Anyway, the idea that isotopes reflected different atomic masses of the same element was pretty well established by the 1920s, even before Chadwick's discovery of the neutron in 1934. In fact, Rutherford had suggested the existence of the neutron ("an atom of mass 1 with zero nucleus charge") in a lecture as early as 1920 (Rutherford, 1920), and by the 1930s the understanding of isotopes had essentially reached its present form. At a symposium on the discovery of deuterium (heavy hydrogen, 2H), Rutherford could even gently chide Soddy for complaining that deuterium hardly met his original definition of “isotope,” since its chemical properties are significantly different from ordinary hydrogen's (Rutherford et al., 1934). Rutherford pointed out that the properties depended on the relative difference in nuclear mass, and the mass ratio of deuterium to hydrogen is 2:1—far greater than any other isotope pair. Of course you'd expect bigger differences!
* * * *
And Back to the Present...
All of which is interesting, but why are isotopes useful? For lots of reasons, which run the gamut from the mundane to the very glamorous. I'll concentrate on geochemical applications, which is what I mostly know, and in which isotopic studies have become extremely important. In fact, probably most folks who call themselves “geochemists” these days are isotope geochemists.
* * * *
Age Dating—and Geochemical Evolution
Radioactive decay provided, for the first time, a direct way to determine the age of geologic materials, and “isotopic” or “radiometric” age dates are now routine. The idea is obvious: when a rock is formed, you start with a certain amount of a radioactive isotope, uranium-235, say, and then you count up how much is left—or rather, how many daughter products have accumulated.
Of course it's not that simple. Measuring the amount of daughter products is a big problem, but a bigger problem turns out to be finding a system in which you know how much pre-existing daughter product you started with. Ideally, of course, you want that to be zero. Certain minerals crystallized from a melt—i.e., those in igneous rocks—tend to be far and away the best candidates, because you know the crystal formed when the magma crystallized. Furthermore, you want a mineral that incorporates the radioactive parent atoms while excluding the daughter products. A nearly perfect example is zircon (zirconium orthosilicate, ZrSiO4). The zircon crystal enthusiastically takes up uranium (the parent) but has little affinity for lead (the daughter).
For all the caveats and cautions, with modern instrumentation and some attention to the geologic setting, the technique works extremely well. You can even get rock samples age-dated commercially! The “traditional” systems use the decay of uranium into lead (with the added feature that the two uranium isotopes allow additional internal consistency checks, because of their very different half-lives).5 Other routine systems are the already-mentioned decay of 40K to 40Ar, and of 87Rb to 87Sr (half-life of some 49 billion years). In addition, though, over the last few decades a number of new long-lived decays have joined them: samarium-neodymium (147Sm-143Nd, t1/2 ~ 100 billion years), lutetium-hafnium (176Lu-176Hf, ~ 40 billion years), and rhenium-osmium (187Re-187Os, ~ 40 billion years). That such extraordinary long-lived decays are now usable is more evidence of the increasing precision of the measurements. After all, with a half-life of 100 billion years, out of a thousand 147Sm nuclei present at the formation of the Earth, 970 are still with us.
Also in the last few decades, the slow change over geologic time in the isotopic composition of a daughter element, as it's added to the original complement from the decay of its precursor, has come to be widely used for the large-scale tracing of geologic processes. For example, over time terrestrial Sr becomes richer in 87Sr, as 87Rb decays away. Hence low, or “primitive,” 87Sr ratios mean a rock body has been geochemically isolated for much of geologic time.
This insight was strikingly applied in lunar geology. The lunar highlands—the ancient, light-colored, and heavily cratered terrane—contains a great deal of calcium feldspar (plagioclase). Strontium, as you might expect, tends to get concentrated in calcium minerals, because it's the chemical cousin of calcium in the periodic table. And the strontium from the lunar highlands has a very primitive 87Sr value. We knew the highlands were relatively old just from the density of craters on them, but this showed they're very old indeed, dating from not long after the accretion of the Moon itself.
