← Back to Business Simulations Semiconductor Metrology & Polarimetry

Spectroscopic Ellipsometry Lab

Simulate thin film optical metrology, ellipsometric angles (ρ = tan(Ψ)·eiΔ), complex refractive index dispersion (N(λ) = n(λ) + i·k(λ)), Bruggeman roughness, Mueller matrix polarimetry, and 20-year fab metrology tool DCF economics.

Metrology Presets:
Ellipsometric Amplitude (Ψ)
21.4°
Phase Difference (Δ)
142.8°
Optical Constants (n, k)
n = 1.457, k = 0.000
Film Thickness + Roughness
2.0 nm (+0.2 nm)
Angle of Incidence
70.0° (Brewster ≈ 75°)
Model Fit Quality (MSE)
0.85 (Good Fit)
Wafer Throughput
45 WPH (3-site inline)
20-Year Metrology Tool NPV
$14.8M

Optical & Metrology Parameters

Film Physical Thickness (d) 2.0 nm
Angle of Incidence (Φ0) 70.0°
Spot Inspection Wavelength (λ) 632.8 nm (He-Ne)
Surface Roughness (Bruggeman EMA) 0.2 nm
Input Polarizer Azimuth Angle (P) 45.0°

Fab Metrology Economics & Tool DCF

Tool Capital Expenditure (CapEx) $2.8M
Wafer Throughput (WPH) 45 WPH
Fab Yield Excursion Scrap Reduction 1.2%
Discount Rate (WACC) 9.0%

Optical & Spectroscopic Visualizer

Internal Rate of Return
34.2%
Discounted Payback
2.4 Years
Return on Investment
428%

Spectroscopic Ellipsometry Engineering & Physics Curriculum

1. Fundamental Equations & Fresnel Reflection Ratios

Spectroscopic ellipsometry measures the change in polarization state of light upon reflection from a planar interface. The fundamental measurement parameter is the complex reflectance ratio ρ:

ρ = rp / rs = tan(Ψ) · eiΔ

where tan(Ψ) is the amplitude ratio of reflected p-polarized (in-plane) to s-polarized (perpendicular) electric field components, and Δ represents the phase shift between the two orthogonal polarizations.

For a multi-layer stack containing ambient air, thin film, and silicon substrate, the total reflection coefficients are determined via the Airy multiple internal reflection formula:

rp = (r01,p + r12,p·e-2iβ) / (1 + r01,p·r12,p·e-2iβ)

where β = 2π(d/λ) √(N12 - N02 sin2Φ0) is the film phase thickness, linking physical thickness d directly to optical interference oscillations.

2. Dispersion Modeling & Bruggeman Roughness

In dielectric thin films transparent across the visible spectrum (such as SiO2, Si3N4, and high-k HfO2), the refractive index dispersion is accurately modeled using the empirical Cauchy relation:

n(λ) = A + B / λ2 + C / λ4

For absorbing films such as amorphous carbon hardmasks, an extinction coefficient k(λ) = α·eβ/λ is added, forming the complex refractive index N = n + ik.

Real wafer surfaces exhibit sub-nanometer roughness that distorts polarization measurements. The Bruggeman Effective Medium Approximation (EMA) treats roughness as an equivalent layer containing 50% film and 50% void (n=1.0):

f1 · (N12 - Neff2) / (N12 + 2Neff2) + fvoid · (1 - Neff2) / (1 + 2Neff2) = 0

Solving for Neff eliminates correlation between surface roughness and bulk film thickness during Levenberg-Marquardt regression.

3. Mueller Matrix Polarimetry & OCD Scatterometry

When thin film layers exhibit optical anisotropy, cross-polarization, or spatial depolarization from patterned 3D nanostructures (FinFET fins, gate all around nanosheets, or 3D NAND holes), conventional ellipsometry is generalized to Mueller matrix polarimetry:

Sout = M · Sin = [mij]4×4 · [S0, S1, S2, S3]T

For non-depolarizing isotropic films, the normalized Mueller matrix elements reduce to N = -cos(2Ψ), C = sin(2Ψ)·cos(Δ), and S = sin(2Ψ)·sin(Δ), satisfying the identity N2 + C2 + S2 = 1.

