Normal Mode Analysis at non-equilbrium structures

[DinosaurChemi]

Hello,
My question stems from a research project of mine involving taking the normal modes of clusters of molecules from a potential. I was curious of the interpretation that can be made of normal mode vectors at a non-equilibrium structure, i.e something following a MD trajectory beginning at equilibrium. Following the normal mode analysis of a such a structure the vectors resemble those of the system at equilibrium. I understand that the lower frequency modes with imaginary frequencies indicate a transitional mode but do the higher frequency retain any meaning full description of the dynamics? Has anyone see this type of procedure used before?

Thanks for your time

 

[DrDu]

The dynamics is mostly described by the non-vanishing gradients and not by their frequency. However, the high frequency modes will more or less adiabatically follow the slow modes' dynamics, i.e. they will remain more or less in equilibrium and the change of their zero point energy will create an effective potential for the slow modes.

 

[JohnRC]

Hello,
My question stems from a research project of mine involving taking the normal modes of clusters of molecules from a potential. I was curious of the interpretation that can be made of normal mode vectors at a non-equilibrium structure, i.e something following a MD trajectory beginning at equilibrium. Following the normal mode analysis of a such a structure the vectors resemble those of the system at equilibrium. I understand that the lower frequency modes with imaginary frequencies indicate a transitional mode but do the higher frequency retain any meaning full description of the dynamics? Has anyone see this type of procedure used before?

Thanks for your time

I am fairly familiar with a similar but simpler approach. It involves changes in the normal mode structure of a single polyatomic molecule along a reaction co-ordinate pathway. That consists in taking as one of the modes of motion the gradient vector at a point along the minimum potential energy valley for the reaction, and then performing a normal mode diagonalization of the remaining directions of motion for the effective potential surface at that point. Changes in the composition of normal modes in this sort of approach can provide interesting information about the statistics of a reaction system -- particularly about the partitioning of energy into translation versus vibration of the products of a dissociation (1 into 2) reaction for a medium sized (5-20 atom) polyatomic molecule or ion.