Abstract

During the last decade, the role of garnet clinopyroxenites (and, more generally, basaltic compositions metamorphosed to high-density assemblages in the mantle) in basaltic petrogenesis has been the subject of several detailed investigations. Experimental studies and geochemical/petrological modeling suggest that pyroxenitic in sensu lato material is a potential source component in basaltic magmatism in general (e.g., Allégre & Turcotte, 1986; Meibom & Anderson, 2003; Anderson, 2005). For example, recycling of former basaltic material (in the form of garnet-clinopyroxenites, granulites, eclogites, websterites, olivine websterites) has been proposed to explain variability in MORB (Mid-Ocean Ridge Basalt) geochemistry (e.g., Hirschmann & Stolper, 1996; Eiler et al., 2000), as well as in HiMu-OIB (High Mu Oceanic Island Basalts, where Mu = μ = ²³⁸U/²⁰⁴Pb) lavas (e.g., Kogiso et al., 2003; Sobolev et al., 2005), both continental and oceanic EMI-like (Enriched Mantle type 1) basalts (e.g., Lustrino et al., 2000; Lustrino & Dallai, 2003; Escrig et al., 2004; Lustrino, 2005a, 2005b), subduction-related magmas (e.g., Schiano et al., 2000), intra-plate igneous rocks (e.g., Gao et al., 2004; Zandt et al., 2004) and continental flood basalts (CFBs; Cordery et al., 1997; Yaxley, 2000).

Here I show that the delamination and detachment of lower continental crust may be able to bring crustal lithologies down to mantle depths. The interaction of partial melts derived from these lithologies (metamorphosed to pyroxenitic/granulitic/eclogitic assemblages) with ambient mantle can explain some geochemical features of volumetrically insignificant though petrologically important magmas such as EMI-like basalts. Interaction of “normal mantle” (but what is “normal mantle”?) with lower-crust-derived partial melts has been proposed, amongst others suggestions, to explain the peculiar composition of the Plio-Quaternary volcanic rocks of Sardinia, Italy (Lustrino et al., 2000, 2005). In particular, the geochemical characteristics of Sardinian rocks (high SiO₂, low CaO and CaO/Al₂O₃, relatively high Ni, relatively low HFSE, low HREE, high Ba/Nb and La/Nb, slightly high ⁸⁷Sr/⁸⁶Sr, unradiogenic ¹⁴³Nd/¹⁴⁴Nd and ²⁰⁶Pb/²⁰⁴Pb ratios) have been linked to derivation from an orthopyroxene-rich mantle source. The origin of this enrichment in orthopyroxene is thought to be a consequence of reaction of the original lithospheric mantle with SiO₂-rich partial melts derived from delaminated and detached ancient lower continental crust.

Introduction

The geochemistry of igneous rocks has often been considered an infallible tool to explain the composition of Earth’s mantle. Major elements, trace elements and the isotope ratios of undifferentiated magmas (i.e. melts directly derived by partial melting of the mantle without modification related to fractional crystallization) reflect the composition of their sources. Moreover, the geochemical characteristics of primitive igneous rocks have often been used to infer the geodynamic setting of formation or, in a more speculative way, the thermal and rheological structure of the underlying mantle (e.g., the presence of mantle plumes). The composition of igneous rocks is the net effect of many complex factors including interconnection of liquids formed at various partial melting degrees at various depths in equilibrium with different restitic mineralogies, heterogeneous composition of mantle sources (i.e., presence of pyroxenitic slices in a peridotitic medium; e.g., Liu et al., 2005) and chromatographic reactions between rising partial melts and ambient mantle (e.g., Kogiso et al., 2004).

The result is that the geochemical message carried by primitive igneous rocks is not easily decipherable because of the many variables required to model the system from the formation of the first drop of melt to the solidification of the last liquid. In this sense, proposing the existence of huge structures in the Earth’s mantle (e.g., mantle plumes with roots anchored in the deep mantle) only on the basis of geochemical argumentations is dangerous. However, it should also be noted that petrology cannot exclude the existence of such thermal anomalies on the basis of geochemical data alone.

