Author Topic: Eugenol -> elemicin and myristicin  (Read 6208 times)

0 Members and 1 Guest are viewing this topic.

GC_MS

  • Guest
Eugenol -> elemicin and myristicin
« on: November 21, 2002, 08:50:00 AM »
Some of the bees here might be familiar with exotic substances as myristicin and elemicin. Myristicin can be used as precursor for MMDA, while elemicin can be applied in the synthesis of TMA-1 (3,4,5-trimethoxyamphetamine, the amphetamine analogue of mescaline). Both substances can be found in essential oils, or can be made synthetically. Rh's site contains information on how one can synthesize 3,4,5-trimethoxybenzaldehyde and myristic aldehyde from vanillin. However, there is no information regarding applying eugenol as precursor. The Hive's archive contains a thread (

Post 184952

(blue: "eugenol >> mmda (bromo snag)", Novel Discourse)
) about applying a bromination procedure. However, eugenol contains an allylic pi-bond which will be affected by bromination agents. So, it is impossible to use eugenol as precursor, isn't it? Well, euhm… Somewhere in a long forgotten and mystic country, there is chemistry library which contains many stinky and dark corners filled with obscure and occult journals and magazines. One of them contained the following:

Reference:
K Visweswara Rao, T R Seshadri, T R Thiruvengadam
Nuclear oxidation in flavones and related compounds. Part XXIV. Synthesis of myristicin and elemicin.
Proceedings of the Indian Academy of Sciences, Section A. 30 (1949) 114-119.

From the Department of Chemistry, Andhra University, Waltair

Received June 13, 1949

In some of the previous part [1] experiments using flavones were described in support of the mechanism of phytochemical ortho-oxidation that it takes place in multiple stages. The work is now extended to simpler benzene derivatives which have biogenetic analogy. Eugenol (I) is a fundamentally important compound in nature. It has the allyl-catechol unit which may be considered to arise in the same way as the nine carbon system comprising the side phenyl nucleus and pyrone carbon atoms of flavones (see Robinson [2]). Myristicin (II) [3] present in nutmeg, mace and parsley and elemicin (III) [4], a component of elemi oil, are derivatives of allyl pyrogallol and may be considered to be related to eugenol just in the same way as myricetin is derived from quercetin and robinetin from fisetin by processes involving ortho-oxidation. These considerations have led to a simple and elegant method of synthesis of these compounds (II and III) which is the subject-matter of the present communication. The suitability of this method of ortho-oxidation for preparing methylene-dioxy compounds has already been illustrated using the sample of kanugin [1].



Originally Mauthner [5] synthesised elemicin by subjecting 1,3-dimethyl-2-allyl pyrogallol (IV) to allyl migration and methylating the resulting hydroxy compound (V). Regarding myristicin Trikojus and White [6] outlined in a preliminary note its synthesis from 1-methyl pyrogallol (VI). It gave a mixture of two liquid monoallyl ethers (VII and VIII). They were separated and characterised as their 3,5-dinitrobenzoates. (VII) Rearranged on pyrolysis to give 3-methoxy-4,5-dihydroxy allyl benzene (IX) which was methylenated with methylene iodide to give myristicin (II).





Eugenol is now found to react with hexamine and give rise to a good yield of eugenol-5-aldehyde (X). This is a crystalline solid and exhibits the properties of an ortho-hydroxy aldehyde, forming a dinitrophenyl hydrazone and a sparingly soluble sodium salt and giving a deep blue colour with ferric chloride. In these reactions it resembles very closely orthovanillin. Its oxidation with hydrogen peroxide produces the corresponding catechol, 5-hydroxy-eugenol (IX). This is a viscous liquid, giving a green colour with ferric chloride and a yellow precipitate with lead acetate. It yields a colourless solid dibenzoate. Methylenation is carried out using methylene bromide and potassium carbonate in dry acetone medium. The product is a liquid whose identity with myristicin (II) is established by a study of its properties and also by preparing the solid dibromomyristicin dibromide. Similarly methylation of the catechol (IX) using dimethyl sulphate yields elemicin. For establishing the identity of the product, besides its properties, is employed isomerisation with alkali and subsequent oxidation with permanganate yielding trimethyl gallic acid. It is interesting to note that all the stages in the above synthesis take place without noticeable isomerisation of the allyl side chain. (See Fodor [7] for benzylation of eugenol without isomerisation).



