Download PDF
Synlett 2020; 31(01): 37-40
DOI: 10.1055/s-0039-1690690
cluster – 9th Pacific Symposium on Radical Chemistry
© Georg Thieme Verlag Stuttgart · New York

Divergent Nickel-Catalysed Ring-Opening–Functionalisation of Cyclobutanone Oximes with Organozincs

Lucrezia Angelini
a   School of Chemistry, University of Manchester, Oxford Road, Manchester, M13 9PL, UK   Email: daniele.leonori@manchester.ac.uk
,
Laia Malet Sanz
b   Eli Lilly and Company Limited, Erl Wood Manor, Windelesham, Surrey, GU20 6PH, UK
,
Daniele Leonori  ∗
a   School of Chemistry, University of Manchester, Oxford Road, Manchester, M13 9PL, UK   Email: daniele.leonori@manchester.ac.uk
› Author Affiliations
Subject Editor: David Nicewicz and Corey Stephenson

D.L. thanks EPSRC for a fellowship (EP/P004997/1), and the European Research Council for a research grant (758427).
Further Information

Publication History

Received: 23 August 2019

Accepted after revision: 10 September 2019

Publication Date:
24 September 2019 (online)

 
 PDF Download Permissions and Reprints All articles of this category


Published as part of the Cluster 9th Pacific Symposium on Radical Chemistry

Abstract

The development of a nickel-catalysed strategy for the remote alkylation, arylation, vinylation and alkynylation of nitriles is presented. The methodology uses electron-poor O-Ar cyclic oximes and organozincs as coupling partners. This redox process proceeds through the generation of an iminyl radical and its following ring-opening reaction.


#

Key words

radical ring-opening - nickel catalysis - nitrogen radicals - C–C bond formation - alkylation - arylation - vinylation - alkynylation

Iminyl radicals are a powerful class of reactive intermediates with a broad reactivity profile spanning intramolecular exo-trig cyclisation, 1,5-hydrogen atom transfer (HAT) and β-fragmentation (Scheme [1]A).[1] Among these transformations, the ring-opening of cyclic iminyl radicals pioneered by Zard[2] has recently received considerable attention as key step in many photoredox and transition-metal-mediated processes.[3] This reactivity pattern carries interesting synthetic potential as it enables easy access to carbon radicals at distal positions from nitrile functionalities and can be used to achieve remote sp3–C functionalisation.

As part of an effort to explore the reactivity of nitrogen radicals,[4] we have recently developed a class of redox-active oximes and used them in oxidative photoinduced radical cascades based on iminyl radical generation, radical ring-opening and SH2 reaction with SOMOphiles like Selectfluor, N-chlorosuccinimide and sulfonyl azide.[5] More recently, we have merged this reactivity with nickel catalysis and extended the range of coupling partners to electrophiles such as aryl and alkyl halides and terminal alkynes (Scheme 1C, left).[6] This dual catalytic approach has enabled remote arylations, alkylations and vinylations of nitriles.

Zoom Image
Scheme 1 Reactivity of iminyl radicals

As the ability to introduce C-based groups at distal sp3–C atoms remains a fundamental synthetic challenge,[7] we recently questioned if the redox requirements for iminyl radical generation and nickel cross-coupling mode could be changed to enable the use of nucleophilic coupling partners. Such an umpolung approach would deliver related products to our previous oxidative approach but it will use a different class of building blocks which might offer complementarity in terms of substrate scope.

From this perspective, we have recently developed a nickel-catalysed protocol for the generation of amidyl radicals from electron-poor O-Ar-hydroxyl amides in the presence of nucleophiles like aryl boronic acids and dialkyl organozincs (Scheme [1]B).[8] This redox activation mode enabled the development of radical cascades based on amidyl exo-trig cyclisation followed by arylation/alkylation via a common diorganyl–Ni(III) intermediate.

We were therefore interested in evaluating whether, by turning the redox nature of the oxime starting material into a strong electron acceptor, we could achieve umpolung ring-opening–functionalisation reactions.

We present here the successful implementation of such a strategy and its use in the divergent functionalisation of nitriles through remote alkylation, arylation, vinylation and alkynylation with organozincs (Scheme [1]C, right).

Our proposed mechanism for ring-opening–functionalisation of iminyl radicals with organozincs (Scheme [2]) is based on a Ni(I)/(II)/(III) redox cycle and starts with the transmetalation of a Ni(I)–X complex (ligand on Ni not shown for simplicity) A with the organozinc coupling partner.[9] The corresponding organyl–Ni(I) species B is expected to be a competent reductant (E ox ~ +1.1 V vs SCE)[10] and should undergo facile ground-state single-electron transfer (SET) with the cyclobutanone oxime C (E red ~ –0.8 V vs SCE).[11] This event will deliver the iminyl radical D with simultaneous extrusion of a stable aryloxy anion. Subsequent fast ring-opening reaction will form the distal nitrile radical E that should have the correct philicity to intercept the oxidised organyl–Ni(II) intermediate F. This radical transmetalation would deliver a diorganyl–Ni(III) species G from which reductive elimination should be facile. In this way we should obtain the desired product H and close the nickel catalytic cycle.

