Written By Aubrey Whymark 2013-2017

The 147Sm-143Nd isotope system

What does it measure?

Samarium-neodymium dating measures the age of the original crustal formation (Shaw and Wasserburg, 1982).

The fundamentals:

Samarium-neodymium dating is based on the decay of 147Sm (with a half-life of 1.06 x 1011 years) to 143Nd. When the Earth’s mantle differentiates to form felsic (silicate-rich) continental crust the Sm and Nd concentrations increase and the Sm/Nd ratio changes from that in the mantle. Samarium is accommodated more easily into mafic minerals and thus during fractional crystallisation the remaining felsic melt is enriched in neodymium.

Once the continental crust is differentiated from the mantle, thus altering the Sm/Nd ratio, the 147Sm continues to decay to 143Nd; but, because the Sm/Nd ratio is now different the 143Nd/144Nd and 147Sm/144Nd ratios will depart from the Chondritic Uniform Reservoir (CHUR) line. The CHUR defines the initial ratios of continental rocks through time as determined from undifferentiated chondritic meteorites. The departures are very small and thus an epsilon notation is used. One epsilon unit represents one part per 10,000 deviation from the CHUR composition. The equations used are beyond the scope of this book, but basically you can see that deviations of 143Nd/144Nd and 147Sm/144Nd ratios from the CHUR will give a model age for differentiation of the continental crust from the mantle.

The Sm/Nd ratios resist metamorphism and subsequent melting of rocks (where differentiation does not take place) and is also generally unaffected by weathering processes (McCulloch and Wasserburg, 1978). The Sm/Nd ratios in a tektite therefore provide the age (or homogenised age in the case of multiple sources) that the tektite-forming rock(s) differentiated from the mantle. The tektite-forming rock might have subsequently been incorporated into numerous sedimentary rocks, been metamorphosed, been weathered, transported, re-deposited and then eventually melted by cosmic impact to produce a tektite, but the Sm/Nd ratio from the time the continental crust differentiated from the mantle will be preserved.

What can be concluded?

The Nd model ages for australites and indochinites range from 1,040 to 1,190 Ma and depleted mantle reservoir Nd model ages range from 1,490 to 1,620 Ma indicating that the source material was dominantly a Proterozoic crustal terrane (Blum et al. 1992). This is in agreement with the conclusion by Shaw and Wasserburg, 1982, that Australasian tektites were ultimately derived from Precambrian craton (which went on to form the target sediment). The relatively narrow ranges in Nd model ages indicate that the source terrains for the parent sediment were of approximately the same age or that transport and sedimentation were effective at homogenizing materials from sources of different ages (Shaw and Wasserburg, 1982).

Rb-Sr and 10Be

If you only have limited time and you're interested in the source crater location then read up on Rb-Sr and 10Be. They are  contradictory, but at the same time very revealing if you think about the logical conclusions. Thorough mixing of surface soils with older jurassic rock clearly did not happen. The rock is poorly mixed in tektites. One has to trust the 10Be results and conclude that the source rock was a very young and rapidly deposited sediment.

The 87Rb-87Sr-86Sr isotope system

What does it measure?

The Rb-Sr model ages define the time of the last significant rubidium enrichment. In an igneous rock this can indicate the time of magmatic differentiation and crustal formation. Sedimentary processes, however, severely disturb the Rb-Sr system, generally by increasing the Rb/Sr ratio in the sediment through preferential Rb uptake, relative to Sr, by clays (Shaw and Wasserburg, 1981, after McCulloch and Wasserburg, 1978). So, the age given by Rb-Sr dating is usually interpreted as being due to a sedimentary process (Whitehead et al., 2000).  The Rb-Sr ratio can therefore be used to constrain the age and provenance of target materials that were impact melted (Shaw and Wasserburg, 1982).

The fundamentals:

(After Blum et al., 1992; Dicken, 2005 and Bjørlykke, 2010).  87Rb has a half-life of 4.88 x 1010 years, decaying to stable 87Sr. When crustal rock forms, magmatic differentiation takes place. Rb substitutes for K in minerals whereas Sr substitutes for Ca. Consequently, as Bowen’s reaction series progresses, the rock becomes increasingly rich in Rb. So, when the crustal rock forms it is enriched in Rb, including the 87Rb isotope, which then decays to 87Sr.

When the rock formed, all the minerals in the rock would have had the same 87Sr:86Sr ratio. The different minerals will contain a different Rb content. The 87Rb decays to 87Sr and so the 87Sr:86Sr ratio changes, but by different amounts in different minerals/rock samples due to the different Rb content. In minerals or rocks of the same age, when 87Sr:86Sr is plotted against 87Rb:86Sr, a straight-line isochron can be drawn through all the samples. Where it hits zero on the 87Rb:86Sr axis the original 87Sr:86Sr ratio is determined. From this the original amount of 87Sr can be determined for each sample and thus the amount of ‘new’ 87Sr can be determined. The exact amount of decay can be calculated and, as the half-life is known for 87Rb, the age of the rock can be calculated.

