Both ends wedged. The axis is defined; standard InChI already handles this.
- New InChI
- /t2-/m0/s1
- Standard InChI
- /t2-/m0/s1
An allene has no stereocentre, yet it can be chiral. This page shows how InChI reads the handedness of the C=C=C axis from a drawing, and what is new: one wedge is enough, and the axis takes or and & labels.
1-Bromo-1-fluoro-3-chlorobuta-1,2-diene, drawn four ways. Strings from /t on:
| Standard InChI | New InChI | |
|---|---|---|
| two wedges | /t2-/m0/s1 | /t2-/m0/s1 |
| one wedge | no stereo | /t2-/m1/s1 |
| or1 | /t2-/m0/s1 | /t2-/s2 |
| &1 | /t2-/m0/s1 | /t2-/s3 |
With two wedges nothing changes. One wedge and the labels are new.
In C=C=C the two π bonds are perpendicular, so the substituents at one end lie in a plane at 90° to those at the other end. If each end carries two different groups, the molecule and its mirror image do not superimpose: the chirality sits on the axis, not on an atom. Chemists call the two forms Ra/Sa (or M/P).
InChI has no special layer for this. It writes the axis as if the central carbon
were a stereocentre, in /t, with /m for the mirror image. So all the
rules for stereocentres apply, including the enhanced labels:
The sample is the drawn enantiomer.
One enantiomer, unknown which.
Both enantiomers, 1:1.
The label goes on the central carbon of the allene.
Example: the allene above, both ends wedged.
A wedge at an end carbon shows which substituent points toward you. The other substituent at that end lies in the paper or behind it.
If an end has two equal groups there is no axis. Longer chains: see what InChI ignores.
Two wedges, one at each end, define the axis. New: one wedge is enough. The
flat end is taken to lie in the paper, the wedged end perpendicular to it, as IUPAC
allows. Only under -EnhancedStereochemistry; standard InChI is unchanged.
No wedge, or two wedges contradicting each other at one end: the axis is undefined and left out.
Canonical numbers in grey. The central carbon is 2: the axis is written as stereo atom 2.
As for a stereocentre, the axis gets a sign, + or −. A lone stereo element is always
written -; /m records whether the drawing had to be mirrored for
that.
| drawing | InChI | meaning |
|---|---|---|
| as drawn | /t2-/m0/s1 | no mirroring needed |
| both wedges reversed | /t2-/m1/s1 | mirrored: the other enantiomer |
/m0 and /m1 tell the two enantiomers apart; they do
not map to Ra/Sa directly.
With &1 on the central carbon, both mirror images are
in the sample, so /m is dropped and /s3 says "racemic":
Both drawings, as drawn or reversed, give this string.
/s2 = or, /s3 = &The prefix InChI=1B marks the string as enhanced.
An allene axis is one /t entry on its central carbon.
/t2-Each end needs two different substituents, or there is no axis.
New, with -EnhancedStereochemistry. Standard InChI usually needs two.
With no label or abs, only the prefix differs from standard InChI.
/t2-/m0/s1or gives /s2, & gives /s3, without /m. A label on
an end carbon is ignored.
/t2-/s3The axis is grouped like any stereocentre: lowest member of each group reads −, groups listed by class.
/m0/s1(5)3(2)No wedge, or contradicting wedges: nothing is written for the axis.
C=C=C only. C=C=C=C=C (four double bonds) is axial too but not supported yet; C=C=C=C (three) is E/Z.
Strings are shown from /t on; the part before is unchanged. "Standard" is the
same file without -EnhancedStereochemistry.
Both ends wedged. The axis is defined; standard InChI already handles this.
Both wedges reversed: the other enantiomer. Only /m changes.
A rotated drawing of case 1 gives the same InChI.
Only the Cl end is wedged; the Br/F end lies flat in the paper. IUPAC accepts this; standard InChI does not.
Only the F end is wedged.
One enantiomer, unknown which.
The racemate: both enantiomers, 1:1.
Labels work on a one-wedge drawing too.
The centre is known; the axis is racemic. The axis is canonical atom 2, the centre 5.
The centre is known; the axis is one configuration, unknown which.
Two independent axes, each unknown on its own.
No wedge: the handedness is not drawn. The axis is left out.
Both substituents of the same end point up: the drawing contradicts itself. Left out.
&1 names an end carbon, not the central one. The label is ignored; the wedges count as absolute.
One end carries two methyls. The molecule equals its mirror image: no axis.
An even number of cumulated carbons: the ends are coplanar. This is E/Z, written in /b.
Axially chiral like an allene, but longer than InChI supports. No stereo is written.
Input. A molfile with a wedge at one or both ends of the allene. To mark the axis
as or or &, put the
label on the central carbon (V3000 collection STERELn /
STERACn); a label on an end atom is ignored.
Command line. Add -EnhancedStereochemistry. The deck allene,
racemic axis:
$ inchi-1 allene.mol -STDIO -AuxNone -Key -EnhancedStereochemistry InChI=1B/C4H3BrClF/c1-3(6)2-4(5)7/h1H3/t2-/s3 InChIKey=GKWKCLIHJCGAMS-NVFNTWQMBA-N
Without the option the label is ignored:
$ inchi-1 allene.mol -STDIO -AuxNone -Key InChI=1S/C4H3BrClF/c1-3(6)2-4(5)7/h1H3/t2-/m0/s1 InChIKey=GKWKCLIHJCGAMS-REOHCLBHSA-N
From a program, pass the option to the InChI library like any other, e.g.
MakeINCHIFromMolfileText(molfile, "-EnhancedStereochemistry", &output)
or in the options string of GetINCHI. The InChI starts with
InChI=1B/: a beta feature, so the format can still change.