5 10 15 20 25 30 35 40 1
10 100 1000 10000
Rh Kβ Rayleigh
Rh Kβ Compton
Rh KαRayleigh
Rh Kα Compton
PTFE_ 40kV_150µA_2500s_ filtered
Simulated/Experimental Counts
Energy (keV)
Experimental Simulated
Figure 35. PTFE experimental over simulated at 40kV, filtered excitation
5 10 15 20 25 30 35
0.1 1 10
Experimetal/Simulated Ratio
Energy (keV)
PTFE ratio_40kV_150µA_2500s_filtered
Figure 36. PTFE experimental over simulated ratio at 40kV, filtered excitation
4.2 Comparison between pXRF measurements and MC
5 10 15 20 25 30 35 100
101 102 103 104
Ag Kβ Ag Kα
Rh Kα Au M
Ag L
Cu Kβ Cu Kα
Au Lγ Au Lβ
ABKMF_40kV_100µΑ_3000s_filtered
Experimental/Simulated Counts
Energy (keV)
experimental simulated Au Lα
Figure 37. ABKMF experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104
Experimental/Simulated Counts
Cu Kα
Ag L Au M
Ag Kβ Ag Kα Rh Kα Au Lγ Au Lβ Au Lα
Cu Kβ
ABSBL_40kV_100µΑ_3000s_filtered
Energy (keV)
experimental simulated
Figure 38. ABSBL experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104
Experimental/Simulated Counts
Ag L Au M
Cu Kβ
Ag Kβ Ag Kα Rh Kα Au Lγ Au Lβ Au Lα Cu Kα
ABLLI_40kV_100µΑ_3000s_filtered
Energy (keV)
Experimental Simulated
Figure 39. ABLLI experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104
Experimental/Simulated Counts
Cu Kβ
Ag Kβ Ag Kα Rh Kα Au Lγ Au Lβ Au Lα
Cu Kα Au M
Ag L
ABQAQ_40kV_100µΑ_3000s_filtered
Energy (keV)
experimental simulated
Figure 40. ABQAQ experimental over simulated at 40kV.
In the following graph, the % difference between experimental and simulated net peak area intensities of the detected characteristic X ray lines are presented, as they were extracted by means of the PyMca Toolkit.
48
-20 0 20
% (exp-sim)/exp
KMF Au: 58.58 ± 0.90 Ag: 29.37 ± 0.43 Cu: 12.05 ± 0.18 LLI
Au:74.96 ± 1.11 Ag: 5.04 ± 0.15 Cu: 20.0 ± 0.3 SBL
Au: 89.99 ± 1.35 Ag: 5.04 ± 0.15 Cu: 4.97 ± 0.15 QAQ
Au: 95.04 ± 1.42 Ag: 2.46 ± 0.15 Cu: 2.50 ± 0.15
Cu-Kα Ag-Kα Au-Lα Au-Mα
Figure 41. The % difference between experimental and simulated net peak area in- tensities of the detected characteristic X-ray lines of gold alloys.
It can be concluded that the simulation overestimates systematically all elements’
major peak intensities between ~5-10%, however, the Au-M intensity is experimentally overestimated by ~10-15%.
4.2.2 Silver alloys
The silver alloys that was used are CNR-91, CNR-92, CNR-141 and CNR-152.
The chemical composition of the silver alloys is reported in detail at the Table II of the chapter 3.2 Reference materials.
49
5 10 15 20 25 30 35 100
101 102 103 104 105
Experimental/Simulated Counts Ag Kβ
Ag Kα
Rh Kβ Rh Kα Pb Lγ
Pb Lβ Pb Lα
Cu Kβ Cu Kα Ag L
CNR91_40kV_100µΑ_3000s_filtered
Energy (keV) experimental
simulated
Figure 42. CNR91 experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104 105
Experimental/Simulated Counts
Rh Kβ Ag Kβ Ag Kα
Rh Kα Pb Lγ
Pb Lβ Cu Kβ
Pb Lα Cu Kα Ag L
CNR92_40kV_100µΑ_3000s_filtered
Energy (keV) Experimental
Simulated
Figure 43. CNR92 experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104 105
Experimental/Simulated Counts
Ag L Cu Kβ Cu Kα
Rh Kβ Rh Kα
Ag Kα Ag Kβ CNR141_40kV_100µΑ_3000s_filtered
Energy (keV) Experimental
Simulated
Figure 44. CNR141 experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104 105
Experimental/Simulated Counts Ag Kβ
Ag Kα
Rh Kβ Rh Kα Cu Kβ
Cu Kα Ag L
CNR141_40kV_100µΑ_3000s_filtered
Energy (keV) Experimental
Simulated
Figure 45. CNR152 experimental over simulated at 40kV.
In the following graph, the % difference between experimental and simulated net peak area intensities of the detected characteristic X ray lines are presented, as they were extracted by means of the PyMca Toolkit.
