Measuring time at the quantum level

Physicists using the Swiss Light Source SLS have found a way to measure the time involved in quantum events and found it depends on the symmetry of the material.

“The concept of time has troubled philosophers and physicists for thousands of years, and the advent of quantum mechanics has not simplified the problem,” says Hugo Dil, a physicist at Paul Scherrer Institute PSI and professor at EPFL. “The central problem is the general role of time in quantum mechanics, and especially the timescale associated with a quantum transition.”

Quantum events, like tunnelling, or an electron changing its state by absorbing a photon, happen at mind‑bending speeds. Some take only a few tens of attoseconds (10-18 seconds), which is so short that light would not even cross the width of a small virus.

But measuring time intervals this small is notoriously difficult, also because any external timing tool can distort the very thing we want to observe. “Although the 2023 Nobel prize in physics shows we can access such short times, the use of such an external time scale risks to induce artefacts,” says Dil. “This challenge can be resolved by using quantum interference methods, based on the link between accumulated phase and time.”

Measuring quantum time without an external clock

Dil and his team from EPFL have now led research that has developed a way to accurately measure time in quantum events. When electrons absorb a photon and leave a material, they carry information in the form of their spin, which changes depending on how the underlying quantum process unfolds. By reading these tiny changes, the researchers could infer how long the transition takes, without ever using an external clock.

Read more on the PSI website

Image: Quantum events can unfold on attosecond timescales, making them notoriously difficult to measure. Researchers have now devised a way to measure the duration of quantum transitions without relying on an external clock.

Credit: © EPFL 2026/iStock (bymuratdeniz)

A new tool in attosecond science

Measuring Angle-Resolved Phases in Photoemission

Photoionization is one of the earliest observations whose explanation led to the establishment of quantum mechanics. The process is fully described by few mathematical quantities—the probability amplitudes—that are of central interest in understanding the electronic structure of matter and its theoretical foundations. Probability amplitudes are complex numbers, which are described by a magnitude and a phase. Phase information (which can be equivalently expressed as a time, i.e., a fraction of the period of the light wave causing ionization) is lost in most measurements.

An international research team from Japan, Germany, Russia, Austria, Hungary, and the local team at the FERMI free-electron laser, combined two-color XUV photoelectron spectroscopy with real-time ab initio simulations to measure phase differences with a precision of few attoseconds. The measurements, in excellent agreement with calculations, revealed a significant anisotropy with the angle of observation of the outgoing photoelectron, particularly when the frequency of the light is nearly resonant with a transition in the atom.

“In atomic and molecular physics, the phase of probability amplitudes can reveal important information about phenomena such as the concerted motion of electrons (electron correlation) in chemical reactions” says Prof. Kevin Prince from Elettra – Sincrotrone Trieste “and our work provides a new tool for attosecond science, i.e., the observation in real time of the motion of electrons inside matter.”

Read more on the ELETTRA website

Image: Scheme of the experiment: Bichromatic, linearly polarized light (red and blue waves), with momentum kg and electric vector Eg, ionizes neon in the reaction volume. The electron wave packets (yellow and magenta waves) are emitted with electron momentum k. The averaged phase difference  between wave packets created by one- and two-photon ionization depends on the emission angle. The photoelectron angular distribution depends on the relative (optical) w‑2w phase f. Lower figures: Polar plots of photoelectron intensity at Ek=16.6 eV for coherent harmonics (asymmetric, colored plot) and incoherent harmonics (symmetric, gray plot).

Credit: Reproduced from You et al., Phys. Rev. X, 10, 031070 (2020) doi: 10.1103/PhysRevX.10.031070; copyright 2020 by the Authors. The original figure has been published under a Creative Commons Attribution 4.0 International license (CC BY 4.0) http://creativecommons.org/licenses/by/4.0/