Valley-polarized Orbital and Spin Magnetism Induced by Femtosecond Optical Pulses in Two-Dimensional Semiconductors

This paper theoretically demonstrates that circularly polarized femtosecond laser pulses can generate and distinctly control valley-polarized spin and orbital magnetism in two-dimensional semiconductors, revealing that orbital dynamics driven by direct electric field coupling are faster and more sensitive to dephasing than the gradually developing spin response mediated by spin-orbit coupling.

The Gist

Imagine a tiny, flat world made of a special kind of material (like a single layer of a sandwich called a transition-metal dichalcogenide). In this world, electrons don't just sit still; they live in two different "neighborhoods" called valleys (labeled K and K'). These valleys are like two sides of a coin that look identical but behave differently depending on how you spin them.

This paper is a theoretical study (a computer simulation) about what happens when you zap this material with an incredibly fast, super-bright flash of light (a femtosecond laser pulse). The researchers wanted to see if they could use this light to create magnetism (a magnetic force) out of nothing, and specifically, if they could control two different "types" of magnetism: Spin and Orbital.

Here is a breakdown of their findings using simple analogies:

1. The Two Types of Magnetism: The "Dancer" vs. The "Spinning Top"

In this material, electrons have two ways to create a magnetic field:

• Spin Magnetism: Think of this like a spinning top. The electron spins on its own axis. In this material, the light doesn't push the top directly. Instead, the light pushes the electron's path, and because of a special rule called "spin-orbit coupling," the top starts to spin slowly. It's an indirect connection.
• Orbital Magnetism: Think of this like a dancer spinning in a circle around a stage. The electron is physically moving in a loop around the atom. The light pushes the dancer directly. Because the light hits the dancer straight on, this movement happens much faster and more violently.

2. The Experiment: Shining the Light

The researchers simulated hitting the material with a laser pulse that is circularly polarized (meaning the light waves spin like a corkscrew as they travel).

• The Result: The light successfully created a magnetic field in the material.
• The Control: By changing the color (energy) of the laser, they could choose which "neighborhood" the electrons went to. This allowed them to pick whether they wanted mostly Spin magnetism or mostly Orbital magnetism. It's like having a remote control where one button turns on the spinning tops, and another button turns on the dancers.

3. The Race: Who Moves Faster?

The study found a huge difference in how fast these two types of magnetism react to the light:

• The Orbital Magnetism (The Dancer): Because the light pushes it directly, it reacts almost instantly. It starts shaking and oscillating (wiggling back and forth) very quickly, like a drum being hit. These wiggles are called "Rabi oscillations."
• The Spin Magnetism (The Spinning Top): Because it relies on the indirect "spin-orbit" rule, it takes its time. It builds up slowly and smoothly, like a heavy wheel slowly gaining speed.

4. The "Noise" Factor (Dephasing)

In the real world, things get messy. Electrons bump into other things (like vibrations in the material), which is called "dephasing" or "noise."

• The Finding: The fast, wiggling Orbital magnetism is very sensitive to this noise. If there is too much noise, the wiggles stop, and the magnetism settles down quickly. Surprisingly, this noise actually helped the orbital magnetism become stronger and more stable than the spin magnetism in some cases.
• The slow Spin magnetism was barely affected by the noise; it just kept building up its speed regardless.

5. The "Magic" of Two-Photon Absorption

The researchers also tried using light that wasn't strong enough to jump the gap between energy levels on its own (below the band gap).

• The Trick: Even with weaker light, the electrons could "team up" and absorb two photons at once to make the jump.
• The Result: This "two-photon" trick still created strong magnetism. It showed that you don't need a super-powerful laser to get this effect; you just need the right timing and color.

Summary

The paper concludes that by using ultrafast laser pulses, we can create and control magnetism in these 2D materials. The key takeaway is that Orbital magnetism (the dancer) and Spin magnetism (the spinning top) are fundamentally different animals. They react to light in different ways, at different speeds, and are affected by noise differently. To build future technologies that use light to control magnets, we must pay attention to the "dancer" (orbital) just as much as the "spinning top" (spin), because they don't behave the same way.

