Sciencemadness Discussion Board - Optimizing Henry reaction conditions (substituted benzaldehydes + nitromethane)

Optimizing Henry reaction conditions (substituted benzaldehydes + nitromethane)

Melgar - 10-9-2017 at 11:14

So, initially I was confused by how every single write-up of a Henry reaction seemed to use a totally different solvent, totally different catalyst, different reaction time, different workup, different reduction method if the nitrostyrene wasn't the desired product, etc. And there never seemed to be an explanation for the choices that were made. So I figured I'd put some effort into understanding and optimizing this reaction. Since the precursors are easy to come by, I figured I'd limit the scope to reacting substituted benzaldehydes with nitromethane, to yield a nitrostyrene.

1) Dehydrating the initial nitroalcohol
Fortunately, dehydrating the nitroalcohol doesn't require much effort at all for benzaldehydes; it's actually harder to not dehydrate it. Simply heating to 40C or more will dehydrate it, if it's not dehydrated already.

2) Use glacial acetic acid as a solvent?
I noticed this being done quite often, and had no idea why. Seeing as this reaction is base-catalyzed, an acidic solvent didn't seem like a good idea. Some research led me to realize that if water isn't present, then GAA doesn't really behave as an acid for the purposes of the Henry reaction, but that still didn't tell me anything about why anyone would choose to use it. It smells quite strongly, and other solvents are much cheaper, after all.

3) How should water be removed to drive the reaction forward?
This too, seemed to be glossed over in any examples I looked at. Acids would stop the reaction by reacting with the base catalyst, so silica gel and polyphosphoric acid were out. Basic desiccants tended to be heterogeneous, and only proceeded very slowly. They also tended to have a lot of the product stick to them. A common solutions was to use a nonpolar solvent, especially an aromatic one like toluene, but this could make workup difficult. Another solution mentioned was to use nitromethane as the solvent, although this seemed slow and rather wasteful.

4) What catalyst to use, and how much of it?
Primary amines seemed to be a popular choice for the reaction of benzaldehydes and nitromethane. They'd be used in anywhere from stoichiometric amounts to very small catalytic amounts. Quantity of the base seems to affect both reaction rate and purity. Catalysts were usually not OTC though, and some examples included methylamine, cyclohexylamine, and octylamine.

5) Temperature?
Lower temperatures seemed to result in slower reactions but better quality yields. About the lowest temperature possible seemed to be roughly 40C, that is, the nitrostyrene dehydration temperature.

For the base catalyst, I had phenylethylamine HCl from Purebulk.com, which came out to about $20 for 500 grams. Neutralizing the HCl with NaOH just resulted in a mess, but melting the salt and adding KOH flakes produced a sticky KCl paste on the bottom and sides of the container, that did not mix with the PEA base at all. The PEA base could then just be poured off into a separate container. It also worked great as a Henry reaction catalyst. PEA can also be obtained by phenylalanine decarboxylation.

When the reaction is complete, it's necessary to add water and acid to neutralize the base. Once the reaction is over, it's then important to keep the nitrostyrene acidic, or else the various base-catalyzed reactions will resume. This seemed to explain why GAA was a common solvent then; you'd only need to add water, at which point the GAA would commence behaving as an acid. That didn't seem like the best explanation though.

I realized why GAA was such a common solvent, when adding the base too fast without stirring immediately. Certain areas of the vessel with high concentrations of solvent began generating reddish-brown side products. Since the dehydration of the nitroalcohol produces water, and water can cause this reaction to generate side-products, it's important to remove the water fast enough that it doesn't interfere with the reaction. For this reason, it's common to use a small enough amount of base that it can't generate water faster than it can be removed. However, another solution is to use GAA as the solvent. In this case, if water begins to accumulate, then the GAA will start to act as an acid, putting the brakes on the reaction. When sufficient water is removed, the GAA stops acting as an acid, and the reaction resumes. Clever.

However, for best yields, the ideal solvent would dissolve the reactants fairly well, but would not dissolve the nitrostyrene well at all. Since methanol, ethanol, and isopropanol all have this property, this is certainly within the realm of possibility. Another property of an ideal solvent would be if it formed an azeotrope with water. That way, water could be removed via its azeotrope with the solvent. Isopropanol has an azeotrope that's 15% water, although a lot of isopropanol is needed to remove water this way. A better solvent for this purpose is n-butanol, with approximately a 50% azeotrope with water. However, it has the disadvantage of dissolving nitrostyrenes well.

Methanol is the only common solvent I'm aware of that forms an azeotrope with nitromethane, and in fact, repeated crystallizations from methanol are required to remove excess nitromethane if it exists at the end of the reaction. However, for this reason, methanol should not be used as a solvent for the Henry reaction with nitromethane. Since ethanol is prohibitively expensive, the logical choice is isopropanol.

