Probe-Photon-Energy-Resolved Carrier Relaxation and Electron–Phonon Interaction in 1T-TiSe2

Abstract

Ultrafast pump–probe spectroscopy is widely used to quantify electron–phonon (e–ph) coupling through the two-temperature model (TTM), which assumes that photoexcited carriers thermalize before transferring energy to the lattice, so that the extracted e–ph coupling is independent of probe photon energy. Using broadband transient-absorption spectroscopy of the layered charge-density-wave (CDW) material 1T-TiSe2, we show that the apparent coupling, gapp, instead depends strongly on probe wavelength: it overestimates the intrinsic value by up to 75% at short wavelengths and converges to g ≈ 6.7 × 1015 W m–3 K–1 only beyond a saturation wavelength λsat ≈ 670 nm. Density-functional calculations attribute this crossover to two excited-state-absorption channels that probe the nonthermal and thermalized parts of the carrier distribution. In the saturated regime, the intrinsic coupling evolves smoothly across the CDW transition with no resolvable anomaly; because an optical probe weights the momentum-averaged coupling, this null result is consistent with a momentum-selective driver of the instability. Our findings reveal a systematic probe-energy bias in a widely used analysis and provide a simple criterion for assessing TTM validity.