Sciencemadness Discussion Board

TEMPO substitute?

 Pages:  1  

chloric1 - 29-5-2005 at 05:49

OK I have recently taking interest in the alternative organic reactions involving TCCA and Oxone as environmentally sound and safe alternatives to traditional oxidants and halogenating reagents. Not to mention the extensive OTC availability of these reagents.:D

The problem? Several reactions such aas conversion of amines to nitriles with TCCA involve TEMPO. This is an exotic catalyst that is virtually unavailable to most home experimenters. The Merk index states the this compound is a stable nitroxide and is actually a "radical trap". With that is mind, I don't feel a simple substitution with benzoyl peroxide or MEK peroxide would work. I someone more experienced in organic synthesis could explain the mechanical happenings in TEMPO mediated reactions and suggest alternatives, I think we could have a worthwhile investigation on our hands.;)

[Edited on 5/29/2005 by chloric1]

Nicodem - 29-5-2005 at 07:56

If I remember corectly TEMPO is prepared from its amine version 2,2,6,6-tetramethylpiperidine by some kind of oxidation (perhaps with H2O2 or peroxyacids?).

Well, 2,2,6,6-tetramethylpiperidine is not OTC, but a similar compund 2,2,6,6-tetramethylpiperidin-4-one can easily be prepared from ammonia and acetone by letting the mixture of both stand for a month or so and then vacuum distilling the wanted product. I tried this once and it works nice. It can be made from very cheap chemicals in large amounts.

The only problem is how the additional carbonyl would interfere in the oxidation to the TEMPO analogue and how such an 4-keto-TEMPO would work as catalyst in radical oxidations. Given that the 4-hydroxy-TEMPO (aka TEMPOL) has some similar use as a radical trap, than perhaps the 4-keto-TEMPO would work as well.

Absolutely splendid!

chloric1 - 29-5-2005 at 08:34

If you could briefly describe the ketoTEMPO synthesis or provide academic information of the synthesis of ketoTEMPO I would greatly appreciate it.

Stupid question but what is a "radical trap"? How or why does it function? As opposed to a simple free radical initiator such as benzoyl peroxide of coarse.

Nicodem - 29-5-2005 at 15:03

I can't give you a literature references as I based my synthesis exclusively on a mention in an very old Czech book from the early beginning of the 20th century where it was said that ammonia and acetone yield 2,2,6,6-tetramethylpiperidin-4-one after the work up of a mixture left standing some time (or some similar short note without any further reference). The reaction mechanism is a simple double condensation of acetone-imine on acetone followed by cyclisation.
I’m sure you can find something on the preparation in the patents as this product is really not that obscure as it sounds by name.

My notes on that experiment are partially lost so I will try to reconstruct by memory. I left standing for a few months a mixture of 200ml acetone, 100ml 25% ammonia and about 1g of ammonium acetate (or acetic acid) in bottle. I later forgot about it otherwise one month should be more than enough, in fact I remember already after a week or two the mixture become red. This mixture was vacuum distilled. The acetone and water that came over first were discarded. The fraction boiling over from 60°C – 120°C (not sure anymore about the temperature) was collected, washed with brine and distilled once more, but this time over a shorter temperature interval. If I remember correctly some 40g of a yellow oil was obtained. It is only partially miscelable with water, soluble in acids and smells like… well, like most amines.

From here on to “4-keto-TEMPO” I found a reference where Oxone is used:

Radical traps are those compound who can “catch a radical” by forming a very stabile one. In case of TEMPO and similar oxylamines, they are more stabile because of the particular electronic configuration of the R2N-O* radicals, but here furthermore also because being surrounded by four methyl groups which hinder access to it (no dimerisation can hapen). Hence the OH hydrogen in TEMPO is very easily abstracted by any other radicals. Google around for a more thorough explanation.

docberto - 29-5-2005 at 19:50

I know BHT, the food preservative, forms a very stable oxy radical. It is used as a radical trap to prevent food from going rancid. The phenol oxygen's hydrogen is readily abstracted by a radical and the oxy radical is stabilized by the aromatic ring. The ortho t-butyls sterically hinder the oxy radical. If BHT would be suitable for this, BHT is very OTC.

Your help is more than appreciated

chloric1 - 30-5-2005 at 05:30

How could I thank you enough? After all that is what this forum is all about. I have to wait until tomorrow to acces the article though because I only have access to online journals from work. But, I must say, the process looks promising. Could other nitroxides do the same thing as TEMPO? In the next few months I will have to do the research here with my own chems. Also, considering lack of reference about condensation of ammonia with acetone followed by cyclation, I will have to test different variables to this. It is hard to believe that mixing ammonia with acetone and waiting a few weeks is the only option.


chloric1 - 16-6-2005 at 13:01

Hey check this out this has been sitting two weeks now. Keep in mind the following, first I only mixed 100ml of acetone with 100 ml of 28% ammonia, second, the bottle was open for several days as I did not have the correct size stopper!:( I since procurred the stopper from my local hardware store!:cool: and I added the 100 ml acetone needed.

If you look at the picture you can see the bright red color along with some solid particles. Are these part of multiple condensation reactions? The ammonia odor is quite a bit diminsihed and there is absolutely no hint of acetone.

[Edited on 6/16/2005 by chloric1]

[Edited on 28-1-2007 by chemoleo]

DSCN1598.JPG - 59kB

Nicodem - 18-6-2005 at 12:18

Yes, I remember that viscious red stuff floating in the mixture (just as on your picture). But it is not really crystaline in my experience. You will see when you will be distilling the stuff that it will form two layers as soon as a little acetone comes over.
I forgot to mention that the vacuum in the distillation was aplied only after the acetone came over. But beware that lots of ammonia will came over at this stage so you better drive the gass somewhere out or in the sink running water (I simply loosely attached the vater jet pump tubing in such a way that it sucks air as well, thus still keeping the normal pressure but sucking noxious gasses).
A considerable amount of an orange-red resin remains in the distillation flask. The distillate must be redistilled as it is still not pure (probably still containing some of the not-yet-cyclisized H2N-C(Me)2-CH2-CO-CH2-C(Me)2-NH2 ).

Interesting developement

chloric1 - 19-6-2005 at 09:57

Nicoderm- Out of boredom the other day, I came across the following patent. I searched under "triacetoneamine" as it is easier to remember. The proceedure sounds promising. It is identical to what we are already doing the only modification is the use of a Ca containing zeolite as a catalyst. Since sodium zeolites are effective ion exchangers, that proceedure would not be to tricky. I have seen the "lab grade" zeolite pellets for removing ammonia from fish tank water but I do not know what zeolite it is. I assume it would be the same as water softeners.

A link to the patent

chloric1 - 19-6-2005 at 09:58

I guess I could attach the link to the patent to.:D

Cannot attach link but it is patent US6646127

[Edited on 6/19/2005 by chloric1]

Nicodem - 19-6-2005 at 13:33

Nice patent with many further references. But I don't think it is very worthwhile to search for any special zeolites even if they are that effective. Unless you already have them or are that easy to obtain, off course. The improvement seams to be in the cyclisation which then happens already in the liquid phase. But you will have to distill the product anyway, though perhaps then one distillation would be enough for the crude product.
But given that both acetone and ammonia are so cheap any complications in using special reagents or catalysts is simply too much of a fuss.

I added some ammonium acetate to be the source of ammonium cations which can then transfer the proton to the acetone imine (Me2C=NH) and make it more electrophylic for the condensation. But even without an acidic catalyst the reaction proceed the same as you noticed (though it is slow I admit).

Another acidic catalyst which might work almost as good as the zeolites is silicagel. This might be worth trying as it is so cheap.

guy - 26-1-2007 at 22:30

From a patent

Seems simple enough.
Example 1

A suspension containt 11 g ammonium chloride and a mixture
of 340 g acetone and 64 g of methanol is saturated over the
course of 12 hours at 13 degrees to 17 degrees with ammonia
gas. The resulting colorless oil is subsequently diluted with
350 g acetone, and maintained, with stirring, for 15 to 20 hours
at 50 - 55 degrees. Excess solvent is removed in vacuo, and
36 g of water is added to the red residue. The
crystallization occuring at 0 - 5 degrees is completed by 2
hours' stirring. The result is 286 g of
2,2,6,6-tetramethyl-4-oxipiperidine hydrate. Melting point 55-60 degrees in the form of slightly yellow crystals.

Wow. The reactants are definitly OTC and the conditions
(ie temperature) are pretty mild. This seems very doable.

[Edited on 1/27/2007 by guy]

[Edited on 1/27/2007 by guy]

Attachment: triacetone amine.pdf (264kB)
This file has been downloaded 1052 times

chemrox - 27-1-2007 at 00:11

What does one do with this stuff?

Did "saturade" mean saturated?

And please forgive my ignorance, I come here to learn as well as share, what is TCCA?

guy - 27-1-2007 at 01:04

Originally posted by chemrox
What does one do with this stuff?
It is a precursor for making TEMPO a very useful selective catalyst for oxidizing things.

Did "saturade" mean saturated?


And please forgive my ignorance, I come here to learn as well as share, what is TCCA?
Trichloroisocyanuric acid

jon - 27-1-2007 at 01:15

Is that a mannich reaction?

guy - 27-1-2007 at 11:38

Yeah kind of.

And about the ammonium chloride, could ammonium sulfate be subsituted?

[Edited on 1/27/2007 by guy]

Nicodem - 27-1-2007 at 12:13

Nearly any ammonium salt will do. I used ammonium acetate since that was what I had nearby. That patent method uses ammonium chloride. But the point is in the catalytic effect of the weakly acidic ammonium cation (NH4+) while the anion part does not play much of a role except for solubility issues (ammonium sulfate will not be very soluble in the reaction mixture unless you use the aqueous ammonia method).

jon - 27-1-2007 at 15:15

can you use a primary amine in that reaction to get n-sustituted 2,2,6,6, tetramethyl piperidine-4-ones?

guy - 27-1-2007 at 15:19

I think so, but then you can't oxidize the subsituted amine to form the nitroxide radical.

jon - 27-1-2007 at 19:06

there was some big discussion about this a while back for use in the prep of variants of 1-propionanilido-4-substituted piperidines.

guy - 28-1-2007 at 14:08

Nicodem, what did your triacetoneamine smell like? I put some ammonia, acetone, and ammonium sulfate in a glass bottle and put a cap on it. I heated at around 50 degrees for 24 hours. The solution was red. Now Im boiling off some water and this smell came out: a very sweet but has an amine odor to it. Is this the right odor?


I attatched a picture of the distillate...Is it too yellow? Should it not be a solid?

[Edited on 1/29/2007 by guy]

[Edited on 1/29/2007 by guy]

P1010007.JPG - 38kB

womble - 29-1-2007 at 00:16

Here is a pdf reference to the use of tempo (the 2,2,6,6, tempremental piperidine whatchacallit). It is useful for hte oxidation of alcohols according to the reference.

guy - 29-1-2007 at 13:56

Should the triacetoneamine form a complex with bisulfite, since it is a ketone?

Nicodem - 29-1-2007 at 14:14

It was several years ago so I don't really remember the smell, except that it was amine-like, but not nearly as unpleasant as piperidine. I guess your description fits. I only identified the compound as "triacetoneamine" from its chemical properties and approximate boiling point during the vacuum distillation with the aspirator (lit. values: 84-88°C at 15 torr; 102-105°C at 18 torr). I remember that the liquid was yellow at the beginning, but after a couple of months deteriorated by air oxidation into some red vicious oil.
Also according to one reference, triacetoneamine can form a monohydrate.

Otherwise, here are two other references for the synthesis of 2,2,6,6-tetramethylpiperidin-4-one:
US2002128482 : 23% yield, 17h/25°C, aq. ammonia, acetone, zeolites (this patent is particularly nice since it also gives a literature review with references to other methods)
Synthetic Communications, 26(19), (1996) 3565-3569 : 65% yield, ammonia, acetone, catalyst

I don't know if it forms a bisulphite adduct, but ketones often do not form very stabile bisulphite adducts unless they are of R-CO-Me type.
If you plan to oxidize it to 4-keto-TEMPO let me know so I can provide you some references.

[Edited on 29-1-2007 by Nicodem]

guy - 29-1-2007 at 14:46

Originally posted by Nicodem
It was several years ago so I don't really remember the smell, except that it was amine-like, but not nearly as unpleasant as piperidine. I guess your description fits. I only identified the compound as "triacetoneamine" from its chemical properties and approximate boiling point during the vacuum distillation with the aspirator (lit. values: 84-88°C at 15 torr; 102-105°C at 18 torr). I remember that the liquid was yellow at the beginning, but after a couple of months deteriorated by air oxidation into some red vicious oil.
Also according to one reference, triacetoneamine can form a monohydrate.

Otherwise, here are two other references for the synthesis of 2,2,6,6-tetramethylpiperidin-4-one:
US2002128482 : 23% yield, 17h/25°C, aq. ammonia, acetone, zeolites (this patent is particularly nice since it also gives a literature review with references to other methods)
Synthetic Communications, 26(19), (1996) 3565-3569 : 65% yield, ammonia, acetone, catalyst

I don't know if it forms a bisulphite adduct, but ketones often do not form very stabile bisulphite adducts unless they are of R-CO-Me type.
If you plan to oxidize it to 4-keto-TEMPO let me know so I can provide you some references.

[Edited on 29-1-2007 by Nicodem]

Thanks for the information. Yes I am planning to oxidize it. I heard it can be done with peroxyacids (oxone?). Those references would be nice. Thanks

I wonder how the zeolite works?

[Edited on 1/30/2007 by guy]

Nicodem - 30-1-2007 at 10:46

References for the N-oxidation of triacetoneamine:

Tetrahedron Letters, 36(31), (1995) 5519-5522 : Oxone, 82% yield (ref already given upthread)

Journal of the American Chemical Society, 117(42), (1995) 10555-10562 : H2O2, Na2WO4/H2O, 85% yield

Tetrahedron Letters, 29(37), (1988) 4677-4680 : DMDO/acetone, 98% yield

Synthetic Communications, 20(4), (1990) 597-601 : m-chloroperbenzoic acid / Et2O

Chemical Papers, 42(2), (1988) 243-248 : H2O2, 65% yield

Makromolekulare Chemie, 187(11), (1986) 2545-2551 : H2O2, 88% yield

Journal fuer Praktische Chemie (Leipzig), 327(6), (1985) 1011-1014 : H2O2, Na2CO3

Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya, (2), (1983) 391-394 : 97%

Chemical & Pharmaceutical Bulletin, 28(11), (1980) 3178-3183 : H2O2, 49% yield

EP1595868 : H2O2, NaHCO3, 4h/60°C, 83% yield

GE4432795 : t-BuOOH/ClCH2CH2Cl, 85% yield

US5087752 : DMDO/acetone, 98% yield

WO8705222 : 84% yield

WO2004085397 : AcOOH/AcOH, buffer

guy - 30-1-2007 at 13:04


[Edited on 1/30/2007 by guy]

Triacetoneamine Synthesis

Klute - 5-4-2008 at 07:28

Bringing this thread to life again!

Over 6 months ago, i prepared a bottle to make some triacetoneamine, to oxidize it to 4-keto-TEMPO as a catalyst of benzylic/secondary alcohols oxydations. Initially, i had planned on leaving it for a couple of weeks, but as time passed i forgot about it, reminding myself to distill it every now and then...
Being very bored because of my incapacity to manipulate after having wounded my left hand, i couldn't resist anymore and decided on distill the mixture. Someone gave me a hand (both litteraly and figuratively :D ) to put the distn setups up, and i was off to isolate my 2,2,6,6-tetramethylpiperidin-4-one.

