| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology Volume 42 Number 2 Pages 377-405 2001
© Oxford University Press 2001
The Turiy Massif, Kola Peninsula, Russia: Isotopic and Geochemical Evidence for Multi-source Evolution
OTTAWA–CARLETON GEOSCIENCE CENTRE, DEPARTMENT OF EARTH SCIENCES, CARLETON UNIVERSITY, OTTAWA, ONT., K1S 5B6, CANADA
Received January 1, 1999; Revised typescript accepted May 17, 2000
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
The Turiy Massif, lying within the Kandalaksha Graben, and on the southern coast of the Kola Peninsula, contains carbonatites, phoscorites, melilitolites, ijolites and pyroxenites within one central and four surrounding satellite complexes. Sr–Nd isotopic data from the central complex phoscorites and carbonatites, and the nearby Terskii Coast kimberlites, combined with other recently published data on the Devonian Kola Alkaline Province, allow us to redefine the position of the Kola Carbonatite Line (KCL) of Kramm (European Journal of Mineralogy 5, 985–989, 1993). We propose that the revised-KCL mantle sources include a lower-mantle plume, and a second enriched source, which also contributed to the Terskii Coast and Archangelsk kimberlites. The Turiy Massif silicate rocks and northern complex carbonatites have more enriched isotopic signatures than the distinct, and depleted signatures of the central complex phoscorites and carbonatites, particularly with respect to
Sr. This is probably due to the contamination of parental magmas, originally derived from the KCL end-members, by crustal material. The phoscorites and carbonatites show unusually enriched stable isotope 
13CPDB values with respect to their conjugate 18OSMOW values. The trace element signatures of the silicate rocks are generally consistent with derivation from the magma sources proposed above. KEY WORDS: carbonatite; Kola; melilitolite; Turiy
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
The geographical and geochemical relationships between silicate and carbonatite magmas are both complex and diverse, and have been at the centre of many scientific studies during the last century [summaries have been given by Bell (1989)
and Bell & Keller (1995)]. Current theories for their origin include the derivation of carbonatites as primary magmas (e.g. Bailey, 1993; Harmer & Gittins, 1998), their differentiation from a silicate parent (e.g. Kjarsgaard, 1997), or their separation from a silicate parent by liquid immiscibility (e.g. Brooker, 1998). To evaluate the role of each of these processes within an individual complex or intrusion, it becomes important to assess the geochemical relationships between spatially related silicate and carbonatite rocks. In addition, trace element and isotopic characteristics from both carbonatite and silicate rocks may help to characterize the mantle source(s) from which the magmas were derived. The Turiy Massif, which has been subjected to little alteration since its intrusion, and for which a comprehensive suite of drill core samples exists, offers an ideal opportunity for such a detailed study. |
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
The Turiy Massif, intruded into Proterozoic charnockites and Riphean sandstones, is located on the southern shore of the Kola Peninsula in NW Russia, just inside the Kandalaksha Graben (Fig. 1). Dominantly Palaeozoic in age, Turiy is a member of a large suite of Upper Devonian alkaline intrusions located within the Kola Peninsula (Kramm et al., 1993
). These include pyroxenite–olivinite (e.g. Afrikanda), ijolite–syenite–carbonatite (e.g. Khibina, Kandagubskii) and pyroxenite–ijolite–melilitolite–carbonatite suites (e.g. Kovdor, Turiy) [the reader is referred to Le Maitre (1989) for a definition of the rock types used in this study, except for the terminology used to define the melilite-bearing rocks, which is taken from Dunworth & Bell (1998)]. Most of these massifs are emplaced along old suture or fracture zones, or reactivated graben structures. The most prominent of these are the NE–SW-trending Kontozero Graben (which controlled the emplacement of the Khibina and Lovozero massifs) and the NW–SE-trending Kandalaksha Graben, which controlled the emplacement of the Sokli, Kovdor, Kandagubskii and Turiy massifs, the Kandalaksha dyke swarms, and the Terskii Coast and Archangelsk kimberlite pipes. The lithospheric thickness within the Kandalaksha Graben during the Devonian was estimated to be 
125 km by Mahotkin et al. (2000) A detailed description of the Palaeozoic magmatism in the Kola Peninsula has been provided by Kukharenko et al. (1965), Arzamastsev (1994), Kogarko et al. (1995) and Mahotkin et al. (2000), and in the summary papers of Gerasimovsky et al. (1974) and Kogarko (1987).
|
The Turiy complex has a total area of 50 km2 and is overlain by 30 m of glacial till. There are five hypabyssal complexes within the massif—one central, and four ‘satellite’ complexes, north, south, east and west (Fig. 2). The western complex is partially exposed along the shore of the Turiy peninsula, whereas the ‘northern complex’ consists of a small group of scattered intrusions to the north of the central complex. Evidence based on information obtained from drill core material shows the following intrusive sequence: (1) pyroxenites (central complex); (2) early ijolites (all complexes); (3) melilitolites (all complexes); (4) later ijolite–urtites (central complex); (5) phoscorites (central complex); (6) carbonatites (central and northern complexes) (Bulakh & Ivanikov, 1984). Carbonatite dykes that also occur on the south coast belong to the youngest set of a three-phase dyke swarm (pre-, syn- and post-massif emplacement) that consist mainly of nephelinites, melilitites and related rocks. An aureole of fenitized country rocks 500–800 m wide surrounds the massif (Bulakh & Ivanikov, 1984). For a comprehensive summary of previous work done on the Turiy Massif and accompanying dyke rocks, the reader is referred to Bulakh & Ivanikov (1984, 1996), Bell et al. (1996) and Ivanikov et al. (1998).
|
Major and trace element data and Sr–Nd isotope analyses for a wide variety of samples from the Turiy Massif are presented in this paper to characterize the magma sources and further elucidate the petrogenetic relationships between the different rock-types. Recent Ar–Ar dating of samples from silicate and carbonatite samples across the massif suggests that most of the massif was intruded over a relatively short time span at 378 Ma, although the northern complex may be slightly older (E. A. Dunworth et al., unpublished data, 1998). A detailed petrographic description of the samples chosen for major and trace element analysis is given in Appendix A. Detailed descriptions of additional samples used for isotopic analysis may be obtained from the authors on request. It should be noted that many of the samples examined from Turiy are texturally heterogeneous. Samples for major and trace element analysis were carefully chosen on the basis of hand-specimen textural homogeneity, a lack of visible or isotopically detectable alteration (including cathodoluminescence studies for carbonate-rich rocks), and consistency of mineral abundances and compositions.
|
ANALYTICAL TECHNIQUES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
All samples used for bulk-rock major and trace element and isotopic analysis were split using a hydraulic jack to obtain
40 g of clean, fresh material, which was pulverized with a steel Braun Chipmunk jaw-crusher and subsequently crushed to a fine powder in an agate shatter box. Duplicate analyses carried out during the course of this study indicate that contamination effects from these facilities are negligible (B. Cousens, personal communication, 1997). Major element data were obtained in the Department of Earth Sciences at the University of Ottawa. All samples were fused, and analyses were obtained using a Philips PW2499 sequential XRF spectrometer. Trace element data determined by inductively coupled plasma mass spectrometry (ICP-MS) were analysed using a Perkin–Elmer–SCIEX ELAN 6000 instrument by Activation Laboratories Ltd in Ancaster, Ontario. The precision of the major element concentrations is <0·1 wt %, and the precision of the trace element analyses by ICP-MS is between 5 and 10% for most elements. Accuracy was monitored for both major and trace elements by duplicate analyses of a variety of rock powder standards.
To obtain mineral separates, the samples were manually crushed using an agate pestle and mortar, then sieved and washed in ethanol and deionized water. The minerals were separated from the (-250 + 149) µm fraction using a Frantz isodynamic magnetic separator, hand-picked to ensure high purity, and rewashed in ethanol and deionized water. Apatite was leached in cold, dilute HCl for 10 min, whereas silicates were leached in warm, dilute HCl for up to 45 min. Samples were dissolved in Savillex beakers using concentrated HCl (calcite), concentrated HNO3 (apatite) or an HNO3–HF mixture (silicate minerals). Sample solutions with <3 mg sample were put through polypropylene columns containing 3 ml Bio-Rad AG50W-X8 cation exchange resin (200–400 mesh) to collect Rb, Sr and rare earth element (REE) fractions. Solutions containing >3 mg sample were loaded onto Teflon columns containing 13 ml resin. A set of polypropylene columns containing 0·8 ml HDEHP resin was used to separate Ca from the Sr fraction. Nd and Sm separation was carried out using HDEHP resin in Bio-Rad Vycor columns. All samples were analysed using a Finnigan-MAT 261 multi-collector solid-source mass spectrometer operated in the static mode at Carleton University. Analyses were corrected to natural fractionation values and normalized to the internationally accepted standard values for NBS 987 87Sr/86Sr and La Jolla 143Nd/144Nd. Average NBS 987 value was 0·710260 ± 12 (n = 22) and La Jolla value was 0·511848 ± 12 (n = 18) during the course of this study. Precision is estimated as ±0·003% for 87Sr/86Sr and ±0·004% for 143Nd/144Nd (Morisset, 1992). The accuracy is ±0·5% for element abundance determination by isotope dilution (Morisset, 1992; Simonetti, 1994). Laboratory blank values were 0·5 ng for each of the elements analysed.
