The following extracts have been taken from Scientific Update’s conference “Modern Synthetic Methods, Reaction to Reality” held in May 2004 at the Biofine Exhibition in Berlin. Any lectures which were not available the time of the conference will be posted on the website.
Copies of the full conference proceedings can be purchased from Scientific Update upon request.
Professor Steve Ley (University of Cambridge) opened the conference with an update on his group’s research on encapsulated catalysts (encats). These are precious metals, which are stabilised and encapsulated in polyurea beads. The resultant catalysts are soft but robust, they don’t scour vessels and they can be recycled. Examples included the use of encapsulated Pd(OAc)2 for Suzuki and Heck reactions, where high yields at low Pd levels are seen. This “encat” is available from Avecia who have scaled examples to 25kg. The same “encat” can be used for catalytic hydrogenation and can be recycled 30 times. In contrast, whilst Palladium-on-charcoal can be re-used or recycled, sometimes a problem of oxygen formation, by reaction with water, can occur – a combination of oxygen and hydrogen in a reaction is a potentially explosive mixture. Other types of encapsulated catalysts are a Pd(O) catalyst, useful for hydrogenation where there are hydrogenolysable groups are present. The advantages of these Pd catalysts over Pd/C is that they are easy to filter and they are not pyrophoric.
Professor Ley also described a new range of catalysts based on perovskite minerals. The palladium containing perovskites are used in catalytic converters in car exhausts and contain not only palladium but lanthanum, iron and cobalt. His group have examined the catalysts for organic synthesis and have found applications in Suzuki reactions. Catalyst loadings as low as 0.0038% Pd and TONs up to 400,000 have been obtained. The catalysts are easily filtered off, washed and reused, and leave low residual Pd in products.
An analogous perovskite copper catalyst is also useful for Ullmann reactions at a 2.5% loading, whereas a combined copper and palladium perovskite has been used for Sonogashira couplings. Although these catalysts appear to be heterogeneous, there may be some metal in solution at elevated temperature, with the perovskite acting as reservoir and scavenger. This helps in catalyst turnover and in preventing product being contaminated by the metal catalyst. For some catalysts there may be an induction period.
Dr. Ulrich Scholz from Bayer Fine Chemicals described the industrial scale operation of Buchwald-Hartwig amination reactions. The work was carried out in collaboration with Prof Buchwald at MIT and with Rhodia Pharma Solutions.
Key issues were the quality of the palladium source, where results varied from vendor to vendor and the quality and particle size of the base (e.g. K3PO4). Catalyst deactivation and precipitation was also a potential problem, but good agitation in a baffled reactor improved the process tremendously. The base must be added in a stepwise manner and water must be present to prevent base fouling during the reaction.
These are challenging reactions for scale up, since the active catalyst may precipitate on the heterogenous base, so good agitation and dispersion of solids is imperative. Entrapment of the catalyst in the product is prevented somewhat by choice of solvent or co-solvent. Toluene-THF and toluene-methanol are the best solvents for scale-up Catalyst stability is also a potential difficulty and is helped by introduction of the catalyst (Palladium acetate, phosphine ligand such as Dave-Phos, S-Phos, X-Phos or Me-Phos) as a preformed solution at elevated temperature.
The methodology was used to prepare on large scale arylpiperazines from aryl chlorides, the main sideproduct being the diarylpiperazine.. At pH 3-4, the byproduct remains in the aqueous layer whilst the desired arylpiperazine can be extracted into the organic phase.
Continuing the theme of industrially viable catalysts, Manfred Reetz (Max Planck Institute) presented his latest work on Directed Evolution of Enzymes for organic synthesis, where enzymes can be designed for stability and enantioselectivity. In the process in scheme 1, a wild type enzyme gives a 55:45 mixture of products. By a fast and efficient mutation method, enzymes which are R-selective (95:5 ratio) or S selective (11:89 ratio) could be obtained. These catalysts also worked with other substrates or even in other reactions.
The methodology relies on the rapid production and screening of thousands of mutants for the desired reaction and with further rounds of mutation or evolution (using the error-prone polymerise chain reaction), the enzyme can be optimised for the particular process. With automated equipment, the production and screening of thousands of enzymes is only a few days work. The methodology has already been applied on a large scale in industry and more applications are expected in the future.
