Skip to main content
Loading Explore guides…
Synthesizing Petrichor: Geosmin Fragrance · Acedra
Home
Explore
Chemistry
Synthesizing Petrichor: Geosmin Fragrance
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
67 segments
Copy
Hide video
0:00
You know, that strong, earthy smell outside after it has rained, especially on a warm day. Can I replicate that in the lab and make some kind of rain smelling fragrance? That was the question I asked myself many years ago, and now it is finally time to do it. So I went online, like any professional chemists would, and asked Google what causes the smell of rain. So the smell is called Petrichor and it seems like the molecule Geosmin is the main contributor to this smell.
0:33
Okay, so Geosmin, how do I make it?
0:46
It seems like this paper does an enantioselective synthesis of deuterated geosmin but I can just ignore the deuterated part and it seems pretty achievable. So this is the order of the reactions. The only thing is, they start with this, now how do I get that?
1:03
After searching around, I found a 30 year old paper that makes it seem pretty achievable. So, that adds some extra steps. But it starts simply from 2-methylcyclohexanone. Now, there's also one reagent that I tried to make before on my channel, which is ethyl vinyl ketone, since I am mad that it didn't work out. I simply refuse to buy it and I still want to make it myself. So after gaining about a year of experience, I found another method for which I have all the chemicals on hand
1:29
and it seems to be pretty straightforward. So why not? So the first thing I will do is prepare ethyl vinyl ketone. So I set up a heating mantle and add a large flask and a stir bar. I add 65 ml of hexanes as a solvent and then 66 ml of diethyl ketone as the first reagent, and then at 40 ml of ethylene glycol as the second reagent and 40 mg of p-Toluenesulfonic acid as the catalyst. Now that all of the reagents have been added, I put a Dean-Stark apparatus on top
1:59
and fill it completely with hexanes until it overflows into the flask. Then on top of that, I add a condenser.
2:07
Now I start heating the mixture so that it begins to boil and leave it running until it's finished. The boiling hexanes will carry over any water that is formed during the reaction. They will condense together in the condenser and then fall back down into the reservoir. Since water does not mix with hexanes and has a higher density, it will sink to the bottom of the reservoir and collect there. So with the Dean-Stark apparatus, we can continuously remove water
2:31
from the mixture, while also tracking how much water has formed. Since we know how much water should be formed theoretically, we can use this to judge whether a reaction is complete. In the reaction the acid catalyzes the reaction between diethyl ketone and ethylene glycol to form the ethylene glycol acetal of diethyl ketone, and water. Since the product can undergo the reverse reaction in the presence of water, we have to continuously remove it by using the Dean-Stark apparatus.
2:57
Otherwise, the reaction will find its equilibrium and stay there and we would not get much of the product. I leave it running for about 8 hours and when I come back we can see a lot of water has collected in the apparatus. With the amount of reagent that I added, the maximum amount of water that could have formed is 11.25 ml. We can see that the water level is sitting a bit above ten ml and there are also some water droplets sticking in other places. This
3:22
pretty much confirms that the reaction was successful and that it is finished. I now add 310 ml of ethylene glycol to dilute the mixture and take it slightly off heat to allow it to cool down to about 35 to 40 C
3:36
Now I can immediately start the second reaction and there is no need to purify the product. So I attach a dropping funnel to the flask and add in 100 grams of bromine that I quickly made myself. I then opened the dropping funnel and dropwise add it to the mixture while keeping the temperature around 40 C. I continuously check the temperature with an IR-thermometer and added a cold water bath to keep it in check. This should prevent any side reactions from happening.
4:02
In the reaction, one bromine of the dibromine takes up the slightly acidic alpha hydrogen to form hydrobromic acid, while the other bromine attaches itself to the carbocation that has formed to then form the mono brominated product. When the addition was complete, the mixture was still orange, which means that unreacted bromine is present. So I set the flask in a heating mantle heat it back to 40 C to force the reaction to completion.
4:26
After a while, the mixture turns clear again, which means that all the bromine has reacted and the reaction is finished. So I remove the dropping funnel and stop stirring. We then see the mixture separate into two layers with the hexanes and the product on top, and ethylene glycol on the bottom. I move it all to a separatory funnel and add more hexanes to take up the products. I then extract the ethylene glycol layer three times more with hexanes. When the extraction was finished,
4:54
I combined all of the extracts in a beaker.
4:57
I add a stir bar and then 12.5 grams of sodium carbonate. This will destroy any remaining hydrobromic acid and maybe stick to any ethylene glycol that might be present. I leave the mixture to stir for a bit, and then filter it all through some cotton, directly into a flask, to remove all of the sodium carbonate. When it is finished, I remove the funnel and attach a short path distillation apparatus. I now start heating and distilling over all of the hexanes under a mild vacuum.
