X-Ray Photoelectron Spectroscopy (XPS)

Read XPS through a binding-energy and chemical-state lens rather than a survey or element-presence lens. This fundamental shift in perspective transforms how we interpret X-ray photoelectron spectroscopy data collected from surfaces, interfaces, and thin films. The measured photoelectron kinetic energy is not an end in itself; it is the entry point to Einstein's photoelectric equation, which allows us to extract the binding energy EB = h times nu minus KE minus phi, the quantity that truly encodes the chemical identity and oxidation state of each element. XPS is therefore not a tool for answering "which elements are present" but rather "what oxidation state and chemical environment does each element occupy." A core-level peak shifted by 3 electron-volts from the elemental reference is not a minor perturbation—it is a definitive sign of chemical bonding or oxidation, and must be reported with the binding-energy reference, the chemical-state assignment, and the concentration (from peak area and relative sensitivity factors).

The binding energy of a core-level peak directly encodes the oxidation state and chemical environment of an atom.

X-ray photoelectron spectroscopy operates via the photoelectric effect: a monochromatic X-ray photon of energy h times nu (where h is Planck's constant and nu is frequency) strikes a sample, ejecting a core-level electron. That photoelectron has kinetic energy KE, which is measured by a hemispherical analyzer. The binding energy EB is then extracted from EB equals h times nu minus KE minus phi, where phi is the spectrometer work function. For aluminum K-alpha X-rays, the photon energy is 1486.6 eV; for magnesium K-alpha, it is 1253.6 eV. A carbon 1s electron at 284.8 eV binding energy (the adventitious hydrocarbon reference) will emerge with kinetic energy approximately 1200 eV from an Al-source spectrum, directly encoding the bonding environment of that carbon atom. Silicon 2p core electrons in elemental silicon show a binding energy around 99.5 eV; in silicon dioxide, the same 2p orbital shifts to approximately 103.5 eV—a chemical shift of 3 to 4 eV that unambiguously identifies the oxidation state change from Si to Si plus four.

Chemical shifts in core-level binding energy scale predictably with oxidation state and enable quantitative oxidation-state assignment without ambiguity.

The magnitude of a chemical shift depends on the effective nuclear charge experienced by the core electron, which changes as the atom's valence-electron configuration (oxidation state) changes. In a reduced state, fewer electrons are stripped away, so the binding energy is lower; in an oxidized state, more valence electrons are transferred to the oxidizing atom, so the effective positive charge rises and the binding energy increases. For silicon, the shift from elemental Si (99.5 eV) to Si plus two (roughly 100.5 eV) to Si plus four in SiO2 (103.5 eV) is approximately 1 eV per oxidation-state change. Oxygen 1s peaks in the SiO2 layer appear at approximately 532 eV binding energy, while oxygen in a water overlayer might appear 1 to 2 eV higher in binding energy due to hydrogen bonding. Nitrogen 1s binding energies shift by tens of millivolts between N-H and N-C bonding environments, enabling subtle bonding-state discrimination. Sulfur 2p peaks shift by several eV between metallic, sulfide, sulfoxide, and sulfate oxidation states, providing unambiguous speciation on buried interfaces and corrosion products.

Peak area (integrated counts), corrected by element-specific relative sensitivity factors, directly measures atomic concentration and composition.

The intensity of a core-level peak is proportional to the number of atoms of that element in the probed volume. The proportionality constant is the relative sensitivity factor (RSF), which depends on the cross-section for photoionization, the angular distribution of ejected electrons, and the transmission efficiency of the analyzer. RSF values for common elements have been tabulated (for example, Si 2p has an RSF near 0.62 on an Al-Kα spectrometer, while O 1s has an RSF near 2.93). By dividing the measured peak area by the RSF, we obtain a quantity proportional to the number of atoms. Ratioing two such corrected peak areas yields the atomic concentration of one element relative to another. For a 10 nanometer SiO2 layer on silicon, the ratio of the integrated O 1s peak area (from the overlayer) to the Si 2p peak area (from both the overlayer and substrate) changes with overlayer thickness, enabling non-destructive depth profiling without removing material. Quantitative compositional analysis of oxides, nitrides, and multi-layer structures relies critically on this peak-area integration and RSF correction.

Peak width (full width at half maximum, FWHM) and lineshape encode crystal quality, chemical inhomogeneity, and sample charging.

A sharp, symmetric core-level peak with a FWHM of 0.5 to 1 eV indicates high crystallinity and a uniform chemical state. A broadened peak (FWHM 1.5 to 3 eV) suggests chemical inhomogeneity—for example, a mixture of oxidation states or disorder in the bonding environment. Satellites and shake-up peaks (satellites appearing 5 to 10 eV higher in binding energy) arise from excitation of outer-shell electrons during the photoionization process and provide additional information about the electronic structure and hybridization. An insulating sample experiencing charge accumulation will exhibit peak shifts of 1 to 10 eV to higher binding energy, and a "charging correction" must be applied (typically by normalizing the C 1s peak to 284.8 eV) before chemical-shift assignments can be made reliably. Many modern spectrometers include a charge-neutralization electron flood gun to suppress charging, ensuring linearity and reproducibility of binding energies.

