Contrasting Aerodynamic Morphology and Geochemistry of Impact Spherules from Lonar Crater, India: Some Insights into Their Cooling History

Abstract

The ~50 or 570 ka old Lonar crater, India, was excavated in the Deccan Trap flood basalt of Cretaceous age by the impact of a chondritic asteroid. The impact-spherules known from within the ejecta around this crater are of three types namely aerodynamically shaped sub-mm and mm size spherules, and a sub-mm sized variety of spherule, described as mantled lapilli, having a core consisting of ash-sized grains, shocked basalt and solidified melts surrounded by a rim of ash-sized materials. Although, information is now available on the bulk composition of the sub-mm sized spherules (Misra et al. in Meteorit Planet Sci 7:1001–1018, 2009), almost no idea exists on the latter two varieties. Here, we presented the microprobe data on major oxides and a few trace elements (e.g. Cr, Ni, Cu, Zn) of mm-sized impact spherules in unravelling their petrogenetic evolution. The mm-sized spherules are characterised by homogeneous glassy interior with vesicular margin in contrast to an overall smooth and glassy-texture of the sub-mm sized spherules. Undigested micro-xenocrysts of mainly plagioclase, magnetite and rare clinopyroxene of the target basalt are present only at the marginal parts of the mm-sized spherules. The minor relative enrichment of SiO₂ (~3.5 wt% in average) and absence of schlieren structure in these spherules suggest relatively high viscosity of the parent melt droplets of these spherules in comparison to their sub-mm sized counterpart. Chemically homogeneous mm-sized spherule and impact-melt bomb share similar bulk chemical and trace element compositions and show no enrichment in impactor components. The general depletion of Na₂O within all the Lonar impactites was resulted due to impact-induced volatilisation effect, and it indicates the solidification temperature of the Lonar impactites close to 1,100 °C. The systematic geochemical variation within the mm-sized spherules (Mg# ~0.38–0.43) could be attributed to various level of mixing between plagioclase-dominated impact melts and ultrafine pyroxene and/or titanomagnetite produced from the target basalt due to impact. Predominance of schlieren and impactor components (mainly Cr, Ni), and nearly absence of vesicles in the sub-mm sized spherules plausibly suggest that these quenched liquid droplets could have produced from the impactor-rich, hotter (~1,100 °C or more) central part of the plume, whereas the morpho-chemistry of the mm-sized spherules induces their formation from the relatively cool outer part of the same impact plume.

1 Introduction

The Lonar asteroid impact crater of Maharashtra state, India (Fig. 1), is one of the few known terrestrial impact craters excavated in the basaltic targets of Deccan Traps (Nayak 1972; Fredriksson et al. 1973; Kieffer et al. 1976; Strobe et al. 1978; Fudali et al. 1980; Rao and Bhalla 1984; Ghosh and Bhaduri 2003; Hagerty and Newsom 2003; Kumar 2005; Osae et al. 2005; Son and Koeberl 2007; Louzada et al. 2008; Misra et al. 2009, 2010; Maloof et al. 2010; Weiss et al. 2010; Jourdan et al. 2011; Arif et al. 2012; Bose et al. 2013; Senthil Kumar et al. 2014; Komatsu et al. 2014). Other recently discovered terrestrial impact craters in basaltic target include the Vista Alegre and Vargeão dome in the Paraná flood basalt (~133–132 Ma), Brazil (Crósta et al. 2010, 2012). The Logancha crater in eastern Siberia, Russia, is also known as a complex crater in basaltic target (Siberaian Trap basalt?) (Reichow et al. 2002), however, little information is available on this crater at present (Feldman et al. 1983; Mironov et al. 1987). In fact, the well-known and fully accessible Lonar crater is an important terrestrial analogue for the large number of craters in basaltic crusts on the surfaces of Moon, Mars, Venus and other planetary bodies (e.g. asteroids) in our solar system and therefore more studies on this crater may have wide application and planetary significance.

Fig. 1

Sketch maps of a India showing the position of Lonar crater, and b Lonar crater with topography (contours are shown in meters). The spherules and impact-melt bomb for the present study obtained at location L-60 close to southeastern crater rim

The Lonar crater was formed by the hypervelocity impact of a chondritic asteroid that struck the preimpact target basalt from the east at an angle between 30° and 45° to the horizon (Misra et al. 2009, 2010). The total duration of shock event at the Lonar crater was suggested to be approximately 1 s (Kieffer et al. 1976), and the stress generated within the target basalts due to the asteroid impact branched out into a major northwest and a southwest component in the downrange direction immediately after the impact (Arif et al. 2012). The target Deccan basalt of the Lonar crater was erupted close to the Cretaceous-Palaeogene boundary approximately at 65 ± 0.9 Ma (Hofmann et al. 2000; Courtillot et al. 2000) or close to 67.4 Ma (Pandey et al. 2004). More recent study reveals that the first extensive phase of the Deccan volcanism, which might have lasted only a few hundred thousand years, occurred approximately at 67.5 Ma at the northern half of the present Deccan outcrops and after a period of approximately 2.5 Ma of quiescence, the second major phase of volcanism erupted close to 65 Ma (Chenet et al. 2007).

The age of the Lonar crater, although attempted in various techniques, is not yet conclusively known. The ¹⁴C dating of the histosol samples occurring below the ejecta blanket indicates the minimum age of this crater as 11.65 ka (Maloof et al. 2010). Storzer and Koeberl (2004) also reported an apparent fission track date of 15.3 ± 13.3 ka for three impact glass samples from the Lonar crater. The very high error of this age is the result of the lower number of tracks counted. The sediments deposited in the saline lake within the Lonar crater were also dated by ¹⁴C dating method and yielded an age between 15 and 30 ka (unpublished data, see Fredriksson et al. 1979; Sengupta et al. 1997). As the lake sediments were contaminated by modern carbon, the radiocarbon dates obtained were interpreted as minimum ages of impact for the Lonar crater. Nakamura et al. (2014) have recently reported the ¹⁰Be and ²⁶Al exposure ages of four quartz pebbles sampled from the summits of the rim and the topographic high on the ejecta blenket of the Lonar crater. One of this four pebbles yielded the ¹⁰Be and ²⁶Al ages close to 37.5 ka, which is interpreted as the minimum age of formation of the Lonar crater. The other three ages produced far young ages (9–5 ka), which are interpreted due to erosional removal of materials at their sampling sites. These authors also have dated the pre-impact soils from the west of the Lonar crater by ¹⁴C dating technique, which yielded a maximum age of the Lonar crater of 40.7 ± 0.7 ka. The thermoluminescence dating of three in situ impact-melt bomb samples collected from the northern, western and eastern parts of the Lonar crater rim from within the ejecta blanket yielded a mean age of 52 ± 6 ka (Sengupta et al. 1997). However, the most precise ⁴⁰Ar/³⁹Ar age of the four impact-melt rocks is much older and is 570 ± 47 ka (Jourdan et al. 2011).

