Periodic Reporting for period 3 - Let-it-Bi (Bismuth Redox Catalysis for Sustainable Organic Synthesis)
Periodo di rendicontazione: 2023-02-01 al 2024-07-31
To tackle this fundamental idea, Let-it-Bi was divided into two main Working Packages. In WP1 we proposed to disclose a high-valent Bi redox catalysis, where the catalyst would revolve between oxidation states +3 and +5. This WP1 included applications of such redox system to a variety of C‒X, C‒O and C‒N bond forming reactions. On the other hand, in WP2 we proposed that a low-valent redox platform should also be feasible, where the Bi now revolves between Bi(I) and Bi(III). As disclosed in Part 2 of this section, both Working Packages have been successfully deployed and achieved, albeit enormous chemical space is still to be investigated. This implies that Let-it-Bi has provided a framework that resulted in a paradigm shift within the field of catalysis and therefore, establish a research group in Europe that is renowned worldwide. Questioning the dogmas or the established beliefs is an arduous task, that normally enables new discoveries in the area or subject of study. Let-it-Bi aimed precisely at this, and is currently harvesting the fruits of 2.5 years of investment in fundamental understanding. Bi redox catalysis is still in its infancy, with our group virtually the only one active in the field at the onset. At present, Let-it-Bi continues to be explored accordingly, and there is no doubt that bismuth will still offer many opportunities in the field of catalysis in the near future (Fig. 1).
During this period we have developed a family of Bi(III) complexes based on a bis-aryl sulfoximine or sulfone backbones, which surround and support the Bi(III) center (Fig. 2A). Investigations on these bismuthine complexes revealed the possibility to perform the canonical organometallic steps of cross-coupling, namely oxidative addition, transmetallation and reductive elimination (Fig. 2B). This reactivity was initially demonstrated in the context of a C‒F bond formation; we provided a Bi-catalyzed fluorination of boronic esters (Fig. 2C). Later on, we expanded this reactivity to the formation of C‒OTf and C‒ONf (Fig. 2D) and more recently, C‒N(SO2R)2 has also been achieved and is currently under investigation (unpublished).
These couplings are currently beyond the scope of transition metals, thus highlighting the uniqueness and orthogonality of this new redox platform. It is important to mention that a great effort has been placed in the mechanistic study of the C‒F bond in particular. Indeed, these investigations revealed that the reductive elimination from high-valent Bi(V) proceeds through a 5-membered ring, thus making the BF4 anion the coupling partner (Fig. 3A). It was also found, that such reductive elimination mode was also applicable to the C‒OTf and the C‒N(SO2R)2 bond (Fig. 3B). This observation is fundamentally very important for several reasons: 1) these coupling partners are usually considered “non-coordinating” counterions, and therefore their reactivity as nucleophiles has always been disregarded; 2) the ligand coupling (aka reductive elimination) proceeds through a 5-membered TS, which differs from the canonical 3-membered TS in transition metal chemistry (Fig. 3C). Whereas the 5-membered TS has been postulated for ligand couplings in stoichiometric reactions, catalytic variants of this step were unknown. The implications of these subtle differences are quite relevant. Reductive elimination is by definition, the microscopic reverse reaction of oxidative addition. Hence, if we have established that a 5-membered reductive elimination is a general pathway for C‒F, C‒O and C‒N, there should exist a reversible oxidative addition process, which proceeds through a 5-memebred ring (Fig. 3C). This would be reminiscent to the SN2’ reactivity of allyl halides in Tsuji-Trost-type reactions but in a concerted fashion.
After establishing the feasibility of three canonical organometallic steps, we became interested in whether another highly coveted step would also be feasible: namely the migratory insertion reaction. In this vein, we discovered that SO2 can insert into Bi‒C, thus forming bismuth sulfinates (Fig. 4A). Based on this fundamentally new step for Bi, we developed a Redox Neutral Catalysis based on Bi(III) complexes which permits the synthesis of aryl sulfonyl fluorides in one-pot, from boronic acids, SO2 and Selectfluor (Fig. 4B). The possibility to have the catalyst in the presence of Selectfluor without side-reactivity is a characteristic of this system, due to the low-energy of the frontier 6s2 electrons for Bi compared to d-block catalysts.
