LEABHARLANN CHOLAISTE NA TRIONOIDE, BAILE ATHA CLIATH TRINITY COLLEGE LIBRARY DUBLIN OUscoil Atha Cliath The University of Dublin Terms and Conditions of Use of Digitised Theses from Trinity College Library Dublin Copyright statement All material supplied by Trinity College Library is protected by copyright (under the Copyright and Related Rights Act, 2000 as amended) and other relevant Intellectual Property Rights. By accessing and using a Digitised Thesis from Trinity College Library you acknowledge that all Intellectual Property Rights in any Works supplied are the sole and exclusive property of the copyright and/or other I PR holder. Specific copyright holders may not be explicitly identified. Use of materials from other sources within a thesis should not be construed as a claim over them. A non-exclusive, non-transferable licence is hereby granted to those using or reproducing, in whole or in part, the material for valid purposes, providing the copyright owners are acknowledged using the normal conventions. Where specific permission to use material is required, this is identified and such permission must be sought from the copyright holder or agency cited. Liability statement By using a Digitised Thesis, I accept that Trinity College Dublin bears no legal responsibility for the accuracy, legality or comprehensiveness of materials contained within the thesis, and that Trinity College Dublin accepts no liability for indirect, consequential, or incidental, damages or losses arising from use of the thesis for whatever reason. Information located in a thesis may be subject to specific use constraints, details of which may not be explicitly described. It is the responsibility of potential and actual users to be aware of such constraints and to abide by them. By making use of material from a digitised thesis, you accept these copyright and disclaimer provisions. Where it is brought to the attention of Trinity College Library that there may be a breach of copyright or other restraint, it is the policy to withdraw or take down access to a thesis while the issue is being resolved. Access Agreement By using a Digitised Thesis from Trinity College Library you are bound by the following Terms & Conditions. Please read them carefully. I have read and I understand the following statement: All material supplied via a Digitised Thesis from Trinity College Library is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of a thesis is not permitted, except that material may be duplicated by you for your research use or for educational purposes in electronic or print form providing the copyright owners are acknowledged using the normal conventions. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone. This copy has been supplied on the understanding that it is copyright material and that no quotation from the thesis may be published without proper acknowledgement. Chiral Hypernucleophilic Acylation Catalysts and Synthesis of Bipyridyls via Palladium-Catalysed Reductive Homocoupling of Chloropyridines By Richard Pio Lee A thesis presented to the University of Dublin for the degree of Doctor of Philosophy December 2000 Trinity College Dublin 0 3 DEC 2001 U ^A R V D 08U N Declaration This thesis has not been submitted as an exercise for a degree at any other university. Except where acknowledged all the work described is original and carried out by the author alone. I agree that the library may lend or copy this thesis upon request. Summary Chapter 1 o f this thesis discusses the explosion o f interest in the non-enzymatic kinetic resolution o f secondary alcohols catalysed by chiral derivatives o f the hypemucleophilic acylation catalyst, 4-dimethylaminopyridine. Particular attention is drawn to the work o f Gregory Fu and Craig Ruble o f the Massachusetts Institute of Technology, Kaoru Fuji o f Kyoto University and Edwin Vedejs o f the University of Wisconsin who have been pioneers in this type of work during the last five years, and who collectively represent the benchmark in terms of non-enzymatic kinetic resolution via enantioselective acylation. Also reviewed is the work o f other more recent entrants into the field such as Alan Spivey of Sheffield University and Tarek Sammakia o f the University of Colorado. Finally, our own tentative steps into this area are discussed as is the structure of the present work, the majority o f which has been carried out during the last three years. Chapter 2 discusses in detail several efficient procedures for the synthesis o f 2-halo and 2,6-dihalo-4-dialkylaminopyridines, precursors to the more elaborate compounds discussed in Chapter 3. Also there is a substantial discussion on other synthetic routes to 2-halo-4-dialkylaminopyridines which did not have the desired outcome, particularly the failed conversion o f 4-(r-pyrrolidinyl)-2-pyridone into 2-chloro-4- ( l ’-pyrrolidinyl)pyridine, and the failure to synthesise 2-amino-4- dimethylaminopyridine in acceptable quantities via the generation o f a heteroaryne from 3 -bromo-4-dimethylaminopyridine. Chapter 3 describes the synthesis o f novel chiral 4-dialkylaminopyridines incorporating the terpene alcohols (-)-menthol and (-)-bomeol from 2-bromo-4- dimethylaminopyridine. Thus, (-)-2-bomyloxy-4-dimethylaminopyridine and (-)-2- menthyloxy-4-dimethylaminopyridine proved to be efficient catalysts in the acylation of 1-phenylethyl alcohol but showed no enantioselectivity. Also researched was the ability o f heterocyclic A^-oxides to act as catalysts for the acylation of 1- phenylethyl alcohol in the presence o f acetic anhydride. O f particular significance was the fact that 2-dimethylaminopyridine A^-oxide is an efficient acylation catalyst whereas the free base is inactive. This implied that potential catalysts may not require a 4-dimethylamino substituent to achieve an acceptable rate o f reaction. A number o f homochiral 7V-oxides were synthesised and indeed they did act as efficient acylation catalysts but showed no enantioselectivity. A detailed kinetic investigation into the acylation o f 1-phenylethyl alcohol in deuteriochloroform was undertaken for all compounds showing catalytic activity. It was shown that 4- dimethylaminopyridine was only 1.2 times better than 4-dimethylaminopyridine N- oxide as an acylation catalyst. 2-Dimethylaminopyridine A^-oxide was shown to be almost twice as effective as pyridine under similar conditions. Chapter 4 deals with an efficient palladium-catalysed synthesis o f bipyridines, which is believed to occur via a reductive homocoupling reaction. The regeneration of palladium(O) was believed to occur via a Wacker reaction involving the alkene present in the reaction mixture although this has not been conclusively demonstrated. Also discussed are initial attempts to synthesise the novel ligand 4,4’- dimethylamino-2,2’-bipyridine from the 2-bromo-4-dimethylaminopyridine synthesised in Chapter 2. Acknowledgements First and foremost I would like to take this opportunity to thank my supervisor Dr. David H. Grayson. I greatly appreciate all the encouragement and help he has given me in that time. Thanks to Enterprise Ireland for funding and Olin Corporation for the generous donation o f 2-chloropyridine A^-oxide. Thanks also to Seal Sands Chemicals and to The Reilly Tar & Chemical Corporation for the kind donation of substantial amounts o f 4-dimethylaminopyridine. I would like to thank my parents for indulging me in my pursuit o f education and knowledge. Without their financial assistance I would never have travelled so far along this long road of enlightenment. Many thanks to my favourite sisters Margaret and Helen who gave me a bed on the many occasions they found me intoxicated and incapacitated on their doorstep. I am also grateful to many o f the technical staff especially Dr. Martin Feeney for running all my mass spectra so quickly, thanks also to John Kelly for repairing all my broken glassware and Brendan Mulvany for repairing all things electrical. Thanks also to Paul Byrne for loans o f various implements over the last three years. Many thanks to Dr. John O ’Brien for the many hundreds o f NMR spectra he promptly ran for me, it is really appreciated. Special thanks to Dr. Sylvia Draper for solving the X-ray crystal structure o f 4-dimethylamino-3-methylsulfanyl-2-pyridone. Special mention is due to the Grayson group both past and present especially Una, Gilles, Eleonora, Philip, Gillian, Marco, Jose, David and Guillaume. Also to the many friends I have made in Trinity especially Damien and Anthony, Conor, Elise, Muriel and Fiona. Many thanks to my other friends particularly Eugene, Thomas, Mervan and the rest o f the lads from Navan who are too numerous to mention. Thanks to Peter, Brian and Jamie and the rest o f the lads in the Applied Chemistry class at DCU. The laughs we share will never be forgotten. God bless us all. Abbreviations Ac acetyl Aq. aqueous b.p. boiling point bs broad singlet COSY correlation spectroscopy d doublet dd double doublet dt double triplet DCM dichloromethane DIBAL diisobutylaluminium hydride DMA A^,7V-dimethylacetamide DMAP 4-dimethylaminopyridine DMF A^,7V-dimethylformamide DMSO dimethylsulfoxide g grams IPA isopropyl alcohol IR infared LDA lithium diisopropylamide LiTMP lithium 2,2,6,6-tetramethylpiperidine mm Hg millimetres o f mercury m.p. melting point ml millilitre MHz megahertz MgS04 magnesium sulphate NBS A^-bromosuccamide Na2C03 sodium carbonate NaHCOa sodium hydrogen carbonate NaN02 sodium nitrite NaOH sodium hydroxide NMR nuclear magnetic resonance ppm parts per million PPY 4-(r-pyrrolidinyl)pyridine quat. quaternary q quartet r.t room temperature rxn reaction s singlet TEA triethylamine THF tetrahydrofuran TMS trimethylsilyl T.l.c thin layer chromatography Table of Contents Chapter 1 Introduction 1.1 Aims of this project........................................................................................... 1 1.2 Enantiomers........................................................................................................2 1.3 Resolution of enantiomers...............................................................................5 1.3.1 Classical resolution..................................................................6 1.3.2 Kinetic resolution..................................................................... 7 1.3.3 Kinetic resolution mediated by enzyme catalysis............... 8 1.3.4 Dynamic kinetic resolution.....................................................9 1.3.5 Non-enzymatic kinetic resolution........................................15 1.4 4-DiaIkyIaminopyridines...............................................................................22 1.4.1 Mechanism of catalysis......................................................... 24 1.4.2 Chiral 4-dialkylaminopyridines............................................27 1.4.3 Chiral 4-dialkyaminopyridine A^-oxides..............................50 1.5 Detailed project description......................................................................... 52 1.6 References........................................................................................................ 56 Chapter 2 Synthesis of 2-halo and 2,6-dihalo-4-dialkvlaminopyridines 2.1 Introduction.....................................................................................................60 2.2 Synthesis of 2-chloro-4-dimethylaminopyridine and 2-bromo- 4-dimethylaminopyridine via diazotisation of 2-amino- 4-dimethylaininopyridine............................................................................60 2.3 Synthesis of 2-bromo-4-dimethylaminopyridine using Fort’s base..72 2.4 Attempted synthesis of 2-chloro-4-(l’-pyrrolidinyl)pyridine via reaction of 4 -( l’-pyrrolidinyl)-2-pyridone with various phosphorus reagents.................................................................................... 76 2.5 Attempted synthesis of 2-amino-4-dimethylaminopyridine via the trapping of a heteroaryne generated from 3-halo-4-dimethylaminopyridines............................................................ 