Textures in Metamorphic Rocks

Grain shapes in metamorpohic rocks depend largely on the crystal structure of the minerals in the assemblage concerned. For aggregate of minerals with structures that are relatively uniform in three dimensions, such as quartz, feldspar, calcite, dolomite, fluorite, scapolite, olivine, galena, pyrite, pyrrhotite, sphalerite, chromite and magnetite, it is common to find the development of a texturallyequilibrated assemblage with adjacent grains having planar boundaries and interfacial angles approximating 120ᵒ. This texture is known as Granoblastic Texture. Such a texture is best developed if the minerals in the assemblage are mostly of similar type. This texture is the result of an attempt to fill 3D space and reduce the interface area to a minimum (Vernon, 2004). In two dimensions, five or six sided grains are most common in an equilibrated assemblage. Larger grains grow at the expense of the smaller grains with few sides, which preogressivelygets smaller, eventually becoming three-sided (in two dimensions), before disappearing completely and thus getting replaced by a new triple junction. Monomineralic igneous assemblage, on slow cooling can develop similar texture. It is often difficult, only by observing a petrographic section and without prior field knowledge, to identify a slowly cooled pyroxene cumulate that has texturally adjusted from a metamorphosed pyroxenite that has been subjected to high grade metamorphism.

Development of equilibrium boundaries may be hampered by presence of other minerals that are strongly anisotropic. For example shapes of mica is strongly dominated by the {001} crystal faces. Thus in a quartz-mica aggregate, the former will often develop straight grain boundaries against the {001} of adjoining mica grains (see fig 6.5.1c). The quartz-quartz interface meets the mica {001}-quartz interface at approximately 90ᵒor more.

In polymineralic aggregates, similar texture may develop but the dihedral angle between grains of different minerals deviate from the ideal 120ᵒof monomineralic aggregates.

If a mineral is considerably larger than the surrounding grains it is called a Porphyroblast. A porphyroblast with well developed crystal faces is an idioblast while one in which only some faces are well developed is a subidioblast. A porphyroblast where none of the grain crystal faces are well formed is a xenoblast. A porphyroblast with inclusions is described as a poikiloblast while too many inclusions in a porphyroblastproduce a sieve texture.

The cause of development of dendritic texture (fig 6.1.5) is debated. Most workers think that these are products of rapid crystal growth under restricted availability of components so that the mineral has grown along whichever direction the required components were available most easily.

Zoning in a mineral may be compositional zoning or growth zoning or both. Compositional zoning is produced during growth of a grain owing to preferential incorporation of certain chemical components in the mineral structure under changing set of physical conditions like temperature, pressure, bulk rock, oxygen fugacity or composition of any fluid permeating the system. Changes in one or more of these conditions are reflected in the composition of the growing mineral through changes in relative preference for chemical components that are incorporated in the crystal structure. In a composition profile prepared from line scan EPMA data, such continuous compositional changes generally show up as smooth curves. Alternately compositional zoning may be produced by exchange of chemical components between the rim of the mineral and adjoining matrix phases due to change in metamorphic conditions like reduction of temperature during retrogression. Chemically zoned grains may also develop through development of overgrowths on existing cores as the process minimizes the energy required for fresh nucleation. In a single episode of metamorphism or in polymetamorphic events different reactions may lead to stabilization of the same mineral, albeit with varying compositions under changing physical conditions. Instead of creating fresh nucleus, the newly formed mineral will tend to overgrow on preexisting grain of same the mineral. Such successive growth zones are often marked by sharp flexures in the composition profile.

Concentric growth zoning, that may or may not be associated with compositional changes in the host porphyroblast, are often optically visible. However, in many minerals like garnet the compositional zones can only be discerned through compositional line profile analyses using EPMA or even better through preparation of X-ray elemental map using arbitrary colours for element concentrations. At times concentric growth zones in a porphyroblastmay be associated with differences in inclusion density or variations in orientation of the inclusions in the growing mineral. Such growth zones with optically identified inclusion pattern variations may or may not be associated with compositional variation. Often differences in growth rate rather than changes in composition can mark growth zones of different inclusion densities.

