You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Partial charges: read every scheme ORCA, Gaussian and NWChem provide, and show them like Mulliken #272
Requested for research use (partial charges). Today ECCE reads Mulliken (NWChem, ORCA, Gaussian, MOPAC, ECCE-QM), ESP fits (NWChem esp, Gaussian MK/CHELPG) and ORCA CHELPG when requested. Everything else the codes provide is ignored.
Not read yet
Scheme
ORCA
Gaussian 16
NWChem
Löwdin
printed by default
—
printed on request
Mayer (atomic charge, bond orders)
printed by default
"Do Mayer BO" offered, output not read
—
Hirshfeld
! Hirshfeld
Pop=Hirshfeld
—
CM5
—
with Pop=Hirshfeld
—
MBIS
! MBIS (ORCA 5+)
—
—
NPA (NBO)
with NBO
Pop=NBO6 already offered, NPA not read
via NBO's file
APT
—
printed with Freq
—
Order (cheapest first):
Read what is printed already: ORCA Löwdin and Mayer, Gaussian NPA (already requestable) and APT.
Offer and read Hirshfeld (ORCA, Gaussian), CM5 (Gaussian), MBIS (ORCA) in the Calculation Editor.
NWChem Löwdin; NBO for ORCA and NWChem if the NBO program is available.
Display: the general PartialCharges panel (table, colour scale, atoms coloured by charge; used for ESPCHARGE) can show each new property by declaring it in data/client/config/PropertyPanelDescriptor.xml (group Charges, class PartialCharges), as long as the parse scripts store the same per-atom table. On top of that:
one "Atomic charges" entry with a scheme choice, instead of one menu entry per scheme;
a comparison table (atoms × schemes), optionally a plot of charge per atom for each scheme;
the same, optionally fixed, colour scale across schemes so colours are comparable.
Verification: each scheme checked against a real output of that code (water, a small ion pair, a transition-metal complex), with sums matching the total charge, as in the teaching set.
d0e1888, 8bcfba9: ORCA Loewdin, Hirshfeld, MBIS, CHELPG, NPA (! NBO) and Gaussian 16 APT, ESP fit, Hirshfeld, CM5, NPA are read, each a property under Charges, shown with the Mulliken panel. NWChem's ecce_print trace carries only Mulliken charges. Values match the outputs and sum to the total charge. Open: ORCA CHELPG appears twice (old Partial Charges and new ESP-Fitted Charges); Gaussian prints the Hirshfeld/CM5 table only with Pop=Hirshfeld or Pop=CM5 on its own.
91ea799: ORCA CHELPG appears once, as ESP-Fitted Charges (orca.espcharge wrote both ESPCHARGE and ESPFIT); ORCA and G16 ESP surfaces now use the fitted charges instead of Mulliken. G16 Hirshfeld/CM5: g16 prints the table when exactly one of Hirshfeld or CM5 is named, beside any other Pop keywords (Pop=(Full,CM5,MK) prints it); naming both, e.g. Pop=(Full,Hirshfeld,CM5), prints nothing. The parser already reads the table wherever it appears; a real g16 output with Pop=(Full,CM5,MK) is now a parser test. ECCE's dialog does not request either keyword yet.
Gaussian AIM (Bader) charges are not read (checked 2026-10-12 on Linux). G16 C.01, CO, rPBE1PBE/6-31G* … AIM=(CP) Pop=(NBO6,savenbos,full) (Theory Details: "Do AIM CP" and NBO ticked). The Viewer's Partial Charges show Mulliken and NPA only. The output has the AIM charges in
III. PROPERTIES OF ATTRACTORS
Attr. Number of electrons Charge
total spin
1 5.109744 0.000000 0.890256
2 8.891752 0.000000 -0.891752
(attractor n = atom n here; the attractor table "I. ATTRACTORS" gives the nucleus each attractor belongs to, which should be used for the mapping). To do: read this table into an AIM charge set shown with the others.
Also missing on the input side: Gaussian's Theory Details can request only NBO (and AIM); Hirshfeld/CM5 (Pop=Hirshfeld), CHelpG (Pop=CHelpG) and Merz–Kollman (Pop=MK) cannot be asked for, though the parser reads them. Proposed: tick boxes for these, combined into one Pop=(…) keyword with NBO.
