Atomic Layer Etching Simulator

Watch atoms disappear layer by layer with angstrom-level precision. Self-limiting chemistry meets digital control for ultimate uniformity and damage-free processing.

Auto-Running Layer-by-Layer Animation

The simulation starts automatically, removing atomic layers one at a time. Observe Langmuir adsorption saturation, self-limiting behavior, and sub-nanometer precision.

ALE Process Parameters

Cycles Completed

0
cycles

Layers Removed

0
layers

Total Depth

0.0
nm

EPC (Etch/Cycle)

0.0
Å

Surface Coverage

0
%

Precision (3σ)

0.0
Å

Throughput

0.0
nm/min

Current Phase

Idle

3D Atomic Lattice - Layer Removal Animation

Atomic Layer Etching Physics

Langmuir Self-Limiting Adsorption

The foundation of ALE is self-limiting surface chemistry:

Surface coverage (fractional): θ = (K × P) / (1 + K × P) K: Equilibrium constant = k_ads / k_des P: Precursor partial pressure (mTorr) At saturation: θ → 1 (100% coverage) Kinetic form: dθ/dt = k_ads × P × (1 - θ) - k_des × θ Solution: θ(t) = θ_sat × [1 - exp(-t/τ)] τ = 1/(k_ads × P + k_des)

Etch Per Cycle (EPC)

Fundamental ALE metric:

EPC = θ_sat × d_monolayer × Y_removal θ_sat: Saturated coverage (ideally 1.0) d_monolayer: Atomic layer thickness Al₂O₃: ~2.8 Å/layer Si: ~3.1 Å SiO₂: ~3.2 Å Y_removal: Removal yield (ideally 1.0) Typical values: Thermal ALE: 0.3-1.5 Å/cycle Plasma ALE: 0.5-3.0 Å/cycle Precision: 3σ < 0.1-0.3 Å

Sequential Ligand Exchange

Example: Al₂O₃ ALE with TMA/HF:

Step 1: Modification Al₂O₃-OH + Al(CH₃)₃ → Al₂O₃-O-Al(CH₃)₂* + CH₄(g) Self-limiting: All -OH sites consumed Step 2: Purge (Ar/N₂) Remove excess TMA and byproducts Step 3: Removal Al₂O₃-O-Al(CH₃)₂* + 2HF → AlF₃(g) + CH₄(g) Self-limiting: Only modified layer removed Step 4: Purge Net: ~2.8 Å Al₂O₃ removed per cycle

Thermal vs Plasma ALE

THERMAL ALE: - Modification: Chemical adsorption - Removal: Chemical reaction - Temperature: 200-400°C - Damage: None (purely chemical) - Anisotropy: Isotropic - EPC control: Excellent (0.05 Å σ) - Rate: 0.5-2 nm/min PLASMA ALE: - Modification: Chemi/physisorption - Removal: Ion bombardment (20-100 eV) - Temperature: 20-300°C - Damage: Minimal (below threshold) - Anisotropy: Directional - EPC control: Good (0.1-0.3 Å σ) - Rate: 2-10 nm/min

Saturation Curve Characterization

Vary t_dose: 0.5 to 20 s Measure EPC vs t_dose Fit: EPC(t) = EPC_sat × [1 - exp(-t/τ_sat)] Extract: - EPC_sat: Maximum EPC - τ_sat: Time constant Criteria: t_dose > 3τ_sat for 95% saturation Optimize: Balance speed vs uniformity

Uniformity and Conformality

Wafer uniformity: σ/μ < 1% (3σ) Mechanism: Self-limiting → rate independent of flux Conformality: Thermal ALE: AR > 100:1 (neutral diffusion) Plasma ALE: AR ≈ 10-20:1 (ion limitations) High AR optimization: - Longer dose times - Pulsed removal - Lower pressure (longer λ_mfp)

Process Window Optimization

Key parameters: 1. t_mod: 3-5 × τ_sat (saturation) 2. t_rem: Just enough for removal 3. t_purge: 3-5 × volume replacement 4. T: Material dependent (200-400°C) Multi-objective optimization: Maximize: EPC, uniformity, selectivity, throughput Minimize: Damage, roughness, cost

Selectivity Mechanisms

Infinite selectivity: Material A: Modified → etches Material B: Not modified → no etch Examples: 1. Al₂O₃ vs TiO₂ (TMA/HF): >100:1 2. Si vs SiO₂ (Cl₂/Ar): 20-50:1 3. SiGe vs Si: Tunable by Ge% Applications: - Gate oxide trim - Fin reveal - Spacer recess - Self-aligned contacts

Damage Characterization

ALE advantages: Physical damage: 10-100× lower than RIE E_ion: 20-100 eV (vs 100-500 eV) Roughness: RMS < 3 Å (atomic smooth) Mechanism: Layer-by-layer vs stochastic Sidewall damage: Minimal RIE: 1-3 nm damaged layer Electrical: D_it < 10^10 cm⁻²eV⁻¹ RIE: 10^11-10^12 Characterization: AFM, TEM, XPS, electrical testing