In studying tektites we wish to utilise this method to determine the age of the sedimentary rock from which the tektite was derived. The principle is basically the same as that outlined above, but to date sedimentary rocks instead of dating minerals with different Rb content we date whole rock samples with different Rb content. Usually these whole rock samples are very carefully selected to encompass only pure claystone and often processed to remove carbonate impurities. Tektites are effectively melted whole rock samples and will therefore encompass a broad array of lithologies present at the impact site.

In acquiring a meaningful geological age for a sedimentary rock, certain criteria must be met. The parent sedimentary rock must have had the Rb-Sr clock reset at the time of deposition. The Rb-Sr clock is reset by the enrichment of rubidium, which typically follows potassium-rich clay minerals. In order to gain a meaningful age it is necessary to homogenise the 87Sr/86Sr value by mechanical mixing. This can produce a horizontal ‘zero’ line at (or close to) the time of deposition. Variable amounts of Rb in the rock then result in a sloped isochron, which yields an age for the rock.

What can be concluded?

Blum et al. (1992) suggests that the last major Rb/Sr fractionation event experienced by target materials, probably coincident with the time of deposition of the sediments, occurred ~170 Ma ago which corresponds to the middle Jurassic. This singular conclusion appears to have heavily focused searches for the source crater on the Indochinese Peninsula. A ~170 Ma age for the source rock would certainly exclude the young, rapidly deposited, sediments in Gulf of Tonkin as a possible source. The influential rubidium/strontium age is therefore worthy of further investigation.
RIGHT: An example isochron diagram showing the evolution of strontium isotopes in the source sedimentary rock of the Australasian tektites. L1 is the original horizontal isochron of the igneous parent rock. Over time the 87Rb decays to 87Sr and, following the white arrows, a sloped isochron (L2), with a flatter slope corresponding to a younger age and steeper slope corresponding to an older age, is formed. The igneous rock is weathered and, following the grey arrows, the rock is enriched in rubidium. The scatter falls between lines L3. The 87Rb (which was enriched) then decays to 87Sr along the black arrows and the scatter falls between lines L4. In order to have gained a meaningful age the 87Sr/86Sr ratios between L3 would need to have been homogenised to a single value by mechanical mixing, forming a single straight line. This would then have evolved to a single straight line L4 instead of the scatter observed. This diagram shows a two-stage rubidium enrichment history for Australasian tektites. Please see text for detailed explanation. After Blum et al. (1992).
Explanation of diagram above right, after Blum et al. (1992):

87Rb is the radiogenic isotope, which decays to 87Sr. 86Sr is a stable isotope. The x axis is the 87Rb (parent isotope) / 86Sr ratio. The y axis is the 87Sr (daughter isotope) / 86Sr ratio.

Line 1 (L1) on the diagram represents the initial line along which the individual mineral constituents of the original igneous rock plot. These are represented by white squares. Note that at this stage different minerals contain variable amounts of rubidium vs. strontium, but all the different minerals have the same homogenous mantle 87Sr/86Sr ratio. Rubidium substitutes potassium in crystal lattices and therefore some minerals rich in potassium, such as orthoclase feldspar and biotite mica, are also enriched in rubidium. A homogenous initial 87Sr/86Sr ratio, i.e. a perfectly horizontal line, is a requirement in order to acquire a meaningful Rb/Sr age.

The white arrows on the diagram represent the decay of 87Rb to 87Sr over time. The amount of 87Sr produced is dependent on the amount of 87Rb in the rock and the amount of time.

Line 2 (L2) on Figure 5.6 represents the isochron. All the different minerals plot along a straight sloped line. These minerals are represented by black squares. The slope of this line represents the age of the original igneous rock. Horizontal is zero age, with the slope becoming steeper / more vertical with increasing age. Obviously the greater the spread of the 87Rb/86Sr ratio, the more accurate the slope and more accurate the age determination. If all minerals had the same 87Rb/86Sr ratio (same rubidium to strontium ratio) then a single dot is produced, with no slope and therefore an age cannot be determined.

The grey arrows represent the weathering of the igneous rock. Weathering does not change the 87Sr/86Sr ratio, but does result in loss of strontium and retention of rubidium as feldspars and mafic minerals are converted to clay minerals. The 87Rb/86Sr ratio can therefore increase significantly.

You will note that at this point that the values, represented by white circles, plot between L3 and L3. The position along the x axis is dependent on the degree of rubidium enrichment. The meaningful age, defined by the straight sloped line isochron, is lost.

This weathering and deposition stage is critical. This is the event we now wish to date, but unless all the white circles plot along a horizontal isochron (i.e. have the same 87Sr/86Sr ratio) the scatter will not produce a straight sloped line and an age cannot be determined. The solution to the problem is reliant on mechanical dispersion and mixing leading to a uniform 87Sr/86Sr ratio. Thorough mixing would, however, also result in a homogenised 87Rb/86Sr ratio. This would not result in a slope, but a single point.

