(Somewhat bad) Synthesis of Piperonal

I reckon piperonal (also commonly referred to as heliotropine) is a compound most of you will probably be familiar with. It is most commonly known for being the precursor to MDA-class drugs, particularly the N-methyl analogue MDMA, on the more legal side it and its derivatives also find lots of use as flavoring ingredients and perfumes though.


Fig.1: Structures of piperonal (left), MDA (middle) and MDMA (right)

There are multiple feasible approaches of making it as the amateur, with ozonolysis (or the Lemieux von-Rufloff reaction) of either piperic acid or isosafrol being most commonly employed from what I can see.


Fig.2: Oxidative cleavage of piperic acid to piperonal by Lemieux von-Rudloff conditions

This has previously been documented on YouTube by both Chemplayer and Chemdelic, so if you want and overview on that method I recommend checking out their videos regarding that topic [1][2].
This approach unfortunately has one major downside though, of both potential alkene precursors being relatively hard or really expensive to get on scales that compensate for the quite unsatisfactory yield of the following cleavage.

An alternative approach would be the methenylation of protocatechulic aldehyde by the Williamson Ether synthesis, which itself can be made from either vanillin by the action of aluminium chloride and pyridine [3] or from bourbonal by the action of concentrated sulfuric acid, second of those methods generally seeming the most accessible, being previously documented in another thread [4].


Fig.3: Synthesis of protocatechulic aldehyde by demethylation of vanillin (left) or deethylation of bourbonal (right)

For the methenylation itself the use of diiodomethane would be ideal, but because of its significantly higher accessibility and lower molecular weight DCM shall be employed. Following [5] on the example of unsubstituted methylenedioxybenzene the use of sodium hydroxide as the base and DMSO as the solvent may be reasonably successful.


Fig.4: Proposed synthesis of piperonal by methenylation of protocatechulic aldehyde

To get things started, inside an 100 ml RBF, 9,12 g of sodium hydroxide (0,23 mol) were dissolved in 10 ml of dest. water and cooled back down to room temperature. This was followed by 20 ml of DMSO and left to stirr for a couple of minutes, during which time things turned a uniform slushy consistency (probably due to NaDMSYL precipitating). A Dimroth-condenser was attached and the flask was warmed to 95-100°C with the help of a heating block, resulting in a clear colorless solution.


Fig.5/6: Slushy of presumably NaDMSYL (left) and eventual clear solution of the base (right)

To this 13,8 g of commercial protocatechulic aldehyde (0,1 mol) were added as a powder, instantly causing the mixture to discolor and eventually take on an almost black color. Heating was continued for 30 more minutes, during which time a separate reflux apparatus containing 12 ml of DCM (0,19 mol) in 30 ml DMSO, heated to 70°C with the help of a glycerol bath, was set up.


Fig.7/8: Strongly discolored solution of the protocatechulic aldehyde (left) and the apparatus used for the preparation (right)

The hot aldehyde solution was slowly added to the refluxing DCM solution in 4-6 ml portions each with the help of a pipette over the course of 45 minutes, and the reaction flask washed out with 5 more ml of DMSO. On each addition vigorous boiling of the DCM could be perceived, which could only be contained by the use of ice water in the condenser.
The temperature of the glycerol bath was risen to 100-105°C, and heating was continued for approximately 3 hours. To isolate the piperonal a steam destillation of the black mixture was performed until the destillate coming over was fully transparent.


Fig.9/10: Final reaction mixture after combining the two solutions (left) and performing the steam destillation (right)

In total almost 100 ml of destillate were collected, with lots of product freezing in the condenser. Things were extracted three times with DCM, the pooled extracts were washed with 2 molar sodium hydroxide solution (this took up all the slight yellow color there was) and saturated brine, and dried over sodium sulfate. The solvent was rotovaped off, and a vacuum of 0,3 mbar was pulled on the product to remove any dimethylsulfide present from decomposition of DMSO. This resulted in a small quantity of crystalline off-white solid (0,97 g; 6,46%) with a melting point of around 30-35°C with a clear residual dimethylsulfide smell.


Fig.11/12: Rotovaping off the DCM (left) and the final crystalline piperonal (right)

Obviously this outcome is far from satisfactory, and there is definitely room for improvement. I‘m not even fully sure where things went wrong this badly, so definitely let me know your suggestions. I‘m personally under the suspicion that there may still be product in the reaction mix that didn’t steam destill over quickly enough, so I‘m probably gonna try further steam destilling the black solution on the weekend.

Besides the yield being absolutely miserable, the presence of an dimethylsulfide contaminant makes the product basically useless for my originally intended purpose of perfumery, so rather then trying to fully purify the already small quantity potentially losing most of it on the way, I decided to perform a Henry reaction with Nitromethane according to Shulgin’s procedure for BOH so I at least end up with some nice colorful crystals [6].


