Synthesis of 3-(p-Hydroxyphenyl)-2-Butanone "raspberry ketone"

Description of GB2416770

Raspberry ketone (p-hydroxyphenyl-2-butanone) is a key flavour molecule with typical raspberry flavour characteristics and a low odour threshold. Raspberry ketone is one of the most expensive flavour components used in the food industry. Up to $20,000/kg may be paid for the natural compound.

Raspberry ketone can be found in raspberries and other fruits (such as peaches, grapes, apples and various berries), vegetables (e.g. rhubarb) and in the bark of tree (e.g. yew, maple and pine).

Raspberry ketone can be used in the aroma formulation of, for instance, strawberry, kiwi, cherry and other berries. However none of these fruits are used to obtain the raspberry ketone as the low content of raspberry ketone in these fruits makes the extraction and purification process unprofitable.

Raspberry ketone can be produced chemically via the condensation of p hydroxybenzaldehyde with acetone.

However, the chemical synthesis of compounds can often result in environmentally unfriendly production processes and in undesirable racemic mixture of the compound of interest (Vandamme and Soetaert 2002; J Chem Techno Biotechnol 77:1323-1332).

In raspberries, the synthesis of raspberry ketone is one part of the phenylpropanod pathway. This pathway has been described by Borejszaysocki and Hrazdina (1994). In the first step, coumaryl CoA (which Is present In many plant tissues) is condensed with one malonyl CoA into benzalacetone (p-hydroxyphenylbut 3-enc-2-one). The enzyme catalysing this step is called benzalacetone synthase (BAS). In the second step, the double bond in benzalacetone is reduced, resulting in raspberry ketone (p-hydroxyphenyl-2-butanone). The enzyme catalysing this step is called benzalacetone reductase (BAR), this enzyme requires the presence of NADPH. Benzalacetone synthase (BAS), EC 2.3.1.-., is a member of the polyketide synthase family.

Benzalacetone reductase condenses one acetone unit from malonyl CoA with one p-coumaric acid to form benzalacetone. Chalcone synthase (CHS), EC 2.3.1.74, is another member of the polyketide synthase family, which condenses three acetate units from malonyl CoA with one p-coumaric acid to form chalcone. Stilbene synthase (STS), EC 2.3.1.146, is another member of the polyketide synthase family, which condenses three acetate units from malonyl CoA with one p-coumaric acid to form stilbene (Zheng et al 2001). The polyketide synthase family is described in detail in Schroder 1999 (Comprehensive natural products chemistry vol 1: polyketides and other secondary metabolites includingfatty acids and their derivatives [U. Sankawa Ed] pp 749-771).

CHS is part of another part of the phenylpropanod pathway, which converts phenylalanine into naringenin chalcone and its derivatives (Weisshaar and Jenkins 1998; Hwang et al 2003). This part of the phenylpropanoid pathway involves the following enzymes: phenylalanne ammonia Iyase (PAL), cinnamate-4-hydroxylase (C4H), 4-coumarate:coenzyme A ligase (4CL) and chalcone synthase (CHS).

As a first step phenylalanine Is deamnated to yield cinnamic acid by the action of PAL. Cinnamic acid is hydroxylatcd by C4H to 4-coumaric acid. 4-coumaric acid is activated to 4-coumaryl-coenzymeA (CoA) by the action of 4CL. CHS catalyses the stepwise condensation of three acetate units from malonylCoA with 4-coumaryl CoA to yield nanngenin chalcone (which is a precursor for some flavonoids). Naringenin chalcone is converted to naringenin by chalcone isomerase (CHI) or naringemn chalcone spontaneously converts to naringenn.

Evidence for the relation between BAS and CHS is provided by BorejszaWysocki and Hrazdna (1994 and 1996). Borejsza-Wysocki and Hrazdina (1994) show that the timing of BAS activity parallels CHS activity. In addition, Borejsza-Wysock and Hrazdina (1996) show that CHS and BAS co-purify, and seem to react with the same antisera. However, CHS and BAS activity in the same enzyme preparation showed a different response to treatments with, for instance, 2-mercaptoethanol and ethylene glycol. Also, CHS activity and BAS activity showed different induction patterns upon treatment of raspberry cell cultures with yeast extract, suggesting that these enzymes are not one and the same molecule.

Soluble enzymes catalysing the reduction of a double bond using NADPH, for example benzalacetone

reductase (BAR), are classified by the International Union of Biochemistry and Molecular Biology as belonging to enzymatic class EC 1.3.1.X. For instance, an enzyme annotated as EC 1.3.1. 11 from Arthrobacter sp. was reported to remove a double bond from coumarate (Levi and Weinstein,

1964). However, no gene has been identified in connection to this enzymatic activity. Other enzymes In the enzymatic class EC 1.3.1.X are orotate reductase, 2-hexadecenal reductase,

cholestenone 5 alpha-reductase etc. for which genes are known. However, none of these enzymes was reported to have benzalacetone reductase (BAR) activity. It is known from literature that 4-hydroxybenzalacetone can be transformed to raspberry ketone by fungi or yeasts such as Pichia, Saccharomyces, Beauveria, Kloeckera, Aureobasidium, Cladosporium Geotrichum, Mucor and Candida spp. (Fuganti and Zucchi, 1998). However, no gene has been identified in connection to this enzymatic activity. Also, to our knowledge, no such activity has been reported for bacteria, neither gram negative nor gram positive.

Attempts to biosynthesise raspberry ketone have been described. Hugueny et al (1995, Boflavour 95

pp 269-273) teach a biotechnological method for producing raspberry ketone. This method comprises culturing a microorganism which has a secondary alcoholdehydrogenase (ADH), such as Candida boidinii, and adding the precursor betulosde to the culture medium. In this cellular environment the secondary ADH dehydrogenatates betuligenol into raspberry ketone.

