The natural thermoluminescence of Antarctic meteorites and their terrestrial ages and orbits: A 2010 update


  • Derek W. G. SEARS,

    Corresponding author
    1. Arkansas Center for Space and Planetary Sciences, University of Arkansas, Fayetteville, Arkansas 72701, USA
    2. Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, Arkansas 72701, USA
    Search for more papers by this author
  • Jordan YOZZO,

    1. Arkansas Center for Space and Planetary Sciences, University of Arkansas, Fayetteville, Arkansas 72701, USA
    2. Department of Geosciences, University of Tulsa, Tulsa, Oklahoma 74104, USA
    Search for more papers by this author
  • Christina RAGLAND

    1. Arkansas Center for Space and Planetary Sciences, University of Arkansas, Fayetteville, Arkansas 72701, USA
    2. Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, Arkansas 72701, USA
    Search for more papers by this author

Corresponding author. E-mail:


Abstract– We have examined the relationship between natural thermoluminescence (TL) and 26Al in 120 Antarctic meteorites in order to explore the orbital history and terrestrial ages of these meteorites. Our results confirm the observations of Hasan et al. (1987) which were based on 23 meteorites. For most meteorites there was a positive correlation between natural TL and 26Al, reflecting their similarity in decay rate under Antarctic conditions and thus in terrestrial age. For a small group with low TL and high 26Al a small perihelion was proposed. Within this group, natural TL decreases with terrestrial age as determined by 36Cl measurements, although the rate of TL decay is faster (half-life approximately 10 ka) and the ages that can be determined are smaller (<200 ka) than for most meteorites. The faster decay rate and lower natural TL levels are a reflection of recent exposure to higher radiation doses and higher temperatures, since this history would populate less stable TL traps with smaller electron densities. We sort the 120 meteorites by perihelion and terrestrial age. The normal perihelion group range up to approximately 1000 ka and the small perihelion group range up to approximately 200 ka. An intermediate perihelion group tends to have short terrestrial ages (20–60 ka). There is acceptable agreement between most (34 out of 43) of our present terrestrial age estimates and those determined by isotopic means, the exceptions reflecting complex irradiation histories, long burial times in the Antarctic, or other issues.


The terrestrial ages of meteorites, the time that has lapsed since they fell to Earth, is of interest because of its relevance to (1) secular variations in the nature of material falling on Earth (Dennison and Lipschutz 1987; Lipschutz and Samuels 1991; Sears et al. 1991; Benoit and Sears 1992, 1993), (2) the nature and behavior of certain major terrestrial stranding surfaces, like the Antarctic ice sheets, prairies, and deserts (Benoit et al. 1992, 1993a, 1993b, 1993c, 1994; Benoit and Sears 1999), and (3) the task of identifying the members of large showers in the highly populated Antarctic and desert strewn fields (Scott 1984, 1989; Benoit et al. 2000).

However, terrestrial age determination is among the most difficult of the dating methods commonly used with meteorites. The isotopic methods available for terrestrial age determination involve measuring the abundance of radioactive isotopes that were produced by cosmic ray bombardment in space and which decay once on Earth (e.g., Jull 2006). Examples are 14C, 36Cl, and 26Al, which are produced by spallation reactions in space and undergo beta-decay with half-lives of 5.7 × 103 yr, 3.0 × 105 yr, and 7.2 × 105 yr, respectively. If the terrestrial age is comparable with the half-life of a given isotope, then dating is feasible using that isotope. Thus radiocarbon is commonly applicable to prairie state meteorites where meteorites tend to have terrestrial ages of 0–40 ka, while 36Cl and 26Al are commonly applicable to Antarctic meteorites where terrestrial ages range up to approximately 1 Ma. Weathering conditions at the fall site largely dictate the maximum terrestrial age of a meteorite on Earth at that site, although sometimes the dynamics of the surfaces on which the meteorites fall (ice sheet movement, wind, and wind blown sand effects) can also be a factor. Over most of the world, iron meteorites tend to survive longer than stony meteorites because of their greater resilience to weathering but this does not appear to be the case for Antarctic meteorites (Sears 1979).

The major challenge of isotopic methods of terrestrial age determination is determining the value of the isotopic abundance at the time of fall. Since this is dependent on cosmic ray fluence, target chemistry, shielding depths, and the possibility of multiple phases of irradiation under different conditions, the value is highly uncertain even after secondary methods for determining and correcting for these are applied.

In the early 1980s a large influx of Antarctic meteorites started to become available to the scientific community (Cassidy et al. 1977; Cassidy 2003). In 1987 the Antarctic Meteorite Working Group (MWG), an advisory group to NASA, NSF, and the Smithsonian Institution, added natural thermoluminescence (TL) measurements to the preliminary examination program for returned Antarctic meteorites. They did this on the basis of arguments summarized by Sutton and Walker (1986) and a pilot study by Hasan et al. (1987). While petrographic and electron microprobe data provided a geological description of the returned meteorites, natural TL and 26Al determination provided information on radiation and thermal histories.

Natural TL measurements were included in the Antarctic Meteorite Newsletter for 14 years (1987–2001) and data for over a thousand meteorites were reported. Most of the data were discussed in depth in primary publications (e.g., Benoit et al. 1992, 1993a, 1993b, 1993c, 1994). Here we revisit the work that resulted in the inclusion of natural TL in the preliminary examination program for Antarctic meteorites, the relationship between natural TL and 26Al, 22 years and 1000 meteorites later.


