Figures and Tables

top

Introduction

The Mo stable isotopic system is a very promising tool to explore both the chemical evolution of the silicate Earth (e.g., McCoy-West et al., 2019

McCoy-West, A.J., Chowdhury, P., Burton, K.W., Sossi, P., Nowell, G.M., Fitton, J.G., Kerr, A.C., Cawood, P.A., Williams, H.M. (2019) Extensive crustal extraction in Earth’s early history inferred from molybdenum isotopes. Nature Geoscience 12, 946–951. https://doi.org/10.1038/s41561-019-0451-2

) and the palaeo-redox conditions of oceans (e.g., Wille et al., 2007

Wille, M., Kramers, J.D., Nägler, T.F., Beukes, N.J., Schröder, S., Meisel, Th., Lacassie, J.P., Voegelin, A.R. (2007) Evidence for a gradual rise of oxygen between 2.6 and 2.5 Ga from Mo isotopes and Re-PGE signatures in shales. Geochimica et Cosmochimica Acta 71, 2417–2435. https://doi.org/10.1016/j.gca.2007.02.019

). Mass balance models associated with both types of applications strongly rely on the Mo isotopic composition (δ^98/95Mo = 1000 × [(⁹⁸Mo/⁹⁵Mo_sample)/(⁹⁸Mo/⁹⁵Mo_standard) − 1]) of the continental crust (CC), especially its upper layer (UCC), because it is highly enriched in Mo and in direct contact with the hydrosphere. While an estimate for the Mo stable isotope composition of the UCC created prior to the Great Oxidation Event (GOE; ∼2.4–2.2 Ga) exists (δ^98/95Mo = +0.03 ‰; Greaney et al., 2020

Greaney, A.T., Rudnick, R.L., Romaniello, S.J., Johnson, A.C., Gaschnig, R.M., Anbar, A.D. (2020) Molybdenum isotope fractionation in glacial diamictites tracks the onset of oxidative weathering of the continental crust. Earth and Planetary Science Letters 534, 116083. https://doi.org/10.1016/j.epsl.2020.116083

), constraints on post-GOE UCC δ^98/95Mo are scarce and conflicting. The difficulty in constraining the δ^98/95Mo of the UCC after the GOE is a consequence of the redox sensitivity of Mo and its fluid-mobility under oxidising conditions. These properties, for instance, prevent the usage of fine grained clastic sedimentary rocks for that purpose (e.g., Greaney et al., 2020

Greaney, A.T., Rudnick, R.L., Romaniello, S.J., Johnson, A.C., Gaschnig, R.M., Anbar, A.D. (2020) Molybdenum isotope fractionation in glacial diamictites tracks the onset of oxidative weathering of the continental crust. Earth and Planetary Science Letters 534, 116083. https://doi.org/10.1016/j.epsl.2020.116083

). One approach to constrain the δ^98/95Mo of modern UCC has been to use the signatures of molybdenites (MoS₂), mostly derived from magmatic-hydrothermal fluids, as proxies for exposed rocks. It was initially thought that isotopic fractionation would be minor in high temperature systems and that, MoS₂ could therefore represent the isotopic composition of their crustal source rocks (e.g., Barling et al., 2001

Barling, J., Arnold, G.L., Anbar, A.D. (2001) Natural mass-dependent variations in the isotopic composition of molybdenum. Earth and Planetary Science Letters 193, 447–457. https://doi.org/10.1016/S0012-821X(01)00514-3

). As studies multiplied, it became clear that the controls on Mo isotopes in hydrothermal systems were more complex than initially thought (e.g., Hannah et al., 2007

Hannah, J.L., Stein, H.J., Wieser, M.E., de Laeter, J.R., Varner, M.D. (2007) Molybdenum isotope variations in molybdenite: Vapor transport and Rayleigh fractionation of Mo. Geology 35, 703–706. https://doi.org/10.1130/G23538A.1

). Nevertheless, because these controls were inferred to result in progressively heavier MoS₂, a global average for MoS₂ δ^98/95Mo was suggested to represent a maximum value for Phanerozoic UCC (Greber et al., 2014

Greber, N.D., Pettke, T., Nägler, T.F. (2014) Magmatic–hydrothermal molybdenum isotope fractionation and its relevance to the igneous crustal signature. Lithos 190–191, 104–110. https://doi.org/10.1016/j.lithos.2013.11.006