Of course, it's possible to have the advantages of all these solvents by mixing them. I've found that using at least 50% isopropanol, then equal parts n-butanol, and glacial acetic acid is pretty close to an ideal solvent system, provided the nitromethane is only in 20% excess or less. (Nitromethane is an excellent solvent of nitrostyrenes.) This allows all the advantages of the constituent solvents, plus nitrostyrene crystals will precipitate out as they form. The reaction can be run slow, since the GAA prevents side-reactions from the presence of water, and the n-butanol removes water via azeotropic evaporation. I ran this reaction for a few days in a TLC jar, and the solution was a yellow-orange color without a hint of the red color that tends to accompany side-reactions. The bottom of the jar was covered with enormous yellow crystals, which I had to remove, because they were insulating the reaction vessel from the heat source underneath. Heat source was a $3 coffee cup warmer I bought on eBay from China. I've included pictures of the crystals I've collected so far, and already more are growing. Of course, with crystals like this, recrystallization seems pointless. Not that I mind.

[Edited on 9/10/17 by Melgar]

alking - 11-9-2017 at 05:11

Awesome write up! Well done. I've wondered about a lot of these things myself, some I figured out, and I think just about the rest you did. I do have one important question that remains however and that is what variables go into choosing a catalyst? I noticed that methylamine works well on unsubstituted catalysts for instance, I can get an ~80% yield, yet the more substituted the nitrostyrene the more side products are formed. On a tri substituted compound for instance there is a considerable amount of polymerization unless GAA is used as a buffer and even then there's quite a lot. I've read from one reference that isopropylamine works better for substituted nitrostyrenes, but no reason or source for the claim was given. I do happen to have a big bag of PEA I purchased years ago as a supplement.

Melgar - 11-9-2017 at 11:49

Quote: Originally posted by alking

Awesome write up! Well done. I've wondered about a lot of these things myself, some I figured out, and I think just about the rest you did. I do have one important question that remains however and that is what variables go into choosing a catalyst? I noticed that methylamine works well on unsubstituted catalysts for instance, I can get an ~80% yield, yet the more substituted the nitrostyrene the more side products are formed. On a tri substituted compound for instance there is a considerable amount of polymerization unless GAA is used as a buffer and even then there's quite a lot. I've read from one reference that isopropylamine works better for substituted nitrostyrenes, but no reason or source for the claim was given. I do happen to have a big bag of PEA I purchased years ago as a supplement.

For the catalyst, you want a strong enough base that the reaction will proceed at a fairly rapid pace. Primary amines work well because with them, the reaction proceeds a bit differently, because water is eliminated at the start when the imine forms. As a result there's less of a chance of it interfering later. In this case, the nitroalkane reacts with an imine, not an aldehyde, but to the same end result. Just, instead of water being eliminated, the original imine is.

I chose PEA because it was aromatic, stable, and a rather strong base for a primary amine. It's also very OTC, and has a high enough boiling point that evaporative losses would be negligible. Like most amines, its solubility in water is significant, and thus it'd wash away almost entirely in a water rinse. It's also a very harmless, and totally legal member of the phenethylamine family, and so having a small amount in your end product will almost certainly not have noticeable negative effects.

I really don't know what the benefits of other catalysts are, other than the ones that you use when the nitroalcohol is what you actually want.

Cryolite. - 11-9-2017 at 12:42

What was the benzaldehyde here? Just unsubstituted stuff, or something phenolic or otherwise electron rich?

Crowfjord - 11-9-2017 at 13:34

I've also recently started investigating 2-phenylethylamine as a catalyst for nitrostyrene formation. In addition to the reasons Melgar gives above, my reasoning for its use also include stearic bulk, which should help it to eliminate more easily from the addition product. I used it as the acetate, formed in situ from the HCl by adding the stoichiometric amount of sodium acetate and a little acetic acid as solvent. I found it to work well with the one benzaldehyde I tested, 5-bromoveratraldehyde.

Melgar - 12-9-2017 at 09:53

Quote: Originally posted by Crowfjord

I've also recently started investigating 2-phenylethylamine as a catalyst for nitrostyrene formation. In addition to the reasons Melgar gives above, my reasoning for its use also include stearic bulk, which should help it to eliminate more easily from the addition product. I used it as the acetate, formed in situ from the HCl by adding the stoichiometric amount of sodium acetate and a little acetic acid as solvent. I found it to work well with the one benzaldehyde I tested, 5-bromoveratraldehyde.

I know I spent a ridiculous amount of time trying to figure out how to isolate the base from the salt without distillation, including doing things like adding reactive metals to the molten salt. All to no avail; salts would precipitate out as soon as the liquid was added to a nonpolar solvent. The magic bullet was potassium hydroxide though. (Potassium carbonate would probably also work) Turns out that potassium salts barely dissolve at all in anything that isn't water, and potassium hydroxide would form a nice clean phase separation, where sodium hydroxide didn't at all. Since I noticed this phenomenon, I've been trying KOH for a lot more things lately.

What you did would probably work, (and if you got it to work, then obviously it did) but it's nice being able to remove as many variables as possible. It's also nice to have a 100 mL reagent bottle with a transparent, viscous yellowish liquid in it that's labeled "phenylethylamine, free base" on it, that I can pipette a few mL out whenever I want.

Quote:

What was the benzaldehyde here? Just unsubstituted stuff, or something phenolic or otherwise electron rich?

There were a few oxygens attached to the ring. No acidic phenol hydrogens though.

[Edited on 9/12/17 by Melgar]

alking - 12-9-2017 at 16:24

Quote: Originally posted by Melgar

Quote: Originally posted by Crowfjord

Why not do an a/b to isolate it? Dissolve HCl in water, basify with NaOH/KOH, extract with some low bp NP, DCM, Ether, Chloroform, etc. Strip the NP then you should have rather pure PEA.

zed - 12-9-2017 at 18:24

Phenethylamine as catalyst! Now, that is a good idea!