92 2.6 Synthesis of 2,6-diiodo-4-dimethylaminopyridine A'-oxide................ 101 2.7 Experimental section...................................................................................104 2.8 References..................................................................................................... 130 Chapter 3 New chiral acylation catalysts 3.1 Synthesis of (-)-2-bornyloxy-4-dimethylaminopyridine and (-)-2- menthyloxy-4-dimethylaminopyridine...................................................133 3.1.1 Possible reaction mechanism for the formation o f 4- dimethylamino-3 -methylsulfanyl-2-pyridone.........................140 3.1.2 (-)-2-Bomyloxy-4-dimethylaminopyridine and (-)-2-menthyloxy-4-dimethylaminopyridine as acylation catalysts...................................................................... 148 3.2 Heterocyclic A^-oxides as acylation catalysts........................................ 152 3.2.1 Synthesis of (-)-2-bomyloxypyridine A^-oxide and (-)-2-menthyloxypyridine A^-oxide and their evaluation as acylation catalysts.............................................. 162 3.3 Synthesis of Ca symmetric acylation catalysts...................................... 167 3.4 Synthesis of 2 -(l’-phenylethylamino)-4-dimethylaminopyridine A^-oxide and ( l ’5 )-(-)-2 -(l’-phenylethylamino)-4-dimethylamino pyridine and their evaluation as acylation catalysts...........................172 3.5 Measurement of rate constants for the acylation of (±)-lphenylethyl alcohol in deuteriochloroform catalysed by compounds with catalytic activity...........................................................................................................179 3.6 Discussion...................................................................................................... 185 3.7 Experimental section...................................................................................187 3.8 References.....................................................................................................203 Chapter 4 A novel synthesis of bipvridines 4.1 Introduction................................................................................................. 204 4.2 Synthesis of bipyridines.............................................................................207 4.3 Optimisation of the reaction conditions................................................. 210 4.4 Possible reaction mechanism..................................................................... 212 4.5 Synthesis of symmetrically substituted bipyrid ines............................ 218 4.6 Prelim inary investigation into the synthesis of the novel ligand 4,4’-dimethylamino-2,2’-bipyridine........................................... 221 4.7 Experim ental section................................................................................... 225 4.8 References..................................................................................................... 232 Chapter 1 Introduction 1.1 Aims o f this project The purpose o f this project is to synthesise chiral catalysts which will act as enantioselective acyl transfer reagents. Typically, these catalysts will incorporate a pyridine ring bearing a 4-dimethylamino moiety within their structure to confer hypemucleophilicity. When the nitrogen atom of such a pyridine ring exists in an asymmetric environment, it follows that each enantiomer o f a racemic alcohol should interact differently with an intermediate jV-acyl pyridinium ion. In an ideal situation one enantiomer would be selectively acylated leaving the other unchanged. This would provide a simple, straightforward method for the resolution of compounds such as alcohols, thiols and amines (Scheme 1.1). It is the realisation o f this hypothesis that is the subject of this project. Scheme 1.1 R r 'C O X (0.5 Mol) ^ (±) R^OH (1 Mol) N Chiral Catalyst R* 'N R* o r‘ X' (-) o r (+) - HX (+) o r (-) R^OH R = N M ej, N E tj, P y rro lid in -l-y l R* = C h ira l substltu ten t Chiral Catalyst X = Cl, GAc 1 1.2 Enantiomers Enantiomers are molecules that possess non-superimposible mirror images (Figure 1.1). Enantiomers have the same physical and chemical properties except that they rotate plane polarised light equally in opposite directions. However, when enantiomers interact with other chiral compounds or with chiral environments different effects are observed due to diastereomeric interactions. Figure 1.1 A '""D A B Enantiomer recognition is very important in biological processes. In order for an enantiomer to exert a biological action it must fit into a chiral receptor at the target site, into which the other enantiomer will not fit. If however, the other enantiomer fits into a chiral receptor elsewhere in the body there may a severe adverse biochemical reaction to it. An example o f enantiomer recognition is limonene which gives orange and lemon skins their characteristic smells. (/?)-Limonene 1 is found in oranges while the other enantiomer (5)-limonene 2 is found in lemons. 2 A further example of how enantiomer recognition operates in biological systems is 2- amino-3-(3,4-dihydroxyphenyl)propanoic acid (DOPA) which is used in the treatment of Parkinsons’ disease. The prodrug DOPA is able to cross the blood brain barrier and reach the site of action where the enzyme dopamine decarboxylase decarboxylates only the (S) enantiomer of DOPA 3 to give the achiral therapeutic agent dopamine. Dopamine itself cannot cross the blood brain barrier. It is therefore inadvisable to administer racemic DOPA as there would be an accumulation of (^)- DOPA 4 in the body. HO HO COOHCOOH 3 4 Similarly, (5)-penicillamine 5 is an antidote for lead, gold and mercury poisoning while (/?)-penicillamine 6 can cause atrophy of the eyes, which may lead to blindness. COOHCOOH 3 The necessity for enantiopure compounds was highlighted tragically in the 1960’s with the drug thalidomide, commercially known as Softenon, which was administered as a sedative. (/?)-Thalidomide 7 acted as a sedative while (S)- thalidomide 8 was teratogenic, inducing malformations in the unborn child.’ This would have been avoided had the individual enantiomers o f thalidomide and the racemate been tested prior to commercialisation. 7 8 In 1992 the US Food and Drug Administration (FDA) and the European Committee for Proprietary Medicinal Products required manufacturers to research and characterise each enantiomer o f all drugs proposed to be marketed as a racemate. Consequently, production of new racemates ceased to be a rational commercial option and instead became a high-risk route for pharmaceutical companies. The regulators also foresaw the redevelopment of existing racemates as single isomer drugs. An example is the anti-obesity drug dexfenfluramine 9 whose racemate was initially developed by Servier. 4 NHEt 9 The single isomer drug (Redux) developed by Intemeuron Pharmaceutical and manufactured by Wyeth Ayerst received FDA approval in 1996 with the advantage o f having reduced side effects. The FDA has announced proposals to extend the market exclusivity for newly approved drugs that are single isomer to five years. Racemates superseded by single isomer drugs are to be withdrawn. Currently, 80% of all drugs entering development are chiral and 75% of all man-made drugs will be single enantiomer by year 2000.^ It is also estimated that the market for single enantiomer drugs will increase from $73 billion in 1996 to over $90 billion by the year 2000.'* Therefore, the search for efficient syntheses o f enantiomerically pure compounds is an active area o f research both in academia and industry. Resolution methods are effective sources of enantiopure compounds, as is the chiral pool and asymmetric synthesis. This project is primarily concerned with kinetic resolution as an effective method for the separation of enantiomers. 1.3 Resolution of enantiomers Most methods of resolution utilise one basic property o f enantiomers: when a pair of enantiomers interact with a chiral reagent, the resulting diastereomers have different properties. Exploitation of this phenomenon allows the enantiomers to be separated. A number o f methods have been devised for the resolution o f enantiomers. 5 1.3.1 Classical resolution The products formed by the interaction o f a pair o f enantiomers with a chiral reagent are diastereomerically related. The resulting diastereomers will have different properties such as melting points and solubilities, and can often be separated by physical methods such as fractional recrystallisation, distillation, extraction, column chromatography and gas liquid chromatography.^ Resolution is achieved if the desired enantiomers can be individually regenerated from the separated diastereomers. This represents a convenient method for the resolution o f racemic acids, bases, amino-acids, alcohols, aldehydes and ketones. Classical resolution is still the method o f choice industrially for the separation o f racemates. DSM/Andeno resolve (±)-phenylglycine with (+)-10-camphorsulfonic acid 10 to give (R)-{-)- phenylglycine 11, an intermediate in the synthesis o f semi-synthetic penicillins such as ampicillin (Scheme 1.2). The resolving agent is easily recovered and the unwanted enantiomer can be racemised and recycled thus making the process commercially viable.^ 10 11 6 Scheme 1.2 (± )-ll (+) -10 Isolation (+)-10 Racemisation 1.3.2 Kinetic resolution When enantiomers react with a chiral reagent, they will react at different rates because the transition states involved are no longer mirror images o f each other; they are diastereomeric rather than enantiomeric. In the case o f two enantiomers A and A interacting with a chiral reagent B, the transition states A B and A B will have different internal energies, consequently they will have different activation energies and hence each reaction will have different rates. Kinetic resolution o f enantiomers exploits this difference in reaction rates. 7 1.3.3 Kinetic resolution mediated by enzyme catalysis Today, kinetic resolution o f racemic substrates by enzyme catalysis has become a standard reaction in organic chemistry. More than 2000 enzymes are known, several hundred o f these are commercially available from biochemical supply houses such as Sigma and Fluka. When an enzyme is enantiomer specific, one enantiomer of a racemate is selectively transformed to product whereas the other is left unchanged, i.e. one enantiomer will fit into the active site o f the enzyme and its reaction will be catalysed. The other enantiomer will not bind as well to the active site and its diastereomeric transition state will be less stable, and so its reaction will be very much slower than the other enantiomer and its reaction will be effectively uncatalysed. The magnitude o f the energy difference, 5AG determines the enantiomeric excess. The obvious limitation with this process is that the maximum theoretical yield o f product is limited to 50% of the starting material (Scheme 1.3). The unwanted enantiomer must be discarded unless it is possible to recycle it by racemisation. Scheme 1.3 OH OH OAc OH Acylating agent Enzyme 50% Enzymes find particular use in effecting the kinetic resolution o f secondary alcohols. Secondary alcohols are an important class of readily available derivatizable compounds that can be incorporated into a variety o f synthetic strategies. The kinetic resolution of these compounds or more often their corresponding acetates has traditionally been achieved by esterases with excellent resu lts/’ ^ As mentioned earlier enzymatic resolution is limited to a maximum yield o f 50%. There are several ways to overcome this problem, (1) use o f meso or prochiral s u b s t r a t e s , (2) stereoinversion of the remaining enantiomer" (e.g. Mitsunobu reaction o f the remaining alcohol) (3) dynamic kinetic resolution or (DKR). 