Table of Contents
1. intent
It not evident if Avogadro2 trades the abstraction of hydrogens from carbon atoms well enough. Departing from the neutral methane molecule, this operation can yield the methyl radical, or – if an additional positive charge is provided – the methyl cation. The issue was addressed as issue Specify atom radical / charge for atypical valence #1116 (), monitored on structures exported as InChI strings with Avogadro (version 1.96.0, AppImage for Linux running in Linux 12/bookworm [then branch testing]).
The thread is given a new momentum by Geoffrey Hutchison's note on . Not explicitly covered – neither by the initial question, nor by Geoffrey's reply – if the geometry optimization provided by openbabel's default force field MMFF94s equally could be affected by this, if geometry optimizations with this force field use different parametrizations for methyl cation, methyl radical, and methyl carbanion.
Hypothesis: MMFS94s does recognizes the situations as different, hence the optimized geometries will be different. The superposition of the two structures by Kabsch then will quantify this difference by a RMSD.
2. tools used
- Linux Debian 13/trixie (branch testing), Emacs org mode extended by Fabrice Niessen's readtheorg theme
- ChemDraw JS, as provided by the sample page. This aims to exclude potential differences between different releases / operating systems.
- obabel, version 3.1.1 (), as provided by DebiChem / the repositories of Linux Debian 13
- Python (version 3.11.5) as packaged by Debian 13/trixie
- Kabsch test, implemented in Python by Jimmy Charnley et al.: rmsd,
(commit
dcca78e, )
3. analysis
3.1. preparation
One by one, the structures of methyl cation, methyl radical, and methyl anion are sketched in ChemdrawJS. Departing from the neutral methane molecule in expanded structure formula, a hydrogen is removed. The structure is individually decorated by "circle plus", "radical", or "circle minus", accordingly, and exported as .sdf (v2000 format).
The local copies are submitted to openbabel for a geometry optimization by a command in pattern of
obabel methyl_carbanion.sdf --gen3d -O methyl_carbanion_obabel_optimized.sdf obabel methyl_cation_obabel_optimized.sdf -O methyl_cation_obabel_optimized.xyz
to eventually yield .xyz one could submit to a Kabsch test.
3.2. comparisons
After the setup of a virtual environment of Python supplemented by the
ancillary modules (pip3 install -r requirements), the analyses are
performed in pattern of
$ python ./calculate_rmsd.py methyl_radical_obabel_optimized.xyz methyl_carbanion_obabel_optimized.xyz 5.362896329588574e-05
to report the corresponding RMSD.
| methyl cation | methyl radical | methyl carbanion | |
|---|---|---|---|
| methyl cation | - | 4.173E-16 | 5.362E-05 |
| methyl radical | 4.173E-16 | - | 5.362E-05 |
| methyl carbanion | 5.362E-05 | 5.362E-05 | - |
The symmetry of the values in the array is anticipated. The difference in geometry between the optimized geometries of methyl cation and methyl radical is much smaller, then the one between the metyl carbanion and either carbacation, or the radical. With molecules this little, the numbers however are tiny.
Hence, the Kabsch test is replicated with the optional --print flag to
report the newly aligned second structure superimposed to the first one:
$ python ./calculate_rmsd.py methyl_carbanion_obabel_optimized.xyz methyl_cation_obabel_optimized.xyz -p > carbanion_cation.xyz $ python ./calculate_rmsd.py methyl_carbanion_obabel_optimized.xyz methyl_radical_obabel_optimized.xyz -p > carbanion_radical.xyz
As eventually (re)discoverd, Jmol permits the simultaneous display of two models in one frame. On Jmol's console:
load models "methyl_carbanion_opbabel_optimized.xyz" "carbanion_cation.xyz"; select 1.1; color red; select 2.1; color blue; model 0; write "superposition_methyl_carbanion_methyl_cation.png";
to eventually display the two (about the methyl carbanion and the methyl cation) at once. Reference of the special visualization method: https://proteopedia.org/wiki/index.php/Jmol/superposition
Figure 1: Superposition (by Kabsch test) of methyl carbanion (red) and methyl cation (blue); geometries optimized by obabel/MMFF86.
Based on VSEPR theory, the two molecular shapes should be different (accounting for the atoms, trigonal pyramidial similar to ammonia vs. flat). But since both models apear flat ironed, the MMFF94s force field seems not to account for the charge (i.e. in case of the carbanion the free electron pair) submitted in the .sdf to perform the geomertical optimization.
4. conclusion
The hypothesis was openbabel's default force field, MMFF94s, would recognize the methyl cation and methyl carbanion defined in a .sdf file (including coordinates and explicit charges) as different enough to yield a different shape when performing an optimization of geometry. This would manifest in a flat ironed shape of the carbacation, and a ammonia like pyramidial one for the carbanion.
There however is no evidence to support the hypothesis, the two shapes eventually reported do not differ significantly which in both instances are flat.
5. archive content
tree
.
├── carbanion_cation.xyz
├── carbanion_radical.xyz
├── methyl_carbanion_obabel_optimized.sdf
├── methyl_carbanion_obabel_optimized.xyz
├── methyl_carbanion.sdf
├── methyl_cation_obabel_optimized.sdf
├── methyl_cation_obabel_optimized.xyz
├── methyl_cation.sdf
├── methyl_radical_obabel_optimized.sdf
├── methyl_radical_obabel_optimized.xyz
├── methyl_radical.sdf
├── readme.html
├── readme.org
└── superposition_methyl_carbanion_methyl_cation.png
1 directory, 14 files