| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Petrology 45(7) © Oxford University Press 2004; all rights reserved
Towards a Better Understanding of the Fibrolite Problem: the Effect of Reaction Overstepping and Surface Energy Anisotropy
1 DEPARTMENT OF MINERALOGY AND PETROLOGY, UNIVERSITY OF PADOVA, C.SO GARIBALDI 37, 35137 PADOVA, ITALY
2 IST. GEOSCIENZE GEORISORSE, CNR, C.SO GARIBALDI 37, 35137 PADOVA, ITALY
RECEIVED DECEMBER 1, 2002; ACCEPTED FEBRUARY 2, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGTHE XENOLITHS IN TRACHYTES...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Trachytes from the Euganean Hills District (Italy) contain metapelitic xenoliths that have been pyrometamorphosed during incorporation in the melt. In xenoliths containing sillimanite crystallized during a previous regional HT/LP metamorphism, fibrolite systematically nucleates at the grain boundaries of sillimanite prisms and within plagioclase crystals. Ternary feldspar thermometry shows that plagioclase in contact with sillimanite plots along the 750°C solvus that reflects near-equilibrium conditions of regional metamorphism. Plagioclase containing fibrolite plots closer towards the 950°C solvus, reflecting the tendency of plagioclase to re-equilibrate at high temperature during pyrometamorphism by a fibrolite-forming reaction:
 |
KEY WORDS: andalusite breakdown; fibrolite; pelitic xenoliths; reaction overstepping; sillimanite
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGTHE XENOLITHS IN TRACHYTES...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Fibrolite is the fine-grained, acicular high-temperature variety of Al2SiO5 and almost invariably forms in metapelites when the sillimanite stability field is approached, either from the kyanite or from the andalusite field. There is a continuum between acicular fibrolite and prismatic sillimanite, so that a clear-cut distinction between them cannot be made on the basis of grain size or habit. However, if potential contribution of grain boundary energy to total free energy is considered (Holdaway, 1971
; Kerrick, 1990; Hemingway et al., 1991), then crystals with a diameter of less than 2 µm should be regarded as fibrolite rather than sillimanite (Kerrick, 1990; Pattison, 1992). Herein we adhere to this distinction. In this definition, it is implicit that fibrolite has a higher molar free energy than coarse-grained sillimanite. Thus, the common observation that fibrolite appears earlier than sillimanite conflicts with equilibrium thermodynamics; consequently the fibrolite ‘problem’ has been debated over recent decades.
Metastable formation of fibrolite is a possible explanation. Hints for metastability come from garnet–biotite geothermometry (Fleming, 1973; Kerrick, 1987), which show that calculated KD values indicative for fibrolite formation occur outside the sillimanite stability field. Epitaxial growth lowering the energy of nucleation for fibrolite is an alternative explanation that is suggested by the widespread observation that fibrolite frequently shows a regular, triangular arrangement pattern within basal sections of biotite (Chinner, 1961; Yardley, 1977). Kerrick (1987) has suggested that fibrolite may inherit the trigonal arrangement of Si and Al chains in the tetrahedral layers of the biotite structure, so that fibrolite nucleation is favoured over andalusite or kyanite nucleation. Therefore, when dehydration reactions affecting biotite are involved in forming the Al2SiO5 polymorph reactant at high temperature, a potentially lower nucleation energy for fibrolite may explain its crystallization instead of sillimanite growth.
However, fibrolite is also found to occur across extremely sharp kyanite–sillimanite isograds (Grambling, 1981; Grambling & Williams, 1985) where there is microstructural evidence of kyanite being replaced by fibrolite. In this case, we are dealing with a real paramorphic transformation.
The above examples clearly show that fibrolite systematically forms before sillimanite. Therefore, the contribution of the surface energy of fibrolite to the total free energy must have no influence on the earlier stability of fibrolite vs sillimanite. At the same time, it is not reasonable to claim that the consistent appearance of fibrolite before sillimanite is explained in every case by metastable or epitaxial growth. It is, therefore, more likely that fibrolite formation is driven by kinetic factors and to understand these we need to focus on examples where such factors can be constrained. This is possible only where fibrolite formation can be linked to specific changes of the physical parameters controlling the metamorphism. Such an opportunity is given by metapelites that have undergone pyrometamorphism when they were included as xenoliths in trachytic magma of the Euganean Hills (NE Italy). The results of our investigations indicate a mechanism of fibrolite formation that implies an initially accelerated growth rate as a result of reaction overstepping of a mineral that is characterized by a strongly anisotropic surface energy.
