Final Report Summary - ATOMICFMR (FERROMAGNETIC RESONANCE AT THE ATOMIC SCALE)
The Marie Curie IEF funded research project AtomicFMR was focused on the interaction between dynamical magnetic properties and transport in nanostructures. In particular, electrically detected ferromagnetic resonance of different nanostructures, including nanocontacts, and dynamical coupling through a nanocontact were investigated.
The principle used in these experiments is the electrical detection of the dc voltage signal generated by the rectification of the induction current in the conducting sample. This dc voltage is caused by the spin-rectification effect that couples the induced rf current in the circuit containing the magnetic nanostructure to any change in resistance at this frequency. In our samples the resistance variation at the rf frequency arises from the anisotropic magnetoresistance, that oscillates due to the precession of magnetization.
To further enhance the sensitivity of the electrical detection of the dynamical properties, we modulate the rf field by a square signal and measure the voltage generated across the magnetic nanostructure at the modulation frequency.
In the presence of nanoconstrictions or atomic contacts, this voltage is dominated by the dynamical behavior of the few atoms composing the contact.
First of all, an optimization of the samples synthesis was performed, including the preparation of the flexible kapton substrates, the UV lithography to prepare the rf antennas and the electron beam lithography technique to design and prepare the nanostructures close to the rf antennas. We have fabricated magnetic nanostructures using permalloy as the magnetic material because it exhibits a sizable anisotropic magnetoresistance contribution and have nearly zero magnetocrystalline anisotropy and magnetostriction, hence low damping.
All micromagnetic simulations in this project have been performed using the code MuMax2 since is freely available and besides it runs on the graphical processing unit, so it is much faster than codes running on the traditional central processing unit.
We have performed comprehensive measurements on several 10 micron long, 1 micron wide and 20 nm thick Py stripes with a 60 nm wide notch, which are positioned with electron beam lithography 500 nm away from the current line of the rf antennas. After performing measurements on several permalloy notched stripes and micromagnetic simulations using MuMax2, we have demonstrated in the frame of this AtomicFMR project that internal domain wall resonances can be detected by an electrical technique. This electrical detection is based, as described above, on the voltage signal generated by rectification of the induced current coupled to the AMR oscillation. This rectification technique has proven to be extremely sensitive as it has allowed us to measure signals from very localized regions within a single pinned domain wall. Micromagnetic simulations have enabled a full understanding of the measured voltage signal and the proper identification of the different resonances we observe.
Our findings have demonstrated for the first time that ferromagnetic domain walls in a nanostructure provide a sufficiently varying internal field so that some resonance can always occur below a cutoff frequency of 6 GHz.
To study in depth the relationship between magnetization dynamics and electronic transport, we have investigated the influence of a dc current on the domain wall resonances by injecting different dc currents into the sample while measuring the rectified voltage. As the current density increases above 8×105 A cm-2, the nucleation peak broadens demonstrating that spin torque starts to play a role on the domain wall ferromagnetic resonance at current densities two orders of magnitude lower than the depinning ones. Therefore the sensitivity of the rectified voltage measurement allows us to conclude that the domain wall resonance behavior is modified at rather low current densities.
Furthermore, preliminary results have been obtained as of the end of this project in the atomic contact limit: Design and fabrication of the samples are completed and the micromagnetic simulations of the behavior of the electrodes when breaking have also been calculated. For a full understanding of the atomic limit behavior, some more measurements are needed and are now being performed by Prof. Michel Viret, the project’s coordinator.
The second theme investigated in the frame of this project is the dynamical coupling of two nanostructures through a nanocontact. In particular, the dynamical coupling of two permalloy discs linked by contact of few tenths of nanometers in size has been studied. First by performing dynamical micromagnetic simulations using the MuMax2 code, and then by performing the rectified voltage measurements as the link between both discs is broken.
As predicted by dynamical micromagnetic simulations, that have allowed us to properly identify the different resonant modes, when breaking the contact between both discs we passed from a situation (at zero static magnetic field) in which both discs were exchange coupled to a situation in which the coupling is dipolar and there is a thin domain wall at the contact.
