Journal of Petrology | Volume 39 | Number 11-12 | Pages 2005-2013 | 1998
© Oxford University Press 1998
Processes of Crustal Carbonatite Formation by Liquid Immiscibility and Differentiation, Elucidated by Model Systems
Division of Geological and Planetary Sciences, California Institute of Technology Pasadena, CA 91125, USA
Received September 30, 1997; Revised typescript accepted May 21, 1998
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
Experimental studies on several silicate–carbonate joins provide a framework in the system CaO–Na2O–(MgO + FeO)–(SiO2 + Al2O3) (+ CO2) which illustrates possible processes for the formation of carbonatites. The two key features are the silicate–carbonate liquidus surface, and the miscibility gap liquidus surface. Crystallizing parental carbonated silicate melts may reach a silicate–CO2 eutectic, a silicate–carbonate field boundary, or a miscibility gap. Some hydrous carbonated silicate melts may bypass the high-temperature miscibility gap and reach the silicate–carbonate field boundary. Immiscible carbonate-rich liquids in model systems simulating magmatic conditions tend to be concentrated near calciocarbonatite compositions (<
80% CaCO3; e.g. nepheline sövite), but may be more alkalic from silicate parents with higher Na/Ca values. An immiscible carbonate-rich liquid separating from the high-temperature parent silicate liquid will cool with the precipitation of silicates only, until it reaches the silicate–carbonate field boundary, where it is capable of precipitating carbonate minerals, which can form carbonatite cumulates. Some parents may reach this boundary by direct crystallization, but most probably traverse the miscibility gap. Along this field boundary, the coprecipitation of calcite drives the liquid toward residual alkali-rich compositions. The carbonate liquidus (>85% CaCO3) is a ‘forbidden volume’ for magmas. Vapor loss from carbonatite magma can introduce alkalis into country rocks, but this does not cause alkali depletion of magma; calcite precipitates to maintain the magma composition. Hydrous magnesiocarbonatite magmas can precipitate cumulate sövites. KEY WORDS: carbonatite; liquid immiscibility; nephelinite; sövite
|
Introduction |
|---|
|
TOPABSTRACTIntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
The dominant view through many years, once the feasibility for the existence of carbonatite magmas was demonstrated experimentally (Wyllie & Tuttle, 1960
), has been that carbonatite magmas were generated from parent alkaline silicate magmas by either fractional crystallization or liquid immiscibility (see books by Heinrich, 1966; Tuttle & Gittins, 1966; Le Bas, 1977; Bell, 1989; Bell & Keller, 1995). The report by Wyllie & Huang (1975) that dolomitic melts were formed by partial melting of model lherzolite–CO2 at high pressures, confirmed by experiments of Eggler (1978), revealed the possibility that primary carbonatite magmas might be generated in the mantle. Gittins (1989) once concluded that calciocarbonatites were primary magmas, but Harmer & Gittins (1997), on the basis of field occurrences of dolomitic carbonatites, recently argued that these were precipitated from primary magmas, and that calciocarbonatites were derived from a dolomitic parent magma. Bailey (1993) built a case from the field occurrence of sövites that some calciocarbonatite magmas were primary. Barker (1996) has also adopted the idea of primary carbonatites in addition to his long-held view that carbonatit had been derived from parent alkaline silicate magmas (Barker, 1989).
In a companion paper (Wyllie & Lee, 1998), we concluded that experimental data have confirmed that primary carbonatite magmas from the mantle must be dominated by calcic dolomite, but if this magma is protected from lherzolite by metasomatic wehrlite, it can react progressively between 70 km and 40 km to reach calciocarbonatite composition (Dalton & Wood, 1993). Despite the possibility of primar carbonatite magmas, we remain persuaded by the petrological evidence reported in the books cited above that many carbonatites in alkaline igneous complexes, especially those where the volume of carbonatite is small, are derived from a silicate parental magma.
Using several phase diagrams for silicate–carbonate–H2O systems, Wyllie (1966) illustrated the three kinds of magmatic processes which may be experienced by a CO2–bearing silicate parent liquid: (1) the silicate liquid may be separated from the carbonate-rich liquid by a thermal barrier, and solidif at a terminal liquidus point with precipitation of silicate minerals and evolution of CO2 vapor; (2) the silicate liquid may follow a path of fractional crystallization to enrichment in carbonate components, eventually coprecipitating a carbonate when it reaches the silicate–carbonate liquidus field boundary; (3) the silicate liquid may reach a silicate–carbonate liquid miscibility gap, producing an immiscible carbonate-rich liquid. In a recent series of experiments in progressively more complex model systems, Lee et al. (1994) and Lee & Wyllie (1994, 1996, 1997a, 1997b, 1998) have defined more closely the positions of the two key phase elements for processes (2) and (3): the silicate–carbonate liquid miscibility gap, and the silicate–carbonate liquidus field boundary, both of which enclose forbidden zones into which liquids derived from silicate–CO2 parents cannot enter.
