Geometry-Driven Selectivity in Macrocyclization via Rigid-Body Simulation
摘要
Macrocyclization under largely irreversible conditions often yields complex mixtures. A recently reported organosilicon system is a striking exception: multiple Si–O bond formations between an aromatic diol and a dichlorosilane produce an almost exclusively square cyclic tetramer at practical concentrations without detectable Si–O exchange. We recast this selectivity problem as geometry-constrained assembly of spherical nodes and cylindrical linkers undergoing stochastic encounters and irreversible bonding. Local preferences for an orthogonal bond angle and coplanarity are encoded either as rigid holonomic pose constraints or as compliant spring–damper bonds that relax toward the target pose. By varying geometric filters (valency, bond angle, coplanarity), post-bond stabilization (rigid vs. compliant), and density, we find: (i) passive geometric selection is necessary to suppress non-viable aggregates but is insufficient to yield the target in abundance; (ii) replacing rigid joints with compliant spring–damper joints shortens the time to the first target and raises the target-object fraction to \({\sim } 7\%\) , while completed cycles remain strongly enriched in squares ( \(S_4 \approx 85\%\) ); and (iii) once non-viable bonding pathways are suppressed, increasing density primarily accelerates productive encounters without eroding ring-size selectivity. These results emphasize that extreme selectivity under irreversible conditions emerges from the conjunction of stringent local geometry, post-bond stabilization, and adequate collision frequency.