The impact- and impact-looking glasses occurring in and around the Lonar crater were classified into two fundamental petrographic classes (Osae et al. 2005). Out of these two classes, only the class I glasses include mm-sized impact-spherules (type ‘a’ subclass) and cm-sized (~≤1 to 30 cm in diameter) impact-melt bombs (type ‘c’ subclass), which were the products of shock-induced melting due to asteroid impact (Fredriksson et al. 1973; Kieffer et al. 1976; Osae et al. 2005). The class I glasses show very restricted occurrences outside the crater rim mostly to the east within the ejecta blanket at present. These are, in general, brown in colour, translucent impact-glasses, showing schlieren and flow structures under the microscope. The unmelted mineral fragments present within these glasses include plagioclase, clinopyroxene and crystals of magnetite of the target basalt. The origin of class II impact-looking glasses, which occur both inside and outside of the crater (Osae et al. 2005; Chakrabarti and Basu 2006), raises uncertainty and these could be of anthropogenic in nature (Misra 2006). These glasses have variable sizes up to 20 cm in diameter; and under microscope these glasses are almost opaque in nature and contain unshocked fragments of plagioclase and clinopyroxene (Osae et al. 2005). Palaeointensity dating suggests these impact-looking glasses are only about 1,150 ± 50 years old (Deenadayalan et al. 2009). More recent investigations establish two more varieties of impact spherules from the Lonar crater. The sub-mm sized impact spherules that occur along with class I glasses show schlieren structure defined by alternate bands of silicate glass and chains of tiny dendritic and octahedral shaped magnetite (Misra et al. 2009). These sub-mm sized spherules contain significant proportions of chondritic impactor component (~12–20 %) that was identified by high enrichment of siderophile elements (e.g., Cr and Ni) in these spherules. The presence of schlieren structure both in the aerodynamic shaped mm- and sub-mm sized impact spherules (Nayak 1972; Fredriksson et al. 1973; Misra et al. 2009) suggests these particles were quenched from silicate liquid droplets during their flight within the impact plume and/or atmosphere immediately following the impact. More recently, a new variety of impact spherules is reported from the Lonar crater; these impact spherules, described as mantled lapilli, are also sub-mm in size with cores consisting of conglomerates of ash-sized mineral grains, shocked basalt or solidified melts and adhering rims with ash-sized mineral grains (Beal et al. 2011). However, detail geochemical investigation on these particles is yet to be revealed. The Lonar impact-melt bomb and sub-mm sized spherule, in general, show target basalt dominated bulk chemistry, except the characteristic depletion in Na₂O (in average ~0.8 times and 0.6 times respectively) (Osae et al. 2005; Son and Koeberl 2007; Misra et al. 2009).

Although, various aspects of the Lonar crater e.g., crater structure and strain analyses (Kumar 2005; Misra et al. 2010), geochemistry (Osae et al. 2005; Son and Koeberl 2007; Misra et al. 2009), rock magnetic properties (Louzada et al. 2008; Maloof et al. 2010; Weiss et al. 2010; Arif et al. 2012), and radiometric ages (Sengupta et al. 1997; Jourdan et al. 2011; Nakamura et al. 2014), impact fragmentation (Senthil Kumar et al. 2014) and drainage system and hydrology (Komatsu et al. 2014) have been intensively studied in the last decade, some important aspects, particularly some more detail geochemistry and isotopes, relevant to this crater evolution are still unknown. These are importantly the bulk geochemistry and internal geochemical variations of the mm-sized spherule, possible geochemical fractionation of impactor asteroid components between the Lonar impact-melt bomb and varieties of impact spherules (cf. Mittlefehldt et al. 1992, 1993; Mittlefehldt and Hörz 1998), abundance of platinum group of elements and Cr isotopic composition particularly of the sub-mm sized spherules etc. The main focus of this work is to examine the bulk geochemical composition (along with some trace elements) of the mm-sized impact spherule, along with a detail study on the backscattered electron (BSE) images of these spherule in order to evaluate (1) the mechanism of formation and cooling history of the Lonar impactites, and (2) the possible fractionation of impactor component(s) among the impact-melt bomb, and mm- and sub-mm sized impact spherules.

2 Geological Setting of Lonar Crater

The Lonar crater is a bowl-shaped, near-circular simple impact crater, with an average diameter of approximately 1,810 m with a circularity of 0.95, and a depth of approximately 150 m (Fredriksson et al. 1973; Fudali et al. 1980; Koeberl et al. 2004; Kumar 2005; Misra et al. 2010). Except for a small sector in the NE, there is a continuous well developed raised rim encircling the crater, which raises to an average elevation of approximately 30 m above the adjacent plains, and the crater floor lies approximately 90 m below the pre-impact surface. The rim is flanked in all directions by a continuous ejecta blanket that extends outward with a gentle slope of 2°–6° to an average distance of approximately 700 m from the crater rim, except to the west where it extends for little more than a kilometer (Fudali et al. 1980; Misra et al. 2010). The crater floor is presently occupied by a shallow hypersaline and alkaline lake (depth <6 m, pH ~9.5–10.5, cf. Komatsu et al. 2014 and references therein); below the lake water a sequence of approximately 100 m thick unconsolidated sediment is reported to overlie the crater base that is made up of highly weathered Deccan Trap basalt (Nandy and Dey 1961; Fudali et al. 1980). Approximately 700 m north of the Lonar crater rim, a shallow depression-like structure known as the Little Lonar (Fig. 1) is present, which has a diameter of approximately 300 m. However, drilling into the structure (Fredriksson, personal communication in 1999) and recent studies on ejecta stratigraphy (Maloof et al. 2010) reveal no evidence favoring an impact origin of this structure. The crater preserves a relatively pristine morphology, although the crater interior shows signs of degradation in the forms of gullies and debris flows (Komatsu et al. 2014).

