Technical Note: Using FIMS to determine mercury content in sewage sludge, sediment and soil samples

Stock standard mercury solution #1, 1000 mg/1, was prepared from Merk Tritisol(R). Stock mercury solutions #2 and #3:10 mg/1 and mg/1 respectively, were prepared by further dilution of the stock standard solution #1. Calibration standards at different mercury levels were prepared from stock solutions #2 and #3 by further dilution in 3% v/v HC1. Calibration standard solutions used for the measurement of sewage samples were 0"00, 5"00, 10"00, 15-00, 20"00, and 30"00 lg/1. Calibration standard solutions used for the measurement of soil and sediment samples were 0"00, 3-00, 5"00, 7"00, and 10"00 tg/1. According to DIN 38 405-E 12, the standard solutions should contain 1% v/v of the stabilizing solution.

The Flow Injection Mercury System (FIMS) is an dedicated system that integrates flow injection mercury cold vapour generation with a very sensitive detector. Instrumental detection limits can be as low as 5 ng/1 using a sample volume of 500 tl. The FIMS permits sample dilution, therefore reducing the likelihood of an interference occurring with complex samples. In this study samples were digested with aqua regia using reflux conditions according to DIN method 38414. After proper dilution of the digested sample solution, Hg was measured interference free using SnC12 as the reductant.
The recoveries of spiked mercury in sewage sludge samples ranged from 96 to 100%. The method was checked by the analysis of standard reference materials. All results were in agreement with certified values. The RSD for three replicates was approximately 2% at 10 tg/1 Hg levels.
The sample solutions were also measured using an FIAS to generate the Hg vapour and the analytical data was collected using an AA spectrometer equipped with a 02 background corrector. The results are in agreement with those obtained with the FIMS, which demonstrates that any background absorption for these determinations was negligible.

Reagents and solutions
All chemicals were at least of analytical reagent grade, and deionized water was used throughout: (1) SnC12 2H20: Pro analysi Merck.
(3) HCI: Pro analysi Merck, 37% (max. 0"0000005% Hg Sample digestion and pretreatment 3"004-0"01 g was weighed into the digestion flask, moistened with a few drops of water, and 21 ml HC1 and 7 ml HNOa were added. 10 ml HNOa was pipetted into the absorbing vessel. The digestion procedure was started under reflux conditions, also in accordance with DIN 38414, Part 12.
After cooling, the solutions were transferred from the digestion flask into a 100 ml volume flask, diluted to the volume with deionized water and mixed well. After allowing the undigested material to settle out or after filtration, ml of the clear supernatant solution was placed in a 10 ml test tube, 100 gl of K2Cr207 stabilizing solution was added, the mixture was diluted to 10 ml and mixed well. This solution was then ready for measurement.
For very reactive samples, the glass type gas-liquid separator should be used. To learn more about the concentration levels of other coexistent metal ions in the samples, the solutions were also analysed semi-quantitatively for Cu, Ni, Pb and Zn using a Perkin-Elmer OPTIMA 3000 ICP Spectrometer.

Discussion
Stabilizing solution was added to the diluted sample solutions to keep the mercury content stable. It was found that the mercury content in diluted sample solution continuously decreased without the addition of the stabilizing solution.
When using SnC12 as a reductant, interferences have been reported for waters containing sulphide, chloride, copper and tellurium. Also, organic compounds which have broad band UV absorbance (around 253"7 nm) are confirmed interferences [1]. Iodide has also been reported to interfere with the measurement [2]. In general there are less severe interferences from heavy metal ions when SnC12 is used as a reductant compared to the use of NaBHa as the reductant. Proper dilution of the sample solutions can alleviate interferences.
It was found that the digested sample solution must be diluted (for example, to 10) prior to measurement.
Otherwise, even when 10% m/v SnC12 was used as the reductant, the measured results were low. Since the FIMS system is highly sensitive and provides improved mercury detection limits, it is possible to measure the low Hg levels even with dilution of the sample solutions.
It has been reported that there is a risk of interference from volatile nitrogen oxides when mercury is determined by FI-CVAAS in digests of samples which have been decomposed by nitric acid [3]. Concentrated aqua regia was used for the sample digestion. Nitrogen oxides were generated during the digestion process, especially for sewage sludge samples, because these samples contain large amounts of organic material. However, no interference or background signals were observed. The background signals of the diluted sample solutions measured with a D 2 background corrector were negligible.

Conclusion
After digestion using the DIN method, the mercury levels in sediment, sewage sludge and soil samples were determined using FIMS with SnC12 as reductant. The measurement is precise, simple and fast.
Relatively high Cu, Ni, Pb and Zn contents were found in the sewage sludge samples. However, by using SnC12 as the reductant, and diluting the sample solutions, the measurements were virtually interference-free. Similarly, no interference from volatile nitrogen oxides or other nonspecific absorption signals were observed.