The first RNA-based life is thought to have emerged during the Late Hadean Eon, a time when the Earth had largely accreted and differentiated into the core, mantle, and athenosphere, but was still subjected to frequent large impacts by projectiles 200-1000 km in diameter and potentially impactors as large as 3000 km in diameter. This presents problems for our understanding of the emergence of life, as such large impacts could potentially have steralised the surface of the Earth. However, it is also quite likely that these impacts played a key role in the development of life, by altering the composition of the atmosphere. The atmosphere of the Earth during the Late Hadean is thought to have been weakly reducing, comprised largely of carbon dioxide, nitrogen, and water, under which conditions RNA is unlikely to form naturally. However, an impact by a large asteroid could potentially change this, delivering substantial amounts of free iron into the atmosphere, and causing the formation of a strongly reducing atmosphere, rich in hydrogen and methane, under which conditions RNA forms readily.
Understanding these events has proven to be difficult. Early attempts at modelling Hadean impacts assumed that 25-50% of the projectile kinetic energy was partitioned towards evaporating a pre-existing ocean, and that all of the projectile iron is available to reduce pre-existing surface water, but it is far from clear how accurate this is, nor how an impact by a bolide 200-3000 km in diameter would actually affect the habitability of the Earth.
In a paper published on the arXiv database at Cornell University on 23 January 2022, and submitted to the Planetary Science Journal, Robert Citron and Sarah Stewart of the Department of Earth and Planetary Science at the University of California Davis, present the results of a study in which they modelled a set of 3D simulations of large impacts on the early Earth in order to better understand what types of impacts would either sterilize the early Earth or provide sufficient iron to sustain a post-impact reducing atmosphere.
Citron and Stewart modelled the potentiall effects of bolides between 1500 and 3400 km in diameter impacting the Earth at a range of velocities and angles, in order to estimate the degree of surface melting, and ocean vaporisation they would cause, as well as the distribution of iron they could deliver. They attempt to take into account the state of the behavior of the iron core and forsterite mantle during such impacts, and how this would affect the degree of vaporisation during a giant impact event, and therefore the subsequent composition of the post-impact atmosphere.
They estimate that sterilising impacts would need to be larger than previously calculated, but nevertheless that they are likely to have happened several times during the late accretion phase. Previous estimates have suggested that a Ceres-sized object (473 km in diameter) would be capable of vaporising several ocean-masses of water (although this might not be completely sterilising), while a Moon-sized object (1737 km in diameter) would completely melt the Earth's surface. Furthermore, only the very largest impacts (of objects around 3400 km in diameter, would deliver enough iron to the Earth's surface to produce a strongly reducing atmosphere, at which size it is unclear how much of the iron would be absorbed by the resultant melt, making it unavailable for subsequent reactions with water. Thus, sterilising impacts are likely to have been common during the late accretion, but forming a strongly reducing atmosphere through impact-delivered iron may require additional mechanisms to prevent that iron being sequestered in the upper mantle.
Habitability of Earth in the aftermath of late accretion impacts depends strongly on the post-impact melt and iron distribution, and the energy deposited into the post-impact atmosphere. (a) In a nominal oblique impact a portion of the mantle is melted and ejecta heating also occurs downrange of the impact. The post-impact Earth has an ambiguous surface boundary of supercritical fluid, and the atmosphere consists of volatiles and a mix of vaporized silicate and projectile iron. Projectile iron is deposited in the melt/mantle/atmosphere, and some fraction also escapes the system with the impact ejecta. (b) In a head-on impact the projectile iron penetrates deep into the mantle, potentially reaching the core for larger impacts, and little material escapes the system. (c) In hit-and-run impacts, less melt is generated by the impact and a significant fraction of the projectile escapes the system. Citron & Stewart (2022).
The Earth's core is believed to have formed during the impact that formed the Moon, removing a large proportion of the Earth's iron and highly siderophile elements (i.e. elements which will completely dissolve in an iron melt) from further interaction with the Earth's outer layers. Based upon the observed content of iron and highly siderophile elements in the Earth's mantle, it is assumed that about 0.5-1% of the Earth's mass (rock and iron) was accumulated after this impact, material sometimes known as the Late Veneer. This provides a minimum mass for the total material deposited onto the Earth during the late accretion, which is equivalent to a single impactor with a diameter of between 2300 and 2900 km, although if some of the material delivered is assumed to have entered the core through extensive melting, then the total deposited could be two to five times as large. If it is assumed that the material impacting the Earth came from objects with a similar size distribution to that found in the Inner Asteroid Belt, then the majority of this material is likely to have been delivered by a single large object, with a diameter in excess of 2500 km. This would also explain the greater proportion of iron seen in the Earth's upper layers than their equivalents on the Moon, as material from a small number of large impactors would be preferentially deposited on the Earth, due to its larger gravitational cross-section.
