fix charge/regulation command


fix ID group-ID charge/regulation cation_type anion_type keyword value(s)
  • ID, group-ID are documented in fix command

  • charge/regulation = style name of this fix command

  • cation_type = atom type of free cations

  • anion_type = atom type of free anions

  • zero or more keyword/value pairs may be appended

    keyword = pH, pKa, pKb, pIp, pIm, pKs, acid_type, base_type, lunit_nm, temp, tempfixid, nevery, nmc, xrd, seed, tag, group, onlysalt, pmcmoves
    pH value = pH of the solution
    pKa value = acid dissociation constant
    pKb value = base dissociation constant
    pIp value = chemical potential of free cations
    pIm value = chemical potential of free anions
    pKs value = solution self-dissociation constant
    acid_type = atom type of acid groups
    base_type  = atom type of base groups
    lunit_nm value = unit length used by LAMMPS (# in the units of nanometers)
    temp value = temperature
    tempfixid value = fix ID of temperature thermostat
    nevery value = invoke this fix every nevery steps
    nmc value = number of charge regulation MC moves to attempt every nevery steps
    xrd value = cutoff distance for acid/base reaction
    seed value = random # seed (positive integer)
    tag value = yes or no (yes: The code assign unique tags to inserted ions; no: The tag of all inserted ions is "0")
    group value = group-ID, inserted ions are assigned to group group-ID. Can be used multiple times to assign inserted ions to multiple groups.
    onlysalt values = flag charge_cation charge_anion.
       flag = yes or no (yes: the fix performs only ion insertion/deletion, no: perform acid/base dissociation and ion insertion/deletion)
       charge_cation, charge_anion = value of cation/anion charge, must be an integer (only specify if flag = yes)
    pmcmoves values = pmcA pmcB pmcI -  MC move fractions for acid ionization (pmcA), base ionization (pmcB) and free ion exchange (pmcI)


fix chareg all charge/regulation 1 2 acid_type 3 base_type 4 pKa 5 pKb 7 lb 1.0 nevery 200 nexchange 200 seed 123 tempfixid fT

fix chareg all charge/regulation 1 2 pIp 3 pIm 3 onlysalt yes 2 -1 seed 123 tag yes temp 1.0


This fix performs Monte Carlo (MC) sampling of charge regulation and exchange of ions with a reservoir as discussed in (Curk1) and (Curk2). The implemented method is largely analogous to the grand-reaction ensemble method in (Landsgesell). The implementation is parallelized, compatible with existing LAMMPS functionalities, and applicable to any system utilizing discrete, ionizable groups or surface sites.

Specifically, the fix implements the following three types of MC moves, which discretely change the charge state of individual particles and insert ions into the systems: \(\mathrm{A} \rightleftharpoons \mathrm{A}^-+\mathrm{X}^+\), \(\mathrm{B} \rightleftharpoons \mathrm{B}^++\mathrm{X}^-\), and \(\emptyset \rightleftharpoons Z^-\mathrm{X}^{Z^+}+Z^+\mathrm{X}^{-Z^-}\). In the former two types of reactions, Monte Carlo moves alter the charge value of specific atoms (\(\mathrm{A}\), \(\mathrm{B}\)) and simultaneously insert a counterion to preserve the charge neutrality of the system, modeling the dissociation/association process. The last type of reaction performs grand canonical MC exchange of ion pairs with a (fictitious) reservoir.

In our implementation “acid” refers to particles that can attain charge \(q=\{0,-1\}\) and “base” to particles with \(q=\{0,1\}\), whereas the MC exchange of free ions allows any integer charge values of \({Z^+}\) and \({Z^-}\).

Here we provide several practical examples for modeling charge regulation effects in solvated systems. An acid ionization reaction (\(\mathrm{A} \rightleftharpoons \mathrm{A}^-+\mathrm{H}^+\)) can be defined via a single line in the input file

fix acid_reaction all charge/regulation 2 3 acid_type 1 pH 7.0 pKa 5.0 pIp 7.0 pIm 7.0

where the fix attempts to charge \(\mathrm{A}\) (discharge \(\mathrm{A}^-\)) to \(\mathrm{A}^-\) (\(\mathrm{A}\)) and insert (delete) a proton (atom type 2). Besides, the fix implements self-ionization reaction of water \(\emptyset \rightleftharpoons \mathrm{H}^++\mathrm{OH}^-\). However, this approach is highly inefficient at \(\mathrm{pH} \approx 7\) when the concentration of both protons and hydroxyl ions is low, resulting in a relatively low acceptance rate of MC moves.

A more efficient way is to allow salt ions to participate in ionization reactions, which can be easily achieved via

fix acid_reaction all charge/regulation 4 5 acid_type 1 pH 7.0 pKa 5.0 pIp 2.0 pIm 2.0

where particles of atom type 4 and 5 are the salt cations and anions, both at chemical potential pI=2.0, see (Curk1) and (Landsgesell) for more details.

Similarly, we could have simultaneously added a base ionization reaction (\(\mathrm{B} \rightleftharpoons \mathrm{B}^++\mathrm{OH}^-\))

fix base_reaction all charge/regulation 2 3 base_type 6 pH 7.0 pKb 6.0 pIp 7.0 pIm 7.0

where the fix will attempt to charge \(\mathrm{B}\) (discharge \(\mathrm{B}^+\)) to \(\mathrm{B}^+\) (\(\mathrm{B}\)) and insert (delete) a hydroxyl ion \(\mathrm{OH}^-\) of atom type 3. If neither the acid or the base type is specified, for example,

fix salt_reaction all charge/regulation 4 5 pIp 2.0 pIm 2.0

the fix simply inserts or deletes an ion pair of a free cation (atom type 4) and a free anion (atom type 5) as done in a conventional grand-canonical MC simulation.