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGTHE XENOLITHS IN TRACHYTES...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
The Euganean Hills District belongs to the wider Venetian Volcanic Province (Fig. 1a), which covers an area of
2000 km2 in NE Italy. The most representative rock types of the Euganean Hills District are Late Eocene to Oligocene trachytes, rhyolites, latites and basalts (Borsi et al., 1969). Geochemical (De Vecchi & Sedea, 1974; De Vecchi et al., 1976; and references quoted therein) and geophysical data (Giese & Buness, 1992) are consistent with an extensional geodynamic context (Fig. 1b). This agrees with the widespread occurrence of numerous extensional tectonic faults with NNE–SSW and NW–SE direction (Fig. 1a) related to the volcanic activity (Piccoli et al., 1981; Barbieri et al., 1991; Castellarin et al., 1992; Zampieri, 1995).
|
Geophysical data indicate the Moho at 30 km depth underneath the Euganean Hills District, giving a lower constraint to the possible source depth for the xenoliths. On the other hand, crystalline basement has been found at 4·7 km in boreholes under the Venice lagoon (Meli & Sassi, 2003
), giving an upper constraint. |
THE XENOLITHS IN TRACHYTES OF THE EUGANEAN HILLS AND THEIR PETROGRAPHIC FEATURES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGTHE XENOLITHS IN TRACHYTES...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Oligocene trachytes from the Euganean Hills include various types of regionally metamorphosed gneissic xenoliths. Currently, age constraints for the regional high-T–low-P metamorphism are not available, but this could be related to magma production in the Euganean Hills District.
Some of the xenoliths have a pelitic composition, and preserve replacement microstructures that indicate a complex metamorphic history involving Al2SiO5 phase transformations. In addition, they also preserve microstructural and petrological evidence that may help to understand one of the long-lasting problems in metamorphic petrology: what drives fibrolite formation?
The xenoliths considered in this paper are slightly elongated and of a few centimetres in length. Two main types are recognized: type (i), sillimanite-bearing andalusite gneiss, consisting of sillimanite, fibrolite, andalusite, corundum, alkali feldspars, plagioclase, biotite, green spinel, magnetite and ilmenite (samples MM182, MM188 and MM189); type (ii), sillimanite-free andalusite–cordierite gneiss, consisting of andalusite, corundum, alkali feldspars, biotite, green spinel, magnetite, ilmenite and cordierite (samples MM29, CZ12 and CZ16).
The gneissic structure of the xenoliths is mainly defined by the orientation of Al2SiO5 crystals. Grain size is highly variable, because of the presence of large andalusite, sillimanite and alkali feldspar (up to 1·5 mm) and small grains of green spinel, magnetite and ilmenite (<50 µm). In some xenoliths both recrystallized and glassy melt pools occur. In detail, several microstructural domains are present within each of the two main xenolith types.
Type (i)—sillimanite-bearing andalusite gneiss
- The occurrence of large, ‘isolated’, inclusion-free porphyroblasts of andalusite is a distinctive microstructural feature of this rock type. The porphyroblasts display rounded edges and are often embayed (Fig. 2a). They are always located within alkali feldspar-rich, plagioclase-bearing, biotite-, spinel- and opaque-free domains.
- Large prismatic sillimanite crystals (up to 3 mm in size), defining the main foliation of the rock, occur together with K-feldspar and plagioclase, and are surrounded by biotite-free, magnetite-, ilmenite- and spinel-rich coronas (Fig. 2b). Crystal boundaries of the sillimanite prisms are invariably irregular, indicating that a sillimanite-consuming reaction has occurred.
- Fibrolite systematically nucleates at the grain boundaries or along cleavage planes of sillimanite prisms (Fig. 2b and c), within plagioclase (Fig. 2d), and extends into the surrounding matrix. Fibrolite is never associated with andalusite.
- Small corundum idioblasts occur within ‘isolated’ andalusite and sillimanite prisms (Fig. 2e) and, more rarely, in the matrix.
- Biotite-rich domains are sillimanite-free, and are always isolated from sillimanite prisms. Biotite is transformed to varying extents to an aggregate of K-feldspar, green spinel, and ilmenite (Fig. 2f).
- Large magnetite crystals show exsolution lamellae of ilmenite and spinel.
|
Type (ii)—sillimanite-free andalusite–cordierite gneiss
- This rock type is characterized by bands of oriented andalusite idioblasts alternating with bands of K-feldspar and cordierite. Both bands are characterized by inclusion trails of spinel and ilmenite that define a former foliation. The inclusion trails are more common within the K-feldspar–cordierite bands, suggesting that they have replaced former biotite-rich bands, whereas the andalusite-rich bands represent former muscovite-rich bands (Fig. 3a).
