Study of the security properties of emerging Si technologies & devise of new countermeasures against physical attacks.

With the growing use of Commercial Off the Shelf products as micro-chips in embedded security devices, it is of utmost importance to design Integrated Circuits (ICs) based on technologies that can be ?intrinsically robust? against physical attacks like side-channel attacks, fault injections or even invasive reverse-engineering/modifications. Several technologies for low power high density IC designs are under development in the laboratories of the Leti. The aim of this thesis is to investigate about the security properties of those cutting-edge technologies (like FD-SOI, FINfet, TSV?): security characterisations of those technologies shall allow us either to devise how they can be used to devise new counter measures against physical attacks or to find out how to influence their development to make them ?intrinsically more robust? against such attacks.

Architecture and Design of an Ultra Low Power Asynchronous Processor Core for Sensor Networks

After the blooming of ?internet for everybody?, is opening the new era of ?internet of things?, were everyday life objects will integrate a micro-system which will monitor, inform and alarm. Various applications are foreseen for these sensor networks: medical monitoring, industrial monitoring, environment sensing, home automation, etc. One of the main technical challenges is to design ultra-low power sensing node, to offer fully autonomous node, through energy harvesting or micro-battery. In this PhD, it is proposed to apply asynchronous logic design techniques (without clocking), in order to design an ultra-low power micro-controller core, adapted for these sensor networks applications. Asynchronous logic offers interesting properties for these applications such as: robustness to working conditions, intrinsic low power logic, efficient and fast wake up and sleep phases. In this PhD subject, the candidate will: - Propose a processor core architecture adapted to these sensor network applications, with innovative design solutions (instruction type, data encoding, adapted arithmetic), - Propose efficient power control mechanisms, including wake-up and sleep, in order to offer reactivity and computation at minimal energy level, and this using a wide voltage range, - Design the processor core using specific CAD tools for async. logic, targetting an advanced CMOS technology (such as FDSOI 20nm, offering low power & ultra-wide voltage range) - Participate to the core integration in a full system, for silicon fabrication, - Measure and compare performance results with state-of-the-art, - Valorize research outcomes through patents and scientific publications. The PhD work will take place within the LETI/DACLE/LISAN laboratory, which is designing digital architectures, with a strong focus on asynchronous logic and advanced CMOS technologies, and the PhD will benefit from expertises of other labs in order to interface with other sensor node elements (RF interface, sensing interface, energy harvesting, and embedded software).

Optimization of communication protocols and associated micro-architecture of optical networks on chip under thermal constraints

Considering ever-increasing bandwidth requirements between processors and memories in massively parallel architectures used in high-performance computing and servers in data-centers, the CEA-Leti institute is considering the use of optical networks between computing chips and memory chips on silicon interposer. In that perspective, it appears necessary to guarantee reliable electro-optical links between CMOS dies whatever the distribution of power supplies and temperature in the package. Indeed, the photonic modulators on silicon of small dimensions required for those optical communications show a large thermal sensitivity. Hence, this PhD thesis aims to propose thermal regulation mechanisms, calibration protocols for the electro-optical link depending of the signal quality, and several architectural and design counter-measures such as spatial and temporal redundancy, and dynamic reconfiguration of data encoding depending on the environmental conditions.

Wireless transmission architectures and circuits in sub-THz band

This research work aims the fabrication of a wireless transceiver chain in the sub-THz frequency band. The aimed throughput of the future RF front end circuit is between 20 Gb/s and 30 Gb/s. The transceivers will be fabricated in Silicon technologies (advanced CMOS or BiCMOS) and will integrate the transmission antennas. As a first step, the frequency bands (120 GHz / 240 GHz) should be identified and the related regulations studied. The available CMOS and BiCMOS technologies will be characterized in order to adapt the transceiver architecture and the circuit design to such high frequency bands. Architecture and system level studies will be performed, coupled to potential modulation schemes in order to identify the most relevant system architecture for each used technology. Link budget calculation will be used for global system and building block specification. Finally, the system will be designed and fabricated in CMOS or BiCMOS technology. Le PhD candidate will use and/or develop simulation tools such as ADS, Fasyle, Matlab and CADENCE. A good knowledge is required on wireless telecommunication systems, Silicon integrated circuit design and microwave engineering.

Radiolocation Algorithms Adapted to High-Precision Ultra Wideband Rake Receivers Using Multiple Antennas

Making possible unprecedented user-centric services (monitoring/smart inventory of personal goods from a smartphone, context-aware indoor navigation, etc.), new radiolocation capabilities appeared in wireless networks are on the verge of modifying in depth our daily life and habits. This revolution might even be comparable to that of GSM cellular communications or GPS-based car navigation in the past. In this context, an Impulse Radio ? Ultra Wideband receiver (IR-UWB) endowed with multiple fingers (RAKE) has been recently integrated with success. The latter claims low consumption (a few 10s of mW in average), while enabling the detection of weak incoming signals and enjoying scalable data rates (<50Mbp) that depends on the targeted transmission ranges (up to several 100s of meters). One more particularity lies in the capacity to estimate precisely the arrival time (TOA) of transmitted pulses, without any high-speed sampling operation on the receiver side. This solution can thus theoretically provide peer-to-peer distance estimates between radio devices within a precision of a few centimetres. Using multiple antennas on both transmitter and/or receiver sides can also help to improve signal sensitivity ahead (and hence, transmission ranges accordingly), as well as to jointly deliver angles of departure/arrival to complement estimated arrival times. However, in most of indoor non-controlled operating conditions (radio device hold in the user's hand or in the pocket), propagation phenomena can play a harmful role on the final localization performance. Practically, the obstruction of radio links by bodies and/or walls, or even an unfavourable device attitude, can lead to significant errors on measured distances and finally on position estimates. But the time/space correlation of received multipath components, as well as the highly timely and localized nature of the most severe obstruction cases, can be beneficial to try to enhance performance. In this PhD work, the main contribution consists in theoretically investigating, designing and evaluating new radiolocation algorithms that could take profits from the specificities of such high-precision IR-UWB devices. In particular, significant multipath components resolved at the receiver fingers shall be exploited further. We will also consider deploying multiple antennas at fixed reference nodes (base stations), as well as at mobile terminals. One goal would be to guarantee performances equivalent to that of optimistic visibility cases. One side objective would be to provide new detection functions (obstructions, wrong device attitude or geometric ambiguities).

Study of an ionization chamber of variable pressure for radiopharmaceuticals

The Laboratoire National Henri Becquerel (LNHB) of the Commissariat à l'Energie Atomique (CEA) is a national metrology laboratory for ionizing radiation. In this context it provides development, maintenance and the transfer of national reference for the international namely the Becquerel and the gray system units. Pressurised well type ionization chambers are used for the Becquerel unit transfer to users. These instruments (so called radionuclide calibrators) are widely used in the nuclear medicine services for activity measurements prior to its administration into patient for diagnostic purposes. LNHB has in charge the calibration of radionuclide calibrators for France. One of the nuclear medicine services difficulties is the access to calibration factors adapted to the clinical measurement conditions in terms of sample geometry. Thesis work consists on one hand in the supervision of the construction of the ionization chamber and in its measurement chain realisation. On the other hand, the work consists in the Monte Carlo simulation of the source-detector system in order to determine calibration factors for specific geometries (adapted do clinical conditions), in several conditions like nature and pressure of the detection gas (Argon, Nitrogen and xenon). Various packaging Radiopharmaceuticals (syringes, vials, ?) calculated calibration factors will then be compared to the experimental ones for three nature of detection gas under a specific pressure.

Open capillary microfluidics on plastic or cellulose substrates

Microassays on paper (paper strips or lab-on-paper) are presently gaining momentum for biologic diagnostics. These systems are attractives because there are simple, user-friendly, and low cost. Moreover, new functionalities are progressively incorporated to make these systems more versatile. However, they have still drawbacks such as absorbing most of the sample liquid before the active region, or relatively long transfer times. In order to develop and improve these systems, it is proposed to link the active regions (absorbing cellulose fibers) by micro-grooves that will trigger fast capillary microflows. Such a device avoids the precedingly cited drawbacks and maintains the advantages of paper-based or open plastric-based systems. Moreover, it is more versatile for the microflows, and enables the use of many microfluidic links on a same substrate.Besides, it seems possible to develop electro-capillary valves using simple embedded electrodes which will still increase the potentialities of the system. The aim of the work is the demonstration of the feasability of such hybrid systems comprising active paper regions and plastic etched grooves, in order to increase the potentialities of already existing chips and to develop new hybrid systems with a wide panel of functionalities.

Sub-nanometer electron tomography for semiconductor devices

FDSOI technology aims at reaching 28nm CMOS transistors by limiting leakage currents and device complexity. It requires a 3D characterisation featuring sub-nanometer resolution. At the moment, no technique allows to perform such a characterization. Still in the semiconductor field, intensive research is carried out on quantum dots, for ultimate transistor and optical applications. The lateral localisation of these quantum dots is today impossible due to the lack of 3D nanoimaging tools. Electron tomography is a 3D imaging technique that allows reconstructing a 3D image of an object from its 2D projections. So far, resolutions achievable are in the nanometer range. This research topic deals with the development of electron tomography in order to reach a resolution of a few Angstroms. Specifically, electron microscopy in STEM mode, that today allows observing atomic columns, will be extended in 3D so that the whole 3D distribution of atoms will be retrieved with an isotropic, deep sub-nanometre resolution. Preliminary works have already been done by few teams in the world. The research work will consist in developing the technique by tuning the electron optics, developing new 3D and efficient sampling and using the last generation of reconstruction algorithms. Of course, the work will be applied to the most recent devices, after properly shaping the samples using ion beam sample preparation.

Study of MEMS / NEMS devices on silicon photonics circuits

These several last years technological advances in the field of silicon micromachining have initiated the industrial growth of Microsystems / Nanosystems Electro Mechanical (M / NEMS) for fabricating sensors or actuators. In the field of NEMS with sub-micron sizes the properties allow for targeting applications in biomedical or biochemical analyses. It has been demonstrated that these nano mass (or force) sensors reach resolutions of the order of zepto gram (10-21g) or pico Newton, hence allowing early diagnosis of certain cancers. Transduction schemes of these systems are currently based on electrical principles. Many teams have nevertheless shown that photonics could operate and detect tiny displacement (fm). This hybrid technology, photonic circuit associated with M / NEMS potentially offers a significant improvement compared to electrical transduction. The purpose of the thesis consists of developing opto-electromechanical model with analytical approach and numerical simulations as well. First different transduction approaches will be studied. In particular the performance of optical detection will be compared to nanoscale electrical detection features. The comparison will be based on objective criteria (sensitivity, noise, crowding) for designing original optomechanical structures. The student will also develop the bench for optical and mechanical characterizations. Measurements on existing components (photonic circuit & NEMS) will allow to better understand the design constrains. This thesis has two parts that are the simulation and characterization. Both parts will run in parallel.

Study and characterization of the passivation layer of photodiodes for the optimization of the next generations of cooled infrared detectors

Binary-level analysis for malware detection

Software security is a major concern of the IT society: cyber-attacks can indeed affect people, organisations or even states. Malware are the most common way of launching an attack [5]. The standard counter-measure relies on detection through syntactic signature and eradication. Unfortunately, syntactic signature are easy to fool using obfuscation methods [6]. The last advances in malware detection are based on more robust "semantic" signatures, like the ones defined at the High-Security Lab (LHS) of LORIA (Nancy, France) [1,4]. However, the signature is still difficult to compute because obfuscation can mask part of the malware code. On the other hand, recent advances made at the Software Security Lab (LSL) of CEA (Saclay, France) allows to recover precise approximations of the control-flow graph (CFG) of (non-obfuscated) programs given as executable codes [2-3]. The broad scope of this PhD proposal is to design original binary-level program analysis techniques geared toward malware comprehension and detection. More precisely, the goal is to study how recent malware detection techniques [1, 4] can benefits from the latest advances in binary-level program analysis [2,3]. The chosen candidate is mainly expected to: 1- extend current binary-level CFG recovery techniques to the case of obfuscated codes, possibly combining static and dynamic methods and exploring trade-offs between completeness and precision; 2- evaluate the gain, if any, of more precise signature computation on malware detection; 3- develop techniques matching the severe speed constraints of real-time malware detection. --- References [1] Beaucamps, P., Gnaedig, I., Marion, J.-Y. : Behavior Abstraction in Malware Analysis. In : Runtime Verification 2010. Springer, Heidelberg (2010) [2] Bardin, S., Herrmann, P. : OSMOSE : Automatic Structural Testing of Executables. Inter- national Journal of Software Testing, Verification and Reliability (STVR), 21(1), 2011 [3] Bardin, S., Herrmann, P., Védrine, F. : Refinement-based CFG Reconstruction from Un- structured Programs. In : VMCAI 2011. Springer (2011) [4] Bonfante, G., Kaczmarek, M., Marion, J.-Y. : Architecture of a morphological malware detector. In : Journal in Computer Virology (2009) [5] Filiol, E. : Les virus informatiques : théorie, pratique et applications. Springer, Heidelberg (2004). [6] Moser, A., Kruegel, C., Kirda, E. : Limits of Static Analysis for Malware Detection. In : IEEE ACSAC 2007. IEEE Computer Society, Los Alamitos (2007)9

Complex wired network topology reconstruction based on reflection coefficients and / or scattering parameters

Cable networks are nowadays of a great importance in many fields of application. In automotive, aeronautics, aerospace or construction, wires are considered as a critical system, subject to constraints that may create defects which themselves can cause serious accidents. Monitoring the state of the wires and performing diagnosis on them is often required during maintenance operations. The best methods are currently based on reflectometry. Embedded Systems Reliability Laboratory (CEA LIST, Saclay research center) is working on the developement and prototyping of quasi full-digital solutions based on FPGA custom boards in order to implement the various reflectometry methods available. These electronic solutions are nothing more than digital reflectometers. While measuring point-to-point wires is quite simple, the knowledge of topology is required to make a correct diagnosis as soon as junctions are present. This input is not always known. One can think for example of the electrical aging wiring of a building for which the maintenance operator has no blueprints of the cable networks. The objective of this thesis is to solve the inverse problem of the one we are accustomed to: the topology reconstruction of the network (assuming it is healthy) from reflection coefficients and scattering parameters measurements.

Formal verification of critical components of operating systems and Cloud hypervisors

The CEA LIST, based in Palaiseau (near Paris, France), is a key software systems and technology research center working on embedded systems (architecture and design of systems, methods and facilities for software and system dependability, and intelligent vision systems), interactive systems, signal detection and processing in partnership with the major industrial players in the nuclear, automotive, aeronautical, defense and medical fields. The Software Safety Lab (LSL) of the CEA LIST develop tools for verification and validation of software and software/hardware systems. One of the ongoing projects consists in developping a formally verified Cloud hypervisor. A first prototype called Anaxogoros was developed by the laboratory LASTRE of CEA LIST. This PhD aims to formally verify the implementation of critical components of operating systems and Cloud hypervisors with the help of static analysis methods (e.g. program proof, value analysis). One of critical components to be verified is the virtual memory service that allows to dynamically allocate memory, create, modify and destroy system objects such as domains, threads, etc. Another example of critical component requiring formal verification is access control module, based on a capacity mechanism. The critical components will be specified and proved using appropriate automatic or interactive proving tools (Jessie, Wp plugins of Frama-C, Coq proof assitant, etc.) The OS prototype Anaxogoros will be used as starting point of the research. First, the case of sequential (non parallel) execution will be studied and verified. An important difficulty of verification for operating systems (particularly important for Cloud hypervisors) is related to parallel execution on a multi-core processor. It will be necessary to specify (and if necessary to adapt algorithms of critical components in order to prove their properties for the parallel execution case. The candidate will have Master's degree on Computer Science, preferably in Software Verification and Validation, and good knowledge of formal methods.

Blood biomarkers dosage using red blood cells immuno-agglutination

The reference method for blood biomarkers dosage is ?enzyme-linked immunosorbent assay? (ELISA) technique. It is performed using blood serum or plasma, and implies a pre-analytical separation step of blood cells, usually achieved via centrifugation. For development of rapid biological test, such as Point-Of-Care (POC), the centrifugation step is not easily integrated, so biomarkers dosage implies the use lateral flow immuno-assays. The objective of this doctoral research work is to set up an alternative dosage method directly on whole blood. We propose to develop blood biomarkers dosage by following continuously an immuno-agglutination reaction of red blood cells using a bi-functional reagent. Owing this approach, red blood cells become the indicator of the dosage reaction, using a bispecific antibody which recognizes on one hand the targeted biomarker (protein, virus, micro-organism, cell) and on the other hand the red blood cells. Agglutination monitoring and detection will be achieved using imaging. This method is easily integrable in a portable device and presents sufficient resolution to visualize cells without any labeling and in real time during agglutination.

Heteroepitaxial Diamond on Iridium for Dosimetry

This PhD work aims to synthesize heteroepitaxial diamond films of very high crystalline quality on iridium buffer layers deposited on SrTiO3 (001) substrates and qualify them as radiotherapy dosimeters. This work will be performed at CEA LIST in collaboration with partners of the French ANR HIRIS project. The Diamond Sensors Laboratory started diamond heteroepitaxy on iridium in 2007. Diamond crystalline quality close to the state of art was still obtained [1]. Two main challenges still remain for heteroepitaxial diamond films on iridium: - First, the crystalline quality must be improved. Indeed, the reported dislocations densities in heteroepitaxial diamond films (> 5x10^8 cm-2) still limit the electronic transport properties. The crystalline quality of diamond layers could be improved by a better control of a homogeneous pattern of domains on iridium (001) substrates. This needs a deeper knowledge of the surface mechanisms occurring at the iridium surface during the nucleation stage using the Bias Enhanced Nucleation (BEN) process. According to the literature, the surface chemistry associated to the unique diamond nucleation mechanism is still not completely understood. After BEN, the surface chemistry (XPS, AES, UPS) and the topography (AFM) modifications of the iridium surface will be investigated using a UHV set-up at CEA LIST. It allows a sequential monitoring of the surface without air exposure. Complementary FEG-SEM observations will be done to characterize the morphology of domains. Specific CVD conditions adapted to the growth of high crystalline quality diamond with a low impurity level will be then applied to achieve thick free-standing heteroepitaxial (001) diamond films with a thickness of 300 µm. - Second, the up-scaling onto substrates 20x20 mm2 may allow reproducibility. After cutting, it will provide a set of 5x5 mm2 diamond samples grown in the same conditions by CVD assisted by Microwave Plasma in a home-made reactor. If these limitations were overcome, heteroepitaxial diamond films could constitute a convincing alternative to single crystal diamond for detection and dosimetry applications. The use of alternative substrates, bulk SrTiO3 (001) or SrTiO3 epitaxial (001) layers grown on Si will allow an easier up-scaling. It is expected to significantly improve the crystalline quality of the iridium buffer layer for diamond heteroepitaxy. The development of heteroepitaxial diamond for medical dosimetry is new. Selected layers will be mounted as detectors at CEA-LIST using a new dosimeter packaging developed by our team. Dosimeters will be tested according to the recommendation advocated by IAEA (International Atomic Energy Agency) to be used in hospital. Their characteristics will be compared to diamond single crystals and other existing devices used for novel radiotherapy techniques. [1] A. Chavanne, J. Barjon, B. Vilquin, J. Arabski, J.C. Arnault, Diam. Relat. Mater. 22 (2012) 52?58.

Cell Disruption Adaptative Device based on Electrical Pulses

Many processes require the opening of microorganisms to extract their contents to produce bio-molecules or to identify them in laboratory tests. The ultimate goal of the proposed thesis is to develop a microfluidic module able to "open" microorganisms for analysis or to exploit their productive potential. This module will be designed to be autonomous, capable of adapting to the physical properties of the sample and should have a low energy consumption. It will easily integrated in analysing tools. Pulsed electric fields can open cell membranes. Easy to integrate, without the need for reagents and without side effects, the technique is promising to improve current processes. We propose to explore pulsed electric fields in an original configuration which should focalise energy into a small volume where microorgisms will be forced to go. Effects should be also amplified to open membranes of the most resistant microorganisms. Furthermore a good knowledge and understanding of the technic should lead to the development of self-adaptating device.

PowerAmplifier-Antenna co-design for high efficiency miniature RF transmitter modules

CEA-LETI currently leads research activities towards integration of RF systems on chip to propose compact and high efficiency solutions for future wireless applications. In this context, new integration schemes have to be investigated and objectives of the thesis will be to study new architectures and circuits in order to propose miniature RF transmitter modules with low loss. The proposed solutions will have to take benefit of a PA-Antenna co-design approach and demonstrators in the 0.5?5 GHz frequency range with high efficiency watt-level characteristics will be developed and integrated in SOI CMOS technology to validate the proposed solutions. The candidate should be highly motivated and self-working, and be familiar with analog and RF CMOS integrated circuits. He should also be familiar with CAD tools such as CADENCE and ADS. In addition, antenna design notions will be highly appreciated.

Analysis of multi-sensors NEMS signals applied to gas chromatography

Devices dedicated to gas analysis have a large class of applications, such as the control of air and natural or industrial gas quality, or the study of respiratory gases. LETI works together with the APIX Company on a new generation of microsystem devices combining sample preparation stages, gas component separation using microchromatography devices integrated on silicium and transduction based on NEMS (Nano Electro Mechanical Systems) detectors. The challenge is to combine high sensitivity to detect very low quantities of compounds, and efficient separation capacity to overcome the very high complexity of the mixtures and to identify the signature of targeted molecules. The objective of this signal processing thesis is to state a relevant formalisation of the analysis issues for the output signals. It will extend our processing methodology based on inverse problem approach associated with a hierarchical model of the analytical chain to gas microchromatrography devices with multiple NEMS sensors. The first targeted application is the monitoring of the calorific power of natural gas. The main contribution will address decomposition of NEMS signals, multi-sensors fusion, auto-calibration, and temporal dynamic monitoring.

Automatic exploit generation from error traces

One of the key points in cyber-security is to ensure that a given piece of software is free of exploitable bugs (vulnerabilities). Verification and automated testing methods are currently adapted to security concerns. However, while these methods are effective at finding bugs, these bugs are not always relevant in security context since they may not be exploitable. The proposed PhD work focuses on the exploitability analysis of a bug, given a faulty execution trace leading to this bug. The goal is to design methods able to automatically classify the "degree" of exploitability of a given bug, by using formal verification techniques for "generalising" the initial (non-exploitable) trace into an exploit. The candidate can re-use former ideas developed by his advisers in terms of binary code analysis and vulnerability detection. The main challenges here are to go beyond state-of-the-art on the following aspects: the generalisation method, the exploitability model and the environment modelling. The successful candidate will be hosted both at CEA (Paris area, France) and University of Grenoble (Grenoble, France). He will be supervised by Sébastien Bardin (CEA, co-supervisor) and Marie-Laure Potet (Uni. Grenoble, supervisor). This work is part of the BINSEC project (2013-2017), funded by ANR (French research agency), a 4-years project gathering top-level academic and industrial partners in order to advance the state-of-the-art of binary-level security analysis.

