Figure 2.

(a) The sub-picosecond time evolution of both the real and imaginary frequency-integrated photoconductivity as a function of the pump–probe delay and the pump wavelength. The used fluences were 200, 246, and 227 μJ/cm² for 1.63, 1.9, and 3.1 eV respectively. (b) Maximum of the real part of the one-dimensional conductivity normalized to the absorbed photon density. The red line is a best fit to a model described in the main text to account for the free carrier generation probability with increasing the pump energy. The model and the corresponding fitting yield an excitonic binding energy of 700 ± 50 meV. (c) Illustration of the model used here to simulate the probability of exciton dissociation from the deep Coulomb potential into free charges at the band edge by thermal excitation.