Contrib Mineral Petrol (2009) 158:675–681 DOI 10.1007/s00410-009-0403-8 ORIGINAL PAPER Experimental boron isotope fractionation between tourmaline and fluid: confirmation from in situ analyses by secondary ion mass spectrometry and from Rayleigh fractionation modelling Horst R. Marschall Æ Christian Meyer Æ Bernd Wunder Æ Thomas Ludwig Æ Wilhelm Heinrich Received: 27 January 2009 / Accepted: 3 April 2009 / Published online: 24 April 2009 Ó Springer-Verlag 2009 Abstract Tourmaline synthesised in an experiment with low boron excess was analysed in situ by secondary ion mass spectrometry. It revealed significant B isotope zonation with 11B/10B ratios increasing in the growth direction of the crystals. Trend, magnitude and absolute values strongly support results from high-B-excess isotope fractionation experiments. Furthermore, the closed system B-isotopic evolution of the experimental fluid was modelled by Rayleigh fractionation. The model results are in excellent agreement with the measured B-isotope composition of the run-product fluid. Consequently, low-elementexcess experiments are proposed as an ideal approach to determine fluid-solid isotope fractionation factors for systems that are characterised by Rayleigh fractionation. Keywords Tourmaline Experimental geochemistry Boron isotopes Fluid Ionprobe Communicated by J. Hoefs. H. R. Marschall (&) Department of Earth Sciences, University of Bristol, Wills Memorial Building, Queen’s Road, Bristol BS8 1RJ, UK e-mail: [email protected] C. Meyer B. Wunder W. Heinrich Department 3, Section 3.3, GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany C. Meyer Institut für Geologische Wissenschaften, Freie Universität Berlin, Malteserstrasse 74-100, 12249 Berlin, Germany T. Ludwig Institut für Geowissenschaften, Universität Heidelberg, Im Neuenheimer Feld 236, 69120 Heidelberg, Germany Introduction Stable isotope systems of light elements, such as Li and B have been widely applied in Earth sciences to unravel processes operating at the surface, as well as in the deeper crust and the mantle (e.g., Leeman and Sisson 2002; Palmer and Swihart 2002; Tomascak 2004). The establishment of these geochemical tools has progressed via two different routes, namely the investigation of natural samples from well-characterised settings and elaborate geochemical modelling. However, both require profound knowledge on equilibrium isotope fractionation among the various phases of interest, such as rock-forming or accessory minerals, fluids and melts. The sole method of acquisition of such high-quality isotope fractionation data is to conduct wellcontrolled experiments, where the phases of interest are equilibrated at a range of temperatures and pressures. Run products recovered from such experiments are analysed separately for their isotopic compositions, from which the T–(P)-dependent isotope fractionation is derived. Recently, a number of such experiments have been conducted using hydrothermal apparatuses to determine equilibrium isotope fractionation between minerals and hydrous fluids for Li (e.g., Wunder et al. 2006, 2007) and B (e.g., Wunder et al. 2005; Meyer et al. 2008). Generally, these experiments are conducted with an ample excess of B (or Li), so that the minerals produced during the experiment do not significantly deplete the fluid in the element of interest. Hence, the synthesised mineral grains grow in contact with a fluid of constant composition and are isotopically unzoned. Fluids and solids recovered from such an experiment, therefore, represent equilibrium isotope fractionation at P and T. In practise, they are independently analysed to obtain isotope fractionation factors. Although this method has been successfully applied 123 676 to a number of fluid-mineral systems, it has the disadvantage of combining two different analytical procedures, i.e., the analyses of solid and liquid run products, respectively. High accuracy on both analytical protocols is, therefore, fundamental for the extraction of data, and limits the accuracy of the resulting isotope fractionation factor. Recently, Meyer et al. (2008) have completed experiments on B isotope fractionation between tourmaline and fluid. While the majority of the experiments were run employing the established method of using a vast excess of B (i.e., 900 mole%), some of the experiments were conducted with a limited B excess of only 10 mole%. Boron, as a major component of tourmaline (Tur *3.3 wt%) is largely removed from the fluid in the course of the experiment. Solid state diffusion of B in Tur is negligible for the experimental run time (Henry and Dutrow 2002), thus no re-equilibration between fluid and early grown Tur will occur. Under these conditions, any B isotopic fractionation between Tur and fluid should produce isotopically zoned crystals and a final fluid that is isotopically different from the initial composition. Actually, Meyer et al. (2008) found the B in the final fluid to be significantly heavier than in the starting composition. They also recovered Tur separates of different grain sizes from the experiment, which had a distinct variability in d11B values [d11 B ¼ ð11 B=10 Bsample =11 B=10 BNBS951 1Þ1;000]. In this contribution, we present in situ analyses of Tur crystals synthesised in a low-excess-B experiment, demonstrating that they are, indeed, isotopically zoned in accordance with assumptions drawn from the bulk analyses (Meyer et al. 2008). In addition, we employ a Rayleigh fractionation model to predict the B isotopic composition of the fluid produced by the experiment with limited excess B. The in situ analyses and the model independently verify the B isotopic fractionation between Tur and fluid as quantified by Meyer et al. (2008) in both trend and magnitude. Experimental methods Tourmaline CM2, analysed in this study was taken from a series of experiments on B-isotope fractionation between Tur and fluid, described in Meyer et al. (2008). Synthesis CM2 produced dravitic Tur grown hydrothermally at 600°C and 200 MPa in the system Na2O–MgO–Al2O3– SiO2– B2O3–H2O (NMASBH). Tourmaline was crystallised directly from the fluid, applying the method of Wunder et al. (2005). The two-chamber Au-capsule method was employed (von Goerne et al. 2001) with starting materials added as oxide mixtures of 25 mg MgO, SiO2, c–Al2O3, as well as solid boric acid (H3BO3) and 45 mg of a NaCl–H2O solution. Excess Si (20 mole%) and 123 Contrib Mineral Petrol (2009) 158:675–681 Na (100 mole%) with respect to the ideal dravite composition were chosen to compensate for high fluid solubilities of these components. Excess B was only 10 mole% in order to monitor B-isotopic evolution of the fluid, as expected from fractional crystallisation of Tur over the course of the experiment. This translates to a mass ratio of B between run-product fluid and Tur of only 0.1. The B isotopic composition of the boric acid in the starting material was d11B = -13.2 ± 0.4%. Applying the twochamber method, SiO2 was mechanically separated from the other compounds within the Au-capsule. Cooling of the hydrothermal cold-seal vessels below 200°C within 3 min was achieved using compressed air. Chemistry and crystal structure of the Tur run products have been determined by X-ray powder diffraction, scanning electron microscopy and electron probe micro analyser. Cell parameters and chemical composition are given in Meyer et al. (2008). The Tur’s chemical composition comprises a solid solution between dravite (62%) and Mg-foitite (35%), and no evidence was found for tetrahedrally coordinated B in the structure refinement. After a run time of 11 days, the fluid extracted from the experiment by piercing and washing the capsule with 80°C hot double-distilled water, had a B-isotopic composition of d11B = -10.0 ± 0.3% (Meyer et al. 2008). Solid by products of Tur synthesis were quartz and traces of talc, which are both known to have very low B partition coefficients towards hydrous fluid (*10-2 to 10-3; Marschall et al. 2006a) and can, hence, be neglected for the B budget of the experiment. Analytical methods Boron isotope ratios of Tur were analysed using a Cameca IMS 3f secondary ion mass spectrometer (SIMS) at the Institut für Geowissenschaften, Universität Heidelberg. The primary ion beam was 16O- accelerated to 10 keV with a beam current of 1 nA, resulting in count rates for 11 B of *2 9 105s-1 and *5 9 104s-1 for 10B on Tur, collected by a single electron multiplier. Diameter of the 1 nA spot was 5–10 lm. The energy window was set to 100 eV and no offset was applied. Fifty cycles were measured on each analysis spot with counting times of 3.32 s and 1.66 s on 10B and 11B, respectively. For each cycle, the 10B signal was integrated for 1.66 s before and after collection of the 11B signal, thus minimising the influence of increasing or decreasing count rates on the measured isotope ratio. Count rates were corrected for the counting system’s dead time of 16 ns. Presputtering lasted for 5 min and settling time between two different masses was 200 ms, resulting in total analysis time for one spot of approximately 10 min. Internal precision of a single analysis was B0.5% (1r). Boron isotopic compositions of Contrib Mineral Petrol (2009) 158:675–681 samples are reported in delta notation (d11B in %) relative to the SRM 951 accepted value (Catanzaro et al. 1970). Instrumental mass fractionation was corrected by using three samples of proposed reference tourmaline (98114: elbaite, 108796: dravite, 112566: schorl; Leeman and Tonarini 2001), yielding a correction factor of 1.0464. Reproducibility of measured isotope ratios during an analytical session is typically ±0.5% (1r). For further details on precision and accuracy of B isotope analysis of the Heidelberg SIMS see e.g., Marschall et al. (2006b) and Rosner et al. (2008). Results Tourmaline synthesised in experiment CM2 shows sheaflike aggregates of euhedral crystals, of *50 lm width and 300–800 lm in length (Figs. 1, 2). The sheaves consist of anhedral, bulky roots with several large euhedral Tur crystals diverging radially away from the central roots. The B-isotope profile analysed by SIMS from root to tip of one large representative crystal (Fig. 2a, b) reveals a rather heterogeneous centre of the root with d11B values scattering between -15.3 and -12.5% (Fig. 3; Table 1). The free-standing crystal itself, however, displays a very systematic increase of d11B values from -15.3% in the root to -12.5% at the tip (Fig. 3; Table 1). In addition, several analyses on tips of other large euhedral crystals (Fig. 2c–f) were completed. These revealed consistently high d11B values between -12.2 and -11.0% (Fig. 3; Table 1). Analysis CM2-17 was probably conducted too close to the edge of the crystal and overlapped the epoxy resin (Fig. 2d), Fig. 1 Scanning electron microscope image of Tur sheaf synthesised in experiment CM2 (T = 600°C; P = 200 MPa; t = 11 days). Note the large euhedral Tur crystals diverging from the bulky root. Talc was produced as a minor by product of experiment CM2. 677 resulting in lower count rates and an unreasonably high d11B value of -6.2%, regarded as an artefact of the lowquality measurement and not further evaluated. Discussion In situ analyses by SIMS We interpret the roots of the Tur sheaves as the nucleation sites of the Tur, based on their petrographic appearance (Figs. 1,2). The tips of the radially diverging Tur crystals are interpreted to represent the last growth phase of Tur before the experiment was quenched. The increase in d11B from root to tip is consistent with the predicted isotopic fractionation between fluid and Tur, with the solid preferentially incorporating the light isotope, and consequently leaving the remaining fluid enriched in the heavy isotope. Subsequent growth zones of Tur, hence, display higher and higher d11B values in temporary equilibrium with the evolving fluid. The magnitude of B-isotope fractionation can be estimated from the difference between the lowest d11B values in the roots of the Tur sheaves (-15.3 to -14.1%) and the initial fluid (-13.2%), resulting in a value of D11 B ¼ d11 Bfluid d11 BTur ¼ 0:9 2:1& . This is in agreement with the value of D11 B ¼ 1:29& derived from B-excess experiments (Meyer et al. 2008). In addition, the difference in d11B values between the tips of the Tur crystals (-12.7 to -11.0%) and the runproduct fluid (-10.0%) results in a value of D11 B ¼ 1:0 2:7&. Again, this is in agreement with the previously established fractionation. It is a technical challenge to find and analyse the very initial Tur that crystallised in the experiment, as well as it is difficult to analyse growth zones that are very close to the edge. This may explain the rather large spread in fractionation data (D11B), and limits the potential of this method in quantification of the B isotope fractionation between the two phases. However, most important, the differences in d11B between roots and tips and the increase along the analysed crystal clearly proves that B isotope fractionation between Tur and hydrous fluid is indeed significant at 600°C, with Tur preferentially incorporating the light isotope 10B. Also, cutting effects may be expected from sections oblique to the crystallographic c axis of the Tur crystals and from sections not cutting the centre of the trigonal prisms. In such cases, one would still observe a systematic increase in B isotopes, but may not expose the earliest grown Tur, and/or may have cut off the tip of the crystals and lose the youngest and isotopically most evolved portions. Note, that SIMS (in contrast to Laser ablation) has the advantage of very high depth resolution. The progress of sputtering into the crystals is at the order of 0.5–1 nm/s, 123 678 Contrib Mineral Petrol (2009) 158:675–681 Fig. 2 Transmitted (a, c, e) and reflected (b, d, f) light microscope images of Tur sheaves synthesised in experiment CM2 (T = 600°C; P = 200 MPa; t = 11 days). SIMS spots for B isotope analyses (*10 lm diameter) are marked with black circles in transmitted-light images and are visible as black spots in reflected light. Number labels refer to d11B values. resulting in a total depth of the sputter crater of \0.5 lm after completion of a 10 min analysis. Consequently, SIMS analyses of Tur represent the portions of the crystals as they are exposed in reflected-light or back-scattered electron images, and effects of zonation in the third dimension can be neglected. The B isotopic heterogeneity of the root (Fig. 3) may be explained by initially large pore space in the roots at the start of the experiment that would be filled by Tur grown at a later stage during the run. Hence, the connection between geometry and growth history in the roots may be obscured, while the euhedral Tur crystals clearly show a steady increase in d11B from root to tip. zones. The fluid is, thus, expected to be progressively enriched in the heavier isotope 11B. The B isotopic fractionation between Tur and fluid in a closed system, such as our experimental capsule, can generally be modelled using a Rayleigh formulation: Rayleigh model where d11 Bi and d11B are the initial and the final d11B values, respectively. In the case of experiment CM2, the initial d11B value was -13.2 ± 0.4%, and the excess amount of B was 10 mole%, translating to a ratio of B in The growing Tur crystals incorporate significant portions of B without subsequent re-equilibration of earlier growth 123 R=Ri ¼ F a1 ð1Þ where R is the final 11B/10B ratio, Ri is the initial 11B/10B ratio, F is the fraction of B remaining in the fluid, and a is the temperature-dependent B-isotope fractionation factor. Replacing isotope ratios by delta values, Eq. 1 can be rewritten as: d11 B ¼ ð1;000 þ d11 Bi ÞF a1 1;000 ð2Þ Contrib Mineral Petrol (2009) 158:675–681 Fig. 3 Boron isotope profile of CM2 Tur analysed by SIMS. One large crystal was analysed from root to tip (Fig. 2a, b). In addition, one root and several tips from other Tur sheaves were analysed and are plotted to the left and right of the profile, respectively. The growth direction of Tur is indicated by the arrow. The starting material (d11B = -13.2 ± 0.4%; TIMS = thermal ionisation mass spectrometry) should produce Tur with d11B = -14.5% (fractionation indicated by arrow No. 1), while the final fluid (d11B = -10.0 ± 0.3%) should produce Tur with d11B = -11.3% (fractionation indicated by arrow No. 2) according to Meyer et al. (2008). Table 1 Boron isotope analyses of CM2 Tur determined by SIMS 679 -9 CM2 tourmaline δ11B analyses by SIMS (all errors 1σ) final fluid (TIMS) -10 2 -11 δ11B (‰) -12 -13 starting material (TIMS) 1 -14 -16 other root -15 growth direction root tip other tips -17 0 100 200 300 400 Distance (µm) Analysis ID Total count rate B ? 10B (kHz) d11B (%) 1RSDmean (%) Type 11 Distance from centre of root (lm) Profile crystal 1 (Fig. 2a, b) CM2-1 227 -14.54 0.48 root CM2-2 271 -12.69 0.42 tip 0 CM2-3 272 -12.55 0.43 tip CM2-4 CM2-5 274 285 -12.73 -12.45 0.47 0.38 424 402 CM2-6 284 -13.46 0.42 379 CM2-7 292 -13.89 0.38 345 CM2-8 271 -13.59 0.44 303 CM2-9 263 -14.23 0.35 263 CM2-10 247 -13.84 0.40 221 CM2-11 266 -13.73 0.44 CM2-12 252 -14.42 0.46 root 118 CM2-13 242 -12.52 0.46 root 66 CM2-14 256 -13.41 0.44 root 28 CM2-15 245 -15.31 0.41 root 93 464 445 170 Other crystals (Fig. 2c–f) 1RSDmean 1 9 relative standard deviation of the mean a Analysis of CM2-17 was probably overlapping the crystal’s edge (Fig. 2d), hence, this isotope ratio is regarded erroneous CM2-16 279 -12.15 0.41 tip CM2-17a 234 (-6.17a) 0.42 tip CM2-18 272 -11.71 0.37 tip CM2-19 CM2-20 278 265 -12.00 -14.06 0.43 0.46 tip root CM2-21 291 -11.75 0.44 tip CM2-22 295 -12.21 0.38 tip CM2-23 288 -11.01 0.41 tip the remaining fluid to B in Tur of 1:10, which can be expressed as F = 1/11 & 0.091. Hence, 9.1% of the initial B remained in the fluid at the end of the experiment. The equilibrium B-isotope fractionation factor a = 1.00129 for T = 600°C derived from B-excess experiments (Meyer et al. 2008), can be tested independently by 123 680 Contrib Mineral Petrol (2009) 158:675–681 10 mole% boron excess -4 ∆11B = -4.20 (1000/T (K)) + 3.52 -6 T = 600 °C (= 873.25 K) δ11B = (1000 + δ11B i ) F α-1 - 1000 δ11B (‰) -8 -10 first Tur M’08 (900%) -12 tou 0.9 0.8 rmaline 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 F (fraction of B remaining in the fluid) Fig. 4 Experiment CM2 with 10 mole% excess B modelled by Rayleigh fractionation. The B isotopic composition of the initial and final fluid and its evolution with progressive Tur growth (stippled curve) are displayed. In addition, the B isotopic composition of the growing Tur is displayed (dashed curve). Note, that the theoretical d11B value of -10.14% calculated from the temperature-dependent fractionation value of Meyer et al. (2008) is in very good agreement with the value of d11B = -10.0 ± 0.3% measured after the experiment. Experiments for the determination of the fractionation factor a completed by Meyer et al. (2008) had 900 mole% B in excess, which is marked by the broken line labeled ‘‘M’08’’ 123 a δ a X = (1000 + δ X i ) F α-1 12.1 ‰ - 1000 10 5.0 ‰ 05 1.0 0‰ 3. 2.0 initial fluid 4.8 ‰ ‰ 2.4 ‰ 1.0 ‰ 1.2 ‰ 0.5 ‰ 0 1.0 7.2 ‰ 4. 0 α 5 9.6 ‰ ‰ ∆aX 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 F (fraction of element X remaining in the fluid) Fig. 5 General Rayleigh fractionation model of the isotopic composition of an element X in a fluid with an initial isotopic composition daX = 0%, evolving with different isotopic fractionation factors a, ranging from 1.0005 to 1.0050 (corresponding to fluid-solid fractionations Da X of 0.5 to 5.0%). Isotopic compositions of the runproduct fluids of experiments with 10 mole% excess of element X are given on the right hand side accuracy of the method is limited by the uncertainty on F, which can be estimated from the initial composition (such as Al2O3/B2O3 ratio in the case of Tur synthesis), but can also be determined more accurately from the concentration of the element in the run-product fluid. The method may be applied to any liquid–solid system that satisfies the requirement of Rayleigh fractionation, i.e. where the element of interest is denied any re-equilibration with the liquid in as small increments as possible. Solids with very slow solid-state diffusion coefficients, such as Tur, will fulfil this requirement. Conclusions fluid -14 1.0 10 mole% excess of X final fluid (theoretical): δ 11B = -10.14 ‰ last Tur (theoretical): δ 11B = -11.43 ‰ initial fluid 15 δaX (‰) applying Eq. 2 to experiment CM2 and comparing the d11B value calculated for the final fluid with the one actually measured after the experiment. The B-isotope evolution modelled for the parameters of experiment CM2 predicts an increase of d11B of both Tur and fluid by *3%, and a d11B value of the final fluid of -10.14% (Fig. 4). This is in excellent agreement with the measured value of d11B = -10.0 ± 0.3%, thus supporting the accuracy of the fractionation factor established by Meyer et al. (2008). In general, this provides a new method of determining isotope fractionation factors between solids and fluids or melts, without analysing the solids. It only requires the determination of F, i.e. the level of excess of the element of interest. The fluid curve in Fig. 4 can be established from Eq. 2, and the isotope fractionation factor a can be derived, as it is the only unknown parameter (Fig. 5). The advantages of this method are (1) only the fluid phase has to be analysed, which reduces uncertainties introduced by limited accuracies, and (2) the isotope fractionation is amplified at low F values, thus reducing the error on the determination of a. This effect is displayed