Synthesis of 2,2,6,6-tetramethylpiperidin-4-one (Triacetoneamine)


I didn't take notes when i mixed the reagents, so i'm not 100% sure of the numbers. Keep in mind that this reaction wasn't done to obtain good yield, but just to obtain some product to work with. A lots of improvements can be done, especially during workup, i just wanted to get my product with the least of efforts, considering my handicap.

In a 500mL bottle, 200mL technical acetone was mixed with 100mL 20% aqeous NH3. Roughly 5g of ammonium chloride and 5g of silicagel were added, and the bottle was shaken. This was left at room temp for over 6 months, shaking occasionally. The colour change started after a week or two. Strangely, the colour was more orange than red as in the picture chloric posted [Picture 1], surely because more dilute ammonia was used. A slight pressure was released a few times. The ammonia smell was less intense as it was as the beggining.

Picture 1

Distillation of the excess reagents

The now fluo (when a light was directed at it) orange mixture was filtered through a cotton plug into a 500mL distn flask. The liquid had a viscosity comparable to that of cooking oil. A simple disnt setup was attached, with a 15% H2SO4 wash bottle connected to the vacuum inlet. As heating started, the air present in the system was chased out, at a flow of 1bubble/1-2sec, which increased as ebullition started. Then ammonia gas was absorbed, as seen but the instant absorbtion of the bubbles and a "cracking" sound. At first, some condensate started passing under 40°C, surely acetone vapors pushed by the ammonia gas. Then a constant take off started, with vapors temp slowly increasing from 50°C to 75°C. An orange oil started crashing out as temp increased above 60°C [Picture 2], and some clear organic oil was seen steam distilling (Note 1). At 75°C vapor temp, the flask was cooled, and the top orange oil (Note 2) was seperated in a seperating funnel [Picture 3][Picture 4].

Picture 2

Picture 3

Picture 4

The orangish aq layer was returned to the distn setup, and distn continued until vapors temp increased over 85°C. A very small amount of oil seperated out. The clear distillate had a small amount of oil, which quickly turned to a white solid (Note 3).

Vacuum fractionnation of the crude oil

The combined organics were directly introduced into a 250mL distn flask [Picture 5], attached a a vacuum fractionnation setup [Picture 6] , using a apirator as vacuum source.

Picture 5

Picture 6

At first some water passed over, with a clear light yellow oil (Note 4). Some white smoke started passing. Vapor temp stabilized at 55°C. Roughly 20mL of water were collected, with ~5mL of oil floating above. Heating was increased, more smoke was produced, to quite an extent [Picture 7]. Then very strong foaming started, to a point were vacuum couldn't be left on for more than a few seconds before the foam dashed up the vigreux. The black/brown residu was transfered to the 500mL flask (heavy smoking in air), and distn started again. This time the foaming was well contained.

Picture 7

Finally, temp increased, smoking diminished, and a light golden yellow oil started passing over [Picture 8]. A first fraction was collected (70-95°C) [Picture 9], weighing 10.44g, followed by another one at 95-105°C [Picture 10] weighing 18.5g. The second fraction was slightly darker. Distn was stopped as head temp increased above 105°C.

Picture 8

Picture 9

Picture 10


Note 1: This was surely the triacetoneamine. This means it could be purified by steam distn if someone is willing to take the time..

Note 2: This oil was pretty viscous, as motor oil, and had a sweetish amine smell, pretty consistant with Guy's description.

Note 3: This is surely the product forming a hydrate with the distd water. I wanted to collect the solid, but when cooling a stupidly forgot to disconnect the wash bottle, and some H2SO4 was sucked back, dissolving the solid.

Note 4: The large amount of water passing could be du to some uncondensed product cyclizating (sp?) and thus generating water. Or it's just that the oil can dissolve large amounts of water. Solvant extraction and drying could be more efficient and avoid excessive degradation (smoke).

A few attempt of making a salt were made, adding some H2SO4 in abs EtOH (10% w/v) in the oil dissolved in a little IPA caused some white solids to appeared but they formed an oily layer after a few seconds. Adding 30% HCl in IPA follwed by acetone didn't produce any cristallization. Any suggestion regarding salt formation are welcome, as i would like to keep some product as a salt for better stability, because I don't want to oxidize everything in one go.

To oxidize this amine, i will surely try H2O2/NaHCO3, and perhaps Oxone if the results aren't great. Any advise on this subject are also welcome.

Hopefully, a very efficient oxydation catalyst is at the reach of home chemist. This procedure requires long reaction times, though a few weeks are surely only needed, but using dry NH3 gas, although more involved, cut reaction times to a few hours (see mentionned patents).
The broad uses and numbers of related publications is very inspiring.

I also want to try reducing the keto- to hydroxy- , to hydroxy-TEMPO, but i'm scared oxidation of the amine would also oxidize the hydroxy group. But things like acetoxy-, methoxy-, etc could be obtained if a easy way of protecting the secondary amine could be found. Any ideas?
Well, perhaps 4-keto-TEMPO is as efficient as the rest, so why bother.

Once the catalyst is ready, the results of the oxidations will be posted in the benzylic/secondary alcohols oxidations thread.

Nicodem - 5-4-2008 at 10:38

Well, what can I say? Nice work as always, Klute.

Was there a jump in distillation temperature rising between the 70-95°C and 95-105°C fractions? Because otherwise it looks like you have triacetoneamine in both fractions and maybe you should redistill for a more pure product (if needed pure at all). Also, it is quite possible a lot of water dissolves in the product or maybe the water distilling out from the crude oil was the consequence of the hydrate dehydrating. That would also explain the wide distillation interval in spite the distillation column.

The hydrochloride has an mp of 186-190°C (CAS: 33973-59-0). It is actually commercially available and described as "white to slightly beige crystalline powder".
You should try neutralizing the triacetoneamine solution in isopropanol with conc. HCl, then dry in vacuum and triturate in diethyl ether (the salt might actually be soluble in acetone).
Triacetoneamine hydrate has an mp of mp 59-61°C. Actually, since the hydrate and hydrochloride are the only commercially available forms, I'm leaning toward believing that the normal (nonhydrated) form might not be particularly stable. This is also in accordance to my experience. My triacetoneamine started deteriorating in few weeks (visibly at least).
The tosylate seems to also exist judging from the fact that it has a CAS number (29334-13-2), but I could find no further information.

Edit: Checked the literature and found these other salts exist as well:
nitrate (; thiocyanate, bromide, iodide, perchlorate, and trifluoroacetate (; sulfate, fumarate, acetate, oxalate, benzoate, picrate… in short, it forms crystaline salts with just about any proper acid.

[Edited on 6/4/2008 by Nicodem]

Klute - 5-4-2008 at 10:51

Thank you, Nicodem. I couldn't have done it without all the ref you provided...

Yes, the freebase seem quite sensible, it has darkened to an orange color overnight! The entire 2nd fraction was subjected to oxidation.
The first fraction justr steadily increased in temperature, i think it is quite pure TAA, just that the smoke entrained the vapors, so the temperature wasn't that of the refluxing vapors... It might contain a bit more impurites than the second though. I will try to crystallize and recrystallize it with aq HCl and IPA as you propose. I'm not too confortable with the acetone/H2SO4 technic mentionned in one of the patents. If it doesn't crystallize out, removing the acetone will surely creat more by-products.
Except for the steam distilled residu, i haven't seen any of the distilled oil form a hydrate, at least not immediatly. I tried adding a drop of water to a mL of TAA just before the oxydation, they simply formed a homogenous yellow solution. I guess it needs to be done in the cold and in more time.

Oxydation of 2,2,6,6-tetramethylpiperidin-4-one (Triacetoneamine) to 4-Oxo-2,2,6,6-tetramethylpiperidine-N-oxyl (4-oxo-TEMPO) with alkaline H2O2

The procedure from US 0256312A1 given above was followed, using AcONa as cocatalyst. The workup was inspired by WO8705222.

All the used glassware was washed with distilled water to avoid contamination with metal salts.

18.5g (119.18 mmol) of distilled 2,2,6,6-tetramethylpiperidin-4-one (Note 1) were placed in a 100mL 3-neck RBF with a stir bar[Picture 1]. 5mL of water were used to rinse the container. This formed a homogeneous orange solution.

Picture 1

In a seperate container, 1.05g (12,5mmol; 0.95eq) of NaHCO3 were covered with 5mL of water, and 0.2g (3.33mmol; 0.27eq to the NaHCO3) of acetic acid were added dropwise, with stirring (CO2 evolution). The erlenmeyer was slightly heated to dissolve the NaHCO3. This solution was added to the RBF, giving two layers [Picture 2], and heated in a oil bath to 60°C.

Picture 2

17.93mL (208.57 mmol); 1.75eq to the TAA) of 35% H2O2 were then added dropwise over 3h, keeping the temperature between 55 and 65°C. The orange color gradually turned dark red over the first hour [Picture 3], and darkened further [Picture 4]. At the end of the addition, stirring was continued for another hour at 60°C [Picture 5].

Picture 3

Picture 4

Picture 5

The flask was cooled to room temp, and 10mL of water was added to the dark red emulsion, followed by 25mL DCM. pH was brought to 2 with 10% HCl (very slight bubbling), and the two layers seperated. The orange aqueous was extracted with 3*10mL DCM. The combined extracts were washed with 50mL water Picture 6, and 50mL brine Picture 7, before been dried over Na2SO4 and placed in the freezer.

The solvent will be removed tomorow.

Picture 6

Picture 7

Final extract


Note 1 : The initially clear yellow freebase had turned orange when kept over night at room temp.


This reaction is very smooth and easy to perform. I'm sure the H2O2 addition can be done more quickly (they add it over 4h in the patent). The demethydioxirane oxidation looks very interesting by the very short reaction time and easy workup, but preparing pur DMD in acetone is tedious. I'm sure adding oxone to acetone and using the suspension directly could work out nicely, filtering before evaporating the solvent.

I'm not too sure how to purify the product, i don't want to do a coluum, being out of chromatographic silica. Apparently the mp is pretty low (~30°C), so recrystallization would be tedious. Any ideas?

Can't wait to try the oxidations out! Even though i have too :)

[Edited on 6-4-2008 by Klute]

guy - 5-4-2008 at 16:22

Wow Klute. I want to try this reaction again. VERY Nice work.

If you want TEMPO, you should try a Wolf-Kishner reduction on your triacetone amine first.

[Edited on 4/5/2008 by guy]

Also, try adding hexanes to your oil and see if solids crystallize out. Ive done stuff in the lab where I get an oil but if I add hexanes, it crashes out.

[Edited on 4/5/2008 by guy]

Here is an IR spectrum from Aldrich in case you wanted to confirm (4-oxo-TEMPO)

[Edited on 4/5/2008 by guy]

[Edited on 4/5/2008 by guy]

Klute - 5-4-2008 at 16:47

Thanks alot of the spectrum. I will compare when i can.
I don't think i would want to try a wolf kishner on it, needing hydrazine and all, i'm pretty sure it will be as efficient as tempo. In any case, a NaBH4 reduction would give 4-hydroxy-TEMPO, often mentionned in patents/articles.

I would definatively encourage you to try it again! Let the 4-oxo-TEMPO spread :)

not_important - 6-4-2008 at 08:07

For a reduction you might do something similar to this:

having a free amine group might lead to problems and polymeric goop, though. Running it in Ac2O should get around that, but Ac2O is a problem in itself.

Klute - 6-4-2008 at 14:27

I would rather use NaBH4, as it seems to be a prety easy reduction; in one of the above patents, they obtain the hydroxy derivative with NaBH4. I guess reduction in methanol is all it takes.
But for alkylating/acylating the hydroxy group without touching the amine, i have no idea. At first, i was thinking of oxydizing the hydroxy derivative to the nitrosyl, and then alkylating with NaH/Mei or EtBr, but i guess the nitrosyl will be affected by those conditions.
How about forming the 4-oxo tosylate, reducing the C=O with NaBH4, alkylating the hydroxy, and basic hydrolysis of the sulfoamide before oxydation? I'm not too sure of the conditions to hydrolyse such a sulfoamide.
In the same way, other alkoxy- and acetoxy- (with Ac2O, I have enough at hand to make a with trials) could be obtained.

It would be nice to compare the efficienty of each catalyst towards oxydation. I think i will be using TCCA as co-oxidant, as it is more stable than NaOCl and would help getting reproducible results. There alot of different reaction conditions appliable here (two-phase, PTC or not for electro-donating cycles, one phase, PTC or not, etc etc) with either TCCA and other chloramines, hypochlorites, etc

Things would surely go much quicker if several people tried this out :) The 4-oxo-TEMPO qsynthesis can take a weekend's work at most if using NH3 gas, or a afternoon if mixing aq ammonia and acetone and forgetting it for a few weeks, counting the distn and oxidation. I means, how much more available can it get? Ammonia and acetone :D !

EDIT: Sorry notimportant, i thought you were talking of the reduction to the hydroxy compound.. As soon as i saw non-amalgamated zinc, i thought of a simple reduction... I might try it out when i do more triacetoneamine. I'm sure it's (TAA) synthesis can be greatly improved..

[Edited on 6-4-2008 by Klute]

Nicodem - 7-4-2008 at 02:51

I doubt the Org. Synth. modification of the Clemmensen reduction works on such substrates, if nothing else I doubt the triacetoneamine hydrochloride is soluble enough in any such aprotic nonpolar solvent system.

For the O-acetyl derivative, I think you could easily prepare it if you make the hydrochloride of 4-hydroxy-2,2,6,6-tetramethylpiperidine and just let it stir in excess acetic anhydride for a while (some heating to ~50°C might be needed). There should be no N-acetylation occurring since N-protonation is generally protection enough.

The O-methylation would best be done on dry 4-hydroxy-2,2,6,6-tetramethylpiperidine base in THF by O-deprotonation with NaH followed by MeI addition at room temperature. There might be some N-methylated side products, but probably irrelevantly little since the nitrogen is sterically hindered. You might be more interested in O-benzylation since this would allow you to follow the alkylation and the N-oxidation reaction by TLC (you would have UV absorbing products).

I would not consider the O-acetylation and O-alkylations of 4-OH-TEMPO since it is a free radical after all and its chemistry might not be so easy.

But first try some alcohol oxidations with plain 4-oxo-TEMPO since I believe in most cases it is similarly effective as TEMPO itself.

Klute - 7-4-2008 at 09:36

It's good to the the acylation can be performed easily. Any acetamide could be easily seperated with a A/B extraction. I could try direct comparaison of the oxo- and acetoxy- catalyst before trying other modifications.

Concerning the oxydations, I'm still hesitating as to the reaction conditions to use. I'm leaning towards using TCCA as primary oxidant as NaOCl solution will have to be titrated or freshly prepared to get any reproducible results. I think DCM 2-phases or with a few drops of water in the cold will be tried first, with 20-50mmol batch of benzyl alcohol. When i get good results, i will try substitued alcohols.
Apparently, electron-donating substituants gives lower yield with 2 phase NaOCl oxidations, so some optmization might be needed with such substituants.

I haven't found any ref with benzoxy-TEMPO used as catalyst, but it could be worth a try latter on. I would like to find a way of protecting the amine though.

Klute - 12-4-2008 at 12:22

After removal of the solvant, 19.4g of a dark brown/black oil is obtained.

TLC (DCM) of the TAA (first, less pur fraction) reveals at least 3 impurities. I am unable to locate the TAA spot...

The oxidation extract shows that most of the stains haven't changed, only the largest one present in the TAA is much smaller, and one above is slightly bigger. I'm not sure if the nitrosyl can be revealed under UV lamp, though i can't find a reason why it would not . There also is a none-eluted brown spot.

