Figure 5: Comparison between theory and experiments.

(a) (circles) Time-evolution of ΔT/T at nm from Fig. 2. Theoretical data from SBE using (dotted line) dynamical screening, (solid line) regularized dynamical screening and (dashed line) static screening. All data are normalized to their maximum, and correspond to a chemical potential ~200 meV (hole-doped sample). (b) As in panel a for 1,500 nm. (c) Time t_max (in fs) at which ΔT/T reaches its maximum. The labelling of the theory data (lines) is the same as in (a,b). Experimental data in Fig. 2a are shown as a colour plot in a continuous optical spectral range for λ≲1,000 nm. The scale bar ranges from 0 to 1. The circles with error bars correspond to three IR measurements. (d) ΔT/T as a function of electron energy at different times. The slope inversion, signature of the initial stage of the dynamics, Fig. 2c, is correctly reproduced by theory. (e) Time-evolution of the electron population per unit cell (in units of eV⁻¹) as derived by the SBE with regularized dynamical screening. The initial hot-electron peak (red) is centred at half the energy of the pump laser. The amplitude of this distribution is divided by 3. The hot-electron peak rapidly broadens into a non-thermal distribution (black), which then thermalises to a hot FD (green). Subsequently, cooling by phonon emission takes place (blue), until thermal equilibrium with the lattice is eventually established (not shown here since other effects neglected in our theory, such as acoustic phonons, are important in the late stages of the dynamics). When the n(ε) peak energy crosses half the energy of the probe, Pauli blocking inhibits absorption. In this case a stronger transmitted PB signal is recorded at the detector. (f) CM (black curves) and CM efficiency ε (red curves), see Methods, as functions of time. Labelling as in panels (a-c). Note the CM suppression in the prediction based on dynamical screening (black dotted line).