Rutherford Backscattering Spectrometry (RBS) Thickness Determination: Principles, Methods, and Appli

Rutherford Backscattering Spectrometry (RBS) thickness determination is a non-destructive ion-beam analysis technique that measures the thickness, areal density, and elemental composition of thin films and coatings with absolute accuracy and without requiring calibration standards. This technical article explains the physical principles of RBS, the types of samples suitable for analysis, the standard measurement workflow, spectrum interpretation, and practical applications across the semiconductor, coating, and advanced materials industries.

Technical Overview

Rutherford Backscattering Spectrometry (RBS) is a quantitative surface and thin-film analysis technique in which a collimated beam of light ions—most commonly helium ions (He+ or He2+) accelerated to energies between approximately 1.0 and 3.0 MeV—is directed onto a specimen under high vacuum. A small fraction of the incident ions undergoes elastic Coulomb scattering with atomic nuclei inside the sample and is deflected backward. A silicon surface-barrier or passivated planar detector, typically positioned at a scattering angle between 150° and 180°, records the energy distribution of these backscattered particles with an energy resolution on the order of 10–20 keV.

Each detected ion energy carries two independent pieces of information. First, the kinematic factor, which is governed by the mass ratio between the projectile ion and the target nucleus, identifies the element from which the ion scattered; heavier target atoms produce backscattered ions of higher residual energy. Second, the continuous energy loss that the ion experiences while penetrating into and escaping from the sample—described by the material's stopping power—converts the detected energy axis into a depth scale. The integrated yield of scattered ions is directly proportional to the areal density of atoms through the Rutherford scattering cross section, which is calculable from fundamental physics.

Because both the scattering cross section and the stopping powers are known from first principles, RBS is one of the few analytical methods capable of standardless, absolute quantification. Once the atomic density of a film is known or assumed, the measured areal density (atoms/cm²) is converted directly into a film thickness (nm or µm). Key performance characteristics of RBS thickness determination include:

• Near-surface depth resolution of approximately 5–20 nm, degrading gradually with depth due to energy straggling
• Probing depth of roughly 1–2 µm for standard helium beams, extendable to several µm with heavier or higher-energy projectiles
• Absolute areal density accuracy of approximately 1–5% without reference samples
• Detection limits in the range of 10¹⁴–10¹⁵ atoms/cm² for medium and heavy elements in light matrices
• Simultaneous determination of thickness, composition, and depth distribution in a single measurement

RBS measurements are performed in accredited third-party laboratories and research facilities worldwide, following the harmonized terminology and specimen-handling practices defined in ISO surface chemical analysis standards, including ISO 18115 (vocabulary), ISO 18116 (preparation and mounting of specimens), and ISO 18117 (handling of specimens prior to analysis).

Testing Samples

Metallic Thin Films and Metallization Layers

Single or stacked metal films such as aluminum, copper, nickel, titanium, platinum, gold, and their alloys deposited on silicon, glass, or polymer substrates are classic RBS specimens. The technique accurately determines film thickness, detects interfacial reactions (for example, silicide formation between a metal layer and a silicon substrate), and quantifies alloy composition gradients.

Semiconductor Thin Films and High-k Dielectrics

Gate oxides, oxynitrides, high-permittivity dielectrics such as hafnium oxide and hafnium silicate, epitaxial compound semiconductor layers, and ion-implanted regions are routinely characterized. RBS thickness determination is particularly valued for ultrathin dielectric films in the 2–50 nm range, where the stoichiometry and areal density must be controlled precisely for electrical performance.

Multilayer Structures and Diffusion Barriers

Alternating multilayers used in X-ray optics, magnetic recording media, and decorative or hard coatings can be profiled layer by layer. Barrier layers such as tantalum, titanium nitride, or ruthenium between a substrate and a conductor are measured for thickness and checked for interdiffusion after thermal processing.

Coatings on Light Substrates

RBS exhibits its highest sensitivity for heavy elements on light matrices. Metallized polymers, anti-reflective silica coatings on glass, anodized layers, conversion coatings, and thin metallic films on carbon or plastic substrates are therefore ideal candidates, offering strong signal contrast and reliable quantification.

Specimen Requirements

• Typical specimen dimensions from a few millimeters square up to full 100–200 mm wafers, depending on the stage configuration
• A flat, reasonably smooth surface; excessive roughness broadens spectral edges and degrades depth resolution
• Vacuum compatibility: the material must be stable under a pressure of approximately 10⁻⁶–10⁻⁷ mbar
• No special conductive coating is required, although insulating samples may need electron flood compensation or a thin conductive contact path to suppress charging

Testing Items

RBS thickness determination services typically cover the following measurable quantities:

• Film thickness of single layers, expressed in nanometers or micrometers, derived from the measured areal density and film density
• Areal density (dose) in atoms/cm², reported as the primary absolute quantity
• Elemental composition and stoichiometry of alloys, oxides, nitrides, and compound films
• Depth profiles showing how composition varies with depth in multilayers and graded layers
• Interface analysis: intermixing, reaction-layer thickness, and diffusion-zone characterization
• Retained ion implantation dose and junction depth verification
• Native oxide or surface contamination layer thickness on metals and semiconductors
• Heavy-element impurity identification and quantification within light matrices
• Crystal quality assessment of epitaxial layers using RBS/channeling, where the aligned versus random yield ratio quantifies lattice disorder

Testing Methods

Specimen Preparation and Mounting

Samples are cleaned if necessary using solvent rinse or nitrogen blowing, avoiding any treatment that could alter the film. The specimen is attached to a conductive holder with carbon tape or mechanical clips and loaded into the analysis chamber. Chamber pressure is brought to the low 10⁻⁶ mbar range or better before beam exposure. Preparation and mounting practices follow the guidance of ISO 18116 and ISO 18117 to ensure that contamination or handling artifacts do not compromise the result.

