Copper Mediated C(sp2)–H Sulfonylation of Aldehydes using a Catalytic Transient Imine Directing Group

Department of Chemistry, Imperial College London, Molecular Sciences Research Hub, White City Campus, Wood Lane, London, W12 0BZ (UK) ABSTRACT: The copper mediated β–C(sp)–H sulfonylation of benzaldehydes with sulfinate salts is accomplished using βalanine as a catalytic transient directing group. A broad range of sulfonylated benzaldehydes are prepared using copper fluoride as both copper source and oxidant. Both b-(ortho) and g-(peri)-sulfonylation are demonstrated. Mechanistic studies indicate the turnover limiting step to be a concerted asynchronous C–H cleavage via a Wheland-type transition state.

strategies for their incorporation and for late-stage diversification. 14 Copper has been used stoichiometrically and catalytically to form sulfones through C-H functionalization with an amide directing group, using bench stable sulfinate salts (Scheme 1b). 15

Scheme 1. C(sp 2 )-H functionalization with transient directing groups and copper mediated sulfonylation
Here we report the first copper mediated, catalytic transient directing group enabled C-H functionalization. This also represents the first C-S bond forming transient C-H functionalization process, enabling the rapid and step-efficient construction of sulfones (Scheme 1c). Kinetics  mechanism of this complex process which is catalytic in amine TDG. Initially, we aimed to establish whether copper mediated C-H functionalization reactions were possible with a catalytic TDG. Studies focused on C-H sulfonylation of benzaldehyde and o-tolualdehyde by assessing the potential of different classes of amines as TDGs. Copper (II) acetate was selected for early studies to effect both a putative concerted metalation deprotonation (CMD) process and subsequent functionalization steps, as well as to provide an inexpensive and non-toxic stoichiometric oxidant. Potassium carbonate base and HFIP, 16 used previously in Pd-mediated transient processes, were employed to screen for successful ortho-sulfonylation using 25 mol% of the transient directing groups (Scheme 2). Scheme 2. Assessing transient directing groups in copper mediated C(sp 2 )-H sulfonylation Yield determined by 1 H NMR using 1,3,5-trimethoxybenzene as an internal standard.
Bidentate transient directing groups TDG1, and TDG8-TDG10 previously reported with Pd in C(sp 2 )-H functionalization were ineffective, as were monodentate aniline and benzylamine. Pleasingly, reactivity was observed when using anthranilic acid (TDG2), forming a presumed [5,6] cupracyclic intermediate, but related orthanilic acid was ineffective. More flexible b-alanine (TDG13) was more effective, giving 40% yield of 3a. Changing the acidic secondary binding site (P(O)(OH)2, SO3H, OH or OMe; TDG14-TDG17) gave lower yield, so TDG13 was selected for further study. We next investigated reaction variables including: copper sources, solvents, base, TDG loading, and other additives. 17 CuF2 was the most effective copper source, and the yield could be further improved with carboxylate additives. Using 0.5 equiv of Cu(OAc) along with 1.5 equiv CuF2 was most effective. Further investigation confirmed K2CO3 and HFIP as the most effective base and solvent respectively. A design of experiment (DoE) optimization focused on CuF2 equivalents, aldehyde equivalents and reaction concentration identified a key positive interaction between increasing equivalents of aldehyde and CuF2. 17a Using catalytic TDG, the aldehyde in excess and 2 equiv of readily available CuF2, aldehyde 3a was isolated in 86% yield ( Table 1, entry 1). Importantly, no reaction was observed in the absence of the TDG (Entry 2). Reduced reactivity was observed when either of the copper salts were omitted (Entries 3 and 4), and when no copper salts were present there was no observable reaction (Entry 5). Using only Cu(OAc)2 (2.5 equiv) in place of CuF2 remained effective but with reduced yield (Entry 6). Low yield was observed in the absence of K2CO3 (Entry 7). Interestingly, in the absence of the sulfinate salt, ether (4) was formed in 33% yield based on the aldehyde, with solvent acting as the coupling partner (Entry 8). 18 We observed complete suppression of sulfone formation on addition of the radical trap TEMPO (Entry 9). The reaction was sensitive to increased water and oxygen as assessed by the Glorius protocol. 17b,19 Reactions performed using 0.2 mmol sulfinate salt.
a Yield determined by 1 H NMR using 1,3,5-trimethoxybenzene as an internal standard. Isolated yield in parenthesis. 1 referes to recoved excess aldehyde.
b Yield based on aldehyde as limiting reagent.
c Yield on 0.5 mmol scale of aldehyde. Note 4 is volatile.
