PROGRESS BY THE CONSORTS OF ANGRA DOS REIS* The Angra dos Reis, Rio de Janeiro, Brazil, achondrite fell in January, 1869, as one stone weighing

1.5kg (1). It is a unique meteorite, an ultramafic pyroxenite, having very primitive 8'Srl"Sr [ADOR (2)], an old A'7Pb-'"fiPb age of 4.555 AE (3), and evidence for extinct 2 4 4 P ~ (4). Because of the importance of this meteorite for understanding the early differentiation of meteorite parent bodies, a comprehensive consor- tium study was initiated. The entire remaining 122.6 gm specimen of Angra dos Reis was made available for inspection and sampling under controlled conditions through the cooperation of the Museu Nacional, Rio de Janeiro, Brazil. Polished thin sections of the rock and of separates were studied microscopically and by electron microprobe. Mineral analyses are in general similar to those reported by Hutchison (5) and Albee and Chodos (6), except that we report for the first time kirschsteinite from a meteorite. There are no grain to grain compositional variations exceeding the precision of the analyses. The meteorite consists of over 90% of clinopyroxene of fassaite composition (Table 1; Fig. I ) , characteriz- ed by high A1 and Ca contents. In the structural formula, A1 in tetrahedral position does not balance the sum of other trivalent cations and an amount of divalent iron was converted into the trivalent state to make up the difference. Olivine ( 5 5% of the meteorite) in grains up to 1 mm in size contains small islands ( 5 100 pm) of kirschsteinite ( 5 0.1% of the meteorite), a member of the isomorphous series of monticellite (Mo) and kirschsteinite (Ks) (Table 1, Fig. 1). The mineral in Angra dos Reis is -Ks62.3 Mo37.7. Magnesian-aluminian hercynite makes up about 0.5%-2% of the rock, with Fe3+ content calculated to balance the structural formula. Noteworthy are the large (mm-sized) grains of whitlockite ( 5 1% of the rock) (Table 1) which can be recognized by their greenish color. Metallic nickel-iron (Fe 95.5, Ni 3.5, Co 1.15, total 100.15) and troilite (Fe 62.9, Co 0.07, S 36.5, total 99.47) are rare. Baddeleyite, plagioclase (bytownite) and Si02 were found for the first time in this meteorite. The major mass of ADOR was brought to Caltech for preparation of samples for a prototype experiment for chemical and isotopic studies:The fragment surfaces contained fusion crust and saw cuts and a 5.8 g fragment was broken off to provide interior material. A mass of 0.8 g was crushed to -75pm and separated using heavy liquids. The 3.3 sinks were distributed to the consorts. The weight of all material with P< 2.8 was -1 mg. The main low density fraction (p=2.8-3.3) weighed 10 mg and was 40% whitlockite and 60% pyroxene. Abundant vacuoles and inclusions were observed in the phosphate. During sample preparation several larger ( -1 mm) yellow-green grains were observed which proved to be whitlockite (Table 1). A 240pm grain was analyzed for Pb to determine the concentration level and isotopic composition; ten clean, clear whitlockite grains ( - 250pm each) were picked with a total weight of -200 pg (estimated from diameters) and analyzed for Pb, U and Th using a Z05Pb tracer and microprocedures (7, 8). INAA analyses of putative ultrapure whitlockite were aborted because it was discovered that not all yellow-green grains were phosphate but were En52 Wo2 Fs46 to the embarrassment of the distributing agent. Each density fraction was analyzed for selected elements including Xe (Table 2). To establish the REE pattern of "pure" pyroxene a sample of 3.3 sinks was split and subjected to INAA (Table 3). One split was leached with acid to dissolve phosphates and the other not treated. The results show no difference in the concen- trations for the REE measured and are somewhat lower than those reported for the total rock (9). We con- clude that the REE are in pyroxene and not in trapped liquid or phosphate. The pyroxene concentrations may thus be used to calculate the composition of the magma from which ADOR crystallized. Using D values from (10) and for Eu at f(02) - 10 - I3 bars from (11) we obtain values for the magma (Table 3). The parent magma appears to be highly enriched relative to chondrites, has a positive Eu anomaly and is the most highly fractionated extraterrestrial magma source so far observed. This requires that the parent magma itself be derived from a previous partial melting process. Rb-Sr analyses were carried out on a 75 mg rock fragment (Table 4). The observed mSr/86Sr and 84Sr/88Sr are given along with the calculated initial values. The 84Sr abundance is not distinguishable from ALL or seawater. The 87Sr/R6Sr clearly confirm the original data (2) which showed this meteorite to have an I value far below BABI. The value for ADOR is confirmed but may possibly be the same as ALL to within limits of error. We conclude that an ancient, highly differentiated parent planet has preserved initialsr that is distinc- tive. Xe was measured in 199 mg of 3.3 sinks and 4.7 mg of 2.8 sinks to establish 2 4 4 P ~ fission (F), trapped (T) and spallation (S) Xe distributions (data in Fig. 2). Xe was extracted in three steps; only 1450°C data are presented. 700°C fractions contained only Xe with air composition. Blank Xe of air com- position contributed 3% of 132Xe of the 3.3 sinks and 11% of 132Xe of the 2.8 sinks in the 1450°C fractions.