Patchy zoning in a mineral is also not uncommon. Such patchy zones can form through partial dissolution and overgrowth in response to changing metamorphic conditions.

Oscillatory zoning in metamorphic rocks is less common but not totally unknown. Interested readers may consult Vernon (2004) for a comprehensive coverage of oscillatory zoning in metamorphic minerals.

In Sector Zoning, different sections or sectors of the same grain shows difference in either composition or inclusion patterns. Compositional sector zoning in staurolite may reflect differences of Ti concentrations along different sectors while that in andalusite is possibly due to variable substitution of Fe+3 for Al+3 (Vernon, 2004).

Texturally sector-zoned porphyroblasts form when passive inclusions are preferentially incorporated along specific crystal faces (growth sectors) and or corners between faces during growth (fig 6.2.1h-i). Inclusion free (or poor) areas in texturally sector-zoned porphyroblasts commonly form when porphyroblast growth is accommodated by dissolution and diffusion of material not needed during growth and or by displacement of insoluble matrix material (Passchier and Trouw, 2005; Ferguson et al., 1981). Development of included areas is generally ascribed to either fast growth with insufficient time for dissolution and diffusion of unneeded material, or to preferential adsorption of impurities along certain crystallographic directions (Frondel, 1934; Barker, 1998; Spry, 1969; Shelly, 1993; Vernon, 2004; Passchier and Trouw, 2005). Sector-hourglass zoning involves preferential incorporation of inclusions along a particular set of crystal faces during growth. In three dimensions this type of zoning ultimately results in one or more pyramid-shaped distributions of inclusions. The apexes of these pyramids are joined at the center of the crystal and each pyramid defines a growth sector for a face that incorporates inclusions. Sector-hourglass zoning may have a number of origins that are unique to specific mineral species. The causes of hour glass zoning also varies from one mineral to another. Chloritoid and biotite have crystallographically similar zoning patterns wherein the non{001} sectors are heavily included and {001} sectors are clear or poorly included, but zoning in chloritoid can develop in the absence or presence of strain whereas zoning in biotite requires a strain field characterized by shortening at a high angle to foliation and growth of a crystal with {001} oriented at a high angle to foliation (Camilleri, 2018). Development of clear {001} sectors in biotite reflects progressive growth by syntaxial precipitation in dilating strain shadows whereas clear sectors in chloritoid are likely produced by slower growth with sufficient time for dissolution and diffusion of unneeded material. Development of sector hourglass zoning in staurolite is enigmatic in that the slower growing sectors are typically included rather than clear as they are in chloritoid.

Sector-boundary zoning occurs in chiastolite (andalusite; fig. 6.2.1h-i), garnet, and staurolite (Pennifield and Pratt, 1894; Harker, 1939; Rast 1965; Hollister and Bence, 1967; Spry 1969; Rice and Mitchell, 1991). Porphyroblasts with sector-boundary zoning commonly contain two different types of inclusions designated type 1 and type 2 (Camilleri, 2018). Type 1 are passive inclusions and type 2 are cylindrical quartz intergrowths that are coprecipitated with the porphyroblast (Anderson, 1984; Burton, 1986; Rice and Mitchell, 1991; Mason et al., 2010). Hybrid sector zoning is a combination of sector-hourglass and sector-boundary zoning.

Replacement zoning is also a possibility in metamorphic minerals. Porphyroblasts that grow by partial or complete replacement of another porphyroblast, or by replacement of matrix with layers of varying composition, may develop replacement zoning.

Interested reader may refer to Camilleri (2018) and Vernon (2004) for a discussion on zoning in minerals.

Twinning has been discussed under the section 1: Igneous Rocks. In deformed and metamorphosed rocks, the repeated lamellar twins of plagioclase are often bent. Sector twinning in microcline is also generally considered an effect of deformation, at least in microdomains, while the major rock body may or may not show significant deformation effects. Lamellar twins in chloritoid, sector twins in cordierite, penetrative twin in staurolite and simple twin in a number of minerals are commonly encountered in metamorphic rocks.

