A photoelectron spectroscopy (PES) spectrum is a plot of electron binding energy against signal intensity, and reading one starts with recognizing that every peak represents electrons ejected from a specific atomic orbital in a specific element. The horizontal axis tells you how tightly those electrons were bound, and the vertical axis tells you how many of them the detector counted. From those two pieces of information you can figure out which elements are present, what chemical states they are in, and roughly how much of each element sits in the surface region of your sample. The details get richer once you learn to read the subtler features hiding in and around those peaks.
What the Axes Tell You
The x-axis of a PES spectrum displays binding energy, measured in electron volts (eV). By convention it runs in reverse: high binding energies sit on the left and low binding energies on the right. This means the most tightly held core electrons appear on the left side of the spectrum, while loosely held valence electrons appear on the right, near zero. If you’ve ever looked at a PES spectrum and felt disoriented because the numbers seem backward, that’s why.
The y-axis shows intensity, typically in counts or counts per second. A taller peak means more photoelectrons reached the detector at that particular binding energy. Peak height matters, but peak area matters more when you are trying to figure out how much of an element is present, because peaks can be narrow and tall or broad and short depending on the element and the instrument’s resolution.
One important thing to keep in mind: the binding energies you see are measured relative to the Fermi level of the sample, which serves as the zero-energy reference point. This convention works cleanly for metals and other conductors but introduces complications for insulators, which we’ll get to when discussing charge referencing.
Identifying Elements From Peak Positions
Every element has a unique set of core-level binding energies, like a fingerprint. Carbon 1s electrons appear near 285 eV, oxygen 1s near 530 eV, silicon 2p near 99–104 eV depending on chemical environment, and so on. When you first look at a spectrum, especially a wide survey scan that covers a broad energy range, you are essentially matching the peaks you see against a reference table of known binding energies to determine which elements are on your sample’s surface.
Peaks are labeled using the orbital notation of the electrons they come from. You will see labels like C 1s, O 1s, Si 2p, Fe 2p, and Au 4f. The number refers to the principal quantum level, and the letter to the orbital shape. For orbitals with angular momentum (p, d, f), you often see doublet peaks rather than single ones. An Fe 2p signal, for instance, splits into 2p₃/₂ and 2p₁/₂ components because of spin-orbit coupling. The spacing and intensity ratio of those doublets are fixed by physics, so if you see a doublet that doesn’t match the expected ratio, something else is going on, maybe an overlap with another element’s peak.
Survey scans cover a wide binding-energy window at lower resolution to give you a quick inventory of what’s on the surface. Once you know which elements are present, you can run high-resolution scans of individual peaks to look more closely at chemical shifts and fine structure.
Reading Chemical Shifts
The exact position of a peak shifts depending on the chemical environment of the atom. A silicon atom bonded to other silicon atoms in a pure crystal will show its Si 1s peak at a different binding energy than a silicon atom bonded to oxygen in silicon dioxide. In high-resolution hard X-ray photoelectron spectroscopy of thin silicon oxide films, for example, the Si 1s spectrum can resolve distinct chemical states at the interface between the oxide layer and the silicon substrate underneath.1Nuclear Instruments and Methods in Physics Research Section A. High-resolution hard X-ray photoelectron spectroscopy: Application of valence band and core-level spectroscopy to materials science The more oxidized the silicon, the higher its binding energy shifts, because the surrounding oxygen atoms pull electron density away from the silicon core.
This makes chemical shifts one of the most powerful features in PES. By measuring exactly where a peak sits, you can distinguish between different oxidation states, different bonding partners, and different functional groups on a surface. A carbon 1s peak from a C–C bond sits at a different energy than one from a C–O bond, which sits at a different energy than one from a C=O bond. Peak fitting software can decompose a broad, asymmetric peak into those individual contributions.
There is an important pitfall here, though. Not every shift in binding energy means the atom’s chemical environment has changed. Since binding energy is measured relative to the sample’s Fermi level, a change in the Fermi level itself will shift every peak by the same amount. In transition-metal oxides, for example, researchers have shown that what looks like an oxidation-state change can sometimes be an “electronic” shift caused by the Fermi level moving rather than a true chemical shift of the core level.2ACS Nano. How to Correctly Analyze 2p X‑ray Photoelectron Spectra of 3d Transition-Metal Oxides: Pitfalls and Principles If you are looking at a material where the electronic structure can vary, like a semiconductor or a metal oxide, you need to check whether the shift you see is from a real change in chemical bonding or just a Fermi-level shift affecting the entire spectrum.
