Sciencemadness Discussion Board

Synthesis of 3,4,5-trimethoxy cinnamic acid using green Knoevenagel condensation

DrDevice - 26-12-2022 at 02:33

A while ago I reported on the conversion of sinapinic acid to its methyl ester and subsequent o-methylation to methyl 3,4,5-trimethoxycinnamic acid.

I have now approached this reaction from another direction, using the Knoevenagel condensation to form 3,4,5-trimethoxycinnamic acid from 3,4,5-trimethoxybenzaldehyde and malonic acid.

Knoevenagel.gif - 9kB

The work is based on that presented in "The green Knoevenagel condensation: solvent-free condensation of benzaldehydes" (Schijndel, Canalle, Molendijk, Meuldijk, doi:10.1080/17518253.2017.1391881). That group used syringaldehyde, I just used the same methodology on 3,4,5-trimethoxybenzaldehyde.

The big difference between the approach I'll describe here and the conventional Knoevenagel is that it uses benign inorganic ammonia salts and (mostly) solid phase reactions, compared to the use of pyridine and piperidine, making it much more accessible to the amateur community.

There were several different ammonia salts used, and some organic amines as well. I worked through several combinations with varying degrees of success, but even with the worst outcome, I never achieved less than 40% conversion, and best result was 73%.
I'll present the approach used for the best outcome, then summarize notable aspects of some of the other experiments.
4.06g (20.7mmol) of 3,4,5-trimethoxybenzaldehyde (MW=196.2) was combined with 1.2 equiv (2.58g) malonic acid (MW=104.06) and 0.4 equiv (0.66g) ammonium bicarbonate (MW=79.056) was added. These solids were placed in a large (150mm x 25mm) test tube and 10ml EtOAc added.

The test tube was then suspended in an oil bath at 140C.

In a short while gas began to evolve and the EtOAc also began to boil. The reaction became so vigourous it almost reached the top of the test tube, but settled after a few minutes to a lower rate of gas generation. The EtOAc boiled off completely as the reaction proceeded.

The reaction mass became steadily more and more viscous, and took on a golden yellow colour.

After 2 hours there was no more gas evolved, there were however several "trapped" bubbles in the now semi-solid reaction mass.

After allowing the product to cool, saturated sodium bicarbonate solution was added to dissolve the solid mass. This took some time to complete as the solid mass only dissolved at the relatively small area exposed to the solution.
(In previous attempts, at this stage, I had tried to remove the colouration by "washing" the aqueous solution with EtOAc, but was mostly unsuccessful).

6M HCl was then added to bring the solution to a pH of 2. An off-white precipitate soon formed. The precipitate was filtered and washed with water.
The precipitate was recrystalized using a 4:1 water:EtOH mixture, resulting in off-white/yellowish crystals.

These crystals were dissolved again in sat. sodium bicarb solution and the resulting solution was washed several times with small quantities of EtOAc, until both the aqueous and organic phases were transparent.

The process of precipitating with 6M HCl and recrystaling from water:EtOH were repeated, leading to a mass of small needles of pure white crystals, which were dried in air.

Obtained 3.61g of 3,4,5-trimethoxycinnamic acid, MW = 238.24, thus 15.15mmmol or 73%.

MP (meas) 124 - 125C, MP (lit) 126 - 128C
TLC with 5:1:1 toluene:acetic acid:ethanol gave a single spot at Rf = 0.57. The benzaldehyde Rf was 0.74 in comparison.
Outcomes from other attempts presented below. The ratio of benzaldehyde:malonic acd:ammonia salts was 1:1.2:0.4, unless otherwise noted. Only a few grams of benzaldehyde were used in each attempt.

I: Attempted evaporation of EtOAc under reduced presure before reacting. Lots of spattered product around the wall of the vessel. Yield = 59%
II: Described in main body.
III: Ran at 90C for 3 hours. Much yellower result, but still obtained 57% yield after workup.
IV: Ran at 140C for 1 hour. Yield 58%
V: Used ammonium sulphate rather than ammonium bicarbonate. Gas reaction slower, and small amount of tarry residue formed. Yield 49%
VI: Used ammonium chloride, and dissolved it in water then added to reaction mass. The product was dark orange/brown after 2 hours, and required tedious washes with EtOAc after the first recrystalization. Yield 40%
VII: Ammonium chloride, with minimal EtOAc. Product brown and tarry after 2 hours, but achieved 47% after workup.
VIII: used ethanolamine rather than inorganic ammonia salt. Approx equimolar quantity of ethanolamine to benzaldehyde. Interestingly, this reaction was exothermic. Although the colour was the same light yellow, yield was only 41%
IX: A slightly bigger scale, and performed in an evaporating dish on a (temperature controlled) hot plate rather than in a test tube in an oil bath. Achieved 59%.
I can't really provide insight as to what lead to the best yield, but I'll speculate that the refluxing EtOAc and confined volume provided better mixing as the reaction continued.

