220 likes | 343 Views
NAA using Kayzero: from peak areas to element concentration. R. van Sluijs , k 0 -ware, Heerlen, the Netherlands D. Bossus, DSM Geleen, the Netherlands F. De Corte, RUG, Ghent, Belgium. Element concentration, ρ. Peak area, A. k 0 -method. Problem? No problem!?.
E N D
NAA using Kayzero:from peak areas to element concentration R. van Sluijs, k0-ware, Heerlen, the Netherlands D. Bossus, DSM Geleen, the Netherlands F. De Corte, RUG, Ghent, Belgium NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Element concentration, ρ Peak area, A k0-method Problem? No problem!? Simple? Perhaps not, but very well documented! nowadays there are many programs to help! NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
No problem,but ..... (1) • How to get the correct peak area? • Use a perfect gamma spectrometry system and keep it in the best possible condition! • Use the perfect gamma spectrum deconvolution program! However: who has this perfect set-up? NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
No problem, but......... (2) • How to select the analytical line • K0-library has data for 68 elements, 144 nuclides and more than 700 gamma-lines • For a panoramic analysis you use preferably 2 irradiations (short and long) each followed by 2-3 measurements • A single element will give data for several nuclides and probably tens of gamma lines. • All we want one single list: Containing the all elements with their concentrations and/or limits of detection. NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
DSM’s/Kayzero’s method User Input Automatic Optimise the peak table! • Peak Table: Area’s and Energies from gamma spectrum deconvolution program Calculation • Concentrations for all lines in the k0- library Evaluation • Identify the nuclides in the sample • Correct for gamma interferences • Calculate averages per nuclide and per element • Average over all measurements Refine the results NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Optimise the peak table! Peak Areas: • Broadened peaks (due to count rate effects, bad pile-up correction, etc..) can be split by the gammaspectrum deconvolution program into two separate peaks. • Sum the false/artificial peaks for a correct result Peak locations: • Linear energy calibration, non-linear is often better • Re-do the energy calibration • Hypermet shifts data 1 channel to the left • Shift back NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Measurement Input screen NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Peak summation screen NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Re-do energy calibration NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
k0-calculations Calculate concentrations for all library lines • Including Corrections for: • background, • dead-time, • gamma-absorption, • burn-up, • neutron self absorption, • True-coincidence correction • No matching peak: the LD concentration is calculated using the spectrum base line. • No result if calculated concentration >100% NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Evaluation of the results:Check the presence of a nuclide NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Evaluation of the results:Check the presence of an element NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Evaluation of the resultsCorrect for gamma interferences N=number of interfering nuclides on line E NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Evaluation of all resultsCalculate averages and standard deviations • over the lines of a nuclide • weighing factor wi= 1/σ2i • σobserved should be approx. equal to σexpected NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Refine the resultsselect the analytical lines Select by hand: • gamma’s to be used • correct for threshold, fission, second-order reaction interferences, • blank correction, Alarm for high epi-thermal self shielding NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Ba-131: large observed SD NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Deselect Ba-131 line NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Fission Correction for La-140 NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Fission Correction NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Blank subtraction NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
The goal of all our work:a (digital) list of concentrations 51 elements reported: 30 with concentration, 21 only detection limit 51 elements reported: - 30 concentrations - 21 only detection limit Final result NAMLS-8, Rio de Janeiro, April 17th-22nd 2005
Conclusion • k0-method (DeCorte/Simonits) is a very complete method, all aspects of INAA are taken into account, elaborate calculations are needed but this is no problem nowadays • The method presented to handle k0-data is based on almost 20 years of experience, analysing all kinds of samples • Evaluation of the data of a panoramic elemental analysis of an unknown sample takes less than half an hour • Although very much automated: the analyst is in control • The competition in our lab is though, NAA is the most expensive technique so quality and reliability is a prerequisite NAMLS-8, Rio de Janeiro, April 17th-22nd 2005