Szczepanik Research Group
Department of Theoretical Chemistry
Faculty of Chemistry, Jagiellonian University
Gronostajowa 2, 30-387 Krakow, Poland
Tel: (+48) 12 686 23 90
E-mail: dariusz.szczepanik@uj.edu.pl
HOME PROJECTS PUBLICATIONS runEDDB X
About EDDB Download Tutorials Manual
The core of runEDDB: it splits the electron density into the part localized on atoms and in two-centre bonds (EDLB) and the part delocalized over the bond system (EDDB). You choose where to look (the whole molecule, a fragment, a pathway or rings), which electrons (all, or the π component), and in which orbital basis the populations are counted. This page shows each choice with an example and explains the tables the program can print.
The scope decides which bonds enter the analysis. The wavefunction is always the whole molecule; only
the populations, tables and cubes are restricted. The scope letter appears in every result
(EDDB_G, EDDB_F, ...).
| Option | Scope | Example |
|---|---|---|
-g (default) | G: the whole molecule, every bonded atom pair | runEDDB -i @naphthalene |
-h | H: the whole molecule without the bonds to hydrogen (cleaner π selection, less C–H hyperconjugation; faster) | runEDDB -i @naphthalene -h |
-f LIST | F: the bonds within a set of atoms | runEDDB -i @naphthalene -f 1:4,9,10 -a |
-p CHAIN | P: only the listed bonds, in order; repeat the first atom to close a cycle | runEDDB -i @naphthalene -p 1-2-3-4-9-10-1 -a |
-r SIZES | R: the bonds of every ring of the given sizes, found automatically | runEDDB -i @coronene -r 6 -a |
1:6 is a range, 1,3,5 a list, combined freely (1:4,9,10).
Atom numbers are those of the input file.-r 6, -r 5,6 or -r 5,6:8 look for rings in the graph of covalent bonds
(Wiberg bond order above 0.30, --ring-thresh). The rings found are written to
<stem>.rings, one -p string per line, so each can be analysed on its own:$ runEDDB -i @naphthalene -r 6 -a $ cat naphthalene.rings 1-2-3-4-9-10-1 4-5-6-7-8-9-4 $ grep -v '^#' naphthalene.rings | while read -r r; do runEDDB -i @naphthalene -p "$r" -a -q; done
Under a restricted scope every density (EDDB, EDLB and BDF) is built only on the selected atoms, so the per-atom tables, the totals and the cube integrals agree with each other.
A closed pathway (-p with the first atom repeated) also gives the cyclic delocalization
index (CDI): the smallest EDDB population on any atom of the cycle, i.e. the largest number of
electrons per atom that can delocalize uniformly all around it. A CDI near 0.9 per atom (π) marks a
strongly aromatic ring, near 0 a non- or antiaromatic one: benzene 0.89
(runEDDB -i @benzene -p 1-2-3-4-5-6-1 -a), tub-shaped cyclooctatetraene 0.03
(runEDDB -i @rings/cot -p 1-2-3-4-5-6-7-8-1 -a). With a π dissection two fractions follow:
$ runEDDB -i @naphthalene -p 1-2-3-4-9-10-1 -a
Total ED_P population: 5.98843 per atom: 0.99807
Total EDLB_P population: 2.53388 per atom: 0.42231
Total EDDB_P population: 3.45455 per atom: 0.57576
Cyclic Delocal. Index (CDI): 0.40233
Delocalized fraction (%del): 57.69 %
Cyclic coherence (%cyc): 40.31 %| Line | Meaning |
|---|---|
CDI | the smallest per-atom EDDB_P along the cycle |
%del | EDDB_P / ED_P: the delocalized fraction of the π electrons of the pathway |
%cyc | N × CDI / ED_P: the part that is delocalized coherently around the whole cycle, limited by its weakest atom |
The gap between %del and %cyc is delocalization that does not close the cycle ("islands"). In
naphthalene the fusion atoms 4 and 9 share their π electrons with both rings, so each ring has
%del = 58 % but %cyc = 40 %. The two fractions are printed only under -a or
-s, where ED_P is a π population.
