1 Preservation of Earth-forming events in the W isotopic composition of modern flood basalts 1 2 Authors: Hanika Rizo1*, Richard J. Walker2, Richard W. Carlson3, Mary F. Horan3, Sujoy 3 Mukhopadhyay4, Vicky Manthos4, Don Francis5, Matthew G. Jackson6 4 5 Affiliations: 6 1Geotop, Departement des sciences de la Terre et de l’atmosphère, UQAM, Montreal, Canada 7 2Department of Geology, University of Maryland, College Park, MD, USA 3Carnegie Institution 8 for Science, Washington DC, USA 4Department of Earth and Planetary Sciences, UC Davis, 9 Davis, CA, USA 5Earth and Planetary Sciences department, McGill University, Montreal, Canada 10 6 Department of Earth Science, UC Santa Barbara, Santa Barbara, CA, USA 11 12 *Corresponding author. E-mail: rizo.hanika@uqam.ca 13 14 Abstract: How much of Earth's compositional variation dates to processes occurring during 15 planet formation remains an unanswered question. High precision W isotopic data for rocks from 16 two large igneous provinces, the North Atlantic Igneous Province and the Ontong Java Plateau, 17 reveal preservation to the Phanerozoic of W isotopic heterogeneities in the mantle. These 18 heterogeneities, caused by the decay of 182Hf in high Hf/W ratio mantle domains, were created 19 during the first ~50 Ma of Solar System history, and imply that portions of the mantle that 20 formed during Earth’s primary accretionary period have survived to the present. 21 22 2 23 Main text: 24 Four and a half billion years of geologic activity has overprinted much of the evidence for 25 processes involved in Earth’s formation and initial chemical differentiation. High-precision 26 isotopic measurements, however, now allow the use of the variety of short-lived radionuclides 27 that were present when Earth formed to provide a clearer view of events occurring during the first 28 tens to hundreds of million years (Ma) of Earth history. Evidence from both the 146Sm-142Nd (t1/2 29 = 103 Ma) and 129I-129Xe (t1/2 = 15.7 Ma) systems show the importance of early mantle 30 differentiation and outgassing events, but provide conflicting evidence on the preservation of 31 early-formed mantle reservoirs to the present day (1-4). Of these short-lived systems, the 182Hf-32 182W (t1/2 = 8.9 Ma) system is uniquely sensitive to metal-silicate separation, and has been used 33 effectively to trace the timing and processes of core formation (e.g. 5), arguably the most 34 important chemical differentiation event to occur on a rocky planet. Only recently, however, have 35 measurement techniques improved to the point of resolving 182W/184W variability in ancient (>2.7 36 Ga) terrestrial rocks, that reflects preservation of compositionally distinct domains in Earth’s 37 interior that were likely created during Earth formation (6-10). Young mantle-derived rocks 38 examined to date have shown neither 142Nd nor 182W isotopic heterogeneity, suggesting that the 39 early-formed compositional domains in Earth’s interior were largely destroyed by mantle mixing 40 processes acting during the first half of Earth history (1-4, 6-10). Here we report 182W/184W ratios 41 in Phanerozoic flood basalts from Baffin Bay and the Ontong Java Plateau, some of which range 42 to the highest ever measured in terrestrial rocks. These results document the preservation of 43 regions within Earth’s interior whose compositions were established by events occurring within 44 the first ~ 50 Ma of Solar System history. This study, consequently, provides new insights into 45 3 the processes at work during planet formation, the chemical structure of Earth’s interior, and the 46 interior dynamics that allowed preservation of chemical heterogeneities for 4.5 billion years. 