Magneto-ionic gating of magnetic tunnel junctions for neuromorphic applications

Magneto-ionics is an emerging field that offers great potential for reducing power consumption in spintronics memory applications through non-volatile control of magnetic properties through gating. By combining the concept of voltage-controlled ionic motion from memristor technologies, typically used in neuromorphic applications, with spintronics, this field also provides a unique opportunity to create a new generation of neuromorphic functionalities based on spintronics devices.

The PhD will be an experimental research project focused on the implementation of magneto-ionic gating schemes in magnetic tunnel junction’s spintronics devices. The ultimate goal of the project is to obtain reliable and non-volatile gate-control over magnetisation switching in three-terminal magnetic tunnel junctions.
One major challenge remains ahead for the use of magneto-ionics in practical applications, its integration into magnetic tunnel junctions (MTJ), the building blocks of magnetic memory architectures. This will not only unlock the dynamic control of switching fields/currents in magnetic tunnel junctions to reduce power consumption, but also allow for the control of stochasticity, which has important implications in probabilistic computing.

Study of electronic processes in nitride LEDs by electro-emission microscopy

Nitride LEDs are universally used for energy-efficient lighting. They are extremely efficient at low indium content and low current density, allowing to produce commercial white LEDs from a blue LED and a phosphor that absorbs blue and re-emits a broad spectrum in the visible range. However, nitride LEDs suffer from a drastic drop in efficiency at higher current densities and higher indium concentrations, for emission in the green or red. This is an obstacle to extending their use, in order to obtain higher efficiencies with less material, as well as better color rendering. These efficiency drops are partly due to an increase in three-particle Auger-Meitner processes, which are strongly impacted by local device heterogeneities, and can be reduced by specific engineering of structural defects in nitride materials. This thesis proposes to study the electronic processes in nitride LEDs in operando, using electro-emission microscopy. In particular, charge injection mechanisms in the active part of the LEDs and Auger-Meitner processes will be investigated and quantified. The spatial resolution of the technique will allow to characterize the role of heterogeneities (defects or alloy disorder) in the loss processes.

Topological superconductivity and Fermi surface in spin-triplet superconductors

Topological superconductivity has become a subject of intense research due to its potential for breakthrough in the field of quantum information. Bulk systems are a promising possibility, with candidates found mainly among unconventional superconductors, which are also strongly correlated electron systems. Today, only a few candidate compounds for topological bulk superconductivity exists, and they are mostly uranium-based heavy fermion superconductors. UTe2 is one of the most prominent candidates. The topological properties of the superconductors depends crucially on the topology of the Fermi surface.
In this project we want to set up a novel technique (for our team) relying on a tunnel diode oscillator circuit. This techniques is very sensitive to quantum oscillations, and to be well both to high magnetic fields and to high-pressure studies. First experiments concentrate on the novel superconductor UTe2, where the Fermi surface is only partly known. In further studies the topological properties of the ferromagnetic superconductors UCoGe and URhGe will be revised.

Sub-critical crack growth in oxide glasses

Material failure is a concern for scientists and engineers worldwide. This includes oxide glasses, which are integral parts of building, electronics, satellites due to multiple advantageous features, including optical transparency, elevated mechanical and thermal properties, chemical durability, biocompatibility and bioactivity, etc. Despite this, oxide glasses have a significate drawback: they are inherently brittle. Oxide glasses are well known to undergo dynamic fracture (crack propagation velocity of ~km/s – as in the case of a glass crashing to the floor and shattering); yet, there is another fracture mode less noticeable that will be studied during this thesis, where crack fronts grow sub-critically. The growth of these crack fronts is aided by environmental parameters including atmospheric humidity and temperature, and the crack front velocity depends on the local stress felt by a crack tip, coined the stress intensity factor.

Currently, our experimental setup tracks the crack front position in time via a tubular microscope equipped with a camera. Post-analysis of images provides the crack front velocity and reveals the environmental limit K_e and region I. However, the current experimental setup cannot capture regions II and III. Several factors play into this limitation: elevated crack front velocity (10e-4 to 1500 m/s), sample size (5×5×25 mm^3), camera acquisition rates, etc.

In recent years, our team has used the potential drop technique to track the crack front velocity when v > 10e-4 m/s in PMMA. This technique involves the deposition of conductive strips on the sample surface. Subsequently, these lines are attached to a high frequency oscilloscope. As the crack front propagates through the sample, the lines are severed resulting in an increase in the electrical resistance. We now wish to adapt this technique to DCDC samples on oxide glasses. The thesis goal is the development and application of the potential drop techniques to DCDC samples. The challenge concerns the spatial temporal resolution (50 µm and 1 ns) in comparison to the crack tip velocity and sample size. The thesis student will take part in all the steps to realize the experiments: designing and depositing patterns (series of strips) on the glass surfaces using a cleanroom, running sub-critical cracking experiments in Region II and III, and analyzing data acquired during the experiment.

Development of solid porous siliceous supports for actinide sorption - Behaviour under irradiation

The aim of this research project is to study the densification of a mesoporous structure under the effect of irradiation damage produced by the presence of an actinide (238Pu) in the porous structure. To achieve this, siliceous materials based on mesoporous silicas modified by the addition of additive elements (B, Al, etc.) will be used. The purpose of adding these elements is to weaken the mesoporous structure in order to promote densification. The characteristics of the mesoporous structure (pore diameter, wall size, symmetry of the pore network) will be other parameters of the study. These materials will be functionalised with phosphonate ligands for actinide adsorption: thorium as a simulant in a preliminary stage, followed by plutonium. The final part of this work, which will continue beyond the thesis, will involve using various techniques (SAXS, BET, microscopy, etc.) to study the evolution of the mesoporous structure under the effect of irradiation damage as the material ages. This fundamental research work could have spin-offs in the field of nuclear waste conditioning materials: ageing of gels on the surface of nuclear glass, support material for decontaminating radioactive effluents. Part of the work will be carried out at CEA Marcoule's Atalante facility.

Modeling of Critical Heat Flux Using Lattice Boltzmann Methods: Application to the Experimental Devices of the RJH

The Lattice Boltzmann Methods (LBM) are numerical techniques used to simulate transport phenomena in complex systems. They allow for the modeling of fluid behavior in terms of particles that move on a discrete grid (a "lattice"). Unlike classical methods, which directly solve the differential equations of fluids, LBM simulates the evolution of distribution functions of fluid particles in a discrete space, using propagation and collision rules. The choice of the lattice in LBM is a crucial step, as it directly affects the accuracy, efficiency, and stability of the simulations. The lattice determines how fluid particles interact and move within space, as well as how the discretization of space and time is performed.

LBM methods exhibit natural parallelism properties, as calculations at each grid point are relatively independent. Although classical CFD methods based on the solution of the Navier-Stokes equations can also be parallelized, the nonlinear terms can make parallelism more difficult to manage, especially for models involving turbulent flows or irregular meshes. Therefore, LBM methods allow, at a lower computational cost, to capture complex phenomena. Recent work has shown that it is possible, with LBM, to reproduce the Nukiyama cooling curve (boiling in a vessel) and thus accurately calculate the critical heat flux. This flux corresponds to a mass boiling, known as the boiling crisis, which results in a sudden degradation of heat transfer.

The critical heat flux is a crucial issue for the Jules Horowitz Reactor, as experimental devices (DEX) are cooled by water in either natural or forced convection. Therefore, to ensure proper cooling of the DEX and the safety of the reactor, it is essential to ensure that, within the studied parameter range, the critical heat flux is not reached. It must therefore be determined with precision.

In the first part of the study, the student will define a lattice to apply LBM methods on an RJH device in natural convection. The student will then consolidate the results by comparing them with available data. Finally, exploratory calculations in forced convection (from laminar to turbulent flow) will be conducted.

Dislocation glide in body-centered-cubic high-entropy alloys

High entropy alloys are single-phase multi-component solid solutions, all elements being present in high concentrations. This class of materials has significant improvements in mechanical properties over "conventional" alloys, particularly their high strength at high temperature. It is commonly accepted that good mechanical performance comes from the interactions of dislocations with the alloying elements and that at high temperature interstitial impurities or interstitial doping, such as oxygen, carbon or nitrogen, play a preponderant role. The study of plasticity in concentrated alloys with a body-centered cubic crystal structure in the high temperature range therefore constitutes the objective of this PhD thesis. The associated technological challenges are important, these alloys being promising structural materials, notably for nuclear applications where operating temperatures above room temperature are targeted.
This work aims to understand and model the physical mechanisms controlling the mechanical strength of these alloys at high temperature, by considering different concentrated alloys of increasing complexity and by using atomistic simulations, in particular ab initio electronic structure calculations. We will first focus on the binary alloy MoNb before extending to the ternary alloys MoNbTi and MoNbTa and studying the impact of oxygen impurities on plastic behavior of these alloys. We will model the dislocation cores and analyze their interaction with interstitial and substitutional elements in order to determine the energy barriers controlling their mobility. Based on these ab initio results, we will develop strengthening models notably allowing us to predict the yield strength as a function of temperature and alloy composition.
This work will be carried out within the framework of the DisMecHTRA project funded by the French National Research Agency, allowing in particular to compare our strengthening models with the data from the experiments which are planned in the project (mechanical tests and transmission electron microscopy), and which will be carried out by the other partners (CNRS Toulouse and Thiais). The PhD thesis, hosted at CEA Saclay, will be co-supervised by a team from CEA Saclay and MatéIS (CNRS Lyon).

Perovskite ferroelectric oxynitride thin films with tunable properties

N-doped oxides and/or oxinitrides constitute a booming class of compounds with a broad spectrum of useable properties and in particular for novel technologies of carbon-free energy production. Indeed, the insertion of nitrogen into the crystal lattice of a semiconductor oxide allows, in principle, to modulate the value of its band gap or to introduce additional electronic states and thus to obtain new functionalities and optical properties. The production of oxynitride single crystalline thin films is highly challenging. In this essentially experimental thesis work, thin films of oxynitrides will be developed by atomic plasma-assisted molecular beam epitaxy. We will start from BaTiO3, which synthesis is well mastered in the laboratory, to realize co-dopings with nitrogen and compensating metals in order to preserve the neutrality of the elementary unit cell. The resulting structures will be studied for their chemical compositions, crystalline structures and ferroelectric characteristics. These observations will be correlated with their performance for the photo-electrolysis of water, which allows the virtuous production of hydrogen. Finally, the corrosion resistance of these new materials will also be studied.
The student will acquire skills in a wide range of ultra-high vacuum techniques, molecular beam epitaxy growth, clean room lithography, ferroelectric measurements and photo-electrolysis of water, as well as in state-of-the-art synchrotron radiation techniques.

Giant magnetoresistance resistors for local characterization of surface magnetic state: towards Non-Destructive Testing (NDT) applications

CIFRE thesis in the field of non-destructive testing using magnetic sensors in collaboration with 3 partners:

Laboratoire de Nanomagnétisme et Oxyde (SPEC/LNO) du CEA Paris-Saclay
Laboratoire de Génie Electrique et Ferroélectricité (LGEF) de l’INSA Lyon
Entreprise CmPhy

Kinetics of segregation and precipitation in Fe-Cr-C alloys under irradiation : coupling magnetic, chemical and elastic effects

Ferritic steels are being considered as structural materials in future fission and fusion nuclear reactors. These alloys have highly original properties, due to the coupling between chemical, magnetic and elastic interactions that affect their thermodynamic properties, the diffusion of chemical species and the diffusion of point defects in the crystal. The aim of the thesis will be to model all of these effects at the atomic scale and to integrate them into Monte Carlo simulations in order to model the segregation and precipitation kinetics under irradiation, phenomena that can degrade their properties in use. The atomic approach is essential for these materials, which are subjected to permanent irradiation and for which the laws of equilibrium thermodynamics no longer apply.

The candidate should have a good background in statistical physics or materials science, and be interested in numerical simulations and computer programming. The thesis will be carried out at CEA Saclay's physical metallurgy laboratory (SRMP), in a research environment with recognised experience in multi-scale modelling of materials, with around fifteen theses and post-doctoral contracts in progress on these topics.

A Master 2 internship on the same subject is proposed for spring 2025 and is highly recommended.

Top