* The ADORABLES are K. Keil (chief consort), M. Prinz, P . F. Hlava (Univ, of New Mexico); C. B. Gomes (Univ. de Sao Paulo); W. S. Curve110 (Museu Nacional, Rio de Janeiro); G. J. Wasserburg, F. Tera, D. A. Papanastassiou, J . C. Huneke (Caltech); A. V. Murali, M.-S. Ma, R. A. Schmitt (Oregon State Univ.); G. W. Lugmair, K. Marti, N. B. Scheinin (UC San Diego); R. N. Clayton (Univ. of Chieago).

0 Lunar and Planetary Institute Provided by the NASA Astrophysics Data System

11'0T LO'OT L00'0T 193 09'b 9bS'P ale4dso4d OL'OT LO'OT LOO'OT E9'P 2 9 3 8P9'P bll

IIlzcz/qdeoz nerdqdsoz (Idsozlqdroz ald"Je~

8'1 91 I 2 1 01 80 9.0 fo 2 . 0 0 .Ala~!13adsal S ~ P O O ' O pue OPLO'O yuew qdboz .lonb!le awes atli u!

pau!wjalap alaM 41 pue n se IlaM se .3u03 pue .duo3 qd !qd,,, ~I!M pay!ds seM aldwes (p

t sI& soo V U D N ~ suod~!cld pue lalaauvs (E .LESC (0~61) I L .dv.3ojd ' ~ e

IewJoN (8 . ( & L ~ L 'OZ sn~e31 l e je Lelg) apually w o ~ j (LO) alnJpuoq3 q3!J IV-83 (L '~ - l e io l JO ~I!M paleall '6ucg i u 6 ! a ~ :I :paleaJlun ' 6 ~ 2 8 146!a~ :n 'am1 )onb!le uo!lnlos pay!dsun uo suo!~!sodwo~-~ palnseaw (g .(0~61) s!saql .o.qd 'no!selseuedy ~aqlo u! pals!l cw3/6~.& < d lo )!ids e s! auaxoJLd .ni lo1 %L 'V 'a W O J ~ (9 .3y g.p JO a68 ue Gu!wnsse UO!I!SO~WO~ le!l!ul (b 'pTL060L'O=JSes/JSis Pue q~ 'n3 'ws 'e l 'eN 'IV JOJ %S-L ale sJoJJa pa~eur!~sg ( L spla!A sun1 latemeas 1ua3a~ @!a 40 a6e la~e !uaA!6 saJn6!) Isel 01 puodsa~ -103 pueDZ ale slolla :UO!I!SO~UIO~ palnseaw (& ' 0 0 1 ~ (2 .B/alow @-OL u! uo!leJ]ua3uo3 (1 EE - 99'0 89'0 ludd n l

9E EL'P Z'P 9'P wdd s2T 8PL9OO'O ZT9L869'0 ZTLL869'0 b10'0 9'L12 010'0 r l l V ILL ~ 8 ' 1 9' 1 wdd n3

a' ~8T9EL900'0 9T08869'0 ST 68869'0 09 ~ 9 . 9 6'9 8.9 udd US - 9T6L869.0 '388869'0 9E1'0 S.LZ1 PLO'O E-1 29 1 E'9 ~ ' 9 wdd e l -

- t r ~ t 8 8 6 9 ' 0 ~ ~ 6 8 8 6 9 ' 0 ~ 8 0 . 0 P'SZL PPO'O 92-1 - - ,920'0 ZEO'O EEO'O (%I O"N

- LTE8869'0 LTZ6869'0 Sb1'O 8.991 L60'0 * I -1 - P'O L 2'6 (%) C O z ~ ~

al!lpuoq3 x I n JS~S/JS,~ ! (JSS~/JS~~) ~(JSS~IJSLS) z~~ss/qtl~n IJS Lqtl ,euBey( ~ 3 0 8 ,auaxo~/(d luawal3

(elep tl3alle3) 'r elqal (elep .n!un aleis uo6alo) 'C e 1 q ~ 1

, I , , , , , ,

.( t< d) ax U ~ I ~ D I I O ~ S

- 5 t 9 :'.,I :J,qE pus uu!iJsij (~.~<i';C.E) I : ' ~ z t ' : 'Jss=L~x~,: , :sax,,., :ddxz,, a18 slunocuw an!je[ax 'umoqs am

-qJol [s$ol uo a s p assa[al [UulaqL aa)szsdas ID lJ lsa l l .paq~!lua-al!qno[l!q~ pue aoaroidd o! a x Lounyy 103 ~wnills!p UO!IU[~UOJ ado~08! aaiqL :Z 'B[d

-

-

(elep walle3) '1 elqsl 'lualuo3 UOJ! lelol palnseaw UOJJ

paie1n31e3 o!iel ,aj pue ,ad . Ielep 43alle3) al!Woll!4M '9 '6L'Z A n '9'E I W 'P'EE dS 'b'9S 3~ '8.8 u 3 !(a1!uL3laq ue!u!unle-ue!sau6eu) lau!dg 'p 'L'LE ow ' r z g sy !al!u!alsqc~s~!y '& '6'9P e j 'L'ES 04 :aU!A!lO '1 '6'1L s j 'L'PS OM ' y&& u3 :(al!esse)) auaxolAd .L

0 ,say!ds WOJJ uo!leu!weluo3 S S O J ~ 101 papa~~o3 s! unJ m 0 3 JOJ anleh p !yuelq JOJ pa13a~~o3 -un (E 'LO'~E=,! pue S P ' S L = ~ ' 'OEY~L=P 4 i ! ~ Yuelq JOJ pa13a~~o3 (2 'salowo3!d ul (1

- 2 0 ' ( , O ! S ~ N U ~ ) PUB ( b o ! ~ a d s 3 ) "1aq"aw puajo srnla? u! a%!u!al8qoar!y jo pus ' (sd) puw (od) JO sluial U! are au!A!Io Jo suo!3!sdmu3 L'8E 090'2 vL'66Z 9L8200'0 S91'2 9bE'L EE8'6 2.0- aieqdsoqd

b . 0 .(OM) pus (ad) (UX) JO s w ~ a l U! a1!n80e3 XU!? 6'b9 9L.L 1 6'PZb Z'bZb P9620.0 62.81 1P'LL 89'P 92 t l l -?old 'wur2s!p auaxadd [u~a?e[!xpsnb :I 'B[d

t41 ~ I l s c a 3U03 duo3 qdm qdsoz qd~oz qdeaa 6~ .UUI.I*IJ.II'I 'A am - 9 a _ p , y , y , ? ~ , ~ , ~ c.= h ,qd al!~oalaw 146!aM aldwes

W O d % W

L'9P 'P'U 89'0 'P'U 'P'U 'p'u 'P'U LL'O 9'68 9L'O

28'1 0.8 'P'U 81'0

61'1 P'BZ - P'E 'P'U LO'O 'P'U E'E 'P'U 9'P9 'P'U S9'0

L9'0 LS'O

'9 'P

X (elgo 43aile3) '5 e l q u .as!3a~d ale sa3uepunqe aA!lelal ! u~ouy llaM IOU 146!aM aldweg (L

'981 (OL6L) b& el3V '3 8 '9 'B~aqua40~ (9 'L9P (~961) E lSd3 ' W W (9 ,su!e~6 1.101 u! paNasqo aleqdsoqd ON 'auaxolLd %66 (p 'luno3 u!e~6 Lq auaxolhd %09 '~I!T~II!~M %OP (E ' ~ ~ 9 1 ( 0 ~ 6 ~ ) I I ' d ~ ' 30~d "le la ( 1 '6/dls 33 ,,,01 JO syun u! 'ax :wdd u! uo!inl!p adolos! Aq suo!ie1lua3uo3 (L

- 2 '1 9.c 9.9 LL.O Z.LL P'LL ZEC 1~0 .0 02 ,d > E.E 82 E'S S'9L 081 08L 9901 81.0 LEZ eE'E> d ) 8.2

- p a x u o l s s ! ~ neE2 - V I - - - - - 826 86.0 OE9 d ' Z > d > S.2

a - '1!4M 10) 9'8L=n/41 '1(16!aM - - - - - 9Eb OL'O OLb 19'2 > d L I 1 I f I I I i , I 1 I I I . I I I . s I aldwes al!y3011!4M 10 asnwaq u!elJa3un (1 - - - - 9'12 EEL LEO'O EL alelo1

, - ! t o 96'1 QZS'O 98b'O

' p'u 90'0 90'0 'P'U 6'82 6'8

ZP'O 2'92 - 'D'U

20.0 EE'O 10.0 9'PE -

'p'u 'p'u sozd 'P'u 20'0> OzeN

PO'O 'P'u O!N 'P'U 'P'U ouz

62'1 b Oe3 E'PZ 9'OL 0 6 ~

09'0 90'0 OUW E'8E L'b 084 - 68'2 .Eozaj .P'U ' F U ' 0 ' ~

EO'O 12'0 '0'13 20'0 O'OL ' Q ~ V 90'0 9L'Z O!l

E'9E 6'9P 'O!S

2 '1

PROGRESS BY THE CONSORTS

ADORABLES

The 1650°C reextracts contained 3.5% and 2% additional Iz6Xe from 3.3 sinks and 2.8 sinks respectively. Xe in both samples is a complex mixture of S, T and F. The position of the data on Fig. 2b along the line joining XeT and ''"u fission Xe indicates large amounts of '"Pu fission Xe in both pyroxene and phosphate. The displacement of 2.8 sinks below the line is caused by substantial addition of XeS. The positions of the two data points in Fig. 2a show that 13%e in pyroxene is primarily 13'XeT and I3'XeF with minor spallation contributions. Spallation contributions to I3'Xe in whitlockite can be quite large and seriously affect partitioning of the heavy isotopes between T and F. Ba/Nd in the 2.8 and 3.3 sinks are similar to Ba/Nd in Apollo 11 high-and low-K rocks respectively (12), and we take XeS compositions in these rocks to determine spallation contributions ('3'XeS/'26XeS=0.9, p >3.3; '3LXeS/'L6XeS=0.15, 3.3> p > 2.8). Remaining 13=Xe is partitioned into T and F using known trapped and !"Pu fission Xe compositions. 66% of the total 136Xe in 3.3 sinks and 94% in 2.8 sinks is '"Pu fission lg6Xe. Blank Xe and pyroxene trapped Xe is sufficient to account for all XeT in the 2.8 sinks; all '36Xe in whitlockite is fission Xe. S, T and F contents are in Table 2. The large Ba enrichment in the whitlockite separate indicates whitlockite comprises only -0.4% of the total rock. From the chemical and isotopic data, it follows that the dominant part of the '"Pu, U and T h in ADOR are in the p > 3.3 separate and almost certainly in the pyroxene. This result on a magmatic rock cannot be directly compared to the more complex chondrites. From the Xe data on whitlockite and the stepwise heating measurements of Hohenberg (4) on a total rock (Fig. 2), which yield points much closer to pure 2 4 4 P ~ fission Xe, we conclude that there must exist a different phase (not pyroxene) rich in 2 4 4 P ~ and poor in REE and Ba. 244PuU/8U-0.006 in the total rock (4) as compared to 0.015 in the St.S&ve'rin chondrite (13). The highly fractionated nature of the parent magma and the cumulate nature of ADOR suggest the iow 244P~/238U most reasonably results from strong elemental fractionation. Since ADOR is a magmatic differentiate it could also be explained by a formation time of ADOR -0.1 AE later than chondrites. The large Xe S contents in both separates and the x4 enrichment in Ba relative to Nd in whitlockite (Table 7) allow a unique determination of XeS from Ba and from REE in a meteorite. Using Nd as a measure for total REE, the relative production is P126(Nd)/P126(Ba) =2.0. The production rate of lL6XeS over the 63 x 106yr cosmic ray exposure age of Angra dos Reis is P126=0.21 x 10 -I2 [Ba + 2.0 Nd] cc STP/g/108 yr, with Ba and Nd in ppm. This result agrees well with the only similar result, on lunar rock 12013 ( 14). Pb-U-Th data are given in Table 5. The radiogenic 207Pb-206Pb ratios for the total meteorite confirm the value reported by Tatsumoto et al. (3). Our (Y =617 is three times larger than they report. These workers washed their samples with acid and also obtain discordant Pb-U-Th ages. The phosphate experiment was very successful and yielded highly radiogenic P b in this phase which is enrich- ed in Th/U. Ages by all Pb-U-Th methods were determined on ADOR. This is the first time that self- consistent Pb-U or Pb-Th ages have been determined on a meteorite. As is manifest from the existing data in the literature, Pb-Pb model ages are not sufficient to define a reliable age [cf. (15)]. All three in- dependent ages on the whitlockite and the total rock ages are in agreement within experimental errors and demonstrate the existence of planetary differentiates a t 4.55 AE within an error of "- 0.05 AE. The gap in time between this age and the P b model ages for the earth and the moon (-4.45 AE) urgently demands a solution. Sm-Nd measurements were carried out a t UCSD to establish the feasibility of determining an internal isochron. Two samples were analyzed-the p > 3.3 fraction and a single grain of whitlockite (-1.2 mg). Fractions of both separates were dissolved and aliquots of each sample spiked. Only the spiked aliquots have so far been analyzed. The Sm/Nd ratio in whitlockite is ex- tremely low (0.215) while the Nd concentration is - 6 0 0 ~ chondritic. The pyroxene exhibits Sm/Nd only 6% higher than found in total rock samples of eucrites. The spread in Sm/Nd between samples is 59%. The slope of the line determined by the data yields an apparent age of 4.42 AE. However, the calculated 143Nd/1"Nd is critically dependent on the assumed 142Nd/'44Nd ratio, and a reliable age can only be assess- ed after measurement of the unspiked aliquots.

Ref. (1) G. Tschermak and E. Ludwig, Tschermak's Min. Pet. Mitt. 8 (1887) 341; 9 (1888) 423. (2) D. A. Papanastassiou, PhD Thesis (1970). (3) M. Tatsumoto et al. Science 180 (1973) 1279. (4) C. M. Hohenberg, G & C. Acta 34 (1970) 185. (5) R. Hutchison, Nature 240 (1972) 58. (6) A. L. Albee and A. A. Chodos, Proc. Ap-11 (1970) 135. (7) F. Tera and G. J. Wasserburg, this vol. (8) F. Tera and G. J. Wasserburg, Anal. Chem 47 (1975) 2214. (9) C. C. Schnetzler and J. A. Philpotts, in Met. Research (1969) 207. (10) J. G. Arth and G. N. Hanson. G & C Acta 39 (1975) 325. (11) Grutzeck, et al. Geophys. Res. Lett. 1 (1974) 273. (12) J. C. Huneke et al., G & C Acta 36 (1972) 269. (13) F. A. Podosek, G & C Acta 36 (1972) 755. (14) Lunatic Asylum, EPSL 9 (1970) 137. (15) J. H. Chen and G. R. Tilton, G & C Acta 1976 (in press). Acknowledgment. This work was supported by NASA and NSF.

0 Lunar and Planetary Institute Provided by the NASA Astrophysics Data System