Next step will be the esterification - that should be straightforward.

Fery - 1-1-2023 at 08:51

Well done, nice yield, good purity!
IX: a lot of benzaldehyde oxidized with air??? Maybe using closed apparatus with smaller death volume (less air present) could help?
Do you plane to make its methyl ester? Using Dean-Stark trap apparatus?
I did cinnamic acid from bare benzaldehyde with acetic anhydride and anhydrous sodium acetate as a catalyst. I got quite poor yield compared to you.

AvBaeyer - 1-1-2023 at 17:53


A very nice piece of experimentation on one of my favorite reactions.


DraconicAcid - 3-1-2023 at 14:15


I did a very similar reaction with p-anisaldehyde, using ethanolamine as the base. I haven't measured yield or mp yet (I'm lazy), but is seems to have worked nicely. Using bipy, on the other hand, gave me the non-eliminated version (anisylidenemalonic acid?), I think.

DrDevice - 3-1-2023 at 17:11

Thanks everyone for the comments.

Fery, I hadn't considered oxidation of the benzaldehyde. I might repeat the experiment with a nitrogen purge then a "rubber glove" finger septum with a tiny hole in it to see if I get a different yield.

Regarding the esterification, I also want to reduce the double bond, so I have two approaches - reduction, then esterification, or esterification, then reduction.

I have done some preliminary "sketches" of a reduction experiment with Co2B/NaBH4, (refer "A Co2B Mediated NaBH4 Reduction Protocol Applicable to a Selection of Functional Groups in Organic Synthesis", doi 10.1002/ejoc.201800440) which resulted in the desired 3,4,5-trimethoxyphenylpropionic acid, but at very poor recovered yield.

I will be going the ester first route next.

Fery - 3-1-2023 at 21:44

Hi DrDevice, when using open vessel there is air exchange and a some benzaldehyde oxidizes. If you use 75 ml tube for the reaction (150 mm length, 25 mm diameter), there could be dead volume of 50 ml of air = 10 ml O2 = 10 mg O2 = 0,3 mmol. Preventing air entering the apparatus is good idea - as you wrote - rubber glove finger septum with a tiny hole in it. But I think vapor of boiling ethyl acetate will destroy the rubber. Also during the reaction the vapor of ethyl acetate reached the top or your tube so expelled the air, most of reactants reacted and there is not too much benzaldehyde to oxidize when the reaction slows down and when all ethyl acetate is evaporated. There is CO2 evolved which also expelled most of air. These are just my thoughts. Anyway the best yield 73 % you achieved is very nice!
Concerning the reduction of double bond in the article doi 10.1002/ejoc.201800440, they tried also H2 with cat. Pd(OH)2/C and methanol solvent, the catalyst was sensitive to poisoning and required high purity of reactants. I have good experience with H2, cat. Pd/C and methanol solvent, but the starting compound was something else

[Edited on 4-1-2023 by Fery]

DraconicAcid - 4-1-2023 at 11:41

This may be a noob question, but is there a version of this reaction that would work with ketones? Benzophenone --> beta-phenylcinnamic acid?

clearly_not_atara - 4-1-2023 at 20:24

I don't think that malonic acid will "fit", sterically, but the Reformatsky reaction should be a go.

DraconicAcid - 5-1-2023 at 00:09

That's a good suggestion. I suppose a Wittig reaction would also work. Thanks.

DrDevice - 25-1-2023 at 04:45

To briefly follow up, I tried a few more reactions with a nitrogen flush and septum. There was no real difference in yield. I also tried a higher catalyst load, but the resulting product was heavily discoloured and yield was lower.

Regarding the reaction with ketones, the Kurti/Czako book "Strategic Applications of named reactions in organic synthesis" states that ketones react much more slowly than aldehydes - but I found no actual reference.

I have also synthesized the methyl ester; report to follow soon.