Particle Physics Division

The proposed PhD is aiming at probing new sources of CP violation in the Universe by studying the Higgs boson properties at the LHC, in particular through a modification of the coupling of the Higgs boson with the heaviest elementary particle: the top quark. The goal of this PhD is to develop a new data analysis strategy within the ATLAS collaboration probing the CP properties of the ttH coupling in pp collisions. The idea would be to design an analysis based exclusively on pure CP observables, meant to complement the existing model-dependent approach relying on machine learning methods that mix observables specific and non-specific to CP.

This new analysis should also have increased sensitivity to other rarer top quark associated Higgs production modes such as tHq.

New observables that will need to be developed will be made flexible enough to be applicable to a large panel of Higgs and top-quark decay channels. They will first be tested on the multilepton channel. A focus will be put on the reconstruction of the heavy particles (Higgs and top quark) in the final state, which is quite challenging in this channel.

The proposed PhD is aiming at measuring the CP properties of the Higgs-top coupling in the multilepton channel within the ATLAS experiment. Discovering new sources of CP violation is one of the most pressing questions in particle physics today. The Yukawa-like interactions in the Higgs boson sector could provide a particularly attractive way for additional sources of CP violation. The goal of PhD is to measure the ttH process in two innovative ways. First new reconstruction algorithms to identify the heavy particles in ttH events will be developed, which is made very challenging by the multilepton final state. Then the analysis will be designed exclusively based on pure CP observables. New observables based on ratios or angles will be explored, and measurements in regions of phase space where the separation between the various Higgs-top processes is less dependent on the CP properties of the Higgs-top-quark coupling will be studied. This new analysis should also have increased sensitivity to other rarer top quark associated Higgs production modes such as tHq.

Time domain and high-energy astrophysics is dealing with the most violent phenomena in the universe. Rapid exchange of information is crucial to detect these transient, i.e. short-lived, events with multiple observatories covering the full multi-wavelength domain and all cosmic messengers. Victim of its own success, the current way of manual reading, analyzing and classifying information shared by astrophysicists in Astronomers Telegrams or circulars within the Gamma-ray Coordinates Network is approaching saturation. One of the most promising novel approaches is to build on the recent progress in artificial intelligence and especially natural language processing (NLP) and feature extraction.

This thesis will bring together leading experts in two exiting domains: Artificial Intelligence and time domain, multi-messenger astrophysics. The project will be part of the UDOPIA program of the Paris-Saclay university and benefit from a rich ecosystem in both astrophysics and AI as well strong ties with leading industry partners.

Very recently a fundamentally new domain of astronomy and astrophysics has shown its first results: multi-messenger and real-time astrophysics. The simultaneous detection of various new astrophysical messengers (gravitational waves, high-energy gamma rays and high-energy neutrinos) and the exchange and combination of data from very different observatories allows to open new windows and provides unprecedented insights into the most violent phenomena ever observed.

New and significant conclusions can be obtained by combining these new messengers. Joint analyses of archival observations in different wavelengths have brought enormous insights in the past and, as this technique provides an assured and certain scientific return, it will also be used in the proposed thesis project. At the same time it has becomes clear that another important step does greatly enhance the sensitivity of multi-messenger searches: the need to gain full access to the wealth of information provided by analyzing and combining the data in real-time. The proposed thesis project will allow opening this new window to the high-energy universe: real-time multi-messenger astronomy at very high energies. The combination of the various particles and radiations in a truly multi-messenger online alert system will resolve several challenges faced in high-energy astrophysics and especially allow detecting and studying violent transient phenomena that are supposed to be at the origin of high-energy cosmic rays. The project will introduce the time domain to high-energy astrophysics and has the potential to cause a paradigm shift in how observations and data analyses are performed.

The core of the proposed project will be H.E.S.S., currently the world’s most sensitive gamma-ray instrument, and CTA, the next generation, global high-energy gamma-ray observatory. We’ll combine their data with events recorded by IceCube, the world’s largest neutrino telescope and the advanced Virgo and Ligo gravitational wave interferometers. The detection of a transient high-energy gamma-ray source in coincidence with gravitational waves or high-energy neutrinos will provide the long sought evidence for their common origin and may resolve the century old quest for the origin of high-energy cosmic rays.

We’ll also collaborate with the world’s most sensitive radio observatories (e.g. the SKA precursors MeerKAT and ASKAP) to search for counterparts to Fast Radio Bursts and in general study a large variety of messengers like Gamma-Ray Bursts or flares from active galactic nuclei. By scanning the data acquired with high-energy gamma-ray observatories in real-time, it will also possible to send alerts to the wider astronomical community to ensure simultaneous observations at other wavelengths.

