Volume 90 Issue 9 | pp. 10-17
Issue Date: February 27, 2012

Cover Story

Dabbling In Fluorine

With their latest synthetic methods, organic chemists help tackle challenges in fluorine chemistry
Department: Science & Technology | Collection: Green Chemistry
News Channels: JACS In C&EN
Keywords: fluorine, organometallics, cross-coupling, drug discovery
3 out of 10 best-selling drugs in 2011 contain fluorine.
These are structures of three of the top 10 best-selling drugs. They all contain fluorine.
3 out of 10 best-selling drugs in 2011 contain fluorine.
7 out of 35 new drugs approved in 2011 contain fluorine.
These structures show the seven drugs that were approved in 2011 that contain fluorine.
7 out of 35 new drugs approved in 2011 contain fluorine.
Credit: Stephen K. Ritter/C&EN
A graphic of various flourinations.
Credit: Stephen K. Ritter/C&EN

There’s a shake-up taking place in fluorine chemistry. Synthetic organic chemists who don’t normally mess with fluorine are stepping in with their toolbox of synthetic methods to broaden the range of fluorination reactions.

Behind the trend are pharmaceutical and agricultural chemical companies, which need fluorine in their bioactive compounds to keep metabolism in check, facilitate delivery to a target, or improve binding to that target. But fluorine’s hard-to-handle notoriety has limited these companies to using simple fluorinated starting reagents, constraining their ability to crank out new lead compounds that could benefit from a well-placed fluorine group.

Enter the organic chemists, who have already used cross-coupling reactions, which were the basis of the 2010 Nobel Prize in Chemistry, to fundamentally shift how companies conduct new-molecule discovery. Cross-coupling and other reactions are allowing discovery chemists to more efficiently create the complex molecular frameworks they need. Now discovery chemists want to use the same strategies to more efficiently install fluorine in molecules—to do fluorinations with less fuss. The contributions of synthetic organic chemists are greatly easing the work of discovery chemists handling fluorine chemistry, although in one key area—practical catalytic fluorinations—success still is elusive.

Fluorine chemists are feeling chagrined by the invasion of their turf, although they acknowledge that this is one case where having too many cooks in the kitchen is a good thing. Chemists in industry are ecstatic because they suddenly are gaining access to new ways of getting fluorine into their molecules, likely accelerating the discovery process. All the kinks haven’t yet been worked out, but the new dynamic is already leading to a burst of synthetic advances.

“The benefit of adding one, two, or three fluorines into a molecule has definitely been recognized by mainstream synthetic organic chemists,” says William R. Dolbier Jr., a seasoned fluorine chemist at the University of Florida. “This trend is playing out because researchers at pharmaceutical and agrochemical companies are interested in finding truly useful ways to get fluorine into organic molecules—methods that are catalytic, inexpensive, easy to use, and have high yields.”

Dolbier’s group is known for using the reducing agent tetrakis(dimethyl­amino)ethylene, or TDAE, with fluorinated precursors such as CF3I to generate in situ fluorinated carbanions that effectively add CF3 groups to molecules. His team also developed trimethylsilyl 2-(fluorosulfonyl)-2,2-difluoroacetate, or TFDA. This reagent has extended the scope of synthetic difluorocarbene chemistry, enabling the construction of difluorocyclopropane substituents, for example.

Dolbier has mixed feelings about the influx of fluorine newcomers. One reservation he has is that some of them are not scrupulous about referencing relevant earlier work. However, he understands how such lapses can happen. Much like the new crop of organic chemists engaging in fluorine research, Dolbier also was an organic chemist before he started making forays into fluorine chemistry in the 1970s. Because fluorine chemistry is an esoteric field, Dolbier says he sometimes found himself “rediscovering the wheel” as a consequence of not being fully aware of the fluorine literature.

“There is one part of you that wants to tell the newcomers: ‘Hey, stay out of our field. Leave the fluorine chemistry to us. We can do it,’ ” Dolbier says. “On the other hand, we have to recognize that they are bringing in the methodologies they use with nonfluorinated compounds and adapting them to fluorine chemistry to create great new reagents. They are making novel, worthwhile contributions to the field.”

Like Dolbier, fluorine veteran G. K. Surya Prakash of the University of Southern California (USC) also has mixed feelings about the new developments. “Chemists who are doing leading organic and organometallic chemistry and traditionally develop methodologies for drug discovery and natural product synthesis have realized there is a bonanza by focusing on fluorine chemistry,” Prakash says. “Fluorine has become a driving force—it’s the new kingpin of drug discovery.”

