Coetaneous catalytic kinetic resolution of alkynes and azides through asymmetric triazole formation

A non-enzymatic simultaneous (coined coetaneous) kinetic resolution of a racemic alkyne and racemic azide, utilising an asymmetric CuAAC reaction is reported. The use of a CuCl (R,R)-Ph-Pybox catalyst system effects a simultaneous kinetic resolution of two racemic starting materials to give one major triazolic diastereoisomer in the ratio 74:12:4:10 (dr 84:16, 90% ee maj). The corresponding control reaction using an achiral copper catalyst gives the four possible diastereoisomers in a 23:27:23:27 ratio, demonstrating minimal inherent substrate control.

A non-enzymatic simultaneous (coined coetaneous) kinetic resolution of a racemic alkyne and racemic azide, utilising an asymmetric CuAAC reaction is reported. The use of a CuCl (R,R)-Ph-Pybox catalyst system effects a simultaneous kinetic resolution of two racemic starting materials to give one major triazolic diastereoisomer in the ratio 74:12:4:10 (dr 84:16, 90% ee maj). The corresponding control reaction using an achiral copper catalyst gives the four possible diastereoisomers in a 23:27:23:27 ratio, demonstrating minimal inherent substrate control.
Catalytic kinetic resolution (KR) occurs when one enantiomer of a racemic substrate is preferentially activated towards reaction by a chiral catalyst (through competing diastereomeric transition states), leading to more rapid formation of one enantiomer of product. At 50% conversion of starting racemic material, effective catalytic KR will have occurred if high ee of product and high ee of unreacted starting material are obtained. The effectiveness of a kinetic resolution may be judged by a criterion named selectivity factor (s). Selectivity factor is the ratio of the rate constants for reaction of each enantiomer in a given asymmetric transformation 11 . Enzymes are capable of performing catalytic KR, albeit under a narrow range of conditions with limited substrate scope 12,13 . Kinetic resolution has been widely studied 11 . Fu and co-workers have championed catalytic KR, applying planar chiral DMAP-derivative catalysts to the successful KR of secondary alcohols [14][15][16] . Catalytic KR has also been successfully employed in copper-catalysed azide-alkyne cycloadditions leading to enantioenriched chiral triazoles and the recovery of enantioenriched starting materials (Scheme 1i and ii) [17][18][19][20][21] , and complete consumption of starting materials in the case of dynamic kinetic resolution 22 . Desymmetrisation by asymmetric triazole formation has also been successfully achieved [23][24][25] .
Parallel kinetic resolution is a well-established field, where a single chiral starting materials' enantiomers undergo simultaneous divergent asymmetric transformations yielding different enantioenriched products from either enantiomer of starting material [26][27][28][29] . For example Fu and co-workers utilised parallel kinetic resolution to resolve 4-alkynals (Scheme 2) 30 .
Herein, we investigate a simultaneous, rather than parallel, kinetic resolution of two racemic substrates, under control of a single chiral catalyst, and coin the term coetaneous resolution to describe it. Upon coetaneous resolution of two racemic substrates, the ideal scenario would be formation, at 50% conversion, of a single diastereoisomer of the product formed from one enantiomer of each substrate. This ideal process would leave the opposite

Results and Discussion
Based on previous work on catalytic kinetic resolution of alkynes and azides, we chose the PyBox ligand family for creation of a chiral copper catalyst and selected (R,R)-Ph-PyBox as a suitable ligand for our investigations 20,25,33,34 . To investigate the potential for coetaneous catalytic kinetic resolution we focused on substrates with demonstrable efficacy in standard catalytic kinetic resolutions. Quaternary oxindole 1 (as employed in Scheme 1i), was selected as the alkyne-containing component. Oxindoles are important, biologically-relevant, scaffolds having found wide application, including as calcium channel blockers 35 , anti-angiogenics 36 , antitumour agents 37-39 and analgesics 35 . Azide 3 (as employed in Scheme 1ii) was chosen owing to its previously reported application in the first catalytic kinetic, copper-catalysed triazole forming, resolution by Fokin and co-workers 20 .
We have previously explored the selectivity of alkyne 1 towards kinetic resolution and found an intriguing solvent dependency upon selectivity 18 , however resolution of azide 3 had not been employed under those same conditions (shown in Scheme 1i). In order to probe this azide 3 was reacted with 0.5 equivalents of phenyl acetylene, 15 mol% L1 and CuCl (12.5 mol%) in acetone-d 6 and the reaction progress was monitored in-situ via proton nuclear magnetic resonance ( 1 H NMR) spectroscopy.
From the resolution of azide 3 in acetone a selectivity factor of s = 7.4 was determined ( Table 1, entry 1). Since, resolution of alkyne 1 had been shown to be superior, in a previous study, when 2,5-hexanedione had been employed as reaction solvent the resolution of 3 was repeated using this dione solvent (unlocked 1 H NMR spectroscopy reaction monitoring, see ESI) which gave conversions and selectivity in line with that when acetone was used as solvent (s = 7.1), ( Table 1, entries 1 and 2 versus 3 and 4) suggesting that the dione solvent-effect is manifest primarily in alkyne rather than azide selectivity.
The coetaneous kinetic resolution of 1 and 3 was then attempted. To our delight, it was found that a mixture of 1:1 of 3 and 1 in the presence of 15 mol% L1 and 12.5 mol% CuCl catalyst provided a 74:12:4:10 diastereoisomeric ratio of product 6 (Scheme 3iii). This showed that indeed the reaction successfully resolved the two starting materials simultaneously.
To enable comparison of the selectivity for each enantiomer in this diastereo-and enantio-selective process the selectivity factors for each enantiomer of alkyne 1 and azide 3 were determined as follows. Using enantiopure 30 Path (i) Reaction of enantiomer (R) to form cyclobutaneone product. Path (ii) Reaction of enantiomer (S) to form cyclopentenone product. substrates as the limiting reagent the selectivity towards that component could be analysed (Table 2). The absolute configuration of 1 was evidenced by X-ray diffractometry of a single crystal of an iodo alkyne derived from one enantiomer of enantiopure alkyne 1 (see compound 7 ESI) and the absolute configuration of 3 by comparison to it and by employing in the synthesis literature data and protocols (see ESI). Thus, also allowing for assignment of the absolute stereochemistry of the products (6, as noted in Scheme 3, and see ESI).

