|U. Mass Lowell||Prof. Nelson Eby||Department of Environmental, Earth, & Atmospheric Sciences|
Chappell and White (1974), based on their study of the major batholiths of the Lachlan Fold Belt, Australia, introduced the first letters, I and S, of the granitoid alphabet soup. According to Chappel and White (1974), I-type granitoids were derived from igneous (or meta-igneous) sources while S-type granitoids were derived from metasedimentary sources. White (1979) defined a third granitoid type (M-type) which was presumably derived from the melting of subducted oceanic crust or the overlying mantle. Loiselle and Wones (1979) added the last letter, A, of the alphabet soup. The A-type (the A denoting anorogenic and/or anhydrous) granitoids occurred along rift zones and within stable continental blocks. The identification of A-type granitoids was based on both tectonic setting and chemical characteristics. This represented a departure from the I and S types which were strictly based on the difference in the sources of the granitoids.
Petrological, geochemical (including INAA trace elements), mineralogical (electron microprobe), and geochronological (including fission track), and isotopic studies of A-type granitoids have been ongoing since the late 1980s. Publications resulting from these studies are listed below.
References for classic papers cited above:
Chappell and White (1974) reprinted in - Chappell, B. W., White, A. J. R., 2001. Two contrasting granite types: 25 years later. Australian Journal of Earth Sciences 48, 489-499.
Loiselle, M. C., Wones, D. R., 1979. Characteristics and origin of anorogenic granites. Geological Society of America Abstracts with Programs 11, no. 7, 468.
White, A. J. R., 1979. Sources of granite magmas. Geological Society of America Abstracts with Programs 11, no. 7, 539.
A-type granites are emplaced in either within plate anorogenic settings or in the final stages of an orogenic event (sometimes referred to as post-orogenic). The universal commonality is that the tectonic environment has become extensional and the granites do not show any tectonic fabric (although such fabric may be introduced by later tectonic events). In a number of provinces A-type granites are associated in space and time with silica-undersaturated lithologies. In some cases, mafic rocks are closely associated with the A-type granites. The A-type granites have been characterized chemically in terms of their Ga/Al ratio versus a variety of trace and major elements and they plot in the Within Plate Granite fields on tectonic discriminant diagrams. They range from peraluminous to peralkaline in composition. In primitive mantle normalized plots they generally show significant enrichment in Cs and Rb, significant depletion in Ba, Sr, Eu, Ti, and P, and no or minor Nb and Ta depletion. Chondrite normalized REE plots are variable from relatively flat patterns with large negative Eu anomalies to more LREE enriched patterns with relatively small Eu anomalies. In most cases the isotopic data indicate mantle involvement. Given the variable settings and associated lithologies, one petrogenetic model cannot be used to explain the origin of A-type granites.
The 130 – 120 Ma Chilwa alkaline province is located at the southern end of the current day East African Rift system. Magmatic activity started with the eruption of basanitic lavas followed a few million years later by the emplacement of nepheline syenites, syenites, granites, and carbonatite. The sequence of emplacement is from silica undersaturated rocks to silica oversaturated rocks. Elemental and isotopic data indicate that the magmas were derived from an OIB-like mantle source. AFC processes led to the variety of felsic rock types seen in the province, with the most contaminated magmas yielding the granites.
The White Mountain igneous province of the northeastern United States is comprised of a number of ring dike complexes and associated stock-like intrusions. The igneous activity occurred in two discrete time intervals – 190 to 160 Ma and ~122 Ma. The large intrusions of the older event are dominantly felsic and silica saturated while the generally smaller intrusions of the younger event have significant associated mafic rocks. The younger event is also time-correlative with the Monteregian Hills alkaline province of Quebec which shows the whole conceivable range of silica-undersaturated to silica over-saturated alkaline rock types. Elemental and isotopic data indicate an OIB-like source, with mafic magmas variably contaminated by crustal material.
The Carboniferous granites of western Argentina were emplaced at the end of a long period of orogenesis and are related to a major shear zone. These granites are peraluminous to metaluminous and elemental and isotopic data indicate a subcontinental lithospheric mantle component, but the magmas were largely derived from the crust.
