ChipFoundryServices
From Thermal Oxidation & Chemical Vapor Deposition to Atomic Layer Epitaxy & Sub-Nanometer Stacks

Film Formation University

The master science and process engineering of semiconductor thin-film formation: Deal-Grove thermal oxidation, LPCVD, PECVD, PVD magnetron sputtering, self-limiting Atomic Layer Deposition (ALD), Selective Epitaxial Growth (SEG), and atomic-scale barrier metallurgy.

7 Levels
Elementary to Fellow
21 Modules
Rigorous Curriculum
7 Sim Labs
Real-Time Engines
7 Diplomas
Industry Fellow Laureate
Academic Level 1 • Ages 6–10
Growing Atomic Blankets
Discover how chipmakers grow ultra-thin glass blankets and metal coats on silicon wafers.
Module 1.1

What is a Thin Film?

Imagine laying a microscopic blanket over a bed. On a silicon chip, a 'thin film' is an ultra-thin layer of material—sometimes just a few atoms thick—spread evenly across the shiny wafer.

These films can be electrical insulators like glass that block electric sparks, or electrical conductors like copper and titanium that let electricity zip through like lightning.

  • Thin Film: A microscopic layer of material deposited on a wafer surface.
  • Insulator vs. Conductor: Insulators prevent short circuits, while conductors carry electrical currents.
$$\text{Film Thickness } d \approx 1\text{ to } 100\,\text{nm} \quad (1\,\text{nm} = 10^{-9}\,\text{m})$$
Module 1.2

Baking Glass: Thermal Oxidation

When you leave an apple slice on a table, oxygen in the air turns it brown. Silicon does something even cooler: when heated to 1,000°C in pure oxygen, its surface turns into pure glass ($SiO_2$)!

This thermal oxide layer is one of the smoothest, most perfect electrical shields ever created by human science, protecting the silicon below from unexpected electrical leaks.

  • Thermal Oxidation: Heating silicon in oxygen to naturally grow a protective silicon dioxide glass film.
  • Passivation: Sealing atomic bonds so electricity cannot leak away.
$$\text{Silicon } (Si) + \text{Oxygen } (O_2) \xrightarrow{1000^\circ\text{C}} \text{Silicon Dioxide } (SiO_2)$$
Module 1.3

Raining Atoms: Vapor Deposition

What if we want to coat the wafer with metals or special crystals? We cannot just paint them on with a brush. Instead, we turn the material into a vapor gas inside a giant vacuum tube!

The floating atoms fly through the empty chamber and land softly onto the cool wafer, freezing into a mirror-like layer that is perfectly smooth from edge to edge.

  • Vapor Deposition: Turning solids into gas so atoms rain down evenly on the wafer.
  • Vacuum Chamber: An airless steel container where atoms can fly in straight lines without bumping into air molecules.
$$\text{Deposition Rate } R = \frac{\text{Thickness } (\Delta d)}{\text{Time } (\Delta t)}$$
⚡ Interactive Laboratory L1
Thermal Oxide Growth Time Calculator
Adjust the furnace temperature and oxidation time to calculate the resulting silicon dioxide film thickness.
Furnace Temperature (°C)1000
Oxidation Time (minutes)45
REAL-TIME SIMULATION TELEMETRY
Interactive physics simulator running client-side transfer models, carrier drift-diffusion kinetics, and boundary potential solvers.
Grown Oxide Thickness
48.5 nm
Interference Color
Light Brown / Tan
🎓 Level 1 Examination
Level 1 Conceptual Mastery Assessment
What happens when a silicon wafer is baked at 1,000°C in an oxygen atmosphere?
Why must vapor deposition chambers be kept under high vacuum?
What is the primary role of an insulating thin film on a silicon microprocessor?

Level 1 Completed: Growing Atomic Blankets Mastery Certificate

Conferred for mastery of Level 1 (Academic Level 1 • Ages 6–10) curriculum, simulation laboratory, and assessment evaluation.