47 48 Flood basalts are the largest volcanic eruptions identified in the geological record. These types of 49 eruptions created both the North Atlantic Igneous Province, which hosts the Baffin Bay locale 50 (11), and the Ontong Java Plateau, western Pacific Ocean (12). We have studied pillow lavas 51 with high MgO, picritic compositions (Tab. S1) from Padloping Island, Baffin Bay (Pi-23 and 52 Pd-2). We targeted these rocks because some Baffin Bay lavas contain the highest 3He/4He ratios 53 ever measured (e.g. 13), along with Pb isotopic compositions (14) and D/H ratios (15) that 54 indicate that their mantle source was relatively primitive and undegassed, consistent with 55 isolation since shortly after Earth formation. Ontong Java is Earth’s largest known volcanic 56 province, and shares chemical and isotopic similarities with the Baffin Bay lavas, consistent with 57 a primitive mantle source (16). The Ontong Java sample (192-1187A-009R-04R) is a basalt 58 (Table S1) collected from the plateau’s eastern flank by Ocean Drilling Project Leg 192. 59 60 We present data from the short-lived 182Hf-182W and 146Sm-142Nd systems because these two 61 systems are variably sensitive to the core formation and mantle differentiation processes that 62 occurred early in Earth history. We compare these data with data from the long-lived U-Th-He, 63 147Sm-143Nd and 187Re-187Os isotope systems, together with W and highly siderophile element 64 (HSE: Re, Os, Ir, Ru, Pt, Pd) concentrations, to better discern early differentiation events from 65 those occurring over all of Earth history. Glassy rim and core pieces of sample Pi-23 (Pi-23a and 66 Pi-23b, respectively), a bulk sample of Pd-2, and a bulk sample of 192-1187A-009R-04R are 67 characterized by high 182W/184W ratios that are well resolved from standards, with μ182W values 68 4 ranging from +10 to +48 (where μ182W = [(182W/184W)sample/(182W/184W)standard -1]x106) (Fig. 1 69 and Table 1, ref. 17). The μ182W values for sample Pi-23a and Pi-23b are in good agreement. This 70 rules out the role of stable W isotope fractionation through interaction of seawater with the pillow 71 rim in creating the measured 182W values (17). Samples 192-1187A-009R-04R and Pd-2 are 72 characterized by the lowest W concentrations and the highest μ182W values (Table 1). The W 73 concentrations of these two samples (23 and 26 ppb, respectively) are broadly consistent with 74 magmas derived by 15-20% partial melting of a mantle source with ~5 ppb W, consistent with a 75 primitive source free of W-rich recycled crust (17, Fig. S2). The geological reference materials 76 VE-32 (mid-ocean ridge glass) and BHVO-1 (Hawaiian basalt) were measured at the same time 77 as the Baffin Bay and Ontong Java samples, and yielded μ182W values of -0.8 ± 4.5 and -2.3 ± 78 7.7, respectively (Table 1). These μ182W values are indistinguishable from the terrestrial Alfa 79 Aesar W standard (μ182W = 0) and other modern rocks (6-10). The 3He/4He ratio measured in 80 olivines of the Baffin Bay samples (Table S3, 17) yielded values up to 48.4 RA (RA being the 81 3He/4He ratio normalized to the atmospheric ratio of 1.39×10−6) which are in agreement of 82 previous findings (13) and indicate that the source of these lavas is relatively undegassed, and 83 possibly isolated since Earth formation. The Baffin Bay and Ontong Java Plateau samples have 84 HSE abundances and initial 187Os/188Os ratios (providing a record of long-term Re/Os ratio) that 85 are indistiguishable from other modern mantle-derived lavas with similar MgO abundances that 86 do not show elevated μ182W (Figure 2, Table S4, 18). 87 88 Variability in 182W/184W ratios reflects Hf/W fractionation while 182Hf was extant. Fractionations 89 in Hf/W are observed in early Solar System materials (e.g. 5), so variable tungsten isotopic 90 compostions in terrestrial samples can reflect the imperfect mixing of late additions of such 91 5 materials (6, 9). The μ182W value of +48 for Baffin Bay sample Pd-2, however, is larger than can 92 be accounted for by this process, and so this possibility is discounted (see Supplementary Text). 