Comment on “Isochron dating of sediments using luminescence of K-feldspar grains” by B. Li et al.

[1] In this paper the authors are attempting to develop a new method of optical dating that does not depend on a knowledge of the environmental radiation dose. The basis is that K-feldspar grains of different sizes will have received different radiation doses because the fraction of the decay energy from the decay of 40K and, to a lesser extent, 87Rb absorbed by the grains increases with grain size. This increase is measured and used to calculate an age. This basic idea is sound, but its application is not. [2] The method proposed is a simpler version of a concept developed in an earlier paper [Li et al., 2007] in which the authors deduced that that portion of the stimulated luminescence subsequent to an internal radiation dose does not fade, whereas that subsequent to an external dose fades. The authors offer no mechanism whereby this separation could be achieved, and none seems possible as it flies in the face of conventional radiation physics. [3] Li et al. [2008] present an impressive comparison of ages obtained by this new method with independent ages to provide proof of the technique; I show that it cannot be so. Also, Li et al. show that K-feldspar optical ages calculated using a simple correction for fading based on a model of quantum-mechanical tunneling [Huntley and Lamothe, 2001] generally do not agree with independent ages; I show that this set of age comparisons does not withstand scrutiny. [4] There is pervasive naturally occurring radiation in the environment resulting mainly from the decay of radioactive isotopes of K, U, and Th. The radiation dose received by a mineral can be determined from the effect the radiation has on that mineral. The dose so determined will be the dose since an event which undoes the effect of any previous radiation dose, such as a sufficient heating or light exposure. The radiation frees electrons from atoms, these electrons lose energy in collisions, freeing more electrons, and they eventually come to rest, trapped at defects in the crystal. A typical beta particle from the decay of a 40K nucleus will create of the order of 104 free electrons as a result of this cascade process. The result of one beta decay then is a new defect at the site of the 40K ion and a large number of electrons trapped at other defects. The number of these trapped electrons is a measure of the radiation dose since those traps were emptied, usually either by heat or by light. In optical dating it is those traps which are readily emptied by sunlight that are made use of, and the event being dated is the last exposure to sunlight. In the case of K-feldspars there is a particular defect with an excited state at 1.4 eV for which the trapped electrons are sampled by measuring the luminescence that results from an exposure to 1.4 eV (infrared) photons. [5] In general there are three classes of trapped electrons: (1) those which are freed as a result of thermal excitation at environmental temperatures (these electrons then wander until they are trapped again), (2) those for which thermal excitation is insignificant and the electrons stay put for approximately a million years or more, and (3) those for which thermal excitation is insignificant but for which there is a nearby defect to which an electron can tunnel. The characteristic tunneling time depends on the energy barrier and the distance to the nearby defect; thus, there is a range of such times, including the laboratory time scale to the geological time scale. These trapped electrons are responsible for anomalous fading, the bane of dating with feldspars as electrons leave some of the traps even though any thermal stability test indicates they should not. If this fading is not allowed for, any age obtained will be too low. It is possible to make a measurement of this effect and correct ages for it for samples young enough that the luminescence is on the linear portion of the dose response curve; in practice this means samples with ages under ∼10–20 ka [Huntley and Lamothe, 2001]. [6] The paper by Li et al. [2008] is an attempt to deal with the problem of fading, and, at the same time, the problem that arises when the dose rate from the surrounding environment is unknown. Their thesis is that in any particular K-feldspar grain the effect of the radiation dose arising from a beta decay within that grain is such that there is no anomalous fading, whereas the effect of the radiation dose arising from a beta decay outside the grain is subject to anomalous fading. They offer no explanation as to how this could occur. The effect of a beta particle cannot depend on whether it originates inside or outside the grain. As to the defect left behind at the site of the 40K, it would be incapable of trapping all of the free electrons generated by the decay of the 40K within the grain, and it is unlikely that this defect would have the same unusual 1.4 eV excited state as have the ones made