Various studies have been carried out to determine the role of polyketide synthase genes in the production of flavonoids using E cold transformed with a raspberry CHS gene (see Zheng et al 2001; Kumar and Ellis 2003), a Glycyrrhiza echinata CHS gene (Hwang et al, 2003) or a Arabidopsis thaliana CHS gene (Watts et al 2004). However these studies teach the in vitro or in vivo synthesis of naringenin.

These studies do not teach the in vivo synthesis of benzalacetone or raspberry ketone.

Abe et al (2001) teach the cloning of rhubarb BAS, expression of the gene in E colt, purification of the recombinant BAS protein and the in vitro synthesis of benzalacetone. However this study does not teach the in vivo synthesis of benzalacetone or raspberry ketone.

SUMMARY OF THE INVENTION

According to a first aspect of the present invention, we provide a host cell comprising a chalcone synthase (CHS) polypeptide sequence and a 4coumarate:CoA 1igase (4CL) sequence in which one or both of the CHS polypeptide sequence and the 4CL sequence is heterologous to thc host cell.

Preferably, the CHS polypeptide sequence is derived from one of the following: raspberry, petunia, grape, Medicago sativa, Arabidopsis thaliana, Antirrhinum majus, Zea mais and Petroselinum crispum, preferably a raspberry CHS sequence. Preferably, the CHS sequence is selected from the group consisting of: accession number AF292367, accession number AF400567, accession number X04080, accession number X76892, accession number L02902, accession number AF112086, accession number X0371O, accession number X60204, accession number V01538, a sequence shown in SEQ ID NO: 2, and a sequence having at least 75% sequence homology thereto.

Preferably, the 4CL sequence is a tobacco 4CL sequence.

Preferably, the 4CL sequence is a sequence having an accession number U50846 or a sequence shown in SEQ ID NO: 3.

Preferably, the host cell is transformed with an expression vector encoding the chalcone synthase (CHS) polypeptide sequence and an expression vector encoding the 4-coumarate:CoA ligase (4CL) sequence, or an expression vector encoding both sequences.

Preferably, the host cell is a microbial host cell selected from the group consisting of F,scherichia spp, Saccharomyces spp, Pichia spp, Beauveria spp, Candida spp, Aspergillus spp, Bacillus spp, Pseudomonas spp, Hansenula spp, Klayveromyces spp, Schizosaccharomyces spp, Streptomyces spp, Lactococcus spp, Lactobacillus spp, Pediococcus spp, Kloeckera spp, Aureobasidium spp, and Streptococcus spp, preferably an E. colt, preferably strain BL21, or a Saccharomyces cerevisiae, preferably strain YPH 499, or a Bacillus subtilis host cell.

Preferably, a polypeptide expressed from the CHS polypeptide sequence has benzal acetone synthase (B AS) acti vity.

Preferably, the host cell is capable of producing benzalacetone when supplied with a precursor of benzalacetone, preferably p-coumaric acid or a source of p coumarc acid.

Preferably, the host cell has benzalacetone reductase (BAR) activity, preferably inherent benzalacetone reductase (BAR) activity.

Preferably, the host cell further comprises a benzalacetone reductase (BAR) sequence, preferably a heterologous BAR sequence, preferably shown as SEQ ID NO: 5.

Preferably, the host cell is capable of producing raspberry ketone when supplied with a precursor of raspberry ketone, preferably benzalacetone or a source of benzalacetone. Preferably, the host cell is capable of producing raspberry ketone when supplied with a precursor of raspberry ketone, preferably p-coumaric acid or a source of p-coumaric acid.

Preferably, the host cell further comprises a cinnamate-4-hydroxylase (C4H) sequence.

Preferably, the host cell Is capable of producing benzalacetone or raspberry ketone, or both, when supplied with cinnamic acid or a source of cinnamic acid.

Reference Information

Reference Information

Thanks for all thoses articles!
I've just realized that there seems to be some confusion in the litterature on the exact nature of the so-called "raspberry ketone": in two of my org.chem books, and in some articles on the net, it is said to be the branched p-hydroxyphenylbutanone, the phenyl ring being attached to the third carbon of the chain, alpha to the carbonyl. See here: Smell Database
But in all the articles/patents you guys have kindly provided, it is said to be the straight-chained isomer.

To my eyes, the linear isomer 4-(para-hydroxyphenyl)butanone seems easier to synth, especially through the p-hydroxybenzaldehyde pathways.

A crossed aldol with acetone is easy and well documented, basic conditions would be more effective here I guess, the benzalacetone should be easy to isolate and purify, and I guess a CTH could reduce the double bond, or NaBH4 even if the carbonyl gets reduced, it could easily be re-oxydized.
US4414417 deals with a CTH using Rhodium and Iridium catalysts that yield the saturated ketone; this could be applied to other catalyst/donors.

The FC option seems equally interesting, but not possessing any b-hydroxybutanone, it doesn't seem readily obtained from MEK or other accesible reagents.

p-hydroxybenzaldehyde could be obtained through a Riemmer-Tiemann as mentionned in the review Solo provided, I remember ready about performing the reaction in alcoholic medium to preveil the para-formylation.
I have ~30g of phenol left, so this could be enough on a small scale, though I want to give the o-formylation a try using Mg(OCH3)2/paraformaldehyde a try again. Maybe a bayer-villiger oxydation on benzaldehyde could give workable yields of phenol. I don't want to melt any benzenesulfonic acid with NaOH .

