Stereochemistry, is the aspect of chemistry concerned with the different spatial arrangement of atoms in molecules and compounds and the effect that those differences have on their physical properties and reactivity. The significance of stereochemistry even in everyday life is famously demonstrated by the two isomers of carvone, where the (S)-(+)-form of carvone provides the odor of caraway while the (R)-(-)-form instead smells like spearmint.
 (S)-(+)-carvone caraway |
 (R)-(-)-carvone spearmint | ||
Chemical structures are inherently three dimensional, but they are commonly depicted on two-dimensional media such as printed paper or electronic computer displays. Any depiction of a three-dimensional object on a two-dimensional surface is going to require some level of distortion.
Chemical information has traditionally been communicated using two-dimensional media such as the printed page and more recently the computer screen. The problems associated with the communication of three-dimensional information in two-dimensional media are far from unique to chemistry. The introduction of perspective, for example, was one of the hallmarks of early Renaissance artwork. Although the use of perspective is commonplace today in pictorial images, it is less commonly used in conjunction with symbolic information: mapmakers use contour lines to indicate elevation or varying shades of blue to indicate ocean depth.
In chemistry, the need to show the true three-dimensional molecular architecture in two-dimensional media has given rise to a variety of structure drawing conventions. One convention depends on perspective to convey spatial relationships, just as a pictorial image would. The use of perspective in structure drawing is discussed here, but in its own section. By far the most common way to represent spatial configurations in chemical structures is through the insertion of special bond types -- bold, hashed, dashed, and/or wedged -- into an otherwise planar depiction of a chemical structure. Each special bond type would indicate the spatial arrangement of two atoms in relation to each other, usually specifying that one or the other atom was closer to or further from the viewer relative to the plane of the diagram. The proper use of hashed wedged and solid wedged bonds occupies the majority of these recommendations.
Historically, the varying conventions for depiction of configuration have caused confusion among chemists. Any one diagram may or may not have had a single interpretation either by the chemist who viewed the information or, as is now becoming more important, by the computer into which the information was stored. This publication contains a self-consistent series of recommendations for the unambiguous depiction of molecules in two dimensions using standards that are for the most part understandable by both human and machine.
Throughout this publication are numerous examples of chemical structures drawn in styles that are labeled as "preferred", "acceptable", "not acceptable", or occasionally "wrong". Since the depiction of chemical structures is something of an art form and will likely remain so, it is worthwhile to clarify the meaning of those terms as they are used here.
A chemical structure diagram is most commonly used simply as a means of identification, a way to answer the implied question, "What is the chemical structure of X?". The styles labeled as "preferred" show how the configuration of a structure should best be indicated in such cases, where there are no other overriding concerns. These depiction styles are generally applicable across many classes of chemical structures.
Sometimes, however, overriding concerns are present. Steroids, for example, must be drawn in a specific fixed orientation. A complex structure might need to be distorted in order to avoid overlap in other parts of the diagram. Bridged ring systems can be particularly interesting, since the topology of the ring system itself can force its bonds into orientations that are not otherwise seen in acyclic systems. Solid wedged and hashed wedged bonds should not be placed between two atoms that are both asymmetric except when literally unavoidable; that restriction alone accounts for many of the exceptional cases in this publication. The diagrams labeled as "acceptable" indicate additional depiction styles that could be considered if the preferred style is inappropriate for some well-considered reason.
Many of the structural depictions included in this document are provided as counterexamples, offering clarification of how structures should not be shown. Those depictions are labeled as "not acceptable", indicating that they should be strongly avoided in normal usage. Where possible, they have been accompanied by further description of why they are not acceptable, and why the alternative depictions are preferred or more acceptable.
Finally, a small number of examples are labeled as simply "wrong". Those show representations that should be avoided in all cases, generally because they depict something that is either self-contradictory or because they accurately represent a substance other than the one intended.
Some structural depictions are described as being formally correct, formally incorrect, or formally ambiguous, referring to whether a depiction might possibly represent the intended configuration from a strictly logical or formal analysis. For example, it is formally incorrect to depict a tetrahedral configuration using four wedged bonds connected to a central atom. Such a depiction would imply that all four substituents are on the same side of the central atom (all "nearer" the viewer relative to the plane of the diagram), whereas the geometrical definition of a tetrahedron precludes such an arrangement. The formal correctness of a structural depiction is related to the acceptability of the depiction, but the two are not exactly the same. Some depictions may be formally incorrect but still preferred because of long-standing convention. Other depictions may be formally correct but not acceptable. Discussions of formal correctness are included principally in cases where the correctness and the acceptability are different.
