Analysis of metal content in water used for consumption and industrial purposes is critical due to fluctuating water quality with respect to geology, geography and pollution. Although techniques like ICP-OES and ICP-MS have gained popularity for measuring minerals in water, flame atomic absorption (AA) spectrometry is still extensively used, as it is a simple, inexpensive and rapid method for water quality analysis.
This article demonstrates the determination of seven non-toxic elements commonly observed in drinking water, using the PerkinElmer PinAAcle™ 500 Flame Atomic Absorption (AA) spectrometer. When other low-level elements need to be measured, techniques such as Graphite Furnace AA, ICP-OES, and ICP-MS are commonly used.
Experimental Procedure
Samples include a certified drinking water standard (Trace Metals in Drinking Water - High-Purity Standards™, Charleston, South Carolina, USA), spring water obtained from a local grocery store, and locally collected municipal and well waters.
Sample preparation involves using 1% HNO3 (v/v) to acidify each water sample, and 0.1% lanthanum chloride is added as an ionization suppressant for sodium (Na) and potassium (K) and as a releasing reagent for calcium (Ca) and magnesium (Mg).
Using the PinAAcle 500 flame AA spectrometer, the analyses were performed under conditions specified in Tables 1 and 2. The burner was rotated by 30 degrees to minimize the signal intensity for mineral analysis due to their high concentrations in the samples.
Furthermore, analysis of K and Na was carried out in emission mode. For this, the PinAAcle 500 Flame AA was auto-configured to expand the analytical range and measure even higher concentrations. As a result, minimal dilution was required for K and dilution was eliminated for Na.
Table 1. PinAAcle 500 instrument and analytical conditions common to all elements
| Parameter |
Value |
| Air Flow (L/min) |
2.5 |
| Acetylene Flow (L/min) |
10 |
| Read Time (sec) |
3 |
| Replicates |
3 |
Table 2. PinAAcle 500 instrument and analytical conditions specific to each element
| Element |
Wavelength (nm) |
Slit (nm) |
Mode |
Burner Angle (degrees) |
Calibration Standards (mg/L) |
Calibration Curve |
| Ca |
422.67 |
0.7 |
Absorption |
30 |
0.5, 1.0, 2.0, 5.0, 10, 20, 40 |
Non-Linear through Zero |
| Cu |
324.75 |
0.7 |
Absorption |
0 |
0.05, 0.10, 0.25, 0.50 |
Linear Through Zero |
| Fe |
248.33 |
0.2 |
Absorption |
0 |
0.05, 0.10, 0.25, 0.50, 1.0 |
Linear Through Zero |
| Mg |
285.21 |
0.7 |
Absorption |
30 |
0.5, 1.0, 2.0, 5.0, 10 |
Non-Linear Through Zero |
| K |
766.49 |
0.7 |
Emission |
30 |
2, 5, 10, 20, 30, 40, 50 |
Non-Linear Through Zero |
| Na |
589.00 |
0.2 |
Emission |
30 |
2, 5, 10, 20, 30, 40, 50 |
Non-Linear Through Zero |
| Zn |
213.86 |
0.7 |
Absorption |
0 |
0.05, 0.10, 0.25, 0.50 |
Linear Through Zero |
A high-sensitivity nebulizer on the PinAAcle 500 spectrometer was used to introduce the samples via self-aspiration. In order to ensure maximum sensitivity, the nebulizer was used without the spacer to determine zinc (Zn), iron (Fe), and copper (Cu). However, the spacer is employed for the analysis of Ca, Mg, K, and Na.
Results and Discussion
The correlation coefficients of all calibration curves were 0.999 or higher. The calibration precision was evaluated through an Independent Calibration Verification (ICV) solution, which was diluted 100 times to be within the calibration curve range. Table 3 shows the results of the ICV, which indicates the accuracy of the calibration curves.