Well, I guess I'm going to keep busy for a while thanks for the help!

Reference Information

UnintentionalChaos, thanks for the advice, but I think I will be getting my vanillin from a flavoring supplier, which offers it in >99% purity for a reasonable price. But I might give it a try with the flavorings at the local supermarket, as the minium is 500g, and I wouldn't know what to do with so much vanillin, even if using it for baking cakes and stuff

As I only screened quickly through the article yesterday, I mixed up some entries, and the fact that the only way of obtaining good yields of pure phenylbutanone is to use a 0.1mol.L solution of Sodium Dodecylsulfate (SDS), which I don't currently have at hand, as a reagent anyway. Apparently, Tetrabutylammonium Bromide can't be a good substitute as it doesn't form micellar emuslions, which apparently "protect" the ketone from reduction by BH4-; it's the formed H2 which is the true reducing agent, catalysed by the formed metal boride. I could try using benzylalkylammonium chlorides, in C10-C12, which would form micellar emuslions and might act similarily. It's starnge as they have a few entries were using MeOH and H2O as solvants give 77% et 51% yield respectively, but latter state that:

"The preparation of pure product 3c [1-PhButanone] was not possible in either MeOH or water, and only use of aqueous solutions of SDS afforded it in good yield (entry 10). It seems that this surfactant ‘‘protects’’ the micellar-bound substrate
from attack by the reactive anion BH4 –, the main reduction
pathway being, therefore, the one carried out by H2."

I guess the product is contaminated by saturated alcohol, and that the 2 must be hard to seperate. A light oxydation could convert any over-reduced material back to the ketone, but could also make more impurities, and give a complex mixture of products. Let's try it one pot.
So if the benzalkonium (pool supply) doesn't work out nicely, I was thinking about generating before-hand the metal boride with stoechimetric amounts of borohydride and the chloride, in the appropriate solvant, and then adding the substrate and performing a atmospheric catalytic hydrogenation. Wouldn't it basicly proceed the same way as in-situ formed H2? I think it is molecular hydrogen that is evolved and not nascent, but don't take my word on this. Can anyone can help me out here? This way, no need of using expensive Rhodium catalyst.

EDIT:I've been kindly given ~30g of reagent-grade vanillin, so I will be able to perform the aldol condensation this weekend.
I've decided on trying a simple atmospheric catalytic hydrogenation with preformed CoB catlyst (from CoCl2 and stoechiometric NaBH4), as I've realized I've only got a couple grams of CoCl2 and no NiCl2... We will see how it turns out. I'm still undecided concerning the amount of catalyst to use... Any advice would be welcome.

[Edited on 10-1-2008 by Klute]

Synthesis of 1-(4-hydroxy-3-methoxyphenyl)but-1-en-3-one by basic crossed-aldol condensation of vanillin and acetone.

Finally started this very easy reaction. I just scaled up 20x the L. R. SMITHChem. Educ. 1(3) "short" synthesis, and plan on leaving it 48h at room temp.

Synthesis

5g (33 mmol) of reagent-grade vanillin [Picture 1] were weighed, and dissolved in 20mL (~280 mmol) of technical acetone, giving a totally limpid solution, in a 100mL capped bottle containing a stir bar [picture 2].
2g (50 mmol) of technical NaOH were dissolved in 20mL dH2O (10% w/v, 2.5 mol/L), and the warm solution was cooled under a stream of cold water. Once at room temp, the solution was slowly poored into the acetone solution with steady stirring. This immediately caused the solution to turn yellow and warm up somewhat [picture 3]. The color gradually darkened, becoming dark orange after 5min, and the bottle cooled down gently [picture 4]. This will be stirred at RT for 48h, then worked up as per the article: addition of 100mL of 3M HCl, filteration and washing of the resulting solids, recrystallization with EtOH/H2O, and caracterisation of the product.
This unsaturated ketone will then be hydrogenated over 10% Pd/C to the corresponding ketone, 1-(4-hydroxy-3-methoxyphenyl)butan-3-one, or Zingerone. Hydrogenation over preformed Co2B might also be tried.

Picture 1





Picture 2





Picture 3






Picture 4





Sorry if the pics are big, I didn't find how to just post a "icon" to the large pictures for display on another window. If this causes too much problems for uploading this thread, I'll diminish the size...

EDIT:
12h later, the mixture is now dark cherry red. I will post a picture tomorow as I don't have the time right now.

[Edited on 17-1-2008 by Klute]

Synthesis of 1-(4-hydroxy-3-methoxyphenyl)but-1-en-3-one - Continued


After 48h stirring at RT [Picture 1], the dark red/black mixture was transfered to a 250mL beaker, and 100mL 10% HCl poored in with magnetic stirring. A black oil crashes out, giving a black heterogenous mixture [Picture 2]. After 2min stirring, a fine yellow solid started precipitating [Picture 3]. Black droplets were still present at the bottom.
After another 5min stirring , the suspension was filtered, care taken of decanting the heavy, dark granules present at the bottom of the beaker. The fine suspension gave a greenish crystalline powder [Picture 4], which was washed with a little water, dried with suction on the buchner, and transfered to a small beaker.
The dark crystalline granules were equally washed, filtered and dried [Picture 5]. They did not agglomerate, and were very easy to collect in a seperate beaker.

The two crops were recrystallized with hot 10mL EtOH, followed by gradual addition of 5-8mL water, and left to cool. The first crop gave a clear, greenish solution, the second gave a dark brown, but limpid solution. The first crop crystallized very quickly. It was filtered, briefly washed with water, and dried on the buchner [Picture 6], giving beautifull canary-yellow crystalline flakes [Picture 7] .
The second crop crysatllized much less, not enough water had been added. The formed crystals were filtered, the filtrate replaced in the beaker, and the cake sparingly washed with water and dried. The crystals formed, although larger than the first crop, were of a more darker yellow [Picture 8].
The two crops of crystals were left to dry in the dessicator, and finally weighed exactly 2.53g each!