|
The Silicate–Carbonate Liquidus Field Boundary |
|---|
|
TOPABSTRACTIntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
Figure 1 shows projections of the vapor-saturated liquidus surface of CaO–MgO–SiO2–CO2 (Wyllie & Huang, 1975
, 1976; Wyllie & Lee, 1998), partly estimated on the basis of experimental data, at mantle and crustal pressures. The heavy shaded area is the liquidus surface for carbonates, including the dashed line marking a temperature minimum extending from the carbonate system (Irving & Wyllie, 1975; Byrnes & Wyllie, 1981). At crustal pressures, the liquidus field for periclase expands at the expense of magnesite (Fig. 1b). The liquidus field for calcite terminates at a field boundary for coprecipitation of wollastonite at low pressures, and o quartz at high pressures (Huang et al., 1980). The bold curve limiting the carbonate liquidus defines the silicate–carbonate field boundary. Subsolidus magnesite and dolomite break down successively with decreasing pressure (Irving & Wyllie, 1975), which is responsible for reduction of the length of this field boundary as shown in Fig. 1b.
|
, positions (ideal composition) of relevant minerals. En, enstatite; Di, diopside; Wo, wollastonite; CC, calcite; MC, magnesite; 2L, oxide–oxide two-liquid field. (b) Estimated (largely schematic) vapor-saturated liquidus surface at 0.5 GPa. F is the carbonate-rich liquid from carbonated wehrlite. Pe, periclase. The dashed lines for Fo–Di and Di–Qz now represent thermal divides in the system.The field boundaries on the silicate liquidus at 0.5 GPa are estimated from related experimental data. The haplobasaltic eutectic liquid coexisting with model lherzolite, Fo + Opx + Cpx, is situated within the composition triangle Fo–En–Di, and its connections to the silicate–carbonate field boundary are characterized by temperature maxima. Eggler (1974)
demonstrated that the effect of CO2 with increasing pressure was to reduce the SiO2 content of this liquid, causing it to cross the thermal divide Fo–Di. Wyllie & Huang (1975, 1976) and Eggler (1978) determined that the expanding orthopyroxene field reached the silicate–carbonate field boundary (at 2.8 GPa), and generated isobaric invariant liquids E and C at higher pressures, giving liquid compositions for vapor-saturated solidus reactions of dolomite–websterite and dolomite-harzburgite, respectively. Figure 1b shows that haplobasaltic liquids crystallize to minerals plus vapor at crustal pressures because they are prevented by the thermal barriers from reaching the silicate–carbonate field boundary and coprecipitating carbonates. Figure 1a shows that with increased pressure CO2 removes the thermal barriers, and crystallizing liquids may follow paths down to the silicate–carbonate field boundary and coprecipitate carbonates (dolomitic). It should be noted that silicate–CO2 liquids cannot fractionate to cross the silicate–carbonate field boundary; the shaded carbonate liquidus is a forbidden zone for derivatives of silicate parent magmas.
|
The Silicate–Carbonate Liquid Miscibility Gap |
|---|
|
TOPABSTRACTIntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
Figures 2 and 3 are Hamilton projections, which have been widely used to represent miscibility gaps between silicate and carbonate liquids (e.g. Freestone & Hamilton, 1980
; Kjarsgaard & Hamilton, 1988 1989a). These are not ternary systems, but compositional slices through complex systems, projected into the triangle. When crystallization of liquids proceeds, the liquid paths represented in the triangle are actually following paths in multicomponent space. Lee & Wyllie (1996, 1997a, 1997b) have determined the topology of such projections for a variety of compositions, and confirmed the occurrence of two different types: (1) the miscibility gap field boundaries do not reach the silicate–carbonate field boundary; (2) the miscibilit gap field boundaries do intersect the silicate–carbonate field boundary. The relative positions of projected field boundaries vary as a function of pressure and bulk composition (e.g. Mg/Ca, K/Na, Al/Si, H2O and CO2 contents and partial pressures).
|
|
Figures 2 and 3 are representations of these two types, with phase fields distorted as required to permit visualization of the key types of phase elements. These are polythermal projections, not the isothermal projections often used t depict the immiscible liquids. The curves e–o, and e–f/g–o are silicate–carbonate field boundaries, with temperatures and invariant points in Fig. 2 being strongly influenced by the liquidus for CaCO3–Na2CO3 (Cooper et al., 1975
). Possible field boundaries separating liquidus Na2CO3 and nyerereite are drawn from the carbonate side, but no attempt has been made to sketch the many field boundaries traversing the silicate liquidus surface. Some experimental details for a projection corresponding to Fig. 2a have been given by Lee & Wyllie (1997b), and for Fig. 2b by Lee & Wyllie (1996). We are satisfied from our experiments that increasing Mg/Ca in the bulk compositions we used between 1 and 2.5 GPa is associated with reduction in size of the miscibility gap, and a transition from the topology in Fig. 2b to that in Fig. 2a (Lee & Wyllie, 1997a). Brooker & Holloway (1997) showed that a similar change in size of the miscibility gap is effected by changing CO2 pressure (Freestone & Hamilton, 1980). B. A. Kjarsgaard & R. Brooker (personal communication, 1997) maintain that the effect of Mg/Ca is insignificant compared with that of other variables.
