Rutherford Backscattering Spectrometry (RBS) for Thin Film Structure Evaluation

Rutherford Backscattering Spectrometry (RBS) is a quantitative, non-destructive ion beam analysis technique widely used to evaluate thin film structure, including elemental composition, layer thickness, stoichiometry, and depth profiles. This technical article explains the physical principles of RBS, typical sample types, measurable parameters, testing methods, and industrial applications. It serves as a practical reference for engineers, researchers, and quality managers seeking reliable thin film characterization at accredited third-party laboratories.

Technical Overview

Rutherford Backscattering Spectrometry (RBS) is an ion beam analysis technique in which a collimated, monoenergetic beam of light ions—most commonly 1.0 to 3.0 MeV helium ions—is directed onto a sample surface under high vacuum. A small fraction of these ions undergoes elastic Coulomb (Rutherford) scattering with atomic nuclei in the near-surface region, and the energy spectrum of the backscattered ions is recorded by a silicon surface-barrier or PIPS detector positioned at a large scattering angle, typically 150° to 180°.

The energy of each backscattered ion is governed by the kinematic factor, which depends on the mass of the target atom and the scattering geometry. Because the kinematic factor increases monotonically with atomic mass, the measured energy spectrum directly identifies the elements present in the film. The absolute scattering cross section is known from fundamental physics, which makes RBS inherently quantitative: areal densities are obtained in atoms per square centimeter without the need for matrix-matched reference standards.

Depth information is derived from the continuous energy loss of ions as they penetrate and exit the sample (electronic stopping power). The cumulative energy loss converts the energy axis of the spectrum into a depth scale, enabling non-destructive depth profiling. Typical depth resolution is 5 to 20 nm near the surface, degrading gradually with depth, and the maximum analyzable depth is approximately 1 to 2 micrometers for helium beams and deeper for proton beams.

RBS offers the highest sensitivity for heavy elements embedded in lighter matrices, with detection limits reaching the ppm to sub-percent range for medium and heavy elements. Sensitivity to light elements such as carbon, nitrogen, and oxygen is lower but can be significantly enhanced using resonant (non-Rutherford) scattering conditions, while hydrogen is measured by the complementary Elastic Recoil Detection (ERD) technique.

• Non-destructive analysis with no sample dissolution or sputtering artifacts
• Standardless quantification with typical accuracy of 1% to 5% for heavy elements
• Areal density measurement independent of material density assumptions
• Depth profiling capability from the surface to approximately 1–2 micrometers
• Compatibility with conductive, semiconducting, and insulating films

Testing Samples

RBS thin film evaluation is applicable to a broad range of materials deposited on silicon, glass, quartz, metals, ceramics, and polymers. Samples are typically analyzed as-received, with minimal preparation, provided the surface is flat, reasonably smooth, and representative of the film structure of interest.

Semiconductor Thin Films and Gate Stacks

High-k dielectric layers such as HfO2, ZrO2, Al2O3, and HfSiON; metal gate electrodes; silicide contacts (NiSi, CoSi2, TiSi2); epitaxial layers; and ion-implanted or annealed wafer sections used in advanced CMOS and power device process development.

Metal Films and Diffusion Barriers

Metallization stacks containing Al, Cu, W, Ti, Pt, Au, and Ni, together with barrier and liner layers such as TiN, TaN, WN, and Ru, where layer integrity, interfacial reaction, and elemental interdiffusion must be verified after deposition or thermal processing.

Optical, Conductive, and Protective Coatings

Transparent conductive oxides (ITO, AZO), antireflection and filter stacks (SiO2, TiO2, Ta2O5, Si3N4), and hard or tribological coatings including diamond-like carbon (DLC), TiN, CrN, and other nitrides and carbides.

Energy and Functional Materials

Photovoltaic absorbers such as CIGS, CZTS, CdTe, and amorphous silicon; solid-state electrolytes (for example, garnet-type LLZO); battery electrode coatings; thermoelectric films; and magnetic multilayers such as Co/Pt and Fe/Cr used in spintronic devices.

Surface-Engineered and Special Materials

Nitrided, carburized, oxidized, or ion-implanted surface layers; coatings on aerospace and automotive components; and heritage or archaeological metal objects where non-destructive compositional depth profiling is required.

Testing Items

RBS thin film structure evaluation reports a set of quantitative structural parameters derived from simulation-assisted spectral analysis. Typical testing items include:

• Elemental composition and stoichiometry of single or multilayer films (reported in atomic percent or atomic ratios)
• Film thickness and areal density (nm and atoms/cm²; areal density is density-independent and directly traceable)
• Composition depth profiles showing how elemental concentrations vary with depth
• Interfacial characterization, including intermixing, reaction layers, and diffusion tails between adjacent layers or substrates
• Quantification of dopants, heavy-element impurities, and surface contamination
• Film density determination by combining RBS areal density with independently measured physical thickness
• Crystalline quality assessment through ion channeling, including lattice disorder, substitutional fraction, and strain-related dechanneling in single-crystal films
• Hydrogen and deuterium content by ERD, and light-element (C, N, O) depth profiling by resonant scattering

Testing Methods

Measurements are performed on accelerator-based ion beam systems housed in accredited third-party laboratories operating under recognized quality frameworks such as ISO/IEC 17025. The laboratory selects beam species, energy, geometry, and detection configuration according to the elements of interest, layer thicknesses, and required sensitivity.