Prakash is a little disappointed, however, that some of the new results are being touted as though fluorine chemistry is something new. “Still,” he adds, “it’s a pretty good thing for the field that an energetic wave of organic chemists is joining in and developing new methodologies.”

Prakash is best known for his work with trifluoromethyltrimethylsilane (TMSCF3). In 1989, he and his colleagues showed it is an ideal reagent for adding CF3 to carbonyl compounds. Now known as the Ruppert-Prakash reagent, TMSCF3 is the most widely used source of CF3 for trifluoromethylation reactions.

Working with USC colleague and Nobel Laureate George A. Olah, Prakash has made dozens of fluorine chemistry discoveries over the past 25 years. Just recently, Prakash and his colleagues reported a method to make a trifluoromethylated version of the pesticide DDT, which may be more potent than plain DDT and biodegradable (Org. Lett., DOI: 10.1021/ol201669a). They also developed a Heck coupling reaction for the synthesis of trifluoromethylstyrenes (Org. Lett.,10.1021/ol300076y).

“This new trend is not a shift where academics are suddenly spending all their time on fluorine chemistry,” notes fluorine newbie Phil S. Baran of Scripps Research Institute. “For our group, it’s more of a hobby, not a main thrust—we’re dabbling in fluorine chemistry.”

Baran explains that during consulting trips to pharmaceutical companies he hears about how medicinal chemists sometimes struggle to add a CF3 in just the right place or put in a CF2H group.

“There is a fundamental desire for organic chemists to understand chemical reactivity with the potential for direct applications,” Baran says. “These forays into fluorine chemistry are a natural response of the academic community to what discovery chemists are saying is useful to them.”

Peter T. Cheng, a research scientist in metabolic diseases discovery chemistry at Bristol-Myers Squibb, says medicinal chemists don’t normally talk with fluorine chemists, although he and his colleagues do read their research papers. On the other hand, medicinal chemists do regularly interact with frontline organic synthetic chemists.

“When we talk with academic chemists, they very curiously ask about the synthetic challenges we are facing,” Cheng says. “They see problems out there that are of importance to everyone that they think they can solve.”

Medicinal chemists generally buy prefluorinated starting materials and then functionalize them further, Cheng notes. But they would prefer to run a few synthetic steps and then add fluorine later, he says.

“We are glad the organic chemists are working on fluorinations,” Cheng says. “Most fluorinations were previously done on unadorned or simply functionalized phenyl rings. But now direct fluorinations are possible at C–H bonds of complex functionalized phenyls and heteroaromatics, which are the truly useful building blocks for drugs. This new trend will ultimately allow us to add fluorine at will under mild conditions.”

Baran’s group now has two fluorine chemistry papers to its credit, and more are on the way. The first one, published last year, reports the trifluoromethylation of pyridines and other nitrogen-based heteroaromatics using CF3 radicals (Proc. Natl. Acad. Sci. USA, DOI: 10.1073/pnas.1109059108).

Fluorine chemists have been preparing and using CF3 radicals for decades, Baran notes. One reaction involves CF3I gas, which is inconvenient to handle and tends to be too harsh for functionally complex compounds, so medicinal chemists don’t like using it, he says.

Baran’s team makes CF3 radicals instead by using tert-butyl hydroperoxide as an oxidant to controllably decompose sodium trifluoromethylsulfinate, NaSO2CF3. This stable, inexpensive solid, known as Langlois reagent for Bernard R. Langlois of Claude Bernard University, in Lyon, France, has often been used for fluorination reactions.

The second paper, published this year, reports a similar process to add CF2H groups to nitrogen heterocycles by generating CF2H radicals from the reaction between tert-butyl hydroperoxide and the new reagent Zn(SO2CF2H)2 (J. Am. Chem. Soc., DOI: 10.1021/ja211422g). This zinc reagent is already available from Sigma-Aldrich.

Generating a CF2H group usually requires deoxyfluorinating reagents. These electrophilic reagents typically store fluorine in an N–F bond and convert carbonyl groups into CF2H or CF2 groups, and alcohols into CH2F groups. More than a half-dozen such reagents are commercially available. But they tend to be harsh and nonselective for discovery chemistry and thus not practical for fluorinations at later stages of a multistep synthesis. Adding a premade CF2H group in the manner Baran’s team has done was previously unheard of.

