In real samples, many concomitant are likely to react competitively with aluminum and fluorine, e.g. ametals with aluminum; metals with fluorine causing the concentration of aluminum mono fluoride (g)
(i.e. the sensitivity of fluorine) to decrease compared to matrix-free solutions.
This interference depends on the matrix concomitant and their concentrations which are mostly unknown.
In this case,
since the use of linear calibration technique with aqueous standards may be risky,
standard addition technique is more appropriate to eliminate any potential interference for quantification of fluorine in all real samples.
Therefore, if the results found by the two techniques are not in agreement,
those found by standard addition technique is more reliable.