PC GAMESS/Firefly-related discussion club

Learn how to ask questions correctly

Alex Granovsky
gran@classic.chem.msu.su

I'm sorry for some delay...

>Great help, as always. The decomposition reactions I am interested in occur in catastrophic events (deflagrations, explosions), so the pressure generated is usually very high. From brief search on the RRKM theory, at high pressures it is equivalent to the transition state theory (for example see here: http://vallance.chem.ox.ac.uk/pdfs/UnimolecularReactionNotes.pdf Is this correct?

yes this is correct.

>>Another point is that you need to be able to describe barriers
>>of different pathways of decomposition within approximately the same
>>accuracy...

>When I use the same level of theory and thight convergence criteria I should be able to get to the same accuracy of enthalpy. The entropy error could be different for different modes because of internal rotations. I was planning to correct that using rotation barriers calculated from unrelaxed PES for all rotors. Is this close enough?

I think it is... However, I see some problems with enthalpy.

First, as you know DFT usually tends to underestimate barriers...,
while HF tends to overestimate them. This error can vary!

Actually, to describe barriers correctly (at least if true chemical
bonds are to be broken and/or formed) and with high precision is a
challenge - and most of single-reference methods are not
working well here... I'd still suggest to use DFT for your
purposes, perhaps adding some (empirical) corrections to the
computed activation energies.

>And last, which reaction paths would I miss?

It's very hard to predict exactly. However, there is always some
finite probability that the decomposition channel that is not
formally allowed because of its high activation energy will get
enough energy as the result of constant redistribution of the
vibrational excitation over (approximate) modes due to
anharmonicity of PES. Anyway, it's a good idea to consider not
only the reactions with barriers lower that e.g. 1000 K but
also these that are above but close enough.

Alex


>Slawomir

>On Tue Oct 20 '09 6:41pm, Alex Granovsky wrote
>----------------------------------------------
>>Hi,

>>>My initial idea is to follow IRC along all vibrations of the starting molecule. Most vibration probably would not result in a decomposition under these conditions and I would concentrate on those with activation energy under 1000 K. Do you think this is reasonable? Would I miss some paths this way?

>>Just my two cents:

>>The idea isn't too bad, however you do can miss some pathways of
>>decomposition. The unimolecular decay is usually a complex process,
>>which includes bimolecular activation stages (e.g., well-known
>>RRKM theory is one of the  possible theoretical approaches to
>>unimolecular reactions) and then the redistribution of
>>vibrational energy.

>>Another point is that you need to be able to describe barriers
>>of different pathways of decomposition within approximately the same
>>accuracy...

>>Best,
>>Alex Granovsky

>>