I figured out i shouldn't have acidified with HCl, as apparently at low pH (<2) the TEMPO nitrosyl can disproportionate to a hydroxylamine and a nitrosonium salt. But i didn't add too much acid, and the fact that i recovered 19.4g product means there wasn't much left in the aq layer.

I will try a alcohol oxidation with the crude product i obtained, and one without to compare. Or beging the oxidation without the catalyst, monitering by TLC, and then add it and see how things change.

I would like to find a way of purifying the TAA, i don't think a A/B could be used here considering how soluble the freebase is in water. Forming a salt isn't easy, i haven't tried Nicodem's proposition yet, but adding dry H2SO4 in EtOH to the amine in either IPA or acetone, or conc HCl to the amine in IPA followed by pet ether (which usually works well) only gave a oil or noting at all, even after freezing.

BTW, the first, less pur fraction of the freebase doesn't solidify in the freezer. I think gassing dry NH3 in acetone could give a cleaner product. Alot of side reactions could have happened during 6 months. Also, the tubing that i use to connect the aspirator to the disnt setup finished BLACK from all the smoke that passed through.

EDIT: I haven't been able to do a IR spectrum on it, and won't be able to do so for another 2 weeks, so i will surely try the oxidations before.

[Edited on 12-4-2008 by Klute]

LSD25 - 13-4-2008 at 07:47

Klute, I've said it before and I'll say it again, your glassware makes me jealous:mad:

So, you suggest adding dry ammonia gas to acetone? Ok, how much ammonia am I adding to how much acetone? How well does acetone hold ammonia?

By the way, this article might be on topic if anyone has access (I'd ask in the right place, but I cannot get there to do so):

Preparation of Triacetoneamine (4-Oxo-2,2,6,6-tetramethylpiperidine), an Improved Method

Synthesis 1976; 1976: 735-736 (Communication) DOI: 10.1055/s-1976-24178


The Abstract is here:

I have also attached an article retrieved by Solo on the Wolff-Kishner on triacetoneamine to tempo on a rather large scale.

Attachment: (99kB)
This file has been downloaded 2923 times

Klute - 13-4-2008 at 11:41

Thanks for the articles! I'm requesting the first one right now.

Concerning the wolf-kishner, i can't perform such a reaction as i haven't got any hydrazine, and don't want to acquire some (toxicity). On the second hand, i'm convinced 4-oxo-TEMPO and 4-hydroxy-TEMPO can lead to catalysts as efficient as TEMPO itself. They can be modified in a large number of ways (reductive amination, alkylation, acylation, etc). I'm sure we could get some nice @SM publications out of these, rather than walking on the paths of all the chemist that have been using TEMPO for the last decade... :)

The procedure using dry NH3 gas is detailed in one of the patents given in this thread. You will find all the information needed there.
I plan on trying it out also, as i'm not satisfied by the purity of the product i isolated. I will post details when i get to perform the reaction.

Please share your results if you try the reaction!

EDIT: Concerning the glassware, i've damadged 10 of those red screw rings you can see at each joint, left them in the base bath for too long (couldn't get them out myself because of my hand), by the time someone was kind enough to get them out (not a nice task i admit), the more than 1-2mm of plastic had been eaten ! I'm not sure what kind of plastic it is, but when adding acid to the rinsing water, fluffy white amorphous solid appeared.... there some of this plastifiant in the base bath too, but if the glassware is directly rinced without adding any acid, it doesn't cause any problem... I guess it some kind of acid monomer (as terephtalic acid or something...). Going to have to buyt some more.Hopefully the teflon joints are not harmed. I guess it made me mearn a lesson: always take them off before puttingthe glassware in the base bath! Usually they handle a few days no problem, though i try to never leave the glassware in for more than 24h.


Sparky very kindly retreived the above article on TAA synthesis. They gas dry NH3 into acetone/CaCl2, at room temp or just under reflux, but do so over 3-4 days! The whole reaction takes a week, counting a few days of stirring after introduction of the NH3. I might try doing it over a shorter period, perhaps 2 days, as they claim no more than 1/4 of the total NH3 could be gassed in during the first day. They claim the crude product has a 85% purity, and that distn affords a product of ~95% purity, which would be more than enough for the oxidation.

Actually, i think i would rather try the procedure followed in US3,959,295, where a suspension of NH4Cl in acetone is gassed with NH3 over 12H at ~15°C, then more acetone is added and reflux maintained for 12H. The way they isolate the hydrate seems very interesting, provided it works.

[Edited on 14-4-2008 by Klute]

LSD25 - 5-5-2008 at 03:36

I am currently trying this reaction with OTC acetone which has been gassed with ammonia in the presence of CaCl2 (I know it seems kinda odd, but have a look online - it seems to be well supported, so we'll see) kinda similarly to the idea in this paper (although I only have access to the front page:

It has been in the cupboard for about a week and all I can see is a clear liquid with some undissolved CaCl2 and a fine, fluffy precipate (also whitish - although it doesn't look completely white). I do however know that the characteristic acetone smell has changed (I cannot smell either it or ammonia but a sharper smell which is kinda hard to describe). I didn't make any effort to dry either the acetone or the ammonia (it seemed kinda dumb in light of what I was using as a catalyst) so the reaction should (going on previous experience) be a long way from anhydrous. I'll report back again in a week or two.

Klute - 5-5-2008 at 12:26

Nice Alice! Good to see other people trying this out! I have put the project aside for the moment, I realized my TAA wasn't more than 50% pur when i isolated the hydrochloride salt.. I want to try gassing over one or two days and distn straight after.

In my first try, it took at least two weeks before the color change was significant, so i wouldn't be worried..

Keep us updated!

LSD25 - 5-5-2008 at 22:01

I actually just looked and while the quantity of the precipate has increased markedly, there has also been a marked colour change to a dull, pale-yellowish hue.

LSD25 - 9-5-2008 at 21:14

Looked in again, the liquid is now a definate clear-yellow type colour - the smell is odd as, a sort of acetone mixed with urine type odour, although the acetone aspect is merely colouring the other smell. I think the absence of water (well the CaCl2 must be doing something, right?) seems to be making this work nicely - there is only the slightest tinge of orange. I'll leave it another week then fractionate it (actually, Klute - you think steam distillation might work better?).


Just went to add more NH3 - must have mistreated my shit-glass (the stuff I use for gas generation - has a wheel on it) - it cracked and sent caustic/MAP sludge everywhere - when I tried to do what I could to prevent spillage I noticed some water/NaOH going into the bottle this was in - the yellow colour was dimmed fairly rapidly, the solution is cloudy - but I have enough cleaning up to do now, I'll just chuck it in the cupboard and see how it looks (I really should start over, but yeah:().

[Edit No.2 - attached crap photo of filtered liquid (my camera is fucking up, time to get a new one)]

[Edited on 10-5-2008 by LSD25]

[Edited on 10-5-2008 by LSD25]

[Edited on 10-5-2008 by LSD25]

oxotempo.JPG - 32kB

Klute - 10-5-2008 at 01:45

You could try working it up now; in one patent and the "improved synthesis" article, they work up 48h after gassing. One way of extracting it is to add NaOH solution, seperate the oil, and extract the aqueous. Remove your solvent and fractionnate.
I'm not sure if some add water would be good for the reaction, and the NaOH could cause over-condensationn products.

Keep us updated!

LSD25 - 10-5-2008 at 04:42

Hang on a minute - the CaCl2 should have at least dried the acetone, right?

Now I want to extract an amine from dry acetone?

What is the solubility of the HCl salt in acetone?


Not fucking much by the look of things - I just set up a HCl gas gen (CaCl2 + HCl + heat + drying train) - then gassed 1/2 the yellow liquid remaining (I diluted it with additional dry acetone).

As seen in the photo - a reddish-grey, seriously cloudy solution resulted - I'll let this settle overnight and then filter off the solvent in the morning

[Edited on 10-5-2008 by LSD25]

Sorry Nicodem, I just now re-read this page - I guess that is what you were hinting at further up the page, huh?

[Edited on 10-5-2008 by LSD25]

oxotemp.hcl.JPG - 22kB

LSD25 - 11-5-2008 at 17:43

I aded in the other half of the TAA (Triacetone Amine) and gassed it - unfortunately it was mother's day here yesterday and obvious family commitments took up rather a lot more time than I had anticipated.

Thus the light-pinkish grey liquid was left for a day - now there are two things to report here, the growth of a cloudy, crystaline layer in the bottle (and a pinkish crystaline layer in the bottom of the bottle) which says to me that the TAA is at least slightly soluble in Acetone.

I took it and carefully (no need as you will see) decanted the pinkish-reddish clear liquid off the crystal layer, which did not move then heated it carefully in a simple distillation setup. A small amount of excess HCl (g) came over first then the pot rapidly heated to about 56C and then stayed there for the first 5/6 of the distillation. A small amount of high-boiling tar appeared as did a rather large amount of crystals (especially in the last 100mL of the distillation). The distillate to this point has come over exclusively between 56-58C and is a clear, volatile liquid with a distinctive odour (rather different to the odour of the crap I put in the pot in fact).

Anyone got any idea what these crystals are able to be washed in? They are well and truly soluble in water, so I was thinking that I'd separate from the oily-tar layer in a sep. funnel and then dry - recrystalize from EtOH/EtOAc? The yields are crud on this scale, I think that I should have added 2-3 more lots of ammonia, but given the fuckup detailed above, working up was about all that could be done.

[EDIT] I am going to have to try that one again - I think I might have to attempt something else - the oily crap has one hell of a pong which is why it was ditched (when diluted with water the oily stuff solidifies into polymeric crap which is kinda soapy), unfortunately I tried extracting without filtration cos I'm daft - the oily shit then dissolved and stunk out every single attempt to do anything thereafter (it appears to exist equally in both the polar & non-polar layers and it bumps badly when you heat both, even with chips).:mad:

Unfortunately without boiling the shit dry it appears to be virtually impossible to rid it of some remnant of acetone. I think I might just boil the next lot dry and we'll see if the crap can be removed. Of course, if the gassed (HCl) liquid was extracted promptly, then I suspect the problem would be much lighter duty.

Fuck it - I think it might be easier to go the fractional distillation route.

[Edited on 11-5-2008 by LSD25]

LSD25 - 14-5-2008 at 04:13

Right, the recovered acetone was gassed again, in this instance three times. Gassing was continued in each case until a white precipate was formed when additional ammonia gas was introduced and a correspondent large amount of white smoke was formed when any bubbles made it to the top of the liquid (if done carefully, rarely will bubbles ever make it to the top). The white precipate I am assuming appears to be a complex between whatever dissolved CaCl2-H2O was present in the acetone and its subsequent reaction with excess ammonia (some H2O is generated in the amination is it not? I know some was there to start with).

This has been done on each of the last four days, immediately after the gassing of the clear, light yellow, yellow, dark yellow* (day 1, day 2, day 3 and today) liquid was stirred at med-high speed for one hour.

*Klute - the liquid is beginning to look a whole lot like your distillate, but until I get the money to spring for some TLC plates, it is hard to know. Workup will probably take place tomorrow - I'll gas it one more time first.

Klute - 14-5-2008 at 09:51

Pet ether works great for washing the TAA.HCl. I recrysatllized it with IPA/Pet ether, or even IPA/Acetone, finishing in the freezer.

In one of the patents on the first or second page, they directly add either conc. H2SO4 or dry HCl gas to isolate the amine from the reaction medium. You need to do it in the cold though, to avoid excessive aldol condensations and leave it to stir/sit for a few hours to complete cristn.

I will trying out the reaction again in a month or so, so we can compare how things go when using different procedures/proportions. It always much more appreciable to be several working on the same project :)

LSD25 - 15-5-2008 at 16:12

I have just spent 2 1/2 hours gassing the crap out of the yellow liquid while it was still cold - there was no colour change or anything else. When placed back in the freezer all that happened was the constant evolution of HCl gas, there is no change in the appearance of the solution. I added some additional dry acetone and gassed again - nothing (more HCl(g) evolved when cooled - obviously solubility related). Obviously the salt is highly soluble in acetone (I'd say the lack of solubility in the non-cooled solution may be related to the solubility of the salt in the mixture of acetone and the aldol condensation products).

Anyone got any suggestions?


Sorry to recant - slight, ultra-fine crystals are apparent (only when mixture is agitated and can only be viewed against strong backlighting) in the yellow solution. Hmmm, I think more time will allow these to come out - oh yeah, I just scratched fuck out of the sides with an old stir-rod, these suckers are sorta reticent but they've started.:D

[Edited on 15-5-2008 by LSD25]

[Edit 2]

I lost patience and simply boiled off the acetone - the yields are crap - seems a lot of the amine didn't come into contact with the HCl (had to boil off the yellow oil 'cos I wasn't prepared to fuck around this time around;)). In the end I got a minute yield of a beige salt, with a small amount of black tar. Unfortunately the 'PYREX' (supposedly) baking dish decided to emit a wee crack and then turn into shrapnel. I collected as much as I could, then started extracting that with 95(ish)% EtOH - the black tar is not seeming to come through the large funnel, while the salt is dissolving and turning the resultant solution yellow as I write. I'll pour the filtrate back over the shattered glass a couple more times and then discard the glass. I'll then remove the EtOH and see how I go (I'll simply reduce it as much as I can via distillation, then crystallise).

[Edited on 16-5-2008 by LSD25]

Klute - 16-5-2008 at 04:38

To crysatllize my oil, I proceeded as Nicodem advised me to: I diluted the oil with some IPA, neutralized with conc. HCl, then distilled 2/3 of the IPA off, until the BP increased over that of the azeotrope. I then left it to cool, at ~40°C I slowly added ~10mL of pet ether as one layer (on the sides of the flask), and left evrythign over night. A tane/pinkish powder had appeared the next moring, covered with a little tar. I vac. filtered it, washed with cold acetone, which remove dthe tar evry well without dissolving the powder (visibly) then washed with pet ether, and dried by suction for 10min.
The very bad yield confirm that the oil I collected wasn't more than 30% TAA, which i feared after seeing the composition by TLC. That's why I had hope the short method with dry NH3 gas would lead to a much more purer product. If that isn't the case, we will just need to do this big scame to obtain a decent amount of product to work with. The oxidation seem clean enough if done with a clean product, whan iwill try this I will purify the 4-oxo-TEMPO via coluum the first time, and see how the pur the crude product is. I'm pretty sure it will be possibly to use the crude product as is, but at least there will be a pur sample to compare to. ALso, whenni have the possibility, I will run an IR on the crude and pur product, and compare with the spectrum Guy kindly provided....

Why don't you try removing most of the acetone after decanting from the CaCl2, then adding dilute NaOH, seperating the oil and cristallizing it with IPA/HCl? I know it's longer work...
Or maybe try with conc. H2So4, possibly the sulfate is less soluble in acetone than the hydrochloride, or the salt forms easier..

Don't disesper Alice, I'll soon join you again!

LSD25 - 16-5-2008 at 05:42

Yeah, this shit is cursed - I have been breaking glassware like it's free (problem being it fucking ain't) every FUCKING time. I 'think' the crude product from dry NH3 gas is not too bad purity wise - it doesn't seem anywhere near as prone to autooxidation and unusual products for a start. What suprised the fuck out of me is the sheer difficulty encountered gassing the amine - I passed fucking shitloads of dry HCl(g) into this crap, while cold, there was nothing - no colour change, no precipitation, nothing, just clouds of HCl (which is right up there in my list of things I don't like). After doing this, I then tried heating the shit to see if I could get a crystaline product - well, something, but each time it was heated more and more HCl gas was given off (obviously held by the Acetone, as one would expect - but if so, why the fuck didn't it react with the amine?).