Liquid-based electrostatic energy harvesters

Mechanical energy harvesting is a growing technology aimed at developing innovative devices able to turn ambient mechanical energies (shocks, vibrations...) into electricity. This thesis will be aimed at developing mechanical energy harvesters exploiting state, shape or permittivity changes?that appear in some liquids when submitted to external effects such as vibrations, pressures, shocks?and to couple these effects with an electrostatic converter: the state change of the liquid induces a variation of capacitance that is then turned into electric energy. Because of their small output powers, these systems are firstly aimed at powering wireless sensor networks for industry, cars, habitat?in order to replace/recharge batteries. Yet, today, researches are targeted on more consumptive applications such as mobile devices recharging (phones, MP3 players?). And in fact, liquid-based energy harvesters seem to be promising for these applications. For more information: S. Boisseau, G. Despesse, B. Ahmed Seddik. Electrostatic conversion for vibration energy harvesting, Small Scale Energy Harvesting, Intech, 2012, http://dx.doi.org/10.5772/51360

Study of strain and electrical properties in Si nanowire transistors

Today novel multi-gate architectures have attracted attention as they allow pursuing Moore's law beyond simple scaling. A good understanding of strain in these 3D devices as well as physics behind it is needed to optimize these innovative devices. The two main challenges of the phD thesis will be to locally measure strain in nanoscaled transistors (Tri-gate, Omega-gate and fnFETs) and to evaluate the effects of stress on the electrical properties. In particular the objectives will be: 1) to locally measure strain in nanoscaled transistors by nano-beam electron diffraction (NBED) 2) study the effects of strain on transport properties by advanced electrical characterization, including extraction of the piezoresisitve coefficients, low temperature measurements... 3) understand and modelize the strain effect on transport properties in these 3D devices. The work will be done within the electrical characterization laboratory in strong interaction with the nanocharacterization platform.

Disruptive processes for ultrathin silicon layer transfer

The Smart-CutTM technology, invented in LETI, is the most efficient process for the production of SOI substrates at an industrial scale. Its principle is based on the fracture in a brittle silicon layer performed by ion implantation of light ions. Due to SOI specifications (Si top silicon layer thikness, uniformity....) provided by the semiconductor roadmap, constraints on the mastery of processes are becoming increasingly stringent. In this context, the objective of this thesis work conducted within the framework of a collaboration between the CEA / LETI and CNRS / CEMES will explore disruptive processes and adapted to these new constraints to obtain silicon fracture. It will therefore develop the new processes and perform studies to characterize them and to understand the mechanisms involved. We are seeking a motivated candidate with a taste for various experiments combining technological work on microelectronics tools and undamental studies to understand the mechanisms of the processes studied.

Modeling and evaluation of data retransmission protocols within a distributed multi-switches architecture

Hard real-time systems, such as the one found in the field of automotive, avionics or distribution of energy, are nowadays developed in the context of distributed architectures. In such systems, data streams have end-to-end network constraints to fulfill between their production and their consumption. The verification of the proper sizing of network resources compared to application needs is a normative requirement given the certification objectives of this type of systems. Existing approaches to design distributed hard real-time system, for example based on the use of an arbitration policy to access to a bus-based network or when using network switches with full-duplex links, requires the oversizing of network resources. In this PhD thesis, we propose to model and characterize a new paradigm to propagate real-time data, and not simply of frames, within a multi-switch architecture in order to increase the utilization made of the network capacities and thus reduces the costs of such system. In particular, the definition and the study of data propagation & aggregation strategies will be carried out taking into account the end-to-end constraints of these data. Validation will be carried out by simulation and possibility through experimental evaluations.

Influence of surface modification on Li distribution and SEI formation in Li-ion silicon-based negative electrodes by combining ToF-SIMS/Auger/XPS characterization

A Li-ion battery is typically composed of a positive electrode and a negative electrode. The latter is generally based on graphite material for its reliable behaviour though a limited specific capacity of 372 mA.h/g. For that reason, silicon appears to be a promising alternative material thanks to a high specific capacity of 3580 mA.h/g. However, the biggest issue - a high volumic expension of more than 300% during Li insertion - favours particle grinding and Solid Electrolyte Interphase (SEI) instability, leading to electrode degradation and electrochemical performance loss with time. In order to get rid of these problems, it is proposed to nanostructure the surface of the electrode material, thus avoiding any contact with the electrolyte. Among the objectives of the PhD research, the influence of surface modification, electrode formulation or test conditions on Li distribution and SEI formation will be evaluated. This will be allowed with preparation tools which are implemented on the Nanocharacterization Platform in Minatec, in particular an in situ FIB implemented on a new ToF-SIMS. Li distribution in a single particle, or throughout the entire electrode, will be easier to determine, by using ToF-SIMS spectroscopy, and also Auger spectroscopy with the help of an UHV transfer suitcase. The results will help in understanding the electrode behaviour, opening ways for improvements relative to negative electrode durability and cycling performance. The research work will consist, in a first step, in developing skills for the use of in situ FIB with ToF-SIMS, in particular for the specific topic described above, and also for UHV transfer based studies between ToF-SIMS/Auger/XPS. In a second step, Li distribution and SEI evolution of reference materials and cycled electrodes will be characterized and correlated with the applied surface modifications. This study will be achieved in collaboration with an industrial partner that will provide the material and the necessary surface modifications.

Biomaterials for the controlled delivery of hydrophilic and/or lipophilic therapeutics; their inclusion in smart bandages.

Several biomaterials, mainly polymers of natural origin, are studied or already used in medical devices. Different systems are designed for the encapsulation and delivery of therapeutic molecules; however chemical drugs and biomolecules (proteins, SiRNAs?) are not always well included in the material, and their kinetics of release is poorly controlled. We have developed, patented, and tested in vitro and in vivo biocompatible lipid nanoparticles able to vectorize lipophilic and/or hydrophilic therapeutics. The objective of the thesis is to include these nanoparticles, with performing vectorization properties, in biomaterial matrices based on polysaccharides, used for their structuration properties. The association of drug nanocargos with the structuring polymer matrix will be made by chemical bonds that can be cleaved specifically by external stimuli such as an electric field or light irradiation. Thus, the systems developed will allow a controlled release of drugs and biomolecules over time and space. In the thesis, the synthesis of these innovative bio-inspired nano-polymer materials, their characterization and their structuration, will be discussed. The new materials will be used to design smart bandages with controlled drugs and biomolecules kinetics of release, as an application demonstrator.

Development and understanding of high performance GaN based layers for improved power transistors

To enhance the development of alternative energies, CEA-LETI is working on high power electrical components, which are a key element of electrical converters needed for photovoltaic solar panels and electric vehicles. CEA-LETI is working on a breakthrough technology using gallium nitride, which has the potential to surpass the theoretical limits of silicon technologies. However, gallium nitride technology is significantly less mature than silicon, in terms of material quality, development of transistor processing, performance of devices and materials understanding. This PhD at CEA-LETI is an exciting opportunity to grow gallium nitride on 200mm substrates using a commercial MOCVD tool, as part of a project to integrate new high performance materials into the cleanroom environment. It is also a great opportunity to develop an understanding of these materials working with our excellent characterisation department. MOCVD is a very powerful and flexible technique for growing high quality semiconductor materials. However, this flexibility also means complexity, and therefore there is a great deal of work to be done to understand the effects of the different parameters on the quality of layers grown, and their performance. This PhD will give a student the rare opportunity of developing high quality advanced layers for transistors, with the support of our world class characterisation suite at LETI, which includes such techniques as KFM, C-AFM , XPS, CL, PL, TEM, SIMS, atom probe etc. In addition, the use of 200 mm substrates allows the reproducible processing of promising layers into full transistors using the silicon quality cleanroom at LETI, to allow correlation of electrical properties to the layers grown.

Numerical modeling of solute transport phenomena and implementation of a mixing system in a crystallization pilot for photovoltaic silicon

Today, photovoltaic cells are mostly fabricated using silicon substrates. The most common technique used in industry involves ingot directional solidification from a liquid bath. Then, these ingots are cut into wafers which are processed into solar cells. The main research objective in the field is focused on the reduction of the production costs with two major trends: an ingot size increasingly bigger and the use of lower quality silicon compared to the electronic industry standard. In this context, the issue of control of convection in the silicon melt becomes decisive, specially regarding impurity segregation issues. The purpose of the thesis is to optimize the convective field in the melt during the growth of silicon ingots by directional solidification. To reach this objective, the candidate will initially identify the most appropriate methods. Then the candidate will have to implement one experimental water set up to characterize qualitatively and quantitatively the flow generated. Finally, the candidate will have to develop the corresponding numerical simulation models and use experimental measurements as a basis for parametric adjustment.

Lightweight Security Protocols for IP-based Wireless Sensor Networks and the Internet of Things

Wireless sensor networks (WSNs) are increasingly used in several industrial domains, such as industrial process monitoring and control, e-health monitoring, and smart grid. Such networks are composed of resource-limited devices equipped with radio communication units and sensing/actuating capability. Conversely, their resource limitations (e.g. in terms of energy, processing power, memory) and the lack of communication reliability make it challenging to use existing cryptographic techniques and common standard protocols for their security. With the aim to reinforce the trend of future networks towards all-IP, the focus of thesis efforts will be to adapt legacy security standards (e.g., EAP, TLS, IPSec) to run in WSNs, and to extend standardized solutions proposed for IP constrained environments (e.g., RPL, CoAP) with security features. Considered security services in WSNs will cover multiple layers: hop-by-hop frame protection, as well as network authentication and access control, and end-to-end data security. In order to evaluate and validate the design, security, and performance of the proposed solutions, a testing framework will be devised to examine the behavior of the secure WSN under realistic attacks.

Embedded broadband measurement of electrical impedance - application to batteries

The context of the study is the monitoring of the state (charge, aging, security) of a battery of an electric vehicle. The thesis aims the embedded identification of the electrical impedance of the battery taking into account the non-linearity of the system. This objective should help to overcome the limitations of existing measurement systems (spectroscopy, linear broadband identification). The design of the architecture of the system identification, the development of the signal processing, and tests will be conducted in close collaboration with the LITEN, CEA laboratory developing innovative accumulators and batteries.

Multi criteria analysis of the 3D sequential integration (Power / Performnace / Area) for 11 and 7 nm node

The PhD goal is to find the process strategies for the 3D sequential integration in order to outperform the standard planar technology (gain of one node in term of Power/ Performance/ Area (PPA) without scaling the transistor). PPA analysis will be runned on processors and will enable to choose the best technological options. The questions of integration will concern the i) partitionning level (N over P or CMOS over CMOS) ii)TYpe of metal line (W or Cu), type of dielectrics (standard or lowk), iii) Vt options (find solutions to adapt the FET threshold voltages, iv) contact technology (autoaligned pre contacts? Shared or through Si contact). The candidate's work ids divided in two tasks: 1- technolgical developement of the above cited technologies and 2-PPA evaluation. The technological studies will enable to define the design rule manual with more details and to anticipate the technological complexity for different options. This methodology will allow the candidate at the end of his PhD to conclude of the integration(s) offerinf the best Power/Performance/Area compromise.

New graph partitioning models for tasks assignment and routing on massively parallel architectures

The emergence, in recent years, of the so-called many-core processor architectures, that is microprocessors integrating hundreds if not thousands cores on a single chip, has created a new field of applications for graph partitioning problems. Indeed, graph partitioning issues are central when one needs to map a network of tasks on such an architecture. Still, the first models which have been studied only coarsely reflect the true complexity of the underlying architectures and new, more complete, models are required. In this context, a first objective of the proposed PhD thesis will be to focus on defining new models which will then induce new variants of the graph partitioning problem. A second objective will be to focus on studying their mathematical structure so as to propose, and experimentally validate, practically relevant resolution algorithms.

Development of sensors monitoring lactates and transfer on imprinted electronic

Non invasive wearable devices allow monitoring in real time and continuously some physiological individual parameters. So, these devices find applications mainly in the military domains, health or sport. However, these systems are limited to the monitoring of very simple parameters such as the pH, the conductivity, the concentration in ions sodium. However, the detection and the monitoring of more complex elements such as small molecules in physiological liquids can be essential for some applications. So, the monitoring of lactates in the sweat allows to have an indication of the muscular activity of the sportsmen. We propose a subject of thesis with the objectives to develop a flexible patch containing electrochemical organic transistors allowing the real time monitoring of the lactates concentration in the sweat of sportsman and to test it on a person.

Design of an hydrogen production plant based on Solid Oxide Electrolysis technology

The overall objective of the thesis is to design thanks to simulation the optimal architecture of an hydrogen production plant or system based on Solid Oxide Electrolysis technology (SOE), which leads to the minimum hydrogen production cost. One characteristic of the SOE technology is that up to several million single repeat units are put to work in a serie/parallel architecture addressing 4 exchange networks which are at the electrical anode/cathode fluidic and thermal level. The challenge consisting in ensuring the overall efficiency as well as a safe working order, the way to drive the installation is not a trivial issue. The final assessment criterion is the miminum hydrogen production cost. The production scale can vary from hundreds of kilowatts up to hundreds of megawatts (100 kg/day à 500 ton/day hydrogen production). The electric conversion function not much studied up to now requires a special attention. The global approach is to use theoretical knowledge in fractals and networks as well as safety methodologies and to be inspired by similar problematics (high speed computer, photovoltaic pannels, batteries), where there are only thermal and electrical issues.

Aging analysis of a battery module under vibration constraints

The objective is to understand the process leading to battery module aging under vibration constraints. The final goal is to have a clear understanding of the test plan needed for battery modules in electric vehicles. The idea is to validate: - the impact of mechanical vibrations on the aging of the module assembly of electrochemical accumulators (electrical connexion loss, increase of internal resistance, electronic board functionality, mechanical integrity...) - the impact of mechanical vibrations on the aging of electrochemical accumulators into a module assembly. The research process will associate experimentation and numerical modelling.

Managing dynamic reconfiguration in distributed, adaptive and energy aware real-time systems.

PhD thesis proposal Phd Title : Managing dynamic reconfiguration in distributed, adaptive and energy aware real-time systems. Sujet de la thèse : Gestion de la reconfiguration dynamique dans les systèmes distribués, adaptatifs, temps réel et à basse consommation. Description of the subject: Smart grids aim to use the current energy resources more efficiently by providing a more resilient and reliable infrastructure. The grid has indeed to manage the interactions between generation, storage and consumption of electrical power, one of its objectives is the increased customer participation and interaction with the overall grid [1]. Within a smart grid, actors in a particular domain (electricity sources, storage, distribution, service providers, markets and customers [2]) often interact with actors in other domains. Main characteristics of the grid are following: ? Heterogeneity: computing nodes vary from small sensors to larger control systems. Communication typically has different characteristics and requirements [3, 4]. ? Long lifetime, yet constant change: the system has a long lifetime (e.g. that of the building), but its components change frequently: for instance new sensors or storage devices might be added or replaced while the system as a whole continues to run. ? Robustness: (temporary) failures of some elements such as sensors must be tolerated (self-repair, self-protection facilities). ? Scalability: the system must support physical growth (e.g. integration of additional nodes, buildings) as well as new/updated applications for interacting with the customers. ? Resource and timing constraints, the system needs to respect end-to-end deadlines (required by control algorithms) and resource requirements, The applicative context of smart grid will be provided through a collaborative project with industrials of the domain: the Energy Positive IT project. In a previous PhD thesis, the laboratory examined systems which have a certain number of known configurations with respect to timeliness during reconfiguration [6]. The work of this new PhD thesis will extend the results, being much more oriented to dynamic changes that have not been planned at development time. Heterogeneity will be addressed by using an existing component based framework (developed by the laboratory) based on components, containers and connectors that abstract from OS and communication protocols. This framework also supports deployment for static systems and have to be extended for dynamically adapted systems. The principle is to compute the deployment configuration and actions from architecture and allocation models describing the grid and expressed by means of a ?smart grid specific modeling language? (to be designed during the PhD as a UML extension) that will: ? Allow for the definition of what applications/devices compose a smart grid and what dependencies link them together. ? Give the possibility to model energy requirements of end-users and their interactions with the overall grid A challenging aspect is that the model would need continuous updates to keep it synchronized with the architecture living at runtime. This model makes it possible to reason about the application state and performs reconfigurations. An adaptation engine, for instance, is thus able to select, test and validate an adaptation scenario on the model, before actually performing the adaptation on the running system [7]. Therefore, it is necessary that a part of the model exists at runtime [5]. It always reflects the current deployment and is the basis for authorizing additions (software or hardware). This model must be stored in the computing nodes, but not each node needs a complete copy, e.g. some may only need to know their neighbors. I.e. different nodes within the grid only have to store different subsets of the model. While selecting suitable subsets enables the respect of resource constraints, decentralized storage enables robustness and scalability. A part of the work is the synchronization between design models and runtime models, another part the runtime-model aware adaptation of the deployed systems. References: [1] Massoud Amin, S.; Wollenberg, B.F., "Toward a smart grid: power delivery for the 21st century," Power and Energy Magazine, IEEE, vol.3, no.5, pp. 34- 41, Sept.-Oct. 2005 [2] Smart Grid conceptual model, http://smartgrid.ieee.org/ieee-smart-grid/smart-grid-conceptual-model [3] Locke G., Gallagher P., NIST Framework and Roadmap for Smart Grid Interoperability Standards, Release 1.0, Office of the National Coordinator for Smart Grid Interoperability, NIST Special Publication 1108, US, 2010. [4] De Mues, M.O.; Alvarez, A.; Espinoza, A.; Garbajosa, J., "Towards a distributed intelligent ICT architecture for the smart grid," 9th IEEE International Conference on Industrial Informatics (INDIN), 2011, pp.745-749, 26-29 July 2011 [5] Blair, G.; Bencomo, N.; France, R.B. "Models@ run.time," Computer , vol.42, no.10, pp.22-27, Oct. 2009 doi: 10.1109/MC.2009.326 [6] Boukhanoufa, M. L.; Radermacher, A.; Terrier F.; Offline validation of real-time application constraints considering adaptation rules, 8th IEEE International Conference on Embedded Software and Systems (ICESS-11), Changsha, China, November 16-18, 2011 [7] Marc Léger, Thomas Ledoux, and Thierry Coupaye. Reliable dynamic reconfigurations in the fractal component model. In ARM '07: Proc of the 6th international workshop on Adaptive and reflective middleware, pages 1?6, Newport Beach, CA, 2007. ACM.

Modelling of block copolymers self-assembly

Nowadays, the use of multi-patterning techniques can no longer be avoided as existing photo lithography pieces of equipment are obviously lacking resolution. Luckily, new promising technologies are on their way to sort this issue out. Amongst them, self-assembly of bloc copolymers is a cost effective technique that allows patterning fine structures guided by patterns exposed using classical lithography. Materials and associated processes are improving at a steady pace so that it's now time to start building the infrastructure required for their use in production. This includes developing fast and reliable models that could be used to define the design rules and functional devices.

Compilation and Analysis of Functionnal Programming Languages for Hybrid Systems

The conception of industrial embedded systems is mainly based on high level graphical programming languages such as Matlab/Simulink. Such tools allow for a quick design of systems, but are not very well suited for a formal analysis of embedded systems. Furthermore, the fact that they are graphical languages make the modularity of programs difficult to exploit. This thesis proposes to define a new language for the modelisation of hybrid systems, based on the functional programming paradigm. The student will also define a compilation scheme for this language in order to simulate the systems, and we will study the use of optimisation techniques on the generated code to improve the simulation. Finally, the analysis techniques developped for the HySon tool(a guaranteed simulation engine developped at CEA) will be extended and implemented for this new language

Wide-field histopathology imaging

Imaging large fields (> 1cm2) of the biopsy is a strong need expressed by clinicists. The objective of this thesis work is to develop simple optical methods for analyzing fingerprint tissues or fluid samples. In particular, the method should allow to define regions of interest for further analysis by means of e.g. mass spectrometry, Raman spectroscopy, IR spectroscopy.

Embedded system for obstacle perception in dynamic environments

The objective of this PhD position is to propose a mixed hardware and software solution for an embedded perception system designed for mobile vehicles (buses, cars, etc?). This embedded system is dedicated to avoid potential collisions of vehicles in dynamic environments. In this context, a set of different type of sensors are fixed on a mobile vehicle. The extracted information is processed in real-time by a Bayesian algorithm called Bayesian Occupancy Filter (BOF). The processing of the BOF algorithm is particularly adapted to parallel architectures due to its matrix representation. Nowadays, this processing is performed by a work-station integrating a powerful GPGPU. The PhD candidate will propose an integration solution of this type of algorithm taking advantage of novel multicore embedded architectures. The algorithm optimisation as well as the hardware/software partitioning will be studied in order to achieve the needed real-time performances at a minimum energy cost. This work will take place at the LIALP laboratory of the DACLE division of LETI in collaboration with INRIA Rhone-Alpes.

Time-resolved multispectral optical tomographie

Optical tomography consists in illuminating biological tissues with infrared light with a large number of source points and in analyzing the medium response so as to reconstruct 3D optical characteristics. Envisaged applications are the diagnosis of disease or the study of biological processes. The subject of the thesis is to use pulsed light jointly with a multispectral approach to increase the informational content. The doctoral student shall implement a supercontinuum laser source associated with a system of wavelength selection and time resolved detection. Signal processing algorithms shall be developed to reconstruct the biological tissues on a material basis. Finally, the contributions of the approach will be assessed.

Development of a conducting and biocompatible polymer for monitoring of cells cultures

Diverse devices, allowing the multiplication of bacteria in micro-colonies forms on culture media in presence or not of antibiotics, exist on the market however these devices, although simple and effective, present important inconveniences such as very long time of incubation and the absence of information on the chemical messengers included during the cellular communication. To answer this problem, we propose a subject of thesis the main objectives of which will consist in i / to develop a conductive polymer biocompatible, permeable (PEDOT) for nutriments according to a process which will allow the obtaining of a rigid material with a fine thickness and transversal porosity and ii / to study the modifications of electric properties material during the cells multiplication.

Massively parallel integrated architecture for NEMS resonator readout

Silicon nano-cantilevers are promising solutions for ultra-high sensitivity mass detectors. Several research groups, including CEA-Leti, have demonstrated the possibility to use such nano-resonators in gas detection or mass spectrometry setups. The detection principle relies on the resonance frequency sensitivity to any mass addition onto the cantilever. In such a system, the electronic part used for resonance frequency tracking is critical. In today's experiments, it is made of discrete, high performance, components. The first integrated solutions are currently being developed to address a few resonators. This Phd work will focus on the following step, i.e. a readout integrated circuit able to address several 100000 resonators. The work will target the best architecture according to signal-to-noise ratio, robustness, speed and power criteria.