Several of the depiction styles include descriptions with specific angular measurements. For the sake of readability, angular measurements are listed with exact numerical values, such as 180°. Unless otherwise specified, all such measurements should be considered to be approximate, and specifying a range within roughly 10° of the listed value. The same applies to textual descriptions of angles, so the term "linear" should be interpreted as "forming an angle between 170° and 190°". In other words, two bonds that look nearly linear should be treated as exactly linear, even that is not exactly true for their actual geometric relationship.
Similarly, any mention of bonds being "adjacent" refers to their appearance in the two-dimensional representation. Any bond in a physical (three-dimensional) tetrahedron is physically adjacent to every other bond, but in a two-dimensional representation it is depicted as adjacent to only two others, and "opposite" to the third.
The recommendations in this publication are intended for use in structural diagrams drawn in the "standard" two-dimensional format where single bonds are represented with one line segment connecting a pair of atoms, double bonds are represented with two parallel line segments connecting a pair of atoms, atoms are labeled with atomic symbols (or not shown at all in the case of carbon atoms), and so on. There are other valid ways to represent structures including Newman projections, ball-and-stick models, and many others. These recommendations should not be overgeneralized as applying to anything beyond the "standard" two-dimensional chemical structure diagrams.
This publication extends and supersedes the section titled "Graphic Representation of Three-Dimensional Structures" in the earlier publication on the Basic Terminology of Stereochemistry [Moss, G. P. Basic Terminology of Stereochemistry. IUPAC Recommendations 1996. Pure and Applied Chemistry 1996, 68, 2193-2222.]. Only issues related to the depiction of stereochemistry are discussed here; a future publication is planned that will make recommendations regarding non-stereochemical aspects of chemical structure depiction.
All single bonds attached to nonstereogenic atoms should normally be drawn as plain bonds,
that is, as simple thin lines that are not bold or dashed or hashed or wavy or
adorned in any other way.
The use of stereobonds (
,
,
,
,
, etc.) at
nonstereogenic atoms should be avoided.
| PREFERRED | NOT ACCEPTABLE | NOT ACCEPTABLE |
 |
 |
 |
It is always acceptable to use stereobonds when emphasizing three-dimensional configuration, whether the associated atoms are stereogenic or not. For example, solid wedged bonds and hashed wedged bonds might be used when designating the pro-R and pro-S substituents on a prochiral tetrahedral center, or when depicting syn, gauche, or anti conformations of a torsion angle. This does not contradict the previous paragraph, as prochiral centers and torsional angles are in fact not normally emphasized as such. The use of solid wedged bonds and hashed wedged bonds in nonstereogenic environments is outside the scope of these recommendations.
Plain bonds should be used for nonstereogenic atoms even though the atom’s substituents are not physically coplanar.
| PREFERRED | NOT ACCEPTABLE |
 |
 |
When attached to stereogenic atoms, plain bonds indicate the set of bonds that is to be considered as being in the plane of the paper. Other solid wedged or hashed wedged bonds are considered to extend above or below this plane. As a general rule, structures should be drawn to maximize the number of plain bonds, although there are exceptions (particularly for inorganic complexes with coordination numbers greater than four).
| PREFERRED | ACCEPTABLE | PREFERRED | ACCEPTABLE |
 |
 |
 |
 |
As discussed above, it is always acceptable to use solid and hashed wedged bonds when emphasizing the perspective of three-dimensional configuration, even if an alternate representation is possible that uses fewer stereobonds. Again, however, such exceptions should be considered only when necessary to accommodate some other aspect of the diagram and are not preferred in the general case.
| ACCEPTABLE |
 |
Simple rings and fused ring systems should be drawn using plain bonds within the rings. If configuration is to be indicated with hashed wedged and solid wedged bonds, those bonds should be restricted to acyclic substituents. This is true even though in most cases the ring atoms will not all be coplanar; the implied coplanarity should be considered on a per-atom basis only, just as with acyclic compounds. Depiction of a solid wedged or hashed wedged bond within a ring is acceptable only in cases where no reasonable alternative is available (as in a spiro fusion, see ST-1.3, or ) or where all acyclic neighbors are themselves stereogenic (ST-0.5).
| PERFERRED | NOT ACCEPTABLE | NOT ACCEPTABLE |
 |
 |
 |
Structural diagrams that depict configuration must be prepared with extra care to ensure
that there is no ambiguity. As discussed above, plain lines depict bonds approximately in the plane of the
drawing. Bonds to atoms above the plane should be shown with a solid wedge
(starting from an atom in the plane of the drawing at the narrow end of the
wedge). A bold bond
has sometimes been used instead of a bold
wedge but is not recommended.