In the last 30 years, the study of the Universe has seen the emergence of a standard model of the Universe based on general relativity. In this model, our Universe is made of ordinary matter, dark matter and a mysterious component called “dark energy”, responsible for the recent acceleration of the expansion of the universe. The upcoming large surveys, such as DESI in the US, will provide a map of the distribution of the galaxies 10 times more precise than the current state-of-the-art. The scientific community is gearing up to define new analysis techniques in order to extract the maximum of information from these surveys and hence enter the era of precision cosmology especially as far as growth of structure measurements are concerned. In this thesis, we propose to study a novel approach based on using the density at large scales to improve the precision on those measurements and to compare it with General Relativity predictions in order to search for possible deviations.

The thesis will take place at Irfu, the Institute for Research on the Fundamental laws of the Universe. The PhD student will join the cosmology group of Irfu/DPhP, composed of 10 physicists, 4 PhD students and 2 post-docs. Actively involved in the eBOSS and DESI experiments, the group also participates in Euclid and has in the past had a strong contribution in the SNLS, Planck and BOSS international collaborations. The future PhD student will be integrated into the DESI collaboration and will benefit from all the group’s expertise acquired on BOSS and eBOSS

Galaxy clusters, located at the node of the cosmic web, are the largest gravitationally bound structures in the Universe. Their abundance and spatial distribution are very sensitive to cosmological parameters. Galaxy clusters thus constitute a powerful cosmological probe. They have proven to be an efficient probe in the last years (Planck, South Pole Telescope, XXL, etc.) and they are expected to make great progress in the coming years (Euclid, Vera Rubin Observatory, CMB-S4, etc.).

Theoretical predictions of the cluster abundance and spatial distribution depend on cosmological parameters and cluster mass. To determine cosmological parameters from cluster surveys, one needs to be able to measure accurately cluster mass. The error on the mass estimation is currently the main systematic error for the determination of cosmological parameters with galaxy clusters. This is the reason why it is crucial to improve on the measurement of the cluster mass and to control associated errors.

The most direct method to measure cluster mass is based on gravitational lensing. It is now used routinely in optical surveys: a cluster induces distortions of the shapes of background galaxies. Using these distortions, it is possible to reconstruct cluster mass. It was shown recently that it is also possible to detect these distortions at millimetre wavelengths in the cosmic microwave background (CMB) instead of using background galaxies, and reconstruct the mass of galaxy clusters. The main advantage of using the cosmic microwave background is because it is located at very high distance allowing for mass measurement of distant clusters; it is not possible to do this measurement with background galaxies, which are too few for distant clusters.



Irfu/DPhP has developed the first tools to measure galaxy cluster masses using gravitational lensing of the cosmic microwave background for the Planck mission. The PhD thesis work will consist in taking hands on the tools and improve them to make them compatible with ground-based data. They will then be applied to public SPT-SZ (https://pole.uchicago.edu) data and SPT-SZ+Planck data jointly. The ACT (https://act.princeton.edu) data has also been made public recently and a joint analysis ACT+Planck will also be made.

In the second part of the thesis, the tools will be used to find observation strategies and compute integration times to measure cluster masses for high resolution ground based experiments such as NIKA2 (http://ipag.osug.fr/nika2/), alone and jointly with Planck.

The current methods are optimal for maps in total intensity and in the low signal-to-noise regime as shown in the Figure above. The future experiments will have lower noise levels and will be very sensitive to polarization. The third part of the thesis will be dedicated to development of new methods to extract the masses for the future low noise cosmic microwave background experiments such as CMB-S4 (https://cmb-s4.org), PICO (arXiv:1902.10541) or CMB Backlight (arXiv: 1909.01592).

Finally, we will study the precision on cosmological parameters that can be reached from galaxy cluster catalogues, given the precision on the mass expected from these future experiments.

DESCRIPTION

Positron emission tomography (PET) is a nuclear imaging technique widely used in oncology and neurobiological research. Decay of the radioactive tracer emits positrons, which annihilate in the nearby tissue. Two gamma quanta of 511 keV energy are produced by positron annihilation and allow one to reconstruct the annihilation vertex and distribution of the tracer activity in the body of the patient.