For C–O stable isotope determination, 10–20 mg of calcite from mineral separates with grain size -250 +149 µm was used. Whole-rock samples from a dolomite–calcite carbonatite (C.DC), and a carbonatite dyke (S.W58) were also analysed. Calcite was reacted at 25°C with 100% H3PO4 (McCrea, 1950) and dolomite at 50°C. The CO2 produced was analysed using a triple-collector VG SIRA 12 mass spectrometer to determine the isotopic composition of the C and O. Routine precision (2
) of a pure carbonate analysis is 0·1
. Fractionation factors used were 1·01025 for calcite at 25°C (Friedman & O’Neil, 1977; Sharma & Clayton, 1965) and 1·01065 for dolomite at 50°C (Rosenbaum & Sheppard, 1986).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
Major element geochemistry
Silicate rocks
The melilitolites, whose modal mineralogy is dominated by melilite, phlogopite, magnetite, and nepheline, have lower SiO2 contents (35·5–37·5 wt %) than the ijolite series rocks, whose modal mineralogy is dominated by pyroxene, nepheline, garnet and tetraferriphlogopite (Table 1). On textural and compositional grounds, samples C.90.29 (pyroxenite) and TUR II (ultramelilitolite) are considered to be cumulates, containing 85% pyxoxene and 65% melilite, respectively. The Al2O3 contents of these two samples are lower (6·7 and 8·2 wt %, respectively) than those of the ijolites and other melilitolites (12·5–15·5 wt % Al2O3). Fe2O3(TOTAL) values range from 10 to 14·4 wt % for all samples, whereas TiO2 and P2O5 contents are in the range 2·0–3·1 wt % and 0·3–1·44 wt %, respectively. MgO concentrations are generally moderate (5·1–7·1 wt % MgO) although the pyroxenite contains 11·3 wt %; mg-number values range from 47 to 61. CaO contents range from 15·9 to 27·1 wt %, and are generally higher in the melilitolites than the ijolites. Na2O + K2O contents range from 1·03 to 11·34 wt % and Na2O/K2O is always >1. The turjite sample (TU 119) has an Na2O/K2O ratio of 2·8, very different from the ‘average turjite’ analysis of Kranck (1928)
with an Na2O/K2O ratio of 0·61.
|
Carbonatites
Three carbonatites analysed for major and trace element chemistry are from the central complex. Two of these carbonatites are calcite carbonatites rich in magnetite (C.49.40, C.TL.344); these, according to the chemical classification of Woolley & Kempe (1989), are ferrocarbonatites although they contain no ferrocarbonate phases. Using the new classification scheme of Gittins & Harmer (1997), they become borderline calciocarbonatite–ferruginous calciocarbonatites. Calcite carbonatite sample C.49.40, which is rich in phlogopite, is the only sample analysed with K2O/Na2O > 1 and has the highest P2O5 content of any of the samples analysed. The third central complex sample is a dolomite–calcite carbonatite [C.DC; defined as a magnesiocarbonatite using the classification scheme of Woolley & Kempe (1989)], containing approximately equal amounts of calcite and dolomite. The MnO content of 0·31 wt % in this sample is relatively low compared with most magnesiocarbonatites (Woolley & Kempe, 1989). The fourth sample (S.W58), a calciocarbonatite dyke from Turiy’s south coast, contains small quantities of apatite and iron sulphides and abundant dolomite-rich veins (see Appendix A). Its MnO content is the highest of all the carbonatites analysed (0·64 wt %).
As a result of the extreme textural heterogeneity of the phoscorite samples studied, none were selected for major and trace element analysis. The mineralogy of these samples is dominated by magnetite, apatite (commonly found as veins), tetraferriphlogopite and calcite. Pyroxene may also occur, whereas olivine has not been found in the samples examined in this study.
Radiogenic isotopes
Rb–Sr and Sm–Nd isotope analyses were obtained in an attempt to determine: (1) whether the individual mineral phases within each sample were in isotopic equilibrium at the time of crystallization; (2) the mantle (and crustal) sources involved in the generation of the magmas; (3) the degree of open-system behaviour during the evolution of the massif.
Samples used for this isotope study were mostly mineral separates. Calcite, apatite and phlogopite were analysed from several carbonatites and phoscorites, and apatite, amphibole, pyroxene, garnet, nepheline and perovskite were analysed from a variety of silicate samples. A few carefully chosen, texturally homogeneous whole-rock samples were also analysed for bulk-rock Sr and Nd isotopic compositions. In addition, data were also obtained from whole-rock powders and mineral separates from the nearby Terskii Coast kimberlite, olivine melilitite and olivine nephelinite pipes, from which similar samples were also described and analysed by Beard et al. (1998).
Isotopic components and relationships between Sr and Nd
Initial 87Sr/86Sr and 143Nd/144Nd ratios and Rb, Sr, Nd and Sm concentrations are given in Table 2. Figure 3a is a plot of Sr(T)–Nd(T) for all the Turiy samples, assuming an age of 378 Ma, based on the dominant Ar–Ar age of the massif (E. A. Dunworth et al., unpublished data, 1998). Most of the Turiy data fall in the upper left-hand quadrant, reflecting derivation from depleted mantle sources; none of the data points fall in the lower right-hand quadrant. Data from some rock-types (e.g. the phoscorites) are distinct and do not overlap with data from any other rock-types analysed in this study.
|
|
The first question to be addressed is whether coexisting minerals were in isotopic equilibrium at the time of crystallization. Our findings suggest that there is little evidence for significant isotopic disequilibrium within individual samples in the Rb–Sr system, except for that shown between phlogopite and calcite in the banded carbonatite sample C.54.x (Fig. 3b). The Sm–Nd data do show within-sample isotopic disequilibrium, as reflected by a lack of Sm–Nd isochrons. Similar isotopic disequilibrium in the Sr and Nd systems has been reported previously from the Jacupiranga complex (Huang et al., 1995
), the Gardiner Complex (Nielsen & Buchardt, 1985) and in extrusive nephelinites (Anthony et al., 1989; Simonetti & Bell, 1993). Andersen (1984) pointed out that true Sm–Nd isochrons are virtually unknown from carbonatites, presumably indicating either extreme sensitivity to post-magmatic alteration, or prolonged open-system behaviour during crystallization.
The phoscorite and carbonatite data from Turiy, together with the Terskii Coast kimberlite data (Table 2, Fig. 3b) and recently published data from other Devonian alkaline complexes (and kimberlites) within the Kola Alkaline Province (KAP) (Kramm & Kogarko, 1994; Zaitsev & Bell, 1995; Beard et al., 1996; Mahotkin et al., 1997, 2000) have been plotted in Fig. 3c. Kramm (1993), using the dataset from the KAP available at that time, proposed the existence of a two-source-component mixing line to explain the Sr–Nd isotope trend, which he named the ‘Kola Carbonatite Line’ (KCL). The larger dataset currently available, and plotted in Fig. 3c, has allowed us to more accurately define the position of this mixing line. Its proposed new position and slope are shown in Fig. 3c, based on the reasonable fit of most of the carbonatite, phoscorite and apatitite data from the Turiy, Kovdor, Khibina, Sokli and Ozernaya Varaka massifs, along with the Terskii Coast and Archangelsk kimberlite data (references as listed above). A small number of additional carbonatite analyses from Turiy (the northern complex samples), Khibina and Kovdor do not follow the trend defined by the majority of the data, and lie above and to the right of the new position of the KCL.
The isotopic data from the Turiy carbonatites, along with the phoscorite and Terskii kimberlite isotopic data, are plotted separately in Fig. 3b. The largest, and most isotopically varied suite of carbonatites, are those from the central complex, whose isotopic signatures lie between those of the phoscorites and Terskii Coast kimberlites on the KCL. Sample S.W58, a member of a suite of 14 calcite carbonatite dykes located on the south coast of the peninsula, plots closer to the enriched end of the KCL than any of the other Turiy carbonatites. The carbonatites from the northern complex include melilite-bearing carbonatites that grade into calcite-rich melilitolites. Data from these samples plot in the upper right-hand quadrant. They have similar Nd(T) values to the phoscorites, but higher Sr(T) values.
Most of the isotope data from the silicate rocks at Turiy lie to the right of the new KCL (Fig. 3a) and clearly have different isotopic signatures from the phoscorites and carbonatites. Data from the pyroxenites and some melilitolites fall closest to the phoscorite data, whereas data from the ijolite-series rocks fall closest to the bulk-Earth–CHUR intersection, near the data from the south coast carbonatite dyke, the Terskii Coast kimberlites and related pipes.
Stable isotope geochemistry
New 18OSMOW and 13CPDB stable isotope measurements have been obtained from a suite of calcite separates from a variety of carbonatites and phoscorites, one calcite melilitolite (N.64.18) and two whole-rock samples—one dolomite–calcite carbonatite (C.DC) and one fine-grained carbonatite dyke from the south coast of the Turiy peninsula (S.W58).
