by Acad. Mikhail KUZMIN, Vinogradov Institute of Geochemistry (Irkutsk)

When was the first continental crust formed on our planet? What was the mechanism of this process? There are a lot of puzzles left, and now only more or less reasoned assumptions are pronounced on the processes at the giant natural laboratory, which took place more than 4 bln years ago, and on the "ingredients" used in these processes. The up-to-date methods of geochemical science of last years enable scientists to better imagine the happenings of so remote past.

Fig. 1. Model of the core formation by "using" the deep magmatic ocean (Wood, Walter, 2006; Wood, 2011).

Fig. 2. Change of the isotope ratio of tungsten isotopes determined in different systems (metal core, carbonaceous chondrites, silicate mantle).

THE NEW COGNITION PHASE

Until recently, in the first half of the 20th century, when the geosincline concept was the basic paradigm of geology or more simply of mobile belts of the Earth, it was considered that the main geological history of the planet covered Phanerozoic (it included Paleozoic, Mesozoic and Cenozoic eras of a total duration of 540 mln years up to the present time). With the advent of the plate tectonics concept* (the 1960s), there started studies of lateral ties of coeval structures of magmatic and ore formations. In this connection it was necessary to determine the age of formation of the asthenosphere**, which is a source of the oceanic magmatic rocks. Of great significance was search for the most ancient ophiolitic complexes***, which had once formed spaces under oceans. Discovery of such complexes comparable with modern rocks under oceans has showed that the plate tectonics similar to the modern one began "to work" on the planet about 2 bln years ago.

Recently astrophysicists after their studies of problems of the early formation of the Solar System and its planets, suggested hypotheses of the Moon's formation and manifestations of the coeval processes in the Earth-Moon system in the distant past.

* See: M. Kuzinin, V. Yarmolyuk, V. Kravchinsky," Plutonic Geodynamics as the Mechanism of Earth Evolution", Science in Russia, No. 6, 2013. - Ed.

** Asthenosphere is a bed of reduced hardness, strength and viscosity of the Earth's upper mantle, which underlies lithosphere. Its upper boundary is at a depth of 200 km and more under the continents, from 40 km under the median ranges and up to 100 km under the ocean bottom while the lower boundary is at a depth of 600 km. - Ed.

*** Ophiolitic complexes are associations of rocks encountered on the continents. They are considered debris of ancient oceanic crust preserved in the folded belts of the Earth. - Ed.

Unfortunately, the national geologists and geochemists are not yet carrying out detailed studies of the beginning of geological history of the Earth. This paper is an attempt to present to the reader the early stages of the Solar System formation and evolution on the basis of contemporary literature. They are divided into a chaotic eon (epoch) and the following Hadean eon, when the Earth's inner structure-core, mantle and the first continental crust-began to form. Worthy of mention is also the start of the Archaean eon (its total duration is from 4 to 2.5 bln years) when the first extant rocks of the continental crust (granite-greenstone associations) were forming.

The new approaches to studies of the origin processes of our planet could have been impossible without perfect scientific instruments. In the past decades there began a new stage in the cognition of the early history of the Earth, when geochemists received at their disposal instruments for conducting local precision quantitative analyses, which help to determine composition of different rare elements and also isotopes in specific points of ultrafine samples of rocks and minerals.

THE CHAOTIC EON

This epoch covers a period from the start of the Solar System formation out of a gaseous-dust protoplanetary disk ~ 4.568 bln years ago to the final formation of the Earth ~ 100-150 mln years later (Wood, 2011). The time limits of the chaotic eon and its main internal events have been established on the basis of data on short-lived isotopes and also geochemical characteristics of elements formed from the nebula, primogenitor of the Solar System.

The composition of the Solar nebula (like similar formations of new stars) is connected with the Big Bang due to which there emerged nuclear reactions, light elements from hydrogen to lithium at the expense of other chemical elements during a stellar nucleosynthesis (Khain, 2003; Lauretta, 2010). Data on them can be obtained from the Solar composition, whose 99 percent of mass correspond to the primary composition of the nebula. According to geochemical characteristics the elements are divided into several groups-refractory elements, including lithophyle*, have similarity with silicates and concentrate in the mantle and the crust. Siderophyle** elements accumulate in the core. As concerns volatile elements they pass into the atmosphere, and their part forming gas is "forced out" to the Solar System periphery.

