Journal of Petrology Advance Access originally published online on September 16, 2004
Journal of Petrology 2004 45(11):2197-2223; doi:10.1093/petrology/egh053
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology 45(11) © Oxford University Press 2004; all rights reserved
Structure and Dynamics of the Laacher See Magma Chamber (Eifel, Germany) from Major and Trace Element Zoning in Sanidine: a Cathodoluminescence and Electron Microprobe Study
GEOWISSENSCHAFTLICHES ZENTRUM GÖTTINGEN. ABT. GEOCHEMIE, GOLDSCHMIDTSTR. 1, 37077 GÖTTINGEN, GERMANY
RECEIVED JUNE 7, 2002; ACCEPTED JUNE 24, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Microtextures and zoning patterns in sanidine phenocrysts from the phonolitic Laacher See Tephra (Germany) were investigated in order to constrain processes occurring in the magma chamber before eruption. The Laacher See Tephra unit is chemically zoned and has been inferred to represent the product of eruption of an inverted, layered magma chamber. Samples from various levels in the deposit were investigated. We used a combination of textural studies (optical microscopy and back-scattered electron (BSE) imaging), electron microprobe (EMP) analysis of major, minor and trace elements (Ca, Na, K, Al, Si, Ba, Sr, Fe and Ti) and element mapping. The samples studied contain two feldspar phases and the ternary composition of the sanidine thus constrains the temperature of crystallization, whereas its trace element content reflects the melt composition. The large diversity in textures found in the sanidine crystals can be classified into three types: composite (C-type), pseudo-oscillatory (PO-type) and resorbed/patchy (R-type). Trace element-poor, lamellar composite alkali feldspars (C-type) are found in samples inferred to represent the top and the middle part of the magma chamber. They grew as composite crystals from the melt at temperatures as low as 650°C in a highly differentiated and volatile-rich boundary layer at the magma chamber roof or wall. Pseudo-oscillatory zoning with resorption surfaces (PO-type) is found in sanidines from samples inferred to represent the middle part and the base of the magma chamber. Repeated, large An variations (1–4 mol %) reflect temperature variations of 100–300°C associated with changes in water content. Variations of minor and trace elements (Ba, Sr and Ti) in sanidine, decoupled from the pseudo-oscillations of the major elements, reflect chemical changes in the melt, and are damped by chemical diffusion in the melt. Both major and minor element variations in PO-type crystals are interpreted as the progressive influence of a more mafic, hotter melt. This may be partly explained by the settling of crystals through the thermal and chemical gradient existing in the magma chamber; however, the additional role of magma recharge may be required to explain the large temperature variations. Resorbed/patchy-zoned crystals (R-type), found mainly in samples corresponding to the middle part of the magma chamber, reflect early growth in a differentiated boundary layer, followed by resorption and overgrowth in the main magma body. Many of the Laacher See sanidine crystals did not crystallize for the most part in the melt in which they were erupted. In the presence of preserved major and trace element zonation in the magma, this observation indicates crystal dispersion within a layered magma chamber without large-scale mixing and overturn.
KEY WORDS: feldspar; zoning patterns; magma chambers; phonolite; trace elements; electron microprobe; geochemistry
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
The interpretation of zoning patterns in phenocrysts has proved a useful tool for the reconstruction of magma chamber processes, such as crystallization, chemical differentiation and magma recharge. It thus provides insight into the thermal and dynamic regimes of magma chambers. Numerous studies (e.g. Anderson, 1983
; Nixon & Pearce, 1987; Pearce et al., 1987; Pearce & Kolisnik, 1990; Singer et al., 1995; Ginibre et al., 2002a) have investigated the major element zoning patterns of plagioclase. They distinguish large discontinuities in the anorthite content of the crystal, reflecting major recharge events in crustal magma chambers, from more subtle changes, such as oscillatory zoning, which may be caused by kinetic effects or convection. The interpretation of these textures is based on experimental studies of crystallization (Lofgren, 1974) and resorption (Tsuchiyama, 1985), as well as numerical modeling (Allègre et al., 1981; Lasaga, 1982; Loomis, 1982; L'Heureux & Fowler, 1994).
A number of recent studies have used minor and trace elements to demonstrate the closed- or open-system behavior of magma chambers (Blundy & Shimizu, 1991; Singer et al., 1995), and to determine the location of crystallization, for example with respect to magma chamber boundary layers (Kuritani, 1998; Ginibre et al., 2002b). In situ Sr isotope data also provide information on recharge and contamination processes in magma chambers (e.g. Davidson & Tepley, 1997; Tepley et al., 1999).
Detailed studies of zoning patterns in magmatic minerals other than plagioclase are relatively rare (e.g. pyroxene: Shimizu, 1990; olivine: Pearce, 1987). Shore & Fowler (1996) showed that oscillatory zoning occurs in many magmatic minerals, which thus could potentially be used in the study of magma chamber processes. Anderson et al. (2000) investigated Ba zoning in sanidine from Bishop Tuff rhyolites, together with glass inclusions in quartz phenocrysts, and showed that most of the crystals sink towards the bottom of the magma chamber. Troll & Schmincke (2002) used zoning patterns in ternary feldspars in rhyolites from Gran Canaria to identify complex mixing events between a trachytic and two rhyolitic end-members in the magma chamber prior to eruption. Knesel et al. (1999) used Sr isotope zoning in sanidine to show the existence of a crustally contaminated magma at the onset of sanidine growth, and the role of subsequent magma chamber recharge. Perini et al. (2003) investigated Sr isotopic zoning in K-feldspar megacrysts to document mixing/mingling processes in trachytes. However, such detailed studies of the range of zoning patterns in sanidine and their implications for magma chamber processes are still scarce and the potential of zoning patterns in sanidine has not been fully explored (e.g. in alkaline-evolved magma systems).
We present, here, results of a study of zoning patterns in sanidine crystals within tephra erupted from a phonolitic magma chamber which lay beneath the Quaternary Laacher See maar volcano in the Eifel volcanic province of Germany. The Laacher See Tephra has been studied extensively. Geological (Bogaard & Schmincke, 1984, 1985), petrological (Wörner & Schmincke, 1984a; Wörner & Wright, 1984; Tait, 1988; Tait et al., 1989), experimental (Berndt et al., 2001), geochemical (Wörner et al., 1983, 1985, 1987; Wörner & Schmincke, 1984b; Harms & Schmincke, 2000) and geochronological data (Bourdon et al., 1994; Bogaard, 1995) are available and provide excellent constraints on the development of the magma chamber. Chemical layering of the magma chamber has been inferred (Wörner & Schmincke, 1984a), as well as the influence of a basanitic magma recharge, either shortly before or during eruption (Wörner & Wright, 1984). The chemical variations within the Laacher See Tephra have been modeled by two-stage fractional crystallization: from basanite to mafic phonolite and from phonolite to highly evolved phonolite (Wörner & Schmincke, 1984b). The development of chemical stratification within the magma chamber was modeled by Tait et al. (1989).
Some questions remain, however, concerning the mechanisms of differentiation and the layering processes, in particular liquid and crystal movements within the magma chamber, and interaction of the magma with the partially crystallized magma chamber wall. For example, the large compositional range of feldspars (and also other minerals, e.g. amphiboles and pyroxenes; Wörner & Schmincke, 1984a) within pumice samples seems to indicate various origins for the phenocrysts and subsequent redistribution. In particular, it has been suggested that most crystals present in the pumice did not crystallize in situ but instead originated at various levels of the magma chamber wall (Tait, 1988; Tait et al., 1989). In order better to understand these processes, we examine the zoning patterns of sanidine from various stratigraphic levels of the deposit inferred to represent different levels in the magma chamber. We focus on sanidines because (1) they show the greatest diversity in composition and zoning pattern, (2) they are abundant in all samples, and (3) sanidine zoning patterns have been less thoroughly investigated up to now as a tool to reconstruct magma chamber evolution.
A combination of imaging methods is used, including optical and cathodoluminescence (CL) microscopy, back-scattered electron (BSE) images and wavelength dispersive X-ray (WDX) element mapping, as well as quantitative analysis by electron microprobe of minor and trace element abundances. Combined with geothermometry and trace element partitioning models, the information recorded in the zoning patterns of sanidine phenocrysts adds to our understanding of the evolution of the Laacher See magma chamber. We have found that it is possible to identify the provenance, growth conditions and growth history of each type of sanidine crystal. We also discuss the possibility of an earlier recharge event, with basanitic magma, during the crystallization of the sanidine phenocrysts.
|
THE LAACHER SEE MAGMA CHAMBER |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
The Laacher See volcano is part of the Quaternary East Eifel volcanic field of Germany (Fig. 1), characterized by intraplate alkaline volcanism (basanites, tephrites, phonolites, leucitites and nephelinites; Schmincke et al., 1983
). The volcano erupted 
6·3 km3 (Harms & Schmincke, 2000) of phonolitic magma during a period of a few days, 12 880 years ago (Litt et al., 2001).
|
The deposit (the Laacher See Tephra: LST, Fig. 2) exhibits vertical zonation in its petrographic characteristics, and major and trace element chemistry, and has been inferred to represent the inverted content of a zoned magma chamber (Wörner et al., 1983
; Wörner & Schmincke, 1984a). This is supported by the very gradual bulk-rock compositional zoning of the LST in all elements and the curved trends in major and trace element variation diagrams (Wörner, 2003), which rule out models involving binary magma mixing (Eichelberger et al., 2000). Three main compositional units have been distinguished (Bogaard & Schmincke, 1984; Wörner & Schmincke, 1984a, 1984b). The Lower Laacher See Tephra (LLST) is almost aphyric, highly differentiated, and has been inferred to represent the volatile-rich top of the magma chamber. The Middle Laacher See Tephra (MLST) is a crystal-poor, differentiated phonolite which appears to have sampled the main volume of the central part of the magma chamber. It has decreasing crystal content and an increasingly differentiated composition towards the lower part of the deposit (i.e. towards the magma chamber top). The Upper Laacher See Tephra (ULST) is a crystal-rich, more mafic phonolite, inferred to represent the base of the magma chamber. Wörner & Wright (1984) described hybrid lavas present in the last erupted products at the top of the deposit, ranging in composition between ULST phonolite and a more mafic composition. They inferred that mingling and mixing with a basanite magma occurred at the bottom of the magma chamber, and suggested a very short time for mingling, implying that it occurred during the eruption and thus possibly did not trigger, but may have been a consequence of, the eruption.
|
Contact metamorphic xenoliths of country rock of Devonian age were mostly erupted with the upper and central part of the magma chamber (LLST and MLST). On this basis, a relatively shallow depth for the Laacher See magma chamber is estimated at 3·5–7 km [based on Wörner & Schmincke (1984b
) and using the revised magma volume estimate of Harms & Schmincke (2000)]. A great variety of mafic to felsic cumulate xenoliths (Tait, 1988) and carbonatitic syenites (Liebsch, 1996) were erupted mainly within the ULST during the last and most explosive stage of the eruption; these are interpreted as samples from the magma chamber wall and its apophyses.
