Low Temperature Thermochronology: Applications and Interlaboratory Calibration

Zattin M., Balestrieri M.L., Hasebe N., Ketcham R., Seward D., Sobel E., Spiegel C.

During the last week of September, 2007, the thermochronology community met in the Research Area of the CNR (the Italian National Research Council) in Pisa. The meeting was organized in the framework of the IGCP543 project: ?Low-temperature thermochronology: applications and inter-laboratory calibration?. The chosen location was highly symbolic since Pisa was the city where 27 years ago the first international meeting on fission-track analysis was organized. The meeting was split into a workshop preceded by a two days short-course in which the basics of the method were presented. The list of participants (45 people) can be found on the web site of the meeting (www.geomin.unibo.it/ft/workshop.htm). The short-course was attended by 18 students. Ten students were from developing and emerging countries and were supported by funds from IGCP and sponsors.

One of the main topics of the workshop, and, more generally, of the project, was related to interlaboratory calibration of track length measurements. Another debated issue was related to the development of new techniques. Much of the discussion was about the use of LA-ICP-MS to avoid the having to employ irradiation. Additionally, Barry Kohn announced and illustrated a methodology for automatic counting of fission tracks. Fabian Kohlmann presented a new facility at the Melbourne University for californium irradiation to increase the number of confined track lengths suitable for measurement. The use of this facility is offered (for payment) to all interested laboratories. Meinhert Rahn gave some useful recommendations for managing radioactive material, and Istvan Dunkl presented a critical review of the statistical treatment of fission-track data. Methodologic basics were also discussed by Eva Enkelmann and Giulio Bigazzi, who examined the use of age standards with a proposal to return to the absolute approach for age determination and the use of neutron fluence dosimetry, respectively. The use of the new European standard glasses IRMM 540 and 541 was recommended to replace the CN5 and CN2 glasses, which are now quite difficult to obtain. Finally, Jocelyn Barbarand and Bart Hendricks discussed the annealing process. An interesting debate on the influence of pressure on annealing processes followed Bart?s talk. During the workshop we had two very welcome interruptions with two down-under Skype connections: Matthias Raab presented the new web-based forum of the fission-track community and Mike Krochmal sent greetings from Autoscan (one of the sponsors of the meeting).
Below we present some additional details about the main issues discussed during the workshop.

Interlaboratory comparison of fission-track length data: do we need some calibration?

The Pisa workshop was an excellent setting for discussing inter-laboratory comparison and compatibility of fission-track length data, the basis for thermal history inversion. It was at the International Fission-Track Workshop in Pisa in 1980 that the possibility of using a zeta calibration for fission-track dating was first proposed as a means of surmounting a number of deep disagreements about the best means of calculating ages, disagreements that were preventing widespread acceptance of the technique by the larger geological community. The zeta calibration procedures that eventually arose served as a springboard for successful application of fission-track dating over the next two and a half decades.
The need for more rigorous length calibration was tested, and ultimately demonstrated, by an informal experiment during the workshop suggested by Richard Ketcham and carried out with the help of Ray Donelick, who provided grain mounts for measurement, and Massimiliano Zattin and Maria Laura Balestrieri, who brought their fission-track microscope systems to the meeting site, and ten volunteers. The volunteers took time off from the workshop proceedings to perform measurements of fission-track lengths and etch pit diameters on two grain mounts, one with unannealed induced tracks and another with a complex distribution of spontaneous tracks. Some discrepancies between measurements were expected, particularly with the spontaneous mount, but the extent of disagreement among results was both surprising and alarming to all workshop participants, as dispersion far exceeded expectations based on standard uncertainty estimates. For example, even for the supposedly simple unannealed mount, the most discrepant mean lengths reported were more than 1.5 ?m apart, or 11 standard errors as estimated from the individual track measurements for each analyst. Even with these two analyses removed the spread among the remaining ones was still more than 6 standard errors. Divergence among the spontaneous mounts was higher in an absolute sense (over 3 ?m), but lower when normalized for experimental uncertainty due to the larger natural spread of lengths. Measurements of etch pit diameters also showed divergence far exceeding expectation. Tabulation and analysis of the data are continuing, with the ultimate goal of studying the influence that these variations exert on thermal history inversion.
It must be noted that some, and likely the larger part, of the variation observed in the experiment was due to its extenuating circumstances, most notably possible unfamiliarity with the measurement equipment, different levels of experience, and the distraction of the workshop and the desire not to miss too much of it. However, the results served to underscore the need for inter-laboratory length calibration.
A first step in this direction is being prepared by Tony Hurford, who has created a set of three samples of Durango apatite with induced tracks that have undergone varying amounts of laboratory annealing. The intention will be to send aliquots of each sample to fission-track laboratories around the world for measurement. The results will provide further information on how to properly construct an effective and practicable inter-laboratory length calibration methodology.