The results of this project have made a significant contribution in the state of the art of low-dimensional spintronics, a strategic field of research in Solid State Physics owing to its great variety of applications in information technology, thus contributing to enhance European research excellence.
The principle used in these experiments is the electrical detection of the dc voltage signal generated by the rectification of the induction current in the conducting sample. This dc voltage is caused by the spin-rectification effect that couples the induced rf current in the circuit containing the magnetic nanostructure to any change in resistance at this frequency. In our samples the resistance variation at the rf frequency arises from the anisotropic magnetoresistance, that oscillates due to the precession of magnetization.
To further enhance the sensitivity of the electrical detection of the dynamical properties, we modulate the rf field by a square signal and measure the voltage generated across the magnetic nanostructure at the modulation frequency.
In the presence of nanoconstrictions or atomic contacts, this voltage is dominated by the dynamical behavior of the few atoms composing the contact.
First of all, an optimization of the samples synthesis was performed, including the preparation of the flexible kapton substrates, the UV lithography to prepare the rf antennas and the electron beam lithography technique to design and prepare the nanostructures close to the rf antennas. We have fabricated magnetic nanostructures using permalloy as the magnetic material because it exhibits a sizable anisotropic magnetoresistance contribution and have nearly zero magnetocrystalline anisotropy and magnetostriction, hence low damping.
All micromagnetic simulations in this project have been performed using the code MuMax2 since is freely available and besides it runs on the graphical processing unit, so it is much faster than codes running on the traditional central processing unit.
We have performed comprehensive measurements on several 10 micron long, 1 micron wide and 20 nm thick Py stripes with a 60 nm wide notch, which are positioned with electron beam lithography 500 nm away from the current line of the rf antennas. After performing measurements on several permalloy notched stripes and micromagnetic simulations using MuMax2, we have demonstrated in the frame of this AtomicFMR project that internal domain wall resonances can be detected by an electrical technique. This electrical detection is based, as described above, on the voltage signal generated by rectification of the induced current coupled to the AMR oscillation. This rectification technique has proven to be extremely sensitive as it has allowed us to measure signals from very localized regions within a single pinned domain wall. Micromagnetic simulations have enabled a full understanding of the measured voltage signal and the proper identification of the different resonances we observe.
Our findings have demonstrated for the first time that ferromagnetic domain walls in a nanostructure provide a sufficiently varying internal field so that some resonance can always occur below a cutoff frequency of 6 GHz.
To study in depth the relationship between magnetization dynamics and electronic transport, we have investigated the influence of a dc current on the domain wall resonances by injecting different dc currents into the sample while measuring the rectified voltage. As the current density increases above 8×105 A cm-2, the nucleation peak broadens demonstrating that spin torque starts to play a role on the domain wall ferromagnetic resonance at current densities two orders of magnitude lower than the depinning ones. Therefore the sensitivity of the rectified voltage measurement allows us to conclude that the domain wall resonance behavior is modified at rather low current densities.
Furthermore, preliminary results have been obtained as of the end of this project in the atomic contact limit: Design and fabrication of the samples are completed and the micromagnetic simulations of the behavior of the electrodes when breaking have also been calculated. For a full understanding of the atomic limit behavior, some more measurements are needed and are now being performed by Prof. Michel Viret, the project’s coordinator.
The second theme investigated in the frame of this project is the dynamical coupling of two nanostructures through a nanocontact. In particular, the dynamical coupling of two permalloy discs linked by contact of few tenths of nanometers in size has been studied. First by performing dynamical micromagnetic simulations using the MuMax2 code, and then by performing the rectified voltage measurements as the link between both discs is broken.
As predicted by dynamical micromagnetic simulations, that have allowed us to properly identify the different resonant modes, when breaking the contact between both discs we passed from a situation (at zero static magnetic field) in which both discs were exchange coupled to a situation in which the coupling is dipolar and there is a thin domain wall at the contact.
The results of this project have made a significant contribution in the state of the art of low-dimensional spintronics, a strategic field of research in Solid State Physics owing to its great variety of applications in information technology, thus contributing to enhance European research excellence.