The above examples indicate the variety of settings in which A-type granites can occur. In all cases there is evidence for a mantle component, although this component seems to be less significant in the case of post-orogenic granites.
Zozulya, D.R. and Eby, G.N.
suite of massif-type anorthosites and peralkaline granites is found in the
Archean Keivy terrane of the NE Baltic shield. The 2660-2680 Ma Keivy
anorthosite complex consists of several large (up to the 170 km2)
lopoliths composed mainly of anorthosite and gabbro-anorthosite and marginal
gabbro-norite and titanomagnetite-rich troctolite bodies. The Keivy anorthosites
have low REE abundances (Ce 5-23 and Yb 1.5-6.8 times chondrites), fractionated
REE distributions (chondrite-normalized La/Yb ratios are 4-10) and positive Eu
anomalies. The comagmatic gabbro-norites have similar REE patterns, but no or
negligible positive Eu anomalies. The rocks show high compatible element (Sc,
25-40 ppm and Sr, 460-670 ppm) abundances. As the chondrite-normalized La/Yb
ratios do not correlate with REE abundances, an enriched source for the primary
magmas is proposed. The enriched source for the Keivy anorthosites has low eNd
(-0.15 to -0.24) and low Y/Nb ratios (0.6-1.3). From the geochemical data it is
inferred that the primary magma was an alkaline/subalkaline basalt magma forming
in a within-plate setting.
Keivy peralkaline granite complex consists of 2650-2660 Ma peralkaline granites,
2670 Ma syenogranites, and 2680 Ma alkaline syenites. The granites form
sheet-like bodies with thicknesses of a few hundred meters, but have vast
exposed areas (100-1200 km2). They are metamorphosed to amphibolite
facies and are bounded by gabbro-anorthosite. The rocks of the Keivy complex are
extremely enriched in Zr (300-1900 ppm), Y (40-150 ppm), Nb (20-150 ppm), REE
(100-1000 times chondrites) and Rb (160-900 ppm), have associated Zr-REE ore
occurrences, are very low in Sc (0.3-1.3 ppm) and Sr (10-30 ppm), show negative
Eu anomalies, have normalized La/Yb ratios of 1.5-13 and
high Ga/Al ratios. On standard trace element discriminant diagrams the Keivy
peralkaline granites plot as within-plate or post-collisional A-type granitoids.
The low Y/Nb and Yb/Ta ratios for associated alkaline syenites point to their
OIB affinities. The least metamorphosed and least evolved rocks plot in the
EM2-field on the eSr
The close temporal and spatial association of the gabbro-anorthosites and the peralkaline granites and their similar magma sources suggest a genetic relationship. One possible model is protracted fractional crystallization of a primary alkaline basalt magma with removal of plagioclase during the early stages of crystallization (forming a Ca- and Al-enriched cumulate, anorthosite) and alkali, iron and HFSE enrichment of the residual melt leading to the peralkaline granites.
Eby, G. N.
Since the inception of the term A-type by Loiselle and Wones (1979), this class of granitoids has proven to be the most controversial and least understood member of the granitoid alphabet soup. Eby (1990, 1992) suggested that there were a variety of granitoids that fell within the A-type classification and that there were multiple petrogenetic pathways that could lead to rocks that met the largely chemical definition of A-type granitoids. The A1-type (often referred to as anorogenic) was a distinct group that had characteristics of magma derived from an OIB source and was inferred to be the fractionation product of an OIB-like basalt magma. The A2-type (often referred to as post-collisional or post-orogenic) represented all A-type granitoids not derived by fractionation of an OIB-like magma. These granitoids were generally emplaced shortly after an orogenic period and may have originated by melting of mantle material with crustal interaction or solely by the melting of crustal material.
The North Nyasa and Chilwa alkaline provinces of Malawi and the White Mountain igneous province of the northeastern US are classic examples of A1-type magmatic provinces. An extreme range of lithologies is found in all three provinces from carbonatites through a variety of both silica-undersaturated and silica over-saturated silicate rocks to alkali granites. Trace element and isotopic data indicate that an OIB-source is an important component in the magmatic history. The White Mountain province is an instructive example. Two periods of A-type magmatism are recognized in this province, the older from 200 to 160 Ma and the younger confined to a narrow time interval centered around 123 Ma.