The target rocks of the Lonar crater are sub horizontal Deccan Trap basalts that overlie the older Precambrian basement and have a thickness of >350 m (Kumar 2005). These basalts contain intertrappean sediments (mainly chert and limestone) of small areal extent of fluviatile and lacustrine origin and of variable thickness up to 3 m (Jhingran and Rao 1958; Venkatesh 1967; Krishnan 1968). Altogether six basalt flows of approximately 8–40 m thickness are identified in and around the Lonar crater, out of which only the four bottom flows of the succession are exposed on the crater wall (Ghosh and Bhaduri 2003). The two topmost flows occur away from the crater, and devoid of any impact-induced deformation. Flows are separated from one another by discontinuous marker horizons like red and green paleosols, chilled and vesicular margins, and vugs filled with secondary minerals. Fresh basalt flows occur only at the top approximately 50 m along the crater wall, whereas below this level, the flows are heavily weathered and friable (Fudali et al. 1980). More recent studies suggest the alteration of these lower basalt flows could be impact-related (Misra et al. 2013). The pre-impact black, sticky, humus-rich soils of approximately 5–90 cm thickness are still preserved at places between flows and overlying ejecta (Fudali et al. 1980; Ghosh and Bhaduri 2003). All the basalt flows generally share a common mineralogy and texture except some minor petrographic differences in the abundance of plagioclase phenocrysts, glass, and opaque minerals (Ghosh and Bhaduri 2003; Osae et al. 2005). The porphyritic basalts contain occasional phenocrysts of plagioclase and rare olivine set in a groundmass of plagioclase, augite, pigeonite, titanomagnetite, palagonite, and secondary minerals, such as calcite, zeolite, chlorite, serpentine, and chlorophaeite (Ghosh and Bhaduri 2003).

3 Samples and Experimental Technique

The samples of impact-melt bomb and spherules were collected from a ~48 cm deep pit dug about 250 m southeast of the Lonar crater rim (19°58.356′N, 76°31.072′E) (Misra et al. 2009). Fine grained (≥250 µm) samples were sieved using mesh sizes of 63 and 125 μm, and the spherules (individual mass ranging from <0.05 to ~0.75 mg) were hand-picked using a binocular lens. Only mm-sized (size varies from 2 to 5 mm) spherules and cm-sized impact-melt bomb were considered for the present study. The samples were cleaned and mounted in epoxy, and finally polished for petrographic and geochemical investigation using an electron microprobe.

A Cameca SX 100 electron microprobe was used for analyses of major, minor and some trace elements (Cr, Co, Ni, Cu and Zn) those could be present in our sample population (cf. Misra et al. 2009) (Table 1, and Tables ST1 and ST2 in online supplement). The instrument was equipped with wavelength dispersive spectrometers (WDS) with large crystals (LPET and LLiF). Quantitative analyses were done using following parameters: 15 keV accelerating voltage, 20 nA sample current, 0–1 µm beam with PAP correction routines. For Si and Na, the counting times were set to 7 s; for Ca, Al, Mg, Fe, Ti, Mn, K and P, the counting times were set to 10 s for each element. The calibration standards used for our analyses were plagioclase and diopside (for Si), jadeite (Na), wollastonite (Ca), kyanite (Al), olivine (Mg), magnetite and/or almandine (Fe), TiO₂ (Ti), orthoclase (K) and apatite (P). Pure metal standards were used for analyses of Cr, Co, Ni, Cu and Zn. For major oxides 2-sigma standard deviation is better than 0.4 wt%. The additional analytical conditions that were used during our microprobe analyses of trace elements (Cr, Co, Ni, Cu and Zn) were as follows: accelerating voltage 15 keV, sample current 80 nA and broad beam size of 15 µm. LLiF crystals were used for Cr, Co, Ni, Cu and Zn to improve the count rates as compared to conventional LiF, e.g. for Ni and Cr the count rates were 90 and 74 cps/nA using conventional LiF crystal, while the improved count rates using the LLiF are 580 and 304 cps/nA respectively. Long peak counting time up to 300 s was used to achieve the low detection limits for Cu and Zn, and up to 350 s for Cr, Co and Ni. With this set up, the detection limits for Cr, Co, Ni, Cu and Zn were achieved up to 42, 52, 55, 78 and 99 ppm respectively. To check the reproducibility, NIST 610 glass standard (Rocholl et al. 1997) was run before, during and after the analyses, which gave values for Cr of ~380, 430 and 420 ppm, all within 8 % of the preferred value of 415 ppm. The detection limits for other minor elements were P: 94 ppm, Mn: 166 ppm, Na: 179 ppm, K: 134 ppm, Ti: 93 ppm with uncertainties of 3, 5, 5, 13 and 6 % respectively. Uncertainties of analyses for trace elements Ni, Cr, Co, Cu and Zn were better than 7, 2, 6, 8 and 10 % respectively. The impact-melt and spherule samples were also examined for their internal morphology using BSE imaging and elemental X-ray imaging techniques.

Table 1 A comparative average composition [with standard deviation (SD)] of Lonar target basalts, impact-melt bomb, mm- and sub-mm spherules

4 Internal Morphology of Impactites

4.1 Impact Spherule

In hand specimen study, the mm-sized impact spherules were black in colour, with vitreous luster and often characterised by highly vesicular surface with various geometric shapes, viz. spherical, rod, ellipsoidal and tear-drop (Nayak 1972; Fredriksson et al. 1973; Osae et al. 2005; Ray et al. 2013a). Under microscope, these spherules were translucent, uniformly brown, although in cases showed schlieren between colourless and brown colour layers along with partially melted mineral inclusions; flow-banding was indicated by contorted schlieren and by trails of minute cross-shaped crystallite, apparently magnetite (Fredriksson et al. 1973). The mm-sized spherules of the present population are characteristically black in colour in hand specimen, having spherical to often ellipsoidal shapes, and vary in sizes from ~2 to 5 mm (Fig. 2a, b).

Fig. 2

Backscattered electron (BSE) images of mm-sized impact-spherule from Lonar crater showing internal morphologies: a homogeneous glassy spherules with vesicles (V) at rim; b highly vesiculated spherule; c enlarged view of the square portion of figure ‘b’ showing partially digested plagioclase (Plag) xenocryst within glassy matrix, d highly vesiculated partially melted plagioclase xenocryst within glassy matrix; e partially digested plagioclase xenocrysts and newly formed Ti-magnetite (Ti-Mgt) grains; f plagioclase xenocryst rimmed by newly formed magnetite (Mgt) grains, g nature of occurrences of partially digested xenocrystic plagioclase, angular fragment and/or sub-rounded clinopyroxene (Cpx) and magnetite (Mgt) xenocrysts

