Sciencemadness Discussion Board

Ethyl Pyruvate from Calcium Pyruvate

tyro - 17-4-2023 at 20:38

I've been working on ways to obtain ethyl pyruvate from calcium pyruvate over the past couple months. It's been a bit of an arduous endeavor balancing obtaining the free keto acid with minimal water and preventing various decomposition/condensation products in the process of purification and transformation to the ester.

Some early attempts involved using mineral acids to liberate pyruvic acid from the calcium salt and concentration using toluene and a dean stark apparatus. But, the process of removing water was pretty tedious, and some attempts at purification through distillation ended with copious amounts of tar and degradation products, some of which I presume to be acetaldehyde (strong, sharp, almost noxious, yet fruity odor).

I'll detail a part of what lead me to the process I've investigated and my most recent attempt at it. The end product has not been fully quantified or qualified, but there are signs that this may be a viable route with some further tuning.

Side note: pyruvic acid has such an interesting odor. In solution, it has an intensely sweet smell, almost that of brown sugar. No smell of acid whatsoever. When concentrated, it is sickly sweet and quite sharp like acetic acid. The sweetness in both cases is almost intoxicating. Very unique.

After the mineral acid runs, I strove to find a way to obtain the pyruvic acid from calcium pyruviate using as little water as possible. The first alternative approach I took was to dry grind calcium pyruvate with oxalic acid dihydrate. This seemed to work pretty well.

An equimolar amount of calcium pyruvate and oxalic acid dihydrate were placed in a mortar and intimately mixed. The texture of the mixture transitioned from powder, to cake batter, to thick icing, and finally to a runny icing. Along the way the smell became increasingly brown-sugar sweet, and at the end was sickly sweet with a pungent acidic odor.

dry_grind2.jpg - 213kB dry_grind3.jpg - 213kB

dry_grind4.jpg - 219kB dry_grind5.jpg - 197kB dry_grind6.jpg - 187kB

Unfortunately, vacuum filtering this mixture was pretty inefficient. It took many hours to affect a reasonable separation of solid from liquid, even when diluting the mixture with ethanol which would later be used in the esterification. Probably something on the order of 6-8 hours on the frit. Not super friendly...

On a hunch, I decided to see if I could get similar, but perhaps more easily filterable, results by mixing oxalic acid and calcium pyruvate in the 95% ethanol which would be used for esterification.

Early trials seemed promising from the color changes and changes to odor which occurred. The mixture would pretty quickly change from a pale yellow to more of a creamy white and the brown sugar smell would become pretty dominant. I haven't conducted rigorous analysis of the conversion, but from the perspective of color and odor, the change is pretty quick. Usually within an hour or so. The mixture obtained in this manner is much more easily filtered than from dry grinding.

Early attempts at performing the esterification involved, again, the use of toluene with a dean stark apparatus to drive off water. This still seemed fairly inefficient as a good deal of the alcohol and toluene would separate out in the lower layer. I instead opted to try removing water to drive the reaction forward by using molecular sieves in a soxhlet extractor. Which brings us to the current experimental run.



To a 1L RBF was added 107.1g calcium pyruvate (mw 214.19g; 0.5 mol), 63g oxalic acid dihydrate (mw 126.07g; 0.5 mol), and 250ml 95% ethanol (4.37 mol). This was mixed at room temperature with magnetic stirring for 9 hours. The mixture was then vacuum filtered, and the filter cake washed with two portions of 50ml 95% ethanol and filtered to relative dryness. At this point 350ml of 95% ethanol had been used (6.1 mol).

mix1.jpg - 193kB mix2.jpg - 194kB mix3.jpg - 203kB

It's a bit hard to see, but the yellow color does fade between pictures 1 and 2.

The filtrate had been collected earlier in another 1L RBF. To this was added 4ml of 93.2% H2SO4. To the 1L RBF was attached a soxhlet extractor filled with 200g of 4A molecular sieves which had previously been dried in a microwave. The mixture was refluxed overnight for about 12 hours.

est1.jpg - 231kB est3.jpg - 244kB est4.jpg - 250kB

The mixture was allowed to cool to room temperature and then was diluted with a number of 100ml portions of water (I din't record the exact amount) until there was a separation of phases. The lower non-aqueous phase was removed in a separatory funnel. This was then washed with another 100ml portion of water, but the volume of the non-aqueous phase seemed to decrease a fair bit. After realizing that ethyl pyruvate has a nontrivial solubility in water (10g/L and likely higher with ethanol in the mix), it was decided to extract the collected aqueous phases with DCM. At this point there was just shy of 1L of aqueous phase separated, so this was split in to two portions and each was extracted with 2 x 20ml and 1 x 10ml portions of DCM.

ex1.jpg - 155kB ex2.jpg - 217kB ex3.jpg - 195kB

The first picture here is the organic layer separating from the aqueous layer. No DCM involved. Second and third pictures depict DCM extractions.

The combined DCM extracts were then dried with anhydrous MgSO4, DCM distilled off and the remaining extract distilled under vacuum. The distillate did not come over at a sharp temperature, but somewhere between 90°C-130°C at 130mbar (the majority came over around 100°C-110°C). Very little distillate was collected above 130°C. In total, about 35g of distillate was collected. At this point the setup was allowed to cool to RT. The non distilled portion was a dark red slightly viscous liquid, while the distillate was a mostly clear slightly oily liquid with a very faint yellow tinge. Both the distillate and non-distilled product were placed into a freezer. The distillate remained liquid, while the red product froze.

dis1.jpg - 201kB dis2.jpg - 248kB dis3.jpg - 181kB

The smell of the distillate is quite pleasant. Very close to rum raisin and fruit, with some sharp notes to it. The red liquid has the rum raisin smell too, but a little less fruit and is a bit smoother.

I have not run a rigorous qualification here, so best I can do is speculate. I find it likely that the distillate is a mixture of ethyl pyruvate and pyruvic acid. When running the esterification, I noted that the distillate entering the soxhlet was slightly acidic. I have had trouble tracking down good azeotrope data on pyruvic acid and ethyl pyruvate, so again hard to say what's happening here. Since the distillate was acidic, it's likely that the molecular sieves were in some way degraded. It's also likely that the molecular sieves were not entirely dry prior to the esterification run, though a fair bit of water was removed. But in any case, I don't think molecular sieves would be the best way to proceed. Perhaps anhydrous MgSO4 should be used in the soxhlet instead.

Additionally, I have not run rigorous analysis on the conversion of calcium pyruvate and oxalic acid dihydrate to pyruvic acid and calcium oxalate and what sort of hydrate of calcium oxalate may or may not form in this process. It is likely that in this particular run, some oxalic acid made it into the filtrate. Perhaps the red liquid contains some amount of oxalic acid ethyl ester. The solidification of the red product in the freezer may be due to the oxalic acid ester and/or degradation/condensation products of pyruvic acid and ethyl pyruvate. Again, speculation. It would be worth investigating longer conversion times for this step, or perhaps adding a slight excess of calcium pyruvate to ensure total conversion.

In any case, assuming the distillate is 100% ethyl pyruvate (which I highly doubt), it would represent a pretty miserable yield (~30%).

I do plan to attempt further purification on the products from this run, as well as tweaking some aspects of the approach as mentioned above.

[Edited on 18-4-2023 by tyro]

Boffis - 18-4-2023 at 08:31

Hi tyro; what a really interesting write-up. I have been working with calcium pyruvate recently and have toyed with the idea of preparing this ester myself but I was a bit concerned that the acid catalyst would cause condensation of the ketone group with the methyl group and give a polymeric product (as with acetone + H2SO4 gives mesitylene by trimerization) yielding things like a tri-ester of benzene-tricarboxylic acid so I am glad to hear you got some yield. I checked out the Org synth web site but the procedure there was only for the methyl ester and required some mind bogglingly complex glassware!

The way I was thinking about trying was to suspend calcium pyruvate in excess absolute alcohol and then add just a little less conc sulphuric acid than theoretically required to precipitate the calcium. Then after filtration make acid with either a little sulphuric acid or HCl gas, reflux, then distil off the excess alcohol. Neutralise the residue and fractionate. I also wonder about just passing HCl gas into a hot suspension of the calcium salt in absolute ethanol. The calcium chloride formed should absorb the water and drive the reaction forward once all of the calcium salt had been converted to chloride. But I haven't tried either method.

Pyruvic acid self condenses into a variety of compounds under alkaline conditions to uvitic acid (5-methyl-isophthalic acid) but I couldn't find any info on acid polymerisation. Most references state that it decomposes to acetaldehyde and CO2.

tyro - 20-4-2023 at 07:41

Hi Boffis! Thanks, this is definitely a tricky one.

With regard to using HCL: I've thought about this and also had one unsuccessful attempt at obtaining free pyruvic acid using aqueous HCL. In this attempt, I tried to replicate an earlier method for freeing propionic acid from its calcium salt where formed CaCl2 drove phase separation of the organic and aqueous layers. Unfortunately, the suspended calcium pyruvate eventually turned into a light yellow rock-hard mass which smelled between brown sugar and honey. I recall it being somewhat different in smell than when using sulfuric acid. Some of this mass eventually dissolved, but there was a fair bit which was quite stubborn and remained. In the end, there was no phase separation like the propionic acid.

Seems that, while both propionic acid and pyruvic acid are miscible with water, pyruvic acid cannot be salted out like propionic acid.

Additionally, I recall reading that CaCl2 can form complexes with ketones and alcohols. At least with acetone and ethanol. So, perhaps there was some reaction here..

As for relying on CaCl2 to drive esterification forward, it seems that this salt not only reacts with ethanol, but is soluble in it. And, even if it precipitated out, it would still be in the reaction mixture. I recall reading that MgSO4 does not help to drive forward esterification when it is within the reaction mixture. The hydrated salt needs to be removed.

Seems that using HCL might not be viable due to the reactivity of CaCl2 and that water, even if drawn towards the salt, is not removed from the reaction mixture?


I'm a little skeptical of my results at the moment. I took the distillate, washed it with a saturated sodium sodium carbonate solution, and attempted to form the bisulfite adduct using a saturated sodium metabisulfite solution. Nothing has formed after stirring overnight..

clearly_not_atara - 20-4-2023 at 19:29

Excessively strong acid may cause pyruvic acid to undergo self-aldol condensation.

Oxalic acid may be the ideal choice for this case. Very insoluble calcium salt, so you can use a dilute solution. Could still be tricky to dry the pyruvate.

EDIT: on further consideration, it makes sense to use a slight molar deficit of oxalic acid, like 0.9 moles

[Edited on 21-4-2023 by clearly_not_atara]

tyro - 21-4-2023 at 15:11

I've seen a few threads here which discuss/demonstrate producing ethyl formate without the aid of strong acid; formic acid being strong enough to catalyze esterification itself. This is advantageous as formic acid can be decomposed into CO with strong acid.

It looks like pyruvic acid is stronger relative to formic acid, based on pKa:

Formic acid: 3.75
Pyruvic acid: 2.50

Would I be correct to assume that a strong acid might not be needed for ethyl pyruvate, thus mitigating condensation/decomposition concerns?

AvBaeyer - 21-4-2023 at 18:40

Have you considered buying ethyl lactate (Amazon, Ebay) and oxidizing to ethyl pyruvate (an Org Syn procedure)? That sure has got to be better than fighting the battle you have on your hands now.


tyro - 21-4-2023 at 19:44

Hi AvBaeyer!

I certainly have considered this, and it very well may come to that. But, I have the reagents on hand, and can't help but enjoy the challenge. I'm not strictly on a mission for the end product by the easiest methods available.

If I do end up going the ethyl lactate route, I'll be sure to post that (I'm not sure I've seen it here, but I'll double check).

clearly_not_atara - 21-4-2023 at 20:09

Would I be correct to assume that a strong acid might not be needed for ethyl pyruvate, thus mitigating condensation/decomposition concerns?

I don't think it's that simple — last I checked pyruvate esters are prone to hydrolysis, while formate esters are stable. Another unique property of formates is that they have lower boiling points than the parent alcohols.