Unfortunately, bonds below the plane of the drawing have
historically been represented in many different ways. Each of those
representations has involved a bond drawn
with small line segments either parallel ()
or perpendicular (,
)
to the main axis of the bond. The two schools of thought that prefer the use of
a hashed wedge bond
assign it
two directly opposite interpretations -- one school says that the atom at the
narrow end should be considered in the plane of the paper, while the other says
that the atom at the wide end should be so considered. Additionally, a dashed line
is often used also to represent a partial bond, delocalization, or a hydrogen
bond. The biggest problem is that there is no way to intuit an author's desired
meaning from a chemical structure drawing alone. For these reasons, chemical structure diagrams
should be created so as to avoid these sorts of bonds
when doing so will not otherwise compromise the clarity of the drawing, and
solid wedged bonds should be used instead:
| PREFERRED | ACCEPTABLE |
 |
 |
Although a solid wedged bond can often be used instead of a hashed wedged bond by drawing the structure slightly differently (as above), that situation is far from universal. There are many cases where the use of a hashed bond produces a more aesthetic structure. One common situation happens with ring systems, where it is often preferable to draw an explicit hydrogen with a hashed bond rather than put a solid wedge on a ring bond. Another common situation arises when showing a series of enantiomers or diastereoisomers where the contrast between the solid and hashed types directly emphasizes the chemical differences.
In cases where such below-the-plane bonds are required, there is no option that will please everyone. Earlier
recommendations proposed the use of an unwedged hashed bond
,
but today such bonds are in fact encountered in the literature least frequently
of all options. This document now recommends the use of a hashed wedged bond
interpreted
in a sense similar to the solid wedged bond (starting from an atom in the plane of the drawing at the narrow end of the
wedge). This recommendation is made primarily because the hashed wedge is easier
to analyze visually than the unwedged type, particularly when it is used
in a sense similar to the wedged bold bond. Strictly speaking, unwedged bold and hashed
lines show that BOTH atoms are above or below the plane of the drawing (as is
used in Haworth drawings of carbohydrates). If these bonds are used to convey
chirality, they require extra effort to determine which atoms are truly
asymmetric and which are not. The wedged hashed bond contains more
-- and more useful -- information than the unwedged version. While the use of an unwedged hashed bond is often unambiguous,
such a bond is not always unambiguous, and in the rare cases that cannot
avoid having a hashed bond between stereocenters (see below), the unwedged
version will never be unambiguous. Additionally, the consistent use of a
wedged bond is simpler: in each case, the sharp end of the bond
shows the stereogenic center, and there is no need to analyze both ends to figure out
what was intended. After careful consideration, it appears that the only
responsible recommendation is that only the hashed wedged bond be used when
representing below-the-plane bonds..
Admittedly, strong cases can be made in favor of other systems. A system based entirely on a perspective-based analysis is possibly the most popular of these other systems. In a perspective-based analysis, more-distant objects are always perceived as smaller than objects that are nearer. Accordingly, a wedged bond is interpreted as extending above the plane of the paper from the narrow end to the wide end, and extending behind the plane of the paper when considered from the wide end to the narrow end. Lin et al. have shown that a system of this type can be internally self-consistent and can be used to represent most types of tetrahedral configurations (Shu-Kun Lin, Luc Patiny, Andrey Yerin, Janusz L. Wisniewski and Bernard Testa. One-Wedge Convention for Stereochemical Representations. Enantiomer 2000, 5, 571-583. http://www.unibas.ch/mdpi/molecules/wedge/wedge2.pdf). Unfortunately, it cannot be used in all cases. Since this system always specifies the configuration of atoms at both ends of every wedged bond, it cannot be used to depict, for example, a tetrahedral center of known configuration connected to four other centers of unknown configuration. Additionally, the One-Wedge system as described by Lin et al. eliminates the use of a hashed wedge bond entirely, thereby discarding a very popular drawing style widely used for many years. This simplification has some appeal in terms of simplicity of usage, but the benefits of a hashed bond -- in particular, the ease with which a hashed bond can be distinguished from a bold bond at a quick glance -- are too great to eliminate.
Others have recommended a perspective-based analysis that preserves the use of a hashed bond. The scheme described earlier, where a hashed wedged bond would be interpreted as "down" from the wide end to the narrow end, is a perspective-based approach. Unfortunately, this system not only suffers from the same weaknesses as the One-Wedge approach above, but additionally results in a scheme where a solid wedged bond and a hashed wedged bond -- while looking quite different -- actually represent exactly the same concept. (for a good example, consider Figures 10a and 11b in Maehr, Hubert. Graphic Representation of Configuration in Two-Dimensional Space. Current Conventions, Clarifications, and Proposed Extensions. J. Chem. Inf. Comput. Sci. 2002, 42, 894-902.)