- In some cases, andalusite appears to be in equilibrium with cordierite and K-feldspar as shown by stumpy granular boundaries (MM29); in other cases it is clearly replaced by K-feldspar (CZ12), indicating that an andalusite-consuming reaction boundary has been crossed.
- Locally, relics of biotite occur, although biotite is largely transformed to an aggregate of K-feldspar, green spinel, magnetite and ilmenite.
- Small corundum idioblasts are included within K-feldspar (CZ12). Their orientation is parallel to that of spinel and ilmenite inclusion trails, suggesting a formation related to the breakdown of a former foliation-forming phase such as muscovite. Locally, corundum porphyroblasts occur that include inclusion trails of spinel (Fig. 3b).
- Corundum also occurs as large blue idioblasts, and is clearly growing at the expense of andalusite (Fig. 3c).
- In some cases (CZ16) andalusite porphyroblasts are extensively replaced by a fine- to very fine-grained aggregate of acicular corundum in a matrix of K-feldspar (Fig. 3d–f).
|
|
MINERAL CHEMISTRY |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGTHE XENOLITHS IN TRACHYTES...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Microprobe analyses were carried out using a Cameca Camebax electron microprobe (IGG-CNR–Department of Mineralogy and Petrology, University of Padova). Operating conditions were: electron beam current 10 nA; accelerating voltage 15 kV; acquisition time 10 s for each element and 5 s on background; beam radius 1 µm. Natural and synthetic standards were used, and the PAP correction procedure was followed. Analytical measurements are affected by a relative uncertainty of 1% for the major elements (>5 wt %) and 4% for the minor elements (<5 wt %). Representative analyses of the main mineral phases are listed in Table 1.
|
Ilmenite
Figure 4 shows the composition of ilmenite (triangles) in the system TiO2–FeO–Fe2O3. With respect to the ideal composition, ilmenites from microdomains containing green spinel and biotite (open triangles, sample MM189) have moderate amounts of Fe3+ (0·13–0·15 a.p.f.u., normalized to three oxygens), whereas ilmenites from K-feldspar-rich domains (filled triangles, sample CZ12) have nearly ideal stoichiometry (Fe3+ = 0·015–0·023 a.p.f.u.). Ilmenite also occurs as exsolved lamellae within magnetite idioblasts (Fig. 5), but the extremely small size of the lamellae precludes electron microprobe analysis.
|
, ilmenite from green spinel- and biotite-bearing microdomains (sample MM189); 
, ilmenite in K-feldspar-rich domains (sample CZ12);
|
Magnetite
Magnetite forms large idioblasts only within sample MM189. Figure 4 shows their composition (open squares). The data points fall within the titanomagnetite field, along the ulvöspinel–magnetite tie-line, but close to the ideal magnetite composition. The magnetites have a moderate Ti content ranging from 0·06 to 0·08 a.p.f.u. (analyses normalized to three cations), corresponding to 2·1–2·9 wt % TiO2. Mg ranges from 0·05 to 0·5 a.p.f.u., corresponding to 0·8–1·4 wt % MgO.
Figure 5 (sample MM189) shows ilmenite and pleonaste spinel lamellae exsolved on {111} planes in titanomagnetite. According to Buddington & Lindsley (1964), ‘trellis-type’ exsolution such as that shown in Fig. 5 could have formed by ‘oxidation–exsolution’ or ‘oxyexsolution’.
Green spinel
Green spinel occurs in different microstructural domains as pleonaste (Fig. 6). Computed Fe3+ is always low, in the range 0·07–0·29 a.p.f.u. (analyses normalized to three cations). XMg [Mg/(Mg + Fe2+)] ranges from 0·27 to 0·55. Higher XMg values are from grains within or near biotite, whereas those with lower XMg are associated with corundum or cordierite. Spinel from sillimanite-rich domains has variable XMg from 0·27 to 0·55.
|
, pleonaste associated with corundum or cordierite; The Mg/Fe ratios in coexisting spinel and biotite have been used to estimate the temperature of biotite decomposition according to Brearley (1987)
. The KD [(Mg/Fe)Spl/(Mg/Fe)Bt] values of samples CZ12 and MM189 range respectively from 0·27 to 0·29 and from 0·29 to 0·37, indicating temperatures from 815 to 845°C.
