazine (A) (B) 20 1.4 [2.2]Paracyclophane Cyclophanes are a class of molecule that consists of at least one aromatic ring joined by an aliphatic chain, between two non-adjacent positions. Cyclophanes constitute a diverse range of structures (Figure 12), from [2.2]paracyclophane 16 and [2.2]metaparacyclophane 17 to natural products such as cylindrocyclophane A 18 and (±)-haouamine A 19.65–68 The most studied cyclophane is probably [2.2]paracyclophane. First discovered in 1949 by Brown and Farthing, it has fascinated chemists due to its unique chemical properties arising from the close proximity of the two aromatic rings.65,69 [2.2]Paracyclophane consists of two benzene rings joined by two ethylene bridges para to one another in a face-to-face conformation. It is a versatile framework used in applications such as asymmetric catalysis,70 stereoselective synthesis,71 and its original application as a precursor to a polymeric surface coating.72 [2.2]Paracyclophane will be used in this strand of research, as the scaffold for a foldamer. Figure 12 – Examples of cyclophanes. [2.2]Paracyclophane 16, the core building block for this strand of research. [2.2]Metaparacyclophane 17, a rare cyclophane explored in later chapters. Cylindrocyclophane A 18, a [7.7]paracyclophane natural product isolated from cyanobacteria and the first class of cyclophane compounds found in nature.68 (±)-Haouamine A (±)-19, a cyclophane compound isolated from Aplidium haouarianum, a marine invertebrate.67 HO HO HO OH OH OH (±)-Haouamine A 19 Cylindrocyclophane A 18 [2.2]Paracyclophane 16 [2.2]Metaparacyclophane 17 OH N OH OH OH 21 Although [2.2]paracyclophane 16 is commonly depicted as two eclipsing rings, the rings are slightly distorted and rotated around 5°, with the ethylene bridges being extended by 0.09 Å more than the usual 1.54 Å for a C-C bond.73,74 The unique bent and battered boat- like structure of [2.2]paracyclophane results in its distinctive properties as the p-electron clouds of each ring overlap.73 This leads to the highest occupied molecular orbital (HOMO) of [2.2]paracyclophane being higher in energy than other alkylbenzenes derivatives such as p-xylene, causing it to have increased reactivity in many reactions.75 The chemistry of [2.2]paracyclophane is full of idiosyncrasies, with many traditional reactions giving surprising products. These differences can be attributed to the steric and p-p interactions. Figure 13 – X-ray crystallography data of [2.2]paracyclophane 16 highlighting structural features and mirror planes.76 An attractive feature of [2.2]paracyclophane as a scaffold is the ease of accessing planar chirality. As shown in Figure 13, [2.2]paracyclophane contains three mirror plans resulting in the parent compound being achiral. However, once a substituent is added to the scaffold the mirror planes no longer exist and the [2.2]paracyclophane derivative will exist as enantiomers. Numerous substituted [2.2]paracyclophane have been reported and a few of the common substitution patterns are highlighted in Figure 14.77 It is important to note some substitutions patterns are inherently chiral, regardless of R groups, while others require unlike substituents (R1 ¹ R2). Additionally, although Figure 14 only highlights mono- and di- substituted [2.2]paracyclophane , tri- and tetra- substitution patterns are also known.78,79 Some substitution patterns are more easily prepared than others. This can be attributed to the unique directing effects of functional groups attached to [2.2]paracyclophane that allow them to direct reactions to the opposite ring. Positions on the second ring are defined by the suffix pseudo. An example of these directing effects 2.78 Å 3.09 Å 1.63 Å 22 can be seen in Chapter 2 where a nitro group can influence the adjacent ring to produce pseudo-geminal derivatives. As a result, the pseudo-geminal position is considered one of the easiest substitution patterns to form. For this reason, this arrangement is used in this strand of research. Figure 14 – Common substitution patterns for [2.2]paracyclophane with chiral substituent patterns highlighted. R H Achiral unsubstituted ortho meta bridged substitution R1 R2 R2 R1 R R pseudo-orthopseudo-meta R2 R2 R1R1 pseudo-parapseudo-geminal mono Inherently chiral Only Chiral if R1 ≠ R2 R R para 1 2 3 4 5 6 7 8 9 10 11 12 1315 16 14 R R R 23 1.4.1 [2.2]Paracyclophanes as Asymmetric Catalysts The growing importance of [2.2]paracyclophane in asymmetric catalysis has been demonstrated by phosphines, such as PhanePhos 20 80 or the imines developed by Stefan Bräse 21 81 or Daniel Glatzhofer 22.82 Each of these catalysts has been thoroughly investigated in the literature and utilised in a variety of reactions including hydrogenation,80 diorganozinc aryl and alkyl transfers,81 metal complexation and cyclopropanation.82 As catalysis is not a large focus of this research, only one example, PhanePhos 20, will be discussed. Figure 15 – Molecular structures of a range of [2.2]paracyclophane asymmetric pre-catalysts. PhanePhos, developed by Roseen and Pye, has had the largest influence on the development of new [2.2]paracyclophane catalysts as it was possibly the first [2.2]paracyclophane catalyst and of the most used.80,83 The catalyst has been reported to have high selectivity and increased efficiency than comparable rhodium and ruthenium- based catalysts.83 As alluded to earlier, PhanePhos 20 is primarily used as a ligand for rhodium or ruthenium catalysed hydrogenations. The cationic rhodium catalyst 27 was found to give increased yields to the catalyst formed in situ and allow for hydrogenations to occur at low temperatures such as -45 °C (Scheme 1).83 As seen in the hydrogenation of tetrahydropyrazine.83 2,2′-Bis(diphenylphosphino)-1,1′-binapthyl (BINAP) 23 had been utilised previously, resulting in only 56% ee and requiring harsher conditions (40 °C at 70 atm for 24 hours), while Et-DuPhos 24 gave 50% ee. PhanePhos gives 86% ee at - 40 °C. PPh2 PPh2 OH N Ph PhanePhos 20 Imines of Bräse 21 Glatzhofer’s salicylidene 22 N HO 24 Scheme 1 – Molecular structures of (S)-BINAP 23 and (S,S)-Et-DuPhos 24, both poorly performing catalyst for the hydrogenation of tetrahydropyrazine. PhanePhos performs the reduction of 25 reliably under mild conditions (2% mol), H2 (1.5 atm), -40 °C, 6 h, while yielding >99%, 86% ee.83 1.4.2 [2.2]Paracyclophanes as a Scaffold for New Materials Not only has [2.2]paracyclophane found applications in asymmetric catalysis but it is also a good scaffold due to its rigidity and ability to position substituents. Some examples are highlighted in Figure 16. Chujo et al. developed an optically active cyclic cage 28 (Figure 16 – A), containing a stacked p-electron system by utilising a pseudo-ortho isomer of [2.2]paracyclophane. While Filichev et al. incorporated enantiomerically pure [2.2]paracyclophane into DNA forming an architecture resembling a zipper (Figure 16 – B). The assembly 29 was shown to be stable at room temperature, which is a crucial component to for these chiral assemblies to be used in applications.84 A tetra-substituted [2.2]paracyclophane 30 has been utilised to form unique hydrogen bonded extended structures (Figure 16 – C).85 Castellano et al. has formed a tetraamide that self-assemblies by exploiting transannular and intramolecular hydrogen bonding interactions.85 By balancing the enthalpically unfavourable face-to-face stacking and the enthalpically favourable formation of hydrogen bonds, the off-set stacking can reduce the enthalpic N N OO O O N N OO O O Ph2 P P Ph2 Rh O O H H TfO 2725 26 > 99 % 86% ee 20 PPh2 PPh2 (S)-BINAP 23 (S,S)-Et-DuPhos 24 P P O O 25 cost producing a larger net binding energy assisting in self-assembly. This is very encouraging for our foldamer project. Figure 16 – A variety of architectures that utilise [2.2]paracyclophane as a key building block. (A) [2.2]Paracyclophane used in the formation of a cage 28.86 (B) [2.2]Paracyclophane functionalised deoxyribonucleic acid 29 (C) A tubular [2.2]paracyclophane sequence 30. 85 Another particular interesting oligomer, is the pseudo-geminal substituted [2.2]paracyclophane scaffold developed by Collard et al 33 (Scheme 2).87 Collard’s group was the first to develop a polymeric system that contained pseudo-geminal substituted [2.2]paracyclophane (31, 34), where the substitution pattern resulted in a U-turn structure, and is an example of an aromatic single-strand ribbon foldamer.87 It shows how oligomers and conjugated polymers exhibit characteristics that allow them to be used as semiconductors in thin-film optoelectronic devices.87 This behaviour is governed by the O O PO O- O O- DNA R O DNA N N N O N R H O NR H O N R H N O H R 29 30 (B) (C) R= Adenine or Uracil HO OH OHHO HO OH 28 (A) 26 delocalisation of charge carries and excitons along the conjugated backbone and by the interchain interactions the p-system of closely packed chains, properties drastically increased by the stacking nature of 33. Scheme 2 – Reaction between 4,13-diethynyl[2.2]paracyclophane 31 and 32 to produce stacked pseudo- geminal [2.2]paracyclophane scaffold 33. Alternative building blocks 34, and 35, have also be used to produce scaffolds to from similar structure to 33. The same group has synthesised various frameworks by linking pseudo-geminal [2.2]paracyclophane with aromatic units easily incorporated using Heck, Sonogashira cross-coupling polymerisation, and Stille coupling of 35 and 32. These scaffolds have been analysed by UV-vis and fluorescence spectroscopies, as well as cyclic voltammetry and compared with their monomeric counterparts. Results showed the polymers had similar absorption maxima to their corresponding unstacked monomers. But the absorption edge of all scaffolds is red shifted by ca. 50 nm. This shows the stacked nature of the scaffolds synthesised has some influence on the electronic ground state. However, R R I I R= OC9H19 3231 R R R R R R R R R= OC9H19 33 S S I I R R R= C8H17 3534 Pd(PPh3)4 CuI THF IPr2NH 27 the emission spectra between the stacked and non-stacked molecules show a significant difference. The unstacked controls show vibronic peaks that are sharp and distinctive, while the emission from the polymers is red shifted and broad. This can be attributed to the fluorescence quenching due to the stacking of the conjugated tiers, a commonly seen trait for thin films comprised of conjugated polymers. Additionally, all stacked polymers show a significantly larger Stokes shift of at least 60 nm, and up to 150 nm. The authors believe the two vinyl groups of [2.2]paracyclophane can be orientated away from the ethano bridge to minimise steric interaction. It is thought multiple orientations of cisoid and transoid conformations through the [2.2]paracyclophane scaffold can provide a break in stacking, limiting the extent of overlap of the p-system resulting in smaller Stokes shifts in comparison to 33 (Figure 17). Figure 17 – Chemical structure of pseudo-geminal phenylene vinylene [2.2]paracyclophane 36 polymer, highlighting the required overlap conformation for communication between planes. 1.4.3 [2.2]Paracyclophane as Amino Acids To form a planar chiral foldamer from [2.2]paracyclophane, an amino acid derivative of the scaffold will be joined together through a series of peptide bonds. Although there are seven possible motifs available (Figure 18), only two existing isomers have been reported in the literature. The most attractive is the pseudo-ortho isomer, but the synthesis is R R R R R R R R R= OC9H19 36 Transoid Cisoid Transoid Transoid 28 challenging, lengthy, and low yielding. Therefore, the pseudo-geminal isomer was chosen as the first target. It also provides the shortest route to an amino-acid derivative; however, it possesses its own difficulties discussed in the next chapter. While the synthesis of both amino acid isomers has been reported, research utilising them has been uncommon. With the literature reporting the use of a proline-[2.2]paracyclophane derivative used in asymmetric catalysis, and the Rowlands group previously investigating pseudo-geminal organocatalyst. Figure 18 – Molecule structures of possible amino acid scaffold, highlighting the isomers seen in literature. It is clear to see the synthesis of a foldamer utilising [2.2]paracyclophane as the core scaffold is an area that is ripe for study. Various architectures have already been produced using [2.2]paracyclophane and have been incorporated into numerous established structures like DNA. By utilising a unique framework that mimics the tools nature has provided, the potential of developing a new form of secondary structure not previously seen in the literature or nature increases. NH2NH2O OH NH2 O OH NH2 O HO O HO NH2NH2 O HO NH2 39 40 41 42 38373 HO O O OH 29 1.5 Aims of this Research As the literature review shows, the chemistry of amino acids, both natural and unnatural, is vast with numerous books written on both. Each topic still inspires chemist with research being completed daily. The goal for each strand of research narrows down to one straightforward question. Are we able to improve on nature? 