A detailed study about the band bending phenomena and its measuring techniques included photoelectron spectroscopy could be found in the following review;
Chem. Rev., 2012, 112 (10), pp 5520–5551
http://pubs.acs.org/doi/abs/10.1021/cr3000626
Furthermore, a comprehensive textbook about the space charge layer with various material aspects could also be found in the following:
Böer, Karl W. Vol. 37. Dordrecht-London-New York,, USA: Springer, 2010.
The first question is if you want to know that your sample has an as prepared band bending or you change the sample and by that you may obtain a band Bendig as aconsequence. In some cases you may observe a shift of your UPS spectra if you Illuminate the sample with additional bias light (hn>Egap). This due tu a surface photovoltage, which might drive your sample for high bias light intensities into saturation at flat band. So the band bending is the difference of the shift between illuminated and dark. The original band Bendig might be higher as your He source also indices a surfacephoto Vortage depending on the doping and the temperature of the sample. The change in band bending determinier from the valence band offset might not always be equal to the change in workfunction as a change in surface dipole may be included. Dipoles are a consequence of reorganisation or adsorbates (change of the Elektron affinity, just 1st layer effect) whereas the band bending is a consequence of reorganisation of the free charges in the surface which might range µm into your crystal depending on the doping, and you observe with UPS only the topmost surface layer. You might see a difference in the position of shallow core level in HeII and XPS due to the different escape depth. In addition conventional unmonochromatised XPS is flooding you sample with additional electrons stemming from the Bremsstrahlung absorbed in the Al window of your X-ray tube (observe a peak shift if you retract the source and reduce the HV of the tube).
For in situ (UHV) modified samples ( adsorbates, Schottky barriers, etc. ) you can estimate the additional band Bendig by comparing the spectra directly.