There are currently two rival models of impact events during the late accretion. The Late Heavy Bombardment suggests a spike in impact activity about 3.9 billion years ago, probably caused by a dynamical instability in the orbits of the giant planets, which scattered trans-Neptunian objects, asteroids, and leftover planetesimals into the Inner Solar System, producing the pattern of impact basins seen on the Moon and other bodies. The alternative tail-accretion scenario suggests that the level of bombardments declined steadily as the Solar System aged and the available material in the Inner Solar System was used up. Both these models could potentially deliver a similar amount of material to the Earth, and altered conditions on the planet by either sterilising its surface or donating enough iron to produce a strongly-reducing atmosphere, but the timing of the impacts is somewhat different.
Logically, the emergence of life on Earth, or at least that life from which modern organisms are descended, must have happened after the last sterilising impact. Since it is reasonable to assume that no subsequent impacts were larger than this event, it is logical to calculate the size of the minimum event that could have sterilised the Earth's surface. If early life was dependent on the presence of a liquid ocean, then the minimum size of a sterilising impact can be assumed to be equivalent to the minimum size of an impactor which could vaporise the Earth's oceans. Alternatively, if life was restricted to the surface layers of such an ocean, then an impact capable of vaporising the upper 200 m of the ocean would be enough to sterilise the Earth's surface. If, however, early life was present in subsurface environments, then vaporising the oceans might not be sufficient to sterilise the planet, particularly since large impacts tend to produce large shockwaves, which could potentially bury organisms deep enough to protect them from any transient atmospheric superheating. Under such a scenario it might be necessary to completely melt the Earth's surface to stelrilise the planet, or, more conservatively, raise it's temperature above the limit for hyperthermophilic organisms (80-110°C). It is unclear if the Earth's surface has been completely melted since the Moon-forming impact, and therefore conceivably possible that the Earth's surface might have been continuously habitable since this event.
The minimum size of an impact needed to vaporise the oceans is also unclear. While a large enough impact will vaporise any water in the region where it impacts, vaporising a global ocean would probably require the planet to be enveloped in a cloud of impact-generated silicate vapour, radiating heat down at the surface of at least 1500 K (1227°C). It has been suggested that the complete vaporisation of an ocean with the same volume as that of the modern Earth could be achieved by an object 350-440 km in diameter, if it is assumed that 25-50% of the impact kinetic energy is directed towards vaporising surface water; but this hypothesis has not been subjected to vigorous mathematical modelling.
Similarly, the size of an impactor needed to melt the Earth's surface is far from clear. Most models of the outcome of large impact events are based upon scaling up the effects of smaller events, but such scaling up cannot really account for the extra volumes of melt generated by impactor in excess of 100 km in diameter. Furthermore, melt generated by large impacts tends to remain within the impact basin, limiting its global impact. It has been suggested that impacts by objects in excess of 100 km might result in a layer of melt 3 km thick covering an area 20-30 times the diameter of the object. If this is realistic, then global surface melting would require an object 1300-2000 km in diameter. However, global melting could potentially be achieved by the deposition of ejecta from an impact, with some models suggesting an impactor as small as 500 km in diameter could result in the Earth's surface being covered by a layer of molten ejecta 200 m thick; however this has only been tested in two dimensional model, and it is not clear if it would actually be enough to cover the whole Earth's surface, or whether areas of unmelted crust might remain. Heating the entire surface of the Earth to above 80-110°C would clearly require a smaller body than completely resurfacing the planet, with some models suggesting that a body as small as 300 km might be able to achieve this, although the models involved relied on grid sizes of 50-100 km, leaving plenty of space for patches of surface to remain habitable.
As well as sterilising the surface of the Earth, large impacts during the late accretion could potentially have donated enough iron to the Earth's surface to temporarily create a strongly reducing atmosphere, suitable for the abiological formation of RNA, which in turn is hypothesized to have predated the modern DNA-based biology. This would have required the formation of a large number of RNA precursor compounds, most of which would form readily under strongly reducing conditions, but which are less likely to have formed in the weekly reducing atmosphere predicted for Late Hadean Earth. This weekly reducing atmosphere is thought to have formed rapidly after the Moon-creating impact led to the differentiation of the core and mantle, removing most of the iron from the mantle, which became oxidized to a state near the fayalite-magnetite-quartz bu ffer, but thus could have potentially been altered by the delivery of a large quantity of new iron to the system from an extra-terrestrial source.
