ERROR: Could not allocate MD atoms to QM atoms or weights in fragment T5H1 in segment T5H1 in molecule T5H1 due to inconsistent number of columns (MD: 93, QM: 93) Weights: 0).

[王建安]

I encountered a problem when I use ctp as follows:
ERROR: Could not allocate MD atoms to QM atoms or weights in fragment T5H1 in segment T5H1 in molecule T5H1 due to inconsistent number of columns (MD: 93, QM: 93) Weights: 0).
NOTE: To define an MD atom without QM counterpart, insert a single ':' in the associated QM-atoms column and specify a mapping weight of 0.
an error occurred:
Inconsistency in mapping file.
I compared the sample mapping.xml, and found that the mapping.xml I prepared has not </weights> . Please see attachment!
I guess that I know how to define weights, I should define them by myself ?

[Denis Andrienko]

Yes, I use pdb2map to construct mapping file from .gro and .xyz file. I found and solved the problem. In the .xyz file,  atom label is like H12 C13 C14, there are unnecessary numbers, so I failed to prepare right mapping file.

Excellent. Atom elements are indeed taken from the xyz file.

a) There is another simple but somewhat uncertain question! If an organic molecular semiconductor contains over 100 atoms, and is also a conjugated molecule,  I don't partition it into different fragments and treat it as a whole. This opinion is right? The partitioning is optional?

If it is one single conjugated molecule, it is Ok. If it has very anisotropic shape, you might want to split it on a few fragments, since the neighborlist is constructed based on the center of mass distance between the fragments.

b) To run the commend of kmcmultiple, it often use one thread, can we add several threads to run the task?

Normally one thread is enough. If you run into problems, either your energetic disorder is too high (more processors will not solve this situation) or you have one-two deep traps and you need another MD snapshot. Have you looked at the distribution of site energies?

c) kmc calculation is time-consuming? I run an amorphous cell containing 100 molecules (100 atoms as a segment, not partitioning), and find that several hours have been spent and it still runs in one thread.

As I said, look at your energetic disorder and at the KMC trajectory. KMC simulations are normally not the bottleneck.

Best,
Denis

[王建安]

Thank you very much!

I have understood  and figured out these problems.

Best wishes!

在 2020年4月14日星期二 UTC+8上午10:14:35，王建安写道：