Fig.13: Synthesis of 3,4-(methylenedioxy)-beta-nitrostyrene by Henry condensation of piperonal and nitromethane

In addition to the piperonal made in the previous procedure a small quantity of an old commercial sample was added to add up to 1,5 g of starting material (10 mmol). This was dissolved in 5 ml of glacial acetic acid by swirling the flask and set in a heating block on a magnetic stirrer. To this 292,2 ul of ethylene diamine (4,37 mmol) was added with the help of an eppendorf pipette, followed by 1 ml of nitromethane (18,46 mmol), instantly causing the colorless mixture to take on a yellow color.


Fig.14/15: Solution of the piperonal in GAA (left) and the reaction mixture after addition of the nitromethane (right)

Stirring was turned on and things were heated to 100°C for 45 minutes, during which time some yellow needles of product started precipitating. The suspension was cooled in the fridge, and the solid was removed by vacuum filtration and washed with a little dest. water.


Fig.16/17: First precipitation of product (left) and the reaction mixture after cooling in the fridge (right)

For further purification the product was recrystallized from 1-propanol, resulting in some gorgeous thin yellow needles of 3,4-(methylenedioxy)-beta-nitrostyrene (1,34 g; 69,44%).


Fig.18/19: The air dried product (left) and its crystals under the microscope (right)


Sources:
[1] https://youtu.be/FjVVX_uASms?si=n0AG51hO9kl3eL6p
[2] https://youtu.be/thWbD1FO2iw?si=n5C4GA6eqfMpwTRx
[3] https://patents.google.com/patent/CN102241575B/en
[4] https://www.sciencemadness.org/whisper/viewthread.php?tid=15...
[5] https://www.erowid.org/archive/rhodium/chemistry/methylenati...
[6] Shulgin A., Shulgin A. (1991). PiHKAL. BOH; ß-METHOXY-3,4-METHYLENEDIOXYPHENYLETHYLAMINE. p.496

Quote: Originally posted by jackchem2001

DCM is said to act poorly as an alkylating agent due to an anomeric-type effect of its most favorable conformation. Perhaps some catalytic iodide in the reaction would help.

Thanks Boffis!
I’ve thought about that as well, tho I never came to a conclusion that fully satisfied me. Most of the really good OTC formylation approaches (particularly Duff reaction) are unfortunately ortho-directing, and the alternative of doing a two-step approach via the Blanc-reaction followed by Kornblum-oxidation (something similar should be documented on Erowid iirc), comes with the significant downside of forming highly carcinogenic BCME as a byproduct in the first step..



Additionally I‘m really not sure about the accessibility of catechol for the average amateur. I know it is occasionally found in oxygen absorption packs, but I feel like that’s a really expensive way of getting a reagent..

Like sure, I could personally use my catechol bought at Sigma and formylate that with my TCT / DMF purchased from TCI, but that feels like defeating the purpose of publishing it here.

Quote: Originally posted by clearly_not_atara

Glyoxylic acid is an interesting possibility. It is an extra step.

Quote: Originally posted by Random

Synthesis of 4-hydroxybenzaldehyde is still how I could say, subject to discussion. ... I say that because there were various attempts at synthesizing it and also according to my personal memory it is not that easy to separate it from other isomer. Is isomer how we call that. [Edited on 6-10-2024 by Random]

Quote: Originally posted by Niklas

On a different point, something that just came to my head. I now think the main issue with the piperonal synthesis probably is the use of NaDMSYL and hydroxide as the base, which may cause the aldehyde to disproportionate in a Cannizzaro like fashion. So by using much weaker potassium carbonate instead, which shouldn’t impact the reactivity all that much considering the acidic nature of the phenol hydrogens, this may result in a significant improvement of yield.

Quote: Originally posted by clearly_not_atara

The difficulty with K2CO3 is finding an appropriate solvent. NMP is a better solvent for potassium carbonate than are DMF or DMSO, and this is what the researchers were using. IIRC, K3PO4 has a better solubility in more common solvents. Beitia et al describe its use in DMF at 100 C, in DMSO at 80 C, and in glycol ether solvents.

getting back your piperonal

Quote: Originally posted by Niklas

Somehow didn’t came across this paper, definitely some useful information right there. It got me thinking about something I remember seeing in PHiKAL a lot.. What do you think of possibly using acetone as the solvent with some kind of quaternary ammonium phase-transfer catalyst added? I think I still have some hexadecyltrimethylammonium bromide around I could test this with. I‘ve never seen it being utilized this way before, but maybe benzalkonium chloride could act as an more generally accessible alternative as well. Not that getting them is too difficult for me, I just really don’t like working with the common solvents you mentioned because of their difficulty of fully being removed from the product. Not that it’s impossible, but it’s definitely a pain compared to a high volatility solvent like acetone. (Admittedly, this may just be me being biased to some extent because of my previous work with polymers, but anyway xd)

You can make Benzodioxole from catechol and Ch2Cl2.
Then react this with glyoxylic acid to get 3,4-methylenedioxymandelic acid.
Then to piperonal using O2 and CuCl2 or using AgNo3.

See attached paper



Attachment: 1488-Bjorsvik.Liguori.Minisci.Synthesis.Vanillin.Isovanillin.Heliotropin160c.pdf (183kB)
This file has been downloaded 44 times