Understanding Angle Resolved Photoemission Spectroscopy in Solids

[nealh149]

Does anybody have a link to an introduction to the theory and techniques surrounding ARPES?

 

[Bob S]

Here is a description of the work on ARPES being done at Stanford University.
http://arpes.stanford.edu/research.html
Many years ago I read (in The Atomic Nucleus by Evans) that the photoelectrons tended to come off at right angles with respect to the photon, and along the E vector for polarized light.

 

[ZapperZ]

Here is a description of the work on ARPES being done at Stanford University.
http://arpes.stanford.edu/research.html
Many years ago I read (in The Atomic Nucleus by Evans) that the photoelectrons tended to come off at right angles with respect to the photon, and along the E vector for polarized light.

Er... no. That would not make any sense since we have normal emission all the time.

Note that in most photoemission, the normal component of the momentum is not conserved. Only the in-plane momentum is, especially in layered, 2D structures.

Zz.

 

[Bob S]

Er... no. That would not make any sense since we have normal emission all the time.

Note that in most photoemission, the normal component of the momentum is not conserved. Only the in-plane momentum is, especially in layered, 2D structures.

Zz.

In Evans, The Atomic Nucleus, in chapter 24, Photoelectric Effect, in paragraph b Directional Distributions of Photoelectrons (page 696), it states "Especially at low photon energies, the photoelectrons tend to be ejected along the electric vector of the incident radiation, hence at right angles to the direction of incidence." Several plots of angular distributions are also shown.

The photoelectric interaction of photons with electrons cannnot occur on free electrons, because energy and momentum cannot be simultaneously conserved. So there has to be something that can absorb recoil momentum. It is also hard to calculate exactly on bound electrons. This is the main reason why the photoelectric cross section drops off so quickly above the binding energy of K-shell electrons, and the Compton cross section becomes relatively larger (until pair production becomes dominant). In Compton scattering, a secondary photon plus the Compton electron together can simultaneously match both the energy and momentum of the incoming photon.

 

[ZapperZ]

In Evans, The Atomic Nucleus, in chapter 24, Photoelectric Effect, in paragraph b Directional Distributions of Photoelectrons (page 696), it states "Especially at low photon energies, the photoelectrons tend to be ejected along the electric vector of the incident radiation, hence at right angles to the direction of incidence." Several plots of angular distributions are also shown.

The photoelectric interaction of photons with electrons cannnot occur on free electrons, because energy and momentum cannot be simultaneously conserved. So there has to be something that can absorb recoil momentum. It is also hard to calculate exactly on bound electrons. This is the main reason why the photoelectric cross section drops off so quickly above the binding energy of K-shell electrons, and the Compton cross section becomes relatively larger (until pair production becomes dominant). In Compton scattering, a secondary photon plus the Compton electron together can simultaneously match both the energy and momentum of the incoming photon.

You need to keep in mind of two things:

1. ARPES occurs in solids, not atoms or molecules. This means that the band structure of the solid plays a role in the momentum distribution. If you look my avatar, that is the RAW ARPES data where the vertical axis represents the energy, while the horizontal axis represents momentum distribution. The center of the band is at the lower end of the dispersion curve. You can orientate the electric field polarization anyway you want, but if the symmetry of the band prohibits it, you'll get NO emission. You also need to keep in mind if it is a single-crystal or a polycrystalline material, because of such symmetry consideration.

2. ARPES are done (as least for now) only within the first 1 eV or so of the Fermi energy. This means that this is NOT core-level photoemission. Why only first 1 eV? Because of energy and momentum resolution! The larger the photon energy that one uses, the more one sacrifice the energy resolution. Many of the material's properties that ARPES are being used to study occur at the low energy range. So your discussion on K-shell electrons, etc. are really not that relevant as far as ARPES are concerned.

One of the standard text for photoemission spectroscopy is Hufner's "Photoemission Spectroscopy" book. There are also 2 good reviews on ARPES technique, with particular application to high-Tc superconductors:

http://arxiv.org/abs/cond-mat/0209476
http://arxiv.org/abs/cond-mat/0208504

Zz.