Research

OverviewNi Cross CouplingNucleophilic FluorinationPhotocatalysisMachine Learning

Doyle Lab Research 

Ni-Catalyzed Cross Coupling

I. Methodology Development

The field of Ni-catalyzed cross coupling has undergone rapid growth in recent years owing to the low cost of Ni, its earth abundance, and its ability to promote unique cross coupling reactions. The Doyle group is interested in developing novel methods for Ni-catalyzed cross couplings that enable direct access to important bioactive motifs. We have developed a number of transformations that leverage the reactivity of unconventional electrophiles, including iminium ions, oxocarbenium ions, and aziridines. These electrophiles (and their precursors) are abundant, stable, and versatile intermediates in synthetic organic chemistry. Additionally, iminium and oxocarbenium ions are prochiral, making them attractive starting materials for asymmetric catalysis. Recent work in the group has centered on developing cross-electrophile couplings, such as a cross coupling of benzylic acetals with aryl iodides, a three-component C–C bond-forming reductive amination from in situ generated iminium ions, and an asymmetric coupling of styrenyl aziridines with aryl iodides. Ongoing research is focused on elucidating the various pathways by which C–C bond formation can occur, expanding the scope of coupling partners for these reactions, and both developing and understanding ligand platforms that can engender high chemo- and stereoselectivity.

Selected References:

  1. Arendt, K. M.; Doyle, A. G. “Dialkyl Ether Formation via Nickel-Catalyzed Cross Coupling of Acetals and Aryl Iodides.” Angew. Chem. Int. Ed201554, 9876-9880. [DOI: 10.1002/anie.201503936]
  2. Lutz, J. P.; Chau, S. T.; Doyle, A. G. “Nickel-catalyzed enantioselective arylation of pyridine” Chem. Sci. 2016, 7, 4105-4109. [DOI: 10.1039/C6SC00702C]
  3. Woods, B. P.; Orlandi, M.; Huang, C.-Y. Sigman, M. H.; Doyle, A. G. “Nickel-Catalyzed Enantioselective Reductive Cross-Coupling of Styrenyl Aziridines” J. Am. Chem. Soc. 2017139, 5688-5691.  [DOI: 10.1021/jacs.7b03448]
  4. Heinz, C.; Lutz, J. P.; Simmons, E. M.; Miller, M. M.; Ewing, W. R.; Doyle, A. G. “Ni-Catalyzed Carbon-Carbon Bond-Forming Reductive Amination” J. Am. Chem. Soc. 2018, 140, 2292-2300. [DOI: 10.1021/jacs.7b12212]

 

II. Catalyst and Ligand Development for Ni

Whereas recent advances in the field of Pd-catalyzed cross coupling have been driven largely by ligand and precatalyst design, related studies for Ni have received less attention. Modern ligand frameworks designed for Pd and other precious metal catalysts (Rh, Ir) have been largely unsuccessful when applied to Ni. To address this need, our lab has developed ligands and precatalysts for Ni and demonstrated that they enable new methodologies not previously possible with existing ligand frameworks. We recently reported a Ni precatalyst, (TMEDA)Ni(o-tolyl)Cl, that has seen broad adoption in the community as it is commercially available, air-stable, inexpensive, and modular with respect to ligand and reaction class.

We have also developed a novel class of phosphines containing bulky aryl groups with substitution at the 3,5-positions that confer high activity upon Ni catalysts for the Suzuki coupling of benzylic acetals. Using parameters to quantify phosphine steric and electronic properties together with regression statistical analysis, we identified that the effectiveness of these ligands is a function of remote steric hindrance, a structural concept relatively unexplored in ligand design. We are continuing to explore the development, mechanistic elucidation, and application of this ligand class across a range of Ni-catalyzed coupling reactions. Our group has also recently developed and investigated electron deficient olefin (EDO) ligands as a tunable framework for challenging Ni-catalyzed cross-coupling reactions. We identified a library of sultam-derived EDOs that enable Ni-catalyzed Negishi cross coupling of 1,1-disubstituted aziridines to forge quaternary carbons with minimal generation of β-hydride elimination side products.

Collaborators: Sigman Group (University of Utah), Bristol-Myers Squibb

Selected References:

  1. Shields, J. D.; Gray, E. E.; Doyle, A. G. “A Modular, Air-Stable Nickel Precatalyst.” Org. Lett. 2015, 17,2166−2169. [DOI: 10.1021/acs.orglett.5b00766]
  2. Huang, C.-Y.; Doyle, A. G. “Electron-Deficient Olefin Ligands Enable Generation of Quaternary Carbons by Ni-Catalyzed Cross Coupling.” J. Am. ChemSoc. 2015, 137, 5638−5641. [DOI: 10.1021/jacs.5b02503]
  3. Wu, K.; Doyle, A. G. “Parameterization of phosphine ligands demonstrates enhancement of nickel catalysis via remote steric effects” Nature Chem. 20179, 779-784. [DOI:10.1038/nchem.2741]