Samples
We have investigated feldspars from five pumice samples used in the study of Wörner & Schmincke (1984a). Samples were chosen from different parts of the deposit in order to represent the evolution between the bottom and the top of the magma chamber (Fig. 2). Detailed mineralogical descriptions have been given by Wörner & Schmincke (1984a).
Sample 1002 (LLST) is a white pumice from the very base of the deposit (top of the magma chamber). Phenocrysts are mainly sanidine (c. 1%) but nepheline, plagioclase, haüyne, clinopyroxene, amphibole, titanomagnetite, titanite, apatite, rare zircons and cancrinite also occur.
Samples 1034 and 1050 are white to light gray pumices from the middle and top of the MLST (middle and base of the central part of the magma chamber). Phenocrysts are still rare in 1034 (<5%) but more abundant in 1050 (c. 5–10%). They are mainly feldspars (sanidine and plagioclase), large euhedral haüyne, titanite, apatite, subordinate clinopyroxene, amphibole and titanomagnetite. Clots, 1–3 cm in size, of a more mafic phonolite with crenular margins, petrographically identical to the more mafic pumice found from higher levels of the deposit, are also found at these sampled levels and occur in various states of disaggregation.
Sample 1088 (ULST) is a mafic phonolite from the top of the deposit (base of the magma chamber) and has a phenocryst content of 20%, mainly feldspars (plagioclase and sanidine), amphibole, clinopyroxene, titanite, apatite and titanomagnetite. There is no evidence for significant mixing with a mafic magma (only few small phlogopite crystals and rare Fo-rich olivines are found in the matrix).
Sample 1099 is a hybrid sample from the top of the deposit (Wörner & Wright, 1984) and contains, in addition to mineral phases similar to those of sample 1088, significant amounts of crystals from the basanitic end-member: olivine, pyroxene megacrysts and phlogopite. Large, resorbed nepheline crystals are also present. Because the feldspar mineralogy is the same as in sample 1088 and represents the mafic phonolitic end-member, we consider both samples together for the study of ULST.
Sanidine and plagioclase occur in all the pumice samples studied, with sanidine always being the more abundant phase. A few plagioclase feldspars were analyzed but are not discussed in detail in this study. Several thin sections of samples 1050, 1088 and 1099 were investigated, whereas mineral separates were used for samples 1002 and 1034 because of their low crystal contents. In each of the samples, 15–50 sanidine crystals were investigated by optical microscopy, 8–27 crystals were chosen for BSE- and CL-imaging and quantitative analysis, and 2–12 for semi-quantitative WDX profiles. The crystals were classified into three different zoning types; the images shown in the subsequent figures are representative of the observed crystal population. Thin sections of cumulate and carbonatitic syenite xenoliths were also investigated by optical microscopy in order to compare the feldspar morphology and zoning patterns to those from the pumices; however, they were not analyzed in the present study, as the feldspars are generally unzoned.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Quantitative point analysis of major, minor and trace elements by electron microprobe
All microprobe analyses were performed on a JEOL JXA 8900 RL electron microprobe (EMP) at Abt. Geochemie, Geoscience Center, Göttingen. Quantitative WDX electron microprobe analyses were performed at 20 kV and 40 nA, using a 2 or 5 µm defocused beam. Alkali elements (K and Na) were measured first, and all major elements (Na, K, Ca, Si and Al) within the first 90 s (measurement time: 16 s on peak). A study of Na loss shows that, during this time, alkali elements have not been significantly lost. Minor and trace elements (Ba, Sr, Fe and Ti) were then analyzed for 4 min on peak. The influence of alkali loss on the minor and trace element measured concentration is believed to be small for these elements in feldspar; the net-intensity/concentration ratio is constant and does not vary with small variations in the major element composition. Detection limits, ranges of concentrations measured in the studied samples and statistical error (2
) for Ba, Sr, Fe and Ti are listed in Table 1. The spatial resolution of WDX analysis is a few microns.
|
Cathodoluminescence (CL) microscopy
Color CL images were obtained using a hot-cathode CL microscope at the Geoscience Center Göttingen, using a 3 µm diameter beam with a 10 kV acceleration voltage and a 0·3 mA current. Cathodoluminescence is visible light that is emitted by the sample under the electron beam. The exact causes of CL in minerals are not yet fully understood [see Marshall (1988)
for a review] but it has been attributed to two types of processes: (1) activation by trace elements or (2) lattice defects. In feldspar, the causes of cathodoluminescence have been attributed to the following elements (Marshall, 1988; Finch & Klein, 1999): Ti4+ acts as an (indirect) activator for blue luminescence (460 nm), Fe3+ for red luminescence (700 nm), and Fe2+ and Mn2+ for green luminescence (550 and 570 nm, respectively). CL gray-scale images were also acquired using a CL detector on the EMP, integrating the light emission from 200–900 nm wavelength as a whole. The contrast is generally lower than with the CL microscope. However, CL images from the EMP can be obtained simultaneously with other signals (back-scattered electrons) and images can be directly compared.
Chemical imaging using EMP
BSEs are emitted from the sample surface and their intensity depends on the mean atomic number of the sample. In plagioclase, the intensity is linearly correlated with An-content (Ginibre et al., 2002a). In sanidine, Or- and Ba-contents influence the BSE intensity. The BSE signal is a good imaging tool for Ba in sanidines that have a large range in Ba concentration and fairly constant major element composition, because of the high atomic number of Ba. In composite alkali feldspars with low Ba-content, the BSE signal is more representative of Ab–Or variations. The spatial resolution of BSE images is better than 1 µm. For a restricted number of crystals, images of 2-D compositional variation were obtained by element mapping of Ba, Ca, Na and K, using semi-quantitative WDX analysis (15 kV, 50 nA). The spatial resolution of WDX analysis (a few microns) is not as good as for BSE imaging (<1 µm).
Line analysis
The same method as for WDX element mapping is applied to line measurements. A 1 µm defocused beam with 20 kV acceleration voltage and 50 nA beam current was used. For all elements (K, Na, Ca, Ba and Sr), X-ray intensity was counted on peak for 5 s per 1 µm pixel along a line across the crystal. No background and matrix corrections were applied but the variations in raw intensity reflect the variations in the element concentration. This allows the semi-quantitative analysis of major and minor elements along a profile with a better spatial resolution and in a shorter time than by fully quantitative analysis. A large number of crystals may thus be compared.
|
ZONING PATTERNS IN LAACHER SEE SANIDINE |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Different zoning types are distinguished on the basis of BSE and CL images, and shown in Figs 3 and 4. Representative quantitative EMP analyses of crystals of the various types, from different samples, are given in Table 2. The complete dataset can be downloaded from the Journal of Petrology website at http://www.petrology.oupjournals.org/. In CL images, Laacher See sanidines show only blue and red luminescence, and the spectra do not show a peak at 550 nm, indicating that Fe is only present as Fe3+. Blue luminescence correlates well with the measured Ti concentration, whereas red luminescence does not correlate with Fe. Actually, it appears to be the Ti/Fe ratio that controls the color (Fig. 5). In contrast to observations by Liebsch (1996)
, the CL colors are well correlated with Ti/Fe in the investigated crystals and not with the Or-content of the sanidine, with red CL being found in the most sodic as well as the most potassic sanidines. We define three types of zoning: PO-type, R-type and C-type. (1) Pseudo-oscillatory sanidines (PO-type) show narrow, concentric resorption surfaces, separating zones of variable composition. This type dominates in the ULST and is found also in the MLST. (2) Resorbed/patchy-zoned sanidines (R-type) have resorbed, red-luminescing, trace element-poor patches, which also occur as cores in some PO-type sanidines. R-type crystals are common in the MLST and are subordinate in the ULST. (3) Composite sanidines (C-type) are composed of two distinct phases—sanidine and albite—forming intergrown lamellae. This type is the only sanidine type present in the most evolved phonolite at the top of the Laacher See magma chamber (sample 1002). In the MLST, it is also found together with other types, but always with resorbed rims. The relative abundance of the various types of sanidine in each unit of the LST is summarized in Table 3. In addition to these main types, a few atypical crystals occur, in particular some unzoned crystals found in the MLST. It is important to note that the most common type of sanidine—PO-type—as well as the R-type are not observed in the cumulates, where most sanidine crystals do not show any zoning patterns. However, some composite textures are seen in the syenite cumulates and will be briefly described further below.
|
|
|
|
|
Pseudo-oscillatory-zoned crystals (PO-type)
Most crystals in the ULST and some from the MLST exhibit near-periodic, concentric zoning patterns that would normally be described as oscillatory. However, as shown below, these ‘oscillations’ are in fact defined by quasi-periodic, rounded resorption surfaces and are therefore referred to as pseudo-oscillatory (PO-type) zoning patterns. Similar patterns were described for andesitic plagioclase from Parinacota Volcano, North Chile, as Saw-Tooth zoning with Resorption (STR) by Ginibre et al. (2002a)
. Pseudo-oscillatory zoning patterns can be observed using various imaging methods, reflecting variations in the concentrations of different elements. As illustrated in Fig. 6 (crystal 1099-S1), the BSE image principally reflects Ba zoning, whereas the optical image reflects major element variations (K and Ca, which are anticorrelated). Figure 7 shows the BSE image and the CL image (reflecting mainly Ti/Fe) of two other crystals (1050b-S3 and 1034-8-S4). PO-type sanidines exhibit blue CL with variable intensity (Fig. 4a and b). PO zoning patterns vary in zone morphology and profile shape between crystals and, within individual crystals, between faces. Unfortunately, the crystals observed in thin sections of pumice are randomly oriented and rarely parallel to a crystallographic plane. The observed zoning in a given crystal section reflects, therefore, both the history of the crystal and the orientation in which it is cut. Like all sanidines (of any type) in the MLST and ULST, PO-types exhibit resorption at the rim and the crystals have rounded shapes. The last resorption surface is followed by a thin (1–5 µm) layer, very rich in Ba, and an outer 10–30 µm overgrowth.