EXPERIMENTAL

Eugenol-5-aldehyde (X)

A solution of eugenol (10 cc) in glacial acetic acid (75 CC) was treated with hexamine (40 g). The mixture was heated with shaking over a wire gauze to get a clear solution (pale brown) and was kept in a boiling water-bath for six hours. The dark brown-red solution was treated while hot with a boiling mixture of concentrated hydrochloric acid (50 cc) and water (100 cc). Heating on the water-bath was continued for another five minutes and the mixture slowly cooled. It was extracted twice with ether and the ether extract washed with water. The clear ether solution was then shaken with 20% sodium hydroxide, added cautiously in small lots of 20 to 30 cc. On shaking, the lower aqueous layer was colourless showing that only acetic acid had been extracted. Two or three such extractions removed all the acetic acid. Further addition of the alkali solution gave a bright yellow crystalline solid. After shaking vigorously the layers were allowed to separate and the lower layer diluted further with 10% alkali in order to complete the precipitation of the yellow sodium salt of eugenol-5-aldehyde. It was filtered and washed with 10% alkali and ether. It was then dissolved in excess of water and filtered and the clear filtrate acidified. After leaving overnight the pale cream coloured solid that separated out was filtered, washed well with water and dried. Yield, 3 g.
The alkaline filtrate from the sodium salt was acidified and extracted with ether. The ether extract was washed with water, concentrated to a small bulk and shaken thoroughly with strong aqueous sodium bisulphite. The crystalline bisulphite addition compound was filtered and washed with ether. It was treated with dilute sulphuric acid and the mixture heated in a boiling water-bath till the solid disappeared giving oily drops. The mixture was cooled in the ice chest and the aldehyde that crystallised out was filtered. Yield, 1 g.
The aldehyde samples thus obtained melted at 51-2°, when recrystallised from petroleum ether it was obtained in the form of pale yellow prisms melting at 53-54°. It was also purified by distillation under reduced pressure (20 mm), the fraction distilling at 148-50° being collected. On cooling in ice and scratching the sides of the tube, it rapidly solidified to give a mass of very pale yellow stout prisms melting at 53-4°. It was easily soluble in common organic solvents but sparingly in cold ligroin. In 10% sodium hydroxide it readily dissolved and the solution soon deposited yellow crystals of the sodium salt. With a drop of ferric chloride in alcoholic solution it gave a deep blue colour which did not change on further addition. (Found: C, 68.5; H, 6.6; C11H12O3 requires C, 68.8; H, 6.3%)

Dinitrophenyl-hydrazone of eugenol-5-aldehyde

An alcoholic solution of the aldehyde (0.2 g) was treated with dinitrophenyl-hydrazine (0.2 g) in alcohol (10 cc) and a few drops of concentrated hydrochloric acid. A copious crystalline precipitate of the dinitrophenyl-hydrazone was formed which was filtered after boiling for a few minutes. It was washed with alcohol and crystallised from ethyl acetate from which it separated as glistening orange red broad rectangular plates melting at 229-30°. (Found: C, 54.4; H, 4.4; C17H16O6N4 requires C, 54.8; H, 4.3%)

5-Hydroxy-eugenol (IX)