1.3.4 Dynamic kinetic resolution Both enzymes and transition metal catalysts have been used for the preparation o f enantiomerically enriched products. In dynamic kinetic resolution the substrate is continuously racemised during the resolution process and this leads to efficient use o f all the starting material. It is therefore possible to obtain in theory a 100% yield o f a desired enantiomer. In DKR three criteria must be met. Firstly, an enzyme that is capable o f effecting kinetic resolution must be identified. Secondly, a catalyst capable o f in situ racemisation o f the starting material and not the product is required, and thirdly, a system needs to be found where the performance o f the 12enzyme and catalyst are not adversely affected by the presence o f each other. This concept was elegantly employed by Allen et al}^ who utilised palladium catalysts in combination with various enzymes to effect the kinetic resolution o f allylic alcohols from allylic acetates via enzymatic hydrolysis (Scheme 1.4). An enzyme was utilised to hydrolyse one enantiomer of an allylic acetate to give the (5)-alcohol and concurrently a palladium catalyst was used to racemise the starting material but not the product, thus the (5)-enantiomer favoured by the enzyme was constantly replenished. Scheme 1.4 enzyme palladium catalysed racemisation enzyme The best results obtained were for the allylic acetates 12 and 14. After 19 days acetate 12 was hydrolysed to the corresponding (5)-alcohol 13 with 96% conversion in 96% enantiomeric excess using the enzyme Pseudomonas fluorescens lipase (PFL). Allylic acetate 14 gave the (5)-alcohol 15 with > 98% conversion and 50 % enantiomeric excess with the same enzyme. The rate determining step is most likely the slow palladium catalysed racemisation o f the allylic acetate, since in the former case 50% conversion is achieved after only 2 days. Ph ^OAc 12 Ph 13 10 Me O Me O 14 15 Backwell et al}'^ report a highly efficient DKR of secondary alcohols in the presence o f the ruthenium catalyst 16 using immobilised Candida antarctica lipase supported on an acrylic resin. This resin is commercially available under the tradename Novozym 435. Ph Ph H Ph Ph PhPh Ru. Ph H COCO 16 Thus, when using vinyl acetate as the acyl donor, racemic 1-phenylethyl alcohol 17 was converted into the (i?)-acetate 18 in the presence of a catalytic amount o f 16 in 50% yield and 99% ee (Scheme 1.5). The rest of the starting material was converted into acetophenone 20 (Table 1.1, entry 1). 11 When isopropenyl acetate was used as the acyl donor a similar result was obtained, 72% of (7?)-acetate and 28% acetophenone (Table 1.1, entry 2). The acetophenone formed arises from overall oxidation o f the starting material. When 4-chlorophenyl acetate was utilised as the acyl donor, no acetophenone was formed and racemic 17 was converted into (i?)-acetate 18 in 100% yield and >99% ee. This can be elegantly explained by considering the mechanism involved. Scheme 1.5 2 mol% 62 Q H Novozym 435 i ROAc ^ 'B uO H ,70"C P h ^ ^ M e (±)-17 (/f)-18 (5)-19 20 Table 1.1 Effect of the acyl donor in the dynamic kinetic resolution of (±)- phenylethyl alcohol catalysed by compound 16 as taken from Backwell et a 0 ‘* Entry“ ROAc (eq.) Time (h) (/?)-18'’ (5)-19'’ 20*’ % ee of (/?)-18 1 Vinyl acetate 17 50 0 50 >99" (5.5) 2 Isopropenyl 24 72 0 28 >99" acetate (5) 3 4-Chlorophenyl 87 100 0 0 >99" acetate (3) ̂ b ^ . Itt NMR and GC. ee of (R) -18 was determined by chiral HPLC. 12 The reaction proceeds through a base-mediated hydrogen abstraction from the substrate to give the ruthenium alkoxy species 21. Abstraction of the a-proton gives the corresponding ketone (acetophenone) and a ruthenium hydride complex. Subsequent readdition of hydrogen to the acetophenone gives the racemised 1- phenylethyl alcohol, thus replenishing the enantiomer preferred by the enzyme. The formation o f the acetophenone in Table 1.1 can be explained by this mechanism. Scheme 1.6 OH Ph Me (5)-17 [Ru] [Ru] 21 O Ph" H + IRu] H -► R a c - \ 1 When vinyl or isopropenyl acetates are utilised as acyl donors, transfer o f the acyl group to the substrate gives the vinyl alcohols 22 and 23, which can readily tautomerise to give acetaldehyde and acetone. These carbonyl compounds can both compete with the acetophenone for the intermediate ruthenium hydride complex. This results in consumption of the hydride complex and overall oxidation of the starting material. OH OH OH 22 23 Cl 24 13 This explains why no acetophenone is produced when 4-chlorophenyl acetate is used as an acylating agent. The 4-chlorophenol 24 produced when the acyl group o f 4- chlorophenyl acetate is transferred to the substrate cannot undergo tautomerism and hence cannot accept a hydride from the ruthenium donor. In the event, 1 -phenylethyl alcohol was converted into its acetate in 100% yield and > 99% ee. The scope o f this reaction was been extended to a range o f aromatic and aliphatic secondary alcohols. In all cases impressive yields and levels o f selectivities were obtained. Backvell et al}^ have also applied this methodology to the resolution o f secondary diols. The meso/dl diol 25 was converted into diacetate 26 in 63% yield as a 86 : 14 mixture o f {R,R)-26 (>99% ee) and meso 26 . This was quite impressive considering that the maximum yield in enzymatic kinetic resolution was circa 25%. O H OAc OAc OH OAc OAc 25 {R^R) - 26 Meso - 26 Reetz et al}^ have successfully used DKR to resolve the enantiomers o f phenylethylamine with immobilised lipase and ethyl acetate as the acyl donor. The non-acylated (5)- enantiomer o f the amine was racemised in situ by palladium on charcoal. Thus, (i?)-jV-acyl-1-phenylethyl amine was isolated in 64% yield and 99% ee. 14 Latest developments in DKR involve trying to combine two enzymes working in the same reaction vessel but independently o f each other. One enzyme for racemisation of the substrate and the other effects the kinetic resolution of the substrate.'^ Stecher 18et al. report that the enzyme mandelate racemase FC.51.2.2 from Pseudomonus putida strain ACTT 12633 catalyses the racemisation o f mandelic acid. The substrate spectrum for the enzyme was found to be very much broader than mandelic acid and was tolerant o f a variety o f functional groups on the aryl ring. This approach has the advantage of using milder conditions, which would fit synergistically with the conditions required for the enzyme catalysing the kinetic resolution. However, to date there are no reports of an effective two enzyme system capable o f a DKR with good yields and high ees. In practice, DKR has become an effective tool in the chemists toolbox to access optically active alcohols, amines and diols. 1.3.5 Non-enzymatic kinetic resolution The non-enzymatic resolution of secondary alcohols has proved to be more difficult than that catalysed by enzymes alone. The only prominent example is the Sharpless resolution o f allylic alcohols.'^ The most straightforward method for the resolution of alcohols is to use a chiral acylating agent such as oxazolidinone 27 (Scheme 1.7). Thus alcohols 28 and 29 were converted to their phenyl esters 30 and 31 with good ee. 15 Scheme 1.7 O Ph O OH CH2 CI2 , 0 ”C, > 90%. Ph O' MeMgBr R 27(1 eq.) (2 eq.) 28 Ri = Ph, R2 = Me 29 Ri = Ph, Rz = /Pr 30 ee = 95% 31 ee = 65% The main drawback with this system is that a stoichiometric amount o f the chiral acylating agent must be employed, i.e. there is no catalyst turnover. The activities o f certain chiral tertiary amines have also been investigated. 1- Phenylethyl alcohol upon treatment with the chiral amine 32 and acetyl chloride as the acyl donor gave the corresponding acetate o f (S)-l-phenylethyl alcohol 19 in 68% optical purity. The ratio o f alcohol to AcCl to catalyst was 2 : 1 : 1 . ^ ’ NMe2 Ph 32 (S)-15 O Cl 33 34 16 22The first example o f a chiral nucleophilic promoter was described by Vedejs et al. The Ci-symmetric phosphine 33 and acetic anhydride gave the acetate of phenylethyl alcohol 19 in 44% conversion and 34% ee. The ratio o f alcohol : acyl donor : catalyst was 1 : 2.5 : 0.16. Racemic 2,2-dimethyl-1-phenylpropanol was acylated with m-chlorobenzoic anhydride under the same reaction conditions to yield ester 34 in 25% conversion, 81% ee, (s = 12-15). * Noyori et a lP have demonstrated the ability of chiral diamine based Ru(II) complexes as catalysts for the reduction of prochiral ketones to the corresponding alcohols with excellent ee. The hydrogen source in this reaction is isopropyl alcohol which is oxidised to acetone. A major flaw o f this reaction is that it is reversible, i.e. the reaction is dependent on the reduction potential o f the alcohols formed. Alcohols such as 1,2,3,4-tetrahydro-l-naphthol and 1-phenylethyl alcohol cannot be prepared directly by this method since they have high reduction potentials and are readily oxidised back to the ketone again. Bearing this in mind and with the correct choice of catalyst it was proposed that this would provide a simple method for the kinetic resolution of these types o f alcohols.^"* Thus, excellent differentiation of enantiomers o f racemic alcohols was achieved with the Ru(II) diamine catalysts 35 and 36. The reaction can be better understood by considering Scheme 1.8. The faster-reacting enantiomer is converted to the ketone while the other remains effectively unchanged, thus providing a mixture consisting of easily separable ketone and alcohol of high optical purity. * Stereoselectivity factor s = (rate of fast reacting enantiomer)/(rate of slow reacting enantiomer) 17 Ts N \ Ru— ^Arene N H (5.5)- 35, Arene = /;-Cymene (5.5)- 36, Arene = Mesitylene Scheme 1.8 1 A number o f alcohols 37 - 47 were resolved by this methodology in good to excellent ee (Table 1.2). In all cases it was the (5)-enantiomer that was converted to the ketone when 35 and 36 were employed as catalysts for the reaction. 18 OH OH OH RO. RO 37 R = H 38 R = MeO 39 R = (CH 3)2N 40 R = CH 3 41 R-R = CH 2 42 R = CHz 43 R = (CH 2 )2 OH Fe OH OH 44 45 R = H 46 R = CH 3 47 An attractive aspect of this methodology is that inexpensive acetone and isopropyl alcohol can be utilised as the hydrogen acceptor and donor respectively. The ketone formed from the substrate can easily be recycled/transformed into desirable products. 19 Table 1.2 R esolution o f alcohols catalysed by chiral diam ine R u(II) com plexes 35 and 36 as taken from N oyori et a l P Alcohol Catalyst Mol % of Catalyst Time h ®/o Recovery of alcohol % ee Configuration S 37 (5,5)-35 0.2 36 50 92 R >80 37 (5’,5)-36 0.2 30 51 94 R >100 38 (5,5)-35 0.2 22 47 92 R >30 39 (5',5)-36 0.2 30 44 98 R >30 40 {S,S)-36 0.2 36 47 97 R > 5 0 41 (S,S)-36 0.2 24 47 97 R >50 42 (S,S)-35 0.2 6 46 97 R >40 43 (S,S)-35 0.2 6 49 99 R >50 44 {S,S)-36 0.075 36 51 98 R >100 45 (S,S)-35 0.2 4.5 43 93 R 14 Oriyama et al. have resolved a number o f secondary alcohols using the chiral diamine 48 derived from (5)-proline. In conjunction with a Lewis acid such as SnBr2, 4 A molecular sieves and an acyl halide as an acyl donor, 48 forms a SnBra-chiral diamine complex which is effective in promoting asymmetric acyl transfer. M e 48 20 Alcohols such as /ra«5-2-phenyl-l-cyclohexanol 50 gave benzoate 52 in 44% yield with excellent enantiomeric excess (97%) together with recovered (li?, 2i?)-50 when such a catalytic system was employed. Scheme 1,9 OU ± Ph SnBr2 , 48 (0.33 Eq.) ► CH2 CI2 , -78 °C 49 n = l 50 n = 2 OCOPh n '""Ph 51 n = 1, ee 86% 52 n = 2 ,ee97% Ph In an investigation with several other alcohols, cyclic substrates generally gave better results than acyclic alcohols. A limitation with this technology was that a high catalyst loading was required (0.33 molar equivalents relative to the alcohol). The above examples are representative o f catalyst systems with low or no turnover. Current thinking into the non-enzymatic kinetic resolution o f alcohols involves the synthesis o f chiral compounds based around the 4-dialkylaminopyridine structure. The last three years has witnessed an explosion o f interest in this area. 4- Dialkylaminopyridines are important catalysts for the acylation o f alcohols. 