A special type of meteorites, carbonic chondrites, correspond to the Earth's "primary" composition, and chemical elements in them are close to the Solar composition, excluding volatile elements (oxygen, carbon, nitrogen) (Wood, 2011; Lauretta, 2010). Among the short-lived isotopes noteworthy is hafnium (¹⁸²Hf) (lithophyle element) passing into tungsten (¹⁸²W) (siderophyle) with a half-life of 8.9 mln years. "Parental" radionuclide

*Lithophyle elements are a group of chemical elements (53 altogether), which form the main mass of minerals of the Earth crust, lithosphere and the Earth mantle.-Ed.

** Siderophyle elements are a group of 11 chemical elements including iron series, platinum metals and also molybdenum and rhenium. In terms of geochemical properties all of them are similar to iron.-Ed.

¹⁸²Hf decays finally in the system 50 mln years later. Of great significance for melting of minor bodies (meteorites, asteroids, and planet embryos) are ferrous radionuclides (⁶⁰Fe→) passing into nickel isotope (⁶⁰Ni) with a half-life of 1.5 mln years and the "parental" aluminum radionuclide (²⁶A1→) with its "daughter" element in decay being magnesium isotope (²⁶Mg) with a half-life period of 0.7 mln years.

Back to the beginning of evolution of the Solar System we should like to note that in the first 1.5-2 mln years owing to volatile elements there formed ice which served as a basic material for the largest planets Jupiter and Saturn. As to the regions where planets of the earth group formed they were poor in volatile elements. Two mln years later in this part of the Solar System there were forming small bodies not more than 10 km in diameter, which isolated from each other 20-100 thous. years later. Collisions brought them to destruction, but as a result a part of them "united" into larger bodies or planet embryos. Approximately 1 mln years later (i.e. 3 mln years after the start of the Solar System formation) there appeared bodies of about 1 percent of the Earth's mass. Since then gravitational energy promoting growth of the embryos of the earth group planets plays a decisive role. Mars and big asteroids formed about 7 mln years later, 63 percent of the Earth's mass formed 11 mln years later, and 93 percent of its mass 30 mln years after the start of the Solar System formation.

Fig. 4. Histogram of local ages of ²⁰⁷Pb/²⁰⁶Pb in sample RSES 96-15.10 from zircon debris in the Jack Hills outcrop subjected to an automatic analysis of lead isotope determination on the basis of SHRIMPI and SHRIMP-Pg (Holden et al., 2009). The ancient age covers a period from 4.34 to 4.1 bln years (second great peak). Thus, the Hadean eon took place 4.5-4.1 bln years ago, and the maximum formation of the Hadean crust - 4.25 bln years. The Hadean-Archaean period includes 7 events when recycling of the Hadean crust took place.

The Earth is a highly differentiated cosmic body with a metal (iron-nickel) core and the surrounding solid silicate shell, i.e. mantle. Differentiation of the planet substance started actually from the earliest moments of its formation, in the process of conception of embryo during collision and merging of small meteorite bodies. Due to short-lived radionuclides (²⁶A1, ⁶⁰Fe and later ¹⁸²Hf) emitting ample quantity of heat at decay, these bodies partly melted, and particles of melted iron were isolated, which could accumulate in the depths of emerging planets. Later at collision of larger bodies with an emerging planet, its silicate substance melted owing to collisions and thus small magmatic (lava) basins were forming. At a high temperature and pressure of 20-23 hPa established equilibrium between silicates and iron melt, they can separate (Wood, 2011). Such oceans (up to 400 km deep) can form during collisions with large asteroids. In case of the collision with a body of the size close to Mars, when the Moon formed, most likely there formed the magmatic ocean up to 700 km deep which covered the whole surface of the Earth. As shown on the model of the Earth core formation (Fig. 1) (Wood, 2011; Lauretta et al., 2011), the asteroids reaching the Earth delivered drops of iron enriched with siderophylic elements to the magmatic ocean. The drops of iron descended to the ocean bottom where a metal "reservoir" was located unstable in relation to the silicate mantle. In this reservoir large globoids formed equilibrium with the silicate substance located below, while the melted metal bodies descended deeper thus building up the core.

