sSPhos: A General Ligand for Enantioselective Arylative Phenol Dearomatization via Electrostatically-Directed Palladium Catalysis

Arylative phenol dearomatization affords complex, cyclohexanone-based scaffolds from simple starting materials, and asymmetric versions allow access to valuable enantioenriched structures. However, bespoke chiral ligands must typically be identified for each new scaffold variation. We have addressed this limitation by applying the concept of electrostatically-directed palladium catalysis whereby the chiral sulfonated ligand sSPhos engages in electrostatic interactions with a phenolate substrate via its associated alkali metal cation. This approach allows access to highly enantioenriched spirocyclohexadienones, a process originally reported by Buchwald and co-workers in a predominantly racemic manner. In addition, sSPhos is proficient at forming two other distinct scaffolds, which had previously required fundamentally different chiral ligands, as well as a novel oxygen-linked scaffold. We envisage that the broad generality displayed by sSPhos will facilitate the expansion of this important reaction type and highlight the potential of this unusual design principle, which harnesses attractive electrostatic interactions.

P henol dearomatization is exceptionally useful for building up three-dimensional molecular complexity. 1 Although the energetic barriers can be high, the products possess versatile functionality and often a new stereocenter.Dearomatization of phenols is typically more challenging than naphthols, indoles, pyrroles, and the like because of lower electron density.While many methods rely on highly electrophilic reagents, transition metal catalysis has recently expanded the breadth of accessible transformations. 2 This has enabled arylative dearomatizations during which a new arene substituent is introduced during the dearomatizing event (early methods for arylative dearomatization relying on stoichiometric lead, 3 bismuth, 4 and iodine 5 arylating reagents possessed various limitations).A pioneering advance was reported by Buchwald and co-workers in 2011 with the palladium-catalyzed intramolecular arylation of phenols, which produced spirocyclohexadienones bearing all-carbon quaternary centers (Figure 1A, upper). 6The scope was explored using an achiral phosphine ligand (L1) but it included two preliminary enantioselective results, one phenol and one naphthol.For the phenol example, L2 allowed 81% ee in the intramolecular dearomatization of 1a (Figure 1A, lower).Since Buchwald's original report, a number of important developments on palladium-catalyzed arylative phenol dearomatization to form different scaffolds, from the groups of You 7 and Tang, 8 have been made. 9This includes asymmetric variants using TADDOL-derived chiral phosphoramidites 7b and P-chiral biaryl monophosphine ligands, 8a respectively.It is evident that success for a new substrate class requires extensive ligand evaluation and tailoring, a feature which hinders more rapid development and widespread use of this important reaction type. 10e recently reported the use of enantiopure sSPhos as an unexplored chiral, bifunctional phosphine ligand, which can be readily obtained by diastereoselective recrystallization of (rac)-sSPhos as its quinidinium salt (Figure 1B). 11Originally reported by Anderson and Buchwald as a water-soluble ligand for cross-coupling, 12 we initially utilized (rac)-sSPhos for control of site selectivity in the cross-coupling of polyhalogenated arenes.Therein, we introduced the concept of electrostatically directed palladium catalysis, whereby an anionic ligand interacts with an anionic substrate via a bridging alkali metal cation through electrostatic interactions (Figure 1C, left). 13,14We subsequently found enantiopure sSPhos to be highly proficient in controlling enantioselectivity in Suzuki− Miyaura couplings to form 2,2′-biphenols, an outcome we tentatively attributed to an organizing network of hydrogen bonds between the ligand sulfonate group and the phenolic hydroxyls on the coupling partners (Figure 1C, right). 11On the basis of these precedents, we hypothesized that enantiopure sSPhos might be an effective ligand for enantiocontrol in the Buchwald arylative dearomatization reaction.Mechanistically, in the presence of a strong base, it is likely that phenolate formation occurs and that the subsequent palladation of this phenolate may be selectivitydetermining.7b,15 We envisaged that an attractive electrostatic interaction might occur between the alkali metal cation of the phenolate and the sulfonate group of the ligand, akin to those we invoked in our prior work, thereby providing organization in a chiral environment (Figure 1D).More broadly, we were optimistic that, by exploiting electrostatic interactions with phenolate intermediates, sSPhos might constitute a generally applicable ligand for enantioselective, Pd-catalyzed arylative phenol deromatization across a diverse range of scaffolds. 