RBS Detection Technology: Rutherford Backscattering Spectrometry for Quantitative Thin Film and Surf

RBS detection technology (Rutherford Backscattering Spectrometry) is a non-destructive ion beam analysis technique that provides absolute, standard-free quantification of elemental composition, film thickness, and depth profiles in thin films and surface layers. This article explains the working principles of RBS, the types of samples and testing items it covers, standard testing methods, typical industrial and research applications, and answers to frequently asked questions.

Technical Overview

Rutherford Backscattering Spectrometry (RBS) is a well-established ion beam analysis technique used to determine the elemental composition, layer thickness, and in-depth distribution of elements in solids. In a typical RBS measurement, a collimated beam of high-energy light ions, most commonly 2 MeV helium ions (He2+), is directed onto the sample surface under high vacuum. A small fraction of these ions undergoes elastic Coulomb scattering with the nuclei of atoms inside the sample and is backscattered toward a detector, usually a silicon surface barrier detector placed at a scattering angle of approximately 160 to 170 degrees.

The energy spectrum of the backscattered ions contains three layers of information, each governed by a fundamental physical quantity:

• Kinematic factor: The energy retained by a backscattered ion depends on the mass of the target nucleus, which allows RBS to identify which elements are present in the sample.
• Stopping power: Ions lose energy continuously as they travel into and out of the sample, so the energy of backscattered particles encodes the depth at which the scattering event occurred, enabling depth profiling without sputtering.
• Rutherford scattering cross section: Because the scattering probability is known from first principles, RBS provides absolute quantitative results without the need for calibration standards.

The key performance characteristics of RBS detection technology include:

• Non-destructive analysis: The sample remains intact and can be re-measured or subjected to subsequent processing.
• Absolute quantification: Atomic concentrations are obtained directly, typically with a relative accuracy of 1 to 5 percent.
• Depth resolution: Approximately 5 to 20 nm near the surface for standard detector geometries, and below 5 nm when high-resolution magnetic spectrometers are used.
• Analysis depth: Roughly 1 to 2 micrometers for helium ion beams, extendable to about 10 to 20 micrometers with hydrogen or heavier probe ions.
• Element sensitivity: The scattering cross section scales approximately with the square of the atomic number, so RBS is highly sensitive to heavy elements (parts-per-million level for heavy metals on light substrates) but less sensitive to light elements such as carbon, nitrogen, and oxygen.

One important limitation is that RBS cannot directly detect hydrogen, because hydrogen nuclei are lighter than the probing ions. For hydrogen quantification, RBS instruments are often coupled with Elastic Recoil Detection Analysis (ERDA). A typical RBS system consists of an ion source, a single-ended or tandem electrostatic accelerator, an analyzing magnet for mass and energy selection, a beamline, a scattering chamber equipped with a multi-axis goniometer, and one or more particle detectors.

Testing Samples

Semiconductor and Microelectronic Samples

Silicon wafers and wafer fragments are the most common RBS samples in the microelectronics industry. Typical cases include high-k gate dielectric films such as hafnium oxide and zirconium oxide, metal silicides such as TiSi2, CoSi2, and NiSi, diffusion barrier layers such as TaN and TiN, metal interconnect stacks, and ion-implanted layers for dose verification.

Metal, Alloy, and Multilayer Films

RBS is well suited to sputtered or evaporated metal films, alloy thin films with unknown stoichiometry, and multilayer stacks in which individual layer thicknesses and interfacial mixing must be determined. Because the backscattering signal of heavy metals is strong, films just a few nanometers thick can be measured with high confidence.

Dielectric, Optical, and Protective Coatings

Oxide and nitride coatings such as SiO2, Si3N4, Al2O3, TiO2, and Ta2O5, as well as optical multilayer coatings and hard protective coatings, are routinely analyzed by RBS to verify stoichiometry and thickness uniformity. Compound semiconductor and photovoltaic absorber films such as GaAs, InGaAs, CIGS, and CdTe also fall within the scope of the technique.

General Sample Requirements

To obtain accurate and reproducible results, samples should satisfy the following conditions:

• Sample size: Typically 5 mm x 5 mm to 20 mm x 20 mm, or small pieces cut from larger wafers and coupons.
• Surface condition: A flat, smooth surface is preferred, since surface roughness degrades depth resolution.
• Vacuum compatibility: Samples must be stable under high vacuum of approximately 10 to the minus 6 torr and must not outgas significantly.
• Conductivity: Conductive samples are ideal; insulating samples can still be measured by using electron flood guns or thin conductive coatings to suppress charge buildup.

Testing Items

RBS detection technology covers a comprehensive set of quantitative measurement items for thin films and surface layers:

• Elemental composition and stoichiometry: Quantitative determination of atomic ratios in compound films, for example the Hf-to-O ratio in HfO2 or the metal-to-silicon ratio in silicides.
• Film thickness and areal density: Layer thickness in nanometers, or equivalently areal density in atoms per square centimeter, obtained independently of reference standards.
• Depth profiling: Concentration versus depth distributions of elements within the top one to two micrometers, reconstructed through spectrum simulation.
• Interface analysis: Detection of interfacial mixing, interdiffusion, and reaction layers between adjacent films after annealing or other thermal treatments.
• Ion implantation verification: Measurement of implanted dose, projected range, and straggling to validate implantation processes.
• Heavy metal contamination: Screening for surface contamination by heavy elements on light substrates such as silicon.
• Crystalline quality assessment: Combined with the channeling technique, RBS quantifies lattice disorder, damage depth distributions, and impurity lattice sites in single crystals.

