Ａ ｄｕｍｂ ｑｕｅｓｔｉｏｎ：
ｗｈｅｎ ｙｏｕ ｄｏ ａ ＴＳＴ ｃａｌｃｕｌａｔｉｏｎ， ｓａｙ， ｐｌａｃｉｎｇ ａ ｍｏｌｅｃｕｌｅ ｏｖｅｒ ａ ｓｕｒｆａｃｅ， ｈｏｗ ｄｏ ｙｏｕ ｃｈｏｏｓｅ ｔｈｅ ｉｎｉｔｉａｌ ｓｔａｔｅ（ｄｏ ｙｏｕ ｊｕｓｔ ｒａｍｄｏｍｌｙ ｐｌａｃｅ ｔｈｅ ｍｏｌｅｃｕｌｅ ５Ａ ａｗａｙ ｆｒｏｍ ｔｈｅ ｓｕｒｆａｃｅ ｏｒ ｄｏｅｓ ｉｔ ｈａｖｅ ｔｏ ｂｅ ａ ｏｐｔｉｍｉｚｅｄ ｇｅｏｍｅｔｒｙ？）
About ＦＳ, a really naive question: anyone would like to share the trick or experience about choosing the interesting reaction pathway to look? cuz it seems to me that there are tons of possible reaction pathways--since you cannot explore each of them, how do you decide which ones you are gonna focus on
I hope my question is clear~~
Happy study~~
TST calculation IS and FS
Moderator: moderators
THE NEB is designed to find a minimum energy pathway between two stable (optimized) minima. This is necessary if you want to relate the barrier along a pathway to a rate.
The second question has more to do with the science than the method. You could find pathways between states which have nothing to do with what the system really does (dynamically). Sometimes, you can use your intuition to guess at important states, and the NEB to map out the energy landscape between the minima that you find. It is good, however, to keep in mind that you could be missing something important.
Alternative saddle point finding methods to the NEB are the min-mode following methods (dimer/Lanczos). This class of method allows you to start from an initial state and find saddle points leading away to various final states. If you find enough of these saddles, you can use kinetic Monte Carlo to simulate the state-to-state dynamics. We have recently implemented this method for use with vasp.
Another approach is to simply run molecule dynamics to see what happens. MD is typically limited to short times, but there are also accelerated dynamics methods (for example hyperdynamics, temperature accelerated dynamics, and parallel replica dynamics) which extend the time scale of these simulations. These are not currently availble in vasp, but they at least provide possible way to calculate molecular dynamics over time scales of chemical reactions.
The second question has more to do with the science than the method. You could find pathways between states which have nothing to do with what the system really does (dynamically). Sometimes, you can use your intuition to guess at important states, and the NEB to map out the energy landscape between the minima that you find. It is good, however, to keep in mind that you could be missing something important.
Alternative saddle point finding methods to the NEB are the min-mode following methods (dimer/Lanczos). This class of method allows you to start from an initial state and find saddle points leading away to various final states. If you find enough of these saddles, you can use kinetic Monte Carlo to simulate the state-to-state dynamics. We have recently implemented this method for use with vasp.
Another approach is to simply run molecule dynamics to see what happens. MD is typically limited to short times, but there are also accelerated dynamics methods (for example hyperdynamics, temperature accelerated dynamics, and parallel replica dynamics) which extend the time scale of these simulations. These are not currently availble in vasp, but they at least provide possible way to calculate molecular dynamics over time scales of chemical reactions.
Thx a lot ~~~
Exactly the kind of of answers i expect~~
Thx a lot for sharing~~~
and
--Happy study!
Thx a lot for sharing~~~
and
--Happy study!