|
|
The chemistry of all analyzed PO-type crystals from samples 1099–1088 (ULST) and 1050 and 1034 (MLST) is shown in Fig. 8. Major elements vary between 60 and 75 mol % Or. The ternary diagram (Fig. 8a) shows three distinct trends within crystals or crystal parts, in which An content decreases with increasing Or content. These trends are roughly parallel to each other between crystals or crystal parts, and depend partly on the sample position in the LST stratigraphy: PO-type sanidines from ULST fall mostly on trends 1 and 2, whereas those MLST fall mostly on trends 2 and 3. Minor and trace elements are shown in Fig. 8b. Sr, Ba and Ti show large variations, whereas Fe is almost constant in all PO-type sanidines, except in some patchy red-luminescing cores, where it is higher.
|
We describe in detail, below, three crystals (1099-S1, 1050b-S3 and 1034-8-S4), which are representative of zoning patterns and compositions of the range of Laacher See PO-type sanidines. These crystals are shown in Figs 6 and 7. In Fig. 8a, crystal 1099-S1 falls on trend 2 in the core and trend 1 at the rim, crystal 1050b-S3 remains on trend 2 and crystal 1034-8-S4 falls on trend 3. WDX semi-quantitative profiles of Ca and Ba for 1050b-S3 and 1034-8-S4 are shown in Fig. 7 and detailed quantitative element profiles for 1050b-S3 and 1099-S1 crystals are shown in Fig. 9. Ca variations are irregular in both profile shapes and in wavelength, but exhibit fairly constant amplitudes. Saw-tooth patterns in Ca profiles (Figs 7 and 9) often occur at the resorption surfaces observed on the images (crystal 1050b-S3, Fig. 7a); from core to rim, Ca increases abruptly at the resorption surface and then decreases smoothly. Typical amplitudes of the oscillations are 1 mol % An, whereas the total variations within individual crystals reach 3 mol % An. In other crystals (1099-S1, Figs 6 and 9), variations are smoother and differences between crystallographic orientations are larger. Crystals from the MLST tend to have lower-amplitude variations in major elements (crystal LS 1034-8-S4, Fig. 8b).
|
In all crystals analyzed, Ba, Ti and Sr (Fig. 9) show two superimposed types of variations: (1) an overall increase followed by a decrease is seen in these elements in most PO crystals from both the ULST and MLST, which is decoupled from the An-content variations. These variations are, here, referred to as large-wavelength (LW) variations. (2) Abrupt variations at the resorption surfaces are correlated with An pseudo-oscillations on a 20–100 µm scale. We refer to these variations in trace elements as PO variations. The largest LW variations are seen in Ba, with concentrations ranging from 0·2 to 2 wt % BaO. Superimposed PO variations are subordinate (0·1% BaO). Sr exhibits both LW and PO variations with amplitudes of about 2000 and 600 ppm, respectively. LW variations in Ti reach 300 ppm, whereas PO variations are only slightly above the analytical uncertainty. For both types of variation, there is a positive correlation between the different elements (Ba, Sr and Ti). Furthermore, the range of PO variation in An- and Ba-content is generally low in the crystal cores. LW and PO variations, in particular the increase in Ba and the onset of Ca pseudo-oscillations, are often correlated. Both types of variations vary between crystals and depend on crystallographic orientation. Correlation of PO-type zoning patterns between crystals is not possible, the number and shape of the peaks being different between crystals, even with similar compositions. Decoupling between PO and LW variations (Ca and Ba) is less clear in the more differentiated composition (trend 3 in Fig. 8a, especially crystal 1034-8-S4). Here, Ba variations correlate with Ca pseudo-oscillations (Fig. 7b).
R-type: resorbed red-CL cores and patchy zoning
‘Patches’ characterized by red-CL, irregular boundaries, variable Or-content (60–75 mol % Or), very low Ti, low Ba, Sr, Ca and relatively high Fe concentration (Fig. 10) form either the resorbed core of PO-type crystals (R-type cores, Figs 3d and 4a) or part of patchy zoned crystals (R-type crystals, Figs 3c–f and 4e–f). In R-type crystals, resorption surfaces with inner or outer embayments are overgrown by blue-luminescing zones, rich in minor elements (Fig. 4e–f). These red-CL patches always form during an early stage of crystal growth and can be considered as resorbed cores, even in patchy crystals such as those seen in Fig. 4e. The Or-content in the cores is higher, and An-content lower, than in the rest of the crystal. In the ternary feldspar diagram, as well as in the trace element versus Or-content diagrams (Fig. 10), the red-CL core is in continuity with the compositional trend of the whole crystal for most PO-type or R-type crystals. The red-luminescing cores are overgrown by blue-luminescing sanidine which, in some cases, is pseudo-oscillatory zoned. R-type crystals define trends parallel to those of PO-type crystals in the ternary diagram, but often at lower An- and Or- and higher Ab-contents (trends 1, 2, 3, 4 and 5 in Fig. 10).
|
Composite feldspars (C-type)
C-type sanidines exhibit red-CL colors (Fig. 4e–f) and consist of composite crystals made of two intergrown alkali feldspar phases (albite and sanidine), where sanidine is generally the more abundant host phase (Figs 3a–b and 11a–d).
|
The principal texture on a 1–500 µm scale consists of elongated domains (1–10 µm x 50–500 µm) of the two phases, oriented parallel to the crystallographic axes of the crystal. The occurrence of shorter (1–10 µm) albite-rich domains, normal to the a- or b-axis, suggests that these lamellae have a smaller extension in this direction and therefore have the shape of platelets. The distribution of these albite lamellae forms three main types of patterns at the crystal scale, as well as a large range of intermediates.
(1) Sector crystals (Fig. 11a). One sector is rich in albite lamellae and has an average composition of c. 20 mol % Or. Lamellae are normal to the crystal surface and cleavage. The other sector has relatively few albite lamellae and an average composition of 38 mol % Or. The density of the lamellae varies in the growth direction of the crystal face, forming concentric zones of variable bulk composition.
(2) Albitic cores and composite rims with lamellae oriented normal to the crystal faces (Fig. 11b). This may be a different section of sector-type crystals, as the chemical composition (see below) cannot be distinguished.
(3) Patches of composite feldspars of variable aspect; in some cases, individual albite lamellae can be distinguished, whereas, in other crystals, albite occurs as homogeneous patches (Fig. 11c and d). Some crystals seem to be intermediate between sector- and patchy-type, with possible poorly defined sectors, as in Fig. 11d.
This classification can be refined using the chemical composition. The chemistry of C-type sanidines is shown in Fig. 11e and f. As the albite lamellae are generally too fine to be resolved by the EMP (1–5 µm), the analyses represent composite values and define mixing lines for major and minor elements between the two phases. Accordingly, the composition of the two end-members is respectively 
15 and 
40 mol % Or. The major element composition of all C-type crystals is more differentiated (displaced toward an Ab-rich composition) than that of the R- and PO-types. It should thus be noted that although both exhibit red CL, C-type crystals are distinct from the R-patches in R- and PO-type sanidines. All C-type crystals have low concentrations of Ti, Sr and Ba, with Sr being below the detection limit (70 ppm). Ba concentration is below the detection limit (50 ppm) in the Ab-rich domains of all C-type crystals. The Or-rich domains have less than 90 ppm Ba in the LLST C-type crystals, and between 70 and 224 ppm in the MLST C-type crystals.
Ca- and Fe-contents define two distinct sub-groups. In the first group, both phases have similar and low Ca and Fe concentrations (1% An and 1000 ppm Fe). This group corresponds to the two first types described above, i.e. crystals with well-defined sectors and crystals with well-developed lamellae at the rim. Patchy composite crystals (including crystals with poorly defined sectors) form the other group, where the sanidine phase has low Ca and Fe concentrations (1% An and 1000 ppm), whereas the albite phase has higher Fe and Ca concentrations (up to 7% An and 1400 ppm Fe). Low-An C-type crystals are found almost exclusively in the LLST (sample 1002), where they are by far the most common. They form flat euhedral or partly rounded crystals. High-An C-type crystals are found both in the LLST (sample 1002) and the MLST (samples 1034 and 1050). In samples 1002 and 1034, they rarely show evidence of resorption, with Ab-, An-, Fe- and Ti-rich overgrowths. In sample 1050, these crystals are strongly resorbed, and the blue-luminescing overgrowth is also richer in Ba and Sr.
Composite crystals also occur in syenitic cumulates, but they have certain different aspects from those observed in the pumices. They have various forms of albite patches, often found at the rim of more potassic crystals, and never show clear sectors. Moreover, the general morphology is different, with elongated crystals forming sheet-like or fan-like crystal aggregates. Similar ‘cumulates crystals’ are also observed rarely in sample 1002, but were not analyzed in this study.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Controls on sanidine composition
Thermodynamic control on the ternary feldspar system
Although this study concentrates on sanidine zoning, it is important to consider the two-feldspar assemblage, if present, because, provided it reflects equilibrium, it gives useful information about the physical conditions of feldspar growth. Two feldspars are present in all Laacher See samples, but there is also evidence for disequilibrium. Our compositional data are compared in Fig. 12 with those of the cumulate xenoliths, which were inferred to represent an equilibrium assemblage by Tait (1988)
. Most compositions of R-type and PO-type sanidines from the pumice samples lie between the sanidine compositions of the two groups of two-feldspar cumulates (groups 2 and 3). This suggests that these sanidines also grew from a liquid saturated in two feldspar phases. C-type sanidines contain two phases within a single crystal and must therefore have been at the solvus at some temperature. Whether this temperature is the crystallization temperature is discussed in a later section.
|
Nekvasil (1992
, 1994) showed that the observed compositional trends in feldspar phases during crystallization provide important qualitative information on the evolution of the two-feldspars + melt system. The compositions of the two coexisting feldspars lie at the intersection line of the liquidus and the solvus surfaces in an Ab–An–Or–temperature (T) diagram, the position of this line being strongly dependent on water content. Two types of trends with falling temperature are possible (Nekvasil, 1992), as shown in the Ab–An–Or diagram in Fig. 13. A temperature drop at constant water activity, associated with crystallization and differentiation of the melt, will shift the composition of the growing sanidine towards Ab-rich, An- and Or-poor compositions (trend 1). By contrast, a temperature drop together with PH2O increase at constant bulk composition shifts the composition of the sanidine towards Ab- and An-poor and Or-rich compositions (trend 2). Intermediate trends can be observed, depending on whether the change in temperature and water content or the melt differentiation is the dominating effect.
|
Further constraints on temperature may be obtained from solvus isotherm curves in the ternary feldspar compositional system. If the positions of the solvus isotherms are known, then the ternary composition of a single feldspar (e.g. sanidine) might be used to determine crystallization temperature, without knowing the composition of the coexisting plagioclase, as long as it is certain that two feldspar phases coexisted and were thermodynamically in equilibrium. Unfortunately, the location of these isotherms varies between models and the uncertainty is relatively large. Nekvasil (1994)
compared solvus isotherm curves in the ternary compositional diagram, calculated using various published thermodynamic models of feldspar solid solution (Ghiorso, 1984; Green & Usdansky, 1986; Nekvasil & Burnham, 1987; Fuhrman & Lindsley, 1988; Lindsley & Nekvasil, 1989; Elkins & Grove, 1990). Sources of uncertainty and variations between these models are because of both the uncertainty on experimental data used to calibrate these curves and the choice of an activity–composition model for the components in the two phases.