1.5.1 Aim of Natural Amino Acid Research Strand Opiorphin could be an effective painkiller, it is a stronger analgesic than most prescribed drugs, but it is limited by undesirable properties. Its low permeability and short duration of action must be addressed before it can be administered as a painkiller. For this strand of research, the aim is to increase opiorphin’s permeability through the BBB by developing a reliable masking technique for the guanidine groups of both arginine residues and thus creating a potential prodrug. Synthesis of a protected arginine 55 will initially be used as a template before attempting reactions on opiorphin as it is a simpler system and more abundant in a lab (Scheme 3). The literature has showed 44 derivatives as a viable masking agent for the guanidine group allowing us to incorporate varying R groups into the framework.27–36 30 Scheme 3 – The synthetic approach to produce the opiorphin prodrug utilising arginine as a prototype. Once the proof of concept has been established, synthesis of an opiorphin prodrug 57 (Scheme 4) will be completed allowing for the protected opiorphin's lipophilicity to be analysed. The synthesis of various derivatives will then be attempted to optimise the Log P between -0.2 to 1.3. The effectiveness and toxicity of the most optimal product will then be analysed by Dr Preet Singh (Massey University). From these results, the R substituent can be further explored to obtain desired results. From there, the prodrug can be finally tested as a painkiller for de-horning of animals. O O N H O O HN H2N NH OH O H2N HN H2N NH HO O O O R OH OH R = (CH2)nCH3 n = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 44-5343 O O N H O O HN HN N O O R 54 55 56 31 Scheme 4 – The synthesis of an opiorphin prodrug, where R is altered to produce an optimal Log P between -0.2 to 1.3 and is non-toxic. 1.5.2 Aim of Unnatural Amino Acid Research Strand This strand of research focused on the development of the first planar chiral foldamer utilising a [2.2]paracyclophane backbone and structural studies to determine its conformation. The foundation of the foldamer, the planar chiral amino acid precursor (±)-58, had to be synthesised first. Then these monomers were able to be joined by peptide bonds (Scheme 5). However, before linking multiple [2.2]paracyclophanes together, the initial monomer required a substituent to act as a cap, this prevented uncontrolled polymerisation. Once the capped derivative 59 was synthesised, we started coupling [2.2]paracyclophane amino acid derivatives to produce dipeptides 60, 61. Once sufficient amounts of the dimer were obtained, the material could be split in half, allowing us to link dimer derivatives together to form a tetramer 62. Once again, we planned to separate the acquired material in half, with each half undergoing reactions to couple them and ultimately produce an octamer, or react a tetramer with a dimer to form the hexamer 63. This would allow us to examine the length required of our foldamer to first see folding. H N O N H O H2N HN H2N NH N H H N OH H2N O O OH O O NH NH2HN 4 N H H N N H H N OH H2N O O O O O OH2N HN HN N OH NH NHN OR OR O O 57 32 Scheme 5 – The initial synthetic approach to synthesis a planar chiral foldamer. O NH R NO2 O N H NH O N O NHCl O R2 O NH O N H O HN O HN OH N O R2 O NH O N H O HN O R2 O H H R1 R1 R1 NO2 O OH (±)-58 59 62 63 16 6160 33 Chapter 2 Results and Discussion This thesis covers a broad range of topics. As such, the results & discussion will be broken up into four parts. Section 1 will discuss the results obtained from the attempted synthesis of a prodrug for opiorphin. Part two discusses our first attempts to couple [2.2]paracyclophane and the work required leading up to dimer formation. Section 3 continues the discussion of the foldamer synthesis building upon the knowledge gained in Section 3 and looks to resolve some of the issues accounted. Lastly, Section 4 investigates a reaction observed during the synthesis of (±)-4-nitro[2.2]paracyclophane. 2.1 Section 1 – Natural amino acids 2.1.1 Synthesis of Hydrate Derivatives Our first goal was to form a greasy hydrate that would be used to protect and alter the properties of opiorphin. Synthesis of 50 was initially attempted by tosylation of the appropriate alcohol followed by substitution with a phenol (Scheme 6).88 Activation of the alcohol was necessary as it is a poor leaving group. Tosylation was attempted by treatment of Et3N and DMAP; however, analysis of the reaction mixture by 1H NMR spectroscopy revealed that no product formation had occurred. Repeating the reaction for an extended period was also unsuccessful. 34 Scheme 6 – First attempt to synthesise a hydrate derivative containing a long alkyl chain. For some reason the alcohol 64 was not reacting under standard conditions. This might be due to the influence of the hydrocarbon. A solvent-free, green approach, was attempted next. This required the reagents to be vigorously ground together for five minutes (Scheme 7).89 Analysis of the 1H NMR of the crude reaction mixture showed that the tosylated product was the minor product with unreacted starting material making up the rest. Usually, the grinding of reagents for prolonged periods would be an automated process; we simply tried to vigorously stir the reaction overnight using ethanol and a magnetic flea as we did not have such equipment. Analysis of the 1H NMR showed no product formation demonstrating that the frictional force was a requirement for product formation in this reaction. Scheme 7 – Second attempt at synthesise of the tosylate 65. As the reaction did not go to completion, an alternative and shorter route was explored. By using the alkyl halide 67, we were able to add the aliphatic chain directly to 43 (Scheme 8). The synthesis of 1-(4-(decyloxy)phenyl)ethan-1-one 68 was achieved with yields ranging from 87-98%, with purification not required. The reaction of 1- bromodecane and 1-(4-hydroxyphenyl)ethan-1-one is a simple SN2 substitution, with the O O O O OH OH TsCl, Et3N, DMAP CH2Cl2, 0 °C, 15 h 64 65 6650 S O O O OH 64 65 S O O O OH TsCl, K2CO3 Grind 5 mins 35 lone pair of electrons on the oxygen acting as the nucleophile, the carbon attached to the bromine leaving group, the electrophile. Scheme 8 – Reaction conditions used to produce the hydrate precursor 68. Synthesis of hydrate 53 was achieved in 77-83% yield through the use of SeO2 (Scheme 9).90,91 Aldehydes establish an equilibrium with H2O to produce a hydrate in the presence of an acid or base; this is known as the gem-diol equilibrium. Generally, the hydrate is unstable and cannot be isolated. However, due to the presence of the ketone, an electron- withdrawing group, the stability of the gem-diol increases, allowing b-keto-gem-diols to be isolated. Entropically, the equilibrium should favour the aldehyde 69 and water due to two molecules being