To a solution of eugenol-5-aldehyde (2.5 g) in pyridine (16 cc) was added one normal aqueous sodium hydroxide (19.5 cc) and the clear yellow solution treated dropwise with 6% hydrogen peroxide (9.6 cc) with vigorous shaking. The colour changed to red and when the solution developed turbidity a little more water was added in order to remove it. After leaving at the room temperature for 1.5 hours with occasional shaking, the solution was acidified with hydrochloric acid while cooling. The hydroxy compound separated as a heavy liquid; this separation was completed by adding common salt and the mixture was extracted thrice with ether. The ether extract was washed successively with hydrochloric acid, aqueous sodium bicarbonate and a small quantity of water. After drying over sodium sulphate it was evaporated to remove the solvent completely; a pale brown viscous oil was left behind (2.0 g). The product from three experiments was collected and distilled under reduced pressure (20 mm); the main fraction distilled at 176°.
5-hydroxy eugenol was a colourless viscous liquid. It was soluble in 5% aqueous sodium hydroxide and 10% aqueous sodium carbonate to give a deep brown solution. With a drop of ferric chloride in alcoholic solution it gave a deep violet colour changing to brownish violet and brown with another drop. After standing for a few minutes the colour changed to a stable olive green. It gave a yellow precipitate with lead acetate. (Found: C, 67.0; H, 6.4; C10H12O3 requires C, 66.7; H, 6.7%)

Benzoylation: dibenzoate of 5-hydroxy eugenol

The above dihydroxy compound was benzoylated by the Schotten-Baumann method. The benzoate separated from the alkaline solution in the form of a white sticky solid. It was extracted with ether, shaken well with 5% aqueous sodium hydroxide followed by water and dried over calcium chloride. On distilling off the ether a viscous liquid resulted which soon became an almost colourless crystalline solid. It was recrystallised from a mixture of ether and light petroleum from which it separated as colourless short rectangular prisms melting at 110-12°. (Found: C, 74.4; H, 5.2; C24H20O5 requires C, 74.2; H, 5.2%)

Methylenation: myristicin (II)

The dihydroxy compound (6 g) was dissolved in anhydrous acetone (200 cc), treated with methylene bromide (7.5 cc) and potassium carbonate (30 g) and refluxed for twenty hours. The solvent was then distilled off, the residue treated with water and the mixture extracted with ether. The ether extract was shaken with aqueous alkali, washed with water, dried over calcium chloride and distilled. The pale yellow oil left behind was distilled under reduced pressure (10 mm) when myristicin passed over at 138°. Yield, 2 g. Its refractive index was 1.5368 at 31° for white light, and 1.5343 at 30° and 1.5412 at 20° for the D line. (Found: C, 68.6; H, 6.6; C11H12O3 requires C, 68.8; H, 6.3%) When treated with a slight excess of bromine in petroleum ether solution (Pickles [8]) it yielded dibromo-myristicin-dibromide which crystallised from acetone-methyl-alcohol mixture as colourless prismatic needles and melted at 128-29° alone or in admixture with a sample made from natural myristicin.

Methylation: elemicin (III)

The dihydroxy compound (6 g) was methylated in anhydrous acetone solution (100 cc) with dimethyl sulphate (9 cc) and potassium carbonate (20 g) by refluxing for six hours. The solvent was distilled off, the residue treated with water and the mixture extracted with ether. The ether extract was shaken with aqueous alkali, washed with water, dried over calcium chloride and distilled. The residual liquid was distilled under reduced pressure (10 mm) when elemicin passed over at 148-9°. It had a refractive index of 1.5245 at 31° for white light and 1.5240 at 30° and 1.5292 at 20° for the D line. Yield, 4 g. (Found: C, 69.3; H, 7.4; OCH3, 45.1; C12H16O3 requires C, 69.2; H, 7.7; OCH3, 44.7%)
The synthetic sample of elemicin was subjected to isomerisation with alcoholic potash and subsequent oxidation with alkaline permanganate (Semmler [9]). The product was found to be trimethyl gallic acid, identical with an authentic sample.

SUMMARY

Employing the two stage process of ortho-oxidation eugenol is converted into 5-hydroxy eugenol. Methylation of this yields elemicin and methylenation myristicin. This constitutes the most convenient synthesis of these naturally occurring compounds and is highly significant form the point of view of biogenesis.
Note added in Proof - After submitting the above paper for publication, the detailed paper of Trikojus and White on the "Synthesis of Myristicin" (JCS 1949, 436) was received. They have now recorded that even the isomeric pyrogallol-1-methyl-3-allyl ether (VIII) yields myristicin, apparently by migration of the allyl group to the meta-position.

