Journal of Petrology

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (33) Request Permissions Google Scholar Articles by MARKS, M. Articles by MARKL, G. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Journal of Petrology Volume 42 Number 10 Pages 1947-1969 2001
© Oxford University Press 2001

Fractionation and Assimilation Processes in the Alkaline Augite Syenite Unit of the Ilímaussaq Intrusion, South Greenland, as Deduced from Phase Equilibria

MICHAEL MARKS and GREGOR MARKL^,*

INSTITUT FÜR MINERALOGIE, PETROLOGIE UND GEOCHEMIE, EBERHARD-KARLS-UNIVERSITÄT, WILHELMSTRASSE 56, D-72074 TÜBINGEN, GERMANY

Received November 22, 1999; Revised typescript accepted March 23, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
The early augite syenite unit in the 1·13-Ga-old Ilímaussaq intrusive complex, South Greenland, consists of a magmatic assemblage of ternary alkali feldspar + fayalitic olivine + augite + titanomagnetite + apatite + baddeleyite ± nepheline ± quartz ± ilmenite ± zircon. Feldspar, nepheline and QUILF thermometry yield T = 1000–700°C, at P = 1 kbar, which is derived from fluid inclusion data from other parts of the complex. Ternary feldspar was the first major liquidus phase. It crystallized at temperatures between 950 and 1000°C from a homogeneous magma with a_SiO2 = 0·8 and f_O2 about 1·5–2 log units below the fayalite–magnetite–quartz (FMQ) buffer. Later, closed system fractionation produced nepheline-bearing assemblages with a_SiO2 = 0·4 and log f_O2 = FMQ – 3 to FMQ – 5. Assimilation of wall rocks produced local variations of melt composition. Four traverses through the unit were sampled parallel to the assumed direction of crystallization. They exhibit significant differences in their mineral assemblages and compositions. The chemical zoning and calculated intensive parameters of four sample suites reflect both closed system fractional crystallization and local assimilation of wall rocks.

KEY WORDS: alkaline magmatism; assimilation; fractionation; redox equilibria; QUILF

	INTRODUCTION

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Magmatic rocks commonly record a prolonged history and various stages of crystallization. In some rocks, the entire solidification interval between liquidus and solidus of a melt can be deduced from one thin section. The detailed mechanisms of solidification, for example, whether it took place in an open or closed system or involved physical separation of minerals, have been extensively discussed in the literature (e.g. McBirney & Noyes, 1979; Stern, 1979; Michael, 1984; Sparks & Huppert, 1984; Sparks et al., 1984; McBirney et al., 1985).

In addition to the solid phase and melt equilibria in the magma, assimilation of material from surrounding rocks may play a significant role in the evolution of a magma (DePaolo, 1981; Huppert & Sparks, 1985). In addition to assimilation of country rock melt, material may be added from hydrous fluids released by dehydration reactions in heated wall rocks (Patchett, 1980). Usually, radiogenic and stable isotopes are needed to decide if assimilation has occurred (e.g. Taylor, 1980; DePaolo, 1981; Wilson & Sørensen, 1994; Tegner et al., 1999). In the present study, however, we investigate how mineral assemblage and mineral composition of the augite syenite unit of the Ilímaussaq intrusive complex, South Greenland, record both fractional crystallization and assimilation processes.

Augite syenite is common in many of the alkaline intrusive complexes of the Proterozoic Gardar rift province of South Greenland (Larsen, 1976; Parsons, 1979, 1981; Stephenson & Upton, 1982; Jones, 1984). In the Kungnat, Nunarssuit and Klokken complexes, it is associated with quartz syenite or alkali granite. Both the augite syenite and these quartz-bearing rock types are interpreted to be derived from fractionation of a parental alkali basaltic magma, possibly with crustal assimilation (Larsen & Sørensen, 1987; Stevenson et al., 1997). The unusual association of augite syenite with both alkali granite and agpaitic (i.e. peralkaline rocks with characteristic Na–Zr silicates such as eudialyte or rinkite) nepheline syenites in the Ilímaussaq intrusion raises the question of the importance of fractionation vs assimilation processes in the evolution of the augite syenite because fractionation alone would produce either quartz-bearing or silica-undersaturated rocks, but not both. Hence, a combination of both fractionation and assimilation in various proportions can be expected to be responsible for the observed divergence. Sm–Nd data (Stevenson et al., 1997) indicate that both processes may have been involved in the petrogenesis of the various rock types at Ilímaussaq, but their relative importance for intensive crystallization parameters is unclear. The present study investigates the evolution of intensive parameters during fractional crystallization, the effects of assimilation upon these parameters and how both fractional crystallization and assimilation are reflected in the phase equilibria of a highly reduced, slightly silica-undersaturated syenitic magma.

	REGIONAL GEOLOGY

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
The Ilímaussaq Intrusive Complex is part of the Gardar Province in South Greenland (Emeleus & Upton, 1976), which is interpreted to be a failed rift (Upton & Emeleus, 1987). It consists of 10 major intrusions that range from alkali granite to nepheline syenite and (rare) carbonatite, along with hundreds of gabbroic to trachytic or phonolitic dykes. These rocks all intrude an early Proterozoic, mostly granitic basement that is about 1·9 Ga old (‘Julianehåb granite’, Kalsbeek & Taylor, 1985). Sandstones and basalts of early Gardar age (the ‘Eriksfjord formation’ of Poulsen, 1964) that rest unconformably on the basement are cut by the Ilímaussaq Intrusion and some other Gardar alkali intrusive rocks.

The Ilímaussaq Intrusion (Fig. 1) is located in the central part of the Gardar Province. It has been dated at 1·13 ± 0·05 Ga by a Sm–Nd mineral isochron (Paslick et al., 1993), and is therefore one of the youngest intrusive complexes of the Gardar Province. It crystallized from three magma batches that were emplaced at depths of 3–4 km (1 kbar, Konnerup-Madsen et al., 1981; Larsen & Sørensen, 1987) into Julianehåb granite in the south (lower structural level) and into Eriksfjord sandstones and basalts in the north (higher structural level, Larsen & Sørensen, 1987). The emplacement of the intrusion was controlled by earlier fault systems (Sørensen, 1966).

View larger version (84K): [in this window] [in a new window]   Fig. 1. Geological sketch map of the Ilímaussaq intrusion (after Ferguson, 1964), showing four sample traverses.

 

The Ilímaussaq Complex is unique in the Gardar Province because it contains both alkali granite and extremely silica-undersaturated, highly differentiated nepheline syenite. Internally, the intrusive complex consists of an early augite syenite sheath and a later series of different nepheline syenites, some of which show spectacular magmatic layering (Ferguson, 1964). In the north part of the Ilímaussaq Complex, which is interpreted to be the top of the intrusion, the augite syenite roof is pierced by alkali granite, which occupies horizons both above and below the augite syenite unit. Stevenson et al. (1997) argued that the alkali granite represents a separate intrusion, which might have been produced from augite syenite by assimilation of silica-rich crustal material. The alkali granite grades into quartz syenite, which is regarded as a hybrid rock (Ferguson, 1964).

The agpaitic nepheline syenites formed by in situ fractionation of a broadly phonolitic, later magma batch (Larsen & Sørensen, 1987). Nielsen & Steenfelt (1979) grouped the agpaitic nepheline syenites into roof cumulates (naujaite, foyaite), floor cumulates (kakortokites) and residual liquids (lujavrites, ‘sandwich horizon’). In contrast, Larsen & Sørensen (1987) argued that the real floor cumulates (which are possibly high-temperature hedenbergite–fayalite syenites) are not exposed, and that the kakortokites and lujavrites represent a later magma batch. Heat flow measurements (Saas et al., 1972) indicate that the agpaitic rocks do not extend to much greater depths. Large positive gravity and magnetic anomalies (Blundell, 1978; Forsberg & Rasmussen, 1978) that are centred on the Ilímaussaq area imply that the intrusion is underlain by dense magnetic rocks, which are interpreted to be the mafic cumulates corresponding to the felsic agpaitic rocks. Because, on the basis of their mafic mineral assemblages, augite syenite, alkali granite and the various types of nepheline syenite are progressively more fractionated, all of these rocks are interpreted to have been derived from one basaltic parent magma of benmoreitic composition that fractionated in a deep-seated magma chamber (Nielsen & Steenfelt, 1979; Larsen & Sørensen, 1987; Stevenson et al., 1997).

