Penciling in details of the Hadean

Some truly remarkable graphite is described by Bell et al. in PNAS (1). Graphite is, of course, the same material as found in pencil tips or in the anode of lithium ion batteries. Graphite is, however, also a very common material in Earth Science, and is often the form of carbon found in very old fossils that have been subjected to substantial heat. The graphite described in the Bell et al. article is remarkable because it is exceptionally old, dating to the Hadean eon. Officially, the Hadean is defined as the time period from the formation of the Earth until 4 billion y ago. Until recently, this has been a seemingly convenient definition, leaving it as the geological eon without a rock record on Earth. Over the past quarter century, however, the discovery and exploration of detrital zircon minerals from the Jack Hills conglomerates of Western Australia (2) have provided a new window into this early time. Jack Hills zircons crystallized in magma chambers at various times as far back as 4.4 billion y ago (3, 4). So far, these zircons and their inclusions are currently our only tangible record of the first half a billion years of Earth history. Based in part on the extreme age of some of these zircon minerals, along with similarly old age dates for a Martian meteorite (5), the Planetary Science and Earth Science communities now appreciate that planets form and cool rather quickly (6). No longer is the Hadean just a placeholder on our timelines between the formation of the Earth and the oldest known rocks. The Hadean Earth is represented by tangible samples that appear to have formed in a continental crust setting, likely above an active subduction zone (7). Specifically, Bell et al. (1) describe two sizable graphite inclusions within a 4.1 billion-y-old zircon from the Jack Hills. This sample implies that a Hadean chunk of organic material was transported by geological processes (including subduction) to a granitic magma chamber, where it was incorporated into the crystalizing minerals. The sheer existence of this new record of early carbon is exciting because it provides a new window into the seemingly accessible Hadean Earth.

The Hadean graphite provides a new constraint on the carbon isotopic composition of reduced carbon likely deposited in sediments at that time. In the upper left portion of Fig. 1, an electron micrograph of the Hadean zircon is shown along with the δ¹³C value of −24±4‰, as measured using an ion microprobe (1). In geochemistry, δ¹³C values are reported relative to an inorganic carbonate reference, meaning that this negative value is quite depleted in ¹³C relative to typical inorganic carbonate rocks. In fact, the value is indistinguishable from biological carbon fractionated by microbial carbon fixation. During carbon fixation, enzymes preferentially incorporate ¹²C, leaving inorganic carbon enriched in ¹³C. Thus, one of the most important geological records for understanding the long habitability of the Earth is the record of the carbon isotopic composition of organic material and inorganic carbonates back through time (8, 9). Broadly, this record convincingly shows the sustained impact of global marine carbon fixation on preserved sedimentary carbon going back at least 3.5 billion y. Throughout this long history, microbial life in the oceans would have had a δ¹³C value of about −25‰ (Fig. 1), quite distinct from inorganic carbon with a δ¹³C value of about 0‰. This record of life has potentially been pushed further back to ∼3.9 billion y ago with the discovery of ¹³C-depleted graphite in Greenland metasediments (10, 11). This new discovery of ¹³C-depleted graphite from 4.1 billion y ago potentially pushes biological carbon into the previously uncharted Hadean. The implication is that there was a substantial amount of potentially biogenic carbon on the Earth 4.1 billion y ago, 200 million y before the next known sedimentary carbon. Extending the biological carbon cycle further back based on this graphite inclusion is simple, consistent, and in line with the principle of uniformitarianism. Thus, this discovery suggests life was flourishing in our oceans near the end of the Hadean, if not earlier. The next clear step will be to test the continuity of possible biogenic graphite preservation in zircon. If this does indeed record ancient sedimentary carbon preserved despite a trip through a magma chamber, then such a record in zircon should be recoverable through the subsequent 4 billion y of Earth history.

Fig. 1.

