Attendance at this event may have been lower than previous years, presumably due to the current economic conditions, but the standards of the presentations were as high as ever.  There were 16 different speakers, from pharmaceutical, fine chemical and related companies in the USA, UK, France, Germany, and Hungary.  A number of companies took the opportunity to have stands in an adjoining exhibition room. As part of the excellent process chemistry presentations there was a special session in memory of Chris Schmid of Eli Lilly, who was also the Associate Editor of the Organic Process and Development Journal.  A Memorial Reception hosted by the American Chemical Society followed the Chris Schmid memorial Session.

Challenges of Design and Development: A Viable Route to AZD6244

Dr John Leonard, AstraZeneca Pharmaceuticals, UK

The conference was opened by John Leonard who presented work on a joint development project carried out by AstraZeneca and Array Biopharma.  The original synthetic route was too long and had some safety and selectivity issues.  The first challenge was to evaluate alternative chemistry and select 1 out of 12 possible alternative routes in a 9 month time period.  The starting point was influenced by the availability of 14kg of a key intermediate, methyl 4-amino-2,3-difluoro-5-nitrobenzoate, from earlier work.  Reaction of this intermediate with ammonia gave the desired 2,4-diamino-ester arising via SnAr, contrary to reports which said the amide would be the major product.  This result meant the difluoronitrobenzoate became a key intermediate in the synthesis.  Ultimately the original 13 step route was reduced to 5 high yielding steps (see Scheme 1) with 4 isolated intermediates and with 3 hazardous stages removed.

Scheme 1: Synthesis of AZD6244

Hypervalent Iodine Reagents for Oxidations at Industrial Scale

Dr Alain Chénedé, Simafex, France

The main topic of this presentation was safe formulations of iodoxybenzoic acid (IBX) which in >60-70% purity is heat sensitive, shock, sensitive and friction sensitive.  However when IBX (47%) is mixed with benzoic acid (23%) and isophthalic acid (30%)a safe and useable formulation is produced -SIBX .  A comparison of IBX and SIBX reactivity showed that the two materials gave almost identical reactivity profiles.  For large / manufacturing scale work the residue can be reused by recycling back to the suppliers who will regenerate the oxidising power and re-supply the material. Similarly iodobenzene the by-product from the other commerically available hypervalent iodine reagents diacetoxyiodobenzene, (DIB, PhI(OAc)2) and bis)trifluoroacetoxy)iodobenzene (BTI, PhI(OCOCF3)2) can also be recycled.

Hypervalent iodine reagents are extremely useful oxidising agents which will carry out standard oxidations (alcohol to aldehyde/ketone) as well as some more challenging reactions such as oxidation of benzylic methyl groups to give carboxylic acids and some unusual reactions such as demethylation of orth-hydroxyanisoles.

The Role of New Technologies in Defining a Manufacturing Process for PPARα Agonist LY518674

Dr John Werner, Eli Lilly, USA

This presentation encompassed the use of a number of modern techniques that were applied to the development of the 4-step synthesis of LY518674 (see Scheme 2).  For example UPLCTM allowed a chromatographic separation of up to 24 components to be developed in less than a day to develop with the same method being applicable to all 5 stages of the synthesis.  FBRM on line monitoring probes were used to improve crystallisation of stages 1, 3, and 4 as well as being used to study the crystal size of the crude product in stage 1.  The crude product crystallised directly from the reaction mixture as a foam initially but this problem was avoided by changing the reaction solvent from water to 1-butanol.  DoE was used to optimise stage 3 (acid chloride formation) and in particular to help find reliable conditions using pyridine as catalyst because the original catalyst, DMF, led to an impurity in the API.