	FIELD OBSERVATIONS AND SAMPLE LOCALITIES

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
The augite syenite unit occurs both as a relatively thin sheath along the southern and southeastern contact of the intrusion (‘augite-syenite sheath’, AS) and as a ‘top sheet’ (TS), which forms a flat-lying unit of 100–150 m thickness at the uppermost part of the intrusion. The presence of basaltic xenoliths of the Eriksfjord formation up to several hundreds of square metres in extent in the TS supports map relations that suggest the TS represents the top of the intrusion (Ferguson, 1964). Crosscutting contacts with later nepheline syenites, and the presence of augite syenite xenoliths in almost all other lithologies suggest that the augite syenite is the oldest unit of the Ilímaussaq intrusive complex. One locality at the shoreline of the south side of Kangerdluarssuk consists of sandstone blocks of the Eriksfjord formation up to 30 m across, which are entrained as xenoliths within the augite syenite. Reaction of the augite syenite melt with these sandstones has produced rims of alkali granite up to 20 cm thick around the xenoliths (Ferguson, 1964, fig. 7). Alkali granite also forms small veins that cut the augite syenite adjacent to this locality. These features provide field evidence that local host rock assimilation occurred during crystallization of the augite syenite unit. The augite syenite was reported to show a fine-grained ‘chilled’ border facies (Ferguson, 1964; Larsen, 1976) at the margins to the host rocks, and to have crystallized inwards from the walls towards the interior of the intrusion.

View larger version (24K): [in this window] [in a new window]   Fig. 7. Minor element variations in clinopyroxene from sample suite AS1. , analyses of Na-rich pyroxenes from sample GM1223; •, analyses from all other samples of this suite.

 
For the present study, four sample suites were collected. Three of these extend from the outer contact of the AS towards the centre of the intrusion, and one transects the TS. We tried to collect samples approximately parallel to the assumed direction of crystallization. Suite AS1 (samples GM1223–29, six samples) is from the western margin of the AS, between the two fjords of Kangerdluarssuk and Tunugliarfik. In this area, the augite syenite intrudes volcanic rocks of the Eriksfjord formation. In places, it forms centimetre- to decimetre-scale apophyses, which penetrate the basalts. The apparent thickness of the augite syenite exposed here is 650 m. Sample GM1223 was collected 2 m from the contact with the Eriksfjord volcanic rocks, sample GM1224 50 m farther toward the inner contact of the augite syenite sheath, and samples GM1226–28 another 150, 300 and 450 m farther inward. The outermost sample (GM1229) was collected 5 m from the interior contact with the nepheline syenite. In this transect, we observed no systematic variation of grain size.

Traverse AS2 (samples GM1267–69 and GM1381, four samples) was collected at the eastern margin of the intrusion where the augite syenite intrudes Julianehåb granite and has an apparent thickness of 200 m. Sample GM1381 was obtained 10 m inward from the contact with Julianehåb granite, GM1268 and GM1269 were successively collected 50 m farther inward, and the innermost sample, GM1267, was taken 30 m away from an agpaitic pegmatite, which forms the inner contact there. The grain size varies systematically from fine grained (GM1381 and GM1269) to medium (GM1268) and coarse grained (GM1267).

Suite AS3 (samples GM1330–1333^*, five samples) was sampled along the Kangerdluarssuk Fjord, at sea level, from the outer contact with Julianehåb granite to the inner contact with the agpaitic border pegmatite. The samples are relatively evenly spaced at 150 m intervals and span the exposed section. Sample GM1331 represents the fine-grained border facies of Ferguson (1964). The remaining four samples increase in grain size from medium (GM1330 and GM1332) to coarse grained (GM1333 and GM1333^*) toward the inner contact. Where the samples were taken, banding or layering is absent, although some hundred metres toward the east, Ferguson (1964) described magmatic banding in the augite syenite. This banding strikes parallel to the wall of the intrusion and dips steeply inwards. Similar banding is seen in other augite syenite bodies within the Gardar Province, most prominently in those of the Klokken Complex (Parsons, 1979; Parsons & Butterfield, 1981).

Sample suite TS (samples GM1359–1366, six samples) was obtained from the top sheet. Sample GM1359 was collected 1 m from the contact with the alkali granite, and the other samples were collected 20–30 m from each other and through the exposed section, between the alkali granite contact and a large basaltic xenolith of the Eriksfjord formation (Fig. 1). Sample GM1366 was collected 2 m from this xenolith. Samples from this suite are medium grained.

	PETROGRAPHY

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
All of the samples show xenomorphic textures and vary from coarse to fine grained. The major mineral phases are alkali feldspar (up to 50–70 vol. %, now with perthitic exsolutions), which represents an early liquidus phase, followed by olivine, clinopyroxene and in most samples titanomagnetite alone, or in a few samples by ilmenite and titanomagnetite (Table 1). Minor minerals are nepheline, quartz, apatite, baddeleyite, zircon, pyrrhotite, pyrite, chalcopyrite and monazite. Principally, four important assemblages can be distinguished, which point to decreasing silica activities:

View this table: [in this window] [in a new window]   Table 1: Mineral assemblages of augite syenite samples

 
fayalite–augite–magnetite–quartz;

fayalite–augite–ilmenite ± magnetite;

fayalite–augite–magnetite;

fayalite–augite–magnetite–nepheline.

Most feldspar is cryptoperthitic, and shows patchy perthitic exsolution textures (Fig. 2a). This texture implies that the grains interacted with a later fluid, possibly at T <500°C (Parsons & Brown, 1988). The exsolution of feldspar may be related to the formation of biotite–amphibole fringes around the mafic phases. Homogeneous and unexsolved feldspar is preserved as cores in some large feldspar grains, but this texture is restricted to the outermost samples of each suite we collected (see also Larsen, 1981).

View larger version (172K): [in this window] [in a new window]   Fig. 2. Photomicrographs (plane-polarized light) and BSE images of mineral textures from the Ilímaussaq augite syenite. Sample numbers appear in the upper right corner. Ab, albite; Or, orthoclase; Alk, alkali feldspar; Bd, baddeleyite; Ap, apatite; Ol, olivine; Cpx, clinopyroxene; Mag, magnetite; Zrn, zircon; Amp, amphibole; Ilm, ilmenite. (a) BSE image of patchy, perthitically exsolved alkali feldspar. (b) Most samples contain small grains of baddeleyite (ZrO2) as inclusions in olivine, clinopyroxene, Fe–Ti oxide and alkali feldspar grains (BSE image). (c) A typical cluster of Fe–Ti oxide minerals, clinopyroxene and olivine in a matrix of alkali feldspar, apatite and interstitial nepheline grains. The boundaries of the apatite grain are outlined to increase its visibility. (d) Zircon is present only as an interstitial phase. (e) Clinopyroxene overgrowth upon olivine. (f) Ilmenite lamellae in a clinopyroxene grain that is surrounded by a thin rim of amphibole.

 
Interstitial nepheline is found in some samples. In sample suite AS1, the distribution of quartz and nepheline is irregular: the two innermost samples contain nepheline, one sample contains quartz and the rest of the suite lacks nepheline and quartz as does the whole suite AS2. In suite AS3, modal abundance of nepheline increases systematically from the nepheline-free outer marginal sample GM1331, towards the inner sample GM1333^*, which contains 5 vol. % nepheline. The TS suite comprises samples with analcime–hydrogrossular–albite–muscovite–sodalite assemblages that pseudomorph euhedral, hexagonal crystals. We interpret these pseudomorphs to have been former nepheline that was altered by reaction with late-magmatic, possibly agpaitic fluids. The other samples of this suite lack both quartz and nepheline.

Fe–Ti oxide grains commonly occur in clusters with clinopyroxene and/or olivine (Fig. 2c). In most samples, former titanomagnetite has been strongly oxy-exsolved (i.e. exsolved and oxidized, presumably in the same process) to ilmenite and magnetite, and exhibits a wide variety of exsolution textures (Fig. 3). Using the terminology of Buddington & Lindsley (1964), trellis-type exsolution is the most common, whereas sandwich- or composite-type exsolutions are more rarely encountered. In sample GM1223, titanomagnetite lacks exsolution features, and two samples (GM1224 and 1361) contain ilmenite rather than titanomagnetite. Textures interpreted to be magmatic two-oxide assemblages of ilmenite and titanomagnetite were observed in samples GM1268 and GM1269 (Fig. 3f).

View larger version (144K): [in this window] [in a new window]   Fig. 3. BSE images of textures of Fe–Ti oxide minerals. In the terminology of Buddington & Lindsley (1964), (a)–(c) and (e) show trellis-type oxy-exsolution, whereas (d) shows sandwich-type oxy-exsolution and (f) displays a composite texture that probably represents a primary magmatic two-oxide assemblage. The numbers indicate the locations of the analyses reported in Table 4. Squares and arrow tips indicate broad-beam analyses and spot analyses, respectively.