In a related development, last year David W. C. MacMillan of Princeton University led a team that made CF3 radicals from a light-induced reaction involving a ruthenium photocatalyst and trifluoromethanesulfonyl chloride, ClSO2CF3, a reagent shown to be useful for radical trifluoromethylations by Nobumasa Kamigata of Tokyo Metropolitan University. MacMillan and coworkers showed that the approach can add CF3 groups to already functionalized aryl groups in molecules such as ibuprofen (Nature, DOI: 10.1038/nature10647).

“The idea we are pursuing is to rapidly append fluorinated groups to diverse collections of compounds for discovery screening, rather than having to build fluorinated molecules one at a time,” Baran says. Baran and his coworkers want to do the reactions in open air at room temperature; using the cheapest, lowest toxicity reagents they can find or invent; using substrates with unprotected functional groups; and all in single-pot reactions. “Coffee cup chemistry—just dump in and stir,” Baran says.

Fluorine newcomer John F. Hartwig of the University of California, Berkeley, thinks the spark that ignited the flurry of fluorine activity came from pharmaceutical companies better articulating their unmet needs.

“Fluorine chemists have already developed methods to take simple fluorinated starting materials and then halogenate, aminate, and do whatever else needs to be done to create the hundreds of fluorinated molecules available in a chemical catalog,” Hartwig notes. “Those contributions are really important.”

But the growing dependence during the past 15 years on catalytic cross-coupling reactions, which are one of Hartwig’s specialties, has created some needs that can’t be met by the existing reagents or the methods to make them, he points out.

Chemists who want to create a collection of compounds, such as a group of drug candidates, might now start with an intermediate containing an aryl halide, Hartwig explains. The aryl group would already have been functionalized with other desired groups in previous steps. The researchers would next do a series of separate cross-coupling derivatization steps, such as Suzuki coupling, Negishi coupling, Heck reaction, C–N coupling, C–O coupling, ketone arylation, and other reactions, to make from that aryl halide intermediate dozens to hundreds of unique molecules to test.

“Discovery chemists also want to make sets of fluorinated analogs,” Hartwig says. “But they don’t want to revert to a chemical catalog and buy 10 or 20 different premade fluorine reagents and start each synthesis from scratch—they want to start from that same prefunctionalized aryl halide intermediate.

“That’s an organic synthesis need that’s different now than it was before people did so much cross-coupling,” Hartwig continues. “But we also have a parallel line of fluorine chemists who have been doing organofluorine chemistry for a long time. What we are now doing is helping make those parallel lines of interests intersect.”

To that end, Hartwig’s group has published two papers on copper-mediated fluorinations, and it plans to publish more. The work stems from mechanistic studies in the Hartwig lab and builds on the work of other research groups, including those of Donald J. Burton of the University of Iowa and Hideki Amii of Gunma University, in Japan. To date, the majority of metal-mediated fluorination reactions use copper reagents.

Hartwig’s team used (phen)CuCF3, a CuCF3 complex with a phenanthroline ligand, to trifluoromethylate a range of aryl compounds under mild conditions (Angew. Chem. Int. Ed., DOI: 10.1002/anie.201100633). The researchers prepare the copper reagent in high yield by sequentially adding 1,10-phenanthroline and then TMSCF3 to [CuOC(CH3)3]4. The phenanthroline ligand helps speed up the reaction and stabilizes CuCF3 against side reactions sparked by its decomposition to a difluorocarbene. The (phen)CuCF3 reagent is now made commercially by Catylix, a company Hartwig cofounded, and it’s available for sale in the Sigma-Aldrich catalog, he says.

Copper and the phenanthroline ligand are cheap, Hartwig says, and the reagent is easy to handle, thermally stable, and operates at room temperature or with mild heating, he notes. Hartwig’s contribution to copper-mediated chemistry is in showing that it can be used on a broad scope, adding perfluoroalkyl groups not only to aryl iodides as others have shown, but also to selected aryl bromides and aryl boronates, including those with electron-rich, electron-deficient, or sterically hindered substituents.

“Our reagent is something a medicinal chemist can pull off the shelf and drop into a round-bottomed flask or into 96 tubes in a rack,” Hartwig says. “Maybe the reaction doesn’t meet all the green chemistry needs. But it does meet the needs of medicinal chemists for milder and more versatile fluorinating reagents.”