Feldspar
Figure 7 shows the compositions of feldspar occurring in the different microstructural domains. Feldspar after biotite has XOr 57–71%, XAb 28–41% and XAn 1–3%. Away from biotite, feldspars are ternary solid solutions, specifically having a more Ab-rich plagioclase composition (XOr = 13–24%, XAb = 61–70%, XAn = 18–19%). In leucocratic domains around andalusite relics (Fig. 2) both plagioclase (XOr = 4–9%, XAb = 55–66%, XAn = 25–41%) and alkali-feldspar (XOr = 66–69%, XAb = 31–33%, XAn = 0·4–0·7%) occur. Plagioclase domains in which sillimanite prisms crystallized have XOr 4–5%, XAb 56–57% and XAn 37–38%, whereas feldspar within which fibrolite nucleates is more alkali-rich (XOr 13–19%, XAb 61–70%, XAn 15–19%). It should be noted that plagioclase domains in contact with andalusite and sillimanite plot along the 750°C ternary feldspar solvus [calibration of Fuhrman & Lindsley (1988)], whereas plagioclase compositions in contact with fibrolite plot between the 850°C and 950°C solvi lines.
|
Alkali-feldspars occur also around melt inclusions and within cordierite, having respectively XOr 60–61%, XAb 37–40%, XAn 0·12–0·16% and XOr 72–74%, XAb 25–27%, XAn 0·7–0·15%.
Corundum
Corundum is homogeneous and contains small amounts of Fe3+ (0·02 a.p.f.u. analyses normalized to three oxygens) within blue-coloured patches.
Aluminium silicates
Andalusite has different Fe2O3 contents within the two xenolith types (see Petrographic Features section). In the sillimanite-free xenoliths, the Fe2O3 content of andalusite is very low, ranging from 0·30 to 0·60 wt %, whereas in the sillimanite–andalusite xenoliths, the pink andalusite cores have Fe2O3 values of 2·7–2·9 wt %, and colourless rims have 1·7–2·0 wt %. Sillimanite and fibrolite Fe2O3 contents are similar, ranging from 0·3 to 1·2 wt %. The Fe2O3 content decreases from the andalusite pink-coloured cores towards the colourless rims; this might reflect either a change of redox conditions during the growth of the andalusite rim or the lower Fe availability owing to the contemporaneous crystallization of an Fe-rich mineral, such as biotite.
Biotite
Biotite compositions show a moderate deviation from the ideal binary trioctahedral phlogopite–annite K2(Fe,Mg)6(Al2Si6)O20(OH)4 series (Fig. 8a). No significant compositional differences occur from core to rim. Therefore, the different composition of biotite may reflect the micro-chemical composition differences of the nucleation sites. TiO2 content is very high, ranging from a minimum of 4·40 to a maximum of 8·74 wt % (0·47–0·97 a.p.f.u; analyses normalized to 22 oxygens), with most of the values (65% of the analyses) being higher than 6·40 wt %. Titanium does not show any variation with Si content. XMg values are relatively constant, ranging from 0·66 to 0·72 (Fig. 8b). The biotites have a moderate Na content ranging from 0·1 to 0·2 a.p.f.u. corresponding to 0·31–1·0 wt % Na2O.
|
Figure 8c shows the VIAl deficiency and IVAl excess with respect to the tschermakite substitution [IVSi, VI(Fe,Mg)
VIAl, IVAl]. IVAl excess could imply a different substitution, i.e. an Fe-Tschermak substitution [VI(Fe,Mg), IVSi = VI(Fe3+) + IVAl]. The Ti content can be explained by a Ti-Tschermak [VI(Fe, Mg) + 2IV(Si4+) VI(Ti4+) + 2IV(Al)] and a Ti-vacancy [2VI(Fe,Mg) VI(Ti4+) + VI(
)] substitution. Figure 8d shows that the biotite compositions plot within a narrow band parallel to the ideal Ti-vacancy substitution (arrow). Deviation of biotite composition from the ideal Ti-vacancy substitution line may be due to contemporary substitutions that promote Ti entry into octahedral sites, such as [2VIAl VITi + VI(Fe,Mg)] and [VIAl + IVSi VITi + IVAl] (Dymek, 1983). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGTHE XENOLITHS IN TRACHYTES...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Metamorphic reaction sequences
Xenoliths MM189 and CZ12 display most of the petrographic features described above and were selected as representative samples for thermodynamic modelling. Sample MM189 is an And–Sil–Pl–Kf–Crn–Bt–Spl–Mag–Ilm assemblage; sample CZ12 is an And–Kf–Crn–Bt–Spl–Crd–Mag–Ilm assemblage, i.e. it does not contain Sil and Pl, but has Crd as an additional phase.