The Hasan et al. (1987) data that led to the inclusion of natural TL in the preliminary examination of Antarctic meteorites are shown in Fig. 1. The samples were selected by the MWG in order to encompass the full range of 26Al activities observed to that date. Nineteen of the samples plot on a trend in which natural TL decreases by a factor of 100, while 26Al decreases by a factor of 6. This is consistent with a number of detailed studies over many years that suggested that natural TL, like 26Al, could be used to estimate terrestrial ages (Sears and Mills 1974; Sears and Durrani 1980; Melcher 1981a; Sears et al. 1990).

Figure 1.

 Plot of natural thermoluminescence (TL) against 26Al activity in 23 Antarctic meteorites chosen by the Antarctic Meteorite Working group for a pilot study of the value of natural TL in the preliminary examination of newly returned Antarctic meteorites. The study and detailed results were reported by Hasan et al. (1987) and led to the inclusion of natural TL data in the initial descriptions of newly returned meteorites reported in the Antarctic Meteorite Newsletter for fourteen years (1987–2001). While most meteorites plot on a correlation line between the two parameters, a small subset have high 26Al and low natural TL suggestive of small perihelion. For simplicity, error bars are not shown but can be seen in the Hasan et al. (1987) paper. Tie lines connect meteorites that were separated on the ice sheet but are thought to be part of the same fall.

A second group of samples, six fragments of four meteorites, in Fig. 1 has lower natural TL (<5 krad) than the others and somewhat higher 26Al. At about the time these data were reported to the MWG the Salem meteorite fell and was also found to have both very low natural TL and high 26Al (Pugh 1983; Nishiizumi et al. 1990). Both the low natural TL and high 26Al were attributed to a particularly close solar passage, which would lower the natural TL by thermally draining the signal while solar cosmic rays (better referred to as solar energetic particles, SEP) would drive 26Al to values much higher than normally encountered, especially if the meteorite was small, because SEP have low energy and small penetrability. It thus came to be understood that, in addition to being small, these samples had experienced small perihelia, a topic that has been discussed at some depth (McKeever and Sears 1980; Melcher 1981a, 1981b; Benoit et al. 1991; Benoit and Sears 1997).

The level of natural TL observed in a sample is the result of a competition between build-up due to radiation exposure and decay due to thermal drainage:


where φ (Gy, 100 Gy = 1 rad, a unit of absorbed dose) is the level of natural TL, φs (Gy) is the value of TL at saturation, dimensionless parameter s is the Arrhenius factor, α is the rate constant (s−1) for de-excitation, R is the dose rate (Gy s−1), E is the trap depth (eV), k is Boltzman’s constant (eV K−1), and T (K) is temperature. This expression is described in various books and publications (for example, see Sears and Hasan 1986). Most of the terms in this expression depend on the nature of the absorbing material, which will be fairly uniform among the ordinary chondrites, physical and compositional differences between H, L, and LL chondrites being minor in importance compared to other factors. Of special importance are the dose rate, R, and temperature, T, which have a major influence on natural TL values and lead to a variety of common applications, including terrestrial age.

Thus the level of natural TL in a meteorite when it enters the Earth’s atmosphere is a function of perihelion and the temperature at perihelion can be calculated from:


where ε = absorbed/emitted radiation and d = solar distance. Most meteorites have perihelia 0.8–1 AU, others 0.5–0.6 AU (McKeever and Sears 1980; Melcher 1981b; Benoit et al. 1991; Benoit and Sears 1997).

Since most meteorites have perihelia close to 1 AU when they entered the atmosphere they have fairly similar natural TL values. Assuming this value can be applied to finds, it is possible to use the natural TL values of the finds to estimate their terrestrial ages, the time since they fell to Earth. For a single TL peak, the rate of decay for the TL is given by


and to calculate the actual decay curve the sum of the peaks must be determined. Several studies have shown that natural TL of ordinary chondrites decreases with increasing terrestrial age and the rate of decay decreases as increasingly stable peaks come into effect (Sears and Mills 1974; Sears and Durrani 1980; Melcher 1981a; Benoit and Sears 1993).

In Fig. 2 we show the natural TL as a function of distance from the Sun from which black body temperature has been calculated assuming typical albedos and emissivities. Relevant dose rates have been measured by spacecraft and are typically 5–10 krad yr−1 (Benoit and Sears 1993). Thus meteorites beyond 1.0–1.3 AU will have saturated (maximum) natural TL levels, but moving closer to the Sun they will drop rapidly (Benoit and Sears 1994). Within approximately 0.8 AU they will have values <5 krad. While we assume black body behavior, we suspect deviations are small compared with uncertainties in emissivity and albedo.

Figure 2.

 Natural TL as a function of perihelion distance for two assumed space radiation dose rates (adapted from Benoit and Sears 1993). The natural TL level of meteorites is strongly dependent on storage temperature, especially the temperatures experienced at perihelion. Thus making reasonable assumptions about the thermal properties of the meteorites in space, it is possible to show that at large distances (>1.2 AU) the natural TL will be at saturation levels, but with decreasing perihelion drops several orders of magnitude for relatively short changes in perihelion. Thus for perihelia within, say, 0.8 AU, the natural TL will decrease by about two orders of magnitude.

