OK, there seem to be multiple "oddities" here:

1) If you calculated your binding energy correctly, the fermi energy is 0 eV, not 2.5 eV. The 2.5 eV value is the cutoff that you would use to calculate the ionization potential (although it looks more like 2.2 to me). If the Fermi energy actually was at 2.5 eV your substance would be metallic since it would have states up to the Fermi level.

2) What kind of radiation source are you using? A standard He-Ka source produces 21.2 eV, not 11. With that value, your spectrum would seem rather "civilized" and your work function would be 4.7 eV which is a quite usual semiconductor work function value.

3) Are you sure about the 15.5 eV value? It looks more like 15.2 or 15.3 to me and that's definitely a precision that should be applied.

3) Regarding the 16 eV peak: do you use metal clamps on your sample holder? Maybe you recorded a little of that.

Hello Jürgen,

Thanks for your answer. I have checked with the engineer and you were right, the source is indeed 21.2 eV. And the 16 eV peak seemed to be due to the roughness of the sample surface. However, could you walk me through your calculation? I'm not very familiar with the technique, so I probably got something wrong, but according to the formula work function=hv-(cutoff-EF), when hv=21.2, cutoff=15.2, EF=0, the work function is 6 eV. Is there something I've missed in the calculation?

No, you didn't miss anything and the formula is correct. The 4.7 eV was a typo (sorry for the confusion), it should have been 5.7 building on the 15.5 eV value you wrote in the question. With the more accurate 15.2 eV value it is indeed 6.0 eV.

That's a relief, thanks very much for your help!