93 Fractionation of Hf/W can also have occurred as the result of endogenous Earth differentiation 94 processes such as magma ocean crystallization (7) and core formation (9). Silicate fractionation 95 processes, however, cannot be responsible for the generation of the anomalous 182W in the 96 sources of Baffin Bay and Ontong Java lavas. If the high μ182W was due to silicate fractionation 97 in a magma ocean while 182Hf was extant, then μ 182W should positively correlate with μ142Nd, the 98 decay product of the short-lived 146Sm (t½ = 103 Ma) isotope system. Instead, the μ142Nd values of 99 the samples are indistinguishable from all other modern basalts so far measured (Fig. S3, Table 100 S5, 17). 101 102 This leaves fractionation of Hf/W as a result of metal-silicate segregation accompanying core 103 formation as the probable cause of the observed anomalies in the Phanerozoic samples. Metal-104 silicate segregation is the most effective process capable of fractionating Hf/W ratios, because Hf 105 is a strongly lithophile trace element, while W is moderately siderophile. The low W 106 concentrations estimated for the mantle source of the flood basalts studied here are consistent 107 with mantle domains that experienced metal-silicate segregation (Table 1, Supplementary Text). 108 Repeated metal-silicate segregation events during planet formation (19) could create one or more 109 mantle domains with distinct μ182W without affecting the Sm-Nd system. Such events would 110 result in variable μ182W due to different times of the metal-silicate segregation events (Figure 3), 111 or different Hf/W ratios in the resulting mantle reservoirs that reflect differing oxidation states, 112 and hence, differing W partitioning into metal (Supplementary Text). The key observation, which 113 is also seen in the results reported here, is that terrestrial samples with 182W excesses do not seem 114 6 to derive from sources depleted in HSE (Figure 2, 6-10). Highly siderophile elements have 115 partition coefficients between metal and silicate of >104 (20), thus, their concentrations in metal-116 depleted mantle domains are expected to be very low. Evolving oxidation states during Earth 117 accretion might explain the decoupling of 182W and HSE, because while W becomes less 118 siderophile under oxidizing conditions, the HSE, even at high oxidation states, are not soluble in 119 silicates (20). This type of model, however, requires subsequent late accreted HSE to be mixed 120 into different mantle domains without the mixing away of tungsten isotopic heterogeneity. 121 Alternatively, the observed decoupling could be explained if some metal from the core of the 122 Moon-forming giant impactor was retained in the mantle, followed by a minor amount of late 123 accretion (9). This type of model requires that a substantial mass of high density metal be 124 retained in the Earth' mantle following the impact, when the mantle was partially or wholly 125 molten, and that the retained metal contained chondritic relative abundances of the HSE. These 126 models and others have been presented to try to explain the apparent decoupling of W isotopic 127 compositions and HSE abundance variations. These models are summarized in the 128 Supplementary Text along with a few additional suggestions. 129 130 Regardless of the origin of the 182W variability, arguably more surprising than the fact that Earth 131 experienced such early differentiation events, is that reservoirs formed by these early processes 132 remain in the mantle today. This conclusion is now supported by data from both W and Xe (1, 133 22) isotopic variability, but not 142Nd, where the evidence suggests that the observed 134 heterogeneity in 142Nd/144Nd ratio was reduced to an unobservable level by the end of the 135 Archean, likely though the mixing caused by mantle convection. Perhaps a key to reconcile these 136 observations is that the 129I-129Xe system primarily reflects mantle outgassing, and the 182Hf-182W 137 7 system metal-silicate separation, whereas the 146Sm-142Nd system is controlled by internal mantle 138 differentiation. For both W and Xe, one