The delocalized density is diagonalised into natural orbitals for bond delocalization (NOBDs); each has an occupation (its share of EDDB) and an s/p/d/f character. Selecting some of them restricts every result to those orbitals.
-a selects the π-type NOBDs automatically: an NOBD is π when its s content is below
10 % and at least half of its p content points along the local normal of the plane at each atom.-s LIST selects NOBDs by number, e.g. -s 1:3. --table nobd shows the
numbers (a star marks the selected ones):$ runEDDB -i @benzene --table nobd Index Sel Occupancy s (l=0) %s p (l=1) %p d (l=2) %d f (l=3) %f 1 * 1.78309 0.00000 0 1.78309 100 0.00000 0 0.00000 0 2 * 1.78303 0.00000 0 1.78303 100 0.00000 0 0.00000 0 3 * 1.78298 0.00000 0 1.78298 100 0.00000 0 0.00000 0 4 * 0.05096 0.01287 25 0.03809 75 0.00000 0 0.00000 0 ... $ runEDDB -i @benzene -s 1:3 # the three pi NOBDs = what -a picks here
-a may pick σ-mixed orbitals, or all of them: on the
force-field model @rings/cyclobutadiene it selects all eight non-zero NOBDs. Look at
--table nobd and use -s when in doubt.By default runEDDB prints only the summary. Tables are requested with --table LIST (comma
separated) and are printed in a fixed order before the summary.
| Token | Table |
|---|---|
nobd | NOBDs: occupation, s/p/d/f character, selection star |
npa | per atom: natural charge, population ED and its delocalized part EDDB |
mo | per canonical MO: energy, occupation, ED / EDLB / EDDB (HF/DFT input) |
bo | bond-order orbitals: the Wiberg bond order of every pair resolved into σ, π, δ, φ components |
ao, subshell, shell | populations per NAO (2px), per subshell (2p) or per shell (2 = 2s + 2p) |
no | natural orbitals: occupation and its core / valence / Rydberg composition |
all | every applicable table |
$ runEDDB -i @benzene --table nobd,npa $ runEDDB -i @benzene --table bo $ runEDDB -i mol.fchk --table no -v # + multireference indices in the -v block
--export-fchk bo), the
total Wiberg bond order and the largest components with their σ/π/δ/φ label. The total equals
the full bond order only with -b nao (benzene C–C 1.440 vs 1.418 in the default basis); the
π share is practically the same. On large molecules use it with -f/-p/-r:
the bond-order orbitals are then built only for the selected pairs. -v writes the complete
balance to <stem>.NBA_DIAG.csv.--export-fchk ao 1:1; the shell table shows
the Rydberg population ladder at a glance.-v the multireference indices follow: Yamaguchi y0 biradical index,
Bofill–Pulay and Head-Gordon unpaired-electron indices, and 〈S2〉 spin contamination
for open shells, each with its DOI.-q:
Calculation completed (with 1 warning!); rerun without -q to read them.Populations are counted in natural atomic orbitals (NAOs), which come in three shells: core, valence
and Rydberg. -b picks which of them are analysed:
| Value | NAOs | Use |
|---|---|---|
nvb (default) | valence | standard for aromaticity and bonding studies; fastest |
nmb | core + valence (natural minimal basis) | the legacy default; the only choice for NBO
.49 files |
nao | all (core + valence + Rydberg) | completeness check; nvb and nmb usually agree with it within a few per cent |
Which atom pairs are analysed does not depend on -b: the bond connectivity is always taken
from the full NAO basis, so a series nvb / nmb / nao differs only in the density, never in the set of bonds.
For correlated densities the layers of natural orbitals are also always formed in the full basis.
| Option | Default | Decides |
|---|---|---|
-w X (--wbo-thresh) | 0.001 | which atom pairs enter the analysis. Deliberately loose: it keeps hyperconjugation, weak hydrogen bonds and through-space contacts. Results are comparable only between runs at the same value; keep the default unless you study the threshold itself. |
--ring-thresh X | 0.30 | which bonds count as covalent when -r searches for
rings. Raise it to suppress rings closed through weak contacts, lower it (e.g. 0.15) to catch a ring closed by
a weak bond. Changes nothing else. |
-w 0 keeps every pair (N(N−1)/2): fine for small molecules, expensive for large ones
(benzene 51 → 66 pairs, a 59-atom RNA fragment 260 → 1711); combine it with a scope.
