An event generator for hard probes of the preequilibrium stages of ion collisions at the LHC
At the Large Hadron Collider (LHC) at Geneva, collisions of nuclei are used to create a thermodynamic system described by fluid dynamics under extreme conditions of temperature and energy density. The temperature of the short-lived system is sufficiently high in order to release the building blocks of matter at a subnucleonic scale, quarks and gluons. This state of matter is commonly called Quark Gluon Plasma (QGP). The space-time evolution of heavy-ion and light-ion collisions at the LHC is described by close-to-ideal hydrodynamics after a short pre-equilibrium lapse of time. However, key features of the early stages of these collisions are largely unknown. These characteristics are crucial to understand the applicability limits of hydrodynamics as well as the thermalization process of a strongly interacting system.
In recent publications, it was pointed out that the dilepton production in the intermediate mass scale between 1.5 and 5 GeV/c² is highly sensitive to the ´thermalization´ time scale towards the equilibrium QGP.
This thesis is dedicated to the development of an event generator based on state-of-the-art theoretical calculations for the production of charm quarks from the preequilibrium in hadronic collisions at the LHC. The generator will allow to generate simulations adapted for the usage in the LHC experiment and to simulate the preequilibrium signal within general purpose event generators. Machine learning methods will be used for the tool optimisation.
ENGINEERING SUPERCONDUCTING SURFACES FOR PARTICLE ACCELERATORS AND QUANTUM DEVICES
This project will establish atomic layer deposition (ALD) as a versatile surface engineering approach for superconducting resonators, addressing two complementary objectives for particle accelerators and quantum technologies.
The first objective is controlled doping of bulk niobium superconducting radiofrequency (SRF) cavities. Conformal films up to 10 nm thick will provide solid dopant sources, followed by high-vacuum annealing to tailor dopant profiles within the RF-active region. This strategy aims to improve uniformity, explore alternative dopants and eliminate subsequent electropolishing. Particular attention will focus on suppressing hydride formation while preserving the superconducting energy gap. Target performances include quality factors of 8 × 10¹° at 2 K and accelerating gradients of 40–45 MV/m. Supported by the European EPITA project, this work addresses requirements for future accelerators, including the Future Circular Collider.
The second objective is developing protective coatings with minimal dielectric losses by mitigating defects, particularly two-level systems. Building on CEA treatments combining dielectric oxide coatings and annealing that achieved record quantum-regime quality factors, alternative compounds will be investigated.
Ultimately, one treatment could enhance performance across quantum and accelerator operating regimes. Tunneling spectroscopy, X-ray spectroscopy and TOF-SIMS will correlate local superconducting properties with surface chemistry and dopant distributions. Two ALD systems will support development on coupons and transfer to resonators, building on demonstrated coating uniformity inside cavities. RF measurements will be conducted at CEA and partner laboratories, with Fermilab evaluating promising treatments at millikelvin temperatures. These investigations will clarify performance-limiting mechanisms and guide the rational design of optimized superconducting surfaces for both application domains.
Structure and reactions of light and medium mass nuclei with the Gamow Shell Model with Coupled Channels
This project investigates the structure and reactions of exotic atomic nuclei near the limits of stability, where nuclei may be weakly bound or unbound. Recent advances in radioactive-ion-beam technology enable experimental study of these rare systems, which often exhibit unusual features such as extended “halo” structures. Traditional nuclear models, effective for stable nuclei, fail to capture the complexities of these open quantum systems. To address this, the project develops and applies the Gamow Shell Model (GSM) and its extension to coupled-channel (GSM-CC) approaches, which treat nuclei as open systems able to exchange particles and energy with their environment. This enables a unified description of both nuclear structure and reactions, including the effects of the quantum continuum that become significant near the driplines. These models yield accurate predictions that align with experimental data, guide future research, and remain flexible enough to apply to areas such as hypernuclei, atomic anions, and engineered quantum systems. A key innovation is integrating structure and reaction theories, overcoming traditional divisions in the field. The frameworks are designed for computational efficiency and scalability, leveraging supercomputing resources and incorporating uncertainty quantification for reliable predictions. The project also fosters international collaborations with leading experimental facilities and contributes to the training of young scientists. By bridging gaps between theory and experiment, structure and reactions, and across quantum sciences, this research advances understanding of matter under extreme conditions, with broad implications for fundamental science and emerging technologies.
