Contents of calcium fluoride in steel making slags and fluorspars were analyzed by determining fluorine using non-destructive 14 MeV neutron activation analysis.
A dual sample irradiation assembly was used to irradiate an analytical sample and an external reference sample of a definite amount of calcium fluoride at the same time. Amounts of fluorine were determined from the ratio of counts of photo peak areas of 0.197 MeV energy of
19O calculated by the TPA method in each spectrum of the analytical and the external reference sample.
Decay curves of gross counts, bakground counts and photo peak area counts were analyzed by the least square method using
19O,
16N,
18F and
28Al as the factors. The results confirmed that
19O was interfered by only
16N when the counts of photo peak were used for calculation.
Experimental conditions obtained from the investigation of the optimum conditions of activation analysis and those obtained from the experimental results of the ratio of photo peak/background were in good agreement.
For the actual analytical conditions, however, the sample decay duration longer than the calculated optimum was chosen to decay
16N. Analytical conditions of 30 s for the sample irradiation, 60 s for decay, and 15 s for the measurement were used. Even if the weight of oxygen present in steel making slags is 30 times as much as fluorine, the contribution of
16N on counts of
19O photo peak area was suppressed down to about 4% of the counts using this analytical condition.
There was a significant difference between the gradients of the calibration curves made by synthetic steel making slags containing various materials and by calcium fluoride standards which did not contain any other material. Therefore, it is necessary to use the calibration curves for both the group of steel making slags and the group of calucium fluoride and fluorspars.
The results obtained by this method were compared with those by the pyro-hydrolytic method showing good agreement in the entire range of fluorine concentrations.
View full abstract