The Background Underneath Your Peaks
If you look at a PES spectrum, you’ll notice that the peaks don’t sit on a flat baseline. The background signal rises in a staircase-like pattern as you move to higher binding energies. This happens because some photoelectrons lose energy through inelastic collisions on their way out of the sample. Those energy-loss electrons contribute a broad, featureless signal that piles up on the high-binding-energy side of each peak.
Before you can measure a peak’s area accurately, you need to subtract this background. How you do that matters more than it might seem. The most common approach is the Shirley background, which models the step-like rise as proportional to the total peak area above the background on the lower-binding-energy side. It works well for many materials, but it has limitations, especially for complex spectra with overlapping peaks or when the true background shape doesn’t follow a simple step.3Journal of Vacuum Science & Technology A. Introductory guide to backgrounds in XPS spectra and their impact on determining peak intensities
More sophisticated methods have been developed to handle tricky cases. One recent approach, the narrow-Shirley method, replaces the standard step function with one that mimics the step near the peak but decays toward zero at higher binding energies, avoiding the artifact of an endlessly rising background far from the peak.4PubMed Central. Characterizing the Intrinsic Background in XPS Using the Narrow-Shirley Approach For routine analysis, the standard Shirley or a linear background often suffice. But if you are doing careful quantitative work or fitting overlapping peaks, the choice of background model can change your results, so it is worth being deliberate about it.
Satellite Peaks and Other Extra Features
Not every peak in a PES spectrum corresponds to a straightforward photoelectron from a core level. You will sometimes see smaller peaks a few electron volts to the higher-binding-energy side of a main peak. These are often shake-up satellites, which arise when the departing photoelectron transfers some of its kinetic energy to a valence electron, exciting it to a higher state. The photoelectron arrives at the detector with less kinetic energy than expected, which the instrument records as a higher binding energy.5Journal of Electron Spectroscopy and Related Phenomena. Review Shake-up satellites in X-ray photoelectron spectroscopy
Shake-up satellites are not just noise to be ignored. Their presence, position, and intensity carry real chemical information. In transition-metal compounds, for instance, the satellite pattern around the 2p peaks can help you distinguish between different oxides of the same metal. Early work demonstrated that shake-up satellites in copper and nickel oxide spectra could be used to analyze oxide layers on metal surfaces.6Nature Physical Science. Use of “Shake-up” Satellites in Photoelectron Spectra for Analysis of Oxide Layers on Metals If you ignore satellite peaks in your peak fitting, you risk misattributing their intensity to other chemical components.
You may also encounter Auger peaks in your XPS spectrum. These come from a secondary relaxation process: after a core electron is ejected, an outer electron fills the hole, and the energy released ejects yet another electron. Auger peaks appear at fixed kinetic energies rather than fixed binding energies, so their position on the binding-energy scale shifts when you change the X-ray source energy. This is actually a useful diagnostic. If you switch from one X-ray source to another and a peak moves, it is an Auger peak, not a photoelectron peak. Auger peaks can overlap with photoelectron peaks and cause confusion if you are not aware of them, but they also provide additional chemical information through the Auger parameter, which combines the Auger peak position with the corresponding photoelectron peak position to give a source-independent measure of chemical state.
Turning Peak Areas Into Composition
Once you have identified your peaks and subtracted the background, you can estimate the elemental composition of the surface. The standard approach treats the sampled region as if it were a uniform slab and applies sensitivity factors to the measured peak areas. Each element and orbital has a different cross-section for photoemission, meaning some orbitals produce stronger signals per atom than others. Sensitivity factors correct for this, converting raw peak areas into values proportional to atomic concentration. The result is usually expressed as an atomic percent, giving you the equivalent composition of the surface if it were perfectly homogeneous.7Surface and Interface Analysis. Surface Analysis Insight Note: Uncertainties in XPS Elemental Quantification
This “equivalent homogeneous composition” is a useful approximation, but it is just that. Real surfaces are rarely homogeneous. They have layers, gradients, and patches of different composition. The sensitivity-factor approach smears all of that into a single average number. If you need to know the actual depth distribution of elements, you’ll need depth profiling, where the surface is gradually sputtered away while spectra are collected at each step. Still, for comparing samples or tracking changes in surface chemistry, the atomic-percent approach works well as a first pass.
