A new type of highly-vaporized microtektite from the Transantarctic Mountains

We report on the discovery of microtektites (microscopic impact glass spherules) in a glacial moraine near Larkman Nunatak in the Transantarctic Mountains, Antarctica. The microtektites were identiﬁed based on their physical and chemical properties. Major and trace element compositions of the particles suggest that they may be related to the Australasian strewn ﬁeld. This would further extend the current strewn ﬁeld (cid:1) 800 km southward. Depletion in volatiles and enrichment in refractory elements in Larkman Nunatak microtektites ﬁt the volatilization trend deﬁned by Australasian microtektites, suggesting that they may represent a new highly vapor fractionated end-member thereof. This observation is supported by their low vesicularity and absence of mineral inclusions. This discovery has signiﬁcant implications for the formation of microtektites (i.e. their evolution with respect to the distance from the source crater). Finally, the discovery of potentially old (i.e. 0.8 Ma) microtektites in moraine has implications for the stability of the East Antarctic Ice Sheet in the Larkman Nunatak area over the last (cid:1) 1 Ma and, as a consequence, the high eﬃciency of such moraines as traps for other extraterrestrial materials (e.g. micrometeorites and meteoritic ablation debris). 2018


INTRODUCTION
Microtektites are the microscopic counterpart of tektites, which are glass objects resulting from the melting and vaporization of the Earth's crust during hypervelocity impacts of extraterrestrial bodies (Glass, 1990;Koeberl, 1994;Artemieva, 2008;Glass and Simonson, 2013). They are usually scattered over regions distal to impact craters called strewn fields (e.g., Glass and Simonson, 2013). To date, four major strewn fields have been discovered on the Earth's surface (i.e. Australasian, Central European, Ivory Coast and North America; Glass and Simonson, 2013). The Australasian strewn field is characterized by its large geographical extent that is at least an order of magnitude greater than that of other strewn fields (i.e. 14,000 km; Fig. 1; Folco et al., 2008;Glass and Simonson, 2013) and its relatively young age (0.8 Ma;Izett and Obradovich, 1992). Despite its recent formation, the source crater of this strewn field has yet to be found. Several studies based on the distribution and/or geochemical properties of tektites and microtektites suggest that the source crater may be located in South East Asia, and probably in Vietnam (e.g., Glass and Pizzuto, 1994;Lee and Wei, 2000; https://doi.org/10.1016/j.gca.2018.02.041 0016-7037/Ó 2018 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Ma et al., 2004;Glass and Koeberl, 2006;Prasad et al., 2007;Folco et al., 2010a.
Australasian microtektites have been found in deep sea sediments of the Indian and Pacific Ocean (hereafter AUS/DSS; e.g., Prasad and Sudhakar, 1999;Glass et al., 2004) and more recently on top of nunataks of the Transantarctic Mountains, Victoria Land, Antarctica (hereafter AUS/TAM; Folco et al., 2008;Folco et al., 2009). The current southernmost limit of the strewn field has been established after the discovery of Australasian microtektites in glacial sediment collected at a low relief crest next to Allan Hills, Victoria Land, Antarctica, which is situated approximately 11,000 km away from the hypothetical source crater location .
Here we describe the discovery of microtektites in glacial moraine collected next to the Larkman Nunatak in the Transantarctic Mountains, Antarctica. We first describe their geochemical affinities with Australasian microtektites (both AUS/DSS and AUS/TAM), suggesting that these two materials are paired. Subsequently, we will show that the microtektites from Larkman Nunatak expand the volatilization trends observed within AUS/DSS and AUS/TAM and, as a result, represent a new highly volatile depleted end-members for these trends.