The final weight was 5.06g (26.35 mmol), representing a yield of 79.86%. Not bad.


Picture 1


Picture 2


Picture 3


Picture 4


Picture 5


Picture 6


Picture 7


Picture 8



Hydrogenation will hopefully be performed next weekend.

@Tinydantzler: Vanillin can also be done by formylating guiacol (2-methoxyphenol) - a cheap and readily available chemical. This can be done by the Reimer Tiemann reaction in the presence of cyclodextrin or tertiary amines (they change the regioselectivity of the reaction thus allowing for the para formylation). Other published methods include the condensation with glyoxalic acid in basic media following by the oxidation of the intermediate mandelic acid to vanillin (but these are two steps). One Japonese patent also describes the Vilsmeier-Haack formylation of guiacol using DMF/POCl3 at 120°C. This last is probably the most suitable for lab scale.

@Klute: Nice work.

Edit: Wrote glycolic acid instead of glyoxalic...

[Edited on 22/1/2008 by Nicodem]

Reference Information

LiALH4 is a bit of an overkill for such a reaction IMHO. If the reduction can be easily done with cheaper, safer reagents, such as NABH4 and H2/Catalysts, using expensive LiALH4 is a bit of waste. But of course that simply depends on your financial aids/work environement .
I'm sure there's lots more one can do with LiAlH4 that can't be done by other means. Oh well, just another article in another journal I guess.

BTW, the two recrysatllized crops had a sharp 127°C melting point (lit: 127–128.5 °C ) , so I think I can safely say it quite pur. The crystals have a lovely, light vanillin-like smell, a bit more fruity.

Klute,

Here is a Pd-CTH enantioselective reduction of a-methylcinnamic acid to the corresponding (S)-2-methyl-3-phenylpropanoic acid...

It proceeds at fairly low temps, seems remarkably simple (in fact seems awfully like, well apart from the time, the one I posted earlier - albeit it is CTH).

I'd dearly love to know how the l-tartaric acid promotes the claimed huge increase in ee's (98% ee or 99% of the (S))

PS Thanks Stoichemetric Steve for this site

I've FINALLY decided on trying to reduce the double bond on the unsaturated zingerone ketone.
I thought it would be a good occasion to start working with atmospheric catalytic hydrogenations. I decided on starting with aregular hydrogenation over recycled Pd/C, and see what gives, before trying out cheaper catalyst, so than I coul dcompare yields, hydrogen uptake rates, ratio ketone:alcohol, etc

I tried a setup using a chromatography column as H2 reservoir, but apparently it didn't work well as the pressure must diminsih too much after a little absorption to let the recation going.
I also did a stupid mistake, adding K2CO3 to a phenol w<hile I didn't want the phenoalte...

Synthesis of Zingerone 4-(4-hydroxy-3-methoxyphenyl)butan-2-one by atmospheric catalytic hydrogenation over Pd/C

Ref: Us patent N° 5,847,225





The following setup was assembled:
A 100mL schlenk was connected via a gas outlet adaptator and plastic tubing to a 350mL chromatography column mounted with a schlenk tap. The bottom of the column was immersed in a beaker containing 500mL water, the column's tap fully opened.
A hydrogen generator was assembled: a 250mL 4-neck RBF with a stir bar was mounted with a 70mL addition funnel, itself mounted with a gas outlet adaptator connected to some tubing.

The column:


The complete setup:


40mg of recycled Pd/C from a previous CTH reduction was charged into the schlenk, along with 10mL AcOEt.

The setup was then charged with hydrogen in the following way:

A very weak vacuum was connected to the schlenk's side arm. The schlenk tap was opened, aswell as the tap on top of the column. The water started rising into the column. Once it was near the top, the tap on top of the column was closed, followed by the schlenk's side arm. The vacuum tubing was disconnected, and the argon tubing tubing attached after purging the side arm a few times (attach the tubing, leave some pressure to build up, briefly discconect the tubing to evacuate the argon/air pressure, etc etc to chase any air from the side arm), then the schlenk tap was opened, followed by the column. The water level fell, and a flow of argon was left to bubble through the setup for a few minutes.

The addition funnel was charged with 20mL 37% H2SO4, and 0.5g of zinc dust was slurried into the RBF with a little dH2O.
The schenk and column were once again evacuated until water level was near the top of the column, acid dripped onto the zinc, and a flow of H2 was left to purge the generator for a min. The closed schlenk side arm was then purged with H2 as previously, the tap opened, and the column opened. The column was filled with H2, and a flow of H2 bubbled through the setup until evolution of H2 slowed down. The column was sealed, filled with H2, and stirring continued for 10min [Note1].

Catalyst during saturation:



The H2 tubing was disconnected, the schlenk briefly evacuated (column still closed), and flushed with argon [Note2]. The schlenk's top was disconnected, and under a flow of argon, 0.39g (2.82 mmol, 0.25eq) of K2CO3 were added [Note3] , followed by a solution of 2.15g (11.20mmol) of 4-(4-hydroxy-3-methoxyphenyl)but-3-en-2-one in 20mL AcOEt [Note4]. The erlen containing the yellow solution was rinsed with 5mL AcOEt.

