DYNAMAG: A Project on Computational Magnetics

From the ancient lodestone to Maxwell’s equations, from the first electrical machines to magnetic hard disks, man’s knowledge about the fascinating properties of magnetic materials and fields has made great strides over the centuries. Today’s emerging fields of research are spintronics and magnonics, which deal with magnetism at the electronic level.

The Electrical Engineering Department of IIT Madras is taking part in a collaborative project on computational magnetics called DYNAMAG. The project is a joint effort by Indian and European universities including IIT Madras, IIT Delhi, IISe Bangalore, SN Bose National Centre for Basic Sciences, University of Exeter and University of Southampton, among others.

Two emerging fields deal with the spins of electrons in magnetic materials – spintronics, and magnonics.

Spintronics is the exploitation of the spin of the electron and its associated magnetic moment for various applications including the encoding of binary information with electron spins. Electrons have two spins (‘up’ and ‘down’), which is well suited to encoding binary data. Recent developments like the Magnetic RAM indicate that conventional data storage could potentially be replaced by spintronic technology which has the advantage of higher speeds.

 

Magnonics: the theory

Magnonics is the second emerging field, and it deals with ‘spin waves’ which are, in layman’s terms, magnetic waves that propagate in the magnetic material. Guru Venkat, a Ph.D. student who is a part of DYNAMAG at the department of EE, IITM, explains, “When one applies an external magnetic field to a magnetic dipole, it begins to precess, i.e., it wobbles about its axis. When the same principle is applied to an ordered array of magnetic moments (which may be found in a ferromagnetic material, for instance), all of them precess, and the moment vector traces a wave, known as a spin wave.

This is the classical picture, and deeper theory lies in quantum mechanics. The frequency of precession determines the frequency of the spin wave.”

Spin waves were discovered more than 40 years ago and differ from conventional EM waves in some important respects. The word ‘magnonics’ is derived from ‘magnon’, which is the quantum analog of a spin wave, just as a photon is the quantum analog of the light wave. Spin waves propagating in magnetic waveguides could be potentially used to store and transfer data. Magnonics has made it possible for researchers to contemplate a transistor-less logic circuit in the future, where magnonics and spin-wave phenomena are used to implement logic circuitry.

The frequencies of spin waves can be in the Gigahertz range, while their wavelengths are in the nanometer range. Thus, they travel at a speed significantly lower than that of light. “The gigahertz frequency range of spin waves, coupled with their low wavelength in nanometres, which will keep device size small, is one of the advantages of magnonics. This is in contrast to photonics, where devices are in the millimetre range. The other advantage of magnonics is that it offers a further degree of control. By changing the magnetic field, the characteristics of the spin wave can be changed,” adds Guru Venkat.

If the precession is damped, the radius of motion of the magnetic moment of the dipole decreases with time until the moment is aligned with the magnetic field . For sustained spin waves, materials which show low damping are chosen, and currently, one field of active research in magnonics is the identification of materials which exhibit desirable damping characteristics 1n the nanometer wavelength range.

The second active component of research in magnonics is that of engineering and probing magnonic waves, and examining phenomena like interference, diffraction, modulation and dispersion in these waves.

 

Project DYNAMAG

Project DYNAMAG aims to develop a theoretical and numerical framework for the analysis of magnonic phenomena. Says Guru Venkat, “We are coming up with tools to analyse spin waves computationally. Precession is governed by a differential equation, and the solution of the equation can be obtained by numerical computer methods. There is a trade-off between complexity and computational power, the resources available, and the size of geometry and duration of the simulation.”

He continues, “My part of the project was to simulate dispersion in spin waves. Just as different frequencies of light travel with different velocities in a medium, so do different components of spin waves travel differently in different media. I had to simulate dispersion, and obtained dispersion diagrams for different geometries.” Once the dispersion characteristics are known, one can then choose frequency components that travel properly in the desired medium.

Next, Guru Venkat plans to test the results of the simulation with measurements to see how well the predictions correlate to observations. He says, “If there is a significant correlation, it would help researchers. In the future, they could run simulations before taking measurements, and the simulations would tell them what to expect in the measurements. This would be useful because the measurement is always very difficult, and time-consuming.”


guru_venkatGuru Venkat is a Ph.D. scholar at the Department of Electrical Engineering, IIT Madras. He is a resident of Sindhu hostel.