 

View this table: [in this window] [in a new window]   Table 4: Microprobe analyses of exsolved magnetite and ilmenite, calculated bulk compositions, and analyses of unexsolved magnetite and ilmenite

 
Olivine appears as anhedral grains that are intergrown with clinopyroxene, feldspar and Fe–Ti oxide minerals (Fig. 2c). In samples GM1363 and GM1366, olivine contains heterogeneous exsolution lamellae that consist of intergrown clinopyroxene and magnetite (Markl et al., 2001). Clinopyroxene grains lack both exsolution lamellae and inclusions of orthopyroxene, but in some samples, they contain needle-shaped ilmenite crystals that closely resemble exsolution textures. These needles are oriented in two directions and are inhomogeneously distributed within crystals (Fig. 2f). The colour of clinopyroxene grains varies from grey to green, with the latter more common toward grain margins. In some samples, clinopyroxene has overgrown olivine (Fig. 2e), but in others, clinopyroxene and olivine appear to have co-crystallized.

Apatite needles are found as abundant inclusions in major minerals. Most samples contain small grains of baddeleyite (ZrO₂). Zircon is found in some samples from the AS1 traverse; sample GM1226 contains both minerals. Typically, baddeleyite is found as euhedral inclusions in olivine, clinopyroxene, Fe–Ti oxide (Fig. 3a) and alkali feldspar (Fig. 2b), whereas zircon is present interstitially (Fig. 2d). This difference in texture indicates that baddeleyite was an early liquidus phase, whereas zircon appeared late in the crystallization sequence.

Dark green or brown amphibole is found as a rare poikilitic phase that probably crystallized from a residual, relatively water-rich liquid, as has been described from the Klokken intrusion by Parsons & Butterfield (1981) and Parsons et al. (1991). In most samples, however, some olivine, Fe–Ti oxide and pyroxene grains are overgrown by small rims of subsolidus brown biotite and/or dark green amphibole. However, grains without such overgrowths are present in every sample, and only these were used to determine the magmatic mineral compositions that were used for the calculations below. Although fringes of biotite and/or amphibole around olivine and Fe–Ti oxide mineral grains, patchy perthite replacement of feldspar grains, pseudomorphic replacement of nepheline by analcime–hydrogrossular–muscovite–albite–sodalite assemblages, and oxidation–exsolution textures of Fe–Ti oxide mineral grains provide evidence for the interaction of the augite syenite with a late-stage water-rich vapour phase comparable with the Klokken complex (Parsons, 1980; Parsons et al., 1991), the stable fluid phase during most of the magmatic stages of the augite syenite was strongly methane dominated (Petersilie & Sørensen, 1970).

	MINERAL CHEMISTRY

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Analytical techniques
Minerals were analysed with the CAMECA SX100 electron microprobe at the Institut für Mineralogie, Petrologie und Geochemie at the University of Freiburg, Germany. Both natural and synthetic standards were used for calibration. Counting times were 20 s on the peak and 10 s on the background except for Zr, which was measured for 60 s on the peak. The emission current was 20 nA, and the acceleration voltage 15 kV. The raw data were corrected by the PAP method (Pouchou & Pichoir, 1984). The bulk compositions of exsolved magmatic feldspars and of very finely oxy-exsolved titanomagnetite grains were measured with a defocused electron beam of 10 µm size. The bulk compositions of coarsely oxy-exsolved titanomagnetite grains were reconstructed by combining image processing (NIH Image software) of back-scattered electron (BSE) images of the exsolved oxide mineral grains with point analyses of exsolved ilmenite and broad beam analyses of exsolved magnetite. The bulk compositions of the grains showing exsolution were recalculated from the area proportions of both exsolved phases, using molar volumes of 44·52 and 31·70 cm³/mol for magnetite and ilmenite, respectively. Bulk compositions were calculated for 3–5 different grains in each sample.

Olivine
In samples GM1223 and GM1361, primary olivine has been altered completely to ‘iddingsite’. In all other samples, olivine is unzoned with respect to Fe, Mg and Mn, and only the Ca content of some grains decreases slightly from core to rim. The Ca zoning is probably an effect of diffusive re-equilibration, rather than of primary chemical zoning. For all samples, olivine compositions range from Fa₇₁ to Fa₉₄ but individual samples rarely show a total variation of more than 3 mol % Fa (Table 2). The most Mg-rich compositions (Fa₇₁) are found in samples of the TS traverse. In suite AS3, the Fa content of olivine increases from Fa₇₉ in the outermost sample GM1331 to Fa₉₁ in the innermost sample GM1333^*. No such systematics are seen in the other suites. Similar to other Gardar intrusive complexes (Stephenson, 1974; Powell, 1978; Parsons, 1979; Jones, 1984; Upton et al., 1985), the MnO content of olivine ranges from 1·7 to 3·0 wt %, and is positively correlated with X_Fa. Olivine from samples GM1363 and GM1366 exhibits higher MnO contents than olivine of similar Fa content from the TS suite. CaO varies between 0·1 and 1·5 wt %, and is not correlated with other elements. These values agree with those of other silica-undersaturated suites from the Gardar Province (e.g. Stephenson, 1974) but they are distinctly greater than in silica-oversaturated suites (e.g. Stephenson & Upton, 1982).

View this table: [in this window] [in a new window]   Table 2: Representative microprobe analyses of olivine

 

Clinopyroxene
Clinopyroxene is the dominant mafic phase in all samples. It is a chemically zoned subcalcic augite (Fig. 4) with typically >90% quadrilateral (Di + Hed + En + Fs) components (Fig. 5; Table 3). Aegirine contents range from 2 to 15 mol % and reach 27 mol % only in sample GM1223 (Fig. 6). In general, X_Fe varies between 0·34 and 0·87, but in single samples, variations as a result of crystal zonation are between 6 and 22 mol % (Fig. 4). X_Fe increases and Ti decreases from core to rim, whereas Wo component is almost constant (Fig. 4). Figure 6 shows pyroxene analyses of the four different sample suites projected into the aegirine–diopside–hedenbergite triangle and compared with several published Gardar pyroxene trends (Stephenson, 1972; Larsen, 1976; Powell, 1978; Parsons, 1979). Our analyses from the TS suite are more Mg rich than the most Mg-rich pyroxenes of Larsen (1976). In most cases, Na exceeds the calculated Fe³⁺, which implies that small amounts of the jadeite molecule are present. Minor elements in pyroxene are Al₂O₃, MnO, TiO₂ and ZrO₂, which reach 2·5, 1, 1·5 and 1·5 wt %, respectively. Mn shows a positive and Al(IV) a negative correlation with Fe (Fig. 7). In most samples, only the outermost rims of pyroxene grains show measurable contents (generally up to 0·6 wt %; in sample GM1223 up to 1·5 wt %) ZrO₂. In sample GM1223, Zr increases with Na (Fig. 4).

View larger version (23K): [in this window] [in a new window]   Fig. 4. Representative zoning profiles through three clinopyroxene grains from various samples. Wo, wollastonite; En, enstatite; Fs, ferrosilite. Formulae are based on four cations and six oxygens; endmember projections were performed after the method of Lindsley (1983).

 

View larger version (24K): [in this window] [in a new window]   Fig. 5. Selected core compositions of pyroxene plotted in the pyroxene quadrilateral (Lindsley, 1983) together with coexisting olivine compositions. The slightly different slopes of the tie-lines within one sample suite should be noted. The Wo-poorest pyroxene compositions are shown, which are likely to reflect the most pristine magmatic conditions.

 

View this table: [in this window] [in a new window]   Table 3: Representative microprobe analyses of clinopyroxene

 

View larger version (14K): [in this window] [in a new window]   Fig. 6. Clinopyroxene composition trends in the four sample traverses plotted in terms of diopside (Di), hedenbergite (Hd) and aegirine (Aeg) components. On the left, generalized pyroxene trends from other Gardar complexes and for the whole Ilímaussaq suite are shown for comparison. Na-rich pyroxene was found only in sample GM1223 ( in suite AS1).

 

Fe–Ti oxide minerals
Primary Ti-rich magnetite has oxy-exsolved to ilmenite and magnetite in grains that show different types of exsolution textures from sample to sample (Fig. 3). Because the reintegrated compositions are among the most ulvøspinel-rich compositions known from any magmatic rock (D. H. Lindsley, personal communication, 2000), what follows is a detailed description of our interpretation of the exsolution textures. In any single sample, the exsolution textures are of the same type in every grain. The most common are trellis-type textures (Fig. 3a–c and e). These, and the less common sandwich-type textures (Fig. 3d) have elsewhere been interpreted to result from oxidation during exsolution (Buddington & Lindsley, 1964). In samples GM1268 and GM1269, however, composite ilmenite and magnetite grains (Fig. 3f) are interpreted to be of primary origin. Coexisting titanomagnetite grains show fine trellis-type exsolutions.

