Rh,
Isolysergic acid referres to the 8-position stereoisomer of lysergic acid. Lysergic acid is (S)-isomer, and iso-lysergic acid is the (R)-isomer.
Yes, you can epimerize lysergic acid and isolysergic acid, since under slightly acidic conditions the carbonyl will undergo a reversible reaction to form the corresponding achiral enol. When the enol reverts to a carbonyl, it may form either the (S)- or the (R)-isomer, but lysergic acid is the more favorable of the two stereoisomers, due to some steric hindrace issues.
However, and I hate to say it, there is just no way you're going to get that double bond to move from the 8,9-position to the 9,10. The problem is that by doing so, you're adding a huge amount of thermodynamically unfavorable ring strain. Go ahead and make a model to try it out: with the 8,9- there's some flexibility in the non-aromatized rings, but by placing the double bond on the 9,10-position, you're forcing the 5, 8, 9, and 10 carbon atoms to basically lie in the same plane as the indole ring - thus eliminating any hope for the two 6-membered non-aromatic rings of conforming to anything resembing a "comfy" chair-configuration. This is one of the reasons why the 9,10-double bond is so prone to attack - adding across it relieves a huge strain on the molecule.
As a visual aide, here are lysergic acid and isolyserigc acid.
lysergic acid:
Molecule: (https://www.the-hive.ws/forum/faq.pl?Cat=#applet)
lysergic acid ("[H][C@@]12Cc3c[nH]c4cccc(C1=C[C@@H](C(=O)O)CN2C)c34")
isolysergic acid:
Molecule: (https://www.the-hive.ws/forum/faq.pl?Cat=#applet)
isolysergic acid ("[H][C@@]12Cc3c[nH]c4cccc(C1=C[C@H](C(=O)O)CN2C)c34")
Indeed, I'm sure the double bond feels quite comfortable conjugated with the indole ring, ignoring the ring strains. ;) I certainly wouldn't disagree with that statement.
Thank goodness for the Hive. This discussion began me thinking more about this paspalic->lysergic business. Yes, there is a great deal of added strain to the molecule, but at the same time, a bond in the 9,10-position would be conjugated with the indole ring, which is certainly an energy incentive. So, if one could find a mechanism by which the reaction could happen (arguments about sterics aside), then one would simply need to find conditions where the conjugation of the double bond would overrule the disadvantages of ring strain...
So, I did a quick literature search, and hit gold.
In 2001, inventor Jean-Claude Gallier, working for Aventis Pharma S.A. of France patented a method of isomerizing paspalic acid into lysergic acid using tetraalkylammonium hydroxide (organic-soluble bases) as the catalyst. The application number is 2001002395, 11 Jan 2001, which corresponds to Patent WO0102395 (http://l2.espacenet.com/dips/viewer?PN=WO0102395&CY=gb&LG=en&DB=EPD)
, FR2795728, EP1196415.
This reaction wasn't available through Beilstein Crossfire, and If it weren't for the sort of idea exchange here at the Hive, I wouldn't have felt any need to question what I had been taught in school a few years ago. I would have never pursued this idea any further without people here inquiring as they did. This was wonderful.
Ever since I learned that paspalic acid could be isomerized into lysergic acid, I've been reexamining what I've thought about ergot chemistry. In particular, I've been thinking about hydergine and other 9,10-dihydrolysergic acid derivatives.
Hydrogenated lysergic acid derivatives are nowhere near as closely regulated as the non-hydrogenated ones. The reason for this is that 9,10-dihydrolysergamides are nowhere near as active or interesting.
But...
What if dehydrogenation really is possible? Sure, H2 isn't going to split off by itself spontaneously, but maybe under the right conditions there is a reaction, or a series of reactions, that would allow this to happen.
I know that DDQ has been used to dehydrogenate benzocyclohexane systems to afford a double bond that is conjugated with the benzene ring, but would it work here? This sort of reaction has been used rather extensively in recent terpenoid syntheses, but how will that pyrrole ring affect it?
Here's a general example of this reaction:
Molecule: (https://www.the-hive.ws/forum/faq.pl?Cat=#applet)
dehydrogenation1 ("[H][C@]24CC[C@]1(C)[C@@H]([R])CC[C@@]1([H])[C@]2([H])CCc3cc(O[R])ccc34>>[H][C@@]23CCc1cc(O[R])ccc1C2=CC[C@]4(C)[C@@H]([R])CC[C@@]34[H]")
Here are some articles related to this:
Tetrahedron Letters, (2002), 41(11), 1729-1731.
Comptes Rendus de lÏAcademie des Sciences, Serie IIc: Chimie, (2001), 4(3), 201-205.
J. Chem. Soc., Perkin Trans.1 (1955), (22), 2813-15
Steroids, (1995), 60(12), 809-11.
Recl. Trav. Chim. Pays-Bas, (1993), 112(12), 627-34.
Synlett, (1992), (10), 821-2.
Patent US4882319 (http://l2.espacenet.com/dips/viewer?PN=US4882319&CY=gb&LG=en&DB=EPD)
Helv. Chim. Acta, (1989), 72(4), 725-30.
Another route might be to take it in two steps: allowing dihydrolysergamides to react with LDA and I2 should add an iodide beta to the carbonyl. If this were followed up with an E2 elimination reaction, you'd have successfully dehydrogenated your starting material.
I don't know. I'm throwing out guesses right now. Does anybody else have another approach to this problem?