Lets assume you need the initial model for geometry optimization, and you would like to use the crystallographic data published somewhere before instead of crafting your own structure based on certain assumptions. I would suggest the following:
1. Look for the original paper. Usually corresponding structural data is provided as CIF within supplementary information. It is particularly handy, as quite often there are some useful comments in the paper itself on how the compound has been synthesized (solvent, temperature, sometimes even pressure are going be used for quantum simulation) and how the structure has been determined.
2. Structural databases.
2.1. The Cambridge Crystallographic Data Centre (CCDC) - http://www.ccdc.cam.ac.uk - provide access to the one of the most comprehensive (about 600k structures) collection of organic-inorganic structures. Access to CSD is not free of charge for you unless the university pays for it. There is a standalone CSD-package including quite powerful tools to search for the structure, and a web-based version http://webcsd.ccdc.cam.ac.uk. As the majority of polyoxometalates are not pure inorganic compounds and being stabilized with organic ligands, most likely you will find your structure here.
2.2. The database for inorganic crystal structures (ICSD) - http://icsd.fiz-karlsruhe.de/ - gives access to about 160k structures of inorganic compounds and minerals. There is also a standalone version called FindIt! and a web-based mirror. Not free of charge.
2.3. Free structural databases such as Crystallography Open Database (COD) - http://www.crystallography.net/ - includes access to about 260k structures. Personally, I would use it as the last resort as many structures are missing, but still it is a great alternative to 2.1. and 2.2. Free of charge.
If you would like to craft your own CIF based on certain assumptions, then you most likely would like to use molecular editor such as GaussView or Avogadro. If the structure is small enough, I guess one might be able to create a direct input for Gaussian in GJF simply using z-matrix.
Lets assume you need the initial model for geometry optimization, and you would like to use the crystallographic data published somewhere before instead of crafting your own structure based on certain assumptions. I would suggest the following:
1. Look for the original paper. Usually corresponding structural data is provided as CIF within supplementary information. It is particularly handy, as quite often there are some useful comments in the paper itself on how the compound has been synthesized (solvent, temperature, sometimes even pressure are going be used for quantum simulation) and how the structure has been determined.
2. Structural databases.
2.1. The Cambridge Crystallographic Data Centre (CCDC) - http://www.ccdc.cam.ac.uk - provide access to the one of the most comprehensive (about 600k structures) collection of organic-inorganic structures. Access to CSD is not free of charge for you unless the university pays for it. There is a standalone CSD-package including quite powerful tools to search for the structure, and a web-based version http://webcsd.ccdc.cam.ac.uk. As the majority of polyoxometalates are not pure inorganic compounds and being stabilized with organic ligands, most likely you will find your structure here.
2.2. The database for inorganic crystal structures (ICSD) - http://icsd.fiz-karlsruhe.de/ - gives access to about 160k structures of inorganic compounds and minerals. There is also a standalone version called FindIt! and a web-based mirror. Not free of charge.
2.3. Free structural databases such as Crystallography Open Database (COD) - http://www.crystallography.net/ - includes access to about 260k structures. Personally, I would use it as the last resort as many structures are missing, but still it is a great alternative to 2.1. and 2.2. Free of charge.
If you would like to craft your own CIF based on certain assumptions, then you most likely would like to use molecular editor such as GaussView or Avogadro. If the structure is small enough, I guess one might be able to create a direct input for Gaussian in GJF simply using z-matrix.
I think that this cif-file will be more preferable for you because in this structure the hydrogen atoms were determined in a difference synthesis and refined in an isotropic approximation with geometrical limitations
(Journal of Structural Chemistry. Vol. 50, No. 2, pp. 296-305, 2009)
Thanks for the above two CIFs which I uploaded to the COD - http://www.crystallography.net/
Anyway, the best would be that every crystallographer upload himself his data... So the "I would use it as the last resort as many structures are missing" would no longer be true.