by Charles Baum
different suite of X-rays of different wavelengths. The task, then, is to
capture these X-rays and measure their diagnostic wavelengths.
Microprobe analysis of most elements has become routine and
automated. The machine is a workhorse, operating 24 hours a day for
analyses of silicon, magnesium, iron, and other rock-forming elements.
But the lightest elements, including the one of greatest interest to us—
the key biochemical element carbon—pose a severe analytical chal-
lenge. Lighter elements tend to be rather inefficient at producing
X-rays, while the relatively few X-rays that are produced have rather
low energies. Both of these factors complicate carbon analysis. What’s
more, we routinely use carbon to coat rock samples prior to probing,
in order to make them electrically conductive. The coat is essential to
prevent the sample from building up an electric charge while being
bombarded by electrons, but it can mask any carbon in our samples. In
short, the Lab’s usual microprobe procedures wouldn’t work. We’d have
to devise new protocols.
The Geophysical Lab’s deviser of protocols is Christos Hadidiacos,
microprobe jockey extraordinaire. For more than 30 years, Chris has
maintained and upgraded the Lab’s electron microprobe. He knows
every trick in the book and constantly invents new ones to push the
limits of analysis. Complex circuit diagrams, cryptic numerical tables,
and other papers decorate his office, while technical manuals and elec-
tronics catalogs fill his bookcases and rows of volleyball trophies line
the shelves above.
I explained the carbon analytical problem to him—one he had
never faced before. The Geophysical Laboratory probe had been used
almost constantly for decades but almost always to study rocks and
experimental run products—samples with mostly heavier elements.
But it took Chris less than a minute to figure out a possible fix. “We can
try raising the current,” he said. “That might work.” He began asking
detailed questions about the expected amount of carbon and what
other elements might be present. Then a frown. “We’re not going to be
able to use a carbon coat, are we?”
I shook my head, knowing that this was a potential deal killer. We
needed an electrically conductive coating, and that meant a metal. But
a carbon coat would mask our fossils, and most metals absorb X-rays
EARTH’S SMALLEST FOSSILS
51
so efficiently that we’d never see the low-energy carbon X-ray signal.
“Any ideas?” I asked.
Again, it took him less than a minute. “Maybe we could use alumi-
num. Just vaporize a bit of aluminum foil.” He was smiling again,
pleased at the simplicity of his solution. Aluminum, element 15, should
be light enough itself to allow most of the carbon signal through. It
was definitely worth a try.
I handed over the first of my fossil specimens, a pair of 2 × 3-inch
rectangular thin sections of 400-million-year-old plant fossils from
Rhynie, Scotland, a classic chert locality. The samples had been col-
lected many years ago as isolated flinty boulders in old stone walls; no
rock outcrop has ever been found. The precious thin sections arrived
courtesy of Kevin Boyce, a bright-eyed, soft-spoken grad student in
Andy Knoll’s group. Kevin had trolled through Harvard’s somewhat
neglected paleobotany collection as part of his thesis work on the evo-
lution of leaves. He hoped that the Rhynie samples, which preserve
cellular structures of some of the oldest known land plants, might re-
veal clues about the chemical evolution of plants.
The first step was to apply the thin aluminum coating. Our anti-
quated but serviceable vacuum coating system consists of a well-worn
metal housing about the size of a washing machine with vacuum
pumps and hoses arranged inside. A 4-inch-square platform with wire
electrodes sits on top, while vacuum gauges and control valves project
from the side.
Chris snipped a 1-centimeter-square piece of aluminum foil,
crumpled it up and placed it in a small wire basket attached to the
electrodes. He arranged the two Rhynie chert thin sections on the metal
platform, and then lowered a 2-foot-tall dome-topped bell jar so that
its rubber-lined base made an airtight seal. The old pumping system
labored for a quarter of an hour to achieve the desired vacuum, but
then it took only a fraction of a second to apply an electric current and
vaporize the aluminum foil. Aluminum atoms flew off in all direc-
tions, coating everything inside the bell jar, including our samples.
Chris released the vacuum, raised the bell jar, and the fossils were ready
for the probe.
The Geophysical Lab electron microprobe is an awkward-looking
tabletop machine that sits in its own small room. [Plate 3] Two chairs
flank the workbench, which is dominated by a 4-foot-high cylindrical
52
GENESIS
tower that looks like a model of some futuristic fortification. The tower
houses the electron gun, the heart of the probe. At the tower’s top, a
coiled tungsten filament generates electrons; a series of ring-shaped
electromagnets focus the electrons into a narrow beam as they acceler-
ate downward onto the sample. The base of the tower is cluttered with
five boxlike attachments, called spectrometers, that measure X-rays,
plus various vacuum lines, power cables, and viewing ports.
