How can we investigate the adsorption mechanism of a gas molecule on the substrate surface (e.g. Silicon) using Gaussian software? By using QST2/3 and IRC only?
In order to use QST2/3, we need to insert final state of the adsorption process. But what if I do not know how the gas molecule will adsorb on the surface (for example: which bond will be broken first and which side will bond to the substrate surface; will it be dissociative adsorption?), what other method can i use?
Thank you.
[Remark: my own experience in mechanism modelling with Gaussian is a bit dated, maybe they improved it in the meantime.]
In mechanism modelling, you are always confonted with the backlash that you actually need an idea where you are going. The same is the case for the transistion state for which you have to provide a guessed geometry in QST3 and almost every other method Gaussian provides (not in QST2, but in my experience that barely works). If you have a lot of CPU power, you might consider a surface MD to find several states, but that's a really expensive calculation plus I'm ot sure whether Gaussian provides that option.
Instead of QST you can also combine all the other algorithms Gaussian provides for transition state searches. Back then I used the Berny algorithm (Gaussian standard, conjugate gradient method) and Eigenvector following (another cg method) and combined those with IRC.
In order to find out what newer methods there might be you'll actually have to go through the not exactly compact Gaussian manual.
Apart from that: does the Imperial College also provide access to other software? After all, Gaussian is not exactly known for fast calculations or as a great surface QC package.
you can the draw various possibility for the TS structure and find out the which one will be the correct using the IRC calculation.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Physical Phenomena
- Luminescence
- Fluorescence