The BSE image of a representative mm-sized spherule shows it is almost homogeneous and glassy, non-vesicular at the central part but vesicular in majority towards the margins, where the maximum size of the vesicle varies up to ~0.4 mm (Fig. 2b). In cases, the vesicles contain secondary infillings of quartz. Spherules with sparse distribution of vesicles are also not uncommon. The central parts of the spherules are devoid of any xenocrystic mineral phases like plagioclase, clinopyroxene or magnetite, which are components of target basalt. However, these components are generally common towards the marginal part of the spherules along with vesicles (Fig. 2b, c). The plagioclase xenocrysts are mostly subhedral and highly vesicular, and locally transformed into vesiculated feldspar glass (Fig. 2d). Shock-induced anhedral plagioclase glasses or melted plagioclases are also noticed along the marginal part of the mm-sized spherules (Fig. 2c–e). The mm-sized spherules characteristically show growth of minute, dendritic, euhedral, magnetite within the homogeneous matrix. These magnetite grains generally form fine trains that encircle undigested plagioclase xenocrysts present at the margin of the spherule (Fig. 2f). The xenocrystic clinopyroxenes are also present and mostly associated with relict plagioclase and Ti-magnetite at the marginal part of the spherule (Fig. 2g). However, the xenocrystic pyroxenes are relatively small in size and less common as compared to the melted plagioclase glass. Importantly, these mm-sized spherules lack well defined schlieren trails defined by newly formed minute euhedral magnetite grains, however, these trails of magnetite are typically present in the sub-mm sized spherules (Misra et al. 2009).

4.2 Impact-Melt Bomb

The centimeter–decimeter sized roughly ellipsoidal pieces of impact-melt bombs (type ‘c’ impactite, Osae et al. 2005), which are black in colour, vitreous, fragile and vary in size up to 30 cm in hand specimen, occur as isolated clasts embedded within a fine grained ejecta (Fig. 3a). In hand specimen, these impact-melt bombs are highly vesicular, sometimes showing ropy structures on the surface (Fig. 3b) and/or flow structures in cross-sections (Fig. 3c). Under the microscope, impact-melt bomb samples consist of alternate layers of brown coloured melt and relatively thin layers of opaque (magnetite) (Fig. 3d). The melt-rich part is sometimes extremely vesicular. Thin layers rich in magnetite are also present within the melt-rich part and these layers alternate in smaller scale with melt-rich layers, and they all together show schlieren structure (Misra and Newsom 2011). The flow layers swerve around sub-circular vesicles.

Fig. 3

Occurrence and texture of type ‘c’ impact melt bombs: a outcrop of ejecta showing an occurrence of an isolated piece of impact-melt bomb (shown by arrow), which is darker than the surrounding matrix, scale is coin with diameter 2.5 cm, Location L-60 southeast of the crater close to the crater rim (Fig. 1); b Ropy structure on the surface of a chilled impact-melt bomb; diameter of coin (scale) is 1.9 cm; c cross-section of an impact-melt bomb showing flow layers. Ripples on flow layer (white arrow) indicate presence of ropy structures, white materials within vesicles (black arrow) are post impact infillings, scale: coin with diameter 1.9 cm, Location- L-60; d microphotograph of impact-melt bomb showing radial growth of microlites from common centres (white arrow head), small magnetite crystals define discontinuous layers in the melt (white arrow)

The BSE images of the impact-melt bomb show some new features, which are present mostly close to the vesicular surface of the impact-melt bomb (Fig. 4a), as follows: (a) the impact-melt bomb shows presence of micro-xenocrystic components, (b) development of radiating crystals around vesicles, (c) concentration of minute opaque crystals, and (d) devitrification of glass to minute radiating crystals. The inner part of the impact-melt bomb, however, is clean and homogeneous. Our microprobe analyses confirm that the micro-xenocrystic components are plagioclase, pyroxene and titano-magnetite belonging to the target basalt. The xenocrystic plagioclases are mostly subhedral prismatic in shape, there are also partially melted plagioclase xenocrysts found to occur as large anhedral grains within the impact-melt bomb (Fig. 4b), and the sizes of these plagioclase xenocrysts vary up to 100 µm. The xenocrystic pyroxenes are mostly subangular to subrounded in shape, and vary in size up to 25 µm (Fig. 4c). The xenocrystic magnetite mostly occurs as subhedral grains that range in size up to 50 µm (Fig. 4c, d). The radial microlites, which are the product of devitrification (Fig. 4e), were not crosscut by flow layers present in the impact-melt bomb, and hence they grew after the cessation of flow in the melts (cf. Kieffer et al. 1976). Additionally, the impact-melt bomb contains fragments of unshocked to highly shocked basaltic clast containing maskelynite and clinopyroxene (Fig. 4f), however, these clasts are devoid of any reaction rims.

Fig. 4

Backscattered electron (BSE) image of impact-melt bomb showing a presence of vesicles (‘b’ in fig), micro-xenocrystic components (as rectangle ‘1’ in fig), development of radiating crystals around vesicles (as rectangle ‘2’ in fig) and concentration of minute opaque crystals and devitrification of glass to minute radiating crystals (as rectangles ‘3’ and ‘4’ in fig), b xenocrystic plagioclase (Plag) with smooth and vesiculated texture and vesicles (V) within glassy groundmass, c xenocrystic magnetite (Mgt) and clinopyroxene (Cpx) in clast-rich area, d xenocrystic Fe–Ti oxides and schlieren of Fe–Ti oxides, e radiating microlites in devitrified glass and f presence of shocked target basaltic clast

The fibrous overgrowths of crystallites are noticeable along the margins of both the plagioclase and pyroxene xenocrysts present within the impact-melt bomb (Fig. 5a) (cf. Kieffer et al. 1976). Rim-like overgrowths of tiny dendritic magnetite are often found around the margins of the xenocrystic magnetites (Fig. 5b). The tiny dendritic magnetites are often observed to form flow banding (Fig. 5c), similar to schlieren as found within the sub-mm sized spherule (Misra et al. 2009).

Fig. 5

Backscattered electron (BSE) image of impact-melt bomb showing a fibrous growth along the rim of plagioclase (Plag) and clinopyroxene (Cpx), b overgrowth along rim of magnetite (Mgt), c xenocrystic plagioclase (Plag), clinopyroxene (Cpx), magnetite (Mgt) and schlieren within the glassy groundmass

5 Geochemistry of Lonar Impactites

Our previous study on mm-sized Lonar impact spherule by Energy Dispersive X-Ray (EDS) technique (for analytical detail see Shyam Prasad and Khedekar 2003) showed that these spherules were essentially homogenous in bulk chemistry. Our present analyses (n = 41) on the central and marginal parts of seven mm-sized spherules also confirm our previous result (Table 2, also see Table ST3 for detail analyses in online supplement). Additionally, these spherules do not show any important size-dependent chemical variation within our studied population of impact spherules (2–5 mm in size).