PES is also inherently surface-sensitive, probing only the outermost few nanometers. Photoelectrons generated deeper in the material lose energy through inelastic scattering before they can escape, so the signal drops off exponentially with depth. The effective sampling depth depends on the kinetic energy of the electrons and the material they’re traveling through, but for typical laboratory XPS it’s roughly 5 to 10 nanometers.8Journal of Electron Spectroscopy and Related Phenomena. Relationships between electron inelastic mean free paths, effective attenuation lengths, and mean escape depths This means your composition numbers describe the surface, not the bulk, which is exactly the point for many applications but a limitation if you want bulk chemistry.
Charge Referencing With Adventitious Carbon
When you measure an insulating sample, something annoying happens: the surface accumulates positive charge as photoelectrons leave, because the insulator cannot replenish them fast enough. This charging shifts all the peaks in your spectrum to higher apparent binding energies by an unknown amount. Without correcting for this, your binding energies are wrong and your chemical-state assignments become unreliable.
The most common fix is to use adventitious carbon as an internal reference. Almost every sample that has been exposed to air picks up a thin layer of carbon-containing contamination from atmospheric organic molecules. The C 1s peak from this adventitious carbon is set to a reference value, and every other peak in the spectrum is shifted by the same amount. The conventional reference value is 284.8 eV for the C–C/C–H component of the adventitious carbon peak.9Surface and Interface Analysis. The Utility of Adventitious Carbon for Charge Correction: A Perspective From a Second Multiuser Facility
How reliable is this? Studies from multi-user XPS facilities have tested the approach across hundreds of samples. One facility-wide survey found an average adventitious carbon C 1s position of about 284.91 eV with a standard deviation of 0.25 eV, and reported that the method gave satisfactory results in roughly 95% of more than 500 cases assessed.10Applied Surface Science. Accessing the robustness of adventitious carbon for charge referencing (correction) purposes in XPS analysis: Insights from a multi-user facility data review A separate study using a fitting method that accounts for different carbon species within the adventitious layer found the C–C/C–H peak position to be 284.81 eV, again with about 0.25 eV of spread.11Applied Surface Science. Defining the nature of adventitious carbon and improving its merit as a charge correction reference for XPS
So the method works well most of the time, but the scatter means you should expect an uncertainty of about ±0.3 eV in your corrected binding energies. For most chemical-state assignments, that’s fine. For very subtle shifts, like distinguishing two states separated by less than half an eV, adventitious carbon referencing may not be precise enough, and you might need an alternative internal reference or the Auger parameter approach. Differential charging, where different parts of a mixed conducting/insulating sample charge up by different amounts, is another common source of trouble. Electrically isolating (floating) the sample during measurement helps reduce this problem.10Applied Surface Science. Accessing the robustness of adventitious carbon for charge referencing (correction) purposes in XPS analysis: Insights from a multi-user facility data review
Depth Profiling in Battery Research
One of the areas where PES spectrum reading has become indispensable is battery research, particularly the study of the solid electrolyte interphase (SEI) that forms on electrode surfaces in lithium-ion and lithium-metal batteries. The SEI is a thin, chemically complex film that controls how lithium ions move between the electrolyte and the electrode. Understanding its composition and layered structure is critical for improving battery lifetime and safety.
