AGE and ORIGIN of IAB IRON METEORITES and THEIR SILICATE INCLUSIONS INFERRED from HF/W CHRONOMETRY Schulz T

Lunar and Planetary Science XXXVII (2006) 1401.pdf

AGE AND ORIGIN OF IAB IRON METEORITES AND THEIR SILICATE INCLUSIONS INFERRED FROM HF/W CHRONOMETRY Schulz T. 1, Münker C. 2, Mezger K. 3 and Palme H. 1, 1 Institut für Geologie und Mineralogie, Universität zu Köln, Zülpicher Str. 49b, 50674 Köln, Germany, 2Mineralogisch-Petrologisches

Institut, Universität Bonn, Poppelsdorfer Schloss, 53115 Bonn, Germany, 3Zentrallabor für Geochronologie, Insti-

tut für Mineralogie, Universität Münster, Corrensstr. 24, 48149 Münster, Germany

Abstract: The decay of 182 Hf (half life of 9 mil- acapulcoite-lodranite group. Although the formation lion years) to 182 W provides a powerful tool to date and thermal history of the IAB parent body remains metal-silicate fractionation in the early solar system. unconstrained, it may be related to condensation in the During segregation of metal the siderophile W parti- solar nebular [1], or to metal-silicate fractionation in tions into the metal phase, whereas the lithophile Hf is the interior of planets or planetesimals, triggered ei- preferably incorporated by silicate phases. The forma- ther by internal heat sources [4-6], external heat tion of iron meteorites during metal-silicate fractiona- sources such as impacts [2,3] or both [7-9]. tion may have occured during condensation in the solar nebula [1] or in the interior of planets and Samples and analytical methods: From each of planetesimals [2-9]. Iron meteorites with silicate in- the selected iron meteorites a few chips were cleaned clusions like the IAB-IIICD group are particularly with steel-free abrasives and put in an ultrasonic bath interesting for the application of the Hf-W system as a before leaching in 0,05M HNO 3 for several minutes. chronometer for silicate-metal fractionation events From all samples, matrix metal and silicates were during the early history of the solar system. We inves- separated, for Caddo County, Landes and Copiapo a tigated this meteorite groups closer by systematic mixture of individual silicate phases was used, analyses of four IAB iron meteorites (Landes, whereas for Campo del Cielo a single inclusion was Copiapo, Campo del Cielo and Caddo County) and extracted. All silicate phases were subsequently pow- their silicate inclusions. Data for metals are in agree- dered in an agate mortar to reduce the grain size. Dur- ment with earlier analyses of metal phases in IAB- ing this procedure magnetic fractions were continu- IIICD irons [10] at around -3 w, whereas silicates ously removed using a hand magnet. In the case of display large excesses of 182 W/ 184 W, between 0.9 and Campo del Cielo the magnetic fraction was visibly 6.6 units. From an internal silicate isochrone for free of silicates. Separation of Hf and W was per- Campo del Cielo an age significantly different from formed by anion exchange methods [17]. All isotope the model age for the metal phase of the same sample compositions were determined using the Micromass can be inferred. The silicates of Copiapo appear to be IsoProbe MC-ICP-MS at Münster [17]. older than the metal phase, whereas the silicates for Campo del Cielo appear to be younger. Results: Figure 1 shows two internal silicate inclusion isochrones for Campo del Cielo and Copiapo. Introduction: Most iron meteorites exhibit geo- In Figure 2 ages for the silicates and model ages of chemical trends that can be explained by fractional metals are plotted relative to Ste. Marguerite (4565 + crystallization during core formation on a chondritic 0.7 Ma) [18]. The isochrones define initials and ages parent body [11,12]. These are magmatic iron meteor- which differ significantly from the model ages of the ites, whereas non-magmatic iron meteorites show ba- metals of the same meteorite. Isochrones for Copiapo sically unfractionated siderophile element patterns. and Campo del Cielo yield absolute ages of 4571.0 + The non-magmatic groups exhibit a unique character- 3.9 Ma and 4555.9 + 2.5 Ma, respectively. Model ages istic in the presence of abundant silicate clasts with up of the matrix metal phases are identical within ana- to a few centimeters in size. They contain planetary- lytical uncertainties. w for inclusion metal in Campo type noble gases [9] and a chondritic to subchondritic del Cielo is indistinguishable from the value for the composition and mineralogy, e.g. the basaltic to ande- matrix metal. sitic silicate phases of Caddo County [12]. Most silicate clasts appear to be unaffected by extensive partial Discussion: Tungsten in silicate inclusions dis- melting and/or fractional crystallization in the interior plays a very different isotopic composition than W in of planets. Based on their oxygen isotope composition, both matrix and in inclusion metals. In addition, the IAB meteorites are related to the non-magmatic isochrones defined by the silicate inclusions do not IIICD meteorites and the primitive achondrites of the pass through matrix metal. This suggests a different winonaite-group [13] and to a lesser extent to the origin of inclusions and metal. The identical composi- Lunar and Planetary Science XXXVII (2006) 1401.pdf

tion of matrix and inclusion metal indicates that in- 860, (1979); [17] Kleine et al., Geochimica et clusion metal is derived from matrix metal and is un- Cosmochimica Acta 68, 2935-2946 (2004) ; [18] related to the host phase of W in the silicate inclu- Polnau et al. LPS XXXII , Abstract #1527 (2001) sions. The age of Copiapo silicates defined by two points is older by 3.8 Ma years than the canonical solar system age. Planetesimals with ages older than

the canonical CAI age were previously predicted by 10 Kleine et al [10] on the basis of low 182 W/ 184 W ratios 8 in iron meteorites. If the weakly magnetic fraction of NM Copiapo were contaminated with inclusion metal, the 6 NM metal free Copiapo silicates would even be older. All 4

ages that are calculated from the W-isotopes are older 2 Copiapo - 4

than typical silicate ages inferred from Rb-Sr and Sm- W m = (1.36 + 0.26) x 10 ε ε ε ε 0 WM i = -5.29 + 1.2 Nd chronometry which may record remelting of sili- WM T = 4571.0 + 2.6 Ma -2 cates. In contrast, Hf-W-ages require silicate metal M 182 184 Campo del Cielo equilibration. The uniform W/ W ratios of the -4 m = (0.42 + 0.08) x 10 - 4 i = -1.38 + 0.77 four matrix metals enables a closer assessment of -6 T = 4555.9 + 2.5 Ma cosmic ray effects on W isotope compositions in iron meteorites. 0 4 8 12 16 20 180 184 Long exposure times of many iron meteorites to cos- Hf/ W mic rays can cause significant cosmic ray effects in 182 182

W, consequently reducing W by neutron capture. Figure 1: Internal isochrons for silicate inclusion of Campo ε Although this effect cannot be ignored, our first re- del Cielo and Copiapo. The w units represent the deviation sults indicate that the effect must be negligible: Caddo of W isotope ratios in the sample from the terrestrial stan- County and Landes metals display, within limits ana- dard value in parts per 10.000. WM = weakly magnetic, 182 184 lytical error, the same W/ W ratios, although they NM = non-magnetic, M = Metal. have drastically different exposure ages, 240-270 my for Landes [14] and 2.8-3.9 my for Caddo County [15]. A more rigorous interpretation of the W-isotope data requires an independent monitor of the neutron flux. We therefore envisage to monitor the thermal neutron flux by analyzing Gd-isotopes in coexisting metal Landes 1 silicates.

Caddo County metal References: [1] Wasson, Proc. Intl. Geol. Cong. 1 24, 161-168 (1972); [2] Wasson et al., Z. Naturforsch. metal 35a, 781-795 (1980) ; [3] Choi et al., Geochim. Cos- Copiapo mochim. Acta 59 , 593-612 (1995); [4] Kelly et al., silicates 1 Geochim. Cosmochim. Acta 41 , 93-111 (1977); [5] Kracher, Geophys. Res. Lett. 9 , 412-415 (1982); [6] Kracher, J. Geophys. Res. 90 (Suppl.), C689-C698 metal Campo del Cielo silicates (1985) [7] Takeda et al., Meteoritics 29 , 830-842 1 (1994); [8] Takeda et al., Geochim. Cosmochim. Acta 64 , 1311-1327 (2000); [9] Benedix et al., Meteoritics -15 -10 -5 0 5 10 15 35 , 1127-1141 (2000); [10] Kleine et al., Geochim. Time (Ma) Cosmochim. Acta 69 , 5805-5818 (2005); [11] Scott, ed., Univerity of Arizona, Tucson , 892-925 (1979); Figure 2: Ages relative to Ste. Marguerite for silicate [12] Stewart et al., Earth and Planetary Science Let- phases in Campo del Cielo and Copiapo and model ages for ters 143 , 1-12 (1995); [13] Clayton et al., Geochim. metal phases. 1 = I/Xe ages for silicates [16]. Cosmochim. Acta 60 , 1999-2017 (1996); [14] Nie-

meyer, Geochim. Cosmochim. Acta 43 , 1829, (1979);

[15] Bogard et al., LPS XXX , Abstract #1253 (1999); [16] Niemeyer, Geochim. Cosmochim. Acta 43 , 843-