GRANITE TYPES AND THEIR FRACTIONATION
 

B.W. Chappell, GEMOC ANU

Studies of granites in the mid-Palaeozoic Lachlan Fold Belt of south-eastern Australia have long shown that those rocks can be subdivided into two dominant groups. These are the I- and S-types, with compositional features that can be related to the Igneous and Sedimentary source rocks from which they were derived by partial melting. The two granite types are most easily recognised at more mafic compositions, where the I-type granites are undersaturated in Al2O3 (metaluminous) and contain hornblende, whereas the S-type granites are peraluminous and contain abundant biotite generally with Al-rich minerals such as cordierite. The mafic S-type granites are peraluminous because they contain unmelted material from the source (restite) that was entrained in a more felsic melt. The mafic I-type granites are generally metaluminous for an analogous reason, but such granites also formed as cumulate rocks that separated from less felsic melts. The more mafic granites that did not form as cumulates have bulk compositions close to those of their source rocks.

At felsic compositions, the compositions of both I- and S-type granites converge to those of minimum-temperature melts. Such compositions represent either primary felsic melts which completely separated from restite crystals, or were the product of fractional crystallisation of a less felsic melt. In cases where there was no significant fractionation of feldspars from that felsic melt, it is difficult to make a compositional distinction between the two types, as both are slightly oversaturated in Al and even the I-type granites may contain small amounts of muscovite.

With fractional crystallisation of a felsic melt, the compositions of the two granite types again diverge, and distinct differences in trace element abundances become apparent. At the same time the major element compositions change very little, being constrained by the equilibrium of quartz and feldspars in the haplogranite system. The fractionation of feldspars from a felsic melt leads to increases in Ga, Rb and Cs, and decreases in Sr, Ba and Eu, in both the I- and S-types. The elements F, Nb, Sn and U also increase, while Zr decreases, with fractionation. The abundances of those elements can be used as indicators of the extent of fractional crystallisation in any felsic granite. However, there are some elements whose response to fractional crystallisation at such felsic compositions differs between I- and S-type granites. Most notably, P contents increase in S-type granites that have undergone progressive fractional crystallisation, and decrease in the equivalent I-type granites; these differences arise because of the much higher solubility of P in strongly peraluminous melts. Those differences in solubility of P also affect the abundances of other elements that reside in P-bearing accessory minerals. Thus Th decreases in abundance in S-type granites, and increases in I-type granites, with fractional crystallisation. The detailed behaviour of the REE is also different between the two types.