Table 4 contains analyses of exsolved ilmenite and magnetite and the calculated bulk composition for reintegrated grains, some of which are shown in Fig. 3. Analyses of primary, unexsolved ilmenite from sample GM1224 and of unexsolved titanomagnetite from sample GM1223 are also given in Table 4. Reintegration produced titanomagnetite compositions between Usp₆₇ and Usp₉₈. The small scatter in these results (Fig. 8) suggests that the calculation method is successful and does not produce erroneous or random results. The two-oxide assemblages in samples GM1267 and GM1268 show Usp_36–39Mt_64–61 coexisting with Ilm_93–94Hem_6–7. The unexsolved titanomagnetite in sample GM1223 has the composition Usp₄₀Mt₆₀, which is significantly less Ti rich than the reintegrated compositions. However, this sample also exhibits unusual clinopyroxene compositions (see above). This titanomagnetite composition probably reflects a significantly different melt composition compared with the other samples. In the two samples in which ilmenite is the only Fe–Ti oxide, it has compositions between Ilm₉₇Hem₃ (sample GM1224) and Ilm₉₉Hem₁ (sample GM1361).

View larger version (16K): [in this window] [in a new window]   Fig. 8. Compositions of Fe–Ti oxide minerals (mol %). , analyses of unexsolved titanomagnetite from samples GM1223, GM1268 and GM1269. Primary ilmenite () was found in samples GM1224, GM1268, GM1269 and GM1361. •, calculated magmatic compositions of exsolved titanomagnetite in all other samples as described in the text.

 

MgO contents of both ilmenite and magnetite are all <0·2 wt %, Al₂O₃ is below microprobe detection limits in ilmenite but reaches 2·5 wt % in magnetite. MnO reaches 3·5 wt % in ilmenite, and 2·1 wt % in magnetite.

Alkali feldspar
Broad-beam measurements of patchily exsolved perthite yielded ternary compositions between Ab₄₀An₆Or₅₄ and Ab₅₅An₁₂Or₃₃ in samples from traverses AS2, AS3 and TS (Fig. 9). In sample suite AS1, estimated bulk compositions are distinctly more Ca poor: these range between Ab₆₇An₅Or₂₈ and Ab₄₂An₃Or₅₅ (Fig. 9).

View larger version (19K): [in this window] [in a new window]   Fig. 9. (a). Feldspar compositions in terms of Ab, An and Or components (mol %) of sample suite AS3. White circles are spot analyses from sample GM1331 (fine-grained border facies). Black circles are spot analyses from all other samples of this suite. The shaded region indicates the field where defocused beam analyses from all samples plot. (b). Total range of reintegrated magmatic feldspar compositions for each sample suite as measured with a defocused electron beam plotted on the temperature-dependent feldspar solvus of Fuhrman & Lindsley (1988).

 

Nepheline
Compositions of interstitial nepheline grains range from Ne₆₇Ks₁₇Qz₁₆ to Ne₇₄Ks₁₁Qz₁₅. The composition of pseudomorphed nepheline crystals in samples from the TS traverse could not be reconstructed. Representative nepheline analyses are shown in Table 5.

View this table: [in this window] [in a new window]   Table 5: Representative microprobe analyses of nepheline

 

	MINERAL EQUILIBRIA

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Various phase equilibria constrain the crystallization conditions of the augite syenite. The variables to be constrained are temperature, silica activity and oxygen fugacity. Minimum liquidus temperatures were estimated based on the ternary feldspar compositions. Solidus temperatures, silica activity and oxygen fugacity were calculated from olivine–pyroxene–(Fe–Ti)-oxide equilibria using the QUILF program of Andersen et al. (1993). Additional estimates of temperature and silica activity were derived from activity-corrected clinopyroxene–nepheline–albite–SiO₂ phase equilibria assuming equilibration of this assemblage with the melt at the time of formation of the last phase in this assemblage. The assemblages used for these calculations are:

fayalite–augite–alkali feldspar–magnetite–quartz;

fayalite–augite–alkali feldspar–ilmenite–magnetite;

fayalite–augite–alkali feldspar–magnetite;

fayalite–augite–alkali feldspar–magnetite–nepheline;

fayalite–augite–alkali feldspar–magnetite–ilmenite–nepheline.

With QUILF, temperature was calculated based on the Fe–Mg exchange between clinopyroxene and olivine and in samples with two magmatic Fe–Ti oxide minerals also based on two-oxide thermometry. The displaced reaction

was then used to calculate a_SiO2 and f_O2.

The assemblage clinopyroxene–alkali feldspar–nepheline constrains temperature and a_SiO2 at fixed pressure by the equilibria

These reactions define a displaced invariant point in a T–a_SiO2 diagram. Activities for albite were calculated from the reintegrated feldspar compositions after the solution model of Fuhrman & Lindsley (1988). At temperatures between 500 and 950°C, a_Ab varies between 0·65 and 0·75. Calculated jadeite activities (after Holland, 1990) lie between 0·02 and 0·04. The solution model of Ghiorso (http://melts.geology.washington.edu) yields activities of nepheline of 0·2–0·3 in the temperature range between 750 and 900°C for all nepheline compositions. The value of 0·3 was chosen for all calculations; but the results for a_Ne = 0·2 are only slightly different. The values of G°(P, T) for the reactions were calculated using the GEOCALC program of Lieberman & Petrakakis (1990) and the Berman (1988) database.

No mineralogical constraints could be placed on the pressure of solidification. However, the fluid inclusion data of Konnerup-Madsen & Rose-Hansen (1982) indicate a crystallization pressure of 1 kbar for the alkali granite. This estimate is consistent with a reconstructed thickness of 3–4 km for the overlying Eriksfjord formation at the time of emplacement (Larsen, 1977). Therefore, all of our calculations were performed at a fixed pressure of 1 kbar.

Temperature
Figure 9b shows reintegrated feldspar compositions plotted on the temperature-dependent feldspar solvus of Fuhrman & Lindsley (1988). Estimated minimum liquidus temperatures are between 750° and 1000°C for all samples. A very similar temperature range (between 750 and 900°C) is also suggested by the nepheline compositions, based on the temperature-dependent NaAlSiO₄–KAlSiO₄–SiO₂ diagram of Hamilton (1964).

Because reaction kinetics between feldspar and silicate melt are slow relative to the assumed cooling rate, feldspar thermometry reveals the conditions during formation of a particular feldspar crystal, and therefore reflects the conditions of formation for feldspar of that composition. This may be at some temperature between the liquidus and the solidus. In contrast, equilibria involving Fe–Mg exchange between olivine, clinopyroxene and the melt are characterized by faster diffusion rates and, thus, record conditions approaching the solidus. These minerals probably record the lowest temperature at which liquid is in contact with the Fe–Mg silicate minerals. Post-magmatic Fe–Mg diffusion or fluid infiltration could alter olivine and clinopyroxene compositions. However, the lack of zoning in olivine and the geologically reasonable temperatures produced by Fe–Mg-silicate-based equilibria suggest that these processes were not of much importance.

The theoretical background for the QUILF calculations has been given by Frost & Lindsley (1992), Lindsley & Frost (1992) and Andersen et al. (1993). In our calculations, average olivine and the most Fe-rich clinopyroxene core compositions were used. These choices minimize the effects of later diffusive re-equilibration, during which clinopyroxene tends to become enriched in Mg (Markl et al., 1998). The larnite component of olivine and Mg components of ilmenite and titanomagnetite were adjusted by QUILF to an equilibrium value because of large scatter in the original data, signs of post-magmatic diffusive changes, and because these elements have the largest analytical uncertainty of those analysed in olivine and the Fe–Ti oxide minerals. X_La calculated by QUILF was fixed in all subsequent calculations. In each calculation, one of the other compositional input parameters (X_Fo, X_En, X_Wo) was also allowed to be adjusted by QUILF. The result was considered reliable if the calculated value was in the range of measured values within any single sample. The accepted range of calculated crystallization parameters is reported in Table 6. In samples GM1223 and GM1361, in which olivine was completely altered to iddingsite, the olivine equilibrium composition was calculated by QUILF.