The largest cause of uncertainty in terrestrial ages determined by isotopic methods is the range of values displayed by observed falls, and this reflects mostly the different levels of shielding experienced by the meteorites in space. The same theoretical and experimental data that are used to determine cosmogenic isotope abundance as a function of depth (e.g., Reedy 1985, 1987; Michel et al. 1996) can be used to calculate total ionization and thus natural TL levels. There are several papers that experimentally explore depth effects on the natural TL levels of meteorites (Valladas and Lalou 1973; Sears 1975a, 1975b; Benoit and Sears 1993). A typical result is shown in Fig. 3 (Benoit and Sears 1993). For an unusually large meteorite, in this case an H5 chondrite, a drop of a factor of about 0.3 is observed for smaller more likely objects the drop is less. This relatively small depth-dependence is not a surprising result, because natural TL is a low energy process. In such cases, the build-up of secondary radiations is offset by the attenuation of the primary radiations.

Figure 3.

 Natural TL as a function of depth inside H chondrites of the sizes indicated. For a large meteorite (approximately 150 cm), the variation is approximately 35%, for smaller meteorites it is less. This is because for natural TL, like low-energy cosmogenic nuclides, the rate of decay due to attenuation of primary cosmic radiation is approximately balanced by the rate of build-up due to secondary radiations. The thermal drainage due to atmospheric heating is restricted to the outer 6 mm or so, and is not significant on these distance scales (Sears 1975a, 1975b). Figure from Benoit and Sears 1993.

Recent laboratory data for 26Al are shown in Fig. 4 (Michel et al. 1996). Since 26Al is a low energy product, the major target being 28Si (with a small mass difference relative to 26Al), there is little depth-dependence in Kynahinya (Fig. 4a). Saturation levels are usually 60 dpm kg−1 in H chondrites and 56 dpm kg−1 for L chondrites, these are the values normally assumed in terrestrial age applications (Jull 2006, quoting values from Evans et al. 1982 and Vogt et al. 1990). In contrast, the small (2.5 cm) Salem meteorite has a very steep profile, the low-energy SEP-like radiations have high flux and small penetrability, and result in 26Al activities as high as 100–150 dpm kg−1 (Fig. 4b).

Figure 4.

 Profiles for 26Al in two meteorites, (a) the L chondrite Knyahinya (radius 45 cm) and (b) the L chondrite Salem (radius 2.5 cm). The data points refer to laboratory measurements, the lines refer to values calculated from theory and laboratory measurements. These images are simplified versions of figures appearing in Michel et al. (1996). For most meteorites, 26Al shows a shallow profile that increases slightly with depth as secondaries contribute more to the 26Al production. For especially small meteorites, solar energetic particles make an important contribution to 26Al production, especially at small perihelia because their flux decreases with an inverse square of solar distance. However, they have lower energy than galactic cosmic rays and thus poorer penetrability, so the profile drops off quickly with depth.

An independent means of examining the value of natural TL in connection with terrestrial age (independent of 26Al) is to compare the natural TL values with terrestrial ages determined by 36Cl (Fig. 5), another isotope for which abundant data exists and which has a decay rate appropriate for Antarctic meteorites (Benoit et al. 1993c). Unlike nuclear decay, the rate of decay of natural TL, once the meteorite has left the relatively high-radiation, low-temperature environment of space and landed on Earth, is temperature dependent (McKeever 1980). Also shown in Fig. 5 are theoretical decay curves for samples at 0 °C and −5 °C. The data for 14 of the samples clearly conform to storage of these meteorites on Earth at these temperatures. The small amount of scatter can be understood in terms of orbital or shielding differences.

Figure 5.

 Natural TL as a function of the terrestrial age of Antarctic meteorites as determined from the abundance of cosmogenic 36Cl. Fourteen of the samples plot along, or reasonably near, theoretical decay curves calculated for Antarctic temperatures of 0 °C and −5 °C. Five meteorites with low terrestrial ages (<200 ka) have especially low natural TL (too low for decay at these temperatures where the natural TL values asymptotically approach approximately 5 krad) and are presumed to be meteorites with small perihelia (Benoit and Sears 1993). Even among these low natural TL, small perihelia, meteorites there appears to be a suggestion of a relationship with terrestrial age, a two orders magnitude decay in natural TL occurring over a time scale of approximately 100,000 yr.

It should be noted that the decay of most meteorites follows the 0 °C to 5 °C curves that tend asymptotically to a natural value of approximately 5 krad. Thus the five meteorites with natural TL<5 krad, some ≪5 krad, cannot be understood by the same kinetics. Instead, the small perihelion that caused their low natural TL values also caused the natural TL to decay with faster kinetics. In fact, we can see from Fig. 5 that in approximately 150 ka the natural TL has decayed by about two orders of magnitude.