REFERENCES

1. Rao, Seshadri. Proc Ind Acad Sci A 1948, 28, 210 - Row, Seshadri, Thiruvengadam. Ibid. 1949, 29, 168.
2. 2. Robinson. Nature 1936, 137, 172 - Robinson et al. Phil Trans Roy Soc 1939, 230B, 149.
3. Thomas. Ber 1903, 3446 - Power, Salway. JCS 1907, 2054 - Baker et al. JCS 1932, 1281.
4. Semmler. Ber 1908, 1768.
5. Mauthner. Ann 1917, 414, 250 - Hahn, Wassmuth. Ber 1934, 696 - Bert. CR 1941, 213, 873.
6. Trikojus, White. Nature 1939, 144, 1016.
7. Fodor. Ber 1943, 1216.
8. Pickles. JCS 1912, 1433.
9. Semmler. Ber 1908, 1918 and 2183.




Swim is thinking of trying the procedure next week, depending on how much time he has available and how long it will take to finish his mescaline MW adventures. The article contains a couple of tweakable parts. We'll see  :) .


Ave Hive, synthetisandi te salutant!

pHarmacist

  • Guest
GC_MS
« Reply #1 on: November 21, 2002, 09:35:00 AM »
Allow me to be the first to say: GREAT FUCKING WORK GC_MS! When you leave the-hive, I'll leave too  ;) . I'm certanly going to give this a shot! Great! One question, can alkylamine (diethylamine) be used instead of pyridine in IX?


"Turn on, Tune in and Drop Out"

Rhodium

  • Guest
Eugenol finally useful
« Reply #2 on: November 21, 2002, 10:04:00 AM »


GC_MS: I have uploaded the article to

https://www.thevespiary.org/rhodium/Rhodium/chemistry/eugenol2myristicin.html



pharmacist: That is just an example of the well-known Dakin reaction. You can use many other reagents for that transformation, such as

Post 377353

(GC_MS: "Sodium percarbonate and the Dakin reaction", Novel Discourse)

GC_MS

  • Guest
Step 1
« Reply #3 on: November 21, 2002, 11:18:00 AM »
Step 1 is a Duff formylation, which is known to be low yielding in most cases. I refer to the review I posted yesterday:

Post 381544

(GC_MS: "Hexamethylenetetramine: a review", Chemistry Discourse)
.
Trifluoroacetic acid might increase the yield, who knows... we won't know till we try, I guess  ;) .

Ave Hive, synthetisandi te salutant!

moo

  • Guest
Luckily it is not the only ortho-formylation of a ...
« Reply #4 on: November 21, 2002, 11:32:00 AM »
Luckily it is not the only ortho-formylation of a phenol in the book, there are others with much higher yields. Nice find!

GC_MS

  • Guest
o-formylation
« Reply #5 on: November 21, 2002, 12:16:00 PM »
Luckily it is not the only ortho-formylation of a phenol in the book, there are others with much higher yields. Nice find!

The problem I see with most other formylation reactions is plenty of secondary reactions. Don't forget we are working with an alkene. There have been improvements on the Duff reaction though... I hope they will work. I just started a test run: 80 mL AcOH, 25 g hexamine and 5 mL eugenol. Was colourless in the beginning, became yellowish after 5 to 10 minutes (the typical beta-nitrostyrene kind of yellow) and after half an hour, the colour is shifting to a yellowish brown. So far, everything seems to go as written. This is just a test run though. Don't expect me to overwhelm you with extravagant yields tomorrow, as my experience tells me I'll fuck this up somewhere  ::) .

Edit:



Ave Hive, synthetisandi te salutant![orange]

Mountain_Girl

  • Guest
Oh-jubilation
« Reply #6 on: November 21, 2002, 11:42:00 PM »
Great & groovy GC_MS!

