APATITE THERMODYNAMIC MODEL
FOR CALCULATING WATER CONCENTRATIONS IN MAGMAS

Instruction:

This model can be used to calculate the abundances of water in silicate melts using concentrations of volatiles (F, Cl, H2O) in apatite, and the equations for the exchange coefficients (KD) for OH-Cl, OH-F, and Cl-F proposed by:

Li W. and Costa F. (2020). A thermodynamic model for F-Cl-OH partitioning between silicate melts and apatite including non-ideal mixing with application to constraining melt volatile budgets. Geochimica et Cosmochimica Acta 269, 203-222. DOI: 10.1016/j.gca.2019.10.035

[Updates: functions in this webpage are included in a python package called "pyAp" at Github (github.com/alexweiranli/pyAp) since February 2022. There's also an excel version that can be requested from Alex (Weiran Li) at: wl413@cam.ac.uk]


1. Calculating chemical formula of apatite:

Enter the measured apatite compositions
Boxes marked as "required" need to be filled in for calculation, whereas those marked as "optional" can be left as empty if the values are unknown.

Required:

 wt. % 
 wt. % 
 wt. % 
 wt. % 

Optional:

 wt. % 
 wt. % 
 wt. % 

Optional:

 wt. % 
 wt. % 
 wt. % 
 wt. % 
 wt. % 
 wt. % 
 wt. % 
 wt. % 
 wt. % 

*If H2O in apatite was not directly measured, leave this box empty, so that our model will calculate it based on stoichiometry, assuming that only F, Cl and OH occupy the anion site.

Chemical formula: Ca10(PO4)6X2
(Using 26 Oxygen)

0
0
0
0



Mole Fraction 

0
0
0
0

**Calculated based on stoichimeotry (e.g Deer et al., 1992)




2. Calculating exchange coefficients and volatile ratios in melt using known temperature

 oC    


Exchange coefficients

KD(OH-Cl)0
KD(OH-F)0
KD(Cl-F)0

Activity coefficients

γ(OH)Ap0
γ(F)Ap0
γ(Cl)Ap0

Calculated volatile ratios in melt using apatite:

0
0
0

3. Calculating H2O in melt using known melt Cl and/or F concentrations

Fill in at least one box below:

 ppm 
 ppm 

Dacitic-rhyolitic magma
Basaltic magma††

0wt. % (Error: ± 30-40%)

0wt. % (Error: ± 30-40%)

Water speciation models from Zhang et al. (1997) and Lesne et al. (2010) ††