Transition Metal Dichalcogenide MoS${}_2$: oxygen and fluorine functionalization for selective plasma processing

This study demonstrates that oxygen and fluorine functionalization, combined with cryogenic temperatures, significantly lowers the sulfur sputtering energy threshold in MoS${}_2$ via the formation of volatile products, thereby widening the ion energy window for selective, damage-controlled chalcogen removal in transition metal dichalcogenide plasma processing.

The Gist

Imagine you have a very delicate, microscopic sandwich made of two slices of bread (Molybdenum atoms) with a layer of filling in the middle (Sulfur atoms). This is MoS₂, a material scientists are excited about because it could power the next generation of tiny, super-efficient electronics.

Now, imagine you want to eat only the filling (the Sulfur) without crushing the bread (the Molybdenum). This is tricky. If you hit the sandwich too hard, you smash the whole thing. If you hit it too gently, nothing happens.

This paper is about finding the "Goldilocks zone" of hitting this sandwich, and discovering a clever trick to make it much easier to remove just the filling.

The Problem: The "Too Hard, Too Soft" Dilemma

In the world of making microchips, scientists use plasma (a super-hot gas of charged particles) to etch away material. They shoot tiny bullets (ions) at the material to knock atoms off.

• The Old Way: If you shoot a bullet at a plain MoS₂ sandwich, you need to hit it with about 30 units of energy to knock out a Sulfur atom. But if you hit it with 100 units, you accidentally smash the Molybdenum bread too.
• The Goal: We want to knock out the Sulfur at a low energy (say, 10 units) so the bread stays perfectly intact. But with plain MoS₂, 10 units isn't enough to do anything.

The Solution: The "Chemical Glue" Trick

The researchers discovered a way to change the rules of the game. Instead of shooting at the plain sandwich, they first coat the top layer with Oxygen or Fluorine atoms. Think of this as putting a special, sticky "chemical glue" on the Sulfur atoms.

Here is what happens when they shoot the bullet (Argon ion) at this new, coated sandwich:

1. The Reaction: When the bullet hits, it doesn't just try to knock the Sulfur out alone. Because of the Oxygen or Fluorine "glue," the Sulfur grabs onto them and forms a new, lightweight molecule (like SO₂ or SF₄).
2. The Ejection: It is much easier to throw a lightweight, fluffy marshmallow (the new molecule) than a heavy, dense rock (a single Sulfur atom).
3. The Result: Because the bullet is now knocking out a "marshmallow" instead of a "rock," it only needs 10 units of energy to succeed. This is a huge drop from the original 30 units!

The Analogy: The Heavy Box vs. The Balloon

• Plain MoS₂: Imagine trying to push a heavy, stuck box off a table. You have to push really hard (30 units of force) to get it to budge. If you push too hard, you knock the table over.
• Functionalized MoS₂: Now, imagine someone ties a giant helium balloon to that box. Suddenly, the box is light and buoyant. You can push it off the table with a gentle nudge (10 units of force). The table (the Molybdenum structure) stays perfectly safe.

The "Temperature" and "Angle" Twist

The paper also found two other cool things:

1. The Wobbly Table (Temperature): If the table is shaking (hot temperature), it's actually easier to knock the box off because the box is already moving a bit. The researchers showed that cooling the material down makes the process more predictable, while warming it up helps the "marshmallows" fly off even easier.
2. The Angle of Attack: If you hit the sandwich straight down (90 degrees), it's okay. But if you hit it at a slight angle (like 30 degrees), it's like sliding a knife under a cookie cutter—it works even better! This is especially true for the Oxygen-coated version, which acts like a perfectly organized grid. The Fluorine version is a bit more chaotic (like a messy pile of LEGOs), so the angle doesn't matter as much.

Why Does This Matter?

This discovery is a game-changer for making better electronics.

• Precision: It allows engineers to remove specific atoms (Sulfur) without damaging the underlying structure.
• Control: By using Oxygen or Fluorine, they can lower the energy required, meaning they can use gentler, more precise tools.
• Versatility: This trick likely works on other similar materials too, not just MoS₂.

In short: By adding a little bit of Oxygen or Fluorine "seasoning" to the material, scientists found a way to gently peel off the top layer without breaking the delicate structure underneath, opening the door to creating smaller, faster, and more efficient electronic devices.

Technical Summary

1. Problem Statement

Transition Metal Dichalcogenides (TMDs), particularly monolayer MoS$_2$, are promising candidates for next-generation electronics due to their direct bandgaps. Plasma processing is a critical technique for etching, cleaning, and doping these materials. However, a major challenge exists in identifying an ion energy window that allows for the selective removal of chalcogen atoms (Sulfur) while preserving the underlying metal lattice (Molybdenum).

• Current Limitation: In pristine MoS$_2$, the threshold energy ($E_{sputt,S}$) required to sputter a sulfur atom is approximately 30 eV, while removing Molybdenum atoms requires energies near 100 eV. While this suggests a window, modern plasma sources can control ion energies with high precision (10–20 eV). The challenge is to lower the sulfur sputtering threshold further to enable highly selective, low-damage processing without risking metal lattice destruction.
• Goal: To identify a mechanism that significantly lowers the sulfur sputtering threshold ($E_{sputt,S}$) below the current ~30 eV baseline, thereby widening the processing window for damage-controlled etching.

2. Methodology

The authors employed Ab-Initio Molecular Dynamics (AIMD) simulations using the CP2K software package to model ion-surface interactions.