Nucleophilic Fluorination

An expansive array of medicines, agrochemicals, and materials contain fluorine due to the unique chemical properties that the element confers on organic molecules. One of the chief obstacles towards the discovery and production of these compounds is the availability of synthetic methods for carbon–fluorine (C–F) bond formation. The most abundant and inexpensive fluorine sources, nucleophilic fluoride salts, typically suffer from low solubility, high hygroscopicity, and strong Brønsted basicity, rendering them recalcitrant reagents for chemical synthesis. An ongoing program of research in the Doyle laboratory is focused on the invention of novel reagents and catalytic strategies to address these limitations. Our group has disclosed the first catalytic and asymmetric catalytic aliphatic fluorination reactions using nucleophilic fluoride reagents, enabling transformations such as enantioselective fluoride ring-opening of epoxides and aziridines, as well as asymmetric allylic fluorination. Not only are the products of these reactions synthons of high value for drug development, but the reactions themselves are characterized by mild conditions and ease-of-operation. We have also invented new reagents to accomplish nucleophilic fluorination such as PyFluor, which is now commercially available as a stable and low-cost deoxyfluorination reagent that fluorinates a broad range of alcohols without substantial formation of elimination side products.

Positron emission tomography (PET) is an imaging modality that enables the visualization of dynamic biological processes at the molecular or cellular level. As one of the most exciting and rapidly growing areas of science, molecular imaging has become an indispensable tool in research and the clinic. However, a roadblock for PET is chemically “tagging” small-molecule tracers with a radioisotope, namely 18F, in order to perform imaging studies. Due to the poor reactivity of fluoride salts such as KF, rates of fluoride incorporation into a tracer are often not competitive with decay of the radioisotope. An additional challenge is that most methods for 18F incorporation fail in the presence of the functionality found in bioactive molecules, but 18F incorporation prior to functional group installation is not possible due to the radioisotope’s short half-life. A goal of the Doyle laboratory is to develop mild, general, robust, and rapid methods for late-stage incorporation of 18F into small molecules. We have translated our methods for Co-catalyzed fluoride ring-opening of epoxides and Cu-catalyzed insertion of HF into diazo carbonyls into mild and robust protocols for radiofluorination. These methods deliver access to clinically validated and experimental PET tracers with fluoride incorporation in the final step.  Furthermore, building upon our discovery of the deoxyfluorinating agent PyFluor, we have identified the first no-carrier-added deoxy-radiofluorination.

Collaborators: Merck, Bristol-Myers Squibb

Selected References:

  1. Kalow, J. A.; Doyle, A. G. “Enantioselective Ring Opening of Epoxides by Fluoride Anion Promoted by a Cooperative Dual Catalyst System.” J. Am. Chem. Soc2010132, 3268–3269. [DOI: 10.1021/ja100161d]
  2. Katcher, M. H.; Sha, A.; Doyle, A. G. “Palladium-Catalyzed Regio- and Enantioselective Fluorination of Acyclic Allylic Halides.” J. Am Chem. Soc. 2011133, 15902–15905. [DOI: 10.1021/ja206960k]
  3. Nielsen, M. K.; Ugaz, C. R.; Li, W.; Doyle, A. G. “PyFluor: A Low-Cost, Stable, and Selective Deoxyfluorination Reagent.” J. Am. ChemSoc. 2015, 137, 9571−9574. [DOI: 10.1021/jacs.5b06307]
  4. Gray, E. E.; Nielsen, M. K.; Choquette, K. A.; Kalow, J. A.; Graham, T. J. A.; Doyle, A. G. “Nucleophilic (Radio)Fluorination of α-Diazocarbonyl Compounds Enabled by Copper-Catalyzed H–F Insertion” J. AmChemSoc. 2016, 138, 10802−10805. [DOI: 10.1021/jacs.6b06770]

Photocatalysis with Ni

I. Ni and Photoredox Catalysis

Dating back to the 1970s and Kochi’s seminal mechanistic studies, Ni catalysts have been identified as both capable of generating radicals from organic halides as well as functionalizing radicals towards the construction of new bonds. We questioned whether it might be possible to separate these two roles of Ni, to capture radicals that have been generated by alternate means. This idea and its recent realization has inspired major advances in the fields of Ni-catalyzed cross coupling and radical chemistry. Numerous classical and modern methods are available for radical generation from abundant, inexpensive and stable feedstocks, however these substrate classes are not often compatible with Ni-catalyzed cross coupling. Our group, alongside the MacMillan and Molander groups, first demonstrated that the merger of nickel and photoredox catalysis can enable a general platform for the cross coupling of abundant carbon sources (C(sp3)–CO2H, C(sp3)–H, C(sp3)–O). By harnessing the energy of visible light with a photoredox catalyst, carbon-centered radicals may be generated under exceptionally mild conditions and subsequently trapped by a nickel catalyst for C(sp3) bond formation. Thus far our lab has reported C(sp3)–H functionalization of cyclic and acyclic anilines and ethers. Current work in the lab seeks to develop new nickel/photoredox-promoted reactions for C–C bond formation with unactivated C(sp3)–H, C(sp3)–O, and C(sp3)–N bonds. Another fundamental goal is to develop strategies to control the site- and stereoselectivity of radical functionalization with nickel via ligand design.