Figure 2a shows the geometrical arrangement of a miscibility gap betwee silicate-rich liquids (k–m), and carbonate-rich liquids (k–n), with consolute liquids connected by tie-lines, and with k being the critical point where the tie-line shrinks to nothing and the two liquids become identical. Figure 2b shows a system where the miscibility gap expands to intersect the silicate–carbonate field boundary, which is thus divided into two portions, e–f and g–n, connected by the tie-line f–g; the silicate liquidus is similarly divided into the two shaded areas. Lee & Wyllie (1996, 1997b) presented evidence that the miscibility gap does not reach the alkali-free side.
Figures 1 and 2 illustrate two forbidden zones for initial silicate–CO2 liquids: the carbonate liquidus surface, as described in Fig. 1, and the miscibility gap. Derivative magmas can occupy only the shaded silicate liquidus in Fig. 2, and they can precipitate carbonates only along the silicate–carbonate field boundaries, e–o or e–/g–o. The carbonate-rich liquids k–n and g–n exsolved from the silicate liquids k–m and f–m precipitate silicate minerals, not carbonates.
|
Paths of Crystallization |
|---|
|
TOPABSTRACTIntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
We emphasize again that these triangles are not ternary systems, but reading them as if they were serves the purpose of illustrating the sequences of phase fields that crystallizing liquids will encounter. Figure 3 is a distorted version of Fig. 2a with illustrations of the types of crystallization paths occurring for initial silicate/CO2 liquids in these systems. Path (1) reaches a terminal point precipitating silicates and giving off a vapor (compare Fig. 1b). Pat (2) reaches the silicate–carbonate liquidus boundary e–o (compare Fig. 1a at E and C), and with precipitation of calcite (or dolomite) it must follow a path toward point o with enrichment in alkalis. Path (3) reaches the miscibility gap, exsolves carbonate-rich liquid at the other end of the tie-line, and both liquids chang composition along the field boundaries k–m and k–n as shown, coexisting with silicate minerals. The composition of the carbonate-rich liquid obviously depends on the alkalinit of the silicate liquid.
The situation is not this clear-cut in complex systems. There is also a range of compositions involving the coexistence of two liquids, silicate minerals, and calcite. This is associated with a multicomponent field boundary connecting points such as f and g in Fig. 2b (Lee & Wyllie, 1997b, 1998). The experiments of Kjarsgaard & Peterson (1991) and Kjarsgaard (1997) have encountered such a boundary.
The process represented by path (4) appears to have been missed by many petrologists As long as the carbonate-rich liquid remains in contact with the consolute silicate liquid, only silicates are precipitated (e.g. Kjarsgaard & Peterson, 1991, fig. 3). Only if the carbonate-rich liquid is physically separated from the silicate liquid (k–m) does it follow path (4) across the silicate liquidus, with precipitation of silicates, to reach the silicate–carbonate field boundary e–o, where carbonate is coprecipitated. The consequence is that immiscible carbonatite magmas initially have temperatures identical to those of the silicate magmas, and these must be separated and cooled through a significant temperature interval, with precipitation of only small amounts of silicate minerals, before they coprecipitate carbonates. There is no superheat involved (Kjarsgaard & Peterson, 1991; Lee & Wyllie, 1997b). This was recognized by Kjarsgaard et al. (1995) in their experiments with peralkaline nephelinite.
Petrologists can fit the petrogenetic interpretations of their rock sequences to these schematic paths in model systems. For example, the paths could represent the differentiation of (1) basalts, (2) melilititic magmas, (3) nephelinitic magmas, and (4) high-temperature immiscible carbonate-rich magmas as they cool down to the field boundary where they precipitate carbonatites; this path has not yet been identified in natural occurrences (to our knowledge).
