“…12,35 Because the lower exciton band of the dimer is forbidden with less intense absorption transition from the ground state for an H-type structure, that as mentioned above, a low energy pre-associated excimer already exist below the lower-energy exciton state, where the relaxation from exciton state to the pre-associated excimer is very fast less than 4.5 ps or even more faster as reported previously, 36,45 and after an excitation, the excited state relaxation in the face-to-face dimer (H-type) will play a main role in excited PDI-hexamer, 42,45,46 while a sequential kinetic model is also appropriate for the global analysis of the timeresolved data of PDI-hexamer as predicted from Kasha model. 28 Althrough, the free dimer (see Figure S5) is not an ideal model for making comparison with the dimer in hexamer, several PDI-dimers with the similar face-to-face π-stacking geometry as the PDI-dimer in our PDI-hexamer have been reported, 17,38,45,46 where the spectral features and excited state dynamics observed in our PDI-hexamer are very similar to the behaviours observed in those reported PDI-dimers. Therefore, together with the conclusion from the steady-state measurements, the sequential scheme could be used for modeling the transient data of both monomer and hexamer, that is, EADS associated with certain kinetic profile represents the spectral features following that dynamics, the corresponding time constant can be regarded as the lifetime of each EADS.…”