Reversible C–C Bond Formation Using Palladium Catalysis

: A widely appreciated principle is that all reactions are fundamentally reversible. Observing reversible transition metal-catalyzed reactions, particularly those that include the cleavage of C – C bonds, are more challenging. The development of the palladium- and nickel- 13 catalyzed carboiodination reactions afforded access to the syn - and anti -diastereomers of the 14 iodo-dihydroisoquinolone products. Using these substrates, an extensive study investigating the 15 catalytic reversibility of the C – C bond formation using a different palladium catalyst was 16 undertaken. A combination of experimental and computational studies led to the discovery of a 17 variety of new methodologies and concepts key to understanding the process of reversible C – C 18 bond formations. Main Text: a C – H rarity interrogate

2 enriched heavy atoms are not required. A reversible process has been found that offers an optimal 42 starting point for further studies in this important reaction manifold (Scheme 1d). With access to both the anti-and syn-diastereomers of the iodinated compounds, we had an ideal 54 opportunity to investigate the reversibility of the carbohalogenation process.

55
By subjecting the anti-diastereomer formed in the nickel-catalyzed carboiodination reaction to a 56 palladium catalyst incapable of performing the C-I reductive elimination, 45 we identified a 57 palladium-catalyzed β-carbon elimination, cleaving the unstrained 6-membered ring containing an 58 all-carbon quaternary stereocenter, followed by reformation of the same C-C bond with the 59 stereochemistry observed in the palladium-catalyzed process. Herein, we describe our studies that 60 shed light on mechanism of this reaction involving a catalytically reversible β-carbon elimination 61 and outline new stereoelectronic factors for this β-carbon elimination process supported by 62 experimental and computational evidence. In addition, we developed an efficient catalytic strategy 63 for the diastereoconvergent formation of indenodihydroisoquinolones via a palladium-catalyzed 64 net epimerization arising from reversible C-C bond formation followed by C-H activation. The These results are in agreement with the proposed β-carbon elimination pathway (Scheme 2d), as a 88 higher yield of product was observed when starting from a later point along the proposed catalytic 89 cycle. 90 We speculated that the transformation initiates via an oxidative addition of the palladium(0) 91 catalyst to the neopentyl iodide, followed by a β-carbon elimination, which yields an aryl To probe the 1,3-palladium shift mechanism, we studied an analogous substrate bearing an ethyl 122 group (Scheme 3a). We could envision 3 possible outcomes: an alkenyl derived product stemming 123 from a 1,3-palladium shift and subsequent β-hydride elimination, a tetracyclic product bearing a 124 methyl group where the previous diastereotopic methylenewould have been, or complete inhibition 125 of the reaction. We observed the expected product of the β-carbon elimination cascade (2b) as the 126 major product, in a yield of 73% from anti-1b, and 82% yield from syn-1b, using only 10 mol % 127 catalyst, supporting that the reaction does not proceed via a 1,3-palladium shift. 6 To exclude the possibility of an aryl epimerization, we employed the enantioenriched starting 130 material anti-1c, prepared from the corresponding aryl iodide lin-1c with the same absolute 131 stereochemistry (Scheme 3b). If the anti-product was undergoing epimerization rather than retro-132 carbopalladation, we would have expected to isolate the enantiomer of the final product.

133
Subjecting both substrates separately to the reaction conditions gave the identical enantiomer, 2c, 134 with no degradation in enantioselectivity. This result supports that stereocenter undergoing the 135 epimerization was the quaternary all-carbon stereocenter.

137
We then turned to density functional theory (DFT) to study the energetic landscape of this reaction 138 (Scheme 4a, see supporting information for computational details). We began our investigation  Although the anti-indenodihydroisoquinolone was never observed experimentally, we 174 investigated the energetics of an anti C-H activation leading to this product (Scheme 4b).

175
Formation of the carbonate complex, 9, from 1 is exergonic by 12.9 kcal mol -1 . The C-H activation  Reactions using the anti-diastereomer typically gave slightly diminished yield when compared to 216 the 1:1 mixture, which is likely due to the greater efficiency of the syn-isomer going to the product 217 (Scheme 5c). Generally, electron-deficient substrates outperformed the electron-rich ones. The p-218 and m-CF3 substrates gave the product in the highest yields (94% and 86% respectively). Even for 219 substrates containing a methyl group at the R 1 position, incorporating a p-CF3 group on the 220 molecule resulted in full conversion of the starting material at only 10 mol % catalyst. The desired 221 product was obtained in 47% yield and the remaining mass balance was the protodemetalation 222 product. In comparison, the parent aryl substrate containing a methyl group gave 13% product and 223 58% of the protodemetalation with ~30% unreacted starting material at 10 mol % catalyst.

224
Unfortunately, it was not possible to explore the reactivity of electron-rich substrates, as the nickel-225 carboiodination reaction failed to yield these parent compounds. that of the methyl (ΔG = 8.8 kcal mol -1 compared to ΔG = 5.7 kcal mol -1 ). Interestingly, our 265 analysis revealed that the activation energy for the β-carbon elimination transition state was 266 lowered for R 1 = Et (ΔG ‡ = 32.6 kcal mol -1 for R = Et and ΔG ≠ = 35.7 kcal mol -1 for R = Me).

267
Based on these results, we suggest that there is an increased level of steric build-up that is relieved 268 upon β-carbon elimination when R = Et. We suspect that off-cycle pathways leading to 269 protodemetalation are negligible due to this rate enhancement.