Sciencemadness Discussion Board


Maja - 7-6-2006 at 04:48


3-Methoxy-4,5-dihydroxybenzaldehyde (5-Hydroxyvanillin) [9] 5-bromovanillin (200 g, 0.91 mol), sodium hydroxide (245 g, 6.1 mol) and copper powder (1 g, 0.016 mol) were slurried into 3 L water. The reaction mixture was heated at reflux for 24-27 hours. Disodium hydrogen phosphate (4.5 g, 0.032 mol) is added for the last half hour of reflux. The reaction is then cooled to less than 50°C, filtered to remove a precipitate of cupric hydrogen phosphate and acidified with hydrochloric acid (460 g). The reaction mixture was placed in a continuous extractor and extracted with ethyl acetate (3 L). The ethyl acetate extract was stirred with activated carbon and filtered. The filtrate was washed with saturated aqueous EDTA solution followed by salt. The solution was then dried over magnesium sulfate and filtered. The ethyl acetate solution was concentrated to a crude solid. The crude product was dissolved in boiling toluene (2 L), treated with activated carbon, filtered and cooled to crystallize. The product, 3-methoxy-4,5-dihydroxybenzaldehyde was isolated in approximately 60% yield (86 g) with a mp 132-133°C (lit. Mp 132-134°C).

This is quote from rhodium. I want to ask a few questions about that synth of Hydroxyvanillin. First of all for what purpose Inert atmosphere is required(what would happen if inert atmosphere wouldn't applied ?). Second, Can nitrogen be replaced with CO2 ? And is there any other route from 5-bromovanillin to 5-hydroxyvanillin without so long extraction and reflux time ? Thanks for answers.
Oh, and one more Q. Can I just apply vacuum to refluxing mixture to remove all air ? That would work ?

wa gwan - 7-6-2006 at 12:38

Originally posted by Maja
This is quote from rhodium. I want to ask a few questions about that synth of Hydroxyvanillin. First of all for what purpose Inert atmosphere is required(what would happen if inert atmosphere wouldn't applied ?). Second, Can nitrogen be replaced with CO2 ? And is there any other route from 5-bromovanillin to 5-hydroxyvanillin without so long extraction and reflux time ? Thanks for answers.
Oh, and one more Q. Can I just apply vacuum to refluxing mixture to remove all air ? That would work ?

To prevent oxidation of the aldehyde to acid.
Use iodovanillin (see and you won't need to worry about using inert gases. It doesn't oxidize.
You'll have to take my word for it on that because I don't have the article with me but it has been discussed before.

garage chemist - 7-6-2006 at 12:51

CO2 and sodium hydroxide... tell me, is H2O or H2CO3 the stronger acid?

But I'd also like to know if there is a less complicated method to convert bromovanillin to hydroxyvanillin.

When trying to make syringaldehyde from bromovanillin, sodium methoxide, CuBr and EtOAc, there was obtained a rainbow colored putty-like mass from which only black gunk could be isolated (though the black gunk smelled very nice). Syringaldehyde from bromovanillin isn't a nice synthesis.

The hydroxyvanillin workup (extraction with ridiculous amounts of solvent) cannot possibly be the only method for this.
Doesn't the hydroxyvanillin crystallize upon cooling after acidification? Maybe after removing some of the water by distillation?

Sandmeyer - 7-6-2006 at 13:17

There is not need for inert atmosphere.

Ullmann - 7-6-2006 at 13:59

Well, IMHO better to do a methoxylation than an hydroxylation... it is easily done and workup is as simple as acidify, extract, wash. I do not think this long soxhlet is necessary, EtOAc should extract well the hydroxyvanillin... i thought it was because of some copper metal retaining product but there is only 1g copper hence it doesnt mazke sence. The recrystallisation in toluene isnt as easy as it seem, larger volume of toluene are needed i fear... there is another way to the hydroxyvanillin that is reaction of hexamine with vanillin to make the isophtaldehyde and then dakin the isophtaldehyde with H2O2, then extract with EtOAc without soxhlet, then purify...

But realy go for methoxylation, find a way to make 3M methoxyde solution (anhydrous), use copper I salt, use an amide cosolvent, concentrate at reflux and it would be a breeze... Syringaldehyde is much more versatile than hydroxyvanillin and less troublesome to make.

inert atmosphere : not needed, the para-OH group protect the aldehyde from oxidation, it has nothing to do with inert atmosphere. CO2 is not needed neither.
Long reaction : maybe it has to do with the very little amount of copper or maybe they overdid it... A vacuum on a refluxing mixture will diminish the temperature of the reflux, it is not useful here.

