Ultramafic rocks Definition: Color Index > 90, i.e., less than 10% felsic minerals. Not to be confused with Ultrabasic Rocks which are rocks with <45 wt.% SiO 2 Classification: See attachment to Lab # 6 Some reasons for studyjng Ultramafic Rocks: ultramafic xenoliths are samples derived from the upper mantle underneath continents and oceans Ultramafic parts of ophiolites represent sub-oceanic mantle Ultramafic rocks are the source rocks that produce basalts on partial melting Types of Ultramafic Rocks Alpine ultramafics (basal parts of ophiolites) and samples dredged from oceanic fracture zones Xenoliths in alkalic basalts and kimberlites Ultramafic lavas - komatiites Ultramafic cumulates in Layered Mafic Intrusions Zoned Ultramafic intrusions, e.g., Duke Island, AK Spinel lherzolite xenolith
Ultramafic Xenoliths Transported to surface in alkali basalts and kimberlites: (+ megacrysts, macrocrysts) Cognate vs. accidental Questions: Which part of mantle do they come from? Is mantle homogeneous or heterogeneous and on what scale? What is average composition? What do trace elements and isotopic analyses tell us? Do xenoliths provide information on mantle flow? Terminology: Mantle depletion, mantle enrichment, mantle metasomatism, fertile mantle, infertile mantle, tectonite fabric Cpx Xenoliths from alkali basalts (oceanic and continental) Spinel lherzolite: [ol (Fo 89-92 )+ opx + cpx + spinel (green/brown)] ± minor mica, amphibole, apatite Average mode: ol 80 opx 10 cpx 8 sp 2 Type 1A: Spinel Lherzolite (sub types: dunite, harzburgite, pyroxenite, wehrlite), granular to sheared, ± tectonite fabric. Ol modes 98% Opx REE/chond 100 CPX 10 1 La NdSm Yb 14 Mantle array 0 ε Nd -20-40 0 ε Sr 80 Type 1A are the most abundant. They are depleted and appear to have been depleted for a long time (model age 1-3 Ga): similar to calculated MORB sources, implying depleted mantle is present below oceans and continents
Type 1B: petrographically and texturally identical to type 1A. However, significant differences in trace element compositions as shown by cpx analyses. Why analyze cpx? REE/chond 100 10 1 La NdSm CPX Yb 14 1B CPX 0 ε Nd Mantle array -20-40 0 ε Sr 80 Type 1B spinel lherzolites appear to show a contradiction. Cpx grains shows a LREEenriched pattern but the isotopic ratios lie in the depleted quadrant. Why is this a contradiction and what is the reason? Type 1B xenoliths are said to be enriched but the enrichment is not expressed petrographically, only chemically. Sometimes called latent metasomatism, presumably involving material added via a fluid phase, which does not precipitate new minerals. What inferences can be drawn regarding the timing and source of the enrichment? Enrichment appears to be recent since it has not produced significant changes in isotopic ratios Types 1A and 1B (metasomatized): Spinel lherzolites containing small amounts of mica, amphiboles, and other exotic minerals formed by an enrichment process that resulted in the growth of new minerals. Process is sometimes referred to as patent metasomatism. What is the medium of metasomatism and what are the chemical effects of this metasomatism? REE/chond 100 10 1 La NdSm CPX Yb 14 0 ε Nd 1B CPX metasomatized Mantle array -20-40 0 ε Sr 80 Isotopic values of metasomatized xenoliths are highly variable. Process of metasomatism may have taken place over a time interval. Cpx is commonly highly enriched in LREE.
Type 2 xenoliths: primarily olivine clinopyroxenites containing olivine, Fe-Ti-Al augite, opx ± spinel ± amphibole ± mica Commonly form veins in Type 1 xenoliths forming composite xenoliths Same composition as host alkalic basalt or basanite Trace elements and isotopes are consistent with crystallization from host basalt/basanite Flow cumulates? What do we mean by the term: tectonized? [100] Sample KH 77-7 Kilbourne Hole, NM equigranular-tabular [010] [001] F Plots show the orientation of olivine grains measured on a universal stage. Contours reflect density of measurements. F = foliation delineated by mineral shape anisotropy. Indicative of flow deformation at high temperature in the mantle, cf. seismic anisotropy Simplified from: G. Bussod MS Thesis University of Washington [100] σ 1 (010) [100] σ 1 [010] σ 1 = principal compressive stress At high T, under penetrative deformation, olivine slips along the (010) plane in the [100] direction (a), producing oriented grains with a shape anisotropy
Xenoliths from kimberlites A. Peridotite-pyroxenite association (<15% garnet: ~95% of all kimberlite xenoliths) Rock type Mineralogy Abundance (%) Dunite ol 0.3 Harzburgite ol-opx±sp 16 Lherzolite ol-opx-cpx±sp 14 Garnet lherzolite ol-opx-cpx-gar 43 Garnet harzburgite ol-opx-gar 18 Pyroxenite cpx-ol±gar 6 Average garnet lherzolite: ol 64 opx 27 cpx 3 gar 6 Other (usually minor) minerals include: phlogopite, amphibole, ilmenite, chromite, sulfides, graphite/diamond, rutile B. Eclogite association (~5% of all kimberlite xenoliths) Rock type Mineralogy Abundance (%) Eclogite gar-cpx 63 2-px eclogite gar-cpx-opx 2 Kyanite eclogite gar-cpx-ky 8 Corundum eclogite gar-cpx-cor 6 Qtz eclogite gar-cpx-qtz 18 Types of garnet lherzolites Mg-rich, cold granular Fo>91 Coarse Fe-rich, cold granular Fo<91 Metasomatized Cold course cold deformed MARID suite: phlogopite + K- richterite + ilmenite + rutile + Craton margin Cape Town Premier Kimberley Deformed Mg-rich: cold / hot Lesotho Fe-Ti-rich, hot sheared
Garnet lherzolite (image on left) Photography of slab of garnet lherzolite, Norway. Light olive green: olivine (~Fo 90 ); Gray: Opx; Bright green: Cpx; Red: pyrope-rich garnet From: Yoder (1976) Generation of basaltic magma. NAS ol Cpx Gnt Garnet lherzolite (image above) Opx 2 mm Photomicrograph (ppl) of hot deformed garnet lherzolite from Jagersfontein, RSA. The fine-grained matrix contains olivine neoblasts. Stripes running E-W are formed by chains of tiny opx neoblasts 1 cm From: Harte (1983) Continental basalts and mantle xenoliths. Shiva
Thermobarometry Determination of the T and P of formation of xenoliths is important because it provides key data points on the mantle geotherm and also enables us to reconstruct fossil geotherms A number of thermobarometers have been applied to xenoliths. In this class, we will discuss only the most widely used ones. 2000 1500 T(ºC) 1000 500 ol+opx +cpx+ plag Subsolidus phase equilibria in peridotite ol+opx+cpx ol+opx +cpx+ spinel 5 solidus ol+opx+cpx +garnet graphite 0 0 15 30 45 60 P (kb) 4 diamond 3 2 Al 2 O 3 in opx isopleths Spinel lherzolites cold course garnet lherzolites hot deformed garnet lherzolites Stability fields: Low pressure: ol+opx+cpx+plag Inter. pressure: ol+opx+cpx+sp High Pressure: ol+opx+cpx+gar The basic assemblage provides an initial estimate of P and T. Pressure is estimated by measuring the Al 2 O 3 content of opx and/or cpx. T determination discussed in a later slide.
Important reactions in the CaO-MgO-Al 2 O 3 -SiO 2 (simple peridotite) system 1. Boundary between plagioclase peridotite and spinel peridotite is defined by the reaction: plag olivine opx spinel cpx CaAl 2 Si 2 O 8 + 2Mg 2 SiO 4 Mg 2 Si 2 O 6 + MgAl 2 O 4 + CaMgSi 2 O 6 In the presence of abundant olivine, plagioclase become unstable at pressures between 9 and 10 kb and breaks down to spinel, etc. Should plagioclase peridotite be common in nature? 2. Boundary between spinel peridotite and garnet peridotite is defined by the reactions: opx spinel olivine pyrope garnet 2Mg 2 Si 2 O 6 + MgAl 2 O 4 Mg 2 SiO 4 + Mg 3 Al 2 Si 3 O 12 cpx opx spinel Py-Gr garnet olivine CaMgSi 2 O 6 + Mg 2 Si 2 O 6 + MgAl 2 O 4 CaMg 2 Al 3 Si 3 O 12 + Mg 2 SiO 4 Garnet becomes stable at the expense of spinel as P increases. In the reactions written above, the V is negative, i.e., the RHS of the equation is favored with increasing pressure 3. Within the garnet peridotite stability field, reactions of the following type occur: opx opx garnet Mg 2 Si 2 O 6 + MgAl 2 SiO 6 Mg 3 Al 2 Si 3 O 12 cpx cpx garnet opx 2CaMgSi 2 O 6 + CaAl 2 SiO 6 Ca 3 Al 2 Si 3 O 12 + Mg 2 Si 2 O 6 Phase rule? In these two reactions the RHS is favored with increasing P, meaning that Al is forced out of the pyroxene and sequestered in the garnet. The Al 2 O 3 content of the pyroxene has been experimentally calibrated as a function of pressure (and temperature).
En (Mg 2 Si 2 O 6 ) - Di (CaMgSi 2 O 6 ) at P = 20 kb (2 GPa) 1800 a b 1600 T(ºC) c 1400 En SS En + Pig L+En SS Pig SS En SS + Di SS L+Pig SS Pig+Di SS L L+Di SS Di SS At pressures above ~2-3 kb, the pyroxene (En-Di) join appears to be binary. At lower pressures, Fo is the liquidus phase. Note the lower stability limit of pigeonite. Note: 3 isobaric equilibria represented by the reactions: 1200 En 20 40 60 80 Wt. % Two-pyroxene geothermometer: a: En ss + L Pig ss b: Pig ss + L Di ss c: Pig En ss + Di ss Widely used in natural assemblages containing orthopyroxene and clinopyroxene Example: Cpx with composition Di 70 En 30 coexisting with Opx of composition En 96 Di 4 would represent a pyroxene pair formed at 1350ºC. What are the limitations in the applications of this geothermometer to natural assemblages? Naturally occurring pyroxenes contain other components, particularly Fe, but also Na, Al, Ti... The location of the solvus should be experimentally determined at other pressures and comp s Essential to have 2 coexisting pyroxenes (usually opx and cpx). Di
Two pyroxene geothermometry (cont.) The figure above shows the pyroxene quadrilateral after Sack and Ghiorso (1994) CMP, 116, 287 (at 1 bar and 15 kb). Figures for other pressures are in the paper cited. If P is known, the compositions of coexisting pyroxenes can be normalized to quad. end members and plotted. The temperature can be read off as shown in the example. Note: The Di ss limb of the solvus is much more precise. The figure above is from Lindsley et al. (1981) Thermodynamics of minerals and melts Springer. It shows computed solvi based on experimental data and thermodynamic calculations. Use QUILF for calcs.