16e began with conditions similar to those identified in Buchwald's study 6 by using [Pd(cinnamyl)Cl] 2 and K 2 CO 3 in dioxane at 110 °C and (R)-sSPhos as the ligand (Table 1, entry 1).Pleasingly, spirocycle 2a was formed in 76% yield with encouraging enantioselectivity.An evaluation of palladium sources (entries 2−4) revealed that Pd 2 dba 3 afforded significant improvement (63% ee), as did switching the base to KOH (entry 5, 84%).Aromatic solvents provided no improvement (entries 6 and 7), but addition of water as a cosolvent increased yield and enantioselectivity in all cases (entries 8−10). 17A PhMe:H 2 O biphasic mixture proved to be optimal and afforded 2a in 98% yield and 92% ee with the absolute configuration determined by X-ray diffraction (entry 10).Reactivity remained excellent at 90 °C but with no improvement in ee (entry 11).Various group 1 metal hydroxides were tested, which gave very similar enantioselectivity outcomes (entries 12−14). 18e evaluated the scope of the dearomatization and were pleased to find that phenols substituted with methyl and phenyl at the meta position also gave a high ee (Scheme 1, 2b, 2c).Conversion to 2c was low, likely because of hindrance at the forming spirocyclic stereocenter.We were curious as to whether substitution at the phenol ortho position would give good outcomes with the facial differentiation being further from the forming stereocenter.Indeed, high enantioselectivities were maintained for these substrates encompassing phenyl (2d), methoxy (2e), and methyl (2f) substituents.The limits of electronic tolerance on the phenol are displayed by an orthofluoro substitution: 2g was obtained in low yield but still remarkably high enantioselectivity given the small size of the differentiating substituent. 19Substitution at two adjacent positions of the phenol ring, including a naphthol, also worked well (2h, 2i), and the tether between the two arenes was extended to afford tetralin derivative 2j.Further extension leading to a seven-membered ring also delivered very high enantioselectivity (2k), although the low yield reflected the present reactivity limit.Substitution of the lower ring gave good outcomes with both electron-poor (2l) and -rich (2m) examples.A substrate bearing both chloride and bromide reacted selectivity at the bromide (2n, 96% ee).A Bocprotected amine was tolerated (2o), as was an ester (2p).A methyl adjacent to the bromide on the lower ring gave a significant ee reduction (2q).Finally, fluorine-containing 2r and 2s were obtained smoothly.When the reaction was scaled to 1 mmol, a small increase in enantioselectivity for 2r was observed (Scheme 1B).This led us to assess a lower 2 mol % loading of Pd (3 mol % sSPhos) at this larger scale with excellent results still obtained for 2l.
Having demonstrated the effectiveness of sSPhos on Buchwald's original arylative dearomatization scaffold, we sought to evaluate how generally applicable it might be.We next targeted arylative dearomatization of the para-aminophenol class of substrates reported by You and co-workers racemically in 2014 7a and enantioselectively in 2020 (Scheme  2).7b These substrates are notable as they map directly onto the skeleton of the Erythrina alkaloids.1b,d,2b Excellent results could be achieved with only 2.5 mol % Pd to give dearomatized 4a in good yield and excellent enantioselectivity.
Usefully, an aryl chloride could also be used as a the starting material.A larger ring in the heterocyclic starting material gave excellent results (4b), and we evaluated several substituents of varying electronic character on the lower ring (4c−4f).Dimethoxy-substituted 4g, upon hydrogenation, leads directly to (−)-3-demethoxyerythratidinone ( 5), as previously demonstrated by You and co-workers. 7o further test the generality of sSPhos, we benchmarked it on a third distinct substrate class, previously reported by Tang and co-workers who elegantly applied it to natural product synthesis (Scheme 3).8a−c,e With little modification to the conditions, excellent results could be obtained for chiral phenanthrenone derivatives related to 7a.Several analogues were demonstrated by varying the phenol para substituent (7b), as well as the lower ring substituent (7c−7e).