I have run such reactions with butylamine as catalyst, a small volume of absolute ethanol as solvent, and a week of gentle heat to promote the reaction. It worked very,very well.

Still, not as well as your procedure seems to work. A bonus, is the large crystals you produced.

Smaller crystals are more pure, but in this reaction......somewhat counter-intuitively, bigger is better.

Rinsing reaction crud off of tiny crystals, inevitably leads to a significant losses of product . Bigger crystals, equals less total surface area, equals higher yield.

Nice! Thank you.

[Edited on 13-9-2017 by zed]

[Edited on 13-9-2017 by zed]

Melgar - 13-9-2017 at 01:33

Quote: Originally posted by alking

Why not do an a/b to isolate it? Dissolve HCl in water, basify with NaOH/KOH, extract with some low bp NP, DCM, Ether, Chloroform, etc. Strip the NP then you should have rather pure PEA.

The problem here is the high affinity that PEA has for water. Any nonpolar solvent with significant amounts of PEA dissolved in it, would then dissolve water as well, on account of the PEA pulling it into that layer. And it gets worse. The water that's pulled into that layer will pull some alkali hydroxide along with it, meaning you can't really count on the alkali hydroxide's desiccating properties, since it's present in both layers. Or to simplify, PEA base is too polar to go into the nonpolar phase preferentially.

However, what I did actually is a variant of the standard A/B extraction, just solventless. Although I don't add water, water does form on account of the HCl neutralization (edit: also, because commercial KOH always contains a percentage of water), and that small amount of water is actually enough to dissolve both KOH and the amine HCl salt, thus allowing the reaction to continue. What I've figured out though, is that whether you use KOH or NaOH is actually quite important, since the main difference between sodium and potassium salts is their solubilities in different solvents. And whether an A/B extraction works or not hinges on those differences.

Oh, also, the way I did it, there is essentially no workup, which I see as a major advantage. I did notice that when rinsing out the potassium salts, the PEA base DID form a separate layer, unlike with the corresponding sodium salts, and even when there was a lot of water present.

Quote:

Phenethylamine as catalyst! Now, that is a good idea!

I have run such reactions with butylamine as catalyst, a small volume of absolute ethanol as solvent, and a week of gentle heat to promote the reaction. It worked very,very well.

Still, not as well as your procedure seems to work. A bonus, is the large crystals you produced.

Smaller crystals are more pure, but in this reaction......somewhat counter-intuitively, bigger is better.

Rinsing reaction crud off of tiny crystals, inevitably leads to a significant losses of product . Bigger crystals, equals less total surface area, equals higher yield.

Nice! Thank you.

You're welcome!

I initially tried to use phenylethylamine when attempting to run the reaction in toluene. I figured I should use a base that was structurally similar to toluene to increase its solubility, and benzylamine has way too many weird exceptions regarding the way it reacts. Phenethylamine could be purchased at health supplement stores, so it just seemed like a really obvious choice, even though I couldn't find any literature examples that used it. I guess if you have access to commercial chemical suppliers, PEA wouldn't be a first choice, but it does work great as an OTC catalyst.

I'm curious how your water was removed in the reaction you described though? Just ordinary evaporation? I'd expect the ethanol to evaporate very quickly, resulting in a solventless reaction, or a reaction where nitromethane was the solvent. Did you use an excess of nitromethane then?

One aspect of this reaction that seemed almost counterintuitive, is that crystals won't form unless the isopropanol concentration is high enough. Initially when I was adding isopropanol, I only added it because it was cheap, and because my reactions that worked best all seemed to include it. But then I noticed how much better the reactants dissolved in it than the product did, and realized that this was a solubility difference that could be exploited. I'm wondering now if it'd be possible to greatly improve yields using this procedure, seeing as the product is removed from the reaction almost as soon as it forms.

I'm debating whether or not to get into nitrostyrene reduction to amines, in this thread. Lately I've been surprised at how well Zn/HCl works for reducing nitro groups and all the associated reduction intermediates, considering nobody seems to take it very seriously. However, zinc has a very high hydrogen overvoltage, meaning that if there's anything around to reduce, it'll reduce that rather than generate hydrogen. Seeing hydrogen bubbles is actually a good indication of reaction completion.

The downside of this is removing zinc salts, but I've been experimenting with using phosphoric acid (combined with HCl to expose new metal) to react with the zinc. Zinc phosphate is very insoluble in just about everything, so it could be removed by filtering.

I would have tried zinc/HCl sooner, except that my results with Zn/GAA were terrible, and I incorrectly assumed that a stronger acid would cause more problems.

[Edited on 9/13/17 by Melgar]

alking - 13-9-2017 at 08:29

Zinc/HCl is strong enough to reduce a nitro group to an amine? I thought it would stop at the imine/oxime stage?

clearly_not_atara - 13-9-2017 at 09:42

Zn/HCl reduces terminal nitrostyrenes efficiently. It fails to efficiently reduce secondary nitroalkenes, instead producing a ketone.

Secondary nitroalkenes can be reduced by activated aluminum or via the oxime (many methods) which is reduced by metallic Na/EtOH.

