Making Metathesis Work

OLEFIN METATHESIS is a fundamental catalytic reaction that stands among a handful of the most versatile ways to make carbon-carbon bonds and build molecules. As the reaction takes its course, carbon-carbon double bonds are broken apart and re-formed with a simultaneous exchange of substituents, ring closing, ring opening, or polymerization. The petrochemical, polymer, and specialty chemical industries have exploited metathesis for nearly half a century to process many simple hydrocarbons.

The Nobel Prize-winning work of Yves Chauvin, Robert H. Grubbs, and Richard R. Schrock underscores the technology's important role in organic synthesis. In the past decade, thanks to the well-defined transition-metal catalysts first developed by Grubbs and Schrock, there's been an explosion in the development of other metathesis catalysts and in the number of published laboratory syntheses using metathesis as a key step. Many of these target natural products and compounds of pharmaceutical interest (Angew. Chem. Int. Ed. 2006, 45, 6086 and 2005, 44, 4490).

Despite its power and applicability, metathesis is still experiencing some growing pains outside industrial chemicals. It sits on the verge of entering commercial pharmaceutical manufacturing. But practitioners are finding that moving the chemistry from the lab to the production plant is about more than just bigger pots. Technical challenges are certainly a big aspect, but so, too, are issues around the commercialization of academic inventions and industry's adoption of new technologies.

Examples appearing in the scientific literature and in patent filings make it clear that pharmaceutical chemists at most of the major drug companies are using these catalysts in lab-scale syntheses. This is not believed to be the case in process development and production scale-up.

"In pharmaceuticals, there are very few industrial uses of these catalysts on larger preparative scales," says Siegfried Blechert, a chemistry professor at the Technical University of Berlin who consults with several companies. Several factors are at play, he explains, including the amount of catalyst needed and its cost, as well as technical issues of activity, substrate specificity, stereoselectivity, and residual metal removal.

"For simple molecules without many functional groups, you do not need large amounts of catalyst," Blechert says, referring to existing industrial-scale applications. But for pharmaceutical syntheses, he adds, "if you're thinking about more interesting small molecules with more functional groups, then you need much more of the catalyst." Quantities required often reach a few mole percent, and Blechert says there's a need for more active catalysts with longer lifetimes and higher turnover numbers that can be used in catalytic parts-per-million amounts.

The hunt for better metathesis catalysts has already yielded plenty of fruit, but only a fraction is available for purchase. "In industry, people want to buy a catalyst to try the reaction, and the ability to do this is restricted to the commercially available catalysts," Blechert says. He also believes these catalysts may not represent the optimal ones, but rather only those unhindered by intellectual property (IP) restrictions.

Although Blechert and others cite a lack of commercially available catalysts, in relative terms the situation is significantly better than it was a few years ago. Strem Chemicals now offers seven ruthenium and six molybdenum catalysts, many of which it sells in collaboration with Degussa, Ciba Specialty Chemicals, Umicore, and Zannan Pharma. Since 2003, Sigma-Aldrich has exclusively offered six Grubbs and Hoveyda-Grubbs catalysts produced by Materia and will soon add another five Materia catalysts.

Materia, based in Pasadena, Calif., was created in 1997 to commercialize metathesis polymer technology. It has an exclusive license to all technology and IP developed in the labs of Grubbs at California Institute of Technology, Schrock at Massachusetts Institute of Technology, and Amir H. Hoveyda at Boston College. All three chemists serve on Materia's scientific advisory board and participate in its efforts to develop and commercialize new metathesis technology.

Metathesis is a "generic transformation," requiring only that there be double bonds present, and thus has broad applicability, says Michael A. Giardello, Materia's chief executive officer. Typically carbon-carbon bond formation is a harsh process that requires a large number of protecting groups on the reactants. "But with olefin metathesis, you can form these bonds even very late in a synthesis because it is very tolerant to functional groups and works under very moderate conditions," Giardello says.