Summary of the few carbon isotopic (δ¹³C) and redox constraints we have for the Earth’s earliest carbon cycle. In the upper left is the newly reported carbon isotopic value for a graphite inclusion in a 4.1 Ga Hadean zircon (−24 ± 4‰). This remarkable graphite provides a new constraint on the global carbon cycle in the Hadean. Also shown are Precambrian diamonds (bimodal at −5‰ and −25‰) and Early Archean microbial biomass (scattered around −25%). The Hadean zircon hosting the graphite comes from a continental crust setting, shown as thick and dark brown. Typically, continental crust forms in magma chambers above a subducting slab of oceanic crust, shown here in black. Finally, various measures of redox suggest that the Earth’s present upper mantle is mildly reducing and has been such for >3.8 billion y. (Inset) Similar constrains for magmatic carbon found in lunar rocks (around −20‰) and Martian meteorites (around −20‰). Both the Moon and Mars appear to have highly reducing mantles. (Inset image adapted from ref. 1.)

This ancient carbon appears to have had a wild ride, having been in a magma chamber, but looking at carbon from extreme geological settings is required to understand the Earth’s carbon cycle. This new isotope measurement is one more constraint added to a substantial body of similar research on materials extruded from deep within the Earth. For example, the new measurement is similar to some mantle diamonds. Fig. 1 includes a summary of several decades of carbon isotopic measurements of diamonds extruded from the Earth’s mantle in violent explosive eruptions. Diamonds show that the mantle is bimodal with respect to carbon isotopes. Most diamonds have carbon isotope values clustered at −5‰, but there is another less-abundant group of diamonds with carbon isotopic values spread around −25‰ (12, 13). Based, in part, on this evidence, the mantle is believed to have an isotopic composition of −5‰, and the prevailing view is that the more ¹³C-depleted diamonds represent biogenic carbon that has been subducted into the mantle (14). In this interpretation, the transition zone diamonds and the Hadean graphite share a similar narrative and a similar wild ride. Again, the story is simple, consistent, and in line with the principle of uniformitarianism.

When beginning to sketch in the details of the Hadean, when does uniformitarianism fail? In other words, at which ancient time intervals do we know that the Earth operated differently? One simple answer is that the Earth was certainly different when if first formed 4.5 billion y ago. The few direct constraints we have for Earth at that time are also shown in Fig. 1. Although the redox state of the upper mantle has been mildly reducing (buffered by a mixture of ferrous and ferric iron) for at least the past 3.8 billion y (15), lunar samples demonstrate that the Moon is strongly reducing (16). Because the Moon formed from a massive impact of the proto-Earth, lunar rocks to a degree demonstrate aspects of the Earth in its final stage of formation. The simplest conclusion is that the Earth’s upper mantle was significantly more reduced in the beginning. Growing evidence indicates that magmatic carbon on Mars (17, 18) and the Moon (19) has a carbon isotopic

Bell et al. describe two sizable graphite inclusions within a 4.1 billion-y-old zircon from the Jack Hills.

composition of about −20‰. This is significantly more ¹³C-depleted than that of Earth’s present upper mantle. At present, the textbook answer is that the Earth and Mars started out with different carbon and that the Moon is just weird, potentially having been influenced by meteorite impacts.

A wild alternative suggestion is that the bulk of Earth is actually more ¹³C-depleted than we think, with an average carbon isotopic value around −20‰. For example, the Earth’s core could contain significant C, resulting in a fractionated mantle (20). In any case, I think it is worth considering that at its beginning, the Earth was highly reducing and had significant ¹³C-depleted carbon in its upper mantle. In other words, the earliest Earth may have looked like our present Moon. Certainly, during the Hadean the Earth’s mantle changed to become more oxidizing and to settle on an upper mantle isotopic composition similar to today. In the beginning, however, a prebiotic soup could have produced substantial sedimentary carbon, and it is possible that the carbon was more ¹³C-depleted than we normally consider. Bell et al. (1) cover their bases by appropriately calling their graphite potentially biogenic. Is the graphite they found from a prebiotic soup? It is possible and would be equally exciting, but I do not think so given that the zircon’s crystallization age is over 400 million y after the Earth’s formation. Four-hundred million years is a long time. In terms of geophysics, it is more time than plate tectonics took to both assemble and break-up the supercontinent Pangea. However, the Jack Hills zircons extend back to about 4.4 billion y ago, roughly 100 million y after the Moon was formed. If graphite continues to be found in more ancient zircons, distinguishing between a dense prebiotic soup and a productive carbon-fixing biosphere could be a challenge.