Practical carbon–carbon bond formation from olefins through nickel-catalyzed reductive olefin hydrocarbonation

1Olefins are important synthons in organic chemistry. They are readily available as stable and inexpensive compounds with great diversity. Simple olefins are both raw materials and products in petrochemical industry. For example, ethylene is produced mostly through steam cracking. They are converted to higher olefins, polyethylene materials and various commodity chemicals. On the other hand, olefin groups are also widely represented in natural products with complex structures and many functional groups. Not only the extensive source but also the unique chemical reactivity of olefins attracts chemists, as the olefin moieties are resistant to a good number of synthetic transformations. Some unique transition metal catalyst systems can activate the olefin double bonds leading to highly elegant as well as useful reactions. Famous examples include the Wacker process, olefin metathesis, olefin hydroformylation and Heck reaction that have been extensively used in the preparation of complex organic molecules both in laboratory and in industry. These reactions establish the central role of olefins in modern synthetic organic chemistry as well as fine chemical industry.

2More recently, unactivated olefins have been used directly as chemical input in some novel cross-coupling reactions (for example, carbon-heteroatom coupling reactions and few examples of carbon–carbon coupling reactions). These findings suggest that olefins can be recognized as nuclephilic radical equivalents or alkylmetallics equivalents from a novel perspective. In some of these emerging methods that involve transition metal catalysts (Cu (ref. ), Fe (refs, ), Co (ref. ), Mn (ref. ) and so on), silanes were used as hydride source as well as reductant. New reactions that use olefins as chemical input are expected to bring new opportunities to organic synthesis. For instance, use of olefins to replace alkylmetallic reagents in traditional cross-coupling reaction fashion (for example, Kumada coupling reaction) with aryl/alkyl electrophiles would have appealing advantages such as better functional group compatibility and broader substrate availability.

3We now report the discovery of a new catalytic reaction of olefins, namely, Ni-catalyzed intermolecular reductive olefin hydrocarbonation between olefins and alkyl/aryl halides in an anti-Markovnikov fashion. This reaction provides an efficient strategy for the construction of carbon–carbon bonds from more stable and less expensive substrates as compared with the existing methods using organometallic reagents. In terms of practicality, the reaction shows high levels of ‘chemo'- and ‘regio'-selectivity, so that a wide range of sensitive functional groups can be tolerated (for example, epoxide, aldehyde and alcohol) in the transformation with minimal substrate protection necessary. As Ni-catalyzed carbon–carbon bond formation processes have enjoyed great success in modern synthesis, the present reaction is expected to find important applications in organic chemistry.

4We screened various Ni catalysts, base, silane and solvents for the reductive olefin hydrocarbonation reaction of 1-octene with 1 (see Table 1 and Supplementary Tables 1–5). The pincer complex L1 was tested first, but only trace amount of desired product was obtained with large amount of alkyl iodide recovered. We then tested the terpyridine ligand L2 and pybox ligand L3. Higher conversion of alkyl iodide was observed but the yield was only slightly improved. To our delight, we observed significant formation of the desired product with the phenanthroline family ligands L4 and L5. We then tested L6 bearing an ortho-methyl group but L6 was inferior. On the other hand, a bipyridine ligand L7 exhibited much better reactivity. Remarkably, when 4,4′-di-tert-butyl-2,2′-bipyridine (L8) was used, the GC yield increased to 96% with an isolated yield of 93% for the desired product. We also tested some monodentate phosphine ligands (L9 & L10) and carbene ligand (L11), but they were not effective.

5The substrate scope of the reductive olefin hydrocarbonation reaction was shown in Table 2. A variety of carbon electrophiles and olefins with different functional groups could be readily converted to the desired products with modest to excellent yields (30–93%). Not only alkyl iodides (for example, 3), bromides (for example, 13) and tosylates (for example, 30) were good substrates, but also aryl iodides (for example, 28) could be transformed successfully. With respect to olefins, both mono-substituted (for example, 26) and 1,1-di-substituted alkenes (for example, 31) could be used. Because of the mild reaction conditions, a wide range of synthetically relevant functional groups could survive the transformation. For instance, ether (4–6), ester (7), fluoride (8), trifluoromethyl (9), carbamate (10–11), sulfonamide (12–13), amine (14), aryl choride (15) and bromide (16) were well tolerated. Heterocycles such as thiophene (17), furan (18), and pyridine (19) could also be used in the reaction. Several base-sensitive groups, such as nitrile (20) and ketone (29, 32) posed no problem. Even more active groups, such as unprotected benzaldehyde (21) and azo groups (22), were compatible with the reaction. As an interesting substrate, 23 containing a pinacol boronate ester could selectively undergo the reductive olefin hydrocarbonation reaction with its carbon–boron bond intact. To our surprise, the reaction could even be conducted in the presence of an epoxide group (24) or an unprotected OH group (25). Noteworthily, an internal alkene (for example, 33) could have been converted in the reaction, although further ligand optimization was needed to improve the yields.

6To further demonstrate the high degree of functional group compatibility of the reductive olefin hydrocarbonation reaction (Fig. 1), we exploited its use as a novel tool for the modification of complex biologically interesting molecules (Fig. 2). As an example, a cholesterol derivative (34) could react with 35 to produce 36 without affecting either the internal alkene or alcohol groups (Fig. 2a). Hecogenin derivative (37), which contained both ketal and ketone groups, was also a good substrate for the modification process (Fig. 2b). Furthermore, calciferol (40) was converted to 42 selectively in the presence of the hydroxyl, internal alkene and even 1,3-diene groups (Fig. 2c).

