Rutherford Backscattering Spectrometry (RBS) is a non-destructive, ion-beam analytical technique that delivers standardless quantitative elemental composition and depth profile information for thin films, coatings, and surface layers. As a core method in surface and interface characterization, RBS testing is widely used in semiconductor process development, thin-film deposition control, materials research, and failure analysis. This article explains the technical principles, sample requirements, measurable parameters, analytical methodology, and typical applications of RBS elemental composition detection services provided by accredited third-party laboratories.
Technical Overview
Rutherford Backscattering Spectrometry (RBS) is a quantitative surface analysis technique in which a collimated beam of light ions—most commonly 1.5 to 3.0 MeV helium ions (He+ or He2+)—is directed onto a sample in high vacuum. A small fraction of these ions undergo elastic (Rutherford) scattering through Coulomb interaction with atomic nuclei near the sample surface. The energy of each backscattered ion is measured with a silicon surface-barrier or PIPS detector, typically positioned at a scattering angle of 150 to 170 degrees.
The measured energy of a backscattered ion carries two pieces of information. First, the fraction of energy retained after the collision, described by the kinematic factor, identifies the mass of the target atom: heavier target elements return ions with higher energies. Second, the additional energy lost as the ion travels into and out of the sample, governed by stopping power, encodes the depth at which the scattering event occurred. An RBS spectrum therefore simultaneously provides elemental identification, quantitative concentration, and depth distribution in a single measurement.
Several characteristics distinguish RBS from other compositional techniques:
- Absolute quantification: because Rutherford scattering cross sections are known from first principles, composition and areal density can be determined without matrix-matched calibration standards, yielding typical accuracy of 1 to 5 percent relative.
- Non-destructive analysis: the probe beam causes negligible damage in most inorganic materials, allowing re-measurement and subsequent analysis by other methods.
- Depth information: near-surface depth resolution of approximately 5 to 20 nm, with a probing depth of roughly 1 to 2 micrometers for standard helium beams.
- Mass sensitivity scaling: detection sensitivity increases with atomic number, making RBS particularly powerful for detecting heavy elements in light matrices, for example heavy-metal contamination on silicon.
- Whole-film integration: the technique measures total areal density (atoms per square centimeter), providing film thickness independently of density assumptions.
RBS terminology and reporting conventions follow the international vocabulary for surface chemical analysis defined in ISO 18115, and laboratory workflows are typically aligned with ISO/IEC 17025 quality management practices in accredited third-party laboratories.
Testing Samples
RBS is best suited to solid, vacuum-compatible samples with reasonably flat surfaces. Typical sample categories submitted for elemental composition detection include:
Thin Films and Functional Coatings
Single-layer and multilayer films deposited by sputtering, evaporation, CVD, ALD, or electroplating, including metals, oxides, nitrides, carbides, and composite layers on silicon, glass, quartz, or metal substrates. Films in the thickness range of a few nanometers to about one micrometer are ideal.
Semiconductor Wafers and Device Structures
High-k dielectric stacks, metal silicides, barrier and seed layers, implanted layers, and gate or contact metallization on patterned or blanket wafers. Pieces as small as a few millimeters square are acceptable.
Multilayer, Interface, and Diffusion Structures
Superlattices, optical multilayer stacks, X-ray mirror structures, diffusion couples, and annealed stacks used to study interfacial reaction, intermixing, and barrier integrity.
Bulk Metals, Alloys, Ceramics, and Glasses
Surface regions of bulk materials, including oxidized or nitrided layers, corrosion products, and surface-modified zones produced by implantation, plasma treatment, or diffusion.
Energy and Optoelectronic Materials
Thin-film photovoltaic absorbers and buffers, transparent conductive oxides, battery electrode coatings, and phosphor or LED-related multilayer structures.
Sample preparation requirements are minimal: the specimen must be compatible with high vacuum (approximately 10-6 to 10-7 mbar), dry, free of loose particulates, and preferably larger than the beam spot, which is typically 1 to 2 mm in diameter. Strongly insulating samples can often still be measured with electron flood compensation or reduced beam current. Organic or volatile samples require prior consultation because beam-induced degradation may occur.
