Spin electronics, thanks to the additional degree of freedom provided by electron spin, enables the deployment of a rich physics of magnetism on a small scale, but also provides breakthrough technological solutions in the field of microelectronics (storage, memory, logic, etc.) as well as for magnetic field measurement.
In the field of life sciences and health, giant magnetoresistance (GMR) devices have demonstrated the possibility of measuring the very weak fields produced by excitable cells on a local scale (Caruso et al, Neuron 2017, Klein et al, Journal of Neurophysiology 2025).
Measuring the information contained in the magnetic component associated with neural currents (or magnetophysiology) can, in principle, provide a description of the dynamic, directional and differentiating neural landscape. It could pave the way for new types of implants, thanks to their immunity to gliosis and their longevity.
The current bottleneck is the very small amplitude of the signal produced (<1nT), which requires averaging the signal in order to detect it.
Tunnel magnetoresistances (TMR), in which a spin-polarised tunnel current is measured, offer sensitivity performance that is more than an order of magnitude higher than GMR. However, they currently have too high a level of low-frequency noise to be fully beneficial, particularly in the context of measuring biological signals.
The aim of this thesis is to push back the current limits of TMRs by reducing low-frequency noise, positioning them as break sensors for measuring very weak signals and exploiting their potential as amplifiers for small signals.
To achieve this objective, an initial approach based on exploring the materials composing the tunnel junction, in particular those of the so-called free magnetic layer, or on improving the crystallinity of the tunnel barrier, will be deployed. A second approach, consisting of studying the intrinsic properties of low-frequency noise, particularly in previously unexplored limits, at very low temperatures where intrinsic mechanisms are reached, will guide the most promising solutions.
Finally, the most advanced structures and approaches at the state of the art thus obtained will be integrated into devices that will provide the building blocks for going beyond the state of the art and offering new possibilities for spin electronics applications. These elements will also be integrated into systems for 2D (or even 3D) mapping of the activity of a global biological system (neural network) and for evaluating capabilities for clinical cases (such as epilepsy or motor rehabilitation).
It should be noted that these improved TMRs may have other applications in the fields of physical instrumentation, non-destructive testing, and magnetic imaging.