Re: PHREEQC questions

["David L Parkhurst" <dlpark@xxxxxxxx>]

> 1) for oxidized Mn minerals containing Mn4+, the major PHREEQC databases
have written the Ksp reactions using Mn3+.  That is perfectly rational as
Mn4+ is not listed as a species in databases.  But why is Mn4+ not
included?  I recognize that equilibrium data for the Mn3+/Mn4+ couple would
be hard to come by and perhaps not terribly meaningful if they were....but
I would appreciate your perspective.  From a practical point of view, does
this mean that in many cases the reported Mn3+ concentrations and
speciation are in fact equivalent to Mn4+ concentrations (or
species????hmmm...)  any references to how this is handled in PHREEQC
documentation would be appreciated

My understanding is that aqueous Mn+4 is virtually non existent. Under
natural conditions, Mn+2 is the only significant manganese species.
Permanganate is metastable and tends to decompose to Mn+2 relatively
quickly. The question seems to imply that because Mn+4 exists in solids, it
must exist in solution, but there is no reason to require this. Sulfur in
pyrite is effectively valence -1, but there is no -1 aqueous species. For
manganese, it simply means that there must be a redox reaction in the
dissolution of +4 minerals, generally reducing manganese and oxidizing
something else, water to oxygen if necessary.

> 2) Both ferric and manganic phases have their  IAP's calculated in
PHREEQC
according to some convention besides that indicated by the mineral reaction
given in PHREEQC.DAT and elsewhere.  They also report Kt that differs from
the database and is following the same alternate reaction.  I am confident
the SI's that are reported are fine but what basis is used for calculating
the IAP's and modifiying the Ksp's from the database values for minerals
like (say) goethite or Fe(OH)3 (a)?

This is a "feature" of the program. I think I just got lazy at some point.
The different K's result from "basis switching", which means the program
rewrites the equations for the minerals to the predominant valence state at
intermediate calculations in the numerical method. Thus if ferric ion is 10
times greater than ferrous ion, the equations are rewritten to ferric,
resulting in a different log K. I should have gone back to the original log
K's in the printout, but didn't.


David Parkhurst (dlpark@xxxxxxxx)
U.S. Geological Survey
Box 25046, MS 413
Denver Federal Center
Denver, CO 80225