Academic Level 2 • Ages 11–13
Chemical and Physical Vapor Deposition
Explore how chemical reactions (CVD) and physical ion bombardment (PVD) build transistor layers.
Module 2.1

Chemical Vapor Deposition (CVD)

In Chemical Vapor Deposition (CVD), we introduce reactive gases into a hot chamber. When the precursor gas molecules strike the hot wafer surface, a chemical reaction breaks them apart, leaving behind a solid film while waste byproducts fly out the exhaust.

For example, silane gas ($SiH_4$) reacting with oxygen produces solid silicon dioxide ($SiO_2$) and harmless hydrogen gas. CVD creates exceptionally conformal coatings that drape neatly over three-dimensional microscopic ridges.

  • Precursor Gas: High-purity gas carrying the chemical elements destined for the wafer.
  • Conformality: The ability of a thin film to coat top, bottom, and sidewalls with identical thickness.
$$SiH_4\text{ (gas)} + O_2\text{ (gas)} \xrightarrow{450^\circ\text{C}} SiO_2\text{ (solid)} + 2H_2\uparrow\text{ (gas)}$$
Module 2.2

Plasma-Enhanced CVD (PECVD)

What happens if our wafer already has delicate aluminum or copper wires that would melt at 700°C? We cannot bake them at high temperatures. Instead, we energize the gas using high-frequency radio waves to ignite a glowing plasma!

The energetic free electrons in the plasma smash into precursor molecules, breaking chemical bonds at room temperature or gentle warmth (200°C–400°C). This allows high-speed deposition without damaging heat-sensitive structures.

  • PECVD: Plasma-Enhanced Chemical Vapor Deposition operating at reduced thermal budgets.
  • Plasma Energy: Using RF electromagnetic fields to energize electrons rather than thermal heat alone.
$$\text{Thermal Budget } \Delta Q = \int k \cdot \exp\left(-\frac{E_a}{k_B T}\right) dt$$
Module 2.3

Physical Vapor Deposition (PVD Sputtering)

When we want to deposit pure metals like titanium, copper, or tungsten, we often use PVD magnetron sputtering. We fill the vacuum chamber with argon gas and turn on high voltage to create heavy argon ions ($Ar^+$).

Strong magnetic fields slam these heavy argon ions like microscopic billiard balls into a solid metal plate (the 'target'). The impact knocks out metal atoms, which fly across the chamber and condense into an ultra-pure metallic layer on the wafer.

  • Sputtering Target: High-purity metallic source disc (99.9999% purity).
  • Magnetron Confinement: Magnetic fields that trap electrons near the target to maximize argon ionization.
$$\text{Sputter Yield } Y = \frac{\text{Atoms Ejected}}{\text{Incident Ion }}$$
⚡ Interactive Laboratory L2
PVD Sputter Deposition Rate Simulator
Tune DC magnetron cathode power and argon working pressure to optimize metal deposition rate and film uniformity.
Magnetron DC Power (kW)8
Argon Working Pressure (mTorr)3.0
REAL-TIME SIMULATION TELEMETRY
Interactive physics simulator running client-side transfer models, carrier drift-diffusion kinetics, and boundary potential solvers.
Metal Deposition Rate
14.2 nm/min
Scattering Regime
Ballistic Transport (Low Scattering)
🎓 Level 2 Examination
Level 2 Conceptual Mastery Assessment
Why is PECVD preferred over conventional thermal CVD when coating wafers with pre-existing metal wiring?
In PVD magnetron sputtering, what physical projectile dislodges metal atoms from the cathode target?
What term describes a thin film that deposits with identical thickness across both horizontal surfaces and vertical sidewalls?

Level 2 Completed: Chemical and Physical Vapor Deposition Mastery Certificate

Conferred for mastery of Level 2 (Academic Level 2 • Ages 11–13) curriculum, simulation laboratory, and assessment evaluation.

Academic Level 3 • Ages 14–18
Kinetics of Film Growth: Deal-Grove & LPCVD
Formulate mathematical growth rates, diffusion limits, and chemical precursor activation.
Module 3.1

The Deal-Grove Oxidation Model

In 1965, Bruce Deal and Andy Grove published the governing equation of silicon oxidation. The reaction starts when oxidant molecules ($O_2$ in dry oxidation, or $H_2O$ in wet oxidation) diffuse across the gas boundary layer, diffuse through the already-grown oxide, and react at the silicon interface.

