Nonthermally activated exciton transport in crystalline organic semiconductor thin films

The temperature dependent exciton transport in the prototypical organic semiconductor di-indeno-perylene (DIP) is investigated by photoluminescence quenching. Analysis by an advanced diffusion model including interference and morphological aspects reveals an exciton diffusion length of about 60 nm at room temperature, which relates to the long-range order induced by the DIP molecular shape. Above 80 K, singlet exciton transport is thermally activated with an energy of 10 to 20 meV. Below 80 K, exciton motion becomes temperature independent and is supported by the crystalline structure of the transport layer in combination with the reduced phonon interaction.

Figure 1

(a) The intensity ratio between specular (dark shaded area) and the diffuse scattering (light shaded area) of the respective rocking curves (upper right inset) has been considered as a measure of the structural quality of the DIP layers. Inset (left): Thickness dependent x-ray diffraction scans of the (001) DIP Bragg peak at $q_z$ = 0.38 ${\AA }^{\mathrm{-1}}$, indicating the high crystallinity of the films along the direction of exciton transport. The red curve demonstrates a fit of the Laue oscillations. (b) Model of the DIP film texture, together with possible radiative (red cross) and nonradiative (black crosses) recombination processes. Integration of the Gaussian distribution of DIP crystallite heights (black curve) yields the error function (red curve).

Figure 2

(a) Sample geometry: Step-wedged layer of DIP half-side covered with a CuPc quencher. The green arrows indicate the laser excitation whereas red arrows indicate the generated PL signals. (b) Structural model of the DIP molecule. (c) Schematic picture of the spatially dependent exciton generation $n_{\mathrm{ex}}$($z$) upon light absorption and the corresponding PL signals from the respective sample areas with and without quencher. The exciton generation profile $n_{\mathrm{ex}}$($z$) sketched by the red areas under the curves is related to the boundary conditions of no quenching (upper half) and of complete quenching at the back side (lower half).

Figure 3

The fit (red line) corresponds to the advanced diffusion model discussed in the text. Accounting for interference effects and penetration at the DIP/quencher interface this model yields an exciton diffusion length $L_{\mathrm{D}}$ of 60 nm (goodness of fit ${\chi }^2$: 0.051). The green and blue curves represent simulations for smaller ($L_{\mathrm{D}}$ = 20 nm; goodness of fit ${\chi }^2$: 0.194) and larger diffusion lengths ($L_{\mathrm{D}}$ = 100 nm; goodness of fit ${\chi }^2$: 0.293) reported in literature [7, 15]. The larger the $L_{\mathrm{D}}$, the smaller the maximum of the relative quenching at 80 nm. The inset displays the spectrally resolved PL signal of a 120 nm thick DIP layer with and without CuPc quencher, demonstrating the pronounced decline of $I_{\mathrm{PL},Q}$ caused by exciton quenching.

Figure 4

Relative quenching $Q$ as a function of temperature for five representative DIP layer thicknesses normalized to the $Q$ value at room temperature. Thin DIP layers show the signature of a thermally activated relative quenching above 80 K, whereas below this temperature, the quenching ratio remains constant. Inset: Evolution of the temperature dependent PL intensity of a 30 nm thick DIP layer normalized to $I_{\mathrm{PL},nQ}$, indicating the distinct quenching with decreasing temperature.