Can a laser flash change a metal's Fermi surface topology?
A femtosecond infrared pulse briefly pushes an empty electron pocket in MoTe2 below the Fermi level, changing its Fermi surface topology for under a picosecond.
Source
Ultrafast dynamical Lifshitz transition
Study at a glance
- Design
- Other — Pump-probe time-resolved multidimensional photoemission on a cooled crystal, combined with time-dependent DFT+U simulations
- N
- No sample N; bulk Td-MoTe2 crystals cleaved and measured at 30 K across pump-probe delays
- Population
- Bulk crystals of the type-II Weyl semimetal Td-MoTe2
- Outcome
- Time evolution of the Fermi surface and energy position of the γ electron pocket after infrared excitation
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The researchers pumped Td-MoTe2 crystals at 30 K with 1030 nm infrared pulses and probed them with 21.7 eV extreme-ultraviolet pulses, recording photoelectrons in energy, both in-plane momenta and time with a momentum microscope. They tracked the electron pockets near the Y point and tested four explanations: simple filling of empty states, light-dressed Floquet states, a structural phase change, or reduced electron-electron repulsion (Hubbard U). Time-dependent DFT+U simulations with a frozen lattice modelled the electronic response.
What they found
After the pump, the γ pockets appeared on the Fermi surface, the signature of a Lifshitz transition, and vanished within about a picosecond. The pocket bottom shifted down by about 70 meV, crossing the equilibrium Fermi level by 17 ± 7 meV, and the shift lagged the pump, ruling out Floquet states. The pump fluence was below the structural transition threshold and heated the lattice only to about 71 K, far below 250 K. Simulations reproduced the transition only when both photoexcited populations and a drop in Hubbard U were included, and the needed drop in U was much smaller than equilibrium calculations would require.
The limits
What it doesn't show
The simulations assume a frozen lattice and omit electron-phonon coupling, so they cannot reproduce the recovery and are only qualitative. Time resolution was limited to about 140 fs by the long pump pulse. The structural-transition scenario is excluded partly by comparison with a separate optical study and a heating estimate rather than a direct structural probe in this experiment. Only one material was studied, so generality to other correlated materials is argued, not shown.
Key terms
- Fermi surface
- The surface in momentum space separating occupied from unoccupied electron states in a metal.
- Lifshitz transition
- An abrupt change in the topology of the Fermi surface, such as a new pocket appearing when a band crosses the Fermi level.
- Hubbard U
- A parameter describing the energy cost of on-site electron-electron repulsion, used to correct DFT for correlations.
- Time-resolved photoemission
- A pump-probe technique measuring electron energies and momenta ejected by a delayed light pulse to film band-structure changes.
- Weyl semimetal
- A material whose conduction and valence bands touch at isolated points acting as sources or sinks of Berry curvature.
- Floquet state
- A light-dressed electronic state that exists only while a periodic driving field is present.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What is a Lifshitz transition?
Common questions
Why is MoTe2 a good candidate?
Equilibrium calculations put it close to a Coulomb-driven Lifshitz transition, so a modest change in correlation can move the pocket across the Fermi level.
How did they rule out laser heating driving a structural change?
Their absorbed fluence was below the known structural threshold, the estimated lattice temperature only rose to about 71 K, and no coherent phonons were seen.
Why does a smaller change in U suffice out of equilibrium?
Nonequilibrium electron populations and the reduced U act together; neither alone produced the transition in simulations, showing nonadiabatic effects matter.
More on Phase transitions