Cordani et al. (2004) studied Holocene pelitic (clay) sediments and their bearing on whole-rock dating. It was found that terrestrial sediments, deltaic sediments and to some extent coastal plain sediments showed no horizontal line, i.e. scatter of the 87Sr/86Sr ratio would result in no meaningful age. It appears that in the restricted depositional conditions of a continental delta, sediments were not adequately dispersed and mixed. In a tidal flat environment a stable 87Sr/86Sr ratio was obtained by mechanical mixing and then the 87Rb/86Sr ratio was altered by halmirolysis, where strontium was partially removed by seawater. This left variably rubidium enriched samples, which produce a horizontal ‘zero’ line. In open marine environments, more time in contact with seawater results in more exchange between clay minerals and seawater. This also results in a flattening of the line to almost horizontal, but 87Sr/86Sr ratios are still very much dependent on mechanical mixing. In deeper water environments, with slower deposition rates and more exposure time to seawater and formation water prior to burial and compaction the 87Sr/86Sr ratios form an almost horizontal line, not far from the seawater 87Sr/86Sr ratios.

In order to date the depositional/diagenetic age of a sedimentary rock one therefore desires to sample a fine-grained ‘mature’ claystone that has been totally homogenised by mechanical mixing and then subsequently variably depleted of strontium by halmirolysis. Claystones are ideal as the constituent clay minerals are enriched in rubidium. Or, even better, a fine-grained claystone in an area of slow deposition where the strontium has been completely exchanged and now reflects a homogenised seawater value.

At this point, one must return to tektites and ask ‘Are tektites formed from melted fine grained claystones?’ Glass et al. (2004) state that the composition of normal Australasian microtektites fall on the border between greywackes and lithic arenites. Lee et al. (2009) concluded that the best fit for tektites from the Wenchang and Kon-Kai areas was a mixture of 47% shale, 23% sandstone, 25% greywacke and 5% quartzite. The most probable environments of deposition were deltaic to shallow marine. In selecting a rock for Rb-Sr dating, if given the choice, one certainly wouldn’t select a whole-rock of tektitic composition.

Blum et al. (1992), on studying tektites from across the Australasian strewn field, concluded that there was no evidence that the samples had the same initial strontium isotopic composition, a basic assumption of Rb-Sr isochron dating is violated, resulting in an age that probably has no geological significance. Blum et al. (1992) then goes on to use four high-SiO2 Muong Nong-type layered impact glass samples collected from one locality. Prior to the impact, these samples were likely to be very closely related stratigraphically and it was concluded that the strontium isotope ratios were consistent. The four samples plotted on an isochron giving an age of 167 ±12 Ma. The fact that these samples were high silica, however, is suggestive of perhaps a higher sand content in the source material. This is not ideal as rubidium enrichment, as found in claystones, is required for dating. If a rock is transported ‘wholesale’, without rubidium enrichment, then the previous sedimentary cycle may be being dated.

So, on the above diagram, the strontium isotopes evolved from the white circles, along the black arrows to the black circles, falling in a scattered pattern between lines L4. Only closely related samples formed a straight line sloped isochron. The age of the event being dated is the last major rubidium enrichment event. This may, or may not, be the last time the sediment was eroded, transported and re-deposited. Given the chemical composition of tektites, the material sampled is far from ideal to obtain the age of the last sedimentary cycle. It is not a pure claystone and may not be enriched in rubidium, it may not be marine and so the process of halmirolysis may or may not have acted and there was likely a very high sedimentation rate, which limits strontium exchange with seawater and might suggest a more immature sediment.

Next, Blum et al. (1992) plotted the reciprocal of the Rb/Sr enrichment factor (1/ƒ) versus the TURSr model age for each sample. Basically for a given uncertainty in the initial 87Sr/86Sr ratio, the uncertainty in the age will be greatest if it has a low 87Rb/86Sr ratio. At infinite 87Rb/86Sr ratio (or 1/ƒ = 0) the uncertainty in the initial 87Sr/86Sr ratio tends towards zero. So, the data plots in a wedge shape, which can be extrapolated to 1/ƒ = 0. Utilising all the data for Australasian tektites the wedge appears to point towards an age of ~170 Ma. The wedge intercept value that best fits the data was then quantified and the minimum wedge angle was calculated at 175 ±15 Ma ago.

The middle Jurassic ages derived by Blum et al. (1992) are not challenged: these ages are evidently correct. What is being dated is, however, questioned. The author questions whether this middle Jurassic age actually represents an averaged age of the sedimentary constituents, effectively the second to last sedimentary cycle. That is to say that the source rock is in fact much younger, in line with 10Be data, and that the source rock was eroded from plus/minus averaged middle Jurassic aged sediments. Caution is suggested in utilising the middle Jurassic age to search for an impact crater. It may be extremely enlightening to sample Plio-Pleistocene sediments in Gulf of Tonkin and surrounding deltaic areas to determine the Rb-Sr age. If the Rb-Sr age of this Plio-Pleistocene sediment is c. middle Jurassic instead of effectively zero then the search for the source crater must be widened to the Gulf of Tonkin.

The Re-Os isotope system

What does it measure?

The Re-Os isotope system has been used to determine and quantify the extra-terrestrial components in impact-derived ejecta (Koeberl and Shirey, 1997).

