RBS data analysis is the process of interpreting energy spectra from Rutherford Backscattering Spectrometry to determine the elemental composition, thickness, and depth distribution of thin films and surface layers. This quantitative technique provides standardless accuracy at the atomic level, making it indispensable in semiconductor, materials science, and thin-film characterization research. This article explains the physical principles, analytical methodology, sample requirements, and typical applications of RBS data analysis.
Technical Overview
Rutherford Backscattering Spectrometry (RBS) is an ion-beam analytical technique that uses a collimated beam of light ions—most commonly helium ions (He+) or helium nuclei (He2+, alpha particles) accelerated to energies between 1.0 and 3.0 MeV—to probe the composition and structure of solid samples. When these ions impinge upon the sample, a small fraction undergoes elastic Rutherford scattering with atomic nuclei in the material and is deflected backward at angles typically between 160 and 180 degrees. A semiconductor detector, usually a silicon surface barrier detector, measures the energy distribution of these backscattered ions.
RBS data analysis converts the measured energy spectrum into quantitative information about the sample. The energy of each backscattered ion depends on the mass of the target atom involved in the collision, governed by the kinematic factor, while the gradual energy loss of ions traveling through the material allows depth information to be reconstructed. Because the scattering cross sections are well known from fundamental physics, RBS is one of the few analytical techniques that can provide absolute quantitative results without calibration standards.
The principal outputs of RBS data analysis include elemental areal density (atoms per square centimeter), film thickness, layer composition in atomic percent, interface sharpness, and depth profiles of heavy elements in lighter matrices. Detection sensitivity scales approximately with the square of the atomic number, making RBS highly sensitive to heavy elements but less effective for detecting light elements such as hydrogen, which is better measured by complementary techniques like Elastic Recoil Detection (ERD).
Testing Samples
Semiconductor Thin Films
RBS is extensively applied to epitaxial layers, deposited metal films, silicides, gate oxides, high-k dielectrics, and ion-implanted wafers. Silicon-based multilayer stacks, germanium-on-silicon structures, and compound semiconductor heterostructures such as GaAs, InP, and GaN are routinely characterized to verify stoichiometry, film thickness, and interfacial mixing.
Metal and Alloy Coatings
Sputtered or evaporated metal films, diffusion barriers, barrier layers, optical reflective coatings, and multilayer structures such as Ni/Pt, Ti/Al, and W/Si are analyzed to determine layer thickness, interdiffusion behavior, and compositional uniformity.
Optical and Functional Coatings
Transparent conductive oxides (ITO, ZnO:Al), dielectric Bragg mirrors, and protective hard coatings can be profiled for thickness and composition. RBS is particularly valuable for verifying stoichiometry in complex oxide films where other techniques struggle with quantification.
Glass, Ceramics, and Geological Materials
Ion-implanted glasses, ceramic surface layers, and mineral surfaces used in nuclear waste studies or diffusion experiments are common RBS samples. The technique is non-destructive for most inorganic materials, allowing repeated measurements on the same specimen.
Sample Requirements
- Sample size: typically 5 x 5 mm to 15 x 15 mm; smaller areas can be measured with micro-beam setups
- Sample thickness: bulk solids up to several millimeters are acceptable; films from a few nanometers to several micrometers are measurable
- Surface condition: must be reasonably flat and smooth; rough surfaces broaden spectral features and degrade depth resolution
- Conductivity: insulating samples may require charge compensation or a thin conductive coating consideration, since charging deflects the beam
- Vacuum compatibility: samples must be stable under high vacuum (approximately 10-6 to 10-7 Torr) and tolerate mild beam heating
- Damage sensitivity: organic or beam-sensitive materials should be evaluated in advance, as ion bombardment can modify them
Testing Items
RBS data analysis delivers the following quantitative results:
- Elemental composition of films and surface layers, expressed in atomic percent, with accuracy typically better than 1-3% relative
- Layer thickness or areal density, converted to physical thickness using assumed or measured density values
- Depth profiles of elemental concentration, with depth resolution of approximately 5-20 nm near the surface
- Interfacial abruptness and interdiffusion assessment in multilayer structures
- Quantification of heavy-element impurities or dopants in light matrices, with sensitivity down to parts-per-thousand or better for high-Z elements
- Stoichiometry verification of compound films such as oxides, nitrides, and silicides
- Ion implantation dose and range verification by comparing implanted profiles with simulated predictions
- Crystalline quality parameters via RBS/channeling, including minimum yield and lattice damage profiles in single crystals
Testing Methods
Standard RBS Measurement
In a standard RBS experiment, a 1.5-2.0 MeV He+ or He2+ beam is directed onto the sample at normal or near-normal incidence. Backscattered ions are detected by a silicon barrier detector placed at a scattering angle of approximately 170 degrees with an energy resolution of 12-20 keV. The detector signal is processed through preamplification, amplification, and multi-channel analysis to produce an energy histogram of backscattered particle counts.
Principles of Spectrum Interpretation
Three physical relationships underpin RBS data analysis:
- Kinematic factor: the ratio of the projectile energy after an elastic collision to its incident energy depends only on the masses involved and the scattering geometry. Heavier target atoms produce higher-energy backscattered ions, enabling mass identification
- Scattering cross section: the probability of backscattering from each element follows the Rutherford formula and scales with the square of the atomic number of the target atom, enabling absolute quantification
- Energy loss: ions lose energy continuously while traversing the sample via electronic stopping. The cumulative energy loss of the incident and exiting ions encodes the depth at which each scattering event occurred, producing the characteristic sloped edges and steps in the spectrum
Spectral Simulation and Fitting
Quantitative RBS data analysis is performed by iterative spectral simulation. The analyst constructs a hypothetical sample model consisting of stacked layers with defined composition and thickness, and simulation software calculates the theoretical backscattering spectrum for the given beam energy, geometry, detector resolution, and stopping powers. The model is refined until the simulated spectrum matches the measured data.