The unsaturated ketone:


The obtained reaction medium:



the schlenk was closed, evacuated, back filled with argon, evacuated, back filled with H2 (adding more zinc dust to the H2 setup) twice, then the schlenk arm was closed, the column opened, and vigorous stirring started. H2 uptake was immediate, but slow. Slight warming was noticed.





After 30min, the reaction medium had turned to a suspension of guacamole-green solid along with the catalyst: the phenolate. Ihad completly forgotten there was a phenol group in the molecule, and that K2CO3 would form the phenolate... Even if this wasn't obviously a problem for the reduction, the solid could cover the catalyst enough to limit H2 uptake, considering the weak loading. So 0.3mL (~5.5mmol) GAA were added (opening the schlenk after performing all the operations described previously), along with a spatula (~1g) of dry AcONa. The green solid dissolved immediatly, leaving a faintly emeraude-green solution and a few orange carbonate particules.


The phenolate:


After adding the acid:



It was noticed that the water level never went above a certain point, even after recharging the column with H2, more absorption occured, but each time the level stopped at the same level (pretty low) [Note5].
It was decided to change the setup (not more than 30mL H2 absorbed over 3h) to a conventional inverted cylinder.

The flask was evacuated, aswell as the column, and back filled with argon. The column was disconnected, schlenk closed, and under a flow of argon the gas tap was attached on the top of the schlenk. A tubing was attached from the top of this tap, and placed in the cylinder, which was inverted under the surface of the water in the beaker. The setup was evacuated to rise empty the column from it's air, just under the tubing, and back filled with argon, doing this twice.

Setup with inverted cylinder:



More zinc was added to the H2 generator, and the setup purged with H2 before filling the column. The schlenk side arm was then closed, and vigorous stirring started. H2 uptake started immediatly, monitering the rate: 1mL/min.

After 20min, it seemed the H2 uptake has ceased. a White inorganic solid had precipitated, and seemed to cover the catalyst....

The inorganic precipitate on the catatalyst (very quick decantation):




The setup was opened (following the directions given previously, and 20mL MeOH containing 1 mL dH2O added. This completely dissolved the solids, but H2 uptake still didn't continue. A spatula tip of potassium formate was then added: there was a slight gas evolution (column closed, side arm opened) for a few minutes, that died off quickly. The column was then opened again after flushing the schlenkwith H2, and H2 uptake started immediatly, quickier than before (2mL/min), and calmed down to 1mL/min over 15min [Note6].

Inorganic dissolved:



The absorption rate then gradually diminished (1mL/2-3min), but didn't stop. The hydrogenation was stopped overnight (sealed column, stopped stirring), and restarted in the morning. The green color had lightened up considerably, and TLC (Eluant: 10:1 DCM:AcOEt) indicated a good amount of saturated ketone was formed, along with a little alcohol [Note7].

Reation medium in the morning:



125mL H2 had been absorbed since the cylinder setup was assembled. I considered that roughly 40mL had been absorbed with the column, so there isn't much left before reaction is finished, but I prefer following the reaction with TLC.

H2 absorption:



Each sample is taken after stopping the stirring and leaving the catalyst enough time to settle, and 0.2mL of the green solution are taken, placed in a vial, and acidified with a drop of GAA. The first sample taken (1H afetr beginning of hydrogenation with the column) stayed green afetr acidification, and only the unsat ketone was detectable by TLC. The sample taken 1h after switching to the inverted cylinder setup cleared up^entirely after acidification, and showed 3 spots (see Note 7). Further samples show that the unsat ketone spots is diminishing, the sat ketone getting larger, and the alcohol staying roughly the same.


TLC samples after acidification:



After a total of 24H hydrogenation, TLC indicated all the unsaturated ketone was consumed, and only a minimal amount of alcohol was present.
After leaving the catalyst decant, it was obvious the solution was completly clear.



The solution was vacuum filtered over 2 filters, the caatlyst washed with 2x5mL AcOEt and 5mL MeOH, and the clear filtrate transfered to a seperating funnel. 50mL of brine was added, and the funnel gently shaken. The very slightly yellow organic was seperated, and the aq. extarcted with 2x10mL AcOEt.




The combined organics were washed with 50mL brine, and dried 20min over Na2SO4. The solevnt was then removed under avcuum at 40°C, to afford a clear pale yellow oil with a very pleasant smell.



The oil was placed in the fridge, covered with a little pet ether, to try and induce cristn, without succes

The crude product will be purified by column chromatography, or by bisulfite adduct formation. Details to come when done.



Notes:

Note 1: The was done to fill the column with hydrogen (250mL required to complete reaction), and to pre-saturate the catalyst with H2.

Note 2: Each time the setup was opened, it was evacuated and flushed with argon to chase any H2, to avoid any fire hazard.

Note 3: the base was added to hinder reduction to the alcohol. Unfortunaly, an acetate should have been used to not form the phenolate.

Note 4: Slight warming was required to acheive complete dissolution.

Note 5: The pressure needs to diminish considerably for the water leevl to rise up more than this level I suppose. maybe at a certain pressure the equilibrium constant diminishes enough to stop the reaction. Prhaps the water leevl rising was simply du to leakes, although the joints were greased and double checked, I decided to follow conventional directions (inverted cylinder), as these were sure to work.

Note 6: apparently formate can reactivate Pd/C catalyst efficiently. this measn saturating Pd/C before a CTH is useless, but adding a little formate before a catalytic hydrogenation could substitue to conventional H2 saturation (no need to open the setup with H2-saturated catalyst: less fire hazard)

Note 7: a few eluants were tried and this variation afforded perfect seperation: from bottom to top: the alcohol as a thin rim, the unsaturated ketone (compared to pur sample), and then the saturated ketone. Rfs to come shortly.

