Inquiry on Using MBX for Water Molecules in Silver Electrode System

[刘赵洋]

Dear developers and everyone,

I hope this email finds you well.

I am working on a simulation with a silver electrode and water molecules, where I aim to apply the MBX polarization force field only to the water molecules, while using Lennard-Jones for the silver electrode. I used the following command in LAMMPS: fix 2 water mbx 1 h2o 2 3 3 2 3 3 json mbx.json, but I encountered the error ERROR: The atom types in fix mbx do not match the atom types in the data file (../fix_mbx.cpp:458). After checking my data file, I confirmed that oxygen is assigned type 2 and hydrogen is assigned type 3, which matches the MBX settings. I would like to know if it is possible to apply MBX only to the water molecules, or if I need to include the silver electrode, and if so, can silver be assigned the dp1 monomer type or is there another way to handle silver in this setup?

Thank you for your time and I look forward to your suggestions.

Best regards,

Sibing Chen

 

[Henry Agnew]

Dear Sibing,
         Based on your question, it sounds like you are using a silver electrode in a simulation with MB-pol water molecules. If this is indeed correct, then yes you will need to perform a hybrid simulation using the dp1 monomer type for the silver electrode.

This setup has some similarities with our MOF+H2O example, where the MOF is treated using dp1 and coul/exclude. https://github.com/paesanilab/MBX/blob/master/plugins/lammps/examples/mof_ff_lammps%2B1_h2o_mbx/lammps.in

Since you are doing an electrode-based system, are you using the electrode with a constant charge or something fancier? MBX calculates all hybrid electrostatics internally using dp1, so it may not be compatible if you are doing something fancier.

Hopefully this answers your question, and let us know if you have any follow-up questions,

Best regards,
 The MBX Team

[刘赵洋]

Dear MBX developers,

I am planning to study how an applied electrode potential affects interfacial water at an Ag–water interface. My current setup is to treat water with MBX/MB-pol and represent the Ag electrode as dp1 in the MBX hybrid scheme, while handling Ag–Ag and Ag–water short-range interactions (e.g., LJ) in LAMMPS. In addition, I would like to use the LAMMPS ELECTRODE package (constant-potential methods such as CONP2 with per-timestep fluctuating charges on electrode atoms) to vary the electrode potential. Could you please advise whether ELECTRODE’s dynamic charge updates are compatible with MBX’s internal hybrid electrostatics for dp1? 

Specifically: (1) does MBX read and use the dp1 atomic charges from LAMMPS every timestep, so that time-dependent electrode charges are correctly seen by MBX; or (2) does MBX assume fixed external charges / compute electrostatics internally in a way that would conflict with constant-potential algorithms? If this combination is feasible, I would greatly appreciate any recommended settings to avoid double counting (e.g., which Coulomb terms should be excluded on the LAMMPS side, or any MBX options/workarounds you recommend). 

Thank you very much for your guidance. 

Best regards, 
Sibing Chen

[Henry Agnew]

Dear Sibing,

        Thank you for the additional information. At the moment, MB-pol is not known to be compatible with the LAMMPS ELECTRODE package, as MB-pol uses variable Partridge and Schwenke charges. Specifically, more coding work is needed in order for the ELECTRODE package and MBX to cross-communicate the electrostatics properly.  While this feature is something we are planning to eventually do in the long-term, it is not currently available.

If you happen to feel inclined to try to implement this feature, we do accept code pull requests at https://github.com/paesanilab/MBX!

Relating to the second part of your question, MBX does read the dp1 atomic charges from LAMMPS every timestep. However, MBX currently needs to calculate all of the electrostatics, so therefore the ELECTRODE package may be missing information from MBX about the MB-pol electrolyte.

Please let us know if you have any additional questions.

Best regards,

The MBX Team

[刘赵洋]

Dear MBX developers, 

  

        Thank you for your clear and patient explanation—this is very helpful for us. We understand the current limitations regarding MB-pol compatibility with the LAMMPS ELECTRODE package.  

Best regards, 
Sibing Chen