Tomographic reconstruction of free-electron quantum states

TL;DR

This paper introduces multiple algorithms for reconstructing the quantum states of swift electrons from experimental data, enabling detailed analysis of their properties and coherence, with implications for ultrafast electron pulse applications.

Contribution

It presents novel algorithms including maximum likelihood, Bayesian, and deep learning methods for quantum state reconstruction of swift electrons.

Findings

Reconstructed quantum states reveal pulse durations around 245 attoseconds.

Predicted a coherence degree of 36% in the electron-induced radiations.

Validated algorithms on experimental data for attosecond electron pulses.

Abstract

We give several algorithms for reconstructing quantum states of swift electrons, using maximum likelihood estimation, Bayesian inversion, and deep learning. We apply these algorithms to data previously recorded for an attosecond electron pulse-train to retrieve the density matrix and to analyse its physical properties. Based on the reconstructed quantum state, we obtain pulse-durations of about 245as and predict a degree of coherence of 36 per cent for radiations and excitations produced by these electrons.