For Thomas M. Stevenson, a research fellow in discovery chemistry at DuPont Crop Protection, the ability of Baran, MacMillan, Hartwig, and others to apply organic synthesis methods to fluorinations is enabling fluorine chemistry in ways not possible before.

“It’s true that we have traditionally bought our fluorine compounds and derivatized them because fluorine chemistry is hard to do, especially at the end of a synthetic sequence,” Stevenson says. “But in a perfect world we would like to have one intermediate that we can derivatize in many different ways. When optimizing a lead compound, we want to try CF3 in almost every position in the molecule.

“That is where this new chemistry comes in handy,” Stevenson continues. “Diversification at the end of a synthesis rather than at the beginning is a real time-saver and is transforming our efficiency—it is really fantastic for our throughput and our ability to carry out optimization programs. And the idea that you can do this in a nonspecialist way—really anyone can do this chemistry—makes our job easier.”

Stevenson says he and his colleagues are trying many of the newly reported methods. For example, he recently put two CF3 groups on a molecule in the last synthetic step, something not possible before. “For me it’s an exciting time to be a discovery chemist because we have the opportunity to make many molecules today that would have been difficult to make 25 years ago.”

Fluorine specialists themselves are taking inspiration from the transition-metal-catalyzed cross-coupling reactions. Among them is Feng-Ling Qing of the Chinese Academy of Sciences’ Shanghai Institute of Organic Chemistry (SIOC).

SIOC has a long history in fluorine chemistry, Qing relates, and it even has a separate organofluorine chemistry department with 10 research groups. The Shanghai area is home to many pharmaceutical and fine chemicals companies, and more are popping up across China, he notes. “Scientists at these companies often require fluorinated compounds, so they come to SIOC to discuss fluorine chemistry with us,” Qing says.

In response, Qing’s group recently developed oxidative trifluoromethylation reactions that use copper and TMSCF3 to introduce fluorinated groups into organic molecules. For example, his team has carried out trifluoromethylations of terminal alkynes via cross-coupling (J. Am. Chem. Soc., DOI: 10.1021/ja102175w), trifluoromethylations of aryl boronic acids via cross-coupling (Org. Lett., DOI: 10.1021/ol1023135), and trifluoromethylthiolations—SCF3 group additions—of aryl boronic acids (Angew. Chem. Int. Ed., DOI: 10.1002/anie.201108663).

Given the functional group tolerance, broad substrate scope, and mild reaction conditions of these methods, Qing believes they can be used to fluorinate highly functionalized compounds at the later stages of a synthetic sequence.

The influx of new methods has the added benefit of expanding the fluorine chemistry skill set of young synthetic organic chemists in industry, University of Florida’s Dolbier points out. “Many of the research chemists working at pharmaceutical and agrochemical companies come out of large academic organic synthesis groups and have no experience in fluorine chemistry,” Dolbier notes. What they know about fluorine, he says, comes from on-the-job training, including the fluorine chemistry short courses Dolbier teaches at pharma and agchem companies.

Even when it comes to process chemistry, when syntheses are scaled up for commercial production, the fluorine chemistry steps are usually farmed out to contract research organizations that specialize in fluorine chemistry. It is unusual for pharmaceutical or agrochemical companies to hire people with any experience with fluorine, Dolbier says. “But that may change now that some synthetic organic groups are working with fluorine.”

One scientist who is showing how fluorine expertise could be more widespread among synthetic organic chemists is Harvard University’s Tobias Ritter. Classically trained as an organic chemist, but with experience in organometallics and fluorine chemistry as well, Ritter has focused on fluorine chemistry during the first few years of his academic career. As several chemists tell C&EN, Ritter is a “human catalyst” helping the new wave of fluorine chemistry pick up speed.

Among the Ritter group’s achievements so far is a new imidazolium-based deoxyfluorinating reagent that transforms substituted phenols into aryl fluorides (J. Am. Chem. Soc., DOI: 10.1021/ja2048072). His team has also carried out silver-mediated reactions that use the commercial electrophilic fluorinating reagent Selectfluor to convert aryl tin compounds into aryl fluorides and aryl tin compounds and aryl boronic acids into trifluoromethoxy derivatives (J. Am. Chem. Soc., DOI: 10.1021/ja105834t and 10.1021/ja204861a).