* * * *
"Cosmogenic” Nuclides
We don't even always have to stick with isotopes long-lived enough to have persisted since the formation of the Earth for age-dating. Several much more ephemeral isotopes are also useful. Most famous is carbon-14, with a half-life of merely 5,570 years. It's formed continuously by the action of cosmic rays on atmospheric nitrogen. A high-energy incoming proton kicks a neutron out of nitrogen-14, the most common nitrogen isotope, to make 14C. In turn, that 14C becomes incorporated into atmospheric carbon dioxide, which is taken up by plants. Then animals eat the plants, so that all living things contain a minute amount of 14C.
Once the organism dies, though, the 14C decays away—and the amount left measures how long's elapsed since the organism's death. So in principle we have a way to date organic matter—provided, of course, that it's young enough that all of the 14C hasn't decayed away completely.
And also provided, of course, that we know how much there was to start with! Contamination by “dead” carbon (i.e., ancient carbon devoid of 14C) can yield anomalously old ages. If you naively date the plants growing in a limestone sinkhole, for example, you'll find they have apparent ages of several thousand years! They've picked up dead carbon from the calcium carbonate making up the limestone.
More interestingly, the amount of 14C, and of other so-called “cosmogenic” nuclides, changes with the cosmic ray flux—and that in turn can reflect changes in the Earth's magnetic field, the Sun's magnetic field, or even the galactic environment. This is another promising field of study—although, of course, it means that an independent method of determining ages is needed!
Carbon-14's the most famous, but several other cosmogenic nuclides are now finding use for age-dating and geochemical tracing: beryllium-10 (t1/2 ~ 1.5 million years) and chlorine-36 (t1/2 ~ 300,000 years), to name two. The systems are not nearly so clear-cut, and the quantities involved are much smaller, so not only is measurement more difficult, but the results are less straightforwardly interpreted.
Of course, as I've already implied, mass specs have come a long way since the days of Thomson, so that we routinely measure isotopes at concentrations orders of magnitude below what he could detect. The latest twist along this line is to use old particle accelerators as extremely high-powered mass specs, in so-called “accelerator mass spectrometry,” or AMS. AMS has extended the practical limit of 14C ages back to 60,000 years or so—roughly 10 half-lives. (The 14C atoms are so rare it's far more practical just to count them, in a mass spec, rather than to try to detect their decay!) And it makes the use of such species as 36Cl or 10Be even conceivable: for example, currently AMS can detect five atoms of 36Cl in 1016 atoms of stable chlorine (e.g., Jannik et al., 1991).
At these exceedingly low concentrations, mass discrimination also becomes a key issue, because there are interfering ions, typically molecules, with the same mass numbers but slightly different masses. 13CH+, for example, also has mass number 14—the same as 14C+. Another advantage of AMS is that such ions are largely broken up by the high energy of the beam, so they're not so much of a problem.
Again, it keeps coming down to how well can we measure.
* * * *
Extinct Nuclides
We've also found traces of geologically ephemeral nuclides in primordial solar-system material. Of course, the processes ("nucleosynthesis") that make the elements don't just make the stable nuclei: they make all nuclei, the amounts of each depending on such things as the temperature, particle flux, cross-section6 for reaction, and so on. And they include nuclides with half-lives up to a few tens of millions of years or so�
�long enough that some were still around when interstellar debris started accreting into the Solar System.
The first such “extinct” nuclide discovered was iodine-129 (t1/2 ~ 16 m.y.), which was fingerprinted by anomalous amounts of its decay product xenon-129 in certain meteorites (Reynolds, 1960). Xenon is a member of the noble, or “inert” gases; it's used to fill xenon flashtubes for strobe lights. Under normal conditions, it does not form chemical compounds. Iodine, however, is chemically reactive and will get incorporated into meteorites. Once it's there, even when it decays into xenon, the xenon remains trapped in the rock. Hence, meteorites coalesced within a few I-129 half-lives of the formation of the elements in the Solar System, because some 129I was still around.