21 1.4 4-DiaIkylaminopyridines The reaction of acetic anhydride/pyridine with hydroxy compounds is a mild, dependable and general method for the preparation o f the corresponding acetates. This was first developed by Verley and Bolsing and was successfully extended by Fischer and Bergmann^^ in their work on carbohydrate chemistry. Steglich and Ho fie found that the addition o f 4-dimethylaminopyridine 54 greatly facilitated the acylation of hindered alcohols with carboxylic acid anhydrides. The most useful aspect of this reagent in acylations is that its action is catalytic. The presence o f less than 2 mol% of this compound enhances the rate o f acylation o f a primary or secondary alcohol by a factor o f circa 10''. Thus for example 1- methylcyclohexanol 53 was not acylated by acetic anhydride and pyridine or triethylamine, but the addition of 0.05 mol equivalents o f 4-dimethylaminopyridine 54 to a mixture o f 1-methylcyclohexanol and acetic anhydride led to the formation of 1-methylcyclohexyl acetate 55 in 86% yield.^^ OH NMe, + AC2O + 'N OAc 53 54 55 22 The superiority o f DMAP to other bases is illustrated in Table 1.3. The relative rates o f benzoylation of w-chloroaniline and benzyl alcohol and the effects o f various bases were investigated.^*^ Table 1.3 Effect of various bases on the relative rates of benzoylation of m- chloroaniline and benzyl alcohol as taken from Scriven et Relative Rate Catalyst pK, m-Chloroaniline Benzyl alcohol 3-Pyridinecarbonitrile 1.39 14 1 2 Quinoline 4.87 138 545 Pyridine 5.23 568 9.29 X 10^ Isoquinoline 5.40 2.62 X 1 0 ^ 3.39 X 10^ 2-Methylpyridine 5.96 29 435 3-Methylpyridine 5.63 1 . 1 2 X 1 0 ^ 2.29 X 10^ 4-Methylpyridine 6.04 2.96 X 10^ 3.98 X 10^ 4-Phenoxypyridine 6.25 4.80 X 10^ 7.98 X 10^ 2,6-Dimethylpyridine 6.72 8 115 DMAP 9.70 3.14 X 10^ 3.45 X 10* TEA 10.65 2 1 - Clearly the use of DMAP has a profound influence on reaction rate o f benzoylation of benzyl alcohol and for m-chloroaniline. Why is this so? 23 1.4.1 Mechanism of catalysis To understand the catalytic activity of DMAP one must understand the mechanism of catalysis. On comparing the basicity of pyridine (jpK̂ = 5.23) with that of DMAP (pa's = 9.70) it could be surmised that the increase in reactivity of DMAP is due to the 2.6 x 10“̂ fold increase in basicity. That this is not the case shown by triethylamine (p^a = 10.65) which is about as active as pyridine, therefore catalysis is not truly reflected by consideration of the p a alone. The only possible reason for the increase in activity is increased nucleophilic catalysis. For example, the hydrolysis of acetic anhydride in the presence of pyridine has been shown to proceed by nucleophilic catalysis and the unstable acetylpyridinium ion 56 was proposed as an 31intermediate. The mechanism was formulated on the basis of kinetic analysis. AC2O H ,0 N N Ac OAc HOAc 'N H 56 The intermediate salt is most readily formed in the case of the more basic pyridine, DMAP. Unlike the case of pyridine, such salts of DMAP may be isolated and are often quite stable. Another factor is n-n overlap that is possible in the 4- dialkylaminopyridines series to give the canonical structures 57 and 58. Structures such as 57 greatly increase the nucleophilicity of the ring nitrogen. 24 NMe, Me, NMe, 'N' N N 57 58 Hassner et al. recorded the chemical shifts o f the P-hydrogens (of the pyridine ring) in a whole series o f substituted 4-dialkylaminopyridines. They found that the greatest shielding occurs in the most effective acylation catalyst, 4-(r-pyrrolidinyl)-pyridine (PPY). They concluded that the catalytic activity is due to a combination o f the donor ability of the amine substituent and the stability o f the intermediate A^-acyl-4- dialkylaminopyridinium species 59. Their results were consistent with the mechanism postulated below (Scheme 1.10). NR, AC2O 'N R = Me R = Et O + R'O CH. Scheme 1.10 NR, OAc r 'o h 'N O" CH3 59 NR, 'N o - R, 'N O V R‘ O h CH, NR OAc 25 They also found that the A^-acyl pyridinium intermediate 59 is sensitive to steric effects since no catalytic activity was observed in the presence o f a series o f 2- dialkylaminopyridines. In summary it can be concluded that; • As a result of their pronounced nucleophilicities bases such as DMAP and 4 -( l’- pyrrolidinyl)pyridine (PPY) form high concentrations o f A^-acyl-4- dialkylaminopyridinium salts with acylating agents even in non-polar solvents and are superior to pyridine and amines in this respect. • Because o f charge delocalisation, the A^-acyl-4-dialkylaminopyridinium salts are present as loosely bound ion pairs, thus greatly facilitating attack o f nucleophiles on the activated acyl group with general base catalysis by the neighbouring anion. This effect also explains why carboxylic anhydrides are better suited for these acylations than the corresponding acyl chlorides. 4-Dialkylaminopyridines were soon found to have general applicability in the catalysis o f acylations and related reactions including alkylations, halogenations, Criegee rearrangements, cyanylations, dehydrations, oxidations, phosphorylations, silylations and sulphonylations. This coupled with the commercial availability of DMAP in larger quantities for the first time stimulated great interest in its use as a catalyst in organic chemistry, polymers, analytical chemistry and biochemistry. 26 1.4.2 Chiral 4-dialkylaminopyridines The first chiral 4-dialkylaminopyridine was reported by Vedejs et who synthesised the chiral acyl transfer reagent 62. This compound was synthesised from commercially available 4-dimethylaminopyridine (Scheme 1.11). 4- Dimethylaminopyridine was reacted with BF3 to give the corresponding DMAP-BF3 adduct in order to increase the acidity o f the protons at the 2 - and 6 - positions, thus increasing the likelihood o f abstraction by a strong base such as LiTMP. Metalation at C2 with lithium tetramethylpiperidine followed by reaction with pivaloyl chloride gave ketone 60 in 61% yield. The ketone was reduced with (-)-5 - chlorodiisopinocampheylborane (ipc2BCl) to give the alcohol 61 in 71% yield and 94% ee. Recrystallisation gave material o f > 99% enantiomeric purity. Methylation with methyl iodide, potassium hydride and 18-crown-6 gave the catalyst 62. Scheme 1.11 NMez NMej N (ii) LiTMP (iii) f-BuOCOCl O 60 ipc2BCl This new compound was non-catalytic and was employed in stoichiometric amounts. Reaction o f 62 with a commercially available chloroformate generated as expected the corresponding A^-acyl-4-dimethylaminopyridinium salt 63. Addition of representative secondary alcohols to 63 did not lead to their acylation. However, the addition o f a tertiary amine together with a Lewis acid (ZnCh or MgBr2) initiated a slow acyl transfer reaction resulting in the formation of the mixed carbonate 64 (Scheme 1.12). 62 Scheme 1.12 o NMe. OMe CCl M e' Me OMe CCl Me Me 63 ROH MgBr2 (2 eq.) TEA (3 eq.) 28 The results of acylating a number o f secondary alcohols showed that several o f the mixed carbonate esters 64 were formed with > 90% enantiomeric purity at conversions in the range of 20-42%. Catalyst 62 was easily recovered and reused many times without any apparent loss of activity. An inherent problem with this method for the kinetic resolution o f alcohols is that the starting material becomes more and more enriched in the slower reacting enantiomer. Therefore, exceptionally large differences in rates between enantiomers are required to obtain high ee values in product as well as the enriched enantiomer o f the starting material, e.g. for conversions at 50%, s = 200, ee = 96%; s = 500, ee = 98%. In an attempt to overcome this problem an ingenious protocol known as parallel kinetic resolution (PKR) was developed by Vedejs and Chen.^^ The underlying principle is o f two competing processes run in parallel with similar rates and selectivities but for opposite enantiomers, thus the optimal 1:1 ratio is maintained throughout the experiment. The advantages of this experiment can be appreciated by considering a PKR experiment in which there are two simultaneous reactions with selectivities for opposite enantiomers of the same racemic alcohol and each with s = 49 (100 % conversion); ideally this system would give both products with ee o f 96%. By contrast a kinetic resolution would require s = 200 (50% conversion). The chiral DMAP derivatives 62 and 65 were chosen and were expected to behave as quasi-enantiomers and to acylate opposite enantiomers o f a racemic secondary alcohol. Thus, when 62 and 65 were treated with representative secondary alcohols 66 - 68 the corresponding carbonates 69 - 71 were formed in good yields and ee. The concept is outlined in Scheme 1.13. 29 Scheme 1.13 NMe OMe NMe -OCOCl CC13 NMe. Me C l' OMe Me CCl (0.56 eq.) OBn (+)-Fenchyl Chloroformate NMe fenchvKJ (0.56 eq.) 66 R = 1-naphthyl 67 R = 2-naphthyI 68 R = o-tolylOH OO OfenchylCCl (S) - 69 46%, ee = 88% (5) - 70 49%, ee = 86% (5)-71 46%,ee = 83% ( R ) (R) - 69 49%, ee = 95% (R) - 70 43%, ee = 93% ( if)-71 46% ,ee = 94% 30 Ruble e t have explored the possibility o f 7i-complexation o f a heterocycle to a transition metal as an effective approach to the development o f chiral analogues o f planar nucleophilic catalysts such as 4-dimethylaminopyridine 54 and imidazole. The resultant complex is chiral by virtue o f there being four different substituents on the nitrogen atom (Figure 1.2). Figure 1.2 Void H N R Ml„ Viewing along the lone pair-nitrogen atom axis, increased differentiation from left to right (H vs R) and from top to bottom (void vs Mlp) corresponds to a more asymmetric environment around the nucleophilic nitrogen. I f either left/right or top/bottom is not differentiated then the complex will be achiral. The compounds r , 2 ’,3’,4’,5’,-pentamethylazaferrocene 72, r , 2 ’,3’,4’,5’,-pentamethyl-2-trimethyl silyloxymethylaza-ferrocene 73, pentamethylcyclopentadienylpyrindinyliron 74 and 4-dimethylaminopyrindinyl -pentamethylcyclopentadienyliron 75 were synthesised, 73 - 75 being chiral. The FeCp fragment was chosen as the Min fragment because it was electron-rich, stable and sterically bulky. The pyrindinyl complexes 74 and 75 were chosen to explore the effect o f nucleophilicity on five- and six- membered t i - bound heterocycles that have different steric and electronic properties. The azaferrocene derivative 73 functioned as an effective acylation catalyst for the kinetic resolution o f chiral secondary alcohols. 31 Me 74 Me Me Me Me Me 73 FeFeMe Me Me Me Me SI2 R 75 R = Me 76 R = Ph Thus, compound 73 catalysed the acylation o f 1-phenylethyl alcohol in benzene employing diketene as the acyl donor to give the corresponding acetate at 58% conversion in 53% ee (s = 3.6). Reaction o f 2-naphthylethyl alcohol under the same conditions afforded (.S)-2-naphthylethyl acetate in 87% ee (s = 6.5) at 67% conversion. Catalyst 75 proved very effective in the acylation o f secondary alcohols but showed no enantioselectivity. Modification of 75 by increasing the steric bulk of the cyclopentadienyl fragment (R = Me vs R = Ph) proved highly successful, thus affording 76 which proved a highly enantioselective acylation catalyst.^^ This catalyst also had the advantage that inexpensive acetic anhydride could be utilised as the acyl donor as opposed to diketene. Thus, in the presence of ether and triethylamine at r.t. a range o f racemic alcohols were resolved with high enantioselectivities and good conversions (Table 1.4). 32 Table 1.4 Catalytic enantioselective acylation of racemic secondary alcohols catalysed by compound 76 as taken from Ruble et al. Entry ̂ Unreacted alcohol major enantiomer R % ee of unreacted alcohol (% Conversion) (Selectivity) 1 2 3 4 5 6 7 OH Ph R OH Me Me Et i-Pr /-Bu CH2CI F OMe OH Me OH Me OH 95.2 (62) 98.8 (62) 97.7 (55) 92.2 (51) 98.9 (69) 99.2 (64) 94.5 (60) 99.7 (63) 99.1 (67) 99.0 (61) 14 20 36 52 12 18 15 22 14 22 (a) Alcohol (3.37 mmol), acetic anhydride (2.53 mmol), ether (6 ml), TEA (2.53 mmol) None of the substrates in Table 1.4 had previously been resolved with a selectivity factor greater than 7 in the presence of a non-enzymatic chiral acylation catalyst. These resolutions were relatively straightforward and were not sensitive to oxygen or moisture. Identical results were obtained for acylations with unpurified reagents exposed to the atmosphere as opposed to reactions conducted in an inert atmosphere with analytical grade reagents. Finally the catalyst was recovered in >98% yield and reused without any loss of activity. A wide ranging solvent study into the acylation of (±)-l-phenylethyl alcohol revealed that both the rate and the enantioselectivity were highly dependent on the solvent.