Measurement Configuration

A typical configuration uses a 2.0 MeV He+ beam with a beam current of 5–50 nA, a beam spot of approximately 1 mm², and a detector at a scattering angle of 165–170° with a solid angle of a few millisteradians. Normal incidence is standard for maximum depth; tilting the sample by 30–60° can improve near-surface depth resolution when very thin films must be resolved. For full quantification, the geometrical parameters (incident angle, scattering angle, solid angle) and the accumulated beam dose are recorded, because they enter directly into the absolute yield calculation.

Spectrum Acquisition

A spectrum is accumulated for 10–30 minutes until the region of interest contains sufficient statistics, typically 10³ counts or more at the film signal edge. Energy calibration of the detection chain is performed using a reference material with known scattering edges. Insulating specimens are monitored for charge buildup, and neutralization measures are applied if peak shifts or instability are observed.

Spectrum Analysis and Thickness Extraction

The measured energy spectrum is interpreted by comparing it with a simulated spectrum generated from a model sample structure. Dedicated simulation software calculates the kinematic factors, scattering cross sections, and energy losses—including straggling—for an assumed stack of layers, and the model is iteratively refined until the simulation matches the measured spectrum. From the converged model, the analyst extracts each layer's areal density, composition, and equivalent thickness. The leading and trailing edges of each elemental signal directly reflect the surface and interface positions, and the energy separation between these edges provides the film's energy loss thickness, which converts to physical thickness through the stopping power and assumed bulk density.

Quality Control and Reporting

Results are reported with the measurement geometry, beam energy, total accumulated charge, the simulated layer model, and estimated uncertainties. Routine quality controls include calibration spectra on reference samples, replicate measurements to verify repeatability, and cross-checks of areal density against complementary methods such as X-ray fluorescence, ellipsometry, or transmission electron microscopy cross-sections where appropriate. Measurement traceability and reporting practices are aligned with internationally recognized surface-analysis methodology published by ISO and ASTM technical committees.

Applications

RBS thickness determination is applied across a broad spectrum of industries and research fields:

• Semiconductor manufacturing: gate dielectric thickness, silicide formation control, high-k/metal gate stack characterization, and implanted dose verification in process development
• Metallization and plating: qualification of sputtered or electroplated layers, barrier-layer integrity, and thickness uniformity validation when optical and stylus methods are unsuitable
• Optical and functional coatings: compositional depth profiling of multilayer interference filters, absorber stacks in solar thermal collectors, and transparent conductive oxides
• Data storage and spintronics: layer-by-layer metrology of magnetic multilayers and tunnel-barrier thickness control
• Corrosion and oxidation science: quantification of native oxide growth, passivation layer thickness, and high-temperature oxide scales
• Photovoltaics: composition and interface analysis of thin-film absorbers such as cadmium telluride, CIGS, and amorphous silicon stacks
• Ion implantation process control: absolute retained-dose measurement, independent of sputtering or matrix effects that affect other techniques
• Cultural heritage and forensics: non-destructive thickness measurement of gilding, plating, and decorative coatings on artifacts
• Academic materials research: calibration standard for other thin-film techniques, thanks to its standardless absolute quantification

FAQ

What thickness range can RBS measure?

RBS comfortably measures films from roughly 1–2 nm up to about 1–2 µm for helium beams. Very thin layers are quantified as areal density even when the front and back edges are not fully separated, while films thicker than the ion range require higher beam energies or complementary cross-sectional methods.

Is RBS destructive?

For metals, semiconductors, ceramics, and most inorganic films, RBS is effectively non-destructive, and the same area can be re-measured. Some organic materials, polymers, and certain salts may suffer beam-induced modification, in which case dose is minimized and the exposure is verified by repeat scans.

Which elements can be detected?

RBS detects every element heavier than the projectile ion. Sensitivity scales approximately with the square of the atomic number, so the technique excels at trace heavy elements in light matrices. Light elements such as carbon, nitrogen, and oxygen are quantifiable but with reduced sensitivity, and hydrogen is not detected by conventional RBS; it is measured instead by the complementary elastic recoil detection (ERDA) variant.

How accurate is RBS thickness determination?

Areal density is determined absolutely, typically within 1–3%, because the Rutherford cross section and stopping powers are known from physics rather than calibration. The conversion to linear thickness carries an additional uncertainty from the assumed film density, yielding an overall thickness accuracy of typically 2–5%.

How does RBS compare with other thickness measurement methods?

Unlike ellipsometry or reflectometry, RBS does not require optical model assumptions; unlike X-ray fluorescence, it provides true depth information; and unlike cross-sectional electron microscopy, it needs no destructive sample preparation. The trade-offs are a depth resolution of only 5–20 nm near the surface and reduced sensitivity for light elements in heavy matrices.

Does surface roughness affect the results?

Yes. Roughness broadens spectral edges, which smears the apparent interface and degrades depth resolution. Areal density remains reliable, but for highly textured or granular films the reported linear thickness should be interpreted as an equivalent thickness based on the assumed density.

What sample sizes are accepted?

Most instruments accept specimens from small fragments a few millimeters across up to whole wafers of 100–200 mm diameter. Samples must be flat, vacuum-stable, and mounted so that the region of interest is accessible to the beam at the chosen incidence angle.

How long does an RBS measurement take?

A single spectrum is typically acquired in 10–30 minutes, and a complete analysis including modeling, reporting, and any replicate or tilt-angle measurements is usually completed within a standard laboratory turnaround of a few working days.