Radiometric Dating Methods

Sean D. Pitman M.D.

© July 2008

History of Radiometric Dating

At the time that Darwin's On the Origin of Species was published, the earth was "scientifically" determined to be 100 million years old. By 1932, it was found to be 1.6 billion years old. In 1947, science firmly established that the earth was 3.4 billion years old. Finally in 1976, it was discovered that the earth is "really" 4.6 billion years old… What happened? ⁴

The study of geology grew out of field studies associated with mining and engineering during the sixteenth to nineteenth centuries.  In these early studies the order of sedimentary rocks and structures were used to date geologic time periods and events in a relative way.   At first, the use of “key” diagnostic fossils was used to compare different areas of the geologic column.   Although there were attempts to make relative age estimates, no direct dating method was available until the twentieth century. 

Following the discovery of radioactivity by Becquerel (1896), the possibility of using this phenomenon as a means for determining the age of uranium-bearing minerals was demonstrated by Rutherford (1906).  In his study Rutherford measured the U and He (He is an intermediate decay product of U) contents of uranium-bearing minerals to calculate an age.  One year later Boltwood (1907) developed the chemical U-Pb method. These first “geochronology studies” yielded the first “absolute ages” from geologic material, which seemed to indicate that parts of the Earth's crust were hundreds of millions of years old. (Boltwood's ages have since been revised).

During this same period of time Thomson (1905), Campbell and Wood (1906) demonstrated that potassium was radioactive and emitted beta-particles.  The first isotopes of potassium (³⁹K and ⁴¹K) were reported by Aston (1921).  Kohlhorster (1930) reported that potassium also emitted gamma radiation. Following theoretical arguments by Klemperer, Newman and Walke (1935) on the existence of ⁴⁰K, which radioactively decayed to ⁴⁰Ca by beta-emission, Nier (1935) discovered ⁴⁰K and reported a value of 8600 for the ³⁹K/⁴⁰K ratio.  Newman and Walke also suggested the possibility that ⁴⁰K could decay to ⁴⁰Ar.  However, it was Von Weizsacker's (1937) argument, based on the abundance of argon in the Earth's atmosphere relative to the other noble gases (He, Ne, Kr, and Xe), that ⁴⁰K also decayed to ⁴⁰Ar by electron capture.  As a test, Von Weizsacker suggested looking for excess ⁴⁰Ar in older K-bearing rocks. By combining Von Weizsacker’s argon abundance arguments with Kohlhorster’s observation that potassium emitted gamma-radiation, Bramley (1937) presented strong evidence that potassium underwent dual decay.  Thompson and Rowlands (1943), using a cloud chamber, confirmed that ⁴⁰Ar was the decay product of ⁴⁰K undergoing electron capture.  The absolute confirmation that ⁴⁰Ar was the decay product of ⁴⁰K came when Aldrich and Nier (1948) measured significantly increased ⁴⁰Ar/³⁶Ar ratios on argon extracted from potassium-rich minerals relative to the atmospheric ⁴⁰Ar/³⁶Ar ratio.  The rapid development of the K-Ar dating method soon followed.

The ⁴⁰Ar/³⁹Ar variation of K-Ar dating grew out of iodine-xenon dating studies of meteorites by Jeffery and Reynolds (1961).  In these studies the isotopic ratios of all the noble gases (He, Ne, Ar, Kr, and Xe) of neutron-irradiated meteorites were measured.  This led to the discovery of ³⁹Ar, which is derived from ³⁹K by Merrihue (1965). The first ⁴⁰Ar/³⁹Ar dating results were presented in a paper by Merrihue and Turner (1966). Further development of the ⁴⁰Ar/³⁹Ar method by Mitchell, (1968), Brereton, (1970), and Turner, (1971) evaluated the interfering argon isotopes derived from potassium and calcium (³⁶ArCa, ³⁹ArCa, and ⁴⁰ArK) and determination of the respective correction factors [ (³⁶Ar/³⁷Ar)Ca, (³⁹Ar/³⁷Ar)Ca, and (⁴⁰Ar/³⁹Ar)K].  The first applications of the ⁴⁰Ar/³⁹Ar dating method of terrestrial rocks compared total fusion ⁴⁰Ar/³⁹Ar ages with conventional K-Ar ages (Mitchell, 1968; Dunham et al., 1968; York and Berger, 1970; Dalrymple and Lanphere, 1971).

It is felt that the ⁴⁰Ar/³⁹Ar dating method offers a significant advantage over the conventional ⁴⁰K/⁴⁰Ar dating technique for several reasons.  However, the most significant advantage of the ⁴⁰Ar/³⁹Ar dating method over the conventional ⁴⁰K/⁴⁰Ar method is the ability to step-heat samples to higher and higher temperatures until the sample is fused, and calculate and ages for each step.  The ⁴⁰Ar/³⁹Ar step-heating method provides information on the internal distribution of potassium relative to argon.  The first ⁴⁰Ar/³⁹Ar step-heating studies of terrestrial samples were by Fitch (1969), Miller (1970), York (1971), Lanphere and Dalrymple (1971), and Brereton (1972).¹

Assumptions

Dating rocks by radioactive timekeepers is simple in theory, but almost all of the different methods (except for the isochron methods - see below) rely on these few basic assumptions: ²¹

• Beginning Conditions Known

• Beginning Ratio of Daughter to Parent Isotope Known (zero date problem)

• Constant Decay Rate

• No Leaching or Addition of Parent or Daughter Isotopes

• All Assumptions Valid for Billions of Years

• There is also a difficulty in measuring precisely very small amounts of the various isotopes

There is, of course, one radiometric dating method that appears to overcome the vital "zero date problem".  The isochron dating method theoretically overcomes the need to know the initial ratio of parent and daughter isotopes.  It will be covered in more detail below.  For now, we will look at those methods that do fall under the above assumptions.