The fix is compatible with LAMMPS sub-packages such as molecule or rigid. That said, the acid and base particles can be part of larger molecules or rigid bodies. Free ions that are inserted to or deleted from the system must be defined as single particles (no bonded interactions allowed) and cannot be part of larger molecules or rigid bodies. If molecule package is used, all inserted ions have a molecule ID equal to zero.

Note that LAMMPS implicitly assumes a constant number of particles (degrees of freedom). Since using this fix alters the total number of particles during the simulation, any thermostat used by LAMMPS, such as NVT or Langevin, must use a dynamic calculation of system temperature. This can be achieved by specifying a dynamic temperature compute (e.g. dtemp) and using it with the desired thermostat, e.g. a Langevin thermostat:

compute dtemp all temp
compute_modify dtemp dynamic yes
fix fT all langevin 1.0 1.0 1.0 123
fix_modify fT temp dtemp

The chemical potential units (e.g. pH) are in the standard log10 representation assuming reference concentration \(\rho_0 = \mathrm{mol}/\mathrm{l}\). Therefore, to perform the internal unit conversion, the length (in nanometers) of the LAMMPS unit length must be specified via lunit_nm (default is set to the Bjerrum length in water at room temperature lunit_nm = 0.71nm). For example, in the dilute ideal solution limit, the concentration of free ions will be \(c_\mathrm{I} = 10^{-\mathrm{pIp}}\mathrm{mol}/\mathrm{l}\).

The temperature used in MC acceptance probability is set by temp. This temperature should be the same as the temperature set by the molecular dynamics thermostat. For most purposes, it is probably best to use tempfixid keyword which dynamically sets the temperature equal to the chosen MD thermostat temperature, in the example above we assumed the thermostat fix-ID is fT. The inserted particles attain a random velocity corresponding to the specified temperature. Using tempfixid overrides any fixed temperature set by temp.

The xrd keyword can be used to restrict the inserted/deleted counterions to a specific radial distance from an acid or base particle that is currently participating in a reaction. This can be used to simulate more realist reaction dynamics. If xrd = 0 or xrd > L / 2, where L is the smallest box dimension, the radial restriction is automatically turned off and free ion can be inserted or deleted anywhere in the simulation box.

If the tag yes is used, every inserted atom gets a unique tag ID, otherwise, the tag of every inserted atom is set to 0. tag yes might cause an integer overflow in very long simulations since the tags are unique to every particle and thus increase with every successful particle insertion.

The pmcmoves keyword sets the relative probability of attempting the three types of MC moves (reactions): acid charging, base charging, and ion pair exchange. The fix only attempts to perform particle charging MC moves if acid_type or base_type is defined. Otherwise fix only performs free ion insertion/deletion. For example, if acid_type is not defined, pmcA is automatically set to 0. The vector pmcmoves is automatically normalized, for example, if set to pmcmoves 0 0.33 0.33, the vector would be normalized to [0,0.5,0.5].

The only_salt option can be used to perform multivalent grand-canonical ion-exchange moves. If only_salt yes is used, no charge exchange is performed, only ion insertion/deletion (pmcmoves is set to [0,0,1]), but ions can be multivalent. In the example above, an MC move would consist of three ion insertion/deletion to preserve the charge neutrality of the system.

The group keyword can be used to add inserted particles to a specific group-ID. All inserted particles are automatically added to group all.


This fix computes a global vector of length 8, which can be accessed by various output commands. The vector values are the following global quantities:

  • 1 = cumulative MC attempts

  • 2 = cumulative MC successes

  • 3 = current # of neutral acid atoms

  • 4 = current # of -1 charged acid atoms

  • 5 = current # of neutral base atoms

  • 6 = current # of +1 charged base atoms

  • 7 = current # of free cations

  • 8 = current # of free anions


This fix is part of the MC package. It is only enabled if LAMMPS was built with that package. See the Build package doc page for more info.

The atom_style, used must contain the charge property, for example, the style could be charge or full. Only usable for 3D simulations. Atoms specified as free ions cannot be part of rigid bodies or molecules and cannot have bonding interactions. The scheme is limited to integer charges, any atoms with non-integer charges will not be considered by the fix.

All interaction potentials used must be continuous, otherwise the MD integration and the particle exchange MC moves do not correspond to the same equilibrium ensemble. For example, if an lj/cut pair style is used, the LJ potential must be shifted so that it vanishes at the cutoff. This can be easily achieved using the pair_modify command, i.e., by using: pair_modify shift yes.


Region restrictions are not yet implemented.


pH = 7.0; pKa = 100.0; pKb = 100.0; pIp = 5.0; pIm = 5.0; pKs = 14.0; acid_type = -1; base_type = -1; lunit_nm = 0.71; temp = 1.0; nevery = 100; nmc = 100; xrd = 0; seed = 0; tag = no; onlysalt = no, pmcmoves = [1/3, 1/3, 1/3], group-ID = all

(Curk1) T. Curk, J. Yuan, and E. Luijten, “Coarse-grained simulation of charge regulation using LAMMPS”, preprint (2021).

(Curk2) T. Curk and E. Luijten, “Charge-regulation effects in nanoparticle self-assembly”, PRL (2021)

(Landsgesell) J. Landsgesell, P. Hebbeker, O. Rud, R. Lunkad, P. Kosovan, and C. Holm, “Grand-reaction method for simulations of ionization equilibria coupled to ion partitioning”, Macromolecules 53, 3007-3020 (2020).