Iodovanillin is good precursor but IMHO bromovanillin is beter as it doesnt lead to as much dehalogenated byproduct and also it is cheaper and easy to make and reactive enough for the ullmann.

PS: do not forget to check the ullmann refs in wanted refs...

wa gwan - 10-6-2006 at 13:26

Sandmeyer, I thought so at the time but wasn't sure. Bromovanillin is also resistant to oxidation so no inert necessary, as you say. I assume it was used in anticipation of oxidation?

Iodovanillin is good precursor but IMHO bromovanillin is beter as it doesnt lead to as much dehalogenated byproduct and also it is cheaper and easy to make and reactive enough for the ullmann.

Its cheap if the iodine is recovered (a must!)

garage chemist - 21-7-2006 at 01:46

A question to the workup of the hydroxyvanillin synthesis: couldn't one, after acidification of the reaction mix, evaporate all the water in vacuum?
Then you'd have a mix of NaBr, NaCl, Cu and your Hydroxyvanillin.
The hydroxyvanillin could then nicely be extracted with ethyl acetate by stirring the residue with it and then filtering warm.
That method of workup sounds much better than extraction of the aqueous solution.
Let's see what mp the resulting hydroxyvanillin has, that will tell if this method does work.

Maja - 21-7-2006 at 08:30

Uemura from Hive did the same and extracted with DCM. He said that DCM is perfect solvent for 5-Hydroxyvanillin in soxhlet because 5-HV has low solubility in it and DCM does not dissolve any other salts. 5-HV precipitates from DCM very easily and forms crystalline crusts...

garage chemist - 21-7-2006 at 12:51

OK, did the hydrolysis of bromovanillin today (very smallscale, with just 2,5g bromovanillin). The reaction mix was very deep green after the refluxing time was over (no inert gas used).
After acidification, the mix turned dark brown and something precipitated. The mix then started to congeal in a strange way, and as ethyl acetate was added and shaken, it emulsified and refused to separate.
So I distilled away the ethyl acetate and then most of the water under coarse vacuum at 50°C. A dark brown smeary sludge remained. I added 50ml ethyl acetate to extract the product, and shook the flask long and hard. There was no apparent change, the ethyl acetate just gained some color (but is still clear).
Tomorrow I'm going to separate the solvent and evaporate it.
I'm going to recrystallize the residue from toluene.

I don't have the equipment to perform a continuous extraction of a liquid (just a soxhlet, which can only be used for extraction of solids).

Did Uemura also encounter the congealing of the acidified reaction mix, and did his batch also become so brown and sludge- like?
Did he use inert gas?

Maja - 21-7-2006 at 13:13

Actually, He gots just gray/white solids ... IRC ,but he ALSO wrote that extracts with Ethyl acetate ended in stubborn emulsions...

Sergei_Eisenstein - 21-7-2006 at 15:33

If the gods are not on your side and you want to make an attempt to get the best out of your emulsion, you may try to filter it over a pad of celite. It can break the emulsion and result in something you can work with.

Sandmeyer - 21-7-2006 at 23:33

If possible use ether next time, generallyh I have never encountered emulsion with it but do with DCM and EtOAc. And trick Dr. Sergei provided is nice indeed...

[Edited on 22-7-2006 by Sandmeyer]

Maja - 10-6-2009 at 05:27


Vanillin (3.04 g., 20 mmole) was dissolved in 1N sodium hydroxide solution (20 ml), and warmed to 80° C. Then a solution of NaI 3 .NaI (2N) in water (10.1 ml) plus 20% aqueous H 2 SO 4 (0.5 ml) was added dropwise over 30 minutes, and the mixture stirred an additional 30 minutes. Sodium hydroxide (7.6 ml of 50% solution), and copper dust (128 mg=10 mole %) were then added and the mixture heated at reflux overnight. The solution was cooled, filtered to recover catalyst, neutralized with 20% aqueous H 2 SO 4 , and extracted thoroughly with chloroform. The organic base was dried (Na 2 SO 4 ) and stripped to yield 3.1 g (99%) of organic material consisting of 75% 5-hydroxyvanillin and 25% vanillin.

The aqueous phase was concentrated under vacuum, and treated with the theoretical quantity of chlorine as a water solution. The purple iodine crystals were removed by filtration. As 87% recover of iodine was achieved."