Interweaving the relative time scale with the atomic time scale poses certain problems because only certain types of rocks, chiefly the igneous variety, can be dated directly by radiometric methods; but these rocks do not ordinarily contain fossils.  Igneous rocks are those such as granite and basalt, which crystallize from molten material called "magma".   There is even some valid question as to if granite could be formed from magma at all since this has never, to my knowledge, been observed or duplicated in the lab.  Radio-halos from rapidly decaying radioactive isotopes in granite seem to indicate that the granites were formed almost instantly.

Most sedimentary rocks such as sandstone, limestone, and shale (which do contain fossils) are related to the radiometric time scale by bracketing them within time zones that are determined by dating appropriately selected igneous rocks in lava flows, or weathered from lava flows.

Potassium - Argon and Argon - Argon dating are based on the current understanding that radioactive Potassium-40 decays to the stable form, Argon-40 with a half-life of approximately 1.25 billion years.  The same principle holds true for the other isotope dating methods.

Radioactive decay occurs at a constant exponential or geometric rate.  The rate of decay is proportional to the number of parent atoms present.  There are some circumstances that can affect this rate such as magnetic fluctuations etc... But in general, this seems to be a constant.

If one starts with an originally pure sample of “parent element,” then the proportion of parent to daughter tells us the number of half-lives, which has been used to find the supposed age of igneous rocks.  For example, if there are equal amounts of parent and daughter isotopes, then one half-life has passed.  If there are three times as many daughter isotopes as parent, then two half-lives have passed, and so on.

Most minerals, which contain radioactive isotopes, are in igneous rocks. The majority of scientists today assume that the dates they give indicate the time the magma cooled.  This also assumes that there was no initial daughter isotopes contained in the magma at the time of cooling.  The assumption is that at least a great majority of the isotope present was the parent isotope.  This parent isotope then degraded to the daughter isotope over time.  Consider the following statement by Dalrymple, a well-known geologist:

“The K-Ar method is the only decay scheme that can be used with little or no concern for the initial presence of the daughter isotope. This is because ⁴⁰Ar is an inert gas that does not combine chemically with any other element and so escapes easily from rocks when they are heated. Thus, while a rock is molten, the ⁴⁰Ar formed by the decay of ⁴⁰K escapes from the liquid.” ²

 

So, according to Dalrymple, K-Ar or Ar-Ar are the only methods that have little or no concern for the presence of initial daughter isotopes.  This means that all the other radioisotope-dating methods (excepting isochron methods) are brought into serious question.  The reason for this is because unless the initial ratio of parent to daughter isotope is known, the current ratio would be worthless as a means of determining elapsed time.  A rock cannot be said to be millions or billions of years old if there is no way of knowing what the original composition of the rock was at the time that it was formed.  The assumption for the K-Ar method is that all argon escapes at the time of rock formation because argon is a gas while potassium is not.  Likewise, the other non-isochron dating methods, such as uranium-lead, also fall short because who is to say when the "zero date" was when there was only parent isotope and no daughter?  Because of this problem, it might be a significant error to simply assume that all original isotopes present in a given rock were parent isotopes. 

"The primary assumption upon which K-Ar model-age dating is based assumes zero ⁴⁰Ar in the mineral phases of a rock when it solidifies. This assumption has been shown to be faulty." CEN Tech. J., Vol. 10, No. 3, p:342 1996

Lets now consider how fossils are dated with many of these methods, such as the potassium-argon method.  The mineralized fossils themselves are not directly datable by radiometric techniques.  The sedimentary rock that buried them is also not datable.  If there is some igneous rock fragments in that sedimentary rock layer, these fragments are dated, most commonly, by the ⁴⁰K/⁴⁰Ar dating method described above.  It is assumed then that the fossil is as old as the igneous rock fragment that it is buried with.  Aside from the zero-date problems noted above, one might consider the possibility that the fossil might not be as old as the sediment that buried it in the first place.  For example, lets say that my pet dog dies.  I decide to bury it in the back yard.  Is the dog as old as the dirt that I buried it in?  Likewise, who is to say that some fossils were not buried in sedimentary material that was weathered from significantly more ancient formations?

Potassium-Argon and Argon-Argon Dating

Since Potassium-Argon and Argon-Argon dating techniques are the most common and are considered, even by geologists, to be among the most accurate of all the radioisotope dating methods, lets consider these in particular detail. 

Argon is a noble gas. The main isotopes of argon in terrestrial systems are ⁴⁰Ar (99.6%), ³⁶Ar (0.337%), and ³⁸Ar (0.063%). Naturally occurring ⁴⁰K decays to stable ⁴⁰Ar (11.2%) by electron capture and by positron emission, and decays to stable ⁴⁰Ca (88.8%) by negatron emission; ⁴⁰K has a half-life of 1.25 billion years.

Most of the argon isotope literature deals with measurement of ⁴⁰Ar for use in ⁴⁰K/⁴⁰Ar dating of rocks. The conventional ⁴⁰K/⁴⁰Ar dating method depends on the assumption that the rocks contained no argon at the time of formation and that all the subsequent radiogenic argon (i.e., ⁴⁰Ar) was quantitatively retained. Minerals are dated by measurement of the concentration of potassium, and the amount of radiogenic ⁴⁰Ar that has accumulated. The minerals that are best suited for dating include biotite, muscovite, and plutonic/high grade metamorphic hornblende, and volcanic feldspar; whole rock samples from volcanic flows and shallow instrusives can also be dated if they are unaltered (Faure, 1986).