5:25
After a while, no more hexanes was coming over and a clear liquid was left behind, which should be mostly the mono-brominated product. I then swapped the receiving flask and increasd heating. I then pull a strong vacuum to distill over the product. The product starts distilling over at 41 C at a pressure below one millibars. Since the vacuum gauge doesn't read between 0 and 1 mbar. I can't know the exact pressure. I assist the distillation with some light
5:50
heating from a heat gun, to make it go faster. And after a while, nothing more comes over and only a tiny bit of cloudy liquid is left in the flask. In the end, I collected 108.94 grams of the product as a clear liquid, which is a yield of 83%. This is very similar to the literature where they achieved a yield of 90%. The liquid feels quite hefty and it probably has a decently thicc density. The smell is surprisingly fruity when dilute.
6:16
It smells similar to pineapples, but smelling it directly makes it smell more like glue.
6:22
Anyhow the yield is quite good and with that I can move onto the next step. So I set up a flask with a stir bar in a heating mantle and add 100 ml of methanol, I then add 58 grams of potassium hydroxide and reflux it to make a solution containing potassium methoxide. This reaction will be in equilibrium, mainly influenced by the amount of water that is present. I then add 280 ml of DMSO as a dilutant and solvent, which causes the potassium hydroxide and methoxide to precipitate out.
6:52
I heat the mixture to 80 C and the solid dissolves again. I then remove the condenser and replace it with a dropping funnel. Now into the dropping funnel, I add in all of the mono-brominated product. I dropwise add it to the flask while keeping the temperature around 80 C. In the reaction, the methoxide ion is a very strong base. It will take up a proton (H+) from the carbon adjacent to the one the bromine is attached to and form methanol. This will cause the formation of a double bond
7:18
which leads to bromine getting kicked off (bromine is a good leaving group). But the remaining potassium ion is there to save the bromine from being alone and together they form potassium bromide, which precipitates out of solution. I continue the reaction and I make sure that the temperature is correct by checking with a thermometer. After a while, all of the mono brominated product has been added and we are left with a dark liquid that has a bunch of suspended potassium bromide.
7:42
I then set the mixture up for short path vacuum distillation to remove all of the methanol. I then swapped the receiving flask and increased the heat to start distilling over a mixture of the vinyl product and DMSO. After a while it looks like no more product is coming over and it seems only pure DMSO is distilling over. So I stopped the distillation. Now I have a mixture of DMSO and the vinyl product, so to remove the DMSO, I can simply wash it out.
8:09
Since it is water soluble but the product is not very water soluble, so I add a 5% sodium chloride solution to the mixture and then move it all to a separatorory funnel. The sodium chloride into solution helps to increase the polarity of the water and so it will dissolve less of the vinyl product. The mixture now has two layers and on top is the vinyl product. I washed it a few more times with a 5% sodium chloride solution and afterward I was left with 50 grams of the vinyl product
8:36
as a clear liquid, which is a yield of 75%. This is also a pretty decent yield. Since the literature had 86%, this product, unlike ethyl final ketone, is quite stable. So it is good to store this product and only start making ethyl vinyl ketone from it when it is needed. The smell of this product is strong, like (mineral) turpentine or some kind of glue. It is not very pleasant and kind of gives me a headache. Anyhow, since I will be needing the ethyl vinyl ketone now, I can make it right away.
9:04
So I set up a flask with a stir bar and add in 55 ml of water.
9:10
I then add 0.85 grams of p-Toluenesulfonic acid and 25 ml of the final product.
9:20
I stopper the flask and then leave it to stir for 3 hours at room temperature. In the reaction, we are converting the acetal group back into a ketone by hydrolysis with the help of the acid catalyst. This will produce the ethyl vinyl ketone and also return the ethylene glycol that was used to make the acetal. After 3 hours I put the flask in an ice bath and then add 0.3 g of sodium bicarbonate to destroy the acid. And then 11.3 grams of sodium chloride.
