Radiometric dating is a method which scientists use to determine the age of various specimens, mainly inorganic matter (rocks, etc.), though there is one radiometric dating technique, radiocarbon dating, which is used to date organic specimens.
How do these dating techniques work? Basically, scientists take advantage of a natural process by which unstable radioactive “parent” isotopes decay into stable “daughter” isotopes spontaneously over time.
Uranium-238 (U238), for example, is an unstable radioactive isotope which decays into Lead-206 (Pb206) naturally over time (it goes through 13 unstable intermediate stages before it finally stabilizes into Pb206).
In this case, U238 is the “parent” and Pb206 is the “daughter.”
Scientists begin by measuring how long it takes for a parent isotope to decay into a daughter isotope. In this particular case, it takes 4,460,000,000 years for half of a sample of U238 to decay into Pb206. It takes another 4,460,000,000 years for half of the remaining sample to decay into Pb206 and then another 4,460,000,000 years for half of what’s then left to decay, and so on.
The time it takes for half of a sample to decay is called a “half-life.”
By measuring radioactive half-lives and measuring how much parent and daughter are present in any given specimen, and by making certain key assumptions, scientists believe they are able to accurately determine the age of a specimen. The measurements involved can be quite accurate. The question is what are the underlying key assumptions and how reliable are they?
The three key underlying assumptions are 1.) The rate of decay of parent into daughter has remained constant throughout the unobservable past; 2) the specimen which we are examining hasn’t been contaminated in any way (that is, no parent or daughter has been added or taken away at any point during the unobservable past), and 3) we can determine how much parent and daughter were present at the beginning of the decay process – not all of the Pb206 present today necessarily came from decaying U238;
Pb206 may have been part of the original constitution of the specimen. If any of these assumptions are wrong, the method cannot accurately determine the age of a specimen.
While the second and third assumptions have always been a bit troublesome, especially the third assumption, which considers the original constitution of a particular specimen, the first assumption was thought to be a pretty safe bet since scientists were not able to vary the decay rates much in a lab. Recently, however, new research has revealed that the decay rates may have been drastically different in the unobservable past. This calls the whole method into question.
When uranium decays to lead, a by-product of this process is the formation of helium, a very light, inert gas which readily escapes from rock. Certain crystals called zircons, obtained from drilling into very deep granites, contain uranium which has partly decayed into lead. By measuring the amount of uranium and ‘radiogenic lead’ in these crystals, one can calculate that, if the decay rate has been constant, about 1.5 billion years must have passed. (This is consistent with the geologic ‘age’ assigned to the granites in which these zircons are found.)
There is a significant amount of helium from that ‘1.5 billion years of decay’ still inside the zircons.
This is at first glance surprising for long-agers, because of the ease with which one would expect helium (with its tiny, light, unreactive atoms) to escape from the spaces within the crystal structure. There should be hardly any left, because with such a slow buildup, it should be seeping out continually and not accumulating. …Results show that because of all the helium still in the zircons, these crystals (and since this is Precambrian basement granite, and by implication the whole earth) could not be older than between 4,000 and 14,000 years. In other words, in only a few thousand years, 1.5 billion years’ worth (at today’s rates) of radioactive decay has taken place.
The point is that radiometric dating is not the sure thing that it has been made out to be over the last century. There still remains a lot of research to do, but, as it currently stands, the accuracy of radiometric dating remains ambiguously suspect at best.