Technical Summary

Technical Summary: Valley-polarized Orbital and Spin Magnetism Induced by Femtosecond Optical Pulses in Two-Dimensional Semiconductors

Problem and Motivation
Conventional magnetic control via external fields is often constrained by slow switching speeds and high energy dissipation. While all-optical helicity-dependent switching using femtosecond circularly polarized light has emerged as a promising alternative, primarily through the inverse Faraday effect (IFE), previous research has largely focused on coherent light-induced magnetization in metallic systems. A significant gap exists in understanding non-perturbative light-matter interactions in nonmagnetic, nonmetallic materials, particularly how these interactions influence coherent magnetization on femtosecond timescales. Furthermore, while orbital magnetization is often neglected in solids due to crystal field quenching, recent studies suggest orbital contributions can be substantial under non-equilibrium conditions. The authors aim to investigate the ultrafast generation of both spin and orbital magnetization in two-dimensional (2D) gapped Dirac systems with spin-orbit coupling (SOC), representative of transition-metal dichalcogenides (TMDs), to identify fundamental differences in their dynamical behaviors.

Methodology
The study employs a theoretical framework based on a minimal tight-binding model for a 2D gapped graphene system with SOC, capturing the valley-contrasting physics of TMDs. The system features a band gap ($\Delta_g$) of 3.0 eV and a spin-flip excitation gap ($\Delta_{sf}^g$) of 4.0 eV. The dynamics are modeled using a time-dependent single-particle density-matrix formalism, solving the quantum Liouville equation with decoherence terms. The laser-matter interaction is treated non-perturbatively within the dipole approximation, where the vector potential is incorporated via minimal coupling.

The simulations utilize a fourth-order Runge–Kutta scheme with a 0.02 fs time step and a $200 \times 200$ k-point mesh in the Brillouin zone. The driving field consists of ultrashort (100 fs) circularly polarized laser pulses with a peak intensity of $10^{11}$ W/cm$^2$. The authors calculate the time evolution of the net magnetic moment per unit cell, distinguishing between spin ($m_{spin}$) and orbital ($m_{orb}$) contributions, defined via the Bloch functions and their derivatives.

Key Results

• Distinct Temporal Dynamics: The study reveals fundamentally different temporal behaviors for spin and orbital magnetization. Orbital magnetization couples directly to the external electric field, resulting in faster dynamics characterized by pronounced Rabi-like oscillations during the pulse interaction. In contrast, spin magnetization develops gradually through indirect coupling mediated by spin-orbit interaction, exhibiting a slower, monotonic increase that saturates after the pulse peak.
• Energy Selectivity: The induced magnetization can be selectively controlled via photon energy.
  • Resonant Excitation: Near the band gap ($\hbar\omega \approx \Delta_g$), the remnant orbital magnetization dominates. Near the spin-flip gap ($\hbar\omega \approx \Delta_{sf}^g$), the remnant spin magnetization dominates.
  • Below-Band-Gap Excitation: Nonlinear multiphoton processes generate substantial magnetization even when photon energy is below the band gap. Specifically, resonant two-photon absorption ($\Delta E = 2\hbar\omega$) drives peaks in orbital response at $\hbar\omega \approx \Delta_g/2$ and spin response at $\hbar\omega \approx \Delta_{sf}^g/2$. Higher-order processes (three- and four-photon) also contribute to both spin and orbital responses at lower energies.
• Valley Polarization: The induced magnetization is directly linked to valley-polarized electronic excitations. Reversing the laser helicity switches the population between the K and K' valleys, thereby inducing magnetization with opposite signs. In the below-band-gap regime, distinct two-photon absorption pathways from different valence bands (VB1 and VB2) to the conduction band selectively drive spin or orbital dynamics in specific regions of momentum space.
• Role of Decoherence: The inclusion of electron-hole dephasing ($T_2 = 10$ fs) significantly alters the results. While spin dynamics remain nearly unchanged, the oscillatory character of the orbital response is strongly suppressed. Paradoxically, this suppression drives the system more rapidly toward saturation, resulting in an induced orbital magnetization that is nearly twice as large as the spin contribution in the presence of decoherence.

Significance and Claims
The authors claim that their work highlights the crucial role of orbital contributions in ultrafast magnetism, which are often overlooked. They demonstrate that spin and orbital magnetization originate from fundamentally different light-matter coupling mechanisms: direct coupling to the electric field for orbital moments versus indirect coupling via SOC for spin moments. This leads to qualitatively dissimilar temporal behaviors and sensitivities to decoherence.

The paper asserts that properly accounting for orbital contributions is essential for future technologies utilizing femtosecond control of magnetism. By establishing a direct correspondence between resultant magnetization and valley-polarized conduction-band populations across both resonant and off-resonant regimes, the study provides a theoretical basis for the selective control of spin and orbital degrees of freedom in 2D semiconductors using tailored optical pulses. The findings suggest that nonlinear light-matter interactions are key to manipulating these degrees of freedom on femtosecond timescales.