Samples
The samples were recovered in 2006 by one of us (MG) from a glacial moraine near Larkman Nunatak (hereafter LKN; 85°46 0 S, 179°23 0 E), along with hundreds of micrometeorites (Fig. 1b, c and d;Van Ginneken et al., 2016). At the time of recovery, the moraine was covered by a 4 cm thick snow cover. The moraine is oriented East-West and extends ca. 1.5 km with a width of 700 m. It rises up to 30 m above the surrounding meteorite-rich blue ice and it is separated from the nunatak by a depression of up to 500 m wide. Samples were collected from the southern edge of a boulder ridge approximately 40 m into the moraine and located approximately half way through the moraine along an East-West traverse. Detailed information on the bedrock and lithology of the moraine is provided in Van Ginneken et al. (2016).
In the laboratory 250 g of moraine samples were first washed in water and hydrogen peroxide to remove evaporite incrustations that prevented the identification of microtektites. They were subsequently dried and size separated using 106, 250, 425, 850 and 2000 lm sieves. Glacial sediment >2000 lm in size was not included. Fifty-two microtektites-like particles >106 lm in size were subsequently hand-picked from the sieved material under a stereomicroscope. Samples were identified on the basis of their spherical shape, pale yellow color and transparency.

Petrography and major element analyses
The microtektites were first mounted on clear adhesive tape and observed using a LEO 1455 environmental Scanning Electron Microscope (SEM) at the Imaging and Analysis Centre (IAC) of the National History Museum (NHM), London, United Kingdom, in order to gather information on their external features. Subsequently, a set of 13 microtektites were embedded in epoxy, sectioned, polished and carbon coated at the NHM. The remaining 39 particles were consumed in unsuccessful Ar-Ar analyses.
The major element composition of the microtektites was determined using a Cameca SX100 electron microprobe at the IAC that is equipped with five wavelength dispersive spectrometers. Bulk compositions were calculated by averaging four point analyses for each particle. A defocused beam 10 lm in size was used in order to reduce the loss of volatile elements. Operating conditions were an accelerating voltage of 20 kV, a 20.0 nA beam current. A number of synthetic and natural standards were used for instrumental calibration. Standards include but were not limited to: forsterite (Mg 2 SiO 4 ) (San Carlos olivine) for calibration of Mg, hematite (Fe 2 O 3 ) for Fe calibration, wollastonite (CaSiO 3 ) for Si and Ca calibration and corundum (Al 2 O 3 ) for Al calibration. The built-in PAP-algorithm (e.g., Pouchou and Pichoir, 1991) was used for correction. The detection limits (in wt.%) are: Si = 0.01; Ti = 0.03; Al = 0.01; Fe = 0.03; Mn = 0.03; Mg = 0.01; Ca = 0.03; Na = 0.03; K = 0.04.

Trace element analyses
Trace element compositions of 11 microtektites were determined by Laser Ablation Inductively Coupled Plasma Mass Spectrometry at the IAC. The instrument was a Agilent 7500 ICP quadrupole mass spectrometer coupled with a ESI NWR193 ArF excimer laser source. The laser was operated at a repetition rate of 10 Hz, the spot size was 45 lm, and the energy at 3.2 mW. Signals for the analytical masses reported in Table 3 were acquired in peak hopping mode with 10 ms dwell time. Analyses consisted of the acquisition of 30 s background signal and one minute ablation signal. Data reduction was performed with the software LamTrace (Jackson, 2008). NIST SRM 612 (Hinton, 1999) and 43 Ca were adopted as external and internal standards, respectively. Precision and accuracy were assessed via repeated analysis of BCR-2 g, resulting better than 7% and ±10%, respectively, at the mg/g concentration level. Mean detection limits at 45 lm spot size for the quadrupole instrument are reported in Table 3.

Overall description
The 52 samples are transparent glass spherules (Fig. 2). The color of all spherules is pale yellow and in all cases their shape has a high degree of sphericity. Only one microtektite exhibits a bubble 10 lm in diameter in its interior (Fig. 2c). The surface of most particles is smooth and featureless ( Fig. 3a), but 8% of the particles show weathering pits ( Fig. 3b) identical to those observed on the surfaces of V-type (i.e. glassy) cosmic spherules extracted from the same glacial moraine (Van Ginneken et al., 2016). The particles also lack microscopic impact craters reported on an Australasian microtektite (Prasad and Sudhakar, 1996). The SEM backscattered images of sectioned samples show constant Z-contrast, suggesting that their chemical compositions are homogeneous ( Fig. 3c and d).
The size of the samples varies between 107 and 388 lm. The size distribution of the particles (Fig. 4) is normal, except for a depletion at around 200 lm that is probably a statistical bias owing to low count statistics. Size fractions of glacial moraine larger than 400 lm were searched for microtektites but none were found, suggesting this is their maximum size limit in the deposit.