The abstract of this paper has actually been posted before:

Post 338053 (missing)

(ChemisTris: "Re: eugenol -> myristicin refs", Chemistry Discourse)


And another abstract which may be useful:

Post 275236 (missing)

(Mountain_Girl: "Re: Benzyl alcohols 2 benzaldehydes w/Fe(NO3)3", Chemistry Discourse)


Hmm, have to wait until Mon for your results  :(

Mountain Boy

GC_MS

  • Guest
hmzz
« Reply #7 on: November 22, 2002, 07:59:00 AM »
Hmmm, didn't see ChemisTris's post, even though SWiM used TFSE. Shame on us that we didn't notice this pathway back when (s)he posted it  ::) .

SWiM started some preliminary runs. He didn't record any yields nor modified too many procedures, since it is a first-time run.
Up till now:
A mixture of 25 g hexamine (the used hexamine was antique, used 5 g extra compared to the original article), 80 mL AcOH and 5 mL eugenol were thoroughly mixed under gentle reflux. The colour was initially yellowish, moved on to orange-brown and red-brown during the next couple of hours. After ca 10 hours, the mixture was red-brown. At this point, a solution of 60 mL concentrated HCl in 100 mL water was added to the refluxing solution (the diluted HCl was warm itself, to prevent a temperature drop of the reaction mixture). The mixture was refluxed for another 15 minutes, after which the RB was removed from the setup and allowed to cool down. The cooled mixture was extracted with 2 x 100 mL diethylether and the combined extracts were washed with 100 mL water. The organic phase was then treated with 20 mL 20% NaOH. The mixture was shaken well and the lower colourless phase was removed. Its smell was acetic acid-like. When another 20 mL NaOH 20% was added, more precipitation was noted. The mixture was shaken vigorously and the lower colourless phase was removed again. A certain amount (ca 150 mL) NaOH 20% was added and a yellow-red precipitate was noticed. Precipitation was allowed to continue for about 30 minutes. The precipitate was isolated via a Buechner setup. It was a yellowish creamy substance, as seen in the picture



This creamy substance was dissolved in 500 mL water. In order to accomplish this, the solution was vigorously stirred. Otherwise, dissolving the solid takes too much time. When dissolved, the yellow mixture (looks very much like beta-nitrostyrene type of yellow) was filtered. I filtered it the regular way, but next time, I'm going to use a Buechner for sure. Filtering seems to be necessary, as there are a couple of insolubles. Then, 50 mL concentrated HCl is added. The effect can be seen in the following picture:



Nice, huh  8) .
I've put it in the fridge now, where it will be allowed to precipitate for the weekend. I hope that the precipitation works out fine, so I can continue with subjecting this substance to the Dakin reaction. I'll use H2O2/pyridine, not percarbonate in sonication bath. I'm not completely sure about the interference of the carbonate/H2O2 system and the allyl bond.
If all goes well, I'll perform a new synthesis including detailed calculations and notes on yields during next week.

Ave Hive, synthetisandi te salutant!

moo

  • Guest
Formylations
« Reply #8 on: November 22, 2002, 08:21:00 AM »
It is obvious Reimer-Tiemann would have no chance of working, but what about the MgCl2/paraformaldehyde/triethylamine formylation, as in

Post 290438

(Lilienthal: "2-hydroxy-5-MeO-BA: The easy way", Chemistry Discourse)
?

GC_MS

  • Guest
CH2O
« Reply #9 on: November 23, 2002, 08:11:00 AM »
It is obvious Reimer-Tiemann would have no chance of working, but what about the MgCl2/paraformaldehyde/ triethylamine formylation, as in Lilienthal: "2-hydroxy-5-MeO-BA: The easy way" (Chemistry Discourse)?

Formaldehyde can add to the alkene bond. Don't know if the use of paraformaldehyde will prevent that.

Ave Hive, synthetisandi te salutant!

GC_MS

  • Guest
update
« Reply #10 on: November 24, 2002, 12:36:00 PM »
OK, here is a small update on my test run results...