Inclusions are often present within metamorphic minerals. While such inclusions may represent early formed phases present in the system, more commonly, inclusions form a part of the reactant assemblage or even the inclusion and the host may haveresulted from the same reaction, with the growth rate of the host phase being greater than that of the inclusion phases. The graphite inclusions in the rim zone of the andalusiteporphyroblastin the figure 6.2.1L is a case of inclusion of preexisting minerals already present in the foliation during growth of andalusiteporphyroblast. Inclusions lead to development of chiastolite structure when passiveinclusions are preferentially incorporated along specific crystalfaces (growth sectors) and or corners between faces duringgrowth (Fig 6.2.1k,L).Similarly the tourmaline inclusion in figure 6.4a has been simply incorporated from the precursor assemblage during growth of the staurolite host. On the other hand,biotite and quartz inclusions in figure 6.4bcan clearly be related to the reactant assemblage from which the porphyroblast developed.

The shape of the included minerals is indicator of structural anisotropy as the interface between the inclusion and its host is also a grain boundary (Vernon, 2004). If both inclusion and host are structurally not too anisotropic, spherical or elliptical shapes of inclusions may result in order to minimize the interfacial energy (fig 6.4d). But if one phase is strongly anisotropic, planar low-energy boundaries appear, these effectively lowering the grain boundary free energy (fig 6.1.2d, fig 6.4b, c & fig 6.6.6h). Even so, corners of inclusions which are sites of high atomic misfit tend to be rounded (fig 6.4b, c). Even in case of strongly anisotropic minerals inclusions, sections approximately parallel to the plane of strong anisotropy (for example the basal section of biotite) tens to develop rounded shapes as all interfaces in such sections have approximately equal energies (Vernon, 2004).

The arrangement of inclusion trails often follows the matrix schistosity as will be elaborated in subsequent sections. However, inclusion shapes and orientations may also be crystallographically controlled as exemplified by the chiastolite structure in andalusite (fig 6.2.1k-L).

This section would be covered in greater details under “Section 7: Deformation”. However a brief introduction here would be relevant in view of the next section which deals with relationship of porphyroblasts to external and internal schistosity.

When metamorphic minerals grow in a stress field they tend to align themselves depending upon the direction of compression. The fabric thus developed is domainal in character due to non-coaxial distribution of strain in rock. Platy or flaky minerals develop a strong preferred alignment. This is called lepidoblastic fabric which anastomose against lensoidal domains of coaxial strain where the minerals that are not strongly anisotropic tend to accumulate. Fabric with prismatic minerals having strong preferred orientation is called nematoblastic fabric. A schistosity can be more or less continuous in phyllosilicate rich rocks or it may be strongly domainal with alternate domains or bands of structurally strongly anisotropic minerals and not so strongly anisotropic minerals. Thus in quartz-mica schist, mica rich domains (M-domains or P-domains) alternate with quartz-rich (Q-domain) or occasionally quartz-feldspar rich domains (QF-domain). Similarly in amphibolites there may be hornblende rich domains alternating with plagioclase rich ones. The development of such domainal character in cleavage is generally attributed to metamorphic differentiation. Experiments suggest that strain partitioning accompanied by removal of soluble minerals from domains of non-coaxial strain plays a major role in metamorphic differentiation. In granulite facies rocks a coarser fabric is generally developed with alternate bands of contrasting mineralogy. This is known as gneissic fabric. Depending upon their crystallographic anisotropy, scarcity of fluid in the system and availability of energy through change in temperature conditions, the minerals in gneissic bands may or may not develop preferred orientation or well developed crystal faces.