Bulk chemistry
The major element bulk compositions of 13 samples is reported in Table 1. Major element concentrations vary from one particle to another, but compositional trends are observed. Most major oxides are inversely correlated to SiO 2 (Fig. 5) that shows concentrations ranging from 43.7 to 64.5 wt%. MgO, Al 2 O 3 , CaO and TiO 2 range from 4.36 to 11.9, 18.9 to 35.1, 3.27 to 9.35 and 1.04 to 1.79 wt %, respectively. Conversely, FeO is positively correlated with SiO 2 and ranges from 1.06 to 3.95 wt%. The Na 2 O and K 2 O contents also slightly increase with SiO 2 and range from 0.07 to 0.39 and 0.10 to 0.81 wt%, respectively.
When compared to known microtektites populations classified according to Glass et al. (2004), only four samples have major element compositions overlapping those of normal AUS/DSS and AUS/TAM microtektites ( Fig. 5; i.e. ''normal microtektites" have major oxide compositions similar to Australasian tektites; Glass et al., 2004). The nine remaining particles show silica contents significantly lower than that of normal AUS/DSS and AUS/TAM microtektites. Conversely, an enrichment in Al 2 O 3 , TiO 2 and CaO is observed. About half of the particles show MgO contents that overlap normal AUS/DSS and AUS/TAM microtektites compositional fields, the rest being enriched. FeO is depleted in only four samples. The low Na 2 O content overlaps that of normal AUS/TAM microtektites in all particles except for particle #LK06-1159, in which it is significantly depleted at 0.07 wt%. Eight samples show K 2 O that is lower than microtektites from other collections. Our samples show major element compositions that plot between the compositional field of normal AUS/DSS and AUS/TAM microtektites and a high-Al AUS/DSS microtektite . Two similar high-Al AUS/TAM microtektites have major element composition overlapping that of our samples . Comparison with differentiated cosmic spherules (i.e. having non-chondritic chemical composition) show that our samples plot in distinct compositional fields, especially when considering FeO and Al 2 O 3 that show systematically higher and lower values, respectively (Fig. 5). Table 2 lists the trace element compositions of 11 samples that was determined by LA-ICP-MS. The major and trace element compositions of our samples,   Nunatak microtektites collected so far. microtektites. The microtektites were collected in the 106-2000 lm size fraction, so the upper limit of this diagram is the actual upper size limit of the microtektites collected. On the other hand, the lower limit is mainly due to the smallest mesh size used for sieving and may not represent the lower size limit of the microtektites.

AUS/DSS and AUS/TAM microtektites and of the
Upper Continental Crust (UCC) were normalized to CI chondrites (Fig. 6). The geochemical patterns of the samples are broadly similar to those of the UCC and AUS/TAM and AUS/TAM microtektites. However, their chemical composition exhibit variations from the UCC common to AUS/DSS and AUS/TAM microtektites. Refractory elements Sc, Cr, Y, Zr, REE, Hf and Th are consistently enriched with respect to the UCC, whereas other refractory and some moderately volatile elements Li, Be, Mg, Al, Si, Fe, Ca, Mn, Nb, Ba and Ta plot close to the UCC, except for Fe in particle #LK06-1159 that is depleted. Conversely, volatile to highly volatile elements Na, K, Zn, Rb, Sr and Cs are significantly depleted compared to the UCC.