preferred. However, both the hydrate and aldehyde are visible in the 1H NMR and the ratio of aldehyde to hydrate changes depending on the length of storage and storage conditions. A 1H NMR taken immediately after the reaction showed a larger proportion of aldehyde, while the proportion of hydrate increased, the longer the sample was left. There is a distinguishable colour change between the two, with the aldehyde being a whitish-green compound, and the hydrate appearing pink. Depending on the humidity, the colour of the compound would change daily between the two. Scheme 9 – Reaction conditions used to produce the desired hydrate 53. The synthesis of 53 occurs by a selenium dioxide-mediated oxidation known as the Riley oxidation, which occurs by a series of sigmatropic rearrangements (Scheme 10). The reaction begins with enol tautomer of 68 attacking the electrophilic centre of the selenium dioxide. A Pummerer-like rearrangement occurs resulting in the loss of H2O, which O O H O9 O O9 SeO2, H2O 1,4 dioxane reflux, 7 h 68 O OH OH O9 69 77-83% 53 HO O Br9 O O9 43 67 K2CO3 DMF, RT, 13 h 68 87-98% 36 attacks the alpha position causing the release of selenic acid and affording 53 in equilibrium with 69. Scheme 10 – Proposed mechanism for the Riley oxidation using SeO2 to afford 53. The equilibrium between the greasy gem-diol and the aldehyde made assignment of the 1H NMR spectrum challenging and caused us to question if the reaction had been successful. This, paired with mass spectroscopy not showing the desired molecular ion for either products, made us sceptical of the reaction outcome. To clarify the 1H NMR spectrum, the decane chain was substituted with a methyl chain to simplify the system. Both reactions (Scheme 11) appeared as successfully as the decane derivatives, with the isolation of 70 in 94-98% yield and the formation of 44 yielding 83-90%. For the synthesis of 70, DMF was substituted for MeCN making the work-up require less brine washes. However, our efforts to simplify data analysis were in vain, as we continued to see the same issues as seen with the decane chain derivate. Efforts to push the equilibrium to the formation of 44 or 53 through the addition of HCl(aq) and water also proved inconclusive. O O O O O O OH O OH H3O+ H2O 69 53 OH O O O Se OH O O O Se O O HH O O H Se O OH 9 R = R R R R R R R 68 O O Se OH2 O HR Se OO H H+ 37 Scheme 11 – Reaction conditions utilised to prepare 70 and the methyl hydrate derivative 44. As the synthesis of both hydrates was uncertain, test reactions to produce 72 or 73 were attempted by reacting the hydrates with NH4OAc (Scheme 12).92 If the hydrate synthesis was successful, either form of the hydrate should react with NH4OAc to produce an imidazole derivative that should be easily identifiable by 1H NMR. However, the test reactions using either 44 or 53 were both unsuccessful. Scheme 12 – Attempted imidazole formation, for the presence of 44 or 53. It was noted, although both products are very soluble in most organic solvents, most 1H NMR recorded in literature were specifically done using DMSO-d6. When the 1H NMR of 44 was recorded in DMSO-d6, the sample showed full conversion to the hydrate form (Figure 18). We suspected this is a result of more significant hydrogen bonding between the hydrogen donor, the gem-diol hydrogen atoms, and hydrogen bond acceptor, the oxygen of the sulfoxide group, aiding the equilibrium to the hydrate. However, the 1H NMR of 53 taken in DMSO-d6 did not have this effect. Although, a larger percentage of O O H O O O SeO2, H2O 1,4 dioxane reflux, 7 h 70 O OH OH O 71 83-90% 44 O O O HO MeI, K2CO3 MeCN, RT, 12 h 43 70 94-98% O OH OH O R N NH O O R R NH4OAc MeOH, RT, 24 h 44, 53 72, 73 9 R = Or 38 the hydrate was seen in the 1H NMR, we suspect the network of hydrogen-bonding is somehow diminished by the larger greasy chain. Figure 18 – The difference in 1H NMR spectrum of 44 run in CDCl3 (top, red) vs DMSO-d6 (bottom, blue) using the same sample. 2.1.2 Synthesis of Protected Arginine The synthesis of 55 was completed in two steps, as shown in Scheme 13. The protection of the N and C termini of the arginine group will prevent any interactions between the hydrate and the arginine’s functional groups, excluding the guanidine group. Using a protected arginine derivative first, we can simplify the chemistry and focus on optimising the guanidine's masking. The synthesis of 55 was attempted using two routes, as shown in Scheme 13. The first route relies on Boc protection of the amine then conversion to a methyl ester, while the second route reverses the order of reactions. DMSO CDCl3 O OH OH O 23 4 5 1 1 2 43 5 DMSO H2O 39 Scheme 13 – The two proposed routes to afford the protected arginine 55 (shown in their non-zwitterionic form). Synthesis of the Boc-protected arginine 74 (Route 1) was attempted using two literature procedures (Scheme 14).93,94 Nevertheless, neither showed any sign of product formation. These reactions were repeated, doubling the reaction time and were analysed by 1H NMR and TLC. However, again, showed no sign of product formation. As these reactions do not appear promising, we focused on the addition of the methyl ester first. HN NHH2N H2N O OH HN NHH2N N H O O HN NHH2N N H O OH HN NHH2N N H O O HN NHH2N H2N O O Route 1 Route 2 54 75 55 74 55 O O O O O O 40 Scheme 14 – Initial attempts to synthesise Boc-L-arginine. Synthesis of methyl L-arginine 75 was achieved as a dichloride salt 77, with yields of 86%-90%. The starting material is sparingly soluble in MeOH, therefore the reaction was stirred for 90 minutes before the addition of SOCl2. Mechanistically, the reaction proceeds in three key steps, as shown in Scheme 15. The lone pair of electrons on the hydroxy group drives nucleophilic attack on the sulfur displacing a chloride ion. Addition of the chloride ion to the carbonyl group is followed by collapse of the tetrahedral intermediate to give producing SO2 and chloride anion. A Lewis basic species present deprotonates the oxonium ion affording the acyl chloride. This intermediate then undergoes nucleophilic attack by the lone pair of electrons of MeOH, followed by the collapse of the tetrahedral intermediate, affording 75. Due to the formation of HCl gas during the conversion of the carboxylic acid to the methyl ester the compound was isolated as a hydrochloride salt. HN NHH2N H2N O OH 54 Boc2O, Amberlyst-15, EtOH, 45 °C, 7 h HN NHH2N N H O OH 74 HN NHH2N H2N O OH 54 Boc2O, 1,4-dioxane, 1 M NaOH, RT, 12 h HN NHH2N N H O OH 74 O O O O 41 Scheme 15 – A proposed mechanism to produce 75. The Boc protection of methyl L-arginine was achieved in a yield of 40-66% (Scheme 16). Under mild conditions, Boc