Zoom Image
Scheme 2 Proposed mechanism for the divergent Ni-catalysed ring-opening–functionalisation via iminyl radicals

We began our study analysing the reaction of aryl-oxime 1, easily accessible through a single-step reaction on multi-gram scale, with freshly prepared dipentylzinc, using the preformed NiBr2 ·dtbpy as the catalyst (Table [1]). Pleasingly, the desired alkylated nitrile 2 was immediately obtained in 65% yield by using one equivalent of the organometallic species, 20 mol% of Ni catalyst in a 1:1 mixture of THF–DMF as the solvent at room temperature (entry 1). Decreasing the amount of catalyst to 10 and 5 mol% still enabled reactivity but had a detrimental effect on the reaction efficiency (entries 2 and 3). The counterion of the Ni catalyst was not important (entry 4), while the absence of the dtbpy ligand completely suppressed the reactivity (entry 5). We also evaluated the possibility of using a larger amount of dipentyl zinc but this led to a decrease in reaction yield suggesting that a 1:1 ratio between iminyl precursor and organozinc is the optimum (entry 6). The solvent plays an important role in this reaction. Indeed, when the DMF co-solvent was replaced with the similarly polar aprotic solvent dimethylacetamide (DMA), a severe reduction in product formation was observed (entry 7). In line with the intramolecular nature of the transformation, a higher dilution affected negatively the reaction outcome (entry 8), although the optimal concentration reaches a plateau and no extra beneficial effect is observed at increased concentration (entry 9).

Table 1 Optimisation of the Ring-Opening–Alkylation Reaction with Use of Oxime 1 and (n-pentyl)2Zn

Entry

[Ni] (mol%)

(n-pentyl)2Zn (equiv)

THF–DMF (M)

Yield (%)

1

dtbpy·NiBr2 (20)

1.0

0.05

65

2

dtbpy·NiBr2 (10)

1.0

0.05

46

3

dtbpy·NiBr2 (5)

1.0

0.05

22

4

dtbpy·NiCl2 (20)

1.0

0.05

64

5

NiBr2 (20)

1.0

0.05

–

6

dtbpy·NiBr2 (20)

2.0

0.05

44

7

dtbpy·NiBr2 (20)

1.0

0.05a

24

8

dtbpy·NiBr2 (20)

1.0

0.025

61

9

dtbpy·NiBr2 (20)

1.0

0.1

50

a DMA was used instead of DMF.

With these optimised conditions in hand, we moved to explore the scope of this ring-opening–functionalisation strategy (Scheme [3]).[12] Oxime 1 was successfully engaged with other dialkyl organozincs like the long-chain (3,7-Me2-octyl)2Zn leading to nitrile 5 in good yield. More importantly, we were able to utilise a secondary cyclic dialkyl zinc as coupling partner albeit giving lower yield (6). Spirocyclic and 3-OBn-containing cyclobutanone-oximes 3 and 4 were also evaluated and proved to be efficient precursors for the ring-opening and coupling with dimethyl (7 and 10), diethyl (8 and 11) and dicyclopropyl zinc (9 and 12) thus enabling the preparation of C-4-disubstituted piperidines as well as aldol-type nitriles, respectively. The successful introduction of Me as well as cyclopropyl groups represents a complementarity aspect of this Ni strategy with respect to our previous dual photoredox-Ni work[6] where coupling partners like Me–Br/I and cyclopropyl–Br/I were found not viable.

Zoom Image
Scheme 3 Scope and limitations for the Ni-catalysed ring-opening–functionalisations of cyclobutanone oximes

Having evaluated the efficiency of this methodology in sp3–sp3 C–C bond formation, we decided to test its applicability in the assembly of sp3–sp2 C–C bonds.[3h] [13] Specifically, the use of diphenyl (13), di-ortho-substituted aromatics (14, 15) as well as a di-hetero-aryl zinc (16) provided the desired products in high yields. More interestingly, we succeeded in engaging divinyl zinc in this strategy and obtained 17 and 18 in moderate and high yields, respectively.

Finally, we prepared diethynyl zinc from the commercially available Grignard reagent in order to evaluate if this strategy would also enable the formation of sp3–sp C–C bonds.[3g] Pleasingly, this was possible and both oximes 1 and 3 provided the corresponding alkyne-containing nitriles 19 and 20 in moderate to high yields. To the best of our knowledge, this represents the first example of a strategy able to introduce a terminal alkyne functionality upon radical ring-opening of iminyl radicals.

This methodology is however not without limitations. At the moment the major hurdle is represented by the iminyl radical scope. In fact, while cyclobutanone oximes can be successfully used, extension of this strategy to cyclopentanone (21) and cyclohexanone (22) oximes was not possible. As the ring-opening of five- and six-membered-ring iminyl radicals is feasible, we propose that difficulties in the radical transmetalation of the corresponding tertiary and benzylic radicals onto the nickel species or the potential formation of an aza-enolate by deprotonation of the oxime from the organozinc might be responsible for this lack of reactivity.

In conclusion, we have demonstrated that highly electron poor and redox-active cyclobutanone oximes can undergo nickel-catalysed iminyl radical generation followed by ring-opening and divergent functionalisation with alkyl, aryl, vinyl and alkynyl organozincs. These results offer an umpolung alternative for the distal functionalisation of nitriles.


#

Acknowledgment

L.A. thanks Eli Lilly for a PhD CASE Award.

Supporting Information

    Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/s-0039-1690690.
  • Supporting Information


   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 Reactivity of iminyl radicals
Zoom Image
Scheme 2 Proposed mechanism for the divergent Ni-catalysed ring-opening–functionalisation via iminyl radicals
Zoom Image
Scheme 3 Scope and limitations for the Ni-catalysed ring-opening–functionalisations of cyclobutanone oximes