Serpentinite dehydration in the subducted lithosphere produces no B isotopic fractionation
Affiliations | Corresponding Author | Cite as | Funding information
Published by the European Association of Geochemistry
under Creative Commons License CC BY 4.0
Keywords: boron isotopes, isotopic fractionation, serpentinite dehydration, subduction fluids
- Share this article
-
Article views:182Cumulative count of HTML views and PDF downloads.
- Download Citation
- Rights & Permissions
top
Abstract
Figures
 Figure 1 Boron concentration versus δ11B of serpentinites from various settings (modified from Martin et al., 2020). Boxes represent antigorite in situ δ11B data from this study, with B concentrations of Zermatt-Saas and Erro-Tobbio samples taken from Ulrich et al. (2024) and from Vesin and Ulrich et al. (2025), respectively. |  Figure 2 In situ δ11B of antigorite and olivine from the studied samples compared to the dataset of other HP serpentinites from the Western Alps (Zermatt-Saas, Aosta, Erro-Tobbio and Monviso, marked by *) from Clarke et al. (2020). Boxes with dotted lines indicate samples from this study with olivine-antigorite O isotopic equilibrium (Ulrich et al., 2024; Vesin and Ulrich et al., 2025). Mineral abbreviations: Ol, olivine; Atg, antigorite. |  Figure 3 (a) Δ11BOl-Atg versus Δ18OAtg-Ol and (b) Δ11BOl-Atg versus DBOl/Atg (δ18O data and DBOl/Atg values from Ulrich et al., 2024). Bars for Δ11BOl-Atg indicate the range from the smallest possible value (δ11BOl min − δ11BAtg max) to the largest possible value (δ11BOl max − δ11BAtg min), while the sample average is indicated by the black dots. For Δ18OAtg-Ol and DB values, the sample average ± the propagated uncertainty (2 s.d.) are reported. The grey vertical band in (a) indicates the range of equilibrium Δ18OAtg-Ol at 550–600 °C, described in Ulrich et al. (2024). δ11B and DB datasets marked with an asterisk (*) in (b) are from Clarke et al. (2020). |
| Figure 1 | Figure 2 | Figure 3 |
top
Boron and Its Isotopes in Serpentinite
Serpentinites play a critical role in fluid transport and element recycling processes within subduction zones, impacting geological and geochemical dynamics on a global scale. Fluid transfer at mantle depths occurs mainly in subduction zones where serpentinites dehydrate and release aqueous fluid into the mantle wedge (Ulmer and Trommsdorff, 1995
Ulmer, P., Trommsdorff, V. (1995) Serpentine Stability to Mantle Depths and Subduction-Related Magmatism. Science 268, 858–861. https://doi.org/10.1126/science.268.5212.858
). Boron is a relatively fluid-mobile element, which is enriched in serpentinites, and its isotopes are a powerful tracer of fluid processes such as fluid production upon dehydration and subsequent fluid transfer. The B isotopic composition of ocean floor serpentinites varies depending on the serpentinisation conditions from δ11B = +5 to +40 ‰ (De Hoog and Savov, 2018De Hoog, J.C.M., Savov, I.P. (2018) Boron Isotopes as a Tracer of Subduction Zone Processes. In: Marschall, H., Foster, G. (Eds.) Boron Isotopes: The Fifth Element. Springer, Cham, 217–247. https://doi.org/10.1007/978-3-319-64666-4_9
and references therein), which is higher compared to the upper mantle value of −7.1 ± 0.9 ‰ (Marschall et al., 2017Marschall, H.R., Wanless, V.D., Shimizu, N., Pogge von Strandmann, P.A.E., Elliott, T., Monteleone, B.D. (2017) The boron and lithium isotopic composition of mid-ocean ridge basalts and the mantle. Geochimica et Cosmochimica Acta 207, 102–138. https://doi.org/10.1016/j.gca.2017.03.028
). Subducted high pressure (HP) serpentinites (also referred to as slab serpentinites) generally maintain a high δ11B (>+10 ‰), whereas mantle wedge serpentinites show an isotopically lighter range of compositions with δ11B = −14 to +10 ‰ (Martin et al., 2020Martin, C., Flores, K.E., Vitale-Brovarone, A., Angiboust, S., Harlow, G.E. (2020) Deep mantle serpentinization in subduction zones: Insight from in situ B isotopes in slab and mantle wedge serpentinites. Chemical Geology 545, 119637. https://doi.org/10.1016/j.chemgeo.2020.119637
). These relatively low δ11B values (δ11B < +10 ‰) likely represent serpentinisation of the mantle wedge by crustal fluids derived from the subducting slab (e.g., Yamada et al., 2019Yamada, C., Tsujimori, T., Chang, Q., Kimura, J.-I. (2019) Boron isotope variations of Franciscan serpentinites, northern California. Lithos 334–335, 180–189. https://doi.org/10.1016/j.lithos.2019.02.004
; Martin et al., 2020Martin, C., Flores, K.E., Vitale-Brovarone, A., Angiboust, S., Harlow, G.E. (2020) Deep mantle serpentinization in subduction zones: Insight from in situ B isotopes in slab and mantle wedge serpentinites. Chemical Geology 545, 119637. https://doi.org/10.1016/j.chemgeo.2020.119637
; De Hoog et al., 2023De Hoog, J.C.M., Clarke, E., Hattori, K. (2023) Mantle wedge olivine modifies slab-derived fluids: Implications for fluid transport from slab to arc magma source. Geology 51, 663–667. https://doi.org/10.1130/G51169.1
). However, negative δ11B values in serpentine from ophicarbonates of oceanic derivation (Cannaò et al., 2024Cannaò, E., Tiepolo, M., Agostini, S., Scambelluri, M. (2024) Fossil hydrothermal oceanic systems through in-situ B isotopes in ophicarbonates (N. Apennines, Italy). Chemical Geology 645, 121899. https://doi.org/10.1016/j.chemgeo.2023.121899
) demonstrate how heterogeneous the input in subduction zones can be.Previous work has shown that dehydration of former ocean floor serpentinites during subduction can be a complex process possibly involving external fluids, which can modify the B isotopic signature and lead to isotopic disequilibrium between serpentinite residue and newly formed minerals, such as metamorphic olivine (Clarke et al., 2020
Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
). Therefore, a microscale approach is necessary to decipher the extent to which B incorporation into metamorphic olivine fractionates B isotopes, and to evaluate whether an external fluid is involved (e.g., De Hoog et al., 2014De Hoog, J.C.M., Hattori, K., Jung, H. (2014) Titanium- and water-rich metamorphic olivine in high-pressure serpentinites from the Voltri Massif (Ligurian Alps, Italy): evidence for deep subduction of high-field strength and fluid-mobile elements. Contributions to Mineralogy and Petrology 167, 990. https://doi.org/10.1007/s00410-014-0990-x
; Clarke et al., 2020Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
