Rutherford Backscattering Spectrometry (RBS) depth profiling is a non-destructive ion beam analysis technique that quantitatively determines elemental compositions and depth distributions in thin films and layered materials with nanometer-scale resolution. This technical article explains the physical principles of the RBS experiment, standard test procedures, applicable sample types, measurable parameters, and industrial and research applications. Accredited third-party laboratories and research institutions worldwide rely on RBS to deliver calibration-free, standards-compliant quantitative results for advanced material characterization and quality control.
Technical Overview
Rutherford Backscattering Spectrometry (RBS) is a quantitative surface and thin-film analysis technique in which a collimated beam of light ions, typically helium ions (He+) accelerated to energies between 1.0 and 3.0 MeV, is directed onto a specimen under high vacuum. A small fraction of these ions undergoes elastic scattering, following the Coulomb interaction described by Lord Rutherford, with atomic nuclei near the surface and within the top micrometer of the sample. The energy spectrum of the backscattered ions, recorded by a silicon surface-barrier or PIPS detector placed at a large scattering angle, encodes both the mass and the depth of the target atoms.
Physical Principle of Elastic Backscattering
Each backscattering event transfers a well-defined fraction of the projectile energy to the target nucleus. This fraction, known as the kinematic factor K, depends on the masses of the projectile ion and the target atom as well as on the scattering geometry. For scattering at 180 degrees, the kinematic factor simplifies to K = [(M2 − M1)/(M1 + M2)]2, where M1 is the projectile mass and M2 is the target atom mass. Because heavier atoms retain more energy after scattering, high-energy edges in the spectrum correspond to heavier elements, allowing elemental identification across nearly the entire periodic table, from boron to uranium.
Depth Information from Energy Loss
Depth profiling is achieved through the continuous energy loss that ions experience while penetrating into and escaping from the sample, described by stopping powers widely tabulated for ion–matter interactions. An ion scattered at a given depth loses more energy along its trajectory than one scattered at the surface, so its detected energy is shifted to lower values. By modeling this energy loss, an RBS spectrum can be converted into a concentration-versus-depth profile. Typical depth resolution is approximately 5 to 15 nanometers at the surface and degrades gradually with depth due to energy straggling. The maximum analysis depth is roughly 1 to 2 micrometers in light matrices such as silicon, and several hundred nanometers in dense heavy-metal films.
Quantification Without Reference Standards
Because the Rutherford scattering cross sections are known from first principles, RBS is an absolute, calibration-free technique for determining areal densities, expressed in atoms per square centimeter. The technique offers a typical accuracy of 1 to 5 percent for film composition and thickness, which is exceptional among surface analysis methods. Combined with independently measured physical thickness, RBS also yields film density. Crystallographic information, including lattice damage and impurity lattice site occupation, can be obtained through the ion channeling variant of the technique.
Testing Samples
RBS depth profiling experiments accommodate a broad range of solid specimens, provided they are vacuum compatible and reasonably flat. Common sample categories submitted to accredited third-party laboratories and university ion beam facilities include the following.
Semiconductor Thin Films and Multilayer Structures
Gate oxides and high-k dielectric stacks, silicide contacts, metal gate electrodes, SiGe epitaxial layers, and ion-implanted wafers are routinely analyzed to verify thickness, stoichiometry, implanted dose, and interface integrity.
Optical, Protective, and Functional Coatings
Antireflective multilayer stacks such as alternating TiO2 and SiO2 films, hard nitride coatings such as TiN and CrN, diamond-like carbon layers, and transparent conductive oxides such as ITO are characterized for layer thickness, composition, and interdiffusion.
Metals, Alloys, and Diffusion Couples
Metallic multilayers, thin metal interconnects, diffusion couples used in kinetic studies, oxidized and corroded surfaces, and plating or metallization layers on various substrates can be profiled to quantify reaction zones and layer growth.
Energy and Photovoltaic Materials
Thin-film solar absorbers such as CIGS and CdTe, solid-state battery thin-film electrodes, solid electrolytes, and barrier layers are analyzed for elemental ratios, uniformity, and interfacial reaction products formed during aging or cycling.