component of the complementary chemical 139 differentiation, the core for W, and the atmosphere for Xe, may not be available for effective 140 recycling and mixing in the mantle (23). By contrast, for Nd, the main reservoirs created during 141 early Earth differentiation may have been in a portion of the mantle that has been effectively 142 mixed by mantle circulation. Estimates for how much of the mantle can remain unmixed depend 143 on the rheological properties assigned to the various materials involved. Some models (e.g. 24) 144 show that as much as 20% of the mantle may remain isolated as distributed masses in the mantle. 145 An important aspect of the results presented here is that both 182W anomalies and elevated 146 3He/4He (Table S3) appear in at least two major flood basalts. These events produce huge 147 volumes of magma that must be derived by melting large volumes of mantle during unusual 148 thermal events in the history of mantle circulation. The large size and sporadic nature of flood 149 basalts is perhaps more indicative of a layer in the mantle that has an appropriate density or 150 rheological properties to keep it from effectively mixing with the rest of the mantle. One 151 candidate for such a reservoir is the large low seismic shear velocity provinces (LLSVP) imaged 152 at the base of the mantle (25). These regions appear to be warmer and compositionally different 153 from surrounding mantle. Estimates of their volume range to as high as 7% of the mantle, or of 154 order 6 x 1010 km3. If the LLSVP are remnants of early differentiation events on Earth, they must 155 have a delicately balanced density contrast compared to the surrounding mantle to allow their 156 survival through 4.5 billion years of dynamic Earth history. 157 158 References and notes: 159 8 1. S. Mukhopadhyay, Early differentiation and volatile accretion recorded in deep-mantle neon 160 and xenon. Nature 486, 101-104 (2012). 161 2. M. Boyet, M. O. Garcia, R. Pik, F. Albarede, A search for 142Nd evidence of primordial mantle 162 heterogeneities in plume basalts. Geophys. Res. Lett. 32, no. 4 (2005). 163 3. V.C. Bennett, A.D. Brandon, A.P. Nutman, Coupled 142Nd-143Nd isotopic evidence for Hadean 164 mantle dynamics, Science 318, 1907-1910 (2007). 165 4. H. Rizo, M. Boyet, J. Blichert-Toft, M.T. Rosing, Early mantle dynamics inferred from 142Nd 166 variations in Archean rocks from southwest Greenland, Earth Planet. Sci. Lett. 377, 324-355 167 (2013). 168 5. T. Kleine, M. et al., Hf–W chronology of the accretion and early evolution of asteroids and 169 terrestrial planets. Geochim. Cosmochim. Acta 73, 5150-5188 (2009). 170 6. M. Willbold, T. Elliott, S. Moorbath, The tungsten isotopic composition of the Earth's mantle 171 before the terminal bombardment. Nature 477, 195-198 (2011). 172 7. M. Touboul, I.S. Puchtel, R.J. Walker, 182W evidence for long-term preservation of early 173 mantle differentiation products. Science 335, 1065-1069 (2012). 174 8. M. Touboul, J. Liu, J. O'Neil, I.S. Puchtel, R.J. Walker, New insights into the Hadean mantle 175 revealed by 182W and highly siderophile element abundances of supracrustal rocks from the 176 Nuvvuagittuq Greenstone Belt, Quebec, Canada. Chem. Geol. 383, 63-75 (2014). 177 9. M. Willbold, S.J. Mojzsis, H.-W. Chen, T. Elliot, Tungsten isotope composition of the Acasta 178 Gneiss Complex. Earth Planet. Sci. Lett. 419, 168-177 (2015). 179 10. H. Rizo et al., Early Earth differentiation investigated through 142Nd, 182W, and highly 180 siderophile element abundances in samples from Isua, Greenland, Geochim. Cosmochim. Acta 181 175, 319-336 (2016). 182 9 11. A.D. Saunders, J.G. Fitton, A.C. Kerr, M.J. Norry, R.W. Kent, The North Atlantic igneous 183 province. Geophys. Mon. Series 100, 45-94 (1997). 184 12. C.R. Neal, J.J. Mahoney, L.W. Kroenke, R.A. Duncan, M.G. Petterson, The Ontong Java 185 Plateau. Geophys. Mon. Series 100, 183-216 (1997). 