The NAOs are built by runEDDB itself. --nao-algorithm chooses the construction; the default
suits every use, and the others exist for comparisons and special studies.
| Algorithm | What it is |
|---|---|
frozen-nmb (default) | the core and valence NAOs are built on a per-atom, angular-momentum-truncated
density and frozen, the Rydberg space is built around them. Well conditioned at any basis size; atomic populations
agree with rigorous within 0.001 e up to cc-pV5Z and stay symmetric where the standard construction
breaks down (cc-pV6Z). |
rigorous | the standard NBO-style construction, directly comparable with NBO 7 and the published
NPA literature (nbo7legacy is an alias) |
cc-layers | Rydberg space built shell layer by shell layer in Dunning's correlation-consistent
order, with monotonic radial occupations; for truncated virtual spaces in correlated calculations
(--table shell shows the layers) |
ano-layers, iano-layers | the same layers with directions from tabulated atomic
natural orbitals (built-in ano-pV5Z, H–Ar, or --ano-basis FILE); iano-layers builds the minimal
basis as intrinsic atomic orbitals |
Only for native input; .49 files carry NBO's own NAOs. The header of an exported NAO file records
which construction made it. Full option list: runEDDB --help eddb.
In strongly polarized basis sets (cc-pVQZ and beyond) the textbook NAOs are stored as large,
near-cancelling coefficient expansions over high-angular-momentum shells and neighbouring atoms. The orbitals
in space stay localized and the populations stable, but most viewers (Avogadro2, IQmol) render the exported
orbitals as delocalized artefacts, and the charges lose numerical conditioning at cardinal number 5 and
above. The frozen-NMB construction has no such coefficients by construction. For pictures of the rigorous
NAOs prefer orbital cubes (--export-cube ao[LIST]) or --export-lproj.
Within one compilation a run repeated with the same number of threads gives the same digits every time; a
different -n may change the last printed digit of the EDLB / EDDB columns (the order of the sums
changes). For bit-identical numbers state the compilation date (printed in the banner) and -n;
-n 1 is the reference. Later compilations may change numbers slightly; every such change is listed
in the changelog with its scale. Archival compilations are available from the author
on request by e-mail.
| Method | Reference |
|---|---|
| Natural atomic orbitals, natural population analysis | A.E. Reed, R.B. Weinstock, F. Weinhold, J. Chem. Phys. 83 (1985) 735, doi:10.1063/1.449486 |
| Frozen-NMB NAO construction (default) | D.W. Szczepanik, J. Comput. Chem. (2026), in preparation |
| Bond-order projection (BOP) | D.W. Szczepanik, E.J. Zak, K. Dyduch, J. Mrozek, Chem. Phys. Lett. 593 (2014) 154, doi:10.1016/j.cplett.2014.01.006 |
| EDDB / EDLB partition | D.W. Szczepanik, Comput. Theor. Chem. 1080 (2016) 33, doi:10.1016/j.comptc.2016.02.003 |
NOBDs and their dissection (-a, -s) | D.W. Szczepanik et al., Phys. Chem. Chem. Phys. 16 (2014) 20514, doi:10.1039/c4cp02932a |
| EDDB of fragments, pathways and rings; CDI | D.W. Szczepanik et al., Phys. Chem. Chem. Phys. 19 (2017) 28970, doi:10.1039/c7cp06114e |
Orbital-resolved EDDB (--table ao) | D.W. Szczepanik, M. Solà, ChemistryOpen 8 (2019) 219, doi:10.1002/open.201900014 |
| ANO guide of ano-layers | F. Neese, E.F. Valeev, J. Chem. Theory Comput. 7 (2011) 33, doi:10.1021/ct100396y |
| Intrinsic atomic orbitals (iano-layers) | G. Knizia, J. Chem. Theory Comput. 9 (2013) 4834, doi:10.1021/ct400687b |