Skip to content

4.5 Bader populations and charge conventions

These are unexecuted teaching inputs and starting settings to test. Original figures are schematics, not computed results. Use licensed VASP and PAW data, replace every placeholder, record the executable version and validate convergence.

4.5.1 Model, units and provenance

Use eV for energy, Å for length and eV/Å for force; 1 kbar = 0.1 GPa. State normalization per atom, molecule, primitive cell or simulation cell. Record PAW identifiers, release, ZVAL, ENMAX and permitted hashes; never redistribute POTCAR. SCF convergence addresses the chosen electronic problem; convergence of the target property requires separate tests.

 Bader populations and charge conventions — original conceptual schematic

Original schematic. Curves explain concepts; blank data areas await verified learner results. No calculation is claimed.

4.5.2 Worked case: procedure, interpretation and checks

Purpose and prerequisites

A Bader analysis partitions space into density basins separated by zero-flux boundaries of the density gradient. Integrating an electron density inside one basin gives an electron population. This is a defensible, explicitly defined way to discuss redistribution, but it is not a measurement of an atom's unique “true charge.” Different partition schemes answer different questions, and Bader populations need not be integers even for a material conventionally described as ionic.

Use a relaxed small ionic crystal, such as NaCl, and a converged static PAW density. You need both VASP and the separate Henkelman-group Bader program; VASP does not run that executable merely because LAECHG is enabled. Use the official project distribution or your institution's verified installation, record its version, and read its help. Do not download a random precompiled executable from a tutorial mirror. Official Bader project

Workflow and bookkeeping

For a static run, retain the converged cutoff and regular mesh, set NSW=0, LCHARG=.TRUE., and LAECHG=.TRUE.. AECCAR0 contains the reconstructed core contribution; AECCAR2 contains the self-consistent all-electron valence contribution. AECCAR1 has a different proto-atomic meaning and must not be substituted for AECCAR2. Because these files are produced at different stages, use a static final-geometry calculation, not an unexamined set from a moving-ion run. LAECHG

With the official helper script already installed, the usual two commands are:

The key distinction is between the density used to find boundaries and the density being integrated. Here the reference includes core information for robust basin finding, while the integrated CHGCAR supplies a valence electron population. Inspect ACF.dat and the program's electron-count and vacuum summaries. Check the atom indices against POSCAR; do not sort one list independently of the other. Official command and VASP notes

For this valence-integration workflow define \(q_i/e=\mathrm{ZVAL}_i-N_i\), using the actual ZVAL of the selected PAW dataset. Positive q means a deficit of valence electrons. A semicore dataset changes ZVAL and must change the reference in this formula. If instead integrating a reconstructed total electron density, use the nuclear charge as the corresponding neutral reference and verify that reconstruction/integration is adequately converged. Mixing an all-electron population with ZVAL creates a large but meaningless charge. State the convention beside every reported table.

Convergence, limitations, and exercise

Converge the fine FFT grid specifically for the partition. At fixed physical settings, increase NGXF/NGYF/NGZF together and monitor individual populations and the global electron-count residual. Check vacuum assignments for slabs and molecules rather than assuming every grid point was assigned to an atom. Small changes between chemically similar atoms may be below the grid uncertainty. An energetically converged VASP calculation can still yield underresolved basin boundaries.

An interface fragment can gain population even when its atoms show mixed individual changes. Sum a defined atom group, compare matching geometries and PAW conventions, and distinguish a difference of populations from a difference-density picture. Neither quantity alone proves a reaction mechanism. For electride-like systems or non-nuclear attractors, verify whether maxima and basins are associated with nuclei as expected; nearest-atom reporting can hide unusual topology. Seek specialist analysis rather than forcing every basin into a simple oxidation-state story.

Exercise: calculate three increasingly fine grids and report electron population, signed net charge, and residual for each. Verify the global balance including any vacuum population. Explain why the formal Na(+I)/Cl(−I) labels are a different chemical convention and need not equal your Bader numbers. Do not insert illustrative numerical charges as if they had been calculated.

4.5.3 Unexecuted inputs and analysis scaffolds

These are unexecuted teaching inputs and starting settings to test. Original figures are schematics, not computed results. Use licensed VASP and PAW data, replace every placeholder, record the executable version and validate convergence.

4.5.3.1 Input block 1

chgsum.pl AECCAR0 AECCAR2
bader CHGCAR -ref CHGCAR_sum

4.5.5 Technical sources