The older White Mountain series essentially consists of silica-saturated felsic (syenite to alkali granite) igneous rocks. Mafic rocks are only significant in the Pliny Range. Nepheline-bearing syenites are found at Red Hill and Rattlesnake, indicating that silica-undersaturated magmas were present at the time the silica-saturated sequences were emplaced. Detailed studies of the largest unit in the series, the White Mountain batholith, suggest that all the igneous rock groups (syenites, metaluminous granties, peralkaline granites and rhyolites) can be related through variable interactions of mantle-derived melts with the subcontinental lithosphere.
The younger White Mountain series and the temporally and spatially related Monteregian Hills province of Quebec, Canada, shows the extreme range of lithologies often typical of the A1-type association. Mafic rocks are significant in these two provinces and basalts and rhyolites are found in several intrusions. Based on trace element and isotopic chemistry, a successful model for the origin of the various lithologies involves various degrees of partial melting of a garnet lherzolite source (depleted mantle based on isotopic characteristics) and subsequent variable interaction of the basaltic magmas with the continental crust.
An example of an A2-type province is the Jurassic granitoids and associated bi-modal volcanics of southern China. These sequences were emplaced in a rift structure that developed shortly after continent-continent collision. The basaltic magmas are continental tholeiites and the rhyolites and the granites apparently formed by differentiation of the basaltic magmas with some crustal contamination.
It is suggested that A-type granitoids can form via three different petrogenetic pathways. The A1-types form by differentiation of a basaltic magma, with variable degrees of crustal contamination, derived from an OIB-like source. The A2-types form either by differentiation of a continental tholeiite, with variable degrees of crustal interaction, or by direct melting of a crustal source that had gone through a previous melting episode. The challenge, in the case of the A2-types, is determine which of these pathways was followed by a particular granitoid.
Zozulya, D. and Eby, N.
The Late Archean (2.6-2.8 Ga) is remarkable for the earliest manifestation of syenite-granodiorite-granite magmatism of alkaline affinity. From the published data this magmatism can be related to either “sanukitoid” (subduction) or “A-type granitoid” (anorogenic) magmatism. Key examples from the Superior, Yilgarn, and Baltic shields were studied to discriminate their origin: 2680-2670 Ma alkaline granites, syenites and associated nepheline syenites of the Abitibi greenstone belt (Sutcliffe et al., 1990; Corfu et al., 1991); 2650-2630 Ma alkaline granites and syenites of the Eastern Goldfields granite-greenstone terrane (Libby, 1989; Smithies & Champion, 1999); and 2610-2680 Ma alkaline granites, syenogranites, and associated nepheline syenites of the Keivy complex of the Central Kola granite-greenstone domain (Mitrofanov et al., 2000; Zozulya et al., 2005).
The first two examples consist of small (10-90 km2)
stocks spatially and temporally associated with potassic volcanics and
lamprophyres, and they have structural and genetic links to greenstone
belts. The Keivy complex consists of several sheet-like bodies of
100-500 m thickness, having vast exposed areas (100-1300 km2),
that are spatially and temporally associated with massif-type
Granitoids from the various provinces have common mineralogical and petrochemical characteristics: anhydrous primary phases, Fe- and Na-rich mafic silicates, low Ca, Mg, Al, and high total alkalis. At the same time the granitoids show different trace element characteristics and fertility types. Coupled with different geological structure this suggests different tectonic settings.
The Superior and Yilgarn felsic alkaline rocks show extremely high concentrations of Ba (c. 500-4500 ppm) and Sr (c. 300-3000 ppm); low Zr, REE (no Eu anomaly), Y, Nb, Ta, and Rb; and low Ga/Al and high Y/Nb and (La/Yb)n ratios. Gold occurrences are detected. Based on these geochemical features the granites were formed in a subduction environment and correspond to sanukitoid suites.