The fundamentals:

Rhenium has two natural isotopes; 185Re, which is stable, and 187Re, which is radioactive. The Re-Os isotopic system is based on the beta-decay of 187Re to 187Os, with a half life of 42.3 ±1.3 Ga. Osmium has seven naturally occurring isotopes, 184Os, 186Os, 187Os, 188Os, 189Os, 190Os and 192Os, all of which are stable. During partial melting of mantle rocks, the osmium remains in the residue, whereas the rhenium is enriched in the melt. Terrestrial crustal rocks therefore have a high Re concentration and low Os concentration. As a result of the decay of 187Re to 187Os, the crustal 187Os/188Os ratio increases rapidly with time. (Koeberl and Shirey, 1997).
Meteorites have a high osmium content, but because the rhenium (including 187Re) has not been concentrated by fractionation, meteorites have a low 187Os/188Os ratio. Eucritic meteorites are an exception to this rule as, like terrestrial crust, they formed through magmatic differentiation. The large contrast between the typical upper continental crust on the Earth, with a high 187Os/188Os ratio, and that of meteorites, with a low 187Os/188Os ratio, can be used to identify a meteoritic component in an impact-derived rock. (Koeberl and Shirey, 1997).

What can be concluded?

Detailed study of Australasian tektites has not been carried, however, Ivory Coast tektites contain only about 0.6% or less extraterrestrial component (Koeberl and Shirey, 1993). Theoretically the percentage of extraterrestrial component should increase with proximity to the source crater.

A study of Ivory Coast tektites by Koeberl and Shirey (1993) indicated that during the impact process significant amounts of both rhenium and, more so, osmium are lost. Osmium and, to a lesser extent, rhenium are extremely volatile elements as oxides (Lovering and Morgan, 1964). In reducing conditions (such as at high atmospheric levels with a very low partial pressure of oxygen), however, osmium oxides are reduced to osmium metal which is less volatile whereas rhenium oxides are more stable and thus remain volatile, resulting in a greater loss of rhenium compared with osmium (Lovering and Morgan, 1964).  Koeberl and Shirey (1993) state that the fraction of target rock-derived osmium in the Ivory Coast tektites did not exceed 10% of the total, so the bulk (>90%) of the osmium content in the tektites came from the impactor. Whilst the total osmium content is altered during the impact process, the osmium isotopic ratios are not likely to be altered significantly.

Due to the high abundance of osmium in meteorites, tektites required the incorporation of only about 0.6% or less extraterrestrial osmium component to lower the 187Os/188Os ratio from the target rock values to near-meteoritic values in tektites.

Meteoritic component in tektites

Based on Ivory Coast Tektites the Osmium content indicates that these proximal tektites contain 0.6% or less meteoritic component, with the remaining 99.4% being terrestrial rock.
ABOVE: The 187Os/188Os ratios of different rock types and of Ivory Coast tektites and their source rock. From Koeberl and Shirey (1993 & 1997).
There is scope for future work on Australasian tektites. Lovering and Morgan, 1964, studied the rhenium and osmium content of tektites, including Australasian tektites. It is evident that the osmium/rhenium ratio increases with proximity to the source. This might be expected as more proximal tektites will contain an increased extra-terrestrial component. The extra-terrestrial component would contain more osmium relative to rhenium.

A 187Os/188Os study of Australasian tektites would be interesting. As one moves towards the source crater, the meteoritic contamination of the tektite (with lower 187Os/188Os  values) would likely increase.
ABOVE: The Os/Re ratios of Australasian tektites. From Lovering and Morgan (1964).

Oxygen isotopes

What does it measure?

Oxygen isotopes, within the realm of tektite studies, help to identify the target lithologies for terrestrial impacts (Chamberlain et al., 1993). They can also provide evidence to link distal ejecta to a source crater.

The fundamentals:

Oxygen has three stable isotopes, which were originally formed in stars: 16O, 17O and 18O. The primary isotope 16O comprises 99.759% of the Earth’s atmosphere. 17O and 18O are secondary isotopes and comprise 0.037% and 0.204% of the Earth’s atmosphere, respectively. The 18O/16O ratios remain essentially constant during igneous differentiation, with end-member acidic rocks, such as rhyolite, increasing by only 2 to 3‰ (or per mil/tenth of a percent) compared with basic rocks, such as basalt (Taylor and Epstein, 1969). Carbonate and silicate sedimentary rocks, however, are significantly enriched in heavy oxygen (18O) compared with the original igneous rocks. This is due to the preferential incorporation of 18O in low-temperature sedimentary minerals. This, in turn, results in the δ18O (a measure of the  18O/16O ratio) seawater value being 0‰, lower than the mean value of  5.5‰ on Earth (Albarède, 2003).

After the 18O/16O ratios of tektites have been measured, adjustments must be made to allow for the fact that porous sediment may contain meteoric (fresh) water. Meteoric water contains less 18O as lighter oxygen isotopes present in water molecules preferentially evaporate, going on to form rain water. This isotopically light oxygen from meteoric waters is mixed into the target material during tektite formation. The porosity of the target rock may be an important factor controlling the oxygen isotope systematics of tektites (Chamberlain et al., 1993).