This rapid decay of the natural TL of small perihelia meteorites is readily interpretable in terms of the expected differences in thermal and radiation history that might be expected for meteorites of different perihelia. The situation is illustrated in Fig. 6. The TL data are obtained as plots (referred to as “glow curves”) of light emitted as a function of temperature as the sample is heated in the laboratory. The presence of a number of discrete peaks reflects different sites in the mineral lattices that can store electrons that were promoted to metastable energy states by ionizing radiation. The higher the temperature needed to release the excited electrons, in general, the more stable the electrons in that site and the longer the half-life for thermal decay. McKeever (1980) performed an examination of meteorite glow curve structure and kinetics. Eight individual peaks within the natural TL glow curves for Soko Banja and Lost City were identified and kinetic parameters determined. In our illustration in Fig. 6, a typical natural TL glow curve is shown and assumed for approximately 1 AU. The main (low temperature) peak at approximately 250 °C in the glow curve decays with a mean-life of approximately 106 yr under Antarctic conditions. On the other hand, higher dose rates closer to the Sun, say at 0.6 AU, will be a factor of nearly 3 higher than at 1 AU by the inverse-square law. We suggest that these higher dose rates populate lower temperature (approximately 200 °C) peaks than otherwise observed that we calculate would have half-lives approximately 104 yr under Antarctic conditions. These values are consistent with the decay rates observed in Fig. 5. However, the higher temperatures experienced by the meteorites at approximately 0.6 AU during irradiation would additionally mean much lower levels of natural TL. These conclusions are not yet fully quantitative, mostly because dose rate profiles throughout the inner solar system are not well known and shielding effects are also being ignored. The solar distances, glow curve temperatures, and decay rates quoted are examples only. However, these conclusions do provide an explanation of the observations, especially the data in Fig. 5.

Figure 6.

 Examples of “glow curves” (light emitted as a function of temperature as the sample is heated in the laboratory) for meteorite samples at two distances from the Sun intended to explain the low natural TL/high 26Al group of meteorites identified by Hasan et al. (1987). At 0.6 AU from the Sun, for instance, the radiation dose from solar energetic particles will be nearly a factor of 3 larger than at 1 AU and this will result in lower temperature peaks in the glow curve being populated. Also closer to the Sun temperatures are higher and TL intensities will be lower.

The Current Situation

There are 120 ordinary chondrites for which both natural TL and 26Al data currently exist, after the removal of probable pairing suggested in the Antarctic Meteorite Newsletter. They are listed in Table 1 and plotted in Fig. 7. The range of data is very similar to that observed in the original study of Hasan et al. (1987), say 0.1–300 krad for natural TL and 10–90 dpm kg−1 for 26Al. Welten et al. (1995) suggested that there was a negative correlation between natural TL and 26Al, but this is clearly incorrect. There was a particularly large group of meteorites in this database, 12 in number, that preliminary examination indicated might be fragments of a single fall which we refer to as the Elephant Moraine (EET) 90053 L6 chondrite pairing group (Table 2, Fig. 8). (Meteorite name abbreviations are given in the footnote to Table 1.) With one exception, these have natural TL values between 7.1 and 37.8 krad, and 26Al between 59 and 82 dpm kg−1. The exception, EET 90054, has natural TL of 0.3 krad and 26Al of 59 dpm kg−1. We will return to the EET 90053 pairing group below.

Table 1. 26Al activities and natural thermoluminescence values for 120 Antarctic meteoritesa.
  1. aData from the Antarctic Meteorite Newsletter. Volume and issue number given under “Ref.” Meteorite name abbreviations are as follows: ALH = Allan Hills; BOW = Bowden Neve; DOM = Dominion range; EET = Elephant Moraine; LEW = Lewis Cliff; MBR = Mount Baldr; MET = Meteorite Hills; RKP = Reckling Peak.

  2. b“(ave)” indicates that paired meteorites have been averaged.