Figure 4 shows extension of the phase elements of Fig. 2 into three dimensions in the pseudoquaternary system
 |
, 1997b) with results in MgO-bearing systems (Baker & Wyllie, 1990; Lee & Wyllie, 1997a). The method of construction for two pressures, 1 and 2.5 GPa, was detailed by Lee & Wyllie (1998), and outlined in the companion paper in this volume (Wyllie & Lee, 1998). Figure 4 is one estimat extrapolated from the high-pressure results (Lee & Wyllie, 1998) for lower pressure, 0.5 GPa, representing crustal conditions. There is uncertainty about the precise shape of the miscibility gap as a function of many complex factors involved in its variation, including pressure, composition (e.g. K/Na, Al/Si), and CO2 content and pressure (e.g. Kjarsgaard & Hamilton, 1989b; Kjarsgaard, 1997; Brooker & Holloway, 1997; P. Ulmer, personal communication, 1997). For example, Kjarsgaard & Hamilton (1989b) determined magnesian immiscible liquids at 0.6 GPa, and Kjarsgaard (1997) reported the exsolution of liquids corresponding to magnesian sövites from CO2-saturated Mg melilitite bulk compositions (0.2–2.5 GPa); both studies indicated consolute tie-lines which cut across the silicate volum of Fig. 4. We have addressed some of these issues (Lee & Wyllie 1998), and emphasize again that the tetrahedron was constructed to illustrate types of processes for the major element compositions which we selected, while providing some compositional constraints, but that precise experiments to locate phase boundar be required for specific rock sequences and particular conditions.
|
The base of Fig. 4 is similar to Fig. 2b, and the front face is similar to Fig. 1b (with additional components). The two field boundaries, silicate–carbonate and consolute liquids, extend upwards from the base as surfaces in the tetrahedron, with shapes depicted by the contours for constant per cent (MgO + FeO*). A horizontal section at 10% or 20% (MgO + FeO*) would have the topology of Fig. 2a. The two surfaces overlap along a curved line rising from the base (the equivalents of points f and g in Fig. 2b). These two surfaces divide the tetrahedron into three volumes: (1) immiscible liquids within the light-shaded dome; (2) the silicate liquidus between the two surfaces; (3) the carbonate/periclase liquidus on the low-SiO2 side of the dark-shaded silicate–carbonate/periclase field boundary (compare Fig. 1 for the expansion of a periclase liquidus volume with decreasing pressure, at the expense of primary carbonate).
Let us consider primitive nephelinites or alkali basalts in the mantle (Lee & Wyllie, 1998, figs 7 and 9). Their projected compositions are situated well within the silicate volume, at a level of about 25–30% (MgO + FeO*). Primary nephelinites and melilitites parental to carbonatites may have lower MgO/CaO (R. Brooker, personal communication, 1997). When considering uprise of primary magmas with concomitant fractionation, many factors enter to complicate th model system. In particular, variations in the CO2 solubility and partial pressure are associated with changes in size of the miscibility gap, different for each initial magma composition, and for different uprise histories. Nevertheless, isobaric Fig. 4 can be used to illustrate a polybaric path in general terms. A primary liquid rising from the mantle would follow a path away from the apex, such as (3), and it may reach the boundary surface of the miscibility gap, a prospect enhanced because our evidence indicates that the volume increases in size with decreasing pressure (Lee & Wyllie, 1997a, 1998). The changes in positions of the two surfaces in Fig. 4 for real systems with decreasing pressure are subject to complex controls, including the factors discussed above. If intersection occurs, the magma compositions then follow paths along the surface, exsolving carbonate-rich magmas, in the pattern illustrated for path (3) in Fig. 3. Kjarsgaard & Hamilton (1989a) explored the changes involved in polybaric fractionation paths, with all changes combining to increase the size of the miscibility gap and thus enhancing the prospect that liquid immiscibility would occur, followed by continued exsolution of carbonate-rich melt, as depicted by path (3) in Figs 3 and 4.
|
|
Differentiation of Carbonatite Magmas |
|---|
|
TOPABSTRACTIntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
Carbonate-rich magmas become carbonatite magmas (precipitating carbonates) only when they reach the silicate–carbonate field boundary. This may occur via fractional crystallization, as in path (2) (Fig. 3), or via liquid immiscibility through one of two routes. Under some conditions, a carbonate-rich liquid may be exsolved from a silicate parent through a temperature interval, while coexisting with both silicate mineral and carbonates, corresponding to point g in Fig. 2b, before becoming separated from the silicate magma parent and following a path along g–o in Fig. 2b. The second route is represented by paths (3) and (4) in Fig. 3, where the immiscible liquids coexist only with silicate minerals (e.g. along k–n or g–n in Fig. 2). Only after the carbonate-rich liquid is separated from the silicate liquid does it differentiate down the liquidus surface (across k–n–e–o, or g–n–o in Fig. 2) to reach the lower-temperature silicate–carbonate field boundary.