In the fine chemicals area, Materia has developed efficient cross-metathesis (CM) routes for synthesizing pheromones (Adv. Synth. Catal. 2002, 344, 728) and launched a business in 2002 around organic building blocks based on metathesis chemistry. "Olefin metathesis allows you to very easily make functionalized medium- and large-sized rings," Giardello adds, "and those are ubiquitous structural elements in pharmaceutical fine chemicals."

Existing catalysts are active enough that metathesis reactions are widely used in medicinal chemistry and discovery work. "In the past several years, metathesis has made its way into development as well," Giardello says. "Several drugs that are in Phase I or II clinical trials use our catalysts in the synthesis of an intermediate that's key to the drug." Although Giardello admits that it's still probably a few years before a drug synthesized via a metathesis step is approved, he says Materia is currently providing bulk quantities of catalysts for several industrial applications as well as for multiple clinical trials.

Meanwhile, Materia continues to expand its catalyst offerings through ongoing research. The catalysts Materia sells in research quantities come with explicit licenses restricting any commercial use of the catalysts or resulting products in manufacturing, patenting, and product or process development, including clinical trials.

"If you want to use a catalyst—maybe start scaling up for development or filing for patents—you then form a direct relationship with Materia," Giardello explains. If a company becomes a licensee or partner, it then gains access to many more of the catalysts that Materia has developed as well as the opportunity to work with its scientists to fine-tune the catalysts to work with different substrates in partner-specific applications.

Materia is flexible in how it sets up these relationships and determines its compensation, Giardello says. Possible arrangements include expanded R&D licenses, development partnerships, full-scale commercial licenses, or even contract manufacturing of intermediates. "We want to share in the value we create, and we take it on a case-by-case basis," he says. Its share can vary, for example, from a portion of the cost savings gained in streamlining a process; an annual fee; a royalty on sales; or a return by building the IP value into the catalyst price.

Many participants in the fine chemicals industry tell C&EN that customers, especially large drug companies, are reluctant to negotiate such licenses and, although they might pay for one particular synthetic step, they may be opposed to paying royalties on their product sales. Late last year, however, Materia licensed its metathesis catalyst platform to Merck for use in discovery and development. Although the terms of the deal were not disclosed, that it exists at all, Giardello says, "speaks to the fact that we can get over these barriers and reach a mutually beneficial solution."

Metathesis catalysts offered in small quantities by Strem are sold for research purposes. "From a research perspective, these catalysts are priced similarly to other catalysts on the market for a variety of different applications," explains Ephraim S. Honig, Strem Chemicals' chief operating officer.

Degussa and Umicore sell bulk quantities under straightforward terms. "The simplest model, and the one that we're told the pharmaceutical companies like the best, is to have one price that covers all of the associated IP, and you simply offer the catalyst on the basis of dollars per kilogram," Honig explains. "You pay up front, and it doesn't matter how much product you make or what the value of your product is; it's simply a one-shot price with no strings attached."

Strem and its partners began offering metathesis catalysts because of what they saw as a growing desire in the marketplace for technology options. "Pharmaceutical researchers are pressed for time, and if there's a research product available for them to order and evaluate, they'll try it," Honig says. "If it's not available, they're not very likely to try to make it themselves; they'll just work around the technology."

As is true of any new catalyst technology, Honig anticipates that, if the catalysts continue to perform and more results are published, the level of interest will continue to rise. "I don't think it's as popular a transformation yet as asymmetric hydrogenation, but it's definitely an important one," he adds. But, at the outset, chemists simply have to have metathesis on their radar screens when they're thinking retrosynthetically.

"We strongly believe that there will be more commercial applications of metathesis catalysts in the near future," agrees Jürgen Krauter, director of marketing and business development for Degussa Catalysts. He offers two main reasons: increased knowledge about metathesis technology as applied to new active pharmaceutical ingredients and lower hurdles to its use under straightforward business models. "Customers don't want to sign complicated license agreements, and they really appreciate Degussa's approach," he adds.