Under some circumstances the requirements for successful ⁴⁰K/⁴⁰Ar dating may be violated. For example, if ⁴⁰Ar is lost by diffusion while the rock cooled, the age-dates represent the time elapsed since the rock cooled sufficiently for diffusive losses to be insignificant.  Or, if excess ⁴⁰Ar is present in the rock, the calculated age-dates are too old.  The ⁴⁰Ar/³⁹Ar method is thought to be able to overcome this problem inherent with the ⁴⁰K/⁴⁰Ar method.

The ⁴⁰Ar/³⁹Ar dating method is based on the formation of ³⁹Ar as a result of the intentional irradiation of K-bearing samples within a nuclear reactor. The bombardment produces various isotopes of Ar, K, Ca, and Cl, but the dominant source of ³⁹Ar is from ³⁹K.  Radioactive ³⁹Ar decays back to ³⁹K by beta emission with a half-life of 269 years, but the decay is slow compared to the analysis time and can be ignored (Faure, 1986).  The principal “advantage” of ⁴⁰Ar/³⁹Ar dating is that argon can be released partially by stepwise heating of irradiated samples, producing a spectrum of dates related to the “thermal history of the rock” (understanding that Argon is a gas while Potassium is not). 

Because of this, it is much easier to determine a ⁴⁰K/⁴⁰Ar ratio and do it in a stepwise fashion with varying amounts of time and heat.  This “stepwise” testing is thought to eliminate the errors caused by “extraneous” argon that might have “contaminated” the rock over time either by a loss or a gain of “outside” argon (ie: atmospheric argon).  The problem with this theory is that who is to know which step, or average of steps in the process represents the “correct” ⁴⁰K/⁴⁰Ar ratio?  How is this calibrated? Also, even if the argon-argon dating method does eliminate the "contamination" problem, it does not solve the problem of original argon.  Did the clock get reset to zero when the volcano erupted?  Or, was there some argon trapped in the rocks originally?  Also, the ⁴⁰Ar/³⁹Ar dating method is not an independent dating method.  It must be first calibrated against a sample of "known age".  This age of this sample is usually determined by, you guessed it, the ⁴⁰K/⁴⁰Ar method.

Recent experiments on volcanoes of known ages have been done using the ⁴⁰Ar/³⁹Ar dating method, which seem to confirm its accuracy.  Recent testing of volcanic material from Mt. Vesuvius was dated accurately with the ⁴⁰Ar/³⁹Ar method to within seven years of the actual event.^{3  40}Ar/³⁹Ar Dating into the Historical Realm: Calibration Against Pliny the Younger was written by P. R. Renne et. al. and published in Science 277: 1279-1280 (1997). Renne tested Ar-Ar dating by checking it against the 79 A.D. eruption of Vesuvius that destroyed Pompeii. Renne and his team noted that “Analysis of single crystals, for example by laser fusion, can obviate xenocrystic contamination, but single crystals are seldom large enough to yield measurable quantities of ⁴⁰Ar^* through radiogenic ingrowth in the Holocene [i.e. last 12,000 years].”  Would Ar-Ar dating methods work such recent material?  It apparently did. The testing returned an age of 1925±94 years. The true age was 1918 years.  The test was off only 7 years. The conclusions of Renne and his team read as follows:

Thus despite the presence of excess ⁴⁰Ar, a sample less than 2000 years old can be dated with better than 5% precision, validating ⁴⁰Ar/³⁹Ar dating as a reliable geochronometer into the late Holocene. These results also demonstrate that excess ⁴⁰Ar can be identified in volcanic sanidine, and while perhaps negligible in pre-Holocene rocks, it has important consequences for sample at the limit of the method’s applicability. Further improvement in precision of ⁴⁰Ar/³⁹Ar analysis of historically dated samples may lead to welcome refinements in the ages of neutron fluence monitors, currently a limitation on the accuracy of the ⁴⁰Ar/³⁹Ar method. Our results also substantiate validity of the ⁴⁰Ar/³⁹Ar method in establishing the eruptive histories of populated active volcanic regions, where such information is vital to volcanic hazard assessment.

Of note however is that this test was not double blinded, and the number of such tests is not statistically significant as far as scientific analysis is concerned.  Although interesting, it is basically a case study report, and as such it has very little scientific weight as far as statistical predictability.

Specific Problems with K-Ar and Ar-Ar Dating

In the first place, I am not primarily concerned with dating meteorites, or Precambrian rocks. What I am more interested in is the fossil-bearing geologic column of Cambrian and later “ages”.    Since ⁴⁰K/⁴⁰Ar and ⁴⁰Ar/³⁹Ar dating are most commonly used to "prove" the ancient age of many life forms, I will discuss these dating methods specifically in more detail and show that they, along with the other common methods of isotope dating, are to be highly questioned.  I will begin this section with a short discussion from Andrew Snelling, an associate professor of geology in El Cajon, California.

 

According to the assumptions foundational to potassium-argon (K-Ar) and argon-argon (Ar-Ar) dating of rocks, there should not be any daughter radiogenic argon (⁴⁰Ar^*) in rocks when they form. When measured, all ⁴⁰Ar^* in a rock is assumed to have been produced by in situ radioactive decay of ⁴⁰K within the rock since it formed. However, it is well established that volcanic rocks (e.g. basalt) contain excess ⁴⁰Ar^*, that is, ⁴⁰Ar which cannot be attributed to either atmospheric contamination or in situ radioactive decay of ⁴⁰K.  This excess ⁴⁰Ar^* represents primordial Ar carried from source areas in the earth's mantle by the parent magmas, is inherited by the resultant volcanic rocks, and thus has no age significance.

However, are all other rocks in the earth's crust also susceptible to "contamination" by excess ⁴⁰Ar^* emanating from the mantle? If so, then the K-Ar and Ar-Ar "dating" of crustal rocks would be similarly questionable.