Porphyroblasts often contain inclusions. A shape preferred orientation of elongate inclusions defines an inclusion trail. Such inclusion trails can be straight or they may be curved or crenulated. The later is described as helicitic texture. Because alignment of minerals typically results from growth in a tectonic foliation, inclusion trails in porphyroblast and its relation to external foliation, if studied with caution, can potentially allow interpretation of the timing of porphyroblast growth relative to local deformation of a specific foliation. A basic set of microstructural criteria was suggested by Zwart (1960, 1962) using geometrical relationships between Si (internal schistosity or inclusion trail) and Se (external schistosity or matrix foliation). Numerous subsequent publications have shown that while such criteria are broadly applicable, it may not always be straightforward to draw a definite conclusion about timing of porphyroblast growth relative to the sequence of regional deformation events. All parts of a terrain are not heated up to the same degree uniformly at the same time in a tectonic event. Thus even assuming bulk rock chemistry to be homogeneous throughout a terrain, a porphyroblast may develop earlier in the areas that receive the thermal flux early and only subsequently in other areas which heat up later in tectonic history. Moreover deformation font varies and may not be synchronous with heat flux font. Thus porphyroblast of the same mineral may record an earlier deformation stage in one part of the terrain and a much more advanced stage of deformation in another part of terrain. Also effects of local strain heterogeneity should not be underestimated while using porphyroblasts to delineate tectonic history. Porphyroblasts may preferentially grow in coaxial strain domains, act as a rigid body subsequent to growth and lead to strain partitioning or even act as a barrier against which tightened crenulations can develop.

Bell and Rubenach (1983) and subsequently Bell et al. (1986) have proposed several hypothesis regarding nucleation and growth of porphyroblast. The salient feature of their work is that the porphyroblasts develop only during active deformation and nucleate in low strain, largely coaxially deformed domains. Vernon (2004) has discussed each of these hypotheses at length and shown that it is extremely difficult to negate the growth of porphyroblasts in absence of stress. Textural evidences support porphyroblast development both prior to or subsequent to a particular deformation event, although this may not necessarily signify cessation of deformation for the entire terrain. Features like (1) random inclusions that are much smaller than average grain size of same phases in the matrix (fig 6.4d, fig 6.6.2h) and (2) porphyroblasts that have over grown and preserved shapes of former randomly oriented mica grains can be taken as evidences of porphroblast growth prior to the development of the external foliation (Vernon, 2004). However, in most cases, growth history of porphyroblast is complex and requires careful consideration of local strain partitioning.

The greatest amount of controversy based on rotation or non-rotation of porphyroblasts during deformation had been centered around snowball texture. These are normally observed in garnet and sometimes in pyrite porphyroblasts. However staurolite and plagioclase porphyroblasts have also been studied by different groups of workers to resolve the controversy of rotation. The debate gathered momentum when Bell (1985) questioned porphyroblast microstructures that had previously been used as evidence for rotation relative to an externally fixed kinematic reference frame, suggesting instead that they may form by growth, without rotation, during crenulation cleavage development. The non-rotation hypothesis was based on a geometrical strain field analysis designed to mimic the so-called millipede microstructure (porphyroblast preserving inclusion trails of opposite concave microfolds or OCMs at their opposite extremities) preserved in and around plagioclase porphyroblasts described by Bell and Rubenach (1980). Suggestion was made that snowball garnets result from the rotation of the foliation while garnet porphyroblasts are kept fixed, even in non-coaxial flow. In such a model, referred to as the strain-partitioning model (Bell, 1985), garnet successively overprints the new generations of near-orthogonal foliations produced by successive episodes of shortening and extension occurring during porphyroblast growth (e.g. Bell & Rubenach, 1983; Bell & Johnson, 1989; Johnson, 1990; Bell et al., 1992; Aerden, 1995; Johnson & Bell, 1996; Hickey & Bell, 1999; Stallard, 2003). The widely variable trend of inclusion trails in the porphyroblasts were explained by proponents of non-rotation as initial variation of foliation orientation in the rock before the porphyroblast engulfed them. The opposing theory of variable amounts of porphyroblast rotation during deformation and growth was supported by Passchier et al., 1992; Jiang, 2001; Vernon, 2004; Johnson et al., 2006; Johnson, 2009. In a study on geometrical analyses, crystal preferred orientation as well as compositional zoning of snowball garnets, Robyr et al. (2009) has shown rotation of as much as 270ᵒ followed by non-rotational overprinting of foliation by growing garnet porphyroblasts. Johnson (2009) stated that although porphyroblasts rotate with respect to each other during deformation, there are several factors that contribute to relatively minor rotation in many instances, including (1) low strain during and after porphyroblast growth in comparison to mylonitic shear zones; (2) small axial ratios combined with relatively low internal vorticity during growth and post-growth deformation; and (3) strain localization at the porphyroblast-matrix interface. Thus, he concluded that given the right circumstances, porphyroblasts may preserve the approximate orientations of deformation fabrics present at the time of their growth, but each case must be individually assessed. The question of whether snowball garnet rotates or not carries particular significance in structural geology, not only because the two end-member models predict either single-phase or polyphase deformation histories but also they insist on diametrically opposite shear senses. Figure 6.64 is the only sketch that has been introduced in this atlas from Robyr et al (2009) in order to clarify the contradiction generated by the rotation vs non-rotation hypotheses.

