Calculation of IR spectra using VMD IRspec Plug-in

Water molecule and four wannier centers
Water mol­e­cule and four wan­nier cen­ters, cor­re­spond­ing to the two lone pairs and two bonds.

The pur­pose of this tuto­r­ial to illus­trate the basic capa­bil­i­ties of VMD IRspec plug-in. On exam­ple of the Wan­nier cen­ter dynam­ics from Car-Parrinello sim­u­la­tion we will cal­cu­late IR spec­tra of water mol­e­cule. Details of CPMD sim­u­la­tions are avail­able as a sep­a­rate tuto­r­ial.

First, we will need IONS+CENTERS.xyz (10mb) from that tuto­r­ial. It includes tra­jec­tory of the ionis and the Wan­nier cen­ters
(labeled as dummy atom X) in XYZ for­mat. After proper visu­al­iza­tion our water mol­e­cule will looks close to the pic­ture on the right. Where Wan­nier cen­ters rep­re­sented by trans­par­ent green spheres.
The final visu­al­iza­tion state water_wannier.vmd file is avail­able.
The upper two Wan­nier cen­ters rep­re­sent two lone pairs of oxy­gen. The other two cen­ters rep­re­sent bonds between oxy­gen and hydro­gen. Please, note there are only 8 elec­trons in our sys­tem since we use pseudopo­ten­tials to describe core elec­trons.
Finally, one can enjoy the dynam­ics on his own.

[main GUI window]
Here is the main GUI win­dow of the IR Spec plug-in.

We need to sup­ply sev­eral para­me­ters from our sim­u­la­tions, namely Time between frames (sim­u­la­tion timestep), Tem­per­a­ture, max­i­mum fre­quency, type of Cor­rec­tion, etc. In order to get the spec­tra, charges have to be assigned for all Wan­nier cen­ters and atoms of inter­est. Despite of clas­si­cal mol­e­c­u­lar dynam­ics there is no topol­ogy file avail­able. The guess from CHARMM does not work, because the names
of the ele­ments do not match. There­fore we have to pre­pare charge.dat file by our­selves. It will looks like following:

# Charges for water molecule + Wannier centers
  O         6.0
  H         1.0
  H         1.0
  X        -2.0
  X        -2.0
  X        -2.0
  X        -2.0

Where num­bers 6.0 and 1.0 rep­re­sent effec­tive atomic charges of cor­re­spond­ing “pseudoatoms”. Hence, the value of this charge for oxy­gen does not cor­re­spond to its true atomic charge 8. As we men­tioned before, two core 1s elec­trons rep­re­sented by pseudopo­ten­tial. The charge of each Wan­nier cen­ter either –2.0 (two elec­trons) or –1.0 in case of spin polar­ized cal­cu­la­tions.
Charge file should be loaded via “Util­i­ties” –> “Load name” -> :Charge map from file” menu in the GUI.

[Water molecule and four wannier centers]
Mul­ti­polt graph of the cal­cu­lated spec­tra. Click on pic­ture to magnify.

You can option­ally select Save to File check­box in order to export spec­tral data in text for­mat. The IR spec­tra will be plot­ted by Mul­ti­plot plug-in.
One might export spec­tra into Post­Script for­mat or Grace pro­gram (on *nix sys­tems) via File menu.
Due to sev­eral rea­sons, our spec­tra is quite coarse and noisy. The water mol­e­cule is extremely small sys­tem, there­fore the ther­mal fluc­tu­a­tions dur­ing mol­e­c­u­lar dynam­ics are high. Time scale of the sim­u­la­tion is small (just few picosec­ond). The fic­ti­tios masses dis­tort the forces, shift­ing some fre­quen­cies by as much as few hudreds wavenum­bers. Partly, this can be resolved by using Born-Oppenheimer Mol­e­c­u­lar Dynam­ics (BOMD), that is far more accu­rat than CPMD.
The lower the fre­quen­cies the longer time you have to sam­ple them. For the fre­quen­cies above 1000 wavenum­bers the tra­jec­tory about 20ps should be suf­fi­cient. Fur­ther­more, the broad group of peaks in 0–500 cm–1 range due to mol­e­c­u­lar libra­tions and does not con­tribute to real IR spec­tra. These motions can be sub­tracted by spe­cial tech­nique [link ??], how­ever this goes beyond the scope of present tutorial.

In the spec­tra we can imme­di­ately iden­tify three modes of water: asym­met­ric stretch in the region 1400–1800 cm–1 and sym­met­ric stretch+bend around 3600 wavenum­bers. These sig­nals are shifed com­par­ing to the exper­i­men­tal val­ues cm–1 1595, 3657 and 3756 cm–1 respec­tively. (DOI here)

For water mol­e­cules it is also pos­si­ble to use par­tial charges from e.g. SPC clas­si­cal water poten­tial (H 0.42; O –0.84). BTW, very nice intro­duc­tion into avail­abe water mod­els is given by Mar­tin Chap­lin. This approach also gives you a rea­son­able approx­i­ma­tion of the spectrum.

[main GUI window]

Com­par­i­son of final spec­tra with dif­fer­ent charge mod­els. Click on pic­ture to magnify.