protection occurs specifically at the N-terminus due to the guanidine group and the amines having different reactivity. However, these conditions do not result in the complete consumption of starting material, even with the addition of excess Boc2O. Purification of 55 was initially completed by silica-gel chromatography eluting 10% MeOH in CH2Cl2. It was later realised the purification of 55 could be simply accomplished by dissolving the crude solid in CH2Cl2 and passing through filter paper, separating the unprotected arginine derivative from 77. Scheme 16 – Reaction conditions to afford Boc protected methyl arginine, 57. Boc protection of 75 occurs in two steps; first, the lone pair of electrons of the terminal nitrogen attack one of the Boc2O carbonyls, resulting in tert-butyl carbonate acting as a leaving group, which will ultimately break down into CO2 gas and tert-butoxide. In the R OH O R O Cl S OH Cl O R Cl O R O Cl S Cl O OH Cl S Cl O R Cl O -SO2 -Cl- H Cl -HCl H2N HN H2N NH R= R Cl O HOR O O -H 54 7675 O O N H HN H2N NH O O H3N HN H3N NH i) Et3N, MeCN, 25 °C, 24 h ii) Boc2O, MeCN, 25 °C, 15 h 2Cl 77 O O 55 40-66% 42 second step, a Lewis basic species abstracts the proton of the protonated amine. We managed to acquire N-p-tosyl-L-arginine methyl ester hydrochloride from a commercial source during the screening of test reactions, which was utilised in some of the test reactions. Scheme 17 – A proposed mechanism for the Boc protection of 75. 2.1.3 Test Coupling Reactions With a protected arginine and hydrate now in hand, we began our test reactions. As previously mentioned, the modification of arginine derivatives with 43 had been previously reported in the literature. Although the reported methods have questionable chemical analysis, they provided a foundation for our reactions. Various conditions were attempted and are reported in Table 1. In summary, the conditions reported in the literature proved to be unsuccessful. Similar to the literature, analysis of crude reaction material using highly sensitive techniques like mass spectroscopy showed the expected peaks for the desired material; however, 1H NMR showed the reaction had not proceeded, with only starting material present. This result was found throughout for all the conditions reported. During the screening of test reactions, we had come across reports of phenyl O O H2N HN H2N NH O O N O O HN H2N NH H H Et3N O O N H O O HN H2N NH O O O O CO2 + 55 75 O O O O O 43 glyoxal derivatives being sensitive to light, therefore some reactions were completed in the dark (indicated in the table entries). However, this had no effect in our results. As these methods proved to be unsuccessful, we looked at exploring the grassroots of the reaction, condensation of the guanidino and hydrate or aldehyde. In principle, the desired product is formed by an imine formation between an amine and aldehyde, catalysed in the presence of acid. As such, we attempted more “conventional” organic reactions (Entry 29-33). The results of these reaction were much more promising than previous conditions, as mass spectrometry suggests the formation of two products, which could be 78, 79, or 80, any of these products were desirable in our situation. More importantly, the 1H NMR spectrum of the crude reaction material indicated signs of the desired product forming in the aromatic and alkyl chain region of arginine. Attempts to optimise the formation of the product began with merely extending the duration of the reaction. This showed the formation of only one desired product, which we believe to be 79 due to the 1H NMR spectrum and analysis of the mass spectrometry results showing the presence of two hydroxy groups. Attempts to purify the crude mixture proved to be unsuccessful. 44 Table 1 – The reaction conditions utilised to join hydrate and arginine derivatives together with the expected product to be formed. * Entries were repeated in the dark. 10 mM NH4OAc (pH 8), RT, 0.5 hours Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 1* -Ts -OMe -Me 79bef 2 -Ts -OMe -C10H21 79beg 3* -Boc -OMe -Me 79cef 4 -Boc -OMe -C10H21 79ceg 100 mM Sodium borate buffer (pH 9), RT, 16 hours Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 5* -Ts -OMe -Me 79bef 6 -Ts -OMe -C10H21 79beg 7* -Boc -OMe -Me 79cef 8 -Boc -OMe -C10H21 79ceg DMF/AcOH (0.1%), RT, 8 hours Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 9* -Ts -OMe -Me 81bef 10* -Ts -OMe -C10H21 81beg 11* -Boc -OMe -Me 81cef N NH O N H N H NH O NO R3 O R3 HN HN R1 R2 O R1 R2 O HN HN R1 R2 O N NH O B O OH O R3 N N H O O NH O OR3 O R3 HN R1 O R2 45, 5354, 55, 75 R3 O OH OH O NHH2N HN R1 N H O 79 83 84 81 82 R2 R1 = O O S O O H O R2 = HO R3 = , , , , N H N HO HO N H O R3 HN R2 O R1 80 N H N N H O R3 HN R2 O R1 HO 78 N H N N H O R3 HN R2 O R1 HO a b c d e f g 45 12 -Boc -OMe -C10H21 81ceg 100 mM Sodium borate buffer (pH 8), MeCN, 37 °C, 16 hours Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 13* -Ts -OMe -Me 81bef 14 -Ts -OMe -C10H21 81beg 15* -Boc -OMe -Me 81cef 16 -Boc -OMe -C10H21 81ceg HCl(aq), H2O, 100 °C, 0.33 h; 60 °C, 24 h Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 17* -H -OH -Me 81adf 18* -H -OH -C10H21 81adg 19* -H -OMe -Me 81aef 20 -H -OMe -C10H21 81aeg 21* -Ts -OMe -Me 81bef 22 -Ts -OMe -C10H21 81beg 23* -Boc -OMe -Me 81cef 24 -Boc -OMe -C10H21 81ceg Phosphate buffer (pH 7.4), 40 °C, 4.5 hours Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 25* -Ts -OMe -Me 79bef 26 -Ts -OMe -C10H21 79beg 27* -Boc -OMe -Me 79cef 28 -Boc -OMe -C10H21 79ceg AcOH, CH2Cl2, RT, 24 hours Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 29* -Ts -OMe -Me 81bef 30 -Ts -OMe -C10H21 81beg 31 -Boc -OMe -Me 81cef 32 -Boc -OMe -C10H21 81ceg 37% HCl(aq), 3 weeks Entry Arginine R1 Arginine R2 Hydrate R3 Expected product 33* -H -H -Me 81adf It is unclear why reactions failed to proceed, although there are a few possible reasons. Firstly, as depicted in Table 1, the reaction is likely to be occurring in equilibrium, and like the issues encountered with the gem-diol hydrate equilibrium a similar issue could be occurring here. This makes analysis of the product through 1H NMR difficult to interrupt. As mentioned earlier, reports in the literature have only characterised derivatised peptides by mass spectrometry or HPLC. In addition, they have rarely quoted yields. It is unclear how effective this reaction is. We could have been chasing an inherently flawed method. 46 Ultimately, this project was abandoned due to the lack of time, and successful results. Further steps required to explore this strand of research and form a prodrug with opiorphin have been expanded on in Chapter 4. 