Classification of magmatic rocks
Scientists have traditionally sought regularity, order, and predictability in their investigations of the natural world. However, for the petrologist, the continuity of rock compositions, the seemingly endless variety of fabric, and the wide range of geologic environments in which rocks form pose formidable obstacles to erecting a well ordered, simple, single rock classification. Unlike in the plant and animal kingdoms, which have discrete species, no such natural divisions exist in rocks. Rocks are more like complex, highly variable biological ecosystems; minerals constituting a rock are like the plant and animal species constituting an ecosystem. Despite the obstacles, a consistent classification of rocks is essential for communication with other petrologists, who should all speak the same language of clasification; a particular rock name should convey thesame meaning to every petrologist, regardless of his or her native tongue. In addition, classification serves as an important means of systematizing information.
Through appropriate and relevant classification, meaningful patterns in composition, fabric, field relations, and, therefore, origin can be perceived. As all classifications of rocks are the fruits of the human mind attempting to erect discrete subdivisions where none exists in the natural, uninterrupted continuum of rock properties, every classification is, to some degree, arbitrary and imperfect. There are many different criteria for classification; consequently, many different labels exist for the very same rock. Each has its own benefit and use; none cancombine the merits of all. “A rock may be given one name on the ground of field occurrence and from hand lens examination, only to require another when it is studied in thin section, and perhaps a third when it is chemically analyzed.
Classification Based on Fabric
We review here only the most fundamental rock terminology based on fabric as it is generally introduced in a beginning geology course. Magmatic fabric is essentially governed by time-dependent (kinetic) processes in the solidifying magma, such as its rate of heat loss, or cooling. Four principal types of fabric occur in magmatic rocks: phaneritic, aphanitic, glassy, and volcaniclastic. The first two refer to the dominant crystal grain size, which ranges over several orders of magnitude, rom 10 -6 to 10 m.
Phaneritic applies to rocks that have mineral grains sufficiently large to be identifiable by eye (minute accessory minerals excepted). This texture is typical of rocks crystallized from slowly cooled intrusions of magma. Aphanitic rocks have mineral grains too small to be identifiable by eye and require a microscope or some other laboratory device for accurate identification. Aphanitic texture is most common in rapidly solidified extruded magma but can also be found in marginal parts of magma intrusions emplaced in the cool shallow crust. Some magmatic rocks contain essentially two grain-size populations and few of intermediate size; such texture is said to be porphyritic. The larger grains are phenocrysts, and the smaller constitute the groundmass, or matrix. Porphyritic aphanitic rocks are far more common than porphyritic phaneritic rocks. Glassy, or vitric, rocks contain variable proportions of glass, in contrast to holocrystalline rocks made entirely of crystals. A vitrophyre is a porphyritic rock that contains scattered phenocrysts in a glassy matrix.
The fabric of volcaniclastic rocks is produced by any fragmenting process that creates broken pieces of volcanic rock and/or mineral grains. Classification of volcaniclasts parallels that of sedimentary clasts according to their particle size, as follows:
<2 mm 2–64 mm >64 mm
volcaniclasts ash lapilli block, bomb
sedimentary clay, granule, cobble,
clasts silt, pebble boulder
sand
Consolidation of volcaniclasts produces volcaniclastic rock types that are classified according to their particle size.
Classification Based on Field Relations
The location where magma was emplaced provides a basis for rock classification. Some petrologists recognize three categories for rocks solidified from magmas emplaced onto the surface of the Earth (volcanic or extrusive), into the shallow crust (intrusive hypabyssal), and into the deep crust (intrusive plutonic). The first and the last categories are readily distinguished on the basis of their field relations but less directly on the basis of their grain size, degree of crystallinity (proportion of crystals to glass), and mineralogical composition.
Magmas emplaced onto the surface of the Earth as coherent lava flows or as fragmental deposits form extrusive, or volcanic rocks. These rocks are typically aphanitic and glassy. Many are porphyritic. Some have fragmental (volcaniclastic) fabric. High-T disordered feldspars are common, so that alkali feldspar, where present, is a clear sanidine. Other minerals that occur only at high-T and low-P in volcanic environments including leucite, tridymite, and cristobalite are found in some volcanic rocks. Amphiboles and biotite, especially where they occur as phenocrysts, are commonly partially altered to fine-grained anhydrous aggregates of Fe oxides, pyroxenes, and feldspars. Phenocrysts of feldspar and quartz commonly contain inclusions of glass.
