MELTS Software Users

A list of possible mineral phases that can be stabilised in the MELTS family of algorithms was requested ages ago (when I was on leave) and I totally forgot until now. So sorry for the delay in getting round to it...

The available solid solution phases and their end-members are listed below, in the order in which they are usually output. The solution models are exactly the same in the MELTS and pMELTS models (and therefore also in pHMELTS, which is based on pMELTS). For the relevant references see here (http://melts.ofm-research.org/CORBA_CTserver/phaseProp.html) and here (http://melts.ofm-research.org/CORBA_CTserver/Webservices/CTserverPLWS.html). Some of the end-member thermodynamic data differ slightly between MELTS and pMELTS, as described in Ghiorso et al., 2002 (http://dx.doi.org/10.1029/2001GC000217).

Quote

• olivine
  

  tephroite	Mn2SiO4
  fayalite	Fe2+2SiO4
  Co-olivine	Co2SiO4
  Ni-olivine	Ni2SiO4
  monticellite	CaMgSiO4
  forsterite	Mg2SiO4

• garnet
  

  almandine	Fe2+3Al2Si3O12
  grossular	Ca3Al3Si3O12
  pyrope	Mg3Al2Si3O12

• melilite
  

  akermanite	Ca2MgSi2O7
  gehlenite	Ca2Al2SiO7
  Fe-akermanite	Ca2Fe2+Si2O7
  soda-melilite	Na2Si3O7

• orthopyroxene (see note * (#post_amphibole); see note | (#post_pyx))
  

  diopside	CaMgSi2O6
  clinoenstatite	Mg2Si2O6
  hedenbergite	CaFe2+Si2O6
  alumino-buffonite	CaTi0.5Mg0.5AlSiO6
  buffonite	CaTi0.5Mg0.5Fe3+SiO6
  essenite	CaFe3+AlSiO6
  jadeite	NaAlSi2O6

• clinopyroxene (see note * (#post_amphibole); see note | (#post_pyx))
  

  diopside	CaMgSi2O6
  clinoenstatite	Mg2Si2O6
  hedenbergite	CaFe2+Si2O6
  alumino-buffonite	CaTi0.5Mg0.5AlSiO6
  buffonite	CaTi0.5Mg0.5Fe3+SiO6
  essenite	CaFe3+AlSiO6
  jadeite	NaAlSi2O6

• cummingtonite
  

  cummingtonite	Mg7Si8O22(OH)2
  grunerite	Fe2+7Si8O22(OH)2

• clinoamphibole (see note * (#post_amphibole))
  

  cummingtonite	Mg7Si8O22(OH)2
  grunerite	Fe2+7Si8O22(OH)2
  tremolite	Ca2Mg5Si8O22(OH)2

• orthoamphibole (see note * (#post_amphibole))
  

  cummingtonite	Mg7Si8O22(OH)2
  grunerite	Fe2+7Si8O22(OH)2
  tremolite	Ca2Mg5Si8O22(OH)2

• hornblende
  

  pargasite	NaCa2Mg4AlAl2Si6O22(OH)2
  ferropargasite	NaCa2Fe2+4AlAl2Si6O22(OH)2
  magnesiohastingsite	NaCa2Mg4Fe3+Al2Si6O22(OH)2

• biotite
  

  annite	KFe2+3Si3AlO10(OH)2
  phlogopite	KMg3Si3AlO10(OH)2

• feldspar
  

  albite	NaAlSi3O8
  anorthite	CaAl2Si2O8
  sanidine	KAlSi3O8

• nepheline
  

  na-nepheline	Na4Al4Si4O16
  k-nepheline	K4Al4Si4O16
  vc-nepheline	Na3Al3Si5O16
  ca-nepheline	CaNa2Al4Si4O16

• kalsilite
  

  na-nepheline	Na4Al4Si4O16
  k-nepheline	K4Al4Si4O16
  vc-nepheline	Na3Al3Si5O16
  ca-nepheline	CaNa2Al4Si4O16

• leucite
  

  leucite	KAlSi2O6
  analcime	NaAlSi2O5(OH)2
  na-leucite	NaAlSi2O6

• spinel
  

  chromite	Fe2+Cr2O4
  hercynite	Fe2+Al2O4
  magnetite	Fe2+Fe3+2O4
  spinel	MgAl2O4
  ulvospinel	Fe2+2TiO4