The precise determination of the position of the positron annihilation vertex is important for an accurate image reconstruction with a good contrast. In particular, it is useful for neuroimaging studies of brain and for pre-clinical studies with animal models (rodents) as well as for full body PET imaging. In this thesis we propose to contribute to an ambitious detector based on Cherenkov/Scintillating crystals. We have selected technologies that are particularly effective for PET imaging. The principles of the detector are patented. They should allow one to produce PET scanner with highly improved performances. The device uses advanced particles detector technologies: a dense scintillator crystal, micro-channel plate photomultipliers, gigahertz bandwidth amplifiers and fast data acquisition modules (WaveCatcher, SAMPIC). Data processing will involve Monté-Carlo simulations and data analysis based on GATE/Geant4 and Root C++ software libraries.



SUPERVISION

The successful candidate will work in the Department of Particle Physics of IRFU in close collaboration with the Department of Electronics Detectors and Computer Science for Physics. The CaLIPSO group includes two physicists and two students. Two Postdoc will join the project next spring. We collaborate closely with CNRS- IJCLabs on fast readout electronic, with CPPM of Marseille and CEA-SHFJ, and CEA-DES for simulations of medical imaging devices and image reconstruction algorithms, and with the University of Munster (Germany).



THE PROPOSED WORK

You will calibrate and optimize the detector prototypes and analyze the measured data. Your work will be focussing on detector time and spatial resolution optimization. This will involve many skills of an instrument scientist : fast photo-detection, fast electronics read-out (analog and digital) with picosecond precision, hardware and detector simulations with GEANT4 and GATE software.



REQUIREMENTS

Knowledge in physics, particle interactions with matter, radioactivity and particle detector principles, a vocation for instrumental (hardware) work, data analysis are mandatory. Being comfortable in programming, having a background in Gate/Geant4 simulation and C++ will be an asset.



ACQUIRED SKILLS

You will acquire skills in particle detector instrumentation, simulation of ionizing radiation detectors, photo-detection, implementation, operation of fast digitizing electronics, and data analysis.

The final goal of this PhD thesis is to obtain the very first 3D image of a nuclear reactor with a non-invasive method, namely the muography. This penetrating imaging technique has shown a rapid development during the last years, with major technological improvements including several innovative contributions from the CEA. Muon telescopes of unprecedented resolution thus unveiled the 2D structures of very large objects like Khufu’s Pyramid or a nuclear reactor. Recently, an algorithm was developed to combine these 2D images in a 3D tomography, despite the small number of available projections and the huge size of the corresponding matrix system. This PhD will then be dedicated to the use of this algorithm to ongoing muography measurements on a nuclear reactor in decommissioning phase. The PhD student will actively participate to the data taking, data analysis and to the corresponding simulations. He/She will first apply the algorithm to smaller objects, in particular nuclear waste containers in various environments, in order to understand and optimize its performance. These intermediate steps, beyond their own interest, will help to better tune the algorithm parameters but also to determine the future measurements (positions, orientations, acquisition times, etc.). The overall goal of this work is thus the development of a 3D, generic, innovative imaging tool in the field of decontamination & decommissioning, with certainly many more applications in the societal and academic domains.

Since the discovery of the Higgs boson efforts are focused on the search for new phenomena, beyond the Standard Model.

One of the important aspects of the comparison between experimental measurements and theory is the need to normalize as

precisely as possible experimental results to theory. This means in practice being able to measure as precisely as

possible the luminosity of the LHC. The goal is to reach a precision better than 1% within the next few years,

a factor two or three better than the precision that has been reached up to now.



After the LHC restart, foreseen in 2022, it is planned to increase the luminosity by a factor of about two. To make

the best out of this luminosity increase, the calorimeter trigger system is being significantly modified and upgraded.

The upgraded trigger system is based on real time analysis by FPGAs of the digitized detector signals.



An essential feature of the upgraded trigger system is its ability to measure the energy deposited in the calorimeter

bunch crossing by bunch crossing. Combined with the stability, excellent linearity and response uniformity of the ATLAS

Liquid Argon calorimeter, the upgraded trigger system offers the potential to measure the luminosity with excellent

linearity and stability performances. A very promising analysis technique would be to use a neural net, that

could be implemented in the core of the FPGA that processes the data.



An other feature of the upgraded trigger system is its ability to keep track of all the interactions taken place in the

detector over a much longer period of time than the main readout. The main readout system is able to keep in memory only

up to four or five consecutive interactions. The trigger system has the capability to keep track of each individual bunch

crossing over a period of time corresponding to several tens of consecutive bunch crossings.