Such a reducing atmosphere can occur through the reaction of iron with water, forming iron oxides and free hydrogen (a highly reducing gas). An small planetary body with a core mass fraction of 0.3 (roughly the same as that of the Earth), and a mass of approximately 0.5% of that of the Earth, would produce about 10 000 000 000 000 000 kilotons of iron, enough to reduce 2.3 ocean masses of water, generating a hydrogen atmosphere with a pressure of several bars. This hydrogen would probably react quickly with carbon to form methane, another reducing gas, and creating conditions ideal for the formation of RNA precursors.
Previous reconstructions of large late accretion impacts have suggested that if a differentiated body 3000 km wide impacted the Earth at an angle of 45°, then 60% of the mass of the core of that body should be available to react with material present on Earth. However, these studies have not taken into account whether this material would be available to react at the Earth's surface or be sequestered into the mantle.
Citron and Stewart's model simulated impactors with masses equivalent to 0.0012, 0.003, 0.006, and 0.012 times that of the Earth, which would correspond to objects with diameters of 1500, 2000, 2700, and 3400 km, or one tenth, one quarter, one half, and the entire mass of the Moon, impacting at velocities of 1.1, 1.5 and 2 times the Earth's escape velocity (which is 11.2 km per second), impacting at angles of 0°, 30°, 45°, and 60°, where 0 corresponds to a head-on impact.
These simulations suggested that when objects impacted the Earth at an oblique angle, this resulted in the formation of a blanket of ejecta in the form of silicate vapour that encircled the Earth. The material from the core was in all cases quickly absorbed into the Earth's core, until the model was adjusted to take into account structural strength of both bodies. This resulted in the majority of the iron being retained in a melt pool at the impact site; Citron and Stewart suspect that much of this material would subsequently find its way to the core, but this was beyond the scope of the model.
Snapshots of a simulation of a 0.006 Earth mass projectile impacting a Earth target at 1.5 times the Earth's escape velocity, at an angle of 45°. Colours track the projectile and target core and mantle materials. Axis units are in present-day Earth radii (6371 km). Citron & Stewart (2022).
For each simulation, Citron and Stewart quantified the potential of the impact to sterilize the early Earth either through vaporization of a pre-existing ocean or globally melting the surface. In most cases the majority of the objects kinetic energy was transferred to the Earth's surface or interior, however in the case of the most high-velocity or oblique impacts, much of this energy escapes the system as unbound ejecta.
For the sake of the model, Citron and Stewart assumed that the vaporisation of any ocean would be caused by thermal radiation caused by the impact, although they acknowledge that this is unlikely to be the case with impacts large enough to melt the Earth's entire surface. Therefore, in order to calculate the ability of an impactor to vaporise all water on the Earth's surface, Citron and Stewart look at that body's ability to create a world-enveloping cloud of hot, silicate vapour.
Citron and Stewart calculate that 1-21% of the projectile kinetic energy is directed into the impact as thermal energy. This is lower than previous estimates, and does not take into account the heating caused by hot iron ejecta on the planet's surface. They further estimate that an atmosphere heated to 2300K (2027°C) would cool within 1 to 100 years due to simple radiative cooling.
All of the modelled impacts produced enough energy to vaporise an ocean with equivalent mass to that of the modern Earth, although this did not take into account energy radiated back into space, which would reduce the energy available for ocean vaporisation. Citron and Stewart estimate that about half of the energy entering the atmosphere following a major impact would be radiated back into space, although the extent of this would depend on the composition and conditions within the atmosphere; cooler initial atmospheres would be likely to radiate more energy out into space following an impact. They also note that factors other than direct heating would contribute to ocean heating, such as the energy contained within silica drops precipitating out of the atmosphere and falling to the surface.
The majority of the modelled impacts produced layer of melt 100s of meters to 100s of kilometres thick, which totally covered the Earth, and which therefore would presumably have sterilised its surface of all life. Only the smallest impacts, with masses equivalent to 0.003 times that of the Earth, produced only localised melting, with more oblique impacts producing additional melting downrange of the impact site. Despite this, Citron and Stewart calculate that only the very largest impactors could reliably sterilise the whole planet, and then only if they impacted it at an acute angle, more oblique impacts are less reliable, with even the largest impactors failing to sterilise the whole planet if they hit at an angle of 60°.