A curious combination of instrument panels controls the electron
gun hardware and X-ray detectors. To the right, a computer monitor
displays all the machine’s vital statistics—beam current, spectrometer
settings, sample position, and more. In sharp contrast, a 1980s-vintage
slant-front console, sporting two antiquated 6-inch black-and-white
video screens and more than a dozen plastic knobs reminiscent of a
classic Star Trek set, dominates the central table. Like an old house that has undergone decades of renovation, the Geophysical Lab probe has
been through a lot of upgrades.
I carefully secured one of the thin fossil sections into a shiny metal
sample holder, closed the sample port, and waited a couple of minutes
for the machine to achieve the high vacuum necessary to stabilize the
electron beam. Meanwhile, Chris fiddled with the computer controls,
raising the electron current to about ten times its normal settings. It
took him a few moments to center and focus the intense beam onto a
carbon-rich portion of our sample. We were about to discover whether
or not we could detect fossil carbon atoms.
It worked! A carbon signal of 600 X-ray counts per second stood
out sharply from the 30-count-per-second background. We were in
business. It was a simple matter to select half a dozen areas, each about
a fiftieth of an inch square, to map. Slowly but surely, the microprobe
beam scanned across the sample, measuring the carbon concentration
point by point. It took about 3 hours to produce one map. We put the
probe on automatic, happy with our rapid progress, and headed out-
side to the Lab’s
sand court for our afternoon game of volleyball.
The analytical procedures took a bit of tweaking. The initial alu-
minum coatings were too thick, the beam settings not quite optimal.
But within a few weeks, we were producing a steady stream of colorful
maps; regions rich in carbon atoms from ancient life-forms stood out
boldly against the carbon-poor fossil matrix. Cellular features less than
a ten-thousandth of an inch across were clearly visible. Armed with
EARTH’S SMALLEST FOSSILS
53
these maps, Kevin Boyce, with his sophisticated botanical eye, was able
to describe and interpret cellular detail never previously seen. [Plate 3]
Making these carbon maps, watching the fine details emerge, is
great fun. Each map is formed from a two-dimensional array of point
analyses, just like the pixels on your computer screen. We typically em-
ploy a quick 400 × 400-point array for reconnaissance, while slower
500 × 500-point arrays yield beautifully detailed maps with colors rep-
resenting the concentration of carbon—red for the highest carbon con-
tent, followed by orange, yellow, and the other spectral colors. We play
with map colors like a high-tech video game to heighten the contrast
and highlight features of special interest.
Maps of the distribution of fossil carbon atoms can be dramatic
and surprising as well as beautiful, revealing subtle cellular-scale de-
tails not previously recognized. In a sense, though, this analytical effort
is little more than an extension of past morphological studies of fossil
size and shape—a slightly more elaborate way to image the specimens.
These carbon-rich fossils preserve far more information than just the
chemical elements that make them up. That’s why, for more than three
decades, geologists have examined ancient life for fossil isotopes.
FOSSIL ISOTOPES
The fascinating discipline of atomic-scale paleontology has blossomed
primarily because all living cells perform a wonderful repertoire of
distinctive chemical tricks. Life transforms any collection of its con-
stituent atoms in subtle and surprising ways. Carbon atoms, for ex-
ample, come in two common varieties—isotopes dubbed carbon-12
and carbon-13. Every carbon atom has exactly six massive positively
charged particles called protons in its nucleus; that atomic number, 6,
is the chemical definition of “carbon.” The distinction between the
two common carbon isotopes lies in the number of neutrons, a sec-
ond kind of massive particle that also resides in the atomic nucleus.
Carbon-12 has six neutrons, while carbon-13 has seven.
The number of neutrons has no bearing whatsoever on carbon’s
chemical behavior. You could live equally well on a pure carbon-12
diet or a pure carbon-13 diet. But there is one important physical dif-
ference: Carbon-13, with its extra neutron, is about 8 percent more
massive than carbon-12. (An average-sized person whose cells were
54
GENESIS
made entirely with carbon-13 atoms would weigh about 2 pounds
more than the same person made entirely with carbon-12 atoms.) As a
consequence of this small mass difference, carbon-13 atoms are also a
little more sluggish than carbon-12 atoms when taking part in some
chemical reactions. So when living cells process carbon-bearing food,
they become slightly enriched in the lighter isotope, carbon-12. That
characteristic isotopic signature of life can be preserved for billions of
years in rock.
Analytical studies of countless carbon-bearing rocks reveal a sharp
dichotomy. Most of Earth’s carbon is locked into mineral deposits,
notably the abundant carbonates that adorn the landscape with bold
limestone cliffs and dissolve to open sublime limestone caverns.