Look, they don't use continious extractions, but they "thoroughly" extract with chloroform.... Maybe someone also tried this ?

ctrlphreak - 12-7-2009 at 04:52

I actually have a couple of grams of theoretical 5-iodovanillin.

It was created by NaOCl + KI + Vanillin.

Here was the procedure:


To a 500ml 3-Neck RBF, there was added, 5g of Vanillin, 6.2g of KI, and 100ml of MeOH. After

a bit of stirring, the vanillin had disolved, and the MeOH had a slight yellow tinge to it,

but the vast majority of the KI did not disolve.

An addition funnel was attached to the 500ml RBF, and was charged with 65ml of 6% household

Bleach. This was allowed to drip in at a rate of about 1-2 drops per 3 seconds. No cooling

was used. It was slightly exothermic, but tolerable to touch. Barely any heat at all

actually. The addition took about 20 minutes.

After the addition, the contents of the flask were almostcompletely in solution, and the

solution had taken on a very light coffee with cream coloring, the contents were allowed to

react still, without cooling, with occassional stirring for about 30min-1hour (was stoned,

so don't know.)

The solution, once at room temperature, was filled with about 50-75ml of tap water, and an

addition of about 1.5g of Sodium ThioSulfate was added to reduce the iodide present. (II'm

aware it's probably, not right, it says a 10%-150ml solution, but that's all I had on


The color is a slightly lighter coffee with cream color, and some percipitate can be seen

before addition of the HCl.

After completion of the HCl addition, the solution has seperated into two apparent 'layers'.

The 'upper half' seemingly being water based, is a VERY deep, almost unable to be seen

through red, and the bottom half is completely covered in a very tan/orange percipitate. The

upper portion was removed by decantation, and set aside for iodine recovery.

The lower portion was isolated, dried, and collected. It was a definate tan colored

substance that weighed approx. 3g.

The entire 3g was disolved in a minimal ammount of alcohol, and hot water, heated up to a

gentle boil, and then thrown into the freezer for about 12 hours.

When examined again, the solution had 'separated' again into two layers, the top part being

a pale yellow/orange/red obviously from free iodine, and the lower half of course was the

whole shebang of, alleged, 5-IodoVanillin.

The bottom part was poured into a filter in a funnel, and the flask flushed with about 50ml

of water to solute, and get all the rest in the flask. This water was added to the

mushy-liquid already in the filter, and allowed to filter.

After filtering, the mass at in the filter was a MUCH paler, slightly cream coloured

substance. It has not been weighed yet as it has not been completely dried.

I'm going to throw the shebang, once it's done drying into a flask, and rig it up for reflux, and attempt a hydrolysis. I also think that copper salts are useable as stated from the patent above, and am intending on linking another document I found that says CuSO4 is sufficient as well. I will be using CuSO4.

5-Hydroxyvanillin success

Scr0t - 16-1-2012 at 10:22

I've recently been playing with this reaction and thought I would share some of my results and observations.
With the stated conditions you can't reduce the reaction time much and yields are in the vicinity of 60% as reported in the literature.

Preparation of 5-Hydroxyvanillin
A solution of 24.5g NaOH and 300ml dH2O in a 500ml flask was prepared.

20.0g 5-Bromovanillin and 0.3g Cu powder (wet, freshly precipitated from aq. CuSO4/l-Ascorbic acid*) were added forming a yellow turbid suspension with some white precipitate.

The mixture was set for reflux and heated while magnetically stirred. As the mixture warmed it was opaque with a very pale green-white color, as the mixture became warmer it became progressively thicker with white solids that halted stirring. As the mixture became hot the white solids broke up and dissolved forming a clear orange-brown solution, all white solids had dissolved after ~10mins of reflux.
By the 3hr point the reaction had become such a deep green it appeared black.

Reflux and stirring was maintained for a total of 27hrs. An inert atmosphere is unnecessary.

Once finished the flask was removed from the heat and allowed to cool causing a grey/green solid to precipitate.

The mixture was acidified with 52ml 36% HCl whereupon it became brown-black with no precipitate visible, it was then filtered to remove the remnants of copper (some of this passed through 595 grade paper) which was black.
The filtrate was extracted with 3x100ml portions of EtOAc that took most of the color into the organic layer (phase interface of first extract hard to see). The solvent was distilled off using magnetic stirring to prevent bumping.

200ml Toluene was added to the black residue and brought to boiling, decanted from black crystalline solids and allowed to crystallize forming brown platelets. The toluene was filtered and returned to the flask with the black solids, heated to boiling, decanted and cooled to yield another crop of brown crystals.
Yield 8.7g of shiny brown flakes with a mp 132-134°C (5-Hydroxyvanillin, yield 60%) with a vanilla-like odor and less dense than Bromovanillin.