Distribution of melt thickness 24 hrs post-impact for several example simulations. In the plotted coordinates the impact direction is east along the equator. The longitude of the impact point 24 hrs post impact varies based on the planetary rotation induced by the impact. Citron & Stewart (2022).
For each simulation Citron and Stewart also calculated the proportions of iron that were deposited into the interior of the planet, deposited on the surface of the planet, remained in the atmosphere, or were ejected into space. In the case of the most acute impacts, the object was burrowed deep into the planets mantle, and were quickly covered over by the mantle material, preventing any further interact with the iron from the impactor then unavailable for further reactions. At oblique angles, much of the material is lost into space as ejecta, but more is also deposited on the surface, with the amount deposited at the surface tending to increase with larger or faster objects.
The ability of any impact to produce a reducing atmosphere depends on the amount of iron deposited in places where it can interact with large volumes of water. It has been suggested that an ocean 1-3 times the mass of the current oceans would need to be completely reduced in order to generate an atmosphere rich enough in hydrogen and methane for the RNA precursors needed to kick-start the RNA world to form. It is unlikely that all of the iron derived from any impactor would become available for reducing such an ocean. Iron which penetrated the mantle would quickly be covered-over by other material, becoming unavailable for reactions with water, and it is likely that a high proportion of material deposited at the surface would be covered over by other ejecta. In fact Citron and Stewart's simulations suggest that much of the iron deposited on the surface following an impact would be deposited in highly melted areas, where it would be easy for it to sink down to deeper layers within the Earth. Such iron might still have an impact on the Earth's atmosphere over a geological timescale, releasing some reducing gasses from ground sources for the next 10-100 million years, which might help sustain a reducing atmosphere once formed, but would be unlikely to result in one forming.
For these reasons, Citron and Stewart suggest that the major reserve of iron for the development of a reducing atmosphere would be iron which is vapourised during any impact, and retained, at least temporarily, within the atmosphere. A post impact atmosphere would be expected to contain a significant amount of vapourised iron, rock, and water. In such an environment iron should react readily with water, either combining to form iron oxides while in a vaporous state, or precipitating out to form droplets which would react with water either while still in the atmosphere or after falling to Earth. The total mass of the iron vapourised into the atmosphere by the impact will provide an upper limit on the amount of iron available to reduce water. Thus, while larger impacts would completely melt the Earth's surface and probably lock the majority of their iron away more rapidly and thoroughly, they are also likely to contribute more iron to the atmosphere, and create a hotter post-impact atmosphere, where vapourised iron might remain for longer.
In Citron and Stewart's simulations, only large impacts at an angle of 45° delivered enough iron into the atmosphere to reduce an ocean-sized body of water. A 0.003 Earth mass object delivered a maximum of enough iron to reduce half an ocean. Most impacts delivered mor iron to the atmosphere than to the surface of the planet, with the exception being head-on or highly incident collisions (0°-30° from perpendicular). All effects taken into acount, the maximum amount of iron any impact could be expected to deliver into the atmosphere would be enough to reduce 1.5 ocean-masses of water.
Citron and Stewart's simulations suggest that large impacts during Late Accretion would have been a highly inefficient way to deliver iron to Earth-systems where it could act as a reducing agent. In most cases the majority of the iron from an impact was buried deep into the crust or mantle, from where larger volumes of iron could potentially sink further, into the core, with much of any iron not sequestered deep within the Earth being ejected into space. They further predict that a large proportion of any iron entering the atmosphere will precipitate out rapidly, falling to the surface while that is still molten, and subsequently being lost from the system.
This presents serious problems for the idea that a large impactor might have completely reduced the Late Hadean Earth's oceans to form a strongly reducing atmosphere, thus facilitating the development of RNA. It has been calculated that almost all the iron from an impactor 2500-3000 km in diameter would be needed to reduce an ocean double the size of the Earth's current ocean (which is predicted for the Late Hadean Earth, as the warmer mantle would lack the retention ability of the modern ocean), but Citron and Stewart's calculations show that even for the very largest impacts, most of this iron would be locked away rapidly rather than becoming available for reactions.
This does not completely rule out this scenario as a possible precursor to the appearance of RNA. It is possible that the Earth could have been impacted by an object significantly larger than is predicted by the observed excess of mantle highly siderophile elements, or by a smaller object with a much larger iron core than we would typically expect.
