Hi  Leandro
Thank you for your helpful advices. The title should be change to "Please give me some advices" , and I am sorry for this error. 
 
The attached .prj file can be opened by Athena rightly by changing the download file's extend name .bin  to .prj, if someone else want to join our discuss.
 
Then go to our topics:
 
The QDs samples were prepared by MBE and XAFS measurements, as you guessed, were performed at a synchrotron facility in fluorescence mode. Since the Si/Ge intermixing happened frequently during growth, there is certainly Si in so-called Ge quantum dots(QDs, or nanocrystal). However, now I am very caring for if the GeOx in the QDs. The XAFS spectra show that there may be GeOx in the Ge QDs, but since the QDs were covered in situ by a 15nm Si cap layer the oxidizing process of QDs by air should not happen. This inconsistency between XAFS and experiments is the real reason for that I want to extract the truth from strict softwae analysis.
 
1, I have the bulk Ge and diluted SiGe standard spectra, but they do not resemble with these spectra as you mentioned. I am sure that the QDs is a combination of Ge and Si elements, and am not clear if the O exists. The growth temperature is beyond 500oC, and annealing was performed from 550-800oC(all temperature uncalibrated with +/- 50oC ). The amorphous Ge seems to be impossible, but can not exclude it. In situ RHEED also can not tell me if the deposited Ge is crystallizing.  
 
2, If I truncate the k to 10 A-1, the fit result is accepted or not?
 
 
Kefan