Written and maintained by CASRAI Editorial Board
Published · Last updated
On this page
Angle-resolved photoelectron spectroscopy (ARPES), also called angle-resolved photoemission spectroscopy, is a technique that shines ultraviolet or soft X-ray light on a single-crystal sample in ultra-high vacuum and records the kinetic energy and the emission angle of the electrons that escape. Together these give the electron’s binding energy as a function of crystal momentum, which is the band structure E(k) of the solid.
Short answer
ARPES is the direct experimental method for studying the momentum-dependent electronic band structure of solids. Where an ordinary photoelectron spectrum tells you which energies electrons occupy, ARPES also tells you which direction they were travelling in, and that direction maps onto momentum inside the crystal. It needs an aligned single crystal, an atomically clean surface and ultra-high vacuum, which is why most ARPES work is done on dedicated end stations rather than on general-purpose surface-analysis instruments.
Definition and scope
ARPES belongs to the photoelectron-spectroscopy family. Every member of the family rests on the photoelectric effect: a photon of energy hν removes an electron from the sample, and the electron’s kinetic energy follows Ekin = hν − φ, where φ is the work function, typically 4–5 eV in metals (Damascelli, 2004). What distinguishes ARPES is that the analyser accepts electrons only within a small cone of emission angles. Scanning that angle, or collecting many angles at once, turns an energy spectrum into an energy-versus-momentum map.
The page Photoelectron spectroscopy (XPS and UPS) covers the chemistry-oriented side of the family, where the question is which elements and oxidation states are present at a surface. ARPES asks a different question: how do electrons in the solid disperse with momentum. The subjects are typically crystalline materials such as those described in quantum materials, where the band structure, Fermi surface and energy gaps are the quantities of interest.
Ultraviolet photoelectron spectroscopy and ARPES
Ultraviolet photoelectron spectroscopy (UPS) uses ultraviolet photons, commonly from a gas-discharge lamp, and probes the valence electrons. ARPES is usually carried out with ultraviolet or soft X-ray photons for the same reason: at these energies the photon excites valence-band electrons, and the low photon momentum lets the measured emission angle be translated into crystal momentum. The helium I line at 21.2 eV is the standard laboratory example, and it is the photon energy at which Damascelli quotes an angular resolution of 0.2° as equal to about 0.5% of the Brillouin zone for cuprates. In practice, UPS without angular resolution gives an angle-integrated valence-band spectrum; ARPES is the same experiment with the angle information kept. Photoemission with an Al Kα X-ray source at 1487 eV is a different regime: the photon momentum is about 50% of the zone, so it can no longer be neglected in the momentum mapping.
How the experiment works in practice
An ARPES end station has four essential parts: a light source, a sample on a manipulator, an electron analyser and a vacuum system. The geometry is simple. Light hits the sample, electrons leave, and the analyser records their kinetic energy for each emission direction. With the emission angle ϑ measured from the surface normal and the azimuth φ, the in-plane momentum components follow from K = √(2m·Ekin)/ħ, for example Kx = (1/ħ)√(2m·Ekin) sin ϑ cos φ (Damascelli, 2004). Because the surface preserves the in-plane momentum of the electron on escape, E(k) along the surface can be read from these angles.
The three-step model and the sudden approximation
The usual framework for interpreting the spectra splits photoemission into three steps: optical excitation of the electron in the bulk, travel to the surface, and escape through the surface barrier into vacuum. Combined with the sudden approximation, this picture links the measured intensity to the electron’s state in the solid, although it disregards many-body effects (Damascelli, 2004). Interaction effects such as electron-boson coupling are nonetheless among the things ARPES is used to study (Vishik, UC Davis).
The electron analyser
Most ARPES analysers are hemispherical. Electrons pass through a multi-element electrostatic lens, enter through a slit, travel between two concentric hemispheres held at a potential difference, and leave through an exit slit to a detector. The energy resolution of the analyser depends on the pass energy, slit width, mean radius and angular acceptance: ΔEa = Epass(w/R0 + α²/4) (Damascelli, 2004). Modern analysers use a two-dimensional position-sensitive detector, built from two micro-channel plates and a phosphor screen, so that a single exposure gives an energy-versus-momentum snapshot instead of a single point.
State-of-the-art instruments quoted in the same review reach energy resolution of a few meV and an angular resolution of about 0.2°. A monochromator on the light source can reach a resolving power equivalent to 1–2 meV at 20 eV.
What the data show
The course material from the Vishik group at UC Davis lists what ARPES is used to measure: band dispersion, Fermi surfaces, energy gaps (including charge-density-wave, superconducting and spin-density-wave gaps), electron-boson coupling, surface states and spin-momentum locking. A Fermi-surface map is built by joining many one-dimensional cuts through momentum space.
Light sources: laser, lamp or synchrotron
The choice of photon source sets the accessible photon energy, the bandwidth and the polarization, and it decides whether an experiment can be done in a home laboratory or needs a beamline. The data table at the end of this page compares the three source types as listed in the Vishik overview. In short, a He discharge lamp is a compact laboratory source with a strong line at 21.2 eV, a laser covers a lower and narrower energy range with very narrow bandwidth, and a synchrotron offers variable photon energy at the cost of beamtime and a different end station for each beamline.