9:49
This will increase the polarity and saturation of the water and kick out most of the ethyl vinyl ketone that has dissolved into it. I then move the mixture to a separatory funnel and extract it three times with o-xylene. In the literature they used p-cymenel to do the extraction. But the closest thing I had was o-xylene. I am not sure if they specifically selected p-Cymene to improve the distillation or if they are just ancient. For old papers, it can be hard to judge if they do something
10:16
because it is necessary, or simply just because it was normal or available at the time. Anyhow, I combined all of the o-xylene extracts in a flask and set it in a heating mantle. I add 80 mg of hydroquinone to the solution as a stabilizer which will scavenge radicals and help prevent ethyl vinyl ketone from getting destroyed. I then attach a fractionating column and a short path distillation apparatus. I pull a strong vacuum and heat the mixture lightly,
10:42
and with the help of a heat gun, I start distilling out the ethyl vinyl ketone. Since it's slightly volatile, I set an ice bath in the receiving flask. It did come over, though it still took quite some time and effort to get it to come out. Despite it being relatively volatile, in the end, I collected enough ethyl vinyl ketone to start the geosmin synthesis. It seems also some residual water had come through and was sitting on the bottom,
11:06
but I will simply avoid taking this up when I'm going to be using it. The smell of ethyl vinyl ketone is like a spicy, sharp vegetable and it also stings the nose a little. Anyhow, now that I have ethyl vinyl ketone, I can move on to the next step. So I have some 2-methylcyclohexane on that I made in a previous video, but that was like a year ago and in the meantime it has turned slightly yellow. I will not use it like this, so I'm going to redistill it.
11:30
So I add all of the dirty 2-methylcyclohexanone to the flask and distill it over through a fractionating column under regular pressure. When that is done, all that is left in the flask is some yellow stuff. I removed the first fraction from the distillate and also the last, and I should now have 9.19 grams of pure 2-methylcyclohexane, which I will use right away. So I set up a heating mantle and a flask with a stir bar.
11:56
So I will also need (S)-(-)-a-methylbenzylamine, which I bought a very long time ago, since I always wanted to do this project and I stored it in the fridge this whole time and I finally get to open it. If anyone is curious, many chemicals that are considered dangerous come in this type of packaging. But before I get to adding it, I first add in 50 ml of Toluene as a solvent. I then add in all of the 2-methylcyclohexanone that I redistilled, and then 10.5 ml of (S)-(-)-a-methylbenzylamine.
12:27
Like earlierm I attach a dean-stark apparatus, but this time I filled it with toluene. I also insulated it with some aluminum foil since it had a hard time coming through. In the reaction, the methylbenzylamine reacts with the 2-methylcyclohexanone to form the corresponding imine product, and water. Since the imine is also water sensitive. We again use the Dean-stark apparatus to remove it. I leave it to react for a day and when I come back, we see some water has collected in the apparatus.
12:58
In theory, a maximum of 1.48 ml of water can form during the reaction. There are also some droplets sticking at the top, so it seems that pretty much the same amount of water has formed, so the reaction should be finished. I then set the mixture up for short path vacuum distillation to remove all of the toluene. In the end I am left with.. yellow?!? Why is it always yellow? Why not just green? Now to this product, I directly at 8.1 ml of ethyl vinyl ketone and flush the flask with nitrogen.
13:31
I then leave it to stir for a day at room temperature. This reaction is a type of stork reaction which is a michael addition, where the nucleophile is an enamine. The imine is an equilibrium with its enamine form. The enamine form can react with the ethyl vinyl ketone, which is a great Michael acceptor, to produce the final product. This happens in a stereoselective manner due to the presence of the bulky methylbenzylamine, and it will favor the R configuration. I come back a day later...
14:00
and it looks the same. I then add 62.5 ml of methanol as a solvent and then 37.5 ml of 10% acetic acid. The addition of water and acetic acid will hydrolyze the product, kicking off the methylbenzylamine and returning the ketone to give the diketone product. I leave it to stir at room temperature for a few hours and after that it has become
14:25
more....(yellow) I then move it to a heating mantle and set it up for short path vacuum distillation to remove all of the methanol.
14:33
When that is done, I dilute the mixture with 100 ml of water and move it all to a separatory funnel. I extract it three times with diethyl ether, and then return the combined ether extracts to the separatory funnel.
14:47
I wash it once with 10% hydrochloric acid, which will force the methylbenzylamine into the water phase by protonating it. I then wash it once with water to wash out the acid and then once with brine to get out most of the water. I then take the ether layer and pull out any remaining water by adding magnesium sulfate. I then filtered the mixture through some cotton to remove the magnesium sulfate again, directly into a flask. I then attach a short path distillation apparatus again and distill off
15:14
all of the ether. When all of it is gone, A light yellow liquid is left behind, which should contain the diketone product. I then add 90 ml of methanol as a solvent and a solution of 0.46 grams of potassium hydroxide and methanol, which will contain some potassium methoxide. I attach a condenser and then heat it to a reflux and leave it for 2 hours. In the reaction the base catalyzes the diketone to cyclisize into the first cyclic product, usually the previous addition of the vinyl
15:43
ketone, which is now followed by an aldol condensation into a fused ring system would be referred to as a type of Robinson annulation. When I come back, the mixture has turned even (more yellow). I
15:56
remove the condenser and add 8.5 ml of glacial acetic acid to neutralize it. I attach a short path distillation apparatus again and remove all of the methanol under reduced pressure. When all of the methanol is gone, all that is left is a ******* dirty yellow liquid. I then add 90 ml of water to dissolve impurities and some diethyl ether to take up the product.
