Geoscience Reference
In-Depth Information
iron also require atmospheric oxygen concentrations much lower than
today's values.
All of this evidence for low atmospheric oxygen levels in the Archean
Eon, while compelling on its own, has largely given way to another line
of reasoning that was so unexpected and so novel that nobody saw it
coming. In 1999 my good friend and colleague James Farquhar was a
postdoc in the lab of Mark Thiemens at the University of California
San Diego. He was working late one night measuring the isotopic com-
position of sulfur compounds on Mark's mass spectrometer. As the re-
sults rolled off the screen, James panicked. Suddenly there was some-
thing so weird that he was sure he had damaged the machine. He shut
down and went home sulking, thinking of how to break the bad news
to Mark. he next morning, with a clearer head, he tried a few more
samples and found that his original observations were correct. After run-
ning a series of standards, he was convinced that the machine was func-
tioning well. After giving the data some thought, he could now tell Mark
that instead of breaking the mass spectrometer, he had found a com-
pletely new way to understand the dynamics of atmospheric oxygen on
early Earth.
5
Let's look at what James did and what he saw. His objective was to
measure the isotopic composition of sulfur compounds in ancient sedi-
mentary rocks. Sulfur has four stable isotopes:
32
S,
33
S,
34
S, and
36
S.
6
In
terms of natural abundance, we find most sulfur as
32
S, accounting for
95.02% of the total. Much less abundant is
34
S, accounting for 4.21 %
of the total;
33
S and
36
S are only present in minor amounts, accounting
for 0.75 % of the total in the case of
33
S, and 0.02 % in the case of
36
S.
For decades, geologists had concerned themselves with measuring only
the ratio between
34
S and
32
S. This is for two reasons. One is that these
isotopes are the most abundant and therefore the easiest to measure.
The other is that by standard thinking, it shouldn't matter which iso-
topes we look at (so why not go for the easiest). The idea is that, as we
saw in
chapter 6
for carbon, if some process favors one isotope over
another (known as a fractionation), then all of the isotopes of the same
element should behave in a predictable manner; here the degree to which
a given isotope is favored (fractionation) depends on the mass of the
isotope. It works like this. There is a one mass unit difference between