Correction of scatter radiation in multi-energy radiographic and tomographic imaging

X-ray imaging is a well-known technique used in the medical diagnostic domain, testing of industrial samples, and security. Conventional X-ray detectors provide a signal proportional to the energy of photons deposited into the detector, integrated on the total energy range. The pixel value is in fact the sum of two signals: the primary flux one, coming linearly from the detector, measuring the photons not attenuated by the photoelectric effect, and the scattered radiation, the interacted photons being modified in direction and energy. The scattered radiation induces modifications of image values, thus quantification bias, and is significant as soon as 2D detectors are used. Numerous methods exist for scatter correction. The recently emerged semiconductor detectors, providing the full energetic spectra for each pixel, imply the evolution of these processing methods. The global objective of the thesis is to develop estimation and correction methods of the scattered radiation when using spectral detectors. Numerical simulation and experimental devices will be used. Generalisation of the existing mono-spectral methods will be tested, then original methods based on all energy bins will be investigated. Both radiographic and tomographic modes will be considered.

Spectrometric detector for mammography

Detectors based on CdTe and CdZnTe have been developed for applications in X-ray and gamma ray high resolution spectrometry at room temperature. Recent advances in this type of detector and the development of fast readout electronics, allow to consider the use of this type of detectors for imaging applications or X-ray tomography. Several research teams have developed X-ray imaging systems with resolutions ranging from 250 to 1000 microns and a number of energy channels between 2 and 5. LDET laboratory has developed a novel solution providing high-resolution spectra of 256 energy channels for each pixel. Studies are underway to expand the concept of matrix detectors and higher resolutions. The main purpose of this thesis is to study the potential of high resolution spectroscopy in mammography. The thesis would be based on a coupling between simulation and experimental study of different architectures of detection and development of methods for processing data issued from these detectors. The needs in terms of information processing in mammography cover different applications, such as detection of nodules, detection of micro calcifications, distinction between tissue types (+ - fibrous) and contrast imaging.

Dielectric filled MEMS resonators for ultrasensitive mass detection in liquid

Learning for tracking: exploiting context and probability reasoning for reinforcing knowledge.

The Laboratory of Vision and Content Engineering (LVIC) brings together 80 researchers around research topics which are the analysis of multimedia data, the augmented reality and the scene analysis. Through these activities, the LVIC works with large groups such as Renault, Thales, SNCF as well as with a network of innovative SMEs DiotaSoft, Arcure ... Through R & D projects, the laboratory develops automatic video analysis systems in real time integrating advanced techniques of objects tracking in the scene. In this thesis, the objective of the laboratory is to segment and track pedestrians in real urban scenes which induces dealing with highly variable appearance and frequent interactions. A strong scientific problem lies in the ability to learn a relevant visual signature of the persons (that is to say, a discriminating and robust one), to achieve an effective tracking function. To do this, one must dispose at the same time of a large number of descriptions as well as accurate and qualitative data. Today, 3D tracking approaches of multiple agents are based on rigid models of shapes and appearances and make use of advanced modules background segmentation, classification and appearance's modeling. These approaches provide a large number of reliable descriptions. In contrast, many recent approaches, concerning the monitoring of an unknown target from an onboard camera, exploit 2D flexible and deformable targets models based on interests points and involve highly dynamic appearance modeling modules. They seem more relevant to obtain accurate and qualitative data. Finaly, the objective is to bring together the best concepts of these two areas. The essence of the thesis is the exploitation of context and probabilistic reasoning to increase knowledge about the appearance of people and be able to distinguish them at all times.

Study of the behaviour of biomass powders in a gasification process

A possible route to produce second generation biofuels consists in gasifying biomass and then upgrading the gas into a liquid fuel (thermochemical conversion). The reactor technology identified to process biomass gasification is entrained flow reactor. This kind of reactors is already used in coal or coke gasification processes. In such reactors, biomass will be injected as fine particles (size below a few hundreds of µm). Fine particles of biomass are known to agglomerate and thus to lead to fouling issues in storage and/or conveying pieces of equipment. This is one of the major issues to be tackled, for developing an industrial process of biomass gasification. The PhD proposed consists in a study of solutions that can be envisaged to improve the flowability of biomass powders. It includes experimental investigations to characterize the powders, as well as interpretation and modeling developments. The objective of this last part is to lead to a method to predict the behavior of a given powder with the knowledge of some key parameters to identify.

Real-time localization of moving, articulated and / or removable objects

The LVIC laboratory is an expert in the augmented reality filed especially in real-time vision-based localization algorithm. While state of the art solutions are limited to localize a known object using its CAD model only (model-based tracking) or estimate the camera displacement in an unknown environment (SLAM ), the solution developed in our laboratory can merge these two approaches into a single framework called constrained SLAM. This resulting localization in a partially known environment achieves a high level of accuracy and robustness. However, it is based on the assumption that the object of interest is rigid, static and that its geometrical configuration is stable over time, which limits many applications such as industrial Maintenance and repair where geometric stability of the object is not verified over time du to disassembly and reassembly of its different components. The goal of this thesis is to propose innovative solutions for real-time vision-based localization of objects by generalizing our constrained SLAM approach to deal with more complex cases such as moving , articulated and removable objects for which no satisfactory solutions exist.

Improvement of magnetic properties of NdFeB materials through microstructural evolutions related to sintering and thermal treatments

Nd-Fe-B permanent magnets have the highest available magnetic performance. They play a strategic role in improving the efficiency and performance of electric motors for hybrid vehicles and wind turbine generators. These magnets are obtained from powders sintered in the presence of a small quantity of liquid phase. A high remanent field (> 1 Tesla) is obtained after sintering however the resistance to demagnetization (coercivity) is strongly influenced by the microstructure, especially by the morphology, the chemistry and the cristallinity of inter-granular phases. Achieve a high coercivity will require a postsinter annealing. The purpose of the PhD is to understand and improve the coercivity mechanism in NdFeB magnets after sintering and post annealing treatments. Dilatometry and thermal analyses (DSC or ATD) will be used to study the sintering behavior and to determine the optimal temperature range for annealing treatments. Microstructure characterizations of sintered samples will be realized in particular by high-resolution scanning electron microscopy (SEM), as well as by X-ray analysis. The magnetic properties of magnets will be characterized by means of a magnetometer.

Code analysis to detect vulnerabilities in secured components

During SmartCard security evaluation, code analysis is performed to identify vulnerabilities usable by hackers. Today, a "manual" analysis is performed, with human error risks and difficulties to prove the work quality. Some studies, using static analysis techniques, have been led on a limited attacks number. Otherwise, dynamic techniques by code modification when running, have been performed. This PhD will study new static analysis techniques and mix them with the dynamic aspects to optimize vulnerabilities researchs. This work could lead to an automatised code analyser. Questions to approach are : - how to model attacks effects on code execution ? - Which kind of optimizations can be applied during a code dynamic analysis ?

Analysis and dynamic modeling of electrolysis systems powered by renewable energy for electricity storage

While electricity storage is increasingly developing, understanding and modeling the dynamic behavior of water electrolysis systems powered by renewable energy become of great interest both at technological and economical levels. Indeed, connecting an electrolysis system to a fluctuating energy source in one side and a hydrogen storage system to the other side imposes high performances in transient conditions which largely depend on technical choices. These performances in dynamic conditions are linked to strongly interacting phenomena at electric, electrochemical, fluidic and thermal levels. Therefore, the understanding of the dynamic behavior of systems components and their modeling is critical in intermittent conditions. Within the frame of this PhD work, dynamic modeling will be done using Bond Graph formalism which is well adapted to dynamic systems modeling. The experimental validation of developed models will be done at a representative scale at CEA. This approach will allow the comparison of various technological solutions and should contribute to innovative design of electrolysis systems.

New medical imaging systems exploiting the energy dispersive X-ray diffraction with spectrometric CdZnTe based detector.

The context of the thesis is the measurement of energy dispersive X-rays diffraction (EDXRD) using spectrometric semiconductor based detectors for medical applications. Today, most of medical imaging systems exploit the X-ray radiation transmitted through tissues. The recent emergence of energy resolved detectors, (semiconductors based in particular) allows considering the exploitation of another type of radiation: the diffracted radiation (or small angle scattered). Measurement of diffracted spectra can provide new information related to the molecular structure of the examined tissues, thus improving the final diagnosis. The modalities can be the detection and analysis of tumors for breast imaging, of kidney stones, bone densitometry, or even the diffraction measurement of nanoparticles used as contrast agents. Spectrometric CdZnTe based detectors, developed in the LDET laboratory, combine performance (energy resolution, sensitivity, spatial resolution, compactness) necessary to measure relevant diffracted spectra in the medical context under clinical condition. The aim of the thesis is therefore to study the feasibility of such a detection system. It will be a global approach of system conception, combining geometric design, and adaptation of technology and architecture of spectrometric detectors.

Hybrid integration of quantum cascade lasers on Silicon for gaz detection applications

Gas sensor technology based on optical detection has already grown to be indispensable for various aspects in our modern society (chemical emission monitoring, air pollution measuring, high sensitivity trace detection, biological sensing applications). However, available gas sensors are bulky, complex, and have a very high cost of ownership. The consequence is that they are not suitable for mass deployment. The objective of the thesis is to realize a miniaturized and low-cost gas sensor, bringing optical gas spectroscopy towards laboratory-on-a-chip size, by hybrid integration of Quantum Cascade Lasers (QCL) on Silicon photonic integrated circuits(PIC). QCL, based on III-V materials, cover indeed a broad portion of the mid-infrared spectral range (MIR, 3-12 µm), where many chemicals of interest for Safety & Security have their strongest absorption lines. The work will consist in the design and modelling of optical coupling between QCLs and Si PIC, taking into account technological constraints for a viable and suitable process flow to merge-and-match III-V and Si technology.

nterconnection reliability of 3D silicon systems on flexible substrate for medical applications

CEA-LETI develops micro-electronic systems with very high integration: this activity consists in stacking and interconnecting components in order to create electronic circuits in three dimensions. This type of 3D integration opens the door to the development of more efficient integrated systems that are smaller, faster and consume less. Thus, the 3D circuits have real benefits not only for mobile applications but also for medical applications. In the context of medical applications, one of the solutions explored is the stacking of silicon components on flexible substrates. Indeed, this type of technology perfectly suits to the development of medical implants or smart dressing. The thesis work will be achieved in the frame of 3D electronic circuit development on flexible substrates for medical applications. The research will focus on applied electro-mechanical characterization, multi-physics simulation and reliability.

Vehicle made and model recognition

Identification of vehicles is a challenging task in intelligent transport system and automatic surveillance. Many vision-based systems for recognizing vehicles in images sequences have been proposed. The number-plate recognition is a task that has been adequately addressed. However, this identification may turn out insufficient, particularly when the license plate number is occulted. As an alternative to the traditional number-plate recognition system, vehicle manufacturer identification has potentially wider applications as the tracking of vehicles, the census of vehicles by types, the high-end automation and control. The aim of this thesis is to design and develop a system for recognizing a vehicle make and model in a video taken from an arbitrary viewpoint. Concretely the following issues will be investigated: (1) vehicle type classification, (2) vehicle's viewpoint detection and pose estimation, (3) 2D-3D vehicle matching using 3D model projection, (4) 3D model based vehicle recognition, (5) vehicle logo recognition. The framework of this thesis is the multi-class classification, which involves two components: feature representation and classification. The first important part, i.e. appropriate feature descriptions of the vehicle images (intensity image and/or depth image), should characterize the differences between distinct makes and models. The second part is to design accurate and robust classification systems. According to the feature description strategies, the state-of-the-art learning algorithms (SVM, boosting, nearest neighbor?) will be used. The developed methods will be evaluated on a vehicle image database. The system will then be tested in real conditions.

Large scale multimedia retrieval

Multimedia databases become ever larger every day although the tools available to explore these data are not fully satisfactory. Several challenges must be overcome to implement such tools. The proposed thesis will consist to explore (i) knowledge extraction from multimedia documents that, in a context of large-scale databases, will require to determine which steps are needed and how they can be optimized; (ii) the indexing or learning process to organize signatures so that can be efficiently used; a particular challenge consists to combine visual and textual information (iii) efficient data access to data and distributed processing and resources, both to parallelize processing during indexing and establish data structures able of distributing a database across multiple computers (iv) special attention will be paid to the evaluation of work developed, which will be conducted on corpora of significant size, recognized by the scientific and industrial community.

Advanced integration technology for power electronics modules

CEA-LETI is strongly committed to develop and transfer advanced technologies to manufacturing. Among its numerous field of expertises, Advanced Substrates Lab ? CEA is focusing its technological expertise to develop and apply direct bonding at wafer level to enlarge and improve semiconductor packaging toolboxes. On its side G2Elab ? UMR CNRS/G-INP/UJF (Grenoble Electrical Engineering Lab) is searching for discovering and improving the knowledge and the techniques related to the production and management of the electrical energy. Both entities are today engaged toward the development of news packaging techniques of power electronics modules for improving their performances, compactness and cost solution. Based on the development and the use of a new technology based on direct bonding at wafer level applied to 3D integration of vertical power devices, Si/SiC and possibly GaN, a PhD position is offered to do research on innovative power converters modules, for automotive and solar energy applications. The objectives of this PhD position is to develop the complete process integration of an innovating module based on advanced direct bonding and layer transfer technologies of power devices on solid metal wafers. Thermal, electrical and mechanical simulations will be realized to define the best material and circuits configuration. Then, the complete process integration technologies will be realized based on LETI's state-of-the-art semiconductors process line. Finally, after the full integration, characterization will be done and results will be analysed to understand the mechanisms involved in the technology and to assess the overall potentiality of the power modules especially in terms of electromagnetic perturbation.

Surface preparation on III-Nitrides materials

III-V semiconductors are widely used in high-power and high-frequency electronics because of their superior electron velocity with respect to the more common semiconductors silicon. They also have a direct bandgap, making them useful for optoelectronics or solar devices. III-V materials are therefore widely investigated at CEA-LETI in many French or European innovative projects with several industrial partners. These projects use production tools dedicated for microelectronic and Si-based technologies. Regarding cleaning and surface preparation of III-V materials, several chemistries borrowed from Si technology have been investigated and are now adopted for organics, particles or metallic contamination removal. But cleaning of III-V compounds remains a challenge compared to Si-based technologies. Interactions between those substrates and chemical solutions as basic surface phenomena mechanisms need to be understood. This thesis will focus on the study of phosphides or arsenides like InP, GaAs, InGaAs, InGaAsP, or gallium nitride GaN or indium antimonide InSb. The thesis will take place in the environment of the technological platform of CEA-LETI and the PhD student will have access to adequate and up to date characterization tools. Initially, the proposed work aims to evaluate fundamental behaviour of these materials in liquid phase. Surface proprieties will be measured using Zeta potential measurements, XPS, SEM, AFM, XRD, ATR, MIR FTIR, TEM and ellipsometry. Different chemistries will be tested and the performances of the wet chemical cleaning sequence developed will be characterize in terms of metallic or particle removal efficiency using SP1 or SP2 tools and TXRF (Total X-ray Reflexion Fluorescence) or ICP-MS (induced coupled plasma ? mass spectrometry) methods.

Diamond microelectrode arrays for analytical and electrophysiological applications

Boron doped diamond microelectrode systems have been developped for several years by the research team at the Diamond Sensors Laboratory of CEA-LIST for various applications including biosensors and implants. This work involves the optimization of the material synthesized in the laboratory by Chemical Vapor Deposition (CVD) processes along with its electrochemical properties. Considerable efforts have also been put into the development of micro-patterning methods and more generally the fabrication processes. Furthermore, an ultrafast activation process was developed that can be applied directly in complex analytical media. This development offers wide opportunities for on-line monitoring for various applications ranging from biomedical diagnostic to environmental monitoring. The topic of the PhD studentship lies in the continuity of these research activities. The student will focus in particular on the immobilization of functional receptors, either in the form of organic receptors of metal catalysts, toward the development of multisensor systems including biochips or electronic tongues.

Microfluidic system for the encapsulation control of biological living cells

At CEA-LETI, the division of systems for Health and Biology develops microfluidic chips for living cell encapsulation. Main applications of these systems are cell therapy or drug delivery. The first targeted application is Diabetes treatment. Indeed, one way of curing Diabetes is to graft either whole pancreas or cluster of pancreatic cells, called islet of Langerhans. This type of graft is always associated with immunosuppressor treatments. These treatments are not only highly expensive but also deleterious for patient's life expectancy. One way to avoid these treatments is to encapsulate the cell to be grafted into a biocompatible capsule. The role of the capsule is to hide the grafted cells from the immune system and also to preserve their integrity. Pancreas is then successfully replaced and glycemia is regulated. At CEA-LETI, we focus on the development of new technologies in order to improve the efficiency and performance of cell encapsulation. This work is done jointly with clinicians, who are international experts. These new microfluidic devices are based on microtechnologies development. These systems can generate capsules with high size precision and allow the automation of the whole process of encapsulation to produce highly reproducible capsules. Although few teams work on that topic, one point of the system is still not optimum. This is the precise control of number of cells in one capsule. This key point is essential to provide to the clinicians a highly reproducible batch of capsules. The aim of this PhD Thesis is to develop an innovative microfluidic device that will answer to that need. The first year of the PhD thesis will be focalised on the study of a CEA patented device. You will have in charge the evaluation of this first system. You will perform microfluidic experiments, analyse the results and propose new microfluidic design to optimise the performance of such a device. Your approach will join experiments and numerical model of the whole system. After these iterations, you will propose an optimise design and couple it with the currently developed automated system. During this project, you will acquire multiple competencies such as microtechnologies, microfluidic, cell culture, surface chemical modification and various characterization technologies (optical, electronic, mechanical and biological).

Projection of linguistic annotations based on parallel and comparable corpora

This thesis falls within the scope of automating the process of building annotated multilingual linguistic resources. The main purpose in this process is to identify lexical, syntactical and semantic equivalences across parallel or comparable corpora in general or specific fields. The proposed subject aims to explore and experiment new approaches for language-independent projection of annotations by using word alignment and terminology extraction for building multilingual lexicons, cross-language information retrieval for extracting aligned sentences from comparable corpora and deep linguistic analysis for corpora annotation.

optimization and integration of control dedicated to new GaN power devices: self-measurement and joint design with silicon SOI technology

To overcome silicon technology limitation in power applications, CEA-LETI develops new GaN devices: large-gap transistors, faster and more efficient. Optimizing the use of these components requires a control circuit in touch with power device parameters to reacts depending on them. Driver circuit design co-integrated in both GaN and silicon, optimizes this command, according to criteria of performance and reliability. Such optimization is already done disorderly on silicon components with the use of integrated current measurements, such as current sense. Unfortunately the separate development of the power component and driver makes driver design dependent of all parasitic present around the power transistor. Thus, the separated development, as in the majority of applications, does not allow efficient implementation. It is therefore proposed in this thesis to measure critical parameters such as current, terminal voltage, and temperature inside the GaN components, and build a dedicated driver on silicon technology. This thesis proposes a bridge between GaN technology developments and design of SOI integrated driver, and therefore requires both skills on the physical structure of the component, and on integrated circuits design. The target is a complete system capable of self-evaluation to optimize its operation

Lithography by directed self assembly of block copolymers and defectivity understanding

Density multiplication of patterned templates by self-assembly of block copolymers (BCP) stands out as a promising alternative to overcome the resolution limitation of conventional optical lithography. Self-assembling materials used in conjunction with the most advanced exposure tools may enable extension of current manufacturing practices to dimensions of 10 nm and less. A diblock copolymer system consists of two covalently bonded polymer chains. These systems form distinct domain structures due to phase separation of the chemically dissimilar polymers during solidification. This project intends to introduce directed self-assembly of block copolymers, as a low cost alternative solution to actual photo lithography techniques. More precisely process implementation and robustness will be addressed.

Suspended microchannel in hollow MEMS sensors, for real-time counting/weighting of cells and nanoparticles

Using tactile information in robotic command - Application to dexterous manipulation

In the rapidly growing field of robotic, catching and manipulating without breaking it are difficult tasks on which the CEA is working for many years. With this aim, the LIST has developed an artificial hand inspired from human one. This artificial hand has 5 fingers, 24 articular degrees of freedom and an optimal control. On the other hand, using MEMS technologies the LETI has developed 3 axis force sensors which size and sensibility allow to use them as tactile sensors. Moreover, we have demonstrated that they have interesting properties for surface analysis and more particularly for friction coefficient estimation, which is involved in sliding. The association of these two technologies allows intending to handle in an optimal way the manipulation of slippery objects. With this objective, the thesis will first have to bring solutions for relevant feature extraction from the sensors signals. Then, fusion of these parameters with those of the hand and of the object to be manipulated will be considered. Finally the command of the hand will have to be adapted to take into account all the available information. The work will be realized partially in the LETI and partially in the LIST. The integration of the sensors into the hand will be realized. Experiments will be conducted and the results will have to be interpreted. Finally, command algorithms will be developed, and will be validated experimentally.

Conception, fabrication and characterization of low electrical power consumption optical modulator for silicon optical interconnects

Electrical interconnects become a real issue for the increase of the integration density of microelectronic integrated circuits. The most promising solution today is to introduce optical links between two chips or inside a chip to ensure high speed transmission. An optical interconnect (or optical link) comprises a light emitter, an optical modulator to encode the information and a photodetector. CEA-Leti and IEF are among the best research groups worldwide in the field of guiding, modulation and photodetection of light on silicon chips which are technology fully compatible with microelectronic chips. Optical modulators are one of the key components of optical links. They have to meet simultaneously a high modulation efficiency, small insertion losses, high data rate of 40 Gb/s and low electrical consumption. The objective of this PhD is to develop a new generation of capacitive optical modulators on silicon meeting all these performance criteria and in particular with a low electrical consumption, which is a major challenge for applications. This PhD work will be done at CEA-Leti with a PhD supervisor at IEF, a research laboratory with which we collaborate since several years. In a first step, the candidate will work on the electro-optical and photonic design of this new generation of modulator, as well as on the RF design of the driving electrodes. He will also follow the fabrication of a demonstrator in the clean room. Finally, he will characterize the devices on our automated 200 mm probers and assess the modulation performances. He will benefit from the strong complementarity between CEA-Leti and IEF in the field of silicon photonics to carry out his research work aiming at the convergence between photonics and microelectroncics.

Oriented growth and positioning of silicon nanowires on 2D supports

This thesis aims to study the oriented growth of Si nanowires by catalytic CVD on any type of substrate. The idea is for instance, to use a thin layer of ordered mesoporous silica as a template for the growth of nanowires. This solution would have the advantage of being compatible with a wide variety of substrates and with conventional microelectronic technologies. Different approaches will be used to relaize the silica with the catalyst positioned in the porosity. Then, the growth of nanowires will be studied and the different nano-objects synthetized will be characterized.