In February 2006, Degussa launched the first member of its catMETium family of metathesis catalysts, catMETium IMesPCy, for pharmaceutical applications. Although sales thus far have been largely for R&D purposes, the company has sold some commercial quantities, Krauter says. The company is working on broadening its metathesis catalyst portfolio and may launch another this year. It collaborates with chemistry professors Karol Grela at the Polish Academy of Sciences' Institute of Organic Chemistry and Steven P. Nolan of the University of New Orleans (UNO), who is now in residence at the Institute of Chemical Research of Catalonia in Tarragona, Spain.

The catMETium IMesPCy catalyst was developed and first patented by chemistry professor Wolfgang A. Herrmann at Technical University Munich, in Germany, and later licensed to Aventis Research & Technology, which Degussa acquired in 2001. Grubbs and Nolan also hold patents on catalysts with similar ligands. Owning IP is beneficial, because commercializing metathesis catalysts entails navigating through a forest of composition-of-matter, method-of-synthesis, and field-of-use claims.

Researchers often are eager to develop new catalysts, but pitfalls abound. For instance, researchers may not realize that if a proprietary catalyst forms as an intermediate in the R&D process of generating a new catalyst , they must get a license from the patent holder because they have gone through protected material, Honig explains. "Someone in industry may be aware of that, but an academic may not realize it."

RESEARCHERS and industry professionals suggest that only large companies have the resources to do the necessary patent review before moving into the market. "It is indeed a challenging task," Krauter says. "Freedom to operate and to have a reliable catalyst vendor partner is crucial for our pharmaceutical customers."

Materia agrees that freedom to operate is an important issue. "There's always competitive technology out there," Giardello points out. "We feel very confident, however, that our issued patents provide a dominant intellectual property position in metathesis and the broadest freedom to operate." Although there has been litigation in this field in the past, both Krauter and Giardello report that none is ongoing.

Meanwhile, more competitors are finding IP gaps in which they can offer alternative catalysts. These new catalysts differ structurally, sometimes subtly, but generally stray little from a basic ruthenium carbene complex developed by Grubbs. Although the molybdenum catalysts are more reactive, they are not as tolerant of air, water, or functional groups, and therefore, industrial use may be more limited. By exchanging or varying ligand structures, developers are altering catalyst sterics and electronics, and consequently the activity, stability, selectivity, and even patentability of new catalysts.

Grubbs's first-generation catalyst, a ruthenium benzylidene complex bearing two tricyclohexylphosphine (PCy₃) ligands, was the first metathesis catalyst to be widely used in organic synthesis. It was followed by a more active second-generation analog, in which an N-heterocyclic carbene (NHC) replaces a PCy₃. Then came the more active and more stable Hoveyda-Grubbs catalysts containing an alkoxybenzylidene ligand and either a PCy₃ (first generation) or NHC (second generation); the latter was also reported by Blechert in 2000.

CatMETium IMesPCy is a ruthenium indenylidene complex with an unsaturated NHC ligand and PCy₃ ligand that Degussa says works particularly well in cross and ring-closing metathesis (Chim. Oggi 2006, 24, 14). Similarly, Umicore's Neolyst M1 catalyst, introduced in late 2004, consists of a ruthenium indenylidene complex and two PCy₃ ligands; it was developed internally and is not patented. Replacing one phosphine ligand with a saturated NHC ligand yielded Neolyst M2, launched in July 2006 under license from UNO. It was followed by Neolyst M3, a ruthenium indenylidene complex with two 2-phosphabicyclononane (phobane) ligands, which Umicore licensed from Sasol Technology.

According to Umicore, M1 is an economical catalyst with excellent air and moisture stability and good activity, and M2 improves upon this performance, especially in ring-closing metathesis (RCM) of tri- and tetrasubstituted olefins. M3 also has greater heat resistance, works well in RCM reactions, and, with a longer lifetime, is expected to show high selectivity at lower catalyst loadings.