7Modification of a cinchonidine derivative (43) resulted in the ‘chemo'-selective formation of 44, while tolerating both the amino group and quinoline structure (Fig. 2d, left). Single-crystal XRD analysis of 44 confirmed that the skeleton of cinchonidine was fully maintained during the modification process. In addition, the coupling of quinine (46) and a fructose derivative (45) enabled the production of highly complex molecules in a convergent fashion (Fig. 2d, right).

8The reductive olefin hydrocarbonation reaction of sclareol (48) proceeded smoothly in the presence of different two tertiary alcohol groups (Fig. 2e). In a more complex example with pleuromulin (50) (Fig. 2f), we obtained the desired product (52) in 28% yield (with 40% recovery of starting material) despite the presence of carbamate, ester, keton, unprotected primary and secondary alcohol groups in the reactant. Therefore, the reductive olefin hydrocarbonation reaction presents attractive opportunities for the modification of natural products or other complex molecules.

9Ethylene, as the simplest and most abundant olefin, has been attracting increasing attentions in synthetic organic chemistry. We were delighted that ethylene as C2 source was indeed a good substrate in the reductive olefin hydrocarbonation reaction (Fig. 3a). Compound 54 was obtained in 62% isolated yield.

10The reductive olefin hydrocarbonation reaction was useful for the synthesis of non-natural amino acids (Fig. 3b). As an example, homoserine-derived iodide 57 could be converted to 59 with a yield of 56%. More interestingly, the reaction of a racemization-prone serine derived iodide 58 was also successful affording 60 in 99% ee. This finding was surprising because in our previous study on the Ni-catalyzed reaction of 58 we observed significant racemization of the amino acid.

11To gain more insights into the reaction mechanism, radical clock experiments were carried out (Fig. 3c). Compound 61 containing a cyclopropyl ring was used as radical clock substrate (Fig. 3c, top). In this coupling reaction, we obtained only the ring-opened product 62 in 34% isolated yield. We also tested the reaction with (Z)-8-Iodooct-3-ene (63) (Fig. 3c, bottom). A mixture of linear coupling product (64a) and ring-cyclized product (64b) was obtained with a ratio of 3:1. The formation of ring-cyclized product (64b) revealed that this reaction proceeds through a radical cyclization process.

12Finally, we took advantage of optical pure secondary alkyl bromide (65) to study the stereochemistry of this reductive olefin hydrocarbonation reaction (Fig. 3d). When (S)-65 was alkylated with 1-hexene, we obtained a racemic product (67) in 37% yield. Furthermore, radical inhibiting experiment using TEMPO (2,2,6,6-tetramethylpiperidinooxy) as a radical trap was carried out (see Supplementary Discussion). The reaction was largely inhibited when 0.2 equiv. TEMPO was added, indicating a radical type reaction mechanism. Nonetheless, details for the mechanism of this reaction are not clear at present. Further investigations are ongoing in our lab.

13In summary, we have developed a practical and user-friendly method for the formation of carbon–carbon bonds through Ni-catalyzed intermolecular coupling of aryl or alkyl electrophiles with olefins under reductive conditions. This newly developed reductive olefin hydrocarbonation reaction provides a useful and general approach for the construction of carbon–carbon bonds by directly using olefins as nucleophile precursors. This reaction exhibited excellent compatibility with varieties of synthetically important functional groups and therefore, provided an efficient new approach for the modification of complex molecules. Our next challenge was the development of asymmetric version of this new carbon–carbon bond forming reaction and its extension to internal olefins.

14For NMR and high-performance liquid chromatography spectra of compounds in this manuscript, see Supplementary Figs 1–119. For details of the synthetic procedures, see Supplementary Methods. For X-ray data see Supplementary Data 1.

15NiBr2·diglyme (7.0 mg, 0.02 mmol, 10 mol%), 4,4′-di-tert-butyl-2,2′-bipyridine (8.0 mg, 0.03 mmol, 15 mol%) and Na2CO3 (42.4 mg, 0.4 mmol, 2.0 equiv.) were added to a Schlenk tube equipped with a stir bar. The vessel was evacuated and filled with argon (three cycles). To these solids, 0.6 ml DMAc was added under argon atmosphere. The reaction mixture was stirred at room temperature for 30 s. To the reaction mixture, electrophile (0.2 mmol, 1.0 equiv.), alkene (0.3 mmol, 1.5 equiv.) and DEMS (0.4 mmol, 2.0 equiv.) were added under a positive flow of argon. The reaction mixture was stirred at 30 °C for 12 h. To remove the DMAc, the reaction mixture was poured into 50 ml of ice water and the resulting mixture was extracted with ethyl acetate (4 × 30 ml). The combined organic layer was dried over Na2SO4, filtered, concentrated in vacuum and purified by column chromatography on silica gel.

16Accession codes: The X-ray crystallographic structures for 39 and 44 reported in this article have been deposited at the Cambridge Crystallographic Data Centre (CCDC), under deposition number CCDC 1439076 and 1439084, see Supplementary Data 1, the data can be obtained free of charge from the Cambridge Crystallographic Data Centre via http://www.ccdc.cam.ac.uk/ data_request/cif.

17How to cite this article: Lu, X. et al. Practical carbon-carbon bond formation from olefins through nickel-catalyzed reductive olefin hydrocarbonation. Nat. Commun. 7:11129 doi: 10.1038/ncomms11129 (2016).