Towards a calibration system for apatite fission track etch pit diameter

The possible need for a calibration of another fission-track parameter, the etch pit diameter, was also discussed. Ed Sobel presented an experiment performed together with Diane Seward in which they addressed the important question of whether sample preparation, particularly etching conditions, influences Dpar measurements. This includes both minor variations in temperature and acid strength as well as larger variations between the three popular strengths of nitric acid that are presently in use within the fission track community. A related question is whether operator variability significantly affects Dpar measurements. As some annealing models explicitly use Dpar as a kinetic indicator (Ketcham et al., 1999), the question of whether Dpar varies according to either operator variability or etching conditions has serious implications.
Ed and Diane are trying to understand the Dpar etching rate of Fish Canyon Tuff (FCT) and Durango apatite in the three commonly used etchants with the aim of addressing whether these samples can be used to establish a calibration system for Dpar. The preliminary results demonstrate that use of different etching conditions clearly affects measured Dpar values. However, even when they used the identical etching conditions reported by Carlson et al. (1999), their results varied slightly from that work, indicating that other experimental factors are also influential.
Ed proposed the following calibration scheme to correct Dpar_user (value measured by a user) to Dpar_Donelick (equivalent value if measured by Donelick; used as input for modeling). This method was chosen based on a discussion with Donelick (2003, pers. com.). First, Dpar_Donelick (Carlson et al., 1999) is plotted vs. Dpar_user for Durango and FCT. The linear correlation of these 2 samples is then used to correct Dpar_user for unknown samples to an equivalent Dpar_Donelick value. The slope of this curve is forced to pass through (0, 0). However, to be a complete solution this strategy must work with multiple types of apatite and, preferably, multiple etchant strengths. Preliminary results support this approach.
Preliminary recommendations include:
- Etch using a temperature-controlled water bath. Do not use room-temperature etching.
- Etch numerous samples together for inter-sample comparison.
- Etch Durango or FCT in every batch for calibration.
- Only use etchants that have been calibrated; preferably only use 5 or 5.5 N etchant.
These are also likely issues for collecting track length data.
The preliminary attempts to calibrate their Dpar data with Carlson et al.(1999) appear promising. They suggest as a first step those wishing to use Dpar values for thermal modeling (i.e., with data calibrated by Carlson et al. (1999)) must either reproduce the published Dpar values for standards or consider the effect of the mismatch of data sets. Use of temperature-controlled etching conditions and consistent etch times is a prerequisite for Dpar-based modeling.

Laser Ablation Inductively Coupled Plasma Mass Spectrometry for the determination of uranium content

The use of Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS) has been illustrated by Noriko Hasebe as an alternative method to thermal neutron irradiation for determining the uranium content of minerals for fission track thermochronometry. The conference participants generally agreed that there are positive sides to the methodology, the best of which is radioactivity-free experimental conditions.
Several aspects which should be addressed prior to routine analysis were proposed. (1) The choice of instrumental settings, including laser wavelength that may contribute the profile of an ablated pit and aerosol characteristics, and their effects on the obtained data must be investigated. (2) The standard reference material, which is necessary to convert the MS signal intensity to a value suitable for the age equation, should be examined. The NIST standard glass is good enough to yield reliable uranium concentration and possibly concentrations of multiple elements for additional information (e.g., U-Pb age, trace element geochemistry, etc.). On the other hand, the use of the same mineral as that being analysed (matrix-matched) would be preferable to reduce an unknown source of error and help simplify the age equation. A few mineral candidates were proposed for further discussion. (3) The age equation, calculation of errors, and the template of data presentation were discussed. Ages could be calculated by means of a method analogous to zeta calibration, or by a so-called absolute approach. Revisiting the basics on track revelation, mineral characteristics (e.g., specific density and major chemistry) and the decay constant, along with storing LA-ICP-MS data for standard samples, would be helpful to direct the development of the methodology for reliable dating techniques.