I allowed the eugenol-5-aldehyde to precipitate over the weekend. When I arrived in my lab, the turbid mixture was completely clear and yellow-orange crystals were noticed on the bottom of the beaker. The crystals were isolated via gravity filtration. They were yellow to orange in colour when they were dried (dry at open air - room temperature - and not in an oven. The crystals will melt and form an oil which is hard to recycle). Some crystals were dissolved in iso-octane and injected on GC-MS (splitt ratio 10:1). There was only one peak (!nice! -> pure product); I am unfamiliar with its mass spectrum, but the most abundant m/z was the molecular ion (M+ = 192 amu). Since the molecular weight of eugenol-5-aldehyde is 192 as well (if the M+ is the most abundant m/z, it probably means that the unknown molecule is rather stable, in most cases stabilized by resonance, e.g. polyphenylic structures), I hereby conclude that the first part of the synthesis was succesful  :) .
I dumped the crystals in 15 cc pyridine and added some NaOH (1 M) and H2O2 (2 mL H2O2 30% diluted with 8 mL H2O). This is the 2nd part of the synthesis; the reaction is also known as the Dakin reaction. And I think something went wrong here... My final ether extracts contain nothing  :( . I must have done something wrong, have to find out what it was. I made small modifications to the paper's procedure, but not in that extent that I'd end up with nothing. Anyway, I'm already satisfied that the formylation worked out fine.
I'm going to start a new run right now. This time, weights and yields will be recorded. I'll keep you posted  ;) .

Edit:

They were yellow to orange in colour when they were dried

I recrystallized a few yellow/orange crystals from petroleum ether, and the colour changed to an attractive fluo yellow. Certainly not pale or near pale yellow. Chromatogram indicates pure product though (or contaminants are not chromatographable).

Ave Hive, synthetisandi te salutant!

GC_MS

  • Guest
update
« Reply #11 on: November 29, 2002, 05:28:00 AM »
I've been busy with this synthesis during the last couple of days, but have to stop it for a week... my heating equipment is kaput  :( .
The really annoying is that I was distilling some impure crystals under reduced pressure to obtain pure 5-formyl-eugenol. The resulting oily substance crystallized in the condenser and was hard to remove. They can be dissolved by using organic solvents, but then I have to recrystallize them again. Anyway, I'll try all over again when I have a new decent heating apparatus.
I made a couple of pictures to visualize the synthesis process. The pure 5-formyl-eugenol is indeed pale yellow in colour; I think the impurities are not GCable. I obtained a nice amount of this substance till now, so it is proven that the method really works (since it has been written by Indians, I was not completely sure about it... have had bad experiences in the past). Don't expect super excellent yields though. The article mentions that 3 g crystalls and 1 g via the bisulfite adduct were obtained using 10 mL eugenol. This is probably correct. I obtained ca 6.5 g of crystallized products (combined) using 15 mL eugenol. This 6.5 g was impure...
I'll keep you posted when I have final results.

Ave Hive, synthetisandi te salutant!

Lego

  • Guest
Oxidation of eugenol to mescaline/MMDA precursor
« Reply #12 on: April 30, 2004, 12:50:00 PM »
Derivatives of methylendioxybenzene, 24th communication
F. Dallacker, R. Sluysmans

Monatshefte, 1969, 100, 560-566

(http://lego.chemistry.tripod.com/Journals/Monatshefte1969.djvu)

[...]

3-methoxy-5-allyl-o-benzoquinone (4) and 3-methoxy-5-allylcatechol (5)

A suspension of 125 g potassium nitroso disulfonate (Fremy's salt) and 1500 ml H2O, buffered with 20 g KH2PO4 which is placed in a 2 l Hüttentrichter [most likely a special seperating funnel] is treated with a solution of 24.6 g of 3-methoxy-4-hydroxy-allyl-benzene (eugenol) and 25 ml acetone.  The solution is shaken vigorously and within minutes the solution turns deep red and after about 10 minutes a deep read product precipitates. It is not recommened to isolate the o-quinone (4). The solution is shaken for further 5 mintes, 1 l ether is added, cold saturated Na2S2O4 [sodium dithionite] solution is added until complete discolouration, the ether phase is seperated, the aqueous phase is extracted several times with ether, the pooled extracts are washed one time with H2O, dried over MgSO4 and evaporated. Yellow-red oil, pure according to GC analysis, b.p.2 111°C.


