Last author

Magnetoelectric materials are very rare. Their magnetic poles switch around when you apply a voltage, and their electrical charge direction is reversed when you apply a large magnetic field. Since the 1960s, a small number of physicists and engineers have sought to use magnetoelectrics in computers. Random access memory (RAM) based on magnetoelectric components would be faster than some types of currently available RAM, and more durable than others. These properties would allow increased computer speed with low power consumption. James Scott, a physicist at the University of Cambridge, UK, reviews the field on page 759 of this issue. He tells us that it has produced some beautiful physics, but that, contrary to considerable hype, its potential for computer applications remains limited.

What's the history of the field?

It began in about 1921, but by the mid-1920s ferromagnetics and ferroelectrics research by a French scientist using nickel and iron had been proved wrong. This cast a cloud over the field for 30 years. In the 1960s, the research picked up again in Moscow, and a young Swiss scientist named Hans Schmid, based in Geneva, found that the best magnetoelectric materials, known as boracites, were expensive and worked only at very low temperatures.

What sparked the field's resurgence?

There was a renaissance three years ago. Researchers wondered why magnetoelectric materials hadn't been commercialized in the 1980s, because they thought they could be used to create a great computer memory. But the magnetoelectric effect works only at very low temperatures and is too weak to be used for error-free memories, and the materials are expensive.

Are today's materials better?

New materials — manganites and terbium compounds — have been found, but they are no better for use in computer memory than the boracites.

So what practical value does the work have?

One area of interest is sensors for measuring ultra-low magnetic fields. These can be used to find underground deposits of minerals or oil, and for highly sensitive lab experiments in solid-state physics.

It sounds like you put a lot of value on an end product.

I do. I think it's important for modern science to have applications in society and industry, to create jobs and benefit the economy.