The Merck process chemist, Ed Grabowski, who through his publications has helped to give the discipline of process R&D “academic respectability”, gave some “Reflections on Process Research” on his 38 years at Merck, USA. He is to retire shortly, and this talk summarised many of his achievements over the years. There were some useful aphorisms such as “the core is our enemy – don’t believe everything that is said, especially when it is said by experts!” The latter comment was in relation to a project requiring an ester to be hydrolysed in the presence of a β-lactam – the experts said it could not be done. However, by careful choice of conditions, 100:1 selectivity for ester hydrolysis versus β-lactam ring opening was obtained.
His lecture emphasised the importance of detailed mechanistic understanding of reactions in the scaling of processes to industrial manufacture. Byproducts formed in reactions give clues as to where the process “goes wrong” and how it can be corrected. Hopefully, an expanded version of this talk will be written up for publication in Organic Process R&D in the future
Nicholas Leadbeater gave a controversial talk on “transition-metal free coupling reactions”. His group has found that under microwave conditions, Suzuki couplings will occur in the absence of palladium catalysts, provided that water is present and a phase-transfer catalyst (PTC) is used. The reaction is very sensitive to the purity of the reagents, particularly the boronic acid and the PTC, and requires good agitation and premixing of the solid reagents.
They have screened boronic acid samples from a number of distributers and found that the purity varies considerably. Some samples had to be purified by aq-organic extraction or column chromatography. If the PTC is highly contaminated, then the reaction will not work either. The temperature must be above 135-140 in the microwave vessel, but does not work if 170 is exceeded, presumably owing to decomposition (loss of B) of the boronic acid. The heterogeneous reaction mixture must be freshly prepared (do not allow to stand overnight!) and must be well mixed before introduction to the microwave vessel.
Leadbeater assured the audience that his reactions contain no extraneous palladium (e.g. new glassware is used and the palladium content of all raw materials is checked) but during a lively question time, it was clear that there were still one or two sceptics who felt that there must still be some palladium contamination in the reaction
Dr. Orlin Petrov from Schering Berlin, one of the conference’s sponsors, presented two case studies, the first being on the design of an epoxidation catalyst for a crucial steroid epoxidation (Scheme 3) and the second part on scale up of an Evans Aldol reaction, where a scale up problem was encountered. This caused him to propose another aphorism “A surprise in research is a discovery; a surprise in development is a disaster”.
In this case, the problem was traced to the instability of the Evans auxiliary in LDA at –78°C over time; on scale up the extended addition time lead to isomerisation.
The entertaining opening talk on the second day from Tony Barrett (Imperial College, London) also included an aphorism “chromatography is the meek acceptance of synthetic failure”. His lecture described the use of ring-opening metathesis polymers (ROMP-gels) in parallel synthesis using the mix-filter-evaporate protocol. A particularly practical use of ROMP gels is in impurity annihilation, for example in the Mitsunobu reaction where product isolation can be extremely difficult.
Kai Rossen from Degussa elaborated a number of development projects including an approach to β amino acids based on the Rodionov condensation followed by enzymatic kinetic resolution (Scheme 4). He also described Degussa’s new catalysts for Suzuki and other reactions.
The importance of the Suzuki coupling reactions in the synthesis of modern drugs was again emphasised by Atul Kotnis from Bristol Myers Squibb. He pointed out that some boronic acids are unstable to the Suzuki reaction conditions (high temp, pH above 10) and may decompose before coupling. The order of mixing of the reagents, as well as the choice of catalyst and base, may be crucial. For the coupling shown in Scheme 5, slow addition of the boronate to a mixture of the iodo oxazole, Pd catalyst and sodium carbonate was the preferred protocol.
Bernhard Rausch from Clariant, Germany described the use of lithium metal as an alternative to butyl lithium in the manufacture of functionalised aromatics. Lithium is used to make aryl lithiums from halo-aromatics, and these can be converted to a variety of products, including boronic acids for Suzuki coupling – the latter is done on tonne scale at Clariant. Lithium metal can also be used for directed ortho metalations, and is much more economical than using butyl lithium. However there are safety issues (Ciba had an explosion and fire at a Swiss plant in 2001) and lithium reacts with THF in an exothermic ring opening reaction, so THF cannot be the solvent. Clearly Clariant have developed expertise in the safe handling of lithium on large scale and during question time Dr Rausch was asked for further details. However, for understandable commercial reasons, Clariant were unwilling to divulge more details which may assist competitors. Lithium metal is obviously cheaper than BuLi, but probably not as easy to charge to a reactor. ( However, the latter difficulty must be solved by the BuLi manufacturers).
In the final lecture of the conference, Barry Trost (Stanford University) presented his work on inventing reactions for atom economy. He described many new reactions which have the potential for industrial use, building complex molecules efficiently using catalytic processes. A key example is shown in Scheme 6.
