e-NEWSLETTER

What’s New in Process Chemistry?

Issue 8

March / April 2003

CONTENTS

NOVEL ENOLATE GENERATION

MICHAEL ADDITIONS CATALYSED BY BISMUTH NITRATE

ALTERNATIVE OXIDISING AGENTS

TOPICS IN STEREOCHEMISTRY – CRYSTALLISATION OF DIASTEREOISOMERS

ADDITION OF GRIGNARD REAGENTS TO NITROARENES

AEROBIC OXIDATION OF AMINES TO NITRILES CATALYSED BY RUTHENIUM ON ALUMINA

CATALYTIC PROPARGYLATION OF AROMATICS

ORGANOCATALYSED ENANTIOSELECTIVE REACTIONS

OXIDATION OF CYCLOHEXENE TO ADIPIC ACID

FOSINOPRIL – SYNTHESIS AND SEPARATION OF ISOMERS

SECOND GENERATION PROCESS DEVELOPMENT

A HIGHLY TOXIC (CARCINOGENIC) IMPURITY IN A DRUG SUBSTANCE REQUIRED A SPECIFICATION AS LOW AS 0.000007%

Novel Enolate Generation

Exposing enones to a rhodium catalyst under 1 atmosphere of hydrogen gas generates a enolate which can react either intermolecularly or intramolecularly with carbonyl compounds.  The work of Michael Krische at University of Texas was reported at the ACS meeting in New Orleans in March 2003 (see C & E News, 2003, April 7^th, p 28-29).  Because of its catalytic approach, the atom-efficient method promises to have industrial applicability.

Michael Additions Catalysed by Bismuth Nitrate

Michael reactions typically use basic or acidic catalysts, often in stoichiometric amounts.  It has now been found by the group of Banik at Univ of Texas, Houston that bismuth nitrate will catalyse the addition of amines, carbonates, imidazoles, indoles and thiols to enones.  The mechanism is not yet understood (C & E News, 2003, April 7^th, pp 28-29).

Alternative Oxidising Agents

IBX has been widely touted as an oxidising agent for oxidising alcohols to carbonyl derivatives and has recently become commercially available in kg quantities from a French company.  Even when pure, however, IBX has a tendency to explode.  The IBX esters (esters of 2-iodoxybenzoic acid), however, are quite stable and have similar reactivity to IBX.  They are made by simple hypochlorite oxidation of o-iodobenzoic acid esters.

The work was carried out in the groups of Zhdankin at Univ of Minnesota, Deluth and Tykwinski at Univ of Alberta, Edmonton.  The esters are white non-explosive crystalline solids and are easily handled under standard lab conditions (C & E News, 7^th April, 2003, pp 28-29).

Topics in Stereochemistry – Crystallisation of Diastereoisomers

The latest volume of topics in Stereochemistry Vol. 23 (ed S E Denmark) Wiley 2003, 369pp, ISBN 0-471-17622-2 (179 Euros) contains a chapter from Kinbara and Sargo on diastereomer separation by racemates.  Other chapters include the asymmetric aldol reaction and the chemistry of enolates.  The earlier vol. 22 (1999, 314pp, ISBN 0-471-25316-2 (119 Euros) included chapters on asymmetric catalytic reactions using chiral lanthanide complexes (Shibasaki and Sasai) and on asymmetric amplification (Fenwick and Kagan)

Addition of Grignard Reagents to Nitroarenes

A short critical review of current literature has appeared (Ricci A et al, Ang Chem Int Edn, 2003, 42, 1444), prompted by recent work from the group of Knochel (J Am Chem Soc, 2002, 124, 9390) on diarylamine synthesis.

This work nicely complements the Buchwald-Hartwig methodology for diarylamine synthesis.  Previously, Bartoli had used the reaction of vinyl Grignard reagents with nitroarenes as a novel method to produce indoles.  This is a complex process involving 6 intermediate steps (see for example Bartoli G et al, J Chem Soc Perkin Trans I, 1991, 2757) but can be a useful synthetic scheme if an ortho substituent is present.

The methodology has recently been extended to produce highly functionlised indoles (Dobbs A, J Org Chem, 2001, 66, 638).

Aerobic Oxidation of Amines to Nitriles Catalysed by Ruthenium on Alumina

Ru/Al₂O₃ is a useful catalyst for the oxidation of alcohols (Yanagushi K et al, Angew Chem Int Ed, 2002, 41, 4538) and the group of Mizuno in Tokyo has continued to explore the use of the catalyst.  A recent communication (Angew Chem Int Ed, 2003, 42, 1480) reports on the facile oxidation of amines to nitriles or imines.

Kinetic studies indicate a zero-order dependence on oxygen pressure and on amine concentration and a first order relationship with the amount of catalyst.  Deuteration studies gave the result below.

