Effects of uniaxial strain on monolayer transition-metal dichalcogenides revisited

Using hybrid density functional calculations, this study reveals that uniaxial tensile strain in monolayer transition-metal dichalcogenides significantly reduces the fundamental band gap and induces a valley drift of band extrema away from the K point, leading to indirect band gaps that explain the experimentally observed decrease in photoluminescence intensity.

The Gist

Imagine a single layer of a material called a Transition-Metal Dichalcogenide (TMD) as a tiny, ultra-thin trampoline made of a honeycomb pattern. This trampoline is so thin it's only three atoms thick, and it's a superstar in the world of electronics because it can turn electricity into light (and vice versa) very efficiently.

The scientists in this paper wanted to understand what happens when you stretch this trampoline. They didn't just pull it in one direction; they pulled it like a rubber band, either along the "armchair" direction or the "zigzag" direction, to see how its internal "energy landscape" changes.

Here is the story of what they found, explained simply:

1. The Setup: A Perfectly Balanced Dance

In its natural, relaxed state, this material is like a perfectly choreographed dance. The "dancers" are electrons (the negative charge carriers) and "holes" (the positive charge carriers).

• The Goal: For the material to glow brightly (emit light), the electron and the hole need to meet up at the exact same spot in their energy map and jump together.
• The Problem: If they are in different spots, they can't meet easily, and the light goes dim.

2. The Experiment: Stretching the Trampoline

The researchers used powerful computer simulations (like a super-accurate virtual wind tunnel) to stretch this material. They pulled it gently, up to 5% of its size.

What happened to the energy?
Think of the material's energy levels as a hill and a valley.

• The Electron Valley (Conduction Band): When you stretch the material, this valley drops down steeply, like a slide.
• The Hole Valley (Valence Band): This one moves, but much more slowly, like a gentle slope.
• The Result: The gap between the top of the hill and the bottom of the valley gets smaller. This means the material needs less energy to work, which is great for making flexible electronics.

3. The Big Surprise: The "Valley Drift"

This is the most important discovery. In the relaxed material, the electrons and holes were dancing in a perfect circle at the center of the map (a point called K).

When you stretch the material, the map itself gets distorted (like stretching a rubber sheet).

• The Drift: The electrons and holes don't stay at the center. They get pushed off-center, drifting away from the original spot.
• The Mismatch: Here's the kicker: They drift at different speeds. The electrons run away faster than the holes.
• The Analogy: Imagine two runners starting at the same line. You tell them to run, but one is on a fast treadmill and the other is on a slow one. After a few seconds, they are far apart. Even though they are both still running, they can't high-five anymore.

4. Why This Matters: The "Dimming" Light

In the real world, scientists have noticed that when you stretch these materials, they stop glowing as brightly.

• Old Explanation: People thought the material was changing its fundamental nature, turning from a "direct" light-emitter to an "indirect" one where the energy levels crossed over completely.
• New Explanation (This Paper): The paper shows that the material doesn't need to change its fundamental nature to stop glowing. Because the electrons and holes have drifted apart in "momentum space" (a fancy way of saying they are now in different locations on the map), they can't meet up to create light easily. It's like trying to catch a ball while running in opposite directions; the harder you stretch, the harder it is to catch.

5. The Takeaway for the Future

This research is like getting a new, highly accurate map for engineers.

• Precision: They didn't just guess; they calculated exactly how much the energy drops and how far the valleys drift for different materials (like Molybdenum Sulfide or Tungsten Selenide).
• Application: If you want to build a flexible phone screen or a sensor that bends without breaking, you need to know exactly how much you can stretch it before it stops working well. This paper gives them the numbers to design those devices perfectly.

In a nutshell: Stretching these tiny materials makes them change shape in a way that pushes their electrons and holes apart. Even though they are still there, they drift too far from each other to create light efficiently. The scientists mapped exactly how this happens, giving us the tools to build better, bendable technology.

Technical Summary

1. Problem Statement

Transition-metal dichalcogenides (TMDs), specifically monolayer $1H$-phase $MX_2$ ($M=$ Mo, W; $X=$ S, Se, Te), are promising materials for flexible electronics and optoelectronics due to their direct band gaps and strong spin-orbit coupling (SOC). While the effects of biaxial strain are well-understood, the effects of uniaxial strain remain ambiguous in existing literature due to several critical theoretical limitations:

• Brillouin Zone (BZ) Distortion: Many studies fail to account for the distortion of the hexagonal BZ under uniaxial loading, leading to incorrect identification of high-symmetry points and band extrema.
• Valley Drift Neglect: Previous works often assume band extrema remain pinned to the high-symmetry $K$ point, ignoring the "valley drift" where extrema move to off-symmetry wave vectors.
• Inaccurate Band Gaps: Reliance on semilocal Density Functional Theory (DFT) functionals often underestimates quasiparticle band gaps and neglects SOC. This leads to erroneous predictions regarding the relative ordering of valence band maxima (e.g., $K$ vs. $\Gamma$ valleys) and spurious predictions of direct-to-indirect transitions.
• Gap Definition Confusion: Comparing calculated fundamental quasiparticle gaps directly to experimental optical gaps (which include exciton binding energy) creates quantitative discrepancies.

