Grazing-Incidence Metrology Block Copolymer DSA

GISAXS Nanopatterning & DSA Metrology Lab

Simulate Grazing-Incidence Small-Angle X-ray Scattering (GISAXS) for Directed Self-Assembly (DSA) pitch determination, Yoneda peak enhancement, DWBA scattering rods, and commercial laboratory DCF modeling.

2D Detector Color Pattern (q_y vs q_z) Yoneda Peak (α_f = α_c) DSA Line-Space Pitch (L₀) & LER 20-Year Laboratory DCF
Full Pitch (L₀)

--

Domain Period
1st Bragg Rod

--

In-Plane Momentum
Correlation (ξ)

--

Lateral Grain Size
Film Critical α_c

--

Yoneda Threshold
Lab 20Y NPV

--

Discounted Cash Flow
Project IRR

--

Internal Return Rate

GISAXS Beam & Pattern Parameters

24.0 nm
6.0 nm (Sub-2nm Node) 60.0 nm
12.0 nm
1.20 nm
0.2 nm (Perfect crystal) 4.0 nm (Disordered)
0.20°
0.10° (Evanescent) 0.45° (Transmission)

GISAXS Metrology Business Model

$650,000
$600
750 scans

2D GISAXS Detector Scattering Pattern

q_y vs q_z Color Map
Extracted Metric Measured Value Physical Significance

1D In-Plane Intensity Cut Along Yoneda Line I(q_y)

20-Year GISAXS Metrology Beamline DCF Cash Flow Schedule

Timeline Scan Volume Gross Revenue Operating Expenses EBITDA Free Cash Flow (FCF) Cumulative Balance

Engineering Guide: Grazing-Incidence Small-Angle X-ray Scattering (GISAXS)

1. Grazing-Incidence Kinematics & Evanescent Penetration

In conventional transmission SAXS, X-rays pass through the entire substrate, creating immense background scatter that obscures thin surface films. GISAXS circumvents this limitation by operating at an incident grazing angle \(\alpha_i\) near the critical angle of total external reflection (\(\alpha_c \approx \sqrt{2\delta} \approx 0.15^\circ\text{--}0.25^\circ\)). The wavefield inside the film becomes an evanescent wave with a shallow penetration depth:

\(\Lambda = \frac{\lambda}{\sqrt{2}\pi} \left[ \sqrt{(\alpha_c^2 - \alpha_i^2)^2 + 4\beta^2} + (\alpha_c^2 - \alpha_i^2) \right]^{-1/2}\)

For \(\alpha_i < \alpha_c\), penetration is restricted to \(5\text{--}10\text{ nm}\), enabling surface-sensitive probing of sub-10nm block copolymer Directed Self-Assembly (DSA) without penetrating deep into the silicon wafer.

2. The Yoneda Peak & Critical Enhancement

When scattered photons emerge at an exit angle matching the film's critical angle (\(\alpha_f = \alpha_{c,\text{film}}\)), the transmitted electric field amplitude inside the film peaks sharply due to boundary resonance. Under the Distorted Wave Born Approximation (DWBA), the transmission function:

\(|T(\alpha_f)|^2 = \left| \frac{2 \sin\alpha_f}{\sin\alpha_f + \sqrt{\sin^2\alpha_f - 2\delta + 2i\beta}} \right|^2\)

reaches a value of \(\approx 4.0\). This creates an intense horizontal band of diffuse scattering across the detector known as the **Yoneda peak**, where signal-to-noise ratios are maximized.

3. Distorted Wave Born Approximation (DWBA) & Bragg Rods

Unlike transmission diffraction governed by simple kinematic Born scattering, GISAXS requires the DWBA framework. Four scattering pathways interfere constructively:

  • Direct scattering from the nanostructure.
  • Reflection from substrate followed by nanostructure scattering.
  • Nanostructure scattering followed by substrate reflection.
  • Double reflection from the substrate before and after scattering.