With these optimized conditions, a wide range of electronically and sterically diverse sulfinate salts were effectively coupled to give sulfones in good to excellent yields (Scheme 3a). Sulfinate salts with electronically neutral or electron donating substituents were well tolerated. Sulfinate salts bearing arenes with inductively electron withdrawing groups (p-CF3,p-halides) were particularly effective giving yields of 63-74% (3e-3h). More sterically encumbered ortho-substituted sulfinates such as 1-napthyl and o-tolyl derivatives were successful. Methyl and cyclopropyl sulfinate salts were both highly effective in this oxidative coupling reacting in 77% and 76% yield respectively. Additionally, bicyclo[1.1.1]pentane (BCP) sulfinate was used to introduce a BCP moiety. No obvious trend between the oxidation potential of the sulfinate salts and the yield of the coupling was observed, 20 indicating direct oxidation of the salt to a sulfonyl radical was likely not a mechanistically significant step. carboxaldehyde gave peri-sulfonylated product 29a exclusively, favoring the required [6,6]-cupracyclic intermediate over the potential ortho-sulfonylation. In the case of 1napthaldehyde, peri-functionalization was again the major product in a 2:1 ratio with the less hindered ortho-product. On blocking the ortho-position with a methyl group, reduced peri-reactivity was observed. In depth mechanistic investigations of copper mediated C-H activation reactions remain rare. Cupracyclic intermediates are relatively unstable and only a handful of stable organocuprates have been reported. 22 Our attempts at direct investigation of the potential organocuprate intermediate were unsuccessful. Therefore, to interrogate the reaction mechanism kinetic and deuteration experiments were undertaken. Visual comparison methods developed by Blackmond (RPKA) 23 and Burés (VTNA) 24 were used to investigate the catalytic performance of the TDG and the orders of all reagents. 17c Initially, same-excess experiments uncovered a product inhibition effect. Studying imine formation by 1 H NMR in the absence of copper indicated a preference for the formation of the starting material imine in a competition between product and starting aldehydes. 17d Therefore, we hypothesize that inhibition results from the product imine being a more effective ligand for copper, so the TDG is temporarily removed from the catalytic cycle. Variable time normalization analysis (VTNA) was used to estimate the order in reagents and catalysts. We identified an order of 1 in aldehyde, transient directing group and K2CO3. Fractional orders were observed for both the sulfinate and CuF2 (0.75 and 0.5 respectively), and a zero order in Cu(OAc)2. Together with the optimization studies is clear that sulfinate has a competing inhibitory effect on the reaction presumably by blocking copper coordination sites. 25 No obvious deuteration was observed when performing the reaction in HFIP-d2, implying that C-H activation is irreversible under the reaction conditions. A significant primary kinetic isotope effect was observed in both parallel (KIEparallel = 2.42) and competition experiments (KIEcompetition = 3.25) which indicates C-H bond cleavage is intimately linked to the turnover limiting step (Figure 1a). 26 Further investigation into the electronics of this step by Hammett analysis revealed a negative correlation (Figure 1b). This implies a buildup of positive charge in the turnover limiting step and therefore an unusual turnover limiting electrophilic mechanism of C-H activation. 27 Studies on bases with comparable pKa showed that those capable of bidentate coordination were more effective, suggestive of an inner sphere deprotonation. 17e Considering these findings together, we propose a concerted asynchronous formation of C-Cu bond and cleavage of the C-H bond which involves nucleophilic attack from the aryl p-system and partial Wheland-type transition state. Overall, we propose the following as a plausible mechanism (Figure 1c). Catalytic amino acid A condenses with the aldehyde to form imine B. Copper and sulfinate coordinate to form intermediate C. Coordination of additional sulfinate (C') may result in inhibition of the reaction. Otherwise, coordination of carbonate to C would form intermediate D which undergoes a concerted irreversible and turnover-limiting C-H activation to form E. Loss of carbonate leads to cupracycle E. Oxidation by Cu(II) leads to formal Cu metallocycle F which would undergo reductive elimination to give copper-ligated product imine G. This imine must then undergo hydrolysis to release the product aldehyde and regenerate the amino acid catalyst. The equilibrium favoring G is likely to be the origin of product inhibition trapping the directing group at high concentration of product. In summary, the copper mediated C(sp 2 )-H sulfonylation of benzaldehydes was achieved using a catalytic b-alanine transient directing group. This demonstrates for the first time the potential to use a transient directing group with copper to promote C-H functionalization. The conditions were applied to a wide range of sulfinate salts and aromatic aldehydes including heteroaromatic aldehydes to effect CuF 2 (2 equiv) Cu(OAc) 2 (50 mol%) β-alanine (25 mol%) K 2 CO 3 (2 equiv) 1,3,5-trimethoxybenzene orthoor peri-functionalization. Kinetic and mechanistic investigations highlighted an unusual turnover limiting electrophilic C-H activation. We expect these results will open further possibilities for the use of copper species for C-H functionalization processes, particularly in combination with transient directing groups, as well as contribute to improved understanding of copper mediated C-H functionalization processes.

Supporting Information
Optimization reactions; details of DoE study; sensitivity screen; deuteration experiments, K.I.E. experiments; kinetic experiments (same excess, different excess, Hammett analysis), experimental procedures and characterization data (PDF)