Standard RBS Measurement

A 1.5 to 3.0 MeV helium beam impinges on the sample at a defined incidence angle, and backscattered ions are detected at a fixed scattering angle. Beam current, integrated charge, and solid angle are precisely recorded so that absolute areal densities can be calculated. Energy calibration of the detection chain is verified using reference samples with known surface elements before each measurement campaign.

Elastic Recoil Detection Analysis (ERDA)

For hydrogen and deuterium, which cannot be measured by conventional RBS, ERDA is employed. A helium beam ejects recoil hydrogen atoms from the sample, which are detected after passing through a thin stopping foil that blocks the scattered helium. This provides quantitative hydrogen depth profiles, critical for evaluating a-Si:H photovoltaic layers, moisture barriers, and hydrogen-rich dielectrics.

Non-Rutherford (Resonance) Scattering

The scattering cross sections of light elements exhibit sharp nuclear resonances at specific helium energies—for example, the well-known 16O resonance near 3.045 MeV and resonances for 14N and 12C at higher energies. Operating at or scanning across these resonances increases sensitivity to oxygen, nitrogen, and carbon by one to two orders of magnitude, enabling accurate stoichiometry of oxides, nitrides, and carbides.

Channeling RBS

For single-crystal films and substrates, the beam is aligned with a major crystallographic axis. Backscattering yields drop dramatically for well-ordered lattices, allowing quantification of lattice disorder, epitaxial quality, and the substitutional fraction of dopants or alloying elements.

Spectrum Simulation and Data Reduction

Measured spectra are fitted with recognized simulation codes such as SIMNRA or RUMP using layer models with assumed compositions. The software calculates theoretical spectra from stopping powers and scattering cross sections, and layer thicknesses and compositions are iteratively adjusted until the simulated spectrum matches the experimental data. Final results are reported with combined uncertainty budgets, typically 1% to 5% for heavy-element composition and a few percent for thickness.

Applications

RBS thin film structure evaluation supports research, development, failure analysis, and process qualification across many industries:

• Microelectronics: qualification of high-k gate stacks, silicide formation studies, barrier-layer integrity after annealing, and verification of ion implantation dose and junction integrity
• Photovoltaics: compositional optimization and uniformity control of CIGS and CdTe absorbers, evaluation of Na incorporation, and interface diffusion analysis in thin film solar cells
• Optics and displays: stoichiometry and thickness verification of multilayer interference filters, AR coatings, and transparent conductive oxides
• Hard and protective coatings: quality control of DLC, nitride, and carbide coatings, including composition drift detection across deposition batches
• Energy storage: characterization of solid electrolyte films, protective coatings on cathodes, and electrode-electrolyte interphases
• Surface engineering: depth profiling of nitrided, carburized, and implanted engineering surfaces for wear and corrosion performance correlation
• Academic and failure analysis studies: diffusion kinetics, thin film reaction studies, and cross-validation of sputter-profiling techniques such as SIMS, AES, and XPS

FAQ

What is the typical depth resolution and maximum analysis depth of RBS?

Near-surface depth resolution is typically 5 to 20 nm depending on beam energy, geometry, and detector resolution, and it degrades gradually with depth due to energy straggling. Standard helium-beam RBS analyzes films up to roughly 1 to 2 micrometers thick; proton beams extend the range but reduce mass resolution.

Can RBS detect hydrogen and other light elements?

Hydrogen and deuterium are measured by Elastic Recoil Detection (ERD) rather than backscattering. Carbon, nitrogen, and oxygen can be quantified by RBS with enhanced sensitivity using resonant scattering at specific beam energies, achieving detection limits far better than standard Rutherford conditions.

Is RBS a destructive technique?

RBS is essentially non-destructive for most inorganic films: the ion dose is low and leaves no crater or sputtered area, and the same sample spot can be reanalyzed or subjected to subsequent processing. Sensitive organic materials and some polymers may experience minor beam-induced modification, which laboratories mitigate by spreading the beam or limiting dose.

How does RBS compare with SIMS, XPS, and TEM?

RBS provides absolute, standardless quantification and is most sensitive to heavy elements, whereas SIMS offers superior trace-level detection limits but requires standards and suffers from matrix effects. XPS surveys the top few nanometers with chemical-state information, and TEM gives atomic-scale imaging but limited quantitative composition. The techniques are complementary, and RBS is frequently used to calibrate or validate depth profiles obtained by sputter-based methods.

What sample size and form are required?

Typical requirements are flat pieces of several millimeters to about two centimeters, sufficiently smooth to avoid geometric broadening. Full wafers can often be accommodated with appropriate holders, and small fragments, diced coupons, or coated pieces are acceptable. Samples must be vacuum compatible.

How accurate are RBS thickness and composition results?

With careful calibration and simulation, elemental fractions of medium and heavy elements are typically determined with 1% to 5% relative uncertainty, and areal density with similar precision. Converting areal density to physical thickness requires either a known or assumed film density; when density is uncertain, laboratories report areal density as the traceable quantity.

How long does an RBS measurement take?

A single spectrum is usually collected in 10 to 30 minutes depending on the required statistics. Including setup, calibration, simulation, and reporting, standard turnkey evaluations are commonly completed within a few working days at accredited third-party laboratories.