When muscovite (a common mineral in crustal rocks) is heated to 740°-860°C under high Ar pressures for periods of 3 to 10.5 hours it absorbs significant quantities of Ar, producing K-Ar "ages" of up to 5 billion years, and the absorbed Ar is indistinguishable from radiogenic argon (⁴⁰Ar^*). In other experiments muscovite was synthesized from a colloidal gel under similar temperatures and Ar pressures, the resultant muscovite retaining up to 0.5 wt% Ar at 640°C and a vapor pressure of 4,000 atmospheres. This is approximately 2,500 times as much Ar as is found in natural muscovite. Thus under certain conditions Ar can be incorporated into minerals which are supposed to exclude Ar when they crystallize.

Because it is known that excess ⁴⁰Ar^* is carried from the mantle by plumes of mafic magmas up into the earth's crust, it is equally likely that much of the excess ⁴⁰Ar^* in crustal rocks could be primordial ⁴⁰Ar. Thus, we have no way of knowing if any of the ⁴⁰Ar^* measured in crustal rocks has any age significance. Additional to the primordial ⁴⁰Ar from the mantle is ⁴⁰Ar^* released from minerals and rocks during diagenesis and metamorphism, so that there is continual migration and circulation of both primordial ⁴⁰Ar and ⁴⁰Ar^* in the crust which is reflected in their presence in CO₂-rich natural gases. Therefore, when samples of crustal rocks are analyzed for K-Ar and Ar-Ar "dating," one can never be sure that whatever ⁴⁰Ar^* is in the rocks is from in situ radioactive decay of ⁴⁰K since their formation, or if some or all of it came from the mantle or from other crustal rocks and minerals. Thus all K-Ar and Ar-Ar "dates" of crustal rocks are questionable, as well as fossil "dates" calibrated by them.¹⁹

 

In summary, many scientists assume that since argon is a gas, all of it should have escaped from the lava before it cooled. Therefore, all the ⁴⁰Ar in the rock should be the result of decay from potassium. Based on the measured potassium, argon, and the decay rate, they calculate an age. That is why it does not matter how long the magma was in the volcano before it erupted. They believe that when the volcano erupts, all the ⁴⁰Ar escapes, and the atomic clock gets reset to zero.

If all the argon escaped from hot lava of volcanoes that erupted long ago, then all the argon should escape from the hot lava of volcanoes that erupt in modern times too. But modern lava does have ⁴⁰Ar in it. This is known as the "excess argon problem".  Scientists are well aware of this problem and use various calibration methods to "correct" for this problem.  However, how are these calibration methods established?  Upon what basis are they validated?

 

"Another issue affecting the ultimate precision and accuracy of the ⁴⁰Ar/³⁹Ar technique is the uncertainty in the decay constants for ⁴⁰K. This uncertainty results from 1) the branched decay scheme of ⁴⁰K and 2) the long half-life of ⁴⁰K (1.25 billion years)." ⁴⁹

 

Some suggest that the ⁴⁰Ar/³⁹Ar dating method, in particular, helps to solve the contamination problem because the sample is step heated and excess argon can be detected once the data is plotted.  This might be true, if the sample weren't evenly contaminated. Beyond this, ⁴⁰Ar/³⁹Ar is not an independent dating technique, but must first be calibrated against a sample of "known age" before it can be used. 

 

 "Once an accurate and precise age is determined for the primary standard, other minerals can be dated relative to it by the ⁴⁰Ar/³⁹Ar method. . . However, while it is often easy to determine the age of the primary standard by the K/Ar method, it is difficult for different dating laboratories to agree on the final age. Likewise, because of heterogeneity problems with the MMhb-1 sample, the K/Ar ages are not always reproducible. This imprecision (and inaccuracy) is transferred to the secondary minerals used daily by the ⁴⁰Ar/³⁹Ar technique." ⁴⁹    

Fission Track Dating

Fission track dating is a radioisotopic dating method that depends on the tendency of uranium (Uranium-238) to undergo spontaneous fission as well as the usual decay process. The large amount of energy released in the fission process ejects the two nuclear fragments into the surrounding material, causing damage paths called fission tracks. The number of these tracks, generally 10-20 µ in length, is a function of the initial uranium content of the sample and of time. These tracks can be made visible under light microscopy by etching with an acid solution so they can then be counted.

The usefulness of this as a dating technique stems from the tendency of some materials to lose their fission-track records when heated, thus producing samples that contain fission-tracks produced since they last cooled down. The useful age range of this technique is thought to range from 100 years to 100 million years before present (BP), although error estimates are difficult to assess and rarely given. Generally it is thought to be most useful for dating in the window between 30,000 and 100,000 years BP.

A problem with fission-track dating is that the rates of spontaneous fission are very slow, requiring the presence of a significant amount of uranium in a sample to produce useful numbers of tracks over time. Additionally, variations in uranium content within a sample can lead to large variations in fission track counts in different sections of the same sample.⁴²

Because of such potential errors, most forms of fission track dating use a form of calibration or "comparison of spontaneous and induced fission track density against a standard of known age. The principle involved is no different from that used in many methods of analytical chemistry, where comparison to a standard eliminates some of the more poorly controlled variables. In the zeta method, the dose, cross section, and spontaneous fission decay constant, and uranium isotope ratio are combined into a single constant." ⁴³

Of course, this means that the fission track dating method is not an independent method of radiometric dating, but is dependent upon the reliability of other dating methods.  The reason for this is also at least partly due to the fact that the actual rate of fission track production. Some experts suggest using a rate constant of 6.85x10^-17 yr^-1 while others recommend using a rate of 8.46x10^-17 yr^-1 (G. A. Wagner, Letters to Nature, June 16, 1977).  This difference might not seem like much, but when it comes to dates of over one or two million years, this difference amounts to about 25-30% in the estimated age value. In other words, the actual rate of fission track production isn't really known, nor is it known if this rate can be affected by various concentrations of U²³⁸ or other physical factors.  For example, all fission reactions produce neutrons. What happens if fission from some other radioactive element, like U²³⁵ or some other radioisotope, produces tracks?  Might not these trackways be easily confused with those created by fission of U²³⁸?   

