Experimental study of U,Th solubility in Earth's core ...

Experimental study of U,Th solubility in Earth's core ...

U Solubility in Planetary Cores: Evidence From High Pressure and Temperature Experiments Xuezhao Bao and Richard A. Secco [email protected]; [email protected] Department of Earth Sciences, University of Western Ontario, London, Ontario, Canada N6A 5B7 Joel E. Gagnon and Brian J. Fryer Department of Earth Sciences, University of Windsor, Windsor, Ontario, Canada N9B 3P4 Introduction > Without U, Th, K40, inner core age is 1.5 Ga (should be >3.5Ga see next) (Labrosse et al. 2001, Inner Core without U, Th, K40 Inner Core with U, Th, K40 Buffett, 2002, 2003; supported by Anderson, 2002, Nimmo, 2004, 2005, etc.). > Perovskite-postperovskite- Mantle 13.5 TW 13.5 TW T Perovskite CMB postperovskite Perovskite Outer core

P Mantl e perovskite forward and reverse transformations within CMB indicate a high heat flow of 13.5 TW (Lay et al., 2006) Current inner core size Tarduno et al., 2007 McElhinny et al., 1980 After Tarduno et al., 2006 The existence of a magnetic field of roughly present-day strength over 3.5 Gyr is more easily reconciled with an old inner core (Christensen & Tilgner, 2004). reversal Two unusually long periods without any reversal N S S N Courtillot, 1998 Computer simulations and field works (Coe and Glatzmaier, 2006) U in the outer core? Introduction Information from geo-neutrino studies > upper limit of 65 TW radiogenic heat from U and Th; central value of 16 TW?.

Outer Core Inner Core Current inner core size Araki et al (2005) Mantl e > Problems: 1) Large range of 65 TW 2) U, Th: in mantle or core? 3) Geochemistry model dependent. U,Th decay Geo-neutrinos KamLAND Experiments and composition analytical methods Earth Sun Mercury S in the core 0~0.2 % Metal in the core . Silicate Venus 0% Mars

0-10 % 14.5 % ~ 65% ~33%? ~ 33% ~ 22% 35% 67%? 67% 78% Our starting materials: mixtures of the following powder Silicate ~ 54 wt% peridotite metal phase - 40wt% Fe or simplified expression Fe-10% S Fe-35wt%S U source ~3 wt% uraninite Righter, 2006 Experiments and composition analytical methods 500 ton Walker module multi-anvil press An 3000 ton multi-anvil press will be coming in this Fall Pressure cell (inside)

Welcome to use this lab. You can email me at: [email protected], or [email protected] Experiments and composition analytical methods Pressure Cell Graphite, MgO Experiments and composition analytical methods Composition Analytical methods > LA-ICP-MS (Laser Ablation Inductively Coupled Plasma Mass Spectrometry) University of Windsor, University of McGill > EM (Electron Microprobe). > SIMS( Secondary Ion Mass Spectrometer) P, T range of experiments Sample container BN Graphite MgO 2400 191 190 g eltin m SiO 4 Mg 2 157 O T ( C)

200 115 2000 205 183 204 330 357 358 2 4 182 332 207 360 201 Ohtani & Kumazawa (1981) 333 334 214 154 Fe 260? 212 Saxena & Dubrovinsky (2000) lting e m Fe

198 1800 1600 0 6 Mg2SiO4 Davis & England (1964) 184 262? 253? 199 2200 202 167 8 P (GPa) 10 12 14 Results and discussion Run 182: 5GPa; 2050 ; metallic phase: Fe ; metallic phase: Fe Backscattered electron image Fe Silicate 500 m

Results and discussion Run 238: 5 GPa; 1972 ; metallic phase: Fe ; metallic phase: Fe-10% S Backscattered electron image silicate Fe-10% S 500 m Results and discussion o Run 198, 0 GPa, 1850 C 80 Background Fe 60 Count (number) 40 20 0 0 20 40 60 80 100 o Run 207, 6.7 GPa, 2010 C

400 350 300 250 Spectra of U in the Fe phases at different pressures analyzed by LA-ICP-MS. 200 150 100 50 0 0 20 40 60 80 100 Time (second, 10 m/s) 120 Results and discussion o Run 199, 8 GPa, 2249 C Background Spectra of U in the Fe phases at different pressures analyzed by LA-ICP-MS.

