Fluorine Rises to the Occasion

Fluorine chemists are fond of pointing out that even though "their" element has one of the smallest atomic radii on the periodic table, it still has the greatest ability to attract electrons. These properties give fluorine the distinction of being called a lot of names, such as "the small atom with a big ego" or "the little atom that could." Another apt description that fluorine chemists have more recently added to their lexicon is that fluorine is ubiquitous.

Fluorine indeed has made major inroads in chemistry, moving beyond commonplace fluoropolymers to become important in other aspects of materials science and to function as an important substituent in agricultural chemicals and active pharmaceutical ingredients. According to University of Southern California chemistry professor G. K. Surya Prakash, chair of the 17th Winter Fluorine Conference, the diverse group of scientists from around the globe and the diversity of the lectures and discussions at this year's meeting "clearly affirmed the theme: 'Ubiquitous Fluorine: From Materials to Medicine.'"

More than 200 fluorine aficionados converged on St. Petersburg Beach, Fla., in January for the weeklong conference, which is sponsored by the American Chemical Society's Division of Fluorine Chemistry. Chemists have been gathering there every other year since the early 1970s in an informal atmosphere to catch up both with each other and with each other's latest research. This year, attendees were treated to sessions on organic synthesis, the role of fluorine in medicine, industrial fluorine applications, materials science, and inorganic chemistry.

Fluorine chemistry has moved away from the original approach of exhaustive fluorination of substrates and continues to shift in the direction of "lightly" fluorinated materials that require increasingly selective fluorination methods, noted Viacheslav A. Petrov, a DuPont chemist and conference cochair. "This is being dictated by the unique properties of fluorinated substituents, our better understanding of the effect of fluorination on the properties of organic substrates, and the continued realization that strategic placement of fluorine in a molecule quite often gives an amazing result," Petrov told C&EN.

Rapid advances in using fluorine as a substituent in pharmaceuticals led to a special feature at this year's conference: The organizers arranged a tutorial on the use of fluorine in drug design, which was given by James R. McCarthy of Eli Lilly. As several attendees pointed out, 10% of newly registered pharmaceuticals and 40% of new agrochemicals contain at least one fluorine atom.

McCarthy described the utility of incorporating a variety of fluorinated groups into marketed drugs. One benefit of adding a single fluorine to a drug is to improve stability of the compound by blocking or slowing its metabolism, he noted. One strategy is to replace a labile hydrogen with fluorine to prevent hydroxylation by cytochrome P450 liver enzymes. For example, the nonfluorinated version of AstraZeneca's anticancer drug Iressa (gefitinib), a tyrosine kinase inhibitor, has a half-life of one hour, McCarthy said. But adding a fluorine atom to a phenyl ring allows the drug to persist at high levels in the bloodstream for 24 hours.

McCarthy cautioned that sometimes a fluorine substituent can make a molecule too stable and present problems with dosing schedules. For example, a fluorine-substituted version of Pfizer's anti-inflammatory drug Celebrex (celecoxib), a COX-2 inhibitor, was found to be stable for days in rat studies. When the fluorine was replaced with a methyl group, the drug's half-life was reduced to about four hours in rats, which turned out to be 11 hours in humans.

Another reason to include fluorine is to modulate the lipophilicity of a drug. The resulting physical and chemical property changes to the drug can affect its ability to enter the bloodstream, cross cell membranes, or cross the blood-brain barrier. Yet another outcome of adding fluorine is to enhance a drug's binding affinity to a target protein.

"Fluorine will continue to have a major impact in the design of biologically active molecules," McCarthy concluded. However, selectively introducing fluorine also will continue to present headaches for process chemists in the pharmaceutical industry, who must take the ideal fluorinated active ingredients created by medicinal chemists and develop ways to produce them on a commercial scale, he added.

Building on McCarthy's theme, a series of talks at the conference focused on the synthesis of fluorinated amino acids. Because there are no known natural amino acids that contain fluorine--and few fluorinated natural products--introducing fluorine atoms into amino acids is a hot topic.