Ritter and coworkers most recently devised a palladium-based electrophilic fluorinating reagent and used it in the late-stage synthesis of 18F-labeled aromatic compounds for positron emission tomography (PET) diagnostic imaging (Science, DOI: 10.1126/science.1212625). Preparing 18F compounds has always been precarious because of the 110-minute half-life of the isotope. The compounds are limited to simple molecules such as deoxyglucose. The new synthesis shows that much more diverse molecules can be used for PET.

The originality of Ritter’s work makes these new methods “real home runs,” Scripps’ Baran says. “Ritter has turned out to be the poster child for the resurgence in fluorine chemistry.” Last year, to begin commercializing some of these successes, Ritter founded Boston-based SciFluor Life Sciences.

When it comes to true catalytic fluorination reactions, however, the new fervor over fluorine has run into a wall. Although many researchers—fluorine chemists and nonfluorine chemists alike—have tried to develop mild catalytic fluorinations that work on complex molecules, success has been limited.

“Just five years ago a chemist would laugh at the idea of palladium-catalyzed nucleophilic trifluoromethylation of aryl halides,” Vladimir V. Grushin notes in a review of aromatic trifluoromethylations (Chem. Rev., DOI: 10.1021/cr1004293). Studies by multiple research groups since then have proven that the reaction is not impossible, but these researchers can’t celebrate using palladium-catalyzed cross-coupling chemistry in drug discovery chemistry just yet.

A former DuPont research scientist who is now at the Institute of Chemical Research of Catalonia, in Spain, Grushin has shown through his research dating back to the mid-1990s that cleanly executed reductive elimination of monofluorinated aryl groups from palladium intermediates is difficult to pull off. In reductive elimination, which is the last step of a catalytic cross-coupling reaction, the two molecular fragments bound to the metal catalyst are released and join to form the product.

Grushin found that standard approaches to palladium-catalyzed cross-coupling using tertiary phosphine ligands favor formation of P–F bonds with the phosphine ligand on the metal, rather than desired C–F bonds with aryl substrates. His group overcame the problem by using a bulkier ligand and eventually reported a breakthrough in 2006: the first aryl trifluoromethyl compounds prepared by palladium-mediated cross-coupling. The reaction showed that catalytic aromatic fluorinations were at least possible.

In 2009, fluorine newcomer Stephen L. ­Buchwald’s group at Massachusetts Institute of Technology built on those and other developments to design a palladium-catalyzed reaction to form C–F bonds on prefunctionalized aryl rings. Buchwald’s team used aryl triflates as the substrate and CsF as a fluorine source to make various monofluorinated aryl compounds (Science, DOI: 10.1126/science.1178239). In 2010, Buchwald’s team followed up with similar chemistry, using the ethyl analog of TMSCF3 as a CF3 source to achieve the first palladium-catalyzed aryl trifluoromethylations (Science, DOI: 10.1126/science.1190524).

Buchwald emphasizes that much credit goes to Grushin for pointing the way. His group’s success, Buchwald says, stems from using one of its trademark bulky biaryl phosphine ligands, named BrettPhos, to create an aryl palladium fluoride complex. BrettPhos appears to prevent formation of P–F bonds and formation of a bridging fluorine in palladium intermediates that can block reductive elimination, he explains.

The catalytic process still needs to be optimized. “There isn’t anything practical yet in terms of getting to scalable processes,” Buchwald says. The ligand is still too expensive, he notes, and the substrate scope needs to be expanded.

Even so, Buchwald is not sure metal-catalyzed reactions for discovery chemistry can compete with the new radical chemistry from the groups of Baran and MacMillan. Their systems are better geared toward discovery research, Buchwald notes. “But everyone is moving in the right direction,” he says. “Hopefully we will be able to kick our palladium fluorination work up a couple of notches so that it will be user-friendly.”

“In terms of achieving metal-mediated fluorine chemistry, we organic chemists are Johnny-come-latelies,” Buchwald admits. “But you see this type of discovery process happen in so many fields.

“People are thinking about the same idea; then if you get a slightly different take on it, the right student, and a little luck, you see the first successful paper,” Buchwald continues. “That’s why it’s important to have a lot of groups from different backgrounds working on topics such as catalytic fluorinations—inorganic chemists, synthetic organic chemists, fluorine chemists, and hybrids of the three.” ◾

Chemical & Engineering News
ISSN 0009-2347
Copyright © American Chemical Society