Elsevier

Chemical Geology

Volume 199, Issues 3–4, 15 September 2003, Pages 317-329
Chemical Geology

Small-scale Sr isotope investigation of clinopyroxenes from peridotite xenoliths by laser ablation MC-ICP-MS—implications for mantle metasomatism

https://doi.org/10.1016/S0009-2541(03)00125-6Get rights and content

Abstract

Although numerous studies have shown that mantle xenoliths are isotopically heterogeneous on a hand specimen scale, experimental data imply that isotopic equilibrium should be attained within single mineral grains in the upper mantle. Until recently, this prediction has not been verified because of the difficulty of analyzing individual grains. We report in situ 87Sr/86Sr compositions of clinopyroxenes in garnet peridotite xenoliths from the Nikos kimberlite, Somerset Island (Arctic Canada) obtained by laser ablation MC-ICP-MS. Results for five different peridotites indicate the existence of large Sr isotope variations within individual xenolith samples, varying from 0.5‰ to as much as 1.1‰ (3 to 8 parts in 7000). This study is the first to document isotopic heterogeneity in peridotite xenoliths at the scale of individual grains. Multiple analyses of the same grain, however, indicate intra-grain Sr isotopic equilibrium.

The Sr isotopic ratios for individual clinopyroxene grains correlate with major element abundances such as SiO2 and TiO2, and trend towards the composition of the host Nikos kimberlite. It thus appears likely that the Sr isotope heterogeneities recorded by the clinopyroxenes are the result of metasomatic interaction with the host kimberlite magma. The preservation of Sr isotopic variability in clinopyroxene separates from different peridotite samples and between clinopyroxene grains of individual xenoliths suggests that metasomatism occurred just prior to or during kimberlite transport. The Sr isotope heterogeneities documented here for a single peridotite phase suggest that isotopic compositions obtained on mineral separates from peridotite xenoliths likely represent weighted averages and could be less homogenous than previously assumed.

Introduction

Radiogenic isotope data from mid-ocean ridge basalts (MORBs) and ocean island basalts (OIBs) indicate that the Earth's mantle is heterogeneous in isotopic composition (see Allègre et al., 1986, Zindler and Hart, 1986, Hart et al., 1992, Hofmann, 1997). Similarly, Nd, Pb, Os and Sr isotope data for peridotite xenoliths hosted by kimberlites and alkali basalts indicate a large variation in isotope compositions for the subcontinental mantle (e.g. Menzies and Murthy, 1980, Kramers et al., 1983, Walker et al., 1989, Hawkesworth et al., 1990). This heterogeneity appears to be the result of multiple episodes of metasomatic activity (e.g. melt or fluid interaction), and has been documented in many peridotite xenolith suites from cratonic areas (e.g. Yaxley et al., 1991, Rudnick et al., 1993). U–Pb ages of metasomatized harzburgite (Kinny and Dawson, 1992) and the preservation of oxygen isotope disequilibrium amongst constituent peridotite minerals (Zhang et al., 2000) have been interpreted to indicate that the timing of mantle metasomatism and kimberlite emplacement may essentially be contemporaneous.

Hofmann and Hart (1978) have argued, on the basis of limited experimental diffusion data, that the mantle is heterogeneous on a regional scale, but homogeneous on the scale of individual grains. Sneering et al. (1984) subsequently reported that at near solidus temperatures, Sr diffusion in peridotite clinopyroxene is sufficiently rapid to easily maintain isotopic equilibrium between grains. Van Orman et al. (2001) have recently proposed that 143Nd/144Nd isotopic equilibrium in clinopyroxene may be achieved in ∼1 Ma at near solidus conditions in peridotite (i.e. ∼1450 °C; 25 kb). However, the same experimental study also indicates that the time required for isotopic equilibration may increase to ∼1000 Ma at lower temperatures and pressures (i.e. ∼1150 °C; 15 kb) of the lithospheric mantle. Confirmation of the implications of these experimental results can only be accomplished by in situ isotopic investigations of mantle mineral(s) in peridotite.

Over the last decade, the introduction of multicollector inductively coupled plasma mass spectrometry (MC-ICP-MS) instruments has provided the ability to accurately measure a larger variety of stable and radiogenic isotope systems. This ability is attributable to the high ion transmission and flexibility of sample introduction in such instruments, the simultaneous detection of ion signals using analogue detectors and flat-top peaks. These features have combined to generate an instrument capable of performing isotope measurements with precisions comparable to those obtained by thermal ionization mass spectrometry (TIMS). Moreover, the coupling of a laser ablation system to MC-ICP-MS instruments has added the potential of obtaining spatially resolved, high precision in situ measurements for a variety of radiogenic isotope systems, including 87Sr/86Sr (Jackson et al., 2001 and references therein). Christensen et al. (1995) first reported the application of laser ablation MC-ICP-MS to Sr isotope studies, obtaining precise and accurate measurements on modern marine gastropods and plagioclase. These analytical data (long ablation times of 13–17 min) documented isotopic disequilibrium between two plagioclase grains from Long Valley basalts that has been interpreted as evidence of magma mixing. More recent laser ablation studies (ablation times of 30–60 s) have investigated the isotope zonation in single plagioclase (e.g. Davidson et al., 2001), and magmatic carbonate and apatite (Bizzarro et al., 2003).

Here we present in situ 87Sr/86Sr analyses obtained by laser ablation MC-ICP-MS on clinopyroxene grains in peridotite xenoliths from the Nikos kimberlite, Somerset Island (Arctic Canada). These results are compared to those obtained by TIMS and solution mode MC-ICP-MS, which both represent weighted average values for a much larger population of clinopyroxene grains from the same rocks. The data document Sr isotopic variability between clinopyroxene grains within the same mantle xenolith, and provide information on the possible interaction with the host rock during kimberlite magmatism. Previous studies using bulk mineral separates have documented isotopic disequilibrium between different mineral phases in peridotite (McDonough and McCulloch, 1987) and clinopyroxene phenocrysts in alkaline magmas (Simonetti and Bell, 1993). To our knowledge, this is the first study to document small-scale Sr isotope variations in individual clinopyroxene grains that have equilibrated at great depths (>100 km) in the Earth's mantle.

The mantle xenoliths of this study were collected from the Nikos kimberlite, which was emplaced into the Archean crystalline basement and Paleozoic cover sequences of Somerset Island on the northern Canadian craton (73°28′N and 90°58′W; Frisch and Hunt, 1993, Pell, 1993). Magmatism occurred between 90 and 105 Ma ago, as indicated by U–Pb dates for perovskites from Somerset Island kimberlites Heaman, 1989, Smith et al., 1989 and a Rb–Sr whole-rock isochron obtained for the Nikos kimberlite (97±17 Ma, MSWD=1.1; Schmidberger et al., 2001). The Nikos mantle xenolith suite is dominated by well-preserved garnet peridotites that contain large crystals (1–10 mm) of olivine, orthopyroxene, clinopyroxene and garnet. Trace amounts (<1%) of phlogopite are present in a small number of peridotites, but amphibole is absent. Estimates of the temperature and pressure of last equilibration (800–1400 °C, 25–60 kb) indicate that the Nikos peridotites were sampled from depths of 80 to 190 km by their host kimberlite (Schmidberger and Francis, 1999). The high magnesium numbers (mg number=Mg/(Mg+Fe)=0.90–0.93) and olivine-rich mineralogy (avg. 80 wt.%) of the peridotites reflect their refractory nature and suggest that these xenoliths are the residues of large degrees of partial melting Schmidberger and Francis, 1999, Schmidberger and Francis, 2001. In contrast to their depleted major element chemistry, however, the peridotite xenoliths are enriched in highly incompatible trace elements (e.g. large ion lithophile elements, LILE; light rare earth elements, LREE) relative to estimates for primitive mantle and chondrites (McDonough and Sun, 1995) as the result of metasomatism (Schmidberger and Francis, 2001). A detailed description of the mineralogy and major and trace element compositions of these mantle xenoliths can be found in Schmidberger and Francis (1999) and Schmidberger and Francis (2001).

The emerald green clinopyroxenes of the Nikos peridotites are chromian diopsides (Schmidberger and Francis, 1999). The incompatible trace element compositions of clinopyroxene correlate with temperature and depth estimates of last equilibration and suggest that the peridotites can be divided into low-temperature (<1100 °C; 80–150 km) and high-temperature xenoliths (>1100 °C; 160–190 km). The clinopyroxenes of the low-temperature peridotites are characterized by high Sr (100–400 ppm) and LREE (e.g. Ce=20–50 ppm) abundances compared to the clinopyroxenes of the high-temperature peridotites (Sr=50–100 ppm; Ce=2–8 ppm; Schmidberger and Francis, 2001). The present study is restricted to clinopyroxene in the low-temperature peridotites because their significantly higher Sr contents facilitated in situ analysis.

The 87Sr/86Sr values (0.7038–0.7046) obtained by TIMS for clinopyroxene separates from the low-temperature peridotites are significantly less radiogenic than their corresponding whole rock compositions (87Sr/86Sr(0.1Ga)=0.7054–0.7066). The Sr isotopic signatures of these clinopyroxenes indicate derivation from a time-integrated depleted mantle (Schmidberger et al., 2001). 143Nd/144Nd(0.1Ga) ratios (0.51256–0.51268) obtained by TIMS on these clinopyroxene separates show little variation and are depleted relative to a chondrite reservoir at the time of kimberlite magmatism.

Section snippets

Analytical methods

Inclusion-free clinopyroxene grains were hand-picked from the low-temperature peridotites for laser ablation analyses and then washed in an ultrasonic bath with acetone, ultrapure water and 2.5 N HCl. The grains were subsequently embedded into epoxy mounts (25 mm diameter) and surfaces were polished to 0.3 μm. In most cases, only one analysis per grain was possible, but two separate analyses were obtained on grains larger than 0.5 mm (see Table 1; Fig. 1).

In situ Sr isotope analyses were

Results

Laser ablation MC-ICP-MS analyses were obtained on 46 grains of clinopyroxene from five low-temperature Nikos peridotites (total of 56 analyses; Table 1; Fig. 3). The results indicate a large variation in the 87Sr/86Sr values within individual samples, with differences ranging from 0.0003 (0.5‰, NK1-4) to 0.0008 (1.1‰, NK2-3; Fig. 3). These intra-sample variations are much larger than the average 2σ internal precision of individual laser analyses (Table 1), and the estimated external

Discussion

Experimental data on Sr diffusion in diopside have shown that isotopic equilibration at the mineral (mm) scale should occur over time periods of 1 to 100 Ma at lithospheric temperatures of 900 to 1000°C (Sneeringer et al., 1984). This is in agreement with previous findings based on diffusion data in mantle minerals, which indicate that isotopic disequilibrium between mineral phases cannot be maintained at 1000 °C for more than a few million years (Hofmann and Hart, 1978). The experimental

Conclusions

This study is the first to report in situ 87Sr/86Sr analyses of individual clinopyroxene grains from peridotite xenoliths. The results demonstrate the ability of a laser ablation MC-ICP-MS system to perform in situ isotope analyses on mantle phases. The Sr isotopic results document large variations in clinopyroxene separates from different samples, and between clinopyroxene grains within individual xenolith samples, which amount to as much as 1.1‰ (8 parts in 7000). The Sr isotopic ratios

Acknowledgments

We thank Joel Visser (Merchantek-New Wave Research) for providing the ablation cell and for technical support. The MC-ICP-MS laboratory at GEOTOP is maintained in part with an NSERC MFA grant, and the research was funded by an NSERC operating grant to D. Francis. S. Schmidberger is grateful for financial support in the form of an NSERC doctoral scholarship. [RR]

References (37)

Cited by (68)

View all citing articles on Scopus
View full text