Since then the spoor of many other extinct nuclides have been found, all with half-lives from around a million to a few tens of millions of years. One of the most interesting is aluminum-26 (26Al), with a half-life of around a million years. After many fruitless searches, its decay product magnesium-26 was found in aluminum minerals in some extremely primitive meteorites, in what appear to have been molten droplets of calcium aluminum silicates. Even more important, though, is that 26Al releases a lot of energy when it decays. The decay of uranium, thorium, and 40K keep the Earth tectonically and volcanically active from the heat they release (see “Refueling a Rundown Planet,” Analog, August 1991). This works fine on a planet the size of the Earth, but a small body like an asteroid cools off much too efficiently to get stirred up by the heat produced by these long-lived radionuclides. With a smidgen of 26Al, though, even asteroid-sized bodies would have melted, back when the Solar System was formed. This probably accounts for iron meteorites: when a body consisting of a haphazard mixture of planet-stuff melts, it chemically differentiates. That is, the rock floats to the outside, and the metal sinks to the inside to form a core. Later, as asteroids collide with each other over time, these differentiated bodies get broken up, to yield pieces both rich in iron metal and in rock. Take note, all you would-be asteroid miners!
One nuclide isn't even quite extinct. Plutonium-244 has a half-life of some 82 million years. (Note that this is a different isotope from the 239Pu used in bombs.) Since the Earth (and Solar System) are some 4.5 billion years old, about 55 half-lives of 244Pu have elapsed since the Solar System formed. That means there should be roughly 2-55 of the Solar System's original complement of Pu-244 left—or about 30 atoms in a sextillion. Not much—but surviving Pu-244 was identified in the early 1970s (Hoffman et al., 1971).
It's an astounding achievement to have identified such a few atoms. Finding a needle in a haystack is trivial by comparison.
Once again, how well can we measure?
* * * *
Fractionation: The Separation of Isotopes
Remember when they said that isotopes have “identical” chemical properties? They lied. They have almost the same chemical properties. As Rutherford pointed out to Soddy, the mass difference does have slight effects, and those effects get bigger as the proportional difference in the mass get bigger. Moreover, the effects can be measured in real-world systems, and they can tell us a lot.
For one thing, molecules containing lighter isotopes evaporate somewhat more readily, as you might expect. Such “kinetic fractionation” is how isotopes are purified commercially, but it also happens naturally: rainwater, for example, is “lighter” (i.e., depleted in deuterium and 18O) than the water from which it evaporated. In fact, the farther inland you go, the lighter the rain gets. (Isotopically, that is!) As the rain soaks into the ground, though, it gets a bit heavier as some of it evaporates back into the atmosphere. Furthermore, the longer groundwater is in the ground, the heavier it tends to get, as the water molecules tend to exchange hydrogen and oxygen with hydrous minerals.
This means that groundwater derived from recent precipitation is noticeably lighter than groundwater that's been there a while. This seems cute but academic until you realize that now it's possible to distinguish groundwater of different origins! Hence a well that is, say, largely drawing from a nearby river instead of a deep aquifer can be identified; not an idle consideration in the arid West with its tangle of water law. (The lawyers are circling....)
The binding energy of the more massive isotope is also slightly greater, and this energy difference also leads to fractionation, since molecules containing the lighter isotope are just slightly more ready to react. For example, living things are routinely lighter isotopically than the background values. With the seething cascades of chemical reactions in a living organism, those slight reactivity differences build up. This “isotopic signature” is how we can infer, for example, whether carbon is probably of biological origin, simply because if it is, it will be significantly depleted in carbon-13.
For another example, isotopically light sulfur, mined from the top of salt domes in the Gulf Coast, is also of biological origin. It results from bacterial activity. More generally, isotope studies are commonly able to constrain the source of sulfur in ore deposits. Since a great many ore minerals are sulfides, this is another application with real-world relevance—and it's all related to how well we can measure!
* * * *
Stardust
Physico-chemical separation is always proportional to the mass difference. So, for example, during water evaporation or biological metabolism the fractionation of 17O, with respect to 16O, is always half that of 18O vs. 16O. Nuclear processes, however, can give utterly different isotopic ratios, because it's the nuclear properties that matter. The original example is from radioactive decay itself, with the resulting build-up of daughter products over time.
