Home » Still In The RBF

Some musings on chemical drawings…   

21 February 2008 15,632 views 39 Comments

I realise this is a break in your usual Tot. Syn. programming, but I’ve been pondering… I know, it’s not healthy, and I should really get back to blogging another synthesis, but this kinda thing interests me. Y’see, after blogging the Danishefsky synthesis last week, and the Corey synthesis on Monday, I was intrigued by the differences in graphical representation of structures. Both were published in the same journal (JACS), in the same month, so one might expect a consistency of style, but not in this case. I’ve taken a pair of screen grabs to illustrate this:

Most obvious is the use of different descriptors for stereochemistry – Danishefky’s is more traditional, with unwedged bonds and more importantly, H-dots. They also differ in their representation of a transformation: as can be seen in the Danishefsky case, he uses lettered representation, which is expanded in the caption. Corey, on the other hand, describes the condition above/below the arrow. I find amazing that both styles were found in the same journal, and this inspired me to delve back into the pre-ChemDraw era, and have a look at Ye Olde Chemical Drawings…

(To get a fair representation of what was standard issue for a particular era, I didn’t take the following snapshots from journals – rather, I nipped along to the Nobel Prize website, and examine the lecture publications of a a variety of chemistry legends. If you haven’t visited the website before, I urge you to take a history lesson!).

Let’s start way back in 1937 (when drawing were done with a hammer, chisel and a block of granite), when Paul Karrer won for his work on Carotenoids, flavins and vitamin A and B2. What we’ve got here is some quite extensive labelling of every atom in the molecule – no ambiguity with this style, but I must say that the representation of single/double bonds is hilarious! I wonder what other punctuation could represent? I think that any of [~ ; ! #] could have interesting interpretations…

Moving forward a decade, now we can examine the work of Sir Robert Robinson, who was lauded for his work on plant extracts and alkaloids in particular. Although the geometry is similar (those hexagons aren’t particularly regular!), things have definately moved forward a bit. The hydrocarbon vertices are unlettered, and most of the bonds are symbolised using lines. However, a personal annoyance apprears here – the numbering of compounds using Roman numerals! Argh…

A bit of a jump to 1965 and we have R.B. Woodwards work on natural products in general, with a lecture focusing on cephalosporin C. Stereochemistry rears it’s head in this example, though using a somewhat depreciated style. I personally love the little “tert” symbol used on the t-butyl carbamate.

Now were talking – these structures taken from Sir Derek Bartons 1969 Nobel lecture on conformation. Again, the stereochemical style is outdated, but the six-membered rings are actually hexagonal! Still numbering like it’s 80BC though…

And now to the modern era. You’ll probably recognise the ligand below, and think of Sharpless – this is from his award in 2001 on “chirally catalysed oxidation reactions“. Everything seems okay in this, except the hashed bonds look a little odd. Still, this is what we’re used to.

Lastly, and as an exception, the snapshot below is from a 2006 paper by Bill Roush (amphidinolide E), which again shows that there is no real consistency to the chemical notation – those dashed bonds are fairly unique to his style. If I get the opportunity, I’ll ask why the journals don’t enforce their standards. BTW, the actual standards are described by IUPAC in their Gold Book. Not a riveting read, but if you’re ever unsure…

Lastly (and with some slight smugness), I’ll admit to being impressed with seeing my particular ChemDraw settings used in journals/presentations. I designed them to be just a little clearer than the ACS standard settings. To make things easier, I’ve recently put the ChemDraw stylesheet on-line – you can download them here.

If you’re still wondering what the hell I was going on about… do not adjust your set… normal service will resume forthwith…

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)


  • Malcolm says:

    I did a publication in carbohydrates and the reviewer was very picky about representation (wedged etc.). I find those non-wedged bonds like Danishefsky confusing because sometimes people use them to highlight bonds rather than show stereochem. what is with that?
    However, I prefer danishefsky’s representation. I was appalled that Corey’s wasn’t slammed for using huge, bold structures with big writing. It looks like a kid did the chemdraw. The whole thing (except the chemistry) was sloppy, pasting structures also in such a way that they expand beyond the margins. Therefore, when you print, much of the structure is missing. I have noticed this in a lot of JACS ASAP lately (Grubbs/Stoltz tot syn comes to mind) and it needs to be addressed. I have spent time trying to figure out mechanisms when the reason I was confused is that the whole stucture did not print!

