# Before the CIP system was created, how did chemists ensure that the D/L system matched the R/S assignments?

The ᴅ/ʟ system I have referred to is the one chemists used before the Cahn-Ingold-Prelog (R/S) system where they arbitrarily assigned the (+) and (-) isomers of glyceraldehyde, 'ᴅ' and 'ʟ' respectively and related other carbohydrates via reactions with retention of configuration. They were eventually found to match the (R) and (S) assignments (for the lowest asymmetric carbon in Fischer projections) of those molecules today (particularly biomolecules), and that is where my question arises.

Even though the molecules were related to glyceraldehyde through reactions that proceeded with retention of configuration, it is also known that the resulting molecules may not always retain the absolute configuration of the original molecules (which depends on priorities of the new groups). And since it was before CIP's (R) and (S) system came into place, there was no way to know if the molecules had changed absolute configuration during the reactions. So how did the chemists manage to retain the configuration of all those molecules with respect to the glyceraldehyde isomers at all?

• Are you only concerned with carbohydrates? Other compound classes can't necessarily be related to glyceraldehyde, unless you are using a convention that I'm not familiar with. – jerepierre Feb 11 '16 at 21:53
• Yes, I am talking about carbohydrates or any other molecule that could be successfully related to the Glyceraldehyde enantiomers, for that matter... – Phill2 Feb 12 '16 at 3:02
• I'm a little confused by the question as the d,l assignment is a physical property and there is an implication (through the mention of "retention of configuration") that an assignment convention is also being used (like the Cahn-Ingold-Prelog system). These assignment systems do not align. – Beerhunter Feb 12 '16 at 21:05

Chemists actually still use D- and L- prefixes today, especially for carbohydrates where using it saves a lot of characters: Compare D-ribose to (2R,3R,4R)-2,3,4,5-tetrahydroxypentanal. The nomenclature is also in common use for amino acids, since all proteinogenic amino acids (except for glycine) in animals are L-configured (Bacteria, smart as they are, have many exceptions), while cysteine has the opposite absolute configuration to all other amino acids due to sulfur’s high priority. But your question revolves around history.

When Emil Fischer first devised the D- and L- nomenclature in 1891, not much was known about stereochemistry at all. It was known that carbon took four bonds and what the elemental composition of monosaccharides was. And methods were known to find out what a compound was, but much less sophisticated than nowadays. You would essentially derive it back to known products via known reactions and measure melting point (solids) or refractive index (liquids) for additional confirmation. Optical activity had been known since 1850, so that was probably performed in Fischer’s lab, too. At the same time, Pasteur performed his experiments, separating a racemic mixture of tartrate crystals into the enantiomers.

The choice Fischer made, when he assigned the D-configuration to (+)-glycerinaldehyde (the plus sign meaning that linear polarised light is rotated clockwise), coming from Latin dexter meaning right, was arbitrary. Not even essentially arbitrary, it was completely arbitrary. He probably knew he had a $50~\%$ chance of getting it wrong but in no way was he going to get proven wrong anytime soon. He decided it fitted nicely: the glycerinaldehyde that polarised to the right (clockwise) shall be the one with the asymmetric hydroxy group pointing to the right. He used that as a basis for all further sugars, knowing that all reactions he performed would take place on the aldehyde and leave the hydroxy group unchanged.

There was a fair bit of understanding how reactions worked even then. The first step to most sugar syntheses starting from glycerinaldehyde is the addition of a cyano group to give a cyanhydrine: Attack to the aldehyde, forming a new carbon-carbon bond and nothing else.

$$\ce{HOH2C-CHOH-CHO + N#C- + H+ -> HOH2C-CHOH-CHOH-C#N}$$

In any case, once a reaction like that was performed, derivatisation back to known compounds was done, and they would have realised if something had gone wrong with the hydroxy group. So we can be pretty sure they knew what wouldn’t influence it.

It wasn’t until half a century later and the X-ray crystal structure of glycerinaldehyde that one actually found out what the configuration of its carbon atom was. When it was proven that he had backed the lucky horse and the hydroxy group of D-glycerinaldehyde indeed pointed rightwards according to his convention, Fischer probably smiled a big smile from up in heaven. But if he had been wrong, he would have shrugged and thought ‘better luck next time.’

Now how does all this connect to the absolute configuration? Not at all! The absolute configuration is assigned on a per asymmetric atom basis. You need to know what the atom’s surroundings are like and can then apply a descriptor directly to it. Most notably, it was not the molecule as a whole that served as the reference but only the neighbouring groups and their priorities (read: atomic charge values). Any similarities to other stereodescriptor schemes, in use or not, was entirely accidental and unintended.

You have to remember that the retention of configuration in atomic terms does not automatically mean that the absolute configuration is retained. Take for example the Mosher acid as depicted below (image was taken from Wikipedia, where a full list of authors is available):

(R) and (S) Mosher acid.

Assigning the (R) configuration to the structure on the left is trivial. Assume you were to turn the (R) Mosher acid into the corresponding acid chloride. Note that although the asymmetric carbon was not touched the acid chloride turns out to be (S)-configured! So absolute configuration is something far more ‘wobbly’ than stereocentres are; make a remote change and influence an absolute configuration although that part of the molecule wasn’t touched.

Incidentally, that is the beauty of the Fischer nomenclature, especially when you consider the age in which it was invented. You don’t need to know which oxygen is pointing where you don’t need to figure out a long chain of priorities. You know you started at D-glycerinaldehyde so your sugar is now D-ribose. And you can take a pen and draw the lowest hydroxy group pointing to the right.

Do note make the mistake and try to correlate Fischer nomenclature with the sign of optical activity or with absolute configuration; all attempts will fail. Each has their use, their place to shine and none treats all similar molecules in the same way. Use wisely.