Kinetics of high-density plasmas generated in Si by 1.06- and 0.53-μmpicosecond laser pulses

The dynamics of high-density (${10}^{19}$${10}^{21}$ ${\mathrm{cm}}^{\mathrm{\ensuremath{-}}3}$), high-temperature (300${T}_{e}$3000 K) plasmas, generated in silicon by single 1.06- and 0.53-\ensuremath{\mu}m picosecond laser pulses, is simulated with a self-consistent model based on the Boltzmann transport equation. Balance equations, which include possible plasma degeneracy effects, are used to obtain the spatial and temporal evolution of the lattice temperature as well as the carrier density and temperature. We compare predictions of the model, which employs parameters extracted from near-equilibrium experiments, with data from recent experiments which measured the melt threshold or time-dependent infrared reflectivity. The model is able to account quantitatively for the weak dependence of the melt threshold on pulse width at 0.53 \ensuremath{\mu}m as well as the strong dependence at 1.06 \ensuremath{\mu}m. In the former case, direct absorption and lattice parameters dominate energy evolution, but at 1.06 \ensuremath{\mu}m two-photon absorption and plasma dynamics dominate. The infrared reflectivity measurements are found to be consistent with a simple Drude model of the dielectric constant and the calculated plasma kinetics when one accounts for nonzero probe pulse widths and plasma spatial inhomogeneity. When this is done there is no need to introduce ultrashort momentum relaxation times or plasmon assisted recombination as speculated on in the recent literature. Finally, contrary to earlier suppositions, it is found that the carrier temperature can exceed the lattice temperature by more than ${10}^{3}$ K during pulses as long as 100 psec even for subpicosecond carrier-phonon relaxation times. It is also found that the carrier temperature exhibits two peaks as a function of time, one due to direct laser heating of the carriers and the other due to Auger heating.