^* In the event, /erf-amyl alcohol was the solvent o f choice for acylations catalysed by 76 (Table 1.5). Interestingly, fer^-amyl alcohol was not acylated to any great extent under these conditions. When compared to identical reactions run in ether a tripling of selectivites was observed. This observation had important practical consequences. It had been determined that acylations catalysed by compound 76 conducted at 0 °C in ether led to higher selectivities. Unfortunately, these reactions took several days, whereas when tert- amyl alcohol was employed as the solvent acylations proceed at a convenient rate and were typically complete within 24 h. 34 Table 1.5 Solvent effect on rate and selectivity for the kinetic resolution of 1-phenylethyl alcohol catalysed by 76 from Ruble et al. Solvent % Conversion after 1 h Selectivity (s) DMF 6 3.4 CHjCN 10 3.6 CH2CI2 14 7.0 Acetone 8 8.7 THF 4 9.6 EtOAc 6 11 Toluene 13 11 EtjO 8 13 Namyl alcohol 36 27 The use o f tert-2imy\ alcohol as solvent led to general increases in selectivity for a wide range o f substrate alcohols. As a result o f such a high selectivity factor both alcohol and ester were now accessible in excellent ee. In general selectivities greater than 50, and ee values o f 99% are possible. This currently represents the best non- enzymatic asymmetric acylation o f arylalkylcarbinols in terms o f enantioselectivity and scope. There is a high catalyst turnover and the catalyst is easily recoverable. Also, the reaction conditions permit unpurified reactants to be used without the need for an inert atmosphere. -}Q Ruble et al. have also extended the use o f this methodology to the synthesis o f protected a-am ino acids from racemic oxazolidinones with moderately impressive results. Because o f their propensity to racemise at room temperature oxazolidinones such as 77 undergo DKR via ring opening by alcohols when subjected to catalysis by 76 to give the protected a-amino acids 78 (Scheme 1.14). Scheme 1.14 Me Ph (±)-77 Fast H-OR Fast Racemisation Me....... Ph Sl^ h -o r ' Me OR N. OH Ph 78 O Me H -N. OR O Ph Addition o f alcohols to oxazolidinones such as (±)-77 almost always gives the L- alanine derivatives 78. The level o f selectivity is found to be solvent dependant with toluene affording the best ee (Table 1.6). Stereoselectivity increases with size of alcohol used, e.g. use o f /-PrOH gave an ee of 78% under these conditions. Unfortunately ring opening is very slow (ti/2 ~ 1 week). 36 Table 1.6 Effect of solvent on enantioselectivity o f the ring opening of aziactones catalysed by compound 76 as taken from Ruble et al. 39 Entry Solvent % ee 1 CH3CN 11 2 PhNOs 17 3 Acetone 2 0 4 THF 31 5 CH2CI2 33 7 EtOAc 40 8 PhOMe 46 9 Toluene 49 Ruble et al.^^ also report that 0-acylated enol lactones such as 79 undergo an enantioselective rearrangement when subjected to catalysis by compound 76 to form the C-acylated isomer thereby generating a new quaternary stereocentre (Scheme 1.15). Scheme 1.15 BnO "Ar 79 2 mol% (-)-76 Namyl alcohol 0 °C, 2-6 h R = Me, Et, CHjPh, allyl, CHjCHMez Ar = 4-MeO-C6H4 BnO Xr 88-95% ee 93-95% yield 37 This reaction gave excellent yields and very good selectivity for an array o f O- acylated enol lactones. The utility of this rearrangement was illustrated by the subsequent conversion of the C-acylated oxazolidinone into the dipeptide 80 and into the protected a-methylserine derivative 81 (Scheme 1.16). An ingenious approach to the kinetic resolution o f secondary alcohols was devised assumed that the reduced steric hindrance in the proximity o f the nitrogen atom would result in high catalytic turnover and that chiral recognition would be possible by remote asymmetric induction, similar to the “induced fif’ mechanism known to operate in enzymes. This method where the stereocontrolling chiral centres are far from the active site overcomes the selectivity-reactivity dilemma, so apparent in the method o f Vedejs et a l} ‘̂ Scheme 1.16 Me BnO Ph OMe BnOjC NHCOAr o O Me 80 NaBH4 NHCOAr 81 by Fuji et who prepared the enantiomerically pure pyridine derivative 82. They 38 OH 82 'H NMR NOE experiments in CDCI3 showed that 82 adopts an “open conformation” in the ground state but, upon reaction with an acyl donor such as isobutyric anhydride, this changes to a “closed conformation” in which the naphthyl moiety sits over the pyridine ring o f the catalyst (Figure 1.3). In the “open conformation” protons Ha, Hb, He and Hd appear as doublets at 5 8.01 and 6.37 ppm respectively. This indicates that there is free rotation about the N (2)-C(l’) bond of the molecule and no significant interaction between the naphthalene ring and the pyridine ring. Upon reaction with the acyl donor, protons Ha, Hb, He and Hd all appear at different shifts at 5 7.45, 8.93, 5.69 and 6.87 ppm respectively. This dramatic change in chemical shifts indicates that there is n -n interaction between the aromatic rings. Also an NOE was observed between Hb and the proton o f the NCOC//(CH 3 ) 2 of the acyl group which suggested that the si face of the carbonyl moiety is blocked by the naphthalene ring. This in effect prevents approach of the alcohol from the si face of the carbonyl group thus leaving the re face exposed to attack. This is the enantiodifferentiating event for this catalyst. 39 Figure 1.3 H""'; Open Conformation OH HO-R si face Closed Conformation ROH When the racemic (±)-cz5-alcohols 83 - 86 were treated with isobutyric anhydride in the presence of a catalytic amount o f 82 (5 mol%), the acetates 87 - 90 were formed in good yields (Scheme 1.17). The optical purity of the products was determined by chiral HPLC on the recovered alcohol (Table 1.7). 40 Scheme 1.17 OCOR OH 83 R = /j-02NC6H4 87 R = /j-02NC6H4 84 R = Ph 88 R = Ph 85 R = /;-MeOC6H4 89 R =/»-MeOC6H4 86 R = /j-Me2NC6H4 90 R =/)-Me2NC6H4 Table 1.7 Resolution of alcohols catalysed by compound 82 as taken from Fuji et al. Entry Substrate % Conversion % ee of alcohols 83 -86 Selectivity 1 83 73 54 2.4 2 84 71 81 4.5 3 85 70 85 5.3 4 86 72 >99 >10.1 Interestingly, the optical purity o f the recovered alcohol is dependent on the electron- donating ability of the aromatic ring in the substrate, possibly indicating that n-n stacking may play a pivotal role in the enantiodifferentiating event (Table 1.7, entries 1 and 4). Recent newcomers to the area o f non-enzymatic kinetic resolution are Spivey et al. who have synthesised a number o f configurationally stable biaryl analogues of ^OCOR .nOCOR (;PrC0)20 (0.7 eq.) ^OH 5 mol% of 82 X)COiPr 41 4-dimethylaminopyridine. The compounds 91 and 92 are configurationally stable at ambient temperature. The chirahty o f these molecules arises from restricted rotation about the Ar-Py bond. l-Methyl-2-pyrrolidino[3,2-c]pyridine 93 was chosen as the backbone o f the catalyst because the C-4-N bond is conformationally rigid and therefore contributes to the barrier o f internal rotation about the central biaryl axis. Also, 93 showed catalytic activity on a par with that o f DMAP itself'*'' MeN- Me MeN Ph 91 92 MeN MeN Br 93 94 Catalysts 91 and 92 were synthesised from the bromo derivative 94 via Suzuki cross coupling with arylboronic acids. The bromo compound 94 was synthesised from commercially available 4-aminopyridine (Scheme 1.18). Catalysts 91 and 92 were 42 designed so that the stereogenic axis was meta to the pyridyl nitrogen such that they retained the high nucleophilicity of the parent amine. NH, 'N NHBoc N Scheme 1.18 NHBoc BocN 'N .OMs IV OBu N ^ O— Li 111 NHBoc OH 'N MeN-MeN Br VI Vll Reagents and Conditions: (i) BocjO, DCM, r.t, 45 min, >99 %; (ii) /-BuLi, THF, -78 °C, 3.5 h; (iii) ethylene oxide, -78 °C r.t, 2 h, 75%, (iv) MsCI, EtjN, DCM, - 10 °C ^ r.t 2 h, > 99%, (v) LHMDS, THF, -78 “C r.t ^ reflux, 95%, (vi) DIBAL, DCM, 0 “C ^ reflux, 20 h, 55%, (vii) NBS, DMF, 0 °C, 90 min, 81%. The racemisation of compounds 91 and 92 were sufficiently slow at ambient temperature to allow for the separation of their constituent enantiomers which were 43 resolved using semipreparative chiral HPLC. In the event catalysts, (+)-91 and (-)-92 showed low levels of enantioselectivity (s = 1.3-1.5 and 2.1 depending on solvent respectively) but retained excellent reactivity. The lack o f enantioselectivity was attributed to insufficient differentiation between the two faces o f the pyridine ring. This differentiation had been elegantly achieved by Fuji et with the catalyst 82. In order to overcome this shortcoming compound 95 was chosen by Spivey as the next lead candidate for testing as an enantioselective acylation catalyst. A more sterically demanding substitutent was anticipated to give much better differentiation between the top and bottom faces of the pyridine ring. MeN- MeMe Me Me(±)-95 Upon testing, compound (-)-95 did give a modest increase in enantioselectivity (s = 4.7) when l-(l-naphthyl) ethanol was utilised as the substrate. In order for this process to be practical selectivity factors o f > 7 are required. In order to differentiate the molecule further, a differentiation from left to right was attempted. So, the compound (-)-96 was synthesised. 44 NEt H -9 6 A number o f secondary alcohols 97 were resolved with good levels o f selectivity when (-)-96 was employed as the catalyst and isobutyric anhydride was used as the acyl donor (Table 1.8). OCOR (±)-97 (5)-98 (R)-99 Although the results obtained were perhaps not as impressive as those obtained by Ruble et (Table 1.4), they nonetheless were still impressive. Further differentiation o f the molecule is still possible by introducing another large biaryl subunit into the 3-position o f the pyridine ring which would give the desirable C2 symmetric compound which may aid further the enantiodiscriminating event. Current work that is ongoing involves further studies into the selectivity-reactivity optimisation. 45 Table 1.8 Kinetic resolution o f alcohols 97 catalysed by biaryl (-)-96 as taken from Spivey et Entry Ar r ' Solvent Time/h (5)-98 % ee (R)-99 % ee s 1 1 - naphthyl Me /Pr PhMe 9.0 49.9 78.1 13 2 1 - naphthyl Me i? T EtOAc 9.3 76.1 70.9 13 3 Ph Me i? T PhMe 7.6 49.9 78.1 13 4 Ph Et /Pr PhMe 9.7 49.9 78.1 13 5 Ph /Pr /Pr PhMe 10.1 29.8 72.7 8.4 6 Ph 'Bu /Pr PhMe 10.5 18.8 88.8 20 7 2-Me(C6H4) Me /Pr PhMe 9.5 60.7 86.0 25 8 2-OMe (C6H4) Me /Pr PhMe 12.1 40.2 81.5 15 9 2,6-di- Me(CeH4) Me /Pr PhMe 8.0 21.3 90.7 25 The use of active esters as acyl equivalents for the acylation o f alcohols has been well documented. Sammakia et have developed 2-form yl-4-(r- pyrrolidinyl)pyridine (FPP) 100 as a catalyst for the hydroxyl-directed methanolysis of hydroxy esters (Scheme 1.19). O 100 46 FPP 100 was designed to have separate binding (2-formyl group) and catalytic (nitrogen o f the pyridine ring) sites. The binding site serves two functions, it brings the ester into close proximity with the active site and activates the catalyst upon binding by converting a aldehyde group which is electron-withdrawing into a hemiacetal functionality which is electron-donating. In kinetic studies, the p - nitrophenyl (PNP) esters o f propionic acid 101, methoxyacetic acid 102 and glycolic acid 103 were chosen as substrates for methanolysis catalysed by FFP (Scheme 1.19). Scheme 1.19 101 102 103 R = Me R = OMe R =O H NO, O Catalyst 5 mol% ► MeOH R OMe From these kinetic studies hydroxy ester 103 underwent methanolysis some 96 times faster than the ester 102 and some 511 times faster than the PNP ester o f propionic acid 101. This difference in rate was interpreted as evidence o f hydroxyl binding o f the alcohol with the binding site as anticipated. However it was uncertain whether the mechanism was nucleophilic or proceeded by general base catalysis (Scheme 1.20). 