Author Topic: Hydrogen bond detection  (Read 2694 times)

0 Members and 1 Guest are viewing this topic.

Offline alex_z

  • Regular QuantumATK user
  • **
  • Posts: 9
  • Reputation: 0
    • View Profile
Hydrogen bond detection
« on: March 1, 2011, 00:04 »
Dear support staff,

I have two questions.

1. I wonder if there is any reliable way to detect hydrogen bond length. I simulate a molecule covalently bonded to the left electrode and attracted to the right hydrogen-terminated carbon electrode. The way I do it is adjust the distance between molecule and the right electrode in hope to see the minimum in total energy. To get the minimum I optimize the system and get E_tot. Then to remove the influence of electrodes I perform SCF without a molecule for the same geometry of electrodes and subtract the total energies. I expect to see the minimum  but the problem is that the energy difference oscillates or stays flat for larger distances. Also, even if I manage to find the minimum this way, hydrogen bond length is too large, of the order of 2.5 Angstroms or even larger.

Also, is there a way to increase the precision of the Total energy? For large systems I have only one digit after decimal point for E_tot and even if there is a minimum, I cannot see it, because E_tot does not change when I move right electrode. Can I get, let's say, three digits after decimal point regardless of how large my system is?

To sum up, is it possible to get correct H-bond length with ATK 10.8.2?   Today I read that in new version ATK 11.2 the distance between electrodes can be optimized together with the central region which means automatic detection of proper bond lengths on both sides. Will it pick up the correct hydrogen bond length?

2. In new ATK 11.2 you added the new feature of calculating the complex band structure. How do you return results for it? Can I save data points and approximate particular branch by some function for imaginary k. I need it to extract effective mass of electron.

Thanks in advance.