The human element is also important here. Fission trackways have to be manually counted.  This is problematic since interpreting what is and what is not a true trackway isn't easy.  Geologists themselves recognize the problem of mistaking non-trackway imperfections as fission tracks.  "Microlites and vesicles in the glass etch out in much the same way as tracks."⁴⁵ Of course, there are ways to avoid some of these potential pitfalls.  For example, it is recommended that one choose samples with as few vesicles and microlites as possible. But, how is one to do this if they are so easily confused with true trackways? Fortunately, there are a few other "hints". True tracks are straight, never curved. They also tend to show characteristic ends that demonstrate "younging" of the etched track. True tracks are thought to form randomly and have a random orientation.  Therefore, trackways that show a distribution pattern tend not to be trusted as being "true".  Certain color and size patterns within a certain range are also used as helpful hints.  Yet, even with all these hints in place, it has been shown that different people count the same trackways differently - up to 20% differently.⁴⁴   Add up the human error with the error of fission track rate and we are suddenly up to a range of error of 50% or so.

This is yet another reason why calibration with other dating techniques is used in fission track dating. It just isn't very reliable or accurate by itself.  And, it gets even worse. Fairly recently, Raymond Jonckheere and Gunther Wagner (American Minerologist, 2000) published results showing that there are two kinds of real fission trackways that had "not been identified previously."  The first type of trackway identified is a "stable" track and the second type is produced through fluid inclusions. As it turns out, the "stable tracks do not shorten significantly even when heated to temperatures well above those normally sufficient for complete annealing of fission tracks."  Of course, this means that the "age" of the sample would not represent the time since the last thermal episode as previously thought.  The tracks through fluid are also interesting. They are "excessively long".  This is because a fission fragment traveling through a fluid inclusion does so without appreciable energy loss. Such features, if undetected, "can distort the temperature-time paths constructed on the basis of confined fission-track-length measurements."   Again, the authors propose measures to avoid such pitfalls, but this just adds to the complexity of this dating "method" and calls into question the dates obtained before the publication of this paper (i.e., 2000).⁴⁶     

These problems have resulted in several interesting contradictions, despite calibration.  For example, Naeser and Fleischer (Harvard University) showed that, depending upon the calibration method chosen, the calculated age of a given rock (from Cerro de Mercado, Mexico in this case) could be different from each other by a factor of "sixty or more" - - "which give geologically unreasonable ages.  In addition, published data concerning the length of fission tracks and the annealing of minerals imply that the basic assumptions used in an alternative procedure, the length reduction-correction method, are also invalid for many crystal types and must be approached with caution unless individually justified for a particular mineral." [emphasis added] ⁴⁷  Now that's pretty significant - being off by a factor of sixty or more?!  No wonder the authors recommend only going with results that do not provide "geologically unreasonable ages".       

Tektites

Another example of this sort of aberrancy comes in the form of glass globs known as "tektites".  Tektites are thought to be produced when a meteor impacts the Earth.  When the massive impact creates a lot of heat, which melts the rocks of the Earth and send them hurtling through the atmosphere at incredible speed.  As these fragments travel through the atmosphere, they become superheated and malleable as they melt to a read-hot glow, and are formed and shaped as they fly along.  It is thought that the date of the impact can be dated by using various radiometric dating methods to date the tektites. For example, Australian tektites (known as australites) show K-Ar and fission track ages clustering around 700,000 years.  The problem is that their stratigraphic ages show a far different picture. Edmund Gill, of the National Museum of Victoria, Melbourne, while working the Port Campbell area of western Victoria uncovered 14 australite samples in situ above the hardpan soil zone. This zone had been previously dated by the radiocarbon method at seven locales, the oldest dating at only 7,300 radiocarbon years (Gill 1965). Charcoal from the same level as that containing specimen 9 yielded a radiocarbon age of 5,700 years. The possibility of transport from an older source area was investigated and ruled out. Since the "Port Campbell australites include the best preserved tektites in the world ... any movement of the australites that has occurred ... has been gentle and has not covered a great distance" (Gill 1965). Aboriginal implements have been discovered in association with the australites. A fission-track age of 800,000 years and a K-Ar age of 610,000 years for these same australites unavoidably clashes with the obvious stratigraphic and archaeological interpretation of just a few thousand years.

"Hence, geological evidence from the Australian mainland is at variance, both as to infall frequency and age, with K-Ar and fission-track dating" (Lovering et al. 1972). Commenting on the above findings by Lovering and his associates, the editors of the book, Tektites, state that, "in this paper they have built an incontrovertible case for the geologically young age of australite arrival on earth" (Barnes and Barnes 1973, p. 214).

This is problematic.  The argument that various radiometric dating methods agree with each other isn't necessarily true. Here we have the K-Ar and fission track dating methods agreeing with each other, but disagreeing dramatically with the radiocarbon and historical dating methods.  These findings suggest that, at least as far as tektites are concerned, the complete loss of ⁴⁰Ar (and therefore the resetting of the radiometric clock) may not be valid (Clark et al. 1966). It has also been shown that different parts of the same tektite have significantly different K-Ar ages (McDougall and Lovering, 1969).  This finding suggests a real disconnect when it comes to the reliability of at least two of the most commonly used radiometric dating techniques.⁴⁸     

In short, it seems like fission track dating is tenuous a best - even when given every benefit of the doubt.  It is just too subjective and too open to pitfalls in interpretation to be used as any sort of independent measure of estimating elapsed time.