For vertically oriented DSA line-space lamellae, the finite vertical boundary condition relaxes the out-of-plane Laue requirement, forming continuous vertical **Bragg rods** at in-plane scattering positions:

\(q_y = \pm \frac{2\pi m}{L_0}\)

4. Paracrystalline Structure Factor & LER Extraction

The 1D horizontal intensity profile along the Yoneda line is modeled by the product of the form factor \(P(q_y)\) and 2D paracrystalline structure factor \(S(q_y)\):

\(S(q_y) = \frac{1 - |p|^2}{|1 - p|^2}, \quad p = \exp\left(i q_y L_0 - \frac{1}{2} q_y^2 \sigma_p^2\right)\)

The peak width directly yields the lateral grain correlation length \(\xi \approx 2\pi / \Delta q_y\), while higher-order harmonic damping quantifies line-edge roughness (\(\sigma_p\)) over billions of nanostructures non-destructively.

Frequently Asked Questions About GISAXS Metrology

Grazing-Incidence Small-Angle X-ray Scattering (GISAXS) illuminates a thin film at an extremely shallow angle of incidence (alpha_i ≈ 0.15° to 0.30°), close to the critical angle of total external reflection (alpha_c). Because the incident beam undergoes total external reflection, the X-ray wave penetrates only 5 to 50 nm as an evanescent wave, dramatically enhancing signal sensitivity from surface nanostructures while suppressing substrate background noise. Collecting diffuse scatter on a 2D area detector reveals nanometer-scale lateral periodicity (q_y) and vertical film morphology (q_z) averaged over macroscopically large surface areas (square millimeters).

The Yoneda peak is an intense horizontal band of diffuse scattering that occurs when the exit angle equals the critical angle of the film or substrate (alpha_f = alpha_c). Under Fresnel optics and the Distorted Wave Born Approximation (DWBA), the transmission coefficient T(alpha_f) of the scattered wave reaches a sharp mathematical maximum at the critical angle due to constructive boundary interference. This field enhancement amplifies diffuse scattering from interfacial roughness and nanostructures, providing a clean signature of electron density and film composition.

In Directed Self-Assembly (DSA), perpendicular lamellar block copolymers form periodic stripes that extend through the thickness of the film (20 to 50 nm). Because the structural ordering is two-dimensional in-plane but truncated out-of-plane by the film interfaces, the out-of-plane Laue diffraction condition is relaxed. As a result, the discrete Bragg diffraction peaks found in 3D crystals elongate into continuous vertical intensity streaks known as Bragg rods at in-plane positions q_y = ±2*pi*m / L_0, where L_0 is the full domain pitch.

The in-plane width (full-width at half-maximum, Delta q_y) of the first-order Bragg diffraction rod is directly governed by paracrystalline lattice disorder and domain grain size. By Scherrer's formula adapted to 2D paracrystals, the lateral correlation length is inversely proportional to peak width: xi ≈ 2*pi / Delta q_y. Furthermore, the intensity decay of higher-order diffraction peaks (2nd and 3rd harmonics) follows a Debye-Waller-like damping factor exp(-q_y^2 * sigma_p^2), enabling non-destructive statistical quantification of line-edge roughness (LER) and pitch fluctuations across billions of DSA features.

A high-performance laboratory GISAXS beamline costs approximately $650,000 in capital expenditures, featuring a microfocus rotating anode or liquid metal jet X-ray source, multilayer Montel collimation mirrors, motorized 6-axis sample goniometer, vacuum flight path, and a hybrid photon-counting 2D detector (such as Dectris EIGER). Operating costs average $54,000 annually. Advanced packaging and sub-2nm foundry clients pay $600 to $900 per qualification scan for non-destructive pitch and defectivity metrology. Operating at 750 to 1,150 scans annually generates $450,000 to $700,000 in gross revenue, delivering an internal rate of return (IRR) exceeding 64% with capital payback in approximately 1.6 years.

Knowledge Check: GISAXS & Nanopatterning Metrology

Related Advanced Characterization & Lithography Labs

SECM Microscopy Lab

Model Scanning Electrochemical Microscopy, UME approach curves, and electrocatalytic kinetics.

SThM Thermal Lab

Model Scanning Thermal Microscopy, 3-omega tip heating, contact resistance, and 20nm hotspot thermometry.

XAFS & EXAFS Lab

Model XANES oxidation states, EXAFS coordination shells, radial distributions, and benchtop XAFS DCF.

CD-SAXS Metrology Lab

Model transmission Critical Dimension SAXS, 3D grating sidewall angles, and sub-2nm node metrology.

EUV Stochastic Defect Lab

Model photon shot noise, line-edge roughness (LER), microbridge pinch-off defects, and die yield.