Pleistocene organic matter modified by the Hiawatha impact, northwest Greenland

Most terrestrial impact structures contain only little organic carbon, recycled from target rocks. The ca. 35 Ma Popigai crater, Siberia, contains diamonds transformed from target-rock graphite (Masaitis et al., 1975). Ejecta from the ca. 23 Ma Haughton crater, Canada (Parnell et al., 2007), and the ca. 15 Ma Ries crater, Germany (Osinski 2003), contain organic carbon derived from sedimentary target rocks, and drill core from the 66 Ma Chicxulub crater, Mexico, contains particles of charcoal (Gulick et al., 2019). If the source of the carbon can be identified, it may serve to constrain the maximum age of cratering. For instance, dispersed graphitic carbon in the ca. 500 Ma Gardnos crater in Norway stems from Cambrian Alum Shale (Gilmour et al., 2003; Parnell and Lindgren 2006). Carbon inclusions in impact glasses linked to the purported ca. 0.8 Ma Darwin crater, Tasmania, preserve biomarkers from contemporaneous vegetation (Howard et al., 2013), and the small, twin ca. 3.2 ka Kaali craters, Estonia, contain locally derived charcoal (Losiak et al., 2016, 2019).

With a diameter of 31 km, the newly discovered Hiawatha crater under the Greenland Ice Sheet in northwest Greenland is one of the 25 largest impact structures on Earth (Fig. 1; Kjær et al., 2018; Impact Earth, 2020, https://impact.uwo.ca/). A large variety of sand-sized impactite grains originating from the crater floor were found in a glaciofluvial outwash sediment sample (HW21-2016 of Kjær et al.; see their table S1 with sample information) deposited at the front of the Hiawatha Glacier just outside the crater rim (Fig. 1). The sand contains shocked quartz, microbreccias, shock-melted mineral glasses, and micromagmatic aggregates; elevated Pt contents and anomalous platinum-group element ratios in this sediment suggest that the impactor was an iron meteorite (Kjær et al., 2018). The bulk mineral assemblage and chemical composition of the sand indicate sourcing from granulite-grade rocks similar to Paleoproterozoic paragneiss exposed in the ice-free foreland to the crater (locally including sulfidic paragneiss with graphite flakes; Dawes, 2004). The crater has not been directly dated but is tentatively referred to the Pleistocene and possibly as young as the last glacial period, based on indirect evidence such as preservation of the subglacial crater topography, a strongly disturbed pre-Holocene ice stratigraphy over the crater identified by airborne radar imaging, and an anomalously high modern-day volume of melt water draining from the impacted area (Kjær et al., 2018).

In this study, we investigate organic matter and its thermal degradation in the impactite grains, charcoal, lignite, and bulk sediment in proglacial outwash from the Hiawatha crater floor. We identify the wood fragments to genera level and discuss the implications of our findings for the maximum age of the Hiawatha impact.

Sediment samples described by Kjær et al. (2018), and charcoal and lignite collected by us from in front of the Hiawatha Glacier were investigated by standard optical and backscattered electron scanning electron microscopy (SEM-BSE; Geological Survey of Denmark and Greenland, Copenhagen, Denmark), pyrolysis, and Raman spectra. The Raman spectra were obtained with a WITec alpha300 R system and a 488 nm laser at the Alfred Wegener institute (Bremerhaven, Germany); see the Supplemental Material¹ for details. Organic carbon geochemistry in bulk sediment sample HW21-2016 and two reference samples was obtained by Hawk pyrolysis at Aarhus University (Aarhus, Denmark) following Carrie et al. (2012), measuring the quantity of total organic carbon (TOC, in weight percent) and proto-hydrocarbons (S2, in milligrams thermally sensitive, labile bio-macromolecule hydrocarbons per gram sediment) released between 300 and 650 °C from the preserved humic matter. Three grain-size fractions of each sample were measured to check potential variability with grain size. The lignite was dated by the Beta Analytic Carbon Dating Service (https://www.radiocarbon.com/) with calendar-year calibration using OxCal v. 4.3 with the IntCal13 calibration curve (Reimer et al., 2013).

We found both macro- and microscale organic matter in the outwash sediments in front of the Hiawatha Glacier. The macroscale organic matter comprises angular, sand- to gravel-sized particles of charcoal in sample HW21-2016, as well as pebble-sized lumps of lignite rounded by water transport on the eastern side of the glacier front; small twigs from dwarf bushes unrelated to the impact occur locally on the glacial outwash plain. The microscale, finely dispersed organic carbon was identified in the impactite grains of sample HW21-2016.

