Hello all:
I want to create amorphous silica (SiO2) and amorphous hafnia (HfO2) and further combine them to perform annealing, using COMB potential. My doubt is regarding the fix qeq/comb command. The following are the concerns.
Do I need to use this fix in conjunction with other fixes while I do the melting, quenching, equilibration, etc.? If that is the case, the computational time will be very high. I did it the other way (by using this fix at the beginning of the minimization) and find that the charges are not getting varied during the subsequent fix runs.
After creating both amorphous structures (without using one fix qeq/comb for each MD run), I tried to make a system with one on top of the other using the PBC. It turns out that, at the interface, the two constituents do not have enough adhesion that one is moving away from the other, even during the NVE run.
I am relatively new to this potential and do not have an idea on how to proceed, despite reading several references.
Please suggest, how I can improve the system?
Thanks in advance.
Yes, fix qeq/comb is an essential component of the COMB potential. … and yes, to slows down things substantially. That is the price to pay for the being able to support systems that are difficult to model well with simpler potentials.
It is impossible to provide a meaningful response to this due to the very limited information provided.
Dear Dr. Axel:
Thank you for your response.
I created two amorphous structures (SiO2 and HfO2) using COMB. I prepared a composite kind of structure, the bottom part is a-SiO2 and the top part is a-HfO2, using the periodic boundary conditions. The top layer is created using the read_data and shift command by shifting the top layer to N units, where N is the size of the bottom layer. Everything goes well until minimizing the structure. When I do an NVE run, the temperature shoots up and the interface deteriorates. Although this implies a bad system, I would like to hear from you the general considerations (if not for this specific problem) that I need to implement for an interface problem.
I understand that still you will not be able to make out exactly what the problem is, even from this description. If you wish to see my script, I would be willing to share it.
Thank you so much for your time and consideration.
Kiran
Exactly, somehow you create a situation where atoms are too close to another - be it directly or through periodic boundaries - and that causes the symptoms you describe.
One thing you could try to do is to write_data just before you start the minimisation, and inspect that configuration visually (for example in Ovito or VMD).