Another example, though, is nucleosynthetic processes. Heavy nuclei can be made in a variety of different stellar furnaces (see “Forging Planet Stuff,” Analog, November 1990), and those different origins yield quite different isotopic mixtures. Solar System material, for example, seems to be dominated by nuclei forged in a hot, Type I supernova, which may have had serious consequences for the evolution of life, intelligence, and technology on Earth (see “Carbonosis,” Analog, March 1993).
Solar System matter overall is homogeneous isotopically, suggesting that it was pretty well stirred up in the nebula from which the planets and Sun formed. In the last few decades, though, tiny grains have been found in the most primitive meteorites that have wildly unearthly (or rather, un-Solar-Systemly) isotopic ratios that also can't be explained by geochemical fractionation. In particular, micron-sized grains of silicon carbide (SiC, the mineral moissanite) have anomalous ratios of both silicon and carbon.
What are they? They're probably truly stardust: the silicon carbide motes most likely condensed in a red giant's atmosphere just like ice crystals in Earth's, and then were swept up by the stellar wind and wafted out to interstellar space, where they eventually wound up in the raw material that became our Solar System. Presumably these are just the tiny fraction of grains that, through happenstance, managed to survive the mixing in the nebula. (Ironically, the grains weren't found sooner because SiC, under the trade name Carborundum, is routinely used for rock polishing! So the exotic SiC was lost in the noise.)
Since then, other such “pre-solar” minerals have been found, most also inferred to be of red-giant origin. See Ott (2001) for a technical review, if you're interested. Another consequence of this recognition of exotic material is that we no longer chauvinistically refer to “cosmic” isotopic composition; now more modestly we refer to “Solar System” composition.
Of course, if you could gather together macroscopic quantities of these grains, the Si and C would have noticeably different atomic weights even as measured by the traditional wet-chemical methods. The atomic weight of terrestrial silicon is 28.0855; that of carbon is 12.011, since both elements are dominated by their lightest isotopes 28Si and 12C, respectively. However, the atomic weights of the SiC-mote elements are going to be more like 29 and 12.5, because both silicon and carbon are strongly enriched in their heavy stable isotopes (29Si and 30Si, and 13
C, respectively). Imagine giving a sample of that stuff to a nineteenth-century chemist, or someone at an equivalent technological level! Maybe there's a story or two there ....
* * * *
Exotic Nuclei and Massive Particles
And finally, there's a notion kicking around that highly massive particles might be left over from the Big Bang. If such “Cahn-Glashow” particles are indeed stable enough to still be around, and if they have a negative charge, they will tend to glom onto an ordinary nucleus by simple electrostatic attraction. The resulting composite nucleus will look like an exceedingly heavy isotope of the element with one less proton (i.e., the next lower atomic number). Searches for such supermassive nuclei have been made (e.g., McInteer & Mills, 1991), so far with negative results—but if, say, there were only a thousand or so dispersed among the 1024-odd molecules in a glass of water, that's still far below the present detection limits.
* * * *
New Measurements —-] New Physics?
So perhaps we've come full circle. In the early twentieth century, measuring isotopic ratios lay at the very cutting edge of technology, and those cutting-edge measurements led to brand new science—science that is now routine. Now, with even better measurements, perhaps there's more new physics to be found. n
Copyright © 2010 Stephen L. Gillett, Ph.D.
* * * *
References
Hoffman, D.C.; Lawrence, F.O.; Mewherter, J.L.; Rourke, F.M., Detection of plutonium-244 in nature, Nature, 234, 132-4, 1971.
Jannik, Nancy Olga; Phillips, Fred M; Smith, George I; Elmore, David, A 36Cl chronology of lacustrine sedimentation in the Pleistocene Owens River system, Geol.Soc.Amer.Bull., 103(9), 1146-1159, 1991
McInteer, B.B., Mills, T.R., Superheavy isotope enrichment using a carbon isotope enrichment plant, Sep.Sci.Technol., 26(5), 607, 1991.
Analog SFF, March 2010 Page 9