• Simulation Setup:
  • System: A 4x4 supercell of monolayer MoS$_2$ (2H phase).
  • Functionalization: The surface was functionalized with atomic Oxygen (O) and Fluorine (F) to simulate high surface coverage ($c \approx 1$).
  • Projectile: Argon (Ar) ions were directed at the surface with varying kinetic energies (10–50 eV) and impact angles ($\theta$).
  • Conditions: Simulations were conducted at various temperatures (1 K to 350 K) to study thermal effects. A cutoff time of 2 ps was established to distinguish between immediate "sputtering" and slow "thermal desorption."
• Theoretical Framework:
  • The authors developed a parameter-free mechanistic theory to predict the temperature dependence of the sputtering threshold. This theory couples the angular dependence of the threshold energy with the thermal fluctuations of surface atoms.
  • Spin States: The study explicitly considered spin states (singlet vs. triplet), determining that the singlet state is the ground state for O-adsorbed MoS$_2$ and that fixed-spin singlet dynamics are sufficient for the collision timescales involved.
• Validation: Results were cross-validated against Density Functional Theory (DFT) functionals (PBE vs. R2SCAN) and dispersion corrections (D3BJ vs. rVV10) to ensure robustness.

3. Key Contributions

1. Discovery of Chemically Enhanced Physical Sputtering: The paper proposes and validates a two-step mechanism where functionalization lowers the energy barrier not just by chemical weakening, but by facilitating the formation of metastable gas-phase molecular fragments (e.g., SO$_2$, SF$_4$) rather than individual atoms.
2. Quantitative Reduction of Threshold Energy: The study demonstrates that O/F functionalization reduces the sulfur sputtering threshold from ~30 eV to ~10 eV (specifically ~14 eV for O and ~9.5 eV for F).
3. Temperature and Angle Dependence Theory: The authors formulated a theory explaining how thermal fluctuations and impact angles interact. They showed that for ordered systems (MoS$_2$O), the threshold is highly sensitive to impact angle and temperature, whereas disordered systems (MoS$_2$F) lose this sensitivity.
4. Generalizability: Preliminary simulations on MoSe$_2$, WS$_2$, and WSe$_2$ suggest this mechanism is generalizable to other TMDs.

4. Key Results

A. Sputtering Threshold Reduction

• Pristine MoS$_2$: $E_{sputt,S} \approx 31 \pm 1$ eV.
• MoS$_2$O (Oxygen-functionalized): $E_{sputt,S} \approx 14.0 \pm 1$ eV.
• MoS$_2$F (Fluorine-functionalized): $E_{sputt,S} \approx 9.5 \pm 0.5$ eV.
• Mechanism:
  • MoS$_2$O: The Ar impact pushes an adsorbed O atom into a neighboring S atom (which also has an O attached), forming SO$_2$. This molecule then desorbs. The formation of SO$_2$ requires breaking fewer bonds and less energy than ejecting a pure S atom.
  • MoS$_2$F: Fluorine creates a more disordered surface (due to "Peierls distortions" from odd electron counts). The mechanism involves the formation of various SF$_n$ species. The high electronegativity of F weakens Mo-S bonds, further lowering the threshold.

B. Impact Angle and Order Sensitivity

• MoS$_2$O (Ordered): Shows strong angular dependence. The threshold drops non-monotonically, reaching a minimum of ~7.5 eV at an impact angle of $\theta = 30^\circ$. This is because the specific geometry allows the O atom to be pushed directly toward a neighbor S atom to form SO$_2$.
• MoS$_2$F (Disordered): Shows no significant angular dependence. The chaotic arrangement of F atoms means the specific impact direction matters less; the threshold remains low (~9.5 eV) regardless of angle.
• Pristine MoS$_2$: Also shows angular dependence, with a minimum around 30°, but the absolute energy remains high (~24–32 eV range).

C. Temperature Dependence

• The authors derived a theory where thermal fluctuations ($\sigma_{xy}$) of surface atoms create a spread in effective impact angles.
• As temperature increases, thermal fluctuations increase, effectively "smearing" the impact angle. If the thermal spread covers the optimal angle for sputtering, the required incident energy decreases.
• Prediction: At cryogenic temperatures (e.g., -200°C), the threshold approaches the theoretical minimum (e.g., ~13 eV for MoS$_2$O). At room temperature, thermal fluctuations lower the effective threshold further (e.g., to ~9 eV) because random thermal motions occasionally align the impact to the optimal sputtering geometry.

D. Selectivity and Damage Control

• Metal Preservation: No Molybdenum back-sputtering was observed up to 50 eV for functionalized surfaces. The layer destruction threshold remains near 100 eV.
• Result: Functionalization widens the "safe" processing window from a narrow 30–100 eV range to a much broader ~10–50 eV range, allowing for precise, low-energy etching.

5. Significance and Implications

• Process Optimization: This work provides a practical roadmap for plasma processing of 2D materials. By pre-functionalizing TMDs with O or F (e.g., via gas exposure or plasma pre-treatment), engineers can use lower-energy ion beams to selectively remove chalcogens without damaging the metal lattice.
• Spatial Control: The findings suggest that selective etching can be achieved by masking specific areas to control O/F coverage, allowing for the creation of sulfur vacancies with high spatial resolution.
• Theoretical Advancement: The development of a parameter-free theory linking thermal fluctuations to sputtering thresholds offers a new tool for predicting damage thresholds in other material systems without running extensive MD simulations for every temperature.
• Cryogenic Processing: The results highlight the importance of temperature control. Cryogenic processing could further refine the energy window, enabling even more selective damage control.

In conclusion, the paper establishes that oxygen and fluorine functionalization transforms the sputtering mechanism of MoS$_2$ from a high-energy atomic ejection process to a low-energy molecular desorption process, fundamentally enabling selective, damage-controlled plasma processing of 2D TMDs.