I was thinking of trying H3PO4 (I have serious problems getting access to conc. H2SO4 (I think it was thrown on peoples faces here once) - I'll make it nice and fresh from polyphosphoric acid) - there is a couple of interesting things to report - the contaminants when I washed the broken glass with EtOH (95%) precipitated when I washed the glass with water in the funnel (obviously not too soluble in less than 90% EtOH), leaving a very pale yellow solution - which I have yet to process.

Seriously, I am really getting the shits with this stuff. How much do you need? Isn't it usually used in catalytic quantities only?

PS When heating the Actone/TAA/TAA.HCl solution, there was the separation of a decent amount of yellow oil, as this was heated further some of it evaporated and some changed to a reddish oil which finally went black and tarry (I'm guessing this is the n-oxide? which suggests autooxidation).

One thing I was thinking of, is the 2,2,6,6-tetramethylpyridine more stable than TAA? If so, save trying to isolate the sucker, how about adding some phosphonic acid and some iodine and just isolate the basic amine at the end? Stripping off the excess acetone first would greatly assist in this endeavour, but I think HI is probably going to be more than enough to desoxygenate the 4-keto position?

[Edited on 16-5-2008 by LSD25]

Klute - 22-5-2008 at 09:11

Considering the impurities present with the TAA direct oxydation of the crude TAA gives a very dirty product. even the oxydation on the distilled freebase gave a very complex mixture of products. But there no evidence this would prevent the 4-oxo-TEMPO present from doing it's job, but you will introduce alot of impurities.

I find it better, when trying out a somewhat new procedure, to get, at least at first, a pur product, and to be sure the problems arising are not du to impurities present. Using a crude product poses to much variables to be sure it works or not.
But once pur 4-oxo-TEMPO has been isolated, and proved to be efficient in catalysing oxidations, the crude product could be tried, if similar results are obtained, it might not be worth the effort of isolating pur intermediates/catalyst. But for now, let's do it cleanly :D

Synthesis of 4-oxo-TEMPO according to US 3,959,295

first step:

In a 250mL 4-neck RBF, equipped with a gas inlet, a condenser, a thermometer and magnetic stirring, 25mL (340 mmol) of acetone and 9mL of MeOh were charged, followed by 1.1g (20.56 mmol) of freshly fused NH4Cl.

A 500mL erlenmeyer was charged with ~100g of humid, crude NH4Cl (byproduct from hexamine hydrolysis), containing roughly 80g (1.5 mol) of NH4Cl, and a stir bar. The erlenmeyer was fitted with a 250mL addition funnel, containing 150mL of 40% w/v NaOH solution (1.5 mol). A gas outlet was connected to the top of the addition funnel, linked to a wash bottle containing solid NaOH pebbles, itself connected to the gas inlet of the RBF.

The NaOH solution was slowly dripped onto the NH4Cl, generating a constant, easily monitered flow of NH3. the acetone/MeOH was gassed at 1bubbles/sec, with vigorous stirring, at room temp (17°C). A latex glove was attached to the top of the condenser, keeping in any unabosrbed NH3. It was not "blown up".

Gassing was continued for 3H, until the mixture was saturated. Near the end, the temp rosed over 30°C, so an ice bath was applied to keep temp under 15°C. Once saturation was definatively attained, the gas inlet was removed.

Second step

75ml (1.02 mol) of acetone are then added, turning the whole mixture milky white as a fine solid precipitates, and agglomerates upon stirring.

The flask is then heated to 50°c with an oil bath. After a few minutes, the solid dissolves completly, offering a water-clear solution. After1 hours stirring, the mixtures takes on a yellow colour, and afetr 15H a dark red solution is obtained.

The condenser is replaced by a fractionnal distn setup, with a 150mm vigreux, and the gas tube connect to the vacuum inlet and immersed in a dilute H2So4 solution. Heating is increased, and the excess acetone removed. After ~75mL are distd, slight vacuum is applied.

A dark red, viscous residu is obtained. It fumes when vacuum is disconnceted.

2.25mL (125 mmol) of water are added to the cooled residu, and the flask is immersed in a ice water bath with slow stirring. After 1 hour, there still isn't any solid forming. apparently the product is too impure to be able too form a hydrate directly.

The residu is then titrated with conc. HCl. Over 20mL are required until acidic to dampened pH-paper. 100mL of IPA are added, and distn under vacuum at 60°C. this is renewed with 50mL IPA. Unfortunaly, the distn was left unattended for too long, and the residu heated up considerable when the IPA was completly removed, but not for long.

The thick black/red syrup was cooled down, thickening alot, and 50mL toluene added. A semi-solid tarry goop is formed, which doesn't solidify when a little sample is triturated under acetone, although it looks likes a cristalline paste, it stays liquid.
30mL acetone are added to the two layers, the solvent top layer taking ona red color, without much effect on the tar. 30mL of IPA are added, which dissolves most of the tar, forming a black/brown homogeneous layer, and a white/yellow solid sperates out immediatly. This is left to sit for 30min, then vacuum filtered. A yellowish powder is obtained, which is generously washed with acetone, and pet ether before being dried by suction, giving a very light, fluffy beige poxder.

A total of 10.7g (55.87 mmol) of triacetoneamine hydrochloride is obtained. The solid will be recrystallized with the old sample, from IPA/acetone, or IPA/pet ether.


The workup is the problem here. The cristallization is far from optimal. The filtrate has been placed in the freezer to see if more product will cristallize, but next time I will rather try adding conc. H2SO4 to the reaction medium directly as advised by the patent, or adding ethanolic oxalic acid to the residu, as advised in some of the articles KMnO4 very kindly posted.
The oxalate salt seems pretty easily isolated, and fairly stable.

It's a pity the hydrate didn't readibly form, this would definatively be th ebest way of stocking TAA for short to medium periods. No need of basifying before oxydation. I guess it could be kept in the fridge for longer stocking time, but as it is commercially avaible under this form, I guess it's pretty stable.

In any case, this seems to be a good reaction: no need of gassing for several consecutive days, waiting a week, etc. Apparently the methanol does a good job at dissolving alto of NH3, as pur acetone doesn't dissolve much by itself, hence the consecutive gassings in sevral articles. The gassing is pretty staright forward and doesn't require constant presence once stable. Just gas the shit out of it until saturated :). I think a little less acetone could be added in the second step, I followed the second series of examples in the patent, but the first one used 1/3 the amount of acetone in the second step.

Considering the ease of the reaction it would be adviseable to do it at a bigger scale. If the TAA can be directly isolated from the reaction medium by conc. H2SO4 as claimed by the patent, no need of distilling the excess acetone, which saves quite some time. The ~8.6g of freebase should yield more than enough 4-oxo-TEMPO, but I might just try another reaction with using either H2SO4 or oxalic acid to isolate the amine.
Once thta's done, oxidation time!

Come people, jump in! This is easy and rather a break through in home-chemistry: a 100$/gram catalyst ( !! ) easily synthesized from scratch!

The second part, checking out the efficienty and range of use of this catalyst is going to be a pretty large project too.... Obviously, i will be starting with benzylic alcohols to aldehdyes oxidations, becasue that's why I decided on try to make the catalyst in the first place, but it can also be used for aliphatic aldehdyes, nitriles from amines, etc

Here is the patent for the lazy people :P :

[Edited on 23-5-2008 by Klute]

Attachment: US 3959295.pdf (264kB)
This file has been downloaded 982 times

detritus - 22-5-2008 at 11:04

i did see one patent that did reactions on the bisulfite addition product of TAA. i will try to dig it up. but it gives some confidence that it would actually form an addition product

kmno4 - 23-5-2008 at 06:11

Nice... I have always wanted to do this, but I am too lazy :P
Are you able to check up melting point of your product, to confirm (let's say) it is indeed triacetonamine hydrochloride ? Aldrich gives 198 °C.

Klute - 23-5-2008 at 06:14

I will do so after recrysatllization, I think there must be a little diacetoneamine and other uncyclized amines in there too.

Synthesis of the acetonine nitroxide radical

chemoleo - 23-5-2008 at 19:05

Will be of interest to most here...

Title: Synthesis of acetonine nitroxide radical and 1-hydroxy-2,2,4,6,6-pentamethyl-1,2,5,6-tetrahydropyrimidine

Tetrahedron Letters 41 (2000) 179–181

Attachment: acetonine nitroxide radical TEMPO.pdf (76kB)
This file has been downloaded 1796 times

Klute - 24-5-2008 at 06:36

Thank you very much Chemoleo! Indeed, it is pretty interesting!
That finally gives us an idea of the product isolated when the recation is done at room temp (its a pity they don't give the physical properties of the acetonine, surely a yellow oil?),b ut the fact it also forms a satble nitrosyl radical means that the crude distillate I oxidized a month ago can act as a catalyst even though it doesn't contain only 4-oxo-TEMPO.
As they claim the acetonine form TAA upon reaction with acetone and ammonium/calcium chloride, I guess this si what happens in the second step of the reaction.

But it's pretty hard to say what exactly is produced when reaction acetone and ammonia at room temp: some say high yileds of diacetoneamine are produced, they say the aectonine is produced in high yields... Surely a complex mixture of boths (and others). The fact that DAA and this acetonine both form salts would suggest that the less recent preparations of DAA would give a product contaminated with the acetonine.
So i guess preparation involving a rather long reflux with excess acetone would be the more appropriate to obtain TAA with the less contamination possible, and adding a stronger lewis acid in the second stage might be beneficial (zinc chloride seems like a good suggestion from the article).

If using the catalyst straight as is, minor contamination with acetonine radical shouldn't be a problem, but it will when trying to derivate the 4-oxo-TEMPO to a solid catalyst.

For the next preparation of TAA, i will try out more elaborate conditions: distilling the acetone over CaCl2 or similar, using recrystallized, fused NH4CL as catalyst, and a more pur grade of NH4Cl to generate the ammonia.

I'm going to request the preparation of TAA referenced in this article, that 90% seems interesting and might containa detail that could greatly help.

EDIT: possible bad news: in patent US6608220, bisnoralcohol is converted to bisnoraldehdye by using NaOCl with catalytic Br- buffered by bicarbonate, and a derivative of TEMPO. They use 4-hydroxy-TEMPO in most case, but give no yields, conversion or selectivity, just "the compound is obtained".

On the other hand, example 6 is more worrying:


Bisnoralcohol (I) to Bisnoraldehyde (II) with 4-oxo-TEMPO

A mixture of bisnoralcohol (I, 6.6 g), 4-oxo-2,2,6,6-tetramethylpiperidine-1-oxyl (18 mg), dichloromethane (30 ml), sodium bicarbonate (180 mg), potassium bromide (238 mg) and water (5 ml) is cooled to 1°. Then aqueous sodium hypochlorite (14.6%, 11.4 ml) is added to the mixture over a 15 min period. The reaction produced the title compound but in only a 7% conversion of bisnoralcohol with 58% selectivity for bisnoraldehyde.

But this odesn't mean 4-oxo-TEMPO is much less efective in catalyzing oxydation of primary alcohols to aldehydes:
First, they add the hypochlorite in 15 in that example, where as they add over 5 h in the other,
Secondly, they oxidize a pretty large and delicate molecule.

The only difference I can think of between the oxo and hydroxy-TEMPO is that the oxonium is slight more stable with the oxo- than with the hydroxy-, though seperated from the nitrogen by two more carbons... But obviosuly there could be lots of other things to consider.

I intend on preparing 4-hydroxy-TEMPO in any case, and compare it efficienty with the oxo-, and also try making methox-TEMPO and ethoxy-TEMPO, as apparently the hydroxy-TEMPO is very easily alkylated.

EDIT 2 :

I really owe KMnO4 : he posted a series of articles on TEMPO derivatives in the ref forum, including a very complete review on hindered piperidines, which really give more perspective of modification of 4-oxo-TEMPO, there are litteraly tons of way of modifying TAA in different ways..

He aslo joined a excellent article detailing the oxidation of TAA, with H2O2/WO4 2- and H2O2/CO3 2-, indentifying all the impurities and their caracteristics (even the Rf's! :) ) and presenting a remarkable workup for the H2O2/CO3 2- (which is much cleaner than the WO4 2-! I wonder why they even want to use versanate and tungsate salts when carbonate works better?! ), that I am attaching because it's so usefull.

Here is the workup:


TAA oxidation in the H20,/Na2C03 system (the modification of the procedure

TAA (20 g, 0.129 mol) is added with stirring and external cooling (cold water) to a solution of anhydrous sodium carbonate (41.7g, 0.393 mol) and 300,:, hydrogen peroxide (200 g, 180 cm5, 1.765 mol). The stirring is continued for two days. Sodium chloride (about 60 g) is added, and the mixture is left overnight in a refrigerator. Red crystals (16.6 g) are filtered off and dried over silica gel in a vacuum dessicator. The crystals are dissolved in 600cm3 of n-hexane and the residue of sodium chloride is filtered off. After evaporation of the solvent, 16.0g (73%) of TAAO (m.p. 35-37°C) is obtained. The impurities 2-4 are not detected (TLC).
The elementar analysis: calculated for C9H,,N02 C 63.50, H 9.47. K 8.23; found C 63.60, H 9.6'5, N 8.22.
The m.s. is the same as in [I.].

I figured out I could simply use the HCl salt in conc. aq solution, and use 1 more eq of Na2CO3, to directly freebase the TAA s it is added, the NaCl produced won't be problematic as more is added after.. I am very happily surprised that the 4-oxo-TEMPO seperates out as crystals so easily! that saves alot of purification steps! But I guess very pur TAA should be used, which involves at least 2 recristn.

Unfortunaly, the silica/SO2Cl2 article showed that this preparation doesn't give a sulfonly chloride functionlized silica, but a sulfate ligand, with two Si-O-S bonds, so no free acid to attach a TEMPO to... Both S-Cl bonds react with a -OH group.
So chlorosulfonic acid is the way to go, but I don't think i will trying this until I get a few derivatives of 4-oxo-TEMPO.

One good news from the rview is that the nitrosyl isn't affected by NaBH4 reduction! That means 4-Oxo-TEMPO can directly be reduced to 4-OH-TEMPO. It might be wiser to start off reducing TAA, and comparing the obtaiend products to see which is the purest. I haveyet to find a procedure detailing the reaction conditions of such a reduction,, I'm not sure fi it would be better to to a usually or reversed addition. I was thinking of adding the TAA in AcOEt to an excess of NaBH4 in MeOH or EtOH.. If anyone know of such a procedure (i checked for reductions of piperidones in general, an didn't find anything specific).

4-amino-TEMPO is easily prepared from 4-oxo-TEMPO: reaction of 4-oxo-TEMPO with hydroxylamine, and reduction of the oxime, and the review also explicites that Wolf-Kishner reduction of 4-oxo-TEMPO proceeds easily to TEMPO, if anyone wants to mess with hydrazine...

Lots of things in perspective :)

[Edited on 25-5-2008 by Klute]

[Edited on 25-5-2008 by Klute]

Attachment: fulltext1.pdf (233kB)
This file has been downloaded 1091 times

Klute - 26-5-2008 at 13:12

I would like to add that MeOH is a excellent recrysatllization solvent for TAA.HCl, a dual solvent with MeOH/acetone yields nice small needles, colorless to yellow. Pictures will come tomorow.