Intrusive, or plutonic, rocks form where magma was intruded into preexisting rock beneath the surface of the Earth as intrusions, or plutons. Plutonic rocks are typically phaneritic. Monomineralic rocks composed only of plagioclase, or olivine, or pyroxene are well known but rare. Amphiboles and biotite are commonly partially altered, usually to chlorite. Some granites contain muscovite, which is exceedingly rare in volcanic rocks. Perthite an intergrowth of sodic and potassic feldspar is widespread and reflects slow cooling and exsolution in initially homogeneous alkali feldspar.
Characteristics of intermediate-depth hypabyssal rocks are not clearly distinct from those of volcanic and plutonic rocks. Many occur in shallow crustal dikes, sills, and plugs that represent feeding conduits for surface extrusions of magma. But dikes and sills are also intruded deep in the crust. Hypabyssal rocks can have fabric similar to that of plutonic and volcanic rocks. Because of these ambiguities, many petrologists tend to categorize magmatic rocks in the field simply as plutonic or volcanic.
Classification Based on Mineralogical and Modal Composition
Mineralogical Mnemonics
Felsic is a mnemonic adjective derived from the words feldspar and silica. It is a useful appellation for rocks that contain large proportions of feldspar with or without quartz and/or its polymorphs, tridymite and cristobalite. Granite and rhyolite made mostly of feldspar and quartz are examples of felsic rocks. The term felsic also applies to rocks containing abundant feldspathoids, such as nepheline, and to these rock-forming minerals as well. Mafic is a mnemonic adjective derived from the words magnesium and ferrous/ferric. Mafic is a less cumbersome term than the synonymous ferromagnesian. It refers to major rock-forming biotite, amphibole, pyroxene, olivine, and Fe-Ti oxide solid solutions as well as rocks that contain large proportions of them, such as basalt. Ultramafic rocks are especially rich in Mg and Fe and generally have little or no feldspar; an example is the olivine-pyroxene rock called peridotite. Silicic rocks contain large concentrations of silica, manifested by an abundance of alkali feldspar, quartz, or glass rich in SiO2. Examples are rhyolite and granite. The term sialic is used less frequently for rocks rich in Si and Al that contain abundant feldspar and is used especially with reference to the continental crust.
Colour is usually the first rock property noticed by the novice. However, a particular rock type can possess a wide range of colors; granites, as just one example, can be nearly white, shades of gray, green, red, and brown. These widely ranging colors reflect equally widely variable colors of the dominant rock-forming feldspars, whose pigmentation is a complex function of minute mineral inclusions, exsolution, and small concentrations of elements such as Fe in solid solution; none of these factors may be petrologically very significant and in any case may be difficult to determine. Color is not a valid basis of rock classification and can, in fact, be highly misleading. Color index has been defined as the modal proportion of dark-colored minerals in a rock. But, in view of the fact that dominant rockforming feldspars can be light- to dark-colored, a more accurate index should be defined on the basis of the proportion of mafic minerals. Leucocratic and melanocratic rocks can be defined as having 0–30% and 60–100% modal mafic minerals, respectively.
Rock Types
The classification of magmatic rocks most familiar to the beginning geology student is that of rock types. In contrast to the broadly defined compositional labels just described, a rock type has a arrowly defined composition and a particular fabric. Familiar rock types include rhyolite, andesite, and basalt (all aphanitic) and granite and diorite (both phaneritic). Many rock-type labels have a long and bscure history stemming from miners’ jargon; many are coined from geographic locales, such as andesite from the Andes Mountains of western South America. About 800 igneous rock-type names are listed in the classic four-volume work of Johannsen (1931–1938), written toward the end of an era when petrology was mostly descriptive petrography and the coining of new rock names was in vogue. Today, most of these names have, fortunately, been abandoned and petrologists need have only a working knowledge of a few dozen major igneous rock-type names.
Regrettably, however, few of these major names have had consistent usage among petrologists. One petrologist’s andesite has been another’s basalt. Personal biases and backgrounds have been strong factors in schemes of classification. If rock compositions were clustered into isolated clumps on any variation diagram it would be a simple matter to draw a line around each cluster and append a rock-type name to it. However, compositions are not clustered but consist of a continuum. There are at least two approaches to nomenclature within this continuum:
- Flexible, loosely defined limits could be defined, leaving the details to the individual petrologist guided by the circumstances and need at hand. However, this approach has over the decades resulted in considerable confusion in the geologic literature.
