A Degenerate Metal-Templated Catalytic System with Redundant Functional Groups for the Asymmetric Aldol Reaction

A degenerate zinc-templated catalytic system containing two bipyridine ligands with redundant functional groups for either enamine or hydrogen bond formation was applied to the asymmetric aldol reaction. This concept led to both a higher probability of reaction and rate acceleration. Thus, the catalyst loading could be decreased to a remarkable 2 mol % in what we think is a general approach.


SYNTHESIS AND CHARACTERIZATION OF
Scheme S1. Coupling between bipyridine 8 and Boc-proline.
1.10 g (1.89 mmol, 1 eq) of Boc-protected 9 was dissolved in a mixture of dry DCM (11.34 mL) and TFA (3.78 mL). Then, 4.65 mL of triisopropylsilane were added and the mixture further stirred at rt for 2 h. Then, the liquids were evaporated and the resulting product was washed with Et2O, dissolved with water and lyophilized. The obtained powder was purified by reversedphase chromatography using a gradient of H2O (100 -0 %) -ACN (0 -100%). The fraction containing the product was collected and the solvents were evaporated. Then, the obtained powder was dissolved in CHCl3 and a solution of NaOH 2 M was added to make the solution weakly alkaline. Afterwards, the solution was extracted with CHCl3 (3 × 50 mL) and the combined organic extracts were dried over anhydrous Na2SO4. The solvent was removed with a rotatory evaporator to afford 124 mg of 9 (56 % yield []D 25 ºC +79.2 (c=0.5, MeOH).
2,2'-bipyridyl-5,5'-dicarboxylate methyl ester (272 mg, 1 mmol) was dissolved in 3 ml CH2Cl2 and 6 ml MeOH and cooled to 0 ºC. Then, hydroxylamine (4 ml 50% in water, 60 mmol) and sodium hydroxide (160 mg, 4 mmol) were added. The reaction was stirred at 0 ºC and left reacting overnight while warming up to rt. A bright yellow precipitate is formed. Ethanol was added and the mixture filtered. The precipitate was washed with ethanol and dried, furnishing 275 mg of dihydroxamic acid 4 (quantitative yield). Described in ref. [2]. A mixture of the three catalyst components was stirred in 300 μL of dry THF and 10 μL of H2O for 1 h. Then, 90 mg of p-nitrobenzaldehyde (0.595 mmol, 1 eq) and 617 μL of cyclohexanone (5.95 mmol, 10 eq) were added. When the reaction finished, H2O was added and the solution was extracted with EtOAc. The organic phase was dried over anhydrous MgSO4, filtered and the solvents evaporated in the rotavap. The crude product was purified by silica gel column chromatography using a gradient of Hexane (100 -50 %) / EtOAc (0 -50 %) to give 110 mg of the aldol product (74 % yield). Chiral analysis was performed using a Chiralpak IB column eluting with n-hexane / isopropanol (90/10) mobile phase, flow 1 mL/min, 30 minutes.  The graphical procedure developed by J. Burés was used.

S21
Since three different zinc complexes with two ligands can arise from a 1:1:1 mixture of Zn:bipyPro2:bipyHA2, ranging from the statistical 25:50:25 distribution to 100% Zn(bipyPro2)(bipyHA2), the validity of the catalyst order calculation could be challenged.
However, firstly, we had found that the equilibrium constant formation for our previous bipyridine based ZnL1L2 complex was 3.5x10 11 M -1 , whereas the Zn(L1)2 and Zn(L2)2 equilibrium constants were 4x10 10 and 1x10 11 M -1 , in line with the literature. These values led to the predominant formation of 71% ZnL1L2, and 14% of Zn(L1)2 and Zn(L2)2 each, overcoming the statistical distribution. [4c] Secondly, as a consequence of these large formation constants the zinc complexes distribution does not change, whatever it is, at least in the range of concentrations used for catalysis (from 4 to 20 mM). This is, the different zinc complexes are not under real reversible equilibrium. See Figure S16 below.
Finally, since the catalyst order determination is therefore not affected by the real concentration of catalyst, which remains proportional to the amount introduced initially into the reaction, we have used for the catalyst order determination the weighed molar amount of the different components.  These equations were derived to obtain rate equations, which furnished the TOF for each catalyst upon division by the ligand (bipyPro or bipyPro2) concentration: For comparison, since bipyPro2 has double the catalytic sites than bipyPro, TOF0[bipyPro2] must be divided by 2 (11 / 2 = 5.5 h -1 vs. 2.2 h -1 from bipyPro), or alternatively, TOF0[bipyPro] must be multiplied by 2 (2.2 x 2 = 4.4 h -1 vs. 11 h -1 from bipyPro2). In any case, the catalytic system containing bipyPro2 is significantly faster. Indeed, the system Zn(TFA)2+bipyPro2+bipyHA2 is 5 times faster, not twice faster as it could be expected by doubling the number of catalytic sites in the ligand.

PROPOSED REACTION MECHANISM AND ORIGIN OF STEREOSELECTIVITY
This proposed mechanism must be understood as based on current and previous research of our group.
[4c] Please, keep in mind that it is a speculative mechanism lacking comprehensive kinetic and theoretical studies. It is presented herein to give the vision of the authors on the catalyst operation (Please, check also Sections 8 and 9 in this Supporting Information). We propose a working mechanism in which a predominant zinc-templated bifunctional catalyst behaves as the catalytically most active species. The Zn(bipyPro2)(bipyHA2) complex would work via an enamine mechanism with cyclohexanone using one prolinamide moiety and attract the aldehyde through hydrogen bonding through one of the hydrazide groups. In this way the asymmetric aldol reaction could take place in a stereocontrolled fashion. The observed diastereo-and enantioselectivity can be rationalized in this way too.