Alternatively, if large amounts of iron were deposited into the crust or upper mantle, but remained in contact with an impact-generated melt pool, then it is possible that this could equilibrate with the atmosphere, allowing this source of iron to contribute to the reduction of atmospheric gasses. This would require the iron in the melt to remain distributed throughout the melt, rather than settling out at the base, something which might happen if the iron was broken down into droplets circulating in an emulsified melt.
Even if iron was trapped within the crust beneath a a surface fusion crust, or layer of non-ferrous melt, it could potentially react with water in the crust, forming iron oxide and releasing hydrogen, which could be released into the atmosphere either as gasseous hydrogen, or methane if it reacts with a source of carbon within the crust.
It is also possible that iron ejected from the atmosphere during an impact could play a role in reducing the atmosphere, if it formed a debris ring which fell back to Earth over time. It has been estimated that 20-30% of the material which was ejected from the atmosphere during a large impact would fall enter such a debris ring, falling back to Earth over the next few tens of millions of years. Citron and Stewart's model suggests that with an impactor hitting the Earth at an angle of 60° from the perpendicular, almost all of the iron would be ejected from the atmosphere. If 20% of this later fell back to Earth, then this would be equivalent to the amount of iron delivered to the atmosphere by a similar sized object impacting the Earth at 45°. Furthermore, because this iron would be falling back to Earth in smaller pieces, it would be far less likely to bury itself in the ground and become unavailable for reaction with the atmosphere, rather many of these objects might 'burn up' in the atmosphere, vaporising before they impact the ground.
An alternative hypothesis is that the Earth's atmosphere was repeatedly transiently reduced by much smaller (100 000 000 000 000-100 000 000 000 000 000 kiloton) impactors. Such impacts would only produce very transient reducing atmospheres, but this might still allow for the formation of nucleobase precursors and ammonia, before returning to equilibrium with the fayalite-magnetite-quartz mantle, where these compounds might further react with compounds present in such an atmosphere and produced by volcanic eruptions. Potentially, an atmosphere which swung back-and-forth between two redox states might have been more beneficial for the emergence of life, as this allows for a greater range of chemical reactions.
If the ancestor of modern life did originate in a post-impact strongly reducing environment in the Late Hadean, then the Earth must have avoided any subsequent sterilising impact event. This is difficult to explain, as Citron and Stewart's model suggests that the size of an impactor needed to produce a strongly reducing atmosphere is considerably larger than the size of an impactor needed to sterilise the Earth. Thus, sterilising impacts should have been more common than impacts producing strongly reducing atmospheres, and if, as seems likely, the scale of impacts reduced slowly over time as larger planetary bodies in the Inner Solar System were swept up, impacts capable of sterilising the Earth should have persisted after impacts with the potential to produce a reducing atmosphere had ceased.
However, Citron and Stewart's model does suggest that complete sterilisation does require a larger body than previously assumed, with a body of 2000 km being the minimum with the potential to melt the entire surface of the planet, and a body of about 2700 km needed to reliably do this. Acheiving the lesser objective of vaporising the oceans, which might sterilise the Earth, would require a body of at least 740 km according to Citron and Stewart's model, whereas previous estimates have assumed a body of 350-450 km could do this. However, Citron and Stewart do note that vaporising only a portion of the ocean could potentially turn the remaining water into a hot, dense, brine, which might be uninhabitable to early organisms, thereby achieving sterilisation.
If an object at the upper end of this range is needed to completely sterilise the surface of the Earth, then this might leave room for an organism evolving in a post-impact reducing atmosphere to survive and leave ancestors still alive today. It is quite plausible that an impact by a body more than 3000 km in diameter could have been followed only by impacts smaller than 2000 km.
Citron and Stewart's model suggests that Late Hadean impacts could have been less likely to sterilise the Earth's surface than previously thought, but also that such impacts are likely to have delivered far less available iron than previously thought. Thus, if life did require a strongly reducing atmosphere in order to originate, then this would probably have required an object of at least 3400 km in diameter to impact the Earth. After such an origin, life would have needed to survive without any subsequent impacts sterilising the Earth, which also might have required a considerably larger impact than previously thought.
See also...
Follow Sciency Thoughts on Facebook.
Follow Sciency Thoughts on Twitter.
Hartbeesfontein Basin (South Africa) Stromatolite textures. (A) Location map. Black square in inset represents map location. (B) Domal stromatolitic chert; hammer head is 2 cm tall. Wilmeth et al. (2022).
