Comparison of different methodologies for the ⁹⁰Sr determination in environmental samples

The paper describes different isolation/separation and detection procedures for ⁹⁰Sr determination in the environmental samples which are routinely used in Laboratories A and B. In this context, four different methods for strontium isolation and two methods for detection were tested and compared by ⁹⁰Sr determination in proficiency test samples (water, soil, vegetation) and animal bone samples. The chromatographic isolation of Sr on Sr resin, AnaLig^®Sr01 resin gel, strong base anion exchange resins in nitrate form and combination of strong base anion exchange and Sr resin were used for the examination of the impact of sample matrix constituents on efficiency of strontium isolation (chemical yield), while Cherenkov counting of ⁹⁰Y and counting of ⁹⁰Sr(⁹⁰Y) on proportional counter were used for the quantitative ⁹⁰Sr determination. The chemical yields obtained with different isolation methods were compared with the emphasis on its influence on reliability of the ⁹⁰Sr determination in different kinds of samples. The results show that the efficiency of strontium isolation depends on type of sample and separation methodology. The strontium yield on Sr resin column decreases with the increase of Sr, Ca and Na concentration. In the presence of 1 g of Ca and 1 g of Na, the yield of 85% was obtained for 5 mg of Sr carrier and dropped below 50% with further increase of Sr and other elements. However, the yield can be increased to 75% if Na and part of Ca are separated from Sr on the anion exchange column with alcoholic solution of nitric acid and by final separation of Ca from Sr on the Sr resin column. In the presence of large amounts of Ca, Na and other elements, isolation efficiency on the Sr resin column significantly decreases in comparison with other methods. The average yield for isolation from vegetation samples on the Sr resin column is only 21%. For the soil samples the highest average yield (78%) is obtained for the isolation in the combination of anion exchange and Sr resin columns. For the isolation from bone samples the average yields over 80% are on AnaLig^®Sr01 and anion exchange resins columns, while Sr resin was not used for separation due to high content of Ca in samples. The results of the ⁹⁰Sr determination in proficiency testing (PT) samples show that the accuracy of the determination does not depend on high chemical yield but depends on accuracy of yield determination. The analysis of z-values shows that 96% of obtained z-values range from 0 to ±2 while 77% of z-values range between 0 and ± 1. Ninety percent of obtained results of ⁹⁰Sr determination deviate less than 20% from assigned values in PT provider reports. The results of ⁹⁰Sr determination in animal bone samples using different methods are in good agreement. The results obtained by Cherenkov counting in both laboratories vary from -3.1-14.5% while results obtained by determination via ⁹⁰Y and counting on i-Matic vary between -10.0 and -2.9%. These deviations are in accordance with deviations obtained with PT samples. Activity concentrations of ⁹⁰Sr in wild boar bone samples range from 4 to 30 Bq kg^-1 while in deer bone samples from 2 to 8 Bq kg^-1.