ARL OPTIM’X WD-XRF Spectrometer
Topics Covered
IntroductionInstrumentationCalibration and Limits of DetectionConclusionAbout Thermo Fisher Scientific-Elemental Analysis
Introduction
The simplest form of glass is the single component fused silica (SiO2).
However it is both difficult to process and expensive. To reduce these
difficulties, some other oxides are added imparting specific properties to the
resultant glass. Most of glasses are composed of about 70 % silica, which is a
glass former, soda as a flux in the form of carbonate and sulfate (about 14 %),
and lime as a stabilizer in the form of limestone (about 10 %). Other types of
oxides like alumina or magnesia improve the physical characteristics of glass,
particularly the resistance to atmospheric conditions.
In-depth coloring is obtained by incorporation of various metallic oxides:
oxides of chromium, iron, manganese or copper.
Instrumentation

An ARL
OPTIM’X XRF spectrometer from Thermo Electron Corporation has been used to
derive limits of detection and precision for the analysis of glasses. The ARL OPTIM’X
is a wavelength dispersive system which
provides superior resolution and
light elements capability. It is fitted with an Air-cooled Rh End-Window Tube
with thin Be window (0.075 mm) and has a maximum power of 50 Watts. Thanks to
close coupling between the X-ray tube anode and the sample the performance of
the ARL OPTIM’X is equivalent to a 200W conventional WD-XRF instrument. The
instrument can be equipped with the unique SmartGonio™, a series of
Multichromators™ or both. Table 1 shows limits of detection for various elements
in soda-lime glasses prepared as pressed powders.

Calibration and Limits of Detection
A series of pressed glass samples have been measured on an ARL OPTIM’X.
Calibration curves have been derived by relating intensities for each oxide (or
element) to concentrations in the standard samples. X-ray fluorescence measures
elements, but the results can be related directly to the oxide forms of these
elements when only one single form is present in the sample. Using the
calibration curves, limits of detection have been derived using the SmartGonio
for the most common oxides/elements found in soda-lime glasses (Table 1).
Table 1: Limits of Detection, using SmartGonio
| Na2O |
Kα1,2 |
AX-06 |
FPC |
100 |
| MgO |
Kα1,2 |
AX-06 |
FPC |
60 |
| Al2O3 |
Kα1,2 |
PET |
FPC |
47 |
| SiO2 |
Kα1,2 |
PET |
FPC |
N.R. |
| P2O5 |
Kα1,2 |
PET |
FPC |
48 |
| SO3 |
Kα1,2 |
PET |
FPC |
23 |
| Cl |
Kα1,2 |
PET |
FPC |
24 |
| K2O |
α1,2 |
LiF 200 |
FPC |
14 |
| CaO |
Kα1,2 |
LiF 200 |
FPC |
12 |
| TiO2 |
Kα1,2 |
LiF 200 |
FPC |
12 |
| Cr2O3 |
Kα1,2 |
LiF 200 |
FPC |
9 |
| MnO |
Kα1,2 |
LiF 200 |
FPC |
9 |
| Fe2O3 |
Kα1,2 |
LiF 200 |
FPC |
9 |
| ZnO |
Kα1,2 |
LiF 200 |
SC |
3.6 |
| SrO |
Kα1,2 |
LiF 200 |
SC |
2.4 |
| ZrO2 |
Kα1,2 |
LiF 200 |
SC |
1.8 |
| BaO |
Lα1 |
LiF 200 |
FPC |
51 |
| PbO |
Lα1 |
LiF 200 |
SC |
9 |
Precision Tests
Precision tests have been carried out by analyzing repeatedly the same
pressed pellet sample for eleven consecutive analyses. Eighteen elements/oxides
are determined using a counting time of 36 seconds per analytical line. The
results are summarized below for two different glass samples (Tables 2 and 3).
In the case when precision should be improved for some elements this counting
time could be increased. Doubling the counting time would improve the precision
by a factor of about 1.4 (square root of 2).