In 9.0.0: charges can be requested and are read for ORCA (CHELPG, Hirshfeld, MBIS, NBO), Gaussian 16 (Hirshfeld/CM5, MK or CHelpG, NBO 6/7, AIM), ECCE-QM (Löwdin, CHELPG, Merz–Kollman; compared with ORCA and Gaussian), NWChem and MOPAC (ESP-fitted). Left: Hirshfeld and CM5 for ECCE-QM.
Activity
Partial charges: read Loewdin, Hirshfeld, MBIS, CM5, NPA, APT and ESP…
Parser tests: ORCA NPA charges from a real NBO 7 run (#272)
ohlincha commentedon Oct 11, 2026
d0e1888, 8bcfba9: ORCA Loewdin, Hirshfeld, MBIS, CHELPG, NPA (! NBO) and Gaussian 16 APT, ESP fit, Hirshfeld, CM5, NPA are read, each a property under Charges, shown with the Mulliken panel. NWChem's ecce_print trace carries only Mulliken charges. Values match the outputs and sum to the total charge. Open: ORCA CHELPG appears twice (old Partial Charges and new ESP-Fitted Charges); Gaussian prints the Hirshfeld/CM5 table only with Pop=Hirshfeld or Pop=CM5 on its own.
Charges: ORCA CHELPG listed once, ESP surface reads ESPFIT, G16 Hirsh…
ohlincha commentedon Oct 11, 2026
91ea799: ORCA CHELPG appears once, as ESP-Fitted Charges (orca.espcharge wrote both ESPCHARGE and ESPFIT); ORCA and G16 ESP surfaces now use the fitted charges instead of Mulliken. G16 Hirshfeld/CM5: g16 prints the table when exactly one of Hirshfeld or CM5 is named, beside any other Pop keywords (Pop=(Full,CM5,MK) prints it); naming both, e.g. Pop=(Full,Hirshfeld,CM5), prints nothing. The parser already reads the table wherever it appears; a real g16 output with Pop=(Full,CM5,MK) is now a parser test. ECCE's dialog does not request either keyword yet.
ohlincha commentedon Oct 12, 2026
Gaussian AIM (Bader) charges are not read (checked 2026-10-12 on Linux). G16 C.01, CO,
rPBE1PBE/6-31G* … AIM=(CP) Pop=(NBO6,savenbos,full)(Theory Details: "Do AIM CP" and NBO ticked). The Viewer's Partial Charges show Mulliken and NPA only. The output has the AIM charges in(attractor n = atom n here; the attractor table "I. ATTRACTORS" gives the nucleus each attractor belongs to, which should be used for the mapping). To do: read this table into an AIM charge set shown with the others.
Also missing on the input side: Gaussian's Theory Details can request only NBO (and AIM); Hirshfeld/CM5 (
Pop=Hirshfeld), CHelpG (Pop=CHelpG) and Merz–Kollman (Pop=MK) cannot be asked for, though the parser reads them. Proposed: tick boxes for these, combined into onePop=(…)keyword with NBO.Gaussian 16: request Hirshfeld/CM5, Merz-Kollman and CHelpG charges (#…
Gaussian 16: read AIM (Bader) charges (#272)
ECCE-QM: Loewdin, CHELPG and Merz-Kollman charges (#272)
ORCA: request Hirshfeld, MBIS and NBO charges in Runtype Details (#272)
MOPAC: ESP-fitted charges (ESP keyword, ESPFIT parse, Details tick bo…
NWChem: ESP-fitted charges tick box (task esp after SCF/DFT, ESPCHARG…
NWChem ESP charges: publish nothing from a .q file cut short
ohlincha commentedon Oct 12, 2026
In 9.0.0: charges can be requested and are read for ORCA (CHELPG, Hirshfeld, MBIS, NBO), Gaussian 16 (Hirshfeld/CM5, MK or CHelpG, NBO 6/7, AIM), ECCE-QM (Löwdin, CHELPG, Merz–Kollman; compared with ORCA and Gaussian), NWChem and MOPAC (ESP-fitted). Left: Hirshfeld and CM5 for ECCE-QM.