47 Scheme 1.20 O N General Base Mechanis™ OH ArO 104 OH OAr100 Nucleophilic Mechanism + N MeOHOH100 OH MeO MeO FPP was designed to operate by the nucleophilic mechanism but this was quickly discredited. Use o f a range of 6-substituted derivatives o f FPP as catalysts served only to speed up the reaction and not hinder it as was observed when Sammakia et al. used a range of 2-alkyl-4-dialkylaminopyridnes to catalyse the acylation of alcohols. For example, the introduction of a TMS moiety into the 6-position o f FPP increased the reaction rate by a factor o f thirteen relative to FPP itself This could only mean that the nitrogen of the pyridine ring was more basic and hence a better proton acceptor. If the nucleophilic mechanism were operating, introduction o f steric bulk near the vicinity of the pyridine nitrogen would greatly hinder the reaction as is observed in the acylation of alcohols with similar type compounds. This is clearly not the case. Further proof was obtained when the hemiacetal intermediate of the general base mechanism 104 was isolated and characterised by NMR spectroscopy. Current work by Sammakia et al. involves attempts to synthesise chiral variants of FPP, which may be capable o f kinetic resolution via enantioselective ester hydrolysis. In a model study in this laboratory Storey"* ̂ synthesised the suite o f four chiral pyridyl alcohols 105 - 108 and all four o f these compounds fulfilled the primary objective o f their synthesis in that they catalysed the enantioselective addition of diethylzinc to benzaldehyde. The derived methyl ether 109 did not act as an acylation catalyst and this result was interpreted as evidence of interaction o f the ether oxygen electron lone pairs with the strongly electron-deficient carbonyl carbon of the intermediate A^-acylpyridinium salt. This in turn deactivates the latter towards nucleophilic attack by the substrate alcohol and so acyl transfer does not take place. Vedejs et also experienced this problem with their chiral 2-subtituted DMAP derivative 62, hence the requirement o f Lewis acids to promote the acyl transfer. 49 H,C OH 105 3 R , l \ R 109 106 3 S , \ \ S 107 3 R , \ \ S 108 35, 11/? In an effort to circumvent this problem o f oxygen lone pair interaction with the N- 48acylpyridinium salt, the fluoro derivative 110 was synthesised by Aubert but this compound also failed to function as an acylation catalyst even under stoichiometric conditions. NMC2 F 110 1.4.3 Chiral 4-dialkyaminopyridine A^-oxides Gallagher"^ ̂discovered that 4-dimethylaminopyridine A^-oxide is an effective catalyst for the acylation o f 1-phenylethyl alcohol with acetic anhydride. This remarkable finding that 4-dimethylaminopyridine N -o \id e is an acylation catalyst opens up a 50 whole new area o f research. The following suite of molecules 111 - 115 were synthesised by Aubert and fully characterised. NMe o- OH NMe- R OH o- OH 111 R = ‘Bu 113 R = 'Bu 112 R = Ph 114 R = Ph NMe OH O - HO 115 Removal of the interactive pendant oxygen atom at the a-carbon of these molecules should allow them to function as hypemucleophilic acylation catalysts. Any steric effects due to bulky substituents at C-2 and/or C-6 should be minimised by the fact that the transferable acyl function should be one atom away from the nitrogen of the pyridine ring. Scheme 1.21 indicates how such an A^-oxide 116 will react in a truly catalytic manner transferring its acyl group via an 0-acyl intermediate 117 to a preferred enantiomer of an racemic alcohol leading to a product mixture containing easily separated optically active alcohol and ester (Scheme 1.21). Scheme 1.21 NMe, AC2O *R + N o - 116 R* H,C- R* = Chiral substituent Me, N R* O O 117 (±)-R‘OH 116 + (+)-R'o CCH3 + ( - ) -r 'o H 1.5 Detailed Project Description The initial aims of this project was to synthesis chiral substituted 4- dimethylaminopyridine like compounds that do not have the handicap o f oxygen- containing functional groups protruding so close to the catalytic site. As discussed earlier, compounds such as 109, 111, 112, 113, 114 and 115 carmot function as acylation catalysts due to the oxygen electron lone pair interaction with the intermediate A^-acyl-4-dialkylaminopyridinium species in the case o f 109 and 52 pendent hydroxyl groups in the case of the other compounds 111 - 115. In order to better facilitate the introduction of suitable chiral groups into the 2- and 6-positions o f 4-dimethylaminopyridine, suitable 2- and 2,6-disubstituted 4- dialkylaminopyridines will be required. It was anticipated that the most suitable type o f compounds would be 2-halo-4-dimethylaminopyridines 118 and 2,6-dihalo-4- dimethylaminopyridines 119. Synthesis of these compounds should allow the successful introduction of chiral alkoxides and amines via nucleophilic displacement of the halogens to give compounds o f type 120,121 and 122, 123. NMC2 NMC2 118 X = Cl, Br, I 119 X = Cl, Br, I NMC2 NMC2 N X X N X R* = Chiral Substitutent 120 X = 0 121 X = N 122 X = 0 123 X = N 53 These compounds which are devoid o f pendant groups which might interact with the active site o f the molecule should function as effective acylation catalysts. Any loss o f basicity at the pyridine nitrogen should be offset by the powerfully electron- donating dimethylamino moiety in the 4-position o f the aromatic ring. It was also intended to subject compounds such as 118 to Heck and Stille crosscoupling reactions with certain alkenes embodying a high degree o f chiral architecture to give compounds o f type 124 and 125 . This would have the effect o f introducing into the ring o f 4-dimethylaminopyridine, chiral substituents free from heteroatoms, thus increasing the basicity o f the ring nitrogen atom and hence the reactivities o f these chiral compounds. NMej NMc2 R* = Chiral Substitutent 125 It was also proposed to investigate whether 2-halo-4-dialkylaminopyridines would undergo Ni-catalysed^'^ coupling with chiral Grignard reagents derived from chiral pool compounds. As mentioned earlier, Gallagher had discovered that 4-dimethylaminopyridine N - oxide was an effective acylation catalyst."*^ It had not been determined if catalytic 54 activity was confined only to 4-dialkylaminopyridine A^-oxides or was characteristic of all A^-oxides. It was proposed to examine a whole series o f A^-oxides as catalysts for the acylation of alcohols and to examine a series o f chiral 7V-oxides as enantioselective acylation catalysts. An important aspect o f this methodology is that, unlike their free bases, A^-oxides are only weakly basic and as such do not require an auxiliary base such as triethylamine to achieve a large number o f catalytic cycles per molecule of A^-oxide. Industrially, for reactions in which catalysts such as DMAP are used, the auxiliary base is sometimes the solvent and is present in considerable excess. Use of DMAP 7V-oxide as a catalyst in conjunction with a cheaper solvent may not require an auxiliary base and thus would result in considerable cost reductions. It was also intended to undertake a detailed kinetic study into the acylation o f 1- phenylethyl alcohol by acetic anhydride catalysed by all compounds synthesised that show catalytic activity. An important reason for doing this was to see how effective each catalyst would be when compared to some reference catalyst such as pyridine. Thus, each new catalyst will have a relative value associated with it which will be indicative o f its effectiveness. 55 1.6 References I. T. Eriksson, S. Bjorkman, B. Roth, A. Fyge and P.Hoglund, Chirality, 1995, 7, 44. 2 . M. L. Smyth, Chem. Eng. News, 1992, 70, 5. 3 . J. C. Gardner and R. L. DiCicco, Spec. Chem., 1994, S9-S12. 4 . S. C. Stinson, Chem. Eng. News, 1997, 75, 38. 5 . P.H. Boyle, Quarterly Reviews, 1971, 323. 6 . E. L. Eliel and S. H. Wilen, Stereochemistry o f Organic Compounds, Wiley, 1994,388. 7 . C. J. Sih and S. H. Wu, Top. Stereochem., 1989, 19, 63. 8 C. H. Wong and G. M. Whitesides, Enzymes in Synthetic Organic Chemistry, Pergamon, New York, 1994, Chapter 4. 9. A. Klibanov, Acc. Chem. Res., 1990, 23, 114. 10. E. Schoffers, A. Golebiowski and A. Johnson, Tetrahedron, 1996, 52, 3769. II. H. Stecher and K. Faber, Synthesis, 1997, 1. 12. R. Stiirmer, Angew. Chem., Int. Ed. Engl.,1997, 36, 1173. 13. J. V. Allen and J. M. J. WiUiams, Tetrahedron Lett., 1996, 37, 1859. 14. B. A. Persson, A. L. E. Larsson, M. Le Ray and J. E. Backvall, J. Am. Chem. Soc., 1999,121, 1645. 15. B. A. Persson, F. Huerta and J. E. Backvell, J. Org. Chem., 1999, 64, 5237. 16. M. T. Reetz and K. Schimossek, Chimia, 1996, 50, 668. 17. M. T. E. Gihani and J. M. J. Williams, Current Opinion in Molecular Biology, 1999, 3, 11. 18. H. Stecher, U Flelfer and K. Faber, J. Biotechnol., 1997, 56, 33. 56 19. D. Seebach and E. Hungerbiihler, Modern Synthetic Methods, Salle and Sauelander, Berlin, 2, 91. 20. D. A. Evans, J. C. Anderson and M. K. Taylor, Tetrahedron Lett.,1993, 34, 5563. 21. P.J. Weidert, E. Geyer and L. Homer, Liebigs Ann. Chem., 1989, 6 533. 22. E. Vedejs, O. Daugulis and S. T. Diver, J. Org. Chem., 1996, 61 , 430. 23. (a) S. Hashiguichi, A. Fujii, J. Takehara, T. Ikariya and R. Noyori, J. Am. Chem. Soc., 1995, 117, 7562; (b) S. Hashiguichi, A. Fuji, J. Takehara, S. Inoue, T. Ikariya and R. Noyori, Chem. Commun., 1996, 233; (c) J. X. Gao, T. Ikariya, R. Noyori, Organometallics, 1996, 15, 1087; (d) K. J. Haack, S. Hashiguichi, A. Fujii, T. Ikariya and R. Noyori, Angew. Chem., Int. Ed. Engl.,\991, 36 , 285; (e) R. L. Chowdhury and J. E. Backwell, J. Chem. Sac., Chem. Comm., 1991, 1063; (f) P. Krasik and H. Alper, Tetrahedron, 1994, 50, 4347. 24. S. Hashiguchi, A Fujii, K. J. Hack, K Matsumura, T. Ikariya and R. Noyori, Angew. Chem. Int. Ed. Engl, 1997, 36, 288. 25. T Oriyama, Y. Hori, K. Imai and R. Sasaki, Tetrahedron Lett., 1996, 37 , 8543. 26. A. Verley and F. Bolsing, Ber., 1901, 34, 3354, 1901; 34 , 3359. 27 E. Fischer and M. Bergmann, Ber., 1917, 50, 1047. 28 W. Steglich and G. Hofle, Angew. Chem. Int. Ed. Engl., 1969, 8, 981; Synthesis, 1972, 619, W. Steglich and G. Ho fie, Angew. Tetrahedron Lett., 1970, 54, 4727. 29. G. Hofle, W. Steghch and H. Vorbriiggen, Angew. Chem., 1978, 90, 569. 57 30. E. F. Scriven, Chem. Soc. Rev., 1983, 12, 129. 31. R. Butler and V. Gold, J. Chem. Soc., 1961, 4362. 32. A. Hassner, L. R. Krepski and V. Alexanian, Tetrahedron, 1978, 34, 2069. 33. “DMAP” - A 24 page publication by Seal Sands Chemicals, A Cambrax Company. 34. E. Vedejs and X. Chen, J. Am. Chem. Soc., 1996,118, 1809. 35. E. Vedejs and X. Chen, J. Am. Chem. Soc., 1997, 119, 2584. 36. J. Ruble and G. C. Fu, J. Org. Chem., 1996, 61, 7230. 37. J. Ruble, J. Tweddell and G. C. Fu, J. Am. Chem. Soc., 1997,119, 1492. 38. J. Ruble, J. Tweddell and G. C. Fu, J. Org Chem., 1998, 63, 2794. 39. J. Ruble, J. Liang and G. C. Fu, J. Org. Chem., 1998, 63, 3154. 40 J. Ruble and G. C. Fu, J. Am. Chem. Soc., 1998, 120, 11532. 41. T. Kawabata, M. Nagoto, K. Takasu and K. Fuji, J. Am. Chem. Soc., 1997, 119,3169 42. A. C. Spivey, T. Fekner, S. E. Spey and H. Adams, J. Org. Chem., 1999, 64, 9430. 43 . A. C. Spivey, T. Fekner, S. E. Spey and S. Spey, J. Org. Chem., 2000, 65, 9430. 44. E. Vedejs and X. Chen, Abst. Pap. Am. Chem. Soc., 1995, 210, 174. 45. J. K. Whitesell, Chem. Rev., 1989, 89, 1581. 46. T. Sammakia and T. B. Hurley, J. Am. Chem. Soc., 1996, 118, 8967. 47. C. B. Storey, Ph.D. Thesis, University of Dublin, 1995. 48. M. Aubert, M.Sc. Thesis, University of Dublin, 1996. 49. E. T. Gallagher, Ph.D. Thesis, University of Dublin, 1997. 50. M. Kumada, K. Tama and K. Sumitani, Org. Synth., 1980, 58, 127; G. J. Quallich, D. E. Fox, R. C. Friedmann and C. William Murtiashaw, J. Org. Chem., 1992,57, 761. 