IPA seems to dissolves only very little product even at boiling temperatures:

10.7g of TAA.HCl were placed in a 250mL erlenmeyer, and 50mL IPA added in portions and heated to boiling on a hot plate. after 10min, hardly any beige poxder had dissolved, although the IPA had takena slightly yellow colour from impurities (an IPA boil could be a good idea to purify crude TAA.HCl ).
MeOH was then gradually added. 13mL were added until complete dissolution, giving a ratehr dark orange solution. Thsi was left to cool, and seeded with a grain of crude TAA.HCl, which initaite cristn immediatly. Small needles started to form.
After 30min, 20mL acetone:pet ether 2:1 was very slowly added from the sides of the erlen, causing cristn on the surface at first. The erlen was left to rest for another 30min, covered, then sealed and placed in the fridge for 3H, then in the freezer overnight.
The crystals will be filtered tomorow; and rinced with a little cold acetone:pet ether.

Bolt - 27-5-2008 at 10:38

Klute, thank you for your dedication; hard work; and illustrated, detailed writeups on this project!

Klute - 27-5-2008 at 10:46

My pleasure Bolt! :) This project is really catching me on! I will not rest until TEMPO catalysts are found in every home-lab :D

Oxydation of triacetoneamine with H2O2/K2CO3

This is a small variation of the methode described in the article KMnO4 retreived lately, that i attached just above.

A made a few modifications, and could have done things a little better; but at least know I know what I will do next time :)

In the procedure, they use anhydrous Na2CO3. At home, I had the choice between very technical (hardware garde) hydrated Na2CO3 (unknown hydratation, less than decahydrate, and of questionnable purity, possibly containing soaps/detergents), and anhydrous 98% K2CO3, so I prefered using K2CO3. It should be noted that in most procedures using H2O2/CO3 2-, the sodium counter ion is used. In the patents, they claim potassium carbonate can used, and a example gives a very high selectivity, so it shouldn't be a problem.


In a 250mL wide neck erlenmeyer, 21.8g (157.73 mmol, ~3.4eq) were dissolved in 64.35mL (631.02 mmol; 13.7 eq), 30% H2O2. This caused some warming up and fizzing. The milky solution was cooled in a ice bath with slow magnetic stirring.

8.83g (46.12 mmol) of recrystallized TAA.HCl were crushed into a finer powder, and added in portions to the solution. AT first, nothing much seemed to happen, no CO2 evolution, the powder jsut formed a suspension.

1.88g (47,00 mmol) of Naoh were weighed and added as a solid in small portions to the suspension. The solids seemed to increase in volume, and whiten. A thick slurry of a milky white fluffy solid was formed: the TAA formed a hydrate as soon as it was based. As O2 bubbles were evolved, the voluminous solid formed a foam, hard to stir in surface. It was broken down regularily with a thermometer, to facilitate magnetic stirring.

A limpid colorless liquid was under the layer of foam, looking alot like egg's white that has been "beaten" to snow (don't know the expression in english, you knwo what I mean :) ). It was continuously stirred with the thermometer to mix up the foam.
the temperature was left to climb up to 20-25°C, sometimes up to 30°C, but kept under that temperature by applyling a ice bath now and then. There wasn't any strong exothermic reaction, the heating was due in part to the hotplate used to stir. There was continuous, but gentle, evolution of bubbles.

After 4h, the liquid started to turn limpid yellow, and the amount of solid/foam seemed to diminish. Stirring was much more easier.

At 6H total reaction time, the white foam had turned to light orange cristallin solids floating on the surafce of the liquid; the temperature ahd increased over 40°C at one point, and was quickly cooled donw to 15°C with the ice bath. There was still noticeable bubbling.

After 20H, a dark orange/red semi-solid tar like gum is obtained. No more bubbling can be seen.

I'm thinking starting the work up tonight, as it doesn't seem to evolve that much now. I hope the semi-solid will solidify entirely whne refrigerated, to be able to filter it effectively, otherwise I will have to do a solvent extraction. In that case, I will neutralize with acetic acid until I have a pH of ~4, so that any unrecated TAA will stay in solution and the nitrosyl radical will not be degraded.


All the glassware was washed with demineralized water to avoid contamination with metal salts, which would promote the decomposition of the H2O2 (note the huge excess).
The reagents were added in a bad order: the H2O2 was added to the weighed K2CO3, which caused some lumps at the begginign, and possibly a little decomposition from the basic conditions when only the first portion of H2O2 was added.

Apparently, potassium carbonate cannot freebase TAA.HCl. I will have the double check the pKa of TAA, but I don't recall it was that high... But, I wasn'texpecting the formation of the hydrate, so the transition could be hard to notice with the carbonate.. If the amine was already freeabsed, adding the NaOh would cause extra decomposition of the H2O2, as it is favored by high pH.

I should have: added stoechiometric amount of NaOH to TAA.HCl in minium water, then added the K2CO3/H2O2 solution to the freebase/hydrate (prepared by adding the K2CO3 gradaully to the stirred and cooled H2O2). I think it is important to cool the recation at the beggining of the addition, as the reaction could spark off and over heat if not controlled.
I'm not sure if the hydrate formation can be avoided if using not-so-cold temperatures, or adding the freebase to a concentrated carbonate solution. Even if it isn't practical, it doesn't pose a big problem to stir, it's just that the layer of foam isn't well mixed with the aqueous if not stirred by a rod to send it into the vortex.

Next time, i might try the bicarbonate procedure again, but use their workup which seems to be a breeze compared to extraction, washing evaporation and recristn.


from US 5,431,901:


A. 2,2,6,6-Tetraperdeuteromethyl-4-oxo-3,5tetradeuteropiperidine (Triacetoneamine-d16) (V)

A mixture of ammonium-d4 chloride (3.45 g, 0.06 mole), acetone-d6 (99.5 atom %D, 12.5 ml, 0.15 mole), anhydrous sodium carbonate (3.18 g, 0.03 mole) and magnesium oxide (3.0 g) was added to a 250 ml round-bottomed flask. The flask was capped with a rubber septum and wired, then the reaction mixture heated in an oil-bath at 50° C. for 3 days. After cooling, 20 ml of acetone was added to the reaction mixture and the resulting mixture was filtered. The recovered solid was crushed into powder, washed with 15 ml of acetone and then filtered with suction filtration. The combined filtrates were concentrated to dryness. The resulting red liquid (7.2516 g) was distilled under reduced pressure to obtain 4.7480 g (56.7%) of a bright yellow liquid (b.p. (boiling point) 54°-55° C./1.9 mm Hg) that solidified when chilled in a dry ice/acetone bath. The solid product subsequently was used without further purification. Recrystallization of an analytical sample from anhydrous diethyl ether yielded white crystals, mp. 57°-58° C. [lit. 58° C.]; IR(KBr, cm-1):3580(m), 3260(m), 2220(m), 1700(s), 1530(w), 1265(s), 1140(m), 1050(m), 930(w); 13 C-NMR (CDCl3): 31.03(m), 53.50(m), 54.88(s), 211.19 (s).

I guess this is basicly the procedure referenced in the zeolite absorption article, but using d-ammonium chloride instead of 15-ammonium sulfate....
No gaseous ammonia, or extended periods (well 3 days, but better than 3 month :) ). i will definatively try this out.

[Edited on 28-5-2008 by Klute]

[Edited on 28-5-2008 by Klute]

[Edited on 28-5-2008 by Klute]

Klute - 28-5-2008 at 15:57

Couldn't wait until tomorow: the first home made 4-oxo-TEMPO :D

It is still contaminated by TAA though... Will see tomorow how to proceed... I'm just to tired right now, between work at work, and work at home, I'm doing well over 18H a day :)

I will post details after a good night's sleep!

The oxidation needs optimizing, or at least waiting 48h total.
I will follow the more recent patents next time.

Klute - 30-5-2008 at 10:54

I restarted a TAA synthesis, following the same procedure with dry ammonia, but at a double scale. Hopefully the yield will be increased by isolating the TAA as a hydrosulfate or oxalate.

The oxidation didn't come out as expected. Apart from the product above, there was alot of unreacted TAA in the mixture. After adding more H2O2 and leaving it to proceed for another 24H, only white solids appeared upon refrigeration. These could be either TAA hydrate, or the hydroxylamine. They will be isolated and analysed by TLC..

If anyone really wants the details on the workup of th eoxidation, I can provide them, but I prefer trying out a new procedure as this one just seems to fall a little under what was expected...

I will post details of the TAA synth when done. Hopefully I will be able to start a NaBH4 reduction of TAA tomorow.

Klute - 31-5-2008 at 18:52

Synthesis of 2,2,6,6-tetramethylpiperidin-4-one according to US 3,959,295, second try

This time I (nearly) followed the patent's procedure, and it's proposed workup with conc. H2SO4. Works like a charm!

First Step

In a 500mL 3-neck RBF, equiped with a condenser fitted with a latex glove fixed with an elastic, a gas inlet, a thermometer and magnetic stirring, 63.2mL (1.377 mol) of technical acetone and 15mL of MeOH were introduced, followed by 2.2g (41.12 mmol) of freshly recrystallized and dried NH4Cl and 1g CaCl2 pellets.
Roughly 150g (2.8 mol) of freshly recrystallized NH4Cl were charged in a 500mL flat bottom flask, with a stir bar. The flask was mounted with a 250mL addition funnel, itself mounted to a gas outlet, connected to the gas inlet of the RBF in series with a solid NaOH washbottle.
A solution of 120g (3 mol) NaOH in 250mL dH2O was charged in the addtion funnel

The NaOH dripped onto the the NH4Cl, producing a slow, constant flow of NH3 gas. The RBF was immersed in a cold water bath, to which ice was periodically added in order to keep the internal temp under 20°C, most of the time between 10 and 15°C.
the flat bottom flask was heated on a hot plate to acheive better desorption of the NH3.

Once all the NaOH solution had been added, the NH4Cl and CaCl2 had completly dissolved into the colorless solution.

Increasing the heating of the flat bottom flask generated increasingly more NH3. Once the solution was nearly saturated (the glove started to get blown up), 50mL of technical acetone were added to the flask's contents. The initially colorless solution immediatly turned milky white, and gradaully cleared up as a white solid agglomerated onto the sides of the flask.

Gassing was continued as long as NH3 was generated. A very rapid flow of gas was now produced by the nearly boiling solution/slurry.
(Remark: In the first try, I dismantled the gas generator shortly after all the NaOH had been added. Apparently, a very large amount of ammonia stays into the NaCl slurry, and extended heating is required for complete desorption. This must surely account for the low yield of the first try.)
Once the flow of ammonia produced came to a near halt, the gas inlet was disconnected and immersed in a dilute H2SO4 solution. There was no suck back at all.
The gassing took a total of 7H to proceed entirely. This period could be shortened by increasing the flow at the beggining of the gassing.
The solution was bubbling from ammonia (saturated) as it heated back up to 25°C (ambient), so a few tiny holes were put through the latex glove to avoid excessive overpressure.

92.2mL of technical acetone were added (total amount of acetone introduced: 63.2 + 50 + 92.2 = 205.4mL; 4.477mol), without apparation of any white solid, and another 1.1g (20.56 mmol) of NH4Cl and 1g of CaCl2 were added. The ammonium chloride dissolved immediatly, the CaCl2 did so more slowly.

The colorless solution was stirred for 2h at room temp to avoid wastefull NH3 release. After that period, a small aqueous layer was noticed at the bottom of the flask.

Second Step

The flask was then heated to 55°C in a oil bath. AS the flask heated up, there was only minimal NH3 release.
After 10H of heating, the solution had taken a light yellow colour.

Following the patent's procedure, 14g of H2So4 were charged into a addition funnel, and slow addition started. There was a extremly vigorous reaction as the first drop touched the surface, generating alot of white fog and a strong sizzling noise.

The acid on the sides of the flask quickly turned to a white solid as it reacted with ammonia. White solids started appearing in the solution. The addition was stopped, and it was decided to continue reflux without addition of acid.
heating was continued for another 5h, at which point the solution had taken a blood red colour.

The flask was cooled with a cold water bath, and maintained in a ice bath while conc. H2SO4 was slwoly dripped in. Again, the reaction was vigorous, producing lots of white fog.

As the addition proceeded, the reaction became less intense, aswell as the generation of the fog. Any acid on the sides of the flask didn't not turn to a solid as previously even after prolonged exposure.
The temperature was kept under 35°C by changing the ice bath regularly. 75mL IPA was added to thin up the slurry and help stirring. A thick red slurry was formed. Once pH was acidic (to dampened universal pH paper, ~5), acid additon was stopped (over 25mL were required), and stirring at 10°C was continued for 45min. The pH was checked at the end o fthe stirring period, and was found to still be acidic.

The thick slurry was vacuum filtered on a large buchner, taking some time to pass (the filtrate started boiling sevral times). The light bordeau cake was washed with 200mL acetone, 75mL 2:1 IPA:Acetone, then 100mL pet ether (filtering was much easier after the first wash). This gave a off white solid cake, and a blood red filtrate.

The filtrate placed in the freezer after been sealed with celophane, and the solid was dried by suction for 20min, then under a lamp.

73g of a light, free flowing off-white/beige powder was obtained.

The powder will be extracted/recrysatllized from methanol to remove ammonium sulfate and other impurities. It has a light smell of the free amine.


What a releif! the workup is definatively much simplier than assumed, and avoids lengthy and destructive distillation of excess acetone and added IPA.
The yield is much better this time, surely because I took time to completly deplete the NH3 generator. Last try I must have wasted at least half the NH4Cl this way, as alot of gas remains in the slurry. Next time, I might consider simply heating 20% ammonia to generate the gas, as I have no more NH4Cl left.

the gradually addition of the acid during the reflux just didn't inspire me. I was scared of neutralizing too much unreacted NH3, and end up with a bunch of ammonium sulfate. Considering that the H2SO4 didn't immediatly react with the atmospher inside the flask as it was introduced during the titration, i think it was a good idea.

I added a extra portion of acetone before finishing the gassing, as the solution was getting saturated, and I didn't want to waste any NH3. Adding a little more methanol could have helped.

Hopefully the product isn't too contaminated with ammonium sulfate, and a simple MeOH extraction and subsequent recrysatllization will be enough. A mp determination is in order, though I have yet to find the litterature value for triacetoneamine hydrogenosulfate.

Conclusion: I would really recommend this recation to anyone desirable to obtaining a fair amount of TAA. The gassing stage isn't such as hassle, as it it virtually takes care of itself, and the workup with H2SO4 (or ethanolic oxalic acid) is very easy and requires very little work. The gassing time can be shortened, it just that I went too slow at first; I'm impressed at how well the mixture absorbs the ammonia, even with a flow of bubbles close to 3-4 bubbles/sec, the glove didn't get blown up entirely until the end were the solution was saturated. It was pretty impressive to see the massive bubbling of the solution from NH3 release once gassing was stopped.

Side note: I have tried the procedure without gaseous ammonia I posted a few post back, using NH4Cl, K2CO3, MgCO3 and CaO (I didn't have any MgO at hand, and was unsure about using either MgCO3 or CaO, so just tossed both in) into a sealed bottle kept at ~40°C. In one day it has already turned blood red, and the solids have turned from very fine powders to a spongeous mush. I will update with the results once workep up.

[Edited on 1-6-2008 by Klute]

Klute - 2-6-2008 at 10:17

MeOH extraction of the above crude bisulfate was tedious, so it was decided to purify the TAA via it's hydrate.

The MeOH-wet powder was dissolved in 300mL warm water, resulting in a yellow suspension of a fine white/beige solid, which was vacuum filtered (calcium sulfate) and washed with a little dH2O.
The yellow solution was basified with conc. NaOH until pH>13, at which point the solution became cloudy, and smelled of ammonia. It was placed in the fridge, then in the freezer for 45min.
the mixture seemed to be entirely frozen. It was left to warm up for 15min, then vacuum filtered. It had turned to a gelatinous mass, which was long to filter. Gelatinous scales slowly collected.