Keivy felsic alkaline rocks are
low in Ba (c. 40-200 ppm) and Sr (c. 5-30 ppm); extremely high in Zr,
REE (except for a distinct negative Eu anomaly), Y, Nb, and Rb; and have
high Ga/Al (for granite) and low (La/Yb)n and Y/Nb (Yb/Ta) (for syenite)
ratios. The granites and syenites host Zr-Y-REE deposits and
occurrences. The granitoids were formed in a within-plate setting and
belong to the A-type granite group (A2 subgroup). The
associated nepheline syenites have geochemical affinities to OIB-derived
Sutcliffe et al. (1990) and Shirey & Hanson (1984), based on Nd isotope studies and elemental constraints, suggested that the “Superior” type granites are derived from depleted mantle sources that were enriched in LILE shortly before melting. Based on Nd and Sr isotopes, the “Keivy” type has a highly evolved enriched mantle source (EM2). It is likely that the Keivy granites are the product of a high degree of fractional crystallization of mantle-plume-derived alkaline basalt magma. Thus the Late Archean felsic alkaline magmatism is of two types: (1) subduction related sanukitoid-like with a depleted mantle source and (2) anorogenic A-type with an enriched mantle source.
Eby, G. N., Woolley, A. R., and Collerson, K.
Cretaceous Chilwa alkaline province lies at the southern end of the
East African rift system. A diverse suite of intrusive and extrusive
alkaline igneous rocks is found in the province. The earliest igneous
activity (ca 133 Ma) is
represented by the extrusion of nephelinites and basanites and the
intrusion of nepheline and sodalite syenites.and carbonatite.
Nepheline syenites and syenites were subsequently intruded ca 126 Ma followed by the emplacement of large peralkaline syenite-granite
intrusions ca 113 Ma (Eby et
al., 1995). Trace element and Sr and Nd isotopic data have been
determined for a subset of the samples reported in Woolley and Jones
OIB-normalized spider diagrams for the basanites are essentially flat with absolute abundances of 1 to 2x OIB. The nephelinites show a more irregular pattern with relative enrichment in LIL elements (3-5x). OIB-normalized spider diagrams for the phonolites, syenites, and granites are generally flat to slightly LIL-enriched with absolute abundances of 1-10x OIB and significant negative Ba, Sr, P, and Ti anomalies. Two groups of phonolites can be identified based on REE patterns – one shows no relative Eu depletion (Group I) while the other shows significant Eu depletion (Group II).
terms of Sr and Nd isotopes the basanitic volcanics and the
nephelinites of Chilwa Island have depleted mantle signatures. The
strongly nepheline-normative syenites also show depleted mantle
signatures. The syenites and granites show evidence of crustal
contamination that can generally be explained by simple mixing with
crustal material. The two phonolite groups are isotopically distinct
– Group I has low Sr initial ratios while Group II has high Sr
basanites and nephelinites can be generated by small percentage melts
of a trace-element enriched mantle source (but significant relative
LIL-enrichment is not required). The isotopic and trace element data
are consistent with the evolution of the Group I phonolites at great
depth, where plagioclase crystallization was suppressed, while the
Group II phonolites evolved by assimilation-fractional crystallization
(AFC) processes at shallow depth. The transition from silica-undersaturated
syenites to silica-saturated syenites and granites is probably the
result of melting and assimilation processes occurring at the base of
the crust. AFC processes, during emplacement at high levels, are
responsible for subsequent minor variations in magma chemistry.
Zozulya, D. and Eby, N.