When investigating the redox couple Bi(III)/Bi(V), we realized that there existed little precedents on high-valent halobismuhonium compounds (Bi(V) halides), in particular, fluorobismuthonium salts. It is for this reason that in order to understand our catalytic systems based on the formation of intermediates of Bi(V), we devoted a severe amount of time to synthesize families of fluoro and chlorobismuthonium compounds to gain insights into their geometry, structure and reactivity. In this line, we have synthesized the rather elusive fluorobismuthonium cations based on mono-, bi- and trimetallic species.
Having established the practical and theoretical foundations of the high-valent catalysis, we are currently exploring the possibility to undergo C‒C bonds, as outlined in our original proposal. This is a much more challenging aspect of this chemistry and it will be reported in due course.
Articles emanated from WP1 are listed below:
1. Planas, O.; Wang, F.; Leutszch, M.; Cornella, J. Science, 2020, 367, 313. (♦)
2. Planas, O.; Peciukenas, V.; Cornella, J. J. Am. Chem. Soc. 2020, 142, 11382.
3. Magre, M.; Kuziola, J.; Nöthling, N.; Cornella, J. Org. Biomol. Chem. 2021, 19, 4922.
4. Planas, O.; Cornella, J. Nachrichten aus der Chemie, 2021, 69, 79 (perspective)
5. Magre, M.; Cornella, J. J. Am. Chem. Soc. 2021, 143, 21497.
6. Moon, H. -W.; Cornella, J. ACS. Catal. 2022, 12, 1382.
7. Kuziola, J.; Magre, M.; Nöthling, N.; Cornella, J. Organometallics 2022, 41, 1754.
8. Planas, O.; Peciukenas, V.; Leutzsch, M.; Noethling, N.; Pantazis, D.; Cornella, J. J. Am. Chem. Soc. 2022, 144, 14489.
9. Magre, M.; Ni, S.; Cornella, J. Angew. Chem. Int. Ed. 2022, 61, e202200904.
Note: the articles containing this symbol ♦ were published during the period comprising the interview and the start of the action.
In Working Package 2, we envisaged to unlock the potential of Bi(I)/Bi(III) redox couple for organic synthesis. Initially, we demonstrated that low-valent Bi(I) complexes supported by a N,C,N pincer ligand were able to catalyze the transfer hydrogenation of azo- and nitroarenes with ammonia borane and H2O (Fig, 5A). Indeed, this process seems to proceed via a redox Bi(I)/Bi(III) cycle, albeit the exact mechanistic details still remained elusive. Based on this reactivity, we initially proposed that activation of H2 and functionalization of olefins was the logical extension of this reactivity. However, after extensive mechanistic investigations, we concluded that H2 formation and further olefin functionalization would require a re-design of the approach due to the differences in mechanism between the two hydrogen sources. Therefore we explored alternatives were the low-valent Bi(I)/Bi(III) redox platform could be applied.
In this sense, we focused our attention in N2O. This gas is a potent greenhouse gas, the third most abundant in the atmosphere behind CO2 and CH4, and ca 300 times the global warming potential than that of CO2. It is for this reason that strategies to degrade this inert gas are highly coveted. When scrutinizing the reports on this topic we realized that there were no examples of catalytic degradation of N2O using a main group element. Various Bi(I) showed excellent reactivity with N2O, permitting a catalytic process to be unfolded, using HBpin as O-accepting reagent. The catalytic system is characterized by the high TON and the mild temperatures and pressures at it operates.
Another catalytic reaction where the low-valent redox catalysis could be applied was in the hydrodefluorination of C‒F bonds (Fig. 6A). We demonstrated that a low-valent Bi(I) can cleave a C‒F bond embedded into a polyfluorinated framework, thus delivering the corresponding C‒Bi(III)‒F. This intermediate is primed to react with a silane, to afford a Si‒F and a C‒Bi(III)‒H. The latter, although is a highly unstable compound, it could be characterized at low temperatures by NMR and HRMS. By exploring such organometallic steps, we could develop the catalytic variant and various molecules could be defluorinated (Fig. 6B).
