Thin film thickness measurements – guidelines for method selection

Thin film thickness is a key parameter that directly influences a material’s electrical, mechanical, and optical behavior. Precise verification of thickness is therefore necessary across a range of industries, including semiconductors, display manufacturing, medical devices, and advanced electronics.

Film thickness can be measured using several methods, such as XRR, cross-sectional SEM and TEM, ellipsometry, AFM, and profilometry. This article summarizes the advantages and limitations of each technique, along with the practical factors that should be considered when selecting the most suitable approach for a given type of film.

Factors to consider during method selection

Expected thickness is typically the primary deciding factor when selecting a measurement technique, as each method operates reliably only within a defined measurement window. Surface roughness and the material composition of the film and substrate also influence the array of suitable methods.

When several methods are suitable, the selection can be based on practical factors, such as the range of additional information provided beyond thickness, the extent of sample preparation required, destructiveness, cost, and throughput. 

Table 1 summarizes the most relevant differences between commonly used measurement techniques. More detailed information is provided in subsequent sections.

Thickness measurements with XRR

X-ray reflectivity (XRR) is a non-destructive method that is particularly useful for determining the thickness, roughness, and density of individual layers in complex multilayer film stacks. It resolves this information by analyzing the interference pattern created when X-rays reflect from the surface and internal interfaces of layers with different densities.

XRR is suitable for a thickness range between 1–250 nm, with the highest accuracy typically achieved below ~150 nm. Above that, the density and surface roughness of thicker films can still be measured, but thickness determination becomes less reliable. 

Accurate XRR results also require a smooth surface, typically with a roughness below ~5 nm, and some prior knowledge of the sample’s composition and layer sequence. Because the analysis relies on model fitting, incorrect assumptions about density or structure can lead to large errors. When these conditions are met, XRR offers high precision across a wide range of thin film materials, including metals, oxides, carbides, and nitrides.

Thickness measurements with cross-sectional SEM

Cross-sectional scanning electron microscopy (SEM) is a direct imaging method for determining thin film thickness in single films and multilayer stacks. It is particularly useful for analyzing complex geometries and very rough films not suitable for XRR. The applicable thickness range is also larger, from approximately 20 nm to 100 µm, making SEM suitable for both nanoscale coatings and comparatively thick functional layers.

In addition to the thickness, SEM also provides information on morphology, grain structure, and interface quality of the film. Elemental composition can also be determined if the microscope is equipped with an energy-dispersive X-ray spectroscopy (EDX) detector.

Cross-sectional SEM is destructive and requires more preparation than surface-based techniques, as the sample must be cut with a method such as focused or broad ion beam (FIB or BIB) to expose the film thickness. Non-conductive films also require a thin conductive coating, such as gold, platinum, or carbon, before they can be imaged.

Thickness measurements with cross-sectional TEM

Cross-sectional transmission electron microscopy (TEM) provides direct, nanometer-scale thickness measurements for single films and complex multilayer stacks. One of its greatest advantages is the ability to image single atomic layers within multilayer structures, a capability that none of the other methods featured here can achieve. The upper end of the measurable thickness range is roughly 500 nm. 

In addition to thickness, TEM  also provides valuable insight into the structure of the material, including grain boundaries, crystallinity, and interface diffusion. Crystal structure and phase information can be acquired with electron diffraction, and elemental composition with either EDX or electron energy loss spectroscopy (EELS), which allows the analysis of light elements such as C, N, and O. 

TEM requires an extremely thin lamella to be prepared with FIB, as samples must be electron-transparent. This means that the comprehensive range of obtainable information comes at a cost of both money and time. It is also important to ensure sample suitability before analysis, as the high-voltage beam can damage polymeric and organic films. 

Thickness measurements with ellipsometry

Spectroscopic ellipsometry is commonly used for fast thickness measurements on transparent and semitransparent single-layer films and multilayer film stacks. It is a non-destructive optical method suitable for film thicknesses ranging from approximately 1 nm to 10 µm, and can be used to map thickness variations across entire wafers. In addition to thickness, ellipsometry provides information on optical parameters such as the refractive index (n), extinction coefficient (k), and anisotropy. 

Ellipsometry’s operating principle is based on measuring changes in the polarization of reflected light and fitting the data to a model of the sample. For reliable fitting, approximate values for the optical constants of each layer and a reasonable estimate of the stack structure should be known in advance. Accurate results also require a relatively smooth surface, with roughness below ~100 nm. Metallic films are generally unsuitable for ellipsometry as they are not transparent, whereas oxides, nitrides, polymers, and other dielectric films are measured routinely.

Thickness measurements with AFM and profilometry

Atomic force microscopy (AFM) and stylus profilometry determine thin film thickness by measuring the height difference between a coated region and an adjacent uncoated part of the substrate. This requires a partial coating that creates a “step” between the substrate and the film. AFM is suitable for thicknesses from approximately 1 nm to 10 µm, while stylus profilometry extends the measurable range up to about 1 mm.

AFM records a three-dimensional topography map using a nanoscale probe, while profilometry traces a height profile with a mechanical stylus. In addition to thickness, both techniques provide surface roughness data and topographical images. Because measurement is based on a physical step, results do not depend on optical properties or density contrast, making the techniques suitable for virtually any material. However, in the case of multilayer film stacks, only the total thickness can be determined rather than that of individual layers. Obtaining reliable results also depends on preparing a clean and well-defined step without edge damage.

One partner for all your thin film analysis needs

Measurlabs offers thin film thickness measurements and related characterization services using all the techniques discussed above. Our experts can also help select the most suitable method for your samples and applications.

You can find the technical details and indicative price ranges for commonly requested services through these links:

• XRR of thin films or coatings

• FIB preparation for cross-sectional SEM or TEM

• SEM-EDX imaging

• HR-TEM imaging

• AFM surface imaging

• 3D optical profilometry

Alternatively, just fill in the form below to describe your samples and testing needs. We will get back to you in one business day.