What can be concluded?

The δ18O value of Australasian tektites ranges between 9 and 11‰. There are a number of methods that have been proposed in the past by which to reach these values. At the time of writing, however, we have a greater dataset available. Various lines of evidence lead to the belief that Australasian tektites are derived from a sedimentary source rock. With reference to moldavites, which have a known source crater and source rock, one can infer the likelihood that Australasian tektites were derived from sandstones and shales that would be expected to have had original δ18O values of ~14‰. The observed values in tektites are ~4‰ lower and this can be explained by either seawater or meteoric pore water incorporation at the time of melting (Blum & Chamberlain, 1992). Note that the water itself is not retained in the tektite as H2O, as tektites are extremely dry, but the oxygen isotopes, more of which are the lighter isotopes in water molecules, may be retained.
RIGHT: Distribution of δ18O values in selected tektites, igneous, metamorphic and sedimentary rocks and in natural water. Combined data from Albarède (2003)* and incorporating data from Taylor & Epstein (1969)† (various original sources). Ivory coast data extended after Chamberlain et al. (1993).
Taylor and Epstein (1969) observed a systematic increase in 18O with decreasing SiO2 content. At the time it was suggested that this was due to mixing of a SiO2-rich igneous component (e.g. quartz grains) and a low-SiO2 component formed at much lower temperatures. With reference to K-T spherules, Blum & Chamberlain (1992) clarify this point by stating that ‘this trend can only be produced by mixing between carbonate and silicate rocks....’.

The narrow range of δ18O values of tektites (see above figure) is likely explained by the fact that only a narrow range of parent materials, i.e. silica enriched sedimentary sandstones and shales and metamorphic equivalents such as quartzites, schists and phyllites are suitable for tektite production. The presence of carbonates increases the δ18O values of tektites, but too much carbonate and insufficient silica will prevent tektite production. Tektites are known to be produced in the earliest stages of impact from the top 100 m, or so, of sediment, as supported by 10Be studies. Invariably, these shallow sediments are likely to be water-rich, thus lowering the ultimate δ18O value (Blum & Chamberlain, 1992).

Notably, Ivory Coast tektites have a slightly higher δ18O range (11.7-12.9‰ in Chamberlain et al., 1993, and 12.8-14.6‰ in Taylor & Epstein, 1966), compared with the average tektite. This can probably be explained by the fact that Ivory Coast tektites were derived from metasediments, as oppose to the other tektite groups being derived from sediments. Metasediments will be of much lower porosity and, assuming an impact on land, the resultant tektites will be a closer reflection of the parent rock as the δ18O values have not been lowered by the presence of ‘light oxygen’ water.

Lead Isotopes

What does it measure?

The lead isotope system can be used to date the ultimate crustal-forming parent rock from unaltered zircon crystals found in layered tektites. Lead isotopes in tektite glass can also be used to identify U/Th enrichment events which are suggestive of source rock type, which, in turn, indicates terrestrial origin. Lead isotopes are also unique amongst different strewn fields and can therefore be used to link macro- and micro-tektites to a known strewn field.

The fundamentals:

Lead has four stable isotopes, all of which are primordial to some extent, being produced by supernovae. Three of these stable isotopes are also produced by radioactive decay. 206Pb forms by the radioactive decay of 238U  with a half-life 704 Ma. 207Pb forms through the radioactive decay of 235U with a half-life 4.47 Ga. 208Pb forms through the radioactive decay of 232Th with a half-life 14.05 Ga. 204Pb is an entirely primordial isotope, meaning that it has existed in its current form since before the Earth was formed. 204Pb can therefore be used to measure the fraction of other lead isotopes in a sample that are also primordial. Any excess of 206Pb, 207Pb and 208Pb is assumed to be radiogenic in origin. This allows various uranium and thorium dating schemes to be used to estimate the age of rocks.

The isotopic ratios of lead (206Pb/204Pb; 206Pb/207Pb; 206Pb/208Pb) can be compared with modern terrestrial leads from various sources. The ratios in an australite were found to be practically the same as modern terrestrial lead (Tilton, 1958). The uranium-lead and thorium-lead ratios can be calculated for the environments which have produced isotopic values in modern terrestrial lead. Next, the actual uranium-lead and thorium-lead ratios in the tektite can be measured and compared. Any depletion or enrichment of uranium or thorium relative to lead will yield information concerning possible recent changes in the ratios which may have occurred in the source material of the australite.

In order to determine the crustal age, the favoured mineral for uranium-lead dating is zircon (ZrSiO4). Uranium easily substitutes the zirconium, but lead is strongly excluded meaning that when zircon forms the clock is truly set at zero. The clock is not easily disturbed by erosion/consolidation into sedimentary rocks or moderate metamorphism. So, the zircon measures the age of the original crustal growth events – the ultimate parent rock of the zircon. Some unaltered zircons have been found in Muong Nong-type georgiaite impact glasses, but unfortunately zircons in Muong Nong-type indochinite impact glasses have been reset to the time of impact (Deloule et al., 2001).

