Postdoctoral Researcher in Heavy-Ion Physics (M/F)

Organisation/Company CNRS Department Laboratoire de physique subatomique et de cosmologie Research Field Physics Researcher Profile First Stage Researcher (R1) Application Deadline 12 Nov 2026 - 23:59 (UTC) Country France Type of Contract Temporary Job Status Full-time Hours Per Week 35 Offer Starting Date 1 Mar 2027 Is the job funded through the EU Research Framework Programme? Not funded by a EU programme Is the Job related to staff position within a Research Infrastructure? No

Offer Description

The ANR BEAUT project, which is led by LPSC in cooperation with IPHC in Strasbourg and IP2I in Lyon, deals with one of the most basic open questions in quantum chromodynamics (QCD), namely what the transport properties of deconfined matter are. Its objective is to determine, for the first time, from the Run 3 Pb–Pb collisions recorded by the ALICE experiment at the LHC (√(s_NN ) = 5.36 TeV), the spatial diffusion coefficient of heavy quarks, D_s (T), in the quark-gluon plasma (QGP). The beauty quark is particularly useful as a probe in this case since its relaxation time is longer than the lifetime of the fireball and therefore the final momentum distribution of the beauty quark keeps a direct reflection of D_s (T), without being contaminated by partial thermalisation or hadronic rescattering effects.
The successful candidate will work on the « beauty jets » aspect of the project, since this is the responsibility of LPSC. As a member of the LPSC ALICE group and under the scientific guidance of the project coordinator, they will have the following duties: (i) develop a machine-learning-based algorithm for identifying jets that arise from beauty quarks (b-tagging), with it being optimised for the high-multiplicity conditions found in Pb–Pb collisions; (ii) measure the nuclear modification factor R_AA and the azimuthal anisotropy v_2 of b-jets, along with their substructure, in order to determine whether the dead-cone effect persists in the dense medium; and (iii) help in the combined Bayesian extraction of D_s (T) from the complete set of observables available to the consortium (exclusive B hadrons, b-jets, and non-prompt J/ψ at forward rapidity).
The contract, which lasts for 36 months, is divided into three consecutive phases: the first phase involves the development and validation of the tagger; the second phase includes the measurement of the b-jet R_AA, v_2 and substructure; the third phase consists of taking part in the D_s (T) extraction, and towards the end of the contract there is preparation for the validation of the analysis tools for Run 4 using the ITS3 detector

Activities
Main activities
Design, train and validate a b-jet tagger based on graph neural networks (GNN) and recurrent neural networks (RNN), following the ParticleNet/DeepJet paradigm, simultaneously combining all available track-level information (impact-parameter significance, secondary-vertex position and mass, displaced-vertex multiplicity, momentum, particle identification), using PYTHIA8 + GEANT4 simulations with full ITS2 digitisation;
Validate the tagger on pp data using beauty-enriched and beauty-depleted samples, then adapt it to Pb–Pb collisions through domain-adaptation techniques; quantify the biases introduced by the use of simulation and assess robustness, notably through a data-driven cross-validation based on exclusive B-hadron reconstruction (in collaboration with IPHC);
Deliver and document an open b-tagging software package (public GitHub repository, Zenodo DOI) compatible with the ALICE O2 framework;
Measure the nuclear modification factor R_AA of b-jets as a function of transverse momentum and centrality in Run 3 Pb–Pb collisions, then their azimuthal anisotropy v_2 and substructure, notably by implementing the semi-inclusive hadron+jet method developed at LPSC, adapted to triggering on a displaced-vertex particle;
Evaluate all systematic uncertainties and their correlations for propagation into the Bayesian inference;
Contribute to the combined Bayesian extraction of D_s (T) (public framework based on the PyMultiNest MCMC sampler) and to the comparison with transport models (TAMU, Duke, LBT), in interaction with the LPSC theory group;
Prepare and publish the relevant papers under the ALICE Collaboration name (covering b-jet R_AA; v_2 and the dead-cone effect; the combined D_s (T) paper) and show the results at collaboration meetings and at international conferences.

Secondary activities
Take part in the monthly consortium coordination meetings, in the annual workshops bringing together experimentalists and theorists, and in bilateral exchanges with the project's PhD students at IPHC and IP2I;
Contribute to adapting the tagger to the ITS3 detector geometry and to preparing Run 4 analyses;
Take part in ALICE Collaboration service tasks and, where appropriate, in the occasional supervision of interns;
Take part in the project's science outreach activities (such as public events and science-communication content).

The Laboratoire de Physique Subatomique et de Cosmologie de Grenoble (LPSC) ( http://lpsc.in2p3.fr ) is a joint research unit that brings together CNRS-IN2P3, Université Grenoble Alpes (UGA), and Grenoble INP. It usually has about 230 people.
Since 2005, the ALICE group of the LPSC has been a member of the ALICE Collaboration and is responsible for the first ALICE measurements of beauty jets; it also has charge over the development of the O2 analysis software and the research and development of the readout electronics for the ITS3 detector. The institute also hosts a theory group (specialising in NLO pQCD predictions) which acts as the interface between experimental observables and transport models.
The BEAUT project is involving three French laboratories which are part of the ALICE Collaboration (LPSC, IPHC, IP2I) and will also hire two PhD students in Strasbourg and Lyon in addition to the current position. It makes use of monthly coordination meetings, cross-site visits, a common code repository, and annual workshops that bring together the consortium and international theorists. The successful candidate will be given the status of member of the ALICE Collaboration (with the associated rights to access data and the computing grid and the right to author publications); he or she will be provided with a dedicated GPU workstation, have access to the resources of CC–IN2P3, and be given a travel budget for attending international conferences and for visits between partners

Technical knowledge and know-how
PhD in experimental particle physics, nuclear physics or heavy-ion physics (or a related discipline); demonstrated experience in collider data analysis;
Strong command of Python and C++, ROOT, and distributed computing environments (grid, computing centre);
Command of machine learning applied to high-energy physics (deep neural networks, preferably GNN/RNN) and of at least one standard framework (PyTorch, TensorFlow), including GPU training;
Knowledge of jet and heavy-quark physics, b-tagging, and the phenomenology of ultra-relativistic heavy-ion collisions;
While it would be helpful to have some knowledge of Bayesian statistical inference (such as that involving MCMC methods), it would be an advantage to have experience obtained either within the ALICE Collaboration or with the O2 framework.

Interpersonal skills
Autonomy, scientific rigour and the ability to carry a research project through from start to finish, evidenced by at least one publication or collaboration note in the field;
Aptitude for collaborative work within a multi-site consortium and a large international collaboration;
Has a perfect command of scientific English, both in writing and in speaking

#J-18808-Ljbffr