4. 20-Year Fab Metrology DCF Valuation Framework

Foundry deployment of spectroscopic ellipsometry tools requires balancing CapEx ($1.5M - $4.5M) and maintenance with catastrophic yield excursion avoidance:

  • Inline Yield Gain: Sub-angstrom gate dielectric and CMP thickness control prevents 0.5% to 2.5% wafer scrap across high-volume production lines.
  • High-Throughput Recipe Automation: Up to 120 wafers per hour with multi-site micro-spot (25 μm) scanning enables 100% statistical process control without bottlenecking lithography clusters.
  • Long-Term DCF Horizon: High reliability optics and diode lasers support 15 to 20 years of operational lifecycle across multiple technology nodes.

20-Year Fab Metrology Tool DCF Cash Flow Schedule

Year Wafer Volume Gross Scrap Savings Tool Maintenance & OpEx Net Free Cash Flow Discount Factor Discounted FCF Cumulative NPV

Formative Knowledge Assessment: Optical Metrology & Polarimetry

Question 1

Why does spectroscopic ellipsometry achieve sub-angstrom thickness sensitivity despite optical wavelengths being hundreds of nanometers?

Question 2

What optical event occurs at the Brewster angle of incidence during reflection?

Question 3

How does the Bruggeman EMA approximation prevent erroneous film thickness measurements?

Frequently Asked Questions: Spectroscopic Ellipsometry

Spectroscopic ellipsometry measures the change in polarization state of light upon oblique reflection from a sample surface. It determines the complex reflectance ratio rho = rp / rs = tan(Psi) * exp(i * Delta), where tan(Psi) represents the relative amplitude attenuation between p-polarized and s-polarized light, and Delta represents the phase difference induced by the reflection. Because ellipsometry measures a ratio rather than absolute intensity, it is self-calibrating and insensitive to source intensity fluctuations.

At the Brewster angle of the substrate (approximately 75 degrees for crystalline silicon at visible wavelengths), the p-polarized reflection coefficient approaches a minimum and undergoes an abrupt 180-degree phase transition. Measuring near this angle maximizes the contrast between rp and rs, yielding extreme sensitivity to sub-angstrom changes in overlayer film thickness and delicate interface layers.

Microscopic surface roughness creates an optical transition layer that can be mistaken for bulk film thickness. The Bruggeman Effective Medium Approximation models roughness as a discrete top layer composed of a 50 percent film material and 50 percent void (air or vacuum) mixture. By fitting both the roughness overlayer and the underlying film thickness simultaneously, optical constants and true layer thickness are extracted without correlation artifacts.

Standard ellipsometry assumes isotropic, non-depolarizing planar films where cross-polarization is zero. Full 4-by-4 Mueller matrix polarimetry measures how the sample alters all four Stokes parameters. This reveals optical anisotropy, sample depolarization caused by non-uniform film thickness or patterning, cross-polarization terms, and grating diffraction parameters essential for Optical Critical Dimension (OCD) scatterometry in 3D NAND and FinFET architectures.

An automated in-line 300mm spectroscopic ellipsometry tool represents an initial CapEx of 1.5 to 4.5 million dollars. Key operating drivers include high throughput (80 to 120 wafers per hour across multiple test sites), sub-angstrom repeatability, and non-destructive recipe automation. Preventing out-of-spec gate oxide or CMP thickness excursions delivers tens of millions of dollars in annual scrap avoidance, yielding rapid payback within 1.5 to 3 years.

Related Semiconductor & Metrology Simulations

X-Ray Photoelectron Spectroscopy Lab

Model XPS core binding energies, ARXPS non-destructive depth profiling, inelastic mean free path, and fab metrology DCF.

Dynamic SIMS & Dopant Profiling Lab

Model Cs/O2 primary ion sputtering, matrix effect ionization, sub-nanometer junction depth resolution, and fab DCF.

Helium Ion Microscopy & GFIS Lab

Model Gas Field Ion Source physics, sub-0.5nm probe sizes, secondary electron yield, and EUV mask repair.

Atomic Layer Deposition (ALD) Lab

Model self-limiting surface chemisorption, precursors, ALD saturation curves, and high-k dielectric fab DCF.

X-Ray Reflectivity (XRR) Lab

Model XRR critical angles, Kiessig interference fringes, film density, interface roughness, and fab DCF.

All Business Simulations & Worksheets →

Browse our comprehensive directory of over 890 engineering, physics, financial, and manufacturing simulation labs.