59 Chapter 2 Synthesis of 2-halo and 2,6-dihalo-4-dialkylaminopyridines 2.1 Introduction This chapter describes the different avenues explored in the search for a successful synthesis of the title compounds. These compounds were required in order to facilitate the successful introduction of suitable chiral substituents into the 2- and 6- positions o f the pyridine ring of 4-dimethylaminopyridine. In the event these compounds were successfully synthesised by several different methods that will now be discussed in more detail. 2.2 Synthesis of 2-chloro- and 2-bromo-4-dimethylaminopyridine via diazotisation of 2-amino-4-dimethylaminopyridine This strategy utilised 2-chloropyridine A^-oxide 126 as the starting material (Scheme 2.1). This compound is commercially available from Olin Corporation. Due to its unstable nature it is supplied as a 21% w/v solution in water. Extraction o f the aqueous solution with DCM afforded 2-chloropyridine //-oxide 126 with excellent recovery. 2-Chloropyridine A^-oxide was easily handled as a solid and was found to be quite stable for prolonged periods at -20 °C. Storage at this temperature during several months revealed no decomposition as revealed by 'H NMR. 2-Chloropyridine A^-oxide 126 was nitrated according to the procedure o f Finger et al} to give 2-chloro-4-nitropyridine A^-oxide 127 in 85% yield. A small amount of the N- deoxygenated 2-chloro-4-nitropyridine (<5%) was also present. Recrystallisation from an ethanol : chloroform 70 : 30 mixture afforded 2-chloro-4- nitropyridine 7V-oxide 127 in excellent purity as determined by ’H NMR. The *H NMR spectrum of 127 revealed that H-3 and H-5 came into resonance at 5 8.37 and 60 6 8.04 ppm, respectively. The proton H-5 was split into a double doublet by ortho coupling with H-6 and meta coupling with H-3. This familiar splitting pattern is characteristic o f 2,4-disubstituted pyridines. The chemical shift for H-3 appears significantly downfield at 5 8.37 ppm as a result o f the strongly electron- withdrawing nature o f the nitro group and o f the chlorine ortho to it. Also, H-6 came into resonance at 5 8.41 ppm. The presence o f the nitro group in the 4-position of 2-chloro-4-nitropyridine A^-oxide 127 greatly facilitates nucleophilic displacement o f the chlorine at the 2-position by amines. Thus, reaction of 2-chloro-4-nitropyridine A^-oxide 127 with concentrated ammonia solution in either isopropyl alcohol or tert-hniyX alcohol gave 2-amino-4- nitropyridine A^-oxide 128 as expected in excellent yield. Introduction o f an amino functionality at the 2-position of the molecule shifts protons H-3 and H-5 upfield by 0.7 and 0.5 ppm to 5 7.64 and 6 7.50 ppm, respectively, when compared to 2-chloro- 4-nitropyridine A^-oxide 127. The chemical shift for H-6 was moved very slightly upfield by 0.1 ppm to 5 8.20 ppm. The amino protons o f 128 came into resonance as ! a broad singlet at 6 5.93 ppm. I 2 The nitrite anion is a good leaving group from sp hybridised carbon atoms and it is i I superior to a chloride ion by a factor of ca. 1000. The nitro group in 4-nitropyridine I I 7V-oxides is quite susceptible to displacement by suitable nucleophiles such as I ' alkoxides and halogens."^ Although the displacement o f the nitro group by amines is not well known, there are some literature precedents for this, although it is generally agreed that yields are poor.^ However, it was still decided to investigate the 61 displacement of the nitro group of 2-amino-4-nitropyridine A'^-oxide 128 by amines such as pyrroHdine. Scheme 2.1 ' N O - 126 NM e, 'N 131 VI C l 'NH, NO NM e, + N O - 130 NM e, 'N NH, IV NO, f N O - 128 Cl + N O - 129 132 X = C1 133 X = Br NH, O 'N H R eagen ts and C on d ition s: (i) Fum ing HNO3/H2SO4, 90 °C, 3 h, 85%; (ii) aqueous NH3 (d 0 .880), 90 °C, pressure tube, 24 h, >90% ; (iii) A cC l, CHCI3, reflux, 18 h, 50%; (iv) aqueous N H M e 2 , pressure tube, 100 °C , 24 h, 80%; (v) Fe/A cO H , 100 °C, 3 h, >95% ; (vi) conc. HCl, N aN O j, -15 °C ^ 80 °C, 2 h, 60% or HBr, Br2 , NaNOz, -15 °C ^ 80 °C, 2 h, 70%. 62 Reaction o f 128 with pyrroHdine in refluxing toluene, did give small amounts of 2- amino-4-(r-pyrrolidinyl)pyridine A^-oxide 134. However, isolation of this compound proved very difficult and was hampered by the fact that there appeared to be a considerable amount of non-aromatic material present in the reaction mixture. It was anticipated that the use o f 2-acetamido-4-nitropyridine A^-oxide 135 would facilitate easier displacement of the nitro group from the 4-position. The rationale behind this was that the displacement might be made easier by converting the electron-donating amino group into its amide, thus the lone pair o f electrons o f the nitrogen in the 2- position would interact with the carbonyl functionality and not the aromatic ring. Successful acylation o f the amino 7V-oxide 128 to give the acetamide 135 was easily achieved using acetic anhydride at room temperature with the aid o f a catalytic amount o f DMAP. o- o- 134 135 Reaction o f the acetamide 135 with pyrrolidine in refluxing toluene resulted only in transamidation to give back 2-amino-4-nitropyridine 7V-oxide 128 and a similar result as described above was obtained. It was decided to abandon this route to 2-amino-4- dialkylaminopyridine A^-oxides since the displacement o f the nitro group from 134 and 135 was neither facile nor straightforward, a result that was in agreement with literature observations.^ Ochiai et al.^ reported the facile displacement o f the nitro group o f 4-nitropyridine N- oxide by acetyl chloride to give the 4-chloro derivative in almost quantitative yield. Thus, in the present work reaction o f 2-amino-4-nitropyridine A^-oxide 128 with excess acetyl chloride in refluxing chloroform gave 2-acetamido-4-chloropyridine N- oxide 129 in moderate yield (50%). Purification was carried out by column chromatography using a DCM : MeOH 90 : 10 mobile phase. The reaction probably proceeds via a two step process, the first being a intramolecular transfer o f an acetyl group from the acylated 7^^-oxide o f 128 to the 2-amino substituent (Scheme 2.2), and the second being displacement o f the nitro group. Scheme 2.2 Cl NONO NHNH + O - C1 N -NOj' NHAc ^ 0 2 o OCOCH. N NHAc OCOCH3 Cl NO, 'N O NHAc AcCl NO, N NHAc OCOCH, H,0 129 64 The rate-determining step in this reaction is the displacement o f the nitro substituent since complete conversion o f 2-amino-4-nitropyridine A^-oxide 128 into its acetamide 134 is observed after only 2 hours. An interesting aspect o f the 'H NM R spectrum o f compound 129 is that H-3 is deshielded considerably and appears downfield as a doublet at 5 8.51 ppm, J 2.5 Hz, while proton H-5 com es into resonance at 6 7.00 ppm whilst H-6 appears as a doublet at 6 8.16 ppm. The N-H o f the acetamide appears downfield at 5 9.96 ppm, and undergoes deuterium exchange in the presence o f D 2O. The displacement o f chlorine from the 4-position o f 6-membered heterocyclic N- oxides is well known.^ Thus, reaction o f 2-acetamido-4-chloropyridine jV-oxide 129 with aqueous dimethylamine solution resulted in the formation o f 2-amino-4- dimethylaminopyridine A^-oxide 130 in excellent yield. Analysis o f the *H NMR spectrum o f 2-amino-4-dimethylaminopyridine A^-oxide 130 showed that the chemical shifts o f H-3 and H-5 had m oved considerably upfield to 5 5.87 and 6 6.00 ppm respectively. The proton H-6 came into resonance at 6 7.8 ppm and the amino protons appeared as a broad singlet at 6 5.55 ppm. The dimethyl amino m oiety came into resonance at 5 2.93 ppm as would be expected, (the chemical shift o f the dimethylamino protons in DM AP is 5 3.00 ppm). The use o f iron to reduce the A^-oxide functionality is a standard reaction in A^-oxide chemistry.* The A^-oxide function o f 2-amino-4-dimethylaminopyridine A^-oxide 130 was easily reduced by iron filings in acetic acid to give the free base 2-amino-4- dimethylaminopyridine 131 in quantitative yield. The 'H NM R spectrum o f 131 was not too dissimilar to that o f the parent A^-oxide 130 . Analysis o f the spectrum showed that proton H-6 had moved downfield by 0.1 ppm to 5 7.75 ppm upon loss o f the N - oxide function. Proton H-5 remained unchanged at 5 6.00 ppm while H-3 resonated at 5 5.67 ppm, a difference o f 0.2 ppm downfield than in the A^-oxide 130 . The amino protons resonated at 6 5.65 ppm compared to 5 4.25 ppm in the A^-oxide. The chemical shift o f the dimethylamino protons remained effectively unchanged at 6 2.93 ppm. With this key compound in hand the stage was set for the synthesis o f the required 2-halo-4-dimethylaminopyridines. Aminopyridines are readily diazotised and behave very much like normal aromatic amines. In general, 2- and 4-aminopyridines tend to be resistant to diazotisation in dilute mineral acids and form the corresponding hydroxy or halogeno derivatives when reaction does occur. The use o f concentrated acids usually avoids or at best minimises these side reactions.^ In the event, 2-amino-4-dimethylaminopyridine 131 underwent diazotization in the presence o f concentrated hydrochloric acid and sodium nitrite at -1 0 °C to yield the expected 2-chloro derivative 132 in 60% yield. This was the only product isolated from this reaction. Analysis o f the 'H NM R spectrum o f 132 (Figure 2.1) showed H- 3 and H-5 very close together at 6 6.49 and 5 6.42 ppm, respectively. H-3 appears further downfield than H-5 since it is adjacent to the electronegative chlorine atom. The chemical shift representing H-6 appears at 6 7.98 ppm and the dimethylamino protons resonate at 5 3.02 ppm. This novel compound was fully characterised by FT- IR, ’H and ’^C NM R (Figure 2.2), melting point and MS. 66 Figure 2.1 *H NMR spectrum of 2-chioro-4-dimethylaminopyridine 132 r o m o o m ^ o o o o o oo r s . . — i / ^ o o o m t o a ^ r s . o ^ o ^ m r v i r v j * — f— i s ! r ^ {O {O ID m S I I o> rs. at 00 LD hs. — ( £ Slr^^ Ct\ nl 8.0 (ppm) 6.60 6.50 6.40 (ppm) A 3.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 (ppm) 3.5 3.0 NMC2 2.5 2.0 1.5 (ppm ) 89 0 0 cn - 148.4434 - 155.6689 151.6794 148.4434 Figure 2.2 C NM R spectrum of 2-chloro-4-dim ethylam inopyridine 132 The bromo derivative, 2-bromo-4-dimethylaminopyridine 133 was also synthesised similarly from concentrated HBr in the presence of bromine and sodium nitrite. The *H NMR spectrum of 133 (Figure 2.3) is very similar to that of 2-chloro-4- dimethylaminopyridine 132 except that H-3 is shifted further downfield by 0.2 ppm. Analysis of the '^C NMR spectrum (Figure 2.4) o f 133 reveals that C-2 resonates some 10 ppm further upfield than in 2-chloro-4-dimethylaminopyridine 132 as expected because chlorine being more electronegative than bromine pulls electron density from C-2, reduces the shielding effect and shifts resonance to a higher value. This synthesis o f 2-halo-4-dialkylaminopyridines 132 and 133 while successful in achieving the desired outcome, was fundamentally flawed due to the low overall yields (15%) and the number of steps required. Since both 2-chloro-4-dimethylaminopyridine 132 and 2-bromo-4-dimethylaminopyridine 133 were potentially direct precursors to more elaborate compounds, a more desirable synthesis of either which would have fewer steps and a higher yield would be very desirable. This was achieved via the use o f Fort’s base. I ! 