A new southernmost extension to the Australasian microtektites strewn field
The glassy spherules collected in the glacial moraine near LKN are identified as microtektites on the basis of several criteria: (1) their pale-yellow color allows their distinction from V-type cosmic spherules found within the same sediment, which usually exhibit darker colors (Genge et al., 2008;Van Ginneken et al., 2016). Furthermore, this pale-yellow color is typical of normal AUS/DSS and AUS/TAM microtektites Folco et al., 2009). However, V-type cosmic spherules exhibit a wide range of color and transparency (Genge et al., 2008), so this criterion should be used in addition with the following criteria; (2) Their major and trace element chemistry is broadly similar to that of the Upper Continental Crust (Taylor and McLennan, 1995), which is also typical of the microtektites from the main known strewn fields ; (3) Their major element chemistry is clearly distinct from that of differentiated cosmic spherules (Taylor et al., 2007;Cordier et al., 2011;Cordier et al., 2012), which represent potential alternatives to microtektites as glassy micro-spherules exhibiting non-chondritic compositions; (4) The total alkali content (Na 2 O + K 2 O) of all samples is lower (0.18-1.20 wt%) than in volcanic glasses for a given silica content, and K 2 O/Na 2 O is always >1. These geochemical features are characteristic of tektitic material (Koeberl, 1990). All these criteria suggest that these glassy spherules recovered at LKN are indeed microtektites.
Pairing microtektites originating from different sampling locations based on their geochemistry alone can be challenging because of important chemical overlaps between populations originating from different strewn fields (Koeberl, 1990;Glass et al., 2004). The major element compositions of normal Australasian, Ivory Coast and North American microtektites, for example, overlap significantly (Fig. 5). However, normal AUS/TAM microtektites are Table 1 Major element bulk compositions (in oxide wt%) of microtektites from Larkman Nunatak. Bulk compositions were determined using EPMA and by averaging four defocused beam (typically 10 lm in diameter) point analyses in each sample. enriched in the refractory element Ca and depleted in alkali elements compared to normal AUS/DSS microtektites due to their more distal deposition relative to the hypothetical source crater and resulting increased volatilization (Folco et al., 2010a). Consequently, AUS/TAM microtektites plot in distinct compositional fields from Ivory Coast and North American microtektites while still overlapping those of AUS/DSS microtektites. Out of 13 LKN microtektites, 9 have major element compositions plotting in-between the composition fields  Folco et al. 2009Folco et al. , 2016 and achondritic V-type cosmic spherules (Taylor et al., 2007;Cordier et al., 2011;Cordier et al., 2012). All values are in wt%.    (Folco et al., 2010a), which is clearly distinct from the trends observed for Ivory Coast and to a lower extent North American microtektites. Similar chemical trends are observed with trace elements in Harker diagrams of volatile against refractory elements (Fig. 8), with LKN microtektites showing affinities with high-Al Australasian microtektites mainly. More importantly, for identical volatile content, refractories La, Hf and Th in LKN microtektites overlap the compositional fields of AUS/DSS and AUS/TAM microtektites only. The chemical similarity of LKN and Australasian microtektites suggest both were generated in the same impact event and have the same source crater. In particular, that LKN spherules lie on the extension of trends in alkali metals, refractory elements and iron observed only in Australasian microtektites is strong evidence for a genetic relationship. The LKN microtektite data testifies to a continued evolution of microtektite compositions, through the same chemical processes that control Australasian microtektites, making it very likely these represent distal ejecta from this impact event.
Until now Allan Hills has been the southernmost extension of the Australasian strewn field, however, the discovery of the LKN microtektites increases it by 800 km, with a new maximum extension from the putative crater of 12,000 km (Fig. 1a). This strengthens the hypothesis by Folco et al. (2016) that the distribution of Australasian microtektites in Antarctica is continental and not only limited to Northern Victoria Land. The maximum diameter of 388 lm of LKN spherules are also compatible with their identity as the southernmost extension to the Australasian strewn field. Although AUS/TAM microtektites are present in the same size range (Folco et al., 2009;Folco et al., 2016), their size distribution peaks at 450 lm, with particles up to 560 lm in diameter in Allan Hills. This decrease in size from AUS/TAM to LKN microtektites further strengthens that the latter were deposited further away from the source location and represent the current maximum extension of the Australasian strewn field.