For ultra-thin films, the rate is limited by surface chemical reaction speed (linear regime). As the oxide thickens, incoming oxygen molecules must journey through a thick glass labyrinth, shifting the growth into a diffusion-limited parabolic regime.

  • Linear Rate Constant ($B/A$): Governed by interface chemical reaction velocity $k_s$.
  • Parabolic Rate Constant ($B$): Governed by oxidant diffusion coefficient $D_{ox}$ through the amorphous $SiO_2$ matrix.
$$x^2 + Ax = B(t + \tau) \implies x(t) = \frac{A}{2}\left[\sqrt{1 + \frac{4B(t+\tau)}{A^2}} - 1\right]$$
Module 3.2

Low-Pressure CVD (LPCVD) & Mass Transport

Operating CVD at low pressures (0.1 to 1.0 Torr) increases the mean free path of gas molecules by a factor of 1,000. Under these conditions, the gas diffusion coefficient $D_g$ escalates dramatically, eliminating mass-transfer gas-phase starvation.

LPCVD operates in the reaction-rate-limited regime. Because chemical reaction speed depends exponentially on temperature via the Arrhenius relation, wafers can be packed tightly like sliced bread in a hot-wall quartz tube while maintaining sub-1% thickness variation across 150 wafers.

  • Reaction-Limited Regime: Growth rate controlled strictly by wafer temperature rather than gas velocity.
  • Polysilicon & Nitride: LPCVD is the workhorse for $Si_3N_4$ masks and doped polycrystalline silicon.
$$R_{dep} = k_0 \cdot \exp\left(-\frac{E_a}{k_B T}\right) \cdot P_{precursor}^n$$
Module 3.3

Step Coverage & Aspect Ratio in Trenches

As transistor dimensions shrink, contact vias and isolation trenches become deep and narrow, quantified by their aspect ratio ($AR = ext{Depth} / ext{Width} > 10:1$). Physical deposition (PVD) struggles because line-of-sight trajectories cause metal atoms to build up at the trench opening ('overhang').

This overhang pinches off the top before the bottom fills, creating an empty cavity called a keyhole void. Chemical vapor deposition overcomes this through surface reactant desorption and re-emission, quantified by the sticking coefficient $S_c$. Lower sticking coefficients yield higher conformality.

  • Sticking Coefficient ($S_c$): Probability that a precursor molecule sticks on first impact versus bouncing back.
  • Keyhole Voiding: Catastrophic trapped vacuum bubble inside a contact plug caused by early top pinch-off.
$$\text{Step Coverage} = \frac{t_{\text{sidewall}}}{t_{\text{top}}} \approx \frac{1}{1 + \frac{3}{4} S_c \cdot AR}$$
⚡ Interactive Laboratory L3
Deal-Grove Wet vs Dry Oxidation Solver
Calculate oxide thickness comparing Dry Oxygen (O2) vs Wet Steam (H2O) across varying thermal durations.
Ambient EnvironmentWet Steam (H2O)
Furnace Temp (°C)1000
Time (Hours)2.0
REAL-TIME SIMULATION TELEMETRY
Interactive physics simulator running client-side transfer models, carrier drift-diffusion kinetics, and boundary potential solvers.
Calculated Oxide Thickness
512.4 nm
Dominant Deal-Grove Regime
Parabolic (Diffusion Limited)
🎓 Level 3 Examination
Level 3 Conceptual Mastery Assessment
In the Deal-Grove model, why does wet oxidation (H2O) exhibit an oxidation rate nearly an order of magnitude faster than dry oxidation (O2)?
Why does Low-Pressure Chemical Vapor Deposition (LPCVD) produce superior thickness uniformity across hundreds of wafers simultaneously?
What phenomenon causes keyhole void formation when filling high-aspect-ratio contacts with directional PVD sputtering?

Level 3 Completed: Kinetics of Film Growth: Deal-Grove & LPCVD Mastery Certificate

Conferred for mastery of Level 3 (Academic Level 3 • Ages 14–18) curriculum, simulation laboratory, and assessment evaluation.

