Thank you for your comments....In Perple_X, I have used two different
amphibole solution models - Amph(DHP), which is modified from Dale, and
GlTrTsPg, which uses parameters from White, Powell & Phillips (2003) and
Wei, Powell, & Zhang (2003). These seem to yield similar minimizations,
although using Amph(DHP) reduces the stability of albite at high
pressure at least in pelite compositions (11.5 kbars with GlTrTsPg at
570C to 8.5 kbars with Amph(DHP))......I have not experimented with
different amphibole solutions using more mafic compositions yet.
Comparing Perple_X with Domino has yielded relatively similar results,
with the exception of cpx stability. In Perple_X, cpx+amph is stable
above ab=jd+qtz with only amph stable below this reaction. Domino
predicts gl (with no cpx) above ab=jd+qtz with cpx + parg stable below
this reaction. The Perple_X result seems more reasonable, and I thought
that a better solution model for amphibole in Domino might change the
results.....
Hope all is well. scott
David Dolejs wrote:
> Dear Scott and others,
>
> These are interesting and important questions. Apart from sgnificant
> differences in computational strategies, the Domino and Perplex
> database formats may not allow you to make identical thermodynamic
> implementations of solid solution models (especially if you use the
> van Laar formalism for amphiboles).
>
> There are several models for amphiboles (Dale et al. 2000, 2005,
> Okamoto & Toriumi 2004, Mäder & Berman), which one do you think is the
> best?
>
> I don't want to discourage you, but I choose either software based on
> the problem at hand. If reciprocal solutions are involved (in your
> case, biotite and amphibole), I perform a few simple minimizations and
> look if assemblage and phase compositions for a given bulk, P and T
> come out from Domino/Theriak and Perplex comparably. If you get
> different results, this is still no surprise at this stage.
>
> Theriak may run into serious problems with reciprocal solutions,
> especially when reciprocal Gibbs free energies are zero (implicit in
> the Holland-Powell ideal models), only sometimes you may get negative
> mole fractions still leading to realistic site fractions. I don't
> think that you can incorporate van Laar size parameters of the latest
> amphibole model in Theriak/Domino. My preference would be Perplex,
> but make sure that you cover adequate composition space (check for
> edenite and ferroedenite, if these are accessible in Perplex, if
> relevant for your problem).
>
> David
>
>
>
> Scott Johnston wrote:
>
>> Hello - I have been using Domino and Perple_X to make pseudosections,
>> and I
>> have a couple of questions with regard to Domino:
>>
>> 1) In domino, is there any way to make isopleths for Mg content in
>> biotite?
>> I have successfully made isopleths for phlogopite, however, these
>> isopleths
>> exclude a lot of Mg that goes into the obiotite solution for biotite,
>> and
>> thus yields isopleths that are less than the true Mg content. Is
>> there any
>> way to get around this and create isopleths that reflect the total Mg
>> content?
>>
>> 2) Perple_X has several solution models for amphibole; I have been using
>> GlTrTsPg. In Domino, I have been using the database tcdb55c2.txt
>> which does
>> not have any amphibole solution models. Is there a database for
>> Domino that
>> I could use that incorporates a solution model for amphibole similar to
>> GlTrTsPg?
>> Thanks, Scott Johnston
>
>
|