The ability to strategically place these fluorinated residues in proteins is expected to pay off with substantial effects on protein stability, protein-protein and ligand-receptor interactions, and the physical properties of protein-based materials, noted Beate Koksch, an organic chemistry professor at Free University Berlin, in Germany. An added benefit is that protein structure and binding can be studied by ¹⁹F nuclear magnetic resonance spectroscopy, she said. The modifications could prove useful to design novel enzymes and therapeutic proteins.

Koksch described her group's work to synthesize fluoroalkyl-substituted amino acids and use chemical and enzymatic methods to incorporate them into peptides and proteins. Her group has further developed "a fast and simple screening system" to investigate the interaction of the fluorinated residues with native amino acids in a polypeptide environment.

"The properties of fluoroalkyl groups, such as space filling, lipophilicity, the ability to take part in hydrogen bonding, and their interaction characteristics with native amino acid side chains, are still controversially discussed in the literature and remain to be investigated in a systematic manner," she told C&EN.

THE BERLIN researchers use an a-helical coiled-coil peptide system that consists of a pair of a-helical peptides that twist around each other. Earlier work by other groups has shown that substituting trifluoro- or hexafluoroleucine for leucine in the hydrophobic surface of an a-helical peptide enhances the thermal stability against denaturization and the structural stability of the resulting coiled coil.

In Florida, Koksch described two screening methods to measure the interactions of a range of fluorinated amino acids with native amino acid side chains in coiled coils [ChemBioChem, 5, 717 (2004)]. One screen uses the melting-point temperature of the coiled-coil dimers as a measure of stability. The second screen takes advantage of the property of a-helical coiled-coil peptides to self-replicate. The influence of the fluorinated amino acids can be determined from the ligation rate of the replicating system. The screens are "sensitive enough to detect the difference of just one fluorine atom in the side chain of the nonnatural amino acid within the hydrophobic environment of the coiled-coil peptide arrangement," she noted.

A third screening method under development is a phage display technique to find the best interaction partners for the fluorinated amino acids out of the pool of 20 natural amino acids. Phage display is a method for quickly evaluating a large number of potentially useful peptides by expressing them on the surface of a bacteriophage, a type of virus. Koksch's method involves displaying one part of the coiled coil on the phage surface and using the other part, synthesized with a terminal biotin molecule, for library screening. The results are pending, she said.

"Koksch has an ideal system for exploring fluorine effects in proteins," commented chemistry professor David O'Hagan of the University of St. Andrews, in Scotland, who attended the conference. The dimer is stable as a function of both steric interactions and its solubility. Thus the organofluorine substituents, which solubilize more readily than the organic substituents, should lead to more stable dimers. "Exciting results have emerged and will continue to emerge from this type of analysis," he added.

A GRAND CHALLENGE for fluorine chemists has been to develop syntheses and fluorinating reagents to impart fluorine to a desired substrate. In an invited lecture, emeritus chemistry professor Richard D. Chambers of the University of Durham, in England, discussed the nuances of using elemental fluorine gas as a reagent. Although it's one of the simplest fluorinating reagents, there is nothing simple about using it, he said.

Direct F₂ reactions with organic substrates are highly exothermic and can be difficult to control, Chambers noted. For safety reasons, F₂ typically is diluted in nitrogen, and special equipment and handling procedures are needed. During the past 20 years, a number of milder fluorinating reagents have become commercially available and provide an alternative to F₂, he said, in particular the so-called N-F class of compounds, where the fluorine is bound to an amino or pyridine nitrogen atom. These reagents are easier and safer to handle, Chambers added, but they still must be prepared from elemental fluorine and they are relatively expensive.

For that reason, some chemists believe it would be more efficient and environmentally friendlier to find a better way to manage using F₂, he added. Chambers went on to describe the efforts of the fluorine research group at Durham, which is headed by senior lecturer Graham Sandford, to better control direct fluorinations by designing and building microreactors.