).Metamorphic olivine, produced during the breakdown of brucite and the serpentine polytype antigorite, can incorporate high contents of B compared to primary mantle olivine (up to 100 μg/g versus <0.11 μg/g; Clarke et al., 2020
Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
and references therein). The extent to which B is retained or lost during dehydration at the lizardite to antigorite transition and at the olivine-in reaction, and how these reactions affect the fractionation of B isotopes, is still controversial (e.g., Scambelluri and Tonarini, 2012Scambelluri, M., Tonarini, S. (2012) Boron isotope evidence for shallow fluid transfer across subduction zones by serpentinized mantle. Geology 40, 907–910. https://doi.org/10.1130/G33233.1
; De Hoog et al., 2014De Hoog, J.C.M., Hattori, K., Jung, H. (2014) Titanium- and water-rich metamorphic olivine in high-pressure serpentinites from the Voltri Massif (Ligurian Alps, Italy): evidence for deep subduction of high-field strength and fluid-mobile elements. Contributions to Mineralogy and Petrology 167, 990. https://doi.org/10.1007/s00410-014-0990-x
; Harvey et al., 2014Harvey, J., Garrido, C.J., Savov, I., Agostini, S., Padrón-Navarta, J.A., Marchesi, C., López Sánchez-Vizcaíno, V., Gómez-Pugnaire, M.T. (2014) 11B-rich fluids in subduction zones: The role of antigorite dehydration in subducting slabs and boron isotope heterogeneity in the mantle. Chemical Geology 376, 20–30. https://doi.org/10.1016/j.chemgeo.2014.03.015
; Cannaò, 2020Cannaò, E. (2020) Boron isotope fractionation in subducted serpentinites: A modelling attempt. Lithos 376–377, 105768. https://doi.org/10.1016/j.lithos.2020.105768
).Besides temperature, B isotopic fractionation is affected by factors such as potential B loss, fluid pH, and B coordination in the mineral. Boron in serpentine substitutes for Si4+ and Al3+ in predominantly tetrahedral coordination (Pabst et al., 2011
Pabst, S., Zack, T., Savov, I.P., Ludwig, T., Rost, D., Vicenzi, E.P. (2011) Evidence for boron incorporation into the serpentine crystal structure. American Mineralogist 96, 1112–1119. https://doi.org/10.2138/am.2011.3709
), whereas B in olivine is thought to be present as B3+ in predominantly trigonal coordination, substituting for Si4+ coupled to an O vacancy (Ingrin et al., 2014Ingrin, J., Kovàcs, I., Deloule, E., Balan, E., Blanchard, M., Kohn, S.C., Hermann, J. (2014) Identification of hydrogen defects linked to boron substitution in synthetic forsterite and natural olivine. American Mineralogist 99, 2138–2141. https://doi.org/10.2138/am-2014-5049
). As the heavy B isotope (11B) is preferentially incorporated in trigonal and the light B isotope (10B) in tetrahedral coordination (Kowalski et al., 2013Kowalski, P.M., Wunder, B., Jahn, S. (2013) Ab initio prediction of equilibrium boron isotope fractionation between minerals and aqueous fluids at high P and T. Geochimica et Cosmochimica Acta 101, 285–301. https://doi.org/10.1016/j.gca.2012.10.007
), olivine is expected to have a heavier B isotopic composition than the coexisting serpentine. However, recent quantum mechanical modelling suggests that, in water-rich environments (such as dehydrating serpentinites), tetrahedral B-Si-H complexes may dominate in olivine (Muir et al., 2022Muir, J.M.R., Chen, Y., Liu, X., Zhang, F. (2022) Extremely Stable, Highly Conductive Boron-Hydrogen Complexes in Forsterite and Olivine. Journal of Geophysical Research: Solid Earth 127, e2022JB024299. https://doi.org/10.1029/2022JB024299
), although B in tetrahedral coordination in olivine has not yet been documented in natural samples. This might impact the estimates on B isotopic equilibrium fractionation, which are currently based on the assumption of B in olivine being trigonally and in serpentine being tetrahedrally coordinated (e.g., Li et al., 2022Li, Y.-C., Wei, H.-Z., Palmer, M.R., Ma, J., Jiang, S.-Y., Chen, Y.-X., Lu, J.-J., Liu, X. (2022) Equilibrium boron isotope fractionation during serpentinization and applications in understanding subduction zone processes. Chemical Geology 609, 121047. https://doi.org/10.1016/j.chemgeo.2022.121047
). Therefore, isotopic equilibrium fractionation between serpentine and metamorphic olivine remains poorly constrained.In this study, we present in situ δ11B data, measured by secondary ion mass spectrometry (SIMS), of antigorite and metamorphic olivine from HP serpentinites from the Western Alps (the Zermatt-Saas unit in Switzerland and the Erro-Tobbio unit in Italy; Fig. S-1). Whole rock strontium (Sr) isotope analyses were performed on eight Zermatt-Saas samples to evaluate if any sediment-derived slab fluids might have been involved, potentially affecting the B isotopic signature of the investigated samples. The δ11B data obtained on antigorite and olivine are coupled with their O isotopic compositions and B contents (Ulrich et al., 2024
Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
; Vesin and Ulrich et al., 2025Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
) to elucidate the fate of B and its isotopes during serpentinite dehydration. Based on our results, we suggest that serpentinite dehydration in the forearc produces no significant isotopic fractionation between antigorite and metamorphic olivine during the brucite-out reaction.top
Zermatt-Saas and Erro-Tobbio HP-Serpentinites
Serpentinite samples were collected from the Zermatt-Saas HP ophiolite (Fig. S-1), which represents a relic of subducted oceanic lithosphere of the Jurassic Tethys Ocean (e.g., Rubatto et al., 1998
Rubatto, D., Gebauer, D., Fanning, M. (1998) Jurassic formation and Eocene subduction of the Zermatt–Saas-Fee ophiolites: implications for the geodynamic evolution of the Central and Western Alps. Contributions to Mineralogy and Petrology 132, 269–287. https://doi.org/10.1007/s004100050421
). The former harzburgites were serpentinised at the ocean floor and reached peak metamorphic conditions of 550–600 °C and 2.2–2.5 GPa during the Eocene subduction (e.g., Kempf et al., 2020Kempf, E.D., Hermann, J., Reusser, E., Baumgartner, L.P., Lanari, P. (2020) The role of the antigorite + brucite to olivine reaction in subducted serpentinites (Zermatt, Switzerland). Swiss Journal of Geosciences 113, 16. https://doi.org/10.1186/s00015-020-00368-0
and references therein). The field relations, petrology, major element, trace element and O isotopic compositions of the samples have been presented in Ulrich et al. (2024)Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
