Elsevier

Chemical Geology

Volume 149, Issues 3–4, 31 July 1998, Pages 211-233
Chemical Geology

High precision Hf isotope measurements of MORB and OIB by thermal ionisation mass spectrometry: insights into the depleted mantle

https://doi.org/10.1016/S0009-2541(98)00036-9Get rights and content

Abstract

The existing Hf isotope database for mid-ocean ridge basalts (MORB) is limited in both quantity and precision. Nevertheless, in Hf–Nd isotope space, MORBs show a wide variation in 176Hf/177Hf over a relatively restricted range in 143Nd/144Nd. The highest 176Hf/177Hf ratios (≥0.283355) within the MORB range are restricted to just four samples (<6.5% of total). Of these high 176Hf/177Hf MORBs, three are from ridge segments adjacent to known active plumes and one is from a ridge segment for which a plume influence has been suggested. By comparison, MORBs from `normal' ridge segments show a more limited range in 176Hf/177Hf ratios (0.283040 to 0.283311). We suggest that NMORB and the depleted MORB mantle reservoir (DMM) are characterised by a similarly limited range in 176Hf/177Hf ratios. Furthermore, we suggest that the high 176Hf/177Hf MORB-like basalts may ultimately be related to mantle plumes and represent melts of a depleted component entrained by the plumes before they traverse the shallow MORB mantle. We illustrate our preferred model with new hafnium isotope data on 11 MORB samples from the Atlantic and Pacific Oceans, two oceanic gabbros from the Indian Ocean (all collected away from known plume localities) and basalts associated with the Iceland and Azores plumes obtained using a new high precision thermal ionisation mass spectrometry (TIMS) technique. The new TIMS technique routinely yields 176Hf beam intensities of 150–700 mV (total Hf beam of 2.8–13.5 V), allowing a routine internal precision on the measured 176Hf/177Hf ratio of 0.002–0.006%, to be achieved using just 1–3 μg of Hf separate. This represents a considerable improvement over the 0.008–0.056% internal precision quoted as typical for conventional single or triple filament TIMS analysis of Hf. The external reproducibility for the international Hf standard JMC 475 has also been significantly improved over conventional TIMS and is currently ∼0.002%. This is comparable with the 0.003% external reproducibility currently obtained on the new Fisons Instruments Plasma 54 at the Ecole Normale Superieure de Lyon.

Introduction

The correlation of 176Hf/177Hf and 143Nd/144Nd ratios in ocean island basalts (OIB) is well established (Fig. 1). In contrast, there appears to be no such correlation between Hf and Nd isotopes in mid-ocean ridge basalts (MORB), suggesting that Lu/Hf and Sm/Nd were decoupled during ancient depletion of the MORB source (Salters and Hart, 1991). This apparent decoupling of Hf and Nd isotopes is clearly of significance for earth evolution models and for the history of the depleted mantle.

Such processes as continental crust formation and mid-ocean ridge magmatism have resulted in a decrease, on average, in the Rb/Sr ratio of the residual, or depleted, mantle throughout geological time, whereas both Lu/Hf and Sm/Nd ratios have increased. As a result, the capacity for variation in 87Sr/86Sr in the depleted mantle due to in-growth of 87Sr by radioactive decay of 87Rb has also decreased. In contrast, the general increase in Lu/Hf and Sm/Nd in the depleted mantle residue means that small differences in the degree of source depletion, and hence parent/daughter ratio, will lead to a greater variation in both Nd and Hf isotopic composition with time. Furthermore, unlike the Sm–Nd isotope system, the fractionation of Lu relative to Hf is sensitive to variations in source mineralogy, specifically the presence or absence of garnet, during partial melting (Salters and Hart, 1991; Beard and Johnson, 1993) leading to a greater range in Lu/Hf than in Sm/Nd. Thus, the range of 176Hf/177Hf ratios in a variably depleted source should be correspondingly greater than for 143Sm/144Nd.

Despite the potential for studying processes of depletion in the mantle with the Lu–Hf isotope system, its application has been limited, as demonstrated by the relatively small number of Hf isotope ratios published for MORB (See Fig. 1 and references therein). The difficulties encountered in analysing Hf isotopes with existing mass spectrometric techniques have been the main cause for this lack of data. By thermal ionisation mass spectrometry (TIMS), these difficulties are largely twofold: (1) the ionisation of Hf is significantly inhibited by the presence of both Ti and Zr, the removal of which requires a complex three-column procedure, and (2) relatively large quantities (≥1 μg) of Hf are required for analysis because the high ionisation potential for Hf (7.09 eV) means that the efficiency of thermal ionisation is low (Patchett and Tatsumoto, 1980a; Walder et al., 1993b).

An improved thermal ionisation mass spectrometry technique is presented here which allows routine high-precision Hf isotope analysis of whole-rock samples. Although 1–3 μg of Hf separate is required, the greatly improved ionisation efficiency of the new procedure allows the 176Hf/177Hf ratio to be determined with a routine internal precision of between 0.002 and 0.006%. This represents a significant improvement over previous thermal ionisation techniques which achieved internal precisions of only between 0.008 and 0.056% on similar sample loads. The external reproducibility on the measured 176Hf/177Hf ratio for the international Hf standard JMC 475 is also considerably better than conventional TIMS, lying between 0.002 and 0.006% 2 SD. For whole-rock samples, the reproducibility is currently between 0.003 and 0.006% 2 SD.

This improved TIMS procedure has been used to obtain high precision Hf isotope data for Atlantic, Pacific and Indian MORB and, in conjunction with published Hf–Nd data for MORB, we argue for the existence of a depleted component distinct from DMM in Hf–Nd isotope space. This depleted component, despite resembling MORB in Sr and Nd isotopic compositions, is characterised by higher 176Hf/177Hf than MORB. With one possible exception, this radiogenic Hf component is spatially related to known active plumes and is further illustrated with Hf isotope data for plume-related basalt samples from Iceland, Reykjanes Ridge, SE Greenland margin (ODP Leg 152). We suggest that high 176Hf/177Hf ridge basalts, previously interpreted as having been derived from MORB source mantle, are in fact derived from a depleted component that is an intrinsic feature of certain mantle plumes.

Section snippets

Sample dissolution and separation

The dissolution procedure is adapted from the techniques of Salters and Hart (1991) and Barrovich et al. (1995). The sample is first dissolved in 15 ml of 29 M Hf (where M=mol/l) and 2 ml of 16 M HNO3 in an open bomb on a hotplate at ∼100°C. Once dry, another 15 ml of 29 M Hf is added to the sample; the bomb is sealed and placed in an oven at 120°C for a minimum of 4 days. After the dissolved sample solution has dried down, it is leached with 2 ml of 4 M Hf in an ultrasonic bath and then

Interlaboratory comparison of JMC 475

There is no real consensus on an actual 176Hf/177Hf ratio for the international Hf standard JMC 475 despite the increased interest in the application of Hf isotopes to geochemical problems, with the advent of plasma-source magnetic sector mass spectrometers. To date, sample data (see references in Appendix B) have been reported relative to JMC 475 176Hf/177Hf ratios ranging from 0.282142 (Schaltegger and Corfu, 1992), 0.282200 (Salters and Hart, 1991) and 0.282161 (Stille and Steiger, 1991).

Hf isotopes in MORB

176Hf/177Hf ratios for 13 new MOR samples are reported in Table 2. All the samples selected for this study come from mid-ocean ridge segments remote from known active plume localities—nine basalts from the East Pacific Rise (EPR; Mahoney et al., 1994), two basalts from MARNOK (Mid-Atlantic Ridge North of the Kane Fracture Zone; Lawson et al., 1996), and two oceanic gabbros from the Indian Ocean (ODP Leg 118, Hole 735B; Kempton et al., 1991). Fig. 4 shows the variation between Hf and Nd isotopic

Discussion

The Hf–Nd isotopic characteristics of MORB are significantly different from OIB. A positive correlation exists between εHf and εNd in ocean island basalts, i.e., εHf≈1.36εNd+3.13 (Johnson and Beard, 1993), as shown in Fig. 1. This positive correlation can be explained by the more compatible nature of the parent isotopes Lu and Sm relative to the daughter isotopes Hf and Nd during partial melting in the upper mantle. However, no such correlation between εHf and εNd is observed for MORB. Instead,

Conclusions

(1) The analysis of Hf isotopes by TIMS may always remain difficult and require large samples compared to magnetic sector plasma-source mass spectrometry (MS-ICP-MS). Nevertheless, our improved double filament TIMS technique shows that high internal precision and external reproducibility equal to that of MS-ICP-MS, can be achieved and that TIMS remains a viable alternative for whole-rock samples in the absence of MS-ICP-MS facilities.

(2) For inter-laboratory standardisation purposes, a JMC 475

Acknowledgements

Graham Pearson and Chris Ottley of Durham University are thanked for allowing GN to run numerous column calibration tests on their ICP-MS. We are grateful to Penny King who kindly provided Nd isotope data and samples from the Azores. We thank J. Vervoort and B. Beard for thorough and constructive reviews. The NERC Isotope Geosciences Laboratory Hf isotope facility was set up under NERC grant #GR9/1260A (PDK, SRN and Julian Pearce) while the work on MORB and the Iceland plume were funded by NERC

References (60)

  • V.J.M. Salters

    The generation of mid-ocean ridge basalts from the Hf and Nd isotope perspective

    Earth Planet. Sci. Lett.

    (1996)
  • V.J.M. Salters et al.

    The mantle source of ocean ridges, island arcs: the Hf-isotope connection

    Earth Planet. Sci. Lett.

    (1991)
  • V.J.M. Salters et al.

    Extreme 176Hf/177Hf in the sub-oceanic mantle

    Earth Planet. Sci. Lett.

    (1995)
  • R.K. Stevenson et al.

    Implications for the evolution of continental crust from Hf isotope systematics of Archean detrital zircons

    Geochim. Cosmochim. Acta

    (1990)
  • P. Stille et al.

    Pb, Sr, Nd, and Hf isotopic constraints on the origin of Hawaiian basalts and evidence for a unique mantle source

    Geochim. Cosmochim. Acta

    (1986)
  • P. Stille et al.

    Hf–Nd isotopic and trace element constraints on the genesis of alkaline and calc-alkaline lamprophyres

    Earth Planet. Sci. Lett.

    (1989)
  • R.N. Taylor et al.

    Isotopic constraints on the influence of the Icelandic plume

    Earth Planet. Sci. Lett.

    (1997)
  • M.F. Thirlwall

    Long-term reproducibility of multicollector Sr and Nd isotope ratio analysis

    Chem. Geol.

    (1991)
  • W.M. White et al.

    Hf–Nd–Sr isotopes and incompatible element abundances in island arcs: implications for magma origins and crust–mantle evolution

    Earth Planet. Sci. Lett.

    (1984)
  • K.M. Barrovich et al.

    Behaviour of isotopic systematics during deformation and metamorphism: an Hf, Nd and Sr isotopic study of mylonitized granite

    Contrib. Mineral. Petrol.

    (1992)
  • K.M. Barrovich et al.

    A chemical method for hafnium separation from high-ti whole-rock and zircon samples

    Chem. Geol.

    (1995)
  • J. Blichert-Toft et al.

    Separation of Hf and Lu for high-precision isotope analysis of rock samples by magnetic sector-multiple collector ICP-MS

    Contrib. Mineral. Petrol.

    (1997)
  • B. Duprè et al.

    Pb–Sr isotope variation in Indian Ocean basalts and mixing phenomena

    Nature

    (1983)
  • T.R. Elliot et al.

    Dynamic melting of the Iceland plume

    Nature

    (1991)
  • Fitton, J.G., Saunders, A.D., Larsen, L.M., Hardarson, B.S., Norry, M.J., 1998. Volcanic rocks from the East Greenland...
  • Fitton, J.G., Hardarson, B.S., Ellam, R.M., Rogerd, G., 1998. Sr-, Nd-, and Pb-isotopic composition of volcanic rocks...
  • B. Hamelin et al.

    Pb–Sr–Nd isotopic data of Indian Ocean ridges: new evidence of large-scale mapping of mantle heterogeneities

    Earth Planet. Sci. Lett.

    (1985)
  • V.L. Hards et al.

    The heterogeneous Iceland plume: new insights from the alkaline basalts of the Snaefell volcanic centre

    J. Geol. Soc. London

    (1995)
  • S.R. Hart

    A large-scale isotope anomaly in the Southern Hemisphere mantle

    Nature

    (1984)
  • S.R. Hart et al.

    Basalts from Iceland and along the Reykjanes Ridge: Sr isotope geochemistry

    Nature

    (1973)
  • Cited by (688)

    View all citing articles on Scopus
    View full text