"These reactors have multiple microchannels that are supplied from single feedstock sources, which is ideal for a small-scale lab but also can be extended easily and inexpensively to a manufacturing scale," Sandford told C&EN. The small size of the channels minimizes the amount of fluorine in the reaction zone, which is a bonus for safety. "There also are well-established benefits of using continuous-flow processes rather than batch processes, such as no downtime," he added.

The reactors are made by cutting 0.5-mm-wide parallel channels in a stainless steel sheet that is about 20 cm long, Chambers explained. So far, the team has made reactors with up to 30 channels. The channels are covered with a transparent polychlorotrifluoroethylene plastic sheet and a sealing plate, and this module is connected to a stainless steel block. The block has reagent reservoirs drilled into it, with additional holes drilled in such a way that the reactants can simultaneously enter all the reactor channels. The reactor core is surrounded by a heat exchanger, and during operation the product is collected in a reservoir at the end of the channels [Lab Chip, 5, 191 (2005)].

The chemists have tested the devices by reacting 10% F₂ in nitrogen with ethyl acetoacetate in formic acid, which forms several mono- and difluorinated products. By varying the reactant ratio and flow rates, the yield of one of the monofluorinated products, ethyl 2-fluoro-3-oxobutanoate, was improved to 81%. This yield compares favorably to the 62% yield the researchers obtain running the same reaction in a traditional bulk reaction. The yield per channel is about 0.2 g per hour, but with 30 channels operating in parallel on one microreactor, the Durham chemists can produce 100 g of product during an overnight run.

"The key advance is the demonstration that scale-out--rather than scale-up--to many microchannels in a simple, efficient manner is possible using devices fabricated from readily available materials and using construction techniques available in most mechanical workshops," Sandford noted. Plus, "fluorine really is a 'green' reagent," Chambers said, "considering the only by-product, HF, can be recycled." The University of Durham has patented the technology and licensed it to Japan's Asahi Glass Co.

Scientists who regularly attend the Winter Fluorine Conference often take the opportunity to update work they described in earlier years. One example this year involved three DuPont scientists who provided an update on the company's development of fluoropolymer photoresists used to make computer chips. Petrov and his colleagues Andrew E. Feiring and William B. Farnham each made a presentation on different aspects of the project.

Photolithographic processes currently use 248-nm light to chemically modify photoresists, the polymeric material deposited on top of silicon wafers to form circuit patterns, Feiring explained. But the semiconductor industry is now transitioning to 193-nm light to create narrower circuit lines for the next generation of chips. After that, the next generation of smaller feature sizes may require 157-nm light.

Each wavelength of light requires a different polymer photoresist, in part so the light can penetrate all the way into the material, he said. Nonfluorinated polymers developed for 248- and 193-nm technology absorb too much light near the surface at 157 nm, Feiring added, so researchers worldwide have been focusing on developing new polymers that contain fluorinated groups or segments, which allow good transparency at the shorter wavelength.

At the Winter Fluorine Conference two years ago, the DuPont researchers and scientists from Asahi Glass and IBM described their approaches to making fluorinated photoresists (C&EN, March 3, 2003, page 44). DuPont's lead polymer at the time was a terpolymer of tetrafluoroethylene, norbornene with a hexafluoroisopropanol substituent, and tert-butyl acrylate, Feiring noted. Each of these components is needed to perform different functions in the photoresist.

The fluoropolymer and fluorinated alcohol group help impart transparency, and the norbornene helps provide etch resistance, he said. The acrylate is converted to a carboxylic acid by protons generated when the resist is exposed to light, which helps in pattern development. But the acrylate causes some undesirable light absorption at 157 nm. The DuPont team recently found that adding a second hexafluoroisopropanol group to the norbornene component "allows elimination of the pesky acrylate, leading to nearly an order of magnitude improvement in transparency," Feiring said.

