Starting Materials
Ergotamine tartrate. Obviously if you've got access to this en masse, then it's safe to conclude some force in the universe wants you to make some acid. For everyone else, plant materials are the most appropriate precursor.
Ergot fungus. Many people claim that one could feasibly grow ergot from culture and harvest, but as far as I know, the strains of ergot which will actually produce ergotamine in these conditions are in a distinct minority. Uncle Fester recommends planting a field of rye and infecting it with ergot. Sounds complicated and drastically ups the investment cost for production. Lysergic acid amides including ergotamine can be extracted from the ergot to serve as starting materials.
Another ergot route would be growing some Claviceps paspali, which will produce paspalic acid in culture. Paspalic acid can be isomerized to lysergic acid with tertbutylammonium hydroxide (or other tertalkylammonium hydroxides), or more crudely with sodium or potassium hydroxide. The big problems with this route is a scarcity in the literature on optimal growth conditions to encourage paspalic acid production. Also I don't even know where you begin to go about obtaining the C. paspali in the first place.
MGS/HBWR The perennial favorite (oops I made a pun), these seeds are commonly available, but if you're buying enough to make LSD, then you might recieve an undesired visit from Johnny Law, so again the only practical method for large-scale production would be growing the flowers and harvesting seeds. Unfortunately, both types of seeds contain significant portions of clavine alkaloids which do not readily seperate from the wanted lysergic acid amides. Consequently, seeds are not a feasible starting material if you're running a synthesis starting from a crude extract of lysergic acid amides, however the clavines can be more readily removed if the lysergic acid amides are processed into lysergic acid and purified.
Synethetic Routes
Once you have your lysergic acid amide or amides, there seem to be two main routes you can follow:
Proceed directly from the amide to LSD: Treat the amide or amides with anhydrous hydrazine to form the hydrazide, followed by actylacetone to forn lysergic acid pyrazole, which forms LSD on treatment with diethylamine.
Proceed to LSD through Lysergic Acid: Most synths proceed through or start from lysergic acid. Lysergic acid is easily produced from lysergic acid amides by treatment with hydroxide, and isolated as the monohydrate. Lysergic acid monohydrate is unsuitable for water-sensitive reactions, but in general it seems easiest to choose a synthetic route that is not sensitive to this water of hydration. If absolutely necessary, the waters of hydration can be removed under 2 mm Hg at 140 degreees C (Uncle Fester doesn't recommend this for reasons he doesn't explain, but probably has something to do with the instability of the lysergic acid).
From lysergic acid, there are a variety of methods available:
- Procede from lysergic acid through an azide intermediate (Hoffman's original method, similar to the above method starting from lysergic acid amide)
- Form the lithium salt of lysergic acid (via KOH), treat with two molar equivalents of sulfuric anhydride (SO3), followed by an excess of diethylamine.
- Procede via trifluoroacetic acid. Another old method, not very popular as far as I can tell.
- Procede with diethylamine via the acid chloride intermediate using POCl3. This is the procedure used by Shulgin in the TiHKAL recipe, and seems to be the best option here.
Final Workup
Whatever procedure was used in the synthesis, the final product generally ends up with a certain portion of iso-LSD. The LSD and iso-LSD are seperated by column chromatography. The LSD is then converted to the tartrate salt. The iso-LSD is racemized with KOH and run through the chromatography column; the process is repeated a couple of times, and the collected LSD fractions also converted to the tartrate salt.
Well, that sums up the main routes that I found. Shulgin also mentions a method using a peptide condensation agent N,N'-carbonyldiimidazole, but this doesn't seem like a generally practical choice... If I've missed a practical route to LSD that someone else can spot, I'd certainly be interested.