What can be concluded?

Tilton (1958) analysed an australite. He found that the isotopic composition of the lead (206Pb/204Pb; 206Pb/207Pb; 206Pb/208Pb) in the australite was that of a modern terrestrial lead. This implied that the tektite material was in an environment with “normal” ratios of thorium and uranium to lead until recent times (204Pb, 207Pb and 208Pb  are decay products). The uranium content, however, was enriched by a factor of 4 with respect to lead and the thorium content was enriched by a factor of 1.4 with respect to uranium. To allow for the normal terrestrial lead isotopic ratios this uranium:lead and thorium:uranium enrichment must have taken place within the last 50-100 million years at most (Tilton, 1958). These observations are very consistent with australites being derived from relatively recently deposited argillaceous sediments. It is difficult to explain the uranium-thorium-lead balance with a lunar origin given the lack of sedimentary processes (due to the almost complete lack of an atmosphere) and lack of recent igneous activity on the Moon. Furthermore, even assuming igneous processes were still active on the Moon, the chemical composition of tektites does not correspond to that of igneous rocks.

Stecher and Baker (2004) revisited lead isotopes. It was noted that Pb isotopic results of tektites from different strewn fields plot in small and clearly separated fields in all Pb isotopic ratio diagrams. Pb isotopic measurements can therefore be used to rapidly assign an unknown tektite to its appropriate strewn field. Only the Australasian strewn field showed a small linear trend, with the last formed proximal indochinites being the least radiogenic and first formed distal australites being the most radiogenic. This information assists in locating the position of the Australasian strewn field crater.

In summary, the age of the crustal forming event for Australasian tektites could not be determined due to the alteration of zircon crystals (Deloule et al., 2001). Recent uranium and thorium enrichment of the Australasian tektite source rock suggests a recent sedimentary argillaceous source rock (Tilton, 1958). The absence of an atmosphere and therefore weathering processes on the Moon indicates that the source rock, as it is likely sedimentary, is terrestrial. Lead isotopic measurements are distinct for each strewn field allowing for rapid assignment of a tektite to a strewn field. Finally, Australasian tektites become increasingly radiogenic with distance from the source location assumed to be in the Indochinese region (Stecher and Baker, 2004).

Cosmogenic Beryllium-10

What does it measure?

The ratio of 10Be, with respect to other radioactive nulclides produced by cosmic rays (principally by interaction with 26Al in rocks and spallation of oxygen (16O) and nitrogen (14N) in the atmosphere) allows the source of 10Be to be established (i.e. atmospheric or due to the tektite having spent significant periods of time in space). Having determined an atmospheric origin it can be used to determine the relative depth/age of the sediment form which the tektite was derived. Utilising data from tektites in different areas of the strewn field the general source area can be identified.

The fundamentals:

10Be is formed by the interaction of cosmic rays and may be produced in situ, in a rock, or within the Earth’s atmosphere and later incorporated into a rock. In space where there is no protective atmosphere, in situ production rates are higher than on the surface of the Earth. The ratios of 10Be against other cosmogenic isotopes are used to establish whether the 10Be was produced in situ or within the Earth’s atmosphere.

On Earth 10Be is predominantly produced in the atmosphere by the galactic and solar cosmic-ray spallation of oxygen (16O) and nitrogen (14N). Oxygen makes up 20.95% of the atmosphere, whilst nitrogen comprises 78.08% of the atmosphere. 10Be will readily dissolve in acid rain waters, which transport the 10Be to the Earth’s surface. As the water becomes more alkaline 10Be will drop out of solution. At the Earth’s surface the 10Be atoms cling to almost any silicate grains. These grains will then be transported and re-deposited. The 10Be decays to stable boron-10 by beta decay, with a half-life of 1.36 million years. If sedimentation rates are reasonably fast, the relatively long half life of 10Be can allow a great thickness of 10Be rich sediment to be built up. Of course in most ‘normal’ locations, the deeper the sediment is buried, the older it will be. The older the sediment is, the more 10Be has decayed, so with depth the amount of 10Be decreases.

What can be concluded?

The presence of 10Be, in the absence of other radioactive nulclides produced by cosmic rays, supports a terrestrial origin for tektites. Furthermore it is suggestive that Australasian tektites formed from young, recently deposited sediments as oppose older sedimentary rocks, igneous or metamorphic rocks  (Ma et al., 2004).

When an asteroid impacts, the first formed tektites are derived from the very uppermost layers of sediment. They are ejected at the highest velocity and lowest angle. In the Australasian strewn field these are the distal australites. As the impact event proceeds, progressively deeper sediment is excavated and ejected as tektitic melt. The deeper (older) sediment has lower abundances of 10Be and will go on to form increasingly proximal tektites as the ejection velocity decreases and ejection angle increases. The last formed ‘tektites’ are the Muong Nong-type impact glasses. Average 10Be contents of Muong Nong-type impact glasses are ~1/3 and ~1/2 those of Australian tektites and splash-form indochinites, respectively (Aggrey et al., 1998). The abundance of 10Be can then be used to create a map of roughly equal concentration levels, which increase with distance from the undiscovered crater. Ma et al. (2004) suggests the most probable source is a single crater in the Bay of Tonkin at 107°E; 17°N. The author would broadly agree with this conclusion.