Researchers use XPS combined with sputter depth profiling to peel back the SEI layer by layer. In one approach applied to commercial lithium-ion cells, a series of sputtering steps with an argon ion beam progressively removed material while XPS spectra were recorded at each stage. This revealed how the relative concentrations of carbon, oxygen, fluorine, lithium, and phosphorus changed with depth through the SEI.12PubMed. Interface investigations of a commercial lithium ion battery graphite anode material by sputter depth profile X-ray photoelectron spectroscopy More recent work has tracked SEI growth over the full lifetime of automotive battery cells using the same sputter-and-measure technique.13Journal of Power Sources. In-depth analysis of the SEI thickness growth over the lifetime of Li-ion cells from an automotive battery
A persistent challenge is that conventional XPS sample preparation can alter the SEI. Rinsing the electrode and drying it before loading it into the spectrometer strips away soluble species, potentially giving an incomplete picture. Cryogenic XPS has emerged as a way around this: the sample is flash-frozen with electrolyte still in place, preserving the native wet chemistry. When coupled with gentle cluster-ion-beam sputtering, cryo-XPS has shown that the native SEI contains substantially more inorganic species like lithium fluoride and lithium carbonate than conventional dry-sample measurements would suggest.14PubMed. Depth-Resolved Probing of Native Solid Electrolyte Interphase Formation and Dynamics in Li Metal Batteries by Cryogenic X-Ray Photoelectron Spectroscopy Reading PES spectra from these experiments requires paying close attention to how the sample was prepared, because the answer you get depends heavily on whether volatile and soluble components were preserved or lost.
Angle-Resolved and Near-Ambient-Pressure Variants
Standard XPS collects photoelectrons from all emission angles together, giving an average over the sampling depth. Angle-resolved photoelectron spectroscopy (ARPES) instead measures both the kinetic energy and the emission angle of each photoelectron. In the context of crystalline materials, this provides a direct map of the electronic band structure, showing how electron energies vary with momentum. ARPES has become a central tool for studying quantum materials, superconductors, and topological insulators, where the arrangement of electronic bands determines the material’s exotic properties.15PubMed. Angle, Spin, and Depth Resolved Photoelectron Spectroscopy on Quantum Materials The spectra look quite different from standard XPS: instead of simple peaks on a binding-energy axis, you see two-dimensional maps of intensity as a function of energy and angle (or momentum), with bright bands tracing the electronic structure.
Another growing variant is near-ambient-pressure XPS (NAP-XPS), which allows spectra to be collected while the sample sits in a gas environment rather than in ultra-high vacuum. Traditional XPS requires pressures below about a billionth of atmospheric pressure because gas molecules scatter photoelectrons. NAP-XPS uses differential pumping and small apertures to maintain vacuum near the detector while keeping the sample at pressures up to a few millibar. This makes it possible to watch catalytic reactions happening on a surface in real time, identifying which chemical species form and disappear as the gas composition or temperature changes.16PubMed. Unraveling Catalytic Reaction Mechanism by In Situ Near Ambient Pressure X-ray Photoelectron Spectroscopy Reading NAP-XPS spectra involves the same peak-identification and chemical-shift skills as conventional XPS, but with added attention to gas-phase peaks that can appear in the spectrum and to the possibility that the surface chemistry is changing while you measure it.
Common Mistakes When Reading a PES Spectrum
Misidentifying an Auger peak as a photoelectron peak is probably the most frequent beginner error. If you see an unexpected peak that doesn’t match any element you’d expect on your sample, check whether it falls at a kinetic energy corresponding to a known Auger transition of an element you’ve already identified. Switching X-ray sources, if your instrument allows it, immediately resolves the question.
Another common mistake is fitting too many components into a single peak envelope. Peak fitting is partly an art, and it is tempting to keep adding Gaussian-Lorentzian components until the residual looks perfect. But every component you add should have a physical justification. A carbon 1s spectrum from a polymer might legitimately contain three or four components representing different functional groups. The same peak from a simple metal carbide probably doesn’t need five.
Ignoring the background model is a subtler error. Two analysts using the same raw data but different background choices can arrive at meaningfully different peak areas, which then propagate into different composition numbers.3Journal of Vacuum Science & Technology A. Introductory guide to backgrounds in XPS spectra and their impact on determining peak intensities Reporting which background you used and why is good practice, and comparing your results with a different background model is a quick sanity check.
Finally, treating chemical shifts as a simple lookup can lead you astray with materials where the Fermi level varies. In transition-metal oxides and semiconductors, always consider whether a peak shift might reflect an electronic shift rather than a genuine change in oxidation state.2ACS Nano. How to Correctly Analyze 2p X‑ray Photoelectron Spectra of 3d Transition-Metal Oxides: Pitfalls and Principles Cross-referencing with the Auger parameter or looking at the overall shift pattern across multiple core levels can help distinguish the two.