- The continuous spectrum could be subdivided along specific, well-defined limits that follow as closely as possible a usage agreed upon by as many petrologists as possible. This is the approach of the International Union of Geological Sciences Subcommission on the Systematics of Igneous Rocks, hereafter referred to as the IUGS. The IUGS system of classification is a universal standard that can eliminate individual biases and contradictions among petrologists.
The IUGS rock-type classification for phaneritic (generally plutonic) rocks, which consist mostly ( 10 modal % but usually more) of felsic minerals. Quartz-rich felsic rocks are also classified; these are collectively referred to as granitic rocks or granitoids. A porphyritic aphanitic to finely phaneritic rock having abundant phenocrysts and occurring in a pluton (intrusion) is called porphyry; depending on its modal composition it may be a granite porphyry, granodiorite porphyry, or other. Uniformly fine-grained phaneritic, very leucocratic granites composed almost entirely of feldspar and quartz that typically occur in thin dikes within a coarser-grained, somewhat more mafic granitic pluton are aplite. Commonly associated with aplite are equally leucrocratic rocks called pegmatite; these are phaneritic rocks of highly variable grain size in which individual crystals are several centimeters to several meters.
Classification Based on Whole-Rock Chemical Composition
There are many advantages of a numerical chemical classification. Insights are provided regarding the nature, origin, and evolution of magmas. Rigorous comparisons can be made between members of suites of rocks and petrotectonic associations. The advantages of chemical classifications are obvious for very finegrained rocks, whose mineralogical compositions may be difficult to determine, and certainly for glassy rocks (but beware of loss of Na and other possibly mobile elements). Aphanitic and glassy volcanic rocks can correspond more closely to the composition of the magma from which they formed than do porphyritic and phaneritic rocks, which may have been derived from magmas that experienced crystal accumulation during their evolution. Magmatic rocks whose characterizing minerals have been obliterated by alteration or metamorphism can be analyzed to reveal their original nature, provided diagnostic chemical elements have not been significantly mobilized during recrystallization. However, an inherent weakness of purely chemical classifications is they have little or nothing to say regarding the effects of geologic processes on fabric and of different P T conditions that govern mineralogical composition.
Aphanitic and Glassy Rock Types
A rigorously quantitative chemical classification of aphanitic and glassy, usually volcanic, rocks must be tempered by the fact that most rock-type names were established decades, and in some instances centuries, ago, when few if any chemical analyses were available and names were based upon mineralogical and modal compositions. All analyses were sorted as to rock-type label, such as “andesite” and “dacite,” irrespective of the classification scheme used.
 |
| IUGS classification of aphanitic and glassy volcanic rock types. |
Overlap between the two fields of andesite and dacite reflects inherent variability in their composition an attribute of all rock types no matter how defined. Nonetheless, averages of these two rock types are quite different. Average compositions of common magmatic rock types, which represent the opinions of thousands of petrologists over many decades. All of the common volcanic rock-type names were so examined and bounding lines drawn on a total alkalies-silica diagram in such a way as to recognize a “consensus” composition. Rock samples to be classified should be as fresh as possible (unweathered and unaltered). Analyses must be recalculated to 100% volatile-free before plotting.
A rock of basaltic composition in which the grain size is marginally phaneritic and transitional into gabbro is diabase (alternatively called dolerite by United Kingdom geologists). Diabase commonly occurs in dikes and sills but also constitutes local lava flows. An olivine-rich basalt or picrobasalt having MgO > 18 wt.% is called picrite if (Na2O + K2O) = 1- 3 wt.% and komatiite if (Na2O + K2O) <1 wt.% and TiO2 is low, generally <1 wt.%. Komatiites are commonly ultramafic and composed essentially of olivine and pyroxene so that they are chemically a peridotite, but their glassy to aphanitic texture precludes use of this phaneritic name.
The chemical classification can be appended to fabric heteromorphs that solidified from chemically similar magmas but have different fabrics. For example, chemically defined rhyolite can be, depending on fabric, rhyolite tuff, rhyolite breccia, rhyolite obsidian (wholly glass), rhyolite vitrophyre, and rhyolite pumice (vesicular glass).
A preliminary IUGS classification for volcanic rocks based upon modal proportions of phenocrysts may be used in the field and before chemical analyses are available. This classification should never be final because the groundmass of porphyritic aphanitic or glassy rocks will always be poorer in plagioclase than the assemblage of phenocrysts because of the way magmas crystallize. A rock containing sparse plagioclase as the dominant or sole phenocryst could be a dacite, rhyolite, or trachyte.
