Experimental melting of phlogopite-peridotite in the garnet stability field

Abstract

International audienceMelting experiments have been performed at 3 GPa, between 1150 and 1450 °C, on a phlogopite-peridotite source in the garnet stability field. We succeeded to extract and determine the melt compositions of both phlogopite-bearing lherzolite and harzburgite from low to high degrees of melting (ϕ = 0.008–0.256). Accounting for the presence of small amounts of F in the mantle, we determined that phlogopite coexists with melt >150 °C above the solidus position (1150–1200 °C). Fluorine content of phlogopite continuously increases during partial melting from 0.2 to 0.9 wt% between 1000 and 1150 °C and 0.5 to 0.6 wt% between 1150 and 1300 °C at 1 and 3 GPa, respectively. The phlogopite continuous breakdown in the lherzolite follows the reaction: 0.59 phlogopite + 0.52 clinopyroxene + 0.18 garnet = 0.06 olivine + 0.23 orthopyroxene + 1.00 melt. In the phlogopite-harzburgite, the reaction is: 0.93 phlogopite + 0.46 garnet = 0.25 olivine + 0.14 orthopyroxene + 1.00 melt. Melts from phlogopite-peridotite sources at 3 GPa are silica-undersaturated and are foiditic to trachybasaltic in composition from very low (0.8 wt%) to high (25.6 wt%) degrees of melting. As observed at 1 GPa, the potassium content of primary mantle melts is buffered by the presence of phlogopite, but the buffering values are higher, from 6.0 to 8.0 wt% depending on the source fertility. We finally show that phlogopite garnet-peridotite melts are very close to the composition of the most primitive post-collisional lavas described worldwide

High Pressure Phase Relations of a Depleted Peridotite Fluxed by CO2‐H2O‐Bearing Siliceous Melts and the Origin of Mid‐Lithospheric Discontinuity

“…In Figure , we show that phlogopite compositions from this study do not vary much when plotted in a K‐Al composition space, and are similar to compositions reported in previous experimental studies in peridotite + H 2 O ± CO 2 ‐bearing systems (Conceicao & Green, ; Condamine et al, ; Condamine & Médard, ; Falloon & Green, ; Fumagalli et al, ; Mallik et al, ; Mengel & Green, ; Thibault et al, ; Tumiati et al, ). For phlogopite compositions found in this study, K content varies between 0.7 and 0.85 while Al varies from 1.12 to 1.3.…”

“…(3) Minerals produced in our experiments are compositionally uniform from core to rim and no evidence of chemical zoning is found. This approach to establish chemical equilibrium was adopted in previous studies on melting of volatile‐bearing peridotite systems (Condamine et al, ; Condamine & Médard, ; Mallik et al, ; Parman & Grove, ). (4) The difference between the nominal temperatures at which the experiments were performed and the corresponding temperatures calculated using the two‐pyroxene thermometer of Brey and Kohler () for most of the experiments in this study vary from 13 to 119°C, indicating approach to equilibrium as demonstrated in a study by Mallik et al ().…”

Cryptic metasomatic agent measured in situ in Variscan mantle rocks: Melt inclusions in garnet of eclogite, Granulitgebirge, Germany

“…The source of RUB and KLA granitic melt is crustal and must be located in the subduction zone. During subduction, the slab and the mantle wedge are in contact and fluid or melt released from the slab can be transferred to the mantle wedge causing metasomatism (Condamine, Médard, & Devidal, ; Fumagalli, Zanchetta, & Poli, ; Scambelluri, Hermann, Morten, & Rampone, ). A similar genetic process may be envisioned for the ultramafic bodies here investigated, where a melt produced via melting of subducted crust migrates to the mantle wedge and causes a metasomatic interaction with an already inhomogeneous peridotite (Figure a, see also Section 6.5 for the KLA and RUB peridotite chemical features).…”

Using Major Element Logratios to Recognize Compositional Patterns of Basalt: Implications for Source Lithological and Compositional Heterogeneities

“…Moreover, FCKANTMS is stronger than FCMS in distinguishing phologopite‐peridotite melts from mafic source melts. For example, melts of phlogopite‐lherzolite BriPhl produced at 3 GPa (Condamine et al, ) have FCKANTMS of 0.03 ± 0.14 (SD, n = 6) which is significantly lower than FCMS of 0.22 ± 0.13 (SD, n = 6). Thus, the compositional difference between melts of lherzolite and mafic lithology in the FCKANTMS space is greater than that in the FCMS space (Figures b and b).…”

“…All the experimental bulk compositions are presented in Table S1. The data are from Baasner et al (), Bartels et al (), Borghini et al (), Conceicao and Green (), Condamine and Médard (), Condamine et al (), Dasgupta et al (, , ), Davis et al (, ), Draper and Johnston (), Dvir and Kessel (), Falloon and Danyushevsky (), Foley et al (), Funk and Luth (), Gaetani and Grove (), Gerbode and Dasgupta (), Ghosh et al (), Grove et al (), Hirose and Kawamoto (), Hirose and Kawamura (), Hirose and Kushiro (), Hirose and Kushiro (), Hirose (, ), Hirschmann et al (), Keshav et al (), Kessel et al (), Kinzler (), Kinzler and Grove (), Kogiso and Hirschmann (), Kogiso et al (), Kushiro (), Lambart et al (, ), Laporte et al (, ), Longhi (), Mallik et al (, ), Medard et al (), Mitchell and Grove (), Parman and Grove (), Pertermann and Hirschmann (), Pickering‐Witter and Johnston (), Pilet et al (), Salters et al (), Schwab and Johnston (), Sobolev et al (), Sorbadere et al (), Spandler et al (), Tenner et al (), Ulmer and Sweeney (), Wagner and Grove (), Walter (), Wasylenki et al (), Whitaker et al (), Yaxley and Green (), and Yaxley and Sobolev ().…”