View this table: [in this window] [in a new window]   Table 6: Input data and results of QUILF calculations

 

Only the calculated temperatures of samples that contain two oxides depend on oxide mineral compositions; in other samples, the calculated temperature depends on the compositions of olivine and coexisting clinopyroxene. The calculated solidus temperatures range from 680 to 960°C (Fig. 10); the uncertainty is estimated at about ±50°C. Temperature systematically decreases from the outer toward the inner contact in the AS1 suite (neglecting sample GM1223), and calculated temperatures are either relatively constant or vary in a non-systematic way in the AS2, AS3 and TS traverses.

View larger version (23K): [in this window] [in a new window]   Fig. 10. Solidus temperatures for each sample as calculated from QUILF. The variation shown for every sample is caused by the variation of different input parameters in the QUILF calculations and it corresponds to the uncertainty of the calculations (about ±50°C). The technique is described in the text and in Table 6. NQC, non-quadrilateral components. (See text for an explanation and discussion of the technique.)

 

Silica activity
Silica activities are based on a reference state of a pure quartz at P and T. They were calculated using both the QUILF program (Table 6) and phase equilibria that use the jadeite component in clinopyroxene, albite component in feldspar and nepheline component in nepheline–kalsilite solid solutions.

Table 6 shows the input data for and the results of the QUILF calculations. The results are also plotted in Fig. 11. On the basis of the comparison between the various methods and between the observed mineral assemblage and calculated silica activity (e.g. in the quartz-bearing sample), the uncertainty of a_SiO2 values is about ±0·05.

View larger version (21K): [in this window] [in a new window]   Fig. 11. Silica activities for each sample, calculated with QUILF. The variation shown for every sample is caused by the variation of the input parameters in the QUILF calculations and it corresponds to the uncertainty of the calculations (about ±0·05). The technique is described in the text and in Table 6. NQC, non-quadrilateral components.

 

The silica activities in suite AS1 vary in a non-systematic way. Sample GM1226 contains quartz (a_SiO2 = 1) and sample GM1223, which contains unexsolved titanomagnetite and Na-rich clinopyroxene, has the lowest calculated silica activity (a_SiO2 = 0.32) of all samples. The other samples of this suite have silica activities between 0·54 and 0·69 (Fig. 11, Table 6). Strictly, however, the QUILF program should not be applied to GM1223, because the pyroxene in this sample has >10% non-quadrilateral components. Sample suite AS3 yields a wide range of silica activities between 0·40 and 0·80. In suites AS2 and AS3, silica activity systematically decreases from the outer towards the inner contact. In suite AS2, a_SiO2 ranges between 0·88 and 0·60. In AS3, the decrease is even larger, from 0·80 to 0·40, and agrees with the increase of the modal content of nepheline towards the inner contact. The calculated a_SiO2 values for the samples from the TS suite range between 0·49 and 0·64, and only the ilmenite-bearing sample GM1361 has a higher a_SiO2 of 0·75.

As shown in Fig. 12, the QUILF and the nepheline–albite–clinopyroxene calculations agree well. Furthermore, in samples that show nepheline as a texturally early liquidus phase (TS suite), the QUILF results plot around the invariant point in Fig. 12. In samples that contain interstitial nepheline or no nepheline at all, most of the QUILF results plot in the clinopyroxene or feldspar fields.

View larger version (37K): [in this window] [in a new window]   Fig. 12. Temperature–aSiO2 plot showing phase equilibria among clinopyroxene, albite and nepheline at 1 kbar. (See text for the activity models and thermodynamic data used.) The small boxes represent the QUILF results for each sample. The good agreement between QUILF results, mineralogy and Ne–Ab–Jd equilibria should be noted.

 

Oxygen fugacity
The oxygen fugacity during crystallization was calculated from equilibria among Fe–Ti oxide and Fe–Mg silicate minerals with the QUILF program of Andersen et al. (1993) using the reintegrated (for oxy-exsolved) or measured (for unexsolved) Fe–Ti oxide mineral compositions. The calculations were carried out by combining the whole range of re-estimated compositions in every sample with the calculated crystallization temperatures. This procedure yielded a range of f_O2 values for each sample, which is reported in Table 6 and shown graphically in Fig. 13. In both, oxygen fugacity is reported as FMQ, i.e. f_O2 of the sample minus f_O2 of the endmember fayalite–magnetite–quartz buffer reaction at the same P and T conditions (all in log units).

View larger version (24K): [in this window] [in a new window]   Fig. 13. Oxygen fugacity in terms of FMQ calculated with QUILF from the whole range of measured or reintegrated Fe–Ti oxide compositions in each sample. The range of calculated oxygen fugacities is taken to represent the uncertainty of the calculations. The dashed lines represent the calculated stability limit of native iron. (See text for discussion.)

 

Most samples record f_O2 values between 1 and 4 log units below FMQ, with a general trend to more reducing conditions when the inner contact of the augite syenite unit is approached. Exceptions are the ‘two-oxide samples’ GM1268 and GM1269, which exhibit relatively high f_O2 values between FMQ – 0·2 and FMQ – 0·6, and the samples GM1333^* and GM1229, which have calculated oxygen fugacities down to FMQ – 6. The significance of the latter two samples will be discussed below.

In sample suite AS1, f_O2 decreases from FMQ – 1·6 in the outermost sample GM1223, to less than FMQ – 4 in the innermost sample GM1229. Although both mineral chemistry and silica activity behave irregularly and although the Fe–Ti oxide assemblages are different in the various samples, calculated oxygen fugacity forms a more or less continuous trend. The four samples of suite AS2 behave very irregularly, but suite AS3 again shows a continuous, regular decrease from 1·6 to 4·6 log units below FMQ towards the inner contact of the unit. A similar trend is also seen in the TS suite, where the two outermost samples indicate log f_O2 between FMQ – 1·2 and FMQ – 2·4, decreasing to FMQ – 2·5 and FMQ – 4·2 in the inner samples.

	DISCUSSION

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Validity of the calculations
We argue that our estimates of magmatic conditions are reliable for two main reasons:

1. there is close agreement between petrography and values of silica activity calculated by QUILF and by nepheline–feldspar–clinopyroxene equilibria in sample suite AS3. This agreement is unlikely to be chance.
   
2. If, as a test, the silica activity for the quartz-bearing sample is calculated from the mafic phase compositions alone, it amounts to 0·96, which is within error of the theoretical value of unity (i.e. Qtz saturated).
   

To test the plausibility of the calculated oxygen fugacities, we calculated at which oxygen fugacities native iron would be stable at the pressures, temperatures and SiO₂ activities we used for our calculations. Native iron is not present in any sample. Two reactions govern the stability of native iron in this system:

Again using the GEOCALC software of Lieberman & Petrakakis (1990) and the database of Berman (1988), activity-corrected phase equilibria among native iron, magnetite, SiO₂, fayalite and oxygen gas were calculated at 1 kbar. An example of such calculations at 800°C is shown in Fig. 14. Then, the distance in terms of log f_O2 between the native iron stability curve and the synthetic FMQ buffer was graphically estimated at the silica activities calculated for the various samples. This ‘Fe stability value’ is shown as dashed lines for each sample in Fig. 13. Most of the samples have a lower limit of ‘permitted’ oxygen fugacity of about FMQ – 3 to FMQ – 4, below which native iron would be stabilized. Of 21 samples, only GM1333^* and GM1229 plot entirely within the stability field of native iron. The discrepancy between calculation and observation could be explained in at least three ways:

View larger version (21K): [in this window] [in a new window]   Fig. 14. The dependence of reactions in the system Fe–Si–O on silica acivity and oxygen fugacity at 800°C and 1 kbar. Fayalite activity was fixed to a value of 0·8 in the calculations. This value is an approximation to the olivine compositions found in the Ilímaussaq augite syenites. Magnetite activity is contoured for values applicable to Ilímaussaq augite syenite samples.

 

1. native iron was once present in these samples, but has not been identified, possibly because of subsequent oxidation;
   
2. native iron did not form, even though it was thermodynamically stable;
   
3. our calculations are wrong because of problems with thermodynamic data, the reintegrated titanomagnetite compositions, the higher than normal non-quadrilateral components in the clinopyroxenes or later diffusive re-equilibration of the mineral assemblage.
   

Most probably, either possibility (1) or (2) is responsible for the observed discrepancy, because as explained above, we most carefully and critically reintegrated the former Fe–Ti oxide compositions, we used the whole range of reintegrated compositions within a particular sample and because our f_O2 estimates closely agree with calculations of Powell (1978) for similar phase assemblages from the nepheline syenitic Igaliko complex in the Gardar Province. Our values are also compatible with the general observation that methane was the stable fluid phase in equilibrium with the melt and the early magmatic mineral assemblage (Petersilie & Sørensen, 1970; Konnerup-Madsen et al., 1979; Konnerup-Madsen & Rose-Hansen, 1982).

