102
R. Capelli et aL
study of the water column was carried out in the
proximity of the buoy ODAS Italia-1 in the
Ligurian sea (43°48.90' N 09°06.80' E), anchored
at a 1270 metre depth. During these off-shore
investigations, density and salinity along the
water column were measured by means of a CTD
Profiler. Samples were collected at different
depths, starting from the surface and down to
600 metres. Depths were chosen by taking into
consideration the thermocline and temperature
and salinity values. Sea water sampling down to a
600 meter depth was carried out by means of the
new MARIKIKI sampling equipment designed
and tested by the Department of Pharmaceutical
and Food Chemistry and Technologies of the
University of Genoa. This previously described
sampling equipment (Capelli et a1.1998) basically consists of a nylon body holding a container, in
the specific case a quartz tube (diameter 5 em,
thickness 1 mm, volume 300-500 ml). A vacuum
can be created simultaneously inside the tube
and the container through a glass capillary that
connects them. The sampler is fastened to the
cable attached to a winch on the research vessel
and lowered to the desired depth. At this point a
messenger slides down along the cable and
breaks the glass capillary. Water is sucked inside
and fills both the quartz tube and the external
container. The sampler is retrieved on board and
the quartz tube full of sea water is sealed with a
cap and either stored or immediately analysed
on board the ship. The "MARIKIKI" sampler has
two nylon bodies and two quartz tubes in order
to collect two samples in the same place and at
the same time. The most significant characteristics of this kind of sampler can be summarised as
follows:
1. Each sample is obtained with a previously
washed quartz tube whose blank value is
known.
2. The tube in which the vacuum is created
does not have any contact with other water
until the moment in which the capillary
breaks.
3. It is possible to obtain two samples in the
same place and, especially, at the same time.
4. The quartz tube can be heated up to 800°C to
ensure its cleanliness and also sterilised in
case of microbiological research.
5. Both the quartz tube and the container can
be substituted with other more suitable
materials to obtain water samples for the
determination of other substances.
Analytical Methods
All analyses were carried out on board the ship
immediately after sampling, to avoid the need to
add adds to prevent loss of ionic mercury due to
absorption onto the container surface. Mercury
was chemically reduced to H~ vapour by means
of 10% (m/v) tin (II) chloride dihydrate in 20%
(v/v) sulphuric acid. The Hgo vapours were
stripped from the solution by a stream of argon
and collected on a gold trap, with 200-300 mg of
0.1 mm diameter gold wire placed in a 4 mm
(inner diameter) quartz tube. After pre-concentration of all mercury from the sample. the trap
was electrically heated and the released H~
vapours were detected with a mercury fluorescence detector (Merlin, PSA Analytical LTD)
connected with a Hewlett-Packard HP3396
Series II Recorder/Integrator. The obtained concentrations of mercury correspond to "reactive
mercury". i.e. the dissolved inorganic mercury
species, labile organo-mercury associations and
mercury that is easily leachable (Gill and
Fitzgerald 1985).
Calibration was carried out by means of the
standard additions method on each sample,
adding 0.1 and 0.2 ng of mercury. Slope of calibrations was checked by comparison with a
direct injection of known amount of mercury
vapour obtained by gas sampling in a container
with metallic Hg in equilibrium with Hg vapour,
kept at a constant temperature.
Each quartz container was add washed, heated at 600°C for 12 hours, and, after cooling, fllied
with ultra-pure (>18MOhm·cm) water. The
water was then analysed with the procedure
described, and the cleaning was repeated until a
mercury level lower than 0.08 nglL was achieved.
The mean (n=37) blank value obtained is 0.06
ngiL with values ranging 0.01-0.08 ng/L and a
standard deviation of 0.02 ng/L. The detection
limit of the method (3 0') is 0.06 ngiL. The blank
value for each quartz container was recorded and
subtracted from the mercury concentration
measured in the sea water.
