How quasiparticles die in a bad metal

Department of Quantum Matter Physics, University of Geneva

Strongly interacting Fermi liquids often turn into bad metals at elevated temperature. How this crossover proceeds is largely unknown, as is the nature of the bad metal state. Here, we address this question by studying the temperature dependence of quasiparticles in the model Fermi liquid Sr₂RuO₄. In contrast to common ARPES folklore, our experiments show that quasiparticles do not disappear via a vanishing residue Z. To the contrary, we find that the residue Z increases with increasing temperature. Quasiparticles eventually disappear not by losing spectral weight but by dissolving via excessive broadening. These findings are in semi-quantitative agreement with dynamical mean field theory calculations.
We further study the non-Fermi-liquid state of Sr₂RuO₄ observed in transport measurements under uniaxial strain. To this end, we introduce a new method for precision ARPES experiments under continuously varying strain. Our data monitor the tuning the of a van Hove singularity across the chemical potential and show that quasiparticles remain intact at the critical strain. This suggests that a non-Fermi liquid behavior emerges in Sr₂RuO₄ from subtle changes in the scattering rate rather than from a breakdown of the concept of quasiparticles.