P–T pseudosections of the KNFMASCH (Fig. 9) and KFMASH (Fig. 10) systems have been calculated by means of the Vertex software package (Connolly, 1990), on the basis of the bulk-rock compositions of samples MM189 and CZ12. In calculating the pseudosections, bulk chemical compositions were obtained by scanning electron microscopy, scanning over a specific area of the thin section, in which the previously mentioned mineral assemblages had been determined optically and by scanning electron microscopy. The bulk composition for MM189 is (in wt %): SiO2 53·28; Al2O3 27·70, FeO 4·54, MgO 3·16, CaO 1·58, K2O 5·35 and Na2O 4·39, and for CZ12 is SiO2 45·34, Al2O3 36·07, FeO 6·68, MgO 2·47 and K2O 9·45.
|
|
The different mineral assemblages and microstructures of the two samples indicate different reaction sequences in rocks of similar bulk composition. The presence of both in the same trachyte indicates that the two reaction sequences relate to metamorphic P–T paths experienced at different crustal levels during regional high-T–low-P metamorphism, prior to their incorporation in the trachytic magma. Thereafter the xenoliths were heated and pyrometamorphosed by the magma so that their respective regional metamorphic parageneses were overprinted.
Regional high-T–low-P metamorphism
Sillimanite-bearing xenoliths (sample MM189)
These xenoliths are characterized by the presence of andalusite and sillimanite, indicating that a set of reactions producing Al2SiO5 occurred in both the andalusite and sillimanite stability fields. The last of these reactions probably relates to the breakdown of the muscovite + quartz + plagioclase assemblage to produce sillimanite + K-feldspar + melt. This constrains the P–T path to cross the Ms + Qtz breakdown curve (b in Fig. 9) within the sillimanite stability field but at P lower than the Al2SiO5 invariant point.
The absence of quartz and the abundance of sillimanite and K-feldspar in the rock indicates that quartz content was relatively low in the protolith. Therefore, when quartz was consumed in the above reaction, further muscovite decomposition could occur only at higher temperatures (i.e. 630°C; c in Fig. 9). Andalusite, which is always located within K-feldspar-rich plagioclase-bearing biotite and spinel-free domains, metastably persisted within the sillimanite stability field. However, it is invariably embayed, indicating that it was resorbed by reaction with a surrounding melt at high temperature. Sillimanite preferentially nucleated in the muscovite-rich domains rather than replacing andalusite, confirming the sluggishness of this polymorphic transformation. The formation of melt and the contemporaneous crystallization of sillimanite in adjacent areas probably made the melt corrosive to the metastable andalusite. Although a general orientation of sillimanite prisms is evident, they have irregular corroded margins indicating that, during prograde metamorphism, a sillimanite-consuming reaction occurred. Prismatic sillimanite, with small corundum inclusions, is located within rectangular K-feldspar-rich, plagioclase-bearing and biotite-free domains that probably represent former sillimanite idioblasts rimmed by a K-feldspar-bearing, biotite-free corona, with abundant spinel and ilmenite. Outside these domains, biotite occurs. This microstructural situation indicates that the regional metamorphism reached peak conditions within the Al2SiO5 + biotite + K-feldspar + spinel + corundum + ilmenite trivariant field (i.e. between d and e in Fig. 9). At these regional metamorphic peak conditions, melt-producing reactions were stopped by dehydration, and the melt pools crystallized producing a K-feldspar, plagioclase and, locally, biotite aggregate.