The responsible recommendation is to reject a perspective-based approach, and rather to interpret the bonds as more-abstract graphical objects. To summarize, a solid wedged bond should be interpreted as projecting above the plane of the paper from the narrow end to the wide end (with the wide end being nearer to the viewer), and a hashed wedged bond should be interpreted as projecting below the plane of the paper, also from the narrow end to the wide end (with the wide end being further away from the viewer).
| PREFERRED |
 |
That said, it is possible to use the other bond types in ways that are unambiguous. In particular, any type of hashed bond is unambiguous when connected to an atom with no other substituents (a hydrogen atom, a chlorine atom), no other non-hydrogen substituents (a hydroxy group, an amino group), or at most one other non-hydrogen substituent (an ethyl group, an ethoxy group). In each of those cases, a viewer would have no difficulty in understanding that the hashed bond projects below the plane of the paper from the other atom to that atom. Accordingly, both the unwedged hashed bond and the hashed wedged bond with reverse directionality are acceptable when used exclusively in those cases.
| ACCEPTABLE | ACCEPTABLE |
 |
 |
However, anyone using bonds of these types should be aware that they are not the most common bond types used to represent behind-the-plane configuration. At best, an author using these types may cause a moment of confusion as a reader figures out which convention was intended. In extreme cases, a reader might actually interpret the structure incorrectly – for example, by interpreting an unwedged hashed bond in the convention where it would represent relative (and not absolute) configuration, as discussed above. Fortunately, such cases are rare. The hashed wedged bond (interpreted from the narrow end to the wide end) remains preferred.
Because the unwedged hashed bond is visually non-directional, it absolutely must not be used when connecting two stereocenters. (See also ST-0.5.)
| NOT ACCEPTABLE |
 |
The dashed bond should be reserved for the indication of partial bonding of various types, including hydrogen bonding. It should not be used to represent configuration.
| NOT ACCEPTABLE |
 |
If configuration is unknown, this can be indicated explicitly by a wavy line
.
The waves in such a line may be rounded or angular, but they should be of
constant amplitude. The use of "wedged wavy" bonds is
discouraged.
In most cases, a wavy bond is conceptually identical to a plain bond, and a plain bond should be used instead. Wavy bonds are most often used in general usage when an author wants to place special emphasis on the unknown nature of the configuration at a specific position; the wavy bond would usually be accompanied by additional explanatory text in that case.
| PREFERRED | ACCEPTABLE |
 |
 |
Historically, wavy bonds have frequently been used to represent a mixture containing several enantiomers or diastereoisomers. That is reasonable, since the configuration of any arbitrary structure within such a mixture is indeed unknown. It is preferred to use a plain bond to depict a mixture, just as it is preferred when depicting single structures. If the nature of the mixture needed further emphasis, a wavy bond would remain acceptable for mixtures as well.
Some specific classes of compounds have further conventions for depiction of unspecified configuration within those classes. In steroid nomenclature, for example, atoms 8, 9, 10, 13, and 14 are always assumed to be in standard configuration unless explicitly denoted with a wavy bond. In Haworth projections of carbohydrates, the anomeric carbon is frequently indicated with a wavy bond when representing a mixture of anomers. These usages of a wavy bond should not be extended to structures outside those classes.
| PREFERRED | PREFERRED |
 |
 |
Stereobonds between stereocenters should be avoided at all costs. This was stated above but bears repeating because it is so important.
| PREFERRED | NOT ACCEPTABLE |
 |
 |
 |
 |
 |
 |
In rare cases -- for example when one stereocenter is completely surrounded by four other stereocenters -- a stereobond must be present between two stereocenters. There is no ideal solution in such cases, as some ambiguity is unavoidable. In this case, like the others, only the atom at the narrow end of a stereobond should be considered as having a specified configuration.
| ACCEPTABLE |
 |
If the central atom has a specific configuration and the surrounding atoms also have specific configurations, additional stereobonds must be added to the surrounding atoms. It is preferred to select a depiction style for the central atom that has as few stereobonds as possible, since that will eliminate completely any ambiguity regarding the intended configuration at some of the adjacent atoms.
| ACCEPTABLE | PREFERRED |
 |
 |
Asterisks have traditionally been used in several contexts: not only to indicate the presence of a stereogenic center, but also to specify isotopic labeling or excited states. Accordingly, the use of an asterisk as an indicator for any one of those items (including as an indicator for stereochemistry) is potentially misleading and should be avoided. Any use of the asterisk should be undertaken with extreme care, and should generally be accompanied by additional descriptive text that explains its meaning in that specific context.
Absolute stereochemical configuration can be described using descriptors derived from the Cahn-Ingold-Prelog (CIP) priority rules. While quite powerful, the CIP rules can also be difficult for even skilled chemists to interpret correctly. CIP descriptors may certainly be present in a diagram to indicate the absolute configuration, but should never be used as a replacement for proper stereochemical representation (hashed wedged and solid wedged bonds). Obviously, if a CIP stereodescriptor is included in a diagram, extra care should be used to ensure that it is accurate.
| PREFERRED | ACCEPTABLE | NOT ACCEPTABLE | WRONG |
 |
 |
 |
 |
| Other items for discussion |