Photon energy and surface sensitivity
Electrons that leave the sample without losing energy come from very near the surface. The electron mean free path has a minimum of about 5 Å at kinetic energies of roughly 20–100 eV, which is exactly the range used in most vacuum-ultraviolet ARPES (Damascelli, 2004). The consequence is that the spectrum reflects the top atomic layers, so it can differ from the bulk, and the surface must be atomically clean and well ordered. Soft X-ray photons of 500–1500 eV raise bulk sensitivity, with the mean free path going from about 5 Å at 120 eV to almost 20 Å at 880 eV, at the cost of energy and momentum resolution.
| Photon regime | What changes | Source |
|---|---|---|
| Vacuum ultraviolet, kinetic energies of about 20–100 eV | Mean free path at its minimum, about 5 Å: very surface-sensitive, best resolution | Damascelli, 2004 |
| Soft X-ray, 500–1500 eV | Greater bulk sensitivity (mean free path about 5 Å at 120 eV to nearly 20 Å at 880 eV), with lower resolution | Damascelli, 2004 |
| Al Kα, 1487 eV | Photon momentum is about 50% of the Brillouin zone, so it cannot be neglected | Damascelli, 2004 |
What a sample must satisfy: a checklist for facility users
Before booking beamtime or time on a laboratory end station, the following conditions come from the two readable sources and are worth confirming with the facility scientist.
- A single crystal, properly aligned. Angle-resolved and momentum-resolved measurement requires a properly aligned single crystal (Damascelli, 2004).
- An atomically clean surface. The Vishik overview names three routes: cleave the crystal in situ, grow the material in situ, or sputter and anneal.
- A surface prepared immediately before the experiment. Damascelli stresses data from atomically clean, well-ordered surfaces prepared right before measurement in ultra-high vacuum, and notes that surface states can differ from the bulk.
- Vacuum in the ultra-high range. The review quotes better than 5×10⁻¹¹ torr for state-of-the-art work. The table below shows why.
- Characterization of the surface. Low-energy electron diffraction and core-level XPS are the techniques named for checking surface order and cleanliness.
| Vacuum regime | Pressure range (torr) | Time to deposit one monolayer |
|---|---|---|
| High vacuum (HV) | 10⁻³ to 10⁻⁹ | 0.006 s to 95 min |
| Ultra-high vacuum (UHV) | 10⁻¹² to 10⁻⁹ | 95 min to 65 days |
The monolayer times come from the Vishik slides, which cite Hüfner. They explain the requirement: at high vacuum a surface can be covered in seconds, so only UHV leaves enough time to measure a prepared surface. Growing the material in situ, for example by molecular beam epitaxy, is one of the routes to a clean surface named in the Vishik slides. For related surface-analysis methods that also demand clean surfaces, see Auger electron spectroscopy.
Common misunderstandings
ARPES is a bulk probe
It is not. With a mean free path of about 5 Å in the common photon range, the signal comes from the first few atomic layers. Soft X-ray photons extend the depth but do not make the method purely bulk, and they reduce resolution.
Any solid sample can be measured
Momentum information needs an aligned single crystal. A powder or a polycrystal averages over directions and removes the angular information that defines the technique.
ARPES and XPS are the same experiment
Both detect photoelectrons, but they answer different questions. XPS, with Al Kα at 1487 eV, is the core-level technique, and Damascelli names core-level XPS as a way to characterize the surface. ARPES uses lower photon energies and an angle-resolving analyser to map valence-band structure.
Good high vacuum is good enough
The monolayer-formation times above show the gap: at the worst end of the high-vacuum range a monolayer forms in 0.006 s, while UHV allows from about 95 minutes to many days.
Where it matters for labs and equipment decisions
For core-facility managers and research groups, ARPES is rarely a stand-alone purchase. It combines a UHV chamber, a sample manipulator, an electron analyser, a photon source and surface-preparation tools. The sources used here describe the components and their performance figures, not prices or vendor choices, so any procurement comparison has to rest on the facility’s own specifications. What the sources do support is a planning list: whether the group needs a laboratory lamp or laser versus a beamline, whether crystals can be cleaved in situ, and whether the analyser’s resolution (a few meV, about 0.2° in state-of-the-art systems) matches the energy scale of the physics being studied.
Sources
- Damascelli, A. (2004). Probing the Low-Energy Electronic Structure of Complex Systems by ARPES. Physica Scripta T109, 61. arXiv:cond-mat/0307085. Read 2026-11-17.
- Vishik, I. Angle-resolved photoemission spectroscopy (ARPES) Overview, Physics 250, UC Davis. Course slides (PDF). Read 2026-11-17.
- Further reading: Damascelli, Hussain and Shen, Reviews of Modern Physics, 75, 473 (2003).
- Further reading: Zhang et al., Angle-resolved photoemission spectroscopy, Nature Reviews Methods Primers 2, 54 (2022).
Data: ARPES light sources compared: laser, gas-discharge lamp and synchrotron
Photon energy, bandwidth and other characteristics of the three light-source types used for angle-resolved photoemission, compiled from a university course overview of ARPES.
Period covered: 2004/2016.
| Light source | Photon energy | Bandwidth | Other characteristics |
|---|---|---|---|
| Laser | 6-11 eV | Can be much narrower than 1 meV | Variable polarization; potentially high intensity |
| Gas discharge lamp (He, Xe, Ne, Ar) | Discrete lines, e.g. 21.2 eV for He I (lines listed span 8.4 to 40.8 eV) | Below 1 meV with a monochromator | Low intensity; random polarization |
| Synchrotron | Variable (tunable) | 0.5 to several meV; tradeoff between bandwidth and intensity | Fixed polarization; different synchrotrons/endstations specialize in different energy ranges |
Sources
- Vishik, ARPES Overview, Physics 250, UC Davis (accessed 2026-11-17)
- Damascelli, Probing the Low-Energy Electronic Structure of Complex Systems by ARPES (accessed 2026-11-17)
Compiled by CASRAI from the sources above. Free to reuse under CC BY 4.0 with attribution.