After testing their insolubility in acetone, the gel-like cake was flooded and triturated under 150mL acetone for 5min, vacuum disconnected, and vacuum filtered again. Beautifull white scales were collected, pressed down and dried by suction for 30min, then layed down under a lamp to air dry.

After 4h, the total weight of pur 2,2,6,6-tetramethyl-4-oxopiperidine hydrate was : 62.0g. :cool:

EDIT: a second crop of the hydrate was obtained by adding K2CO3 to the filtrate and placing it in the freezer. The white crystals formed were vacuum filtered, and washed with acetone before being dried by suction and under a lamp. The solid is much more "soggy" and is possibly contaminated by K2CO3, although it melts around 50-60°C.
I'm unsure of which solvent to use to recrysatllize it. I think I recall undried ether been mentionned somewhere, but can't find the mention again. What could I use, water? It might not be the most effective way to remove K2CO3..

Well, next step is NaBH4 reduction! I now have some TAA hydrate and recrystallized bisulfate (from the MeOH cristn) that will be properly characterized.

BTW, can anyone suggest a easy and good yielding method of reducing oximes? I was thinking of doing a CTH to reduce the oxime of 4-oxo-TEMPO or TAA (depending on reducer used, hydrogenation reduces the nitrosyl radical IIRC) as I've got some experience in the field, but if I could find a another good reduction that doesn't involve Pd/C (or messy zinc), I would be happy. I have heard one unreferenced proposition of using NaBH4 alone, but haven't found much info so I doubt it a bit. Any suggestions welcome.

Still the crude crop to purify and the unrecrystallized TAA.HSO4 to add.

[Edited on 3-6-2008 by Klute]

guy - 2-6-2008 at 19:19

wow beautiful pure crystals!!

Klute - 2-6-2008 at 23:23

Thank you very much Guy :)
Apparently when purified via a salt, the TAA anhydrous freebase is pure enough to crisatllize at room temp, and forms beautfillu small white needle! The problem is the mp is slightly depressed and in the day when it's hot they completly melt to a yellow oil, and crisatllize back at night :)
The hydrate stays as is though, and seems like the ideal way of storing TAA, no need of basifying for any further reactions, and if the nahydrous form is needed, a reflux in toluene with a dean stark or similar should be just enough!
Gaseous ammonia is definatively the way to go!

EDIT: after some though, I'ev decided on posting all my TEMPO-related work in this thread, instead of making a thread for each compound, to regroupe all the informations on theses derivatives in one big thread.

Synthesis of 4-hydroxy-2,2,6,6-tetramethylpiperidine by NaBH4 reduction of triacetoneamine

Effect of electrostatic factors on the stereochemistry of the hydride reduction of ketones of the piperidine series
É. A. Mistryukov
Russian Chemical Bulletin (14)10; 1788-1792 (1965)

Very kindly provided by Solo in the ref forum.

Ina 100mL 3-neck RBF, equipped with a condenser, a addition funnel, a thermometer and magnetic stirring, 4.33g (25,00 mmol) of 2,2,6,6-tetramethyl-4-oxopiperidine mono-hydrate (preparation detailed a few posts above) were dissolved in 32ml of dH2O. the dissolution took a few minutes and was endothermique. The solution was then cooled down to under 5°C in a ice bath.

A solution of 0.53g (14.03 mmol, 2.24 hydride equivalents) NaBH4 in 13mL dH2O containing a few mg of NaOH (to stabilize the hydride) was prepared (a little fizzing at first, that quickly stopped) and charged in the addition funnel. With steady stirring, the hydride was added dropwise over 10min to the cooled TAA solution. There was no noticeable appearance change, or visible bubbling. The addition funnel was rinsed with a few mL of MeOH, which caused some TAA hydrate to crash out of solution. A little dH2O was added until complete dissolution. Once the addition was finished, the flask was stirred in the ice bath for another 10min, then at RT (15°C) for 2h30.

5mL of 31% HCl were then charged in the addition funnel, and added dropwise, evry slwoly at first as each drop caused massive H2 evolution. The flask was immersed in a cold water bath, and the temperature kept under 20°C. After ~3mL of acid ahd been added, the gas evolution calmed down at each drop then ceased completly. The rest of the acid was added in one portion. The completly colorless solution had a pH under 5 (universal pH paper), and had a smell pretty similar to that of TAA freebase, but noticeably different.

The solution was transfered in a 100mL single neck RBF, and place din the fridge awaiting vacuum distn of half the water tomorow.

Update will be posted when I get time to, possibly not until a day or two (that's why I preferred starting now, pretty busy at work at the moment).

The reaction looks very easy and manageable. Hopefully removal of most the water could be avoided by using more concentrated solutions (or adding solid NaBH4) and saturating with NaCl or K2CO3, but for the first time I'm sticking to the article by the letter (well nearly :) ).

Direct oxidation to the nitrosyl in a one pot reaction could be possible, I don't think the borate salts can cause problems, actually perborate formed by action of H2O2 on borax could be a good oxidant for this type of reaction!
It would just take neutralizing the excess hydride with hydrochloric or acetic acid (directly generating acetate co-catalyst in situ), adding (bi)carbonate, and slowly adding H2O2 at 50°C.. Both reactions could be done pretty quickly, the reduction is surely finished in less than a hour (TLC could confirm that), and the oxidation hardly taking more time than that. Place in the fridge, filter the next evening, recrysatllize et voilà! Let's cross fingers.

[Edited on 4-6-2008 by Klute]

detritus - 4-6-2008 at 08:00

Wow. You are really the TEMPO master, Klute!

I just wanted to chime in that it seems gaseous NH3 is really not needed from what I'd found. I'm sure using gas would be better, but for those who are too lazy to make a gas generator, fair results can be obtained by simply mixing the ammonium salt and a strong base under acetone.

I ran a few small-scale experiments comparing various ammonia salts and various bases and found that almost any ammonia salt and almost any base mixed in situ with acetone will (after time) turn yellow, orange, then deep red. No heating was used, just time.

Lack of water seems to help the condensation but some water seems to help get a complete degradation of the ammonium salt. So hydrated salts like ammonium acetate combined with your standard KOH or NaOH in an excess of acetone seem to work well, going bright yellow in just 1-2 days at room temp. Color goes to orange after 4 days and is deep red after 8. I expect NH4Cl and NaOH would be even better, but I didn't make up any chloride.

I did not work any of these up as I was just looking for the speed of color change, but I did notice that with an excess of acetone, two layers easily form if a little water is added. Maybe I'll run another test soon with the sulfate workup.

Pardon my bad form, please! This is just a intellectual exercise for me as I don't really have any use for TEMPO, so I only hope to contribute a little to those who are interested but don't have all the fancy glass like Klute. As far as I can tell, this condensation is pretty easy - so more of us should give it a try!

[Edited on 4-6-2008 by detritus]

Klute - 4-6-2008 at 11:10

Very interesting! I am currently trying method using only ammonium chloride potassium carbonate and MgO or CaO, it's looking good (keeping it at 50°C for afew days..).

Beware, color change doesn't necessairly imply triacetoneamine formation! At room temp the formation of acetonine and diacetoneamine are predominant! Very long times (weeks) at RT or some heating is apparently needed to obtain good amoutns of triacetoneamine.

Indeed, the reaction seems to go well enough even with water, certain methods use NH4OH as ammonia source with (claimed) good yields). I'm sure refluxing NH4OH with acetone and a catalyst would be a viable way of obtaining TAA, but you cna easily salt it out like with anhydrous conditions. At least one tedious ddistn would be needed.

I agree using gaseous ammonia isn't very practical for most (even if it's much more simple that mmost thing, and relatively odorless with enouhg precautions), but I consider this to be the best way I've tried until now. Doig it on a rather large scale doesn't imply that much extra work, and then you are fitted with TAA for a loong time.. (keep in mind the oxidations are quite high yielding and the amount of TEMPO needed is rarely more than 1-5% molar..)

I'd be delighteed if anyone tried out more preparations and isolated some TAA with much less effort!

Polverone - 4-6-2008 at 11:21

Originally posted by detritus
I just wanted to chime in that it seems gaseous NH3 is really not needed from what I'd found. I'm sure using gas would be better, but for those who are too lazy to make a gas generator, fair results can be obtained by simply mixing the ammonium salt and a strong base under acetone.

I ran a few small-scale experiments comparing various ammonia salts and various bases and found that almost any ammonia salt and almost any base mixed in situ with acetone will (after time) turn yellow, orange, then deep red. No heating was used, just time.

Acetone will condense under the influence of bases to form progressively darker colored products as you describe even in the absence of ammonia. The procedure you describe might work but I would be wary of making conclusions on the basis of color.

Klute - 4-6-2008 at 15:40

Hum, bad surprise with the NaBH4 reduction!

2/3 of the water from the acidified solution was removed under vacuum, between 35 and 45°C (fluctuating because of irregular take off). the totally colorless solution was then basified with K2CO3 until CO2 evolution stopped, and more carbonate added to acheive saturation. The solution turned milky, and was extracted with 3x10mL DCM. A very fine white solid caused a thick emulsion as it stuck to the interface. The biphasic mixture was filtered through a cotton plug to remove most of it, the collected solid was then extracted with 5mL DCM. The extracts were directly dried, and solvent removed by simple distn, leaving.... nothing.

Nada, no a trace of solid or oil. A empty flask. That rarely happens! Confused, I though maybe the carbonate wasn't basic enough to freebase the compound, although it is employed in the procedure I worked with, so I added some NaOH to the aqueous mixture. It turned milky, the distd DCM used to extract the solution again, which turned clear, the solid (I mixed it with aqueous back again) filtered through a cotton plug, and the extract dried and distd. TLC indicated a very eluted spot (AcOEt), so I considered that this time it had worked, and the NaOH was required to base the amine.
Upon removal of the solvent, there seemed to have a very small amount of oil left (could be mainly DCM, I didn't want to apply to much vacuum to avoid eventual sublimation). The flask was cooled, and a little (1-2mL) pet ether added. There seemed to be a infinite amount of white solid appear, but it could just be dust. After 30min in the fridge, still nothing (mp of the coumpound 127°C)

As a last ressort, I considered that the white solid that I considered to be borate salts from the hydride, was the product,a nd that it was insoluble in DCM (unlikely). It didn't melt above 130°C, and had the appearance of an inorganic compound when heated on the hotplate. Insoluble in water, DCM, or alcohol.

Were did my product go? Have I just discovered a reaction that is an exception to the law of conservation of matter? I'd prefer to get some 4-hydroxytetramethylpiperdine actually, more usefull to me right now than revoking the basics of chemistry :P

The only thing I can think of is that either:
-removal of the water in presence of acid, hence heating to 50°C for a long period ina cidic media, somehow degarded my product (condensations, polymerisations?), giving a neutral water-soluble compound. Maybe it is more sensible than the compounds the authors of the article used (I doubt so: 3-methylpiperidone)
-the freebase is EXTREMELY soluble in water and not possible to extarct with DCM (doubt so, I would ahev at least extracted a minute amount).
-The TAA hydrate is not TAA hydrate (doubt so, mp is perfect). I will try a marquis test tomorow to confirm (shoudl react is it's not an amine, no?), and possibly a hinsberg test with TsCl, should give an insoluble product upon action of a base.
-I have managed to make matter disappear in a vortex or a 9th dimension or something. In that case I wish to name this phenomenon "Klute's dilema effect".

I will try the reaction again using methanol. I'm not sure how to dry the hydrate, heating the solid under vacuum before adding solvent? Distn with dry IPA? Dean stark with a little toluene, then diluted with methanol?

I'm really surprised the reaction didn't work out, and really can't see where the problem comes from. The litterature made me confident this reaction would be a breeze, I followed the procedure exactly...

Any suggetsions?

[Edited on 5-6-2008 by Klute]

Nicodem - 5-6-2008 at 00:25

One of my professors often used to say: "You ain't a real organic chemist until one day you manage to isolate sodium sulfate and believe it to be your product!"

Sodium sulfate decahydrate: mp = 32.4 °C, insoluble in acetone

2,2,6,6-Tetramethyl-4-piperidone monohydrate: mp = 59 - 61 (°C)

So, the question is: Are you are real organic chemist or not? :D

Klute - 5-6-2008 at 09:43

Hehe.. Well that aready happened to me in the past, so I'm already an organic chemist :D

I did a hinsberg test on my supposed TAA hydrate and a blank with Na2SO4. I think my TsCl was bad becasue in the two test a white precipitate formed (much more with the TAA though), but even with the blank test it didn't dissolve in 10% NaOH, eevn with 20min heatinbg at 60°C..

I did a marquis test, the TAA dissolve dentirely inthe conc. H2SO4/formol, the Na2SO4 just formed a thick insoluble white block, with absolutely no colour change (I used a tainted sulfuric acid though, so hard to see minor colour change).

I melted hydrated Na2SO4 and TAA hydrate one next to each other on the hotplate (cold at first), the Na2SO4 mellted a few minute sbefor ethe TAA, and then formed an dry white solid (the anhydrous salt), were as the TAA melted and the quickly satrted to turn brown and fume.. I really doubt the TAA isn't what it seems to be. But i would like to be able to determine is there are any inorganic salts present, especially in the second crop (there surely, considering the change in apparance).

I purified the 4-oxo-TEMPO by column chromatography, beautifull blue substance came first, surely a unsaturated nitro- ring opening product described in the article KMnO4 posted. I've kept the fraction to isolate it on rainy day to see if it's the crisatllin one or the oil. I guess removing the ethyl acetate at atmospheric in presence of acetic acid caused some ring opening. Next time I will use DCM and wash the extract before removing solvent.

Pictures to come in a few days; I will need to put a halt to reactions for a while , going to Paris for a few days.

Klute - 7-6-2008 at 15:17

I used 5% AcOEt in toluene at the beggining, to elute most of the blue fraction, than gradually came up to 20% AcOEt in toluene to elute the 4-oxo-TEMPO: worked like a charm, and offered a very good seperation (there was a blank fration, no UV-visible compounds for ~50mL between the blue and orange fraction, using 5uL samples on the plate (twice the amount usually used, most of the time hardly 0.5uL is needed to get a spot with 1% solutions).

The solvent was removed under vacuum, to give a red oil that crisatllized once excess toluene had evaporated off at RT to give a low-melting red solid. It will be recrysatllized with pet ether to try and get a sharper mp.

Pictures of the column:

First blue fraction:

it's surely one of the two unsaturated nitro ring-opening compounds mentionned in the article KMnO4 posted and out of which I used the procedure for the oxydation. They claim ring-opening can happen on workup of the product, the ethyl acetate was removed at atm pressure and contained some acetic acid. It seems that's all it takes. I will remove the solvent latter on and see which compoud it is (one is an oil, the other a crisatllin substance)

The orange fraction (pur by TLC) once solvent is removed under vacuum.

Picture of the pur product will arrive when done.

Klute - 8-6-2008 at 12:47

Ok, I have tried another oxidation with the supposed TAA hydrate, and complete lack of isolated product has lead me to think the product is indeed a mixture of sodium sulfate and potassium carbonate, along with a little amount of TAA hydrate, enough to give a faint amine odour and a colour change with a marquis... God damn it!
I have trie dthe oxidation with the hydrosulfate I isolated, definitevly is the right compound as once basified it release this pale orange oil with the caarcteristic TAA smell... So I have 50g of beautiufull hydrated Na2SO4 and no more TAA, great!

