Does charging of a sample create shifts, or can it also alter spectral features in XPS or UPS spectra? Also what is the role of biasing in UPS and why is it necessary?
Dear Monu, usually sample charging only shifts the spectra (you will observe larger binding energies if the sample is charged). In the case of UPS, however, this can be crucial because some of the excited electrons - those with very small kinetic energies - will not be able to leave the sample any more, because the level of their kinetic energy is below the vacuum level of the detector (energy analyser). In that sense, a bias voltage may help to overcome these issues (see the attached chart). Hope this has helped, Dirk
Charging will also cause the peaks to broaden. Often an electron flood gun is used to reduce charging effects.
Like Dirk mentioned charging is a severe problem in XPS, specially in monochromatic XPS where you don't have the secondaries from the Al window for compensation. In semiconductors an additional case of charging may occur when the photo generated carriers exceed the number of majority carriers. You have a splitting of the Fermi level in the one at the back contact and the quasi Fermi levels at the surface under investigation (surface photo voltage, see Alonso, Cimino, Horn ). Biassing the sample in UPS is not only to overcome the difference in work function between analyzer and sample, it is necessary because at the cutoff electrons have a kinetic energy of 0, so they are in the vicinity of the cutoff very sensitive to remaining fields in your chamber, giving them some eV by applying a bias helps.
in brief, the reasons for biasing sample are two-fold:
1. a negative bias is to ensure the local vacuum level of sample is negative than that of the detector,
2. bias can seperate the the "secondary electrons" cut-off of sample from that of the detector. "secondary electrons" cut-off is important when you calculate the work function of your sample.
On the topic of biasing a sample:
A photoelectron analyser with a grounded potential is usually the standard setup and thus transmission function and peak intensities behaves as expected under these comditions.
when adding a bias you add an additional electrostatic lens into your system and hence you change the conditions for the analyser and peak intensity will change. However, as a first approximation the energy position will not be altered, except for the bias energy added.
As a consequence of the added lens one can also expect different behavior of spectra dependent on the sample orientation.
Aa a hemispherical analyser usually have a grounded sample it is usually not possible to measure a zero energy, as there is no element accelerating the electron into the analyser. The bias is then added to create this effect. The bias should then be applied such that the kinetic energy of the electrons are such that the lenses, lens tables, work: each lens table with a pass energy has a given working range.
Just to add to the previous answers, there are two other effects not mentioned here, which can be a large hassle (hard or impossible to be removed in post-experimental data processin): temporal and spatial variations of charging (the latter also known as local charging). They both can skew the shape of the peaks or cause adventitious splittings, potentially to such a level that the spectrum becomes unusable.
Practical note: The temporal instability of charging is often an indicator of omitted ground connection - it is a good practice for students getting acquainted with XPS to let them measure a spectrum once with sample intentionally not grounded and then again with the proper ground. :)
- Badges
- Science topic
- Similar topics
- Philosophy
- Philosophy Of Science
- Reasoning