Cosmogenic Aluminium-26

What does it measure?

The ratio of 26Al:10Be indicates whether the 10Be formed in the Earth’s atmosphere (subsequently being incorporated into terrestrial sediment and then impact melted to form a tektite) or within a solid object in space. Terrestrial surface exposure age (time the tektite has spent at or near the surface) can then be calculated by change in the 26Al:10Be ratio.

The fundamentals:

Aluminium-26 is produced by the interaction of galactic and solar cosmic rays with the atmosphere and surface rocks (Middleton et al., 1987). 26Al has a half life of 717,000 years. On Earth, about 4000 times less 26Al is produced in the atmosphere compared to 10Be. This is because 26Al is produced from 40Ar in the atmosphere by spallation caused by cosmic-ray protons. Argon only makes up 0.93% of the Earth’s atmosphere. The 26Al makes its way to the Earth’s surface in rain water in much the same way as 10Be. 26Al can also be formed in situ in solid objects on the ground. The principal reactions producing 26Al are neutron-induced reactions with Al and Si (Jha & Lal, 1981). Muon (an elementary particle similar to an electron) capture by 28Si constitutes about 10% of the total 26Al at the Earth's surface, but becomes increasingly important with depth because of the more-penetrating nature of muons compared with neutrons (Middleton et al., 1987).

Within the field of tektites, the primary use of 26Al, in conjunction with 10Be, is to prove a terrestrial origin. Ratios can be compared to atmospheric abundances, thus proving the source rock is terrestrial sediment. Tektites on the Earth’s surface are shielded somewhat by the atmosphere, but tektites that have been buried or are underwater and further shielded from surface exposure to cosmic rays produce the most accurate values as no additional 26Al is accumulated. Extra-terrestrial material contains a much higher 26Al:10Be ratio than that found in tektites. This is because in extraterrestrial rocks, which are not buried too deeply, the 26Al readily forms in the solid object, whereas 10Be is formed at a much lower rate.

A secondary use of 26Al:10Be ratios is to determine the surface exposure age, i.e. the time a tektite has spent on or very close to the Earth’s surface (within ~1 metre) (Yiou et al., 1984). Production of 26Al in terrestrial surficial rocks is low (compared to the surface of asteroids or the Moon), due to the attenuation of the cosmic-ray flux by the atmosphere (Middleton et al., 1987). Despite this low production rate on the surface of the Earth, the additional 26Al can significantly affect the 26Al:10Be ratio. This is firstly due to the very low initial concentrations of 26Al. Secondly, the in situ production of 26Al and 10Be in solid objects barely affects the concentration of 10Be, but greatly increases that of 26Al, thereby increasing the 26Al:10Be ratio. By understanding the solid object production rates of 26Al and 10Be, the time since the tektite formed and the initial (atmospheric) 26Al:10Be ratio, corrected for decay, the surface exposure age can be calculated. The surface exposure age will be the same time elapsed since its formation or less, dependent on the tektite’s burial history or shielding by overlying water bodies.

What can be concluded?

The 26Al:10Be ratio provides very strong evidence that the tektite source rock is terrestrial sediment. Tera et al. (1983) concluded that the values closely matched those of sediments formed on continental margins. 26Al:10Be ratios prove tektites have not spent any significant time in space. The excess 26Al allows surface exposure age to be calculated on individual specimens – these ages are equal to or less than the formation age, thereby giving a minimum possible age. An australite measured by Middleton et al. (1987) was calculated to have lain on the surface of the Earth for about 0.5 Ma or, on average, had been within the top 23 cm of the surface 0.7 Ma. (the age estimate at the time of writing is closer to 0.788 Ma).

Cosmogenic Manganese-53

What does it measure?

53Mn provides a test for the presence of matter in tektites derived from an iron impactor. (Englert et al., 1984).

This is important because 10Be measurements alone are highly suggestive of the tektite source material being terrestrial sediment. 10Be measurements alone, however, do not identify the location where 10Be production took place, for instance a 1% component of chondritic or lunar material would be sufficient to account for the 10Be content of tektites (Raisbeck et al., 1983).

To close this loophole 26Al/10Be ratios were analysed by Raisbeck et al. (1983), which provides a means of testing the terrestrial or extraterrestrial origin of these isotopes. Basically, 26Al is sparse relative to 10Be in terrestrial rocks, whereas in space (in the absence of an atmosphere) 26Al is much more abundant in rocks relative to 10Be. Results indicated that even if the source of the 26Al was chondritic or lunar (i.e. from the impactor), the bulk of the 10Be and hence the bulk of the tektite was terrestrial, i.e. the amount of 10Be was high relative to 26Al, so even if the 26Al were extraterrestrial it would only account for a small percentage of the 10Be in the tektite.