Academic Level 4 • Undergraduate (Freshman–Sophomore)
Atomic Layer Deposition (ALD) Fundamentals
Master self-limiting surface chemistry, sequential precursor pulsing, and sub-angstrom control.
Module 4.1

Self-Limiting Surface Reactions

When transistor dimensions approached single nanometers, conventional CVD could no longer deliver the atomic precision required. Atomic Layer Deposition (ALD) revolutionized thin films by decoupling chemical reactions into two self-terminating half-cycles.

Instead of mixing precursors simultaneously, ALD injects Precursor A into the chamber. Gas molecules chemisorb onto available surface active sites until every site is occupied. Once saturated, no more molecules can bind—the reaction stops automatically, regardless of exposure time!

  • Self-Limiting Adsorption: Saturation of surface ligands guarantees exactly one sub-monolayer per pulse.
  • Precursor Purge: Inert gas ($N_2$ or $Ar$) evacuates all unreacted precursor before the next reactant arrives.
$$\text{Coverage } \theta(t) = 1 - \exp\left(-s_0 \cdot \frac{P \cdot t}{\sqrt{2\pi m k_B T}}\right)$$
Module 4.2

The Binary ALD Cycle ($Al_2O_3$ Paradigm)

The canonical ALD process is the synthesis of aluminum oxide ($Al_2O_3$) using trimethylaluminum ($Al(CH_3)_3$, TMA) and water vapor ($H_2O$). Pulse 1 introduces TMA, reacting with surface hydroxyl ($-OH$) groups, liberating methane ($CH_4$) gas, and leaving methyl-terminated aluminum atoms.

Following an inert nitrogen purge, Pulse 2 introduces water vapor ($H_2O$). Water molecules react with the methyl ligands, releasing more methane and regenerating fresh hydroxyl ($-OH$) sites, completing one full cycle and advancing the film by roughly 0.1 nm (1 Ångström).

  • TMA Reaction: $\text{Si-OH} + Al(CH_3)_3 \rightarrow \text{Si-O-}Al(CH_3)_2 + CH_4\uparrow$
  • Water Hydrolysis: $\text{Si-O-}Al(CH_3)_2 + 2H_2O \rightarrow \text{Si-O-}Al(OH)_2 + 2CH_4\uparrow$
$$\text{Growth Per Cycle (GPC)} \approx 0.9 - 1.1\,\text{Å/cycle} \quad (T_{substrate} = 200^\circ\text{C} - 300^\circ\text{C})$$
Module 4.3

The ALD Thermal Window & Conformality

ALD operates within a distinct temperature plateau known as the 'ALD Temperature Window'. Below this window, precursors may condense onto the wafer or react too sluggishly due to activation energy barriers. Above the window, precursors decompose thermally (reverting to uncontrolled CVD) or desorb prematurely.

Because the reaction is completely self-limiting, precursor molecules can diffuse into microscopic crevasses, 3D FinFET fins, and deep vertical holes with aspect ratios exceeding 100:1, coating every square nanometer with 100% conformal step coverage.

  • Thermal Window: Operating temperature range where GPC remains constant and self-limiting.
  • 3D Step Coverage: Flawless thickness equality on planar surfaces and deep vertical sidewalls.
$$\text{Total Thickness } d = N_{\text{cycles}} \times \text{GPC} \quad (100\%\text{ Conformal})$$
⚡ Interactive Laboratory L4
ALD Cycle and Saturation Simulator
Simulate TMA/H2O precursor pulse duration, purge efficiency, and cycle count to evaluate atomic thickness and unreacted ligand impurities.
Number of ALD Cycles150
Precursor Pulse Time (ms)80
Substrate Temp (°C)250
REAL-TIME SIMULATION TELEMETRY
Interactive physics simulator running client-side transfer models, carrier drift-diffusion kinetics, and boundary potential solvers.
Accumulated Film Thickness
15.0 nm
ALD Process State
Ideal Self-Limiting ALD Window
🎓 Level 4 Examination
Level 4 Conceptual Mastery Assessment
What foundational characteristic separates Atomic Layer Deposition (ALD) from conventional Chemical Vapor Deposition (CVD)?
What happens if an ALD chamber substrate temperature is set above the upper boundary of the ALD temperature window?
In the TMA/H2O ALD reaction, what gaseous reaction byproduct is exhausted during both the metal and oxidizer half-cycles?