This long term memory feature gives the possibility to compensate real time the effect of charge space accumulation,

which will be crucial for data taken after 2025, at very high luminosity. More importantly, this also opens up the

possibility to detect particles reaching the detector long (several tens or even hundreds of ns, to be compared to the

25 ns between two consecutive bunch crossings) after their production. Such particles are slow and very heavy, and can be

detected almost up to the kinematic limit of 7 TeV. This is significantly higher than the limits reachable by more

classic techniques. Such particles typically appear in many classes of supersymmetric models.

The thesis will start in Autumn 2021. ATLAS, one of the major experiment at the LHC, is preparing for the expected increase of luminosity for Run3 and HL-LHC. The first part of the thesis is dedicated to a qualification task that could either consist in participating to the commissioning of the new muon detectors which are integrating the experiment, or taking part in the muon momentum calibration effort in view of Run3, that will start in 2022. Both options are closely related to the main thesis subject. The thesis will be followed by a measurement of precision physics in the field of the Z boson with ATLAS data.

The subject is focused on electroweak precision physics in ATLAS. The aim is to measure with the best possible precision the electroweak mixing angle, as well as the mass of the Z boson, using Run2 and Run3 data. The explored channel is that of the Z boson decaying into a muon-antimuon lepton pair. The student will work on muon momentum calibration using the J/Psi resonance as a standard candle, and will also reduce, through advanced fitting methods, the uncertainties related to the parton distributions functions (PDFs). These measurements should lead to a high improvement in the electroweak fit and thus significantly constrain the Standard Model, as well as Beyond Standard Model physics.



The CEA ATLAS group is part of the Department of Particle Physics (DPhP) of the Institute of Research into the Fundamental Laws of the Universe (IRFU) at CEA Paris-Saclay.

DPhP comprises about 110 physicists. DPhP scientific themes include elementary components of matter at the highest energies at the CERN LHC collider, R&D for future accelerators, study of antimatter, neutrino physics, gamma ray astronomy, study of gravitational waves, observational cosmology and instrumentation for medical applications. The group has a world-leading expertise in electroweak physics, namely with measurements of Z, W and Higgs boson cross sections and measurement of the W boson mass, achieved for the first time at the LHC. It builds on competences in muon reconstruction and muon spectrometer alignment and electron/photon identification.

This PhD proposes to apply artificial intelligence algorithms to big data in two innovative ways by exploiting the large proton-proton collision dataset collected by the ATLAS experiment at the Large Hadron Collider (LHC). The challenge is to extract processes that are both rare and complex from the huge amount of LHC data. Cutting edge deep learning techniques will be explored first to reconstruct complex final states with underconstrained kinematics. This would allow reconstructing final state particle energies and momenta knowing some conservation laws. Second deep learning will be implemented to extract rare signals. These new developments will be applied to two very rare and complex processes (ttH and 4-top). These two processes combined will allow testing the true nature of the coupling between the cornerstone of the Standard Model, the Higgs boson, and the heaviest elementary particle, the top quark, and could reveal new sources of asymmetry between matter and anti-matter. First unsupervised training will be tested for the first time for final state reconstruction. Observables based on the fully or partially reconstructed final state particles should then improve the ability to extract rare signals using for instance Graph Neural Network classifiers newly in high energy physics.

Exploring these new strategies for event reconstruction and classification will pave the way to understanding how the increased amount of data expected in the next phase of the LHC can be exploited in an optimal way.

The goal of the thesis is a measurement of a fundamental parameter in the Standard Model of particle physics, the W boson mass, with the ATLAS detector at the LHC, using leptonic W boson decays. The analysis will be based on dedicated low-pile-up data samples, which have limited integrated luminosity but optimal experimental resolution in the reconstruction of missing transverse energy, which is a requirement in the analysis of final states with neutrinos.



The candidate will participate in the installation and the commissioning of the New Small Wheel, an upgraded muon detector for the ATLAS endcaps. IRFU has played a leading role in its construction and will get strongly involved in its scientific exploitation. In addition, the candidate will calibrate the muon momentum with sufficient precision for the measurement. The second phase of the project consists in improving the QCD aspects of the modelling of W-boson production and decay, and optimizing the analysis to minimize the resulting measurement uncertainty. After completion, the measurement will be interpreted in terms of compatibility with the Standard Model or as a hint of New Physics.

Description and context :



The discovery of neutrino oscillations proved that neutrinos are not massless. This cannot be explained in the framework of the Standard Model of Particle Physics.