Table 2 Average chemical analyses of central and marginal parts of mm-sized impact spherules (7 in number) from Lonar crater

All the Lonar impactites, including the mm- and sub-mm sized impact spherules and impact-melt bomb, are characteristically depleted in Na₂O (Table 1) [we have only considered sample numbers L-22, L-23, L-29, GSI-1, SG-1, SG-2 and SG-3 of impact-melt bomb from Table 2 of Osae et al. 2005 for our present and other comparisons made in this paper. The other analyses in the Table 2 of Osae et al. (op. cit.) perhaps belong to samples, which were anthropologically made by ancient people (cf. Misra 2006)], which varies in average abundances from ~0.6 to 0.8 times to that in the target basalt (Fig. 6 and online supplement SF1). Moreover, the average mm-sized spherule is distinctly enriched in K₂O (~1.6 times) and marginally depleted in MgO, MnO and P₂O₅ (~0.9 times). The mm-sized spherule and impact-melt bomb are similar in their average bulk chemical compositions except the former is marginally depleted in MnO. In contrast, the sub-mm sized spherule is distinctly different in bulk composition by having importantly low K₂O (~0.85 times) and P₂O₅ (~0.25 times), and relatively high FeO_t, MgO (~1.1 times) and MnO (~1.3 times) in comparison to those in the target basalt. The average major oxide composition of the Lonar impact spherule described in Fredriksson et al. (1973) has higher FeO_t and MgO (average 16 and 6.8 wt% respectively, Table 1), and similar in composition to that of the magnetite-rich zones in sub mm-sized spherule described in Misra et al. (2009).

Fig. 6

Major oxides spidergram showing variations in bulk composition of mm- and sub-mm sized spherule and impact-melt bomb in average, data from Table 1. Note the similarity of composition among mm-sized spherule and impact-melt bomb. The relative variations (SD%) in abundance of oxides are shown in figure SF1 in online supplement

The Mg# [mole Mg/(moleMg + moleFe²⁺)] of the Lonar mm-sized spherule and impact-melt bomb shows relatively restricted range of variation (0.38–0.44, average ~0.41) in comparison to those in the target basalt (0.40–0.49, average = ~0.43, Osae et al. 2005) that overlap with the range shown by the sub-mm sized spherule (Table 1). In Mg# versus SiO₂ plot, mm-sized spherule and impact-melt bomb are slightly SiO₂-rich (~50–54 wt%) compared to the Lonar basalt and many of the sub-mm sized spherule (~45 to 52 wt%; Fig. 7a). The composition of the sub-mm sized spherule mostly overlaps with the target basalt in this diagram and shows no trend of variation. However, SiO₂ in majority of mm-sized spherule shows a decreasing linear trend with increasing Mg#. Irrespective of their nature, all Lonar impactites show a distinct depletion in Na₂O (~2–2.5 wt%) in comparison to those in the Lonar target basalt (~2.5–3.5 wt%, Fig. 7b). The nature of depletion is significantly high in sub-mm sized spherule, and comparatively less in the impact-melt bomb and mm-sized spherule. Both the sub mm-sized and mm-sized spherule show sub-parallel gentle linear increasing trends for Na₂O with decreasing Mg#, and the amount of increment of Na₂O for both the cases is ~0.5 wt%. No specific trend of variation is observed for the target basalt. In Mg# versus K₂O plot, the target Lonar basalt shows a two-step linear increasing trend for K₂O with decreasing Mg# (Fig. 7c). First there is a gentle increasing trend of K₂O with decreasing Mg# from ~0.49 to 0.42 (trend TB1 in Fig. 7c), followed by a shape linear increase of K₂O with a very little variation of Mg# between ~0.43 and 0.40 (trend TB 2). The trend of variation of the sub-mm sized spherule exactly overlaps with the gentle trend of variation shown by the target basalt. The mm-sized spherule, which has higher K₂O than those in the sub-mm sized spherule, shows a sub-parallel linear increasing trend of variation with the latter. The majority of the mm- and sub-mm sized spherule show relatively low range of variation of K₂O of 0.2 wt%. The impact-melt bomb has the highest K₂O, and it also shows steep linear increasing trend of K₂O with decreasing Mg# and the range of variation of K₂O in this case is ~0.4 wt%. The mm-sized spherule and impact-melt bomb show a wide variation in P₂O₅ (~0.1–0.4 wt%), which is partly comparable to those in the target basalt but mostly restricted within two separate partly overlapping elliptical fields (Fig. 7d). The mm-sized spherule shows a broad linear increasing trend of P₂O₅ with decreasing Mg#. The P₂O₅ content of impact-melt bomb also shows a sub-parallel linear increasing trend for P₂O₅ with decreasing Mg#. The sub-mm sized spherule shows lower P₂O₅ content than the target basalt, and both the sub-mm sized spherule and target basalt show no important variation of P₂O₅ with decreasing Mg#. The sub-mm sized spherule does not show any variation in MnO with increasing Mg#, however they have in general relatively high MnO content compared to those in the target basalt (Fig. 7e). The target basalt and mm-sized spherule along with impact-melt bomb show overlapping plot without any definite trend of variations, and their distribution in this diagram partly overlaps the field of sub-mm sized spherule close to their lowest Mg#.

Fig. 7

Bivariant plots of target basalt, mm- and sub-mm spherule and impact-melt bomb from Lonar crater in Mg# versus a SiO₂, b Na₂O, c K₂O, d P₂O₅ and e MnO plots, all oxides in weight percent, note two different trends (TB₁ and TB₂) of the target basalt in Mg# versus K₂O plot

In Mg# versus transitional trace element plots, Cr contents of mm-sized spherule and impact-melt bomb fall in an overlapping field along with the majority of target Lonar basalt (Fig. 8a). These two Lonar impactites show no variation in Cr with Mg#. However, Cr content of sub-mm sized spherule is considerably higher and show a broad moderately linear decreasing trend as Mg# decreases. In Mg# versus Zn plot (Fig. 8b), target basalt shows almost a constant concentration of Zn (~125 ppm), while the mm- and sub-mm sized spherule show relatively high Zn contents in majority of cases and the amount of Zn is found to be highest in the sub-mm sized spherule. The mm-sized and sub mm-sized spherule are plotted within two separate partly overlapping ellipses, which show sub-parallel linear increasing trends for Zn with decreasing Mg#. Cu content of the majority of mm-sized spherule and the impact-melt bomb are marginally lower than that in the target basalt, and show different sharp linear decreasing trends with decreasing Mg#, however, almost no variation in Cu is observed for the target-basalt (Fig. 8c).