From results of several experimental studies, we conclude that for many magmatic systems, when carbonate-rich liquids begin to precipitate calcite, they tend to cluster the composition range 70–80 wt % CaCO3, 10–20% Na2CO3, and 5–10% silicates. The residual liquids then follow the silicate–carbonate boundary (the e–o and g–o equivalent of Figs 2 and 3), toward enrichment in alkalis (Lee & Wyllie, 1996, 1997b). Kjarsgaard et al. (1995), in experiments with carbonated peralkaline nephelinite, generated immiscible alkali-rich, carbonate-rich liquids directly at low pressures and moderate temperatures, which with minor fractionation could yield natrocarbonatite magma.
The effect of H2O
The most obvious effect of adding H2O under pressure is to lower both liquidus and solidus temperatures, and this has other consequences. Here we illustrate the effect of H2O on silicate–carbonate liquid immiscibility, and also on the silicate–carbonate field boundaries. Watkinson & Wyllie (1971) concluded from experiments in NaAl-SiO4–CaCO3–H2O at 0.1 GPa at temperatures from 960°C to 625°C that fractional crystallizatio of hydrous nepheline-normative silicate melts could yield low-temperature, carbonatite-like liquids. Kjarsgaard & Hamilton (1988) reported a high-temperature miscibility gap across the join NaAlSiO4–CaCO3 (H2O-free), concluding that the two sets of data were ‘in conflict’. Lee & Wyllie (1994) repeated one selected run near the silicate–calcite–aqueous vapor field boundary at 833°C, where the calcite had just begun to crystallize, using analytical facilities unavailable in 1970. This confirmed the reported phase assemblage of Watkinson & Wyllie (1971), and revealed no sign of the liquid immiscibility existing at higher temperatures. They analyzed the quenched liquid coexistin with melilite, cancrinite, and calcite, with detailed evaluation of the effects of crystallization during the quench; the liquid composition (recalculated anhydrous) was found to lie within the known high-temperature miscibility gap in the Hamilton projection.
Figure 5 shows an interpretation in an isobaric TX prism for the Hamilton projection. The shaded volume represents the miscibility gap, with isothermal tie-lines shown at high temperature, and a projection of its limits on the bottom triangle. The high-temperature miscibility gap closes with decreasing temperature in the presence of H2O vapor [as observed by the experiment; see Lee & Wyllie (1994) for supporting evidence]. The bold lines represent schematic paths of crystallization for different silicate liquids, shown in TX space and in projection o the bottom triangle. Path (3) corresponds to high-temperature, low-H2O path (3) in Figs 3 and 4, reaching the surface of the miscibility gap, and exsolving a carbonate-rich liquid. In contrast, path (5) for a hydrous liquid, corresponding to the path determined by Lee & Wyllie (1994), passes below the miscibility gap. The single analyzed liquid composition marks the change in direction shown in the projected path (5), because this liquid has just begin to precipitate calcite in addition to silicates. The silicate–calcite–aqueous vapor field boundary is therefore shifted in composition further away from CaCO3 than those boundaries depicted in the dry systems (at higher pressures) (compare Figs 2 and 4).
|
The effect of H2O on silicate–carbonate field boundaries is shown in Fig. 6. Figure 6a is the simplest system (CaO–CO2–H2O, Wyllie & Tuttle, 1960
) showing the effect of H2O on calcite melting temperatures. With addition of H2O, the composition of melted calcite (CC) changes toward portlandite (CH), with marked decrease in liquidus temperature from 1310°C to 645°C. The coexisting vapor phase changes progressively from pure CO2 to almost pure H2O (F) at the eutectic temperature, with partition of H2O strongly favoring the vapor. This result provides a simple hydrous liquid capable of precipitating only calcite through hundreds of degrees, with the potential to yield a pure sövite cumulate. The final liquid at E would also precipitate portlandite, a mineral not characteristic of carbonatites.
|
The liquids in Fig. 6a are incomplete in representation of carbonatite magmas because they lack silicates, and liquids derived from a silicate parent magma cannot enter the forbidden zones of carbonate liquidi. The simplest system including silicates is CaO–SiO2–CO2–H2O, and Fig. 6b shows a projection of the vapor-saturated liquidus surface onto the plane Ca2SiO4–CaCO3–Ca(OH)2 (Wyllie & Haas, 1965
). The orientation of CaCO3–Ca(OH)2 corresponds to that in Fig. 6a, and the orientation with respect to silicate–calcite corresponds to that of the Hamilton projection. The shaded area is the liquidus field for calcite, and the bold line is the silicate–calcite–aqueous vapor field boundary (compare Fig. 1). Wyllie & Haas (1965) showed that for initial liquids on the SiO2 side of Ca2SiO4, there was a thermal barrier preventing differentiation to the low-temperature synthetic carbonatite magma. But carbonated, hydrous silicate liquids with less SiO2 than Ca2SiO4 could follow differentiation paths to the silicate–carbonate field boundary between calcite and spurrite–calciochondrodite, which exists through a wide temperature interval, from 1180°C to 637°C, down to temperatures at the lower end of path (5) in Fig. 5. The haplocarbonatite magma composition is limited by the silicate–carbonate field boundary, and cannot enter the shaded forbidden zone. The maximum CaCO3 content is 65%, decreasing at lower temperatures as the hydrous component increases and the silicate component decreases.