Level 4 Completed: Atomic Layer Deposition (ALD) Fundamentals Mastery Certificate

Conferred for mastery of Level 4 (Academic Level 4 • Undergraduate (Freshman–Sophomore)) curriculum, simulation laboratory, and assessment evaluation.

Academic Level 5 • Advanced Undergraduate (Junior–Senior)
High-k Dielectrics & Barrier Metallurgy
Engineered multi-layers: hafnium oxides, metal gate electrodes, and film stress mechanics.
Module 5.1

High-k Dielectric Stacks ($HfO_2$, $ZrO_2$)

When gate oxide thickness reached 1.2 nm in $SiO_2$, quantum mechanical direct tunneling current caused catastrophic power dissipation. The solution was replacing $SiO_2$ ($\kappa pprox 3.9$) with High-k dielectrics like Hafnium Dioxide ($HfO_2$, $\kappa pprox 25$) or Zirconium Dioxide ($ZrO_2$).

A high dielectric constant enables a physically thicker insulating layer that suppresses quantum tunneling while delivering an ultra-thin Equivalent Oxide Thickness (EOT). ALD is universally required to deposit $HfO_2$ with sub-angstrom uniformity on top of a 0.5 nm $SiO_x$ interfacial chemical oxide.

  • Equivalent Oxide Thickness (EOT): Physical thickness of $SiO_2$ that would yield identical gate capacitance.
  • Direct Tunneling Suppression: Increasing physical barrier thickness $t_{phys}$ reduces leakage exponentially.
$$\text{EOT} = t_{\text{high-k}} \cdot \left(\frac{\kappa_{\text{SiO}_2}}{\kappa_{\text{high-k}}}\right) = t_{HfO_2} \cdot \left(\frac{3.9}{25}\right)$$
Module 5.2

Work-Function Metal Gates & Diffusion Barriers

High-k dielectrics suffered from Fermi level pinning and phonon scattering when paired with legacy polysilicon gates. Advanced nodes use Replacement Metal Gate (RMG) stacks consisting of titanium nitride (TiN), tantalum nitride (TaN), and aluminum-doped alloys (TiAlC).

These metal layers set the transistor threshold voltage ($V_{th}$) by aligning the metal work function ($\Phi_m$) near the silicon conduction band (nFET: $\sim 4.1\, ext{eV}$) or valence band (pFET: $\sim 5.1\, ext{eV}$). Ultra-thin ALD TiN also serves as an impermeable barrier preventing copper diffusion into silicon.

  • Work Function Tuning ($\Phi_m$): Precision composition engineering to adjust transistor switching thresholds.
  • Copper Diffusion Barrier: Amorphous ALD TaN/TiN preventing Cu migration and deep-level trap formation.
$$V_{th} = \Phi_{ms} + 2\phi_F + \frac{\sqrt{4\epsilon_s q N_A \phi_F}}{C_{ox}} - \frac{Q_{ox}}{C_{ox}}$$
Module 5.3

Thin-Film Stress, Wafer Bow & Stoney's Equation

Every deposited film develops residual mechanical stress—either tensile (contracting, curling the wafer concave) or compressive (expanding, bowing the wafer convex). This stress arises from thermal expansion mismatch ($\Delta lpha \Delta T$) and intrinsic microstructure growth defects.

Uncontrolled film stress causes wafer warpage, delamination, cracking, and photolithography defocus. Stoney's equation calculates the exact film stress $\sigma_f$ by measuring the change in wafer radius of curvature before and after film deposition.