- SIMNRA: widely used software developed for the simulation of ion beam spectra, supporting RBS, ERD, and nuclear reaction analysis with extensive stopping power and cross-section databases
- RUMP: an established simulation and fitting package for RBS spectrum modeling with graphic editing of layer structures
- Other codes such as DataFurnace and NDF perform automated fitting of multiple spectra acquired at different geometries
Variable-Geometry Measurements
Depth resolution can be improved by tilting the sample to increase the ion path length within a given depth interval. Combining spectra acquired at normal incidence and grazing incidence enhances the resolution of buried interfaces. Grazing exit geometries similarly amplify depth sensitivity for near-surface features.
RBS/Channeling Analysis
For single-crystal samples such as epitaxial semiconductor layers, the beam is aligned with a major crystallographic axis. Regular lattice atoms shadow each other, drastically reducing the backscattering yield from the bulk crystal. The normalized minimum yield indicates crystal quality, while excess yield versus depth localizes lattice damage, implantation disorder, or interstitial impurity atoms. Angular scans around axial or planar directions further reveal lattice strain and site occupancy of dopant atoms.
Complementary Techniques
Because RBS has limited sensitivity for light elements, hydrogen analysis is performed with Elastic Recoil Detection using heavier projectiles such as 2-3 MeV helium or higher-energy heavy ions. For elements with overlapping spectral features, Particle Induced X-ray Emission (PIXE) may be combined with RBS to resolve ambiguities in elemental identification.
Applications
RBS data analysis serves a broad range of industries and research fields:
- Semiconductor manufacturing: qualification of deposited film thickness and stoichiometry, calibration of ion implantation dose, evaluation of rapid thermal annealing and dopant activation, characterization of silicide formation and contact metallization
- High-k gate dielectrics: determination of HfO2, ZrO2, and aluminide film composition and interfacial layer growth
- Photovoltaics: compositional profiling of CdTe, CIGS, and perovskite absorber layers, diffusion studies of buffer layers and contacts
- Surface engineering: verification of nitriding, carburizing, and oxidation treatments, thickness and composition of PVD and CVD hard coatings
- Optoelectronics: analysis of III-V epitaxial heterostructures, quantum well thickness, and superlattice periodicity
- Metallurgy and corrosion science: oxide scale thickness, passivation layer composition, and diffusion couple interfacial analysis
- Nuclear materials: helium and heavy-ion implantation studies, radiation damage profiling, and inert matrix fuel research
- Cultural heritage and archaeology: non-destructive depth profiling of ancient glass weathering layers, gilded surfaces, and patinas
- Academic research: absolute benchmarking of composition for novel thin-film materials reported in publications
FAQ
What is the detection depth of RBS?
Using 2 MeV helium ions, RBS can typically probe depths of up to approximately 1-2 micrometers in typical materials, limited by the increasing energy spread of the beam as it penetrates the sample. Best depth resolution of roughly 5-10 nm is achieved near the surface and degrades with depth.
What elements can RBS detect?
RBS can detect all elements heavier than the projectile, effectively from carbon and nitrogen upward with helium beams. Sensitivity improves strongly with atomic number; heavy elements such as gold or platinum can be quantified at concentrations orders of magnitude lower than light elements. Hydrogen is invisible to standard RBS and requires ERD.
Why is RBS considered a standardless quantitative technique?
The scattering probability per unit solid angle is given exactly by the Rutherford cross section, which depends only on fundamental constants, the projectile energy, and the scattering angle. Detector solid angle and accumulated beam charge can be measured independently, so absolute areal densities are obtained without reference standards.
How is film thickness derived from the spectrum?
The energy width of a spectral feature corresponding to a film equals the product of the film areal density and the effective stopping cross section along the combined incident and exit paths. Simulation software converts the measured energy width into areal density, which is then divided by the assumed mass density to yield physical thickness.
Is RBS destructive?
For most inorganic materials, RBS is effectively non-destructive because the MeV helium beam deposits little energy into the lattice and does not sputter material. Beam-sensitive organic materials, however, may suffer bond breaking or cross-linking under prolonged irradiation.
Can RBS measure multiple layers simultaneously?
Yes. A single spectrum from a multilayer sample contains superimposed contributions from every layer. Simulation software deconvolutes the spectrum into a layer model reporting thickness and composition of each layer, provided adjacent elements or layers produce sufficiently distinct spectral edges.
What limits the accuracy of RBS data analysis?
Accuracy is governed by knowledge of stopping powers, detector energy calibration and resolution, geometric angle accuracy, beam charge integration, counting statistics, and sample flatness. For routine measurements, composition accuracy of 1-3% relative and thickness accuracy of 3-5% are commonly achieved.
How does RBS compare with SIMS and XPS?
RBS offers absolute quantification and non-destructive depth profiling but poorer trace sensitivity than SIMS and shallower information depth. XPS provides chemical-state information that RBS lacks. The three techniques are complementary: RBS for absolute composition and thickness, SIMS for trace-level depth profiles, and XPS for bonding chemistry at the extreme surface.
What information does RBS channeling add?
Channeling converts RBS into a crystal-quality probe. The minimum yield quantifies the fraction of atoms displaced from regular lattice sites, depth-resolved damage profiles locate implantation-induced disorder, and angular scans can determine strain and lattice site locations of impurities in epitaxial films.