Start([Sample Preparation])
Start --> Clean["Clean surface in situ or ex situ"]
Clean --> Mount["Mount sample on sample holder"]
Mount --> Evacuate["Evacuate to UHV: below 1e-8 torr"]
Evacuate --> Align["Align sample to X-ray and analyzer"]
Align --> Survey["Acquire survey scan: 0-1200 eV binding energy"]
Survey --> Identify["Identify elements and oxidation states"]
Identify --> HighRes["Acquire high-resolution core-level scans"]
HighRes --> PassEnergy["Set pass energy 20-100 eV for high resolution"]
PassEnergy --> Measure["Measure kinetic energy of photoelectrons"]
Measure --> CalculateBE["Calculate binding energy from Einstein equation"]
CalculateBE --> FitPeaks["Fit core-level peaks to Lorentzians and satellites"]
FitPeaks --> ChemicalShift["Extract chemical shift and assign oxidation state"]
ChemicalShift --> Quantify["Quantify atomic concentration from peak area and RSF"]
Quantify --> ChargeCorrect["Apply charge correction if needed (C 1s at 284.8 eV)"]
ChargeCorrect --> End([Chemical-state and composition report])

X-ray photoelectron spectroscopy remains the gold standard for surface and interfacial chemical analysis in semiconductor, materials, and catalysis research. Commercial systems from Keysight, Keithley, Semilab, and NIST-traceable facilities typically employ monochromatic aluminum or magnesium X-ray sources, hemispherical analyzers with pass energies from 5 eV to 200 eV, and energy resolutions of 0.2 eV or better for high-resolution core-level spectroscopy. Modern spectrometers integrate charge-neutralization electron flood guns, variable-angle capabilities (enabling depth profiling via changes in take-off angle), and even spatially resolved micro-XPS probes with spot sizes of 10 micrometers to 200 micrometers, enabling chemical mapping across wafers and devices. Time-resolved XPS (using pulsed X-rays and a time-of-flight analyzer) can achieve microsecond-level temporal resolution, and resonant XPS (exciting specific near-edge resonances) can enhance chemical selectivity. Cross-validation with complementary techniques—four-point probe for conductivity, ellipsometry for film thickness, AFM for surface morphology, deep-level transient spectroscopy for defect levels, and SIMS for subsurface profiling—provides a complete chemical and electrical picture of interfaces and buried layers.

Depth profiling via angle-resolved XPS or sputtering-based XPS permits non-destructive (or destructive with depth control) access to buried interfaces and composition gradients.

Taking XPS measurements at different take-off angles (0 degrees to 80 degrees from the surface normal) alters the probing depth via the attenuation length: at grazing angles (high take-off angle), only the topmost 1 nanometer to 2 nanometers contributes to the signal; at normal incidence (low take-off angle), the probing depth is 5 nanometers to 10 nanometers, depending on material and kinetic energy. This angular variation enables non-destructive depth profiling by comparing core-level peak areas at different angles, reconstructing the composition profile. Destructive depth profiling uses ion sputtering (typically 1-5 kilovolts of argon ions) to gradually remove surface layers while repeatedly acquiring XPS spectra, creating a depth-versus-composition profile. Ion-sputtering rates must be calibrated (typically 1 nanometer to 10 nanometers per minute, depending on ion current and impact angle), and sputter damage must be minimized via low ion flux (below 1 microampere per cm squared) and argon sputtering (rather than xenon) on sensitive samples. An alternating sequence of XPS acquisition (10 seconds to 60 seconds per measurement) and ion sputtering (30 seconds to 300 seconds per cycle) yields compositional depth profiles with nanometer-scale resolution.

We read XPS through a binding-energy and chemical-state lens, interpreting core-level peak positions as definitive measures of oxidation state and chemical environment, peak areas as quantitative atomic concentrations (after RSF correction), and peak widths as indicators of chemical homogeneity and sample quality. This lens reveals why a seemingly small 1 eV or 2 eV shift in peak position is not noise but critical evidence of a different bonding or oxidation state, and why peak-fitting accuracy and binding-energy referencing are essential for valid chemical interpretation. XPS remains indispensable in interface engineering, oxidation-state determination, and subsurface composition profiling because it uniquely combines surface sensitivity (probing only the outermost 1-10 nanometers), chemical selectivity (via binding-energy shifts), and quantitative composition measurement (via peak areas and RSFs).