Umicore has seen a great amount of interest in the market, says Oliver Briel, marketing director for precious metals chemistry. "In 2007, we plan to introduce M4 and possibly M5 catalysts." In addition to getting its catalysts into pharmaceutical development pipelines, Briel says there are applications in the electronic chemicals and specialty polymer sectors.

Umicore is working with Nolan to better understand the properties of its catalysts and with other metathesis researchers based mainly in Europe. "We also have signed an exclusive licensing and cooperation agreement with Viacatt, a spin-off company of the University of Ghent, in Belgium, founded by chemistry professor Francis Verpoort," Briel adds. "Viacatt has new metathesis catalysts in its pipeline, and Umicore will commercialize them."

Specialty chemicals firm Johnson Matthey launched a new class of ruthenium catalysts late last year; licensed from Ciba Specialty Chemicals, its entries contain bidentate pyridine-alkoxide ligands and either a PCy₃ or triisopropylphosphine ligand.

In 2002, Grela created a version with a 5-nitro-substituted alkoxybenzylidene ligand that has been used with much success (J. Am. Chem. Soc. 2004, 126, 9318). A collaboration among Grela, Ligand Chemie GmbH, and Boehringer Ingelheim led to a series of catalysts with different electron-withdrawing group (EWG) substituents, including sulfones, keto groups, and phosphine oxide.

Zannan Pharma, meanwhile, is offering a version of a Grela-type EWG-activated catalyst with a 5-dimethylaminosulfonyl-substituted alkoxybenzylidene ligand and an NHC ligand. The company also has reported a derivative in which the NHC is replaced with a PCy₃ ligand.

Boehringer Ingelheim has licensed the so-called nitro catalyst and anticipates commercializing it this year. Its plan is to provide the catalyst through catalog suppliers to develop research and industrial interest. The company's interest in metathesis catalysts arose from its own process chemistry work in scaling up reactions to make macrocyclic drug candidates (J. Organomet. Chem. 2006, 691, 5163).

Although Boehringer Ingelheim scientists succeeded in using metathesis at a production scale (see page 38), they faced many hurdles related to catalyst loading, reaction conditions, and residual metal. They originally purchased catalyst from Materia under an agreement whose details they say they cannot disclose. Optimized conditions use a proprietary catalyst owned by Boehringer Ingelheim, the researchers tell C&EN.

It is often preferable for major pharmaceutical companies to engage in the search for a new catalyst internally or through academic contacts than to pay expensive royalties to catalyst producers with proprietary technology, they comment. Catalyst suppliers should keep this in mind when in-licensing new, appealing catalytic technology from academia, they add.

Many academic research groups continue to work on sterically and electronically "tuning" modular catalyst structures to improve performance and broaden their scope. A common problem in catalysis design in general is balancing several desirable, yet simultaneously hard to achieve, properties, Grela says. "It's quite easy to activate a catalyst, but it's very difficult at the same time to have a very selective and stable one and achieve these goals within the same structure."

Grela and collaborators have created what he calls a new class of "Scorpio carbenes," in which an alkoxybenzylidene ligand is doubly chelated to the ruthenium center through coordination of both the ether and an attached ester group (J. Am. Chem. Soc. 2006, 128, 13652). The extra chelation makes the catalysts remarkably stable, Grela reports, while also offering high turnover numbers and 10-times higher activity in a test reaction.

Grela and other collaborators have created a catalyst containing a quaternary ammonium EWG on the benzylidene fragment (Green Chem. 2006, 8, 685). "This is our favorite catalyst now," he says. "It's very easy to prepare, the salts happen to be patent-free, and so we're free to develop and test it." The ammonium group activates the catalyst, allows it to work in solvent-water mixtures, and, because of its affinity to silica gel, allows it to be separated out by filtration. Grela, with chemistry professor Andreas Kirschning at Leibniz University Hannover, developed a similar catalyst that has a protonated amino group and that can be noncovalently attached and immobilized on glass-polymer composite rings (J. Am. Chem. Soc. 2006, 128, 13261).