Melgar - 13-9-2017 at 13:40

Quote: Originally posted by alking

Zinc/HCl is strong enough to reduce a nitro group to an amine? I thought it would stop at the imine/oxime stage?

As far as I can tell, that combination will reduce virtually any nitroalkane to an amine, and even nitrostyrenes (but not nitropropenes) to amines. Reduction of nitrostyrenes isn't very clean unless temperatures are controlled, although an ice/salt bath is reported to work well.

Corrosive Joeseph - 13-9-2017 at 14:24

Zinc-HCl reduces aromatic nitro and aliphatic terminal nitro groups to amines very well.

Secondary nitroalkenes give a mix of products which hydrolyze in excess acid to the ketone.

Don't forget this - http://orgsyn.org/demo.aspx?prep=cv1p0413

Attached here is by far the best review available on the net for nitrostyrene/nitroalkane reduction.

Apologies for slight topic drift. I will whip myself later.

/CJ

Attachment: An_Experimental_Evaluation_of_the_Zinc_Hydrochloric_Acid_Reduction_of_Nitrostyrenes_Final.pdf (265kB)
This file has been downloaded 1694 times

AvBaeyer - 14-9-2017 at 05:29

CJ:

What is the source of the review? It is well done and extremely interesting.

AvB

kmno4 - 14-9-2017 at 07:09

Quote: Originally posted by Melgar

Optimizing Henry reaction conditions (substituted benzaldehydes + nitromethane)

Hm.....
I read this and I cannot see any "optimizing". First thing, Henry reaction gives nitroalcohols, their dehydration has nothing to do with same condensation and its conditions.
How many substituted benzaldehydes have you tested, 10, 20... ?
How many amines have you tested ? Yields ?
It is known, that experimental results strongly depend on kind of benzaldehyde/amine or base/temperature/solvent(s) system and its acidity.
Some conditions are more universal, others are not.
There are many cheap amine catalysts (ammonium acetate, urotropine... etc).
What you have written here, has nothing to do with the tile above. Literature about various variants of Henry reaction is very extensive.

Melgar - 14-9-2017 at 11:50

Quote: Originally posted by kmno4

The idea was to come up with a set of considerations that are important when deciding your reaction conditions, and then describe in a general sense, how they interact with each other. I tried to include as much information as I could that was particularly hard for me to find, and tried to keep it open-ended enough to use as guidelines for optimizing this reaction generally. And I obviously don't know everything about this reaction, but I thought it would be useful to start a conversation. That's why the title was "Optimizing Henry reaction conditions" and not "Optimized Henry reaction conditions".

Also, if you're saying that the reaction name is wrong, then please enlighten me as to what the dehydrating version of the nitroaldol reaction is called? When a primary amine is used as a catalyst, it wouldn't be accurate to call it a Henry reaction with subsequent dehydration, which is what I usually see it called. I know it wouldn't be technically correct to refer to it as a nitroaldol reaction, since there's no nitroaldol product, so I'd think it'd be most accurate to refer to it as a variant of the Henry reaction, no?

I'd like to work more on figuring out how functional groups affect optimal reaction conditions, since I have yet to find any good information that explains this. Hell, the nitroaldol reaction page on wikipedia says that solvent and catalyst selection doesn't affect yields at all! And I've never even seen a description of the nitroaldol/Henry/Knoevenagel reaction where it's indicated that water is being removed via azeotropic evaporation, even though there are writeups of reactions where this is obviously the case.

[Edited on 9/15/17 by Melgar]

Crowfjord - 14-9-2017 at 12:52

I think it's a Knoevanagel condensation when an amine catalyst is used. This might affect the information found when gathering data.

Melgar - 14-9-2017 at 22:54

Quote: Originally posted by Crowfjord

I think it's a Knoevanagel condensation when an amine catalyst is used. This might affect the information found when gathering data.

Knoevanagel condensations can apparently refer to any aldol-type reaction catalyzed by any type of amine, unless I'm mistaken. And if that is what they should be called, well, there are plenty of books and published papers out there that don't call them that.

There seems to be a debate on what to call this reaction in this Vespiary post too, which as a bonus, gives reaction conditions for a lot of benzaldehyde + nitromethane reactions. Those conditions aren't all optimal, of course, but it's certainly good to have conditions that have been published and proven to work at least moderately well:

https://www.thevespiary.org/talk/index.php?topic=8824.0

edit: A convenient modification to this reaction, especially for those nitroalkenes that oxidize in air, is to rinse them with an ascorbic acid solution before putting them aside for later. The acidity and antioxidant properties of the ascorbic acid are very helpful for preventing degradation.

[Edited on 9/15/17 by Melgar]

Crowfjord - 15-9-2017 at 07:13

Quote: Originally posted by Melgar

Yeah, you're right. Knoevanagel condensations can use aromatic or tertiary amines, so that doesn't really capture this specific reaction; the mechanisms can differ depending on catalyst chosen. For example, if a nitroalkane is reacted with an aldehyde using a tertiary amine as base, the nitroalcohol is formed, at least as an intermediate. Nicodem has asserted in the past that the primary amine-catalyzed condensation is a Knoevanagle, for example in this thread. This reaction is probably still a subset of Knoevanagel condensations, but the literature is loose on the naming conventions in this particular case. It certainly complicates the research.

