ELEMENTAL DISTRIBUTION: OVERVIEW
Y. Nozaki
w , University of Tokyo, Tokyo, Japan
Copyright & 2001 Elsevier Ltd.
Introduction
More than 97% of liquid water on the earth exists in
the ocean. The ocean water contains approximately
3.5% by weight of dissolved salt. What is the elemental composition of the salts, how does it vary from
place to place and with depth, and why? These are
fundamental questions for which chemical oceanographers have sought answers. Despite more than a
hundred years of intense investigation by modern
chemical oceanography, the answers have not been
fully elucidated. Nevertheless, we are now approaching complete understanding of the chemical
composition of sea water and its variability in the
ocean.
Historical Review
By the late nineteenth century it was well-established
that the major components of sea water are extremely
constant in their relative abundance, and comprise
some ten constituents including Cl
À , Na
þ , Mg
2þ
,
SO 4
2À (see Conservative Elements). The analytical
results reported by W. Dittmar in 1884 for waters
collected during the British RMS Challenger Expedition (1872–1876) from the world’s oceans were almost the same as today’s values. The constancy of
major chemical composition has led oceanographers
to define ‘salinity’ as a fundamental property together
with temperature to calculate the density of sea water.
It was routine for classic physical oceanographers to
titrate sea water for chloride (plus bromide) ion with
silver nitrate standard solution, until the mid 1960s
when salinity could be determined more practically
by measurement of conductivity.
On the other hand, for minor elements, there has
been little information gained since the establishment
of major chemical composition of sea water. Measurements of trace constituents in sea water are difficult because of their very low abundance. There was
a clear tendency for the reported concentrations of
many trace elements to become lower and lower as
time elapsed. This trend was, of course, not real but
an artifact. It is a famous story that, to aid Germany’s
national deficit after World War I, the Nobel Prize
winning chemist F. Haber attempted to recover gold
from sea water which according to the current literature occurred at about 5mg m
À3
. He completely
failed however, but, after long and rigorous examination, he found that the concentration was B1000
times less than that expected. Incidentally, Haber’s
value of gold concentration was two orders of magnitude higher compared to later reports (Table 1).
Another good example may be found in the measurements of lead in sea water by Patterson and his
associates (see Anthropogenic Trace Elements in the
Ocean). The vertical profile of Pb in the North Pacific
obtained by Schaule and Patterson in 1981 had concentrations about two orders of magnitude lower
than those reported earlier (B1970) by the same
workers although their 1981 values are believed to be
accurate and real (Table 1).
Technical Challenge
It is now known that the most obvious reason for
these trends is the continuous improvement in removing sources of contamination during sampling,
handling, storage, and analysis. Significant efforts and
advances in such field and laboratory techniques had
been made until the GEOSECS (Geochemical Ocean
Section Study) program started at around 1970. For
example, polyvinyl chloride Niskin-bottle multisampling system together with CTD (conductivity–
temperature–depth) sensors has routinely been employed in the hydrocasts, replacing the serial Nansen
(metallic) bottle sampling method most widely used
prior to that time. Yet, this was not enough for many
trace metals except for barium, and an intercalibration exercise made in the early stage of the program
did not produce any congruent results between laboratories. It was a significant and wise decision of
the GEOSECS leaders that they focused on radionuclides and stable isotopes, which are almost free
from contamination, and did not get involved in trace
element geochemistry. Obviously, without having the
real concentration data, any arguments that might be
built upon them would be meaningless.
Obtaining clean (uncontaminated) water samples
from various depths of the ocean is of prime
importance in the study of trace metals. In this regard, various types of sampling bottles have been
developed both domestically and commercially.
They include the Cal-Tech Patterson sampler,
w Deceased.
7
Y. Nozaki
w , University of Tokyo, Tokyo, Japan
Copyright & 2001 Elsevier Ltd.
Introduction
More than 97% of liquid water on the earth exists in
the ocean. The ocean water contains approximately
3.5% by weight of dissolved salt. What is the elemental composition of the salts, how does it vary from
place to place and with depth, and why? These are
fundamental questions for which chemical oceanographers have sought answers. Despite more than a
hundred years of intense investigation by modern
chemical oceanography, the answers have not been
fully elucidated. Nevertheless, we are now approaching complete understanding of the chemical
composition of sea water and its variability in the
ocean.
Historical Review
By the late nineteenth century it was well-established
that the major components of sea water are extremely
constant in their relative abundance, and comprise
some ten constituents including Cl
À , Na
þ , Mg
2þ
,
SO 4
2À (see Conservative Elements). The analytical
results reported by W. Dittmar in 1884 for waters
collected during the British RMS Challenger Expedition (1872–1876) from the world’s oceans were almost the same as today’s values. The constancy of
major chemical composition has led oceanographers
to define ‘salinity’ as a fundamental property together
with temperature to calculate the density of sea water.
It was routine for classic physical oceanographers to
titrate sea water for chloride (plus bromide) ion with
silver nitrate standard solution, until the mid 1960s
when salinity could be determined more practically
by measurement of conductivity.
On the other hand, for minor elements, there has
been little information gained since the establishment
of major chemical composition of sea water. Measurements of trace constituents in sea water are difficult because of their very low abundance. There was
a clear tendency for the reported concentrations of
many trace elements to become lower and lower as
time elapsed. This trend was, of course, not real but
an artifact. It is a famous story that, to aid Germany’s
national deficit after World War I, the Nobel Prize
winning chemist F. Haber attempted to recover gold
from sea water which according to the current literature occurred at about 5mg m
À3
. He completely
failed however, but, after long and rigorous examination, he found that the concentration was B1000
times less than that expected. Incidentally, Haber’s
value of gold concentration was two orders of magnitude higher compared to later reports (Table 1).
Another good example may be found in the measurements of lead in sea water by Patterson and his
associates (see Anthropogenic Trace Elements in the
Ocean). The vertical profile of Pb in the North Pacific
obtained by Schaule and Patterson in 1981 had concentrations about two orders of magnitude lower
than those reported earlier (B1970) by the same
workers although their 1981 values are believed to be
accurate and real (Table 1).
Technical Challenge
It is now known that the most obvious reason for
these trends is the continuous improvement in removing sources of contamination during sampling,
handling, storage, and analysis. Significant efforts and
advances in such field and laboratory techniques had
been made until the GEOSECS (Geochemical Ocean
Section Study) program started at around 1970. For
example, polyvinyl chloride Niskin-bottle multisampling system together with CTD (conductivity–
temperature–depth) sensors has routinely been employed in the hydrocasts, replacing the serial Nansen
(metallic) bottle sampling method most widely used
prior to that time. Yet, this was not enough for many
trace metals except for barium, and an intercalibration exercise made in the early stage of the program
did not produce any congruent results between laboratories. It was a significant and wise decision of
the GEOSECS leaders that they focused on radionuclides and stable isotopes, which are almost free
from contamination, and did not get involved in trace
element geochemistry. Obviously, without having the
real concentration data, any arguments that might be
built upon them would be meaningless.
Obtaining clean (uncontaminated) water samples
from various depths of the ocean is of prime
importance in the study of trace metals. In this regard, various types of sampling bottles have been
developed both domestically and commercially.
They include the Cal-Tech Patterson sampler,
w Deceased.
7