Investigating the Gluon Content of the Pion through Exclusive Vector Meson Production with CLAS12
The pion is a key particle for understanding Quantum Chromodynamics (QCD), the theory of the strong interaction. However, its gluon content, remains poorly understood. This thesis will explore a novel way to investigate the gluon content of the pion, combining vector meson production and the Sullivan process, which allows the pion to be studied in electron-proton collisions without requiring a dedicated pion target.
The student will analyse existing data from the CLAS12 experiment at Jefferson Lab to study exclusive phi-meson production near threshold on the pion. The goal is to establish the feasibility of isolating the Sullivan process and accessing the pion’s gluon gravitational form factors, which provide insight into its internal mass and pressure distributions. The analysis will involve particle identification, event reconstruction and background suppression and development of a dedicated event generator.
The project will also include feasibility studies for future measurements at the Electron-Ion Collider (EIC), where higher collision energies could open new opportunities to investigate the pion’s gluon structure, including through J/psi production.
Investigating alpha-clustering in neutron-rich Oxygen isotopes
This thesis aims to investigate the formation of alpha particles as a function of the neutron abundance of oxygen isotopes and their influence on nuclear structure. An experiment at GANIL will enable the measurement of nuclear reactions and the identification of the resulting products using the FAZIA and INDRA detectors. The results will contribute to a better understanding of the correlations between nucleons, the distribution of neutrons within the nucleus and the properties of nuclear matter.
Towards Intelligent VAMOS++: Physics-Informed AI for Adaptive and Reproducible Nuclear Instrumentation
Recent developments at GANIL have demonstrated the opportunities opened by AI methods for trajectory reconstruction, particle identification and self-supervised detector calibration in the VAMOS++ magnetic spectrometer and associated detectors for fission studies.
This PhD project will extend these advances into an integrated framework for intelligent instrumentation, combining calibration, reconstruction, and performance monitoring. The objective is to develop robust, scalable and transferable methods that adapt to changing experimental conditions while preserving the reliability of physical observables.
Hybrid (classical-quantum) algorithms for Green's functions in nuclear physics
High-performance computing is nowadays at the heart of research activities. In particular, fundamental research in theoretical nuclear physics relies on efficient numerical codes to carry out ab initio calculations of atomic nuclei. These calculations make it possible to establish systematic predictions, both for fundamental studies of nuclear systems and to support experimental activities. Considerable progress has been made in this field in recent years, leading the CEA team to perform, in 2026, the first ab initio calculation of the heavy nucleus uranium-238.
Despite these advances, ab initio calculations of nuclear structure and, in particular, calculations of reactions involving heavy nuclei remain extremely challenging. In this context, quantum computing is emerging as a new computational paradigm with the potential to transform current approaches and help overcome some of the existing computational bottlenecks. With this long-term perspective in mind, the CEA team launched a research activity several years ago dedicated to quantum algorithms applied to the nuclear many-body problem.
Building on this work, the aim of this PhD project is to develop quantum algorithms for calculating many-body Green’s functions. Green’s functions are a central quantity in many-body theory: they capture correlations between the constituents of a system and provide easy access to observable quantities. Furthermore, these functions are relevant not only to the study of the static properties of atomic nuclei, but are also naturally suited to describing dynamical processes, such as nuclear reactions. These latter properties make them particularly promising for quantum computing, where the dynamical evolution of a quantum system can be encoded naturally.
The project will focus on hybrid computational protocols that combine quantum and classical operations. The algorithms will be implemented and tested on quantum emulators and, where possible, on digital quantum computing platforms. The results of this project will help shape the computational strategies to be implemented in future ab initio nuclear theory research.
