RBS Surface Analysis: Rutherford Backscattering Spectrometry for Thin Film and Depth Profiling Chara

RBS surface analysis (Rutherford Backscattering Spectrometry) is a powerful, non-destructive ion beam technique that delivers standardless quantitative measurement of elemental composition, film thickness, and depth profiles in thin films and layered materials. With absolute accuracy typically within a few percent and depth resolution on the order of 10–20 nanometers, RBS has become the reference method for compositional verification in semiconductors, optical coatings, and advanced functional materials. This article explains the principles, sample requirements, measurable parameters, analytical variants, and typical applications of RBS surface analysis.

Technical Overview

Rutherford Backscattering Spectrometry (RBS) is a quantitative surface and thin film analysis technique in which a beam of high-energy ions, most commonly 1.5–3 MeV helium ions (He²⁺ or He+), is directed onto a sample surface under high vacuum. A small fraction of these ions undergo elastic collisions with the nuclei of atoms in the sample and are scattered backward toward a detector, typically placed at a scattering angle of about 160–170 degrees. Because the energy of each backscattered ion depends on the mass of the target atom it collided with and on the depth at which the collision occurred, the measured energy spectrum encodes both the elemental identity and the depth distribution of atoms in the sample.

The quantitative power of RBS rests on three well-established physical relationships. First, the kinematic factor determines how much energy an ion loses when it backscatters from a nucleus of a given mass; heavier target atoms cause less energy loss, so their signal appears at higher energies in the spectrum. Second, ions lose energy continuously as they travel into and out of the sample through electronic stopping, which converts the energy axis into a depth scale. Third, the Rutherford scattering cross section is known precisely and scales approximately with the square of the atomic number, which allows absolute quantification of atomic areal densities without calibration standards.

Key performance characteristics of conventional RBS include a probing depth of roughly 1–2 micrometers in typical materials, a near-surface depth resolution of approximately 5–20 nanometers depending on geometry and beam energy, and detection limits that range from about 0.001 at% for very heavy elements to several atomic percent for light elements. The technique is essentially non-destructive, requires no matrix-matched standards, and provides results that are traceable to fundamental physical constants, which is why RBS is frequently used to certify reference materials and to cross-check other analytical methods.

Complementary Ion Beam Techniques

Because standard RBS is relatively insensitive to very light elements, it is often combined with related ion beam methods in the same analytical session. Elastic Recoil Detection Analysis (ERDA or ERD) detects forward-recoiled atoms and enables quantification of hydrogen and other light elements. Nuclear Reaction Analysis (NRA) uses ion-induced nuclear reactions to measure light elements such as nitrogen, oxygen, carbon, and boron with high selectivity. Heavy-ion RBS (HI-RBS) improves mass resolution and light-element sensitivity by using heavier probing ions. Ion channeling, discussed below, extends RBS to crystallographic analysis.

Testing Samples

Thin Films and Coated Substrates

The ideal RBS sample is a smooth, flat, laterally uniform film on a flat substrate. Typical examples include metal films, metal silicides, oxides, nitrides, high-k dielectrics, and diffusion barrier layers on silicon wafers, glass, sapphire, or quartz. Film stacks from a few nanometers up to roughly one micrometer are within the direct probing range, and thicker multilayers can often be characterized by combining measurements at multiple beam energies.

Multilayer and Graded Structures

Multilayer stacks, superlattices, compositionally graded layers, and ion-implanted materials are well suited to RBS because the technique resolves overlapping layers through their characteristic energy signatures. Interfaces can be evaluated for intermixing, interdiffusion, and reaction after thermal processing.

Bulk Solids and Other Forms

Bulk solids, polished cross sections, ceramics, and glasses can be analyzed directly as long as they are vacuum compatible. Powders must be pressed into pellets and results should be interpreted with care due to surface roughness. Liquids, volatile materials, and biological tissues are generally not suitable for routine RBS without special preparation.

Sample Requirements and Preparation

• Recommended lateral size: approximately 5 mm x 5 mm to 15 mm x 15 mm; larger samples may be accommodated depending on the chamber stage.
• Surface roughness below a few nanometers is preferred, since roughness broadens spectral features and degrades depth resolution.
• Samples must be compatible with high vacuum (approximately 10^-6 to 10^-7 mbar) and free of outgassing oils, loose particles, and contaminants.
• Strongly insulating samples can usually be measured using charge integration or electron flood compensation to manage charging effects.
• No conductive coating or chemical staining is required, preserving the sample for further testing.

Testing Items

RBS surface analysis delivers several distinct, quantitative results from a single spectrum:

• Elemental composition and stoichiometry (atomic percentages) of films and layers, with absolute accuracy typically better than 3–5%.
• Film thickness and areal density expressed in atoms per square centimeter, convertible to nanometers using the measured composition and assumed or known density.
• Depth concentration profiles of elements from the surface to approximately 1–2 micrometers depth.
• Interface analysis, including interdiffusion, intermixing width, reaction layer formation, and layer abruptness.
• Implanted ion dose (atoms/cm²) and implanted depth distribution, used to verify ion implantation processes.
• Atom areal density of heavy trace elements with detection limits approaching 10¹³–10¹⁴ atoms/cm² for the heaviest elements on light substrates.
• Hydrogen and deuterium content via ERDA, and light-element quantification via NRA when required.
• Crystalline quality assessment via ion channeling: minimum yield, lattice disorder profiles, impurity lattice location, and thickness of amorphized layers.

