Thèse Rf-Scope Caractérisation et Outils de Simulation Optimisé pour les Détecteurs Rf H/F - Doctorat.Gouv.Fr

Établissement : Université Grenoble Alpes École doctorale : EEATS - Electronique, Electrotechnique, Automatique, Traitement du Signal Laboratoire de recherche : Centre de Radiofréquences, Optique et Micro-nanoélectronique des Alpes Direction de la thèse : Pietro MARIS FERREIRA ORCID 0000000200389058 Début de la thèse : 2026-10-01 Date limite de candidature : 2026-08-31T23:59:59 Cette recherche doctorale porte sur la modélisation multiphysique et l'optimisation des composants à semi-conducteurs III-V (GaN, GaAs) pour les réseaux 5G et futurs 6G. Afin d'atteindre une efficacité élevée (100 Gbits/J) et des fréquences de l'ordre des THz. Le projet se base sur un solveur Monte Carlo pour simuler des effets nanométriques tels que le transport balistique, la formation du 2DEG ou encore l'auto-échauffement (points chauds). Les objectifs du travail de recherche doctoral sont : l'optimisation des performances du solveur via Python/C++/Fortran, l'intégration des parasites électromagnétiques (HFSS/Q2D) aux simulateur, le développement d'un banc de test expérimental jusqu'à 110 GHz pour valider le flux de simulation et la conception de détecteurs RF innovants. In the industrial telecommunications sector, for 5G (and future 6G) network architectures, performance and energy autonomy are the two primary criteria for electronic component design. Signal detection by the receiver must therefore be performed by integrated diodes or transistors featuring ultra-wide bandwidth and low power consumption.
III-V Schottky diodes and HEMT (High Electron Mobility Transistor) technologies, in Gallium Nitride (GaN) or Gallium Arsenide (GaAs), are at the heart of current research aimed at reaching cutoff frequencies exceeding 1 THz and energy efficiency levels on the order of 100 Gbits/J. For comparison, Wi-Fi operates at a few GHz with an efficiency of 200 Mbits/J. GaN HEMTs, which offer very high power density and withstand junction temperatures above 200°C, are used in the power amplification (PA) stage of base station transmitters (Macro-cells) [Liu2026]. Their level of miniaturization also enables the creation of active antenna arrays (Massive MIMO) that direct the signal precisely toward the user (Beamforming) by integrating hundreds of small amplifiers directly behind each radiating element of the antenna.GaAs HEMTs (pHEMTs) are predominantly used in mobile devices to switch between different frequency bands (4G, 5G, Wi-Fi). Their low insertion loss helps preserve battery life. Finally, Schottky diodes are utilized as RF power detectors, and innovative applications are also considering the use of HEMT components as RF detectors at the front end of receiver stages [Artillan 2024].
The design of these III-V semiconductor high-frequency (RF) transistors requires the use of a Monte Carlo solver to understand electronic transport at the nanometric scale. Unlike classical models that treat electrons as a fluid, the Monte Carlo method tracks thousands of individual particles subjected to random collisions, making it possible to capture complex physical phenomena. The key phenomena simulated include the creation of the two-Dimensional Electron Gas (2DEG) at the heterojunction, the saturation velocity (predicting the maximum speed at which electrons cross the channel under an intense electric field), the ballistic transport (in ultra-short components, electrons cross without any collisions), valley transfer or Gunn effect (typical of GaAs, where electrons 'jump' to higher energy valleys, creating negative differential resistance used to generate microwaves; this modeling is essential for dimensioning RF sources). Only Monte Carlo captures this 'shoot-through' effect, which drift-diffusion models ignore. Furthermore, self-heating analysis can be modelled [Garcia2025]: a complete solver couples electron transport with phonon transport (lattice vibrations), specifically to identify 'hot spots' near the gate. Since 2004, University of Salamanca (Spain) and, since 2022, CROMA laboratory join their efforts to develop such a solver [Mateos2004, Sanchez2022, Garcia2026].
Nevertheless, several identified challenges remain to be addressed. The primary issues include optimizing the solver's server-side computation time, incorporating access parasitics obtained through electromagnetic simulation, and developing a comprehensive component simulation method. Additionally, the experimental validation of the entire simulation flow must be carried out.
To achieve the ambitious target, the doctoral research involves advanced multi-physics modeling -encompassing semiconductor physics, thermal and electrical domains- to optimize the integration of sensing and amplification stages.
Doctoral Research Program (36 months):
1) State of the Art: (0M-6M) comprehensive literature review of III-V technology detectors (architectures and performance metrics).
2.a) Electromagnetic Simulation (0M-6M): mastery of EM simulation suites (HFSS, Q2D Extractor, etc.) and extraction of parasitic access elements for HEMT (High Electron Mobility Transistor) structures.
2.b) Monte Carlo Solver Integration (6M-24M): hands-on development with the homemade Monte Carlo solver and coupling electromagnetic parasitic data with transport simulations.
2.c) Computational performance optimization (6M-24M): enhancing solver efficiency and scalability through high-performance programming (Python, C++, and Fortran).
3.a) Experimental benchmarking (3M-12M): development of a characterization testbed for RF detectors operating up to 110 GHz.
3.b) Experimental validation (12M-36M): empirical validation of the end-to-end simulation workflow.
4) Design and simulation of innovative structures (24M-36M): conceptualization and modeling of novel RF detectors and HEMTs fabricated by CROMA's strategic partners.

The candidate profile required for the project is a young professional holding a master's degree in Electrical or Electronics Engineering, interested in the scientific field of microwaves technologies. He/She must be motivated, passionate about research in a multidisciplinary field and an organized person using scientific methods. He/She must justify good academic tracks in maths and applied physics; an experience in programming; linguistic competence in English (C1 written and spoken).