16:19
I move it all to a separatory funnel and extract the water phase three times with ether. I combined all of the ether extracts, dried it with magnesium sulfate, and then filtered it again. I again remove all of the ether and then start distilling over the product. At first some lower boiling liquid is coming over, which smells very similar to 2-methylcyclohexanone So it's probably just the residual 2-methylcyclohexanone. After a while, that stops distilling over and I change the short path
16:47
for another smaller one. I then start to distill over to cyclic product, with the help of a heat gun, I squeeze out every last drop that I can. And afterward, I am left with only 1.7 grams of the cyclic product. As a clear oil. I'm not sure which step was so low yielding, but unfortunately the yield turned out to be only 11.6%. Well, at least it's something. So let's just continue and move onto the next paper. I set up a flask with a stir bar and dissolve all of the cyclic product
17:14
into 125 ml of DCM and transfer that to the flask. I then add 1.61 grams of sodium bicarbonate and set the flask in an ice bath. I flush the flask with nitrogen. I then slowly started adding in 2.46 grams of m-chloroperbenzoic acid while keeping it under nitrogen. When everything was added, I stopper the flask and leave it to stir for 2 hours on ice and then for two days at room temperature. This reaction is a typical Prilezhaev epoxidation where an alkene
17:44
double bond reacts with a peroxy acid to form the corresponding epoxide. The sodium bicarbonate was added to counteract the acidity of the perbenzoic acid, which could lead to side products. After a while, a white precipitate formed, which is the insoluble sodium salt of m-chlorobenzoic acid. Now I wash the reaction mixture two times with 100 ml of a 1 to 1 mixture of a saturated sodium thiosulfate solution and a saturated sodium bicarbonate solution.
18:11
I do it inside of the flask and decant the upper water layer. Because the solid soaked up the DCM. It can’t be done inside a separatory funnel. I wash it one time with some brine and then add a bunch of sodium sulfate to hold onto any remaining water. I shake it strongly and then filter it all through some cotton. After 100 years, it all filtered through and I distill off all of the DCM with short path vacuum distillation. When only a bit of the DCM is left, I add 4 g of celite
18:38
which will adsorb the product, and then continue distilling off all of the DCM. When all of the DCM is gone, only the celite that contains product is left. Now to purify the product, I have to do column chromatography. So I set up a column and add in 100 grams of spherical silica gel that I mix with 20% diethyl ether in hexanes. I allow the silica to pack and then add a layer of sand on top. I run the solvent level into the sand and then add all of the celite with the product on top.
19:05
I add some sand to the flask to scrub lose any remaining celite and add that on top as well. Since there were some air bubbles, I try to loosen them by tapping the column with a cork ring, but I went a little too hard and the top layer became uneven, though for this operation it doesn't matter too much since I'm only trying to remove the bottom half of the spots on the TLC. I run the column by applying pressure with a pump and after that I collected fractions of varying sizes.
19:30
On the TLC we see fractions 3 to 8 containing the upper half of the components and the other fractions only the lower half. So I take fractions 3 to 8 and combine them in a flask.
19:42
I then distill off all of the solvent and afterward I am left with a little bit of product that has a very sweet and floral smell. Now for the next reaction I dissolve the product in 20 ml of dry THF and then set it aside for a second. I set up another flask in a heating mantle with a nitrogen line. I then connect a condenser and add in 60 ml of dry THF and add in 1gram of lithium aluminum hydride. I heat the mixture to a reflux and attach a dropping funnel. Then to the dropping funnel,
20:09
I add the product-THF mixture and dropwise add it to the flask. I then leave it for 2 hours. In the reaction, lithium aluminum hydride will, transfer a hydride to the ketone and epoxide ,causing both to be reduced to an alcohol and give the corresponding diol product. When that is finished, I add 30 ml of a saturated sodium sulfate solution to the now empty dropping funnel and gradually add that to the flask. This will destroy any remaining lithium aluminum hydride as well as protonate
20:37
the product, the sodium sulfate also helps THF separate out of the water by salting out. I then add 5 ml of 1% hydrochloric acid, which will make sure that all the product has protonated and also neutralize the solution by reacting with formed lithium hydroxide. I then move it all to a separatory funnel and extract it with ether and wash it with brine. I take the combined ether extracts and dry it with sodium sulfate. I then filtered it all through some cotton
21:04
and do the exact same as in the previous step. And distill over part of the ether, and then add in three grams of celite. I run the column again in the same manner, but this time the separation is a lot easier. The product is the lower spot on the TLC. It looks suspiciously long for it all to elute off the column, which is a thing when solubility of the product is low in the eluent. But it wasn't a problem, because the product
21:25
is the last spot on the TLC so nothing else could come through anyway. I combined all the fractions containing the product and i distill off all of the solvent again. When only very little is left a white solid already precipitates out of solution. After a while, it is completely dry and only the product is left as a white solid. Now to start the last step, I dissolve it in 160 ml of chloroform and add 3.2 ml of Pyridine.