Circular Calibration Methods

  There is a methodological problem connected with the manner in which geologists infer the argon-retention abilities of different minerals. Concerning the suitability of different minerals for K-Ar dating, Faure (1986, p. 72) writes "The minerals beryl, cordierite, pyroxene, and tourmaline frequently contain excess 40Ar, while hornblende, feldspar, phlogopite, biotite, and sodalite contain such excess 40Ar only rarely ... ." And how is this known? By comparing the K-Ar dates yielded by such minerals with the expected ones. Thus the correctness of the geologic time scale is assumed in deciding which minerals are suitable for dating. For example, concerning the use of glauconies for K-Ar dating, Faure (1986, p. 78) writes, "The results have been confusing because only the most highly evolved glauconies have yielded dates that are compatible with the biostrategraphic ages of their host rocks whereas many others have yielded lower dates. Therefore, K-Ar dates of 'glauconite' have often been regarded as minimum dates that underestimate the depositional age of their host." All of the choices are made in order to obtain dates that are more in agreement with each other. 

It is also interesting that Faure (1986, pp. 345-6) mentions that fission track dating is calibrated (the "zeta calibration") using rocks of "known" ages.   However, if these "known" ages are incorrect, then fission track dating that is based on these ages is also incorrect.  Thus fission track dating is not an independent test that helps to verify the accuracy of other tests.  The result is that radiometric dating in general is in danger of being based on circular reasoning.²⁵

Examples of Problems with Radiometric Dating Techniques

Dalrymple's work early work on 26 historic lava flows showed that many of them had excess argon and were not set to zero at the eruption of the volcano.  The following is the data from these tests: ⁵

• Hualalai basalt, Hawaii (AD 1800-1801)                    1.05 to 1.19 million years

• Mt. Etna basalt, Sicily (122 BC)                                   100,000 years

• Mt. Etna basalt, Sicily (AD 1972)                                 150,000 years

• Mt. Lassen plagioclase, California (AD 1915)         130,000 years

• Sunset Crater basalt, Arizona (AD 1064-1065)       210,000 to 220,000 years

• Glass Mountain (BP 130-390)                                     130,000 years in the future

• Mt. Mihara (AD 1951)                                                     70,000 years in the future

• Sakurajima (AD 1946)                                                   200,000 years in the future

Dalrymple comments on such findings by saying, "With the exception of the Hualalai flow, the amounts of excess ⁴⁰Ar and ³⁶Ar found in the flows with anomalous ⁴⁰Ar/³⁶Ar ratios were too small to cause serious errors in potassium-argon dating of rocks a few million years or older. However, these anomalous ⁴⁰Ar/³⁶Ar ratios could be a problem in dating very young rocks. If the present data are representative, argon of slightly anomalous composition can be expected in approximately one out of three volcanic rocks."

Dalrymple may have a point.  It seems like rocks dating within one or two million years cannot be accurately dated by K-Ar techniques just because of the relatively wide ranges of error.  However, can rocks that are tens or hundreds of millions of years be more accurately dated?  Perhaps, if these rocks were in fact closed systems and were not subject to contamination by external argon. 

Investigators also have found that excess ⁴⁰Ar is trapped in the minerals within lava flows.^{7, 8, 9} Several instances have been reported of phenocrysts with K-Ar "ages" 1-7 millions years greater than that of the whole rock, and one K-Ar "date" on olivine phenocrysts in a recent (<13,000 year old) basalt was greater than 110 Ma.¹⁰ Laboratory experiments have tested the solubility of argon in synthetic basalt melts and their constituent minerals, with olivine retaining 0.34 ppm ⁴⁰Ar.^{11, 12}  It was concluded that the argon is held primarily in lattice vacancy defects within the minerals.

The obvious conclusion most investigators have reached is that the excess ⁴⁰Ar had to be present in the molten lavas when extruded, which then did not completely degas as they cooled, the excess ⁴⁰Ar becoming trapped in constituent minerals and the rock fabrics themselves. However, from whence comes the excess ⁴⁰Ar, that is, ⁴⁰Ar which cannot be attributed to atmospheric argon or in situ radioactive decay of ⁴⁰K? It is not simply "magmatic" argon? Funkhouser and Naughton found that the excess ⁴⁰Ar in the 1800-1801 Hualalai flow, Hawaii, resided in fluid and gaseous inclusions in olivine, plagioclase, and pyroxene in ultramafic xenoliths in the basalt, and was sufficient to yield "ages" of 2.6 Ma to 2960 Ma.¹³ Thus, since the ultramafic xenoliths and the basaltic magmas came from the mantle, the excess ⁴⁰Ar^* must initially reside there, to be transported to the earth's surface in the magmas.

Many recent studies confirm the mantle source of excess ⁴⁰Ar. Hawaiian volcanism is typically cited as resulting from a mantle plume, most investigators now conceding that excess ⁴⁰Ar in the lavas, including those from the active Loihi and Kilauea volcanoes, is indicative of the mantle source area from which the magmas came. Considerable excess ⁴⁰Ar measured in ultramafic mantle xenoliths from Kerguelen Archipelago in the southern Indian Ocean likewise is regarded as the mantle source signature of hotspot volcanism.¹⁴ Indeed, data from single vesicles in mid-ocean ridge basalt samples dredged from the North Atlantic suggest the excess ⁴⁰Ar in the upper mantle may be almost double previous estimates, that is, almost 150 times more than the atmospheric content (relative to ³⁶Ar).¹⁵ Another study on the same samples indicates the upper mantle content of ⁴⁰Ar could be even ten times higher.¹⁶

Further confirmation comes from diamonds, which form in the mantle and are carried by explosive volcanism into the upper crust and to the surface. When Zashu et al. obtained a K-Ar isochron "age" of 6.0±0.3 Ga for 10 Zaire diamonds, it was obvious excess ⁴⁰Ar was responsible, because the diamonds could not be older than the earth itself.¹⁴ These same diamonds produced ⁴⁰Ar/³⁹Ar "age" spectra yielding a ~5.7 Ga isochron.¹⁷ It was concluded that the ⁴⁰Ar is an excess component which has no age significance and is found in tiny inclusions of mantle-derived fluid.