Testing Items
Standard RBS elemental composition detection typically covers the following measurable quantities and analytical deliverables:
- Elemental identification of all constituents from beryllium (or carbon, in favorable cases) through uranium, subject to matrix effects.
- Quantitative elemental composition in atomic percent or weight percent, determined standardlessly from Rutherford cross sections.
- Film thickness in nanometers, together with areal density in atoms per square centimeter, obtained independently of assumed film density.
- Stoichiometry verification of compounds such as oxides, nitrides, silicides, and complex quaternary chalcogenides.
- Depth profiles of elemental concentration from the surface to approximately 1 to 2 micrometers, revealing gradients, layer interfaces, and interdiffusion.
- Interfacial analysis: degree of intermixing, interface abruptness, reaction-layer formation after annealing, and barrier-layer effectiveness.
- Heavy-element impurity and contamination detection in light matrices, with detection limits typically around 1018 to 1019 atoms per cubic centimeter (approximately 0.1 to 1 atomic percent depending on the element pair).
- Ion implantation dose verification and retained-dose measurement for semiconductor process qualification.
- Hydrogen and deuterium content by Elastic Recoil Detection Analysis (ERDA), offered as a companion technique because conventional RBS is not sensitive to hydrogen.
- Crystal quality assessment, defect density, and lattice site occupation of dopants in single-crystal samples by RBS/channeling.
Testing Methods
The reference RBS measurement configuration used by most accredited laboratories is described below. Individual parameters are adjusted according to the sample system and the analytical question.
Standard RBS Measurement
A tandem accelerator or Van de Graaff accelerator produces a 2.0 MeV He+ (or He2+) beam with a typical current of 5 to 50 nA. The beam is collimated onto the mounted sample inside a scattering chamber at high vacuum. Backscattered ions are detected by a passivated implanted planar silicon (PIPS) or surface-barrier detector with an energy resolution of approximately 12 to 20 keV, placed at a scattering angle near 160 to 170 degrees in a Cornell or IBM geometry. Each spectrum typically accumulates a total collected charge of 10 to 40 microcoulombs, corresponding to measurement times of 10 to 30 minutes per spectrum.
Enhanced-Resolution and Multi-Geometry Measurements
When improved near-surface depth resolution is required, the sample is tilted toward grazing incidence to lengthen the ion path through the surface region, improving depth resolution to a few nanometers at the cost of probing depth. Measuring the same sample at normal incidence and grazing incidence, or at two detector angles, resolves overlapping spectral features and refines layer thickness determination.
RBS/Channeling
For single-crystal samples, the beam is aligned with a major crystallographic axis. Backscattering yield reduction relative to random incidence quantifies crystalline quality, and the angular dependence of dopant or impurity signals identifies their lattice location. Channeling is also used to suppress substrate yield and improve sensitivity to light surface layers.
ERDA for Hydrogen Analysis
In Elastic Recoil Detection Analysis, the detector is placed forward of the sample so that recoiled hydrogen or deuterium nuclei are measured, with the scattered primary beam blocked by a thin foil. This provides quantitative hydrogen depth profiles that RBS cannot supply directly.
Calibration, Simulation, and Data Reduction
Energy calibration of the detector chain is performed using spectra from certified reference materials and national-metrology-institute-traceable standards, such as known silicon dioxide films on silicon or heavy-element implants with certified fluence. Measured spectra are interpreted by fitting simulated spectra generated with dedicated simulation codes (for example SIMNRA or RUMP) that model kinematic factors, Rutherford and non-Rutherford cross sections, stopping powers based on widely used stopping-power databases, straggling, and detector resolution. Iterative fitting yields layer compositions, thicknesses, and interface profiles, with reported uncertainties reflecting counting statistics, stopping-power uncertainty, and geometry calibration.
Quality Assurance
Accredited third-party laboratories implement method validation, periodic intermediate checks with reference films, control charts, and documented uncertainty budgets in accordance with ISO/IEC 17025 requirements, ensuring traceable and reproducible results.