Fanelli et al. (1986; in Wyllie, 1989) showed that H2O lowered the liquidus temperature of CaCO3–MgCO3, with the result that dolomitic melts precipitated sövites (Lee et al. 1998). Harmer & Gittins (1997) showed that Na2CO3 had a similar effect.
Differentiation toward alkali depletion by vapor loss?
There have been recurrent suggestions that an alkali-rich carbonatite magma could lose alkalis through vapor loss, thus forming a calciocarbonatite magma. Model phase diagrams illustrate that although a vapor phase can transfer alkalis and other dissolved components from magma into country rocks (explaining fenites), this does not cause alkali depletion of the magma, but simply results in the precipitation of more calcite.
Pseudoternary system Fig. 7 is diagrammatic. CaCO3–(Na,K)2CO3 (Cooper et al., 1975) is in the same orientation as in the Hamilton projections, Figs 2 and 3, with volatile components represented by the lower left corner. The curve 2–3 is the vapor-saturated liquidus field boundary limiting the liquidus for calcite; the curve 2'–3' is the corresponding vaporus field boundary, indicating (Na,K) solubility higher than that of Ca. Crystallization of calcite from liquids on the field boundary 2–3 changes the liquid toward alkali-rich carbonatite magma. A magma of similar composition could be formed by liquid immiscibility (Figs 2 and 3). The fractionation of suc a magma can be illustrated by liquids on the field boundary in Fig. 7.
Point (1) is an initial liquid which precipitates calcite, and follows the path 1–2 to the vapor-saturated liquidus field boundary, with compositions of the phases when liquid reaches 2 given by the corners of the three-phase triangle CC–2–2'. Under equilibrium conditions, with no loss of vapor, calcite would continue to precipitate while liquid and vapor changed compositions along the paths 2–3 and 2'–3', respectively, with both phases being enriched in alkalis as calcite was precipitated. If fractionation occurred and the vapor (2') were removed at constant P and T, liquid (2) would not change composition. When the temperature decreased, it would simply precipitate calcite and release vapor, and the liquid and vapor would again follow the paths 2–3 and 2'–3'. Vapor (3') has much higher (Na + K)/Ca than does liquid (3), but if the vapor (3') were removed this would still have no effect on the composition of the liquid (3). The liquid would not be depleted in alkalis by loss of vapor. What if the vapor loss were associated with decreas in pressure? The liquidus field boundary would be shifted away from the volatile components, so the liquid would change composition slightly, but there would be no extensive loss of alkalis, nor reversal of crystallization direction of the liquid.
|
Summary of Conclusions |
|---|
|
TOPABSTRACTIntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
Most carbonatite magmas are derived from carbonated silicate parents, which could follow crystallization paths (1) to a terminal point where they precipitate silicates and evolve CO2 without carbonatites (e.g. basaltic), (2) to a field boundary wher silicates and carbonates are coprecipitated (e.g. melilititic), or (3) to a miscibility gap where a carbonate-rich magma is exsolved (e.g. nephelinitic). Carbonatite magmas do not precipitate carbonate-rich rocks until they reach the silicate–carbonate field boundary, whether this be by fractional crystallization as in Fig. 1a, path (2) in Fig. 3, and varied paths in Fig. 6b; or via a miscibility gap as in paths (3) in Figs 3, 4 and 5. The silicate–carbonate field boundary appears to limit the compositions of carbonatite magmas to <
85 wt % CaCO3. The phase diagrams indicate that crystallization along this field boundary causes precipitation of calcite along with silicates, and enrichment of the residual liquids in hydrous components, alkali carbonates, and surely fluorine as well (e.g. Jago & Gittins, 1991). Vapor loss does not cause alkali depletion of carbonatite magma. We see no prospect in the phase relationships that a parent natrocarbonatite liquid could follow a crystallization path toward residual calciocarbonatite, a process that has been vigorously debated (see Barker, 1996). Carbonatites with compositions within the forbidden carbonate liquidus volume cannot represent original liquid compositions, and they must represent cumulates. Hydrous magnesiocarbonatite (dolomitic) magmas can precipitate cumulate sövites.
|
Acknowledgements |
|---|
We thank D. R. Baker, D. H. Eggler, B. A. Kjarsgaard and an anonymous reviewer for their critical comments. This research was supported by the Earth Science section of the US National Science Foundation, grant EAR-9218806. This is Contribution 8572 of the Division of Geological and Planetary Sciences, California Institute of Technology.