  • Tensile vs Compressive Stress: Tensile pulls inward (+); compressive pushes outward (-).
  • Stoney's Formulation: Relates substrate Young's modulus $E_s$, Poisson ratio $ u_s$, and radius of curvature $R$.
$$\sigma_f = \frac{E_s \cdot t_s^2}{6(1 - \nu_s) \cdot t_f} \left(\frac{1}{R_{\text{post}}} - \frac{1}{R_{\text{pre}}}\right)$$
⚡ Interactive Laboratory L5
High-k EOT & Quantum Tunneling Leakage Solver
Calculate Equivalent Oxide Thickness (EOT) and direct tunneling gate leakage comparing SiO2 vs HfO2 dielectrics.
Dielectric MaterialHfO2 (k = 25.0)
Physical Thickness t_phys (nm)2.8
REAL-TIME SIMULATION TELEMETRY
Interactive physics simulator running client-side transfer models, carrier drift-diffusion kinetics, and boundary potential solvers.
Equivalent Oxide Thickness (EOT)
0.44 nm
Estimated Tunneling Leakage
1.2 x 10^-3 A/cm² (Immense Savings)
🎓 Level 5 Examination
Level 5 Conceptual Mastery Assessment
Why was Hafnium Dioxide (HfO2, k ≈ 25) introduced to replace Silicon Dioxide (SiO2, k = 3.9) in sub-45nm CMOS gates?
According to Stoney's equation, how does the substrate thickness (t_s) affect the measured wafer radius of curvature (R) for a given film stress?
In Replacement Metal Gate (RMG) architecture, what is the primary role of work function setting metals like TiN, TaN, and TiAlC?

Level 5 Completed: High-k Dielectrics & Barrier Metallurgy Mastery Certificate

Conferred for mastery of Level 5 (Academic Level 5 • Advanced Undergraduate (Junior–Senior)) curriculum, simulation laboratory, and assessment evaluation.

Academic Level 6 • Master of Science (M.S.) & Graduate
Epitaxy, MOCVD & Superlattice Stacks
Explore single-crystal selective epitaxial growth, bandgap engineering, and GAA sheet release precursors.
Module 6.1

Selective Epitaxial Growth (SEG) & Channel Strain

Unlike amorphous or polycrystalline deposition, epitaxial growth aligns newly arriving atoms perfectly with the underlying crystalline lattice. Selective Epitaxial Growth (SEG) utilizes chlorine chemistry ($HCl$) mixed with silane ($SiH_4$) or germane ($GeH_4$).

Chlorine selectively etches unnucleated clusters on dielectric oxide masks while allowing monocrystalline growth to proceed inside exposed silicon windows. Growing Silicon-Germanium ($Si_{1-x}Ge_x$) in pFET source/drain cavities exerts compressive lattice strain in the silicon channel, boosting hole mobility by over 300%.

  • Lattice Mismatch ($\Delta a / a$): Germanium atoms are larger than silicon, inducing lateral pseudomorphic strain.
  • Mobility Enhancement: Compressive strain splits the heavy and light hole valence subbands, slashing effective mass.
$$\epsilon_{\parallel} = \frac{a_{sub} - a_{epi}}{a_{epi}} \quad \implies \quad \mu_{p} \propto (1 + \beta \cdot \sigma_{xx})$$
Module 6.2

Si / SiGe Superlattices for GAA Nanosheet Release

Gate-All-Around (GAA) nanosheet transistors require alternating epitaxial superlattices of pure Silicon ($Si$) and Silicon-Germanium ($Si_{0.7}Ge_{0.3}$) grown with monolayer interface sharpness. Typically 3 to 4 periods are deposited continuously in reduced-pressure epitaxial reactors.

During later processing, a wet chemical or dry vapor etch selectively dissolves the sacrificial $SiGe$ sheets with a selectivity exceeding 150:1 over pure $Si$. This suspends isolated silicon nanosheet ribbons in midair, ready for ALD high-k metal gate wrapping around all four surfaces.