Corrosive Joeseph - 16-9-2017 at 12:03

Quote: Originally posted by AvBaeyer

CJ:

What is the source of the review? It is well done and extremely interesting.

AvB

@ AvB - I pulled it from another forum but I believe the document was originally authored and posted on the HyperLab forum by a user known as 'persona'.

" an experimental evaluation of the Zinc Hydrochloric Acid reduction of nitrostyrenes - 23.11.2007 - www.hyperlab.info - Post #514742"

Unfortunately the forum is in Russian but the quality of work and write-ups there is second to none.

/CJ

Melgar - 16-9-2017 at 18:17

Saying this reaction is actually a Knoevanagel condensation is like saying that a platypus is actually a vertebrate. While it's technically true, it's not a very useful way to categorize it.

Although I had anticipated this being the case, it has become more obvious that the crystals you collect will be smaller in subsequent crops, and the last one will require extracting a rather small amount of nitrostyrene from a rather large amount of solvent. Nitrostyrenes only have a medium-to-low solubility in this solvent mixture, but that's a lot more than zero. However, if you have plenty of time, I see no reason that it wouldn't be possible to start on a small scale with only a fraction of your reactants, then top off all your solvents and add more reactants every time you collect crystals from it. I left for about a week, and when I came back, it had changed from the color of about apple juice, to the color of tea. However, I didn't smell any butanol, so my suspicion is that it ran out, and wasn't able to get rid of water as efficiently. Or something like that. N-butanol seems to play an important beneficial role that I don't quite understand yet, probably via the formation of an azeotrope that I don't know about.

zed - 18-9-2017 at 18:04

Actually, I primarily used absolute ethanol and butylamine, in the condensation of nitroethane with benzaldehydes. Very good yields were obtained without physical removal of water from the reaction medium.

When it comes to H2O; ethanol has a powerful grip. It simply rips off the water, and refuses to give it back. Produces a big solid block of crystals and a small amount of residual liquid ethanol/butylamine/H2O/with a little dark crud......in it. With large nitroalkene crystals created, a little ice-cold ethanol rinses the residual liquid/gunk right off, without very much loss of product.

Not as interesting as what you are doing. You have evolved a very thoughtful approach.

New to me. Different solvent, different catalyst, different way of driving the reaction.

[Edited on 19-9-2017 by zed]

[Edited on 19-9-2017 by zed]

Gl3n - 30-6-2018 at 13:19

I just came across this while utfse, sorry if I’m bringing a dead post up but I just read this and did have what might be additional useful information.

It was mentioned that this reaction might not run below 40C, I have had it run at no more then ambient temperatures 32-50F? in an uninsulated garage in the middle of winter in the north-east US. I followed the methods as listed on Rhodiums Chemistry page under Ultrasound Promoted Nitrostyrene Synthesis. While the true temperatures at the site of the reaction may be much greater then the ambient temperature I had the reaction run to 90% of the theoretical yield after about 24 -48hrs with the US bath not exceeding 25C. This method utilizes ammonium acetate as a catalyst, GAA/nitromethane as solvents and in my experience, compared to refluxing using published amounts of the same reactants results in a much cleaner reaction without the nasty side products. I am extremely interested in combining this methodology with US as in a closed system it would eliminate solvent loss and if the results are anything like what have been shown it would make for a simple, odor free, low cost system.
It does however leave the question of whether Ultrasound results in a nitroalcohol or nitrostyrene? My MP determination was based on flame sealed capillary tubes and a remote sensing digital barbecue thermometer. The MP was sort of all over the place so I don’t know that I did not actually produce a beta hydroxy nitrostyrene instead of a nitrostyrene. Now that I think about it I have been interested in making some beta hydroxy compounds but I guess it’s possible that’s what I had all the time. To be honest I never did anything with the results of those experiments so maybe I should test them and try to figure out what I actually produced.

Gl3n - 4-7-2018 at 17:39

Melgar,
What molar ratio of PEA did you utilize? Or volume per molar reaction size etc? I have also run a Leminger reaction to at least 80%, higher yields resulting from a slower, temp controlled, reaction. No non-polar utilized only dry IprOH washes after basification, The salts prevent the water/alcohol layers from mixing and the dry IprOH prevent ZnOH salts and hydrates from forming. No clogged up filters or emulsions etc.
Also how bad did the liquefaction of PEA HCl smell? If this were attempted should it be carried out in a fume hood?

[Edited on 5-7-2018 by Gl3n]

alking - 5-7-2018 at 11:22

PEA should not smell much due to the high bp.

happyfooddance - 5-7-2018 at 12:26

Quote: Originally posted by alking

PEA should not smell much due to the high bp.

This makes no sense. Phenethylamine salts fume profusely when heated to their melting point. Additionally, PEA HCl smells quite a bit at room temperature. Most importantly, the boiling point of PEA freebase is lower than the melting point of the the hydrochloride salt. So when you free the base with KOH in the melt, your PEA will be superheated and boiling. On top of it all, adding KOH is probably exothermic in this case.

So I would suggest do it in a hood, or outside, and try it on a small scale first (like any sensible chemist would).

alking - 6-7-2018 at 08:05

I thought you were asking if PEA Freebase smells at RT or with use in this reaction, not if you boil it off lol. Of course it would smell a lot if you're boiling it.