Alternative methods for epitaxial growth of GaN on Silicon for LED applications

Blue Leds, used for Solid State Lighting, are based on the use of semiconducting materials from the III-Nitride family (GaN and related alloys). In commercial LEDs, the optically active layers are deposited mainly on sapphire substrates, but industrial companies and R&D laboratories are presently assessing silicon as an alternative substrate. Indeed, silicon is less expensive and can be made in very large dimensions (up to 12 inches). Nevertheless, using silicon as a substrate meets two difficulties. The first is related to the high dislocation density resulting from the large lattice parameter difference between nitride materials and silicon, the second being related to their large difference in Coefficient of Thermal Expansion (CTE) which leaves the layer into high tensile stress conditions, leading to crack formation. . To deal with these problems, we want to use silicon substrates on which a mask will be deposited, in the holes of which the growth of GaN will be initiated locally. In order to reach the objectives of dislocation and tensile stress reduction, the thesis work will consist in defining and optimising the mask itself and the growth parameters. As such, the proposed thesis consists in a comprehensive work mainly based on MOVPE epitaxial growth of the GaN layers, complemented by characterization studies to assess the morphological, structural and optical quality of the layers and to follow the process evolution. The proposed work is part of the research carried out in the « Laboratoire des Composants pour Eclairage » (LCE) in LETI, whose objective is to propose innovative solutions for solid stale lighting, and will be done in close collaboratyion with the reserachers and engineers of the lab.The work will also be carried out in close collaboration with the CNRS team from LTM (Laboratoire des Technologies pour le Microélectronique), lab in which the growth process on patterned substrates has been initially developped.

Analytical characterization of distributed cache coherence protocols in a 3D MpSoC architecture

The objective of this thesis is to characterize the performances of different memory consistency protocols alternatives, in the context of distributed multiprocessor systems. Starting from actual 3D architectural realizations, the candidate will develop a hybrid analytical method, combining parametric models for the cache/processor couples, and a more systematic, abstract approach (e.g. queuing theory, Markov chains) for the interactions between processor groups, central memory and the rest of the infrastructure. The significant criteria are transfer latency, throughput and energy consumption. The results will be validated by means of simulation.

Increasing physics engine performances using multi- and many-core hybrid approaches

In response to physical limitations of the increase of modern processors frequency, chips integrating multiple processors have appeared . Along with this evolution of CPUs, the HPC market has seen the emergence of GPUs with hundreds of processing cores. The introduction of new programming models such as CUDA has also allowed their use in various application contexts. The purpose of this thesis is twofold. First develop algorithms for mechanical simulation that can dynamically migrate on the most suitable architecture. Second, it will use adaptive approaches to perform only the calculations required for the user and ensure scalability.

Pressure effect on High Temperature Steam Electrolysis (HTSE) to produce hydrogen and store new intermittent energies

High Temperature Steam Electrolysis (HTSE) is one of the most promising ways to produce massively hydrogen with no carbon dioxide emission and also to store renewable energies. This process with solid oxide cells runs at high temperature (>700°C). The use of the vapor phase and the high temperature conditions allow a significant decrease in the electrical energy consumption for the electrolysis step. The substitution of part of this electrical energy by cheaper thermal energy leads to a better global process output. Additionally, performing HTSE under pressure is of major interest, since on one hand it should allow increasing the cell performance and on the other hand at the system level it avoids a hydrogen compression step that is energetically costly. The CEA aims to demonstrate the role of this pressure increase. This thesis work provides an experimental part to collect steam electrolysis data under high pressure (30 bars), a microstructure part to characterize the electrochemical cells and modeling part.

Perovskite thin films optimisation for microsystems : doping and new electrodes

Thin films of perovskite structure as Pb(Zr,Ti)O3 (PZT) are becoming genuinely important for microsystems: high-K capacitors, piezoelectric devices, electrocaloric systems or pyroelectric sensors. Although they require very elaborated technology, their properties are so interesting that more and more people from both academic and industry study and develop them. One of the main drawbacks of these materials is that bulk still exhibits better properties than thin films. The main ideas to improve these lasts are 1) subtle composition modification by doping and 2) electrodes nature. In this PhD, we propose to perform PZT thin films with different dopants together with uncommon electrodes. The aim is to simplify the technological route by getting rid of the standard platinum electrodes that hinder the spreading of perovskite materials into microsystems. As it has been observed for bulk, doping can help understanding and improving the dielectric constant, the piezoelectric coefficients and the reliability of PZT. To do so, several labs from CEA that have been studying perovskite films for more than 10 years will be involved in this work: DCOS (Grenoble) for films elaboration and simulation, DTSI (Grenoble) and IRAMIS (Saclay) for all advanced characterizations (XPS, XPEEM, SIMS, XRD).

Mitigation Methods of Ionizing Dose Effects on Emerging Deep-Submicron CMOS Technologies

Commercial electronic devices will generally survive for a very short time when subjected to ionizing radiation. That is the main reason for using "hardened" electronic systems, i.e. systems designed to withstand this kind of environment. It has been shown that highly integrated advanced CMOS circuits are more robust to cumulative dose of ionizing radiation. However, the robustness of these components must still be improved to fit the constraints of the civilian nuclear industry. The goal of this thesis is to study and develop effective methods for mitigating the effects of dose on emerging technologies such as the FDSOI CMOS node. Preferred strategies are based on a combination of degradation monitoring and regeneration schemes. As an example, component regeneration may be based on high temperature annealing. The objective is to design a self-healing circuit based on advanced technologies with a high degree of robustness to ionizing dose.

Linear temporal logic and controller synthesis

It has been shown in litterature that some kind of temporal logics can be used both for specifying hazards but also for the automatic synthesis of automatic observers having the ability to detect those hasards in a flow of dated events. Generally the detection is transformed into an alarm signal[1] which is adressed to a supervisor which places the equipment in the more adapted running mode. Rising alarm seems a poor usage of temporal logics. By introducing action features into those logics we could improve the impact of those formalisms in order to design controllers in a simple manner, since, like controllers based on fuzzy logics [2], one can express its control strategy just as if one could act in the place of the controller. The Phd student should: 1) Provide a state of the art concerning rules based controller synthesis 2) Improve temporal logics used in diagnostic in order to bring them action features 3) Build a demonstrator 4) Evaluate the benefits of the approach [1] M. Batteux, P. Dague, N. Rapin et P. Fiani : Diagnosability study of technological systems. Modern Approaches in Applied Intelligence - 24th International Conference on Industrial Engineering and Other Applications of Applied Intelligent Systems, IEA/AIE 2011, Syracuse, NY, USA, June 28 - July 1, 2011 [2] Kevin M. Passino and Stephen Yurkovich, Fuzzy Control, Addison Wesley Longman, Menlo Park, CA, 1998.

Overlay errors modelisation for MAPPER MATRIX multibeam lithography

Performance improvement of integrated circuits is historically built on the size reduction of basic components. Part of the CEA-LETI, the lithography laboratory (LLIT) is at the heart of this ?more Moore? approach by studying low cost lithographic solutions for futures technological nodes. Especially, a pre-alpha electronic multibeam lithographic platform form Mapper Lithography has been recently acquired: this platform will integrate at the end 13000 beams running individually and in parallel. This technology benefits from both high resolution from electronic lithography and high throughput from multibeam parallelism. This tool will also integrate a mark reading system. It will be thus possible to stack and align several technological levels and it opens the possibility to realize advanced electronic devices. The proposed work in this thesis aims to investigate the alignment strategy associated with this technology to meet aggressive production specifications (<7nm for14nm node). In particular, the different sources of misalignment will be quantified and modeled and innovative solutions for overlay corrections will be implemented in order to manufacture an electronic component demonstrator.

Error correction exploiting the properties of emerging digital memories

Storage memory is one of the major components of modern electronic systems (smart phone, digital camera ?). Besides being non-volatile, an ideal storage memory must be fast, dense, power frugal and cheap. Several new non-volatile memories are currently developed (M-RAM, PC-RAM, CB-RAM, Ox-RAM). For some of them, electrical parameters (charge, resistance) used to encode the binary or multi-valued information in storage cells are distributed over a continuous range. This property may degrade memory reliability and yield due to the inherent variability of memories with high integration densities. On the other hand, the mentioned property could provide new opportunities to provide error detection and correction. The goal of this thesis is to propose and develop new solutions to improve the yield and reliability of emerging memories. These solutions may rely on new types of algebraic error correcting codes (ECCs) or on new decoding approaches of classical algebraic ECCs which will allow to beneficiate from the special properties of the emerging memories in order to increase the error correction power without affecting the check-bit number (without increasing the storage overhead) and/or improve the error correction speed. The PhD student will perform his research and development activities within a research group of CEA LIST which is very engaged in the field of integrated circuit reliability and emerging memory technologies. This work can make the object of projects developed in cooperation with industrial and academic partners.

Contribution to Critical System Design through incremental and cooperative safety analysis

The objective of the thesis work is to define a methodology to support incremental design of complex systems submitted to safety regulations. The orientation chosen is to combine safety analysis, and reuse of safety related mitigation patterns within system design in a cooperative way. The system design will be performed using UML/SysML and annotated with safety requirements resulting from a safety analysis, then composition with safety patterns will be performed. Safety analyis will thus provide hints on possible solutions of mitigation patterns corresponding to recommendations from normative standards, then a trade-off analyis will perform multi-criteria optimization in order to select best architectural candidates for implementation.The goal of the work is to propose a methodology, prove the soundness of solutions proposed and develop tools to support the implementation of the methodology. The approach will use formal support providing abstraction/projection operations developed in the laboratory. This work will be integrated within an already rich modeling and design environment of the laboratory (www.eclipse.org.papyrus) and will exploit analysis tools from academic environment such as AltaRica (http://altarica.labri.fr). Key Words : System modeling and analysis, Safety, architectural optimisation, critical systems

Definition and Implementation of a Novel Peroperative Multi-Spectral Imaging Technique

Biomedical optics is a very promising and expanding domain that is already impacting medical diagnostics and surgical procedures. One of the recently developed medical techniques is the measurement of the endogenous contrast in human tissues. The primary objective of this thesis is the development of a, multi-spectral based, novel 2D imaging technique for tissues characterization, focusing on the exploitation of the photons currently used for video color imaging. The research work will derive from the instrumentation and methodologies already existing within the laboratory and will target applications in the fields of open and minimally invasive surgery.

Optimisation and integration of catalytic porous structure into structured reactor for CO2 hydrogenation to methane.

With the current development of renewable energy at the European level such as wind and solar, the electricity storage is becoming a major issue for decades to come. On the other hand, the energy and the industry sectors are high CO2 emitters and these emissions represent a non-exploited source of carbon. A complementary and sustainable route to carbon capture and storage is the CO2 valorisation for energy storage. The efficiency of CO2 hydrogenation into methane with hydrogen produced by electrolysis from renewable energy is a major stake for the ?power to gas? approach with methane storage into the natural gas network. This hydrogenation reaction is highly exothermic, and the catalyst integration into catalytic reactor is of utmost point to ensure high yield of the catalytic process. The efficiency of the catalytic process is linked to the thermal control of the catalyst (Ni based) in the reactive area in order to promote high yield and low catalyst deactivation. The objective of this thesis is to contribute to the development of efficient structured catalytic reactor for the methanation reaction. In structured reactors the catalyst is embedded in open porous structures (eg. honeycomb, foam) inserted inside cm-sized channels through which reactant gases are circulated. Heat transfer is ensured by integrated cooling channels surrounding reactive channels. However, the application of this technology to highly exothermic reactions requires that thermal continuity is ensured between the porous structure and the reactor wall. The work will consist in the development of innovative porous structures enabling a high catalyst load and improved thermal continuity with the reactor walls. The design of the porous structure and channels sizes will be supported by multi-scale modelling (from catalyst particles to reactor channel levels) taken into account reaction kinetic, heat and mass transfer and hydrodynamics. This approach will ensure a control of temperature homogeneity, high space velocity of gas and catalyst productivity. The efficiency of the catalytic process will be validated by experiments on model lab-scale reactor. This thesis will be a work in collaboration between the research centre of (i) VITO in Belgium who has developed innovative porous structure and material catalyst integration by wet processes and the (ii) CEA-Liten who has competences in multi-scale modelling and has developed compact structured reactor. This PhD will be co-funded by the CEA-Liten in France and by VITO in Belgium.

Design, Fabrication and Test of o a 1Tbps-optoelectronic Transceiver for very short reach transmissions

Wavelength Division Multiplexing techniques, and more recently, Multi-level signaling formats have established themselves in optical data transmissions over long distances. Implemented with standalone, big-footprint-components, the techniques are energyvorous, and they can't address yet, shorter reach applications, despite the need for higher transmission rates. CMOS photonics (compatible with the low-cost-fabrication technology of CMOS-electronics) is an emerging technology allowing the integration of optical and analog/digital electronic functions within a unique module, through 3D CMOS processes. Lower cost, lower footprint, and lower power solutions are expected for short reach data-transmissions ranging from rack-to-rack, board-to-board and module-to-module (e.g.: switch). The targeted applications are High Performance Computers, and Data centers. In order to establish itself as the technology of choice for such applications, Silicon Photonics must demonstrate its increased power efficiency as compared with the existing technologies, while showing its ability for higher transmission data rates. Relying on the Silicon photonic components developed in the lab (lasers, (de)multiplexers, photo-detectors, ?), and making use of the electronic drivers developed by our partner at Caltech, the doctoral fellow will propose and develop innovative solutions for the transmission and reception of 1Tbps-data, implementing different multiplexing and modulation techniques. The main activities will consist in - design of transmitters and receivers scalable to 1Tbps, and modeling of the optical transmission - implementation of the solutions (circuit design including layouts according to the Design Rule Manual of the CMOS fabrication) - characterization of the optical transmission implementing the developed transceivers (emitter/receiver)

Impact study of the surface mechanical properties on direct bonding energy

This phd study concerns the surface mechanical property impact on direct bonding energy. For example, the direct bonding energy of a Si/SiO2 structure can be as high as 4J/m² whereas only 30% of the bonded surface is really in contact. As the silicon bulk limit is around 5J/m², this is a paradox. One way to explain this result is to suppose that a part of the debonding energy is consumed inside the SiO2 layer. The crack is then not fragile but ductile. This is not in accordance with literature classical glass crack mechanism. If the crack is not fragile, a certain material volume should be impacted. The direct bonding energy dependency on the SiO2 layer thickness should then reveal this phenomenon. On a more general note, this phd study will look at this impact of the mechanical properties of the bonding layer on the direct bonding energy and mechanism. Indeed, the influence of many surface treatments as the plasma, the CMP, the UV or the ozone could be explain by the mechanical modification of the bonding layer sub-surface.

Algebraic topology and structure of computation

Computing processes, whether they are computer programs or more abstractly rewriting systems, consist in iteratively modifying data according to predefined transformation rules. We are interested in the algebraic structure of the space of possible transformations that they induce, the computation space, in order to better understand those computating processes: which problems can they solve? how can we simplify and optimize them? what are their intrinsic properties (termination, coherence, complexity, etc.)? The goal of this PhD thesis is to apply and adapt tools coming from algebraic topology (Morse reduction theory, homological algebra) in order to extract invariants from computation spaces which will help to provide answer to those questions. On the one hand, the candidate will - extend homological computations known for string rewriting systems (developed by Squier and Kobayashi) to higher-dimensional rewriting systems (rewriting on terms and in 2-categories) in order generalize the results showing the impossibility to decide the word problem, and to obtain lower bounds to "topological obstructions" (minimal number of generators, relations, or critical pairs), - implement homological computations in the framework of higher-dimensional rewriting. On the other hand, those methods will be applied to the study of tasks that can be realized by concurrent processes with failures, starting from the work of Herlihy who showed impossibility results based on topological criteria on the state spaces (modeled as simplicial sets).

Synthesis of organic materials for electrode of Lithium battery

Nowadays lithium batteries use metal based electrode materials. Although their performances are satisfying, they present several important drawbacks. Indeed these compounds are hard to recycle and expensive because they are prepared at very high temperature from rare precursors. Moreover, some metals are toxic and their reactivity leads to safety issues in lithium batteries. As some organic compounds can reversibly react with lithium, it is possible to use them as electrode materials. Interestingly these molecules can be easily prepared using classical organic chemistry techniques from low cost precursors which can even come from biomass. But until now, their use in electrodes is still challenging especially due to their solubility in common electrolytes. The work of this thesis will focus on the synthesis and the grafting of these molecules on organic or inorganic support. Various techniques will be tested in order to prepare electrodes and their electrochemical performances will be determined in lithium batteries.

Model Based Testing of real time distributed systems

Symbolic Transition Systems (STS) are composed of a data part and of a state-transition graph part. They specify behaviours of reactive systems with some benefits compared to the use of classical labelled transition systems. Models are often smaller and it is even possible to finitely denote systems having an infinite number of states. In [GAS06], we have defined a model testing algorithm for particular STS called Input Output Symbolic Transition Systems (IOSTS). The symbolic aspect of IOSTS make possible to exploit a particular analysis technique, the so-called symbolic execution, to define a test selection strategy. This technique has been first defined to compute program executions according to some constraints expressed on input values. The main idea is to use symbols instead of concrete data as input values and to derive a symbolic execution tree in order to describe all possible computations in a symbolic way. We have later adapted this algorithm to take into account timing constraints to test real time systems ([EGL11], [BGLE12]). Real time systems are systems in which the respect of timing constraints on computations is as important as the results of computations. We have therefore improved our IOSTS model to express such timing constraints in so-called Timed Input Output Transition Systems (TIOSTS). TIOSTS are simply timed automata ([AD94]) enriched with mechanisms to handle data in a symbolic manner. We have transposed our algorithm with mechanisms that take TIOSTS as entry, and that permit to test the satisfaction of timing constraints appearing in them. However all those techniques are dedicated to test systems with a centralized interface, that is, a unique interface with which the tester interacts. He/She may then completely order events (inputs sent to the system, output received from the system) occurring at this interface. Consequently interactions with the system can be completely modelled as traces (sequences of inputs and outputs, possibly separated by delays in the timed framework). The goal of this thesis is to extend those existing testing techniques to manage distributed real time systems with several interfaces. Interacting with such systems requires exchanging values with it by means of several interfaces in the same testing process. Different events occurring at different interfaces can not be ordered anymore since it is not possible to compare their instants of occurrences (there is no global clock in a distributed system). Therefore testing techniques may not rely on a simple notion of trace anymore. Automatically testing such real time distributed systems is a first-order challenge because lots of nowadays critical systems are representative of this class of systems (e.g. train or plane control systems, automatic trading systems, embedded systems in cars?.) [GAS06] Christophe Gaston, Pascale Le Gall, Nicolas Rapin, Assia Touil, ?Symbolic execution techniques for test purpose definition?, Testing of Software and Communicating Systems: 18th IFIP TC 6/WG 6.1 International Conference, TestCom 2006. Lecture Notes in Computer Science. [EGL11] José Pablo Escobedo, Christophe Gaston, Pascale Le GALL P. Timed Conformance Testing for Orchestrated Service Discovery. 8th International Symposium on Formal Aspects of Component Software: FACS 2011, September 14-16 2011. Springer [BGLE12] Boutheina Bannour, Christophe Gaston, Pascale Le Gall and José Pablo Escobedo, Off-line test case generation for timed symbolic model-based conformance testing, 24th IFIP Int. Conference on Testing Software and Systems: ICTSS'12, November 19-21 2012. Springer [AD94] Rajeev Alur and David L. Dill. A theory of timed automata. Journal of Theoretical Computer Science, 126(2):183?235, 1994.

Real-time diagnosis of PEMFC performance and durability using impedance spectroscopy

PEMFC are very promising converters of chemical energy into electrical energy for stationary and transport applications. For many years, LITEN works on the development of MEA, stacks and PEMFC systems. This work led to the development of fuel cells based on bipolar plates stamped in perfect adequation with the market for this technology. At the same time, the development and testing of prototypes in actual operating conditions have allowed acquire, for LITEN, an important feedback on these systems and to highlight the impact of fuel cells operating conditions on their performances and their lifetime. It seems to be interesting to monitor in real time, the evolution of different operating parameters to quickly control the command-control and avoid degradations linked with extreme operation conditions. The PhD thesis deals with development of a diagnostic tool in real time based on fuel cell electrochemical impedance. In this work, the development of an electric model based on physical phenomena will be made. This model will then be streamlined in in order to online diagnosis system implementation. In a second step, electrochemical impedance spectroscopy measurements will be performed in nominal operating conditions and in degraded conditions (humidity, pressure and stoichiometry defects). The correlation between model and impedance measurements will be analysed. Finally, the impedancemeter will be implemented on a fuel cell system in order to validate the relevance of laboratory measurements and to evaluate the advantages of the embedded real-time diagnosis hardware for the command-control of fuel cell system.

Organized nanostructures studied by small angle X-ray scattering

The study of nano-objects requires the development of dedicated techniques in order to characterize their form, their size and their spatial organization. Complementary to other microscopy techniques (TEM, SEM, AFM, STM) - the Small Angle X-ray Scattering at Grazing Incidence (GISAXS) known in this context a boom. It makes it possible to obtain a non-destructive, statistical information averaged over the whole sample and allows a meaningful comparison with the physical properties observed at the macroscopic scale. In addition, through modulation of the X-ray penetration into the material with the grazing angle, islands deposited on a substrate as well as nanostructures buried at different depths can be characterized. Finally, the technique can be performed in situ and in real time under many environments. Within the LETI, applications of this technique are numerous: lithography, advanced interconnections ... The thesis will focus on the development of GISAXS in the case of well-organized nanostructures. This will include the developments of tools of modeling/simulation of the of X-ray diffusion by these structures. These developments will be applied to the ?state of the art? grapho-epitaxial copolymers developed in LETI and which have already been studied in-situ by GISAXS at the ESRF (European Synchrotron Radiation Facility - Grenoble). The objective is to better understand the ordering of these materials. Finally, the applicants could combine the GISAXS results with other characterization tools (CD-AFM, CD-SEM ...) already performed on these copolymers. Experiments will take place both at the ESRF and in the LETI laboratory which will benefit from a new SAXS/GISAXS instrument in end 2013. This topic is the opportunity to work in the environment of a European synchrotron and the most advanced materials developed MINATEC for micro-nano technologies

Manufacturing of a SiGe nanowires based thermoelectric generator

Thermoelectrics is based on physical phenomena allowing the conversion of thermal energy in electrical energy, and reciprocally. Thermoelectrics can be applied to some applications as thermoelectric cooling, energy recovery or sensors. The use of nanostructured materials for thermoelectrics is a key step to increase performances. Indeed, the structuring at nanoscale allows developing new materials with new chemical and physical properties, radically different from those of conventional bulk materials. Thus, the state of the art has recently shown that nanowires can have very high ZT (factor calibrating the TE performances of materials). The goal of this work is to develop TE devices for energy harvesting based on SiGe nanowires. It is a real technological challenge because the manipulation of nanowires for their characterization or their integration into a device is very delicate. Thus specific characterization tools for the measure of TE properties and specific technological process for the manufacturing of TE devices will be developed. In parallel, to optimize the performances of these devices, the understanding of electrical and thermal transport mechanims in such materials will be studied. This work will allow the candidate to achieve a strong experience in materials growth and their characterization. The manufacturing of high sensitive measure tools will also allow to achieve excpetionnal experience in fine electrical and thermal measurement.