Several experiments are currently investigating neutrino oscillations properties such as T2K or NOvA. These experiments study a beam of muon neutrinos produced at particle accelerators.

DUNE is one of the main projects under development in this field. The target mass of DUNE of 4$\times$17 kt and a larger beam power and longer baseline will allow a very precise measurement of the CP violating phase (deltaCP) and of the neutrino mass hierarchy. The measurement of CP violation in the leptonic sector would be a major discovery. It is connected, in several models for new physics, to the matter-antimatter asymmetry observed in the Universe. T2K has already set the first constraint on the value of (deltaCP), excluding at 90\% the conservation of CP in the leptonic sector (deltaCP=0).



The concept of DUNE is very similar to existing experiments. Neutrinos are produced in the interaction of a beam with a target and detected at two locations: the near detector site at ~210 m and the far detector site at ~1300 km. The comparison between the spectra observed at both sites allows the parameters governing neutrino oscillations to be measured. The synergy between the detectors on both sites enables the reduction of the uncertainties coming from the flux or the neutrino interaction cross-sections.



The construction of the near detector of DUNE should start at the end of 2026 or at the beginning of 2027. This makes the future years crucial for the development of the technologies to be deployed in the detectors and of the data analysis techniques necessary to their scientific exploitation.



The beam monitoring for DUNE will be performed by the SAND detector, largely inspired from the one of T2K. The student will take part in the development and the optimisation of the resistive Micromegas detectors for the charge readout of the time projection chambers performing track measurements in SAND. The work will take place inside the DUNE international collaboration with a possible contribution to the tests and commissioning of the detectors for the near detector of T2K: ND280-Upgrade.



Description of the group, institute and supervision :



The DUNE collaboration is an international collaboration constituted of more than a thousand members with a strong contribution in Europe. The accelerator neutrino group at IRFU/DPhP has been actively involved in the collaboration for several years, namely thanks to the developments for the charge readout of the dual-phase technology for the far detector.



The activities on the near detector SAND directly benefit from the expertise acquired from the design of the resistive Micromegas detectors for the near detector of T2K. In this context, the accelerator neutrino group is constituted of 6 permanent scientists and 2 students involved in the resistive Micromegas activities.



The student will also benefit from the collaboration with the detector and technical services at IRFU/DEDIP, one of the international leaders in the development of micro-pattern gaseous detectors. Very advanced tools on detectors, DAQ, slow-control and electronics will therefore be available.



Proposed activities :

The student will take part in the optimisation on test bench data and with simulations of the design of the resistive Micromegas detectors for the TPCs of SAND. Cosmic rays tests and beam tests will be necessary to characterise the detectors in real conditions. These tests will be performed at the beam lines of CERN and/or DESY and on the equipment at IRFU. A large contribution to the detectors installation on site and to their operation is foreseen. The analysis of the data acquired during these tests, which will be done in collaboration with a post-doctoral researcher, will be one of the main activities and could result in a publication. The construction of the near detector of T2K is also supposed to be completed by Summer 2022. This will give the opportunity to the student, already during the internship, to take part in the testing of the resistive detectors in the context of the production of a final detector.



The other main activity proposed will be the production and analysis of simulation data of the entire SAND detector. This work will allow the theoretical systematic uncertainties to be reevaluated using new neutrino-nucleon interaction models, an important contribution to the sensitivity of SAND for the oscillation measurements with DUNE. This analysis on the nuclear models, currently being developed in the group for T2K, will need to be adapted to DUNE while incorporating the specific instrumental developments based on the various prototypes available. The development of the track reconstruction algorithms for the TPCs of SAND will also be carried out in tight collaboration with the members of the group currently involved in this activity for the near detector of T2K.



Education and skills required :



A Master's degree in particle physics with knowledge of the Standard Model is a requirement for this thesis. A strong interest for instrumental activities is expected and a motivation for neutrino physics is a very positive point. Knowledge of C++ and ROOT will also be very useful.



Acquired skills :



The student will have by the end of his/her PhD a good understanding of detectors and computing tools used in a particle physics collaboration thanks to his/her involvement in their developments. He/she will be able to promote his/her technical skills on detectors and on data analysis methods in other contexts.



Collaboration/Partenariats :

The student will work within a large international collaboration. This will provide a very good experience in particle physics and an important visibility through the participation to physics schools, workshops and conferences where he/she will present his/her results.