Testing Methods

The standard RBS workflow consists of sample loading, chamber evacuation to high vacuum, beam alignment, spectrum acquisition, and quantitative spectrum simulation. Several measurement configurations are selected according to the analytical objective:

Normal Incidence RBS

The ion beam strikes the sample at normal incidence and backscattered particles are collected at a fixed angle near 170 degrees. This geometry offers the highest detection efficiency and is the default choice for composition and thickness measurement of most thin film samples.

Grazing Angle RBS

By tilting the sample so that the beam enters at a grazing angle, the effective path length inside the film is increased, which improves near-surface depth resolution. Grazing exit geometries provide a similar benefit and are used when ultra-shallow depth profiling is required.

RBS Channeling Analysis

For single crystalline samples such as silicon or GaAs wafers, the sample is aligned with a crystallographic axis using a goniometer so that ions channel along the rows of atoms. Comparing the channeled spectrum with a random-orientation spectrum reveals the degree of lattice damage, implantation-induced disorder, and amorphous layer thickness.

ERDA for Hydrogen Detection

When hydrogen content must be quantified, the detector is placed in the forward direction and recoil hydrogen atoms ejected by the incident beam are measured, often through a thin foil that blocks scattered primary ions. This complementary method extends RBS platforms to full elemental coverage including hydrogen.

Data Analysis and Reporting

Acquired energy spectra are fitted with simulation software such as SIMNRA or RUMP, which models the sample as a stack of layers with defined compositions and thicknesses. The simulation iteratively adjusts layer parameters until the calculated spectrum matches the measured one. The final report typically includes the measured energy spectra, layer-by-layer composition in atomic percent, thickness in nanometers and atoms per square centimeter, and estimated uncertainties.

Applications

Thanks to its standard-free quantification and non-destructive nature, RBS detection technology serves a wide range of industries and research fields:

Semiconductor Process Control

RBS is used extensively for metrology of gate dielectrics, silicide formation, barrier layer integrity, and implant dose monitoring. It serves as a reference method against which inline optical and X-ray metrology tools are calibrated.

Photovoltaics and Compound Semiconductors

Absorber layers in thin film solar cells, including CIGS and CdTe, require precise control of composition ratios that directly affect device efficiency. RBS provides the absolute composition data needed for process optimization and certification.

Thin Film Coatings and Surface Engineering

Manufacturers of optical coatings, hard protective layers, and decorative films use RBS to verify stoichiometry, detect oxygen or nitrogen incorporation, and evaluate coating uniformity across deposition batches.

Advanced Materials Research

In universities and national laboratories, RBS supports research on ion beam modification of materials, high-entropy alloys, battery electrode interfaces, corrosion layers, and diffusion studies. Its ability to measure atomic areal density makes it a benchmark technique for published thin film data.

FAQ

Is RBS a destructive testing method? No. The MeV helium ion beam deposits very little energy per unit area during a typical measurement, and the sample remains physically and chemically unchanged, allowing repeated measurements or subsequent device processing.

Can RBS detect hydrogen in my sample? Not directly, because hydrogen is lighter than the probe ions and cannot backscatter them. Hydrogen is measured on the same instrument using the ERDA technique, which detects recoil protons in the forward direction.

What is the detection limit of RBS? Sensitivity depends strongly on atomic number. For heavy elements such as gold, tantalum, or hafnium on a silicon substrate, detection limits reach the parts-per-million range. For light elements such as oxygen, nitrogen, and carbon, detection limits are typically on the order of 0.1 to 1 atomic percent.

How deep can RBS analyze? With standard 2 MeV helium ions, the practical analysis depth is about 1 to 2 micrometers. Using proton beams, the probing depth can be extended to roughly 10 to 20 micrometers at the cost of reduced mass resolution.

How does RBS compare with SIMS, XPS, and EDS? Unlike SIMS, RBS does not sputter the sample and does not require matrix-matched standards. Compared with XPS, RBS probes much deeper and provides absolute quantification. Compared with EDS in an electron microscope, RBS offers far better sensitivity to trace heavy elements and genuine depth information rather than a beam-integrated average.

What sample size and preparation are needed? A piece as small as 5 mm x 5 mm is sufficient. No special preparation is required beyond ensuring the surface is clean, dry, and flat. Full wafers can often be accommodated directly depending on the chamber design.

How accurate are RBS results? Composition results are typically accurate to within 1 to 5 percent in relative terms, and thickness results to within about 5 percent, limited mainly by stopping power database uncertainties and surface roughness.

How long does a typical RBS measurement take? A single standard spectrum can be acquired in 10 to 30 minutes, while a complete analysis including channeling alignment or multiple geometries usually takes one to two hours per sample.