  • Hap says:

    1) I think Prof. Corey has used the same ChemDraw settings for at least 20 years – the structures look similar to those in Logic of Chemical Synthesis and pretty much every paper his group has published in the last 15(?) years. Stoltz probably got the settings from Prof. Corey. I get tired of them but they are at least clear.

    2) I don’t really like the solid dot/no dot for beta/alpha-oriented protons at ring junctions – no dot can mean unknown stereo. If the space is too small to fit the proton in a drawing, than the squiggly bonded H will also be difficult to fit, but the stereo is useful to know. Somoetimes people will draw steroids in which all of the stereo at protonated fusions is omitted, figuring that the stereo can be assumed; I don’t really like that, either.

    3) I guess I don’t mind Roush’s wedged/dashed stereo structures, but they may be confusing if the structures are reduced in size (is that a badly printed single bond or actual stereo?). The horizontally dashed wedge bonds are unambiguous, while the dashed bonds could be used to mark a connection in certain cases. They shouldn’t be used for disconnections, ever.

  • Pete says:

    I like it when the structures, especially the more extravagant ones, are drawn three dimensionally, as the grad student/post doc/PI seems it…Phil Baran gave a talk here not too long ago (2.5 yrs ago) and one of his slides was an actual image from his grad students (I believe Richter’s) notebook of welwitindolinone drawn three dimensionally. If thats how he is perceiving the molecule and the transformations being used to make it, then isn’t that what we should see in the paper?

  • GYA says:

    I totally disagree and would be surprised if most people don’t as well (in paticular folks who’ve enjoyed forays in total synthesis). I feel Corey’s (and Stotlz’s as well) are almost the best settings out there. The structure is highlighted and clear with nice bold bonds and letters. Personally, I’m a fan of wedges and dashes that are tapered and not the ones that remain the same size start to finish. I’m also a huge fan of drawing with perspective. I think its essential in certain contexts and a plus in others (Corey’s recent synthesis clearly highlights the later). Also, I think you need to figure out how to use a printer Malcolm (never had those issues with Corey’s work).

  • Wavefunction says:

    In general I love Corey’s style and representations.

  • excimer says:

    I prefer Corey’s style, with the bold style. I despise the lettering reagent system, which forces me to look elsewhere, typically in an annoyingly long figure caption, for the actual reagents. (Then again, it’s sometimes fun to guess what they used in the first place, but when you’re trying to follow a complex synthesis, regents above/below arrows flow much more readily.

  • GYA says:

    Second the motion for reagents in schemes above/below arrows!

  • Big_C says:

    Am I alone in liking the reagent lettering system?

    C’mon folks, its just a twitch of an eye away. Besides, the paper looks much sleeker when its lettered.

  • antiaromatic says:

    I’m not a big fan of the lettered stuff when they’re in schemes that go from a-m. At that point, it really gets annoying when you try to find a reagent. I do understand, however, that there are cases when you have a communication and to include all of the reagents above the arrows would be space prohibitive.

  • milkshake says:

    The fat Chemdraw template setting comes from the times of Xerox-produced overhead transparencies. Normal-style drawings turn spidery or nearly invisible when projected with a transparency projector.

    Oh, the kids these days – they don’t know what the electromechanic analytic balances look like.(Cue in the rustle of hand playing with a pocketful of change)

  • sjb says:

    Part of the issue I think here is what does the depiction of a hashed bond actually mean? If you draw a hash like Corey’s at the 6,5 ring junctions, does that mean that the bond is going away from you, or from a perspective point of view towards you? I know some people who subscribe to one interpretation, and others to the opposite. Makes life interesting when you try and report to both. ;)


  • ipa says:

    Instead of voting for my favorite chemdraw settings I feel this post has brought up a good point. WHY are schemes from two papers in the same journal different? If there is a standard across the board (IUPAC) then shouldn’t all schemes in all journals regardless of publisher be the same? I don’t know the answers nor do I want to speculate but I am interested in seeing what others think about this.

  • Potstirrer says:

    I’m a big fan of placing the relevant reagent, catalyst, solvent, and or temperature above and below the arrow with a letter referencing a complete legend below. Could be just “Swern” or “Grubbs’ II” or TMSOTf, -78 C”. This way you get a quick idea of the reaction conditions without having to search, the scheme isn’t too cluttered, but you can also gather all of the relevant information if you want to.

