Stillwater Anorthosites: A lunar analog?

Abstract

Samples of anorthosite from Stillwater Complex units AN‐I and AN‐II (terminology of McCallum et al (1980)) were analyzed by instrumental neutron activation analysis (INAA) for FeO, CaO, Na2O, REE, and other trace elements. No compositional trend with stratigraphic position was observed. Mean concentrations for most elements are about the same for three traverses (two for AN‐II) but CaO, FeO, and Sr concentrations are slightly different for the three traverses. The elements are not normally or lognormally distributed among samples from a single traverse. Strong interelement correlations can be understood in terms of sample mineralogy. Little evidence for compositional evolution of minerals during precipitation of the entire Middle Banded Zone is apparent. As in lunar anorthosites, the component of trapped residual liquid is relatively insignificant (less than 5%). In addition to plagioclase, which makes up, on average, 90% of the anorthosites, the rocks contain interstitial pyroxene (augite and inverted pigeonite) and, in some cases, quartz, titano‐magnetite, sulfides, and rare apatite and allanite. There is an inverse relationship between the amounts of pyroxene and quartz. General compositional trends, exemplified by La‐Sc systematics, indicate that the bulk of the poikilitic pyroxene appears to be cumulus‐adcumulus, presumably formed at the crystallization front from a magma saturated in plagioclase and pyroxene. A small fraction of the pyroxene and plagioclase plus all the quartz and other trace minerals formed from the minor trapped liquid component. The irregular distribution of pyroxene oikocrysts indicates that interstitial melt migrated substantially prior to final consolidation of the rock.

Chronology, geochemistry, and petrology of a ferroan noritic anorthosite clast from Descartes breccia 67215: Clues to the age, origin, structure, and impact history of the lunar crust

“…The Sm-Nd mineral isochron ages of 60025 (4.44 ± 0.02 Gyr), 67215c (4.40 ± 0.11 Gyr), and 67016c (4.54 ± 0.12 Gyr) are all consistent with this age within the error limits, while the nominal isochron age of 62236 is resolvably younger (4.29 ± 0.06 Gyr). The relatively large variation of Sm/Nd in the mafic fractions from these rocks would not be unusual for a cogenetic suite of igneous cumulates, as shown by the diverse LREE patterns observed in clinopyroxene separates from anorthosites in the Stillwater igneous complex (Salpas et al 1983) and may reflect variable proportions of cumulus phases and trapped liquid. The higher Mg# and more fractionated Sm/Nd in pyroxenes from 60025 and 62236 compared to 67215c and 67016c may be consistent with a greater proportion of cumulus pyroxene in the former.…”

Platinum-group elements in the Boulder Bed, western Bushveld Complex, South Africa

“…This led them to suggest a poorly defined process of ''insitu secretion'' for the origin of the boulders. One of the characteristics of many anorthosites in the Bushveld (and the Stillwater Complex, Salpas et al 1983) is their paucity in many incompatible trace elements (Maier and Eales 1997). This contrasts with the relative enrichment of these components in the boulders.…”

The parental magma for some rocks from the Norite 1 subzone of the Stillwater Complex: A lunar analog study

“…(Note, for instance, the complexities shown in the petrographic part of the present study in a comparison of orthopyroxenes from two successive layers: They have different compositions, yet both have subhedral orthopyroxenes that are almost unzoned in major elements but show different zoning in minor elements.) Furthermore, the ever-present negative Eu anomalies suggest a magma system dominated by low pressure fractional crystallization rather than repeated intrusions of separate magmas from a mantle source undergoing dynamic partial melting as suggested by Lambert [1982], A second set of comparisons can be made with the data of Salpas et al [1983] for rocks from the anorthosite zones. Parents calculated for the averages of Anorthosite 1 (AN-1), assuming three different minerals, and for a plagioclase separate from Anorthosite II (AN II PLAG) are shown in Figure 5.…”

“…5. Calculations of parent liquids from orthopyroxene separates of Lambert [1982] and anorthosites of Salpas et al [1983], using Weill and McKay [1975] orthopyroxene and plagioclase distribution coefficients, except that Eu is adjusted to be nonanomalous in the opx coefficients. If the Eu coefficient for orthopyroxene of Weill and McKay [1975] is used, the calculated Eu anomalies decrease slightly.…”

“…For the Lambert samples, % refers to the assumed amount of trapped magma; in the lower left (TAZ) box, all were assumed free of trapped magma. For the Salpas et al [1983] samples, the numbers refer to percentages of plagioclase, clinopyroxene, and trapped magma, respectively; the shaded band is limited by trappedmagma-free and 5%-trapped-magma calculations, as shown. Calculations for anorthosite S-l (this study) using the same plagioclase coefficients are shown for comparison.…”