* Corresponding author. Telephone: (626)-395-6239. Fax: (626)-568-0935. e-mail: wjl{at}gps.caltech.edu
|
References |
|---|
|
TOPABSTRACTIntroductionThe Silicate-Carbonate Liquidus...The Silicate-Carbonate Liquid...Paths of CrystallizationDifferentiation of Carbonatite...Summary of ConclusionsReferences |
|---|
Bailey D. K. Carbonate magmas. Journal of the Geological Society, London (1993) 150:637–651.
Baker M. B., Wyllie P. J. Liquid immiscibility in a nephelinite–carbonate system at 25 kbars and implications for carbonatite origin. Nature (1990) 346:168–170.
Barker D. S. Field relations of carbonatites. In: Carbonatites: Genesis and Evolution—Bell K., ed. (1989) London: Unwin Hyman. 38–69.
Barker D. S. Carbonatite volcanism. In: Undersaturated Alkaline Rocks: Mineralogy, Petrogenesis, and Economic Potential. Mineralogical Association of Canada, Short Course—Mitchell R. H., ed. (1996) 24:45–61.
Bell K. Carbonatites: Genesis and Evolution (1989) London: Unwin Hyman.
Bell K., Keller J., eds. Carbonatite Volcanism: Oldoinyo Lengai and the Petrogenesis of Natrocarbonatites. IAVCEI Proceedings in Volcanology 4 (1995) Berlin: Springer-Verlag.
Brooker R., Holloway J. R. The role of CO2 saturation in silicate–carbonatite magmatic systems. In: Geological Association of Canada–Mineralogical Association of Canada Annual Meeting, Abstract Volume (1997) A–18.
Byrnes A. P., Wyllie P. J. Subsolidus and melting relations for the join CaCO3–MgCO3 at 10 kb. Geochimica et Cosmochimica Acta (1981) 45:321–328.[Web of Science]
Cooper A. F., Gittins J., Tuttle O. F. The system Na2CO3–K2CO3–CaCO3 at 1 kilobar and its significance in carbonatite petrogenesis. American Journal of Science (1975) 275:534–560.
Dalton J. A., Wood B. J. The compositions of primary carbonate melts and their evolution through wallrock reaction in the mantle. Earth and Planetary Science Letters (1993) 119:511–525.[Web of Science]
Eggler D. H. Effect of CO2 on the melting of peridotite. Carnegie Institution of Washington, Yearbook (1974) 73:215–224.
Eggler D. H. . American Journal of Science (1978) 278:305–343.
Fanelli M. F., Cava N., Wyllie P. J. Calcite and dolomite without portlandite at a new eutectic in CaO–MgO–CO2–H2O, with applications to carbonatites. In: Morphology and Phase Equilibrium of Minerals. Proceedings 13th General Meeting, International Mineralogical Association, Bulgarian Academy of Science, Sofia (1986) 313–322.
Freestone I. C., Hamilton D. L. The role of liquid immiscibility in the genesis of carbonatites—an experimental study. Contributions to Mineralogy and Petrology (1980) 73:105–117.[Web of Science]
Gittins J. The origin and evolution of carbonatite magmas. In: Carbonatites: Genesis and Evolution—Bell K., ed. (1989) London: Unwin Hyman. 580–600.
Harmer R. E., Gittins J. The origin of dolomitic carbonatites: field and experimental constraints. Journal of African Earth Sciences (1997) 25:5–28.
Heinrich E. W. The Geology of Carbonatites (1966) Chicago, IL: Rand McNally.
Huang W.-L., Wyllie P. J., Nehru C. E. Subsolidus and liquidus phase relationships in the system CaO–SiO2–CO2 to 30 kbar with geological applications. American Mineralogist (1980) 65:285–301.[Abstract]
Irving A. J., Wyllie P. J. . Geochimica et Cosmochimica Acta (1975) 39:35–53.[Web of Science]
Jago B. C., Gittins J. The role of fluorine in carbonatite magma evolution. Nature (1991) 349:56–58.
Kjarsgaard B. A. Carbonatites in context: differentiation trends of carbonated alkaline ultrabasic silicate magmas based on field and experimental studies. In: Geological Association of Canada–Mineralogical Association of Canada Annual Meeting, Abstract Volume (1997) A-79.
Kjarsgaard B. A., Hamilton D. L. Liquid immiscibility and the origin of alkali–poor carbonatites. . Mineralogical Magazine (1988) 52:43–55.[Web of Science]
Kjarsgaard B. A., Hamilton D. L. The genesis of carbonatites by immiscibility. In: Carbonatites: Genesis and Evolution—Bell K., ed. (1989a) London: Unwin Hyman. 388–404.