The time of the core formation can be computed using hafnium isotope (¹⁸²Hf) which produces, as was mentioned above, tungsten isotope (¹⁸²W). The half-life of ¹⁸²Hf → ¹⁸²W-8.9 mln years, and the whole ¹⁸²Hf disappears from the parental reservoir in 50 mln years. The major part of the core formed in 20 mln years from the start of the Solar System formation. The ratio of ¹⁸²W/¹⁸⁴W in the mantle and carbonic chondrites was equal at that time (Fig. 2). Upon formation of the major part of the core W, the siderophylic element passed into the core, and ¹⁸²Hf remained, as the lithophylic element, in the mantle. The mantle enrichment with ¹⁸²Hf is determined by a higher ratio of ¹⁸²W/¹⁸⁴W in the Earth's silicate mantle as compared with carbonic chondrites. 50 mln years later the whole of ¹⁸²Hf passed into ¹⁸²W, therefore consistency of Hf/W is observed both in the mantle and chondrite meteorites. It is the time of the final formation of the core.

The formation of the Earth's satellite, the Moon, which was important for the further development of the Earth, was a big event in the chaotic eon. Most of the researchers believe that the collision of the Earth with a large cosmic body whose mass was close to Mars caused the Moon's formation. Such "accident" took place 30-40 mln years after the Solar System had formed when it had the mass equal to 60-70 percent as compared with the modern one. By that time there already existed a substantial part of the Earth core, i.e. a process of its differentiation took place to a certain extent. In the course of the impact an ample quantity of the earth's silicate substance was thrown to the region of the Moon's orbit, which was located at that

Fig. 5. Standardized distributions of rare-earth elements in the Hadean zircons (A) (Maas et al., 1992); in lunar zircons (B) (Taylor et al., 2009).

time much closer to the Earth than today. The lunar rocks lack siderophylic elements, which proves that our planet did not split apart at the impact, but the collision brought only a considerable outburst of the silicate substance. The Moon has a small core, and its silicate part lacks isotope ¹⁸²W, concentrated in the Earth's core. Thus, the Moon greatly resembles the substance of the terrestrial mantle, i.e. debris of the substance of our planet were probably a basis for the Moon's formation. To draw a final conclusion on the sources of the lunar substance it is necessary to go on with studies of the nature of this celestial body comparing it with the terrestrial nature, and also identification of peculiarities of such collision (Wood, 2011).

We have insufficient knowledge yet about the time of separate events at that early development stage of the Solar System. Use of different isotopes can provide different dates of the final accretion (formation of the disk) of the Earth (Wood et al., 2006; Wood, 2011). But most of the researchers believe that 4.5 bln years ago the chaotic eon was over, and the Hadean stage started in the history of our planet.

THE HADEAN EON

The new epoch began after the Earth-Moon system had formed, which is confirmed by discovery of zircon or zirconium silicate (Zr[SiO₄]) of the same age on the both cosmic bodies. But identification of the Hadean eon by the geological scientific community as a separate epoch started after discovery of unusual minerals on a bed rock exposure at the Jack Hills outcrop in the west of Australia (Fig. 3). The zircon debris whose central part refers to the age of 4.376 bln years was found in sedimentary rocks located at the periphery of the Yilgarn craton*. It is more than by 500 mln years older of the most ancient rocks discovered on our planet earlier! The "deposit" of ancient

* Craton is a region of the continental terrestrial crust, which is not experiencing considerable folded deformations.-Ed.

zircons of about 4x4 m area is represented by metamor-phized sediments with debris of different size consolidated by a fine-grained basic mass. More than 100,000 grains (particles of 20-40 µm) of zircons are already recovered from the outcrop and studied. The ancient age is identified only in the central parts of grains, and the samples of the peripheral zone are much younger, their age is about 3.1 bln years. (In connection with the Australian discovery it should be noted that the most intensive studies of zircons started in the 21st century when it became possible to conduct local precision analyses of separate elements and isotopes, i.e. analyses in a specific point.)

The work of the Swedish scientist Peter Holden and his colleagues published in 2009 proves that the fastest growth of Hadean zircons, and consequently of the Hadean Earth crust, took place 4.25 bln years. The histogram (Fig. 4) submitted by the researchers shows a peak corresponding to the age of 4.1 bln years. These data suggest that the Hadean eon should have covered a period from 4.5 to 4.1-4.0 bln years ago. The time of the end of the epoch is 4.1 bln years ago, which is confirmed by regularities of distribution of hafnium isotopes in zircons of the Jack Hills outcrop (Blichert-Toft, Albarede, 2008). Recently new findings of ancient zircons were discovered close to this outcrop but no detailed research was undertaken as yet. Perhaps later on we shall receive new information on the history of our planet in the Hadean epoch.

Studies of distribution of rare elements in the Hadean zircons provides valuable information on the magma nature and genesis. The first such information was obtained by the German geochemist R. Maas and his colleagues in 1992, which was confirmed later by other specialists. The petrographic research proved that zircons included impurities of potassium feldspar, quartz, plagioclase, mona-zite and apatite. This implies a conclusion on a granite composition of zircons, i.e. melts typical of the terrestrial continental crust.

Fig. 6. Schematic chart of Hf isotopes (expressed by symbol ε_Hf) to the age (Nebel et al., 2014). Separate points illustrate results of zircon analyses taken from the single crust area. Median through such population can be used to obtain the ratio Lu/Hf required for isotope evolution. High values of Lu/Hf point to a mafic crust, while low values imply a felsitic crust. These features are determined by separation of Lu and Hf during partial melting of the primary mantle.

Identification of the oxygen isotope ratio (in the basic ¹⁶O and ¹⁸O) in the Hadean zircons helps to evaluate many conditions of the environment during formation of the first continental crust. The detailed research (Nebel et al., 2014) provided two important conclusions. The first implies that the weathering processes (changes of primary rocks) in the Hadean epoch were similar to the present ones. This suggests that the Earth's early atmosphere was damp and had a high potential of oxygen. The second conclusion: formation of zircons shows that partial melting of mantle rocks as a source of magma for the crust formation in the Hadean epoch took place in the conditions of closeness to the terrestrial surface.

Of great importance for understanding of tectonic conditions existing on our planet in the Hadean time is discovery of zircons on the Moon, whose samples were brought to the Earth by the members of the Apollo 14 (1971) and Apollo 17 (1972) expeditions. The age determination of the lunar zircons was carried out at the Australian National University where the minerals from the Jack Hills outcrop were also studied. The age of the lunar zircons is in the range of 4.0-4.35 bln years (Taylor et al., 2009), i.e. they are coeval with the Hadean zircons. And that is the end of their similarity. The lunar zircons formed at high temperatures of 920-1, 140°C while the Hadean zircons at mean temperatures of ~ 700°C. Their charts of content of standardized rare-earth elements (Fig. 5) also differ. The positive anomaly of cesium (Ce) is clearly demonstrated in the earth samples. As compared with other rare-earth elements it can have valency Ce⁴⁺, therefore in the conditions of oxidation it can easily be a part of zircons. These data confirm the oxidation conditions on the Earth in the Hadean eon, which is proved also by distribution of oxygen isotopes. The lunar analogs were formed in the reducing environment-they have no cesium anomaly.

Finally, the studies of the lunar zircons microstructure (Grande et al., 2013) revealed the presence of local areas of recrystallization, local amorphous areas, plastic deformations, ruptures and fissures, i.e. typical traces of the structures of impact genesis. In all appearance these crystals formed as a result of meteorite hits of the lunar surface. Such structures are not observed in zircons of the Hadean epoch found in Australia. As mentioned above they crystallized mostly from magmatic melt. But it can be assumed that the Moon and the Earth, as closely spaced bodies of the Solar System, were subjected to meteorite hits at the same time.

Lu-Hf isotope analyses (Fig. 6) are essential for solving the problem of the Hadean zircons origin. Lutecium during decomposition turns into hafnium (¹⁷⁶Lu→¹⁷⁶Hf). The latter can isomorphically be a part of zircon substituting zirconium. Zircons can contain up to several dozens of thousands of ppm of hafnium. The half-life of lutecium is 35.7±1.2 bln years, therefore it is clear that practically the whole of Hf "freezes" in zircons. Small age correlation is needed to obtain important geochemical information on the nature of primary melts and consequently also of rocks for the test zircons (Nebel et al., 2014).

At present specialists are studying samples received from different cratons such as the Canadian gneisses in Acasta, the Mississippi river sands where they have accumulated during destruction of the North American craton, from the ancient rocks of Greenland and the Archaean sediments of North China. All of them are connected with processing of the only source, i.e. Hadean zircons. This implies important conclusions. First, the Hadean crust was widespread

Fig. 7. Global compilations of zircon debris analyses (modernized Nebel-Jacobsen et al., 2010). These data prove that the Hadean crust existed not only under the craton in Western Australia, but also under other cratons, for example, in Canada, North America, Greenland and North China, i.e. practically under all Archaean cratons, which perhaps were parts of a single supercontinent at that time.

covering large regions of the Earth (Fig. 7). Secondly, perhaps the ancient cratons, known now on different continents, previously were located together and fell within the single continent (Glukhovsky et al., 1994, 2013). The latter was subjected to heavy meteorite hits which caused destruction of the first Hadean crust and its sinking to the mantle depth. Later on, as a result of geological events of the Archaean epoch, it could serve as a "primogenitor" of zircons and perhaps also of the rocks, which are a base of the Archaean formations, i.e. grey gneisses. Thus, the findings of ancient zircons with traces of processing of the first continental crust of the Earth are evidence of the early Archaean varied tectonic environment, whose nature still needs to be studied in detail.

Unfortunately no ancient zircons were found on the North Asian continent. It is evident that the Russian geologists should take part in such studies and be equipped with modern analytical instruments, which will enable them to carry out local high precision quantitative analyses.

PRELIMINARY SCENARIO

In conclusion we shall sum up our knowledge of the ancient history of our planet. As regards formation of the first continental crust it must be emphasized that it was constantly growing in a time span from 4.4-4.5 to 4.1-4.0 bln years ago or the end of the Hadean eon. The crust formation peak was 4.25 bln years ago. Perhaps 4.1-4.0 bln years were a turning point in the early geological history of the Earth after which the subsequent Archaean era started. Almost all Hadean zircons have a zonal (heterogeneous in composition and properties) structure, which suggests different geological events in the early history of our planet.

It was established basing on the analysis of an element composition of the Hadean zircons that their parent rocks were mainly of acid composition, most likely granites melted in cold (700-800 "C) conditions. These rocks were subjected to weathering close to the terrestrial surface. The further history of zircons points to burial of primary rocks deep in the mantle, which is indicative of geodynamic activities of the Earth in Hadean-early Archaean history. It is obvious that the main mechanism of activities was in the form of heavy meteorite bombardment which covered both the Earth and the Moon, where these processes are well recorded. These very reasons caused formation of large masses of basic rocks emerging from the mantle on the planet. Their differentiation is a cause of formation of small quantities of acid melts in which the Hadean zircons formed and are preserved up to now.

Of course, this is a preliminary scenario of the crust formation. It is obvious that only further studies of the ancient rocks and their minerals will enable us to understand the nature of geological processes of the Hadean eon, connection of the processes of the meteorite bombardment of the Moon with the fall of meteorites and asteroids to the already differentiated Earth.

These studies were supported by Integration Project No. 87 of the RAS Siberian Branch and grant of the Russian Fund of Fundamental Research No. 13-05-12026-ofi_m.

The author expresses gratitude to RAS Corresponding Member Vladislav Shatsky for assistance in drafting this paper.