• rhm-oxide
  

  giekelite	MgTiO3
  hematite	Fe3+O3
  ilmenite	Fe2+TiO3
  pyrophanite	MnTiO3

• ortho-oxide
  

  psuedobrookite	Fe3+2TiO5
  ferropseudobrookite	Fe2+Ti2O5
  karrooite	MgTi2O5

• alloy-solid
  

  Fe-metal	Fe
  Ni-metal	Ni

• alloy-liquid
  

  Fe-liquid	Fe
  Ni-liquid	Ni

* Note that in some MELTS software there is just one amphibole solid solution, rather than separate clino- and ortho- phases. Also, early releases of MELTS (e.g. prior to the development of the pMELTS algorithm) had a phase called 'pyroxene'. Although all current versions of MELTS have 'clinopyroxene' and 'orthopyroxene' you may see 'pyroxene' in some older documentation (http://melts.ofm-research.org/manual.html).

| Note that the compositions of the essenite, buffonite and alumino-buffonite end members mean that calculations for bulk compositions that contain just two of Fe₂O₃, TiO₂ and Al₂O₃ are likely to fail (e.g. no TiO₂ in the bulk requires that buffonite and alumino-buffonite cancel out exactly).

Although the solution models are the same in all the algorithms, the names of a few solid solutions differ between MELTS and pMELTS when using the graphical user interfaces (i.e. standalone MELTS, CORBA MELTS or Java MELTS):

Quote

MELTS name	pMELTS name
nepheline	nepheline ss
leucite	leucite ss
kalcilite	kalcilite ss
ortho-oxide	ortho oxide
rhm-oxide	rhm oxide

Note: alphaMELTS (and adiabat_1ph) can accept either form in files - its input routine replaces spaces or hyphens within phase names with underscores, before making any string comparisons. When entering phase names at the menu prompt (e.g. for phase diagram mode) use the MELTS form to avoid problems with spaces.

Note also: early releases of MELTS (e.g. prior to the development of the pMELTS algorithm) used what are now the pMELTS names so that the input file (http://melts.ofm-research.org/Melts-ftp/Morb.melts) that accompanies this old tutorial (http://melts.ofm-research.org/Manual/UnixManHtml/examples-test.html) will cause read errors if used in the current graphical user interfaces without modification.

The pure phases included in all versions of MELTS and pMELTS are shown below (the phases in brackets are the ones that appear immediately before and after, in case you are trying to anticipate the output order):

Quote

sphene	CaTiSiO5	(olivine; garnet)
aenigmatite	Na2Fe5TiSi6O20	(clinopyroxene; cummingtonite)
muscovite	KAl2Si3AlO10(OH)2	(biotite; feldspar)
quartz	SiO2	(feldspar; tridymite)
tridymite	SiO2	(quartz; cristobalite)
cristobalite	SiO2	(tridymite; nepheline)
corundum	Al2O3	(leucite; rutile)
rutile	TiO2	(corundum; perovskite)
perovskite	CaTiO3	(rutile; spinel)
whitlockite	Ca3(PO4)2	(ortho-oxide; apatite)
apatite	Ca4(PO4)3(OH)	(whitlockite; water)
water	H2O	(apatite; alloy-solid)

alphaMELTS (and adiabat_1ph) has these extra pure phases, as they are needed for the pHMELTS model. They all appear at the end of the output list:

Quote

andalusite	Al2SiO5
sillimanite	Al2SiO5
kyanite	Al2SiO5
coesite	SiO2
talc	Mg3Si4O10(OH)2
anthophyllite	Mg7Si8O22(OH)2
cordierite	Mg2Sl4SI5O18
lawsonite	CaAl2Si2O6(OH)4
antigorite	Mg48Si34O85(OH)62
chrysotile	Mg3Si2O5(OH)4
brucite	Mg(OH)2

Hope that helps.  If you spot any errors or omissions please let me know.

Paula

Mullite, no possible??

There is pure Al₆Si₂O₁₃ mullite in the Berman (1988) dataset. I can add that to the next alphaMELTS 2 / alphaMELTS for MATLAB/Python releases (coming in the next few weeks) if it will help.

However, natural mullites are rather variable in composition. Without accounting for the excess Al₂O₃, it is likely that mullite stability in MELTS will be underestimated. We were vaguely thinking about adding a mullite solid solution model a while back - before Covid - so I suppose we could revisit that idea. It would require at least one ordering parameter, so not completely trivial to implement.

We are busy with the upcoming alphaMELTS 2 workshop (https://magmasource.caltech.edu/workshops/goldschmidt/) and I have other things lined up for July / early August. But if you give me a little more information on what you're trying to model I'll consider putting mullite ss on the to-do list for after that!

Paula