Testing Methods

Standard RBS Measurement Procedure

In a typical experiment, the sample is mounted on a goniometer stage inside a scattering chamber connected to a particle accelerator such as a single-ended or tandem electrostatic accelerator. A collimated beam of 2 MeV He²⁺ ions with a current of a few to tens of nanoamperes is rastered over an analysis area of roughly 1 mm² to ensure representative sampling. Backscattered ions are detected by a silicon surface barrier or passivated implanted planar silicon (PIPS) detector at a fixed angle near 170 degrees. The collected charge is measured precisely to normalize the spectrum, and the resulting energy histogram is converted into concentration versus depth through model-based simulation.

Spectrum Simulation and Data Analysis

Quantitative interpretation is performed by comparing the measured spectrum with simulated spectra generated by software such as SIMNRA or RUMP. The analyst constructs a layered model containing layer thicknesses and compositions, and iteratively refines the model until simulated and measured spectra agree. This approach allows deconvolution of overlapping elemental signals and yields error estimates for every fitted parameter. Certified reference materials or well-characterized standards may be included in the batch to confirm accuracy.

Channeling RBS

For single-crystal samples such as Si, GaAs, GaN, or epitaxial films, the sample is aligned so that the ion beam travels along a major crystallographic axis. Backscattering from perfectly aligned lattice atoms is strongly suppressed, so the minimum yield (chi-min) quantifies crystalline quality, while excess yield at depth locates defects, damage, or interstitial impurities. Channeling is widely used to evaluate ion implantation damage and annealing recovery.

ERDA and NRA Configurations

For hydrogen determination, the detector is placed at a forward angle and a thin foil stops scattered helium ions while letting recoiled protons reach the detector, enabling depth-profiling of hydrogen. In NRA mode, the beam energy is tuned to excite a nuclear reaction specific to a light isotope, and the emitted charged particles or gamma rays are counted, providing selective quantification of elements such as N, O, C, F, and B.

Measurement Uncertainty and Limitations

Principal sources of uncertainty include the accuracy of stopping powers, detector energy calibration, charge collection, and counting statistics. Mass resolution decreases for heavier elements, and neighboring heavy elements can be difficult to separate at conventional beam energies. Elements lighter than the projectile are either undetectable by standard RBS or detectable only with reduced sensitivity, requiring ERDA or NRA. Lateral resolution is limited to the beam spot size, typically on the order of one millimeter for standard systems, though microbeam systems can reduce this to a few micrometers.

Applications

Semiconductors and Microelectronics

RBS is the reference technique for verifying the stoichiometry and thickness of gate dielectrics such as HfO₂, ZrO₂, and their silicates; for measuring metal silicide layers (NiSi, CoSi₂, TiSi₂) formed during salicide processing; and for confirming diffusion barrier integrity in interconnect stacks. Ion implantation dose verification and post-anneal damage assessment through channeling are routine applications in process development.

Optical and Protective Coatings

Multilayer optical coatings, dielectric mirrors, transparent conductive oxides such as ITO, and hard protective nitride coatings are characterized for layer thickness, stoichiometry, and interfacial mixing, all of which directly influence optical and mechanical performance.

Energy and Functional Materials

Thin film photovoltaic absorbers such as CIGS and CdTe, solid electrolytes, battery electrode coatings, thermoelectric films, superconducting layers, and magnetic multilayers for spintronics are routinely analyzed by RBS to correlate composition gradients and interface quality with device performance.

Surface Treatment, Corrosion, and Heritage Science

RBS quantifies the growth of passive oxide films, the uptake of implanted species in plasma-treated surfaces, and nitrided or carburized case layers. Because it is non-destructive, the technique is also used in archaeometry and museum science to analyze glass corrosion layers, gilding, and patinas on valuable artifacts.

Method Validation and Certification

Owing to its standardless absolute quantification, RBS serves as an arbitration method for certifying film thickness and composition reference materials, and as a benchmark against which SIMS, XPS, EDS, and ellipsometry results are validated.

FAQ

Can RBS detect hydrogen?

Standard RBS cannot detect elements lighter than the projectile ion, including hydrogen. Hydrogen and deuterium are measured quantitatively with depth resolution using Elastic Recoil Detection Analysis (ERDA) performed on the same type of instrument.

Is RBS destructive?

RBS is effectively non-destructive for most inorganic materials. Beam currents are low, and He ions at MeV energies cause negligible sputtering or ablation, so samples can be re-measured or subjected to subsequent analyses. Some polymers and organic materials may suffer beam-induced modification.

What is the detection limit of RBS?

Detection limits scale with the square of atomic number. Heavy elements such as Au or Pt in a light matrix can be detected down to approximately 0.001–0.01 at%, while light elements such as oxygen or nitrogen in a heavy matrix may require several at% to be quantified reliably. Absolute areal density sensitivity for heavy surface contaminants reaches about 10¹³–10¹⁴ atoms/cm².

What depth resolution and analysis depth can be achieved?

Typical near-surface depth resolution is 10–20 nm in normal incidence geometry and can be improved to a few nanometers using grazing-angle geometries. The maximum analysis depth is approximately 1–2 micrometers for common materials, limited by the energy loss of the probing ions.

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

Unlike XPS, RBS is not a spectroscopy of chemical states but provides far more accurate absolute quantification and much greater analysis depth. Compared with SIMS, RBS is quantitative without standards but has lower sensitivity and poorer depth resolution. Compared with EDS in an electron microscope, RBS offers substantially better light-element capability in many matrices, true depth information, and absolute accuracy suitable for certification work.

What sample information should be provided before testing?

Providing the expected layer sequence, nominal thicknesses, nominal composition, substrate material, and process history (deposition method, annealing, implantation energy and dose) greatly accelerates spectrum modeling and improves the reliability of the final depth profile.