Classification of granites according their magmatic origin results in the formation of two contrasting groups, S-types and I-types. S-types result from the partial melting of metasedimentary source rocks, a process called anatexis or ultrametamorphism. I-types are derived from source rocks of igneous composition that have not gone through the surface weathering process, or from crystal fractionation of magmas. It is more often than not very difficult to determine whether a granite has a particular origin. As a result, many techniques have been developed to determine whether a granite is an S-type or an I-type. Some of these techniques are very simple, only requiring a study of mineral assemblages and inclusions; however, to determine beyond doubt whether a granite is of a particular type a combination of different analysis must be used.
S- and I-type granites from the Tasman Orogenic Zone of Eastern Australia were studied in detail by Chappell and White (1974) and others. The granites, which are of widespread occurrence, may be distinguished by chemical, mineralogical, field and other criteria. White, Chappell and their coworkers (1974, 1978, Hine et al, 1978) have carried out a complete study of magma provenance in this area, particularly in the Lachlan Fold Belt. They recognized a group of early, metamorphically harmonious plutons largely composed of S-type granites, which probably originated from the remelting of metasediments, and a younger group of mostly I-type granites with aureoles derived from remelting of deep seated igneous material.
Geochemical Methods
It has been shown that the crustal S-type granites are compositionally restricted and the mantle I-type granites are compositionally expanded (Chappell and White, 1974). This is reflected by marked differences in geochemical parameters. Chappell and White (1974) used a number of chemical properties to distinguish between S- and I-type granites. Hine et al (1978) uses a more detailed chemical analysis to confirm and expand upon Chappell and Whites findings.
Chappell and White distinguish S-type and I-type granites using numerous chemical parameters. I-types have relatively high sodium, Na2O greater than 3.2% , in felsic varieties, decreasing to more than 2.2% in mafic types. S-types have relatively low sodium, Na2O normally less than 3.2% in rocks with approximately 5%K2O, decreasing to less than 2.2% in rocks with approximately 2% K2O. Furthermore, S-types have been determined to have a Mol Al2O3/(Na2O+K2O+CaO) ratio of greater than 1.1 and I-types less than 1.1. Another distinctive chemical property determined by Chappell and White is the normative corundum in S-type granites, being greater than 1% CIPW corundum. In I-type granites, less than 1% CIPW diopside is present. I-type granites have regular inter-element variations within plutons with linear or near linear variation diagrams. The variation diagrams of S-type granites are more irregular.
These chemical properties result from the removal of sodium into sea water, or evaporites and calcium into carbonates, during sedimentary fractionation (weathering). Subsequent, relative enrichment of the sediment in aluminium must have occurred.. S-type granites com from a source which has been subjected to this sedimentary fractionation.
Hine et al (1978) used the Kosciusko Batholith to illustrate chemical differences between S-type and I-type granites. A Na2O/K2O plot of the Kosciusko batholith illustrates the fundamental chemical difference. The more potassium rich S-types are lower in sodium, a very distinctive phenomenon. Differences in this type are useful in recognizing S- and I-type granitoids and were an important factor in deciding what the overall S- and I-type characteristics were inherited from different sources (Chappell and White, 1974). The formation of shales by chemical weathering processes enriches Al relative to Na and Ca, since Na is removed into sea water and Ca is concentrated in carbonates. Pelitic rocks have high K/(Na+Ca) ratios and this is reflected in the high K/Na ratios of S-type granites and their relatively low Ca contents.. They directly result in the higher Al/(Na+K+Ca/2) ratio of S-type granites.
Hine also used oxidation states of iron in his attempt to classify the Kosciusko Batholith granitoids. Flood and Shaw (1975) suggest the presence of carbon or sulphur in the sedimentary source rocks results in S-types being much more reduced than I-types. This is consistent with the Kosciusko Batholith rocks studied by Hine et al.
Chappell and White suggest S-type granites are restricted in composition to high SiO2 types, while I-type granites have a wide composition from felsic to mafic. These characteristics are a consequence of S-type granitoids having been derived from a more SiO2 rich source. Therefore, granitoids containing less than 655% SiO2 can generally be assumed to be I-type. This is consistent with the work done by Hine et al.
Another important geochemical feature of S- and I-type granites is the various isotope compositions. The initial Sr87/Sr86 ratios being higher in S-types because have been through an earlier sedimentary cycle.. A critical boundary of 0.7060 is suggested by Kistler(1974) and Armstrong et al (1977). Chappell et al (1974) suggest a boundary of 0.708. I-type granites have initial Sr87/Sr86 ratios between 0.704 and 0.706 (Chappell and White, 1974). Isochrons of I-types give a regular linear set of points whereas those of S-types show a scatter of points within a broad envelope, reflecting variations in initial Sr87/Sr86 within a single pluton as a result of a heterogeneous source material. Classification based Sr87/Sr86 ratios are valid.