Fe 1400 1200 1000 Count (number) 800 600 400 200 0 0 20 40 60 80 100 o 10000 Run 191, 8.5 GPa, 2287 C 8000 6000 Summary: U in the pure Fe: from 0 to 7.5 GPa, 50 to 10, 000 counts/s Oxides and silicates 4000 2000

Fe 0 0 20 40 60 80 Time (second, 10 m/s) 100 Results and discussion Run products with Fe in metal phase: DU= UFe / Usilicate DU(U in metal/U in silicate) 0.09 199 Container TTsilicate melt BN Graphite 0.08 0.04 262 Silicate

U P Fe 191 167 0.03 190 0.02 253 260 0.01 200 204 0.00 198 0 205 115 2 201 182 183 4 202 214 207 157 154

6 8 Pressure (GPa) 10 12 14 Results and discussion Run Products With Fe-S: DU= UFe-10%S / Usilicate LA-ICP-MS DU (Umetal or metal-sulfide/Usilicate) 0.14 T>Silicate melting line 236 0.12 Silicate U P 199 0.10 0.08 262 161 0.06 234 235 0.04

224 238 0.02 0.00 Fe10%S 231 198 232 115 243 123 239 237 190 191 207 113 240 260 T

10 12 14 16 Results and discussion Possible mechanism for U entry the cores Silicate Fe 2 Silicate Silicate Fe+Si Fe-10wt% S this study Fe-35wt% S this study log fo2(IW) -5 0-28wt% S in Fe phase -4 -1 ~ -6 Wheeler -3 Malavergne

log DU -2 this study -1 ~ -6 Fe20FeS80 0 Murrell+Burnett -7~-3 1 -1 Low fo2 is the key for U entry metal phases. U is very soluble in metal Fe in situations without O and other oxidative elements. Fe-S -2 Study Fe0 + U0 + Si0 reducing UFe10Si2 Possible mechanism for U entry the cores 262(8686 ppm)

199(2303 ppm) 1000 260 191 metal Fe with BN capsule metal Fe with C capsule metal Fe-10wt%S with MgO capsule metal Fe with MgO capsule The others after Fig. 3 800 167 600 190 400 201 207 202 200 183 214 198 0 -7 -6 -5 205

182 253 154 204 200 157 -4 -3 212 333 -2 Oxygen fugacity (IW) -1 0 Results and discussion Run products with Fe: Si increases with P(consistent with previous studies of Ito et al., 1995, Gessmann et al., 2001 and Malavergne et al., 2004).Based on log(wt%)Si fo2 relation fo2 below the IW buffer decreases with P, coinciding with U increase with P in Fe (Kilburn and Wood,1997). 0.5 184 212 202 260 167

253 201 214 0.0 198 207 191 205 204 199 157 Container: BN Graphite 150 182 -1.0 log D(S) 154 190 153 -0.5 IW = -6.0 262 183

115 -1.5 200 -2.0 0 2 IW = -2.0 4 6 8 10 12 14 16 log D(Si) 18 P (GPa) For geo-neutrino study: it may be impossible for significant K in the core. Possible mechanism for U entry the cores 0 -1 358 357

-2 360 115 200 204 -3 212 330 182 205 -5 -6 -8 -9 190 253 191 262 154 157 199 260 202 207 201 198

226 371 183 137 -4 -7 333 153 167 214 Oxides and silicates 150 184 metal Fe with BN capsule metal Fe with C capsule metal Fe-10wt%S with MgO capsule metal Fe with MgO capsule Smaller signs identify samples with very small metal phases. Their U concentrations in metal phases are difficult to obtain with LA-ICP-MS analysis. 0 2 4 6 8

P (GPa) 10 12 14 Fe 16 Possible mechanism for U entry the cores According to Arora (2000) increase P, reactions proceed to left, fo2 increase. But in fact, increase P, fo2 decrease. 0 -1 358 357 -2 360 115 200 204 -3 212 330 182 205 -5

-7 Fe can be oxidized even at room T (Kubaschewski et al., 1962) -8 -9 183 190 198 253 191 262 154 157 199 260 202 207 201 -6 226 371 137 -4

O,Fe-O Fe 333 153 167 214 150 184 metal Fe with BN capsule metal Fe with C capsule metal Fe-10wt%S with MgO capsule metal Fe with MgO capsule Smaller signs identify samples with very small metal phases. Their U concentrations in metal phases are difficult to obtain with LA-ICP-MS analysis. 0 2 4 6 8 P (GPa) 10 12 14 16 Possible mechanism for U entry the cores

A) Si , Al 4+ 3+ impact Si , Al 0 Al-O 0 Si-O Yakovlev (1993) discovered metallic Si and Al in the impacted products. B) Serghiou et al., (1992) found at room T, high P (19.2 GPa) can decrease the O content of Tetragonal Nb2O5, and it was amorphized. Al O Si O O-Nb-O O-Nb,O Nb cant be oxidized by O below 200oC (Kubaschewski et al., 1962) Possible mechanism for U entry the cores P Confirmed by Drickamer et al., 1969 and Burns, 1993, etc. in a wide variety of compounds.

P P Burns (1993). Similar elements: Mn, Cu, Ti, Cr and Ni Core-mantle boundary If Burns (1993) is right, then it is natural Fe--O P P Fe O Possible mechanism for U entry the cores McCammon & Kopylova, 2004 Xenoliths from the mantle under northern Canada (Mossbauer data) Woodland & Koch, 2003 Xenoliths from the mantle under Lesotho and South African (Mossbauer data) Possible mechanism for U entry the cores 0 -1 358 357 -2 360 115 200 204

-3 212 182 205 -5 -6 -8 -9 183 190 253 191 262 154 157 199 260 202 207 201 198 226 371 137

-4 -7 333 330 153 167 214 150 184 metal Fe with BN capsule metal Fe with C capsule metal Fe-10wt%S with MgO capsule metal Fe with MgO capsule Smaller signs identify samples with very small metal phases. Their U concentrations in metal phases are difficult to obtain with LA-ICP-MS analysis. 0 2 4 6 8 10 12 14 16

P (GPa) Average IW +4, basalts on the surface (Karner et al., 2006) ~0.6/GPa (Ballhaus, 1995) Possible mechanism for U entry the cores Proto-planetary disk or nebula ~IW -6.8 (Average, IW -6.8 (Average, Grossman et al., 2008) ~IW -6.8 (Average, IW -13~8.9 (enstatite chondrite (EC), Grossman et al., 2008) EC model: based on composition (Javoy, 1995,1998), and O(Clayton,1993),Cr(Lugmair, 1998), and Mo(Dauphas, 2002) isotope compositions. Possible mechanism for U entry the cores ~IW -6.8 (Average, IW -6.8 (Average, Grossman et al., 2008) ~IW -6.8 (Average, IW -13~8.9 (enstatite chondrite (EC), Grossman et al., 2008) Proto-planetary disk or nebula Fe-rich region Condensation T of Fe is higher than that of silicates at the relevent pressures of the solar nebula (~10-4 atm, Grossman, 1972) Inner planet cores may have formed completely (Hwaung, 2000) or at least partially (Grossman, 1972), from the Fe-rich materials in the Fe-rich region. Metal

core Earth Possible mechanism for U entry the cores Core formation during Earths accretion Wood et al.,2008 Core has formed entirely under highly reducing conditions (Wnke, 1981; Javoy,1995; Allgre et al., 1995) Core has formed mainly under an early highly reducing conditions (ONell, 1991; Javoy,1995; Wade & Wood, 2005; Wood et al.,2008 ) Possible mechanism for U entry the cores 262(8686 ppm) 199(2303 ppm) 1000 260 191 metal Fe with BN capsule metal Fe with C capsule metal Fe-10wt%S with MgO capsule metal Fe with MgO capsule The others after Fig. 3 800 167 600 190 400 201 207 202

200 183 214 198 0 -7 -6 -5 205 182 253 154 204 200 157 -4 -3 212 333 -2 Oxygen fugacity (IW) -1 0 Possible mechanism for U entry the cores O2 and water contents (McCammon et al., 2004 ) are the control factors of fo2. Other oxidative volatiles are also important. Element Density Melting

point Compound Density Melting point Compound Density Melting point Fe 7.86 1535 Ni 8.90 1453 U 19.05 1132 Th 17.70 1700 K4ThOX4.4H2O soluble in water ThO2 10.0 3220 ThN 10.6 2500 ThS 1905 ThCl4 600 Th(NO3)4.5H2O soluble in water UO2 UCl UN UF6 UO3 10.95 2800 4.86 567

14.32 2800 4.68 ~0 7.29 soluble in water The melting point (oC) and density (g/cm3) of U, Th, their compounds and complexes and core related elements (after Bao and Zhang, 1998). Oxides and silicates U Metal O or other volatiles High density(19 g/cm3) Small density (0.00143g/cm3) Implication for planetary dynamics Earth Only planet with life; only planet with plate tectonics; only planet with asthenosphere; only planet with granitoid continental crust; with magnetic field; with ocean Pacific ocean Continent Circum-pacific volcanic belt magmatism U,Th rich zone Volatiles or their ions U - ne- U n+

Th - ne- Th n+ The distribution of U and Th in the outer core and its influence on the formation of deep mantle plumes and subducted lithosphere plates in the Pacific Ocean.A-lithosphere; BU,Th richer sphere or asthenosphere; C- mantle; D''- core-mantle boundary; D- outer core; E- inner core; O- center of Earth; M- center of geomagnetic field and thermal convection; the black points represent the relative concentration of U and Th (after Bao, 1999). Implication for planetary dynamics Average U, Th concentration (ppm) Oceanic Continental crust crust Th 5.6 U 1.4 U Th 0.1 0.2 Taylor & Rudnick & Cumulative volume of McLennan, Fountain, continental crust 1985 1995 extracted through time from different researcher groups (After Abbott et al., Earth Application to Earths core Run product with Fe Magma Ocean

DU(U in metal/U in silicate) 0.2 Container TTsilicate melt BN Graphite 0.17 x 20 ppb U in Primitive Silicate Earth (PSE, based Cl chondrite) (McDonough, 2003) DU= 0.17 T > silicate melting 3.4 ppb U 0.1 DU= 0.047 Percolation 0.047 x 20 ppb U in PSE T < silicate melting 1 ppb U 0.0 0 2 4 6

8 10 12 14 Pressure (GPa) 16 18 20 22 24 26 0.21 x 20 ppb U in PSE(to CMB?) 4.2 ppb U Earths major, minor, and isotopic composition is unlike any known chondrite group or mixture of chondrite group (Righter, 2003, Caro et al., etc.) Run product with Fe-10% S Application to Earths cores T > silicate melting line the co re 10.2 ppb U in the core?

3.4 p pb U in Mars 5 ppb U in the core th E ar 22 ppb U at CMB? T < silicate melting line 1 ppb U in the core 4.2 ppb at CMB? Mercury & Venus Influence of S concentration in metal phases on the extrapolated D U values at 26 GPa, the pressure at the bottom of a proposed magma ocean. Implication for planetary dynamics P, T distribution in the protoplanetary disk from different researchers (after Fegley, 1999). O and volatile contents are controlled by two factors: planetary heliocentric distances (Lewis, 1974a, 1984) and size (Ahrens, 1993) Implication for planetary dynamics Mercury

Giant impact: lost most its mantle and almost all its Fe has entered its core (Benz et al., 1988; Palme et al., 2003) 639 km Heat loss Liquid outer core Mantle 1800 km core Inner Fe-Ni core? U, Th Structure of Mercury O-Si-Mg (FeO<0.3%, Inferred spectrometry) (After Strom and Sprague, 2003) Formation Model of the magnetic field in Mercury. (After Stanley et al., 2005) Implication for planetary dynamics Mars Mars possesses no internal magnetic field, but has old (>4 Ga) Magnetic records (After Solomon et al., 2005) Martian core is at least partially liquid; confirmed by solar tide study (Yoder et al., 2003) Magnetic records implies that a dynamo was present in its early

stage but must have be absent after 4.0 Ga (Nimmo &Tanata, 2005; Fei & Bertka, 2005; Solomon et al.,2005), but why? Implication for planetary dynamics Hemispheric dichotomy structure on Martian crust Mars (Solomon et al., 2005) < 3.7 Ga, lowlands, basaltic composition > 3.7 Ga, less dense highlands, andesite or basaltic andesite, high radioactive elements (Nimmo&Tanaka, 2005) 4.12 Ma (Nimmo & Tanata, 2005) A relatively small impact can drive off Martian volatiles (1034-1036.5 ergs). For Earth 1038 ergs (Ahrens, 1993) Today: dry Martian mantle from Martian meteorites (Dreibus & Wanke, 1985, 1987; Carr & Wanke, 1992; Ghosal et al., 1998; Reese & Solomatov, 2002; Herd et al., 2002, Jones, 2004 ); Infrared mineralogical mapping (Mustand et al., 2005; Bibring et al., 2005). The content of interior volatiles are 2000 times less than that of Earth (Beaty et al., 2005 ). Implication for planetary dynamics Mars 14.5% S in the core (Fei & Bertka,2005) U, Th(?) has entered

the core andesite or basaltic andesite U rich zone Volatiles or their ions U-4e U4+ A-lithosphere; B-asthenosphere (U,Th rich sphere); C-mantle; D-coremantle boundary; D-outer core; E-inner core (after Bao, 1999) Similar internal circulation system and plate tectonics might have developed before 4.12 Ga? Giant impacts have drove off Martian volatiles and then these systems stopped to work and the dynamo are dead since then. Conclusions For P<15 GPa, T< 2500 ; metallic phase: Fe , DU in FeS (including Fe-35% S and Fe10% S) and Fe melts increase with P,T. DU is 3-5 times larger when silicate was molten than silicate was solid. For a 800 km thick magma ocean, Earths core could contain about ~10 ppb U? from the extrapolated result of Fe-10% S samples or 3.4 ppb U from Fe samples. For percolation core formation, Earths core could contain 5ppb -22 ppb (?) U from the extrapolated result of Fe-10% S samples or 1ppb - 4 ppb(?) U from Fe samples. Inside Earth, U may be a radioactive heat source to slow down the cooling and crystallization of the inner core. U is also a possible energy source for maintaining the magnetic fields of Earth and Mercury. Acknowledgments Thanks to: NSERC, CFI Dr. G. Young for partial funding. Drs. C. Cermignani and M. Liu for EM analyses Drs. W. Minarik and J. Gagnon for LA-ICP-MS analysis. Dr. A. Pratt and Mr. G.Gord for SIMS analysis. Dr. D. Liu and Mr. R. Tucker for help in sample preparation Mr. G. Wood for help in sample cutting and polishing. Thank you! Implication for planetary

dynamics Calculated fraction of atmosphere blown off versus impactor energy for Earth, Venus, and Mars. Lower and higher energy curves for each planet correspond to assumed polytropic exponent of ideal gas of =1.1 and 1.3, respectively (After Ahrens, 1993). Implication for planetary dynamics Water-rich atmosphere formed by the impact-induced dehydration of water-bearing minerals by planetesimal impacts (After Ahrens, 1993) 2400 191 -2.31 190 ing t l e m iO 4 S g M 2 O T ( C) 2000 -1.19 198 1800 200 -2.48 115 204

357 330 -2.48 -2.13 358 -0.97 -2.31 -1.59 -3.83 157 333 202 214 205 183 201 334 154 -4.01 182 332 207 212 360 -1.83 -1.80 -1.14 -1.83 fo2(IW)(Wheeler et al., 2006) 0 2 4 6 -1.92 -1.98

260? lting e m Fe Sample container BN Graphite MgO -2.02 1600 -4.69 184 262? 253? 199 2200 -2.06 167 8 10 12 14 P (GPa) Fe can be oxidized even at room T (Kubaschewski et al., 1962) O,Fe-O

Fe

Recently Viewed Presentations

  • Modeling Whole Plant Water Transport: Introduction to Plant

    Modeling Whole Plant Water Transport: Introduction to Plant

    Most of the biologically available energy in the biosphere is derived from the conversion of sunlight into energy-rich carbon compounds. ... This is a SEM of two vessels running side by side in a bottle tree. For water to move...
  • Master Slide - Leeds School of Business

    Master Slide - Leeds School of Business

    Step 1: Establish the Facts Gather all relevant facts Irrelevant facts may become relevant later Constantly be aware that new facts may be needed as the research progresses Significant Facts Client's tax entity Client's personal (family) situation Client's marginal tax...
  • Orientation Workshop - Ryerson

    Orientation Workshop - Ryerson

    You are an Ambassador for Ryerson, Toronto and Canada. The value of your degree is directly related to Ryerson's reputation. If you hurt that reputation, the value of your degree is lessened. Think about your actions - a positive impression...
  • My Shopping, My Way

    My Shopping, My Way

    Tested hypotheses about the future of shopping. Survey goals: Understand how / why consumers use technology. Learn how technology is shaping consumers' behavior / expectations. Test delivery of technology-enabled shopping experiences that empower customers. Why Cisco Conducted the "My Shopping,...
  • NHS England Presentation Template

    NHS England Presentation Template

    Deaths per 1,000 prisoners by apparent cause, 12 months ending December 2000 to 12 months ending June 2016, England and Wales . In the 12 months to June 2016 there were 321 deaths in prison custody, an increase of 74...
  • Introduction to PsychToolbox in MATLAB

    Introduction to PsychToolbox in MATLAB

    Problem solving, detective work, deductive reasoning. Debugging may take longer than initial writing. Enjoy it! ... integer: no decimal places 345. Numbers. Basic calculations in Matlab. Matlab is a powerful graphical calculator. Square root. Exponential.
  • Two Types of Library Facilities

    Two Types of Library Facilities

    Beirut, Lebanon truck, TNT with gas enhancement 12000 Feb 1993 World Trade Center. New York, USA van, urea nitrate and hydrogen gas 2000 Apr 1995 Murrah Federal Bldg. Oklahoma City, USA truck, ammonium nitrate fuel oil 5000 Jun 1996 Khobar...
  • Kerr-lens Mode Locking Without Nonlinear Astigmatism Shai Yefet,

    Kerr-lens Mode Locking Without Nonlinear Astigmatism Shai Yefet,

    For continuous-wave (CW) operation, the astigmatism is linear and can be fully compensated by correctly choosing the folding angles of the focusing elements. For mode-locked (ML) operation induced by the nonlinear Kerr effect, an additional nonlinear astigmatism from the Kerr...