Applications
RBS elemental composition detection supports research, development, qualification, and failure analysis across many industries:
- Semiconductor and microelectronics: stoichiometry and thickness control of high-k dielectrics, silicide phase and thickness verification, ion implant dose certification, barrier-layer integrity, and epi-layer characterization.
- Thin-film photovoltaics: composition and gradient analysis of CIGS, CZTS, CdTe, and silicon-based stacks, verifying absorber stoichiometry and diffusion behavior after process optimization.
- Optical and protective coatings: layer thickness, interface quality, and compositional uniformity of antireflection stacks, filters, hard coatings, and decorative PVD films.
- Data storage and magnetics: multilayer thickness and interfacial mixing in spin-valve and magnetic tunnel junction structures.
- Metallurgy and corrosion science: oxide-scale growth kinetics, surface alloying, coatings on metals, and hydrogen ingress studied by ERDA.
- Advanced materials research: ion-beam modification studies, irradiation damage, diffusion couple analysis, and validation of deposition models.
- Batteries and energy storage: characterization of electrode coatings, solid electrolyte layers, and interface reactions.
- Cultural heritage and forensics: non-destructive analysis of gilding, glass composition, and pigment layers where sampling is prohibited.
- Academic and publication support: providing quantitative, referee-acceptable composition and thickness data for peer-reviewed research.
FAQ
Which elements can RBS detect?
RBS can, in principle, detect all elements from helium to uranium. In practice, the mass separation of the kinematic factor and signal overlap with the substrate determine feasibility. Heavy elements in light matrices are detected with the best sensitivity; light elements (below silicon) in heavy matrices are difficult or impossible to quantify reliably. Hydrogen is invisible to conventional RBS and requires ERDA.
What are the detection limits of RBS?
For heavy elements on a light substrate such as silicon, detection limits of approximately 1018 to 1019 atoms per cubic centimeter, corresponding to roughly 0.1 to 1 atomic percent, are routinely achieved. Sensitivity degrades for lighter elements and in heavy matrices, where limits may exceed several atomic percent.
What depth resolution and maximum analysis depth can be expected?
Depth resolution at the surface is typically 5 to 20 nm for standard geometry and can be improved to a few nanometers using grazing-incidence geometry. The maximum analyzable depth is approximately 1 to 2 micrometers for helium beams, limited by ion energy loss and spectral pile-up from the substrate.
Is RBS a destructive test?
No. For metals, semiconductors, ceramics, and most inorganic films, the helium beam dose used causes no measurable damage or sputtering, and samples remain fully usable for subsequent analyses. Organic materials and certain alkali halides may exhibit beam-induced modification and require evaluation before testing.
Why choose RBS instead of EDS, XPS, or SIMS?
EDS probes micrometer-scale volumes and cannot resolve nanometer films; XPS is highly surface-sensitive (top 5 to 10 nm) and needs standards for accurate quantification; SIMS offers excellent trace sensitivity but has matrix-dependent quantification. RBS uniquely combines standardless absolute quantification, full film thickness coverage up to about 1 micrometer, and non-destructive depth profiling, making it the reference method for quantitative thin-film composition. The techniques are frequently used together as complementary tools.
What accuracy and precision can be reported?
Elemental composition accuracy is typically 1 to 5 percent relative, dominated by stopping-power uncertainties, and areal density can often be determined to better than 3 percent. Layer thickness uncertainty depends on density assumptions or is reported directly as areal density when appropriate.
What sample size and form are accepted?
Common sample forms are wafer pieces, coated coupons, cleaved fragments, or pellets roughly 5 to 20 mm across, with a reasonably flat surface and vacuum compatibility. Full wafers can usually be accommodated only in specially configured chambers, so submitting pieces of appropriate size is recommended.
How should results be interpreted?
A standard RBS report includes the measured spectrum with the fitted simulation, layer-by-layer composition in atomic percent, areal density and thickness of each layer with uncertainties, measurement conditions (beam energy, species, geometry, collected charge), and any observations on interface broadening or impurities. Laboratory engineers can assist with interpretation and with designing complementary measurements when needed.