186 13. N.A. Starkey et al., Helium isotopes in early Iceland plume picrites: Constraints on the 187 composition of high 3He/4He mantle. Earth Planet. Sci. Lett. 277, 91-100 (2009). 188 14. M.G. Jackson et al., Evidence for the survival of the oldest terrestrial mantle reservoir. Nature 189 466, 853-856 (2010). 190 15. L. J. Hallis et al., Evidence for primordial water in Earth’s deep mantle. Science 350, 795-797 191 (2015). 192 16. M.G. Jackson, R.W. Carlson, An ancient recipe for flood-basalt genesis. Nature 476, 316-319 193 (2011). 194 17. Materials and Methods are available as Supplementary Materials on Science Online 195 18. C.W. Dale et al., Osmium isotopes in Baffin Island and West Greenland picrites: implications 196 for the 187Os/188Os composition of the convecting mantle and the nature of high 3He/4He mantle. 197 Earth Planet. Sci. Lett. 278, 267-277 (2009). 198 19. D. Rubie, F. Nimmo, H.J. Melosh, Formation of Earth’s core. Treatise on Geophysics, 199 Second Edition, 2015. 200 20. H. St. C. O’Neill et al., Experimental petrochemistry of some highly siderophile elements at 201 high temperatures, and some implications for core formation and the mantle's early history. Chem. 202 Geol. 120, 255-273 (1995). 203 10 21. J. M. D. Day et al., Origin of felsic achondrites Graves Nunataks 06128 and 06129, and 204 ultramafic brachinites and brachinite-like achondrites by partial melting of volatile-rich primitive 205 parent bodies. Geochim. Cosmochim. Acta, 81, 94-128 (2012). 206 22. M.K. Pető, S. Mukhopadhyay, K.A. Kelley, Heterogeneities from the first 100 million years 207 recorded in deep mantle noble gases from the Northern Lau Back-arc Basin. Earth Planet. Sci. 208 Lett. 369, 13-23 (2013). 209 23. M.G. Jackson et al., Peridotite xenoliths from the Polynesian Austral and Samoa hotspots: 210 Implications for the destruction of ancient 187Os and 142Nd isotopic domains and the preservation 211 of Hadean 129Xe in the modern convecting mantle. Geochim. Cosmochim. In press (2016). 212 24. J.P. Brandenburg, E.H. Hauri, P.E. van Keken, C.J. Ballentine, A multiple-system study of 213 the geochemical evolution of the mantle with force-balanced plates and thermochemical effects. 214 Earth Planet. Sci. Lett. 376, 1-13 (2008). 215 25. E.J. Garnero, A.K. McNamara, Structure and dynamics of Earth’s lower mantle, Science 320, 216 626-628 (2008) 217 26. M. Horan et al., Highly siderophile elements in chondrites. Chem. Geol. 196, 5-20 (2003). 218 27. T. J. Ireland, R.J. Walker, M.O. Garcia, Highly siderophile element and 187Os isotope 219 systematics of Hawaiian picrites: implications for parental melt composition and source 220 heterogeneity, Chem. Geol. 260, 112-128 (2009). 221 28. W. B. Tonks and H. J. Melosh, Magma ocean formation due to giant impacts, J. Geophys. 222 Res. : Planets 98, E3, 5319-5333 (1993). 223 29. F. R. Boyd, S.A. Mertzman, Magmatic processes: Physicochemical principles, 13-24 (1987). 224 30. M. Touboul, R. J. Walker, High precision tungsten isotope measurement by thermal 225 ionization mass spectrometry. Int. J. Mass Spectrom. 309, 109 (2011). 226 11 31. E. Gayer, S. Mukhopadhyay, B.J. Meade, Earth Planet. Sci. Lett. 266, 303-315 (2008). 227 32. A.S. Cohen, F.G. Waters, Separation of osmium from geological materials by solvent 228 extraction for analysis by thermal ionisation mass spectrometry. Anal. Chim. Acta 332, 269-275 229 (1996). 230 33. J.L. Birck, M. Roy-Barman, F. Capman, Re‐Os isotopic measurements at the femtomole level 231 in natural samples. Geostandards Newsletter 20, 19-27 (1997). 232 34. M. Rehkämper, A.N. Halliday, Accuracy and long-term reproducibility of lead isotopic 233 measurements by MC-ICP-MS using an external method for correction of mass discrimination. 234 Int. J. Mass Spectrom. 181, 123-133 (1998). 235 35. I.S. Puchtel, et al, Insights into early Earth from the Pt–Re–Os isotope and highly siderophile 236 element abundance systematics of Barberton komatiites. Geochim. Cosmochim. Acta 125, 394-237 413 (2014). 