Description

RESEARCH OPPORTUNITY: The electronic von Neumann architecture has today reached its energy limits, hindering the deployment of high-performance AI at the edge. To overcome this bottleneck, neuromorphic electronics based on Spiking Neural Networks (SNNs) and the In-Memory Computing (IMC) paradigm have emerged as the leading solution. While the community has so far favored resistive IMC using memristors, this approach faces physical barriers when scaling up: high static currents, Joule heating losses, leakage, and voltage drops (IR drop). This PhD project proposes a paradigm shift by turning to capacitive IMC via voltage summation. Thanks to the emergence of ferroelectric capacitive memories (memcapacitors), this approach offers an ultra-low-energy alternative, free from the imperfections of the resistive model and well-suited for SNN applications. We are seeking a motivated PhD candidate passionate about materials science, microelectronics, and artificial intelligence to develop high-performance memcapacitors and exploit them as artificial synapses for capacitive neuromorphic computing enabling ultra-low-energy AI.

 

SCIENTIFIC CHALLENGES: Within an ecosystem of experts, you will lead the following research axes:

• Microfabrication & material innovations: Development of cleanroom processes to fabricate memcapacitors in crosspoint and crossbar architectures. You will explore doped HfO₂ (Gd, Zr, La) as the reference ferroelectric layer and optimize performance (memory window, multi-level capacitance programming, endurance) through thin-film engineering (interlayers, superlattices) and electrode design (Ru, Mo).
• Advanced characterizations: Memcapacitors will be studied using advanced physico-chemical (XRD, XRR, XPS, TEM) and electrical characterization techniques. You will develop semi-automated characterization protocols (I-V, C-V, pulsed) on samples and 300 mm wafers. You will analyze the write and read variability of capacitance levels, a critical factor for physical SNN performance.
• Contribution to neuromorphic demonstrators: You will actively participate in the development of capacitive neuromorphic circuits and demonstrations of capacitive IMC (charge or voltage summation). This work will be part of a collective dynamic, involving constant interaction with other team researchers (PhD students, postdocs, interns) working on system integration and modeling.

 

AN EXCEPTIONAL ENVIRONMENT IN EUROPE: The PhD will take place at the Laboratory of Microelectronics Technology (LTM), a core member of the LabEx Microelectronics, located in the heart of Europe’s largest R&D hub for nanotechnologies.

• Technology platforms: You will benefit from direct access to state-of-the-art cleanrooms at the Upstream Technological Platform (PTA) and CIME Nanotech, working on fabrication lines up to 200 mm, a rarity in the academic world.
• Network & collaborations: Your work will leverage strategic collaborations with laboratories within the LabEx (CEA-Leti, CROMA, TIMA, LMGP) for process development, advanced characterizations, and neuromorphic circuit design.
• Excellence training: You will develop a unique hybrid profile (materials, fabrication, neuromorphic electronics) within a world-renowned interdisciplinary ecosystem that fosters innovation and scientific excellence. Includes the opportunity for a 6-month internship in another academic or industrial research laboratory specializing in ferroelectric devices and neuromorphic electronics.

Skills required

Requires a Master’s degree or Engineering Diploma in Microelectronics, Materials Science, or Nanotechnologies. You demonstrate academic excellence and confirmed experimental skills in cleanroom processing and/or electrical characterization. Scientific rigor, autonomy, curiosity and fluent technical English are required. A strong interest in emerging memories and brain-inspired technologies for low-power AI is essential.

Keywords

ferroelectric memories, neuromorphic microelectronics, microfabrication, applied physics, AI