Scheme 2: Synthesis of LY518674:

 

 

Development of Highly Practical and Large Scale Ring-Closing Metathesis and economical and Greener synthesis for BI HCV Protease Inhibitor

Dr Chris Senanayake, Boehringer-Ingelheim, USA

This talk centred on the development of the ring-closing metathesis chemistry used to construct the 15-membered ring in Boerhinger-Ingelheim’s HCV protease inhibitor clinical candidate BILN 2061.  By understanding the reaction profile and mechanism, studying the impurities generated, using a modified substrate, and improved catalyst the chemists at BI were able to dramatically improve the process.  A change of protecting groups in the substrate eliminated a troublesome epimerisation side product, whilst changing the catalyst allowed the throughput of the reaction to be increased 10-fold whilst still maintaining an assay yield of >90%.

Development of Reactions Used to Make Key Improvements in API Synthesis – Process Development of CP-944629.

Dr Bryan Li, Pfizer, USA.

CP-944629 is a potent p38a inhibitor is a 4,5-disubstituted oxazole (full name: 3-tert-butyl-6-(4-(2,4,5-trifluorophenyl)oxazol-5-yl)-[1,2,4]triazolo[4,3a]pyridine).  There were two main problems to be solved in developing a scaleable synthesis – use of anhydrous hydrazine to form the triazolopyridine ring and finding an efficient and selective method to form the oxazole ring.  The hydrazine problem was solved by reacting 2,5-dibromopyridine with hydrazine hydrate in PEG-300/water/butanol at reflux giving 2-hydrazino-5-bromopyridine which could pivalated and cyclized to form the triazolopyridine moiety.

Various approaches to the oxazole ring were attempted such as the use of 2,4,5-trifluorophenyl-Tosmic which is not stable enough for isolation.  The use of Tosmic resulted in the formation of a triazolopyridyloxazole intermediate in 82% yield which could be regioselectively brominated (regioselectivity 200:1 after significant optimisation work) in 87% yield and the synthesis was completed by a Suzuki-Miyaura coupling.  Some of this work has been published in Org. Process Res. Dev., 2007, 11, 951-955.

Organometallic Fluorine Chemistry: Studies toward Aromatic Fluorination

Vladimir Grushin, DuPont, USA.

Dr Grushin presented work on studies of the use of palladium fluorides to effect aromatic fluorination thus providing an alternative to the Balz-Schiemann reaction.  However despite making numerous Pd-F complexes and investigating many intermediate complexes direct fluorination of iodides and bromides does not work well.  Aromatic trifluoromethylation was successful, as was direct fluorination either via benzyne intermediates or using copper fluoride – tetramethylethylenediamine complex.  However palladium fluoride complexes are not completely unreactive.  They will catalyse fluorocarbonylation of aromatic halides and in the presence of bis(triphenylphosphine)iminium chloride produces “naked” fluoride that is very nucleophilic and will substitute the chloride of dichloromethane at room temperature.

Improvements in the Preparation of Key Intermediates in Manufacturing Processes of Torcetrapib and Ziprasidone API

Dr Durgesh Nadkarni, Pfizer, USA

In this presentation Dr Nadkarni described improvements to the first 2 steps of the synthesis of both these products.  The main issues with the first two steps of the Torcetrapib process were the cost of the starting material (4-bromobenzotrifluoride), the ligand used in the Buchwald-Hartwig amination step (tri-tert-butylphosphine) which is pyrophoric and subject to patent issues, the use of expensive cesium carbonate as base, and the throughput in the subsequent nitrile hydrolysis step.  Potassium phosphate appeared to be a good alternative base but the variability in the quality of commercial batches of material mitigated against it’s use on scale.  Potassium hydroxide was found to work well provided a biphasic system including water was used as the reaction medium.  The cheaper chloride starting material could be used but necessitated a ligand screen to find a suitable catalyst system.  Throughput in the nitrile hydrolysis step was improved by a two prong approach – DoE to optimise the reaction conditions (water: sulphuric acid ratio) and the use of ammonium hydroxide rather than sodium hydroxide for neutralisation which is significantly more volume efficient due to the higher water solubility of ammonium sulphate over sodium sulphate.

