Origin of carbonatites—liquid immiscibility caught in the act

Abstract

Carbonatites are rare but worldwide occurring igneous rocks and their genesis remains enigmatic. Field studies show a close spatial but controversially debated genetic relationship with alkaline silicate rocks, and petrological and experimental studies indicate liquid immiscibility from mantle-derived magmas being one viable model for the generation of carbonatites. However, unaltered carbonatitic melts are rare and the composition of primary carbonate liquids and their silicate conjugates is poorly constrained. Here we show an example of primary Ca-carbonatitic melt formed by liquid immiscibility from a phonolitic magma of the Laacher See volcano (Eifel, Germany). The conjugate blebs of carbonate-silicate liquids are found in hauyne-hosted melt inclusions. The Ca-carbonatite melts are moderately alkali-rich and contain high F and Cl at elevated SiO2 and Al2O3 concentrations. Such carbonatite liquids are viable parental magmas to the globally dominating intrusive Ca-carbonatite complexes and may provide the missing link to extrusive Na-carbonatitic magmas.

Reply to: Silica is unlikely to be soluble in upper crustal carbonatite melts

“…As the solidus and the liquidus of a volatile element-rich system critically depend on pressure, we are not surprised that the single experiment of A&G did not melt at all. However, we agree with A&G that experiments are needed to map out the new immiscibility gap between phonolitic melts and SiO 2 -rich Ca-carbonatites 1 . However, these are difficult experiments that need to be well documented and they need to be carried in a well-calibrated high-pressure apparatus.…”

“… 8 , 9 ). It is quite obvious that these experimental results, and our observations at the Laacher See volcano 1 , do not agree with a simplistic model (e.g., refs. 2 , 10 , 11 ) that aims to explain the origin of Ca-rich carbonatites, in essence, by loss of Na (and gain of Si amongst other elements) from a parental Na-rich carbonatite melt to the surrounding wall rocks.…”

“…Nevertheless, as hauyne is the host mineral of the melt inclusions care must be taken to avoid mixed analyses. The hauyne contains around 12 wt.% SO 3 and a hauyne contribution of 10% to the overall carbonatite signal would result in SO 3 -contents of >1.6 wt.%, clearly above the Lc average SO 3 concentration of 0.38 wt.% 1 . A significant contribution of hauyne to the carbonatite analyses can thus be ruled out.…”

“…In a recent paper, we present evidence for immiscible Ca- and Si-rich carbonatite and phonolite melts that were found in hauyne-hosted melt inclusions from the Laacher See volcano in the Eifel, Germany 1 . In their comment, Anenburg and Guzmics (hereafter A&G) describe one experiment performed to assess whether they could replicate the observed immiscible carbonatite liquid from the Laacher See volcano 1 . Their experimental run product shows that the starting material did not melt, that the run was quite obviously subsolidus and, consequently, contained no carbonatitic liquids.…”

“…As to the droplets, some of the melt inclusions reported by ref. 1 show the incomplete physical separation of silicate and carbonate melts as they represent early stages of liquid immiscibility. However, A&G falsely conclude that all carbonatite melts contain such silicate droplets and that all analyses could be contaminated.…”