The Embedded Atom Method (EAM) potential is a widely used interatomic potential for modeling metallic systems. It is particularly effective for describing the properties of metals and alloys, including their mechanical behavior, defects, and surface properties. However, its validity under low-pressure conditions depends on the specific system and the parameters used in the potential.

For some materials, low-pressure conditions may lead to phase transitions or changes in bonding behavior that the EAM potential might not accurately describe. For example, EAM potentials are less reliable for materials with significant covalent or ionic bonding contributions.

EAM potentials are semi-empirical and rely on approximations. They may not fully capture electronic effects or long-range interactions that become significant at very low pressures.

Dear Jamal Zrinej ,

Thank you very much for your reply. Do you have any suggestions for simulating a Pt cluster at such low pressure?

Best regards,

Bahareh

Dear Bahareh Bamdad,

Can you describe the collapse of the Pt cluster in more detail? I'm afraid you may have forgotten to attach the images, or else it's ResearchGate or my browser acting strangely and not showing them.

I'm also interested in your model of Pt-C and C-C interactions. Is the carbon in diamond phase, or maybe graphene, graphite, or amorphous? If diamond, which surface orientation is exposed to the Pt cluster? Furthermore, by application of pressure, I assume you mean using a barostat (fix npt or similar) over the periodic boundaries of the carbon slab - is this correct? What temperature do you have? Did you also try the simulation at fixed volume (without barostat)?

Best regards,

Jyri

Dear Jyri Kimari,

Thank you very much for your response. The attached images show the results at both low and ambient pressure. As you can see in the picture, while the simulation works fine at ambient pressure, it collapses at low pressure. I need to simulate the interaction between a Pt cluster and amorphous carbon. I have attached the input file in case you need more details.

Best regards,

Bahareh

Dear Bahareh Bamdad,

The first thing to stand out in your input script is the freezing of C atoms. This will cause at least two features behave in an unexpected way:

You need to have carbon atoms moving freely to be able to calculate the total pressure of the system and to impose pressure control. The AIREBO and Lennard-Jones parameters you set for C atoms are not currently being used at all.

(When C atoms will be unfrozen, you should assign them also initial velocities by velocity create)

The second crucial thing is to set the correct npt parameters for your system, as it represents a 2-dimensional surface: namely, you should not have the "iso" keyword in npt, which means applying isometric pressure control, i.e., the same box size increase/decrease factor in all three dimensions. In the y-direction (if I'm interpreting correctly your images) you have an open surface, meaning usually permanently negative pressure, which would cause the box to shrink in that direction - you don't want this, because you want to have vacuum facing the surface. Use separate x and z keywords to apply independent x and z direction pressure control (fix ... npt temp ${temp} ${temp} 0.5 x ${press} ${press} 10.0 z ${press} ${press} 10.0)

I'm suspecting that the "collapse" you observed was due to the y-direction of the box shrinking due to the isometric pressure control, and the Pt cluster being pushed into the frozen C. Because only the Pt atoms affect the pressure, and they form a cluster with open surfaces in all direction, the pressure will be negative and the box will tend to shrink. You should never apply a barostat in a direction where open surfaces are exposed, it will act inappropriately.

If you want to use a direct barostat for studying clusters on surfaces at different pressures, the only available pressure you have is the 2-dimensional pressure of the substrate, and the only way to control it is to have the substrate unfrozen.

(If you want to, instead, study the effect of the pressure of some gas above the surface (such as Pt vapor?), you can use frozen substrate, but it will be much more complicated. You will have to include those free gas phase atoms in the model, and set their number with respect to the box size to produce the pressure you want (given the temperature)... But you may end up needing very, very large boxes with very few atoms, giving almost useless statistics for any runs that are of reasonable length...)

Best regards,

Jyri