O’Neil et al (1977) and Taylor (1977) have illustrated the importance of oxygen isotope ratios, and to a lesser extent, hydrogen isotope ratios in distinguishing between S- and I-type granites. In general, sedimentary rocks are much richer in O18 than primitive igneous rocks, therefore, if the granitoids have retained the characteristics of their presumed source material, O18 composition should be a useful tool for differentiating between S- and I-type granitoids.
O’Neill et al, studying the Berridale Batholith, determined that oxygen isotope compositions of whole rock samples are an excellent discriminant between the two types. O’Neill determined the average composition of O18 for S- and I-type granites of the Berridale Batholith to 9.9 to 10.5 and 7.9 to 9.4 respectively. A drawback in his argument is that he assumes the oxygen isotope compositions to be a remnant of the original material from which was partially melted to form the granitoid. Earlier, he says, “… the effect of alteration is to lower the O18 content of the rock. If this alteration took place in the presence of aqueous fluid that was in near oxygen isotope equilibrium with these cooling plutons (i.e., a magma derived deuteric fluid), the lower temperature chloritization process should have increased the O18 content of the rock”. Clearly, the use of O18 compositions to classify granitoids is restricted to unaltered terrains. In cases where whole rock analysis does not clarify the problem, O’Neil says the analysis of mineral separates can often provide answers.
Discrimination based on hydrogen isotope composition is less precise than using oxygen isotopes, but average hydrogen isotope compositions lower than -80 almost certainly indicate I-types. S-types have average values clustering around -62.
Petrographic Methods
Some of the geochemical aspects of S- and I-type granitoids are reflected in the mineralogy. Hornblende is common in the more mafic I-types and is generally present in the felsic types also. In the felsic S-types, hornblende is absent, but muscovite is common, while in the mafic S-types, biotite is often very abundant. Monazite is the usual accessory in the S-types whereas sphene is common in the I-types. Garnet and cordierite may occur in S-type xenoliths as well as in the granites themselves. Apatite inclusions are common in biotite and hornblende of I-types, but occur in larger individual crystal in S-types. Thus, I-types characteristically contain biotite+hornblende plus/minus sphene plus/minus monazite. S-types contain biotite plus/minus muscovite plus/minus cordierite plus/minus garnet plus/minus ilmenite plus/minus monazite.
A fairly detailed petrological analysis of S- and I-type granites was carried out by Hine et al (1978) of the Kosciusko Batholith. Hine and his coworkers determined the S-type granitoids to be dominantly quartz rich adamellites and granodiorites with a few felsic granites and mafic tonalites. A modal distribution shows the composition in S-types to be dissimilar to I-types, with S-types falling in the granite/adamellite range and I-types in the granodiorite/tonalite/monzodiorite range.
Field Methods
Observable features in the field may at times be distinctive. Chappell and White (1974) determined the more mafic I-types to contain mafic hornblende bearing xenoliths of igneous appearance whereas hornblende bearing xenoliths are rare in S-types. Metasedimentary xenoliths are common in S-types. Chappell and White observed that S-types are usually early in the intrusive sequence and often have a strong secondary foliation. They apply the term “restite” to relict, or residual, source material. They go on to say large, restite milky quartz inclusions are common in S-type granites. Restite occurs as xenoliths, clots or xenocrysts and may be used to distinguish between S- and I-type granites.
Conclusion
The methods that may be used to distinguish between S- and I-type granites include geochemical, petrographic and field methods. Not one single method may be used universally to make a clear distinction since outside influences such as post orogenic alteration and reaction with wall rock material may drastically change some of these criteria. In general, geochemical methods seem to be the most successful contrasting the differences between S- and I-type granites. Petrographic and field observations are extremely useful to supplement geochemical data.
References
Chappell, B.J. and White, A.J.R., 1974, “Two Contrasting Granite Types”. Pac. Geol., v8, pp.173-174.
Flood, R.H. and Shaw, S.E., 1975,”A Cordierite bearing granite suite from The New England Batholith”. Contr. Mineral. and Petrol., v52, pp.157-164.
Hine, R., Williams, I.S., Chappell, B.W. and White, A.J.R., 1978, “Contrasts Between I- and S-Type granitoids of the Kosciusko Batholith”, J. Geol. Soc. Aust., V25, pp.219-234.
O’Neill, J.R. and Chappell, B.W., 1977, “Oxygen and Hydrogen Isotope Relations In the Berridale Batholith”, J. Geol. Soc. Lond., V.133, pp.559-571.
Pitcher, W.S., 1979, “The Nature, Ascent and Emplacement of Granitic Magmas”, J. Geol. Soc. Lond., V136, pp. 627-662.
White, A.J.R. and Chappell, B.W., “Ultrametamorphism and Granite Genesis”, Tectonophysics, V.43, pp. 7-22