The results are listed in Table 3 and displayed in Fig. 4, which compares the stable carbon and oxygen isotope results obtained in this study with data from ‘primary igneous carbonatites’ (Hoefs, 1987), the Khibina massif (Zaitsev, 1996), the Kovdor massif (Zaitsev & Bell, 1995), the Sokli Massif (Whattham, 1998) and carbonatite dykes from Kandalaksha (Beard et al., 1996). In addition, fields of data from Alnö (Deines & Gold, 1973) and the Fen carbonatite complex (Andersen, 1984), which are also located within the Fennoscandian Shield, are shown, along with a field representing mantle CO2 as found in oceanic basalts (Deines, 1989) and the most recent data from the Oldoinyo Lengai carbonatite volcano in Tanzania (Keller & Hoefs, 1995). It can be seen that many of the samples analysed in this study have, in general, heavier 13CPDB ratios than the ‘primary’ (fresh, unaltered) mantle-derived carbonatites of Hoefs (1987) but similar 18OSMOW signatures. Fifty percent of all carbonatites have average values of 18OSMOW 6–9, and 91% of all carbonatites have 13CPDB -2 to -8 (Deines, 1989). However, samples with heavy 13CPDB values normally have similarly heavy 18OSMOW values—a feature that is not seen in the Kandalaksha Graben carbonatites. Dolomite–calcite fractionation factors are 
18O(dolomite - calcite) = +1 to -1 and 13C(dolomite - calcite) = 0 to +1 (Deines, 1970), which explains the relatively heavy 13CPDB value of the Turiy dolomite–calcite carbonatite sample (C.DC).
|
|
|
|
|
Trace element geochemistry
The trace element data are given in Table 1. Of particular interest and concern in the interpretation of the trace element data, is the ability to discern between magma source characteristics, fractionation or accumulation effects, and the results of magma mixing. In carbonatite systems where traditionally ‘incompatible’ elements frequently become compatible, and where field relationships show clear evidence of accumulation, layering and veining, this task is made significantly more difficult. However, here we briefly discuss REE patterns and some key trace element ratios, before using spidergrams to draw comparisons with other samples and rock-types world-wide.
Rare earth elements (REE)
The samples show varying degrees of light REE (LREE) enrichment (Figs 5 and 6). Their chondrite-normalized (La/Yb)CN ratios decrease in the order melilitolites (321–1100) > carbonatite dyke (213) > melilitite dyke (95) > central massif carbonatites (46–87) > (pyroxenite, olivine melteigite and ijolites) (24–36) > fenite (19) > turjite (9). Total REE contents decrease in the order carbonatite dyke (3893 ppm) > melilitolites (798–942 ppm) > (ijolites and turjite) (532–610 ppm) > melilitite dyke (465 ppm) > central complex carbonatites (249–342 ppm) > (pyroxenite, olivine melteigite and fenite) (171–184 ppm).
|
The central massif calcite carbonatites at Turiy have significantly lower REE contents than most carbonatites (Woolley & Kempe, 1989; Hornig-Kjarsgaard, 1998) and, as seen above, contents lower than most of the silicate rocks analysed from Turiy, although, interestingly, they have REE contents similar to those of the Terskii Coast kimberlites (Beard et al., 1998). Demaiffe et al. (1997) reported values of 255–886 ppm for calcite carbonatites and 120 ppm REE for magnesiocarbonatite from Kovdor.
Element ratios
Trace element ratios, which are particularly effective at discriminating between the different rock types or suites at Turiy, have been plotted in Fig. 6, where values have been mantle-normalized (MN) or chondrite-normalized (CN) as indicated. The plot of (P/Ce)MN vs (U/Ce)MN demonstrates the extremely low contents of U, with respect to LREE contents, of the central complex carbonatites compared with the south carbonatite dyke and the silicate samples. Most of the silicate samples show a curving trend from P depletion in the LREE-enriched melilitolites, to P enrichment in the ijolite and turjite. Samples that do not follow the main silicate trend include the central complex melteigite sample (C.IJ.122), which has an unusually strong P depletion for a given U/Ce ratio, and the pyroxenite sample, which is depleted in U compared with the other silicate samples. The points from the melilitite dyke, carbonatite dyke, the olivine melteigite and two of the melilitolites overlap near the centre of the plot, perhaps indicating a source composition.
Plotting (Y/Ho)MN vs (Nb/Sr)MN ratios demonstrates the extreme heavy REE (HREE) fractionation within the melilitolite samples (given that Y tends to show similar behaviour to the HREE). The low (Nb/Sr)MN ratios in the central complex carbonatites are due to strong depletion in Nb (3–11 ppm) compared with the south coast carbonatite dyke (S.W58) (391 ppm) and the silicate rocks (19–505 ppm).
Rudnick et al. (1993) suggested that Zr enrichment (with respect to Hf) within a mantle melt could reflect carbonate metasomatism within the mantle source, and that a decreasing Ti/Eu ratio within any mantle melt was also indicative of carbonate metasomatism, as a result of the preferential incorporation of Eu over Ti and Zr over Hf by carbonatite magmas (Green 1994; Veksler et al., 1998b). Figure 6 is generally consistent with this claim inasmuch as it shows a direct correlation between increasing (Zr/Hf)MN and decreasing (Ti/Eu)MN with increase in carbonate content of the Turiy samples.
The high field strength elements (HFSE) exhibit distinctive behaviour within and between rock-types at Turiy. The carbonatite dyke (S.W58) has (Nb > Ta)MN, which is consistent with derivation by liquid immiscibility (Veksler et al., 1998b), an origin also proposed by Ivanikov et al. (1998) for Turiy’s coastal carbonatite dykes, and typical of the ‘average calcite carbonatite’ of Woolley & Kempe (1989). However, the (Nb/Ta)MN ratios of the central massif calcite carbonatites are similar to those of the silicate rocks plotted in Fig. 6, which is inconsistent with derivation by typical silicate–carbonate immiscibility. Only the turjite and fenite share the strong (Nb/Ta)MN enrichment of the coastal carbonatite dyke. Of particular note is the enrichment of Zr and Hf in the turjite, pyroxenite and melilitite dyke (TU 107) with respect to Ti; the significant depletion of Zr and Hf in the melilitolites; ratios of (Zr/Hf)N close to unity in all silicate samples; and high Zr/Hf ratios in the carbonatites. The melilitolites and carbonatite sample C.49.40 are the only samples with both (Hf/Ti)MN and (Zr/Ti)MN < 1.
Spidergram comparisons
The simplest way of comparing and contrasting a range of trace element data from the Turiy rocks is to present them in the form of a series of mantle-normalized spidergrams in which the elements are plotted in order of increasing compatibility. The samples have been grouped on the basis of their trace element and isotopic similarities.
Carbonatites
The four carbonatites selected for bulk-rock analysis were chosen on the basis of their very different petrology, and hence major element chemistry characteristics. Similarly, their trace element signatures are also varied (Fig. 7a). In general, the carbonatite dyke (sample S.W58) exhibits the most enriched signature of all of the samples analysed and has an almost identical trace element signature to the ‘average calcite carbonatite’ of Woolley & Kempe (1989). The central massif dolomite–calcite carbonatite (C.DC) has a similar spidergram pattern, but at much lower element concentrations (except Sr and P), to the ‘average Mg-carbonatite’ of Woolley & Kempe (1989) (Fig. 7b), and also shows a similar pattern to S.W58, but at lower degrees of enrichment. All three central massif samples (C.49.40, C.TL.344 and C.DC) are generally much less enriched in trace elements than the carbonatite dyke (S.W58) (Fig. 7a). Of particular note are the relatively low abundances of Nb, Ta, Zr and Hf; enrichment in Sr and P relative to the REEs, and (Ta/Nb)N > 1 in all three central complex samples, which is unusual in most carbonatites, as discussed above.
|
Pyroxenite–melilitolite–melilitite
With the exception of Zr, Hf and Th, the pyroxenite (C.90.29), melilitolites, and a melilitite dyke from the south coast (TU 107) share similar trace element spidergram patterns, with the least compatible normalized ratios increasing in the order pyroxenite < melilitite dyke < melilitolites (Fig. 8a).
|
The reasons for the extreme variations in Zr–Hf behaviour in these samples are not entirely clear. Pyroxenes crystallizing from alkaline silicate melts at crustal pressures tend to show Zr–Hf–HREE compatibility (Dunworth et al., 2001
). Thus, the Zr–Hf enrichment and HREE depletion in pyroxenite sample C.90.29 is unlikely to be due to pyroxene accumulation. The depletion in Zr, Hf and HREE in the melilitolites reflects the partitioning behaviour of two of the most trace-element-rich minerals in these samples, perovskite and melilite (H. Mirnejad, A. Zanetti, personal communications, 1999), suggesting that Zr and Hf, and some HREE could have been removed from the melilitolites during open-system fractional crystallization into a residual or immiscible liquid (e.g. carbonatite or possibly phoscorite; Dunworth, 1997). The general HREE depletion seen in most samples is most likely to be a source effect.
Olivine melteigite–ijolite–turjite
Some similarity is also seen between the spidergram patterns of the olivine melteigite (C.AG.48A), ijolite (C.IJ.14) and turjite (TU 119) (Fig. 8b). Overall, the normalized spidergram patterns are relatively flat and exhibit no significant enrichment or depletion in any particular group of trace elements. The P depletion in sample C.AG.48A is presumably due to apatite fractionation. The spidergram pattern of the fenite is also plotted in Fig. 8b, and it can be seen that, apart from K and Rb, the pattern is remarkably similar to that of the turjite. The reasons for this will be discussed in a later section.
Other silicate-rock comparisons
An interesting problem, in terms of field relationships, petrology and geochemistry, is understanding the link between the two most common silicate rock-types at Turiy, the ijolites and the melilitolites. Figure 8c includes spidergrams for a central massif turjaite (C.5.118), melteigite (C.AG.122) and ijolite (C.IJ.14). The most incompatible elements (left-hand end of the spidergram) are very similar, in terms of both pattern and concentration, for all three samples. This is considered significant, given that these elements are so incompatible and mobile. The principal differences are, yet again, in Zr, Hf and the HREE. This is likely to be a fractionation effect, perhaps as a result of Zr, Hf and HREE retention by garnet in the ijolite and melteigite, and Zr–Hf–HREE rejection by all phases (but especially melilite and perovskite) in the melilitolite.