The conventional K-Ar dating method was applied to the 1986 dacite flow from the new lava dome at Mount St. Helens, Washington. Porphyritic dacite which solidified on the surface of the lava dome in 1986 gives a whole rock K-Ar 'age' of 0.35 ± 0.05 million years (Ma). Mineral concentrates from the dacite which formed in 1986 give K-Ar 'ages 'from 0.34 ± 0.06 Ma (feldspar-glass concentrate) to 2.8 ± 0.6 Ma (pyroxene concentrate). These dates are, of course, preposterous. The fundamental dating assumption (no radiogenic argon was present when the rock formed) is brought into question.  Instead, data from the Mount St. Helens dacite argue that significant "excess" argon was present when the lava solidified in 1986. Phenocrysts of orthopyroxene, hornblende and plagioclase are interpreted to have occluded argon within their mineral structures deep in the magma chamber and to have retained this argon after emplacement and solidification of the dacite. The amount of argon occluded is probably a function of the argon pressure when mineral crystallization occurred at depth and/or the tightness of the mineral structure. Orthopyroxene retains the most argon, followed by hornblende, and finally, plagioclase. The lava dome at Mount St. Helens dates very much older than its true age because phenocryst minerals inherit argon from the magma. The study of this Mount St. Helens dacite brings yet another question to mind:  How accurate are K-Ar "ages" from the many other phenocryst-containing lava flows world-wide?¹⁸

The Contamination Argument

Potassium is about 2.5% of the earth's crust. About 1/10,000 of potassium is ⁴⁰K, which decays into ⁴⁰Ar with a half-life of 1.25 billion years. Actually, only about 1/10^th of the⁴⁰K decays to Argon, and the rest decays to calcium.  Argon is about 3.6 x 10^-4 % of the earth's crust. We can assume then that the magma is probably about 2.5% potassium and about 0.00025% of the radioactive form, Potassium-40 (⁴⁰K).  Now, Lets say we are trying to date a one billion year old rock.  How much of it would be ⁴⁰K?  Starting with 0.00025% as the modern concentration of ⁴⁰K in magma, we would have to divide by roughly two (About one half-life).  This would leave us with a 0.000125% of ⁴⁰K.  Now, about 90% of the decay product is calcium and only about 10% is Ar-40.  This gives about 0.0000125% ⁴⁰Ar in the total make-up of the rock.  This is about one ten millionth of the mass of the rock, a very tiny fraction.  If the rock weighed one gram, the Ar-40 in the rock would weight one ten millionth of a gram.  And yet, with a relatively large amount of argon in the air, argon filtering up from rocks below, excess argon in lava, the fact that argon and potassium are water soluble, and the fact that argon is mobile in rock and is a gas, we are still expecting this wisp of argon gas to tell us how old the rock is?  The percentage of ⁴⁰Ar is even less for younger rocks. For example, it would be about one part in 100 million for rocks in the vicinity of 50-60 million years old.  However, to get just one part in 10 million of argon in a rock in a thousand years, we would only need to get one part in 10 billion entering the rock each year. This would be less than one part in a trillion entering the rock each day, on the average. This would suffice to give a rock an average computed potassium-argon age of over a billion years.  Some geochronologists believe that a possible cause of excess argon is that argon diffuses into certain minerals progressively with time and pressure.  Significant quantities of argon may be introduced into a mineral even at pressures as low as one bar.

We can also consider the average abundance of argon in the crust.  If we assume that a rock has 1/400,000 ⁴⁰K, that is, 2.5 x 10^{-6 40}K, and 3.6 x 10^{-6 40}Ar, then eight times this much ⁴⁰K must have decayed, thus about 28.8 x 10^-6 parts of ⁴⁰K have decayed, so there is less than 1/10 of the original ⁴⁰K left. This implies a radiometric age of over 4 billion years.  So a rock can get a very old radiometric age just by having average amounts of potassium and argon.  It seems reasonable to me that the large radiometric ages are simply a consequence of mixing, and not related to ages at all, at least not necessarily the ages of the rocks themselves. The rates of exchange that would mess up “dates” are very small. It seems to me to be a certainty that water and gas will enter most, if not all, volcanic type rocks through tiny openings and invalidate almost all K-Ar ages.  Rocks are not sealed off from the environment.  Even if magma was set to “zero time” at the eruption of a volcano, over the course of eons of time and exposure to atmospheric and other sources of extraneous argon, it would seem that contamination would be inevitable.  This contamination would seem to be more and more of a problem the older the rock became.

Let me illustrate the circulation patterns of argon in the earth's crust. About 2.5 percent of the earth's crust is believed to be potassium, and about 1/10,000 of this is ⁴⁰K, which decays to ⁴⁰Ar with a half-life of about 1.25 billion years. So argon is being produced throughout the earth's crust, and in the magma, all the time. In fact, it probably rises to the top of the magma, artificially increasing its concentration there. Now, some rocks in the crust are believed not to hold their argon, so this argon will enter the spaces between the rocks. Leaching also occurs, releasing argon from rocks. Heating of rocks can also release argon. Argon is released from lava as it cools, and probably filters up into the crust from the magma below, along with helium and other radioactive decay products.  All of this argon is being produced and entering the air and water in between the rocks, and gradually filtering up to the atmosphere.  But, we know that some minerals absorb argon (“correction factors” are applied for this when using K-Ar dating). So this argon that is being produced will leave some rocks and enter others.  The various pressures, temperatures, moisture, nature of the materials and a variety of other factors all play together to challenge the validity of K-Ar and/or Ar-Ar dating.