69 (ppm ) 0/. Integra 1 .OOOQ- f 1.0527- f CTi b ■D ^ 1 - b / 1.0000 -7.9391 7.2831 ■7.9542 ■7.9391 7.2831 r6 .6 5 2 1 ^ 6.6458 T -6 .4 5 5 2 V-6.4489 '-6.4401 '-6 .4338 1.0356 1.0527 -6.6521 -6.6458 6.4552 6.4489 6.4401 6.4338 6 .3 4 i2 -r -3.0017 Figure 2.3 H NM R spectrum of 2-brom o-4-dim ethylam inopyridine 133 1 \ L 1 5 5 .3598 148 .8008 142 .6282 o o o O H ■1 55 .3598 - 148 .8008 - 142 .6282 ■ 108 .8093 105 .7472 ^ 7 6 .8 8 3 8 76 .5 6 5 0 ^ 7 6 .2462 38 .7 4 7 0 Figure 2.4 ” C NM R spectrum of 2-brom o- 4-dim ethylam inopyridiue 133 2.3 Synthesis of 2-bromo-4-dimethylaminopyridine 133 using Fort’s base The reaction o f pyridines with reagents such as butyUithium and tetramethylethylenediamine (TMEDA) usually results in Chichibabin type reactions, thus limiting their use in the synthesis o f functionalised pyridines. This is due in some part to the high nucleophilicity o f the reagents. An important aspect to the design and development o f strong bases is how to increase basicity without a corresponding increase in nucleophilicity. Fort et al}^ have developed a complex superbase BuLi»LiO(CH 2 )2NM e2 abbreviated BuLi»LiDMAE, which is easily obtained by the reaction o f 2 equivalents o f butyllithium and 1 equivalent o f N,N- dimethylaminoethanol in dry hexane. This base can facilitate the functionalisation o f pyridines and, to lesser extent, quinolines via simple metallation. The base functions by forming an aggregate between the BuLi and the lithiated alcohol which can then form a complex with the pyridine ring to effect the metallation at the 2- and/or 6- position (Scheme 2.3). Scheme 2.3 2BuLi 72 Thus, using BuLi«LiDMAE as the hthiating agent and carbon tetrabromide as the halogenating agent Fort et al. ” synthesised 2-bromopyridine from pyridine in 85% yield. The use o f carbon tetrabromide as a halogen equivalent is well known and documented.'^ In a personal communication from Fort it was revealed that the same complex base would metallate more basic pyridines such as 4- dimethylaminopyridine. Thus, 4-dimethylaminopyridine 54 was successfully metallated by Fort and then reacted with dimethyl disulphide as an electrophile to give 4-dimethylamino-2-methylsulfanylpyridine 136 and 4-dimethylamino-2,6- dimethylsulfanyl pyridine 137 in 66% and 24% yields, respectively. This experiment was successfully repeated in our laboratory. NMe It was decided to investigate the application of this technology to the synthesis of 2- bromo-4-dimethylaminopyridine 133 based on the evidence that, (a) 2- bromopyridine had been successfully synthesised in excellent yield by this method and (b) the complex base could metallate the more basic pyridine 4- dimethylaminopyridine. Thus, treatment o f 4-dimethylaminopyridine 54 with this base in dry hexane in the presence o f carbon tetrabromide as the halogenating agent (Scheme 2.4) gave 2- bromo-4-dimethylaminopyridine 133 in 57% yield as the sole reaction product, identical in every respect to the material prepared earlier (Scheme 2.1). There was no evidence to suggest that any 2,6-dibromo-4-dimethylaminopyridine had been formed in this reaction. Scheme 2.4 N N Br 54 133 Reagents and Conditions: (i) BuLi/DM AE, C Br4, - 10 °C -> 0 °C, 3 h, 57%. This method affords 2-bromo-4-dimethylaminopyridine 133 from 4- dimethylaminopyridine 54 in workable yield in only one step. When compared to the strategy in Scheme 2.1, there is almost a 3 fold increase in yield and five fewer steps. While the method of Fort et a /." did allow access to 2-bromo-4- dimethylaminopyridine 133 the reaction had some major drawbacks. A large amount of an unknown brown solid was invariably present in the crude material after aqueous workup. This resulted in the crude mixture having to be chromatographed (sometimes twice) to obtain material o f high purity. Given that the crude material had to be chromatographed and considering that 2 equivalents o f butyllithium were required to generate the superbase the process may not have been so economical. 74 Also the reaction was not amenable to scale up to a level that was synthetically useful {i.e. > 1 g). A recent paper by Sammakia et al}^ describes the synthesis o f 2-brom o-4-(l’- pyrrolidinyl)pyridine 139 from 4-(-r-pyrrolidinyl)pyridine 138 (Scheme 2.5). 4 -( l’- pyrrolidinyl)pyridine 138 was complexed with the Lewis acid boron trifluroide diethyl etherate and the resulting adduct was metallated using butyllithium. The metallated species was reacted with elemental bromine to afford 2-bromo-4- pyrrolidinopyridine 139 in good yield. Scheme 2.5 138 139 Reagents and Conditions; (i) BF3 (O Et)2 , 0 °C, (ii) BuLi, -78 °C, (iii) Br2 , 58%. Repetition of this experiment with 4-dimethylaminopyridine gave 2-bromo-4- dimethylaminopyridine 133 in, typically, 70% yield. Unlike the method o f Fort et a /.” small amounts (5-8%) o f the highly desirable 2,6-dibromo-4- dimethylaminopyridine 140 were also obtained from this reaction. Chromatography easily separated the two compounds. The NMR spectrum of 140 was very straightforward. Since the molecule is symmetrical only two signals were observed in the 'H NMR spectrum, at 5 6.63 and 3.00 ppm. These two chemical shifts were 75 easily assigned to protons H-3 and H-5 and to the dimethylamino protons respectively. NMC2 140 Attempts to make 140 the major product from this reaction by using 2 equivalents of 13 •BuLi proved unsuccessful. This methodology developed by Sammakia et al. is by and far the method o f choice for synthesis of the monobromo derivative 133. This reaction avoids the need for excessive chromatography and overall uses less butyllithium when compared to the method of Fort et al}^ It has also proved very amenable to scale up (5 g o f reactant) without any apparent loss of yield, which cannot be said about the method o f Fort et a /." 2.4 Attempted synthesis of 2-chloro-4-diaIkyiaminopyridines via reaction of 4 -( l’-pyrrolidinyl)-2-pyridone with various phosphorus reagents. The starting point in this synthesis (Scheme 2.6) was again 2-chloropyridine jV-oxide 126. The chlorine in the 2-position o f this compound underwent facile displacement with methoxide ion to give 2-methoxypyridine A^-oxide 141 in quantitative yield. The NMR spectrum of 141 showed that protons H-3 and H-5 came into resonance at 5 6.82 and 6.86 ppm. The resonance for H-6 appeared at 5 8.13 ppm. The proton H-4 appears as a distinct double triplet at 6 7.21 ppm. The methoxy methyl appears as a singlet at 5 3.99 ppm. An inherent difficulty with this 76 com pound was its lack o f solubility in non-alcoholic solvents. 2-M ethoxypyridine N - oxide 141 was only partially soluble in chloroform . N (V) OMe Schem e 2.6 146 X = NM ei 147 X = I ’-pyrrolidinyl Cl N 145 (iv) OMe OMe R, = H, Ri = NO2 R, = NO2, R2 = H ( 111) NH, N 144 'OMe (V I ) N (vii) N ^ O H 148 N N 149 Cl R eagen ts and C on d ition s: (i) NaOMe, MeOH, re flu x , 95% ; (ii) fu m in g HNO3/H2SO4 75 °C , 3 h 4 7% ; (iii) Fe/AcOH, 100 °C , 2 h, 98% ; (iv) conc. HCI, NaNOj, -10 °C ^ 60 °C , 2 h , 80% ; (v) p y rro lid in e , A ^-m ethyl-2-pyrrolidinone, 100 °C , 80% ; (v i) B B r3, DCM, 0 °C , 9 0% , (v ii) POCI3/PCI5, re flux 4 h. 77 2-Methoxypyridine A^-oxide 141 was nitrated according to the procedure o f Den Hertog et al}^ to afford a crude mixture whose composition was a 9 : 1 ratio of 2- methoxy-4-nitropyridine A^-oxide 142 (47% yield overall) and 2-methoxy-5- nitropyridine A^-oxide 143 (5.2% yield overall), and were separable by chromatography. The formation of the 5-nitro isomer was surprising since Den Hertog et al}^ claimed that the 4-nitro isomer was the sole reaction product. The *H NMR spectrum of 142 showed 3 aromatic protons at 6 8.36 (H-6), 7.79 (H-3) and 7.74 (H-5) ppm. The methoxy methyl group appeared at 6 4.19 ppm. The 'H NMR spectrum of 2-methoxy-5-nitropyridine A^-oxide 143 showed some interesting features. The chemical shift representing H-6 was dramatically shifted downfield to 5 9.16 ppm and showed very fine meta coupling with H-4, J 2.5 Hz. Proton H-3 appeared as a doublet at 5 7.00 ppm and H-4 at 5 8.10 ppm as a double doublet while the methoxy methyl group resonated as a singlet at 5 4.21 ppm. Miller at al}^ reported a synthesis o f 2-methoxy-4-nitropyridine A^-oxide 142 which claimed an improvement in yield could be achieved by using a more elevated reaction temperature and a longer reaction time. The formation of 2-methoxy-5- nitropyridine A^-oxide 143 was observed by Miller et al}^ On repeating this work no apparent increase in yield was observed under these conditions. However, the appearance o f another two compounds, the deoxygenated 2-methoxy-4-nitropyridine 150 and 2-methoxy-5-nitropyridine 151 to which Miller et al. make no reference, was observed (Scheme 2.7). Scheme 2.7 OMe N OMe 144 152 Reagents and Conditions: (i) Fuming HNO3/H2SO4, 95 °C, 3 h; (ii) H2, Pd/C, EtOAc, (iii) Fe/AcOH. The formation of 151 can be rationalised in one o f two ways; (a) initial deoxygenation of 2-methoxypyridine A^-oxide 141 at the elevated temperature to give 2-methoxypyridine, whose methoxy group directs the nitronium ion to the 5-position o f 2-methoxypyridine, or (b) the increased reaction temperature results in larger 79 amounts of the A^-oxide 143 being formed which then undergoes deoxygenation to give 151. The formation o f 150 can only be explained by the deoxygenation of 2- methoxy-4-nitropyridine A^-oxide 142 after nitration has taken place. The formation of deoxygenated products in this reaction parallels the formation o f small amounts of 2-chloro-4-nitropyridine in the nitration o f 2-chloropyridine A^-oxide 126 (See page 60, Scheme 2.1). The formation o f 2-methoxy-5-nitropyridine and 2-methoxy-4-nitropyridine were not merely undesirable side-products but constituted a large percentage o f the recovered material. The combined mass o f compounds 143, 150 and 151 usually equalled that of the desired product 142. 2-Methoxy-5-nitropyridine 7V-oxide 143 was of no synthetic importance and was discarded after chromatography. Compounds 150 and 151 were not separable from each other by column chromatography. This reaction step represented a severe bottleneck within the overall strategy and limited the amount o f the desirable TV-oxide 142 available for further synthetic steps. Thus, the original nitration conditions of Den Hertog et «/.''* were more appropriate under these circumstances. I 2-Methoxy-4-nitropyridine A^-oxide 142 was easily reduced with iron in acetic acid to yield 4-amino-2-methoxy pyridine 144 in excellent yield (98%). Introduction of i j the amino group at the 4-position shifts protons H-3 and H-5 up field to 6 5.95 and 6 i 6.23 ppm respectively. The signal representing H-6 appears as a distinct doublet at 5 7.84 ppm while the amino protons came into resonance as a broad singlet at 5 4.10 ppm. The methoxy group appeared as a singlet at 5 4.90 ppm. 80 As mentioned earlier, 2-methoxy-4-nitropyridine 150 and 2-methoxy-5-nitropyridine 151 are not separable by column chromatography, so in order to increase the amount of 4-amino-2-methoxypyridine 144 available for subsequent chemistry it was decided to reduce the nitro groups of compounds 150 and 151, since 2-methoxy-4- nitropyridine 150 is a direct precursor to 4-amino-2-methoxypyridine 144. Reduction of these compounds was achieved by hydrogenation over palladium on charcoal in ethyl acetate at atmospheric pressure to give a mixture o f 4-amino-2-methoxy- pyridine 144 and 2-methoxy-5-aminopyridine 152, which were separable by chromatography (Scheme 2.7). The 4-amino-2-methoxypyridine 144 prepared in this way was identical in every respect with the 4-amino-2-methoxypyridine prepared by the reaction o f 2-methoxy- 4-nitropyridine A^-oxide 142 with iron and acetic acid. 2-Methoxy-4-aminopyridine 144 underwent diazotisation with concentrated hydrochloric acid and sodium nitrite at -10 °C, to yield 4-chloro-2-methoxypyridine 145. This compound was a low-melting point solid and was either a liquid or a colourless solid depending on the temperature in the laboratory. It had a melting point o f 24 °C, which was in agreement with literature observations.'^ The chlorine in the 4-position of 145 was quite resistant to displacement with amines. Initial attempts to displace it with an aqueous solution o f dimethylamine at elevated temperature (ca. 130 °C) proved to be not very fhiitful. A dimethylamino 1 7moiety was successfully mtroduced mto 145 usmg the method o f Cho et al. who reported a very efficient dimethylamination of activated aromatic halides using DMF and for example, diethanolamine to give the corresponding dimethylamino derivative in excellent yield. The reaction was very successful with 2-chloropyridine as the aryl halide. A proposed mechanism is outlined below (Scheme 2.8). An hydroxyl group activated by the neighbouring amino function at