I am trying to understand the fundamentals, otherwise it will just be try & error and with an approach like this the only thing I need is luck. I am pleased about everyone who is helping to understand the process.
So far we have a lot of theories and great ideas on this board from forward thinkers. But none of those ideas really matched my experiences (with pTSA, phosphoric acid and ZnCl2)
I started to dive into this theory and found this paper. This is another context, but they labeled d8 as thermodynamic and d9 as kinetic product, too. (Synthetic pathways to tetrahydrocannabinol (THC): an overview - Organic & Biomolecular Chemistry (RSC Publishing) DOI:10.1039/D0OB00464B)
For example, phosphoric acid. Our observations can be explained with this model.
If you pour phosphoric acid into boiling solvent with CBD. Within a few minutes you will reach your maximum of D9 (kinetic product, which forms the fastest). Since your system has energy in excess, your pathway 1 is reversible and therefore your D9 will start to decrease over time. This is not like a gear box where you shift from gear 1 into gear 2. This happens simultaneity. That explains why your timing is so important on regular reflux setups when you aim for high d9. At the same time it is nearly impossible to reach very high d9 numbers in this setup.
Rogue is playing around with pressure and made some amazing yields with it. Why? This can be explained with this model, too. He is raising the boiling point and therefore increasing the energy on this setup. This means: pathway 1 is for sure reversible, but it also increases the speed of the kinetic path. This way you are able to convert the biggest part of CBD into D9, before D8 is formed in big amounts.
My TLC plate above is from a 8 bar pressure run, too.
@moronnabis Thanks for pointing out stereoselectivity! This was not on my radar. So this is the explanation, why our results vary when switching solvents and/or catalysts. It has an influence on the difference between the activation energies for Δ8 and Δ9. So you have solvent/catalyst mix A that has a bigger ΔEa than solvent/catalyst mix B. I do believe that all Δ8/Δ9 catalyst can be either turned into high Δ8 or high Δ9, based on setup conditions. I can confirm that you are able to produce high Δ9 AND high Δ8 with pTSA/toluene, but high Δ9 is much more challenging for me. For sure there are catalyst that perform better or worse and ND Δ9/Δ8 won’t be achievable with many.
Does this maybe also relate to the formation of isomers?
Compare the results of pTSA with different solvents at RT in this paper.
Hexane is the only solvent, which led to the formation of Δ8-iso-THC.
So there seem to be other pathways, which get influenced by the choice of solvent/catalyst.