47 2.2 Section 2 – Unnatural Amino acid Section 2 of results and discussion is split into two parts, in part 1, each step to produce the amino acid precursor (±)-58 will be discussed in detail. While part two discusses coupling reactions involving the planar chiral amino acid reactions. This is done as the steps to produce the amino acid precursor (±)-58, plays a critical role and is the foundation of the results of Sections 3 and 4. 2.2.1 Part 1 – Synthesis of the Amino Acid Precursor As noted in Chapter 1 the synthesis of (±)-4-nitro[2.2]paracyclophane-13-carboxylic acid has been reported in the literature. The previous method prepared (±)-3 in five steps, the method used in this strand of research has been developed by members of the Rowlands group, and produces the desired precursor of (±)-3 in three steps. Synthesis of (±)-58 was achieved by nitration producing (±)-85, a directed formylation to give aldehyde (±)-86 followed by oxidation to the desired amino acid precursor (Scheme 18). Scheme 18 – Synthetic pathway for the synthesise of (±)-4-nitro[2.2]paracyclophane-13-carboxylic acid. Nitration The first step of the synthesis of (±)-58 was the electrophilic substitution of [2.2]paracyclophane to produce (±)-4-nitro[2.2]paracyclophane (±)-85. Nitration of [2.2]paracyclophane is a capricious reaction and, in our hands, led to an unexpected rearrangement that will be discussed in Section 4. Various methods have been reported in literature and they cover a bewildering spectrum of conditions. Some reactions cool the reaction -72 °C while others heat to 100 °C.95,96 Time can vary from 10 minutes to 4 days.97 The Rowlands group developed an alternative procedure utilising a 4:2 stoichiometry of H2SO4 to HNO3 added separately in CH2Cl2 at 0 °C and stirred overnight. Lower temperature initially results in slower formation of side product, while the NO2 O OH NO2 ONO2 H 16 (±)-85 (±)-86 (±)-58 48 increased equivalents of reagents ensures the reaction proceeds to completion. However, these conditions posed their own difficulties in the formation of a black-tar. This is likely a product of [2.2]paracyclophane polymerisation, and not further investigated. The tar was decanted before quenching the reaction, as the presence of this material results in the formation of an emulsion, making the work-up difficult and decreasing yield. Under these conditions the yield varied between 67-80%, but purification is simple using standard column chromatography. Scheme 19 – Reaction conditions used to synthesise (±)-85. In an effort to reduce the formation of the black-tar substance the reaction was further optimised by developing a more controlled procedure. The H2SO4 was decreased to 2 equivalents and added to HNO3 at 0 °C for 15 minutes before adding to the solution of [2.2]paracyclophane in CH2Cl2. Additionally, monitoring the reaction by TLC showed consumption of starting material after 8 hours even while keeping the reaction at 0 °C, rather than bringing it to RT overnight. The decreased ratio of reagents, reaction time and temperature prevented the formation of the black tar and other side products. These conditions afforded yields in the range of 80-87%. Nitration of [2.2]paracyclophane occurs through electrophilic aromatic substitution as shown in Scheme 20. The nitronium ion is produced by dehydration of HNO3 with H2SO4. Electrophilic aromatic substitution occurs via a carbocation intermediate known as an arenium ion intermediate, the aromaticity is then restored by the removal of the proton through the nucleophilic attack of any Lewis basic species present in solution. The mono- substitution of [2.2]paracyclophane leads to the symmetry being broken, resulting in formation of a mixture of enantiomers. NO2H2SO4, HNO3 CH2Cl2, 0 °C, 16 h (±)-85 80-87% 16 49 Scheme 20 – Proposed mechanism for the nitration of [2.2]paracyclophane using H2SO4 and HNO3. Formylation Synthesis of (±)-4-nitro-13-formyl[2.2]paracyclophane (±)-86 was achieved through Rieche formylation (Scheme 21). First reported by Alfred Rieche in 1960, the Rieche formylation utilises MeOCHCl2 and TiCl4 in a Friedel-Crafts-like aromatic substitution.98 Formylation of (±)-85 occurs successfully with yields ranging from 73-84% and can be purified by column chromatography. Slow evaporation of the eluent (40% EtOAc, 60% hexane) produces crystals suitable for X-ray crystallography analysis. Although this reaction works well, the use of TiCl4 required careful handling and isn’t user friendly due to it being sensitive to air. In an effort to find an alternative reagent, a range of Lewis acids, including AlCl3, FeCl3, or ZrCl4 were screened. Although the reagents are significantly easier and safer to handle, all three reagents gave lower yields (17-34%), increased side products, and required longer reaction times. Scheme 21 – Rieche formylation of (±)-85 to afford (±)-86. Formylation of (±)-85 begins with the Lewis acidic TiCl4 activating MeOCHCl2. Elimination of a chloride ion generates a powerful electrophilic oxonium species. The O N O O O SH OHO O O N O OH2 O N O H2O H + O N O NO2 H NO2 16 (±)-85 NO2 NO2 O H TiCl4, MeOCHCl2 CH2Cl2, -10 °C, 16 h (±)-85 (±)-86 73-84% 50 regioselectivity can be explained by the transannular effect. This postulates that the initial addition is reversible and occurs anywhere on the rings, but the re-aromatisation of the Wheland intermediate is faster at C13 and this funnels the product down this route. Aromaticity is restored by an internal deprotonation by an oxygen atom on the nitro group. An additional chloride ion is lost, and H2O is added to the oxonium electrophile. Protonation of the hemiacetal results in the oxonium MeOH species, which acts as a good leaving group. Finally, deprotonation of the oxonium species affords the desired product (±)-86. Scheme 22 – Mechanism of Riche formylation of (±)-4-nitro[2.2]paracyclophane (±)-86. NO2 O (±)-86 N H Cl O O O NO2 O Cl H NO2 O H2O NO2 O O NO2 OH O H NO2 O H -MeOH Cl N Cl O O O H H H Cl H+ H Cl ClCl O TiCl4 N O Cl O Cl TiCl5 O Cl (±)-85 + O O 51 Formylation occurs specifically at the C13 position of (±)-85 and with the regioselectivity explained by the transannular effect. The transannular effect isn’t fully understood but a working model suggests it is a combination of the electronic properties of one aromatic ring influencing the other and an internal deprotonation. This can easily be seen by comparing two electron withdrawing groups, the nitro group and a nitrile. Normally, the rate determining step of aromatic substitution is the breaking of the aromaticity when the ring attacks a suitable electrophile. In [2.2]paracyclophane with an electron withdrawing substituent, this is probably not the case. Formation of the cationic intermediate is reversible and the ability to lose the hydrogen and regain aromaticity within the system appears to be key. A nitro group can aid internal deprotonation as there is a conformation in which the oxygen lone pair can act as a base. This accelerates irreversible re- aromatisation, causing substitution at the C13 position to be favoured. A nitrile group is also electron withdrawing but it does not direct to the pseudo-geminal position suggesting electronics are not that important. This can be attributed to the nitrile group being linear, forcing the lone pair to be orientated outwards so that it cannot participate in internal deprotonation. Other functional groups that can rotate can cause internal deprotonation and direct to pseudo-geminal position. This means the transannular effect is a result of selective internal deprotonation. It is further presumed the first step occurs in some form of equilibrium. This would allow the initial step to occur at alternative positions on the ring, but only productively proceed when internal deprotonation allowed. All other intermediates reform starting material. It is only at the C13 positions a rapid rearomatisation occurs, which drives the equilibrium to the product side. Figure 19 – Visual representation showing how the lone pair orientation and rotation aids in internal deprotonation. N R H N R H O O 52 Oxidation The synthesis of (±)-4-nitro[2.2]paracyclophane-13-carboxylic acid (±)-58 was achieved with yields ranging from 10-53%, while returning 28%-43% of starting material (±)-86 (Scheme 23). Scheme 23 – Reaction conditions used to afford (±)-58 Initially, it was believed that the reaction did not go to completion due to the oxidation of acetone to acetic acid. This was presumed due to a strong aroma of vinegar emitted during the work up and later confirmed by the presence of sodium acetate crystals. A lower concentration of acetone (0.4 M rather than 0.2 M) was used in following reactions with the intention of resolving this issue, increasing the average yield by 8%. However, this was not a significant increase in yield, so an alternative water-soluble solvent was attempted. However, no other solvent improved the yields. This suggests that oxidation of acetone was not the real issue. Currently, it is not known why oxidation is not occurring efficiently, but we believe optimisation of pH should be investigated. Purification by acid/base extraction gives a simple method to obtain clean product, as well recycling starting material (±)-86. An X-ray crystal structure of (±)-58 was acquired by slow evaporation of CD3OD. The connectivity of the compound has been confirmed, but the data quality was not sufficient for full structural characterisation. NO2 NO2 O OH KMnO4, H2O acetone, RT, 16 h (±)-86 (±)-58 10-53% H O 53 Figure 20 – X-crystal structure of (±)-58 acquired through slow evaporation of CD3OD. Carboxylic acid hydrogen omitted. C=grey, H=white, N=blue, O=red. Thermal ellipsoid probability level 50%. Oxidation of (±)-86 with KMnO4 begins with the nucleophilic addition of H2O to the electrophilic carbonyl forming a hydrate (Scheme 24). Followed by the negatively charged oxygen to attack KMnO4 allowing for proton transfer to oxidise the tetrahedral carbon producing the desired carboxylic acid (±)-58. Scheme 24 – Mechanism for the oxidation of (±)-4-nitro-13-formyl[2.2]paracyclophane to of (±)-4- nitro[2.2]paracyclophane-13-carboxylic acid (±)-58. NO2 O H H2O NO2 O H OH2 NO2 O H OH Mn O O O O NO2 O H OH Mn O O OO -H+ (±)-86 NO2 O OH (±)-58 NO2 O OH (±)-60 54 This provided a short route of the preparation of amino acid surrogates. In this nitro form, the amine is effectively protected allowing selective reactions of the acid. Alternatively, reduction would allow access to the amine. 55 2.2.2 Part 2 – Synthesis of a Dimer To couple the carboxylic acid, we needed to activate it. As a reaction between acid and base results in salt formation, conversion to an acyl chloride is a required step before reacting (±)-58 with an amine. Conversion to an acyl chloride allows for the substitution of the halide when reacted with an amine, resulting in the formation of an amide. Additionally, the hydroxy group of a carboxylic acid is a poor leaving group, thus conversion to a more labile group is required. However, the lability of an acyl chloride can be a double-edge sword as (±)-87 can be readily converted back to starting material in the presence of H2O, which can be difficult to avoid due to moisture in the air. The synthesis of (±)-4-nitro[2.2]paracyclophane-13-carbanoyl chloride was completed using two procedures (Scheme 25). Initially, a mild procedure utilising (COCl)2 with a sub- stoichiometric amount of DMF in CH2Cl2 was used. The 1H NMR spectrum obtained from the following transformation when using the (COCl)2 and DMF pathway showed the presence of remaining starting material, demonstrating the reaction does not go to completion. As such, the procedure was changed to refluxing (±)-58 in SOCl2. Unsure of the stability of (±)-87, collection of 1H NMR was not attempted and the acyl chloride was used immediately for the following reactions. Although the yields for either procedur cannot be reported, yields for the following transformations were at least 15% higher when using SOCl2. Scheme 25 – Reaction conditions used to synthesise (±)-87. The mechanism for the initial procedure begins with a reaction between DMF and (COCl)2 to ultimately form the imidoyl chloride known as the Vilsmeier reagent. This imidoyl chloride derivative is the active chlorinating agent, which reacts with the lone pair of electrons on the carboxylic acid. Following the loss of hydrogen, the lone pair on the nitrogen forms a double bond causing a chloride to be eliminated. The chloride then NO2 O OH NO2 O Cl Initial procedure - (COCl)2, DMF in CH2Cl2, RT, 1 h Final procedure - SOCl2, reflux, 15 h (±)-58 (±)-87 56 attacks the carbonyl group producing the desired product. The mechanism utilising SOCl2 can be seen in Section 1 Scheme 15, however stops at the acyl chloride intermediate. Scheme 26 – The proposed mechanism to synthesise the Vilsmeier reagent and (±)-87. The synthesis of the first peptide bond is a crucial step in the synthesis of the foldamer as it produces the end terminus of the molecule, preventing uncontrolled polymerisation. Three carboxamides were synthesised (Scheme 27). Initially (S)-1-phenylethylamine was used; however, this was changed to (R)-1-phenylethylamine due to the limited quantity of the reagent present in the lab. The synthesis of 88a was achieved in 56% yield while 88b gave a range