Spin separation of Dirac electrons by twisted and vector optical beams

In the world of modern electronics, the ability to control the tiny magnetic orientation of an electron, known as its spin, is a holy grail. This property allows data to be stored and processed more efficiently than with traditional electrical charges alone. For decades, scientists have tried to manipulate these spins using light, hoping to create faster, cooler, and more powerful devices. The challenge has always been that light waves are incredibly large compared to the atoms they hit. While a beam of light might twist or swirl on a scale visible to the eye, the interaction with a single electron happens in a space so small that the light looks like a flat, uniform sheet. Because of this size mismatch, trying to use the complex shapes of light to directly twist an electron's spin has proven to be a very inefficient process, often yielding almost no result.

A team of researchers at the Ioffe Institute in St. Petersburg has now proposed a different way to solve this puzzle, one that bypasses the need for a direct, microscopic handshake between light and electron. They suggest that instead of trying to twist the electron directly, we can use specially shaped beams of light to create a flow of electrons that naturally separates by spin. Imagine a river where the water itself is not swirling, but the banks are shaped in a way that forces the water to split into two distinct streams. In their study, the researchers used two types of structured light beams: "twisted" beams that carry a spiral phase, and "vector" beams where the direction of the light's vibration rotates as you move across the beam. Crucially, neither of these beams uses the circular polarization that typically spins electrons up or down. Instead, they rely on the spatial pattern of the light to generate currents that push electrons with opposite spins in different directions.

The researchers focused on a class of materials called two-dimensional Dirac materials, specifically looking at monolayers of transition metal dichalcogenides. These are atom-thin sheets of matter that behave like a unique type of fluid for electrons. When these sheets are illuminated by the structured beams, the light creates a textured pattern of spin currents. These are flows of electrons where the direction of movement is tied to their spin. The study shows that these currents do not flow in a straight line; they diverge, spreading out from the center of the beam. As these currents spread, they encounter resistance and eventually slow down, a process known as relaxation. This interplay between the spreading flow and the slowing down causes the electrons to pile up in specific areas, creating a steady accumulation of spin. The result is a map of spin density that mirrors the intricate patterns of the light beam itself.

To understand how this works, the team looked at two specific scenarios. In the first, they used twisted beams that carry orbital angular momentum, which gives the light a corkscrew-like structure. In the second, they used vector beams where the polarization, or the direction the light waves vibrate, rotates around the center. In both cases, the light is locally linear, meaning at any single point, the light vibrates in a straight line, but the direction of that line changes as you move across the beam. The researchers found that this changing direction is the key. It generates a spin current that depends on how the polarization changes from point to point. When these currents diverge, they leave behind a trail of separated spins. The team calculated that this indirect method is significantly more efficient than trying to transfer the light's twist directly to the electron, potentially by a factor related to how far an electron travels before hitting a defect in the material.

The researchers performed detailed numerical simulations to see what these spin patterns would look like in real materials. They used parameters relevant to transition metal dichalcogenide layers, such as a band gap of 2 electron volts and a spin relaxation time of 10 nanoseconds. The simulations revealed distinct spatial patterns. For the twisted beams, the spin density formed rings and spirals that depended on the number of twists in the light. For the vector beams, the pattern was determined by how many times the polarization vector rotated as you went around the center. Interestingly, if the polarization was aligned perfectly radially or azimuthally, no spin separation occurred. However, when the polarization was set at a 45-degree angle, the separation was most efficient, creating a strong, localized spin density. The study also showed that the size of these spin patterns is not just a copy of the light pattern but is smoothed out by the diffusion of electrons, which can travel several micrometers before their spin relaxes.

The findings suggest a new mechanism for imprinting the complex geometry of light onto the spin of electrons without needing circular polarization. This means that the twist of a light beam or the vorticity of its polarization can be translated into a tangible magnetic pattern in a solid material. The researchers note that these spin textures could be detected experimentally using established techniques like Faraday or Kerr rotation, which measure how light reflects off a magnetic surface. While the work is currently theoretical and based on calculations, it provides a clear roadmap for how to use the shape of light to control electron spins in future spintronic devices. By moving away from the difficult task of direct transfer and embracing the power of structured currents, this approach offers a promising path toward more efficient control of quantum information.