Comments

Well, the selectivity of Pd/C is very encouraging, contamination with the amcohol is only minor by TLC, but nevertheless enough to prevent cristn. On the other hand, the recation is much too long. Is it the low loading? The fact that's it's a recycled catalyst? I haven't tried any used catalyst for other recations up to now, maybe the regeneration method isn't that efficient... I will see with another reaction.

I still plan on trying out other catalysts. I'm particularily interested by the Cu/SiO2, as it can be easily and cheaply made at home. On the other hand, the authors of the patent use fumed silica, which I don't have. I will surely try chromatography silica (60-200 mesh, acros) and home-made silica (less porous, but of varying quality).

I still have 2.15g of dehydrozingerone to play with, and enough vanilline to make some more (considering how easy it is..), but I would like to obtain some 4-hydroxybenzaldehyde. I definitively can't find some at a decent price, and am out of phenol. I'm afraid I will not be able to buy some until a little time,a s I've been spending too much lately and have other projects to attend to (TEMPO derivatives, etc).

If anyone find a relatively cheap source of 4-hydroxybenzaldheyde, please contact me! I would be delighted to obtain ~50gr at a reasonable price.. I'm sure we could work out a compassion

[Edited on 16-6-2008 by Klute]

Great contribution!
(Though I have to admit I have not read it all since I don't have much time at the moment)

Thank you for your comments, both of you.

Nicodem, thank dfor the advice regarding the bisulfite adduct, especially I would have though it woulo dhave reacted pretty well being a methyl ketone but... I guess I'll just do a quick column: I am one the rare chemists at work that actually LIKES doing columns

Alice, that synth looks pretty nice, but I don't have any dichromates and am not particularily found of using them. But the degradation of tyrosine with TCCA (IIRC?) that you proposed looks very interesting. If I can find some tyrosine locally, I wil give it a try.
At one moment I will have to get some more phenol, as I'm out of salicylaldehyde (had problems forming the N-ethyldiethylenetriamine ligand), so I will try the Riemmer-Tiemann with triethylamine..

Klute, I am on fucking drugs I swear - I meant to write p-methoxybenzaldheyde (anisaldehyde) not p-methoxyphenol

I also meant to say fucking nice job

PS Wouldn't the permanganate oxidation of anethole give the same result?

http://www.erowid.org/archive/rhodium/pdf/isosafrole2piperon...

http://designer-drugs.com/pte/12.162.180.114/dcd/chemistry/b...

The tyrosine variant uses Silver oxide as the oxidant which is going to make it too expensive.

PPS I also added something which Solo only just got today, the demethylation of substituted porphyrins with anilinium chloride, used in preference to the more usual pyridine.HCl, HBr, etc.

[Edited on 17-6-2008 by LSD25]

Attachment: Demethylation.AniliniumChloride.pdf (210kB)
This file has been downloaded 1454 times

After purification on a short column, using 50:50 AcOEtet Ether, 1.61g (8.30 mmol) of practically colorless zingerone was obtained, giving a 74.11% yield from the unsaturated ketone, and 59.18% from vanillin.

The oil is practically pur by TLC, only a very faint spot of the alcohol from the last fraction, but nevertheless is very hard to cristallize; it is very slowly starting to cristallize in the freezer, with 2 nucleation points, very slowly propagating (over several hours). They look like mold growing in the oil


The compound has a very pleasant, slightly pungeant though faint smell of ginger. Hurray!

Hopefully Cu/SiO2 will be more selective than Pd/C, and there will be no need of seperating the ketone from the alcohol, and it will possibly cristallize more readily.

EDIt: I forgot to add the Rf's:

Eluant: 10:1 DCM:AcOEt

Saturated ketone: 0.65
Unsaturated ketone: 0.56
Alcohol: 0.45

[Edited on 18-6-2008 by Klute]

An update:

After considering the Orgsyn prep Filemon indicated, and the notes in OrgSyn preparation of benzylacetophenone, I decided to try reducing the unsaturated ketones with Zn dust: the procedure seemed ridiculousy simple, and sued very easily accesable reagents. This would really make the method of choice for preparation of phenolic phenylbutanones...

I gathered a few old Ber. refs on the reduction of bezalacetone and benzalacetophenone to their saturated ketones, and with the unvaluable help of Woelen at translating german, managed to figure out the sparse details furnished... basicly, the zinc dust was added to the unsaturated ketone in GAA, the mixture heated to reflux for 3-4H, the solids filtered off, GAA diluted with water,a nd products extarcted with ether.

Now, the Zn reduction of a,b-unsaturated ketones is said to produce varying amounts of a dimer, and higher polymers, which are only slightly soluble in ether. This is with the non-phenolic unsaturated ketones...


I performed a few reductions of 4-(4'-hydroxyphenyl)but-3-en-2-one, here are my notes and some pics:



Preparation of 4-(4'-hydroxyphenyl)but-3-en-2-one



(no pictures, but basicly very similar to dehydrozingerone)

5,0 g (40.94 mmol) of p-hydroxybenzaldehyde are dissolved in 20mL (272.04 mmol) freshly distilled acetone inside a colorless 100mL bottle, giving a dark orange solution, with a very small amount of some fine amber solids in suspension.
2,0g (50 mmol) of NaOH are dissolved in 20mL dH2O (10% w/w; 2.5N), and the cooled solution is slowly added to the stirred aldehyde solution. At first, a beige voluminous solid is formed, but it quickly dissolves to give a very light yellow limpid solution, gradually darkening to a deep orange.
The solution turns dark orange/brown over 1H, and is left to stand for 24H, with occasional stirring. A dark red solid appears after two hours, giving a dense cristallin slurry.