  • Monolayer Interface Abruptness: Transitioning between $Si$ and $SiGe$ within $< 0.5\, ext{nm}$.
  • Etch Selectivity Ratio: Rate of sacrificial $SiGe$ removal versus preservation of active $Si$ channel nanosheets.
$$\text{Selectivity } S = \frac{R_{\text{SiGe}}}{R_{\text{Si}}} > 150:1 \quad (\text{Vapor } CH_3COOH / H_2O_2 / HF)$$
Module 6.3

Metalorganic CVD (MOCVD) for Heterostructures

When integrating optical interconnects or compound semiconductors (III-V materials like $GaAs$, $InGaAs$, and $GaN$) onto silicon, Metalorganic Chemical Vapor Deposition (MOCVD) is the dominant industrial method. Precursors consist of volatile organometallic complexes like trimethylgallium ($Ga(CH_3)_3$) and arsine ($AsH_3$).

Heteroepitaxial growth across mismatched lattices requires graded buffer layers or aspect ratio trapping (ART) inside narrow dielectric trenches to bend and annihilate threading dislocations before they penetrate active device channels.

  • Aspect Ratio Trapping (ART): Trapping crystalline dislocations on vertical dielectric sidewalls.
  • Threading Dislocation Density (TDD): Must be suppressed below $10^6\, ext{cm}^{-2}$ for reliable device operation.
$$TDD = \frac{N_{\text{dislocations}}}{\text{Area}} \quad (\text{Critical Thickness: } h_c \approx \frac{b}{8\pi f}\frac{1-\nu \cos^2\theta}{(1+\nu)\cos\lambda}\ln\frac{h_c}{b})$$
⚡ Interactive Laboratory L6
SiGe Superlattice Critical Thickness & Strain Simulator
Calculate critical Matthews-Blakeslee thickness and channel stress as a function of Germanium molar fraction (x).
Germanium Mole Fraction x (%)30
SiGe Layer Thickness (nm)12
REAL-TIME SIMULATION TELEMETRY
Interactive physics simulator running client-side transfer models, carrier drift-diffusion kinetics, and boundary potential solvers.
Critical Thickness (h_c)
14.2 nm
In-Plane Channel Stress
-1.18 GPa (Compressive)
🎓 Level 6 Examination
Level 6 Conceptual Mastery Assessment
How does chlorine (HCl) enable selectivity during Selective Epitaxial Growth (SEG) of silicon in oxide windows?
What happens when an epitaxial film exceeds the Matthews-Blakeslee critical thickness (h_c)?
Why is an ultra-high selectivity (>100:1) required when etching sacrificial SiGe layers during GAA nanosheet channel release?

Level 6 Completed: Epitaxy, MOCVD & Superlattice Stacks Mastery Certificate

Conferred for mastery of Level 6 (Academic Level 6 • Master of Science (M.S.) & Graduate) curriculum, simulation laboratory, and assessment evaluation.

Academic Level 7 • Ph.D., Research Scientist & Technical Fellow
Frontier Atomic Layer Epitaxy & 2D Materials
Design area-selective ALD, 2D transition metal dichalcogenides, and sub-0.3nm EOT interfaces.
Module 7.1

Area-Selective Atomic Layer Deposition (AS-ALD)

At sub-2nm nodes, lithographic overlay errors ($< 1.5\, ext{nm}$) threaten edge placement yield. Area-Selective ALD (AS-ALD) achieves bottom-up self-aligned material placement by exploiting differential surface chemisorption kinetics between adjacent materials.

By applying Self-Assembled Monolayers (SAMs) or chemical passivation inhibitors (e.g. small molecule fluoro-silanes) that selectively bind to dielectric surfaces, precursor nucleation is blocked on dielectrics while proceeding unhindered on exposed metal features, eliminating alignment errors.

  • Passivation Inhibitors: Small-molecule blocking agents providing nucleation incubation delays exceeding 100 cycles.
  • Self-Aligned Via (SAV): Bottom-up metal growth preventing short circuits to adjacent metal interconnect lines.
$$\text{Selectivity } S = \frac{\theta_{\text{growth}} - \theta_{\text{non-growth}}}{\theta_{\text{growth}} + \theta_{\text{non-growth}}} \quad (S \to 1.00 \text{ at zero incubation loss})$$
Module 7.2

2D Transition Metal Dichalcogenides ($MoS_2$, $WS_2$)

As silicon nanosheets scale below 3 nm physical thickness, surface roughness scattering and quantum confinement collapse carrier mobility. 2D semiconductors—specifically transition metal dichalcogenides (TMDs) like monolayer $MoS_2$ and $WS_2$ ($0.65\, ext{nm}$ atomic thickness)—offer the ultimate scaling channel.