of 52-56%, both resulting in the formation of a 1:1 mixture of diastereoisomers. The use of a chiral amine was intended to give diastereoisomers in the hope that we could resolve the planar chirality of the [2.2]paracyclophane amino acid. We needed to resolve the planar chirality before we started linking [2.2]paracyclophane units in order to prevent the synthesis of mixtures of stereoisomers. Unfortunately, the Rf values of the two were too close and resolution was impossible. It was noted during the N Cl H NO2 O OH NO2 O O-H+ Cl H N NO2 O O H N Cl N O H NO2 O Cl N O Cl OCl O N O O Cl O Cl N O O Cl O Cl -CO -CO2 -Cl- N Cl H (±)-58 (±)-87 57 purification that a precipitate had formed in the collected fractions, and when analysed by 1H NMR, the precipitate showed a greater proportion of one diastereoisomer to the other. With further experimentation, recrystallisation of a single diastereoisomer may have been possible. As the separation of the phenylethylamine diastereoisomers was not achieved, it was decided to simply form the amide as a racemic mixture. This would simplify the 1H NMR spectrum. Furthermore, in an attempt to increase yields CH2Cl2 was substituted for THF, which allowed the reaction to be heated at a higher temperature (Scheme 27). The isopropyl derivative (±)-88c was formed as a racemate in yields between 64-76% with no complications; this procedure showed an increase in yield by 17% compared to the diastereomeric mixture earlier. Scheme 27 – Reaction conditions used to synthesise various (±)-4-nitro[2.2]paracyclophane carboxamide derivatives. Amide formation is a standard acyl substitution reaction involving an addition- elimination mechanism. The first step is the nucleophilic addition of the amine to the partially positive carbon of the carbonyl. The elimination proceeds by the collapse of the tetrahedral intermediate to reform the carbonyl group with the expulsion of the chloride ion, followed by the deprotonation of the nitrogen by any Lewis basic species present to afford the product. NO2 O Cl NO2 O N H RRNH2, Et3N Initial procedure - CH2Cl2, 0 °C, 12 h Final procedure - THF, 0 °C 1 h; reflux, 13 h (±)-87 (±)-88c 64-76% 88a 56% 88b 52-56% 58 Scheme 28 – A proposed mechanism to produce amide monomer building blocks. Reduction of the nitro amides to amino amides has been achieved for three compounds with yields ranging from 50-86%. The nitro group had effectively behaved as a protecting group up to this point, reduction to the amine unmasks the nucleophile required in subsequent couplings. The reduction of 88a was acquired in 50% while 89b was acquired with yields from 57%-64%. Reduction of (±)-88c resulted in a 86% yield, with no further purification required. Scheme 29 – Reaction conditions used to synthesise 4-amino[2.2]paracyclophane carboxamide derivatives. NO2 O Cl RH2N NO2 O Cl H2N R NO2 O N R H H Cl NO2 O N H R (±)-87 NO2 O N H NH2 O N H RZn, NH4Cl MeOH, reflux, 16 h (±)-89c 86% R 89b 57-64% 89a 50% 88a, 88b, (±)-88c 59 The mechanism of the reaction can be seen in Scheme 30. Conversion of the nitro to the amine begins with the protonation of the hydroxy group followed by a series of single- electron transfers (SET) and protonation to produce unstable oxonium species. Loss of H2O gives a nitroso group that is protonated and reduced by a second series of SETs. A strong acid is required to protonate the oxygen of the resulting hydroxylamine and not the nitrogen. Loss of H2O and reduction by yet another series of SETs creates a nitrenium ion, which is short-lived in an acid medium, being protonated to give the desired amine. Scheme 30 – Proposed mechanism to for the reduction of the nitro group to amino for (±)-89c. Amide formation between the capped amino amide derivatives and (±)-87 is arguably the most fundamental reaction for this project as it allows for the coupling of monomers. The initial attempts at coupling [2.2]paracyclophane derivatives gave poor yields, with both phenylethyl derivatives resulting in a crude yield of 4%. As separation of diastereomers had not been achieved, the coupling results in up to eight stereoisomers. The 1H NMR R N O O R N OH O R N OH O R N OH O R N OH OH2 R N O R N H OH R N H R N H H Nitro Nitroso Amine Hydroxylamine Nitrenium ion H+ H+ O N H Zn HZn Zn R N O H Zn R N O H R = R N H OH2 Strong acid required to protonate O and not N H+ R N H Zn R N H Zn (±)-88c (±)-89c -H2O -H2O 60 was inconclusive as it contained a mixture of potential products and impurities. The mass spectrometry analysis of the crude reaction mixture showed the desired mass and isotopic abundance splitting of a predicted C42H39N3O4 (Figure 21), however, purification was not attempted due to the limited amount of material. With both derivatives giving poor results, we knew we required better coupling conditions. Exploring alternative methods originated from these results and lead to the procedure discussed earlier using THF and refluxing. Scheme 31 – Reaction conditions used to synthesise [2.2]paracyclophane dimer derivatives. The changes paid off and resulted in the isolation of (±)-90c obtaining a yield of 34% after purification, showing a significant increase in yield. Analysis of 1H NMR showed signs of hindered rotation of the isopropyl group, as the methyl groups are no longer equivalent to one another in the 1H NMR spectrum. However, further research is required to determine whether this restricted rotation is due to the adjacent [2.2]paracyclophane. This could explain the low yields acquired for 90a and 90b, as phenyl groups are much bulkier than isopropyl groups and thus hinder the reaction from occurring. NH2 O N H Initial procedure - CH2Cl2, 0 °C, 12 h Final procedure - THF, 0 °C 1 h; reflux 13 h (±)-90c 34% R NO2 O Cl (±)-87 NO2 N H O NH O R 90b 4% 90a 4% 89a, 89b, (±)-89c Et3N 61 Figure 21 – Mass spectrometry of 90b highlighting the isotopic splitting and predicated splitting. The work-up for these reactions was complicated by the formation of gels. Adding extra organic solvent during separation failed to help the situation. These emulsions were observed in later coupling reactions (see compounds (±)-103 and 110). We wonder if these amides act as gelators analogous to the b-amino acids derivatives reported by Hirose et al.; or the phenylaniline derivatives synthesised by Díaz et al. The task of examining the gel is beyond the scope of this research.99,100 2.3 Section 3 – Learning from our First Attempts to Synthesise a Foldamer Although, there were promising signs of progress up to this stage, we had not addressed a few critical problems. Firstly, the resolution of [2.2]paracyclophane derivatives had not been completed. As mentioned e