50
-20 -15 -10 -5 0 5 10 15 20 25 30
% (exp-sim)/exp
Cu-Kα Pb-Lα Ag-Ka Ag-La
CNR-91
Ag: 97 ± 1 Cu: 1.5 ± 0.1 Pb: 1.5 ± 0.1
CNR-92
Ag: 92 ± 1 Cu: 6.5 ± 0.3 Pb: 1.5 ± 0.1
CNR-141
Ag: 92.5 ± 1 Cu: 7.5 ± 0.4
CNR-152
Ag: 96.5 ± 1 Cu: 3.5 ± 0.1
Figure 46. The % difference between experimental and simulated net peak area inten- sities of the detected characteristic X-ray lines of silver alloys.
In silver alloys, the Ag is the predominant component and the Ag-Kα lines are accurately simulated with differences below ~6% between the experimental and simu- lated intensities. The Cu-Kα lines are mostly overestimated. In CNR-91, where the Cu concentration is 1.5% the Cu-Kα intensity is overestimated by ~21%, whereas for other concentrations the observed differences are less than ~5%. The deviation observed for the Cu-Kα intensity for CNR-91 reference silver alloy could be attributed to a possible inhomogeneity of the Cu distribution within the alloy. Finally, the Ag-Lα lines are un- derestimated by the simulation for about 20%.
4.2.3 Copper alloys
As copper alloys was used the BCRA, BCRB, BCRD and BCRE. The chemical composition of the copper alloys is reported in detail at the Table I of the chapter 3.2 Reference materials.
51
5 10 15 20 25 30 35 100
101 102 103 104 105
Experimental/Simulated Counts
Sn Kβ Sn Kα
Rh Kβ Rh Kα
As Kβ Pb Lγ Pb Lβ
Pb Lα As Kα
Zn Kβ Cu Kβ Zn Kα Cu Kα
Sn L
BCRA_40kV_100µΑ_3000s_filtered
Energy (keV)
Experimental Simulated
Figure 47. BCRA experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104 105
As Kβ Pb Lγ Pb Lβ Pb Lα As Kα
Sn Kβ Sn Kα Rh Kβ Rh Kα Zn Kβ
Cu Kβ Zn Kα Cu Kα
Sn L
BCRB_40kV_100µΑ_filtered
Simulated/Experimental Counts
Energy (keV)
Experimental Simulated
Figure 48. BCRB experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104 105
Experimental/Simulated Counts
As Kβ
Sn Kβ Sn Kα
Rh Kβ Rh Kα Pb Lγ Pb Lβ Pb Lα As Kα
Zn Kβ Cu Kβ Cu Kα
Sn L
BCRD_40kV_100µΑ_3000s_filtered
Energy (keV)
Experimental Simulated
Figure 49. BCRD experimental over simulated at 40kV.
5 10 15 20 25 30 35
100 101 102 103 104 105
Experimental/Simulated Counts
Pb Lβ Pb Lα As Kα
Zn Kβ Cu Kβ Cu Kα
Sn L
As Kβ
Rh Kβ Rh Kα
Sn Kα Sn Kβ BCRE_40kV_100µΑ_3000s_filtered
Energy (keV)
Experimental Simulated
Figure 50. BCRE experimental over simulated at 40kV.
In the following graph, the % difference between experimental and simulated net peak area intensities of the detected characteristic X ray lines are presented, as they were extracted by means of the PyMca Toolkit.
52
-50 -40 -30 -20 -10 0 10 20 30 40 50 60
% (exp-sim)/exp
Cu-Kα Zn-Kβ As-Kβ Pb-Lβ Sn-Kα Sn-La
BCR-A Cu: 78.726 As: 0.194 ± 0.010 Pb: 7.9 ± 0.7 Sn: 7.16 ± 0.21 Zn: 6.02 ± 0.22
BCR-B Cu: 82.651 As: 0.099 ± 0.010 Pb: 0.39 ± 0.03 Sn: 2.06 ± 0.07 Zn: 14.8 ± 0.5
BCR-D Cu: 80.267 As: 0.285 ± 0.022 Pb: 9.2 ± 1.7 Sn: 10.1 ± 0.8 Zn: 0.148 ± 0.024
BCR-E Cu: 92.445 As: 0.194 ± 0.020 Pb: 0.204 ± 0.018 Sn: 7.0 ± 0.6 Zn: 0.157 ± 0.025
Figure 51. The % difference between experimental and simulated net peak area in- tensities of the detected characteristic X-ray lines of copper alloys.
The elements Zn and As were analyzed and compared by their respective Kβ lines in order to avoid systematic errors in the analysis of their Kα peak areas due to spectral interferences (Cu-Kβ with Zn-Kα and As-Kα with Pb-Lα). In the cases where the con- centration of both elements is below 1% (close to detection limits) the experimental measurements significantly overestimated the simulated counts by ~50-54% for Zn and
~26% for As. The Sn-Kα lines are overestimated by the simulation by ~9.7%. The dif- ferences observed regarding the intensity of Cu-Kα line is generally below ~10-14%.