The mixture was then added to 100mL of 10% HCl in a 250mL beaker, with good stirring. White fumes are evolved, and the dark orange/brown mixture warms up. the beaker is cooled in a ice water bath with vigorous stirring, and the oily material crystallizes as a dark brown solid after a couple of minutes. The suspension is stirred for another 10min at 5°C, then vacuum filtered over a glass frit. The brown mass is washed with a little amount of ice-cold water, giving a dark brown filtrate and light green cristillin solid. The solid is dried by suction for 20min, and dried in a CaCl2 dessicator over night.

the solids are added to a 100mL erlenmeyer, and covered with 25mL toluene. The toluene is heated to ~80°C on a hotplate, and AcOEt added dropwise until total dissolution of the solids (2-3mL required). An equal volume of toluene is added, and the dark green mixture is left to cool slowly. The next morning, beautifull yellow prisms are obtained. The crystals are vacuum filtered, washed with cold toluene:AcOEt 10:1, and pet ether, and dried by suction.

After 24h drying in a desicator, the crystals weighed 4.9g (30.21 mmol), 73.79%, and pur by TLC (single spot, eluant 3:1 AcOEtet ether).

By evaporating part of the mother liquors and adding pet ether, some more orange solids are obtained. They are recrysatllized by toluene/AcOEt to offer 1.13g (6.97 mmol) of TLC pur yellow crystals, bringing the yield to 90.81%.





Notes:
-The first few times I left the reaction proceed for 48H, but it seems the yields are even better with 24h reaction time as they are less impurites.
-Dissolving the crude solid in a solvent, and washing with cold brine (the unsaturated ketone is sparingly soluble in H2O) could be a more efficient approach.
-The recrystallization system is really adequate, I had trouble finding a good solvent system as the product kept on oiling out of solution with EtOH/H2O, AcOEt/Pet ether, etc. Having the crude product as clean as possible gives better results.
-the product has a very slight, pleasant smell, much lighter than dehydrozingerone (which gives 80-90% yields by following this procedure using 5.0g vanillin, all other reagents in same quantities)





Synthesis of 4-(4'-hydroxyphenyl)butan-2-one (Rheosmin)






1.62g (10.00 mmol) of the unsaturated ketone previously prepared were crushed to a fine powder and dissolved in 30mL GAA in a 100mL 3-neck 14/19 RBF, equipped with a CaCl2-guarded condenser, a thermometer and amgnetic stirring.








1.32g (20.00 mmol) of zinc dust were weighed out, and added in small portions to the stirred solution over 30min, at room temp. There was no exothermic reaction of noticeable bubbling, and the yellow solution took a greenish tint before enough zinc was added to give a light grey suspension.





20min after addition:




The mixture was then heated to 50-60°C in an oil bath, gentle H2 evolution started, and the green/yellow color disappeared in 10min.

40min after addition (leaving the zinc decant before taking picture):


1H:


2H:



after 2h reaction time, the flask was cooled, and the unreacted zinc was filtered on a glass frit. the zinc was thoroughly washed with 2x10mL GAA.






The slightly yellow, cloudy filtrate was then added to 150mL ice-cold water, in a beaker immersed in a ice bath. The mixture immediatly turned pink, and some ice was added.



Unfortunaly, a viscous semi-solid was obtained, but smelling very strongly of raspberries. Solid NaCl was added to precipitate even more of this viscous paste.





The paste was dissolved in a small volume of MeOH, NaCl filtered off, and solvent evaporated. Pet ether boils on the obtained viscous liquid didn't yeild any cristillin product. The slightly pink aqueous mixture was extarcted with AcOEt and the solvent evaporated afetr been washed with brine, togive a viscous, raspberry-smelling paste. Attempts to cristallize this with various solvents and conditiosn were all fruitless.. A column chroamtography could be done, but th eobjective is to find a easy reduction with acceable reagents giving a pur enough product..

TLC (3:1 AcOEtet ether) indicated saturated ketone (major stain), a little unsat ketone, and two other spots (dimers).

Comments:

I performed a first reduction using 4x zinc, but directly extracting the diluted AcOH solution. After washings, drying and removal of the solvent, as viscous dark purple goo was obtained, which smelled very strongly (and nicely) of raspberries. There again, it was impossible to obtain any cristilline material from this substance ( mp pur rheosmin= 80°C).

The pink colour is a very curious thing: when collecting samples during the reaction, the pipet used to withdraw the supernatant turned pinkish after a few minute sexposiiton to air. On the TLC plates, the saturated ketone stain turns dark pink whne exposed to UV! (and not before!) It gets darker and darker when exposed several times... Quinonic material? Adding trace amounts of reductant makes the colour disappear, and when in solution in AcOEt, it turns light red/orange....


I'm afraid the reaction conditions are not suited to phenolic unsaturated ketones, which must polymerise/dimerise much more easily than benzylacetones and acetophenones... But the material does smell very nice!

So I'm going to work on other reductions, Cu/SiO2 seems to be the most promising, as CoCl2/NaBH4 requires 5x weight of the unsat ketone in CoCl2....

It's really a pity as this seme dto be the ideal reduction: very cheap Zn dust, easy reaction conditions and workup, etc Just too good to be true.. i might try leaving the recation perform at room temp for 12h and see what happens.. Maybe dimerisation can be reduced enough to obtain a cristallizeable product...

Also, this reduction could be applied to the ethers of phenolic unsaturated ketones: methylzingerone is said to smell of ginger too, p-methoxyphenylbutanone is used in perfumery (not sure how it can smell though).
On the latest discoveries in the so-called Cassinone, 3,4-methylenedioxyphenylbutanone, a natural compound which is said to smell of blackcurrant, and is used commecially. I think piperonal is fairly regulated in my country, so I will surely go via the benzyl halide and acetoacetate anion for this one... That I will make in small quantities considering the lengthy preparation of the precursor.