Scheme 2. Parallel Kinetic Resolution of 4-Alkynals, Fu and co-workers
From the data presented in Table 2 it can be seen that consumption of racemic alkyne is faster and more selective (in a reaction catalysed by a catalyst derived from L1) in combination with (S) azide. As well as the observation that consumption of racemic azide is essentially equally rapid in combination with either enantiomer of alkyne (in a reaction catalysed by a catalyst derived from L1), with only a slight difference in selectivity, it being subtly better with (R) alkyne. These, and earlier, data reveal two important features, firstly issues of selectivity are more dependent upon alkyne than azide stereochemistry (in this example) and secondly that across these four experiments with a single enantiomer component the faster reacting and more selective examples involve (S) azide and (R) alkyne, which corresponds to the major product (of Scheme 3 iii) being formed from these same two enantiomers. Thus, adding support for the hypothesis that a coetaneous kinetic resolution is taking place.

conclusions
These preliminary findings are to the best of our knowledge the first non-enzymatic example of two racemic starting materials being successfully kinetically resolved by the same catalyst to an enantioenriched diastereomeric product. We recognise that in this first study substrate scope is limited and hope this strategy can be applied to other types of substrates and increase the efficiency of resolution procedures. It is interesting to consider if this kind of selectivity may be operating in any systems of nature and we hope to be able to explore the scope and mechanistic aspects of this reaction.
General procedure for the kinetic resolution of 3 with Alkyne 1. To an oven dried Radley's multi-reactor tube was added L1 (6.70 mg, 0.018 mmol, 15.0 mol%) and CuCl (1.50 mg, 0.015 mmol, 12.5 mol%) followed by 2,5-hexanedione (1 mL), the resulting solution was allowed to stir at rt for 1 h. After this time compound 1 (16.7 mg, 0.06 mmol, 0.5 equiv.) dissolved in 2,5-hexanedione (0.5 mL) was added. The reaction mixture was allowed to stir at rt for a further 15 mins after which it was cooled to 0 °C in an ice bath and stirred for a subsequent 15 mins. After this time had passed azide 3 (17.8 mg, 0.12 mmol, 1 equiv.) dissolved in 2,5-hexanedione (0.5 mL) was added. The reaction mixture was stirred for 96 h at 0 °C before being quenched by the addition of aqueous ammonia 5% v/v (5 mL). The resulting solution was extracted with EtOAc (2 × 10 mL), the combined organic fractions were dried over MgSO 4 and concentrated under reduced pressure. Chiral GC was carried out on the crude recovered material to measure the ee of the remaining azide 3. The remaining crude material was the purified by automated flash column chromatography Combiflash Rf (0-100% hexane/EtOAc gradient, 12 mins). The dr and ee of the triazolic product was then determined by chiral HPLC.
General procedure for simultaneous kinetic resolution of 1 and 3. To an oven dried Radley's multi-reactor tube was added L1 (6.70 mg, 0.018 mmol, 15.0 mol%) and CuCl (1.50 mg, 0.015 mmol, 12.5 mol%) followed by 2,5-hexanedione (1 mL), the resulting solution was allowed to stir at rt for 1 h. After this time compound 1 (33.4 mg, 0.12 mmol, 1.00 equiv.) dissolved in 2,5-hexanedione (0.5 mL) was added. The reaction mixture was allowed to stir at rt for a further 15 mins after which it was cooled to 0 °C in an ice bath and stirred for a subsequent 15 mins. After this time had passed azide 3 (17.8 mg, 0.12 mmol, 1 equiv.) dissolved in 2,5-hexanedione (0.5 mL) was added. The reaction mixture was stirred for 96 h at 0 °C before being quenched by the addition of aqueous ammonia 5% v/v (5 mL). The resulting solution was extracted with EtOAc (2 × 10 mL), the combined organic fractions were dried over MgSO 4 and concentrated under reduced pressure. Chiral GC was carried out on the crude recovered material to measure the ee of the remaining azide 3. The remaining crude material was the purified by automated flash column chromatography Combiflash Rf (0-100% hexane/EtOAc gradient, 12 mins). The dr and ee of the triazolic product 6 and ee of the recovered alkyne 1 was then determined by HPLC with a chiral stationary phase.
Supplementary information is available that includes detailed experimental procedures, NMR spectrums, HPLC traces and X-ray crystallographic information. Citations therein should be referred to in relation to published procedures [40][41][42][43][44] and a pre-peer reviewed preprint was submitted prior to peer assessment of this manuscript 45 .