Extending the Krogh Legacy: Development and Applications of "Chemical Abrasion TIMS" Analysis of Zircon
Quite simply put, Tom Krogh has been the single most influential figure in zircon U-Pb geochronology in the past 40 years. Fundamental breakthroughs such as the 1973 paper on hydrothermal decomposition of zircon, and the 1982 paper on air abrasion have huge citation records. But equally impressive is the fact that Tom's entire career was essentially a continuous record of improvements, both large and small, in techniques and understanding of the U-Pb system, chiefly in zircon. This role model is as important a part of Tom's legacy as his specific breakthroughs. Tom's original techniques for zircon dissolution continue unchanged, except for miniaturization, in all labs around the globe. Tom's air abrasion technique was so successful, that it almost completely halted research on all other approaches to "creation of more concordant systems…" such as chemical leaching or partial dissolution. Air abrasion works by physically removing the outer, usually more U+Th-rich zones of zircons. However, some zircons contain high-U+Th regions deep within the grains, and Pb loss associated with these domains could not be removed by air abrasion. Attempts at leaching or partial dissolution were commonly defeated by problems with preferential leaching of Pb relative to U, evidently due the radiation-damaged sites occupied by radiogenic Pb. "Chemical Abrasion TIMS" (CA-TIMS) uses a crucial high-T step that anneals most of the natural radiation damage, completely eliminating U-Pb leaching effects for all but very badly damaged zircon, but without disturbing the U-Pb system itself. This allows the most damaged (and most discordant) zircon to be dissolved preferentially, leaving behind a residue of "closed- system" zircon for single or multi-step analysis. The CA-TIMS technique has now been tested in numerous labs, yielding precise and accurate results for demanding studies ranging from time-scale work using single zircon crystals from volcanic ashes, to evaluation of the decay constants for 235U and 238U.
Geochronology from the Masses - a Pb-Pb Manifesto
The techniques introduced by Tom Krogh made it possible to extract precise (ca. ±0.1%) ages from zircon, thereby revolutionizing our ability to understand geologic time. These methods rely on the use of isotope dilution thermal ionization mass spectrometry (ID-TIMS). A number of other zircon dating methods have since been introduced but up to now ID-TIMS has remained the mainstay for precise geochronology. It is also by far the most time-consuming method, requiring specialized skills for working in an ultra-clean environment. Advances in laser ablation and multi-collector inductively coupled plasma mass spectrometry (LA-MC-ICPMS) promise eventual in situ dating of zircon at ±0.1% with far less time and effort. Use of TIMS can continue only if the method is made much more efficient or if the routine precision of ages is increased by another order of magnitude. Both objectives can be readily achieved for most of geologic time (the Precambrian) with a little imagination and a shift in perspective. U isotopic information makes an insignificant contribution to the determination of most Precambrian zircon ages. This is because 207Pb/206Pb ages can be measured more precisely than 206Pb/238U ages on very old rocks and are not sensitive to recent isotopic disturbance. In practice, one tries to obtain data that are as concordant as possible and then effectively average the 207Pb/206Pb ages. U information provides a level of redundancy that allows us to determine discordance but, because data can be biased by slightly older inheritance in addition to Pb loss, the reproducibility of the 207Pb/206Pb ages is our only assurance of accuracy with or without U. This is important to realize because zircon geochronology by TIMS is potentially much easier without the 'ID'. Pb is a relatively volatile metal, making possible the vapour transfer of analyte at >1000°C under totally clean conditions and with little effort, as opposed to aqueous transfer, which requires laborious clean chemistry. Furthermore, zircon is almost ideally suited to this approach because instead of melting, it undergoes a transition to ZrO2, rapidly releasing SiO2 and radiogenic Pb above 1500°C. The temperature of this transition is lower in altered than in fresh zircon, so isotopically disturbed Pb can be differentially removed at the atomic scale and undisturbed zircon subsequently dated as the reaction front moves from the surface to the interior of the crystal at higher temperature. The Pb evolved in this way is almost entirely radiogenic and should not fractionate because of the low diffusivity of Pb in zircon on the laboratory time scale, so the accuracy of ages is potentially limited only by the size of the sample. This approach has been realized for over two decades and is the basis of the Kober Pb evaporation method. However, Kober method was only developed to a basic level where it produces ages of moderate precision. It has been generally ignored by the specialist geochronology community, in large part because it produces only Pb-Pb ages. Improvements to this approach, which can potentially allow age precisions approaching the limit imposed by initial isotopic disequilibrium in the U decay chains, will be discussed. Their practical realization is technically difficult, but far less challenging than the problems of building plasma and secondary ion-based instruments. The utility of increased age precision will show itself as such data are applied, revealing answers to questions that we have not yet thought to address.