The type-locality turjaite (WKT) and turjite (TU 119) from Turiy are compared in Fig. 9a with the trace element compositions of the type-locality okaite from Oka and bergalite from Kaiserstuhl (Hornig, 1988). Le Maitre (1989) considered bergalite the hypabyssal equivalent of, and okaite the haüyne-bearing equivalent of turjaite. The spidergram patterns of the okaite and turjaite show some similarity, including the significant depletion in Zr, although the okaite does not share the extreme HREE depletion seen in the turjaite. The Kaiserstuhl bergalite shows similarities in LREE, MREE and P with the turjaite, and in the ratios of the LILE and the HREE contents with the turjite.
|
Figure 9b shows the close parallelism between the trace element signatures of the Turiy pyroxenite (C.90.29) and the pyroxenites from Jacupiranga (Huang et al., 1995
). A spidergram pattern from a hornblende pyroxenite sample from the Precambrian Cargill carbonatite complex in Canada (Sharpe, 1987) lies parallel to the Turiy and Jacupiranga data, but with a greater degree of trace element enrichment. |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
Mantle reservoirs
Isotopic data from carbonatites and their accompanying silicate rocks are useful in helping to unravel their relationships, and impinge on the problem of whether carbonatites are primary or whether they have been produced by igneous differentiation. In addition, the carbonatites themselves provide insights into the secular evolution of the sub-continental mantle. In the early 1980s it was noted that data from several young carbonatites (<120 Ma) and related silicate rocks from the East Africa Rift define a mixing line, just below, and to the left of the mantle array defined by many oceanic island basalts (Faure, 1986
), called the ‘East Africa Carbonatite Line’ (EACL) by Bell & Blenkinsop (1987). This linear array is very similar to the Lo–Nd array defined by Hart et al. (1986) based on the isotopic signatures of modern-day oceanic island basalts. The Devonian ‘Kola Carbonatite Line’ (KCL) defined by Kramm (1993), the position of which has been redefined in this study, has a much steeper slope than the EACL and thus cannot represent mixing between the same two end-members as those implied for the EACL (Bell & Blenkinsop, 1987).
Kramm (1993) suggested that one end-member of his original KCL was perhaps similar to an EM I component, as it lies within the enriched quadrant, and the other, in the depleted quadrant, similar to the source that generated the Canadian Grenville and Superior Province carbonatites (data from Bell & Blenkinsop, 1987). Stein & Hofmann (1994), in an examination of the isotopic characteristics of a suite of large igneous provinces emplaced since 3·5 Ga, showed a linear trend in Nd–time space, which they attributed to lower-mantle plume contributions from a reservoir termed PREMA (PREvalent MAntle). The top end of the KCL lies just above the PREMA evolution line of Stein & Hofmann (1994), and this fact, combined with the lower-mantle 3He isotopic signature detected in several of the Devonian alkaline massifs from Kola by Marty et al. (1998), suggests that the isotopically depleted end of the KCL represents a plume derived from lower-mantle material, initiated at either the core–mantle boundary (Brandon et al., 1998) or the 1600 km discontinuity (Kellogg et al., 1999), which impinged on the base of the Baltic Shield during the Devonian. Hilton et al. (1999), in a recent study of 3He-enriched Icelandic basalts, suggested that this lower-mantle reservoir, the so-called ‘5th component’ in mantle plumes, best fits the original definition of mantle component ‘FOZO-2’ [Focal Zone 2 of Hauri et al. (1994b)]. Thus, it is suggested that a Devonian equivalent of ‘FOZO-2’ represents our best estimate of the isotopically primitive component of the KCL.
The isotopically enriched end of the KCL, as now redefined in Fig. 3, is dominated by data from the Terskii Coast and Archangelsk basaltic kimberlites (this study; Mahotkin et al., 1997, 2000; Beard et al., 1998), and overlaps neither the position of the enriched end of the original KCL (Kramm, 1993), nor the position of the EM I in the Devonian, calculated according to the model of Dostal et al. (1998) (not shown). Beard et al. (1998) argued for derivation of the Terskii Coast (Group I) kimberlites from a garnet harzburgite source within the asthenosphere where carbonate phases are stable, although this model presumed that the kimberlites represented 100% melt, which was deemed unlikely by Mahotkin et al. (2000), who claimed the presence of melt-metasomatized lithospheric material within the Archangelsk basaltic kimberlites.
It was felt important to try to correlate the trace element geochemistry of the Turiy magmas with melts derived from well-characterized mantle sources, as was done with the isotope geochemistry above, although all the Turiy samples show varying signs of fractionation-dependent trace element partitioning, which partially obscures the trace element signature of their primary parental magma(s). Tentative comparisons have been made between some of the more mafic Turiy magmas and the trace element ratios of oceanic basalts derived from well-known modern mantle reservoirs (HIMU, EM I) (Weaver et al., 1987; Chauvel et al., 1992), Hawaiian basalts [which are believed to be primarily extracted from plumes derived from the lower mantle (Garcia et al., 1995; Hilton et al., 1999)], and data from one of the Terskii kimberlites (Beard et al., 1998). The results, plotted in Fig. 10, suggest that the trace element signatures of the isotopically depleted melilite-bearing Turiy samples are most closely allied with modern alkaline basalts derived from the HIMU reservoir. The most mafic member of the more isotopically enriched ijolite series, the olivine melteigite (C.AG.48A), shows some similarity to the trace element characteristics of the Hawaiian alkalic basalts. Both the melilitolites and the olivine melteigite share the strongly HREE-depleted Terskii kimberlite signature.
|
Phoscorites and carbonatites
The relationship between the carbonate and silicate rocks within the Turiy Massif remains unclear, as no silicate rock has yet been found within the intrusive massifs that can match the isotopically depleted signatures of the phoscorites, or the signatures of the most depleted central complex carbonatites. Thus, any model involving simple closed-system differentiation or immiscibility between the silicate and carbonatite magmas that formed the massif can be ruled out, at least on the basis of the data currently available.
Open-system immiscibility involving simultaneous crustal contamination has been successfully modelled by Ray (1998) in a process called AFCLI (assimilation–fractional crystallization–liquid immiscibility). There are insufficient data available at present to carry out detailed AFCLI calculations involving the Turiy rocks, but early results suggest that this process may provide a plausible explanation of the relationship between the isotope data from the central complex silicate rocks and the carbonatites and phoscorites.
It can be seen from Fig. 3c that Sr–Nd isotope data from some of the Kola carbonatites, and a number of micaceous kimberlites and associated rocks from Archangelsk, lie below the KCL. The points fall on a fairly linear array between the top of the KCL and a point in the lower left-hand quadrant at approximately Nd -7. Mahotkin et al. (2000) attributed the origin of the low-Nd Archangelsk kimberlite data to the assimilation of carbonate–phlogopite-metasomatized, low Rb/Sr, low Sm/Nd lithospheric material, and thus it is possible that this lithospheric material has also played a role in the formation of the sub-KCL carbonatite data.
The stable isotope signatures of the Turiy carbonatites and phoscorites (Fig. 4), and in particular, the origin of the heavy 13CPDB values (relative to their conjugate 18OSMOW values) warrant discussion, especially when such enrichment is common to many carbonatites within the Baltic Shield area, and the Kandalaksha Graben in particular. Several factors have been proposed to produce 13CPDB enrichment (with constant or lower conjugate 18OSMOW values in CO2 or carbonate phases), including re-equilibration with meteoric water or seawater (Andersen, 1984), oxidation of mantle diamond (Deines & Gold, 1973) or mixing of two (unspecified) mantle components (Zaitsev & Bell, 1995).
Figure 4 shows that the re-equilibration of magmatic carbonate with Devonian seawater (13CPDB +1·5 and 18OSMOW +4, Lohmann & Walker, 1989) would produce the stable isotope signatures shown by most Turiy data. However, evidence of hydrous alteration, particularly in the Turiy rocks is, on the whole, not seen, particularly in those carbonatites and phoscorites located within the central complex.
Deines & Gold (1973) predicted that oxidation of mantle diamond could produce magmatic carbonate phases that would be 2 heavier than the original diamond 13CPDB values. New isotope fractionation data (Rosenbaum, 1994) now allow a more quantitative examination of this model. The average 13CPDB value for mantle-derived diamonds, of all ages, brought to the surface by Russian kimberlites is -7·1 (Galimov, 1984), and 90% of the world’s diamonds have 13CPDB values of -5 to -7 (Deines, 1989). Calculated values of the stable isotopic composition of carbonate (e.g. calcite) produced from the oxidation of mantle diamond, using C–CO2 and CO2–calcite fractionation factors at 900°C, are shown in Table 3b These results show that the CO2 and calcite produced from diamond oxidation are more enriched in 13CPDB than typical mantle CO2 and the calcite produced from it, as predicted by Deines & Gold (1973). Although it is realized that 900°C is not representative of mantle conditions, these calculations provide a clue to the relationship between the possible values of primary mantle CO2 and oxidized diamond (or graphite). Rayleigh fractionation is also an important influence on the stable isotope compositions of slowly cooled, intrusive carbonatite bodies. The calcite produced by equilibrium oxidation and fractionation of mantle diamond (Table 3) is not sufficiently enriched in 13C to produce the 13CPDB values of -3·1 to -2·1 seen in Turiy, but Rayleigh fractionation of the initial carbonate produced by oxidation of mantle diamond or graphite would produce values similar to those found in the Turiy and Kovdor carbonatites. However, although in terms of mass balance it would also be unrealistic to suggest that all the carbonate present in 13CPDB-enriched carbonatites is produced from mantle diamond or graphite, the fact that diamond-bearing kimberlites at Terskii and Archangelsk, (Beard et al., 1998; Mahotkin et al., 2000) show Sr–Nd isotopic similarities to the Turiy carbonatites, as well as the similarity in trace element signatures between some of the Terskii kimberlites and central complex carbonatites, suggests that there may a significant link between the 13CPDB-enriched Kola carbonatites and a carbon-bearing mantle source that is being actively tapped within or beneath the Baltic Shield during this period of magmatism.