Geochemical constraints

During the last two decades some basic geochemical and petrological concepts have been reconsidered. For example, the concept that the source of MORB is homogeneous and depleted contrasts with new data (e.g., Hofmann, 2003; Workman & Hart, 2005) that are reconcilable only with heterogeneous sources (i.e., pyroxenitic slices dispersed in a depleted matrix; Allegre & Turcotte, 1986; Meibom & Anderson, 2003; Anderson, 2005). Also other types of magmatism (e.g., related to OIB) have revealed a very wide range of chemical compositions. Several acronyms have become part of geochemical jargon (DMM, HiMu, EMI, EMII, FoZo, PHeM, C, PreMa and so on) implying the existence of geochemical mantle end-members not immediately associable with physical entities (i.e., volumes in the mantle; Zindler & Hart, 1986; Hofmann, 1997, 2003, Lustrino, 2001; Lustrino & Dallai, 2003; Workman et al., 2004; Stracke et al., 2005). For those not skilled in igneous geochemistry these acronyms may lead to confusion. To clarify matters, it may be said that most of such acronyms have been invoked to explain the most extreme isotopic compositions (mostly based on Sr-Nd-Pb isotope systematics, plus a few constraints based on He-Os-Hf-O isotopes) of basalts emplaced in oceanic settings (i.e., away from places where the continental lithosphere can act as potential contaminant). Trace- and major elements are much more variable (e.g., Lustrino & Dallai, 2003) than originally supposed (e.g., Weaver, 1991) and large overlaps may exist between single end-members. To complicate the situation still further, the isotopic composition of some end-members as proposed in literature is continuously changing (e.g., the isotopic definition of FoZo has changed four times during the last 13 years; Stracke et al., 2005) and in some cases it is impossible to propose a fixed isotopic composition (e.g., the ²⁰⁷Pb/²⁰⁴Pb isotopic ratios of EMI; Mahoney et al., 1996; Lustrino & Dallai, 2003).

What may be said with certainty is that virtually all the geochemical end-members (as revealed by the composition of primitive or only slightly differentiated basalts) can be explained by an origin from mantle sources that have experienced interaction with crustal material introduced at depth. Another key point is that geochemistry cannot be considered as a proof for or against the existence of mantle plumes coming from deep reservoirs (e.g., Lustrino, 2005a). Detailed study of the geochemical characteristics of primitive igneous rocks has inspired several petrological models that can explain isotopic and trace element characteristics without invoking deep and isolated reservoirs (e.g., Foulger et al., 2005 and references therein; Lustrino, 2001, 2005a; Anderson, 2005; Stracke et al., 2005).

Mantle plumes

Recently the classical model of a mantle plume, considered to be a thermal anomaly coming from a deep reservoir, has been subject to strong criticism (e.g., Sheth, 1999; Lustrino, 2001, 2005; Foulger et al., 2005, and references therein). The concept of the mantle plume has been considered suitable to explain the production of the huge volumes of magma associated with CFBs and Large Igneous Provinces (LIPs). Mantle plumes have also been proposed to explain magmatic activity away from active subduction systems and passive margins. At the heart of the concept of the mantle plume is that it is associated with a thermal anomaly coming from the deep mantle.

The fundamental objection to the classical hypothesis of the mantle plume, proposed to explain both high- and low-volume magmatism in both in continental and oceanic settings, is that the observed anomaly is chemical rather than thermal. High melt productivity from mantle sources can be obtained at “normal” temperatures if basaltic material (metamorphosed to high-density assemblages), or partial melt from basaltic material, is interbedded in “normal” peridotite (e.g., Cordery et al., 1997; Yaxley, 2000; Meibom & Anderson, 2003; Anderson, 2005). Basaltic material (metamorphosed to garnet-clinopyroxenitic, granulitic or eclogitic assemblages, depending on the starting composition) is characterized by solidus temperatures about 100°C lower than peridotite at the same pressure.

Most of the basaltic material in the mantle is probably introduced at subduction zones. However, another mechanism able to transport shallow basaltic material to greater depth in the mantle is the delamination and detachment of lower continental crust (see Lustrino, 2005a, for a detailed discussion).

The lower crust and mantle portions of continental lithosphere exert strong structural and chemical controls on basaltic magmatism. Apart from the obvious controls of subduction on magmatic geochemical budgets and the role of buoyancy in trapping basaltic magmas, I advance an alternative model that a combination of lower crust and lithospheric mantle controls the composition of both continental and oceanic intra-plate magmatism, and magmatism associated with continental break-up, particularly basalts with EMI affinity.

