Dear Prof Fred, Thank you for your reply. I collected the data in fluorescence mode. We have indeed anticipated the self-absorption effect and analysed the same standards mixtures that were diluted so that the S content is less than 10%. The spectrum of disulphide, however, showed different features as those found in undiluted one (disulphide-10 vs disulphide), as you can see in the ATHENA file I attached here (the diluted samples were labelled with -10 after their name) which eventually alters the samples' spectra. I have also done the LCF on these diluted samples and the fitting results are not as what it should be: S2: ~20% disulphide and 80% thiophene. LCF reported 1.6% disulphide and 98.4% thiophene. S3: ~60% disulphide and 40% thiophene. LCH reported 11.2% disulphide and 88.8% thiophene. S4: ~80% disulphide and 20% thiophene. LCH reported 82.4% disulphide and 17.6% thiophene. (This surprisingly gave a better fitting than the undiluted one, as I mentioned in the previous mail). I noticed that the intensity of samples S2-10 and S3-10 is higher than both of the standard spectra (thiophene-10 and disulphide-10) which I presumed to be the main cause for the incorrect fitting results obtained for these samples. Do you have any suggestions on how to tackle this issue? You most probably right that the samples are not well-mixed (the thiophene we analyzed is dibenzothiophene and it is in powder form). Also we only use a small amount of it for the analysis. Regards, Niken On 8/02/2012 6:24 PM, fred.mosselmans@diamond.ac.uk wrote:
Dear Niken, To my untutored eye your normalisation and fitting look OK so thus it comes down to the data or whether the treatment is valid: Did you collect the data in fluorescence mode? Two possible problems occur to me. As thiophene is a liquid at RT unless you did the experiment cryogenically I think 2 is unlikely.
1) Most probably there is an inherent self-absorption problem with S XANES LCF which was written about by G. Almkvist, K. Boye and I. Persson J. Synchrotron Rad. (2010). 17, 683-688 [ doi:10.1107/S0909049510022946 ] K-edge XANES analysis of sulfur compounds: an investigation of the relative intensities using internal calibration. The self-absorption seen can affect the results of simple LCF data quite alot.
2) the samples were not well enough mixed, hence the data collected does not reflect the real mixture but a fraction that is more like the values you obtain by the LCF. S K-edge measurements don't sample much sample (sorry for that) so mixing solid standards is not necessarily trivial but for liquids/solutions this should not apply.
Cheers Fred
Prof. J F W Mosselmans Principal Beamline Scientist I18 Diamond Light Source Diamond House Harwell Campus Didcot Oxon OX11 ODE UK
T 00 44 1235 778568 M 00 44 7785510211 E fred.mosselmans@diamond.ac.uk F 00 44 1235 778448
Never mind the W it's the tax that counts