I will work up the TAA preparation with NH4Cl, K2CO3 and Mgo/CaO and see what gives, and might consider trying out a new synth, but without gaseous ammonia. Refluxing NH4OH with acetone and CaCl2/NH4Cl overnight might be enough, IIRC Guy tried out a simialr thing. Guy, if you are reading this, how did that run go? How did you isolate your product?

Details on the oxidation to come. I've decied on oxidizing evrything in one go, and thenr educing the nitrosyl to the hydroxy-TEMPO, the oxime and the amine will wait for a new TAA batch...

Nicodem - 9-6-2008 at 03:37

I have yet to see a amine hydrogensulfate that is insoluble in alcohols, particularly methanol. And you recrystallized the crude product from methanol, so perhaps that was the crucial error.

Klute - 9-6-2008 at 05:59

Well, the TAA.HSO4 isolated from the methanol seems definatively free of ammonium salts, as when basified no ammonia smell was detectable (there could haev been minutes amounts covered by the smellof TAA).
It was just too tedious as the fine solid remaining was very fine and was long to decant, so I just did one extraction with ~150mL MeOH (recovered less than 100mL), which furnished the 6.9g I used for the oxidations.
I'm sure if I had continued these extractions I would have recovered more TAA. When I dissolved the solid, i used too much water: the formed TAA hydrate must have stayed solubilized in the water, or partially precipitated with the massive amount of Na2SO4.10H2O and other salts, hence the smell and the colour change with the marquis.

I'm going to try more basic methods like refluxing concentrated ammonia with acetone and ammonium salts and stick to that, satrting to get bored of using ammonia gas, I would be happier if the recation could take care of itself while I do other things.. I have still the MgO/CaO run to work up, but that was one a rather small scale any way.

LSD25 - 9-6-2008 at 06:26

I'll grab some benzoate salts and try the ammonium benzoate method, at least I should get something

Here is the other ammonium salt article:

Apparently ammonium benzoate gives predominantly the TAA-benzoate when refluxed with acetone IIRC, I have to look the article up again

BTW Here is something of interest:

See page 18/60 where it describes the effect of adding alkali metal alcoholates to triacetoneamine, ie. the formation of the alkali metal salt (by one Emanuel Merck no less). Now I seem to remember reading somewhere that dry alcohols mixed with alkali metal hydroxides exist predominantly as alcoholic solutions of the alkali metal alcoholate? Would this explain why this is acting weird?

Klute - 9-6-2008 at 06:33

Was the ammonium benzoate used stoechiometrically or catalytically? If used insteochiometric amounts, hopefully a less expensiveve salt can be used.. I have a little by-product benzoic acid at hand, i could try using it as a catalyst, hoping that it will be more effective than simple ammonium chloride..

I'm not sure what you are implying with the alkoxide? There si no way any methanolate could be produced in the reaction with ammonia, presence of water, no strong base, large amount of acetone. Or where you talking of something else?

LSD25 - 9-6-2008 at 07:32

Too tired to think it through at the present time, but it struck me as odd, we have an amine which forms a sodium salt... too fucking odd, there might be something there but I'll have a look in the morning. Probably something to do with it acting as an acid, rather than a base, but the eyes are shutting and its time for bed (In the CA they specifically referred to it as a salt).

The ammonium benzoate was used directly instead of ammonia and other ammonium salts, I think I pointed you toward the article once before? IIRC it was suggested as a simple, lab-scale procedure for preparing small quantities of TAA. Sodium(?) benzoate isn't expensive, so I was thinking of grabbing 1kg of it, then forming the ammonium salt of it.

BTW Happy birthday Queen Elizabeth for yesterday, though why we celebrate it on Queen Victoria's birthday is fucking beyond me.

Klute - 9-6-2008 at 09:04

Well, you had mentionned the use of ammniumbenzoate, salicylate and anthralinate, but you couldn't find the article or something similar.
I might try the procedure with calcium zeolites, using ammonium nitrate and ammonium hydroxide, but I don't know where to find such zeolites. I will have a look at aquarium stores or similar, and might try with silica gel or straight calcium chloride..

Klute - 23-6-2008 at 08:47

Preparation of triacetoneamine according to US 6,646,127

The patent's procedure in example 22 was followed, using CaCl2 as catalyst.

NH4NO3 was prepared by reacting 53% HNO3 with 30% NH4OH (very vigorous reaction), evaporating the water and drying the salt by melting it.

Acetone was technical grade, CaCl2 pellets were powdered prior to addition.


In a 500mL 3-neck RBF, 5g (45mmol) of powdered CaCl2 and 40g ( 0.5 mol) of dry NH4NO3 wee suspended in 220mL acetone. 31mL (0.5 mol) of 30% NH4OH (Panreac) were charged in the addition funnel, and slowly added dropwise, over 20min.

At first there was a little caking of the solids, then gradual dissolution. The flask warmed up to 40-45°C, so an ice bath was applied to keept temp under 30°C. Once the addition was finished, most of the solids had dissolved to give a limpid solution.

The reaction medium was then stirred at RT for 7H total. The slightly yellow solution gradually turned orange then red, with that caracetistic "fluorescent" red tint near the end.

The condenser was then replaced by a fractional distn setup, with a short (200mm vigreux column). The acetone was removed under slight vacuum at 45°C over a 80°C oil bath. the reaction mixture had a strong smell of TAA. The colorless distillate smelled strongly of ammonia. Distn was continued until no more acetone was distilled, and water started to take-off.n A dark red residu remained.

After cooling down to 30°C, the very viscous blood red residu whic smelled strongly of TAA was transfered to a 1L seperating funnel, and extarcted with 4*50mL toluene. The extarcts came out golden orange, leaving the blood red color in the viscous aq. layer. After the 4 extraction, the aq. didn't really smell as TAA anymore, but more as ammonia.

The toluene extracts were combined, washed with 50mL conc. K2CO3, and brine, thend ried over K2CO3 in a stoppered erlenmyer for 20min. The extract was then transfered to a 500mL beaker, and acidified with 20% v/v H2SO4 in IPA. At first, a white precipitate formed, but as acidification continued, a dark brown tar_like resin preicpitated also, very similar to the other TAA workups.

There was still a certain amount of white precipitate in suspension. After standing for 5min, most of the suspension was vacuum filtered, leaving most of the resin in the beaker. 100mL acetone:IPA 50:50 were added, and the resin triturated, dissolving most of theresin and freeing more clear solid. The suspension was vacuum filtered, thoroughly triturated on the buchner with 50mL acetone, and 50mL acetone:pet ether 50:50 before drying by suction for 5min. The fine beige precipitate was then placed in a glass cup and dried under a lamp to afford 10.81g (41mmol) of crude 2,2,6,6-tetramethylpiperidone hydrosulfate.

The aqueous layer was re-extracted with 4*50mL AcOEt. These extracts took on a much redder colour. The combined extarcts were washed with 50mL co,nc. K2CO3, brine, and dried over Na2SO4.

This extract was also acidified until acidic on dampened pH-paper with the same H2SO4/IPA, giving a very fien white precipitate at first, then a brown/purple voluminous solid. This solid was much harder to correctly clean, giving a sticky paste, and required larger amounts of acetone, IPA and AcOEt to get it relatively clean. The sticky paste finally dried, giving a crunchy brown solid (4.2g).

The two batches will be recrysatllied from MeOH/pet ether or MeOH/AcOEt.


Thsi reaction is pretty simple compared to the use of gaseous NH3, but the claime dyields are still not attained. Toluene seems to be a much more selective solvent for the extarction than AcOEt is, giving a much cleaner product. The workup still needs optimizing though. Direct fractionnation isn't possibble becasue of the large amounts of inorganic solids present in the syrup. Using diethyl ether as in the patent's procedure, or DCM wasn't possible because of the ambient temp >35°C here (hot...). Tolueneseems pretty eefctive, but still requires large amounts of solvent. I guess a gradaully distn of the solvent, and extarction with the recycled toluene could be the best option, as the extract seems pretty clean. Fractionnation of the TAA freebase might be a better option than direct acidification, it is pretty clear there is a resinous product appering near the end of the acidification, covering the TAA.HSO4.

Forming the hydrate or oxalate after fractionnation should give a pure product, stable as is. I will try another variation of the reaction, possibly letting it heat up during the addition s is mentionned in the patent. I'm sure the yield is good, and that the isolation of the product is problematic (TAA freebase is pretty soluble in water).

I think we are in the good direction though. Using only ammonia could be more practical, or at least catalytic amount s of ammonium nitrate (apart from the extarction, producing dry NH4NO3 is the lengthy part).

I still have some NH4NO3 left, so will try another recation with a smaller NH4NO3/NH4OH ratio, but same acetone/ammonia donor ratio.

According to the patent, excellent selectivities to TAA are possible at RT, the acetone/ammonia ratio seems to be more important than the temp regarding acetonine/DAA formation.

[Edited on 23-6-2008 by Klute]

kmno4 - 24-6-2008 at 11:03

I tried to bypass gaseous/liquid NH3 by mixing :
25g acetone
5g H2O
18g Ca(OH)2 , containing ~15% CaO - I just have very old CaO :)
30g NH4Cl
Ammoniacates of CaCl2* are not very stable at room temp (at least higher than [Ca(NH3)2]Cl2), so I thought that Ca(OH)2/NH4Cl can serve as NH3 in situ source. At the begining, mixture did not look good (semi-solid), almost not able to stirr. But after ~1 hour of shaking and stirring with glass rod, consistency became more liquid than solid. It turn out to be stirable without problems with magnetic stirrer in a stopped flask. After ~24 hours of stirring, mixture became white-red: red thick layer and semi-solid white reminder. There was some problem with removing this red something, but it turn out (it was my 3rd run :P) that it can be removed with aid of dioxane. I just added 4x20 ml of dioxane with stirring and careful pipetted off liquid after each portion of dioxane.
After evaporation there was ~20g of brown-red, very thick remainder, smelling amine (for me it was like morpholine smell), soluble in water, not soluble in toluene, of very bitter taste. I supposed it was mixture of more than 10 substances but hoped there was large amount of acetonine in it.
So I treated it as acetonine and tried to convert into triacetonamine by procedure from some patent (acetonine, NH4Cl, KI, acetone, water).
Next, I evaporated acetone from water bath - I got brown-red sticky mass looking and smelling like input "acetonine" :(.
After adding 30% NaOH, formed oganic layer (red-brown) was taken, mixed with HCl(aq) and evaporated**. After very long evaporating from water bath, I was not able to gain any crystals of hydrohloride. I got honey-like, brown-red mass.
That is all from me. Very messy job :mad:
*if somebody is interested, I can share some articles about these ammoniacastes (U2U)
** propably I oversimplifyied this procedure - I should have extracted organic and aquenous layer with toluene (as Klute did)

Klute - 24-6-2008 at 12:12

Thank you for sahring! I am glad this subject is starting to interest others! :)

I have also tried a method using CaO/MgCO3, acetone , K2CO3 and NH4Cl, it turned the same as you describ it, a slurry turning to a paste under red supernatant... It's been some time now (weeks?) an dI still haven't bothered doing the workup i admit. I will post a pic of it soon.

A efficiçent, practical preparation of TAA at a reasonable scale is still missing... I will try some more modification of the use of conc. aq. ammonia, as this eems the most direct path. It's annoying becasue as long as this isn't solved, UI cannot do any furhter work on the nitrosyl radicals and their modification. I would like to obtain a deacnt supply of TAA and work on from there, rather than preparing 5-10g each time..

Ullmann - 1-8-2008 at 05:47

Very good work Klute and very inspiring and detailed procedures you are showing us, thank you!

I have a little something to add to the TEMPO field:

The reduced reactivity of the keto analog ask us to change that keto group in an OTC fashion and I think the best we can do is not to wolff-kishner it nor reduce it but ketalize it!

check those few patents (no journal refs here, sorry)
EP1443049 & EP1457491 for ketalisation
EP1595868 & EP0574666 and US2005256312 for oxydation of the ketalized TEMPO
US5821374 and US5631366 for the oxydation of alcohols to aldehyde using ketalized TEMPO and TCCA or NaOCl

The most obvious choice and an example in the precited documents is the ethyleneglycol ketal. The yields with it are as good as with TEMPO and it is the most straightforward to make in your garage! Only a few mol% are needed as catalyst for great and quick results! With a few g of it you can make readily kg of aldehydes!

Be the TEMPO be with you!

Klute - 1-8-2008 at 09:34

Thanks alot :)

Indeed, the acetals weren't something I focused on, but do look promising. there aren't further ways of modifying them after, but they would definatively be something to include in the oxydation test to come. I think I will get started again on the subject ins eptember or so, for now I'm focusing on more time consuming projects as I'm on vacation. I would be very happy if you would like to join me and a few other in trying out various derivatives, so then we could each report our results and see which are the most avantageous catalysts.. I plan on using the benzyl alcohol oxidation asa comparasion test, because I have enough on hand, but other substartes could be used, like certain aliphatic alcohols.

Before any of this, a easy, high yielding method of preparing TAA would be needed to generate all the various derivatives, preferably withotu the need of dry NH3 gas...

ziqquratu - 17-1-2011 at 20:31

I've just performed the synthesis of triacetonamine in accordance with Klute's post above and patent US 6,646,127. Due to the glassware I had available, I did it on half of Klute's scale.

CaCl<sub>2</sub> (2.5 g) and NH<sub>4</sub>NO<sub>3</sub> (20 g) were stirred with technical acetone (110 mL) in a 3N, 250 mL RBF fitted with a reflux condenser, a thermometer and an addition funnel. Ammonium hydroxide (28 %, 15.5 mL) was added, with mild cooling, over ca. 20 min, keeping the temperature below 20 °C. The cooling was removed and the mixture stirred at r.t. (20-25 °C) for 16 h. The originally clear, colourless solution slowly became yellow, and was dark red by the end of the reaction.

The acetone was removed on a rotovap, and the resulting viscous aqueous layer was extracted with Et<sub>2</sub>O (5 x 60 mL). The combined organics were dried over Na<sub>2</sub>SO<sub>4</sub>, filtered and the solvent removed to afford a yellow oil. The oil was transferred to a smaller flask (it began to solidify in the process, so more ether was used to assist the transfer) and distilled through a short (~10 cm) Vigreux column to give the product ketoamine (15.60 g, b.p. ~80-87 °C/11 mmHg) as a yellow oil, which solidified to yellow needles on standing.

<sup>1</sup>H NMR (300 MHz, CDCl<sub>3</sub>;) 1.23 (s, 12, CH<sub>3</sub>;), 2.26 (s, 4 H, CH<sub>2</sub>;)

The distillation was somewhat bumpy - a bit of extra headspace would be an advantage here (in my case, the column JUST saved me when using a 50 mL RBF - 100 mL would have been better). Also, following the extraction and removal of the ether, I ran an NMR, and the material was already quite clean - just a little baseline rubbish. It would probably have been usable at that point without distillation. So perhaps ether is the best solvent for the job (but, obviously, it isn't easily available to everyone).

I intend to perform the Wolff-Kishner reduction of the ketone in the near future. I'm unlikely to provide images, but I'll give a write-up if/when I get to it!

mr.crow - 18-1-2011 at 10:06

Great work, and excellent thread. Now that it has been bumped to the top I say we sticky it

Has anyone preformed an oxidation using the 4-oxo-TEMPO? Is it necessary to remove the ketone group?