Sillimanite-free xenoliths (sample CZ12)
Another group of rocks are characterized by the assemblage And–Kf–Crd–Spl–Crn–Bt–Mag–Ilm. In these rocks also Al2SiO5 reactions occurred prior to their incorporation in the trachytic magma as xenoliths. The presence of nearly mono-mineralic andalusite bands suggests that they represent former muscovite-rich, biotite-bearing layers, where inclusion trails of spinel and ilmenite mark the orientation of a foliation. The andalusite indicates that the muscovite + quartz and the muscovite breakdown reactions (respectively b and c in Fig. 10) occurred in the andalusite stability field, and therefore at pressures lower than those indicated by the intersection of these reaction curves with that of the andalusite 
sillimanite transition (a in Fig. 10). Spinel and ilmenite inclusion trails are also present within polygonal blasts of K-feldspar and cordierite, indicating that their nucleation was probably controlled by the presence of biotite. Small elongated idioblasts of corundum along the foliation and the absence of quartz in these xenoliths indicate a quartz-poor protolith and that all available quartz was consumed by the muscovite + quartz reaction. Higher-temperature breakdown of muscovite to form corundum according to the reaction muscovite corundum + K-feldspar + vapour is indicated by the small elongated corundum idioblasts in inclusion trails and corundum porphyroblasts within K-feldspar. With a further rise of temperature (650°C), the andalusite + biotite assemblage reacted to form K-feldspar + cordierite + spinel + ilmenite + corundum (d in Fig. 10). The extent of andalusite decomposition depends on the amount of biotite present. Where biotite is abundant, andalusite is nearly completely transformed into the higher-temperature mineral assemblage, whereas when biotite is scarce, andalusite is only partially replaced by corundum.
Pyrometamorphism
Sillimanite-bearing xenoliths: the formation of fibrolite (sample MM189)
The incorporation of the xenoliths in the trachytic melt caused an instantaneous rise of temperature. Ternary feldspar thermometry (Fuhrman & Lindsley, 1988) implies temperatures in the interval 750–950°C (Fig. 7). However, our thermodynamic modelling suggests that temperatures probably did not exceed that of the Opx-in reaction (f in Fig. 9), i.e. approximately 840°C, as Opx has been never observed in the xenoliths. The range of temperature estimates reflects the tendency of alkali-feldspar and plagioclase to re-equilibrate at higher temperatures in response to the increased solubility along the feldspar solvus (Fig. 6), through a reaction such as K-feldspar (1) + plagioclase (1) K-feldspar (2) + plagioclase (2) + fibrolite. This reaction has been balanced and calculations are shown in Table 2. The reaction is confirmed by the common growth of fibrolite within plagioclase in the presence of K-feldspar (Fig. 2d). Fibrolite also nucleates on sillimanite. At the same time, biotite reacts to form K-feldspar + spinel + ilmenite, although the formation of fibrolite is not related to this reaction, because it never occurs in contact with biotite. Spinel–biotite thermometry (Brearley, 1987) indicates temperatures in the interval 815–845°C for this reaction. We interpret that the growth of fibrolite is related to significant overstepping of the above-mentioned reaction involving feldspars. The extent of overstepping may have been as much as 150°C, which is the difference between the regional high-T–low-P metamorphic peak and the pyrometamorphic conditions. In the context of the biotite breakdown reaction a second melting event occurred. The small melt pockets that formed can be recognized as small glass pools.
|
Sillimanite-free xenoliths: the formation of acicular corundum (sample CZ12)
In these xenoliths, fibrolite did not form, the main reason being the absence of plagioclase, so that the reaction K-feldspar (1) + plagioclase (1)
K-feldspar (2) + plagioclase (2) + fibrolite could not occur. Although the strong temperature rise during the pyrometamorphism probably resulted in a significant overstepping of the andalusite stability field, fibrolite also did not form, confirming the sluggishness of this polymorphic transformation. Instead, andalusite prisms are replaced by oriented tiny fibrous corundum needles and K-feldspar.
Fibrolite growth as an effect of reaction overstepping and anisotropic surface energy
The stability field of fibrolite relative to that of sillimanite is a topic that continues to be controversial. The main reason for this is that, because of the potential contribution of grain boundary energy to the total free energy (Holdaway, 1971; Kerrick, 1990; Hemingway et al., 1991), it is predicted that fibrolite should be stabilized at higher temperatures than sillimanite (Kerrick, 1990; Pattison, 1992). With the exception of a few cases (e.g. Ahmad & Wilson, 1982; this study), it is observed that fibrolite formation systematically precedes sillimanite growth. This suggests that kinetic factors, rather than epitaxy or metastability, control fibrolite growth.
When surface energy is considered as a contributor to the overall Gibbs' free energy of a newly forming mineral assemblage, then it has to be balanced against other sources of energy acting in the system, and it has to be considered how these influence the reaction kinetics. Heterogeneous reaction kinetics are directly correlated to 
G (Fyfe et al., 1958; Fisher & Lasaga, 1981), so that the net rate (r) of a heterogeneous reaction can be expressed as (Fisher & Lasaga, 1981)
 | (1) |
G is the difference in Gibbs' free energy of the system in the actual state and in the equilibrium state. Equation (1) shows that for any forward reaction, the term involving G yields a progressive increase in r with rising T above the equilibrium boundary.
