RBS coating analysis (Rutherford Backscattering Spectrometry) is a powerful ion beam technique for quantitatively measuring the thickness, elemental composition, and depth profile of thin films and plated layers without destroying the sample. As a standardless, absolute quantification method, RBS is widely used in semiconductor manufacturing, electroplating quality control, optical coatings, and advanced materials research. This article explains the technical principles of RBS coating detection, the sample types accepted, testing items, methods, and practical applications.
Technical Overview
Rutherford Backscattering Spectrometry (RBS) is an ion beam analysis technique used to determine the elemental composition, areal density, and depth distribution of coatings and thin films. In a typical RBS measurement, a beam of high-energy ions, most commonly 2 MeV helium ions (He+ or He2+), is directed at the sample surface. A small fraction of these ions undergo elastic collisions with the nuclei of atoms in the coating and are backscattered toward a detector positioned at a known angle, typically 160 to 170 degrees relative to the incident beam.
The energy of each backscattered ion carries two key pieces of information. First, the kinematic factor, which depends on the mass ratio between the projectile ion and the target atom, identifies the mass of the scattering nucleus and therefore the element. Second, the incremental energy loss of the ion as it travels into and out of the sample provides depth information. By analyzing the complete energy spectrum, analysts can reconstruct the layer structure of the coating, including thickness, composition, interface sharpness, and interdiffusion between layers.
RBS is distinguished from most surface analysis techniques by its absolute quantification capability. The scattering cross sections are well known from fundamental physics, so the measurement of areal density in atoms per square centimeter does not require calibration standards. Sensitivity scales with the square of the atomic number, which makes RBS extremely sensitive to heavy elements such as gold, platinum, and tantalum in light matrices such as silicon, carbon, or oxygen, with detection limits reaching parts-per-million levels for favorable element combinations. Depth resolution at the surface is typically 10 to 20 nanometers for standard beam geometries, and the total analyzable depth for helium beams extends to approximately 1 to 2 micrometers depending on the matrix.
- Absolute, standardless quantification of film composition and areal density
- Non-destructive analysis of coatings from sub-nanometer to micrometer scale
- Excellent sensitivity for heavy elements in light matrices
- Simultaneous measurement of thickness, stoichiometry, and depth profiles
- Widely recognized as a reference method for validating other analytical techniques
Testing Samples
RBS coating analysis is applicable to a broad range of sample types. Because the technique is non-destructive and requires minimal sample preparation, both production parts and dedicated witness coupons can be analyzed. Typical sample requirements include a flat analysis area of at least a few millimeters by a few millimeters, vacuum compatibility, and reasonably smooth surfaces for optimal depth resolution.
Metallic Plating and Electroplated Layers
Common submissions include gold, silver, nickel, copper, tin, zinc, palladium, and rhodium plating on substrates such as copper alloys, steel, kovar, and ceramic. RBS accurately measures plating thickness and detects intermetallic compound formation, such as nickel-tin or copper-gold interdiffusion, that affects solderability and contact resistance.
Semiconductor Thin Films
Silicides, high-k dielectrics, barrier layers, metal gate stacks, and epitaxial layers are routinely characterized. Examples include TiN, TaN, HfO2, TiSi2, CoSi2, SiGe, and GaN-based heterostructures, where precise stoichiometry and interface abruptness are critical to device performance.
Optical and Conductive Transparent Coatings
Indium tin oxide (ITO), fluorine-doped tin oxide, TiO2, SiO2, Nb2O5, and multilayer dielectric stacks are analyzed for composition control, which directly governs refractive index, transparency, and sheet resistance.
Hard and Protective Coatings
PVD and CVD coatings such as TiN, CrN, TiAlN, Al2O3, and diamond-like carbon (DLC) can be measured for elemental ratios, nitrogen or oxygen content, and coating thickness uniformity.
Photovoltaic and Energy Materials
CIGS, CdTe, CZTS, and perovskite absorber layers are analyzed for compositional grading and sodium or potassium diffusion, which strongly influence device efficiency.
Testing Items
RBS coating detection services typically cover the following measurable quantities:
- Coating thickness expressed as areal density (atoms/cm²) and, with known or assumed film density, as physical thickness in nanometers or micrometers
- Elemental composition and stoichiometry of single or multilayer films, including atomic ratios for compounds such as nitrides, oxides, and alloys
- Depth concentration profiles showing how elemental composition varies from the surface into the film and substrate
- Interface characterization, including intermixing, diffusion zones, and intermetallic layer formation between coating and substrate
- Heavy element impurity detection in light matrices, for example trace gold, tungsten, or platinum contamination in silicon or carbon-based films
- Ion implantation dose verification and implanted depth profile confirmation
- Surface oxidation or contamination layer quantification
- Hydrogen content measurement when RBS is combined with elastic recoil detection analysis (ERDA)
- Crystal quality assessment, such as defect density and lattice site occupancy, using RBS in channeling mode for single-crystal samples
Testing Methods
The standard RBS measurement protocol begins with sample mounting on a goniometer stage inside a high-vacuum analysis chamber. A collimated beam of helium ions accelerated to approximately 2 MeV by a particle accelerator is scanned across the analysis area, typically a region of 1 square millimeter. Backscattered ions are detected by a silicon surface barrier or passivated implanted planar silicon (PIPS) detector, and pulse height analysis converts each event into an energy channel, building the backscattering spectrum.
