Fig. 2 shows the excitation and emission spectra of Pyrene films with thicknesses of 1 and 50 nm. As is evident from the figure, the shape of the emission and excitation spectra of the films is independent of the excitation and emission wavelengths, respectively. This indicates fast relaxation to a common emissive species, most probably a relaxed singlet exciton state. The differences between the absorption and excitation spectra of the film suggest that some excited states relax toward this common emissive species with a much lower yield. The presence of additional emissive species that cannot be directly excited (i.e., excimers or charge-transfer states) cannot be ruled out based on the examination of the steady-state spectra alone and will be discussed later. Figure 3 shows the normalized room–temperature PL spectra of T4DIM films with thicknesses of 1, 2, 5, 15, and 50 nm. Up to 600 nm, the shape of the PL spectrum is independent of the film thickness. However, at wavelengths longer than 600 nm, the relative PL intensity increases with decreasing film thickness. It is worth noting that this trend cannot be attributed to fluorescence reabsorption, viz. to an inner filter effect, since fluorescence reabsorption causes the emission spectrum to become less intense at short wavelengths and more intense at long wavelengths. In other words, if the PL spectra of the films were being affected by fluorescence reabsorption, increasing the thickness of the film would cause the red side of the PL spectrum to become more prominent. Exactly the opposite of what we observe. Fig. 4 shows the normalized PL spectra of a 50-nm-thick film at temperatures ranging from 29 to 298 K. On the one hand, the relative intensity of the 0-0 PL transition increases with increasing temperature, as expected for an H-type aggregate. On the other hand, a shoulder centred at about 650 nm is found to decrease with increasing temperature. As reported by Spano, in the case of an H-type aggregate the intensity of the 0-0 transition increases with increasing dynamic (thermal) or structural disorder. Hence, the trend observed herein confirms that the molecules adopt an H-type packing arrangement. Figures 3 and 4 show that the shoulder centred at about 650 nm increases in relative intensity both with decreasing dynamic disorder and with increasing film thickness. Since the effects of dynamic and structural disorder are frequently found to be similar, this observation suggests that structural disorder increases with decreasing film thickness. Bearing in mind that the PL and excitation spectra of the films are independent of the excitation and emission wavelengths (Fig. 2), it must be concluded that the emission band centred at 650 nm arises from an excited state that is not directly accessible (i.e., a state that cannot be directly excited and is populated exclusively through an excited-state process). To shed further light on the mechanism whereby this state is populated we conducted time-resolved PL spectroscopy measurements. The room-temperature time-resolved PL spectra of a 50-nm-thick film can be found in Fig. 5.