Quantum Magnetoelectronic Phenomena
|
Magnetoelectronic phenomena such as giant magnetoresistance ,
tunnel magnetoresistance , and spin transfer torque ,
play a central role in the existing and emerging applications of magnetic systems for sensing, memory, and information processing. These phenomena can be interpreted in terms of the interactions between
conduction electrons and magnetization, with the latter usually thought of as a classical vector field. The central question we are exploring in our experimental and computational studies of quantum magnetoelectronic phenomena is whether there
are new effects, or important contributions to the existing effects, that arise specifically from the non-classical aspects of magnetism. In a 2017 Phys. Rev. Lett. ,
we demonstrated a contribution to the spin transfer effect arising from the quantum fluctuations of magnetization. Recently, we used a fully quantum simulation of spin transfer to show that the laws
of energy and momentum conservation govern the characteristics of magnons (spin wave quanta) generated in the spin transfer process, and also the orbital electron scattering associated with spin transfer.
|
Current-induced nonequilibrium phonon states
|
Joule heating is an increase of temperature of materials when electrical current flows through them, as explained on Wikipedia and in the textbooks. The question we are adressing in our research is whether the
lattice vibrations induced by electric current can be actually described by a temperature - in other words, whether the lattice remains in a thermalized (equilibrium) state, albeit at an elevated temperature. A related question is - when electronic nanostructures or nanodevices
are destroyed by large currents, can these catastrophic processes be described as a "burn-out", or is the dynamics of destruction more complicated? In a 2020 Phys. Rev. X paper, we showed that the dependence
of resistance of a variety of metallic nanostructures on current indicates a strongly non-equilibrium distribution of current-generated phonons that cannot
be described by a temperature. We are developing electronic spectroscopy techniques to learn more about the current-driven phonon states, and about the mechanisms underyling the destruction of
nanodevices by high currents.
|
Complex states of matter for neuromorphic applications
|
Artificial neuromorphic systems - hardware emulating the functionality of the brain - may transform the society and change the ways we interact with each other and with the world.
At the heart of the brain functionality is a network of neurons extensively connected by synapses. Viewed as a material, this system is characterized by a multitude of metastable states.
Transitions among very different states can be triggered by small perturbations. From this perspective, brain is reminiscent of glasses - complex disordered materials that have a multitude of very long-lived metastable
states. If a glass is brittle, it can be shattered by even small perturbations, resulting in avalanches leading to a very different state. We are developing
complex materials and devices based on them that take advantage of these similarities to achieve neuromorphic functionalities. In a series of experimental papers (PRB 2015,PRB 2016,
PRB 2018, JMMM 2019), we showed that thin-film magnetic heterostructures
formed by materials with incompatible magnetic ordering (e.g., ferromagnets and antiferromagnets) can be engineered to become spin glasses with properties, such as the glass transition temperature, readily controllable by their structure.
Our calculations (see arXiv:2006.07996) show that these structures can be utilized to build one of the central elements of artificial neuromorphic networks - an "ideal" memristor.
Our research focuses on the experimental demonstration of such devices, studying and optimizing the materials utilized for their operation, and extending the functonality beyond memristance.
|
|