At RT it has a faint odor, not anything to worry about. You don't need to melt the HCL salt or superheat it. The reaction at RT/moderately heated would be exothermic, but not anything approaching the bp. I've done this with just enough water and heat to get it to react efficiently, separated the PEA, and then dried it with Na2SO4 to pull any water out. It worked fine and there's less smell than a can of tuna.

Gl3n - 7-7-2018 at 18:33

To all involved:
I took into consideration that PEA HCl liquifies at at greater temp then the freebase will volatilize. I answered this issue with Melgars original info. Using all modern safety and regulations... And INSIDE AN APPROVED FUME HOOD, I made a 20molar KOH solution, to a reaction vessel( 150x25mmborosilicate) added a 1/20 molar qty of PEA HCl. I then added 1.5 Mol qty of 20m solution. It slowly and if anything endothermically reacted. It got cold and separated into a thick salt layer on bottom and lighter amine on top. The melting of an HCl with a higher Bp then adding a salt to form a lower bP amine sounded dangerous. I’m happy to report in essence using a saturated KOH solution works as well as liquid HCl salt

Gl3n - 7-7-2018 at 19:21

I had attempted to utilize Melgars ratios for large single crystals. I will admit it was ambiguous as fuck add 50%IPrOH, 25%1- butanol, 25%GAA. No mention of nitroalkanes besides “below 20% excess”. I honestly hold all of the old members in extremely high regard but some things do not lend themselves to repeatability. I’m not trying to mock him at ALL but I think all scientists published or not need to ensure their methods are repeatable and easily understood hence moles and shit like that. While here in the US our research stations crumble around the fume hoods for lack of funding

Metacelsus - 7-7-2018 at 19:51

I find that the refluxing acetic acid + ammonium acetate method is more reliable than the alcohol / base method. Nitromethane is cheap enough that a 5x excess can be used, in which case yields are quite good (typically 75–85%).

So far I have not attempted reductions of any nitrostyrenes. However, the Zn/HCl method seems like the most useful one, given that LiAlH4 is much harder to get (and also pretty dangerous).

[Edited on 7-8-2018 by Metacelsus]

Melgar - 19-7-2018 at 08:45

For anyone frustrated by all the missing details in my writeup, I guess the reason for that is that it wasn't really supposed to be a writeup. It was more like "I was experimenting with synthesizing nitrostyrenes using various solvent systems, and seem to have figured out something new, though I'm not sure what. Anyone feel like having a discussion about this?"

Quote: Originally posted by happyfooddance

Quote: Originally posted by alking

PEA should not smell much due to the high bp.

Okay, so you know how I said that I melted the PEA base and added KOH? Not only does this produce KCl and PEA base, it also produces substantial amounts of water. Since water has a much lower boiling point than PEA base, you will indeed get substantial amounts of water evaporating, and this keeps everything well below PEA base's boiling point. Not only is there water forming from the reaction, but KOH is always going to have a significant fraction of its mass be water initially, so this isn't something you really have to worry about.

I prefer doing it this way because the salt layer ends up as a slushy consistency that's more solid than liquid. You can just pour the PEA base off straight from a beaker, and the K-salt layer will stay stuck to the sides. This also combines several steps, and eliminates the need for titrating your KOH to figure out how much water it's absorbed and how much of it has converted to the carbonate salt. This means I can use my "old" KOH, in which the flakes have a powdery white carbonate coating from spending too much time exposed to air. It's always nice when you can find a use for your less pure reagents.

Anyway, you just add a moderate excess of KOH beyond what would be needed to neutralize the HCl salt. KOH is a strong desiccant too, so whatever doesn't react with the amine HCl salt will pull water out of it, which is important because PEA base is very soluble in water. I should also mention that I heated the vessel that I did this in, to just over water's boiling point, which sped things up a lot. Then I let it sit for a few hours to cool off, and give the K-salt layer sufficient time to desiccate the PEA base.

PEA HCl does not smell at all, or at least mine doesn't. This would be expected, since ammonium chloride doesn't smell, and neither does any pure amine HCl salt I know of. Any smell is almost certainly due to impurities. The free base does smell, although it's not particularly strong or unpleasant. Lots of things smell far below their boiling points or melting points. Vanillin, for example.

Quote: Originally posted by Gl3n

I had attempted to utilize Melgars ratios for large single crystals. I will admit it was ambiguous as fuck add 50%IPrOH, 25%1- butanol, 25%GAA. No mention of nitroalkanes besides “below 20% excess”. I honestly hold all of the old members in extremely high regard but some things do not lend themselves to repeatability. I’m not trying to mock him at ALL but I think all scientists published or not need to ensure their methods are repeatable and easily understood hence moles and shit like that. While here in the US our research stations crumble around the fume hoods for lack of funding

Yes, I admit to running like ten experiments at once, and not being especially diligent with documenting them. I enjoyed visualizing the progression of this reaction using different solvent systems. And I certainly left out a lot of information in my writeup.