Microbreccia grains contain splinters of silicate minerals including shocked quartz as well as partially melted mineral fragments, set in an opaque matrix of amorphous organic matter mixed with altered glass largely derived from shock-melted alkali feldspar (Fig. 2A). Impact glass grains contain appreciable dispersed organic carbon; new magmatic microcrysts within them are commonly surrounded by carbon sheaths (Fig. 2B). A predominantly carbonaceous grain with tiny, angular mineral fragments and clayey material is shown in Figures 2C and 2D. Further examples of the association between shocked quartz, shock melts, and organic matter are presented in the Supplemental Material.

Sand- and gravel-sized pieces of charcoal from sample HW21-2016 have a largely amorphous structure (“glassy coal”) but also preserve relict cell structures (Fig. 3A). Their reflectance in incident white light is high (R_o ≤ 3.5%), within the range of modern charcoal (Braadbaart and Poole, 2008; McParland et al., 2009).

Pebble-sized lumps of lignite collected at the eastern Hiawatha Glacier front (Fig. 3B) have cellular structures and a low random reflectance (R_o ∼0.2%–0.5%). The ash content is only ∼0.5 wt%. One of these lumps yielded a non-finite ¹⁴C age of >43,500 yr B.P. (our sample Beta-471661). Most of the lignite consists of woody material with rows of well-preserved, even-sized cells with central voids, spiraled fibrous cell walls, and bordered pit pairs (Fig. 3C); layers of cork cells have filled interiors. Alternating growth zones of spring and summer wood are up to a dozen cells thick. These cell structures are diagnostic of conifer wood, most likely Pinus or Picea, whereas few fragments contain wood cells of uneven size and distinct perforation plates, which is the characteristic of the angiosperm genus Betula (Hoadley, 1990). Some conifer cork layers have greatly inflated cells with numerous spherical voids (Fig. 3D), which we interpret as results of expansion by heating with sudden release of volatile components from the lipid-rich cork cells. Common branching and tapering systems of shrinkage cracks are likewise interpreted as evidence of rapid loss of volatile components by heating (Fig. 3E). Other lumps display flattened cells with homogeneous (gelatinous) interiors and low reflectance (R_o = 0.1%–0.2%; Fig. 3C), interpreted as remnants of decaying and mildly compacted wood. Conglomerate-like masses of variably decomposed and disrupted wood particles mixed with small vitrinous particles (Fig. 3F) are interpreted as results of normal degradation of lignite and/or impact-related mixing and redeposition. Only very sporadic remnants of plant spores and/or algae have been found, and no leaf cuticula.

The organic matter was characterized using laser Raman spectrography (Figs. 4A and 4B). The positions and shapes of the carbon peaks in the silicate glasses and microbreccias show that the carbon ordering is mostly very low and variable, as is the case for the lignite (Fig. 4A). The highest degree of ordering (narrow Raman peak width) is found in charcoal with R_o = 2%–3.7%, and in small vitrinite particles in the lignite (R_o = 0.5%–0.6%). The carbon peaks in some shock-melted glasses changed shape during analysis. At first, a weak and very broad carbon peak appeared, which then abruptly intensified and narrowed on continued analysis or increased laser-beam intensity, and finally became permanent. This observation shows that the carbon in the glass is principally unordered and finely dispersed and suggests that accumulation of laser beam energy led to localized ordering and crystallization of platy, more ordered carbon.

The quantity and geochemical composition of the organic carbon in the bulk sediment sample HW21-2016 and reference samples HW12-2016 and HW13-2016 (Kjær et al., 2018) were analyzed by a Hawk carbon analyzer (Figs. 1 and 4C). Sample HW21-2016 directly draining the impact area is depleted in thermally sensitive, labile bio-macromolecule hydrocarbons (S2) compared to the two reference samples.

The organic matter is directly linked to the Hiawatha impact by its close association with impactite grains of microbreccias with shocked quartz and mineral melt glass. The charcoal itself documents a high-temperature event, presumably impact-induced incineration of surficial and/or originally shallowly buried plant material (Fig. 3A); we have not discovered formerly airborne organic particles originating from wildfires. The textural evidence of expansion and contraction in lignite (Figs. 3D and 3E) also suggests that some of it was subjected to a distinct heating episode, although its low reflectance shows that it was not affected by direct incineration (e.g., Belcher et al., 2018). In the possible scenario where the impact took place through the Greenland Ice Sheet, any lignite preserved at the margin of the crater would have been further protected by being embedded in permafrozen sediments just like the present Kap København Formation (Funder et al., 2001). Losiak et al. (2016, 2019) reported low-reflecting charcoal at the Kaali craters, without interpreting this observation.