Experimental uncertainties are caused either by equilibrium problems, which are known to occur in experiments using crystals as a starting material (Johannes, 1978), or by analytical problems (typically mixed-analyses with glass, as a result of the small crystal size). In the experimental study of Elkins & Grove (1990), variations within one given run are up to 1·8% An, which corresponds to temperature variations of c. 100°C, based on their thermodynamic model. In the experiments of Berndt et al. (2001) on Laacher See samples, some of the sanidine compositions from experiments at 880°C plot near the 1300°C isotherm calculated by Nekvasil (1994), but these analyses have much higher An-contents (up to 20%) than the natural LST sanidines. Analytical and equilibrium problems lead both to an overestimate of the An-content of the equilibrium sanidine composition at a given temperature and thus to an underestimate of the temperature for a given composition. Because the uncertainty in the experiments of Elkins & Grove (1990) seems to be higher at high temperatures than at low temperatures, the temperature intervals inferred from the solvus curves based on these experiments may be also underestimated.
The other source of uncertainty is the choice of a thermodynamic model. Although the various thermodynamic models are based on two different experimental datasets (Seck, 1971; Elkins & Grove, 1990), the results at low temperatures (700°C) are consistent to ±50°C. We use the isotherm curves calculated by Nekvasil (1994), based on the experimental data and thermodynamic model of Elkins & Grove (1990), which give reasonable results for the two-feldspar thermometer in natural samples. At high temperatures, the uncertainty is higher. Two models (Ghiorso, 1984; Nekvasil & Burnam, 1987) would result in lower estimated temperature and smaller temperature ranges, but these two models were discarded by Nekvasil (1994) because they do not predict reasonably the experimental values for An solubility in Or and Or solubility in An. All other models would infer even higher temperatures for any given sanidine composition, so temperature ranges inferred from the model of Elkins & Grove (1990) are minimum estimates.
Even if the equilibrium isotherms are well determined, an additional uncertainty may be caused by the influence of kinetic effects on the composition of feldspars grown from melts in natural systems. The effect of growth rate on An-content in sanidine is not known. Smith & Lofgren (1983) and Muncill & Lasaga (1987, 1988) showed the effect of growth kinetics on the composition of the growing crystal for plagioclase. However, growth in natural magma systems is expected to occur at low degrees of undercooling (Cashman, 1992) and disequilibrium effects are less likely than in experiments. If this assumption is correct, we can infer temperatures from the major element composition of sanidine and the solvus curves. At low temperatures (700°C), the temperature estimates are thought to be accurate to within 50°C for Ab-rich compositions, but only to 100°C in Or-rich composition, because of narrowed isotherms, whereas, at 900°C, the uncertainty could be +100°C. The temperatures ranges inferred from An variations are believed to be minimum values.
Control on trace elements in sanidine
Minor and trace element concentrations in sanidine are related to their concentration in the host melt through equilibrium mineral–melt partition coefficients (Ki for element i). Most partitioning data for sanidine in the literature are for Ba, Sr, Rb and Cs. In plagioclase, partition coefficients for Ba and Sr were shown by Blundy & Wood (1991) to depend mainly on the An-content of the growing crystal and, to a lesser extent, on temperature. Ba and Sr become less compatible with increasing An-content of the plagioclase. This result has, however, been questioned by Ewart & Griffin (1994). In sanidine, Ba and Sr partitioning and its controlling parameters may be even more complicated.
Highly variable empirical partition coefficients were determined for Laacher See sanidine and interpreted as evidence for disequilibrium by Wörner et al. (1983). Other empirical studies (Mahood & Hildreth, 1983; Mahood & Stimac, 1990; Stix & Gorton, 1990; Ewart & Griffin, 1994) report variable Ba partition coefficients in sanidine and invoke the influence of the bulk melt composition. However, there is no evidence in these empirical studies that the phenocrysts are in equilibrium with their host melt. We therefore cannot use empirical data; hence, we base our discussion on experimental data.
Sr and Ba are both compatible in sanidine. Published experimental partitioning data for Ba and Sr in sanidine are highly variable and inconsistent, varying between 1 and >50 for Ba and between 5 and >100 for Sr (Long, 1978; Carron & Lagache, 1980; Guo & Green, 1989; Icenhower & London, 1996). However, values lower than 5 for Ba are reported only at high pressure (15 kbar, Guo & Green, 1989) or for ternary feldspar with low Or-contents (<15 mol %) and An-contents between 15 and 35 mol %, which are significantly different from our compositions (Icenhower & London, 1996). Therefore, we consider partition coefficients for Sr and Ba between sanidine and melt in the Laacher See magma chamber to be 5.
The effects of pressure (P), temperature, water content, melt composition and crystal composition on partition coefficients are difficult to distinguish in the available experimental data because they are not independent. The effect of pressure on KBa, between 10 and 25 kbar, is around 0·08 unit/kbar (Guo & Green, 1989), which is negligible for the Laacher See magma chamber, in which pressure variations are likely to be smaller than 1 kbar. KBa is correlated positively with Or-content in all of the experiments, but to different degrees. In contrast, KSr dependence on Or-content is not clear (Long, 1978; Carron & Lagache, 1980). KBa correlation with temperature appears to be negative (Guo & Green, 1989). Other parameters shown to be correlated with mineral–melt partition coefficients are PH2O (negatively, Guo & Green, 1989) and crystal growth rate (negatively, Long, 1978). There is no agreement on the role of major and minor element composition of the melt.
Fe3+ is known as a possible substitute for Al (Petrov & Hafner, 1988) and the red-CL color indicates the presence of Fe3+ only. The partition coefficient of Fe is thus expected to increase with fO2.
Based on these considerations, we can use trace- and minor-element concentrations in sanidine to reconstruct the evolution of the host–melt compositions during the history of their crystallization.
Interpretation of the different sanidine types in the Laacher See magma chamber
In the following discussion, for each type of sanidine crystals, variations in the ternary compositions of sanidine are used to infer physical changes in crystallization conditions (temperature and water content). Other elements (Ba, Sr, Ti and Fe) provide additional information on the bulk melt composition, based on the available experimental data on mineral–melt partition coefficients, as well as information on possible kinetic effects and oxygen fugacity.
PO-type: repeated temperature changes and continuous compositional evolution
Oscillatory zoning has been mainly studied in plagioclase but has been shown to occur also in other minerals. Two types of models for oscillatory zoning exist (Shore & Fowler, 1996): (1) local kinetic control, involving a diffusion boundary layer at the crystal–melt interface; (2) high-frequency variations in the physical and chemical environment of the growing crystal, such as P, T, composition and water content.
The presence of wavy or rounded resorption surfaces, separating individual growth zones, as seen in crystals 1050b-S3 and 1034-8-S4 (Fig. 7), shows that the crystal underwent periodic destabilization, which underlines that these oscillations are pseudo-oscillatory. Development of a chemical boundary layer at the crystal interface, depleted in the sanidine components, cannot explain the observed PO variations with correlations between all minor and trace elements, especially Ca, which is incompatible in sanidine, with Sr, which is compatible. This clearly indicates that the predominant factors controlling these oscillations are external factors, such as the variations of temperature, melt composition or water content in the magma (model 2) and not local kinetic effects of crystal growth (model 1).
The ULST is believed to be water-undersaturated (Harms & Schmincke, 2000) but the water content increases with differentiation. Each of the parallel trends defined by the PO-type crystals in ternary composition diagrams (Fig. 8a) is typical for changes of crystallization temperature with water content, without other significant changes in the bulk composition (trend 2 in Fig. 13). By contrast, the shift towards Ab-rich compositions between different trends is explained by increasing differentiation of the bulk melt (trend 1 in Fig. 13). The fact that the major element composition within each crystal remains on the same trend over several zones (Fig. 8a) is clear evidence that changes in the bulk composition do not play a significant role in the oscillations. A possible evolution with time of water content, T, An and Or is proposed below and illustrated schematically in Fig. 14. A resorption surface can be explained by local mixing with a less differentiated, hotter, drier melt. Possible causes for such mixing are discussed in the next section. Effects that result from changes in temperature may be seen almost immediately, before those that result from changes in water content, because of the slow diffusivity of water compared with heat conduction. This will facilitate sanidine resorption. Homogenization of melt water content and, possibly, cooling then occur, until sanidine begins to grow again. The slow decrease in Ca-content reflects the return to initial conditions, e.g. if the crystals are transferred again to a cooler, wetter magma.
|
From the amplitude of An variations in the ternary diagram (Fig. 13), temperature variations (associated with variations in magma water content) can be quantified within the limit of the uncertainty on the solvus curves discussed above. We find temperatures for PO-type sanidine ranging between 700°C or slightly below (R-cores) and 1000°C within individual crystals. Temperature variations at resorption surfaces are about 100°C. As discussed earlier, these values depend on which thermodynamic model is most appropriate and are believed to be minimum ranges. Therefore, the results suggest a strong temperature gradient in the lower part of the magma chamber during the crystallization of PO-type sanidine.
We have argued above that PO zoning in major elements does not reflect significant changes in melt composition. Therefore, the increase in trace element concentrations at resorption surfaces (PO variations) should be caused mainly by changes in partition coefficients caused by a coupled rise in T and increase in An-content, and a decrease in H2O and Or. Our observations suggest that the net effect of these changes on Ba, Sr and Ti is an increase in their partition coefficients. In contrast, LW variations in minor and trace elements, especially Ba, are not seen in major elements and thus cannot be explained by variations in temperature and water content because these parameters also control the major element composition of the crystal. Therefore, LW variations in minor elements must be caused either by changes in the bulk melt composition or by kinetic effects. The dependence of KBa (and KSr) on growth rate is not well known, but does not seem to be large (Long, 1978), leaving compositional changes in the melt as the probable cause.