One thing I like about this reaction is that it can be run under a wide variety of solvent systems. Another thing I like about it is that you can tell how the reaction is progressing by monitoring the color change of the solution. Nitrostyrenes are typically yellow, and when a lot are present, the solution tends to look orange. Side products of this reaction are reddish-brown, and seem to form in the presence of water. Since the initial step involves an imine formation and the ensuing production of a water molecule, I started with about 2% molar mass (compared to benzaldehyde) of PEA free base. I figured this would be low enough that most of the water would be long gone (via azeotropic evaporation) before the reaction would progress to the point where it could interfere. If you're attempting to speed this reaction up though (I wasn't), then perhaps you could add PEA in portions, separated by a few hours. The goal is to make sure that the quantity of catalyst is low enough so that water concentration is as low as possible at any given time. By opting for a slow reaction, you also give your crystals plenty of time to grow, and thus you can get pretty big crystals this way.

However, because I was removing water via azeotropic evaporation, it seemed to matter a lot that I occasionally replace solvents. The shape of the container also seemed to be important, and ones with narrow openings obviously lost solvent at a lower rate. The reaction that I ran with n-butanol, GAA, and IPA had the most impressive results, but there seemed to be a complex system of azeotropes that I wasn't sure what to make of. IPA seemed to evaporate the most quickly, so I had to add new IPA more often than the other two. IPA is critical to this method, because it's responsible for the low solubility of the product, which will then precipitate crystals and remove itself from the reaction.

Of course, if you're removing water via azeotropic evaporation, you have to worry about the reactants evaporating too. The only one that evaporated significantly was nitromethane. So if you start with no excess of nitromethane, then there will not be enough of it for the reaction. But if you start with too much, then it will significantly dissolve the formed nitrostyrene, preventing it from forming huge crystals.

So, to conclude, I was vague was because I'd been topping off the solvents as they evaporated, since my goal was to combine the benefits of GAA (becomes acidic in the presence of water, stopping the reaction) and of IPA (high solubility of reactants, low solubility of product) and of n-butanol (forms a useful azeotrope with water).

[Edited on 7/20/18 by Melgar]

morganbw - 19-7-2018 at 11:35

Quote: Originally posted by Gl3n

I think that if you look at his post not as how to do things exactly but as a soup of what might be possible. I took his post, not as a how-to, but more as to what are the possibilities. Much more a post to request a discussion rather than to state that he has all the answers.

This could be a good conversation. I personally would use well-vetted literature to guide me in an effort to do a somewhat similar synthesis.

This is a subject which does interest me, sometimes discussions are just that and if lucky a rare jewel is discovered.

Melgar - 20-7-2018 at 07:14

Quote: Originally posted by morganbw

Yeah, that was how I'd intended it. The photos and the description of the reaction conditions I ran was more of a proof-of-concept than anything, indicating that this could be a very fruitful direction for future research. However, if someone wanted to do this in a more systematic way that's more conducive to replication, I'd recommend starting by researching what azeotropes form from a mixture of water, isopropanol, GAA, and nitromethane. Then see if a different azeotrope forms when n-butanol is added to the mixture. Once it's established what azeotrope is removing water, then the solvent could be pre-mixed at that ratio, and added to the reaction flask as it evaporated. I would imagine that some nitromethane should be added to the pre-mixed solvent too, to account for the slow loss of it from the system. My theory is that the azeotrope that removes water will dominate when water is present, but some other azeotrope that removes nitromethane also manifests at times. Isopropanol, for instance, has a very significant azeotrope with nitromethane, which is something I learned only recently.

edit: Also, for anyone curious as to the reaction mechanism, the base that actually does the catalyzing is almost certainly the imine, NOT the amine. Imines are actually fairly basic in their own right, and certainly basic enough to catalyze this reaction. So if you're worried about making sure that there's enough free amine to catalyze this reaction, don't be. This is actually why methylamine works fairly well as a catalyst despite having a very low boiling point: it's almost never present as a free amine.

[Edited on 7/20/18 by Melgar]

alking - 28-7-2018 at 07:03

When I've done this reaction before with i-PrOH/GAA as the solvent I have always used a moisture trap. Nothing to remove the water created, but just to prevent more from absorbing into the solution since it's (initially) anhydrous.

If you do this in an open vessel over a matter of days wouldn't that become an issue? I would think at only 40-50C it would absorb atmospheric moisture faster than it drives it off as an azeotrope.

monolithic - 7-9-2018 at 10:41

Quote: Originally posted by Corrosive Joeseph

Zinc-HCl reduces aromatic nitro and aliphatic terminal nitro groups to amines very well.

Secondary nitroalkenes give a mix of products which hydrolyze in excess acid to the ketone.

Don't forget this - http://orgsyn.org/demo.aspx?prep=cv1p0413

Attached here is by far the best review available on the net for nitrostyrene/nitroalkane reduction.

Apologies for slight topic drift. I will whip myself later.

/CJ

Has anyone here run the reaction as described in this paper (low temperature without any catalytic amine)? It seems superior to the traditional conditions of elevated temperatures and presence of a catalyst, in terms of simplicity and yield, but I can't seem many other cases of people carrying it out like this.

Corrosive Joeseph - 14-9-2018 at 02:36

Quote: Originally posted by monolithic

OrgSyn procedures are usually rock-solid. The 'low temperature without catalytic amine' works for nitromethane but not for longer chain nitroalkanes. Just don't forget this -

"5. The alkaline solution must be added slowly to the acid, for the reverse procedure always forms an oil containing a saturated nitro alcohol. A large excess of acid at room temperature is used, conditions which facilitate the formation of the desired unsaturated nitro compound."

