Chiral Metal Surfaces

A research topic across the group’s papers.

A metal surface can be made enantioselective in two ways: by adsorbing chiral modifier molecules on an achiral surface, or by using an intrinsically chiral surface. Intrinsically chiral surfaces are high Miller index planes of a single crystal, such as fcc(643), whose kinked step edges have no mirror symmetry; the two mirror-image cuts can bind the two enantiomers of a chiral molecule with different energies. These surfaces are of interest for enantioselective separations and catalysis, since enantiopure compounds are needed for pharmaceuticals.

Decorating kinks with a second metal (2009)

Han, Kitchin and Sholl proposed tuning chiral surfaces by decorating their kinks with a second metal and used DFT to ask when the added atoms would stay at the kinks instead of moving to terraces or into the bulk. On Cu(643) and other surfaces they found stable kink decoration for Ag on Cu; Cu and Pt on Rh; Ag on Pd; and Pd, Pt, Cu and Rh on Ir. A descriptor-based surface segregation correlation extended this to 347 pairs, of which 206 favor kink decoration, with decoration favored on many Mo, Ru, W, Re, Os and Ir hosts and unfavorable on almost all Ag and Au hosts. Kinetics, aggregation and the actual effect on enantiospecific adsorption were left for future work (Han et al. 2009). Summary: han2009; see also Bimetallic surfaces.

Modeling enantiospecific kinetics across all Cu orientations (2024)

Abdelmaqsoud, Radetic, Fernández-Cabán, Widom, Kitchin and Gellman modeled the enantiospecific decomposition of D- and L-tartaric acid on chiral Cu surfaces, using half-time differences measured at 433 K on 507 orientations of three curved surface structure spread single crystals centered on (100), (110) and (111). A ridge regression on 58 generalized coordination number features (one model per diastereomeric pairing, since these features cannot distinguish R from S) and a ridge regression on five chiral cubic harmonics (symmetry-adapted functions antisymmetric under inversion and reflection) both reproduced the measured symmetry, with test MAEs of 138 s and 133 s. They predicted Cu(11,3,1) and Cu(17,5,2), only 1.8 degrees apart, as the most enantiospecific orientations, and the GCN model tied high enantiospecificity to a kink motif of atoms with GCN values 4.8, 5.6, 6.4 and 9.3. The data also show that surface handedness alone does not determine which enantiomer reacts faster (Abdelmaqsoud et al. 2024). Summary: abdelmaqsoud2024; see also Generalized coordination numbers.

Papers on this topic