Shake-up Peaks #
In XPS, shake-up peaks are extra features that appear at slightly higher binding energy (BE) than the main core-level photoelectron peak. They are not new elements or contaminants — instead, they are satellite peaks caused by electronic excitations during the photoemission process.
When an X-ray photon ejects a core electron, the remaining atom is left in an excited, ionized state. Normally, the ejected electron carries away the difference between photon energy and the binding energy.
However, sometimes:
-
As the electron is emitted, another valence electron is simultaneously excited to a higher unoccupied orbital (like a π* or d* state).
-
This costs extra energy, so the photoelectron comes out with slightly less kinetic energy → meaning it appears at a higher binding energy in the spectrum.
This extra energy loss produces the shake-up peak, offset by a few electron volts from the main line.
Shake-up peaks are strongly associated with systems containing delocalized or partially filled electronic states:
-
Transition metal oxides → e.g. Cu²⁺ in CuO shows strong shake-up satellites in Cu 2p.
-
Aromatic/π-conjugated molecules → e.g. benzene, polymers, graphite (π→π* excitations).
-
Certain open-shell ions → where unpaired electrons can be promoted to higher energy states.
Position: They always occur at a fixed energy separation from the main peak (specific to material/electronic structure).
Shape: Typically weaker and broader than the main peak.
Not contamination: Unlike adventitious carbon or oxygen peaks, shake-ups track with the main peak intensity.
Element-specific: For example, Cu²⁺ shows strong satellites in Cu 2p, while Cu⁰ and Cu⁺ do not — making shake-ups useful for oxidation state identification.
Shake-off peaks are a special type of shake-up, where the excited electron is removed from the atom.
-
Prominent in atoms with delocalized or weakly bound electrons, since those valence electrons are easier to ionize during core-hole creation.
-
More noticeable in light elements (like C, N, O) compared to transition metals.
-
Often contributes to the “background” in XPS spectra, and is accounted for in background subtraction methods (e.g. Shirley or Tougaard backgrounds).
They look slightly dissimilar to shake-ups:
-
Shake-up = discrete, material-specific satellite peaks → great for identifying oxidation states and conjugated systems.
-
Shake-off = continuous, broad background → not diagnostic by itself, but explains the asymmetric tails in many XPS peaks.