The proposed PhD is aiming at observing for the first time the production of four top quark (tttt) at the LHC using the multilepton signature in data collected with the ATLAS experiment. The production of four top quarks is one of the most spectacular final states which became accessible at the LHC. While expected to be small in the Standard Model, the tttt cross section is predicted to be strongly enhanced in many new physics scenarios. The tttt process is also sensitive to the detailed properties of the Yukawa-like interaction between the top quark and the Higgs boson. Probing further this class of events over the next years will therefore be crucial to understand better the true nature of the Higgs-top-quark interaction and potentially to pinpoint subtle deviations from the Standard Model.

Several innovative ways will be pursued to reach the observation of this tttt process. First a better separation between the signal and the different background processes will be studied by designing various multivariate discriminants and by a better understanding of the modeling of the ttW process. Another path towards observation will be to study the possibility to reconstruct the different top quarks in the final state, which is particularly challenging in this channel.

Achieving a measurement of the tttt production will allow to probe a key property of the top-Higgs interaction, i.e. the CP nature of the top-Higgs coupling. Ultimately, the tttt process should also be measured separately from the production of three top quarks, which is currently totally unexplored experimentally.

This PhD topic is about the NUCLEUS experiment, which aims at precisely measuring coherent elastic neutrino-nucleus scattering (CEvNS) at the Chooz nuclear power plant (France). Although at ~MeV energies, CEvNS is the predominant interaction process of neutrinos with matter, it has remained unobserved for a very long time because of the difficulty to measure the very low energy nuclear recoils it induces. It was only 40 years after its first prediction that this process was observed in 2017 with neutrinos of a few tens of MeV at the Oak Ridge laboratory (Tennessee). The first detection of CEvNS at a nuclear reactor remains to be achieved, especially because the corresponding nuclear recoils lie in an energy regime (~100 eV) which is difficult to measure with conventional detection technologies, and also because of the unfavorable background conditions nuclear power plant environments generally offer. The NUCLEUS collaboration is therefore working on the design of an innovative detection system using two cryogenic calorimeter arrays capable of reaching ~10 eV energy thresholds, and surrounded by a twofold system of instrumented cryogenic vetoes. This set of cryogenic detectors will be protected by an external passive shielding and by a muon veto to improve the identification and discrimination of backgrounds. With this system, NUCLEUS aims at a precise measurement of CEvNS in order to push the study of the fundamental properties of the neutrino as well as the search for beyond standard model physics towards the low energy frontier. Interestingly, CEvNS also exhibits a cross-section 10 to 1000 times larger than the usual ~MeV neutrino detection channels (inverse beta decay reaction, neutrino-electron scattering process), making it possible to miniaturize future long-range neutrino detection setups. The first phase of the NUCLEUS experiment will for instance deploy an array of cryogenic calorimeters made of sapphire (Al2O3) and calcium tungstate (CaWO4) crystals, totaling 10 g of detector.



In addition to the characterization and preparation of the experimental site at Chooz, our team at Irfu is taking a leading role in the project through several hardware and software developments. In particular, the DPhP is strongly involved in the realization of one of the instrumented cryogenic shielding of the experiment, here called the cryogenic outer veto. This detector consists of an arrangement of high-purity Germanium crystals, erected around the two cryogenic calorimeter arrays, and operated in ionization mode. This detection system will play a central role in the identification and discrimination of external backgrounds, such as ambient gamma radioactivity or atmospheric muons resulting from the interaction of primary cosmic rays in the atmosphere. The exploitation of the data delivered by this detector is then a natural entry in the global analysis effort to extract a first CEvNS signal at a reactor, with first background data collected in 2021/2022 during the blank assembly phase at the Technical University of Munich, and with data collected during the first physics run planned in 2023 at Chooz.



The work proposed in this PhD thesis is focused on the external cryogenic veto of the experiment, with the ultimate goal of achieving a comprehensive understanding of the backgrounds in the CEvNS region of interest, between 0.01 and 1 keV. The priority at the beginning will be given to the realization and commissioning of the external cryogenic veto system during the blank assembly phase in Munich. This work includes the assembly of the different detector elements (crystals, support mechanics, readout electronics, etc.) in the cryostat of the experiment, and includes all the necessary tests to validate and quantify the performances of this detector. In a second step, the student will ramp up in the collaboration analysis effort by contributing to the development of analysis and simulation tools. These tools will be used to interpret the background and detector calibration data acquired during the blank assembly phase and during the first physics run. He (she) will focus on the study of a specific source of external background, and quantify its impact on the physics potential of the experiment. This work will require a good understanding of the processes governing radiation interactions in matter and of the solid-state physics driving the behavior of cryogenic detectors (e.g. phonon propagation). Finally, the student will use the first data from the physics run at Chooz to conduct a search for new physics beyond the standard model (measurement of the weak mixing angle at low energies, search for new neutrino couplings, constraints on the electromagnetic properties of the neutrino, etc.). This work will require the implementation of advanced statistical methods for interpreting the data, in order on the one hand to understand the impact of the various sources of uncertainty on the constraints obtained, and on the other hand to guarantee the reliability of the results.

New artificial intelligence techniques are attracting growing interest in handling the massive volume of data collected by particle physics experiments, particularly at the LHC collider. This thesis proposes to study these new techniques for the simulation of the background noise of rare events originating from the decay into two muons of the Higgs boson as well as to set up a new artificial intelligence method to extract these rare events from the gigantic dimuon background noise.

In 2012, the Higgs boson, a fundamental part of the Standard Model of particle physics, was discovered at the LHC. The demonstration of its decay in dimuon is now at the heart of the LHC program to measure the coupling of the Higgs boson to 2nd generation particles.

Simulating the dimuon background with sufficient statistics is the first challenge of this analysis. The thesis proposes to test, for the first time, the use of very promising artificial intelligence models as a simulation method using "Generative Adversarial Networks (GANs)" with an architecture of two competing networks. In addition, the thesis also foresees a complete redesign of the analysis in order to implement new data processing methods (Deep Neural Networks) to optimize the extraction of the weak signal.

In July 2012, the ATLAS and CMS collaborations announced the discovery of a new particle with a mass of about 125 GeV, at the Large Hadron Collider (LHC) at CERN. Since this discovery, the two collaborations have been actively studying the properties of this new particle, so far consistent with those of the Standard Model Higgs boson.

In the Standard Model, the Brout-Englert-Higgs mechanism predicts that the Higgs boson will interact with particles of matter (quarks and leptons, called fermions) with a force proportional to the mass of the particle. It also predicts the Higgs boson will interact with the force carrier particles (W and Z bosons) with a strength proportional to the square of the particle's mass. Therefore, by measuring the Higgs boson decay and production rates, which depend on the interaction strength to these other particles, one can perform a fundamental test of the Standard Model.

The ATLAS and CMS collaborations have already observed the decay of the Higgs boson into tau lepton, belonging to the third “generation” of fermions. Since muons are much lighter than tau leptons, the decay of the Higgs boson into a muon pair is expected to occur about 300 times less often than that of a lepton-tau pair. Despite this scarceness, the H ? µµ decay provides the best opportunity to measure the interaction of the Higgs boson with second generation fermions at the LHC, providing new information on the origin of mass for different generations of fermions. The ATLAS and CMS collaborations recently presented results on this decay using the dataset during the 2nd phase of the LHC (Run-2 from 2015 to 2018). Figure 1 shows the mass distribution of muon pairs while Figure 2 shows an example of a candidate event to be a Higgs boson decaying into two muons as recorded by the ATLAS detector. The study of this process is one of the main objectives of the third phase of the LHC (Run-3).

The aim of this thesis is the search for a Higgs boson decaying into two muons by analyzing the whole dataset from Run-3 and combining them with the previous data from the second phase (Run-2) in order to establish the discovery of the decay of the Higgs boson into two muons and constrain possible theories of physics beyond the Standard Model that would affect this decay mode of the Higgs boson. The thesis will also include work on the performance evaluation of the ATLAS muon spectrometer. Particular interest will be paid to the understanding, analysis and operation of MicroMegas type gas detectors. The purpose of phase-I of the ATLAS detector upgrade is to prepare for the high luminosities that will supply the LHC. In this context, the 2 large detection planes called NSW (New Small Wheel) will be equipped with new MicroMegas type detectors and will replace part of the ATLAS muon spectrometer and be operational for the restart of the LHC in 2022.

The vector-boson scattering process, pp?VV+jj+X, where V=W,Z, is characterized by the presence of

leptons from the W or Z decay, and high-energy forward jets in the final state. WW, WZ and ZZ scattering all have

been observed during the LHC Run2, with partial

datasets. The next step for the community in this field, using improved event selections and event classification, and larger data samples, is to enhance the visibility of the signal and confront the SM predictions with more precise analysis results.

This measurement constitutes a fundamental test of the coupling between the Higgs and the vector bosons, and of the Standard Model as whole.