ALHA7600651 ± 5  0.4 ± 0.11 (3)
ALHA7600811 ± 1  8.5 ± 0.31 (3)
ALHA7600965 ± 5 10.4 ± 0.11 (3)
ALHA77001 (ave)69 ± 6  2.02 ± 0.142 (1)
ALHA7700230 ± 3 17.2 ± 0.44 (1)
ALHA77004 (ave)54 ± 4 33 ± 0.21 (3), 2 (1), 3 (1)
ALHA7700932 ± 2 30 ± 13 (1)
ALHA7711120 ± 7  5 ± 0.16 (2)
ALHA7711229 ± 1 24.5 ± 0.26 (2)
ALHA7715571 ± 3 25.4 ± 0.12 (1)
ALHA7718241 ± 4  1 ± 0.12 (1)
ALHA7725829 ± 2 48 ± 12 (1)
ALHA7726136 ± 4 14 ± 0.32 (1)
ALHA7726247 ± 5 65 ± 32 (1)
ALHA7726851 ± 2  1.2 ± 0.43 (1)
ALHA7727139 ± 2 46.3 ± 0.31 (3)
ALHA7728538 ± 4 31.2 ± 0.12 (1)
ALHA7729463 ± 2  4.8 ± 0.12 (1)
ALHA7804338 ± 3 11 ± 0.13 (2)
ALHA7804733 ± 1  1.2 ± 0.16 (2)
ALHA7807652 ± 4 58 ± 23 (2)
ALHA7810235 ± 3 23.4 ± 0.43 (1)
ALHA7810561 ± 7 45.3 ± 0.53 (1)
ALHA7811242 ± 3 27.9 ± 0.73 (2)
ALHA7811438 ± 2 14.9 ± 0.83 (2)
ALHA7811543 ± 3 48 ± 23 (2)
ALHA7812834 ± 2  4.4 ± 0.23 (2)
ALHA7813461 ± 3 64.8 ± 0.23 (2)
ALHA7825156 ± 6 49.6 ± 0.53 (1)
ALHA7900234 ± 2 60 ± 14 (1)
ALHA7900771 ± 4 27.6 ± 0.14 (1)
ALHA7902553 ± 3  0.33 ± 0.074 (1)
ALHA7903372 ± 4  0.2 ± 0.14 (1)
ALHA7905463 ± 8164.7 ± 0.44 (1)
ALHA8012676 ± 7  3.9 ± 0.15 (1)
ALHA8103756 ± 4 65.7 ± 0.66 (1)
ALHA8109980 ± 4  2.6 ± 0.16 (2)
ALH 8406673 ± 3  0.4 ± 0.19 (1)
ALH 8501643 ± 5 40 ± 310 (1)
ALH 8501766 ± 7  3.6 ± 0.610 (1)
ALH 8501969 ± 5 17.9 ± 0.110 (1)
ALH 8502050 ± 6 39 ± 210 (1)
ALH 8502159 ± 3  0.13 ± 0.0110 (1)
ALH 8502247 ± 4 55 ± 410 (1)
ALH 8502764 ± 4 67 ± 410 (1)
ALH 8502857 ± 4 18 ± 110 (1)
ALH 8502956 ± 3 28.2 ± 0.910 (1)
ALH 85030 (ave)53 ± 4 35.5 ± 1.210 (1)
ALH 8503343 ± 3258 ± 310 (1)
ALH 8503451 ± 2 40 ± 310 (1)
ALH 8503542 ± 2  6 ± 0.110 (1)
ALH 8503659 ± 4 41.7 ± 0.210 (1)
ALH 8503756 ± 3  6.2 ± 0.410 (2)
ALH 8503849 ± 4 27.2 ± 0.310 (2)
ALH 8503952 ± 4 26.6 ± 0.310 (2)
ALH 8504049 ± 3 40.5 ± 0.210 (2)
ALH 8504257 ± 4 48 ± 110 (2)
ALH 8504347 ± 5107 ± 1110 (2)
ALH 8504426 ± 3 20.2 ± 0.310 (2)
ALH 8504527 ± 3 63 ± 210 (2)
ALH 8504665 ± 3 13 ± 310 (2)
ALH 8505450 ± 5  7.7 ± 0.612 (1)
ALH 8506238 ± 2 59 ± 610 (2)
ALH 8507648 ± 6 16.9 ± 0.410 (2)
ALH 8507944 ± 2 92 ± 110 (2)
ALH 8508047 ± 3 57 ± 510 (2)
ALH 8508356 ± 3 52 ± 510 (2)
ALH 8509749 ± 4 93 ± 212 (1)
ALH 8510449 ± 4  0.56 ± 0.0912 (1)
ALH 8512466 ± 3  6.9 ± 0.310 (2)
ALH 8512952 ± 4 36 ± 210 (2)
ALH 8513056 ± 5  0.5 ± 0.110 (2)
ALH 8513348 ± 4 57 ± 412 (1)
ALH 8514252 ± 3 26.2 ± 0.212 (1)
ALH 8514558 ± 3  0.24 ± 0.0412 (1)
ALH 8660345 ± 3 80 ± 211 (2)
ALH 9041158 ± 4 20.5 ± 0.114 (2)
BOW 8580047 ± 4 43 ± 310 (2)
DOM 8550193 ± 6  4 ± 0.510 (1)
DOM 8550259 ± 4 34 ± 210 (1)
DOM 8550439 ± 3 52 ± 110 (1)
DOM 8550654 ± 3 56 ± 110 (2)
DOM 8550749 ± 3  0.7 ± 0.110 (2)
DOM 8550937 ± 3 54.1 ± 0.710 (2)
EET 8753565 ± 3 17 ± 211 (2)
EET 8753633 ± 2  0.54 ± 0.0812 (1)
EET 8753938 ± 2 11.8 ± 0.412 (1)
EET 8754444 ± 2 69.1 ± 0.212 (1)
EET 8754771 ± 4  2.8 ± 0.212 (1)
EET 8754946 ± 3 86.9 ± 0.812 (1)
EET 8755455 ± 2 90 ± 112 (1)
EET 8755562 ± 2 44 ± 112 (3)
EET 8755668 ± 4  8.6 ± 0.212 (1)
EET 8755759 ± 2 35.3 ± 0.512 (1)
EET 8755860 ± 4  0.78 ± 0.0512 (1)
EET 8756458 ± 3 31.6 ± 0.512 (1)
EET 8756656 ± 2 59 ± 612 (1)
EET 8757348 ± 3 53.1 ± 0.912 (1)
EET 8757674 ± 4 58.2 ± 0.512 (1)
EET 8757859 ± 3 25.5 ± 0.312 (1)
EET 8761360 ± 5 12.2 ± 0.112 (3)
EET 8774447 ± 3103 ± 212 (3)
EET 8780542 ± 3  6.4 ± 0.112 (3)
EET 8781850 ± 4135 ± 412 (3)
EET 9001261 ± 4 11.6 ± 0.114 (2)
EET 9003080 ± 6 12.6 ± 0.114 (2)
EET 9003161 ± 4 26.1 ± 0.114 (2)
EET 9005152 ± 3 31 ± 0.315 (2)
EET 90053 (ave)72 ± 4 19 ± 0.115 (2)
LEW 8531955 ± 4  7.3 ± 0.110 (1)
LEW 8532431 ± 1 32 ± 310 (1)
LEW 8601231 ± 2 50.8 ± 0.911 (1)
LEW 8601336 ± 2 93 ± 211 (1)
LEW 8601541 ± 2122 ± 611 (1)
LEW 8602558 ± 5  0.9 ± 0.111 (1)
MBRA7600173 ± 4 10.4 ± 0.31 (3)
META7800350 ± 3 38.6 ± 0.63 (2)
META7800660 ± 4  1.6 ± 0.23 (1)
META7802856 ± 3 22.7 ± 0.73 (1)
RKPA78001 (ave)56 ± 4  3.48 ± 0.0554 (1, 2)
Figure 7.

