Background and Summary: PFAS (per- and polyfluoroalkyl substances) are also called “forever chemicals”. The resistance of PFAS to biodegradation is in part related to the thermodynamic strength of C–F bonds (110-130 kcal/mol), but also because polyfluorinated compounds are inherently foreign to biology, and evolutionary adaption over 3.6 billion years of life on earth has occurred in their absence. As a result, these compounds can exist for hundreds to thousands of years, and harmful levels of PFAS can be found in 31 percent of groundwater samples remote from any obvious source of contamination. An example of PFAS in everyday life can be the pots and pans we use, which contain coatings with this material, such as Teflon™, and they work well under harsh conditions due to their chemical inertness. However, PFAS compounds have accrued in the environment as a persistent contaminant, where they pose a threat to human health, and have been linked to cancer and changes in hormone regulation. 

In this article published in Nature, a collaboration between the Miyake and Paton laboratories at Colorado State University and the Damrauer group at the University of Colorado Boulder, a new photocatalytic system is described that generates a powerful photo-reductant (-3.64 V vs SCE) capable of cleaving all of the C–F bonds in a variety of PFAS compounds. The paper, “Photocatalytic C‒F bond activation in small molecules and polyfluoroalkyl substances,” describes synthetic studies along with photophysical and computational mechanistic studies to develop and understand this new catalytic protocol, the result of which can be seen through aryl and alkyl defluorination reactions, cross-couplings, and the ability to degrade fluorinated polymers. A key discovery was that the cyclic imide in the pre-catalyst undergoes hydrolytic ring-opening, route to forming a stable dianionic Meisenheimer-like complex that acts as a potent reductant after it is photoexcited. 

Over the past decade, photoredox catalysis has become a valuable tool for both small molecule transformations as well as polymerizations. A particular challenge we aim to address with this work is the development of robust organic photocatalysts, which have a lower environmental cost for scalability compared to rare earth metal-based photocatalysts. This is one of the central goals of the newly formed NSF Center for Sustainable Photoredox Catalysis (SuPRCat), involving researchers at several universities. 

Scientific Breakthrough Built on a Strong Foundation: In 2020 the Miyake group demonstrated a modified benzo[ghi]perylene monoimide (BPI) catalyst in the presence of visible light and sacrificial electron donors, was able to achieve reduction potentials low enough to reduce benzene – a challenging feat due to a reduction potential of -3.42 V vs SCE.¹ This discovery provided a more sustainable and safer approach to the Birch reduction, avoiding the dissolving of alkali metals in ammonia. The impact of this work set the stage for a deeper understanding and development of organic dyes functioning as highly reducing photocatalysts in single electron reductions. 

A breakthrough occurred in this work when we were able to reduce simple fluorinated compounds such as fluorobenzene using a modified BPI photocatalyst. After optimization, we were able to apply the reactivity towards the hydrodefluorination of PFAS and other fluorinated polymers. However, we sought to understand how the catalyst functioned on a deeper level to expand the reactivity to not just hydrodefluorination, but also as a tool for cross-coupling reactions such as heteroarylations, which are commonly performed using transition metals.

To understand the mechanism, we tracked the nature of the catalyst. While a cyclic imide is the starting structure of the catalyst, we found that this undergoes ring-opening under the reaction conditions. We were able to characterize the ring-opened catalyst spectroscopically (¹H-, ¹³C-, 2D-NMR, and UV-vis) and by studying the mechanism computationally. This form of the catalyst, however, also showed no quenching in the presence of fluorobenzene, indicating that further reduction is required. We found that irradiation in the presence of excess nBu₄NF results in the formation of a radical dianion. The presence of alcohol also proves to be critical for this electron-transfer to occur: with DFT calculations we found that clusters of tAmylOH and fluoride can be oxidized at ca. 1 V vs SCE. Further, the role of alcohol is also essential for subsequent H-atom transfer to the catalyst structure, forming the key dianionic singlet species.

The excited state of this dianion is strongly reducing, with a reduction potential up to –3.64 V vs. SCE. ¹H-NMR experiments support the formation of an air stable intermediate, for which photoluminescence quenching is observed with aryl fluorides. We found a non-linear power dependence of aryl fluoride conversion on the irradiation density, which is consistent with a closed-shell mechanism involving an overall multi-photon excitation process.The discovery of multiple elementary steps to convert the pre-catalyst into the active super-reducing catalyst involved multiple experimental, analytical and computational investigations, and revealed that this system was far more complex than what we initially had hypothesized.

Reflections: Earlier this year, the same groups also worked together to highlight the limitations of the BPIs, and developed a model system to help understand the catalytically active species. A breakthrough discovery showed that the model catalyst underwent a two electron/one proton activation to generate a closed shell species which explained the highly reductive nature of the photocatalyst.² By applying what we had all learned on the mechanistic side of a similar system, coupled with the discovery of the defluorinations using BPIs, we were able to realize the full potential of this organo-catalytic system towards solving a global environmental problem in breaking down PFAS. We are very grateful to have interdisciplinary skillsets come together in this developing area of chemistry and are continuing to explore the capabilities of highly reducing photocatalysts.

References:

1. Cole, J. P.; Chen, D. F.; Kudisch, M.; Pearson, R. M.; Lim, C. H.; Miyake, G. M. Am. Chem. Soc. 2020, 142, 13573-13581. DOI: 10.1021/jacs.0c05899

2. Sau A.; Pompetti,, N. F.; Green, A. R.; Popescu, M. V.; Paton, R. S; Miyake, G. M.; Damrauer, N. H. ACS Catal. 2024, 14, 2252–2263. DOI: 1021/acscatal.3c05386