WATER-SOLUBLE metathesis catalysts promise greener approaches to this chemistry, which typically uses large quantities of solvents to achieve the highly dilute conditions required. Grubbs and graduate student Soon Hyeok Hong reported a variant of the second-generation Hoveyda-Grubbs catalyst that is stable and highly active in water (J. Am. Chem. Soc. 2006, 128, 3508). They did so by attaching a poly(ethylene glycol) chain to the NHC.

Efforts to improve separation and recovery of the catalysts would also make metathesis chemistry greener. To this end, many groups have attempted to immobilize metathesis catalysts by using solid phases, polymers, tagging, and ionic liquids. Recyclable heterogeneous catalysts would also address some of the cost issues surrounding metathesis. But what may be more important, especially for pharmaceutical applications, is that immobilized catalysts would avoid the formation of ruthenium by-products in solution and allow easy separation of residual catalyst from the valuable reaction product.

Immobilized catalysts also would open up opportunities for metathesis reactions in high-throughput and continuous-flow reactors. Blechert's group and collaborators at Merck KGaA in Germany recently have described new mono- and disubstituted fluorocarboxylate ruthenium complexes for both homogeneous and heterogeneous processes (J. Organomet. Chem. 2006, 691, 5267). But in general, immobilized catalysts often end up being less active, Blechert says. "There's a need for highly efficient catalysts in heterogeneous form, but this only makes sense if you have catalysts with excellent turnovers in the presence of many functional groups, and this is still a problem."

Despite having created what's often referred to as one of the most efficient and stable ruthenium complexes containing a phenyl-substituted alkoxybenzylidene ligand, Blechert suggests that metathesis catalyst development is still evolving. "We're learning more and more about details of the metathesis reaction, but we're not yet actually able to 'design' catalysts," he says, "There's a lot of trial and error, and this makes the process complicated."

More consistent and better understood experimental results, along with increased knowledge of reaction mechanisms, might enable researchers to vary substituents in meaningful ways to achieve different reactions, Blechert says. He also expresses concern about the lack of comparability of experimental results in testing catalysts. This limitation stems from differences in the conditions and substrates used in various studies and on interpretations of yield and turnover. Along these lines, Grubbs has proposed a standard system of six reactions for comparing catalysts (Organometallics 2006, 25, 5740).

Varying substituents can also yield changes in diastereoselectivity. "A remaining challenge is to solve the problem of selectivity around double bonds," Blechert says. To this end, Blechert's group has made catalysts with saturated unsymmetrical NHC ligands, a relatively rare modification that gave comparable E/Z ratios and improved diastereoselectivites of up to 2:1 compared with existing catalysts (Organometallics 2006, 25, 25). "What we need is a catalyst for exclusively Z or exclusively E configurations, but this is still not possible," Blechert says.

Some of the most exciting advances yet to come in metathesis chemistry are likely to be around enantioselective catalysis (Adv. Synth. Catal. 2007, 349, 23 and 25). Schrock reported the first chiral metathesis catalyst in 1993 for polymer synthesis. Hoveyda and Schrock, who continue to collaborate in this area, published the first efficient enantioselective RCM with a chiral molybdenum catalyst in 1998. Grubbs then reported the first enantioselective metathesis reaction using chiral ruthenium catalysts in 2001, and Hoveyda followed with more in 2002 (J. Am. Chem. Soc. 2006, 128, 1840).

In 2006, Hoveyda received a patent on his chiral ruthenium catalysts containing a saturated unsymmetrical NHC ligand with a bulky binaphthalene substituent. The catalyst is active, stable, highly enantioselective, and can be recovered and reused. Meanwhile, Strem sells one enantioselective Schrock-Hoveyda molybdenum catalyst. With the importance of chiral compounds for pharmaceutical applications, it may be only a matter of time until such catalysts move further toward commercialization and then possibly industrial use.

[Top of Page]