CHAPTER 19 • PIXE Analysis for Trace Elements in Marine Environments
The technique (Luke 1986) is based on the high stability of the carbamate of several elements and on the low solubility in water that is determined by pH. Keeping the
pH values in a suitable range, a large number of carbamate elements precipitate and
can be collected by filtration on a thin backing using a coprecipitant agent (that acts
also as internal standard).
At pH 9, with sodium-diethyldithiocarbamate (Na DDTC) as chelant (Aprilesi et al.
1984) the largest number of elements, 41, precipitate (as carbamates mainly, as carbamates and hydroxides or as hydroxides alone). This coprecipitation effect represents an
advantage when diluted solutions (under ppb (Ilg kg-') level} are treated, as is the case
with open sea water, where the carbamate hydroxide precipitate may also act as a "scavenger" for constituents in solution. However it becomes a problem with fresh water or
ground water, because, even if Na DDTC does not form chelates with alkaline earth
metals, at pH 9 some major cations, like Ca, precipitate following the equilibra of their
other salts. This phenomenon strongly affects the features of the resulting target.
A modified target preparation procedure was performed that takes place at pH 4
with the synergistic effect of two chelating agents, Na DDTC and ammoniumpirrolydine-dithiocarbamate (APDC), which also do not chelate the alkaline earths
(Cecchi et al. 1987).
Palladium is used as the coprecipitant agent and internal standard, given the high
stability of its carbamate complexes and for its rarity in the tested specimens. In the
case of spectral interferences, Cu is used as the internal standard.
Precipitation is carried out as follows: 100 Ilg of Pd (from Pd solution) is added to
the sample. An ammonia solution is added to keep the pH to 4. The carbamate elements are precipitated by adding 1 ml of a freshly prepared Na DDTC APDC (1% each
one) solution which has been previously purified by shaking with an equal volume of
freon (1,1,2-trichlorine-2,2,1-trifluorine ethane). The sample is allowed to stand for a
certain time depending on the treated volume (from 30 minutes to 12 hours) and is
then filtered through a Nuclepore PC membrane (0.4 11m pore size, 10 11m thickness,
1 mg cm- 2 surface density and 10 8 pore cm- 2 , filtration area of 0.79 cm 2 ). The possible
non-homogeneity of the precipitate deposit can be integrated by performing PIXE
measurements with a uniformly diffused beam larger than the filtration area.
This procedure has been widely tested and calibrated with mono and multi -element
solutions (fresh and artificial sea water) and with reference materials for Ti, V, Cr, Mn,
Fe, Co, Ni, Cu, Zn, Se, Hg, Pb, Bi, Mo, Y, Tl, Sb, Ag, Cd in the ppt-ppm range, with
preconcentration of variable sample volumes, from soo to IS ml. The recovery efficiency
(Cecchi et al. 1990b) remains generally higher than 80% for Fe, Co, Ni, Cu, Zn, Se, Hg,
Pb, Ag, Cd, higher than so% for Ti, V, Cr, Mo, Sb, and lower than 2S% for Y and T1.
The detection limits (for lSIlC of 1.8 MeV protons in an averaged blank spectrum
for the calibrated K and L X-rays) are reported in Table 19.1. The values in the blank,
with respect to the detection limits for the Nuclepore backing alone, are affected by
the slightly increased target thickness and by the behaviour of each element as carbamate.
Satisfactory reproductiveness of the results have been verified, with measurement
errors generally lower than 10%. The PIXE measurement error for these samples mainly
depends on the uncertainty of the chemical treatement and on the variance of a suitable number of blank samples measured under the same experimental conditions; in
minor part it depends on the uncertainty of the experimental setup condition.
The technique (Luke 1986) is based on the high stability of the carbamate of several elements and on the low solubility in water that is determined by pH. Keeping the
pH values in a suitable range, a large number of carbamate elements precipitate and
can be collected by filtration on a thin backing using a coprecipitant agent (that acts
also as internal standard).
At pH 9, with sodium-diethyldithiocarbamate (Na DDTC) as chelant (Aprilesi et al.
1984) the largest number of elements, 41, precipitate (as carbamates mainly, as carbamates and hydroxides or as hydroxides alone). This coprecipitation effect represents an
advantage when diluted solutions (under ppb (Ilg kg-') level} are treated, as is the case
with open sea water, where the carbamate hydroxide precipitate may also act as a "scavenger" for constituents in solution. However it becomes a problem with fresh water or
ground water, because, even if Na DDTC does not form chelates with alkaline earth
metals, at pH 9 some major cations, like Ca, precipitate following the equilibra of their
other salts. This phenomenon strongly affects the features of the resulting target.
A modified target preparation procedure was performed that takes place at pH 4
with the synergistic effect of two chelating agents, Na DDTC and ammoniumpirrolydine-dithiocarbamate (APDC), which also do not chelate the alkaline earths
(Cecchi et al. 1987).
Palladium is used as the coprecipitant agent and internal standard, given the high
stability of its carbamate complexes and for its rarity in the tested specimens. In the
case of spectral interferences, Cu is used as the internal standard.
Precipitation is carried out as follows: 100 Ilg of Pd (from Pd solution) is added to
the sample. An ammonia solution is added to keep the pH to 4. The carbamate elements are precipitated by adding 1 ml of a freshly prepared Na DDTC APDC (1% each
one) solution which has been previously purified by shaking with an equal volume of
freon (1,1,2-trichlorine-2,2,1-trifluorine ethane). The sample is allowed to stand for a
certain time depending on the treated volume (from 30 minutes to 12 hours) and is
then filtered through a Nuclepore PC membrane (0.4 11m pore size, 10 11m thickness,
1 mg cm- 2 surface density and 10 8 pore cm- 2 , filtration area of 0.79 cm 2 ). The possible
non-homogeneity of the precipitate deposit can be integrated by performing PIXE
measurements with a uniformly diffused beam larger than the filtration area.
This procedure has been widely tested and calibrated with mono and multi -element
solutions (fresh and artificial sea water) and with reference materials for Ti, V, Cr, Mn,
Fe, Co, Ni, Cu, Zn, Se, Hg, Pb, Bi, Mo, Y, Tl, Sb, Ag, Cd in the ppt-ppm range, with
preconcentration of variable sample volumes, from soo to IS ml. The recovery efficiency
(Cecchi et al. 1990b) remains generally higher than 80% for Fe, Co, Ni, Cu, Zn, Se, Hg,
Pb, Ag, Cd, higher than so% for Ti, V, Cr, Mo, Sb, and lower than 2S% for Y and T1.
The detection limits (for lSIlC of 1.8 MeV protons in an averaged blank spectrum
for the calibrated K and L X-rays) are reported in Table 19.1. The values in the blank,
with respect to the detection limits for the Nuclepore backing alone, are affected by
the slightly increased target thickness and by the behaviour of each element as carbamate.
Satisfactory reproductiveness of the results have been verified, with measurement
errors generally lower than 10%. The PIXE measurement error for these samples mainly
depends on the uncertainty of the chemical treatement and on the variance of a suitable number of blank samples measured under the same experimental conditions; in
minor part it depends on the uncertainty of the experimental setup condition.