AMBER researchers have identified 22 new magnets and a new discovery procedure, facilitating the fast discovery of new advanced materials for applications ranging from electronics to aerospace. Prof Stefano Sanvito reports

The love affair between mankind and magnetic materials has been continuing for two thousand years and does not show any sign of losing steam. From the invention of the compass, which enabled the 11th-century Chinese dynasty of Song to embark on long-distance navigation, magnetic materials have been key to developments in human technology.

Today, magnets are everywhere in our life. The hard disk of our computer is composed of billions of tiny magnets and the device that reads and writes the data on the disk (the read/write head) is made of a complex magnetic nanostructure.

Wind turbines are made of strong permanent magnets, as are the electrical motors in our cars, in our kitchen blenders and in our lawn mowers.

Yet, nature is stingy when it comes to magnets. The first report of a magnetic material dates back to 79AD and it is contained in what was probably the first scientific encyclopaedia ever, the Naturalis Historia (Natural History) from the Roman natural philosopher Plinius the Elder.

History of magnetism

In the Historia, Plinius, who perished in the eruption of the Vesuvio in 79AD, reports the legend of Magnes the shepherd, who discovered two metallic stones attracting each other. These were likely made of magnetite (Fe3O4), a mineral abundant in the earth crust and one of the main iron ores.

It took about 2,000 years to understand the microscopic origin of magnetism, which was possible only after the formulation of quantum mechanics. During all this time, we have discovered about 2,000 materials that behave as magnets. In other words, our discovery speed is one magnet per year. Interestingly, only a handful of them are used in mainstream applications [1].

Why it is so difficult to discover new magnets, and even more difficult to find new useful ones? Mostly because magnetism is a relatively rare event in the world of materials. For a compound to be magnetic, it must contain some ions that are able to sustain a spontaneous magnetic moment. This requires a partially filled electronic shell where the Hund’s coupling is strong.

Thus, with very few exceptions, only 3d transition metals (mostly Mn, Fe, Co and Ni), rare earths (e.g. Gd, Nd) and some 4d ions (e.g. Ru) can give rise to magnetism. Then, when magnetic ions feature in a compound, they also need to interact in such a way that their magnetic moments align with respect to each other.

Microscopic mechanisms to achieve alignment

There are several microscopic mechanisms to achieve such alignment, depending on whether the material is a metal or an insulator. However, there are many cases where no mechanism is at play or it is not strong enough and even the presence of magnetic ions does not produce magnetism. Thus, of the 100,000 inorganic materials documented, only about 2,000 are magnetic.

For a magnet to be useful, many other properties need to be present in the same material. In general, different applications have different requirements so a magnet that is good for one application (e.g. an electrical motor) may not be good for another one (e.g. an electrical transformer). However, there are several features that are common to all `useful’ magnets.

Firstly, a magnet should have a critical temperature approximately 100C above the temperature one wants to use it. The critical temperature, known also as the Curie temperature, is the temperature above which the magnet loses its magnetic properties, namely its magnetisation vanishes.

Thus, if one wants to use the magnet as the media in a hard-disk drive its critical temperature should be at least 200C, if instead it is part of an electrical motor in a hybrid car, the critical temperature should not be below 300C.

Secondly, in many applications, magnets are like commodities: the electric motor of a car contains 2kg of magnets, a wind turbine about 250Kg per MW. For this reason, they must be cheap. They should not contain toxic, radioactive or explosive elements.

Finally, if one looks at a specific application, additional properties need to be present. In an electrical motor, for instance, both the magnetisation and the anisotropy must be large (these are characteristics that establish how strong a magnet is).

In a hard-disk drive, the anisotropy should be large enough for the data not to be spontaneously erased, but not too large or it would be impossible to write the data. When one combines all these requirements, it is not too surprising that only a couple of dozen magnets are used in modern technology.

Speed of discovery of new magnets

In an article just appeared in Science Advances [2], we have demonstrated that the speed of discovery of new magnets can be accelerated by a factor of twenty. We have developed a computational protocol, already largely used in the pharmaceutical industry for drugs discovery, to identify novel promising magnetic compounds.

This consists of generating an extremely large database containing the properties of 300,000 new hypothetical materials, calculated with advanced electronic structure methods. Then, we evaluated the possibility of fabricating such compounds.

This is achieved by evaluating the stability phase diagram of the newly predicted magnets, an operation that requires the evaluation of all possible decomposition products. The result of such a selection process is knowing which material has a high chance of being made, this knowledge is then passed to our experimental colleagues.

Out of the 300,000 prototypes, we have evaluated for stability about 40,000 and predicted 20 to be stable and magnetic. Then, our experimental colleagues attempted the fabrication of four, succeeding in two cases. One of the newly made magnets, Co2MnTi, displays an ordering temperature of about 940C.

This is a remarkable discovery in itself, since no more than two dozen among the previously known magnets remain magnetic at such a high temperature. Most importantly, the protocol designed by our team can be applied to other material classes beside magnets (e.g. thermoelectrics, light absorbers, superconductors, etc.), effectively opening a new avenue for fast materials discovery [3], of potential use to a wide range of industries.

The discovery of new magnets is important because they form part of everyday applications, from computers, to wind turbines, the electrical motors in our cars, kitchen blenders and lawn mowers. However, there are several technologies for which we still need to find the ideal magnet, which could provide for example more energy-efficient non-volatile magnetic storage, such as hard discs and more energy efficient motors in hybrid cars.

Prof Stefano Sanvito, School of Physics, AMBER Centre and CRANN Institute, Trinity College Dublin

[1] J.M.D. Coey, Magnetism and Magnetic Materials. Oxford University Press, (Oxford, 2009).
[2] S. Sanvito, C. Oses, J. Xue, A. Tiwari, M. Zic, T. Archer, P. Tozman, M. Venkatesan, J.M.D. Coey and S. Curtarolo. ‘Accelerated discovery of new magnets in the Heusler alloy family.’ Science Advances  14 Apr 2017: Vol. 3, no. 4, e1602241. DOI: 10.1126/sciadv.1602241
[3] S. Curtarolo, G.L.W. Hart, M.B. Nardelli, N. Mingo, S. Sanvito and O. Levy. ‘The high-throughput 
highway to computational materials design.’ Nature Materials 12, 191-201 (2013). O'RiordanElecAMBER,research,TCD
The love affair between mankind and magnetic materials has been continuing for two thousand years and does not show any sign of losing steam. From the invention of the compass, which enabled the 11th-century Chinese dynasty of Song to embark on long-distance navigation, magnetic materials have been key to...