I am trying to generate the litharge (PbO) Raman spectra in GULP, to compare with the experimental Raman data found from Litharge - RRUFF Database: Raman, X-ray, Infrared, and Chemistry.
Experimental data shows that there should be a spectra, however I keep getting zero for all values in the Raman susceptibility tensors, and for the values in the IR and Raman intensities. I know from the manual that I need a shell model and two- three- or four- body potentials for the Raman susceptibility to not be zero, which I think I have. Is there maybe an obvious rookie error in my input file?
I have attached my input file;
opti prop phon conv noden inten Raman
Pb core 0.0
O core 0.0
O shell 0.0
3.9744 3.9744 5.0220 90.0 90.0 90.0
O shell O shell 22764.0000000 0.149000 27.879000 0.0 12.0 1 0 0
O shell Pb core 2000.0000000 0.205000 50.6600000 0.0 12.0 1 0 0
Pb core Pb core 2825.27046900 0.365590 50.000000 0.0 12.0 1 0 0
O 74.92 0
O c 0.750000000 0.750000000 1.000000000
O c 0.250000000 0.250000000 0.000000000
Pb c 0.750000000 0.250000000 0.235800000
Pb c 0.250000000 0.750000000 0.764200000
O s 0.760000000 0.740000000 0.999000000
O s 0.240000000 0.260000000 0.000100000
Thank you for your time!
I am sure Julian will reply as soon as the time difference allows (most likely pointing out that there are zero charges in your model). FWIW, this is a small unit cell so quantum-mechanical calculations are very cheap. A quick CASTEP run gives the peaks at 145 and 342 cm-1, in very good agreement with the spectrum data at https://rruff.info/repository/sample_child_record_raman/by_minerals/Litharge__R060959__Raman__532__0__unoriented__Raman_Data_Processed__13392.txt (145.6 and 336.5). Calculated intensities are about 3:1 for the two peaks, while it’s more like 10:1 in the experiment. One can look deeper in the peak width and maybe compare integrated peak areas or try to extract Raman activities. Anyway, it seems like a pretty good result in about 10 minutes.
Victor is absolutely correct - if you have zero charge then not only would your model not work and optimise the structure, you are guaranteed to have zero intensity for the IR spectra since this depends on the change in dipole with the vibrational mode & if there are no charges you can’t have a dipole. Raman is a bit more subtle, but if you have no shell charge you will again have no intensity. You should note that PbO is a structure that most force field models will fail to describe because of the lone pair effect which goes beyond simple classical models & so you would need to develop a quite specialist model to handle this case. Therefore I again agree with Victor that using a quantum mechanical approach, such as CASTEP, makes far more sense for such materials with very small unit cells.