and are summarised in Table S-1. The antigorite serpentinite samples are from two nearby sites, the Lower and the Upper Theodul Glacier (LTG and UTG, respectively). Antigorite from the two sites differ in arsenic (As), antimony (Sb) and B concentrations and the O isotopic composition (Ulrich et al., 2024Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
). The positive B-(As-Sb)-δ18O correlation within each site has been interpreted as temperature dependent B uptake during oceanic serpentinisation (Ulrich et al., 2024Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
). Samples contain from approximately 25–40 vol. % metamorphic olivine produced in situ by the brucite-out reaction (Atg + Brc = Ol + fluid) to up to 90 vol. % metamorphic olivine produced during subsequent reactive fluid flow in shear bands, shear zones and veins. Where olivine is produced in situ by the brucite-out reaction, olivine and antigorite are in O isotopic equilibrium (Ulrich et al., 2024Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
). On the other hand, olivine in shear bands, shear zones and veins is in O isotopic equilibrium with antigorite in samples from the UTG outcrop, but in O isotopic disequilibrium with antigorite in samples from the LTG outcrop, indicating channelling of an external serpentinite-derived fluid with low δ18O between +5 and +6 ‰ (Ulrich et al., 2024Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
).The serpentinites from the Erro-Tobbio metaophiolite complex are also former harzburgites from the Jurassic Tethys oceanic lithosphere that reached peak metamorphic conditions of 500–600 °C and 1.8–2.5 GPa (Scambelluri et al., 1991
Scambelluri, M., Hoogerduijn Strating, E.H., Piccardo, G.B., Vissers, R.L.M., Rampone, E. (1991) Alpine olivine‐ and titanian clinohumite‐bearing assemblages in the Erro‐Tobbio peridotite (Voltri Massif, NW Italy). Journal of Metamorphic Geology 9, 79–91. https://doi.org/10.1111/j.1525-1314.1991.tb00505.x
). Here, we report in situ δ11B of antigorite, whereas metamorphic olivine in situ δ11B data are from Vesin and Ulrich et al. (2025)Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
. Metamorphic olivine was analysed from an HP-vein and an olivine shear band, both consisting of about 90 vol. % olivine and showing O isotopic equilibrium with coexisting antigorite (Vesin and Ulrich et al., 2025Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
). More details about field relations, petrology, in situ trace element concentrations, and in situ O isotopic compositions of the samples are in Vesin and Ulrich et al. (2025)Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
and Table S-1. Analytical details regarding SIMS analyses are in Section 1 of the Supplementary Information. The δ11B analyses of the matrix-matched reference materials are given in Tables S-2a and S-2b, and the δ11B data and SIMS spots locations are given in Table S-3a and Figure S-2.top
δ11B of Subducted Serpentine Inherited from Ocean Floor Alteration
The two Zermatt-Saas HP ophiolite localities UTG and LTG, which are suggested to have experienced different serpentinisation conditions based on trace elements and O isotopes (Ulrich et al., 2024
Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
), contain antigorite with two distinct δ11B signatures (Fig. 1). Samples from the UTG outcrop have antigorite δ11B average values of +13 to +19 ‰ (Table S-3b). These high δ11B values overlap with literature data from oceanic, forearc, and other subducted serpentinites (Fig. 1). Antigorite from the LTG outcrop has significantly lower δ11B values, mostly ranging between −5 and +10 ‰ (sample δ11B averages range from −2 to +5 ‰ except for sample ZS-32). Such low values are comparable to mantle wedge serpentinites that have experienced interaction with slab-derived metamorphic fluids from altered ocean crust and/or sediments (Yamada et al., 2019Yamada, C., Tsujimori, T., Chang, Q., Kimura, J.-I. (2019) Boron isotope variations of Franciscan serpentinites, northern California. Lithos 334–335, 180–189. https://doi.org/10.1016/j.lithos.2019.02.004
; Martin et al., 2020Martin, C., Flores, K.E., Vitale-Brovarone, A., Angiboust, S., Harlow, G.E. (2020) Deep mantle serpentinization in subduction zones: Insight from in situ B isotopes in slab and mantle wedge serpentinites. Chemical Geology 545, 119637. https://doi.org/10.1016/j.chemgeo.2020.119637
; De Hoog et al., 2023De Hoog, J.C.M., Clarke, E., Hattori, K. (2023) Mantle wedge olivine modifies slab-derived fluids: Implications for fluid transport from slab to arc magma source. Geology 51, 663–667. https://doi.org/10.1130/G51169.1
), even though the Zermatt-Saas serpentinites are clearly of oceanic origin (e.g., Kempf et al., 2020Kempf, E.D., Hermann, J., Reusser, E., Baumgartner, L.P., Lanari, P. (2020) The role of the antigorite + brucite to olivine reaction in subducted serpentinites (Zermatt, Switzerland). Swiss Journal of Geosciences 113, 16. https://doi.org/10.1186/s00015-020-00368-0
).Figure 1 Boron concentration versus δ11B of serpentinites from various settings (modified from Martin et al., 2020
Martin, C., Flores, K.E., Vitale-Brovarone, A., Angiboust, S., Harlow, G.E. (2020) Deep mantle serpentinization in subduction zones: Insight from in situ B isotopes in slab and mantle wedge serpentinites. Chemical Geology 545, 119637. https://doi.org/10.1016/j.chemgeo.2020.119637
). Boxes represent antigorite in situ δ11B data from this study, with B concentrations of Zermatt-Saas and Erro-Tobbio samples taken from Ulrich et al. (2024)Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
and from Vesin and Ulrich et al. (2025)Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
, respectively.The whole rock 87Sr/86Sr of the Zermatt-Saas samples (Supplementary Information Section 1 for analytical details and Table S-4 and Fig. S-3 for data) from both sites are mostly close to the Sr isotopic composition of Jurassic sea water. Some samples (mainly UTG-samples ZS-02, ZS-10 and ZS-03 with the exception of LTG-sample EK-MF) are slightly more radiogenic while having a higher δ11B antigorite. Sample ZS-02 with the highest 87Sr/86Sr values indicate a sediment contribution of maximum 2 %, which seems not to affect the B isotopic composition. The other samples show insignificant interaction between the subducted sediment end member and the HP serpentinites. Thus, we exclude that the low δ11B in antigorite of the LTG-samples is due to input from sediment-derived fluids.
Low (<+10 ‰) to negative δ11B values in subducted serpentinite may also be the result of B loss and associated depletion of heavy 11B during the lizardite to antigorite transition (Cannaò, 2020
Cannaò, E. (2020) Boron isotope fractionation in subducted serpentinites: A modelling attempt. Lithos 376–377, 105768. https://doi.org/10.1016/j.lithos.2020.105768
; Cannaò and Debret, 2024Cannaò, E., Debret, B. (2024) Variable δ11B signatures reflect dynamic evolution of the Mariana serpentinite forearc. Geochemical Perspectives Letters 30, 13–19. https://doi.org/10.7185/geochemlet.2416
). Our sample set does not provide evidence for B loss being responsible for the distinct δ11B in antigorite among the two Zermatt-Saas localities, as the B content in antigorite from the two sites overlaps. However, the B concentration in antigorite from each site shows a positive correlation with the As and Sb content and the O isotopic composition (Ulrich et al., 2024Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
), but an inverse correlation with its B isotopic composition. The correlation between B content and B isotopes is also observed for the two Erro-Tobbio samples, where antigorite with the higher δ11B (δ11B of +26 ± 2 ‰ versus +14 ± 2 ‰) contains less B and has lower δ18O (δ18O of +5.8 ± 0.4 ‰ versus +6.4 ± 0.6 ‰; Vesin and Ulrich et al., 2025Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
). Such negative correlation between δ11B and B content and furthermore O isotopes is also described by Vesin et al. (2024)Vesin, C., Rubatto, D., Pettke, T. (2024) The history of serpentinisation at mid-ocean ridges: Insights from in situ trace elements coupled with oxygen and boron isotopes. Chemical Geology 654, 122060. https://doi.org/10.1016/j.chemgeo.2024.122060
in oceanic serpentine from IODP samples (Mid Atlantic Ridge and Hess Deep).We therefore conclude that the different δ11B compositions of antigorite from the Zermatt-Saas and Erro-Tobbio samples are most likely inherited from compositional differences obtained during ocean floor serpentinisation because (i) the HP serpentine shows the same element isotope correlation as its oceanic equivalents, and (ii) low to negative δ11B values have been documented in oceanic serpentinites produced by oceanic hydrothermal vent fluids (Cannaò et al., 2024
Cannaò, E., Tiepolo, M., Agostini, S., Scambelluri, M. (2024) Fossil hydrothermal oceanic systems through in-situ B isotopes in ophicarbonates (N. Apennines, Italy). Chemical Geology 645, 121899. https://doi.org/10.1016/j.chemgeo.2023.121899
). We acknowledge that pH and oxidation state also can have an effect on the B isotopic composition, which, however, cannot be assessed with our data.top
δ11B in Metamorphic Olivine and Implications for B Isotopic Fractionation
The metamorphic olivine in the Zermatt-Saas samples shows different isotopic composition according to locality (Fig. 2): (i) UTG samples have δ11B olivine averages ranging from +12 ± 1 ‰ to +18 ± 2 ‰ (2 s.d.), coexisting with high δ11B antigorite (sample averages of +13 ± 2 ‰ to +19 ± 4 ‰); (ii) LTG samples contain olivine with even higher δ11B values (sample averages range from +27 to +31 ‰), coexisting with generally lower δ11B antigorite (−5 to +10 ‰, with one sample ranging up to +20 ‰). This results in different apparent mineral-mineral B fractionation (Δ11BOl-Atg = δ11BOl − δ11BAtg) for the two localities. By comparing Δ11BOl-Atg with Δ18OAtg-Ol from the same samples (Ulrich et al., 2024
Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
), we find that samples exhibiting olivine-antigorite O isotopic equilibrium generally have Δ11BOl-Atg near zero (−2.3 to +0.4 ‰ for individual samples with an overall average of −0.7 ± 3.7 ‰; Fig. 3a, Table S-3b).Figure 2 In situ δ11B of antigorite and olivine from the studied samples compared to the dataset of other HP serpentinites from the Western Alps (Zermatt-Saas, Aosta, Erro-Tobbio and Monviso, marked by *) from Clarke et al. (2020)
Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
. Boxes with dotted lines indicate samples from this study with olivine-antigorite O isotopic equilibrium (Ulrich et al., 2024Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
; Vesin and Ulrich et al., 2025Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
). Mineral abbreviations: Ol, olivine; Atg, antigorite.Figure 3 (a) Δ11BOl-Atg versus Δ18OAtg-Ol and (b) Δ11BOl-Atg versus DBOl/Atg (δ18O data and DBOl/Atg values from Ulrich et al., 2024
Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
). Bars for Δ11BOl-Atg indicate the range from the smallest possible value (δ11BOl min − δ11BAtg max) to the largest possible value (δ11BOl max − δ11BAtg min), while the sample average is indicated by the black dots. For Δ18OAtg-Ol and DB values, the sample average ± the propagated uncertainty (2 s.d.) are reported. The grey vertical band in (a) indicates the range of equilibrium Δ18OAtg-Ol at 550–600 °C, described in Ulrich et al. (2024)Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
. δ11B and DB datasets marked with an asterisk (*) in (b) are from Clarke et al. (2020)Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
.In the two Erro-Tobbio samples, metamorphic olivine is in O isotopic equilibrium with antigorite (Vesin and Ulrich et al., 2025
Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
) and has high δ11B of +13 ± 2 ‰ and +24 ± 1 ‰, respectively (Fig. 2). Antigorite in these samples also has high δ11B of +14 ± 2 ‰ and +26 ± 2 ‰, and thus Δ11BOl-Atg is near zero (−0.9 ± 2.8 ‰), as for the equilibrated Zermatt-Saas samples (Fig. 3a). This strengthens the conclusion that a near zero Δ11BOl-Atg value represents equilibrium B isotopic fractionation during serpentinite dehydration. A near-zero Δ11BOl-Atg (−0.7 ± 3.4 ‰, Zermatt-Saas and Erro-Tobbio samples combined) contrasts with studies that have inferred larger isotopic fractionation values between olivine and antigorite of +10 to +15 ‰ by considering olivine with B only in trigonal coordination (Cannaò, 2020Cannaò, E. (2020) Boron isotope fractionation in subducted serpentinites: A modelling attempt. Lithos 376–377, 105768. https://doi.org/10.1016/j.lithos.2020.105768
; Clarke et al., 2020Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
; Li et al., 2022Li, Y.-C., Wei, H.-Z., Palmer, M.R., Ma, J., Jiang, S.-Y., Chen, Y.-X., Lu, J.-J., Liu, X. (2022) Equilibrium boron isotope fractionation during serpentinization and applications in understanding subduction zone processes. Chemical Geology 609, 121047. https://doi.org/10.1016/j.chemgeo.2022.121047
). Our results conflict with these speciation models, but are supported by ab initio calculations by Muir et al. (2022Muir, J.M.R., Chen, Y., Liu, X., Zhang, F. (2022) Extremely Stable, Highly Conductive Boron-Hydrogen Complexes in Forsterite and Olivine. Journal of Geophysical Research: Solid Earth 127, e2022JB024299. https://doi.org/10.1029/2022JB024299
), which proposed that B in olivine is dominantly in tetrahedral coordination in water-saturated systems at subduction zone conditions, which, in turn, might lead to limited B isotope fractionation between olivine and serpentine. Our results are consistent with a tetrahedral coordination of B in olivine, although this still needs to be confirmed by crystallographic measurements.A large positive Δ11BOl-Atg of up to +25 ‰ is found in the Zermatt-Saas LTG samples (Fig. 3a), where metamorphic olivine and antigorite are not in O isotopic equilibrium. In these samples, olivine formation is attributed to the migration of 11B-enriched external fluids in shear bands, shear zones and veins that act as fluid channels (Ulrich et al., 2024
Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
). External in this context means a fluid derived from serpentinites that has a different O and B isotopic composition.Sample EK-FA is an exception to the general trend (Figs. 2 and 3a) as it shows a large Δ11BOl-Atg even when olivine and antigorite are in O isotopic equilibrium, and it is mineralogically similar to the other studied samples. A possible explanation is that the B isotopic system is more sensitive than O and could record in this sample a minor influence of the external fluids that were not sufficient to modify the isotopic composition of a major element such as O. An alternative explanation is that B in olivine in this sample is in trigonal instead of tetrahedral coordination, resulting in a larger positive Δ11BOl-Atg. However, the variability in Δ11BOl-Atg in sample EK-FA is quite large and rather suggests B isotopic disequilibrium. Furthermore, a different B coordination might be unlikely as P-T conditions, which have a main control on the B coordination, are similar to the other samples. Thus, we favour the first explanation.
In contrast, negative Δ11BOl-Atg values are reported in another unit of the Western Alps, the Monviso serpentinites, and are attributed to the influx of slab-derived mafic or sedimentary fluids during the olivine-in reaction (Clarke et al., 2020
Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
; Fig. 3b). The external fluids in the Monviso also explain the high B content in the metamorphic olivine, and, in turn, high olivine-antigorite B partitioning values (DBOl/Atg), exceeding the estimated sample-internal equilibrium range between 0.6 and 0.9 (Clarke et al., 2020Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
). DBOl/Atg values in our samples range from 0.5 to 3.1, regardless of whether olivine and antigorite are in O and apparent B isotopic equilibrium (Ulrich et al., 2024Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
; Fig. 3b). The variable DBOl/Atg values may be due to fluids passing through serpentinite rocks having variable B contents, while the isotopic composition might, or might not, be relatively homogeneous (referred to as internal and external migration, respectively). The migration of these fluids in shear zones and veins (open system) leads to B enrichment in metamorphic olivine.In summary, we propose that fluid production and internal fluid migration produces Δ11BOl-Atg values close to zero, whereas non-zero Δ11BOl-Atg values (LTG and Monviso) likely indicate fluid migration from other sections of the lithosphere, e.g., metasediments, altered ocean crust or serpentinite having a different isotopic composition.
We further conclude that, at the temperatures of the olivine-in reaction (500–600 °C; Scambelluri et al., 1991
Scambelluri, M., Hoogerduijn Strating, E.H., Piccardo, G.B., Vissers, R.L.M., Rampone, E. (1991) Alpine olivine‐ and titanian clinohumite‐bearing assemblages in the Erro‐Tobbio peridotite (Voltri Massif, NW Italy). Journal of Metamorphic Geology 9, 79–91. https://doi.org/10.1111/j.1525-1314.1991.tb00505.x
; Kempf et al., 2020Kempf, E.D., Hermann, J., Reusser, E., Baumgartner, L.P., Lanari, P. (2020) The role of the antigorite + brucite to olivine reaction in subducted serpentinites (Zermatt, Switzerland). Swiss Journal of Geosciences 113, 16. https://doi.org/10.1186/s00015-020-00368-0
), metamorphic olivine has a similar B isotopic composition as the subducted serpentine from which it forms, and that serpentinite dehydration will not result in significant changes in the B isotopic composition of the slab. Therefore, subducted serpentinites retain their (heterogeneous) δ11B sea floor signatures and their dehydration in subduction zones contributes to the heterogeneity of arc lavas.top
Acknowledgements
We thank Anne-Sophie Bouvier and Thomas Bovay for setting up the B isotope measurements of serpentine and olivine at the SwissSIMS facility. Enrico Cannaò, Jörg Hermann, Martin Wille, Qasid Ahmad, Sebastian Flöter and Jesse Walters are warmly thanked for insightful scientific discussions. We acknowledge the financial support by the Swiss National Science Foundation (project Nr. 191959) to D. Rubatto. Enrico Cannaò and an anonymous reviewer are thanked for their valuable comments. We further thank Horst Marschall for editorial handling.
Editor: Horst R. Marschall
top
References
Benton, L.D., Ryan, J.G., Tera, F. (2001) Boron isotope systematics of slab fluids as inferred from a serpentine seamount, Mariana forearc. Earth and Planetary Science Letters 187, 273–282. https://doi.org/10.1016/S0012-821X(01)00286-2
Boschi, C., Dini, A., Früh-Green, G.L., Kelley, D.S. (2008) Isotopic and element exchange during serpentinization and metasomatism at the Atlantis Massif (MAR 30°N): Insights from B and Sr isotope data. Geochimica et Cosmochimica Acta 72, 1801–1823. https://doi.org/10.1016/j.gca.2008.01.013
Boschi, C., Bonatti, E., Ligi, M., Brunelli, D., Cipriani, A., Dallai, L., D’Orazio, M., Früh-Green, G.L., Tonarini, S., Barnes, J.D., Bedini, R.M. (2013) Serpentinization of mantle peridotites along an uplifted lithospheric section, Mid Atlantic Ridge at 11° N. Lithos 178, 3–23. https://doi.org/10.1016/j.lithos.2013.06.003
Cannaò, E. (2020) Boron isotope fractionation in subducted serpentinites: A modelling attempt. Lithos 376–377, 105768. https://doi.org/10.1016/j.lithos.2020.105768
Show in context The extent to which B is retained or lost during dehydration at the lizardite to antigorite transition and at the olivine-in reaction, and how these reactions affect the fractionation of B isotopes, is still controversial (e.g., Scambelluri and Tonarini, 2012; De Hoog et al., 2014; Harvey et al., 2014; Cannaò, 2020).
View in article
Low (<+10 ‰) to negative δ11B values in subducted serpentinite may also be the result of B loss and associated depletion of heavy 11B during the lizardite to antigorite transition (Cannaò, 2020; Cannaò and Debret, 2024).
View in article
A near-zero Δ11BOl-Atg (−0.7 ± 3.4 ‰, Zermatt-Saas and Erro-Tobbio samples combined) contrasts with studies that have inferred larger isotopic fractionation values between olivine and antigorite of +10 to +15 ‰ by considering olivine with B only in trigonal coordination (Cannaò, 2020; Clarke et al., 2020; Li et al., 2022).
View in article
Cannaò, E., Debret, B. (2024) Variable δ11B signatures reflect dynamic evolution of the Mariana serpentinite forearc. Geochemical Perspectives Letters 30, 13–19. https://doi.org/10.7185/geochemlet.2416
Show in context Low (<+10 ‰) to negative δ11B values in subducted serpentinite may also be the result of B loss and associated depletion of heavy 11B during the lizardite to antigorite transition (Cannaò, 2020; Cannaò and Debret, 2024).
View in article
Cannaò, E., Tiepolo, M., Agostini, S., Scambelluri, M. (2024) Fossil hydrothermal oceanic systems through in-situ B isotopes in ophicarbonates (N. Apennines, Italy). Chemical Geology 645, 121899. https://doi.org/10.1016/j.chemgeo.2023.121899
Show in context However, negative δ11B values in serpentine from ophicarbonates of oceanic derivation (Cannaò et al., 2024) demonstrate how heterogeneous the input in subduction zones can be.
View in article
We therefore conclude that the different δ11B compositions of antigorite from the Zermatt-Saas and Erro-Tobbio samples are most likely inherited from compositional differences obtained during ocean floor serpentinisation because (i) the HP serpentine shows the same element isotope correlation as its oceanic equivalents, and (ii) low to negative δ11B values have been documented in oceanic serpentinites produced by oceanic hydrothermal vent fluids (Cannaò et al., 2024).
View in article
Clarke, E., De Hoog, J.C.M., Kirstein, L.A., Harvey, J., Debret, B. (2020) Metamorphic olivine records external fluid infiltration during serpentinite dehydration. Geochemical Perspectives Letters 16, 25–29. https://doi.org/10.7185/geochemlet.2039
Show in context Previous work has shown that dehydration of former ocean floor serpentinites during subduction can be a complex process possibly involving external fluids, which can modify the B isotopic signature and lead to isotopic disequilibrium between serpentinite residue and newly formed minerals, such as metamorphic olivine (Clarke et al., 2020).
View in article
Therefore, a microscale approach is necessary to decipher the extent to which B incorporation into metamorphic olivine fractionates B isotopes, and to evaluate whether an external fluid is involved (e.g., De Hoog et al., 2014; Clarke et al., 2020).
View in article
Metamorphic olivine, produced during the breakdown of brucite and the serpentine polytype antigorite, can incorporate high contents of B compared to primary mantle olivine (up to 100 μg/g versus <0.11 μg/g; Clarke et al., 2020 and references therein).
View in article
In situ δ11B of antigorite and olivine from the studied samples compared to the dataset of other HP serpentinites from the Western Alps (Zermatt-Saas, Aosta, Erro-Tobbio and Monviso, marked by *) from Clarke et al. (2020).
View in article
The grey vertical band in (a) indicates the range of equilibrium Δ18OAtg-Ol at 550–600 °C, described in Ulrich et al. (2024). δ11B and D B datasets marked with an asterisk (*) in (b) are from Clarke et al. (2020).
View in article
A near-zero Δ11BOl-Atg (−0.7 ± 3.4 ‰, Zermatt-Saas and Erro-Tobbio samples combined) contrasts with studies that have inferred larger isotopic fractionation values between olivine and antigorite of +10 to +15 ‰ by considering olivine with B only in trigonal coordination (Cannaò, 2020; Clarke et al., 2020; Li et al., 2022).
View in article
In contrast, negative Δ11BOl-Atg values are reported in another unit of the Western Alps, the Monviso serpentinites, and are attributed to the influx of slab-derived mafic or sedimentary fluids during the olivine-in reaction (Clarke et al., 2020; Fig. 3b).
View in article
The external fluids in the Monviso also explain the high B content in the metamorphic olivine, and, in turn, high olivine-antigorite B partitioning values (D B Ol/Atg), exceeding the estimated sample-internal equilibrium range between 0.6 and 0.9 (Clarke et al., 2020).
View in article
De Hoog, J.C.M., Savov, I.P. (2018) Boron Isotopes as a Tracer of Subduction Zone Processes. In: Marschall, H., Foster, G. (Eds.) Boron Isotopes: The Fifth Element. Springer, Cham, 217–247. https://doi.org/10.1007/978-3-319-64666-4_9
Show in context The B isotopic composition of ocean floor serpentinites varies depending on the serpentinisation conditions from δ11B = +5 to +40 ‰ (De Hoog and Savov, 2018 and references therein), which is higher compared to the upper mantle value of −7.1 ± 0.9 ‰ (Marschall et al., 2017).
View in article
De Hoog, J.C.M., Hattori, K., Jung, H. (2014) Titanium- and water-rich metamorphic olivine in high-pressure serpentinites from the Voltri Massif (Ligurian Alps, Italy): evidence for deep subduction of high-field strength and fluid-mobile elements. Contributions to Mineralogy and Petrology 167, 990. https://doi.org/10.1007/s00410-014-0990-x
Show in context Therefore, a microscale approach is necessary to decipher the extent to which B incorporation into metamorphic olivine fractionates B isotopes, and to evaluate whether an external fluid is involved (e.g., De Hoog et al., 2014; Clarke et al., 2020).
View in article
The extent to which B is retained or lost during dehydration at the lizardite to antigorite transition and at the olivine-in reaction, and how these reactions affect the fractionation of B isotopes, is still controversial (e.g., Scambelluri and Tonarini, 2012; De Hoog et al., 2014; Harvey et al., 2014; Cannaò, 2020).
View in article
De Hoog, J.C.M., Clarke, E., Hattori, K. (2023) Mantle wedge olivine modifies slab-derived fluids: Implications for fluid transport from slab to arc magma source. Geology 51, 663–667. https://doi.org/10.1130/G51169.1
Show in context These relatively low δ11B values (δ11B < +10 ‰) likely represent serpentinisation of the mantle wedge by crustal fluids derived from the subducting slab (e.g., Yamada et al., 2019; Martin et al., 2020; De Hoog et al., 2023).
View in article
Such low values are comparable to mantle wedge serpentinites that have experienced interaction with slab-derived metamorphic fluids from altered ocean crust and/or sediments (Yamada et al., 2019; Martin et al., 2020; De Hoog et al., 2023), even though the Zermatt-Saas serpentinites are clearly of oceanic origin (e.g., Kempf et al., 2020).
View in article
Harvey, J., Garrido, C.J., Savov, I., Agostini, S., Padrón-Navarta, J.A., Marchesi, C., López Sánchez-Vizcaíno, V., Gómez-Pugnaire, M.T. (2014) 11B-rich fluids in subduction zones: The role of antigorite dehydration in subducting slabs and boron isotope heterogeneity in the mantle. Chemical Geology 376, 20–30. https://doi.org/10.1016/j.chemgeo.2014.03.015
Show in context The extent to which B is retained or lost during dehydration at the lizardite to antigorite transition and at the olivine-in reaction, and how these reactions affect the fractionation of B isotopes, is still controversial (e.g., Scambelluri and Tonarini, 2012; De Hoog et al., 2014; Harvey et al., 2014; Cannaò, 2020).
View in article
Ingrin, J., Kovàcs, I., Deloule, E., Balan, E., Blanchard, M., Kohn, S.C., Hermann, J. (2014) Identification of hydrogen defects linked to boron substitution in synthetic forsterite and natural olivine. American Mineralogist 99, 2138–2141. https://doi.org/10.2138/am-2014-5049
Show in context Boron in serpentine substitutes for Si4+ and Al3+ in predominantly tetrahedral coordination (Pabst et al., 2011), whereas B in olivine is thought to be present as B3+ in predominantly trigonal coordination, substituting for Si4+ coupled to an O vacancy (Ingrin et al., 2014).
View in article
Kempf, E.D., Hermann, J., Reusser, E., Baumgartner, L.P., Lanari, P. (2020) The role of the antigorite + brucite to olivine reaction in subducted serpentinites (Zermatt, Switzerland). Swiss Journal of Geosciences 113, 16. https://doi.org/10.1186/s00015-020-00368-0
Show in context The former harzburgites were serpentinised at the ocean floor and reached peak metamorphic conditions of 550–600 °C and 2.2–2.5 GPa during the Eocene subduction (e.g., Kempf et al., 2020 and references therein).
View in article
Such low values are comparable to mantle wedge serpentinites that have experienced interaction with slab-derived metamorphic fluids from altered ocean crust and/or sediments (Yamada et al., 2019; Martin et al., 2020; De Hoog et al., 2023), even though the Zermatt-Saas serpentinites are clearly of oceanic origin (e.g., Kempf et al., 2020).
View in article
We further conclude that, at the temperatures of the olivine-in reaction (500–600 °C; Scambelluri et al., 1991; Kempf et al., 2020), metamorphic olivine has a similar B isotopic composition as the subducted serpentine from which it forms, and that serpentinite dehydration will not result in significant changes in the B isotopic composition of the slab.
View in article
Kowalski, P.M., Wunder, B., Jahn, S. (2013) Ab initio prediction of equilibrium boron isotope fractionation between minerals and aqueous fluids at high P and T. Geochimica et Cosmochimica Acta 101, 285–301. https://doi.org/10.1016/j.gca.2012.10.007
Show in context As the heavy B isotope (11B) is preferentially incorporated in trigonal and the light B isotope (10B) in tetrahedral coordination (Kowalski et al., 2013), olivine is expected to have a heavier B isotopic composition than the coexisting serpentine.
View in article
Li, Y.-C., Wei, H.-Z., Palmer, M.R., Ma, J., Jiang, S.-Y., Chen, Y.-X., Lu, J.-J., Liu, X. (2022) Equilibrium boron isotope fractionation during serpentinization and applications in understanding subduction zone processes. Chemical Geology 609, 121047. https://doi.org/10.1016/j.chemgeo.2022.121047
Show in context This might impact the estimates on B isotopic equilibrium fractionation, which are currently based on the assumption of B in olivine being trigonally and in serpentine being tetrahedrally coordinated (e.g., Li et al., 2022).
View in article
A near-zero Δ11BOl-Atg (−0.7 ± 3.4 ‰, Zermatt-Saas and Erro-Tobbio samples combined) contrasts with studies that have inferred larger isotopic fractionation values between olivine and antigorite of +10 to +15 ‰ by considering olivine with B only in trigonal coordination (Cannaò, 2020; Clarke et al., 2020; Li et al., 2022).
View in article
Marschall, H.R., Wanless, V.D., Shimizu, N., Pogge von Strandmann, P.A.E., Elliott, T., Monteleone, B.D. (2017) The boron and lithium isotopic composition of mid-ocean ridge basalts and the mantle. Geochimica et Cosmochimica Acta 207, 102–138. https://doi.org/10.1016/j.gca.2017.03.028
Show in context The B isotopic composition of ocean floor serpentinites varies depending on the serpentinisation conditions from δ11B = +5 to +40 ‰ (De Hoog and Savov, 2018 and references therein), which is higher compared to the upper mantle value of −7.1 ± 0.9 ‰ (Marschall et al., 2017).
View in article
Martin, C., Flores, K.E., Vitale-Brovarone, A., Angiboust, S., Harlow, G.E. (2020) Deep mantle serpentinization in subduction zones: Insight from in situ B isotopes in slab and mantle wedge serpentinites. Chemical Geology 545, 119637. https://doi.org/10.1016/j.chemgeo.2020.119637
Show in context Subducted high pressure (HP) serpentinites (also referred to as slab serpentinites) generally maintain a high δ11B (>+10 ‰), whereas mantle wedge serpentinites show an isotopically lighter range of compositions with δ11B = −14 to +10 ‰ (Martin et al., 2020).
View in article
These relatively low δ11B values (δ11B < +10 ‰) likely represent serpentinisation of the mantle wedge by crustal fluids derived from the subducting slab (e.g., Yamada et al., 2019; Martin et al., 2020; De Hoog et al., 2023).
View in article
Such low values are comparable to mantle wedge serpentinites that have experienced interaction with slab-derived metamorphic fluids from altered ocean crust and/or sediments (Yamada et al., 2019; Martin et al., 2020; De Hoog et al., 2023), even though the Zermatt-Saas serpentinites are clearly of oceanic origin (e.g., Kempf et al., 2020).
View in article
Boron concentration versus δ11B of serpentinites from various settings (modified from Martin et al., 2020).
View in article
Muir, J.M.R., Chen, Y., Liu, X., Zhang, F. (2022) Extremely Stable, Highly Conductive Boron-Hydrogen Complexes in Forsterite and Olivine. Journal of Geophysical Research: Solid Earth 127, e2022JB024299. https://doi.org/10.1029/2022JB024299
Show in context However, recent quantum mechanical modelling suggests that, in water-rich environments (such as dehydrating serpentinites), tetrahedral B-Si-H complexes may dominate in olivine (Muir et al., 2022), although B in tetrahedral coordination in olivine has not yet been documented in natural samples.
View in article
Our results conflict with these speciation models, but are supported by ab initio calculations by Muir et al. (2022), which proposed that B in olivine is dominantly in tetrahedral coordination in water-saturated systems at subduction zone conditions, which, in turn, might lead to limited B isotope fractionation between olivine and serpentine.
View in article
Pabst, S., Zack, T., Savov, I.P., Ludwig, T., Rost, D., Vicenzi, E.P. (2011) Evidence for boron incorporation into the serpentine crystal structure. American Mineralogist 96, 1112–1119. https://doi.org/10.2138/am.2011.3709
Show in context Boron in serpentine substitutes for Si4+ and Al3+ in predominantly tetrahedral coordination (Pabst et al., 2011), whereas B in olivine is thought to be present as B3+ in predominantly trigonal coordination, substituting for Si4+ coupled to an O vacancy (Ingrin et al., 2014).
View in article
Rubatto, D., Gebauer, D., Fanning, M. (1998) Jurassic formation and Eocene subduction of the Zermatt–Saas-Fee ophiolites: implications for the geodynamic evolution of the Central and Western Alps. Contributions to Mineralogy and Petrology 132, 269–287. https://doi.org/10.1007/s004100050421
Show in context Serpentinite samples were collected from the Zermatt-Saas HP ophiolite (Fig. S-1), which represents a relic of subducted oceanic lithosphere of the Jurassic Tethys Ocean (e.g., Rubatto et al., 1998).
View in article
Scambelluri, M., Tonarini, S. (2012) Boron isotope evidence for shallow fluid transfer across subduction zones by serpentinized mantle. Geology 40, 907–910. https://doi.org/10.1130/G33233.1
Show in context The extent to which B is retained or lost during dehydration at the lizardite to antigorite transition and at the olivine-in reaction, and how these reactions affect the fractionation of B isotopes, is still controversial (e.g., Scambelluri and Tonarini, 2012; De Hoog et al., 2014; Harvey et al., 2014; Cannaò, 2020).
View in article
Scambelluri, M., Hoogerduijn Strating, E.H., Piccardo, G.B., Vissers, R.L.M., Rampone, E. (1991) Alpine olivine‐ and titanian clinohumite‐bearing assemblages in the Erro‐Tobbio peridotite (Voltri Massif, NW Italy). Journal of Metamorphic Geology 9, 79–91. https://doi.org/10.1111/j.1525-1314.1991.tb00505.x
Show in context The serpentinites from the Erro-Tobbio metaophiolite complex are also former harzburgites from the Jurassic Tethys oceanic lithosphere that reached peak metamorphic conditions of 500–600 °C and 1.8–2.5 GPa (Scambelluri et al., 1991).
View in article
We further conclude that, at the temperatures of the olivine-in reaction (500–600 °C; Scambelluri et al., 1991; Kempf et al., 2020), metamorphic olivine has a similar B isotopic composition as the subducted serpentine from which it forms, and that serpentinite dehydration will not result in significant changes in the B isotopic composition of the slab.
View in article
Spivack, A.J., Edmond, J.M. (1987) Boron isotope exchange between seawater and the oceanic crust. Geochimica et Cosmochimica Acta 51, 1033–1043. https://doi.org/10.1016/0016-7037(87)90198-0
Tonarini, S., Leeman, W.P., Leat, P.T. (2011) Subduction erosion of forearc mantle wedge implicated in the genesis of the South Sandwich Island (SSI) arc: Evidence from boron isotope systematics. Earth and Planetary Science Letters 301, 275–284. https://doi.org/10.1016/j.epsl.2010.11.008
Ulmer, P., Trommsdorff, V. (1995) Serpentine Stability to Mantle Depths and Subduction-Related Magmatism. Science 268, 858–861. https://doi.org/10.1126/science.268.5212.858
Show in context Fluid transfer at mantle depths occurs mainly in subduction zones where serpentinites dehydrate and release aqueous fluid into the mantle wedge (Ulmer and Trommsdorff, 1995).
View in article
Ulrich, M., Rubatto, D., Hermann, J., Markmann, T.A., Bouvier, A.-S., Deloule, E. (2024) Olivine formation processes and fluid pathways in subducted serpentinites revealed by in-situ oxygen isotope analysis (Zermatt-Saas, Switzerland). Chemical Geology 649, 121978. https://doi.org/10.1016/j.chemgeo.2024.121978
Show in context The δ11B data obtained on antigorite and olivine are coupled with their O isotopic compositions and B contents (Ulrich et al., 2024; Vesin and Ulrich et al., 2025) to elucidate the fate of B and its isotopes during serpentinite dehydration.