Spectrum interpretation is performed by comparing the measured data with simulated spectra generated by software packages such as SIMNRA or RUMP. The analyst constructs a layer model with trial compositions and thicknesses, and the software iteratively refines the model until the simulation matches the experimental spectrum. Because scattering kinematics and stopping powers are physically well characterized, the converged solution yields absolute areal densities and atomic fractions with typical uncertainties of 3 to 5 percent.
Several methodological variations extend the capability of standard RBS:
- Variable incidence geometry, in which tilting the sample to a glancing angle improves depth resolution for ultra-thin films
- Higher energy or heavier projectiles, using beams above 3 MeV or heavier ions to increase sensitivity for specific elements
- ERDA, in which the detector is placed forward of the sample to recoil and quantify light elements, especially hydrogen
- RBS channeling, which aligns the beam with a crystallographic axis of single-crystal substrates to assess epitaxial quality and defect levels
- Combined ion beam analysis with PIXE, which adds trace element capability to complement the quantitative layer information from RBS
Measurements are commonly performed in accordance with recognized practices, and results are reported as layer models with composition tables, areal densities, calculated thicknesses, and fitted spectra overlays. When absolute film density is required, RBS results are frequently cross-referenced with profilometry, ellipsometry, or X-ray reflectometry.
Applications
RBS coating detection serves quality control, process development, failure analysis, and research across many industries:
- Semiconductor and microelectronics: verifying barrier layer integrity, silicide formation, high-k dielectric stoichiometry, and implant dose accuracy in process qualification
- Electroplating and connector manufacturing: certifying precious metal plating thickness, detecting underlayer diffusion that degrades contact resistance, and optimizing bath chemistry
- Optical coating production: controlling refractive index through film composition in antireflection, filter, and mirror coatings
- Decorative and functional finishing: confirming chromium, nickel, and gold layer structures on consumer and luxury hardware
- Photovoltaics: profiling compositional gradients in thin film absorbers and monitoring alkali diffusion through buffer layers
- Hard coating and tooling: validating nitrogen content and layer thickness of wear-resistant nitride coatings on cutting tools
- Aerospace and automotive: qualifying thermal barrier, anodized, and corrosion protection coatings
- Academic and R&D laboratories: calibrating deposition systems, benchmarking other analytical techniques, and publishing traceable film composition data
- Failure analysis: identifying thickness shortfall, interfacial contamination, oxidation, or interdiffusion as root causes of coating-related product failures
FAQ
What is the measurable thickness range of RBS coating analysis? RBS accurately measures coatings from a fraction of a nanometer up to roughly 1 to 2 micrometers for common matrices using standard 2 MeV helium beams. Thicker multilayer structures can be characterized when heavier or higher-energy beams are applied.
Is RBS a destructive test? No. RBS is essentially non-destructive. The ion dose is low enough that most materials suffer no measurable damage or heating, allowing the same sample to be retested or subjected to subsequent analysis.
What are the detection limits? Detection limits depend strongly on the element pair. Heavy elements in a light matrix, such as gold or tantalum in silicon, can be detected at levels below 0.1 atomic percent, while light elements such as oxygen, nitrogen, and carbon have higher limits of roughly 1 to 10 atomic percent due to background from heavier substrate elements.
How accurate is the thickness measurement? Areal density, the fundamental RBS result, is typically accurate to within a few percent without standards. Conversion to physical thickness introduces additional uncertainty from the assumed or measured film density.
What sample size and preparation are required? A flat area of a few millimeters square is sufficient, and the sample must be vacuum compatible. No coating, polishing, or sectioning is needed, and full-sized parts can often be accommodated by cutting or using suitable sample holders.
How does RBS compare with SIMS, XPS, and cross-sectional SEM? RBS provides absolute quantitative composition and thickness without standards but with modest depth resolution, whereas SIMS offers excellent depth resolution and trace sensitivity with less direct quantification, XPS is highly surface sensitive within about 10 nanometers, and SEM cross sections give direct images of layer thickness but limited compositional quantification. The methods are complementary, and RBS is often used as the reference technique to calibrate the others.
Can RBS detect light elements such as hydrogen? Standard RBS cannot quantify hydrogen because of kinematic constraints, but the companion ERDA technique measures hydrogen content and depth distribution with good sensitivity.