The oldest known examples of felsic alkaline magmatism are from the Superior province, Yilgarn Craton, and Fennoscandian Shield. These are the 2680-2670 Ma alkaline granites and associated nepheline syenite stocks of the Abitibi greenstone belt (Sutcliffe et al., 1990; Corfu et al., 1991): the 2650-2630 Ma alkaline granites and syenites of the Mount Monger, Emu, Claypan, and Ninnis suites of the Eastern Goldfields granite-greenstone terrane (Libby, 1989; Smithies, Champion, 1999): and the 2610-2680 Ma alkaline granites, syenogranites, and nepheline syenites of the Keivy complex of the Central Kola granite-greenstone domain (Mitrofanov et al., 2000; Zozulya et al., 2001). The Superior and Yilgarn felsic alkaline rocks form small (10-90 km2) stocks that are spatially and temporally associated with potassic volcanics and lamprophyres. The Kola examples form sheet-like bodies with thicknesses of a few hundred meters, but have vast exposed areas (100-1300 km2). They are metamorphosed to amphibolite facies and are closely associated with large gabbro-anorthosite bodies. While the various provinces have the common mineralogical (anhydrous primary phases, Fe- and Na-rich mafic silicates) and petrochemical (low Ca, Mg, Al, and high total alkalis) characteristics of alkaline granites they show different trace element characteristics and mineralization types. The Superior and Yilgarn alkaline granites and syenites have extremely low Ga/Al and high (La/Yb)n ratios, no Eu anomaly, and related Au mineralization. Based on these geochemical features the granites were formed in a subduction environment and correspond to volcanic arc granites. In contrast, the rocks of the Keivy complex are extremely enriched in Zr, Y, Nb, REE, and Rb, have associated Zr-REE ore occurrences, are very low in Ba and Sr, show distinct negative Eu anomalies, and have low (La/Yb)n and high Ga/Al similar to A-type granitoids forming in within plate settings. The low Y/Nb and Yb/Ta ratios for the associated nepheline syenites point to their OIB affinities. The felsic alkaline magmatism of the Superior and Yilgarn provinces is related to the final stages of greenstone-belt formation (the ages are 2725-2680 Ma and 2720-2675 Ma, respectively). The Keivy complex was formed long after the development of the adjacent greenstone belt (the formation age is 2920-2830 Ma) and reflects the influence of a mantle plume.
Waight, T. E., Weaver, S. D., Maas, R., and Eby, G. N.
The Hohonu Dyke Swarm and French Creek Granite represent contemporaneous and cogenetic alkaline magmatism generated during crustal extension in the Western Province of New Zealand. The age of 82 Ma for French Creek Granite coincides with the oldest oceanic crust in the Tasman Sea and suggests emplacement during the separation of New Zealand and Australia. The French Creek Granite is a composite A-type granitoid, dominated by a subsolvus biotite syenogranite with high silica, low CaO, MgO, Cr, Ni, V and Sr and elevated high-field-strength elements (Zr, Nb, Ga, Y). Subordinate varieties of French Creek Granite include a hypersolvus alkali amphibole monzogranite and a quartz-alkali feldspar syenite. Spatially associated rhyolitic dykes are considered to represent hypabyssal equivalents of French Creek Granite. The Hohonu Dyke Swarm represents mafic magmatism which preceded, overlapped with, and followed emplacement of French Creek Granite. Lamprophyric and doleritic varieties dominate the swarm, with rare phonolite dykes also present. Geochemical compositions of French Creek Granite indicate it as an A1-subtype granitoid and suggest derivation by fractionation of a mantle-derived melt with oceanic island basalt - like characteristics. The hypothesis that the French Creek Granite represents fractionation of a Hohonu Dyke Swarm composition, or a mantle melt derived from the same source, is tested. Major- and trace-element data are compatible with derivation of the French Creek Granite by fractionation of amphibole, clinopyroxene and plagioclase from mafic magmas, followed by fractionation of alkali and plagioclase feldspar at more felsic compositions. Although some variants of the French Creek Granite have Sr and Nd isotopic compositions overlapping those of the Hohonu Dyke Swarm, most of the French Creek Granite is more radiogenic than the Hohonu Dyke Swarm, indicating the involvement of a radiogenic crustal component. Assimilation-fractional crystallisation modelling suggests isotopic compositions of French Creek Granite are consistent with extreme fractionation of Hononu Dyke Swarm magmas with minor assimilation of the Greenland Group metasediments.
Eby, G. N., Krueger, H. L., and Creasy, J. W.
The White Mountain batholith, central New Hampshire, is a member of the older White Mountain igneous province. The batholith consists of a number of overlapping centers of felsic magmatism comprised of quartz syenite porphyry ring dikes, alkali granite, metaluminous biotite granite, and felsic volcanics. Igneous activity occurred over an extended period of time from ca. 200 Ma to 155 Ma.
Both chemically and mineralogically the various suites of the batholith are A-type granitoids. Absolute abundances of alkalis, total Fe, REE (except Eu), Y, Nb, Ta, Hf, Zr, Th, and U are high relative to other granitoid types. The mafic mineralogy is characterisitc of A-type granitoids: ferrohedenbergite, fayalite, ferrohastingsite, and riebeckite.