I propose a model that posits a role for lower-crustal delamination and detachment in the upper-mantle geochemical budget. This model can explain the geochemical signature of volumetrically insignificant, but petrologically important, EMI-like magmas, and is an alternative to crustal recycling in subduction zones.

Lower crustal delamination and detachment

The model is based on the development of gravitative instability following continent-continent collision along mobile belts (Figure 1). Continent-continent collisions force lower crustal rocks (with their original basaltic/gabbroic composition) to higher pressures and temperatures. Under such conditions, the lithospheric keel becomes gravitationally unstable and may detach, sinking into the asthenospheric mantle [Ed: see also other Lithospheric thinning pages]. Metamorphic reactions occurring in the lower continental crust as a consequence of continent-continent collision can lead to density increases, with densities rising to up to 3.8 g/cm³ with the appearance of garnet in the metamorphic assemblage (basalt – garnet – clinopyroxenite/eclogite). This may lead to gravitative instability of the overthickened lithospheric keel (the lower crust + lithospheric mantle) which may detach from the uppermost lithosphere and sink into the upper mantle. The volume formerly occupied by the lithospheric keel is replaced by asthenosphere melts (transformed into new lower crust) and asthenospheric restite (which is transformed to lithospheric mantle, i.e., the mechanical boundary layer, by cooling from above). The lower crust, sinking to great depths, may partially melt, yielding liquids of dacitic/rhyolitic composition that react with the newly formed lithospheric mantle, forming orthopyroxene-rich layers. After freezing, such metasomes may also be reactivated several million years after lithospheric delamination occurred (Lustrino, 2005a). The metasomatized lithosphere may thereby acquire strong crustal geochemical imprints as commonly observed in CFB (Continental Flood Basalts; e.g., Peate et al., 1999), as well as in continental (e.g., Lustrino et al., 2000, 2005) and oceanic (e.g., Mahoney et al., 1996; Borisova et al., 2001; Frey et al., 2002) intra-plate basalts, and, rarely, mid-ocean ridge magmas (e.g., Kamenetsky et al., 2001).