Worldwide, this mineral-locked carbon has a well-defined uniform
ratio of carbon-12 to carbon-13 of about 99:1—the standard refer-
ence value, which is designated as 0. By contrast, living cells are invari-
ably isotopically lighter, with a higher proportion of carbon-12 than
in limestone. This difference between limestone and life arises from
chemical reactions in cells, which more readily incorporate the lighter
carbon isotope.
On the geochemist’s peculiar scale, a 1 percent deficiency of car-
bon-13 relative to standard limestone is called “–10 per mil,” a 2 per-
cent drop “–20 per mil,” and so on. Such “light” carbon in a rock sample
thus carries a negative number value and with it a strong presumption
of biological activity. Thousands of ancient fossil specimens, from
mammoth bones (about –21) to the 3.1-billion-year-old microbial
mats in South African sandstones and the fossils in Western Australian
chert (between –25 and –27), bear out this simple relationship. So does
fossil coal, the transformed remains of 300-million-year-old swamp
life, which typically ranges from –24 to –25. Carbon isotope studies of
soft-bodied fossils from the Burgess Shale, a 540-million-year-old Brit-
ish Columbian locality, made famous by Stephen Jay Gould’s Wonder-
ful Life, display a similarly narrow range of values between –25 and
–27. The conclusion: If a rock holds an ancient inventory of carbon
atoms from once-living cells, then the carbon invariably will be light,
even if all morphological signs of life are gone.
The most ancient microbial samples, which often consist of black,
carbon-rich splotches in limestone, shale, or other sedimentary rock,
have received special attention in recent years. Dozens of studies on
billion-year-old rocks from Africa, Australia, Europe, and North
EARTH’S SMALLEST FOSSILS
55
America reveal consistently negative carbon isotope values, but they
also point to a significant scatter among the dozens of known micro-
bial fossils more than a billion years old. Photosynthetic microbes,
which live on sunlight, tend to lie in the –20 to –30 range. These organ-
isms have dominated the fossil record since about 2 billion years ago,
when Earth’s atmosphere became oxygen-rich. However, many types
of more primitive single-celled organisms that live off the Earth’s
chemical energy are much more efficient at concentrating the light car-
bon isotope, carbon-12. Values as low as –50 have been found in 3.8-
billion-year-old sediments. While these differences help paleontologists
interpret the varied lifestyles of ancient microbes, all unambiguous cel-
lular fossils contain some proportion of light carbon.
It’s amazing how nerve-wracking waiting for a machine to pro-
duce a single isotope value can be. I was given my most memorable
carbon-rich sample in the summer of 2002, during a lecture tour to
Australia. A side trip to Sydney’s northwestern suburbs brought me to
the campus of Macquarie University. There, in beautiful green land-
scaped grounds, is the home of the Australian Centre for Astrobiology,
whose director is paleontologist Malcolm Walter.
Walter’s work on ancient Australian microfossils was well known
to me, and it was a delight to meet
him. He welcomed me with a strong
handshake, keen eyes, and ready smile and gave me a quick tour of the
facility, a tweed jacket his only protection from the mild Sydney winter.
My research was unknown to him, though he was well aware of the
well-funded NASA Astrobiology Institute, of which I was part. He lis-
tened as I described our various research projects, but his ears really
perked up when I recounted my collaborations at Carnegie with Andy
Knoll and his students. I outlined our procedures for mapping carbon
atoms and described some recent carbon-isotope results.
“You might be interested in this,” he said, handing me a small cloth
bag containing several black rock fragments. “It’s Strelley Pool Chert, a
new find from Trendall in Western Australia. It’s almost as old as the
Apex.”
The chance to hold, much less study, one of Earth’s oldest rocks is
a rare privilege. My response was rather pointed and less than subtle:
“We’d be happy to do the carbon work, if you’d like. I could do it next
week when I get back.” Few laboratories in Australia had the facilities
to analyze the ancient rock, while Carnegie was set up and ready to go.
A brief cloud of concern seemed to pass across Walter’s face, but
56
GENESIS
he hesitated only a moment before extracting a thumbnail-sized
sample, a tiny fraction of the valuable hoard. “Perhaps you could have
a look at this,” he said, and told me that he was eager to find out as
soon as possible whether or not the rock’s carbon was light. I slipped it
into my pocket, hardly believing my good fortune in having acquired a
piece of Earth’s earliest history. The sample would be a top priority on
my return, I assured him, and any data would be his to announce or
publish as he wished. Our conversation shifted to less scientific mat-
ters: the constant stress of raising money for the Institute, and the glo-
ries of his sheep farm in the countryside, where he spends his weekends.
For the rest of that Australian trip, the tiny, 3-ounce sample
weighed heavily on my mind, and it rose right to the top of a long “to
do” list on my return. When I need a carbon-isotope analysis at the
Carnegie Institution, I turn to Marilyn Fogel, a biologist who has
amassed an impressive arsenal of analytical hardware. Marilyn and her