 Plot of natural TL against 26Al activity for Antarctic meteorites measured over the duration of the natural TL survey of Antarctic meteorites, 1987–2001. Data are taken from the Antarctic Meteorite Newsletter and have appeared in a number of primary publications (Benoit et al. 1992, 1993a, 1993b, 1993c, 1994). The data appear to define a diamond-shaped field with (natural TL, 26Al) values of (200, 50), (10, 90), (0.1, 60), and (10, 10).

Table 2. 26Al activities and natural thermoluminescence values for the EET 95003 L6 paired groupa.
Name26Al (dpm kg−1)NTL (krad)
  1. aData from the Antarctic Meteorite Database.

EET 9005459 ± 30.3 ± 0.1
EET 9005382 ± 47.1 ± 0.1
EET 9020767 ± 47.2 ± 0.2
EET 9011558 ± 37.3 ± 0.1
EET 9015780 ± 410.7 ± 0.1
EET 9012174 ± 415.4 ± 0.1
EET 9015280 ± 517.1 ± 0.1
EET 9015878 ± 419.8 ± 0.1
EET 9007159 ± 432.6 ± 0.1
EET 9020474 ± 533.1 ± 0.1
EET 9013878 ± 635.7 ± 0.2
EET 9017771 ± 537.8 ± 0.1
Figure 8.

 Among the Antarctic meteorites for which we have both natural TL and 26Al data is the large group of “paired” L6 meteorites associated with Elephant Moraine (EET). (Paired meteorites are those that were separate on the ice but thought to have been on the same object when it entered the atmosphere.) Given that these were part of the same object in space, the data provide an indication of the spread in data expected for a single meteorite and therefore the uncertainty on individual data. On the basis of these data, we suggest that EET 90054 is not paired with the others.

In Fig. 9 we compare the full set of data with the data of Hasan et al. (1987), representing the earlier data as two ellipses, one outlining the “main” field (in which natural TL and 26Al activity showed a positive correlation) and another enclosing the meteorites thought to have experienced small perihelia. In addition, the field enclosing the EET 90053 pairing group (excluding the outlier) is shown, the EET 90053 pairing group clearly plots away from the main group having fairly high natural TL and 26Al activities. There are also a group of previously unknown meteorites with natural TL approximately 1 krad and 26Al approximately 60 dpm kg−1. In fact, we suggest that the small perihelia meteorites are plotting on another positive trend with slightly steeper slope than the main trend (Fig. 5). In other words, there is nothing in the existing database to contradict the findings of Hasan et al. (1987) based on 23 meteorites and on which the 14 yr natural TL survey of Antarctic meteorites was based.

Figure 9.

 An analysis of the present natural TL versus 26Al plot based on the work of Hasan et al. (1987) and Fig. 5. The trend through the “normal” meteorites identified by Hasan et al. (1987), the upper trend line in this figure, agrees approximately with the theoretical curves and data for normal meteorites shown in Fig. 5, with the natural TL decaying by just over an order of magnitude in approximately 750 ka (one half-life of 26Al). The lower trend line in the above figure, passing through the EET 90053 group and the small perihelia group of Hasan et al. (1987) agrees with the trend through the small perihelia group of Fig. 5, the natural TL decaying by two orders of magnitude in about one-third of a half-life of 26Al.

Terrestrial Age and the Sorting of Antarctic Meteorites by Orbit

Figure 10 shows the perihelion values of near-Earth asteroids in the JPL online database as of August 20, 2009. Perihelia values range from about 0.1 AU to about 1.3 AU, with a maximum at approximately 1.0 AU. Objects in the 1 AU peak are analogous to the meteorites we would consider “normal” and belonging in the main trend on the natural TL–26Al plot (Fig. 9). Toward the sunward side of this peak is a gradual downward trend in the number of asteroids in each perihelion bin, with slight suggestions of steps at about 0.8 AU and 0.6 AU. If we assume that the step at approximately 0.6 AU represents the small perihelia group of meteorites identified in Fig. 9, then the decrease in natural TL is approximately as expected (a factor of 50–100 is suggested by Figs. 2 and 5) and the number of meteorites involved is essentially as expected (approximately 20%), these “expectations” allowing for the uncertainties in dose rates and the statistics of small numbers.

