Maryland Physicists Stacked a Two-Dimensional Magnet on a Two-Dimensional Ferroelectric and Got a Room-Temperature Multiferroic That Switches Its Magnetism With a Small Voltage, 3,000 Times Without Wearing Out.
Neither long-range magnetic order nor ferroelectric order survives easily in a sheet a few atoms thick, and getting both to coexist at 19°C was thought to be extraordinarily hard. Cheng Gong's team did it by building each order in its own layer and pressing the layers together.
Researchers at the University of Maryland have built a two-dimensional material that is both a ferromagnet and a ferroelectric at room temperature, and in which a small applied voltage flips the magnet between two stable states that persist after the voltage is removed. The work, published in Science, is a demonstration of a room-temperature multiferroic in a van der Waals heterostructure, a stack of atomically thin flakes held together by weak surface forces, and it points toward memory and sensor devices that store magnetic information using almost no current.
Multiferroics are materials that carry two or more "ferroic" orders at once, most usefully a permanent magnetization and a permanent electric polarization. When the two are coupled, an electric field can rewrite a magnetic bit, which is far cheaper in energy than driving a current through a wire to generate a magnetic field. Engineers have wanted such materials for decades for low-power non-volatile memory, spintronic logic, neuromorphic hardware and ultra-sensitive field sensors. The obstacle has always been that the conditions favoring magnetism and the conditions favoring ferroelectricity tend to exclude each other, and in two-dimensional sheets, where thermal fluctuations can destroy long-range order entirely, the problem gets worse.
"Long-range ferroic orders such as ferromagnetic order and ferroelectric order are not easy to be achieved in 2D systems," Cheng Gong, the paper's senior author, told Phys.org. "Integrating several orders together in a 2D system is even harder. Realizing multiple ferroic orders in a 2D system, under the threat of thermal agitations at room temperature, is imaginably challenging. However, once achieved, the reward would be big." His team's strategy was to stop trying to force both orders into one crystal. "Our approach is to realize 2D ferromagnetic order and 2D ferroelectric order in two separate 2D layers and then merge them together to form a heterostructure," he said.
The magnet is triiron gallium ditelluride, Fe3GaTe2, one of the few van der Waals materials that stays ferromagnetic above room temperature. The ferroelectric is copper indium thiophosphate, CuInP2S6, whose copper ions can be shifted up or down through the layer by an electric field. The team peeled thin flakes of each from bulk crystals, laid the CuInP2S6 on top of the Fe3GaTe2 and added a transparent indium tin oxide electrode above and a chromium-gold electrode below so a voltage could be applied straight through the stack. At 292 kelvin, about 19 degrees Celsius, the structure showed both orders simultaneously.
Applying a small voltage switched the stack between states with different magnetic coercivity, the field strength needed to reverse the magnetization, and the new state held after the voltage was removed. The researchers cycled it more than 3,000 times without measurable fatigue. The effect was strongest in the thinnest Fe3GaTe2 flakes and faded as the magnetic layer got thicker, which is what one expects if the ferroelectric layer is acting on the magnet's interface rather than its bulk. "The obvious advantage of our approach is that it breaks down the challenge of realizing a few different orders in one single-phase material," Gong said. "The ensuing advantage is that it gives unlimited freedom for researchers to combine different 2D ferroics together," by choosing different materials, thicknesses, twist angles and stacking registries.
Gong said the work "could have implications for the development of energy-efficient spintronic devices, integrated nonlinear and nonreciprocal photonics, and ultracompact and supersensitive electric and magnetic field sensors." The immediate next steps are to try other pairings of 2D ferroics and to push the switching from a change in coercivity to a full, deterministic reversal of the magnetization by voltage alone, which is the operation a working memory cell needs.
Originally reported by Phys.org.