Reply Number 4 to Dr. Henke

Reply Number 4 to Dr. Henke

December 20, 1999

I do set My bow in the cloud

This article is a response to the article entitled "ANOTHER RESPONSE TO DR. PLAISTED'S ATTACKS ON RADIOMETRIC DATING" by Dr. Kevin Henke that was e-mailed to me on December 16, 1999 and also posted to the talk.origins newsgroup. First let me state that my object is not to attack radiometric (isotopic) dating, but to clarify its assumptions. Also, I wonder if it is nearly time to bring this exchange to a close.

Excess Argon

Excess argon can cause K-Ar dates to be too old. Since K-Ar dates constitute a large portion of the dates on the Phanerozoic (Cambrian and later strata), this phenomenon could be invalidating much of the geological time scale if it is prevalent.

In my last message, I gave quotes from Woodmorappe (1999) about the prevalence of excess argon. These characterized excess argon as "widely recognized in dated materials," a "significant problem," "not uncommon," "often ... observed," "much more common than previously suspected," and "a common phenomenon in some minerals, especially biotite," among others. Dr. Henke attempted to explain all of these away. One wonders what kind of a quotation would possibly convince Dr. Henke that excess argon is "not uncommon." And these quotes are only referring to argon recognized by geologists as excess; there might be much more than they recognize. It is true, as Dr. Henke states, that geologists believe they can recognize this excess argon in many cases, but that was not my point.

Even biotite can have large amounts of excess argon (Woodmorappe, 1999, p. 76):

The ability of biotite to incorporate large amounts of excess 40Ar has been recognized for many years. (McDougall and Harrison 1988, p. 110.)
If even biotite can have large amounts of excess argon, is there any mineral that cannot?

It is interesting that Snelling (Creation Ex Nihilo 20(1) 1997-1998 pp. 24-27) gives an example of basalt in a coal mine, estimated to have an age of about 30 million years old, containing wood reliably dated by three laboratories at 29,000 to 45,000 years old using carbon 14 dating. The K-Ar dates of the surrounding basalt ranged from 36 million to 58 million years. The carbon 14 dates were "staunchly defended by the laboratories as valid." Another example was also found by Snelling and reported in Creation Ex Nihilo 22(1) (1999-2000 pp. 18-21). He dated lava flows from Mt. Ngauruhoe, New Zealand. Dr. Snelling dated lava flows from 1949 and 1954 and obtained K-Ar dates from 1 to 3.5 million years from four flows, with one date from each flow of under .27 or .29 million years. I don't know if these samples had obvious xenoliths or xenocrysts, or whether a K-Ar isochron would have revealed the excess argon.

In my last reply I stated

However, inherited argon can be present in a rock without any evidence of weathering, metamorphism, or chemical change, and all the diagnostic tests for it have limitations.
Dr. Henke questioned this statement. As justification, suppose that we have two samples of lava, one that cools in a high partial pressure of argon 40 and the other that cools in a low partial pressure of argon 40. The first sample will end up with much more excess argon than the second, but there will be virtually no other chemical or physical differences between them, since argon 40 is an inert gas. The only hope of detecting this excess argon is a K-Ar isochron or some other test based on the distribution of argon itself.

The tests for excess argon, such as K-Ar isochrons and others, are not completely reliable. For all we know, the tests for excess argon may only rarely detect all the excess argon, and virtually all the argon in most Phanerozoic rocks may be excess argon.

I believe that the argon in Phanerozoic strata is of three types:

(1) Excess argon from the mantle, which consists of argon 40. (See Faure (1986), pp. 72-73.)

(2) Excess argon that consists of a mixture of argon 36 and argon 40, and is sometimes detected by K-Ar isochrons.

(3) A very small amount of argon 40, which has been generated by decay of potassium 40 since the rock was formed or cooled.

This would imply that essentially all of the argon in Phanerozoic strata is excess argon, and that K-Ar dates there are uniformly much too old. In this view, argon from the mantle is giving ages that often agree with the current geological time scale, and this argon is what is generally being measured in K-Ar dating. I will give some justification for this belief below. In general, there is currently not much evidence in favor of this view or against it. Whether radiometric dates are more accurate for pre-phanerozoic rocks or rocks that have no fossils below them I do not know.

K-Ar Isochrons

K-Ar isochrons are one test for excess argon. But mixings can invalidate isochrons. There is a test for mixings. But the test is not reliable. This means that isochrons are not reliable, which means that we cannot know how much excess argon is present. This means that we cannot be sure that all of these K-Ar isotopic dates are not much too old.

If there is a mixing, some of the inherited argon will be perceived as radiogenic argon. The remainder of the inherited argon will be perceived as inherited argon. Thus, even though some inherited argon will be detected, additional inherited argon will go undetected, resulting in anomalously old ages. Geologists agree that mixing is a common phenomenon, so it is reasonable to assume that many K-Ar isochrons are being invalidated in this way, and that many K-Ar ages are too old. The fact that many isochrons accepted by geologists have correlations suggesting mixings adds plausibility to this scenario. Further support is given by excess argon in historic volcanoes, as well as argon acknowledged by geologists to be excess argon. This latter situation is often because the K-Ar ages are older than geologists expect, but even the argon regarded by geologists as radiogenic could be excess, meaning that excess argon could be much more common and abundant than geologists believe.