Figure 10.

 Histogram of perihelia for near-Earth asteroids taken from the online database maintained by NASA’s JPL (retrieved 20 August 2009). While values range all the way down to approximately 0.1 AU the frequency of occurrence peaks at approximately 1.0, with slight shoulders at approximately 0.8 and approximately 0.6 AU. Assuming NEA and meteorites share similar orbital properties, these data are consistent with the range of orbits suggested for Antarctic meteorites on the basis of their natural TL properties, and with the distribution of meteorites over these perihelia.

Knowing the rates of decay of natural TL and 26Al, we can assign ballpark figures for the terrestrial ages of the meteorites in the present database, and we can say something about perihelion values. We indicate the ages in Fig. 11, and we list values obtained this way in Table 3. In Table 3 we also indicate which perihelion group, “normal” (say, 0.9–1.0 AU), “intermediate” (approximately 0.8 AU), and “small” (approximately 0.6 AU), to which each meteorite belongs.

Figure 11.

 The present data with a “calibration” overlaid on the data based in terrestrial ages from Fig. 5 and 26Al, allowing for differences in perihelion. From placement in this grid the nature of the orbit (“normal,”“intermediate,” and “small” perihelion) and an indication of terrestrial age can be assigned to each meteorite. These are listed in Table 3.

Table 3.   Terrestrial ages and perihelion descriptions for meteorites in the present studya.
NameTerrestrial age (ka)Perihelion (AU)
  1. aData read from Fig. 12.

  2. bAverage of several paired fragments.

ALH 84066100–200Small
ALH 8501670–80Normal
ALH 8501750–60Small
ALH 8501920–30Small
ALH 8502040–50Medium
ALH 85021200–300Small
ALH 8502240–50Normal
ALH 85027<20Medium
ALH 8502840–50Medium
ALH 8502930–40Medium
ALH 85030b40–50Medium
ALH 8503340–50Normal
ALH 8503450–60Medium
ALH 85035100–200Medium
ALH 8503620–30Medium
ALH 8503760–70Small
ALH 8503850–60Medium
ALH 8503940–50Medium
ALH 8504050–60Medium
ALH 8504220–30Medium
ALH 8504340–50Normal
ALH 85044200–300Normal
ALH 85045200–300Normal
ALH 8504630–40Small
ALH 8505480–90Medium
ALH 8506290–100Normal
ALH 8507670–80Medium
ALH 8507940–50Normal
ALH 8508040–50Normal
ALH 8508320–30Medium
ALH 8509730–40Normal
ALH 85104200–300Small
ALH 8512430–40Small
ALH 8512940–50Medium
ALH 85130100–200Small
ALH 8513340–50Normal
ALH 8514240–50Medium
ALH 85145200–300Small
ALH 8660340–50Normal
ALH 9041130–40Medium
BOW 8580040–50Normal
DOM 8550130–40Small
DOM 8550220–30Medium
DOM 8550490–100Normal
DOM 8550630–40Medium
DOM 85507200–300Small
DOM 85509100–200Normal
EET 8753530–40Small
EET 87536500–600Medium
EET 87539100–200Medium
EET 8754450–60Normal
EET 8754730–40Small
EET 8754950–60Normal
EET 8755420–30Normal
EET 8755520–30Medium
EET 8755630–40Small
EET 8755730–40Normal
EET 87558100–200Small
EET 8756430–40Medium
EET 8756620–30Medium
EET 8757350–60Normal
EET 87576<20Small
EET 8757830–40Medium
EET 8761340–50Small
EET 8774430–40Normal
EET 87805100–200Medium
EET 8781820–30Normal
EET 9001230–40Small
EET 9003020–30Small
EET 9003130–40Medium
EET 9005140–50Medium
EET 90053b20–30Small
LEW 8531960–70Small
LEW 85324200–300Normal
LEW 86012100–200Normal
LEW 8601390–100Normal
LEW 8601550–60Normal
LEW 86025100–200Small

We can pursue this argument a little further and ask whether the proposed step at approximately 0.8 AU is reflected in the natural TL–26Al data. A possibility is indicated in Fig. 12 where there is a cluster of meteorites between the proposed main group and the small perihelion group trends. What is interesting about this third group is that there is no trend, suggesting that while they are intermediate in perihelion, they are relatively recent falls. These meteorites are listed in Table 4, along with their classifications.

Figure 12.

 Plot of natural TL against 26Al for the present data indicating the two trend lines discussed above but with a fairly tightly grouped number of intermediate meteorites circled. Based on our presently favored interpretations of the significance of the two trends we suggest these are a cluster of meteorites with intermediate orbits and small terrestrial ages. These meteorites are identified in Table 4.

Table 4.   Details of meteorites in the “medium orbit” and “low terrestrial” age clustera.
NameClassTerrestrial age (ka)
  1. aData read from Fig. 12.