Revisiting High Temperature Annealing to Improve the Chemical Abrasion Method
Accurate age determination of zircon using U-Pb and Pb-Pb systematics is often challenged by radiation damage leading to alteration and Pb loss. This issue is particularly important in dating Precambrian zircon with pervasive damage where differential expansion of high-U zones results in cracking and penetrative access for fluids that alter metamict crystal domains. Various methods have been developed to counter this problem. One of the most effective, the chemical abrasion method (CA-TIMS), was recently pioneered by Mattinson (2005 Chem Geol Vol 220, 47). Grains are annealed at 1000° to 1100° for 1 to 2 days and altered domains along with discordant Pb are subsequently dissolved by stepwise leaching in HF over progressive time-temperature windows. This method works extremely well for Phanerozoic zircon but for highly damaged Precambrian zircon there remains a potential danger of excessive or complete dissolution of sample while HF leaching due to incomplete annealing. Moreover following a complete set of leaching steps is time-consuming so most users are likely to try to achieve this with a single leach under conditions used for dissolution of whole zircon but for a limited time. High-U (>1000 ppm), Neoarchean and Paleoproterozoic zircon annealed at 1000° for periods from 1 to 10 days was observed to develop a greenish tinge after annealing and in some cases dissolved in less than 1 hour in conc. HF at 200°. We are carrying out experiments to test the effectiveness of annealing at much higher temperature. Unannealed zircon grains and grains annealed at 1000° for periods from 1 to 10 days (depending on the nature of damage) were subsequently annealed at 1450° for 1 hr in vacuum. At this temperature zircon slowly breaks down into ZrO2, evaporating silica (Chapman and Roddick, 1994; Ansdel and Kyser, 1993). The residual grain is coated with a thin layer of ZrO2 crystals. U-Pb analysis of several low-U (ca. 100 ppm) grains (using the annealing technique mentioned above) gave accurate Pb-Pb ages, indicating that essentially all the discordant Pb had evaporated, but the discordance is around 10%, indicating that Pb was preferentially removed relative to U. Leaching in conc HF at 200° for 1 hour removes all visible trace of ZrO2, leaving clear grains with a rough, scalloped surface texture. U-Pb analysis of these grains gave concordant data, showing that U-Pb isotopic equilibrium can be restored by this process. Experiments with high-U zircon were less successful. HF leaching of Sudbury Norite zircon dissolved visible ZrO2 but residual grains were white in colour and had a fine-grained, almost porous surface texture. The dissolution rate of these grains was slowed by the annealing process but they dissolved in a few hours. U-Pb analysis of leached grains produced highly (40-60%) discordant data, although Pb-Pb ages were correct within error. Our preliminary interpretation is that annealing zircon at 1450° for 1 hour is slightly less effective at healing radiation damage than annealing at 1000° for several days but still effective enough to permit effective chemical abrasion of zircon with low to moderate amounts of radiation damage. Zircon with high levels of radiation damage may not be reparable at any temperature but it can still potentially be dated using only 207Pb/206Pb age information. Ref: Ansdell K.M., Kyser T.K. 1993. American Minerologist.78, 36-41 Chapman H.J., Roddick J.C. 1994. Earth and Planetary Science Letters 121, 601-611 Mattison J.M. 2005. Chemical Geology. 220, 47-66
High-precision U-Pb geochronology of Cretaceous-Paleocene: Challenges and promise