Continuous trends in the sample traverses: the effects of fractional crystallization in a closed system
Some of our sample suites show trends that are consistent with the interpretations of Larsen (1976) that the augite syenite shell crystallized from the contact with the host rocks inward, and that it shows the effects of continuous fractional crystallization. This is particularly true for suite AS3, which we collected at approximately the same localities as the samples of Larsen (1976). The continuous decrease in silica activity and oxygen fugacity and the changes in mineral compositions are indicative of relatively undisturbed closed system fractional crystallization. Nepheline crystallization in the interstitial liquid is compatible with fractionation of alkali feldspar from an initially undersaturated liquid. A possible equilibrium liquid evolution path is shown in Fig. 15 (after Schairer, 1950). In this diagram, the whole-rock analysis of the chilled border facies (comparable with our sample GM1331) of Ferguson (1964) was plotted. In agreement with the textures and the calculated parameters, it plots below the feldspar join. After extensive feldspar crystallization, it reaches the nepheline–feldspar cotectic, to finally evolve towards the thermal minimum of alkali feldspar and nepheline at 1040°C (at 1 bar).

View larger version (26K): [in this window] [in a new window]   Fig. 15. Ne–Ks–Qtz diagram of Schairer (1950) showing a possible equilibrium fractionation path of the augite syenite chilled border (black circle, analysis from Ferguson, 1964) as a proxy for the augite syenite parental melt. The range of magmatic feldspar and nepheline compositions as measured in the present study in sample GM1331 from the chilled border is shown as black bars. Temperatures are in °C.

 

Frost et al. (1988) showed that a system closed to oxygen can evolve towards decreasing or increasing f_O2 values, but that a system open to oxygen exchange should evolve parallel to an appropriate buffer curve. The buffer curve is likely to be a displaced FMQ equilibrium in the fayalite + titanomagnetite-bearing augite syenite. Whereas in silica-saturated systems, the change of the Fe/Mg ratio of the silicates controls the oxygen fugacity, differentiation in silica-undersaturated systems involves both iron enrichment and decrease in a_SiO2. Decrease of a_SiO2, however, will also cause f_O2 to decrease as a result of their connection in the displaced FMQ equilibrium (Fig. 14). The low f_O2 values recorded by some samples may reflect an extreme variety of this process at very low and still decreasing silica activities. The continuous trends seen in the various traverses imply that the augite syenite crystallized under essentially closed system conditions after emplacement at upper-crustal levels.

Differences in phase compositions and mineral assemblages among the sample traverses
The four sample traverses, which were collected at different areas in the augite syenite unit, show similar textures, general mineralogy, and mineral compositions. However, in detail they reveal some significant differences:

1. the samples of suite AS1 contain either quartz or nepheline or neither of these minerals. Clinopyroxene in the two outermost samples is relatively rich in Na. Some samples from traverse AS1 contain baddeleyite, others only zircon and one contains both of these minerals. The two outermost samples show either unexsolved titanomagnetite or ilmenite whereas all other samples contain exsolved titanomagnetite.
   
2. Three of the four samples of suite AS2 contain two magmatic Fe–Ti oxide minerals, and only the innermost sample contains exsolved titanomagnetite alone. All of the samples contain baddeleyite and lack zircon, nepheline and quartz.
   
3. All but the outermost sample of suite AS3 contain interstitial nepheline. All samples contain baddeleyite and exsolved titanomagnetite, but quartz, zircon and magmatic ilmenite are lacking.
   
4. Suite TS displays the most Mg-rich olivine compositions of the four localities. The content of MnO in olivine and clinopyroxene in two samples is significantly higher than in the others. These two samples were collected close to a basaltic xenolith. Ilmenite is the only Fe–Ti oxide in one sample, whereas exsolved titanomagnetite is present in all other samples. Baddeleyite is the only Zr mineral.
   

Heterogeneities in the sample traverses: the effects of assimilation
Although we propose above that the augite syenite melt was closed to oxygen exchange during crystallization for most samples in suites AS1, AS3 and TS, and although the samples were collected perpendicular to the outer contact and parallel to the assumed direction of crystallization (Larsen, 1976), both mineral compositions and phase assemblages change in non-systematic ways in each sample suite. Obviously, the augite syenite melt was modified on a local scale during solidification and we invoke local changes in bulk composition and/or intensive parameters such as f_O2 or a_SiO2 to explain these inhomogeneities. Assimilation of host rocks is the most likely way for this to happen. Three types of host rocks could be contaminants: the Julianhåb granite, the Eriksfjord sandstone and the Eriksfjord basaltic to trachytic volcanic rocks. Interaction of augite syenite melt with sandstone can produce quartz-saturated melts (Ferguson, 1964), and the same may be true for assimilation with quartz-bearing Julianhåb granite. Assimilation of these materials would easily explain the scatter in SiO₂ activities in some traverses. The same effect (see, e.g. Juster et al., 1989) or the contribution of upper-crustal fluids may be responsible for drastic changes of f_O2, such as those seen in the AS2 suite or in some samples of the other suites. Assimilation of Na- and Zr-rich trachytic or phonolitic rocks from the Eriksfjord formation, which were reported by Watt (1966), could explain the unusual Na and Zr enrichment in pyroxenes and the low calculated SiO₂ activity of sample GM1223. Compositional changes caused by Na–Zr-rich fluids related to the later agpaitic rocks is unlikely, because the agpaites post-date the solidification of the augite syenite, and the textures of sample GM1223 show no sign of any later recrystallization. The variable MnO contents at equal X_Fe in olivine and clinopyroxene of the TS suite are probably related to the assimilation of basaltic Eriksfjord lavas. According to Ferguson (1964) and Watt (1966), these rocks have lower X_Fe but similar MnO contents compared with the augite syenite melt (X_Fe^basalts = 0·25–0·45, X_Fe^{augite syenite} = 0·75–0·85; MnO^basalts = MnO^{augite syenite} = 0·2 wt %) and assimilation with such a basalt would lower X_Fe in the crystallizing olivine without much effect on X_Mn. The field relations support host rock assimilation at the site of intrusion, because in most traverses the outermost sample or samples exhibit features that differ from the rest of the sample suite, and because the contaminants proposed for the various samples above are all found in the immediate vicinity. However, our results indicate that these changes owing to assimilation were very localized and the effects of fractional crystallization are still discernible.

	SUMMARY AND CONCLUSIONS

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
The augite syenite unit of the Ilímaussaq intrusion of South Greenland crystallized at a high crustal level (about 1 kbar) in a system closed to oxygen. Baddeleyite crystallized first, feldspar crystallization began at 1000°C and continued down to probably 750°C. Both feldspar–clinopyroxene–nepheline and olivine–augite equilibria constrain the crystallization interval to temperatures between 1000 and 700°C. The latter is probably the temperature at which silicate minerals equilibrated with the last residual melt. Crystallization progressed from the outer margin of the intrusion inwards. Fractional crystallization in the principally closed system decreased silica activity from about 0·8 to 0·4 and oxygen fugacity from about FMQ – 1·5 to FMQ – 4. Assimilation of blocks of host rocks (quartzite, granite, trachyte, basalt) more or less in situ during final emplacement led to significant, but only localized, effects that are recorded by mineral assemblages and compositions. In some places, quartz saturation was reached; in others, nepheline crystallized. In every traverse, the outermost samples (close to the host rocks or to a stoped block) differ in mineral compositions from the rest of the same sample suite, which is closer to the pluton interior.

Our results provide evidence that highly reduced, strongly silica-undersaturated alkaline melts can be produced by closed system fractional crystallization of augite syenitic melt. Our data are also consistent with the model of Larsen & Sørensen (1987), which relates the various intrusive events at Ilímaussaq to the emplacement of a highly evolved, continuously fractionating alkali-basaltic magma at depth.

Localized effects of assimilation can be recorded in great detail by plutonic rocks, although these effects may be visible in silica-undersaturated rocks only. In an undersaturated melt, assimilation of a quartz-bearing rock will have a large effect on silica activity. Changes in silica activity will then change both the oxygen fugacity and the compositions of the minerals present (see Fig. 14), whereas in silica-saturated rocks, these changes will not be large and therefore will not be recorded. Although we could not quantify the amount of assimilated material necessary to produce the changes of many features we observed, our study does reveal details of the physical process of intrusion: within a single intrusion (the augite syenite, before intrusion of the later magmas), compositional gradients on a scale of metres or tens of metres were not homogenized during solidification. These can be detected today in the mineral compositions of many rock-forming minerals. The preserved heterogeneities argue for a quiet, undisturbed period of crystallization and lack of convection in the augite syenite melt.