238 36. M. Boyet, R.W. Carlson, 142Nd evidence for early (> 4.53 Ga) global differentiation of the 239 silicate Earth. Science 309, 576-581 (2005). 240 37. H. Tazoe, H. Obata, T. Gamo, Determination of cerium isotope ratios in geochemical samples 241 using oxidative extraction technique with chelating resin. J. Analytical Atomic. Spectroscopy 22, 242 616-622 (2006). 243 38. Hirahara Y. et al., Detailed bathymetric features in the outer-arc high off the northwest 244 Sumatra-results from KY09-09 cruise. JAMSTEC Rep. Res. Dev. 15, 27-33 (2012). 245 39. R.W. Carlson, M. Boyet, M. Horan, Chondrite barium, neodymium, and samarium isotopic 246 heterogeneity and early earth differentiation. Science 316, 1175-1178 (2007). 247 12 40. R.J. Walker, Highly siderophile elements in the Earth, Moon and Mars: update and 248 implications for planetary accretion and differentiation. Chem. Erde-Geochem. 69, 101-125 249 (2009). 250 41. M.A. Bouhifd et al., Superchondritic Sm/Nd ratio of the Earth: Impact of Earth's core 251 formation. Earth Planet. Sci. Lett. 413, 158-166 (2015). 252 42. J. Wade, B.J. Wood, J. Tuff, Metal–silicate partitioning of Mo and W at high pressures and 253 temperatures: evidence for late accretion of sulphur to the Earth. Geochim. Cosmochim. Acta 85, 254 58-74 (2012). 255 43. J.M.D. Day et al., Late accretion as a natural consequence of planetary growth. Nature 256 Geoscience 5, 614-617 (2012). 257 44. J.H. Jones, M.J. Drake, Geochemical constraints on core formation in the Earth. Nature 322, 258 221-228 (1986). 259 45. J. Li and C. B. Agee, Geochemistry of mantle–core differentiation at high pressure. Nature 260 381, 686-689 (1996). 261 46. M.L.G. Tejada et al., Pin-pricking the elephant: Evidence on the origin of the Ontong Java 262 Plateau from Pb-Sr-Hf-Nd isotopic characteristics of ODP Leg 192 basalts. Geol. Soc. London 263 Spec. Publ. 229, 133-150 (2004). 264 47. Brenan J. M. and McDonough W. F. (2009) Core formation and metal-silicate fractionation 265 of osmium and iridium from gold. Nature Geosciences, 2, 798-801. 266 48. Wheeler, K.T., Walker, D., and McDonough, W.F. (2010) Pd and Ag metal-silicate 267 partitioning applied to Earth differentiation and core-mantle exchange. Meteoritics and Planetary 268 Science 46, 199-217. 269 13 49. R. Arevalo, W.F. McDonough, Tungsten geochemistry and implications for understanding 270 the Earth's interior. Earth Planet. Sci. Lett. 272, 656-665 (2008). 271 50. S. König et al., The Earth’s tungsten budget during mantle melting and crust formation. 272 Geochim. Cosmochim. Acta 75, 2119-2136 (2011). 273 51. D. Francis. The Baffin Bay lavas and the value of picrites as analogues of primary magmas. 274 Contrib. Mineral. Petr. 89,144-154 (1985). 275 52. A. Corgne et al., Silicate perovskite-melt partitioning of trace elements and geochemical 276 signature of a deep perovskitic reservoir. Geochim. Cosmochim. Acta 69, 485-496 (2005). 277 53. K. Righter, C. K. Shearer, Magmatic fractionation of Hf and W: constraints on the timing of 278 core formation and differentiation in the Moon and Mars. Geochim. Cosmochim. Acta 67, 2497-279 2507 (2003). 280 54. M. Boyet, R.W. Carlson, M. Horan, Old Sm–Nd ages for cumulate eucrites and 281 redetermination of the solar system initial 146Sm/144Sm ratio. Earth Planet. Sci. Lett. 291, 172-181 282 (2010). 283 55. C. Vockenhuber et al., 182Hf, a new isotope for AMS. Nuclear Instruments and Methods in 284 Physics Research Section B: Beam Interactions with Materials and Atoms 223, 823-828 (2004). 285 56. G. Caro, B. Bourdon, Non-chondritic Sm/Nd ratio in the terrestrial planets: consequences for 286 the geochemical evolution of the mantle–crust system. Geochim. Cosmochim. Acta 74, 3333-287 3349 (2010). 288 289 Acknowledgments: We thank T. Mock for assistance with mass spectrometry. This work was 290 improved by helpful discussions with J. O’Neil, S. Shirey and B. Wood. We thank S. Shirey who 291 provided sample VE-32. We also thank the helpful comments and suggestions from three 292 14 anonymous reviewers and the editor B. Grocholski. We thank M. Garçon who developed the 4-293 step 142Nd acquisition method. This work was supported by NSF-CSEDI grant EAR1265169 to 294 R.J. Walker, and by EAR1250419 to S. Mukhopadhyay. All data are available in the main 295 manuscript and supplementary materials. 