Post deformation growth of porphyroblast was argued against by Bell et al. (1986). According to their hypotheses only deformation can lead to conditions conducive to material transport and porphyroblast growth. Even in contact metamorphism, stress generated in microdomains are responsible for development of porphyroblasts. Vernon (2004) have given textural evidences for post deformation porphyroblast development. However cessation of deformation in a domain within a rock does not necessarily imply absence of deformative stress in the entire terrain or even a few meters off. Thus only a complete study encompassing both field and laboratory can decide upon the real condition of a porphyroblast growth.

When an inequigranular rock or a rock having minerals of different rheology is deformed and undergoes grain size refinement, the constituent minerals behave differently according to their rheological properties. Thus while softer phases or finer grained ones are completely recrystallized, others, although marginally recrystallized and thus shape modified, continue to persist as coarser grains within the recryatallized fine garined matrix. These minerals are called porphyroclasts as they are larger than the surrounding finer matrix materials. A porphyroclast may be of any origin. It might be the remnant of a coarse phenocryst from an igneous rock that has undergone grain size refinement through deformation and recrystallization or it may be a metamorphic porphyroblast that has, during a subsequent deformation, undergone a process similar to the one described above.

When porphyroblasts of high aspect ratio develops in absence of any regional or local stress, they tend to be randomly oriented. Such a texture is known as hornfelsic texture. Care should be taken to study an outcrop in three dimensions before naming a texture as truly hornfelsic as it might happen in plain strain domain that while the longest axes of porphyroblasts are parallelly oriented, the two shorter axes in another dimension shows random orientations.

Depletion halo sometimes develop around growing porphyroblast due to complete consumption of any reactant and presence of excess of another. Effectively the porphyroblast is separated from the surrounding matrix by a rim of excess reactants. Genetically the depletion haloes are quite different from pressure shadows and should not be confused with the later.

All metamorphic minerals in a microdomain may or may not belong to the same assemblage. Due to decrease in temperature leading to a dearth of energy as well as lack of fluid in the system during retrograde metamorphism of high grade terrains, the metamorphic reactions are often not brought to completion. This leads to metastable persistence of one or more of the reactant phases in the system. Such metastable persistence of reactants are useful in delineating metamorphic reactions.

The product phases of a metamorphic reaction may nucleate and form independent grains in the matrix, may overgrow on preexisting grains of similar mineral already present in the system or may partially to completely replace a reactant phase. Figures 6.6.6b-d demonstrates how the porphyroblasts have replaced the M-domains of schistosity leaving the Q-domains that pass continuously from matrix into the porphyroblast as inclusion trails. At times a product phase or an aggregate of grains of the product phase can so replace a reactant that the shape of the reactant is preserved. This happens normally in strain free domains. Such completely replaced reactant without any distortion of its shape is called pseudomorph. In most cases, however, replacement is only partial.

Exsolution occurs when a homogeneous solid solution becomes unstable and breaks down into two minerals (Vernon, 2004). Exsolutions are common in both igneous and metamorphic rocks and occurs generally with decreasing temperature. Some common examples of exsolution textures in metamorphic rocks include perthite, antiperthite, exsolution lamellae in pyroxene, lamellae of hercynite in magnetite or vice versa, exsolution lamellae of rutile in corundum, of ilmenite in magnetite or of sulphides in each other. Exsolution textures can be used to deduce cooling temperature through appropriate applications of geothermometry. Exsolution of clinopyroxene in majorite garnet has tectonic significance.

Relict igneous or sedimentary textures can at times be easily identified in metamorphic rocks, particularly in low grade metamorphic terrains.

Symplectites are vermicular intergrowth of minerals that grow simultaneously in a solid state reaction. These textures are mostly found in metamorphic rocks subjected to retrogression and thereby decrease of temperature. However symplectites developed during prograde metamorphism have also been reported. Similarly de Haas et al (2002) have mentioned symplectite development in the presence of small amounts of residual magma in cooling igneous rocks, such as olivine plagioclase symplectite in some gabbros. A special type of symplectite is kelyphitic rim where garnet porphyroblast in mantle peridotite that is subjected to rapid exhumation, reacts with adjoining olivine and develop a radiating symplectite of orthopyroxene, clinopyroxe and spinel (grains often in topotactic relation) around the garnet.

The greatest significance of symplectic textures is the fact that due to their development under retrograde conditions, these reactions are often incomplete with portions of reactant left in the system. Thus they are excellent reaction indicators. Development of symplectites are generally attributed to low diffusion rate of at least for one of the minerals involved, so that elongate and somewhat dendritic grain shapes are favoured than more equant grains (Vernon, 2004). As a certain volume of one phase grows through breakdown of the reactants, it expels components from the reactants that are taken up by the other product phases, so that two or more phases grow simultaneously. However simple breakdown of reactants may not necessarily fully explain the symplectite texture. Mass balance calculations suggest that major or trace component elements of symplectite phases cannot often be explained through breakdown of reactants only and ion transport over some distances in the rock may be necessary.

Symplectites commonly occur as lobes projecting into the grain being replaced or at times form a rim of symplectite called symplectic corona (fig 6.12.1q, 6.12.1v and 6.13.1f). Although the symplectite appears as fine grained vermicular intergrowths, large sections are often in optical continuity. The size of the intergrowth phases generally decrease from base to the projecting tip of symplectic lobe suggesting decrease in temperature and therefore decreasing diffusion rate during retrograde metamorphism. Fig 6.12f offers such an example where, although the reactant garnet has been completely pseudomorphed by the symplectite assemblage, a decrease in grain size of the advancing lobes from garnet rim to garnet core is easily observed.

Symplectites grow in low strain localities, as evidenced by random arrangement of delicate shapes, which could not survive in high strain environments. They become distorted and recrystallize into granular aggregates under strain.

A common variety of symplectite is myrmekite – a vermicular intergrowth of quartz and sodic plagioclase replacing potash feldspar. Myrmekite typically occur as lobes projecting into the potash feldspar grains from adjoining plagioclase. The feldspar in myrmekite is often in optical continuity with the adjoining plagioclase on which it has nucleated. Rarely myrmekite can nucleate on quartz grains as well.

The mechanism of myrmekitization is not simple replacement. It involves supply of Na and Ca as well as removal of K from the site. The potassium is generally incorporated in muscovite which results in muscovitization of adjoining plagioclase grains. The Na and Ca, along with SiO2 released in the process, are used up in development of the myrmekitic phases. Volume of quartz in myrmekite is related to the calcic component in the replacing plagioclase - the more calcic the plagioclase, the less SiO2 it accommodates and the more the proportion of quartz in the myrmekite. Thus complex reactions and ion transfer in presence or absence of fluid is involved in myrmekite development processes.

Myrmekites are generally associated with deformed granitoids suggesting some sort of connection between deformation and myrmekite development. Possibly deformation facilitate fluid permeation and therefore ion transfer to and from the site of replacement.

A corona is the rim or mantle of product phase developed on the reactant in course of a metamorphic reaction. Corona texture is common in retrograde metamorphism when decreasing temperature do not allow completion of the reactions. In other words, corona is a reaction frozen in time. Often more than one corona may form on a reactant, the sequence depends on the relative mobility of cations from reactant to product. Often coronae themselves are constituted of symplectic intergrowths. These are known as symplectic corona. Although common in metamorphic rocks, corona may develop in igneous rocks as a result of incomplete reactions between early crystallized phase and evolved late magma in the system.