Wikipedia:Reference desk/Archives/Science/2020 May 4
Science desk | ||
---|---|---|
< May 3 | << Apr | May | Jun >> | May 5 > |
Welcome to the Wikipedia Science Reference Desk Archives |
---|
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages. |
May 4
editPalladium pentafluoride
editI'm don't sure that PdF5 will have dark red in color. So what is the color of this compound? Thanks for much (Sorry if you don't understand, because my English is not good). --Ccv2020 (talk) 10:23, 4 May 2020 (UTC)
- If you have a reference for this compound even existing, please post here so we can track down other details. DMacks (talk) 10:30, 4 May 2020 (UTC)
- It is expected to disproportionate in to PdF4 and PdF6 and there is no good evidence that PdF6 actually exists. IR spectra have been calculated but not UV-vis (TD-DFT isn't very trustworthy anyway).Pelirojopajaro (talk) 10:49, 4 May 2020 (UTC)
- You've been asking a lot of questions about transition metal complex colors. Is there some specific goal we can help you with? It might help to know what you need this information for... --OuroborosCobra (talk) 15:07, 4 May 2020 (UTC)
- These questions make a lot of sense if someone is developing a model for predicting the colours of chemical compounds in their various states of matter. --Lambiam 08:35, 5 May 2020 (UTC)
- The problem is, that isn't actually possible. It's the quantum equivalent of solving the many body problem for a quantum system. It basically involves solving the wave function quantitatively for a system consisting of several nuclei and dozens of electrons. We don't even have a good quantitative solution for the Helium atom, and that's only a 3-particle system. There are things like the Hartree–Fock method, which provides a decent enough approximation of the wave function for complex systems, but not at the fine detail necessary to pick out exact colors. --Jayron32 13:15, 5 May 2020 (UTC)
- I think that may be understating modern capabilities a bit. I've done TD-DFT modeling of heme systems, and gotten pretty good estimations of at least the Soret band that were within about 10 nm of my experimental measurements. Granted, that was pi-pi* transition within the conjugated pi system of the ligand, but it was a far larger system than a helium atom. This was not a trivial calculation, mind you, and I needed access to a supercomputer cluster to do it, but it was possible. This is, of course, not the same as different colors arising from things like d-orbital splitting due to ligand interaction with transition metals... but sometimes, you don't even need TD-DFT for that. In a high spin ferric compound, for example, you can calculate the energies of the d-orbitals without examining excited states, as all of the iron 3d orbitals will be occupied by at least one electron in the ground state. Additionally, with enough similar experimental data of a particular transition metal and other ligands, you can make some decent qualitative guesses at color using spectrochemical series and applying ligand field theory. Is it perfect? No, but we are only talking qualitative. All of that said, I'm still unclear as to the goal of Ccv2020, as they seem to be randomly throwing out transition metal complexes (many of which haven't even existed at all) and asking us to guess the colors. You don't build a model by randomly throwing darts. --OuroborosCobra (talk) 14:57, 5 May 2020 (UTC)
- I'll have to defer to you on this then. It is clear you've done more detailed work in the field, mine is a more general exposure to it. It's plain you have access to better sources of information than I do, and I will defer to your knowledge on these matters. --Jayron32 17:56, 5 May 2020 (UTC)
- I agree that for these structually simple transition-metal cases, we can often approximate it without fully solving the many-body problem. A Tanabe–Sugano diagram or similar analysis can be a good approximation if some related cases are known. DMacks (talk) 05:22, 6 May 2020 (UTC)
- I think that may be understating modern capabilities a bit. I've done TD-DFT modeling of heme systems, and gotten pretty good estimations of at least the Soret band that were within about 10 nm of my experimental measurements. Granted, that was pi-pi* transition within the conjugated pi system of the ligand, but it was a far larger system than a helium atom. This was not a trivial calculation, mind you, and I needed access to a supercomputer cluster to do it, but it was possible. This is, of course, not the same as different colors arising from things like d-orbital splitting due to ligand interaction with transition metals... but sometimes, you don't even need TD-DFT for that. In a high spin ferric compound, for example, you can calculate the energies of the d-orbitals without examining excited states, as all of the iron 3d orbitals will be occupied by at least one electron in the ground state. Additionally, with enough similar experimental data of a particular transition metal and other ligands, you can make some decent qualitative guesses at color using spectrochemical series and applying ligand field theory. Is it perfect? No, but we are only talking qualitative. All of that said, I'm still unclear as to the goal of Ccv2020, as they seem to be randomly throwing out transition metal complexes (many of which haven't even existed at all) and asking us to guess the colors. You don't build a model by randomly throwing darts. --OuroborosCobra (talk) 14:57, 5 May 2020 (UTC)
- The problem is, that isn't actually possible. It's the quantum equivalent of solving the many body problem for a quantum system. It basically involves solving the wave function quantitatively for a system consisting of several nuclei and dozens of electrons. We don't even have a good quantitative solution for the Helium atom, and that's only a 3-particle system. There are things like the Hartree–Fock method, which provides a decent enough approximation of the wave function for complex systems, but not at the fine detail necessary to pick out exact colors. --Jayron32 13:15, 5 May 2020 (UTC)
- These questions make a lot of sense if someone is developing a model for predicting the colours of chemical compounds in their various states of matter. --Lambiam 08:35, 5 May 2020 (UTC)
This is the existence of this compound (also PdF6): [1] — Preceding unsigned comment added by Ccv2020 (talk • contribs) 07:42, 5 May 2020 (UTC)
- That's not the compound PdF5, that's the complex ion PdF51-. Ions are not compounds. --Jayron32 13:08, 5 May 2020 (UTC)
- Another entry is for the claimed product of the reaction as:
- Pd + KrF2 → PdF5
- in anhydrous HF or BrF5, described as "unconfirmed". It is cited to:
- Sokolov, V. B.; Drobyshevskii, Yu. V.; Prusakov, V. N.; Ryzhkov, A. V.; Khoroshev, S. S. (Dokl. Akad. Nauk SSSR 229 [1976] 641/4; Dokl. Chem. Proc. Acad. Sci. USSR 226/231 [1976] 503/5)
- I just did a lit search for PdF5 and see discussions of its structure and spectral features in: doi:10.1021/acs.inorgchem.5b02273, doi:10.1016/j.solidstatesciences.2012.08.024, doi:10.1021/jp1022949. DMacks (talk) 05:42, 6 May 2020 (UTC)
- Another entry is for the claimed product of the reaction as:
- That's not the compound PdF5, that's the complex ion PdF51-. Ions are not compounds. --Jayron32 13:08, 5 May 2020 (UTC)
Have you see page 48 in the link I'm added? --Ccv2020 (talk) 06:49, 6 May 2020 (UTC)
- That is exactly what I summarized in my comment about the Sokolov ref. DMacks (talk) 17:40, 6 May 2020 (UTC)