These results limited the fractions of lunar or chondritic material in Australasian tektites, but they did not rule out iron meteorites as a source of the 10Be (Englert et al., 1984). This problem arises due to the low production rate of 26Al in irons. The production rate of 53Mn in iron meteorites is much higher than that of 26Al and thus 53Mn can be used to further close the loophole and prove the 10Be is predominantly terrestrial.

The fundamentals:

Manganese has one stable isotope and 25 radioisotopes. 53Mn forms due to cosmic ray induced disintegration of iron. 53Mn is the most stable isotope, decaying to 53Cr with a half-life of 3.7 Ma. Due to this short half-life 53Mn only occurs in small amounts.

What can be concluded?

Englert et al. (1984) concluded that the 53Mn content of Australasian tektites at the present day does not exceed 8 dpm/kg Fe (disintegrations per minute per kilogram of iron) and is probably lower by a factor of two. The estimated production rate of 53Mn at the Earth’s surface was calculated as 1.3 dpm/kg Fe.

A weighted average 53Mn/10Be ratio of 2.8 ±3.6 was calculated for Australasian tektites, compared to 260 ±100 for iron meteorites. The 53Mn/10Be ratio for terrestrial rock is poorly constrained, but is believed to be around 100 times smaller than that of iron meteorites. With the assumption that 53Mn and 10Be do not fractionate during tektite formation, it was concluded that iron meteorites supplied less than 1% of the 10Be found in the Australasian samples studied. The 10Be (or at least the vast majority)  must therefore be terrestrial in origin.

Cosmogenic Neon-21

What does it measure?

The presence of 21Ne indicates cosmic ray bombardment, which will primarily occur in space. The amount of  21Ne can be used to determine the amount of time the tektite spent in space.

The fundamentals:

21Ne is a cosmogenic isotope produced in spallation reactions on Mg, Na, Si and Al due to cosmic ray bombardment (Dicken, 2005). It was known that neon can migrate through glass by diffusion. It was therefore necessary to determine the rate of diffusion through glass. In a 1 cm radius sphere, at room temperature the neon diffusion half life was 1.1 million years. Another isotope formed by cosmic ray bombardment was 3He, but unlike neon, helium had a diffusion half life of 5.6 years under the same conditions. 21Ne was therefore a potentially measurable isotope in Australasian tektites, whereas 3He was not (Reynolds, 1960).

In order to calculate the cosmic ray exposure age the production of 21Ne in space must be known and this can theoretically be calculated by bombarding a specimen with high energy protons. The age of the specimen must also be known so the amount of 21Ne can be adjusted for loss.

What can be concluded?

In Reynolds (1960) a maximum flight time of one of the australites was 28,000 years and for a philippinite it was 40,000 years. Minimum flight time was, of course, zero.

Cosmic ray induced tracks

What does it measure?

Tracks produced by cosmic rays are distinctive and allow the time the tektite spent in space, since the last significant thermal event, to be calculated. It assumes the tektite was a discrete body and not part of the interior of a much larger body.

The fundamentals:

Sourced from Fleischer et al. (1965) with reference to Maurette and Walker (1964). High energy protons produce etchable tracks by inducing fission in heavy-element impurities - primarily Th and U (and, to a lesser extent Pb). These induced events have a characteristic appearance which allows them to be distinguished from spontaneous fission tracks such as those produced by the decay of 238U. Spontaneous or thermal-neutron-induced tracks result in a straight line (effectively a 180 degree ‘V’) as the two products of fission move in opposite directions. High energy reactions, involving cosmic-ray particles, transfer sufficient momentum to the struck nucleus that fission takes place whilst in motion. The two fission products therefore diverge in a ‘V’ shape as oppose to diverging at 180 degrees.

The number of V-tracks formed by terrestrial cosmic-ray events can be separated from space cosmic-ray events by comparing track density of events in the interior of the tektite with track density in the flange, which was formed during the passage of the tektite through the atmosphere. It has been proven that the heating of the body of the tektite during atmospheric passage was insufficient to erase tracks, except in the frontal 3 mm of the ablated tektite surface. Cosmic ray events on the Earth’s surface are relatively sparse due to the protective atmosphere and even less frequent after burial of the specimen. If the tektite had spent any significant time in space, where no protective atmosphere exists, it would be expected to exhibit a greater density of V-tracks on the interior.

In order to count the tracks an experimental process similar to that of 238U fission track dating is applied. A tektite surface is polished and then etched in hydrofluoric (HF) acid. The process differs in that the etch time is longer (around 85 seconds in 48% HF) so that the etching penetrates the full length of the V-track. In order to calibrate the V-tracks one slice of the tektite is proton irradiated. This gives the production rate observable, which is dependent on the concentration of heavy elements in the tektite. Around one third of the proton-induced fissions yield etch pits that are noticeably v-shaped (Fleischer et al., 1965).

What can be concluded?

Fleischer et al. (1965) noted that their failure to find such events in a group of tektites leads to a probable cosmic ray exposure age of less than 300 years for the tektites in their present form. It is apparent that tektites spent little or no time in space, limiting the tektite source to the Earth-Moon system.