Energy harvesting in thermoelectric systems optimized by µfluidic

Thermoelectrics (TE) is based on physical phenomena allowing the conversion of thermal energy in electrical energy, and reciprocally. Thermoelectrics can be applied to some applications as thermoelectric cooling, energy recovery or sensors. At very low scale, for example in microelectronics, dimensions are so low that it is very difficult to maintain a temeprature difference at the TE system extremities. Thus, the innovating association of TE device with cooling µchannels could heat to be evacuated, and so to have a high temeprature difference and so to harvest energy. Power between some hundreds µW and some mW can be obtained. The goal of this work is to develop µsystems combining TE and µchannels. Two main fiels will be studied : the manufacturing of a technological process, based on system modeling, and the characterization of such systems. This work will be done in the frame work of a collaboration between CEA-Liten, specialized in TE µsystems, and the University of Sherbrooke, specialized in µfluidics, et will be done in the two sites. Thus the candidate will be able to achieve a strong and varied experience in different fields, such as energu harvesting, thermoelectrics, microelectronics and microfluidics.

Co-design of image sensor architectures and advanced optical functions for novel acquisition of images

The micro-mirrors in video projectors, microlenses of plenoptic cameras, or Spatial Light Modulators (SLM) are some examples of the micro and nanotechnologies contribution to novel optical functions, which constitute breakthrough with respect to the conventional optics. In addition to new features, these processes often allow a low-cost, yet extremely accurate manufacturing, making them compatible with consumer applications and/or with high volume production. However, the acquisition of the images or the light signals is realized with standard image sensors, which limit the fields of applications. Based on Léti's works in smart CMOS image sensors, the objective of the thesis consists in proposing original images sensors' architecture, as well as novel processing, to be integrated in advanced CMOS technology. This combination will allow the emergence of novel functions and breakthrough in terms of performances.

Organosilicate materials like SiOCH are an original way for the realization of light sources for applications in the field of lighting or Si photonics. These materials present the advantage of being harmless because these compounds are based on chemical elements such as silicon, carbon, oxygen or nitrogen present in large quantities on earth. PECVD SiOCH or SiOCNH thin films developed at CEA-LETI are new and original materials with light emission properties still poorly understood. A deep investigation of their optical properties as a function of elaboration parameters and their structural/physico-chemical properties is essential to consider their future use as light sources for applications. The objective of this thesis is to study their fundamental properties regarding their potential applications, in particular the emission of light in the visible range or the realization of integrated silicon-based light sources or microlasers emitting at the telecom wavelength of 1.5 microns by Er doping.

Energy harvesting device on ultrathin substrate for biomedical applications

The PhD research will deal with thin-film photovoltaic cells on ultrathin and mechanically flexible substrates, for powering energy-autonomous miniature devices, in particular for biomedical applications. The ultrathin, flexible and robust photovoltaic cells will allow new integration concepts, and will be able to supply power for wearable or even implantable medical devices. The objective of the PhD research will be two-fold. First, the PhD student will focus on novel types of flexible substrates, namely ultrathin glass or ceramic foils. Such substrates present significant advantages compared to the more conventional metal or polymer foils. An important part of the study will focus on the properties of the back electrical contact deposited on ultrathin glass or ceramic foils. Several scientific and technological challenges have to be addressed, in particular controlling mechanical stress, microstructure, chemical composition and electrical properties of the back electrode. Second, the PhD student will explore a new crystallization annealing technique for photovoltaic thin film absorber synthesis. Compared to conventional processes, this new technique leads to a decrease of the number of process steps and a reduction of raw material consumption, hence a significantly lower manufacturing cost. A crucial aspect of the research will be to study the physical mechanisms related to reactive gas control in a prototype equipment which is available at CEA LITEN.

Hybrid memory hierarchy and dynamic data handling for embedded multicore architectures

Due to technological limitations on CMOS operating frequency, the race for ever increasing processing performance in embedded systems today translates to a rapid increase of the number of cores per device. But the over-proportional raise of power consumption is increasingly hindering this performance race. Regarding the processing cores, recent proposals target to limit the average power consumption of the system by adding smaller cores specifically optimized for low power and dynamically activate the most appropriate ones depending on the system load. The most known concept of this type is the ?big.LITTLE? proposal from ARM. It is admitted that a large part of the total power consumption is coming from the ?memory hierarchy? needed to supply data to the processors. Today's memory configurations are mainly optimized for compute performance and dimensioned to handle the most complex access types. Several studies in literature are addressing optimized usage of the existing memory resources [1], [2]. The work of this PhD thesis will focus on read-only data and propose a power consumption optimized memory hierarchy using specific extensions to manage and store this type of data. A coherent implementation approach will be elaborated, covering all steps from compilation down to the dynamic handling of data allocation, to enable an easy real-life implementation of this optimization. [1] A.Badawy, A.Aggarwal, D.Yeung and C.Tseng, "The Efficacy of Software Prefetching and Locality Optimizations on Future Memory Systems", J. Instruction-Level Parallelism, 2004 [2] Pfrimmer, J.; Li, K.F.; Rakhmatov, D., "Balancing scratchpad and cache in embedded systems for power and speed performance", IEEE-NEWCAS Conference, 2005. The 3rd International, vol., no., pp. 107- 110, 19-22 June 2005

Automated abstraction of models for symbolic analysis

Formal analysis of big complex systems in order to simulate them, to prove properties and to make automated test generation, is generally confronted to the combinatorial explosion of calculus. The use of symbolic calculus reduces efficiently this explosion but the state of the art limits the results. The symbolic approach must be often completed by an abstraction process on the models in order to increase the chances of reaching the complete analysis of big systems. The goal of the thesis is to identify these existing techniques of abstraction (abstract interpretation, qualitative simulation, etc.) and then to integrate them into a methodology and an associated tool, which will be able to increase the performance of the symbolic tool of the LISE laboratory of CEA LIST. The use of these techniques must be done in order to allow the users to exploit the results without knowledge on the mathematical techniques which are necessary. The benefit must be qualitative (exhibition of the behaviors, proof of properties, impossible to process without these abstractions) and quantitative (less time necessary to process the analysis).

Channel emulation and "Over-The-Air" test for sub-GHz frequencies and vehicular applications

The performance evaluation of future wireless systems requires a full assessment of the communication chain, taking into account the characteristics of the propagation channel for different conditions of mobility depending on the application (e.g. vehicular, cellular, MIMO ...) and referred standards (e.g. LTE, 3GPP, IEEE 802.x..). Radiofrequency channel can reproduce the propagation characteristics in a controlled and repeatable way, avoiding expensive on-site measurement campaigns. In particular, the so-called "Over-The-Air" (OTA) tests reproduce the radiated conditions of a multipath channel in a controlled environment, like an anechoic chamber. The objective of the PhD thesis is to investigate, model, and implement channel emulation and OTA test methodologies at frequencies below 1 GHz, for different types of applications (e.g. vehicular). The work includes theoretical studies related to electromagnetic waves propagation in different contexts, as well as experimental implementation and validation of the methods developed by the PhD student.

ECoG signal processing for Brain Computer Interface with multiple degrees of freedom for clinical applications

The thesis will be carried out within the frame of the interdisciplinary project ?Brain Computer Interface? (BCI) at CEA/LETI/CLINATEC® (Grenoble, France). The objective of the project is to improve the quality of life of individuals with severe motor disability thanks BCI systems. The BCIs will be based on the measurement and processing of neuronal activity from the cerebral cortex of subject (Electrocorticogram). The particular goal of the thesis will be development of BCI signal processing block including signal preprocessing, feature extraction, and modeling to predict state of system. The prediction is used to send the command to an effector (fragments of exoskeleton). For stable and robust functioning in natural environment different control strategies will be studded. Sequences of commands will be optimized balancing efficiency and robustness of control. To achieve a large number of degrees of freedom and in order to pilot effectors such as exoskeleton the algorithms of real time calculation and feature selection will be improved. Finally adaptive calibration during the control session will be introduced. Algorithms will be implemented on MATLAB and C/C++ and integrated to the ?BCI-CLINATEC? software platform. Finally, clinical study will allow improving the algorithms.

Programmability for the many-core area, beyond stream-programing

A new area of electronic progress is opening with an updated Moore's Law: now, the number of available processing elements on a chip will double every two years or so. Hence, before the end of the decade, Many-core systems offering thousands of processors will be widely available. But the programing concept currently used like OpenMP will not scale up well. New programing languages are arising that can fit the requirements of parallelism, determinism and data reuse of these systems: stream programing. Nonetheless, these programing languages are often too static to fit the programing needs of emerging application like cognitive radios, augmented reality, autonomous vehicles, etc. Such applications require a context-dependent dynamic adaptation. This is a new challenge. In the embedded-HPC team of the LASTRE laboratory (Laboratory of real-time and embedded systems) of DRT/DACLE, we developped several stream languages, including compilers, runtime generators and execution support that would be a good basis for the candidate's research work. From that new ways can be opened for the languages that could fit the challenges of the manycore area. In this context, the PhD Candidate will explore novel ways for enhancing the programmability of many-core computing systems based from stream languages developped at the lab.

Definition and implementation of protocols intended for the energy evaluation of buildings and for the information of their users

The complexity of buildings does not allow to totally control the knowledge of the level of consumption based only on numerical simulations. On the other hand, understanding building behavior only from measurements would imply a huge investment to deploy enough probes to get valuable information. A compromise solution consists in coupling the real time simulation with a metrology limited to a small number of sensors. The work will consist in defining scenarios of building thermal solicitations in order to experimentally determine the time shift or the damping effect on temperature for an existing building. This approach will then be tested within the framework of an international competition called the Solar Decathlon Europe which will consist in comparing the energy consumption in real time on 20 houses coming from all over the world.At the end we will have an objective and inexpensive evaluation method that could be implemented in a more general context to evaluate the performance of new or retrofitted buildings.

Photonics: a novel approach to hyperspectral detection.

We are looking for talented students in the fields of optics and optical engineering willing to join the optoelectronics insitute (LETI) of the CEA (the Atomic Energy and Alternative Energy Commission, Commissariat à l'énergie atomique et aux énergies alternatives) to participate into its plasmonics and nanophotonics effort. In recent years, combinations of metallic and dielectric structures have been used to sort photons according to their wavelengths, build high performance filters and enhance the absorption of photons into detector materials. We exploit these capabilities in a variety of high performance infrared imaging devices. Work is underway to understand the fundamentals and produce new detector designs, in close collaboration with industry. Our Institue has state of the art optical simulation and device fabrication tools. The student will take part to the conception and realization of devices and will help to develop new mathematical tools to further our fundamental understanding of this exciting new field.

Methodological Framework for Dealing with Model Viewpoints

One of the primary purposes of models in engineering is to support communications between various stakeholders in a system. Well-constructed engineering models can greatly increase the likelihood that there is a common understanding of both requirements and corresponding solutions. This is because a model, by definition, reduces the amount of information that needs to be absorbed, while emphasizing the aspects that are of concern. Of course, different stakeholders will have different concerns. These are invariably tied to how the system in question relates to their specific needs. (e.g., designers with its internal structure and workings and regulatory agencies with its conformance to standards and regulations). The worldviews, or viewpoints [3], of these varied concerns can be radically different from each other. Hence, to maximize the communication value of a model, it is necessary not only to extract and incorporate into the model stakeholder-relevant information but also to present it in a form that most readily understood by that stakeholder. On the other hand, it is well known that duplicating information can cause serious problems such as difficulties in maintaining data consistency (which can be a source of other even more serious problems) and duplication of effort. Modern model-based engineering methods have the potential to help us cope with these problems through various automatic or semi-automatic techniques, such as automated model filtering (queries), model merging, and the use of custom model transformations. The purpose of this PhD work will then to exploit the full potential of those aforementioned approaches and technologies which are still largely underutilized: there are no clearly defined and agreed on methods for how to exploit these techniques to deal with the problem of capturing, representing and using viewpoints.

Physical layer security for cloud-assisted wireless networks

Over the past few years, Cloud computing has emerged as a new paradigm in distributed/mobile computing environments by providing access to on-demand services and resources to the clients. Cloud storage services appear as an inexpensive and scalable solution for storing large amounts of data and making it pervasively available to users. Today, the concept of cloud is moving closer to users by considering cloud over mobile distributed nodes and devices with some computing and storage capabilities, offering thus the same cloud services to users, nearby. The users access then to these online distributed resources through the available wireless communication links, to store their data or to offload their computation. In the absence of an infrastructure controlling the distributed nodes, the security issues for distributed communication and storage become more crucial in the wireless environment. In this cloud-assisted wireless networks context, this PHD addresses the physical layer security of distributed storage by exploiting the randomness of the wireless channels, and taking advantage of the different resources available at nearby distributed nodes. The aim is to investigate practical secure coding schemes for distributed communication and storage resilient against eavesdropping attacks. In such attacks, the eavesdropper can read or modify actively the data stored in a subset of storage nodes or it can passively access the data downloaded during repair/replacement of node failures. According to the scenarios, the designed coding schemes will consider LDPC (Low Density Parity Check) codes based graph structures or nested lattice codes that can resist to eavesdropping and showed to be robust in wiretap networks.

Interface optimization of PIN OLED architectures

The purpose of the PhD is to develop an electric model taking into account the effects of interface in a multilayer white OLED typical type PIN. Work will be carried out on electrical simulation software available at the CEA. Work characterizations of interfaces will be performed at IMS in Bordeaux on simplified stacks. Full OLED stacks will be made at CEA to validate the models. CEA tasks to understand the structure and behavior of interfaces in the OLED architecture will include: - Study of the behavior of PIN OLED structures in high temperature environment - Study of the behavior of PIN OLED structures under IVL pulsed characterization: estimation of response time, measurement of charge carrier mobility - Study of the molecular orientation of small molecules evaporated in order to improve the charge transport, and the light extraction efficiency.

High frequency nano electro mechanical systems (NEMS HF); a breakthrough in infrared imaging technology

Uncooled microbolometer infrared focal plane arrays (U-IRFPA) have gained a strategic position in the infrared imaging business due to a progressive shift from the single military application market to a dual market, mainly driven today by some new low-cost/high-volume applications (thermography, automotive for night vision?). However, these newly appeared commercial demands require significant improvements and a dramatic cost reduction of the technology. One way of addressing those issues is to scale down the pixel size, another way is to rethink the design of the pixel itself. The last is the framework of this PhD which overall objective is to develop an innovative high frequency thermal nano resonator, sensitive to IR and piezoelectrically actuated. We expect from this new architecture a real performance breakthrough for scaled pixel in comparison with the current amorphous silicon based technology. The candidate work can be basically divided into three successive tasks. The first task is focused on design in order to obtain the most suited NEMS device for the microbolometer function (bibliography, preliminary measurements on existing devices, multi physics modelling). The second task is devoted to technology developments; a fully functional demonstrator will be processed on the CEA-LETI 200mm silicon line. The candidate is responsible for the technology definition (process and route), the device CAD, and the process supervision. The integration of a piezoelectric film (made from aluminium nitride ? AlN ? as an example) embedded in an IR self supporting membrane, could be a major and original contribution of this PhD to high frequency NEMS achievement. Finally, the third task is to carry out electrical tests by measuring the demonstrator performances (frequency shift versus temperature, frequency stability, thermal, optical and electro-optical measurements). This study will be conducted in-between two different CEA-LETI research labs, respectively expert in infrared imaging technology and physics of micro & nano electro mechanical systems (NEMS). At the end of this three years study, the candidate will have performed a complete validation of a new NEMS based microbolometer architecture for addressing the next UIRFPA generation.

Topological insulator manufacturing, application to spintronics

Topological insulators are a new class of materials. They are insulators in the bulk but present surface electronic conduction by Dirac fermions with topologically protected wavefunctions. These particles attract lots of interest because of their extraordinary properties and new quantum physics effects are expected, as well as new spintronics devices utilizing the spontaneous spin polarization of these systems. Strained HgTe/CdTe epitaxial structures are very good candidates with a very convincing demonstration of a 3D topological insulator by CEA-Leti and Néel Institute last year. The present thesis objectives are to realize these topological insulators and to utilize their new properties in novel spintronic devices. The success in this approach depends on the mastering, optimization and control of the epitaxial and processing steps. Particular attention will be paid to the control of surfaces and interfaces properties. CEA-Leti molecular beam epitaxy systems associated with nanocharacterization will be used for material optimization. Devices design and realization will be carried out using the extensive processing capabilities of both DOPT/STM and CNRS-PTA technological platforms. The devices and associated spin currents measurements will be fully characterized in dilution cryostats at the Néel Institut in Grenoble.

Active Antenna design for future compact and high efficient RF front-end

Since several years, CEA-Leti is working on ?smart' antennas by adding new functions which will allow power efficiency and miniaturisation to the future wireless applications. In this context, new integration schemes supporting reconfigurable components will be investigated. The objectives of the thesis will be to study new architectures for the RF front-end and antenna strategies in order to propose miniature transmitter modules with low loss. The proposed solutions will have to take benefit of a Power Amplifier-Antenna co-design approach to avoid the standard common 50 ohms reference impedance between them. The antenna development will take into account the constraints of Leti power amplifier team. The candidate should be highly motivated and self-working, and be familiar with antenna and RF standard architecture. He (she) should also be familiar with EM simulator and Matlab.

Prostethic Synapses based on Resistive Memory Technologies

In neurobiology there is a strong need of advanced neuro-prostethic systems able to communicate with the brain cortex, serving both to map and locally stimulate the neuronal activity. These will enable researchers to take samples in real time, giving them a precise picture of the neuronal activities during certain processes such as Parkinson's disease. Resistive Random Access Memory devices by default become a very elementary or simplistic electrical model of the biological synapse for the following reasons: (a) Two ? Terminal, Nanoscale ; (b) Conductance/Resistance Modulation (i.e. can be programmed with electrical pulses analogous to neuron action potential) ; (c) changes (dynamic) and stores (non-volatile) simultaneously. The main idea of this Master stage will be thus to develop a specialized neural probe with intelligent RRAM array, with the following functionalities: 3D spatial mapping of neuron activity through different layers of cortex/brain tissue, offsite storage of synaptic weights/patterns in response to a stimuli, possibility of by-passing a real synapse with an artificial RRAM synapse. The object of the stage will be to make a proof of concept of the prostetic synapses by integration of RRAM neuron/synapses in NeuroPXI, the new real-time data acquisition system for neurosciences studies from the Health Department from CEA-LETI. This brain-interface platform already well-established will allow to test the RRAM demonstrator in-vitro and in-vivo (in collaboration with CLINATEC lab).

Performance modeling of PEMFC coupling discrete and continuous approaches.

The objective of this thesis is to develop a new generation of more realistic performance model of PEM Fuel Cells based on Pore Network Modeling (PNM) that allow 1) to have a better determination of fluid transport (gas, vapor, liquid) in the MEA (Membrane Electrode Assembly) which is crucial for the performance and durability 2) to better link the structures of the components to their performance. This work will be based on the original PNM models developed in the laboratory for the study of transfers in the diffusion layers (gas and liquid) and in the active layers (gas, liquid, heat, charge and electrochemistry), by direct coupling or by homogenization. An experimental validation will be done on tests performed in parallel. Improvements of components (pore size distribution, wettability, material compounds distribution...) will be discussed and proposed. This thesis will be done in close collaboration between the CEA / LCPEM recognized as a specialist in component development and modeling of Fuel Cells, and CNRS/IMFT recognized as a specialist in the development of the approach pore networks in porous media.

Study and Design of a data memory scheme for embedded vision systems

The embedded vision systems are growing rapidly due to evolution of microelectronics technologies and are used in many fields (robotics, automation, personal assistance, industrial control, driving assistance, surveillance). These evolutions allow the miniaturization of vision systems and provide the opportunity of major changes in their architecture. The purpose of this thesis is to revisit the architecture and organization of memory in embedded vision systems to facilitate access to both images and data structures specific to image processing, with the aim of increasing computing power and efficiency of these systems. Multiple aspects will guide this study: first analysis of applications of image processing and vision (wavelet transform, multi-resolution processing, recognition, indexing) will be performed to extract the data structures involved and the cost in performance and power consumption of data access and this considering the constraints of the various types of vision systems such as SiP, SoC or 3D integrated systems. Based on this analysis and according to specific needs of vision systems and the capabilities of considered integration technologies, an innovative and efficient memory organization will be proposed.

Solutions for machine learning and decision on autonomous platform in a real-time context

The human brain can realize high-level tasks, such as pattern recognition, in real-time, with a low energy consumption and a non-extensible volume. Nevertheless, even supposedly powerful hardware architectures cannot fully address these problems that remain true technological challenges, despite an extensive literature. Therefore, there is much room for improvement that we aim to explore on so-called embedded systems, which do not have the extended computing power neither the quasi-infinite memory capacity of a CPU-GPU platform, connected to online data-bases. The thesis works will aim to revisit existing methods through approximate computing. We will try to formalize this computing technics and to use them within decision and machine learning applications, on an autonomous platform. These works will also try to take advantage of a smart encoding of the knowledge in order to allow the system to learn ?online? and to keep its adaptation ability, even when its whole memory will be used. Eventually, the chosen applications will have to illustrate the possibility to adapt the QoS (Quality of Service), with respect to its context, letting the system switch dynamically from a precise computing state to an approximate one (and reciprocally), when temporal constraints strengthen.

Automatic pedestians detection by active sensors

Although the automatic person recognition improved a lot, the problem is still open for embedded systems working in complex environments. In order to improve recognition rates an approach was recently proposed which is based on the fusion of 3D and appearance (i.e. image) informations. These 3D informations are usually obtained by stereoscopic cameras. The goal of this PHD is to use a 3D active sensor like Microsoft's Kinect in order to obtain 3D informations in conditions leading passive sensors to fail. The work will begin by calibrating and synchronising the different data sources. The candidate will then be able to work on the fusion method and on a learning algorithm for joint 3D and image data taking into account all the sensor's specificities. The second part of this PHD will consist in catching up the weaknesses of this mass market system, which was designed for domestic applications, in order to create an embedded system for industrial applications, i.e. in diverse lightning conditions and with higher distances.