DuPont had expected that the new polymers would be commercially available before 2006. But Feiring disclosed that prospects for 157-nm imaging have dimmed. "A process called immersion lithography may allow the industry to use 193-nm imaging to create the features previously assumed to require 157-nm light," he said. This would result in a significant savings in equipment costs for semiconductor manufacturers. Consequently, DuPont is now targeting its fluoropolymer photoresist effort on 193-nm imaging.

Fluorine isn't required for transparency at 193 nm, Feiring said, but the good news for DuPont is that modified versions of its 157-nm resists should work well at the longer wavelength. The new polymers the DuPont scientists described contain several acrylates, which can be tolerated at 193 nm, and may offer significant advantages over recently commercialized all-methacrylate polymers designed for 193-nm light. DuPont is working with a major resist supplier to possibly commercialize a fluorinated 193-nm resist within the next year.

In a meeting as diverse as the Winter Fluorine Conference, few presentations universally stand out. But a talk on the preparation of anhydrous tetrabutylammonium fluoride (TBAF) by associate chemistry professor Stephen G. DiMagno of the University of Nebraska, Lincoln, impressed many conference attendees.

SOME OF THE simplest fluorinating reagents are anhydrous organic fluoride salts, such as phosphonium or tetraalkylammonium compounds, where the "naked" fluoride ion acts as a nucleophile. These reagents are prepared as hydrated compounds, and the water is subsequently driven off by heating under a vacuum or by azeotropic distillation. One limitation is that the organic cations can decompose. For example, TBAF tends to degrade to bifluoride (HF₂^–) and tributylamine. Thus, a decade ago it was claimed that pure anhydrous tetraalkylammonium fluoride salts likely had never been produced.

DiMagno and research assistant professor Haoran Sun have shown that pure anhydrous TBAF can, in fact, be made by a low-temperature nucleophilic aromatic substitution reaction in an aprotic solvent. The researchers treat hexafluorobenzene with tetrabutylammonium cyanide in tetrahydrofuran, acetonitrile, or dimethyl sulfoxide solvent, which results in a moderately exothermic reaction to form TBAF and hexacyanobenzene in better than 95% yield [J. Am. Chem. Soc., 127, 2050 (2005)].

The reaction required careful choice of the nucleophile, DiMagno noted, because the C–F bond in hexafluorobenzene is exceptionally strong. It turns out that only diffusely charged anionic nucleophiles capable of forming a strong bond to carbon will work, and cyanide ion turned out to be an excellent candidate.

"The approach is ingenious," commented Suzanne T. Purrington, an emeritus professor at North Carolina State University who attended the meeting. "There is never an opportunity for water to contaminate the reagent," she said, adding emphatically that the hexacyanobenzene even scavenges any adventitious water to form the strong acid pentacyanophenol. Purrington received the Fluorine Division's Distinguished Service Award at the conference.

DiMagno said the anhydrous TBAF can be prepared in situ at room temperature and used without isolation or purification, although it's stable for weeks at low temperature. The researchers have used the reagent for direct fluorination of several substrates, such as converting primary alkyl halides and tosylates (RCH₂X) to fluoromethyl groups (RCH₂F). They also have used it in aromatic halogen exchange, an important industrial route to make fluorinated aromatic building blocks. Their results show that the anhydrous TBAF is comparable to or exceeds the reactivity of other nucleophilic fluorinating reagents. The University of Nebraska has filed a patent on the technology.

So it went at another Winter Fluorine Conference. Santos Fustero of the University of Valencia, in Spain, summed up the spirit of fluorine chemists with a quote as he closed out his talk on the synthesis of fluorinated amino acids. He highlighted a passage in a review paper from several years ago by chemistry professor Manfred Schlosser of the University of Lausanne, in Switzerland: "Fluorine leaves nobody indifferent; it inflames emotions, be that affections or aversions. As a substituent it is rarely boring, always good for a surprise, but often completely unpredictable."