High-precision U-Pb geochronology has become a crucial tool for examining rates of change in deep time for a broad range of earth system processes, from climate to biologic evolutionary patterns. It is now possible to determine the age of ash-beds intercalated with sedimentary rocks at the precision level of 0.1% or better. When integrated with astronomically tuned cyclostratigraphic, magnetostratigraphic, biostratigraphic, and chemostratigraphic records there is great potential for even higher resolution. However, this approach will require much effort to resolve systematic biases between the U-Pb and 40Ar/39Ar isotopic systems, as well as high-precision calibration/testing of astronomical and geomagnetic timescales. For zircons from a ca. 50 Ma ash bed it is possible to achieve weighted mean 206Pb/238U dates with internal uncertainties as low as 30 kyr (0.06%). At this level of precision, accurate corrections for Th-disequilibrium and laboratory blank, as well as recognition of small amounts of Pb-loss and pre-eruptive zircon history are all crucial. Despite the potential for high levels of precision, interlaboratory bias often exceeds analytical precision. The EARTHTIME initiative is a community based effort attempting to eliminate interlaboratory bias through cooperation and the use of shared, calibrated tracers for ID-TIMS analysis. If successful the next decade will see a major re-sequencing of the last ca. 100 Ma of earth history. The latest Cretaceous through Eocene is characterized by major biotic events including the K-T extinction, the radiation of mammals, and the emergence of modern flora. This period also saw dramatic climatic variations, which provide invaluable context for understanding the relationships between marine and terrestrial records and the dynamics and evolution of greenhouse climate systems. High-precision sequencing and calibration is critical for determining the timescales of these events. Existing age models rely on a small number of 40Ar/39Ar age constraints on the Geomagnetic Polarity Time Scale (GPTS) complemented by interpolation and/or astrochronological models. Large uncertainties exist due to a general lack of dates from key magnetic polarity/fossil intervals that limit robustness of interpolation between tie-points; major improvements are possible. We have made considerable progress in calibrating Cretaceous-Eocene history of the Rocky Mountain region of North America using an integration of U-Pb geochronology, magnetostratigraphy, and astrochronology.
Time-scale calibration by U-Pb geochronology: Examples from the Triassic Period
U-Pb zircon geochronology, pioneered by Tom Krogh, is a cornerstone for the calibration of the time scale. Before Krogh's innovations, U-Pb geochronology was essentially limited by laboratory blank Pb (typically hundreds of nanograms) inherent in the then existing zircon dissolution and purification methods. The introduction of high pressure HF dissolution combined with miniature ion exchange columns (1) reduced the blank by orders of magnitude and allowed mass-spectrometric analyses of minute amounts of material (picograms of Pb and U). Krogh also recognized the need for minimizing the effects of Pb loss, and the introduction of the air-abrasion technique was the method of choice for two decades (2), until the development of the combined annealing and chemical abrasion technique resulted in essentially closed system zircons (3). These are the prerequisite for obtaining precise (permil-level) and accurate radio-isotopic ages of individual zircons contained in primary volcanic ash deposits, which are primary targets for the calibration of the time scale if they occur within fossil bearing sediments. A prime example is the calibration of the Triassic time scale which improved significantly using these techniques. The ages for the base and the top of the Triassic are constrained by U-Pb ages to 252.3 (4) and 201.5 Ma (5), respectively. These dates also constrain the ages of major extinction events at the Permian-Triassic and Triassic-Jurassic boundaries, and are statistically indistinguishable from ages obtained for the Siberian Traps and volcanic products from the Central Atlantic Magmatic Province, respectively, suggesting a causal link. Ages for these continental volcanics, however, are mostly from the K-Ar (40Ar/39Ar) system which requires accounting and correcting for a systematic bias of ca 1 % between U-Pb and 40Ar/39Ar isotopic ages (the 40Ar/39Ar ages being younger) (6). Robust U-Pb age constraints also exist for the Induan- Olenekian boundary (251.2 Ma, (7)) and the Early-Middle Triassic (Olenekian-Anisian) boundary (247.2 Ma, (8, 9)), resulting in a surprisingly short duration of the Early Triassic which has implications for the timing of biotic recovery and major changes in ocean chemistry during this time. Furthermore, the Anisian-Ladinian boundary is constrained to 242.0 Ma by new U-Pb and 40Ar/39Ar ages. Radio-isotopic ages for the Late Triassic are scarce and the only reliable and biostratigraphically controlled age is from an upper Carnian tuff dated to 230.9 Ma (10), yielding a duration of more than 35 Ma for the Late Triassic. The resulting time-scale is at odds with the most recent compilation (11) but arguably more accurate because it is entirely based on U-Pb analyses applied to closed-system zircons with uncertainties at the permil level or better. 