296 297 Figure 1. μ182W values measured for Baffin Bay and Ontong Java Plateau samples, the 298 geological reference materials VE-32 and BHVO-1 and the Alfa Aesar W standard. The 299 values are expressed as parts in 106 or ppm deviations from the average measured from the W 300 standard. The grey shaded area represents the 2 standard deviation (2σ) for the W standard 301 measurements. Errors for each data point are 2σ. 302 303 Figure 2. Highly siderophile element abundances (HSE) for the Baffin Bay and Ontong 304 Java Plateau samples. Abundances are normalized to the HSE of carbonaceous chondrites (CI) 305 of ref. 26. Grey shaded area shows the range of HSE abundances for type-2 Hawaiian picrites 306 (ref. 27). 307 308 Figure 3. Model for the creation of distinct W isotopic mantle reservoirs. (A) Early core 309 formation leaves the proto-Earth’s mantle with a high Hf/W ratio that, with time, evolves to a 310 high μ182W value (i). (B) The impact of a large body affects the Hf/W ratio and W isotopic 311 composition of a portion of the proto-Earth’s mantle. (C) Evolution of the portion of the mantle 312 (ii) affected by the impact of a large body, involving some degree of isotopic equilibration 313 between impactor materials and mantle. The core of the impactor subsequently merges with the 314 core of the proto-Earth. (D) Possible scenario after isostatic adjustment (e.g. 28), and creation of 315 15 a mantle with heterogeneous μ182W through impacts of large bodies. Mantle domains affected by 316 impacts that occur after the extinction of 182Hf, will no longer generate radiogenic 182W, so their 317 182W/184W ratios will change only by mixing with other terrestrial reservoirs or with late-accreted 318 chondritic material. (E) Late accretion representing ~ 0.5% of Earth’s mass decreases the 319 182W/184W ratio of all the earlier formed reservoirs by ~ 15 ppm. This last accretion is responsible 320 for endowing the modern mantle with chondritic relative abundances of the HSE. 321 322 323 324 Table 1. Tungsten concentrations and isotopic compositions measured for Baffin Bay samples 325 Pi-23a, Pi-23b, Pd-2, the Ontong Java Plateau sample 192-1187A 009R 04R, the MORB glass 326 sample VE-32, the BHVO-1 basalt standard. Uncertainties are 2σ. More details are given in ref. 327 17 and Table S2. 328 329 Supplementary Materials: 330 Materials and Methods 331 Figures S1-S7 332 Tables S1-S5 333 Table 1. Tungsten isotopic compositions and concentrations measured for Baffin Bay and Ontong Java Plateau flood basalts, the VE-32 MORB glass and BHVO-1 basalt standards Locality Geodynamic context Sample µ182W (ppm) ± 2σ (ppm) W (ppb) Baffin Bay Flood basalt Pi-23 a 11.9 5.9 62 Baffin Bay Flood basalt Pi-23 b 8.3 5.6 62 Baffin Bay Flood basalt Pd-2 48.4 4.6 26 Ontong Java Plateau Flood basalt 192-1187A 009R 04R 23.9 5.3 23 East Pacific Rise Mid-ocean ridge basalt VE-32 -0.8 4.5 54 Hawaii Ocean island basalt BHVO-1 -2.3 7.7 274 16 References (29-56) 334 335 Figures 336 337 338 339 -10.0 0.0 10.0 20.0 30.0 40.0 50.0 VE-32 BHVO-1 μ182W (ppm) Alfa Aesar W Standard Pi-23b Pi-23a Pd-2 OJP Figure 1 17 340 341 342 343 344 0.001 0.010 0.100 Os Ir Ru Pt Pd Re Hawaiian picrites OJP Pd-2 Pi-23 H SE / C ho nd rit es Figure 2 Core Hf/W ~ 0 High Hf/W high μ182W Core Hf/W ~ 0 Impact μ182W= -200 Core Hf/W ~ 0 Mantle domain created after impact Core Hf/W ~ 0 μ1 82 W Time Late accretionμ1 82 W Time μ1 82 W Time μ1 82 W Impact μ1 82 W Time Time A B C D E Figure 3 i i i ii i ii i ii i i ii ii i i Core Hf/W ~ 0 ii i Late accretionHigh Hf/W high μ182W High Hf/W high μ182W Proto-Earth’s mantle Proto-Earth’s mantle Proto-Earth’s mantle Mantle domain created after impact Heterogeneous μ182W mantle