|
|
Tutorial: Calibration Curves |
|---|
There are times when a linear calibration curve does not give a good fit of the calibration data. It is common in such cases to use a nonlinear function for the calibration curve. A second- or higher-order polynomial is often used in these situations. See also the tutorial "Polynomial regression in MS Excel."
You wish to analyze the lead concentration in tap water using graphite furnace AAS. The following data was collected. Report the concentration of lead in the tap water in the form of a confidence interval.
| lead conc, ppb | signal, A-s |
|---|---|
| blank | 0.006 |
| 10.0 | 0.077 |
| 20.0 | 0.138 |
| 30.0 | 0.199 |
| 40.0 | 0.253 |
| 50.0 | 0.309 |
| 60.0 | 0.356 |
| tap water sample | 0.278 |
Answer: 44.3 +/-
1.0 ppb (95% CI)
It is perhaps worth noting that the calibration data in the above example are "real" - actually collected in an AA experiment - and are not "made up" for the purposes of illustration. Thus, the procedure described is a practical, realistic way to account for apparent nonlinearity in calibration data. Atomic absorption calibration curves are particularly notorious in that regard.