Questions, Representations and Evidence
I transform three-dimensional structures and electronic states into information for both predicting reaction selectivity and interpreting its physical-organic origins.
Question
When reaction faces, sites, or substrates differ in reactivity, I examine which spatial regions of steric and electronic effects account for the difference. No single descriptor is treated as a mechanism by itself; experiments, transition states, three-dimensional fields, and statistical models are compared.
Electron density, ESP, orbitals, interaction fields, and steric descriptors are aligned in a common coordinate system to represent differences between reaction faces.
Nucleophilic additions to ketones, asymmetric reductions, and competition experiments are analyzed through product ratios, rate ratios, and transition-state energies.
Regularized regression, PCA, and SHAP are used to balance external prediction and chemical interpretation in small-sample, high-dimensional data.
From Question to Evidence
Begin with explicit structures and targets, then connect three-dimensional fields and statistical models to independent chemical evidence. Conditions are fixed at each stage, and validation can send the analysis back to an earlier stage.
Define
Fix reaction face, product identity, rate ratio, temperature, and sign.
Explore
Search conformers and competing pathways, then examine coverage.
Calculate
Evaluate density, ESP, orbitals, and NCI/IGMH in common coordinates.
Model
Pass steric descriptors, regional integrals, and voxel fields to validated models.
Validate
Compare with transition states, substituent effects, energy decomposition, and further experiments.