Because the effect of surface energy on chemical potentials is very small in comparison with modest temperature overstepping (Wheeler, 1991), surface energy will be an important factor only when T is close to equilibrium. Therefore, as long as a reaction temperature has been overstepped, surface energy cannot dominate the system.
It should also be mentioned that acicular growth of a mineral does not in itself imply high surface energy. The surface energy of a crystal is dependent on the orientation of a crystal surface relative to its lattice (Wheeler, 1991). Therefore, if some faces have markedly lower energies than others, these will be the most developed.
Another aspect of surface energy is that its contribution to the total free energy may be reduced by changing the shape of a growing crystal, because surface energy is proportional to the surface area. However, for the growth of larger crystals at the expense of long and thin ones, surface energy must become a major driving force. This will be the case only when fast reaction progress as a result of overstepping has brought the system back to equilibrium.
We therefore interpret that, when the reaction rate is high because of overstepping, sillimanite forms strongly elongate crystals (fibrolite) because the energy of the {110} faces is low (Vernon, 1976) and their growth rate is high. In this respect, fibrolite will always form when a sillimanite-forming reaction has been significantly overstepped, regardless of whether this occurs at the lower temperatures of regional metamorphism or at the higher temperatures equivalent to the central part of the sillimanite stability field, as in the pyrometamorphosed sillimanite- and fibrolite-bearing xenoliths. The same seems to hold also for the growth of acicular corundum pseudomorphing andalusite prisms, in the sillimanite-free xenoliths.
|
ACKNOWLEDGEMENTS |
|---|
The authors are greatly indebted to R. H. Grapes and C. V. Guidotti for the thorough review of the manuscript. F. P. Sassi is acknowledged for fruitful discussions and for his critical reading of an earlier version of the manuscript. Financial and analytical support have been provided by the Istituto di Geoscienze e Georisorse (IGG-CNR).
|
FOOTNOTES |
|---|
* Corresponding author. Telephone: +39-049-8272019. Fax: +39-049-8272010. E-mail: raffaele.sassi{at}unipd.it.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGTHE XENOLITHS IN TRACHYTES...MINERAL CHEMISTRYDISCUSSIONREFERENCES |
|---|
Ahmad, R. & Wilson, C. J. L. (1982). Microstructural relationships of sillimanite and fibrolite at Broken Hill, Australia. Lithos 15, 49–58.[CrossRef][Web of Science]
Barbieri, G., De Zanche, V. & Sedea, R. (1991). Vulcanismo Paleogenico ed evoluzione del semigraben Alpone-Agno. Rendiconti della Società Geologica Italiana 14, 5–12.
Borsi, S., Ferrara, G. & Piccoli, G. (1969). Determinazione col metodo del K/Ar dell'età delle eruzioni euganee. Rendiconti della Società Italiana di Mineralogia e Petrologia 25, 27–34.
Brearley, A. J. (1987). A natural example of the disequilibrium breakdown of biotite at high temperature: TEM observations and comparison with experimental kinetic data. Mineralogical Magazine 51, 93–106.
Buddington, A. F. & Lindsley, D. H. (1964). Iron–titanium oxide minerals and synthetic equivalents. Journal of Petrology 5, 310–357.
Castellarin, A., Cantelli, L., Fesce, A. M., Mercier, J. L., Picotti, V., Pini, G. A., Prosser, G. & Selli, L. (1992). Alpine compressional tectonics in the Southern Alps. Relationships with the N-Apennines. Annales Tectonicae 6, 62–94.
Chinner, G. A. (1961). The origin of sillimanite in Glen Clova, Angus. Journal of Petrology 2, 312–323.
Connolly, J. A. D. (1990). Multivariable phase diagram: an algorithm based on generalized thermodynamics. American Journal of Science 290, 666–718.
De Vecchi, G. P. & Sedea, R. (1974). Sui basalti Eocenici dei Colli Euganei. Memorie degli Istituti di Geologia e Mineralogia dell'Università di Padova 31, 3–26.
De Vecchi, G. P., Gregagnin, A. & Piccirillo, E. M. (1976). Aspetti del vulcanismo terziario veneto. Memorie degli Istituti di Geologia e Mineralogia dell'Università di Padova 30, 3–32.