Fig. 8

Plots of Lonar impactites in Mg# versus a Cr, b Zn and c Cu diagrams, all trace elements in ppm, symbols as in Fig. 7

The incompatible trace element geochemistry of the target basalt, impact-melt bomb and mm-sized impact spherule are available in Osae et al. (2005). The chondrite normalized average incompatible trace element spidergrams of these three litho-types are very similar in appearance, which show strong depletion in Rb and U and moderate in Sr (Fig. 9 and online supplement SF2). The three spidergrams show moderately high negative slope [(Ba/Lu)_N ~3.5]. The impact-melt bomb and mm-sized spherule are found to have overlapping composition, and these impactites are relatively enriched in Rb (~30–50 times), Ba (~30 times), Th (~25 times), La (~20 times) and Ce (~15 times) over the Lonar target basalt.

Fig. 9

Chondrite-normalised incompatible trace element spidergram of target basalt, mm-sized spherule and impact-melt bomb. Chondrite normalised data taken from McDonough and Sun (1995). The relative variations (SD %) in abundance of elements are shown in figure SF2 in online supplement

6 Discussion

6.1 Possible Chemical Alteration of Lonar Impactites

The terrestrial and planetary basaltic rocks could be examined using CaO + Na₂O + K₂O + MgO (CNKM)–Al₂O₃–FeO^T plot (after Nesbitt and Wilson 1992) in order to evaluate the effect of chemical alteration due to surface weathering (e.g. Newsom et al. 2010; Ray et al. 2013b). When the Lonar target basalt, impact-melt bomb and mm-sized spherule are examined in this diagram, they form an overlapping point concentration on the feldspar line close to the center of this triangular plot (Fig. 10). This plot suggests that the effect of surface weathering on the Lonar samples is minimal at least in terms of their bulk composition. Additionally, Osae et al. (2005), with their extensive geochemical data sets on Lonar target basalt and impactite samples from within the ejecta around the crater rim, also discussed that the effect of weathering on the Lonar impactites was minimal. However, the marginal elevations of As, Zn, Sb and Br in the Lonar impactites over the target basalt were interpreted as limited local weathering effect. Highly mobile LIL trace elements in target basalt, impact-melt bomb and mm-sized impact spherule (Osae et al. 2005) show limited range of variations, therefore, indicating limited mobilisation. For example, Cs and Rb vary only within limited ranges of ~0.7 ppm (between 0.05 and 0.74 ppm) and ~27 ppm (between ~2 and 29 ppm) respectively. The restricted range (~1 ppm) of variation of highly mobile U (between 0.25 and 1.50 ppm) in the samples of target basalt, impact-melt bomb and mm-sized impact spherule (Th contents of the samples vary only between 0.8 and 3 ppm) (Osae et al. 2005) along with absence of any noticeable Ce anomaly in the impactites (Son and Koeberl 2007) argue against any extensive aqueous alteration of the Lonar samples. Therefore, the effect of post-impact near-surface aqueous alteration during weathering on the Lonar samples collected from around the crater rim was minimal, and hence their major and trace element geochemistry have genetic connotations.

Fig. 10

Ternary CaO + Na₂O + K₂O + MgO (CNKM)-FeO_t-Al₂O₃ plots (after Nesbitt and Wilson, 1992) for target basalt, mm-sized spherule and impact-melt bomb showing point concentration of data on the feldspar line in the central part of the triangle indicating against any surface alteration of Lonar impactites by surface weathering

6.2 Viscosity Contrast Between Parent Liquid Droplets and Quenching Temperature of Lonar Spherules

The aerodynamically shaped mm- and sub-mm sized Lonar impact spherule are found to be distinct in their internal morphologies (Fig. 2, also Misra et al. 2009). While the former shows presence of vesicles and xenocrysts close to their surfaces and absence of schlieren structure inside the spherule; the sub-mm sized spherule shows, in general, absence of vesicles and micro-xenocrysts, and smooth surface morphology and presence of schlieren within the spherules. These morphological dissimilarities between these two groups of spherules definitely suggest noticeable difference in viscosity between the fast rotating parent liquid droplets from which these spherules solidified in atmosphere/impact plume just after the impact. It is understood that the parent liquid droplets of the sub-mm sized spherule could have relatively low viscosity, which allowed forming the schlieren structure within these spherules, and the low viscosity along with the smaller sizes of these spherules perhaps induced major loss of volatile elements (e.g., Na, K, P and perhaps Cu) from their parent liquid droplets during their evolution. For example, the schlieren have also been observed in the impact spherules from Greenland, some of which are within the size range of the spherules from the Lonar crater, and it has been interpreted that this structure was formed within the very fast rotating parent liquid droplets of these spherules having low viscosity (Jones et al. 2005). However, the exact physio-chemical factors that controlled the viscosity of the parent liquid droplets of these spherules are not clearly understood at this stage, and these could be due to the differences in temperature between the parent liquid droplets of these two types of impact spherules (cf. Keays and Lightfoot, 2004), and the presence/absence of unmelted solid phases (cf. Giordano et al. 2008). However, minor enrichment of SiO₂ and P₂O₅ in mm-sized spherule (~52 and ~0.24 wt% in average respectively) over sub-mm sized spherule (~48 and ~0.07 wt% respectively) (Fig. 7) could have some positive effect on the viscosity of the former (cf. Shaw 1972; Toplis et al. 1994).

Volatiles (mainly H₂O, CO₂) perhaps had minimum role in controlling the viscosity of the Lonar impact spherules because most of the volatiles escaped from their parent liquid droplets (note very low Loss Of Ignition values of impact-melt bomb analyses, which is ~10 times lower in average than that in target basalts, Osae et al. 2005), which could have formed at very high impact-induced temperature of ~1,600 °C (Son and Koeberl 2007). Relatively high K₂O and low Na₂O in the Lonar impactites (Fig. 6), in general, have been attributed to volatilization or post-impact hydrothermal processes (Purra et al. 2004). As the effect of post-impact hydrothermal alteration on the Lonar impactites is minimal or least (Fig. 10), the loss of Na from the Lonar impactites compared to that in target basalt can be attributed to volatilization. The other important volatile element that was lost from the impact-melt bomb and mm-sized impact spherule was perhaps Cu (Fig. 8c). The volatility temperature of Na in basaltic magma at atmospheric or low pressure (cf. Storey 1973) suggest the solidification temperature of the Lonar impact spherules and impact-melt bomb (which are dominantly basaltic in composition, Table 1) could be close to ~1,100 °C. A comparison with experiments at 1 bar (cf. Muan and Osborn 1956) also suggests that the newly formed magnetite crystals within the Lonar impactites (Figs. 2, 3, 4, 5) could form at temperature range of ~1,150–1,140 °C, which in turn indicates the upper temperature range of quenching of Lonar impactites.

6.3 Mechanism of Formation and Chemical Fractionation Within Lonar Impactites

It is well established that the Lonar impact-melt bomb was formed by plagioclase-dominated partial melting of the target basalt (Kieffer et al. 1976; Ghosh and Bhaduri 2003; Misra and Newsom 2011), and as the major oxides and incompatible trace element chemistry of the average mm-sized impact spherules is almost similar to that of impact-melt bomb (Figs. 7, 9; Table 1), we can argue that these impact spherules were also formed by the similar type of process. The presence of highly vesiculated partially digested plagioclase xenocrysts or glass within the mm-sized spherules (Fig. 2c–g) also confirms the plagioclase-dominated melting of the target basalt during their formation, and the shock pressure of melting can be estimated between 80 and 60 GPa when compared with experiments (Kieffer et al. 1976; Ray et al. 2013a).

Although plagioclase dominated melting of the target basalt is an accepted mechanism for the formation of the impact-melt bomb and spherules, no clear idea presently exists on how did the pyroxene of target basalt get involved in the composition of the impact melt? In their detail observation on class 5 impact glasses from the Lonar crater, Kieffer et al. (1976) reported rounded pyroxene grains with innumerable skeletal crystals, and these relict grains only showed reduction of size by fracturing and edge melting. Our additional observation on the BSE images of Lonar impact-melt bomb and spherules also suggests the presence of sub-rounded and fragmented clinopyroxene and angular fragments of titano-magnetite within the Lonar impact-melt glasses. However, we are unable to find out any partially melted pyroxene micro-xenocrysts within the Lonar impact glasses after several searches. Possibility is that during the impact-induced melting of the target basalt, the plagioclase was melted but the pyroxene and titano-magnetite were perhaps mechanically fragmented to ultrafine grains and finally mixed and reacted with the plagioclase dominated melt. However, more detail observation on the micro-xenocrysts and their chemistry of the Lonar impact glasses are in progress and will be published in future.

The exact reason for the observed chemical variation within the mm-sized Lonar impact spherules (Fig. 7) is a matter of debates. However, it is certain that the impact-induced spinning of the parent liquid droplets of these spherules was not responsible for this fractionation because our previous EDS and present microprobe analyses on the central and marginal parts of these spherules confirm against any fractionation (Table 2). Instead, relatively more abundance of xenocrystic components along the marginal part of the mm-sized spherules suggests that the impact-induced spin of the parent-liquid droplets of these spherules could have a role, which resulting the partial separation between the melt and micro-xenocrystic components towards the central and marginal parts of the spherules respectively.

The target basalt dominated bulk composition of the Lonar impact-glasses (Nayak 1972; Fredriksson et al. 1973; Kieffer et al. 1976; Osae et al. 2005; Son and Koeberl 2007; Misra et al. 2009; Beal et al. 2011) (Table 1) also rules out the control of minor compositional variations of the target on the chemical variation observed within the Lonar impact spherules (Fig. 7). The only remaining possibility could be the mixing of the impact-induced plagioclase-dominated melt with the impact-generated ultrafine pyroxene and/or titanomagnetite components of the target basalt. This possibility has been examined in the following MgO–Al₂O₃–(FeO^T + TiO₂) plot (after Kieffer et al. 1976) (Fig. 11). In the MgO–Al₂O₃–(FeO^T + TiO₂) plot of Kieffer et al. (1976), the colourless and brown coloured glasses of the impact-melt bomb are plotted on a mixing line joining Al₂O₃ apex and the point MgO₂₅(FeO^T + TiO₂)₇₅ on the base of this triangle. The impact-melt bomb shows a wide variation in Al₂O₃ from 100 % down to 35 % on this mixing line. Our studied population of the Lonar impactites are also plotted on this mixing line joining plagioclase apex and point B on the base of the triangle between clinopyroxene and titano-magnetite (Fig. 11 and online supplement ST4). The overlapping fields of the Lonar impactites show a very restricted variation of plagioclase (Al₂O₃ in the diagram) only between ~25 and 45 %. It is apparent from this plot that the contribution of the clinopyroxene of the target basalt was perhaps more in the formation of the Lonar impactites, and the proportions of plagioclase: clinopyroxene: titanomagnetite vary between ~43:47:10 and ~25:62:13.

Fig. 11

Plots of Lonar impactites in MgO-Al₂O₃-(FeO^T + TiO₂) plot after Kieffer et al. (1976); plagioclase, pyroxene and titanomagnetite of the target basalt (for data see Table ST 4 in online supplement) are plotted in the diagram for reference. Note that the Lonar impactites form an overlapping field on a mixing line joining Al₂O₃ apex (plagioclase) and the point ‘B’ on the base line of the triangle with MgO₂₅(FeO^T + TiO₂)₇₅, Al₂O₃ content of Lonar impactites vary between ~25 and 45 wt%

6.4 Fractionation of Impactor Components Within Lonar Impactites

The abundances and ratios of siderophile elements (Cr, Co, Ni) can be used for identification of meteoritic components if the meteoritic contribution in impactites is high (exceeding 0.1 wt%) and the composition of the target rocks are well known (Son and Koeberl 2007). The sub-mm sized impact-spherule recently reported from the Lonar crater contains relatively high Cr (average—310 ppm, standard deviation—150 ppm) and Ni [1,300 ppm (560 ppm)], which are thought to be the impactor asteroid components (Misra et al. 2009). On the other hand, the present observation on low Ni and Cr (Table 1; Fig. 8a) in the mm-sized spherule and impact-melt bomb from the Lonar crater negates any fractionation of meteoritic components within these impactites (cf. Osae et al. 2005). Extensive analyses of Son and Koeberl (2007) showed that Cr [89 ppm (22 ppm)] and Ni [94 ppm (52 ppm)] of the Lonar impactites are comparable to those in the target basalts (Table 1). A recent investigation (Newsom et al. 2010) using a hand-held Bruker Tracer X-ray Fluorescence spectrometer over a hundred impactite samples of shocked and melted materials failed to find a sample with a substantial enrichment in Ni comparable to those in the Wabar or Henbury crater samples.