use of in dating K-feldspars. Thus, free electrons generated by the decay of the 40K within the grain will be trapped just like ones arriving from outside the gain. I conclude that the thesis advanced is unrealistic. [7] One method of experimentally verifying the thesis is to plot the ratio of the uncorrected K-feldspar age to the true age as a function of the external dose rate for a particular grain size and fading parameter; if the thesis is correct, then extrapolation of the data should yield unity at zero external dose rate. The authors did such a test and it failed [Li et al., 2008, Figure 10], thus disproving the thesis. [8] So where have Li et al. [2008] gone astray? Some ideas follow from an examination of the ages provided by Li et al. [2008, Table 1]. [9] An age is calculated primarily from the slope of the line F [Li et al., 2008, Figures 4 and A2]. This line consists of values of the equivalent doses measured for different sizes of K-feldspar grains, hence different internal dose rates. In the absence of anomalous fading this slope should yield the correct age even in the absence of any information about the environmental dose rate. Because of anomalous fading these equivalent dose values are expected to be too small, and hence, the slope too small. It is apparently not too small since it gives correct ages. Something must be increasing the slope to yield the correct age. Some suggestions are as follows. [10] 1. Table 1 in the paper by Li et al. [2008] contains K-feldspar ages obtained for 11 samples. For 6 of them (D4, Dgw5, Sm1, Sm2, Sm3, and Sm4), the uncorrected ages agree with the independent ages; this must not be so because of anomalous fading, so there is something amiss. It seems likely that there was something wrong with the equivalent dose determinations, yielding values larger than they should have been. If these values were too large, then the slopes of the lines F would be too large in proportion. [11] 2. Because this is a difference technique, small effects are amplified. For example, in Figure A2 in the paper by Li et al. [2008], a relative error of 3% going from the smallest grains to the largest would yield a 22% error in the slope. There are several ways that such a small error could occur. Earlier I suggested this could arise from a small change in fading rate with grain size, occasioned by different source material for different grain sizes [Huntley, 2011]; the reply to this suggestion [Li et al., 2011] does not disprove it. It seems improbable that this effect would occur consistently for so many samples, so it can be ruled out as a general explanation for all of them. A consistent change in fading rate with grain size could occur if there was an unknown effect associated with the surface to volume ratio, perhaps the result of weathering or laboratory chemical treatment. Another potential source of a small systematic error is the dependence of the laboratory beta dose rate on grain size, though Li et al. [2007] appear to have ruled this out. Other sources of error of this magnitude are possible. I should note that for two samples (Sm0404 and Sm5) the independent ages are not secure as they are based on an unreliable interpretation of magnetic susceptibility measurements. [12] I now turn to the ages presented for 10 of the samples using the conventional technique with a correction for fading using the Huntley and Lamothe formula [Li et al., 2008, Table 1 and Figure 9]. Of these 10 comparisons there are the following. [13] 1. Three samples (WG3, HLD3, and SY3) for which the fading-corrected ages are in satisfactory agreement with the independent age. [14] 2. Four samples (Sm3, Sm4, Sm0404, and Sm5) for which the Huntley and Lamothe formula is applied even though the equivalent doses are too large for the formula to be applicable. As well, for Sm0404 and Sm5 the independent ages are not secure as indicated above. [15] 3. Three remaining samples: Dgw5, Sm1, and Sm2. For Dgw5 the independent age is 8.3 ± 0.3 ka, and the reported fading rate is 2.6 ± 0.1% per decade. With this fading rate the uncorrected feldspar age should be significantly less than the true age (6.5 ka using the Huntley and Lamothe formula) instead of the reported 8.9 ka, which is larger; therefore, at least one of the independent age, the feldspar age, or the fading rate must be wrong. Similar comments apply to Sm1 and Sm2. It is unlikely that these samples do not exhibit anomalous fading, so there is clearly something is amiss with the determination of either the independent ages or the K-feldspar ages. [16] Thus, there is nothing here that throws doubt on the Huntley and Lamothe formula, and the comparisons shown by Li et al. [2008, Figure 9] that throw such doubt are misleading. [17] I conclude that the method proposed by Li et al. [2008] is not supportable, and that there are alternative explanations of their data that do not require the invocation of unconventional radiation physics.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.