Finding a More Efficient Method for Forming Carbon-Fluorine Bonds
Dr. Melanie Sanford presents her team’s new research
On Thursday, Feb. 4, Professor of Chemistry at the University of Michigan, Ann Arbor, Dr. Melanie Sanford, gave a seminar describing her work on the formation of Carbon-Fluorine (C-F) bonds. C-F bonds are incredibly useful to synthetic chemistry, as they are found nowhere in nature and see applications from agriculture to pharmaceuticals. As can be expected from a bond never found in nature, C-F bonds are incredibly hard to make.
For a long time, the only process was nucleophilic aromatic substitution. Under high temperatures, and very slowly, you can replace a chlorine atom that is bonded to a carbon ring called an aryl group with a fluorine atom, forming an aryl-fluoride. This reaction is hard to do and only works with a very limited set of aryl groups, and didn't work with a variety of R groups. An R group is an additional attachment to the aryl group, and being able to work with a large variety of R groups would let a C-F bonding reaction be used to fluoridate a wider range of molecules. Sanford was doing research on how to form aryl-fluorides without using this difficult reaction. Her specialty is metallic inorganic chemistry, so she was looking at using a metal catalyst that could bind to both the aryl group and the fluorine, then be removed to bind the fluorine and aryl to each other.
A researcher named Vladymir Grushin had been working for 50 years trying to make this reaction happen and finally published a 2010 paper saying that it couldn't be done; successfully removing the metal so that the aryl-fluoride formed was impossible. In 2009, however, researcher Steven Buchwald was able to do it by binding a very big and fancy molecule as a ligand, a molecule that binds to a metal. By adding this ligand to a palladium atom, he was able to catalyze the synthesis of an aryl-fluoride group successfully without using nucleophilic aromatic substitution. Sanford's team was working with really similar reactions, and one of her students, Nick Ball, was able to forego the fancy ligand and form the aryl-fluoride by oxidizing the palladium to a higher oxidation state, increasing its reactivity and allowing it to bind the flouride to the aryl group.
“If you talk to them, they’ll tell you, the people at Dow, that I basically tried to fire them.”
At this point, Sanford was approached by Dow Chemical. Dow was manufacturing herbicides that contained aryl-fluoride groups and wanted a cheaper, more efficient way to produce these bonds, so Sanford's work was very interesting to them. Dow liked the reaction they had found with the oxidized palladium group, but palladium was very cost prohibitive to use; they wanted a cheaper catalyst at all. Sanford's team got to work using copper because it was much cheaper than palladium, but still close to it on the periodic table, and there was some literature on copper forming C-F bonds in some very specific conditions.
One of Sanford's students, Yinda Ye, was able to find a combination of an aryl source, an oxidizing fluorine source, and a copper source that was able to produce aryl-fluorides with a wide variety of R groups under very mild conditions. Dow was not excited. Even though Sanford's team had lowered costs by using copper instead of palladium, the oxidizing fluorine source was prohibitively expensive. Ye and another of Sanford's students, Naoko Ichiishi, created more reactions that worked with a variety of R groups, used mild conditions, and each reaction continued to lower cost, yet all used at least one reagent that was cost prohibitive on an industrial scale.
At this point, Sanford was frustrated; she felt that her students' work wasn't being respected when Dow kept calling their papers "uninteresting." She said in her seminar that, "If you talk to them, they'll tell you, the people at Dow, that I basically tried to fire them." She told them, "I don't want to be funded by you anymore unless you are going to tell us what you actually want." Dow told Sanford exactly what they wanted. Dow didn't care about having the ability to use lots of different R groups, they wanted one specific R group; they didn't want to be able to use lots of different starting materials, they had one specific molecule that they wanted to fluorinate, and they didn't want to use any metal catalyst. This was new to Sanford, she admits maybe in part because of her lack of attention in her first meetings with Dow, but she said it was helpful to have the new focus. What was confusing to Sanford was that Dow said they could already make that reaction happen, with the exact starting material they wanted, the addition of the flouride, and no metal catalyst. It used the old method of fluorine bonding that Sanford showed at the beginning of her lecture, nucleophilic aromatic substitution.
Dow used cesium fluoride as their fluorine source, which is rare and expensive, and at the scale Dow wanted, there may not be enough cesium on Earth to carry it out. The cesium reaction took up lots of time, lots of heat, and created side products that reduced the product's purity and yield. Potassium fluoride (KF) would be ideal, but the yield they got when using KF was a paltry 5%. Sanford found this new problem uninteresting, but at least straight forward.
After investigating the reaction, Sanford saw that one of the biggest problems with using KF was that it wasn't very soluble in the organic solvents that had to be used for the reaction. Sanford's first approach to the problem was to find a reagent or catalyst that would help the KF dissolve. One of Sanford's students, Laura Allen, was developing and testing all sorts of fancy catalysts to help the KF dissolve and none of them were working. Catalysts are normally used in very small amounts relative to the other reagents, but for one experiment, Allen used a large amount of catalyst, just as much catalyst as there was reagent, and got a good yield (85%) from her reaction. This was the first paper that really excited Dow, as they found that this approach would bring down costs on an industrial scale.
This reaction wasn't perfect. It required high temperatures, its yield wasn't incredible, and there were still some undesirable side products forming. Sanford's team noticed that many of the side products formed because of the heat the reaction had to be kept at. They figured that by lowering the heat of the reaction, they could get rid of the side products that were reducing their yield.
The catalyst they were using to dissolve the KF underwent a reaction into a slightly different molecule, and Sanford thought that if they could skip the original catalyst and add purely the molecule it turns into, they could lower the heat of the reaction. The problem is that this product molecule is more or less impossible to create on its own outside of reactions like the one Sanford was unhappy with. Instead, Sanford's team found a very similar molecule that could be acquired, and in fact synthesized by Dow, very easily and at a low cost. Two of Sanford's students, Sarah Ryan and Megan Cismesia, tried the reaction, and it worked! The reaction used cheap materials, worked at room temperature, and gave 99% yield without side effects.
In the subsequent paper, Sanford's team saw that their new reaction worked for a variety of starting materials, but only ones very similar to the one Dow had wanted the reaction to work on. This was great for Dow, but Sanford's team wanted to find a better, more general reaction.
The basic problem with nucleophilic aromatic substitution was that when adding the fluorine, they created an intermediate stage where both the flourine and the chlorine it was replacing were bonded at the same time. This concentrates lots of negative charges on the aryl group and stresses it. Only aryl groups with certain R groups could withstand these stresses. Sanford's team theorized that if they could avoid having a fluorine and another electronegative atom both bonded to the aryl at the same time, they could use a wider variety of R groups. By using an aryl with a sulfate bonded to it as the base material, Sanford theorized that they could push the fluorine into the aryl group, and the intermediate stage would never involve both the flourine and the sulfate bonded to the aryl at the same time. This would also be beneficial to Dow because aryl sulfates would be even cheaper and more readily synthesized than the ones used in Sanford's last reaction. After trying this reaction, Sanford could make a wide variety of C-F bonded products, including ones that were impossible to synthesize with nucleophilic aromatic substitution.
Sanford's aryl-sulfate was structurally similar to another molecule, aryl-triflate, the same molecule Steven Buchwald had used in his palladium catalyzed reactions. By using a different mechanism, Sanford was able to do this reaction that had previously required metal catalysis, fancy ligands, and high temperatures without any of those expensive complications. This was what Dow had wanted, and Sanford’s team had previously believed would never work.