The Standard Model (SM) of particle physics has been intensively tested experimentally over the past few decades and accurately fits the measurements. The latest major success of the SM is the discovery of the Higgs boson at the LHC in 2012. However, in the SM all is not clear. Many of the properties of the interactions and particles cannot be explained by the SM, such as the number of particle families or the differences in scale of their masses. Moreover, the SM has many free parameters to be determined experimentally, indicating that it is a theory that is only effective at low energy. More importantly, the SM can explain neither dark matter nor dark energy. While theorists are building new models to fill in these gaps, the experimentalists are trying to highlight physics beyond the SM, or ”new physics”, on the one hand by carrying out high-precision measurements to detect inconsistencies in the SM, and on the other by directly looking for new particles.



In 2012, both the ATLAS and CMS collaborations observed a new boson with a mass of approximately 125 GeV whose properties are at present compatible with those of the SM Higgs boson. The analyses of data in the diphoton final state (one of the two most sensitives modes) leading to this discovery probed an invariant mass range extending from 110 to 150 GeV. However, physics beyond the SM (BSM) can also provide a Higgs boson that is compatible with the observed 125 GeV boson. The extended parameter space of several BSM models, for example generalized models containing two Higgs doublets (2HDM) and the next-to-minimal supersymmetric model (NMSSM) gives rise to a rich and interesting phenomenology, including the presence of additional Higgs bosons, some of which could have masses below 125 GeV. Such models provide good motivation for extending searches for Higgs bosons to masses as far below mH = 110 GeV as possible.



Another mystery in the SM is the strong CP problem: why does the quantum chromodynamic (QCD) Lagrangian conserve CP symmetry, to within extraordinarily strict experimental limits, in absence of a correspondent fundamental symmetry? A possible way to make it natural, as introduced by Peccei and Quinn in 1977, is to add an extra global symmetry into the theory, the U(1) symmetry, that is spontaneously broken at some high energy fa. Such a symmetry leads in turn to the prediction of a new light pseudo-scalar particle: the axion, coupling to photons and gluons. A light axion (with a mass ma below the eV) could solve the strong CP problem. More generally, ”axion-like particles” (ALP) appear in any theory with a spontaneously broken global symmetry and can be searched for at particle colliders.



It is thus of primary importance to look for light resonances at the LHC and it is encouraged by many theorists. The X -> gamma gamma decay channel provides a clean final-state topology that allows the mass of a potential new resonance to be reconstructed with high precision. Present published direct searches at LHC in the diphoton decay channel cover a mass range down to about 65 to 70 GeV. Interesting limits could also be achieved in the lower mass range down to about 10 to 20 GeV. For example in ref [1], a conservative re-interpretation of inclusive diphoton cross section at LHC allows to put limits in this still unexplored region. Extending the mass range in the diphoton decay channel down to as lower masses as possible is what proposed in this thesis. Such a search is specially ambitious because of the difficulty to trigger on the signal. There is already a thesis about this on ATLAS, although the result is still not yet public.



This is the right moment to perform this search at LHC. Run 3 should start in the first half of 2022 and there should be about 133 fb-1 of new data available for the thesis. This very low mass analysis (below 70 GeV) has never been pursued in CMS. The triggers in run 2 were not optimised for this search, their pT thresholds were too high to have a good efficiency. In run 3, there is a possibility to adapt the trigger to gain in efficiency. Also, the run 2 data could still be used to set limits using the available boosted events. Since no limit at all is available at the moment in this mass region, even a mild one would be of interest. The thesis will be divided into 4 parts:

- During the 6 first months and before the start of run 3, the student will work on optimising a trigger for run 3 to gain in efficiency for this very low mass search. This means lowering the photons pT thresholds without taking too much bandwidth.

- Then, during the following year, the student will carry out and optimise the analysis of run 2 data.

- During the next 6 months, once the data in run 3 are recorded with a better suited trigger, the analysis of run 3 data will be carried out as well end eventually combined with the results from run 2.

- Finally, the last six months will be used to write the thesis.



The IRFU CMS group has a great expertise in photon energy measurement, as it has been involved in the ECAL construction and design and has a leading role in its calibration. There were recently several important responsibilities in the group relating to ECAL calibration (ECAL detector performance group conveners, electron/photon physics object group convener, ...). The IRFU CMS group did play an important role in H -> gamma gamma nalysis at 13 TeV. Two members of the saclay group were conveners of the H -> gamma gamma group during run 2. The group is also in close contact with the main authors of the current low mass analysis, based in Lyon (IP2I, IN2P3). The student will greatly benefit from group’s knowledge to lead these studies.