	ACKNOWLEDGEMENTS

 
H. Müller-Sigmund provided invaluable help during microprobe measurements. Comments of S. Sorensen, R. Frost, D. Lindsley, H. Nekvasil, D. Barker, I. Parsons, M. Westphal and an anonymous reviewer led to substantial improvement of an earlier version of the manuscript and are gratefully acknowledged, as is funding for this work by the Deutsche Forschungsgemeinschaft (grant Ma2135/1-1). This paper is Contribution to the Mineralogy of Ilímaussaq No. 112.

	FOOTNOTES

 
^*Corresponding author. E-mail: markl{at} uni-tuebingen.de

	REFERENCES

TOP ABSTRACT INTRODUCTION REGIONAL GEOLOGY FIELD OBSERVATIONS AND SAMPLE... PETROGRAPHY MINERAL CHEMISTRY MINERAL EQUILIBRIA DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Andersen, D. J., Lindsley, D. H. & Davidson, P. M. (1993). QUILF: a PASCAL program to assess equilibria among Fe–Mg–Mn–Ti oxides, pyroxenes, olivine, and quartz. Computers and Geosciences 19, 1333–1350.

Berman, R. (1988). Internally consistent thermodynamic data for minerals in the system Na₂O–K₂O–CaO–MgO–FeO–Fe₂O₃–Al₂O₃–SiO₂–TiO₂–H₂O–CO₂. Journal of Petrology 29, 445–522.[Abstract/Free Full Text]

Blundell, D. J. (1978). A gravity survey across the Gardar igneous province, SW Greenland. Journal of the Geological Society, London 135, 645–654.

Buddington, A. F. & Lindsley, D. H. (1964). Iron–titanium oxide minerals and synthetic equivalents. Journal of Petrology 5, 310–357.[Abstract/Free Full Text]

DePaolo, D. J. (1981). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallisation. Earth and Planetary Science Letters 53, 189–202.[Web of Science]

Emeleus, C. H. & Upton, B. G. J. (1976). The Gardar Period in South Greenland. In: Escher, A. & Watt, W. S. (eds) Geology of Greenland. Copenhagen: Grønlands Geologiske Undersøgelse, pp. 152–180.

Ferguson, J. (1964). Geology of the Ilímaussaq alkaline intrusion, south Greenland. Bulletin Grønlands Geologiske Undersøgelse 39, 82 pp.

Forsberg, R. & Rasmussen, K. L. (1978). Gravity and rock densities in the Ilímaussaq area, South Greenland. Rapport Grønlands Geologiske Undersøgelse 90, 81–84.

Frost, B. R. & Lindsley, D. H. (1992). Equilibria among Fe–Ti oxides, pyroxenes, olivine, and quartz: Part II. Application. American Mineralogist 77, 1004–1020.[Abstract]

Frost, B. R., Lindsley, D. H. & Andersen, D. J. (1988). Fe–Ti oxide–silicate equilibria: assemblages with fayalitic olivine. American Mineralogist 73, 727–740.[Abstract]

Fuhrman, M. L. & Lindsley, D. H. (1988). Ternary-feldspar modeling and thermometry. American Mineralogist 73, 201–205.[Abstract]

Hamilton, D. L. (1964). Nephelines as crystallisation temperature indicators. Journal of Geology 69, 321–329.

Holland, T. J. B. (1990). Activities of components in omphacitic solid solutions; an application of Landay theory of mixtures. Contributions to Mineralogy and Petrology 105, 446–453.

Huppert, H. E. & Sparks, R. S. J. (1985). Cooling and contamination of mafic and ultramafic magmas during ascent through the continental crust. Earth and Planetary Science Letters 74, 371–386.[Web of Science]

Jones, A. P. (1984). Mafic silicates from the nepheline syenites of the Motzfeldt centre, south Greenland. Mineralogical Magazine 48, 1–12.

Juster, T. C., Grove, T. L. & Perfit, M. R. (1989). Experimental constraints on the generation of Fe–Ti basalts, andesites and rhyodacites at the Galapagos spreading center, 85 degrees W and 95 degrees W. Journal of Geophysical Research 94, 9251–9274.[Web of Science]

Kalsbeek, F. & Taylor, P. N. (1985). Isotopic and chemical variation in granites across a Proterozoic continental margin—the Ketilidian mobile belt of South Greenland. Earth and Planetary Science Letters 73, 65–80.

Konnerup-Madsen, J. & Rose-Hansen, J. (1982). Volatiles associated with alkaline igneous rift activity: fluid inclusions in the Ilímaussaq intrusion and the Gardar granitic complexes (south Greenland). Chemical Geology 37, 79–93.

Konnerup-Madsen, J., Larsen, E. & Rose-Hansen, J. (1979). Hydrocarbon-rich fluid inclusions in minerals from the alkaline Ilímaussaq complex, South Greenland. Bulletin de Minéralogie 102, 642–653.

Konnerup-Madsen, J., Rose-Hansen, J. & Larsen, E. (1981). Hydrocarbon gases asociated with alkaline igneous activity: evidence from compositions of fluid inclusions. Rapport Grønlands Geologiske Undersøgelse 103, 99–108.

Larsen, J. G. (1977). Petrology of the late lavas of the Eriksfjord Formation, Gardar province, South Greenland. Bulletin Grønlands Geologiske Undersøgelse 125, 31 pp.

Larsen, L. M. (1976). Clinopyroxenes and coexisting mafic minerals from the alkaline Ilímaussaq Intrusion, South Greenland. Journal of Petrology 17, 258–290.[Abstract/Free Full Text]

Larsen, L. M. (1981). Chemistry of feldspars in the Ilímaussaq augite syenite with additional data on some other minerals. Rapport Grønlands Geologiske Undersøgelse 103, 31–37.

Larsen, L. M. & Sørensen, H. (1987). The Ilímaussaq Intrusion—progressive crystallization and formation of layering in an agpaitic magma. In: Fitton, J. G. & Upton, B. G. J. (eds) Alkaline Igneous Rocks. Geological Society, London, Special Publications 30, 473–488.

Lieberman, J. & Petrakakis, K. (1990). TWEEQU thermobarometry, analysis of uncertainties and applications to granulites from western Alaska. Canadian Mineralogist 29, 857–887.

Lindsley, D. H. (1983). Pyroxene thermometry. American Mineralogist 68, 477–493.[Abstract]

Lindsley, D. H. & Frost, B. R. (1992). Equilibria among Fe–Ti oxides, pyroxenes, olivine, and quartz: Part i. Theory. American Mineralogist 77, 987–1003.[Abstract]

Markl, G., Frost, B. R. & Bucher, K. (1998). The origin of anorthosites and related rocks from the Lofoten islands, Northern Norway: I. Field relations and estimation of intrinsic variables. Journal of Petrology 39, 1425–1452.

Markl, G., Marks, M. & Wirth, R. (2001). The influence of T, a_SiO2, and fO₂ on exsolution textures in Fe–Mg olivine: an example from augite syenites of the Ilímaussaq Intrusion, South Greenland. American Mineralogist 86, 36–46.[Abstract/Free Full Text]

McBirney, A. R. & Noyes, R. M. (1979). Crystallisation and layering of the Skaergaard intrusion. Journal of Petrology 20, 487–545.[Abstract/Free Full Text]

McBirney, A. R. Baker, B. H. & Nilson, R. H. (1985). Liquid fractionation, Part 1: Basic principles and experimental simulations. Journal of Volcanology and Geothermal Research 24, 1–24.

Michael, P. J. (1984). Chemical differentiation of the Cordillera Paine granite (southern Chile) by in situ fractional crystallization. Contributions to Mineralogy and Petrology 87, 179–195.

Nielsen, B. L. & Steenfelt, A. (1979). Intrusive events at Kvanefjeld in the Ilímaussaq igneous complex. Bulletin of the Geological Society of Denmark 27, 143–155.

Parsons, I. (1980). Alkali-feldspar and Fe–Ti oxide exsolution textures as indicators of the distribution and subsolidus effects of magmatic ‘water’ in the Klokken layered syenite intrusion, South Greenland. Transactions of the Royal Society of Edinburgh: Earth Sciences 71, 1–12.

Parsons, I. (1979). The Klokken gabbro–syenite complex, South Greenland: cryptic variation and origin of inversely graded layering. Journal of Petrology 20, 653–694.[Abstract/Free Full Text]

Parsons, I. (1981). The Klokken gabbro–syenite complex, South Greenland: quantitative interpretation of mineral chemistry. Journal of Petrology 22, 233–260.[Abstract/Free Full Text]

Parsons, I. & Brown, W. L. (1988). Sidewall crystallization in the Klokken intrusion: zoned ternary feldspars and coexisting minerals. Contributions to Mineralogy and Petrology 98, 431–443.