21:49
I then add 8 g of p-toluenesulfonyl chloride and flush the flask with nitrogen.
21:56
I stopper the flask and then leave it to react for four days at room temperature. In the reaction the p-toluenesulfonylchloride reacts with one alcohol group to form the corresponding tosylate and hydrochloric acid. The tosylate is a good leaving group and can therefore be used to remove the alcohol. The pyridine in the reaction mixture will take of the hydrochloric acid to form pyridine hydrochloride. When I come back it is still a clear liquid and I replace the stopper with a dropping funnel.
22:22
I then add in a suspension of 0.8 grams of lithium aluminum hydride in THF. I add it slowly to the flask and then replace the dropping with a condenser. I reflux the mixture and then allow it to react for 5 hours. In this reduction reaction, the lithium aluminum hydride will remove the tosylate group and replace it with a hydrogen. The remaining pyridine in the mixture is able to neutralize the p-toluenesulfonic acid that is formed. When that is done, I take it off heat and allow it to cool
22:50
down to room temperature. We are now left with a disgusting yellow solution with some solid bits. I add 85 ml of diethyl ether to take up some of the product and then add water to destroy remaining lithium aluminum hydride. I then add 85 ml of 1% hydrochloric acid which will protonate the product and react with any remaining pyridine. I then move it all to a separatory funnel and separate the layers. I return the organic layer and wash it with a 5% sodium bicarbonate
23:17
solution to destroy remaining acid and then once with some brine, to take out the majority of the water, the organic layer becomes clear and I separate the layers. I take the organic layer and add some sodium sulfate to dry it. I filter it through some cotton to remove the sodium sulfate and then set the filterate up for short path vacuum distillation. I removed all of the solvent and a yellow liquid with some solid bits is left behind.
23:45
I add some ether to redissolve it, but it isn't very soluble, so I add some DCM to get everything to dissolve. I then add 3 grams of celite and distill off all of the solvent again. I then set up a column again, but this time I'm using diol functionalized silica gel with a small particle size. This will improve the separation of the compounds and allow me to get pure geosmine. I also use 60% diethyl ether and hexanes as the eluent run the column the exact same as before.
24:14
When the column is finished, I combine all the fractions that should contain geosmine and set it up for short path vacuum distillation to remove all of the hexanes and ether. When that is done, I am left with a fluffy white solid, which should be geosmine. I put it all into a vial, but the weight was higher than the maximum possible yield, which means there is still some solvent in there. I tried to dry it a little, but it was very slow and the geosmin starts to sublime.
24:38
It's likely that the solvent is trapped inside of the solid. But it doesn't really matter too much because smelling this compound pure is like putting wet soil inside of your nose, regardless of any bit of solvent. The smell is extremely earthy and it really depends on the temperature and concentration. What exactly it smells like. It definitely has some similarities to the smell of rain, but smelling it pure like this is not really the same experience.
25:01
It just smells wet and damp and a little musky, so they give it a more of a chance. I will have to dilute it. So I added about 50 miles of dipropylene glycol to the residue in the flask. It dissolved really slowly. So I left it to stir for a day and after it is all dissolved, it is time to smell it again. So this time the smell is a lot more pronounced. It is very, very earthy and it smells more like rain. It probably had some time to build up in the flask
25:26
and the concentration is way beyond what you would ever smell outside. It probably needs to be even more dilute, but that way I could make endless dilutions. So maybe it would be better for you to experience it yourself. So if anyone wants to get a sample to smell it for themselves, I will send you some. The only thing you have to do is support me in the 25 bucks tier on patron or via PayPal and I will then send you a decent sized sample to smell for yourself, pretty much anywhere in the world.
25:50
And then you can tell me what you think it smells like. Anyhow, I'm very happy that it worked out and I finally got to do this. Thanks for watching. And as always, a special thanks to all my patrons. Cya.
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat