Chlorine isotope ratios record magmatic brine assimilation during rhyolite genesis
Affiliations | Corresponding Author | Cite as | Funding information
Published by the European Association of Geochemistry
under Creative Commons License CC BY-NC-ND 4.0
Keywords: chlorine isotopes, hydrosaline fluids, magmatic brine, magmatic volatiles, silicic magma, magma mush
- Share this article
-
Article views:599Cumulative count of HTML views and PDF downloads.
- Download Citation
- Rights & Permissions
top
Abstract
Figures
 Figure 1 Chlorine isotope variations vs. (a) Cl concentrations and (b) SiO2. Silicic rocks in Iceland have lower δ37Cl values than basalts (data from Halldórsson et al., 2016), overlapping with rhyolites from the Mono Craters, USA (Barnes et al., 2014). Arrows in (a) indicate the effects of the small equilibrium isotope fractionations caused by fractional crystallisation and degassing, and the large kinetic isotope fractionation during magmatic brine exsolution (Fortin et al., 2017) and assimilation on the δ37Cl and Cl composition of silicic melts. The gray curve in (a) shows the effect of magmatic brine assimilation (in wt. %) on the Cl-δ37Cl values of a hypothetical rhyolite melt with an average propagating rift basalt δ37Cl value of +0.7 ‰. The negative δ37Cl shifts between silicic rocks and basalts are illustrated in (b) by arrows anchored at the average SiO2 concentrations and δ37Cl values of the rift, propagating rift and off-rift basalts. The 1σ uncertainty is ±0.2 ‰ for δ37Cl. |  Figure 2 The δ18O-δ37Cl systematics of silicic rocks. Low δ18O values of rift-related silicic rocks result from assimilation of hydrothermally altered, δ18O-depleted crust. Lack of correlation between δ18O and δ37Cl indicates that negative δ37Cl values of rhyolites are not caused by crustal assimilation. The gray line shows the effect of magmatic brine assimilation in wt. % (Supplementary Information S-4). Mono Craters field is drawn after data reported in Newman et al. (1988) and Barnes et al. (2014). The 1σ uncertainty is ±0.2 ‰ for δ37Cl, and smaller than the size of the symbols for δ18O. |  Figure 3 (a) A model of magmatic brine formation and assimilation in a long lived upper crustal magma mush. (b) Magmatic fluids exsolve from silicic melts during late stage crystallisation and acquire negative δ37Cl values through kinetic fractionation (Fortin et al., 2017). (c) Decompression-driven phase separation of a supercritical fluid produces a NaCl-rich brine and a NaCl-poor vapour with a maximum fractionation of Δ37Clliquid-vapour = ±0.5 ‰ (Liebscher et al., 2006). |
| Figure 1 | Figure 2 | Figure 3 |
top
Introduction
Magmatic volatiles play a fundamental part during silicic magma genesis and the formation of associated ore deposits. Chlorine is among the most abundant volatile elements in igneous rocks and may become concentrated enough in late stage silicic melts to exsolve and form hydrosaline liquids, i.e. high density Cl-enriched aqueous fluids or hydrosaline brines (Webster, 2004
Webster, J.D. (2004) The exsolution of magmatic hydrosaline chloride liquids. Chemical Geology 210, 33–48.
). As chlorine is a hydrophile element, its isotopic fingerprint has been used to trace volatile sources in igneous rocks and hydrothermal fluids (Barnes et al., 2008Barnes, J.D., Sharp, Z.D., Fischer, T.P. (2008) Chlorine isotope variations across the Izu-Bonin-Mariana arc. Geology 36, 883–886.
; Li et al., 2015Li, L., Bonifacie, M., Aubaud, C., Crispi, O., Dessert, C., Agrinier, P. (2015) Chlorine isotopes of thermal springs in arc volcanoes for tracing shallow magmatic activity. Earth and Planetary Science Letters 413, 101–110.
). Lavas associated with subduction zones and oceanic islands have a range of δ37Cl values from −3 to +3 ‰, likely due to incorporation of subduction fluids, recycled marine sediments and altered oceanic crust into the mantle (John et al., 2010John, T., Layne, G.D., Haase, K.M., Barnes, J.D. (2010) Chlorine isotope evidence for crustal recycling into the Earth’s mantle. Earth and Planetary Science Letters 298, 175–182.
; Barnes and Sharp, 2017Barnes, J.D., Sharp, Z.D. (2017) Chlorine isotope geochemistry. Reviews in Mineralogy and Geochemistry 82, 345–378.
). In contrast, the depleted upper mantle (DMM) has a restricted δ37Cl variability of −0.2 ± 0.3 ‰ (Sharp et al., 2013Sharp, Z.D., Mercer, J.A., Jones, R.H., Brearley, A.J., Selverstone, J., Bekker, A., Stachel, T. (2013) The chlorine isotope composition of chondrites and Earth. Geochimica et Cosmochimica Acta 107, 189–204.
), reflecting the limited δ37Cl fractionation from high temperature magmatic processes (Schauble et al., 2003Schauble, E.A., Rossman, G.R., Taylor Jr., H.P. (2003) Theoretical estimates of equilibrium chlorine-isotope fractionations. Geochimica et Cosmochimica Acta 67, 3267–3281.
). The majority of chlorine isotope studies on igneous rocks have been conducted on basaltic rocks, whereas published δ37Cl data for silicic rocks is limited, with 40 out of 44 published analyses coming from a single volcanic system, the Mono Craters, USA (Barnes et al., 2014Barnes, J.D., Prather, T.J., Cisneros, M., Befus, K., Gardner, J.E., Larson, T.E. (2014) Stable chlorine isotope behavior during volcanic degassing of H2O and CO2 at Mono Craters, CA. Bulletin of Volcanology 76, 805.
). This study was designed to explore if and how δ37Cl systematics can provide new insights into silicic magmatic processes such as assimilation and brine-melt interaction, using Iceland as a test site.top
Chlorine Isotope Systematics in Silicic Rocks
We present new δ37Cl and δ18O data for a sample set (n = 16) focusing on neovolcanic extrusive silicic (SiO2 > 65 wt. %) and intermediate (SiO2 = 52-65 wt. %) rocks from Iceland (Tables S-1, S-2). Together with previously published δ37Cl and δ18O data on Icelandic basalts (Halldórsson et al., 2016
Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
), the samples represent the full chemical range between subalkaline-tholeiitic rift zone, and transitional to alkaline propagating rift and off-rift magma suites in Iceland (Fig. S-1), spanning a SiO2 range of 44.4-77.7 wt. % and Cl concentrations between 17 and 3988 ppm (Figs. 1, S-2). The samples cover the main types of silicic rocks in Iceland, i.e. dacites and alkaline and subalkaline rhyolites (Jónasson, 2007Jónasson, K. (2007) Silicic volcanism in Iceland: Composition and distribution within the active volcanic zones. Journal of Geodynamics 43, 101–117.
), and include both obsidians and tephras (i.e. products of effusive vs. explosive eruptions). All studied volcanoes are situated on land and are free of seawater influence (Halldórsson et al., 2016Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
).Figure 1 Chlorine isotope variations vs. (a) Cl concentrations and (b) SiO2. Silicic rocks in Iceland have lower δ37Cl values than basalts (data from Halldórsson et al., 2016
Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
), overlapping with rhyolites from the Mono Craters, USA (Barnes et al., 2014Barnes, J.D., Prather, T.J., Cisneros, M., Befus, K., Gardner, J.E., Larson, T.E. (2014) Stable chlorine isotope behavior during volcanic degassing of H2O and CO2 at Mono Craters, CA. Bulletin of Volcanology 76, 805.
). Arrows in (a) indicate the effects of the small equilibrium isotope fractionations caused by fractional crystallisation and degassing, and the large kinetic isotope fractionation during magmatic brine exsolution (Fortin et al., 2017Fortin, M.A., Watson, E.B., Stern, R. (2017) The isotope mass effect on chlorine diffusion in dacite melt, with implications for fractionation during bubble growth. Earth and Planetary Science Letters 480, 15–24.
) and assimilation on the δ37Cl and Cl composition of silicic melts. The gray curve in (a) shows the effect of magmatic brine assimilation (in wt. %) on the Cl-δ37Cl values of a hypothetical rhyolite melt with an average propagating rift basalt δ37Cl value of +0.7 ‰. The negative δ37Cl shifts between silicic rocks and basalts are illustrated in (b) by arrows anchored at the average SiO2 concentrations and δ37Cl values of the rift, propagating rift and off-rift basalts. The 1σ uncertainty is ±0.2 ‰ for δ37Cl.Significant Cl variation is present at any given SiO2 content in the basaltic (17–1269 ppm), intermediate (130–942 ppm) and silicic (282–3988 ppm) samples (Fig. S-2). These ranges are similar to published Cl concentrations in melt inclusions (MIs) from corresponding locations (Fig. S-3). The δ37Cl value of all analysed samples (n = 14) vary from −1.9 to +1.3 ‰ (1σ =±0.2 ‰) (Fig. 1a,b). The basaltic (n = 3) and intermediate (n = 4) samples have δ37Cl values between −0.4 and +1.3 ‰, overlapping with the known range of Icelandic basalts of −0.6 to +1.4 ‰ (Halldórsson et al., 2016
Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
). In contrast, the δ37Cl values of the silicic samples from this study (n = 8) and those previously published (n = 3; Halldórsson et al., 2016Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
) deviate from the basaltic-intermediate range towards more negative values of −1.9 to −0.6 ‰ (Fig. 1b), except for a single outlier (SAL-74) with δ37Cl = +0.9 ‰.Local δ37Cl variability in Icelandic rhyolites appears to be small (≤0.5 ‰ for Hekla: H3, H4, H5; and Askja: ASD1L, ASD14L) compared to the large range of −1.9 to 0.0 ‰ reported for the Mono Crater rhyolites (Barnes et al., 2014
Barnes, J.D., Prather, T.J., Cisneros, M., Befus, K., Gardner, J.E., Larson, T.E. (2014) Stable chlorine isotope behavior during volcanic degassing of H2O and CO2 at Mono Craters, CA. Bulletin of Volcanology 76, 805.
) (Fig. 1). Rift, propagating rift and off-rift samples define distinct fields in the SiO2-δ37Cl and δ18O-δ37Cl diagrams (Figs. 1b, 2), suggesting a possible correlation between volcano-tectonic setting and δ37Cl (see Supplementary Information S-2).Figure 2 The δ18O-δ37Cl systematics of silicic rocks. Low δ18O values of rift-related silicic rocks result from assimilation of hydrothermally altered, δ18O-depleted crust. Lack of correlation between δ18O and δ37Cl indicates that negative δ37Cl values of rhyolites are not caused by crustal assimilation. The gray line shows the effect of magmatic brine assimilation in wt. % (Supplementary Information S-4). Mono Craters field is drawn after data reported in Newman et al. (1988)
Newman, S., Epstein, S., Stolper, E. (1988) Water, carbon dioxide, and hydrogen isotopes in glasses from the ca. 1340 AD eruption of the Mono Craters, California: constraints on degassing phenomena and initial volatile content. Journal of Volcanology and Geothermal Research 35, 75–96.
and Barnes et al. (2014)Barnes, J.D., Prather, T.J., Cisneros, M., Befus, K., Gardner, J.E., Larson, T.E. (2014) Stable chlorine isotope behavior during volcanic degassing of H2O and CO2 at Mono Craters, CA. Bulletin of Volcanology 76, 805.
. The 1σ uncertainty is ±0.2 ‰ for δ37Cl, and smaller than the size of the symbols for δ18O.top
Origin of Large δ37Cl Variability: Sources Versus Processes
Our dataset demonstrates that silicic rocks in Iceland have more negative δ37Cl values relative to associated basalts and intermediate rocks. Whereas basalts inherit the δ37Cl signatures of their mantle sources (Halldórsson et al., 2016
Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
), the shift to more negative δ37Cl values in silicic rocks must reflect a process or a combination of processes taking place during rhyolite genesis, such as mineral-melt fractionation, degassing and/or assimilation.Rayleigh δ37Cl fractionations between HCl(g), minerals and silicic melt are expected to be small, based on theoretical equilibrium fractionation factors of Schauble et al. (2003)
Schauble, E.A., Rossman, G.R., Taylor Jr., H.P. (2003) Theoretical estimates of equilibrium chlorine-isotope fractionations. Geochimica et Cosmochimica Acta 67, 3267–3281.
extrapolated to magmatic temperatures (Δ37Clmineral-melt ≈ Δ37Clvapour-melt ≈ 0.2 ‰ at 600 °C). Thus, even extreme cases of 90 % Cl removal by fractional apatite crystallisation or open system degassing of HCl(g) only fractionate the δ37Clmelt values by about −0.5 ‰. However, modal apatite abundances in our samples are low (<2 %) and similar Cl concentrations in MIs and matrix glasses imply an insignificant degree of syn-eruptive chlorine degassing (Fig. S-3). Therefore, the combined effect of fractional crystallisation and degassing on δ37Clmelt values of our samples is negligible (<0.2 ‰). Moreover, similar δ37Cl values of both obsidians and tephras indicate that δ37Cl fractionation is independent of eruption type and occurs within the crustal magma domain prior to eruptions.To test if assimilation of altered basaltic crust causes negative δ37Cl shifts between rhyolites and basalts, we analysed the δ18O compositions of our samples (Fig. 2). In Iceland, low δ18Orock values relative to pristine basaltic values (+4.8 to +5.8 ‰; Thirlwall et al., 2006
Thirlwall, M.F., Gee, M.A.M., Lowry, D., Mattey, D.P., Murton, B.J., Taylor, R.N. (2006) Low δ18O in the Icelandic mantle and its origins: Evidence from Reykjanes Ridge and Icelandic lavas. Geochimica et Cosmochimica Acta 70, 993–1019.
) are used to recognise assimilation (or partial melting) of altered crust, which has been shifted to low δ18O signatures (≤+2 ‰) by hydrothermal alteration with low δ18O meteoric water (Gautason and Muehlenbachs, 1998Gautason, B., Muehlenbachs, K. (1998) Oxygen isotopic fluxes associated with high-temperature processes in the rift zones of Iceland. Chemical Geology 145, 275–286.
). We note that basalts and intermediate rocks from all three volcanic settings display δ18O values between +3.4 and +5.2 ‰ (Fig. 2), typical for Icelandic basalts (Thirlwall et al., 2006Thirlwall, M.F., Gee, M.A.M., Lowry, D., Mattey, D.P., Murton, B.J., Taylor, R.N. (2006) Low δ18O in the Icelandic mantle and its origins: Evidence from Reykjanes Ridge and Icelandic lavas. Geochimica et Cosmochimica Acta 70, 993–1019.
). Silicic rocks from the propagating rift and off-rift zones have basalt-like δ18O values of +4.0 to +6.1 ‰, whereas the lower and more variable δ18O values from −0.5 to +4.7 ‰ in the rift zone rhyolites (Fig. 2) indicate variable degrees of crustal assimilation.However, there is no correlation between δ37Cl and δ18O (Fig. 2). For example, silicic samples with the most negative (H3a) and positive δ37Cl values (SAL-74) have normal δ18O values, while the two samples with the lowest δ18O values (ASD1L, ASD14L) show relatively small δ37Cl shifts. This indicates that the negative δ37Cl shifts in Icelandic rhyolites are not caused by assimilation of δ18O-depleted altered crust, but by an additional process. Conversely, this suggests that hydrothermally altered crust in Iceland has a basalt-like δ37Cl range, consistent with the basalt-like δ37Cl values in Icelandic hydrothermal fluids (Stefánsson and Barnes, 2016
Stefánsson, A., Barnes, J.D. (2016) Chlorine isotope geochemistry of Icelandic thermal fluids: Implications for geothermal system behavior at divergent plate boundaries. Earth and Planetary Science Letters 449, 69–78.
) and the lack of δ37Cl fractionation resulting from hydrothermal alteration (Cullen et al., 2019Cullen, J.T., Hurwirz, S., Barnes, J.D., Lassiter, J.C., Penniston-Dorland, S., Kasemann, S.A., Thordsen, J.J. (2019) Temperature-Dependent variations in mineralogy, major element chemistry and the stable isotopes of boron, lithium and chlorine resulting from hydration of rhyolite: Constraints from hydrothermal experiments at 150 to 350° C and 25 MPa. Geochimica et Cosmochimica Acta 261, 269–287.
). In contrast, boron another fluid-mobile element, displays anomalous positive δ11B values in Icelandic silicic rocks that correlate with decreasing δ18O, and that have thus been explained by crustal assimilation (Rose-Koga and Sigmarsson, 2008Rose-Koga, E.F., Sigmarsson, O. (2008) B–O–Th isotope systematics in Icelandic tephra. Chemical Geology 255, 454–462.
).Extensive previous work on the Hekla volcano demonstrates that for non-volatile element stable isotope systems studied thus far, fractionations between rhyolites and basalts are either negligible or can be explained by fractional crystallisation (Supplementary Information S-4). Our silicic Hekla samples (H3a, H4, H5) display the largest δ37Cl shifts (up to −2.9 ‰) compared to corresponding basalts (Fig. 1). This comparison highlights that δ37Cl selectively records a process relating to the pre-eruptive volatile history of silicic magmas that is not recorded by other, non-volatile stable isotope systems. Indeed, a complicated pre-eruptive volatile history is also reflected by high Cl variability in Icelandic propagating rift and rift rhyolites (50 to 2600 ppm) (Fig. S-2), likely reflecting a combination of fractional crystallisation, partial melting, accumulation of fractional melts from volatile heterogeneous sources as well as episodic exsolution and resorption of magmatic volatile phases, including magmatic brines (Webster et al., 2019
Webster, J.D., Iveson, A.A., Rowe, M.C., Webster, P.M. (2019) Chlorine and felsic magma evolution: Modeling the behavior of an under-appreciated volatile component. Geochimica et Cosmochimica Acta 271, 248–288.
; Supplementary Information S-3; Fig. 3a).Figure 3 (a) A model of magmatic brine formation and assimilation in a long lived upper crustal magma mush. (b) Magmatic fluids exsolve from silicic melts during late stage crystallisation and acquire negative δ37Cl values through kinetic fractionation (Fortin et al., 2017
Fortin, M.A., Watson, E.B., Stern, R. (2017) The isotope mass effect on chlorine diffusion in dacite melt, with implications for fractionation during bubble growth. Earth and Planetary Science Letters 480, 15–24.
). (c) Decompression-driven phase separation of a supercritical fluid produces a NaCl-rich brine and a NaCl-poor vapour with a maximum fractionation of Δ37Clliquid-vapour = ±0.5 ‰ (Liebscher et al., 2006Liebscher, A., Barnes, J., Sharp, Z. (2006) Chlorine isotope vapor–liquid fractionation during experimental fluid-phase separation at 400 C/23 MPa to 450 C/42 MPa. Chemical Geology 234, 340–345.
).Chlorine isotope systematics provide a strict constraint on the nature of a potential assimilant, which must have a negative δ37Cl and elevated concentrations of Cl compared to the rhyolites. These criteria best match a fluid assimilant with high Cl concentrations (Classimilant/Clrhyolite >> 1) and low δ37Cl (<−3 ‰) (Fig S-4). Magmatic hydrosaline fluids have by definition high Cl concentrations and may acquire highly negative δ37Cl values during exsolution from dacitic (and more silicic) melts due to kinetic diffusion effects, that cause considerable fractionation of up to Δ37Clfluid-melt = −5 ‰ even at high temperatures (Fortin et al., 2017
Fortin, M.A., Watson, E.B., Stern, R. (2017) The isotope mass effect on chlorine diffusion in dacite melt, with implications for fractionation during bubble growth. Earth and Planetary Science Letters 480, 15–24.
) (Fig. 3b). Therefore, in terms of predicted Cl-δ37Cl values, magmatic brine is an assimilant that near-perfectly matches the observed δ37Cl shifts in Icelandic rhyolites (Fig. S-4). Anomalously negative δ37Cl values of down to −5.6 ‰ have been reported for saline fluid inclusions in porphyry copper and iron oxide-copper-gold deposits, showing that low δ37Cl brines do exist in magmatic-hydrothermal environments (Gleeson and Smith, 2009Gleeson, S.A., Smith, M.P. (2009) The sources and evolution of mineralising fluids in iron oxide–copper–gold systems, Norrbotten, Sweden: Constraints from Br/Cl ratios and stable Cl isotopes of fluid inclusion leachates. Geochimica et Cosmochimica Acta 73, 5658–5672.
; Nahnybida et al., 2009Nahnybida, T., Gleeson, S.A., Rusk, B.G., Wassenaar, L.I. (2009) Cl/Br ratios and stable chlorine isotope analysis of magmatic–hydrothermal fluid inclusions from Butte, Montana and Bingham Canyon, Utah. Mineralium Deposita 44, 837.
).top
Assimilation of Magmatic Brines
The presence of magmatic brine is a widely reported phenomenon associated with silicic and intermediate magmas in the upper crust. For example, brines are found in magmatic fluid inclusions and are implicated in the formation of magmatic-hydrothermal ore deposits globally (Audétat et al., 2008
Audétat, A., Pettke, T., Heinrich, C.A., Bodnar, R.J. (2008) The composition of magmatic-hydrothermal fluids in barren and mineralized intrusions. Economic Geology 103, 877–908.
), including in Iceland (Kremer and Bird, 2018Kremer, C.H., Bird, D.K. (2018) Fluid origin and evolution of Cu-Pb-Zn mineralization in rhyolite breccias in the Lón area, southeastern Iceland. Journal of Volcanology and Geothermal Research 349, 177–191.
). Crystallisation of silicic melts that stall in the crust lead to late stage exsolution of magmatic brines or hydrosaline fluids. Magmatic brines may form by direct exsolution from melts with moderate Cl/H2O ratios (>0.05 for granitic melts; Webster, 2004Webster, J.D. (2004) The exsolution of magmatic hydrosaline chloride liquids. Chemical Geology 210, 33–48.
) at pressures below about 1.5 kbar, by phase separation of a magmatic fluid into low NaCl vapour and a high NaCl brine (up to NaCl >85 wt. %) during decompression (Fig. 3c), or by condensation of magmatic vapour (Webster and Mandeville, 2007Webster, J.D., Mandeville, C.W. (2007) Fluid immiscibility in volcanic environments. Reviews in Mineralogy and Geochemistry 65, 313–362.
). Magmatic brines are less dense and stable to lower temperatures compared to melts, and once formed, may accumulate in pore space or pool in roof zones of magma mushes forming lenses (Fig. 3a) that can stay stable for over 1 Myr (Blundy et al., 2015Blundy, J., Mavrogenes, J., Tattitch, B., Sparks, S., Gilmer, A. (2015) Generation of porphyry copper deposits by gas–brine reaction in volcanic arcs. Nature Geoscience 8, 235.
; Afanasyev et al., 2018Afanasyev, A., Blundy, J., Melnik, O., Sparks, S. (2018) Formation of magmatic brine lenses via focussed fluid-flow beneath volcanoes. Earth and Planetary Science Letters 486, 119–128.
; Edmonds and Woods, 2018Edmonds, M., Woods, A.W. (2018) Exsolved volatiles in magma reservoirs. Journal of Volcanology and Geothermal Research 368, 13–30.
). Individual accumulations of silicic melts form over short time scales (0.01 to 1 kyr time scales) compared to the long lifetimes of the silicic magma mushes that they are part of (100 kyr – 1 Myr time scales) (Padilla et al., 2016Padilla, A.J., Miller, C.F., Carley, T.L., Economos, R.C., Schmitt, A.K., Coble, M.A., Wooden, J.L., Fisher, C.-M., Vervoort, J.D., Hanchar, J.M. (2016) Elucidating the magmatic history of the Austurhorn silicic intrusive complex (southeast Iceland) using zircon elemental and isotopic geochemistry and geochronology. Contributions to Mineralogy and Petrology 171, 69.
; Cooper, 2019Cooper, K.M. (2019) Time scales and temperatures of crystal storage in magma reservoirs: Implications for magma reservoir dynamics. Philosophical Transactions of the Royal Society A 377, 20180009.
). Thus, cycles of silicic melt production and crystallisation lead to repeated production and accumulation of magmatic brines in long lived magma mushes.Our samples are chlorine undersaturated, similar to the majority of felsic melt inclusions globally (Webster et al., 2019
Webster, J.D., Iveson, A.A., Rowe, M.C., Webster, P.M. (2019) Chlorine and felsic magma evolution: Modeling the behavior of an under-appreciated volatile component. Geochimica et Cosmochimica Acta 271, 248–288.
). We envision that such chlorine undersaturated melts may, prior to or during eruptions, assimilate ambient low δ37Cl magmatic brines that have been formed by previous generations of silicic intrusions within the same, long lived silicic magma mush (Fig. 3a). Our bulk assimilation model shows that modest amounts (ca. 0.5 wt. %) of addition of magmatic brines with NaClequivalent = 16.5 wt. % and δ37Cfluid = −4 ‰ is sufficient to explain the maximum observed δ37Cl shift of −2.9 ‰ between silicic rocks and basalts in our samples (Figs. 1b, S-4, S-5). Assimilation of brines has been previously demonstrated to take place in submarine basalts, that may directly assimilate seawater-derived brines (Kendrick et al., 2013Kendrick, M.A., Arculus, R., Burnard, P., Honda, M. (2013) Quantifying brine assimilation by submarine magmas: Examples from the Galápagos Spreading Centre and Lau Basin. Geochimica et Cosmochimica Acta 123, 150–165.
), and in silicic melts, where surplus Cl contents have been interpreted as assimilation of hydrosaline fluids of unknown origin (Webster et al., 2019Webster, J.D., Iveson, A.A., Rowe, M.C., Webster, P.M. (2019) Chlorine and felsic magma evolution: Modeling the behavior of an under-appreciated volatile component. Geochimica et Cosmochimica Acta 271, 248–288.
).The common association of silicic magmas with brines and the dominantly negative δ37Cl signatures observed in silicic volcanic rocks that are difficult to reconcile with other known magmatic processes suggest that magmatic brine assimilation may be a fundamental process in silicic, long lived magma mushes. Our results highlight that little is still known about the storage and evolution of hydrosaline liquids in magma mushes. The details of physical and chemical interactions between brines and melts should be a fruitful target of future research aiming to improve our understanding of silicic magmatism. Finally, we note that low δ37Cl magmatic vapours and/or liquids residing in the roof zones of magma mushes may become incorporated in eruption clouds or shallow hydrothermal systems (Fig. 3a). This process could, instead of direct degassing of magmatic Cl, offer an alternative explanation to the association of volcanic activity with negative δ37Cl signatures in thermal springs and fumaroles in Guadeloupe, Martinique (Li et al., 2015
Li, L., Bonifacie, M., Aubaud, C., Crispi, O., Dessert, C., Agrinier, P. (2015) Chlorine isotopes of thermal springs in arc volcanoes for tracing shallow magmatic activity. Earth and Planetary Science Letters 413, 101–110.
) and the Izu-Bonin-Mariana arc (Barnes et al., 2008Barnes, J.D., Sharp, Z.D., Fischer, T.P. (2008) Chlorine isotope variations across the Izu-Bonin-Mariana arc. Geology 36, 883–886.
) as well as volcanic gases in Stromboli, Italy (Liotta et al., 2017Liotta, M., Rizzo, A.L., Barnes, J.D., D’Auria, L., Martelli, M., Bobrowski, N., Wittmer, J. (2017) Chlorine isotope composition of volcanic rocks and gases at Stromboli volcano (Aeolian Islands, Italy): Inferences on magmatic degassing prior to 2014 eruption. Journal of Volcanology and Geothermal Research 336, 168–178.
).top
Acknowledgements
ER acknowledges support from NordVulk and the University of Iceland Research Fund. SAH acknowledges support from the Icelandic Research Fund (Grant #196139-051). We would like to thank Guðmundur H. Guðfinnsson for assistance with EPMA analysis and Enikő Bali for help with FTIR. We thank Maja Bar Rasmussen, Edward W. Marshall and Olgeir Sigmarsson for fruitful discussions. Níels Óskarsson is thanked for sharing previously unpublished data for Askja and Hekla samples. Cin-Ty Lee is acknowledged for smooth editorial handling of the manuscript. We are grateful to Shanaka de Silva and an anonymous reviewer for constructive comments that helped improve the manuscript. Isabelle Chambefort and two anonymous reviewers are thanked for helpful comments on a previous version of the manuscript.
Editor: Cin-Ty Lee
top
References
Afanasyev, A., Blundy, J., Melnik, O., Sparks, S. (2018) Formation of magmatic brine lenses via focussed fluid-flow beneath volcanoes. Earth and Planetary Science Letters 486, 119–128.
Show in context Magmatic brines are less dense and stable to lower temperatures compared to melts, and once formed, may accumulate in pore space or pool in roof zones of magma mushes forming lenses (Fig. 3a) that can stay stable for over 1 Myr (Blundy et al., 2015; Afanasyev et al., 2018; Edmonds and Woods, 2018).
View in article
Audétat, A., Pettke, T., Heinrich, C.A., Bodnar, R.J. (2008) The composition of magmatic-hydrothermal fluids in barren and mineralized intrusions. Economic Geology 103, 877–908.
Show in context For example, brines are found in magmatic fluid inclusions and are implicated in the formation of magmatic-hydrothermal ore deposits globally (Audétat et al., 2008), including in Iceland (Kremer and Bird, 2018).
View in article
Barnes, J.D., Sharp, Z.D. (2017) Chlorine isotope geochemistry. Reviews in Mineralogy and Geochemistry 82, 345–378.
Show in context Lavas associated with subduction zones and oceanic islands have a range of δ37Cl values from −3 to +3 ‰, likely due to incorporation of subduction fluids, recycled marine sediments and altered oceanic crust into the mantle (John et al., 2010; Barnes and Sharp, 2017).
View in article
Barnes, J.D., Sharp, Z.D., Fischer, T.P. (2008) Chlorine isotope variations across the Izu-Bonin-Mariana arc. Geology 36, 883–886.
Show in context As chlorine is a hydrophile element, its isotopic fingerprint has been used to trace volatile sources in igneous rocks and hydrothermal fluids (Barnes et al., 2008; Li et al., 2015).
View in article
This process could, instead of direct degassing of magmatic Cl, offer an alternative explanation to the association of volcanic activity with negative δ37Cl signatures in thermal springs and fumaroles in Guadeloupe, Martinique (Li et al., 2015) and the Izu-Bonin-Mariana arc (Barnes et al., 2008) as well as volcanic gases in Stromboli, Italy (Liotta et al., 2017).
View in article
Barnes, J.D., Prather, T.J., Cisneros, M., Befus, K., Gardner, J.E., Larson, T.E. (2014) Stable chlorine isotope behavior during volcanic degassing of H2O and CO2 at Mono Craters, CA. Bulletin of Volcanology 76, 805.
Show in context The majority of chlorine isotope studies on igneous rocks have been conducted on basaltic rocks, whereas published δ37Cl data for silicic rocks is limited, with 40 out of 44 published analyses coming from a single volcanic system, the Mono Craters, USA (Barnes et al., 2014).
View in article
Silicic rocks in Iceland have lower δ37Cl values than basalts (data from Halldórsson et al., 2016), overlapping with rhyolites from the Mono Craters, USA (Barnes et al., 2014).
View in article
Local δ37Cl variability in Icelandic rhyolites appears to be small (≤0.5 ‰ for Hekla: H3, H4, H5; and Askja: ASD1L, ASD14L) compared to the large range of −1.9 to 0.0 ‰ reported for the Mono Crater rhyolites (Barnes et al., 2014) (Fig. 1).
View in article
Mono Craters field is drawn after data reported in Newman et al. (1988) and Barnes et al. (2014).
View in article
Blundy, J., Mavrogenes, J., Tattitch, B., Sparks, S., Gilmer, A. (2015) Generation of porphyry copper deposits by gas–brine reaction in volcanic arcs. Nature Geoscience 8, 235.
Show in context Magmatic brines are less dense and stable to lower temperatures compared to melts, and once formed, may accumulate in pore space or pool in roof zones of magma mushes forming lenses (Fig. 3a) that can stay stable for over 1 Myr (Blundy et al., 2015; Afanasyev et al., 2018; Edmonds and Woods, 2018).
View in article
Cooper, K.M. (2019) Time scales and temperatures of crystal storage in magma reservoirs: Implications for magma reservoir dynamics. Philosophical Transactions of the Royal Society A 377, 20180009.
Show in context Individual accumulations of silicic melts form over short time scales (0.01 to 1 kyr time scales) compared to the long lifetimes of the silicic magma mushes that they are part of (100 kyr – 1 Myr time scales) (Padilla et al., 2016; Cooper, 2019).
View in article
Cullen, J.T., Hurwirz, S., Barnes, J.D., Lassiter, J.C., Penniston-Dorland, S., Kasemann, S.A., Thordsen, J.J. (2019) Temperature-Dependent variations in mineralogy, major element chemistry and the stable isotopes of boron, lithium and chlorine resulting from hydration of rhyolite: Constraints from hydrothermal experiments at 150 to 350° C and 25 MPa. Geochimica et Cosmochimica Acta 261, 269–287.
Show in context Conversely, this suggests that hydrothermally altered crust in Iceland has a basalt-like δ37Cl range, consistent with the basalt-like δ37Cl values in Icelandic hydrothermal fluids (Stefánsson and Barnes, 2016) and the lack of δ37Cl fractionation resulting from hydrothermal alteration (Cullen et al., 2019).
View in article
Fortin, M.A., Watson, E.B., Stern, R. (2017) The isotope mass effect on chlorine diffusion in dacite melt, with implications for fractionation during bubble growth. Earth and Planetary Science Letters 480, 15–24.
Show in context Arrows in (a) indicate the effects of the small equilibrium isotope fractionations caused by fractional crystallisation and degassing, and the large kinetic isotope fractionation during magmatic brine exsolution (Fortin et al., 2017) and assimilation on the δ37Cl and Cl composition of silicic melts.
View in article
Magmatic fluids exsolve from silicic melts during late stage crystallisation and acquire negative δ37Cl values through kinetic fractionation (Fortin et al., 2017).
View in article
Magmatic hydrosaline fluids have by definition high Cl concentrations and may acquire highly negative δ37Cl values during exsolution from dacitic (and more silicic) melts due to kinetic diffusion effects, that cause considerable fractionation of up to Δ37Clfluid-melt = −5 ‰ even at high temperatures (Fortin et al., 2017) (Fig. 3b).
View in article
Edmonds, M., Woods, A.W. (2018) Exsolved volatiles in magma reservoirs. Journal of Volcanology and Geothermal Research 368, 13–30.
Show in context Magmatic brines are less dense and stable to lower temperatures compared to melts, and once formed, may accumulate in pore space or pool in roof zones of magma mushes forming lenses (Fig. 3a) that can stay stable for over 1 Myr (Blundy et al., 2015; Afanasyev et al., 2018; Edmonds and Woods, 2018).
View in article
Gautason, B., Muehlenbachs, K. (1998) Oxygen isotopic fluxes associated with high-temperature processes in the rift zones of Iceland. Chemical Geology 145, 275–286.
Show in context In Iceland, low δ18Orock values relative to pristine basaltic values (+4.8 to +5.8 ‰; Thirlwall et al., 2006) are used to recognise assimilation (or partial melting) of altered crust, which has been shifted to low δ18O signatures (≤+2 ‰) by hydrothermal alteration with low δ18O meteoric water (Gautason and Muehlenbachs, 1998).
View in article
Gleeson, S.A., Smith, M.P. (2009) The sources and evolution of mineralising fluids in iron oxide–copper–gold systems, Norrbotten, Sweden: Constraints from Br/Cl ratios and stable Cl isotopes of fluid inclusion leachates. Geochimica et Cosmochimica Acta 73, 5658–5672.
Show in context Anomalously negative δ37Cl values of down to −5.6 ‰ have been reported for saline fluid inclusions in porphyry copper and iron oxide-copper-gold deposits, showing that low δ37Cl brines do exist in magmatic-hydrothermal environments (Gleeson and Smith, 2009; Nahnybida et al., 2009).
View in article
Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
Show in context Together with previously published δ37Cl and δ18O data on Icelandic basalts (Halldórsson et al., 2016), the samples represent the full chemical range between subalkaline-tholeiitic rift zone, and transitional to alkaline propagating rift and off-rift magma suites in Iceland (Fig. S-1), spanning a SiO2 range of 44.4-77.7 wt. % and Cl concentrations between 17 and 3988 ppm (Figs. 1, S-2).
View in article
All studied volcanoes are situated on land and are free of seawater influence (Halldórsson et al., 2016).
View in article
The basaltic (n = 3) and intermediate (n = 4) samples have δ37Cl values between −0.4 and +1.3 ‰, overlapping with the known range of Icelandic basalts of −0.6 to +1.4 ‰ (Halldórsson et al., 2016).
View in article
In contrast, the δ37Cl values of the silicic samples from this study (n = 8) and those previously published (n = 3; Halldórsson et al., 2016) deviate from the basaltic-intermediate range towards more negative values of −1.9 to −0.6 ‰ (Fig. 1b), except for a single outlier (SAL-74) with δ37Cl = +0.9 ‰
View in article
Whereas basalts inherit the δ37Cl signatures of their mantle sources (Halldórsson et al., 2016), the shift to more negative δ37Cl values in silicic rocks must reflect a process or a combination of processes taking place during rhyolite genesis, such as mineral-melt fractionation, degassing and/or assimilation.
View in article
Silicic rocks in Iceland have lower δ37Cl values than basalts (data from Halldórsson et al., 2016), overlapping with rhyolites from the Mono Craters, USA (Barnes et al., 2014).
View in article
John, T., Layne, G.D., Haase, K.M., Barnes, J.D. (2010) Chlorine isotope evidence for crustal recycling into the Earth’s mantle. Earth and Planetary Science Letters 298, 175–182.
Show in context Lavas associated with subduction zones and oceanic islands have a range of δ37Cl values from −3 to +3 ‰, likely due to incorporation of subduction fluids, recycled marine sediments and altered oceanic crust into the mantle (John et al., 2010; Barnes and Sharp, 2017).
View in article
Jónasson, K. (2007) Silicic volcanism in Iceland: Composition and distribution within the active volcanic zones. Journal of Geodynamics 43, 101–117.
Show in context The samples cover the main types of silicic rocks in Iceland, i.e. dacites and alkaline and subalkaline rhyolites (Jónasson, 2007), and include both obsidians and tephras (i.e. products of effusive vs. explosive eruptions).
View in article
Kendrick, M.A., Arculus, R., Burnard, P., Honda, M. (2013) Quantifying brine assimilation by submarine magmas: Examples from the Galápagos Spreading Centre and Lau Basin. Geochimica et Cosmochimica Acta 123, 150–165.
Show in context Assimilation of brines has been previously demonstrated to take place in submarine basalts, that may directly assimilate seawater-derived brines (Kendrick et al., 2013), and in silicic melts, where surplus Cl contents have been interpreted as assimilation of hydrosaline fluids of unknown origin (Webster et al., 2019).
View in article
Kremer, C.H., Bird, D.K. (2018) Fluid origin and evolution of Cu-Pb-Zn mineralization in rhyolite breccias in the Lón area, southeastern Iceland. Journal of Volcanology and Geothermal Research 349, 177–191.
Show in context For example, brines are found in magmatic fluid inclusions and are implicated in the formation of magmatic-hydrothermal ore deposits globally (Audétat et al., 2008), including in Iceland (Kremer and Bird, 2018).
View in article
Li, L., Bonifacie, M., Aubaud, C., Crispi, O., Dessert, C., Agrinier, P. (2015) Chlorine isotopes of thermal springs in arc volcanoes for tracing shallow magmatic activity. Earth and Planetary Science Letters 413, 101–110.
Show in context As chlorine is a hydrophile element, its isotopic fingerprint has been used to trace volatile sources in igneous rocks and hydrothermal fluids (Barnes et al., 2008; Li et al., 2015).
View in article
This process could, instead of direct degassing of magmatic Cl, offer an alternative explanation to the association of volcanic activity with negative δ37Cl signatures in thermal springs and fumaroles in Guadeloupe, Martinique (Li et al., 2015) and the Izu-Bonin-Mariana arc (Barnes et al., 2008) as well as volcanic gases in Stromboli, Italy (Liotta et al., 2017).
View in article
Liebscher, A., Barnes, J., Sharp, Z. (2006) Chlorine isotope vapor–liquid fractionation during experimental fluid-phase separation at 400 C/23 MPa to 450 C/42 MPa. Chemical Geology 234, 340–345.
Show in context Decompression-driven phase separation of a supercritical fluid produces a NaCl-rich brine and a NaCl-poor vapour with a maximum fractionation of Δ37Clliquid-vapour = ±0.5 ‰ (Liebscher et al., 2006).
View in article
Liotta, M., Rizzo, A.L., Barnes, J.D., D’Auria, L., Martelli, M., Bobrowski, N., Wittmer, J. (2017) Chlorine isotope composition of volcanic rocks and gases at Stromboli volcano (Aeolian Islands, Italy): Inferences on magmatic degassing prior to 2014 eruption. Journal of Volcanology and Geothermal Research 336, 168–178.
Show in context This process could, instead of direct degassing of magmatic Cl, offer an alternative explanation to the association of volcanic activity with negative δ37Cl signatures in thermal springs and fumaroles in Guadeloupe, Martinique (Li et al., 2015) and the Izu-Bonin-Mariana arc (Barnes et al., 2008) as well as volcanic gases in Stromboli, Italy (Liotta et al., 2017).
View in article
Nahnybida, T., Gleeson, S.A., Rusk, B.G., Wassenaar, L.I. (2009) Cl/Br ratios and stable chlorine isotope analysis of magmatic–hydrothermal fluid inclusions from Butte, Montana and Bingham Canyon, Utah. Mineralium Deposita 44, 837.
Show in context Anomalously negative δ37Cl values of down to −5.6 ‰ have been reported for saline fluid inclusions in porphyry copper and iron oxide-copper-gold deposits, showing that low δ37Cl brines do exist in magmatic-hydrothermal environments (Gleeson and Smith, 2009; Nahnybida et al., 2009).
View in article
Newman, S., Epstein, S., Stolper, E. (1988) Water, carbon dioxide, and hydrogen isotopes in glasses from the ca. 1340 AD eruption of the Mono Craters, California: constraints on degassing phenomena and initial volatile content. Journal of Volcanology and Geothermal Research 35, 75–96.
Show in context Mono Craters field is drawn after data reported in Newman et al. (1988) and Barnes et al. (2014).
View in article
Padilla, A.J., Miller, C.F., Carley, T.L., Economos, R.C., Schmitt, A.K., Coble, M.A., Wooden, J.L., Fisher, C.-M., Vervoort, J.D., Hanchar, J.M. (2016) Elucidating the magmatic history of the Austurhorn silicic intrusive complex (southeast Iceland) using zircon elemental and isotopic geochemistry and geochronology. Contributions to Mineralogy and Petrology 171, 69.
Show in context Individual accumulations of silicic melts form over short time scales (0.01 to 1 kyr time scales) compared to the long lifetimes of the silicic magma mushes that they are part of (100 kyr – 1 Myr time scales) (Padilla et al., 2016; Cooper, 2019).
View in article
Rose-Koga, E.F., Sigmarsson, O. (2008) B–O–Th isotope systematics in Icelandic tephra. Chemical Geology 255, 454–462.
Show in context In contrast, boron another fluid-mobile element, displays anomalous positive δ11B values in Icelandic silicic rocks that correlate with decreasing δ18O, and that have thus been explained by crustal assimilation (Rose-Koga and Sigmarsson, 2008).
View in article
Schauble, E.A., Rossman, G.R., Taylor Jr., H.P. (2003) Theoretical estimates of equilibrium chlorine-isotope fractionations. Geochimica et Cosmochimica Acta 67, 3267–3281.
Show in context In contrast, the depleted upper mantle (DMM) has a restricted δ37Cl variability of −0.2 ± 0.3 ‰ (Sharp et al., 2013), reflecting the limited δ37Cl fractionation from high temperature magmatic processes (Schauble et al., 2003).
View in article
Rayleigh δ37Cl fractionations between HCl(g), minerals and silicic melt are expected to be small, based on theoretical equilibrium fractionation factors of Schauble et al. (2003) extrapolated to magmatic temperatures (Δ37Clmineral-melt ≈ Δ37Clvapour-melt ≈ 0.2 ‰ at 600 °C).
View in article
Sharp, Z.D., Mercer, J.A., Jones, R.H., Brearley, A.J., Selverstone, J., Bekker, A., Stachel, T. (2013) The chlorine isotope composition of chondrites and Earth. Geochimica et Cosmochimica Acta 107, 189–204.
Show in context In contrast, the depleted upper mantle (DMM) has a restricted δ37Cl variability of −0.2 ± 0.3 ‰ (Sharp et al., 2013), reflecting the limited δ37Cl fractionation from high temperature magmatic processes (Schauble et al., 2003).
View in article
Stefánsson, A., Barnes, J.D. (2016) Chlorine isotope geochemistry of Icelandic thermal fluids: Implications for geothermal system behavior at divergent plate boundaries. Earth and Planetary Science Letters 449, 69–78.
Show in context Conversely, this suggests that hydrothermally altered crust in Iceland has a basalt-like δ37Cl range, consistent with the basalt-like δ37Cl values in Icelandic hydrothermal fluids (Stefánsson and Barnes, 2016) and the lack of δ37Cl fractionation resulting from hydrothermal alteration (Cullen et al., 2019).
View in article
Thirlwall, M.F., Gee, M.A.M., Lowry, D., Mattey, D.P., Murton, B.J., Taylor, R.N. (2006) Low δ18O in the Icelandic mantle and its origins: Evidence from Reykjanes Ridge and Icelandic lavas. Geochimica et Cosmochimica Acta 70, 993–1019.
Show in context In Iceland, low δ18Orock values relative to pristine basaltic values (+4.8 to +5.8 ‰; Thirlwall et al., 2006) are used to recognise assimilation (or partial melting) of altered crust, which has been shifted to low δ18O signatures (≤+2 ‰) by hydrothermal alteration with low δ18O meteoric water (Gautason and Muehlenbachs, 1998).
View in article
We note that basalts and intermediate rocks from all three volcanic settings display δ18O values between +3.4 and +5.2 ‰ (Fig. 2), typical for Icelandic basalts (Thirlwall et al., 2006).
View in article
Webster, J.D. (2004) The exsolution of magmatic hydrosaline chloride liquids. Chemical Geology 210, 33–48.
Show in context Chlorine is among the most abundant volatile elements in igneous rocks and may become concentrated enough in late stage silicic melts to exsolve and form hydrosaline liquids, i.e. high density Cl-enriched aqueous fluids or hydrosaline brines (Webster, 2004).
View in article
Magmatic brines may form by direct exsolution from melts with moderate Cl/H2O ratios (>0.05 for granitic melts; Webster, 2004) at pressures below about 1.5 kbar, by phase separation of a magmatic fluid into low NaCl vapour and a high NaCl brine (up to NaCl >85 wt. %) during decompression (Fig. 3c), or by condensation of magmatic vapour (Webster and Mandeville, 2007).
View in article
Webster, J.D., Mandeville, C.W. (2007) Fluid immiscibility in volcanic environments. Reviews in Mineralogy and Geochemistry 65, 313–362.
Show in context Magmatic brines may form by direct exsolution from melts with moderate Cl/H2O ratios (>0.05 for granitic melts; Webster, 2004) at pressures below about 1.5 kbar, by phase separation of a magmatic fluid into low NaCl vapour and a high NaCl brine (up to NaCl >85 wt. %) during decompression (Fig. 3c), or by condensation of magmatic vapour (Webster and Mandeville, 2007).
View in article
Webster, J.D., Iveson, A.A., Rowe, M.C., Webster, P.M. (2019) Chlorine and felsic magma evolution: Modeling the behavior of an under-appreciated volatile component. Geochimica et Cosmochimica Acta 271, 248–288.
Show in context Indeed, a complicated pre-eruptive volatile history is also reflected by high Cl variability in Icelandic propagating rift and rift rhyolites (50 to 2600 ppm) (Fig. S-2), likely reflecting a combination of fractional crystallisation, partial melting, accumulation of fractional melts from volatile heterogeneous sources as well as episodic exsolution and resorption of magmatic volatile phases, including magmatic brines (Webster et al., 2019; Supplementary Information S-3; Fig. 3a).
View in article
Our samples are chlorine undersaturated, similar to the majority of felsic melt inclusions globally (Webster et al., 2019).
View in article
Assimilation of brines has been previously demonstrated to take place in submarine basalts, that may directly assimilate seawater-derived brines (Kendrick et al., 2013), and in silicic melts, where surplus Cl contents have been interpreted as assimilation of hydrosaline fluids of unknown origin (Webster et al., 2019).
View in article
top
Supplementary Information
The Supplementary Information includes:
- S-1: Samples and Methods
- S-2: Geological Setting
- S-3: Chlorine Contents in Icelandic Melt Inclusions
- S-4: Bulk Assimilation Model
- S-5: Quantifying Magmatic Brine Formation in the Crust
- Tables S-1 to S-3
- Figures S-1 to S-5
- Supplementary Information References
Download Table S-3 (Excel).
Download the Supplementary Information (PDF).
Figures
Figure 1 Chlorine isotope variations vs. (a) Cl concentrations and (b) SiO2. Silicic rocks in Iceland have lower δ37Cl values than basalts (data from Halldórsson et al., 2016
Halldórsson, S.A., Barnes, J.D., Stefánsson, A., Hilton, D.R., Hauri, E.H., Marshall, E.W. (2016) Subducted lithosphere controls halogen enrichments in the Iceland mantle plume source. Geology 44, 679–682.
), overlapping with rhyolites from the Mono Craters, USA (Barnes et al., 2014Barnes, J.D., Prather, T.J., Cisneros, M., Befus, K., Gardner, J.E., Larson, T.E. (2014) Stable chlorine isotope behavior during volcanic degassing of H2O and CO2 at Mono Craters, CA. Bulletin of Volcanology 76, 805.
). Arrows in (a) indicate the effects of the small equilibrium isotope fractionations caused by fractional crystallisation and degassing, and the large kinetic isotope fractionation during magmatic brine exsolution (Fortin et al., 2017Fortin, M.A., Watson, E.B., Stern, R. (2017) The isotope mass effect on chlorine diffusion in dacite melt, with implications for fractionation during bubble growth. Earth and Planetary Science Letters 480, 15–24.
) and assimilation on the δ37Cl and Cl composition of silicic melts. The gray curve in (a) shows the effect of magmatic brine assimilation (in wt. %) on the Cl-δ37Cl values of a hypothetical rhyolite melt with an average propagating rift basalt δ37Cl value of +0.7 ‰. The negative δ37Cl shifts between silicic rocks and basalts are illustrated in (b) by arrows anchored at the average SiO2 concentrations and δ37Cl values of the rift, propagating rift and off-rift basalts. The 1σ uncertainty is ±0.2 ‰ for δ37Cl.
Figure 2 The δ18O-δ37Cl systematics of silicic rocks. Low δ18O values of rift-related silicic rocks result from assimilation of hydrothermally altered, δ18O-depleted crust. Lack of correlation between δ18O and δ37Cl indicates that negative δ37Cl values of rhyolites are not caused by crustal assimilation. The gray line shows the effect of magmatic brine assimilation in wt. % (Supplementary Information S-4). Mono Craters field is drawn after data reported in Newman et al. (1988)
Newman, S., Epstein, S., Stolper, E. (1988) Water, carbon dioxide, and hydrogen isotopes in glasses from the ca. 1340 AD eruption of the Mono Craters, California: constraints on degassing phenomena and initial volatile content. Journal of Volcanology and Geothermal Research 35, 75–96.
and Barnes et al. (2014)Barnes, J.D., Prather, T.J., Cisneros, M., Befus, K., Gardner, J.E., Larson, T.E. (2014) Stable chlorine isotope behavior during volcanic degassing of H2O and CO2 at Mono Craters, CA. Bulletin of Volcanology 76, 805.
. The 1σ uncertainty is ±0.2 ‰ for δ37Cl, and smaller than the size of the symbols for δ18O.
Figure 3 (a) A model of magmatic brine formation and assimilation in a long lived upper crustal magma mush. (b) Magmatic fluids exsolve from silicic melts during late stage crystallisation and acquire negative δ37Cl values through kinetic fractionation (Fortin et al., 2017
Fortin, M.A., Watson, E.B., Stern, R. (2017) The isotope mass effect on chlorine diffusion in dacite melt, with implications for fractionation during bubble growth. Earth and Planetary Science Letters 480, 15–24.
). (c) Decompression-driven phase separation of a supercritical fluid produces a NaCl-rich brine and a NaCl-poor vapour with a maximum fractionation of Δ37Clliquid-vapour = ±0.5 ‰ (Liebscher et al., 2006Liebscher, A., Barnes, J., Sharp, Z. (2006) Chlorine isotope vapor–liquid fractionation during experimental fluid-phase separation at 400 C/23 MPa to 450 C/42 MPa. Chemical Geology 234, 340–345.
).