The above discussion suggests that PO variations, seen mainly in An-content variations, reflect correlated changes in T and water content, whereas LW variations, seen mainly in Ba concentrations, document the compositional evolution of the host melt. However, in the magma chamber, gradients in temperature and water content and compositional gradients are associated with hotter, drier and more mafic melts, having higher Ba, Sr and Ti (Wörner & Schmincke, 1984a). Variations in the composition and temperature of the host melt will be recorded by the crystals if sufficient local mixing occurs. In this case, coupled changes in temperature, water content and chemical composition are expected. Thus, changes in minor elements should be correlated with changes in major elements, which is not observed in most PO crystals (except in the most differentiated). The most plausible explanation is that chemical diffusion is much slower than the transport of heat and water in the melt (e.g. Singer et al., 1995). Chemical changes are thus decoupled from changes in temperature and water content and, consequently, the crystal records processes at different time-scales. Mechanical mixing is required for chemical mixing and a probable cause is convection, which, in turn, can explain local temperature variations. The better correlation of smooth Ca and Ba variations in some PO-type crystals, especially in the MLST (crystal 1034-8-S4, Fig. 7), suggests only limited decoupling between thermal and chemical effects and, thus, probably slower changes in growth conditions. This is consistent with the smaller amplitude of the compositional variations in these crystals. To summarize, most PO-type sanidines appear to record in their major and trace element composition the progressive influence of a hotter and more mafic magma, which is more marked in the less differentiated compositions. This is then followed by differentiation dominated by sanidine crystallization.
R-type: core from a low-temperature, wet mafic magma
The composition of red-luminescing resorbed cores in patchy-zoned or oscillatory-zoned sanidine corresponds in the ternary feldspar system to higher water content and lower temperature (600–700°C), which suggests growth from a more differentiated melt than for the PO-type crystals. This is consistent also with the lower Ti, Sr and Ba concentrations. The red-CL color is partly because of higher Fe-content, present as Fe3+, in turn, possibly a result of the higher oxygen fugacity. However, as the R-cores have more than 60 mol % Or, the melt from which they grew has a major element composition close to the melt that is in equilibrium with PO-type crystals. Overgrowths with compositions richer in An and minor elements (oscillatory or not) are characteristic of a hotter, drier and Ba- and Sr-richer melt, similar to that of PO-type crystals.
C-type: differentiated environment
The intergrowth of two feldspar phases that characterizes the C-type crystals can be interpreted either as an exsolution or a growth feature. This is an important issue, because the first case implies a long subsolidus cooling history, and thus that these are xenocrysts, e.g. from the previously solidified magma chamber wall, whereas the second does not. A large range of lamellar and patchy intergrowth textures of two feldspar phases have been produced in growth experiments (Lofgren & Gooley, 1977; Morse & Lofgren, 1978; Petersen & Lofgren, 1986), and these are very similar to the range of textures observed in Laacher C-type feldspars. Conversely, although a few of these textures might be similar to natural microperthites and unequivocal criteria for distinguishing exsolution from composite growth do not exist, most of these textures do not resemble typical perthitic textures described in the literature (e.g. Brown & Parsons, 1988; Waldron et al., 1994; Lee et al., 1995). Therefore, we consider it more likely that the observed composite textures are primary magmatic features. However, if these primary textures are allowed to cool slowly, then they may re-equilibrate compositionally without significant changes in the texture (Petersen & Lofgren, 1986). The compositions of the two phases may therefore reflect temperatures ranging from magmatic to possibly subsolidus, if the crystals experienced a subsolidus history in the already crystallized magma chamber wall.
The interpretation of temperaturess in C-type crystals is unfortunately made difficult by the small size of the albite domains, which implies that many analyses, especially those in the Ab-rich composition, are mixed analyses of both phases and thus inappropriate for geothermometry. For this reason, the compositions of C-type crystals reported in Fig. 13 are only the most extreme compositions and those where the size of albitic patches is believed to be large enough to ensure a pure analysis of the albite phase. The chosen analyses also give a temperature from the sanidine composition, which is consistent with that given by the albite phase, albeit much less precise because of the respective spacing of the isotherms in the various compositional domains.
The two compositional groups of C-type crystals define distinct trends and allow us to characterize the melt from which they formed. These crystals have very low Ba- and Sr-contents, whereas major elements, as well as Fe and Ti, reflect different degrees of differentiation. The low-An group represents a highly differentiated liquid, whereas the high-An group grew in a slightly less differentiated melt. Temperatures of just below 800°C and 650°C are estimated for high-An and low-An C-type crystals, respectively. The estimated temperatures may be overestimated if the analysis of the albite patches was affected by the adjacent sanidine phase. Even in this case, it seems fairly likely that the high-An group represent magmatic temperatures, whereas the low temperatures of the low-An C-type feldspars may reflect subsolidus re-equilibration. However, these crystals do not show the typical morphology and textures of the feldspars from syenite cumulates, which suggests that they still might have grown and remained suspended in the melt at the magma chamber margin. Unfortunately, experimental data on the solidus temperatures of Laacher See magma compositions are lacking. This temperature of 650°C is lower than the temperatures of 720 ± 20 and 750–760°C determined experimentally for LLST by Berndt et al. (2001) and Harms et al. (2001), respectively, and, even with an uncertainty of 50°C, suggests a possible thermal disequilibrium between these crystals and the melt of sample 1002.
Trace element concentrations in sample 1002 also suggest a possible slight chemical disequilibrium between the crystals and the melt. The Ba whole-rock concentration in 1002 is low (20 ppm, Wörner & Schmincke, 1984a). As feldspar crystals are very rare in this pumice and have low Ba-content, the whole rock is considered a good approximation for the Ba concentration in the melt. Therefore, low Ba concentrations in 1002 sanidine, below 90 ppm, imply that the apparent Ba partition coefficient is less than 4·5. This value is lower than typical experimental values (>5) and suggests that the parent melt of these crystals may have had an even lower Ba-content than sample 1002, and thus was still more differentiated. Clear disequilibrium between C-type crystals and MLST melts is also evidenced by the even lower apparent partition coefficient (down to 1·4) in sample 1034, which has 49 ppm Ba in the matrix (Wörner & Schmincke, 1984a), and the strong resorption of C-type sanidine in sample 1050.
Magma chamber dynamics inferred from feldspar zoning
The discussion above identifies the environments in which sanidine phenocrysts grew. We discuss next the possible location of the various growth environments within the magma chamber, as well the relative time of sanidine crystallization for different zoning types. This can, in turn, provide information on exchanges that occur between different parts of the magma chamber and on the dynamic regime of the magma chamber prior to eruption. Summarizing our results, we develop the history of crystallization of the Laacher See Magma chamber, as shown schematically in Fig. 15. We discuss below the processes occurring in the main magma body and at the magma chamber margin, as well as the exchanges between the various parts of the magma chamber.
|
Magma chamber margin
The involvement of a double diffusive crystallizing boundary layer at the magma chamber wall in the formation of magma chamber layering has been suggested by models of liquid fractionation based on analog experiments (Turner, 1980
; Baker & McBirney, 1985; McBirney et al., 1985; Nilson et al., 1985). Tait (1988) and Tait et al. (1989) used observations from cumulates to show that such a model can be applied to the Laacher See magma chamber; the distinct mineral assemblages and compositions in various cumulate xenoliths, better correlated with the composition of interstitial glass than the pumice assemblages, were inferred to represent crystallization at various levels of the magma chamber wall. In their model, the crystallization takes place essentially in a boundary layer, in which volatile-rich, differentiated melt rises along the magma chamber wall, transporting crystals upwards and bringing them into the main magma body.
Our observations are broadly consistent with this model, but allow us to refine it by identifying more precisely the environment of crystal growth. We have not observed sanidine types typical of the Laacher See pumice (sector C-type, PO-type, R-type) in the suite of cognate cumulate xenoliths and carbonatitic syenites. It thus appears that the vast majority of the crystals from the pumice relate directly to the zoned magma chamber, without any significant number of feldspars from cumulates. Therefore, the interpretation of Bourdon et al. (1994) that U–Th isochrons have been disturbed by cumulate-derived phases can probably only apply to the mafic minerals. We have found in the pumice samples only few sanidine crystals that are similar to those found in syenites or cumulates. Old Ar–Ar ages in sanidine (Bogaard, 1995) suggest that at least some of these older crystals are brought into the pumice only shortly (hours, days) prior eruption, because otherwise they would have degassed their Ar.
However, C-type crystals are good candidates for a growth near or at the magma chamber wall. In fact, the slight similarity between some C-type crystals from the pumice and crystals from the syenite cumulate xenoliths may indicate a relatively similar growth environment. We showed in the previous section that C-type crystals grew in an environment that was much more differentiated than the main magma body, and, for low-An C-type, record low temperatures (below 700°C). The growth environment has thus to be chemically fairly isolated from the main magma body. Therefore, the most probable environment would be the crystallizing boundary layer at the magma chamber margin. This is also consistent with the variable chemical and thermal disequilibrium between these crystals and their present host magma. Red-luminescing cores of R-type or PO-type crystals also show low temperature and low Ba- and Sr-content in the equilibrium melt, and may, therefore, also be related to the magma chamber margin. This is consistent with the low temperature and high water content inferred from their major element compositions, and may also explain a possible higher oxygen fugacity, resulting in higher Fe-content in sanidine. The differences between low-An C-type, high-An C-type and R-type cores may be explained by spatial variations within this boundary layer. As suggested by Tait (1988), the most differentiated compositions (low-An C-type) probably relate to the highest level, whereas the higher temperatures and less differentiated composition (high-An C-type) may be found lower at the magma chamber wall, as well as, possibly, at a different distance from the wall. Whereas R-type cores in R- or PO-type from the MLST and the ULST always have a large overgrowth, typical for the main magma body (see below), the small overgrowth on high-An C-type crystals in MLST samples suggests that they have been brought into their present host melt only relatively shortly prior to eruption. This means that C-type and R-type cores sampled the boundary layer not only at different places but probably also at different times.
Because high-An C-type crystals are found in 1002 but hardly any low-An C-type crystals in MLST, the overall movement of crystals from the boundary layer appears to be upwards, as found by Tait (1988) and Tait et al. (1989). The differences between sanidine from cumulates on the one hand and C-type crystals and R-cores on the other hand may be explained by slightly different location during growth, as crystals sampled by the melt are likely to have grown further towards the interior of the magma chamber than those eventually trapped in the cumulates.
Main magma body
As shown in a previous section, PO-type crystals and the blue-luminescing overgrowths of R-type crystals grew in a relatively hot and little differentiated environment undergoing convection. We thus infer this growth environment to be the main magma body. The variations in major element differentiation between crystals of similar size suggest that the magma chamber was already (at least partially) chemically layered at the time of sanidine growth. Convection thus occurred within the individual layers. PO-type crystals have been shown above to record periodic changes in temperature and water contents decoupled from the overall chemical influence of a more mafic magma, followed by a final differentiation and massive sanidine crystallization. The decoupling between physical and chemical parameters is less pronounced in more differentiated compositions. Repeated temperature variations and progressive chemical mixing probably represent two aspects of the same process: mechanical mixing driven by convection in a thermal/chemical gradient. Two models can explain this influence of a more mafic magma.