Many thanks to Woelen for help with translations, and all the very kind people at the ref forums that helped me out with bibliographic research on this subject...

Info on the Cu/SiO2 reduction will be summerized here, or a link to the Cu/SiO2 thread will be placed here.

Anyone with further suggestions on a easy, efficient reduction is very welcome to comment further..


[Edited on 23-9-2008 by Klute]

Good news!


I have tried out the atm hydrogenation of dehydrorheosmin over Pd/C, using 25% mol/mol AcONa as an additive, in AcOEt, and after 14H all the unsaturated ketone was reduced to the saturate dketone, without any traces of the alcohol!

I'm unsure if it is this substrate that is much more ressitant to reduction of the carbonyl, or if using sodium acetate is very effective as suppressing 1,2-hydrogenation, but the reduction was text book perfect: regular H2 absorption, and theoritical amount consumed.

Using 1:1 AcOEtet ether as an eluant proved to give a superior seperation than with the eluant used during the Zn reduction. I am unsure of which side-products are formed during the reduction, but I'm pretty sure Zn/AcOH can't be used as a preparative procedure for phenolic butanones..

Sorry, still no photos I haven't got the camera issu dealt with yet..

Workup was simple: filtration of the catalyst and suspended AcONa, washing on the colorless filtrate, followed by drying and removal of the solvent..

The isolated product smells much nicer than the Zn reduction product! Much more subtle, "natural" aroma.

I can't wait to try the NiB hydrogeantion on the substrate. Even if it is clear than Pd/C is an ideal catalyst for such hydrogeantion, it si quite expensive and not always accesable. A cheap, easily made catalyst would really be usefull to many.

Ok, I just HAD to find a camera to take pictures of theses and show them to you guys... I have fallen in love of my Rheosmin and have been showinbg it to evryone I cross








96.6% Yield by H2/ Pd/C hydrogenation (actually I realized absorption was finished after 4h, as the theorical volume was 150mL! I guess the other 100mL were lost from small leaks or something. The absorption rate had greatly diminished afetr 150mL mark...

Ideal TLC eluant: 50:50 AcOEt: Pet ether. Will get Rf if someone wants them (saturated ketone higher than the unsaturated one, no traces of alcohol or other by-products!)

Let's hope NiB can do as good! I feel like making dozen grams of the stuff, my room smells beautifull Pretty strangely, the nice crystals smell much less than the crude powder, before recrystallization.

Thank you

They really are part of my favourites! Both the smell and the apparence!

Recently I have got few grams of 10% Pd/alumina, so I have tried hydrogenation of benzylideneacetone. Unfortunately, suspension of catalyst (0,3g) in 30g unsaturated ketone+50 cm3 of toluene does not want to "eat" hydrogen (under almost atmospheric pressure and 30-40 C). I also tried ethyl acetate instead of toluene - results the same.
This Pd/alumina is a little old - is it possible that it needs to be activated somehow ? I mean H2 stream nad ~300 C....
I do not know, I am not familiar with hydrogenations on Pd

Why not give sodium dithionite a try for the reduction of the double bond?

It works for isatylideneacetone, it might as well for your substrate. The procedure is so straightforward that it can't hurt giving it a try.



Tet. Lett. 49, 21 p 4741-4758 (1993)




Attachment: isatylideneacetones.pdf (1.8MB)
This file has been downloaded 1220 times

other lingand no problem....

http://pubs.acs.org/doi/abs/10.1021/om050625b

but i think this one might be carcinogen and it need's (silght) overpressure... would prefer the [Fe(CN)5NH3]3- which possibly is not toxic at all since the toxic effekt of the [Fe(CN)5NO]2- is induced by the elimination of NO

I've tryed to make [Fe(CN)5NH3]3- some time ago (wanted to use it on dehydrozingerone) but instead of obtaining nice red crystals i obtained a brown slurry, if someone made this already i would be thankfully for some hints ^^

Quote: Originally posted by Sydenhams chorea

Why not give sodium dithionite a try for the reduction of the double bond? It works for isatylideneacetone, it might as well for your substrate. The procedure is so straightforward that it can't hurt giving it a try. Tet. Lett. 49, 21 p 4741-4758 (1993)

Here's a little something for anyone interested. I ran the condensation of Vanillin with acetone (24h at room temperature), and recrystallized my product from 50:30 water:isopropanol. I then brought it over to school for use as a Michael acceptor in undergrad research. Here's a 400MHz proton NMR for you guys. Looks pretty good for all over the counter chemicals, huh?

The peak at 1.6ppm is water, but after running a blank of CDCl3 through the NMR, it's contamination of the chloroform, not the sample.

2.35ppm is the methyl ketone.

3.9ppm is the methoxy group

5.9ppm is the phenol hydrogen

The "triplet" with integration of 2 at 7.1ppm is the overlap of the singlet from the hydrogen in the 2-position on the ring with what I suspect is the 6 position hydrogen's doublet.

Unrelated, but has anyone seen what this stuff costs from Aldrich?

http://www.sigmaaldrich.com/catalog/ProductDetail.do?lang=en...



[Edited on 9-23-09 by UnintentionalChaos]

Quote: Originally posted by Klute

Nice! Thanks for sharing! It's a rather clean spectra you've got there! Looks like the reaction is pretty selective and no impurirites remain after a single recrystallization.. Do you intend on reducing it to the butanone? A spectra of the butanone would be great too!