The substrates so far have generated products possessing allcarbon (Schemes 1 and 3) and α-tertiary amine (Scheme 2) quaternary stereocenters.We questioned whether this might be extended to O-linked substrates to form α-tertiary ethers at the spirocyclic stereocenter.Such motifs have not, to the best of our knowledge, been formed so far using arylative phenoldearomatization, even racemically.The resulting scaffold features in a number of natural products, including members of the Journal of the American Chemical Society urnucratin 20 and kadsulignan 21 families and parvifloral F. 22 Pleasingly, methyl-substituted 9a and methoxy-substituted 9b were obtained in 82% and 83% ee and 9c, which bears a chloride on the lower ring and methyls on the upper, was obtained in 93% ee (Scheme 4).The low to moderate yields are attributed to decomposition of the electron-rich starting material under the reaction conditions.Nevertheless, these results underline the generality of sSPhos as a chiral ligand for arylative phenol dearomatization, in the context of an as-yetunexplored scaffold.
We sought to probe the interactions responsible for the effectiveness of sSPhos.The anticipated pK a difference between a phenol and KOH would suggest that the potassium phenolate salt is formed under the reaction conditions, a scenario supported by NMR studies (see the Supporting Information).Phenolate formation would mean it is unlikely that ligand−substrate hydrogen bonding is occurring.We carried out the reaction in anhydrous toluene by comparing the standard phenol with a TMS-protected variant (Scheme 5A).The identical ee values provide further evidence against hydrogen bonding playing a role in selectivity because the latter conditions feature no feasible proton source.Furthermore, use of the preformed potassium phenolate salt as the substrate returned the ee to the exact value (92%) obtained when running the reaction under the optimized conditions with water.We speculate that the presence of water in the optimized conditions assists in rapid potassium phenolate formation, which is crucial for high yield and enantioselectivity.We next sought evidence for the proposed electrostatic interaction involving a bridging metal cation (Figure 1D).During optimization, no significant variation in enantioselectivity had been observed between the various alkali metal cations when they were evaluated in toluene/water (Table 1).However, differences between them were observed in dioxane, which suggests possible involvement in the selectivitydetermining step. 18Crucially, replacement of the alkali metal cation with either tetrabutylammonium or tetrabutylphosphonium was found to be detrimental to both yield and ee, thereby suggesting that favorable organization in the enantiodetermining transition state cannot be maintained with these bulky cations (Scheme 5B).Accordingly, reduction of the length of the alkyl chains in tetramethylammonium hydroxide largely restored both metrics.
We further probed the importance of the alkali metal cation by the addition of stoichiometric crown ethers of varying size (Scheme 5C). 23In toluene, almost no effect on ee was observed by the addition of 12-crown-4, as expected, given its small size relative to K + (entry 2 vs 1). 24However, 15-crown-5 and 18-crown-6 both gave reduced ee, which suggests that binding to the cation disrupts the substrate−ligand organization to some extent (entries 3 and 4).A similar outcome was observed in dioxane (entries 5−8).Finally, we sought to remove the charge from the ligand altogether to rule out the possibility that sSPhos might be exerting enantiocontrol through simple steric repulsion.Accordingly, a neopentyl sulfonate ester derivative of the ligand gave only −6% ee (Scheme 5D).The absolute stereochemistry of the products from Schemes 1, 25 2, 7b and 3 8a could all be reliably determined.In all cases, use of (R)-sSPhos is consistent with arylation occurring from the lower face of the phenol when it is depicted with its substituted side to the left and the unsubstituted to the right (Scheme 5E).
In summary, enantiopure sSPhos, easily obtained via resolution, is an extremely general chiral ligand for the Pdcatalyzed intramolecular arylative dearomatization of phenols.Using Buchwald's pioneering report, which afforded spirocyclohexadienones bearing all-carbon quaternary centers in a predominatly racemic manner, as a forum for demonstrating its effectiveness, we subsequently extended this to two other substrate classes.We also report several oxygen-linked substrates, which have not to date been explored, that give

Figure 1 .
Figure 1.Previous arylative phenol dearomatization and use of sSPhos as a bifunctional ligand.

Scheme 1 .a
Scheme 1. Scope of Arylative Phenol Dearomatization on Substrates Related to 1a a

Scheme 5 .
Scheme 5. Control Experiments to Probe Ligand−Substrate Interactions and Predictive Model

Table 1 .
Reaction Optimization a Yields determined by 1 H NMR with reference to a dibromomethane internal standard.Value in parentheses refers to isolated yield.b ee determined by SFC analysis of the crude reaction mixture, except entry 10. c Reaction temperature 90 °C.