Mining LEP data for fragmentation: a TMD-oriented analysis of pi+pi- pairs in e+e- collisions
Fragmentation, the process by which quarks and gluons turn into hadrons, cannot be computed from first principles in QCD and has to be extracted from data. Transverse-momentum-dependent fragmentation functions (TMDFFs) are particularly poorly known. The cleanest observable for constraining them is the transverse-momentum distribution of nearly back-to-back pi+pi- pairs in e+e- annihilation, but no such measurement exists. The recent public release of the DELPHI data and software through CERN Open Data now makes it possible. The student will first set up access to the archived LEP data (DELPHI or ALEPH) and then extract pion-pair distributions. Pythia8 simulations will be used to identify the observable most sensitive to TMDFFs, and the measurement will come with a full uncertainty budget. The result will enter the MAP Collaboration’s global TMD fits as their first e+e- input, giving the most precise TMDFFs to date. These are key inputs for the EIC and FCC-ee physics programmes. The project is co-supervised by Valerio Bertone (theory) and Francesco Bossù (experiment) at CEA/Irfu. It involves David d’Enterria (CERN) and Ronan McNulty (UCD), with CERN and the MIT group that leads the current LEP data reanalysis directly involved. A highly motivated candidate, who did his M1 with the team and is now doing his M2 in Dublin, has already been identified.
PIPERADE: Offline commissioning and first online experiments at DESIR
PIPERADE (PIèges de PEnning pour les RAdionucléides à DESIR) is a double Penning trap being developed in Bordeaux, which will be installed at the DESIR facility at GANIL.
This instrument has two objectives: 1) perform precision mass measurements of exotic nuclei, providing valuable information on nuclear structure, the forces acting within the atomic nucleus (strong and weak interactions), and stellar nucleosynthesis, 2) perform the mass selection of nuclei of interest among a large number of contaminants, with very high resolving power (> 106) and a high capacity (> 104 ions per bunch). It will therefore provide to all other experimental setups in the DESIR hall an ultra-pure ion beam, greatly improving the precision of measurements of various observables, such as lifetimes, charge radii or electromagnetic moments.
The characterization of PIPERADE using a stable-ion beam has been ongoing at LP2iB since 2019 and will be completed in 2027, before the instrument is transferred to GANIL in summer 2027. This PhD at GANIL will therefore be perfectly synchronized with the installation of PIPERADE at DESIR. The student will actively contribute to the installation and offline commissioning of the instrument, i.e. investigating the systematic uncertainties on the masses as well as studying the different purification techniques. He/she will also participate to the first online experiment at DESIR with PIPERADE. Depending on the timeline of the project, he/she will also have the possibility to analyse online mass spectrometry data from other facilities, such as from IGISOL/JYFL.
Machine Learning-Based 3D Dose Reconstruction from Limited Scintillation Projections for Proton Therapy Dosimetry
Radiotherapy is an important modality in treatment cancer. In this domain, proton beams have ballistic superiority against photon beams. Among the existing delivery techniques, pencilbeam scanning is one of the most modern ones and presents many advantages in terms of flexibility and dose conformation. Nevertheless, the use of pencil beams (PBs) still requires well adapted dosimetry tools for small irradiation fields quality assurance. To answer this issue, an innovative dosimetry system has been developed. It is based on a scintillating block of 10 × 10 × 10 cm3, a mirror and an ultra-fast camera recording the scintillation from two orthogonal points of view to reconstruct 3-dimensional dose maps [1].
The recorded images require some extensive preprocessing to correct hot pixels caused by scattered radiation directly hitting the camera sensor, optical artifacts[2] and the scintillation quenching [3].
3D reconstruction usually relies on numerous projections, not achievable in dosimetry in radiotherapy and many reconstruction algorithms have been developed for imaging purpose.
A reconstruction method based on only two projections and relying on the radial symmetry of PBs has been developed and showed good performances. Nevertheless, some steps of the preprocessing and the reconstruction algorithm are computing time consuming and the radial symmetry limits its application to PBs.
Different experimental setups and machine Learning (ML) methods will be studied in this PhD thesis to address these limitations.
In a first step, two different setups (the previous setup and one including an additional scintillator sheet) generating different scintillation projections will be simulated to study different reconstruction strategies.
These digital setups will be fed to a treatment planning system (TPS) to generate 3D and 2D dose distributions for the two different setups. In parallel, LET and quenching correction maps will be computed to convert the dose distributions into scintillation distributions.
A projection matrix corresponding to the studied setups will allow to convert the 3D and 2D scintillation distributions into projection images and optical artifacts and noise (characterized in a preliminary M2 internship) will be added create a realistic images database.
The collection of 3D dose distribution and corresponding computed images will constitute the database used to study and evaluate different ML strategies for 3D dose reconstruction. The implemented reconstruction methods will be evaluated with experimental measurements performed at the Cyclhad proton therapy center, compared to conventional reconstruction methods and confronted to other beams such as not circular photon beams.