Intermolecular selective carboacylation of alkenes via nickel-catalyzed reductive radical relay

The development of catalytic carboacylation of simple olefins, which would enable the rapid construction of ketones with high levels of complexity and diversity, is very challenging. To date, the vast majority of alkene carboacylation reactions are typically restricted to single- and two-component methodologies. Here we describe a three-component carboacylation of alkenes via the merger of radical chemistry with nickel catalysis. This reaction manifold utilizes a radical relay strategy involving radical addition to an alkene followed by alkyl radical capture by an acyl-nickel complex to forge two vicinal C−C bonds under mild conditions. Excellent chemoselectivity and regioselectivity have been achieved by utilizing a pendant weakly chelating group. This versatile protocol allows for facile access to a wide range of important β-fluoroalkyl ketones from simple starting materials.


Supplementary Discussion
Evaluation of the stirring rates:

Supplementary Table 2. Evaluation of the stirring rates
Conclusion: A stirring rate of 1500 rpm is necessary for high reaction efficiency and reproducibility. We also sometimes observed that the reaction mixtures remained clear with low stirring rates, which led to no formation of products.

Other successful and unsuccessful substrates:
Acyclic internal alkenes: The reaction of (E)-but-2-en-1-yl benzoate under the standard condition afforded the desired product S9 in 40% yield as well as the regioisomeric product S10 in 28% yield (r.r. 10:7).

S48
The reaction of CF 3 I under the standard condition afforded the desired trifluoromethylacylated product S11 in 29% yield (37% 19 F NMR yield).

Tertiary alkyl halides:
Simple tertiary alkyl bromides were also viable coupling partners under slightly modified conditions. For example, the reaction of ethyl 2-bromo-2-methylpropanoate gave the desired carboacylated product S12 in 36% yield.

Regioselectivity discussions:
Regarding the regioselectivity issue, we have carefully reevaluated all NMR data, GC-MS data, and several crude NMR and GC-MS data, and can conclude that regioselectivity is very high for these terminal alkenes. Some of the relevant analysis has been highlighted bellows:

Crude NMR analysis:
Since the 19  has been observed, which is the same as the final pure product compound (dt).

Supplementary Figure 10. The crude 19 F NMR of ICF 2 CO 2 Et
For the case of CF 3 I, the CF 3 group for the desired product is triplet (t), while the CF 3 group for the regioisomer should be doublet (d).
The crude 19 F NMR of ICF 2 CO 2 Et:

S53
In the crude 19 F NMR, the desired CF 3 -product is present (-64.25 ppm, triplet, (J = 11.3 Hz)). There are some unidentified byproducts which appear as singlet peaks in the crude 19 F NMR, which does not suggest the presence of the regioisomer.
Besides CF 3 H (-78.41 ppm, doublet, J = 39.5 Hz), only one double peak (-68.97 ppm, doublet, J = 2.8 Hz) has been observed in low yield (1% yield in 19 F NMR). This doublet peak was inseparable to the desired product, and the ratio of these two compounds is 1:150 in the final spectrum of the isolated product. We couldn't rule out the possibility that this minor peak is the regioisomeric product. Nevertheless, the regioselectivity for this transformation is still high (grear than the limit of precision by NMR analysis).
The crude 19 F NMR of CF 3 I:

F NMR of CF 3 I
For the case of C 4 F 9 I, only one set of 19 F NMR peaks, which is the same as the final product, has been observed in the crude 19 F NMR of the reaction mixture.
Supplementary Figure 12. The crude 19 F NMR of C 4 F 9 I

Characterization of impurities
One reviewer questioned whether the minor impurities showed in the spectra of several products are the regioisomeric products or not.
For the cases of 19 and 25, we have successfully isolated the primary impurities appeared in 1 H NMR spectra. The primary impurities are the b-eliminated byproducts of the final products, not the regioisomeric products.
The crude 19 F NMR of C 4 F 9 I:

S58
Being different from pathway I, in pathway II, alkyl radical P3 could be captured by Ni(0) to form Ni I complex P7, followed by oxidative addition of acyl chloride to deliver the crucial Ni III adduct P4.

Stoichiometric reaction of isolated Ac-Ni II complex
The stoichiometric reaction of Ac-Ni II complex 55 (prepared according to the previously reported procedure 3 ) with alkene and C 4 F 9 I in the presence of Mn dust gave the desired coupling product 16 in 42% yield.

Radical inhibition experiment
Addition of TEMPO (1.0 equiv.) completely shut down the desired transformation, and TEMPO-C 4 F 9 adduct was detected in 79% 19

Radical clock experiment
Diene 52 underwent radical addition and cyclization, furnishing the expected coupling product 53 in 23% yield as well as alkyl iodide 54 in 49% yield.     In the absence of nickel catalyst, alkene remained untouched. Mn(0) couldn't promote the generation of C 4 F 9 radical. We assumed that Ni(0) is responsible for the generation of C 4 F 9 radical in this transformation.

Evaluation of the chelating group effect
Six-membered and seven-membered chelating substrates gave the optimal yields.
For the case of the eight-membered chelating substrate, the formation of C 4 F 9 -alkene byproduct (36% isolated yield) as well as C 4 F 9 -alkane byproduct (detected by GC-MS) has also been observed. For the substrate without directing group, only a trace amount of the desired product was observed, major byproducts were C 4 F 9 -alkene and C 4 F 9 -alkane (detected by GC-MS).

Supplementary Figure 22. Reaction of pent-4-en-1-ylbenzene
These results indicate that the chelating effect does not promote the generation of C 4 F 9 radical. We envision that the chelating effect would facilitate the capture of alkyl radical to Ni complex.