Expanding the volatilization trend of the Australasian microtektites
A common volatilization trend exists between AUS/ DSS and AUS/TAM (Folco et al., 2010a) and is typified by the depletion in volatiles and enrichment in refractory elements with increasing distance from the hypothetical source crater. If the microtektites from Larkman Nunatak are indeed an extension of the Australasian strewn field, translating to a 7% increase in distance from the hypothetical source crater, then it might be expected that the geochemistry of our microtektites will show a notable increase in volatilization with respect to AUS/DSS and AUS/TAM particles.
The most refractory major elements Ti and Al (Lodders, 2003) are not affected by volatilization during microtektite formation and thus, increased chemical fractionation due to volatilization at high temperature will not modify the TiO 2 /Al 2 O 3 ratio, which is representative of the target material. As a result, on a TiO 2 versus Al 2 O 3 diagram, Australasian microtektite values plot along a linear trendline (e.g. Folco et al., 2010a). Fig. 7a shows that TiO 2 /Al 2 O 3 values of LKN microtektites plot along the same trendline of normal AUS/DSS and AUS/TAM and have a TiO 2 /Al 2 O 3 ratio of 0.052 (±0.003), similar to that of AUS/DSS and AUS/TAM of 0.056 (±0.005) (Folco Fig. 7. Al 2 O 3 vs. TiO 2 (a) and Na 2 O/Al 2 O 3 K 2 O/TiO 2 (b) diagrams showing how Larkman Nunatak microtektites fit into the volatilization trend of AUS and AUS/TAM microtektites defined by refractory and volatile/refractory elements (Folco et al., 2010a). Symbols for LKN, AUS/DSS and AUS/TAM microtektites of Fig. 6 were used. et al., 2016). More importantly, eight of our particles significantly extend the trendline towards more refractory compositions and in the compositional field previously defined by the high-Al AUS/DSS and AUS/TAM microtektites.
Additionally, the Ti and Al contents of LKN microtektites are inversely correlated to the content in alkali Na 2 O and K 2 O (Table 1 and Fig. 7b). Once again, LKN microtektites fall along the trend defined by AUS/DSS and AUS/TAM microtektites, with most of our particles plotting towards more volatile-poor contents. In particular, three particles exhibit the lowest values on the trendline, indicating that they are severely depleted in volatiles and enriched in refractory elements. This is consistent with the constant depletion in K 2 O and Na 2 O content with distance from the hypothetical source crater (Fig. 9), which is particularly pronounced in microtektites recovered in Antarctica. Thus, major elements support the contention that LKN microtektites are highly vapor fractionated and extend the common volatilization trend observed in Australasian microtektites.
Regarding trace elements, the LKN microtektites are depleted in volatile (Rb, Cs and Zn) and significantly enriched in refractory trace elements (La, Hf, Th) with respect to normal AUS/DSS and AUS/TAM microtektites (Fig. 8). The only exception being the refractory element U that is strongly enriched in most normal AUS/DSS microtektites. However, the range of U content in LKN particles is similar to that of normal AUS/TAM particles and strongly depleted compared to normal AUS/DSS and UCC (Figs. 6 and 8). Wasson et al. (1990) explained this depletion in U by volatilization, with possible enhancement by high pO 2 and preferential location of U in the reduced target's carbonaceous matter. Consistently with the major element discussed above, LKN microtektites plot in the Fig. 8. Harker diagrams of volatile against refractory trace elements in Larkman Nunatak micrometeorites compared to data from the literature Folco et al. 2009Folco et al. , 2016. All values are in lg/g. Symbols for LKN, AUS/DSS and AUS/TAM microtektites of Fig. 6 were used. same compositional fields of high-Al AUS/DSS and AUS/ TAM particles, except for Zn that is significantly higher in the latter and extremely depleted in LKN particles. Glass et al. (2004) argued that the high Zn content of high-Al AUS/DSS microtektite suggests that they may not represent a severely vapor fractionated end-member of the AUS microtektites, but rather a different population of microtektites, similarly to high-Mg ''bottle green" microtektites (Glass, 1972;Folco et al., 2009). However, the severely depleted Zn content in LKN particles suggests that they may not be related to high-Al microtektites recovered closer to the hypothetical source crater and represent a new type of highly volatile depleted microtektites.
The occurrence of only one vesicle amongst recovered LKN microtektites (Fig. 2) indicates that vesicularity in this population is extremely low. This further supports a link with Australasian microtektites, as vesicularity has been observed to decrease significantly in AUS/DSS and AUS/TAM with the distance from the source crater (Folco et al., 2010b possibly owing to ''bubble-stripping"  Cassidy et al. (1969), Glass et al. (2004), Glass and Koeberl (2006) and Folco et al. (2010a)) with distance from the hypothetical source crater in the Indochina region. mechanism that was proposed to explain the loss of volatile in tektites and microtektites (Melosh and Artemieva, 2004).
No partially dissolved silica-rich lechatelierite-like inclusions, that are commonly observed in AUS/DSS and AUS/TAM microtektites (Glass, 1990;Folco et al., 2009;Folco et al., 2010b) were observed in the LKN spherules. Although their presence cannot be entirely excluded, they are at most extremely rare and may suggest that in LKN particles, possible target material was completely melted and digested during microtektite formation, consistently with more intense and/or longer heating compared to AUS/DSS and more particularly AUS/TAM microtektites.
Constant depletion of volatile and enrichment in refractory elements coupled with an extremely low vesicularity and likely absence of mineral inclusions suggest that LKN microtektites may represent the most vapor fractionated and intensely heated end-members of the Australasian strewn field discovered so far.

Accumulation mechanism of microtektites in the moraine
Knowing the accumulation mechanism of microtektites is critical to understanding whether they were deposited directly within the moraine at the time of formation or rather transported from another locality via the ice and/ or wind. A recent study of the accumulation mechanisms of micrometeorites found within LKN moraine alongside the microtektites suggests that we cannot exclude the possibility that LKN microtektites were first deposited at a different locality before being windblown and trapped within the moraine (Suttle et al., 2015). The relatively high concentration of microtektites within the moraine (200 particles/ kg) rules against this hypothesis, as we would except such a scenario to quickly dilute the particles over large areas. Another possible accumulation mechanism is the recent release of microtektites in the moraine by sublimation of ice present directly underneath it. The age of the surface ice in the vicinity of LKN is unknown however, as well as the terrestrial age of meteorites recovered from the moraine. However, this would require ice at least as old as the microtektites, that is 0.8 Ma old considering that they are related to the Australasian strewn field. Such old ice surfaces have not been identified in the Transantarctic Mountains, which usually exhibit ice younger than 200-150 ka based on modeling of ice flow and exposure age of meteorites (Grinsted et al., 2003). Two scenarios are therefore envisioned: (1) direct infall of microtektites in the moraines at the time of their formation and (2) sublimation of earlier generations of ice that contained microtektites, before their preservation in the moraine. Similarly to Australasian microtektites recovered from Allan Hills , the lack of evidence of abrasion on any of the LKN microtektite suggests that a likely accumulation scenario is direct infall at the time of formation. The observation of abundant weathering pits on some LKN microtektites (Fig. 3b) suggests that they have been exposed to liquid water over long periods of time, similarly to V-type cosmic spherules (Van Ginneken et al., 2016), suggesting a long-lasting presence within the moraine, which is in agreement with a direct infall.