(1) Model 1: overall movement of the crystal towards the less differentiated bottom of the magma chamber by crystal settling.
(2) Model 2: recharge of the layered phonolitic magma chamber with a hot basanitic magma, triggering the formation of convective plumes within the mafic phonolite layers into the more evolved magma layers situated above. Direct chemical mixing between basanite and phonolite is probably restricted to the latest stages and the lowermost layers of the Laacher See magma chamber.
There is evidence in the LST for feldspar accumulation at the bottom of the magma chamber, both in the trace element bulk compositions of ULST samples, showing enriched Sr and Ba, and in the higher crystal content of the ULST (Wörner & Schmincke, 1984a). However, it is not clear whether this is caused by settling of individual crystals, or by downwards movement of a crystal-laden melt (Tait et al., 1989). We try, below, to estimate the importance of crystal settling. Crystal settling in a magma chamber depends on the Stokes' velocity, i.e. the velocity of the crystal with respect to the host liquid, and the strength of the convection (Marsh & Maxey, 1985; Martin & Nokes, 1988). The crystal Stokes' velocity is given by the formula
 | (1) |
is the density difference between crystal and liquid, r the radius of the crystals, µ the magma viscosity, and g is the acceleration of gravity.
Taking a minimum density of the liquid of 2300 kg/m3 (Wolff et al., 1990), a maximum density of sanidine of 2700 kg/m3 (high-Ba sanidine contains, at most, 2 mol % celsian and only in restricted zones), the maximal density contrast is 400 kg/m3. The maximal radius of crystals is 5 mm, minimum viscosity of the magma 104 poise (Wolff et al., 1990). The maximum value for Stokes' velocity is thus: 4 x 10–5 m/s, the minimum velocity (1 mm, 10 kg/m3, 106 poise) is 4 x 10–9 m/s1, whereas a more probable value is 3·5 x 10–7 m/s. A crystal starting to grow in the middle of the magma chamber thus needs between 0·8 and 8000 years to settle to the base of the magma chamber (about 1 km). This is broadly consistent with the time-scales inferred for the formation of the layering in the Laacher See magma chamber, of less than 20 000–10 000 years by Bourdon et al. (1994) and approximately 3000 years by Tait et al. (1989).
However, as is shown by the An variation, convection occurred in the magma chamber. Convection velocities are poorly known. Singer et al. (1995) calculated laminar convection velocities of 3 x 10–7 m/s. If convection is turbulent, higher but non-uniform velocities would be expected, so that a crystal may remain for a long time in the main magma body but settle eventually, which is consistent with the significantly higher crystal content in the ULST (the base of the magma chamber). In any case, settling velocities seem to be possibly lower than convection velocities and variable from one crystal to another if convection is turbulent. In conclusion, some crystal settling is expected and observed; however, it is not clear how it affects each individual crystal and over what time-scale.
In fact, the settling path of individual crystals may be assessed from their major element chemistry. The ternary composition of crystal 1099-S1 in Fig. 8a changes from trend 2 to 1 (less differentiated) from core to rim, probably documenting settling into a more mafic magma. However, crystal 1050b-S3 remains on trend 2 and shows the same (or larger) Ba LW variations compared with crystal 1099-S1, which suggests that crystal settling alone cannot explain fully the chemical variations seen by the crystals. Furthermore, model 1 implies that the thermal and compositional variations at the crystal–melt interface are a result of convection through stable, existing thermal and chemical gradients in the crystallizing magma chamber. As the Laacher See magma chamber was chemically layered at the time of sanidine growth, the vertical thermal contrast within each layer cannot have been as high as that deduced from the sanidines where variations are commonly more than 100°C. Higher temperature gradients are expected between the cooler boundary layer near the magma chamber walls and its central part. However, movements of crystals occurring only near this boundary layer may not be sufficient to explain the abundance of PO-type sanidine in the main magma body, in which case, large, periodic, lateral movement would be required. Therefore, we believe that although crystal settling does probably occur in the magma chamber, the second model, i.e. heating by a recharge magma, might be necessary to explain all the chemical variations observed.
Model 2 involves the injection of new mafic magma as a source of heat and compositional variations. Phonolite–basanite magma mingling is known to have occurred during or shortly before the Laacher See eruption, as evidenced by the presence of hybrid rocks at the top of the deposit (Wörner & Wright, 1984), and this recharge event probably caused the strong resorption observed at the rims of PO-type crystals. A similar, but distinct, recharge event at an earlier stage of the magma chamber evolution, during which sanidine crystallized, could explain the PO-zoning patterns. The arrival of a basanitic magma at 1080°C (Wörner & Wright, 1984)—much hotter than the mafic phonolite—would produce a temperature contrast at the bottom of the magma chamber and, therefore, provide the necessary driving force for secondary convection and local mixing in the overlying phonolitic magma. Huppert & Sparks (1980) produced such effects in analogue experiments, and Couch et al. (2001) proposed that a similar process, referred to as ‘self mixing’, may explain disequilibrium in natural magmas. Such processes are also the likely explanation for the presence of abundant and partly disintegrated mafic phonolite clots in the higher, more evolved magma layers. However, strong chemical zonation and curved trace element co-variation trends in the deposit (Wörner, 2003) exclude wholesale mixing between the phonolite magma layers and, therefore, limited small plumes or finger-type convection is a more likely scenario. Gradients in temperature and chemical composition then may propagate from layer to layer towards higher levels, where the effect decays. This would well explain the variations observed with stratigraphic position in the deposit and sanidine composition in the PO-zoning patterns, including reduced decoupling between Ba and Ca towards the magma chamber top. After the initial increase in minor and trace element concentrations during this mixing stage, the subsequent decrease represents reduced convection and the decaying effect of the recharge magma, and subsequent dominating differentiation.
Crystal dispersion
As shown in the discussion above, most of the different crystal types record the distinct environments within the magma chamber in which they originated, and which they ‘visited’ during their growth. Together with the diversity of sanidine types and zoning patterns in each pumice sample, this implies fairly extensive exchange of crystals between different parts of the magma chamber. Various mechanisms have been described in natural systems that may lead to crystal dispersion, including the disintegration of crystal wall mush (Kuritani, 1998), small-scale exchanges (Troll & Schmincke, 2002) and replenishment (Tepley et al., 1999). In the present case, we have shown that the crystal mush at the magma chamber wall plays a role, whereas large-scale mixing during replenishment is ruled out by the preserved magma chamber zonation. A possible mechanism for the transfer of crystals from the boundary layer into the main magma body is the rise of magma and its crystals at the cool boundary layer to its level of buoyancy in the zoned magma. Then, the melt spreads horizontally into the interior of the magma body, from where the crystals begin to sink into the hotter magma immediately below. Some crystal exchange within the layered main magma body, both upwards and downwards, is recorded by PO-type crystals, as shown by the occurrence of variously differentiated major elements within each of the samples investigated. These exchanges most probably occur on a small scale at the interface between magma layers. The importance of crystal exchange may also explain partly the chemical and isotopic disequilibrium observed in the LLST (Wörner et al., 1983, 1987), which was attributed to a late-stage (post-crystallization) contamination of the magma.
The most common type of sanidine—PO—did in fact crystallize in the main magma body (except for its R-type cores). However, a minority of crystals from the margins of the magma chamber are also found in the main magma body. This conclusion is similar to that reached by Ginibre et al. (2002b) for an andesitic magma chamber at Parinacota, Chile, in which plagioclase crystals grew both in the main magma body and under the influence of a volatile-rich boundary layer. In the present case, it is important to note that the history of crystals indicates often an early growth in a cooler and more differentiated environment relative to the melt in which they were at the time of eruption. Therefore, one effect of the process of crystal dispersion is to incorporate crystals from the cool crystallizing boundary layer into the main volume of the evolving magma chamber, where they may continue to grow.
|
SUMMARY AND CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Numerous studies have shown the potential of zoning patterns in plagioclase for the reconstruction of magma chamber history. This study shows that sanidine also provides an excellent tool for this purpose. BSE and CL images, combined with quantitative analyses of major, minor and trace elements using the electron microprobe, are useful for documenting the zoning patterns in sanidine.
Provided that two feldspars coexist stably during crystallization, the ternary composition of sanidine can be used as a ‘one-feldspar thermometer’. The evolution of the ternary composition within and between crystals also documents the influence of chemical differentiation within the magma chamber or changes in water content with cooling. Trace and minor element (Ba, Sr and Ti) zoning patterns record the chemical variations in the melt surrounding the growing crystals. These variations are partly decoupled from the temperature variations, reflecting different rates of chemical and thermal diffusion and, thus, recording processes at different time-scales.
The application of these methods to the chemically zoned Laacher See magma chamber allows us to refine the existing models for its structure and dynamics, provide independent constraints on temperature and thermal structure and suggest temperature ranges that are higher than previously reported. Distinct crystal types represent distinct environments within the magma chamber. A volatile-rich, highly differentiated crystallizing boundary layer at the top or wall of the magma chamber crystallized composite crystals at temperatures as low as 700°C, or possibly even lower. On the other hand, pseudo-oscillatory-zoned crystals record convection within the chemical layers of the main magma body, with local temperatures possibly ranging up to 1000°C. This convection, involving large temperature gradients, has probably been caused by the presence of an underlying hot mafic magma, possibly basanitic, similar to that which intruded the magma chamber later, during the eruption.
We have refined the dynamical model of the Laacher See magma chamber proposed by Tait et al. (1989). Most crystals found in the pumice are not, as they assumed, directly derived from the magma chamber wall. Whereas some derive from the cooler but mobile and convective boundary layers along the magma chamber walls, many seem also to have crystallized in the main magma body and to have moved from one layer to another within the zoned magma chamber. The main type (PO-type) of crystals have been shown to have crystallized in the already layered main magma body, and are, therefore, younger than the formation of the layering (103–104 years, Tait et al., 1989). This is consistent with the finding of Bourdon et al. (1994) that ‘phenocrysts’ in pumices have ages identical within error (±1500 years) to the eruption age. This study shows that phenocrysts are not what they are generally assumed to be: crystals formed by cooling in the melt in which they are found. Rather, they record complex processes and many distinct paths and histories of growth.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Supplementary data for this paper are available at Journal of Petrology online.
|
ACKNOWLEDGEMENTS |
|---|
We thank A. M. van den Kerkhof for help with the cathodoluminescence microscope and G. Bergantz for useful discussion. A discussion with H. Nekvasil was decisive for the interpretation of ternary sanidine compositions. Useful comments by B. Singer, F. Tepley, M. Wilson, G. Zellmer and an anonymous reviewer helped to improve the manuscript. This study was part of the DFG-funded project SFB 468.