1. T. E. Krogh, Geochimica et Cosmochimica Acta 37, 485 (1973); 2. T. E. Krogh, Geochimica et Cosmochimica Acta 46, 637 (1982); 3. J. M. Mattinson, Chemical Geology 220, 47 (2005); 4. R. Mundil, K. R. Ludwig, I. Metcalfe, P. R. Renne, Science 305, 1760 (2004); 5. U. Schaltegger, J. Guex, A. Bartolini, B. Schoene, M. Ovtcharova, Earth and Planetary Science Letters 267, 266 (2008); 6. R. Mundil, P. R. Renne, K. K. Min, K. R. Ludwig, in Eos Trans. AGU, Fall Meet. Suppl. (2006), vol. 87(52), pp. V21A-0543; 7. T. Galfetti et al., Earth and Planetary Science Letters 258, 593 (2007). 8. M. Ovtcharova et al., Earth and Planetary Science Letters 243, 463 (2006). 9. J. Ramezani et al., Earth and Planetary Science Letters 256, 244 (2007). 10. S. Furin et al., Geology 34, 1009 (2006); 11. J. G. Ogg, in A Geologic Time Scale 2004 F. M. Gradstein, J. G. Ogg, A. G. Smith, Eds. (University Press, Cambridge, 2004) pp. 271-306.
The decay constant of 87Rb
Despite dozens of measurements of the decay constant of 87Rb (λ87), uncertainty surrounding the value remains. Mounting evidence [e.g. 1,2,3] suggests that the actual value is 1-2% lower than the conventional value of 1.42 × 10-11a-1 . Increased precision and accuracy are crucial if meaningful comparisons are to be made between Rb-Sr and U-Pb ages. We have been working on measuring the decay constant by the accumulation of radiogenic 87Sr (87Sr*) in a RbClO4 salt. Our original measurements by this method had large errors [5,6] and tended to agree with the conventional value. Because the samples contained very little common Sr, it was impossible to properly correct for instrumental fractionation, with the result that both precision and accuracy were compromised. Furthermore, the concentration of the 84Sr spike was not determined reliably, which likely affected the accuracy. In order to overcome this, a new 84-86Sr double-spike was prepared, and the experiment was repeated. The spike was calibrated against three different Sr reference solutions. Two were prepared from Sr metal and the third from SrCl2. The isotopic abundance ratios of the 84-86Sr double-spike are: 84/86 = 0.93252, 87/86 = 0.01033, and 88/86 = 0.02240. The concentration was determined to be 832.95 ± 0.26 ng Sr/g solution (MSWD = 2.5). Seventeen measurements of the decay-constant were made by measuring 87Sr* ingrowth in a RbClO4 salt over approximately 32 years. 87Sr* ranges from 125 - 616 pg. The two highest points are eliminated: one due to high procedure blank and the second due to abnormal fractionation behaviour. A weighted average of the remaining fifteen measurements yields a decay constant of 1.3981 × 10-11a-11 ± 0.0009 (0.062%; and a high MSWD = 106. The 2σ standard deviation is 0.004). The data scatter outside of their analytical errors. Recent geological calibrations [1,2] and a carefully controlled decay counting measurement  yield λ87 values from 1.395 ± 0.006 to 1.398 ± 0.003. There are no systematic differences between the results of these methods, suggesting that the systematic errors do not exceed random analytical errors. The current measurement also lies in that range, and is the most precise among the published values. The conventional λ87 value, however, lies outside of this range, and requires re-evaluation in light of the current work and the additional recent measurements by other methods.  Nebel et al. (2006) Fall AGU abst. V21A-0558.  Amelin and Zaitsev (2002) GCA 66, 2399.  Kossert (2003) Appl. Rad. Isot. 59, 377.  Steiger and Jäger (1977) EPSL 36, 359.  Rotenberg et al. (2004) 32nd Int. Geol. Congr. abst. 172-5.  Rotenberg et al. (2005) Goldschmidt abst. p. A326.