Implications
This new type of highly volatile-depleted microtektites puts new constraints on the formation and deposition of the Australasian strewn field. As mentioned earlier, LKN microtektites represent the southernmost extension of the Australasian strewn field. This strewn field is characterized by its tri-lobed shape (or butterfly pattern; Glass and Simonson, 2013). Larkman Nunatak microtektites thus currently represent the maximum extension of the main ejecta ray oriented toward SSE. This is in contrast with the two lateral lobes that extend WSW and ESE and have only been observed in deep sea cores in the Indian and Pacific Oceans (Glass and Simonson, 2013). Such an extension from the source crater further confirms that the impactor most likely came from a very oblique (i.e. <45°) NNW trajectory (e.g. Artemieva, 2008;Artemieva, 2013). Furthermore, cratering models suggest that ejecta thrown furthest from the source crater along ballistic trajectories were extracted nearest from the target area (Melosh, 1989). Assuming that distal ejecta are normally graded in terms of particle-sizes (e.g. Glass and Simonson, 2013), this would suggest that LKN microtektites, which are notably smaller than other Australasian microtektites (including AUS/TAM microtektites), currently represent the ejecta that was ejected closest to the target area (i.e. contact surface). Thus, they suffered the highest temperature regimes during their formation. This is consistent with their smaller size compared to other Australasian microtektites, high-temperature melts having smaller surface tensions that results in smaller particles (e.g. Artemieva et al., 2002). Higher temperature regimes are also consistent with the severe volatile depletion, the almost complete lack of vesicularity and mineral inclusions of the LKN microtektites. This, in turn, confirms the suggestion by Folco et al. (2010aFolco et al. ( , 2016) that volatile-content in microtektites decreases with increasing distance from the source crater (Fig. 9). In conclusion, the presence of Australasian microtektites as far as 12,000 km from the hypothetical source crater give new constraints for future cratering models aiming at finding the actual crater location and/or studying the deposition of the Australasian strewn field (e.g. Artemieva et al., 2002).
Another important implication concerns the age of the newly discovered collection of micrometeorites that were found along with microtektites in LKN glacial moraine (Suttle et al., 2015;Van Ginneken et al., 2016). If we consider that LKN microtektites are part of the Australasian strewn field, that would imply that they were deposited in the moraine 0.8 Ma ago. This would make the LKN micrometeorite collection as old as the TAM collection, which is consistent with the similar ranges of weathering states observed in both collection (Van Ginneken et al., 2016). Furthermore, the discovery of Australasian microtektites in sediments recovered from a low relief crest sampled at Allan Hills suggest that the East Antarctic Ice Sheet has been extremely stable in this area over the last 1 Ma . Note that this is consistent with the discovery of numerous meteorites in blue ice fields close to LKN (e.g., Corrigan et al., 2014). More importantly, this would imply that glacial moraines located close to nunataks in the Transantarctic Mountains are efficient sampling sites for micrometeorites and other extraterrestrial materials (e.g., meteoritic ablation debris). For example, similar areas could be surveyed to confirm the continental distribution of meteoritic ablation debris related to a large airburst event 480 ka ago that were discovered in the TAM, Dome Fuji and Dome Concordia (Van Ginneken et al., 2010).

CONCLUSIONS
We report the discovery of microtektites in a glacial moraine near Larkman Nunatak. Geochemical evidence, both in term of major and trace elements, suggest that they may be related to the Australasian strewn field. This would further extend the Australasian strewn field by 800 southward. Continuous depletion in volatiles and enrichment in refractory elements in Larkman Nunatak microtektites prolong the volatilization trend defined by Australasian microtektites, suggesting the former represent a new type of highly vapor fractionated microtektites. The fact that they are a new end-member to the volatilization trend is further supported by their very low vesicularity (i.e. almost complete loss of volatiles through boiling of the silicate melt) and absence of mineral inclusions (i.e. complete melting of target material due to extremely high and/or long temperature regimes). This discovery has strong implication for tektites/microtektites formation. Another important implication of the discovery of Australasian microtektites in the Larkman Nunatak area is that the East Antarctic Ice Sheet in this area may have been stable over the last 1 Ma, confirming a similar observation the Allan Hills area, in the Transantarctic Mountains. Finally, old age and stability of glacial moraine at Larkman Nunatak suggest that such glacial moraines related to nunataks in the Transantarctic Mountains may be efficient sampling sites for infalling microscopic extraterrestrial matter, such as micrometeorites and meteoritic ablation debris.