When exploring the reactivity of low-valent Bi(I) complexes, we realized of an incredibly fascinating reactivity. When Bi(I) is oxidized with bulky oxyaryl one-electron oxidants, it leads to a Radical Equilibrium Complex (Fig. 7A). These complexes are characterized by staying in equilibrium between the Bi‒O bond and its homolysis product (Bi and O radicals within the solvent cage). We envisaged that such electronically unique situation would be suitable in the context of small molecule activation (Fig. 7B). Indeed, when these complexes are mixed with either O‒H or N‒H bonds, activation of the N‒H bond occurs immediately at room temperature in seconds, delivering the corresponding Bi‒OH or Bi‒NH2 (Fig. 7C). Ammonia and water were also smoothly activated. This activation mode for main group is rather unusual with only one example for Ge known using light excitation. In this case, the relativistic effects of Bi permit the homolysis of the Bi‒O bond at room temperature without the need of additional light. The resulting Bi(II) weakens the BDFE of the N‒H or O‒H bond permitting the O radical to undergo rapid H atom abstraction. This coordination bond-weakening properties are mostly reported for transition metals. Yet, the fact that Bi is able to perform in this manner opens the door to consider this main group element in the context of ammonia activation or even, N2 reduction.
Up until now, all the catalytic redox processes developed in the low-valent front are restricted to a two-electron manifold. However, very recently, we have found that our Bi(I) pincer complexes are able to engage in one-electron oxidative addition (Fig. 8A). This new property is truly unconventional for main group elements, which usually engage lone pairs and therefore, two electrons at a time. However, catalytically exchanging solely one electron is a revolutionary new reactivity paradigm for main group. We have already initiated exploring this possibility further and have recently developed a C‒N bond formation reaction, where Bi(I) converts the same starting materials for an amide bond formation into a decarboxylative C‒N bond formation (Fig. 8B). These results have not been yet published in a peered review journal.
This new reactivity will certainly be explored further during the second part of this action and the results will be communicated in due course.
Articles emanated from WP2 are listed below:
1. Wang, F.; Planas, O.; Cornella, J. J. Am. Chem. Soc. 2019, 141, 4235. (♦)
2. Pang, Y.; Leutzsch, M.; Nöthling, N.; Cornella, J. J. Am. Chem. Soc. 2020, 142, 19473
3. Pang, Y.; Leutzsch, M.; Nöthling, N.; Katzenburg, F.; Cornella, J. J. Am. Chem. Soc. 2021, 143, 12487
4. Pang, Y.; Cornella, J. Organometallic Compounds of Arsenic, Antimony and Bismuth, Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, Elsevier, 2021 (book chapter)
5. Moon, H. -W.; Cornella, J. ACS. Catal. 2022, 12, 1382
6. Yang, X.; Reijerse, E.; Bhattacharyya, K.; Leutzsch, M.; Kochius, M.; Noethling, N.; Busch, J.; Schnegg, A.; Auer, A. A.; Cornella, J. J. Am. Chem. Soc. 2022,144, 16535.
7. Mato, M.; Spinnato, D.; Leutzsch, M.; Moon, H. –W.; Reijerse, E.; Cornella, J. ChemRxiv, 2022, doi:10.26434/chemrxiv-2022-dqj5r
Note: the articles containing this symbol ♦ were published during the period comprising the interview and the start of the action.
Due to our continuous efforts in this front, a new chapter for Bismuth Redox Catalysis has recently also opened. While exploring the possibilities of Let-it-Bi, we have encountered that Bi compounds can also engage in radical catalysis; in other words, they are able to revolve not only through the postulated two electron manifold in the catalytic redox cycle, but one electron at a time. With this new reactivity paradigm, Bismuth places itself as a unique member of the main group family, with properties that are closer to a first row transition metal than a p-block element.
Having established the redox neutral catalysis and the two-electron redox catalysis in both high and low-valent manifolds, we are now venturing into this new field of radical catalysis, which was not expected in the initial proposal of Let-it-Bi. However, whereas there is still plenty to discover in the two-electron front, we believe that this new area of expertise can open up a plethora of new and unexpected opportunities for synthesis and catalysis, and is therefore important to continue the exploration of this unexpected reactivity.