Precision spectroscopy of fast, hot, exotic isotopes using machine-learning-assisted event-by-event Doppler correction

Open Access

Precision spectroscopy of fast, hot, exotic isotopes using machine-learning-assisted event-by-event Doppler correction

S. M. Udrescu, D. A. Torres, and R. F. Garcia Ruiz

Phys. Rev. Research 6, 013128 – Published 31 January 2024

Abstract

We propose an experimental scheme for performing sensitive, high-precision laser spectroscopy studies on fast exotic isotopes. By inducing a stepwise resonant ionization of the atoms traveling inside an electric field and subsequently detecting the ion and the corresponding electron, time-, and position-sensitive measurements of the resulting particles can be performed. Using a mixture density network, we can leverage this information to predict the initial energy of individual atoms and thus apply a Doppler correction of the observed transition frequencies on an event-by-event basis. We conduct numerical simulations of the proposed experimental scheme and show that kHz-level uncertainties can be achieved for ion beams produced at extreme temperatures ( $> 10^{8}$ K), with energy spreads as large as 10 keV and nonuniform velocity distributions. The ability to perform in-flight spectroscopy, directly on highly energetic beams, offers unique opportunities to study short-lived isotopes with lifetimes in the millisecond range and below, produced in low quantities, in hot and highly contaminated environments, without the need for cooling techniques. Such species are of marked interest for nuclear structure, astrophysics, and new physics searches.

Received 23 April 2023
Revised 21 November 2023
Accepted 21 December 2023

DOI:https://doi.org/10.1103/PhysRevResearch.6.013128

Published by the American Physical Society under the terms of the Creative Commons Attribution 4.0 International license. Further distribution of this work must maintain attribution to the author(s) and the published article's title, journal citation, and DOI.

Published by the American Physical Society

Physics Subject Headings (PhySH)

Atomic & molecular processes in external fields Atomic & molecular structure Nuclear shapes and moments Single- and few-photon ionization & excitation

Atomic & molecular beams Exotic atoms & molecules

Machine learning

Atomic, Molecular & OpticalNuclear PhysicsNetworks

Authors & Affiliations

S. M. Udrescu ^1,*, D. A. Torres ², and R. F. Garcia Ruiz ¹

¹Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA
²Departamento de Física, Universidad Nacional de Colombia, Bogotá 111321, Colombia

^*sudrescu@mit.edu

Article Text

Click to Expand

References

Click to Expand

Issue

Vol. 6, Iss. 1 — January - March 2024

Subject Areas

Reuse & Permissions

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
Illustration of the experimental setup. The neutral atoms resulting from the neutralization process (not shown) enter the interaction region where they are excited to a higher-lying electronic state by a collinear laser (I). The excited atoms are then ionized by a standing-wave laser inside an optical cavity (II). The resulting electrons are detected by a position-sensitive detector located above the ionization point (III), while the ions continue their trajectories in the electric field produced in between two parallel plates, until they reach a second position-sensitive detector (IV). The direction and magnitude of the electric field experienced by the ions is shown in light blue.
Reuse & Permissions
Figure 2
Architecture of the employed MDN. The initial ( $x_{i}$ ) and final ( $x_{f}$ ) positions of the ions as well as their time of flight ( $t$ ) are used as inputs (green disks). They are passed to a hidden layer containing ten nodes (orange disks), each with an exponential linear unit (ELU) activation function, shown in the magnified view. For each input, the MDN predicts a mean and standard deviation for the energy distribution (blue disks), from which the energy of the event can be sampled.
Reuse & Permissions
Figure 3
Results of the energy and frequency reconstruction. The predicted energy error for ${}^{120}{Sn}$ over all events is normally distributed with a standard deviation of 2.3 eV (a). This corresponds to an event-by-event reconstructed rest frame transition frequency with an error around 2 MHz when 100 target atoms events are detected, which is further decreased to only 2.5 kHz for $10^{8}$ events (b). For ${}^{8}B$ , an energy spread of 0.4 eV is obtained (c), corresponding to an uncertainty on the rest frame transition frequency of $\sim 1.2$ MHz for $10^{4}$ events (d). For ${}^{48}{Ni}$ , the energy uncertainty is reduced to 77 eV (e), leading to a reconstructed rest frame transition frequency uncertainty at the MHz level, for $10^{6}$ events (f).
Reuse & Permissions
Figure 4
Results of the frequency extraction for ions with non-Gaussian distributed initial energies for the ${}^{120}{Sn}$ case. The energy histograms when the reference and target atom having the same and different initial energy distributions are shown in (a) and (b), while the associated reconstructed rest-frame frequency of the target atom using our proposed method, relative to the true rest-frame frequency, $ν_{0}$ , is shown in (c) and (d). (e) Difference, $Δ ν$ , between true and reconstructed frequency, when using an MDN (blue circles) and when inferred from the distance and time information of the target atom only (red filled circles). The results are shown for different tilt angles of one electrode plate relative to the other (see main text for details).
Reuse & Permissions

Physical Review Research