Initially, model reaction by condensation of
γ-bis-methyl-
α-ethylfuranone
1a and benzal acetone
2a was used to optimize the reaction condition. The corresponding results are shown in
Table 1. We firstly investigated the activity of catalysts
Ⅰ-Ⅸ, a range of chiral primary amine catalysts. We envisaged that a catalyst with neighboring bulky substituents of the primary amine group could improve the enantioselectivity. Therefore, the (1
S, 2
S)-cyclohexane-1,2-diamine
Ⅰ was chosen as the catalytic backbone, which is readily available from cheap amino acids, and the
tert‑butyl group was introduced into the catalyst backbone to investigate the effect of secondary and tertiary amines on enantioselectivity and product yield (
Table 1, entries 1–9). Unfortunately, changing the substituent groups at the R
1 and R
2 positions of catalyst
Ⅱ and
Ⅲ, which with a
tert‑butyl group at the R
3 position, did not improve the enantioselectivity (
Table 1, entries 2 and 3). Moreover, the catalysts
Ⅳ, Ⅴ, Ⅵ and
Ⅶ with different substituents also failed to increase the efficiency (
Table 1, entries 4–7). To our delight, the use of (1
R, 2
R)−1,2-diphenylethane-1,2-diamine
Ⅷ as catalyst resulting in good yield, enantioselectivity and diastereoselectivity (
Table 1, entry 8). Next, the tridentate N
3-catalyst
Ⅸ provided ineffective (
Table 1, entry 9). Subsequently, we optimized the reaction by using various additives in the presence of catalyst
Ⅷ and found that both weak acid additives AcOH and PhCO
2H and stronger acid CF
3CO
2H and
p-TsOH gave lower enantioselectivity and yields than N-Boc-L-Phg (
Table 1, entries 10–13). Further screening of the solvent indicated that
iPrOH is better than DMSO, PhMe, MeOH, EtOH, and THF (
Table 1, entries 14–18). Finally, control experiments demonstrated that the organo-catalyst and acid additive are both essential for the success of this transformation (
Table 1, entries 19 and 20).