Absolute Concentration of Silica
Except for the very rare carbonatites, silica (SiO2) is the principal oxide constituent of magmatic rocks and serves as a basis for broadly defined classifications. Some petrologists use a classification based on silica concentration in the rock analysis, as follows:
Silica concentration (wt.%) Name
66 acid
52 to 66 intermediate
45 to 52 basic
45 or less ultrabasic
As defined here, acid and basic have no reference whatsoever to hydrogen ion content, or pH, as used in chemistry. (Long ago it was erroneously believed that SiO2 occurred as silicic acid and metallic oxide components, such as CaO and FeO, as bases in magmas.) These four categories have no direct correlation with modal quantity of quartz in the rock, although as a general rule, acid rocks do contain quartz and ultrabasic ones do not. Two rocks having identical concentrations of silica may have widely different quantities of quartz,and two rocks of similar quartz content may have different silica concentrations, depending upon the composition and quantity of other minerals in the rock. Roughly speaking, acid rocks are silicic, basic are mafic, and ultrabasic are ultramafic.
The CIPW Normative Composition
Near the beginning of the 20th century, three petrologists and a geochemist devised an elegant procedure (from whose surnames the acronym CIPW is formed) for calculating the chemical composition of a rock into a hypothetical assemblage of water-free, standard minerals. These standard normative minerals are designated in italics, such as Q , An, Ol, to distinguish them from the actual rock-forming minerals in the rock. Normative minerals are some of the simple end members of the complex solid solutions the actual minerals in the rock comprise. A complex solid solution, such as hornblende, is represented by several simpler normative minerals.
What are the benefits of the normative calculation? Because of extensive solid solution in the major rock forming minerals, substantial variations in whole-rock chemical composition may not be evident in any obvious variations in mineralogical or modal composition. Basaltic rocks are an example. Much the same assemblage of plagioclase, clinopyroxene, olivine, and Fe-Ti oxides can constitute basalt, trachybasalt, and basanite. The norm facilitates comparisons between these basaltic rocks as well as others in which solid-solution minerals conceal whole-rock chemical variations. Aphanitic and, especially, glassy rocks are readily compared. Mica- and amphibole-bearing rocks that crystallized from hydrous magmas can be compared with rocks lacking hydrous minerals that crystallized from dry magmas of otherwise similar chemical composition. Moreover, rock compositions cast as norms can be easier to relate to the results of experimental laboratory studies of simplified, or model, rock systems.
Silica Saturation
In its allocation of silica first to normative feldspars and then to pyroxenes and finally to quartz, the normative calculation (Appendix B) emphasizes the concentration of SiO2 relative to oxides of K, Na, Ca, Mg, and Fe in the rock. The relative amounts of these oxides are compared on a molecular, rather than weight, basis. If there is insufficient silica in the rock to make normative pyroxenes from the amounts of these other oxides, then some FeO and MgO is instead allocated to normative olivine, which requires relatively less silica than Fe-Mg pyroxene; the silica deficiency is thus compensated. This chemical balance may be seen in the reaction
(Mg,Fe)2SiO4 + SiO2 = 2(Mg,Fe) SiO3
olivine orthopyroxene
Note that there are equal molar proportions (1:1) of SiO2 and (Mg,Fe)O in orthopyroxene, but half as much SiO2 as (Mg,Fe)O in olivine, or SiO2:(Mg,Fe)O = 1:2. (In the norm, orthopyroxene is represented by the normative mineral hypersthene, Hy.) In rocks that still have a deficiency of silica after eliminating all of the orthopyroxene, some silica must be reassigned from albite to nepheline, a silica-poor mineral. This chemical balance may be seen in the reaction
NaAlSiO4 + 2SiO2 = NaAlSi3O8
nepheline albite
Once again, note the difference in relative molar proportion of SiO2:Na2O 6:1 in albite and 2:1 in nepheline. Creating one mole of nepheline from one mole of albite liberates more silica than does conversion of one mole of orthopyroxene to one mole of olivine. Hence, modest silica deficiencies in rocks are manifest by olivine in lieu of orthopyroxene, whereas greater deficiencies are manifest by nepheline in lieu of sodic plagioclase.
