Part 3
Of all the elements that are part of the useful parent-daughter systems, only potassium is common enough to be analyzed by conventional chemical techniques. All the other elements, especially the radiogenic ones, are present in such small quantities that special processes had to be developed to measure them. The most valuable and generally used process is called ISOTOPE DILUTION.
Isotope Dilution
This is a process for analyzing an unknown material by incorporating uniformly into it a small amount of a radioactive test substance and determining how much the tracer radioactivity is altered by dilution in the original material.
It works like this: Let’s say that we have an unknown number of atoms, _x_, of a given element. The normal isotopic composition of this element is accurately known, as it is for most elements, and the ratio of two of its isotopes can be expressed as A/B. We now add to _x_ a known (but usually smaller) amount, _c_, of the same element. This quantity has a drastically different isotopic ratio, A′/B′. We mix _x_ and _c_ thoroughly together. The ratio A′/B′ can have almost any value, but must be different from A/B and we must know exactly what it is. (There are many ways of determining this chemically, or we can use a sample isotope of known composition obtained from the U. S. Atomic Energy Commission’s Oak Ridge National Laboratory at Oak Ridge, Tennessee.) The substance added is known colloquially as the SPIKE.
After the original material and the spike are thoroughly mixed we have:
_x_(A/B) + _c_(A′/B′) = (_x_ + _c_) (A″B″)
in which A″/B″ will be the ratio of the two isotopes in the mixture. With this information in hand, we can perform any chemical purification or transfer process with the material (see photo on page 22), without having to worry about loss. (Even if 90% of the material should be lost in some operation, the isotopic composition would not be changed, and that is all we are interested in.) Now we can place the material containing the isotopic mixture in a MASS SPECTROMETER, which will determine the ratio A″/B″. When we have that, we may substitute the value of A″/B″ in the equation and quickly calculate _x_, the unknown concentration of atoms in the original sample.
Mass Spectrometry
The mass spectrometer measures isotopic abundances using a magnetic field to sort electrically charged particles into groups according to their masses. It works this way: A small drop of material to be analyzed is placed on a metal filament and dried. The filament, in its holder, is placed inside the mass spectrometer, and heated electrically in a vacuum, like the filament in a light bulb. As the wire begins to glow, some of the sample begins to radiate, or “boil off”, losing an electron or two in the process. In other words, some of the atoms will be changed into positive IONS.
An alternative method is to introduce the sample material into the vacuum chamber in the form of a gas (like argon, for example), and then bombard the gas with electrons streaming from a hot filament. The electron stream will knock some electrons off the gas molecules and this also will produce positive ions. Either process of ion production is satisfactory, depending on the problem to be tackled, but the mass spectrometers for the two methods are naturally quite different.
Whichever way the ions were produced, they are next exposed to a strong electric field, accelerated, and electrostatically focused into a beam. These charged particles are directed into a magnetic field between the pole faces of an electromagnet. The magnet does the analyzing by the principle of magnetic deflection that was known to André Ampere and Michael Faraday more than a century ago. Any moving electric charge has a magnetic field associated with it. This field interacts with the field of the analyzing magnet to impress a deflecting force on the charge. The force acts at right angles to the direction the charge travels and also at right angles to the direction of the impressed magnetic field. The pull of this force depends only on the electric charge and the speed of each particle: A light single-charged particle will be deflected more than a heavier particle with the same charge. In this way, the ions in the beam are sorted out into a number of separate beams, each made up of particles of the same charge/mass ratio. Each beam contains one isotope of the original material, because isotopes differ on the basis of their mass. By adjusting the current in the electromagnet we can direct these separate beams into a “collector” and electrically measure their intensity one by one. This gives the relative abundance of the separate isotopes in the sample.
Minerals That Can Be Dated
Measuring age by one of the long-lived radioisotopes requires a closed system. Usually this is some kind of crystal formed in a period of time that is short, compared to the time that has elapsed since, and that has remained unchanged since it formed. Specifically, neither the parent isotopes can have been added nor the daughter isotopes removed by any process other than radioactive decay.
The earth is a dynamic system, however. Things are always changing and moving—not very rapidly, perhaps, but fast enough, in geologic time, to raise mountains and shift oceans. Solutions are moving around, dissolving something here and depositing it again somewhere else. Temperatures are changing as one place is denuded by erosion and another area buried under layers of sediment. Under such conditions, few systems remain closed. It is perhaps surprising that we find any closed systems at all. Let us look at a few that are known to be reliable. (They are listed in Table I on page 4.)
Potash Feldspar
In the early 1950s, when the potassium-argon (parent-daughter) method was being developed by scientists at the University of Chicago, it was thought that the potash-bearing variety of the mineral feldspar would be an ideal closed system, because it was usually optically clear and free of flaws. This widely shared, logical, and perfectly scientific deduction soon turned out to be quite wrong. The scientific workers discovered that when feldspar and mica from the same rock (and thus of the same age) were analyzed side by side, the mica always came out older! Investigation showed that feldspar “leaked” argon (lost some of its radiogenic argon) even at room temperature, but the mica retained all or nearly all of the argon that had been generated in it.
Mica
With the development of the rubidium-strontium (parent-daughter) method by L. T. Aldrich and his co-workers at the Carnegie Institution of Washington, came the realization that mica was also very useful for this analysis, for it usually contains ample rubidium and not much original strontium that would mask the presence of the radiogenic strontium. As a result, mica, especially black mica (the mineral biotite), has enjoyed great popularity as a good and easy-to-find closed system.
Everything has its limits, and mica is no exception: Even mica tends to leak argon at elevated, but still relatively low (geologically speaking), temperatures. These effects also depend on pressure and other factors, not all of which are well known; these elevated temperatures, pressures and other conditions of course act to some extent on all rocks buried in the earth’s crust. It is known that at only about 300°C at moderate pressures argon is leaked from mica faster than it is being generated in it by the decay of radioactive potassium. The temperature needed to cause the rapid loss of strontium from mica is not much higher. Mica, especially biotite, will recrystallize and lose all its radiogenic constituents (argon and strontium) at temperatures where many other minerals show little or no change.
That means that we cannot always rely on mica to give the date of the _original_ crystallization of a rock—the time when it cooled from a molten state. Instead, mica will tell us when the rock _last_ cooled from, say, several hundred degrees centigrade, regardless of what may have happened to the rock before that. The mica may have been reheated as a result of being buried under a few miles of sediment, for example. The mica will show when the rock last cooled—in other words, when it came up again.
Low-Strontium Feldspar
In spite of early disappointments with potash feldspar for argon dating, some of it is useful for rubidium-strontium procedures. It all depends on how much original strontium the potash feldspar contains. Most feldspars, unfortunately, contain far too much, but rapid screening by X-ray fluorescence or flame photometry methods can weed these out and identify specimens low enough in original strontium to be useful. Otherwise, feldspar is an excellent closed system for rubidium and strontium; it remains closed even at temperatures high enough to melt many other minerals. It is not affected at all by the same degree of heating that will drive argon out of biotite. The rubidium-strontium age of feldspar usually comes close to the time of original crystallization of the rock.
Obviously, here is a geologically important tool. If we find feldspar and biotite in one rock, and if feldspar, tested by the rubidium-strontium method gives the same age as biotite tested by potassium-argon decay, then we can say with confidence that the rock has not been reheated since shortly after it crystallized. Conversely, if the biotite comes out much younger than the feldspar, we can be sure that something _has_ happened to this rock long after it first crystallized. Such information is not only valuable to pure science—it can also be useful in locating areas favorable for ore prospecting and in other practical ways.
Zircon
Another very interesting mineral is zircon (a silicate of zirconium), one of the accessory minerals found in small quantities in many crystalline rocks. Zircon usually occurs in very small grains and is heavy and hard, so that it can be separated from the other rock without much difficulty, even though it may take 100 pounds of rock to supply a gram of zircon.
Zircon usually contains a fair amount of uranium and very little lead. It holds radiogenically produced lead well, even at relatively high temperatures. But that is not all. Even if some of the lead is lost, there is a mathematical way of correcting for it. This technique is called CONCORDIA ANALYSIS and was developed by G. W. Wetherill, a physicist then at the Carnegie Institution of Washington. It is based, again, on the fact that natural uranium has two long-lived isotopes—²³⁸U and ²³⁵U—and that the lighter one, ²³⁵U, decays faster than the heavier. The daughter products of both uranium decay processes are isotopes of the same element, lead—²⁰⁸Pb and ²⁰⁷Pb, respectively. Heavy isotopes are not separated to any significant degree by chemical processes, so that if radiogenic lead has been lost from a system for any reason, the other lead isotopes also will have been lost in whatever proportion they were present originally.
If we plot a graph of the radiogenic ²⁰⁶Pb/²³⁸U ratio against the radiogenic ²⁰⁷Pb/²³⁵U ratio for concordant (closed) systems _of all ages_, we obtain the curved line shown in the figure below. The curve is the locus of _all_ concordant U-Pb ages and is called _Concordia_. Then if we test two or more particular zircons of the same age that have lost different amounts of lead, at about the same time, the plot of their ²⁰⁶Pb/²³⁸U ratios against their ²⁰⁷Pb/²³⁵U ratios will fall on a straight line that is a chord of the Concordia curve. The upper intersection of this chord with the curve then will mark the true age of the zircons. This is an elaborate technique utilizing difficult chemical procedures, but it has proved invaluable in solving some important geologic problems.
Hornblende
The mineral hornblende provides another useful system. Hornblende is a complex silicate of sodium, calcium, iron, magnesium, and aluminum, and usually contains a few tenths of 1% of potassium. It is unusual in that it tenaciously retains its radiogenic argon, even at relatively high temperatures.
Sanidine
Still another good system is the rare feldspar, sanidine, which is excellent for both potassium-argon and rubidium-strontium age determination. Sanidine usually is found in volcanic ash falls and has been important in the establishment of the geologic time scale, as we shall see.
Whole Rocks
Finally there is still another way of obtaining a closed system by using the whole rock, not just a crystal of a single mineral within it. A large body of granite or similar rock may contain a number of minerals, some or none of which may be closed systems. Yet as long as this body of rock remains impermeable to solutions (which in nature means mostly to water), no substance will be able to move very far in it because diffusion in solids is so slow. Consequently it will remain a closed system, as a whole, regardless of what happens to the individual mineral grains.
If we take a piece from near the middle of this body of rock and if this piece is much larger than the largest constituent grain in it, then we have a fair sample of a closed system—the whole rock. The only difficulty arises from the fact that few rocks are sufficiently impermeable to solutions to retain argon, and many rocks contain so much common strontium that rubidium-strontium analysis is impractical. Still, we can use the rapid survey methods as for feldspar, selecting the few rocks that would be useful. This work has been done frequently, and the results have been fruitful for rubidium-strontium analysis. The whole-rock rubidium-strontium age dates the time when the rock became impermeable.
SOME INTERESTING RESULTS
The Old Man from Olduvai
One of the most talked-about age measurements in recent years was the determination of the unexpectedly great age of fossil ancestors of man, found by the British anthropologist, Dr. L. S. B. Leakey, in Olduvai Gorge in Tanzania. The measurements were made by Garniss H. Curtis and Jack F. Evernden at the University of California in Berkeley by the potassium-argon method. The age came out a little less than 2 million years, about twice as old as it “should be” in the view of many scientists. Human remains of such great antiquity had never been found before, and much doubt was raised about the validity of the figures.
Time periods as short as two million years are not easy to measure by potassium-argon. The amount of argon produced in that time is extremely small, and contamination by argon from the air is a serious problem. Still, the measurements were repeated, the rocks were studied again, and the result did not change: The fossils were still about 2 million years old.
In cases like this, one tries to find some other method to check the results in an independent way. After many attempts it was discovered that the same rock strata dated by potassium-argon also contained some pumice—a porous volcanic glass—and that this glass was suitable for uranium fission-track dating. The measurements were made in the General Electric Research Laboratory. What was the result? Just about 2 million years!
When such altogether different techniques give the same number, one can have some confidence that the number is exact. It would be difficult to imagine a disturbance in nature that would cause these unrelated methods to give the same wrong number—in both cases by a factor of two. The double check simply means the Olduvai man is 2 million years old. There is not much doubt about it.
The Geologic Time Scale
Much of historical geology is based on a relationship called the LAW OF SUPERPOSITION. This simply means that when some rock formation was placed on top of some other formation by natural processes (sedimentation or volcanic eruption, for example), the layer on top must be younger than the one on the bottom. Such a conclusion may now seem obvious, but the concept was not even expressed until the very end of the eighteenth century and was still a matter of scientific controversy when Abraham Lincoln was a boy. It was the law of superposition, however, that led the early geologists to establish the first geologic time scales and to realize the enormous extent of geologic time.
In essence the system of establishing age by this concept is this: Somewhere a large and easily recognized layer of sedimentary rock was known. It had a characteristic color, texture, gross composition, and overall appearance. Let us call it _bed M_. This bed could be traced across the countryside until a place was reached where it could be seen that _bed M_ rested on another, different layer of rock, which we might call _bed L_. _Bed L_ could also be traced some distance and ultimately could be observed resting on a still different stratum, which we shall call _bed K_. Some of these beds had fossils in them, and it was eventually realized that rocks with the same kinds of fossils are of the same age, even though they may differ in other respects—in color or composition, for instance.
If _bed M_ was followed in another direction, perhaps a point was reached where it dipped down a little, and here, was found still another layer—call it _bed N_—on top of _M_. Obviously, the sequence of beds from oldest to youngest was _K-L-M-N_. Their relative ages were now established. Over the years, hundreds of geologists described various rock layers and identified the fossils in them. By the middle of the nineteenth century, this “layer-cake” structure of sedimentary rocks was well-known in western Europe. (One may say, parenthetically, that America was geologically a vast unknown at that time. Today we know much more about the geology of Antarctica than anyone a hundred years ago knew about the geology of the United States.)
A system of nomenclature for the “layer-cake” was developed and refined. Gradually this nomenclature was accepted internationally. Long-range correlations between beds of the same age, distant halfway around the globe from each other, were made possible as the science of PALEONTOLOGY developed. The relative age of almost any rock containing even poorly preserved fossils could be determined anywhere in the world with precision. That is, the age of rock layers in relation to one another was known. But the real age—the absolute age—remained unknown until knowledge of radioactivity provided the necessary clocks.
Even this process wasn’t entirely without problems. The difficulty lay in the fact that ordinary sedimentary rocks (shale, sandstone, and limestone) cannot be dated by the usual nuclear methods because they present no suitable closed systems. Only some volcanic sediments can be reliably dated by the mica, feldspar, and zircon they contain; but these ancient ashfalls are rare, usually are only a few inches thick, and are not easy to identify at the surface because they weather quickly to clay. Only about a dozen volcanic beds have been accurately dated in North America. The time points they established are mainstays of the present geologic time scale, but they have had to be supplemented by indirect information.
The most important of these indirect time points are furnished by what geologists call BRACKETED INTRUSIVES. It often happens in geologic history that a mass of rock becomes molten at a great depth and forces its way up through several layers of sedimentary rocks. The sedimentary layers are usually bent and twisted (folded) by the upthrust, and where the cooler rock comes in contact with the molten mass, cooler material is burned (recrystallized). Geologists call that process CONTACT METAMORPHISM because the sedimentary rock forms are changed, or metamorphosed, to have another form or composition. Contact-metamorphosed rocks, in spite of the damage they have suffered, may contain recognizable and accurately datable fossils. Thereby the metamorphic rock establishes a lower limit for the age of the intrusion: It must be younger than the fossils in the youngest of the metamorphic rocks it touched.
Now on top of the intrusive rock, we may find another sedimentary rock, deposited on top of the intrusive after it cooled and was exposed by the erosion of overlying materials, perhaps millions of years later. This new sediment may also contain fossils and thus furnish an upper limit for the age of the intrusion. The measured age of the intrusive rock thus can be used to set upper and lower limits on the absolute ages of the two sediments. If we are lucky, the two sediments will bracket a relatively short interval, making our measurement quite precise.