point interaction, bounded region, principal eigenvalue, optimalization