Dr. Henke questioned how excess argon could go undetected by a K-Ar isochron if the concentration of potassium varied. In the first place, we don't know how many K-Ar isochrons with variable potassium concentration contribute to the Phanerozoic time scale. However, one can obtain K-Ar isochrons by mixing, just as one obtains Rb-Sr isochrons or any other kind of an isochron. This will result in an isochron with a variable potassium concentration. The difference with K-Ar isochrons is that some of the argon may escape during cooling. But if a roughly constant proportion of argon escapes from all samples during cooling, the isochron property will be preserved, and some excess argon will go undetected. This will result in the isochron measuring a spurious, partially inherited age from an earlier event.

Dr. Henke asserts that phenocrysts would also invalidate the constancy of the concentration of potassium:

Most granitic magmas are relatively cool (around 700C) so that they're usually only partially molten. The minerals in the magma (phenocrysts) would tend to have very different potassium concentrations from each other and from their parent magma. Potassium would tend to segregate into muscovite and K-feldspars, according to Bowen's Reaction Series. Mafic magmas may be hot enough to be entirely or almost entirely molten, but they are potassium-poor.
I agree that if a K-Ar isochron includes phenocryst minerals with widely varying potassium concentrations, no mixing would fool it. This is one reason why it would be good to know how many of the dates from Harland et al (1990) on the Phanerozoic are from multiple mineral isochrons. However, if the phenocrysts were really xenocrysts, then they might inherit an earlier K-Ar isochron. If they only partially degas, and all of them lost a roughly similar proportion of their argon, then a spurious earlier age might be partially inherited in the isochron.

There may also be other ways to get spurious K-Ar isochrons; see for example Dickin (1995, p. 250). Despite this, Dickin states that K-Ar isochrons are a useful test for excess argon.

It is also possible that one cannot get an isochron because the points do not line up. In this case, excess argon cannot be detected.

Agreements among Methods

There are only a limited number of locations where datable samples are found that can unambiguously be assigned to a geological period and thus can potentially contribute to the geological time scale. This means that the geological time scale on the Phanerozoic is based on a very small selection of samples (less than 800 for the time scale in Harland et al (1990), which includes all the dates from earlier studies that they could).

Concerning their procedure for selecting dates from earlier collections and papers, Harland et al (1990, p. 79) state

If we have erred it is in the direction of including items that more critical review have excluded as unsuitable. Our approach is a somewhat `democratic' one. As many time scale items as possible are allowed to influence the calibration.
However, they excluded items that are "clearly anomalous" or that "lie more than two to five time scale divisions away from any likely time scale." They ended up with less than 800 dates, after including all these dates from earlier papers and studies that they could.

With such a small sample size, and a number of degrees of freedom in selecting the dates, it is not surprising that geologists can construct a plausible geological time scale. Some of the dating methods, such as K-Ar dating, may tend to give older dates with increasing depth on this small set of samples, for reasons that have nothing to do with a true age. Also, geologists can choose which minerals to date, and to some extent can choose how they are dated. They can also explain away dates that do not agree with others as inherited or influenced by later events. Some dates may not even be published, and some published dates can be regarded as anomalous. With so many layers of selection, interpretation, and rationalization, there is a substantial danger that the pattern perceived is a product of the measurement process and not of the data itself.

Magnetic separation is a technique applied to select zircon grains for dating. Another technique applied is to remove the outer layers of the grains. Between the two techniques, there are four possibilities for obtaining dates from zircons. Even if these techniques have reasonable justifications and are applied consistently (which would surprise me), there is still a substantial opportunity for choosing the combination that gives dates most in agreement with expectations and justifying this choice "a posteriori."

Even after all of these techniques have been applied, the results still do not completely agree. For example, Harland et al (1990, p. 109) state "The glauconite dates in the database systematically give younger chronogram ages than the non-glauconite dates (Table 5.2)." The largest difference in the time scale would be 17 million years out of about 148 million years. (page 110). Many other examples of discordances are known, as well.

Without an exhaustive survey of all published and unpublished dates, we really do not know how often the various methods agree, or how often they disagree by various percentages. Nor do we know the prevalence of excess argon, as evidenced by K-Ar ages older than the current geological time scale permits.

If there are correlations between different dating methods, they can be caused by common physical and chemical properties of the various systems. Woodmorappe (1999, p. 52) states

Thus, Nd, Pb, and Sr isotopes are subject to similar fractionation processes within the earth .... There are also geochemical similarities between the behaviors of many different isotopes on a subcrustal scale. For instance, the elements K and Rb tend to behave similarly during granulite-facies metamorphism .... It is also interesting to consider the fact that the daughter elements (at least Sr and Pb) are much less mobile than the parent elements (Rb, U, and Th) ... . All of these facts add up to a pattern of concerted behavior among the isotopes used in isotopic dating.
Such common properties might not result in dates that agree, but they might tend to make one system give higher dates when another system does, thus resulting in a correlation between different dating systems.

References

Dickin, Radiogenic Isotope Geology, Cambridge University Press, 1995, 452 pp.

Faure, Principles of Isotope Geology, Wiley and Sons, 1977, 589 pp.

Harland, W.B.; R.L. Armstrong; A.V. Cox; L.E. Craig; A.G. Smith; and D.G. Smith, 1990, "A Geologic Time Scale 1989," Cambridge University Press, Cambridge, 263 pp.

McDougall, I. and T. M. Harrison. 1988. Geochronology and Thermochronology by the 40Ar-39Ar Method. New York: Oxford University Press, 212 pp.

Woodmorappe, J., The Mythology of Modern Dating Methods, Institute for Creation Research, 1999, 118 pp.

Back to home page.