  • Smitty says:

    I was about to post on the same thing as Potstirrer, except I don’t like that format. For relevant examples, see all of Nicolaou’s syntheses from ACIEE this year (the epothilone papers aren’t as consistent). What he does is write “TBSOTf, 2,6-lutidine” (as an example) on the arrow and then references the caption that gives equivalents, solvents, temperature, reaction times, yields (hiding them down there?), etc. I don’t like it because I think it looks very cluttered and is somewhat redundant. I only say somewhat redundant and not completely redundant because sometimes the information on the arrows is not enough to figure out the transformation. For example, in the chiral pool based synthesis of platensimycin, scheme 2, step C is an oxymercuration reaction. If you look at the arrows, there is no water in the reaction, thus leading to some confusion. The caption explains that it is done in a 1:1 v:v mix of THF/water – maybe it can be assumed, I’ve never worked with mercury acetate, but I think part of the point of a scientific paper is giving as much complete info as possible so there is no gray area – I’m not an oxymercuration expert, maybe the reaction can be done without water to afford some sort of hydrogenation product? There are a few other minor examples where equivalents would help a reader understand the transformation, but aren’t provided on the arrow (so does that means it can be assumed to be all 1:1:1…?). This is a very minor point, but why give some information and not all?

    I prefer Tot. Syn.’s chemdraw settings and conventions, such as seen in his other posts – the least ambiguous, scales up and down well, doesn’t have too many assumptions.

  • Hap says:

    A twitch of an eye? More like a nervous tic as I spend five minutes trying to find steps a-c and then find a i) ii) iii)… With communications it’s necessary to draw schemes in that manner because of space constraints, but it makes the schemes hard to follow. Most schemes where the reactions are given in that manner have at least five or six lines of dense print listing conditions, which also makes it hard to find what you want (and for detail you inevitably have to look to the SI or elsewhere, anyway).

    I have seen two different types of dashed bonds with the wedges pointing in different ways, but I don’t think I actually seen it used, probably because it is so confusing. The only time the normal dashed bond connection is ambiguous is when stereocenters are connected by one, and that is deprecated most everywhere.

  • Radical says:

    The non-wedged bonds are used to indicate relative stereochemistry, such as in Danishefsky’s synthesis, which is racemic, whereas Corey’s wedged bonds indicate absolute stereochemistry. I don’t think many people, including young professors, know and still use this. I think it’s unfortunate, since it is a very useful convention that can provide quick information about a synthesis and the compounds in the scheme.

  • Hap says:

    I’ve seen the use of wedged and dashed bonds to imply relative or absolute stereochemistry, but it doesn’t seem to be used consistently enough to be useful (which makes it difficult to gather anything from its use in a scheme).

  • For all of you reading about graphical representations of chemical structures you will likely be interested in this article published this month:

    (IUPAC Recommendations 2008)


  • TWYI says:

    Prefer Corey’s structure drawings/reagents over the arrow

  • GYA says:

    Danishefsky’s synthesis is actually enantioselective–not racemic.

  • TWYI says:

    TWYI says NO to dots in structures as a way of depicting stereochemistry.

  • Radical says:

    GYA: You’re correct, it isn’t racemic… However, it is asymmetric, not enantioselective, since they start with achiral starting materials, and chirality is introduced from an external source (i.e. reagent control).

  • Malcolm says:

    thanks for the style sheet man. I actually wrote an exam yesterday and used your style of displaying retro disconnections!!

  • GYA says:

    Lets get technical! The synthesis IS enantioselective (many also use chiral and asymmetric in this context interchangeably). In this case, enantioselective is a better term as one enantiomer of the DA adduct is SELECTED for rather than having one enantiomer removed by separation (this kind of synthesis would be better termed asymmetric or chiral). An asymmetric synthesis doesn’t mean that you have to start with achiral starting materials either. In fact, an asymmetric synthesis can derive its asymmetry by using the chiral pool (not achiral SM), utilizing asymmetric induction, or by using asymmetric catalysis. Also, the key step is enantioselective and it is CATALYST controlled and not reagent controlled.

  • Malcolm says:

    “Oh, the kids these days – they don’t know what the electromechanic analytic balances look like.(Cue in the rustle of hand playing with a pocketful of change)”

    yeah i looked online and i couldn’t even find a picture of what that is?

  • TWYI says:

    GYA – In future, perhaps Danishefsky should run his papers by you before submitting them?

  • milkshake says:

    Malcom: Here is one model, side-view


    the Chief is rather fond of them

  • antiaromatic says:

    GYA – I’m pretty sure that catalysts are considered reagents. In the context of that reaction, I agree that the catalyst is what promotes the asymmetry, but that doesn’t change the fact that it is REAGENT controlled. It is not SUBSTRATE controlled. If it were, he wouldn’t require the external source for stereochemical induction.