Isotopic and trace-element data indicate that a number of the suites evolved by closed system crystal fractionation. The evolution of these suites was largely controlled by feldspar (perthite, K-feldspar, plagioclase), fayalite, ferrohedenbergite, and ilmenite fractionation. Locally contaminated phases attest to limited open system fractionation at high crustal levels.
While Sr isotopic ratios are generally significantly greater than mantle values, trace-element ratios are typical of magmas derived from oceanic-island-basalt-like sources. Modeling shows that the high 87Sr/86Sr ratios can be explained by limited contamination of highly evolved magmas by radiogenic crust.
The preferred model for the origin of the batholith involves the emplacement of mantle-derived melts into the base of the crust. These magmas evolve, with some crustal contamination, at this level. The resulting homogeneous magmas are then emplaced at higher levels in the crust where further crystal fractionation and contamination can occur. This is essentially the MASH model advanced by Hildreth and Moorbath (1988).
Eby, G. N.
The A-type granitoids can be divided into two chemical groups. The first group (A1) is characterized by element ratios similar to those observed for oceanic-island basalts. The second group (A2) is characterized by ratios that vary from those observed for continental crust to those observed for island-arc basalts. It is proposed that these two types have very different sources and tectonic settings. The A1 group represents differentiates of magmas derived from sources like those of oceanic-island basalts but emplaced in continental rifts or during intra-plate magmatism. The A2 group represents magmas derived from continental crust or under-plated continental crust that had been through a cycle of continent-continent collision or island-arc magmatism.
Eby, G. N.
A variety of granitoid suites, which have been classified as A-types, are reviewed in this paper. Based on this review, the general characteristics of the A-type granitoids are summarized as follows. The A-type granitoid suites vary in composition from quartz syenties to peralkaline granites and their respective volcanic equivalents. These suites are emplaced into non-orogenic settings - both within plate and along plate margins during the waning stages of subduction-zone-related magmatism. With respect to I- and S-type granitoids, the A-types are characterized by their relatively high alkali contents and low CaO contents (at SiO2 = 70%: Na2O+K2O = 7-11%, CaO<1.8%), high FeOT/MgO = 8-80), and often elevated halogen, particularly F, contents (F = 0.05-1.7%). The major element chemistry is reflected in the mineralogy by the occurrence of iron-rich micas, amphiboles and pyroxenes and in the peralkaline varieties by the occurrence of alkali-rich amphiboles and pyroxenes. Trace element abundances, particularly elevated concentrations of high-field-strength cations, are distinctive.
Y/Nb and Yb/Ta ratios are relatively constant for each A-type suite and thus serve as useful indices for chemical comparisons. A-type suites with Y/Nb<1.2 are derived from sources chemically similar to those of oceanic island basalts while suites with Y/Nb>1.2 are derived from sources chemically similar to island arc or continental margin basalts. In combination with isotopic and other trace element data, these relationships suggest that A-type granitoids are generated by a variety of processes including fractionation from mantle-derived basaltic magmas, interaction between these mantle-derived magmas and continental crust and in some cases by the formation of anatectic, halogen-rich, melts during the remelting of a terrane from which a previous melt had been extracted.
Eby, G. N. and Kochhar, N.
Major and trace element data have been obtained for the ca. 730 Ma granites of the Siwana and Jalor complexes, Rajasthan and the Tosham complex, Haryana. On the basis of geologic, geochronologic and chemical data these granites are classified as anorogenic, and are believed to have been formed during a wide-spread thermal event centered around 800 Ma. All of the complexes show evidence of differentiation at high levels through fractionation of alkali feldspar, plagioclase, and amphibole. The granite from the Siwana complex is peralkaline and shows high rare-earth (REE) contents with little fractionation between LREE and HREE. This granite may have been derived as a high temperature melt from an anhydrous granulitic source region from which a previous melt had been extracted or by melting of a metasomatized lower crustal source. The Jalor quartz syenites and granites are more primitive, metaluminous and have more fractionated REE patterns. These rocks are most likely differentiates of a mantle-derived magma. The Tosham granites are metaluminous to peraluminous and show isotopic and chemical characteristics typical of magmas derived from a high-grade metasedimentary source. These observations indicate that a variety of granite types can be formed during a period of anorogenic magmatism.