KrysHalide - 8-6-2013 at 10:00

This review is of interest for anyone aiming the synthesis and derivation of 4-oxo-TEMPO:

C. Bruckner
Organic Preperations and Procedures Int 36 (); 1-31 (2004)

The H2O2/WO4 oxydation procedure was tried with succes:

In a 300ml erlenmeyer, 300mg of NaWO4.2H2O were dissolved in 50mL dH2O, followed by 4g of brown colored triacetoneamine crystals (>95%). A turbid brown/orange solution was obtained.
3mL of 30% H2O2 were added with good stirring over 5 min, O2 bubbles appearing quickly.

The brown/orange solution was stirred two hours at RT, the colour fading to a limpid orange color.

The solution was saturated with K2CO3, a orange oil crashing out and quickly agglomerating to small orange crystals. These were extracted from the milky aqueous with 3*25mL Et2O, the extracts washed with brine (with a spatula of NaHSO4 added), dried over K2CO3, and the ether removed on the rotavap.

A clear dark orange oil was obtained. It was topped with hexane, and placed at -18°C. Orange cubic crystals were obtained, filtered, dried, to give the product in a 87% yield, mp 35-36°C, pur by TLC (1/1 ether:toluene)

AndersHoveland - 8-6-2013 at 23:11

Nitric oxide is another stable radical with catalytic properties, though of course it has to be kept away from air.

For TEMPO, the wiki entry states that "The stability of this radical is attributed to the steric protection provided by the four methyl groups adjacent to the nitroxyl group", but I do not think this is true.

Those methyl groups just prevent vicinal hydrogen atoms from ionizing off. An example is the oxidation of N,N-dialkyl-hydroxylamines by hydrogen peroxide.

So I think N,N-di-tert-butyl nitric oxide would also be a stable radical. (tBu)2NO•

[Edited on 9-6-2013 by AndersHoveland]

KrysHalide - 9-6-2013 at 08:53

Ok, they made it before Klute could: a solid-supported TEMPO catalyst, with a simple as it get's preperation:

Silicagel, 4-hydroxy-TEMPO, and 1 eq of NaCl, RT then annealing..

Reuseable, easily recovered, efficient oxydation catalyst with no leaching.. Sounds promising and easily accesable.

"Novel [4-Hydroxy-TEMPO + NaCl]/SiO2 as a Reusable Catalyst for Aerobic Oxidation of Alcohols to Carbonyls"
N. Tamura, T. Aoyama, T. Takido, M. Kodomari, Synlett, 2012, 1397-1407.
DOI: 10.1055/s-0031-1290980

Let's hope other secondary oxydants can be used, although simply bubbling air looks like an economic alternative to traditional oxydants..

Could someone retreive the article? I've got 5g of 4-oxo-TEMPO waiting next to the NaBH4...

[Edited on 9-6-2013 by KrysHalide]

KrysHalide - 12-6-2013 at 04:26

I have optimized the oxydation a bit, as I have noticed my first product was contaminated from unreacted TAA..

So the amine is dissolved in water, the tungsten salt added, and the 4eq of H2O2 added dropwise at RT. After an hour os stirring at RT, the solution is gradually heated to 40°C (I once added the H2O2 at 40-50°C, there was an immediate color change but some resinous tar was also produced in small quantities..).. After 2h at 40°C, NaCl is added until saturation, aswell as a spatula tip of NaHSO4 to keep Ph slightly acidic.. The mixture is stirred vigorously as the NaCl dissolves, an red oil crashes out, letting the mixture to cool to RT. It is then placed the fridge for 30min: the oil solidifies to a red disk. It is removed with a spatula, leaving an yellowish aqueous layer.

The red solid is dissolved in Et2O, diluted with an equal volume of hexane or any hydrocarbon (petroleum ether, ligroin, etc), dried with NaHSO4, which absorbs a little brown color, decanted from the clumped salt, and the ether removed on the rotavap. The resulting red solution is placed in the freezer, and turns to a thick paste of a pale orange solid, which filtered while cold, washed with a little cold hydrocarbon, and dried by suction. The product is free from any unreacted amine.

The aqueous NaCl saturated layer is then stirred at RT, and more H2O2 added. Any unreacted TAA while have stayed in this aqueous layer, and gets further oxydized on addition of further H2O2 as witnessed by the color change. I even added K2CO3 to this aqueous, and the white cristallin TAA hydrate appeared. The biphasic mixture is stirred at 40°C for an hour or two, and treated as the first time. In this way, +95% yields of quasi-pur 4-oxo-TEMPO are obtained. Traces of degradation products are detected by TLC, but in minute amounts. Longer reaction times at RT could prevent this, but it seems complete conversion of the amine at RT is VERY long..

My plan is to reduced the 4-oxo-TEMPO the the hydroxy compound, then esterify it with benzoyl chloride, or alkylate it. I will also try the silica-supported catalyst is someone is kind enough to retreive the article I mentionned upthread!!

On the other hand, I will also go by another route, directly reducing the triacetoneamine with NaBH4, and then oxydizing the 4-hydroxy-TEMPO before esterification/alkylation, to compare global yields. The oxydation might be smoother with the piperidol.

By the way, I tried sublimating a gram of 4-oxo-TEMPO, which was a complete failure, surely due to the fact that i did it at atm pressure.. The red/orangfe solid was slowly heated in a RBF with a cold finger, but only a liquid vapour was deposited, with evident decomposition (turned to tar, generation of water vapour), and these drops never solidified. I was scared that this would happen with vacuum applied, considering the very low mp of these TEMPO derivatives..
I am not familiar with sublimations, is vacuum indispensable? Or do I have to heat less, and leave it to it own device for several hours?

Dr.Bob - 12-6-2013 at 05:35

Quote: Originally posted by KrysHalide  
I have optimized the oxydation a bit, as I have noticed my first product was contaminated from unreacted TAA..

I am not familiar with sublimations, is vacuum indispensable? Or do I have to heat less, and leave it to it own device for several hours?

For most compounds, a high vacuum is indispensable. Otherwise the compound merely oxidizes from oxygen in the air or decomposes from the heat before it sublimes. The lower the vacuum, the easier most compounds will sublime at lower temps.

Unfortunately, I can't get access to that article right now, but it looks interesting. There are several companies that sell other oxidants covalently bound to resins, but this looks very simple, inexpensive, and easy to use. Very nice find.

[Edited on 12-6-2013 by Dr.Bob]

KrysHalide - 12-6-2013 at 10:44

okay, thanks for the explaination! But, as vacuum decreases the bp, will it not also decrease the mp? Meaning that under reduced pressure, 4-oxo-TEMP will stay as a liquid even at -10°c (cold finger temp)? Or is the mp not affected?

watson.fawkes - 12-6-2013 at 12:34

Quote: Originally posted by KrysHalide  
okay, thanks for the explaination! But, as vacuum decreases the bp, will it not also decrease the mp? Meaning that under reduced pressure, 4-oxo-TEMP will stay as a liquid even at -10°c (cold finger temp)? Or is the mp not affected?
Melting points aren't much different between atmosphere pressure and vacuum, since the absolute pressure difference isn't much as pressure differences go. Nor does it change the partial pressure of the material very much. What it does is to change the transport dynamics for sublimation. When there's ordinary pressure at a surface, any molecule that comes off the surface won't travel very far (see "mean free path") before colliding with some other gas molecule . There's a deposition-sublimation equilibrium that gets set up. Some molecules diffuse away, but that's a slow process. This is called the "viscous flow" region. You have to wait for them to eventually diffuse to the other side.

With vacuum, the mean free path increases. If that length increases to be larger than the size of your geometry, you move into what's called the "ballistic flow" region, where there are very few collisions. In a sublimation apparatus, molecules sublime off the hot surface (in random directions), fly through vacuum with no collisions, and deposit on the cold surface. If you try this without vacuum, you're in the viscous flow region. Transfer rates can be orders of magnitude lower. For some substances that readily sublime, this may be adequate.

As a rule, you want to run the hot surface as hot as you can without either decomposition or overwhelming your cooling system, you want to run the cold surface at just under the deposition temperature (to minimize radiative heat transfer), and you want to ensure that you're vacuum is low enough that you're in the ballistic flow region for the gap between the two. Higher vacuum beneath that threshold is mostly wasted.

solo - 12-6-2013 at 18:25

Requested by Klute

Novel [4-Hydroxy-TEMPO + NaCl]/SiO2 as a Reusable Catalyst for Aerobic Oxidation of Alcohols to Carbonyls
N. Tamura, T. Aoyama
2012; 23(9): 1397-1401
DOI: 10.1055/s-0031-1290980

Attachment: phpGGs8p7 (207kB)
This file has been downloaded 1119 times

Dr.Bob - 13-6-2013 at 09:16

Thanks Solo, that is an interesting paper, and gives a pretty simple way to prepare a useful oxidant with very simple chemicals. I have used some similar solid supported reagents, and they are very handy for some uses, where having to purify something from the oxidants would be a pain. I may have to try that sometime for making aldehydes.

KrysHalide - 13-6-2013 at 13:11

Yes, thank you very much Solo!!

I have good news: I will be able to try this procedure very soon, as my first bath of 4-hydroxy-TEMPO is drying right next to the laptop!!

Reduction was very smooth, and gave a very clean product in quantitative yields!!

Here's what I did:

3.2g of 4-oxo-TEMPO were dissolved in 30mL abs EtOH. The dark red solution was cooled in a ice water bath, and 0.55g of NaBH4 were added over 5 min. there was slight H2 evolution, and a exothermic reaction. The ice bath was removed, and the reaction medium stirred at RT for 3H. There was a subtil color change, from dark red to a more orangish red.

Most of the ethanol was then removed on the rotavap at atmospheric pressure (vacuum pump down), and 20mL of water were added, along with some solid K2CO3 to attempt crashing the product out as during the TAA oxydations. But there was too much EtOH left, and a orange organic layer seperated. The aq was diluted with a little more water, and the product extracted with ether, which did the job very well. The combined extracts were washed with 2*25mL saturated K2CO3, and dried over solid K2CO3. The ther was then removed at atm. Hexane was added to the residu, but this time two layers formed, the product oiled out unlike 4-oxo-TEMPO. A little Et2O was added to try and obtain a homogenous solution, and the flask placed in the freezer. No cristallization occured even after 1H, so the flask connected to the rotavap, and as much solvent was removed as possible. Petroleum ether was added to the residu, and while cooling the flask under a stream of cold water, the product solidified nearly instantly, giving a pale orange solid, looking alot like 4-oxo-TEMPO, but of a much lighter pale orange color. The flask was left in the freezer for 30min to insure complete solidification, and the product vacuum filtered. The chunks of solid were crushed as fine as possible, triturated with some petroleum ether, and filtered again. After air drying, 3.2g of a light cristalline pale orange powder were obtained, surely still a little solvant present.

TLC ( Et2O) reveals no 4-oxo-TEMPO, and very faint impurities spots in addition to the product.

Succes :)

Tomorow, part of the product will be reacted with benzoyl chloride, and a small scale experiment will be conducted, the oxydation of benzylalcohol with 4-hydroxy-TEMPO. I haven't decided on the conditions I will use, I have TCCA, Ca(OCl)2, NaBrO3 and a few other secondary oxydants I could use..

I guess i will go along the sides of substrate dissolved in DCM, catalyst added, and solid oxydant graddually added, in presence of TBAB if need be. I will have to check the different procedures published, haven't read them in while.

I will also try the SiO2 supported version, but with a slight modification of the published procedure in that I will not use atmospheric oxygen as sec. oxydants. I guess this will circonvent the need for NaCl additive.

Any comments, suggestions, welcomed.

KrysHalide - 14-6-2013 at 10:42

Ok, the oxydation proceeded well.

1.52g (10mmol) of 3,4-methylenedioxybenzyl alcohol were diluted in 25mL of DCM, 50mg of 4-hydroxy-TEMPO were added. Then a solution of 520mg of NaHCO3 and 120mg of KBr and 80mg of TBAB in 20mL H2O was added. The biphasic mixture was cooled to 0°C in a ice bath, and 7mL of 10.2% NaOCl solution was slowly dripped in with good stirring. The first drops caused a quick milkyness, and appearance of a yellow color in the aq, which dissipated after a few seconds. Surely formation of the oxonium salt, and it's disappearance after reaction with the alcohol? The addition rate was adjusted to let time for each drop to correctly react, and took 30min. The bottom organic layer displayed a yellow/green color, and a white emulsion/fine solid appeared gradually. The reaction was followed by TLC, and was slower than expected. The benzylic alcohol was present even after 1H reaction time, so 3 more mL of the NaOCl solution was slowly added.

TLC reavealed formation of two compounds, with similar Rf, one darkening after exposure to UV lamp. A small uneluted spot also appeared near the end (carboxylic acid?). The aldehyde spot beacme larger with room temp stirring.

After 2H30 total reaction time, the biphasic mixture was filtered on a N°4 glass fritt to remove the white emulsion "blobs", the organic seperated and washed with NaHCO3 solution two times, and was then dried over Na2SO4. It was then filtered over a 2cm plug of silicagel, to remove the catalyst, which was quite effective by the color of the silicagel. The plug was rinced with DCM, and the solvent removed, to afford a green oil, that crystallized on cooling in the fridge. The mp was depressed, surely beacause of minor impurities. A 74% yield of crude product was obtained.

The aldehyde was directly used in a aldol condensation with acetone catalyzed by NaOH, which is currently stirring.

Basicly, this works well, but I was surprised by the slow reaction rate after reading numerous articles claiming that the reaction is over in minutes. The reaction conditions will be optimized, using different secondary oxydants, and TEMPO derivatives.

I also obtained some 2,2,6,6-tetramethylpiperidine, that I will oxydize to TEMPO, so I can compare the efficienty of each derivative to the original catalyst.

Unfortunaly, it will be on another substrate, as I only have 500mg of piperonyl alcohol left.

[Edited on 14-6-2013 by KrysHalide]

DJF90 - 11-9-2013 at 12:40

So, I decided to have a pop at this myself, following US patent 3959295. This is the same patent that Klute made use of at some point in this thread, although I cannot find mention of using CaCl2 in the patent. I'll spare the details on the preparation, but at the end of the reaction (I used judgement based on what the patent says) I filtered off the ammonium chloride (it never went into solution, another difference to Klute's replication) and concentrated to a residue. On standing the dark red-brown residue crystallised to a semisolid. I figured I'd bypass the issues with precipitating a solid, and directly vacuum distilled the semicrystalline residue at ca. 7mbar. Three fractions were taken, over the range 64-84*C and 6-9mbar. All three fractions were yellow oils, and the third crystallised on standing. Small samples were dissolved in methanol and TLC'd in 50%v/v EtOAc/n-Hexane, and visualised with UV, Ninhydrin and 2,4-DNPH (see attached image). The spots are a little strong, but what I suspect is the product (mostly fraction 3) does not stain particularly well with ninhydrin. I was actually suprised that there were so many impurities.

Where to go from here? Well I guess theres the option to purify further by hydrate or salt formation. I've not really decided what to do yet though, so any suggestions/comments are welcome. Its worth noting that Ziquarratu (sp?) also went the route of distilling the product directly, although he did not comment on the purity of his distillate.

2013-09-11 18.05.50.jpg - 234kB

ziqquratu - 11-9-2013 at 19:24

It was quite some time ago, and I've moved to a new work-place and thus don't have the data or the sample any more. However, from memory, the purity following distillation was excellent (and I suspect the slight amount of garbage - invisible by NMR, but a small high-running spot by TLC) came from my poor choice of distillation apparatus (too small with lots of bumping).

One thing I do recall, however, is that the compound was quite unstable - stored in an air-tight bottle flushed with argon, it darkened quite quickly, becoming black (but still crystalline, for the most part) within a month or so. Could have been due to impurities, or perhaps light was the problem - I never did investigate as I ended up not needing the material.

 Pages:  1