64
River and Stream Sediments
not produce methylmercury on incubation with organic mercury, suggesting a microbiological origin for the methylmercury. A control experiment was carried out in
which identical samples were collected and homogenized. Some of the samples were
sterilized by treatment with an approximate 4 wt % solution of formaldehyde. Several
samples of both sterilized and unsterilized sediments were analysed at intervals and
all of the samples were stored at ambient room temperature (18 °C) in the laboratory.
It can be seen from Fig. 2.9 that there is a difference in behaviour between the
sterilized and unsterilized samples. Some of the samples were separately inoculated
into various growth media to test for microbiological activity.
This work suggests that the application of laboratory-derived results directly to
natural conditions could, in these cases, be misleading: analytical results for day 10, if
extrapolated directly, might lead to the conclusion that natural methylmercury levels
and rates of methylation are much greater than in fact they really are. Work in this
area, with model or laboratory systems, needs to be interpreted with particular caution.
Bartlett et al. [174] used the method ofUthe et al. [175] for determining methylmercury. Sediment samples of 2-5 g were extracted with toluene after treatment with
copper sulphate and an acidic solution of potassium bromide. Methylmercury was
then back extracted into aqueous sodium thiosulphate. This was then treated with
acidic potassium bromide and copper sulphate following which the methylmercury
was extracted into pesticide grade benzene containing approximately 100 I-lg I-I of
ethyl mercuric chloride as an internal standard. The extract was analysed by electron
capture gas chromatography. The detection limit was 1-2 ~ kg-I.
A method [176] has been described for the determination of down to 2.5I-lgkg-1
alkylmercury compounds and inorganic mercury in river sediments. This method
uses steam distillation to separate methylmercury in the distillate and inorganic
mercury in the residue. The methylmercury is then determined by flameless atomic
absorption spectrophotometry and the inorganic mercury by the same technique after
wet digestion with nitric acid and potassium permanganate [177]. These workers
considered the possible interference effects of clay, humic acids, and sulphides, all
possible components of river sediment samples, on the determination of alkylmercury compounds and inorganic mercury, and devised methods of overcoming interference effects.
Jurka and Carter [178] have described an automated determination of down to
O.ll-lg I-I mercury in river sediment samples. This method is based on the automated
procedure of EI-Awady et al. [179] for the determination of total mercury in waters
and wastewaters, in which potassium persulphate and sulphuric acid were used to
digest samples for analysis by the cold vapour technique. These workers proved that
the use of potassium permanganate as an additional oxidizing agent was unnecessary.
There was no significant interference due to sulphide in the solutions containing
10 mg sulphide I-I. However, a negative interference was observed for both organic
and inorganic standards containing 100 mg sulphide I-I which is equivalent to
25,000 mg sulphide kg- I in the sediment. This interference was overcome by ensuring
that an excess of dichromate was present during the automated analysis.
This automated procedure was estimated to have a precision of 0.13-0.21 mg Hg
kg- I at the 1 mg Hg kg- I level with standard deviations varying from 0.011 to 0.02 mg
River and Stream Sediments
not produce methylmercury on incubation with organic mercury, suggesting a microbiological origin for the methylmercury. A control experiment was carried out in
which identical samples were collected and homogenized. Some of the samples were
sterilized by treatment with an approximate 4 wt % solution of formaldehyde. Several
samples of both sterilized and unsterilized sediments were analysed at intervals and
all of the samples were stored at ambient room temperature (18 °C) in the laboratory.
It can be seen from Fig. 2.9 that there is a difference in behaviour between the
sterilized and unsterilized samples. Some of the samples were separately inoculated
into various growth media to test for microbiological activity.
This work suggests that the application of laboratory-derived results directly to
natural conditions could, in these cases, be misleading: analytical results for day 10, if
extrapolated directly, might lead to the conclusion that natural methylmercury levels
and rates of methylation are much greater than in fact they really are. Work in this
area, with model or laboratory systems, needs to be interpreted with particular caution.
Bartlett et al. [174] used the method ofUthe et al. [175] for determining methylmercury. Sediment samples of 2-5 g were extracted with toluene after treatment with
copper sulphate and an acidic solution of potassium bromide. Methylmercury was
then back extracted into aqueous sodium thiosulphate. This was then treated with
acidic potassium bromide and copper sulphate following which the methylmercury
was extracted into pesticide grade benzene containing approximately 100 I-lg I-I of
ethyl mercuric chloride as an internal standard. The extract was analysed by electron
capture gas chromatography. The detection limit was 1-2 ~ kg-I.
A method [176] has been described for the determination of down to 2.5I-lgkg-1
alkylmercury compounds and inorganic mercury in river sediments. This method
uses steam distillation to separate methylmercury in the distillate and inorganic
mercury in the residue. The methylmercury is then determined by flameless atomic
absorption spectrophotometry and the inorganic mercury by the same technique after
wet digestion with nitric acid and potassium permanganate [177]. These workers
considered the possible interference effects of clay, humic acids, and sulphides, all
possible components of river sediment samples, on the determination of alkylmercury compounds and inorganic mercury, and devised methods of overcoming interference effects.
Jurka and Carter [178] have described an automated determination of down to
O.ll-lg I-I mercury in river sediment samples. This method is based on the automated
procedure of EI-Awady et al. [179] for the determination of total mercury in waters
and wastewaters, in which potassium persulphate and sulphuric acid were used to
digest samples for analysis by the cold vapour technique. These workers proved that
the use of potassium permanganate as an additional oxidizing agent was unnecessary.
There was no significant interference due to sulphide in the solutions containing
10 mg sulphide I-I. However, a negative interference was observed for both organic
and inorganic standards containing 100 mg sulphide I-I which is equivalent to
25,000 mg sulphide kg- I in the sediment. This interference was overcome by ensuring
that an excess of dichromate was present during the automated analysis.
This automated procedure was estimated to have a precision of 0.13-0.21 mg Hg
kg- I at the 1 mg Hg kg- I level with standard deviations varying from 0.011 to 0.02 mg