  • GYA says:

    antiaromatic-you’re right. defining it as catalyst control is just more specific (a catalyst is obviously a reagent, but a reagent isn’t necessarily a catalyst). imagine doing an enantioselective allylation of an aldehyde with a chiral allyl borane, the reaction is under reagent control (the chiral element from the borane is not conserved in the product).

  • Tot. Syn. says:

    Exactly what I was getting at in the previous comments. However, GYA explained it really quite nicely :)

  • Radical says:

    GYA has explained it quite nicely, except that he/she is completely wrong.

  • TheEdge says:

    From the IUPAC webpage, which contains a tremendous number of definitions for everyone’s enlightenment:
    stereoselective synthesis
    A chemical reaction (or reaction sequence) in which one or more new
    elements of chirality are formed in a substrate molecule and which
    produces the stereoisomeric
    (enantiomeric or diastereoisomeric) products in unequal
    amounts. Traditionally called asymmetric synthesis.

    Here’s the webpage for those who would like to use their interweb connection for something other than bickering over what are apparently the same term ( http://www.iupac.org/publications/compendium/index.html ) FWIW, they do not list the term “enantioselective synthesis”, although it’s meaning can be derived as a specific case from the definition above.

    This man (http://www.uottawa.ca/publications/interscientia/inter.4/eliel/eliel.html) who claims to have written books on stereochemical terms raises an interesting point about the term “asymmetric synthesis”. There have been some truly beautiful C2-symmetric synthesis, and many of us use C2-symmetric catalysts. Can these be said to be asymmetric?

    To further belabor the point, SD’s synthesis is, at best, enantioenriched, not enantioselective. This doesn’t in anyway take away from the beauty of the chemistry, though.

  • I dunno how important it is to have a defined style. Sloppy drawings should be excluded, but if you know what you’re looking at then there’s no reason to enforce uniformity across a journal. I think most Chemists like to treat their “brand” as art, and no artist likes being told to conform to the establishment.

  • Leo says:

    Personally, I find Danishefsky’s better, although both styles are acceptable. I’m uncomfortable about dashed bonds being also wedged. In my opinion, that’s a bit OTT -tapered should do. Whether the caption contains the letter a, (a), A. etc. (with conditions somewhere beneath) or whether the caption should be self-explanatory (a la Corey) is probably just journal dependent. Correct me if I’m wrong! Regards, leo

  • Tot. Syn. says:

    Leo, I’m afraid you are wrong. As I explained, both the screengrabs are from the same issue of the same journal, so it’s not a journal-specific thing at all. And according to IUPAC, Danishefsky’s drawing style is “unacceptable”… though I agree it is perfectly understandable…

  • Peter W says:

    As I am writing an organic chemistry mechanism workbook, and I struggle with the format issue. What size should I use and how to best present a reaction? Since I am writing a workbook, I need to leave enough room for the bold writer (who writes with a crayon?) and the petite writer. Currently, I have stayed with a modification of the default CD settings, but I expect to change that in the future.

    Looking at the two reactions, the Danishefsky drawing would scale much larger with a fixed bond length. He shouldn’t follow the standard. Corey’s is written as organic chemists think. However, I think it looks overly bold. Now that CD has provided more latitude in their settings, I would narrow the bond width and the hashing. I think the reagent font could be smaller as well.

    However, I like journals to have a standard. It is a place to start, but it should not upheld. (The format that Tot. Syn. is so proud of is better suited to the web and I do not think it serves print as well.) The objective is communication and it is the journals objective to constrain this to the least number of pages possible. If there is a rule as to how this is to be done, then I think that rule should be allowed to be broken. Indeed, in a science in which innovation is valued, standardization should not be rigidly upheld. The history of the structures bears evidence to that. I am pleased the structures of the 1920s, ’30s, or ’40s are not the standards of today.

  • Tot. Syn. says:

    Agreed – my template is best for screen display (powerpoint included!). As for style in text books, I again have to lean towards what I know best – Stuart Warren’s “Organic Synthesis”. The diagrams in this book are exceptionally easy-on-the-eye, unlike March or Carey & Sunberg for instance. Kurti and Czako’s “Strategic Applications of Named Reactions in Organic Synthesis” is also laudible, but veers a little too much into rainbow territory for my liking (though the extensive use of colour does make the mechanisms very clear).

  • Taitauwai says:

    Yes, agree with you. Have been using your ChemDraw style eversince… Thanks so much for the stylesheet. Domoarigato!