Unknown concentrations are determined by inserting K values into the equation for b in the place of concentrations and finding initial concentrations. These new concentrations are reinserted to find new bs and the iteration is repeated until convergence in achieved. Two or three iterations usually result in values very close to the theoretical composition. In this example, the "unknown" will be a synthetic grossularite garnet, Ca3Al2Si3O12. The standards that were used are:
|
|
Oxide Weight Fraction |
||
|
|
Al2O3 |
SiO2 |
CaO |
|
Corundum |
1.000 |
|
|
|
Quartz |
|
1.000 |
|
|
Wollastonite |
|
0.517 |
0.483 |
The appropriate values of a (Albee and Ray 1970) are:
|
|
a |
||
|
|
Al2O3 |
SiO2 |
CaO |
|
Al |
1.00 |
1.05 |
1.18 |
|
Si |
1.54 |
1.00 |
1.07 |
|
Ca |
1.08 |
1.10 |
1.00 |
Note that all a >= 1, indicating that absorption dominates.
Calculating the bs for the standards yields:
|
Corundum |
b-Al = 1.00 x 1.000 = 1.000 |
|
Quartz |
b-Si = 1.00 x 1.000 = 1.000 |
|
Wollastonite |
b-Ca = 1.10 x 0.517 + 1.00 x 0.483 = 1.052 |
The measured X-ray counts corrected for background and deadtime were:
|
|
Unknown |
Standard |
|
Al |
29146 |
140126 |
|
Si |
19764 |
56471 |
|
Ca |
26971 |
34358 |
First, we need to make an initial approximation of the composition of the unknown. To do this we calculate initial K factors for the unknown:

So,
|
K-Al2O3 |
(29146 / 140126) x (1.000 / 1.000) = 0.2080 |
|
K-SiO2 |
(19764 / 56471) x (1.000 / 1.000) = 0.3500 |
|
K-CaO |
(26971 / 34358) x (0.483 / 1.052) = 0.3604 |
These K-factors total to 0.9184. Next, we make an approximation of the bs using these K values and the as above. For this initial approximation, the K values are used as estimates of the concentrations:
|
Al2O3 |
(1.00 x 0.2080 + 1.05 x 0.3500 + 1.18 x 0.3604) / 0.9184 = 1.090 |
|
SiO2 |
(1.54 x 0.2080 + 1.00 x 0.3500 + 1.07 x 0.3604) / 0.9184 = 1.150 |
|
CaO |
(1.08 x 0.2080) + 1.10 x 0.3500 + 1.00 x 0.3604 / 0.9184 = 1.056 |
Now we can use these bs and the K values to calculate a better approximation to the actual concentrations:
|
Al2O3 |
1.090 x 0.2080 = 0.2267 |
|
SiO2 |
1.150 x 0.3500 = 0.4025 |
|
CaO |
1.056 x 0.3604 = 0.3806 |
The new sum is 1.0098. This total can be improved by iteration. We refine the estimates of the bs using the weights just calculated:
|
Al2O3 |
(1.00 x 0.2267 + 1.05 x 0.4025 + 1.18 x 0.3806) / 1.0098 = 1.088 |
|
SiO2 |
(1.54 x 0.2267 + 1.00 x 0.4025 + 1.07 x 0.3806) / 1.0098 = 1.148 |
|
CaO |
(1.08 x 0.2267 + 1.10 x 0.4025 + 1.00 x 0.3806) / 1.0098 = 1.058 |
And use the refined beta-factors to get new concentrations:
|
Al2O3 |
1.088 x 0.2080 = 0.2263 |
|
SiO2 |
1.148 x 0.3500 = 0.4018 |
|
CaO |
1.058 x 0.3604 = 0.3816 |
The new total is 1.0094. Continued iteration would improve this number a little. Compare the values for the ideal composition:
|
|
Weight |
|
Al2O3 |
0.2263 |
|
SiO2 |
0.4002 |
|
CaO |
0.3735 |
In Bence-Albee corrections, standards of compositions similar to the unknowns are often used to minimize Z effects. In addition, because the effect of fluorescence in silicates is very small, only absorption need be considered. Bence-Albee calculations require very little computing time or memory, but with more powerful on-line computer this is no longer a significant advantage and the more complicated ZAF-type corrections can be made.
Back: 5.3. Bence-Albee Matrix Corrections | Next: 5.4. ZAF Matrix Corrections | Home: Course Overview
Copyright 1997-2003, James H. Wittke
Last update: 01/18/2006 01:47 PM.