Table 3. Results for independent calibration verification (ICV)
| Element |
Concentration (mg/L) |
Experimental (mg/L) |
% Recovery |
| Ca |
5.00 |
4.86 |
97 |
| Cu |
0.25 |
0.26 |
104 |
| Fe |
1.00 |
1.00 |
100 |
| Mg |
5.00 |
4.88 |
98 |
| K |
5.00 |
4.78 |
96 |
| Na |
5.00 |
5.12 |
102 |
| Zn |
0.20 |
0.21 |
105 |
A reference material was assessed first to evaluate the accuracy of the methodology, with the results being shown in Table 4. It is evident from the results that all recoveries fall within 10% of the certified value, establishing the accuracy of the methodology.
Table 4. Results for reference material (all units in mg/L)
| Element |
Experimental (mg/L) |
Certified (mg/L) |
% Recovery |
| Ca |
33.4 |
35.0 |
95 |
| Cu |
0.022 |
0.020 |
110 |
| Fe |
0.095 |
0.100 |
95 |
| Mg |
8.69 |
9.00 |
97 |
| K |
2.28 |
2.50 |
91 |
| Na |
5.90 |
6.0 |
98 |
| Zn |
0.070 |
0.070 |
100 |
Upon establishing the accuracy of the methodology, the analysis of various drinking water samples collected from different sources was performed. The municipal and well water samples were directly obtained from a faucet. The spring water samples were transferred from the purchased bottles. Table 5 shows the results of the analysis of different samples.
Table 5. Results for samples (all units in mg/L)
| Element |
Municipal Water (mg/L) |
Well Water-1 (mg/L) |
Well Water-2 (mg/L) |
Well Water-3 (mg/L) |
Spring Water-1 (mg/L) |
Spring Water-2 (mg/L) |
| Ca |
17.7 |
0.148 |
35.3 |
32.4 |
3.43 |
19.2 |
| Cu |
0.048 |
< DL |
0.052 |
0.017 |
< DL |
< DL |
| Fe |
< DL |
< DL |
0.019 |
< DL |
< DL |
< DL |
| Mg |
6.43 |
0.026 |
4.90 |
5.12 |
0.799 |
6.09 |
| K |
< 0 |
233* |
4.89 |
4.10 |
0.73 |
0.69 |
| Na |
38.4 |
3.63 |
10.9 |
42.9 |
6.60 |
7.25 |
| Zn |
0.008 |
0.043 |
0.010 |
0.023 |
< DL |
< DL |
*Sample required a 10x dilution
The four samples collected from the faucet were seen to have Cu and Zn, which is most probably due to leaching caused by copper pipes, fittings, and solder. Well Water-1 is unique due to the lowest concentrations of all elements, except K, out of all samples. The residence from which this sample was obtained uses a water softener that eliminates high quantities of Ca and Mg from the well water by using K as the counter-ion.
Cu and Zn were found to be absent in the spring water, as anticipated; the sample consisted of only the minerals. The changes in mineral concentration indicate different geologies from where the water originates.
The detection limits for Cu, Fe, and Zn were determined as three times the standard deviation of 10 blank measurements (i.e. 1% HNO3), as displayed in Table 6. The detection limits for the minerals like Na, Mg, K, and Ca were not measured due to their elevated levels. Moreover, the instrument was detuned for analyzing these elements as they generally occur at elevated concentrations. Hence, detection limits have no significance.
Table 6. Detection limits
| Element |
Detection Limit (mg/L) |
| Cu |
0.002 |
| Fe |
0.006 |
| Zn |
0.004 |
Conclusion
This article illustrates the use of the PinAAcle 500 Flame AA to determine mineral elements present in various drinking water samples such as spring, well and municipal waters. With features like emission mode measurement and burner rotations, PinAAcle 500 can easily measure both minor and mineral elements.
In addition, the presence of Syngistix Touch™ software enables exclusive operation of the instrument via a touchscreen interface. The PinAAcle 500 Flame AA spectrometer also has an on-board computer option to run Syngistix™ for AA software to achieve better flexibility. With these capabilities, the PinAAcle 500 is a suitable solution for drinking water analysis.
This information has been sourced, reviewed and adapted from materials provided by PerkinElmer.
For more information on this source, please visit PerkinElmer.