Catalytic Propargylation of Aromatics

Ruthenium complexes catalyse the propargylation of aromatic and heteroaromatic compounds giving high yields of adducts (Nishibayahi Y et al, Angew Chem Int Edn, 2003, 42, 1495).

Organocatalysed Enantioselective Reactions

Over the last few years several groups have focussed on the use of organic catalysts to promote C-C bond forming reactions.  In the USA, the groups of MacMillan, List (both Europeans!) have predominated, whereas in Europe, the work of Jorgensen at Aarhus, Denmark has been well received.  These reactions are of great industrial interest since they build up highly functionlised molecules using simple procedures in which the catalyst can easily be reused.  Jorgensen has recently published on the inverse – demand Hetero-Diels-Alder reaction (Juhl K et al, Angew Chem Int Edn, 2003, 42, 1498).  The cycloaddition of aldehydes to unsaturated carbonyls gives the dihydropyran, which is the presence of silica yields ring opened products.  The silica also helps catalyst turnover.

Oxidation of Cyclohexene to Adipic Acid

The combination of aqueous hydrogen peroxide and a molecular sieve catalyst (titanium aluminophosphate number 5 in the IZA classification, TAPO-5) allows slow oxidation of cyclohexane to adipic acid in the absence of organic solvent (Lee S-O et al, Angew Chem Int Ed, 2003, 42, 1520).  The mechanism is via cyclohexane oxide and cyclohexanediols but more than one mechanism may operate.

100% conversion can be achieved at 80^o for 72 hrs with a turnover number of 310, but as yet selectivity is 30%.  The paper describes experiments designed to understand the mechanism more fully.

Fosinopril – Synthesis and Separation of Isomers

Fosinopril is a drug which came off patent in 2002, although several process patents have extended the protection to 2007.  For the generic drug companies, one challenge is either a stereoselective synthesis or a novel strategy for separation of the isomers.  At the recent Scientific Update conference on Organic Process R & D, Dr. Girij Pal Singh of Lupin Laboratories, India described the separation of the SRSS from the RRSS, RSSS and SSSS isomers by making the cesium salt dihydrate which crystallised easily, leaving the other isomers behind.  The process is operated on 100 kg scale – all the unwanted isomers are recycled back into the system.  Fosinopril sodium crystallises in 2 forms depending on the water content of the solvent.  From a keto or hydroxylic solvents with > 2% water form A is produced, whereas form B is produced when the water is < 0.2%.  However Lupin have developed a slow crystallisation process which gives the desired form A but in a low water content solvent.  Lupin have also an improved process for preparing phosphorous acids under mild conditions.

Second Generation Process Development

Many pharmaceutical companies do little process development after a drug is launched partly because of the fear of regulatory issues such as new impurities arising.  Roche (and prior to the merger Syntex) have for many years taken the view that “second generation” process development, which includes synthetic route change is some cases, can lead to large cost savings, and really does pay for itself.

At Scientific Update’s Organic Process R & D conference in New Orleans in March 2003, Peter Harrington from Roche spoke about several examples including the drug bosentan.

The last step in the synthesis involved a replacement of a chlorine by the 2-hydroxy ethoxy group, but a number of by-products arose.  By the use of ethyleneglycol t-butyl ether (ETB), a new sequence was introduced which had a crystalline intermediate suitable for cleaning up impurities, which allowed further improvements and economies to be made earlier in the sequence.  The new route is shown below.

The effect on the overall efficiency of the process is shown from the table below comparing the old and new processes.

Characteristics Isolations Drying Operations Sensitizers Isolated Solvents Yield Purity

Glycol 6 5 2 6 67 99.3

ETB 3 2 0 2 84 99.7

Second generation processes are often patentable and can usefully extend the product lifetime of successful products by limiting the options for generic competition.

In a similar manner, the process for orlistat involved a late stage resolution without the opportunity to recycle the unwanted isomer.  However as shown on the right hand side of the Scheme    , a resolved “raw material” would allow a shorter synthesis of orlistat.  The “raw material” was synthesised from a simpler precursor by the route shown below.

A late stage resolution has a much lower space-time-yield and therefore involves more capital expenditure (larger equipment) than a route involving an early resolution, an asymmetric synthesis or synthesis from a readily available chiral precursor.

A Highly Toxic (Carcinogenic) Impurity in a Drug Substance required a specification as low as 0.000007%

During the synthesis of the drug PGE-6134091, a highly potent a-2 adrenergic Receptor Agonist, a batch of drug substance failed the Ames test.  It was subsequently found that phenazene impurities in the ppm range were responsible, and after discussions with FDA, a limit of 70 ppb (0.000007%) was set for these impurities.  (J Randall, presented at OPRD, New Orleans, March 2003).

The carcinogenic phenazine impurities arise during the early stages and could be deliberately synthesised by the method shown below.

If a transfer hydrogenation was used, however, low levels of phenazines were produced.