Enriched mantle or crustal contamination?
One of the most important questions about the genesis of the Turiy rocks, in terms of their Sr and Nd isotope ratios, is whether the spread of isotope data to the right of the KCL, as shown in Fig. 3, reflects the presence of additional mantle source components other than those proposed to explain the formation of the KCL, or whether it reflects subsequent lithospheric contamination of the ascending magmas. In particular, the data from many of Turiy’s silicate rocks lie to the right of the KCL, similar to carbonatite–silicate-rock relationships seen in the East Africa Rift (Rudnick et al., 1993) and the Deccan Province (Ray, 1998). As discussed above, AFCLI modelling (Ray, 1998) may provide an explanation of the relationship between the central complex carbonate-rich and silicate rocks at Turiy. However, there are a number of Turiy samples with high 87Sr/86Sr ratios, including both silicate rocks and carbonatites from the satellite complexes as well as some of the south coast dykes, and both the oxygen and strontium isotopic signatures of the northern carbonatite and melilitolite samples (N.63.4, N.64.18m and c) are significantly enriched (Figs 3 and 4). The fact that samples from the satellite complexes appear to be more affected than samples from similar rock-types within the central complex suggests that, if these enriched values are due to additional crustal contamination (either by wall-rock assimilation or hydrothermal reaction in situ with the charnockitic host rocks), and if the entire massif was initially sourced via a central plumbing system, then the additional satellite-complex contamination occurred once the plumbing systems of the different complexes had diverged.
To evaluate these possibilities, basic isotopic mixing calculations were carried out between a whole-rock central complex melilitolite sample (C.190.105), which has one of the most depleted isotopic ratios of all the melilitolite analyses and lies close to the KCL, and a variety of crustal end-members, to see if the isotopic signatures of the contaminated melilitolites from the satellite complexes and related carbonatites from the northern complex could be produced by crustal contamination. These calculations involved mixing sample C.190.105 with four different components: (1) the Turiy fenite (sample NC.153.82); (2) charnockite (Johansson et al., 1993); (3) average upper crust (AUC) (McCulloch & Chappell, 1982); (4) average Precambrian lower-crustal granulite (ALC) (DePaolo et al., 1982). The results are shown in Fig. 11. The percentage of ‘pure’ charnockite that is needed to contaminate the satellite complexes to produce the highest 87Sr/86Sr isotopic signatures from the silicate rocks from Turiy is 
20%, which, in terms of bulk mixing or even AFC(LI), seems unrealistic. The melilitolite–AUC line shows that smaller quantities of ‘average upper crust’ (10%) would also be sufficient to produce the most contaminated isotopic signatures of the northern complex carbonatites by bulk mixing, although the charnockites forming the upper crust in this area may not be considered typical of ‘average upper crust’. In addition, the melilitolite–fenite mixing line, which should provide the best estimate of the real contamination process at magma-emplacement levels, has a very different slope in Fig. 11 from either the AUC or charnockite mixing lines. However, this may be due to the composition of the Swedish charnockite chosen for the calculation, as the use of Zairean charnockites (de Mulder et al., 1986) produces a line close to, and parallel to the fenite line in Fig. 11 (not shown), although with higher percentage-mixing estimates. Given these unrealistically high percentage estimates for contamination, it is considered likely that trace-element-rich, syn-magmatic, hydrothermal fluids, rather than true crustal assimilation, have been responsible for much of the contamination, as supported by the extensive fenitization surrounding the massif. The position of the high 143Nd/144Nd northern complex carbonatites in Fig. 3b may be explained by similar hydrothermal contamination processes affecting a magma derived from a source similar to those that produced the most isotopically depleted central complex phoscorite magmas.
|
Fractionation modelling
The data collected in this study have enabled us to carry out major and trace element modelling to impose some constraints on the origin of the two type-locality rocks from Turiy, turjaite and turjite. As discussed above, the turjaites and other melilite-bearing rocks are associated with the pyroxenites, whereas the turjite is associated with the more silica-rich melteigite–ijolite suite of samples, in isotopic space. These associations have been taken into account in the modelling.
Two parent magmas (starting compositions) were chosen for each of the two sets of calculations. One of the parental compositions chosen for both calculations was the average composition of the olivine melanephelinite dykes located around the coast of the Turiy Peninsula: OMN-11 (mg-number 69) (Ivanikov et al., 1998). More mafic parental compositions represented by dykes associated with the Kaiserstuhl carbonatite in the Rhinegraben were also chosen for both calculations. An olivine melilitite sample (RH 1) from Mahlberg Castle, close to the Kaiserstuhl carbonatite (mg-number 72) (Wilson et al., 1995), was used as a low-SiO2 mafic starting composition for the turjaite calculation, and Kais-1, a more Si-rich nephelinite from Kaiserstuhl (Keller et al., 1990), was chosen as a more mafic parent melt for the turjite calculations. The geochemical and tectonic similarities between the Turiy Massif and associated dykes within the Kandalaksha Graben, and the Kaiserstuhl carbonatite complex and associated hypabyssal silicate rocks within and close to the Rhinegraben (Keller et al., 1990), justify comparison of the magmatism within the two rift systems.
Table 4 shows the results of the major element model calculations, which attempt to produce the composition of the central complex turjaite (C.190.105) from RH 1 and OMN-11, and the turjite dyke (TU 119) from Kais-1 and OMN-11. The whole-rock and mineral data from Turiy used in the modelling were chosen to be as representative as possible of the compositions and modal mineralogy of the suites of samples involved (Dunworth, 1997). Additional components include the olivine composition (Fo86) used in the turjaite calculations, taken from a melilite-bearing olivinite from Afrikanda (E. A. Dunworth, unpublished data, 1997) and the natrocarbonatite from Oldoinyo Lengai (Keller & Spettel, 1995). It can be seen that the fit of the ‘Result’ columns to the corresponding turjaite and turjite analyses is good, in all four sets of calculations. However, the fractionation pathways taken to produce these results are rather different, depending on the starting compositions used. For example, only 37% fractionation of the Rhinegraben melilitite RH 1 is needed to produce the turjaite, whereas 62% fractionation of the Turiy olivine melanephelinite is needed to produce the same result. The turjaite shows the need for less fractionation if sample OMN-11 is used as a parental composition rather than Kais-1. At present, an accurate assessment of the volume of cumulate material lying beneath the Turiy Massif is difficult to obtain, and it would be premature to speculate on the validity of the varying percentage-fractionation estimates. However, what is noticeable in all the models is the need for the addition of extra components, and, in particular, carbonatite components, to balance the calculations.
The turjite sample, TU 119, has chemical characteristics very different from those of any of the other rock types studied, and it is clearly not in textural or chemical equilibrium (Bell et al., 1996; Dunworth, 1997). On the basis of the textural and chemical features of its mineralogy, Bell et al. (1996) suggested that turjite might be a natural example of the experimental ijolite–carbonatite reaction documented by Verwoerd (1978). One of the most fascinating results seen here is the correlation between the use of natrocarbonatite in the experiments of Verwoerd (1978), and its successful inclusion in the calculations in Table 4 to produce both the turjite and the melilitolite. No natrocarbonatites have yet been found in the Turiy Massif, although Veksler et al. (1998a) reported natrocarbonatite melt inclusions in both the Gardiner (Greenland) and Kovdor (Kola) melilitolite–carbonatite intrusions, and, given the similarity between the Kovdor and Turiy melilitolites, there remains the possibility that such results could yet be obtained from the Turiy rocks. It is also possible that alkali-rich brines could have been involved in hydrothermal circulation during intrusion and contributed the requisite alkalis and trace elements, perhaps from alkali-rich fenitizing fluids derived from the carbonatites (Veksler & Keppler, 2000), or from hydrothermal leaching of the surrounding country rocks. The involvement of alkali-rich fluids such as these might explain the similarity in spidergram patterns between the fenite sample NC.153.82 and the turjite sample TU 119, given the ‘natrocarbonatite’ component in the turjite model, as described above. Hydrothermal circulation of alkali-rich fluids between the massif and surrounding country rocks is borne out by the extensive country-rock alkali-fenitization, and the simultaneous crustal contamination shown by the isotopic signatures of the silicate rocks in the satellite complexes (Fig. 11).
The results of trace element modelling, based on the results of Table 4 and shown in Fig. 12, have been obtained using the partition coefficients listed in Appendix B, where appropriate. It can be seen that the close fit of the major element data in Table 4 is not reproduced as effectively by the trace element spidergrams in Fig. 12. This is likely to be due to (1) the use of inappropriate partition coefficients, and/or (2) the fractionation of trace-element-rich minor phases found in the Turiy rocks, but not included in Table 4, such as celsian, baddelyite, pyrochlore and strontianite (Dunworth, 1997). Further work should allow greater refinement of the model.
|
SUMMARY MODEL |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
Evaluation of the isotopic signatures from the Turiy samples shows that they cannot have been generated from a single, isotopically homogeneous magma, or generated by discrete partial melting of an isotopically homogeneous source. Interaction among sources or melts with different signatures is an essential part of any model proposed for the origin and evolution of the Turiy complex. Isotopically distinct sources may all be located within the convecting mantle, but this would invoke the necessity for an extremely heterogeneous source area. Alternatively, as outlined above, and as shown by the xenolith content of the nearby Archangelsk kimberlites (Mahotkin et al., 2000
), there may have been some contribution from the lithosphere.
