Attribution: You must attribute the work in the manner specified by the author or licensor ( but no in any way that suggests that they endorse you or your use of the work).Noncommercial ‒ you may not use this work for commercial purpose.No Derivative works ‒ You may not alter, transform, or build upon this work.Sharing ‒ Individual scientists are hereby granted permission, without fees or further requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in other subsequent works and to make unlimited photo copies of items in this journal for noncommercial use in classrooms to further education and science.

Klemme (2004) describes Ca- and F-rich inclusions in peridotite xenoliths (Kakanui, New Zealand) and advocates the concept of the fluoride-silicate liquid-liquid immiscibility at upper-mantle conditions for their origin. He quotes experimental results in the CaF2-SiO2 system to support the existence of the fluoride-silicate liquid-liquid immiscibility. The miscibility gap in the CaF2-SiO2 system is located above 1390 °C at 1 atm pressure (Hillert, 1970), exceeding current magmatic processes. The addition of network-modifiers (CaO, MgO, FeO, Al2O3), which bring the system composition closer to natural melts, causes a rapid closure of the miscibility gap (Ueda and Maeda, 1999); no fluoride-silicate immiscibility exists in the binary systems: wollastonite-CaF2 (Mukerji, 1965), enstatite-CaF2, forsterite-CaF2 (Berezhnoi, 1951; Ueda et al., 2001), and anorthite-CaF2 (Færøyvik et al., 1999).

The coexisting silicate glass analyzed by Klemme (2004) has a peraluminous hypersthene-normative composition and a very small CaO content (0.9 wt%), both are surprising for a mafic system saturated with clinopyroxene and a Ca- and F-rich liquid. In addition, the fluorine content of the precursor liquid is incorrect, a mixture of 8.9 and 40.2 wt% F cannot produce 4.0 wt% F.

The fluoride liquids, with 39.9 wt% Ca and 40.2 wt% F (Klemme, 2004), are molten ionic salts which neither form nor quench to glasses (Grande et al., 1995); present “fluoride glasses” may represent interstitial isotropic solid phases.

The single analysis of the fluoride phase has a total of 94.5 wt%, which “indicate[s] significant H2O, CO2, carbonate, or B dissolved in both [fluoride and silicate] melts” (Klemme, 2004, p. 442). The low molecular weights of hydrogen, carbon, boron (and sulfur) require the weight-percent amounts of these elements, necessary to ensure the charge balance, to be small (wt%: H, 0.9; C, 2.6; B, 3.1; or S[6+], 4.6). Thus, none of these elements can account for the deficit of 5.5 wt%.

Fluorine as a low-polarizing anion produces a strong short-range order with low-field strength cations (Zeng and Stebbins, 2000). That is, the fluoride liquid is expected to incorporate Na, K, and Ca, whereas the coexisting silicate liquid should be enriched in Mg, Fe, Mn, and Al. These trends are not indicated by the presented data.

Klemme (2004) illustrates metasomatic effects by fluoride melts on the peridotite assemblage by a reaction (his Equation 2) which, however, is not balanced (missing oxygen on the left). Written correctly, 3MgSiO3 + CaF2 + 1/2 O2 = CaMgSi2O6 + Mg2SiO4 + F2, the oxygen fugacity of mantle peridotites (e.g., Dalton and Wood, 1995) defines the fluorine fugacity, fF2 = 10−52.0 to 10−30.5 at 400–900 °C and 1 GPa (Holl and Powell, 1998; Chase, 1998). The fluorine gas constitutes a very negligible fraction of the supercritical fluid and the melt devolatilization cannot remove fluorine from the system. Similarly, the equation 3MgSiO3 + CaF2 = CaMgSi2O6 + Mg2SiO4 + F2O–1 refutes the author's hypothesis that “secondary clinopyroxene may in fact be generated by fluoride melts” (Klemme, 2004, p. 442) because the fluorination of the clinopyroxene-olivine assemblage must produce fluoride liquids or solids coexisting with orthopyroxene.

The Ca- and F-rich nature of the fluoride “liquid” suggests a misinterpretation for interstitial fluorite; the only analysis has a Ca:F molar ratio of 1:2.13. I have calculated melting temperatures of fluorite and fluoride-rich liquids to high pressures (Tsytsenko et al., 1993; Chase, 1998; Jingu et al., 2002): (1) fluorite exhibits a high melting temperature, 1418–1840 °C (1 atm to 4 GPa) and is stable over a wide range of subarc conditions, (2) Klemme's fluoride liquid with 77.4 mol% CaF2 (on the CaF2-MO2 basis) has activity(CaF2) ~0.8; it can only exist above 1256–1561 °C (1 atm to 4 GPa), and (3) along the subducting-slab geotherm, the maximum fluorine concentrations in the melt are 0.03 mol% CaF2 (400 °C, 1 GPa) to 0.9 mol% CaF2 (600 °C, 2 GPa). These levels, less than 0.5 wt% F, clearly show that neither the precursor melt (4 wt% F), nor the conjugate silicate (8.9 wt% F) and the fluoride melts (40.2 wt% F) can exist in a subduction zone.

Thus, several chemical and thermodynamic principles argue against Klemme's interpretations: Ca- and F-dominated liquids do not form glasses; the fluoride-silicate immiscibility does not exist in multicomponent peridotite or basaltic systems; the formation of fluorine-rich melts in subduction zones is unlikely due to the fluorite crystallization and buffering, and hence the analyzed “liquids” are probably misinterpreted fluorite grains, which could have precipitated directly from the silicate melt.