Oh BTW, I have heard great things regarding ethanolamine/acetic acid catalyst @ RT, based on the attached paper.

/CJ

Attachment: alizadeh2010.pdf (1MB)
This file has been downloaded 652 times

monolithic - 14-9-2018 at 06:15

Quote: Originally posted by Corrosive Joeseph

Quote: Originally posted by monolithic

Awesome find. Thank you for your contributions.

Melgar - 29-10-2018 at 08:52

Quote: Originally posted by alking

When I've done this reaction before with i-PrOH/GAA as the solvent I have always used a moisture trap. Nothing to remove the water created, but just to prevent more from absorbing into the solution since it's (initially) anhydrous.

If you do this in an open vessel over a matter of days wouldn't that become an issue? I would think at only 40-50C it would absorb atmospheric moisture faster than it drives it off as an azeotrope.

Nope. The way azeotropes work is by having a lower boiling point than either of the constituent compounds. That also translates to having a higher vapor pressure, when the solvents are below the azeotropic boiling point. So any water that entered into the system would raise the vapor pressure, increasing the evaporation rate until it was gone, at which point the system would go back to having the lower vapor pressure consistent with an anhydrous solvent system.

The other key part here, is that you have to make sure that the solvent is warmer than the environment, so water can't condense on its surface. Also, that the evaporation rate is high enough to remove the water as an azeotrope as it forms.

So as long as solvent is evaporating at a positive rate, and the solvent is warmer than its surroundings, that shouldn't be a problem.

Cactuar - 13-11-2018 at 15:31

Quote: Originally posted by Melgar

This is not correct. An azeotrope can also be higher boiling than either of its constituents. And an azeotrope (or atleast the values you find online) is measured at mixture reflux. At room temperature or slightly above these values would be completely different or even be zeotropic.

Melgar - 18-11-2018 at 09:48

Quote: Originally posted by Cactuar

Let's just assume I specified "low-boiling azeotropes", since that's what I meant, and move on to the next part, ok?

My understanding of azeotropic evaporation is that what often happens is that the azeotropic solvent mixture cools to significantly below ambient temperature, due to heat loss from evaporation. This can induce water to condense on its surface, which changes the solvent composition, and throws off the azeotropic evaporation rate. Therefore, it should be possible (if this is correct, I could easily be wrong) to prevent this from happening by simultaneously limiting the exposed surface area, and ensuring that the solvent mixture is always ~5°C+ above ambient temperature.

I'm not exactly sure where this notion came from, but I'd appreciate it if someone could confirm a) whether it's definitively true or false b) if this phenomenon happens with azeotropic solvent mixtures that are hydrophobic or do not form azeotropes with water and c) if this is known to happen under an inert, anhydrous atmosphere. A quick literature search seems to indicate that evaporation rates of azeotropes are related in some capacity to their boiling-point azeotropes, though I'm not seeing much that's definitive on the subject.

Experimentally, 95% ethanol was left in two vessels, until half the original volume was gone. One vessel was a flat metal motion-picture film canister that was left at ambient temperature. The other was a much narrower still-picture film canister that was placed on the transformer for a router power supply, about 5°C above room temperature (room temperature fluctuated between 60°F and 66°F). When half of the alcohol was gone from each vessel, an attempt was made to ignite a piece of cotton dipped in each one. (The narrow vessel took longer for half its volume to evaporate, neither was timed) Both ignited, although the warmer one in the narrower vessel ignited noticeably more easily. Both were allowed to evaporate until only a small amount of liquid was left. The narrow, heated vessel ignited easily, but the liquid from the flat, unheated vessel would not ignite. There did not seem to be any drop in concentration in the heated alcohol in the narrow vessel.

Apologies for not having much in the way of equipment and/or instrumentation. This seems to at least demonstrate that water can be absorbed from the atmosphere such that it has an effect on azeotropic evaporation, if nothing else.

Cactuar - 23-11-2018 at 14:24

Quote: Originally posted by Melgar

My understanding of azeotropic evaporation is that what often happens is that the azeotropic solvent mixture cools to significantly below ambient temperature, due to heat loss from evaporation. This can induce water to condense on its surface, which changes the solvent composition, and throws off the azeotropic evaporation rate. Therefore, it should be possible (if this is correct, I could easily be wrong) to prevent this from happening by simultaneously limiting the exposed surface area, and ensuring that the solvent mixture is always ~5°C+ above ambient temperature.

I'm not sure about temperature dropping significantly below RT. I guess it depends on many things like the ambient temperature, the latent heat needed for vaporization, the area exposed and the vapor pressure of the liquid. If you drop acetone, DCM or ether on a watch glass this is clearly the case. However I don't think isopropanol in a beaker would cool down enough to make water start condensing on the walls and greatly change the solvent composition. It does seems logical though that a warmer mixture would have slightly less water in it, although any dry solvent, warm or not, would get wetted when exposed to the atmosphere.

Even if the mixture was zeotropic it wouldn't really matter in an open beaker, water would still evaporate along with isopropanol. I guess it would reach some sort of equilibrium where there is an equal chance of a water molecule being absorbed and one being evaporated. I don't know how they analyze compositions of azeotropes but I'm guessing it has to pass some fractionating column first.