Table 2: Precision Tests - Glass Sample A
| Time (s) |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
| Run 1 |
13.98 |
0.185 |
1.79 |
72.59 |
0.588 |
10.85 |
0.330 |
582 |
579 |
| Run 2 |
13.93 |
0.193 |
1.81 |
72.60 |
0.582 |
10.82 |
0.333 |
640 |
563 |
| Run 3 |
13.97 |
0.177 |
1.80 |
72.64 |
0.588 |
10.82 |
0.330 |
608 |
563 |
| Run 4 |
14.01 |
0.178 |
1.80 |
72.64 |
0.582 |
10.87 |
0.330 |
645 |
581 |
| Run 5 |
13.95 |
0.182 |
1.80 |
72.60 |
0.588 |
10.83 |
0.329 |
576 |
564 |
| Run 6 |
13.94 |
0.177 |
1.81 |
72.61 |
0.589 |
10.82 |
0.329 |
573 |
569 |
| Run 7 |
13.86 |
0.185 |
1.80 |
72.64 |
0.588 |
10.83 |
0.330 |
658 |
569 |
| Run 8 |
13.92 |
0.186 |
1.81 |
72.59 |
0.585 |
10.84 |
0.331 |
652 |
566 |
| Run 9 |
13.94 |
0.184 |
1.81 |
72.63 |
0.591 |
10.82 |
0.334 |
651 |
579 |
| Run 10 |
13.98 |
0.183 |
1.80 |
72.63 |
0.586 |
10.87 |
0.332 |
617 |
526 |
| Run 11 |
13.95 |
0.188 |
1.78 |
72.62 |
0.588 |
10.83 |
0.330 |
619 |
561 |
| Avg. |
13.95 |
0.183 |
1.80 |
72.62 |
0.587 |
10.84 |
0.331 |
620 |
565 |
| Std.Dev. |
0.04 |
0.005 |
0.01 |
0.02 |
0.003 |
0.02 |
0.0015 |
32 |
15 |
| Time (s) |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
| Run 1 |
166.9 |
113.4 |
93.6 |
48.4 |
101.6 |
127.4 |
209.1 |
454.3 |
228.2 |
| Run 2 |
146.3 |
129.5 |
91.4 |
44.7 |
101.5 |
124.8 |
204.6 |
392.6 |
218.9 |
| Run 3 |
193.3 |
111.2 |
91.1 |
43.8 |
95.9 |
127.1 |
207.0 |
361.8 |
197.8 |
| Run 4 |
199.2 |
104.6 |
96.2 |
29.9 |
103.8 |
127.0 |
205.4 |
375.7 |
234.6 |
| Run 5 |
158.1 |
111.8 |
94.6 |
41.8 |
103.7 |
122.7 |
204.7 |
385.2 |
228.5 |
| Run 6 |
171.3 |
107.9 |
85.2 |
49.5 |
95.0 |
126.3 |
203.8 |
355.4 |
194.8 |
| Run 7 |
203.6 |
113.4 |
88.9 |
40.3 |
96.0 |
125.0 |
205.4 |
434.1 |
234.4 |
| Run 8 |
190.4 |
135.6 |
94.5 |
44.7 |
96.8 |
125.9 |
203.4 |
315.1 |
207.3 |
| Run 9 |
150.7 |
110.1 |
88.6 |
43.1 |
114.9 |
127.1 |
206.6 |
401.2 |
220.8 |
| Run 10 |
255.0 |
104.0 |
83.6 |
41.9 |
99.2 |
125.8 |
206.2 |
402.2 |
214.6 |
| Run 11 |
218.3 |
97.9 |
80.6 |
38.9 |
97.7 |
126.8 |
203.1 |
429.8 |
197.8 |
| Avg. |
186.6 |
112.7 |
89.9 |
42.4 |
100.5 |
126.0 |
205.4 |
391.6 |
216.1 |
| Std.Dev. |
32.5 |
11 |
5 |
5.2 |
5.7 |
1.4 |
1.8 |
39.7 |
15 |
Table 3: Precision Tests - Glass Sample B
| Time (s) |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
| Run 1 |
13.35 |
0.180 |
1.67 |
73.07 |
0.556 |
10.77 |
773.3 |
0.177 |
556.0 |
| Run 2 |
13.33 |
0.180 |
1.68 |
73.08 |
0.564 |
10.76 |
757.9 |
0.181 |
568.0 |
| Run 3 |
13.28 |
0.186 |
1.67 |
73.08 |
0.554 |
10.81 |
789.6 |
0.180 |
555.1 |
| Run 4 |
13.28 |
0.185 |
1.66 |
73.11 |
0.559 |
10.83 |
768.2 |
0.186 |
587.2 |
| Run 5 |
13.35 |
0.181 |
1.67 |
73.05 |
0.554 |
10.79 |
763.9 |
0.181 |
594.7 |
| Run 6 |
13.32 |
0.172 |
1.67 |
73.11 |
0.566 |
10.80 |
767.3 |
0.186 |
541.4 |
| Run 7 |
13.33 |
0.185 |
1.67 |
73.06 |
0.554 |
10.79 |
758.9 |
0.180 |
570.3 |
| Run 8 |
13.26 |
0.185 |
1.69 |
73.04 |
0.555 |
10.78 |
771.7 |
0.185 |
565.2 |
| Run 9 |
13.33 |
0.180 |
1.64 |
73.11 |
0.561 |
10.82 |
775.7 |
0.183 |
553.7 |
| Run 10 |
13.30 |
0.193 |
1.68 |
73.08 |
0.556 |
10.80 |
764.1 |
0.188 |
572.6 |
| Run 11 |
13.31 |
0.184 |
1.66 |
73.06 |
0.561 |
10.78 |
785.3 |
0.186 |
566.0 |
| Avg. |
13.31 |
0.183 |
1.67 |
73.08 |
0.558 |
10.80 |
770.5 |
0.183 |
566.4 |
| Std.Dev. |
0.03 |
0.01 |
0.01 |
0.03 |
0.004 |
0.02 |
10 |
0.003 |
15 |
| Time (s) |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
36 |
| Run 1 |
200.6 |
100.7 |
63.4 |
9.3 |
118.2 |
122.4 |
227.2 |
883.9 |
895.6 |
| Run 2 |
159.5 |
111.2 |
65.9 |
3.9 |
112.0 |
122.2 |
226.1 |
960.6 |
914.5 |
| Run 3 |
193.3 |
115.7 |
64.7 |
18.2 |
115.6 |
119.3 |
225.3 |
925.4 |
911.5 |
| Run 4 |
156.6 |
103.5 |
74.7 |
8.7 |
105.6 |
126.5 |
225.1 |
891.3 |
900.0 |
| Run 5 |
187.4 |
97.9 |
63.6 |
12.3 |
106.8 |
126.3 |
226.2 |
948.8 |
904.8 |
| Run 6 |
183.0 |
114.6 |
59.5 |
14.7 |
115.2 |
125.5 |
226.3 |
960.5 |
904.6 |
| Run 7 |
193.3 |
101.8 |
67.9 |
6.1 |
113.5 |
124.9 |
226.9 |
960.6 |
910.1 |
| Run 8 |
191.8 |
95.2 |
66.9 |
9.3 |
101.6 |
124.7 |
227.4 |
918.0 |
919.7 |
| Run 9 |
219.7 |
113.4 |
61.5 |
9.3 |
109.6 |
124.6 |
222.9 |
916.9 |
913.4 |
| Run 10 |
243.2 |
101.3 |
64.4 |
1.0 |
103.4 |
123.0 |
226.0 |
980.8 |
912.7 |
| Run 11 |
191.8 |
109.0 |
70.8 |
13.8 |
106.9 |
122.1 |
227.7 |
950.9 |
875.6 |
| Avg. |
192.8 |
105.8 |
65.8 |
9.7 |
109.9 |
123.8 |
226.1 |
936.2 |
905.7 |
| Std.Dev. |
24.2 |
7.2 |
4.3 |
4.9 |
5.4 |
2.2 |
1.4 |
31.2 |
12 |
Conclusion
All limits of detection obtained show that the ARL OPTIM’X can deliver
adequate analysis results, notably for bottle glass application. Repeatability
of analysis is excellent for major and minor elements even for Na2O and MgO.
Longer counting time may be used in case elements present below 100 ppm need to
be controlled precisely. These results show that the ARL OPTIM’X spectrometer is
well suited to produce precision results for the determination of the main
oxides and the coloring agents in glasses.
About Thermo Fisher Scientific-Elemental Analysis
For over 75 years, Thermo
Fisher Scientific has been a worldwide supplier of spectrochemical
instrumentation to major industries including steel, transportation, cement,
construction, food, pharmaceuticals, chemicals, academic research, petroleum and
electronics. They offer unsurpassed capabilities in the areas of optical
emission (OE), X-ray fluorescence (XRF), X-ray diffraction (XRD) and automation
of spectrometers.

This information has been sourced, reviewed and adapted from materials
provided by Thermo Fisher Scientific- Elemental Analysis.
For more information on this source, please visit Thermo Fisher Scientific-
Elemental Analysis.