John Gittins

The origin of dolomitic carbonatites: field and experimental constraints

Abstract Carbonatites are most commonly either calcitic or dolomitic/ankeritic with very few types in between - what is generally referred to as the bimodal distribution. There is a widely held view that dolomitic carbonatites are less abundant than calcitic carbonatites and that these arise as sub-solidus alteration products of primary calcitic carbonatites. It is demonstrated here that dolomitic types are far more common than is sometimes appreciated and that they are particularly abundant in old Precambrian cratonic regions such as the Zimbabwean and Kaapvaal Cratons and the Archaean parts of the Canadian Shield. The field, petrographic and chemical features favour these dolomitic carbonatites being magmatic rather than arising from sub-solidus replacement of calcitic carbonatites. Experimental studies show that partial melting of carbonated mantle peridotite produces a carbonate liquid with high Mg# and MgO content and an alkali content of up to 6%. On ascent through the mantle from its original generation site this primitive carbonatite will be destroyed by reaction with lherzolite and harzburgite if it remains in equilibrium with the surrounding mantle. If, however, the melt is shielded from the surrounding mantle by a lining of metasomatic wehrlite on the conduits or if it rises too rapidly for equilibrium to be maintained, it is able to escape the mantle and rise into the crust. Reaction between primitive magnesian carbonate melt and wall-rock wehrlite shifts the composition of the melt to more Ca-rich (“calcitic”) compositions. It is argued that such liquids are capable of generating the complete range of carbonatite compositions recognised at the surface. Dolomite melts incongruently at low pressures and so will only crystallise from a magnesian carbonate magma at temperatures below the dolomite dissociation reaction. These conditions are dictated by the P-T trajectory of the ascending melt as well as the nature and concentration of minor “fluxing” constituents in the melt such as fluorine and alkalis. As a result calcite is the liquidus phase over a wide range of P-T-X conditions. Several authors have suggested that many calcite-rich carbonatites formed as cumulate-enriched crystal mushes. Such calcite mushes could be readily generated from magnesian, essentially “dolomitic”, parental magmas. It is argued that no reasonable petrogenetic mechanism exists whereby magnesian carbonatite magmas could be generated from calcitic parental melts: it is argued that the reverse is true.

What is ferrocarbonatite? A revised classification

The term “ferrocarbonatite” has been in use for about twenty years but is not adequately defined. The IUGS system of igneous rock nomenclature defines it mineralogically as a carbonatite in which “the main carbonate mineral is iron rich” and chemically as a carbonatite in which (in weight percent) CaO:CaO + MgO + FeO + Fe2O3 + MnO 0.75, magnesiocarbonatite: CCMF 1.0, ferruginous calcicarbonatite: 0.5 < CCMF < 0.75; MgO/FeO∗ < 1.0, ferrocarbonatite: CCMF < 0.5; MgO/FeO∗ < 1.0 where CCMF is the molar ratio CaO/(CaO + MgO + FeO∗ + MnO) and FeO∗ refers to molar FeO if FeO and Fe2O3 are both determined and total Fe as FeO if not. It is proposed that the term ferrocarbonatite only be used in this modified chemical sense: carbonatites in which the main carbonatite is Fe rich are adequately described using the modal mineralogy.

The petrology of the Mont Saint Hilaire complex, southern Quebec: An alkaline gabbro-peralkaline syenite association

Abstract The Mont Saint Hilaire complex consists of an older (133 Ma) suite of layered cumulates of titanaugite, kaersutite, plagioclase and titaniferous magnetite, and two younger (122 Ma) suites, one a thick ring dyke of nepheline-olivine diorite to monzonite, and the other a pipe or funnel-like mass of peralkaline nepheline-sodalite syenite and porphyry associated with a variety of breccias. Trace element data suggest derivation of the older suite from a garnet-bearing source. Only minor amounts of possible liquid compositions are preserved in this suite. The nepheline and olivine-bearing suite followed a course of fractionation from gabbroic to monzonitic compositions involving fractionation of pyroxene, magnetite, apatite and plagioclase. Field and trace element data suggest mixing of the evolved liquid with a saline brine at crustal depths produced the strongly nepheline-normative peralkaline magma. Rich in Na and Cl, the brine was poor in other major and trace elements, and had a high initial Sr ratio. The localization and extended time of emplacement of the complex appear to be due to upward migration of a thermal anomaly from the base of a lithosperic plate.