* Corresponding author: Present address: Department of Earth Sciences, University of Durham, South Road, Durham DH1 3LE, UK. E-mail: Catherine.ginibre{at}durham.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONTHE LAACHER SEE MAGMA...METHODSZONING PATTERNS IN LAACHER...DISCUSSIONSUMMARY AND CONCLUSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Allègre, C. J., Provost, A. & Jaupart, C. (1981). Oscillatory zoning: a pathological case of crystal growth. Nature 294, 223–228.[CrossRef]
Anderson, A. T., Jr (1983). Oscillatory zoning of plagioclases: Nomarski interference contrast microscopy of etched polished sections. American Mineralogist 68, 125–129.[Abstract]
Anderson, A. T., Davis, A. M. & Lu, F. (2000). Evolution of the Bishop Tuff rhyolitic magma based on melt and magnetite inclusions and zoned phenocrysts. Journal of Petrology 41, 449–473.
Baker, B. & McBirney, A. R. (1985). Liquid fractionation. Part III: Geochemistry of zoned magmas and the compositional effect of liquid fractionation. Journal of Volcanology and Geothermal Research 24, 55–81.[CrossRef][Web of Science]
Berndt, J., Holtz, F. & Koepke, J. (2001). Experimental constraints on storage conditions in the chemically zoned phonolitic magma chamber of the Laacher See volcano. Contributions to Mineralogy and Petrology 140, 469–486.[CrossRef][Web of Science]
Blundy, J. D. & Shimizu, N. (1991). Trace element evidence for plagioclase recycling in calc-alkaline magmas. Earth and Planetary Science Letters 102, 178–197.[CrossRef][Web of Science]
Blundy, J. D. & Wood, B. J. (1991). Crystal–chemical controls on the partitioning of Sr and Ba between plagioclase feldspar, silicate melts, and hydrothermal solutions. Geochimica et Cosmochimica Acta 55, 193–209.[CrossRef][Web of Science]
Bogaard, P. v. d. (1995). 40Ar–39Ar ages of sanidine phenocrysts from Laacher See Tephra (12 900 yr BP): chronostratigraphic and petrological significance. Earth and Planetary Science Letters 133, 163–174.[CrossRef][Web of Science]
Bogaard, P. v. d. & Schmincke, H.-U. (1984). The eruptive center of the late Quaternary Laacher See Tephra. Geologische Rundschau 73, 935–982.
Bogaard, P. v. d. & Schmincke, H.-U. (1985). Laacher See Tephra: a widespread isochronous late Quaternary tephra layer in central and Northern Europe. Geological Society of America Bulletin 96, 1554–1571.
Bourdon, B., Zindler, A. & Wörner, G. (1994). Evolution of the Laacher See magma chamber: evidence from SIMS and TIMS measurement of U–Th disequilibria in minerals and glasses. Earth and Planetary Science Letters 126, 75–94.[CrossRef][Web of Science]
Brown, W. L. & Parsons, I. (1988). Zoned ternary feldspars in the Klokken intrusion: exsolution microtextures and mechanisms. Contributions to Mineralogy and Petrology 98, 444–454.[CrossRef][Web of Science]
Carron, J. P. & Lagache, M. (1980). Étude expérimentale du fractionnement des éléments Rb, Cs, Sr, et Ba, entre les feldspaths alcalins, solutions hydrothermales et liquides silicatés dans le système Q.Ab.Or.H2O à 2 kbar entre 700 et 800°C. Bulletin de Minéralogie 103, 571–578.
Cashman, K. V. (1992). Textural constraints on the kinetics of crystallization of igneous rocks. In: Nicholls, J. & Russell, J. K. (eds) Modern Methods of Igneous Petrology, Understanding Magmatic Processes. Mineralogical Society of America, Reviews in Mineralogy 24, 259–314.
Couch, S., Sparks, R. S. J. & Caroll, M. R. (2001). Mineral disequilibrium in lavas explained by convective self-mixing in open magma chambers. Nature 411, 1037–1039.[CrossRef][Medline]
Davidson, J. P. & Tepley, F. J. (1997). Recharge in volcanic system: evidence from isotope profiles of phenocrysts. Science 275, 826–829.
Elkins, L. T. & Grove, T. L. (1990). Ternary feldspar experiments and thermodynamic models. American Mineralogist 75, 544–559.[Abstract]
Eichelberger, J. C., Chertkoff, D. G., Dreher, S. T. & Nye, C. J. (2000). Magmas in collision: rethinking chemical zonation in silicic magmas. Geology 28, 603–606.
Ewart, A. & Griffin, W. L. (1994). Application of proton microprobe data to trace element partitioning in volcanic rocks. Chemical Geology 117, 251–284.[CrossRef][Web of Science]
Finch, A. A. & Klein, J. (1999). The causes and petrological significance of cathodoluminescence emissions from alkali feldspars. Contributions to Mineralogy and Petrology 135, 234–243.[CrossRef][Web of Science]
Fuhrman, M. L. & Lindsley, D. H. (1988). Ternary-feldspar modeling and thermometry. American Mineralogist 73, 201–215.[Abstract]
Ghiorso, M. S. (1984). Activity/composition relation in ternary feldspars. Contributions to Mineralogy and Petrology 87, 282–296.[CrossRef][Web of Science]
Ginibre, C., Kronz, A. & Wörner, G. (2002a). High-resolution quantitative imaging of plagioclase composition using accumulated backscattered electron images: new constraints on oscillatory zoning. Contributions to Mineralogy and Petrology 142, 436–448.[Web of Science]
Ginibre, C., Wörner, G. & Kronz, A. (2002b). Minor and trace element zoning in plagioclase: implications for magma chamber processes at Parinacota Volcano, N. Chile. Contributions to Mineralogy and Petrology 143, 300–315.[Web of Science]
Guo, J. & Green, T. H. (1989). Barium partitioning between alkali feldspar and silicate liquid at high temperature and pressure. Contributions to Mineralogy and Petrology 102, 328–335.[CrossRef][Web of Science]
Green, N. L. & Usdansky, S. I. (1986). Ternary-feldspar mixing relations and thermobarometry. American Mineralogist 71, 1100–1108.[Abstract]
Harms, E. & Schmincke, H.-U. (2000). Volatile composition of the Laacher See Phonolite magma (12 900 yr BP): implications for syneruptive degassing of S, F, Cl, H2O. Contributions to Mineralogy and Petrology 138, 84–98.[CrossRef][Web of Science]
Harms, E., Gardner, J. M. & Schmincke, H.-U. (2001). Preeruptive storage conditions of the highly differentiated phonolitic Laacher See Magma (East Eifel, Germany). Journal of Conference Abstracts 6, 788–789.
Huppert, H. E. & Sparks, R. S. (1980). The fluid dynamics of a basaltic magma chamber replenished by influx of hot dense ultrabasic magma. Contributions to Mineralogy and Petrology 75, 279–289.[CrossRef]
Icenhower, J. & London, D. (1996). Experimental partitioning of Rb, Cs, Sr, and Ba between alkali feldspar and peraluminous melt. American Mineralogist 81, 719–734.[Abstract]
Johannes, W. (1978). Melting of plagioclase in the system Ab–An–H2O at PH2O = 5 kbars, an equilibrium problem. Contributions to Mineralogy and Petrology 66, 295–303.[CrossRef][Web of Science]
Knesel, K. M., Davidson, J. P. & Duffield, W. A. (1999). Evolution of silicic magma through assimilation and subsequent recharge: evidence from Sr isotopes in sanidine phenocrysts, Taylor Creek Rhyolite, NM. Journal of Petrology 40, 773–786.[CrossRef][Web of Science]
Kuritani, T. (1998). Boundary layer crystallization in a basaltic magma chamber: evidence from Rishiri Volcano, Northern Japan. Journal of Petrology 39, 1619–1640.[CrossRef][Web of Science]
Lasaga, A. C. (1982). Toward a master equation in crystal growth. American Journal of Science 282, 1264–1288.
Lee, M. R., Waldron, K. A. & Parsons, I. (1995). Exsolution and alteration microtextures in alkali feldspar phenocrysts from the Shap granite. Mineralogical Magazine 59, 63–78.[Abstract]
L'Heureux, I. & Fowler, A. D. (1994). A nonlinear dynamical model of oscillatory zoning in plagioclase. American Mineralogist 79, 885–891.[Abstract]
Liebsch, H. (1996). Die Genese der Laacher See-Karbonatite. Ph.D. thesis, University of Göttingen.
Lindsley, D. H. & Nekvasil, H. (1989). A ternary feldspar model for all reasons. EOS Transactions, American Geophysical Union 70, 506.
Litt, T., Brauer, A., Goslar, T., Merk, J., Balaga, K., Mueller, H., Ralska-Jasiewiczowa, M., Stebich, M. & Negendank, J. F. W. (2001). Correlation and synchronisation of Lateglacial continental sequences in northern Central Europe based on annually laminated lacustrine sediments. In: Bjorck, S., Lowe, J. J. & Walker, M. J. C. (eds) Integration of Ice Core, Marine and Terrestrial Records of Termination 1 from the North Atlantic Region. Quaternary Science Reviews 20: 1233–1249.
Lofgren, G. E. (1974). An experimental study of plagioclase crystal morphology: isothermal crystallization. American Journal of Science 274, 243–273.[Abstract]
Lofgren, G. E. & Gooley, R. (1977). Simultaneous crystallization of feldspar intergrowths from the melt. American Mineralogist 62, 217–228.[Abstract]
Long, P. E. (1978). Experimental determination of partition coefficients for Rb, Sr, and Ba between alkali feldspar and silicate melt. Geochimica et Cosmochimica Acta 42, 833–846.[CrossRef][Web of Science]
Loomis, T. P. (1982). Numerical simulations of crystallization processes of plagioclase in complex melts: the origin of major and oscillatory zoning in plagioclase. Contributions to Mineralogy and Petrology 81, 219–229.[CrossRef][Web of Science]
Mahood, G. & Hildreth, W. (1983). Large partition coefficients for trace elements in high-silica rhyolite. Geochimica et Cosmochimica Acta 47, 11–30.[CrossRef][Web of Science]
Mahood, G. & Stimac, J. A. (1990). Trace element partitioning in pantellerites and trachytes. Geochimica et Cosmochimica Acta 54, 2257–2276.[CrossRef][Web of Science]
Marsh, B. D. & Maxey, M. R. (1985). On the distribution and separation of crystals in convecting magma. Journal of Volcanology and Geothermal Research 24, 95–150.[CrossRef][Web of Science]
Marshall, D. J. (1988). Cathodoluminescence of Geological Materials. Boston, MA: Unwin Hyman.