Sample Requirements and Constraints
- Specimen size is typically 5 to 20 millimeters across, mounted on a multi-position holder or goniometer.
- Surfaces should be smooth and flat, since roughness broadens spectral edges and degrades depth resolution.
- Insulating samples such as glasses, oxides, and polymers may require electron flood compensation or thin conductive coatings to suppress charging.
- Samples must tolerate high vacuum of approximately 10−6 to 10−7 millibar and low beam currents of roughly 10 to 50 nanoamperes.
Testing Items
An RBS depth profiling experiment yields a comprehensive set of quantitative results. Standard deliverables include:
- Elemental identification of constituents from boron through uranium, with the highest mass resolution achieved at backscattering angles near 170 to 180 degrees.
- Quantitative areal density of each element in atoms per square centimeter, derived without reference standards.
- Layer thickness in nanometers or atomic density times thickness, and film density when combined with profilometry or ellipsometry data.
- Stoichiometric ratios of compound films, such as HfO2, TiN, Ta2O5, and complex quaternary absorbers.
- Depth concentration profiles of individual elements, resolving layered structures, gradients, and intermixed zones.
- Interface position, sharpness, and interdiffusion or reaction-layer thickness between adjacent films.
- Retained ion implantation dose and projected range profiles for process qualification.
- Heavy-element impurity quantification with detection limits around 0.01 to 1 atomic percent, depending on matrix and geometry.
- Crystalline quality indicators, including lattice damage depth profiles and minimum channeling yield, obtained in channeling RBS mode.
Testing Methods
Specimen Preparation and Mounting
Preparation is minimal compared with many surface techniques because sputtering and sectioning are not required. Specimens are cleaned with solvents or compressed nitrogen to remove particulate contamination, then fixed on a conductive sample holder using carbon tape or mechanical clips. Reference positions are recorded so that multiple spots can be analyzed to assess lateral uniformity.
Instrumentation and Beam Conditions
A typical experimental setup consists of a Van de Graaff, pelletron, or tandem electrostatic accelerator producing a 1.5 to 3.0 MeV He+ or He2+ beam, magnetic or electrostatic beam conditioning, a scattering chamber with a precision goniometer, and one or more silicon detectors positioned at scattering angles between 150 and 170 degrees. A normal-incidence geometry with a detector at about 170 degrees is the most common configuration, while grazing-angle geometries can enhance depth resolution when required. Beam current is integrated on the holder to normalize the collected charge, typically 10 to 50 microcoulombs per spectrum.
Data Acquisition
Backscattered ions are energy-analyzed by the detector, and pulse-height spectra of counts versus channel are accumulated using multichannel electronics. Energy calibration is performed with reference samples of known surface elements. For crystalline specimens, the goniometer is used to align the beam with a major crystal axis, and both aligned and random spectra are collected to quantify damage. Where hydrogen or light-element quantification is needed, the same accelerator infrastructure can simultaneously or sequentially perform elastic recoil detection analysis (ERDA) or nuclear reaction analysis (NRA).
Spectrum Simulation and Depth Reconstruction
Raw spectra are interpreted through dedicated simulation software that forward-models the scattering process. The analyst constructs a trial layer model, and the software computes the expected spectrum using kinematic factors, stopping powers, scattering cross sections, and straggling corrections, then iteratively refines layer thicknesses and compositions until the simulation matches the measured spectrum. Best practice includes evaluating the covariance of fitted parameters, applying corroborated stopping-power data consistent with internationally recognized tabulations, and reporting uncertainties for every fitted quantity.
Quality Assurance and Reporting
Reputable providers operate under quality frameworks aligned with ISO/IEC 17025 general requirements for the competence of testing and calibration laboratories, and use terminology consistent with ISO 18115, the international vocabulary standard for surface chemical analysis. Reports typically present measured and simulated spectra, tabulated layer models, elemental areal densities, molar fractions, thicknesses, and expanded uncertainties, together with full experimental conditions to ensure reproducibility.