Different Dating Methods Agree

It is often said that a great many dating methods, used on a single specimen, will agree with each other, thus establishing the accuracy of the date given.  In reality, the overwhelming majority of measurements on the fossil bearing geologic column are all done using one method, the K-Ar method (Recall that both potassium and argon are water soluble, and argon (a gas) is mobile in rock.)

"The construction of this time scale was based on about 380 radioisotope ages that were selected because of their agreement with the presumed fossil and geological sequences found in the rocks. Radioisotope ages that did not meet these requirements were rejected on the basis of presumed chemical and/or physical modifications that made the "ages" unreliable indicators of real time. About 85% of the selections were K-Ar date s, 8% rubidium-strontium dates, and 4% uranium-lead dates."

Thus the agreement found between many dates does not necessarily reflect an agreement between different methods, but rather the agreement of the K-Ar method with itself (Especially noting that Dalrymple suggested that only K-Ar dating methods were at all trust worthy).  I have seen no good double-blinded research studies that say otherwise.  One would think that if this were a good science, then such studies would be done and published, but they are strangely lacking.

Also, specific differences are known and have been known to exist between different dating methods.  For example, Isotopic studies of the Cardenas Basalt and associated Proterozoic diabase sills and dikes have produced a geologic mystery. Using the conventional assumptions of radioisotope dating, the Rb-Sr and K-Ar systems should give concordant "ages". However, it has been known for over 20 years that the two systems give discordant "ages", the K-Ar "age" being significantly younger than the Rb-Sr "age". 

The "argon reset model" was the first explanation proposed for the discordance. A metamorphic event is supposed to have expelled significant argon from these rocks. The reset model is unable to reconcile the new data, leading to a metamorphic event which is excessively young and inconsistent with the conventional stratigraphic interpretation.

The "argon leakage model" also attempts to explain why these rocks have about half the argon which seems to be required by the Rb-Sr system.  The leakage model supposes an incredible improbability. Both the old and new data imply that the rocks leaked argon in nearly exact proportion to the abundance of potassium producing a "leakage isochron", an explanation not supported by a quantity of an appropriate mineral or mesostasis phase. Strong negative correlation between K-Ar model age and K₂O in the upper portion of the Cardenas Basalt is not easily explained in a consistent manner. Furthermore, reset and leakage models have difficulty explaining the abundance of initial ³⁶Ar in the rocks, especially the abundance of ³⁶Ar in those rocks which supposedly leaked the most ⁴⁰Ar.

Three alternatives are suggested to the two argon loss models. The "argon inheritance model" and "argon mixing model" simply propose that argon is positively correlated with potassium from its magma source or produced by a mixing process, and that the linear relationship on a plot of ⁴⁰Ar versus ⁴⁰K is an artifact of the magma, not produced by radioisotope decay within these rocks. The inheritance of argon seems to be a better model than is the mixing model. The "change of decay model" goes to the physics of radioisotope decay and proposes a fundamental change in ⁸⁷Rb and/or ⁴⁰K decay. All three explanations offered as alternatives to the argon loss models invalidate using the K-Ar system as conventional geochronology would assume. ²³

Isochrons

The word "isochron" basically means "same age".  Isochron dating is based on the ability to draw a straight line between data points that are thought to have formed at the same time.  The slope of this line is used to calculate an age of the sample in isochron radiometric dating.  The isochron method of dating is perhaps the most logically sound of all the dating methods - at first approximation.  This method seems to have internal measures to weed out those specimens that are not adequate for radiometric evaluation. Also, the various isochron dating systems seem to eliminate the problem of not knowing how much daughter element was present when the rock formed.

Isochron dating is unique in that it goes beyond measurements of parent and daughter isotopes to calculate the age of the sample based on a simple ratio of parent to daughter isotopes and a decay rate constant - plus one other key measurement.  What is needed is a measurement of a second isotope of the same element as the daughter isotope.  Also, several different measurements are needed from various locations and materials within the specimen.  This is different from the normal single point test used with the other "generic" methods.   To make the straight line needed for isochron dating each group of measurements (parent - P, daughter - D, daughter isotope - Di) is plotted as a data point on a graph.  The X-axis on the graph is the ratio of P to Di.  For example, consider the following isochron graph: ²¹

Obviously, if a line were drawn between these data points on the graph, there would be a very nice straight line with a positive slope.  Such a straight line would seem to indicate a strong correlation between the amount of P in each sample and the extent to which the sample is enriched in D relative to Di.  Obviously one would expect an increase in the ratio of D as compared with Di over time because P is constantly decaying into D, but not into Di.  So, Di stays the same while D increases over time.  

But, what if the original rock was homogenous when it was made?  What if all the minerals were evenly distributed throughout, atom for atom?  What would an isochron of this rock look like?  It would look like a single dot on the graph.  Why?  Because, any testing of any portion of the object would give the same results.  

The funny thing is, as rocks cool, different minerals within the rock attract certain atoms more than others.  Because of this, certain mineral crystals within a rock will incorporate different elements into their structure based on their chemical differences.  However, since isotopes of the same element have the same chemical properties, there will be no preference in the inclusion of any one isotope over any other in any particular crystalline mineral as it forms.  Thus, each crystal will have the same D/Di ratio of the original source material.  So, when put on an isochron graph, each mineral will have the same Y-value.    However, the P element is chemically different from the D/Di element.  Therefore, different minerals will select different ratios of P as compared with D/Di.  Such variations in P to D/Di ratios in different elements would be plotted on an isochron graph as a straight, flat line (no slope).