No explicite yield is given but the authors claim that the product is obtained in high yields.




As Fremy's salt is too expensive for this purpose here a two methods for synthesis at home:

http://www.chem.umn.edu/class/4711/mann02s/sylschsup/expa.pdf


Thanks

Post 300203

(Natrix: "Fremy's Salt", Chemistry Discourse)

____ ___ __ _

Oxidations with potassium nitrosodisulfonate (Fremy's radical)
The Teuber reaction

Hans Zimmer, David C. Lankin, Stephen W. Horgan
Chem. Rev., 1971, 71, 229-246


1 . Preparation of Fremy’s Salt
Sodium nitrite ( 5 M, 100 ml) is placed in a 1-l beaker and cooled in an ice bath. Chopped ice (200 g) is added and the solution tirred steadily during the addition of fresh sodium bisulfite solution (100 ml, 3 5 z w/v), followed by glacial acetic acid (20 ml). Reaction is complete in 2-3 min, as shown by the momentary darkening in color of the reaction mixture and by its failure to decolorize iodine solution. After addition of concentrated ammonia solution (25 ml, sp gr 0.88), the mixture is again cooled in an ice bath, and fresh ice added whenever necessary to keep some present in the reaction mixture throughout the next stage. Ice-cold 0.2 M potassium permanganate (400 ml) is now added dropwise with continued stirring, during ca. 1 hr. 
The precipitated manganese dioxide is removed by gravity filtration (Whatman No. 5,24 cm), using two or more funnels in parallel to reduce the time required. The filtrate is allowed to come to room temperature as filtration proceeds, but any unfiltered suspension is kept in an ice bath. A portion of the filtrate (10-15 ml) is treated with an equal volume of saturated potassium chloride solution to precipitate some Fremy’s salt for seeding the main batch. The bulk of the filtrate is stirred steadily, while saturated potassium chloride solution (250 ml) is added dropwise over a period of about 45 min. Small portions of the previously prepared suspension are added from time to time during this period until the solid persists in the bulk solution. Precipitation is completed by stirring the bulk solution cooled in ice for a further 45 min.
The orange solid is collected on a Biichner funnel but is not sucked dry. It is washed with ammoniacal saturated potassium chloride solution (containing ca. 5 v/v 0.88 ammonium hydroxide), twice with ammoniacal methanol (containing ca. 5% v/v 0.88 ammonium hydroxide), and finally with acetone. Only after the whole washing process is all the liquid sucked away, but even then air is not drawn through. The solid is spread on a watch glass and the acetone allowed to evaporate for 10-15 min. Finally, the orange crystals are stored in a desiccator over calcium oxide, in the presence of ammonium carbonate in a separate dish to provide an ammoniacal atmosphere. Under these conditions even this relatively crude material is stable for several months (crude yield, based on bisulfite, 81-82 %).

2. Recrystallization of Fremy’s Salt
For bulk recrystallization, Fremy’s salt is suspended in a solution, 2 M in potassium of whichever potassium salt is selected. Solution is completed by heating, if necessary up to 50°, when the solubility is of the order of 3 g per 100 ml of solution. The resulting solution is filtered and cooled overnight and crystallization completed by cooling in ice for 2 hr. The filtered solid is washed with methanol (twice) and acetone (twice), dried in air, and stored in a dry ammoniacal atmosphere as before. By carrying out the recrystallization in about six 10-g batches, using the same liquor throughout, it is possible to attain a yield of 62-65 %, relative to thc original bisulfite, of analytically pure product. The product can contain both crystalline modifications of Fremy’s salt.