Phonon-mode-specific spin-flip scattering dictated ultrafast demagnetization in a van der Waals ferromagnet

In the world of modern technology, the ability to store and process information relies heavily on the magnetic properties of materials. For decades, scientists have sought to make these magnetic devices faster and more efficient by manipulating the tiny magnetic moments of electrons, a property known as spin. When a magnetic material is hit with a pulse of light, its magnetism can vanish in a fraction of a second, a process called ultrafast demagnetization. Understanding exactly how this happens is crucial for building the next generation of high-speed memory. For a long time, researchers viewed the atoms inside a solid material as a chaotic, undifferentiated crowd of vibrations, or phonons, that simply absorbed energy and helped cool the system down. However, a new perspective suggests that not all vibrations are the same; some specific types of atomic shaking might be far more effective at scrambling magnetic spins than others, acting as the true drivers of how quickly magnetism disappears.

A team of researchers at Nanjing University and their international collaborators has now provided direct evidence for this idea using a special type of magnetic material called a van der Waals ferromagnet, specifically a thin film of chromium telluride. These materials are made of atomically thin layers held together by weak forces, making them ideal for studying how light, electricity, and magnetism interact on a microscopic scale. The scientists wanted to solve a mystery: why does the speed at which this material loses its magnetism change depending on the temperature? By combining high-speed laser experiments with detailed computer simulations, they discovered that the answer lies in which specific atomic vibrations are doing the work. They found that at lower temperatures, one type of vibration dominates the process, but as the material gets warmer, a different type of vibration takes over, slowing down the loss of magnetism.

The researchers began by growing a high-quality, ten-layer-thick film of chromium telluride on a gallium arsenide substrate. They then used a technique called time-resolved magneto-optical Kerr effect, which involves firing a short laser pulse to disturb the material's magnetism and a second, weaker pulse to measure how the magnetism recovers over time. They observed a characteristic two-step process: the magnetism drops almost instantly, followed by a slower, more gradual decline. To understand the physics behind this, they applied a theoretical framework that treats the electrons, the spins, and the atomic lattice as three separate but connected systems exchanging energy. By fitting their experimental data to this model, they calculated the probability of a spin flip—the event where an electron's magnetic direction reverses—occurring during a collision with a vibrating atom. Their calculations showed that as the temperature rose from 80 to 200 Kelvin, the probability of these spin flips dropped significantly, meaning the spins became harder to scramble at higher temperatures.

To figure out why this probability changed, the team turned to Raman spectroscopy, a method that measures how the material scatters light to reveal the specific frequencies of its atomic vibrations. They identified two main types of vibrations in the chromium telluride film: one where the atoms move back and forth within the plane of the layers, and another where they move up and down, perpendicular to the layers. By analyzing how the strength of the interaction between electrons and these specific vibrations changed with temperature, they found a clear shift. At lower temperatures, the in-plane vibrations were the primary cause of spin flips. However, as the material warmed up, the contribution from the up-and-down vibrations grew, while the role of the in-plane vibrations diminished. This redistribution of activity suggested that the material was switching from one mechanism of demagnetization to another as it heated up.

To confirm that these different vibrations actually had different abilities to destroy magnetism, the researchers performed complex computer simulations based on quantum mechanics. They modeled the material and artificially induced the specific atomic movements of both the in-plane and up-and-down vibrations to see what would happen to the electron spins. The results were decisive: the in-plane vibrations caused a much larger and faster loss of magnetic moment than the up-and-down vibrations. The simulations showed that the in-plane motion created a stronger disturbance in the electronic structure, leading to a more efficient transfer of energy that scrambled the spins. In contrast, the up-and-down motion was less effective at disrupting the magnetic order. This confirmed that the specific type of vibration matters immensely; some are simply better at turning magnetic order into disorder than others.

The team then connected these findings to explain the temperature dependence they observed in the lab. As the material heats up, it expands slightly, much like a metal bridge expanding on a hot day. This tiny expansion changes how the atoms overlap and how their electrons interact, a phenomenon known as orbital hybridization. The researchers determined that this thermal expansion makes the up-and-down vibrations more effective at coupling with the electrons' spin states, even though they are naturally less efficient than the in-plane vibrations. Consequently, at higher temperatures, the material relies more on this less efficient up-and-down mechanism, which explains why the overall process of demagnetization slows down and the spin-flip probability decreases. This insight moves beyond the old view of the atomic lattice as a uniform bath of energy and instead reveals a nuanced landscape where specific vibrational modes play distinct, temperature-dependent roles.

This work fundamentally changes how scientists understand the speed limits of magnetic materials. By showing that the efficiency of spin relaxation depends on which specific atomic vibration is active, the study provides a new framework for controlling ultrafast magnetic processes. The findings suggest that by engineering materials to favor or suppress specific vibrational modes, it may be possible to tune the speed of magnetic switching for future computing devices. The researchers have established that the path from light to lost magnetism is not a single, uniform road but a complex network of channels, each with its own speed and efficiency, dictated by the precise way the atoms move.