Martin, D. & Nokes, R. (1988). Crystal settling in a vigorously convecting magma chamber. Nature 332, 534–536.[CrossRef]
McBirney, A. R., Baker, B. H. & Nilson, R. H. (1985). Liquid fractionation. Part I: Basic principles and experimental simulations. Journal of Volcanology and Geothermal Research 24, 1–24.[CrossRef][Web of Science]
Morse, S. A. & Lofgren, G. E. (1978). Simultaneous crystallization of feldspar intergrowths from the melt: a discussion. American Mineralogist 63, 419–421.[Abstract]
Muncill, G. E. & Lasaga, A. C. (1987). Crystal growth kinetics of plagioclase in igneous systems: one-atmosphere experiments and application of a simplified growth model. American Mineralogist 72, 299–311.[Abstract]
Muncill, G. E. & Lasaga, A. C. (1988). Crystal growth kinetics of plagioclase in igneous systems: isothermal H2O-saturated experiments and extension of a growth model to complex silicate melts. American Mineralogist 73, 982–992.[Abstract]
Nekvasil, H. (1992). Ternary feldspar crystallization in high temperature felsic magmas. American Mineralogist 77, 592–604.[Abstract]
Nekvasil, H. (1994). Ternary feldspar/melt equilibria: a review. In: Parson, I. (ed.) Feldspars and their Reactions. Dordrecht: Kluwer Academic, pp. 195–219.
Nekvasil, H. & Burnham, C. W. (1987). The calculated individual effects of pressure and H2O content on phase equilibria in the granite system. In: Mysen, B. (ed.) Physicochemical Principles. Geochemical Society Special Publication 1, 433–446.
Nilson, R. H., McBirney, A. R. & Baker, B. H. (1985). Liquid fractionation, Part II: Fluid dynamics and quantitative implications for magmatic systems. Journal of Volcanology and Geothermal Research 24, 25–54.[CrossRef][Web of Science]
Nixon, G. T. & Pearce, T. H. (1987). Laser-interferometry study of oscillatory zoning in plagioclase: the record of magma mixing and phenocryst recycling in calc-alkaline magma chambers, Iztaccíhuatl volcano, Mexico. American Mineralogist 72, 1144–1162.[Abstract]
Pearce, T. H. (1987). The theory of zoning patterns in magmatic minerals using olivine as an example. Contributions to Mineralogy and Petrology 97, 451–459.[CrossRef][Web of Science]
Pearce, T. H. & Kolisnik, A. M. (1990). Observation of plagioclase zoning using interference imaging. Earth-Science Reviews 29, 9–26.
Pearce, T. H., Griffin, M. P. & Kolisnik, A. M. (1987). Magmatic crystal stratigraphy and constraints on magma chamber dynamics: laser interference results on individual phenocrysts. Journal of Geophysical Research 92, 13745–13752.[CrossRef]
Perini, G., Tepley, F. J., Davidson, J. P. & Conticelli, S. (2003). The origin of K-feldspar megacrysts hosted in alkaline potassic rocks from central Italy: a track for low-pressure processes in mafic magmas. Lithos 66, 223–240.[CrossRef][Web of Science]
Petersen, J. S. & Lofgren, G. E. (1986). Lamellar and patchy intergrowths in feldspars: experimental crystallization of eutectic silicates. American Mineralogist 71, 343–355.[Abstract]
Petrov, I. & Hafner, S. S. (1988). Location of trace Fe3+ ions in sanidine, KAlSi3O8· American Mineralogist 73, 97–104.[Abstract]
Schmincke, H.-U., Lorenz, V. & Seck, H. A. (1983). Quaternary Eifel volcanic fields. In: Fuchs, K., von Gehlen, K., Maelzer, H., Murawski, H. & Semmel, A. (eds) Plateau Uplift; The Rhenish Shield; A Case History. Berlin: Springer Verlag, pp. 139–151.
Seck, H. A. (1971). Koexistierende Alkalifeldspäte und Plagioklase im System NaAlSi3O8–KAlSi3O–CaAl2Si2O8–H2O bei Temperaturen von 650°C bis 900°C. Neues Jahrbuch für Mineralogie, Abhandlungen 115, 315–345.
Shimizu, N. (1990). The oscillatory trace element zoning of augite phenocrysts. Earth-Science Reviews 29, 27–37.
Shore, M. & Fowler, A. D. (1996). Oscillatory zoning in minerals: a common phenomenon. Canadian Mineralogist 34, 1111–1126.[Web of Science]
Singer, B., Dungan, M. A. & Layne, G. D. (1995). Textures and Sr, Ba, Mg, Fe, K, and Ti compositional profiles in volcanic plagioclase: clues to the dynamics of calc-alkaline magma chambers. American Mineralogist 80, 776–798.[Abstract]
Smith, R. K. & Lofgren, G. E. (1983). An analytical and experimental study of zoning in plagioclase. Lithos 16, 153–168.[CrossRef][Web of Science]
Stix, J. & Gorton, M. P. (1990). Variations in trace elements partitions coefficients in the Cerro Toledo Rhyolite, Jemez Mountains, New Mexico: effects of composition, temperature and volatiles. Geochimica et Cosmochimica Acta 54, 2697–2708.[CrossRef][Web of Science]
Tait, S. R. (1988). Samples from crystallizing boundary layer of a zoned magma chamber. Contributions to Mineralogy and Petrology 100, 470–783.[CrossRef][Web of Science]
Tait, S. R., Wörner, G., Bogaard, P. v. d. & Schmincke, H.-U. (1989). Cumulate nodules as evidence for convective fractionation in a phonolite magma chamber. Journal of Volcanology and Geothermal Research 37, 21–37.[CrossRef][Web of Science]
Tepley, F. J., Davidson, J. P. & Clynne, M. A. (1999). Magmatic interactions as recorded in plagioclase phenocrysts of Chaos Crags, Lassen volcanic center, California. Journal of Petrology 40, 787–806.[CrossRef][Web of Science]
Troll, V. R. & Schmincke, H.-U. (2002). Magma mixing and crustal recycling recorded in ternary feldspar from compositionally zoned peralkaline ignimbrite ‘A’, Gran Canaria, Canary Islands. Journal of Petrology 43, 243–270.
Tsuchiyama, A. (1985). Dissolution kinetics of plagioclase in the melt of the system diopside–albite–anorthite, and origin of dusty plagioclase in andesites. Contributions to Mineralogy and Petrology 89, 1–16.[CrossRef][Web of Science]
Turner, J. S. (1980). A fluid-dynamical model of differentiation and layering in magma chambers. Nature 285, 213–215.[CrossRef]
Waldron, K., Lee, M. R. & Parsons, I. (1994). The microstructures of perthitic alkali feldspars revealed by hydrofluoric acid etching. Contributions to Mineralogy and Petrology 116, 360–364.[CrossRef][Web of Science]
Wolff, J. A., Wörner, G. & Blake, S. (1990). Gradients in physical parameters in zoned felsic magma bodies: implications for evolution and eruptive withdrawal. Journal of Volcanology and Geothermal Research 43, 37–55.[CrossRef][Web of Science]
Wörner, G. (2003). Processes of magma evolution, magmatic suites. In: de Vivo, B., Grasemann, B. & Stüwe, K. (eds) Geology, in Encyclopedia of Life Support Systems (EOLSS), developed under the auspices of the UNESCO. Oxford, UK: Eolss Publishers, http://www.eolss.net.
Wörner, G. & Schmincke, H.-U. (1984a). Mineralogical and chemical zonation of the Laacher See Tephra Sequence (East Eifel, W. Germany). Journal of Petrology 25, 805–835.
Wörner, G. & Schminke, H.-U. (1984b). Petrogenesis of the zoned Laacher See Tephra. Journal of Petrology 25, 836–851.
Wörner, G. & Wright, T. L. (1984). Evidence for magma mixing within the Laacher See magma chamber (East Eifel, Germany). Journal of Volcanology and Geothermal Research 22, 301–327.[CrossRef][Web of Science]
Wörner, G., Beusen, L. M., Duchateau, N., Gijbels, R. & Schmincke, H.-U. (1983). Trace element abundances and mineral/melt distribution coefficients in phonolites from the Laacher See Volcano (Germany). Contributions to Mineralogy and Petrology 84, 152–173.[CrossRef][Web of Science]
Wörner, G., Staudigel, H. & Zindler, A. (1985). Isotopic constraints on open system evolution of the Laacher See magma chamber (Eifel, West Germany). Earth and Planetary Science Letters 75, 37–49.[CrossRef][Web of Science]
Wörner, G., Harmon, R. S. & Hoefs, J. (1987). Stable isotope relations in an open magma system, Laacher See, Eifel (FRG). Contributions to Mineralogy and Petrology 95, 343–349.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  I. Parsons, D. A. Steele, M. R. Lee, and C. W. Magee Titanium as a cathodoluminescence activator in alkali feldspars American Mineralogist, May 1, 2008; 93(5-6): 875 - 879. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. C. Ramos and F. J. Tepley III Inter- and Intracrystalline Isotopic Disequilibria: Techniques and Applications Reviews in Mineralogy and Geochemistry, January 1, 2008; 69(1): 403 - 443. [Full Text] [PDF]
|
|
|
|
|
|
M. J. Streck Mineral Textures and Zoning as Evidence for Open System Processes Reviews in Mineralogy and Geochemistry, January 1, 2008; 69(1): 595 - 622. [Full Text] [PDF]
|
|
|
|
 |
|
 C. Ginibre, G. Worner, and A. Kronz Crystal Zoning as an Archive for Magma Evolution Elements, August 1, 2007; 3(4): 261 - 266. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Lee, I. Parsons, P. R. Edwards, and R. W. Martin Identification of cathodoluminescence activators in zoned alkali feldspars by hyperspectral imaging and electron-probe microanalysis American Mineralogist, February 1, 2007; 92(2-3): 243 - 253. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Kennedy and J. Stix Magmatic processes associated with caldera collapse at Ossipee ring dyke, New Hampshire Geological Society of America Bulletin, January 1, 2007; 119(1-2): 3 - 17. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Schmitt Laacher See revisited: High-spatial-resolution zircon dating indicates rapid formation of a zoned magma chamber Geology, July 1, 2006; 34(7): 597 - 600. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Parsons and M. R. Lee MINERALS ARE NOT JUST CHEMICAL COMPOUNDS Can Mineral, December 1, 2005; 43(6): 1959 - 1992. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. GAGNEVIN, J. S. DALY, G. POLI, and D. MORGAN Microchemical and Sr Isotopic Investigation of Zoned K-feldspar Megacrysts: Insights into the Petrogenesis of a Granitic System and Disequilibrium Crystal Growth J. Petrology, August 1, 2005; 46(8): 1689 - 1724. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||