The normative calculation serves as a model for a crystallizing magma and illustrates the concept of the degree of silica saturation. Consider a simple hypothetical magma consisting only of O, Si, Al, and Na. If there is an excess of molar SiO2 relative to that needed to make albite from Na2O, that is, SiO2/Na2O >6, then the magma can crystallize quartz in addition to albite. (In a natural magma, the albite would be in solid solution in plagioclase and/or alkali feldspar.) This magma and the corresponding rock are silica-oversaturated. If the magma contains SiO2 andNa2O in the exact ratio of 6, then these two constituents can only combine into albite; the magma and rock are silica-saturated. If the molar ratio SiO2/Na2O <6 but >2 in the magma, then there is insufficient SiO2 to combine with all of the Na2O into albite and some nepheline is created instead; the magma and rock are silica-undersaturated. If the molar ratio SiO2/Na2O = 2 in the magma, then there is insufficient SiO2 to combine with the Na2O to create any albite at all and only nepheline can be produced; the magma and rock still qualify as silica-undersaturated.
In real magmas and corresponding rocks that contain Mg, Fe, Ca, K, Ti, and so on, in addition to O, Si, Na, and Al, the concept of silica saturation still applies.
In the classification that follows, the degree of saturation is manifested in normative minerals (shown in italic letters) and with less accuracy by real minerals (in parentheses).
- Silica-oversaturated rocks contain Q (quartz or its polymorphs—cristobalite and tridymite), such as granite.
- Silica-saturated rocks contain Hy, but no Q, Ne, or Ol (no quartz, feldspathoids, or olivine), such as diorite and andesite.
- Silica-undersaturated rocks contain Ol and possibly Ne (Mg-olivine and possibly feldspathoids, analcime, perovskite, melanite garnet, and melilite), such as nepheline syenite.
Alumina Saturation
Al2O3 is the second most abundant constituent in most magmatic rocks and provides another means of classification, especially for felsic rocks, such as granitic ones. The alumina saturation index is defined as the molecular ratio Al2O3/(K2O + Na2O + CaO), which equals 1 in feldspars and feldspathoids. In magmas crystallizing feldspars and/or feldspathoids, any excess (ratio > 1) or deficiency ( <1), respectively, of Al2O3 must be accommodated in mafic or accessory minerals. In alumina-oversaturated, or peraluminous, rocks, excess alumina is accommodated in micas, especially muscovite, in addition to Al-rich biotite, and in aluminous accessory minerals such as cordierite, sillimanite, or andalusite, corundum, tourmaline (requires boron), topaz (fluorine), and almandine-spessartine garnet. (But beware: The latter three minerals also occur as vapor-phase precipitates in some metaluminous rocks.) After allocation of CaO for apatite, peraluminous rocks containnormative corundum, C. In alumina-undersaturated, or metaluminous, rocks, deficiency in alumina is accommodated in hornblende, Al-poor biotite, and titanite (but its stability also depends on other compositional properties of the magma including oxidation state). After allocation of CaO for apatite, metaluminous rocks contain normative anorthite, An, and diopside, Di (or wollastonite, Wo). A further constraint on metaluminous rocks is that they have Al2O3/(K2O + Na2O) > 1, whereas peralkaline rocks have Al2O3/ (K2O + Na2O) < 1. In peralkaline rhyolites and granites the alumina deficiency (alkali excess) is accommodated in alkali mafic minerals such as aegirine end-member pyroxene (NaFe3 +Si2O6) and the alkali amphiboles riebeckite richterite, and aenigmatite in which Fe2O3 and TiO2 substitute for Al2O3. Peralkaline rocks contain normative acmite or sodium metasilicate (Ac or Ns) and lack normative An.
Real feldspars in peralkaline rocks contain little of the anorthite end member. Peralkaline rhyolites can be further subdivided into comendites in which Al2O3 > 1.33 FeO + 4.4 (on a wt.% basis), and pantellerites, in which Al2O3 <1.33 FeO + 4.4. Peralkaline rocks can be silica-oversaturated, -saturated, or -undersaturated, as in, for example, comenditic and pantelleritic trachytes. An inherent weakness of classifications depending on the ratios of alumina or silica to alkalies is that Na and K can be mobilized and transferred out of a magma by a separate fluid phase. For example, escaping steam from cooling hot lava flows carries dissolved Si, Na, and K. However, Al tends to be less mobile. Initially metaluminous magma can, therefore, become peraluminous after alkali loss. Glasses can also lose alkalies relative to Al during high-T alteration or during weathering. A clue to preferential alkali loss is the presence of metaluminous minerals as phenocrysts, formed prior to extrusion, in a glassy matrix.