View in article
The field relations, petrology, major element, trace element and O isotopic compositions of the samples have been presented in Ulrich et al. (2024) and are summarised in Table S-1
View in article
Antigorite from the two sites differ in arsenic (As), antimony (Sb) and B concentrations and the O isotopic composition (Ulrich et al., 2024).
View in article
The positive B-(As-Sb)-δ18O correlation within each site has been interpreted as temperature dependent B uptake during oceanic serpentinisation (Ulrich et al., 2024).
View in article
Where olivine is produced in situ by the brucite-out reaction, olivine and antigorite are in O isotopic equilibrium (Ulrich et al., 2024).
View in article
On the other hand, olivine in shear bands, shear zones and veins is in O isotopic equilibrium with antigorite in samples from the UTG outcrop, but in O isotopic disequilibrium with antigorite in samples from the LTG outcrop, indicating channelling of an external serpentinite-derived fluid with low δ18O between +5 and +6 ‰ (Ulrich et al., 2024).
View in article
The two Zermatt-Saas HP ophiolite localities UTG and LTG, which are suggested to have experienced different serpentinisation conditions based on trace elements and O isotopes (Ulrich et al., 2024), contain antigorite with two distinct δ11B signatures (Fig. 1).
View in article
Boxes represent antigorite in situ δ11B data from this study, with B concentrations of Zermatt-Saas and Erro-Tobbio samples taken from Ulrich et al. (2024) and from Vesin and Ulrich et al. (2025), respectively.
View in article
However, the B concentration in antigorite from each site shows a positive correlation with the As and Sb content and the O isotopic composition (Ulrich et al., 2024), but an inverse correlation with its B isotopic composition.
View in article
By comparing Δ11BOl-Atg with Δ18OAtg-Ol from the same samples (Ulrich et al., 2024), we find that samples exhibiting olivine-antigorite O isotopic equilibrium generally have Δ11BOl-Atg near zero (−2.3 to +0.4 ‰ for individual samples with an overall average of −0.7 ± 3.7 ‰; Fig. 3a, Table S-3b).
View in article
Boxes with dotted lines indicate samples from this study with olivine-antigorite O isotopic equilibrium (Ulrich et al., 2024; Vesin and Ulrich et al., 2025).
View in article
(a) Δ11BOl-Atg versus Δ18OAtg-Ol and (b) Δ11BOl-Atg versus DBOl/Atg (δ18O data and DBOl/Atg values from Ulrich et al., 2024).
View in article
The grey vertical band in (a) indicates the range of equilibrium Δ18OAtg-Ol at 550–600 °C, described in Ulrich et al. (2024). δ11B and DB datasets marked with an asterisk (*) in (b) are from Clarke et al. (2020).
View in article
In these samples, olivine formation is attributed to the migration of 11B-enriched external fluids in shear bands, shear zones and veins that act as fluid channels (Ulrich et al., 2024).
View in article
DBOl/Atg values in our samples range from 0.5 to 3.1, regardless of whether olivine and antigorite are in O and apparent B isotopic equilibrium (Ulrich et al., 2024; Fig. 3b).
View in article
Vesin, C., Rubatto, D., Pettke, T. (2024) The history of serpentinisation at mid-ocean ridges: Insights from in situ trace elements coupled with oxygen and boron isotopes. Chemical Geology 654, 122060. https://doi.org/10.1016/j.chemgeo.2024.122060
Show in context Such negative correlation between δ11B and B content and furthermore O isotopes is also described by Vesin et al. (2024) in oceanic serpentine from IODP samples (Mid Atlantic Ridge and Hess Deep).
View in article
Vesin, C., Ulrich, M., Rubatto, D., Hermann, J., Scambelluri, M. (2025) Chemical and isotopic exchanges in serpentinites during hydration and dehydration events in the Erro-Tobbio Massif (Italy): from ocean to subduction. Journal of Petrology 66, egaf088. https://doi.org/10.1093/petrology/egaf008
Show in context The δ11B data obtained on antigorite and olivine are coupled with their O isotopic compositions and B contents (Ulrich et al., 2024; Vesin and Ulrich et al., 2025) to elucidate the fate of B and its isotopes during serpentinite dehydration.
View in article
Here, we report in situ δ11B of antigorite, whereas metamorphic olivine in situ δ11B data are from Vesin and Ulrich et al. (2025).
View in article
Metamorphic olivine was analysed from an HP-vein and an olivine shear band, both consisting of about 90 vol. % olivine and showing O isotopic equilibrium with coexisting antigorite (Vesin and Ulrich et al., 2025).
View in article
More details about field relations, petrology, in situ trace element concentrations, and in situ O isotopic compositions of the samples are in Vesin and Ulrich et al. (2025) and Table S-1
View in article
Boxes represent antigorite in situ δ11B data from this study, with B concentrations of Zermatt-Saas and Erro-Tobbio samples taken from Ulrich et al. (2024) and from Vesin and Ulrich et al. (2025), respectively.
View in article
The correlation between B content and B isotopes is also observed for the two Erro-Tobbio samples, where antigorite with the higher δ11B (δ11B of +26 ± 2 ‰ versus +14 ± 2 ‰) contains less B and has lower δ18O (δ18O of +5.8 ± 0.4 ‰ versus +6.4 ± 0.6 ‰; Vesin and Ulrich et al., 2025).
View in article
Boxes with dotted lines indicate samples from this study with olivine-antigorite O isotopic equilibrium (Ulrich et al., 2024; Vesin and Ulrich et al., 2025).
View in article
In the two Erro-Tobbio samples, metamorphic olivine is in O isotopic equilibrium with antigorite (Vesin and Ulrich et al., 2025) and has high δ11B of +13 ± 2 ‰ and +24 ± 1 ‰, respectively (Fig. 2).
View in article
Vils, F., Tonarini, S., Kalt, A., Seitz, H.-M. (2009) Boron, lithium and strontium isotopes as tracers of seawater–serpentinite interaction at Mid-Atlantic ridge, ODP Leg 209. Earth and Planetary Science Letters 286, 414–425. https://doi.org/10.1016/j.epsl.2009.07.005
Yamada, C., Tsujimori, T., Chang, Q., Kimura, J.-I. (2019) Boron isotope variations of Franciscan serpentinites, northern California. Lithos 334–335, 180–189. https://doi.org/10.1016/j.lithos.2019.02.004
Show in context These relatively low δ11B values (δ11B < +10 ‰) likely represent serpentinisation of the mantle wedge by crustal fluids derived from the subducting slab (e.g., Yamada et al., 2019; Martin et al., 2020; De Hoog et al., 2023).
View in article
Such low values are comparable to mantle wedge serpentinites that have experienced interaction with slab-derived metamorphic fluids from altered ocean crust and/or sediments (Yamada et al., 2019; Martin et al., 2020; De Hoog et al., 2023), even though the Zermatt-Saas serpentinites are clearly of oceanic origin (e.g., Kempf et al., 2020).
View in article
top
Supplementary Information
The Supplementary Information includes:
- Analytical Methods
- Supplementary Tables
- Supplementary Figures
- Supplementary Information References
Download the Supplementary Information (PDF)