The first two steps of the Ziprasidone synthesis, acylation of 6-chlorooxindole with chloroacetyl chloride and reduction of the carbonyl group with triethylsilane, gave concern because of the isolation of the intermediate chloroketone which is a strong skin sensitiser and the use of corrosive trifluoroacetic acid as solvent in the second step.  Both these issues were circumvented by changing the reducing agent in the second step to tetramethyldisiloxane which works best in the presence of a Lewis acid such as aluminium chloride.  This meant that at the end of the acylation step no isolation was required and the reducing agent could simply be added to the crude reaction mixture to effect the complete reduction of the carbonyl group.

Practical Synthesis of a Chiral 2-Morpholino Investigational New Drug Candidate

Dr. Michael Kopach, Eli Lilly, USA

The synthesis of the target compound, an NERI inhibitor, was slightly unusual in that it included three Grignard reactions (see Scheme 3).  Based on the precedent from previous compounds in the series the first Grignard reaction was between racemic 2-cyano-N-benzylmorpholine and the Grignard reagent derived from 4-chlorotetrahydropyran.  However the major product arose form self-condensation of 2 moles of nitrile with only 10-15% of the desired product being formed at best.  Various other substrates were evaluated for the Grignard reaction including the ester, carboxylic acid lithium salt, but the best results were obtained using the morpholine amide as substrate.  And the reaction could be carried out on chiral morpholine amide, produced by classical resolution with di-p-toluoyl tartaric acid.  It was important to add the morpholine amide substrate to the Grignard reagent (smooth Grignard initiation was achieved at 40ºC by addition of Dibal and iodine to speed up the initiation) to avoid racemisation of the chiral centre.

The chiral ketone was then reacted with the Grignard reagent formed from 2-methoxy-5-fluorobenzyl bromide, itself produced via a Grignard reaction between 2-methoxy-5-fluorophenyl magnesium bromide and dimethyl formamide (followed by subsequent reduction and bromination).  This benzyl bromide is a strong lachrymator and so avoiding isolation is preferred.  In this case in situ Grignard formation gave 10% more active Grignard reagent than the standard Grignard formation with isolated benzyl bromide, mainly due to lower levels of Wurtz coupling.  The synthesis of the chiral ketone intermediate has now been published (see M.E. Kopach et al, Org. Process Res. Dev., 2009, 13, 209-224).

Scheme 3: Synthesis of chiral 2-morpholino API:

An Efficient Process for the Synthesis of Methyl 1-(2,3,5-tri-O-acetyl-β-L-ribofuranosyl)-1,2,4-triazole-3-carboxylate

Dr. P. Zhang, Roche Carolina, USA

Dr Zhang’s presentation showed how and an extremely unwieldy low yielding process can be significantly improved by careful process development.  The title compound is prepared by peracetylation of L-ribose followed by reaction with methyl 1,2,4-triazole-3-carboxylate.  β-Tetraacetyl-L-ribofuranose is commercially available but the purchase price was prohibitively expensive ($660-760/kg).  Making the tetra-acetate requires 3 steps, formation of the methyl acetal, tri-acetylation and finally acetolysis to form the tetra-acetate.  However the literature procedure for these 3 steps requires 16 extraction steps and 8 distillations to dryness giving an overall yield of 57%.  By understanding the individual reaction steps, changing solvents, and reagents the process was optimised.  Thus sulphuric acid was used in place of hydrochloric acid for the acetylation step which was neutralised with lithium carbonate rather than multiple pyridine additions and strips to dryness.  Solvent switch from methanol to acetic acid and addition of acetic anhydride gave the tri-acetate which was subjected to acetolysis by addition of sulphuric acid.  By careful adjustment of the acetic anhydride charge and the reaction temperature the solution yield of the tetra-acetate was increased to 95%.  The tetra-acetate solution was neutralised, concentrated and then extracted in to dichloromethane prior to reaction with the triazole methyl ester to give the title compound in an overall isolated yield of 74% on 100kg scale.