Recrystallization from 300ml Toluene failed to remove the coloration, activated carbon would probably help here although the melting point was good.

Filtering once acidified helps prevent the emulsion with EtOAc.

A run that involved a short reflux time (9hrs) led to a different extraction experience; much light brown solids precipitated on acidification, much of the black color taken up into the Toluene and low yield (<30%) of an impure product of purple color.

A run that was 24hrs long was successful however it did produce a small amount of precipitate on acidification which caused emulsion problems during extraction (not filtered after acidification). This run was filtered while still hot and basic to give a reflective purple copper coating on the filter paper. This reaction gave a product with a lighter brown color than the 27hr run with a 54% yield (extraction procedure was less efficient).

Glassware was easily cleaned with dilute aq. NaOH.

*Preparation of the copper powder.
10g CuSO4·5H2O was dissolved in 100ml H2O and heated to boiling while magnetically stirred. 5g l-Ascorbic acid was added once boiling. The solution instantly became dark green and rapidly developed a cloudy brown appearance. The mixture was filtered and rinsed a couple of times with dH2O. The material was used immediately. A yield of ~1g Cu powder (~40%) was obtained.
The powder had a low ohmic resistance but lost its conductivity after a few hours of standing in air.

turd - 16-1-2012 at 23:17

Well done! Great first post. :) I will try that one day.

Haber - 25-4-2014 at 03:58

Im not sure if im correct but i would assume alot of the folks who wish to synthesize 5-hydroxyvanillin is going after the trimethoxy substituted aldehyde.
If so why not just purchase the commercially available syringaldehyde instead? To cut down the number of steps necessary.

Or am i missing something obvious?

And sorry for the bump, I wasnt sure if it was okey to bump threads when you dont actually have anything substantial to add on the topic.

Scr0t - 25-4-2014 at 09:42

Because it's more fun!

plus 3,4,5-trimethoxybenzaldehyde is commercially available too but such purchases may draw unwanted attention from the 'authorities' or maybe it isn't available to somebody due to their location.

Anyway 5-hydroxyvanillin can be methylenated to give myristicinaldehyde while syringaldehyde cannot.

[Edited on 25-4-2014 by Scr0t]

Methyl.Magic - 2-5-2014 at 04:52


Ive done myrticinaldehyde. Bromovaniline was transformed to 5-hydroxyvaniline with CuSO4 and recrystallised in toluene. Yield = 40% yield of tan slightly purple powder. A lot of toluene is needed to recrystallise it. recrystallisation is needed absolutly. Synthesis from cu powder didnt work. Cu +1 or +2 salt are much better. I prefer CuSO4 because of its good solubility. CuOH2 is the best according to a publication i've seen.

Myrticinaldehyde have been made by reactiong 5-HV with dibromomethane in DMSO at 70°C for 24h. Yield = ~65% after reacrystallisation. this latter was not easy due to low solubility of myrt. in any alcohol. A lot of solvent have been use for a few grams.

The best thing to do for hydroxylating vanillin is use a ligand for the copper. a lot of different ligand have been used. The best ones are : 1,10-Phenanthroline, 1,3-di-t-butylpropan-1,3-dione , 2-2'-bipyridine, glycolic acid and 8-OH-quinoline and 8-OH-quinaldine, the latter two being the best ! Other ligand such as EDA, L-proline, picolines or salens have poor yields.

For the base CsOH is usually used but KOH or NaOH works also good, a little less but there are much lest costly... The best base is apparently tetrabutylammonium hydroxide but it can be more difficult to get than simply alkali hydroxides.

Solvent is usually DMSO/H2O in proportion 3:1 or 2:1.

temperature are around 100-120°C

Yield are perfect for all iodo compounds. Also perfect for bromo compounds containing deactivating groups in ortho/para such as nitro. Reaction time are usually less than 24 h. A few catechol syntesis have been reported but the yield are less than the other compounds this is probably due to the electron donating hydroxy group in the ortho position.

I guess Iodovanilin have no problem to be hydroxylated but bromovaniline can be more harsh to react but I bet that longer reaction time and high temperature and a little more catalyst amount can give decent yields.

