Re: Effect on melting temperature of mantle upon addition of water

Water, which is present in rocks in small quantities, is usually treated in its dissociated form 
as OH-.  It is a "network modifier" which means that it disrupts the silicon-oxygen bonds 
and therefore disrupts the polymerization of silica tetrahedra that form the backbones of 
silicate minerals.  Once disrupted, the minerals are much easier to melt, which simply 
involves further disruption.  Since the structures take less energy to melt, they can melt at 
lower temperatures.  Silicate melts do not consist of isolated ions or ionic groups, rather 
they contain long and short fragments of chains, sheets, and framework structures.  These 
structures interfere with each other as the melt flows and so contribute to viscosity.  As 
water is added, it not only lowers density, it also lowers viscosity.

Water percentages in melts are usually given as weight percent.  This can be misleading, 
since water has such  low molecular weight compared to  minerals.  The mole percent water 
may be much higher.

The equations needed to predict the thermodynamic relationships of minerals and melts are 
derived from simple thermodynamic relationships such as those for ideal gases.  
Unfortunately, the relationships for minerals are not simple, in part because the materials 
are not ideal and in part because the materials have such complex chemistries.  The 
treatment of those equations is well beyond what I am capable of here.  Whole books have 
been written on this subject.  Go to GeoRef and search on mineral thermodynamics or melt 
thermodynamics.  Go to a science bookstore or your professors and look for 
'Thermodynamics of Minerals and Melts' or 'Physics and Chemisrty of Earth Materials', both by 
Alexandra Navrotsky, or Krauskoff's 'Thermodynamics for Geologists' (not sure if that is 
exactly the right name) or others.

David Smith, Ph.D.
Discovery Center of Science and Technology, Bethlehem, PA