The low degree of carbon ordering in Hiawatha lignite, charcoal, and shock-melted glasses compared with graphite in the Hiawatha foreland rocks and Gardnos breccia (Figs. 2 and 4A; Gilmour et al., 2003) and the pyrolysis data from the three bulk sediment samples suggest that the bulk of the organic carbon at the Hiawatha crater is impact related and not reworked graphite from the crystalline bedrock. The range of Raman peak positions and widths in the right side of Figure 4A and the variation trend toward initial ordering are the same as observed in Antarctic micrometeorites displaying various degrees of heating on entry into the atmosphere but no further thermal history (Fig. 4A; Dobrică et al., 2011). With subsequent maturation and metamorphism, the Hiawatha carbon peaks would be expected to have obtained still-narrower peak widths and reverted to lower peak positions, as observed in the Gardnos data. There is a pronounced loss of thermally sensitive hydrocarbons (S2) in the bulk sediment sample HW21-2016 relative to the reference sites, located farther away from the impact crater (Fig. 4C). This relative depletion of S2 in sample HW21-2016 is attributed to an intense, externally derived temperature excursion at the impact site, which appears to have incinerated virtually all humic substances. Proto-hydrocarbons are typically preserved through the diagenetic phase under burial temperatures of <60 °C (Taylor et al., 1998). Therefore, the distinct loss of labile proto-hydrocarbons in sample HW21-2016 cannot be explained by normal terrestrial diagenetic processes, but rather by an intense, localized external heating event that is well constrained with the previously reported compelling evidence of the Hiawatha impact from shocked quartz and mineral melt glasses.

Our observations and analyses of the organic carbon in front of the Hiawatha Glacier show that it stems from organic-rich beds formed at a time when tree growth at this high northern latitude was possible. Pliocene to early Pleistocene deposits at ∼80°N are known, e.g., from the 2.4 Ma Kap København Formation in North Greenland (Funder et al., 2001) and the adjacent ca. 3 Ma Beaufort Formation at Meighen Island, Canada (Fyles et al., 1991). The likely genera recorded in the Hiawatha lignite, Picea-Pinus and Betula, are common in the Pliocene-Pleistocene as well as present northern boreal forests, and therefore do not allow correlation with specific deposits. However, a correlation seems obvious with wood commonly found along rivers on the adjacent Washington Land, 150 km northeast of the Hiawatha Glacier (Fig. 1; Bennike, 1998, 2000). This wood stems from small trees, transported by melt water from an unknown source under the Greenland Ice Sheet, and comprises both Picea and Pinus, but not Betula, which is less resistant to wear. Similar sediments might well exist under the Greenland Ice Sheet where it borders Inglefield Land, although no timber has been found along the rivers here. The early Pleistocene, 2.4 Ma, Kap København Formation in North Greenland is the youngest known occurrence of forest at these high latitudes (see Funder et al. [2001] regarding the 2.4 Ma age determination), and the absence of Pinus does not preclude contemporaneous growth in Inglefield Land some 200 km to the south. In summary, the age of the organic carbon at Hiawatha is probably 3–2.4 Ma, and we favor the younger, 2.4 Ma age as the simplest interpretation and a realistic maximum age of the impact.

Charcoal and abundant dispersed organic carbon in the impactite grains of glaciofluvial sand draining the Hiawatha crater come from local, thermally degraded conifer trees with a probable late Pliocene to early Pleistocene age of ca. 3–2.4 Ma. The carbon cannot have been derived from the well-crystallized graphite in the crystalline bedrock in the target area. The age of the wood is also the maximum age of the impact, and this study provides the first solid evidence that the Hiawatha crater is the youngest known large impact crater on Earth.

We sincerely thank Adrian Jones, Timmons Erickson, two anonymous reviewers, Paula Lindgren, Anna Losiak, and Tod Waight for constructive and precise comments; Nynke Keulen for assistance with SEM imaging; and John Parnell for a Raman spectrograph of Inglefield Land reference graphite. This research was funded by the Carlsberg Foundation (Copenhagen, Denmark) and the Aarhus University Research Foundation (Denmark).