Results and Discussion
Reactive mercury concentrations obtained are
reported in Fig. 1, together with salinity and temperature data. No correlation between mercury
concentration and salinity. temperature or depth
R. Capelli et aL
study of the water column was carried out in the
proximity of the buoy ODAS Italia-1 in the
Ligurian sea (43°48.90' N 09°06.80' E), anchored
at a 1270 metre depth. During these off-shore
investigations, density and salinity along the
water column were measured by means of a CTD
Profiler. Samples were collected at different
depths, starting from the surface and down to
600 metres. Depths were chosen by taking into
consideration the thermocline and temperature
and salinity values. Sea water sampling down to a
600 meter depth was carried out by means of the
new MARIKIKI sampling equipment designed
and tested by the Department of Pharmaceutical
and Food Chemistry and Technologies of the
University of Genoa. This previously described
sampling equipment (Capelli et a1.1998) basically consists of a nylon body holding a container, in
the specific case a quartz tube (diameter 5 em,
thickness 1 mm, volume 300-500 ml). A vacuum
can be created simultaneously inside the tube
and the container through a glass capillary that
connects them. The sampler is fastened to the
cable attached to a winch on the research vessel
and lowered to the desired depth. At this point a
messenger slides down along the cable and
breaks the glass capillary. Water is sucked inside
and fills both the quartz tube and the external
container. The sampler is retrieved on board and
the quartz tube full of sea water is sealed with a
cap and either stored or immediately analysed
on board the ship. The "MARIKIKI" sampler has
two nylon bodies and two quartz tubes in order
to collect two samples in the same place and at
the same time. The most significant characteristics of this kind of sampler can be summarised as
follows:
1. Each sample is obtained with a previously
washed quartz tube whose blank value is
known.
2. The tube in which the vacuum is created
does not have any contact with other water
until the moment in which the capillary
breaks.
3. It is possible to obtain two samples in the
same place and, especially, at the same time.
4. The quartz tube can be heated up to 800°C to
ensure its cleanliness and also sterilised in
case of microbiological research.
5. Both the quartz tube and the container can
be substituted with other more suitable
materials to obtain water samples for the
determination of other substances.
Analytical Methods
All analyses were carried out on board the ship
immediately after sampling, to avoid the need to
add adds to prevent loss of ionic mercury due to
absorption onto the container surface. Mercury
was chemically reduced to H~ vapour by means
of 10% (m/v) tin (II) chloride dihydrate in 20%
(v/v) sulphuric acid. The Hgo vapours were
stripped from the solution by a stream of argon
and collected on a gold trap, with 200-300 mg of
0.1 mm diameter gold wire placed in a 4 mm
(inner diameter) quartz tube. After pre-concentration of all mercury from the sample. the trap
was electrically heated and the released H~
vapours were detected with a mercury fluorescence detector (Merlin, PSA Analytical LTD)
connected with a Hewlett-Packard HP3396
Series II Recorder/Integrator. The obtained concentrations of mercury correspond to "reactive
mercury". i.e. the dissolved inorganic mercury
species, labile organo-mercury associations and
mercury that is easily leachable (Gill and
Fitzgerald 1985).
Calibration was carried out by means of the
standard additions method on each sample,
adding 0.1 and 0.2 ng of mercury. Slope of calibrations was checked by comparison with a
direct injection of known amount of mercury
vapour obtained by gas sampling in a container
with metallic Hg in equilibrium with Hg vapour,
kept at a constant temperature.
Each quartz container was add washed, heated at 600°C for 12 hours, and, after cooling, fllied
with ultra-pure (>18MOhm·cm) water. The
water was then analysed with the procedure
described, and the cleaning was repeated until a
mercury level lower than 0.08 nglL was achieved.
The mean (n=37) blank value obtained is 0.06
ngiL with values ranging 0.01-0.08 ng/L and a
standard deviation of 0.02 ng/L. The detection
limit of the method (3 0') is 0.06 ngiL. The blank
value for each quartz container was recorded and
subtracted from the mercury concentration
measured in the sea water.
Results and Discussion
Reactive mercury concentrations obtained are
reported in Fig. 1, together with salinity and temperature data. No correlation between mercury
concentration and salinity. temperature or depth