Collaborators: MacMillan Group (Princeton), Rovis Group (Columbia University), Merck, Celgene

Selected References:

  1. Zuo, Z.; Ahneman, D.; Chu, L.; Terrett, J.; Doyle, A. G.; MacMillan, D. W. C. “Merging photoredox with nickel catalysis: Coupling of α-carboxyl sp3-carbons with aryl halides.” Science 2014, 345, 437-440. [DOI: 10.1126/science.1255525]
  2. Joe, C. L.; Doyle, A. G. “Direct Acylation of C(sp3)−H Bonds Enabled by Nickel and Photoredox Catalysis” Angew. Chem. Int. Ed. 2016554040-4043. [DOI: 10.1002/anie.201511438]
  3. Shields, B. J.; Doyle, A. G. “Direct C(sp3)−H Cross Coupling Enabled by Catalytic Generation of Chlorine Radicals” J. Am. Chem. Soc. 2016138, 12719−12722.  [DOI: 10.1021/jacs.6b08397]
  4. Stache, E. E.; Rovis, T.; Doyle, A. G. “Nickel-photoredox catalyzed enantioselective desymmetrization of meso cyclic anhydrides” Angew. Chem. Int. Ed. 2017563679-3683. [DOI:10.1002/anie.201700097]

 

II. Ni Photochemistry and Photophysics

To date, little is known about the photophysics of organometallic Ni complexes relevant to cross coupling, despite their prevalence as intermediates in Ni/photoredox applications. In a collaboration with the Scholes research group, we explored the photophysics and photochemistry of the Ni(II) aryl halide complexes that are common to Ni/photoredox reactions. Computational and ultrafast spectroscopic studies revealed that these complexes feature unexpectedly long-lived 3MLCT excited states. Given the broad interest in the development of sustainable photocatalysts using earth-abundant first-row transition metals, this finding implicates Ni as an underexplored alternative to precious metal photocatalysts. Moreover, we have shown that 3MLCT Ni(II) engages in bimolecular electron transfer with ground-state Ni(II), which enables access to Ni(III) in the absence of external oxidants or photoredox catalysts. As such, we have shown that it is possible to facilitate Ni-catalyzed C−O bond formation solely by visible light irradiation, thus representing an alternative strategy for catalyst activation in Ni cross coupling reactions.

Collaborators: Scholes Group (Princeton)

Selected Reference:

  1. Shields, B. J.; Kudisch, B.; Scholes, G. D.; Doyle, A. G. “Long-Lived Charge Transfer States of Nickel(II) Aryl Halide Complexes Facilitate Bimolecular Photoinduced Electron Transfer” J. Am. Chem. Soc. 2018140, 3035-3039. [DOI: 10.1021/jacs.7b13281]

Machine Learning for Reaction Prediction

Due to the multidimensionality of chemical reactivity and structure, vast resources and time are currently expended on the translation of reported catalytic chemistry to new chemical entities. Reports of catalytic reactions often include a brief survey of the dimensions under study and substrate scope. For the sake of simplicity in interpretation and adoption, only a single set of conditions are reported for a given transformation even though it is well-recognized that a single set of conditions will not be applicable nor optimal for a wide range of substrates. For a chemist needing to utilize these reported methods to prepare new structures, this incomplete and fragmented data set constitutes a significant barrier to utilization. Our group has become interested in leveraging machine learning (ML) to address this limitation through quantitative prediction of reaction performance. While ML, defined as the study and application of computer algorithms that can learn from data, has been successfully applied in various fields of science and technology, its use in synthetic organic chemistry has been limited. In collaboration with Merck, we have gathered a large multi-dimensional data set for a Buchwald–Hartwig amination reaction using Merck’s ultra-High-Throughput Experimentation (HTE) technology. Using software created by our group, the reaction components were parameterized via the automated collection of atomic, molecular and vibrational descriptors. Various machine-learning algorithms, available in R, were evaluated for their ability to accurately predict yields in a training and test set. We found that a Random Forest (RF) algorithm was able to match patterns in data. This random forest model was also successfully applied to sparse training sets and out-of-sample prediction, suggesting its value in facilitating adoption of synthetic methodology.

Collaborators: Merck

Selected Reference:

  1. Ahneman, D. T.; Estrada, J. G.; Lin, S.; Dreher, S. D.; Doyle, A. G. “Predicting Reaction Performance in C-N Cross-Coupling Using Machine Learning” Science 2018360, 186-190[DOI: 10.1126/science.aar5169]
  2. Nielsen, M. K.; Ahneman, D. T.; Riera, O.; Doyle, A. G. “Deoxyfluorination with Sulfonyl Fluorides: Navigating Reaction Space with Machine Learning” J. Am. Chem. Soc. 2018140, 5004-5008. [DOI: 10.1021/jacs.8b01523]