Impact of the glass transition on exciton dynamics in polymer thin films

In the development of organic electronics, unlimited design possibilities of conjugated polymers offer a wide variety of mechanical and electronic properties. Thereby, it is crucially important to reveal universal physical characteristics that allow efficient and forward developments of new chemical compounds. In particular for organic solar cells, a deeper understanding of exciton dynamics in polymer films can help to improve the charge generation process further. For this purpose, poly(3-hexylthiophene) (P3HT) is commonly used as a model system, although exciton decay kinetics have found different interpretations. Using temperature-dependent time-resolved photoluminescence spectroscopy in combination with low-temperature spectroscopic ellipsometry, we can show that P3HT is indeed a model system in which excitons follow a simple diffusion/hopping model. Based on our results we can exclude the relevance of hot-exciton emission as well as a dynamic torsional relaxation upon photoexcitation on a ps time scale. Instead, we depict the glass transition temperature of polymers to strongly affect exciton dynamics.

Figure 1

(a) Normalized steady-state PL spectra of rrP3HT. (b) Refractive index $n$ and extinction $k$ of rrP3HT at 10 and 350 K, respectively.

Figure 2

Top: Peak PL intensity as a function of temperature, normalized to the maximum emission spectrum. rraP3HT films were excited at 405 nm. Middle: effective Huang-Rhys factor extracted from fits with Eq. (1) on steady-state spectra of rrP3HT. Bottom: 0-0/0-1 transition ratio extracted from fits with Eq. (1) on steady-state PL spectra of rrP3HT; lines are a guide for the eye and error bars represent the standard deviation obtained from fits.

Figure 3

(a) Time-resolved PL of rrP3HT measured at 77, 250, and 350 K using an excitation wavelength of 405 nm. (b) ln($I_0$/$I$) decay dynamics of the 0-1 transition extracted from streak images. Solid lines represent stretched exponential fits using Eq. (2). The initial decay is characterized by a simple exponential ($\beta =1$), while for longer times β converges to 0.5. (c) PL decay of the 0-1 transition on a semilogarithmic scale.

Figure 4

(a) Time-resolved PL spectra of 50 nm rrP3HT films at selected temperatures (77, 250, and 350 K) representing the low-temperature ($<200$ K), intermediate-temperature (200–300 K), and high-temperature regimes (>350 K). (b) 0-0 to 0-1 transition ratio as a function of decay time. (c) Dynamic progression of the 0-0 transition. (b) and (c) are obtained from fits on streak camera spectra using Eq. (1).

Figure 5

Time-resolved PL decay dynamics taken of a 50 nm (a) rraP3HT film excited at 405 nm and (b) rrP3HT excited at 405 and 532 nm for various temperatures. The gray line shows the IRF, while solid lines on experimental data represent biexponential fits. (c) Fit results using an excitation wavelength of 405 nm; top: Amplitude $A_1$ of the first decay component; middle: decay times for rraP3HT; bottom: decay times for rrP3HT evaluated for the 0-0 and 0-1 transition. Lines are included as a guide for the eye and error bars represent the standard deviation obtained from fits.

Figure 6

Model of exciton diffusion, which is described by two regimes, namely exciton diffusion via energy transfer and exciton hopping to neighboring sites within a broad DOS. The latter process takes place at temperatures sufficient for thermal activation above the transport energy $E_{\mathrm{r}}$ (red dot) and is strongly limited at low temperatures (blue dot).