Out-of-plane orbital torque from orbital-to-spin conversion in a low-symmetry system

In the quest to build faster, more energy-efficient computers, scientists have long relied on the tiny magnetic spins of electrons to store information. Imagine these spins as microscopic compass needles that can point up or down to represent the ones and zeros of digital data. To write new information, researchers typically use electric currents to nudge these needles into a new position. However, a major hurdle remains: in many common materials, the laws of physics prevent these currents from flipping the needles without an external magnetic field to help guide them. This requirement for an extra magnetic field makes devices bulky and power-hungry, blocking the path toward the ultra-compact memory chips of the future.

Recently, a new avenue has opened up by looking not just at the spin of electrons, but at their orbital motion. Just as the Earth spins on its axis while also orbiting the Sun, electrons possess a similar dual nature. While spin has been the workhorse of modern electronics, the orbital motion of electrons offers a fresh, powerful way to generate magnetic forces, often without needing heavy, rare elements. The challenge is that orbital motion does not directly talk to magnetic needles; it must first be converted into a spin force. For years, scientists believed that the symmetry of a crystal's structure strictly limited how this conversion could happen, effectively forbidding the kind of force needed to flip magnetic needles without help. A team of researchers at Beihang University has now shown that this rule can be broken, revealing a new way to control magnetic memory using the hidden geometry of atoms.

The researchers focused on a specific sandwich of materials: a layer of ruthenium metal placed on top of a cobalt-platinum alloy. They chose ruthenium because it is excellent at generating a flow of orbital motion when an electric current passes through it. The cobalt-platinum layer, however, was the key to the experiment. While it is a disordered alloy on a large scale, the team engineered it so that at the interface where it meets the ruthenium, the atomic arrangement lacks a specific type of mirror symmetry. In simpler terms, the atomic landscape at this boundary is lopsided in a very specific, three-sided pattern. This broken symmetry is the crucial ingredient that allows the physics to behave differently than in standard, highly symmetrical materials.

When the team sent an electric current through the ruthenium layer, it generated a transverse flow of orbital motion, which then traveled into the cobalt-platinum layer. In a standard, symmetrical material, this flow would simply convert into a sideways spin force, which is not strong enough to flip a magnetic needle standing upright on its own. But in their low-symmetry setup, the researchers observed something unexpected. The orbital flow was converted into a spin force that pushed not sideways, but straight up or down. This vertical push, or torque, was strong enough to flip the magnetic needles from pointing up to pointing down, and vice versa, without any external magnetic field. The result was a clean, deterministic switch of the magnetic state, driven entirely by the electric current and the unique geometry of the material interface.

To prove that this effect was real and not just a fluke, the team tested how the switching behaved when they changed the direction of the electric current. They found that the ability to flip the magnet followed a distinct three-fold pattern. As they rotated the current direction, the switching strength rose and fell in a rhythm that repeated every 120 degrees. This pattern perfectly matched the three-sided symmetry of the cobalt-platinum crystal structure, confirming that the effect was indeed rooted in the material's geometry. Furthermore, they discovered that the strength of this vertical force grew as they made the ruthenium layer thicker, up to a point. The effect peaked when the ruthenium layer was about 18 to 21 nanometers thick, suggesting that the orbital motion was being generated deep within the bulk of the ruthenium and then efficiently converted at the interface.

The researchers also used powerful computer simulations to look inside the material at the atomic level. These calculations revealed that the conversion process relies on a specific "swapping" effect that occurs only when the symmetry is broken. In the bulk of the cobalt-platinum material, this effect is weak, but at the interface with the ruthenium, it becomes dramatically stronger. The simulations showed that the electrons' orbital motion is effectively swapped into a vertical spin force right at the boundary, a process that is forbidden in symmetrical crystals. This microscopic picture confirmed that the interface itself acts as a filter, selecting and amplifying the specific type of force needed for field-free switching.

The implications of this discovery are significant for the design of future memory devices. By combining a material that generates orbital motion with a low-symmetry layer that converts it, the team demonstrated a way to write magnetic data using less energy and fewer components than current methods require. Their device achieved a level of efficiency that is roughly twice as high as previous attempts using similar materials without the ruthenium layer. This work suggests that by carefully engineering the symmetry of atomic interfaces, scientists can unlock new ways to manipulate magnetic information. It opens a path toward creating memory chips that are not only smaller and faster but also capable of operating without the bulky magnetic fields that have long been a bottleneck in the field. The study establishes that the geometry of the atomic world is just as important as the materials themselves in determining how we can control the flow of information in the next generation of electronics.