PHOSPHATE MINERALS IN SEMARKONA (LL3.0). J.S. Goreva and D.S. Lauretta, Lunar and PlanetaryLaboratory, University of Arizona, Tucson AZ 85721, jgoreva@lpl.arizona.edu

Introduction. Direct observations of phosphatesin unequilibrated ordinary chondrites (UOC) indicatethat prior to metamorphism, phosphorus wasassociated with the metal phase (e.g. [1]-[3]). Thealloying of P in Fe-Ni metal is a predictedconsequence of calculations of gas-solid equilibriumunder solar nebular conditions (e.g. [4]). Parent bodyconditions appear to have been sufficiently oxidizingthat phosphates formed early even in the mostprimitive chondrites. Presumably, P reacts with anexternal source of Ca and O on the metal grainboundaries to form Ca-phosphates [2]. Both apatiteand merrillite are observed in UOC. However, theconditions and timing of the phosphate nucleation isnot constrained.

The distribution and chemistry of P-phases inUOCs is highly complex. Through the petrographicstudy and electron microprobe analyses in this workwe attempt to look at the details of the phosphatedistribution in the Semarkona (LL3.0) chondrite.

Figure 1. Phosphorus x-ray map (red pixels) overlaidon a). BSE image, b). Fe x-ray map of SemarkonaLL3.0 chondrite.

Results and Discussion. Figure 1a is an overlayof P-Ka x-ray map of the Semarkona USNM 1805-14 thin section on a BSE image. It is evident that

there are number of well-defined P-rich regions on inand outside chondrules. We will discuss the 3 mostcommon occurrences of phosphates in this section –Ca-phosphate rims around chondrules and chondrulefragments, Ca-phosphates in the mesostasis of type-IIchondrules, and P-rich phases associated withmagnetite in the matrix.Chondrule Rims. Thin Ca-phosphate rims surroundmany chondrules and chondrule fragments inSemarkona. These rims are not necessarily associatedwith metal phase, however, they do seem to be moreprominent around fragmented and/or plasticallydeformed FeO-rich chondrules (Figure 1b). Figure 2illustrates an example of such a rim. In this case Camost likely was supplied by a Ca-rich matrix (in thelower right corner of the image). The quantitativeanalysis of phases smaller than 2 mm did not producesatisfactory results, however larger fragments wereidentified as merrillite, Ca-phosphate with up to 7 wt.% of Fe, Mg and Na oxides. Cl content is at or belowdetection limit.

Figure 2. Merrillite in chondrule rims

Type II chondrules. P-distribution in type IIchondrules of Semarkona is very complex. Jones(1990) reported up to 0.17% P2O5 in olivine withintype IIA Semarkona chondrules. Grossman (2000)describes P-zoning in type II chondrules within type3 ordinary chondrites. Although such zoning was notobserved in the Semarkona section the latter authorstudied, our observations support Grossman’sinterpretation of the ability of phosphate-rich phasesto form not only at the chondrule boundaries duringthe parent body alteration or metamorphism but alsowithin microcrystalline mesostasis. Figure 3 showsmerrillite dendrites (similar in composition to thephosphate rims discussed above) within Si, Al, Ca,Fe, P – rich mesostasis.

Lunar and Planetary Science XXXV (2004) 2065.pdf

Figure 3. Merrillite in mesostasis of type IIchondrules.

Fe-rich assemblages. Third most commonoccurrence of P-rich phases is associated with thickFe-rich rims surrounding type-I chondrules as well asFe-rich assemblages within Semarkona Matrix. Theassemblage in Figure 4 consists of magnetite and Fe-Ni metal core with layers of Fe-rich silicates and Ni-rich sulfides. Phosphorus is dissolved in the olivine(up to 5 wt.% P2O5), and forms individual phosphatestoo small for a quantitative analyses. Although wedid not detect carbides, this assemblage is similar tothe carbide-magnetite assemblages CMAs describedin type 3 ordinary chondrites before. Taylor et al.(1981) attributed the formation of CMAs to gas-solidreactions in the solar nebula. Krot et al. (1997)argued that metal and troilite have been replaced bycarbides and magnetite in the parent body by low-temperature reaction with a C-H-O fluid.

Figure 4. Merrillite and P-rich silicates (olivine) inmagnetite assemblages within matrix.

Conclusions. The Semarkona matrix consists largelyof phyllosilicates (mainly smectite) indicating low-temperature (likely below 250oC, [7]) aqueous

alteration of the asteroid. Formation of thephosphates on the chondrule boundaries and withinmagnetite assemblages in the matrix can be attributedto this process.

However, mineralogy and chemistry of chondrulesdoes not seem to be affected by aqueous alteration(see [8] for review). Nevertheless phosphates formwithin type-II chondrule mesostasis as well on thechondrule boundary.

It is likely that those type-II chondrules that wereplastically deformed (presumably during theaccretion) were heated sufficiently to drive out partof the phosphorus and immediately deposit Ca-phosphates in the thin rims discussed above (Figure2). On the other hand, phosphorus from the residualmelts within chondrules could have formed animmiscible component within the feldspatic meltproducing Ca-phosphates as in Figure 3.Alternatively, phosphates crystallized during thereheating of mesostasis either prior or after to theaccretion.

Interestingly, all our microprobe analyses ofphosphate rims around chondrules and withinmagnetite assemblages within matrix consistentlycame up to ~92 wt. % totals. Such low totals mightbe explained by the presence of light elements, notmeasured during the analyses. Although in the thinsection studied we did not see carbides andcarbonates observed in Semarkona ([8] for review),presence of carbonate for phosphate substitution aswell as OH could account for differences in chemicalanalyses. If holds true, this further supports Krot’s et.al. (1997) low-temperature origin of the magnetiteassemblages due to the reactions of UOC with C-H-O-bearing fluids. Further study is in progress. Phosphate minerals are of a primary importance inpertrogenesis of ordinary chondrites. They are majorsinks for the rare earth elements and actinides.However, petrographic observations to date could notdistinguish which Ca-phosphate phase (apatite ormerrillite) formed first. In this study did not find anyCl-apatites, so abundant in ordinary chondrites ofhigher metamorphic grades. This leads us to believethat merrillite nucleated first, however the question inwhat form, when and where Cl came from remainsunanswered.

[1] Murrel & Burnett (1985) GCA 47, 1999-2014, [2]Rubin & Grossman (1985) Meteoritics 20, 479, [3] Yabukiand El Goresy (1986) Proc. Tenth Symp. Ant. Met.,235–242, [4] Grossman & Olsen (1974) GCA 38, 173 [5]Taylor et al. (1981) LPS XII, 1076-1078 [6] Krot et al.(1997) GCA 61, 219-237 [7] Alexander et al (1989) EPSL95, 187-207, [8] Brearly & Jones (1998) in PlanetaryMaterials [9] Grossman (2000) LPS XXXI, 1599

Lunar and Planetary Science XXXV (2004) 2065.pdf