Is Ne-21 worth bothering with for exposure dating? Part I
Recently there has been more than the usual amount of attention to cosmic-ray-produced Ne-21, including a bunch of new production rate estimates (Goethals, Niedermann, others; Balco and Shuster; Amidon, Farley, others.) and a somewhat smaller number of new applications (for example, a paper by Balco and Shuster on Al-26/Be-10/Ne-21 burial dating; a paper by Codilean and others that made Ne-21 measurements on numerous single clasts in Namibian rivers; and several papers that seek to use Al-26/Be-10/Ne-21 measurements to unscramble complex exposure histories).
Ne-21 is a potentially extremely useful cosmogenic nuclide for a couple of reasons, the main ones being as follows. First, like Al-26 and Be-10, it is produced in quartz, which means that: i) because quartz is so common, it is nearly always possible to apply Ne-21 measurements in your area; and ii) it is potentially useful for multiple-nuclide applications with Al-26 and Be-10. Second, it is relatively easy to measure with a standard noble gas mass spectrometer such as one might use for argon dating or helium thermochronometry. This makes for a cheaper, faster, and more accessible measurement than accelerator mass spectrometry, and opens the possibility of projects (like the Codilean paper noted above) that involve enough measurements on individual clasts, surfaces, etc. to characterize statistical distributions.
Ne-21 is also a pain in the neck, for a couple of reasons. The most important one is that nearly all minerals contain significant amounts of non-cosmogenic Ne-21. Thus, it is not enough to just quantify the total amount of Ne-21 in a mineral sample — you also have to determine how much of the Ne-21 is actually cosmogenic in origin, and how much is either trapped or nucleogenic. This requires measurement of all three Ne isotopes, that is, Ne-20, Ne-21, and Ne-22, and deconvolution of the various Ne components based on the isotope ratios of Ne produced in various ways. This, in turn, is complicated (to various extents, depending on your mass spectrometer) by isobaric interferences on both mass 20 (doubly charged Ar-40) and 22 (doubly charged carbon dioxide).
The interference issue can be dealt with in a couple of different ways: some mass spectrometers resolve Ar from Ne at mass 20, and one can use a variety of correction schemes based on knowing peak intensities of singly charged Ar and CO2 as well as the double/single charge ratio (see the Balco and Shuster production rate paper for an example). So actually measuring all three Ne isotopes is not that big a problem.
The more serious problem is deconvolving the various sources of Ne-21. Even if straightforward, this process always exacts a penalty on the measurement precision of cosmogenic Ne-21. If we measure 10 units of total Ne-21 at 1% precision, but then find that only one unit is cosmogenic Ne-21 and the other 9 are trapped Ne-21, our 1% precision on the total rapidly gets worse, to something like 10% precision on quantifying the cosmogenic fraction alone. Thus, it gets harder and harder to measure cosmogenic Ne-21 precisely as the ratio of cosmogenic to non-cosmogenic Ne-21 decreases. In practice this means that cosmogenic Ne-21 measurements are very easy when exposure ages are millions of years and erosion rates are a meter per million years. They are very hard when exposure ages are tens of thousands of years and erosion rates are hundreds of meters per million years. In this and subsequent blog entries I will discuss under what conditions it is actually feasible to use cosmogenic Ne-21 to measure erosion rates, exposure ages, and burial ages, that is, when it is worth bothering to make the measurement.
