I am absolutely no expert in XPS, but I would assume a higher stability of the oxidation state of +3 for iron in the Fe2O3 rather than in the zinc ferrite and hence maybe narrower peaks. Only an idea. Why don't you prepare well-defined references and compare them to your data of the unknown sample?

i think the oxidation state of iron in hematite and zinc ferrite +3 !, i made a comparison with more than a reference, but no result.

i found paper suggest that the mixture takes the form of hematite and zinc oxide, also there is another paper proved the mixture was  a composite of hematite and zinc ferrite!! using approximately the same XPS data !!

Hm. Maybe, XPS is the wrong method then. If your sample is polycrystalline, I would suggest XRD for phase analysis.

Check these links .........

http://www.xray.cz/ms/bul99-2/sepelak.pdf

http://journal.chemistrycentral.com/content/8/1/40

http://connection.ebscohost.com/c/articles/14910482/microstructural-evolution-fe2o3-znfe2o4-during-sonochemical-synthesis-zinc-ferrite

http://espace.library.uq.edu.au/view/UQ:163225

http://www.sciencedirect.com/science/article/pii/S1293255810000592

Also check the attachment , Each element [Fe,O,Zn ] has specific binding energy in the XPS spectra , especially for analysing O [ O1s ] you can deconvolute the XPS spectra in the range 520-550 eV.

zinc element in the form of oxide has a binding energy in the XPS spectra resembles or doesnot differ from that of ferrite form !! also calcium !!