Does anyone have access to SPARC? They seem to have a very nice modeling software for Liquid-Liquid-Extraction (LLE).

Can anyone run logD simulations as seen here: http://www.archemcalc.com/video/logD.htm ?

Specific request (4 combos):

1) Substance psilocybin
2) Solvents DCM, Xylene, Hexane, limonene
3) Salt ammonium sulfate, salt

At the pI (~4), What is the expected distribution coefficient logD? Are there better solvent options out there that give a larger logD? Repeat calculations for psilocin at its isoelectric point (~9). As a bonus check Harmalol too.

Also for:

1) Substance psilocybin
2) Solvent acetone
3) Salt ammonium sulfate

What does the logD vs pH plot look like? Does the psilocybin tend to stay in the water layer overall?

Below are notes from using the free Marvin software. Unfortunately, one cannot change the solvent (standard octanol is used) nor the salt (NaCl) and the salt concentration does not go above 0.25M. The free Marvin software does not plot logD vs pH, but the plot can be generated by writing down the numbers it gives as the salt concentration is changed in the logP calculation under the "tools" menu.

Since the fee Marvin software is limited, any SPARC calculations would be appreciated. Thanks!

Ok, here is an update using the free demo of Marvin for some of our favorite zwitterions. If anyone has access to SPARC would be very interested in that simulation using different solvents and salts, especially for psilocin (thank you).

Limitations in Marvin:

1) Must use octanol as a solvent. Wish we could change it to xylene/toluene/DCM/IPA or other more practical solvents for the home alchemist.
2) Limited to ionic strength of 0.25 molar. Also limited to NaCl/KCl as the ion provider. In practice one can get to ~6 molar for NaCl (35% solution) and ~11 molar for ammonium sulfate (~70% solution). That means we have a very limited view of the x-axis.

With that being said, this is what Marvin gives for the log D (y-axis) at various ionic strengths in moles/liter (x-axis). The data is simulated at the iso-electric pH (pI) where the molecule is neutral and more likely to move to the solvent (for non zwitterios like DMT there is no pI point, since the molecule is simply neutral in alkaline solutions). Note that log D is simply the log of the ratio of the solvent concentration to water concentration. For example if log D = 1, then there is 10 times more of the product in the solvent than in the water (theoretically). If log D = 0, then the concentrations are the same. if log D = 2, then there is 100 times more concentration in the octanol.

We are limited to ocatnol as the solvent, but these plots may give an idea of how easily the zwitterions will leave the water and how the ionic strength affects them.

Notes:

1) Psilocybin (first image), heavily affected by ionic strength. However, it seems to not want to leave the water and log D never gets to 1 in our ionic strength scale. In the second plot Psilocybin is removed to take a closer look at the scale of interest for other zwitterions.
2) Psiclocin, affected by ionic strength. Could be pushed out when salting the water. This may be good news for the FASA approach with xylene proposed by Orion in 2013.
3) Bufotenin, slightly affected by ionic strength. Effects of salting may be limited at high molarity compared to psilocin.
4) Harmalol, hardly affected by ionic strength. Brute-force salting may not work, a suitable solvent may be needed (Fisher used chloroform in the early 20th century to extract it from rue).

General note: For zwitterions it could be that cholorform and DCM are good solvents to pull out of water. This post is mainly looking at the salting out potential of different zwitterions. Tentative conclusion is that psilocin can be helped out of the water at high ionic strengths (assuming the trend continues beyond 0.25 M).

Focusing on magic mushroom zwitterion alkaloids in this post.

The phosphorylated zwitterions psilocybin, baeocystin, and norbaeocystin don't seem very suitable for LLE (Liquid-Liquid-Extraction). Marvin gives at 0.25 ion molarity log D = 0.49, -.04, and -0.95. Calculated pI for these alkaloids is 4.09, 5.88, and 5.86 respectively.

The corresponding dephosphorilated zwitterions seem more suitable for LLE at their isoelectric point. The plot below shows log D vs Ion (M) for the dephosphorylated forms of psilocybin, baeocystin, and norbaeocystin (called psilocin, "baeoctin", and "norbaeoctin" respectively). Conclusion from these Marvin calculations is that at a pH ~ 9.4 magic mushroom dephosphorilated zwitterions seem like potential LLE candidates at high ionic strength (since lod D > 0.5). Well, at least in theory...

4/24 Update: Nomenclature changed to baeocin and norbaeocin per downwardsfromzero excellent suggestion.

Great that you've been doing the homework!


Just a tangent on nomenclature (that being my thing somewhat): I've been using "baeocin" and "norbaeocin" as my preferred trivial names for the dephosphorylated compounds for a number of years, in analogy with the psilocybin/psilocin pairing. I see your logic too but feel the "-oct-" phonic component misleadingly hints towards eight-carbon compounds. The "-cin" ending is also very marginally quicker to type so join me in the wa'er if you like and drop your t's

...

It will be most splendid when this idea is tested out - pulling psilocin into limonene (for example) could really be a thing.

Interesting, too, that "baeocin" shows the steepest slope on the graph. Does this relate to the internal hydrogen bonding mentioned in the other thread as well, I wonder?

Great suggestion on the notation. Previous post updated.

Yes, I think we can explain the increased slope of logD of baeocin (vs psilocin) by the same mechanism. This time there is no methyl group to get in the way, so the internal interaction you proposed may be easier, so the salt can "squeeze" the molecule together now. Unfortunately, since the methyl is replaced by a hydrogen, there is more water solubility at the beginning, so that is why log D is so much lower at zero ionic strength.

The slope of psilocybin is dramatic because maybe the phosphate group is large and a little bit of torque moves it to close the "ring-like" interaction you mentioned?.

If we are right baeocysting should an even larger slope than psilocybin (dangling phosphate plus no methyl group to get in the way of the "ring-like" interaction).

I think it all makes sense and fits your interesting theory at this time. Will also run the molecules you mentioned in the other thread through the lod D calculation. All this also explains harmalol's poor slope since it looks a lot "stiffer" (which in practice means adding salt won't help harmalol LLE extraction as much as other zwitterions in theory).

I can definitely try limonene instead of xylene and then do a FASI. On my list to try, thanks for the suggestion

Thanks for helping to keep home chemistry fun and interesting!

I went ahead and added Limonene among several tests that are going on (test tube #4). It took longer than with xylelene, but a good amount of precipitate formed on the limonene pull done at pH 9.3 (near the isoelectric point of psiclocin) after FASI. Based on all the tests we have pulling at this pH is giving the most promising analytical results with a possible psiclocin peak. Also, over time (1 month +) a dark purple/brown layer forms at the top of the precipitate which could be another promising sign.

So a "PILIE" Tek may work for mushroom extraction one day. Will keep on working on it to see. Looks like it would be a "food safe" process too compared to PIXIE (but the name is not as cool ).

Thanks for the experimental suggestion