). This, however, is at odds with a recent Phanerozoic UCC composition derived from igneous rock compositions (δ^98/95Mo = +0.14 ± 0.07 ‰; Yang et al., 2017

Yang, J., Barling, J., Siebert, C., Fietzke, J., Stephens, E., Halliday, A.N. (2017) The molybdenum isotopic compositions of I-, S- and A-type granitic suites. Geochimica et Cosmochimica Acta 205, 168–186. https://doi.org/10.1016/j.gca.2017.01.027

), since the latter is visibly heavier than the most recent MoS₂ δ^98/95Mo averages (+0.04 ‰ in Breillat et al., 2016

Breillat, N., Guerrot, C., Marcoux, E., Négrel, Ph. (2016) A new global database of δ⁹⁸Mo in molybdenites: A literature review and new data. Journal of Geochemical Exploration 161, 1–15. https://doi.org/10.1016/j.gexplo.2015.07.019

; −0.04 ‰ in Willbold and Elliott, 2017

Willbold, M., Elliott, T. (2017) Molybdenum isotope variations in magmatic rocks. Chemical Geology 449, 253–268. https://doi.org/10.1016/j.chemgeo.2016.12.011

). Clearly, current constraints on Phanerozoic UCC do not converge, and deriving a robust estimate will require a better understanding of magmatic-hydrothermal systems.

One geological process having the potential to create the discrepancy described above is Mo isotopic fractionation during fluid exsolution in silicic systems. The dominant igneous rock types in the UCC are plutonic silicic rocks, and these lithologies were shown to be depleted in Mo (20–90 % depletion; mean 60 %) compared to fluid immobile elements of similar incompatibility (e.g., Ce and Pr; Greaney et al., 2018

Greaney, A.T., Rudnick, R.L., Gaschnig, R.M., Whalen, J.B., Luais, B., Clemens, J.D. (2018) Geochemistry of molybdenum in the continental crust. Geochimica et Cosmochimica Acta 238, 36–54. https://doi.org/10.1016/j.gca.2018.06.039

). In some plutonic suites, correlations between Mo and fluid soluble elements exist, suggesting fluid exsolution as a dominant control over the Mo depletions observed (Greaney et al., 2018

Greaney, A.T., Rudnick, R.L., Gaschnig, R.M., Whalen, J.B., Luais, B., Clemens, J.D. (2018) Geochemistry of molybdenum in the continental crust. Geochimica et Cosmochimica Acta 238, 36–54. https://doi.org/10.1016/j.gca.2018.06.039

). Significant Mo partitioning in magmatic fluids is also supported by a large number of experimental (e.g., Candela and Holland, 1984

Candela, P.A., Holland, H.D. (1984) The partitioning of copper and molybdenum between silicate melts and aqueous fluids. Geochimica et Cosmochimica Acta 48, 373–380. https://doi.org/10.1016/0016-7037(84)90257-6

) and empirical (e.g., Zajacz et al., 2008

Zajacz, Z., Halter, W.E., Pettke, T., Guillong, M. (2008) Determination of fluid/melt partition coefficients by LA-ICPMS analysis of co-existing fluid and silicate melt inclusions: Controls on element partitioning. Geochimica et Cosmochimica Acta 72, 2169–2197. https://doi.org/10.1016/j.gca.2008.01.034

) studies. Whether isotopic fractionation is associated with this process remains unclear, with only one empirical study suggesting a possible enrichment of heavy Mo isotopes in fluids in a Mo porphyry deposit (Questa, USA; Greber et al., 2014

Greber, N.D., Pettke, T., Nägler, T.F. (2014) Magmatic–hydrothermal molybdenum isotope fractionation and its relevance to the igneous crustal signature. Lithos 190–191, 104–110. https://doi.org/10.1016/j.lithos.2013.11.006

). This is based on the very low Mo and light δ^98/95Mo of a porphyry rhyolite dike compared to the molybdenites in the nearby and contemporaneous Questa mineralisation. However, the porphyry dike is altered and may not derive from the same magma as that from which the fluids of the mineralisation are derived (Greber et al., 2014

Greber, N.D., Pettke, T., Nägler, T.F. (2014) Magmatic–hydrothermal molybdenum isotope fractionation and its relevance to the igneous crustal signature. Lithos 190–191, 104–110. https://doi.org/10.1016/j.lithos.2013.11.006

and references therein). Furthermore, theoretical constraints suggest that, in the occurrence of isotopic fractionation, a preferential enrichment of light Mo isotopes in the fluids should occur. This is because Mo coordination in silicate melts is tetrahedral, while both tetrahedral and octahedral coordination have been inferred for Mo in magmatic-hydrothermal fluids (e.g., Borg et al., 2012

Borg, S., Liu, W., Etschmann, B., Tian, Y., Brugger, J. (2012) An XAS study of molybdenum speciation in hydrothermal chloride solutions from 25–385 °C and 600 bar. Geochimica et Cosmochimica Acta 92, 292–307. https://doi.org/10.1016/j.gca.2012.06.001

). Given that most MoS₂ measured thus far have crystallised from fluids exsolved from silicic magmas or melts highly enriched in those fluids (Breillat et al., 2016

Breillat, N., Guerrot, C., Marcoux, E., Négrel, Ph. (2016) A new global database of δ⁹⁸Mo in molybdenites: A literature review and new data. Journal of Geochemical Exploration 161, 1–15. https://doi.org/10.1016/j.gexplo.2015.07.019

), a preferential enrichment in light Mo isotopes in the fluids could explain the lighter δ^98/95Mo of MoS₂ averages, compared to average silicic rocks (δ^98/95Mo = +0.16 ‰ in Yang et al., 2017

Yang, J., Barling, J., Siebert, C., Fietzke, J., Stephens, E., Halliday, A.N. (2017) The molybdenum isotopic compositions of I-, S- and A-type granitic suites. Geochimica et Cosmochimica Acta 205, 168–186. https://doi.org/10.1016/j.gca.2017.01.027

) and the associated discrepancies in UCC constraints. It is therefore the aim of this contribution to establish the first experimental constraints of the fluid/melt equilibrium fractionation value of Mo stable isotopes (Δ^98/95Mo_fluid-melt = δ^98/95Mo_fluid − δ^98/95Mo_melt) at temperatures, fluid salinities, melt compositions and oxygen fugacities relevant to supercritical fluid exsolution in upper crustal silicic magmatic systems.

top

Methods

Twelve experiments were conducted at McGill University and Institut des Sciences de la Terre d’Orléans (ISTO) to simulate the equilibration of exsolved fluids with upper crustal felsic magmas. Conditions for each experiment are shown in Table 1. Cold seal pressure vessels were used, except for two experiments (Mo21 and Mo27) requiring the usage of internally heated pressure vessels (IHPV) to allow for higher temperature (900 °C) or higher oxygen fugacity (∼FMQ+3). Powders of Mo-doped haplogranitic glasses (n = 11) or non-doped natural obsidian (n = 1) were loaded into gold capsules together with Mo-free fluids in similar proportions (1:1) (see Supplementary Information for details on the experimental approach). The experiments were equilibrated at 200 MPa, a pressure relevant for most long lived magma chambers (e.g., Huber et al., 2019

Huber, C., Townsend, M., Degruyter, W., Bachmann, O. (2019) Optimal depth of subvolcanic magma chamber growth controlled by volatiles and crust rheology. Nature Geoscience 12, 762–768. https://doi.org/10.1038/s41561-019-0415-6

). At such conditions, a single magmatic fluid phase with low to intermediate salinity (e.g., 1–5 wt. % NaCl in Rusk et al., 2004

Rusk, B.G., Reed, M.H., Dilles, J.H., Klemm, L.M., Heinrich, C.A. (2004) Compositions of magmatic hydrothermal fluids determined by LA-ICP-MS of fluid inclusions from the porphyry copper–molybdenum deposit at Butte, MT. Chemical Geology 210, 173–199. https://doi.org/10.1016/j.chemgeo.2004.06.011

), a supercritical fluid, exsolves from magmas (Burnham, 1979

Burnham, C.W. (1979) Magmas and hydrothermal fluids. In: Barnes, H.L. (Ed.) Geochemistry of Hydrothermal Ore Deposits. Second Edition, Wiley, New York, 71–136.

). We therefore set the range of salinity of our experimental fluids accordingly (0.5–1.5 M (Na,K)Cl). The effects of changing temperature, melt aluminium saturation index (ASI) and oxygen fugacity (fO₂) on Δ^98/95Mo_fluid-melt were also assessed over ranges relevant for upper crustal silicic magma chambers (see Table 1). In similar experiments, chemical equilibrium is typically reached within a maximum of 10 days (e.g., Jiang et al., 2021

Jiang, Z., Shang, L., Guo, H., Wang, X.-s., Chen, C., Zhou, Y. (2021) An experimental investigation into the partition of Mo between aqueous fluids and felsic melts: Implications for the genesis of porphyry Mo ore deposits. Ore Geology Reviews 134, 104144. https://doi.org/10.1016/j.oregeorev.2021.104144

and references therein). Time scales for chemical and isotopic equilibrium were therefore assessed via a series of identical experiments performed at 800 °C with durations ranging from 1 to 20 days.

Experiment ID	Starting material*	T (oC)	Duration (days)	Starting glass ASI	(Na,K)Cl in fluid (mol/L)	fO2	δ98/95Momelt (‰)**	CMo,melt (μg/g)	δ98/95Mofluid (‰)**	CMo,fluid (μg/g)	Dfluid-melt	Δ98/95Mofluid-melt (‰)**
Mo15	HG 1A	700	20	1	1	∼FMQ+1	0.12	171	−0.31	47	0.28	−0.43
Mo22	HG 1A	800	1	1	1	∼FMQ+1	0.08	197	−0.23	43	0.22	−0.30
Mo02	HG 1A	800	5	1	1	∼FMQ+1	0.06	184	−0.25	48	0.26	−0.31
Mo03	HG 1A	800	10	1	1	∼FMQ+1	0.09	176	−0.24	56	0.32	−0.32
Mo16	HG 1A	800	10	1	1	∼FMQ+1	0.07	188	−0.26	52	0.28	−0.33
Mo07	HG 1A	800	20	1	1	∼FMQ+1	0.09	171	−0.26	52	0.30	−0.36
Mo21	HG 1A	900	7	1	1	∼FMQ+1	0.11	175	−0.24	59	0.34	−0.35
Mo05	HG 1A	800	10	1	0.5	∼FMQ+1	0.07	187	−0.24	51	0.27	−0.32
Mo33	HG 1A	800	11	1	1.5	∼FMQ+1	0.05	170	−0.26	63	0.37	−0.31
Mo27	HG 1B	800	10	1	1	∼FMQ+3	0.03	269	−0.19	102	0.38	−0.22
Mo08	HG 0.8	800	10	0.80	1	∼FMQ+1	0.05	245	−0.13	148	0.60	−0.17
Mo24	Nat. obsi.	800	10	0.85	1.5	∼FMQ+1	0.19	5.8	−0.04	2.4	0.40	−0.23

Abbreviations: ASI, aluminum saturation index; fO₂, oxygen fugacity relative to the fayalite-magnetite-quartz buffer (FMQ).
*HG 1A: haplogranite with ASI = 1, 254 μg/g Mo and δ^98/95Mo = −0.01 ± 0.05 ‰. HG 1B: haplogranite with ASI = 1, 383 μg/g Mo and δ^98/95Mo = −0.01 ± 0.05 ‰. HG 0.8: haplogranite with ASI = 0.8, 421 μg/g Mo and δ^98/95Mo = −0.01 ± 0.05 ‰. HG 1A, HG 1B and HG 0.8 contain 78.7 wt. % SiO₂. Nat. obsi.: natural obsidian with ASI = 0.85, 76.7 wt. % SiO₂, 8.8 μg/g Mo and δ^98/95Mo = 0.13 ± 0.05 ‰.
**Error on δ^98/95Mo is 0.05 ‰; error on Δ^98/95Mo_fluid-melt is propagated (0.07 ‰).

Mo isotope compositions and Mo concentrations of starting glasses, final (quenched) glasses and fluids were measured using double spike MC-ICP-MS at the University of Göttingen. The detailed description of the analytical approach is shown in the Supplementary Information. The uncertainty presented for the δ^98/95Mo of each sample corresponds to twice the standard deviation of the replicate analyses of the Japan Geological Survey reference materials (±0.05 ‰).

top

Results

δ^98/95Mo values and Mo concentrations of starting materials and experimental results are presented in Table 1, together with the calculated Mo partition coefficients between fluid and melt (D_fluid-melt = C_Mo,fluid /C_Mo,melt) and the Δ^98/95Mo_fluid-melt for each pair. The D_fluid-melt and Δ^98/95Mo_fluid-melt of identical experiments produced at different durations (the time series experiments; Figs. 1, 2) suggest that both chemical and isotopic equilibrium were reached between 5 and 10 days at 800 °C. This is consistent with chemical equilibrium time scales observed in similar experimental studies (e.g., Jiang et al., 2021

Jiang, Z., Shang, L., Guo, H., Wang, X.-s., Chen, C., Zhou, Y. (2021) An experimental investigation into the partition of Mo between aqueous fluids and felsic melts: Implications for the genesis of porphyry Mo ore deposits. Ore Geology Reviews 134, 104144. https://doi.org/10.1016/j.oregeorev.2021.104144

and references therein). Therefore, with the exception of the 1 and 5 days experiments, all presented experimental data represent equilibrium values.

Full size image

Full size image

At equilibrium, D_fluid-melt range from 0.27 to 0.60 and overlap the literature range for similar experiments (e.g., compilation in Fang and Audétat, 2022

Fang, J., Audétat, A. (2022) The effects of pressure, fO₂, fS₂ and melt composition on the fluid–melt partitioning of Mo: Implications for the Mo-mineralization potential of upper crustal granitic magmas. Geochimica et Cosmochimica Acta 336, 1–14. https://doi.org/10.1016/j.gca.2022.08.016

). Δ^98/95Mo_fluid-melt range from −0.43 ± 0.07 ‰ to −0.17 ± 0.07 ‰ (Fig. 2) and therefore indicate the preferential incorporation of light Mo isotopes in fluids under all investigated conditions, and whether or not the starting glass was doped. The data set shows a clear control of the melt ASI over Δ^98/95Mo_fluid-melt (linear regression with R² = 0.99; Fig. 3a) with the greatest isotopic differences observed in the samples with the least peralkaline compositions. No effect of fluid salinity is observed (Fig. 3b). Apparent systematics suggest that Δ^98/95Mo_fluid-melt could be larger at lower temperatures and lower oxygen fugacities (Fig. 3c, d), but this is not resolvable in this data set, and further experimental investigation will be required to test these possible correlations.

Full size image

top

Interpretation and Discussion of Experimental Results

The experimental data suggest that silicic melts with geologically realistic ASI will preferentially lose light Mo isotopes to exsolved supercritical fluids. In theory, both a difference in coordination and valence state of Mo between fluid and melt could induce the isotopic fractionation observed, i.e. higher coordination and/or lower valence state of Mo in the fluid than in the melt (e.g., Urey, 1947

Urey, H.C. (1947) The Thermodynamic Properties of Isotopic Substances. Journal of the Chemical Society (Resumed) 1947, 562–581. https://doi.org/10.1039/jr9470000562

). However, thermodynamic and empirical evidence suggests that the valence state of Mo should be similar in silicate melts and associated magmatic-hydrothermal fluids, i.e. hexavalent (e.g., Kaufmann et al., 2021

Kaufmann, A.K.C., Pettke, T., Wille, M. (2021) Molybdenum isotope fractionation at upper-crustal magmatic-hydrothermal conditions. Chemical Geology 578, 120319. https://doi.org/10.1016/j.chemgeo.2021.120319

; Willbold and Elliott, 2017

Willbold, M., Elliott, T. (2017) Molybdenum isotope variations in magmatic rocks. Chemical Geology 449, 253–268. https://doi.org/10.1016/j.chemgeo.2016.12.011

and references therein). Therefore, the most likely driver behind the Δ^98/95Mo_fluid-melt observed in our experiments is higher Mo coordination in the fluid compared to the melt. In silicate melts, Mo dominantly occurs as tetrahedral molybdate species (MoO₄²⁻). On the other hand, in high temperature fluids, the coordination of Mo remains debated. Some studies suggested that Mo dominantly occurs as Na-K molybdate, monochloride or thio-molybdate at modest salinity, and as Mo-oxy-hydroxy complexes at low salinity (Zhang et al., 2012

Zhang, L., Audétat, A., Dolejš, D. (2012) Solubility of molybdenite (MoS₂) in aqueous fluids at 600–800 °C, 200 MPa: A synthetic fluid inclusion study. Geochimica et Cosmochimica Acta 77, 175–185. https://doi.org/10.1016/j.gca.2011.11.015

; Tattitch and Blundy, 2017

Tattitch, B.C., Blundy, J.D. (2017) Cu-Mo partitioning between felsic melts and saline-aqueous fluids as a function of X_NaCleq, f_O2, and f_S2. American Mineralogist 102, 1987–2006. https://doi.org/10.2138/am-2017-5998

). In such cases, Mo would be tetrahedrally coordinated, and no isotopic fractionation should be observed between fluids and melts, which is inconsistent with our experimental results. However, others have suggested the occurrence of species involving octahedrally coordinated Mo. For instance, Ulrich and Mavrogenes (2008)

Ulrich, T., Mavrogenes, J. (2008) An experimental study of the solubility of molybdenum in H₂O and KCl–H₂O solutions from 500 °C to 800 °C, and 150 to 300 MPa. Geochimica et Cosmochimica Acta 72, 2316–2330. https://doi.org/10.1016/j.gca.2008.02.014

suggested the presence of a chloro-oxo Mo (VI) complex in high salinity (>20 % KCl) fluids (at 490 °C and ≥150 MPa). Borg et al. (2012)

Borg, S., Liu, W., Etschmann, B., Tian, Y., Brugger, J. (2012) An XAS study of molybdenum speciation in hydrothermal chloride solutions from 25–385 °C and 600 bar. Geochimica et Cosmochimica Acta 92, 292–307. https://doi.org/10.1016/j.gca.2012.06.001

, based on experiments performed at lower temperatures and pressures (up to 385 °C, 60 MPa), found that octahedral species were becoming predominant in increasingly acidic solutions (pH < 5), with species such as molybdic acid and chloro-oxo Mo complexes. They also showed that increasing temperature favoured the formation of oxo-chloro complexes and suggested that these could be responsible for Mo transport in less acidic solutions at higher temperature (e.g., 700 °C), as proposed by Ulrich and Mavrogenes (2008)

Ulrich, T., Mavrogenes, J. (2008) An experimental study of the solubility of molybdenum in H₂O and KCl–H₂O solutions from 500 °C to 800 °C, and 150 to 300 MPa. Geochimica et Cosmochimica Acta 72, 2316–2330. https://doi.org/10.1016/j.gca.2008.02.014

. Based on the lack of correlation between Δ^98/95Mo_fluid-melt and starting fluid salinity in our experiments, a control of Mo isotopes by chloro-oxo Mo complexes alone seems unlikely. Hence, at least one other species in which Mo coordination is greater than tetrahedral, perhaps molybdic acid, is required in the fluids to explain the systematics observed in our experiments. Finally, the negative correlation between Δ^98/95Mo_fluid-melt and ASI suggests that the latter exerts a strong influence on the coordination of Mo.

top

Implications for the Composition of the Upper Continental Crust

Most large and long lived silicic magma chambers are located at upper crustal levels corresponding to lithostatic pressures of ∼200 MPa (Huber et al., 2019

Huber, C., Townsend, M., Degruyter, W., Bachmann, O. (2019) Optimal depth of subvolcanic magma chamber growth controlled by volatiles and crust rheology. Nature Geoscience 12, 762–768. https://doi.org/10.1038/s41561-019-0415-6

). There, significant amounts of exsolved supercritical fluids accumulate and equilibrate with silicic magmas. Slow fluid exsolution associated with the cooling and crystallisation of the melt is punctuated by repetitive fast exsolution events associated with decompression of incoming recharging melt (e.g., Edmonds and Woods, 2018

Edmonds, M., Woods, A.W. (2018) Exsolved volatiles in magma reservoirs. Journal of Volcanology and Geothermal Research 368, 13–30. https://doi.org/10.1016/j.jvolgeores.2018.10.018

). Our data indicate that the extraction of these exsolved fluids likely results in the removal of light Mo from magma bodies, including precursor bodies of the silicic plutonic rocks making ∼50 % of the UCC (e.g., Wedepohl, 1995

Wedepohl, K.H. (1995) The composition of the continental crust. Geochimica et Cosmochimica Acta 59, 1217–1232. https://doi.org/10.1016/0016-7037(95)00038-2

). The Mo depletion of these plutons (e.g., Greaney et al., 2018

Greaney, A.T., Rudnick, R.L., Gaschnig, R.M., Whalen, J.B., Luais, B., Clemens, J.D. (2018) Geochemistry of molybdenum in the continental crust. Geochimica et Cosmochimica Acta 238, 36–54. https://doi.org/10.1016/j.gca.2018.06.039

), together with the large number of experimental constraints, suggests that a significant Mo fraction must have been removed from their precursor melts via exsolved fluids. Since UCC silicic plutonic rocks are dominantly peraluminous (ASI ≈ 1.25; Wedepohl, 1995

Wedepohl, K.H. (1995) The composition of the continental crust. Geochimica et Cosmochimica Acta 59, 1217–1232. https://doi.org/10.1016/0016-7037(95)00038-2

), our experiments suggest that resolvable Δ^98/95Mo_fluid-melt likely applied during the devolatilisation of their precursor melt. In other words, the δ^98/95Mo of UCC silicic plutons should be heavier than their precursor, un-degassed melts. Based on our D_fluid-melt, the difference between undegassed and degassed silicic magmas might, however, not be large. For instance, using the highest D_fluid-melt in our experiments and assuming an extreme scenario whereby a hydrous melt with 10 wt. % volatiles undergoes fluid exsolution with a Δ^98/95Mo_fluid-melt of −0.51 ‰ (extrapolation of the linear regression to ASI = 1.25 in Fig. 3a), would result in a residual melt that is only 0.03 ‰ heavier than the undegassed melt (see Supplementary Information for calculation). This is smaller than our analytical uncertainty. However, the range of published D_fluid-melt for similar experiments is large and includes significantly higher values (e.g., compilation in Fang and Audétat, 2022

Fang, J., Audétat, A. (2022) The effects of pressure, fO₂, fS₂ and melt composition on the fluid–melt partitioning of Mo: Implications for the Mo-mineralization potential of upper crustal granitic magmas. Geochimica et Cosmochimica Acta 336, 1–14. https://doi.org/10.1016/j.gca.2022.08.016

). It is therefore best to consider the average δ^98/95Mo of UCC silicic plutonic rocks (Yang et al., 2017

Yang, J., Barling, J., Siebert, C., Fietzke, J., Stephens, E., Halliday, A.N. (2017) The molybdenum isotopic compositions of I-, S- and A-type granitic suites. Geochimica et Cosmochimica Acta 205, 168–186. https://doi.org/10.1016/j.gca.2017.01.027

) as a maximum value for the signature of the total Mo contribution to the UCC from undegassed silicic magmatism. In turn, UCC δ^98/95Mo estimates derived from igneous rocks should also be viewed as maxima.

Based on our experiments, the average δ^98/95Mo of UCC silicic rocks should also be clearly heavier than global δ^98/95Mo average for UCC MoS₂, since MoS₂ measured thus far are from magmatic-hydrothermal systems. There, MoS₂ crystallises from both brines and low salinity vapours that unmix from supercritical fluids once they reach a miscibility gap in the NaCl-H₂O system during ascent in the shallowest part of magmatic-hydrothermal complexes (typically <140 MPa and 400–700 °C; e.g., Bodnar et al., 1985

Bodnar, R.J., Burnham, C.W., Sterner, S.M. (1985) Synthetic fluid inclusions in natural quartz. III. Determination of phase equilibrium properties in the system H₂O-NaCl to 1000°C and 1500 bars. Geochimica et Cosmochimica Acta 49, 1861–1873. https://doi.org/10.1016/0016-7037(85)90081-X

). The average UCC silicic rock composition (δ^98/95Mo = +0.16 ‰) of Yang et al. (2017)

Yang, J., Barling, J., Siebert, C., Fietzke, J., Stephens, E., Halliday, A.N. (2017) The molybdenum isotopic compositions of I-, S- and A-type granitic suites. Geochimica et Cosmochimica Acta 205, 168–186. https://doi.org/10.1016/j.gca.2017.01.027

is indeed ∼0.18 ‰ and ∼0.12 ‰ heavier than the two most extensive and recent MoS₂ global averages of Willbold and Elliott (2017)

Willbold, M., Elliott, T. (2017) Molybdenum isotope variations in magmatic rocks. Chemical Geology 449, 253–268. https://doi.org/10.1016/j.chemgeo.2016.12.011

and Breillat et al. (2016)

Breillat, N., Guerrot, C., Marcoux, E., Négrel, Ph. (2016) A new global database of δ⁹⁸Mo in molybdenites: A literature review and new data. Journal of Geochemical Exploration 161, 1–15. https://doi.org/10.1016/j.gexplo.2015.07.019

(−0.04 ‰ and +0.04 ‰, respectively), in agreement with the direction of isotopic fractionation in the experiments. The isotopic difference between UCC silicic rocks and UCC magmatic-hydrothermal MoS₂ averages is however smaller than suggested by the experimental Δ^98/95Mo_fluid-melt values. In the experiments, equilibrated melts are up to 0.4 ‰ heavier than associated supercritical fluids, and while this will need to be confirmed in future experiments, the negative correlation between Δ^98/95Mo_fluid-melt and ASI suggests that even greater values could apply during exsolution from more peraluminous melts, such as those of the plutons dominating the UCC. While other factors are possible, this smaller isotopic difference could simply be the consequence of the timing of fluid exsolution relative to mineral fractionation in silicic magma chambers. Most recent studies suggest that long lived silicic magma reservoirs are largely crystallised between recharge events (e.g., Schmitt et al., 2010

Schmitt, A.K., Stockli, D.F., Lindsay, J.M., Robertson, R., Lovera, O.M., Kislitsyn, R. (2010) Episodic growth and homogenization of plutonic roots in arc volcanoes from combined U–Th and (U–Th)/He zircon dating. Earth and Planetary Science Letters 295, 91–103. https://doi.org/10.1016/j.epsl.2010.03.028

). Hence, over the lifespan of a silicic magma body, most exsolved fluids will equilibrate with interstitial melts of crystal mushes. These interstitial melts are expected to be enriched in Mo, since it is an incompatible element. More importantly, based on the Mo coordination (octahedral) in all minerals capable of carrying significant amount of Mo in the mush (Ti-bearing oxides, biotite, amphibole, K feldspar; Greaney et al., 2018

Greaney, A.T., Rudnick, R.L., Gaschnig, R.M., Whalen, J.B., Luais, B., Clemens, J.D. (2018) Geochemistry of molybdenum in the continental crust. Geochimica et Cosmochimica Acta 238, 36–54. https://doi.org/10.1016/j.gca.2018.06.039

), interstitial melts (in which Mo is tetrahedral) are most likely isotopically heavier than the bulk mush. Exsolved fluids in equilibrium with these melts should therefore be heavier than hypothetical fluids in equilibrium with bulk mushes or fully crystallised equivalents. This would explain the smaller isotopic difference between UCC silicic rocks and UCC MoS₂ averages, than expected based on the experiments herein.

Overall, our results provide a solution for the discrepancy between the UCC δ^98/95Mo estimate derived from exposed igneous UCC rocks and constraints obtained from average UCC MoS₂. They also stress the lack of straight forward assessment of the UCC δ^98/95Mo via MoS₂ and suggest that UCC δ^98/95Mo derived from igneous rock compositions should be considered as a maximum value.

top

Acknowledgements

We thank Horst Marschall for the editorial handling of the manuscript and Alex McCoy-West and an anonymous reviewer for constructive reviews. This work was supported by the German Research Foundation grants BE 6670/1-1 and BE 6670/2-1 (RB) and Fundamental Research Funds for the Central Universities, China University of Geosciences (Wuhan) CUG230610 (HG).

Editor: Horst R. Marschall

top

References

top

Supplementary Information

The Supplementary Information includes:

Experimental and Analytical Methods

Details for Modelled Impact of Fluid Exsolution in Discussion

Supplementary Table S-1

Supplementary Information References

Download the Supplementary Information (PDF)