A preliminary model for the evolution of the Turiy complex is outlined in Fig. 13. The minimum number of components that are needed to explain the distribution of data points shown in Fig. 3 is four, two of which define the KCL. One is characterized by a relatively primitive isotopic signature and is believed to be representative of a lower-mantle plume, probably derived from the core–mantle boundary or 1600 km discontinuity (Kellogg et al., 1999). It forms the isotopically depleted end-member of the KCL, and contributes to both the carbonatite–phosocorite and silicate magmas; it could well be the Devonian equivalent of the modern-day FOZO-2 component of Hauri et al. (1994b). The second marks the lower end of the KCL, and is dominated by a component that has also contributed to basaltic kimberlites from the Terskii Coast and Archangelsk regions, within the Kandalaksha graben. The third component lies beneath the KCL and contributes to a few Turiy carbonatites, the Telyachi carbonatites (Beard et al., 1996), and the micaceous Archangelsk kimberlites and related carbonatites (Mahotkin et al., 2000). It may either be directly derived from or have entrained significant quantities of metasomatized lithospheric material (Mahotkin et al., 2000). Lastly, it is believed that crustal contamination from the surrounding Proterozoic charnockites, either by syn-magmatic hydrothermal circulation or AFCLI, or both, is responsible for most of the Turiy data points that lie above the KCL.
|
The sequence of Devonian magma intrusion within the Turiy Massif suggests that plume-generated silicate magmas fractionated within the crust, producing an underlying body of olivinite and pyroxenite, but also intruding pyroxenites closer to the surface. Saturation in, or reaction of the parental melts with carbonatite magmas appears to have taken place during intrusion, producing the Ca-enriched melilitolites and the turjite. Crustal contamination, probably dominated by selective hydrothermal dissolution processes within the surrounding charnockites, occurred during the formation of the satellite complexes.
The central complex carbonatites and phoscorites were predominantly derived from lower-mantle-plume–kimberlite source mixing, with the possible addition of a third, sub-KCL component to the dolomitic carbonatites. The central complex carbonatites and phoscorites cannot be attributed to closed-system magma differentiation or immiscibility from the silicate magmas within the massif, although both appear to share the plume end-member source as one of their source components.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
The Turiy Massif has evolved under partial open-system conditions, both at source, during magma intrusion, and during crystallization. The range of
Sr(T)–Nd(T) signatures indicates that at least four components have contributed to the Turiy magmatism, and that closed-system magmatic differentiation of a single parental magma cannot be responsible for producing the wide variation in rock types seen within the massif. Two of the mantle sources, including a lower-mantle plume, define the ends of the redefined Kola Carbonatite Line. The isotopic data from rocks from the outer complexes suggest that the satellite-complex magmas were subjected to crustal contamination, before or during magma emplacement.
The pyroxenite–melilitolite group of rocks share depleted isotopic signatures but enriched trace element signatures. The melilitolites have the steepest REE patterns (LREE-enriched, HREE-depleted) of all the samples analysed from the Turiy complex, and are unusually depleted in Zr and Hf. The olivine melteigite–ijolite–turjite group of rocks share more enriched isotopic signatures but only moderately enriched trace element signatures. These samples can also be related by fractional crystallization and reaction or mixing with carbonatite. The most mafic members of the ‘melilitolite’ and ‘ijolite’ sample groups share the strong HREE depletion seen in the Terskii kimberlite sample (Fig. 10) (Beard et al., 1998).
A combination of trace element data, mineral compositions, petrographic textures and major and trace element modelling suggest that the turjite dykes may be similar to the products produced by the reaction of an ijolite with a carbonatite, possibly including a small amount of natrocarbonatite or Na-rich hydrothermal fluid, in keeping with the experimental results of Verwoerd (1978).
The phoscorites exhibit the most depleted Sr–Nd isotopic signatures of any of the samples analysed from Turiy, and, like the central-complex carbonatites, show no isotopic overlap with any other rock type from the massif. Thus, they cannot be pure residual or conjugate liquids from any other samples analysed in this study.
The carbonatites at Turiy have been divided into three main groups, based on spatial distribution, isotopic, trace element, mineralogical and petrographic variations. The central complex carbonatites have depleted trace element signatures and moderately depleted isotopic signatures. They have low REE concentrations, similar to those of the Terskii Coast kimberlites (Beard et al., 1998), low Ce/Yb ratios and (Nb/Ta)N < 1. The northern complex carbonatites have enriched 87Sr/86Sr signatures and lie above the KCL; mixing calculations suggest that they could have been affected by crustal contamination. The coastal carbonatite dykes include sample S.W58, which exhibits an enriched Sr–Nd isotopic signature and a relatively enriched trace element pattern, similar to that of the ‘average calcite carbonatite’ of Woolley & Kempe (1989).
The stable isotope compositions of the carbonatites and phoscorites exhibit unusually high 13CPDB signatures, which may be explained by magma re-equilibration with Devonian seawater, or a contribution to the magmatic carbon budget from oxidized mantle carbon.
|
APPENDIX A: SAMPLES USED FOR MAJOR AND TRACE ELEMENT ANALYSIS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
|
|
APPENDIX B: TRACE ELEMENT PARTITION COEFFICIENTS AND MINERAL COMPOSITIONS USED IN FRACTIONATION CALCULATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
|
|
ACKNOWLEDGEMENTS |
|---|
The authors are especially grateful to Andrei Bulakh and Valeriy Ivanikov for providing samples, fieldwork assistance and many stimulating discussions during the course of this study, and to Andrei Arzamastsev for providing the Terskii Coast samples. We would also like to thank the many additional people from the University of St Petersburg who made our trips to Turiy, in 1991 and 1995, possible, productive and enjoyable. Brian Cousens, Ron Hartree, Ina de Jong, Jeff Rosenbaum and Tony Simonetti are thanked for technical and intellectual assistance. We would also like to thank Daniel Demaiffe, Don Francis, an anonymous reviewer and Else-Ragnhild Neumann for providing extremely constructive criticism on earlier presentations of this work. This work was partly funded by NSERC Grant A7813 to K. Bell.
|
FOOTNOTES |
|---|
*Corresponding author. E-mail: lizzyann{at}magma.ca
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGANALYTICAL TECHNIQUESRESULTSDISCUSSIONSUMMARY MODELCONCLUSIONSAPPENDIX A: SAMPLES USED...APPENDIX B: TRACE ELEMENT...REFERENCES |
|---|
Andersen, T. (1984). Secondary processes in carbonatites: petrology of ‘rodberg’ (hematite–calcite–dolomite carbonatite) in the Fen central complex, Telemark (South Norway). Lithos 17, 227–245.
Anthony, E. Y., Segalstad, T. V. & Neumann, E.-R. (1989). An unusual mantle source for nephelinites from the Oslo Rift, Norway. Geochimica et Cosmochimica Acta 53, 1067–1076.
Arzamastsev, A. A. (1994). Unique Palaeozoic Intrusions of the Kola Peninsula. Apatity: Russian Academy of Sciences, pp. 1–80.
Bailey, D. K. (1993). Carbonate magmas. Journal of the Geological Society, London 150, 637–651.
Beard, A. D., Downes, H., Vetrin, V., Kempton, P. D. & Maluski, H. (1996). Petrogenesis of Devonian lamprophyre and carbonatite minor intrusions, Kandalaksha Gulf (Kola Peninsula, Russia). Lithos 39, 93–119.
Beard, A. D., Downes, H., Hegner, E., Sablukov, S. M., Vetrin, V. R. & Balogh, K. (1998). Mineralogy and geochemistry of Devonian ultramafic minor intrusions of the southern Kola Peninsula, Russia: implications for the petrogenesis of kimberlites and melilitites. Contributions to Mineralogy and Petrology 130, 288–303.
Bell, K. (ed.) (1989). Carbonatites, Genesis and Evolution. London: Unwin Hyman, 618 pp.
Bell, K. & Blenkinsop, J. (1987). Nd and Sr isotopic compositions of East African carbonatites: implications for mantle heterogeneity. Geology 15, 99–102.
Bell, K. & Keller, J. (1995). Carbonatite volcanism: Oldoinyo Lengai and the petrogenesis of natrocarbonatites. IAVCEI Proceedings in Volcanology 4, 210 pp.
Bell, K., Dunworth, E. A., Bulakh, A. G. & Ivanikov, V. V. (1996). Alkalic rocks of the Turiy Massif, Kola Peninsula, including type-locality turjaite and turjite: a review. Canadian Mineralogist 34, 265–280.[Web of Science]
Bottinga, Y. (1969). Carbon isotope fractionation between graphite, diamond and carbon dioxide. Earth and Planetary Science Letters 5, 301–307.[Web of Science]
Brandon, A. D., Walker, R. J., Morgan, J. W., Norman, M. D. & Pritchard, H. M. (1998). Coupled 186Os and 187Os evidence for core–mantle interaction. Science 280, 1570–1573.
Brooker, R. A. (1998). The effect of CO2 saturation on immiscibility between silicate and carbonate liquids: an experimental study. Journal of Petrology 39, 1905–1915.
Bulakh, A. G. & Ivanikov, V. V. (1984). The Problems of Mineralogy and Petrology of Carbonatites. St Petersburg: Leningrad University Publishing House (in Russian).
