Fluvial Transports of Mercury Pollution in the III River Basin (Northeastern France)
503
tributaries (Vieille Thur, Lauch, Fecht and Giessen; Fig. 1). For each water period,
the samples were collected the same day for all stations; the hydrological
conditions were thus the same for all the stations.
River water samples were immediately filtered in the field under nitrogen
pressure using O.45-~Lm Millipore filters. For mercury analysis in the solute
phase, the filtrate was put in a 500 ml Pyrex glass bottle with 20 ml of HN0 3
(35%) and 20 ml of H2 S0 4 (50%). To avoid mercury losses when the bottle is
opened just before analyzing the filtrate, the bottle is closed with Parafilm and a
cap. Some drops of KMn04 (6%) are added to the filtrate just before analyzing to
oxidize mercury. It is important to note that KMn04 is added at the last moment
rather than in the field after filtration because it could be observed that by adding
KMn04 in the field, a manganese precipitate is formed and traps mercury
between the field and the laboratory.
Part of the filtrate is treated with HN0 3 to analyze trace elements by ICP-AES
(Samuel and Rouault 1983) or ICP-MS using a quick scanning method (± 10%
accuracy). The same techniques were used to analyze trace elements present in
sediments. Major elements were determined for part of the filtrate without
treatment in the field using the following classical techniques (Krempp 1988;
Probst et al. 1992): atomic absorption for Ca H , MgH, Na+ and K+ (accuracy 1
~Lmol!l), ionic chromatography for Cl ,SO~ -, NO~ ,NO~ and PO! (accuracy
1 ~Lmol!l), Gran titration with HCl for alkalinity (accuracy 1 ~lmol!l) and
colorimetric method for H 4 Si0 4 (accuracy 1 ~Lmol!l).
The suspended matters and the bottom sediments were dried at room
temperature to avoid mercury loss. As one can see in Fig. 2, when the drying
temperature of the sediments increases, the mercury loss increases and the
Temperature eC)
Fig. 2. Effect of the drying temperature of sediment on loss of mercury
503
tributaries (Vieille Thur, Lauch, Fecht and Giessen; Fig. 1). For each water period,
the samples were collected the same day for all stations; the hydrological
conditions were thus the same for all the stations.
River water samples were immediately filtered in the field under nitrogen
pressure using O.45-~Lm Millipore filters. For mercury analysis in the solute
phase, the filtrate was put in a 500 ml Pyrex glass bottle with 20 ml of HN0 3
(35%) and 20 ml of H2 S0 4 (50%). To avoid mercury losses when the bottle is
opened just before analyzing the filtrate, the bottle is closed with Parafilm and a
cap. Some drops of KMn04 (6%) are added to the filtrate just before analyzing to
oxidize mercury. It is important to note that KMn04 is added at the last moment
rather than in the field after filtration because it could be observed that by adding
KMn04 in the field, a manganese precipitate is formed and traps mercury
between the field and the laboratory.
Part of the filtrate is treated with HN0 3 to analyze trace elements by ICP-AES
(Samuel and Rouault 1983) or ICP-MS using a quick scanning method (± 10%
accuracy). The same techniques were used to analyze trace elements present in
sediments. Major elements were determined for part of the filtrate without
treatment in the field using the following classical techniques (Krempp 1988;
Probst et al. 1992): atomic absorption for Ca H , MgH, Na+ and K+ (accuracy 1
~Lmol!l), ionic chromatography for Cl ,SO~ -, NO~ ,NO~ and PO! (accuracy
1 ~Lmol!l), Gran titration with HCl for alkalinity (accuracy 1 ~lmol!l) and
colorimetric method for H 4 Si0 4 (accuracy 1 ~Lmol!l).
The suspended matters and the bottom sediments were dried at room
temperature to avoid mercury loss. As one can see in Fig. 2, when the drying
temperature of the sediments increases, the mercury loss increases and the
Temperature eC)
Fig. 2. Effect of the drying temperature of sediment on loss of mercury