The paper aims to resolve these inconsistencies by providing a rigorous, quantitative analysis of uniaxial strain effects on the electronic structure of monolayer TMDs.

2. Methodology

The authors employed a high-accuracy first-principles computational approach:

• Electronic Structure: Calculations were performed using Hybrid DFT (HSE06) with a Hartree-Fock mixing parameter of $\alpha = 0.40$, explicitly including Spin-Orbit Coupling (HSE$\alpha$+SOC). This ensures accurate quasiparticle band gaps and correct valence band splitting.
• Structural Relaxation: Equilibrium lattice parameters were obtained using the SCAN meta-GGA functional. Under strain, structures were fully relaxed (ionic coordinates and cell shape) to minimize total energy, ensuring the correct Poisson ratio and transverse strain response.
• Strain Modeling: Uniaxial tensile strain (up to 5%) was applied along both armchair and zigzag directions. Crucially, the calculations were performed directly on the distorted primitive hexagonal cell to preserve the correct reciprocal lattice geometry, avoiding the errors associated with mapping to rectangular supercells.
• Valley Tracking: The band structures were computed along strain-adapted paths in the distorted BZ ($\Gamma \to Q \to Q' \to K \to K' \to M \to M'$) to accurately track the movement of band extrema.
• Reference Frame: Band edge energies were referenced to the vacuum level to provide absolute shifts, facilitating comparison with photoemission experiments.

3. Key Contributions

• Valley Drift Mechanism: The study definitively demonstrates that under uniaxial strain, both the Conduction Band Minimum (CBM) and Valence Band Maximum (VBM) drift away from the nominal $K$ point to nearby off-symmetry wave vectors.
• Asymmetric Drift Rates: While time-reversal symmetry ensures that the $K$ and $K'$ valleys remain degenerate in energy, the electron and hole valleys drift at different rates. This creates a growing momentum mismatch ($|\delta k_e - \delta k_h|$) between the CBM and VBM.
• Mechanism for PL Quenching: The paper provides a natural explanation for the experimentally observed decrease in photoluminescence (PL) intensity under strain. It is not solely due to an energetic crossover to an indirect $\Gamma-K$ gap, but primarily due to momentum-space indirectness caused by valley drift, which pushes the lowest-energy excitons outside the radiative light cone.
• Orbital Character Analysis: The study links the differential drift rates to the distinct orbital characters of the band edges (e.g., $d_{z^2}$ vs. $d_{x^2-y^2}/d_{xy}$), explaining their varying sensitivities to strain-modified bonding.

4. Key Results

• Band Gap Reduction: Tensile uniaxial strain causes a pronounced, systematic reduction in the fundamental band gap across all $MX_2$ compounds. This reduction is dominated by the downward shift of the CBM, while the VBM shifts more weakly.
• Valley Drift Quantification:
  • For MoS$_2$ at 5% strain:
    • Armchair: Electron valley drifts by $0.064 \, \text{\AA}^{-1}$; Hole valley by $0.052 \, \text{\AA}^{-1}$.
    • Zigzag: Electron valley drifts by $0.080 \, \text{\AA}^{-1}$; Hole valley by $0.065 \, \text{\AA}^{-1}$.
  • This drift is approximately linear with strain.
• Indirect Gap Formation: Even when the band edges remain $K$-derived in energy, the momentum mismatch renders the fundamental gap momentum-indirect. This occurs at lower strains than the energetic crossover to a $\Gamma$-valley dominated gap.
• Vacuum-Referenced Shifts: The authors provide absolute energy shifts for CBM and VBM relative to the vacuum level. The CBM shifts downward significantly (increasing electron affinity), while the $\Gamma$-valence maximum ($\Gamma_v$) shifts in the opposite direction (upward) due to its different orbital character ($d_{z^2}/p_z$) and sensitivity to layer thickness changes.
• Poisson Ratio: The Poisson ratio is found to be weakly dependent on strain magnitude but anisotropic (e.g., $\nu \approx 0.22$ for armchair vs. $0.27$ for zigzag in MoS$_2$).

5. Significance

• Resolution of Experimental Discrepancies: The findings explain why PL intensity drops continuously under strain before any major energetic crossover occurs, attributing it to momentum-space indirectness rather than just energy-level crossing.
• Quantitative Design Rules: By providing vacuum-referenced band-edge slopes and accurate quasiparticle gaps, the study offers a "response map" for engineering band alignments in TMD-based devices.
• Methodological Benchmark: The work establishes a rigorous protocol for studying strained 2D materials, emphasizing the necessity of including SOC, using hybrid functionals, and correctly modeling the distorted Brillouin zone.
• Applications: These insights are critical for the development of strain-tunable optoelectronics, flexible electronics, and quantum-defect applications where precise control over band structure and carrier mobility is required.