Bulakh, A. G. & Ivanikov, V. V. (1996). Carbonatites of the Turiy Peninsula, Kola: role of magmatism and of metasomatism. Canadian Mineralogist 34, 403–409.[Web of Science]
Chauvel, C., Hofmann, A. W. & Vidal, P. (1992). HIMU-EM: the French Polynesian connection. Earth and Planetary Science Letters 110, 99–119.[Web of Science]
Deines, P. (1970). The carbon and oxygen isotopic composition of carbonates from the Oka carbonatite complex, Quebec, Canada. Geochimica et Cosmochimica Acta 44, 943–961.
Deines, P. (1989). Stable isotope variations in carbonatites. In: Bell, K. (ed.) Carbonatites, Genesis and Evolution. London: Unwin Hyman, pp. 301–359.
Deines, P. & Gold, D. P. (1973). The isotopic composition of carbonatite and kimberlite carbonates and their bearing on the isotopic composition of deep-seated carbon. Geochimica et Cosmochimica Acta 37, 1709–1733.
Demaiffe, D., Verhulst, A., Weis, D., Andre, L. & Nivin, V. (1997). Geochemical (major and trace elements) and Nd–Sr isotopic study of the Kovdor carbonatites (and related rocks) from the Kola Peninsula. GAC–MAC Annual Meeting, Ottawa, Ont., Abstracts Volume, A.37.
de Mulder, M., Hertogen, J., Deutsch, S. & Andre, L. (1986). The role of crustal contamination in the potassic suite of the Karisimbi volcano (Virunga, African rift valley). Chemical Geology 57, 117–136.
DePaolo, D. J., Manton, W. I., Grew, E. S. & Halpern, M. (1982). Sm–Nd, Rb–Sr and U–Th–Pb systematics of granulite facies rocks from Fyfe Hills, Enderby Land, Antarctica. Nature 298, 614–618.
Dostal, J., Cousens, B. & Dupuy, C. (1998). The incompatible element characteristics of an ancient subducted sedimentary component in ocean island basalts from French Polynesia. Journal of Petrology 39, 937–952.[Web of Science]
Dunworth, E. A. (1997). The Turiy Massif, Kola Peninsula, Russia: open-system disequilibrium. Ph.D. thesis, Carleton University, Ottawa, Ont., 488 pp.
Dunworth, E. A. & Bell, K. (1998). Melilitolites: a new scheme of classification. Canadian Mineralogist 36, 895–903.[Web of Science]
Dunworth, E. A., Neumann, E.-R. & Rosenbaum, J. (2001). The Skien Lavas, Oslo Rift: petrological disequilibrium and geochemical evolution. Contributions to Mineralogy and Petrology (in press).
Exley, R. A., Mattey, D. P., Clague, D. A. & Pillinger, C. T. (1986). Carbon isotope systematics of a mantle ‘hotspot’: a comparison of Loihi Seamount and MORB glasses. Earth and Planetary Science Letters 78, 189–199.[Web of Science]
Faure, G. (1986). Principles of Isotope Geology. New York: Wiley, 589 pp.
Friedman, I. & O’Neil, J. (1977). Compilation of stable isotope fractionation factors of geochemical interest. In: Friedman, I. & O’Neil, J. R. (eds) Data of Geochemistry, 6th edn. US Geological Survey, Professional Paper 440-KK.
Galimov, E. M. (1984). The relationship between formation conditions and variations in isotopic composition of diamonds. Geokhimiya 1984, 1091–1118.
Garcia, M. O., Foss, D. J. P., West, H. B. & Mahoney, J. J. (1995). Geochemical and isotopic evolution of Loihi volcano, Hawaii. Journal of Petrology 36, 1647–1674.
Gerasimovsky, V. K., Volkov, V. P., Kogarko, L. N. & Polyakov, A. I. (1974). Kola Peninsula. In: Sorensen, J. (ed.) Alkaline Rocks. New York: Wiley, pp. 206–221.
Gittins, J. & Harmer, R. E. (1997). What is ferrocarbonatite? A revised classification. Journal of African Earth Sciences 25, 159–168.
Green, T. H. (1994). Experimental studies of trace-element partitioning applicable to igneous petrogenesis—Sedona 16 years later. Chemical Geology 117, 1–36.[Web of Science]
Harmer, R. E. & Gittins, J. (1998). The case for primary, mantle-derived carbonatite magma. Journal of Petrology 39, 1895–1903.
Hart, S. R., Gerlach, D. C. & White, W. M. (1986). A possible new Sr–Nd–Pb mantle array and consequences for mantle mixing. Geochimica et Cosmochimica Acta 50, 1551–1557.
Hauri, E. H., Wagner, T. P. & Grove, T. L. (1994a). Experimental and natural partitioning of Th, U, Pb and other trace elements between garnet, clinopyroxene and basaltic melts. Chemical Geology 117, 149–166.[Web of Science]
Hauri, E. H., Whitehead, J. A. & Hart, S. R. (1994b). Fluid dynamic and geochemical aspects of entrainment in mantle plumes. Journal of Geophysical Research 99, 24275–24300.[Web of Science]
Hilton, D. R., Grönvold, K., Macpherson, C. G. & Castillo, P. R. (1999). Extreme 3He/4He ratios in northwest Iceland: constraining the common component in mantle plumes. Earth and Planetary Science Letters 173, 53–60.[Web of Science]
Hoefs, J. (1987). Stable Isotope Geochemistry. Berlin: Springer, 241 pp.
Hornig, I. (1988). Spurenelementuntersuchungen an Karbonatiten mit hilfe der ICP-atomenissionsspektroskopie. Ph.D. dissertation, University of Freiburg.
Hornig-Kjarsgaard, I. (1998). Rare earth elements in sövitic carbonatites and their mineral phases. Journal of Petrology 39, 2105–2121.
Huang, Y.-M., Hawkesworth, C. J., van Calsteren, P. & McDermott, F. (1995). Geochemical characteristics and origin of the Jacupiranga carbonatites, Brazil. Chemical Geology 119, 79–99.
Ivanikov, V. V., Rukhlov, A. S. & Bell, K. (1998). Magmatic evolution of the melilitite–carbonatite–nephelinite dyke series of the Turiy Peninsula (Kandalaksha Bay, White Sea, Russia). Journal of Petrology 39, 2043–2059.
Johansson, Å., Meier, M., Oberli, F. & Wikman, H. (1993). The early evolution of the Southwest Swedish Gneiss Province: geochronological and isotopic evidence from southernmost Sweden. Precambrian Research 64, 361–388.
Keller, J. & Hoefs, J. (1995). Stable isotope characteristics of recent natrocarbonatites from Oldoinyo Lengai. In: Bell, K. & Keller J. (eds) Carbonatite Volcanism: Oldoinyo Lengai and the Petrogenesis of Natrocarbonatites, Vol. 4. Berlin: Springer, pp. 113–123.
Keller, J. & Spettel, B. (1995). The trace element composition and petrogenesis of natrocarbonatites. In: Bell, K. & Keller, J. (eds) Carbonatite Volcanism: Oldoinyo Lengai and the Petrogenesis of Natrocarbonatites, Vol. 4. Berlin: Springer, pp. 70–86.
Keller, J., Brey, G., Lorenz, V. & Sachs, P. (1990). IAVCEI 1990 Pre-Conference Excursion 2A: Volcanism and Petrology of the Upper Rhinegraben (Urach–Hegau–Kaiserstuhl). Mainz: IAVCEI, 60 pp.
Kellogg, L. H., Hager, B. H. & van der Hilst, R. D. (1999). Compositional stratification in the deep mantle. Science 283, 1881–1884.
Kjarsgaard, B. (1997). Carbonatites in context: differentiation trends of carbonatite alkaline ultrabasic silicate magmas based on field and experimental studies. GAC–MAC Annual Meeting, Ottawa, Ont., Abstract Volume, A.79.
Kogarko, L. N. (1987). Alkaline rocks of the eastern part of the Baltic Shield (Kola Peninsula). In: Fitton, J. G. & Upton, B. G. J. (eds) Alkaline Igneous Rocks. Geological Society, London, Special Publication 30, 531–544.
Kogarko, L. N., Kononova, V. A., Orlova, M. P. & Woolley, A. R. (1995). Alkaline Rocks and Carbonatites of the World Part 2: Former USSR. London: Chapman & Hall.
Kramm, U. (1993). Mantle components of carbonatites from the Kola Alkali Province, Russia and Finland: a Nd–Sr study. European Journal of Mineralogy 5, 985–989.[Web of Science]
Kramm, U. & Kogarko, L. N. (1994). Nd and Sr isotope signatures of the Khibina and Lovozero agpaitic centres, Kola Alkaline Province, Russia. Lithos 32, 225–242.
Kramm, U., Kogarko, L. N., Kononova, V. A. & Vartiainen, H. (1993). The Kola Alkaline Province of the CIS and Finland: precise Rb–Sr ages define 380–360 Ma age range for all magmatism. Lithos 30, 33–44.
Kranck, E. H. (1928). On turjaite and the ijolite stem of Turja, Kola. Fenniae 51, 1–104.
Kretz, R. (1983). Symbols for rock-forming minerals. American Mineralogist 68, 277–279.[Abstract]
Kukharenko, A. A., Orlova, M. P., Bulakh, A. G., Bagdasarov, E. A., Rimskaya-Korsakova, O. M., Nefedov, E. I., Illinskiy, G. A., Sergeev, A. S. & Abakumova, N. B. (1965). The Caledonian Ultramafic Alkalic and Carbonatitic Complexes of the Kola Peninsula and Northern Karelia. Moscow: Nedra (in Russian).