Dymek, R. F. (1983). Titanium, aluminium and interlayer cation substitutions in biotite from high-grade gneisses, west Greenland. American Mineralogist 68, 880–899.[Abstract]
Fisher, G. W. & Lasaga, A. C. (1981). Irreversible thermodynamics in petrology. In: Lasaga, A. C. & Kirkpatrick, R. J. (eds) Kinetics of Geochemical Processes. Mineralogical Society of America, Reviews in Mineralogy 8, 171–209.
Fleming, P. D. (1973). Mg–Fe distribution between coexisting garnet and biotite, and the status of fibrolite in the andalusite zone–staurolite zone of the Mt. Lofty Ranges, South Australia. Geological Magazine 109, 477–482.[Abstract]
Fuhrman, M. L. & Lindsley, D. H. (1988). Ternary-feldspar modelling and thermometry. American Mineralogist 73, 201–215.[Abstract]
Fyfe, W. S., Turner, F. J. & Verhoogen, J. (1958). Metamorphic Reactions and Metamorphic Facies. Geological Society of America, Memoirs 73, 259 pp.
Giese, P. & Buness, H. (1992). Moho depth, Atlas Map 2. Deep crustal roots of the Alps and the Northern Apennines. In: Blundell, D., Freeman, R. & Müller, S. (eds) A Continent Revealed: the European Geotraverse. Cambridge: Cambridge University Press, pp. 11–13.
Grambling, J. A. (1981). Kyanite, andalusite, sillimanite, and related mineral assemblages in the Truchas Peaks region, New Mexico. American Mineralogist 66, 702–722.[Abstract]
Grambling, J. A. & Williams, M. L. (1985). The effects of Fe3+ and Mn3+ on aluminum silicate phase relations in North–Central New Mexico, U.S.A. Journal of Petrology 26, 324–354.
Hemingway, B. S., Robie, R. A., Evans, H. T., Jr & Kerrick, D. M. (1991). Heat capacities and entropies of sillimanite, fibrolite, andalusite, kyanite, and quartz and the Al2SiO5 phase diagram. American Mineralogist 76, 1597–1613.[Abstract]
Holdaway, M. J. (1971). Stability of andalusite and the aluminum silicate phase diagram. American Journal of Science 271, 97–131.[Abstract]
Kerrick, D. M. (1987). Fibrolite in contact aureoles of Donegal, Ireland. American Mineralogist 72, 240–254.[Abstract]
Kerrick, D. M. (ed.) (1990). The Al2SiO5 Polymorphs. Mineralogical Society of America, Reviews in Mineralogy 22, 406 pp.
Kretz, R. (1983). Symbols for rock-forming minerals. American Mineralogist 68, 277–279.[Abstract]
Meli, S. & Sassi, R. (2003). Petrography, age and geochemistry of the buried ‘Venice Granodiorite’ (Northern Italy). Schweizerische Mineralogische und Petrographische Mitteilungen 83, 57–61.[Web of Science]
Milani, L., Beccaluva, L. & Coltorti, M. (1999). Petrogenesis and evolution of the Euganean Magmatic Complex, Veneto Region, North-East Italy. European Journal of Mineralogy 11, 379–399.
Pattison, D. R. M. (1992). Stability of andalusite and sillimanite and the Al2SiO5 triple point: constraints from the Ballachulish Aureole, Scotland. Journal of Geology 100, 423–446.[Web of Science]
Piccoli, G., Sedea, R., Bellati, R., Di Lallo, E., Medizza, F., Girardi, A., De Pieri, R., De Vecchi, G. P., Gregnagnin, A., Piccirillo, E. M., Norinelli, A. & Dal Pra', A. (1981). Note illustrative della carta geologica dei Colli Euganei alla scala 1:25000. II Edizione. Memorie di Scienze Geologiche 34, 523–566.
Vernon, R. H. (1976). Metamorphic Processes, Reactions and Microstructure Development. New York: Wiley, 247 pp.
Wheeler, J. (1991). A view of texture dynamics. Terra Nova 3, 123–136.[Web of Science]
Yardley, B. W. D. (1977). The nature and significance of the mechanism of sillimanite growth in the Connemara schists, Ireland. Contributions to Mineralogy and Petrology 65, 53–58.[CrossRef][Web of Science]
Zampieri, D. (1995). Tertiary extension in the southern Trento Platform, southern Alps, Italy. Tectonics 14, 645–657.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Sassi, G. Cruciani, C. Mazzoli, L. Nodari, and J. Craven Multiple titanium substitutions in biotites from high-grade metapelitic xenoliths (Euganean Hills, Italy): Complete crystal chemistry and appraisal of petrologic control American Mineralogist, February 1, 2008; 93(2-3): 339 - 350. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||