Figure 1: (A) Initial situation with thin oceanic lithosphere between thick continental plates. The continental lithosphere is divided according to rheological behavior: 1) brittle SiO2-rich upper continental crust; 2) ductile mafic lower crust; 3) lithospheric mantle, reaching an average depth of about 80 km, the crustal portion accounting for less than half. (B) Under compressive stress, the oceanic lithosphere (colder and denser than continental lithosphere) is subducted beneath one of the two continental plates, with relatively little deformation of the lower continental crust. (C) Continent-continent collision. The oceanic lithosphere has been completely subducted. (D) The upper continental crust is piled up and thrusted, leading to thickening of the entire lithosphere, including the lower continental crust. Depending on the extent of pressure increase, the original lower crustal basaltic/gabbroic paragenesis may be transformed to amphibolite (P ≤1 GPa) or to granulite/eclogite/garnet clinopyroxenite assemblages at higher pressures (~2-3 GPa; e.g., Wolf & Wyllie, 1994; Rapp & Watson, 1995; Jull & Kelemen, 2001). The net effect is a density increase to up to 3.5 g/cm3 (e.g., Wolf & Wyllie, 1993). These metamorphic reactions are not strictly isochemical. During the formation of new phases, Rb and U are preferentially concentrated in the melt compared to Sr and Pb, respectively, while Sm and Nd are not strongly fractionated. Restites are therefore characterized by low to very low Rb/Sr and U/Pb and relatively unchanged (low) Sm/Nd ratios. The restite eclogite/garnet-clinopyroxenite thus evolves with low time-integrated 87Sr/86Sr, 204Pb/206Pb and 143Nd/144Nd. (E) The density increase leads to gravitative instability of the overthickened lithospheric keel. In particular, dense garnet-rich lower-crustal restite causes detachment of the lithospheric mantle from the higher lithosphere levels and it sinks into the asthenosphere. The high densities of average lower crust contrast significantly with those of the lithospheric mantle and asthenosphere, and there is a strong tendency for the overthickened lithospheric keel to sink. (F) Detail of (E). During sinking, the lower crust is likely to undergo partial melting producing liquids of TTG (Tonalitic, Trondhjemitic, Granitic) and adakitic affinity (e.g., Springer & Seck, 1997; Defant & Kepezhinskas, 2001; Zegers & van Keken, 2001; Xu et al., 2002). During lithospheric detachment, new asthenospheric mantle replaces the region vacated by delaminated lithospheric mantle and lower crust. According to this model, the asthenosphere accretes to the remaining lithosphere and becomes transformed to lithosphere on cooling. Isostatic uplift and formation of intermontane troughs accompany delamination as a consequence of upward impingement of hot, buoyant asthenosphere. Partial melts of the asthenospheric megalith underplate the remaining lithosphere to form the new lower crust and basaltic magmatism at surface. Lithospheric mantle formed from former asthenosphere is metasomatized by SiO2-rich melts derived from the coeval sinking of lower crust. Such highly reactive melts would form orthopyroxene-rich zones, still peridotitic in composition, and will therefore be able to yield SiO2-undersaturated melts at relatively high pressures (for more details see Yaxley, 2000). (G) After delamination of the lower crust and lithospheric mantle, asthenospheric counter flow, contemporaneous partial melting of the lower crust and mechanical accretion of the asthenosphere to the remaining lithosphere, the new lithospheric mantle comprises variably depleted peridotite, heterogeneously metasomatized, with orthopyroxene-rich (lherzolite, olivine-pyroxenite, websterite) lithologies. This mantle source may then remain unsampled for several m.y. During this period, lithospheric mantle metasomes evolve with peculiar crustal isotopic features, i.e.: 1) elemental Sr originally present in plagioclase is transferred to metasomatic melts during lower-crustal partial melting; 2) the presence of residual garnet in sinking lower crust produces partial melts with strongly fractionated LREE/HREE evolving with very low 143Nd/144Nd isotopic ratios, and; 3) low m (m = 238U/204Pb) crustal partial melts evolve with low 206Pb/204Pb isotopic ratios. (H) Metasomatized lithospheric mantle may be reactivated several m.y. after lower crustal delamination occurred. Partial melts of such regions are likely to inherit lower-crust-like metasomatic attributes, characterized by typical EMI-like geochemical features (e.g., low radiogenic Pb ratios (206Pb/204Pb< 17), slightly radiogenic Sr isotopes (87Sr/86Sr ~0.706) unradiogenic Nd (143Nd/144Nd ~0.5121), unradiogenic Hf (176Hf/177Hf ~0.2826), slightly radiogenic Os (187Os/188Os ~0.135-0.145); Lustrino & Dallai, 2003). Relative mantle-normalized Ba, Pb, Eu or Sr anomalies and variation of Ba/Nb ratios (3.5-47.4), Ce/Pb (1.2-24.6), Nb/U (10.5-71.8), Sr/Nd (6.2-36.4) and Eu/Eu* (0.83-1.25) ratios in EMI-type basalts (Lustrino & Dallai, 2003) reflect the effects of 1) a lower-crust starting composition, 2) metamorphic paragenesis and PT parameters conditioning lower-crust partial melting, 3) metasomatic reactions between SiO2-rich melts and peridotite, 4) a cratonization style of asthenosphere, 5) partial melting of newly-accreted lithospheric mantle, and 6) fractional crystallization (coupled to potential crustal assimilation) of lithospheric melts.

Summary

Sources in the lowermost mantle and anomalously high temperatures are not necessary to explain the geochemistry of OIB and CFB magmas or to achieve the requisite large melt fractions. An alternative model is suggested by the presence of lower crustal metasomatic signatures and related isotopic growth prior to partial melting. Crustal material (either continental lower crust or subducted oceanic slab) stored in the upper mantle or the mantle transition zone is predisposed to melting at ambient mantle temperatures.

Both the crustal and ultramafic parts of the lithosphere may sink when cold, but become buoyant when heated to ambient mantle temperatures. The sinking of subducted lithospheric mantle may be arrested at the 660 km discontinuity, where crustal components begin to melt. Because of their buoyancy, these melts trigger plume-like upwellings without the need for lower mantle convective upwelling. The presence of lithospheric material in the upper mantle can also explain the isotopic features of both CFB and their oceanic counterparts.

In conclusion, the role of sinking lower crust and lithospheric mantle (and of basaltic material recycled back into the mantle in general) is critical to explaining the geochemical attributes in CFB, OIB and LIP magmas. Delamination and detachment of lower crust and lithospheric mantle are supported by geophysical, geological, geochemical and petrological considerations.

Acknowlegments

I warmly thank Gill Foulger for her enthusiastic approach to the problem of the chemical composition and the structure of upper mantle. Her energy has been a continuous stimulus for my recent studies.

References

• Allegre J.-C., Turcotte D.L., 1986. Implications of a two-component marble-cake mantle. Nature, 323, 123-127.
• Anderson D.L., 2005. The layered mantle revisited. An eclogite reservoir. http://www.mantleplumes.org/Eclogite.html
• Borisova A.Yu., Belyatsky B.V., Portnyagin M.V., Sushchevskaya N.M., 2001. Petrogenesis of olivine-phyric basalts from the Aphanasey Nikitin Rise: evidence for contamination by cratonic lower continental crust. J. Petrol., 42, 277-319.
• Cordery M.J., Davies G.F., Campbell I.H., 1997. Genesis of flood basalts from eclogite-bearing mantle plumes. J. Geophys. Res., 102, 20179-20197.
• Defant M.J., Kepezhinskas P., 2001. Evidences suggests slab melting in arc magmas. Eos, Trans. Am. Geophys. Un., 82, 65-69.
• Eiler J.M., Schiano P., Kitchen N. Stolper E.M., 2000. Oxygen-isotope evidence for recycled crust in the sources of mid-ocean-ridge basalts. Nature, 403, 530-534.
• Escrig S., Capmas F., Dupré B., Allégre C.J., 2004. Osmium isotopic constraints on the nature of the DUPAL anomaly from Indian mid-ocean-ridge basalts. Nature, 43, 59-63.
• Foulger G.R., Natland J.H., Presnall D.C., Anderson D.L. (Eds.), 2005. Plates, plumes & paradigms. Geol. Soc. Am. Special paper, 388 (in press).
• Frey F.A., Weis D., Borisova A.Yu., Xu G., 2002. Involvement of continental crust in the formation of the Cretaceous Kerguelen Plateau: new perspectives from ODP Leg 120 sites. J. Petrol., 43, 1207-1239.
• Gao S., Rudnick R.L., Yuan H.-L., Liu X.-M., Liu Y.-S., Xu W.-L., Ling W.-L., Ayers J., Wang X.-C., Wang Q.-H., 2004. Recycling lower continental crust in the north China Craton. Nature, 432, 892-897.
• Hirschmann M.M., Stolper E.M., 1996. A possible role for garnet pyroxenite in the origin of the "garnet signature" in MORB. Contrib. Mineral. Petrol., 124, 185-208.
• Hofmann A.W., 1997. Mantle geochemistry: the message from oceanic volcanism. Nature, 385, 219-229.
• Hofmann A.W., 2003. Sampling mantle heterogeneity trough oceanic basalts: isotopes and trace elements. In: R.W. Carlson, H.D. Holland and K.K. Turekian (Ed.) Treatise on geochemistry Vol. 2: The mantle and core, 61-101.
• Kamenetsky V.S., Maas R., Sushchevskaya N.M., Norman M.D., Cartwright I., Peyve A.A., 2001. Remnants of Gondwanan continental lithosphere in oceanic upper mantle: evidence from the South Atlantic Ridge. Geology, 29, 243-246.
• Kogiso T., Hirschmann M.M., Frost D.J., 2003. High-pressure partial melting of garnet pyroxenite: possible mafic lithologies in the source of ocean island basalts. Earth Planet. Sci. Lett., 216, 603-617.
• Kogiso T., Hirschmann M.M., Petermann M., 2004. High-pressure partial melting of mafic lithologies in the mantle. J. Petrol., 45, 2407-2422.
• Jull M., Kelemen P.B., 2001. On the conditions for lower crustal convective instability. J. Geophys. Res., 106, 6423-6446.
• Liu Y., Gao S., Aeolus Lee C.-T., Hu S., Liu X., Yan H., 2005. Melt-peridotite interactions: links between garnet pyroxenite and high-Mg# signature of continental crust. Earth Planet. Sci. Lett., 234, 39-57.
• Lustrino M., 2001. Debated topics of modern igneous petrology. Per. Mineral., 70, 1-26.
• Lustrino M., 2005a. How the delamination and detachment of lower crust can influence basaltic magmatism. Earth-Sci. Rev. (in press).
• Lustrino M., 2005b. Pyroxenites everywhere. Comment on “High-pressure melting experiments on garnet clinopyroxenite and the alkalic to tholeiitic transition in ocean-island basalts” by Keshav et al. [Earth Planet. Sci. Lett. 223 (2004) 365-379]. Submitted to Earth Planet. Sci. Lett.
• Lustrino M., Dallai L., 2003. On the origin of EM-I end-member. N. Jahrb. Miner. Abh., 179, 85-100.
• Lustrino M., Melluso L., Morra V., 2000. The role of lower continental crust and lithospheric mantle in the genesis of Plio-Pleistocene volcanic rocks from Sardinia (Italy). Earth Planet. Sci. Lett., 180, 259-270.
• Lustrino M., Melluso L., Morra V., 2005. The uniqueness of Plio-Quaternary volcanic rocks of Sardinia in the circum-Mediterranean Cenozoic Igneous Province. Submitted to Geol. Soc. Am. Spec. Paper, L. Beccaluva, G. Bianchini and M. Wilson (Eds.).
• Mahoney J.J., White W.M., Upton B.G.J., Neal C.R., Scrutton R.A., 1996. Beyond EM-1: lavas from Afanasy-Nikitin Rise and the Crozet Archipelago, Indian Ocean. Geology, 24, 615-618.
• Meibom A., Anderson D.L., 2003. The statistical upper mantle assemblage. Earth Planet. Sci. Lett., 217, 123-139.
• Rapp R.P., Watson E.B., 1995. Dehydration melting of metabasalt at 8-32 kbar: implications for continental growth and crust-mantle recycling. J. Petrol., 36, 891-931.
• Sheth H.C., 1999. Flood basalts and large igneous provinces from deep mantle plumes: fact, fiction, and fallacy. Tectonophysics, 311, 1-29.
• Sobolev A.V., Hofmann A.W., Sobolev S.V., Nikogosian I.K., 2005. An olivine-free mantle source of Hawaiian shield basalts. Nature, 434, 590-597.
• Springer W., Seck H.A., 1997. Partial fusion of basic granulites at 5 to 15 kbar: implications for the origin of TTG magmas. Contrib. Mineral. Petrol., 127, 30-45 .
• Stracke A., Hofmann A.W., Hart S.R., 2005. FOZO, HIMU, and the rest of the mantle zoo. Geochem. Geophys. Geosyst., 6, doi:10.1029/2004gc000824.
• Weaver B.L., 1991. Trace element evidence for the origin of ocean-island basalts. Geology, 19, 123-126.
• Wolf M.B., Wyllie P.J., 1993. Garnet growth during amphibolite anatexis: implications of a garnetiferous restite. J. Geol., 101, 357-373.
• Wolf M.B., Wyllie P.J., 1994. Dehydration-melting of amphibolite at 10 kbar: the effects of temperature and time. Contrib. Mineral. Petrol., 115, 369-383.
• Workman R.K., Hart S.R., 2005. Major and trace element composition of the depleted MORB mantle (DMM). Earth Planet. Sci. Lett., 231, 53-72.
• Workman R.K., Hart S.R., Jackson M., Regelous M., Farley K.A., Blusztajn J., Kurz M., Staudigel H., 2004. Recycled metasomatized lithosphere as the origin of the Enriched Mantle II (EM2) end-member: evidence from the Samoa volcanic chain. Geochem. Geophys. Geosyst., doi:10.1029/2003gc000623.
• Xu J.-F., Shinjo R., Defant M.J., Wang Q., Rapp R.P., 2002. Origin of Mesozoic adakitic intrusive rocks in the Ningzhen area of East China: partial melting of delaminated lower continental crust? Geology, 30, 1111-1114.
• Yaxley G.M., 2000. Experimental study of the phase and melting relations of homogeneous basalt + peridotite mixtures and implications for the petrogenesis of flood basalts. Contrib. Mineral. Petrol., 139, 326-338.
• Zandt G., Gilbert H., Owens T.J., Ducea M., Saleeby J., Jones C.H., 2004. Active foundering of a continental arc root beneath the southern Sierra Nevada in California. Nature, 431, 41-46.
• Zegers T.E., van Keken P.E., 2001. Middle Archean continent formation by crustal delamination. Geology, 29, 1083-1086.
• Zindler A., Hart S., 1986. Chemical Geodynamics. Ann. Rev. Earth Planet. Sci., 14, 493-571.