Le Maitre, R. W. (ed.) (1989). A Classification of Igneous Rocks and Glossary of Terms. London: Blackwell Scientific, 193 pp.
Lohmann, K. C. & Walker, J. C. G. (1989). The 18O record of Phanerozoic abiotic marine calcite cements. Geophysics Research Letters 16, 319–322.
Mahotkin, I. L., Sablukov, S. M., Zhuravlev, D. Z., Zherdev, P. U., Thompson, R. N. & Gibson, S. A. (1997). Plume–lithosphere interaction as a geodynamic model of origin of the Archangelsk diamondiferous province. Transactions of the Russian Academy of Science 353, 1–5.
Mahotkin, I. L., Gibson, S. A., Thompson, R. N., Zhuravlev, D. Z. & Zherdev, P. U. (2000). Late Devonian diamondiferous kimberlite and alkaline picrite (proto-kimberlite?) magmatism in the Arkhangelsk region, NW Russia. Journal of Petrology 41, 201–227.
Marty, B., Tolstikhin, I., Kamensky, I. L., Nivin, V., Balaganskaya, E. & Zimmermann, J.-L. (1998). Plume-derived rare gases in 380 Ma carbonatites from the Kola region (Russia) and the argon isotopic composition in the deep mantle. Earth and Planetary Science Letters 164, 179–192.
McCrea, J. (1950). On the isotope chemistry of carbonates and a paleotemperature scale. Journal of Chemical Physics 18, 849–857.
McCulloch, M. T. & Chappell, B. W. (1982). Nd isotopic characteristics of S- and I-type granites. Earth and Planetary Science Letters 46, 323–334.
McDonough, W. F. & Sun, S.-S. (1995). The composition of the Earth. Chemical Geology 120, 223–253.[Web of Science]
Morisset, N. (1992). Stable isotope and radioisotope geochemistry of the Panda Hill carbonatite, Tanzania. M.Sc. dissertation, Carleton University, Ottawa, Ont., 91 pp.
Nielsen, R. L., Forsythe, L. M., Gallahan, W. E. & Fisk, M. R. (1994). Major- and trace-element magnetite–melt equilibria. Chemical Geology 117, 167–191.
Nielsen, T. F. D. & Buchardt, B. (1985). Sr–C–O isotopes in nephelinitic rocks and carbonatites, Gardiner Complex, Tertiary of East Greenland. Chemical Geology 53, 207–217.
Onuma, N., Ninomiya, S. & Nagasawa, H. (1981). Mineral/groundmass partition coefficients for nepheline, melilite, clinopyroxene and perovskite in melilite–nepheline basalt, Nyiragongo, Zaire. Geochemical Journal 15, 221–228.[Web of Science]
Rass, I. T. (1982). Rare-earth elements in the rock-forming minerals of melilitic rocks in alkaline–ultrabasic complexes. Geochimica et Cosmochimica Acta 46, 1477–1488.
Ray, J. S. (1998). Trace element and isotope evolution during concurrent assimilation, fractional crystallization, and liquid immiscibility of a carbonated silicate magma. Geochimica et Cosmochimica Acta 62, 3301–3306.
Ronenson, B. M., Afanas’vey, B. V. & Levin, V. Y. (1978). Turjaite parageneses of Tur’ya Peninsula. International Geology Review 23, 535–543.
Rosenbaum, J. & Sheppard, S. (1986). An isotopic study of siderites, dolomites and ankerites at high temperatures. Geochimica et Cosmochimica Acta 50, 1147–1150.
Rosenbaum, J. M. (1994). Stable isotope fractionation between carbon dioxide and calcite at 900°C. Geochimica et Cosmochimica Acta 58, 3747–3753.
Rudnick, R. L., McDonough, W. F. & Chappell, B. W. (1993). Carbonatite metasomatism in the northern Tanzanian mantle: petrographic and geochemical characteristics. Earth and Planetary Science Letters 114, 463–475.[Web of Science]
Sharma, T. & Clayton, R. (1965). Measurement of 18O/16O ratios of total oxygen of carbonates. Geochimica et Cosmochimica Acta 29, 1347–1353.
Sharpe, J. (1987). Geochemistry of the Cargill carbonatite complex, Kapuskasing, Ontario. M.Sc. dissertation, Carleton University, Ottawa, Ont.
Simonetti, A. (1994). Comparative Nd, Pb and Sr isotopic study of alkalic complexes from East Africa and India; implications for mantle source regions, melt dynamics and fluid processes. Ph.D. dissertation, Carleton University, Ottawa, Ont., 255 pp.
Simonetti, A. & Bell, K. (1993). Isotopic disequilibrium in clinopyroxenes from nephelinitic lavas, Napak volcano, eastern Uganda. Geology 21, 243–246.
Stein, M. & Hofmann, A. W. (1994). Mantle plumes and episodic crustal growth. Nature 372, 63–67.
Veksler, I. V. & Keppler, H. (2000). Partitioning of Mg, Ca and Na between carbonatite melt and hydrous fluid at 0·1–0·2 GPa. Contributions to Mineralogy and Petrology 138, 27–34.
Veksler, I. V., Nielsen, T. F. D. & Sokolov, S. V. (1998a). Mineralogy of crystallized melt inclusions from Gardiner and Kovdor ultramafic alkaline complexes: implications for carbonatite genesis. Journal of Petrology 39, 2015–2031.
Veksler, I. V., Petibon, C., Jenner, G. A., Dorfman, A. M. & Dingwell, D. B. (1998b). Trace element partitioning in immiscible silicate–carbonate liquid systems: an initial experimental study using a centrifuge autoclave. Journal of Petrology 39, 2095–2104.
Verwoerd, V. I. (1978). Liquid immiscibility and the carbonatite–ijolite relationship: preliminary data on the join NaFe3+Si2O6–CaCO3 and related compositions. Carnegie Institution of Washington Yearbook 77, 767–774.
Weaver, B. L. & Tarney, J. (1984). Empirical approach to estimating the composition of the continental crust. Nature 310, 575–577.
Weaver, B. L., Wood, D. A., Tarney, J. & Joron, J. L. (1987). Geochemistry of ocean island basalts from the South Atlantic: Ascension, Bouvet, St. Helena, Gough and Tristan da Cunha. In: Fitton, J. G. & Upton, B. G. J. (eds) Alkaline Igneous Rocks. Geological Society, London, Special Publication 30, 253–267.
Whattham, S. A. (1998). A comparison of carbon and oxygen composition of the Sokli carbonatite massif, northern Finland, and the Siilinjarvi carbonatite complex, southern Finland, with carbonatite data from selected parts of the Kola Alkaline Province. B.Sc. thesis, Carleton University, Ottawa, Ont., 48 pp.
Wilson, M. & Lyashkevich, Z. M. (1996). Magmatism and the geodynamics of rifting of the Pripyat–Dnieper–Donets rift, East European platform. Tectonophysics 268, 65–81.
Wilson, M., Rosenbaum, J. M. & Dunworth, E. A. (1995). Melilitites: partial melts of the thermal boundary layer? Contributions to Mineralogy and Petrology 119, 181–196.[Web of Science]
Woolley, A. R. & Kempe, D. R. (1989). Carbonatites: nomenclature, average chemical compositions, and element distribution. In: Bell, K. (ed.) Carbonatites, Genesis and Evolution. London: Unwin Hyman, pp. 1–14.
Zaitsev, A. & Bell, K. (1995). Sr and Nd isotope data of apatite, calcite and dolomite as indicators of source, and the relationships of phoscorites and carbonatites from the Kovdor massif, Kola peninsula, Russia. Contributions to Mineralogy and Petrology 121, 324–335.
Zaitsev, A. N. (1996). Rhombohedral carbonates from carbonatites of the Khibina Massif, Kola Peninsula, Russia. Canadian Mineralogist 34, 453–468.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. S. Ray Radiogenic Isotopic Ratio Variations in Carbonatites and Associated Alkaline Silicate Rocks: Role of Crustal Assimilation J. Petrology, October 1, 2009; 50(10): 1955 - 1971. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. W. Marks, J. Schilling, I. M. Coulson, T. Wenzel, and G. Markl The Alkaline-Peralkaline Tamazeght Complex, High Atlas Mountains, Morocco: Mineral Chemistry and Petrological Constraints for Derivation from a Compositionally Heterogeneous Mantle Source J. Petrology, June 1, 2008; 49(6): 1097 - 1131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. Mitchell CARBONATITES AND CARBONATITES AND CARBONATITES Can Mineral, December 1, 2005; 43(6): 2049 - 2068. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. MARKS, T. VENNEMANN, W. SIEBEL, and G. MARKL Quantification of Magmatic and Hydrothermal Processes in a Peralkaline Syenite-Alkali Granite Complex Based on Textures, Phase Equilibria, and Stable and Radiogenic Isotopes J. Petrology, July 1, 2003; 44(7): 1247 - 1280. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Dunworth, E. A. Dunworth, and K. Bell The Turiy Massif, Kola Peninsula, Russia: mineral chemistry of an ultramafic-alkaline-carbonatite intrusion Mineralogical Magazine, June 1, 2003; 67(3): 423 - 451. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. N. Zaitsev and A. R. Chakhmouradian CALCITE AMPHIBOLE CLINOPYROXENE ROCK FROM THE AFRIKANDA COMPLEX, KOLA PENINSULA, RUSSIA: MINERALOGY AND A POSSIBLE LINK TO CARBONATITES. II. OXYSALT MINERALS Can Mineral, February 1, 2002; 40(1): 103 - 120. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||