Data Quality

Concentration data for Zn-70 appears to be inconsistent compared to what was expected and what was seen for other zinc isotopes. This is likely due to interferences from ⁴⁰Ar¹⁴N¹⁶O. Because of this we are choosing to emit Zn-70 from our analysis of the data quality of the other isotopes. 

Rho values for Zn-64, Zn-66, Zn-67, Zn-68 and As-75 are all 1 and the rho value for Zn-70 is 0.994, indicating that the calibration curves for all of these isotopes of interest are extremely linear. 

The accuracy of the calibration curve was almost exact for the highest calibration standard (990 ppb) for all zinc isotopes. The accuracy was within 7% for the next highest calibration standard (49.32 ppb) for all zinc isotopes except for Zn-70. The accuracy was within 4% for the next highest calibration standard (24.67 ppb), except for Zn-70. The accuracy was within 8% for the next highest calibration standard (2.47 ppb), excluding Zn-70. The accuracy for the lowest calibration standard (0.247 ppb) ranged from within 2% for Zn-66 in helium mode to within 400% for Zn-67 in helium mode. This large range of accuracy makes sense for the lowest calibration standard because there are fewer counts, and therefore, fewer counting statistics.

Arsenic had accuracies within 1% for the two highest calibration standards (101 ppb and 5.03 ppb), within 12% for the next calibration standard (2.52 ppb), and within 60% for the two lowest calibration standards (0.25 and 0.03 ppb). It makes sense that the accuracies for the two low standards are not as close because there are so few counts and limited counting statistics.

The NIST 1643 f standard was diluted 10x and run as a QC standard with calculated  concentrations of 7.47 ppb Zn and 5.77 ppb As. Zn-64 in helium mode and Zn-66 in no gas mode were within 7% of the actual concentration. Zn-66 in helium mode was within 16% of the actual concentration. Zn-67 in helium mode, Zn-68 in both modes, and Zn-70 in helium mode were not as accurate, with values within 132%, 55%, 26%, and 280% of the actual concentration, respectively. This indicates that there may be an interference in the NIST standard that wasn’t present in the calibration standard, causing the accuracy to drop for some of the zinc isotopes. As-75 run in helium mode was within 3% of the actual concentration. The precision, represented as the %QRSD, for the Zn isotopes except for Zn-70 was within 6% of the average. The precision for As was within 7.1% of the average. 

Our B stock L4 was calculated to contain 25.43 ppb Zn and 2.53 ppb As. The samples had values very close to the calculated concentrations, with sample concentrations within 5% of the actual concentrations for all isotopes except for Zn-70. The precision for the zinc isotopes was within 6% of the average for the zinc isotopes except for Zn-70.

Both Zn-66 and Zn-68 produced better results in no gas based on the LOD and %QRSD’s listed in table 5. Zn-67, Zn-64, Zn-70 and As-75 were only measured in helium.

In terms of comparing the measured analyte concentrations to the calibration standards, the concentrations of As-75 fell between the first and second calibration standard, but they were intended to fall between the second and third calibration standard. Concentrations of Zinc all fell below the calibration range, except for Zn-70, and were all intended to fall within the blank and first calibration standard based on previous zinc concentration LOD’s.

The uncertainty of the spikes can be seen in table 6 where the %QRSD represents one standard deviation of the mean.

To determine that the measurement of spikes occurred under the same analytical conditions as the standards, internal standard variation can be assessed. As seen in Figure 1 internal standard variation is limited to ±25% over the course of the run. 

The calculated LODs from the blanks can be seen in Table 7 ranging from 0.05 ppb to 0.25 ppb excluding Zn-70. 

Based on the limit of detection calculated from the blanks, the arsenic detection limit is 0.05 ppb. But, the standard deviation for this measurement is ±0.02 ppb, which is almost half of the detection limit. Additionally, looking at the precision, there’s a lot of variation in the measurements for arsenic in the blanks, which indicates that the repeatability is not as good as it is for other isotopes. The practical limit of quantification will be somewhere above the 0.05 ppb limit of detection.