Hydroxylation of bromovaniline :


- Catalyst CuOH2>CuCl2 = CuSO4 = CuO = CuI >> Cu powder, 10-20 mol%

- Ligand : 8-HO-quinaldine > 8-HO-quinoline >> 1,10-ohenanthroline > glycolic acid > 1,3-diketone > 2-2' bipyridine > 1,2-diamine 20-40 mol%

- Solvent : DMSO in H2O : 50-80%

- Temperature : 100-120 °C (refluxing)

- Time : 36-48 h

- Work up : adding the mixture to Ice/water, acidification to pH <4 if amine are present, filtration, recrystallisation in toluene

BETTER : **2 IN 1** :

At the end of the reaction : add 1.5 eq of methylene bromide/iodide or 3 eq. of methylene chloride or 3 eq of MeI/Me3PO4 or any suitable alkylating agent, keep reacting at good temperature depending of the alkylating reagent for 24h. Add into ice/water, turn it to high pH to get rid of the unreacted bromovaniline, 5-hydroxyvaniline, monoalkylated 5-HV or 5-bromovaniline, filtrate, wash with acidic water to get rid of the ligand, wash with clean water to pH around 7, dry and recrystallise in toluene. DONE !

Scr0t - 2-5-2014 at 11:12


I've noticed several others who've had problems using Cu powder catalyst. However the results obtained when using freshly prepared Cu from CuSO4/ascorbic acid ("nano-copper") gave consistent yields at ~60% 5HV.

When I first attempted purification of myristicinaldehyde I tried steam distillation which was a mistake. It took several gallons to steam distill only ~15g taking me about 3days, also steam distilling a solid of mp 131°C unsurprisingly tends to clog the condenser.
Another irritating thing was that the crude aldehyde tended to form small clumps in the distilling flask slowing the distillation rate of the aldehyde.

zed - 2-5-2014 at 14:37

ScrOt, this is very interesting to me. Tell me, you have manufactured Myristicinaldehyde, but have you tasted it? Rare it is, in nature, and not often synthesized. I don't have any on hand, and in order to obtain some, I'd probable have to make it. Which I might be willing to do, if its flavor merits it. Heliotropine (aka piperonal) is sometimes used in ice-creams as a premium vanilla-like flavor. Is this aldehyde also especially tasty? Vanilla like, or not?

[Edited on 2-5-2014 by zed]

Scr0t - 3-5-2014 at 10:22

No, unfortunately I've not tasted it however I can tell you it was disappointingly odorless.

chemrox - 23-5-2014 at 15:03

I think piperonal is a little more like root beer than doctor is always haranguing me about tasting my products.

zed - 29-5-2014 at 22:01

Way I remember it, Heliotropine isn't anything like root beer. The aroma itself is intoxicating. I had a few kilos, back in the day. Can't recall what I did with them. Do you suppose I might have thrown them away?

Anyway, the current buzz is, that Heliotropine is used as a flavoring in some premium vanilla ice creams.

I'm a simple man. I have ordinary aspirations. Is it so wrong that I would like to discover, patent, manufacture and market... the next big flavor breakthrough?

Chocolate, Vanilla, Almond, Cinnamon, and Zedelicious!

Crowfjord - 21-7-2015 at 11:47

I found a few relevant references that may prove useful for the amateur toward synthesis of 5-hydroxyvanillin from 5-bromo or iodo vanillin.

First off, I think that the paper that Methyl.Magic refers to above is Organic Letters 2011, 13(16) 4340-4343. That one is all fine and well, but the authors use ligands that aren't really available to most amateur experimenters.

I found a couple relatively recent papers that report high yields of phenols from the corresponding halobenzene. In JOC 2013, 78, 5804-5809, the authors report high yields of phenols in 6 hours using 5 mol% Cu(OH)2 in 1:1 DMSO:water with 0.3 equivalents of glycolic acid as the ligand.

Chem. Commun., 2011, 47, 6692-6694 reports high yields in the standard 24 hour time frame using Cu(OAc)2 with D-glucose as the ligand, also in 1:1 DMSO:water. This one has the added benefit of a soluble catalyst, so that it is easily removed in the workup.

Though neither report mentioned 5-halovanillin as one of their substrates, it stands to reason that since the reaction did not work for any of the authors without ligand (as has been done with some success with bromovanillin), the reaction using a ligand of some sort should likely show some improvement here as well.

If I get time, I might try one or both variations sometime this summer. Glucose is easily available from the brew shop, and I have some lactic acid that I could try in the place of glycolic acid.

I have posted these papers in the references section, in the thread "Recent journal articles of interest."

[Edited on 21-7-2015 by Crowfjord]

[Edited on 22-7-2015 by Crowfjord]