Problems inherent in the application of calcite-dolomite geothermometry to carbonatites

Calcite-dolomite geothermometry has been used extensively to determine the temperature attained during regional metamorphism of limestones. Several attempts have been made to apply the technique to carbonatites. Although doubts have been expressed recently about the realiability of the method for limestones, the difficulties inherent in using it to estimate carbonatite magma temperatures are so profound that it is of very little value, and published carbonatite magma temperatures based on the method are dubious. These studies have tended to overlook the fact that the highest temperature that can be obtained by the method is still below the liquidus temperature. They have further tended to overlook the fact that Mg diffusion from calcite into coexisting dolomite continues during sub-solvus cooling and that in carbonatites this diffusion is likely to be far more extensive than in metamorphic marbles because of the ubiquitous presence of an alkali-H2O-CO2-halogen fluid. This diffusion is very variable within single “perthitic” carbonate grains and from grain to grain. The technique of dissolution of carbonatites in cold dilute HCl leads to difficulties and should be avoided. Electron microprobe analysis can be used only on unexsolved calcite or on calcite that has exsolved only very fine dolomite lamellae. The closest approach to magmatic temperatures is obtained by wet chemical analysis of coarse calcite-dolomite “perthites”. Published carbonatite magma temperatures based on calcite-dolomite geothermometry are misleadingly low and tend to overemphasize the 300–500 ° C temperature range, whereas evidence is presented for temperatures of about 900 ° C in one Ontario carbonatite. Except in rare cases, calcite-dolomite geothermometry cannot usefully be applied to carbonatites.

Phlogopitization of pyroxenite; its bearing on the composition of carbonatite magmas

Phlogopitization of pyroxenite is common in contact zones between clinopyroxenites and carbonatite dikes of the Cargill ultramafic rock—carbonatite complex near Kapuskasing, Ontario. The most typical development is a mica zone 1–10 cm wide but phlogopite is also developed in a more pervasive manner throughout the groundmass of several types of ultramafic rock. Fenitization is most commonly thought of as a process whereby aegirine and riebeckitic amphiboles are formed in the host rock while feldspar is recrystallized and silica progressively removed. Phlogopitization of pyroxenite can properly be referred to, however, as a type of fenitization. It is clearly related to the intrusion of carbonatite into pyroxenite and is further testimony to the fact that many carbonatite magmas are initially alkalic but lose alkalies to the surrounding rocks and crystallize as calcitic and dolomitic carbonatite with alkali contents restricted to the amounts that could be fixed as micas, pyroxenes or amphiboles. This in turn is controlled by the silica and alumina activity of the carbonatite magma. Abundant evidence for considerable amounts of fluorine in carbonatite magmas suggests that alkalies may be transported into the country rocks as fluorides. It is further suggested that late-stage feldspathization in carbonatite complexes is explained by the abstraction of potassic halide solutions from the crystallizing carbonatite magma. The conclusion seems inescapable that alkali carbonatite magmas, far from being the curiosity thought by many petrologists, are in fact very common during the evolutionary history of carbonatites. The common calcitic and dolomitic carbonatites have not generally crystallized from a magma of the same composition but are the residue remaining after the abstraction of an alkali-rich aqueous fluid. Consequently, there is a need to redesign the experimental phase equilibrium approach to problems of carbonatite genesis in order to take account of the presence of alkalies in most carbonatite magmas.

Intergrowths of nepheline-potassium feldspar and kalsilite-potassium feldspar: A re-examination of the ‘pseudo-leucite problem’

AbstractThe alkalic ultramafic Batbjerg intrusion of East Greenland contains rocks in which nepheline and leucite are important constituents. In addition, there are vermicular, ‘finger print’ intergrowths of nepheline with potassium feldspar, and patchy to micrographic intergrowths of kalsilite with potassium feldspar. The history of the ‘pseudoleucite problem’ is reviewed, and it is suggested that the term pseudoleucite be restricted to intergrowths of nepheline with alkali feldspar that appear to be pseudomorphs with the crystal morphology of leucite. It is further suggested that flame-like or feather-like finger print intergrowths of nepheline with alkali feldspar, that are either interstitial to the other minerals of the rock or have grown perpendicularly on relative large and often euhedral nepheline grains are an entirely different problem and are best explained by late-stage magmatic crystallization within the system NaAlSiO4-KAlSiO4-SiO2-H2O.In the Batbjerg intrusion the early crystallization of nepheline was followed by the co-crystallization of nepheline with leucite, or in some cases by nepheline and a silica-rich leucite. Although the magma was essentially dry, as indicated by the dominantly pyroxenitic character of the rocks, water pressure rose toward the late stages of crystallization as indicated by the presence of phlogopite and occasionally both amphibole and zeolite. Shrinkage of the leucite stability field attendant upon this rise in

Significance of Strontium Isotope Ratios in Theories of Carbonatite Genesis

P_{H_2 O}

The role of fluorine in carbonatite magma evolution

left the liquid that was crystallizing nepheline and leucite stranded on the nepheline-alkali feldspar cotectic. Shrinkage occurred too rapidly for the liquid to remain at the reaction point of the system, and leucite, therefore, was not resorbed. The remaining liquid crystallized rapidly as ‘flames’ of vermicular intergrowth of nepheline with potassium feldspar (composition Ne 24.0, Ks 45.9, Qz 30.1), a texture that might be attributable to supercooling. Silica-rich leucite compositions (Ks 68.8, Qz 31.2) decomposed to intergrowths of kalsilite with potassium feldspar but reaction kinetics, or possibly variations in