Abstract
Chiral ligands coordinated to metal ions exert a selectivity towards the additional coordination of racemic substrates. Experimentally determined equilibria distributions of [Co(L^3^)~2~]^3+^ and [Co(L^3^)(L^2^)(X)]^n+^ are compared with calculated data based on strain‐energy minimization (L^3^: trap = propane‐1,2,3‐triamine; 1,2,4‐trab = butane‐1,2,4‐triamine; 1,2,3‐trab = butane‐1,2,3‐triamine; 1,3,4‐trpe = pentane‐1,3,4‐triamine; 1,3,4‐tmeb = 2‐methylbutane‐1,3,4‐triamine; 1,2,4‐trpe = pentane‐1,2,4‐triamine; L^2^: en = ethane‐1,2‐diamine; pn = propane‐1,2‐diamine; X: NH~3~, OH~2~, OH^−^). Equilibration of Co(III) complexes was achieved by oxygenation of aqueous solutions of Co(II) salts in presence of the ligands. Quantitative isomer distribution was investigated with HPLC, and quantitative analysis of the enantiomeric excess (ee) of the racemic substrate (present in a two‐fold excess) was studied, after chromatographical recovery, by ^1^H‐NMR analysis of its Mosher‐acid derivative. There is good agreement between calculated and experimental data. Systems with L = 1,2,4‐trab are, as expected, relatively poorly discriminating (ee([Co(1,2,4‐trab)~2~]^3+^) ∼ 5%; ee([Co(1,2,4‐trab)(pn)(X)]^n+^) ∼ 10%). Calculations indicate that Me substitution of the ligand backbone of 1,2,4‐trab (and trap) leads to an increased enantioselectivity (with practically constant isomer selectivity), and at the optimum site for substitution ∼ 90% ee is predicted.