The first part of this is to look into exactly how much trapped, i.e. non-cosmogenic, Ne-21 exists in typical quartz (and other mineral) separates. The following figure shows trapped Ne-21 concentrations in quartz and sanidine samples from a variety of studies. These include data from the BGC lab as well as a number of published papers; the common theme is that measurements of all three Ne isotopes on all of these samples lay on a mixing line between atmospheric and cosmogenic Ne. Thus, the Ne in these samples consists of only two components — trapped Ne of atmospheric composition and cosmogenic Ne — so can be easily decomposed into these components. So here is the figure:
These are histograms of total trapped Ne-21 — summed over all temperature steps for step-heating measurements — for quartz and sanidine from a variety of sources. The top two panels are measurements made at BGC on quartz from the Antarctic Dry Valleys; the upper panel is sedimentary quartz in unconsolidated deposits — colluvium, till, etc. — and the second panel is quartz in Beacon Group sandstone bedrock. There is a lot of trapped Ne-21 in some of this quartz, up to 100 million atoms/g. However, these particular samples also have million-year-plus exposure ages and as much or more cosmogenic Ne-21. The third panel shows measurements (also ours at BGC) on quartz from igneous and metamorphic bedrock and erratics collected by John Stone in the southern Transantarctic Mountains near Reedy Glacier. Below that, we have measurements of Ne-21 in sanidine made at BGC (on Yucca Mountain and Coso sanidines) and by Florian Kober (in a 2005 paper) on Chilean ignibrites. Continuing down we have sedimentary quartz from Missouri (analysed at BGC as part of a burial-dating project). Sedimentary quartz in a colluvial deposit from central Missouri has relatively low trapped Ne, but a sample of chert from a cave in southwest Missouri has the most we have ever observed — nearly 250 Matoms/g trapped Ne-21. The lower three panels are sedimentary quartz from the Rio Lluta system, Chile (another paper by Florian Kober) and quartz in the Fish Canyon (Julie Libarkin’s work) and Bandelier (Bill Phillips’ work) tuffs in the western U.S.
The summary of all this is that there is a very wide range of trapped Ne-21 concentrations. Leavind aside the chert example, sedimentary quartz — here represented by Beacon Group sandstones from the Transantarctic Mountains — appears to be the worst. This is probably because i) there might be some trace minerals other than quartz in the samples, and ii) it seems likely that intergranular silica cement might be more likely to incorporate atmospheric Ne during precipitation in relatively near-surface environments. Pure igneous and metamorphic quartz appears to be the best. Remember, all these samples had undetectable amounts of nucleogenic Ne, so the age or U/Th concentrations should be irrelevant here. Sanidine, at least the few samples that have been measured, is better than sedimentary quartz and seems to be similar to igneous/metamorphic quartz.
So the first thing we learn here is that it’s probably a good idea to avoid sedimentary quartz if you can.
Subsequent posts will go into more depth on the relationship between how much trapped Ne-21 there is and how low a level of cosmogenic Ne-21 you can actually measure.
Amidon W.H., Rood D.H., Farley K.A., 2009. Cosmogenic 3He and 21Ne production rates calibrated against 10Be in minerals from the Coso volcanic field. Earth and Planetary Science Letters, Volume 280, Issues 1-4, 15 April 2009, Pages 194-204
Balco G., Shuster D.L., 2009. Al-26 – Be-10 – Ne-21 burial dating. Earth and Planetary Science Letters 286, p. 570.
Balco G., Shuster, D.L., 2009. Production rate of cosmogenic Ne-21 in quartz estimated from Be-10, Al-26 and Ne-21 concentrations in slowly eroding Antarctic bedrock surfaces. Earth and Planetary Science Letters 281, pp. 48-58
Codilean, A.T. and Bishop, P. and Stuart, F.M. and Hoey, T.B. and Fabel, D. and Freeman, S.P.H.T. (2008) Single-grain cosmogenic 21Ne concentrations in fluvial sediments reveal spatially variable erosion rates. Geology, 36 (2). pp. 159-162.
Kober F., S. Ivy-Ochs, I. Leya, H. Baur, T. Magna, R. Wieler, P.W. Kubik. In situ cosmogenic 10Be and 21Ne in sanidine and in situ cosmogenic 3He in Fe–Ti-oxide minerals
Earth and Planetary Science Letters, Volume 236, Issues 1-2, 30 July 2005, Pages 404-418
Goethals M.M., R. Hetzel, S. Niedermann, H. Wittmann, C.R. Fenton, P.W. Kubik, M. Christl, F. von Blanckenburg. An improved experimental determination of cosmogenic 10Be/21Ne and 26Al/21Ne production ratios in quartz. Earth and Planetary Science Letters, Volume 284, Issues 1-2, 30 June 2009, Pages 187-198.