Kjarsgaard B. A., Hamilton D. L. Carbonatite origin and diversity. Nature (1989b) 338:547–548.[Web of Science]
Kjarsgaard B., Peterson T. Nephelinite–carbonatite liquid immiscibility at Shombole Volcano, East Africa: petrographic and experimental evidence. Mineralogy and Petrology (1991) 43:293–314.[Web of Science]
Kjarsgaard B. A., Hamilton D. L., Peterson T. D. Peralkaline nephelinite/carbonatite liquid immiscibility: comparison of phase compositions in experiments and natural lavas from Oldoinyo Lengai. In: Carbonatite Volcanism: Oldoinyo Lengai and the Petrogenesis of Natrocarbonatites. IAVCEI Proceedings in Volcanology 4—Bell K., Keller J., eds. (1995) Berlin: Springer-Verlag. 163–190.
Le Bas M. J. Carbonatite–Nephelinite Volcanism (1977) London: John Wiley.
Lee W.-J., Wyllie P. J. Experimental data bearing on liquid immiscibility, crystal fractionation, and the origin of calciocarbonatites and natrocarbonatites. International Geology Review (1994) 36:797–819.
Lee W.-J., Wyllie P. J. Liquid immiscibility in the join NaAlSi3O8–CaCO3 to 2.5 GPa and the origin of calciocarbonatite magmas. Journal of Petrology (1996) 37:1125–1152.
Lee W.-J., Wyllie P. J. Liquid immiscibility between nephelinite and carbonatite from 2.5 to 1.0 GPa compared with mantle melt compositions. Contributions to Mineralogy and Petrology (1997a) 127:1–12.[Web of Science]
Lee W.-J., Wyllie P. J. Liquid immiscibility in the join NaAlSiO4–NaAlSi3O8–CaCO3 at 1.0 GPa: implications for crustal carbonatites. Journal of Petrology (1997b) 38:1113–1135.
Lee W.-J., Wyllie P. J. Petrogenesis of carbonatite magmas from mantle to crust, constrained by the system CaO–(MgO+ FeO*)–(Na2O + K2O)–(SiO2+ Al2O3+ TiO2)–CO2. Journal of Petrology (1998) 39:495–517.
Lee W.-J., Wyllie P. J., Rossman G. R. CO2-rich glass, round calcite crystals and no liquid immiscibility in the system CaO–SiO2–CO2 at 2.5 GPa. American Mineralogist (1994) 79:1135–1144.[Abstract]
Lee W.-J., Fanelli M. F., Cava N., Wyllie P. J. Calciocarbonatite and magnesiocarbonatite rocks and magmas represented in the system CaO–MgO–CO2–H2O at 0.2 GPa. In: Contributions to Mineralogy and Petrology (1998).
Tuttle O. F., Gittins J. Carbonatites (1966) New York: Interscience.
Watkinson D. H., Wyllie P. J. Experimental study of the join NaAlSiO4–CaCO3–H2O and the genesis of alkalic rock–carbonatite complexes. Journal of Petrology (1971) 12:357–378.
Wyllie P. J. Experimental studies of carbonatite problems: the origin and differentiation of carbonatite magmas. In: Carbonatites—Tuttle O. F., Gittins J., eds. (1966) New York: John Wiley. 311–352.
Wyllie P. J. Origin of carbonatites: evidence from phase equilibrium studies. In: Carbonatites: Genesis and Evolution—Bell K., ed. (1989) London: Unwin Hyman. 500–545.
Wyllie P. J., Haas J. L. Geochimica et Cosmochimica Acta (1965) 29:871–892.[Web of Science]
Wyllie P. J., Huang W.-L. Peridotite, kimberlite, and carbonatite explained in the system CaO–MgO–SiO2–CO2. Geology (1975) 3:621–624.
Wyllie P. J., Huang W.-L. Carbonation and melting reactions in the system CaO–MgO–SiO2–CO2 at mantle pressures with geophysical and petrological applications. Contributions to Mineralogy and Petrology (1976) 54:79–107.[Web of Science]
Wyllie P. J., Lee W.-J. Model system controls on conditions for formation of magnesiocarbonatite and calciocarbonatite magmas from the mantle. Journal of Petrology (1998) 39:1885–1893.
Wyllie P. J., Tuttle O. F. The system CaO–CO2–H2O and the origin of carbonatites. Journal of Petrology (1960) 1:1–46.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Berger, N. Ennih, J.-C. C. Mercier, J.-P. Liegeois, and D. Demaiffe The role of fractional crystallization and late-stage peralkaline melt segregation in the mineralogical evolution of Cenozoic nephelinites/phonolites from Saghro (SE Morocco) Mineralogical Magazine, June 15, 2009; 73(1): 59 - 82. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||