Ellipsometry measures the thickness and optical constants of thin films, coatings, and photoresist layers on semiconductor wafers. The technique is non-contact and non-destructive. It works by analyzing the change in polarization state of light reflected from a layered surface. This article explains the measurement principle, sample preparation, workflow, accuracy, and applications.

Principle & mechanism

Ellipsometry measures how polarized light changes its polarization state when it reflects from a sample surface. The instrument records two quantities: psi and delta. These describe the amplitude ratio and phase difference between the parallel and perpendicular polarization components of the reflected light. The values depend on the optical constants of each layer. They also depend on interference between light reflected at the film surface and at the film–substrate interface. The phase information is highly sensitive to film thickness on the nanometer scale. This lets the technique resolve layers far thinner than the wavelength of the probing light. A single wavelength or a spectroscopic scan may be used. Spectroscopic ellipsometry sweeps a broad wavelength range and gives richer information for multilayer stacks. No physical contact with the sample is required, so the measurement leaves the film intact.

Sample

Types and Preparation Requirements
Ellipsometry applies to many sample types, provided the surface is smooth enough to produce specular reflection. Common specimens include semiconductor wafers with dielectric or metallization layers, photoresist-coated wafers, optical coating stacks, polymer films on glass or silicon, and protective coatings on metals. Samples should be flat. The reflective area must be at least as large as the measurement spot. Patterned or rough surfaces scatter light and degrade the signal. Before measurement, the surface must be free of dust, residues, and moisture. Clean handling with gloves and, where appropriate, filtered nitrogen blowing is standard practice. For polymer and resist layers, storage history matters. Residual solvent, post-exposure baking, or standing time can change the refractive index. A witness coupon or unpatterned region of the wafer is often preferred, since patterned areas complicate modeling.

Measurement

Workflow and Modeling Methods
The workflow begins with instrument calibration against a reference sample. This is typically a bare silicon wafer with a known native oxide layer. The operator then defines the measurement spots, sets the incident angle, and records the spectra. Data reduction is an inverse problem. A layer model is constructed with an assumed thickness and a dispersion relation. The Cauchy equation is common for transparent films. Absorbing layers need a more complex oscillator model. Fitting software adjusts the model parameters until the calculated psi and delta curves match the measured data. The quality of fit is judged by the mean squared error. Thickness, refractive index, and extinction coefficient are obtained simultaneously from the converged model. For photoresist layers, a separate measurement of the bare substrate is usually taken first. Where several films are stacked, fitting proceeds layer by layer. Ambiguous solutions are checked by measuring at multiple angles.

Accuracy, Repeatability, and Sensitivity

Thickness results from ellipsometry are model-dependent. Accuracy depends on how well the optical model represents the real layer structure. On smooth, well-defined single films, sub-nanometer sensitivity is routinely achievable. Repeatability on a stable sample is typically at the angstrom level. Several factors limit accuracy. These include surface roughness not included in the model, lateral thickness non-uniformity within the measurement spot, uncertainty in substrate optical constants, and absorbing layers whose thickness and extinction coefficient are correlated. Sensitivity also varies with film thickness. Very thin layers constrain mainly the optical constants. Films thicker than several wavelengths can generate multiple valid solutions. Good practice includes multi-angle or multi-wavelength measurement. Report the fit error together with the thickness value. Repeat measurements at several positions to characterize uniformity across the wafer.

Co-Testable Film Parameters

Beyond thickness, a single ellipsometric measurement yields the optical constants of the film. These are the refractive index and the extinction coefficient as functions of wavelength. Additional characteristics can be derived from them. The dispersion behavior of a transparent layer can indicate material density or porosity differences between process conditions. For absorbing films, the absorption edge and band-gap-related parameters can be estimated from spectroscopic data. Surface roughness can be modeled as an effective medium layer and reported as an equivalent roughness thickness. In multilayer stacks, the thickness of each layer can be extracted separately. This requires sufficient optical contrast between adjacent layers. For resist materials, index changes after exposure or baking reveal the degree of chemical conversion. This links the measurement to process control in Photoresist Testing workflows.

Application

Scenarios and Acceptance Criteria
Ellipsometry is widely used in semiconductor process development and qualification. Gate dielectrics, antireflection coatings, and resist thickness must be verified against target values. Optical coating manufacturers use it to confirm layer designs. Laboratories apply it to polymer and functional films, an area adjacent to Plastic Film Testing services. Acceptance is typically defined as a tolerance band around the nominal thickness. This may be a few percent or a few nanometers, depending on the layer and its function. A specified uniformity across the wafer or batch is also required. The report should state the model used, the fitted thickness and optical constants, the measurement angles and wavelength range, and the goodness of fit. When a result will gate a production lot, agree the measurement conditions and acceptance limits with the requesting party befOre testing begins.

Quick Answers

Frequently Asked Questions

01

What

does an ellipsometry report for thin film and resist thickness include, and how is it used?
The report typically covers measured film thickness, refractive index values, the optical model applied, and goodness-of-fit indicators. Customers use these results to verify process deposition or lithography targets, compare batches, and support quality control or failure analysis decisions within their manufacturing or research workflows.

02

How should samples be submitted for thin film and resist thickness measurement by ellipsometry, and what should be communicated beforehand?

Submit samples with clean, reflective surfaces meeting preparation requirements. Specify the films or resist layers of interest, their approximate thickness ranges, and substrate details. Communicating the layer stack, measurement locations, and intended data use ensures appropriate modeling methods and measurement workflows are selected for reliable results.

03

What

factors affect the cost of ellipsometry testing for thin film and resist thickness?
Cost depends on the number of film layers and whether known optical constants are available. Complex multilayer stacks require more elaborate modeling and analysis time. Sample count, required measurement locations, and whether additional co-testable parameters such as refractive index or absorption are requested also influence overall testing cost.

← Previous Article Seat belt testing
Next Article → Security door testing

Ready to Discuss Your Testing Needs?

Contact our team for a customized quote and expert consultation on your Ellipsometry for Thin Film and Resist Thickness: Principles testing requirements.

Contact Our Team
n n n n n n n n n n