Because 2D TMDs possess pristine, dangling-bond-free surfaces, they exhibit near-zero surface roughness scattering and maintain high mobility even at sub-1nm body thickness. Synthesizing single-domain 300mm wafer-scale TMD films via low-temperature MOCVD or atomic layer epitaxy remains the premier grand challenge.

  • Monolayer Thickness: Single S-Mo-S atomic sandwich ($\sim 6.5 ext{ \AA}$) eliminating short-channel effects.
  • Contact Resistance ($R_c$): Overcoming Schottky barriers at metal-TMD junctions via semi-metallic bismuth ($Bi$) or antimony ($Sb$) contacts.
$$E_g^{1L} \approx 1.8\,\text{eV} \quad (\text{Direct Bandgap}) \quad \text{vs.} \quad E_g^{bulk} \approx 1.2\,\text{eV} \quad (\text{Indirect})$$
Module 7.3

Sub-0.3nm EOT Scaling & Ferroelectric HfZrO2 (FeFETs)

For 3D CFETs and advanced logic, gate dielectrics must scale below $EOT = 0.4\, ext{nm}$. This requires scavenging the interfacial $SiO_x$ layer using catalytic metal capping (e.g. titanium or aluminum scavenging gates) down to a zero-interfacial layer (zero-IL) regime.

Simultaneously, doping $HfO_2$ with zirconium ($Hf_{0.5}Zr_{0.5}O_2$, HZO) stabilizes an orthorhombic non-centrosymmetric crystalline phase ($Pca2_1$) that exhibits spontaneous, reversible ferroelectric polarization. This enables negative capacitance field-effect transistors (NC-FETs) that break the Boltzmann tyranny ($SS < 60\, ext{mV/decade}$) and non-volatile embedded FeFET memory.

  • Interfacial Scavenging: Remote oxygen extraction reducing the low-k interfacial oxide to zero.
  • Orthorhombic Phase Stabilization: Rapid thermal annealing with capping mechanical stress to induce ferroelectricity.
$$P(E) = P_s \tanh\left(\frac{E \pm E_c}{2E_0}\right) \quad \implies \quad SS = \left(\frac{\partial \log_{10} I_d}{\partial V_g}\right)^{-1} < 60\,\text{mV/dec}$$
⚡ Interactive Laboratory L7
Area-Selective ALD Selectivity & Incubation Simulator
Simulate AS-ALD precursor nucleation delay on inhibited dielectric vs active metal surfaces to maximize edge placement yield.
SAM Inhibitor Packing Density (%)99.5
Deposition Cycles60
REAL-TIME SIMULATION TELEMETRY
Interactive physics simulator running client-side transfer models, carrier drift-diffusion kinetics, and boundary potential solvers.
Calculated Selectivity (S)
0.992 (Yield Clean)
Parasitic Nucleation Density
2.4 defects/µm²
🎓 Level 7 Examination
Level 7 Conceptual Mastery Assessment
In Area-Selective ALD (AS-ALD), what metric defines the operational process window before loss of selectivity occurs?
Why are 2D transition metal dichalcogenides (like monolayer MoS2) considered the ultimate successor to silicon nanosheets in sub-1nm CMOS channels?
What crystallographic phase must be stabilized in doped HfO2 (such as HZO) to induce non-volatile ferroelectric polarization in FeFET devices?

Level 7 Completed: Frontier Atomic Layer Epitaxy & 2D Materials Mastery Certificate

Conferred for mastery of Level 7 (Academic Level 7 • Ph.D., Research Scientist & Technical Fellow) curriculum, simulation laboratory, and assessment evaluation.

🏅
Distinguished Fellow in Thin-Film Kinetics & Atomic Layer Deposition
Highest academic honor conferred by ChipFoundryServices OS for demonstrated mastery across all 7 curriculum tiers, interactive simulation laboratories, and verified examination standards.