in Fig. 5: at 5% fractionation between solid and fluid (a = 1.005), the final fluid containing 10 mole% of the initial budget of the element will have a d value increased by 12.1%. The SIMS analyses of Tur synthesised with limited excess B (10 mole%) revealed a significant d11B zonation from relatively light B in the early grown roots to heavy B in the tips. The trend, magnitude and the absolute d11B values measured by SIMS verify the B-isotopic fractionation derived from high-B-excess experiments published recently by Meyer et al. (2008). Modelling the B-isotope fractionation in the experimental run by Rayleigh fractionation, reproduces the measured d11B value of the run-product fluid well within the margins of analytical uncertainty. This encourages the completion of Rayleigh-fractionation experiments for the determination of mineral-fluid isotope fractionation factors with better precision and accuracy compared to conventional equilibrium experiments. Contrib Mineral Petrol (2009) 158:675–681 Acknowledgments We are grateful to R. L. Romer for fruitful discussions. Comments from two anonymous reviewers are appreciated. This study was financially supported by a European Union Marie-Curie Fellowship awarded to HRM (ID 025844: ‘‘Isotopes in subduction zones—the metamorphic perspective’’), which is greatly acknowledged. References Catanzaro FJ, Champion CE, Garner EL, Marinenko G, Sappenfield KM, Shields WR (1970) Boric acid: isotopic and assay standard reference materials. NBS (US) Spec Publ 260:1–70 Henry DJ, Dutrow BL (2002) Metamorphic tourmaline and its petrologic applications. In: Grew ES, Anovitz LM (eds) Boron: mineralogy, petrology and geochemistry, vol. 33 of Rev. Mineral., 2nd edn, chap 10. Mineralogical Society of America, Washington DC, pp 503–557 Leeman WP, Sisson VB (2002) Geochemistry of boron and its implications for crustal and mantle processes. In: Grew ES, Anovitz LM (eds) Boron: mineralogy, petrology and geochemistry, vol 33, 2nd edn. Mineralogical Society of America, Washington DC, pp 645–708 Leeman WP, Tonarini S (2001) Boron isotopic analysis of proposed borosilicate mineral reference samples. Geostand Newsl 25:399– 403 Marschall HR, Altherr R, Ludwig T, Kalt A, Gméling K, Kasztovszky Zs (2006a) Partitioning and budget of Li, Be and B in highpressure metamorphic rocks. Geochim Cosmochim Acta 70:4750–4769 Marschall HR, Ludwig T, Altherr R, Kalt A, Tonarini S (2006b) Syros metasomatic tourmaline: evidence for very high-d11B fluids in subduction zones. J Petrol 47:1915–1942 Meyer C, Wunder B, Meixner A, Romer RL, Heinrich W (2008) Boron-isotope fractionation between tourmaline and fluid: an 681 experimental re-investigation. Contrib Mineral Petrol 156:259– 267 Palmer MR, Swihart GH (2002) Boron isotope geochemistry: an overview. In: Grew ES, Anovitz LM (eds) Boron: mineralogy, petrology and geochemistry, vol. 33 of Rev. Mineral., chap 13, 2nd edn. Mineralogical Society of America, Washington DC, pp 709–744 Rosner M, Wiedenbeck M, Ludwig T (2008) Composition-induced variations in SIMS instrumental mass fractionation during boron isotope ratio measurements of silicate glasses. Geostand Geoanal Res 32:27–38 Tomascak PB (2004) Developments in the understanding and application of lithium isotopes in the earth and planetary sciences. In: Johnson CM, Beard BL, Albarède F (eds) Geochemistry of non-traditional stable isotopes, vol. 55 of Rev. Mineral., chap 5. Mineralogical Society of America, Washington DC, pp 153–195 von Goerne G, Franz G, Heinrich W (2001) Synthesis of tourmaline solid solutions in the system Na2O–MgO–Al2O3–SiO2–B2O3– H2O–HCl and the distribtion of Na between tourmaline and fluid at 300 to 700°C and 200 MPa. Contrib Mineral Petrol 141:160– 173 Wunder B, Meixner A, Romer RL, Wirth R, Heinrich W (2005) The geochemical cycle of boron: constraints from boron isotope patitioning experiments between mica and fluid. Lithos 84:206– 216 Wunder B, Meixner A, Romer RL, Heinrich W (2006) Temperturedependent isotopic fractionation of lithium between clinopyroxene and high-pressure fluids. Contrib Mineral Petrol 151:112– 120 Wunder B, Meixner A, Romer RL, Feenstra A, Schettler G, Heinrich W (2007) Lithium isotope fractionation between Li-bearing staurolite, Li-mica and aqueous fluids: an experimental study. Chem Geol 238:277–290 123
© Copyright 2025 Paperzz