Parsons, I. & Butterfield, A. W. (1981). Sedimentary features of the Nunarssuit and Klokken syenites, S Greenland. Journal of the Geological Society, London 138, 289–306.[Abstract/Free Full Text]

Parsons, I., Mason, R. A., Becker, S. M. & Finch, A. A. (1991). Biotite equilibria and fluid circulation in the Klokken Intrusion. Journal of Petrology 32, 1299–1333.[Abstract/Free Full Text]

Paslick, C. R., Halliday, A. N., Davies, G. R., Mezger, K. & Upton, B. G. J. (1993). Timing of Proterozoic magmatism in the Gardar Province, southern Greenland. Geological Society of America Bulletin 105, 272–278.[Abstract/Free Full Text]

Patchett, P. J. (1980). Thermal effects of basalt on continental crust and crustal contamination of magmas. Nature 283, 559–561.

Petersilie, I. A. & Sørensen, H. (1970). Hydrocarbon gases and bituminous substances in rocks from the Ilímaussaq alkaline intrusion, South Greenland. Lithos 3, 59–76.

Pouchour, J. L. & Pichoir, F. (1984). A new model for quantitative X-ray microanalyis. Part 1: Applications to the analysis of homogeneous samples. Recherche Aerospatiale 3, 13–38.

Poulsen, V. (1964). The sandstones of the Precambrian Eriksfjord Formation in South Greenland. Rapport Grønlands Geologiske Undersøgelse 2, 16 pp.

Powell, M. (1978). The crystallisation history of the Igdlerfigssalik nepheline syenite intrusion, Greenland. Lithos 11, 99–120.

Saas, J. H., Nielsen, B. L., Wollenberg, H. A. & Munroe, R. J. (1972). Heat flow and surface radioactivity at two sites in South Greenland. Journal of Geophysical Research 77, 6435–6444.[Web of Science]

Schairer, J. F. (1950). The alkali feldspar join in the system NaAlSiO₄–KAlSiO₄–SiO₂. Journal of Geology 58, 512–517.

Sørensen, H. (1966). On the magmatic evolution of the alkaline igneous province of South Greenland. Rapport Grønlands Geologiske Undersøgelse 7, 19 pp.

Sparks, R. S. J. & Huppert, H. E. (1984). Density changes during fractional crystallisation of basaltic magmas: fluid dynamic implications. Contributions to Mineralogy and Petrology 85, 300–309.

Sparks, R. S. J., Huppert, H. E. & Turner, J. S. (1984). The fluid dynamics of evolving magma chambers. Philosophical Transactions of the Royal Society of London, Series A 310, 511–534.

Stephenson, D. (1972). Alkali clinopyroxenes from nepheline syenites of the South Qoroq Centre, south Greenland. Lithos 5, 187–201.

Stephenson, D. (1974). Mn and Ca enriched olivines from nepheline syenites of the South Qoroq Centre, south Greenland. Lithos 7, 35–41.

Stephenson, D. & Upton, B. G. J. (1982). Ferromagnesian silicates in a differentiated alkaline complex: Kungnat Fjeld, South Greenland. Mineralogical Magazine 46, 283–300.

Stern, C. (1979). Open and closed system igneous fractionation within two Chilean ophiolites and the tectonic implication. Contributions to Mineralogy and Petrology 68, 243–258.

Stevenson, R., Upton, B. G. J. & Steenfelt, A. (1997). Crust–mantle interaction in the evolution of the Ilímaussaq Complex, South Greenland: Nd isotopic studies. Lithos 40, 189–202.

Taylor, H. P., Jr (1980). The effects of assimilation of country rocks by magmas on ¹⁸O/¹⁶O and ⁸⁷Sr/⁸⁶Sr systematics in igneous rocks. Earth and Planetary Science Letters 47, 243–254.

Tegner, C., Robins, B., Reginiussen, H. & Grundvig, S. (1999). Assimilation of crustal xenoliths in a basaltic magma chamber; Sr and Nd isotopic constraints from the Hasvik layered intrusion, Norway. Journal of Petrology 40, 363–380.

Upton, B. G. J. & Emeleus, C. H. (1987). Mid-Proterozoic alkaline magmatism in southern Greenland: the Gardar province. In: Fitton, J. G. & Upton, B. G. J. (eds) Alkaline Igneous Rocks. Geological Society, London, Special Publications 30, 449–471.

Upton, B. G. J., Stephenson, D. & Martin, A. R. (1985). The Tugtutoq older giant dyke complex: mineralogy and geochemistry of an alkali gabbro–augite–syenite–foyaite association in the Gardar Province of South Greenland. Mineralogical Magazine 49, 624–642.

Watt, S. W. (1966). Chemical analyses from the Gardar Igneous Province, South Greenland. Rapport Grønlands Geologiske Undersøgelse 6.

Wilson, R. & Sørensen, H. (1994). Fractionation and assimilation processes in the Fongen–Hyllingen layered intrusion, Norway. In: Magmatic Processes; do the Answers Lie in the Rocks? Mineralogical Society Bulletin 105, 15–31.

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				J. Schonenberger and G. Markl The Magmatic and Fluid Evolution of the Motzfeldt Intrusion in South Greenland: Insights into the Formation of Agpaitic and Miaskitic Rocks J. Petrology, September 1, 2008; 49(9): 1549 - 1577. [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]

				M. S. Riishuus, D. W. Peate, C. Tegner, J. R. Wilson, and C. K. Brooks Petrogenesis of Cogenetic Silica-Oversaturated and -Undersaturated Syenites by Periodic Recharge in a Crustally Contaminated Magma Chamber: the Kangerlussuaq Intrusion, East Greenland J. Petrology, March 1, 2008; 49(3): 493 - 522. [Abstract] [Full Text] [PDF]

				G. Graser and G. Markl Ca-rich Ilvaite-Epidote-Hydrogarnet Endoskarns: a Record of Late-Magmatic Fluid Influx into the Persodic Ilimaussaq Complex, South Greenland J. Petrology, February 1, 2008; 49(2): 239 - 265. [Abstract] [Full Text] [PDF]

				L. N. Kogarko, C. T. Williams, and A. R. Woolley Compositional evolution and cryptic variation in pyroxenes of the peralkaline Lovozero intrusion, Kola Peninsula, Russia Mineralogical Magazine, August 1, 2006; 70(4): 347 - 359. [Abstract] [Full Text] [PDF]

				I. Parsons, P. Thompson, M. R. Lee, and N. Cayzer Alkali Feldspar Microtextures as Provenance Indicators in Siliciclastic Rocks and Their Role in Feldspar Dissolution During Transport and Diagenesis Journal of Sedimentary Research, September 1, 2005; 75(5): 921 - 942. [Abstract] [Full Text] [PDF]

				R. HALAMA, T. VENNEMANN, W. SIEBEL, and G. MARKL The Gronnedal-Ika Carbonatite-Syenite Complex, South Greenland: Carbonatite Formation by Liquid Immiscibility J. Petrology, January 1, 2005; 46(1): 191 - 217. [Abstract] [Full Text] [PDF]

				R. C. PRICE, A. F. COOPER, J. D. WOODHEAD, and I. CARTWRIGHT Phonolitic Diatremes within the Dunedin Volcano, South Island, New Zealand J. Petrology, November 1, 2003; 44(11): 2053 - 2080. [Abstract] [Full Text] [PDF]

				I. M. Coulson and I. M. Coulson Evolution of the North Qoroq centre nepheline syenites, South Greenland: alkali-mafic silicates and the role of metasomatism Mineralogical Magazine, October 1, 2003; 67(5): 873 - 892. [Abstract] [Full Text] [PDF]

				M. Marks, M. Marks, and G. Markl Ilimaussaq 'en miniature': closed-system fractionation in an agpaitic dyke rock from the Gardar Province, South Greenland (contribution to the mineralogy of Ilimaussaq no. 117) Mineralogical Magazine, October 1, 2003; 67(5): 893 - 919. [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]

				G. MARKL, M. MARKS, G. SCHWINN, and H. SOMMER Phase Equilibrium Constraints on Intensive Crystallization Parameters of the Ilimaussaq Complex, South Greenland J. Petrology, December 1, 2001; 42(12): 2231 - 2257. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (33) Request Permissions Google Scholar Articles by MARKS, M. Articles by MARKL, G. Search for Related Content GeoRef GeoRef Citation Social Bookmarking          What's this?

Online ISSN 1460-2415 - Print ISSN 0022-3530

Copyright © 2009 Oxford University Press

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics