Fig. 2: Monitoring charge separation dynamics using pump–probe (PP) spectroscopy.

a–c Normalized PP transient absorption response (∆T/T) for pristine P3TEA, P3TEA:SF-PDI₂, and P3TEA:PCBM near the absorption edge at room temperature (photoexcited at 670 nm, 1.5 µJ cm⁻² per pulse). Non-normalized data over full detection range can be found in Supplementary Figs. 8 and 9. At early times (0.2 ps; panel a), the non-fullerene blend shows similar PP response to the pristine polymer which is associated with photogenerated excitons, while a negative ∆T/T signal from ~720 nm onwards associated with charges (hole polarons) has already emerged for the fullerene blend. As detailed in the main text, this negative signal is due to the growth of a transient electroabsorption (EA) signal caused by charge separation across the D/A interface as well as polaron absorption. With increasing time delay (20 ps; panel b), a similar negative ∆T/T signal emerges in the non-fullerene blend, but not evolving to match the fullerene ∆T/T signal until ~200 ps (panel c) at which both signals are similar to the quasi-steady-state EA response measured in a diode structure (Ref. EA). This indicates that charge separation in the non-fullerene blend takes much longer time to complete compared to the fullerene blend. d PP response of P3TEA:SF-PDI₂ at reduced temperatures. At early time (1 ps; inset) the response (associated mostly with singlet excitons) is largely unchanged with respect to that of the pristine polymer (see Supplementary Fig. 10). We note that the slight spectral red-shift (≤20 nm) of the ground-state bleach with reduced temperature is due to lowering of the optical gap. At later times (~200 ps), we find much weaker signs of transient EA signal created by charge separation at reduced temperatures (as evidenced by the much reduced spectral blue-shift), thus implying that it is indeed an endothermic process. In contrast, reducing temperature has insignificant effect on charge separation in the fullerene blend, where a clear EA signal is seen even at ~20 K (see Supplementary Fig. 10).