ALH 85036H620–30
ALH 85042H520–30
ALH 85083L620–30
DOM 85502L620–30
EET 87555L620–30
EET 87566L620–30
ALH 85029L630–40
ALH 90411L3.730–40
DOM 85506LL530–40
EET 87564L430–40
EET 87578L630–40
EET 90031LL630–40
ALHA77004 (ave)H440–50
ALH 85020H640–50
ALH 85028H640–50
ALH 85030 (ave)H640–50
ALH 85039L640–50
ALH 85129LL640–50
ALH 85142H540–50
EET 90051H640–50

In Fig. 13 we compare the terrestrial ages we obtained from natural TL using Fig. 12 and listed in Table 3 with terrestrial ages obtained by isotopic methods. Of the 43 meteorites for which we were able to obtain terrestrial ages that have been determined by isotopic methods, 34 show agreement between the two methods that is within a factor of 2. In fact, these 34 meteorites plot on a regression line that is indistinguishable from the unity line. Given the generally large uncertainty associated with terrestrial age determinations, this is encouraging agreement. Of the nine meteorites not showing agreement within a factor of 2, one (Allan Hills [ALH] A76008) is noted for having a complex irradiation history (e.g., Polnau et al. 1999), and we suggest this might be true of the three others lying well clear of the unity line (EET 87536, ALH 85130, and AHLA78105). For the sample lying below the trend, ALHA78105, there is another possibility, which is that the meteorite spent much of its terrestrial life buried below the ice. In this case, isotopic measurements would reflect the true age while natural TL reflects time on the surface of the ice. Below the ice, temperatures are sufficiently low that natural TL fading would be reduced several orders of magnitude.

Figure 13.

 Plot of terrestrial ages determined from natural thermoluminescence in the present work (see Table 3) against terrestrial age determined from cosmogenic isotope measurements. For the terrestrial age determined from natural TL, the middle of the range (which covers a factor of 1.5–2.0) is plotted, while for isotopic ages the reported value is plotted and this has a typical uncertainty of 20–70%. Data for isotopic terrestrial ages are from Jull et al. (1998), Nishiizumi et al. (1979, 1989), and Michlovich et al. (1995). Nine outliers are indicated by open symbols and are discussed in the text. The remaining 34 meteorites (out of 43 on the plot) lie within a factor of 2 of a regression line through the data, which is indistinguishable from the unity line.

The cluster of samples numbered in Fig. 13 (ALH 85062, ALH 85038, ALHA77294, EET 87549, ALH 85040) seems to form a discrete population. These five meteorites have very 14C low terrestrial ages (5.5 ± 1.3 for EET 87549 to 11.5 ± 1.3 ka for ALH 85062) compared to their natural TL ages (50–60 ka for EET 87549 to 90–100 ka for ALH 85062). This might be explained by a particularly low shielding that would give low 14C ages. Higher than expected natural TL values seems inconsistent with the data in Fig. 9 which shows that these meteorites plot with the main band described by Hasan et al. (1987). Aside from their anomalously low 14C ages, these five meteorites have little in common, being a mixture of classes and types and from various find sites.


The initial interpretation of the natural TL versus 26Al plot suggested by Hasan et al. (1987) on the basis of 23 meteorites, and consistent with earlier natural TL studies (Sears and Mills 1974; McKeever and Sears 1980; Sears and Durrani 1980; Melcher 1981a, 1981b), is confirmed by the present study of 120 meteorites. Most meteorites (approximately 80%) show a positive correlation between natural TL and 26Al (on a log-linear plot), and this reflects the decay of both as a function of terrestrial age. While the decay of 26Al is first-order and the decay of TL is second-order, both reach detection limits in about 1000 ka for meteorites entering the atmosphere from normal orbits (perihelia 0.9–1.0 AU, say). About approximately 20% of the meteorites have low natural TL and high 26Al consistent with these meteorites having experienced small perihelia. In small perihelia, the 26Al reaches high values (>60 dps kg−1) due to interaction with solar energetic particles, and the lower (less stable) temperature TL is populated, but levels are much lower than for normal meteorites because of much lower net doses. For lower-temperature TL traps, decay rates are much faster, so we observe a trend in natural TL with terrestrial age (determined independently from 36Cl) for the small perihelion meteorites whereby natural TL decreases by a factor of approximately 50 in 100 ka. Thus we are able to sort the meteorites into terrestrial age and orbital types and these are listed in Table 3. In addition to the “normal” and “small” perihelia groups is a group of meteorites with intermediate orbit and small terrestrial age. That a large number of independent meteorites should share these common properties of similar terrestrial age and orbit, suggests their origin on a common asteroid. Estimates of the terrestrial age based on isotopic measurements exist for 43 of the present samples and 34 show acceptable agreement (within a factor of approximately 2) with the natural TL-based estimate. The remaining nine may have experienced complex irradiation histories, extended burial below the ice before recovery, or remain unexplained.

Acknowledgments— We thank Tim Jull and an anonymous reviewer for comments that helped clarify and otherwise improve the paper, Hazel Sears for an internal review and proofing of the paper, and NASA (cosmochemistry) and NSF (REU site) programs for support. We also thank our many colleagues interested in thermoluminescence for many years of support and encouragement during the natural TL survey of Antarctic meteorites and Charles Melcher, Steve Sutton, Robert Walker, Louis Rancitelli, Bill Cassidy, Don Bogard, Hermann Zimmerman, Fouad Hasan, Paul Benoit, and the Meteorite Working Group for their various contributions to establishing the survey and enabling us to participate in this way to one of the great science adventures, the recovery of meteorites from Antarctica.

Editorial Handling— Dr. Edward Scott