Evidence for fractionation of impactor components within the terrestrial impact melt bomb and spherules from other craters are not new. Observations on impactite samples from around 64 ky old Wabar crater, Saudi Arabia, which has a diameter of around 90 m and formed by impact of a IIIAB iron meteorite into a quartz-dominated target, show that the impact spherules contain >10 % meteoritic component, compared to <5 % for the massive impact melts (Mittlefehldt et al. 1992). Recent work by Boslough and Crawford (2008) and Newsom and Boslough (2008) have suggested that these materials could have formed by an airbrust mechanism. Studies on mm- to cm-sized impactites from around 49 ky old and 1.2 km diameter Meteor (Barringer) crater, Arizona, USA, which was formed by the impact of a IAB iron meteorite on sandstone–siltstone–limestone sequence, also established 15–22 % meteoritic material in the impactites, with the glasses containing a relatively high CaO content and a low meteorite component (Mittlefehldt et al. 1993; Mittlefehldt and Hörz 1998). A similar history of evolution is also observed for the Lonar crater, India, and in addition we observed fractionation of asteroid components within the impact spherules depending in the sizes. While the sub-mm sized Lonar impact spherules contain significant proportion (~12–20 wt%) of impactor asteroid component (Co, Cr, Ni) (Misra et al. 2009), the impact-melt bomb and mm-sized impact spherule from this crater have very similar siderophile element geochemistry to that of the target basalts and do not show any enrichment of impactor components (Table 1; Fig. 8a).

The exact mechanism, which resulted in the fractionation of meteoritic components between the impact-melt bomb and impact spherules from the Lonar crater, is still poorly constrained. Clearly the sub-mm- and mm-sized Lonar impact spherule differ in their morpho-chemistry (Fig. 2; Table 1; Misra et al. 2009). While the sub-mm sized spherule shows relative enrichments in Cr and Zn, the mm-sized spherule has relatively high concentration of Si, Na, K and P. To explain this heterogeneity in impactite composition we propose the following hypothesis.

We propose that the impact ejecta plume generated above the Lonar crater just after the impact gradually became inhomogeneous in terms of temperature and contents of vaporised impactor (chondritic) components with time during the excavation stage of the crater (Fig. 12). The temperature and impactor components were higher at the central part of the fast uprising impact plume, and the concentration of vaporised components of the impactor was perhaps the highest at the plume head (Misra et al. 2009) (we presume that the chondrite impactor of the Lonar crater was completely vaporized after impact because no fragment/partially melted part of chondrite has been reported either under petrological microscope or in the field after several searches). As the uprising plume head was gradually expanding with time, the concentration of impactor components and the temperature dropped at the peripheral part of the plume. The predominance of schlieren and impactor components, and nearly absence of vesicles in the sub-mm sized spherule (Misra et al. 2009) suggest that the parent liquid droplets of these spherules could have formed within the impactor-rich, hotter (higher than the volatilization temperature of 1,100 °C of Na in basaltic magma, Storey 1973) central part of the plume where these liquid droplets incorporated maximum amount of impactor components from the vapour phase and lost relatively more volatile elements like Na, K and P due to high temperature. On the other hand, the morpho-chemistry of the mm-sized spherules suggests possibility of their formation from the relatively cool, impactor component less/absent outer part of the same plume.

Fig. 12

A hypothetical model showing evolution of Lonar impact plume with time: Stage I initial plume at the beginning of the excavation stage expanding in all directions, Stage II the expanding plume at the end of the excavation stage showing a hot, impactor component-rich central part surrounded by a relatively cool peripheral part, PM- arrows within plume showing movement of gaseous matters within plume, T_SMS, T_MS, T_IMB—trajectories of sub-mm-, and mm-sized spherule, and impact-melt bomb respectively within the impact plume and its surroundings

These morpho-chemical differences between the sub-mm and mm-sized Lonar impact spherules could be caused by their trajectories within the expanding impact plume (Fig. 12). It is a fact that all the components of the plume and parent liquid droplets of the Lonar impactites gained the same impact-induced kinetic energy (K. E. = ½mv², m—mass, v—velocity) just after the impact. However, their trajectories within the impact plume were controlled by their masses. The parent liquid-droplets of the sub-mm sized spherule for their tiny masses perhaps managed to get sub-vertical trajectory that reached to the highest elevation of the plume and passed through the hot central part of the expanding plume just after the impact (Fig. 12b). By contrast, the parent liquid droplets of the mm-sized spherule and impact-melt bomb, due to their higher masses followed more gentle trajectories and remained within the peripheral part of the plume without any significant meteoritic components.

It appears that the mm-sized spherule and impact-melt bomb may have lost Cu during their formation or cooling (Fig. 8c), and the temperature of the peripheral plume could be constrained higher than the condensation temperature of Cu of ~760 °C (Lodders 2003).

7 Conclusions

• 1. In contrast to the morphology of the sub-mm sized spherule (Misra et al. 2009), the mm-sized spherule from the Lonar crater contains vesicles and micro-xenocrystic plagioclase, pyroxene and titano-magnetite of the target basalt, which are more common towards the marginal part of these spherules, indicating the role of impact-induced spin of their parent liquid droplets within the impact-plume for partial separation of melts from the volatile and micro-xenocrystic phases. These spherules also lack the schlieren structure defined by tiny magnetite crystals, which are common in sub-mm sized spherules.

• 2. The sub-mm sized Lonar spherule shows enrichment in impactor components (e.g. Cr, Ni), whereas the mm-sized spherule is relatively enriched in Si, Na, K and P, although all the Lonar impactites, in general, are relatively depleted in Na compared to that in the target basalt.

• 3. The morpho-chemical difference between the Lonar impact-spherules suggest progressively heterogeneous nature of the Lonar impact-plume in terms of temperature and vapourized impactor component particularly at the end of excavation stage of the crater. The central part of the plume was hot (could be more than 1,100 °C) and enriched in vapourized impactor component mostly towards the head of the plume. There was a peripheral part of the impact-plume, which was relatively cool and devoid of impactor components.

• 4. The trajectory of the sub-mm sized spherule was perhaps sub-vertical and mostly restricted within the hot central part of the plume where they incorporate most of the impactor components through mixing of their parent liquid droplets with the vaporized impactor (and perhaps other) components. On the other hand, the trajectory of the mm-sized spherules was relatively gentle and mostly restricted towards the peripheral part of the impact plume where their parent liquid droplets chilled at relatively low temperature and thus contain relatively more volatile components like Si, Na, K and P but without any component of the impactor.

• 5. The general loss of Cu from the mm-sized spherule and impact-melt bomb constrain the lower temperature of the peripheral plume more than 760 °C.