Applications
RBS depth profiling occupies a unique position between high-sensitivity trace techniques and imaging methods, providing absolute quantification that anchors and cross-validates other measurements. Key application areas include:
- Semiconductor process development and qualification: verifying high-k dielectric composition, metal gate integrity, silicide formation, and implanted dopant dose retention against specification.
- Optical coating certification: confirming layer thicknesses and refractive-index-relevant stoichiometry in multilayer interference filters and mirror stacks.
- Photovoltaic research: optimizing compositional gradients in CIGS and CdTe absorbers and detecting diffusion of buffer or contact elements during annealing.
- Battery and energy storage studies: profiling interfacial reaction layers in thin-film batteries and solid electrolytes after formation and cycling.
- Metallurgy and corrosion science: measuring oxide scale thickness, passivation layer growth kinetics, and elemental interdiffusion in bonded or coated components.
- Ion implantation metrology: auditing retained dose and range distributions for implanters and surface modification processes.
- Cultural heritage and forensic science: non-destructive determination of composition and thickness of gilding, glazes, and patinas on artifacts.
- Academic materials research: benchmarking thin-film deposition reproducibility and validating composition inputs for property modeling.
FAQ
What elements can RBS detect, and what are the detection limits?
With a helium beam, RBS detects all elements heavier than the projectile, effectively from boron upward to uranium. Sensitivity scales approximately with the square of atomic number, so heavy elements in light matrices, such as gold in silicon, can be measured down to roughly 10 to the power of 13 atoms per square centimeter, corresponding to about 0.01 atomic percent, while light elements in heavy matrices have poorer limits, often near 1 atomic percent.
Can RBS detect hydrogen?
No. Hydrogen is lighter than the helium projectile and cannot backscatter it. Hydrogen depth profiling is instead performed with elastic recoil detection analysis (ERDA) using the same accelerator, and isotopes of light elements such as lithium, boron, nitrogen, and oxygen are best quantified by nuclear reaction analysis (NRA).
Is the RBS experiment destructive?
For most inorganic materials, RBS is effectively non-destructive, with negligible beam-induced modification at standard currents and doses. Sensitive organic or polymeric specimens may experience slight beam damage, which laboratories mitigate by spreading the beam or rastering over fresh areas.
What depth resolution and maximum analysis depth can be achieved?
Typical near-surface depth resolution is 5 to 15 nanometers for a 2 MeV helium beam at a 165 to 170 degree detector angle. The practical probing depth is approximately 1 to 2 micrometers in silicon and lighter matrices, and a few hundred nanometers in dense heavy-metal films, set by the energy loss of the incoming and outgoing ions.
How accurate are RBS composition and thickness results?
Because scattering cross sections are known analytically, RBS quantification typically achieves 1 to 5 percent relative accuracy for areal densities and stoichiometry without external standards, which is why the technique is frequently used to certify reference films for other methods.
How does RBS compare with SIMS and XPS?
Secondary Ion Mass Spectrometry offers much lower detection limits in the parts-per-billion range and superior ultimate depth resolution, but it is destructive and requires standards for quantification. X-ray Photoelectron Spectroscopy provides chemical-state information within the top few nanometers but probes a shallower region and needs sensitivity-factor calibration. RBS complements both by providing absolute, standardless quantification over a micrometer depth range, though without chemical bonding information.
What information does channeling RBS add?
In single-crystal samples, aligning the beam with a crystal axis dramatically reduces the backscattering yield from the lattice. Comparing aligned and random spectra quantifies lattice disorder, amorphized layer thickness, and the substitutional versus interstitial site occupation of impurities, which is essential for implantation damage studies and epitaxial film quality assessment.
What should be considered when submitting samples?
Submitters should provide approximate layer structure and expected elements, declare any vacuum-sensitive features, ensure samples fit within holder dimensions, and indicate flat, representative measurement areas. Multiple specimens can usually be mounted together to maximize throughput and cost efficiency.
