Experimental studies of nickel partitioning

Irvine (1974) proposed a slight modification of the equilibrium represented by eq. (8.14) as follows 

As the melt becomes more siliceous, the equilibria for Ni2+ shift strongly in favour of the crystallizing olivine phase due to decreased CFSE s attained in more highly polymerized silicate liquids. Apparently, the proportion of tetrahedral sites in the melt rises with increasing K20 + Si02 components. 

---