Design of state observors for the command of a fuel cell system for transport application in order to optimize performances and durability

The PEMFC fuel cell (Proton Exchange Membrane Fuel Cell) for transport applications, needs a complex system to feed and evacuate the reactive gases (oxygen and hydrogen), to evacuate the heat and to manage the electric power. Many auxiliaries are used (compressor, valves, pump, heat exchanger and converter) to dynamically control the inlet and outlet conditions of the fuel cell. However, it is the fuel cell local conditions that determine the fuel cell performance and durability. In order to have a complete optimal command, it is necessary to develop state observers to estimate the fuel cell local conditions from measurement of the global parameters. This command must provide an optimal and stabilize operation (fluidic, thermal, water management) and minimize the fuel cell ageing process. The thesis work includes an important modeling part with dedicated experiments to characterize specific parts of the system of the fuel cell, but also a validation part on a real fuel cell system. SOH (State of Health) indicators will be generated to take into account the durability constraints in the command.

High temperature steam electrolysis: analysis of the cell degradation by a multi-scale and multi-physic approach

Because of its great potential, hydrogen production by high temperature steam electrolysis has received an increasing national and international interest in recent years. The hydrogen produced by water electrolysis if taking advantage of sustainable energy sources would constitute an energy carriers with low carbon footprint and would allow limit greenhouse gas emission. However, this scheme will be relevant only if the durability and the reliability of the electrolyser is improved in order to reduce the cost of the produced hydrogen. In the thesis, it is proposed to analyse the mechanisms responsible of the electrolyser cells' degradation. Firstly, some durability tests will be carried out on elementary cells to measure their efficiency decrease over time. The degradation mechanisms will be then analysed by coupling a multi-physic modelling approach associated to a fine characterisation of the electrodes material properties. A special attention will be given to investigate the morphological and physical destabilisation of the electrodes by using X-ray adsorption methods at the European synchrotron Radiation Facility in Grenoble (ESRF). This work aims at proposing solutions in terms of microstructure optimisation and strategy of the electrolyser operation to limit the degradation.

Micrometer scale lead free solder connexions for advanced 3D integration : metallurgical mechanisms study and optimization

3D integration, technology which consists in stacking vertically several chips of various technologies requests the soldering of these chips using lead free materials and more specifically Tin/Silver. The soldering metallurgy of such materials on interconnect metals (Copper, Nickel) has recently been widely studied and is well known for macroscopic scales. However, the increasing in the connexion density between different chips or between chip and carrier (board, Interposer) leads to smaller and smaller connexions request. For these connexions, Only few studies regarding the size scale effect on the formation of the intermetallic compound such as NixSny or CuxSny as well as their impact on solder reliability have been carried out. These materials are generally deposited through consecutive electroplating processes through resist mask and process conditions optimization allows the modification of each material and total stack characteristics thus leading to optimum behaviour conditions.

Adiabatic piezoelectric DC/DC converter

The aim of this PhD is to design a piezoelectric structure able to translate an input electrical energy on a first voltage into a mechanical energy and then translate that mechanical energy in an electrical energy on a second voltage. The energetic cycle applied on the piezoelectric structure will be similar to a thermo dynamical cycle by replacing the pressure/temperature by an electrical charge/voltage on a piezoelectric device. A significant part of the work will be to study the energetic cycle to apply on the piezoelectric structure assuming that the energetics transfers have to be done adiabatically (i.e. without losses). An electronic power management will be then designed to applied this ideal energetic cycle, that electronic will enclose several regulation loops to ensure the system stability and regulate the electrical output power converted.

Understanding of the mechanisms of thermoelectric assemblies interfaces

In the current context of energy savings, the CEA-LITEN works on the research and development of components using thermoelectric materials for energy harvesting. The principle is to convert a part of a heat flow crossing a component in electric power using the Seebeck effect. A component is a series of pairs of thermoelectric material legs connected by metallic tracks, the quite clamped between two electrically isolated patches working as thermal heat exchanger. The subject of this thesis is the fine characterization of the interfaces between the thermoelectric material and the electric tracks to accurately measure the physical phenomena acting there, to model them and to propose technology solutions to decrease their contact resistances which penalize the efficiency of the heat flow conversion into electricity. Several materials for legs and tracks will be to study. The purpose of the thesis is, for each leg-track couple, to find a solution of specific assembly and to validate it thanks to these measurements of the interface quality. This thesis will allow the candidate to acquire an expertise on a wide range of measurement techniques including several fields of the physics: thermics, electricity, chemistry and physics.

Postural stability for active lower-limbs exoskeletons

Active robotics exoskeletons are developed for different uses in the military, civil and medical field. Although having very different function, the postural stability is crucial during walking, standing and changing positions. The PhD will address this problem by developing control schemes and algorithm improving the postural stability and preventing fall of persons wearing an active exoskeleton. It will be important for some application to preserve the ability of the person to move and to react. Of particular interest for this research are the compliant control methods developed for bipedal robots, as they are more robust and based on force control. The Phd research will include an important experimental work on an active exoskeleton developed in the Interactive Robotics Laboratory. It will use also real-time physical simulation with a software developed by a companion laboratory.

Developement of new technology of flexible array probes for the detection of flaws in ferromagnetic pieces

The CEA-LIST develops high performance electromagnetic probes for the non destructive testing (NDT). These probes are arrays of microcoils on flexible foild generating magnetic fiels and inducing eddy currents in the inspected part. Receivers measure the perturbations of the field due to tje persence of flaws. The applications are numerous in different industrial sectors such as the aeronautics, steel industry and the nuclear energy. This thesis will consist in conceiving new arrays probes adapted to the inspection of ferromagnetic materials and based on coils or magneto resistors sensors (GMR, GMI, ?). The PhD student will get involved in the design of such probes, more efficient than present commercial probes. Steps of design by simulations, optimisations and experimental validations will be carrying out. Algorithms of signal and images processing will be implemented in order to enhance the detection of flaws in the complex structures.

Multi-level modeling for hybrid energetic systems including renewable generation: Combining electric and thermal modeling for energy management

The large scale integration of renewable energy generation into power systems may be challenging due to the variability and limited predictability of the resource. In order to reduce such impacts, renewable sources may be combined into hybrid power systems, which include controllable thermal and electric units. The main objective of the present work is to develop multi-energy models for the energy management of such hybrid systems. The first step consists in combining thermal and electric models for the energy management of hybrid systems, such as a building including photovoltaic generation. These energetic models are used to compute the system optimal control, which defines optimal setpoints for the controllable units, from forecasts of the renewable generation and of the consumption. The second step consists in combining the given energetic models with physical models for the simulation of the system operation. These models will inform about the component operation simulation based on the setpoints which are given by the energy management models. Also, these physical models will be used to validate the energy management models and algorithms in order to deploy them for real-time operation.

Domain decomposition and 'hp' discretization methods for the electromagnetic integral equations applied to non destructive testing

Non-destructive testing (NDT) gathers various industrial processes to detect and characterize defects in a piece without altering it. The modeling of this type of control is increasingly used as it is a valuable help to understand the physical phenomena, to design new sensors, to optimize control procedures and to demonstrate their performance. The Department of Imaging, Simulation and Control (DISC) of the CEA LIST, which is a major player in the community of the NDT at the international level, developed for several years the simulation platform CIVA dedicated to NDT. This thesis concerns the study of a specific numerical approach dedicated to the simulation of non destructive electromagnetic testing. This approach is based on a surface integral representation of electromagnetic fields. Its main innovation lies in the introduction of a non-conventionnal approximation space, to work on non-conforming meshes that are more adapted to complex configurations and more prone to a future integration into CIVA.

Characterization and Quantification of Membrane/Electrode Assembly Degradation in Proton Exchange Membrane Fuel Cell

The proton exchange membrane fuel cell (PEMFC) is one of the most promising candidates as zero emission alternative power sources for transport and stationary applications. While their performance has been greatly improved in recent years, their cost and lifetime remain the two major issues for a wide commercialization. To achieve this objective, it is necessary to optimize the membrane-electrode assembly (MEA) that constitutes the core of the fuel cell and consequently to study in details their degradation mechanisms. The MEA consists of a diffusion layers and a membrane coating with two active layers where the electrocatalyst reactions take place. These active layers exhibit a multiscale structure of aggregated carbon particles supporting platinum nanoparticles, the catalyst, and covered by ionomer. During the fuel cell operation, these components are degradated. The aim of this PhD work is to analyze, understand and quantify the MEA degradation. To carry out this study, the PhD student will mainly use the state-of-the-art electron microscopes available on the Nano-characterization platform at CEA-Grenoble (PFNC) as well as electrochemical characterization techniques.

Biooil production from food processing residues

The aim of this PhD thesis is to study and develop a liquefaction process of food processing and agricultural residues to produce a bio-oil. This bio-oil will be further be transformed into a biofuel. Hydrothermal liquefaction is done with water between 250 and 350°C, 7-20 MPa with or without a catalyst. In these conditions, the water solubility of sels and hydrocarbons but also the reactivities are modified. The bio-oil produced has a higher heating value (30-36 MJ/Kg) with a low oxygen content (10-20%). This process is well suited to convert biomasses with a high water content. Water is a solvent but takes also part in the reactions. This process requires wet reactants and the biomass resources do not need drying. The process is known for some wet feeds, but still needs optimization for the majority of biomasses. The chemistry of the liquefaction process is not well understood and current research is very empirical. In this PhD work the approach will be to use model mixtures regarding to the lipids, proteins and carbohydrates composition of the residues and also minerals. Experiments will be done on the laboratory devices : batch reactors and the continuous pilot developed during the ANR LIQHYD project. An analytical work will be necessary to characterize the different products (gaz, liquid, water and solid phases). The objective is to build a global chemical mechanism for this hydrothermal liquefaction and to optimize the bio-oil production.

Failed fuel detection system for generation 4 sodium-cooled fast reactors

The Sodium-cooled Fast Reactors (SFR) will able the fuel breeding and the close fuel cycle. It is a promising Generation 4 reactor type for a future sustainable and competitive nuclear industry. This thesis concerns the development of nuclear instrumentations for fuel cladding integrity monitoring. The CEA works on SFR since the 50's with RAPSODIE, PHENIX, SUPERPHENIX, EFR and now ASTRID projects. Since the last SFR project, nuclear instruments have been improved notably by the use of digital electronics allowing the implementation of advanced signal processing. The CEA LIST develops innovative sensors and electronics for nuclear measurements. A high resolution gamma spectrometer operating into hard-radiation conditions living close to primary coolant system has been developed and qualified. The potential access to an isotopic knowledge about the released fission products will be a new improvement for SFR safety and operating. The student will have to understand the industrial needs and to develop physical models in order to simulate the coolant activation, the fission products releases, the transport of the radionuclides and the measurement. The work will require an important collaboration with the team in charge of the ASTRID project. The student will also have to set-up an experimental test loop at the SAPHIR linear accelerator facility in order to validate sodium activation models. The student will conclude on the feasibility of the gamma spectroscopy system and will design an optimized system and evaluate its performance for SFR safety and operating.

Detection of particles by magnetic phase transition

This PhD thesis concerns interaction of particles with matter and more particularly the study of magnetic properties of Gadolinium during interaction with neutrons or gamma rays. Indeed Gadolinium is a particular element regarding interaction with neutrons and also it shows some magnetic transition properties at room temperature. Thanks to its high thermal neutron cross-section, Gadolinium is often used in neutron sensors. Moreover it has its Curie point at approximately 20 °C. This second remarkable property has never been considered for particle detection. Nevertheless a Gadolinium-based bolometer could potentially show interesting neutron-sensitivity and discrimination properties under gamma field which could find many industrial applications. The student will start with a bibliographical study to assimilate the state of the art about interaction of particles and magnetic transient. Thus a feasibility study will be conducted trhough simulation and experimentation. Finally, a prototype will be designed, developed and tested. The study of a detection instrument based on an innovative principle is an important scientific challenge and could fit some strong industrial needs (neutron active dosimetry, NPPs decommissioning...)

Back contact silicon heterojunction solar cells optimization for high efficiency PV modules

The enhancement of solar cells efficiency is one way to reduce the cost of PV (photovoltaic) electricity. This may be achieved through an innovative solar cell design which combines the strength of the two highest efficiency c-Si solar cells (?24%) commercially available (SunPower MaxeonTM and Sanyo HITTM). Several labs and companies worlwide are interested in this new technology called IBC (Interdigitated Back Contact) Si-HJ (Silicon Heterojunction). The Laboratory of PV Components (LCP) at CEA-LITEN study IBC Si-HJ solar cells since a few years and this led to several publications and the fabrication of demonstrators with up to 20% efficiency. To fill the gap between this result and the theoretical efficiency above 25%, the devices has to combine perfect light trapping, surface passivation schemes and reduced resistive losses. These optimizations have to be be evaluated in this PhD, supported by device modelling as well as thin layers and interface characterizations. There will be a collaboration within this work with the Laboratory of PV Modules (LMPV) of CEA-LITEN.

Adaptive Medium Access Control for Virtual Sensor Network

The emerging technologies of the Internet of Things begin to connect physical world objects and places with people transforming citizens' interactions with the surrounding world and influencing how they live, travel, and spend their leisure time. They can also help in improving the quality of life of elderly people by supporting integrated living. The candidate will face the challenge of supporting re-programmability of virtual sensors and multiapplication sharing of heterogeneous sensor network components. To this end, the candidante will first design methods for application request aware sensor network components characterization. Then, the candidate will design MAC Aware sensors' characteristics protocols based on a recipe approach.

Methods of characterization of electrodes or packaging technics to be used in Active Implantable Medical Devices

The Active Implantable Medical Devices benefit from the miniaturization thanks to recent development in the field of micro and nanotechnologies, more and more functions are embedded in minimally invasive volumes. In particular electrodes used for neural recording or stimulation can be miniaturized in order to increase their number or their density or to reduce the volume of the device. Down scaling of devices requires a new assessment of stimulation paradigm and the technology at stake have to remain biocompatible. The PhD will take place at CEA/LETI/CLINATEC on the Minatec campus (Grenoble, France). Clinatec is a new biomedical research center at the CEA in partnership with the Grenoble Hospital, the French medical research agency (INSERM) and the University of Grenoble. Our mission is to demonstrate the proof of concept of preclinical and clinical medical devices based on micro and nanotechnologies. For 6 years, the LETI has been involved in the development of implantable packaging and electrodes (silicon or polymer based) for in-vitro or in-vivo devices. The characteristic dimension of electrodes goes from tens to hundreds of microns. The in-vivo behavior of these systems must be deeply investigated to specify their characteristic and their ?harmfulness?. The local tolerance of the electrodes and the packaging must be assessed by direct methods such as histology or indirect methods such as impedance spectroscopy or markers of inflammatory reaction. The research consists in developing a set of physical and biological methods using existing and innovative devices so as to fully characterize devices and to be able to better specify future component for recording and stimulation. The candidate has a Master of Science with knowledge in electrochemistry, clean room processes, electronic systems and eventually a background in biology.

Developing "Normally-Off" Architectures of Gallium Nitride HEMT transistors

Current transistors based on Gallium Nitride GaN are "Normally-On" HEMT (High Electron Mobility Transistors) that is to say that a current flows in the absence of gate bias. To meet the application specifications, it is necessary to develop new architectures of "normally-off" HEMT transistors with a positive threshold voltage without any degradations of others electrical characteristics of the transistor. The aim of the thesis is to identify the physical and electrical phenomena involved in the modulation of the electron density into the two-dimensional gas of the stressed heterostructure AlGaN / GaN and InAlN / GaN and leading to the realization of "Normally-Off" devices. For this purpose, different "Normally-Off" architectures will be evaluated based on the 200mm CEA-LETI clean room and on the existing "Normally-On" technology. This work will include: 1. State of the art of "Normally-Off" structures and identification of the most promising architectures based on TCAD. 2. The definition of process and integration variations in order to modulate electrons density in the channel starting from 200mm "Normally-On" technology in relation with process and integration teams. 3. Process follow-up of 200mm lots in the CEA-LETI clean room. Physico-chemical characterization of stacks and electrical characterization of new components (static and dynamic performance). 4. Analysis and interpretation of results. Identification of key physical and electrical phenomena correlated with TCAD simulation. 5. Summary of results and drafting the manuscript

Healthmonitoring for photovoltaic power converter

PV power plants are developing as important power sources that need to be improved, especially for operation and maintenance phases. Today several parameters are available to power producers in order to schedule preventive maintenance and identify the cause of failures. Inverter reliability is a major issue since inverter has the highest failure rate among the solar plant components and its failure is not predictable. The aim of this PhD proposal is to identify a set of parameters to monitor the inverter state of health in real time. Two major failure mechanisms are known: bond wire lift-off and degradation of the chip substrate. Nowadays it is possible to observe the evolution of these failures when the defaulting module is outside the inverter, ie you need to take the module out of the inverter in order to evaluate its state of health. Conversely we aim to conceive a method to determine the state of health of the inverter during operation, without arresting it, in order to ensure preventive maintenance and avoid abrupt loss of safety functions.

Interconnection methods for high efficiency photovoltaic cells : materials, process, characterization of electrical performance and reliability.

High efficiency cells are developed at INES. The aim is to keep high efficiency at module level. The interconnection of the cells is a key point, as the standard interconnection technique is not compatible with the materials used for the cell. Innovative interconnection technologies will be studied. The materials and process will be developed and characterized. Electrical performance of modules will be measured after module fabrication and after accelerated tests. The aim is to obtain high efficiency and reliable modules. The work is mostly experimental: development of processes and characterization.

Realtime computation of ultrasonic fields on massively parallel architecture

The CIVA software developed by CEA LIST offers different modules for simulation and treatment for non-destructive testing (NDT). In recent years, the use of simulation spread at all levels of expertise, however, the computation time remains an obstacle to an even wider use. On ultrasonic methods, different propagation and interaction models are developed by the CEA-LIST. These models are based on semi-analytical methods, and involve the simulation of the beam propagation of ultrasonic waves. Techniques currently used are based on pencil tracing and ray tracing modules. The subject of this thesis is to provide from the state of the art of simulation algorithms developed at CEA-LIST, a real-time calculation unit of the field propagated by one transducer phased array through a heterogeneous geometry with isotropic and anisotropic materials. For this, new hardware architectures : ManyCore CPU and GPU will be targeted. These architectures will require new algorithms and finely tuned source code to be developed in order to fully exploit their potential. Precise and rigorous benchmarks will be conducted on the targeted hardwares.

Design, fabrication and characterization of the electrical properties of a silicon nano-micro interposer for a CMOS / nano-object integration

The DCOS unit (Département des Composants Silicium / Unit of Silicon devices) at the CEA-LETI works on R&D projets which are focused on advanced CMOS devices, microsystems (sensors, MEMS, NEMS) and 3D integration. One of these projects aims at developing the integration of nano-objects on a silicon interposer and their electrical interconnection with CMOS devices (which is the basic idea of a CEA-LETI patent): this work axis puts the CEA-LETI in a leadership position in the international context. There are various potential fields addressed by this problematic, among which are molecular electronics (as addressed by the Atmol ICT project) and gas molecules detection with ultra-dense NEMS sensors (as addressed by the Delphins ERC project). The leading idea is to design, fabricate and characterize a single and innovative interposer enabling the exploitation and development of both microelectronic knowledge and techniques, and nano-objects knowledge (nano-objects being for example carbon nanotubes, graphene, nanowires, molecules with electronic or biological functions, bio-inspired materials)(cf figure). These application fields converge towards the design of a common hybrid nano-object / CMOS circuit, and the fabrication of a connective path meeting both the criteria from the nano-object integration (surface properties, electrodes densities and interconnections properties, nano-object environment) and the criteria from the CMOS performance. This is the context of this PhD thesis, which aims at developing the connective part of a generic silicon interposer, enabling nano-objects integration for their electrical characterization. The work proposed in the PhD thesis is focused on the design, technological development and characterization of a common 3D connective part between a CMOS device and a nano-object (or a nanoelectrodes array), that is to say from nanoscale (nanovias) to microscale (microvias). First technological developments are being studied and aim at fabricating such a connective path while meeting the surface criteria required for the integration of the desired nano-objects: it uses nanovias and micro through silicon vias (TSV), the first ones being filled with silicon while the second ones are metallic. A standard technology is planned for the CMOS part, and may be adapted if required for the integration of nano-objects. This PhD thesis will focus on the design of a 1st CMOS / nano-object circuit and the fabrication of the connective part, bringing together 3D technology, CMOS technology and nano-objects specificities. The proposed work will include the study of the interposer electrical and mechanical properties (electrical and mechanical compatibility of the materials, interfaces). Three main axes will be considered: - Electrical modeling and fabrication of a hybrid CMOS/ nano-object circuit, - Interposer technological design and fabrication, - Electrical and morphological characterization of the test vehicles. The PhD candidate will be welcome both in the LECA lab (Laboratoire Empilements des Composants Avancés / 3D Integration lab) and the LDI lab (Laboratoire des Dispositifs Innovants / Advanced CMOS devices lab), in the DCOS unit. The candidate will also have access to technological and characterization facilities from other units. Candidate profile: Master in Electronics, Nanoelectronics or equiv. Communication skills Experience in one of these fields is an asset: semiconductor physics, microelectronic techniques (work in cleanroom), nano-objects manipulation.

Epitaxial growth of AlGaN alloys on silicon substrates for LEDs in the UV-C domain

For water and air purification applications, UV LEDs are gradually replacing mercury based lamps, because of their high prices and low energy efficiency. The mùaterials involved in such LEDs are semiconducting alloys of the Ga1-xAlxN family. However, for the considered wavelength (around 270 nm), the UV LEDs have not yet reached the required performances because of their low conversion efficiencies. This limit is mainly due to the very high dislocation density present in the epitaxial layer, that act as non radiative recombination centers. We propose to investigate new epitaxial growth processes to try and circumvent the problem. These will rely on using either patterned silicon surfaces or localised growth through a mask on the silicon substrate surface, in order to limit the dislocation density as much as possible. We will work on the definition/optimisation of the pattern on the silicon surface and on the mask definition, together with the optimisation of the growth processes in both cases. Once this is achieved, we will then turn to epitaxial growth of the whole LED structure and characterize its electrooptical properties. The work mainly concerns epitaxial growth, but will also deal with optical and structural characterization(photoluminescence, cathodoluminescence, electrooptical spectra and efficiency, X-ray diffraction). The work will take place in the Lighting Component Laboratory from CEA/LETI in close collaboration with CNRS labs, CRHEA in Sophia Antipolis and GES in Montpellier.

Design and validation of predictive control algorithms to control digital human in interactive time

This thesis is part of the Digital Human theme for industry, and more specifically for the design, maintenance, operator training, workstation design and ergonomics. Unlike the approaches currently used in the industry, our approach has been, from the beginning of the theme development (2006), to propose a dynamic digital human, physically realistic, controlled by its own actuators, that behaves at most like a real human without the need for accurate tracking of the operator movements - the avatar can then be treated as a collaborator rather than a simple projection of the operator. Digital human controllers today have a purely reactive behavior and do not anticipate the task. This involves the use of high servo gains, which do not guarantee, in case of complex and dynamic movements, stability of the digital human, and consequently, the success of the task. This thesis will focus on the complex and dynamic movements of digital human (quick posture change (standing - crouching), technical movements, walking, or running, ...) and will aim to propose interactive principles and control algorithms based on prediction, in order to obtain a better robustness to both, human behavior and control parameters.

New high capacity electrode materials for all-solid-state lithium microbatteries

The recent technological development of the stand-alone microsystems is limited by the miniaturized source able to power such systems. All-solid-state thin film batteries are attractive systems to power such devices. Their thickness do not exceed 10 microns, and they are built of several layers including a glassy lithium ion conductor as electrolyte, deposited by PVD or CVD techniques. Nevertheless, their surface capacity is generally less than 100 µAh.cm-2.µm-1. The aim of this PhD thesis is to study new positive electrode materials, called conversion materials in order to greatly improve this surface capacity. Hence, transition metal sulfides will be prepared in a thin film form by sputtering. The influence of sputterring parameters on both the composition and the structure of the films will be studied. Electrochemical performances and lithium insertion/deinsertion mechanisms in these materials will be comprehensively studied. A wide panel of characterization techniques will be used to study these materials and all-solid-state cells including them: ICP, RBS, EPMA, XPS, Auger spectroscopy for determination of the chemical composition; XRD, Raman and Mössbauer spectroscopies,HRTEM to determine the structure; SEM to study the morphology and electrochemical methods (EIS, Galvanostatic cycling, cyclic voltammetry).

Beyong lithium microbatteries: new materials for Li-free systems

New families of miniaturized stand-alone and wireless devices are emerging, embarking their own energy-harvesting (photovoltaic, piezoelectric, thermoelectric) and energy storage (microbattery) systems. Lithium all-solid-state microbatteries are performant electrochemical systems. Nevertheless, their operation voltage is quite high for these specific applications which require quite low voltage (ca. 1V) to achieve improved energy recovery yields. This implies to consider new all-solid-state electrochemical cells free of lithium and potentially less reactive towards humidity (convenient for implantables sytems). The aim of this PhD thesis will be to prepare and evaluate new couples of electrode materials able to fullfill these requirements and compatibles with an integration in all-solid-state microbatteries.'Simple' systems based on a metallic anode and a intercalation compound of same metal ions (mono or divalent) will be studied at first. The study will deal with both powder and thin film materials (PVD). A wide panel of characterization techniques will be used to study these materials and all-solid-state cells including them: ICP, RBS, EPMA, XPS, Auger spectroscopy for determination of the chemical composition; XRD, Raman and Mössbauer spectroscopies,HRTEM to determine the structure; SEM to study the morphology and electrochemical methods (EIS, Galvanostatic cycling, cyclic voltammetry).

New solid electrolyte based on ionic liquid: realisation by wet method and impedance spectroscopy characterization

One of the main issues for the development of portable technologies is the energy storage. For many applications (hearing implants, back up energy sources, etc.), the commonly used battery is the coin cell. However, this type of battery is oversized compared to its use. Today, one of the most promising solutions would be the use of solid state microbatteries. Developed for twenty years, these micro-batteries are made by successive thin films deposition of less than ten microns. They are composed by a solid electrolyte which avoids the risk of leakage compared with classical batteries using liquid electrolytes. Currently, they are mainly performed by vacuum deposition processes like PVD or CVD. However, these techniques have a restrictive application (high vacuum, expensive equipment, etc.). Moreover, obtained thin films exhibit a low conformity which not allows the deposition on rough or textured substrates. The aim of this PhD thesis is to develop innovative solid electrolytes by wet method. These electrolytes will be based on a polymer matrix (like UV polymerized material) with a lithium salt and an ionic liquid. These types of electrolytes are stable under ambient atmosphere and high temperatures. Moreover, they can be deposited by wet deposition techniques like dip-coating or spin-coating. This study will begin by the choice of the optimum precursors and deposition technique for the elaboration of thin films. In a first period, the various precursors will be characterized in terms of conductivity, viscosity or IR-spectra signature. Then, deposited thin films will be electrochemically characterized by impedance spectroscopy and integrated in an all solid state micro-battery. This thesis will be done in collaboration between the CEA of Grenoble and the ICMMO laboratory of Paris Sud University.

Advanced methods for defect characterization using ultrasonic phased array techniques

Techniques for ultrasonic non-destructive testing based on phased array probe are now increasingly used in industry. This technology offer many advantages, in particular it provides access to collection of data richer and more complete than those provided by traditional techniques. The optimal use of acquired data for detection and characterization of defects requires the implementation of new fast and robust imaging techniques. Imaging methods such as synthetic focusing algorithms are now increasingly being developed. These methods provide a large number of images, each of which may contain different and additional information on the location, nature and geometry of defects. However, tools for use and interpretation of such set of information do not yet exist. The objective of this thesis is to develop new tools and methods, mainly based on the simultaneous use of information provided by existing imaging methods, for defect characterization in terms of size, position and shape.

Study of the stability in aqueous medias of catalytic systems used for hydrolysis of NaBH4

Laboratory of Chemistry and Safety of Nanomaterials (LCSN) is strongly involved in the development of energy sources based on micro fuel cell. They relate mainly to low power applications for nomad devices. The main advantage of these energy sources is their theoretical energy density potentially much greater than the lithium batteries to increase the autonomy of electronic systems. Fuel and more particularly the "fuel cartridge" remains the main obstacle to the emergence of fuel cells in the field of nomad applications. Today, many industrial and research laboratories, including the CEA / LITEN focus on hydrogen fuel and the means to store or produce it in-situ. The LCSN, in close collaboration with a renowned industrial, works for ten years on the hydrogen generation by hydrolysis of sodium borohydride NaBH4. Thus, many studies are being conducted in the laboratory on the implementation of this reaction with the ultimate goal of designing and developing a passive generator to generate the hydrogen required for fuel cell functioning. The use of catalysts to activate the hydrolysis of NaBH4 in aqueous solution is required to generate sufficient flow of hydrogen. This thesis, on the study of the stability of the catalytic systems in aqueous solutions, will focus on the following themes: - Understanding of the mechanisms of aging non-noble catalysts (study fundamental, electrochemistry ...) - Determination of the optimal conditions of aging (nature of the solution, the pH of the solution, addition of additive atmospheric aging ...) - Evaluation in real conditions. The stakes are both scientific and technological.

Development of a 3D method based on phase contrast imaging with synchroton and laboratory X-ray source

In radiography X conventional, photons X crossing a sample are differently absorbed according to the nature of the material. The image trained(formed) in exit(release) of the sample so represents a distribution(casting) of the absorptions in the sample. Another approach is the radiography by contrast of phase which represents a variation of the phase of the beam appearing from the sample. During these last ten years, research has been active in this field and several techniques of phase contrast imaging were proposed. These techniques require the use of a beam of X-rays which can arise from a strongly coherent source(spring) (typically synchrotron) or partially coherent (typically classic X-ray tube). The purpose of this thesis is to develop a 3D imaging method based on phase contrast imaging. This method will be developed in the aim of dealing with beams produced by classical laboratory sources.

Advanced concepts for structural modification of amorphous silicon layers and their application on heterojunction solar cells

Amorphous / crystalline silicon heterojunction solar cell technology (HET) has become a very promising because it combines easily industrialized processes avec very high efficiency (> 23%). The key of heterojunction solar cells is the deposition of hydrogenated amorphous silicon (a-Si: H) on a substrate of monocrystalline silicon. This allows obtaining excellent surface passivation and reaching open circuit voltage values (Voc) higher than 730mV. This thesis is focused on using innovative methods to dope and more generally to improve the structural properties of a-Si:H. The three processes approached for this goal are: ion implantation, irradiation and micro or full Laser crystallization. The aim of the work is to get stacks of a-Si: H with better passivation, improved conductivity and higher transparency for high efficiency HET cells by these innovative technologies. The application will be the single or bifacial heterojunction cells and rear contact heterojunction solar cells.

Development of 3D modelling tools for simulation of ultrasonic testing of complex welds

Non Destructive Testing (NDT) techniques can provide information on the health of a component or a structure without inducing detrimental changes to its future use. Among the NDT methods, techniques based on ultrasonic elastic waves are widely used, especially to control the volume of structures. Among these structures, the controllability of welded joints is a major industrial challenge to meet the requirements of safety and service life of structures. Based on research activities on this subject, this thesis aims to improve the modelling of wave propagation phenomena in welds. The axis of reseaech are: 1. The extension of models in order to take into account 3D properties of welds coming from either macrographic description or welding numerical simulation codes; 2. The modelling of the effects on the ultrasonic wavefront due to the polycrystalline structure of welds: attenuation of the ultrasonic beam and a backscattered noise. The models will be validated by comparison with experiments in cases representive of industrial applications.

Modelling of critical elastodynamic phenomena in the asymptotic ray theory. Applications to the simulation of ultrasonic testing

The CEA LIST develops the CIVA software platform that proposes simulation tools for non destructive testing (NDT) techniques and in particular for ultrasound. The models that are currently used to simulate the ultrasonic wave propagation are based on the dynamical ray theory, which is a high-frequency approximation valid in a wide range of practical situations. Nevertheless this theory does no account for interference between the bulk waves and the so-called head waves. This thesis aims at developing new efficient model taking into account these critical phenomena and allowing to accurately describe the interaction between an ultrasonic beam and a defect. Two different approaches will be studied, a numerical integration and a high frequency asymptotic evaluation specific of the Rayleigh-Sommerfeld integral.This method should respect limits in computation time and should be adaptable to different control situations.

3D Tomographic Reconstruction on Non-standard Trajectoires

The tomographic reconstruction was initially developed for a 360° circular trajectory. Advances in the medical field lead to the use of helicoidal paths and therefore some reconstruction algorithms were adapted. In the field of non destructive testing (NDT), the technological advances lead to the use of robotic positioning systems in the process of automatization the testing procedure. For this case, the acquisition trajectories may vary and in the same time mechanical constraints may impact the trajectory. The main objective of this work is to propose, adapt and optimize CT reconstruction methods for non-standard trajectories. Both analytical and iterative algorithms are of interest, especially GPU accelerated versions. Two important problems arise: missing data when compared to traditional acquisitions implies the ill-posedness of the inversion problem and secondly the positioning of the source-object-detector system becomes critical. For the first aspect pre-processing techniques such as inpainting can be used or a direct handling of the missing data by the reconstruction algorithm will be investigated. For the second aspect, detection and calibration techniques should be implemented in order to ensure the correct positioning. The developed algorithms will be validated on synthetic and experimental data generated and acquired with the tools and devices available in our laboratory.

Modelling of the radiation of ultrasonic guided waves in a thin structure in view of its nondestructive examination

Elastic guided waves (GW) propagate at long range, permitting the examination of large structures. Their complex behaviour makes it necessary to conduct simulation studies to help interpreting results and to improve nondestructive testing (NDT) methods that use them. The CEA LIST develops the CIVA software for NDT simulation and the models on which simulations are based. One module simulates GW-NDT of structures having a single guiding direction (pipeline, rail). Models are based on semi-analytic finite element (SAFE) techniques for the guided propagation, coupled to finite elements for calculating local interaction with defects. The thesis aims at extending this module to the case of thin parts that guide waves in all directions. A model for predicting the field radiated by ultrasonic transducers will be developed and by reciprocity, will predict their sensitivities in reception. A formulation for coupling this model to a model of interaction of a GW with a defect in such structures, in progress, is to be developed too. The overall model shall be able to deal with testing configurations involving curved anisotropic parts. Validation experiments will be made in collaboration with experimentalists at CEA. The theoretical work will be developed in the framework of a long term collaboration with specialists of aapplied mathematics.

Scattering correction in flat panel detector X-ray tomography

X-ray imaging is a non destructive or medical method that allows the 3D estimation of the density of an object of a patient from its 2D projections obtained at different angles. This method allows the detection of defects in the sample, such material inhomogeneities, cracks, occlusion and the geometrical control of the sample or patient (morphology, dimension, material, fractures). When using a 2D X-ray detector, the image is blurred by photons scattered by the object itself. The magnitude of this signal can reach 2 to 5 times the desired signal, depending on X-ray energy, material of the object and geometrical configuration. The scattered signal leads to incorrect estimation of reconstructed image. Due to the high performances of X-ray modelling accuracy, as exists in the software Civa, the objective of the thesis is to rely on scattering modelling for scatter correction and compensation in the 2D projections, dynamically in the reconstruction process or directly in the projector. The first stage of the thesis will consist in taking over the X-ray simulation tools and reconstruction methods used in X-ray tomography. The second stage will lead to the development of an iterative method to compensate X-ray scattering, for a short number of geometries. The third stage will be the validation of the method on experimental data.

Estimation of the imaging dose to the patient for embedded X-ray computed tomography in image-guided radiotherapy and integration in the treatment planning

Protocols for cancer treatment using intensity-modulated radiation therapy (IMRT) allow to target the tumor with an increased precision. They require accurate anatomical information of the patient just before the treatment, which can be obtained using on-board imaging systems mounted on the linear accelerator delivering the treatment beam. These systems called kV-CBCT combine a kV X-ray tube with a 2D planar detector and are very recently introduced. Their use is expected to grow in next years. However, these kV-CBCT exams are also responsible for an additional dose of ionizing radiation which is far to be negligible et which is suspected to be the cause for acute secondary effects and radiation-induced second cancers for treated patients. The PhD project aims at developing a simulator based on the Monte Carlo method in order to estimate accurately the delivered doses during kV-CBCT exams. The developed simulator will then be used to study different strategies to take into account for these imaging additional doses during the treatment planification process, with the objective of getting all the best from kV-CBCT imaging without bringing additional risks to the patient. This study will be conducted on selected patient cases and will allow to draw optimized and patient-specific protocols for clinical routine. This PhD project will be held in collaboration between the DCSI/LM2S (CEA Saclay, DOSEO platform) for the simulator development and the Medical Physics Unit of the Centre Eugène Marquis in Rennes for the clinical evaluation of the simulator.

Efficient contact determination based on boundary representations (B-Rep)

The simulation of multibody systems with intermittent contacts has numerous and various applications in the engineering of mechanical and mechatronic systems, as well as in biomechanics, where contact phenomena may endorse a crucial importance. For particular applications, involving high sliding velocities for instance, the respect of the smoothness of the contact surfaces may turn into a necessity: indeed, the discontinuities in the normal direction at contact points, induced by polyhedral approximations, can introduce spurious mechanical impact phenomena, which may result in dramatic modifications in the behavior of the system being simulated. The goal of this thesis is to propose efficient algorithms for the determination of contacts, enabling the use of boundary representations (B-Rep) for the geometrical shapes of the bodies. These representations could come either from CAD models, compatible with B-Rep standards such as IGES or STEP, or be based on shape functions derived from finite element models for the simulation of deformable bodies. The LMD++ contact determination code, which is a module of the XDE Physics multibody simulation code developed at CEA LIST, will be used as an existing basis for the implementation of the algorithms explored in this thesis.

Study of extrinsic doping in high operating temperature (HOT) HgCdTe photodetectors

CEA/Leti is, in collaboration with the Grenoble based company Sofradir, one of the world leaders in the research and development of Infrared photodetectors made in the II-VI semi-conductor HgCdTe. One of the main challenges in this field is the increase in operating temperature of the detectors. Theoretically, the highest operating temperature in HgCdTe photodiodes should be achieved in extrinsically doped material. The characteristics of such detectors made so far have however failed to meet the expected performance level. The reason for this has been attributed to defects that are generated by the introduction of the dopants. The present thesis will be made in the context of the development of new means to incorporate and activate dopants in HgCdTe. In the beginning, the work will be concentrated on the structural (x-ray and electron microscopy) and physical (minority carrier life-time and hall-effect measurements) characterisation of samples prepared by the different techniques. In a second phase, the student will participate in the design and characterisation of innovative high-operating temperature photodiodes based on the developed material.

Modeling the ultrasonic response of cracks using BEM and coupling with a propagation model, application to nondestructive testing

The CIVA simulation platform for non destructive testing is developed at CEA LIST. It offers simulation tools for usual inspection techniques. The general objective of this PhD work is to provide an additional method for the modeling of ultrasonic inspections. In this purpose, the elastodynamic behavior of a flaw with a complex geometry will be modeled using the Fast Multipole Boundary Element Method (FM-BEM). This model will be coupled with tools existing in CIVA in order to account for the propagation of diffracted ultrasounds in the part and towards the receiving transducers.

Ultra wideBand power amplifier spectrum calibration method and architecture

In the context of smart environment, Ultra-WideBand (UWB) technology receives a great interest as a promising technology capable of both very accurate localization and high-rate short range communication network. Nowadays the Federal Communication Commission (FCC) has specified the emission mask of UWB devices within the 3.1-10.6GHz. In order to enable the worldwide dissemination of UWB devices the need of transmitter able to guarantee the output signal spectral occupation is mandatory. Most of test probers are not intrinsically able to verify this specification because of the impulse nature of the signal and because of the time, and therefore the cost, needed for spectral analysis. This PhD aims to address this specific problem in order to reduce the time, coat and complexity of the UWB transmitter test. In addition, like pre-distortion loop already existing for continuous time power amplifier (PA) the PhD aims to provide a mean to control the output spectrum of the UWB PA. In literature a few PA using different techniques to provide UWB pulse can be found. Most of them are versatile and provide different control to manually modify the signal spectrum. But none of them enable an automatic control. As UWB PA always works in transient nonlinear regime, the first part of the PhD will focus on the output signal theory in this specific regime. Then, some metrics will be identified in order to provide, using correlation techniques, a mean to estimate and servo-control the output spectrum. after this theoretic and control part of the PhD, a design part will produce probe circuits as less invasive as possible to provide identified metrics and the loop back control of the output spectrum.

Modelling of the plasma-surface interaction during LIPON deposition by radiofrequency magnetron sputtering

The LCMS laboratory (Laboratory of Micro Storage Components) of the CEA/LITEN (Grenoble) has developed for more than ten years thin layers microbatteries for microelectronics applications. Microbatteries are constituted mainly by two thin layers, the anode and the cathode of the device, separated by a solid electrolyte. Thin layers of LIPON, the most used solid electrolyte, are deposited from a Li3PO4 target by radiofrequency magnetron sputtering (f= 13.56 MHz) in pure nitrogen plasma. It is important to control the deposition of this layer in order to optimize microbatteries and to obtain reproducible performances. The proposed thesis work consists in modelling the RF magnetron nitrogen plasma and its interaction with surfaces. This work will be done with the collaboration of the LPGP (Laboratory of Gas and Plasma Physics) in Orsay. The results of this model will be compared to experiments which will be done in LPGP, using R&D deposition chambers, in order to validate this model, and in the CEA/LITEN in an industrial-type deposition chamber. This thesis work will be mainly located in the LPGP in Orsay.

Planning in adaptive radiotherapy dose using images from embedded X-ray computed tomography

This thesis aims to demonstrate the feasibility of dose planning in radiotherapy using images of the patient obtained by X-ray computed tomography under the linac (kV-CBCT imaging). To reach this goal, a simulator based on the Monte Carlo code PENELOPE will be developed to enable the analysis and characterization of the beams from the imager. A particular attention will be paid to the scattered fraction. In the mean time experimental acquisitions will be performed to relate the characteristics of beams with measured Hounsfield units for different acquisition protocols. The purpose of this comparison is to determine the acquisition parameters for dose planning which lead to the same precision than the planning in virtual simulation.

Development of algorithms for the Tracking of Optimum Point of Charge for electrochemical storage technologies. Application to Li-Ion and ZEBRA batteries chargers

The fast charge algorithms of electrochemical batteries decrease inevitably the global charge/discharge efficiency. The previous studies held by the CEA-Tech have permitted to determine a parameter for setting the charging rate at an optimum value that can reduce the charging time while minimizing the losses. This optimal rate called OPC (Optimum Point of Charge) varies dynamically with the states of the battery. This thesis aims to extend these initial studies operated on Lithium-Ion cells to storage systems of Li-Ion packs and High temperature Sodium batteries (ZEBRA) including reversible inverters. The purpose of this thesis is the development and the implementation of OPC tracking algorithms into the inverters.

Synthesis and characterization of high energy density layered oxides for Li-ion batteries

degradation diagnosis of planar breathing micro fuel cells

LCSN Laboratory (CEA Grenoble LITEN) develops planar breathing fuel cells for low power supply (1-50Watt) since ten years. The work, done in a close relationship with an industrial partner, has led to the development of the fuel cell basic building blocks. First commercialization stage is targeted within 3 to 4 years. In this context, it is crucial for us to study / analyze and improve the performances and the ageing of these prototypes. The PhD work will mainly consist in studying the ageing of fuel cell core as well as of functional prototypes. It will also be necessary to participate in improving the devices, to propose innovative solutions. This work will be articulated according to three phases - Test strategy: the test strategy will be developed in agreement with the functional devices constraints, in a real working mode. - Test analyses: the results exploitation will be made through electrochemical characterization methods. In addition, ?post mortem? analyses of the devices will be done via characterization of the core fuel cell materials as well as the integration materials. Characterization techniques such as SEM, TEM, RBS and XPS will be used. It will also be an objective to study the impact of various contaminants on the fuel cells behavior. - Integration / materials: from these tests and characterization, new materials and/or architecture will be proposed and explored.

Intelligent onboard electronics for Concentrator Photovoltaic technologies

In an energetic market mainly based on fuel energies, renewable energies are gaining more and more market share thanks to economics and politics support. Among this new market and technologies, photovoltaic is one of the more promising ones. The CEA-INES is working for long years in the development of new PV technologies and more recently is involved in high efficiency photovoltaic, also Concentrator Photovoltaic (CPV). CPV modules consist in an optical element concentrating a big amount of sunlight onto small area and high efficiency photovoltaic cells. The use of an optical system leads to the use of tracker which role is to align the module to the sun in order to form an image of the sun on the cell. CPV technologies are emerging compared to classic PV technologies, however, with conversion efficiencies twice higher, they are considered as a very promising alternative. In CPV modules, the cells are series connected and mounted along a by-pass diode. This by-pass diode is used to disconnect the cells showing degraded performances. Due to the exigent assembling tolerances needed in CPV, some alignment problem between cells and optics can induce some inhomogeneity in terms of efficiency. In this case, the by-pass diode can ?disconnect? the affected cell. This phenomenon avoids extracting the maximum power of the CPV module. However, CPV modules are usually of great size and filled by air. This means that they physically room to include new functionalities such as onboard intelligent electronics. In the frame of this PhD, we proposed to design an intelligent electronic board able to make each cell of the CPV module working at its maximum power point instead of disconnecting the cell with defects such as misalignment. Moreover, this electronic could be coupled to a small energy storage component in order to create an energy buffer (from the inverter point of view) to reduce the constraint applied in the case of a cloudy day. At the end of this PhD, a prototype will realized and integrated to a CPV module for an infield characterization. This PhD is aimed to work on a very foreground technology. For this reason, this thesis will have a great potential in terms of innovation, publications and patents.

Discriminant dictionnary learning for sparse coding of signals

Interactive analysis and validation of multi-agent based simulations with fuzzy logic

Multi-agent based simulations (MABS) have been widely used in complex systems modeling like in socio-spatial (urban dynamics), sociotechnical (smart-grids) or biological phenomena. These simulations can be very complex to analyze when there is a large number of agents, behaviors and interactions involved. However, only few tools enable to analyze and validate these simulations, none of them enabling an analysis guided by the user. This thesis aims at tackling this issue by proposing new analysis and validation tools in order to ease the understanding of the underlying phenomena. These tools will be based on the use of fuzzy logic that permits to get intelligible results automatically or interactively, as well as visualization tools that enable to get quick and intuitive analysis results of one or several simulation. The definition of fuzzy concepts, that gives sense to numerical values, as well as fuzzy rules, making causal relations, will be used to analyze the behavior of homogeneous groups of agents identified by clustering in the simulation, test hypotheses for a specific scenario and search the configurations that meet user or expert requirements.

Modelization, conception and experimental investigation of a novel high temperature thermal energy storage system based on packed beds. Applications to CSP systems

CSP power plants use mirors to concentrate the sunlight to heat a heat transfer fluid used as hot source in a thermodynamic systems to produce electricity. Due to intermittent solar ressource, the integration of a thermal energy system is needed to optimize a CSP power plants. The use of the thermal energy storage system allows * to maintain constant electricity production during cloudy periods; * to reduce start and stop of the electricity production device such as turbine; * to optimize the turbine operating conditions; * to match the electricity production to the demand; * to increase the global capacity of the CSP power plant. The thermal energy storage studied here is a regenerator using solid storage material. The heat transfer fluid depends on the temperature level: * for moderate temperature (300-450°C), a liquid heat transfer fluid such as oil or moltel salt is used; *for high temperature, the heat transfer fluid is generally air. Optimization of a CSP power plant with thermal energy storage needs deep investigation and understanding of the hydraulic and thermal behaviours of the thermal energy storage system. In addition, specific work has to be made on the operating control. To meet these objectives, the PhD student will use two different approaches, based on experimental test loop already operating (STONE using thermal oil and CLAIRE using air as heat transfer fluid) and 1D and 3D numerical models.

Development of new aqueous Li-Ion technology

After a first PhD, this subject concerns the improvment and the set-up of new Li-Ion technology working totally in aqueous media. The scope of the study is mainly concentrated on the development of solutions allowing overcoming the restricted stability window of the water, which widthness is only 1.23V. The work will benefit from the promising results of the first PhD and will have to go beyond in terms of stability and ideally should allow touching the problematics of life duration at the end of the thesis. The works will be jointly led at LEPMI facilities, campus of Grenoble (INPG, St Martin d'Hères) and at CEA-LITEN-LBA. The solutions under study concern both the use of additives within the electrolyte and both surface deposits on the electrode materials in order to passivate and to delay the electrolysis reactions. The work could go from the synthesis of the additives if they are not commercials to the different electrochemical tests to qualify the selected solutions (voltametry for instance)

Determination of L X-ray emission intensities of radionuclides with Z > 80 using high energy resolution metallic magnetic microcalorimeters

The Laboratoire National Henri Becquerel (LNHB) is the French national laboratory of ionizing radiation metrology. It oversees the definition and improvement of national standards in the area of radioactivity and strives to improve measurement methods, as well as publishing tables of recommended radionuclide data. X-ray photon emission intensities are essential radionuclide data for quantitative analysis through X-ray spectrometry. However, the intensities of many L X-rays are not well known because i) experimentally, the L X-ray energy spectra cannot be resolved using conventional semiconductor detectors, and ii) theoretically, important approximations are required in order to perform intensity calculations. The goal of the thesis is to measure L X-ray intensities between 5 and 25 keV with unequalled high precision for radionuclides with an atomic number Z > 80. This will be possible using high energy resolution metallic magnetic calorimeters (MMCs). MMCs are thermal cryogenic detectors that have excellent energy resolution (18 eV at 6 keV for example) when operated at extremely low temperature (~ 20 mK). The programme of work for the thesis includes the following steps: 1) A bibliographic study to identify the key radionuclides for which insufficient data have currently been published, hence more detailed measurements are required. 2) The experimental set-up will be optimized and installed in the dilution cryostat of the LNHB, and characterized in order to provide quantitative measurements of L X-ray intensities. 3) The L X-ray intensities of the key radionuclides will be measured with unprecedented precision and details.

Thermo-mechanicam modelling of the hydride breathing for a safe and efficient design of hydrogen solid state storage tanks

The thesis concerns the hydrogen storage through hydride materials. The team "Hydrogen Storage" in the laboratory of hydrogen technologies is responsible for developing this type of technology called solid storage, which has the advantage over the other two common storage route (as compressed gas or liquid form) to be more secure and compact. The hydride material is in the form of a dry granular medium. Hydrides that work well in a reversible manner are currently intermetallic materials, with a grain size of a few microns to several hundred microns. The generic phenomenon of hydride breathing (swelling/shrinking during the absorption/desorption of hydrogen) is a known phenomenon, but only empirically for the moment. Yet it is important to control it for two reasons: the efficiency of the heat exchanger in which the hydride is contained (as the hydriding/déhydruration phenomenon is exo/endothermic) and the mechanical integrity of the container. This topic closely implies the field of mechanics of granular media with an aspect that has been very little studied until now. The fundamental issue is to answer the question: how does a granular material behave knowing that each grain can see its volume change by 20% to 30% ? The main objective is to predict the level of load created on the walls of the container. The thesis aims firstly to better characterize the mechanical behaviour of a material during hydride cycles of absorption-desorption of hydrogen. For this, an experimental apparatus has just been developed in the laboratory that permits a mechanical characterization under tri-axial of revolution state of stress of the hydride under hydrogen with strain measurement by optical method. The bench is operational, but has not provided results yet. In a second time, the work consists in developing a predictive model of the mechanical behaviour of the hydride (or even a thermomechanical model, because coupling to the thermal effect is expected). The first method is to use the discrete element simulation that can take into account the physics of the phenomenon better, but we will also try to develop a continuum model type, more suitable for "engineer" design of real tanks, on the basis of existing models developed in the field of compaction of dry granular media (eg Drucker Prager Cap Model). It is also desirable to look at analytical models such as silo models (Jensen).

Critical scenarios identifying in railways safety procedures using a dysfunctional analysis

The strategic aim of the PhD work is to propose a dysfunctional analysis methodology for identifying safety critical scenario ensuing from the specification. In other words, this is a contribution to the robustness evaluation of procedures contained in the SRS ERTMS in case of failure occurrences. This work takes place in the framework of the Perfect ANR Project which deals with the integration of the ERTMS procedures in a French national context. The human factor is taken into account. In fact, even if the control is assisted by a global automation system, hypothesis of human errors cannot be eliminated. The railways European industry uses a global norm which specification is grouped under the name ERTMS for the on board materials. Concerning the non-rolling materials the specification is given by the national rules for each European state. Some ERTMS modeling works can be found in the state of the art. There are some models for national rules too. The framework of the PhD study considers two different models built separately in order to claim that a system respects the global specification or not. A particular focus on dysfunctional analysis is assumed. The aim is to identify the critical scenario which may not respect the specification and which may not achieved the required global level of safety. It is also assumed that the modeling point of view to b used is the safety requirement one. The technology to be used in order to identify the critical scenarios is the one of model based automatic test generation. The first step is to identify the appropriate modeling tool and methodology for the considered rules. An oriented state of the art concerning dysfunctional analysis is to be made taking into account the know-how of both laboratories ESTAS and LISE. The originality of scenarios is to be discussed and validated regarding the truthfulness of human operator behaviors. Then an implementation using the papyrus software modeling tool and the test generation tools of the LISE laboratory will be made. A final link with the ERTMS simulation software environment can be made at the end of the PhD work.

Development of silicon highly integrated capacitors

The IPDiA company (Caen, France) is manufacturing silicon integrated capacitors with outstanding performances of reliability and linearity (temperature, voltage) suitable for very high-end markets like medical, aeronautics, or petrol. This PhD aims to optimize the capacitance density (goal of 1µF/mm2), the electrical performances and the robustness of a new generation of integrated capacitors. Differents dielectrics stacks have been identified but their silicon integration as well as the design of the capacitors need to be optimized. The PhD will be part of a common lab between IPDiA and CEA-Leti, academic supervisor will be G2ELab laboratory.

Advanced Patterning process development for sub 14nm CMOS technology

For sub 14nm CMOS technology, Optical lithography based Patterning techniques are limited in terms of minimum dimension and minimum pitch. Alternative techniques, such as ebeam lithography, spacer patterning & direct self assembly, are able to overcome the sub 14nm challenges. One of the key points of such patterning techniques is to have innovative plasma etch processes that will enable the integration of these solutions. Plasma etching processes are able to transfer mask dimension into multi layers stack with a good control of the anisotropy & the selectivity towards the used materials. The goal of this PhD is to study & develop plasma etching processes for alternatives patterning techniques (Ebeam, DSA, Spacer patterning) to reach the sub 14nm dimension for line/space features.

Experimental and numerical study of 3D metallic powder densification using Hot Isostatic Pressing

Main objectives of this PhD work is to study powder mechanical behavior at low (particle reorganization) and high temperature (viscoplasticity) of a metallic powder, with the aim of the modelling of a complex 3D parts compaction by Finite Element Method. The PhD will contain an experimental part (instrumented HIP experiments, traction and compaction experiments,SEM TEM XRD observations, tomography or/and FIB), a numerical part (Discret Element Method, Finite element method, ODE resolution) and finally a more applied part consisting in a complex part compaction simulation using new modeles developped in this three years work.

Shot-noise parameters estimation, application to nuclear instrumentation

The shot noise, i.e. a marked and filtered Poisson process, is particularly relevant for modelling detector signals in nuclear instrumentation. But the domain of application is much broader than particle physics. The shot-noise very specific properties, like the absence of the analytical expression of its density, explain that very little is known about optimal estimation in this context. The classical maximum likelyhood (ML) method and all the related results cannot be directly applied to this problem. This thesis aims at deriving algorithms that outperform the methods presently used in nuclear instrumentation. This will be made possible by exploiting the complete shot-noise distribution instead of only taking into account the first cumulants.

Design and Characterization of Memory Circuits and Macro-Cells based on organic Materials for Printed Electronics

The work of this PhD is devoted to de design and characterization of a non-volatile memory circuit in the frame of the memory and logic technologies developed within the Laboratory of Printed Components (LCI) of the CEA-LITEN. Improvements in material synthesis for organic electronics enable now the fabrication of circuits, intended for applications on large surfaces and on substrates of all types (flexible or conformable, etc). In this context the building block for information storage are still required for the autonomous circuits (Internet of Things, RFID, Sensor Tags). The aim this PhD is to design and characterize Circuits and macro cells for memory application based on the logic and ferroelectric and resistive nonvolatile memory printable technologies developed at CEA-LITEN.

Development of advanced interconnections for power applications and study of their reliability

To respond to requirements for environmental protection and alternative energies (solar panels, wind energy, electrical vehicles), CEA-LETI is developing power devices based on HEMT GaN technology (High Electron Mobility Transistor on Gallium Nitride semiconductor). On top of these GaN transistors, a metallic network (also called interconnects) is fabricated to bring current and voltage to the source and drain electrodes of the components. These interconnects must withstand high current (100A), high voltage (600V), and elevated working temperatures (>250°C). Such operating conditions can lead to metallic reliability issues (diffusion, electro-migration?). This work is focused on the development of advanced interconnections for GaN power transistors and the study of their reliability. After an analysis of the specifications required for these interconnects (current density, electrical field, temperature, geometries...), the work will consist of identifying appropriate materials (eg: aluminium, copper?) and technologies to fabricate them. The interconnect structures produced will be characterized by several physico-chemical and electrical methods. Accelerated aging tests will help to study failure mechanisms. Through optimization of the technological process steps, this work will permit the fabrication of efficient and reliable interconnects.

Characterization and modelling of organic photodiodes : application to device's reliability

The field of organic electronics is a fast growing field of activities since the recent years. Initially pushed by the development for flat panel displays based on Organic Light Emmiting Diodes (OLED), the organic electronic field is now adressing new applications such as organic solar cells or organic photodetectors. This multi-disciplinary work that wioll be done in the frame of this thesis will be at the interface between material science and electronic. It aims at understanding physical phenomena that play a role for optimization of the printed organic photodiode and for understanding the failure mechanism as well as the shift in performances during aging. The main target is thus to understand and improve device's performances as well as to understand how these performances are shifting with device aging. This will be done by combining deep physical analysis of materials, opto-eletrical characterization and modelling work. The goal is to obtain an equivalent opto-eletrical model of the photodiode in order to analyse the electrical measurements and to optimize the devices, both in terms of device architecture and fabrication process.

Wireless and Privacy: Device Free Motion Detection into the wild

Wireless networks are nowadays deployed everywhere. If their primary goals is to transmit data, it seems possible to subvert radio waves for other purposes such as motion tracking or localization. During this thesis, we want to investigate how intrusive are the wireless indicators like the Received Signal Strength Indicator (RSSI) and Link Quality Indicator (LQI). How much information on a person can be inferred from these indicators ?

Synthesis and functionalization of 1D nanomaterials (metallic nanowires fabricated by solution synthesis and/or electrospinning). Fabrication of flexible transparent electrodes with low temperature printing techniques and integration into functional devices.

The transparent conductive layers currently used for photoelectric devices are essentially based on TCO (Transparent Conductive Oxides). These materials are made of indium, which is an expensive and poorly abundant metal, moreover ITO has no flexibility properties. For technical and economic reasons the development of new electrodes is the subject of considerable research. Some alternatives appear relevant today, including the use of conductive materials of nanometric size. For example, nanomaterials such as metallic nanowires, nanoparticles of zinc oxide or carbon nanotubes can be used for this purpose and the first results in this direction portend excellent development prospects. The metal nanowires appear particularly promising and will be the focus of this thesis. The study will address the entire process, from the synthesis of nanowires with different techniques up to their integration into functional materials.

Multi strained channels integration for 10nm FDSOI CMOS

This PhD thesis is focused on the evaluation of different strain techniques for boosting the electrical performance of FDSOI CMOS for sub 10nm generation nodes.

High quality graphene growth with or without laser assistance and transport properties evaluation for spintronic and other applications

The chemical vapour deposition (CVD) of graphene material is a key point for applications like advanced interconnects for future integrated circuits, transparent electrodes for solar cells or portable electronic devices and spintronic devices which uses spin current and might replace conventional CMOS circuits in future nano electronic components. All these applications are demanding in term of graphene quality especially in term of transport properties. These performances means that the best structural properties must be achieved on the grown graphene materials (low defect density and large flake size for example). The original way we want to develop in this PhD thesis is the assistance of pulsed UV laser during CVD growth in order to master the nucleation density and to achieve materials elaborated at a high equivalent temperature. The subject has already been started in the laboratory and the whole set up is already operational. The student will have to develop and optimize the process to achieve the best possible materials for the targeted applications. Particularly he will realize and characterize the spintronic devices to measure the mobility and the spin diffusion length. The other characterization tools will be used to understand the growth mechanisms and explore the electrical optical and mechanical properties of the materials. This work will be performed in collaboration between two laboratories of CEA Grenoble on the basis of 50% work on each laboratory.

Stydu of the acoustic answer of direct and temory bonding.

Direct and temporay bonding are more and more used by several type of applications in Microelectronic (SOI elaboration, Baside side imager, 3D technology?). It is then mandatory to be able to characterize, inline in a non-destructive manner, the bonding quality. Nowadays several techniques could be employed (Infra red microcopy, X-Ray observation?) but to be precise and fast enough, and with easy interpretation, only the reflexion acoustic microscopy is suitable. Moreover, this characterization technique allows to perform a 2D defect mapping which could be very useful for defect study. This characterization technique is already used for these bonding types but a deep understanding still needed. Indeed, fine interpretation and acoustic signal modelisation still not implemented. In a first time, very simple samples will be used to test acoustic answer models of direct/polymer bonding. Next, real inline production wafer of CEA and ST will be characterize and simulate in order to optimize the bonding process. This study will mainly be performed on longitudinal acoustic waves. But in parallel, bonding characterization using Rayleigh wave will be studied. This Rayleigh wave study should allow us to characterize the mechanical properties of the bonding interface and then maybe the bonding energy.

Development of nanocatalyst for PEM Fuel Cell

PEM fuel cell is a promise technology in the context of new energy development. Indeed, it allows feeding an electrical device for transport application (automotive) or stationary application (back-up) without any greenhouse gas emission. However, to reach the objectives of power requested by the utilisation, the use of catalyst made of platinum is required. Platinum is a noble and expensive metal which is a brake of the widespread dissemination of the technology. To support the dissemination of this technology, new synthesis of nanostructured catalyst has to be developed. This thesis is focused on the synthesis, characterisation and performance tests of new catalysts which contain less platinum than thus currently used. These catalysts based on nanostructured alloy will be supported on carbon nanotubes (or other nanostrutured carbon). The nanostructuration allows a better catalyst use and thus decrease the platinum loading of the active layer. Characterisation by electron microscopy (SEM, TEM) will allow validating the nanostructuration.

Modeling of diffraction effects to calculate echoes from the specimen geometry in ultrasonic NDT

The proposed thesis is part of the development at CEA LIST of integrated models in the CIVA software platform of expertise in Non Destructive Testing. The thesis precisely aims at extending the models of components inspection by ultrasonic waves, in order to develop by a ray method a complete simulation tool of the ultrasonic response of the specimen geometry. Complete inspection modeling requires simulating echoes from the geometry or structures of the inspected components. The existing ultrasonic field model is based on the ray formalism of geometrical acoustics. The objective of this thesis is to extend this model in order to deal with the diffraction from structures edges and with multiple interactions (specular reflections and diffractions). To model the diffraction from edges, ?uniform corrections? of the Geometrical Theory of Diffraction will be developed. The model will be validated by comparison with experimental results and with simulation results obtained by a finite elements method.

Characterization of nano-connectors by Scanning Tunneling Spectroscopy (STS) and application to the electronic characterization of atomic wires and circuits

This PhD thesis is proposed within the scope of a larger study on the integration and electrical characterization of nano-objects on silicon substrates (molecular logic gates, atomic circuits, biological objects). It more specifically focuses on the study of the size and electrical properties of the connecting path to these nano-objects. One element of this connecting path, called nano-connector, is based on doped silicon and is prone to geometry variation with thermal treatments, due to dopants lateral diffusion. This is a problem as it can generate leakage current, or short circuits between nano-connectors. This PhD thesis will thus aim at studying these nano-connectors with Scanning Tunneling Spectroscopy / Microscopy (STS/STM). The work can be divided into three parts. The first one is dedicated to the characterization of the nano-connectors dopants distribution, with low temperature STM/STS, in order to control their geometry. The second part deals with the nano-connectors electrical characterization and understanding of transport properties. A specific STM equipment will therefore be used. Finally, the objective of the third part is to propose a connecting path based on dopants and enabling the electrical connection of atomic wires, molecular logic gates, proteins?

Surface micro-nanostructuring by local enhancement of the laser field

The interaction between a light beam and particles of very small size is an area of innovation in the field of surface micro-nanostructuring. Indeed, under certain conditions of laser irradiation, there is formation of a photonic nanojet under the irradiated particle associated to a concentration of electromagnetic field. The local reinforcement of the laser field can lead to an ablation of the substrate, localized melting or curing of photosensitive species. The objective of this thesis is to control the formation and characteristics of photonic nanojet forming underneath different types of particles and in interaction with the substrate. Through this approach, generic processes of surface structuring, multi-material, multi-scale, multi-functional, large area and low cost are targeted. Work will take place at CEA/LITEN whose applications are related to the fields of energy and nanomaterials in close collaboration with the Hubert Curien laboratory for its expertise in the area of optics, modeling and lasers. Two research axes will be carried out at the theoretical and experimental levels, one aiming at structuring by removal of material and the other one by adding material on the host substrate.

Whole-body motion generation for an active exoskeleton

Active exoskeletons can be used in rehabilitation to help a person with motor impairments to move. The movements of the exoskeleton must be adapted to the person and to the exoskeleton. The aim of the thesis is to generate these movements for a four limbs (arms and legs) exoskeleton developed by CEA. The thesis will be based on path planning methods and techniques developed by Gepetto LAAS team, for generating dynamic movements. These methods consist in solving the system inverse dynamics under constraints, taking into account a stack of equalities and linear inequalities, using a digital hierarchical quadratic programming solver. They can generate complex multi-objective movements and may include several non-coplanar contacts with the environment, while considering the geometric and dynamic limits related to the task execution. These methods have been successfully implemented on the humanoid robot HRP-2. Adapting them to the exoskeleton will bring challenging problems ranging from dynamic modeling of the system, to its state observation and control. This will be done at LAAS, based on a close collaboration between experts from LAAS GEPETTO team and CEA Interactive Robotics Laboratory.

A methodology for verification of Cloud hypervisors by a combination of formal methods and testing

The CEA LIST, based in Palaiseau (near Paris, France), is a key software systems and technology research center working on embedded systems (architecture and design of systems, methods and facilities for software and system dependability, and intelligent vision systems), interactive systems, signal detection and processing in partnership with the major industrial players in the nuclear, automotive, aeronautical, defense and medical fields. The Software Safety Lab of the CEA LIST develops tools for verification and validation of software and software/hardware systems. One of the ongoing projects consists in developping a verified Cloud hypervisor. A first prototype called Anaxagoros has been developed by the Embedded Real-Time System Laboratory of CEA LIST. Security of Cloud hypervisors is one of the most challenging issues of Cloud Computing. This PhD aims to develop a methodology for verification of the implementation of critical components of Cloud hypervisors with the help of a combination of formal methods and testing. One of the difficulties is related to optimized code and data structures that often cannot be successfully verified by static analysis tools alone. However, optimized code can be executed, so testing methods can be applied in combination with formal methods to ensure the reliability of the hypervisor. Another important aspect of verification is modularity of the proof. This PhD will study a methodology of structuring the specification so that the impact of a source code modification remains limited and do not require a new verification (including specification and proof for the whole program) to be started from the beginning. A third research direction is related to hardware. This PhD will study how to take into account the properties of the hardware for a combined verification software/hardware. Verification tools developed at CEA LIST (Frama-C, PathCrawler) can be used during this PhD. The proposed methodology will be applied to the Anaxagoros hypervisor. The candidate will have Master's degree in Computer Science, and good knowledge of Software Verification and Validation.