Hi,
I’m aiming to build a coarse grained model of core-shell polymer nanoparticles.
I want to do this using Moltemplate. Which would be the best example to modify? Where best to start?
I’m a complete beginner to LAMMPS!
As a “complete beginner” you should not start with an “advanced” model like the core-shell polarization model. You should first build your general LAMMPS and MD simulation skills with much simpler models, so you can better assess whether any issues you encounter (and there will be issues with advanced models) are due to lack of understanding or bad choices or settings, parameters or geometry. Thus I strongly recommend you start your LAMMPS journey with the tutorials at https://lammpstutorials.github.io and do not proceed to core-shell models until you are comfortable with creating and running simulations with LAMMPS.
The next steps would then be to study the information available in the LAMMPS manual and the corresponding examples in the LAMMPS source distribution. Again, I strongly advise against jumping ahead; it will just make everything more painful and likely filed with more errors and frustrations.
Thanks Axel.
I’ll work on that and move to core-shell models ASAP!
I’ll get back if I have further questions.
Hi Axel.
For clarification, though I’m a complete beginner to LAMMPS, I am not to molecular simulation.
I have experience in GROMACS and NAMD for protein simulation.
It’s my first-time modelling materials, but while I understand what you are saying with regards to rushing things. I am working on a time-limited project, so do want to start modelling core-shell polymer nanoparticles ASAP.
I have completed the main tutorials, and feel fairly confident with them. Are there any tutorials precisely targeted to polymers, so I can build with them before targeting the core-shell particles?
That doesn’t make a difference in this case. If you had that much more experience in simulations that it would make a difference, you would already know that all available tutorials that we know about are posted on the LAMMPS homepage and any additional instructional material in the Howto section of the LAMMPS manual. You would also know and understand the “mechanics” of who writes tutorials and for what purpose (usually by people that regularly have to educate people about the basics of using LAMMPS and thus about mainstream topics) therefore that you are not likely going to find a lot of educational material for an advanced and uncommon topic, especially if applied to a specific context. You would have already known that the odds are strongly stacked against you.
If you are under time constraints, then it is unlikely that you will be able to sufficiently educate yourself on your own in a timely fashion. When you are training yourself, things always take much longer than anticipated (I can personally vouch for that, I just spend over two months educating myself about the basics of computer graphics for some project that I expected to complete in two to three weeks.) Thus, in my opinion, the only viable option would be to recruit a sufficiently experienced collaborator.
That’s fine, thanks Axel.
Knowing my situation though, would you have a collaborator you would recommend?
I specialize in scientific software development and maintenance and high-performance computing. I have no experience in your area of research, so any recommendation of mine would be tainted by me having to guess somebodies capabilities. It would be far more meaningful to survey the published literature related to your project and identify competent research groups with experience in your field that have also used LAMMPS and then reach out to them. In fact, I would assume that you have already done that kind of survey, because that is usually the first thing to do if you enter a new area of research, so you would just need to check on whether they used LAMMPS in their work and then focus on those research groups.
Random idea: maybe you can find something on how people a configuration for core shell polymer nanoparticles from a publication where they are simulating it (I mean, at least someone must give some details on the methodology section or in further supplementary material). 
PS: maybe in the process the create_atoms command can be useful if summoned for a spherical region (which you can define using the region command)
1 Like
I think @hfarouq23 and @Dr_Haider_Farouq want to model “core-shell nanoparticles”, as defined in this review, and not using the polarizable core-shell model for inorganic solid (see the extensive work of Richard Catlow and coworkers, e.g. in this collection).
Semantics clarified, I can only suggest to follow this example on how to build a short polymer chain and put in a water box using the OPLS-AA force field.
To create a core-shell nanoparticle, you can follow the procedure described e.g. in this article of mine if the polymer is chemically bound to the surface. Use the cone algorithm to find suitable sites on the surface of a nanoparticle and mark them for substitution. Then, replace the atom (or group of atoms) with a polymer chain. If the polymer is just adsorbed on the surface, use the cone algorithm to identify the vector along which to align the polymer chain, assuming they are radially oriented around the nanoparticle.
Whatever the case, this is advanced stuff that will likely require writing your own tools to handle the creation of the starting structure. For the force field, I would use an existing library that is suitable for most organic compounds –at least to begin with.
1 Like
Thanks, will explore my options and see how I get on!
Thanks, will take a look and see how I get on!
Could you confirm what you mean exactly by “modelling core-shell polymer nanoparticles”? You said that you would have looked at core-shell models, but I still am not sure if you really want to use a core-shell model, which was developed to account for polarisation in inorganic crystals, or to model a core-shell nanoparticle, as I pointed out in my previous comment. That’s the kind of feedback one would like to receive; not individual replies that do not advance the discussion.
1 Like
Hi, just as you rightly point out, the core-shell model isn’t applicable here, as you mention it is for inorganic crystals.
In the project, an experimental chemist is developing core-shell polymer nanoparticles and I want to build a model to guide the project. I am looking to do this by making a coarse-grained model, as a fine-grained model will be too detailed and irrelevant. I am seeking to use dissipative particle dynamics to build the model and at the moment trying to see how to best do it with LAMMPS.
I said I’ll take a look at your previous comment, and have done so now. It’s an all-atom example of the polymer, as it is in your article.
Do you have better advice for building a coarse-grained model?
Your problem can be broken down into two major tasks:
- Create a suitable coarse-grained parametrisation for the materials of interest.
DPD is a sensible choice for polymers. As far as I am concerned, your nanoparticle core could be just a big finite-size sphere with a Lennard-Jones potential to make it sticky enough, and you are good to go. It is your job to find suitable parameters in the literature, or to work them out yourself from potential energy curves obtained with a finer-grained model. - Create the initial geometry for your sample.
If the nanoparticle’s core has certain features that you want to reproduce, e.g. elongated shape, inhomogeneous distribution of binding sites, then the coarse-grained model should use more sites to approximate the nanoparticle shape –hence the method described to attach molecules radially to a surface applies, even if it was used with an atomistic model.
I pointed out to the Moltemplate tutorial as it shows how to hierarchically build a polymer from monomer fragments. The same concept (and syntax) applies to coarse-grained models, as you can see from the many examples covering different materials and models.
1 Like
Thanks for your input Otello.
I already have suitable parameters which I obtained during my literature search.
I will make a start and consult my supervisors on the best way forward.