Knowledge Base

Doublet Separations

• No relevant non-S-orbital emissions

The Energies Listed are Binding Energies!

• O 1s: 530 eV
• O 2s: 20 eV
• O 2p: 8 eV

The Energies Listed are Binding Energies!

Overlaps for O 1s (primary emission)

• Na KLL (Al kα X-rays) (500 eV)
• V 2p (512 eV)
• Dy MNN (Al kα X-rays) (517 eV)
• Sb 3d (528 eV)
• Pd 3d (531 eV)
• At 4d (533 eV)

Energies listed are Kinetic Energies!

O KLL: ~ 505 eV

Oxygen spectra of metal oxide systems ranges from simple to highly complex, particularly in the case of mixed metal oxide systems. Often times the dominant signal may be attributed to the lattice oxygen – for simple systems such as SiO₂ or TiO₂ (Figure 1). These tend to be accompanied by a secondary signal at higher binding energy associated with surface hydroxyls.

Figure 1: Lattice O and surface hydroxyls of P-25 TiO₂ and SiO₂ nanospheres.¹

Appreciation of the lattice peak relative to the hydroxyl may in som cases be tricky, with real surfaces rarely presenting themselves free from surface terminal groups. CryoXPS of liquid metal oxides however, presents the capability to isolate the M-O species free from additional complications, for the right system (Figure 2).

Figure 2: CryoXPS of Titanium tert-Butoxide^1,2

Oxygen analysis of mixed metal oxides may prove advantageous when considering quantification analysis of inhomogeneous metal oxides – for example titania monolayers grown onto silica supports,³ since the two species have comparable kinetic energies rendering differing escape depths (of the Ti 2p and Si 2p core lines) a redundant factor when considering surface composition (Figure 3).

Figure 3: O 1s region for a Ti oxide overlayer on Si oxide support.¹

Surface basicity

Analysis of the oxygen auger region may yield crucial information regarding surface basicity, particularly crucial when considering solid base catalysts. The auger parameter provides insight into the surface polarizability of the solid material, and hence may be correlated quantifiably with surface basicity (Figure 4).⁴

Figure 4: Correlating O 1s – KLL auger parameter with surface basicity.⁴

O 1s spectra can be used as one way of estimated material band gaps, albeit with some limitations, first recorded by Nichols et al for low-k organosilicate dielectrics.[5]

Measurement of a material band gap using the O 1s core emission and onset off loss structure

Since, this method has been used to study many materials, including alumina,[6] AlGa,[7] and HfO₂.[8] It should, however, be noted that this method may not be applicable to small band gap materials, due to the natural line width of O 1s emissions, and potential formation of surface defects/hydroxyls resulting in a high energy shoulder.

It has become an unfortunately entrenched approach to attribute high energy photoemission intensity to the presence of oxygen vacancies in metal oxide, semiconductor systems. Simply put, a vacancy cannot produce a photoelectrons signal, because it is a lack of an atom – and therefore does not have electrons to emit! A number of recent works have tackled this problem in the literature (See, Morgan [9], Idriss [10], and Wang, Mueller and Crumlin [11]). It is of far more use to study the oxidation state of the metal component and determine vacancies through this lens.

Now this is not to say that XPS may not be used in the inference of a changing vacancy content by analysis of an O 1s region, however a deep understanding of the surface is required, and of the possibly states that may form – and is still best performed alongside a complementary technique (e.g. EPR and FTIR) to ensure the model developed is applicable and accurate.

For example, in the case of TiOx – the presence of hydroxyls and adsorbates has been used to infer changing vacancy contents, and compared with VB, and metal core line XPS.[12] Frankcombe and Liu [13] have determined theoretically that such adsorbates on ZnO represent a more robust appreciation of the actual surface, compared to assignment as vacancies.