Why solid-state harmonic models need femtosecond dephasing times
In plain words
Computer models of harmonic light from crystals only match experiments if the electrons lose their quantum phase within about a femtosecond. That is much faster than measured loss of phase in the same materials, and the cause is not agreed.
Precise statement
Semiconductor Bloch equation fits to solid HHG spectra require an interband dephasing time $T_2$ of roughly 1 to a few fs to produce clean harmonic peaks, while carrier dephasing measured by other ultrafast methods in the same crystals is much longer. Question: which physical process (spatial averaging over the focal intensity distribution, propagation, electron-electron scattering at high excitation, real-space trajectory interference) produces the apparent short $T_2$. An answer is a calculation with $T_2$ set by independent data that reproduces measured spectra, or the identification of the process that mimics short dephasing.
What would settle it
A simulation with no fitted dephasing, including macroscopic propagation and focal averaging, that reproduces measured harmonic spectra and their contrast.
Status in the literature
Kilen et al. (Physical Review Letters 2020) showed that propagation effects can mimic fs dephasing; whether propagation, focal averaging or real-space trajectory interference accounts for the short $T_2$ in general is not agreed.