LO-TO splitting correction can activated by the keyword LO in the input.txt file; the keyword must be followed by the name of the file containing the LO-TO data: a first column specifying the mode numbers (for those modes affect by LO-TO splitting only) and (second column) the corresponding values of the split (in $cm^{-1}$). The present example concerns pyrope and such file is named LO.txt
The presence of the LO keyword in the input file produces corrected frequencies for the modes affected by LO-TO splitting. Precisely, the new frequencies $\nu_c$ are computed as follows
$$\nu_c = 2/3\cdot \nu_{TO} + 1/3\cdot \nu_{LO}$$where $\nu_{LO}=\nu_{TO}+\Delta\nu$ and $\Delta\nu$ is the LO-TO split value. The TO values are found in the frequencies file specified by the keyword FREQ.
The corrected frequencies $\nu_c$ are internally stored in the class lo, along with a copy of the original TO frequencies, and used in all the subsequent calculations instead of the TO frequencies.
Start the program as usual:
%run bm3_thermal_2.py
The info.show command print the information concerning the LO-TO splitting:
info.show()
Compute the equation of state (here, the bulk_dir is used).
Note: The LO-TO split values are considered to be constant as the cell volume is changed; the impact that such LO-TO splitting correction might have on the EoS (or on the thermal expansion) is therefore not due to possible different Gruneisen parameters with respect to those corresponding to the TO modes; rather, it is due to the generally different phonon populations of the splitted modes as an effect of their different frequency values.
bulk_dir(300)
Compute thermal expansion, specific heat and entropy (Kp is fixed to the optimized value from bulk_dir)
set_fix(4.11)
thermal_exp_p(300,0)
compare_exp()
The lo.off method can be used to switch off the LO-TO splitting correction (the original TO frequencies are restored). The method lo.on can be used to switch on again the correction.
lo.off()
reset_fix()
info.show()
Then repeat the computation above without the LO-TO splitting correction:
bulk_dir(300)
set_fix(4.11)
thermal_exp_p(300,0)
compare_exp()
As shown by the results, the correction had a very little impact on bulk modulus and themal expansion, however the entropy and the specific heat changed by some units: In particular, at 298.15 K, the entropy decreases from 267.02 J/mol K, without correction, to 265.65 J/mol K, with correction.